Venom Yield, Regeneration, and Composition in the Centipede

Transcription

Loma Linda University
TheScholarsRepository@LLU: Digital Archive of Research,
Scholarship & Creative Works
Loma Linda University Electronic Theses & Dissertations
6-2014
Venom Yield, Regeneration, and Composition in
the Centipede Scolopendra Polymorpha
Allen Michael Cooper
Follow this and additional works at: http://scholarsrepository.llu.edu/etd
Part of the Biology Commons
Recommended Citation
Cooper, Allen Michael, "Venom Yield, Regeneration, and Composition in the Centipede Scolopendra Polymorpha" (2014). Loma
Linda University Electronic Theses & Dissertations. Paper 192.
This Dissertation is brought to you for free and open access by TheScholarsRepository@LLU: Digital Archive of Research, Scholarship & Creative
Works. It has been accepted for inclusion in Loma Linda University Electronic Theses & Dissertations by an authorized administrator of
TheScholarsRepository@LLU: Digital Archive of Research, Scholarship & Creative Works. For more information, please contact
[email protected].
LOMA LINDA UNIVERSITY
School of Medicine
in conjunction with the
Faculty of Graduate Studies
____________________
Scolopendra Polymorpha
by
____________________
A Dissertation submitted in partial satisfaction of
the requirements for the degree
Doctor of Philosophy in Biology
____________________
June 2014
© 2014
All Rights Reserved
Each person whose signature appears below certifies that this dissertation in his/her
opinion is adequate, in scope and quality, as a dissertation for the degree Doctor of
Philosophy.
, Chairperson
William K. Hayes, Professor of Biology
Leonard R. Brand, Professor of Biology and Paleontology
Stephen G. Dunbar, Associate Professor of Biology
Kevin E. Nick, Associate Professor of Geology
Zia Nisani, Professor of Biology, Antelope Valley College
iii
ACKNOWLEDGEMENTS
Earning this degree has been a journey that would not have been possible without
the assistance and mentoring of many people. For this reason, I would like to express my
deepest appreciation to all of the individuals who helped me complete this journey. I am
especially grateful to all the members of my dissertation committee who invested so
much energy to help me succeed. I would like to thank Dr. William Hayes for his kind
mentoring, warm friendship, and patient encouragement. If he had not shared his many
skills in research design, statistics, and scientific writing, this dissertation would not have
come to fruition. I am also grateful to Drs. Leonard Brand, Stephen Dunbar, Kevin Nick,
and Zia Nisani for their advice and assistance. I would like to thank Wayne Kelln for
sharing his vast technical expertise in proteomics on a near-daily basis in the lab.
I would like to thank my wife, Jamey Cooper, for her active support, be it chasing
and lassoing squirming centipedes, feeding centipedes, or assisting with venom
extractions. Without her help and commiseration as a fellow grad student I would not
have survived. I am grateful to my parents, Clyde and Sandra Cooper, for all their
encouragement and many forms of support throughout my life that have made higher
education attainable. My father-in-law and mother-in-law, Larry and Cindy Hiday, also
deserve recognition for their enthusiasm for the research and for helping to keep two grad
students fed and clothed.
In addition to advisors and family, I would like to express my appreciation to the
many students whom I have had the privilege of working alongside. Thanks to Gerad Fox
for googling answers to all my questions, finding obscure articles written in strange
languages in forgotten reaches of the internet, helping to edit manuscripts, and swatting
iv
fruit flies. Thanks to David Nelsen for help with research design and proteomic methods,
as well as for YouTube study breaks (ski ballet anyone?). Thanks to Eric Gren for many
valuable proteomics discussions and for help in bringing long hair back into vogue.
Thanks also to Chip Cochran, without whom my knowledge of the snake venom
literature would be seriously impoverished.
v
DEDICATIONS
I would like to dedicate my dissertation to my parents
who have given me so many opportunities,
&
to my wife, Jamey, whose love and support
never cease to amaze me.
I love you all!
vi
CONTENTS
Approval page.................................................................................................................... iii
Acknowledgements............................................................................................................ iv
Dedications ........................................................................................................................ vi
Table of contents............................................................................................................... vii
List of tables...................................................................................................................... xii
List of figures................................................................................................................... xiv
List of abbreviations ....................................................................................................... xvii
Abstract ..............................................................................................................................xx
Chapter
1. Introduction..............................................................................................................1
Specific objectives .............................................................................................8
References........................................................................................................10
2. Variation in venom yield and protein concentration of the centipede
Scolopendra polymorpha and Scolopendra subspinipes .......................................18
Abstract ............................................................................................................19
Introduction......................................................................................................20
Materials and methods .....................................................................................25
Centipedes..................................................................................................25
Venom extraction and venom volume determination................................32
Protein quantification.................................................................................35
Statistical analyses .....................................................................................35
Results..............................................................................................................37
Total venom volume yield .........................................................................37
Body length and volume yield .............................................................37
Species and volume yield.....................................................................41
Additional variables influencing volume yield in S.
polymorpha ..........................................................................................41
vii
Body length and number of productive shocks....................................46
Pulse-related venom volume yield.............................................................46
Body length and number of measurable venom pulses .......................47
Body length and venom volume per pulse...........................................50
Body length and percent of total volume yielded in first pulse ...........50
Pulse number and percent of total volume of venom extracted...........51
Venom protein concentration.....................................................................55
Body length and venom protein concentration ....................................55
Species and venom protein concentration............................................55
Additional variables influencing venom protein concentration
in S. polymorpha ..................................................................................55
Pulse-related venom protein concentration................................................59
Comparison of variation ............................................................................62
Discussion ........................................................................................................62
Venom collection, venom attributes, and venom supply...........................62
Factors influencing venom yield................................................................66
Body length..........................................................................................66
Species .................................................................................................71
Geographic origin ................................................................................71
Relative body mass ..............................................................................72
Milking history ....................................................................................72
Sex........................................................................................................73
Factors associated with venom depletion (successive shocks) ..................75
Factors influencing protein concentration .................................................76
Geographic origin ................................................................................77
Milking history ....................................................................................78
Body length..........................................................................................78
Functional aspects of venom supply ..........................................................80
Conclusions................................................................................................85
Acknowledgments............................................................................................86
References........................................................................................................87
3. Venom regeneration in the centipede Scolopendra polymorpha: evidence
for asynchronous venom component synthesis....................................................104
viii
Abstract ..........................................................................................................105
Introduction....................................................................................................106
Materials and methods ...................................................................................112
Centipedes................................................................................................112
Experimental design.................................................................................113
14-day venom regeneration trial ........................................................113
48-hour venom regeneration trial.......................................................113
7-month follow-up to 14-day regeneration trial.................................114
Venom extraction and venom volume determination..............................114
Protein quantification...............................................................................115
RP-FPLC analysis....................................................................................115
Pre-treatment and analysis of RP-FPLC chromatograms ........................116
Statistical analyses ...................................................................................117
Results............................................................................................................119
Initial milking and general trends in venom regeneration .......................119
48-hr trial .................................................................................................123
14-day trial ...............................................................................................134
7-month follow-up ...................................................................................140
Analysis of chromatograms from 48-hr trial ...........................................144
General analysis of chromatograms...................................................144
Analysis of chromatograms by region ...............................................147
Analysis of chromatograms by peak..................................................152
Discussion ......................................................................................................157
Electrical milking.....................................................................................157
Venom regeneration during the 48-hr and 14-day trials..........................157
7-month follow-up study..........................................................................159
Factors influencing venom regeneration..................................................161
Venom regeneration: comparison with other venomous animals............162
Asynchronous regeneration of venom components in 48-hr trial............164
Implications of venom regeneration for venom extraction protocols......166
Functional aspects of venom regeneration...............................................166
Conclusions..............................................................................................169
Acknowledgments..........................................................................................170
References......................................................................................................171
ix
4. Partial characterization of the venom proteome of the centipede
Scolopendra polymorpha .....................................................................................182
Abstract ..........................................................................................................183
Introduction....................................................................................................185
Material and methods.....................................................................................191
Centipedes................................................................................................191
Venom extraction and venom pools ........................................................191
Protein quantification...............................................................................193
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE)......................................................................................................194
RP-FPLC..................................................................................................194
Pre-treatment and analysis of individual venom sample RPFPLC chromatograms ........................................................................195
Mixed-location venom pool protein characterization by MALDI
TOF MS, LC-MS/MS, MALDI TOF/TOF MS/MS, and BLASTp
sequence similarity searching ..................................................................196
MALDI TOF MS analysis .................................................................196
LC-MS/MS analysis ..........................................................................198
MALDI TOF/TOF MS/MS analysis..................................................199
Sequence similarity searching............................................................200
Statistical analyses ...................................................................................200
Results............................................................................................................200
Venom yields ...........................................................................................200
SDS-PAGE ..............................................................................................201
MALDI TOF MS analysis of unfractionated venom...............................205
RP-FPLC of mixed-location venom pool ................................................208
MALDI TOF MS analysis of RP-FPLC fractionated venom from
mixed-location venom pool .....................................................................211
MS/MS analyses for protein identification and sequence similarity
searching ..................................................................................................226
Comparison of venom from S. polymorpha from Arizona and
California by RP-FPLC of individual samples ........................................231
Discussion ......................................................................................................234
SDS-PAGE ..............................................................................................234
MALDI TOF MS analysis of unfractionated venom...............................234
RP-FPLC of mixed-location venom pool ................................................235
x
MALDI TOF MS analysis of RP-FPLC fractionated venom from
mixed-location venom pool .....................................................................236
MS/MS analyses for protein identification and sequence similarity
searching ..................................................................................................242
Comparison of venom from S. polymorpha from Arizona and
California .................................................................................................253
Conclusions..............................................................................................254
Acknowledgements........................................................................................255
References......................................................................................................257
5. Conclusions..........................................................................................................276
Future directions ............................................................................................281
References......................................................................................................285
Appendices
A. Strategic deployment of predatory and defensive emissions: a review of
the literature .........................................................................................................286
xi
TABLES
Tables
Page
Chapter Two
1. Correlations of variables in standard multiple regression model for venom
volume yield from Scolopendra polymorpha (n = 177) ........................................44
2. Standard multiple regression results for venom volume yield from
Scolopendra polymorpha (n = 177) .......................................................................45
3. Correlations of variables in standard multiple regression model for venom
protein concentration from Scolopendra polymorpha (n = 168) ...........................57
4. Standard multiple regression results for venom protein concentration from
Scolopendra polymorpha (n = 168) .......................................................................58
Chapter Three
1. Correlations of variables in standard multiple regression model for percent
of initial milk venom volume regenerated from Scolopendra polymorpha
(n = 71) in 48-hr trial ...........................................................................................125
2. Standard multiple regression results for percent of initial milk venom
volume regenerated from Scolopendra polymorpha (n = 71) in 48-hr trial ........126
of initial milk venom protein mass regenerated from Scolopendra
polymorpha (n = 71) in 48-hr trial .......................................................................128
protein mass regenerated from Scolopendra polymorpha (n = 71) in 48-hr
trial .......................................................................................................................129
of initial milk venom volume regenerated from Scolopendra polymorpha
(n = 79) in 14-day trial .........................................................................................135
6. Standard multiple regression results for percent of initial venom volume
regenerated from Scolopendra polymorpha (n = 79) in 14-day trial ...................136
xii
of initial milk venom protein mass regenerated from Scolopendra
polymorpha (n = 79) in 14-day trial.....................................................................138
protein mass regenerated from Scolopendra polymorpha (n = 79) in 14day trial ................................................................................................................139
9. Results of two-way ANOVA of clr-transformed relative peak area of
chromatographic region by milking interval and geographic origin for 10
chromatographic regions (Fig. 5) from RP-FPLC of electrically extracted
Scolopendra polymorpha venom .........................................................................151
10. Results of two-way ANOVA of clr-transformed relative peak area by
milking interval and geographic origin for 28 peaks (Fig. 5) from RPFPLC of electrically extracted Scolopendra polymorpha venom........................156
Chapter Four
1. Molecular masses of RP-FPLC fractionated peptides and proteins from
electrically extracted Scolopendra polymorpha venom pool as determined
by MALDI TOF MS. Masses with relative intensity ≥ 75% of base peak in
a given fraction are given, with the predominant species indicated in bold ........212
2. Potential correspondence between Scolopendra polymorpha venom
protein masses from MALDI TOF MS with published Scolopendra venom
protein masses......................................................................................................224
3. Characterization of components from the RP-FPLC fractionated venom
pool of Scolopendra polymorpha (Fig. 3) by LC-MS/MS and MALDI
TOF/TOF MS/MS of trypsin-digested proteins followed by BLASTp
sequence similarity searching ..............................................................................227
xiii
FIGURES
Figures
Page
Chapter Two
1. Ventral and dorsal views of (A) Scolopendra polymorpha (female,
Cochise Co., Arizona), and (B) Scolopendra subspinipes (male,
purportedly from Vietnam) ....................................................................................27
2. (A) Ventral view of the head of Scolopendra polymorpha showing
measurements of forcipule width (a) and length (b). (B) Venom extraction
from S. polymorpha using an electrified scoopula positioned between the
centipede’s forcipules ............................................................................................30
3. Centipede body length-specific venom yields in terms of total volume (A)
and protein concentration (B) from Scolopendra polymorpha and S.
subspinipes by electrical stimulation. ....................................................................39
4. Relationships between centipede body length and (A) number of
measurable (i.e., >0.05 µL) venom pulses, (B) mean volume of venom per
measurable pulse, and (C) first pulse volume as a percent of total volume
yield from Scolopendra polymorpha (n = 16) and S. subspinipes (n = 6) by
electrical stimulation. ............................................................................................48
5. Percent of total venom volume (mean ± 1 S.E.) associated with individual
venom pulses resulting from consecutive electric shocks of Scolopendra
centipedes. (A) and (B) portray data from within-subjects analyses of
variance (ANOVAs) for S. polymorpha (n = 14, including previously
milked animals; p = 0.011) and S. subspinipes (n = 6; p = 0.037),
respectively. (C) and (D) portray a between-subjects comparison of data
(i.e., assuming independence of pulses) from S. polymorpha (sample size
varied with pulse, excluding previously milked animals) and S.
subspinipes (sample size varied with pulse), respectively.....................................53
6. Protein concentration (mean ± 1 S.E.) associated with individual venom
pulses resulting from consecutive electric shocks of Scolopendra
centipedes. (A) and (B) portray data from within-subjects analyses of
variance (ANOVAs) for S. polymorpha (n = 10, including previously
milked animals; p = 0.63) and S. subspinipes (n = 6; p = 0.55),
respectively. (C) and (D) portray a between-subjects comparison of data
(i.e., assuming independence of pulses) from S. polymorpha (sample size
varies with pulse, excluding previously milked animals) and S.
subspinipes (sample size varies with pulse), respectively .....................................60
xiv
Chapter Three
1. Ventral (A) and dorsal (B) views of Scolopendra polymorpha (female,
Cochise Co., Arizona)..........................................................................................110
2. Percent of initial milk venom protein mass and volume extracted on
second milk after specified time interval from Scolopendra polymorpha
centipedes.............................................................................................................121
3. Protein concentration of venom electrically extracted after specified time
interval from initial milk from Scolopendra polymorpha centipedes..................132
4. Comparison of initial milk (1st milk) venom with venom extracted seven
months later (3rd milk) from Scolopendra polymorpha centipedes. (A)
Volume yield. (B) Protein mass yield. (C) Protein concentration .......................142
5. RP-FPLC of electrically extracted venom from Scolopendra polymorpha.
Separation of the initial (0 hr) venom (40 µg protein) from a single animal
(Arizona specimen #46). Numbers indicate the locations of the 28 peaks
analyzed for changes in relative area among milking intervals. The 10
chromatographic regions analyzed for changes in relative peak area among
milking intervals are indicated below the trace ...................................................145
6. Variation in centered logratio-transformed (clr) (column A) and
untransformed (column B) relative peak area of chromatographic region
(region numbers illustrated in Fig. 5) among milking intervals and
between geographic origins from RP-FPLC of electrically extracted
7. Variation in centered logratio-transformed (clr) relative peak area by peak
(peak identification numbers are from Fig. 5) among milking intervals and
between geographic origins from RP-FPLC of electrically extracted
Chapter Four
1. SDS-PAGE analysis of Scolopendra polymorpha venom (50 µg/lane) in
reduced (lanes 1–3) and non-reduced (lanes 5–7) forms .....................................203
2. MALDI TOF mass spectra of electrically extracted, unfractionated venom
from Scolopendra polymorpha. Spectrum based on pooled venom from S.
polymorpha from (A) combined locations (n = 18 AZ, n = 15 CA), (B)
Arizona (n = 26), (C) California (n = 36) ............................................................206
xv
3. RP-FPLC separation of a pool of electrically extracted venom
(approximately 596 µg protein) from Scolopendra polymorpha (n = 18
AZ, n = 15 CA) ....................................................................................................209
4. Distribution of venom component molecular masses determined from
MALDI TOF MS as a function of RP-FPLC hydrophobicity from an
electrically extracted pool of venom from Scolopendra polymorpha
centipedes (n = 18 from Arizona, n = 15 from California). (A) includes all
935 molecular species detected, while (B) includes only the 76 most
abundantly formed ions (intensity ≥ 20,000 counts) ...........................................215
5. 3D plots of Scolopendra polymorpha venom components constructed by
plotting MALDI TOF MS peak intensities (y-axis, counts) as a function of
mass (z-axis, m/z) and hydrophobicity (x-axis, RP-FPLC fraction
number). (A) 3D plot based on MALDI TOF MS conducted in the
window m/z 1,000–20,000, and (B) in the window m/z 20,000–100,000...........218
6. (A) Frequency of Scolopendra polymorpha venom components detected
by MALDI TOF MS of RP-FPLC fractions according to hydrophobicity
(% buffer B). (B) Histogram of S. polymorpha venom components sorted
into 2,000 Da mass classes. Overlaid curves show the cumulative total
number of protein masses (935) identified from MALDI TOF analyses of
RP-FPLC fractions...............................................................................................221
7. Comparison of relative peak area of 46 individual peaks from RP-FPLC of
electrically extracted venom between Scolopendra polymorpha from
Arizona (n = 5; Cochise County) and southern California (n = 4; Riverside
and San Diego counties). .....................................................................................232
xvi
ABBREVIATIONS
S.E.
Standard error
CO2
Carbon dioxide
cm
Centimeter
mm
Millimeter
V
Volts
mA
Milliamps
AC
Alternating current
Sec
Second(s)
µL
Microliter
PCR
Polymerase chain reaction
BSA
Bovine serum albumin
PBS
Phosphate-buffered saline
CV
Coefficient of variation
CI
Confidence interval
mL
Milliliter
ln
Natural log
ANOVA
Analysis of variance
ANCOVA
Analysis of covariance
MANCOVA
Multivariate analysis of covariance
IV
Independent variable
DV
Dependent variable
LSD
Least significant difference
xvii
ELISA
Enzyme-linked immunosorbent assay
µg
Microgram
mg
Milligram
g
gram
pg
Picogram
LD
Lethal dose
ED
Effective dose
RP-FPLC
Reversed-phase fast protein liquid chromatography
TFA
Trifluoroacetic acid
ACN
Acetonitrile
clr
Centered logratio
LAAO
L-amino acid oxidase
MALDI TOF MS
Matrix assisted laser desorption/ionization time-of-flight
mass spectrometry
LC-MS
Liquid chromatography mass spectrometry
MS/MS
Tandem mass spectrometry
BLASTp
Basic local alignment search tool (protein)
SDS-PAGE
Sodium dodecyl sulfate polyacrylamide gel electrophoresis
Da
Dalton
kDa
Kilodalton
Kv
Voltage-gated potassium channel
PLA2
Phospholipase A2
CRISP
Cysteine-rich secretory protein
xviii
Cav
Voltage-gated calcium channel
Nav
Voltage-gated sodium channel
aa
Amino acid
NCBInr
National Center for Biotechnology Information nonredundant
CHCA
α-cyano-4-hydroxy cinnamic acid
m/z
Mass-to-charge ratio
RP-HPLC
Reversed-phase high-performance liquid chromatography
cDNA
Complementary deoxyribonucleic acid
ESI-Q-TOF MS
Electrospray ionization quadrupole time-of-flight mass
spectrometry
IC50
Half maximal inhibitory concentration
nM
Nanomolar
MMP
Matrix metalloproteinase
ECM
Extracellular matrix
ADAM
A disintegrin and metalloproteinase
SCP
Sperm-coating glycoprotein
Ag5
Antigen 5
DRG
Dorsal root ganglion
CCAP
Crustacean cardioactive peptide
kDa
Kilodalton
xix
ABSTRACT OF THE DISSERTATION
Scolopendra Polymorpha
by
Allen M. Cooper
Doctor of Philosophy, Graduate Program in Biology
Loma Linda University, June 2014
Dr. William K. Hayes, Chairperson
In this dissertation, I investigated yield, regeneration, and composition of
centipede venom. In the first of three empirical studies, I investigated how size
influenced venom volume yield and protein concentration in Scolopendra polymorpha
and S. subspinipes. I also examined additional potential influences on yield in S.
polymorpha, including relative forcipule size, relative mass, geographic origin, sex, time
in captivity, and milking history. Volume yield was positively linearly related to body
length in both species; however, body length and protein concentration were
uncorrelated. In S. polymorpha, yield was most influenced by body length, but was also
positively associated with relative forcipule length and relative body mass. In the second
study, I investigated venom volume and total protein regeneration during the 14-day
period subsequent to venom extraction in S. polymorpha. I further tested the hypothesis
that venom protein components, separated by RP-FPLC, undergo asynchronous
synthesis. During the first 48 hours, volume and protein mass increased linearly.
However, protein regeneration lagged behind volume regeneration, with only 65–86% of
venom volume and 29–47% of protein mass regenerated during the first 2 days. No
significant additional regeneration occurred over the subsequent 12 days. Analysis of
xx
chromatograms of individual venom samples revealed that five of 10 chromatographic
regions and 12 of 28 peaks demonstrated changes in percent of total peak area among
milking intervals, indicating that venom proteins are regenerated asynchronously. In the
third study, I characterized the venom composition of S. polymorpha using proteomic
methods. I demonstrated that the venom of S. polymorpha is complex, generating 23
bands by SDS-PAGE and 56 peaks by RP-FPLC. MALDI TOF MS revealed hundreds of
components with masses ranging from 1014.5 to 82863.9 Da. The distribution of
molecular masses was skewed toward smaller peptides and proteins, with 72% of
components found below 12 kDa. BLASTp sequence similarity searching of MS/MSderived amino acid sequences demonstrated 20 different sequences with similarity to
known venom components, including serine proteases, ion-channel activators/inhibitors,
and neurotoxins. In Appendix A, I reviewed how animals strategically deploy various
emissions, including venom, highlighting how the metabolic and ecological value of
these emissions leads to their judicious use.
xxi
CHAPTER ONE
INTRODUCTION
Centipedes, comprising the class Chilopoda, are terrestrial arthropods from the
subphylum Myriapoda, the taxonomic group that also includes the Diplopoda
(millipedes), Symphyla, and Pauropoda (Edgecombe and Giribet, 2007). These latter
three myriapod classes make up the clade Progoneata, having the genital opening
positioned anteriorly on the trunk, behind the second pair of legs (Edgecombe and
Giribet, 2007). In contrast, Chilopoda are opisthogoneate, with the genital opening
located terminally on the body (Edgecombe and Giribet, 2007). Chilopoda is divided into
five extant orders, Scutigeromorpha, Lithobiomorpha, Craterostigmomorpha,
Geophilomorpha, and Scolopendromorpha, and one extinct order, Devonobiomorpha
(Edgecombe and Giribet, 2007). Centipedes are among the oldest extant terrestrial
arthropods (Murienne et al., 2010; Shear and Edgecombe, 2010), with a fossil record
spanning 420 million years from the late Silurian to the recent (Shear and Edgecombe,
2010).
The number of currently described species of centipedes is close to 3300;
however, the estimated number of species world-wide approaches 7,000 (Adis and
Harvey, 2000; Edgecombe and Giribet, 2007). Centipedes are known from all continents
except Antarctica, with the greatest diversity occurring in the tropics and warm temperate
regions (Edgecombe and Giribet, 2007), and can be found from sea level to high
1
elevations (Ruppert et al., 2004). Centipedes dwell in forests, grasslands, deserts, caves,
and even the littoral zone (Edgecombe and Giribet, 2007), and are frequently found in
soil, leaf litter, or under stones and logs (Ruppert et al., 2004). Some centipede species
are synanthropic (Blackburn et al., 2002; Lesniewska et al., 2008).
Centipedes are soft-bodied, often dorsoventrally flattened, bilaterally
symmetrical, metamerically segmented animals, having 15 to 191 pairs of walking legs,
one pair per trunk segment (Edgecombe and Giribet, 2007; Lewis, 1981). The number of
leg-bearing trunk segments varies between taxa (Lewis, 1981). Adult body length ranges
from 4–300 mm, with most species measuring 10–100 mm (Edgecombe and Giribet,
2007). Of the five extant orders of centipedes, Scolopendromorpha contains the largest
animals (Mundel, 1990). The epidermis secretes a non-living layer, the cuticle, which
forms rigid sclerites separated by flexible arthrodial membranes (Lewis, 1981). The
cuticle is shed periodically to allow growth in a process known as ecdysis (Lewis, 1981).
The anterior part of the body is comprised of a head bearing a pair of antennae, a pair of
mandibles, and two pairs of jointed legs modified to form mouthparts (Lewis, 1981). The
legs of the first trunk segment, functionally incorporated into the head, are modified to
form the forcipules, the anteromedially curving, sharp-tipped, venom-delivery
appendages used to stab and hold prey with a powerful pinching, forceps-like motion
(Dugon and Arthur, 2012a; Dugon et al., 2012).
The forcipules are one of the most prominent features of centipedes and are
evolutionarily very old. Forcipules are found in the oldest known centipede fossils, dated
from the late Silurian (~420 Ma) (Edgecombe, 2011; Murienne et al., 2010). The
centipede venom apparatus thus represents one of the oldest extant venom systems
2
known among terrestrial animals (Undheim and King, 2011). Each forcipule contains a
venom system composed of a glandular epithelium arranged around a cuticular duct,
which opens at an orifice located laterally on the subterminal part of the forcipule (Dugon
and Arthur, 2012a). The glandular epithelium is composed of a multitude of glandular
units arranged radially around the porous proximal part of the duct (Dugon and Arthur,
2012a). Longitudinal, circular, and radial muscle fibers under control of the central
nervous system regulate discharge of the venom (Antoniazzi et al., 2009).
Centipedes are primarily nocturnal carnivores (Lewis, 1981), exploiting a variety
of animal foods including earthworms, termites, grubs, woodlice, spiders, slugs, and
occasionally small vertebrates such as frogs, toads, lizards, snakes, birds, and mice
(Jangi, 1984). With the exception of the scutigeromorphs, which have unusually large,
compound eyes, the vision of centipedes is restricted to distinguishing between light and
darkness, and appears to play little role in hunting (Elzinga, 1994; Jangi, 1984). Rather,
detection of prey is primarily through tactile senses (Elzinga, 1994; Jangi, 1984). Once
prey is detected, the centipede raises up the forepart of its body and then pounces on its
prey with great speed, seizing the prey tightly with multiple pairs of legs while
simultaneously impaling it with the forcipules and presumably injecting venom (Elzinga,
1994; Jangi, 1984). Some Scolopendromorphs will grab prey with their prehensorial anal
legs and then rapidly bend the body in half to bring the forcipules to bear on the prey
(Bücherl, 1971c). The forcipules and second maxillae hold the prey while the mandibles
and first maxillae bite and chew (Brusca and Brusca, 2003; Lewis, 1981). Typically,
when a centipede grasps its prey with the forcipules, it holds on tenaciously (Menez et al.,
1990; Molinari et al., 2005). Chemoreceptors located on the forcipules (Jangi and Dass,
3
1977) may distinguish palatability of prey once it has been captured and pierced, and/or
chemoreception of the internal fluids of the prey could signal penetration of the victim by
the venom claw and stimulate expulsion of venom (Dass and Jangi, 1978; Menez et al.,
1990). Although no studies have been published on how centipedes employ their venom
in predatory stinging, the general assumption throughout the literature is that predatory
stinging is accompanied by venom use (e.g., Bucherl, 1946; Cloudsley-Thompson, 1955;
Dass and Jangi, 1978; Forti et al., 2007; Jangi, 1984; Malta et al., 2008; Menez et al.,
1990; Molinari et al., 2005; Peng et al., 2010; Stankiewicz et al., 1999; Stewart, 1997).
However, several authors have contended that venom use may be conditional (Bücherl,
1971c; Cornwall, 1916; Jangi, 1984), and that the amount of venom deployed can be
regulated (Antoniazzi et al., 2009; Cornwall, 1916; Dass and Jangi, 1978).
The forcipules are also employed in defense (Davis, 1993; Demange, 1981;
Maschwitz et al., 1979; Neck, 1985). Most centipedes, including the large
scolopendromorphs, are not naturally vicious, preferring to run and hide rather than sting
unless molested (Cloudsley-Thompson, 1968; Cornwall, 1916; Freyvogel, 1972;
McMonigle, 2003; Tan and Kretzschmar, 2009). Defensive stinging, based on
symptomology of the stings in humans and animals, may (e.g., McKeown, 1930) or may
not (Anthony et al., 2007; Bush et al., 2001; Cornwall, 1916; Klingel, 1960) be
accompanied by envenomation. The factors influencing defensive stinging in centipedes
remain largely unexplored; however, it is relatively clear that nearly all cases of human
stings by centipedes are likely the result of defensiveness due to intentional or
unintentional rough “handling” (e.g., squeezing, McMonigle, 2003), treading upon (Balit
et al., 2004; Freyvogel, 1972; Jangi, 1984; Logan and Ogden, 1985; Ozsarac et al., 2004;
4
Rodriguez-Acosta et al., 2000; Tan and Kretzschmar, 2009), sitting on (Lewis et al.,
2010), encountering within the confines of bedding (Bush et al., 2001; Lewis et al., 2010;
Mohri et al., 1991; Remington, 1950; Rodriguez-Acosta et al., 2000), or trapping between
clothing and skin (Bush et al., 2001; Mebs, 2002; Mohri et al., 1991; Rodriguez-Acosta et
al., 2000). While most centipede stings of humans are not life threatening (Burnett et al.,
1986; Bush et al., 2001), scolopendromorph stings can be medically significant (Balit et
al., 2004; Jangi, 1984; Lewis, 1981; Logan and Ogden, 1985; Mohri et al., 1991).
Centipedes serve an ecologically important role as soil and leaf litter predators
(Albert, 1983; Robertson et al., 1994; Wallwork, 1976). As generalist predators, they can
exert a strong influence on meso- and macro-invertebrate soil communities (Albert, 1983;
Robertson, 1993). In addition, centipedes function as prey (sometimes dangerous prey,
Jangi, 1984) for numerous invertebrates (Cloudsley-Thompson, 1949, 1968; Crawford,
1990; Davis, 1993; Funasaki et al., 1988; Lewis et al., 2010) and vertebrates (CloudsleyThompson, 1949; Cloudsley-Thompson and Crawford, 1970; Crawford, 1990; Curry,
1986; Davis, 1993; FitzSimmons, 1962; Hamilton and Pollack, 1955; Hoffman, 1954;
Jangi, 1966; Lewis, 1981; Lewis et al., 2010; Ryan and Croft, 1974; Shelley, 2002;
Voinstvenskii et al., 1977), some of the latter including toads, monitor lizards, snakes,
armadillos, mongooses, foxes, burrowing owls, and mockingbirds.
Venom is defined as a toxic substance (comprised of one or more toxins) causing
dose-dependent physiological injury that is passively or actively transferred from one
organism to the internal milieu of another organism via a delivery mechanism and
mechanical injury (Nelsen et al., 2013). Venoms are found in a broad phylogenetic range
of animals, and are used for defense, competitor deterrence, and predation (Fry et al.,
5
2009). Venoms are typically complex mixtures composed of proteins, peptides, salts,
polyamines, amino acids, and neurotransmitters (Fry et al., 2009). In centipedes, known
venom constituents include acid and alkaline phosphatases, phospholipase A2, esterases,
hyaluronidases, metalloproteases, non-metalloproteases, cardiotoxins, CRISPs,
disintegrins, haemolysins, myotoxins, neurotoxins, histamine, and serotonin (Undheim
and King, 2011). Despite recent progress in characterizing centipede venoms (Liu et al.,
2012a; Liu et al., 2012b; Peng et al., 2010; Rates et al., 2007; Undheim et al., 2012; Yang
et al., 2012; Yang et al., 2013), we have only begun to explore the venom diversity of this
taxon. Only a handful of the estimated nearly 7000 species of centipedes have had their
venoms analyzed using large-scale transcriptomic and/or proteomic methods (Liu et al.,
2012b; Undheim et al., 2012; Yang et al., 2012). Considering the diversity of centipedes
and the fact that over 500 different proteins were reported from the venom of a single
species (Liu et al., 2012b), a potentially enormous pharmacological reservoir remains to
be explored in centipede venoms.
Natural selection can be expected to fine-tune the amount, rate of production, and
composition of venom in a given species because venom can be metabolically expensive
to synthesize and store (Billen, 1990; Inceoglu et al., 2003; McCue, 2006; Nisani et al.,
2007, 2012; Pintor et al., 2010, 2011), and because an insufficient venom supply or an
ineffective venom composition can translate into high ecological costs of lost prey
capture opportunities or diminished defense capabilities (Currier et al., 2012; Haight and
Tschinkel, 2003; Hayes, 2008; Malli et al., 1998; Mirtschin et al., 2002).
Despite the diversity of centipedes, the ancient nature of the venom system, and
the medical significance, ecological importance, and promising pharmacological potential
6
of the venom of centipedes, few studies have pursued an understanding of basic aspects
relating to the roles of venom in centipedes, or the factors that influence venom supply
and the timing of venom regeneration. Likewise, few studies have thoroughly
characterized centipede venom components. These kinds of investigations, however,
promise many benefits. Knowledge of venom yields may give insight into the range of
prey types and sizes that a given centipede might be capable of taking (Wigger et al.,
2002), and the effectiveness of defense it could mount. Enhanced knowledge may also
guide physicians in treating centipede-sting patients. Understanding the timing of venom
regeneration is an important step in learning how venom supply might impact the timing
of foraging and other activities that expose the centipedes to predators, and may be useful
for researchers devising venom extraction protocols. Characterizing centipede venom
components will lead to many benefits, including understanding the mechanisms
involved in prey capture and defense, and unlocking a potential trove of bioactive
molecules with insecticidal and therapeutic potential.
One of the overarching objectives of this dissertation is to contribute to the
growing foundation of basic knowledge of centipede venoms in an effort to build a solid
understanding of the behavioral ecology of venom use in centipedes. A second objective
is to provide data that can guide researchers aiming to develop applied uses for centipede
venoms. In recent years, researchers have begun to realize the potential of centipede
venom components as potential leads for the development of new drugs and
bioinsecticides (Bhagirath et al., 2006; Hou et al., 2013; Kong et al., 2013; Kwon et al.,
2013; Peng et al., 2010; Schwartz et al., 2012; Smith et al., 2013; Xu et al., 2013; Yang et
al., 2012; Yang et al., 2013). Knowledge of venom yields, regeneration timetables, and
7
species- or population-specific venom composition can help guide development of
centipede venom extraction protocols and narrow the search for pharmacologically active
biomolecules.
Specific Objectives
In this dissertation, I begin in Chapter 2 by describing how size influences volume
yield and protein concentration of electrically extracted venom in the centipedes
Scolopendra polymorpha and S. subspinipes. I hypothesized a positive association
between body length and volume of venom extracted. I found that volume yield was
positively and linearly related to body length, and that S. subspinipes yielded a larger
length-specific volume than S. polymorpha. Body length and protein concentration were
uncorrelated. Several other factors were found to influence volume yield and protein
concentration, indicating that venom supply can be complex to predict.
In Chapter 3, I investigate venom volume and total protein regeneration during
the 14-day period subsequent to venom extraction in Scolopendra polymorpha. In the
experiment, I further tested the hypothesis that venom protein components, separated by
reversed-phase fast protein liquid chromatography (RP-FPLC), undergo asynchronous
(non-parallel) synthesis. I found that during the first 48 hours, volume and protein mass
increased linearly. However, protein regeneration lagged behind volume regeneration,
with only 65–86% of venom volume and 29–47% of protein mass regenerated during the
first 2 days. No significant additional regeneration occurred over the subsequent 12 days.
Rate of venom regeneration suggests either a lengthy replenishment cycle or a damaging
effect of electrical extraction on the venom glands. Analysis of chromatograms of
8
individual venom samples demonstrated that venom proteins are regenerated
asynchronously.
In Chapter 4, I characterize the venom composition of S. polymorpha using
proteomic methods. I demonstrated that the venom of S. polymorpha is complex,
generating 23 bands by SDS-PAGE and 56 peaks by RP-FPLC. MALDI TOF MS
revealed hundreds of components with masses ranging from 1014.5 to 82863.9 Da.
BLASTp sequence similarity searching of MS/MS-derived amino acid sequences
demonstrated 20 different sequences with similarity to known venom components,
including serine proteases, ion-channel activators/inhibitors, and neurotoxins.
In Chapter 5, I summarize and discuss the results from my research, and include
suggestions for future studies.
In Appendix A, I review the strategic deployment of predatory and defensive
emissions in animals, including venoms, predatory glues, and non-venomous defensive
secretions. I present examples from many taxa that together illustrate the following trends
related to the utilization of these valuable emissions: emissions are deployed (1) only
under certain conditions, (2) in amounts that can vary with circumstances, (3) from the
location on the emitter’s body most proximate to the triggering stimulus (when multiple
emission locations are possible), (4) specifically aimed toward the intended receiver, and
(5) in a manner that allows for recovery or reuptake of emitted material.
9
References
Adis, J., Harvey, M.S., 2000. How many Arachnida and Myriapoda are there world-wide
and in Amazonia? Stud. Neotrop. Faun. Environ. 35, 139-141.
Albert, A.M., 1983. Characteristics of two populations of Lithobiidae (Chilopoda)
determined in the laboratory and their relevance with regard to their ecological
role as predators. Zool. Anz. 211, 214-226.
Anthony, C.D., Hickerson, C.A.M., Venesky, M.D., 2007. Responses of juvenile
terrestrial salamanders to introduced (Lithobius forficatus) and native centipedes
(Scolopocryptops sexspinosus). J. Zool. 271, 54-62.
Antoniazzi, M.M., Pedroso, C.M., Knysak, I., Martins, R., Guizze, S.P.G., Jared, C.,
Barbaro, K.C., 2009. Comparative morphological study of the venom glands of
the centipede Cryptops iheringi, Otostigmus pradoi and Scolopendra viridicornis.
Toxicon 53, 367-374.
Balit, C.R., Harvey, M.S., Waldock, J.M., Isbister, G.K., 2004. Prospective study of
centipede bites in Australia. J. Toxicol. Clin. Toxicol. 42, 41-48.
Bhagirath, T., Chingtham, B., Mohen, Y., 2006. Venom of a hill centipede Scolopendra
viridicornis inhibits growth of human breast tumor in mice. Indian J. Pharmacol
38, 291.
Billen, J., 1990. A survey of the glandular system of fire ants, in: Vander Meer, R.K.,
Jaffe, K., Cedeno, A. (Eds.), Applied myrmeclogy: a world perspective. Westview
Press, Boulder, pp. 85-94.
Blackburn, J., Farrow, M., Arthur, W., 2002. Factors influencing the distribution,
abundance and diversity of geophilomorph and lithobiomorph centipedes. J. Zool.
256, 221-232.
Brusca, R.C., Brusca, G.J., 2003. Invertebrates. Sinauer Associates, Sunderland, Mass.
Bucherl, W., 1946. Acao do veneno dos Escolopendromorfos do Brasil sobre alguns
animais de laboratorio. Mem. Inst. Butantan 19, 181-197.
Bücherl, W., 1971. Venomous chilopods or centipedes, in: Bucherl, W., Buckley, E.E.,
Deulofeu, V. (Eds.), Venomous invertebrates. Academic Press, New York, pp.
169-196.
Burnett, J.W., Calton, G.J., Morgan, R.J., 1986. Environment versus man — centipedes.
Cutis 37, 241.
Bush, S.P., King, B.O., Norris, R.L., Stockwell, S.A., 2001. Centipede envenomation.
Wilderness Environ. Med. 12, 93-99.
10
Cloudsley-Thompson, J.L., 1949. The enemies of myriapods. Naturalist (Lond.)137-141.
Cloudsley-Thompson, J.L., 1955. Some aspects of the biology of centipedes and
scorpions. Naturalist (Lond.) 147-153.
Cloudsley-Thompson, J.L., 1968. Spiders, scorpions, centipedes, and mites. Pergamon
Press, Oxford.
Cloudsley-Thompson, J.L., Crawford, C.S., 1970. Water and temperature relations, and
diurnal rhythms of scolopendromorph centipedes. Entomol. Exp. Appl. 13, 187193.
Cornwall, J.W., 1916. Some centipedes and their venom. Indian J. Med. Res. 3, 541-557.
Crawford, C.S., 1990. Scorpions, Solifugae and associated desert taxa, in: Dindal, D.L.
(Ed.), Soil biology guide. John Wiley & Sons, New York, pp. 421-475.
Currier, R.B., Calvete, J.J., Sanz, L., Harrison, R.A., Rowley, P.D., Wagstaff, S.C., 2012.
Unusual stability of messenger RNA in snake venom reveals gene expression
dynamics of venom replenishment. PLoS One 7, e41888.
doi:10.1371/journal.pone.0041888.
Curry, R.L., 1986. Galapagos mockingbird kleptoparasitizes centipede. Condor 88, 119120.
Dass, C.M.S., Jangi, B.S., 1978. Ultrastructural organization of the poison gland of the
centipede Scolopendra morsitans Linn. Indian J. Exp. Biol. 16, 748-757.
Davis, J.M. 1993. Burrow plugging in centipedes: an anti-invasion function [M.S. thesis].
Arlington: University of Texas at Arlington. 51 p.
Demange, J.M., 1981. Les Mille-pattes, Myriapodes. Societe Nouvelle des Editions
Boubee, Paris.
Dugon, M.M., Arthur, W., 2012. Comparative studies on the structure and development
of the venom-delivery system of centipedes, and a hypothesis on the origin of this
evolutionary novelty. Evol. Dev. 14, 128-137.
Dugon, M.M., Black, A., Arthur, W., 2012. Variation and specialisation of the forcipular
apparatus of centipedes (Arthropoda: Chilopoda): a comparative morphometric
and microscopic investigation of an evolutionary novelty. Arthropod Struct. Dev.
41, 231-243.
Edgecombe, G.D., 2011. Chilopoda–fossil history, in: Minelli, A. (Ed.), The Myriapoda.
Vol. 1. Brill, Leiden, pp. 355-361.
11
Edgecombe, G.D., Giribet, G., 2007. Evolutionary biology of centipedes (Myriapoda:
Chilopoda). Annu. Rev. Entomol. 52, 151-170.
Elzinga, R.J., 1994. The use of legs as grasping structures during prey capture and
feeding by the centipede Scolopendra viridis Say (Chilopoda, Scolopendridae). J.
Kans. Entomol. Soc. 67, 369-372.
FitzSimmons, V.F.M., 1962. Snakes of southern Africa. Macdonald, London.
Forti, L.R., Fischer, H.Z., Encarnacao, L.C., 2007. Treefrog Dendropsophus elegans
(Wied-Neuwied, 1824) (Anura: Hylidae) as a meal to Otostigmus tibialis
Brolemann, 1902 (Chilopoda: Scolopendridae) in the tropical rainforest in
southeastern Brazil. Braz. J. Biol. 67, 583-584.
Freyvogel, T.A., 1972. Poisonous and venomous animals in East Africa. Acta Trop. 29,
401-451.
Fry, B.G., Roelants, K., Champagne, D.E., Scheib, H., Tyndall, J.D.A., King, G.F.,
Nevalainen, T.J., Norman, J.A., Lewis, R.J., Norton, R.S., Renjifo, C., de la Vega,
R.C.R., 2009. The toxicogenomic multiverse: convergent recruitment of proteins
into animal venoms. Annu. Rev. Genomics Hum. Genet. 10, 483-511.
Funasaki, G.Y., Lai, P.-Y., Nakahara, L.M., Beardsley, J.W., Ota, A.K., 1988. A review
of the biological control introductions in Hawaii: 1890-1985. Proc. Hawaii.
Entomol. Soc. 28, 105-160.
Haight, K.L., Tschinkel, W.R., 2003. Patterns of venom synthesis and use in the fire ant,
Solenopsis invicta. Toxicon 42, 673-682.
Hamilton, W.J., Jr., Pollack, J.A., 1955. The food of some crotalid snakes from Fort
Benning, Georgia. Chicago Academy of Science and Natural History
Miscellaneous publication No. 140, 1-4.
Hayes, W.K., 2008. The snake venom-metering controversy: levels of analysis,
assumptions, and evidence, in: Hayes, W.K., Beaman, K.R., Cardwell, M.D.,
Bush, S.P. (Eds.), The biology of rattlesnakes. Loma Linda University Press,
Loma Linda, pp. 191-220.
Hoffman, R.L., 1954. Distributional records for two species of Plethodon in the southern
Appalachians. Herpetologica 10, 191-193.
Hou, H.H., Yan, W.L., Du, K.X., Ye, Y.J., Cao, Q.Q., Ren, W.H., 2013. Construction
and expression of an antimicrobial peptide scolopin 1 from the centipede venoms
of Scolopendra subspinipes mutilans in Escherichia coli using SUMO fusion
partner. Protein Expression Purif. 92, 230-234.
12
Inceoglu, B., Lango, J., Jing, J., Chen, L.L., Doymaz, F., Pessah, I.N., Hammock, B.D.,
2003. One scorpion, two venoms: prevenom of Parabuthus transvaalicus acts as
an alternative type of venom with distinct mechanism of action. Proc. Natl. Acad.
Sci. U.S.A. 100, 922-927.
Jangi, B.S., 1966. Scolopendra (the Indian centipede). Zoological Society of India,
Calcutta.
Jangi, B.S., 1984. Centipede venoms and poisoning, in: Tu, A.T. (Ed.), Handbook of
natural toxins, volume 2: insect poisons, allergens, and other invertebrate venoms.
M. Dekker, New York, pp. 333-368.
Jangi, B.S., Dass, C.M.S., 1977. Chemoreceptive function of the poison fang in the
centipede Scolopendra morsitans L. Indian J. Exp. Biol. 15, 803-804.
Klingel, H., 1960. Vergleichende verhaltensbiologie der chilopoden Scutigera
coleoptrata L. (Spinnenassel) und Scolopendra cingulata Latreille (Skolopender).
Z Tierpsychol 17, 11-30.
Kong, Y., Shao, Y., Chen, H., Ming, X., Wang, J.B., Li, Z.Y., Wei, J.F., 2013. A novel
factor Xa-inhibiting peptide from centipedes venom. Int. J. Pept. Res. Ther. 19,
303-311.
Kwon, Y.N., Lee, J.H., Kim, I.W., Kim, S.H., Yun, E.Y., Nam, S.H., Ahn, M.Y., Jeong,
M., Kang, D.C., Lee, I.H., Hwang, J.S., 2013. Antimicrobial activity of the
synthetic peptide Scolopendrasin II from the centipede Scolopendra subspinipes
mutilans. J. Microbiol. Biotechnol. 23, 1381-1385.
Lesniewska, M., Lesniewski, P., Szybiak, K., 2008. Effect of urbanization on centipede
(Chilopoda) diversity in the Wielkopolska-Kujawy Lowlands of western Poland.
Biologia 63, 711-719.
Lewis, J.G.E., 1981. The biology of centipedes. Cambridge University Press, Cambridge.
Lewis, J.G.E., Daszak, P., Jones, C.G., Cottingham, J.D., Wenman, E., Maljkovic, A.,
2010. Field observations on three scolopendrid centipedes from Mauritius and
Rodrigues (Indian Ocean) (Chilopoda: Scolopendromorpha). Int. J. Myriap. 3,
123-137.
Liu, W.H., Luo, F., He, J., Cao, Z.J., Miao, L.X., 2012a. Molecular cloning and
characterization of a new cDNA sequence encoding a venom peptide from the
centipede Scolopendra subspinipes mutilans. Mol. Biol. 46, 508-513.
13
Liu, Z.C., Zhang, R., Zhao, F., Chen, Z.M., Liu, H.W., Wang, Y.J., Jiang, P., Zhang, Y.,
Wu, Y., Ding, J.P., Lee, W.H., 2012b. Venomic and transcriptomic analysis of
centipede Scolopendra subspinipes dehaani. J. Proteome Res. 11, 6197-6212.
Logan, J.L., Ogden, D.A., 1985. Rhabdomyolysis and acute renal failure following the
bite of the giant desert centipede Scolopendra heros. West. J. Med. 142, 549-550.
Malli, H., Imboden, H., Kuhn-Nentwig, L., 1998. Quantifying the venom dose of the
spider Cupiennius salei using monoclonal antibodies. Toxicon 36, 1959-1969.
Malta, M.B., Lira, M.S., Soares, S.L., Rocha, G.C., Knysak, I., Martins, R., Guizze,
S.P.G., Santoro, M.L., Barbaro, K.C., 2008. Toxic activities of Brazilian
centipede venoms. Toxicon 52, 255-263.
Maschwitz, U., Lauschke, U., Wurmli, M., 1979. Hydrogen cyanide-producing glands in
a scolopender, Asanada n.sp. (Chilopoda, Scolopendridae). J. Chem. Ecol. 5, 901907.
McCue, M.D., 2006. Cost of producing venom in three North American pitviper species.
Copeia 2006, 818-825.
McKeown, K.C., 1930. Centipedes and centipede bites. Aust. Mus. Mag. 4, 59-60.
McMonigle, O., 2003. Giant centipedes: the enthusiast's handbook. Elytra & Antenna,
Brunswick Hills.
Mebs, D., 2002. Venomous and poisonous animals: a handbook for biologists,
toxicologists and toxinologists, physicians and pharmacists. CRC Press, Boca
Raton.
Menez, A., Zimmerman, K., Zimmerman, S., Heatwole, H., 1990. Venom apparatus and
toxicity of the centipede Ethmostigmus rubripes (Chilopoda, Scolopendridae). J.
Morphol. 206, 303-312.
Mirtschin, P.J., Shine, R., Nias, T.J., Dunstan, N.L., Hough, B.J., Mirtschin, M., 2002.
Influences on venom yield in Australian tigersnakes (Notechis scutatus) and
brownsnakes (Pseudonaja textilis: Elapidae, Serpentes). Toxicon 40, 1581-1592.
Mohri, S., Sugiyama, A., Saito, K., Nakajima, H., 1991. Centipede bites in Japan. Cutis
47, 189-190.
Molinari, J., Gutierrez, E.E., De Ascencao, A.A., Nassar, J.M., Arends, A., Marquez,
R.J., 2005. Predation by giant centipedes, Scolopendra gigantea, on three species
of bats in a Venezuelan cave. Caribb. J. Sci. 41, 340-346.
14
Mundel, P., 1990. Chilopoda, in: Dindal, D.L. (Ed.), Soil biology guide. John Wiley &
Sons, New York, pp. 819-834.
Murienne, J., Edgecombe, G.D., Giribet, G., 2010. Including secondary structure, fossils
and molecular dating in the centipede tree of life. Mol. Phylogen. Evol. 57, 301313.
Neck, R.W., 1985. Comparative behavior and external color patterns of two sympatric
centipedes (Chilopoda: Scolopendra) from central Texas. Tex. J. Sci. 37, 253-255.
Nelsen, D.R., Nisani, Z., Cooper, A.M., Fox, G.A., Gren, E.C.K., Corbit, A.G., Hayes,
W.K., 2013. Poisons, toxungens, and venoms: redifining and classifying toxic
biological secretions and the organsims that employ them. Biol. Rev. doi:
10.1111/brv.12062.
Nisani, Z., Boskovic, D.S., Dunbar, S.G., Kelln, W., Hayes, W.K., 2012. Investigating
the chemical profile of regenerated scorpion (Parabuthus transvaalicus) venom in
relation to metabolic cost and toxicity. Toxicon 60, 315-323.
Nisani, Z., Dunbar, S.G., Hayes, W.K., 2007. Cost of venom regeneration in Parabuthus
transvaalicus (Arachnida: Buthidae). Comp. Biochem. Physiol., A: Mol. Integr.
Physiol. 147, 509-513.
Ozsarac, M., Karcioglu, O.U., Ayrik, C., Somuncu, F., Gumrukcu, S., 2004. Acute
coronary ischemia following centipede envenomation: case report and review of
the literature. Wilderness Environ. Med. 15, 109-112.
Peng, K.F., Kong, Y., Zhai, L., Wu, X.F., Jia, P., Liu, J.Z., Yu, H.N., 2010. Two novel
antimicrobial peptides from centipede venoms. Toxicon 55, 274-279.
Pintor, A.F.V., Krockenberger, A.K., Seymour, J.E., 2010. Costs of venom production in
the common death adder (Acanthophis antarcticus). Toxicon 56, 1035-1042.
Pintor, A.F.V., Winter, K.L., Krockenberger, A.K., Seymour, J.E., 2011. Venom
physiology and composition in a litter of common death adders (Acanthophis
antarcticus) and their parents. Toxicon 57, 68-75.
Rates, B., Bemquerer, M.P., Richardson, M., Borges, M.H., Morales, R.A.V., De Lima,
M.E., Pimenta, A.M.C., 2007. Venomic analyses of Scolopendra viridicornis
nigra and Scolopendra angulata (Centipede, Scolopendromorpha): shedding light
on venoms from a neglected group. Toxicon 49, 810-826.
Remington, C.L., 1950. The bite and habits of a giant centipede (Scolopendra
subspinipes) in the Philippine Islands. Am. J. Trop. Med. Hyg. 30, 453-455.
15
Robertson, L.N., 1993. Influence of tillage intensity (including no till) on soil-dwelling
pests and predatory soil animals in central Queensland crops, in: Covey, S.A.,
Dall, D.J., Milne, W.M. (Eds.), Pest control and sustainable agriculture. CSIRO,
Canberra, pp. 349-352.
Robertson, L.N., Kettle, B.A., Simpson, G.B., 1994. The influence of tillage practices on
soil macrofauna in a semiarid agroecosystem in northeastern Australia. Agric.
Ecosyst. Environ. 48, 149-156.
Rodriguez-Acosta, A., Gassette, J., Gonzalez, A., Ghisoli, M., 2000. Centipede
(Scolopendra gigantea Linneaus 1758) envenomation in a newborn. Rev. Inst.
Med. Trop. Sao Paulo 42, 341-2.
Ruppert, E.E., Fox, R.S., Barnes, R.D., 2004. Invertebrate zoology : a functional
evolutionary approach. Thomson-Brooks/Cole, Belmont.
Ryan, G.E., Croft, J.D., 1974. Observations on the food of the fox, Vulpes vulpes (L.) in
Kinchega National Park, Menindee, NSW. Australian Wildlife Research 1, 89-94.
Schwartz, E.F., Mourao, C.B.F., Moreira, K.G., Camargos, T.S., Mortari, M.R., 2012.
Arthropod venoms: a vast arsenal of insecticidal neuropeptides. Biopolymers 98,
385-405.
Shear, W.A., Edgecombe, G.D., 2010. The geological record and phylogeny of the
Myriapoda. Arthropod Struct. Dev. 39, 174-190.
Shelley, R.M., 2002. A synopsis of the North American centipedes of the order
Scolopendromorpha (Chilopoda). Virginia Museum of Natural History,
Martinsville.
Smith, J.J., Herzig, V., King, G.F., Alewood, P.F., 2013. The insecticidal potential of
venom peptides. Cell. Mol. Life Sci. 70, 3665-3693.
Stankiewicz, M., Hamon, A., Benkhalifa, R., Kadziela, W., Hue, B., Lucas, S., Mebs, D.,
Pelhate, M., 1999. Effects of a centipede venom fraction on insect nervous
system, a native Xenopus oocyte receptor and on an expressed Drosophila
muscarinic receptor. Toxicon 37, 1431-1445.
Stewart, J.W. 1997. Centipedes and millipedes. Texas Agricultural Extension Service:
Texas A & M University.
Tan, K., Kretzschmar, H., 2009. On a meeting between the Horn Viper and a centipede in
the Peloponnese, Southern Greece or the biter, bit. Adv. Sci. Lett. 2, 10-13.
Undheim, E.A.B., Jones, A., Holland, J.W., Morales, R.A.V., Winnen, B., Fry, B.G.,
King, G.F., 2012. Centipede venoms: old and unusual. Toxicon 60, 126.
16
Undheim, E.A.B., King, G.F., 2011. On the venom system of centipedes (Chilopoda), a
neglected group of venomous animals. Toxicon 57, 512-524.
Voinstvenskii, M.A., Petrusenko, A.A., Boyarchuk, V.P., 1977. Trophic relations of the
rook (Corvus frugilegus L.) in steppe ecosystem: 1 - Nutrition (diet composition).
Vestn. Zool. 6, 9-17.
Wallwork, J.A., 1976. The distribution and diversity of soil fauna. Academic Press,
London.
Wigger, E., Kuhn-Nentwig, L., Nentwig, W., 2002. The venom optimisation hypothesis:
a spider injects large venom quantities only into difficult prey types. Toxicon 40,
749-752.
Xu, S.Y., Zhang, F., Wang, H.Y., Liu, Y., Li, D., Wu, Z., Liu, Z.H., 2013. The venom of
the centipede Scolopendra subspinipes mutilans inhibits the growth of
myelogenous leukemia cell lines. Lett. Drug Des. Discov. 10, 390-395.
Yang, S., Liu, Z., Xiao, Y., Li, Y., Rong, M., Liang, S., Zhang, Z., Yu, H., King, G.F.,
Lai, R., 2012. Chemical punch packed in venoms makes centipedes excellent
predators. Mol. Cell. Proteomics 11, 640-650.
Yang, S., Xiao, Y., Kang, D., Liu, J., Li, Y., Undheim, E.A.B., Klint, J.K., Rong, M., Lai,
R., King, G.F., 2013. Discovery of a selective Nav1.7 inhibitor from centipede
venom with analgesic efficacy exceeding morphine in rodent pain models. Proc.
Natl. Acad. Sci. U.S.A., www.pnas.org/cgi/doi/10.1073/pnas.1306285110.
17
CHAPTER TWO
VARIATION IN VENOM YIELD AND PROTEIN CONCENTRATION OF THE
CENTIPEDES SCOLOPENDRA POLYMORPHA AND SCOLOPENDRA SUBSPINIPES
Allen M. Coopera,*, Gerad A. Foxa, David R. Nelsena, William K. Hayesa
a
Department of Earth and Biological Sciences, Griggs Hall #101, Loma Linda University,
24941 Stewart St., Loma Linda, CA, 92350, USA
Published in Toxicon: http://dx.doi.org/10.1016/j.toxicon.2014.02.003
*Corresponding author:
Allen M. Cooper
Department of Earth and Biological Sciences
Griggs Hall #101
24941 Stewart St.
Loma Linda, CA 92350
USA
Tel: 1-707-225-5018
Fax: 1-909-558-0259
Email: [email protected]
18
Abstract
Venom generally comprises a complex mixture of compounds representing a nontrivial metabolic expense. Accordingly, natural selection should fine-tune the amount of
venom carried within an animal’s venom gland(s). The venom supply of
scolopendromorph centipedes likely influences their venom use and has implications for
the severity of human envenomations, yet we understand very little about their venom
yields and the factors influencing them. We investigated how size, specifically body
length, influenced volume yield and protein concentration of electrically extracted venom
in Scolopendra polymorpha and Scolopendra subspinipes. We also examined additional
potential influences on yield in S. polymorpha, including relative forcipule size, relative
mass, geographic origin (Arizona vs. California), sex, time in captivity, and milking
history. Volume yield was linearly related to body length, and S. subspinipes yielded a
larger length-specific volume than S. polymorpha. Body length and protein concentration
were uncorrelated. When considering multiple influences on volume yield in S.
polymorpha, the most important factor was body length, but yield was also positively
associated with relative forcipule length and relative body mass. Scolopendra
polymorpha from California yielded a greater volume of venom with a higher protein
concentration than conspecifics from Arizona, all else being equal. Previously milked
animals yielded less venom with a lower protein concentration. For both species,
approximately two-thirds of extractable venom was expressed in the first two pulses, with
remaining pulses yielding declining amounts, but venom protein concentration did not
vary across pulses. Further study is necessary to ascertain the ecological significance of
the factors influencing venom yield and how availability may influence venom use.
19
Introduction
Many animals depend on venoms to procure food, defend themselves, or deter
competitors (Mebs, 2002). Maintaining a sufficient venom supply is essential to avoid the
serious costs of venom depletion, including lost prey capture opportunities and
diminished defensive capabilities (Currier et al., 2012; Haight and Tschinkel, 2003;
Hayes, 2008; Malli et al., 1998). Because venom is generally a complex mixture of
compounds (Fry et al., 2009; Rodriguez de la Vega et al., 2010; Undheim and King,
2011), representing a non-trivial metabolic expense (Billen, 1990; McCue, 2006; Nisani
et al., 2012; Nisani et al., 2007; Pintor et al., 2010, 2011), natural selection should finetune the amount of venom carried within an animal’s venom gland(s) (Mirtschin et al.,
2002). The amount of venom an animal possesses may be influenced by ultimate factors
such as prey type, prey size, and rates of prey encounter and venom regeneration
(Mirtschin et al., 2002).
One measure of the amount of venom an animal possesses is venom yield, the
quantity of venom expelled, either voluntarily or involuntarily, from an intact, live
animal. Venom yield in arthropods has been measured most commonly in terms of dry
mass (Herzig, 2010; Herzig et al., 2008), volume (de Roodt et al., 2012; Fox et al., 2009),
and wet mass (Rocha-e-Silva et al., 2009; Sahayaraj et al., 2006). Venom yield often
refers to the maximum amount of venom that can be expelled using a given extraction
technique, such as electrical milking (Herzig et al., 2004; McCleary and Heard, 2010),
glandular massage (Mackessy, 1988), administration of saliva-inducing chemicals such
as pilocarpine (Hill and Mackessy, 1997), and spontaneous ejection (Hopkins et al., 1995;
Sahayaraj et al., 2006; Tare et al., 1986). Venom yield can be influenced by diverse
factors, including body size (Fox et al., 2009; Vapenik and Nentwig, 2000), age (Brown,
20
1973; Malli et al., 1993), sex (Atkinson, 1981; Glenn and Straight, 1977; Rocha-e-Silva
et al., 2009), season (Bücherl, 1953; Vieira et al., 1988; Wiener, 1956, 1959), temperature
(Gregory-Dwyer et al., 1986; Kochva, 1960; Morgan, 1969), humidity (Kristensen,
2005), geographic population (Binford, 2001; Mirtschin and Davis, 1992; Mirtschin et
al., 2002), health (Brown, 1973; Klauber, 1997b), and number and frequency of milkings
(Kristensen, 2005; Perret, 1977b; Sissom et al., 1990).
The class Chilopoda, part of the subphylum Myriapoda, is divided into five living
orders (and 1 extinct): Scutigeromorpha, Lithobiomorpha, Craterostigmomorpha,
Scolopendromorpha, and Geophilomorpha (Edgecombe and Giribet, 2007). This diverse
group of terrestrial arthropods (an estimated 3500 species) serves an ecologically
important role as soil and leaf litter predators (Edgecombe and Giribet, 2007; Robertson
et al., 1994; Trucchi et al., 2009; Undheim and King, 2011; Wallwork, 1976). Despite
their importance, our knowledge of the natural history of centipedes is very limited (Forti
et al., 2007; Molinari et al., 2005). Anatomically, centipedes possess a long, segmented
body with one pair of legs per segment, and a head containing a pair of long antennae
(Lewis, 1981). The legs of the first trunk segment are modified to form the characteristic
forcipules that are used to grasp and envenomate prey (Bonato et al., 2010; Lewis, 1981;
Undheim and King, 2011). The forcipules are also employed in defense (Davis, 1993;
Demange, 1981; Maschwitz et al., 1979; Neck, 1985). Although prey immobilization
(Undheim and King, 2011) comprises the primary role of the relatively complex venom
(Liu et al., 2012a; Liu et al., 2012b; Undheim and King, 2011), a digestive function has
also been suggested (Jangi, 1984; Martin, 1971; Minton, 1974), but remains unclear
21
(Bücherl, 1971a; Dugon and Arthur, 2012b). To date, we understand very little about
venom yields and the factors influencing them in centipedes.
Of the five extant orders of centipedes, the fleet-footed Scolopendromorpha,
ranging in adult length from 1 to 30 cm (Edgecombe and Koch, 2008), contains the
largest (Mundel, 1990), most fiercely predatory (Edgecombe and Koch, 2008), and most
medically important (Balit et al., 2004; Jangi, 1984) of all centipedes. Within the
Scolopendromorpha, the family Scolopendridae comprises powerfully muscled and
stoutly built centipedes that potentially pose serious health hazards to humans due to their
venomous sting (Jangi, 1984). In this study we investigated venom yields of two
scolopendrids, Scolopendra polymorpha and S. subspinipes.
Scolopendra polymorpha inhabits desert, dry grassland, and forest habitats from
the Great Plains westward to California, ranging up the Pacific states into Oregon, and
throughout the desert southwest into northern Mexico (Crabill, 1960; Crawford and
Riddle, 1974; Shelley, 2002). The sting of S. polymorpha causes temporary sharp pain in
humans (Baerg, 1924; Maldonado, 1998; Turk, 1951). Scolopendra subspinipes is
cosmopolitan in tropical and subtropical regions of the world (Kronmuller, 2012; Lewis
et al., 2010). The sting of S. subspinipes reportedly causes intense pain, burning,
swelling, and erythema (Bouchard et al., 2004; Bush et al., 2001; Mohri et al., 1991;
Veraldi et al., 2010).
There is a growing body of knowledge regarding scolopendromorphs and their
venoms (reviewed in Undheim and King, 2011). While there have only been a few
studies focusing on scolopendromorph behavioral use of venom (e.g., Dugon and Arthur,
2012b; Formanowicz and Bradley, 1987), recent studies have shed light on the venom
22
apparatus (Antoniazzi et al., 2009; Chao and Chang, 2006; Dugon and Arthur, 2012a;
Dugon et al., 2012; Jarrar, 2010), venom transcriptome (Liu et al., 2012a; Liu et al.,
2012b; Undheim et al., 2012; Yang et al., 2012), and venom components and
biochemistry (Gonzalez-Morales et al., 2009; Guo et al., 2013; Jarrar, 2010; Liu et al.,
2012a; Liu et al., 2012b; Malta et al., 2008; Peng et al., 2010; Rates et al., 2007; Yang et
al., 2012, 2013). Other recent studies reported the effects of venom on invertebrates
(Rates et al., 2007; Yang et al., 2012), non-human vertebrates (Malta et al., 2008; Menez
et al., 1990), and humans (Chaou et al., 2009; Haddad et al., 2012; Lewis et al., 2010;
Othong et al., 2012; Uzel et al., 2009; Veraldi et al., 2010). With the exception of
incidental observations described by Malta et al. (2008), no quantitative data have been
published regarding the quantity of venom centipedes have at their disposal (Minton,
1974), or the factors that influence venom yield or protein concentration, despite venom
being a key element in the predatory behavior of centipedes (Dugon and Arthur, 2012b;
Quistad et al., 1992).
Knowledge of venom yields and factors influencing them in centipedes are
important for a fuller understanding of centipede venom use, and may prove useful to the
medical community. The amount of venom an animal produces likely influences the
severity of envenomation following a sting or bite (Janes et al., 2010; Mirtschin et al.,
2002), and thus understanding venom yields may give insight into the range of prey types
and sizes that a given centipede might be capable of taking (Wigger et al., 2002), as well
as the effectiveness of defense it could mount. While most centipede stings of humans are
not life threatening (Burnett et al., 1986; Bush et al., 2001), scolopendromorph stings can
be medically significant (Balit et al., 2004; Jangi, 1984; Lewis, 1981; Logan and Ogden,
23
1985; Mohri et al., 1991). If knowledge of venom yields in S. polymorpha and S.
subspinipes serves as a foundation for an understanding of yields in other
scolopendromorphs, then perhaps the characteristics of an offending centipede can help
guide physicians in their treatment of the patient (Hayes and Mackessy, 2010; Janes et
al., 2010).
Evidence from centipede anatomy supports the hypothesis that venom yields in
scolopendromorphs relate to body size. The venom glands are located inside the external
lateral face of each forcipule (Antoniazzi et al., 2009) and extend from the
trochanteroprefemoral (Bonato et al., 2010; Lewis et al., 2005) segment of the forcipule
to the basal part of the tarsungulum (Jangi, 1984; Lewis et al., 2005; Menez et al., 1990),
although exceptions exist (Edgecombe and Koch, 2008). In the three scolopendromorphs
studied by Antoniazzi et al. (2009), venom gland length represented 5–6% of the total
adult animal length.
Circumstantial evidence from the effects of centipede stings in humans further
suggests that larger centipedes possess and deploy more venom than smaller centipedes.
Reports indicate that larger centipedes cause more painful stings (Balit et al., 2004;
Harwood et al., 1979; McFee et al., 2002; Norris, 2007; Undheim and King, 2011) with a
higher incidence of swelling (Balit et al., 2004; Maldonado, 1998). Duration of pain also
reportedly varies with centipede size (Gomes et al., 1982a). However, with the exception
of the comments of Maldonado (1998) and Gomes et al. (1982a), which indicate an
intraspecific relationship between centipede size and sting-symptom severity, it remains
unclear whether sting severity varies with size within a given species or whether venom
differences between species of differing size are contributing to the reported size-
24
symptom relationship. Without further intraspecific studies it is difficult to draw firm
conclusions regarding the relationship between size of centipede of a given species and
quantity of venom possessed or injected.
We designed the present study to determine if and how size, specifically body
length, influenced yield and protein concentration of electrically extracted venom in the
centipedes S. polymorpha and S. subspinipes. We also compared venom yields and
venom protein concentrations between these two species, and investigated additional
potential sources of venom yield and protein concentration variation in S. polymorpha,
including relative forcipule size, relative mass, geographic origin, sex, time in captivity,
and milking history.
Materials and Methods
Centipedes
We purchased live S. polymorpha (n = 153; collected from Cochise County, AZ;
Fig. 1A) and S. subspinipes (n = 6; purportedly from Vietnam; Fig. 1B) from Bugs of
America LLC (Portal, Arizona, USA); additional S. polymorpha (n = 40) were collected
from southern California (San Bernardino, Riverside, and San Diego counties). We
determined sex of the centipedes under a Nikon SMZ-10A stereo dissecting scope (Nikon
Corp., Melville, NY, USA) by visual inspection of genitalia (Jangi, 1966). Final sample
sizes for analysis were n = 177 (88 AZ males, 49 AZ females; 18 CA males, 22 CA
females) S. polymorpha (mean ± SE and range of time in our captivity prior to venom
extraction: 242 ± 18, 24–968 days), and n = 6 (5 male, 1 female) S. subspinipes (27 ± 3,
22–36 days). For S. polymorpha, 10 animals (7 AZ males, 3 AZ females) had been
previously milked twice, with milkings occurring 5 and 3 weeks prior to the collection of
25
data analyzed here. The remaining 167 S. polymorpha and all 6 S. subspinipes had never
before been milked.
26
27
28
We housed the centipedes individually in plastic containers (18 x 11 x 22 cm L x
W x H) with a soil substrate, water dish, and hide box. They were fed a cricket (Acheta
domesticus) once every 2 weeks, and misted weekly. We obtained morphometric
measurements prior to venom extraction from digital photographs of CO2-anesthetized
animals. Measures included: body length (distance between the anterior margin of the
cephalic plate and the posterior margin of the postpedal tergite, excluding the caudal legs,
to nearest 1 mm; Shelley, 2002; Simaiakis et al., 2011); forcipule width (mean of left and
right forcipule widths, measured as distance between the medial and lateral margins of
the trochanteroprefemur in line with the anterior margin of the forcipular coxosternite, to
nearest 0.01 mm; Fig. 2A; cf. Dugon et al., 2012); and forcipule length (mean of left and
right forcipule lengths, measured as distance between a line drawn through the most
posterior attachment points of the trochanteroprefemora with the forcipular coxosternite
and a parallel line at the anterior tip of the tarsungulum, to nearest 0.01 mm; Fig. 2A).
Although forcipule length, by our definition, varies with degree of articulation of
forcipular joints, all photographs were of anesthetized animals; therefore, forcipule length
was consistently measured with forcipules in the resting (i.e., folded rather than
extended) position. For centipedes with a forcipule whose tarsungulum appeared
obviously blunted (presumably from injury and subsequent regeneration) but which still
yielded venom (n = 2), forcipule length was based solely on the length of the uninjured
forcipule rather than on the average of both forcipules.
29
30
31
We chose body length as the primary measure of centipede size in analyses
throughout this paper, as we believe it better reflects long-term fitness than mass, which
more likely represents short-term variations in fitness (Vapenik and Nentwig, 2000).
Precedent also establishes use of body length as a preferred indicator of centipede size
(Hayden et al., 2012; Lewis, 1981; Maldonado, 1998; Pilz et al., 2007; Shelley, 2002).
Body lengths used in the final data set for S. polymorpha (n = 177; see section 2.2)
ranged from 5.6 to 13.8 cm (possibly including both sexually mature and immature
individuals; Maldonado, 1998), whereas S. subspinipes lengths ranged more narrowly
from 14.1 to 15.8 cm. Scolopendra polymorpha averaged shorter than S. subspinipes (8.9
± 0.1 vs. 14.8 ± 0.2 cm, respectively; t-test adjusted for unequal variance: t7.05 = 23.08, p
< 0.001, Cohen's d = 4.43; mean difference and 95% CI = 5.9 [5.3–6.5]), with no overlap.
The narrow size range of S. subspinipes rendered the intraspecific relationship between
body size and venom yield unreliable. However, in some analyses we pooled the two
species to examine the general effect of body size within the genus and to compare the
species while accounting for body size differences. Body lengths for S. polymorpha after
excluding previously milked specimens averaged 8.8 ± 0.1 cm (range 5.6–11.5 cm).
Venom Extraction and Venom Volume Determination
Venom extraction and collection were carried out under a stereo dissecting scope.
We anesthetized centipedes using CO2 for 5 min. We extracted venom by electrical
stimulation using repeated shocks (15V, 7.6mA, AC) of 2.5-sec duration at 10-sec
intervals. We applied electrical stimulation to the bases of the forcipules and the posterior
cephalic region. To do so, we slid the tip of a tapered metal scoopula between the tips of
the centipede’s forcipules while grasping the centipede at the base of the head with a pair
32
of forceps, with current traveling between the forceps and scoopula (Fig. 2B). Positioning
the scoopula in this way allowed us to isolate venom secretion from possible
contamination by saliva and regurgitated digestive fluids. We increased conductivity by
placing a few drops of saline solution on the bottom of the scoopula and on the forceps.
Each productive shock elicited a single “pulse” of venom, defined as the total volume of
venom expelled simultaneously from both forcipules. Not all shocks were productive;
thus, any shock that resulted in venom expulsion (of any amount) was defined as a
productive shock. We collected venom using graduated 5-µL Drummond PCR
Micropipets (0.246 mm radius; PGC Scientifics, Garner, NC, USA). We measured
venom either by viewing the micropipet under a Carson Linen-Test Magnifier (Carson
Optical Inc., Hauppauge, NY, USA; used for volumes up to 7.3 µL), or by taking a
digital photograph of the venom-filled micropipet using the dissecting scope at high
power immediately after milking and then measuring the venom column in the
photograph using Adobe Illustrator (Adobe Systems Inc., San Jose, CA, USA; used for
volumes up to 3.3 µL). Volume of venom (V) was calculated from the length of venom
column (L) using the formula V = (L) × (0.2462) × (3.14159). To establish accuracy and
precision of these measuring techniques, we pipetted (0.1–2.5 µL Eppendorf Research;
0.5–10.0 µL Eppendorf Research, Eppendorf, NY, USA) five replicates of known
volumes at 0.5-µL increments (0.1 µL and 0.5–7.5 µL [Carson Linen-Test Magnifier] or
0.5–3.5 µL [digital photographs]) of a solution of bovine serum albumin (BSA; SigmaAldrich, St. Louis, MO, USA; same protein concentration as the average for S.
polymorpha venom, 165 µg/µL) in phosphate-buffered saline (PBS; Mediatech,
Manassas, VA, USA) directly into micropipets. For the Carson Linen-Test Magnifier
33
method, percent error for individual measurements averaged ≤3.4% (1 S.E. ≤3.8%) at
each volume increment, with coefficients of variation (CVs) ≤3.1% with the exception of
that for the smallest volume (0.1 µL, CV = 8.6%). For the measuring method employing
digital photographs, errors averaged ≤4.3% (1 S.E. ≤3.5%) at each volume increment,
with CVs ≤2.7% with the exception of that for 0.1 µL (CV = 8.1%). Further, there was no
systematic bias in measurements using either method.
For some subjects (26 S. polymorpha, six S. subspinipes), we used separate
micropipets to collect venom from each shock until <0.05 µL was elicited, after which
the same micropipet was used to collect all subsequent venom. Following each shock we
collected the venom expelled simultaneously from both forcipules in the same
micropipet; we did not measure venom yield from each forcipule separately. We
