"Extracellular Matrix". In: Current Protocols in Cell Biology

Transcription

CHAPTER 10
Extracellular Matrix
INTRODUCTION
M
ost cells of multicellular organisms interact routinely with extracellular matrix
(ECM) molecules. The ECM provides structural support, as well as important
regulatory signals governing cellular growth, metabolism, and differentiation. Epithelial
cells of all types generally require various ECM components in the basement membranes
to which they adhere for maintaining their characteristic polarized organization, differentiated state, and speciÞc gene expression. Connective tissue cells nestle within matrices
of collagens, proteoglycans, and other ECM components. Even circulating blood cells
such as lymphocytes can interact extensively with ECM as they extravasate from blood
vessels and localize in tissues during recirculation and inßammation. Two particularly
dynamic tissue-remodeling processes in which ECM becomes critically important are
cell movements during embryonic development and wound repair. In addition, tumor
cells often must invade through basement membranes and connective tissue in order to
metastasize (see UNIT 19.1).
Throughout these various cell-biological processes, ECM can act both as a structural
scaffolding for cell adhesion and migration and as a trigger for signaling through ECM
receptors. Binding of speciÞc ECM molecules to their plasma-membrane receptors activates signal-transduction responses that can include activation of various tyrosine and
serine-threonine kinase families, MAP kinase systems, ion ßuxes, or phosphoinositide
and arachidonic acid pathways. The study of these complex processes has blossomed
recently due to the availability of individual puriÞed ECM molecules, along with the
realization that ECM modulates many crucial cell-biological functions.
reviews key ECM functions, as well as the biochemistry of the major classes of
ECM molecules. This comprehensive review provides a solid framework for understanding the initially somewhat bewildering complexity of ECMs, which range from basement
membranes to loose connective tissue, cartilage, and ligaments.
UNIT 10.1
Basement membranes are crucial for normal epithelial cell biology. UNIT 10.2 describes
the preparation of the basement-membrane extract termed Matrigel and the isolation
of two important components, laminin and type IV collagen. Although these proteins
are also available commercially, preparing them within the laboratory is much more
economical for larger-scale studies. These puriÞed proteins can be used in cell adhesion
assays (UNITS 9.1 & 9.2) and for studying other cell-biological responses such as migration
and differentiation.
Because extracellular matrices are generally three-dimensional, and their effects can be
greater than the sum of their isolated components, cell biologists also use ECM gels.
UNIT 10.3 details the preparation and use of gels of puriÞed collagen and of Matrigel
basement-membrane extracts. Collagen gels and Matrigel can be used to study the threedimensional behavior of epithelial, Þbroblastic, and other cells. Matrigel can also be used
in animals for angiogenesis assays (UNIT 10.3) and in tumor cell invasion assays (UNIT 12.2),
as well as to promote the survival and growth of primary tumor cells that would not
otherwise grow in vivo (UNIT 10.3).
Extracellular
Matrix
Current Protocols in Cell Biology 10.0.1-10.0.3, December 2009
Published online December 2009 in Wiley Interscience (www.interscience.wiley.com).
DOI: 10.1002/0471143030.cb1000s45
C 2009 John Wiley & Sons, Inc.
Copyright 10.0.1
Supplement 45
An alternative approach is to use intact extracellular matrices assembled by living cells
in culture; the cell monolayers are extracted away to leave just the three-dimensional
extracellular matrix. UNIT 10.4 describes the preparation of two forms of such complex
matrices, which can be used as substrates for cell biological studies in vitro.
A particularly widely used approach involves puriÞed extracellular matrix molecules as
substrates. Both Þbronectin and vitronectin can be puriÞed by afÞnity chromatography
from human plasma, providing large quantities of an individual adhesive protein useful
for cell culture studies using many types of cells. UNIT 10.5 presents simple, reliable
methods for purifying substantial quantities of Þbronectin from plasma. Also included
are two different protocols for purifying the cellular form of Þbronectin, by extraction
from cell surfaces or from Þbronectin secreted by cells into serum-free medium; this
form of Þbronectin contains additional peptide sequences and has moderately enhanced
adhesive activity. UNIT 10.6 describes a simple protocol for the isolation of vitronectin,
which is a cell adhesion protein used by many cells—e.g., by most adherent cells cultured
in medium that contains 10% serum.
Even though proteoglycans are a major class of diverse extracellular and cell-surface
molecules with important roles in a myriad of regulatory and structural functions, many
researchers have hesitated to study them because they are technically difÞcult to analyze.
These large, highly charged, readily aggregated molecules are now known to be present
not only in extracellular matrices, but also linked to membranes by phospholipid, in
transmembrane locations (e.g., the syndecans), or even in intracellular locations in some
cells. UNIT 10.7 provides numerous protocols and valuable tips for isolating and analyzing
these widely distributed molecules implicated in signal transduction, adhesion, and ECM
structure.
The glycoproteins and proteoglycans of the extracellular matrix are not static, but are
instead in a state of dynamic balance between synthesis and degradation. The continuous
turnover and active remodeling of the extracellular matrix in embryonic development,
growth, and tissue repair depend critically on carefully regulated degradation by proteolytic enzymes, particularly by the matrix metalloproteinases (MMPs). The MMPs are
regulated by activation and by their inhibitors (the TIMPs, as well as α2-macroglobulin).
UNIT 10.8 provides a series of methods for studying MMPs and their inhibitors, including
a live-cell collagen degradation assay, analysis by either direct or reverse zymography,
and enzyme-capture techniques.
The ECM of living organisms is often organized into a Þbrillar, three-dimensional matrix. Although collagen gels and Matrigel provide valuable three-dimensional cell culture
model systems, recent Þndings suggest that some cells require a matrix that more closely
mimics their natural microenvironment in vivo in order to form normal cell adhesions;
such matrices also enhance cell attachment, migration, and proliferation. UNIT 10.9 provides methods for generating a Þbroblast-derived three-dimensional ECM that closely
resembles in vivo matrices. It also describes protocols for assessing cell adhesive signaling and morphological responses to such matrices, as well as methods for generating
two-dimensional substrate controls.
Introduction
ECM proteins in a group termed ‘matricellular’ proteins play complex and interesting
roles in regulating the interaction of cells with other ECM molecules. These proteins
include SPARC (also known as osteonectin; UNIT 10.11), thrombospondin (UNIT 10.10),
tenascin, and other molecules that function to modulate a wide range of biological
responses to the ECM. UNIT 10.11 provides comprehensive protocols for isolating and
purifying SPARC, a matricellular protein that can regulate cell adhesion, proliferation,
and other processes. This unit presents methods for its puriÞcation from a cell line or
10.0.2
Supplement 45
Current Protocols in Cell Biology
platelets, as well as for purifying recombinant SPARC from E. coli or insect cells. Finally,
it also describes assays for its inhibitory effects on cell proliferation and adhesion.
Assembly of the extracellular matrix is currently thought to be a cell-directed process.
For example, even though collagens can self-polymerize, their patterns of deposition
are controlled by cells. The cell-mediated formation of a Þbronectin Þbrillar matrix is
a particularly dramatic example, since cells have to convert soluble Þbronectin dimers
into an insoluble Þbrillar matrix. UNIT 10.12 provides detailed methods for analyzing this
process of Þbronectin matrix assembly based on measuring conversion of this protein to
a detergent-insoluble form after polymerization. It describes the isolation and analysis
of the resultant insoluble Þbronectin matrix from cultured cells, as well as assays using
exogenously added Þbronectin or endogenous labeling. Because some laboratories may
prefer to avoid using radioactivity, UNIT 10.13 provides an alternative method of quantifying
Þbronectin matrix assembly using biotin-labeled Þbronectin. This protocol includes an
internal control to ensure equal recovery of detergent-insoluble material, as well as
methods for parallel analyses of signal transduction processes that are involved in matrix
assembly or result from this important process.
Cell interactions with three-dimensional matrices help to regulate cell growth, migration,
and differentiation in vivo. Researchers now have increasing numbers of options for natural and synthetic extracellular matrices that can mimic various types of three-dimensional
microenvironments. A new type of synthetic matrix termed Extracel is described in UNIT
10.14 that consists of hyaluronan cross-linked to gelatin. It forms a hydrogel that is an
effective culture system for both normal and malignant cells. This biodegradable scaffold
can be designed to deliver growth factors to cells, and it provides an excellent environment
for growth of tumor xenografts.
In order to study the roles of patterns and topology of matrix molecules, the newly
developed process termed micro photopatterning provides a simple and highly ßexible
methodology to generate any protein pattern on a substrate. As described in detail in UNIT
10.15, this new technique uses a standard two-photon confocal microscope controlled by
its built-in software to ablate any desired pattern in a polyvinyl alcohol thin Þlm. Any
protein can then be coated onto the exposed pattern, and the process can be repeated
multiple times to generate complex patterns of multiple proteins positioned as close as a
few micrometers from each other. Another advantage of this method compared to others
is that it readily permits ßuorescence and total internal reßection ßuorescence (TIRF)
microscopy.
The units in this chapter span topics from individual puriÞed ECM proteins to complex
three-dimensional matrices consisting of many molecules interacting to form gels. They
provide the opportunity to study the functions and mechanisms of the whole range of
types of cell-ECM interactions, which are now recognized to play crucial roles in cell
biology and pathology.
Kenneth M. Yamada
Extracellular
Matrix
10.0.3
Supplement 45
Overview of Extracellular Matrix
In all multicellular organisms, development
is influenced by the interactions between cells
and their extracellular matrix (ECM). Information contained in the ECM provides the cell
with temporal and positional clues, such as
where it is, where it should be going, how old
it is (in terms of cellular differentiation), and in
some instances, when it is time for it to die
(through apoptosis). It is not surprising, then,
that there has been a great deal of interest in
defining the extracellular signals as well as the
cell surface receptors that interact with these
molecules and interpret the information. Now
more than ever, understanding cell biology requires understanding the ECM.
Studying the ECM, however, is not for the
faint of heart! In most instances, the functional
form of matrix macromolecules is a large, sparingly soluble aggregate that cannot easily be
solubilized or dissociated into component units.
Even when dissociated matrix components are
obtained, the biological properties of the constituent chains often differ from the intact form.
To complicate matters, most ECM macromolecules participate in supramolecular assemblies
where their biological properties are modified
by the molecules with which they interact.
These unusual physical properties create serious problems for matrix characterization using a standard “wet chemistry” approach. They
also create some, though not many, unique
advantages. For example, the multimeric,
cross-linked nature of ECM imparts an element
of stability that is not found in other proteins.
This is most obvious if one takes a historical
look at techniques used for matrix purification
(Partridge, 1962; Piez, 1997). In the early days
of matrix biology, “connective tissue” was purified using extraction protocols that relied on
the ability of the target matrix component to
withstand relatively harsh conditions: acid solutions were used for purifying collagen, chaotropic agents for mucopolysaccharides (now
called proteoglycans), and, the harshest of them
all, boiling sodium hydroxide for purifying
elastin. It is quite remarkable that so much of
what we know about these three matrix classes
resulted from experiments using products purified in this way. Although purification strategies are now a bit more sophisticated, modifications of these basic protocols are still used
today. The use of molecular biology and mouse
genetics has quickened the pace of matrix char-
UNIT 10.1
acterization and opened the door to functional
studies of complex matrices that were unthinkable several years ago.
One of the most important properties of
ECM is its functional diversity (Kleinman,
1993). Some components are designed to be
rigid, others elastic; some wet, others sticky. All
have modular designs that impart diverse roles,
yet allow for highly specialized functions. The
formation of a basement membrane, for example, requires the assembly of ECM molecules
that have significant tensile strength (collagen),
can act as charged molecular sieves (proteoglycans), and facilitate cell attachment (laminin).
These molecules are woven together through
processes that involve self-assembly and interactions with molecules that are specifically
designed to serve as molecular bridges or linkers (nidogen/entactin; Yurchenco, 1994).
It is not possible to ascertain the functional
properties of a complex matrix such as basement membrane without studying its individual
components. At the same time, however, it is
also clear that the functional complexity of the
assembled basement membrane is greater than
the sum of its component parts. To comprehend
this greater sum requires shifting one’s view away
from a reductionist biochemical approach to
one focused on cell and developmental biology.
Here the cell becomes the reagent, interpreting
informational signals contained in the ECM
and adjusting its physiology accordingly. The
researcher’s task is to understand the readout.
The sections below contain an overview of
the major classes of ECM. Molecules have been
selected to illustrate specific functional or
structural properties that are common to a matrix class or to ECM macromolecules generally.
Where possible, recent reviews with references
to more detailed literature are cited. Although
somewhat dated, the text Cell Biology of Extracellular Matrix (Hay, 1991) provides an excellent overview of ECM biology. More detailed reviews can be found in various volumes
of the Biology of Extracellular Matrix series,
published by Academic Press.
COLLAGENS
Structure of Collagens
Collagen is the most ubiquitous ECM protein and is designed to provide structure and
resiliency to tissues. It is defined by the presExtracellular
Matrix
Contributed by Robert P. Mecham
Current Protocols in Cell Biology (1998) 10.1.1-10.1.14
Copyright © 1998 by John Wiley & Sons, Inc.
10.1.1
ence of a triple-helical domain containing peptide chains with repeating Gly-Xaa-Yaa triplets, and by the presence of hydroxyproline and
hydroxylysine (Kühn, 1987; Prockop and Kivirikko, 1995). To date, nineteen distinct genetic
collagen types have been identified. The characteristic molecular form of collagen is a triple
helix made up of three polypeptides, called α
chains, that coil into a right-handed triple helix.
Collagens exist either as homotrimers composed of three identical α chains (α1)3 or as
heterotrimers consisting of two ([α1]2α2) or
three (α1α2α3) α chains.
The nomenclature for the collagen superfamily consists of an indication of their genetic
type (a Roman numeral that generally denotes
the chronological order in which the collagens
were characterized) together with the α-chain
composition. Type I collagen, for example, is a
heterotrimer of two α1 chains and one α2 chain,
and is indicated as (α1[I])2α2(I). Type II collagen is a homotrimer of three α1 chains and is
written (α1[II])3. Other collagens consist of
three different α chains and (using type IX as
an ex ample) ar e written in the form
α1(IX)α2(IX)α3(IX). It is important to note
that each α chain within a collagen type is a
distinct gene product; that is, an α1 chain in one
collagen type is not the same protein as the α1
chain in any other collagen type. It is critical,
therefore, to indicate the collagen type when
referring to a particular α chain (e.g., the α1
chain of type I collagen).
Synthesis of Collagens
Collagen α chains are synthesized on membrane-bound ribosomes (ER) as large precur-
sors, called pre-pro-α chains. In addition to the
signal peptide (the “pre” part of the name)
required for transport into the ER, each collagen precursor has extension peptides (the “pro”
part) on both its N- and C-terminal ends (Fig.
10.1.1). Each pro-α chain combines with two
others in the lumen of the ER to form the
triple-helical molecule. The extension peptides
are required for correct triple helix formation
and remain with the triple-helical unit throughout the secretory pathway.
In the triple helix, the side chain of every
third α-chain residue is directed towards the
center of the helix, shifted by 30° from the
preceding central residue of the same chain
(Brodsky and Ramshaw, 1997). Steric constraints dictate that the center of the helix be
occupied only by glycine residues; side chains
of any other amino acid would perturb the
triple-helical conformation.
Hydroxylation of proline residues in the Yaa
position occurs as a post-translational modification in the lumen of the ER. The side-chain
hydroxyl group of hydroxyproline stabilizes
the helix through the formation of intermolecular hydrogen bonds. In fact, hydroxylation of
∼100 prolyl residues is essential for the three
pro-α chains of fibrillar collagens to form a
triple helix that is stable at body temperatures.
Hydroxylation of α chains is catalyzed by prolyl 4-hydroxylase, a tetrameric enzyme consisting of two α and two β subunits (α2β2; Kivirikko and Myllyharju, 1998). Interestingly, the
β subunit is protein disulfide isomerase, an ER
protein that catalyzes thiol-disulfide interchange during protein folding (Koivu et al.,
1987). The hydroxylation reaction catalyzed by
N-propeptide
C-propeptide
N-telopeptide
C- telopeptide
NH2
terminal
Triple-helical doman (Gly-Xaa-Yaa)n
Overview of
Extracellular
Matrix
α1
α 2 COOH
α1 terminal
Figure 10.1.1 Functional domains of the type I procollagen molecule. Following cleavage of the
propeptide domains in the extracellular space, collagen units assemble in a quarter-stagger
arrangement to form a fibril.
10.1.2
prolyl 4-hydroxylase requires Fe2+, 2-oxoglutarate, O2, and ascorbate. Conditions that prevent proline hydroxylation (such as nutritional
deficiency of iron or of vitamin C) affect helix
formation or stability. In scurvy, a human disease caused by a dietary deficiency of vitamin
C, the nonhydroxylated pro-α chains are unstable and the skin and blood vessels become
extremely fragile.
A second post-translational modification of
procollagen that is crucial to its function is the
hydroxylation of lysine residues. This reaction,
which also occurs in the ER, is catalyzed by the
enzyme lysyl hydroxylase. The active enzyme
is a homodimer and, like prolyl hydroxylase,
requires Fe2+, 2-oxoglutarate, O2, and ascorbate. Hydroxylysine residues have two important functions: their hydroxy groups act as attachment sites for carbohydrate units, and they
are essential for the stability of the intermolecular collagen cross-links that occur in the extracellular space after secretion. The glycosylation of hydroxylysine is unusual, consisting
of a single galactose residue or a glucosylgalactosyl disaccharide attached to the hydroxyl group. The amount of carbohydrate
added to procollagen varies greatly among different types of collagen, and its function is
unknown.
Assembly of Collagens
After secretion into the extracellular space,
the extension peptides of procollagen are removed by specific proteolytic enzymes. Both
the N and C proteinases are members of the zinc
metallopeptidase family and contain domains
that suggest the ability to interact with cells and
other matrix components (Kessler et al., 1996;
Colige et al., 1997). Removal of the extension
peptides converts the procollagen molecules to
collagen (once called tropocollagen). Triplehelical collagen units then come together in the
extracellular space to form the much larger
collagen fibrils. The process of fibril formation
is driven, in part, by the tendency of the collagen
molecules to self-assemble. The fibrils form
close to the cell surface, however, and it seems
likely that the cell regulates the sites and rates
of fibril assembly. The nonfibrillar collagens
(see below) undergo only limited proteolytic
processing prior to assembly. Here it is important to distinguish between collagen and gelatin. As stated above, collagen is the triple-helical form of the protein and can exist as single
triple-helical units or triple-helical units polymerized into fibrils. Gelatin is denatured collagen. The individual α chains are no longer in
a triple helix but can nevertheless polymerize
into a random gel under appropriate conditions
of temperature and ionic strength.
Collagen fibrils are greatly strengthened by
covalent cross-links within and between the
constituent collagen molecules. The types of
covalent bonds involved are unique to connective tissue and are formed through deamination
of certain lysine and hydroxylysine residues to
yield highly reactive aldehyde groups. The aldehydes then undergo classical condensation
reactions to form covalent bonds with each
other or with other lysine or hydroxylysine
residues. The extent and type of cross-linking
varies from tissue to tissue, depending on tissue
requirements. For example, collagen is highly
cross-linked in tendons, where tensile strength
is crucial. Lysyl oxidase, the enzyme that catalyzes cross-link formation, requires copper and
molecular oxygen. If cross-linking is inhibited,
collagenous tissues become fragile, and structures such as skin, tendons, and blood vessels
tend to tear.
Collagen Classification
The polymeric structures formed by members of the collagen family vary depending on
collagen type (Prockop and Kivirikko, 1995).
The structures formed result, in large part, from
the nontriple-helical “modules” found within
many of the nonfibrillar collagens (Brown and
Timpl, 1995). Based on structural similarities,
the collagen superfamily can be divided into
the following classes.
Fibril-forming collagens: types I, II, III, V,
and XI. These collagens (Kühn, 1987; Kadler,
1994) all share a long triple-helical segment
with a continuous Gly-Xaa-Yaa repeat over its
entire length. They assemble into cross-striated
fibers upon cleavage of N and C propeptides,
with the individual units adopting a one-quarter
stagger relative to their neighbors in the fibril.
Types II and XI collagen undergo alternative
splicing, and hybrid molecules containing both
types V and XI collagen have been identified
in some tissues.
Network-forming collagens: types IV, VIII,
and X. α chains in the type IV collagen family
(Hulmes, 1992; Kühn, 1994; Yurchenco, 1994)
contain a large collagenous domain that is frequently interrupted by short noncollagenous
sequences (i.e., something other than Gly-XaaYaa). Noncollagenous domains are also found
at the N and C termini of the chain, with the
C-terminal domain being the larger of the two.
Monomers associate at the C termini to form
dimers and at the N termini to form tetramers.
Extracellular
Matrix
10.1.3
Overview of
Extracellular
Matrix
The triple-helical domains intertwine to form
supercoiled structures, resulting in a net-like
structure. Type VIII collagen is found in Descemet’s membrane in the eye and forms a stack
of hexagonal lattices. A similar structure is
formed by type X collagen synthesized by hypertrophic chondrocytes in the deep-calcifying
zone of cartilage.
Fibril-associated collagens with interrupted triple helices (FACIT): types IX, XII,
XIV, XVI, and XIX. These collagens (Mayne
and Brewton, 1993; Olsen et al., 1995) are
characterized by short triple-helical segments
interrupted by short noncollagenous domains.
They attach to the surface of fibril-forming
collagens and do not form fibrils themselves.
Type IX collagen is found on the surface of type
II collagen, to which it is covalently bound. An
unusual property of this collagen is the presence
of a glycosaminoglycan (GAG) chain attached
to a noncollagenous domain of the α2(IX)
chain. Types XII and XIV collagen show structural similarities to type IX, including an attached GAG side chain. Types XVI and XIX
have not been fully characterized but show
similarities in structure to other members of the
family.
Beaded filaments and anchoring fibrils:
types VI and VII. Among the collagens of this
family (Burgeson, 1993; Timpl and Chu, 1994),
type VI collagen is characterized by α chains
containing large N- and C-terminal globular
domains separated by a small triple-helical segment. Alternative splicing produces variants of
the α2(VI) and α3(VI) chains. Type VI collagen
forms small beaded filaments in the ECM. Type
VII collagen forms anchoring fibrils that link
basement membranes to anchoring plaques of
type IV collagen and laminin in the underlying
ECM. Type VII collagen contains the longest
triple helix of any known collagen, with only
small interruptions throughout. The NC1 domain of type VII collagen binds to collagen
types I and IV, fibronectin, and laminin 5.
Collagens with a transmembrane domain:
types XIII and XVII. Types XIII and XVII
collagen (Li et al., 1996) are unique in having
a transmembrane domain with its N terminus
predicted to be in the cytoplasm. Type XIII
collagen undergoes extensive alternative splicing. Type XVII collagen is found primarily in
the hemidesmosomes of the skin and is one of
the antigens that produces the autoimmune disease bullous pemphigoid.
Other nonfibrillar collagens: types XV and
XVIII. Types XV and XVIII collagen (Rehn
and Pihlajaniemi, 1994) have large N- and C-
terminal globular domains and a highly interrupted triple helix. Their large number of potential N- and O-linked glycosylation sites suggests that both types have the potential to be
highly glycosylated.
ELASTIN AND MICROFIBRILLAR
PROTEINS
Elastin
During evolution, with the advent of the
closed circulatory system, came the requirement for blood vessels to accommodate the
pulsatile blood flow of the heart. Vessels made
mostly of collagen were too stiff, so in its place,
we see the emergence of a matrix protein that
has the properties of elastic recoil. This protein,
elastin, is the predominant protein component
of the elastic fiber that is of particular importance to the structural integrity and function of
tissues in which reversible extensibility or deformability are crucial, such as the major arterial vessels, lungs, and skin.
In contrast to the genetic diversity evident
in the collagen gene family, elastin is encoded
by only one gene. Like collagen, elastin maturation in the ECM involves the assembly of a
soluble precursor molecule (tropoelastin) into
a highly cross-linked polymer. This assembly
process is more complex than that for collagen,
however, because the ability to self-assemble
does not appear to be an intrinsic property of
tropoelastin. Instead, elastin assembly requires
helper proteins to align the multiple cross-linking sites on elastin monomers (Mecham and
Davis, 1994).
Two functional domains repeat along the
tropoelastin molecule (Fig. 10.1.2). One domain, related to cross-link formation, is an α
helix containing alanine and lysine. The other,
related to extensibility, is enriched in glycine,
valine, and proline. The hydrophobic amino
acids in this domain are arranged in repeating
sequences that form a succession of β turns.
The stacked β turns form a β spiral with a
hydrophobic core. Stretching the elastin polymer exposes the hydrophobic core to water.
Recoil occurs when the leaves of the β spiral
contract to shield the hydrophobic amino acids
from the aqueous microenvironment. Mature,
cross-linked elastin is extremely hydrophobic
and insoluble under most conditions (including
when boiled in sodium hydroxide; Partridge,
1962). Its unusual physical properties make
insoluble elastin one of the most stable proteins
in the body—lasting the lifetime of the organism. Two polyfunctional cross-links, desmos-
10.1.4
tropoelastin
NH2
* * *
hydrophobic
domain
K-Ptype
cross-linking
domain
*
**
- COOH
K-A type
cross-linking
domain
*alternatively spliced domains
fibrillin-1
RGD
P
COOH
NH2
fibrillin-2
RGD
RGD
G
COOH
NH2
eight-Cys domain (CCC)
calcium-binding EGF-like domain
eight-Cys domain (CC)
EGF-like domain
nine-Cys domain (CC)
potential glycosylation site
G
Gly-rich domain
P
Pro-rich domain
Figure 10.1.2 Domain map of tropoelastin and the fibrillins. Tropoelastin is secreted as a peptide
of ∼70 kDa and undergoes extensive covalent cross-linking during incorporation into the elastic
fiber. Fibrillin-1 and fibrillin-2 each have a molecular weight of ∼350 kDa and are the major structural
elements of 10- to 12-nm-diameter microfibrils. Abbreviations: K-A, alanine-rich cross-linking
domain; K-P, proline-rich cross-linking domain; RGD, Arg-Gly-Asp; CC, Cys-Cys sequences; CCC,
Cys-Cys-Cys sequences; EGF, epidermal growth factor.
ine and isodesmosine, are unique to elastin and
can be used as specific markers for this protein.
Fibrillin
Microfibrils were first identified as components of elastic fibers. They are found in greatest abundance in elastic tissues or in the ciliary
zonules of the eye, although their distribution
is widespread. Fibrillin-1 and -2 play key roles
in microfibrillar architecture. These 350-kDa
glycoproteins are highly homologous (Fig.
10.1.2), with modular structures consisting of
repeating calcium-binding epidermal growth
factor (EGF)–like domains interspersed between 8-cysteine domains similar to those
found in the latent transforming growth factorβ (TGF-β)–binding protein family (Lee et al.,
1991). Tandemly arranged EGF domains form
a structural motif found frequently in ECM
macromolecules (e.g., laminin, fibulin, latent
TGF-β-binding protein, nidogen). When
stacked together, these tightly folded, disulfidebonded loop structures form a rigid, rod-like
arrangement stabilized by interdomain calcium
binding and hydrophobic interactions (Downing et al., 1996). The precise function of microfibrils is unclear, although their association
with developing elastic fibers suggests a role in
elastin assembly. Both fibrillin-1 and fibrillin-2
interact with the αvβ3 integrin through an ArgGly-Asp (RGD) sequence (see Adhesive Glycoproteins).
ADHESIVE GLYCOPROTEINS
Most, if not all, ECM macromolecules interact with binding proteins on the surface of
cells. In many instances, this is through a
unique sequence motif that is accessible as part
Extracellular
Matrix
10.1.5
N-terminal domain of fibronectin also mediates
fibronectin’s binding to gram-positive bacteria
through type I modules. The type I module
contains ∼45 amino acids with four cysteines
forming two disulfide bonds (Potts and Campbell, 1994). This module has also been found
in a number of other proteins. In addition to
type I repeats, the collagen-binding domain
contains the only type II repeats found in fibronectin. Like type I repeats, these motifs
contain two disulfide bonds, but they are larger
than type I motifs.
The predominant structural feature of fibronectin consists of type III repeats, accounting for more than 60% of the sequence. No
disulfide bonds are present in this structure,
although two of the repeats contain a free cysteine. The cell-binding RGD sequence is located in the tenth type III repeat. This sequence
is recognized by many members of the integrin
family, including α5β1, αvβ1, αvβ3, αvβ5,
αvβ6, αIIbβ3, and α8β1. Other cell-binding
regions include the C-terminal heparin-binding
domain and the type III–connecting segment
(IIICS), including the CS1 region. The type III
consensus sequence is frequently found in other
proteins.
Only one gene for fibronectin has been identified, but mRNAs for fibronectin have been
shown to give rise to multiple versions of the
protein through variable patterns of RNA splicing during gene transcription. Alternative splicing occurs predominately at three sites, termed
extra type III domain A (EDA or EIIIA), extra
type III domain B (EDB or EIIIB), and the
of the protein’s folded functional structure, or
cryptic and exposed only when the protein
undergoes a conformational change induced by
binding to another protein or as the result of
degradation or denaturation. One such “recognition motif” is the well-known RGD sequence
that is recognized by several members of the
integrin family.