immediately placed most of these samples (17 S. polymorpha, six S. subspinipes) into
separate vials of 500 µL chilled PBS (Mediatech, Inc., Herndon, VA, USA), and froze
them at -20 C to later determine protein concentrations (nine S. polymorpha samples were
frozen for future studies). For the remaining 151 S. polymorpha, we collected the full
venom amount from each individual using a single micropipet, and placed these in
separate vials of 50 µL Nanopure water before freezing. We terminated venom extraction
after the third unproductive or "dry" shock. Following Dass and Jangi (1978), centipedes
were not fed 2 weeks prior to venom extraction to ensure replete venom glands.
Of the 193 S. polymorpha tested, 12 (6%; range of body lengths: 3.7–11.5 cm)
provided no venom at all and four (2%; range of body lengths: 3.6–7.7 cm) gave total
volumes too small to accurately measure using our methods. Of the 12 centipedes
yielding no venom, two had both tarsungula obviously blunted (presumably from injury),
34
two had one tarsungulum blunted, and the remaining eight animals had no obvious
damage to the forcipules. These 16 individuals were excluded from further analyses, thus
providing a final sample size for S. polymorpha of n = 177. All six S. subspinipes yielded
measurable volumes of venom. All centipedes included in analyses yielded venom from
both forcipules.
Protein Quantification
We determined venom protein concentrations using the Coomassie blue dyebinding method (Bradford, 1976) with BSA in either PBS or Nanopure water (depending
on which the venom sample was combined with) as the standard. Coomassie Protein
Assay Reagent and BSA were purchased from Thermo Fisher Scientific (Rockford, IL,
USA). All assays using the manufacturer's 1–25 µg/mL protocol resulted in high
coefficients of determination (r2 > 0.96), indicating reliability. To further establish
accuracy and precision of the protein assay, we assayed three replicates of known protein
concentrations (4, 8, 12, 16, 20, and 25 µg/mL BSA in PBS). Percent error for individual
measurements averaged ≤8.0% (1 S.E. ≤2.1%) at each concentration, with CVs ≤3.3%.
There was no systematic bias.
Statistical Analyses
We relied on parametric tests (Field, 2005; Mertler and Vannatta, 2004) when
assumptions were met, including Pearson correlation (r), linear regression, Chow test
(Chow, 1960; Gujarati, 1970), t-tests, and analyses of variance (ANOVAs) and
covariance (ANCOVAs) followed by Tukey’s Least Significant Difference (LSD) for
35
multiple comparisons. To better meet assumptions, we transformed some measures,
unless otherwise indicated, using natural log (ln; number of individually measurable
pulses) or square root (total venom volume, mean volume of venom per pulse, and
percent of total venom volume). For multiple regression, we tested the absence of
multicollinearity by examining tolerance values (>0.1) and variance inflation factor
scores (<10; Mertler and Vannatta, 2004). For ANCOVAs, we always tested the
assumption of homogeneity of regression slopes by including an interaction term, and
then removed the term from the final model if non-significant (Mertler and Vannatta,
2004).(Field, 2005) For models that failed the within-subjects assumption of sphericity,
we applied Greenhouse-Geisser adjustments to the degrees of freedom. We also
employed non-parametric Spearman correlation (rs; Field, 2005) when parametric
assumptions were not met. Individual models are specified in the Results section.
We further computed effect sizes, which are independent of sample size (in
contrast to statistical significance) and more readily compared among different data sets
and different studies. For pairwise comparisons (t-tests), we relied on Cohen’s d using
pooled standard deviation (Hojat and Xu, 2004), for which values of ~0.2, ~0.5, and ≥0.8
are generally considered small, moderate, and large, respectively (Cohen, 1988). For
ANOVA and ANCOVA models, we computed eta-squared (η2) and partial η2, with
values of ~0.01, ~0.06, and ≥0.14 loosely regarded as small, moderate, and large,
respectively (Cohen, 1988). Because partial η2 values for a single model never summed
to >1.0, no adjustments were applied to these values. We expressed effect sizes for
bivariate correlations (r, rs) as coefficients of determination (r2), with values of ~0.01,
~0.09, and ≥0.25 deemed small, moderate, and large, respectively (Cohen, 1988). For
36
multiple regression, we report adjusted R2 (R2adj) for the full models, and semipartial
correlations (sr2) for individual predictors (Cohen et al., 2003; Mertler and Vannatta,
2004). With the exception of Cohen’s d, these effect size estimators roughly indicate the
approximate proportion of variance explained.
Statistical tests were conducted using SPSS 20.0 for Windows (Statistical
Package for the Social Sciences, Inc., Chicago, 2011), with α = 0.05. Following
Nakagawa (2004), we chose not to adjust α for multiple tests. The mean difference and
associated 95% CI limits given for analyses conducted with transformed data are backtransformed values. Unless indicated otherwise, measures of central tendency presented
are mean ± 1 S.E. For ANCOVAs utilizing untransformed data, we present adjusted
means ± 1 S.E. For ANCOVAs utilizing transformed data, we present back-transformed
adjusted means with 95% confidence limits, rather than standard error, as backtransformed error values would be misleading (Sokal and Rohlf, 1995).
Results
Total Venom Volume Yield
Body Length and Volume Yield
For S. polymorpha (n = 167, excluding previously milked animals), linear
regression of the form V = aL + b (where V = untransformed volume of venom extracted
in µL, a = slope, L = body length in cm, b = y-intercept) fit the data well (r2 = 0.49, F1,165
= 160.80, p < 0.001; Fig. 3A), indicating that 49% of the variance in venom volume was
explained by body length. The unstandardized coefficient (slope) for body length was
0.36 (95% CI = 0.30–0.42, t165 = 12.68, p < 0.001). The regression equation, V = 0.36(L)
37
- 2.08, revealed that for every 1-cm increase in body length (within the size range of
specimens studied), an additional 0.36 (95% CI = 0.30–0.42) µL venom could be
extracted. Linear regression using pooled data from both species was also significant (r2
= 0.68, F1,171 = 368.03, p < 0.001; Fig. 3A). Curve-fitting models (exponential: r2 = 0.61;
power: r2 = 0.63) and log-log transformation (r2 = 0.63) did not improve the model fit.
The unstandardized coefficient (slope) for body length was 0.52 (95% CI = 0.46–0.57,
t171 = 19.18, p < 0.001). The regression equation, V = 0.52(L) - 3.42, revealed that for
every 1-cm increase in body length (within the size range of specimens studied), 0.52
(95% CI = 0.46–0.57) µL more venom could be extracted. A Chow test indicated that the
regression models for S. polymorpha and pooled data were significantly different (F2,336 =
7.98, p < 0.001, partial η2 = 0.05), with a significant difference between both slopes (t336
= 3.88, p < 0.001; mean difference and 95% CI = 0.16 [0.08–0.24]) and y-intercepts (t336
= -3.68, p < 0.001; mean difference and 95% CI = -1.34 [-2.06– -0.63]).
38
39
40
Species and Volume Yield
Scolopendra polymorpha (n = 167, excluding previously milked animals) yielded
significantly less venom, in terms of total volume, than S. subspinipes (1.1 ± 0.05 vs. 5.4
± 0.5 µL, range 0.05–3.3 and 4.1–7.3 µL, respectively; t171 = 10.08, p < 0.001, Cohen's d
= 4.21; mean difference and 95% CI = 1.7 [1.1–2.5]). The ANCOVA analysis revealed
that body length significantly influenced venom volume (F1,170 = 190.67, p < 0.001,
partial η2 = 0.53), and that a significant difference existed between species (F1,170 = 4.47,
p = 0.036, partial η2 = 0.03), with a smaller length-specific yield from S. polymorpha
than S. subspinipes (adjusted means, 1.1 [1.0–1.1] vs. 1.7 [1.1–2.3] µL, respectively;
mean difference and 95% CI = 0.1 [0.0003–0.2]), although the effect size for the latter
was small. However, the results of this ANCOVA (and all other ANCOVAs comparing
species utilizing body length as a covariate) should be interpreted with caution, as there
was no body length overlap between species and adjusted means were evaluated at a
value of the covariate (9.0 cm) that is not represented in the S. subspinipes data set.
Additional Variables Influencing Volume Yield in
S. Polymorpha
A standard multiple regression analysis was conducted to evaluate how well body
length, relative forcipule length, relative forcipule width, relative mass, geographic origin
of centipede, sex, time in captivity, and milking history (previously milked or unmilked)
predicted venom volume yield (untransformed) in S. polymorpha (n = 177, including
previously milked animals). We used unstandardized residual scores (Mirtschin et al.,
2002; Schulte-Hostedde et al., 2005) from the general linear regression of forcipule width
versus body length as an index of relative forcipule width, and forcipule length versus
41
body length for relative forcipule length. Furthermore, we used unstandardized residual
scores from the linear regression of ln(body mass) versus ln(body length) as an index of
relative body mass. A negative residual score for the regression of forcipule width versus
body length, for example, indicates a centipede with a forcipule width smaller than
expected from its body length, whereas a positive residual score indicates a centipede
with a forcipule width larger than expected from its body length.
The correlations of the variables included in the multiple regression model are
shown in Table 1. Multicollinearity was not a problem. Regression results indicated that
the overall model significantly predicted volume yield (R2adj = 0.64, F8,168 = 39.77, p <
0.001), and accounted for 64% of the variance in volume. Five of the eight variables
contributed significantly to the model (Table 2). Venom volume was primarily predicted
by body length (sr2 = 0.506), and to a lesser extent by relative forcipule length (sr2 =
0.100), and relative body mass (sr2 = 0.016). Venom volume was positively associated
with each of these independent variables (IVs), with b indicating the increase in volume
(µL) for every 1-unit change in the IV. Geographic origin was also a significant predictor
(sr2 = 0.019), with yield for S. polymorpha higher from southern California than from
Arizona (adjusted means, 1.4 ± 0.1 vs. 1.0 ± 0.04 µL, respectively; t168 = 3.01, p = 0.003;
mean difference and 95% CI = 0.3 [0.1–0.5]) with all other variables held constant.
Milking history was also a significant predictor (sr2 = 0.014), with yield from unmilked
centipedes (n = 167) higher than from previously milked animals (n = 10; adjusted
means, 1.1 ± 0.03 vs. 0.7 ± 0.1 µL, respectively; t168 = 2.66, p = 0.009; mean difference
and 95% CI = 0.4 [0.1–0.7]). Of the significant predictors, body length had the greatest
relative influence on volume (β = 0.786), with yield increasing 0.39 (95% CI = 0.34–
42
0.44) µL for every 1-cm increase in body length. With all other variables held constant,
there was no difference between males and females in venom volume (adjusted means,
1.1 ± 0.04 vs. 1.1 ± 0.05, respectively; t168 = 1.09, p = 0.28; mean difference and 95% CI
= 0.1 [-0.1–0.2]). Time in captivity and relative forcipule width were also non-significant.
43
44
45
Body Length and Number of Productive Shocks
Knowing that longer centipedes yielded greater volumes of venom, we sought to
determine whether this was due to longer centipedes yielding venom over the course of a
greater number of shocks. For S. polymorpha (n = 167, excluding previously milked
animals, rs2 = 0.06, p = 0.002) and pooled data (n = 173, rs2 = 0.04, p = 0.007), the
correlation between number of productive shocks (shocks yielding any quantity of
venom) and body length, while significant and positive, was weak. The ANCOVA results
indicated that body length significantly influenced number of productive shocks (F1,170 =
7.95, p = 0.005, partial η2 = 0.05), but again the effect was small. Once body length was
controlled for, S. polymorpha yielded significantly more productive shocks than S.
subspinipes (F1,170 = 6.55, p = 0.011, partial η2 = 0.04; adjusted means 11.2 ± 0.5 vs. 1.4
± 3.7 µL, respectively; mean difference and 95% CI = 9.7 [2.2–17.2]), though again the
effect size was small.
Pulse-Related Venom Volume Yield
The venom extracted was transparent and colorless. However, the venom
sometimes became more viscous and sticky at the end of an extraction sequence for
pulses having very minute yields that were not individually measured but still included in
total venom yield. Otherwise, we saw no visible differences in venom associated with
body size or other variables, including successive pulses.
46
Body Length and Number of Measurable Venom
Pulses
Similar to total number of productive shocks, the number of measurable venom
pulses (individual pulses >0.05 µL) increased with body length of S. polymorpha (n = 16,
excluding previously milked animals; rs2 = 0.71, p < 0.001; Fig. 4A). When data for both
species were pooled, number of measurable pulses clearly increased with body length (n
= 22, rs2 = 0.84, p < 0.001; Fig. 4A). In the ANCOVA model, body length significantly
influenced number of venom pulses (F1,19 = 39.80, p < 0.001, partial η2 = 0.68). Once
body length was controlled for, S. polymorpha yielded significantly more measurable
pulses than S. subspinipes (F1,19 = 6.70, p = 0.018, partial η2 = 0.26; adjusted means 3.7
[2.9–4.6] vs. 1.5 [0.9–2.6], respectively; mean difference and 95% CI = 2.4 [1.2–4.9]).
47
48
49
For the n = 16 S. polymorpha and n = 6 S. subspinipes for which we measured
individual pulses (>0.05 µL) along with the sum of all individually unmeasurable pulses
(<0.05 µL), the total additional volume of venom yielded in these latter small pulses
contributed a small percentage of the total yield (mean ± SE: 15.1 ± 5.7% for S.
polymorpha; 1.5 ± 1.2% for S. subspinipes).
Body Length and Venom Volume Per Pulse
Analyses revealed that the mean volume of venom per measurable pulse increased
significantly with body length for S. polymorpha (n = 16, excluding previously milked
animals; r2 = 0.33, p = 0.020) and when data for both species were pooled (n = 22, rs2 =
0.56, p < 0.001; Fig. 4B). The ANCOVA results demonstrated that body length did not
significantly influence volume of venom per pulse (F1,19 = 4.16, p = 0.056, partial η2 =
0.18), although it approached significance and the effect size was large. Once body length
was controlled for, mean volume per measurable pulse was smaller for S. polymorpha
than S. subspinipes (F1,19 = 5.38, p = 0.032, partial η2 = 0.22; adjusted means 0.3 [0.2–
0.4] vs. 0.6 [0.4–0.9] µL, respectively; mean difference and 95% CI = 0.1 [0.0006–0.2]).
Body Length and Percent of Total Volume Yielded
in First Pulse
To determine whether longer centipedes gave larger proportions of total venom
yield per pulse, we examined the correlation between body length and volume of the first
pulse as a percent of total volume yield (Fig. 4C). We chose to use percent of total
volume yielded in the first pulse rather than the mean percent of total volume per pulse
because the latter measure is calculated based on number of measurable pulses, which
50
itself increases with increasing body length. Using percent of total volume yielded in the
first pulse thus allows for a more equitable comparison of relative yield across body
sizes. Correlation analysis, applied to untransformed data, indicated that percent of total
volume yielded in the first pulse was independent of body length for S. polymorpha (n =
16, excluding previously milked animals; r = -0.30, r2 = 0.09, p = 0.27) and for both
species pooled (n = 22, r = -0.32, r2 = 0.10, p = 0.15). Thus, at least when considering the
first pulse, longer centipedes did not give a greater percent of total yield per pulse despite
the fact that longer centipedes gave greater absolute mean volume yields per pulse.
Pulse Number and Percent of Total Volume of
Venom Extracted
Because some S. polymorpha had been previously milked, we investigated how
the proportion of total venom volume yield varied across pulses by conducting a separate
one-way ANOVA for each species, with pulse number treated as a within-subjects factor,
and including the previously milked S. polymorpha. Few S. polymorpha (30% of n = 26)
yielded >3 measurable pulses of venom during extraction; thus, for S. polymorpha we
limited our analysis to the first three pulses in order to maintain a reasonable sample size
(n = 14). There was a significant difference in the percent of total volume of venom
yielded among pulses (F1.2, 15.5 = 7.72, p = 0.011, partial η2 = 0.37; Fig. 5A), with an LSD
post hoc test revealing that percent of total volume yielded was significantly lower for the
third pulse than for either the first (p = 0.040) or second (p < 0.001) pulses. The ANOVA
for S. subspinipes (n = 6), using untransformed data and examining the first four pulses,
demonstrated a significant difference among pulses (F3,15 = 3.67, p = 0.037, partial η2 =
0.42; Fig. 5B), with an LSD post hoc test indicating that percent of total volume yielded
51
was significantly lower for the third pulse than for either the first (p = 0.027) or second (p
= 0.035) pulses. For S. polymorpha and S. subspinipes, the first two pulses represented
69% and 63% of extractable venom, respectively. A between-subjects comparison of data
(i.e., assuming independence of pulses) indicated an obvious linear or possibly quadratic
trend for both S. polymorpha (sample size varied with pulse, previously milked animals
excluded; Fig. 5C) and S. subspinipes (sample size varied with pulse; Fig. 5D), but
because some but not all data were related (from the same individual), we were unable to
test the nature of the decline statistically.
52
53
54
Venom Protein Concentration
Body Length and Venom Protein Concentration
The correlation between body length and protein concentration (Fig. 3B) was not
significant for S. polymorpha (n = 158, excluding previously milked animals; r2 = 0.02, p
= 0.11) or for both species combined (n = 164, rs2 = 0.007, p = 0.29).
Species and Venom Protein Concentration
Protein concentration of venom from S. polymorpha (n = 158, excluding
previously milked animals) exceeded that of S. subspinipes (165 ± 3 vs. 113 ± 11 µg/µL,
range 83–292 and 75–140 µg/µL, respectively; t162 = 3.42, p = 0.001, Cohen's d = 1.43;
mean difference and 95% CI = 52 [22–82]).
Additional Variables Influencing Venom Protein
Concentration in S. Polymorpha
We conducted standard multiple regression analysis to evaluate how well the IVs
of body length, relative body mass, geographic origin, sex, time in captivity, number of
productive shocks, and milking history (previously milked or unmilked) predicted venom
protein concentration (untransformed) in S. polymorpha (n = 168, including previously
milked animals). Correlations of the variables are shown in Table 3. Multicollinearity
was not a problem. The overall model significantly predicted protein concentration (R2adj
= 0.29, F7,160 = 10.68, p < 0.001), and accounted for 29% of the variance in protein
concentration. Three of the seven variables contributed significantly to the model (Table
4). Venom protein concentration was primarily predicted by geographic origin (sr2 =
0.129), with protein concentration higher for S. polymorpha from southern California (n
55
= 40) than from Arizona (n = 128; adjusted means, 188 ± 6 vs. 152 ± 3 µg/µL,
respectively; t160 = 5.50, p < 0.001; mean difference and 95% CI = 36 [23–49]), with all
other variables held constant. Milking history was also a significant predictor (sr2 =
0.099), with protein concentration higher from unmilked centipedes (n = 158) than from
previously milked animals (n = 10; adjusted means, 164 ± 3 vs. 105 ± 12 µg/µL
respectively; t160 = 4.81, p < 0.001; mean difference an 95% CI = 59 [35–84]). Body
length was also a significant predictor (sr2 = 0.032), with protein concentration increasing
6 (95% CI = 2–10) µg/µL for every 1-cm increase in body length. With all other
variables held constant, there was no difference in venom protein concentration between
males and females (adjusted means, 160 ± 3 vs. 162 ± 4, respectively; t160 = 0.42, p =
0.67; mean difference and 95% CI = 2 [-9–13]). Number of productive shocks and
relative body mass were also non-significant.
56
57
58
Pulse-Related Venom Protein Concentration
Because some S. polymorpha had been previously milked, we investigated how
venom protein concentration varied across pulses by conducting a separate one-way
ANOVA for each species, with pulse number treated as a within-subjects factor, and
including the previously milked S. polymorpha. Few S. polymorpha (35% of n = 17)
yielded >3 measurable pulses of venom during extraction, thus for S. polymorpha we
limited our analysis to the first three pulses in order to maintain a reasonable sample size
(n = 10). The ANOVA for S. polymorpha revealed no significant difference in protein
concentration among pulses (F2,18 = 0.48, p = 0.63, partial η2 = 0.05; Fig. 6A). Likewise,
ANOVA examining the first four pulses from S. subspinipes (n = 6) demonstrated no
significant difference in protein concentration among pulses (F1.19,5.95 = 0.48, p = 0.55,
partial η2 = 0.09; Fig. 6B). A between-subjects comparison of data (i.e., assuming
independence of pulses) indicated little change in protein concentration across pulses for
either S. polymorpha (sample size varied with pulse, excluding previously milked
animals; Fig. 6C) or S. subspinipes (sample size varied with pulse; Fig. 6D), but because
some but not all data were related (from the same individual), we were unable to test for a
difference statistically.
59
60
61
Comparison of Variation
Of the two aspects of venom yield that we studied for S. polymorpha (excluding
previously milked animals), total venom volume showed considerably more variation
(CV = 60%, n = 167) than protein concentration (CV = 22%, n = 158).
Discussion
This study of two scolopendromorph species represents the most comprehensive
study to date of venom yield in any centipede. In this section, we begin by discussing
venom collection methods, venom attributes, and inferred venom supply. We then
consider the factors that influence venom yield, venom depletion, and protein
concentration. Lastly, we infer some functional aspects of the centipede's venom supply.
Venom Collection, Venom Attributes, and Venom Supply
We used electrical stimulation to obtain venom from the centipedes S.
polymorpha and S. subspinipes. Similar to some spider (e.g., Celerier et al., 1993;
Schanbacher et al., 1973) and centipede (e.g., Dugon and Arthur, 2012b) milking
methods, we found the use of a saline solution on the forceps and the scoopula to be vital
in achieving consistent conduction of electricity to the centipedes’ forcipules. Our
experience indicates that CO2 anesthetization and electrical venom extraction were
neither fatal nor overtly detrimental to the centipedes.
Several specimens of S. polymorpha were omitted from this study because they
yielded no venom despite strong adduction contractions of the forcipules. In some cases,
this appeared to be related to injury to the centipedes’ forcipules (as detected by a blunted
shape of the tarsungula), a phenomenon that is only minimally documented (e.g., Barber,
62
2011; Frund, 1992) but not uncommon. Forcipule injury has the potential to affect
forcipule size and venom yield; depending on the extent of the injury, the forcipule may
be smaller and the venom system rendered non-functional upon regeneration (Verhoeff,
1940, cited in Lewis, 1981). In our investigation, forcipules with blunted tarsungula
frequently did not yield venom. Inspection of unproductive, blunted forcipules typically
revealed the lack of a meatus (Dugon and Arthur, 2012a; Dugon et al., 2012) or venom
pore (Menez et al., 1990), the opening of the venom duct located subterminally on the
tarsungulum. Presumably, occlusion of the meatus or the venom duct occurred during
regeneration. In the rare cases (n = 2) in which venom was collected from a blunted
tarsungulum the blunting was minor and distal to the meatus. In these cases, as we did not
measure venom yield from the uninjured forcipule and the blunted forcipule separately,
we are unable to comment on the influence, if any, of forcipule injury on venom yield.
However, given the apparently minor extent of the injuries we assume yield was
minimally impacted. Several centipedes with apparently uninjured forcipules did not
yield venom. We do not know why such specimens yielded no venom, although temporal
proximity to molting may be involved, as has been demonstrated in spiders (Herzig,
2010; Rocha-e-Silva et al., 2009). The possibility also exists that these were naturally
venomoid specimens, presumably arising from a pathogenetic condition. One of us has
observed a naturally venomoid adult rattlesnake (WKH, unpublished observation).
Our observations that the extracted venom from both species was clear and
colorless agrees with the findings of several earlier authors who have noted the clear,
colorless nature of centipede venom in general (Minton, 1974) and of S. polymorpha
venom specifically (Baerg, 1924). However, Martin (1971) described S. subspinipes
63
venom as opalescent rather than clear. Jangi (1966) described the venom of S. amazonica
as clear, but also noted the color as yellowish. Lewis (1981), citing Duboscq (1898),
described Scolopendra venom as a clear, yellowish, homogenous liquid.
Our results are likely an approximation of the quantity of venom present in the
venom glands. Venom milking may not result in complete emptying of the venom glands,
as has been noted in honeybees (Owen, 1978), spiders (Malli et al., 2000), ants (Blum
and Callahan, 1960; Haight, 2002), and snakes (Kochva et al., 1982; McCleary and
Heard, 2010). Furthermore, venom yields from electrical milking should not be
interpreted as indicative of the amount of venom animals expend in natural stings and
bites. This is illustrated by the spider Cupiennius salei, in which electrical milking
yielded 7–15 µL per spider (Kuhn-Nentwig et al., 2004), but ELISA results demonstrated
that only 0.5–2.5 µL was injected when attacking large crickets (Malli et al., 1998).
Likewise, gland massage and electrical milking of snakes produce a greater yield than a
natural bite or spontaneous ejection (di Tada et al., 1976; Hayes et al., 1992; Klauber,
1997b; Tare et al., 1986). Despite these obvious limitations, this study provides an
estimation of venom capacity of these two centipede species, which can help us better
understand venom deployment under normal contexts.
Although electrical venom extraction has been routinely used for centipedes (e.g.,
Dugon and Arthur, 2012b; Gonzalez-Morales et al., 2009; Liu et al., 2012b; Malta et al.,
2008; Rates et al., 2007; Yang et al., 2012), venom yields are seldom reported, and even
when they are, differences in how yield is measured can make comparisons challenging.
Average venom yield for S. subspinipes mutilans was approximately 1.7 µL based on
total volume of venom collected divided by number of animals used (Peng et al., 2010).
64
This yield is only about 30% of the mean yield of S. subspinipes (5.4 µL) and 150% of
the mean yield of S. polymorpha (1.1 µL) from our study, but Peng et al. (2010) did not
report the size range of their specimens. Comparison of our volumetric yields with the
mean wet mass yields reported by Malta et al. (2008) for Cryptops iheringi (17.5 µg; 60–
100 mm body length), Otostigmus pradoi (25 µg; 40–80 mm), and S. viridicornis (100
mg; 160–200 mm) are difficult without venom density measurements. Likewise, since we
did not measure venom dry mass, we cannot directly compare our data to Yang et al.’s
(2012) yield of 0.2 mg venom per adult S. s. mutilans or Rates et al.’s (2007) yield of
approximately 0.05 mg per S. viridicornis nigra (10 cm).
Venom yields can also be inferred from published venom gland dimensions, with
gland volume calculated as a prolate spheroid (PS; cf. Menez et al., 1990) or a cylinder
(C; Antoniazzi et al., 2009). We estimated total gland volume (both glands combined) for
the following three species: S. polymorpha, PS = 4.2 µL, C = 6.3 µL (glands 4 mm length
x 1 mm diameter, n = 1 specimen 14.5 cm long including caudal legs; Baerg, 1924); S.
viridicornis, PS = 4.3–6.4 µL, C = 6.4–9.7 µL (glands 8–12 mm x 716 ± 50 µm, n = 3
specimens 16-20 cm long; Antoniazzi et al., 2009) or PS = 7.3–7.9 µL, C = 11.0–11.8 µL
(glands 7–7.5 mm x 1 mm, no specimen details; Bücherl, 1971c); and Ethmostigmus
rubripes, PS = 5.2–11.8 µL, C = 7.9–17.7 µL (glands 5 mm x 1–1.5 mm, no specimen
details; Menez et al., 1990). Although considerable venom storage is afforded by the
secretory body ("extracellular space", Rosenberg and Hilken, 2006) of each secretory unit
(Undheim and King, 2011), and large secretory vacuoles occupy much of the cytoplasm
volume of the secretory cells (Antoniazzi et al., 2009), the gland obviously is not merely
a hollow venom storage reservoir. Thus, similar to the relationship between gland volume
65
and yield observed in the spider C. salei (Malli et al., 1993), we expected gland volume
estimations to be over-estimates of true venom capacity. Indeed, gland volume estimates
for S. polymorpha were nearly 400–600% greater than our mean venom yield, and about
125–190% greater than our maximum venom yield. Baerg’s (1924) S. polymorpha
specimen used for computing venom gland size was larger (14.5 cm) than our mean S.
polymorpha length (8.9 cm), but similar to our maximum length (13.8), although our
length measurement did not include the caudal legs. Estimates treating glands as prolate
spheroids for Antoniazzi et al.'s (2009) S. viridicornis were about 80–120% of our mean
yield for S. subspinipes, though their animals (16–20 cm) were slightly larger than our
largest S. subspinipes (15.8 cm). These comparisons imply that electrical venom
extraction can remove a high percentage of available venom from the glands.
Factors Influencing Venom Yield
Body Length
Centipede body length was the single most important factor determining venom
volume yield, generating approximately 2-fold (S. polymorpha) to 4.8-fold (pooled data)
differences in the simple linear regression equation. The increase in venom yield with
body length presumably corresponds to a concomitant increase in venom gland size
(Antoniazzi et al., 2009).
A similar positive relationship between size and venom yield obtained by
electrical stimulation or other forced (non-voluntary) extraction methods exists in other
taxa. In spiders, for example, venom volume corresponds linearly to prosoma length (C.
salei; Vapenik and Nentwig, 2000); wet mass linearly to body mass (Vitalius dubius;
Rocha-e-Silva et al., 2009); dry mass by fourth-order power function to prosoma length
66
(Phoneutria nigriventer; Herzig et al., 2004); and volume to size in general (Loxosceles
reclusa, Morgan, 1969; Pterinochilus sp., Perret, 1973). In scorpions, venom yield is
positively correlated with scorpion size in several species (Whittemore et al., 1963), and
exponentially related to body length in Hadrurus arizonensis (Fox et al., 2009). In the
reduviids Rhynocoris marginatus and Catamiarus brevipennis, more massive bugs yield
larger wet masses of venom (although sex may be a confounding factor, as females are
larger than males; Sahayaraj et al., 2006). In venomous snakes, the pattern of increase in
venom yield with body size is probably universal (Mirtschin et al., 2002). Studies of
snakes generally reveal an exponential relationship (Huang and Mackessy, 2004;
Mackessy, 1988; Mackessy et al., 2003, 2006; McCue, 2006; Mirtschin et al., 2002), but
several have reported a linear relationship (Abdel-Aal and Abdel-Baset, 2010; de Roodt
et al., 1998; Kochva et al., 1982; McCleary and Heard, 2010).
Regression analyses showed the slopes of the lines relating body length to venom
volume yield for S. polymorpha and pooled data were 0.36 (95% CI = 0.30–0.42) and
0.52 (95% CI = 0.46–0.57), respectively. Thus, for every unit increase in centipede size
there was between approximately 1/3–1/2 unit increase in venom yield. These slopes
were small compared to those for electrical extractions of the spiders V. dubius (2.9, body
mass vs. wet mass; Rocha-e-Silva et al., 2009) and C. salei (1.8, prosoma length vs.
volume; Vapenik and Nentwig, 2000). The length-yield slopes were also lower than that
of electrical venom extraction in the snake Agkistrodon piscivorus conanti (1.0, snoutvent length vs. volume; McCleary and Heard, 2010), but greater than those of manual
venom extraction in the snakes Bothrops alternatus and Crotalus durissus terrificus (0.29
and 0.06, respectively, body mass vs. dry mass; de Roodt et al., 1998). Mirtschin et al.
67
(2002) posited that natural selection may adjust the relationship between body length and
venom production in snakes based on the relationship between body size and factors that
determine probable venom use, including prey sizes, prey types, and feeding frequencies.
Perhaps the relatively small increase in venom yield with increasing body size in these
centipedes indicates relatively small differences in prey sizes, prey types, or feeding
frequencies between young and old centipedes compared to those experienced during
ontogeny in spiders and some snakes.
Numerous proximate anatomical and physiological factors presumably influence
the relationship between ontogeny and venom yield in centipedes. One may be how the
volume of the venom glands increases during development, which depends on the shape
of the glands (often cylindrical or ovoid in scolopendrids; Antoniazzi et al., 2009;
Gopalakrishnakone, 1992; Jangi, 1984; Menez et al., 1990), and the rate at which the
glands expand in each dimension. Limiting the possibilities, in scolopendrids, are the
growth rates in three-dimensional space of the forcipules housing the glands. Although a
detailed study relating venom yield to measured gland volumes and forcipule dimensions
would be necessary to confirm the likely complex interrelationships, we can still make
inferences based on our data. Our multiple regression model predicting volume yield in S.
polymorpha showed that relative forcipule width was not a significant predictor, implying
that all else being equal, a wider forcipule may not be indicative of a larger diameter
venom gland. In contrast, relative forcipule length was a significant predictor of yield and
had the second highest relative influence on yield (β = 0.501) following body length.
Thus, a longer forcipule may be indicative of a longer venom gland, with forcipule length
exerting a greater effect on gland size than forcipule width. This relationship, combined
68
with the fact that forcipule length increases more rapidly with an increase in body length
than forcipule width (data not shown), may explain why the relationship between yield
and body length is linear rather than exponential.
Of the two aspects of venom yield that we studied, volume showed considerably
more variation (CV = 60%) than protein concentration (CV = 22%). Although the
relationship between centipede body length and venom volume yield was strong (r2 =
0.68, pooled data), there was still considerable variation in volume extracted from a given
length of centipede. For example, for S. polymorpha between 9.9 and 10.1 cm in length
from Arizona (n = 6), volume yield varied between 1.4 and 2.4 µL, a nearly 2-fold
variation. This may result from a variation in venom gland size in similarly sized
centipedes, as was shown to be the case in same-sized worker fire ants (Solenopsis
invicta, Haight, 2002). Although every effort was made to ensure all centipedes were
healthy, fed on the same schedule prior to venom extraction, and had access to water, it is
possible that, as in other venomous animals, variations in health (Klauber, 1997b), age
(Malli et al., 1993), proximity to molting (Rocha-e-Silva et al., 2009), nutritional status
(Kuhn-Nentwig and Nentwig, 1997), or hydration status (Mirtschin et al., 2002) may
have contributed to venom yield variation.
Although it would be useful in relating ontogeny to venom yield to understand the
relationship between body length and developmental stage in S. polymorpha and S.
subspinipes, a lack of data prevents such a correlation. Perhaps none of the
morphological characters that change significantly with age (e.g., body length, mass,
number of antennal segments) show sufficiently abrupt changes to act as characters to
differentiate post-larval stadia, as was shown for S. morsitans (Lewis, 1968). The lack of
69
ontogenetic data is a consequence of the trend in the scolopendromorph developmental
biology literature to focus nearly exclusively on descriptive embryology and larval
development (e.g., Brunhuber, 1970; Heymons, 1901; Jangi, 1966; Lawrence, 1947;
Radl, 1992), neglecting, with the exception of work by Lewis (1968, 1970, 1972),
detailed examination of the characteristics of post-larval stadia (Lewis, 1972, 1981). In
general, after leaving the brood chamber, growth is thought to be continuous, gradual (cf.
Lewis, 1972), and chiefly associated with increase in size (Cloudsley-Thompson, 1968).
The exact number of molts passed through by young scolopendromorphs after leaving
their mother remains unknown, but is thought to be a considerable number (CloudsleyThompson, 1968). Scolopendra polymorpha sexually matures in an estimated two years
(Crawford, 1990) after exceeding 6.8 cm in length (Maldonado, 1998), has an estimated
lifespan of 5 years (Crawford, 1990), and attains a maximum length of 11.1 cm (Shelley,
2002) to 14.5 cm (Maldonado, 1998). Lewis (1970) reported that S. subspinipes may molt
up to 10 times over the course of 2.5 years in captivity. The large Scolopendra species,
including S. subspinipes, become adults only in the third or even fourth year of age, and
can live for more than 10 years (Bücherl, 1971c). The reported maximum length of S.
subspinipes varies between 10.1 and 25 cm (Bücherl, 1971c; Sandefer, 1998; Shelley,
2002; Turk, 1951) depending on subspecies recognition (Kronmuller, 2012). Thus, until
the gaps in our knowledge of scolopendromorph development are filled, relating venom
yield tightly to ontogeny will remain problematic.
70
Species
In terms of absolute venom volume, differences in yield between S. polymorpha
and S. subspinipes appear to be largely the result of a difference in size, although
ANCOVA indicated a greater length-specific venom yield from S. subspinipes (but effect
size was small). While the preliminary ANCOVA indicated absence of a factor-covariate
interaction (p = 0.99, partial η2 = 0.00001), indicating homogeneity of regression slopes
for the relationship between body length and volume yield for the two species, a separate
linear regression model did not fit the S. subspinipes data well, leaving unclear the exact
nature of the relationship between length and yield in this species. Results of the Chow
test demonstrating different slopes and y-intercepts for the regression models for S.
polymorpha data and for pooled data suggest the body length-volume yield relationship is
different between species. A larger data set for S. subspinipes is needed to determine
whether the coefficient linking body length and venom volume yield is different for these
two species.
Geographic Origin
Multiple regression revealed that, on average and with all else being equal, venom
yield from California S. polymorpha was about 40% greater than that from Arizona. An
intraspecific difference in venom yield based on population has been observed in other
taxa, including snakes (Klauber, 1997b; Mirtschin and Davis, 1992; Mirtschin et al.,
2002) and the spider Tegenaria agrestis (Binford, 2001). Previous authors have attributed
geographic variation in yield to a combination of genetic differences among populations
and the direct influence of environmental factors (Mirtschin et al., 2002). Because
centipedes in our study had been maintained in captivity for varying amounts of time, we
71
included duration of captivity in the multiple regression model, which proved to be a
non-significant predictor. Although the effects of proximate environmental conditions on
yield presumably diminish with time in captivity, the potential effects of environmental
conditions are of unknown duration. Thus, the relative contributions of genetics and
environment to this source of variation in yield remain unclear.
Relative Body Mass
According to the multiple regression model for S. polymorpha, relative body mass
had a small but significant influence on venom yield (β = 0.129). Relative body mass, or
body condition, has been used as an estimate of an animal’s nutritional state and
ultimately fitness (Jakob et al., 1996; Schulte-Hostedde et al., 2005). Little has been
written specifically about the effect of body condition on venom yield in any taxon,
although Fairly and Splatt (1929) attributed a three-fold difference in venom yields
between two populations of tigersnakes to a diminished body condition in one
population. Because relative mass is not expected to influence size of the venom glands,
we suspect that it influences venom yield via degree of filling of the gland, representing a
differential investment in venom production.
Milking History
The multiple regression model for S. polymorpha indicated that milking history
(whether the centipede had been previously milked or not) significantly influenced
venom yield (β = -0.137), with unmilked centipedes having approximately 57% higher
yields than previously milked animals, with all else being equal. Two possibilities could
72
explain why venom volume was lower for previously milked animals. First, venom
volume regeneration following emptying of the glands may require more than three
weeks (the time between the previous milking and the milking generating the data
analyzed here). Second, electrical milking may damage the venom glands or venom gland
musculature, as proposed by other investigators. Among spiders, Argiope bruennichi
failed to yield venom after a single milking (Friedel and Nentwig, 1989), and volume
yield declined with additional milkings of Agelenopsis aperta (Kristensen, 2005). Sissom
et al. (1990) suggested that scorpions can only be milked, on average, four times before
the muscles of the gland stop responding to electrical stimulation. In some cases,
electrical milking may even kill the animal (Sahayaraj et al., 2006). In contrast, repeated
electrical venom extractions did not reduce yield in the spider Coremiocnemis tropix
(Herzig, 2010), the scorpion H. arizonensis (Fox et al., 2009), or in snakes (Marsh and
Whaler, 1984; McCleary and Heard, 2010). Presumably, the potential for damage to an
animal or its tissues increases with larger voltage-current combinations, and varies
depending on where the shock is applied. Although electrical venom extraction did not
appear overtly detrimental to the centipedes, further study of the effects of repeated
milking on yield, preferably incorporating histological examination of the venom glands,
is necessary to determine the mechanism by which repeated electrical milking leads to
reduced venom volume yields in S. polymorpha.
Sex
Sex of S. polymorpha was not a significant predictor of venom volume or protein
concentration in our multiple regression models. Anecdotal wisdom from Trinidad claims
a sting from a female centipede (species unspecified) is worse than a sting from a male
73
(Cloudsley-Thompson, 1968). Based solely on a comparison of venom availability and
protein concentration, our data from S. polymorpha fail to support this assertion.
However, females may employ a greater volume of venom in a defensive sting, or
possess venom components with greater toxicity.
The biological significance of these factors influencing venom yield is difficult to
determine without additional study to elucidate what relationship, if any, might exist
between venom capacity and the amount of venom a centipede actually deploys in any
given sting. Anatomical and anecdotal evidence support a level of control over venom
expulsion in centipedes. Histological evidence of a mechanism involving a sphincter and
a nozzle-like non-return valve that would work together, under neuromuscular control, to
regulate venom discharge from each of the secretory cells has been reported (Antoniazzi
et al., 2009; Dass and Jangi, 1978; Menez et al., 1990; Rosenberg and Hilken, 2006).
Furthermore, Lewis (1981) claimed that scolopendrids could vary the amount of venom
they inject, citing Klingel (1960) who found that S. subspinipes that had not fed for
several days (and thus were assumed to have venom available) could sting humans
harmlessly. Reports of human envenomations note that stings can vary significantly in
pain intensity (Bush et al., 2001), with severity greater when stings came from centipedes
who were injured or protecting young (Lewis et al., 2010). Antoniazzi et al. (2009) have
speculated that regulation of venom discharge may contribute to the varied clinical
presentations following human envenomation by centipedes of the same species.
Evidence of dry defensive stings of targets by S. polymorpha (Cooper, unpublished data)
further supports the likelihood of flexible venom expulsion. Such context-dependent
expenditure of venom would conform to the venom-metering (Hayes, 2008; Hayes et al.,
74
2002) or venom-optimization (Wigger et al., 2002) hypotheses. Investigation into
flexibility of venom use must precede any conclusions about the biological significance
of factors such as species or population differences in venom yields.
Factors Associated With Venom Depletion (Successive Shocks)
The relationships between body length and number of productive shocks and
body length and number of measurable pulses (Fig. 4A) indicated that more shocks were
required to deplete the yield of longer centipedes. Furthermore, correlation analysis of
volume per pulse revealed that longer centipedes yielded greater volumes of venom per
pulse, although they did not give a greater proportion of total venom in the first pulse.
All three measures associated with venom depletion (number of productive
shocks, number of measurable pulses [individual pulses >0.05 µL], and mean volume of
venom per measurable pulse) were positively correlated with centipede body length.
Once body length was controlled for, S. polymorpha yielded more productive shocks,
more measurable pulses, but a smaller mean volume per measurable pulse than S.
subspinipes. For a given body length, S. polymorpha yielded a greater number of very
small pulses that we were unable to individually measure (individual pulses <0.05 µL)
but still counted as productive shocks. These very small pulses of venom occurred at the
end of an extraction sequence, and summed to a small amount of additional volume. The
significantly higher total venom yield from S. subspinipes (section 3.1.2) derived from its
larger size, with the concomitantly greater number of productive shocks, measurable
pulses, and greater mean volume per measurable pulse. For S. subspinipes, the positive
relationship between body length and these variables outweighed the additional volume
75
accruing to S. polymorpha from a greater length-specific number of productive shocks
and measurable pulses, many of the former being too small to individually measure.
Examining the effect of pulse number (treated as a within-subjects factor) on
percent of total volume of venom extracted for both species (Fig. 5A, 5B) revealed that
about two-thirds of the extractable venom (69%, S. polymorpha; 63%, S. subspinipes)
was expressed in the first two pulses. In the first pulse (with pulse treated as a betweensubjects factor; Fig. 5C, 5D), 48% and 36% of extractable venom was emitted from S.
polymorpha and S. subspinipes, respectively. For comparison, in the scorpion
Centruroides limpidus tecomanus, approximately 51% of the telson's venom content was
emitted during the first electrical stimulation (Whittemore et al., 1963). The apparent
decreasing trend in percent of total volume across pulses (Fig. 5C, 5D) suggests that our
venom milking methods were successful in extracting the majority of available venom.
Factors Influencing Protein Concentration
Because of the important roles of proteins in most venoms, measuring venom
protein concentration is common (e.g., Celerier et al., 1993; Chacon et al., 2012;
Sahayaraj et al., 2006). While not all venom proteins are toxins (Chavez-Olortegui et al.,
1997; Kuhn-Nentwig et al., 2011), venom toxicity may be correlated with protein content
(Boeve et al., 1995; Mirtschin et al., 2002; Oukkache et al., 2013; Perret, 1977b), but this
is not always the case (Malli et al., 1993). However, protein content of venom can vary
with extraction method (Brochetto-Braga et al., 2006; Mackessy, 2002; Oukkache et al.,
2013). We recognize that venom elicited by electrical milking cannot be assumed to be
identical in composition to that used during natural prey capture or defense (Kristensen,
2005), and that attempting to remove all venom from the glands may introduce increased
76
amounts of cellular material (McCleary and Heard, 2010). Even so, this method
represents an improvement over the use of forcipule or gland extracts.
The variation in venom protein concentration for S. polymorpha (3.5-fold) and S.
subspinipes (2-fold) was relatively large. Similar variation (>3.5-fold) was observed in
the venom of the tarantula Eurypelma californicum (Savel-Niemann and Roth, 1989).
Authors who have milked scolopendromorphs seldom report the venom protein
concentrations, precluding direct comparisons. The 126.6 mg/g protein concentration of a
venom extract from ground forcipules of S. morsitans (Mohamed et al., 1983) makes for
a poor comparison due to inclusion of forcipule tissues. Venom protein concentration was
greater for S. polymorpha than S. subspinipes, but given the possibility of a compositional
difference as well, speculation regarding the functional significance of the difference
would be unwise.
Geographic Origin
Multiple regression revealed that geographic origin had the greatest relative
influence on venom protein concentration (β = 0.391) in S. polymorpha, with southern
California animals having higher concentration than those from Arizona. In a similar
vein, population differences in percent solids were found in venoms of tigersnakes and
brownsnakes, although the authors did not speculate on the biological meaning of such
variation (Mirtschin et al., 2002). Further investigation is necessary to ascertain whether
the difference in protein concentration relates to compositional differences or has a
functional significance.
77
Milking History
Multiple regression also indicated that milking history had a relatively large
influence on venom protein concentration (β = -0.357), with unmilked animals having
56% higher protein concentration than previously milked animals. As with the influence
of milking history on volume yield, two possible explanations may be given for the
reduced protein content of venom from previously milked animals, including insufficient
time since last milking for complete regeneration, and potential damaging effects of
electrical extraction on the venom glands. By way of comparison, repeated electrical
milking of the scorpion H. arizonensis (five milkings at 3-week intervals) did not alter
venom protein concentration (Fox et al., 2009). Similarly, repeated electrical milking did
not affect venom protein concentration in the snake Bitis nasicornis (milking frequency
varying from 2 to 21 day intervals; Marsh and Glatston, 1974), or venom enzyme content
of B. gabonica (Marsh and Whaler, 1984). In contrast, concentration of venom solids
increased with decreasing milking (manual extraction) frequency in the snake Daboia
palaestinae (Kochva, 1960).
Body Length
Venom protein concentration, when subjected to bivariate correlation, was
independent of body length for S. polymorpha and pooled data from both species,
indicating no change in concentration with size within the range of lengths studied. In
contrast, multiple regression demonstrated that body length was a significant predictor of
protein concentration (although with the lowest relative influence on the DV [β = 0.201])
once the other variables in the model were controlled for. Protein concentration in several
78
snake species increased during ontogeny (Fiero et al., 1972; Mackessy et al., 2006; Meier
and Freyvogel, 1980), whereas other species showed the opposite trend (Furtado et al.,
1991). In the spider C. salei, there was a slight (but non-significant) increase in protein
concentration during ontogeny (Malli et al., 1993). Additional study is required to
understand why larger centipedes have a higher concentration of protein in their venom.
Venom color changes across successive defensive stings in Parabuthus
transvaalicus have been linked to variation in venom composition (Inceoglu et al., 2003;
Nisani and Hayes, 2011). In this scorpion, a small quantity of transparent "prevenom"
containing a high concentration of potassium (K+) salt and relatively low protein
concentration was followed by an increasingly milky and proteinaceous "venom" in
subsequent stings (Inceoglu et al., 2003). Along the same lines, the first approximately 1
µL of venom (representing about 10% of total yield) electrically extracted from the spider
C. salei was found to have reduced protein concentration relative to subsequently
extracted venom, although no color change was observed (Boeve et al., 1995). In S.
polymorpha and S. subspinipes, although venom became stickier at the end of an
extraction sequence, there was no variation in color during milking. Additionally, venom
protein concentration for both species did not vary across pulses (Fig. 6A, 6B). Our data
do not support the presence of a prevenom (distinguished on the basis of protein
concentration) in the scolopendromorphs we examined; however, even if heterogeneous
storage of venom (the suggested cause of the reduced protein content of prevenom or
initial venom; Boeve et al., 1995; Nisani and Hayes, 2011) did occur in the glands of
these centipedes, the segregation may exist on a scale too fine to be detected given the
size of the first venom pulse (33–46% of total volume) elicited with our milking method.
79
Furthermore, composition of the initial venom in a natural sting may differ from that
extracted using electrical stimulation (Zlotkin and Shulov, 1969).
Functional Aspects of Venom Supply
Determining venom yield makes it possible to estimate the number of theoretical
“doses” of venom a given animal possesses based on published lethal (LD) or effective
(ED) doses for a given prey species. Such calculations show that the killing power
represented by centipede venom yields can be great. As an example, the LD50 (24 hr) and
ED50 (1 hr) of S. polymorpha venom in Manduca sexta larvae (3rd instar) have been
determined as 0.083 and 0.017 µL/g, respectively (Quistad et al., 1992). For a 50-mg M.
sexta (the size used by Quistad et al., 1992), we calculate the LD50 as 0.0042 µL of
venom. An S. polymorpha 8.9 cm long (mean body length from the present study) is
estimated to have 1.1 µL of venom by simple regression. Thus, this centipede possesses
262 LD50 doses of venom, and the capacity to kill 131 larvae. For a 50-mg M. sexta, we
calculate the ED50 as 0.00085 µL of venom. Thus, an 8.9-cm S. polymorpha is estimated
to have nearly 1300 ED50 doses, and the capacity to paralyze about 650 larvae. Based on
our feeding observations, we believe that even the largest of our S. polymorpha would be
satiated and refusing to attack prey long before consuming 6.5 g (about 290% of its own
body mass, LD50) to 32 g (1429% of its own body mass, ED50) of prey! Yang et al.
(2012) determined the LD50 (48 hr) of S. s. mutilans venom for blowflies (15–25 mg),
blowfly larvae (35–45 mg), mealworms (Tenebrio molitor larvae; 190–210 mg), and
cockroaches (Periplaneta americana; 700–900 mg) as 5, 1174.9, 10.5, and 346.7 pg/g,
respectively. Based on Yang et al.'s (2012) yield of 0.2 mg venom per adult centipede
80
(0.4 mg venom in a 2-week period with weekly milking), we calculate the yield from a
single S. s. mutilans could kill as many as 8 x 109 blowflies, or 1.9 x 106 blowfly larvae,
or 4.5 x 107 mealworms, or 3.2 x 105 cockroaches. Similarly, Rates et al. (2007)
determined that the venom yield from a single 10-cm S. v. nigra could instantly kill
100,000 house flies.
Although the above calculations imply that these centipedes possess a large
excess of venom, a more biologically realistic consideration of their venom supply must
take into account a group of factors that may have influenced the evolution of these
centipedes’ venom stores, including selection for speed of kill/immobilization, frequency
of prey capture, variation in prey susceptibility to venom, selection for adequate
defensive capacity, and cost of venom production, regeneration, and storage. Although
we couch our discussion in terms of venom capacity (i.e., the amount of venom the
centipedes possess), we recognize that selection pressures for venom capacity are likely
deeply intertwined with selection pressures on venom composition, toxicity, and
quantities delivered during stinging.
Lethal dose measures are useful for comparing the relative lethality of different
toxins or venoms (e.g., Quistad et al., 1992) and, in combination with yield, for gaining a
rudimentary perspective on the potential killing power of a venomous animal. However,
such measures are virtually meaningless from the animal’s perspective, as they may not
reflect the normal quantities and routes of venom administration required to immobilize
and kill prey within a time frame relevant for securing a meal (Hayes, 2008). For
example, the amount of venom snakes inject into prey often exceeds the lethal dose by
100–1000 times (Mebs, 2001; Young et al., 2002), and the amount of venom C. salei
81
injected into crickets was three times the LD50 and thought to exceed the LD100 (Malli et
al., 1998; cf. Wigger et al., 2002). Such “overkill” in the quantity of venom expended
may be likened to "bet hedging" on the part of the predator, acting as a form of insurance
in the event of an ineffective bite (Hayes, 1992a). Observations in our lab of centipedes
preying on house crickets (A. domesticus) indicate that prey consumption typically begins
shortly after the first sting and progresses relatively rapidly. Such a short elapsed time
from sting to consumption means that if venom is playing a role in subduing prey it must
kill or paralyze quickly. Injected venom quantities exceeding the reported LD50 values
would be needed to achieve such fast action. Likewise, observations of very rapid
paralyzation or death of prey following other scolopendromorph stings (Keegan, 1963;
Lewis, 1981; Minton, 1974; Rates et al., 2007; Vijayakumar et al., 2012) suggest they are
using larger amounts of venom than doses calculated to be lethal after 24 or 48 hours.
However, given the potential digestive function of centipede venom (Dugon and Arthur,
2012b; Jangi, 1984), selection may act on the venom supply and expenditure for reasons
other than killing (Hayes, 2008; Mirtschin et al., 2002). Nonetheless, given the rapidity
with which scolopendromorphs immobilize prey, S. polymorpha and S. subspinipes likely
carry fewer biologically relevant “doses” of venom than estimated by the lethal dose
calculations above.
In addition to being influenced by the need to kill or immobilize prey rapidly, the
venom supplies of centipedes may also be influenced by selection for the ability to
subdue multiple prey in a short time. Little is known about scolopendromorph foraging
behavior under natural conditions (Menez et al., 1990; Molinari et al., 2005; Wallwork,
1982), including feeding frequency, largely due their cryptic lifestyle (Vijayakumar et al.,
82
2012). However, in a rare investigation of scolopendromorph hunting behavior,
Formanowicz and Bradley (1987) found that S. polymorpha in the lab actively searched
for prey (T. molitor larvae) at low prey density but switched to ambush tactics at high
prey density, killing approximately five (low prey density) to ten (high prey density) prey
over a 6-hour time period. Unfortunately, the authors did not report numbers of prey
consumed or fed upon, or the mean mass of prey consumed for the six-hour period,
leaving unclear the relationship between killing and consumption. Nonetheless, this study
demonstrated that S. polymorpha is capable of killing multiple prey in a short span of
time. Similarly, S. heros in captivity can attack and consume multiple prey in succession
(Campbell, 1932). However, while capable of killing multiple prey in a short time,
satiated scolopendrids may be reluctant to attack additional prey for several days (Jangi,
1966). Additionally, venom availability may influence predatory behavior, a phenomenon
also observed in spiders and included in the venom-optimization hypothesis (Hostettler
and Nentwig, 2006; Wullschleger and Nentwig, 2002). Scolopendra s. mutilans with
depleted venom supplies attacked prey less frequently than those with replete glands
(Dugon and Arthur, 2012b). Another situation in which the need to envenomate multiple
prey in succession might arise is when centipedes lose prey after an initial attack
(Carpenter and Gillingham, 1984; Cloudsley-Thompson, 1968; pers. obs.). If prey
escaped after being injected with venom, maintaining sufficient venom capacity would
ensure that venom remained to capture additional prey. However, it remains unclear
exactly what role venom plays in prey capture. Scolopendromorphs may rely to some
extent on the powerful piercing and cutting power of their sharp, heavily chitinized
83
forcipules (Bücherl, 1971c; Jangi, 1984; Manton, 1977) and coxosternal tooth-plates
(Manton, 1964) to subjugate prey without envenomation.
The variety of prey taken and their various susceptibilities to venom have likely
influenced the evolution of centipedes’ venom supplies. Scolopendromorphs are
generalized predators (Malta et al., 2008; Menez et al., 1990) primarily capturing live
prey (Bücherl, 1971a; Lewis, 1981), although scavenging may occur (Lewis et al., 2010;
Manton, 1964). Arthropods comprise the bulk of the diet (Lewis, 1981; Menez et al.,
1990), but larger scolopendromorphs are also implicated in attacks on vertebrates
including frogs (Forti et al., 2007), toads (Carpenter and Gillingham, 1984; CloudsleyThompson, 1968), lizards (Bauer, 1990; Orange, 1989), snakes (Cloudsley-Thompson,
1968; Easterla, 1975; Orange, 1989), rodents (Clark, 1979; Shugg, 1961), bats (Molinari
et al., 2005), and birds (Anonymous, 1985; Cloudsley-Thompson, 1968; Cumming,
1903). The diet of S. polymorpha includes a variety of vertebrates and invertebrates
(Formanowicz and Bradley, 1987), but has not been carefully documented in the field
(Crawford, 1990). The diet of S. subspinipes has likewise not been carefully documented
outside the lab, although it has been observed in the field to consume lizards (La Rivers,
1948), slugs (Lawrence, 1934), and various winged insects (Remington, 1950). Given the
potentially broad scope of natural prey of scolopendromorphs, a wide range of prey
susceptibilities is possible. The wide range (two orders of magnitude) of LD50 values for
S. s. mutilans venom in a variety of invertebrates (Yang et al., 2012) supports this
assertion. If some prey have low susceptibilities to scolopendromorph venoms,
expenditure of relatively large amounts of venom may be required to achieve rapid prey
immobilization or death. The need to quickly subdue prey whose defenses pose a
84
potential threat may also select for the accumulation and expenditure of relatively large
amounts of venom (Hayes et al., 2002; Malli et al., 1999; Wigger et al., 2002).
Maintenance of sufficient venom stores for defense, especially in cases where
defense must follow shortly after predatory venom expenditure, may be a selection
pressure on venom supplies in scolopendromorphs. Predation pressure by relatively large
vertebrates, including toads, monitor lizards, snakes, armadillos, mongooses, foxes,
burrowing owls, and mockingbirds (Crawford, 1990; Curry, 1986; Jangi, 1966; Lewis,
1981; Lewis et al., 2010) that may require sizeable amounts of venom to deter, may select
for larger venom supplies. However, it is possible that, as with "dry bites" in snakes (De
Rezende et al., 1998; Gibbons and Dorcas, 2002; Russell et al., 1997) and spiders
(Celerier et al., 1993), and dry stings in scorpions (Nisani and Hayes, 2011),
scolopendromorph dry stings (Bush et al., 2001; Cornwall, 1916; Klingel, 1960) may
have their own deterrent power.
Whereas the above factors might select for larger venom supplies, other pressures
could select for smaller venom stores in scolopendromorphs. Chief among these would
be the metabolic costs associated with producing (McCue, 2006; Nisani et al., 2007;
Pintor et al., 2010; Pintor et al., 2011) and storing (Enzor et al., 2011; Inceoglu et al.,
2003) large quantities of venom.
Conclusions
In summary, our data show that while several factors influenced venom yield and
venom protein concentration in scolopendrid centipedes, body length and geographic
origin were of primary influence on yield and protein concentration, respectively.
However, the challenge remains to determine the ecological significance of these
85
influences. The implications for centipede-sting risk are simple to infer given the
importance of body length in determining venom yield: severity of envenomation is
likely greater with larger centipedes, all else being equal. However, further investigation
is required to better understand the functional consequences of centipede venom yields.
We need to learn more about the amounts of venom deployed in natural predatory and
defensive contexts, and the influence of venom availability on behavioral decisions
regarding venom use.
Acknowledgments
We thank Dr. P. Duerksen-Hughes and her students in the LLU Biochemistry
Department for access to their equipment. We are also grateful to Charles Kristensen of
Spider Pharm Inc. for his valuable instruction regarding venom extraction. This work was
supported by the Department of Earth and Biological Sciences, Loma Linda University.
Conflict of Interest
The authors declare that there are no conflicts of interest involved in this work.
Ethical Statement
This paper represents a series of experiments carried out under the standard
procedures of scientific ethics. The animal experiments were conducted in accordance
with Loma Linda University’s Policy on the Care and Use of Laboratory Animals.
86
References
Abdel-Aal, A., Abdel-Baset, A., 2010. Venom yield and toxicities of six Egyptian snakes
with a description of a procedure for estimating the amount of venom ejected by a
single snake bite. Sci. J. King Faisal Univ. (Basic and Appl. Sci.) 11, 169-184.
Anonymous. 1985. Dangerous Australians. Bay Books, Sydney.
Toxicon 53, 367-374.
Atkinson, R.K., 1981. Comparisons of the neurotoxic activity of the venom of several
species of funnel web spiders (Atrax). Aust. J. Exp. Biol. Med. Sci. 59, 307-316.
Baerg, W., 1924. The effect of the venom of some supposedly poisonous arthropods
(centipedes and scorpions). Ann. Entomol. Soc. Am. 17, 343-352.
Barber, A.D., 2011. Lithobius forficatus (Linn., 1758) with apparently massive scar tissue
on damaged forcipules. Bulletin of the British Myriapod and Isopod Group 25,
52.
Bauer, A.M., 1990. Gekkonid lizards as prey of invertebrates and predators of
vertebrates. Herpetol. Rev. 21, 83-87.
Binford, G.J., 2001. An analysis of geographic and intersexual chemical variation in
venoms of the spider Tegenaria agrestis (Agelenidae). Toxicon 39, 955-968.
Blum, M.S., Callahan, P.S., 1960. Chemical and biological properties of the venom of the
imported fire ant (Solenopsis saevissima var. richteri Forel) and the isolation of
the insecticidal component, in: Pavan, M., Eisner, T. (Eds.), XI. Internationaler
kongress fur entomologie, Wien 1960. Istituto di Entomologia Agrara Dell'
Universita di Pavia, Pavia, Italy, pp. 290-293.
Boeve, J.L., Kuhn-Nentwig, L., Keller, S., Nentwig, W., 1995. Quantity and quality of
venom released by a spider (Cupiennius salei, Ctenidae). Toxicon 33, 1347-1357.
87
Bonato, L., Edgecombe, G.D., Lewis, J.G.E., Minelli, A., Pereira, L.A., Shelley, R.M.,
Zapparoli, M., 2010. A common terminology for the external anatomy of
centipedes (Chilopoda). ZooKeys 69, 17-51.
Bouchard, N.C., Chan, G.M., Hoffman, R.S., 2004. Vietnamese centipede envenomation.
Vet. Hum. Toxicol. 46, 312-313.
Bradford, M.M., 1976. Rapid and sensitive method for quantitation of microgram
quantities of protein utilizing principle of protein-dye binding. Anal. Biochem. 72,
248-254.
Brochetto-Braga, M.R., Palma, M.S., Carmona, E.C., Chaud-Netto, J., Rodrigues, A., da
Silva, G.P., 2006. Influence of the collection methodology on the Apis mellifera
venom composition: peptide analysis. Sociobiology 47, 759-770.
Brown, J.H., 1973. Toxicology and pharmacology of venoms from poisonous snakes.
Charles C. Thomas, Springfield.
Brunhuber, B.S., 1970. Egg laying, maternal care and development of young in the
scolopendromorph centipede, Cormocephalus anceps anceps Porat. J. Linn. Soc.
Lond. Zool. 49, 225-234.
Bücherl, W., 1953. Dosagem comparada da atividade dos extratos glandularis e do
veneno pro de Phoneutria nigriventer (Keyserling 1891). Mem. Inst. Butantan 25,
1.
Bücherl, W., 1971a. Introduction, in: Bücherl, W., Buckley, E.E. (Eds.), Venomous
animals and their venoms: volume III, venomous invertebrates. Academic Press,
New York, pp. xix-xxii.
Bücherl, W., 1971b. Venomous chilopods or centipedes, in: Bucherl, W., Buckley, E.E.,
169-196.
Cutis 37, 241.
Campbell, M., 1932. Some observations on Scolopendra heros under laboratory
conditions. Trans. Kans. Acad. Sci. 35, 80-81.
Carpenter, C.C., Gillingham, J.C., 1984. Giant centipede (Scolopendra alternans) attacks
marine toad (Bufo marinus). Caribb. J. Sci. 20, 71-72.
88
Celerier, M.L., Paris, C., Lange, C., 1993. Venom of an aggressive African
Theraphosidae (Scodra griseipes): milking the venom, a study of its toxicity and
its characterization. Toxicon 31, 577-590.
Chacon, D., Rodriguez, S., Arias, J., Solano, G., Bonilla, F., Gomez, A., 2012.
Maintaining coral snakes (Micrurus nigrocinctus, Serpentes: Elapidae) for venom
production on an alternative fish-based diet. Toxicon 60, 249-253.
Chao, J., Chang, H., 2006. Variation of the poison duct in Chilopoda centipedes from
Taiwan. Nor. J. Entomol. 53, 139-151.
Chaou, C.H., Chen, C.K., Chen, J.C., Chiu, T.F., Lin, C.C., 2009. Comparisons of ice
packs, hot water immersion, and analgesia injection for the treatment of centipede
envenomations in Taiwan. Clin. Toxicol. 47, 659-662.
Chavez-Olortegui, C., Kalapothakis, E., Ferreira, A., Ferreira, A.P., Diniz, C.R., 1997.
Neutralizing capacity of antibodies elicited by a non-toxic protein purified from
the venom of the scorpion Tityus serrulatus. Toxicon 35, 213-221.
Chow, G.C., 1960. Tests of equality between sets of coefficients in two linear
regressions. Econometrica 28, 591-605.
Clark, D.B., 1979. Centipede preying on a nestling rice rat (Oryzomys bauri). J. Mammal.
60, 654.
Press, Oxford.
Cohen, J., 1988. Statistical power analysis for the behavioral sciences, 2nd ed. Erlbaum,
Hillsdale, NJ.
Cohen, J., Cohen, P., West, S.G., Aiken, L.S., 2003. Applied multiple
regression/correlation analysis for the behavioral sciences. Lawrence Erlbaum
Associates, Mahwah, NJ.
Crabill, R.E., 1960. A new American genus of cryptopid centipede, with an annotated
key to the scolopendromorph genera from America north of Mexico. Proc. U.S.
Natn. Mus. 111, 167-195.
Crawford, C.S., Riddle, W.A., 1974. Cold-hardiness in centipedes and scorpions in New
Mexico. Oikos 25, 86-92.
89
Cumming, W.D., 1903. The food and poison of centipedes. J. Bombay Nat. Hist. Soc. 15,
364-365.
De Rezende, N.A., Torres, F.M., Dias, M.B., Campolina, D., Chavez-Olortegui, C.,
Amaral, C.F.S., 1998. South American rattlesnake bite (Crotalus durissus sp)
without envenoming: insights on diagnosis and treatment. Toxicon 36, 20292032.
de Roodt, A.R., Dolab, J.A., Galarce, P.P., Gould, E., Litwin, S., Dokmetjian, J.C., Segre,
L., Vidal, J.C., 1998. A study on the venom yield of venomous snake species
from Argentina. Toxicon 36, 1949-1957.
de Roodt, A.R., Laskowicz, R., Saavedra, S., Vucharchuc, M., Lanari, L.C., Reati, G.,
Beltramino, J., Varni, L., Lopez, R., Bazan, E.E., de Oliveira, V.C., 2012.
Comments on the venom yield of Tityus trivittatus, considering two
methodologies of extraction. Toxicon 60, 183.
Boubee, Paris.
di Tada, I.E., Martori, R.A., Doucet, M.E., Abalos, J.W., 1976. Venom yield with
different milking procedures, in: Rosenberg, P. (Ed.), Toxins: animal, plant and
microbial. Pergamon Press, Oxford, pp. 3-7.
Duboscq, O., 1898. Recherches sur les chilopodes. Arch. Zool. Exp. Gen. 6, 481-650.
Dugon, M.M., Arthur, W., 2012a. Comparative studies on the structure and development
90
Dugon, M.M., Arthur, W., 2012b. Prey orientation and the role of venom availability in
the predatory behaviour of the centipede Scolopendra subspinipes mutilans
(Arthropoda: Chilopoda). J. Insect Physiol. 58, 874-880.
41, 231-243.
Easterla, D.A., 1975. Giant desert centipede preys upon snake. Southwest. Nat. 20, 411.
Edgecombe, G.D., Koch, M., 2008. Phylogeny of scolopendromorph centipedes
(Chilopoda): morphological analysis featuring characters from the peristomatic
area. Cladistics 24, 872-901.
Enzor, L.A., Wilborn, R.E., Bennett, W.A., 2011. Toxicity and metabolic costs of the
Atlantic stingray (Dasyatis sabina) venom delivery system in relation to its role in
life history. J. Exp. Mar. Biol. Ecol. 409, 235-239.
Fairly, H., Splatt, B., 1929. Venom yields in Australian poisonous snakes. Med. J. Aust.
1, 336-348.
Field, A., 2005. Discovering statistics using SPSS. Sage Publications, London.
Fiero, M.K., Bonilla, C.A., Seifert, M.W., Weaver, T.J., 1972. Comparative study of
juvenile and adult prairie rattlesnake (Crotalus viridis viridis) venoms. Toxicon
10, 81-82.
Formanowicz, D.R., Bradley, P.J., 1987. Fluctuations in prey density: effects on the
foraging tactics of scolopendrid centipedes. Anim. Behav. 35, 453-461.
Fox, G.A., Hayes, W.K., Cooper, A.M., Nelsen, D.R., 2009. Venom yield and
characteristics in the desert hairy scorpion (Hadrurus arizonensis). Bull. South.
Calif. Acad. Sci. 108, 108.
Friedel, T., Nentwig, W., 1989. Immobilizing and lethal effects of spider venoms on the
cockroach and the common mealbeetle. Toxicon 27, 305-316.
91
Frund, H., 1992. The occurence and frequency of scars in centipedes, in: Meyer, E.,
Thaler, K., Schedl, W. (Eds.), Advances in myriapodology: proceedings of the 8th
international congress of myriapodology. Universitatsverlag Wagner, Innsbruck,
pp. 269-275.
Furtado, F.D., Maruyama, M., Kamiguti, A.S., Antonio, L.C., 1991. Comparative study
of 9 Bothrops snake venoms from adult female snakes and their offspring.
Toxicon 29, 219-226.
Gibbons, J.W., Dorcas, M.E., 2002. Defensive behavior of cottonmouths (Agkistrodon
piscivorus) toward humans. Copeia, 195-198.
Glenn, J.L., Straight, R.C., 1977. The midget faded rattlesnake (Crotalus viridis
concolor) venom: lethal toxicity and individual variability. Toxicon 15, 129-133.
Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1982. Occurrence of histamine
and histamine-release by centipede venom. Indian J. Med. Res. 76, 888-891.
Gonzalez-Morales, L., Diego-Garcia, E., Segovia, L., Gutierrez, M.D., Possani, L.D.,
2009. Venom from the centipede Scolopendra viridis Say: purification, gene
cloning and phylogenetic analysis of a phospholipase A2. Toxicon 54, 8-15.
Gopalakrishnakone, P., 1992. Light and electron microscopic features of the venom
apparatus of the centipede Scolopendra morsitans. Toxicon 30, 514.
Gregory-Dwyer, V.M., Egen, N.B., Bosisio, A.B., Righetti, P.G., Russell, F.E., 1986. An
isoelectic focusing study of seasonal variation in rattlesnake venom proteins.
Toxicon 24, 995-1000.
Gujarati, D., 1970. Use of dummy variables in testing for equality between sets of
coefficients in 2 linear regressions: a note. Am. Stat. 24, 50-52.
Guo, J., Zhang, R., Wang, J., Miao, L., 2013. Molecular cloning and characterization of a
new cDNA encoding a trypsin-like serine protease from the venom gland of
Scolopendra subspinipes mutilans. Afr. J. Pharm. Pharmacol. 7, 1054-1060.
Haddad, V., Cardoso, J.L.C., Lupi, O., Tyring, S.K., 2012. Tropical dermatology:
venomous arthropods and human skin: part II. Diplopoda, Chilopoda, and
Arachnida. J. Am. Acad. Dermatol. 67, 347.e1-347.e9.
92
Haight, K.L. 2002. Fire ant venom economy: patterns of synthesis and use [M.S. thesis].
Tallahassee: Florida State University. 75 p.
Harwood, R.F., James, M.T., Herms, W.B., 1979. Entomology in human and animal
health. Macmillan, New York.
Hayden, L., Parkes, G., Arthur, W., 2012. Segment number, body length, and latitude in
geophilomorph centipedes: a 'converse-Bergmann' pattern. Biol. J. Linn. Soc. 107,
166-174.
Hayes, W.K., 1992. Factors associated with the mass of venom expended by prairie
rattlesnakes (Crotalus v. viridis) feeding on mice. Toxicon 30, 449-460.
Hayes, W.K., Herbert, S.S., Rehling, G.C., Bennaro, J.F., 2002. Factors that influence
venom expenditure in viperids and other snake species during predatory and
defensive contexts, in: Schuett, G.W., Hoggren, M., Douglas, M.E., Greene, H.W.
(Eds.), Biology of the vipers. Eagle Mountain Publishers, Eagle Mountain, pp.
207-233.
Hayes, W.K., Kaiser, I.I., Duvall, D., 1992. The mass of venom expended by prairie
rattlesnakes when feeding on rodent prey, in: Campbell, J.A., Brodie, E.D., Jr.
(Eds.), Biology of the pitvipers. Selva, Tyler, Texas, pp. 383-388.
Hayes, W.K., Mackessy, S.P., 2010. Sensationalistic journalism and tales of snakebite:
are rattlesnakes rapidly evolving more toxic venom? Wilderness Environ. Med.
21, 35-45.
Herzig, V., 2010. Ontogenesis, gender, and molting influence the venom yield in the
spider Coremiocnemis tropix (Araneae, Theraphosidae). J. Venom Res. 1, 76-83.
Herzig, V., Khalife, A.A., Chong, Y., Isbister, G.K., Currie, B.J., Churchill, T.B., Horner,
S., Escoubas, P., Nicholson, G.M., Hodgson, W.C., 2008. Intersexual variations in
Northern (Missulena pruinosa) and Eastern (M. bradleyi) mouse spider venom.
Toxicon 51, 1167-1177.
Herzig, V., Ward, R.J., dos Santos, W.F., 2004. Ontogenetic changes in Phoneutria
nigriventer (Araneae, Ctenidae) spider venom. Toxicon 44, 635-640.
93
Heymons, R., 1901. Die entwicklungsgeschichte der Scolopender. Zoologica 13, 1-244.
Hill, R.E., Mackessy, S.P., 1997. Venom yields from several species of colubrid snakes
and differential effects of ketamine. Toxicon 35, 671-678.
Hojat, M., Xu, G., 2004. A visitor's guide to effect sizes: statistical significance versus
practical (clinical) importance of research findings. Adv. Health Sci. Educ. 9,
241-249.
Hopkins, C., Grilley, M., Miller, C., Shon, K.J., Cruz, L.J., Gray, W.R., Dykert, J.,
Rivier, J., Yoshikami, D., Olivera, B.M., 1995. A new family of Conus peptides
targeted to the nicotinic acetylcholine receptor. J. Biol. Chem. 270, 22361-22367.
Hostettler, S., Nentwig, W., 2006. Olfactory information saves venom during preycapture of the hunting spider Cupiennius salei (Araneae: Ctenidae). Funct. Ecol.
20, 369-375.
Huang, P., Mackessy, S.P., 2004. Biochemical characterization of phospholipase A2
(trimorphin) from the venom of the Sonoran lyre snake Trimorphodon biscutatus
lambda (family Colubridae). Toxicon 44, 27-36.
Sci. U.S.A. 100, 922-927.
Jakob, E.M., Marshall, S.D., Uetz, G.W., 1996. Estimating fitness: a comparison of body
condition indices. Oikos 77, 61-67.
Janes, D.N., Bush, S.P., Kolluru, G.R., 2010. Large snake size suggests increased
snakebite severity in patients bitten by rattlesnakes in southern California.
Calcutta.
Jarrar, B.M., 2010. Morphology, histology and histochemistry of the venom apparatus of
the centipede, Scolopendra valida (Chilopoda, Scolopendridae). Int. J. Morphol.
28, 19-25.
94
Keegan, H.L., 1963. Centipedes and millipedes as pests in tropical areas, in: Keegan,
H.L., Macfarlane, W.V. (Eds.), Venomous and poisonous animals and noxious
plants of the Pacific region. Pergamon, London, pp. 161-163.
Klauber, L.M., 1997. Rattlesnakes: their habits, life histories, and influence on mankind.
Volume II. University of California Press, Berkeley.
Kochva, E., 1960. A quantitative study of venom secretion by Vipera palaestinae. Am. J.
Trop. Med. Hyg. 9, 381-390.
Kochva, E., Tonsing, L., Louw, A.I., Liebenberg, N.V., Visser, L., 1982. Biosynthesis,
secretion and in vivo isotopic labelling of venom of the Egyptian cobra, Naja haje
annulifera. Toxicon 20, 615-636.
Kristensen, C., 2005. Comments on the natural expression and artificial extraction of
venom gland components from spiders. Toxin Rev. 24, 257-271.
Kronmuller, C., 2012. Review of the subspecies of Scolopendra subspinipes Leach, 1815
with the new description of the South Chinese member of the genus Scolopendra
Linnaeus, 1758 named Scolopendra hainanum spec. nov. Spixiana 35, 19-27.
Kuhn-Nentwig, L., Nentwig, W., 1997. Venom of the hunting spider Cupiennius salei
(Ctenidae). Toxicon 35, 813.
Kuhn-Nentwig, L., Schaller, J., Nentwig, W., 2004. Biochemistry, toxicology and
ecology of the venom of the spider Cupiennius salei (Ctenidae). Toxicon 43, 543553.
Kuhn-Nentwig, L., Stocklin, R., Nentwig, W., 2011. Venom composition and strategies
in spiders: is everything possible?, in: Casas, J. (Ed.), Advances in insect
physiology. Academic Press, Burlington, pp. 1-86.
La Rivers, I., 1948. Some Hawaiian ecological notes. Wasmann Collector 7, 85-110.
Lawrence, R.F., 1947. Some observations on the post-embryonic development of the
Natal forest centipede, Cormocephalus multispinus (Kraep.). Ann. Natal Mus. 11,
139-156.
Lawrence, T.C., 1934. Notes on the feeding habits of Scolopendra subspinipes Leach
(Myriapoda). Proc. Hawaii. Entomol. Soc. 8, 497-498.
95
Lewis, J.G.E., 1968. Individual variation in a population of the centipede Scolopendra
amazonica from Nigeria and its implications for methods of taxonomic
discrimination in the Scolopendridae. J. Linn. Soc. Lond. Zool. 47, 315-326.
Lewis, J.G.E., 1970. The biology of Scolopendra amazonica in Nigerian guinea
savannah. Bull. Mus. Hist. nat., Paris 41, 85-90.
Lewis, J.G.E., 1972. The life histories and distribution of the centipedes Rhysida nuda
togoensis and Ethmostigmus trigonopodus (Scolopendromorpha: Scolopendridae)
in Nigeria. J. Zool. (Lond.) 167, 399-414.
123-137.
Lewis, J.G.E., Edgecombe, G.D., Shelley, R.M., 2005. A proposed standardized
terminology for the external taxonomic characters of Scolopendromorpha
(Chilopoda). Fragm. Faun. 48, 1-8.
Mackessy, S.P., 1988. Venom ontogeny in the Pacific rattlesnakes Crotalus viridis helleri
and Crotalus viridis oreganus. Copeia, 92-101.
Mackessy, S.P., 2002. Biochemistry and pharmacology of colubrid snake venoms. J.
Toxicol.: Toxin Rev. 21, 43-83.
Mackessy, S.P., Sixberry, N.A., Heyborne, W.H., Fritts, T., 2006. Venom of the brown
treesnake, Boiga irregularis: ontogenetic shifts and taxa-specific toxicity.
Toxicon 47, 537-548.
Mackessy, S.P., Williams, K., Ashon, K.G., 2003. Ontogenetic variation in venom
composition and diet of Crotalus oreganus concolor: a case of venom
paedomorphosis? Copeia, 769-782.
96
Maldonado, J.V. 1998. The taxonomy and biology of southwestern U.S. scolopendrid
centipedes (Chilopoda: Scolopendridae) [M.S. thesis]. El Paso: University of
Texas at El Paso. 107 p.
Malli, H., Kuhn-Nentwig, L., Imboden, H., Moon, M.J., Wyler, T., 2000.
Immunocytochemical localization and secretion process of the toxin CSTX-1 in
the venom gland of the wandering spider Cupiennius salei (Araneae: Ctenidae).
Cell Tissue Res. 299, 417-426.
Malli, H., Kuhn-Nentwig, L., Imboden, H., Nentwig, W., 1999. Effects of size, motility
and paralysation time of prey on the quantity of venom injected by the hunting
spider Cupiennius salei. J. Exp. Biol. 202, 2083-2089.
Malli, H., Vapenik, Z., Nentwig, W., 1993. Ontogenic changes in the toxicity of the
venom of the spider Cupiennius salei (Araneae, Ctenidae). Zool. Jahrb., Abt. allg.
Zool. Physiol. Tiere 97, 113-122.
Manton, S.M., 1964. Mandibular mechanisms and the evolution of arthropods. Philos.
Trans. R. Soc. Lond., Ser. B: Biol. Sci. 247, 1-183.
Manton, S.M., 1977. The arthropoda: habits, functional morphology, and evolution.
Clarendon Press, Oxford.
Marsh, N., Glatston, A., 1974. Some observations on the venom of the rhinoceros horned
viper, Bitis nasicornis Shaw. Toxicon 12, 621-628.
Marsh, N.A., Whaler, B.C., 1984. The Gaboon viper (Bitis gabonica): its biology, venom
components and toxicology. Toxicon 22, 669-694.
Martin, C.M., 1971. Scorpions and centipedes of the Hawaiian Islands: their medical
significance. Hawaii Med. J. 30, 95-98.
97
McCleary, R.J.R., Heard, D.J., 2010. Venom extraction from anesthetized Florida
cottonmouths, Agkistrodon piscivorus conanti, using a portable nerve stimulator.
Toxicon 55, 250-255.
Copeia, 818-825.
McFee, R.B., Caraccio, T.R., Mofenson, H.C., McGuigan, M.A., 2002. Envenomation by
the Vietnamese centipede in a Long Island pet store. J. Toxicol. Clin. Toxicol. 40,
573-574.
Mebs, D., 2001. Toxicity in animals. Trends in evolution? Toxicon 39, 87-96.
Raton.
Meier, J., Freyvogel, T.A., 1980. Comparative studies on venoms of the fer-de-lance
(Bothrops atrox), carpet viper (Echis carinatus) and spitting cobra (Naja
nigricollis) snakes at different ages. Toxicon 18, 661-662.
Morphol. 206, 303-312.
Mertler, C.A., Vannatta, R.A., 2004. Advanced and multivariate statistical methods:
practical application and interpretation, 3rd ed. Pyrczak Publishing, Los Angeles.
Minton, S.A., 1974. Venom diseases. Thomas, Springfield.
Mirtschin, P., Davis, R., 1992. Snakes of Australia: dangerous and harmless. Hill of
Content, Melbourne.
Mohamed, A.H., Abusinna, G., Elshabaka, H.A., Elaal, A.A., 1983. Proteins, lipids,
lipoproteins and some enzyme characterizations of the venom extract from the
centipede Scolopendra morsitans. Toxicon 21, 371-377.
47, 189-190.
98
Morgan, P.N., 1969. Preliminary studies on venom from the brown recluse spider
Loxosceles reclusa. Toxicon 6, 161-165.
Nakagawa, S., 2004. A farewell to Bonferroni: the problems of low statistical power and
publication bias. Behav. Ecol. 15, 1044-1045.
Physiol. 147, 509-513.
Nisani, Z., Hayes, W.K., 2011. Defensive stinging by Parabuthus transvaalicus
scorpions: risk assessment and venom metering. Anim. Behav. 81, 627-633.
Norris, R. 2007. Centipede envenomation. In: Slapper, D., VanDeVoort, J., Farina, G.,
Halamka, J., Plantz, S., editors. eMedicine from WebMD. Boston: eMedicine
Corporation.
Orange, P., 1989. Incidents of predation on reptiles by invertebrates. Herpetofauna 19,
31-32.
Othong, R., Wananukul, W., Vohra, R., 2012. Centipede envenomation: 104 cases from
Bangkok, Thailand. Toxicon 60, 142.
Oukkache, N., Chgoury, F., Lalaoui, M., Alagon Cano, A., Ghalim, N., 2013.
Comparison between two methods of scorpion venom milking in Morocco. J.
Venom. Anim. Tox. incl. Trop. Dis. 19, doi:10.1186/1678-9199-19-5.
Owen, M.D., 1978. Venom replenishment, as indicated by histamine, in honey bee (Apis
mellifera) venom. J. Insect Physiol. 24, 433-437.
99
Perret, B.A. 1973. Biologie, gift und giftigkeit der orthognathen spinne Pterinochilus
spec. (fam. Theraphosidae) [Ph.D. thesis]: University of Basel. 80 p.
Perret, B.A., 1977. Venom regeneration in tarantula spiders—I: analysis of venom
produced at different time intervals. Comp. Biochem. Physiol., A: Mol. Integr.
Physiol. 56, 607-613.
Pilz, C., Melzer, R.R., Spelda, J., 2007. Morphometrics and SEM analysis of the species
pair Lithobius mutabilis L. Koch, 1862 and L. glacialis Verhoeff, 1937
(Chilopoda: Lithobiomorpha). Org. Divers. Evol. 7, 270e1-270e20.
Quistad, G.B., Dennis, P.A., Skinner, W.S., 1992. Insecticidal activity of spider
(Araneae), centipede (Chilopoda), scorpion (Scorpionida), and snake (Serpentes)
venoms. J. Econ. Entomol. 85, 33-39.
Radl, R.C., 1992. Brood care in Scolopendra cingulata Latreille, in: Meyer, E., Thaler,
K., Schedl, W. (Eds.), Proceedings of the 8th international congress of
myriapodology, Innsbruck, Austria, July 1990. Berichte des
Naturwissenschaftenlich Medizinischen Vereins in Innsbruck, Supplementum 10,
123-127.
Rocha-e-Silva, T.A.A., Sutti, R., Hyslop, S., 2009. Milking and partial characterization of
venom from the Brazilian spider Vitalius dubius (Theraphosidae). Toxicon 53,
153-161.
100
Rodriguez de la Vega, R.C., Schwartz, E.F., Possani, L.D., 2010. Mining on scorpion
venom biodiversity. Toxicon 56, 1155-1161.
Rosenberg, J., Hilken, G., 2006. Fine structural organization of the poison gland of
Lithobius forficatus (Chilopoda, Lithobiomorpha). Nor. J. Entomol. 53, 119-127.
Russell, F.E., Walter, F.G., Bey, T.A., Fernandez, M.C., 1997. Snakes and snakebite in
Central America. Toxicon 35, 1469-1522.
Sahayaraj, K., Kumar, S.M., Anandh, G.P., 2006. Evaluation of milking and electric
shock methods for venom collection from hunter reduviids. Entomon 31, 65-68.
Sandefer, C., 1998. The giant centipedes of the genus Scolopendra: their captive care and
husbandry. Privately printed.
Savel-Niemann, A., Roth, D., 1989. Biochemical analysis of tarantula venom (Eurypelma
californicum). Naturwissenschaften 76, 212-213.
Schanbacher, F.L., Lee, C.K., Hall, J.E., Wilson, I.B., Howell, D.E., Odell, G.V., 1973.
Composition and properties of tarantula Dugesiella hentzi (Girard) venom.
Toxicon 11, 21-29.
Schulte-Hostedde, A.I., Zinner, B., Millar, J.S., Hickling, G.J., 2005. Restitution of masssize residuals: validating body condition indices. Ecology 86, 155-163.
Martinsville.
Shugg, H.B., 1961. Predation on mouse by centipede. West. Aust. Nat. 8, 52.
Simaiakis, S.M., Giokas, S., Korsos, Z., 2011. Morphometric and meristic diversity of the
species Scolopendra cingulata Latreille, 1829 (Chilopoda: Scolopendridae) in the
Mediterranean region. Zool. Anz. 250, 67-79.
Sissom, W.D., Polis, G.A., Watt, D.D., 1990. Field and laboratory methods, in: Polis,
G.A. (Ed.), The biology of scorpions. Stanford University Press, Stanford, pp.
445-461.
Sokal, R.R., Rohlf, F.J., 1995. Biometry: the principles and practice of statistics in
biological research. W. H. Freeman & Co., New York.
Tare, T.G., Sutar, N.K., Renapurkar, D.M., 1986. A study of snake venom yield by
different methods of venom extraction. Amphib-reptil. 7, 187-191.
101
Trucchi, E., Pitzalis, M., Zapparoli, M., Bologna, M.A., 2009. Short-term effects of
canopy and surface fire on centipede (Chilopoda) communities in a semi natural
Mediterranean forest. Entomol. Fenn. 20, 129-138.
Turk, F.A., 1951. Myriapodological notes. III. The iatro-zoology, biology, and
systematics of some tropical "myriapods". Ann. Mag. Nat. Hist. 12, 35-48.
Uzel, A.P., Steinmann, G., Bertino, R., Korsaga, A., 2009. Necrotizing fasciitis and
cellulitis of the upper limb resulting from centipede bite: two case reports. Chir.
Main. 28, 322-325.
Vapenik, Z., Nentwig, W., 2000. The influence of hunger and breeding temperature on
the venom production of the spider Cupiennius salei (Araneae, Ctenidae).
Toxicon 38, 293-298.
Veraldi, S., Chiaratti, A., Sica, L., 2010. Centipede bite: a case report. Arch. Dermatol.
146, 807-808.
Verhoeff, K.W., 1940. Chilopoden-Kieferfuss-Regenerate in freier. Natur. Z. Morph.
Okol. Tiere 36, 645-650.
Vieira, E.G.J., Rolim-Rosa, R., Iizuka, H., Furtado, M.d.F.D., Fernandes, W., 1988.
Influencias sazonal e do processo de extração sobre a produção, toxicidade do
veneno e sobrevida de Bothrops jararaca (Wied, 1824). Mem. Inst. Butantan 50,
29-35.
Vijayakumar, K., Muthuraman, M., Jayaraj, R., 2012. Predation of stingless bees
(Trigona iridipennis: Apidae, Meliponinae) by centipede (Scolopendra hardwicki:
Chilopoda: Scolopendromorpha). Int. J. Adv. Lif. Sci. 5, 156-159.
London.
Wallwork, J.A., 1982. Desert soil fauna. Praeger, New York.
Whittemore, F.W., Keegan, H.L., Fitzgerald, C.M., Bryant, H.A., Flanigan, J.F., 1963.
Studies of scorpion antivenins: 2. Venom collection and scorpion colony
maintenance. Bull. W.H.O. 28, 505-511.
102
Wiener, S., 1956. The australian red back spider (Latrodectus hasseltii): II. Effect of
temperature on the toxicity of venom. Med. J. Aust. 43, 331-335.
Wiener, S., 1959. The Sydney funnel-web spider (Atrax robustus): II. Venom yield and
other characteristics of spider in captivity. Med. J. Aust. 46, 678-682.
749-752.
Wullschleger, B., Nentwig, W., 2002. Influence of venom availability on a spider's preychoice behaviour. Funct. Ecol. 16, 802-807.
Young, B.A., Lee, C.E., Daley, K.M., 2002. Do snakes meter venom? Bioscience 52,
1121-1126.
Zlotkin, E., Shulov, A.S., 1969. A simple device for collecting scorpion venom. Toxicon
7, 331-332.
103
CHAPTER THREE
VENOM REGENERATION IN THE CENTIPEDE SCOLOPENDRA POLYMORPHA:
EVIDENCE FOR ASYNCHRONOUS VENOM COMPONENT SYNTHESIS
Allen M. Coopera,*, Wayne J. Kellna, William K. Hayesa
a
Submitted for publication in Zoology
Allen M. Cooper
Griggs Hall #101
24941 Stewart St.
USA
Tel: 1-707-225-5018
Fax: 1-909-558-0259
104
Abstract
Venom regeneration comprises a vital process in animals that rely on venom for
prey capture and defense. The timing of venom regeneration in scolopendromorph
centipedes likely influences their ability to subdue prey and defend themselves, and may
influence the quantity and quality of venom extracted by researchers investigating the
venom’s biochemistry. We investigated venom volume and total protein regeneration
during the 14-day period subsequent to venom extraction in the North American
centipede Scolopendra polymorpha. We further tested the hypothesis that venom protein
components, separated by reversed-phase fast protein liquid chromatography (RP-FPLC),
undergo asynchronous (non-parallel) synthesis. During the first 48 hours, volume and
protein mass increased linearly. However, protein regeneration lagged behind volume
regeneration, with only 65–86% of venom volume and 29–47% of protein mass
regenerated during the first 2 days. No significant additional regeneration occurred over
the subsequent 12 days, and neither volume nor protein mass reached initial levels 7
months later (93% and 76%, respectively). Centipede body length was negatively
associated with rate of venom regeneration. Analysis of chromatograms of individual
venom samples revealed that five of 10 chromatographic regions and 12 of 28 peaks
demonstrated changes in percent of total peak area (i.e., percent of total protein) among
milking intervals, indicating that venom proteins are regenerated asynchronously. The
considerable regeneration of venom occurring within the first 48 hours, despite the
reduced protein content, suggests that predatory and defensive capacities are minimally
constrained by the timing of venom replacement.
105
Introduction
Many animals depend on venom to procure food and/or defend themselves. For
such animals, a reduced or depleted venom supply could represent a serious cost in terms
of lost prey capture opportunities or diminished defense capabilities (Currier et al., 2012;
Haight and Tschinkel, 2003; Hayes, 2008; Malli et al., 1998). Thus, a vital component in
the lives of virtually all venomous animals is the process of venom regeneration
subsequent to venom usage. Because venom is generally a complex mixture of
compounds (Abdel-Rahman et al., 2009; Fry et al., 2009; Undheim and King, 2011) and
represents a non-trivial metabolic expense (Billen, 1990; McCue, 2006; Nisani et al.,
2007, 2012; Pintor et al., 2010, 2011), natural selection should not only fine-tune the
amount of venom carried within an animal’s venom gland(s), but also the rate of
production of venom (Mirtschin et al., 2002). If encounter rates with prey vary with
species, habitat, sex and body size of predator, as they do in snakes (da Silva and Aird,
2001; Daltry et al., 1997), then such factors could potentially influence the rate at which
venom is produced (Mirtschin et al., 2002). Moreover, the rate of venom regeneration
may vary with biochemical complexity (Nisani, 2008; Nisani et al., 2012).
Venom regeneration has been studied in several groups of organisms, including
snakes (Brown et al., 1975; Currier et al., 2012; De Lucca et al., 1974; Klauber, 1997b;
Kochva, 1960; Kochva et al., 1982; Luna et al., 2009; McCue, 2006; Oron et al., 1978;
Pintor et al., 2010, 2011; Rotenberg et al., 1971; Willemse et al., 1979) and, to a lesser
degree, in invertebrates such as spiders (Boeve et al., 1995; Freyvogel et al., 1968;
Galindo et al., 2009; Kaire, 1963; Perret, 1977b; Uzenbaev and Lyabzina, 2009),
scorpions (Alami et al., 2001; Nisani et al., 2007, 2012; Pimenta et al., 2003), and
hymenopterans (Beard, 1971; Haight, 2012; Haight and Tschinkel, 2003; Owen, 1978).
106
However, as yet, there has been no study of venom regeneration in centipedes. The class
Chilopoda, part of the subphylum Myriapoda, is divided into five living orders (and 1
extinct) of centipedes: Scutigeromorpha, Lithobiomorpha, Craterostigmomorpha,
Scolopendromorpha, and Geophilomorpha (Edgecombe and Giribet, 2007). Centipedes
are multi-legged terrestrial arthropods recognized as an ecologically important group of
soil and leaf litter predators (Edgecombe and Giribet, 2007; Robertson et al., 1994;
Trucchi et al., 2009; Undheim and King, 2011; Wallwork, 1976). Centipedes,
approximately 3,500 species strong (Edgecombe and Giribet, 2007), are primarily
nocturnal predators that use the forcipules (modified pair of legs of the first trunk
segment) to grasp and envenomate live prey (Bonato et al., 2010; Lewis, 1981; Undheim
and King, 2011; Voigtlander, 2011) and defend themselves (Davis, 1993; Demange,
1981; Lewis, 1981; Maschwitz et al., 1979; Neck, 1985). Each forcipule houses a venom
gland, the histology of which has been reviewed by Undheim and King (2011).
The only published information immediately relevant to the timing of centipede
venom regeneration comes from passing references made to the frequency of milking of
members of the order Scolopendromorpha, which includes the largest (Mundel, 1990),
most fiercely predatory (Edgecombe and Koch, 2008), and most medically important
(Balit et al., 2004; Jangi, 1984) of all centipedes. Rates et al. (2007) repeatedly milked S.
viridicornis nigra and S. angulata using electrical stimulation at variable intervals
ranging from weekly to monthly, but did not mention if or how milking interval
influenced venom yield. Similarly, using electrical extraction, S. subspinipes dehaani was
milked every 20 days (Liu et al., 2012b), S. s. mutilans was milked weekly (Kong et al.,
2013; Yang et al., 2012, 2013), and S. viridicornis, Otostigmus pradoi, and Cryptops
107
iheringi were milked monthly (Malta et al., 2008), but none of the authors mentioned
whether these time intervals were chosen to maximize venom yield or if shorter intervals
were tried.
Understanding the rate of venom regeneration in scolopendromorphs is important
for two reasons. First, knowledge of the kinematics of venom regeneration may yield
insights into the largely unknown foraging ecology of scolopendromorphs (Menez et al.,
1990; Molinari et al., 2005; Wallwork, 1982), as well as their defensive behavior.
Although total depletion of venom supply (the aim in most venom regeneration studies,
including this one) likely never occurs under natural conditions, it is possible that
envenomation of multiple prey items (Campbell, 1932; Formanowicz and Bradley, 1987)
or an intense encounter with a predator might decrease venom supply to the point that the
centipede is constrained to wait for venom regeneration before additional prey can be
captured or a potent defense can once again be mounted. Thus, understanding the timing
of regeneration is an important step in learning how venom supply might impact the
timing of foraging and other activities that expose the animal to predators. Second,
understanding the rate of venom regeneration is important for researchers wishing to
further study the biochemistry of centipede venoms, a potentially rich source of novel
bioactive molecules (Liu et al., 2012b; Peng et al., 2010; Rates et al., 2007; Undheim et
al., 2012; Yang et al., 2012, 2013). Knowing when venom volume and venom protein
concentration return to normal levels will help researchers devise venom extraction
protocols that maximize yields and ensure they are studying venom samples
representative of replete glands.
108
For this investigation, we used the centipede Scolopendra polymorpha
(Chilopoda: Scolopendromorpha: Scolopendridae; Fig. 1). This centipede inhabits desert,
dry grassland, and forest habitats of the southern half of the United States from the Great
Plains westward to California, ranging up the Pacific states into Oregon, and throughout
the desert southwest into northern Mexico (Crabill, 1960; Crawford and Riddle, 1974;
Shelley, 2002). Relatively little is known about the ecology of this animal (Crawford,
1990; Wallwork, 1982), although the cold-hardiness and overwintering physiology
(Crawford and Riddle, 1974; Crawford et al., 1975), circadian activity patterns
(Cloudsley-Thompson and Crawford, 1970), burrowing behaviors (Davis, 1993), and
water loss (Hadley et al., 1982) have been investigated, and some of its predators
(including burrowing owls, scorpions, snakes, and centipedes) have been identified or
suggested (Cloudsley-Thompson and Crawford, 1970; Crawford, 1990; Davis, 1993;
Tennant, 1985). The diet of S. polymorpha includes a variety of invertebrates and
vertebrates (Formanowicz and Bradley, 1987; Punzo, 2000), but has not been carefully
documented in the field (Crawford, 1990). In the only study to date on scolopendromorph
hunting behavior, Formanowicz and Bradley (1987) found that, in the lab, S. polymorpha
actively searched for prey (Tenebrio molitor larvae) at low prey density, but switched to
ambush tactics at high prey density. The sting of S. polymorpha causes temporary sharp
pain in humans, lasting up to 3 hours with no associated swelling (Baerg, 1924;
Maldonado, 1998; Turk, 1951).
109
110
111
To date, no studies of venom regeneration in centipedes have been reported in the
literature. Thus, we designed the present study to determine the extent of venom volume
and total protein regeneration over the course of 14 days subsequent to venom gland
emptying in the centipede S. polymorpha. We hypothesized that regeneration of venom
volume would initially exceed regeneration of venom protein, as observed in other taxa
(Boeve et al., 1995; Kochva et al., 1982; Nisani et al., 2007). We also tested the
hypothesis that venom protein components, separated by reversed-phase fast protein
liquid chromatography (RP-FPLC), regenerate asynchronously (non-parallel).
Centipedes
We initially acquired S. polymorpha centipedes from Arizona (Cochise County;
Bugs of America LLC, Portal, AZ, USA). We later collected additional specimens from
southern California (San Bernardino, Riverside, and San Diego counties, CA, USA).
Centipedes (n = 156; body length range: 6.2–11.5 cm) were housed individually in plastic
containers with a mixed sand/soil substrate and water dish. They were offered a cricket
once every two weeks, and misted weekly. Both sexes of centipedes were used in this
study, and because sex did not influence venom yield or protein concentration in S.
polymorpha (Cooper et al., 2014), there was no a priori reason to believe venom
regeneration might differ between sexes. We obtained morphometric measurements,
using methods previously described (Cooper et al., 2014) from digital photographs of
CO2-anesthetized animals prior to venom extraction. Forcipule injury (often detected by a
blunted shape of the regenerated tarsungula), while only minimally documented (e.g.,
Barber, 2011; Frund, 1992), is not uncommon. Because of the potential for forcipule
112
injury to affect forcipule size and venom yield (Verhoeff, 1940, cited in Lewis, 1981),
only animals whose forcipules appeared uninjured were included in this study.
Experimental Design
14-Day Venom Regeneration Trial
Using a repeated-measures design, we milked 84 centipedes from Arizona on day
0, and then re-milked subsets of 12 centipedes on days 2, 4, 6, 8, 10, 12, and 14. We
chose to sample the 14-day period because most venom regeneration occurs within this
time frame in other arthropods (Beard, 1971; Boeve et al., 1995; Haight, 2012; Nisani et
al., 2012; Nisani et al., 2007; Perret, 1977b). None of these centipedes were subjected
previously to milking procedures. Duration in captivity prior to the trial ranged from 81–
414 days (mean ± SE: 145 ± 14 days). Centipedes were assigned to each of the seven
milking intervals in a balanced design with respect to centipede body length (i.e., all
milking interval groups had a similar range and normal distribution of body lengths).
Final sample size for analysis was n = 79 (body length range: 6.4–11.5 cm; five
centipedes were dropped from analysis due to milking difficulties). We determined
venom volume and protein content for both the initial and second milks of each animal.
Milking procedures and venom measurement are described in subsequent sections.
48-Hour Venom Regeneration Trial
Following inspection of data from the 14-day trial, we decided to investigate
venom regeneration over the first 48 hours in more detail. Using a repeated-measures
design, we milked 72 centipedes (n = 32 from Arizona, n = 40 from California) at time 0,
and then re-milked subsets of 12 centipedes at times 8, 16, 24, 32, 40, and 48 hours. None
113
of these centipedes were subjected previously to milking procedures. Duration in
captivity prior to the trial ranged from 53–968 days (mean ± SE: 338 ± 33 days).
Centipedes were assigned to each of the six milking interval groups in a balanced design
with respect to centipede body length. Final sample size for analysis was n = 71 (body
length range: 6.2–11.1 cm; one centipede was dropped from analysis due to milking
difficulties). We determined venom volume and protein content for the initial and second
milks of each animal. We subsequently analyzed venom samples using RP-FPLC.
For RP-FPLC analysis, we chromatographed the initial (0 hr) venom from 9 (n =
5 from Arizona; n = 4 from California) individual centipedes, and regenerated venom
from the second milk of individual centipedes at the remaining milking intervals (n = 10–
11 for each milking interval). Thus, each group, including the initial venom group, was
comprised of unique individuals, and therefore was independent. Each group had
individuals of similar body length and both geographic origins.
7-Month Follow-Up to 14-Day Regeneration Trial
Centipedes (n = 69) from the 14-day trial that were still alive seven months later
were re-milked, and the volume, protein mass, and protein concentration of the venom
determined. This allowed us to assess the long-term effects of the two venom extraction
procedures (for initial and regenerated venom).
Venom Extraction and Venom Volume Determination
Venom extraction and volume determination were conducted according to
previously described methods (Cooper et al., 2014). Briefly, we extracted venom by
electrical stimulation to the bases of the forcipules and collected venom in graduated 5114
µL Drummond PCR Micropipets (PGC Scientifics, Garner, NC, USA). We immediately
placed the extracted venom from each individual into a separate microcentrifuge tube
with 50 µL of chilled nanopure water, and stored it at -80 C. Following Dass and Jangi
(1978), centipedes were not fed for 2 weeks prior to venom extraction to ensure,
presumably, replete venom glands.
We determined protein concentration of individual venom samples (in nanopure
water) using the Coomassie blue dye-binding method (Bradford, 1976), with bovine
serum albumin (BSA) in nanopure water as the standard. Coomassie Protein Assay
Reagent and BSA were purchased from Thermo Fisher Scientific (Waltham, MA, USA).
Using the manufacturer’s 1–25 µg/mL protocol, all assays resulted in high coefficients of
determination (r2 > 0.98), indicating reliability. We then derived total protein and protein
concentration of the individual venom samples based on the original milked volume of
venom.
RP-FPLC Analysis
We conducted RP-FPLC on the portions of 48-h trial venom samples remaining
after protein quantification. Separations were performed on an ÄKTA FPLC (GE
Lifesciences, Pittsburgh, PA, USA). Venom samples (up to 402 µg, comprising all of the
available sample from each individual) were diluted with 175 µL buffer A (0.065% TFA,
2% ACN in water) and centrifuged at 10,000 xg for 10 min to remove cellular debris. A
SOURCE 15 RPC ST 4.6/100 polystyrene/divinyl benzene reversed-phase column (GE
115
Lifesciences) was equilibrated in buffer A, and 100 µL of the diluted sample injected
onto the column. Proteins were eluted at a flow rate of 0.5 mL/min, in a 40-columnvolume linear gradient of 0–100% buffer B (0.05% TFA, 80% ACN in water), and the
elution monitored at 214 nm using Unicorn 5.0 (GE Lifesciences) software.
Pre-Treatment and Analysis of RP-FPLC Chromatograms
Because retention time/volume shifts can impair results of analyses using
chromatographic data sets (Liang et al., 2010; Malmquist and Danielsson, 1994; Tomasi
et al., 2011; Zhang et al., 2011), we aligned chromatograms from individual venom
samples using wavelet pattern matching and differential evolution (Zhang et al., 2011)
prior to analysis. We exported chromatograms to an Excel file from Unicorn using 2x
downsampling and retention normalization. Each chromatogram was then baseline
corrected and aligned with the same reference chromatogram (generated by conducting
RP-FPLC on a pool of venom from 27 S. polymorpha from California from a separate
experiment) using the open-source software alignDE, implemented in the R programming
environment (Zhang et al., 2011). Parameters used in peak detection and baseline
correction of the reference chromatogram were: gapTH = 3, skip = 2, SNR.Th = 3,
ridgeLength=5, λ=500). Parameters used in peak detection and baseline correction of
target chromatogram were: gapTH = 3, skip = 2, ridgeLength = 10, peak shape threshold
= 0.5, λ = 500. The SNR.Th parameter for the target chromatogram was customized for
each alignment. Parameters used in peak alignment were: slack = 60, NP = 120, itermax
= 200. Following alignment, we imported chromatograms back into Unicorn and
integrated them to obtain relative peak areas. Despite alignment using alignDE, some
116
additional visual alignment of chromatographic peaks was necessary prior to analysis.
Visual alignment criteria included retention time, relative peak height, and peak shape
and area. We assumed that aligned peaks across samples contained the same components;
however, more detailed analyses (mass spectrometry or sequence analysis) are necessary
to verify this assumption. Visual alignment preceded any analyses, and therefore we
deemed the analyses unbiased.
We analyzed the RP-FPLC chromatograms using both a coarse-grained and a
fine-grained approach. The coarse-grained approach involved breaking each
chromatogram into the same 10 elution regions (13.6–16.2, 16.2–18.3, 18.3–21.7, 21.7–
25.3, 25.3–30.1, 30.1–33.0, 33.0–38.7, 38.7–45.0, 45.0–49.4, 49.4–55.0 mL)
encompassing the entire retention volume of 13.6–55.0 mL (14.7–76.7% buffer B).
Region boundaries, although arbitrary, were chosen after careful visual inspection of all
chromatograms so that each region contained one or more peak clusters. For each
chromatogram, we determined the percent of total protein for each region (i.e., peak area
in each region relative to total peak area). We then subjected percent of total protein
(relative peak area) for the 10 regions to factorial analysis of variance. The fine-grained
approach was similar but focused on individual peaks. We chose to compare those peaks
representing ≥0.5% of total protein and present in at least five samples of any given
milking interval group. We identified 28 peaks for the analysis, excluding a ubiquitous
peak resulting from buffer mixing at retention volume 7.8 mL (5.9% buffer B).
We relied on parametric tests (Mertler and Vannatta, 2004) when assumptions
were met, including Pearson correlation (r), standard multiple linear regression, t-tests,
117
analyses of variance (ANOVAs) and covariance (ANCOVAs), and multivariate analyses
of covariance (MANCOVAs). For ANCOVAs and MANCOVAs, we always tested the
assumption of homogeneity of regression slopes by including an interaction term, and
then removed it from the final model if no interaction existed (Mertler and Vannatta,
2004). For models that failed the assumption of sphericity, we applied GreenhouseGeisser adjustments to the degrees of freedom (Field, 2005). We also employed several
non-parametric tests (Zar, 1996), including Kruskal-Wallis ANOVA. Individual models
and data transformations are specified in the Results section.
We further computed effect sizes, which are independent of sample size (in
contrast to statistical significance) and more readily compared among different data sets
and different studies (Hojat and Xu, 2004; Nakagawa and Cuthill, 2007). For pairwise
comparisons (t-tests), we relied on Cohen’s d using pooled standard deviation (Hojat and
Xu, 2004), for which values of ~0.2, ~0.5, and ≥0.8 are generally considered small,
moderate, and large, respectively (Cohen, 1988). For ANOVA, ANCOVA, and
MANCOVA models, we computed eta-squared (η2), partial η2, and multivariate partial
η2, respectively, with values of ~0.01, ~0.06, and ≥0.14 loosely regarded as small,
moderate, and large, respectively (Cohen, 1988). Because partial η2 values for a single
model never summed to > 1.0, no adjustments were applied to these values. We
expressed effect size for bivariate correlation (r) as the coefficient of determination (r2),
with values of ~0.01, ~0.09, and ≥0.25 deemed small, moderate, and large, respectively
(Cohen, 1988). For multiple regression, we report R2adj (which accounts for the tendency
for R2 to overestimate the population value) for the full models, and squared semipartial
correlations (sr2; interpreted as the proportion of variance that a given IV accounts for in
118
the DV after taking into account the effects of other IVs) for individual predictors (Cohen
et al., 2003; Mertler and Vannatta, 2004). Cohen’s d provides a standardized unit of
difference, whereas the other effect size estimators roughly indicate the approximate
proportion of variance explained.
For multiple regressions, we computed relative forcipule width and relative
forcipule length using unstandardized residual scores (Mirtschin et al., 2002; SchulteHostedde et al., 2005) from the regressions of forcipule width versus body length, and
forcipule length versus body length. We further computed unstandardized residual scores
from the regression of ln(body mass) versus ln(body length) as an index of relative body
mass. We tested the absence of multicollinearity by examining tolerance values (>0.1)
and variance inflation factor scores (<10; Mertler and Vannatta, 2004).
Statistical tests were conducted using SPSS 20.0 for Windows (Statistical
Package for the Social Sciences, Inc., Chicago, 2011), with α = 0.05. Following
Nakagawa (2004), we chose not to adjust α for multiple tests. Unless indicated otherwise,
measures of central tendency presented are mean ± 1 S.E. The mean difference and
associated 95% CI limits given for analyses conducted with transformed data are backtransformed values.
Results
Initial Milking and General Trends in Venom Regeneration
Venom volume from the initial milking of all centipedes (n = 150) in the 48-hr
and 14-day trials averaged 1.1 ± 0.1 µL (range 0.1–3.3 µL). Within 2 days, 65 ± 6% (48hr trial) to 86 ± 10% (14-day trial) of initial milk venom volume was regenerated (Fig.
119
2). Volume regeneration did not increase substantially over the subsequent 12 days,
remaining at approximately 86% of the initial milking level.
120
121
122
Venom protein mass from the initial milking of all centipedes averaged 190 ± 10
µg (range 8–604 µg). Protein concentration from the initial milking averaged 165 ± 3
µg/µL (range 83–292 µg/µL), with California specimens (188 ± 6 µg/µL; n = 40) having
higher concentrations than those from Arizona (157 ± 3 µg/µL; n = 110; t148 = 4.91, p <
0.001, d = 0.91; mean difference and 95% CI = 31 [19–44]). Within 2 days, 29 ± 3% (48hr trial) to 47 ± 9% (14-day trial) of protein mass was regenerated (Fig. 2). Protein mass
regeneration did not increase substantially over the next 12 days, remaining at about 46%
of the initial milking level. Thus, protein regeneration lagged behind volume
regeneration, with protein concentration averaging about 40% (70 µg/µL, 48-hr trial) to
50% (79 µg/µL, 14-day trial) of the initial milking concentration.
Mean body lengths (untransformed) were similar for the 48-hr and 2-day groups
(8.7 ± 0.4 vs. 8.9 ± 0.4 cm, respectively; t22 = 0.323, p = 0.75, Cohen's d = 0.1; mean
difference and 95% CI = 0.2 [-0.9–1.3]), although the range was shifted slightly lower for
the 48-hr group relative to the 2-day group (6.7–11.0 and 7.0–11.3 cm, respectively).
48-Hr Trial
We used multiple regression to examine the relative contributions of milking
interval, geographic origin of centipede (dummy variable for Arizona, California), time in
captivity, body length, relative forcipule width, relative forcipule length, and relative
body mass in predicting percent of initial venom volume regenerated. Correlations among
variables included in the model are shown in Table 1. Multicollinearity was not a
problem. The overall model significantly predicted volume regeneration (R2adj = 0.34,
F7,63 = 6.08, p < 0.001), accounting for 34% of the variation. Two of the seven predictors
123
were significant (Table 2): milking interval (sr2 = 0.259, β = 0.529) and body length (sr2
= 0.080, β = -0.320). Percent volume regenerated increased 0.9% (95% CI = 0.6–1.3%)
for every 1-hour increase in milking interval, and decreased 6.0% (95% CI = -10.1– -1.9)
for every 1-cm increase in body length.
124
125
126
We used a similar regression model to examine the percent of initial venom
protein mass regenerated. Correlations are provided in Table 3. Multicollinearity was not
a problem. The overall model significantly predicted percent protein mass regenerated
(R2adj = 0.22, F7,63 = 3.89, p = 0.001), and accounted for 22% of the variation. Two of the
seven predictors were significant (Table 4): milking interval (sr2 = 0.172, β = 0.431) and
body length (sr2 = 0.064, β = -0.284). Percent protein mass regenerated increased 0.4%
(95% CI = 0.2–0.6%) for every 1-hour increase in milking interval, and decreased 2.8%
(95% CI = -5.1– -0.5%) for every 1-cm increase in body length.
127
128
129
We included significant predictors from the regression analyses (milking interval,
6 levels, between-subjects factor; body length, covariate) in a MANCOVA model that
examined their effect on the combined dependent variable of percent volume
(untransformed) and percent protein mass (untransformed) regenerated. Both predictors
were significant, but the relationship was stronger for milking interval (Wilks’ Λ = 0.55,
F10,126 = 4.47, p < 0.001, multivariate partial η2 = 0.26) than body length (Wilks’ Λ =
0.90, F2,63 = 3.45, p = 0.038, multivariate partial η2 = 0.10; note the differences in effect
sizes). Follow-up ANCOVAs showed milking interval differences were stronger for
percent volume regenerated (F5,64 = 6.14, p < 0.001, partial η2 = 0.32) than percent
protein mass regenerated (F5,64 = 4.07, p = 0.003, partial η2 = 0.24), and that body length
influenced percent volume regenerated (F1,64 = 6.74, p = 0.012, partial η2 = 0.10) more so
than percent protein mass regenerated (F1,64 = 4.12, p = 0.047, partial η2 = 0.06). Both
volume and protein regeneration showed linear trends, with increases across successive
milking intervals (percent volume regenerated: F1,65 = 27.60, p < 0.001, partial η2 = 0.30;
percent protein mass regenerated: F1,65 = 15.27, p < 0.001, partial η2 = 0.19).
Because no relationship existed between second milk venom protein
concentration and body length (r2 = 0.0003, p = 0.89), we used a factorial ANOVA
(disregarding body length) to investigate the effect of milking interval and geographic
origin on protein concentration (untransformed) of second milk venom. ANOVA
indicated a significant main effect of geographic origin (F1,59 = 6.39, p = 0.014, partial η2
= 0.10), with protein concentration lower for Arizona animals than for California animals
(60 ± 6 vs. 80 ± 5 µg/µL; mean difference and 95% CI = 20 [4–36]), but no difference
among milking intervals (F5,59 = 1.46, p = 0.22, partial η2 = 0.11), though the effect size
130
was relatively large but with no discernable pattern (Fig. 3). The interaction was not
significant (F5,59 = 0.68, p = 0.64, partial η2 = 0.05). Protein concentrations of 60 µg/µL
(Arizona) and 80 µg/µL (California) represented 39% and 42% of initial protein
concentration, respectively.
131
132
133
14-Day Trial
We applied the same multiple regression model used in the 48-hr trial to the 14day trial data for percent of initial venom volume regenerated. This omnibus model with
seven predictors did not significantly predict venom volume regeneration (R2adj = 0.02,
F7,71 = 1.27, p = 0.28). Because analysis of 48-hr trial data revealed that only milking
interval and body length significantly predicted venom regeneration, we restricted our
multiple regression model for the 14-day trial to include only these two predictors.
Correlations among variables included in the multiple regression model are shown in
Table 5. Multicollinearity was not a problem. The overall model significantly predicted
volume regeneration (R2adj = 0.06, F2,76 = 3.40, p = 0.038), accounting for 6% of the
variation. One of the two predictors was significant (Table 6): body length (sr2 = 0.081, β
= -0.284). Percent volume regenerated decreased 7.9% (95% CI = -14.1– -1.8%) for
every 1-cm increase in body length.
134
135
136
We employed the same omnibus multiple regression model to predict percent of
initial venom protein mass regenerated. This omnibus model with seven predictors was
not significant (R2adj = 0.07, F7,71 = 1.84, p = 0.09). Multiple regression was then
performed using milking interval and body length as predictors. Correlations are provided
in Table 7. Multicollinearity was not a problem. The overall model significantly predicted
percent protein mass regenerated (R2adj = 0.11, F2,76 = 5.63, p = 0.005), and accounted for
11% of the variation. Again, one of the two predictors was significant (Table 8): body
length (sr2 = 0.108, β = -0.329). Percent protein mass regenerated decreased 7.3% (95%
CI = -12.0– -2.6%) for every 1-cm increase in body length.
137
138
139
We included both variables from the regression analyses (milking interval, 7
levels, between-subjects factor; body length, covariate) in a MANCOVA model that
examined their effect on the combined dependent variable of percent volume
(untransformed) and percent protein mass (untransformed) regenerated. Because the
assumption of homogeneity of covariance matrices was violated and group sizes were
unequal, Pillai’s Trace was utilized when interpreting MANCOVA results (Mertler and
Vannatta, 2004). Although milking interval was not significant (Pillai’s Trace = 0.19,
F12,142 = 1.22, p = 0.27, multivariate partial η2 = 0.09), body length significantly
influenced the combined dependent variable (Pillai’s Trace = 0.11, F2,70 = 4.12, p =
0.020, multivariate partial η2 = 0.11). Percent venom volume regenerated when averaged
across all intervals was 86 ± 4% (range of interval means, 78–99%), whereas percent
protein mass regenerated was 46 ± 3% (range of interval means, 36–60%).
As with the 48-hr data, because no relationship existed between second milk
venom protein concentration and body length (r2 = 0.02, p = 0.20), we used one-way
ANOVA (disregarding body length) to investigate the effect of milking interval on
protein concentration (untransformed) of second milk venom. ANOVA indicated no
difference in second milk protein concentration among milking intervals (F6,72 = 1.29, p =
0.28, η2 = 0.10; Fig. 3). Protein concentration when data were pooled across all intervals
averaged 79 ± 2 µg/µL, representing 50% of initial protein concentration.
7-Month Follow-Up
Venom volume (square root-transformed) was significantly higher for the initial
milk than the third milk obtained seven months later (1.2 ± 0.1 vs. 1.0 ± 0.1 µL
140
respectively; t68 = 3.10, p = 0.003, d = 0.38; mean difference and 95% CI = 0.01 [0.002,
0.04]; Fig. 4A), with the third milk yielding 93 ± 7% of initially extractable volume. We
used multiple regression to examine the relative contributions of body length and time in
captivity in predicting third milk volume as a percent of initial milk volume.
Multicollinearity was not a problem. The overall model did not significantly predict
percent volume regenerated (R2adj = 0.03, F2,67 = 1.94, p = 0.15).
141
142
143
The same pattern was seen for protein mass (square root-transformed) of initial
and third milks (202 ± 16 vs. 130 ± 11 µg respectively; t68 = 5.65, p < 0.001, d = 0.68;
mean difference and 95% CI = 8 [3, 14]; Fig. 4B), with the third milk yielding 76 ± 6%
of initially extractable protein. We used the same multiple regression model employed for
volume to investigate protein mass regenerated as a percent of initial milk protein mass.
One multivariate outlier (determined by Mahalanobis distance; Mertler and Vannatta,
2004) was removed. The overall model did not significantly predict percent protein mass
regenerated (R2adj = 0.05, F2,65 = 2.63, p = 0.080).
Similarly, venom protein concentration was significantly higher for the initial
milk than the third milk (158 ± 4 vs. 128 ± 3 µg/µL respectively; t68 = 7.17, p < 0.001, d
= 0.87; mean difference and 95% CI = 30 [22, 38]; Fig. 4C), with the third milk
characterized by 83 ± 3% of initial protein concentration.
Analysis of Chromatograms from 48-Hr Trial
General Analysis of Chromatograms
A typical RP-FPLC chromatogram of initial (0 hr) venom from a single S.
polymorpha (from Arizona) is shown in Figure 5. RP-FPLC was able to separate on
average 74, 70, 75, 74, 73, 75, and 71 different peaks from the 8, 16, 24, 32, 40, 48-hr,
and initial venom groups, respectively.
144
145
146
Analysis of Chromatograms By Region
We conducted an omnibus factorial ANOVA to investigate the effect of
chromatographic region (10 levels, within-subjects), milking interval (7 levels, betweensubjects), and location (2 levels, between-subjects) on relative peak area summed across
each chromatographic region. We chose to analyze centered logratio-transformed (clr)
data (Filzmoser et al., 2009; Pawlowsky-Glahn and Egozcue, 2006; Schilling et al., 2012)
despite not strictly meeting parametric assumptions. The model yielded a complex,
difficult to interpret three-way interaction. Because including chromatographic region in
the model was uninformative (since relative area obviously differed among regions, and
this was not of primary interest), we subsequently ran two-way (location x interval)
ANOVAs for each region separately. We present graphs of clr-transformed data (Fig.
6A), as these show the data as they are analyzed in coordinate space (Ulbrich, 2011), and
also include graphs of untransformed relative peak area (Fig. 6B). Five of ten
chromatographic regions exhibited a significant main effect of milking interval or an
interaction between milking interval and location (Table 9), suggesting asynchronous
regeneration of venom components. Five also showed a significant main effect of
location or an interaction (Table 9), confirming that location needed to be controlled for
as a confounding factor. For the eight chromatographic regions having no interaction,
nonparametric Kruskal-Wallis ANOVAs with milking interval as the independent
variable yielded results (not shown) identical to the parametric ANOVAs. This
correspondence between parametric and non-parametric results supports our
interpretation of the parametric tests.
147
148
149
150
151
Analysis of Chromatograms by Peak
The 28 peaks selected for analysis collectively represented an average of 54.1,
60.1, 66.2, 61.7, 66.1, 68.4, and 72.9% of total peak area for the 8, 16, 24, 32, 40, 48-hr,
and initial venom time groups, respectively.
Similar to analysis of chromatographic regions, we conducted an omnibus
factorial ANOVA on clr-transformed data to investigate the effect of peak (28 levels,
within-subjects), milking interval (7 levels, between-subjects), and location (2 levels,
between-subjects) on relative peak area. Because the relative peak areas of the 28
analyzed peaks never summed to 100% of total peak area for any of the cases, we
obtained the 28-part subcomposition for each case by dividing each part by the sum of
the 28 parts for that case (Pawlowsky-Glahn and Egozcue, 2006). The model yielded two
significant two-way interactions: peak x milking interval (F92.44,862.74 = 1.83, p < 0.001,
partial η2 = 0.16), interpreted as asynchronous venom regeneration, and peak x location
(F15.41,862.74 = 3.33, p < 0.001, partial η2 = 0.06), interpreted as geographic variation. As
explained for the coarse analysis of chromatographic regions, we ran two-way (location x
interval) ANOVAs for each peak separately, with graphs of clr-transformed data by peak
presented in Figure 7. Twelve of the 28 peaks (~42%) revealed a significant main effect
of milking interval or interaction (Table 10), suggesting asynchronous venom
regeneration. Nine of the 28 peaks (~32%) showed a significant main effect of location or
an interaction, suggesting geographic variation. Nonparametric Kruskal-Wallis ANOVAs
with milking interval as the independent variable yielded identical results (not shown) for
20 of 26 peaks lacking an interaction, again supporting our interpretation from parametric
tests that asynchronous venom regeneration existed.
152
153
154
155
156
Discussion
This investigation of S. polymorpha represents the most comprehensive study to
date of venom regeneration in any centipede. We begin by discussing venom extraction,
trends in venom regeneration across trials, and variation between trials. We then consider
factors influencing regeneration and make comparisons to venom regeneration in other
venomous animals. Lastly, we discuss asynchronous regeneration of venom components
and infer some functional aspects of venom regeneration.
Electrical Milking
Although we assume our milking procedure fully depleted the venom supply, we
may not have done so for every centipede. Venom milking may result in incomplete
emptying of the venom glands, as noted in honeybees (Owen, 1978), spiders (Malli et al.,
2000), ants (Blum and Callahan, 1960; Haight, 2002), and snakes (Kochva et al., 1982;
McCleary and Heard, 2010). Even so, because we consistently employed the same
milking effort, the data reflect regeneration from a similar state of “emptiness” even if
glands were not absolutely devoid of venom.
Venom Regeneration During the 48-Hr and 14-Day Trials
In the 48-hr trial, both percent volume and percent protein mass regenerated
increased linearly as the milking interval increased (Fig. 2); however, protein mass
regeneration lagged behind volume. Both volume and protein mass regeneration slowed
two days after initial milking, as no significant differences existed among milking
intervals in the 14-day trial. The phenomenon of venom protein regeneration lagging
behind volume regeneration appears to be common, as it has been observed in spiders
157
(Boeve et al., 1995; Perret, 1977b), scorpions (Nisani et al., 2007), and some snakes
(Brown et al., 1975; Klauber, 1997b; Kochva, 1960; Schenberg et al., 1970; Willemse et
al., 1979). Exceptions, however, were reported for two snakes, the puff adder (Bitis
arietans; Currier et al., 2012) and the rhinoceros horned viper (B. nasicornis; Marsh and
Glatston, 1974), for which venom protein concentration returned to its initial milking
level within one and two days, respectively. Differences among taxa could result from
methodological rather than species differences, especially if some researchers achieved
greater gland depletion than others.
Although venom volume was more quickly replenished than venom protein (Fig.
2), the protein concentration remained consistent over time within each trial; in the 48-hr
trial, concentration remained at 60 µg/µL (Arizona) and 80 µg/µL (California),
respectively, while in the 14-day trial concentration remained at 79 µg/µL. This may be
because a larger proportion of fluid relative to protein moved back into the secretory
body within the first 8 hours of emptying, and then replenishment of both volume and
protein mass proceeded at similar rates thereafter (Fig. 2). Protein concentration of
regenerated venom was higher for California S. polymorpha than Arizona animals at each
milking interval (Fig. 3); the same relationship was observed for initial venom. It would
be premature to speculate why this might be the case.
The percentage of venom volume regenerated at 48 hours differed between the
48-hr trial (65 ± 6%) and the 14-day trial (86 ± 10%). The same was also true for protein
mass regenerated (29 ± 3 vs. 47 ± 9%, respectively; Fig. 2). No obvious explanation
exists for these differences. Differences between the two trials may stem from differences
in duration of captivity, or in the composition of the groups with regard to geographic
158
origin of subjects, but neither of these variables was a significant predictor of volume or
protein regeneration within the 48-hr or 14-day trials. Body lengths were also similar for
the 48-hr and 2-day groups (8.7 ± 0.4 vs. 8.9 ± 0.4 cm, respectively). As the 48-hr and
14-day trials were conducted in January and July, respectively, a seasonal effect might
exist. According to aktograph studies, S. polymorpha demonstrates circadian control over
its activity (Cloudsley-Thompson and Crawford, 1970), but the existence or persistence
of seasonal cycles in activity under invariant light/dark and temperature conditions has
not been investigated. Because all centipedes were in captivity at least seven weeks prior
to milking, and were maintained under invariant lighting and temperature conditions, an
effect of season seems unlikely unless an endogenous rhythm exists.
7-Month Follow-Up Study
Although volume and protein mass regeneration slowed two days after initial
milking, the mean values for percent volume and percent protein mass regenerated in the
14-day trial (86% and 46%, respectively) were not asymptotes, as they increased to 93%
and 76%, respectively, at the 7-month follow-up. These differences suggest that while
volume regenerates more quickly than protein mass during the initial 14 days after
milking, the reverse is true thereafter. This pattern would be expected of a venom gland
nearing repletion but undergoing continued protein synthesis. The 7-month multiple
regression analyses indicated that, unlike in the 48-hr and 14-day trials, body length was
not a useful predictor of venom regeneration over this extended period of time.
Several possibilities could explain why venom volume and protein mass did not
return to initial levels within 7 months. First, venom regeneration following complete
emptying of the glands may require more than 7 months, which seems unlikely. Second,
159
the decline could have resulted from senescence, but multiple regression analyses
indicated that time in captivity (48-hr trial: range 53–968 days; 14-day trial: 81–414
days) was not a useful predictor of venom regeneration. Furthermore, time in captivity
was not a significant predictor of venom volume yield or protein concentration (Cooper
et al., 2014). Third, electrical milking may damage the venom glands or venom gland
musculature, as proposed by other investigators. Among spiders, Argiope bruennichi
failed to yield venom after a single milking (Friedel and Nentwig, 1989), and volume
yield declined with additional milkings of Agelenopsis aperta (Kristensen, 2005). Sissom
et al. (1990) suggested that scorpions can only be milked, on average, four times before
the muscles of the gland stop responding to electrical stimulation. In some cases,
electrical milking may even kill the animal (Sahayaraj et al., 2006). In contrast, repeated
electrical venom extractions did not reduce yield in the spider Coremiocnemis tropix
(Herzig, 2010), the scorpion Hadrurus arizonensis (Fox et al., 2009), or in snakes (Marsh
and Whaler, 1984; McCleary and Heard, 2010). Presumably, the potential for damage to
an animal or its tissues increases with larger voltage-current combinations, and varies
depending on where the shock is applied. Our experience indicated that CO2anesthetization and electrical venom extraction were neither fatal nor overtly detrimental
to the centipedes. However, results of a separate study (Cooper et al., 2014) using S.
polymorpha showed that previous milking was associated with decreased venom volume
yield and protein concentration. A comparison of venom yields between our experimental
group of animals and a control group of animals never milked until the 7-month followup would have been informative.
160
Factors Influencing Venom Regeneration
Multiple regression models for the 48-hr trial revealed two primary predictors of
venom regeneration: milking interval and body length. Milking interval had the greatest
relative influence on percent venom volume and percent protein mass regenerated, as
discussed above. Centipede body length was also a significant predictor but, intriguingly,
was negatively associated with venom regeneration in both the 48-hr and 14-day trials.
Body length is positively associated with venom yield (Cooper et al., 2014), but it
remains unclear why venom regeneration is slower for longer, presumably older,
centipedes. Perhaps the ratio of venom secretory cells to venom storage volume (cf.
"venom vacuoles", Dugon and Arthur, 2012a; "extracellular space", Rosenberg and
Hilken, 2006; "secretory body", Undheim and King, 2011) decreases with centipede
length, with the result that the capacity to regenerate venom does not increase as fast as
volume for venom storage.
Several significant predictors of venom volume yield (relative forcipule length,
relative body mass, geographic origin) and protein concentration (geographic origin)
from a prior study (Cooper et al., 2014) were not significant predictors in the models for
regeneration of volume and protein mass. Although relative body mass may be an
indicator of increased energy reserves, and thus of increased fitness (Jakob et al., 1996;
Schulte-Hostedde et al., 2005), the lack of a relationship between regeneration and
forcipule length and body mass reinforces our conclusion that body size, beyond that
predicted by body length, does not influence venom regeneration. That geographic origin
was not a significant predictor of regeneration, combined with the fact that California
animals yielded larger volumes of venom with higher protein concentration (Cooper et
al., 2014), suggests that absolute rates of volume and protein mass regeneration must be
161
greater for California S. polymorpha. Perhaps the two populations experience differences
in selection pressures related to feeding frequency or predation intensity.
Venom Regeneration: Comparison with Other Venomous Animals
Direct comparisons of rates of venom regeneration among different taxa are
limited by factors such as variation in venom composition and size of venom gland(s),
and can be complicated by studies in which the potential effects of repeated milkings may
confound interpretation of regeneration (e.g., Currier et al., 2012; Klauber, 1997b;
Kochva, 1960; Perret, 1977b). Even so, comparison of S. polymorpha to other animals
provides a contextual framework for understanding the dynamics of venom regeneration.
Venom regeneration in S. polymorpha occurred rapidly during the first 48 hours
(volume: 65–86% of initial milk; protein mass: 29–47%) and then plateaued during days
2–14. Initial volume regeneration (at 48 hours) was relatively rapid compared to some
venomous animals, including tarantulas (Dugesiella hentzi, 50% in 2-3 days;
Aphonopelma chalcodes, 50% in 3–7 days; Perret, 1977b), the Egyptian cobra (Naja haje
annulifera, 70% in 21 days; Kochva et al., 1982), and several rattlesnakes (C. durissus,
69–80% in 25–32 days; De Lucca et al., 1974; George, 1930; Crotalus atrox, 73% in 38
days; C. viridis, 50% in 28 days; Klauber, 1997). However, volume regeneration in S.
polymorpha early on was slower than for the wasp Bracon brevicornis (100% in <3
hours; Beard, 1971), the spider Cupiennius salei (57% in 24 hours; Boeve et al., 1995),
and the scorpion Parabuthus transvaalicus (100% in 3–8 days; Nisani et al., 2007, 2012).
At 14 days, venom volume regeneration was similar for S. polymorpha (86%) as for the
ant Harpegnathos saltator (80–95%; Haight, 2012). While S. polymorpha had faster
regeneration early on, the tarantulas more quickly reached the upper end of regeneration,
162
with >90% volume regeneration in 28 days (A. chalcodes) and nearly 100% in 14 days
(D. hentzi) (Perret, 1977b). If it takes more than 7 months for S. polymorpha to achieve
complete volume regeneration, this is longer than required for complete regeneration in
several spider species: 1 week for Selenops mexicanus and Tegenaria atrica, 4 weeks for
Atrax robustus and Brachypelma albopilosum (Friedel and Nentwig, 1989), and 3–4
months for Pterinochilus (Freyvogel et al., 1968).
Compared to several species of snakes, the period of rapid venom regeneration in
S. polymorpha appeared to be much briefer, with an apparent plateau reached much
sooner (2 days). In C. d. terrificus, the continual increase in venom (wet mass) was slow
from day 0 to day 10, and then became faster from day 10 to day 32, without yet
plateauing (De Lucca et al., 1974). In N. h. annulifera, venom regeneration (both wet
mass and protein mass) was linear during the first 20 days, and then approached a plateau
after 30 days (Kochva et al., 1982). In Daboia palaestinae, rapid regeneration was
followed by a plateau after about 16 days (Rotenberg et al., 1971).
Over the course of the 48-hr trial, venom protein concentration for S. polymorpha
remained the same at 60 µg/µL (Arizona) and 80 µg/µL (California), representing 39%
and 42% of initial protein concentration, respectively. Over the course of the 14-day trial,
the protein concentration remained consistent at 79 µg/µL (average for all specimens),
representing 50% of initial protein concentration. By comparison, protein concentration
reached 25% of initial concentration in 3 days in the scorpion P. transvaalicus (Nisani et
al., 2007), 27% and 40% of initial concentration on days 1 and 4 in the spider C. salei
(Boeve et al., 1995), and 78% in 15 days in the snake D. palaestinae (Brown et al., 1975).
163
Asynchronous Regeneration of Venom Components in 48-Hr Trial
Five of ten chromatographic regions and 12 of 28 peaks (Fig. 5) from
chromatograms from the 48-hr trial demonstrated a change in percent of total protein
(percent of total peak area) among milking intervals, suggesting that venom proteins are
regenerated asynchronously (or non-parallel) in S. polymorpha. A caveat to consider is
that a given chromatographic peak may represent several different co-eluting venom
proteins (Palagi et al., 2013; Wang et al., 2013). In this circumstance, the area of a peak
would reflect the summation of regeneration of multiple venom components whose
individual rates of regeneration could vary, potentially confounding a determination of
venom regeneration synchrony/asynchrony. However, even in the case of co-eluting
proteins, changes in relative peak areas among milking intervals are evidence of
asynchronous regeneration of suites of venom components. Asynchronous regeneration
of venom components has been observed in other taxa. In two tarantulas, venom
hyaluronidase levels were restored more slowly than total protein (Perret, 1977b). In the
scorpion P. transvaalicus, some of the "prevenom" constituents (parabutoxins, 25 kDa
group, and possibly the parakinins in the initial clear secretion from the gland) were
resynthesized rapidly (~4 days), whereas other components represented only in the
"venom" (long-chain neurotoxins present in the subsequent opaque secretion) took longer
(6–8+ days) to reappear (Nisani et al., 2012). Similarly, in the scorpion Tityus serrulatus,
smaller venom peptides among the major peptide peaks took longer to regenerate than the
larger peptides (Pimenta et al., 2003). There is also considerable evidence for
asynchronous regeneration of venom proteins in snakes (Guo et al., 2009; Luna et al.,
2009; Oron et al., 1978; Taylor et al., 1986; Willemse et al., 1979). For example, in
Bothrops jararaca, the detection of L-amino acid oxidase (LAAO), phospholipase A2
164
inhibitor, metalloproteinase inhibitor, and disintegrin in venom gland tissue was higher 4
and 7 days post-milking compared to unmilked controls, whereas metalloproteinase was
lower at 4 days but higher at 7 days post-milking (Luna et al., 2013). Furthermore,
phospholipase A2 was higher than controls at 4 days but undetectable at 7 days postmilking (Luna et al., 2013). A separate study on the venom of B. jararaca demonstrated
that 5’-nucleotidase activity regenerated slower than phosphodiesterase and caseinolytic
activities (Schenberg et al., 1970). In D. palaestinae, Brown et al. (1975) concluded that
the venom enzymes LAAO, phosphodiesterase, and benzoylarginine ethyl esterase were
secreted at independent rates during regeneration. In contrast, evidence suggests
synchronous regeneration of venom components in the common death adder
(Acanthophis antarcticus; Pintor et al., 2011) and in the African puff adder (B. arietans;
Currier et al., 2012). An asynchronous pattern of venom component regeneration may be
the result of differences in rates of synthesis, in rates and routes of intracellular transport,
or in rates of expulsion into venom storage areas (Brown et al., 1975).
Given the relatively small quantities of protein present in many of the individual
S. polymorpha venom samples, and the large number of samples that we
chromatographed, we did not attempt to identify proteins from individual venom
samples. We subsequently characterized venom proteins of a pooled venom sample using
matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI
TOF MS), and liquid chromatography mass spectrometry (LC-MS/MS) and MALDI
TOF/TOF MS/MS followed by BLASTp sequence similarity searching. This
investigation will be described in a separate paper. As more centipede venom
components become fully characterized and added to databases, future research utilizing
165
chromatography and mass spectrometry may identify and illuminate the time course of
regeneration of important proteins in the venom of S. polymorpha.
Implications of Venom Regeneration for Venom Extraction Protocols
Intervals for electrical venom extraction from scolopendromorphs reported in the
literature vary from 1 week (Rates et al., 2007; Yang et al., 2012, 2013) to 20 days (Liu et
al., 2012b), and up to 1 month (Malta et al., 2008; Rates et al., 2007). Unfortunately, as
neither volume nor protein mass regeneration reached 100% during our investigation, our
data do not permit us to specify a milking interval that allows for complete regeneration
in S. polymorpha. A milking interval of between 2–14 days would allow a researcher to
obtain a major percentage of initial volume; however, venom milked at intervals of 14
days or less are predicted to have reduced protein content in comparison to initially
milked venom. Furthermore, as venom protein components are regenerated
asynchronously, pools of venom derived from repeated milkings at short intervals may
show artificial enrichment of the more rapidly regenerated components in comparison to
initially milked venom. It must be stressed that the effects of repeated extractions on
venom yield and venom composition in centipedes remain largely unexplored.
Functional Aspects of Venom Regeneration
Natural selection should adjust the rate of venom regeneration by balancing the
benefits of venom availability (e.g., additional prey capture and effective defense) and the
costs of rapid replenishment (dedicating metabolism towards venom synthesis, including
catabolizing and mobilizing endogenous materials, upregulating genetic material, protein
synthesis and secretion, and venom storage; McCue, 2006). The selection pressures
166
influencing venom regeneration are likely intertwined with those acting on venom
capacity, composition, toxicity, and quantities delivered during stinging. Because so little
is known about scolopendromorph behavioral use of venom, including prey capture
frequency and amount of venom employed in stings (predatory or defensive), it is
difficult to estimate what constitutes a “normal” venom regeneration scenario in S.
polymorpha (e.g., how quickly the amount of venom in a single natural deployment is
regenerated). Although one of the goals of this investigation of venom regeneration was
to completely empty (as far a possible) the venom glands, such an exhaustive depletion of
venom may never take place under natural conditions. Scolopendrids have the capacity to
kill multiple prey in a short time (e.g., up to 10 T. molitor larvae in six hours;
Formanowicz and Bradley, 1987), presenting the possibility of repeated and perhaps
extensive predatory venom deployment. Likewise, considerable defensive venom
deployment may be inferred in the case of a large aggressor (4-year-old bull terrier) that
died following a centipede sting (McKeown, 1930). However, the amount of venom
expended in such attacks is unclear.
Given the large ratio of scolopendrid venom capacity (Cooper et al., 2014) to
invertebrate LD50s (Quistad et al., 1992; Rates et al., 2007; Yang et al., 2012), the
tolerance of S. polymorpha to a lengthy between-feeding interval (our animals typically
live for several years on a diet of a single house cricket every two weeks), and our data
showing substantial volume regeneration of venom over a few days following a drastic
reduction, we speculate that S. polymorpha seldom goes without prey for want of venom.
Dugon and Arthur (2012b) showed that, in the lab, the frequency of successful attacks
(i.e., seize and hold) on a relatively small prey, Gryllus assimilis, by S. s. mutilans at 6
167
hours after complete venom extraction (16%) was lower than for unmilked controls
(96%), but had nearly returned to control levels after 24 hours (92%). Accordingly, the
authors concluded that the centipedes produced enough venom in 24 hours to subdue this
small prey. Employing a much larger prey, Schistocera gregaria, revealed that while
successful attacks were more frequent after 48 hours (56%) than after 24 hours (8%),
success at 48 hours remained below that of the control group with replete glands (92%).
The authors suggested, in accordance with the venom-optimization hypothesis (Hostettler
and Nentwig, 2006; Wullschleger and Nentwig, 2002), that the centipedes may take into
account the amount of venom regenerated and the size of the prey before making a
“decision” to attack. Thus, at least for S. s. mutilans, the effect of depleted venom supply
on prey capture is relatively brief. In terms of defense, our data suggest that after a
massive depletion of venom, S. polymorpha’s window of increased vulnerability may last
days or weeks depending on the quantity and quality of venom necessary to deter typical
predators. While venom availability influences scolopendrid predatory behavior (Dugon
and Arthur, 2012b), it remains to be seen how venom availability may influence activity
levels, shelter-seeking, and other behaviors which might indicate that centipedes attempt
to reduce exposure to predators due to lack of venom.
Scolopendromorphs may rely to some extent on the powerful piercing and cutting
power of their sharp, heavily chitinized forcipules (Bücherl, 1971c; Jangi, 1984; Manton,
1977) and coxosternal tooth-plates (Manton, 1964) to subjugate prey without
envenomation. Likewise, even dry stings by scolopendromorphs (Bush et al., 2001;
Cornwall, 1916; Klingel, 1960) may have considerable deterrent power for defense.
Accordingly, although a regenerated venom supply may allow centipedes to subdue
168
larger prey more quickly and with less risk of retaliation, or to mount a more robust
defense, the extent to which centipedes are constrained by venom regeneration and how
that might influence behavior require further study.
Conclusions
Venom regeneration in S. polymorpha occurred most rapidly during the first 48
hours, with volume regenerating faster than protein mass. While venom volume was
largely regenerated after 14 days, protein content remained low. Even 7 months after the
second milking, neither venom volume nor protein mass was at its initial value,
indicating a long regeneration cycle or perhaps a negative impact of electrical extraction
on future yields. Body length was negatively associated with regeneration of both venom
volume and protein mass, but other factors such as geographic origin, time in captivity,
relative forcipule size, and relative body mass did not influence regeneration. In
comparison to venom regeneration in other animals, venom regeneration in S.
polymorpha did not stand out as extremely slow or extremely fast; while the initial rate of
regeneration was faster than in some animals, S. polymorpha was slower to approach
complete regeneration than many animals. When regenerated venom from different
milking intervals in the 48-hr trial was subjected to RP-FPLC, 42% of analyzed peaks
showed changes in percent of total protein, suggesting that venom proteins are
regenerated asynchronously. The timing of venom regeneration in S. polymorpha
indicates that venom extraction protocols in which scolopendromorphs are milked at
intervals of 14 days or less may result in the collection of venom with reduced protein
content relative to an initial milking. The considerable extent of venom regeneration
occurring within the first 48 hours, despite the reduced protein content, may imply that
169
prey capture and defensive capacity are seldom seriously constrained by the timing of
venom replacement.
Acknowledgments
The authors thank Dr. P. Duerksen-Hughes and her students in the LLU
Biochemistry Department, and Dr. U. Soto-Wegner in the LLU Basic Sciences
Department for access to their equipment. We are also grateful to Charles Kristensen of
Spider Pharm Inc. for his valuable instruction regarding venom extraction. This work was
supported by the Department of Earth and Biological Sciences, Loma Linda University.
Conflict of Interest Statement
The authors declare no conflicts of interest with respect to the authorship and/or
publication of this article.
Ethical Statement
with Loma Linda University’s Policy on the Care and Use of Laboratory Animals. All
authors have read the manuscript, agree to its publication in Zoology, and agree that it
has followed the rules of ethics presented in Elsevier’s Ethical Guidelines for Journal
Publication.
170
References
Abdel-Rahman, M.A., Omran, M.A.A., Abdel-Nabi, I.M., Ueda, H., McVean, A., 2009.
Intraspecific variation in the Egyptian scorpion Scorpio maurus palmatus venom
collected from different biotopes. Toxicon 53, 349-359.
Alami, M., Ouafik, L., Ceard, B., Legros, C., Bougis, P.E., Martin-Eauclaire, M.F., 2001.
Characterisation of the gene encoding the alpha-toxin Amm V from the scorpion
Androctonus mauretanicus mauretanicus. Toxicon 39, 1579-1585.
52.
Beard, R.L., 1971. Production and use of venom by Bracon brevicornis (Wesm.). In: de
Vries, A., Kochva, E. (Eds.), Toxins of animal and plant origin. Gordon &
Breach, New York, pp. 181-190.
Billen, J., 1990. A survey of the glandular system of fire ants. In: Vander Meer, R.K.,
the insecticidal component. In: Pavan, M., Eisner, T. (Eds.), XI. Internationaler
248-254.
171
Brown, R.S., Brown, M.B., Bdolah, A., Kochva, E., 1975. Accumulation of some
secretory enzymes in venom glands of Vipera palaestinae. Am. J. Physiol. 229,
1675-1679.
Bücherl, W., 1971. Venomous chilopods or centipedes. In: Bucherl, W., Buckley, E.E.,
169-196.
Hillsdale, NJ.
Cooper, A.M., Fox, G.A., Nelsen, D.R., Hayes, W.K., 2014. Variation in venom yield
and protein concentration of the centipedes Scolopendra polymorpha and
Scolopendra subspinipes. Toxicon
http://dx.doi.org/10.1016/j.toxicon.2014.02.003.
Natn. Mus. 111, 167-195.
Crawford, C.S., 1990. Scorpions, Solifugae and associated desert taxa. In: Dindal, D.L.
Crawford, C.S., Riddle, W.A., Pugach, S., 1975. Overwintering physiology of centipede
Scolopendra polymorpha. Physiol. Zool. 48, 290-294.
172
da Silva, N.J., Aird, S.D., 2001. Prey specificity, comparative lethality and compositional
differences of coral snake venoms. Comp. Biochem. Physiol. C-Toxicol.
Pharmacol. 128, 425-456.
Daltry, J.C., Wuster, W., Thorpe, R.S., 1997. The role of ecology in determining venom
variation in the Malayan pitviper, Calloselasma rhodostoma. Symp. Zool. Soc.
Lond. 70, 155-171.
De Lucca, F.L., Haddad, A., Kochva, E., Rothschi.Am, Valeri, V., 1974. Protein
synthesis and morphological changes in secretory epithelium of venom gland of
Crotalus durissus terrificus at different times after manual extraction of venom.
Toxicon 12, 361-368.
Boubee, Paris.
173
Filzmoser, P., Hron, K., Reimann, C., 2009. Univariate statistical analysis of
environmental (compositional) data: problems and possibilities. Sci. Total
Environ. 407, 6100-6108.
Freyvogel, T.A., Honegger, C.G., Maretic, Z., 1968. On the biology and toxicity of the
East African spider Pterinochilus spec. Acta Trop. 25, 217-55.
Frund, H., 1992. The occurence and frequency of scars in centipedes. In: Meyer, E.,
pp. 269-275.
Galindo, J., Galindo, G., Irwin, J.T., 2009. The effects of temperature on metabolic rate,
venom synthesis and potency in Peucetia viridans (Araneae: Oxyopidae). Integr.
Comp. Biol. 49, E231-E231.
George, I.D., 1930. Notes on the extraction of venom at the serpentarium of the
Antivenin Institute at Tela, Honduras. Bull. Antivenin Inst. Am. 4, 57-59.
Guo, Y.W., Liu, H.W., Lin, Y.C., Liang, M.C., 2009. Non-parallel expression of a
triflavin-like disintegrin venom protein in the main glands of Trimeresurus
mucrosquamatus. Toxin Rev. 28, 266-270.
Hadley, N.F., Stuart, J.L., Quinlan, M., 1982. An air-flow system for measuring total
transpiration and cuticular permeability in arthropods: studies on the centipede
174
Haight, K.L., 2012. Patterns of venom production and temporal polyethism in workers of
Jerdon's jumping ant, Harpegnathos saltator. J. Insect Physiol. 58, 1568-1574.
assumptions, and evidence. In: Hayes, W.K., Beaman, K.R., Cardwell, M.D.,
241-249.
20, 369-375.
Jangi, B.S., 1984. Centipede venoms and poisoning. In: Tu, A.T. (Ed.), Handbook of
Kaire, G.H., 1963. Observations on some funnel-web spiders (Atrax species) and their
venoms, with particular reference to Atrax robustus. Med. J. Aust. 2, 307-311.
Klauber, L.M., 1997. Rattlesnakes: their habits, life histories, and influence on mankind.
Trop. Med. Hyg. 9, 381-390.
175
303-311.
Liang, Y.Z., Xie, P.S., Chau, F., 2010. Chromatographic fingerprinting and related
chemometric techniques for quality control of traditional Chinese medicines. J.
Sep. Sci. 33, 410-421.
Wu, Y., Ding, J.P., Lee, W.H., 2012. Venomic and transcriptomic analysis of
Luna, M.S., Valente, R.H., Perales, J., Vieira, M.L., Yamanouye, N., 2013. Activation of
Bothrops jararaca snake venom gland and venom production: a proteome
approach. J. Proteomics http://dx.doi.org/10.1016/j.jprot.2013.10.026.
Luna, M.S.A., Hortencio, T.M.A., Ferreira, Z.S., Yamanouye, N., 2009. Sympathetic
outflow activates the venom gland of the snake Bothrops jararaca by regulating
the activation of transcription factors and the synthesis of venom gland proteins. J.
Exp. Biol. 212, 1535-1543.
Malmquist, G., Danielsson, R., 1994. Alignment of chromatographic profiles for
principal component analysis: a prerequisite for fingerprinting methods. J.
Chromatogr. 687, 71-88.
176
Toxicon 55, 250-255.
Copeia 2006, 818-825.
Morphol. 206, 303-312.
Mundel, P., 1990. Chilopoda. In: Dindal, D.L. (Ed.), Soil biology guide. John Wiley &
177
Nakagawa, S., Cuthill, I.C., 2007. Effect size, confidence interval and statistical
significance: a practical guide for biologists. Biol. Rev. 82, 591-605.
Nisani, Z. 2008. Behavioral and physiological ecology of scorpion venom expenditure:
stinging, spraying, and venom regeneration [Ph.D. dissertation]. Loma Linda,
California: Loma Linda University.
Physiol. 147, 509-513.
Oron, U., Kinamon, S., Bdolah, A., 1978. Asynchrony in the synthesis of secretory
proteins in the venom gland of the snake Vipera palaestinae. Biochem. J. 174,
733-739.
Palagi, A., Koh, J.M.S., Leblanc, M., Wilson, D., Dutertre, S., King, G.F., Nicholson,
G.M., Escoubas, P., 2013. Unravelling the complex venom landscapes of lethal
Australian funnel-web spiders (Hexathelidae: Atracinae) using LC-MALDI-TOF
mass spectrometry. J. Proteomics 80, 292-310.
Pawlowsky-Glahn, V., Egozcue, J.J., 2006. Compositional data and their analysis: an
introduction. In: Buccianti, A., Mateu-Figueras, G., Pawlowsky-Glahn, V. (Eds.),
Compositional data analysis in the geosciences: from theory to practice.
Geological Society of London, London, pp. 1-10.
Perret, B.A., 1977. Venom regeneration in tarantula spiders—I: analysis of venom
Physiol. 56, 607-613.
Pimenta, A.M.C., Almeida, F.D., de Lima, M.E., Martin-Eauclaire, M.F., Bougis, P.E.,
2003. Individual variability in Tityus serrulatus (Scorpiones, Buthidae) venom
elicited by matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry. Rapid Commun. Mass Spectrom. 17, 413-418.
178
Punzo, F., 2000. Desert arthropods: life history variations. Cloudsley-Thompson, J.L.,
editor. Springer, Berlin.
Rotenberg, D., Bambereger, E.S., Kochva, E., 1971. Studies on ribonucleic acid synthesis
in venom glands of Vipera palaestinae (Ophidia, Reptilia). Biochem. J. 121, 609612.
Schenberg, S., Pereira-LIma, F.A., Nogueira-Schiripa, L.N., Nagamori, A., 1970.
Unparalleled regeneration of snake venom components in successive milkings.
Toxicon 8, 152.
Schilling, K., Oberdick, J., Schilling, R.L., 2012. Toward an efficient and integrative
analysis of limited-choice behavioral experiments. J. Neurosci. 32, 12651-12656.
179
Martinsville.
Sissom, W.D., Polis, G.A., Watt, D.D., 1990. Field and laboratory methods. In: Polis,
445-461.
Taylor, D., Iddon, D., Sells, P., Semoff, S., Theakston, R.D.G., 1986. An investigation of
venom secretion by the venom gland cells of the carpet viper (Echis carinatus).
Toxicon 24, 651-659.
Tennant, A., 1985. A field guide to Texas snakes. Gulf Publishing, Houston.
Tomasi, G., Savorani, F., Engelsen, S.B., 2011. icoshift: an effective tool for the
alignment of chromatographic data. J. Chromatogr. 1218, 7832-7840.
Ulbrich, H.F., 2011. Graphing and communicating compositional data in high
dimensions. In: Egozcue, J.J., Tolosana-Delgado, R., Ortego, M.I. (Eds.),
Proceedings of the 4th international workshop on compositional data analysis
(CoDaWork 2011), Girona, Spain, 9-13 May 2011.
Uzenbaev, S.D., Lyabzina, S.N., 2009. An experimental study of the effects of spider
venom on animals. Entomol. Rev. 89, 479-486.
Okol. Tiere 36, 645-650.
Voigtlander, K., 2011. Chilopoda—ecology. In: Minelli, A. (Ed.), The Myriapoda. Brill,
Leiden, pp. 309-325.
London.
180
Wang, H.Y., Zhang, F., Li, D., Xu, S.Y., He, J., Yu, H., Li, J.Y., Liu, Z.H., Liang, S.P.,
2013. The venom of the fishing spider Dolomedes sulfurous contains various
neurotoxins acting on voltage-activated ion channels in rat dorsal root ganglion
neurons. Toxicon 65, 68-75.
Willemse, G.T., Hattingh, J., Karlsson, R.M., Levy, S., Parker, C., 1979. Changes in
composition and protein concentration of puff adder (Bitis arietans) venom due to
frequent milking. Toxicon 17, 37-42.
Zar, J.H., 1996. Biostatistical analysis, 3rd ed. Prentice Hall, Upper Saddle River.
Zhang, Z.M., Chen, S., Liang, Y.Z., 2011. Peak alignment using wavelet pattern
matching and differential evolution. Talanta 83, 1108-1117.
181
CHAPTER FOUR
PARTIAL CHARACTERIZATION OF THE VENOM PROTEOME OF
THE CENTIPEDE SCOLOPENDRA POLYMORPHA
Allen M. Coopera,*, Wayne J. Kellna, William K. Hayesa
a
Allen M. Cooper
Griggs Hall #101
24941 Stewart St.
USA
Tel: 1-707-225-5018
Fax: 1-909-558-0259
182
Abstract
Applying proteomic tools to the investigation of venom composition is an
important preliminary step toward understanding venom’s biological functions and
potential applied uses. Recent efforts to characterize centipede venoms utilizing
proteomic and transcriptomic methods have revealed these venoms as potential leads for
the development of new drugs and bioinsecticides. Despite recent progress in
characterizing the centipede venom proteome, much remains to be learned about
centipede venoms. We characterized the venom composition of Scolopendra
polymorpha, a centipede widely distributed across the western United States and into
northern Mexico, using SDS-PAGE, RP-FPLC, LC-MS/MS, MALDI TOF MS, and
BLASTp amino acid sequence similarity searching. We further conducted an
intraspecific comparison of venoms from Arizona and California animals. In the first
investigation of the venom of this species, we demonstrate that the venom is complex,
generating up to 23 bands by SDS-PAGE and 56 peaks by RP-FPLC. MALDI TOF MS
analyses of fractionated venom revealed hundreds of components with masses ranging
from 1,014.5 Da to 82,863.9 Da. The distribution of molecular masses was skewed
toward smaller peptides and proteins, with 72% of components found below 12 kDa.
However, there was also a prevalent group of proteins between 20 kDa and 26 kDa,
corresponding to about 9% of all components. Arizona and California S. polymorpha
venoms were largely similar, with subtle, primarily quantitative differences apparent by
RP-FPLC and MALDI TOF MS. Although Mascot searching following both LC-MS/MS
and MALDI TOF/TOF MS/MS yielded no significant centipede- or venom toxin-related
protein hits, BLASTp sequence similarity searching of MS/MS-derived amino acid
183
sequences demonstrated 20 different sequences with similarity to known venom
components, including serine proteases, Kv channel inhibitors, PLA2, CRISP-like
proteins, a Cav channel activator, and several putative neurotoxins of unknown function.
Given the complexity of the venom and the number of RP-FPLC fractions yielding no
similarity matches, the venom of S. polymorpha promises to yield many novel protein
components.
184
Introduction
The diverse group of terrestrial arthropods known as centipedes comprises the
class Chilopoda, part of the subphylum Myriapoda, and is divided into five living orders
(and 1 extinct): Scutigeromorpha, Lithobiomorpha, Craterostigmomorpha,
Scolopendromorpha, and Geophilomorpha (Edgecombe and Giribet, 2007). Centipedes,
among the oldest extant terrestrial arthropods (Murienne et al., 2010; Shear and
Edgecombe, 2010), serve an ecologically important role as soil and leaf litter predators
(Albert, 1983; Robertson et al., 1994; Wallwork, 1976). Despite their importance, our
knowledge of the natural history of centipedes is very limited (Forti et al., 2007; Molinari
et al., 2005), perhaps due in part to their cryptic, and often nocturnal, lifestyles (Lewis,
1981). Anatomically, centipedes are characterized by a long, segmented body with one
pair of legs per segment, and a head containing a pair of long antennae (Lewis, 1981).
The legs of the first trunk segment are modified to form the characteristic forcipules that
are used to grasp and envenomate prey (Bonato et al., 2010; Hayden and Arthur, 2013;
Lewis, 1981; Undheim and King, 2011). The forcipules are also employed in defense
(Davis, 1993; Demange, 1981; Maschwitz et al., 1979; Neck, 1985). Prey immobilization
(Undheim and King, 2011) comprises the primary role of the relatively complex venom
(Liu et al., 2012a; Liu et al., 2012b; Undheim and King, 2011); however, a digestive
function has also been suggested (Jangi, 1984; Martin, 1971; Minton, 1974), but remains
unclear (Bücherl, 1971a; Dugon and Arthur, 2012b).
As with the initial study of other venomous animals, a large proportion of the
literature on centipedes has focused on the effects of venom on non-human animals
(Kimura et al., 2013; Undheim and King, 2011) and on the symptoms (Othong et al.,
2012; Undheim and King, 2011) and treatment (Bush et al., 2001; Chaou et al., 2009;
185
Haddad et al., 2012) of human envenomations. However, following a similar historical
trajectory as spider venom toxinology (Escoubas et al., 2006; Palagi et al., 2013), there is
a growing movement to elucidate the venom proteomes of centipedes (Liu et al., 2012a;
Liu et al., 2012b; Peng et al., 2010; Rates et al., 2007; Undheim et al., 2012; Yang et al.,
2012; Yang et al., 2013), spurred in part by the potential of centipede venom components
as potential leads for developing new drugs and bioinsecticides (Bhagirath et al., 2006;
Hou et al., 2013; Kong et al., 2013; Kwon et al., 2013; Peng et al., 2010; Schwartz et al.,
2012; Smith et al., 2013; Xu et al., 2013; Yang et al., 2012; Yang et al., 2013). If these
efforts are rewarded as they have been for the study of other venomous organisms,
including cone snails, hymenopterans, spiders, scorpions, and snakes, then this close
examination of centipede venom composition promises to unlock the roles of venom in
prey incapacitation (Kuhn-Nentwig et al., 2011; Lewis et al., 2012; Mackessy, 2010) and
defense (Casewell et al., 2013; Li et al., 2013), aid in understanding phylogenetic and
taxonomic relationships (Calvete et al., 2007a; Souza et al., 2008; Tashima et al., 2008),
facilitate means of countering the deleterious effects of human envenomation (Calvete,
2011, 2013; de Graaf et al., 2009; Espino-Solis et al., 2009; Gutierrez et al., 2009;
Oukkache et al., 2008), and provide a source of bioactive molecules with insecticidal
(Chaim et al., 2011; King and Hardy, 2013; Schwartz et al., 2012; Smith et al., 2013;
Yang et al., 2012) and therapeutic potential (Almaaytah and Albalas, 2014; Baron et al.,
2013; Brady et al., 2013; de Oliveira et al., 2013; Fox and Serrano, 2007; Georgieva et
al., 2008; King, 2011; McCleary and Kini, 2013; Prashanth et al., 2012; Santos et al.,
2011; Yang et al., 2013).
186
The relatively late development of intensive centipede venom research likely
relates to several factors. First, centipedes are often nocturnal and spend much time
hidden under logs, leaf litter, and stones (Lewis, 1981), or within small burrows (Davis,
1993), and thus frequently go unnoticed by humans. Even when noticed, the speed
(Manton, 1977) and agility (Cloudsley-Thompson, 1955; Jangi, 1984; Molinari et al.,
2005; Remington, 1950; Vijayakumar et al., 2012) of centipedes often prevents direct
interaction with humans. Second, even when human-centipede interactions do occur, only
a few of the large taxa of centipedes are capable of causing serious envenomations (Balit
et al., 2004; Haddad et al., 2012; Medeiros et al., 2008). Given the diversity of other
venomous creatures that can deliver life-threatening bites or stings (Balhara and Stolbach,
2014; Borges et al., 2012; Chippaux and Goyffon, 2008; Cruz et al., 2009; Forrester et