Fibronectin
A great deal of biochemical work has led to
a model of the fibronectin molecule in which
the protein’s binding functions and its structure
are clearly correlated (Hynes, 1990). The molecule is secreted as a dimer consisting of two
similar subunits joined together at the C terminus by disulfide bonds (Fig. 10.1.3). Each chain
has a molecular weight of ∼220 to 250 kDa and
is subdivided into a series of tightly folded
domains. Each domain is responsible for one
of fibronectin’s binding functions. In plasma,
fibronectin exists as a soluble dimer, but in the
ECM it is found as an insoluble multimer.
Amino acid sequence analysis of fibronectin
shows that the molecule is made up mostly of
three repeating motifs, referred to as types I, II,
and III repeats. These repeats are organized into
functional domains that contain binding sites
for ECM proteins and cell surface receptors
(see Fig. 10.1.3). For example, there are two
fibrin-binding domains consisting of multiple
type I repeats on each subunit of the protein.
Type I repeats are also found in the collagenbinding domain, and the first five type I repeats
play an important role in matrix assembly. The
collagen,
gelatin
NH2
f1
f2
heparin, fibrin,
matrix assembly,
S. aureus binding
f1
f1
f3
EDB
f3
EDA
f3
RGD
f3
matrix
assembly
f2
type 1 repeat
(∼45 aa)
Overview of
Extracellular
Matrix
cell binding
cell binding
f3
type 2 repeat
(∼60 aa)
f3
IIICS
f1
f3 f3
heparin,
CS-PG
COOH
S S
fibrin
S S
COOH
f3
type 3 repeat
(∼90 aa)
alternatively spliced
Figure 10.1.3 Domain map of fibronectin. The subunits of fibronectin vary in size between ∼235
and 270 kDa. Alternative splicing occurs at three positions: EDA, EDB, and IIICS. Binding sites for
other molecules and cells are indicated. Abbreviations: EDA, extra type III domain A; EDB, extra
type III domain B; IIICS, connecting segment between the fourteenth and fifteenth type III repeats;
RGD, Arg-Gly-Asp; CS-PG, chondroitin sulfate proteoglycan; aa, amino acids.
10.1.6
connecting segment between the fourteenth
and fifteenth type III repeats (IIICS or V).
Splicing within the IIICS segment produces
five variants, such that twenty different fibronectin subunits can result from splicing
within the three segments.
Subunits of plasma fibronectin produced by
adult hepatocytes contain neither EDA nor
EDB segments, and one subunit lacks the entire
IIICS domain. Cultured fibroblasts, however,
typically produce a form of fibronectin, referred to as cellular fibronectin, that contains
the EDA and/or EDB segments. Fibronectins
expressed in fetal and tumor tissues contain a
greater percentage of EDA and EDB segments
than those expressed in normal adult tissues.
The biological functions of fibronectin isoforms are only poorly understood, despite having been studied extensively. Differences in
solubility have been demonstrated, but it has
been difficult to detect functional differences
between plasma and cellular fibronectin in their
ability to promote cell adhesion and spreading.
Vitronectin
Vitronectin (also called serum spreading
factor, S-protein, and epibolin) is a multifunctional protein found in plasma and ECM. It is
synthesized as a single chain that undergoes N
glycosylation, tyrosine sulfation, and phosphorylation prior to secretion. In plasma, vitronectin circulates in two forms: a single chain of
∼75 kDa and a proteolytically cleaved, twochain form that dissociates into 65- and 10-kDa
fragments upon reduction. It is present in fibrillar form in the ECM of a variety of tissues,
where it sometimes colocalizes with fibronectin and elastic fibers. While little vitronectin
immunoreactivity is detectable in most normal
tissues, increased deposition has been observed
in areas of tissue injury and necrosis. Tissue
vitronectin was believed to be plasma derived, but
recent studies indicate that extrahepatic cells
have the biosynthetic potential to produce
vitronectin and that its synthesis can be regulated
under inflammatory conditions (Seiffert, 1997).
The cell attachment activity of vitronectin
results from an RGD sequence that is recognized by a wide variety of integrins. Most of
the cell adhesive activity of serum used for
tissue culture can be attributed to vitronectin.
Laminin and Basement Membranes
Like fibronectin, the laminins are modular
proteins with domains that interact with both
cells and ECM (Ekblom and Timpl, 1996).
They constitute a family of basement mem-
brane glycoproteins that affect cell proliferation, migration, and differentiation. Eleven different laminins have been identified, each containing an α, β, and γ chain (Fig. 10.1.4).
Electron microscopy has revealed that all laminins have a cross-like shape with three short
arms and one rod-like long arm, a shape well
suited for mediating interactions between sites
on cells and components of the ECM (Beck et
al., 1990; Maurer and Engel, 1996). The rodlike regions separating the globular units of the
short arms are made up of repeating EGF-like
domains. The long arm is formed by all three
component chains folding into an α-helical
coiled-coil structure, and is the only domain
composed of multiple chains. It is terminated
by a large globular domain composed of five
homologous subdomains formed by the C-terminal region of the α chain.
Along with type IV collagen, laminins are a
major structural element of the basal lamina
(Timpl, 1996). The molecular architecture of
these matrices results from specific binding
interactions among the various components.
The structural skeleton is formed by type IV
collagen chains that assemble into a covalently
stabilized polygonal network. Laminin self-assembles through terminal domain interactions
to form a second polymer network. Nidogen
(Mayer and Timpl, 1994) binds laminin near its
center and interacts with type IV collagen,
bridging the two. A large heparan sulfate proteoglycan (HS-PG), perlecan, binds laminin
and type IV collagen through its GAG chains
and forms dimers and oligomers through a
core-protein interaction. Perlecan is important
for charge-dependent molecular sieving, one of
the critical functions of basement membrane.
Other components that are sometimes found
associated with basement membranes but may
not be intrinsic components include fibronectin, type V collagen, fibulin, osteonectin
(also known as BM-40 or SPARC), and chondroitin sulfate proteoglycans.
Cells attach to laminin through specific interaction sites created by its multidomain structure. For example, sites for receptor-mediated
cell attachment and promotion of neurite outgrowth reside in the terminal region of the long
arm. A second cell-attachment site and a cellsignaling site with mitogenic action are localized in the short arms. Cell binding to laminin
occurs via a variety of receptors, including
non-integrins (Mecham and Hinek, 1996) and
integrins (Aumailley et al., 1996). The β1 family includes most of the laminin-binding integrins (α1β1, α2β1, α3β1, α7β1, α9β1). Other
Extracellular
Matrix
10.1.7
LEs
LEs
NH2- LN
LEs
L4
cc
L4
LEs
NH2-
E
E
cc
NH2LEs
E
LEs
LEs
cc
L4
LN
- COOH α1, α2
LGs
- COOH α3
LGs
- COOH α4
LGs
- COOH α5
cc
L4
E
LGs
NH2
LEs
NH2- LN
LEs
G
LEs
NH2- LN
LEs
NH2- LN
cc
- COOH β1, β2
cc
cc
- COOHβ3
LEs
L4
LEs
NH2-
cc
LEs
L4
laminin-1 α1β1γ1
laminin-2 α2β1γ1
α
laminin-3 α1β2γ1
laminin-4 α2β2γ1
γ laminin-5 α3β3γ2
β
laminin-6 α3β1γ1
laminin-7 α3β2γ1
laminin-8 α4β1γ1
laminin-1 laminin-9 α4β2γ1
laminin-10α5β1γ1
laminin-11α5β2γ1
cc
- COOH γ1
cc
- COOH γ2
LN N-terminal domain VI LG G-domain
L4 domain IV
structure undefined
LE EGF-like domain
CC coiled-coil domain
Figure 10.1.4 Domain map of laminin chains. Three polypeptide chains (α, β, and γ) form the
laminin cross. The chain composition of known laminin types is shown in the insert. Abbreviation:
EGF, epidermal growth factor.
Overview of
Extracellular
Matrix
integrins that bind laminin include αvβ3 and
α6β4. Basement membrane can also have an
indirect effect on cells by binding and sequestering growth and differentiation factors, such
as fibroblast growth factor (FGF), platelet-derived growth factor (PDGF), and TGF-β.
The importance of laminin to cell differentiation and migration has been demonstrated in
developmental studies. Isoforms of laminin assembled from different chains are focally and
transiently expressed and may serve distinct
functions at early stages of development even
before being deposited as components of basement membranes. Laminin is present at the
two-cell stage in the mouse embryo, making it
one of the first ECM proteins detected during
embryogenesis.
MATRICELLULAR PROTEINS
The term “matricellular” has been applied
to a group of extracellular proteins that function
by binding to matrix proteins and to cell surface
receptors, but do not contribute to the structural
integrity of the ECM (Bornstein, 1995). Proposed members of this group include the thrombospondins, members of the tenascin protein
family, SPARC/osteonectin (Lane and Sage,
1994), and osteopontin. These proteins are frequently called “antiadhesive proteins” because
of their ability to induce rounding and partial
detachment of some cells in vitro (Sage and
Bornstein, 1991). Their ability to interact with
many different matrix proteins and cell surface
receptors may explain their complex range of
biological functions.
10.1.8
Thrombospondin
Tenascin
The thrombospondin (TSP) family consists
of five secreted glycoproteins (Adams et al.,
1995). TSP-1 and TSP-2 have identical domain
structures and are secreted as disulfide-bonded
homotrimers (Fig. 10.1.5). TSP-3, TSP-4, and
TSP-5/COMP (cartilage oligomeric matrix
protein) are pentamers whose expression is
more limited than that of TSP-1 and TSP-2.
TSP-1 binds HS-PGs, various integrins, the
integrin-associated protein, and CD36. It also
binds plasminogen, fibrinogen, fibronectin,
urokinase, and TGF-β (which it can also activate). TSP-1 exhibits variable effects on cell
adhesion and cell proliferation (Bornstein,
1995). For example, TSP-1 promotes proliferation of vascular smooth muscle cells, yet inhibits proliferation of endothelial cells. It supports
attachment and spreading of skeletal myoblasts
but expresses antiadhesive activity toward endothelial cells. Thrombospondin is the most
abundant protein of platelet alpha granules and
is released when platelets are activated.
The tenascins constitute a gene family consisting of four members: tenascins-C, -R, -X,
and -Y (Erickson, 1993; Chiquet-Ehrismann,
1995). Tenascin-C (early names include
GMEM, cytotactin, J1, hexabrachion, and
neuronectin) was the first form discovered and
exists as a hexamer of disulfide-bonded subunits. Each subunit consists of a cysteine-rich
N-terminal domain involved in oligomerization, EGF-like repeats, fibronectin type III–like
repeats, and a fibrinogen-like globular domain
(Fig. 10.1.5). The number of fibronectin type
III–like repeats varies as a result of alternative
splicing. Like TSP, tenascin-C has diverse biological effects when applied to cells. Both stimulation and inhibition of cellular proliferation
have been observed in response to tenascin-C.
In terms of cell adhesion, some cells do attach
to tenascin, but weakly. In most instances, tenascin does not allow cell adhesion and can even
inhibit cell attachment to other matrix proteins
such as fibronectin and laminin. The finding
that tenascin-C contains defined cell attach-
thrombospondin-1
NH2
NH2
PC
I
I
I
III
III
III
III
III
III
COOH
COOH
RGD
TGF-β
collagen IV
CD36
decorin
syndecans
HS-PG/CS-PG
sulfatides
PC
procollagen homology
I
type I thrombospondin
or properdin repeat
integrin
binding
Type II (EGF-like) repeat
III
type III calciumbinding repeat
tenascin-C
f3 f3 f3 f3 f3 f3 f3 f3 f3 f3 f3 f3 f3 f3 f3
NH2
COOH
EGF-like domain
f3 fibronectin type 3 domain
fibrinogen-like domain
f3 fibronectin type 3 domain
(alternatively spliced)
Figure 10.1.5 Domain map of thrombospondin-1 and tenascin-C. The functional form of thrombospondin is a homotrimer of ∼420 kDa. It interacts with numerous matrix proteins and modulates
cell attachment by interacting with various cell-surface receptors. Tenascin-C monomers form a
hexameric structure joined at their N termini by disulfide bonds. Alternative splicing leads to subunits
of differing molecular weights. Abbreviations: HS-PG, heparan sulfate proteoglycan; CS-PG,
chondroitin sulfate proteoglycan; TGF-β, transforming growth factor-β; RGD, Arg-Gly-Asp; EGF,
epidermal growth factor.
Extracellular
Matrix
10.1.9
ment sites suggests that the overall antiadhesive
properties of the glycoprotein are effected by
separate domains that override the attachment
domains.
PROTEOGLYCANS
The proteoglycans (once called acid mucopolysaccharides) constitute a number of genetically unrelated families of multidomain
proteins that have covalently attached GAG
chains. To date, more than 25 distinct gene
products have been identified that carry at least
one GAG chain (Iozzo and Murdoch, 1996).
Like other matrix components discussed in this
review, proteoglycans exist as structural variants, further increasing their functional and
structural diversity.
For historical reasons, proteoglycans are
named based on the type of attached GAG
chain(s): (1) chondroitin sulfate and dermatan
sulfate, consisting of a repeating disaccharide
of galactosamine and either glucuronic acid or
iduronic acid; (2) heparin and heparan sulfate,
consisting of a repeating disaccharide of glucosamine and either glucuronic acid or iduronic
acid; and (3) keratan sulfate, consisting of a
repeating disaccharide of glucosamine and galactose. Hyaluronate is also a repeating disaccharide but is not sulfated and not bound to a
core protein. GAG chains are usually attached
through O-glycosidic linkages to serine residues in the proteoglycan core protein. A characteristic feature of GAG chains is that at physiological pH they contain one to three negative
charges per disaccharide due to carboxylate and
sulfate groups.
Knowledge of the structure and function of
proteoglycans increased dramatically when
molecular biology was used to study the core
proteins (Hassell et al., 1993). The heterogeneity of this family of matrix proteins also became
evident with the finding that there are no structural domains common to all proteoglycans.
There are, however, distinguishing characteristics that allow them to be grouped into four
broad categories.
Large Proteoglycans that Form
Aggregates by Interaction with
Hyaluronan
Overview of
Extracellular
Matrix
These proteoglycans interact with strands of
hyaluronate to form a very-high-molecularweight aggregate. A structural trait shared by
these proteoglycans is the presence of three
functional domains: a globular hyaluronanbinding domain at the N terminus, a central
extended region that carries most of the GAG
chains, and a modular C-terminal domain containing two EGF repeats, a C-type lectin domain, and a complement-regulatory-proteinlike motif (Iozzo and Murdoch, 1996).
The largest member of this family is versican
(Zimmermann and Ruoslahti, 1989), a major
proteoglycan in blood vessels that is also expressed in nonvascular tissues. Aggrecan, the
large aggregating proteoglycan of cartilage, has
a smaller core protein than versican but contains nearly 3-fold more GAG chains (Fig.
10.1.6). The high charge density of aggrecan
results in each monomer occupying a large
hydrodynamic volume. Aggrecan’s GAG
chains result in a high density of fixed charge
in cartilage, producing an osmotic swelling
pressure that is balanced by tension in the
collagenous network. The reversible redistribution of proteoglycan-bound water under loading gives cartilage the ability to absorb compressive loads (Wight et al., 1991). Two other
members of this family include neurocan
(Rauch et al., 1992) and brevican (Yamada et
al., 1994), both found in brain tissues.
Basement Membrane Proteoglycans
HS-PGs appear to be ubiquitous components of all basement membranes. Perlecan is
the largest basement membrane proteoglycan,
with a modular core protein of 467 kDa (Fig.
10.1.6; Iozzo et al., 1994). It provides the basement membrane with a negative charge that is
important to its sieving properties. The heparan
sulfate chains of perlecan also bind growth
factors and cytokines and sequester them into
the basement membrane, where they may function as a reserve to be released during tissue
repair. The interaction of heparan sulfate with
the FGFs has been extensively studied (Aviezer
et al., 1994). Perlecan interacts with other components of the basement membrane, particularly laminin and nidogen. The multidomain
structure of perlecan core protein is reminiscent
of other ECM proteins, and includes EGF repeats and repeats of structures found in the
low-density lipoprotein receptor, laminin
chains, and neural cell adhesion molecule.
Agrin was originally isolated from torpedo
ray electric organ and was found to induce
acetylcholine receptor aggregation. It is secreted by motor neurons and deposited in the
synaptic cleft basement membrane. Agrin may
also play a role in the sequestration of growth
factors in the basement membrane. Like perlecan, agrin is a multidomain protein with regions
of EGF and laminin G-domain homology.
Agrin is found predominantly in the brain, but
10.1.10
aggrecan
HABR
Lec
G
KS
NH2
COOH
CS
perlecan
HS
LEs
NH2-
L4
LEs
L4
LEs
LG LG LG
L4
LEs
COOH
HS
HABR
G
HA-binding Lec lectin domain
domain
globular
domain
LG
laminin
G-domain
immunoglobulin
superfamily
LE
laminin
EGF-like
LDL receptorlike domain
LE
complement
regulatory-like
Figure 10.1.6 Domain map of two representative large proteoglycans. Aggrecan is the core
protein of the aggregating proteoglycan found in cartilaginous tissues. The molecular weight of the
aggrecan core protein is 210 to 250 kDa. There are 100 to 150 keratan sulfate chains and many
more chondroitin sulfate chains that contribute to the 2500-kDa molecular weight of the mature
proteoglycan. The glycosaminoglycans are attached to repetitive sequences in the middle two-thirds
of the molecule, including several types of repeats containing Ser-Gly, the linkage site for chondroitin
sulfate. Perlecan is the largest of the basement membrane proteoglycans and has two or three
attached heparan sulfate side chains. Removal of heparan sulfate side chains by heparatinase
produces a core protein of 400 to 450 kDa on SDS-PAGE. Abbreviations: HA, hyaluronate; KS,
keratan sulfate; CS, chondroitin sulfate; HS, heparan sulfate; LDL, low-density lipoprotein; EGF,
epidermal growth factor.
has also been localized to smooth and cardiac
muscle.
Cell Surface Heparan Sulfate
Proteoglycans
HS-PGs on the cell surface influence several
important biological functions, including cell
adhesion; the sequestration of heparin-binding
ligands on the plasma membrane; and the promotion of dimerization/oligomerization of
bound ligands, which enhances activation of
primary signaling receptors.
Cell-associated HS-PGs have been divided
into two major families, syndecan-like integral
membrane HS-PGs (SLIPs) and glypican-related integral membrane HS-PGs (GRIPs;
David, 1993; Carey, 1997). The SLIPs are
transmembrane HS-PGs with a conserved intracellular domain that likely interacts with
cytoskeletal and regulatory proteins. The
GRIPs are linked to the cell surface by glycosyl
phosphatidyl inositol in the outer leaflet of the
lipid bilayer.
The syndecans, the principal form of cellsurface HS-PG, are synthesized by many cells.
Syndecans bind a variety of extracellular ligands via their covalently attached heparan
sulfate chains and are thought to play important
roles in cell-matrix and cell-cell adhesion, migration, and proliferation. To date, four homologous syndecan core proteins have been
cloned from vertebrate cells. All syndecans are
type I transmembrane proteins, with an N-terminal signal peptide, an ectodomain that contains several consensus sequences for GAG
attachment, a single hydrophobic transmembrane domain, and a short C-terminal cytoplasmic domain. The majority of GAG chains
added to syndecan core proteins are of the
heparan sulfate type, although syndecan-1 and
syndecan-4 have chondroitin sulfate chains attached as well. Syndecans act as cell surface
Extracellular
Matrix
10.1.11
receptors for a number of matrix molecules,
thereby mediating cell attachment and tissue
organization. They influence the interactions of
basic FGF and other growth factors with their
receptors on cells and are responsible for the
maintenance of a nonthrombogenic surface on
endothelial cells.
Small Leucine-Rich Proteoglycans
Small leucine-rich proteoglycans (SLRPs)
comprise a class of secreted proteoglycans that
include five structurally related members:
decorin, biglycan, fibromodulin, lumican, and
epiphycan (see Fig.10.1.7). Each has a leucinerich core protein that assumes an arch-shaped
structure with a concave surface capable of
interacting with various other proteins. The
N-terminal region contains one (decorin) or
two (biglycan and epiphycan) GAG chains that
can be either dermatan or chondroitin sulfate.
Instead of GAG chains, fibromodulin and lumican have tyrosine sulfate in the N terminus,
which provides an analogous negatively
charged domain. These two SLRPs also contain
N-linked keratan sulfate chains in their central
domain.
SLRPs interact with numerous ECM proteins (e.g., fibronectin, TSP, fibrillin, microfibril-associated glycoprotein) and act to orient
and order collagen fibers during development
decorin
CS/DS
CHO CHO CHO
NH2
COOH
biglycan
CS/DS
CHO CHO
NH2
COOH
fibromodulin
NH2
CHO
COOH
N-linked carbohydrate
chondroitin sulfate or
dermatan sulfate GAG
tyrosine sulfate
keratan sulfate GAG
leucine-rich domain
Overview of
Extracellular
Matrix
Figure 10.1.7 Domain map of representative members of the small leucine-rich proteoglycans.
Decorin contains a single chondroitin or dermatan sulfate chain attached near the N terminus. The
core protein is ∼38 kDa. Decorin is heterogeneous with respect to glycosaminoglycan (GAG) chain
size, such that the secreted proteoglycan shows a range of molecular weights centered between
100 and 250 kDa. The core protein of biglycan is similar in size to that of decorin, except biglycan
contains two chondroitin or dermatan sulfate chains. The GAG chains are also heterogeneous in
size, resulting in a broad band on SDS-PAGE centered anywhere from 200 to 350 kDa. Removal
of GAG chains with chondroitin ABC-lyase results in a 45-kDa band. Fibromodulin has a core protein
size of 42 kDa. Four of the five potential N-glycosylation sites in the leucine-rich region of the
molecule are substituted with keratan sulfate chains. Five to seven closely spaced tyrosine sulfate
residues are found in the N-terminal domain. Abbreviations: CS, chondroitin sulfate; DS, dermatan
sulfate.
10.1.12
and tissue remodeling. Interactions with matrix
proteins occur through the leucine-rich core
which, in the case of type I collagen, influences
collagen fibrillogenesis by binding to the surface of the collagen fibril at the d-band with the
highly charged GAG chain extending out to
regulate interfibrillar distances. Like other proteoglycans, SLRPs bind to growth factors (e.g.,
TGF-β) and thereby likely influence cellular
differentiation and matrix synthesis. Decorin
has recently been shown to directly regulate cell
growth by activating the EGF receptor (Moscatello et al., 1998).
CONCLUSIONS
The furious pace of advances in the molecular biology of ECM has greatly expanded the
knowledge of individual matrix components.
The structure of many matrix macromolecules,
for example, was determined from cloned
cDNAs or genes long before complete protein
information was available. With this increased
knowledge as background, there is a growing
realization that the information contained in the
ECM is not a monosyllabic message encoded
by individual molecules, but a complex and
intricate arrangement dictated by the combinatorial organization of the supramolecular structure. As the focus of biological research
changes from the letters to the message, understanding how cells read and interpret this information will undoubtedly reveal more about the
letters in the code.
LITERATURE CITED
Burgeson, R.E. 1993. Type VII collagen, anchoring
fibrils, and epidermolysis bullosa. J. Invest. Dermatol. 101:252-255.
Carey, D.J. 1997. Syndecans: Multifunctional cellsurface co-receptors. Biochem. J. 327:1-6.
Chiquet-Ehrismann, R. 1995. Inhibition of cell adhesion by anti-adhesive molecules. Curr. Opin.
Cell Biol. 7:715-719.
Colige, A., Li, S.W., Sieron, A.L., Nusgens, B.V.,
Prockop, D.J., and Lapiere, C.M. 1997. cDNA
cloning and expression of bovine procollagen I
N-proteinase: A new member of the superfamily
of zinc-metalloproteinases with binding sites for
cells and other matrix components. Proc. Natl.
Acad. Sci. U.S.A. 92:2374-2379.
David, G. 1993. Integral membrane heparan sulfate
proteoglycans. FASEB J. 7:1023-1030.
Downing, A.K., Knott, V., Werner, J.M., Cardy,
C.M., Campbell, I.D., and Handford, P.A. 1996.
Solution structure of a pair of calcium-binding
epidermal growth factor-like domains: Implications for the Marfan syndrome and other genetic
disorders. Cell 85:597-605.
Ekblom, P. and Timpl, R. 1996. The Laminins.
Harwood Academic Publishers, New York.
Erickson, H.P. 1993. Tenascin-C, tenascin-R, tenascin X: A family of talented proteins in search of
functions. Curr. Opin. Cell Biol. 5:869-876.
Hassell, J.R., Blochberger, T.C., Rada, J.A., Chakravarti, S., and Noonan, D. 1993. Proteoglycan
gene families. In Extracellular Matrix. Advances
in Molecular and Cellular Biology (H.K. Kleinman, ed.), Vol. 6, pp. 69-113. JAI Press, Greenwich, Conn.
Hay, E.D. (ed.) 1991. Cell Biology of Extracellular
Matrix, 2nd ed. Plenum, New York.
Hulmes, D.J.S. 1992. The collagen superfamily:
Diverse structures and assemblies. Essays Biochem. 24:49-76.
Adams, J.C., Rucker, R.P., and Lawler, J. 1995. The
Thrombospondin Gene Family. R.G. Landis,
Austin, Tex.
Hynes, R.O. 1990. Fibronectins. Springer-Verlag,
New York.
Aumailley, M., Gimond, C., and Rouselle, P. 1996.
Integrin-mediated cellular interactions with
laminins. In The Laminins (P. Ekblom and R.
Timpl, eds.) pp. 127-158. Harwood Academic
Publishers, New York.
Iozzo, R.V. and Murdoch, A.D. 1996. Proteoglycans
of the extracellular environment: Clues from the
gene and protein side offer novel perspectives in
molecular diversity and function. FASEB J.
10:598-614.
Aviezer, D., Hecht, D., Safran, M., Eisinger, M.,
David, G., and Yayon, A. 1994. Perlecan, basal
lamina proteoglycan, promotes basic fibroblast
growth factor receptor binding, mitogenesis, and
angiogenesis. Cell 79:1005-1013.
Iozzo, R.V., Cohen, I.R., Grässel, S., and Murdoch,
A.D. 1994. The biology of perlecan: The multifaceted heparan sulphate proteoglycan of basement membranes and pericellular matrices. Biochem J. 302:625-639.
Beck, K., Hunter, I., and Engel, J. 1990. Structure
and function of laminin: Anatomy of a multidomain glycoprotein. FASEB J. 4:148-160.
Kadler, K. 1994. Extracellular matrix 1: Fibril-forming collagens. Protein Profile 1:519-638.
Bornstein, P. 1995. Diversity of function is inherent
in matricellular proteins: An appraisal of thrombospondin 1. J. Cell Biol. 130:503-506.
Brodsky, B. and Ramshaw, J.A. 1997. The collagen
triple-helix structure. Matrix Biol. 15:545-554.
Brown, J.C. and Timpl, R. 1995. The collagen superfamily. Int. Arch. Allergy Immunol. 107:484490.
Kessler, E., Takahara, K., Biniaminov, L., Brusel,
M., and Greenspan, D.S. 1996. Bone morphogenetic protein-1: The type I procollagen C–proteinase. Science 271:360-362.
Kivirikko, K.I. and Myllyharju, J. 1998. Prolyl 4hydroxylases and their protein disulfide isomerase subunit. Matrix Biol. 16:357-368.
Extracellular
Matrix
10.1.13
Kleinman, H.K. (ed.) 1993. Extracellular Matrix.
Advances in Molecular and Cellular Biology,
Vol. 6. JAI Press, Greenwich, Conn.
Koivu, J., Myllylä, R., Helaakoski, T., Pihlajaniemi,
T., Tasanen, K., and Kivirikko, K.I. 1987. A
single polypeptide acts both as the β subunit of
prolyl 4-hydroxylase and as a protein disulfideisomerase. J. Biol. Chem. 262:6447-6449.
Kühn, K. 1987. The classical collagens: Types I, II,
and III. In Structure and Function of Collagen
Types (R. Mayne and R.E. Burgeson, eds.) pp.
1-42. Academic Press, San Diego.
Kühn, K. 1994. Basement membrane (type IV) collagen. Matrix Biol. 14:439-445.
Lane, T.F. and Sage, E.H. 1994. The biology of
SPARC, a protein that modulates cell-matrix interactions. FASEB J. 8:163-173.
Lee, B., Godfrey, M., Vitale, E., Hori, H., Mattei,
M.-G., Sarfarazi, M., Tsipouras, P., Ramirez, F.,
and Hollister, D.W. 1991. Linkage of Marfan
syndrome and a phenotypically related disorder
to two different fibrillin genes. Nature 352:330334.
Li, K., Tamai, K., Tan, E.M., and Uitto, J. 1996.
Cloning of type XVII collagen. Complementary
and genomic DNA sequences of mouse 180kilodalton bullous pemphigoid antigen (BPAG2)
predict an interrupted collagenous domain, a
transmembrane segment, and unusual features in
the 5′-end of the gene and the 3′-untranslated region
of the mRNA. J. Biol. Chem. 268:8825-8834.
Maurer, P. and Engel, J. 1996. Structure of laminins
and their chain assembly. In The Laminins (P.
Ekblom and R. Timpl, eds.) pp. 27-49. Harwood
Academic Publishers, New York.
Mayer, U. and Timpl, R. 1994. Nidogen: A versatile
binding protein of basement membranes. In Extracellular Matrix Assembly and Structure (P.D.
Yurchenco, D.E. Birk, and R.P. Mecham, eds.)
pp. 318-416. Academic Press, San Diego.
Mayne, R. and Brewton, R.G. 1993. New members
of the collagen superfamily. Curr. Opin. Struct.
Biol. 5:883-890.
Mecham, R.P. and Davis, E.C. 1994. Elastic fiber
structure and assembly. In Extracellular Matrix
Assembly and Structure (P.D. Yurchenco, D.E.
Birk, and R.P. Mecham, eds.) pp. 281-314. Academic Press, San Diego.
Mecham R.P. and Hinek, A. 1996. Non-integrin
laminin receptors. In The Laminins (P. Ekblom
and R. Timpl, eds.) pp. 159-183. Harwood Academic Publishers, New York.
Moscatello, D.K., Santra, M., Mann, D.M., McQuillan, D.J., Wong, A.J., and Iozzo, R.V. 1998.
Decorin suppresses tumor cell growth by activating the epidermal growth factor receptor. J. Clin.
Invest. 101:4406-4412.
Olsen, B.R., Winterhalter, K.H., and Gordon, M.K.
1995. FACIT collagens and their biological
roles. Trends Glycos. Glycotech. 7:115-127.
Partridge, S.M. 1962. Elastin. Adv. Protein Chem.
17:227-302.
Piez, K.A. 1997. History of extracellular matrix: A
personal view. Matrix Biol. 16:85-92.
Potts, J.R. and Campbell, I.D. 1994. Fibronectin
structure and assembly. Curr. Opin. Cell Biol.
6:648-655.
Prockop, D.J. and Kivirikko, K.I. 1995. Collagens:
Molecular biology, diseases, and potentials for
therapy. Annu. Rev. Biochem. 64:403-434.
Rauch, U., Karthikeyan, L., Maurel, P., Margolis,
R.U., and Margolis, R.K. 1992. Cloning and
primary structure of neurocan, a developmentally regulated, aggregating chondroitin sulfate
proteoglycan of brain. J. Biol. Chem. 267:1953619547.
Rehn, M. and Pihlajaniemi, T. 1994. Alpha 1
(XVIII), a collagen chain with frequent interruptions in the collagenous sequence, a distinct tissue distribution, and homology with type XV
collagen. Proc. Natl. Acad. Sci. U.S.A. 91:42344238.
Sage, E.H. and Bornstein, P. 1991. Extracellular
proteins that modulate cell-matrix interactions.
J. Biol. Chem. 266:14831-14834.
Seiffert, D. 1997. Constitutive and regulated expression of vitronectin. Histol. Histopath. 12:787797.
Timpl, R. 1996. Macromolecular organization of
basement membranes. Curr. Opin. Cell Biol.
8:618-624.
Timpl, R. and Chu, M.L. 1994. Microfibrillar collagen type VI. In Extracellular Matrix Assembly
and Structure (P.D. Yurchenco, D.E. Birk, and
R.P. Mecham, eds.) pp. 207-242. Academic
Press, San Diego.
Wight, T.N., Heinegard, D.K., and Hascall, V.C.
1991. Proteoglycans: Structure and function. In
Cell Biology of Extracellular Matrix, 2nd ed.
(E.D. Hay, ed.) pp. 45-78. Plenum, New York.
Yamada, H., Watanabe, K., Shimonaka, M., and
Yamaguchi, Y. 1994. Molecular cloning of brevican, a novel brain proteoglycan of the aggrecan/versican family. J. Biol. Chem. 269:1011910126.
Yurchenco, P.D. 1994. Assembly of laminin and
type IV collagen into basement membrane networks. In Extracellular Matrix Assembly and
Structure (P.D. Yurchenco, D.E. Birk, and R.P.
Mecham, eds.) pp. 351-358. Academic Press,
San Diego.
Zimmermann, D.R. and Ruoslahti, E. 1989. Multiple domains of the large fibroblast proteoglycan,
versican. EMBO J. 8:2975-2981.
Contributed by Robert P. Mecham
Washington University School of Medicine
St. Louis, Missouri
Overview of
Extracellular
Matrix
10.1.14
3UHSDUDWLRQRI%DVHPHQW0HPEUDQH
&RPSRQHQWVIURP(+67XPRUV
81,7
!"#$
%#
#
& %'! ("
%
)* +"
&
,
#
%
%&
##
%&
% % &
##
&)
%
%
%& '! #%
%$
%
&
%)*%%'-
.
%
$-
#$-%-'-&
!"
("&
'!
%)*
&'!
%)*
%#%
'!
%&/
#
''!%)*
&
35(3$5$7,212)0$75,*(/
%
%%#
###
&0
%
#$
&'
!11
&
%$6,&
35272&2/
0DWHULDOV
!"
+&234"#2°4
("#2°4
'
.$33(1',;$"#2°4
4
&&#5#6 )"
718
9&&# %"
4
#2°4
,
%
%$33(1',;"
&RQWULEXWHGE\+\QGD..OHLQPDQ
&XUUHQW3URWRFROVLQ&HOO%LRORJ\!::;"!1&(&!'!1&(&!1
4%<!::;%=>%?#)&
([WUDFHOOXODU
0DWUL[
127(,
2°4#
2°4&
!& )9#$!11"(11+&234
∼!&9
$9&
$3RO\WURQKRPRJHQL]HUZRUNVZHOO'RQRWKRPRJHQL]HIRUWRRORQJDVLWLVLPSRUWDQWQRW
WRZDUPWKHWLVVXH
(& 4!@;111×J#2°4#&5
-
9!"&
7KHFHQWULIXJHDQGURWRUPXVWEHFROGEHIRUHDQGGXULQJXVHIRUDOOPDQLSXODWLRQV
+& ,
!11(
9
&2°4&
2& 4(1(+#111×J#2°4#
%&&
3RXURIIWKHWKLFNVXSHUQDWDQWJHQWO\DVWKHSHOOHWPD\GHWDFKIURPWKHERWWRPRIWKHWXEH
@& 9 @1 ( 2&
4
&
A& 5%9(!1B≥2
$33(1',;"&%(
&
7KLV VWHS VWHULOL]HV WKH LQVLGH RI WKH EDJ FKORURIRUP ZLOO NLOO EDFWHULD DQG VSRUHV
6XEVHTXHQWVWHSVUHPRYHWKHXUHDDQGFKORURIRUP:KHQGLVFDUGLQJWKHFKORURIRUPFRQ
WDLQLQJGLDO\VLVEXIIHULWLVEHVWWRGHFDQWWKHEXIIHUDQGGLVFDUGWKHUHPDLQLQJFKORURIRUP
ZKLFKVHWWOHVWRWKHERWWRPRIWKHYHVVHOLQDFKHPLFDOZDVWHFRQWDLQHU
7& 5%9$#≥(#(
(
&
0DQXDOO\URWDWHWKHEDJVEHWZHHQFKDQJHVWRIDFLOLWDWHWKHH[FKDQJHRIVROXWLRQV
;& >-
%$#%
$718#$#
%
&
:& 5%&&#C$%
%.$33(1',;"&5
"
!(−(1°4&
0DWULJHOFRQFHQWUDWLRQVKRXOGEHWRPJPO
,WLVQRWGHVLUDEOHWRIUHH]HDQGWKDZ0DWULJHOWRRPDQ\WLPHVVRJHQHUDOO\DQGPO
DOLTXRWVDUHPRVWXVHIXO7KHVL]HRIWKHDOLTXRWVVKRXOGEHEDVHGRQQHHG
3UHSDUDWLRQRI
%DVHPHQW
0HPEUDQH
&RPSRQHQWVIURP
(+67XPRUV
4 4%
35(3$5$7,212)/$0,1,1
C'!
#
##
$%
&)
#
$
$'!&C'!
#
%)*
+"&
%$6,&
35272&2/
0DWHULDOV
!"
+&234"#2°4
1&@34"#2°4
,
$33(1',;$"#2°4
34
4
9&&# %"
4
#2°4
,
%
%$33(1',;"
127(,
2°4#
2°4&
!& $ !11 (11 +&2 34 ∼! &9$9
&
7KHYROXPHVJLYHQLQDOOWKHSURWRFROVDUHEDVHGRQJRIWLVVXH/DUJHURUVPDOOHU
DPRXQWVRIWXPRUFDQEHXVHGZLWKSURSRUWLRQDOYROXPHVRIEXIIHUV
$3RO\WURQKRPRJHQL]HUZRUNVZHOO'RQRWKRPRJHQL]HIRUWRRORQJDVLWLVLPSRUWDQWQRW
WRZDUPWKHWLVVXH
(& 4!@;111×J#2°4#&5
-
9!"&
+& ,
(111&@34
9&
2°4&
2& 4 !@ ;111 × J# 2°4& &6+
&
7KHWXPRUKRPRJHQDWHSHOOHWFDQEHXVHGWRSUHSDUHW\SH,9FROODJHQVHH%DVLF3URWRFRO
LI WKH WXPRU WLVVXH ZDV WDNHQ IURP ODWK\ULWLF DQLPDOV 2WKHUZLVH WKH SHOOHW VKRXOG EH
GLVFDUGHG
@& 1&@34#$%
+18
!A&2B!11"$&!
&
A& 4!@;111×J#2°4&
7KHSHOOHWVVKRXOGEHODUJHDQGZKLWHZLWKDUHDVRQDEO\FOHDUVXSHUQDWDQW7KHVXSHU
QDWDQWVKRXOGEHVDYHGXQOHVVLWLVFHUWDLQWKDWDOOWKHODPLQLQSUHFLSLWDWHG
7& 5'!+11
%9≥((
$33(1',;"&6
%$(&
([WUDFHOOXODU
0DWUL[
4 4%
;& %
%##
$%
34#
$#!&7&4!@;111×J#2°4&
7KHVROXWLRQLVDOUHDG\01D&OVR01D&ORUJVKRXOGEHDGGHGSHUOLWHU
7KHSHOOHWVKRXOGEHYHU\VPDOO7KLVVWHSUHPRYHVWKHFRQWDPLQDWLQJW\SH,9FROODJHQ,I
WKHSHOOHWLVODUJHUHSHDWVWHSDQGRPLWVWHS
:& 5%9 7"& /%#
@B
%9'!&
)
#
%$&
!1& >-
%$#%
$718#$#
%
&
!!& 5$
%&&#C$%%.
$33(1',; "&5
'!"!'
−71°4(2&
7KHSURWHLQFRQFHQWUDWLRQVKRXOGEHWRPJPO
%$6,&
35272&2/
35(3$5$7,212)7<3(,9&2//$*(1
%)*
&%)*
≥($- ("&
0DWHULDOV
(#2"#
("
1&@34"#2°4
(&1
⋅4"#2°4
(&1
⋅4+(B
5"#2°4
1&@
(;&@
B$"#2°4
4
#2°4
,
%
%$33(1',;"
127(,
2°4#
2°4&
!& +111&@34&4!@
;111×J#2°4&5
&
7KHVHVWHSVDQGWKHVXEVHTXHQWRQHVUHPRYHFRQWDPLQDWLQJODPLQLQ
(& 9(&1
⋅4
!@;111×
J#2°4&5
$&
+& 9(@1(&1
⋅4B5
2°4&
2& 4!@;111×J#2°4&
&
3UHSDUDWLRQRI
%DVHPHQW
0HPEUDQH
&RPSRQHQWVIURP
(+67XPRUV
@& 9 (@1 (&1 ⋅4 B5& 4
%!@;111×J#2°4&5
$2&
4 4%
A& 5%91&@
≥(
$33(1',;"&
,IWKHFROODJHQFRQFHQWUDWLRQLV!PJPOLWZLOOSUHFLSLWDWHDQGLWLVGLIILFXOWWRUHFRYHU
WKH SUHFLSLWDWHG PDWHULDO 7R HQVXUH WKDW WKH HQWLUH EDWFK GRHV QRW SUHFLSLWDWH GLDO\]H
RQHWHQWK RI WKH VDPSOH WKURXJK WKUHH FKDQJHV EHIRUH SURFHVVLQJ WKH UHPDLQGHU ,I WKH
FROODJHQGRHVSUHFLSLWDWHWKHUHPDLQLQJVXSHUQDWDQWVKRXOGEHGLOXWHGE\KDOIZLWKFROG
GLVWLOOHGZDWHUDQGWKHQGLDO\]HG7KHPDWHULDOWKDWKDVSUHFLSLWDWHGFDQEHFROOHFWHGE\
FHQWULIXJDWLRQUHGLVVROYHGLQWZLFHWKHRULJLQDOYROXPHRI0JXDQLGLQH⋅+&OEXIIHU'77
VROXWLRQDQGWKHQGLDO\]HGDJDLQVW0DFHWLFDFLG
7& 5%
%%%9!1
$&5
%)*"!'
−(1°4(2&
7KHVROXWLRQVKRXOGEHFOHDU7KHSUHVHQFHRIDSUHFLSLWDWHLQGLFDWHVWKDWFROODJHQKDVFRPH
RXWRIVROXWLRQ7KHSUHFLSLWDWHFDQEHUHPRYHGE\FHQWULIXJDWLRQOHDYLQJFRQVLGHUDEOH
W\SH,9FROODJHQLQWKHVXSHUQDWDQW
0$,17(1$1&($1'+$59(672)(+6780256
& )
#
%$
%)* ("&
#$
, % 4 4 6833/,(56 $33(1',;"& , 6&*&)99
=&6&"&/2'
%D&
%
+2$-&
6833257
35272&2/
0DWHULDOV
E
'
.$33(1',;$"
!110B
!11µB%
4@7CBA
,&&#+&(8,"
5%
5/"
718
@1'
(1'+1'
+'%
!A'F
4
#2°4
127( , $
% ) , 4 0 4 ),404" $ %
%&
!& G2"$&
,
7 -(1'+1'%$
&
E - % %& !A'F
%
&
+DYHVHYHUDOLFHEXFNHWVKDQG\WRNHHSWKHWXPRUFROG+DQGOHHDFKWXPRUVHSDUDWHO\
XVLQJFOHDQWXEHVDQGV\ULQJHVLQFDVHWKHUHLVWRRPXFKEDFWHULDOFRQWDPLQDWLRQZLWK
EORRGRUSXVLQRQHRIWKHWXPRUV1XPEHUWKHWXEHVDQGUHVSHFWLYHV\ULQJHVZKHQZRUNLQJ
ZLWKODUJHQXPEHUVRIWXPRUV7KLVVWHSFDQEHPHVV\JORYHVODEFRDWVDQGPDVNVDUH
UHFRPPHQGHG
([WUDFHOOXODU
0DWUL[
4 4%
(& ,
21 %-
&,$
$#@!1
&
5
%&
+& 6(&,
%
!1&
$WWKLVVWDJHWKHSUHSDUDWLRQVKRXOGEHDWKLFNVOXUU\
2& C%9%D(11µ,&)D!
%#%%#!A'F
$+'%&
,WLVSRVVLEOHWRXVHWZRLQMHFWLRQVLWHVSHUDQLPDOWRLQFUHDVHWKH\LHOG:DWFKWKHDQLPDOV
FDUHIXOO\DVWKHWXPRUVJURZWREHVXUHWKDWWKHDQLPDOVFDQREWDLQIRRGHWF
)HPDOHPLFHDUHPRUHGRFLOHDQGHDVLHUWRKDQGOHWKDQPDOHVDQGPD\EHSUHIHUUHGDV
WXPRUUHFLSLHQWV7KHFDJHVVKRXOGEHPDUNHGZLWKWKHWXPRUQXPEHUWRDOORZWUDFNLQJLQ
FDVHRIFRQWDPLQDWLRQLQWKHWXPRU
'LVSHUVHGWXPRUFDQDOVREHIUR]HQDQGVWRUHGIRUIXWXUHXVHDGG3%6FRQWDLQLQJ
'062WRGLVSHUVHGWXPRUWUDQVIHUPODOLTXRWVWRIUHH]LQJYLDOVSODFHLQD−°&IUHH]HU
IRUKUDQGWKHQVWRUHLQOLTXLGQLWURJHQ,WJHQHUDOO\WDNHV!ZHHNVIRUWKHLQLWLDOWXPRU
WREHFRPHIXOO\JURZQZKHQIUR]HQPDWHULDOLVLQMHFWHGSUREDEO\EHFDXVHVRPHFHOOVGLH
GXULQJWKHIUHH]LQJDQGWKDZLQJSURFHVV:KHUHSRVVLEOHLWLVEHVWWRZRUNZLWKIUHVKWXPRU
WLVVXH
@& +$-&9%4/(%
718&4%%'-&
@1'
#9
&
7XPRUVL]HVKRXOGQRWH[FHHGFPLQOHQJWKKHLJKWRUZLGWKDVODUJHUWXPRUVZLOOEHFRPH
QHFURWLF 7XPRUV WKDW DUH YHU\ EORRG\ RU IXOO RI \HOORZ SXV VKRXOG EH GLVFDUGHG )RU
FRQYHQLHQFHIUHH]HWXPRUVDWJEDJDQGODEHOWKHEDJVZLWKWXPRUZHLJKWDQGGDWH
,IQRWXVHGLPPHGLDWHO\IUHVKWXPRUFDQEHIUR]HQLQSRZGHUHGGU\LFHRULPPHUVHGGLUHFWO\
LQOLTXLGQLWURJHQDQGWKHQVWRUHGIRU≤\HDUDW−°&LQIUHH]HUEDJV
6833257
35272&2/
0$,17(1$1&(2)0,&(21$/$7+52*(1,&',(7
'-.%)*
%
%#$%
&)
H+$-"#%
&)
($-#
%%9%
9%
&
0DWHULDOV
C$"
&$87,21C$.#-#
$$
&
!& E
$A
%D&6
$$$
"A
%&
3UHSDUDWLRQRI
%DVHPHQW
0HPEUDQH
&RPSRQHQWVIURP
(+67XPRUV
7KHUHLVQRQHHGWRIHHGPLFHODWKURJHQLFGLHWXQWLOWKHWXPRUVWDUWVWRJURZ$QLPDOVPD\
VSUHDGWKHIRRGDURXQGWKHFDJHVRWKHOHYHOVRIIRRGVKRXOGEHFKHFNHGGDLO\
(& I
≤+$-#$
&
$IWHUZHHNVRQWKHGLHWPLFHVKRXOGEHPRQLWRUHGIRUERQHIUDFWXUHVZKLFKDUHREVHUYHG
ZLWKWKLVGLHW0LFHLQGLVWUHVVVKRXOGEHVDFULILFHGDQGWKHWXPRUVKDUYHVWHG
4 4%
5($*(176$1'62/87,216
8VHGHLRQL]HGRUGLVWLOOHGZDWHULQDOOUHFLSHVDQGSURWRFROVWHSV)RUFRPPRQVWRFNVROXWLRQVVHH
$33(1',;$IRUVXSSOLHUVVHH6833/,(56$33(1',;
*XDQLGLQH⋅+&OEXIIHU0
!:!&1
⋅4("
!(&!1&1@"
(/!;11
,
D7&2$(4
,
(/(
/DWKURJHQLFFKRZ
!11β'"
(9
"
1&(%"
@-
$
5
+11
$
$
$(1&A2°4&
:KHQ DGGLQJ WKH VROXWLRQ VSULQNOH LW RQ GLIIHUHQW DUHDV RI WKH FKRZ WR KHOS WKH PL[LQJ
SURFHVV
4,0)/3J*ORYHVPDVNDQGODEFRDWVKRXOGEHZRUQZKHQKDQGOLQJWKHVHWR[LFPDWHULDOV
1D&OEXIIHU0
7&@341&@"
A&1@1&1@"
(/:11
,
D7&2$(34
,
(/!
A
1D&OEXIIHU0
+:734+&2"
+5,1&11@"
@1'%
3.1&1("
!(&!1&1@"
(/!;11
,
D7&2$(34
,
(/(
8UHDEXIIHU0
(21("
!(&!1&1@"
(/!;11
,
D7&2$(34
,
(/(
A2°4
([WUDFHOOXODU
0DWUL[
4 4%
&200(17$5<
%DFNJURXQG,QIRUPDWLRQ
3UHSDUDWLRQRI
%DVHPHQW
0HPEUDQH
&RPSRQHQWVIURP
(+67XPRUV
'
/- &# !:77"&
##
# !:;7. I &#
!::+"& D#'!
%)*
#
%
% %&'
%%
% &
'
%
%)*
'-
%
&)#
% )* % &
$
$
!:21&
)-$&)$%
9
'
%%&)!:77#
$ $%)*$
D /- &# !:77"& % '
%
& , %#'
%#
%
&# !::1. K %# !::1.
$#!::2"&
#
α!#β!#
γ!#
'!&#
!:7:.&#!::2"&Eα#β#
$γ
&
#
# & F
% & , '( α(β!γ!" %&%'
' '
&
% )* $ #
(J! # $ α!)*"(α()*"/-&#!:77.I
&#!:;("&,
%)*'
#
α+#
α2#α@#
αA&C-#
9#
% &
'
&
# # $'
4% &# !:;A. I &#
!:;A"&
%&
%'-
$ # $
% %'- $;1'
C&#
!:;A.I&#!:;;"&
% $ %4%&#!:;A"&
#
$
%
'
&4
$
$$
%
&#!:;@.81,7"&)#
&#
!::("
- $ E
&#
!::!"& % , &# !:;7. 81,7
"& C'! % &)
#
$# # $
&# !:;(. I &#
!::+"&)
$&%
)*
$
%
%&
&ULWLFDO3DUDPHWHUVDQG
7URXEOHVKRRWLQJ
(+6WXPRU ,D$'
&)'
%9
$%#%$#
&)
&)9
%
% - $&)
# $
%
'
4 4%
& F% H2 & E
#
9−71°4≤!%
$
&
0DWULJHO -% # % %&
%'
≤2
%&)
- &
/
#-
2°4#'
&0
%-
& L:B.$'
#$
&
5
% 218
#$
%
%&
/DPLQLQ ) $ %'
$%'!
&)%-
'!&
7\SH ,9 FROODJHQ , D$
%)*&
$'
H1&;B$
%9
#
& %
%#
$%
.
%
$%&
%)*
&
E#%)*
&%
%
&
$QWLFLSDWHG5HVXOWV
)#!%
∼!&@
!1'!
@
%)*&
7LPH&RQVLGHUDWLRQV
%∼2
+$-D
&)
($-&)%
% %
9%
&'
# '!# % )*
≤!$-&
,
&
/LWHUDWXUH&LWHG
,# ,&# )$# K&# I# &I&# #
F&6&#I9$-#=&&#
$#6&3&!:;7&
,
%'
& &DQFHU 5HV
27J+(+:'+(2@&
-#I&##)&#
#=&!::1&
J,%
'
&)$6(%-2J!2;'!A1&
#6&&#4#&#59#6&#-'
# &# # =&# I# &I&# #
F&6&# 99# F&# # &# # =&#
# 6&# %# I&# K
# K&# K# &&!::2&,$
&0DWUL[%LRO!2J(1:'(!!&
4%#5&=&#
#&&#
6%#4&&!:;A&
$%J)
%'
-'
&-&HOO%LRO&!1(J((@2'((A+&
# ,&&# I# &I&# /# &#
$9#=& &#
5'5#&!:;(&E'
$ % & - 1HXURVFL 5HV
;J!7:'!:2&
E
#6&#I%#&4&#6%#C&&#M#&#
$%#&&#=#5&C&#K#=&6&#'
#F&6&#
I#&I&!::!&
" '
$ % D
& - 1DWO &DQFHU ,QVW
;+J7A:'772&
%# &,&# %# &>&# 9'N# 4&,&#
I#&I&#
5%#&!:;@&'
#
# '
&-&HOO%LRO!1!J!@!!'!@((&
I#&I&#F%#&C&#C#C&,&#F'
6%# &#%#I&#
#F&6&
!:;(& ) 9 % )* & %LR
FKHPLVWU\(2JA!;;'A!:+&
I# &I&# F%# &C&# # =&6&#
#*&C&#4#E&&#C#F&>&#
'
#F&6&!:;A&
$ %& %LRFKHPLVWU\ (@J+!('
+!;&
I# &I&# I%# &4&# # >&&#
K# I&# >-# &&# F# 5&&
!::+&J,%'
%
'
&9LWDPLQV+RUPRQHV
27J!A!'!;A&
I# K&# I# &I&# # F&6&# C$%#&=&!:;;&6
'
%'-&-&&HOO
%LRO&!17J!@;:'!@:7&
C# C&# ,# &=&# E# 5&,&# # 4&#
#=&6&#
#&=&!:;A&)
'
%&3URF
1DWO$FDG6FL86$;2J!+A'!21&
([WUDFHOOXODU
0DWUL[
4 4%
#F&6&
#6&!:;7&C
& $QQX 5HY
&HOO%LRO+J@7';@&
# 6& $# =&4& !::2& % & %LR(VVD\V
!;J!(+'!+(&
/-#6&>&#F# &#F
$#&&##
F&6&#*#&#
$#6&!:77&,'
'
&-([S0HG&!2@J(12'((1&
# 6&# 6
# &# 6%# &F&# 6
# &)&#
E
#=&&#
#F&6&!:7:&CO
, % & -
%LRO&KHP&(@2J::++'::+:&
#,&#%#6&&# #6&#F#K&#C#
&*&#%#=&,&# %#6&6&#F#5&&#
# F&6& !::(& , # '
'
#
$
&/DE,QYHVW&A7J@!:'@(;&
K# &5&
%#=&4&!::1&
& )$6(% -&
2J!@77'!@:1&
4
%%
I&I
3)56B3)
#%
3UHSDUDWLRQRI
%DVHPHQW
0HPEUDQH
&RPSRQHQWVIURP
(+67XPRUV
4 4%
3UHSDUDWLRQRI*HOOHG6XEVWUDWHV
81,7
%)
#$
$%
&
% & % ) %$
&
%#
$
& %
.
%!1&+&!"&5
%
#
$
$
'
&
#$#
&
%) !"
("
&F
%)
&5
%)
&,
$$
,
!"
$
$, ("&
127(,
$#
%&
35(3$5$7,212)7<3(,&2//$*(168%675$7(6
F
%%$
&F
%)
%%#
% & 5
%
-&%)
$ $# # %# -
%#
&
%$6,&
35272&2/
0DWHULDOV
+@B%)
&&#*!11#4"#
!1×'
#7&2!1× .$33(1',;$"!1×
&&#
!::#C"#
1&!3/#
1&!4#
5
J(;&@
B
9
(/
4
#+7°4
>-
4
3UHSDULQJJHOOHGW\SH,FROODJHQVXEVWUDWHV
!& 0
-#;%)
$!!1× #7&2#
!1&!3/
"&,
D7&2$
1&!3/1&!4
##
&
&RQWULEXWHGE\+\QGD..OHLQPDQ
&XUUHQW3URWRFROVLQ&HOO%LRORJ\!::;"!1&+&!'!1&+&:
4%<!::;%=>%?#)&
([WUDFHOOXODU
0DWUL[
7\SH,FROODJHQFRPPHUFLDOO\REWDLQHGDWWRPJPOLQGLOXWHDFLGVKRXOGQRWEHIXUWKHU
GLOXWHGPRUHWKDQQHFHVVDU\RUWKHILQDOJHOZLOOEHWRRZHDN%HORZPJPOWKHJHOZLOO
EHYHU\ZHDNDQGZLOOQRWZLWKVWDQGWKHXVXDOPDQLSXODWLRQVRISODWLQJDQGIHHGLQJWKH
FHOOV7\SH,FROODJHQREWDLQHGDVDSRZGHUFDQEHVROXELOL]HGDWWRPJPOE\VWLUULQJ
RYHUQLJKWDW°&LQGLOXWHDFHWLFDFLG
$OWHUQDWLYHO\×PHGLXPVDOWVFDQEHXVHGLQSODFHRIWKH3%6
:KHQDGULHGILOPRIILEULOODUFROODJHQLVGHVLUHGFRDWWKHGLVKZLWKDVPDOODPRXQWRI
FROODJHQDQGDLUGU\WKHPDWHULDOXVXDOO\RYHUQLJKWLQDVWHULOHFXOWXUHKRRG6NHOHWDO
PXVFOHFHOOVDUHFXOWXUHGRQGULHGFROODJHQWRHQFRXUDJHFHOOGLIIHUHQWLDWLRQLHIXVLRQWR
P\RWXEHV
(& $
!+7°4&
7KHFROODJHQVKRXOGFRPSOHWHO\FRYHUWKHVXUIDFHRIWKHGLVKPOLVJHQHUDOO\VXIILFLHQW
IRUDPPGLVK7KHPDWUL[XVXDOO\WXUQVZKLWHZKHQLWJHOV
&2LVQRWLPSRUWDQWIRUJHOOLQJRIWKHFROODJHQEXWVXEVHTXHQWLQFXEDWLRQVZLWKFHOOV
VKRXOGEHSHUIRUPHGLQDKXPLGLILHG°&&2LQFXEDWRU
+& #
$
$
&,
&
:DUPDOOVROXWLRQVDQGJHQWO\SLSHWRQWRWKHFXOWXUHGLVK
,QVRPHFDVHVRQHPD\ZDQWWRJURZFHOOVLQVLGHDW\SH,FROODJHQJHO7KLVFDQEHGRQH
E\RYHUOD\LQJWKHFROODJHQDIWHUWKHFHOOVKDYHDWWDFKHGRUE\PL[LQJWKHFHOOVLQWKHOLTXLG
W\SH,FROODJHQSULRUWRJHODWLRQVHH$OWHUQDWH3URWRFRO
)ORDWLQJFROODJHQJHOVFDQEHREWDLQHGE\JHQWO\DJLWDWLQJDURXQGWKHHGJHRIWKHFROODJHQ
JHOZLWKDVSDWXODDIWHUWKHFHOOVKDYHDWWDFKHG:KHQW\SH,FROODJHQJHOVDUHXVHGDV
IORDWLQJJHOVZLWKILEUREODVWVDWWDFKHGVRPHFRQWUDFWLRQRIWKHJHOZLOORFFXU+HSDWRF\WHV
KDYHDOVREHHQVXFFHVVIXOO\FXOWXUHGRQIORDWLQJFROODJHQJHOV
3UHSDULQJQRQJHOOHGW\SH,FROODJHQVXEVWUDWHV
!& 5+@B1&!B$
1&!4&
:KHQGLUHFWO\GU\LQJFROODJHQRQWLVVXHFXOWXUHGLVKHVZLWKRXWJHODWLRQXVXDOO\PXFKOHVV
PDWHULDOLVQHHGHGWKDQIRUJHOOHGVXEVWUDWHV*HQHUDOO\WRµJFROODJHQSURYLGHVJRRG
DFWLYLW\RQPPGLDPHWHUZHOOZHOOGLVKHV
7KHFROODJHQFDQEHGLOXWHGZLWK0+&ODVDQDOWHUQDWLYHWRDFHWLFDFLG
(& 1&!B
&
,$
%&
)RUH[DPSOHµORIDPJPOFROODJHQVROXWLRQFDQEHXVHGIRUHDFKPPGLDPHWHU
ZHOORIDZHOOGLVKWKHVROXWLRQVKRXOGFRPSOHWHO\FRYHUWKHERWWRPRIWKHZHOO
+& #$
$"&,
&
:DUPDOOVROXWLRQVDQGSLSHWJHQWO\RQWRWKHFXOWXUHGLVK
3UHSDUDWLRQRI
*HOOHG6XEVWUDWHV
4 4%
35(3$5$7,212)*(//('0$75,*(/68%675$7(6
%81,7"&#
#
##
$$&
$ # $ &
%$6,&
35272&2/
0DWHULDOV
81,7.