al., 2012; Kuruppu et al., 2008; Mebs, 2002; Weinstein et al., 2013), it’s not surprising
that this relatively low-threat group of animals (Chaou et al., 2009; Haddad et al., 2012;
Othong et al., 2012) is only now receiving attention. Third many investigators may not
have previously had easy access to the larger and more medically significant species,
many of which are tropical in origin (Adis, 2002; Bush et al., 2001; Jangi, 1984; Lewis,
1981). Finally, difficulty collecting sufficient amounts of venom for analysis has often
proven challenging (Gonzalez-Morales et al., 2009; Rates et al., 2007; Stankiewicz et al.,
1999; Yang et al., 2012). Venom yield from S. polymorpha (mean body length 8.8 cm),
for example, averaged only 1.1 µL (Cooper et al., 2014). Analytical resources appropriate
for investigating such small amounts of material have become available only recently
(Escoubas et al., 2006). Modern separation techniques coupled with high sensitivity mass
spectrometry are opening up new opportunities to characterize venoms of many smaller
187
animals, including centipedes. Even so, investigations of centipede venoms employing
multiple separation steps prior to mass spectrometry and/or functional assays often
require venom from hundreds to thousands of animals (Liu et al., 2012b; Peng et al.,
2010; Yang et al., 2012) to generate a sufficient sample, a labor-intensive process.
The known constituents of centipede venoms were reviewed by Undheim and
King (2011), and include acid and alkaline phosphatases, phospholipase A2 (PLA2),
esterases, hyaluronidases, metalloproteases, non-metalloproteases, cardiotoxins, CRISPs,
disintegrins, haemolysins, myotoxins, neurotoxins, histamine, and serotonin. Since this
review, proteomic and transcriptomic investigations have reported the presence of more
than 500 venom components, with pharmacological properties including voltage-gated
sodium, potassium, and calcium channel activity, anticoagulant activity, PLA2 activity,
and trypsin inhibition activity (Liu et al., 2012b; Yang et al., 2012). Venom component
masses reportedly range from ~7 kDa to >200 kDa by SDS-PAGE (Gutierrez et al., 2003;
Liu et al., 2012b; Malta et al., 2008) and from 554.3 Da (Kong et al., 2013) to 22,559.3
Da (Liu et al., 2012b) by mass spectrometry. In general, centipede venoms appear to
differ from the venoms of other arthropods in the abundance of high molecular weight
components (Undheim et al., 2012). Additionally, many centipede venom components
characterized thus far are novel (Liu et al., 2012b; Rates et al., 2007; Undheim et al.,
2012; Yang et al., 2012; Yang et al., 2013). Recently, Yang et al. (2013) demonstrated
that the peptide Ssm6a from S. subspinipes mutilans was a highly selective inhibitor of
Nav1.7 channels, proving to be an effective analgesic in rodent pain models and thereby
showing promise for drug development.
188
Despite recent progress in characterizing the centipede venom proteome,
centipede venoms still represent a largely unexplored landscape. Only approximately
3,300 species of centipede out of an estimated 6,950 species have been described (Adis
and Harvey, 2000; Edgecombe and Giribet, 2007), and while all centipedes are venomous
(Dugon and Arthur, 2012a; Lewis, 1981), investigations of their toxins have focused
almost exclusively (cf. Undheim et al., 2012) on centipedes in the order
Scolopendromorpha. This is not surprising given that scolopendromorphs, ranging in
adult length from 1 cm to 30 cm (Edgecombe and Koch, 2008), are the largest (Mundel,
1990), most fiercely predatory (Edgecombe and Koch, 2008), and most medically
important (Balit et al., 2004; Jangi, 1984) of all centipedes. Even so, of the estimated 800
species in the order Scolopendromorpha, only slightly more than 600 have been
described (Adis and Harvey, 2000), and the venoms of less than a dozen species have
received much study (Liu et al., 2012b; Undheim et al., 2012; Undheim and King, 2011;
Yang et al., 2012). Large-scale characterizations of venom components employing
transcriptomic and/or proteomic methods have been limited to Cormocephalus
westwoodii, Ethmostigmus rubripes, S. alternans, S. angulata, S. morsitans, S.
subspinipes dehaani, S. s. mutilans, and S. viridicornis (Liu et al., 2012b; Undheim et al.,
2012; Yang et al., 2012). The transcriptomic investigation of S. s. dehaani venom alone
revealed 543 different proteins (Liu et al., 2012b). If the number of proteins in S. s.
dehaani venom is any indication of the complexity of centipede venoms in general, then
coupling this complexity with measures of centipede taxonomic diversity underscores the
potentially enormous pharmacological resource present in centipede venoms.
189
For this investigation, we used the centipede S. polymorpha (Chilopoda:
Scolopendromorpha: Scolopendridae). This centipede inhabits desert, dry grassland, and
forest habitats of the southern half of the United States from the Great Plains westward to
California, ranging up the Pacific states into Oregon, and throughout the desert southwest
into northern Mexico (Crabill, 1960; Crawford and Riddle, 1974; Shelley, 2002).
Scolopendra polymorpha is the most abundant and widespread scolopendrid centipede
throughout the desert southwest, and the most common centipede in the Chihuahuan
Desert region of west Texas (Maldonado, 1998). Relatively little is known about the
ecology of this animal (Crawford, 1990; Wallwork, 1982), although the cold-hardiness
and overwintering physiology (Crawford and Riddle, 1974; Crawford et al., 1975),
circadian activity patterns (Cloudsley-Thompson and Crawford, 1970), burrowing
behaviors (Davis, 1993), water loss (Hadley et al., 1982), and venom regeneration
(Cooper et al., submitted) have been investigated. The diet of S. polymorpha includes a
variety of invertebrates and vertebrates (Formanowicz and Bradley, 1987; Punzo, 2000),
but has not been carefully documented in the field (Crawford, 1990). The sting of S.
polymorpha causes temporary sharp pain in humans (Baerg, 1924; Maldonado, 1998;
Turk, 1951).
To date, there have been no investigations of the venom composition of S.
polymorpha. In our efforts to lay the groundwork for greater understanding of the
biological roles and potential applied uses of the venom of this abundant and widely
distributed species, we designed the present study to characterize venom components
using reversed-phase fast protein liquid chromatography (RP-FPLC) combined with
liquid chromatography mass spectrometry (LC-MS/MS) and matrix assisted laser
190
desorption/ionization time-of-flight mass spectrometry (MALDI TOF MS), as well as
amino acid sequence similarity searching. In addition, we aimed to compare venom from
S. polymorpha originating from two separate locations, Arizona and California. We
demonstrate that the venom of S. polymorpha is very complex, comprising hundreds of
proteins, some of which show similarity to reported serine proteases, ion-channel
activators/inhibitors, and neurotoxins.
Centipedes
We acquired S. polymorpha centipedes from Arizona (n = 49, collected from
Cochise County, AZ; Bugs of America LLC, Portal, AZ, USA) and southern California
(n = 55, collected from San Bernardino, Riverside, and San Diego counties, CA).
Centipedes were housed individually in plastic containers with a mixed sand/soil
substrate and water dish. They were offered a cricket once every two weeks, and misted
weekly.
Venom Extraction and Venom Pools
We procured venom from S. polymorpha following the methods of Cooper et al.
(2014). Briefly, we extracted venom by electrical stimulation applied to the bases of the
forcipules and collected venom in graduated 5-µL Drummond PCR Micropipets (PGC
Scientifics, Garner, NC, USA). Following Dass and Jangi (1978), centipedes were not fed
for 2 weeks prior to venom extraction to ensure, presumably, replete venom glands.
We created three venom pools for analyses. First, we obtained two separate
location-specific pools of venom, one from AZ specimens (n = 12 females and 14 males;
191
body length 7.0–11.1 cm; mean ± 1 SE and range of time in captivity prior to venom
extraction: 395 ± 65, 53–964 days), and one from CA specimens (n = 20 females and 16
males; 6.2–10.9 cm; 288 ± 31, 72–496 days). Individual samples were immediately
transferred to a microcentrifuge tube with 50 µL of chilled nanopure water, and stored at
-80 C. From a different group of animals, we collected a third, mixed-location venom
pool by combining samples from AZ (n = 10 females and 8 males) and CA (n = 8
females and 7 males) specimens (collectively 7.5–11.5 cm; 658 ± 62 days, 371–1282
days). We transferred the extracted venom from each individual into a single, common
microcentrifuge tube with 50 µL of chilled buffer A (0.065% TFA, 2% ACN in water),
and then centrifuged the pooled sample at 10,000 xg for 10 min, collected the
supernatant, and stored it at -80 C. In addition to these venom pools, we also analyzed
venom samples from individual specimens from AZ (n = 3 females and 2 males; 8.2–10.7
cm; 536 ± 205, 54–968 days) and CA (n = 2 females and 2 males; 6.3–8.6 cm; 284 ± 121,
72–496 days; Riverside and San Bernardino counties) using RP-FPLC. None of these
centipedes were subjected previously to milking procedures.
To prepare the location-specific samples for SDS-PAGE, we thawed each sample,
centrifuged the solution at 10,000 xg for 5 min, removed 1 µL of the venom solution
supernatant from each microcentrifuge tube and pooled these, according to location (AZ
or CA), in 300 µL of nanopure water. Both venom pools were then lyophilized. We
treated the mixed-location venom pool similarly, but added just 1 µL of the supernatant
to the 300 µL of nanopure water. We calculated the mass of protein in each venom pool
(section 2.3), and reconstituted the lyophilized venom pools with nanopure water to
approximately 11 µg/µL.
192
To prepare the location-specific samples for MALDI TOF MS (section 3.3), we
thawed and centrifuged the individual samples as before, but removed 2 µL of
supernatant from each microcentrifuge tube and pooled these in 168 µL (AZ) or 148 µL
(CA) of buffer A (0.065% TFA, 2% ACN in water). We treated the mixed-location
venom pool similarly, but placed 2 µL of the supernatant in 210 µL buffer A. Each
venom pool was de-salted by RP-FPLC. We equilibrated a RESOURCE RPC 1 ml
polystyrene/divinylbenzene reversed-phase column (GE Lifesciences, Pittsburgh, PA,
USA) in buffer A, and injected 100 µL (approximately 100 µg) of sample onto the
column. Proteins were eluted at a flow rate of 0.5 mL/min, in a 1-column-volume linear
gradient of 0–100% buffer B (0.05% TFA, 80% ACN in water), and the elution
monitored at 214 nm using Unicorn 5.0 (GE Lifesciences) software. We discarded the
flow-through and manually collected the eluted proteins in three 1-mL fractions. Protein
fractions were subsequently combined and lyophilized.
We determined protein concentration of individual venom samples (in nanopure
water) used to create the location-specific venom pools using the Coomassie blue dyebinding method (Bradford, 1976), with bovine serum albumin (BSA) in nanopure water
as the standard. We purchased Coomassie Protein Assay Reagent and BSA from Thermo
Fisher Scientific (Waltham, MA, USA). Using the manufacturer’s 1–25 µg/mL protocol,
all assays resulted in high coefficients of determination (r2 > 0.98), indicating reliability.
We measured protein concentration of the mixed-location venom pool (dissolved in
buffer A) by absorbance at 280 nm using a NanoDrop 2000 UV-Vis spectrophotometer
193
(Thermo Fisher Scientific), assuming an extinction coefficient (εpercent) of 10, per
manufacturer's recommendation.
Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDSPAGE)
We subjected three venom pools (AZ, CA, mixed-location; 50 µg protein each) to
SDS-PAGE using a NuPAGE 4–12% Bis-Tris gel (Life Technologies, Grand Island, NY,
USA) under reducing and non-reducing conditions following the manufacturer’s
protocol. After separation of proteins by electrophoresis, we stained the gel using
SimplyBlue SafeStain (Life Technologies). Novex Sharp Protein Standard (Life
Technologies) was used for molecular mass markers. We imaged the stained gel using a
BioSpectrum 500 Imaging System (UVP, Upland, CA, USA) and used VisionWorksLS
Image Acquisition and Analysis Software (ver. 6.8, UVP) for mass determination and
densitometric analysis.
RP-FPLC
We performed all separations on an ÄKTA FPLC (GE Lifesciences, Pittsburgh,
PA, USA). For fractionation of the mixed-location venom pool, we diluted 12 µL of
venom pool solution with 210 µL buffer A and centrifuged at 10,000 xg for 10 min. We
equilibrated a SOURCE 15 RPC ST 4.6/100 polystyrene/divinylbenzene reversed-phase
column (GE Lifesciences) in buffer A, and injected 100 µL of sample (approximately
596 µg protein) onto the column. Proteins were eluted at a flow rate of 0.5 mL/min, in a
40-column-volume linear gradient of 0–100% buffer B (0.05% TFA, 80% ACN in
water), and the elution monitored at 214 nm using Unicorn 5.0 (GE Lifesciences)
194
software. During separation of the mixed-location venom pool, we collected fractions
manually for further analysis. We also conducted RP-FPLC on the portions of the
individual venom samples (AZ and CA) remaining after protein quantification. Individual
venom samples (ranging between 25 µg and 402 µg protein, comprising all of the
available sample from each individual) were diluted with 175 µL buffer A, centrifuged as
described, and 100 µL of the diluted sample injected onto the column. Separation
involved the same column, buffers, and elution program as for the venom pool.
Pre-Treatment and Analysis of Individual Venom
Sample RP-FPLC Chromatograms
Because retention time/volume shifts can impair results of analyses using
chromatographic data sets (Liang et al., 2010; Malmquist and Danielsson, 1994; Tomasi
et al., 2011; Zhang et al., 2011), we aligned chromatograms from individual venom
samples using wavelet pattern matching and differential evolution (Zhang et al., 2011)
prior to integration and analysis according to methods already described (Cooper et al.,
submitted). We assumed that aligned peaks across samples contained the same
components; however, more detailed analyses (mass spectrometry or sequence analysis)
are necessary to verify this assumption.
We compared venom composition between Arizona (n = 5) and California (n = 4)
S. polymorpha by analyzing relative peak areas of individual RP-FPLC peaks in the
retention volume range 15.22–55.00 mL that were ≥1% of total peak area in at least one
sample. Relative peak area derived from monitoring eluate at the absorption wavelength
of the peptide bond (190–230 nm) is a reliable method for quantifying the relative
195
abundance of different venom components in the reversed-phase chromatogram (Calvete,
2013).
Mixed-Location Venom Pool Protein Characterization by MALDI
TOF MS, LC-MS/MS, MALDI TOF/TOF MS/MS, and BLASTp
Sequence Similarity Searching
We analyzed mixed-location venom pool RP-FPLC fractions for whole protein
molecular masses and protein identification/similarity. For whole protein molecular
masses, we lyophilized 50 µL of each fraction. We subjected the remaining volume of
fractionated venom to reduction and alkylation prior to enzymatic digestion using
dithiothreitol and iodoacetamide, respectively. Reduced and alkylated proteins were
digested with proteomics-grade porcine pancreatic trypsin (Sigma-Aldrich, St. Louis,
MO, USA). We purified the tryptic peptides with Zip TipC18 (EMD Millipore, Billerica,
MA, USA) according to manufacturer’s instructions, and then lyophilized them. Tryptic
peptides were then analyzed for protein identification using LC-MS/MS and MALDI
TOF/TOF MS/MS.
MALDI TOF MS Analysis
MALDI TOF MS and MALDI TOF/TOF MS/MS analyses were conducted by the
Center for Education in Proteomics Analysis (CEPA), Integrated Research in Materials,
Environments, and Society at California State University, Long Beach. Data were
acquired using an ABI 4800 MALDI TOF/TOF mass spectrometer (Applied
Biosystems/MDS SCIEX, Foster City, CA, USA) in linear, positive ion mode using Data
Explorer (ver. 4.9) and GPS Explorer (ver. 3.6) software (Applied Biosystems, Foster
City, CA, USA), respectively. For whole protein molecular masses of RP-FPLC
196
fractionated and unfractionated venom, proteins were directly spotted onto MALDI plates
with sinapinic acid as matrix. MS spectra were collected using 2,000 laser shots/spectrum
from m/z 1,000–20,000, and m/z 20,000–100,000. To avoid matrix interference, we did
not take low-mass constituents (m/z <1,000) into account (Oukkache et al., 2008; Palagi
et al., 2013). Prior to analysis, external calibration was performed using Opti-TOF™ Cal
Mix 3 Plus High Mass Calibration Insert (AB SCIEX, Framingham, MA, USA). Only
mass peaks exceeding a signal-to-noise ratio of 10 were retained.
For MALDI TOF MS of whole proteins from RP-FPLC fractionated venom, we
defined masses within ±1.0 Da in adjoining fractions as identical proteins in this study,
and such neighboring identical masses, reflecting incomplete chromatographic
separation, were removed from analysis from the fraction in which the mass had the
lower absolute intensity (Batista et al., 2007; Palagi et al., 2013). Furthermore, masses
representing apparent multiply charged or dimer species were removed from all analyses.
We constructed three-dimensional (3D) plots similar to “venom landscape” plots
(Escoubas et al., 2006; Palagi et al., 2013; Wang et al., 2013) from RP-FPLC and
MALDI TOF MS analyses using SPSS software (ver. 20.0 for Windows, 2011; Statistical
Package for the Social Sciences, Inc., Chicago, IL, USA) for each MS window (m/z
1,000–20,000 and m/z 20,000–100,000) by associating RP-FPLC fraction number (a
measure of hydrophobicity; x-axis), masses detected by MALDI TOF MS (m/z; z-axis),
and the absolute mass signal intensities (counts; y-axis). We also created twodimensional (2D) plots of MALDI TOF MS masses according to RP-FPLC percent
buffer B. Amino acid (aa) estimates were determined using the molecular mass (111.1254
Da) of an average amino acid, averagine, with the formula
197
C4.9384H7.7583N1.3577O1.4773S0.0417 based on the statistical occurrences of amino acids in
proteins (Senko et al., 1995).
LC-MS/MS Analysis
For LC-MS/MS analysis, we resuspended tryptic peptides in 20 µL of LCMS/MS mobile phase A (0.1% formic acid, 2% ACN in water). We analyzed tryptic
peptides with a ThermoFinnigan LCQ Deca XP spectrometer (Thermo Fisher Scientific)
equipped with a PicoView 500 nanospray ionization source (New Objective, Woburn,
MA, USA) using Xcalibur software (ver. 1.3; Thermo Fisher Scientific) for instrument
control and data acquisition. Separation was performed on a 10 cm x 75 µm i.d. C18
Biobasic bead column (New Objective), injecting 20 µL samples. Mobile phase B
consisted of 98% acetonitrile and 0.1% formic acid in water. The gradient program was:
0% B at 0.18 mL/min for 7.5 min, 0% B at 0.35 mL/min for 0.5 min, linear gradient to
20% B at 15 min at 0.35 mL/min, linear gradient to 75% B at 55 min at 0.3 mL/min (flow
rate constant for remainder of program), linear gradient to 90% B at 60 min, hold at 90%
B till 85 min, linear gradient to 0% B at 90 min, hold at 0% B till 120 min. Spectra were
acquired in positive ion mode with scan range m/z 300–1,500, default charge state of +2,
and minimum MS signal required set to 100,000. Each MS scan triggered three MS/MS
scans with collision energy of 30%, and minimum MS/MS signal required was set to
5,000. A 25 entry exclusion list was populated with peaks ±1.5 Da that were seen more
than twice within a 15 second window. Peaks were removed from this MS/MS exclusion
list after 1 minute. We converted MS/MS data into peaklist files using Extract_msn
implemented in Bioworks (ver. 3.1; Thermo Fisher Scientific) using the following
198
parameters: peptide molecular weight range 300–3,500, threshold 100,000, precursor
mass tolerance 1.4, minimum ion count 35. We conducted MS/MS database searches
using Mascot (licensed, ver. 2.2, Matrix Science, Boston, MA, USA) against the National
Center for Biotechnology Information non-redundant (NCBInr) database (3/6/2013,
23,038,794 sequences, 7,909,901,449 residues) in the taxon Metazoa with a parent
tolerance of 1.20 Da, fragment tolerance of 0.60 Da, and two missed trypsin cleavages
allowed. We specified carbamidomethylation of cysteine and oxidation of methionine in
Mascot as fixed and variable modifications, respectively.
MALDI TOF/TOF MS/MS Analysis
For MALDI TOF/TOF MS/MS analysis, tryptic peptides were mixed with αcyano-4-hydroxy cinnamic acid (CHCA) matrix and directly spotted onto MALDI plates.
MS spectra were collected using 1000 laser shots/spectrum, and MS/MS spectra from
3,000 shots/spectrum. Peptides with signal-to-noise ratio above 15 in MS mode were
selected for MS/MS analysis, with a maximum of 15 MS/MS spectra allowed per spot.
Internal calibration was achieved using TOF/TOF Calibration Mixture (AB SCIEX).
MS/MS data were searched against the NCBInr database (5/6/2013, 25,455,905
sequences, 8,764,053,280 residues) in the taxon Metazoa using GPS Explorer running the
Mascot (licensed, version 2.1) search engine with a peptide tolerance of 300 ppm,
MS/MS tolerance of 0.8 Da, and one missed cleavage allowed. Carbamidomethylation of
cysteine was specified as a fixed modification, and the following as variable
modifications: carbamyl, Glnpyro-Glu (N-term Q), and Glu pyro-Glu (N-term E).
199
Sequence Similarity Searching
Amino acid sequences with ion scores exceeding the Mascot homology threshold
(from LC-MS/MS analysis) or the 70% ion C.I. (from MALDI MS/MS analysis) that
were not matched to typical contaminants (keratin, albumin, human proteins, etc.) or
trypsin were subjected to sequence similarity searching using the protein Basic Local
Alignment Search Tool (BLASTp) from NCBI (http://blast.ncbi.nlm.nih.gov/Blast.cgi)
against the NCBInr database. We blasted sequences against the unrestricted database,
Scolopendromorpha only, and the deduced amino acid sequence of mature trypsin-like
serine protease Ssmase reported by Guo et al. (2013).
We employed non-parametric Mann-Whitney U tests (Zar, 1996) to compare
relative peak areas of individual RP-FPLC peaks between Arizona and California
centipedes. Statistical tests were conducted using SPSS 20.0 for Windows, with α = 0.05.
Following Nakagawa (2004), we chose not to adjust a for multiple tests. Unless indicated
otherwise, measures of central tendency presented are mean ± 1 S.E.
Results
Venom Yields
Milking 33 S. polymorpha for the mixed-location venom pool yielded a total of
34.9 µL of venom, with a protein concentration of 268 µg/µL. The 26 S. polymorpha
milked for the AZ venom pool yielded a total of 36.7 µL venom with a mean protein
concentration of 159 ± 8 µg/µL. The 36 S. polymorpha milked for the CA venom pool
200
yielded a total of 29.2 µL venom with a mean protein concentration of 193 ± 6 µg/µL.
Animals milked for RP-FPLC of individual samples yielded 0.7–2.5 µL (n = 5, AZ) and
0.1–0.9 µL (n = 4, CA) of venom at protein concentrations of 87–175 µg/µL (AZ) and
106–225 µg/µL (CA).
SDS-PAGE
When submitted to SDS-PAGE analysis (Fig. 1), reduced (lanes 1–3) and nonreduced (lanes 5–7) venom demonstrated up to 14 and 23 stained bands, respectively. In
both reduced and non-reduced forms, bands ranged from approximately 5 kDa to 240
kDa. Reduced venom showed no substantial differences among venom pools, and a
comparison of non-reduced venom revealed only minor differences, including bands at
approximately 216 kDa and 51 kDa present in the Arizona pool (lane 6) but not
distinguishable in the California pool (lane 7), and a band at approximately 17.5 kDa
present in the California pool but not distinguishable in the Arizona pool. Additionally, a
possible difference near approximately 11 kDa was noted, where the band appeared
higher in the Arizona pool than the California pool. In all three pools, non-reduced
venom showed large, densely-staining regions in the ranges 79–105 kDa and 19–26 kDa,
representing about 20% and 30% of total protein signal, respectively. Non-reduced
venom also showed faint bands at approximately 240 kDa and 205 kDa. In reduced form
(lanes 1–3), the large, densely-staining region at 79–105 kDa was absent, and the region
in the range 39–52 kDa became more diffuse and densely stained, representing about
40% of total protein signal. In addition, a prominent band appeared at 67 kDa, and the
band at approximately 6 kDa became darker and more diffuse. The densely-stained
201
region in the range 19–26 kDa was still present in the reduced venom, and faintly staining
bands at approximately 220 kDa and 240 kDa were observed.
202
203
204
MALDI TOF MS Analysis of Unfractionated Venom
An overview of the most abundant components of the venom, or the ones that
ionized most efficiently (Escoubas et al., 2006), from the three venom pools (mixed
location, AZ, CA; Figs. 2A, 2B, and 2C, respectively) was achieved by MALDI TOF MS
analysis of unfractionated venom. Although not identical, the venom mass spectra are
largely similar. The number of peaks detected in the range m/z 1,000–20,000 for AZ, CA,
and mixed pools was 23, 18, and 21, respectively. In this m/z window, peaks for both AZ
and CA pools were primarily localized between m/z 2,500 and 8,500. One considerable
difference between the AZ and CA spectra in this m/z window was the prominent peaks
between approximately m/z 3,400 and 3,600 in the AZ spectrum that are greatly reduced
in the CA spectrum. Number of peaks detected in the range m/z 20,000–100,000 for AZ,
CA, and mixed pools was 4, 2, and 9, respectively. The most noticeable feature in all
three spectra in this m/z window was the prominent peak near m/z 20,700. Additional
peaks were detected at higher m/z values but were relatively small and poorly resolved. A
low signal-to-noise ratio of unknown cause for the CA pool may have prevented
detection of some of these additional small, high m/z peaks.
205
206
207
RP-FPLC of Mixed-Location Venom Pool
RP-FPLC of the mixed-location venom pool separated 56 peaks, from which we
collected 51 fractions (Fig. 3) representing 95.1% of total peak area. The majority of
fractions were eluted with retention volumes in the range 18.5–45.5 mL, corresponding to
22.1–62.7% buffer B. Relative peak area of fractions ranged between 0.01% and 15.3%,
with the majority (88%) of fractions individually representing <5% of total peak area.
Fifteen relatively hydrophobic fractions (fractions 33–47; eluting in the range 33.2–45.5
mL, 44.2–62.7% buffer B) represented the majority of protein (75.2% of total peak area
combined).
208
209
210
MALDI TOF MS Analysis of RP-FPLC Fractionated Venom From
Mixed-Location Venom Pool
Using the mixed-location venom pool, we performed a RP-FPLC separation step
followed by offline analysis of the chromatographic fractions to gain additional
information on venom composition and to potentially reduce ion suppression effects
known to occur in MALDI of complex mixtures such as venoms (Biass et al., 2009;
Escoubas et al., 2006; Palagi et al., 2013). MALDI TOF analysis of the 51 RP-FPLC
fractions (Fig. 3) resulted in 1,043 distinguishable masses. From this cumulative total, 71
masses representing the same component (same mass ± 1.0 Da) in consecutive RP-FPLC
fractions, and 37 masses presumably representing multiply charged or dimer species,
were removed from further analyses. The resultant estimate of the total number of
components in the venom was therefore 935, with masses ranging from 1,014.5 Da to
82,863.9 Da. The major molecular mass species in each RP-FPLC fraction are listed in
Table 1.
211
212
213
Integrating the results of RP-FPLC and MALDI TOF MS into a 2D plot (Fig. 4A)
revealed that multiple masses were detected in all fractions analyzed, with some fractions
containing as many as 50 masses, indicating that multiple components were co-eluted in
each fraction. Even when considering only the most abundantly formed ions (intensity ≥
20,000 counts, n = 76 masses; Fig. 4B), multiple components per fraction were common.
The complexity of the venom is underscored by the number of components identified
from even small RP-FPLC fractions; MS on fraction 15, a small peak comprising 0.25%
of total peak area, revealed 16 peptides in the range m/z 1,086–5,991.
214
215
216
Considering all detected masses (Fig. 4A), the distribution of masses was only
weakly dependent on hydrophobicity. Peptides with masses below 5 kDa were distributed
throughout all fractions, and peptides and proteins with masses in the range 5–10 kDa
were observed in all but the earliest eluting fractions. Proteins with masses between 10
kDa and 40 kDa were likewise detected across a wide range of fractions. However, for
the most part, proteins with masses greater than ~45 kDa were only observed in late
eluting fractions, including fractions 34–42 (46.5–56.9% buffer B). Considering only the
most abundantly formed ions (intensity ≥ 20,000 counts; Fig. 4B), a general increase in
mass with increasing hydrophobicity was apparent.
Including MALDI TOF peak intensities to create 3D plots (Fig. 5), it became
clear that many of the most abundantly formed ions from the venom were in the range
m/z 1,000–10,000 and eluted prior to fraction 30 (42.5% buffer B; Fig. 5A), and the
range m/z 20,700–20,900, eluting between fractions 34–40 (46.5–53.7% buffer B; Fig.
5B). In comparison to these more abundant m/z 20,700–20,900 ions in the m/z 20,000–
100,000 window (Fig. 5B), most higher mass ions (i.e., m/z > 21,000) were present at
low abundance.
217
218
219
Plotting the number of detected masses as a function of hydrophobicity (Fig. 6A)
revealed a component-rich region in the 34–44% buffer B range (up to 84 components
eluted per 2% buffer B increase; fractions 21–32), and smaller component-rich regions in
the 22–24% (55 components eluted; fractions 9–12) and 46–52% (up to 60 components
eluted per 2% buffer B increase; fractions 34–39) buffer B ranges.
220
221
222
The distribution of molecular masses (Fig. 6B) was heavily skewed toward
smaller peptides and proteins, with 72% of components found below 12 kDa,
representing components less than ~107 amino acids in length. However, there was also a
prevalent group of proteins in the range 20–26 kDa (~179–233 aa), corresponding to
about 9% of all components. Only approximately 7% of detected components were above
40 kDa (> ~359 aa).
As a preliminary validation of our MALDI TOF MS data, we compared masses
for S. polymorpha venom to those of previously reported Scolopendra venom
constituents (Table 2), taking into consideration, when available, the elution behavior
(i.e., % ACN) of source fractions when determining the feasibility of mass matches. We
found 32 potential mass matches (all m/z discrepancies <1.9) spanning fractions 11–42,
with matches of m/z values <10,000 most common.
223
224
225
Comparing masses identified from the MALDI TOF MS spectrum of
unfractionated S. polymorpha venom (in the window m/z 1,000–20,000; Fig. 2C) to m/z
values determined based on MALDI TOF MS of RP-FPLC fractions revealed reasonable
correspondence (m/z discrepancy mean ± 1 SE = -0.11 ± 0.93, n = 13) at lower values
(m/z <5,000); however, MS of unfractionated venom appeared to underestimate masses
relative to MS of fractionated venom at higher m/z values. Of the 21 peaks (m/z 1,000–
20,000) from MS of unfractionated venom, 13 (62%) were represented as the
predominant mass in one of the RP-FPLC fractions (Table 1). While a few of the more
intense peaks (i.e., m/z 20,715.8 and 41,369.1) in the m/z 20,000–100,000 window from
MS of unfractionated venom could be matched to m/z values from MS of fractionated
venom, it was difficult to match the low intensity peaks. These comparisons must be
interpreted cautiously as the identities of potentially corresponding masses were not
verified (e.g., by sequencing).
MS/MS Analyses for Protein Identification and Sequence Similarity
Searching
Mascot searching following both LC-MS/MS and MALDI TOF/TOF MS/MS did
not lead to any significant centipede- or venom toxin-related protein hits. However,
BLASTp sequence similarity searching of amino acid sequences exceeding the Mascot
ion score homology threshold (LC-MS/MS) or 70% ion C.I. (MALDI TOF/TOF MS/MS)
revealed 20 different sequences from 18 RP-FPLC fractions showing moderate to high
similarity with known venom components. BLASTp results are shown in Table 3.
226
227
228
229
230
Comparison of Venom from S. Polymorpha from Arizona and
California by RP-FPLC of Individual Samples
We compared venom composition between Arizona (n = 5) and California (n = 4)
S. polymorpha by analyzing relative peak areas of individual RP-FPLC peaks in the
retention volume range 15.22–55.00 mL that were ≥1% of total peak area in at least one
sample. These criteria identified 46 peaks (Fig. 7). Taken together, these 46 peaks
represented on average 91.7% of total peak area for both Arizona and California animals.
Prior to analysis, a centered logratio transformation (Filzmoser et al., 2009; PawlowskyGlahn and Egozcue, 2006; Schilling et al., 2012) was applied to the data. Five of 46
peaks (11%), including peaks 4 (24.8 ± 0.04% buffer B), 6 (30.0 ± 0.04% buffer B), 10
(35.7 ± 0.03% buffer B), 20 (45.7 ± 0.06% buffer B), and 46 (64.2 ± 0.07% buffer B),
differed significantly between locations in clr-transformed relative area by MannWhitney U tests. The difference in peak 20 represented the only significant qualitative
difference between the venom profiles, with the peak absent from all Arizona venoms but
present in all California venoms (% total peak area: 2.6 ± 1.3% [mean ± SE]). Given the
uncertain homology (i.e., matching identity) of the peaks, these comparisons should be
verified with mass spectral or sequence information.
231
232
233
Discussion
SDS-PAGE
Comparison of SDS-PAGE results (Fig. 1) revealed more bands from S.
polymorpha venom (23, non-reduced) than from Mexican Scolopendra sp. (16, nonreduced; Gutierrez et al., 2003). Venom from S. polymorpha was similar in complexity to
that of S. viridicornis, Otostigmus pradoi, and Cryptops iheringi (Malta et al., 2008), but
had fewer high molecular mass proteins. Whereas S. polymorpha venom showed only a
few faintly staining bands above 110 kDa, the other three venoms studied by Malta et al.
(2008) exhibited numerous densely-staining bands above 100 kDa. Both the non-reduced
and reduced venoms of S. polymorpha and S. viridicornis showed prominent bands in the
20–30 kDa region. And prominent bands in the 80–115 kDa region of both unreduced
venoms disappeared in the reduced venom, with new bands appearing in the range 40–50
kDa, suggesting some of the higher mass components are polymeric proteins.
Comparison of reduced venoms from S. polymorpha and S. s. dehaani (Liu et al., 2012b;
Fig. 1A inset) revealed similarities including: (1) presence of several bands at high mass
(>116 kDa), (2) densely staining region between approximately 40 kDa and 50 kDa, and
(3) prominent signal in the 18–25 kDa region.
MALDI TOF MS Analysis of Unfractionated Venom
The MALDI TOF MS spectra of unfractionated venom (Fig. 2) represent the first
whole venom spectra published for S. polymorpha, and to our knowledge, the first for
any scolopendromorph. Although these spectra are useful in providing clues about S.
polymorpha venom composition, some components likely cannot be distinguished due to
ion suppression and poor resolution (broad-based peaks) (Pimenta et al., 2001). It is too
234
early to speculate on what functional differences, if any, may result from the venom
variation between S. polymorpha populations represented by the greatly reduced m/z
3,400–3,600 peaks in the CA spectrum. Identification (i.e., purification and sequencing)
and bioassays of venom components in this mass range from both populations would be
useful to determine whether these differences are qualitative or quantitative, and whether
they impact venom function.
Comparing the MALDI TOF spectra of unfractionated venom (Fig. 2) to SDSPAGE results (Fig. 1) indicated that several venom components which show up
prominently on the gel in the 10,000–20,000 Da range are not apparent by MALDI TOF
MS conducted in the m/z 1,000–20,000 range. Furthermore, venom components
appearing on the gel in the 60,000-100,000 Da range were not visible by MALDI TOF
MS conducted in the m/z 20,000–100,000 range. These components may not have been
detected by MALDI TOF as suppression effects in MALDI may prevent ionization of
some molecular species, and the high dynamic range of venom constituents can induce
the suppression of minor components during ionization (Escoubas et al., 2008; Escoubas
and Rash, 2004). Despite these differences, both methods corroborated the presence of
several prominent components in the 3.5–10 kDa range.
RP-FPLC of Mixed-Location Venom Pool
RP-FPLC of the S. polymorpha mixed-location venom pool revealed a relatively
complex chromatographic profile (Fig. 3), which was similar in complexity to the
venoms of S. viridis (Gonzalez-Morales et al., 2009) and S. s. mutilans (Yang et al.,
2013) separated by HPLC on C18 reversed-phase analytical columns. The number of
peaks (56) we detected was remarkably similar to the number of peaks (54) detected by
235
Gonzalez-Morales et al. (2009). Rates et al. (2007) found 62 and 65 components from S.
viridicornis nigra and S. angulata venoms, respectively, using a two-dimensional
chromatographic analysis (cation-exchange chromatography followed by RP-HPLC). By
comparison, using a low-resolution HPLC separation (semi-preparative C8 reversedphase column) these investigators separated approximately 24 peaks (Rates et al., 2007,
Fig. 8) in the crude venom of S. v. nigra. Many of the major peaks from separations of S.
polymorpha (Fig. 3), S. viridis (Gonzalez-Morales et al., 2009) and S. v. nigra (Rates et
al., 2007) venoms were late-eluting (relatively hydrophobic). Similarly, two of the five
peaks (peaks II and III, Liu et al., 2012b, Fig. 1A) separated by gel filtration of S. s.
dehaani venom yielded many relatively late eluting peaks when subjected to C18 RPHPLC (Liu et al., 2012b).
MALDI TOF MS Analysis of RP-FPLC Fractionated Venom from
Mixed-Location Venom Pool
The total number of masses detected (935) in the venom of S. polymorpha by RPFPLC and offline MALDI TOF MS was large, but likely represents an overestimation of
the number of distinct components. First, some molecular masses may represent the same
polypeptide chain having undergone post-translational modifications and/or unspecific
cleavages (Pimenta et al., 2001). Second, redundancy in mass assignments can occur due
to unspecific chromatographic behavior, in which the same molecule can be found spread
over many elution fractions (Pimenta et al., 2001). To combat this possibility we removed
components of the same mass (± 1.0 Da) in adjacent fractions. However, if a component
exhibited highly unspecific chromatographic behavior and eluted across many fractions
with sufficient variability in measured mass, our methods would not have detected and
236
removed such inapparent redundant masses. Third, contamination of venom by cellular
debris can occur as a result of electrical extraction (McCleary and Heard, 2010; Pimenta
et al., 2001), with some of the measured molecular masses originating outside of the
venom itself (Pimenta et al., 2001). Fourth, some masses could be related to experimental
adducts (Palagi et al., 2013; Pimenta et al., 2001) or even spurious fragments (Rates et
al., 2008; Rodriguez de la Vega et al., 2010). Finally, since data were obtained with a
pooled sample reflecting centipedes from two locations and likely a range of ontogenetic
stages, the number of molecular species may overestimate the number of different
proteins in any individual centipede, and should be regarded as providing a general
picture of the actual venom component diversity at the species level (Rodriguez de la
Vega et al., 2010).
Even taking into account some degree of redundancy in the data, the number of
proteins in the venom of S. polymorpha certainly reaches several hundred, on par with the
number reported for other centipedes, and indicative of considerable venom complexity.
For example, 543 different proteins/peptides were deduced from venom gland cDNA
from S. s. dehaani (Liu et al., 2012b). The total number of molecular species identified in
S. polymorpha venom was considerably larger than that reported for venoms of S.
viridicornis nigra (62 components) and S. angulata (65 components) (Rates et al., 2007).
However, considering only the most abundant masses (intensity ≥ 20,000 counts, Fig.
4B), our value of 76 molecular species is comparable with the findings from S. v. nigra
and S. angulata. Differences in numbers of components identified may be associated with
a difference in MS methods employed by this investigation (MALDI TOF) and that of
Rates et al. (2007), who used ESI-Q-TOF (Batista et al., 2006; Biass et al., 2009;
237
Escoubas et al., 2008). The range of molecular masses identified from S. polymorpha
venom (1,014.5–82,863.9 Da) is also considerably larger than that reported for S. v. nigra
(3,019.62–20,996.94 Da) or S. angulata (1,304.74–22,639.15 Da, Rates et al., 2007), or
for S. s. dehaani (3,233.7–22,559.3 Da, Liu et al., 2012b), given the presence (although at
low levels) of a number of relatively high mass (m/z > 23,000) components (Fig. 6B).
The presence of these high mass components in S. polymorpha venom was supported by
SDS-PAGE results (Fig. 1). When limited to the most abundantly formed ions, the range
for S. polymorpha venom (1,033.8–20,877.8 Da) was similar to that reported for other
scolopendrids. Thus, the venom of S. polymorpha appears to be at least as complex, if not
more so, than previously studied scolopendrid venoms. In comparison to the many
scolopendromorph venom components identified by mass spectrometry in this
investigation and others (Liu et al., 2012b; Rates et al., 2007), or deduced by
transcriptomics (Guo et al., 2013; Liu et al., 2012b; Yang et al., 2012), relatively few
have been definitively characterized functionally, underscoring that the diversity of
centipede venom remains largely unexplored.
Using the same method of liquid chromatography followed by offline analysis of
chromatographic fractions using MALDI TOF MS, similarly large venom complexity has
been observed in other animals. The number of venom components can reach 889 in cone
snail venom (Biass et al., 2009), 1,018 within a single spider species (Escoubas et al.,
2006), 120–800 in other spider species (Liao et al., 2007; Palagi et al., 2013; Tang et al.,
2010), and 205–632 in scorpions (Batista et al., 2006; Nascimento et al., 2006; Pimenta et
al., 2001; Schwartz et al., 2008). Additionally, just as we found numerous components by
MALDI TOF MS in even small RP-FPLC peaks, MALDI TOF MS investigations
238
following single-dimension liquid chromatography often reveal a multiplicity of
components per fraction (Biass et al., 2009; Escoubas et al., 2006; Palagi et al., 2013;
Wang et al., 2013).
Examining all MALDI TOF MS masses from RP-FPLC fractions (Fig. 4A)
revealed only a weak association between molecular mass and hydrophobicity for S.
polymorpha venom components, with 5–10 kDa components observed in many fractions.
The prevalence of these small components across a broad range of RP-FPLC fractions
was consistent with the frequency histogram (Fig. 6B), and suggests a wide underlying
structural diversity in scolopendrid venom components of similar size. A similarly low
correlation between chromatographic retention time and molecular mass was observed in
venom proteins from Australian funnel-web spiders (Escoubas et al., 2006) and tarantulas
in the genus Brachypelma (Escoubas et al., 1997). However, looking only at the most
abundantly formed ions (Fig. 4B) revealed a clearer positive association between
hydrophobicity and m/z. This phenomenon is more reminiscent of results obtained from
the venom of the Chinese scorpion Mesobuthus martensii, in which RP-HPLC retention
times clearly separated out “short” peptides acting on potassium channels, “short
insectotoxins”, and “long” peptides acting on sodium channels (Escoubas et al., 2008;
Romi-Lebrun et al., 1997). Such a phenomenon can occur when venom components,
despite varying in mass, share a similar structure (e.g., all tightly packed and stabilized by
disulfide bridges, Batista et al., 2007).
Venom protein structural families, whose members often share related
pharmacologies, have been identified based on the clustering of venom protein masses on
plots of hydrophobicity (% ACN or fraction number) versus mass in Australian funnel-
239
web spiders (Atrax and Hadronyche, Escoubas et al., 2006) and in buthid scorpions
(Nascimento et al., 2006). In contrast, the positions of tarantula (Brachypelma) toxins in
hydrophobicity versus mass plots showed no correlation with their pharmacological
activity (Escoubas et al., 1997). Rates et al. (2007) suggested that centipede venom
protein structural families may also be inferred from clusters formed on these plots, and
hypothesized 22 putative protein families based on analysis of the venoms of S. v. nigra
and S. angulata. Examining our hydrophobicity (% buffer B) versus m/z plot including
all detected masses (Fig. 4A), it was difficult to distinguish potential clusters due to the
sheer number of data points in close proximity to one another. However, when data were
restricted to the most abundantly formed ions (intensity ≥ 20,000 counts; Fig. 4B),
potential clusters became visible, and the general appearance of the plot was reminiscent
of Rates et al.’s (2007) plots for S. v. nigra and S. angulata. We did not verify structural
or functional similarity of clustered masses, and therefore the possibility of identification
of protein families by clustering in S. polymorpha venom seems purely speculative. We
have refrained from attempting to identify specific clusters.
The distribution of molecular masses in S. polymorpha venom was skewed toward
smaller peptides and proteins (Fig. 6B). In comparison to the distribution of masses from
the venoms of S. v. nigra and S. angulata (Rates et al., 2007), S. polymorpha venom was
not as dramatically bimodal, and was more heavily weighted towards components below
m/z 4,000. Furthermore, components in the range m/z 10,000–12,000 were relatively
more abundant in S. polymorpha venom. The venoms of S. polymorpha, S. v. nigra, and
S. angulata were similar in having a prominent group of proteins in the range m/z
20,000–22,000. Consistent with our findings for S. polymorpha venom (Fig. 6B), a large
240
percent of Scolopendra venom component masses determined thus far have been below
12 kDa: 89% in S. v. nigra, and 77% in S. angulata (Rates et al., 2007); 54% in S. s.
dehaani (Liu et al., 2012b); and 100% in S. s. mutilans (Yang et al., 2012). Similarly, the
prevalent group of proteins (about 9% of masses) in the range m/z 20,000–26,000
observed in S. polymorpha venom appears consistent with the prevalence of these masses
reported in the literature: 11% in S. v. nigra and 23% in S. angulata (Rates et al., 2007);
and 31% in S. s. dehaani (Liu et al., 2012b). Although protein masses above 22.7 kDa
have been reported from SDS-PAGE analyses of scolopendromorph venoms (e.g., 60
kDa, Gomes et al., 1983; >190 Da, Malta et al., 2008) and from transcriptomic
investigations as theoretical molecular masses (e.g., 27,199.02 Da, Liu et al., 2012b), to
date this is the first mass spectrometry investigation to report masses above m/z 22,700.
Perhaps this is because of a bias toward investigating components that demonstrate
physiological activity, that are easier to chromatographically purify, or that are present at
high abundance. Regardless, it appears that future mass spectrometry studies of
scolopendromorph venoms aimed at systematic investigation of potential masses above
23 kDa could yield novel results.
The comparison between MALDI TOF MS masses from S. polymorpha venom
and potential mass matches from the Scolopendra literature (Table 2) is necessarily
tentative at best, requiring protein sequences to confirm component identities. Although
many of the potential mass matches are to peptides and proteins of unknown function,
these comparisons raise the possibility of the presence of K+ channel inhibitors, a Ca2+
channel inhibitor, and a venom allergen 3-like protein in S. polymorpha venom.
241
MS/MS Analyses for Protein Identification and Sequence Similarity
Searching
There is a paucity of genome/protein sequence information available for
Scolopendromorpha (NCBI accessed 2/7/14: 2,522 nucleotide sequences, 668 protein
sequences) in general, and for the genus Scolopendra specifically (1,543 nucleotide
sequences, 281 protein sequences). For S. polymorpha in particular, there are even fewer
sequences (68 nucleotide sequences, 64 protein sequences), none of which represent
venom toxins. Only 23 centipede venom toxins were present in the NCBInr database at
the time of searching. Adding to the challenge, the many protein sequences determined
by Liu et al.’s (2012b) work on the venom of S. s. dehaani were unavailable for
searching, presumably due to the “unverified” status in the NCBInr database (Benson et
al., 2012) of their submissions. Given the limited number of centipede venom toxins
available for searching, and the novel nature of many centipede venom components (Liu
et al., 2012b; Rates et al., 2007; Undheim et al., 2012; Yang et al., 2012; Yang et al.,
2013), our lack of Mascot protein matches was not surprising.
Even when species-specific genomic or protein sequence information is not
available in databases, proteins still may be identified through cross-species protein
identifications (de Graaf et al., 2009; Liska and Shevchenko, 2003; Shevchenko et al.,
2009; Ward et al., 2010). Often, homologous proteins from closely related species will
only differ by a few amino acids, allowing identification by MS/MS data (Shevchenko et
al., 2009; Ward et al., 2010). However, as organisms become more phylogenetically
distant, identification using automated database searching is hampered as orthologous
proteins retain a lower percentage identity (Liska and Shevchenko, 2003; Ward et al.,
2010). In such circumstances, de novo sequencing and database blasting are often
242
employed (Junqueira et al., 2008; Shevchenko et al., 2009; Ward et al., 2010; Waridel et
al., 2007). Characterization of venom components by blasting peptide sequences (from
MS/MS data or from Edman degradation) has proven useful in venom exploration when
species-specific database sequences are scarce (Calvete et al., 2007b; Carregari et al.,
2013; de Graaf et al., 2010; Guercio et al., 2006; Liao et al., 2007). However, great
caution should be exercised when interpreting similarity between protein sequences of
phylogenetically distant species, as sequence similarity does not necessarily imply
functional similarity (Junqueira et al., 2008; Shevchenko et al., 2009).
Even using sequence similarity searching, Rates et al. (2007) encountered a
challenge in their “structure-to-function” proteomic approach of characterizing the
venoms of S. v. nigra and S. angulata; when the N-termini of representatives of 10
protein/peptide families were sequenced, sequences from nine families showed no
significant similarity to protein sequences deposited in the Swiss-Prot database.
Similarly, Liu et al. (2012b) demonstrated that 308 of 543 (56.7%) proteins/peptides
deduced from cDNAs of S. s. dehaani venom, and most of the 40 proteins/peptides
purified from the venom, showed no significant similarity to existing database sequences.
In our investigation, only 18 out of 51 RP-FPLC fractions yielded amino acid sequences
with similarity to existing proteins by BLASTp searching. Especially lacking were
characterizations of the more hydrophobic components eluting in the range 51–63%
buffer B (fractions 39–47). The number of uncharacterized components suggests the
presence of novel molecules in the venom of S. polymorpha, and highlights the need for
expansion of databases by sequencing of centipede genomes and venom proteins.
243
BLASTp sequence similarity searching of amino acid sequences from LCMS/MS and MALDI TOF/TOF MS/MS (Table 3) revealed four unique sequences in six
fractions that showed similarity to serine proteases. These serine protease-like proteins
eluted towards the first half of the chromatography run, in the range of fractions 13–26
(24.9–37.6% buffer B). Serine proteases have been identified as venom components in
other scolopendromorphs, including S. s. mutilans (Guo et al., 2013) and S. s. dehaani
(Liu et al., 2012b). Venom serine proteases were expressed in low numbers in S. s.
dehaani, with only two different molecules found by transcriptomic analysis (Liu et al.,
2012b). Additionally, a serine protease with potent fibrinolytic activity was found in the
whole-body extract of S. s. mutilans (You et al., 2004). The role of centipede venom
serine proteases in envenomation has not been established, though snake venom serine
proteases affect various physiological functions including blood coagulation, fibrinolysis,
blood pressure, and platelet aggregation (Phillips et al., 2010; Serrano and Maroun,
2005). Spider venom serine proteases likely contribute to dermonecrotic injury (Veiga et
al., 2000), causing local tissue destruction by degrading collagen type-I and fibronectin
(Devaraja et al., 2008), and interfering with hemostasis (Devaraja et al., 2010, 2011).
Scorpion venom serine proteases are suspected to act as spreading factors, increasing
tissue permeability and facilitating the spread of other venom proteins, or to be involved
in the post-translational processing of toxins (Almeida et al., 2002). Serine proteases are
also components of hymenopteran venoms (Santos et al., 2011); in the bumblebee
(Bombus ignitus), bee venom serine protease (Bi-VSP) induces a lethal melanization
response in target insects by modulating the innate immune response, and in target
mammals acts as a fibrin(ogen)olytic enzyme to facilitate the spread of bee venom
244
components throughout the bloodstream (Choo et al., 2010). The majority of proteolytic
activity in centipede venoms is believed to stem from metalloproteases (Malta et al.,
2008; Undheim and King, 2011); however, the weak gelatinolytic activity from nonmetalloproteases detected in the venoms of O. pradoi and C. iheringi by Malta et al.
(2008) was attributed to serine proteases (Undheim and King, 2011). Because serine
protease activities were weak, Undheim and King (2011) suggested the enzymes play a
role in toxin activation by cleaving precursor peptides to create mature toxins rather than
acting to disrupt a victim’s hemostasis. Indeed, the centipede venoms tested by Malta et
al. (2008) did not interfere with coagulation factors. In fractions 24–26 of S. polymorpha
venom, the amino acid sequence NDIALLRLQK showed similarity to both the potent
fibrinolytic enzyme (scolonase) isolated from S. s. mutilans whole-body extracts by You
et al. (2004), and the novel putative trypsin-like serine protease (Ssmase) from cDNA
characterization of S. s. mutilans venom by Guo et al. (2013). In the latter case, sequence
alignment placed the aspartic acid residue from the S. polymorpha sequence as part of the
conserved catalytic triad, specifically Asp108.
Three different amino acid sequences, one in fraction 15 (26.8% buffer B) and
two in fraction 28 (39.4% buffer B), showed similarity to disulfide-rich, voltage-gated
potassium (Kv) channel inhibitors from S. s. mutilans venom (Yang et al., 2012). The
sequence from fraction 15 aligned with the 24-residue propeptide of κ-SLPTX-Ssm2a,
which is removed during post-translational processing (Yang et al., 2012). The
determined mass of κ-SLPTX-Ssm2a was 3,465.8 Da, very close to the mass of one of
the molecular species (3,464.6 Da) we detected in fraction 15. However, κ-SLPTXSsm2a eluted in RP-HPLC at about 50% ACN, whereas our 3,464.6 Da species in
245
fraction 15 eluted at only 26.8% buffer B. κ-SLPTX-Ssm2a inhibited K+ currents in rat
dorsal root ganglion (DRG) neurons with an IC50 of approximately 570 nM, and had
potent insecticidal activity, with injected insects showing signs of neurotoxicity including
twitching, paralysis, and body contraction (Yang et al., 2012). Two additional peptide
neurotoxins acting on Kv channels have been identified in the venom of S. s. mutilans
(Yang et al., 2012). The relative importance of neurotoxins in centipede venoms remains
unknown; however, neurotoxic effects of the venoms are reported (Undheim and King,
2011). In green mambas and black mambas, potent modulators of Kv channels called
dendrotoxins cause convulsion and death when injected centrally in mice (Harvey, 2006;
Harvey and Robertson, 2004). In spiders, disulfide-rich neurotoxic peptides, including
modulators of Kv channels, are the major contributors to the venoms’ insecticidal
activities (King and Hardy, 2013; Kuhn-Nentwig et al., 2011). Peptide neurotoxins
targeting Kv channels are also components of venoms of scorpions (de la Vega and
Possani, 2004; Schwartz et al., 2013), bees (Castle et al., 1989; Dreyer, 1990), sea
anemones (Castaneda and Harvey, 2009; Frazao et al., 2012), and cone snails (Lewis et
al., 2012; Terlau and Olivera, 2004).
Two different amino acid sequences in five fractions ranging from fraction 15
(26.8% buffer B) to fraction 37 (49.8% buffer B) showed similarity to two putative
neurotoxins (putative neurotoxins 8 and 4) of unknown function identified by Yang et al.
(2012) from S. s. mutilans venom. The sequence from fraction 15 was similar to one of
the four families of putative neurotoxins, while the sequence from fractions 32, 33, 34,
and 37 was similar to a second family.
246
Only one amino acid sequence, from fraction 24 (36.5% buffer B), showed
limited similarity to a PLA2, in this case to the PLA2 from Mexican S. viridis named
Scol/Pla (Gonzalez-Morales et al., 2009). A phylogenetic analysis of the Scol/Pla
sequence showed that it was more similar to snake phospholipases than to those of
arthropods (Gonzalez-Morales et al., 2009). Low phospholipase activity has been
reported for the venoms of S. viridicornis and O. pradoi, and Malta et al. (2008) posited
that PLA2 could contribute to the myotoxicity observed in S. viridicornis and O. pradoi
venoms. No PLA2 activity was detected in the venom of the cryptopid C. iheringi,
suggesting that PLA2 may not be ubiquitous in centipede venoms (Malta et al., 2008;
Undheim and King, 2011). In contrast to centipedes, PLA2 enzymes feature prominently
in snake venoms, playing an important role in immobilization and capture of prey, and
causing a wide variety of pharmacological effects including neurotoxicity, myotoxicity,
cardiotoxicity, anticoagulant effects, interference with platelet function, hemolysis,
edema, and internal hemorrhage (Doley et al., 2010). PLA2s are also found in the venoms
of spiders (Kuhn-Nentwig et al., 2011), scorpions (Costal-Oliveira et al., 2012;
Hariprasad et al., 2013), hymenopterans (de Lima and Brochetto-Braga, 2003; Santos et
al., 2011), sea anemones (Frazao et al., 2012), and cone snails (Terlau and Olivera, 2004).
One amino acid sequence, from fraction 24 (36.5% buffer B), showed similarity
to a PLA2 inhibitor from the serum of the ringed water snake, Sinonatrix annularis. To
our knowledge, PLA2 inhibitors have not been reported from scolopendromorph venoms.
PLA2 inhibitors have been isolated from the blood plasma or serum of numerous
venomous snakes (Neves-Ferreira et al., 2010), where they are believed to contribute to
the natural resistance of snakes to their own venom, and perhaps to that of other
247
venomous snakes (Dunn and Broady, 2001; Lizano et al., 2000). These inhibitors are
typically secreted from the liver (Kinkawa et al., 2010; Lima et al., 2011); however, they
have also been detected in the transcriptomes of the venom glands and in the venom
(Cidade et al., 2006; Lima et al., 2011; Luna et al., 2013; Sousa et al., 2001). The
presence of the inhibitor may provide a mechanism to control venom toxicity during
synthesis and storage (Luna et al., 2013). The role, if any, that PLA2 inhibitors play in
envenomation remains unknown (Cidade et al., 2006).
With a BLASTp search limited to the taxon Scolopendromorpha, one amino acid
sequence, from fraction 27 (38.2% buffer B), showed similarity to one (Scolopendra
20,566.01 Da toxin) of the five toxins identified by Rates et al. (2007) in the Scolopendra
toxin 10 family (also called the 20 kDa protein family, or the CRISP-like family) from S.
angulata. Although the Scolopendra 20,566.01 Da toxin showed similarity to vespid
allergen V and to several snake venom CRISPs, our particular sequence did not.
Scolopendromorph venom toxins with similarity to wasp allergens are interesting given
that patients with allergies to centipede venom also display allergic reactions to wasp,
honey bee, and yellow jacket venoms (Harada et al., 2005). With an unrestricted
BLASTp search, the sequence from fraction 27 showed similarity to a matrix
metalloproteinase (MMP) from the freshwater cnidarian Hydra vulgaris, aligning near
the N-terminal region of the hemopexin domain (Leontovich et al., 2000). This MMP was
shown to digest Hydra extracellular matrix (ECM), which is similar to that seen in
vertebrates (Leontovich et al., 2000). Present in both invertebrates and vertebrates,
MMPs comprise a major enzyme group involved in tissue remodeling and repair via
ECM degradation (Murphy and Nagase, 2008). MMPs have been grouped within the
248
metazincins, together with snake venom metalloproteinases, astacins, serralysins, and
ADAMs (a disintegrin and metalloproteinase; Bode et al., 1993). Such zinc-dependent
metalloproteinases are common constituents of the venom proteomes and venom gland
transcriptomes of viperid and colubrid snakes (Gutierrez et al., 2010). Metalloproteinases
targeting extracellular matrix have also been reported in the venoms of spiders (TrevisanSilva et al., 2010; Veiga et al., 2001).
Two different amino acid sequences, one in fraction 30 (42.5% buffer B) and the
other in fraction 33 (45.3% buffer B), showed similarity to toxins in the Scolopendra
toxin 3 family identified by Rates et al. (2007) from S. angulata and S. v. nigra,
respectively. Annotation in the NCBInr database identified these toxins as putative
neurotoxins.
One amino acid sequence, in fraction 32 (43.8% buffer B), showed similarity to
the antimicrobial peptide scolopin-2, one of two antimicrobial peptides (scolopins)
identified by Peng et al. (2010) from S. s. mutilans. Both scolopins showed strong
antimicrobial activities against Gram-positive and Gram-negative bacteria, as well as
against yeast. The scolopins also induced histamine release by mast cells, and showed
moderate hemolytic activities against both human and rabbit red blood cells. Peng et al.
(2010) suggested that the scolopins, along with venom phospholipases, might take part in
the hemolytic functions (e.g., Deng et al., 1997; Malta et al., 2008) of centipede venoms.
As scolopins induce release of histamine, a known transmitter for pain signaling, the
authors further proposed that scolopins may contribute to the nociceptive effects of
centipede stings. Wenhua et al. (2006) demonstrated that potent antimicrobial agents,
including the peptide scolopendrin I, could be induced in the venom following injection
249
of S. s. mutilans with Escherichia coli. The primary function of these cytotoxic peptides
may be to prevent infection of the venom duct and gland lumen by pathogenic
microorganisms (Undheim and King, 2011; Wenhua et al., 2006). Antimicrobial
components have been identified in the venoms of numerous taxa, including snakes,
spiders, scorpions, and hymenopterans (Kuhn-Nentwig, 2003; Samy et al., 2006), and
may serve multiple roles, including incapacitating prey (often as a spreading agent),
preventing infection of the venom glands during venom injection, disinfecting and
possibly preserving prey, and protecting hymenopteran broods from microorganisms
(Kozlov et al., 2006; Kuhn-Nentwig, 2003).
One amino acid sequence, in fractions 34 (46.5% buffer B) and 35 (48.0% buffer
B), showed high similarity to a hypothetical protein from the planktonic crustacean
Daphnia pulex, aligning within the predicted protein’s SCP-like extracellular protein
domain. The sperm-coating glycoprotein (SCP) protein family (Pfam PF00188) is also
known as the CAP (cysteine-rich secretory proteins [CRISPs], antigen 5 [Ag5], and
pathogenesis-related 1 [Pr-1]) superfamily of proteins, and the structurally conserved
CAP (or SCP) domain is characteristic and definitive of the entire superfamily (Gibbs et
al., 2008). CRISPs are particularly enriched in reptilian venom ducts, and although the
main function of CRISPs in envenomation is poorly understood, they may act to block
various ion channels (Sunagar et al., 2012). The Ag5 proteins are abundant, immunogenic
proteins present in the venom of some hymenopterans (Henriksen et al., 2001; King and
Spangfort, 2000), but their role in envenomation remains unclear (King and Spangfort,
2000). Venom allergen 3 from fire ants is included in the CAP superfamily (Cantacessi et
al., 2009), and the amino acid sequence from fractions 34 and 35 also had high similarity
250
to a venom allergen 3-like protein from the bee Megachile rotundata. Centipede venom
proteins with similarity to hymenopteran venom allergens may explain the allergy-like
complications recorded after centipede envenomations (Undheim and King, 2011).
One amino acid sequence, in fractions 37 (49.8% buffer B) and 38 (50.4% buffer
B), showed similarity to a ω-SLPTX-Ssm1a neurotoxin precursor from S. s. mutilans,
aligning within the mature toxin sequence of this Cav channel activator (Yang et al.,
2012). The peptide neurotoxin ω-SLPTX-Ssm1a was shown by Yang et al. (2012) to
increase Cav currents in DRG neurons by 70% at 1 mM concentration. The authors also
identified a second Cav modulator from the venom of S. s. mutilans, ω-SLPTX-Ssm2a,
which inhibited Cav channel currents in DRG neurons with an IC50 of 1,590 nM. Many
neurotoxins act on Cav channels, which play an important role in cardiac, muscular, and
neuronal function (Kuhn-Nentwig, 2003). Cav channel modulators are present in snake
venoms (Sousa et al., 2013), and modulation of Cav channels comprises one of the
dominant pharmacologies of spider-venom toxins (Smith et al., 2013). Modulators of Cav
channels are also found in the venoms of scorpions (Chuang et al., 1998; OlamendiPortugal et al., 2002). A range of peptides (ω-conotoxins) from cone snail venom
preferentially inhibits Cav channels (Lewis et al., 2012).
One amino acid sequence, from fraction 38 (50.4% buffer B), showed similarity
to L-cystatin, a Family 2 cystatin, from the Japanese horseshoe crab (Tachypleus
tridentatus), aligning within the cystatin-like domain. The cystatins are a superfamily of
cysteine protease inhibitors found in a wide range of organisms and involved in many
biological functions (Kordis and Turk, 2009; Ochieng and Chaudhuri, 2010). Cystatins
have been identified from snake venoms and show high sequence identity with the
251
Family 2 cystatins (Richards et al., 2011). In fact, protein sequence parsimony analysis
suggested that the cystatins from Bitis arietans and Naja naja atra, together with T.
tridentatus cystatin and human cystatin M, form a new subfamily within the cystatin
Family 2 (Brillard-Bourdet et al., 1998). There is no evidence that cystatins are directly
involved in venom toxicity, and their biological role in snake venom is therefore unclear
(Brillard-Bourdet et al., 1998; Richards et al., 2011). Cystatins may protect venom
proteins from proteolytic inactivation by proteases of the snakebite victim (BrillardBourdet et al., 1998), however housekeeping or regulatory roles have been suggested
(Richards et al., 2011).
One amino acid sequence, from fraction 49 (65.1% buffer B), showed high
similarity to crustacean cardioactive peptide (CCAP), originally known as
cardioacceleratory peptide 2a (Cheung et al., 1992), from the tobacco hornworm (moth)
Manduca sexta. CCAP and related peptides represent multifunctional regulatory
neurotransmitters, modulators, and hormones found in invertebrates (Dircksen, 1998;
Wasielewski and Skonieczna, 2008). Cardioacceleratory effects have been described
when CCAP was applied in vitro and in vivo (Dulcis et al., 2005; Estevez-Lao et al.,
2013). A CCAP-related peptide (conoCAP-a) has been isolated from the venom of Conus
villepinii, and cloning of its cDNA precursor revealed two additional conoCAPs (Moller
et al., 2010). The Conus CAPs showed up to 78% sequence homology with CCAP
(Moller et al., 2010). While arthropod CCAP is a cardio-accelerator, conoCAP-a
decreased the heart frequency in Drosophila larvae, decreased heart frequency and blood
pressure in rats (Moller et al., 2010), and decreased heart rate in Danio rerio embryos
(Miloslavina et al., 2010). Moller et al. (2010) suggested that the worm-hunting snail may
252
use conoCAPs for targeting a specific receptor in their prey, thus facilitating prey
capture. Cardiotoxic effects have been demonstrated using forcipule extracts from the
centipedes S. morsitans and S. subspinipes (Gomes et al., 1982b, 1983; Mohamed et al.,
1980), and Undheim and King (2011) speculated that cardiotoxins are likely present in
other Scolopendra, as cardiovascular-related symptoms have been reported in humans
following stings from S. heros (Bush et al., 2001) and other centipedes (Chaou et al.,
2009; Ozsarac et al., 2004; Yildiz et al., 2006).
Comparison of Venom from S. Polymorpha from Arizona and
California
Results of SDS-PAGE, MALDI TOF MS, and RP-FPLC analyses of venom from
Arizona and California populations of S. polymorpha indicated relatively subtle
differences in venom composition. Palagi et al.’s (2013) analysis of venom from different
populations of the Australian funnel-web spider H. infensa showed similar minor
variations, leading the authors to contend that habitat factors such as seasonal
microclimate, local topography and geology, and prey species diversity and abundance
only subtly influence the expression and levels of venom peptides of geographic variants
of the same species. Similarly, venom from a Swiss population of the hobo spider
Tegenaria agrestis was found to differ only subtly in composition (based on RP-HPLC)
from UK and US venoms, with several quantitative differences but only a few qualitative
differences (Binford, 2001). In contrast, the rattlesnake venom proteome exhibits
considerable intraspecific geographic variation, presumably the result of selection arising
from prey differences among populations (Boldrini-Franca et al., 2010; Mackessy, 2010;
Sunagar et al., 2014). Venom differences between Arizona and California populations of
253
S. polymorpha may similarly be associated with subtle dietary differences, as suggested
by Antoniazzi et al. (2009) for venom variation among Brazilian scolopendromorphs (S.
viridicornis, O. pradoi, and C. iheringi) detected by Malta et al. (2008).
Conclusions
The venom of S. polymorpha is highly complex, generating up to 23 bands by
SDS-PAGE, 56 peaks by RP-FPLC, and hundreds of molecular mass species by MALDI
TOF MS. SDS-PAGE indicated venom component masses as high as approximately 240
kDa. MALDI TOF masses also ranged widely (approximately 1–83 kDa), but the
frequency distribution of masses was dominated by peptides and proteins below 12 kDa.
Many of the most abundantly formed ions from the venom were below m/z 10,000 and
eluted prior to fraction 30 (42.5% buffer B), while the majority of the high mass ions
(i.e., m/z > 21,000) were present at low abundance. Comparing venom pools of S.
polymorpha from Arizona and California by SDS-PAGE and MALDI TOF MS, and
individual samples from both locations by RP-FPLC, revealed minor, largely
quantitative, differences. Twenty different amino acid sequences from 18 RP-FPLC
fractions showed similarity to known venom components, including serine proteases, Kv
channel inhibitors, PLA2, CRISP-like proteins, a Cav channel activator, and several
putative neurotoxins of unknown function. The complexity of the venom coupled with
the number of uncharacterized RP-FPLC peaks suggests numerous novel components
remain to be identified in the venom of S. polymorpha.
Future studies on the venom of S. polymorpha would benefit from the added
physicochemical information available from additional separation steps orthogonal to
reversed-phase chromatography (i.e., 2D-PAGE, ion exchange chromatography, size
254
exclusion chromatography) prior to mass spectrometry analysis. With the potential to
provide new dimensions of information (e.g., isoelectric point, ionic behavior) on
molecular species, these steps will also facilitate the purification of components, leading
to improved mass spectra and setting the stage for protein sequencing and functional
assays (Vetter et al., 2011). Such multidimensional pre-MS separation steps for decomplexing venoms have proved valuable in achieving deep proteomic coverage of snake
venoms (Fox et al., 2006). Given the important role that matrix selection can play in
MALDI TOF MS (Watson and Sparkman, 2007), including analyses of venom (Batista et
al., 2004; Escoubas et al., 1997; Quinton et al., 2007), future investigations should also
consider whether additional S. polymorpha venom components might be characterized by
employing different matrices. In addition, a more complete picture of venom composition
could be achieved by also analyzing fractionated venom for whole protein masses using
electrospray ionization mass spectrometry (ESI-MS); a study of C. consors (cone snail)
venom comparing MALDI TOF MS and ESI-MS found the techniques are
complementary, with only 21% of masses being common to both data sets (Biass et al.,
2009). Future analyses should also aim to derive an S. polymorpha venom gland
transcriptome to function as a database to search venom protein mass spectrometry data
against. Finally, if methods can be devised to obtain venom from S. polymorpha without
electrical stimulation, it would be enlightening to compare this secretion with electrically
extracted venom.
Acknowledgments
The authors thank Dr. YuanYu Lee, Center for Education in Proteomics Analysis
(CEPA), Integrated Research in Materials, Environments, and Society at California State
University, Long Beach, for MALDI TOF MS analyses. We thank Dr. P. Duerksen255
Hughes and her students in the LLU Biochemistry Department, and Dr. U. Soto-Wegner
in the LLU Basic Sciences Department for access to their equipment. We are also
grateful to Charles Kristensen of Spider Pharm Inc. for his valuable instruction regarding
venom extraction.
Conflict of Interest Statement
The authors declare no conflicts of interest with respect to the authorship and/or
publication of this article.
Ethical Statement
with Loma Linda University’s Policy on the Care and Use of Laboratory Animals.
256
References
Adis, J., 2002. Myriapoda: identification to classes, in: Adis, J. (Ed.), Amazonian
Arachnida and Myriapoda. Pensoft, Sofia [Bulgaria], pp. 457-458.
Almaaytah, A., Albalas, Q., 2014. Scorpion venom peptides with no disulfide bridges: a
review. Peptides 51, 35-45.
Almeida, F.M., Pimenta, A.M.C., De Figueiredo, S.G., Santoro, M.M., Martin-Eauclaire,
M.F., Diniz, C.R., De Lima, M.E., 2002. Enzymes with gelatinolytic activity can
be found in Tityus bahiensis and Tityus serrulatus venoms. Toxicon 40, 10411045.
Toxicon 53, 367-374.
Balhara, K.S., Stolbach, A., 2014. Marine envenomations. Emerg. Med. Clin. N. Am. 32,
223-243.
Baron, A., Diochot, S., Salinas, M., Deval, E., Noel, J., Lingueglia, E., 2013. Venom
toxins in the exploration of molecular, physiological and pathophysiological
functions of acid-sensing ion channels. Toxicon 75, 187-204.
Batista, C.V.F., D'Suze, G., Gomez-Lagunas, F., Zamudio, F.Z., Encamacion, S., Sevcik,
C., Possani, L.D., 2006. Proteomic analysis of Tityus discrepans scorpion venom
and amino acid sequence of novel toxins. Proteomics 6, 3718-3727.
Batista, C.V.F., del Pozo, L., Zamudio, F.Z., Contreras, S., Becerril, B., Wanke, E.,
Possani, L.D., 2004. Proteomics of the venom from the Amazonian scorpion
Tityus cambridgei and the role of prolines on mass spectrometry analysis of
toxins. J. Chromatogr. B 803, 55-66.
257
Batista, C.V.F., Roman-Gonzalez, S.A., Salas-Castillo, S.P., Zamudio, F.Z., GomezLagunas, F., Possani, L.D., 2007. Proteomic analysis of the venom from the
scorpion Tityus stigmurus: biochemical and physiological comparison with other
Tityus species. Comp. Biochem. Physiol. C-Toxicol. Pharmacol. 146, 147-157.
Benson, D.A., Karsch-Mizrachi, I., Clark, K., Lipman, D.J., Ostell, J., Sayers, E.W.,
2012. GenBank. Nucleic Acids Res. 40, D48-D53.
38, 291.
Biass, D., Dutertre, S., Gerbault, A., Menou, J.L., Offord, R., Favreau, P., Stocklin, R.,
2009. Comparative proteomic study of the venom of the piscivorous cone snail
Conus consors. J. Proteomics 72, 210-218.
Bode, W., Gomisruth, F.X., Stockler, W., 1993. Astacins, serralysins, snake-venom and
matrix metalloproteinases exhibit identical zinc-binding environments
(HEXXHXXGXXH and MET-turn) and topologies and should be grouped into a
common family, the 'metzincins'. FEBS Lett. 331, 134-140.
Boldrini-Franca, J., Correa-Netto, C., Silva, M.M.S., Rodrigues, R.S., De La Torre, P.,
Perez, A., Soares, A.M., Zingali, R.B., Nogueira, R.A., Rodrigues, V.M., Sanz,
L., Calvete, J.J., 2010. Snake venomics and antivenomics of Crotalus durissus
subspecies from Brazil: assessment of geographic variation and its implication on
snakebite management. J. Proteomics 73, 1758-1776.
Borges, A., Miranda, R.J., Pascale, J.M., 2012. Scorpionism in Central America, with
special reference to the case of Panama. J. Venom. Anim. Tox. incl. Trop. Dis.
18, 130-143.
248-254.
Brady, R.M., Baell, J.B., Norton, R.S., 2013. Strategies for the development of
conotoxins as new therapeutic leads. Mar. Drugs 11, 2293-2313.
258
Brillard-Bourdet, M., Nguyen, V., Ferrer-Di Martino, M., Gauthier, F., Moreau, T., 1998.
Purification and characterization of a new cystatin inhibitor from Taiwan cobra
(Naja naja atra) venom. Biochem. J. 331, 239-244.
Bücherl, W., 1971. Introduction, in: Bücherl, W., Buckley, E.E. (Eds.), Venomous
Calvete, J.J., 2011. Proteomic tools against the neglected pathology of snake bite
envenoming. Expert Rev. Proteomics 8, 739-758.
Calvete, J.J., 2013. Snake venomics: from the inventory of toxins to biology. Toxicon 75,
44-62.
Calvete, J.J., Escolano, J., Sanz, L., 2007a. Snake venomics of Bitis species reveals large
intragenus venom toxin composition variation: application to taxonomy of
congeneric taxa. J. Proteome Res. 6, 2732-2745.
Calvete, J.J., Juarez, P., Sanz, L., 2007b. Snake venomics. Strategy and applications. J.
Mass Spectrom. 42, 1405-1414.
Cantacessi, C., Campbell, B.E., Visser, A., Geldhof, P., Nolan, M.J., Nisbet, A.J.,
Matthews, J.B., Loukas, A., Hofmann, A., Otranto, D., Sternberg, P.W., Gasser,
R.B., 2009. A portrait of the "SCP/TAPS" proteins of eukaryotes—developing a
framework for fundamental research and biotechnological outcomes. Biotechnol.
Adv. 27, 376-388.
Carregari, V.C., Floriano, R.S., Rodrigues-Simioni, L., Winck, F.V., Baldasso, P.A.,
Ponce-Soto, L.A., Marangoni, S., 2013. Biochemical, pharmacological, and
structural characterization of new basic PLA2 Bbil-TX from Bothriopsis bilineata
snake venom. Biomed Res. Int. DOI: 10.1155/2013/612649.
Casewell, N.R., Wuster, W., Vonk, F.J., Harrison, R.A., Fry, B.G., 2013. Complex
cocktails: the evolutionary novelty of venoms. Trends Ecol. Evol. 28, 219-229.
Castaneda, O., Harvey, A.L., 2009. Discovery and characterization of cnidarian peptide
toxins that affect neuronal potassium ion channels. Toxicon 54, 1119-1124.
Castle, N.A., Haylett, D.G., Jenkinson, D.H., 1989. Toxins in the characterization of
potassium channels. Trends Neurosci. 12, 59-65.
Chaim, O.M., Trevisan-Silva, D., Chaves-Moreira, D., Wille, A.C.M., Ferrer, V.P.,
Matsubara, F.H., Mangili, O.C., da Silveira, R.B., Gremski, L.H., Gremski, W.,
259
Senff-Ribeiro, A., Veiga, S.S., 2011. Brown Spider (Loxosceles genus) venom
toxins: tools for biological purposes. Toxins 3, 309-344.
Cheung, C.C., Loi, P.K., Sylwester, A.W., Lee, T.D., Tublitz, N.J., 1992. Primary
structure of a cardioactive neuropeptide from the tobacco hawkmoth, Manduca
sexta. FEBS Lett. 313, 165-168.
Chippaux, J.P., Goyffon, M., 2008. Epidemiology of scorpionism: a global appraisal.
Acta Trop. 107, 71-79.
Choo, Y.M., Lee, K.S., Yoon, H.J., Kim, B.Y., Sohn, M.R., Roh, J.Y., Je, Y.H., Kim,
N.J., Kim, I., Woo, S.D., Sohn, H.D., Jin, B.R., 2010. Dual function of a bee
venom serine protease: prophenoloxidase-activating factor in arthropods and
fibrin(ogen)olytic enzyme in mammals. PLoS One 5, e10393. DOI:
10.1371/journal.pone.0010393.
Chuang, R.S.I., Jaffe, H., Cribbs, L., Perez-Reyes, E., Swartz, K.J., 1998. Inhibition of Ttype voltage-gated calcium channels by a new scorpion toxin. Nat. Neurosci. 1,
668-674.
Cidade, D.A.P., Simao, T.A., Davila, A.M.R., Wagner, G., Junqueira-De-Azevedo,
I.D.M., Ho, P.L., Bon, C., Zingali, R.B., Albano, R.M., 2006. Bothrops jararaca
venom gland transcriptome: analysis of the gene expression pattern. Toxicon 48,
437-461.
Costal-Oliveira, F., Duarte, C.G., de Avila, R.A.M., Melo, M.M., Bordon, K.C.F.,
Arantes, E.C., Paredes, N.C., Tintaya, B., Bonilla, C., Bonilla, R.E., Suarez, W.S.,
Yarleque, A., Fernandez, J.M., Kalapothakis, E., Chavez-Olortegui, C., 2012.
General biochemical and immunological characteristics of the venom from
Peruvian scorpion Hadruroides lunatus. Toxicon 60, 934-942.
260
Natn. Mus. 111, 167-195.
Cruz, L.S., Vargas, R., Lopes, A.A., 2009. Snakebite envenomation and death in the
developing world. Ethn. Dis. 19, 42-46.
de Graaf, D.C., Aerts, M., Danneels, E., Devreese, B., 2009. Bee, wasp and ant venomics
pave the way for a component-resolved diagnosis of sting allergy. J. Proteomics
72, 145-154.
de Graaf, D.C., Brunain, M., Scharlaken, B., Peiren, N., Devreese, B., Ebo, D.G.,
Stevens, W.J., Desjardins, C.A., Werren, J.H., Jacobs, F.J., 2010. Two novel
proteins expressed by the venom glands of Apis mellifera and Nasonia vitripennis
share an ancient C1q-like domain. Insect Mol. Biol. 19, 1-10.
de la Vega, R.C.R., Possani, L.D., 2004. Current views on scorpion toxins specific for
K+-channels. Toxicon 43, 865-875.
de Lima, P.R., Brochetto-Braga, M.R., 2003. Hymenoptera venom review focusing on
Apis mellifera. J. Venom. Anim. Toxins 9, http://dx.doi.org/10.1590/S167891992003000200002.
de Oliveira, N.G., Cardoso, M., Franco, O.L., 2013. Snake venoms: attractive
antimicrobial proteinaceous compounds for therapeutic purposes. Cell. Mol. Life
Sci. 70, 4645-4658.
Boubee, Paris.
261
Deng, F., Fang, H., Wang, K., 1997. Hemolysis of Scolopendra toxins. J. Chi. Med.
Mater. 20, 36-37.
Devaraja, S., Girish, K.S., Devaraj, V.R., Kemparaju, K., 2010. Factor Xa-like and
fibrin(ogen)olytic activities of a serine protease from Hippasa agelenoides spider
venom gland extract. J. Thromb. Thrombolysis 29, 119-126.
Devaraja, S., Girish, K.S., Gowtham, Y.N.J., Kemparaju, K., 2011. The Hag-protease-II
is a fibrin(ogen)ase from Hippasa agelenoides spider venom gland extract:
purification, characterization and its role in hemostasis. Toxicon 57, 248-258.
Devaraja, S., Nagaraju, S., Mahadeswaraswamy, Y.H., Girish, K.S., Kemparaju, K.,
2008. A low molecular weight serine protease: purification and characterization
from Hippasa agelenoides (funnel web) spider venom gland extract. Toxicon 52,
130-138.
Dircksen, H., 1998. Conserved crustacean cardioactive peptide (CCAP) neuronal
networks and functions in arthropod evolution, in: Coast, G.M., Webster, S.G.
(Eds.), Recent advances in arthropod endocrinology. Cambridge University Press,
Cambridge, UK, pp. 302-333.
Doley, R., Zhou, X., Kini, R.M., 2010. Snake venom phospholipase A2 enzymes, in:
Mackessy, S.P. (Ed.), Venoms and toxins of reptiles. CRC Press, Boca Raton, pp.
173-205.
Dreyer, F., 1990. Peptide toxins and potassium channels. Rev. Physiol. Biochem.
Pharmacol. 115, 93-136.
Dulcis, D., Levine, R.B., Ewer, J., 2005. Role of the neuropeptide CCAP in Drosophila
cardiac function. J. Neurobiol. 64, 259-274.
Dunn, R.D., Broady, K.W., 2001. Snake inhibitors of phospholipase A2 enzymes.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1533, 29-37.
262
Escoubas, P., Celerier, M.L., Nakajima, T., 1997. High-performance liquid
chromatography matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry peptide fingerprinting of tarantula venoms in the genus
Brachypelma: chemotaxonomic and biochemical applications. Rapid Commun.
Escoubas, P., Quinton, L., Nicholson, G.M., 2008. Venomics: unravelling the complexity
of animal venoms with mass spectrometry. J. Mass Spectrom. 43, 279-295.
Escoubas, P., Rash, L., 2004. Tarantulas: eight-legged pharmacists and combinatorial
chemists. Toxicon 43, 555-574.
Escoubas, P., Sollod, B., King, G.F., 2006. Venom landscapes: mining the complexity of
spider venoms via a combined cDNA and mass spectrometric approach. Toxicon
47, 650-663.
Espino-Solis, G.P., Riano-Umbarila, L., Becerril, B., Possani, L.D., 2009. Antidotes
against venomous animals: state of the art and prospectives. J. Proteomics 72,
183-199.
Estevez-Lao, T.Y., Boyce, D.S., Honegger, H.W., Hillyer, J.F., 2013. Cardioacceleratory
function of the neurohormone CCAP in the mosquito Anopheles gambiae. J. Exp.
Biol. 216, 601-613.
Environ. 407, 6100-6108.
Forrester, J.A., Holstege, C.P., Forrester, J.D., 2012. Fatalities from venomous and
nonvenomous animals in the United States (1999-2007). Wilderness Environ.
Med. 23, 146-152.
Fox, J.W., Ma, L., Nelson, K., Sherman, N.E., Serrano, S.M.T., 2006. Comparison of
indirect and direct approaches using ion-trap and Fourier transform ion cyclotron
263
resonance mass spectrometry for exploring viperid venom proteomes. Toxicon
47, 700-714.
Fox, J.W., Serrano, S.M.T., 2007. Approaching the golden age of natural product
pharmaceuticals from venom libraries: an overview of toxins and toxinderivatives currently involved in therapeutic or diagnostic applications. Curr.
Pharm. Des. 13, 2927-2934.
Frazao, B., Vasconcelos, V., Antunes, A., 2012. Sea anemone (Cnidaria, Anthozoa,
Actiniaria) toxins: an overview. Mar. Drugs 10, 1812-1851.
Georgieva, D., Arni, R.K., Betzel, C., 2008. Proteome analysis of snake venom toxins:
pharmacological insights. Expert Rev. Proteomics 5, 787-797.
Gibbs, G.M., Roelants, K., O'Bryan, M.K., 2008. The CAP superfamily: cysteine-rich
secretory proteins, antigen 5, and pathogenesis-related 1 proteins—roles in
reproduction, cancer, and immune defense. Endocr. Rev. 29, 865-897.
Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1982. Pharmacodynamics of
venom of the centipede Scolopendra subspinipes dehaani Brandt. Indian J. Exp.
Biol. 20, 615-618.
Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1983. Isolation, purification
and pharmacodynamics of a toxin from the venom of the centipede Scolopendra
subspinipes dehaani Brandt. Indian J. Exp. Biol. 21, 203-207.
Guercio, R.A.P., Shevchenko, A., Lopez-Lozano, J.L., Paba, J., Sousa, M.V., Ricart,
C.A.O., 2006. Ontogenetic variations in the venom proteome of the Amazonian
snake Bothrops atrox. Proteome Sci. 4, DOI: 10.1186/1477-5956-4-11.
Gutierrez, J.M., Lomonte, B., Leon, G., Alape-Giron, A., Flores-Diaz, M., Sanz, L.,
Angulo, Y., Calvete, J.J., 2009. Snake venomics and antivenomics: Proteomic
tools in the design and control of antivenoms for the treatment of snakebite
envenoming. J. Proteomics 72, 165-182.
Gutierrez, J.M., Rucavado, A., Escalante, T., 2010. Snake venom metalloproteinases:
biological roles and participation in the pathophysiology of envenomation, in:
264
Mackessy, S.P. (Ed.), Handbook of venoms and toxins of reptiles. CRC Press,
Boca Raton, pp. 115-138.
Gutierrez, M.D., Abarca, C., Possani, L.D., 2003. A toxic fraction from Scolopendra
venom increases the basal release of neurotransmitters in the ventral ganglia of
crustaceans. Comp. Biochem. Physiol., C: Toxicol. Pharmacol. 135, 205-214.
Harada, S., Yoshizaki, Y., Natsuaki, M., Shimizu, H., Fukuda, H., Nagai, H., Ikeda, T.,
2005. Three cases of centipede allergy—analysis of cross reactivity with bee
allergy. Arerugi 54, 1279-1284.
Hariprasad, G., Srinivasan, A., Singh, R., 2013. Structural and phylogenetic basis for the
classification of group III phospholipase A2. J. Mol. Model. 19, 3779-3791.
Harvey, A.L., 2006. Snake venom peptides, in: Kastin, A.J. (Ed.), Handbook of
biologically active peptides. Academic Press, Burlington, pp. 355-362.
Harvey, A.L., Robertson, B., 2004. Dendrotoxins: structure-activity relationships and
effects on potassium ion channels. Curr. Med. Chem. 11, 3065-3072.
Hayden, L., Arthur, W., 2013. Expression patterns of Wnt genes in the venom claws of
centipedes. Evol. Dev. 15, 365-372.
Henriksen, A., King, T.P., Mirza, O., Monsalve, R.I., Meno, K., Ipsen, H., Larsen, J.N.,
Gajhede, M., Spangfort, M.D., 2001. Major venom allergen of yellow jackets,
Ves v 5: structural characterization of a pathogenesis-related protein superfamily.
Proteins 45, 438-448.
265
Junqueira, M., Spirin, V., Balbuena, T.S., Thomas, H., Adzhubei, I., Sunyaev, S.,
Shevchenko, A., 2008. Protein identification pipeline for the homology-driven
proteomics. J. Proteomics 71, 346-356.
Kimura, L.F., Prezotto-Neto, J.P., Tavora, B., Antoniazzi, M.M., Knysak, I., Guizze,
S.P.G., Santoro, M.L., Barbaro, K.C., 2013. Local inflammatory reaction induced
by Scolopendra viridicornis centipede venom in mice. Toxicon 76, 239-246.
King, G.F., 2011. Venoms as a platform for human drugs: translating toxins into
therapeutics. Expert Opin. Biol. Ther. 11, 1469-1484.
King, G.F., Hardy, M.C., 2013. Spider-venom peptides: structure, pharmacology, and
potential for control of insect pests, in: Berenbaum, M.R. (Ed.), Annual Review of
Entomology, Vol 58. Annual Reviews, Palo Alto, pp. 475-496.
King, T.P., Spangfort, M.D., 2000. Structure and biology of stinging insect venom
allergens. Int. Arch. Allergy Immunol. 123, 99-106.
Kinkawa, K., Shirai, R., Watanabe, S., Toriba, M., Hayashi, K., Ikeda, K., Inoue, S.,
2010. Up-regulation of the expressions of phospholipase A2 inhibitors in the liver
of a venomous snake by its own venom phospholipase A2. Biochem. Biophys.
Res. Commun. 395, 377-381.
303-311.
Kordis, D., Turk, V., 2009. Phylogenomic analysis of the cystatin superfamily in
eukaryotes and prokaryotes. BMC Evol. Biol. 9, DOI: 10.1186/1471-2148-9-266.
Kozlov, S.A., Vassilevski, A.A., Feofanov, A.V., Surovoy, A.Y., Karpunin, D.V.,
Grishin, E.V., 2006. Latarcins, antimicrobial and cytolytic peptides from the
venom of the spider Lachesana tarabaevi (Zodariidae) that exemplify
biomolecular diversity. J. Biol. Chem. 281, 20983-20992.
Kuhn-Nentwig, L., 2003. Antimicrobial and cytolytic peptides of venomous arthropods.
Cell. Mol. Life Sci. 60, 2651-2668.
Kuruppu, S., Smith, A.I., Isbister, G.K., Hodgson, W.C., 2008. Neurotoxins from
Australo-Papuan elapids: a biochemical and pharmacological perspective. Crit.
Rev. Toxicol. 38, 73-86.
266
Leontovich, A.A., Zhang, J.S., Shimokawa, K., Nagase, H., Sarras, M.P., 2000. A novel
Hydra matrix metalloproteinase (HMMP) functions in extracellular matrix
degradation, morphogenesis and the maintenance of differentiated cells in the foot
process. Development 127, 907-920.
Lewis, R.J., Dutertre, S., Vetter, I., Christie, M.J., 2012. Conus venom peptide
pharmacology. Pharmacol. Rev. 64, 259-298.
Li, R.L., Zhang, L., Fang, Y., Han, B., Lu, X.S., Zhou, T.E., Feng, M., Li, J.K., 2013.
Proteome and phosphoproteome analysis of honeybee (Apis mellifera) venom
collected from electrical stimulation and manual extraction of the venom gland.
BMC Genomics 14, DOI: 10.1186/1471-2164-14-766.
Sep. Sci. 33, 410-421.
Liao, Z., Cao, J., Li, S.M., Yan, X.J., Hu, W.J., He, Q.Y., Chen, J.J., Tang, J.Z., Xie, J.Y.,
Liang, S.P., 2007. Proteomic and peptidomic analysis of the venom from Chinese
tarantula Chilobrachys jingzhao. Proteomics 7, 1892-1907.
Lima, R.M., Estevao-Costa, M.I., Junqueira-de-Azevedo, I.L.M., Ho, P.L., Diniz,
M.R.V., Fortes-Dias, C.L., 2011. Phospholipase A2 inhibitors (beta PLIs) are
encoded in the venom glands of Lachesis muta (Crotalinae, Viperidae) snakes.
Toxicon 57, 172-175.
Liska, A.J., Shevchenko, A., 2003. Expanding the organismal scope of proteomics: crossspecies protein identification by mass spectrometry and its implications.
Proteomics 3, 19-28.
267
Lizano, S., Angulo, Y., Lomonte, B., Fox, J.W., Lambeau, G., Lazdunski, M., Gutierrez,
J.M., 2000. Two phospholipase A2 inhibitors from the plasma of Cerrophidion
(Bothrops) godmani which selectively inhibit two different group-II
phospholipase A2 myotoxins from its own venom: isolation, molecular cloning
and biological properties. Biochem. J. 346, 631-639.
Mackessy, S.P., 2010. Evolutionary trends in venom composition in the Western
Rattlesnakes (Crotalus viridis sensu lato): toxicity vs. tenderizers. Toxicon 55,
1463-1474.
Toxicon 55, 250-255.
McCleary, R.J.R., Kini, R.M., 2013. Non-enzymatic proteins from snake venoms: a gold
mine of pharmacological tools and drug leads. Toxicon 62, 56-74.
Raton.
268
Medeiros, C.R., Susaki, T.T., Knysak, I., Cardoso, J.L.C., Malaque, C.M.S., Fan, H.W.,
Santoro, M.L., Franca, F.O.S., Barbaro, K.C., 2008. Epidemiologic and clinical
survey of victims of centipede stings admitted to Hospital Vital Brazil (Sao Paulo,
Brazil). Toxicon 52, 606-610.
Miloslavina, A., Ebert, C., Tietze, D., Ohlenschlager, O., Englert, C., Gorlach, M., Imhof,
D., 2010. An unusual peptide from Conus villepinii: synthesis, solution structure,
and cardioactivity. Peptides 31, 1292-1300.
Mohamed, A.H., Zaid, E., Elbeih, N.M., Elaal, A.A., 1980. Effects of an extract from the
centipede Scolopendra moristans on intestine, uterus and heart contractions and
on blood-glucose and liver and muscle glycogen levels. Toxicon 18, 581-589.
Moller, C., Melaun, C., Castillo, C., Diaz, M.E., Renzelman, C.M., Estrada, O., Kuch, U.,
Lokey, S., Mari, F., 2010. Functional hypervariability and gene diversity of
cardioactive neuropeptides. J. Biol. Chem. 285, 40673-40680.
Murphy, G., Nagase, H., 2008. Progress in matrix metalloproteinase research. Mol. Asp.
Med. 29, 290-308.
Nascimento, D.G., Rates, B., Santos, D.M., Verano-Braga, T., Barbosa-Silva, A., Dutra,
A.A.A., Biondi, I., Martin-Eauclaire, M.F., De Lima, M.E., Pimenta, A.M.C.,
2006. Moving pieces in a taxonomic puzzle: venom 2D-LC/MS and data
clustering analyses to infer phylogenetic relationships in some scorpions from the
Buthidae family (Scorpiones). Toxicon 47, 628-639.
269
Neves-Ferreira, A.G.C., Valente, R.H., Perales, J., Domont, G.B., 2010. Natural
inhibitors: innate immunity to snake venoms, in: Mackessy, S.P. (Ed.), Venoms
and toxins of reptiles. CRC Press, Boca Raton, pp. 259-284.
Ochieng, J., Chaudhuri, G., 2010. Cystatin superfamily. J. Health Care Poor Underserved
21, 51-70.
Olamendi-Portugal, T., Garcia, B.I., Lopez-Gonzalez, I., Van Der Walt, J., Dyason, K.,
Ulens, C., Tytgat, J., Felix, R., Darszon, A., Possani, L.D., 2002. Two new
scorpion toxins that target voltage-gated Ca2+ and Na+ channels. Biochem.
Biophys. Res. Commun. 299, 562-568.
Oukkache, N., Rosso, J.P., Alami, M., Ghalim, N., Saile, R., Hassar, M., Bougis, P.E.,
Martin-Eauclaire, M.F., 2008. New analysis of the toxic compounds from the
Androctonus mauretanicus mauretanicus scorpion venom. Toxicon 51, 835-852.
introduction, in: Buccianti, A., Mateu-Figueras, G., Pawlowsky-Glahn, V. (Eds.),
Phillips, D.J., Swenson, S.D., Markland, F.S., Jr., 2010. Thrombin-like snake venom
serine proteinases, in: Mackessy, S.P. (Ed.), Handbook of venoms and toxins of
reptiles. CRC Press, Boca Raton, pp. 139-154.
Pimenta, A.M.C., Stocklin, R., Favreau, P., Bougis, P.E., Martin-Eauclaire, M.F., 2001.
Moving pieces in a proteomic puzzle: mass fingerprinting of toxic fractions from
the venom of Tityus serrulatus (Scorpiones, Buthidae). Rapid Commun. Mass
Spectrom. 15, 1562-1572.
270
Prashanth, J.R., Lewis, R.J., Dutertre, S., 2012. Towards an integrated venomics
approach for accelerated conopeptide discovery. Toxicon 60, 470-477.
Quinton, L., Demeure, K., Dobson, R., Gilles, N., Gabelica, V., De Pauw, E., 2007. New
method for characterizing highly disulfide-bridged peptides in complex mixtures:
application to toxin identification from crude venoms. J. Proteome Res. 6, 32163223.
Rates, B., Ferraz, K.K.F., Borges, M.H., Richardson, M., De Lima, M.E., Pimenta,
A.M.C., 2008. Tityus serrulatus venom peptidomics: assessing venom peptide
diversity. Toxicon 52, 611-618.
Richards, R., St Pierre, L., Trabi, M., Johnson, L.A., de Jersey, J., Masci, P.P., Lavin,
M.F., 2011. Cloning and characterisation of novel cystatins from elapid snake
venom glands. Biochimie 93, 659-668.
Romi-Lebrun, R., Martin-Eauclaire, M.F., Escoubas, P., Wu, F.Q., Lebrun, B., Hisada,
M., Nakajima, T., 1997. Characterization of four toxins from Buthus martensi
scorpion venom, which act on apamin-sensitive Ca2+-activated K+ channels. Eur.
J. Biochem. 245, 457-464.
Samy, R.P., Pachiappan, A., Gopalakrishnakone, P., Thwin, M.M., Hian, Y.E., Chow,
V.T.K., Bow, H., Weng, J.T., 2006. In vitro antimicrobial activity of natural
toxins and animal venoms tested against Burkholderia pseudomallei. BMC Infect.
Dis. 6, DOI: 10.1186/1471-2334-6-100.
271
Santos, L.D., Pieroni, M., Menegasso, A.R.S., Pinto, J., Palma, M.S., 2011. A new
scenario of bioprospecting of Hymenoptera venoms through proteomic approach.
J. Venom. Anim. Tox. incl. Trop. Dis. 17, 364-377.
Schwartz, E.F., Bartok, A., Schwartz, C.A., Papp, F., Gomez-Lagunas, F., Panyi, G.,
Possani, L.D., 2013. OcyKTx2, a new K+- channel toxin characterized from the
venom of the scorpion Opisthacanthus cayaporum. Peptides 46, 40-46.
Schwartz, E.F., Camargos, T.S., Zamudio, F.Z., Silva, L.P., Bloch, C., Caixeta, F.,
Schwartz, C.A., Possani, L.D., 2008. Mass spectrometry analysis, amino acid
sequence and biological activity of venom components from the Brazilian
scorpion Opisthacanthus cayaporum. Toxicon 51, 1499-1508.
385-405.
Senko, M.W., Beu, S.C., McLafferty, F.W., 1995. Determination of monoisotopic masses
and ion populations for large biomolecules from resolved isotopic distributions. J.
Am. Soc. Mass Spectrom. 6, 229-233.
Serrano, S.M.T., Maroun, R.C., 2005. Snake venom serine proteinases: sequence
homology vs. substrate specificity, a paradox to be solved. Toxicon 45, 11151132.
Martinsville.
Shevchenko, A., Valcu, C.M., Junqueira, M., 2009. Tools for exploring the
proteomosphere. J. Proteomics 72, 137-144.
Sousa, J.R.F., Monteiro, R.Q., Castro, H.C., Zingali, R.B., 2001. Proteolytic action of
Bothrops jararaca venom upon its own constituents. Toxicon 39, 787-792.
Sousa, S.R., Vetter, I., Lewis, R.J., 2013. Venom peptides as a rich source of Ca(v)2.2
channel blockers. Toxins 5, 286-314.
272
Souza, G., Catharino, R.R., Ifa, D.R., Eberlin, M.N., Hyslop, S., 2008. Peptide
fingerprinting of snake venoms by direct infusion nano-electrospray ionization
mass spectrometry: potential use in venom identification and taxonomy. J. Mass
Spectrom. 43, 594-599.
Sunagar, K., Johnson, W.E., O'Brien, S.J., Vasconcelos, V., Antunes, A., 2012. Evolution
of CRISPs associated with toxicoferan-reptilian venom and mammalian
reproduction. Mol. Biol. Evol. 29, 1807-1822.
Sunagar, K., Undheim, E.A.B., Scheib, H., Gren, E.C.K., Cochran, C., Person, C.E.,
Koludarov, I., Kelln, W., Hayes, W.K., King, G.F., Antunes, A., Fry, B.G., 2014.
Intraspecific venom variation in the medically significant Southern Pacific
Rattlesnake (Crotalus oreganus helleri): biodiscovery, clinical and evolutionary
implications. J. Proteomics DOI:10.1016/j.jprot.2014.01.013.
Tang, X., Zhang, Y.Q., Hu, W.J., Xu, D.H., Tao, H., Yang, X.X., Li, Y., Jiang, L.P.,
Liang, S.P., 2010. Molecular diversification of peptide toxins from the tarantula
Haplopelma hainanum (Ornithoctonus hainana) venom based on transcriptomic,
peptidomic, and genomic analyses. J. Proteome Res. 9, 2550-2564.
Tashima, A.K., Sanz, L., Camargo, A.C.M., Serrano, S.M.T., Calvete, J.J., 2008. Snake
venomics of the Brazilian pitvipers Bothrops cotiara and Bothrops fonsecai.
Identification of taxonomy markers. J. Proteomics 71, 473-485.
Terlau, H., Olivera, B.M., 2004. Conus venoms: a rich source of novel ion channeltargeted peptides. Physiol. Rev. 84, 41-68.
Trevisan-Silva, D., Gremski, L.H., Chaim, O.M., da Silveira, R.B., Meissner, G.O.,
Mangili, O.C., Barbaro, K.C., Gremski, W., Veiga, S.S., Senff-Ribeiro, A., 2010.
Astacin-like metalloproteases are a gene family of toxins present in the venom of
different species of the brown spider (genus Loxosceles). Biochimie 92, 21-32.
273
Veiga, S.S., da Silveira, R.B., Dreyfuss, J.L., Haoach, J., Pereira, A.M., Mangili, O.C.,
Gremski, W., 2000. Identification of high molecular weight serine-proteases in
Loxosceles intermedia (brown spider) venom. Toxicon 38, 825-839.
Veiga, S.S., Zanetti, V.C., Braz, A., Mangili, O.C., Gremski, W., 2001. Extracellular
matrix molecules as targets for brown spider venom toxins. Braz. J. Med. Biol.
Res. 34, 843-850.
Vetter, I., Davis, J.L., Rash, L.D., Anangi, R., Mobli, M., Alewood, P.F., Lewis, R.J.,
King, G.F., 2011. Venomics: a new paradigm for natural products-based drug
discovery. Amino Acids 40, 15-28.
London.
Ward, D.A., Sefton, E.M., Prescott, M.C., Webster, S.G., Wainwright, G., Rees, H.H.,
Fisher, M.J., 2010. Efficient identification of proteins from ovaries and
hepatopancreas of the unsequenced edible crab, Cancer pagurus, by mass
spectrometry and homology-based, cross-species searching. J. Proteomics 73,
2354-2364.
Waridel, P., Frank, A., Thomas, H., Surendranath, V., Sunyaev, S., Pevzner, P.,
Shevchenko, A., 2007. Sequence similarity-driven proteomics in organisms with
unknown genomes by LC-MS/MS and automated de novo sequencing.
Wasielewski, O., Skonieczna, M., 2008. Pleiotropic effects of the neuropeptides CCAP
and myosuppressin in the beetle, Tenebrio molitor L. J. Comp. Physiol. BBiochem. Syst. Environ. Physiol. 178, 877-885.
274
Watson, J.T., Sparkman, O.D., 2007. Introduction to mass spectrometry: instrumentation,
applications and strategies for data interpretation. John Wiley & Sons, Chichester.
Weinstein, S.A., White, J., Keyler, D.E., Warrell, D.A., 2013. Non-front-fanged
colubroid snakes: a current evidence-based analysis of medical significance.
Toxicon 69, 103-113.
Wenhua, R., Shuangguan, Z., Daxiang, S., Kaiya, Z., Guang, Y., 2006. Induction,
purification and characterization of an antibacterial peptide scolopendrin I from
the venom of centipede Scolopendra subspinipes mutilans. Indian J. Biochem.
Biophys. 43, 88-93.
Yildiz, A., Biceroglu, S., Yakut, N., Bilir, C., Akdemir, R., Akilli, A., 2006. Acute
myocardial infarction in a young man caused by centipede sting. Emerg. Med. J.
23, e30.
You, W.K., Sohn, Y.D., Kim, K.Y., Park, D.H., Jang, Y., Chung, K.H., 2004.
Purification and molecular cloning of a novel serine protease from the centipede,
Scolopendra subspinipes mutilans. Insect Biochem. Mol. Biol. 34, 239-250.
275
CHAPTER FIVE
CONCLUSIONS
In this dissertation, I investigated a number of factors expected to impact the
behavioral ecology of venom use in the centipedes Scolopendra polymorpha and S.
subspinipes, including venom yield, timing of venom regeneration, and venom
composition. To date, my work represents the first and most comprehensive studies of
venom yield and venom regeneration in any centipede, and the first proteomic
investigation of the venom of S. polymorpha. In this chapter, I will revisit the most
important conclusions from each major study.
In Chapter 2, I investigated how size, specifically body length, influenced volume
yield and protein concentration of electrically extracted venom in S. polymorpha and S.
subspinipes. I also examined additional potential influences on yield in S. polymorpha,
including relative forcipule size, relative mass, geographic origin (Arizona vs.
California), sex, time in captivity, and milking history. Centipede body length was the
single most important factor determining venom volume yield, generating approximately
2-fold (S. polymorpha) to 4.8-fold (pooled data) differences in the simple linear
regression equation. Consistent with my hypothesis, volume yield in S. polymorpha
(range 0.05–3.3 µL) was positively and linearly related to body length (range 5.6–11.5
cm). The regression equation, V = 0.36(L) - 2.08, revealed that for every 1-cm increase in
body length (within the size range of specimens studied), an additional 0.36 (95% CI =
276
0.30–0.42) µL venom could be extracted. The narrow size range of S. subspinipes (14.1–
15.8 cm) rendered the intraspecific relationship between body size and venom yield
(range 4.1–7.3 µL) unreliable. Using pooled data from both species, the regression
equation, V = 0.52(L) - 3.42, suggested that, for every 1-cm increase in body length
(within the size range of specimens studied), 0.52 (95% CI = 0.46–0.57) µL more venom
could be extracted. Data also revealed a significantly smaller length-specific yield from S.
polymorpha than S. subspinipes (adjusted means, 1.1 [1.0–1.1] vs. 1.7 [1.1–2.3] µL,
respectively). Body length and protein concentration were uncorrelated. When
considering multiple influences on volume yield (i.e., multiple regression model) in S.
polymorpha, the most important factor was body length, but yield was also positively
associated with relative forcipule length and relative body mass. Scolopendra
polymorpha from California yielded a greater volume of venom than conspecifics from
Arizona, (adjusted means, 1.4 ± 0.1 vs. 1.0 ± 0.04 µL, respectively), as well as a more
concentrated venom (adjusted means, 188 ± 6 vs. 152 ± 3 µg/µL, respectively), all else
being equal. Milking history was also a significant predictor of yield and protein
concentration, with both measures higher from unmilked centipedes than from previously
milked animals (volume yield: adjusted means, 1.1 ± 0.03 vs. 0.7 ± 0.1 µL, respectively;
protein concentration: adjusted means, 164 ± 3 vs. 105 ± 12 respectively). For both
species, approximately two-thirds of extractable venom was expressed in the first two
pulses, with remaining pulses yielding declining amounts. Venom protein concentration,
in contrast, did not vary across pulses. Collectively, these findings demonstrate that
venom yield and concentration are influenced by numerous factors, but that body length
can be the most useful predictor of venom yield. Additionally, the implications for
277
centipede-sting risk are simple to infer given the importance of body length in
determining venom yield: severity of envenomation is likely greater with larger
centipedes, all else being equal. Combined with previously published LD50 and ED50
values, my venom yield data indicate that S. polymorpha has the potential to kill a large
number of prey.
In Chapter 3, I investigated venom volume and total protein regeneration during
the 14-day period subsequent to venom extraction in S. polymorpha. During the first 48
hours, volume and protein mass increased linearly. However, protein regeneration lagged
behind volume regeneration, with only 65–86% of venom volume and 29–47% of protein
mass regenerated during the first 2 days. No significant additional regeneration occurred
over the subsequent 12 days, with volume and protein concentration remaining at 86%
and 46% of initial levels, respectively. Neither volume nor protein mass reached initial
levels 7 months later (93% and 76%, respectively). In the first 48 hours, percent volume
regenerated increased 0.9% (95% CI = 0.6–1.3%) for every 1-hour increase in milking
interval, and decreased 6.0% (95% CI = -10.1– -1.9) for every 1-cm increase in centipede
body length. Also in the first 48 hours, percent protein mass regenerated increased 0.4%
(95% CI = 0.2–0.6%) for every 1-hour increase in milking interval, and decreased 2.8%
(95% CI = -5.1– -0.5%) for every 1-cm increase in body length. Over 14 days, percent
volume and percent protein mass regenerated decreased 7.9% (95% CI = -14.1– -1.8%)
and 7.3% (95% CI = -12.0– -2.6%), respectively, for every 1-cm increase in body length.
Analysis of chromatograms of individual venom samples revealed that five of 10
chromatographic regions, and 12 of 28 peaks, demonstrated changes in percent of total
peak area (i.e., percent of total protein) among milking intervals, indicating that venom
278
proteins are regenerated asynchronously, as hypothesized. The majority of literature
examples from other venomous animal taxa also suggest asynchronous venom
component regeneration. Taken together, these results indicated that, while volume
regenerates more quickly than protein mass during the initial 14 days after milking, the
reverse is true thereafter, a pattern that would be expected of a venom gland nearing
repletion but undergoing continued protein synthesis. Explanations for why venom
volume and protein mass did not return to initial levels within 7 months could include a
long replenishment cycle, senescence, or damage to the venom glands caused by
electrical milking. It remains unclear why venom regeneration is slower for longer,
presumably older, centipedes, although I speculate that it may be related to a change in
the ratio of venom secretory cells to venom storage volume during ontogeny.
If these results can be extrapolated to other centipedes, there are two major
implications for venom extraction protocols. First, a milking interval of between 2 and 14
days would allow a researcher to obtain a major percentage of initial volume; however,
venom milked at intervals of 14 days or less are predicted to have reduced protein content
in comparison to initially milked venom. Second, as venom protein components are
regenerated asynchronously, pools of venom derived from repeated milkings at short
intervals may show artificial enrichment of the more rapidly regenerated components in
comparison to initially milked venom.
In Chapter 4, I characterized the venom composition of S. polymorpha using
proteomic methods, including SDS-PAGE, RP-FPLC, mass spectrometry, and amino
acid sequence similarity searching. The venom of S. polymorpha proved to be highly
complex, generating up to 23 bands (approximate range: 5–240 kDa) by SDS-PAGE. In
279
reduced form, the large, densely-staining region at 79–105 kDa was absent, and the
region between 39–52 kDa became more diffuse and densely stained, a prominent band
appeared at 67 kDa, and the band at approximately 6 kDa grew darker and more diffuse.
These data indicate some of the higher mass components are polymeric proteins. MALDI
TOF MS analysis of unfractionated venom showed approximately 20 peaks primarily
localized between m/z 2,500 and 8,500 in the m/z range 1,000–20,000, and only a few
peaks, including a prominent peak near m/z 20,700, from the m/z 20,000–100,000
window. Analysis of venom by RP-FPLC led to the collection of 51 fractions, with
fifteen relatively hydrophobic fractions (33–47; eluting between 33.2–45.5 mL, 44.2–
62.7% buffer B) representing the majority of protein (75.2% of total peak area
combined). MALDI TOF MS of unfractionated venom revealed an estimated 935
components, with masses ranging from 1,014.5 to 82,863.9 Da. Many of the most
abundantly formed ions from the venom ranged between m/z 1,000–10,000 and eluted
prior to fraction 30 (42.5% buffer B), whereas most higher mass ions (i.e., m/z > 21,000)
were present at low abundance. The distribution of molecular masses was heavily skewed
toward smaller peptides and proteins, with 72% of components found below 12 kDa,
representing components less than ~107 amino acids in length. Although Mascot
searching following both LC-MS/MS and MALDI TOF/TOF MS/MS yielded no
significant centipede- or venom toxin-related protein hits, BLASTp sequence similarity
searching of MS/MS-derived amino acid sequences demonstrated 20 different sequences
with similarity to known venom components, including serine proteases, Kv channel
inhibitors, PLA2, CRISP-like proteins, a Cav channel activator, and several putative
neurotoxins of unknown function. As a whole, these findings add to the small but
280
growing literature demonstrating the complexity and novelty of centipede venoms.
Furthermore, the findings underscore the need for more complete database sequence
information based on additional proteomic and transcriptomic investigations of centipede
venoms.
Future Directions
While this dissertation is intended to provide insights into venom yield,
regeneration, and composition in centipedes, in the process it has also highlighted areas
for future research.
In Chapter 2, I found a linear relationship between centipede body length and
venom volume yield for S. polymorpha. It would be fascinating to determine what shape
this relationship takes in other centipedes, and if linear, whether the slopes are similar.
Comparing how venom supply varies with size across several species may highlight
differences in gland morphology during ontogeny, and perhaps shed light on differences
between species in selection pressures related to predation and defense with age. To
relate venom yield to ontogeny, a need also exists for investigating the relationship
between centipede size and developmental stage; future studies relating morphometrics to
post-larval stadia would be valuable in determining the relative contributions of size and
ontogeny to venom yield.
In Chapter 2, I also found a positive relationship between relative forcipule length
and volume yield. Future studies could use 3D tissue imaging techniques to explore how
venom gland shape and size vary with centipede body length, and how gland dimensions
influence venom yield. Future research, perhaps employing venom gland dissections and
histological preparations, could also examine the extent to which “complete” electrical
281
venom extraction (the goal in my investigations) actually empties the venom glands of
centipedes.
Differences in venom volume yield and protein concentration between California
and Arizona populations of S. polymorpha raise several questions for future exploration:
What differences in selection pressures may be operating in the two populations leading
to these venom differences? Do differences in yield and protein concentration exist
amongst other populations of S. polymorpha across its range?
I further found that previously milked centipedes yielded a smaller volume of
venom with a lower protein concentration than unmilked animals. It would be useful for
future research to determine the short-term and long-term effects on yield of repeated
venom extractions at various intervals. Such data would be helpful to researchers
devising venom extraction protocols.
In Chapter 3, I investigated venom regeneration. Future studies in this area could
compare rates of regeneration among centipede species; differences in the timing of
regeneration may be associated with differences in the frequency and quantity of venom
expenditure. It would also be interesting to determine why venom regeneration is slower
for longer centipedes. Perhaps careful anatomical and histological comparisons of venom
glands from a wide size range of centipedes combined with transcriptomic analyses might
provide hints as to the cause of this phenomenon. Future investigations could also strive
to determine whether electrical venom extraction is damaging to the centipedes’ venom
glands or venom gland musculature, which my data hinted at. Gland dissections and
histological preparations would likely play a key role in such studies. In addition, if it
were possible to extract venom from centipedes without electrical shock, it would be
282
enlightening to compare regeneration under both extraction conditions. Such an effort
would be helpful in determining to what extent the electrical extraction technique itself
may be influencing venom synthesis.
I also found evidence of asynchronous regeneration of venom components based
on changes in relative areas of chromatographic regions and individual peaks from RPFPLC. As more centipede venom components become fully characterized and added to
databases, future research utilizing transcriptomics and imaging mass spectrometry may
illuminate the time course of regeneration of important proteins in the venom of S.
polymorpha. It would be helpful to understand which particular toxins are regenerated
more quickly than others, and the reasons for the differences. Regeneration rate may
relate to protein size or complexity (Nisani et al., 2012), or to functional needs, with the
most critical toxins regenerating more quickly. Examining gene expression would shed
light on these issues.
In Chapter 4, I characterized the venom of S. polymorpha using proteomic
methods. Future studies on the venom composition of S. polymorpha would be improved
through additional separation steps orthogonal to reversed-phase chromatography (i.e.,
2D-PAGE, ion exchange chromatography, size exclusion chromatography) prior to mass
spectrometry analysis. These steps will facilitate the purification of components, leading
to improved mass spectra and setting the stage for protein sequencing and functional
assays. Given the important role that matrix selection can play in MALDI TOF MS,
future investigations should also consider whether additional S. polymorpha venom
components might be characterized by employing different matrices. In addition, a more
complete picture of venom composition could be achieved by also analyzing fractionated
283
venom for whole protein masses using electrospray ionization mass spectrometry (ESIMS), a technique known to be complementary to MALDI TOF MS. Future analyses
should also aim to derive an S. polymorpha venom gland transcriptome to function as a
database to search venom protein mass spectrometry data against. Finally, if methods can
be devised to obtain venom from S. polymorpha without electrical stimulation, it would
be enlightening to compare the composition of this secretion with electrically extracted
venom.
284
References
285
APPENDIX A
STRATEGIC DEPLOYMENT OF PREDATORY AND DEFENSIVE EMISSIONS:
A REVIEW OF THE LITERATURE
Allen M. Cooper
Corresponding author:
Allen M. Cooper
Griggs Hall #101
24941 Stewart St.
USA
Tel: 1-707-225-5018
Fax: 1-909-558-0259
286
Abstract
Due to the metabolic cost of their synthesis and their value in subduing prey and
combating aggressors, emissions such as venoms, predatory glues, and non-venomous
defensive secretions should be discharged in ways that increase their effectiveness and
avoid waste. Based on assessment of external environment and internal state, animals are
expected to deploy an emission (1) only under certain conditions, (2) in an amount that
can vary with circumstances, (3) from the location on the emitter’s body most proximate
to the triggering stimulus (when multiple emission locations are possible), (4) specifically
aimed toward the intended receiver, and (5) in a manner that allows for recovery or
reuptake of emitted material. For many venomous animals, including scorpions, spiders,
hymenopterans, and snakes, both the threshold for venom deployment and the amount
deployed are related to prey size and intensity/duration of prey struggle. Spitting cobras
demonstrate great accuracy in aiming venom spat in defense. For scytodid spiders and
onychophoran worms, the threshold for predatory glue squirting and the amount of glue
deployed depend on prey size and struggle. Animals employing non-venomous defensive
secretions often perform evasive behaviors prior to releasing chemical defenses, and
repeated harassment, threats directed at the body (as opposed to appendages), and
frequently encountered predators are most likely to elicit discharges of defensive
secretions. The amount of defensive secretion deployed is most often varied by the
number of discrete emissions, with the number of emissions matching the persistence of
threat stimuli. In many cases where multiple points of emission of defensive secretions
are possible, the gland(s) responding are those nearest the location of attack. Typically,
animals that spray defensive secretions accurately target whichever parts of their bodies
are under attack. Several invertebrates use eversible glands or cuticular projections to
287
present defensive secretions in a manner that allows for reuptake of secretion when threat
abates.
288
CONTENTS
Introduction ....................................................................................................................291
Strategic deployment of venom .....................................................................................296
Scorpions.................................................................................................................298
Cost of venom in scorpions............................................................................298
Selective sting/venom use by scorpions ........................................................299
Amount of venom deployed by scorpions .....................................................302
Delivery location and aiming of venom by scorpions ...................................304
Spiders.....................................................................................................................306
Cost of venom in spiders................................................................................306
Selective venom use by spiders .....................................................................306
Amount of venom deployed by spiders .........................................................309
Delivery location of venom by spiders ..........................................................317
Hymenopterans .......................................................................................................320
Wasps.............................................................................................................320
Cost of venom in wasps ........................................................................320
Selective sting/venom use by wasps .....................................................321
Amount of venom deployed by wasps..................................................323
Delivery location and aiming of venom by wasps................................326
Ants ................................................................................................................328
Cost of venom in ants ...........................................................................328
Selective sting/venom use by ants ........................................................329
Amount of venom deployed by ants .....................................................333
Delivery location of venom by ants ......................................................337
Snakes .....................................................................................................................338
Cost of venom in snakes ................................................................................338
Selective venom use by snakes ......................................................................339
Amount of venom deployed by snakes..........................................................342
Delivery location and aiming of venom by snakes........................................343
Strategic deployment of non-venomous predatory emissions .........................................346
Glue.........................................................................................................................346
289
Scytodid spiders .............................................................................................346
Velvet worms .................................................................................................348
Strategic deployment of non-venomous defensive secretions .........................................351
Cost of non-venomous defensive secretions...........................................................352
Selective deployment of non-venomous defensive secretions................................353
Amount of non-venomous defensive secretion deployed.......................................369
Location of emission from the emitter’s body........................................................385
Aiming of defensive secretions...............................................................................390
Reuptake of defensive secretions............................................................................399
Conclusion .......................................................................................................................401
References........................................................................................................................403
290
Introduction
Animals employ a variety of emissions in a variety of ways to procure food and
defend themselves. Some of the better-studied examples include venom injection and
discharge of chemical repellents for defense. In many cases where an emission represents
a valuable resource there is evidence that animals perform strategic emissions, that is,
deployment of emissions in a manner that avoids waste and maximizes their effectiveness
(Clements and Li, 2005; Hayes, 2008; Morgenstern and King, 2013; Nolen et al., 1995;
Read and Hughes, 1987). The purpose of this review is to explore and present examples
of some of the many ways in which a diversity of animals, including invertebrates and
vertebrates, strategically deploy predatory and defensive emissions. Rather than
attempting to be exhaustive, this review instead serves to highlight some of the major
areas of research that have uncovered evidence of strategic emission. This review will
consider strategic deployment of predatory and defensive venom, predatory glues, and
non-venomous defensive secretions, including various repellents and toxins,
hemolymph/blood, mucus, ink, glue, and slime.
In the field of behavioral ecology, cost-benefit analysis is an important component
in the evaluation of behavior, with natural selection favoring strategies having an
advantageous cost-benefit ratio (Cuthill and Houston, 1997). Predatory and defensive
emissions are expected to be strategically deployed, in a functional (ultimate) sense,
because natural selection favors behaviors that maximize available resources and
minimize costs. Discharging an emission is expected to entail a cost for the emitting
organism, whether the emission is for the purpose of predation (Morgenstern and King,
2013; Nisani et al., 2007; Read and Hughes, 1987; Suter and Stratton, 2012), or for
defense (Berenbaum, 1995; Higginson and Ruxton, 2009; Ruxton et al., 2004). If the
291
emission is synthesized by the organism itself, there is a metabolic cost in generating the
emission (e.g., Enzor et al., 2011; Inceoglu et al., 2003; McCue, 2006; Morgenstern and
King, 2013; Nisani et al., 2007; Pintor et al., 2010), and whether the emission is made by
the organism or acquired exogenously, there may be costs associated with storage
(Higginson and Ruxton, 2009; Inceoglu et al., 2003), prevention of autotoxicity (if the
emission is toxic; Bowers, 1992; McCue, 2006), and delivery/discharge of the emission
(Berenbaum, 1995; Bowers, 1992). In addition to metabolic costs, there are also
ecological costs associated with emission discharge. Unnecessary discharge, be it at
inappropriate times or in excessive amounts, represents a squandering of resources
(energy and materials) that can negatively influence the ability of the organism to defend
itself or take advantage of future opportunities to procure prey (Currier et al., 2012;
Haight and Tschinkel, 2003; Hayes, 2008; Malli et al., 1998). Because of the metabolic
and ecological costs associated with discharge of emissions, selection pressures are
expected to tune organisms to deploy emissions strategically (Morgenstern and King,
2013) in terms of circumstance, amount, discharge location, accuracy of delivery, and
recovery. In other words, for the purposes of this review, strategic deployment may
involve any of the following: discharging an emission (1) only under certain conditions,
(2) in an amount that can vary with circumstances, (3) from the location on the emitter’s
body most proximate to the triggering stimulus (when multiple emission locations are
possible), (4) specifically aimed toward the intended receiver, and (5) in a manner that
allows for recovery or reuptake of emitted material.
In order for an organism to deploy an emission in a strategic manner, as defined
above, it must be able to accurately assess both its external environment, including the
292
intended receiver of the emission, as well as its own internal state (Higginson and
Ruxton, 2009; Hostettler and Nentwig, 2006; Wullschleger and Nentwig, 2002).
Obviously, which factors are important to assess will vary based on the type of emission
to be deployed. External factors assessed for deployment of predatory emissions may
include type of prey (Carlin and David, 1989; Cushing and Matherne, 1980; Hayes,
1992b; Wigger et al., 2002), prey size/mass (Budriene and Budrys, 2005; Casiraghi et al.,
2001; Clements and Li, 2005; Daly-Schveitzer et al., 2007; Dejean, 1990; Edmunds and
Sibly, 2010; Hayes et al., 1995; McCormick and Polis, 1990; Steiner, 1986), and prey
struggle intensity/duration (Clements and Li, 2005; Dejean, 1988; Djieto-Lordon et al.,
2001; Malli et al., 1999; Read and Hughes, 1987; Steiner, 1986). External factors
assessed for deployment of defensive emissions may include type of predator/aggressor
(Carlin and David, 1989; Derby, 2007; Jeanne and Keeping, 1995; Moore and Williams,
1990), degree of threat (Haight, 2006; Haight and Tschinkel, 2003; Machado and Pomini,
2008; Nisani and Hayes, 2011; Nolen et al., 1995; Rehling, 2002; Woodring and Blum,
1965), location of threat (Bateman and Fleming, 2009; Blum and Woodring, 1962;
Eisner, 1965; Lim et al., 2006), persistence of threat (Eisner et al., 1963a; Eisner et al.,
1971; Fink, 1984; Machado et al., 2000; Maschwitz et al., 1981; Whitman et al., 1991),
and presence of conspecifics (Tobach et al., 1965). Internal factors assessed related to
emission deployment may include the amount of emission remaining in storage (Nolen
and Johnson, 2001; Wullschleger and Nentwig, 2002) and hunger level (Hayes, 1993;
Read and Hughes, 1987). In many cases, the emitter may physically interact with the
intended emission-recipient prior to discharge, thus enabling assessment (e.g., struggling
prey receive more stings/bites; Malli et al., 1999; Quinlan et al., 1995; Steiner, 1986).
293
However, in other cases, such as with snakes that inject venom during a single brief bite,
assessment of the target and decisions regarding deployment may be made prior to
contact with the target/receiver (Hayes et al., 2002).
Although strategic deployment of an emission may comprise numerous
behavioral components (as numbered above), one of the most important aspects of
strategic deployment is varying the amount of emission discharged (including foregoing
discharge altogether) according to circumstance. In this context, the first level of strategic
deployment results simply from the dichotomy of whether an emission, in any amount, is
released or not. In other words, although deployment may be warranted in some
circumstances it may not be appropriate in others. In a predatory context, the ofteninterrelated variables of prey type, size, and struggle intensity are important determinants
of emission deployment (Bücherl, 1971a; Dejean and Bashingwa, 1985; Djieto-Lordon et
al., 2001; Hayes et al., 2002; Malli et al., 1998; McCormick and Polis, 1990; Rein, 1993;
Rodriguez-Robles, 1992; Rodriguez-Robles and Leal, 1993; Schenberg and PereiraLima, 1978; Steiner, 1986). In terms of defense, there is typically a threshold intensity or
duration of stimulus that provokes an emission (Dettner et al., 1985; Eisner, 1958, 1970;
Krall et al., 1999; Machado and Pomini, 2008; Nolen and Johnson, 2001), with subthreshold stimuli evoking alternate, presumably less energetically costly, behaviors
(Blum, 1981, 1985; Duffield et al., 1981; Eisner, 1960; Machado et al., 2000; Schmidt,
1990), including fleeing (Duffield et al., 1981; Gibbons and Dorcas, 2002; Heiss et al.,
2010; Krall et al., 1999; Machado et al., 2000), thanatosis (Duffield et al., 1981; Eisner et
al., 1963a; Moore and Williams, 1990), postural defenses (Blum, 1981), threat displays
(Dettner et al., 1985; Gibbons and Dorcas, 2002; Starr, 1990; Verts et al., 2001),
294
retaliatory pinching/biting (Eisner, 1960; Eisner et al., 1963a; Heatwole, 1967), and sham
strikes (Hayes, 2008) and dry bites/stings (Herzig, 2010; Morgenstern and King, 2013;
Nisani and Hayes, 2011) of venomous animals.
Once an animal commits to the release of an emission, the animal may control the
amount of emission deployed in a number of ways. One way is by varying the number of
discrete, sequential or simultaneous emissions (of either fixed or variable amount). An
example of controlling the amount of emission deployed by number of sequential
emissions can be found in many social wasps, which respond to continued struggle of
their prey by repeated stinging until the prey becomes motionless (Steiner, 1986). Control
of the amount of emission deployed by the number of simultaneous emissions is
exemplified by millipedes, which vary the number of glands releasing defensive secretion
depending on the intensity of the threat stimulus (Eisner and Meinwald, 1966; Woodring
and Blum, 1965). A second way an animal may control the amount of emission deployed
is by altering the characteristics of a single emitting act (e.g., varying the rate and/or
duration of discharge by altering the pressure on the emission reservoir), resulting in a
quantitative variation in the amount of emission from one discharge to the next. For
example, rattlesnakes vary the amount of venom injected into different sized prey,
injecting, with a single bite, more venom into large than small mice (Hayes et al., 1995),
likely due to differential contraction of venom gland musculature (Hayes, 2008).
Although varying the amount of emission discharged is only one aspect of strategic
deployment, this component, in the form of the venom-metering (Hayes, 2008; Hayes et
al., 2002) and venom-optimization hypotheses (Wigger et al., 2002), has received
considerable attention among venom researchers.
295
Strategic Deployment of Venom
Venom can be defined as a toxic substance (comprised of one or more toxins)
causing dose-dependent physiological injury that is passively or actively transferred from
one organism to the internal milieu of another organism via a delivery mechanism and
mechanical injury (Nelsen et al., 2013). Venoms are complex secretions (Biass et al.,
2009; Escoubas et al., 2006) that are composed of many active constituents, including
variable combinations of proteins, salts, and other organic molecules such as polyamines,
amino acids, and neurotransmitters (Almaaytah and Albalas, 2014; Casewell et al., 2013;
Fry et al., 2009; Kuhn-Nentwig et al., 2011; Smith et al., 2013). Many different animals
possess venom (Fry et al., 2009; Mebs, 2002; Morgenstern and King, 2013) and may
employ it in predation, defense, and competitor deterrence (Fry et al., 2009) via injection
(Mebs, 2002; Wuester, 2010) or topical application (Dejean, 1988; Holldobler and
Wilson, 1990; Jeanne and Keeping, 1995; Westhoff et al., 2010) as a toxungen (Nelsen et
al., 2013). Since survival depends on an organism’s ability to feed and defend itself,
venom may be a commodity of great importance to venomous animals (Fry et al., 2013;
Haight, 2002). As a commodity, venom functions in a venom economy in which
availability and use can be analyzed in terms of the benefits and costs of its production
and deployment (Haight, 2002). In terms of benefits, the utilization of venom is
associated with minimization of energetic expenditure on hunting, minimization of risk of
injury from dangerous prey and aggressors, and maximization of success in prey
acquisition (Pintor et al., 2010). Furthermore, the efficacy of venom use may be related to
the amount of venom deployed (Hayes et al., 1995; Whiffler et al., 1988). On the flip side
of the coin, venom’s value can be viewed in terms of the metabolic costs of replacing it
(Haight and Tschinkel, 2003; McCue, 2006; Nisani et al., 2007; Pintor et al., 2010; Pintor
296
et al., 2011) and the ecological costs (vulnerability to predation, inability to envenomate
additional prey) of a depleted venom supply (Haight and Tschinkel, 2003; Hayes, 2008;
Malli et al., 1998; Nisani et al., 2007). Due to the valuable nature of venom, it may be
advantageous for venomous animals to be judicious when deploying their venom. For
example, the venom-metering hypothesis in snakes (Hayes, 2008) and the venomoptimization hypothesis in spiders (Wigger et al., 2002) have suggested that venomous
animals should use their venom as economically as possible. Morgenstern and King
(2013) have argued that the consistency of experimental evidence across several taxa
strongly suggests a convergence in economy of venom use, and have called for a unified
hypothesis of venom optimization. A review of the venom literature reveals the judicious
or economic use of venom may involve using venom only under certain conditions,
varying the amount used depending on circumstances, delivering the venom to specific
targeted areas, and aiming sprayed discharges at predators. Research has shown that
venom is strategically deployed to minimize costs and maximize benefits by a diverse
range of animals including scorpions, spiders, wasps, ants, and snakes.
297
Scorpions
Cost of Venom in Scorpions
Scorpions (order Scorpiones) use venom to subdue prey and to ward off predators
(McCormick and Polis, 1990; Rein, 1993), although there are interspecific differences in
how the sting is employed for predation (Cloudsley-Thompson, 1955; McCormick and
Polis, 1990; McDaniel, 1968; Quinlan et al., 1995; Williams, 1987) and defense
(Heatwole, 1967). Evidence suggests the ecological costs of depleted venom glands could
be long-lasting, as venom regeneration may take several weeks (Baerg, 1961; Fox, 2010).
In addition to the ecological costs of venom depletion, it has been suggested that
production and storage of the protein-rich venom is an expensive metabolic investment,
especially for scorpion species adapted to survive in extreme ecosystems on scarce
resources (Inceoglu et al., 2003). Data collected by Nisani et al. (2007) demonstrated that
venom was in fact expensive to regenerate and thus a valuable commodity; in Parabuthus
transvaalicus (Buthidae) the first 72 hours of venom regeneration following venom
extraction required a 39% increase in metabolic rate compared to the unmilked control
condition. Furthermore, although venom volume returned to original levels by 72 hours,
venom protein content remained at 25% of initial milking levels, indicating that venom
regeneration was not yet complete. In fact, although venom lethality (as measured by
cricket bioassay) returned within four days, it took eight days for major venom peptides
to return to original levels (Nisani, 2008; Nisani et al., 2012). Scorpions may be
disadvantaged by a depleted venom supply, and the need for biochemically efficient
venom could explain the lack of surface activity reported in post-ingestive scorpions
(Nisani et al., 2007). The expense associated with venom regeneration has been posited
as key to understanding selective sting use in scorpions (Rein, 1993).
298
Selective Sting/Venom Use by Scorpions
In terms of predation, selective sting use in scorpions is widely recognized (e.g.,
Cloudsley-Thompson, 1955; Cushing and Matherne, 1980; Nisani et al., 2007; Rein,
1993; Williams, 1987), and numerous authors suggest that selective sting use is
advantageous for the conservation of metabolically expensive venom (Casper, 1985;
Cushing and Matherne, 1980; Nisani et al., 2007; Rein, 1993). In scorpions, venom use is
dependent on the often-related prey characteristics of size and struggle (Baerg, 1961;
Kaltsas et al., 2008; McCormick and Polis, 1990; Pocock, 1893; Rosin and Shulov, 1963;
Stahnke, 1966). For example, Bucherl (1971a) noted that scorpions use the sting only
when the prey is large and vigorously defends itself, whereas small prey are captured
directly with the pedipalps without use of venom. More specifically, whether a scorpion
employs the sting depends on the ratio of prey size to scorpion pedipalp size (McCormick
and Polis, 1990). In combination with prey size and struggle, the hardness of the prey’s
integument may influence stinging, with hard-bodied prey such as grasshoppers stung
more often than soft-bodied prey such as termites (Cushing and Matherne, 1980).
Selectivity of sting use may also depend on scorpion ontogeny (Casper, 1985; Cushing
and Matherne, 1980) and species (McDaniel, 1968). Several examples of selective sting
use in scorpions are described below.
In one of the more detailed studies of selective sting use in scorpions, Rein (1993)
examined sting use in relation to prey size and activity in the scorpions Parabuthus
liosoma and P. pallidus. In this study, sting use was compared after presentation of three
different types of prey that differed in size and morphology, including small (10–18 mm)
and large (24–32 mm) larvae of Tenebrio molitor and a centipede, Lithobius forficatus
(26–35 mm). The author found that P. liosoma used the sting significantly less against
299
small larvae (sting attempted in < 5% of 18 trials) than against the large larvae (~83% of
17 trials) and centipedes (~87% of 23 trials). Similar results were found for P. pallidus,
which likewise attempted to sting small larvae (~20% of 29 trials) significantly less than
large larvae (100% of 23 trials) and centipedes (100% of 28 trials). Prey were usually not
stung immediately after being seized by the pedipalps, but rather the sting was brought to
bear only after the prey struggled and resisted capture. In fact, immediate use of the sting
occurred in only 14.7% and 26.3% of trials in which the sting was used by P. liosoma and
P. pallidus, respectively. Although quickly accepted as prey, large freshly killed T.
molitor larvae (i.e., non-resistant prey) were never stung by either species of scorpion.
In another study of selective sting use, Edmunds and Sibly (2010) investigated
sting use in relation to relative prey size and activity using the scorpion Hadrurus spadix
(Caraboctonidae) and crickets (Acheta domesticus) as prey. Prey were divided into six
relative size classes based on the ratio of prey length to scorpion pedipalp patella length,
ranging from approx. 0.5 to 1.5. Data showed that sting use by H. spadix, as measured by
the percentage of cases in which the sting was employed, increased with increasing
relative prey size. For example, when the smallest prey were offered, stings occurred in
only 29% (7/24) of cases, but this rose to 100% (48/48) of cases when the prey items
were larger than the scorpions’ patella length. To test the effect of prey activity on sting
use, the authors offered prey that had been cooled for short periods (5, 10, or 15 min).
Data indicated that sting use increased during encounters with more active prey.
In a study by Jiao and Zhu (2009) examining the prey capture behavior of the
scorpion Heterometrus petersii (Scorpionidae) feeding on T. molitor larvae (28–32 mm,
ca. 0.1 g) and Zophobas morio larvae (48–52 mm, ca. 1.0 g), it was found that scorpions
300
never used the sting to subdue the smaller T. molitor prey (0/20 cases), but occasionally
(~7% [2/30] of cases) stung the larger Z. morio prey. Additionally, the authors noted that
the scorpion only stung actively struggling Z. morio and never stung passive prey. Taken
together, these results for predatory sting use by scorpions indicate that venom
deployment depends upon the size and resistance of the prey, as presumably determined
by the scorpions during interactions with the prey.
Selective sting use by scorpions for defense has also been documented. Heatwole
(1967) investigated the defensive behavior of the scorpion Opisthacanthus lepturus
(Hemiscorpiidae) and discovered this scorpion relied chiefly on the pedipalps for
defensive weapons, but with repeated stimulation it occasionally used the sting in
conjunction with them. Specifically, in only about 7% (3/42) of cases did O. lepturus
employ the sting, whereas in the remaining cases responding with pedipalps only.
Sometimes even when the stinger is used, the scorpion might not deploy venom; Nisani
and Hayes (2011) investigated venom expenditure over a series of stings in the scorpion
P. transvaalicus and reported that, on average, 12.5% (15/120) of defensive stings were
dry stings. These dry stings often occurred early in a stinging sequence, supporting the
hypothesis that lack of venom emission was due to scorpions exercising selective venom
deployment rather than simply experiencing a lack of available venom.
In a study of the defensive venom-squirting behavior of P. transvaalicus, Nisani
(2008) found that among juvenile scorpions (n = 8), the proportion that squirted under a
high-threat condition (87.5%; metasoma grasped with forceps combined with forcefullyblowing air) was significantly greater than under a low-threat condition (0%; metasoma
grasped with forceps). Similarly, among adults (n = 8) there was a higher incidence of
301
venom squirting under high-threat (87.5%) compared to low-threat (25%) conditions,
although the difference only approached significance. The author concluded that the
scorpions were judicious in whether or not they squirted venom, doing so only as a
measure of last resort when the additive stimuli (contact and airborne) were likely to
confirm the proximity and relative threat of a predator. Thus, results for defensive sting
use by scorpions indicate that venom deployment is influenced by both persistence and
degree of a threatening stimulus.
Amount of Venom Deployed by Scorpions
Although the question has been raised whether scorpions can control the amount
of venom they inject into their prey as a means of venom conservation (Edmunds and
Sibly, 2010), little quantitative data has been published on predatory venom metering in
scorpions. So far, control of venom deployment for predation has only been noted at the
level of number of stings. As with other venomous organisms such as wasps and ants,
scorpions base repeated stinging behavior on prey struggle. For example, when Hadrurus
arizonensis (Caraboctonidae) was offered cockroaches (Periplaneta americana) and
crickets (A. domesticus), all prey were stung once, and when the scorpion stung
repeatedly it was usually in response to the struggling of the prey (Bub and Bowerman,
1979). Similarly, when offered P. americana, the scorpions Urodacus novaehollandiae
and U. armatus (Scorpionidae) typically stung prey once (when the sting was used) but
re-stung prey that continued to struggle (Quinlan et al., 1995). Clearly, then, scorpions
meter venom in terms of number of stings delivered to prey based on persistence of prey
struggle.
302
In terms of defense, Nisani (2008) and Nisani and Hayes (2011) demonstrated
venom metering in the scorpion P. transvaalicus, showing this scorpion perceives risk
and regulates venom expenditure during stinging according to level of threat. In this
study each scorpion (n = 6) was tested in each of two conditions, high threat (five
provocations separated by five second intervals) and low threat (five provocations
separated by five minute intervals). The provocation consisted of touching the scorpion’s
dorsum with the edge of a parafilm-covered plastic cup (also used to collect the expended
venom). The authors found that P. transvaalicus regulated venom expenditure at three
levels: (1) dry versus wet stings, (2) volume of venom in wet stings, and (3) venom
composition. Of the 120 stings analyzed, 12.5% were dry stings, and the authors
contended these scorpions chose between delivering a wet or dry sting. In terms of
regulating venom volume, when all five successive stings were considered, data showed
that scorpions expended 2.2-fold more venom per sting in the high-threat condition
compared to the low-threat condition. When the authors compared only the first three wet
stings (to distinguish between decisions involving venom release [dry vs. wet stings] and
quantity of venom released), the effect of threat level on venom expenditure was no
longer significant; however, the substantial effect size suggested that scorpions still
injected more venom (1.9-fold) per sting during high-threat compared to low-threat
conditions. Concerning venom composition, the authors found that progression through
the sequence of clear, potassium-rich prevenom, to opalescent, to milky, protein-rich
venom depended on the quantity of venom expended, with milky secretion appearing
only after the limited quantity of prevenom was exhausted. Additionally, the sequence of
venom categories expulsed varied with threat level; in the high-threat condition scorpions
303
more quickly escalated their delivery of milky venom, doing so earlier within the
sequence of stings compared to the low-threat condition. The authors concluded that
these scorpions make decisions regarding usage (dry vs. wet sting), quantity, and,
indirectly, composition (prevenom or venom) of venom injected, providing support for
the venom-metering hypothesis.
The defensive venom-squirting behavior of P. transvaalicus also suggests
scorpions control the amount of venom expended. Although Nisani (2008) did not
measure the quantities of venom squirted in this experiment, he did document variation in
the duration and velocity of venom flow of successive squirts, suggesting that scorpions
can regulate venom gland contraction. Regulation of venom gland contraction has been
argued as evidence in support of the venom-metering hypothesis in snakes (Hayes, 2008;
Hayes et al., 2008). Taken together, data indicate scorpions control the amount of venom
deployed by number of stings and volume per sting, with scorpions delivering more
stings to more vigorously struggling prey, and a greater volume of venom per sting when
threatening stimuli occur in close temporal proximity.
Delivery Location and Aiming of Venom by
Scorpions
Unlike for some venomous animals (e.g., some spiders, wasps, and ants), for
scorpions, the location where the sting is delivered to the prey’s body plays a minor role,
if any, in strategic deployment of venom. When the sting is used it is usually drawn
across the prey’s body in a probing motion until a soft area is found for insertion (Bub
and Bowerman, 1979; Casper, 1985; Quinlan et al., 1995; Rein, 2003). Unlike the
positive correlation between sting sites and major ganglia observed when digger wasps
304
subdue cricket prey (Steiner, 1976), Bub and Bowerman (1979) found no such correlation
in the sting sites of H. arizonensis on either cricket or cockroach prey. Rather, the sting
site distribution reflected the sites of the first penetrable tissue encountered. For example,
cockroaches received 94% of stings on their ventral surface even though the dorsal
surface (protected by heavy wings and thoracic sclerites) was usually encountered first by
the telson. In contrast, cricket prey received 75% of their stings dorsally; the dorsal aspect
was usually encountered first by the telson and since the relatively small cricket wings
did not extend over the abdomen, immediate penetration could occur.
As in the context of predation, there is no evidence that venom used for defense is
specifically targeted by scorpions. Nisani (2008), studying the venom squirting behavior
of P. transvaalicus, found initial direction of the squirt varied considerably and
concluded that sprayed venom was not aimed at the threat stimulus. However, the author
contended that the rapid, simultaneous, and independent movement by the metasoma
and/or telson during squirting increases the dispersion of sprayed venom and thereby
increases the likelihood of venom contacting a predator’s eyes.
305
Spiders
Cost of Venom in Spiders
Spider (order Araneae) venoms are complex, multi-component mixtures of
biologically active substances that play important roles in both attack (killing or
paralyzing prey) and defense (Atkinson and Wright, 1992; Bettini and Brignoli, 1978;
Foelix, 1996; Kuhn-Nentwig et al., 2011; Quintero-Hernandez et al., 2011; Vassilevski et
al., 2009). Despite venom’s value to spiders, there exists no quantitative data on the
metabolic cost of its regeneration in this taxon, although Malli et al. (1999) suggested
venom production is energetically expensive. In addition, data on the often-lengthy
timing of venom regeneration suggests spiders may incur an ecological cost for venom
expenditure. Since venom production may take weeks (Boeve et al., 1995; Perret, 1977b)
to months (Freyvogel et al., 1968), and spiders may capture several prey items per day, it
is expected that spiders would strictly control venom release to avoid the metabolic
expense of regeneration and the depleted reserves which could leave the spider
vulnerable to predation or unable to deal with subsequent prey (Boeve and Meier, 1994;
Clements and Li, 2005; Malli et al., 1998). In addition, secondary losses of venom use in
spiders, such as in uloborid spiders which kill their prey not with venom but by wrapping
them tightly in hackled silk, further indicate that venom use comes with a biochemical
price (Morgenstern and King, 2013).
Selective Venom Use by Spiders
Strategic deployment of venom by spiders includes using this valuable
commodity selectively. In the context of predation, the factors which influence whether
venom is deployed by spiders are those which contribute to the difficulty this animal
306
encounters when subduing prey, and are likely to include an interaction of prey size, prey
defensive capabilities, and prey struggle. In some cases spiders may opt not to deploy
venom by foregoing biting prey altogether. For example, Robinson (1969) noted that
when the spider Argiope argentata (Araneidae) attacked prey less than 1/100th of the
spider’s weight, the prey were not bitten but simply seized in the jaws and carried back to
the web’s hub. However, when the prey were greater than 1/100th of the spider’s weight,
the prey were bitten and it was assumed that venom was used in these bites (Malli et al.,
1998; Robinson, 1969). Alternately, selective venom use may be the result of decoupling
biting from envenomation. The ability to use or withhold venom independently from the
biting act stems from spider anatomy; the venom glands are surrounded by striated
muscle under nervous control, allowing deployment of venom at the volition of the spider
(Boeve et al., 1995; Bücherl, 1971b; Malli et al., 2000; Schenberg and Pereira-Lima,
1978). Prey size may be an important factor influencing venom deployment; spiders
routinely seize and chew small arthropods without applying venom, relying instead on
the chelicerae to crush or chew them, reserving venom for larger prey (Eisner et al., 2005;
Freyvogel et al., 1968; Kuhn-Nentwig et al., 2011). For example, when Malli et al.
(1998) quantified the venom dose injected by Cupiennius salei (Ctenidae) into crickets of
various size classes ranging from 100–660 mg, they found the spiders did not inject
venom into 22% (7/32) of the crickets in the smallest size class (100–110 mg). The
authors contended that C. salei does not rely exclusively on its venom when feeding on
small prey, but can accomplish the job through mechanical damage alone inflicted by the
chelicerae. Prey size is also important in determining whether venom is used by the
spider Phoneutria nigriventer (Ctenidae), which only injects venom into excessively
307
large prey, relying on mechanical damage caused by the chelicerae to kill small insects
(Schenberg and Pereira-Lima, 1978). According to Wigger et al. (2002), the difficulty a
spider encounters in overwhelming prey, which can vary with prey species, may also
determine whether spiders use their venom. Wigger et al. (2002) noted selective venom
use by C. salei while using ELISA (enzyme-linked immunosorbent assay) to quantify the
amount of venom the spider injected into four different prey (blowflies, crickets, stick
insects, and ground beetles). The authors found that no venom could be detected in 32%
(6/19) of crickets attacked, whereas all prey items of the other prey types that were
attacked had been envenomated. The authors suggested that sometimes the spider relies
on its strong chelicerae to kill soft prey susceptible to mechanical damage. However, it
remains less clear why, if C. salei often withheld venom from crickets, the spider did not
also occasionally withhold venom from stick insects, which the authors argued were also
a soft, unproblematic prey type. Other factors that may influence whether spiders deploy
venom include prey struggle (Bücherl, 1971b) and the prey’s defensive capabilities
(Kuhn-Nentwig et al., 2011).
Spiders may selectively deploy venom in defense as well as for predation. For
example, P. nigriventer employs venom only when the spider finds no way to escape
attack (Schenberg and Pereira-Lima, 1978). In female mouse spiders, Missulena bradleyi
(Actinopodidae), aggravation by experimenters led to voluntary expression of venom
from only 15% of spiders, suggesting most spiders’ threshold for venom expenditure was
not reached under these conditions (Herzig et al., 2008). Defensive dry bites are another
example of selective use of venom. Herzig (2010) argued some dry bites by the mouse
spider (Missulena spp.) investigated by Isbister (2004) might be explained by the
308
voluntary decision of the spider not to deploy venom during a bite in order to save the
metabolic expense of venom synthesis. In addition, while analyzing methods of venom
extraction from the African tarantula Scodra griseipes (Theraphosidae), Celerier et al.
(1993) observed that the spiders could bite a lure many times without emitting venom.
Similarly, Freyvogel et al. (1968) noted that the baboon spider Pterinochilus
(Theraphosidae) often actively withheld venom during milking attempts. Taken together,
these data indicate that whether spiders deploy venom or not may depend on several
factors, including prey size, prey type, and threat level.
Amount of Venom Deployed by Spiders
When it comes to the study of how venom deployment varies with circumstance,
perhaps the best-studied group, among invertebrates, is the spiders. Spiders have been
compared to snakes in their ability to control the amount of venom delivered (Schenberg
et al., 1970), and evidence indicates that degree of venom gland emptying is at the
volition of the spider (Boeve et al., 1995; Maretic, 1987). Using indirect measures of the
amount of venom deployed, investigators have found that prey size and prey struggle
intensity are important factors influencing predatory venom expenditure in spiders. For
example, Perret (1977a), comparing volume of venom electrically milked before and
after spiders were fed, found that the tarantulas Aphonopelma chalcodes and Dugesiella
hentzi released more venom in the first bite when feeding on adult (1–2 g) cockroaches
(P. americana) than when feeding on adult (0.1 g) mealworm beetles (T. molitor). When
attacking cockroaches, A. chalcodes injected, on average, 1.7 µL of venom (25% of
available venom), but injected no venom into mealworm beetles. Similarly, D. hentzi
309
injected 1.7 µL of venom (28% of available venom) into cockroaches, but only 0.1 µL of
venom (2% of available venom) into mealworm beetles. As the venom was delivered in a
single bite, the mechanism by which venom deployment was controlled is likely related
to extent of venom gland compression. In another study, based on mass gain of bitten
prey, Pollard (1990) found that the New Zealand crab spider Diaea sp. (Thomisidae)
injected more venom, on average, into a larger fly, Pegohylemyia sp., than into a prey
item of half the mass, the fruit fly Drosophila immigrans (0.108 vs. 0.067 mg venom,
respectively). In only 7% (3/43) of cases was a prey item bitten more than once,
indicating that number of bites was not the primary mechanism for controlling amount of
venom deployed. Although the author noted that Diaea can regulate the amount of venom
injected based on prey size, he also suggested that the spider may use tactile information
from captured, struggling prey to help assess prey size. In a study based on the mortality
rates of multiple series of crickets (A. domesticus) of different mass classes attacked by
the spider C. salei, Boeve (1994) showed that for prey of small or medium mass (less
than 40 mg) the spider injected an amount of venom proportional to the mass of the prey,
but that for large prey items the spider injected all its venom. Furthermore, by
interrupting bites on the different prey mass classes at various intervals and analyzing the
state of the bitten prey, the author showed that the spiders varied the rate of venom
delivery in a single bite based on prey mass. Although acknowledging the relationship
between prey size and venom dose injected, the author also speculated about the
influence of prey struggle, suggesting the possibility that each escape attempt made by
the prey may have stimulated the spider to inject a discrete amount of venom. In another
study using C. salei, and based once again on the conditions of series of singly-bitten
310
prey, Boeve et al. (1995), demonstrated that larger crickets received larger venom doses
than smaller crickets. Furthermore, the authors showed that C. salei injected larger
venom doses into “difficult-to-handle” prey (cricket Grilloides sigillatus) than into
“easy” prey (cricket Gryllus bimaculatus), with the dichotomy essentially reflecting a
difference (not quantified) in struggle intensity after attack. The authors concluded that C.
salei could empty its glands partially as well as completely, resulting in dosed injections
of venom. Thus, there are several examples using indirect measurements of venom
expenditure showing the amount of venom deployed by spiders varies with prey size and
prey struggle intensity.
The first to directly quantify the amount of venom spiders delivered to various
size classes of prey was Malli et al. (1998). These authors performed ELISA on whole
cricket (A. domesticus) homogenate using monoclonal antibodies to the main toxin in C.
salei venom, CSTX-1. The authors discovered that when mature C. salei females
attacked crickets (n = 128 attacks) of four size classes (100–110, 290–320, 420–460, and
600–660 mg) there was a significant relationship between the size of prey and the
quantity of venom expended (r = 0.80), with mean venom quantities ranging from 0.15
µL for the smallest prey to 1.53 µL for the largest. Multiple comparisons indicated that C.
salei released significantly more venom with increasing size of cricket (p < 0.01 for all
comparisons). Although a clear relationship between venom dose and prey size was
found, the authors acknowledged that it was not clear if the spiders injected more venom
into larger prey simply because of their size, or if this was a consequence of greater
struggle by larger prey. The authors also suggested the pattern of venom deployment
observed could result from the combination of C. salei injecting venom gradually until
311
the prey is motionless and a size-based difference in venom susceptibility. The authors
did not emphasize the number of bites delivered to prey, although multiple bites did
occur in at least 5% of attacks. Thus, a metering mechanism based on a graduallydelivered dose of venom from a single bite was more common than one based on
multiple bites.
In a follow-up study, Malli et al. (1999) further investigated the influence of prey
size on venom expenditure by mature female C. salei, once again using ELISA to
quantify the amount of venom released. To disentangle the effects of prey size and prey
struggle intensity on venom dosage, the authors used anesthetized crickets (A.
domesticus) in four size classes (100–110, 290–320, 420–460, and 600–660 mg) that
were artificially caused to struggle at the same intensity and for the same duration (5
min). Quantity of venom released varied widely within a size class, and prey size and
quantity of venom expended were weakly correlated (r = 0.23, p < 0.05). Multiple
comparisons showed that C. salei injected significantly less venom into the smallest size
class, whereas the null hypothesis (equal amounts of venom) could not be rejected for
comparisons between the other three size classes. The authors concluded that prey size
alone is not likely to be an important cue for effectively regulating venom injection.
Further, they argued that the results of Malli et al. (1998), in which larger prey received
larger venom doses, were a consequence of predator-prey interactions during
envenomation which, though increasing with the size of a given prey species, did not
depend on the size of the prey itself.
In addition to prey size, Malli et al. (1999) also investigated the effects of prey
struggle intensity and prey struggle duration on the amount of venom C. salei injected
312
into cricket prey. Using ELISA to quantify the venom injected into anesthetized crickets
of the same mass (290–320 mg) artificially caused to struggle at four different intensities
(no movement [control], low, medium, and high), the authors found a highly significant
relationship between intensity of prey movement and quantity of venom expended.
Multiple comparisons indicated that, with the exception of the difference between control
and low-intensity prey movement, C. salei released significantly more venom as the
intensity of prey movement increased. The authors suggested that injection of larger
quantities of venom into vigorously resisting prey would serve for rapid immobilization,
thus preventing the spider from injuries or from losing its prey. Additionally, the authors
noted that C. salei saved up to 50% of its venom by discriminating between highintensity and low-intensity prey movements.
Malli et al.’s (1999) study of the influence of prey struggle duration on venom
expenditure by C. salei yielded similar results. In this experiment the crickets (290–320
mg) were vibrated at the same intensity (medium) but for different lengths of time (0
[control], 1, 2.5, or 5 minutes) following the initial bite. Data indicated that duration of
prey movement and quantity of venom expended were positively correlated (r = 0.61, p <
0.01). Multiple comparisons showed that, with the exception of the difference between
the control and 1-minute treatments, C. salei released significantly more venom with
increasing duration of prey movement. Malli et al. (1999), based on this series of
experiments, concluded that C. salei injects venom gradually in response to stimuli
generated during the course of envenomation. The authors speculated that perhaps tactile
hairs and slit sense organs found on the chelicerae and base of the claws serve a
vibrosensitive function in controlling the release of venom during envenomation. For all
313
experiments, as prey were held in the chelicerae (i.e., single bite) for the duration of each
trial, the mechanism for varying venom expenditure was independent of number of bites.
Following in the footsteps of Malli et al. (1999), Wigger et al. (2002)
demonstrated differential venom expulsion by C. salei based on difficulty in
overwhelming prey, which varied with prey species. The authors used ELISA to quantify
the amount of venom injected by adult female C. salei into four prey species: blowflies
(Protophormia sp.), larval crickets (A. domesticus), stick insects (Carausius morosus),
and ground beetles (Poecilus cupreus). All prey were of a uniform (but unreported) size
class. Results indicated that ground beetles received significantly more venom than either
of the other three prey species. The authors argued that the blowflies, crickets, and stick
insects, all lacking thick chitinization, were relatively soft and thus unproblematic prey
types, resulting in a relatively low dose of venom. In contrast, the heavily sclerotized
ground beetles represented difficult to overwhelm prey because the spiders were forced
by the beetles’ mechanical protection to inject their neurotoxic venom into the prey’s
abdomen, an injection site requiring more venom to subdue the prey than a bite to the
head or thorax normally would. In fact, the authors suggested that the lengthy handling
time for the ground beetles may have been the stimulus leading to greater venom
expenditure. Although the number of bites C. salei delivered to prey was not explicitly
stated, it appears that the spider held prey in its chelicerae (i.e., single bite) for the full 5
minutes of each trial, indicating that the mechanism controlling venom expenditure was
independent of number of bites.
Risk of prey escape, which may vary with prey species, may also influence
spiders’ venom deployment. Robinson’s (1969) findings suggest this possibility, and
314
Boeve et al. (1995) interpreted them in this way, citing them as evidence that more
venom is injected into easily escaping insects. Robinson (1969), while studying the
predatory behavior of A. argentata, noted that lepidopteran prey, which were bitten prior
to silk wrapping, received a statistically longer bite than other prey types which were first
wrapped in silk and then bitten. It was suggested that the short bite might involve a
smaller dose of venom (Robinson, 1969; Robinson et al., 1969; Robinson and Olazarri,
1971) since it would be wasteful to use biologically expensive secretions unnecessarily
on a wrapped prey (Robinson, 1969). However, it is possible that the long bite may be
long simply because the spider must wait for the venom to take effect before it can safely
release the prey and commence wrapping (Robinson et al., 1969). The adaptive
significance of the long bite lies in its ability to cause the most rapid restraint of prey with
high escape potential, such as lepidopterans (Robinson, 1969; Robinson and Olazarri,
1971). Although the duration of the long bite delivered to lepidopterans varied
dramatically (e.g., from 1 sec to 527 sec when attacking live moths; Robinson and
Olazarri, 1971), there was no systematic relationship between length of bite and weight
of prey for either the long or short bite (Robinson, 1969).
Wullschleger and Nentwig (2002) studied the predatory behavior of the spider C.
salei and, though they didn’t measure venom expenditure, these authors demonstrated
C. salei knows how much venom is available in its venom glands, chooses the
appropriate type of prey according to this information, and distinguishes between prey
types with different venom sensitivities. Hostettler and Nentwig (2006) showed that C.
salei uses olfactory information to identify prey type and distinguish venom sensitivity,
presumably in order to conserve venom.
315
Although much of the work examining venom deployment in spiders has been
performed using C. salei in circumstances involving, for the most part, a single sustained
bite to prey, it should be noted that spiders may bite prey multiple times (Gilbert and
Rayor, 1985; Malli et al., 1998; Parks et al., 2006; Schenberg and Pereira-Lima, 1978;