%
5-C$"
#2°4#-$-"
4
#+7°4
>-
4
!& $2°4&
7KDZLQJFDQEHDFFRPSOLVKHGE\OHDYLQJWKH0DWULJHOLQWKHUHIULJHUDWRURQLFHRYHUQLJKW
RUE\UROOLQJWKHERWWOHLQWKHKDQGVXQWLOLWLVWKDZHG'RQRWDOORZWKHVROXWLRQWREHFRPH
ZDUP DV LW ZLOO JHO DQG FDQQRW EH UHFRYHUHG ,W LV QRW GHVLUDEOH WR WKDZ DQG UHIUHH]H
0DWULJHOWRRPDQ\WLPHVVRLWVKRXOGEHVWRUHGLQDSSURSULDWHVL]HGDOLTXRWVEDVHGRQWKH
TXDQWLWLHVJHQHUDOO\QHHGHG
,WLVEHVWWRDOZD\VKDYHIUHVKZHWLFHLQDEXFNHWZKHQZRUNLQJZLWK0DWULJHOWRNHHSLW
FROG
(& )
#
&
7KH0DWULJHOVKRXOGEHDW≥PJPOIRUPD[LPXPJHOOLQJ&RQFHQWUDWLRQVEHORZWKDWZLOO
IRUPZHDNJHOVWKDWPD\IDOODSDUW0DWULJHOFDQEHGLOXWHGWRPJPODQGZLOOVWLOOJHO
EXWWKHJHOZLOOEHZHDN,QVRPHFDVHVKRZHYHUDZHDNJHOLVQHHGHG'LOXWHWKHVWRFN
0DWULJHOZLWKFROGPHGLXPVDOWVDQGPL[ZHOO
,IWKHUHDUHEXEEOHVLQWKH0DWULJHOZKLFKFDQFDXVHSUREOHPVRQFHLWJHOVOHWVWDQGIRUD
IHZPLQXWHVRQLFHWRDOORZWKHPWRGLVVLSDWH.HHSLQJWKHGLVKOHYHOTXLFNO\SLSHWWKH
0DWULJHOLQWRWKHGLVKHVZKHQSHUIRUPLQJPXOWLSOHSLSHWWLQJVWHSVGRWKHVHTXLFNO\DQG
FKDQJHWKHSLSHWVLIWKH0DWULJHOVWDUWVWRJHOLQVLGHWKHP)RUPPGLDPHWHUZHOOVXVH
µOSHUZHOOWKLVYROXPHLVQHFHVVDU\EHFDXVHDODUJHPHQLVFXVIRUPVDWWKHHGJHRI
WKHZHOODQGGLVWRUWVWKHVXUIDFH)RUODUJHUGLVKHVDWKLQQHUFRDWLQJZRUNVZHOOEXWRQH
PXVWEHDZDUHRIWKHFXUYHLQWKHJHODWWKHHGJHVRIWKHGLVK)RU≥PPGLDPHWHUGLVKHV
SLSHWWKH0DWULJHOVRWKDWWKHERWWRPRIWKHGLVKLVFRYHUHGWKHQSRXURIIWKHH[FHVVDQG
XVHLWIRUDQRWKHUGLVK
+& 4
+1A1+7°4$
&
6LQFH0DWULJHOZLOOJHODWURRPWHPSHUDWXUHSODFHGLVKHVLQLQFXEDWRULPPHGLDWHO\
2& F%
$
&6
-
&4
#-
%&
)RUH[DPSOHIRULQYLWURDVVD\VRIDQJLRJHQHVLVWXEHIRUPDWLRQLQFXEDWHHQGR
WKHOLDOFHOOVLQµOFRPSOHWHPHGLXPLQPPGLDPHWHUZHOOVFRQWDLQLQJµOJHOOHG
0DWULJHO)L[WKHFXOWXUHDQGVWDLQZLWK'LII4XLN%D[WHUXVLQJRQO\WKHIL[DQGVROXWLRQ
,,XVHRIVROXWLRQ,ZLOOVWDLQWKH0DWULJHODQGPDNHLWGLIILFXOWWRGLVWLQJXLVKWKHFHOOV
7KHDVVD\FDQEHTXDQWLILHGZLWKDQLPDJHSURFHVVRUHJ1,+,PDJH
([WUDFHOOXODU
0DWUL[
4 4%
$/7(51$7(
35272&2/
*52:7+2)&(//6,16,'(0$75,*(/
*%
$
&E#
%
'-$$
&
0DWHULDOV
81,7.
%
5-C$"
4
#+7°4
>-
4
!& $2°4&
7KDZLQJFDQEHDFFRPSOLVKHGE\OHDYLQJWKH0DWULJHOLQWKHUHIULJHUDWRURQLFHRYHUQLJKW
,WLVEHVWWRDOZD\VKDYHIUHVKZHWLFHLQDEXFNHWZKHQZRUNLQJZLWK0DWULJHOWRNHHSLW
FROG
(& & 4 @ !71 × J# #&5#
#
$&,
$
#$#
&
7DNHFDUHWRIXOO\PL[WKHFHOOVLQWRWKH0DWULJHOZLWKRXWFDXVLQJEXEEOHV7KLVFDQEH
DFFRPSOLVKHG E\ JHQWO\ WLSSLQJ WKH WXEH ° VHYHUDO WLPHV $SSUR[LPDWHO\ µO RI
0DWULJHOLQDPPGLDPHWHUZHOOZRUNVZHOOZLWKHQGRWKHOLDOFHOOV
7KHQXPEHURIFHOOVDQGYROXPHRI0DWULJHOQHHGHGIRUFHOOGLIIHUHQWLDWLRQYDU\DPRQJFHOO
W\SHV,WPD\EHQHFHVVDU\WRWHVWYDULRXVQXPEHUVRIFHOOVLQDFRQVWDQWYROXPHRI0DWULJHO
$JRRGUDQJHWRWHVWLQLWLDOO\LVWRFHOOVSHUµO
7KHFHOOVZLOOEHWUDSSHGLQVLGHWKH0DWULJHOZKHQLWJHOV3UHYLRXVVWXGLHVKDYHIRXQGWKDW
VRPHFHOOVZLOOPLJUDWHWRZDUGVHDFKRWKHUDQGIRUPGLIIHUHQWLDWHGVWUXFWXUHV+LVWRORJLFDO
VHFWLRQLQJDQGRUHOHFWURQPLFURVFRS\LVQHFHVVDU\WRIXOO\REVHUYHWKHVHVWUXFWXUHV
+& 4
+1A1+7°4$&
6LQFH0DWULJHOZLOOJHODWURRPWHPSHUDWXUHSODFHGLVKHVLQLQFXEDWRULPPHGLDWHO\
2& F%
$
&6
-&
4
#-
&
$/7(51$7(
35272&2/
3UHSDUDWLRQRI
*HOOHG6XEVWUDWHV
86(2)0$75,*(/,19,92)25$1*,2*(1,&$66$<6$1'
78025*52:7+
%$%$
&)
$
D#$#
$!$-&
% % &
4 4%
)
%
$
D
# - $ &%
$
%&
0DWHULDOV
81,7.
%
5-C$"
4@7CA%RU%
$
%RU
&&#%"
#%
%-&&#5-6I#"
%#
&&#3))"
>-
4
!'+'%
(+'(@'F
127( , $
% ) 4 0 4 ),40C" $ % %&
!& $2°4&
7KDZLQJFDQEHDFFRPSOLVKHGE\OHDYLQJWKH0DWULJHORQLFHLQWKHUHIULJHUDWRURYHUQLJKW
,WLVEHVWWRDOZD\VKDYHIUHVKZHWLFHLQDEXFNHWZKHQZRUNLQJZLWK0DWULJHOWRNHHSLWFROG
(& $%$
$&
6HH$OWHUQDWH3URWRFROVWHSIRUGHWDLOVRIPL[LQJFHOOVDQG0DWULJHO7KHILQDOLQMHFWHG
YROXPHVKRXOGQRWH[FHHGPO,WLVEHVWQRWWRUHGXFHWKHFRQFHQWUDWLRQRI0DWULJHOEHORZ
PJPO
$QJLRJHQLFIDFWRUVIRUWHVWLQJDQGWXPRUFHOOVSLHFHVVKRXOGEHVXVSHQGHGLQDVVPDOOD
YROXPHDVSRVVLEOH%DVLFILEUREODVWJURZWKIDFWRUDWWRQJPOVKRXOGEHLQFOXGHGZLWK
0DWULJHOLQVRPHSOXJVDVDSRVLWLYHFRQWURO
,I XVLQJ WXPRU ELRSV\ SLHFHV WKH PDWHULDO PXVW EH ILQHO\ PLQFHG RU GLVVRFLDWHG ZLWK
SURWHDVHVLQFOXGLQJWU\SVLQRUFROODJHQDVHVRWKDWLWZLOOSDVVWKURXJKWKHV\ULQJH
+& 4!'+'%
$(+'(@'F
%-
&C
%$!B
B
&E%#
%
D%$
$&E'$
#
D'
%
&,D#
∼+1
& $%
9
-D-%%$%"&
)HPDOHPLFHDUHHDVLHUWRKDQGOHWKDQPDOHPLFH0LFHGRQRWQHHGWREHDQHVWKHWL]HG
,IGRLQJPXOWLSOHDQLPDOVLWPD\EHQHFHVVDU\WRFKDQJHQHHGOHVRIWHQDVWKH0DWULJHOZLOO
FORJWKHQHHGOH
([WUDFHOOXODU
0DWUL[
4 4%
2& 6XEFXWDQHRXVDQJLRJHQHVLVDVVD\!$-&F%-
$%
#$$%
-
%%&
%-
"
%%&
)RUKLVWRORJ\IL[WKHSOXJVVWDLQWKHPZLWK0DVVRQWULFKURPHDQGTXDQWLI\DQJLRJHQHVLV
XVLQJDQLPDJHSURFHVVRUE\FRXQWLQJWKHQXPEHURIYHVVHOVDQGRUPHDVXULQJWKHDUHD
RFFXSLHG
2& 7XPRUJURZWK%%($-#
3)++#H(&
&200(17$5<
3UHSDUDWLRQRI
*HOOHG6XEVWUDWHV
$
&'
$
'
%%
$ #!:7(.&#!:7:.I
&#!:;!"&4
%
'
#$
"
&
4%
% &5
%
-
I#!:AA"&
'
$ 9
'
-$ $ $
%
$
I&#!:;A#!:;7"&'
%
& 4
'
!1&+&!"&
%&E#
%'-
$ # I&#!:;A.I&#
!:;;"&%
'
&#!::A"
*-&#!::1"&
'
#
#$'
$
%
4%&#!:;A"&,
$
$ % %"
$$&
,
$'
#'
#
'
#$
&C-'
% % &%
)
#'
+18
%$'
"+@B$
$
&%)
#
$
%
&
$ %&#!:;@"&,
#
'
%
$
'
&E#
% # !@'
$ 9
$
&>
'
#
'-'
$
9
&)#
% $ % #
%
#
&#!::1.*-&#!::("&'
$%
-
%
&/%%'
$ &
%
&E'
# % #
%
'
&E#
% '
% % 4
&# !::+"&
#'
#$
%#
4 4%
7DEOH (IIHFWRI0DWULJHORQ&HOODQG([SODQW'LIIHUHQWLDWLRQ
4B
6
5
E
/$
%
E%'-$
*-&#!::1
4%&#!:;A
I&#!:;;
%
%
% 2@1
%
E
#
6
3
/$
/% /
/
E$$
9
,$%
%
,
!@'
#
%
%
%))% 4
#
%
%
%$%
'
%&
7URXEOHVKRRWLQJ
%)'
2°4 $ & E
'H!$-#
%
#%$%
&)%
%
&
>$%#%'
%%
# # '
# " &5
& % '
%
'
&
&K
3&
#&&
'MP&#!:;;
C&#!:;A
I&#!:;A
9
%#!:;;
4&#!:;:
=#!::!
%&#!:;A
&#!::A
%&#!:;@
F
%#!:;A
&#!:;;
$QWLFLSDWHG5HVXOWV
% #'
'
&)#$#
% & 5
%B#
%%)
.%'-
$I&#!:;;"&
3$$$
$
'
& $ # #%+'A'
'
E
&#!::!"&E#
-
'
'
J ∼@8 $ H@18 $ &
([WUDFHOOXODU
0DWUL[
4 4%
7LPH&RQVLGHUDWLRQV
% ) $ +1 # -+7°4
$
&
%&
!;
%'-
(2I&#!:;;"&)'
#%
%
'-&#
!::A"& % $ &E$
% % % @ %. &#!::("&F'
%%
7
%D&
/LWHUDWXUH&LWHG
%
&&HOO!!J!1:'!!:&
E
#6&#I%#&4&#6%#C&&#M#&#
$%#&&#=#5&C&#K#=&6&#'
#F&6&#
I#&I&!::!&
" '
$%D
& - 1DWO &DQFHU ,QVW&
;+J7A:'772&
F#F&
%#&5&!:;A&
%
%'-
J 4 '
%
& - &HOO %LRO&
!1@J(@1&
%# &,&# %# &>&# 9'N# 4&,&#
I#&I&#
5%#&!:;@&'
#
&
-&HOO%LRO&!1!J!@!!'!@((&
-#&5&
I#)&6&!:AA&'
&3URF1DWO$FDG6FL86$@@J!!:'!(A&
%#&#5#&,&#
F#5&!:;A&
'
J"'
'
9&
(XU-&HOO%LRO2(JA1'A7&
# & &# I%# &4&# C# =&=&# I# &I& !::A& 6 '! FEβ+
'
%
&-&HOO6FL&
!1:J(!1+'(1(!&
'MP#,&#6#F&&##3&6&C&#
E#&6&!:;;&4'
''
%
%- %
%&3URF1DWO$FDG6FL
86$;@AJ(!A!'(!A@&
=#&&!::!&F$
% $ &-([S=RRO&
(A1J((:'(+;&
9# && %# &5& !:;;& 3 +5 9
# & 'HY %LRO
!(@J!:'++&
#&#F9$-#5&&#
#M&,&##
&&#
-$-#6&&!:;;&E%(
% '- '
'
&-&HOO3K\VLRO!+AJ(1+'(!2&
4%#5&=&#
#&&#
6%#4&&!:;A&
$%J)
%'
-'
&-&HOO%LRO&!1(J((@2'((A+&
4
#&4&#F#5&&##F&&#,#
6&#E#,&&#
I#&I&!::+&)
'
$%&-
&OLQ,QYHVW&:!J:77':;@&
4# &F&# 4# 6&# -# I&# #
&C&#
$
$#6&5&!:;:&4
% $
&3URF1DWO$FDG6FL
86$;@J!71('!7!7&
#&
#=&!:7(&4
&-&HOO%LRO&@2JA(A'A+7&
# =&&# $# &=&# -# 5&6&
!:7:& '
I#&I&#I#6&=&#
#F&6&!:;!&
6
$&-&HOO%LRO&;;J27+'2;@&
I# &I&# F%# &C&# # =&6&#
#*&C&#4#E&&#C#F&>&#
#
F&6&!:;A&$
%&%LRFKHPLVWU\(@J+!('+!;&
I#&I&#F#=&#)$#K&#I#F&&#
/#6&4&#-#&#K
#K&##F&6&#
C-'
#C&!:;7&6
&$QQ1<$FDG
6FL&@!+J!+2'!2@&
I# K&# I# &I&# # F&6&# C$%#&=&!:;;&6
'
%'-&-&HOO
%LRO!17J!@;:'!@:7&
C# C&# ,# &=&# E# 5&,&# # 4&#
#=&6&#
#&=&!:;A&)
'
%&3URF
1DWO$FDG6FL86$;2J!+A'!21&
#,&#%#6&&# #6&#F#K&#C#
&*&#%#=&,&# %#6&6&#F#5&&#
# F&6& !::(& , # '
'
#
$&/DE
,QYHVW&A7J@!:'@(;&
3UHSDUDWLRQRI
*HOOHG6XEVWUDWHV
4 4%
#&#>#K&#99%#&&#
I#
&I&!::1&
$'
$
-
%'
''
&3URF1DWO$FDG6FL86$;7J211('211A&
*-# &# C%# E& &# I# &I&# 6
#,&&!::1&5'
-$-%'
&
&HOOA+J2+7'22@&
*-# &# I# &I&# C%# E& &#
6# ,&&# 6# 3&&# 6
# ,&&
!::(& )
$
%'
& ([S
&HOO5HV&(1(J!';&
4
%%
I&I
3)56B3)
#%
([WUDFHOOXODU
0DWUL[
4 4%
3UHSDUDWLRQRI([WUDFHOOXODU0DWULFHV
3URGXFHGE\&XOWXUHG&RUQHDO(QGRWKHOLDO
DQG3)+5(QGRGHUPDO&HOOV
81,7
4 % $ # # &
$4"
$4
##
&$4
%
4" E'6:
&4'4%4
# $ 4 E'6: % 4%"&
$ $ 4# $ %
4#4P9#
B & E # 4'
%&E$
%
#
%9%$Q'!11
32/#
%4#
#
%
&4$
$ 2°4&0
%&&#
##
"A&
4'4'
& ,
% 4
$ % $ "
4
E'6:6:'4"&%
4# $
'
##
%4&
127( , +7°4# !18 4/(
$
&
127(,
$#
%&
35(3$5$7,212)%29,1(&251($/(1'27+(/,$/&(//(&0
%&((&0
##
-%
%'
4"#$'
$4
%&
4%
#
%
.
%
##
'
#
.
%%&D
#%####
-'
)"&
%$6,&
35272&2/
,
%%
#
4'
&&#$
"
%
%+@'
&#
%$!1'
%-
([WUDFHOOXODU
0DWUL[
&RQWULEXWHGE\,VUDHO9ORGDYVN\
&XUUHQW3URWRFROVLQ&HOO%LRORJ\!:::"!1&2&!'!1&2&!2
4%<!:::%=>%?#)&
!
"#$
%$
+@'
%''
#%
4'
"&
0DWHULDOV
!1(1%%
$
:@8B"
$33(1',;$"
$'5'!1"
6$EFE."
%B5,"
521"
B32/%"
FC"
E9.C"
!;'F
!1'#A'#
+@'
F
'
"#
+7°4#!18B"4/($!118
%
!1''
"
(VWDEOLVKDQGPDLQWDLQVWRFN%&(FHOOFXOWXUHV
!& /!1(1%%
$&
2°4
$(2&
7KHH\HVFDQEHVWRUHGDW°&
(& %
##%'&>%$:@8
9
%
%&
+& D
$
!;'F
&)
#
#
!1'
&
2& >%$ #$
-
%
&
@& F%
$
#
%&
$SSO\LQJSUHVVXUHZKLOHVFUDSLQJWKHHQGRWKHOLXPPD\FDXVHRQHWRSHQHWUDWHWKH'HVFH
PHW©VPHPEUDQHDQGSLFNXSVWURPDOFHOOV
A& %
A'
@
$'5'!1&6@
A
$&
9LVXDOH[DPLQDWLRQRIWKHSODWHVKRXOGUHYHDOVPDOOZKLWLVKWLVVXHIUDJPHQWVIORDWLQJLQWKH
PHGLXP:KHQH[DPLQHGXQGHUDSKDVHFRQWUDVWPLFURVFRSHWKHWLVVXHIUDJPHQWVDSSHDU
WREHFRPSRVHGRIFORVHO\DSSRVHGKH[DJRQDOFHOOVRUJDQL]HGLQWRDKRQH\FRPESDWWHUQ
7& 62A#%
&
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
;& )+7°4!184/($!118
%&C
@ %# '
&
$IWHU WR GD\V VHYHUDO WLVVXH IUDJPHQWV DWWDFK WR WKH SODVWLF DQG FHOOV FDQ EH VHHQ
PLJUDWLQJRXWRIWKHH[SODQWV
!
4 4%
:& /
%@A#
$
EFE
%
%!B&
!1& / % ; !(# A' % # ( %B5,#
+7°4@
&
%\GD\WRZHOOGHYHORSHGFRORQLHVDUHFOHDUO\YLVLEOHDQGHDFKFPGLVKFRQWDLQV∼
×FHOOV
2QH SULPDU\ FXOWXUH LV VHOHFWHG IRU SUHSDUDWLRQ RI VWRFN FXOWXUH GLVKHV 2WKHU SULPDU\
FXOWXUHGLVKHVFDQEHSURFHVVHGLQSDUDOOHORUPDLQWDLQHGLQWKHLQFXEDWRUIRUXSWRZHHNV
!!& ,%B5,
@
$''
5'!1&
!(& @'∼!×!1A"'
!1'
@
$'5'
!1&,
EFE!B%
%%@7×
!1AB
"&
*HODWLQFRDWHGGLVKHVDUHXVHGLIQHFHVVDU\WRLPSURYHWKHSODWLQJHIILFLHQF\RI%&(FHOOV
!+& /
%2A%"#
#
@
%B5,#
+7°4@
&
,WLVEHVWWRSDVVDJHWKHFHOOVZKHQVXEFRQIOXHQWEHFDXVHDIWHUUHDFKLQJFRQIOXHQFHWKH
FHOOVSURGXFHDODUJHDPRXQWRI(&0DQGKHQFHEHFRPHOHVVVXVFHSWLEOHWRWU\SVLQL]DWLRQ
DQGSURQHWRGDPDJHLQWKHSURFHVV7KLVZLOOEHUHIOHFWHGLQSRRUSODWLQJHIILFLHQF\DQG
FHOOJURZWKDVZHOODVLQDEHUUDQWFHOOPRUSKRORJ\
!2& ,%B5,
!!
$''
5'!1&
!@& ,
!!1''
#!1
$'5'!1$!B
EFE&,
EFE!B%
%%@7
×!1AB
"&
7KHVXUIDFHDUHDRIDFPVWRFNGLVKLVDSSUR[LPDWHO\HTXDOWRWKDWRIHLJKWPPGLVKHV
7KHUHIRUHWRSUHSDUHDEDWFKRIWZRKXQGUHGPPGLVKHVIRUSURGXFWLRQRI(&0VHH
VWHSLWLVDGYLVDEOHWRXVHWKHFHOOVIURPVWRFNGLVKHV$JURXSRIHOHYHQFPVWRFN
GLVKHVFDQEHXVHGVRWKDWRQHGLVKLVNHSWDVDVWRFNGLVKDQGFHOOVRIWKHUHPDLQLQJWHQ
FPGLVKHVDUHXVHGWRSURGXFHWZRKXQGUHGPP(&0FRDWHGGLVKHV
,I (&0 LV SURGXFHG RQ D URXWLQH EDVLV WU\ WR SUHSDUH SULPDU\ FXOWXUHV RQFH D PRQWK
0DLQWDLQFRQIOXHQW%&(FXOWXUHVIRUXSWRZHHNVLQWKH&2LQFXEDWRUZLWKQRPHGLXP
FKDQJH7KHFHOOVUHPDLQYLDEOHDQGFDQEHUHDGLO\VXEFXOWXUHGDWDQ\WLPH7KHFHOOVFDQ
DOVREHIUR]HQLQOLTXLGQLWURJHQXVLQJYYGLPHWK\OVXOIR[LGH'062LQVXSSOH
PHQWHGORZJOXFRVH'0(0EXWWKHUHVXOWVDUHRIWHQQRWDVJRRGDVWKRVHREWDLQHGZLWK
IUHVKO\LVRODWHGFHOOV
3UHSDUH(&0FRDWHGPPGLVKHV
!A& @11
$'5'!1$28$B"
21
!BEFE&
'H[WUDQ 7 LV DGGHG WR WKH JURZWK PHGLXP WR LQFUHDVH LWV YLVFRVLW\ DQG KHQFH WKH
SKDJRF\WRVLVH[RF\WRVLVDFWLYLW\RIWKH%&($VDUHVXOWWKHDPRXQWWKLFNQHVVRI(&0
GHSRVLWHGLVDWOHDVWWZRIROGKLJKHUZLWKGH[WUDQ7
([WUDFHOOXODU
0DWUL[
4 4%
!
!7& %9 !1' - % #
@%B5,
#
+7°4@
&
%HVWSURGXFWLRQRI(&0LQWHUPVRIWKLFNQHVVDQGKRPRJHQHLW\LVREWDLQHGZLWKVHFRQGDU\
RUHDUO\SDVVDJH%&('RQRWXVHFHOOVEH\RQGWKHLUVL[WKSDVVDJHLQFXOWXUH
!;& , %B5, @ $'5'!1& @11
$'5'!1$28
!BEFE&
!:& 5('∼(×!1@"$
+@'
& ,
! B EFE % % %$2A
%"&
(1& )
@;
%$
&
5HPRYHFHOOV
(!& >B32/%!1(1+7°4&,
+@'
!%&
,QFHUWDLQH[SHULPHQWV7ULWRQ;DQG1+2+VKRXOGEHDYRLGHGVHH&ULWLFDO3DUDPH
WHUV,QWKLVFDVHFHOOVDUHUHPRYHGZLWK0XUHDLQVXSSOHPHQWHGORZJOXFRVH'0(0
IRUWRPLQDW°&*RVSRGDURZLF]HWDO
((& )
+@$-
'&
*HQHUDOO\FRPSOHWHFHOOO\VLVRFFXUVZLWKLQWRPLQDQGWKH(&0UHPDLQVILUPO\DWWDFKHG
WRWKHWLVVXHFXOWXUHSODVWLF,IFHOOQXFOHLUHPDLQERXQGWRWKHERWWRPRIWKHGLVKLQFXEDWH
IRUDQDGGLWLRQDOWRPLQDW°&WRREWDLQFRPSOHWHVROXELOL]DWLRQRIWKHFHOOOD\HU
(&0SURGXFHGE\%&(FHOOVDSSHDUVDVDGHOLFDWHXQLIRUPQHWZRUNRIDPRUSKRXVPDWHULDO
FORVHO\DVVRFLDWHGZLWKWKHSODVWLF7KHKRPRJHQHRXVQDWXUHDQGWKLFNQHVVRIWKH(&0
ODWWLFHFDQEHDSSUHFLDWHGE\VFUDWFKLQJWKHSODWHZLWKDQHHGOHZKLFKPD\GHWDFKDQGOLIW
WKH(&0PHPEUDQHDORQJWKHHGJHRIWKHVFUDWFK$PRUHILEULOODUQRQKRPRJHQRXVW\SH
RI(&0LVOHVVDFWLYHDQGLVRIWHQSURGXFHGE\ODWHSDVVDJHFHOOV
'HWDFKHG(&0FDQEHXVHGIRUSXULILFDWLRQRUELRFKHPLFDODQDO\VLVKRZHYHUSXULILFDWLRQ
RIDVLQJOHFRPSRQHQWIURPERXQGRUGHWDFKHG(&0LVQRWVWUDLJKWIRUZDUG
(+& >$( &
(2& 44'
$( $@1µB
1&(@µBE9&22°4&
$/7(51$7(
35272&2/
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
35(3$5$7,212)0(7$%2/,&$//</$%(/('(&0
#
% 4& % 4 % 3(+@/2"# R+S R!24S%"# &&#
R!24S"
&
4%9
9
%
4
%
9%#
%&E#'
4'
'-%
% % %
*
-%&#!:;+#!::(#!::@.3-&#!:;2"&
'
4'4# 21 µ4 3(+@/2 @21 @:1 4B.
,"+@'
%(
@
%'
!