Willey et al., 1992), and varying the number of bites may be one way in which spiders
could control the amount of venom deployed. In such cases, continued prey struggle may
be the stimulus for additional bites (Gilbert and Rayor, 1985).
In the context of defense, the green lynx spider Peucetia viridans (Oxyopidae)
studied by Fink (1984) presents an example of a spider that varies the amount of venom
expended based on extent of provocation, and may vary the amount of venom in
individual spits. The female spiders were observed to eject venom straight forward from
their chelicerae up to a distance of 20 cm when approached or when their legs were
pulled. Although a single spit was most common, the spiders would spit several times in
succession if repeatedly provoked. The quantity of venom in a spit was variable, from
trace amounts up to more than 5 µL. Another example of defensive variable venom
expenditure in a spider comes from Perret’s (1977a) investigation into the amount of
venom released in a single bite by tarantulas. In defensive bites against mice (30 g; n = 2
cases), A. chalcodes injected, on average, 2.3 µL of venom (36% of available), about the
same amount of venom (1.7 µL, 25% of available) as spiders injected when killing
cockroaches, but more venom than injected into mealworm beetles. The author suggested
that since the cockroaches and mice were of considerably different sizes, and since the
spiders displayed typical defensive behavior toward the mice, it was possible that the
spider calculated venom injection differently in defensive rather than predatory
316
situations.
In consideration of this large amount of evidence, there can be little doubt that
some spiders have the capacity to vary the amount of venom expended dependent on the
context of the predatory situation. Perhaps unsurprisingly, response of the bitten prey is
an important cue influencing how much venom spiders inject. In comparison to the
evidence for predatory venom metering, evidence for venom metering by spiders in a
defensive context is weaker, with persistence of threat and type of threat potentially
important cues.
Delivery Location of Venom by Spiders
Delivery location of venom may be an important part of strategic venom
deployment in some spiders. However, without data on prey morphometrics it can be
difficult to distinguish spider targeting preference from an unequal, but still random, bite
distribution stemming from unequal surface areas of body parts available for biting
(Morse, 1999). Furthermore, for spiders, evidence suggests that the initially attempted
bite location may often play a smaller role in prey incapacitation than the final location of
envenomation (Malli et al., 1998; Pollard, 1990). One might suspect that a strategic site
of venom injection used by spiders, given the potent neurotoxins present in their venoms
(King and Hardy, 2013), would be as near to the prey’s central nervous system as
possible, typically the thorax or head. In general, thorax envenomations are common
(Foelix, 1996), and several investigators contend that targeting the thorax or head gives
the fastest effects on prey (Malli et al., 1998; Malli et al., 1999; Pollard, 1990; Wigger et
al., 2002). In fact, as Morse (1999) pointed out, such a “neckbite” pattern of
envenomation is often reported by general spider sources. Even so, data on bite location
317
on prey are relatively scarce.
Much of the limited data on the location of predatory spider bites comes from
studies of the spider C. salei. Malli et al. (1999) reported that when C. salei was offered
CO2-anesthetized crickets (A. domesticus) the majority of the prey were bitten in the
thoracic region (87.4% [420/480] of bites), whereas only 12% and 0.6% of bites were
delivered to the abdomen and head, respectively. In a separate study investigating C. salei
preying on four size-classes of un-anesthetized crickets (A. domesticus), Malli et al.
(1998) reported the frequency of bites to a given site varied among prey size classes, but
that most crickets were bitten either in the thorax (66–85%) or in the pronotum right
behind the head (9–28%). In contrast, few bites were aimed at the soft abdomen (3–9%)
or the hard, chitinized head capsule (0–3%). Furthermore, except in one case, all crickets
first bitten in the abdomen were subsequently bitten in the thorax. Given that a cricket’s
thorax is smaller than its abdomen, the reported distribution of bites indicates a
preference for thorax envenomation. Malli and colleagues (Malli et al., 1998; Malli et al.,
1999) argued that bites to the thorax may decrease the amount of venom needed for
paralyzation, and also reduce time to immobilization and therefore reduce the spider’s
risk of injury. In contrast to the high frequency of thorax bites when preying on crickets,
when C. salei preyed on heavily sclerotized ground beetles (P. cupreus), the spiders, after
trying to inject venom into the thorax or head, most often ended up placing bites in the
abdomen (Wigger et al., 2002). This study demonstrates that strategic deployment of
venom by delivery location can be constrained by prey characteristics.
Other spiders besides C. salei target their venom. Pollard (1990) noted that more
than half a dozen species of crab spiders are known to envenomate prey principally in the
318
head and thorax, and that several of these species have been observed, after capturing
prey by their posterior abdomen, to re-envenomate these prey in the head or thorax. The
author hypothesized that the crab spiders are re-envenomating prey in the thorax to
achieve faster immobilization. The orb-web spider Argiope aurantia may also
strategically envenomate certain prey by placement of the bite (Harwood, 1974). After
wrapping orthopteran prey that have struck its web, this spider delivers a series of short
bites and a single sustained bite. Data for 51 bites indicated that the majority (~80%) of
the sustained bites were on the anterior half of the prey. For the orb-web spider A.
argentata, non-lepidopteran prey were bitten after wrapping, and in these cases the bite
was often directed at the head or thorax (Robinson and Olazarri, 1971). However, in
some cases initial bite location is more a matter of happenstance; when prey were bitten
before wrapping, the case for lepidopteran prey, the bite of A. argentata was often
delivered to the first point of contact with the prey. Even so, if the initial bite happened to
be located on a wing or other appendage, the bite was transferred to a more “substantial”
part of the prey, possibly as a result of sensory information received directly by the
chelicerae. In contrast to the above examples in which spiders targeted their venom
towards the prey’s central nervous system, the bites of other spiders may be directed
toward peripheral targets such as legs or antennae (Parks et al., 2006; Suter and Stratton,
2012), or directed toward prey in proportion to the surface area of the prey’s body parts
(Morse, 1999). Taken together, the evidence indicates that some spiders demonstrate a
preference for envenomating some types of prey in the thorax or pronotum, presumably
to effect rapid prey immobilization, but that targeted venom delivery in spiders is not
strict or universal.
319
Hymenopterans
Hymenopteran venoms are complex mixtures of biochemically and
pharmacologically active components such as biogenic amines, peptides and proteins,
some of which are recognized as important allergens (de Souza et al., 2009; dos Santos et
al., 2010). The venom contained in the hymenopteran sting apparatus comes from the
venom reservoir, a modified accessory reproductive gland (Schmidt, 1982). This venom
is used to procure prey and for defense (Schmidt, 1982). In fact, the hymenopteran sting
and venom have provided a key enabling mechanism for these small organisms to defend
themselves against vertebrate predators (Schmidt, 1990).
Wasps
Cost of Venom in Wasps
Wasps are often grouped into the social wasps, which employ venom for defense
(Steiner, 1986), and solitary (predatory) wasps, whose venoms are nonlethal paralyzing
fluids intended for inactivating their prey while maintaining it alive as a food source for
the wasps’ larvae (Schmidt, 1990). However, despite the important roles of wasp venom,
there has been no research into the metabolic cost of venom production in wasps.
However, in regards to the stings of predatory wasps, Budriene and Budrys (2005)
suggested that each sting is associated with significant physiological costs, and that those
costs may vary depending on the amount of venom injected per sting. Furthermore, the
authors suggested the possibility that venom supply may limit the number of stings that
can be deployed for provisioning each offspring or for a given provisioning time unit. In
addition to the metabolic cost of venom regeneration, wasps stinging in defense against
320
vertebrates (usually the case when social wasps employ venom) face a cost in terms of
the risk of retaliation from an adversary much larger than themselves (Starr, 1990).
Selective Sting/Venom Use in Wasps
Wasps strategically deploy venom by using the resource selectively, reserving its
use for occasions when other methods of predation and defense are insufficient. Among
social wasps, biting is generally used to kill prey and the sting is primarily a defensive
weapon only rarely employed in predation (Akre, 1982; Evans and West-Eberhard, 1970;
Olson, 2000). However, these wasps will use the venomous sting for predation if
grappling with particularly large or vigorously struggling prey and the risk of
counterattack or escape becomes imminent (Edwards, 1980; Hermann, 1984; Spradbery,
1973). For example, several species of Pacific Northwest yellow jackets (Vespinae) will
attempt to sting if they attack an arthropod that physically overpowers or injures them
during the attack (Hermann, 1984). Similarly, the Giant Hornet of Japan, Vespa
mandarinia (Vespinae), will destroy large honeybee colonies by biting workers, but will
use their stings in fights to subdue nests of other Vespa or Vespula species (Matsuura and
Sakagami, 1973). Likewise, Spradbery (1973), citing Bordas (1917), noted the case of
Vespa crabro (Vespinae) using its sting to immobilize a large grasshopper before
dismembering it.
Selective sting use is also present in solitary wasps, which usually do inject
venom during prey capture (Andrietti, 2011; Steiner, 1986). For example, although
sphecids nearly always paralyze their prey with venom, wasps in the subfamily
Pemphredoninae effectively “paralyze” (thoroughly disable but do not kill) particularly
321
small and helpless aphids by merely squeezing them gently with their mandibles, but use
the sting to subdue leafhoppers and related prey (Steiner, 1986). In addition to selective
sting use, it may be possible for some wasps to employ the sting without delivering
venom; Steiner (1979) reported some predatory stings of the wasp Oxybelus uniglumis
(Crabroninae) appeared to be “dry” (i.e., sting was not followed by venom inoculation)
judged on the lack of paralysis of fly prey. In addition to selective venom use for
predation, wasps may occasionally bring the sting into use for non-predatory aggression.
For example, although Polistes (Polistinae) wasps seldom use their sting when hunting,
they do use it against conspecifics, such as in fights between foundresses in pleometrotic
spring associations (Hermann, 1984).
Because wasps should prefer any tactic that can repel a predator (especially a
large vertebrate predator) without the risk of direct contact, venom use is generally the
last line of defense for wasps (Schmidt, 1990). To avoid or delay the use of the sting,
wasps may rely on other defenses including crypsis, aposematism, a tough integument,
association with better defended species, protean escape flight, chemical barriers, graded
threat displays, biting, and kicking (Schmidt, 1990; Starr, 1990). Even when venom is
employed for defense it may be used selectively. For example, the social wasp
Parachartergus colobopterus (Polistinae) sprayed venom in defense against human
aggressors but failed to sting or spray intruding ants (Jeanne and Keeping, 1995). When
considered as a whole, the evidence clearly indicates that wasps are selective in
deployment of venom, bringing it to bear in response to larger, more vigorous prey, and
against specific aggressors.
322
Amount of Venom Deployed by Wasps
Wasps strategically deploy venom not only by selectively using their
sting/venom, but also by controlling the amount of venom deployed when emission is
warranted. For wasps, evidence for the control of amount of venom deployed has only
been measured at the level of number of individual stings delivered. In terms of
predation, many authors have reported that stinging by solitary wasps is repeated until the
prey ceases to move; thus, responses of the attacked prey quantitatively affect stinging
(Casiraghi et al., 2001; Evans, 1966; Rathmayer, 1978; Steiner, 1986). For example,
Steiner (1986), citing Williams (1929), noted that when Ampulex canaliculatus
(Ampulicidae) captured cockroaches, re-stinging frequently occurred in response to
growing or residual resistance of the prey. Similarly, Steiner (1986), citing Berland
(1928) and Malyshev (1968), described how Sclerodermus wasps (Bethylidae), when
attacking beetle larvae, first stung the mouth and then added innumerable additional
stings, including abdominal stings, until all prey movements ceased. In another example,
stinging can require up to eight hours for Dibrachoides dynastes (Pteromalidae) which
delivers 3–100 stings until its host (lepidopteran and coleopteran larvae) becomes
completely motionless (Clausen, 1940). Steiner (1986), citing Janvier (1930), noted that
the spider wasps Haploneurion minus and H. apogonum (Pompilidae) stung their spider
prey once between the legs, but that re-stinging (up to seven times) occurred whenever
resistance to transport of the incompletely paralyzed prey was felt. Some species of the
genus Exeristes (Ichneumonidae) may also sting host larvae multiple times, and there is
evidence that additional stings result in the injection of additional venom as repetition of
323
stinging frequently results in death of the host whereas single stinging causes only
paralysis (Clausen, 1940).
In addition to (or likely related to) prey struggle, there is also evidence that prey
size may influence strategic venom expenditure by wasps in terms of number of stings
delivered. Perhaps one of the most quantitative studies examining factors influencing
wasp stinging behavior was performed by Budriene and Budrys (2005) who examined
stinging behavior in relation to offspring provisioning in eight species of Eumeninae
wasps. In a field study including 4,642 prey specimens taken from 347 nests, the authors
investigated how the variables of prey mass and diameter of nesting cavity (as an indirect
measure of provisioning wasp size) influenced the absolute stinging effort (number of
stings on a prey specimen) and relative stinging effort (number of stings on a prey
specimen divided by its mass). In terms of absolute stinging effort, results indicated that a
weak (r2 = 0.01–0.19) positive dependence of the absolute stinging effort on the mass of
a prey specimen was significant in seven (and marginally significant in one) out of the
eight studied wasp species. Thus, the greater the victim’s mass the more stings it
received. Additionally, a weak negative dependence of absolute stinging effort on nest
cavity diameter (indirect measure of wasp size) was found for the wasps Symmorphus
allobrogus and Ancistrocerus antilope; therefore, in general, larger females of these
species delivered fewer stings to prey than smaller females. Relative stinging effort
depended negatively on prey mass in all eight species, indicating larger prey received
proportionately fewer stings than smaller prey. What’s more, for S. allobrogus there was
a negative association between relative stinging effort per victim and nest cavity
diameter, indicating that prey of a given mass were stung fewer times by larger wasps
324
than smaller wasps. Taken together these results indicate (1) although mass alone cannot
be considered the most essential prey feature determining stinging effort, the wasps can
flexibly adjust stinging effort (and thus presumably venom deployment) based on prey
mass, and (2) some wasps are able to assess prey size in relation to their own bodies,
thereby reducing stinging effort. Given the latter result, it would be intriguing to know
whether venom dose (venom volume per sting) is related to wasp size, as this could shed
light on how reduced stinging effort in these wasps is associated with total venom
deployment.
Other authors have also described the influence of prey size on stinging behavior
in wasps. For example, Casiraghi et al. (2001) examined the stinging behavior of the
digger wasp Ammophila sabulosa (Sphecidae) preying on caterpillars to provision its
nest. The authors found the total number of stings inflicted on a single prey was
significantly correlated with prey volume (r = 0.64, n = 42). The authors suggested the
correlation may reflect the need of an increased amount of venom to paralyze larger prey.
Similarly, Rathmayer (1978), citing the work of Bridwell (1920), noted the number of
stings inflicted by the bethylid wasp Scleroderma immigrans depended on the size of the
beetle larvae (Cerambycidae) prey. Likewise, Steiner (1986), citing Ferton (1897),
suggested that number of stings delivered might depend on the size of the prey, as the
sphecid wasp Tachysphex julliani stung small versus large specimens of its mantid prey
once or several times, respectively. Thus, by varying the number of stings, wasps vary the
amount of venom deployed based on the predatory situation, delivering more venom to
longer-struggling and larger prey.
325
Delivery Location and Aiming of Venom by Wasps
In addition to controlling the amount of venom used, wasps strategically deploy
venom by targeting specific sites on prey for venom delivery. The stinging behavior of
many solitary predatory wasps is thought to be a stereotyped sequence of motions closely
adapted to the anatomy and physiology of prey, although variations in stinging behavior
are recognized (Andrietti, 2011; Evans, 1966; Rathmayer, 1978; Steiner, 1986). In many
cases, sting locations match the locations of prey ganglia involved with locomotion,
attack, or defense, allowing the wasp to quickly and safely incapacitate its prey
(Andrietti, 2011; Steiner, 1986). An example of the precise targeting of venom can be
seen in the stinging of the cockroach P. americana by the wasp Ampulex compressa.
Using a combination of liquid scintillation and light microscopy autoradiography, Haspel
and colleagues (Haspel and Libersat, 2003; Haspel et al., 2003) showed that after A.
compressa stung a cockroach, 14C radiolabeled amino acids in the venom of the wasp
were localized in the cockroach’s first thoracic ganglion and specific regions of head
ganglia. The authors described this precise anatomical targeting of the venom as akin to
the most advanced stereotaxic delivery of drugs. In a similar example of precise targeting
of venom, Gnatzy and Otto (1996) provided evidence that the wasp Liris nigra
(Sphecidae) stung its cricket prey in the thoracic and subesophageal ganglia; using a
tethered wasp forced to sting a tethered prey, the authors used visual observation to
confirm the penetration of the sting into the nervous system. For numerous additional
examples of the specificity of the location of stinging in the solitary wasps, the reader is
referred to the reviews of Andrietti (2011) and Steiner (1986). Although the precise
localization of predatory venom delivery in solitary wasps is well known, there can also
326
be flexibility in delivery location. For example, Steiner (1986), citing Soyer (1938),
described how Anoplius concinnus (Pompilidae) adapts its stinging method to the kind of
prey attacked; Lycosa spiders, which live in shelters, are stung once and slowly near the
base of a leg, whereas running spiders like Pardosa are stung quickly near the mouth.
Wasps also target venom delivery in the context of defense. For example, Eisner
(1970) and Jeanne and Keeping (1995), citing Maschwitz (1964), noted that the social
wasps Vespa germanica and V. crabro sometimes ejected as an aimed spray the venom
they ordinarily inject with the sting; when grasped with forceps the wasps flexed the
gaster toward the stimulus and sprayed a stream of venom up to a distance of 3 cm. Since
the venom contains volatile alarm substances, the venom spraying alerts other wasps to
the presence of an enemy that has been topically “labeled” with the venom. Furthermore,
the sprayed venom may have intrinsic deterrent potential since it contains kinin and
histamine which could be topically irritating to vertebrates if the venom impinges on
sensitive surfaces such as the eyes (e.g., Jeanne and Keeping, 1995). Similarly, when the
wasp P. colobopterus was alarmed by scratching or vibration of its nest, workers on the
surface of the nest responded to objects moving within a few centimeters of the nest by
bending the gaster laterally and forward and aiming an atomized jet of venom (likely a
mucus membrane and eye irritant) in the direction of movement (Jeanne and Keeping,
1995). With their abilities to inject venom into specific regions of prey ganglia and spray
aimed venom discharges in defense, wasps clearly perform some of the most precise
venom deliveries among all venomous animals.
327
Ants
Cost of Venom in Ants
Use of venom among ants (Formicidae) is widespread (Blum, 1992; Blum and
Hermann, 1978; Holldobler and Wilson, 1990) and ants are known to use venom for
predation, defense, and brood tending (Holldobler and Wilson, 1990; Obin and Vander
Meer, 1985). Despite the relative ubiquity of venomous habits among ants, publications
addressing the direct costs of venom synthesis and use in these animals are scarce.
However, Haight’s (2002) consideration of the venom economy in colonial aculeate
hymenopterans helps to frame the issue. In these organisms the venom economy must be
considered at both the colony and individual levels. At the colony level, enough venom
must be produced and available to meet the colony’s needs without undue sacrifices to
growth and reproduction. At the individual level, workers must be able to generate, store,
and deliver venom sufficiently to meet the colony’s needs without unduly reducing their
usefulness to the colony in other areas. For example, a potentially severe ecological cost
could be incurred if workers expended too much venom subduing prey and as a result had
insufficient venom to act as useful nest defenders. For the fire ant, Solenopsis invicta
(Myrmicinae), although venom production by a worker makes up less than 6% of the
overall energetic cost of producing a worker, venom is expensive and important enough
that its use-rate is regulated (Haight and Tschinkel, 2003). The status of venom as a
valuable commodity for this species is due in part to its limited supply, as venom
synthesis ability is limited to a worker’s early life (Haight and Tschinkel, 2003). For a
number of species of the myrmicine genus Pogomyrmex (e.g., P. comanche and P.
maricopa), use of the venomous sting represents an extreme cost to the individual; often
employing their sting in defense against vertebrates, these animals leave their stinging
328
apparatus, together with its ganglia and glands, lodged in the their opponent, in a suicidal
yet involuntary altruistic act advantageous to the colony as a whole (Hermann, 1984).
Thus, evidence supports the idea that venom is a valuable commodity for ants and it’s use
comes at a cost.
Selective Sting/Venom Use by Ants
One way ants strategically deploy venom is by using it selectively, employing it
only when warranted, as determined by characteristics of the prey or aggressors. In terms
of predation, prey struggle (frequently correlating with prey size) often plays a role in
determining whether venom is deployed. In any given species of ant, selective sting use is
often not perfectly discrete in the sense that all ants always employ the sting for prey with
a specific set of characteristics (size/mass/type/propensity to struggle) while never
employing the sting for prey with a different set of characteristics. Rather, more
frequently, selective sting use manifests itself as a spectrum of differences in stinging
frequency of prey with varying characteristics. It seems reasonable to assume that interindividual variations in sting-use thresholds and subtle differences in individual prey
behavior during any given predator-prey interaction contribute to such observed stinging
frequencies.
The often-related prey characteristics of size, struggle intensity, and species all
appear to influence whether ants use their venom. For example, the African weaver ant,
Oecophylla longinoda (Formicinae), occasionally used sprayed venom when capturing
large prey (grasshoppers, 15–19 mm long; venom used in 15% [5/35] of trials), but never
used venom while capturing small (Drosophila, < 4 mm; venom used in 0% [0/294] of
329
trials) or medium-sized (Calliphoridae fly, 7–10 mm; venom used in 0% [0/532] of trials)
prey (Dejean, 1990). The arboreal ponerine ant Platythyrea conradti also demonstrated a
measure of selective sting use; small prey were sometimes retrieved without being stung
(28.4% [21/74] and 5.9% [4/68] for termite workers and small tettigonid larva,
respectively), whereas large prey (termite soldiers and large tettigonids) were always
stung (Dejean, 2011). Selective sting use also occurs in the African ponerine ant
Platythyrea modesta, as described in the study by Djieto-Lordon et al. (2001). When this
ant attacked small termite prey, Microcerotermes fuscotibialis (workers 2–3 mm, and
soldiers 3–4 mm), it frequently killed the weakly struggling termites using mandible
pressure and only rarely employed its stinger (10% [6/60] and 18% [11/60] of trials with
workers and soldiers, respectively). However, when P. modesta attacked small (2–4 mm)
workers of a different termite, Macrotermes bellicosus, the ant always employed the sting
(100% [60/60] of trials). Platythyrea modesta likewise frequently stung larger, more
vigorously struggling prey including M. bellicosus soldiers (6–8 mm, stung 87% [52/60]
of trials) and three sizes of unidentified grasshoppers (8–12 mm, stung 70% [70/100] of
trials; 12–16 mm, stung 100% [60/60] of trials; 16–20 mm, stung 80% [48/60] of trials).
Thus, P. modesta varied its stinging behavior based on prey type (e.g., Microcerotermes
vs. Macrotermes workers) and the related variables of prey size and struggle intensity
(Djieto-Lordon et al., 2001). Media workers of the carpenter ant Camponotus maculatus
(Formicinae) sprayed venom facultatively based on prey struggle when attacking small
prey, but always sprayed larger prey, which always struggled after being attacked
(Dejean, 1988). For example, after relatively small prey, such as live termite workers
(Allognathotermes, 4–6 mm), were bitten by C. maculatus, 56.7% (17/30) of the termites
330
struggled and 29% (5/17) of these struggling prey were sprayed with venom. In contrast,
of the 43.3% (13/30) of termites subdued by biting alone (no visible struggle) only 15%
(2/13) were sprayed with venom. For comparison, when offered recently killed termite
workers, C. maculatus did not spray any (0/20) of these non-struggling prey prior to
transport. When preying on larger prey such as T. molitor larvae (15–18 mm), 100%
(20/20) of the prey struggled after being bitten and C. maculatus sprayed venom on all of
these struggling prey. Thus, prey struggle is a key characteristic assessed by C. maculatus
in determining whether to use venom. In fact, Dejean (1988) noted that for several groups
of ants (Myrmecia, Ponerinae [Odontomachus, Hypoponera, Mesoponera], Myrmicinae
[Aphaenogaster, Dacetini]) prey struggle is the stimulus that elicits stinging behavior.
Further evidence in support of the key role of prey struggle is found in the fact that ants
frequently do not sting anesthetized or dead prey items (Daly-Schveitzer et al., 2007;
Dejean, 1985, 1986; Orivel et al., 2000; Robertson, 1971). In addition to prey struggle,
distance of prey from the ants’ nest and colony hunger level may influence sting use
(Cerda and Dejean, 2011; Dejean, 1985).
Predatory selective sting use is also demonstrated by the trap-jaw ants in the
genus Odontomachus (Ponerinae), as described by De la Mora et al. (2008). For example,
O. opaciventris, when presented with several types of small prey that differed in
morphological or defensive characteristics, occasionally stung dealated tephritid fruit
flies (Anestrepha oblique, 6.8–8.2 mm; 28% [11/40] of trials), rarely stung sclerotized
tenebrionid beetle larvae (T. molitor, 13–18 mm; 6.7% [2/30] of trials), but never stung
termite workers (Nasutitermes sp., 4.4 mm; 0% [0/30] of trials) or chemically defended
soldiers (Nasutitermes sp., 3.8 mm; 0% [0/20] of trials). The authors suggested that
331
stinging of flies might have been the result of the prey’s bulky shape and jumping escape
movements, whereas the absence of stinging of termites likely resulted from their small
size relative to the ant. Similar selective sting use occurred in O. troglodytes faced with
prey of different sizes and thus struggling capacities; anesthetized prey
(Allognathotermes workers, 5–6 mm) and small prey (Microtermes workers, 2 mm)
immobilized (stunned or killed) by the ant’s trap-jaw were never stung, medium-sized
prey (Allognathotermes workers, 5–6 mm; T. molitor larvae, 5–6 mm) were stung only if
still struggling after the trap-jaw blow, and large prey (T. molitor larvae, 10–11 mm),
which invariably struggled after impact of the trap-jaw, were nearly always stung (Dejean
and Bashingwa, 1985). Selective sting use has also been reported for the ant O. bauri,
which relies on its trap-jaw to subdue termite prey (Ehmer and Holldobler, 1995) but
sometimes uses its sting in defense against ants from different conspecific colonies (Jaffe
and Marcuse, 1983). Similarly, O. ruginodis will sting some prey such as ant larvae but
will rely on its trap-jaw to subdue termite prey such as Nasutitermes and Reticulitermes
(Carlin and David, 1989). Selective use of the sting by Odontomachus ants during
predation, and relying when possible on their powerful mandible strike instead of venom
deployment, would limit the energetic cost of predation, preserving the workers’ ability
to use their sting against difficult-to-handle prey or for colony defense (De la Mora et al.,
2008).
Ants may also employ their sting/venom selectively in the context of defense. For
example, workers of O. ruginodis guarding the nest used their stings on conspecifics
from different colonies and on vertebrate nest-intruders, but relied on their trap-jaws to
repel other ants, including species larger than themselves such as Camponotus floridanus,
332
C. tortuganus, and Pogonomyrmex badius (Carlin and David, 1989). The study of Marlier
et al. (2004) also sheds light on selective venom use in ants, specifically the ant
Crematogaster scutellaris (Myrmicinae), which, as with other Crematogaster ants, has a
reduced, spatulate sting used to apply venom topically onto the integument of enemies. In
dyadic encounters between a resident worker of C. scutellaris and a homo- or
heterospecific intruder, C. scutellaris preformed classical aggressive behaviors such as
grips and grip attempts, but also performed gaster flexions, sometimes emitting venom
and attempting to apply it on the enemy. During these aggressive encounters, C.
scutellaris used its venom with parsimony; although gaster flexions were frequently
observed, a droplet of venom was present at the tip of the abdomen in only 10-30% of
them, and a worker rarely emitted more than one droplet per encounter even though each
ant is capable of producing up to 90 droplets. Additionally, the use of venom increased
with the aggressiveness of the intruder, from 26% (76/293) of encounters with C.
lateralis minors, the least aggressive ant, to 52% (56/108) of encounters with C.
cruentatus, one of the most aggressive species. In contrast, C. scutellaris never used
venom (0/326 encounters) during aggressive intraspecific encounters or during prey
capture (0/9 trials). Clearly ants are selective in their venom deployment, with the
propensity to sting higher for larger, more vigorously struggling prey, and for more
aggressive potential predators.
Amount of Venom Deployed by Ants
Strategic deployment of venom by ants includes not only selective sting/venom
use, but also control over the amount of venom deployed when emission is warranted. In
333
terms of predation, evidence for the control of amount of venom deployed has only been
measured at the level of number of individual stings delivered. However, it’s possible
that duration of stinging may also play a role (Dejean and Lachaud, 2011).
Prey size and weight are important factors influencing the amount of venom
deployed by ants as measured by number of stings. For example, when the ponerine ant
Pachycondyla pachyderma attacked various centipedes, the number of stings an ant
delivered increased with prey size, with the smallest centipedes (lithobiomorphs)
receiving 0.9 stings, on average, and the largest centipedes (scolopendromorphs)
receiving 1.9 stings (Dejean and Lachaud, 2011). Another detailed example is found in
the study of Daly-Schveitzer et al. (2007), in which the ant Gnamptogenys sulcata
(Ectatomminae) varied the number of stings delivered based on prey size and type.
Gnamptogenys sulcata retrieved prey using either a solitary or collective strategy
depending on prey size, with mean mass and length of prey retrieved solitarily
significantly less than that of prey retrieved collectively. Regardless of prey attributes and
retrieval strategy, a single ant performed the attack, with the predator stinging the prey
until it was immobilized. The number of stings required to immobilize prey retrieved by
the collective strategy (n = 27) was significantly higher than for solitarily retrieved (n =
30) prey (3.7 vs. 1.8 stings, respectively). What’s more, for prey retrieved solitarily,
number of stings was positively correlated with mass (ρ = 0.51) and length (ρ = 0.51). In
addition, prey type influenced stinging behavior, as repeated stinging was nearly twice as
frequent against small mealworms than against small crickets (59.4% vs. 34.4% of the
sequences, respectively). For prey retrieved collectively, number of stings was positively
correlated with length (ρ = 0.44). When investigators artificially manipulated prey weight
334
while holding prey size constant, G. sulcata delivered more stings to heavier prey. For
example, when ants were presented with “small” prey (live T. molitor larvae, 16 mm) of
two different weight classes, the heavier larvae artificially weighted with lead weights,
ants delivered a significantly greater number of stings, on average, to the heavier prey
than the lighter prey (3.6 vs. 1.9 stings, respectively). When presented with small but
“infinitely heavy” prey (live T. molitor larvae, <16.5 mm, pinned to substrate), G. sulcata
delivered significantly more stings to “infinitely heavy” prey than to the larvae weighted
with lead (7.7 vs. 3.6 stings, respectively). A similar result was obtained when G. sulcata
was presented with “large” prey (live T. molitor larvae, 21 mm) of two different weight
classes, with the weight of the lighter group manipulated by hemolymph removal; ants
delivered half as many stings, on average, to the lighter prey than the heavier,
unmanipulated, prey (1.7 vs. 3.2 stings, respectively).
In addition to the influence of prey size and weight, the amount of venom
deployed by ants, as controlled by number of stings delivered, is often associated with
prey struggle. For example, the ant Myrmecia gulosa (Myrmeciinae) attacking blowflies
and mealworms was observed to continue stinging the prey as long as any movement
persisted, and it was suggested that tactile perception was the most probable releaser of
continued stinging behavior (Robertson, 1971). Similarly, intensity of prey struggle
(related to prey size) was associated with number of stings delivered by P. modesta
attacking termites and grasshoppers (Djieto-Lordon et al., 2001). Likewise, Metapone
madagascarica (Myrmicinae), attacking the termite Cryptotermes kirbyi, stung the prey
repeatedly, between 3 and 11 times, until the prey stopped moving (Holldobler et al.,
2002). In a similar fashion, when the ant Ectatomma ruidum (Ectatomminae) attacked T.
335
molitor larvae, repeated stinging occurred when prey continued to struggle (Schatz et al.,
1997). The ant Plectroctena minor (Ponerinae) behaved similarly when attacking
spirostreptid millipedes, stinging repeatedly until the prey became motionless (Dejean et
al., 2001). Thus, prey struggle influences the number of stings and the amount of venom
ants deliver to their prey.
In the context of defense, there is evidence that ants can control the amount of
venom deployed by number of emissions as well as volume per emission. An example of
the former is seen in workers of the ant Pachycondyla tridentata which emits a venom
foam when disturbed, typically releasing 3–8 mm3 of foam at each stimulation
(Maschwitz et al., 1981). When the ant is stimulated intensively for long periods it is
capable of producing foam volumes of up to 26 mm3 over the course of 20 foam releases.
The capacity to control venom deployment by volume per emission was described by
Obin and Vander Meer (1985) in their study of the fire ant S. invicta. This ant controlled
the quantity of venom released as an aerosol during gaster flagging depending on context;
more venom (up to 500 ng) was released during gaster flagging at heterospecifics in a
foraging arena than was released when brood tending (~1 ng), where venom is assumed
to have an antibiotic function. Exactly how the ants control the amount deployed is
unknown. S. invicta also controls the volume of venom deployed per sting (measured as
volume/headwidth3) based on context (Haight, 2002; Haight and Tschinkel, 2003). For
example, Haight (2006) discovered that S. invicta workers (n = 225) flooded from their
nest and rafting on the water’s surface delivered a significantly larger average venom
dose (1.41 nL/mm3) than workers stinging in defense pre-flood (0.76 nL/mm3). The
author suggested S. invicta increased defensiveness while rafting as a result of increased
336
vulnerability, and the larger venom doses deployed by ants in rafting colonies may reduce
their chances of being damaged by encounters with other animals. In a similar study, S.
invicta workers were found to deliver significantly greater venom volumes per sting in
spring (corresponding to the timing of production of sexuals) than in the fall and winter.
In fact, venom dose was more than 55% greater in spring (1.15 nL) than in the rest of the
year (0.74 nL, average of summer, fall, and winter), with majors delivering doses in
spring 2.7 times those they did in the summer (1.35 vs. 0.50 nL, respectively; Haight,
2002; Haight and Tschinkel, 2003). The authors suggested that increased venom doses in
spring may represent an increased investment in protecting the extremely valuable
reproductive castes, as higher venom doses should repel offending organisms more
quickly and effectively. Taken together, the evidence is conclusive that ants vary the
amount of venom deployed dependent on context; ants deploy more venom via a greater
number of stings to larger, heavier, and longer-struggling prey, and more venom via a
larger volume per sting when vulnerability to predators is higher.
Delivery Location of Venom by Ants
In addition to controlling when and how much venom used, ants strategically
deploy venom by targeting where the venom is delivered. For example, P. conradti
preying on tettigonid larvae and termites (M. bellicosus) was observed in all cases to
sting prey on their ventral surface where the neural chain passes, thus hastening paralysis
(Dejean, 2011). When P. minor attacked spirostreptid millipedes (30-35 mm long), nearly
90% were stung on the anterior half of the body. Notably, the more anteriorly the
millipedes were seized and stung, the shorter the duration between stinging and the end
337
of transport to the nest, representing less effort handling prey (Dejean et al., 2001). In a
strategy presumably employed to secure quick prey paralysis, the ant G. sulcata often
targeted its stings (Daly-Schveitzer et al., 2007). Although there was no distinct stinging
pattern for solitarily retrieved G. assimilis nymphs, by contrast, G. sulcata performed
highly non-random stinging for solitarily retrieved T. molitor larvae. For these latter prey,
stings were directed toward the cephalic region more frequently (63% of stings) than
predicted by chance. When G. sulcata retrieved prey using a collective strategy, the
stinging pattern was nonrandom for both prey species; the majority of stings to both G.
assimilis and T. molitor were located in the cephalic region (91% and 65% of stings,
respectively). Thus, for some ants, delivery of venom to specific locations on prey is
likely an important aspect of strategic venom deployment.
Snakes
Cost of Venom in Snakes
In a recent review of the venom optimization hypothesis, Morgenstern and King
(2013) summarized the limited research to date on the metabolic cost of venom
regeneration in snakes, and concluded the cost of venom production is not trivial. For
example, working with three species of North American pitvipers, McCue (2006) found
the snakes demonstrated an 11% increase in resting metabolic rates during the first 72
hours of venom replenishment. Given that venom regeneration in snakes can take several
weeks (Kochva et al., 1982; Oron and Bdolah, 1973; Rotenberg et al., 1971), venom
regeneration represents a significant metabolic load (Morgenstern and King, 2013).
Although Pintor et al. (2010) contended that venom regeneration in the death adder
Acanthophis antarcticus (Elapidae) represented a relatively small cost (26% of the cost of
338
digesting a small prey and 6% of the cost of shedding), Morgenstern and King (2013)
argued that only 3–4% of the snakes’ total venom supply had been depleted in the Pintor
study. When Morgenstern and King (2013) adjusted the data to reflect full depletion of
venom, the cost of venom regeneration was significant, at 5.8 times higher than prey
digestion and 1.7 times higher than shedding. It is this considerable metabolic cost of
venom regeneration, along with the ecological costs of a depleted venom supply,
including reduced ability to procure additional prey or mount a sufficient defense, that are
hypothesized as the selection pressures driving venom metering in snakes (Hayes, 2008;
Hayes et al., 2002).
Selective Venom Use by Snakes
Strategic deployment of venom by snakes includes selective use of this valuable
resource, employing it, presumably, only when circumstances demand it. Venomous
snakes do not always employ venom when taking prey. In addition to venom, some of
these snakes also use constriction as a weapon in their offensive arsenal (Rochelle and
Kardong, 1993; Shine, 1985). In some cases, such as when prey is released from the bite
before death, venom use and constriction may complement each other in prey capture
(Mackessy et al., 2006; Morgenstern and King, 2013; Rochelle and Kardong, 1993).
However, under certain circumstances, venomous snakes may rely solely on constriction
for killing prey. For example, Rodriguez-Robles (1992) reported that the Puerto Rican
Racer, Alsophis portoricensis (Colubridae), sometimes killed mice solely by constriction.
However, the snake also killed mice by venom alone or using a combination of venom
and constriction. It is unclear what cues the snake used in determining which
immobilization technique to employ, but the author speculated that a combination of prey
339
size and prey type was involved. Hayes (2002), interpreting the work of Jones (1988),
argued that the colubrid Trimorphodon biscutatus uses its venom to paralyze lizards, but
relies largely on constriction to kill mice. Even in cases where constriction is not used as
an alternative to envenomation, snakes may use venom selectively. In a separate
investigation of A. portoricensis, this time preying on lizards (Anolis cristatellus) and
frogs (Eleutherodactylus coqui), it was found that although the snakes never constricted
either prey type, they envenomated most lizards (76% [13/17]) but only one frog (7%
[1/14]; Rodriguez-Robles and Leal, 1993). The authors suggested that the snakes used
venom to subdue lizards rather than frogs because the lizards retaliated to attacks with
biting and tail lashing whereas frogs exhibited few antipredatory behaviors. Visual and
chemical stimuli were also suggested as signals used by the snake to “decide” whether to
envenomate prey (Rodriguez-Robles and Leal, 1993). In addition to these examples,
Hayes et al. (2002) pointed out that there are numerous anecdotal observations (e.g.,
Klauber, 1997a; Radcliffe et al., 1980; Savitzky, 1992) suggesting that venomous snakes
swallow without envenomation a number of prey types (invertebrates, neonatal
vertebrates, fish, amphibians) that can be ingested with minimal risk or struggle.
In snakes, as in other venomous animals such as scorpions and spiders, selective
venom use may include using the venom delivery apparatus without deploying venom
(Morgenstern and King, 2013). Offensive dry bites have not been documented, however
bites deploying very small amounts of venom have been noted, with prey surviving for
long periods before expiring (Morgenstern and King, 2013). Dry defensive bites,
however, are well documented in snakes (De Rezende et al., 1998; Whitaker et al., 2000),
and the frequency of defensive dry bites has been estimated to be as high as 50% (Hayes
340
et al., 2002). Based on the high frequency of dry bites and the lack of correlation between
duration of fang penetration and total venom injected during predatory bites, Morgenstern
and King (2013) argued that although some defensive dry bites might be due to kinematic
constraints, most result from a decision aimed at conservation of venom. Defensive sham
or bluff strikes, in which the snake closes its mouth and does not bite, are also
documented (Rowe and Owings, 1990; Whitaker et al., 2000). Hayes (2008) contended
that sham strikes are consistent with the interpretation that snakes make decisions about
venom use, either pre- or mid-strike. Another example of defensive selective venom
deployment in snakes is seen in the spitting cobras Naja nigricollis and N. pallida
(Elapidae), whose sophisticated sensory systems enable them to differentiate between
potential targets, saving venom expenditure for the best targets and the most likely
threats. If threatened, these snakes can aim their venom, ejecting it from their fangs as
distinct jets or a fine spray (Westhoff et al., 2005). Investigators discovered that these
snakes could be triggered to spit venom at a moving human face or real size photo of a
human face, but would not spit at a stationary human face (real or photo) or a moving or
stationary human hand (Westhoff et al., 2005). The results suggested that cobras can
visually differentiate between hands and faces, and the authors argued that since the
venom of the spitting cobras only has an impact on an aggressor if it hits the eyes, this
defensive strategy of only spitting at moving faces is adaptive. Data from a separate study
demonstrated that target shape was important in eliciting spitting from cobras; oval- and
round-shaped targets resulted in a higher spitting frequency (80%), whereas triangles
with the same surface area hardly elicited spitting (23.9%). Such findings further
contribute to an understanding of how snakes may differentiate between potential
341
aggressors and inanimate moving objects (e.g., leaves or branches moving in the wind) in
the environment, and thus avoid wasteful deployment of venom (Berthe et al., 2013). As
a whole, these results support the idea that snakes selectively deploy their venom, using it
more often for prey types more apt to struggle or retaliate, and sometimes separating its
use from the biting act employed for defense. Snakes’ keen senses are important in
assessing whether venom deployment is warranted.
Amount of Venom Deployed by Snakes
Of all venomous animals, the capacity to control the amount of venom injected
has been most extensively studied in snakes (Morgenstern and King, 2013). In fact,
several reviews on venom metering in snakes are available (e.g., Hayes, 2008; Hayes et
al., 2002; Morgenstern and King, 2013; Young, 2008; Young et al., 2002), and the reader
is referred to these sources for a deeper discussion of the topic. Although objections have
been raised against the venom-metering hypothesis, and the mechanisms leading to the
control of venom deployment remain unknown, an increasing amount of evidence
supports the contention that snakes can control venom expulsion based on circumstances
(Hayes, 2008; Morgenstern and King, 2013). Research demonstrates that snakes vary the
amount of venom injected based on numerous factors, including prey size, prey type, risk
of prey escape, hunger level, predatory vs. defensive context, and threat level (Hayes,
2008; Hayes et al., 2002).
342
Delivery Location and Aiming of Venom by
Snakes
It is unclear what role targeted venom injection may play in the strategic
predatory deployment of venom by snakes. Given the possibility that struck and released
prey might escape beyond recovery range of the snake if not efficiently envenomated
(Kardong, 1986b), it is expected that delivery location of venom could be strategically
important if it influenced time to death of prey (Hayes, 1991). Indeed, for some snakes,
time to death of prey does vary with site where venom is injected (Hayes, 1991; Kardong,
1986a; Minton, 1969). However, for other snakes no such relationship exists (Rochelle
and Kardong, 1993). In support of a role for targeted venom delivery, an investigation by
Kardong (1986a) of the predatory strike behavior of the rattlesnake Crotalus viridis
oreganus (Viperidae) preying on mice, found that most mice (71% [549/768]) were
struck on the head/thorax, with fewer mice struck on the mid-region (posterior edge of
the thorax to the pelvic girdles, 23%) and the rump (6%). Mice struck in the head/thorax
died more quickly than mice struck elsewhere. What’s more, when struck in the rump,
some mice (18%) occasionally successfully delivered a retaliatory bite to the snake;
fewer of the mice struck in the head/thorax or mid-region bit the snake (3% and 9%,
respectively). The author speculated that venom absorbed by the heavily vascularized
lungs was the basis for quicker death in mice envenomated in the head/thorax. The author
argued that although the exact roles played by vision and thermoreception in directing the
strike to the head/thorax are not known, the high rate of strikes to this region suggests
that snakes targeted the most vulnerable site on the mouse. In another study, the attack
behavior of the Puerto Rican Racer (A. portoricensis) on the lizard A. cristatellus was
investigated; results indicated the highest frequency of attacks (50%) was aimed at the
343
body of the prey (as opposed to head, limbs, or tail), and that body attacks gave the
greatest attack success (82.4% success per strike) (Rodriguez-Robles, 1992). The role
played by the Duvernoy’s gland venom in these attacks, however, was unclear.
Furthermore, the reason why the body was the most common site struck on the anole may
be that it was the longest part of the lizard (presenting the largest target), or because the
snakes specifically targeted this area based on previous successes (Rodriguez-Robles and
Leal, 1993). However, even if snakes do target bites to specific regions of their prey,
snake strikes are not always accurate. In fact, snakes sometimes miss prey with one or
both fangs (Kardong, 1986b). Even so, an unsuccessful strike may be followed
immediately by a readjusted strike or by fang repositioning after the jaws have made
contact with the prey (Kardong and Bels, 1998). Furthermore, it has been speculated that
snakes may make rapid movements of the head during a strike to “fine tune” the impact
point of the snake’s fangs on the target (Young et al., 2009).
In the case of spitting cobras, which eject venom to a distance of up to 3 meters
toward the face of an aggressor (Berthe et al., 2009), it is more obvious that venom
delivery is targeted. Venom spitting is used only for defense, and although the venom has
little effect on unbroken skin, even small amounts of venom can damage eyes (Westhoff
et al., 2010; Young and O'Shea, 2005). In fact, many reports claim that spitting cobras
aim at the eyes of an aggressor (Berthe et al., 2009; Young et al., 2009). A handful of
studies have quantitatively investigated spitting behaviors, showing that rapid movements
of the snake’s head during venom expulsion produce the spatial-dispersal patterns of spit
venom (Westhoff et al., 2010; Young et al., 2009). A study by Westhoff et al. (2005)
investigating the distribution of spit venom on the eyes and face of real and photographic
344
targets found that N. nigricollis hit at least one of the target’s eyes in 80% (28/35) of
trials, and that N. pallida hit at least one eye in 100% (10/10) of trials. The authors
suggested the snakes aimed either at the middle of the face or at the area between the
eyes. Data from a separate study suggested that cobras do not intentionally try to hit the
eyes of an aggressor, but rather at the center of the body part closest to the eyes (Berthe et
al., 2013). Another investigation, using N. pallida and N. nigricollis, demonstrated cobras
adjusted their spitting behavior according to target distance (Berthe et al., 2009). The
authors hypothesized that to optimize the spitting act the snakes should decrease their
spitting angles (i.e., amplitude of head movements) with increasing target distance so the
venom would cover the face of the antagonist but not exceed its width and height. Data
revealed the cobras did indeed decrease horizontal and vertical spitting angles with
increasing target distance, although on average snakes made small systematic errors such
that the spitting pattern was slightly larger than the actual target size. Finally, research
has shown that spitting cobras can accurately track the movements of a potentially
threatening vertebrate, and by anticipating its subsequent short-term movements direct
their venom to maximize the likelihood of striking the target’s eye (Westhoff et al.,
2010). As a whole, data is relatively thin in support of targeted delivery of snake venom
during predation, although some evidence appears to show that snakes may target heavily
vascularized prey regions. In defense, however, spitting cobras demonstrate great
accuracy in venom delivery.
345
Strategic Deployment of Non-Venomous Predatory Emissions
Glue
Scytodid Spiders
Spitting spiders (Scytodidae) eject a mixture of silk, glue, and perhaps venom
(hereafter referred to as spit) from a pair of complex venom glands opening from their
chelicerae, and use the sticky mix to entangle prey prior to envenomation, as well as for
defense (McAlister, 1960; Suter and Stratton, 2012). As the spit is a limited resource that
takes time to replenish (Clements and Li, 2005), is energetically expensive to produce
(Suter and Stratton, 2012), and its depletion leaves the spider vulnerable to predation and
unable to deal with subsequent prey, scytodids are thought to use the spit judiciously
(Clements and Li, 2005). Scytodid spiders strategically deploy their sticky spit by
selectively using spit for large prey, and by regulating the amount of spit discharged
based on prey size and struggle intensity, the latter being supported by evidence
indicating the spider’s ability to vary the characteristics of the spiting act. Entangling
prey with the spit at a distance underscores the spider’s ability to aim, another facet of the
strategic deployment of spit.
The first level of control that scytodids manage over the amount of spit expended
is whether to discharge spit at all. Limited evidence suggests that factors such as prey
type, prey size, and prey struggle intensity may play roles in determining whether spiders
use their valuable spit. For example, Scytodes sp. did not always use spit to capture
stemborer moths (Chilo suppressalis) prior to seizing and envenomating these prey (Li et
al., 1999). Also, a number of spiders were excluded from Clements and Li’s (2005)
analysis of spit expenditure related to prey size and struggle intensity because the spiders
did not spit. Over 80% (13/16) of non-spitters in this study were spiders presented with
346
small prey and/or prey with low struggle intensity, and of these, nearly 62% (8/13) chose
to directly envenomate prey instead of spitting. The authors argued that direct
envenomation of small prey without spitting is advantageous, as it would be energetically
wasteful to expend spit on small prey. Furthermore, spiders have individual control over
discharge from each fang; although apparently rare, spiders can spit from a single fang at
a time (Li et al., 1999).
When spitting spiders do spit, they have sophisticated control over the
composition and quantity of their spit. Regarding the former, investigators postulate that
just as the spiders can discharge venom without silk and glue, they may also be able to
eject silk and glue without venom (Suter and Stratton, 2012), and this may explain why
the spit is not toxic to entangled prey (Clements and Li, 2005). In addition to controlling
the composition of the venom glands’ discharge, the spiders also vary the amount of spit
deployed based on prey size and struggle intensity. Li et al. (1999), studying Scytodes sp.
from the Philippines, first noted that the spiders varied their spitting behavior with prey
size and struggle intensity. When spiders spat at small prey, they usually spat only once,
whereas the spiders spat multiple times (up to 8 spits in succession) at large and
vigorously struggling prey. These authors also documented that the amount of fluid
ejected per spit varied with prey size. Additional investigation using large and small
crickets artificially vibrated at two different amplitudes (simulating different struggle
intensities) revealed that Scytodes pallida spat a significantly greater mass of spit at
larger prey and also at prey vibrated at higher amplitude (Clements and Li, 2005). The
larger amount of spit ejected onto larger and more vigorously struggling prey may be a
response to the necessity of securing a body with a larger surface area, and to prevent the
347
spiders from being injured by or losing highly animated prey (Clements and Li, 2005).
Since scytodids have poor vision, prey size and struggle intensity may be gauged by
using their long tarsi, which allow probing of prey at a safe distance (Clements and Li,
2005). Work by Suter and Stratton (2009) with Scytodes thoracica indicated that multiple
spits by the same spider varied in duration, as well as the length, velocity, volume, and
ejection rate of the spit thread. There was also considerable variation in these parameters
between individual spiders. Taken together, these data suggest the range of possible
spitting responses is large, and that spiders may vary several dynamics of the spitting act.
Finally, as spiders may entangle prey with spits traveling as far as 60 mm (Li et al.,
1999), the spiders’ aim helps ensure that spit is deployed so as to maximize its
effectiveness. As a whole, data on glue use by scytodids indicate these spiders
differentiate between different prey types, sizes, and struggle intensities, discharging
more gluey spit onto larger and more vigorously struggling prey.
Velvet Worms
Velvet worms (phylum Onychophora) are terrestrial, many-legged, invertebrate
carnivores, which capture their prey (including isopods, termites, and spiders) and defend
themselves by squirting a slimy, entangling glue (Blaxter and Sunnucks, 2011; Dias and
Lo-Man-Hung, 2009; Read and Hughes, 1987). Velvet worms strategically deploy their
valuable glue by squirting at some prey but not others, modifying their firing threshold
depending on hunger level, and varying the amount of glue discharged with prey struggle
intensity and size.
The velvet worm’s glue, comprising up to nearly 13% of body mass, is produced
and stored in large glands situated on each side of the gut within the body cavity, and is
348
ejected via a pair of modified limbs, the slime papillae (Baer and Mayer, 2012; Read and
Hughes, 1987). The glue gland reservoirs, where glue is stored prior to ejection, are
surrounded by prominent layers of muscle that, along with contractions of the somatic
musculature, provide the force to squirt the glue a distance of up to 4 cm (Baer and
Mayer, 2012; Read and Hughes, 1987). The glue is composed of unique, high-molecularweight proline-rich proteins (200+ kDa), lectins, and small peptides, with protein making
up 55% of the glue’s dry mass (Haritos et al., 2010). Investigators contend that the glue is
energetically expensive to produce given the large amount of protein involved, although
the metabolic cost of glue secretion has not been measured (Haritos et al., 2010; Read
and Hughes, 1987). The lengthy period of time required for replenishing a depleted glue
supply (~24 days; Read and Hughes, 1987) lends credence to the hypothesis that glue is
metabolically expensive. As further evidence that glue is a valuable resource, velvet
worms re-ingest large amounts of their expended glue, and even when their prey escape,
velvet worms will return to the site of attack to eat the glue left on the ground (Read and
Hughes, 1987). Beyond the metabolic cost of regenerating glue constituents, there is also
an ecological cost to squirting glue, in that depleted glue reserves render velvet worms
less capable of attacking further prey or of defending themselves (Read and Hughes,
1987). In fact, velvet worm feeding frequency may be limited by the rate of glue
replenishment (Read and Hughes, 1987). Given the cost and importance of their glue,
velvet worms are expected to deploy their glue judiciously (Read and Hughes, 1987).
Although not stringently tested, there is some evidence indicating that part of
strategic glue use in velvet worms involves squirting at larger prey but not at smaller
prey. For example, Read and Hughes (1987) found that the velvet worm Macroperipatus
349
torquatus (Peripatidae) always squirted glue at larger prey such as crickets and
cockroaches, but sometimes seized smaller prey, woodlice, with only its jaws, capturing
and consuming these smaller prey without deployment of any glue. How velvet worms
assess prey characteristics such as size prior to squirting glue is unknown, however they
likely use sense organs on the antennae. In addition to being selective in their glue use,
velvet worms vary their selectivity based on hunger, lowering their threshold for glue
discharge when they’re hungry. Normally M. torquatus squirted glue only at prey within
a distance of about 0.5 cm and which were not moving extensively; however, when
starved, the velvet worm sometimes squirted at rapidly moving prey from a distance of
up to 4 cm. Selective use of glue dependent on prey size, and variation of discharge
threshold with hunger both indicate the velvet worm regulates glue expenditure.
Velvet worms also strategically deploy their glue by varying the amount
discharged dependent on prey struggle intensity and size. Read and Hughes (1987),
studying the behavior of M. torquatus preying on crickets, cockroaches, spiders, and
woodlice, found that one way the velvet worm varied the amount of glue delivered to
prey was by the number of squirts. With prey that was relatively quiescent, one squirt of
glue was often sufficient to subdue the prey. However, violently struggling prey would
receive multiple squirts, including squirts directed at the limbs. Spiders, the most active
and potentially dangerous prey, received up to 30 squirts of glue. The authors noted that
the amount of glue M. torquatus squirted was significantly correlated with the relative
size of the prey, and that there were significant differences in this relationship among
prey types. Glue expenditure could be extensive, with up to 80% of glue reserves,
representing more than 10% of body mass, squirted at large Aclodes and Lutosa crickets.
350
The mass of glue used per unit mass of prey decreased asymptotically with increasing
relative prey size, and this relationship differed among prey types, with Aclodes crickets
requiring the least and spiders (unidentified ctenid) the most glue at similar sizes. Given
evidence that velvet worms are capable of complex behavior (e.g., organization of social
groups into female dominance hierarchies, collective hunting; Reinhard and Rowell,
2005), perhaps it should not be surprising that velvet worms strategically deploy glue
based on characteristics of their prey. Taken together, the evidence indicates that velvet
worms take into account prey characteristics including size and struggle intensity when
determining whether to deploy glue, and how much to expend.
Strategic Deployment of Non-Venomous Defensive Secretions
Upon disturbance many animals often squirt, ooze, or otherwise release noxious
substances at potential aggressors (Berenbaum, 1995; Eisner and Meinwald, 1966;
Rosenberg et al., 1984; Whitman et al., 1990). These chemical defenses, substances
produced to reduce the risk of bodily harm, are widespread among animals, and their
presence, in general, reflects the probability of attack and relative risk of damage
(Berenbaum, 1995; Pasteels et al., 1983; Ruxton et al., 2004). Thus, chemical defenses
are less abundant among parasitic organisms and organisms at the top of the food chain,
but organisms that cannot run away from potential predators, including many
invertebrates, are well represented among the chemically defended (Berenbaum, 1995;
Ruxton et al., 2004). Defensive secretions used by organisms for protection against
potential aggressors are often complex mixtures, and are generally chemically reactive
products of secondary metabolism (Berenbaum, 1995; Pasteels et al., 1983; Ruxton et al.,
2004). It is widely understood that defensive secretions are judiciously conserved to
351
avoid the metabolic and ecological costs of depleted defensive stores (Blum, 1981, 1985;
Eisner and Meinwald, 1966; Fescemyer and Mumma, 1983; Higginson and Ruxton,
2009; Krall et al., 1999; Whitman et al., 1990).
Cost of Non-Venomous Defensive Secretions
A recurrent theme in the literature is that chemical defenses confer a benefit but
typically also exact a cost (cf. Kearsley and Whitham, 1992; Ruxton et al., 2004)),
although measuring these has proven exceedingly difficult (Berenbaum, 1995; Ruxton et
al., 2004). As with venoms, the emission of non-venomous defensive secretions are
associated with costs of synthesis, transportation, storage, and prevention of autotoxicity
(Berenbaum, 1995; Bowers, 1992; Higginson and Ruxton, 2009; Ruxton et al., 2004).
Furthermore, maintaining systems for external discharge, delivery, or activation of
secretions likely represents a non-trivial metabolic cost (Bowers, 1992; Higginson and
Ruxton, 2009). Resources allocated for maintaining defensive secretions represent energy
that can’t be invested in growth and adult size, which in turn might negatively impact
adult survival and reproductive success (Bowers, 1992; Higginson and Ruxton, 2009;
Ruxton et al., 2004; Sato et al., 2009). Significant metabolic costs of defensive secretions
might also be inferred based on the often lengthy amount of time required for
regeneration of secretions, varying from days (Blum, 1978; de Jong et al., 1991; Eisner
and Meinwald, 1966; Eisner et al., 1961; Nolen et al., 1995) to weeks (Eisner, 1960,
1965; Fescemyer and Mumma, 1983; Read and Hughes, 1987; Whitman et al., 1992), and
even months (Carrel, 1984; Krall et al., 1999; Rossini et al., 1997). In contrast, relatively
rapid regeneration periods for some defensive secretions (Eisner et al., 1971; Roth and
Eisner, 1962) may imply that not all defensive secretions are metabolically expensive.
352
The cost of emitting defensive secretions can also be seen in the ecological cost of
increased vulnerability to aggressors when secretions run low (Higginson and Ruxton,
2009; Slobodchikoff, 1979). Authors have noted the presumed costliness of defensive
secretions in a number of animals, including harvestmen (Gnaspini and Hara, 2007),
beetles (Hetz and Slobodchikoff, 1990; Holloway et al., 1991), sea snails (Bancala,
2009), sea hares (Derby, 2007), and horned lizards (Sherbrooke and Middendorf, 2001).
As was the case with venoms discussed above, due to the valuable nature of defensive
secretions, it may be advantageous for chemically defended animals to strategically
deploy their secretions by emitting (1) only under certain conditions, (2) in an amount
that can vary with circumstances, (3) from the location on the emitter’s body most
proximate to the triggering stimulus (when multiple emission locations are possible), (4)
specifically aimed toward the intended receiver, and (5) in a manner that allows for
recovery or reuptake of emitted material. Each of these deployment strategies of nonvenomous defensive secretions are discussed with examples below.
Selective Deployment of Non-Venomous Defensive Secretions
Many animals deploy defensive secretions only when stimuli indicate a clear and
immediate danger, and often as a last resort after all other defensive options (e.g., fleeing,
thanatosis, retaliatory pinching/biting) have failed (Blum, 1981, 1985; Crabb, 1948;
Duffield et al., 1981; Eisner, 1960; Eisner et al., 1963a; Heiss et al., 2010; Krall et al.,
1999; Machado and Pomini, 2008; Moore and Williams, 1990). This is especially true of
many terrestrial arthropods, including insects (Eisner, 1970; Krall et al., 1999). In fact, as
a general rule, arthropods discharge only in response to direct contact stimulation (Eisner,
1970), although exceptions exist (e.g., Edwards, 1962). Here a number of animals are
353
described briefly that selectively deploy defensive secretions, releasing secretion in some
contexts but not in others.
Millipedes (class Diplopoda) demonstrate selective use of their defensive
secretion. Millipedes are well defended both physically, via their hard cuticle, and
chemically, via defensive glands called ozopores, located on most body segments (Blum,
1981; Whitman et al., 1990). Depending on species, the ozopores discharge a diverse
array of repellent substances, including benzoquinones, cresols, hydrogen cyanide,
benzaldehyde, and alkaloids (Blum, 1981; Whitman et al., 1990). Some millipedes even
forcibly spray their secretions, which are powerfully irritating to potential predators’ eyes
(Eisner et al., 1978). A variety of millipede species roll into a ball or spiral when
molested and this behavior, in combination with an extremely hard, deflective cuticle,
will often provide sufficient protection from attacks by small predators (Blum, 1981).
However, continued harassment of the coiled millipede usually results in the discharge of
the defensive glands (Blum, 1981). For example, when initially disturbed, the millipede
Glomeris marginata (Glomerida: Glomeridae) coiled itself into a ball of cuticular-plated
armor, but if prodded the coiled millipede discharged a secretion from middorsal
glandular pores that entangled and repelled small arthropods (Blum, 1981). Furthermore,
although ants are among the chief natural enemies of the millipede Abacion magnum
(Callipodida: Abacionidae), this millipede preferred death-feigning rather than
discharging its defense secretion when it encountered ants (Eisner et al., 1963a).
Although there may be individual variation in propensity to deploy chemical defenses, A.
magnum usually tolerated considerable prodding and prolonged handing before
discharging; but even the least responsive individuals discharged when the stimulus was
354
more traumatic, such as persistent leg pinching or touching the body with a hot needle
(Eisner et al., 1963a). Once A. magnum had been induced to discharge, subsequent
discharges were more readily induced (e.g., by scratching with a cold needle) without the
strong trauma (e.g., persistent pinching of legs, cautery) required to evoke the first
discharge (Eisner et al., 1963a).
Opiliones (Arachnida: Opiliones), the so-called harvestmen or daddy longlegs,
secrete a variety of ketones and quinones from paired cephalothoracic defensive glands
(Whitman et al., 1990). The secretion is oozed, sprayed, or spread along specialized
integumental grooves, and in some cases may be mixed with oral discharges of enteric
fluid and dabbed onto attackers with the legs (Gnaspini and Hara, 2007; Whitman et al.,
1990). Defensive secretion and enteric fluid can be emitted independently (Gnaspini and
Cavalheiro, 1998; Gnaspini and Hara, 2007), and the relatively expendable enteric fluid,
which has no repellent power of its own (Eisner et al., 1971), is often used to dilute and
distribute the presumably more costly defensive secretions (Gnaspini and Hara, 2007).
The defensive fluid enables harvestmen to deter predators including ants (Duffield et al.,
1981; Eisner et al., 1971), spiders (Gnaspini and Hara, 2007), and lizards (Duffield et al.,
1981). However, release of defensive secretion is a survival strategy of last resort, with
harvestmen first exhibiting evasive behaviors including running in retreat and deathfeigning (Duffield et al., 1981; Gnaspini and Cavalheiro, 1998; Willemart and Gnaspini,
2004). Whether harvestmen deploy their defensive secretion is influenced by location and
persistence of threat stimuli. When the harvestman Camarana flavipalpi was tested under
three increasing levels of threat (seizure of the distal region of a single leg, seizure of the
basal regions of two legs simultaneously, or simultaneous seizure of the dorsum and
355
venter), no animals (0/30) secreted from the defensive glands when subjected to the first
two threat conditions, but over 50% (8/15) of animals released chemical secretions when
subjected to the highest level of threat (Machado and Pomini, 2008). The threshold for
oral discharge of enteric fluid was lower than that for secretion from the defensive glands
in some cases, as several harvestmen (3/30) emitted enteric fluid (but not defensive gland
secretion) when subjected to the first two threat levels (Machado and Pomini, 2008). In
another example of selective use of defensive fluids, the harvestman Acanthopachylus
aculeatus did not respond with chemical defenses when merely prodded or picked up
gently with forceps, but when the body was squeezed or appendages pinched, the animal
deployed its defensive gland secretion mixed with enteric fluid (Eisner et al., 2004).
Similarly, researchers found that the Brazilian cave harvestman Goniosoma albiscriptum
released enteric fluid from the mouth when handled by the legs or illuminated with
headlamps, but if handled by the abdomen or cephalothorax the harvestman would
deploy secretion from its defense glands in addition to enteric fluid (Willemart and
Gnaspini, 2004). In addition to the location of stimuli on the body, the frequency of
disturbance plays a role in triggering release of chemical defenses in harvestmen, with
continuous disturbance increasing the chances of release (Machado and Pomini, 2008).
For example, when harassed, Goniosoma longipes emitted enteric fluid and tried to flee,
but when persistently disturbed the harvestman discharged secretion from the defensive
glands (Machado et al., 2000). Thus, harvestmen clearly demonstrate selective use of
their defensive secretions, showing a higher propensity to discharge when stimulated on
the body (as opposed to appendages) and when repeatedly harassed.
356
Studies of the armored ground cricket Acanthoplus discoidalis (Orthoptera:
Tettigoniidae) by Bateman and Fleming (2009) demonstrated that this insect shows a
degree of selective use of its defensive secretion, hemolymph that is autohemorrhaged
when the insect is threatened. Although some examples of selective secretion highlight
“either-or” cases that contrast consistent and complete lack of defensive secretion use
under some conditions with ubiquitous and prodigious use in other conditions, the case of
autohemorrhaging in A. discoidalis underscores a subtler version of selective use of a
defensive secretion, in which the propensity to deploy an emission varies between
contexts but discharge occurs in both. When A. discoidalis was “attacked” from the side
by investigators using forceps to grab the legs, the insect was less likely to
autohemorrhage (63% [68/108] of attacks) than when grabbed by the pronotum from
above (84% [91/108] of attacks). The authors suggested that the difference in propensity
to autohemorrhage was related to the cricket’s ability to use biting as an alternative
defense. When attacks came from the side the crickets could bite the forceps; however
when attacks came from above the crickets could not physically reach the forceps with
their mouths. An alternative hypothesis is that the crickets perceived attacks on the
pronotum as a higher threat than attacks on the legs, and thus were more likely to use
defensive secretion in a higher threat situation.
Part of the strategic deployment of defensive emissions exhibited by the lubber
grasshopper Romalea guttata (Orthoptera: Romaleidae), studied by Whitman et al.
(1991), includes selectively using the phenolic defensive secretion deployed from its
metathoracic spiracles. In this grasshopper, the tracheal trunks leading to the
metathoracic spiracles are specialized for the storage and discharge of secretion produced
357
in an overlying glandular epithelium. When sufficiently stimulated by a potential
aggressor, the secretion is forcibly ejected as a spray through the metathoracic spiracles
with an audible hiss. Individual variation in propensity to eject secretion was noted,
however investigators found that R. guttata did not discharge its defenses when only
subjected to visual stimulation; the grasshopper did not deploy defensive secretion when
experimenters approached within 30 cm of the grasshopper, or when experimenters
rapidly moved an open hand toward the grasshopper five times in 10 seconds. Rather,
contact stimulation was necessary to elicit secretion discharge. Furthermore, type of
contact stimulus was important in determining use of secretion; over 50% of both sexes
ejected secretion in response to antennal or leg squeezing, however all insects discharged
when squeezed on the anterior abdomen. In addition, females exhibited a lower
disturbance threshold than males; when sharply poked, 65% of females discharged versus
only 20% of males. Copulating grasshoppers sprayed more readily than non-copulating
individuals. Temporal and spatial summation of stimuli occurred. The more times an
insect was subjected to the same stimulus, the more likely it was to discharge. Squeezing
legs and antennae was more likely to elicit discharge than squeezing just the antennae.
Based on the relatively slow replenishment of lost defensive secretion in this species,
Whitman et al. (1992) suggested that lubbers should be under selection pressure to
conserve their defensive secretion and use it only as a last resort, for example, only after
strong tactile stimulation.
Among the Carabidae, several genera of beetles (including Brachinus,
Stenaptinus, and Pheropsophus) are known as “bombardier beetles” (Eisner et al., 2005).
These beetles chemically defend themselves by spraying hot 1,4-benzoquionones, potent
358
irritants, from their abdominal glands at aggressors (Eisner et al., 2005). Several other
taxa of carabids (including the subfamily Paussinae and the tribe Crepidogastrini) have a
similar defense (Eisner and Aneshansley, 1982; Eisner et al., 2001). Experiments by
Eisner (1958) with the bombardier beetle Brachinus ballistarius demonstrated that these
beetles were selective in deployment of their defensive secretion. When beetles’
appendages were pinched, pulled, or touched with a hot needle, the beetles discharged.
However, when appendages were only prodded instead of pulled, the beetles did not
spray. Brachinus ballistarius did spray defensively when the beetle Galerita janus seized
the antenna of the bombardier in its mandibles. In contrast, casual encounters with G.
janus failed to elicit discharges. Thus, spraying in response to mild stimulation was rare.
If the cost of replenishing depleted defensive secretion supplies can be inferred from the
time required for regeneration, then selectively deploying the spray likely represents a
considerable savings for bombardier beetles. Based on a comparison of the number of
discharges from replete glands with the number that could be elicited 11 hours after total
depletion, time to fully replenish expendable stores may be several days (Eisner, 1958).
The carabid beetle Chlaenius cordicollis (subfamily Licininae), studied by Eisner
et al. (1963a), sprays a secretion containing m-methylphenol in defense. The spray
originates from a pair of glands that open submarginally on the hypopygium a short
distance behind the terminal spiracles, and effectively repels ants. Chlaenius cordicollis
strategically deploys its defensive secretion by reserving it until other defenses fail. When
beetles were placed near the entrance of a colony of ants (P. badius), they were quickly
attacked by the ants. During the early stages of an attack, C. cordicollis first relied on its
359
mandibles to bite attacking ants. Only as the ant attack progressed did the beetle abandon
use of its mechanical defense and discharge its chemical defense.
Rove beetles (Coleoptera: Staphylinidae) possess a diversity of abdominal glands
that emit a complex array of defensive compounds, including quinones, hydrocarbons,
lactones, aldehydes, and esters (Whitman et al., 1990). When harassed, these beetles twist
their abdomens and smear defensive exudates against aggressors (Whitman et al., 1990).
The rove beetle Deleaster dichrous, investigated by Dettner et al. (1985), selectively
deployed its defensive secretions, a mixture of the toxin p-toluquinone and an iridodialbased glue (from separate gland systems), depending on the type of aggressor and
intensity of stimulus. When approached by conspecifics, D. dichrous would bend its
abdominal tip dorsally but never secreted. Mere contact with ants (Myrmica) likewise
evoked no secretion. However, when bitten by ants, the beetle smeared secretion onto the
aggressors. Deleaster dichrous immediately smeared defensive secretion when contacted
by Drosophila melanogaster, leading to the death of the flies. Thus, the rove beetle only
secreted in response to strong irritations, and secreted more readily against some
aggressors than others.
The staphylinid beetle Drusilla canaliculata scavenges on dead ants and uses its
tergal gland secretion of alkanes, alkenes, aliphatic aldehydes, and quinones to defend
itself from ants (Brand et al., 1973). However, investigators noted that these beetles only
secreted their defensive fluid as a last resort after other defensive mechanisms had failed
(Blum, 1981, 1985). In fact, Drusilla only utilized its defensive secretion when it was
subjected to sustained molestation by ants (Brand et al., 1973). Researchers contended
360
that the rove beetle’s secretory frugality represented an effective conservation mechanism
(Brand et al., 1973).
Stink bugs (Heteroptera: Pentatomidae) are well known for the odorous volatiles
they emit when molested (Krall et al., 1999). The secretions of these insects typically
contain mixtures of n-alkanes, alkenyl acetates, alkenols, and alkenals with a proven
defensive function against potential predators (Krall et al., 1999). Krall et al. (1999)
investigated the defensive behavior of the stink bug Cosmopepla bimaculata and found
that these animals are selective in their deployment of the defensive secretion they emit
from their paired ventral metathoracic glands. Tests conducted in the field using a
wooden dowel to approach or gently prod the bugs found on foliage resulted only in
evasive behaviors by C. bimaculata, including walking away, hiding, dropping to the
ground, or flying away, with no secretions emitted. However, when the investigators
pinched the bugs on the antennae or legs, or placed the bugs in their mouths and squeezed
them between the tongue and palate, the bugs emitted their defensive secretion. The
threshold for release of secretion was lower when animals were agitated by prior
squeezing. When bugs were calm they secreted only when strongly squeezed on an
appendage or on the body, however, once agitated, bugs would discharge when even
lightly stroked with a fine paintbrush. The authors suggested that in addition to
preventing waste of valuable defensive secretion, there may be another benefit in the bug
not discharging prematurely: secreting in a predator’s mouth might be a more effective
deterrent than secreting in response to mere approach or initial investigative touches.
The nymph of the stonefly Pteronarcys dorsata (Plecoptera: Pteronarcyidae),
studied by Moore and Williams (1990), strategically deployed its autohemorrhaged
361
hemolymph by discharging it at some aggressors but not others. Interactions between the
stonefly nymph and co-existing pelagic (trout) and benthic predators (sculpins, suckers,
and crayfish), as well as with aggressive conspecifics, demonstrated that the nymph used
autohemorrhaging as a defense only when retreat from the predator failed, and almost
exclusively when attacked by crayfish. The nymph typically responded to attacks by
benthic and pelagic fish by freezing and death feigning, with autohemorrhaging never
occurring in response to benthic fish attack (0/144), and only once in 111 trout attacks
(with no effect). In contrast, when attacked by crayfish, all nymphs (44/44) responded by
autohemorrhaging, forcibly expelling hemolymph into the water as a milky cloud from
pores located on the trochanteral segments of the metathoracic pair of legs. Nymphs were
capable of multiple discharges. The defensive discharge was effective in repelling attack,
as crayfish immediately released the autohemorrhaging nymphs and retreated. The
defensive secretion coated crayfish antennae with a viscous film and, rather than being
toxic, the secretion is thought to be repellent based on a cloaking effect on the crayfish’s
antennal sense organs. Because of the nymph’s size and preferred habitat, crayfish likely
pose a greater threat than the other predators tested. For this reason, the authors proposed
that the nymph employs a more costly (in terms of energy investment) defense,
autohemorrhaging, to avoid predation by crayfish.
The European earwig, Forficula auricularia (Dermaptera: Forficulidae),
strategically deploys its benzoquinone-containing defensive secretion by holding the
secretion in reserve and first using its pincers to ward off aggressors. Eisner (1960)
studied this insect and described the conditions in which it sprayed its defensive secretion
from the two pairs of glands situated dorsally in the abdomen, and opening on the
362
posterior margins of the third and fourth abdominal tergites. Experiments indicated that
the earwig used its pincers in precedence to defensive secretion. When pinched with
forceps or touched with a warm needle the earwig would revolve its abdomen and
quickly and accurately bring its pincers to bear on the offending agent. Given that the
pincers are sharp enough to pierce human skin, they function as effective defensive
weapons. Only after continued irritation and persistent attempts at using the pincers failed
to bring relief did the earwig finally discharge its spray. However, when subjected to a
more violent stimulus, such as when the head or abdomen was pinched with hot forceps
or when the animal was seized between the fingers, the earwig sprayed with little or no
delay. Similarly, when exposed to a group of ants (P. badius), with the ants initially
attacking singly or in small groups, earwigs responded by grabbing and removing biting
ants rapidly with the pincers, and no discharges of defensive secretion were observed.
However, when earwigs began to be swarmed by biting and stinging ants, earwigs finally
sprayed their benzoquinones, instantaneously dispersing the attacking ants. Typically in
less than a minute ants returned and reinitiated attacks. Again, the earwigs first defended
themselves with the pincers alone, but eventually discharged another spray of defensive
secretion to repel the ants. The ecological cost of depleted defensive secretion was severe
in these experiments; earwigs with exhausted supplies of benzoquinones were overrun
and killed by the ants. Additionally, the metabolic cost of regenerating the secretion may
be significant, given the regeneration period from depletion was greater than five days.
The walkingstick insect, Anisomorpha buprestoides (Phasmatodea:
Pseudophasmatidae), sprays a lachrymogenous terpene dialdehyde (anisomorphal) from
openings just behind the head when disturbed (Eisner, 1965). The defensive secretion is
363
valuable, with depleted glands requiring up to 2 weeks for replenishment. Anisomorpha
buprestoides strategically deploys its defensive secretion by spraying at most aggressors
only upon physical contact but spraying at birds from a distance before being attacked.
As a rule, A. buprestoides only sprayed when touched, as for example, when tapped,
prodded, or pinched with forceps. Likewise, A. buprestoides sprayed when contacted by
predators including ants, beetles, mice, and a mouse-opossum. However, in most
encounters (15/21) with blue jays, A. buprestoides sprayed the jays as soon as they landed
beside the insects, before the birds ever touched the insects. It wasn’t clear what sensory
modalities the insects used to identify the jays, but assessment was relatively advanced;
attempts to elicit discharges by waving objects in the vicinity of the walkingsticks, or by
tapping the substrate around them, or doing both simultaneously, met with failure.
Some gastropods release mucus in defense against predators, and strategically
deploy their mucus by varying propensity to discharge with degree of threat or by using it
as a last resort. In field and lab interactions between the predatory sea snail Thais
tuberosa (Gastropoda: Muricidae) and its prey, the topshell sea snail Trochus niloticus
(Gastropoda: Trochidae), Castell and Sweatman (1997) found that T. niloticus discharged
mucus significantly more often when in contact with the predator than when located 10
cm away. For example, in lab experiments, only 16% of cultured T. niloticus released
mucus in the presence of the predator, whereas 75% released mucus when in contact with
the predator. Trochus niloticus could release multiple discharges of mucus, although this
occurred less frequently than single discharges. Mucus discharge caused the predator to
turn away and become inactive. Experiments with the lamellose ormer Haliotis
tuberculata (Gastropoda: Haliotidae) by Bancala (2009) showed that this gastropod could
364
differentiate between a predatory starfish (Marthasterias glacialis) and a non-predatory
starfish (Echinaster sepositus), with 75% (15/20) of individuals discharging mucus in
interactions with M. glacialis, but no individuals (0/20) using mucus in interactions with
E. sepositus. In addition, mucus was rarely expelled during the first contact with M.
glacialis and usually many attacks were necessary to stimulate mucus release.
Furthermore, H. tuberculata released mucus as one of the last defensive responses,
typically after performing covering, twisting, and running behaviors. The author
suggested that mucus may be costly to produce and that H. tuberculata may behave so as
to conserve this resource.
Sea hares (Gastropoda: Aplysiidae), including species of the genus Aplysia,
secrete ink and opaline, from separate glands, when attacked by predators (Johnson et al.,
2006). Sea hares strategically deploy their defensive secretions by maintaining high
thresholds for secretion release and by varying their propensity to discharge based on
circumstances. Ink is a purple fluid containing a diversity of molecules including redalgal-derived pigments, amino acids, and protein, and acts against crustacean predators
through phagomimicry and/or sensory disruption (Johnson et al., 2006). Opaline is a
complex, whitish, viscous material that defends sea hares through phagomimicry and/or
sensory disruption, inhibits ingestion, and provides substance to the defensive secretions
(Johnson et al., 2006). When attacked, the sea hare releases ink and opaline into the
mantel cavity and then pumps the secretions out of the siphon toward the attacker
(Johnson et al., 2006). Inking appears to occur at the volition of the sea hare as all
defensive behaviors associated with inking may be elicited without ink release (Nolen
and Johnson, 2001). Deployment of ink is likely costly (Derby, 2007) due to its limited
365
quantity, the several days required to regenerate depleted stores, and the sea hare’s
increased vulnerability to attack if secretions are depleted (Nolen and Johnson, 2001).
Because of these costs, investigators contend that sea hares reserve their chemical
defenses for emergencies by maintaining high thresholds for releasing ink and opaline
(Walters and Erickson, 1986). For example, Aplysia californica did not ink when a
mechanical suction device was applied to its parapodium, but did release its chemical
defense when lifted off the substrate by the vacuum apparatus (Nolen and Johnson,
2001). Furthermore, the probability of inking increased as a function of the area of skin
mechanically stimulated and, in a separate experiment, the amperage of shock applied
(Nolen and Johnson, 2001). Similarly, low intensity shock (10 mA) evoked mantle cavity
pumping in A. californica, but higher intensity shock (40 mA) was necessary to elicit ink
release (Walters and Erickson, 1986). Duration of stimulus is also important in eliciting
inking; propensity to ink increased the longer the duration of shock stimulus applied to a
sea hare’s head (Shapiro et al., 1979). In addition to high thresholds for chemical
discharge, propensity to ink may vary with circumstances. For example, the threshold
amount of stimulation necessary to cause inking increased for sea hares with low supplies
of ink (Nolen and Johnson, 2001). Propensity to ink was also influenced by the amount of
ambient tactile stimulus in the environment; sea hares in calm water environments
demonstrated a higher incidence of inking in response to being poked by a pin than sea
hares in rough water environments (Carew and Kupfermann, 1974). Furthermore, sea
hares in the field were more likely to ink when in a group than alone (Tobach et al.,
1965), and more likely to ink when burrowed than not in a burrow (Aspey and
Blankenship, 1976). Finally, a noxious stimulus triggering ink release reduced the
366
threshold for subsequent ink secretion (Illich et al., 1994). Additional strategic
deployment of the sea hare’s chemical defenses is seen in the way the sea hare varies
which gland(s) respond based on the type of predator attacking. Ink and opaline glands
are under separate neural control and the sea hare can release the secretions together
(most common) or independently (Derby, 2007; Illich et al., 1994; Walters and Erickson,
1986). Experiments with A. californica showed that the sea hare differentiated between
predators, using both ink and opaline in defense against spiny lobsters (Panulirus
interruptus) but only ink against a predatory sea anemone (Anthopleura sola; Derby,
2007).
Some lizards of the genus Phrynosoma (Squamata: Phrynosomatidae), including
the Texas horned lizard P. cornutum, expel a stream of blood from blood sinuses around
their eyes as an antipredator defense (Sherbrooke and Middendorf, 2004). Cost of the
blood-squirting defense can vary with the number of squirts, but was estimated to be high
in artificial predatory trials in which as much as 53% of total body blood was squirted
over the course of seven days, with an average of 27 squirts per individual per day
(Sherbrooke and Middendorf, 2001). Phrynosoma cornutum strategically deploys
squirted blood by selectively utilizing this defense against canid predators (Sherbrooke
and Middendorf, 2004). For example, 22 of 28 horned lizards engorged blood sinuses or
squirted blood before, or within 5 seconds, of contact by a Kit Fox (Sherbrooke and
Middendorf, 2004). Squirted blood negatively affects oral receptors of canid predators,
reducing attacks and likely increasing survival of the lizards (Sherbrooke and Mason,
2005). In contrast, P. cornutum did not squirt blood but rather relied on postural defenses
and counter attacks when subjected to attacks by other potential and known predators
367
including Greater Roadrunners (Geococcyx californianus), Southern Grasshopper Mice
(Onychomys torridus), and a number of snakes (Sherbrooke, 1990, 1991; Sherbrooke and
Middendorf, 2004). Sherbrooke and Middendorf (2004) suggested that P. cornutum
visually identifies and categorizes its potential predators, reserving blood squirting for
defense against canid attacks.
The Australian genus of geckos, Diplodactylus (Squamata: Gekkonidae), includes
a distinctive subgenus Strophurus, members of which possess caudal glands that release a
sticky, noxious defensive secretion (Rosenberg and Russell, 1980; Savitzky et al., 2012).
These tail-squirting geckos are selective in their discharge of defensive secretion, having
a high threshold for release. The caudal gland secretion has not been characterized
chemically, but some species in the subgenus feed on termites and beetles, which
constitute a potential source of toxins that the geckos may sequester (Savitzky et al.,
2012). The secretion is thought to reduce palatability to vertebrate predators and perhaps
act as a mechanical defense against large spiders (Rosenberg and Russell, 1980).
Investigators noted that at least one of these geckos, Diplodactylus spinigerus, exhibited a
reluctance to squirt, even after rough handling (Rosenberg and Russell, 1980), with
ejection occurring only after extreme physical provocation (Richardson and Hinchliffe,
1983). However, the gecko did squirt in response to strong pinching and when its neck
was firmly grasped with a pair of forceps and pressure applied repeatedly (Rosenberg and
Russell, 1980).
Skunks, including the spotted skunks of the genus Spilogale and the striped
skunks of the genus Mephitis, spray a pungent thiol-based secretion in defense (Acorn,
1996; Crabb, 1948; Cuyler, 1924; Verts et al., 2001). Skunks strategically deploy their
368
defensive spray in a number of ways, including spraying only as a last resort. Many
authors have noted that skunks are reluctant to use their defensive weapon and avoid
promiscuous discharge (Crabb, 1948; Cuyler, 1924; Verts et al., 2001). As skunks carry
enough secretion for only six sprays (approximately 15 mL total), and complete
regeneration of secretion takes up to 10 days (Acorn, 1996), it is likely advantageous for
skunks to warn possible predators off without expending their valuable spray. Thus,
threatened skunks first attempt escape or engage in hissing, teeth clicking, growling, footstamping, and tail-raising threat postures before resorting to spraying (Acorn, 1996;
Crabb, 1948; Verts et al., 2001). In addition, skunks may have a lower threshold for
spraying when threatened on open ground as opposed to in a whole or under rocks
(Cuyler, 1924).
Upon examination of the circumstances in which these various animals deploy
their defensive secretions, several similarities in response stand out. First, evasive
behaviors (e.g., running, hiding, thanatosis, threat displays, pinching, and biting) are
often performed prior to releasing chemical defenses. Second, all animals appear to have
thresholds for release that are related to threat persistence, intensity (e.g., generalized vs.
local, or number of molestations per unit time), location, and type (i.e., species of
predator). Repeated harassment (especially within a short timeframe), threats directed at
the body (as opposed to appendages), and frequently encountered predators were most
likely to elicit discharge of defensive secretions.
Amount of Non-Venomous Defensive Secretion Deployed
As might be expected, the amount of defensive secretion necessary to deter a
potential aggressor varies according to the characteristics of the predator and the intensity
369
of attack (Hetz and Slobodchikoff, 1990; Higginson and Ruxton, 2009). Evidence
indicates that many animals control the amount of defensive secretion deployed
depending on the nature of the aggression aimed against them.
Millipedes vary the amount of defensive secretion discharged based on threat
intensity, persistence, and degree of localization. Millipedes control the amount of
defensive secretion deployed primarily by the number of glands they discharge (Eisner et
al., 1963a). However, as some millipedes can discharge from a single gland many times
in succession, another level of control over amount deployed is afforded by the number
of individual gland emissions (Woodring and Blum, 1965). Orthocricus arboreus
(Spirobolida: Rhinocricidae), studied by Woodring and Blum (1965) varied the amount
of quinoidal secretion deployed based on stimulation intensity, oozing secretion from a
single gland when lightly stimulated, spraying from a single gland (up to 30 cm) when
strongly stimulated, and spraying from several segments simultaneously when very
strongly stimulated. The authors speculated that oozing was likely sufficient to repel an
ant, whereas a spray would be required to repel a vertebrate predator. Investigators noted
that complex neuromuscular regulation is required for the millipede to vary the amount of
secretion, number of glands recruited, and forcefulness of ejection (ooze vs. spray). The
millipede A. magnum demonstrated that the amount of defensive secretion deployed is
also related to the persistence of stimulation; whereas initial targeted stimulation evoked
discharge from only the ozopores of the stimulated diplosegment, continued stimulation
at the same locus led to discharge of secretion from several adjacent segments on both
sides of the area stimulated (Eisner et al., 1963a). Likewise, in the millipede Narceus
gordanus (Spirobolida: Spirobolidae), a localized stimulus caused only the nearest glands
370
to emit benzoquinones, but after persistent or generalized disturbance the glands
discharged in large numbers (Eisner and Meinwald, 1966). Similarly, Blum and
Woodring (1962) reported that Pachydesmus crassicutis (Polydesmida: Xystodesmidae)
secreted benzaldehyde and hydrogen cyanide only from the segments stimulated unless
the millipede was highly irritated at which point the discharge became general. Thus,
millipedes tend to deploy more defensive secretion when harassment is intense,
persistent, and generalized across the whole body.
Several species of harvestmen (Opiliones) are known to exercise considerable
control over the quantity and composition of defensive fluid emitted, with the capacity to
regulate the output of both enteric fluid and glandular defensive secretion (Eisner et al.,
1971; Gnaspini and Hara, 2007; Machado and Pomini, 2008). Glandular secretion can be
released independently from one gland at a time or both simultaneously (Duffield et al.,
1981; Machado and Pomini, 2008), can be released into the enteric fluid in intermittent
pulses, and can be oozed (in which case it mixes with enteric fluid) or sprayed in a pure
form (Gnaspini and Hara, 2007). In some species, elaborate musculature associated with
the glands and gland openings are thought to form the basis of neuromuscular control
over emission of glandular secretion (Gnaspini and Hara, 2007). This considerable level
of control of emissions enables harvestmen to adjust their defensive chemical response to
the intensity of stimulation (Eisner et al., 1971; Machado and Pomini, 2008). For
example, studies of Vonones sayi (Opiliones: Cosmetidae) by Eisner et al. (1971)
demonstrated that localized stimulation, as when individual legs were pinched with
forceps, caused only limited emission of defensive fluid, usually less fluid than could be
measured as weight loss (<0.03 µL). In contrast, when the body was persistently held,
371
prodded, and tapped with forceps, V. sayi responded with massive, enterically depleting
discharges, emitting up to a total of 1.65 µL of fluid. Presumably, harvestmen use their
defensive fluids judiciously so that adequate levels of these materials are maintained for
future encounters with potential predators (Machado and Pomini, 2008).
Members of the family Thelyphonidae (Arachnida: Thelyphonida), commonly
called whipscorpions, uropygids, or vinegarroons, spray an acetic-acid-based defensive
secretion in response to physical disturbance from a pair of glands opening on the knoblike postabdomen that forms the stalk of the “whip” (Eisner et al., 1961). The spray is an
effective defense against both arthropod and vertebrate predators (Eisner et al., 1961).
The whipscorpion Mastigoproctus giganteus, studied by Eisner et al. (1961), is capable
of multiple sequential sprays and thus can control the amount of defensive spray released
by the number of discharges. When a leg was pulled repeatedly with forceps, M.
giganteus sprayed repeatedly at intervals of several seconds. Adults with presumably
replete glands sprayed up to 19 consecutive times when harassed. The whipscorpion also
sprayed multiple times when defending itself from ants, solpugids, and grasshopper mice,
depending on persistence of attack. Nearly a full day after having its glands depleted, M.
giganteus only discharged 2–4 times, suggesting regeneration of this valuable defensive
secretion may take between several days to over a week.
Another example of controlling the amount of defensive secretion deployed can
be seen in the autohemorrhaging behavior of the armored ground cricket A. discoidalis
studied by Bateman and Fleming (2009). Upon simulated attack from the side (legs
gripped with forceps by investigators), the crickets released, on average, less hemolymph
(13 mg, n = 68) than when attacked from the top (pronotum gripped with forceps from
372
above; 19 mg, n = 91). The authors noted that the ability to control how much
hemolymph is released based on different predatory approaches indicated that
autohemorrhaging is a carefully regulated defense response under the cricket’s central
control.
The lubber grasshopper R. guttata, studied by Whitman et al. (1991), varies the
amount of phenolic defensive secretion deployed from its metathoracic spiracles both by
number and size of discharges in response to variation in stimulation. When squeezed on
the leg or antenna, discharge from the grasshopper was often low in force and volume. In
contrast, when squeezed on the anterior abdomen, the grasshopper discharged large
amounts of secretion. Multiple mild stimulations could elicit up to 30 consecutive, small
discharges. Conversely, when R. guttata was strongly squeezed on the thorax it would
emit four or five massive discharges. The metathoracic tracheal gland is devoid of
muscles and cannot forcefully discharge secretion by itself; rather, ejection is effected by
pneumatic and hemostatic pressure. Although ejection of secretion occurs due to active
neuromuscular abdominal contraction, some portion of ejection force may be generated
from external pressure to the abdomen or thorax when the grasshopper’s body is
squeezed forcefully by an aggressor.
Bombardier beetles spray hot benzoquinones in defense, and can control the
amount of spray deployed by varying the number of sprays ejected. Persistence of threat
stimuli is an important influence on the number of discharges beetles emit. In
experiments with B. ballistarius, Eisner (1958) found that the beetle discharged only once
when briefly pinched, but when an appendage was persistently pulled the beetle released
up to four discharges in quick succession. A single beetle could eject up to 29 separate
373
discharges. The capacity for multiple discharges has also been demonstrated in
bombardiers of the Crepidogastrini tribe (Eisner et al., 2001) and the Paussinae subfamily
(Eisner and Aneshansley, 1982; Eisner et al., 2000). Eisner (1958) noted that the ability
to discharge repeatedly and in rapid succession made the bombardier’s defense a refined
weapon. Some bombardier beetles may be able to control the amount of secretion in a
given discharge by varying the length of the discharge event. When brachinines
discharge, the spray is ejected as a series of very rapid pulses reflecting the discontinuity
of the chemical events in the beetle’s reaction chamber and triggered by the periodic
infusion of the reactants into the chamber (Dean et al., 1990; Eisner et al., 2001).
Duration of a discharge is likely determined by the quantity of reactants fed into the
reaction chamber, which is controlled by neuromuscular action (Dean et al., 1990; Eisner
et al., 2000). Experiments with Stenaptinus insignis showed wide ranges in discharge
duration (2.6–24.1 ms) and number of pulses per discharge (2–12) from similarly
stimulated beetles (pinch of left foreleg with forceps; Dean et al., 1990), suggesting
beetles may be able to vary spray duration. Dean et al. (1990) hypothesized that beetles
adjust the length of the discharge pulse train to the magnitude and duration of an attack,
although this was not explicitly tested. When attacked by the wolf spider Lycosa
ceratiola, the bombardier Pheropsophus aequinoctialis’s defensive discharge was shorter
(43 ms on average) than the mean time to release by the spider (58 ms), leading Eisner et
al. (2006) to note that the beetle emitted secretion long enough to secure its release, but
for no longer than necessary. A comparison with P. aequinoctialis’s discharge duration in
response to a sustained simulated attack would have been informative in determining
374
whether the beetle varied the duration, and thus amount, of spray based on duration of
attack.
Several species of carabid beetle, including Galerita lecontei, G. janus, Platynus
brunneomarginatus, P. ovipennis, and Calathus ruficollis, eject a toxic, formic-acidcontaining spray in defense (Attygalle et al., 1992; Eisner, 1970; Rossini et al., 1997;
Will et al., 2010), and control the amount of defensive secretion released by the number
of sprays ejected. The number of sprays a beetle releases is related to the persistence of
threat stimuli. These beetles have a pair of glands that open marginally near the
abdominal tip to the sides of the anus (Eisner, 1970; Rossini et al., 1997). Compressor
muscles thickly envelop the sac in which the secretion is stored, allowing the beetles to
forcefully eject the defensive secretion as a spray (Rossini et al., 1997). When subjected
to a pinch of a leg or antenna with forceps, the beetles responded by discharging their
spray toward the offending stimulus (Eisner, 1970; Rossini et al., 1997; Will et al., 2010).
When subjected to repeated pinching, beetles ejected up to 9 consecutive discharges
before chemical defenses were depleted, although 4–6 discharges were more common
(Rossini et al., 1997; Will et al., 2010). The amount of fluid a beetle released when
induced to spray by a repeated standardized stimulus tended to decrease with each
subsequent spray event, with the quantity of fluid in the gland reservoir apparently an
important factor in determining the quantity of fluid sprayed (Will et al., 2010). Even so,
an increase in spray quantity between the first and second spray events in more than half
the beetles tested suggested the possibility that beetles could vary the muscular effort
driving spray ejection (Will et al., 2010). The average quantity of formic acid released in
a single spray was less than the amount lethal to ants, but the total quantity from multiple
375
sprays from a single individual was sufficient to kill an ant aggressor (Will et al., 2010).
Will et al. (2010) proposed the possibility that different levels of attack severity might
elicit variation in quantity of fluid sprayed, although this was not tested.
The carabid beetle C. cordicollis, studied by Eisner et al. (1963a), sprays a
defensive secretion containing m-methylphenol in defense. The beetle can control the
amount of spray deployed by the number of sprays released and by whether one or both
glands respond. For example, pinching individual pro- and metathoracic legs in turn
elicited four consecutive discharges from C. cordicollis. Similarly, intermittently
tightening the grasp on a single leg elicited multiple discharges. The beetle responded to
stimulation on one side of the body with ejection of spray from only the gland on the
same side, but when the head was touched on both sides simultaneously, or when the
abdomen was grasped with broad-tipped forceps, the discharge was synchronous from
both glands.
Rove beetles smear a variety of defensive exudates on potential attackers. The
rove beetle D. dichrous, studied by Dettner et al. (1985), controlled the amount of
secretion deployed by the number of secretions performed, and showed the potential to
vary the amount of secretion released in a single discharge. When molesting the beetles
with forceps, the authors noted that the beetles secreted multiple times in response to
multiple stimulations. Additionally, there was a considerable range in the amount of
secretion released in a single discharge, suggesting the beetles may control the amount
released. One specimen with completely filled glands discharged 2.5% of its gland
reservoir 1 (containing the toxin p-toluquinone) and simultaneously 5% of its gland
reservoir 2 (containing an iridodial-based glue), whereas another beetle with partly
376
emptied gland reservoirs exuded 44.6% (gland reservoir 1) and 74.6% (gland reservoir 2)
of its secretions. The authors speculated that the amount of secretion released might
depend on the intensity of the irritating stimulus.
Stink bugs vary the amount of defensive secretion they deploy depending on the
intensity of threat stimulus. Although quantitative measurements were not made, Krall et
al. (1999) reported that the stink bug C. bimaculata could control the volume of secretion
discharged by the amount emitted per gland and by the number of emitting glands. When
agitated bugs were lightly stroked, only one of the paired metathoracic glands would
discharge. However, pressure applied to the head or abdomen elicited discharge of both
glands. Furthermore, when lightly stimulated, the droplet of defensive secretion was
small, but when strongly stimulated (crushing an appendage) the bug emitted large
droplets. Stink bugs are known to possess complex musculature that facilitates secretion
discharge, and C. bimaculata’s control of secretion highlights that discharge is not
passive but under neuromuscular control (Krall et al., 1999).
In addition to selectively deploying its benzoquinone-containing defensive
secretion, the earwig F. auricularia also demonstrated the ability to control the amount of
secretion discharged by the number of sprays released. Eisner (1960) found that the
earwig could discharge up to six consecutive times when repeatedly irritated, although
the amount of secretion ejected decreased progressively as the supply was depleted.
The walkingstick, A. buprestoides (Phasmatodea), sprays a terpene dialdehyde
when disturbed (Eisner, 1965), and can spray multiple times, thereby determining the
amount of secretion released by number of sprays. Up to five consecutive bilateral
discharges were elicited from an adult female by repeatedly pinching with forceps before
377
the glands were depleted. Likewise, when attacked by a mouse-opossum (Marmosa
demararae), the walkingstick sprayed repeatedly. Once depleted, regeneration of the
valuable defensive spray took up to 2 weeks.
The Malayan cockroach Archiblatta hoeveni (Blattodea: Blattidae), studied by
Maschwitz and Tho (1978), sprays a secretion of p-cresol in defense. The cockroach
strategically deploys its secretion by varying the number of emissions with the
persistence of threat. The secretion is produced in a large bilobed gland that opens in the
intersegmental membrane between the 6th and 7th abdominal sternites. The cockroaches
sprayed in response to physical contact, and occasionally to just the approach of a hand.
Cockroaches could spray repeatedly, ejecting up to 20 individual sprays when repeatedly
touched. Furthermore, the mode of spraying was variable; cockroaches could eject either
a few far-reaching drops or a spray of many tiny droplets. However, the authors did not
investigate whether variation in stimulation influenced mode of spraying.
The assassin bug Platymeris rhadamanthus (Hemiptera: Reduviidae), investigated
by Edwards (1961, 1962), spits its salivary gland secretion containing proteases,
hyaluronidase, and phospholipase in defense, and controls the amount of secretion
emitted by varying the number of spits ejected according to intensity of threat. When
injected into arthropod prey the salivary gland secretion functions as a venom, however
when spat in defense against vertebrates, the same secretion functions as a toxungen
(Nelsen et al., 2013), causing intense local pain, vasodilation, and edema when contacting
eyes or nose. The defensive spit is thought to provide protection against a number of
vertebrate predators, including reptiles, birds, and particularly monkeys. Intensity of
stimulation was important in determining the amount of spit P. rhadamanthus released;
378
when the assassin bug was briefly stimulated, it only discharged one or two spits,
however, when highly irritated, the animal delivered as many as 15 successive spits at a
rate of 3–5 spits/second. Up to 0.17 mg and 2 mg of saliva were discharged in a single
spit and a single series of spits, respectively.
Many caterpillars of the family Notodontidae (Lepidoptera), including
Heterocampa manteo, Schizura unicornis, S. badia, and S. concinna, discharge a formicacid-based spray when disturbed, which is an effective deterrent to birds, lizards, toads,
and spiders (Attygalle et al., 1993; Eisner et al., 1972; Weatherston et al., 1979). The
spray originates from the cervical gland that opens ventrally in the neck region just
behind the head (Attygalle et al., 1993). These caterpillars can eject multiple sprays and
vary the number of sprays based on persistence of threat. Caterpillars freshly taken from
the field responded to repeated pinches to the body with fine forceps with between 3 and
10 consecutive sprays (Attygalle et al., 1993; Eisner et al., 1972).
Caterpillars of the family Papilionidae (Lepidoptera) possess a defensive gland
called the osmeterium that is situated middorsally just behind the head, and the extent of
its use is determined by the intensity of stimulus (Eisner and Meinwald, 1965). Eisner
and Meinwald (1965) studied the use of this gland and its secretion in the caterpillar of
the swallowtail butterfly, Papilio machaon. The osmeterium is an eversible, two-pronged
invagination of the neck membrane that is ordinarily tucked away invisibly beneath the
integument, but can be forcibly everted when the caterpillar is disturbed. When everted,
the two “horns” are covered with an odorous secretion composed primarily of isobutyric
and 2-methylbutyric acids. After the disturbance subsides, the caterpillar retracts the
horns. In laboratory experiments, the caterpillars exercised control over the way the gland
379
was employed, minimizing evaporative loss of defensive secretion by extruding only as
much of the osmeterium as was warranted by the conditions of the attack (Eisner, 1970).
Thus, for example, mild disturbance such as poking the animal gently with a blunt probe
elicited no more than incipient evagination (Eisner and Meinwald, 1965). In contrast,
more complete or even total eversion was only elicited when considerable trauma was
applied, such as pinching the caterpillar’s body with forceps. When attacked by ants
singly or in small numbers, the caterpillar easily repelled the aggressors; in response to an
ant’s bite, a caterpillar would revolve its front end and wipe its extruded osmeterium
against the attacker and the ant, covered with secretion, would flee.
Soldiers of termites in the subfamily Nasutitermitinae, including the genera
Nasutitermes and Tenuirostritermes, squirt a defensive secretion of volatile terpenoids
dissolved in a matrix of isoprenoids through the long, pointed rostrum (a.k.a. “fontanellar
gun”) in response to direct physical contact (Deligne et al., 1981; Eisner et al., 1976;
Lubin and Montogomery, 1981; Nutting et al., 1974). The sticky, strong-smelling
secretion, produced in the cephalic gland, acts essentially as a glue, physically entangling
arthropod predators while simultaneously functioning as a topical irritant and an alarm
pheromone (Eisner et al., 1976; Lubin and Montogomery, 1981). Nasute soldiers do not
deplete their glue in a single squirt, but rather dole out their secretion across multiple
squirts, keeping some of the gluey irritant in reserve to combat future aggressions (Eisner
et al., 1976; Nutting et al., 1974). Persistence of motion by aggressors is an important
influence on the number of discharges emitted by soldiers. In unstaged encounters with
live ants fixed to glass slides, nasute soldiers usually only fired once at an ant (Nutting et
al., 1974). However, when investigators prodded and excited the ants, individual soldiers
380
fired up to six additional shots onto the ants (Nutting et al., 1974). When multiple shots
were ejected, the sticky threads of the first shots had the greater diameter and length, with
the volume of fluid discharged decreasing as the defensive gland was depleted (Nutting et
al., 1974). In the field, once sprayed and partially immobilized, ants struggling in place
and not moving about did not elicit additional squirts from “surveillant” soldiers;
however, if such ants regained mobility, the soldiers would squirt the ants again (Eisner
et al., 1976). Thus, both direct physical contact with the termite as well as predator
movement were important stimuli in eliciting multiple squirts.
Sea hares secrete ink and opaline when attacked by predators (Johnson et al.,
2006). In addition to maintaining a high and variable threshold for inking, sea hares
strategically deploy their chemical defenses by varying the amount of ink released
according to persistence and intensity of threat stimuli. Sea hares do so by altering the
number of inking episodes, and potentially the amount of ink deployed in a single
episode. Although it was once thought that sea hares released nearly all of their ink in a
single deployment once sufficiently stimulated, abundant evidence now indicates that sea
hares can ink multiple times in close succession (Nolen and Johnson, 2001; Nolen et al.,
1995; Walters and Erickson, 1986). It is estimated that a well-fed sea hare with a full ink
gland can release 4–6 defensive salvos, although up to 12 consecutive inkings were
recorded when Aplysia brasiliana was repeatedly shocked (45mA) on the neck (Nolen
and Johnson, 2001). Multiple ink releases were recorded when sea hares were subjected
to mechanical, electrical, and live predator (anemone) stimuli (Nolen and Johnson, 2001).
Nolen and Johnson (2001) noted the sea hare’s most efficient use of ink would be to
deploy the smallest amount that effectively deterred a predator, and then if the first
381
release of ink was not adequate, additional deployments would still be possible without
having wasted the entire ink supply in the first response. Since sea hares may encounter a
predator several times in rapid succession, controlling ink release through multiple
deployments should increase survival (Nolen and Johnson, 2001). In addition to
controlling ink release by the number of ink deployments, sea hares may control the
amount of ink in a single deployment. In simulated tide pool encounters between the sea
hare A. californica and anemones, Nolen et al. (1995) reported the sea hare released
variable amounts of ink depending on the kind of interaction it had with the anemone;
encountering a single anemone tentacle elicited only a small amount of ink, but large ink
emissions occurred whenever several tentacles grabbed the sea hare and lifted it towards
the predator’s oral disc. Furthermore, the amount of ink released in successive inkings in
response to an electrical stimulus were variable and did not merely decline steadily as
supply was reduced, suggesting control over the amount deployed (Nolen and Johnson,
2001). However, the amount of ink deployed was not related to the area of skin
stimulated; sea hares subjected to a mechanical stimulus (weak suction) over a large area
of skin (113.1 mm2) did not release more ink than animals stimulated over a small area of
skin (12.6 mm2; Nolen and Johnson, 2001). By controlling the amount of ink deployed,
sea hares avoid wasting valuable ink (Nolen and Johnson, 2001).
Hagfish (class Myxini) can release large amounts of slime when harassed (Lim et
al., 2006; Strahan, 1959), and their ability to produce multiple discharges of slime in
response to persistent threat (Strahan, 1959) gives them the capacity to strategically
control the amount of slime released. In the Pacific hagfish, Eptatretus stouti
(Myxiniformes: Myxinidae) there are approximately 150 slime-producing glands spaced
382
in two linear rows along its ventrolateral sides (Downing et al., 1981). Each slime gland
is connected to the epidermal surface by a short duct, and the entire gland is surrounded
by a connective tissue capsule and skeletal muscle fibers (Downing et al., 1981). When
irritated, hagfish forcefully eject slime as coherent jets from the slime glands (Lim et al.,
2006). Hagfish slime is composed of muscins and fine fibers called “slime threads” that
lend the slime strength and cohesion (Fudge et al., 2005). The slime is thought to clog the
gills of gill-breathing predators (Lim et al., 2006). Experiments with the Atlantic hagfish,
Myxine glutinosa (Myxiniformes: Myxinidae), demonstrated these animals did not
discharge their entire supply of slime at once when harassed; rather, repeated severe
irritation (squeezing and chemical irritants) was required to induce the hagfish to eject all
of their slime (Strahan, 1959).
The fire salamander, Salamandra salamandra (Caudata: Salamandridae), sprays a
defensive secretion containing the neurotoxic alkaloid samandarine and related toxins
that irritate mucous membranes, affect the central nervous system, and can cause death
by respiratory paralysis (Brodie and Smatresk, 1990). The secretion contains cholesterol
derivatives that are energetically expensive to produce. Evidence suggests this
salamander may control the amount of defensive secretion deployed by regulating the
number of glands that discharge. The salamander’s spray is generated in the middorsal
granular glands, which are recessed into greatly modified epaxial musculature and
positioned in a double row running from the rear of the head to the tip of the tail. The
aimed spray travels at high velocity and can be evoked in response to simulated predator
attack (prodding or pinching with forceps). When exposed to simulated attack, the
salamanders more frequently responded with spraying from a single gland (65% [15/23]
383
of trials) than from two (18%), three (11%), or six (5%) glands. Unfortunately, although
various levels of stimulation were reportedly applied to the salamanders (prodding with
grass stems, or with dull or sharp probes, and pinching on the legs or body with blunt
forceps), the authors did not analyze the relationship between stimulation level and
number of spraying glands. Further investigations should seek to establish clearly
whether S. salamandra sprays from more glands when stimulated intensely than when
mildly provoked. Even so, S. salamandra has the capacity to control the number of
spraying glands; defensive spraying is known to be a two-stage process in which multiple
glands are pressurized by contraction of the epaxial muscles and then active regulation of
the gland pore controls whether an individual gland will discharge. The authors suggested
that by spraying from only a few glands to repulse predators rather than from a large
number of glands (as in some other salamanders), S. salamandra conserves its energyrich glandular contents.
Skunks spray a thiol-containing secretion in defense from a pair of glands, the
walls of which are composed of circular muscle layers supplied with voluntary nerve
fibers, and which open via ducts near the anus (Cuyler, 1924; Verts et al., 2001). Skunks
can control the left and right glands independently, and can release multiple sprays in a
sequence (Acorn, 1996; Cuyler, 1924). Furthermore, skunks can control whether the
secretion reaches the target as a fine spray or a solid stream; usually the spray is delivered
as a fine spray, but when the skunk is profoundly agitated it may be delivered as a stream
(Cuyler, 1924). Thus, skunks can control the amount of spray expended by number of
ejections, and also likely by the amount per spray. Additional study is necessary to
384
explicitly demonstrate the relationship between intensity of stimulation and the quantity
of spray discharged by skunks.
When considered together, these many examples of strategic deployment of
defensive secretions indicate that the most common method, measured thus far, of
varying the amount of secretion deployed is by the number of discrete ejections.
Furthermore, increasing the amount of defensive secretion deployed (by increasing the
number of ejections) is a common response to a persistent threat stimulus. Varying the
number of defensive glands ejecting secretion is another oft-employed strategy, with
more glands responding to generalized (whole body) than to localized (appendage)
stimulation.
Location of Emission from the Emitter’s Body
In some cases animals may have serially arranged defensive glands along their
bodies, and among the chemically protected arthropods this is not unusual (Eisner et al.,
1981). Even animals not equipped with such a serial arrangement often have paired left
and right glands. Frequently, animals with more than one site of discharge for their
defensive secretion display the ability to discharge from only the gland(s) closest to the
site of attack, and to vary the number of glands responding based on the magnitude and
degree of localization of an attack (Eisner et al., 1981; Eisner and Meinwald, 1966;
Whitman et al., 1990). Limiting discharge to the site of stimulation adds considerably to
the efficiency of defensive chemical weapons (Eisner and Meinwald, 1966).
As noted above, millipedes strategically deploy their defensive secretions by
discharging only under certain conditions, and by controlling the amount of secretion
emitted. However, another method of strategic deployment utilized by millipedes is to
385
discharge their secretions from the body location closest to the site of attack, and to only
employ generalized emission as a response to generalized assault (Blum and Woodring,
1962; Eisner, 1970; Eisner et al., 1978; Eisner et al., 1963a; Eisner and Meinwald, 1966;
Woodring and Blum, 1965). Several groups of millipedes, including chordeumoids,
juoids, spiroboloids, spirostreptoids, polydesmoids, and polyzenoids, are known to
conform to this habit of restricting discharge to the region stimulated (Eisner et al., 1978;
Eisner et al., 1963a), and it appears to be a general rule in millipedes (Eisner et al.,
1963a). As an example, when P. crassicutis was exposed to worker fire ants (Solenopsis
saevissima), the millipede discharged defensive secretions only in the segmental areas
contacted by ants, effectively repelling the attackers (Blum and Woodring, 1962). Eisner
et al. (1978) noted that control over localization of discharge made millipede defenses
especially refined for efficient operation.
The armored ground cricket A. discoidalis autohemorrhages in defense, and the
location from which it bleeds is related to the location at which it is attacked (Bateman
and Fleming, 2009). In simulated attacks on its legs from the side, A. discoidalis tended
to release hemolymph from seams in the connective tissue at the base of the legs nearest
to where they were grasped by the investigator’s tweezers. In contrast, when grasped by
the pronotum from above, the cricket discharged hemolymph from pores under the
pronotum.
Bombardier beetles, which discharge benzoquinones in defense, are found in both
the brachinoid and paussoid lineages of Carabidae (Eisner et al., 2001). In paussoid
bombardiers, the defensive glands open separately at some distance from the abdominal
tip, rather than together on the tip itself as they do in brachinoids (Eisner et al., 2000). As
386
part of the strategic deployment of their defensive secretion, paussoid bombardiers do not
discharge simultaneously from both glands as do the brachinoids, but rather eject their
defensive spray from one gland or the other, depending from which side they are attacked
(Eisner, 1980). Thus, for example, when subjected to leg pinching with a pair of forceps,
Metrius contractus always discharged ipsilaterally to the stimulated leg (Eisner et al.,
2000). Similarly, Goniotropis nicaraguensis only sprayed from the gland of the side of
the appendage (leg or antenna) pinched (Eisner and Aneshansley, 1982).
A number of carabid beetles spray a defensive secretion composed predominantly
of formic acid from a pair of glands emptying on either side of the anus (Eisner, 1970;
Rossini et al., 1997; Will et al., 2010). These beetles strategically deploy their emission
by releasing the spray from only one of the two defensive glands depending on the side of
the beetle that is stimulated (Eisner, 1970; Rossini et al., 1997). For example, when
appendages of G. lecontei or G. janus were pinched with forceps, the beetles always
responded with a unilateral spray from the gland of the side of the appendage stimulated
(Rossini et al., 1997).
Chlaenius cordicollis, a carabid beetle studied by Eisner et al. (1963a), sprays mmethylphenol in defense from paired glands opening on either side of the body behind the
terminal spiracles. The beetle strategically deploys the secretion by matching which
gland(s) respond to the location of the stimulus. Chlaenius cordicollis responded to a
unilateral stimulus (pinching of a leg, or touching one side of the abdomen or head with a
hot needle) with ejection of spray from only the gland on the same side. However, when
the head was touched on both sides simultaneously, or when the abdomen was grasped
with broad-tipped forceps, the discharge was synchronous from both glands.
387
Studying the stink bug C. bimaculata, Krall et al. (1999) found that the bug could
independently secrete defensive secretion from either of the paired metathoracic glands.
When stroked on the right side, for example, the bug only discharged the right gland.
Only when extensively stimulated by pressure applied to the head or abdomen or the
crushing of an appendage would the bug discharge both glands simultaneously.
The walkingstick, A. buprestoides, sprays a terpene dialdehyde when disturbed
(Eisner, 1965), and ejection is from one gland or from both depending on whether the
stimulus is applied unilaterally or bilaterally. Bilateral discharges were elicited by tapping
the dorsal thorax, touching both antennae with a heated probe, and pinching the rear of
the abdomen. In contrast, unilateral discharges were induced by pinching individual legs,
and when stimuli were unilateral, discharge was from the gland on the same side as the
stimulus.
The cockroach Diploptera punctata (Blattodea: Blaberidae) sprays a
benzoquinone secretion from its second pair of abdominal spiracles in defense and it can
fire from either of its defensive glands independently depending on which side of its body
is being attacked (Roth and Eisner, 1962; Whitman et al., 1991). Glandular tissue
surrounding the tracheae leading to the second spiracles produce the benzoquinone
secretion, and the secretion is stored in the tracheal lumen (Whitman et al., 1991). When
disturbed, depending on which side of the roach is stimulated, the spiracular valve on that
side opens and the benzoquinones are ejected as air passes through the tracheal system
(Whitman et al., 1991).
Caterpillars of butterflies of the family Papilionidae possess an eversible
defensive gland, the osmeterium, with two extrudable horns bearing isobutyric and 2-
388
methylbutyric acids (Eisner and Meinwald, 1965). Working with the caterpillar of the
swallowtail butterfly, P. machaon, Eisner and Meinwald (1965) found that the caterpillar
controlled the two horns independently when airing its odorous repellent. If a stimulus
was applied to only one side of the caterpillar’s body, then the horn of that side was
everted further than its partner. Matched eversion of the repellent-bearing horns only
occurred in response to generalized trauma such as rough handling (causing total eversion
of both horns) or a more restricted but balanced bilateral stimulus. The horns are
evaginated by blood pressure and withdrawn by special retractor muscles.
Hagfish release large amounts of slime in defense from approximately 150 slimeproducing glands located along their ventrolateral sides (Downing et al., 1981). When the
hagfish E. stouti was pinched with forceps, it discharged slime from only those glands
near the point of contact, as opposed to global release from all of the glands (Lim et al.,
2006). Whereas a global release of slime might suggest an attempt by the hagfish to
create a protective shroud of slime, the local release suggests an active role in which the
hagfish may be targeting the gills of an attacking fish predator (Lim et al., 2006).
The fire salamander S. salamandra sprays a neurotoxic defensive secretion from a
double row of glands running from the rear of the head to the tip of the tail (Brodie and
Smatresk, 1990). The salamander demonstrated a degree of strategic deployment by
controlling which glands responded to simulated attack. When subjected to prodding and
pinching with forceps, the salamander sprayed from the gland row nearest the stimulus in
62% (26/42) of trials in which only a single gland discharged. When multiple glands
sprayed simultaneously, they were often from widely different parts of the dorsal gland
rows, suggesting complex control over the spraying process.
389
When multiple points of emission of defensive secretions are possible, the trend
observed is an ability to vary the number of defensive glands responding based on
whether trauma is localized or generalized, and when localized, to conserve secretion by
discharging from the gland(s) nearest to the site of attack. If defensive secretions are
costly to produce, matching the location of response to the location of threat should be an
efficiency-increasing behavior favored by natural selection.
Aiming of Defensive Secretions
Many animals can eject defensive secretions as a spray, and often those that spray
possess the ability to aim the spray with a high degree of accuracy (Eisner, 1970;
Matthews and Matthews, 2010). In fact, aiming is the rule rather than the exception
among arthropods that spray (Eisner et al., 1963a), with some characterized as “infallible
marksmen” (Eisner, 1970). The ability to aim the spray is a clear advantage in terms of
conservation of secretion, as it provides for maximum effectiveness (i.e., predator is met
by the full impact of a discharge) with minimum expenditure (Eisner and Meinwald,
1966; Roth and Eisner, 1962). Spray aiming is important in repelling small predators
such as ants that might otherwise be missed, and also advantageous when defending
against larger vertebrate predators which likely attack face-first, exposing their sensitive
eyes, nose, and mouth to the defender’s spray (Roth and Eisner, 1962). Although there
are multiple ways in which aiming is accomplished, directionality of spraying is often
determined by postural adjustments (Eisner, 1970).
When harvestmen are antagonized sufficiently to deploy their chemical defenses,
they may create a chemical shield around their bodies by oozing defensive gland
secretion into streams of orally discharged enteric fluid, they may dab this mixture with
390
the legs onto an aggressor, or some may squirt the undiluted gland exudate as a jet
(Gnaspini and Cavalheiro, 1998; Gnaspini and Hara, 2007). The spray can be fired up to
a distance of 10 cm or more, and can be emitted in any direction, including forward
(Gnaspini and Cavalheiro, 1998). In fact, investigators found that whichever region of the
harvestman’s body was handled, the jet emitted usually reached the observer’s hands
(Gnaspini and Cavalheiro, 1998).
When disturbed, whipscorpions spray an aimed discharge comprised
predominantly of acetic acid from paired openings located on the knob-like postabdomen
(Eisner et al., 1961). When appendages of M. giganteus, the whipscorpion investigated
by Eisner et al. (1961), were pinched with forceps, or dorsal parts of the body touched
with a hot needle, the whipscorpion discharged an aimed spray that, although broadly
dispersed, was always directed with sufficient precision to thoroughly douse the area
stimulated. Aiming was accomplished by revolving the postabdominal knob so the tip
bearing the gland openings pointed toward the stimulus. To spray the prosoma or anterior
appendages, M. giganteus also adjusted the position of the opisthosoma as a whole,
revolving it at its base and bending it sharply upwards. However, due to the inability of
the postabdomen to bend downward and under the opisthosoma, the entire ventral surface
of M. giganteus was inaccessible to the spray. The aimed discharges of M. giganteus
commonly ranged from 20 to 40 cm, with a maximum range of 80 cm. The accurate
ejection of the spray allowed M. giganteus to spray and repel individual attacking ants.
When it comes to spraying defensive secretions, Eisner et al. (1961) considered
whipscorpions among the best marksman.
The nymphs of several species of grasshoppers in the genus Poekilocerus
391
(Orthoptera: Pyrgomorphidae) can forcefully eject an aimed defensive spray from a gland
opening dorsally on the first abdominal tergum (Hingston, 1927; Roth and Eisner, 1962;
Von Euw et al., 1967). For example, when immature Poekilocerus pictus was pinched on
the head, it bent its abdomen dorsally and fired a jet of defensive fluid in the forward
direction (Hingston, 1927). However, when the tip of the abdomen was pinched, P. pictus
sprayed toward the rear (Hingston, 1927). In addition to aiming, P. pictus controlled the
mode of spraying, with some sprays delivered as short jets and others as steady streams
(Hingston, 1927). The aimed spray, which could be discharged multiple times in
succession, was ejected several inches from the nymph (Hingston, 1927). In Poekilocerus
bufonius, the spray, which can be launched up to 60 cm from the grasshopper nymph,
contains poisonous cardenolides derived from the animal’s milkweed diet (Von Euw et
al., 1967).
All species of carabid beetle studied in detail that spray defensive secretions aim
their ejections (Rossini et al., 1997). That is certainly the case with bombardier beetles.
Bombardier beetles spray benzoquinones in defense, and all bombardiers, whether from
the brachinoid or paussoid branch of Carabidae, aim their discharges, accurately targeting
any part of the body that is subjected to assault (Eisner and Aneshansley, 1982; Eisner et
al., 2001; Matthews and Matthews, 2010; Roth and Eisner, 1962). However, the method
of aiming differs between brachinoids and paussoids (Eisner and Aneshansley, 1982). In
the brachinoid bombardiers, the two defensive glands open close together on the tip of
the abdomen and aiming occurs by rotation of the abdominal tip, which projects beyond
the shortened elytra (Eisner and Aneshansley, 1982; Eisner et al., 2001). For example, in
the crepidogastrine bombardiers Crepidogaster ambreana and C. atrata, no matter which
392
leg was pinched with forceps, the beetles directed the tip of the abdomen toward the
stimulus and accurately sprayed that leg (Eisner et al., 2001). In the paussoid
bombardiers, in contrast, the defensive glands open laterally at some distance from the
abdominal tip, the elytra cover the entire abdomen, and aiming of defensive discharges is
accomplished by vertical deflection of the abdomen and selective engagement of elytral
flanges or tracks that serve as launching guides for the ejections (Eisner et al., 2000;
Eisner et al., 2001). For example, when the beetles G. nicaraguensis and Ozaena magna
ejected toward a rear appendage, they did so by depressing the abdominal tip and
spraying obliquely downward (Eisner and Aneshansley, 1982). On the other hand, when
they discharged toward a foreleg, they maintained the abdominal tip appressed against
the elytra and the jet of fluid was bent into an anteriorly directed trajectory by adherence
to the outer curvature of a flange on the elytron (Eisner and Aneshansley, 1982). The
ability of bombardiers to aim their spray increases the efficiency of the defensive
weapon, especially against small targets such as ants that might otherwise be missed if
the discharge was not directed accurately (Eisner, 1958).
Several species of carabid beetles eject a formic-acid-containing spray in defense,
and they all perform accurately aimed ejections (Eisner, 1970; Rossini et al., 1997), thus
maximizing the effectiveness of the spray. For example, G. lecontei and G. janus
responded to pinching of appendages with forceps with precisely targeted sprays to the
area of stimulus (Rossini et al., 1997). Aiming is achieved by flexing of the abdominal tip
(Eisner, 1970; Rossini et al., 1997). Slow-motion video of spraying indicated that the
continuous narrow jet of secretion oscillated in direction through a narrow sweep
393
multiple times during the 116 ms ejection (Rossini et al., 1997). Investigators argued that
the oscillation ensured the spray hit the stimulated appendage.
The carabid beetle C. cordicollis, investigated by Eisner (1963a), can accurately
aim its defensive spray of m-methylphenol. The spray is discharged from a pair of glands
opening near the margins of the hypopygium and just behind the terminal spiracles, and
was ejected a distance of up to 50 cm. At the moment of discharge, a short nozzle
evaginates from each glandular pore and the secretion is expelled as a jet of finely
dispersed spray. The beetle aims by varying the degree of flexion of the abdominal tip;
when anterior legs were stimulated, the beetle bent the tip downward so that the
projecting nozzles pointed forward, but when middle or hind legs were stimulated the
bending was less pronounced such that the nozzles pointed downward. Although C.
cordicollis cannot revolve the abdominal tip upward so as to spray its back, the beetle
accurately sprayed all other parts of the body when gripped by forceps. Eisner (1963a)
considered aiming an adaptive refinement of this beetle’s defensive weapon.
The carabid beetle Calosoma prominens sprays a defensive secretion of
salicylaldehyde from glands opening at the tip of the abdomen when disturbed (Eisner et
al., 1963b). The aimed ejections can travel a distance of up to a foot or more. When
individual appendages were pinched with forceps C. prominens aimed its spray with
accuracy toward the appendage stimulated, with aiming accomplished by revolving the
end of the abdomen so that the glandular openings at its tip were pointed toward the
stimulus. Ability to aim was poorest in the forward direction; unlike the bombardier
beetles in the genus Brachinus, C. prominens was unable to bring the full impact of its
spray to bear on the front of its body.
394
The earwig F. auricularia discharges a benzoquinone-containing secretion from
two pairs of glands on its abdomen when sufficiently threatened. Eisner (1960) noted the
gland contents do not ooze out, but instead are forcibly ejected as a spray that the animal
aims with considerable precision toward the region of the body subjected to stimulation.
The mechanism used for aiming the secretion is linked with the defensive use of the
pincers, since by revolving the abdomen at its base while brining the pincers toward the
stimulus, the gland openings are automatically pointed in the proper direction.
Anisomorpha buprestoides (walkingstick insect), sprays a terpene dialdehyde
when disturbed (Eisner, 1965), and its marksmanship is precise. Whenever an ant or
beetle bit one of the walkingstick’s appendages, A. buprestoides responded with an aimed
discharge that struck and repelled the aggressors. When pinched on any appendage with
forceps, A. buprestoides released an aimed spray that invariably drenched the offending
instrument.
The Malayan cockroach A. hoeveni sprays a secretion of p-cresol in defense
(Maschwitz and Tho, 1978). In addition to the ability to eject multiple sprays, the
cockroach strategically deploys its defensive secretion by aiming it. When touched on
their bodies, legs, or antennae, cockroaches altered their posture accordingly by turning
and lifting their bodies so that the spray ejected from between the 6th and 7th sternites was
aimed toward the offending stimulus. In 30 tests with three animals, the direction sprayed
never deviated more than 45 degrees from the stimulus direction regardless of which part
of the body had been touched. The cockroaches could spray their aimed secretion up to
60 cm.
Not only does the assassin bug P. rhadamanthus, investigated by Edwards (1961,
395
1962), control the amount of saliva deployed in defense, it is also able to spit with
considerable accuracy. When stimulated to spit, P. rhadamanthus raises one side of its
body up while lowering the other, rotates its head, and deflects the penultimate segment
of the rostrum so that the rostral tip is directed over and to one side of the body. Accuracy
in aiming of the saliva toward the source of disturbance is then achieved by deflecting the
terminal segment of the rostrum as each jet of saliva is discharged. Thus, the mobile
penultimate and ultimate rostral segments, together with their well-developed retractor
muscles, are integral in directing the saliva spray, which can travel a distance of up to 30
cm.
Many notodontid caterpillars spray a formic-acid-based discharge when harassed,
and all of these spraying caterpillars aim their spray with accuracy (Attygalle et al., 1993;
Eisner et al., 1972; Weatherston et al., 1979). When prodded or pinched, the caterpillar
raises and revolves the front end of the body so as to bring the cervical gland opening to
face the site stimulated, and discharges on the target (Attygalle et al., 1993; Eisner et al.,
1972). The range of the spray can exceed 20 cm, indicating the considerable force of the
aimed ejections (Attygalle et al., 1993).
Termite soldiers of the subfamily Nasutitermitinae squirt a sticky, toxic defensive
secretion of volatile terpenoids, and can aim the noxious glue with high precision
(Deligne et al., 1981; Eisner et al., 1976). Although blind, the nasute soldiers are thought
to aim their glue by orienting toward aggressors by olfactory or auditory cues (Nutting et
al., 1974). When firing, soldiers sometimes oscillated the head back-and-forth one or
more times; this behavior enhanced the effectiveness of the discharge by throwing one or
more loops in the sticky thread and increasing the area covered (Eisner et al., 1976;
396
Nutting et al., 1974). When soldiers were seized by a leg with forceps, or poked on the
back of the abdomen with a pin, they aimed their snout toward the site stimulated and
discharged (Eisner et al., 1976). Soldiers can aim in nearly any direction, the head
functioning as a veritable turret and capable of rotating even posteriorly (Eisner et al.,
1976). Straight-line range of ejection exceeded the soldier’s length by a factor of three
(Eisner et al., 1976). Nutting et al. (1974) considered the fontanellar gun of the nasute
soldiers to be the apex of sophistication among termite chemical defense mechanisms.
When sea hares deploy ink and opaline in defense, they aim their defensive
chemicals at the attacking stimulus (Derby, 2007; Walters and Erickson, 1986). Walters
and Erickson (1986) have described the process of aiming in the sea hare A. californica.
Ink from the ink gland’s vesicles empties into a duct that opens into the sea hare’s mantle
cavity. The ink is then forced out of the mantle cavity by parapodial and gill contractions
and is directed towards the eliciting stimulus by coordination of the siphon and anterior
parts of the parapodia. In fact, noxious stimulation of any point on the sea hare’s body
triggers adjustments of the diameter and angle of the siphon, and of the posture of the
mantle and parapodia, that determine whether defensive secretions are aimed to the front,
back, or both.
The fire salamander S. salamandra aims the defensive secretions it sprays from a
double row of glands running from the rear of the head to the tip of the tail (Brodie and
Smatresk, 1990). Prior to spraying, the salamanders altered their posture, elevating and
tilting their bodies toward the simulated predatory stimulus (prodding with a probe). In
96% (50/52) of trials, salamanders sprayed toward the simulated attack, often striking the
hand holding the probe. When subjected to simulated attacks to the head, neck, or
397
forelimbs, S. salamandra arched the pelvic region so that spray discharged from the
pelvic region was launched over the head of the salamander.
Some geckos in the genus Diplodactylus squirt a sticky, noxious defensive
secretion from glands on the tail when sufficiently stimulated (Rosenberg and Russell,
1980). In addition to selectively discharging the secretion, the geckos aimed the squirted
secretion by positioning of the tail. When pinched by the neck with forceps, D. spinigerus
elevated the tail and squirted one or two streams of caudal secretion, accurately striking
the forceps and the part of the body being pinched (Rosenberg and Russell, 1980). This
gecko ejected the defensive secretion with a range of up to 50 cm and could squirt
multiple times (Richardson and Hinchliffe, 1983). The gecko also used the tail to smear
ejected secretion on the offending forceps (Rosenberg and Russell, 1980).
Skunks reportedly discharge their defensive spray with great accuracy over a
distance of several meters (Crabb, 1948; Cuyler, 1924). The spray is released from
paired, muscle-encapsulated glands that empty via two papillae on either side of the anus
(Cuyler, 1924; Verts et al., 2001). By contracting the muscles surrounding the glands, the
defensive secretion is ejected in two streams (Cuyler, 1924). Skunks, according to
Cuyler’s (1924) description, can control ejection from the left and right glands
independently, allowing the animals to vary the blending of the two streams based on
distance to the target, and contributing to their aiming ability. The distance of the target
from the skunk determines the distance at which the skunk blends the two streams; the
closer the object the shorter the distance before fusion. Furthermore, when the target is
not directly behind the skunk, the skunk will spray a larger stream from the gland on the
opposite side of the skunk as the target. In addition, by adjusting the position of the hind
398
part of their bodies, skunks can accurately spray straight up as well as forward. Cuyler
(1924) related that he was once accurately sprayed by a skunk at a distance of 20 feet.
The weight of evidence is considerable in support of aiming as a key component
in the strategic deployment of defensive secretions in a wide range of animals. It is
logical that natural selection has favored behaviors that maximize the impact of defensive
secretions on aggressors and thereby increase the probability of survival of the
chemically defended. In general, animals that spray defensive secretions have a wellrefined capacity, akin to proprioception, to sense the position of body parts being
stimulated and to accurately discharge secretions from another part of the body in order
to strike the region under attack. The ability to vary the continuity of the spray (e.g., short
jets vs. steady stream) or oscillate the firing mechanism may contribute to the aimed
spray’s effectiveness.
Reuptake of Defensive Secretions
Defensive secretions are not always lost after discharge. In several taxa, strategic
deployment of defensive secretion includes presenting the secretion in such a way that
the chemical weapon can be reclaimed after danger passes. Presumably, the resulting
conservation of defensive secretion is a result of selection pressures acting to reduce
wasteful expenditure of a valuable resource.
Many leaf beetle larvae (Coleoptera: Chrysomelidae) possess paired eversible
defensive glands on the dorsum of the meso- and metathorax and the first seven
abdominal segments (Wallace and Blum, 1969). When disturbed, the larvae evaginate the
glands by blood pressure, emitting salicylaldehyde droplets from a series of paired
tubercles along its back (Eisner, 1970; Wallace and Blum, 1969). The repellent droplets
399
are aired only for the duration of an attack, and when the disturbance subsides, the larvae
salvage much of the secretion by drawing it back into the glands by muscular retraction
of the gland reservoirs (Eisner, 1970; Garb, 1915; Wallace and Blum, 1969). Larvae are
capable of discharging their odorous secretion numerous times in succession, and vary
the duration of airing of the secretion to match the duration of disturbing stimulation
(Garb, 1915). Conservation of defensive secretion may be an adaptive consequence of the
leaf beetles’ eversible glands (Garb, 1915; Wallace and Blum, 1969), as it has been
assumed that reclaiming the droplets represents an energy savings (Kearsley and
Whitham, 1992). However, data from a study of the leaf beetle Chrysomela confluens
indicated production of salicylaldehyde for defense may be metabolically inexpensive
(Kearsley and Whitham, 1992).
Part of the strategic deployment of defensive secretion in stink bugs includes
reuptake of secretion when danger passes. In C. bimaculata, when the secretion from the
metathoracic glands is discharged it forms a spherical droplet that is held in place by a
cuticular projection at the gland orifice (Krall et al., 1999). When harassed, bugs “aired”
these droplets for several seconds and then drew the fluid back into their bodies.
However, if further harassed, bugs touched a leg to the droplet, causing it to spread and
wet the cuticle of the leg and thorax, providing a protective film of secretion.
The eversible defensive gland, the osmeterium, of papilionid caterpillars
represents another example of reuptake of defensive secretions. In these caterpillars the
osmeterium is composed of two extrudable horns, which when evaginated air an odorous
repellent of isobutyric and 2-methylbutyric acids (Eisner and Meinwald, 1965). In
addition to matching the extent of evagination of the horns to the intensity of attack
400
stimulation (Eisner and Meinwald, 1965), another aspect of strategically deploying their
defensive secretion is retracting the repellent-covered horns back into the body when a
disturbance subsides. In cases where the horns of the osmeterium are not wiped on an
aggressor but merely aired, having the ability to retract the horns back into the body
while still covered with defensive secretion enables that secretion to be conserved for
future deployment. In these cases, evaporation represents the only loss of the caterpillar’s
chemical repellent. Thus, several invertebrates deploy defensive secretions by exposing
volatile liquid repellents when harassed. Although these secretions may be wiped on
aggressors and thus sacrificed to the cause of survival, apparently they are often potent
enough when simply aired that their recovery for future use has become a part of the
strategic deployment of these animals’ chemical defenses.
Conclusions
Venoms, predatory glues, and non-venomous defensive secretions are all valuable
emissions in terms of their metabolic costs of synthesis, storage, and discharge, and in
terms of the vital roles they play in acquiring food and warding off predators. Because of
their value, natural selection is expected to favor strategic deployment of these emissions,
selecting for behaviors that maximize their effectiveness and minimize their waste.
Considering many examples of how venoms, predatory glues, and non-venomous
defensive secretions are deployed, several trends become clear. First, strategic
deployment of emissions is a result of animals being keen perceivers both of their
external environments and their internal states, with the ability to assess and integrate
information about their prey (type, size, mass, struggle intensity and duration, location,
trajectory, susceptible anatomy), threats (type, intensity/size, location, persistence), and
401
themselves (quantity of emission remaining, hunger). Second, most animals have
relatively high thresholds for discharging emissions, often relying first on mechanical
means to dispatch prey or to combat aggression. High thresholds for discharge also
translate into a preference for escape over chemical retaliation. Third, animals typically
vary the amount of emission deployed according to the demands of a particular situation.
Thus, animals do not usually discharge their entire supply of emission in one large
ejection, but rather may discharge several small ejections, with multiple ejections a
response to continued prey struggle or persistent or generalized threat stimuli. In some
cases, animals vary the amount of emission deployed in a single emitting act in
proportion to prey characteristics (size, type, struggle intensity) or threat level. Fourth,
when multiple emitting locations are possible for defensive secretions (i.e., serially
arranged or paired glands), animals tend to respond with only the gland(s) closest to the
site of attack for localized stimuli, but may discharge all glands in response to
generalized threats. Fifth, when emissions can be projected forcefully over a distance
(i.e., sprayed or squirted), they are typically aimed with considerable accuracy toward
prey or offending stimuli. Lastly, in a few invertebrates, eversible glands or cuticular
projections are used to present defensive secretions in a manner that allows for
reclaiming the valuable resources back into the animals’ bodies after a threat subsides.
Given the variety of animals employing emissions and the importance of the emissions to
survival, perhaps its not surprising to find that natural selection has shaped several
behaviors contributing to the strategic deployment of emissions.
402
References
Abdel-Aal, A., Abdel-Baset, A., 2010. Venom yield and toxicities of six Egyptian snakes
with a description of a procedure for estimating the amount of venom ejected by a
single snake bite. Sci. J. King Faisal Univ. (Basic and Appl. Sci.) 11, 169-184.
Abdel-Rahman, M.A., Omran, M.A.A., Abdel-Nabi, I.M., Ueda, H., McVean, A., 2009.
Intraspecific variation in the Egyptian scorpion Scorpio maurus palmatus venom
collected from different biotopes. Toxicon 53, 349-359.
Acorn, B., 1996. Agri-facts: practial information for Alberta's agriculture industry.
Biology and control of skunks. Department of agriculture, food and rural
development, Alberta, Canada Agdex 684-5,
http://www1.agric.gov.ab.ca/$department/deptdocs.nsf/all/agdex4663.
Adis, J., 2002. Myriapoda: identification to classes, in: Adis, J. (Ed.), Amazonian
Arachnida and Myriapoda. Pensoft, Sofia [Bulgaria], pp. 457-458.
Akre, R.D., 1982. The social wasps, in: Hermann, H.R. (Ed.), Social Insects. Academic
Press, New York, pp. 1-105.
Alami, M., Ouafik, L., Ceard, B., Legros, C., Bougis, P.E., Martin-Eauclaire, M.F., 2001.
Characterisation of the gene encoding the alpha-toxin Amm V from the scorpion
Androctonus mauretanicus mauretanicus. Toxicon 39, 1579-1585.
Almaaytah, A., Albalas, Q., 2014. Scorpion venom peptides with no disulfide bridges: a
review. Peptides 51, 35-45.
Almeida, F.M., Pimenta, A.M.C., De Figueiredo, S.G., Santoro, M.M., Martin-Eauclaire,
M.F., Diniz, C.R., De Lima, M.E., 2002. Enzymes with gelatinolytic activity can
be found in Tityus bahiensis and Tityus serrulatus venoms. Toxicon 40, 10411045.
Andrietti, F., 2011. The art of managing weapons: the stinging behaviour of solitary
wasps in the eyes of past, present and future research, in: Polidori, C. (Ed.),
Predation in the Hymenoptera: an evolutionary perspective. Transworld Research
Network, Kerala, India, pp. 123-198.
Anonymous. 1985. Dangerous Australians. Bay Books, Sydney.
403
Anthony, C.D., Hickerson, C.A.M., Venesky, M.D., 2007. Responses of juvenile
terrestrial salamanders to introduced (Lithobius forficatus) and native centipedes
(Scolopocryptops sexspinosus). J. Zool. 271, 54-62.
Toxicon 53, 367-374.
Aspey, W.P., Blankenship, J.E., 1976. Aplysia behavioral biology: I. A multivariate
analysis of burrowing in A. brasiliana. Behavioral Biology 17, 279-299.
Atkinson, R.K., 1981. Comparisons of the neurotoxic activity of the venom of several
species of funnel web spiders (Atrax). Aust. J. Exp. Biol. Med. Sci. 59, 307-316.
Atkinson, R.K., Wright, L.G., 1992. The modes of action of spider toxins on insects and
mammals. Comp. Biochem. Physiol. C-Pharmacol. Toxicol. Endocrinol. 102,
339-342.
Attygalle, A.B., Meinwald, J., Eisner, T., 1992. Defense mechanisms of arthropods. 108.
Defensive secretion of a carabid beetle, Helluomorphoides clairvillei. J. Chem.
Ecol. 18, 489-498.
Attygalle, A.B., Smedley, S.R., Meinwald, J., Eisner, T., 1993. Defensive secretion of 2
notodontid caterpillars (Schizura unicornis, S. badia). J. Chem. Ecol. 19, 20892104.
Baer, A., Mayer, G., 2012. Comparative anatomy of slime glands in onychophora (velvet
worms). J. Morphol. 273, 1079-1088.
Baerg, W.J., 1961. Scorpions: biology and effect of their venom. Agricultural Experiment
Station, Division of Agriculture, University of Arkansas, Fayetteville.
Balhara, K.S., Stolbach, A., 2014. Marine envenomations. Emerg. Med. Clin. N. Am. 32,
223-243.
Bancala, F., 2009. Function of mucus secretion by lamellose ormer, Haliotis tuberculata
lamellosa, in response to starfish predation. Anim. Behav. 78, 1189-1194.
404
52.
Baron, A., Diochot, S., Salinas, M., Deval, E., Noel, J., Lingueglia, E., 2013. Venom
toxins in the exploration of molecular, physiological and pathophysiological
functions of acid-sensing ion channels. Toxicon 75, 187-204.
Bateman, P.W., Fleming, P.A., 2009. There will be blood: autohaemorrhage behaviour as
part of the defence repertoire of an insect. J. Zool. 278, 342-348.
Batista, C.V.F., D'Suze, G., Gomez-Lagunas, F., Zamudio, F.Z., Encamacion, S., Sevcik,
C., Possani, L.D., 2006. Proteomic analysis of Tityus discrepans scorpion venom
and amino acid sequence of novel toxins. Proteomics 6, 3718-3727.
Batista, C.V.F., del Pozo, L., Zamudio, F.Z., Contreras, S., Becerril, B., Wanke, E.,
Possani, L.D., 2004. Proteomics of the venom from the Amazonian scorpion
Tityus cambridgei and the role of prolines on mass spectrometry analysis of
toxins. J. Chromatogr. B 803, 55-66.
Batista, C.V.F., Roman-Gonzalez, S.A., Salas-Castillo, S.P., Zamudio, F.Z., GomezLagunas, F., Possani, L.D., 2007. Proteomic analysis of the venom from the
scorpion Tityus stigmurus: biochemical and physiological comparison with other
Tityus species. Comp. Biochem. Physiol. C-Toxicol. Pharmacol. 146, 147-157.
Bauer, A.M., 1990. Gekkonid lizards as prey of invertebrates and predators of
vertebrates. Herpetol. Rev. 21, 83-87.
Beard, R.L., 1971. Production and use of venom by Bracon brevicornis (Wesm.), in: de
Vries, A., Kochva, E. (Eds.), Toxins of animal and plant origin. Gordon &
Breach, New York, pp. 181-190.
Benson, D.A., Karsch-Mizrachi, I., Clark, K., Lipman, D.J., Ostell, J., Sayers, E.W.,
2012. GenBank. Nucleic Acids Res. 40, D48-D53.
Berenbaum, M.R., 1995. The chemistry of defense: theory and practice. Proc. Natl. Acad.
Sci. U.S.A. 92, 2-8.
Berland, L., 1928. Faune de France. Hymenopteres vespiformes. II. (Eumenidae,
Vespidae, Masaridae, Bethylidae, Dryinidae, Embolemidae). Lechevalier, Paris.
Berthe, R.A., de Pury, S., Bleckmann, H., Westhoff, G., 2009. Spitting cobras adjust their
venom distribution to target distance. J. Comp. Physiol. A -Neuroethol. Sens.
Neural Behav. Physiol. 195, 753-757.
405
Berthe, R.A., Westhoff, G., Bleckmann, H., 2013. Potential targets aimed at by spitting
cobras when deterring predators from attacking. J. Comp. Physiol. A -Neuroethol.
Sens. Neural Behav. Physiol. 199, 335-340.
Bettini, S., Brignoli, P.M., 1978. Review of the spider families, with notes on the lesserknown poisonous forms, in: Bettini, S. (Ed.), Arthropod venoms (Handbook of
experimental pharmacology, vol. 48). Springer-Verlag, Berlin, pp. 101-120.
38, 291.
Biass, D., Dutertre, S., Gerbault, A., Menou, J.L., Offord, R., Favreau, P., Stocklin, R.,
2009. Comparative proteomic study of the venom of the piscivorous cone snail
Conus consors. J. Proteomics 72, 210-218.
Blackburn, J., Farrow, M., Arthur, W., 2002. Factors influencing the distribution,
abundance and diversity of geophilomorph and lithobiomorph centipedes. J. Zool.
256, 221-232.
Blaxter, M., Sunnucks, P., 2011. Velvet worms. Curr. Biol. 21, R238-R240.
Blum, M.S., 1978. Biochemical defenses of insects, in: Rockstein, M. (Ed.),
Biochemistry of insects. Academic Press, New York, pp. 465-513.
Blum, M.S., 1981. Chemical defenses of arthropods. Academic Press, New York.
Blum, M.S., 1985. Exocrine systems, in: Blum, M.S. (Ed.), Fundamentals of insect
physiology. Wiley, New York, pp. 535-579.
Blum, M.S., 1992. Ant venoms: chemical and pharmacological properties. J. Toxicol.Toxin Rev. 11, 115-164.
the insecticidal component, in: Pavan, M., Eisner, T. (Eds.), XI. Internationaler
406
Blum, M.S., Hermann, H.R., 1978. Venoms and venom apparatuses of the Formicidae:
Myrmeciinae, Ponerinae, Dorylinae, Pseudomyrmecinae, Myrmicinae, and
Formicinae, in: Bettini, S. (Ed.), Arthropod venoms (Handbook of experimental
pharmacology, vol. 48). Springer-Verlag, Berlin, pp. 801-869.
Blum, M.S., Woodring, J.P., 1962. Secretion of benzaldehyde and hydrogen cyanide by
millipede Pachydesmus crassicutis (Wood). Science 138, 512-513.
Bode, W., Gomisruth, F.X., Stockler, W., 1993. Astacins, serralysins, snake-venom and
matrix metalloproteinases exhibit identical zinc-binding environments
(HEXXHXXGXXH and MET-turn) and topologies and should be grouped into a
common family, the 'metzincins'. FEBS Lett. 331, 134-140.
Boeve, J.L., 1994. Injection of venom into an insect prey by the free hunting spider
Cupiennius salei (Araneae, Ctenidae). J. Zool. 234, 165-175.
Boeve, J.L., Meier, C., 1994. Spider venom and ecoethological implications - a simple
mathematical model. Ecol. Model. 73, 149-157.