4 4%
#!:"#
$
&$
%
#4
#(!(2"&
% 4# EP $ "5+@'4'
&E'
P
C"
$'5'!1
28$B"
!BEFE&-
"&,;184'
%
%#%%&
35(3$5$7,212)+5(&0
)4-
&)%##
$-$
4&
E'6:" $ 4&#!:77"
,44%%
$-$I
*#!:;2.6D&#!:;:"&0-4#
E'6:
##%
4
4
%4"&
>4'4
#6:'4
%
% & D 4 #
# % )*# %& 0- 4'4#
6:'4 # % )))# %#
EFEF
$9&#!:;2.6D&#!:;:"&
) # $ -
%F
$9&#!:;2"#
$ 4!(6D&#!:;:"&
$/7(51$7(
35272&2/
+@'
#%
-& ) # $%
$4'4
6:'4&
$GGLWLRQDO0DWHULDOVDOVRVHH%DVLF3URWRFRO
E'6:,44.4&#!:77.I
*#!:;2"
'5'!1"
,
"
%
E'
"
!& - E'6:
'5'!1
!1'
+7°4
#!184/(&5
-$%B5,@+7°4"
$-%
!J!1&
3)+5FHOOVFDQEHIUR]HQLQOLTXLGQLWURJHQLQKLJKJOXFRVH'0(0FRQWDLQLQJYY
)%6DQGYY'062
(& 5!1'- E'6:$%B5,@
+7°4"# # !11 '
5'!1&
2QHVWRFNFXOWXUHSURYLGHVHQRXJKFHOOVIRUILIW\PPGLVKHV
+& (×!1@+@''
'5'!1$28$B"
21&
([WUDFHOOXODU
0DWUL[
4 4%
!
7KHVHGLVKHVZLOOEHFRDWHGZLWK+5(&03UHFRDWLQJZLWKILEURQHFWLQHQIRUFHVDILUP
DGKHVLRQRIWKH+5(&0WRWKHSODVWLFVXEVWUDWXP
2& )@A
%#
@1µB
%(
2&
@& >B32/%!1(1+7°4&,
+@'
!%&
,QFHUWDLQH[SHULPHQWVGHQXGDWLRQZLWK7ULWRQ;DQG1+2+VKRXOGEHDYRLGHGVHH
&ULWLFDO3DUDPHWHUV
A& ) + @ $ - '&
*HQHUDOO\FRPSOHWHFHOOO\VLVRFFXUVZLWKLQWRPLQDQGWKH(&0UHPDLQVILUPO\DWWDFKHG
WRWKHWLVVXHFXOWXUHSODVWLF,IFHOOQXFOHLUHPDLQERXQGWRWKHERWWRPRIWKHGLVKLQFXEDWH
IRUDQDGGLWLRQDOWRPLQDW°&WRREWDLQFRPSOHWHVROXELOL]DWLRQRIWKHFHOOOD\HU
7KHDSSHDUDQFHRIWKH+5(&0LVVLPLODUWRWKDWRI%&((&0VHH%DVLF3URWRFROVWHS
H[FHSWWKDWLWRIWHQFRQWDLQVFHOOXODUGHEULVDQGLVVOLJKWO\PRUHILEULOODU
7& ,9
$$( &
,I QHHGHG UHPRYH FHOOXODU GHEULV E\ SLSHWWLQJ ,I SODWHV KDYH QRW EHHQ SUHFRDWHG ZLWK
ILEURQHFWLQWKH+5(&0PD\FRPHRIIWKHSODWHVGXULQJZDVKLQJ
;& ,
! @1µB
1&(@µBE9
+@'6:'4&2°42&
6833257
35272&2/
&(//$77$&+0(17$66$<
,#
71+#
% 4& 44C%/-
#4,"#
/
36
60%4&71+
%$9
*
-%
F
$9#!:;!"&/
'
#$%
&)#%
@4#
%
*
-%&#!:;1.*
-%
F
$9#
!:;!.E&!1&2&!"&,
71+
##
%
4&
4'4$
6:'4&
0DWHULDOV
71+*
-%&#!:;1.*
-%
F
$9#!:;!"
'5'!1"
%B5,"
+@'4'
.,
("
$33(1',;$"
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
+@'
4
!& -71+
'5'!1
+7°4#!184/(&
!
4 4%
)LJXUH $GKHVLRQDQGIODWWHQLQJRIKXPDQFRORQFDUFLQRPDFHOOV+67RQ%&((&0$,QWDFW
(&0%¦ZRXQGHG§(&0ZKHUHWKH(&0FRDWLQJZDVVFUDSHGIURPSDUWRIWKHGLVK/RRVHO\DWWDFKHG
DJJUHJDWHV RI VSKHURLG FHOOV DUH VHHQ RQ SODVWLF ZRXQGHG (&0 DV FRPSDUHG WR ILUPO\ DWWDFKHG DQG
IODWWHQHGFHOOVRQ(&03KDVHFRQWUDVWPLFURJUDSKV×ZHUHWDNHQKUDIWHUVHHGLQJ
(& 5$%B5,@+7°4#$
"%
&
+& ! × !1@ +@' 4'
&
2& ,@A1
#
%
$$( &
@& 5%
$!%B5,&
4
$4
P
&
$OWHUQDWLYHO\SUHODEHOVWRFNFXOWXUHVZLWK>+@WK\PLGLQHµ&LPOKUDQGSURFHHGDV
GHVFULEHGDERYH,QVWHDGRIFRXQWLQJFHOOVVROXELOL]HWKHILUPO\DWWDFKHGFHOOVZLWK1
1D2+PLQ°&DQGPHDVXUHWKHUDGLRDFWLYLW\E\OLTXLGVFLQWLOODWLRQVSHFWURVFRS\
)ULGPDQHWDO
&(//352/,)(5$7,21$66$<
,4.
F
$9#
!:7A"
$'%4'4&>
%+11B+@'
"#
4E&!1&2&("#
E&!1&2&(,.F
$9&#!:;1#.6D&#!:;:"&
6:'4&
0DWHULDOV
,%
$
5B4J
$'5'!1"$E
$
$!18B"
%B5,"
+@'
4'4'
"
+&78
%
+78"
1&!(@8$B"%"
6833257
35272&2/
([WUDFHOOXODU
0DWUL[
4 4%
!
A
B
)LJXUH &ORQDO JURZWK RI ERYLQH DRUWLF HQGRWKHOLDO FHOOV RQ $ SODVWLF RU % (&0 &HOO
FRORQLHVZHUHIL[HGDQGVWDLQHGZLWKFU\VWDOYLROHWWHQGD\VDIWHUVHHGLQJFHOOVPP
GLVK
!& )
,4"%
F
$9&#!:7A"#
5B4
+7°4#!184/(&
(& 5-'%@
+7°4$%B5,&6$
&
+& ,
( 5B4 +@' 4'4'
#
$+11#
4/(&
2& ,!1!(
%
#
!
$!+&78
%
&
@& @$!1&!(@8%
$
$&
$WOHDVWFHOOFRORQLHVHDFKFRQWDLQLQJ!FORVHO\DSSRVHGSRO\JRQDOFHOOVVKRXOG
EHVHHQRQ(&0,QFRQWUDVWXQFRDWHGGLVKHVVKRXOGKDYHQRFHOOFRORQLHVRUYHU\VPDOO
DQGOLJKWO\VWDLQHGFRORQLHVFRPSRVHGRIVHQHVFHQWFHOOV
6833257
35272&2/
&(//',))(5(17,$7,21$66$<
% 4!(#%
'
%#
'%
4'4*
-%&#!:;(.6D&#!:;:"&
6:'4&
0DWHULDOV
4!(,44"
'5'!1"
%B5,"
+@'
4'4'
"
!& 4!(
'5'!1
+7°4#
!184/(&
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
)RU3&FHOOV)%6LQWKHPHGLXPVKRXOGEHKHDWLQDFWLYDWHG$33(1',;$
(& 5 ' % @ +7°4 $
%B5,
$
&
!
4 4%
+& ,
(
'5'!1+@'
4'
#
$!×
!1@#
4/(&
2& %$&
([WHQVLRQRIQHXULWHVWRDWOHDVWILYHIROGWKHOHQJWKRIWKHFHOOERG\LVVHHQZLWKLQWR
GD\VRQ(&0,QFRQWUDVWWKHUHLVDOPRVWQRRXWJURZWKRQUHJXODUWLVVXHFXOWXUHSODVWLF
5($*(176$1'62/87,216
&U\VWDOYLROHWVROXWLRQZY
5(&@%!11&
>&+&2&5!B(1(/
%&
)LEURQHFWLQFRDWHGWLVVXHFXOWXUHGLVKHV
(@µB" $33(1',;
$"&,
!
+@'
!
&6
$ &
$ !2°4&6&
*HODWLQFRDWHGWLVVXHFXOWXUHGLVKHV
1&(8$B" $33(1',;$"&,#
#
1&('µ&,
@
!1'
#
≥+2°4&6
$
$!1 &
$ !2°4&6
&
6XSSOHPHQWHGORZJOXFRVH'0(0
'XOEHFFR©VPRGLILHG(DJOHPHGLXP'0(0FRQWDLQLQJJOLWHUJOXFRVHVXSSOH
PHQWHGZLWK
!18B"E$33(1',;$"
@8B"$
@1µB
1&(@µBE9.C"
!2°4
6XSSOHPHQWHGKLJKJOXFRVH'0(0
'XOEHFFR©VPRGLILHG(DJOHPHGLXP'0(0FRQWDLQLQJJOLWHUJOXFRVHVXS
SOHPHQWHGZLWK
!18B"E$33(1',;$"
@10B
@1µB%
!2°4
+HDWLQDFWLYDWLRQRI)%6LVRQO\UHTXLUHGIRU3&FHOOVVHH6XSSRUW3URWRFRO
7ULWRQ1+2+FHOOO\VLVVROXWLRQ
3%6$33(1',;$"FRQWDLQLQJ
1&@8B"Q'!11
(132/
+2°4
([WUDFHOOXODU
0DWUL[
4 4%
!
7U\SVLQ('7$VROXWLRQ
(&@%
1&(5,
;34
1&2I4
!
1&+@34/+
1&1!
(/!
+−(1°4
&200(17$5<
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
49
'
#%#
%#'
%
$##
'
I&#!:;!"&'
%# 4 $ %
9 %
& '
% $ %
# %
4 $''
F
$9 &# !:7;.
)#!::1.%#!::(.5-%
>#
!::("&
$%
%
4 #
%4
&#!::@"&)
$
'
'4
%-# '
%#
%
4&#!::7"&
#9
4%
#
#'
###
&#!::+.*
-%
&#!::+.,&#!::7"&
)4
%
"&&#'
)*# # B# %'
"
&&#$#
9%" 9
4 % '
# % % F. 6 K# !::!. *
-% &# !::+"&
'
%
%4
4#
'
% )# )))# *"#
##
#
'
&
4
-#%
&
4
%
'
$
-#
'
-
4" E'6:
%
#
# - 4
%*
-%
&# !:;1. F
$9 &# !:;2"& 4
%
4%
% & )
%%%)))
)*#$%)
*"#
%%
'
%# $ %"#
###
&*'
%
$
4
$ EFE" B $'
'
F
$9&#!:;1#.*
-%
&#!:;("&0-
&&#"#'
$4'4
F
$9 )#
!:;1"&
!
4 4%
E'6: & ,
#%
4
%
%)*#
%# # &# !:7:. C &# !:;("& 4 '
4'
%4&>4'4
#6:'4
%
%
&0-
4'4#6:'4
#%)))#
%#
EFE6D
&#!:;:"&
, % % 4#4
$$
%9
$B32/&
%4'
% &&# #
"$
$
%&
)#
#
# E'6:
4 # %
%&'
%$
4
&)
# % 4$
4%
% E'6:&#
4'4$
%'
%
&
4%4
4 $ %
%
#
*
-%&#!:;1.F'
$9 &# !:;1#"& %- 4 $ $ % ' '
%#%#'
# '
'
#9'
++(@; F
$9&#!:;+"&'
#4
(2'
4 R+S%
$
&C-$#
'
$49'
B32/" '
4$
4(2&3
4
F
$9
)#!:;1"&
0- 4'4# $%
6:'4&
%
E'6:
$&4'
4
% % &
4
6:'4 % $%
%%
$
B32/&
D
%%F,F"4
%4
'
$ %
9
%
4& , EFE#
4'4
*
-%&#!:;7"
E-&#
!:;;"& EFE $ % '
4
'4
&# !::1.
*
-%&#!::+"&
EFE4'
#
'
#
*
-%&#!::!#
!::+"&
4'
$
%
'
%.
%'
##
'
# % '
*
-% &# !::+". ##
'
*
-%&#!:;+#!::(#
!::@"&4'
%
%$
'
$%$'
F
$9
#
!:;1. F
$9 &#!:;1#"&)
#
4
'
&
,&#!::7"&4'
$ '
'
# F
$9
)#!:;1.F
$9
&#!:;1##!:;2"&%'
([WUDFHOOXODU
0DWUL[
4 4%
!
%'
#$$%
&# !:;+"& 4'4
*
-%&#
!:;(.C9-'I&#!:;+"&
#'
% 4!(#'
% ED &# !:;(. *'
-%&#!:;("&)#$
4#$
&
6:'4%
& ) -
% #%$
F
$9&#!:;2"&6:'4
$-
%'
' '
.
$ '
%"
'
$F
$9&#!:;2"&
&ULWLFDO3DUDPHWHUV
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
% $ # % # -&%
$-
4'
%#$4
$ & 4# - %# '
$
%%'4&
4 %
& ,# 28 21
$ % %B% % 4#
$ 4
& 0- 4'4# E'6: &
6:'4
%
$ & # $6:'4
#
4& ) # $4'4&
)#
$'
Q'!11
32/
&E
#%
'
EFE
4
4# %
4$
%
$
32/&E#
4'
% %&
%'
%(5!1(1
+7°4F
$9&#!:;+"&0
# #
#
%'
#'
%4%
'
*
-% &# !:;7# F'
$9&#!:;+"&
$QWLFLSDWHG5HVXOWV
% $
#%
4#
9 4'4" 6:'4"
& '
$ $
4 B9%
%
4
'
9&
/%4
A'
!1'
-
#%
$
+@'
&E E'
6: # !1' - %+@'
&
7LPH&RQVLGHUDWLRQV
E 4'4# - ; !( % %#;!1
%
-#
∼!1
% 4& ,
4
% E'6: '
&6:'4
@
A
%
#-
4'4#
-EFE&
/LWHUDWXUH&LWHG
,#5&##)&#,9#6&#*
-%#)&#
,
#,&!::7&5
'
&&XUU%LRO7J2+'@!&
#&#$9#,&&#E-#M&#
*
-%#
)&!:;+&'
%
'
$ '
% % ' & ,QW - &DQFHU
+!J@@7'@AA&
!
4 4%
# 3&# %# 4&# # &=& !::@&
EJ6''
%4&7UHQGV&HOO
%LRO@J!'2&
4
'4
#4&#*
-%#)&#9'E
'
#,&#-#5&#
E-#M&!::1&'
$
&/DE,QYHVWA+J;+(';21&
4#4&&#-#&##>
#F&&#
)#5&&!::7&F
&6FLHQFH(7AJ!2(@'!2(;&
4#,&&##C&&#9-#&#-#
=&#C9#4&#
4#4&,&!:77&'
'
% 442'E&&DQFHU
5HV+7J(17('(1;!&
5-%#4&&
>#M&!::(&
'
%'
&4'
&&XUU2SLQ&HOO%LRO2J77('7;!&
E-# =&# I# &# # =&# >
9-#
&#)#5&#
*
-%#)&!:;;&,'
'
O
$O
$'
&$P-3DWKRO!+1J+:+'211&
E
#6&#,#K&#5D-%#6&#E-#M&#
*
-%# )& !:;@& 4 $ % & ([S &HOO 5HV !@;J2A!'
27A&
ED#5&I&##&C&##3&#
F'
$9#5&!:;(&3$
'
% % 4!( $&-1HXURVFL
(J!!@7'!!7@&
F
$9#5&
)#4&6&!:;1&5
$T3URF1DWO$FDG6FL86$77J(7(A'
(7+1&
F
$9#5&
#=& &!:;1&F$
&(QGRFULQH5HY
!J(1!'((7&
F
$9#5&##=&##5&#
'
$#4&6&!:7A&4$'
JE
$ & 3URF 1DWO
$FDG6FL86$7+J2!(1'2!(2&
F
$9# 5&# F# F&# $#
4&6& !:7;& 5 %
$
$&&DQFHU5HV+;J2!@@'
2!7!&
F
$9#5&#5
#5&#
*
-%#)&
!:;1& '
'& 3URF 1DWO
$FDG6FL86$77J21:2'21:;&
F
$9# 5&# *
-%# )& # 3&
!:;1&
'
&-6XSUDPRO6WUXFW
!+J++:'+7(&
F
$9#5&#F99#6&#
ED#5&I&
!:;+& , #
$
'
% T-&HOO3K\VLRO!!2J!:!'(1(&
F
$9#5&#C#=&##&#
>
# )& !:;2& 4 % %'
'
E'
6':
'
-
% &-&HOO%LRO::J:27':A!&
%#6&/&!::(&)J*%#
&&HOOA:J!!'(@&
)# 5&& !::1& E %
$ % & 3URF 1DWO $FDG 6FL 86$ ;7J+@7:'
+@;+&
I#&I&#I#6&=&#
#F&6&!:;!&
6
$&-&HOO%LRO;;J27+'2;@&
I#&6&
*#I&F&!:;2&
'
'
% & - 1DWO &DQFHU ,QVW
7(J;;:';::&
C#)&#,#I&#6#C&#*#,&##6&#
># =& !:;(& %'
%)*'
'
'
&
([S&HOO5HV!+7J!@'(+&
C9-'I#)&#/
#&#*
-%#)&#E-#
M&# ,-%# /& !:;+& $
% J 0 % & %UDLQ
5HV(A7J!@!'!@@&
# =&& =&# E9# &# $9# &,&
::+&
&0RO%LRO&HOO2J:@+':A!&
3-#K&#4#)&6&#E-#M&#
*
-%#
)&!:;2&,
%%
'
'
%
&
1DWXUH+!1J(2!'(2+&
6D#&#I#&#,9#6&#I-$#
&#9#,&#E-#M&#
*
-%#)&
!:;:& $ '
'
4!(&-&HOO
%LRO!1:J;(+';+!&
6#&
K#K&!::!& %'
$ &
&HOOA2J;A7';A:&
# 6&# 6
# &# 6%# &F&# 6
# &)&#
E
#=&&#
#F&6&!:7:&CO
, % & -
%LRO&KHP(@2J::++'::+7&
*
-%#)&
F
$9#5&!:;!&6'
&1DWXUH(;:J+12'+1A&
([WUDFHOOXODU
0DWUL[
4 4%
!
*
-%# )&# C# F&&# F
$9# 5&
!:;1&#$'
%%
&
&HOO!:JA17'A!A&
*
-%#)&#C#,&#C#)&##=&#
E-#M&!:;(&)
'
%
%%%
&'HY%LRO:+J(;@'+11&
*
-%#)&#E-#M&#'3#&#,#K&#
#*&!:;+&C%
% J 6'
& &DQFHU 5HV
2+J(712'(7!!&
*
-%# )&# E-# =&# #6&#E
#
6&#)'#6&##=&#
I#
& !:;7& '
'
$J%
& 3URF
1DWO$FDG6FL86$;2J((:('((:A&
*
-%# )&# '# 6&# )'# 6&#
-# &# E-# M& !::!& $
J , % T 7UHQGV %LR
FKHP6FL!AJ(A;'(7!&
*
-%#)&#
#,&#9'E
#,&#
9# K&# )'# 6&# C# &#
-# &#C
#/&#3-#K&#4#)&6&#
E-#M&!::(&%
%J &,QYDVLRQ0HWDVWDVLV!(J!!('!(7&
*
-%#)&#'#6&#I#F&#
E-#
M&!::+&'
$'
#9%
&,Q
4 , 5&&6
6&#
&"&+(7'+2+&
,
#5&
*
-%#)&##&#C
#/&##4&&#
-#&' &#*
#&#)'#6&#
9#&!::@&)
%&,Q
YDVLRQ0HWDVWDVLV!2J(:1'+1(&
.H\5HIHUHQFHV
F
$9&#!:;1&&
6HHQRWHIRU9ORGDYVN\HWDO
F
$9&#!:;2&&
)LUVWDUWLFOHGHVFULELQJWKHHIIHFWVRI+5(&0RQ
FHOOSUROLIHUDWLRQDQGGLIIHUHQWLDWLRQ
*
-%&#!:;1&&
7KLVDQG*RVSRGDURZLF]HWDODDUHWKHILUVW
WZR DUWLFOHV GHVFULELQJ WKH SUHSDUDWLRQ FRPSRVL
WLRQDQGELRORJLFDOHIIHFWVLQGXFWLRQRIFHOODWWDFK
PHQWIODWWHQLQJDQGSUROLIHUDWLRQRI%&((&0RQ
QRUPDODQGPDOLJQDQWFHOOV
*
-%&#!::+&&
5HYLHZ DUWLFOH HPSKDVL]LQJ WKH ELRORJLFDO VLJQLIL
FDQFH RI (&0UHVLGHQW E)*) DQG RWKHU DFWLYH
PROHFXOHV
4
%)*
-%
'$0%
=#)
3UHSDUDWLRQRI
([WUDFHOOXODU
0DWULFHV
3URGXFHGE\
&XOWXUHG&HOOV
!
4 4%
3XULILFDWLRQRI)LEURQHFWLQ
81,7
%
%
#
#
&%#
#$
##
$%
&
!"9
"
&,@&@
%
$
&
$
'
#&
4 %
("
%, !"
'
'
, ("&
&$87,21 > $- $ # # # %$
&
%
$)*&$%
#%#
$
& $%
&5$
$#
U
V&&#
-
$"-&,%
-$
$!18
%
'
>
% &
385,),&$7,212)3/$60$),%521(&7,1
& ) % % #
# -# # & %
$$
A18&
%
-#
%
%&
!
%
%&
%$6,&
35272&2/
0DWHULDOV
!111+2"
4C'2, "
F', "
A
'
"J1&:8$B"34B!1⋅4#7&@$33(1',;$"
1&(5,#7&1
1&( E"
ε','Q'
4,
"
4"
1&@
#
,#C"
5P $33(1',;$"
4, B."
@''
#≥@#9
$, "
@''
#≥!1#9
$,
FRQWLQXHG
&RQWULEXWHGE\6WHYHQ.$NL\DPD
&XUUHQW3URWRFROVLQ&HOO%LRORJ\!:::"!1&@&!'!1&@&!+
4%<!:::%=>%?#)&
([WUDFHOOXODU
0DWUL[
(
@1'%'
(@1'@11'
5%!(#111!2#111>4/."
'
$F'+
F,"
,
5' ,F81,7"
3UHSDUHFROXPQV
!& -@
≥@"$!114C'2&
>$+11,%%$&
7KLVZLOOEHXVHGDVDSUHFROXPQWRUHPRYHSODVPDFRPSRQHQWVWKDWELQGQRQVSHFLILFDOO\
WRJHODWLQRU6HSKDURVH7KHIORZUDWHFDQEHLQFUHDVHGE\XVLQJWXELQJFRQQHFWHGWRWKH
ERWWRPRIWKHFROXPQWRVLSKRQGRZQWKHFROXPQHOXDWHPRUHUDSLGO\
(& -@
#!1"$(11'&>
'$;11A#;11
,&
%RWKRIWKHSUHFHGLQJVWHSVFDQEHFRPSOHWHGXSWRVHYHUDOGD\VEHIRUHWKHIROORZLQJVWHSV
3UHSDUHSODVPD
+& >+7°4&
$OOVXEVHTXHQWVWHSVDUHSHUIRUPHGDWURRPWHPSHUDWXUHH[FHSWZKHUHQRWHG
2& ,
$J
(@1&(5,@"
@1&( E!"
A&Aε''Q'
@1"&
@& 4!@!1#111×J##
&
8VHWRPOERWWOHVDSSURSULDWHIRUD6RUYDOO*6$RU*6URWRURUHTXLYDOHQW7KH
URWRUDQGWKHFHQWULIXJHVKRXOGEHDWURRPWHPSHUDWXUHSULRUWRDQGGXULQJFHQWULIXJDWLRQ
$IILQLW\SXULI\SODVPDILEURQHFWLQ
A& ,%#
%%
$'$∼@11
B&
7UHDWWKHIORZWKURXJKIUDFWLRQVRIWKHJHODWLQ6HSKDURVHFROXPQDVELRKD]DUGRXVZDVWH
DQGGLVFDUGLQDPDQQHUFRQVLVWHQWZLWKORFDODQGLQVWLWXWLRQDOUHJXODWLRQV
7KH SUHFROXPQ UHPRYHV SODVPD FRPSRQHQWV WKDW ELQG QRQVSHFLILFDOO\ WR WKH JHODWLQ RU
6HSKDURVHEHDGV
7& >
%$;11#$211
,&
;& $
21'
&4-
%%
$(;1"& '#
%
2
1&@
9&
7KLVLVDSRVVLEOHVWRSSLQJSRLQW1HXWUDOL]HGILEURQHFWLQVROXWLRQVFDQEHVWRUHGXSWR
VHYHUDOGD\VDW°&
3XULILFDWLRQRI
)LEURQHFWLQ
7KH\LHOGRIILEURQHFWLQFDQEHLQFUHDVHGE\IROORZLQJWKHFLWUDWHEXIIHUHOXWLRQE\DVHFRQG
HOXWLRQZLWK0XUHDLQ7%67KHXUHDHOXWHGILEURQHFWLQZLOOEHRIORZHUSXULW\DQGXVXDOO\
FRPSULVHV OHVV WKDQ KDOI WKH \LHOG RI WKH FLWUDWHHOXWHG ILEURQHFWLQ KRZHYHU LW LV VWLOO
SRWHQWLDOO\XVHIXOHVSHFLDOO\DVVWDUWLQJPDWHULDOIRUWKHSUHSDUDWLRQRIILEURQHFWLQIUDJ
(
4 4%
PHQWV 7KH XUHD VKRXOG EH UHPRYHG LPPHGLDWHO\ E\ GLDO\VLV EHIRUH WKH ILEURQHFWLQ LV
FRQFHQWUDWHGE\DPPRQLXPVXOIDWHSUHFLSLWDWLRQDVGHVFULEHGLQVWHS
7RVDYHWKHJHODWLQ6HSKDURVHFROXPQIRUUHXVHZDVKWKRURXJKO\ZLWK7%6FRQWDLQLQJ
VRGLXPD]LGHLPPHGLDWHO\DIWHUWKHXUHDHOXWLRQDQGVWRUHLWDW°&7KHFROXPQFDQEH
XVHGUHSHDWHGO\IRUPDQ\\HDUVLILWLVNHSWK\GUDWHG5HXVHGJHODWLQ6HSKDURVHFDQDFWXDOO\
SURYLGHKLJKHU\LHOGVRIILEURQHFWLQWKDQ¦QHZ§JHODWLQ6HSKDURVHEHFDXVHRILUUHYHUVLEOH
RFFXSDQF\RIQRQVSHFLILFELQGLQJVLWHVRQWKHEHDGV
&RQFHQWUDWHILEURQHFWLQE\SUHFLSLWDWLRQRSWLRQDO
:& ) %#%$%
1&(:!
@18"$
%
&C'
$(&
!1& 4!@!1#111×J#2°4&6
%9
% 2°4 $ +&@ % &6
%$+&@&
$ZLGHUDQJHRIFRPPRQEXIIHUVFDQEHXVHGWRVROXELOL]HDQGGLDO\]HWKHILQDOSODVPD
ILEURQHFWLQSURGXFW,WLVJHQHUDOO\HDVLHVWWRXVH7%6RU'XOEHFFR©V3%6,WLVSRVVLEOHWR
DFKLHYHILEURQHFWLQFRQFHQWUDWLRQVJUHDWHUWKDQ∼PJPOLQPRVWSK\VLRORJLFDOEXIIHUV
,IKLJKHUFRQFHQWUDWLRQVDUHQHHGHGILEURQHFWLQLVVROXEOHDWRYHUPJPOLQ&$36VDOLQH
,IQHFHVVDU\ILEURQHFWLQFDQEHVWHULOL]HGE\ILOWUDWLRQDOWKRXJK\LHOGVZLOOEHORZGXHWR
ELQGLQJRIILEURQHFWLQWRWKHILOWHU
!!& 5&
7KLVLVPRVWHDVLO\GHWHUPLQHGE\PHDVXULQJWKH$DQGXVLQJDQH[WLQFWLRQFRHIILFLHQWRI
PO⋅PJ⋅FP−
!(& ,%9%5' ,F81,7"&
,QWKHSUHVHQFHRIUHGXFLQJDJHQWSODVPDILEURQHFWLQDSSHDUVDVDGRXEOHWRIDQG
N'D
!+& E9
%
−;1°4&
6WRUHGDWWKLVWHPSHUDWXUHILEURQHFWLQFDQUHPDLQDFWLYHIRU\HDUVRUPRUH
,IVWRUDJHDW−°&LVQRWDYDLODEOHWKHVHFRQGEHVWRSWLRQLVWRVWRUHILEURQHFWLQDVDOLTXLG
VROXWLRQRQZHWLFH6WRUDJHRIILEURQHFWLQDW−°&VKRXOGEHDYRLGHG¥DWWKLVWHPSHUDWXUH
ILEURQHFWLQWHQGVWRIRUPDFORWWKDWFDQQRWEHUHGLVVROYHG7KDZIUR]HQILEURQHFWLQDW°&
ZLWKPLQLPDOVKDNLQJRUDJLWDWLRQ$YRLGUHSHDWHGIUHH]LQJDQGWKDZLQJ
385,),&$7,212)),%521(&7,1)520&8/785('&(//6
$D'
%
J!"'
% (" % % %
, ("&
# -# # & , % % # '
% & % %&E
%
#,%44&K
%
$
%
%
#%&E
#+8
%-%
%$6,&
35272&2/
([WUDFHOOXODU
0DWUL[
4 4%
(
& ) # ' %
'&
%
;@1'(&,%#!@1'(
!7@'(-
# $ D
%& %
#$$
%
%
&
%&&#-%"%
% % # % $ $&
4%
%'
$!(&)!
#%
$$
%&$(%
%
#%$
&0
+@' A1' %
%
&
,(!#
&
# $ ( %
&
0DWHULDOV
E$
'
-P
.$33(1',;$"
5P
P
5"
!11
1&( E"
0#C)43"
,#C"
1&(4, #!!&1"
4, B"
6
6
'
@1'%'
5%!(#111!2#111>4/."
!&@'%%#@
!(×7@%%
,
5' ,F81,7"&
([WUDFWFHOOV
!& >(111
!@115
$A1!11
2"
!@1&( E("+7°4&
$SSUR[LPDWHO\PO+%66DQGPOVXSSOHPHQWHG'0(0ZLOOEHQHHGHGSHUERWWOH
(& F%$@1$
&5
&
+& ,
(@B5$2
( E
!
+7°4$!
%
&
3XULILFDWLRQRI
)LEURQHFWLQ
(
4 4%
2& (5%
+15'
$2(@1&>+7°4&=
#
(&@1&( E("&
@& 5
'
#
(1B(5$
2
( E
2"
(+7°4
$!&
7KLVVWHSH[WUDFWVWKHILEURQHFWLQIURPWKHFHOOVXUIDFH8VXDOO\0XUHDLVQHHGHGWRH[WUDFW
KXPDQILEURQHFWLQEXW0XUHDFDQEHXVHGZLWKVRPHFHOOVIRUH[DPSOHFKLFNHPEU\R
ILEUREODVWV$OWHUQDWLYHO\0XUHDZLOOH[WUDFWKXPDQILEURQHFWLQDOEHLWDWORZ\LHOGV
+RZHYHU ILEURQHFWLQ H[WUDFWHG ZLWK 0 XUHD LV RIWHQ VXIILFLHQWO\ SXUH WKDW DIILQLW\
SXULILFDWLRQLVQRWQHFHVVDU\
A& 4
@1'%
!@!@#111×J#2°4&#$#
&
3UHFLSLWDWHILEURQHFWLQ
7& $%
1&27(
718"$&,
#∼!$&
;& 4
@1'%
!@!@#111×J#2°4&5
%
%
∼!11&(4, #!!&1&6
D
!!&1$3/
$
&
'LDO\]HSUHFLSLWDWH
:& 5%9
+&@4, B
2°4$&6$&
&DUHPXVWEHWDNHQWRVHDOWKHGLDO\VLVEXIIHUIURPH[SRVXUHWRDLUDVZHOODVSRVVLEOHXVLQJ
6DUDQZUDSRUHTXLYDOHQW7RPLQLPL]HWKHDLUVSDFHLQWKHEHDNHUDERYHWKHGLDO\VLVEXIIHU
PDNHVXUHWKH6DUDQZUDSLVWRXFKLQJWKHWRSRIWKHEXIIHU2WKHUZLVHFDUERQGLR[LGHIURP
WKHDLUFDQGLVVROYHLQWKHEXIIHUORZHULQJLWVS+DQGSUHFLSLWDWLQJWKHFHOOXODUILEURQHFWLQ
!1& 4(1!@#111×J#2°4#%
&
!!& 5%$(;1
!&(;⋅'!⋅'!
%9%5' ,F81,7"&
&HOOXODUILEURQHFWLQVKRXOGSURGXFHDEDQGDW∼N'D
!(& 6
%9$
&
7RDYRLGSDUWLDOQHXWUDOL]DWLRQRIWKH&$36EXIIHUGRQRWXVHGU\LFH&HOOXODUILEURQHFWLQ
VKRXOGEHVWRUHGLQOLTXLGQLWURJHQRUDW−°&,QOLTXLGQLWURJHQFHOOXODUILEURQHFWLQZLOO
UHPDLQDFWLYHIRUVHYHUDO\HDUVLILWLVQRWUHSHDWHGO\IUR]HQDQGWKDZHG&HOOXODUILEURQHFWLQ
VKRXOGEHWKDZHGDW°&ZLWKDPLQLPXPRIVKDNLQJRUDJLWDWLRQ
$)),1,7<385,),&$7,212)(;75$&7('&(//8/$5),%521(&7,1
4 $ &
$#
$#%(
.%&
'
.$33(1',;$"
F', "
@'9
%%
$/7(51$7(
35272&2/
([WUDFHOOXODU
0DWUL[
4 4%
(
!& % (#!A"&
(& -@'9
%%&
'RQRWXVHDQH[FHVVLYHO\ODUJHDIILQLW\FROXPQ5HFRYHULHVRIFHOOXODUILEURQHFWLQWHQGWR
EHJUHDWHUDWKLJKHUUDWLRVRIILEURQHFWLQWRJHODWLQ6HSKDURVHGXHDUHODWLYHGHFUHDVHLQ
QRQVSHFLILFELQGLQJWRWKHEHDGV$POEHGYROXPHLVVXIILFLHQW
+& >'$(&@;#$(@&
2& 5!J+$1&@&
@& ,%%
+'&
A& > '
% $ # - $(;1 #$(;1L1&1@&
7& $;
+'&
%-$(;1&
;& $
% (#:!("&
$/7(51$7(
35272&2/
385,),&$7,212)+80$1&(//8/$5),%521(&7,1)520
&21',7,21('0(',80
5%
&
' $ # '
&'
$&E
%
'
#
%
;@1'(%
%&
&
$%
'
'
464'+1$%C&0'
%#
#
'
P
E'!(
5
$#%#
#
###
,&)#
%
464'+1#$-
%&,$
'
%
%$
&
'
J&&#!J!5BPE!(.
"
F', "
,#%+×%9
.4"
'
.$33(1',;$"
0#C"
4, B"
,
(&@'%%9
#@9
$
,# "
!'%
%%
+'%%
3XULILFDWLRQRI
)LEURQHFWLQ
(
4 4%
3UHSDUHFHOOVDQGFROOHFWFRQGLWLRQHGPHGLXP
!& >(111'
+7°4&
(& F% ' $ @1 $
'
&
+& ,
@1B'
!+7°4$
!
%
&
7KHLQFXEDWLRQSHULRGLQWKLVVWHSFDQEHDVORQJDVRYHUQLJKWZLWKQRGHOHWHULRXVHIIHFWV
2& 5
'
&,
!11'
(22;+7°4$!&
@& 4'#
!11&(
E!111
("&
A& 4
%%!@
(@#111×J#2°4&#$#
&
7KH FRQGLWLRQHG VHUXPIUHH PHGLXP IURP VHYHUDO ERWWOHV FDQ EH SRROHG DQG VWRUHG DW
−°&RUXVHGLPPHGLDWHO\LQWKHQH[WVHFWLRQ
6HFUHWHGFHOOXODUILEURQHFWLQFDQRIWHQEHUHSHDWHGO\KDUYHVWHGIURPWKHVDPHUROOHUERWWOHV
E\DGGLQJPRUHVHUXPIUHHPHGLXP8VXDOO\WKHFHOOVZLOOHYHQWXDOO\VWDUWWRGHWDFKIURP
WKH UROOHU ERWWOHV :KHQ WKLV KDSSHQV WKH\ FDQ EH UHJURZQ WR FRQIOXHQFH E\ DGGLQJ
VHUXPFRQWDLQLQJPHGLXPEHIRUHUHSHDWLQJVWHSVWR$VWKHILEUREODVWVDJHWKHDPRXQW
RIILEURQHFWLQWKH\VHFUHWHZLOOGHFUHDVHVRWKLVSURFHVVFDQQRWEHUHSHDWHGLQGHILQLWHO\
:KHQWKH\LHOGRIILEURQHFWLQEHJLQVWRGHFOLQHWKHFHOOVVKRXOGEHGLVFDUGHG
$IILQLW\SXULI\VHFUHWHGFHOOXODUILEURQHFWLQ
7& '%& -!1'
(&@'%%9
&
;& >'$+1@B,&$
$@1;$
%@1&
7KH FROXPQ LV SUHZDVKHG ZLWK %6$ LQ RUGHU WR EORFN QRQVSHFLILF LUUHYHUVLEOH SURWHLQ
ELQGLQJWRWKHJHODWLQ6HSKDURVH
:& !,'
'$!
(22°4&
7KLVLVPRVWHDVLO\GRQHE\PL[LQJWKHEHDGVDQGPHGLXPWRJHWKHULQOLWHUF\OLQGULFDO
SRO\SURS\OHQHERWWOHVSODFHGRQDUROOHUERWWOHDSSDUDWXVZLWKURWDWLRQDWUSPDW°&
'XULQJWKLVWLPHWKHILEURQHFWLQLQWKHFRQGLWLRQHGPHGLXPVKRXOGELQGWRWKHLPPRELOL]HG
JHODWLQ
$YRLGPHFKDQLFDOVWLUULQJRUH[FHVVLYHDJLWDWLRQRIWKHJHODWLQ6HSKDURVHERWKRIZKLFK
FDQEUHDNXSWKH6HSKDURVHEHDGVGHFUHDVLQJVXEVHTXHQWFROXPQIORZUDWHVDQGWKH\LHOG
RIILEURQHFWLQ
5HFRYHUILEURQHFWLQ
!1& ,$ # # '
(&@'
9
%%&
!!& >'
$@1&
!(& $;#+'&
%-$(;1#$(;1L1&1@&
([WUDFHOOXODU
0DWUL[
4 4%
(
&RQFHQWUDWHILEURQHFWLQRSWLRQDO
!+& ) '
%H!B#
!A&)
#
%9
+&@4, B
$+&6$&
!2& %%-##
$%
1&27(718"$&
,
#!
$&
!@& 4
%%!@
!@#111 ×J#2°4&5
%
%
1&(4, #!!&1&6
D!!&1$
3/
$
&
'LDO\]HILEURQHFWLQ
!A& 5%9 +&@ 4, B 2°4 $
&6$&
&DUHPXVWEHWDNHQWRVHDOWKHGLDO\VLVEXIIHUIURPH[SRVXUHWRDLUDVZHOODVSRVVLEOHXVLQJ
SODVWLF6DUDQZUDS7RPLQLPL]HWKHDLUVSDFHLQWKHEHDNHUDERYHWKHGLDO\VLVEXIIHUPDNH
VXUHWKH6DUDQZUDSLVWRXFKLQJWKHWRSRIWKHEXIIHU2WKHUZLVHFDUERQGLR[LGHIURPWKH
DLUFDQGLVVROYHLQWKHEXIIHUDQGORZHULWVS+
!7& 4%
%(1!@#111×J#2°4&
!;& 5%$(;1
!&(;⋅'!⋅'!
%9%5' ,F81,7"&
!:& E9$
&
7RDYRLGSDUWLDOQHXWUDOL]DWLRQRIWKH&$36EXIIHUGRQRWXVHGU\LFH&HOOXODUILEURQHFWLQ
VKRXOGEHVWRUHGLQOLTXLGQLWURJHQRUDW−°&,QOLTXLGQLWURJHQFHOOXODUILEURQHFWLQZLOO
UHPDLQDFWLYHIRU\HDUVRUPRUHLILWKDVQRWEHHQUHSHDWHGO\IUR]HQDQGWKDZHG&HOOXODU
ILEURQHFWLQVKRXOGEHWKDZHGDW°&ZLWKDPLQLPXPRIVKDNLQJRUDJLWDWLRQ
5($*(176$1'62/87,216
&$36EXIIHUS+0
((&!4, %%
"
∼2@1(/
,
D!!&1$A33/
,
(/@11
$-
$IWHUWRPRQWKVFKHFNWKHS+DQGUHDGMXVWZLWK1D2+LIQHFHVVDU\
&$36VDOLQH
;&;
∼:11(/
@11&(4, #!!&1"
!!
(/!111
!
3XULILFDWLRQRI
)LEURQHFWLQ
$IWHUPRQWKFKHFNWKHS+DQGUHDGMXVWZLWK1D2+LIQHFHVVDU\
(
4 4%
&LWUDWHHOXWLRQEXIIHU
!!&A34
(:&2
#
%
"
∼!;11(/
((18$B"
9
-"
,
D@&@$A34
,
(/(111
!
&RPSRVLWLRQ01D&ODQGP0VRGLXPFLWUDWHS+
&ROXPQEXIIHU$
!(&!⋅4
!+&!ε''Q'
!!&;
#
%
"
((18$B"
9
-"
(/∼!:11
,
D7&A$A34
(/(111
%
$(
%
&RPSRVLWLRQP07ULV⋅&OP0εDPLQRQFDSURLFDFLGDQGP0VRGLXPFLWUDWH
S+
&ROXPQEXIIHU%
@;&2
A&1⋅4
A&@ε''Q'
@&:
#
%
"
∼;11(/
!(18$B"
9
-"
,
D7&A$A34
,
(/!111
%
$(
%
&RPSRVLWLRQ01D&OP07ULV⋅&OP0DPLQRKH[DQRLFDFLGDQGP0VRGLXP
FLWUDWHS+
306)VWRFNVROXWLRQ0
5A:&A E%%%
"(%
"&
%#
%−(1°4&
6RGLXPD]LGH
5(1
9
"$!11&
&
&200(17$5<
%
'
%#!::1.#!:;:.,
K
#!::@. 4#!::2.
,-%&#!::@"&E$
% %
&E#
% '
'
,'F%',6F5"'
''',', 63"
%%$
6F5
α@β!
6# !:;2. K
I
%#!:;2.,&#!::2"&,'
([WUDFHOOXODU
0DWUL[
4 4%
(
3XULILFDWLRQRI
)LEURQHFWLQ
# '
'
%
)))4*"
I&#
!:;@" α2β! & , $
D $ ))'
)4
%O
C','*C5*"
,'F'
,'*65*"&#
!:;7.I%&#!::!"&0-'
6F5
63#C5*
65* %% '
#65*
∼!8C5*"&
E%
'
% # %
&
-$
#'
# -$ @1 %# '
%#
% U
V 4)
&# !:2;"& 5 %
#
& '
# $#
##
%
%
$
&
E
'
##
#$
& ) % " .#%
% % % %
6
!:77"& ! $
%--&!:;("&)
@&@%
$ $ '
&'
#
%
#
'
#
#$%
%%
&
4
%
I &# !:;@"& 4 '
'
& '
$
-%
# $ '
%%K
/
#!:7;#'
'0 &# !::7. &#
!::7"&
,
%%9
# # ##%#
'
%%#!:7+.=
#!:7;.
K
>#!:72.4&#!:77.
5&#!:7;.E&#!:7;.6'
&# !:7+"& 4'
%
#
&*%
%'
9
#%
$
%
'
& '
$
$
&
7URXEOHVKRRWLQJ
% # %
& $
$ %
# $ %
'
&/
%'
%&
$
#
'
∼((°4". %
9
$&
'
%(;1$(;1"
%5' ,F81,7 "
$ 4 6'(@1 &
) # %$'
O%≤!B&,'
$
%9
%
%
'
& %#
'% (
4 4%
4 4'" & E %&
#%
%'
%% E. $33(1',;
%"& E % % %$
#
(11 #$
& E $-
&$# E
# # %&,
E $(1
&
)
#%$
&$%
-#
B'
&
% #
%%%9
-
&,
%
''
# # &
E 9
&)#
%
9
$&E
'9 $
% '
"
−;1°4& E9 −(1°4 %
'
# % $ 9
% & E'
$
% +7°4
$&
1°4 $- 9
$
9−(1°4&
* & # $ #$$
# '
% & 4
% #
%
%
'
%U&V
%!(
%#
%$-%
%'
&
E
$%'
&#$%
%
'
%&
>
#%
% % % & '
$%
7
1&1@8
9
$& E# %
#
U
V$
#
%%
&#$'
% '
9
&
)
- %
'
%#
%
#
&
$ %
'
%'%'
&-%
&4
%&#
$
$& / $% $%'$2
$ $& '
'
$'
# $ &
4 % !!& $& -
#
%
%#
#
$%
# $ &
/%# $ %
%$&'
UV%#%
$$&)#
%%
-'
#$
$&
%
'
([WUDFHOOXODU
0DWUL[
4 4%
(
$%O
O
%
%
K
!:;+"&)#
'
$2&
%
'
' $
$ 2 & $
$% $1&+
& , - @18#%
% &
%$'
%'%
&
'
'
&
$QWLFLSDWHG5HVXOWV
!
%
!11+11
#$%''
&
%%%#
$%9
%&
9
'
#
%
% & ) %
##
$
"%
&
)
%
##
% $ % & C$'' '
;@1'( %%
$!1
(1&
D%'
'4&
# & ,
%
- '
- )# !:;2"& ' % '
% '
'
-#!:;:.
C&#!:;;"&
7LPH&RQVLGHUDWLRQV
3XULILFDWLRQRI
)LEURQHFWLQ
,-≤!
%
'& !$%
'
(+
%&0%
!⁄(
%
- $ & %
%
%& ##
$-
%&
/LWHUDWXUH&LWHG
,-%#&I&#/
#I&#
K
#I&&!::@&
E
'
&&DQFHU0HWDVW5HY!2J!7+'!;:&
,#&
K
#I&&!::@&E
J%'
'
&$GY(Q]\PRO5HODW$UHDV0RO%LRO
71J!'(!&
,#&#39#&#
K
#I&&!::2&
''',',
'
&-%LRO&KHP(A:J(27@A'(27A!&
4# C&&# # 3&# F# &&# 5# =&I& !:77& 5 #
# ' &([S&HOO5HV!1AJ+:'2A&
5#>&##=&##6&#=-#E&#
-#I&!:7;&%
% '
&-&HOO%LRO7:J+2('+@@&
#&
6#&!:77&
'
# '
#&,QW-&DQFHU(1J!'@&
-#& &!:;:&J,'
9
%
&$QQX5HY&HOO%LRO@J7!'
:(&
-#& &
)#=&C&!:;2&,'
&1DWXUH+!!J(A7'(A:&
E# C&&# # 5&E&# >
'
4#F&!:7;&)%9'
C"
CA%JC
&&HOO!+J(A+'(7!&
'0#&#K#K&M&#
#F&#
,-# )& !::7& % )))
))),
&- &HOO6FL
!!1J((7!'((;1&
%#&
K
#I&&!:;+&5'
%' - %LRO &KHP (@;J+++('
++21&
# &=&# I%# ,&# ,-%# &I&#
/
#I&#
K
#I&&!:;7&)
'
$
%)))'
%
&-%LRO
&KHP(A(JA;;A'A;:(&
%#6&/&!:7+&,'
& %LRFKLP
%LRSK\V$FWD2@;J7+'!17&
(
4 4%
%# 6&/& !::1& E& '*#
3$K-&
# 5&E&# & !:;:& E& ,
#3$K-&
=# &,& # 5&E& !:7;& % %
&-([S
0HG!27J!77:'!7:!&
# &5& 6# & !:;2& 4
%
'
%%'
&1DWXUH+1:J+1'++&
I%# ,&# F# C&=&# # &# K
#
&&#K
#I&&#
#&=&!::!&
D
%
4!"$'
%'
% ))) &-
%LRO&KHP(AAJ!@17@'!@17:&
I#,&6&#09$#I&#*' #I&#
#E&&!:;@& %'
J5%'
!1%
&
(0%2-2J!7@@'!7@:&
C#*&,&#6D# &#
-#& &
!:;;& B # # & - &HOO %LRO
!17JA11&
# 6&# /-# 3&# # &# E-
# &# -#I&!::7&
'
%%%
5,&-&HOO%LRO!+:J(:@'+17&
--#&)&#)#I&4&#
#5&!:;(&
6
&7KURPE5HV(7J!'!2&
# =&6& 4# )&5& !::2& E
%& &XUU 2SLQ &HOO %LRO
AJA2;'A@@&
6#&#*#,&#I# &#
C
#&
!:7+&EJ,$
&%LRFKLP%LRSK\V$FWD+((J+@('+@;&
K
#I&&
I
%#5&>&!:;2&5
'
% '
&-&HOO%LRO::J(:'+A&
K
#I&&
/
#I&!:7;&EO
%
&
1DWXUH(7@J!7:'!;2&
K
#I&&
>#=&,&!:72&)
D % '
&3URF1DW$FDG6FL86$7!J+2@('
+2:A&
4
%I&,-%
4#34
# &6&#
#=&&#
#&F&!:2;&
Q*)))& '
)&-$P
&KHP6RF7AJ+!1+'+!1;&
([WUDFHOOXODU
0DWUL[
4 4%
(
3XULILFDWLRQRI9LWURQHFWLQ
%
%
& - U
V%
&
'
'
&
'
"'
%&
'
%%&
81,7
%$6,&
35272&2/
, # 211 '
#$
%
!1(1
%H:;8"%&,#
# & % % %
&
127(/%'
#'
#
$&
#%$
9
$33(1',; "& ,%# O
%%# %# %%O %% %
"& %%
&
&$87,21 > $- $ # %$
&
$)*
&,$%$$%#
%#
&I!18#
>
%# %
&
0DWHULDOV
4C'2,# "
',# "
BB5,"
(34BB5,"
∼211!("'
1&(5,#7&1$33(1',;$"
1&(%%
E.81,7"
:@8#
0
;BB5,"
;B(34BB5,"
('
;B1&+34BB5,"
5P'
5 .$33(1',;$"
(1'
@1'"(&@''
%%9
A11'
!111'%%-
%%
>&(
!1'%%
N99
$, "
5%!(#111!2#111>4/"
&RQWULEXWHGE\6WHYHQ.$NL\DPD
&XUUHQW3URWRFROVLQ&HOO%LRORJ\!:::"!1&A&!'!1&A&@
4%<!:::%=>%?#)&
([WUDFHOOXODU
0DWUL[
(
3UHSDUHFROXPQV
!& (&@''
# ;' 4C'2 & > $ (2 BB5,&
)ROORZPDQXIDFWXUHU©VLQVWUXFWLRQVWRSUHSDUH6HSKDURVH&/DQGKHSDULQ6HSKDURVH
$OOFROXPQFKURPDWRJUDSK\VWHSVLQWKLVSURWRFRODUHDWURRPWHPSHUDWXUH
(& (&@''
#(1''&
>$;1(34BB5,#$;1'
BB5,&
7KHFROXPQVFDQEHZDVKHGXSWRVHYHUDOGD\VLQDGYDQFHDQGNHSWDWURRPWHPSHUDWXUH
LIVRGLXPD]LGHLVDGGHGWRWKHSKRVSKDWHVDOLQH('7$
+& 211'
#
(&@1&(5,
1&@1&( E$
2°4&
2& )
#
@&)
#
A11'%%--
%
2;
'
:@8$&4!1&
7KLVVWHSSUHFLSLWDWHVILEULQRJHQLQSODVPD
@& 4!1!(#111×J#2°4&5
&
A& >&(2°4&
7KLVLVDSRVVLEOHRYHUQLJKWVWRSSLQJSRLQW7KHILOWUDWHVKRXOGEHVWRUHGDW°&
5HPRYHKHSDULQELQGLQJSURWHLQV
7& ,% # # 4C'2%%
&,
#$$(1
BB5,
$&
;& ,%
$
$7'&
, '#$
$@1BB5,
$&
$WWKLVSRLQWWKHSURFHGXUHFDQEHVWRSSHGXQWLOWKHQH[WGD\RUWKHYLWURQHFWLQFDQEH
SXULILHGLPPHGLDWHO\,IWKHYLWURQHFWLQLVQRWWREHSXULILHGLPPHGLDWHO\ZDVKWKHKHSDULQ
6HSKDURVHZLWKPORI01D&OLQSKRVSKDWHVDOLQH('7$WRUHPRYHKHSDULQDVHVRU
JO\FRVLGDVHVWKDWFRXOGGHJUDGHWKHEHDGVDQGVWRUHWKHFROXPQWUHDWHGSODVPDFRQWDLQLQJ
WKHYLWURQHFWLQDW°&RYHUQLJKW7KHFROXPQFDQEHVWRUHGDWURRPWHPSHUDWXUHLI
VRGLXPD]LGHLVDGGHGWRWKH01D&O
7KLVVWHSUHPRYHVSODVPDFRPSRQHQWVVXFKDVILEURQHFWLQWKDWELQGWRKHSDULQLQWKHLU
QDWLYHIRUP
$FWLYDWHWKHYLWURQHFWLQ
:& !111'%%-
+;2$&
;11$BB5,&C
!(
&
7KLVVWHSDFWLYDWHVWKHYLWURQHFWLQVRWKDWLWFDQELQGWRKHSDULQ7KHILQDOXUHDFRQFHQWUDWLRQ
LV0
3XULILFDWLRQRI
9LWURQHFWLQ
!1& >'
#$'
;$!(1;B(34BB5,#
$!(1;BB5,&
(
4 4%
8UHDLVXQVWDEOHLQDTXHRXVVROXWLRQDWURRPWHPSHUDWXUH7KHUHIRUHDOOXUHDFRQWDLQLQJ
VROXWLRQVPXVWEHSUHSDUHGIUHVKHDFKGD\DQGVXEVHTXHQWO\GLVFDUGHGLIQRWXVHGWKHVDPH
GD\WKH\DUHSUHSDUHG
$IILQLW\SXULI\YLWURQHFWLQ
!!& ,%'
!!'$
≤!11B&
7KLVVWHSFDQSURFHHGRYHUQLJKWDVORQJDVSUHFDXWLRQVDUHWDNHQWRHQVXUHWKDWWKHFROXPQ
ZLOOQRWGU\RXW
,I VWHS SURFHHGV RYHUQLJKW SUHSDUH IUHVK VROXWLRQV RI 0 XUHD LQ SKRVSKDWHVD
OLQH('7$VHHUHFLSHDQG0XUHD01D&OLQSKRVSKDWHVDOLQH('7$VHHUHFLSH
!(& >$211;BB5,&
!+& ,
!21µ('(11;BB5,&
!2& >$!11;BB5,
$('&
,IWKLVZDVKLQJVWHSWDNHV≤KUVWRSWKHFROXPQVRWKDWWKHWRWDOH[SRVXUHRIWKHFROXPQ
WR PHUFDSWRHWKDQRO LV DW OHDVW KU WR HQVXUH UHPRYDO RI FRQWDPLQDQWV VHQVLWLYH WR
UHGXFWLRQ/RQJHUWLPHVXSWRRYHUQLJKWFDQDOVREHXVHG
,I VWHS SURFHHGV RYHUQLJKW SUHSDUH IUHVK VROXWLRQV RI 0 XUHD LQ SKRVSKDWHVD
OLQH('7$VHHUHFLSHDQG0XUHD01D&OLQSKRVSKDWHVDOLQH('7$VHHUHFLSH
!@& > $ !11 ; BB5,
$('&
7KHPLQLPDOH[SRVXUHWLPHRIWKHFROXPQWRPHUFDSWRHWKDQROVKRXOGEHKU
!A& > $ !11 ; BB5, $
('&
!7& $;B1&+34BB5,&4
!1'& D
%$(;1
9
9&$$(;1H1&1@&
,WLVUHFRPPHQGHGWKDWWKHIUDFWLRQVFRQWDLQLQJKLJKHUFRQFHQWUDWLRQV$!EHNHSW
VHSDUDWHIURPWKRVHFRQWDLQLQJORZHUFRQFHQWUDWLRQV
'LDO\]HIUDFWLRQV
!;& 5%9$$(;1H1&@"2°4+
%
+&@ 5 # %& 5%9 $'
2°4+
%+&@!185 !5 $:
(/"#
%&
!:& 4(1!(#111×J#2°4#%
&4
&
(1& &
7KHIDVWHVWZD\WRHVWLPDWHWKHSURWHLQFRQFHQWUDWLRQLVWRPHDVXUHWKH$DQGFDOFXODWH
WKHFRQFHQWUDWLRQDVVXPLQJDQH[WLQFWLRQFRHIILFLHQWRIPOPJFP
(!& ,@1!11µ%%
%9&
C $ $ # # &
:KHQWKHWXEHVDUHUHFRQVWLWXWHGZLWKGHLRQL]HGZDWHUWKHWXEHVWKDWFRQWDLQHGYLWURQHFWLQ
LQ'3%6VKRXOGEHUHFRQVWLWXWHGWRRQHWHQWKWKHRULJLQDOYROXPHWR\LHOGKLJKHU
FRQFHQWUDWLRQV RI YLWURQHFWLQ LQ IXOOVWUHQJWK '3%6 5HFRQVWLWXWHG YLWURQHFWLQ FDQ EH
VWHULOL]HGE\DXWRFODYLQJLIQHFHVVDU\
([WUDFHOOXODU
0DWUL[
4 4%
A
5($*(176$1'62/87,216
01D&OLQSKRVSKDWHVDOLQH('7$
(:&234
@1&@
#7&7"
A&+1&(5,#7&1$33(1',;$"
1&(@(18$B"
9
81,7"
59
(/(@1
$-
3KRVSKDWHVDOLQH('7$
7&A
!+1"
(11&@
#7&7.!1"
(@1&(5,#7&1$33(1',;$.@"
!(18$B"
9
1&1(8.81,7"
59
(/!111
$-&
6RGLXPSKRVSKDWHS+0
7:&73( /2⋅!((/
2&(:3( /2⋅((/
59
(/!111
0XUHDLQSKRVSKDWHVDOLQH('7$
2;1
7&A
(11&@
#7&7"
(@1&(5,#7&1$33(1',;$"
59
(/!111
7KHVKHOIOLIHRIWKLVVROXWLRQLVGD\
59
(/@11
0XUHD01D&OLQSKRVSKDWHVDOLQH('7$
A1
!2&734
(&@1&@
#7&7"
+&(1&(5,#7&1$33(1',;$"
59
(/!(@
0XUHD01D&OLQSKRVSKDWHVDOLQH('7$
:A
+&@34
21&@
#7&7"
@1&(5,#7&1$33(1',;$"
59
(/(11
3XULILFDWLRQRI
9LWURQHFWLQ
A
4 4%
&200(17$5<
*#-$
#'#
#
'
%
!11
211µB
$ # !::!"& *
$9&'
A@
7@ -5 & $%&*
,'F%',$#
#'
'
& $'
#
'
%
&4%
% β! $
%
β+&
$
'
% K & !:;;" % % & %
&)D
#$
%
% %
9#!:;+"&
-
# - '
#
$;
& # %%
#
%
%%&
7URXEOHVKRRWLQJ
C$ %
$%
%''
&E#
' '
&#$
%%&,#
-%'
& ) '
# % % #
%
%
&,'
$
$
%$&#
%#'
%
# & 4 $ % %$
$&,$
%-
UV
%&)%'
%
!