Boldrini-Franca, J., Correa-Netto, C., Silva, M.M.S., Rodrigues, R.S., De La Torre, P.,
Perez, A., Soares, A.M., Zingali, R.B., Nogueira, R.A., Rodrigues, V.M., Sanz,
L., Calvete, J.J., 2010. Snake venomics and antivenomics of Crotalus durissus
subspecies from Brazil: assessment of geographic variation and its implication on
snakebite management. J. Proteomics 73, 1758-1776.
Bordas, M.L., 1917. Sur le regime alimentaire de quelques Vespinae (Vesp crabro).
Insecta, Rennes 7, 5-7.
Borges, A., Miranda, R.J., Pascale, J.M., 2012. Scorpionism in Central America, with
special reference to the case of Panama. J. Venom. Anim. Tox. incl. Trop. Dis.
18, 130-143.
Bouchard, N.C., Chan, G.M., Hoffman, R.S., 2004. Vietnamese centipede envenomation.
Vet. Hum. Toxicol. 46, 312-313.
Bowers, M.D., 1992. The evolution of unpalatability and the cost of chemical defense, in:
Roitberg, B.D., Isman, M.B. (Eds.), Insect chemical ecology: an evolutionary
approach. Chapman & Hall, New York, pp. 359.
407
248-254.
Brady, R.M., Baell, J.B., Norton, R.S., 2013. Strategies for the development of
conotoxins as new therapeutic leads. Mar. Drugs 11, 2293-2313.
Brand, J.M., Blum, M.S., Fales, H.M., Pasteels, J.M., 1973. Chemistry of defensive
secretion of beetle, Drusilla canaliculata. J. Insect Physiol. 19, 369-382.
Bridwell, J.C., 1920. Some notes of Hawaiian and other Bethylidae with a description of
a new genus and species. Proc. Hawaii. Entomol. Soc. 4, 291-314.
Brillard-Bourdet, M., Nguyen, V., Ferrer-Di Martino, M., Gauthier, F., Moreau, T., 1998.
Purification and characterization of a new cystatin inhibitor from Taiwan cobra
(Naja naja atra) venom. Biochem. J. 331, 239-244.
Brochetto-Braga, M.R., Palma, M.S., Carmona, E.C., Chaud-Netto, J., Rodrigues, A., da
Silva, G.P., 2006. Influence of the collection methodology on the Apis mellifera
venom composition: peptide analysis. Sociobiology 47, 759-770.
Brodie, E.D., Smatresk, N.J., 1990. The antipredator arsenal of fire salamanders: spraying
of secretions from highly pressurized dorsal skin glands. Herpetologica 46, 1-7.
Brown, J.H., 1973. Toxicology and pharmacology of venoms from poisonous snakes.
Charles C. Thomas, Springfield.
Brown, R.S., Brown, M.B., Bdolah, A., Kochva, E., 1975. Accumulation of some
secretory enzymes in venom glands of Vipera palaestinae. Am. J. Physiol. 229,
1675-1679.
Brunhuber, B.S., 1970. Egg laying, maternal care and development of young in the
scolopendromorph centipede, Cormocephalus anceps anceps Porat. J. Linn. Soc.
Lond. Zool. 49, 225-234.
Brusca, R.C., Brusca, G.J., 2003. Invertebrates. Sinauer Associates, Sunderland, Mass.
Bub, K., Bowerman, R.F., 1979. Prey capture by the scorpion Hadrurus arizonensis
Ewing (Scorpiones, Vaejovidae). J. Arachnol. 7, 243-253.
Bucherl, W., 1946. Acao do veneno dos Escolopendromorfos do Brasil sobre alguns
animais de laboratorio. Mem. Inst. Butantan 19, 181-197.
Bücherl, W., 1953. Dosagem comparada da atividade dos extratos glandularis e do
veneno pro de Phoneutria nigriventer (Keyserling 1891). Mem. Inst. Butantan 25,
1.
408
Bücherl, W., 1971a. Introduction, in: Bücherl, W., Buckley, E.E. (Eds.), Venomous
Bücherl, W., 1971b. Spiders, in: Bücherl, W., Buckley, E.E. (Eds.), Venomous animals
and their venoms: volume III, venomous invertebrates. Academic Press, New
York, pp. 197-277.
Bücherl, W., 1971c. Venomous chilopods or centipedes, in: Bucherl, W., Buckley, E.E.,
169-196.
Budriene, A., Budrys, E., 2005. Effect of prey size and number of prey specimens on
stinging effort in predatory wasps. Acta Zool Litu 15, 330-340.
Cutis 37, 241.
Calvete, J.J., 2011. Proteomic tools against the neglected pathology of snake bite
envenoming. Expert Rev. Proteomics 8, 739-758.
Calvete, J.J., 2013. Snake venomics: from the inventory of toxins to biology. Toxicon 75,
44-62.
Calvete, J.J., Escolano, J., Sanz, L., 2007a. Snake venomics of Bitis species reveals large
intragenus venom toxin composition variation: application to taxonomy of
congeneric taxa. J. Proteome Res. 6, 2732-2745.
Calvete, J.J., Juarez, P., Sanz, L., 2007b. Snake venomics. Strategy and applications. J.
Cantacessi, C., Campbell, B.E., Visser, A., Geldhof, P., Nolan, M.J., Nisbet, A.J.,
Matthews, J.B., Loukas, A., Hofmann, A., Otranto, D., Sternberg, P.W., Gasser,
R.B., 2009. A portrait of the "SCP/TAPS" proteins of eukaryotes—developing a
framework for fundamental research and biotechnological outcomes. Biotechnol.
Adv. 27, 376-388.
Carew, T.J., Kupfermann, I., 1974. Influence of different natural environments on
habituation in Aplysia californica. Behavioral Biology 12, 339-345.
409
Carlin, N.F., David, G.S., 1989. The "bouncer" defense of Odontomachus ruginodis and
other Odontomachine ants (Hymenoptera: Formicidae). Psyche 96, 1-19.
Carpenter, C.C., Gillingham, J.C., 1984. Giant centipede (Scolopendra alternans) attacks
marine toad (Bufo marinus). Caribb. J. Sci. 20, 71-72.
Carregari, V.C., Floriano, R.S., Rodrigues-Simioni, L., Winck, F.V., Baldasso, P.A.,
Ponce-Soto, L.A., Marangoni, S., 2013. Biochemical, pharmacological, and
structural characterization of new basic PLA2 Bbil-TX from Bothriopsis bilineata
snake venom. Biomed Res. Int. DOI: 10.1155/2013/612649.
Carrel, J.E., 1984. Defensive secretion of the pill millipede Glomeric marginata, 1. Fluid
production and storage. J. Chem. Ecol. 10, 41-51.
Casewell, N.R., Wuster, W., Vonk, F.J., Harrison, R.A., Fry, B.G., 2013. Complex
cocktails: the evolutionary novelty of venoms. Trends Ecol. Evol. 28, 219-229.
Casiraghi, M., Martinoli, A., Bosco, T., Preatoni, D.G., Andrietti, F., 2001. Nest
provisioning and stinging pattern in Ammophila sabulosa (Hymenoptera,
Sphecidae): influence of prey size. Ital. J. Zool. 68, 299-303.
Casper, G.S., 1985. Prey capture and stinging behavior in the Emperor Scorpion,
Pandinus imperator (Koch) (Scorpiones, Scorpionidae). J Arachnol 13, 277-283.
Castaneda, O., Harvey, A.L., 2009. Discovery and characterization of cnidarian peptide
toxins that affect neuronal potassium ion channels. Toxicon 54, 1119-1124.
Castell, L.L., Sweatman, H.P.A., 1997. Predator-prey interactions among some intertidal
gastropods on the Great Barrier Reef. J. Zool. 241, 145-159.
Castle, N.A., Haylett, D.G., Jenkinson, D.H., 1989. Toxins in the characterization of
potassium channels. Trends Neurosci. 12, 59-65.
Celerier, M.L., Paris, C., Lange, C., 1993. Venom of an aggressive African
Theraphosidae (Scodra griseipes): milking the venom, a study of its toxicity and
its characterization. Toxicon 31, 577-590.
Cerda, X., Dejean, A., 2011. Predation by ants on arthropods and other animals, in:
Polidori, C. (Ed.), Predation in the Hymenoptera: an evolutionary perspective.
Transworld Research Network, Kerala, India, pp. 39-78.
Chacon, D., Rodriguez, S., Arias, J., Solano, G., Bonilla, F., Gomez, A., 2012.
Maintaining coral snakes (Micrurus nigrocinctus, Serpentes: Elapidae) for venom
production on an alternative fish-based diet. Toxicon 60, 249-253.
410
Chaim, O.M., Trevisan-Silva, D., Chaves-Moreira, D., Wille, A.C.M., Ferrer, V.P.,
Matsubara, F.H., Mangili, O.C., da Silveira, R.B., Gremski, L.H., Gremski, W.,
Senff-Ribeiro, A., Veiga, S.S., 2011. Brown Spider (Loxosceles genus) venom
toxins: tools for biological purposes. Toxins 3, 309-344.
Chao, J., Chang, H., 2006. Variation of the poison duct in Chilopoda centipedes from
Taiwan. Nor. J. Entomol. 53, 139-151.
Chavez-Olortegui, C., Kalapothakis, E., Ferreira, A., Ferreira, A.P., Diniz, C.R., 1997.
Neutralizing capacity of antibodies elicited by a non-toxic protein purified from
the venom of the scorpion Tityus serrulatus. Toxicon 35, 213-221.
Cheung, C.C., Loi, P.K., Sylwester, A.W., Lee, T.D., Tublitz, N.J., 1992. Primary
structure of a cardioactive neuropeptide from the tobacco hawkmoth, Manduca
sexta. FEBS Lett. 313, 165-168.
Chippaux, J.P., Goyffon, M., 2008. Epidemiology of scorpionism: a global appraisal.
Acta Trop. 107, 71-79.
Choo, Y.M., Lee, K.S., Yoon, H.J., Kim, B.Y., Sohn, M.R., Roh, J.Y., Je, Y.H., Kim,
N.J., Kim, I., Woo, S.D., Sohn, H.D., Jin, B.R., 2010. Dual function of a bee
venom serine protease: prophenoloxidase-activating factor in arthropods and
fibrin(ogen)olytic enzyme in mammals. PLoS One 5, e10393. DOI:
10.1371/journal.pone.0010393.
Chow, G.C., 1960. Tests of equality between sets of coefficients in two linear
regressions. Econometrica 28, 591-605.
Chuang, R.S.I., Jaffe, H., Cribbs, L., Perez-Reyes, E., Swartz, K.J., 1998. Inhibition of Ttype voltage-gated calcium channels by a new scorpion toxin. Nat. Neurosci. 1,
668-674.
Cidade, D.A.P., Simao, T.A., Davila, A.M.R., Wagner, G., Junqueira-De-Azevedo,
I.D.M., Ho, P.L., Bon, C., Zingali, R.B., Albano, R.M., 2006. Bothrops jararaca
venom gland transcriptome: analysis of the gene expression pattern. Toxicon 48,
437-461.
Clark, D.B., 1979. Centipede preying on a nestling rice rat (Oryzomys bauri). J. Mammal.
60, 654.
Clausen, C.P., 1940. Entomophagous insects. McGraw-Hill, New York.
411
Clements, R., Li, D.Q., 2005. Regulation and non-toxicity of the spit from the pale
spitting spider Scytodes pallida (Araneae: Scytodidae). Ethology 111, 311-321.
Cloudsley-Thompson, J.L., 1949. The enemies of myriapods. Naturalist (Lond.)137-141.
Press, Oxford.
Hillsdale, NJ.
Costal-Oliveira, F., Duarte, C.G., de Avila, R.A.M., Melo, M.M., Bordon, K.C.F.,
Arantes, E.C., Paredes, N.C., Tintaya, B., Bonilla, C., Bonilla, R.E., Suarez, W.S.,
Yarleque, A., Fernandez, J.M., Kalapothakis, E., Chavez-Olortegui, C., 2012.
General biochemical and immunological characteristics of the venom from
Peruvian scorpion Hadruroides lunatus. Toxicon 60, 934-942.
Crabb, W.D., 1948. The ecology and management of the Prairie Spotted Skunk in Iowa.
Ecol. Monogr. 18, 201-232.
Natn. Mus. 111, 167-195.
412
Cruz, L.S., Vargas, R., Lopes, A.A., 2009. Snakebite envenomation and death in the
developing world. Ethn. Dis. 19, 42-46.
Cumming, W.D., 1903. The food and poison of centipedes. J. Bombay Nat. Hist. Soc. 15,
364-365.
Cushing, B.S., Matherne, A., 1980. Stinger utilization and predation in the scorpion
Paruroctonus boreus. Great Basin Nat. 40, 193-195.
Cuthill, I.C., Houston, A.I., 1997. Managing time and energy, in: Krebs, J.R., Davies,
N.B. (Eds.), Behavioural ecology. Blackwell Publishing, Oxford, pp. 97-120.
Cuyler, W.K., 1924. Observations on the habits of the striped skunk (Mephitis mesomelas
varians). J. Mammal. 5, 180-189.
da Silva, N.J., Aird, S.D., 2001. Prey specificity, comparative lethality and compositional
differences of coral snake venoms. Comp. Biochem. Physiol. C-Toxicol.
Pharmacol. 128, 425-456.
Daltry, J.C., Wuster, W., Thorpe, R.S., 1997. The role of ecology in determining venom
variation in the Malayan pitviper, Calloselasma rhodostoma. Symp. Zool. Soc.
Lond. 70, 155-171.
Daly-Schveitzer, S., Beugnon, G., Lachaud, J.P., 2007. Prey weight and overwhelming
difficulty impact the choice of retrieval strategy in the Neotropical ant
Gnamptogenys sulcata (F. Smith). Insectes Soc. 54, 319-328.
413
de Graaf, D.C., Aerts, M., Danneels, E., Devreese, B., 2009. Bee, wasp and ant venomics
pave the way for a component-resolved diagnosis of sting allergy. J. Proteomics
72, 145-154.
de Graaf, D.C., Brunain, M., Scharlaken, B., Peiren, N., Devreese, B., Ebo, D.G.,
Stevens, W.J., Desjardins, C.A., Werren, J.H., Jacobs, F.J., 2010. Two novel
proteins expressed by the venom glands of Apis mellifera and Nasonia vitripennis
share an ancient C1q-like domain. Insect Mol. Biol. 19, 1-10.
de Jong, P.W., Holloway, G.J., Brakefield, P.M., de Vos, H., 1991. Chemical defence in
ladybird beetles (Coccinellidae). II. Amount of reflex fluid, the alkaloid adaline
and individual variation in defence in 2-spot ladybirds (Adalia bipunctata).
Chemoecology 2, 15-19.
De la Mora, A., Perez-Lachaud, G., Lachaudab, J.P., 2008. Mandible strike: the lethal
weapon of Odontomachus opaciventris against small prey. Behav. Processes 78,
64-75.
de la Vega, R.C.R., Possani, L.D., 2004. Current views on scorpion toxins specific for
K+-channels. Toxicon 43, 865-875.
de Lima, P.R., Brochetto-Braga, M.R., 2003. Hymenoptera venom review focusing on
Apis mellifera. J. Venom. Anim. Toxins 9, http://dx.doi.org/10.1590/S167891992003000200002.
De Lucca, F.L., Haddad, A., Kochva, E., Rothschi.Am, Valeri, V., 1974. Protein
synthesis and morphological changes in secretory epithelium of venom gland of
Crotalus durissus terrificus at different times after manual extraction of venom.
Toxicon 12, 361-368.
de Oliveira, N.G., Cardoso, M., Franco, O.L., 2013. Snake venoms: attractive
antimicrobial proteinaceous compounds for therapeutic purposes. Cell. Mol. Life
Sci. 70, 4645-4658.
De Rezende, N.A., Torres, F.M., Dias, M.B., Campolina, D., Chavez-Olortegui, C.,
Amaral, C.F.S., 1998. South American rattlesnake bite (Crotalus durissus sp)
without envenoming: insights on diagnosis and treatment. Toxicon 36, 20292032.
de Roodt, A.R., Dolab, J.A., Galarce, P.P., Gould, E., Litwin, S., Dokmetjian, J.C., Segre,
L., Vidal, J.C., 1998. A study on the venom yield of venomous snake species
from Argentina. Toxicon 36, 1949-1957.
de Roodt, A.R., Laskowicz, R., Saavedra, S., Vucharchuc, M., Lanari, L.C., Reati, G.,
Beltramino, J., Varni, L., Lopez, R., Bazan, E.E., de Oliveira, V.C., 2012.
414
Comments on the venom yield of Tityus trivittatus, considering two
methodologies of extraction. Toxicon 60, 183.
de Souza, B.M., da Silva, A.V.R., Resende, V.M.F., Arcuri, H.A., Cabrera, M.P.D., Neto,
J.R., Palma, M.S., 2009. Characterization of two novel polyfunctional mastoparan
peptides from the venom of the social wasp Polybia paulista. Peptides 30, 13871395.
Dean, J., Aneshansley, D.J., Edgerton, H.E., Eisner, T., 1990. Defensive spray of the
bombardier beetle: a biological pulse jet. Science 248, 1219-1221.
Dejean, A., 1985. An ecoethological study of predation in the ant genus Smithistruma
(Formicidae, Myrmicinae, Dacetini) 3. Prey capture in S. emarginata. Insectes
Soc. 32, 241-256.
Dejean, A., 1986. Study of the predatory behavior in the genus Strumigenys (Formicidae,
Myrmicinae). Insectes Soc. 33, 388-405.
Dejean, A., 1988. Prey capture by Camponotus maculatus (Formicidae - Formicinae).
Biology of Behaviour 13, 97-115.
Dejean, A., 1990. Prey capture strategy of the African weaver ant, in: Vander Meer, R.K.,
Dejean, A., 2011. Prey capture behavior in an arboreal African ponerine ant. PLoS One 6.
Dejean, A., Bashingwa, E.P., 1985. Predatory behavior in Odontomachus troglodytes
Santschi (Formicidae, Ponerinae). Insectes Soc. 32, 23-42.
Dejean, A., Lachaud, J.P., 2011. The hunting behavior of the African ponerine ant
Pachycondyla pachyderma. Behav. Processes 86, 169-173.
Dejean, A., Suzzoni, J.P., Schatz, B., 2001. Behavioral adaptations of an African
ponerine ant in the capture of millipedes. Behaviour 138, 981-996.
Deligne, J., Quennedey, A., Blum, M., 1981. The enemies and defense mechanisms of
termites, in: Hermann, H.R. (Ed.), Social insects. Academic Press, New York, pp.
1-76.
Boubee, Paris.
Deng, F., Fang, H., Wang, K., 1997. Hemolysis of Scolopendra toxins. J. Chi. Med.
Mater. 20, 36-37.
415
Derby, C.D., 2007. Escape by inking and secreting: Marine molluscs avoid predators
through a rich array of chemicals and mechanisms. Biol. bull. 213, 274-289.
Dettner, K., Schwinger, G., Wunderle, P., 1985. Sticky secretion from two pairs of
defensive glands of rove beetle Deleaster dichrous (Grav) (Coleoptera,
Staphylinidae): gland morphology, chemical constituents, defensive functions,
and chemotaxonomy. J. Chem. Ecol. 11, 859-883.
Devaraja, S., Girish, K.S., Devaraj, V.R., Kemparaju, K., 2010. Factor Xa-like and
fibrin(ogen)olytic activities of a serine protease from Hippasa agelenoides spider
venom gland extract. J. Thromb. Thrombolysis 29, 119-126.
Devaraja, S., Girish, K.S., Gowtham, Y.N.J., Kemparaju, K., 2011. The Hag-protease-II
is a fibrin(ogen)ase from Hippasa agelenoides spider venom gland extract:
purification, characterization and its role in hemostasis. Toxicon 57, 248-258.
Devaraja, S., Nagaraju, S., Mahadeswaraswamy, Y.H., Girish, K.S., Kemparaju, K.,
2008. A low molecular weight serine protease: purification and characterization
from Hippasa agelenoides (funnel web) spider venom gland extract. Toxicon 52,
130-138.
di Tada, I.E., Martori, R.A., Doucet, M.E., Abalos, J.W., 1976. Venom yield with
different milking procedures, in: Rosenberg, P. (Ed.), Toxins: animal, plant and
microbial. Pergamon Press, Oxford, pp. 3-7.
Dias, S.C., Lo-Man-Hung, N.F., 2009. First record of an onychophoran (Onychophora,
Peripatidae) feeding on a theraphosid spider (Araneae, Theraphosidae). J.
Arachnol. 37, 116-117.
Dircksen, H., 1998. Conserved crustacean cardioactive peptide (CCAP) neuronal
networks and functions in arthropod evolution, in: Coast, G.M., Webster, S.G.
(Eds.), Recent advances in arthropod endocrinology. Cambridge University Press,
Cambridge, UK, pp. 302-333.
Djieto-Lordon, C., Orivel, J., Dejean, A., 2001. Predatory behavior of the African
ponerine ant Platythyrea modesta (Hymenoptera : Formicidae). Sociobiology 38,
303-315.
Doley, R., Zhou, X., Kini, R.M., 2010. Snake venom phospholipase A2 enzymes, in:
Mackessy, S.P. (Ed.), Venoms and toxins of reptiles. CRC Press, Boca Raton, pp.
173-205.
dos Santos, L.D., Santos, K.S., Pinto, J.R.A., Dias, N.B., de Souza, B.M., dos Santos,
M.F., Perales, J., Domont, G.B., Castro, F.M., Kalil, J.E., Palma, M.S., 2010.
Profiling the proteome of the venom from the social wasp Polybia paulista: a clue
to understanding the envenoming mechanism. J. Proteome Res. 9, 3867-3877.
416
Downing, S.W., Salo, W.L., Spitzer, R.H., Koch, E.A., 1981. The hagfish slime gland: a
model system for studying the biology of mucus. Science 214, 1143-1145.
Dreyer, F., 1990. Peptide toxins and potassium channels. Rev. Physiol. Biochem.
Pharmacol. 115, 93-136.
Duboscq, O., 1898. Recherches sur les chilopodes. Arch. Zool. Exp. Gen. 6, 481-650.
Duffield, R.M., Olubajo, O., Wheeler, J.W., Shear, W.A., 1981. Alkylphenols in the
defensive secretion of the nearctic opilionid, Stygnomma spinifera (Arachnida,
Opiliones). J. Chem. Ecol. 7, 445-452.
41, 231-243.
Dulcis, D., Levine, R.B., Ewer, J., 2005. Role of the neuropeptide CCAP in Drosophila
cardiac function. J. Neurobiol. 64, 259-274.
Dunn, R.D., Broady, K.W., 2001. Snake inhibitors of phospholipase A2 enzymes.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1533, 29-37.
Easterla, D.A., 1975. Giant desert centipede preys upon snake. Southwest. Nat. 20, 411.
Edgecombe, G.D., 2011. Chilopoda–fossil history, in: Minelli, A. (Ed.), The Myriapoda.
Vol. 1. Brill, Leiden, pp. 355-361.
Edmunds, M.C., Sibly, R.M., 2010. Optimal sting use in the feeding behavior of the
scorpion Hadrurus spadix. J. Arachnol. 38, 123-125.
417
Edwards, J.S., 1961. Action and composition of saliva of an assassin bug Platymesis
rhadamanthus Gaerst (Hemiptera, Reduviidae). J. Exp. Biol. 38, 61-77.
Edwards, J.S., 1962. Spitting as a defensive mechanism in a predatory reduviid.
Proceedings of the 11th International Congress of Entomology, Vienna 1960 3,
259-263.
Edwards, R., 1980. Social wasps: their biology and control. Rentokil Limited, East
Grinstead.
Ehmer, B., Holldobler, B., 1995. Foraging behavior of Odontomachus bauri on Barro
Colorado Island, Panama. Psyche 102, 215-224.
Eisner, T., 1958. The protective role of the spray mechanism of the bombardier beetle,
Brachynus ballistarius Lec. J. Insect Physiol. 2, 215-220.
Eisner, T., 1960. Defense mechanisms of arthropods. II. The chemical and mechanical
weapons of an earwig. Psyche 67, 62-70.
Eisner, T., 1965. Defensive spray of a phasmid insect. Science 148, 966-968.
Eisner, T., 1970. Chemical defense against predation in arthropods, in: Sondheimer, E.,
Simeone, J.B. (Eds.), Chemical ecology. Academic Press, New York, pp. 157217.
Eisner, T., 1980. Chemistry, defense, and survival: case studies and selected topics, in:
Locke, M., Smith, D.S. (Eds.), Insect biology in the future. Academic Press, New
York, pp. 847-878.
Eisner, T., Alsop, D., Hicks, K., Meinwald, J., 1978. Defensive secretions of millipeds,
in: Bettini, S. (Ed.), Arthropod venoms (Handbook of experimental
pharmacology, vol. 48). Springer-Verlag, Berlin, pp. 41-72.
Eisner, T., Aneshansley, D., del Campo, M.L., Eisner, M., Frank, J.H., Deyrup, M., 2006.
Effect of bombardier beetle spray on a wolf spider: repellency and leg autotomy.
Eisner, T., Aneshansley, D.J., 1982. Spray aiming in bombardier beetles: jet deflection by
the coanda effect. Science 215, 83-85.
Eisner, T., Aneshansley, D.J., Eisner, M., Attygalle, A.B., Alsop, D.W., Meinwald, J.,
2000. Spray mechanism of the most primitive bombardier beetle (Metrius
contractus). J. Exp. Biol. 203, 1265-1275.
418
Eisner, T., Aneshansley, D.J., Yack, J., Attygalle, A.B., Eisner, M., 2001. Spray
mechanism of crepidogastrine bombardier beetles (Carabidae; Crepidogastrini).
Eisner, T., Eisner, M., Siegler, M., 2005. Secret weapons : defenses of insects, spiders,
scorpions, and other many-legged creatures. Belknap Press of Harvard University
Press, Cambridge, Mass.
Eisner, T., Hill, D., Goetz, M., Jain, S., Alsop, D., Camazine, S., Meinwald, J., 1981.
Defense mechanisms of arthropods, 67. Antifeedant action of Z-dihydromatricaria
acid from soldier beetles (Chauliognathus spp.). J. Chem. Ecol. 7, 1149-1158.
Eisner, T., Hurst, J.J., Meinwald, J., 1963a. Defense mechanisms of arthropods. XI. The
structure, function, and phenolic secretions of the glands of a chordeumoid
millipede and a carabid beetle. Psyche 70, 94-116.
Eisner, T., Kluge, A.F., Carrel, J.E., Meinwald, J., 1971. Defense of phalangid: liquid
repellent administered by leg dabbing. Science 173, 650-652.
Eisner, T., Kriston, I., Aneshansley, D.J., 1976. Defense mechanisms of arthropods. 47.
Defensive behavior of a termite (Nasutitermes exitiosus). Behav. Ecol. Sociobiol.
1, 83-125.
Eisner, T., Meinwald, J., 1966. Defensive secretions of arthropods. Science 153, 13411350.
Eisner, T., Meinwald, J., Carrel, J.C., Kluge, A.F., 1972. Defense mechanisms of
arthropods, 34. Formic-acid and acyclic ketones in spray of a caterpillar. Ann.
Entomol. Soc. Am. 65, 765-766.
Eisner, T., Meinwald, J., Monro, A., Ghent, R., 1961. Defence mechanisms of
arthropods. 1. The composition and function of the spray of the whipscorpion,
Mastigoproctus giganteus (Lucas) (Arachnida, Pedipalpida). J. Insect Physiol. 6,
272-298.
Eisner, T., Meinwald, Y.C., 1965. Defensive secretion of a caterpillar (Papilio). Science
150, 1733-1735.
Eisner, T., Rossini, C., Gonzalez, A., Eisner, M., 2004. Chemical defense of an opilionid
(Acanthopachylus aculeatus). J. Exp. Biol. 207, 1313-1321.
Eisner, T., Swithenbank, C., Meinwald, J., 1963b. Defense mechanisms of arthropods.
VIII. Secretion of salicylaldehyde by a carbid beetle. Ann. Entomol. Soc. Am. 56,
37-41.
419
Elzinga, R.J., 1994. The use of legs as grasping structures during prey capture and
feeding by the centipede Scolopendra viridis Say (Chilopoda, Scolopendridae). J.
Kans. Entomol. Soc. 67, 369-372.
Enzor, L.A., Wilborn, R.E., Bennett, W.A., 2011. Toxicity and metabolic costs of the
Atlantic stingray (Dasyatis sabina) venom delivery system in relation to its role in
life history. J. Exp. Mar. Biol. Ecol. 409, 235-239.
Escoubas, P., Celerier, M.L., Nakajima, T., 1997. High-performance liquid
chromatography matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry peptide fingerprinting of tarantula venoms in the genus
Brachypelma: chemotaxonomic and biochemical applications. Rapid Commun.
Escoubas, P., Quinton, L., Nicholson, G.M., 2008. Venomics: unravelling the complexity
of animal venoms with mass spectrometry. J. Mass Spectrom. 43, 279-295.
Escoubas, P., Rash, L., 2004. Tarantulas: eight-legged pharmacists and combinatorial
chemists. Toxicon 43, 555-574.
Escoubas, P., Sollod, B., King, G.F., 2006. Venom landscapes: mining the complexity of
spider venoms via a combined cDNA and mass spectrometric approach. Toxicon
47, 650-663.
Espino-Solis, G.P., Riano-Umbarila, L., Becerril, B., Possani, L.D., 2009. Antidotes
against venomous animals: state of the art and prospectives. J. Proteomics 72,
183-199.
Estevez-Lao, T.Y., Boyce, D.S., Honegger, H.W., Hillyer, J.F., 2013. Cardioacceleratory
function of the neurohormone CCAP in the mosquito Anopheles gambiae. J. Exp.
Biol. 216, 601-613.
Evans, H.E., 1966. Behavior patterns of solitary wasps. Annu. Rev. Entomol. 11, 123154.
Evans, H.E., West-Eberhard, M.J., 1970. The wasps. University of Michigan Press, Ann
Arbor.
Fairly, H., Splatt, B., 1929. Venom yields in Australian poisonous snakes. Med. J. Aust.
1, 336-348.
Ferton, C., 1897. Nouvelles observations sur l'instinct des Pompilides. Actes Soc. Linn.
Bordeaux 52, 1-34.
Fescemyer, H.W., Mumma, R.O., 1983. Regeneration and biosynthesis of dytiscid
defensive agents (Coleoptera, Dytiscidae). J. Chem. Ecol. 9, 1449-1464.
420
Fiero, M.K., Bonilla, C.A., Seifert, M.W., Weaver, T.J., 1972. Comparative study of
juvenile and adult prairie rattlesnake (Crotalus viridis viridis) venoms. Toxicon
10, 81-82.
Environ. 407, 6100-6108.
Fink, L.S., 1984. Venom spitting by the Green Lynx spider, Peucetia viridans (Araeae,
Oxyopidae). J. Arachnol. 12, 372-373.
FitzSimmons, V.F.M., 1962. Snakes of southern Africa. Macdonald, London.
Foelix, R.F., 1996. Biology of the spiders. Oxford University Press, New York.
Forrester, J.A., Holstege, C.P., Forrester, J.D., 2012. Fatalities from venomous and
nonvenomous animals in the United States (1999-2007). Wilderness Environ.
Med. 23, 146-152.
Fox, G.A., 2010. [Venom regeneration in the scorpion Hadrurus arizonensis].
Unpublished raw data.
Fox, J.W., Ma, L., Nelson, K., Sherman, N.E., Serrano, S.M.T., 2006. Comparison of
indirect and direct approaches using ion-trap and Fourier transform ion cyclotron
resonance mass spectrometry for exploring viperid venom proteomes. Toxicon
47, 700-714.
Fox, J.W., Serrano, S.M.T., 2007. Approaching the golden age of natural product
pharmaceuticals from venom libraries: an overview of toxins and toxinderivatives currently involved in therapeutic or diagnostic applications. Curr.
Pharm. Des. 13, 2927-2934.
421
Frazao, B., Vasconcelos, V., Antunes, A., 2012. Sea anemone (Cnidaria, Anthozoa,
Actiniaria) toxins: an overview. Mar. Drugs 10, 1812-1851.
Freyvogel, T.A., 1972. Poisonous and venomous animals in East Africa. Acta Trop. 29,
401-451.
Freyvogel, T.A., Honegger, C.G., Maretic, Z., 1968. On the biology and toxicity of the
East African spider Pterinochilus spec. Acta Trop. 25, 217-55.
Frund, H., 1992. The occurence and frequency of scars in centipedes, in: Meyer, E.,
pp. 269-275.
Fry, B.G., Undheim, E.A.B., Ali, S.A., Jackson, T.N.W., Debono, J., Scheib, H., Ruder,
T., Morgenstern, D., Cadwallader, L., Whitehead, D., Nabuurs, R., van der
Weerd, L., Vidal, N., Roelants, K., Hendrikx, I., Gonzalez, S.P., Koludarov, I.,
Jones, A., King, G.F., Antunes, A., Sunagar, K., 2013. Squeezers and leaf-cutters:
differential diversification and degeneration of the venom system in toxicoferan
reptiles. Molecular & Cellular Proteomics 12, 1881-1899.
Fudge, D.S., Levy, N., Chiu, S., Gosline, J.M., 2005. Composition, morphology and
mechanics of hagfish slime. J. Exp. Biol. 208, 4613-4625.
Funasaki, G.Y., Lai, P.-Y., Nakahara, L.M., Beardsley, J.W., Ota, A.K., 1988. A review
of the biological control introductions in Hawaii: 1890-1985. Proc. Hawaii.
Entomol. Soc. 28, 105-160.
Furtado, F.D., Maruyama, M., Kamiguti, A.S., Antonio, L.C., 1991. Comparative study
of 9 Bothrops snake venoms from adult female snakes and their offspring.
Toxicon 29, 219-226.
Galindo, J., Galindo, G., Irwin, J.T., 2009. The effects of temperature on metabolic rate,
venom synthesis and potency in Peucetia viridans (Araneae: Oxyopidae). Integr.
Comp. Biol. 49, E231-E231.
422
Garb, G., 1915. The eversible glands of a Chrysomelid larva, Melasoma lapponica.
Journal of Entomology and Zoology 7, 88.
George, I.D., 1930. Notes on the extraction of venom at the serpentarium of the
Antivenin Institute at Tela, Honduras. Bull. Antivenin Inst. Am. 4, 57-59.
Georgieva, D., Arni, R.K., Betzel, C., 2008. Proteome analysis of snake venom toxins:
pharmacological insights. Expert Rev. Proteomics 5, 787-797.
Gibbons, J.W., Dorcas, M.E., 2002. Defensive behavior of cottonmouths (Agkistrodon
piscivorus) toward humans. Copeia, 195-198.
Gibbs, G.M., Roelants, K., O'Bryan, M.K., 2008. The CAP superfamily: cysteine-rich
secretory proteins, antigen 5, and pathogenesis-related 1 proteins—roles in
reproduction, cancer, and immune defense. Endocr. Rev. 29, 865-897.
Gilbert, C., Rayor, L.S., 1985. Predatory behavior of spitting spiders (Araneae,
Scytodidae) and the evolution of prey wrapping. J. Arachnol. 13, 231-241.
Glenn, J.L., Straight, R.C., 1977. The midget faded rattlesnake (Crotalus viridis
concolor) venom: lethal toxicity and individual variability. Toxicon 15, 129-133.
Gnaspini, P., Cavalheiro, A.J., 1998. Chemical and behavioral defenses of a neotropical
cavernicolous harvestman: Goniosoma spelaeum (Opiliones, Laniatores,
Gonyleptidae). J. Arachnol. 26, 81-90.
Gnaspini, P., Hara, M.R., 2007. Defense mechanisms, in: Pinto-da-Rocha, R., Machado,
G., Giribet, G. (Eds.), Harvestmen: the biology of Opiliones. Harvard University
Press, Cambridge, MA, pp. 374-399.
Gnatzy, W., Otto, D., 1996. Digger wasp vs cricket: Application of the paralytic venom
by the predator and changes in behavioural reactions of the prey after being stung.
Naturwissenschaften 83, 467-470.
Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1982a. Occurrence of
histamine and histamine-release by centipede venom. Indian J. Med. Res. 76, 888891.
Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1982b. Pharmacodynamics of
venom of the centipede Scolopendra subspinipes dehaani Brandt. Indian J. Exp.
Biol. 20, 615-618.
Gomes, A., Datta, A., Sarangi, B., Kar, P.K., Lahiri, S.C., 1983. Isolation, purification
and pharmacodynamics of a toxin from the venom of the centipede Scolopendra
subspinipes dehaani Brandt. Indian J. Exp. Biol. 21, 203-207.
423
Gopalakrishnakone, P., 1992. Light and electron microscopic features of the venom
apparatus of the centipede Scolopendra morsitans. Toxicon 30, 514.
Gregory-Dwyer, V.M., Egen, N.B., Bosisio, A.B., Righetti, P.G., Russell, F.E., 1986. An
isoelectic focusing study of seasonal variation in rattlesnake venom proteins.
Toxicon 24, 995-1000.
Guercio, R.A.P., Shevchenko, A., Lopez-Lozano, J.L., Paba, J., Sousa, M.V., Ricart,
C.A.O., 2006. Ontogenetic variations in the venom proteome of the Amazonian
snake Bothrops atrox. Proteome Sci. 4, DOI: 10.1186/1477-5956-4-11.
Gujarati, D., 1970. Use of dummy variables in testing for equality between sets of
coefficients in 2 linear regressions: a note. Am. Stat. 24, 50-52.
Guo, Y.W., Liu, H.W., Lin, Y.C., Liang, M.C., 2009. Non-parallel expression of a
triflavin-like disintegrin venom protein in the main glands of Trimeresurus
mucrosquamatus. Toxin Rev. 28, 266-270.
Gutierrez, J.M., Lomonte, B., Leon, G., Alape-Giron, A., Flores-Diaz, M., Sanz, L.,
Angulo, Y., Calvete, J.J., 2009. Snake venomics and antivenomics: Proteomic
tools in the design and control of antivenoms for the treatment of snakebite
envenoming. J. Proteomics 72, 165-182.
Gutierrez, J.M., Rucavado, A., Escalante, T., 2010. Snake venom metalloproteinases:
biological roles and participation in the pathophysiology of envenomation, in:
Mackessy, S.P. (Ed.), Handbook of venoms and toxins of reptiles. CRC Press,
Boca Raton, pp. 115-138.
Gutierrez, M.D., Abarca, C., Possani, L.D., 2003. A toxic fraction from Scolopendra
venom increases the basal release of neurotransmitters in the ventral ganglia of
crustaceans. Comp. Biochem. Physiol., C: Toxicol. Pharmacol. 135, 205-214.
424
Haight, K.L., 2006. Defensiveness of the fire ant, Solenopsis invicta, is increased during
colony rafting. Insectes Soc. 53, 32-36.
Haight, K.L., 2012. Patterns of venom production and temporal polyethism in workers of
Jerdon's jumping ant, Harpegnathos saltator. J. Insect Physiol. 58, 1568-1574.
Hamilton, W.J., Jr., Pollack, J.A., 1955. The food of some crotalid snakes from Fort
Benning, Georgia. Chicago Academy of Science and Natural History
Miscellaneous publication No. 140, 1-4.
Harada, S., Yoshizaki, Y., Natsuaki, M., Shimizu, H., Fukuda, H., Nagai, H., Ikeda, T.,
2005. Three cases of centipede allergy—analysis of cross reactivity with bee
allergy. Arerugi 54, 1279-1284.
Hariprasad, G., Srinivasan, A., Singh, R., 2013. Structural and phylogenetic basis for the
classification of group III phospholipase A2. J. Mol. Model. 19, 3779-3791.
Haritos, V.S., Niranjane, A., Weisman, S., Trueman, H.E., Sriskantha, A., Sutherland,
T.D., 2010. Harnessing disorder: onychophorans use highly unstructured proteins,
not silks, for prey capture. Proc. R. Soc. B-Biol. Sci. 277, 3255-3263.
Harvey, A.L., 2006. Snake venom peptides, in: Kastin, A.J. (Ed.), Handbook of
biologically active peptides. Academic Press, Burlington, pp. 355-362.
Harvey, A.L., Robertson, B., 2004. Dendrotoxins: structure-activity relationships and
effects on potassium ion channels. Curr. Med. Chem. 11, 3065-3072.
Harwood, R.F., James, M.T., Herms, W.B., 1979. Entomology in human and animal
health. Macmillan, New York.
Harwood, R.H., 1974. Predatory behavior of Argiope aurantia (Lucas). Am. Midl. Nat.
91, 130-139.
Haspel, G., Libersat, F., 2003. Wasp venom blocks central cholinergic synapses to induce
transient paralysis in cockroach prey. J. Neurobiol. 54, 628-637.
425
Haspel, G., Rosenberg, L.A., Libersat, F., 2003. Direct injection of venom by a predatory
wasp into cockroach brain. J. Neurobiol. 56, 287-292.
Hayden, L., Arthur, W., 2013. Expression patterns of Wnt genes in the venom claws of
centipedes. Evol. Dev. 15, 365-372.
Hayden, L., Parkes, G., Arthur, W., 2012. Segment number, body length, and latitude in
geophilomorph centipedes: a 'converse-Bergmann' pattern. Biol. J. Linn. Soc. 107,
166-174.
Hayes, W.K., 1991. Ontogeny of striking, prey-handling and envenomation behavior of
Prairie Rattlesnakes (Crotalus v. viridis). Toxicon 29, 867-875.
Hayes, W.K., 1992a. Factors associated with the mass of venom expended by prairie
rattlesnakes (Crotalus v. viridis) feeding on mice. Toxicon 30, 449-460.
Hayes, W.K., 1992b. Prey-handling and envenomation strategies of Prairie Rattlesnakes
(Crotalus v. viridis) feeding on mice and sparrows. J. Herpetol. 26, 496-499.
Hayes, W.K., 1993. Effects of hunger on striking, prey-handling, and venom expenditure
of Prairie Rattlesnakes (Crotalus v. viridis). Herpetologica 49, 305-310.
Hayes, W.K., Herbert, S.S., Harrison, J.R., Wiley, K.L., 2008. Spitting versus biting:
differential venom gland contraction regulates venom expenditure in the BlackNecked Spitting Cobra, Naja nigricollis nigricollis. J. Herpetol. 42, 453-460.
Hayes, W.K., Herbert, S.S., Rehling, G.C., Bennaro, J.F., 2002. Factors that influence
venom expenditure in viperids and other snake species during predatory and
defensive contexts, in: Schuett, G.W., Hoggren, M., Douglas, M.E., Greene, H.W.
(Eds.), Biology of the vipers. Eagle Mountain Publishers, Eagle Mountain, pp.
207-233.
Hayes, W.K., Kaiser, I.I., Duvall, D., 1992. The mass of venom expended by prairie
rattlesnakes when feeding on rodent prey, in: Campbell, J.A., Brodie, E.D., Jr.
(Eds.), Biology of the pitvipers. Selva, Tyler, Texas, pp. 383-388.
Hayes, W.K., Lavinmurcio, P., Kardong, K.V., 1995. Northern Pacific Rattlesnakes
(Crotalus viridis oreganus) meter venom when feeding on prey of different sizes.
Copeia, 337-343.
426
Hayes, W.K., Mackessy, S.P., 2010. Sensationalistic journalism and tales of snakebite:
are rattlesnakes rapidly evolving more toxic venom? Wilderness Environ. Med.
21, 35-45.
Heatwole, H., 1967. Defensive behaviour of some Panamanian scorpions. Caribb. J. Sci.
7, 15-17.
Heiss, E., Natchev, N., Salaberger, D., Gumpenberger, M., Rabanser, A., Weisgram, J.,
2010. Hurt yourself to hurt your enemy: new insights on the function of the
bizarre antipredator mechanism in the salamandrid Pleurodeles waltl. J. Zool.
280, 156-162.
Henriksen, A., King, T.P., Mirza, O., Monsalve, R.I., Meno, K., Ipsen, H., Larsen, J.N.,
Gajhede, M., Spangfort, M.D., 2001. Major venom allergen of yellow jackets,
Ves v 5: structural characterization of a pathogenesis-related protein superfamily.
Proteins 45, 438-448.
Hermann, H.R., 1984. Defensive mechanisms in social insects. Praeger, New York.
Herzig, V., Khalife, A.A., Chong, Y., Isbister, G.K., Currie, B.J., Churchill, T.B., Horner,
S., Escoubas, P., Nicholson, G.M., Hodgson, W.C., 2008. Intersexual variations in
Northern (Missulena pruinosa) and Eastern (M. bradleyi) mouse spider venom.
Toxicon 51, 1167-1177.
Herzig, V., Ward, R.J., dos Santos, W.F., 2004. Ontogenetic changes in Phoneutria
nigriventer (Araneae, Ctenidae) spider venom. Toxicon 44, 635-640.
Hetz, M., Slobodchikoff, C.N., 1990. Reproduction and the energy cost of defense in a
Batesian mimicry complex. Oecologia 84, 69-73.
Heymons, R., 1901. Die entwicklungsgeschichte der Scolopender. Zoologica 13, 1-244.
Higginson, A.D., Ruxton, G.D., 2009. Dynamic state-dependent modelling predicts
optimal usage patterns of responsive defences. Oecologia 160, 399-410.
Hill, R.E., Mackessy, S.P., 1997. Venom yields from several species of colubrid snakes
and differential effects of ketamine. Toxicon 35, 671-678.
Hingston, M.R.W.G., 1927. The liquid-squirting habit of oriental grasshoppers.
Transactions of the Royal Entomological Society of London 75, 65-68.
Hoffman, R.L., 1954. Distributional records for two species of Plethodon in the southern
Appalachians. Herpetologica 10, 191-193.
427
241-249.
Holldobler, B., Oldham, N.J., Alpert, G.D., Liebig, J., 2002. Predatory behavior and
chemical communication in two Metapone species (Hymenoptera: Formicidae).
Holldobler, B., Wilson, E.O., 1990. The ants. Belknap Press of Harvard University Press,
Cambridge.
Holloway, G.J., de Jong, P.W., Brakefield, P.M., de Vos, H., 1991. Chemical defence in
ladybird beetles (Coccinellidae). I. Distribution of coccinelline and individual
variation in defence in 7-spot ladybirds (Coccinella septempunctata).
Cemoecology 2, 7-14.
Hopkins, C., Grilley, M., Miller, C., Shon, K.J., Cruz, L.J., Gray, W.R., Dykert, J.,
Rivier, J., Yoshikami, D., Olivera, B.M., 1995. A new family of Conus peptides
targeted to the nicotinic acetylcholine receptor. J. Biol. Chem. 270, 22361-22367.
20, 369-375.
Huang, P., Mackessy, S.P., 2004. Biochemical characterization of phospholipase A2
(trimorphin) from the venom of the Sonoran lyre snake Trimorphodon biscutatus
lambda (family Colubridae). Toxicon 44, 27-36.
Illich, P.A., Joynes, R.L., Walters, E.T., 1994. Response-specific inhibition during
general facilitation of defensive responses in Aplysia. Behav. Neurosci. 108, 614623.
Sci. U.S.A. 100, 922-927.
Isbister, G.K., 2004. Mouse spider bites (Missulena spp.) and their medical importance: a
systematic review. Med. J. Aust. 180, 225-227.
428
Jaffe, K., Marcuse, M., 1983. Nestmate recognition and territorial behavior in the ant
Odontomachus bauri Emery (Formicidae, Ponerinae). Insectes Soc. 30, 466-481.
Janes, D.N., Bush, S.P., Kolluru, G.R., 2010. Large snake size suggests increased
snakebite severity in patients bitten by rattlesnakes in southern California.
Calcutta.
Jangi, B.S., Dass, C.M.S., 1977. Chemoreceptive function of the poison fang in the
centipede Scolopendra morsitans L. Indian J. Exp. Biol. 15, 803-804.
Janvier, H., 1930. Recherches biologiques sur les predateurs du Chili. Ann. Sci. Nat.
Zool. 13.
Jarrar, B.M., 2010. Morphology, histology and histochemistry of the venom apparatus of
the centipede, Scolopendra valida (Chilopoda, Scolopendridae). Int. J. Morphol.
28, 19-25.
Jeanne, R.L., Keeping, M.G., 1995. Venom spraying in Parachartergus colobopterus: a
novel defensive behavior in a social wasp (Hymenoptera, Vespidae). J. Insect
Behav. 8, 433-442.
Jiao, G.-B., Zhu, M.-S., 2009. Prey capture behavior in Heterometrus petersii (Thorell,
1876) (Scorpiones: Scorpionidae). Euscorpius 80, 1-5.
Johnson, P.M., Kicklighter, C.E., Schmidt, M., Kamio, M., Yang, H.C., Elkin, D.,
Michel, W.C., Tai, P.C., Derby, C.D., 2006. Packaging of chemicals in the
defensive secretory glands of the sea hare Aplysia californica. J. Exp. Biol. 209,
78-88.
Jones, K.B., 1988. Influence of prey on the feeding behavior of Trimorphodon biscutatus
Lambda (Colubridae). Southwest. Nat. 33, 488-490.
Junqueira, M., Spirin, V., Balbuena, T.S., Thomas, H., Adzhubei, I., Sunyaev, S.,
Shevchenko, A., 2008. Protein identification pipeline for the homology-driven
proteomics. J. Proteomics 71, 346-356.
429
Kaire, G.H., 1963. Observations on some funnel-web spiders (Atrax species) and their
venoms, with particular reference to Atrax robustus. Med. J. Aust. 2, 307-311.
Kaltsas, D., Stathi, I., Mylonas, M., 2008. The foraging activity of Mesobuthus gibbosus
(Scorpiones : Buthidae) in central and south Aegean archipelago. J. Nat. Hist. 42,
513-527.
Kardong, K.V., 1986a. Predatory strike behavior of the rattlesnake, Crotalus viridis
oreganus. J. Comp. Psychol. 100, 304-314.
Kardong, K.V., 1986b. The predatory strike of the rattlesnake: when things go amiss.
Copeia 1986, 816-820.
Kardong, K.V., Bels, V.L., 1998. Rattlesnake strike behavior: kinematics. J. Exp. Biol.
201, 837-850.
Kearsley, M.J.C., Whitham, T.G., 1992. Guns and butter: a no cost defense against
predation for Chrysomela confluens. Oecologia 92, 556-562.
Keegan, H.L., 1963. Centipedes and millipedes as pests in tropical areas, in: Keegan,
H.L., Macfarlane, W.V. (Eds.), Venomous and poisonous animals and noxious
plants of the Pacific region. Pergamon, London, pp. 161-163.
Kimura, L.F., Prezotto-Neto, J.P., Tavora, B., Antoniazzi, M.M., Knysak, I., Guizze,
S.P.G., Santoro, M.L., Barbaro, K.C., 2013. Local inflammatory reaction induced
by Scolopendra viridicornis centipede venom in mice. Toxicon 76, 239-246.
King, G.F., 2011. Venoms as a platform for human drugs: translating toxins into
therapeutics. Expert Opin. Biol. Ther. 11, 1469-1484.
King, G.F., Hardy, M.C., 2013. Spider-venom peptides: structure, pharmacology, and
potential for control of insect pests, in: Berenbaum, M.R. (Ed.), Annual Review of
Entomology, Vol 58. Annual Reviews, Palo Alto, pp. 475-496.
King, T.P., Spangfort, M.D., 2000. Structure and biology of stinging insect venom
allergens. Int. Arch. Allergy Immunol. 123, 99-106.
Kinkawa, K., Shirai, R., Watanabe, S., Toriba, M., Hayashi, K., Ikeda, K., Inoue, S.,
2010. Up-regulation of the expressions of phospholipase A2 inhibitors in the liver
of a venomous snake by its own venom phospholipase A2. Biochem. Biophys.
Res. Commun. 395, 377-381.
Klauber, L.M., 1997a. Rattlesnakes: their habits, life histories, and influence on mankind.
Volume I. University of California Press, Berkeley.
430
Klauber, L.M., 1997b. Rattlesnakes: their habits, life histories, and influence on mankind.
Trop. Med. Hyg. 9, 381-390.
303-311.
Kordis, D., Turk, V., 2009. Phylogenomic analysis of the cystatin superfamily in
eukaryotes and prokaryotes. BMC Evol. Biol. 9, DOI: 10.1186/1471-2148-9-266.
Kozlov, S.A., Vassilevski, A.A., Feofanov, A.V., Surovoy, A.Y., Karpunin, D.V.,
Grishin, E.V., 2006. Latarcins, antimicrobial and cytolytic peptides from the
venom of the spider Lachesana tarabaevi (Zodariidae) that exemplify
biomolecular diversity. J. Biol. Chem. 281, 20983-20992.
Krall, B.S., Bartelt, R.J., Lewis, C.J., Whitman, D.W., 1999. Chemical defense in the
stink bug Cosmopepla bimaculata. J. Chem. Ecol. 25, 2477-2494.
Kronmuller, C., 2012. Review of the subspecies of Scolopendra subspinipes Leach, 1815
with the new description of the South Chinese member of the genus Scolopendra
Linnaeus, 1758 named Scolopendra hainanum spec. nov. Spixiana 35, 19-27.
Kuhn-Nentwig, L., 2003. Antimicrobial and cytolytic peptides of venomous arthropods.
Cell. Mol. Life Sci. 60, 2651-2668.
Kuhn-Nentwig, L., Nentwig, W., 1997. Venom of the hunting spider Cupiennius salei
(Ctenidae). Toxicon 35, 813.
Kuhn-Nentwig, L., Schaller, J., Nentwig, W., 2004. Biochemistry, toxicology and
ecology of the venom of the spider Cupiennius salei (Ctenidae). Toxicon 43, 543553.
431
Kuruppu, S., Smith, A.I., Isbister, G.K., Hodgson, W.C., 2008. Neurotoxins from
Australo-Papuan elapids: a biochemical and pharmacological perspective. Crit.
Rev. Toxicol. 38, 73-86.
La Rivers, I., 1948. Some Hawaiian ecological notes. Wasmann Collector 7, 85-110.
Lawrence, R.F., 1947. Some observations on the post-embryonic development of the
Natal forest centipede, Cormocephalus multispinus (Kraep.). Ann. Natal Mus. 11,
139-156.
Lawrence, T.C., 1934. Notes on the feeding habits of Scolopendra subspinipes Leach
(Myriapoda). Proc. Hawaii. Entomol. Soc. 8, 497-498.
Leontovich, A.A., Zhang, J.S., Shimokawa, K., Nagase, H., Sarras, M.P., 2000. A novel
Hydra matrix metalloproteinase (HMMP) functions in extracellular matrix
degradation, morphogenesis and the maintenance of differentiated cells in the foot
process. Development 127, 907-920.
Lesniewska, M., Lesniewski, P., Szybiak, K., 2008. Effect of urbanization on centipede
(Chilopoda) diversity in the Wielkopolska-Kujawy Lowlands of western Poland.
Biologia 63, 711-719.
Lewis, J.G.E., 1968. Individual variation in a population of the centipede Scolopendra
amazonica from Nigeria and its implications for methods of taxonomic
discrimination in the Scolopendridae. J. Linn. Soc. Lond. Zool. 47, 315-326.
Lewis, J.G.E., 1970. The biology of Scolopendra amazonica in Nigerian guinea
savannah. Bull. Mus. Hist. nat., Paris 41, 85-90.
Lewis, J.G.E., 1972. The life histories and distribution of the centipedes Rhysida nuda
togoensis and Ethmostigmus trigonopodus (Scolopendromorpha: Scolopendridae)
in Nigeria. J. Zool. (Lond.) 167, 399-414.
432
123-137.
Lewis, J.G.E., Edgecombe, G.D., Shelley, R.M., 2005. A proposed standardized
terminology for the external taxonomic characters of Scolopendromorpha
(Chilopoda). Fragm. Faun. 48, 1-8.
Lewis, R.J., Dutertre, S., Vetter, I., Christie, M.J., 2012. Conus venom peptide
pharmacology. Pharmacol. Rev. 64, 259-298.
Li, D.Q., Jackson, R.R., Barrion, A.T., 1999. Parental and predatory behaviour of
Scytodes sp., an araneophagic spitting spider (Araneae : Scytodidae) from the
Philippines. J. Zool. 247, 293-310.
Li, R.L., Zhang, L., Fang, Y., Han, B., Lu, X.S., Zhou, T.E., Feng, M., Li, J.K., 2013.
Proteome and phosphoproteome analysis of honeybee (Apis mellifera) venom
collected from electrical stimulation and manual extraction of the venom gland.
BMC Genomics 14, DOI: 10.1186/1471-2164-14-766.
Sep. Sci. 33, 410-421.
Liao, Z., Cao, J., Li, S.M., Yan, X.J., Hu, W.J., He, Q.Y., Chen, J.J., Tang, J.Z., Xie, J.Y.,
Liang, S.P., 2007. Proteomic and peptidomic analysis of the venom from Chinese
tarantula Chilobrachys jingzhao. Proteomics 7, 1892-1907.
Lim, J., Fudge, D.S., Levy, N., Gosline, J.M., 2006. Hagfish slime ecomechanics: testing
the gill-clogging hypothesis. J. Exp. Biol. 209, 702-710.
Lima, R.M., Estevao-Costa, M.I., Junqueira-de-Azevedo, I.L.M., Ho, P.L., Diniz,
M.R.V., Fortes-Dias, C.L., 2011. Phospholipase A2 inhibitors (beta PLIs) are
encoded in the venom glands of Lachesis muta (Crotalinae, Viperidae) snakes.
Toxicon 57, 172-175.
Liska, A.J., Shevchenko, A., 2003. Expanding the organismal scope of proteomics: crossspecies protein identification by mass spectrometry and its implications.
433
Lizano, S., Angulo, Y., Lomonte, B., Fox, J.W., Lambeau, G., Lazdunski, M., Gutierrez,
J.M., 2000. Two phospholipase A2 inhibitors from the plasma of Cerrophidion
(Bothrops) godmani which selectively inhibit two different group-II
phospholipase A2 myotoxins from its own venom: isolation, molecular cloning
and biological properties. Biochem. J. 346, 631-639.
Lubin, Y.D., Montogomery, G.G., 1981. Defenses of Nasutitermes termites (Isoptera,
Termitidae) against Tamandua anteaters (Edentata, Myrmecophagidae).
Biotropica 13, 66-76.
Luna, M.S.A., Hortencio, T.M.A., Ferreira, Z.S., Yamanouye, N., 2009. Sympathetic
outflow activates the venom gland of the snake Bothrops jararaca by regulating
the activation of transcription factors and the synthesis of venom gland proteins. J.
Exp. Biol. 212, 1535-1543.
Machado, G., Pomini, A.M., 2008. Chemical and behavioral defenses of the neotropical
harvestman Camarana flavipalpi (Arachnida: Opiliones). Biochem. Syst. Ecol.
36, 369-376.
Machado, G., Raimundo, R.L.G., Oliveira, P.S., 2000. Daily activity schedule,
gregariousness, and defensive behaviour in the Neotropical harvestman
Goniosoma longipes (Opiliones: Gonyleptidae). J. Nat. Hist. 34, 587-596.
Mackessy, S.P., 1988. Venom ontogeny in the Pacific rattlesnakes Crotalus viridis helleri
and Crotalus viridis oreganus. Copeia, 92-101.
Mackessy, S.P., 2002. Biochemistry and pharmacology of colubrid snake venoms. J.
Toxicol.: Toxin Rev. 21, 43-83.
Mackessy, S.P., 2010. Evolutionary trends in venom composition in the Western
Rattlesnakes (Crotalus viridis sensu lato): toxicity vs. tenderizers. Toxicon 55,
1463-1474.
Mackessy, S.P., Sixberry, N.A., Heyborne, W.H., Fritts, T., 2006. Venom of the brown
treesnake, Boiga irregularis: ontogenetic shifts and taxa-specific toxicity.
Toxicon 47, 537-548.
434
Mackessy, S.P., Williams, K., Ashon, K.G., 2003. Ontogenetic variation in venom
composition and diet of Crotalus oreganus concolor: a case of venom
paedomorphosis? Copeia, 769-782.
Malli, H., Kuhn-Nentwig, L., Imboden, H., Nentwig, W., 1999. Effects of size, motility
and paralysation time of prey on the quantity of venom injected by the hunting
spider Cupiennius salei. J. Exp. Biol. 202, 2083-2089.
Malli, H., Vapenik, Z., Nentwig, W., 1993. Ontogenic changes in the toxicity of the
venom of the spider Cupiennius salei (Araneae, Ctenidae). Zool. Jahrb., Abt. allg.
Zool. Physiol. Tiere 97, 113-122.
Malyshev, S.I., 1968. Genesis of the Hymenoptera and the phases of their evolution.
Richards, O.W., Uvarov, B., editors. Methuen, London.
Maretic, Z., 1987. Spider venoms and their effect, in: Nentwig, W., Aitchison-Benell,
C.W. (Eds.), Ecophysiology of spiders. Springer-Verlag, Berlin, pp. 448.
435
Marlier, J.F., Quinet, Y., de Biseau, J.C., 2004. Defensive behaviour and biological
activities of the abdominal secretion in the ant Crematogaster scutellaris
(Hymenoptera: Myrmicinae). Behav. Processes 67, 427-440.
Maschwitz, U., 1964. Gefahrenalarmstoffe und gefahrenalarmierung bei sozialen
Hymenopteren. Z. Vergleich. Physiol. 47, 596-655.
Maschwitz, U., Jessen, K., Maschwitz, E., 1981. Foaming in Pachycondyla - a new
defense mechanism in ants. Behav. Ecol. Sociobiol. 9, 79-81.
Maschwitz, U., Tho, Y.P., 1978. Phenols as defensive secretion in a Malayan cockroach,
Archiblatta hoeveni Vollenhoven. J. Chem. Ecol. 4, 375-381.
Matsuura, M., Sakagami, S.F., 1973. A bionomic sketch of the giant hornet, Vespa
mandarinia, a serious pest for Japanese apiculture. Journal of the Faculty of
Science, Hokkaido University. Series 6, Zoology 19, 125-162.
Matthews, R.W., Matthews, J.R., 2010. Insect behavior, 2nd edition. Springer, Dordrecht.
McAlister, W., 1960. The spitting habit in the spider Scytodes intricata Banks (Family
Scytodidae). Texas Journal of Science 12, 17-20.
Toxicon 55, 250-255.
McCleary, R.J.R., Kini, R.M., 2013. Non-enzymatic proteins from snake venoms: a gold
mine of pharmacological tools and drug leads. Toxicon 62, 56-74.
McCormick, S.J., Polis, G.A., 1990. Prey, predators, and parasites, in: Polis, G.A. (Ed.),
The biology of scorpions. Standford University Press, Stanford, pp. 295-320.
436
Copeia 2006, 818-825.
McDaniel, M.M., 1968. Notes on the biology of California scorpions. Entomol. News 79,
278-284.
McFee, R.B., Caraccio, T.R., Mofenson, H.C., McGuigan, M.A., 2002. Envenomation by
the Vietnamese centipede in a Long Island pet store. J. Toxicol. Clin. Toxicol. 40,
573-574.
McMonigle, O., 2003. Giant centipedes: the enthusiast's handbook. Elytra & Antenna,
Brunswick Hills.
Mebs, D., 2001. Toxicity in animals. Trends in evolution? Toxicon 39, 87-96.
Raton.
Medeiros, C.R., Susaki, T.T., Knysak, I., Cardoso, J.L.C., Malaque, C.M.S., Fan, H.W.,
Santoro, M.L., Franca, F.O.S., Barbaro, K.C., 2008. Epidemiologic and clinical
survey of victims of centipede stings admitted to Hospital Vital Brazil (Sao Paulo,
Brazil). Toxicon 52, 606-610.
Meier, J., Freyvogel, T.A., 1980. Comparative studies on venoms of the fer-de-lance
(Bothrops atrox), carpet viper (Echis carinatus) and spitting cobra (Naja
nigricollis) snakes at different ages. Toxicon 18, 661-662.
Morphol. 206, 303-312.
Miloslavina, A., Ebert, C., Tietze, D., Ohlenschlager, O., Englert, C., Gorlach, M., Imhof,
D., 2010. An unusual peptide from Conus villepinii: synthesis, solution structure,
and cardioactivity. Peptides 31, 1292-1300.
Minton, S.A., 1969. The feeding strike of the timber rattlesnake. J. Herpetol. 3, 121-124.
437
Mirtschin, P., Davis, R., 1992. Snakes of Australia: dangerous and harmless. Hill of
Content, Melbourne.
Mohamed, A.H., Abusinna, G., Elshabaka, H.A., Elaal, A.A., 1983. Proteins, lipids,
lipoproteins and some enzyme characterizations of the venom extract from the
centipede Scolopendra morsitans. Toxicon 21, 371-377.
Mohamed, A.H., Zaid, E., Elbeih, N.M., Elaal, A.A., 1980. Effects of an extract from the
centipede Scolopendra moristans on intestine, uterus and heart contractions and
on blood-glucose and liver and muscle glycogen levels. Toxicon 18, 581-589.
47, 189-190.
Moller, C., Melaun, C., Castillo, C., Diaz, M.E., Renzelman, C.M., Estrada, O., Kuch, U.,
Lokey, S., Mari, F., 2010. Functional hypervariability and gene diversity of
cardioactive neuropeptides. J. Biol. Chem. 285, 40673-40680.
Moore, K.A., Williams, D.D., 1990. Novel strategies in the complex defense repertoire of
a stonefly (Pteronarcys dorsata) nymph. Oikos 57, 49-56.
Morgan, P.N., 1969. Preliminary studies on venom from the brown recluse spider
Loxosceles reclusa. Toxicon 6, 161-165.
Morgenstern, D., King, G.F., 2013. The venom optimization hypothesis revisited.
Toxicon 63, 120-128.
Morse, D.H., 1999. Location of successful strikes on prey by juvenile crab spiders
Misumena vatia (Araneae, Thomisidae). J. Arachnol. 27, 171-175.
438
Murphy, G., Nagase, H., 2008. Progress in matrix metalloproteinase research. Mol. Asp.
Med. 29, 290-308.
Nakagawa, S., Cuthill, I.C., 2007. Effect size, confidence interval and statistical
significance: a practical guide for biologists. Biol. Rev. 82, 591-605.
Nascimento, D.G., Rates, B., Santos, D.M., Verano-Braga, T., Barbosa-Silva, A., Dutra,
A.A.A., Biondi, I., Martin-Eauclaire, M.F., De Lima, M.E., Pimenta, A.M.C.,
2006. Moving pieces in a taxonomic puzzle: venom 2D-LC/MS and data
clustering analyses to infer phylogenetic relationships in some scorpions from the
Buthidae family (Scorpiones). Toxicon 47, 628-639.
Nelsen, D.R., Nisani, Z., Cooper, A.M., Fox, G.A., Gren, E.C.K., Corbit, A.G., Hayes,
W.K., 2013. Poisons, toxungens, and venoms: redifining and classifying toxic
biological secretions and the organsims that employ them. Biol. Rev. doi:
10.1111/brv.12062.
Neves-Ferreira, A.G.C., Valente, R.H., Perales, J., Domont, G.B., 2010. Natural
inhibitors: innate immunity to snake venoms, in: Mackessy, S.P. (Ed.), Venoms
and toxins of reptiles. CRC Press, Boca Raton, pp. 259-284.
Nisani, Z. 2008. Behavioral and physiological ecology of scorpion venom expenditure:
stinging, spraying, and venom regeneration [Ph.D. dissertation]. Loma Linda,
California: Loma Linda University.
Physiol. 147, 509-513.
Nisani, Z., Hayes, W.K., 2011. Defensive stinging by Parabuthus transvaalicus
scorpions: risk assessment and venom metering. Anim. Behav. 81, 627-633.
Nolen, T.G., Johnson, P.M., 2001. Defensive inking in Aplysia spp.: multiple episodes of
ink secretion and the adaptive use of a limited chemical resource. J. Exp. Biol.
204, 1257-1268.
439
Nolen, T.G., Johnson, P.M., Kicklighter, C.E., Capo, T., 1995. Ink secretion by the
marine snail Aplysia californica enhances its ability to escape from a natural
predator. J. Comp. Physiol. A-Sens. Neural Behav. Physiol. 176, 239-254.
Norris, R. 2007. Centipede envenomation. In: Slapper, D., VanDeVoort, J., Farina, G.,
Halamka, J., Plantz, S., editors. eMedicine from WebMD. Boston: eMedicine
Corporation.
Nutting, W.L., Blum, M., Fales, H.M., 1974. Behavior of the North American termite
Tenuirostritermes tenuirostris, with special reference to the soldier frontal gland
secretion, its chemical composition, and use in defense. Psyche 81, 167-177.
Obin, M.S., Vander Meer, R.K., 1985. Gaster flagging by fire ants (Solenopsis spp.):
funtional significance of venom dispersal behavior. J. Chem. Ecol. 11, 1757-1768.
Ochieng, J., Chaudhuri, G., 2010. Cystatin superfamily. J. Health Care Poor Underserved
21, 51-70.
Olamendi-Portugal, T., Garcia, B.I., Lopez-Gonzalez, I., Van Der Walt, J., Dyason, K.,
Ulens, C., Tytgat, J., Felix, R., Darszon, A., Possani, L.D., 2002. Two new
scorpion toxins that target voltage-gated Ca2+ and Na+ channels. Biochem.
Biophys. Res. Commun. 299, 562-568.
Olson, E.J., 2000. Parachartergus fraternus (Gribodo) (Hymenoptera: Vespidae:
Polistinae) uses venom when taking caterpillar prey. Psyche 103, 85-93.
Orange, P., 1989. Incidents of predation on reptiles by invertebrates. Herpetofauna 19,
31-32.
Orivel, J., Souchal, A., Cerdan, P., Dejean, A., 2000. Prey capture behavior of the
arboreal ponerine ant Pachycondyla goeldii (Hymenoptera: Formicidae).
Sociobiology 35, 131-140.
Oron, U., Bdolah, A., 1973. Regulation of protein synthesis in venom gland of viperid
snakes. J. Cell Biol. 56, 177-190.
Oron, U., Kinamon, S., Bdolah, A., 1978. Asynchrony in the synthesis of secretory
proteins in the venom gland of the snake Vipera palaestinae. Biochem. J. 174,
733-739.
Oukkache, N., Chgoury, F., Lalaoui, M., Alagon Cano, A., Ghalim, N., 2013.
Comparison between two methods of scorpion venom milking in Morocco. J.
Venom. Anim. Tox. incl. Trop. Dis. 19, doi:10.1186/1678-9199-19-5.
440
Oukkache, N., Rosso, J.P., Alami, M., Ghalim, N., Saile, R., Hassar, M., Bougis, P.E.,
Martin-Eauclaire, M.F., 2008. New analysis of the toxic compounds from the
Androctonus mauretanicus mauretanicus scorpion venom. Toxicon 51, 835-852.
Parks, J., Stoecker, W.V., Kristensen, C., 2006. Observations on Loxosceles reclusa
(Araneae, Sicariidae) feeding on short-horned grasshoppers. J. Arachnol. 34, 221226.
Pasteels, J.M., Gregoire, J.C., Rowell-Rahier, M., 1983. The chemical ecology of defense
in arthropods. Annu. Rev. Entomol. 28, 263-289.
introduction, in: Buccianti, A., Mateu-Figueras, G., Pawlowsky-Glahn, V. (Eds.),
Perret, B.A. 1973. Biologie, gift und giftigkeit der orthognathen spinne Pterinochilus
spec. (fam. Theraphosidae) [Ph.D. thesis]: University of Basel. 80 p.
Perret, B.A., 1977a. Determination of amount of venom released by striking tarantulas. J.
Ga. Entomol. Soc. 12, 329-333.
Perret, B.A., 1977b. Venom regeneration in tarantula spiders—I: analysis of venom
Physiol. 56, 607-613.
Phillips, D.J., Swenson, S.D., Markland, F.S., Jr., 2010. Thrombin-like snake venom
serine proteinases, in: Mackessy, S.P. (Ed.), Handbook of venoms and toxins of
reptiles. CRC Press, Boca Raton, pp. 139-154.
441
Pilz, C., Melzer, R.R., Spelda, J., 2007. Morphometrics and SEM analysis of the species
pair Lithobius mutabilis L. Koch, 1862 and L. glacialis Verhoeff, 1937
(Chilopoda: Lithobiomorpha). Org. Divers. Evol. 7, 270e1-270e20.
Pimenta, A.M.C., Almeida, F.D., de Lima, M.E., Martin-Eauclaire, M.F., Bougis, P.E.,
2003. Individual variability in Tityus serrulatus (Scorpiones, Buthidae) venom
elicited by matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry. Rapid Commun. Mass Spectrom. 17, 413-418.
Pimenta, A.M.C., Stocklin, R., Favreau, P., Bougis, P.E., Martin-Eauclaire, M.F., 2001.
Moving pieces in a proteomic puzzle: mass fingerprinting of toxic fractions from
the venom of Tityus serrulatus (Scorpiones, Buthidae). Rapid Commun. Mass
Spectrom. 15, 1562-1572.
Pocock, R.I., 1893. Notes upon the habits of some living scorpions. Nature 48, 104-107.
Pollard, S.D., 1990. The feeding strategy of a crab spider, Diaea sp. indet
(Araneae,Thomisidae): post-capture decision rules. J. Zool. 222, 601-615.
Prashanth, J.R., Lewis, R.J., Dutertre, S., 2012. Towards an integrated venomics
approach for accelerated conopeptide discovery. Toxicon 60, 470-477.
Quinlan, T.G., Smith, G.T., Calver, M.C., 1995. Relationships between morphology and
feeding behavior in the syntopic scorpions Urodacus armatus Pocock and
Urodacus novaehollandiae Peters (Scorpiones, Scorpionidae). J. Aust. Entomol.
Soc. 34, 277-279.
Quintero-Hernandez, V., Ortiz, E., Rendon-Anaya, M., Schwartz, E.F., Becerril, B.,
Corzo, G., Possani, L.D., 2011. Scorpion and spider venom peptides: gene
cloning and peptide expression. Toxicon 58, 644-663.
Quinton, L., Demeure, K., Dobson, R., Gilles, N., Gabelica, V., De Pauw, E., 2007. New
method for characterizing highly disulfide-bridged peptides in complex mixtures:
application to toxin identification from crude venoms. J. Proteome Res. 6, 32163223.
442
Radcliffe, C.W., Chiszar, D., Oconnell, B., 1980. Effects of prey size on poststrike
behavior in rattlesnakes (Crotalus durissus, C. enyo, and C. viridis). Bull Psychon
Soc 16, 449-450.
Radl, R.C., 1992. Brood care in Scolopendra cingulata Latreille, in: Meyer, E., Thaler,
K., Schedl, W. (Eds.), Proceedings of the 8th international congress of
myriapodology, Innsbruck, Austria, July 1990. Berichte des
Naturwissenschaftenlich Medizinischen Vereins in Innsbruck, Supplementum 10,
123-127.
Rates, B., Ferraz, K.K.F., Borges, M.H., Richardson, M., De Lima, M.E., Pimenta,
A.M.C., 2008. Tityus serrulatus venom peptidomics: assessing venom peptide
diversity. Toxicon 52, 611-618.
Rathmayer, W., 1978. Venoms of Sphecidae, Pompilidae, Mutillidae, and Bethylidae, in:
Bettini, S. (Ed.), Arthropod venoms (Handbook of experimental pharmacology,
vol. 48). Springer-Verlag, Berlin, pp. 661-690.
Read, V.M.S., Hughes, R.N., 1987. Feeding behavior and prey choice in Macroperipatus
torquatus (Onychophora). Proc. R. Soc. Lond. Ser. B. Biol. sci. 230, 483-507.
Rehling, G.C. 2002. Venom expenditure in multiple bites by rattlesnakes and
cottonmouths [M.S. thesis]. Loma Linda, California: Loma Linda University.
Rein, J.O., 1993. Sting use in 2 species of Parabuthus scorpions (Buthidae). J. Arachnol.
21, 60-63.
Rein, J.O., 2003. Prey capture behavior in the East African scorpions Parabuthus
leiosoma (Ehrenberg, 1828) and P. pallidus Pocock, 1895 (Scorpiones: Buthidae).
Euscorpius 6, 1-8.
Reinhard, J., Rowell, D.M., 2005. Social behaviour in an Australian velvet worm,
Euperipatoides rowelli (Onychophora : Peripatopsidae). J. Zool. 267, 1-7.
443
Richards, R., St Pierre, L., Trabi, M., Johnson, L.A., de Jersey, J., Masci, P.P., Lavin,
M.F., 2011. Cloning and characterisation of novel cystatins from elapid snake
venom glands. Biochimie 93, 659-668.
Richardson, K.C., Hinchliffe, P.M., 1983. Caudal glands and their secretions in the
western spiny-tailed gecko, Diplodactylus spinigerus. Copeia 1983, 161-169.
Robertson, L.N., 1993. Influence of tillage intensity (including no till) on soil-dwelling
pests and predatory soil animals in central Queensland crops, in: Covey, S.A.,
Dall, D.J., Milne, W.M. (Eds.), Pest control and sustainable agriculture. CSIRO,
Canberra, pp. 349-352.
Robertson, P.L., 1971. Pheromones involved in aggressive behaviour in ant, Myrmecia
gulosa. J. Insect Physiol. 17, 691-715.
Robinson, M.H., 1969. Predatory behavior of Argiope argentata (Fabricius). Am. Zool.
9, 161-173.
Robinson, M.H., Mirick, H., Turner, O., 1969. The predatory behavior of some araneid
spiders and the origin of immobilization wrapping. Psyche 76, 487-501.
Robinson, M.H., Olazarri, J., 1971. Units of behavior and complex sequences in the
predatory behavior of Argiope argentata (Fabricius): (Araneae: Araneidae).
Smithsonian Institution Press, Washington.
Rocha-e-Silva, T.A.A., Sutti, R., Hyslop, S., 2009. Milking and partial characterization of
venom from the Brazilian spider Vitalius dubius (Theraphosidae). Toxicon 53,
153-161.
Rochelle, M.J., Kardong, K.V., 1993. Constriction versus envenomation in prey capture
by the Brown Tree Snake, Boiga irregularis (Squamata, Colubridae).
Herpetologica 49, 301-304.
Rodriguez-Acosta, A., Gassette, J., Gonzalez, A., Ghisoli, M., 2000. Centipede
(Scolopendra gigantea Linneaus 1758) envenomation in a newborn. Rev. Inst.
Med. Trop. Sao Paulo 42, 341-2.
Rodriguez-Robles, J.A., 1992. Notes on the feeding behavior of the Puerto Rican Racer,
Alsophis portoricensis (Serpentes: Colubridae). J. Herpetol. 26, 100-102.
444
Rodriguez-Robles, J.A., Leal, M., 1993. Effects of prey type on the feeding behavior of
Alsophis portoricensis (Serpentes, Colubridae). J. Herpetol. 27, 163-168.
Romi-Lebrun, R., Martin-Eauclaire, M.F., Escoubas, P., Wu, F.Q., Lebrun, B., Hisada,
M., Nakajima, T., 1997. Characterization of four toxins from Buthus martensi
scorpion venom, which act on apamin-sensitive Ca2+-activated K+ channels. Eur.
J. Biochem. 245, 457-464.
Rosenberg, H.I., Russell, A.P., 1980. Structural and functional aspects of tail squirting: a
unique defense mechanism of Diplodactylus (Reptilia, Gekkonidae). Can. J.
Zool.-Rev. Can. Zool. 58, 865-881.
Rosenberg, H.I., Russell, A.P., Kapoor, M., 1984. Preliminary characterization of the
defensive secretion of Diplodactylus (Reptilia, Gekkonidae). Copeia 1984, 10251028.
Rosin, R., Shulov, A.S., 1963. Studies on the scorpion Nebo hierochonticus. Proceedings
of the Zoological Society of London 140, 547-575.
Rossini, C., Attygalle, A.B., Gonzalez, A., Smedley, S.R., Eisner, M., Meinwald, J.,
Eisner, T., 1997. Defensive production of formic acid (80%) by a carabid beetle
(Galerita lecontei). Proc. Natl. Acad. Sci. U.S.A. 94, 6792-6797.
Rotenberg, D., Bambereger, E.S., Kochva, E., 1971. Studies on ribonucleic acid synthesis
in venom glands of Vipera palaestinae (Ophidia, Reptilia). Biochem. J. 121, 609612.
Roth, L.M., Eisner, T., 1962. Chemical defenses of arthropods. Annu. Rev. Entomol. 7,
107-136.
Rowe, M.P., Owings, D.H., 1990. Probing, assessment, and management during
interactions between ground squirrels and rattlesnakes. 1. Risks related to
rattlesnake size and body temperature. Ethology 86, 237-249.
Ruppert, E.E., Fox, R.S., Barnes, R.D., 2004. Invertebrate zoology : a functional
evolutionary approach. Thomson-Brooks/Cole, Belmont.
Russell, F.E., Walter, F.G., Bey, T.A., Fernandez, M.C., 1997. Snakes and snakebite in
Central America. Toxicon 35, 1469-1522.
445
Ruxton, G.D., Sherratt, T.N., Speed, M.P., 2004. Avoiding attack: the evolutionary
ecology of crypsis, warning signals, and mimicry. Oxford University Press,
Oxford; New York.
Ryan, G.E., Croft, J.D., 1974. Observations on the food of the fox, Vulpes vulpes (L.) in
Kinchega National Park, Menindee, NSW. Australian Wildlife Research 1, 89-94.
Samy, R.P., Pachiappan, A., Gopalakrishnakone, P., Thwin, M.M., Hian, Y.E., Chow,
V.T.K., Bow, H., Weng, J.T., 2006. In vitro antimicrobial activity of natural
toxins and animal venoms tested against Burkholderia pseudomallei. BMC Infect.
Dis. 6, DOI: 10.1186/1471-2334-6-100.
Sandefer, C., 1998. The giant centipedes of the genus Scolopendra: their captive care and
husbandry. Privately printed.
Santos, L.D., Pieroni, M., Menegasso, A.R.S., Pinto, J., Palma, M.S., 2011. A new
scenario of bioprospecting of Hymenoptera venoms through proteomic approach.
J. Venom. Anim. Tox. incl. Trop. Dis. 17, 364-377.
Sato, S., Kushibuchi, K., Yasuda, H., 2009. Effect of reflex bleeding of a predatory
ladybird beetle, Harmonia axyridis (Pallas) (Coleoptera: Coccinellidae), as a
means of avoiding intraguild predation and its cost. Appl. Entomol. Zool. 44, 203206.
Savel-Niemann, A., Roth, D., 1989. Biochemical analysis of tarantula venom (Eurypelma
californicum). Naturwissenschaften 76, 212-213.
Savitzky, A.H., Mori, A., Hutchinson, D.A., Saporito, R.A., Burghardt, G.M., Lillywhite,
H.B., Meinwald, J., 2012. Sequestered defensive toxins in tetrapod vertebrates:
principles, patterns, and prospects for future studies. Chemoecology 22, 141-158.
Savitzky, B.A.C., 1992. Laboratory studies on piscivory in an opportunistic pitviper, the
cottonmouth, Agkistrodon piscivorus, in: Campbell, J.A., Brodie, E.D., Jr. (Eds.),
Biology of the pitvipers. Selva, Tyler, pp. 347-368.
Schanbacher, F.L., Lee, C.K., Hall, J.E., Wilson, I.B., Howell, D.E., Odell, G.V., 1973.
Composition and properties of tarantula Dugesiella hentzi (Girard) venom.
Toxicon 11, 21-29.
Schatz, B., Lachaud, J.P., Beugnon, G., 1997. Graded recruitment and hunting strategies
linked to prey weight and size in the ponerine ant Ectatomma ruidum. Behav.
Ecol. Sociobiol. 40, 337-349.
446
Schenberg, S., Pereira-Lima, F.A., 1978. Venoms of Ctenidae, in: Bettini, S. (Ed.),
Arthropod venoms (Handbook of experimental pharmacology, vol. 48). SpringerVerlag, Berlin, pp. 215-245.
Schenberg, S., Pereira-LIma, F.A., Nogueira-Schiripa, L.N., Nagamori, A., 1970.
Unparalleled regeneration of snake venom components in successive milkings.
Toxicon 8, 152.
Schmidt, J.O., 1982. Biochemistry of insect venoms. Annu. Rev. Entomol. 27, 339-368.
Schmidt, J.O., 1990. Hymenopteran venoms: striving toward the ultimate defense against
vertebrates, in: Evans, D.L., Schmidt, J.O. (Eds.), Insect defenses: adaptive
mechanisms and strategies of prey and predators. State University of New York,
Albany, pp. 482.
Schwartz, E.F., Bartok, A., Schwartz, C.A., Papp, F., Gomez-Lagunas, F., Panyi, G.,
Possani, L.D., 2013. OcyKTx2, a new K+- channel toxin characterized from the
venom of the scorpion Opisthacanthus cayaporum. Peptides 46, 40-46.
Schwartz, E.F., Camargos, T.S., Zamudio, F.Z., Silva, L.P., Bloch, C., Caixeta, F.,
Schwartz, C.A., Possani, L.D., 2008. Mass spectrometry analysis, amino acid
sequence and biological activity of venom components from the Brazilian
scorpion Opisthacanthus cayaporum. Toxicon 51, 1499-1508.
385-405.
Senko, M.W., Beu, S.C., McLafferty, F.W., 1995. Determination of monoisotopic masses
and ion populations for large biomolecules from resolved isotopic distributions. J.
Am. Soc. Mass Spectrom. 6, 229-233.
Serrano, S.M.T., Maroun, R.C., 2005. Snake venom serine proteinases: sequence
homology vs. substrate specificity, a paradox to be solved. Toxicon 45, 11151132.
Shapiro, E., Koester, J., Byrne, J.H., 1979. Aplysia ink release: central locus for selective
sensitivity to long-duration stimuli. J. Neurophysiol. 42, 1223-1232.
447
Martinsville.
Sherbrooke, W.C., 1990. Predatory behavior of captive greater roadrunners feeding on
horned lizards. Wilson Bull. 102, 171-174.
Sherbrooke, W.C., 1991. Behavioral (predator-prey) interactions of captive grasshopper
mice (Onychomys torridus) and horned lizards (Phrynosoma cornutum and P.
modestum). Am. Midl. Nat. 126, 187-195.
Sherbrooke, W.C., Mason, J.R., 2005. Sensory modality used by coyotes in responding to
antipredator compounds in the blood of Texas horned lizards. Southwest. Nat. 50,
216-222.
Sherbrooke, W.C., Middendorf, G.A., 2001. Blood-squirting variability in horned lizards
(Phrynosoma). Copeia 2001, 1114-1122.
Sherbrooke, W.C., Middendorf, G.A., 2004. Responses of kit foxes (Vulpes macrotis) to
antipredator blood-squirting and blood of Texas Horned Lizards (Phrynosoma
cornutum). Copeia 2004, 652-658.
Shevchenko, A., Valcu, C.M., Junqueira, M., 2009. Tools for exploring the
proteomosphere. J. Proteomics 72, 137-144.
Shine, R., 1985. Prey constriction by venomous snakes: a review, and new data on
Australian species. Copeia 1985, 1067-1071.
Shugg, H.B., 1961. Predation on mouse by centipede. West. Aust. Nat. 8, 52.
Simaiakis, S.M., Giokas, S., Korsos, Z., 2011. Morphometric and meristic diversity of the
species Scolopendra cingulata Latreille, 1829 (Chilopoda: Scolopendridae) in the
Mediterranean region. Zool. Anz. 250, 67-79.
Sissom, W.D., Polis, G.A., Watt, D.D., 1990. Field and laboratory methods, in: Polis,
445-461.
Slobodchikoff, C.N., 1979. Utilization of harvester ant (Hymenoptera, Formicidae) debris
by tenebrionid beetles (Coleoptera, Tenebronidae). Environ. Entomol. 8, 770-772.
448
Sokal, R.R., Rohlf, F.J., 1995. Biometry: the principles and practice of statistics in
biological research. W. H. Freeman & Co., New York.
Sousa, J.R.F., Monteiro, R.Q., Castro, H.C., Zingali, R.B., 2001. Proteolytic action of
Bothrops jararaca venom upon its own constituents. Toxicon 39, 787-792.
Sousa, S.R., Vetter, I., Lewis, R.J., 2013. Venom peptides as a rich source of Ca(v)2.2
channel blockers. Toxins 5, 286-314.
Souza, G., Catharino, R.R., Ifa, D.R., Eberlin, M.N., Hyslop, S., 2008. Peptide
fingerprinting of snake venoms by direct infusion nano-electrospray ionization
mass spectrometry: potential use in venom identification and taxonomy. J. Mass
Spectrom. 43, 594-599.
Soyer, B., 1938. Notes sur les Sphegiens et les Pompiles I. L'Anoplius concinnus
(Dahlbom). Bull. Soc. Entomol. Fr. 42, 29-30.
Spradbery, J.P., 1973. Wasps: an account of the bioogy and natural history of solitary and
social wasps. Dales, R.P., Martin, A.W., editors. University of Washington Press,
Seattle.
Stahnke, H.L., 1966. Some aspects of scorpion behavior. Bull. South. Calif. Acad. Sci.
65, 65-80.
Starr, C.K., 1990. Holding the fort: colony defense in some primitively social wasps, in:
Evans, D.L., Schmidt, J.O. (Eds.), Insect defenses: adaptive mechanisms and
strategies of prey and predators. State University of New York, Albany, pp. 422463.
Steiner, A.L., 1976. Digger wasp predatory behavior (Hymenoptera, Sphecidae). II.
Comparative study of closely related wasps (Larrinae: Liris nigra, Palearctic; L.
argentata and L. aequalis, Nearctic) that all paralyze crickets (Orthoptera,
Gryllidae). Z Tierpsychol 42, 343-380.
Steiner, A.L., 1979. Digger wasp predatory behavior (Hymenoptera, Sphecidae): fly
hunting and capture by Oxybelus uniglumis (Crabroninae, Oxybelini); a case of
extremely concentrated stinging pattern and prey nervous system. Can. J. Zool.Rev. Can. Zool. 57, 953-962.
449
Steiner, A.L., 1986. Stinging behaviour of solitary wasps, in: Piek, T. (Ed.), Venoms of
the Hymenoptera. Academic Press, London, pp. 63-160.
Stewart, J.W. 1997. Centipedes and millipedes. Texas Agricultural Extension Service:
Texas A & M University.
Strahan, R., 1959. Slime production in Myxine glutinosa Linnaeus. Copeia 1959, 165166.
Sunagar, K., Johnson, W.E., O'Brien, S.J., Vasconcelos, V., Antunes, A., 2012. Evolution
of CRISPs associated with toxicoferan-reptilian venom and mammalian
reproduction. Mol. Biol. Evol. 29, 1807-1822.
Sunagar, K., Undheim, E.A.B., Scheib, H., Gren, E.C.K., Cochran, C., Person, C.E.,
Koludarov, I., Kelln, W., Hayes, W.K., King, G.F., Antunes, A., Fry, B.G., 2014.
Intraspecific venom variation in the medically significant Southern Pacific
Rattlesnake (Crotalus oreganus helleri): biodiscovery, clinical and evolutionary
implications. J. Proteomics DOI:10.1016/j.jprot.2014.01.013.
Suter, R.B., Stratton, G.E., 2009. Spitting performance parameters and their
biomechanical implications in the spitting spider, Scytodes thoracica. J Insect Sci.
9.
Suter, R.B., Stratton, G.E., 2012. Predation by spitting spiders: elaborate venom gland,
intricate delivery system, in: Nentwig, W. (Ed.), Spider ecophysiology. SpringerVerlag, Heidelberg, pp. 241-251.
Tan, K., Kretzschmar, H., 2009. On a meeting between the Horn Viper and a centipede in
the Peloponnese, Southern Greece or the biter, bit. Adv. Sci. Lett. 2, 10-13.
Tang, X., Zhang, Y.Q., Hu, W.J., Xu, D.H., Tao, H., Yang, X.X., Li, Y., Jiang, L.P.,
Liang, S.P., 2010. Molecular diversification of peptide toxins from the tarantula
Haplopelma hainanum (Ornithoctonus hainana) venom based on transcriptomic,
peptidomic, and genomic analyses. J. Proteome Res. 9, 2550-2564.
Tare, T.G., Sutar, N.K., Renapurkar, D.M., 1986. A study of snake venom yield by
different methods of venom extraction. Amphib-reptil. 7, 187-191.
Tashima, A.K., Sanz, L., Camargo, A.C.M., Serrano, S.M.T., Calvete, J.J., 2008. Snake
venomics of the Brazilian pitvipers Bothrops cotiara and Bothrops fonsecai.
Identification of taxonomy markers. J. Proteomics 71, 473-485.
Taylor, D., Iddon, D., Sells, P., Semoff, S., Theakston, R.D.G., 1986. An investigation of
venom secretion by the venom gland cells of the carpet viper (Echis carinatus).
Toxicon 24, 651-659.
450
Tennant, A., 1985. A field guide to Texas snakes. Gulf Publishing, Houston.
Terlau, H., Olivera, B.M., 2004. Conus venoms: a rich source of novel ion channeltargeted peptides. Physiol. Rev. 84, 41-68.
Tobach, E., Gold, P., Ziegler, A., 1965. Preliminary observations of the inking behavior
of Aplysia (Varria). Veliger 8, 16-18.
Trevisan-Silva, D., Gremski, L.H., Chaim, O.M., da Silveira, R.B., Meissner, G.O.,
Mangili, O.C., Barbaro, K.C., Gremski, W., Veiga, S.S., Senff-Ribeiro, A., 2010.
Astacin-like metalloproteases are a gene family of toxins present in the venom of
different species of the brown spider (genus Loxosceles). Biochimie 92, 21-32.
Ulbrich, H.F., 2011. Graphing and communicating compositional data in high
dimensions, in: Egozcue, J.J., Tolosana-Delgado, R., Ortego, M.I. (Eds.),
Proceedings of the 4th international workshop on compositional data analysis
(CoDaWork 2011), Girona, Spain, 9-13 May 2011.
Uzel, A.P., Steinmann, G., Bertino, R., Korsaga, A., 2009. Necrotizing fasciitis and
cellulitis of the upper limb resulting from centipede bite: two case reports. Chir.
Main. 28, 322-325.
Uzenbaev, S.D., Lyabzina, S.N., 2009. An experimental study of the effects of spider
venom on animals. Entomol. Rev. 89, 479-486.
Vapenik, Z., Nentwig, W., 2000. The influence of hunger and breeding temperature on
the venom production of the spider Cupiennius salei (Araneae, Ctenidae).
Toxicon 38, 293-298.
451
Vassilevski, A.A., Kozlov, S.A., Grishin, E.V., 2009. Molecular diversity of spider
venom. Biochem.-Moscow 74, 1505-1534.
Veiga, S.S., da Silveira, R.B., Dreyfuss, J.L., Haoach, J., Pereira, A.M., Mangili, O.C.,
Gremski, W., 2000. Identification of high molecular weight serine-proteases in
Loxosceles intermedia (brown spider) venom. Toxicon 38, 825-839.
Veiga, S.S., Zanetti, V.C., Braz, A., Mangili, O.C., Gremski, W., 2001. Extracellular
matrix molecules as targets for brown spider venom toxins. Braz. J. Med. Biol.
Res. 34, 843-850.
Veraldi, S., Chiaratti, A., Sica, L., 2010. Centipede bite: a case report. Arch. Dermatol.
146, 807-808.
Okol. Tiere 36, 645-650.
Verts, B.J., Carraway, L.N., Kinlaw, A., 2001. Spilogale gracilis. Mammalian Species
674, 1-10.
Vetter, I., Davis, J.L., Rash, L.D., Anangi, R., Mobli, M., Alewood, P.F., Lewis, R.J.,
King, G.F., 2011. Venomics: a new paradigm for natural products-based drug
discovery. Amino Acids 40, 15-28.
Vieira, E.G.J., Rolim-Rosa, R., Iizuka, H., Furtado, M.d.F.D., Fernandes, W., 1988.
Influencias sazonal e do processo de extração sobre a produção, toxicidade do
veneno e sobrevida de Bothrops jararaca (Wied, 1824). Mem. Inst. Butantan 50,
29-35.
Voigtlander, K., 2011. Chilopoda—ecology, in: Minelli, A. (Ed.), The Myriapoda. Brill,
Leiden, pp. 309-325.
Voinstvenskii, M.A., Petrusenko, A.A., Boyarchuk, V.P., 1977. Trophic relations of the
rook (Corvus frugilegus L.) in steppe ecosystem: 1 - Nutrition (diet composition).
Vestn. Zool. 6, 9-17.
Von Euw, J., Fishelso.L, Parsons, J.A., Reichste.T, Rothschi.M, 1967. Cardenolides
(heart poisons) in a grasshopper feeding on milkweeds. Nature 214, 35-39.
Wallace, J.B., Blum, M.S., 1969. Refined defense mechanisms in Chrysomela scripta.
Ann. Entomol. Soc. Am. 62, 503-506.
452
London.
Walters, E.T., Erickson, M.T., 1986. Directional control and the functional organization
of defensive responses in Aplysia. J. Comp. Physiol. A-Sens. Neural Behav.
Physiol. 159, 339-351.
Ward, D.A., Sefton, E.M., Prescott, M.C., Webster, S.G., Wainwright, G., Rees, H.H.,
Fisher, M.J., 2010. Efficient identification of proteins from ovaries and
hepatopancreas of the unsequenced edible crab, Cancer pagurus, by mass
spectrometry and homology-based, cross-species searching. J. Proteomics 73,
2354-2364.
Waridel, P., Frank, A., Thomas, H., Surendranath, V., Sunyaev, S., Pevzner, P.,
Shevchenko, A., 2007. Sequence similarity-driven proteomics in organisms with
unknown genomes by LC-MS/MS and automated de novo sequencing.
Wasielewski, O., Skonieczna, M., 2008. Pleiotropic effects of the neuropeptides CCAP
and myosuppressin in the beetle, Tenebrio molitor L. J. Comp. Physiol. BBiochem. Syst. Environ. Physiol. 178, 877-885.
Watson, J.T., Sparkman, O.D., 2007. Introduction to mass spectrometry: instrumentation,
applications and strategies for data interpretation. John Wiley & Sons, Chichester.
Weatherston, J., Percy, J.E., Macdonald, L.M., Macdonald, J.A., 1979. Morphology of
the prothoracic defensive gland of Schizura concinna (JE Smith) (Lepidoptera,
Notodontidae) and the nature of its secretion. J. Chem. Ecol. 5, 165-177.
Weinstein, S.A., White, J., Keyler, D.E., Warrell, D.A., 2013. Non-front-fanged
colubroid snakes: a current evidence-based analysis of medical significance.
Toxicon 69, 103-113.
Wenhua, R., Shuangguan, Z., Daxiang, S., Kaiya, Z., Guang, Y., 2006. Induction,
purification and characterization of an antibacterial peptide scolopendrin I from
the venom of centipede Scolopendra subspinipes mutilans. Indian J. Biochem.
Biophys. 43, 88-93.
453
Westhoff, G., Boetig, M., Bleckmann, H., Young, B.A., 2010. Target tracking during
venom 'spitting' by cobras. J. Exp. Biol. 213, 1797-1802.
Westhoff, G., Tzschatzsch, K., Bleckmann, H., 2005. The spitting behavior of two
species of spitting cobras. J. Comp. Physiol. A -Neuroethol. Sens. Neural Behav.
Physiol. 191, 873-881.
Whiffler, L.A., Drusedau, M.U.H., Crewe, R.M., Hepburn, H.R., 1988. Defensive
behavior and the division of labor in the African honeybee (Apis mellifera
scutellata). J. Comp. Physiol. A-Sens. Neural Behav. Physiol. 163, 401-411.
Whitaker, P.B., Ellis, K., Shine, R., 2000. The defensive strike of the Eastern
Brownsnake, Pseudonaja textilis (Elapidae). Funct. Ecol. 14, 25-31.
Whitman, D.W., Billen, J.P.J., Alsop, D., Blum, M.S., 1991. Anatomy, ultrastructure, and
functional morphology of the metathoracic tracheal defensive glands of the
grasshopper Romalea guttata. Can. J. Zool.-Rev. Can. Zool. 69, 2100-2108.
Whitman, D.W., Blum, M.S., Alsop, D.W., 1990. Allomones: chemicals for defense, in:
Evans, D.L., Schmidt, J.O. (Eds.), Insect defenses: adaptive mechanisms and
strategies of prey and predators. State University of New York, Albany, pp. 482.
Whitman, D.W., Jones, C.G., Blum, M.S., 1992. Defensive secretion production in lubber
grasshoppers (Orthoptera, Romaleidae): influence of age, sex, diet, and discharge
frequency. Ann. Entomol. Soc. Am. 85, 96-102.
Whittemore, F.W., Keegan, H.L., Fitzgerald, C.M., Bryant, H.A., Flanigan, J.F., 1963.
Studies of scorpion antivenins: 2. Venom collection and scorpion colony
maintenance. Bull. W.H.O. 28, 505-511.
Wiener, S., 1956. The australian red back spider (Latrodectus hasseltii): II. Effect of
temperature on the toxicity of venom. Med. J. Aust. 43, 331-335.
Wiener, S., 1959. The Sydney funnel-web spider (Atrax robustus): II. Venom yield and
other characteristics of spider in captivity. Med. J. Aust. 46, 678-682.
749-752.
Will, K.W., Gill, A.S., Lee, H., Attygalle, A.B., 2010. Quantification and evidence for
mechanically metered release of pygidial secretions in formic acid-producing
carabid beetles. J Insect Sci. 10.
454
Willemart, R.H., Gnaspini, P., 2004. Spatial distribution, mobility, gregariousness, and
defensive behaviour in a Brazilian cave harvestman Goniosoma albiscriptum
(Arachnida, Opiliones, Gonyleptidae). Anim. Biol. 54, 221-235.
Willemse, G.T., Hattingh, J., Karlsson, R.M., Levy, S., Parker, C., 1979. Changes in
composition and protein concentration of puff adder (Bitis arietans) venom due to
frequent milking. Toxicon 17, 37-42.
Willey, M.B., Johnson, M.A., Adler, P.H., 1992. Predatory behavior of the basilica
spider, Mecynogea lemniscata (Araneae, Araneidae). Psyche 99, 153-168.
Williams, F.X., 1929. Notes on the habits of the cockroach-hunting wasps of the genus
Ampulex ses. lat., with particular reference to Ampulex (Rhinopsis) caniculatus
Say. Proc. Hawaii. Entomol. Soc. 7, 315-329.
Williams, S.C., 1987. Scorpion bionomics. Annu. Rev. Entomol. 32, 275-295.
Woodring, J.P., Blum, M.S., 1965. Anatomy, physiology, and comparative aspects of
repugnatorial glands of Orthocricus arboreus (Diplopoda, Spirobolida). J.
Morphol. 116, 99-108.
Wuester, W., 2010. What's your poison? Heredity 104, 519.
Yildiz, A., Biceroglu, S., Yakut, N., Bilir, C., Akdemir, R., Akilli, A., 2006. Acute
myocardial infarction in a young man caused by centipede sting. Emerg. Med. J.
23, e30.
You, W.K., Sohn, Y.D., Kim, K.Y., Park, D.H., Jang, Y., Chung, K.H., 2004.
Purification and molecular cloning of a novel serine protease from the centipede,
Scolopendra subspinipes mutilans. Insect Biochem. Mol. Biol. 34, 239-250.
455
Young, B.A., 2008. Perspectives on the regulation of venom expulsion in snakes, in:
Hayes, W.K., Beaman, K.R., Cardwell, M.D., Bush, S.P. (Eds.), The biology of
rattlesnakes. Loma Linda University Press, Loma Linda, pp. 181-190.
Young, B.A., Boetig, M., Westhoff, G., 2009. Functional bases of the spatial dispersal of
venom during cobra "spitting". Physiol. Biochem. Zool. 82, 80-89.
Young, B.A., Lee, C.E., Daley, K.M., 2002. Do snakes meter venom? Bioscience 52,
1121-1126.
Young, B.A., O'Shea, M., 2005. Analyses of venom spitting in African cobras (Elapidae:
Serpentes). Afr. Zool. 40, 71-76.
Zlotkin, E., Shulov, A.S., 1969. A simple device for collecting scorpion venom. Toxicon
7, 331-332.
456

Venom Yield, Regeneration, and Composition in the Centipede

Transcription

Similar documents

2015-16 snowmobile race program application

effects of hogna carolinensis and phidippus octopunctatus spider

a potent hallucinogen of animal origin

pages 27

American Brown Recluse Spider

Walwa World October 2011

English I - Texas Education Agency

MEERKATS

Improving the Weighing Process of Young Swine

Management of venomous snakebite