( & * $ % & # %
%% 9
#
-#
$
& 0 $%
&E%#
−;1°4$%9#
$%%%
$&
$QWLFLSDWHG5HVXOWV
%
!1
(1%%
%H:;8"& %-
%5'
,F 81,7 "& %
A@
7@-5&
7LPH&RQVLGHUDWLRQV
(+
%&
$!
(
%&4'
&
/LWHUDWXUH&LWHG
# 5&>& 9# =& !:;+& ) & - %LRO &KHP&
(@;J!(@2;'!(@@(&
#I&&!::!&
&$QQX5HY&HOO%LRO7J(7@'+!1&
K#&#)9#&#I$#&#
%#
&!:;;&3
%%'
%&&HOO6WUXFW)XQFW!+J(;!'(:(&
4
%I&,-%
4#34
([WUDFHOOXODU
0DWUL[
4 4%
A
3URWHRJO\FDQ,VRODWLRQDQG$QDO\VLV
81,7
% % %%&
$-$
9#
#
%
$'
&
'
%& %%9&&#
≈!11-5"%
&&#≈!×!1A-5"& %&&#
"#-
%
&&#%"#
&&#%
"#
81,7
"&,#%
9%'
%&&#%'%"
%'
%%"&%%%9
%
$ 9% $
% 9' % 5' ,F81,7"#
81,7"&
&$87,21>$-$
#-
&4%
$%
#$
%
%$33(1',;'"&
127($%%
#
&)
#%
!#:'
%%
%" %% % &&#
%F%%,%#.,4
"&
,62/$7,21$1'385,),&$7,212)3527(2*/<&$16
%
&
,VRODWLRQRI3URWHRJO\FDQVIURP&XOWXUHG&HOOV
$ % "# $ '
% " %"&
%#
, !#$%
#%'
%'
-$%-
&)%
%
#
%
$
$
%
&
%$6,&
35272&2/
% & $ %
# % %# !1''
;
B
"&
127(,
$#
%&
127(,
#+7°4#@8!18
4/(
"$
&
&&#
5"%
4/(7&2&
&RQWULEXWHGE\$QQH:RRGVDQG-RKQ5&RXFKPDQ
&XUUHQW3URWRFROVLQ&HOO%LRORJ\(111"!1&7&!'!1&7&!:
4%<(111%=>%?#)&
([WUDFHOOXODU
0DWUL[
7
0DWHULDOV
6
!1'
"
$33(1',;$"
0#
9"$
$Q'!11
!& 4
%
@1'
&6%$$2 #
$
&
(& 4!1;11×J##%
&
+& ,
2&
2°4.
9&0%&
$OO IXUWKHU VWDJHV ZLWK VXSHUQDWDQW SURWHRJO\FDQV PD\ EH SHUIRUPHG ZLWKRXW GHWHUJHQW
SUHVHQWLQWKHEXIIHUV7KHDXWKRUV©H[SHULHQFHKRZHYHULVWKDWORVVHVDUHGHFUHDVHGE\WKH
SUHVHQFHRIYY7ULWRQ;
2& >%$$(1 &
@& ,
(9
&
!@'&
'RQRWVXEVWLWXWH7ZHHQIRU7ULWRQDVWKLVGHWHUJHQWLVQRWDVHIIHFWLYHLQVROXELOL]LQJ
V\QGHFDQV
A& 6%
%$$(9
$ & ) +1 $ -&
7& 4!1;11×J#&
7KHVXSHUQDWDQWVKRXOGDSSHDUFOHDU,ILWLVQRWFRQWLQXHVKDNLQJDQGLQFXEDWLQJXQWLO
SDUWLFXODWHVDUHQRORQJHUVHHQ3URWHRJO\FDQVHVSHFLDOO\V\QGHFDQVDJJUHJDWHUHDGLO\
DQGPD\IRUPODUJHFRPSOH[HVWKDWFDQEHSHOOHWHGDWKLJKHUJIRUFHLIQRWVROXELOL]HG
;& 5!11'
%
!J@$9'
$
1&(8&
:& 2°4.
9&0%&
$/7(51$7(
35272&2/
,VRODWLRQRI3URWHRJO\FDQ3RROV
%
#
#
##$%
- $ %- & >
&!:;@"&
127(,
$#
%&
127(,
#+7°4#@8!18
4/(
"$
&
&&#
5"%
4/(7&2&
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
'
%'4 )' C4.#
"
C%"$
$
7
4 4%
,VRODWHFRQGLWLRQHGPHGLXPSRRO
!& >
%$$(1 !
+7°4$'
1&!0 )' C4&
7KLV UHOHDVHV JO\FRV\OSKRVSKRWLG\OLQRVLWRO *3,¤OLQNHG SURWHRJO\FDQV 6FKPLGWFKHQ HW
DO
(& 5 !@' & 6 % $ $ (&@ '
$9%
&
+& ,
2 # # &
4
%!!"$(2"&
,VRODWHO\VDWHSRRO
2& ,
(%
!@+7°4&
@& %!@'#%$$!%#
$&
,IGHVLUHGO\VDWHVIURPXSWRWKUHHGLVKHVFDQEHSRROHGLQDVLQJOHPOWXEHDQGXSWR
WZHOYHFDQEHFRPELQHGLQDPOWXEH
A& ,
2
&
7& 5!J!$%$
1&!8&)
&4
%!!"$(2"&
7KLVLVWKH¦O\VDWH§PDWHULDOZKLFKZLOOFRQWDLQK\GURSKRELFSURWHRJO\FDQVQRWUHWDLQHG
GXH WR OLQNDJH WR F\WRVNHOHWRQPDWUL[ DQG SURWHRJO\FDQV XQGHUJRLQJ GHJUDGDWLRQ RU
V\QWKHVLV&HOOOD\HUVZLOOORVHSKDVHGHQVLW\EXWQXFOHLDQGVWUHVVILEHUVVKRXOGUHPDLQ
,VRODWHF\WRVNHOHWDOPDWUL[SRRO
;& 6
$(1 #2°4&
:& ,
!9
&
!1& 5!J!1$9$
1&!8&
)
&4
%!!"$
(2"&
7KLV LV WKH F\WRVNHOHWDOPDWUL[ IUDFWLRQ ZKLFK ZLOO FRQWDLQ K\GURSKRELF SURWHRJO\FDQV
OLQNHGWRWKHF\WRVNHOHWRQPDWUL[DVZHOODVPDWUL[SURWHRJO\FDQV
&RQFHQWUDWHDQGSXULI\SRROV
!!& 4
%%5,'%
("&
7KHVWHSVDERYHDOOFRQWDLQSURWHDVHLQKLELWRUVLQWKHEXIIHUVEHFDXVHFHOOO\VLVLVRFFXUULQJ
+RZHYHULWLVHVVHQWLDOWRSHUIRUPDQLRQLFH[FKDQJHFKURPDWRJUDSK\GLUHFWO\DIWHULVROD
WLRQRIWKHWKUHHSRROVRUSURWHRO\VLVZLOOVWLOORFFXU
5DGLRODEHOLQJRI3URWHRJO\FDQVZLWK62RU>+@*OXFRVDPLQH
%%%"%%
#%
%
$+@/2&
%
21µ4B(+@/2!1@1
!A114B.33C"
$
#
$
$D&&#:@8"&'
#
#
&%
#
&)
%#
&
6833257
35272&2/
([WUDFHOOXODU
0DWUL[
4 4%
7
F
%
%%#R+S'
21A14B.,"
%
%& % !1 µ4B R+S $%&
127(,
$#
%&
127(,
#+7°4#@8!18
4/(
"$
&
&&#
5"%
4/(7&2&
%$6,&
35272&2/
$QLRQ([FKDQJH &KURPDWRJUDSK\ 3XULILFDWLRQ RI 3URWHRJO\FDQV ZLWK '($(
6HSKDFHO
,' % % % !"
, !"&5%%#%
$%5,
&
0DWHULDOV
% !, !"
5%%5,"J@1J@1B"% $33(1',;$"
5,'"
4
#
$(22;
5,$'"
5,"
(×1&@
!& 5,'( ×1&@$∼(+
%
@1J@1%5,' &
7KLVLVVXIILFLHQWIRULVRODWLRQRISURWHRJO\FDQVIURPWKHHTXLYDOHQWRIILYHFPWLVVXH
FXOWXUH IODVNV )RU ODUJHU DPRXQWV RI PDWHULDO¥HJ IURP WR OLWHUV RI FXOWXUH
VXSHUQDWDQWV¥XVHWRPOJHOYROXPHSHUOLWHU
(& ,& $
$!1 %!AB&
$QHFRQRPLFV\VWHPIRUPLQLFROXPQVLVWRSODFHWZRV\ULQJHQHHGOHV*WKURXJKD
UXEEHUEXQJ%UHDNRQHQHHGOHDWWKHPHWDOSODVWLFMXQFWLRQDQGDWWDFKWXELQJ3ODFHD
V\ULQJHRQWKHRWKHUQHHGOHDQGXVHWKLVWRUHPRYHDLUDQGSXOOOLTXLGGRZQWKHWXELQJLQWR
WKHFROXPQ)RUVPDOOHUYROXPHVDUHVHUYRLUFDQEHLQVHUWHGDWWKHFROXPQPRXWK
+& $(5,'&
2& 4
%
(1
(#&
@& >$$@5,'&
A& >$$@5,$'&
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
6LQFHWKLVVWHSUHPRYHVPRVWPDWHULDORWKHUWKDQSURWHRJO\FDQVDOHVVVWULQJHQWSURWHDVH
LQKLELWRUFRFNWDLOLHODFNLQJεDPLQRFDSURLFDFLGDQGEHQ]DPLGLQH⋅&OFDQEHXVHGLQ
WKHORZS+DQGHOXWLRQEXIIHUV
7& >$$@5,&
7
4 4%
;& 6+&
:& C
% & ) #
2°4&
!1& 6$@
A&
!!& $!'5,&
7\SLFDOO\WKHSURWHRJO\FDQVEHJLQWRHOXWHDIWHUPODQGDUHIXOO\HOXWHGE\PO$
VROXWLRQRI0JXDQLGLGLXP⋅&OFDQEHXVHGLQVWHDGRIKLJKVDOWLIWKHSURWHRJO\FDQLVWR
EHFRQFHQWUDWHGE\HWKDQROSUHFLSLWDWLRQ,IVRXUHDVDOWV('7$1(0DQG306)VKRXOG
EHUHPRYHGE\ZDVKLQJZLWK'($(ORZS+EXIIHUZLWKRXWWKHVHLQJUHGLHQWV
!(& %
%
#%
%'
%
"&4%
2°4&59&
'($(6HSKDFHO FROXPQV FDQ EH UHXVHG LQGHILQLWHO\ 7KH WRS WR PP RI JHO FDQ EH
GLVFDUGHGEHIRUHUHXVHLILWEHFRPHVKHDYLO\FRQWDPLQDWHGJUH\LQFRORU&ROXPQVVKRXOG
EHVWRUHGXSULJKWDW°&LQHTXLOLEUDWLRQEXIIHUFRQWDLQLQJZYVRGLXPD]LGH5LQVH
FROXPQVZLWKWZRYROXPHVHTXLOLEUDWLRQEXIIHUSULRUWRXVH,IVWRUHGORQJWHUPPRQWKV
UHFRQGLWLRQDVLQVWHSVWRSULRUWRXVH
,VRODWHGSURWHRJO\FDQVFDQEHGLDO\]HGLQWRQHZEXIIHUVRUFRQFHQWUDWHGE\SUHFLSLWDWLRQ
DQGUHVXVSHQGHGLQWKHEXIIHURIFKRLFH)RUGHWDLOVVHH%DVLF3URWRFROVWHSVDDQGE
,VRODWLRQRI+\GURSKRELF3URWHRJO\FDQV
% '
% %
%
&)
(#$
%%
%B%
%&%
%
(
$&)
#'
5,'
(" & $- $
%
%%%'&
$/7(51$7(
35272&2/
5,'
#RU
("
5,$'"$34
%
%"
%
%"
%
%$"
%
%"
/%' "J@1J@1B"% 1&!8
Q'!11
1&!8B"Q'!11
/%'"$
$34
/%'$"
/%')"
/%'))"
([FKDQJHGHWHUJHQW
!& 55,'!J!1B"$5,$'$34
341&(&
([WUDFHOOXODU
0DWUL[
4 4%
7
(& 65,'$(%
'
%&
7KHFROXPQLVUHHTXLOLEUDWHGWRLQFUHDVHWKHS+DQGWKXVSUHYHQWSUHFLSLWDWLRQRIFKRODWH
+& ,%
&
$OWHUQDWHO\VWDUWH[FKDQJHZKLOHSURWHRJO\FDQVUHPDLQRQ'($(6HSKDFHOIROORZLQJVWHS
RI%DVLF3URWRFRO5LQVHWKHFROXPQZLWKK\GURSKRELFFKURPDWRJUDSK\HTXLOLEUDWLRQ
EXIIHUWRLQFUHDVHS+DQGWKHQSURFHHGZLWKVWHSEHORZ
2& %'
$%$%!11%
'
%&
7KLVLVDVORZSURFHVVRIH[FKDQJHDQGLVEHVWDFKLHYHGXVLQJJUDYLW\UDWKHUWKDQDSXPS
7KHFROXPQZLOOWXUQDEURZQFRORUGXHWRERXQGGHWHUJHQW
@& >$!1%
%$&
A& $(1%
%#!''
&
7& % % "# %&
7KHPRVWFRQFHQWUDWHGIUDFWLRQVZLOOEHEURZQ7KHVHDUHW\SLFDOO\IUDFWLRQVWR
3XULI\K\GURSKRELFSURWHRJO\FDQV
;& %'$
+2%
A;@1J@1% 1&!8Q'!11&
:& >$(11&!8Q'!11#
$!11
%'&
7KLVFROXPQLVVXIILFLHQWIRUDQDO\VLVRULVRODWLRQRIWRFSPK\GURSKRELF
SURWHRJO\FDQVIURPO\VDWHRUF\WRVNHOHWDOPDWUL[SRRO$OWHUQDWH3URWRFROSXULILHGE\
'($(6HSKDFHO%DVLF3URWRFRO,QGLYLGXDOHOXDWHIUDFWLRQVRUSRROHGIUDFWLRQVPD\EH
XVHG
!1& 5
7"1&A34$%'
$34#
%'&)2°4&
7KHPRVWHIILFLHQWELQGLQJLVDFKLHYHGE\DOORZLQJOLTXLGDERYHWKHFROXPQWRGUDLQWKHQ
DGGLQJWRPOVDPSOHDQGDOORZLQJWKLVWRGUDLQLQWRWKHFROXPQEHIRUHFORVLQJWKH
FROXPQ
!!& >$$%!'%'$
&
!(& >$$%!'%')
%
%&
!+& %
% $ $% !' %'
))#!'&
'XULQJWKLVSURFHGXUHWKHRFW\O6HSKDURVHZLOOVZHOODQGVKULQN7KLVLVQRUPDO
)ROORZLQJFRQFHQWUDWLRQE\K\GURSKRELFFKURPDWRJUDSK\SURWHRJO\FDQVPD\EHKLJKO\
DJJUHJDWHG7KLVLVQRWUHYHUVHGE\ERLOLQJ
!2& 0&
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
,IQHFHVVDU\VDPSOHVFDQEHVWRUHGDW°&EXWQRWIRUPRUHWKDQWRGD\V6DPSOHV
VKRXOGQRWEHIUR]HQ
7
4 4%
$1$/<6,62)3527(2*/<&$16
,%%%&
%99
%
&
$QDO\VLVRI3URWHRJO\FDQVE\6L]H([FOXVLRQ&KURPDWRJUDSK\
%
%
%%#%
5' ,F&/
%9''
%2 81,7.@&@&("&
(, (#$
!1"&)
!,
!"# % # % $ &
%$6,&
35272&2/
0DWHULDOV
2 "
4"
%"
-
%J!B
"
1'(#2'
%
53 ""
:1×!
4
!& :1 × ! 71' 2
P& -:B$4
&
(& 5AB&
7KLVUDWHVKRXOGEHXVHGIRUDQDO\VLVWRSUHYHQWFROODSVHRIWKHFROXPQ
+& 6
$
$D
%&,
!%%9
&
)RUDQDO\WLFDOSXUSRVHVWRFSPVKRXOGEHXVHG
2& ,$#%
(4
%&
)DLOXUHWRULQVHZLOOUHVXOWLQVPHDULQJRIVDPSOHWKURXJKWKHJHO
@& 6
AB&4;1!'
%
%'
%
"&
A& 5
91.
"
9.
"%(11µ
%;11µ"
$.%
"
53 '$
$.%
"#
$
%&
9LVWKHYROXPHDWZKLFKEOXHGH[WUDQHOXWHVSHDNRIEOXHFRORUDQG9WLVWKHYROXPHDW
ZKLFK'13DODQLQHHOXWHVODVWIUDFWLRQFRQWDLQLQJ\HOORZFRORU
6HH81,7IRUGHWDLOVRQFDOLEUDWLRQZLWKPDUNHUVIURPDVHSDUDWHUXQ
7& 5.".W9−91"B9−91"#$
9$%-&
([WUDFHOOXODU
0DWUL[
4 4%
7
,I+3/&HTXLSPHQWLVDYDLODEOHD76.FROXPQ7RVR+DDVFDQEHXVHGWRPRQLWRU
SURWHRJO\FDQV+RZHYHUORVVHVFDQEHODUJHLIDSUHILOWHULVXVHGVLQFHPDQ\SURWHRJO\FDQV
ZLOOWHQGWRELQGQRQVSHFLILFDOO\WRWKHSUHILOWHU
%$6,&
35272&2/
$QDO\VLVRI*O\FRVDPLQRJO\FDQ6L]H)ROORZLQJ$ONDOLQH(OLPLQDWLRQ
E%%F,F"
%-"
$ @" % & ,-
#%%&
0DWHULDOV
%, (%"
"
!3/B(32
!1
5%#>4/!('!(#111"
!& )RUGLDO\VLV%
$%
%9
!11$(22°4&0
%$>4/
!('!(#111&
!& )RU SUHFLSLWDWLRQ % $ 2 ;18 −(1°4#$$;18#
%&6
1@
$&
(& ,
!3/B(32&)!A2°4&
+& 39$!1
#-$≈7&
2& 6$#
&9$&
@& ,%9F,F9
%%9'% +.
81,7" C4&
3/J,IVDPSOHVDUHQRWQHXWUDOL]HGVWHSDQDO\WLFDOFROXPQVZLOOEHUXLQHG
%$6,&
35272&2/
$QDO\VLVRI*O\FRVDPLQRJO\FDQ6L]H)ROORZLQJ3DSDLQ'LJHVWLRQ
#%%&)
9&
0DWHULDOV
%, (%"
"
(1B"
A@°4$
,
%
2"
!& %
2#
!!"&
(& )$(B!AA@°4&
+& ,%9F,F9
%%9'% +.
81,7" C4&
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
1RHQ]\PHLQDFWLYDWLRQLVUHTXLUHG
7
4 4%
$QDO\VLV RI *O\FRVDPLQRJO\FDQ &RQWHQW DQG 3URWHLQ &RUH E\ *$* 'HJUDGDWLRQ
ZLWK/\DVHV
%$6,&
35272&2/
%9%%
%%&
)))$
%%&4
,4
%$
%%&
4
,4))%$
%'
%&4
%$
&9%
%%#
)))
$$-%$%%
%9
&/
"%
!11µB
%&-
2
@"#$%9
&
0DWHULDOV
%, (%"
4
"
)))"
4
%4
A'
."
4
,442&(&(&2"#,4))42&(&(&@"#4"
--,"
)))&-&&).42&(&(&;.--
,"
,
%
2"
)RUWUHDWPHQWZLWKFKRQGURLWLQDVHV
!& %
2#!
!"&
6DPSOHYROXPHVGHSHQGRQWKHSURWHRJO\FDQFRQFHQWUDWLRQLQWKHVDPSOHDQGWKHIXWXUH
PHWKRG RI DQDO\VLV )RU LPPXQREORWWLQJ D PLQLPXP YROXPH LV QHHGHG LH UHVXVSHQG
SUHFLSLWDWHGVDPSOHIURP%DVLF3URWRFROVWHSELQWRµORIDSSURSULDWHEXIIHU
)RU6(&RU+3/&XVHWRPO≥FSPLIUDGLRODEHOHG
6LQFH FKHPLFDO DPRXQWV RI LVRODWHG SURWHRJO\FDQV DUH XVXDOO\ VPDOO µJ OHYHOV WKH
DPRXQWVRIFDUULHUDQGHQ]\PHGRQRWQHHGWREHYDULHG
,ISHUIRUPLQJGXDOGLJHVWVZLWKFKRQGURLWLQDVHVDQGKHSDULQDVHXVHKHSDULQDVH,,,EXIIHU
&KRQGURLWLQDVH%LVLQKLELWHGE\SKRVSKDWH
(& ,
∼(µ
%4
,4,4))#
%
&
7KLVLQFUHDVHVWKHHIILFLHQF\RIGLJHVWLRQDQGUHFRYHU\LIORZDPRXQWVRIPDWHULDOHJ
UDGLRODEHOHGSUHSDUDWLRQVDUHXVHG
+& ,
1&@!0
,4#,4))#
+7°4&
,QFXEDWLRQ WLPHV GHSHQG RQ WKH DPRXQW RI SURWHRJO\FDQ SUHVHQW )RU XS WR ∼ µJ
SURWHRJO\FDQDQLQFXEDWLRQRIWRKUVKRXOGVXIILFHZKHUHDPRXQWVDUHKLJKHUDGGLQJ
DVHFRQGDOLTXRWRIHQ]\PHIRUDIXUWKHUKUPD\EHQHHGHG
2& %%
%9'%
+" C4
$$9"#%'
;"&
([WUDFHOOXODU
0DWUL[
4 4%
7
)RUWUHDWPHQWZLWKKHSDULQDVH,,,
!& %))) 2#!
!"&
6DPSOHYROXPHVGHSHQGRQWKHSURWHRJO\FDQFRQFHQWUDWLRQLQWKHVDPSOHDQGWKHIXWXUH
PHWKRG RI DQDO\VLV )RU LPPXQREORWWLQJ D PLQLPXP YROXPH LV QHHGHG LH UHVXVSHQG
SUHFLSLWDWHGVDPSOHIURP%DVLF3URWRFROVWHSELQWRµORIDSSURSULDWHEXIIHU
)RU6(&RU+3/&XVHWRPO≥FSPLIUDGLRODEHOHG
6LQFH FKHPLFDO DPRXQWV RI LVRODWHG SURWHRJO\FDQV DUH XVXDOO\ VPDOO µJ OHYHOV WKH
DPRXQWVRIFDUULHUDQGHQ]\PHGRQRWQHHGWREHYDULHG
(& ,
(@µ$%&
'RQRWDGGKHSDUDQVXOIDWHLIDOVRGLJHVWLQJZLWKFKRQGURLWLQDVHV7KHSURGXFWVRIKHSDUDQ
VXOIDWHO\DVHVDUHVOLJKWO\LQKLELWRU\WRFKRQGURLWLQDVHV'XDOGLJHVWLRQVPD\WKHUHIRUHQRW
EHDVHIILFLHQWDVVHSDUDWHLQFXEDWLRQV
+& ,
!(0)))
+7°4&
7LPHRILQFXEDWLRQYDULHVDVLQGLFDWHGIRUFKRQGURLWLQDVHWUHDWPHQW
2& %%
%9'%
+" C4
$$9"#%'
;"&
%$6,&
35272&2/
7UHDWPHQWZLWK1LWURXV$FLGWR'HJUDGH+HSDUDQ6XOIDWH
3
$
%%&
&$87,21
#'
&
0DWHULDOV
!3(/2
1&!!2B3/("(
%, (%"
!133/
!& !!3(/21&@"$!1&!!2B3/("(!1
#-%&
(& 6/2%(@11×J#2°4&'
&
7KHVXSHUQDWDQWLVQLWURXVDFLGDWS+7KLVLVVWDEOHIRUDIHZKRXUVRQLFH
+& !J2B"
(&
7KHSURWHRJO\FDQVDPSOHVKRXOGEHXQEXIIHUHGSUHIHUDEO\LQZDWHU
2& 39$!133/≈7#-$&
@& % 9' % +" C4
$$9"#%'
;"&
3/J,IVDPSOHVDUHQRWQHXWUDOL]HGVWHSDQDO\WLFDOFROXPQVZLOOEHUXLQHG
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
7
4 4%
$QDO\VLVRI*$*7\SHDQG&RUH3URWHLQ
5' ,F81,7"
81,7"
%9F,F%
%&
% $ $ % #%
%&4
%
UV
%%'
&
--,"
∆"
F,F
$
∆
'1"$
2A
∆
'2∆
'A"&$
&
% % & %
% 5' ,F %# #$
''$
%$$%#
$%
F,F$&
%$6,&
35272&2/
3%
#%$
%& %
$9%
#%%&
4#$#
%'
%&
9'%
%
%9
%&)F,F$91
F'@1
#$
%9%$
9&5%
%
"F,F
&) C4#I2111 C4"%
&
$QDO\VLVRI*$*6L]H
%
C4#
F'(11#4C'A#
%
C4&
%$6,&
35272&2/
0DWHULDOV
C4"
%#9%%%
2
@"
5,' ("
5'!1', "
'
C4"$I2111
"
!& )F,F%$
2 @"# ' % 5,'
("&
(& 5 5'!1'
P'
#
$#
%9&
+& I2111 C4$ C4&
2& 9 C4&C
!11@11µ
$1&@B&41&@'&
,VRODWHGJO\FRVDPLQRJO\FDQVGRQRWELQGWKHSUHILOWHU
([WUDFHOOXODU
0DWUL[
4 4%
7
@& ,%
%
%
"&
$QDOWHUQDWH+3/&UXQQLQJEXIIHUVHHUHFLSHFRQWDLQLQJ0JXDQLGLXP⋅&OPD\EHXVHG
WRSUHYHQWVHOIDVVRFLDWLRQRI*$*V
*$*VWDQGDUGVRINQRZQVL]HVKRXOGEHXVHGVLQFHJOREXODUSURWHLQVDUHLQDSSURSULDWH
IRUFDOLEUDWLRQ
%$6,&
35272&2/
,PPXQRSUHFLSLWDWLRQRI3URWHRJO\FDQV
,%
%
81,7"#&E#%
%
#%%
/&#!::7"&
%
%
'#
$&#!:::"&
#F,F%
%
&0%
%
%#$
%9
,#%%
&,%#F,F
%
9%%&%
%F,F#&
5($*(176$1'62/87,216
$OWHUQDWH+3/&UXQQLQJEXIIHU
+;(&!(
⋅42"
A&1@@1"
(/:11
,
DA&1$!4
,
@Q'!111&@8B"
,
(/!
!2°4
&KRQGURLWLQDVHEXIIHU
A&1@@1"
2&1;
+1"
1&!,.1&!B"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D;&1$!
,
!1&(%%%
E"$
%
1&("
,
!Q'!11%
%.1&!8B"
$%
,
(/!
!$-2°4
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
'($(HOXWLRQEXIIHU
;7&AA34!&@"
2&!1(
@1"
(212"
FRQWLQXHG
7
4 4%
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D2&1$!
,
!1&(%%%
E"$
%
1&("
,
!Q'!111&!8B"
$%
,
(/!
!$-2°4
'($(ORZS+EXIIHU
!!&A:341&("
2&!1(
@1"
(212"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D2&1$!
,
!1&(%%%
E"$
%
1&("
,
!Q'!111&!8B"
$%
,
(/!
!$-2°4
'($(6HSKDFHOHTXLOLEUDWLRQEXIIHU
!!&A:341&("
@1⋅4#;&1$33(1',;$"
(212"
!&(@1'%
3.!1"
1&7;+9
⋅4@"
!+&!(ε'
1&!"
7&225,(1"
(/:11
,
D;&1$!4
,
!1&(%%%
E"$
%
1&("
,
!Q'!111&!8B"
$%
,
(/!
!$-2°4
+HSDULQDVH,,,EXIIHU
!&+A
1&!"
1&111!@;1&1!"
!!8$B"
1&11@@1µB"
1&1!@79
⋅4!"
(/:1
FRQWLQXHG
([WUDFHOOXODU
0DWUL[
4 4%
7
,
D7&1$!
,
!11µ1&(%%%
E"$'
%1&("
,
!11µQ'!11%
%.1&!8B"
$%
,
(/!11
(
%2°4
+3/&UXQQLQJEXIIHU
!+&AI( /2#A&11&!"
(:&(341&@"
(M$,
.1&(8$B"
(/!
!2°4
+\GURSKRELFFKURPDWRJUDSK\HOXWLRQEXIIHU
2
1&28$B"
!7@&(34+"
(&2(((1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D7&+$!4
,
!1&(%%%
E"$
%
1&("
,
(/!
!$-2°4
+\GURSKRELFFKURPDWRJUDSK\HTXLOLEUDWLRQEXIIHU
@&;2341&!"
(&2(((1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D7&+$!4
,
!1&(%%%
E"$
%
1&("
,
(/!
!$-2°4
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
+\GURSKRELFFKURPDWRJUDSK\H[FKDQJHEXIIHU
(1
(8$B"
(&2(((1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D7&+$!4
,
!1&(%%%
E"$
%
1&("
,
(/!
!$-2°4
7
4 4%
+\GURSKRELFFKURPDWRJUDSK\ZDVKLQJEXIIHU
2
1&28$B"
(&2(((1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D7&+$!4
,
!1&(%%%
E"$
%
1&("
,
(/!
!$-2°4
/\VLVEXIIHU
+&1+(@"
!&(@1'%
3.!1"
1&7;+9
⋅4@"
!+&!(ε'
1&!"
7&225,(1"
(/:11
,
D7&@$!4
,
!1&(%%%
E"$
%
1&("
,
(Q'!111&(8B"
$%
,
(/!
D
2FW\O6HSKDURVHHOXWLRQEXIIHU,
%'$"#
D34+
&
2FW\O6HSKDURVHHOXWLRQEXIIHU,,
!7@&(34+"
(&2(((1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D7&+$!4
,
!1&(%%%
E"$
%
1&("
,
!1Q'!11!8B"
$%
,
(/!
!$-2°4
2FW\O6HSKDURVHHTXLOLEUDWLRQEXIIHU
2
1&28$B"
+@&12341&A"
(&2(((1"
!&(@1'%
!1"
FRQWLQXHG
([WUDFHOOXODU
0DWUL[
4 4%
7
7&225,(1"
(/:11
,
D7&+$!4
,
!(1%%%
E"$
%
1&("
,
(/!
!$-2°4
2FW\O6HSKDURVHZDVKLQJEXIIHU
@&;2341&!"
(&2(((1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D7&+$!4
,
!1&(%%%
E"$
%
1&("
,
(/!
!$-2°4
3DSDLQGLJHVWLRQEXIIHU
!!&A:34("
1&7225,!1"
1&2!1(
@1"
1&1!7@%⋅41&1!"
(/:1
,
D@&@$!
(/!11
D
6L]HH[FOXVLRQFKURPDWRJUDSK\6(&UXQQLQJEXIIHU
!51&!8$B"
(1&27341&+@"
A&1A@1"
!&(@1'%
3.!1"
7&225,(1"
(/:11
,
D;&1$!4
,
!1&(%%%
E"$
%
1&("
,
(/!
($-
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
6ROXELOL]DWLRQEXIIHU
!1Q'!11!8B"
(212"
+&1+(@"
!&(@1'%
3.!1"
7&225,(1"
1&7;+9
⋅4@"
FRQWLQXHG
7
4 4%
!+&!(ε'
1&!"
(/:11
,
D7&@$!4
,
!1&(%%%
E"$
%
1&("
(/!
!
%2°4
&200(17$5<
% % $%
9
%
$
&0%%#$
%%
%%
& 3$# 9
#
%%
%
%
'
%%&$#
%%%
-#
%%
%%
$
&
%%%
%
%9%
'
&$#
'
%%
O%9'
# # O %
&E
#
*&
!:::"&
)%#%'
%#$%&
'%'
$
#
$
$ % '
$
>
4#!::;.'
&#!:::. 81,7"&)
$'
%
#
'
$&
&ULWLFDO3DUDPHWHUV
%'
%# 9 '
& ) '
##
%
'
% & ) # %'
% ##%
%
&,#
&
7URXEOHVKRRWLQJ
4#%
%
#5'#
$&
$% 5'
%%
$& $ '
&)
'
.'
%&
9%
%'
%
% %
&
>%
%#9%
%
#
#
9
&)#
9%
-
'
%
&
P
'
&
$QWLFLSDWHG5HVXOWV
%DVLF3URWRFROK
&,'
(#
$%
'
∼2×!1A"#$'
%
!×!1A&
%
%'
% !1'
"&
$OWHUQDWH3URWRFROK
'
'
& , (# $%
∼2×!1A"#$%%
'
$ ∼:@1#111 "# %-B %
∼!@1#111 "& % ([WUDFHOOXODU
0DWUL[
4 4%
7
%
% !1'
"&
6XSSRUW3URWRFROVDQG%
%9$%
%&E#'
%
@A×!1AB@
&
%DVLF 3URWRFRO ,
%
!
, !&
$OWHUQDWH3URWRFROC
%'
5,'
&)
#
$
&K
$
$%&%%#
%#218A18
% % (18 218 %'
-B%#'
%#%'$
'
&
%DVLF3URWRFRO)
!'%
%9'%#
'
%&
%DVLF 3URWRFRO ,- % %%
&
%DVLF3URWRFRO%'
$
%%&
%DVLF3URWRFRO%'
$%$
%
$
'
#
%
&
%DVLF3URWRFRO%'
$
$''$
&
%DVLF 3URWRFRO ) $
$ %% %&
%DVLF 3URWRFRO C4 % $
F,F9&
%DVLF 3URWRFRO %
#%
$&
7LPH&RQVLGHUDWLRQV
3URWHRJO\FDQ
,VRODWLRQDQG
$QDO\VLV
%DVLF3URWRFRODQG$OWHUQDWH3URWRFRO
$
L+1 - % &
/
5,'
$
$$'
#%-
&
6XSSRUW3URWRFRO,
!@#+2
%%
$%&
%DVLF3URWRFRO
%
5,$
&
H!AB&>
'
-∼+1&,
#
#
$
A"&
$OWHUQDWH 3URWRFRO - "
% 5,'
$
&6
H!A B & % '
& C
%'
#
$
∼+
&
%DVLF3URWRFRO9''
%-∼!A&
% &
%DVLF 3URWRFRO,-'
!A#$@!19&
%DVLF3URWRFRO !A&
%DVLF3URWRFRO)
%&E
!1µ%#(+
&
>#
9%(%
&
%DVLF 3URWRFRO $ (&@&
%DVLF 3URWRFRO 9% '
(&@ & 5' ,F# # '
(2&
%DVLF 3URWRFRO C4 % ∼2#$4!A!;&
%DVLF3URWRFRO)'
∼2&
/LWHUDWXUH&LWHG
# &# F# &# -# &>&# 69# /&#
E9
# &C&# C# =&# M-# &
!:::&E
%&$QQX5HY%LRFKHPA;J7(:'77;&
/# &&# >
# ,&# 4# =&6& !::7&
%
'2%
'
% - 4& - %LRO
&KHP(7(J;!++';!+A&
# ,&# # 6&# E#C&,&
!::1&,%'
$ 71 -5 -
%& *O\FRFRQMX
JDWH-7J@A+'@7(&
# *&# C3# E&# # &# ,9# 5&/&#
C9#E&##4&#
6#&!:::&
7
4 4%
' 45;!Bα2β!#45!@!Bα+β!#45!@!BαAβ!"'
&
%LRFKHP-+21J!1+'!!!&
*#6&6&# %#5&,&##&#$
#
=&=&#C
#0&#
F#=&&!:::&'
%
&%LRFKHP-++:J7A7'77+&
>
#,&
4#=&6&!::;&%
J
% & 7UHQGV
&HOO%LRO;J!;:'!:(&
>
# ,&# 4# =&6&# XX-# & !:;@&
% %
& , %
% - %'
- & - %LRO
&KHP(A1J!1;7('!1;7:&
.H\5HIHUHQFHV
)99#6&*&!::;&%JE'
& $QQX 5HY
%LRFKHPA7JA1:'A@(&
$QLQGHSWKLQIRUPDWLYHUHYLHZRIH[WUDFHOOXODUPD
WUL[SURWHRJO\FDQV
C
#0&#I'F#&#
IDY#C&
!::;&6
%&-
%LRO&KHP(7+J(2:7:'(2:;(&
$FRQFLVHUHYLHZRIKHSDUDQVXOIDWHJO\FRVDPLQR
JO\FDQFRPSOH[LW\DQGELRV\QWKHVLV
,QWHUQHW5HVRXUFHV
$$$&%&&D
$:HEVLWHVXSSRUWHGE\6HLNDJDNX&RUSWKDWFRQ
WDLQVDUWLFOHVRQJO\FRVFLHQFHLQFOXGLQJSURWHRJO\
FDQVDQGOLQNVWRRWKHUVLWHV
&#!:::&&
$Q H[FHOOHQW UHYLHZ RI FHOO VXUIDFH JO\SLFDQ DQG
V\QGHFDQSURWHRJO\FDQV
F#!::7&'%
&%LRFKHP6RF7UDQV(@J!(1A'
!(1:&
4
%,>
=6&4
0%,
#,
$QH[SHUWUHYLHZRIKHSDUDQVXOIDWHJO\FRVDPLQR
JO\FDQVWUXFWXUHDQGIXQFWLRQ
([WUDFHOOXODU
0DWUL[
4 4%
7
Matrix Metalloproteinases
1
1
UNIT 10.8
2
Henning Birkedal-Hansen, Susan Yamada, Jack Windsor, Anne Havernose
Pollard,3 Guy Lyons,4 William Stetler-Stevenson,5 and Bente Birkedal-Hansen5
1
National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland
University of Indiana School of Dentistry, Indianapolis, Indiana
3
University of Copenhagen School of Dentistry, Copenhagen, Denmark
4
Kanematsu Laboratories, Royal Prince Alfred Hospital, Sydney, Australia
5
Center for Cancer Research, National Cancer Institute, National Institutes of Health,
Bethesda, Maryland
2
ABSTRACT
Matrix metalloproteinases are a class of enzymes that play an important role in the
remodeling of the extracellular matrix in development and cancer metastasis. This unit
describes a set of methods—cell-mediated dissolution of type-1 collagen fibrils, direct
and reverse zymography, enzyme capture based on α2-macroglobulin and TIMP-1 and
-2, and demonstration of cryptic thiol groups in metalloproteinase precursors—that are
used to characterize the functions of matrix metalloproteinases and their inhibitors. Curr.
C 2008 by John Wiley & Sons, Inc.
Protoc. Cell Biol. 40:10.8.1-10.8.23. Keywords: matrix metalloproteinases r type-1 collagen r zymography r
α2-macroglobulin r TIMP-1 and -2
This unit describes a set of methods that are relatively unique to studies of matrix
metalloproteinases (MMPs) and their inhibitors (TIMPs, α2M), including cell-mediated
dissolution of type I collagen fibrils (see Basic Protocol 1), direct and reverse zymography
(see Basic Protocols 2 and 3), enzyme capture techniques based on α2-macroglobulin
(α2M) and TIMP-1, and -2 (see Basic Protocol 4 and Alternate Protocol), and detection
and demonstration of cryptic thiol groups in MMP precursors (see Basic Protocol 5).
Support Protocols are included for preparation (see Support Protocol 1) and labeling of
collagen with a fluorophore (see Support Protocol 2).
DISSOLUTION AND DEGRADATION OF COLLAGEN FIBRILS BY LIVE
CELLS
BASIC
PROTOCOL 1
Comparatively few methods allow detailed analysis of how live cells orchestrate MMP
and inhibitor functions in the degradation and remodeling of extracellular matrices. The
methods described in this protocol were developed to study the function of matrix metalloproteinases (MMPs) in the degradation of type I collagen fibrils by live cells under
controlled but readily variable conditions. In its simplest form, cells are seeded on a
few-micron-thick film of reconstituted collagen fibrils, then incubated for a period of
1 to 7 days. The progressive dissolution of the film under the cell layer—in response,
e.g., to changing environmental conditions, inducing agents, or inhibitors—may be monitored directly and related to the level of expression of key components of the requisite
proteolytic machinery. The system is readily manipulated in a number of ways: by induction/repression of transcription of components of the signaling and effector systems;
by transfection of new genes of potential importance to the process; or by selective or
specific blocking strategies using antisense-, MMP-specific inhibitor–, or antibody-based
approaches. The limited susceptibility of type I collagen fibrils to cleavage and dissolution by MMPs permits one to narrow the scope of the investigation to a small number
Extracellular
Matrix
Current Protocols in Cell Biology 10.8.1-10.8.23, September 2008
Published online September 2008 in Wiley Interscience (www.interscience.wiley.com).
DOI: 10.1002/0471143030.cb1008s40
Copyright 10.8.1
Supplement 40
of (“collagenolytic”) enzymes. This characteristic also makes it a realistic objective to
dissect the entire sequence or set of reactions involved in cell-mediated dissolution of collagen fibrils, starting from the initial engagement of cell surface receptors by cytokines,
growth factors, and other catabolic reagents, through the final enzymatic cleavage, dissolution, and disposal of the substrate. Important questions that may be addressed using
this approach include the following:
a. What enzymes are actually involved in the cleavage reaction itself and in the
precursor activation steps?
b. How do cells regulate the activity of the enzymes?
c. What role is played by TIMPs in modulating, containing, and blocking the response?
d. What is the ultimate fate of the collagen chains and peptides generated as a result
of proteolysis?
Recent studies have shown that type I collagen (in solution or in reconstituted fibrillar
form) may be cleaved by a larger number of enzymes than previously anticipated,
including the three classical “collagenases,” MMP-1, MMP-8, and MMP-13 (BirkedalHansen et al., 1993; Knäuper et al., 1996). In addition, reports suggest that MMP14 (Ohuchi et al., 1997) and TIMP-free MMP-2 may also dissolve collagen fibrils at
meaningful rates under physiologic conditions (Aimes and Quigley, 1995). It is of note
that although the three classical collagenases (MMP-1, MMP-8, and MMP-13) were
discovered because of their ability to dissolve reconstituted fibrils of type I collagen,
no definitive proof has yet been rendered that cleavage of collagen fibrils is indeed the
exclusive or even prevailing biologic function of any of these enzymes. Admittedly, the
evidence seems compelling based on a large number of in vitro studies.
Earlier versions of this method have been published (Birkedal-Hansen, 1987; BirkedalHansen et al., 1989, 1993; Lin et al., 1987). The isolation and purification techniques of
type I collagen and the methods for formation of reconstituted hydrated gels of type I
collagen have been described elsewhere in detail (Birkedal-Hansen, 1987). The method
relies on the ability of neutral solutions of type I collagen in an appropriate concentration
range (0.1 to 5 mg/ml) to form hydrated gels of reconstituted fibrils by heating to 37◦ C.
The method also takes advantage of the observation that such loose hydrated gels may be
collapsed by gentle air-drying into a thin film of uniform, densely packed, randomly oriented fibrils which remain as highly resistant to proteolysis by enzymes such as trypsin,
chymotrypsin, and plasmin as hydrated gels or natural fibrils (Fig. 10.8.1). Trypsin, which
is often used as a standard for testing the resistance of collagen fibrils to “unspecific” proteolytic cleavage, is unable to dissolve the collagen fibril films prepared as described. The
same is true for a large number of proteinases of all four classes, and it is this unique resistance to proteolysis which renders this assay system particularly valuable, as it greatly
reduces the number of proteinases that are involved in the cleavage/dissolution reaction.
Several variants of the method may be used. While the authors often prefer (for ease of
presentation and interpretation) to seed the cells in a small button in the middle of a much
larger dish (35 mm; Fig. 10.8.1A, middle) in order to maintain medium excess, it is also
possible to seed the cells over the entire collagen-coated surface, although a confluent
monolayer rapidly exhausts the medium. The collagen may be used in its natural state or
labeled either with radioactive or fluorescent tags to facilitate monitoring (see Support
Protocol 2), retrieval, and quantification of dissolved collagen chains and fragments.
Matrix
Metalloproteinases
Depending on the casting conditions, collagen films may be generated with a thickness
down to 1 to 2 μm, which is approximately the thickness of a single layer of well-spread
cells. Most cell types seeded on this film spread within minutes to hours, although often
10.8.2
Supplement 40
A
B
C
Figure 10.8.1 Reconstituted collagen fibril film. (A) Rat tail tendon type I collagen is polymerized by heat
gelation. The gel is air dried and reduced in thickness to a few microns. Cells are seeded in the middle of
the plate and incubated with culture medium. After incubation, the cells are removed and a clearing beneath
the cell layer is exposed by staining with Coomassie blue. (B) The air-dried collagen fibril film consists of
uniform, randomly oriented reconstituted fibrils. (C) Detail of cell attached to the collagen fibril film.
Figure 10.8.2 Dissolution of collagen fibrils by live adherent cells. Fibroblasts seeded in the
center of the well dissolve the underlying collagen fibril film. Upper left panel shows scanning
electron micrograph of fibroblast attached on collagen fibril film. Recreated from HavemosePoulsen et al. (1998).
more slowly than on plastic. Cells that express an appropriate complement of MMPs
either constitutively or after exposure to cytokines and growth factors (or phorbol ester)
progressively dissolve the underlying fibril coating, and, within 24 to 96 hr, clear a path
to the plastic surface (Fig. 10.8.1A, lower; Fig. 10.8.2). Coomassie blue staining of the
residual collagen fibril film after removal of the cells is usually sufficient to visualize the
dissolution of the underlying film (Fig. 10.8.2).
Materials
3 mg/ml rat tail tendon type I collagen in 13 mM HCl (see Support Protocol 1)
13 mM HCl, 4◦ C
Neutralizing buffer (see recipe), 4◦ C
Extracellular
Matrix
10.8.3
Supplement 40
Phosphate-buffered saline (PBS) without Ca2+ and Mg2+ (CMF-PBS; APPENDIX 2A)
supplemented with 100 U/ml penicillin G and 100 μg/ml streptomycin sulfate
Cells of interest (e.g., fibroblasts, keratinocytes, or tumor cells)
DMEM (APPENDIX 2A) supplemented with 100 U/ml penicillin G and 100 μg/ml
streptomycin sulfate with and without 10% (v/v) FBS (or other medium
appropriate for cell type)
Growth factors/cytokines: e.g., IL-1β, TNF-α, TGF-α, or TPA; or phorbol ester
(12-O-tetradecanoylphorbol-13-acetate, TPA, or phorbol myristate acetate,
PMA)
1% (v/v) Triton X-100
0.05% (w/v) trypsin/0.53 mM EDTA (Invitrogen)
Coomassie blue stain (see recipe)
6-well cell culture plates
Additional reagents and equipment for trypsinizing and counting cells (UNIT 1.1)
Prepare collagen-coated plates
1. To cast one 6-well plate, dilute 1 ml of 3 mg/ml type I collagen stock solution with
7 ml of 13 mM HCl at 4◦ C. Mix the collagen solution with 2 ml of cold neutralizing
buffer in a precooled test tube either by gently pipetting up and down while avoiding
formation of air bubbles (which will form defects in the gel) or by gently inverting
the tube several times.
The neutralizing buffer is designed to bring the pH of the solution to 7.4 (check
with pH paper). The concentration of this buffer is 0.2 M inorganic phosphate (as
Na2 HPO4 /NaH2 PO4 ) and 0.47 M NaCl. The final collagen concentration is 300 μg/ml
in 40 mM Pi /∼0.10 M NaCl. Since pH dramatically influences the gelling properties of
the collagen solution it is often advantageous to first test the efficacy of the neutralizing
buffer by mixing 4 volumes of 13 mM HCl with one volume of neutralizing buffer and
checking the final pH (7.4).
The final thickness of the collagen film depends on the concentration of the collagen
solution. A 300 μg/ml solution dispensed at a volume of 1.5 ml per (35-mm) dish yields
a film of 1.5 to 2.0 μm in thickness after drying. Higher concentrations yield thicker
films. The lower concentration limit for proper gelling is around 100 μg/ml using rat
tail tendon collagen prepared as described (see Support Protocol 1) but somewhat higher
(500 μg/ml) with commercial type I collagen preparations.
2. Immediately after mixing, add a 1.5-ml aliquot of neutralized collagen solution to
each well of the 6-well culture plate. Rotate the plate to permit the collagen to cover
the entire well bottom evenly. Incubate in humidified incubator for 2 hr at 37◦ C.
Avoid movement of gel and plate during gelling.
3. Remove plate from incubator, remove lid, and place at room temperature in an air
stream (laminar flow hood) overnight (during this process the gel dries down to a thin
film). Wash three times with distilled water, each time for 30 min at room temperature
or 37◦ C, to remove salt crystals formed during the drying (check efficacy of washing
step using a phase-contrast microscope). Dry again overnight in laminar flow hood
and check for absence of residual salt crystals.
It is important that all salt crystals be removed by washing before the plates are used.
4. Add 2 ml CMF-PBS or DMEM supplemented with penicillin/streptomycin. Store in
this solution in incubator at 37◦ C or in refrigerator at 4◦ C in closed plastic bag to
prevent evaporation.
The plate can be stored in this manner for up to 2 weeks as long as evaporation is avoided.
Matrix
Metalloproteinases
5. Immediately before seeding cells, remove medium from wells by aspiration and wash
with 2 ml distilled water for 30 min. Remove water and leave plate to air dry in hood.
10.8.4
Supplement 40
Plate cells
6. Trypsinize and count cells (see UNIT 1.1), then dilute cell suspension to the appropriate
concentration in DMEM/10% FBS, or in medium appropriate for the cell type being
used.
Best results are obtained with 10,000 to 50,000 cells in a 25-μl aliquot, using a cell
suspension of 4 × 105 to 2 × 106 cells/ml, somewhat depending on cell size. The intent is
to form a coherent monolayer in a small central button (Fig. 10.8.1A, middle).
7. Deliver a 25-μl aliquot to the center of the well without touching the fragile collagen
film. Fill plate volume between wells with distilled water to avoid evaporation
during seeding and attachment. Place plate in plastic box on wet paper towels to
avoid evaporation, and then place in incubator for 5 hr or overnight at 37◦ C to allow
cells to attach.
8. Add to each well 2 ml DMEM/10% FBS or appropriate medium and incubate
overnight at 37◦ C to allow cell spreading.
Some cell types can be transferred immediately to serum-free medium while others require
overnight incubation in serum-supplemented medium.
Once the cells are spread, incubation may be performed either with or without serum.
The result depends somewhat on cell type. Some cells tend to detach in the absence of
serum while others can be maintained for 2 to 3 days in complete absence of serum while
degrading the collagen fibril matrix.
9. If the experiment is to be performed in the absence of serum, thoroughly and repeatedly wash with CMF-PBS or serum-free DMEM for 10 min at 37◦ C, to remove
remnants of serum.
Some cells may require special media formulations, i.e., keratinocytes. Most fibroblast
strains do well under serum-free conditions either in DMEM or DMEM/F12 (1:1).
Induce expression of MMPs
10. Induce cells for expression of MMPs at this stage by including in the medium
cytokines such as IL-1β (10−9 M), TNF-α (10−8 M), TGF-α (10−8 M), or TPA (1 to
2× 10−7 M).
Alternatively, cytokine or TPA induction may be achieved during the last 24 hr before
trypsinization and seeding. If incubated under serum-free conditions, plasminogen may
be added to the medium. Some cells respond to exposure to plasminogen by greatly
accelerating the rate of dissolution, while others do not. If desired, plasminogen is added
from a stock solution in CMF-PBS to give a final concentration of 4 μg/ml. Human
plasminogen is either purchased from one of several commercial sources (i.e., Pharmacia
Hepar or Sigma-Aldrich) or prepared as described (Deutsch and Mertz, 1970) from
outdated human plasma by lysine-Sepharose chromatography.
11. Incubate the plates at 37◦ C for 1 to 4 days (or up to 7 days) depending on the experimental design. Follow the progress of the process with a phase-contrast microscope.
To avoid evaporation it may be advantageous to fill the volume between the wells with
sterile distilled water.
Stain plate and quantitate results
12. In order to visualize the dissolution of the film beneath the cell layer, remove the
cells either by dissolution in 1% (v/v) Triton X-100, by 0.5% trypsin/0.53 mM EDTA
(10 min, 37◦ C), or by a combination thereof.
Avoid use of SDS, which dissolves the collagen fibril film as well as the cells.
13. Rinse the wells with distilled water.
Extracellular
Matrix
10.8.5
Supplement 40
14. Stain with Coomassie blue stain for 5 to 15 min to visualize residual collagen film,
then wash three times with distilled water.
15. Destain in distilled water for 30 min (or perform three quick washes with water) and
finally allow plates to air dry.
After drying the plates, they can be stored indefinitely (Fig. 10.8.2).
In order to follow the progressive dissolution of the collagen fibril film it is advantageous
to terminate sample wells on consecutive days and to contrast the dissolution after 1, 2,
3 ... days.
If desired the plates can be scanned directly into Adobe Photoshop using a scanner
capable of scanning transparent originals.
16. Determine the extent or rate of dissolution of the substrate
The degree of dissolution at the conclusion of the experiment may be measured photometrically in Coomassie blue–stained plates by measuring the absorption of light in a
conventional light microscope equipped with a exposure (photo)meter as described in
Havemose-Poulsen et al. (1998). The relationship between amount of collagen present on
the plate and exposure time is strictly linear at least up to three times the collagen layer
thickness used in this protocol.
Alternatively, if the cells are seeded evenly as a confluent monolayer over the entire
collagen-coated well bottom (see below), progression may be monitored daily by removal
of aliquots of medium and measuring the release of collagen chains and peptides. To
this end the collagen may be labeled either with 3 H (Birkedal-Hansen, 1987; BirkedalHansen and Danø, 1981) or with fluorescent tags (Ghersi et al., 2002). This approach is
less useful if the cells are seeded in a small 2- to 4-mm button at the center of the well,
because the background release of radioactivity and fluorescent label from the entire film
compromises the sensitivity of the analysis (typically only 10% to 20% of the collagen
fibril film is covered by cells in this variation).
SUPPORT
PROTOCOL 1
PREPARING RAT TAIL TENDON COLLAGEN TYPE I
Methods for isolation and preparation of rat tail tendon type I collagen have been described in detail elsewhere (Birkedal-Hansen, 1987; Birkedal-Hansen and Danø, 1981).
Alternatively, rat, bovine or human type I collagen may be purchased from Becton
Dickinson Biosciences Discovery Labware. Briefly, tendons teased from rat tails are
washed with distilled water and with 0.5 M NaCl. The acid-soluble collagen fraction is
then extracted in 0.5 M acetic acid, and type I collagen is purified by sequential salt precipitation at neutral to slightly alkaline pH, first with 5% NaCl, then (after redissolution
in acetic acid) with 0.02 M Na2 HPO4 .
NOTE: All protocols using live animals must first be reviewed and approved by an
Institutional Animal Care and Use Committee (IACUC) or must conform to governmental
regulations regarding the care and use of laboratory animals.
Materials
Tails of ∼400 g rats (freshly removed or stored frozen at −80◦ C)
0.5 M NaCl in 50 mM Tris·Cl, pH 7.4 (see APPENDIX 2A for Tris·Cl)
5 mM, 50 mM, and 0.5 M acetic acid
NaCl (solid)
0.02 M Na2 HPO4
13 mM HCl
Neutralizing buffer (0.2 M NaPi )
Matrix
Metalloproteinases
10.8.6
Supplement 40
Glass wool or cheesecloth
500-ml centrifuge bottles
High-speed centrifuge (Sorvall with SS-34 and GSA rotors, or equivalent
centrifuge and rotors)
10,000 to 14,000 MWCO dialysis membrane
One large (25-liter) or several smaller (4-liter) dialysis tanks
Sterile scissors
125-ml glass Wheaton bottles
Additional reagents and equipment for dialysis (APPENDIX 3C)
Extract collagen
1. Skin 10 to 20 rat tails and place tails on ice. Break tails at joints and tease out
individual collagen fibers. Wash in large volume distilled water (2 to 3 liter) for 1 hr
with agitation. Change wash water three to four times.
The yield is 200 to 400 mg collagen per rat.
2. Extract overnight at 4◦ C with agitation in 2 liters of 0.5 M NaCl/50 mM Tris·Cl, pH
7.4. Discard extract and repeat step.
3. Discard second salt extract and wash collagen fibers extensively (over a 3-hr period
with change two to three times per hr) in distilled water to remove salt.
4. Extract overnight at 4◦ C with slow agitation in 2 liters of 0.5 M acetic acid.
5. Remove insoluble remnants by filtration through glass wool or cheesecloth, then
centrifuge in 500-ml bottles for 30 min at 11,000 × g, 4◦ C. Add solid NaCl little by
little to a final concentration of 5% w/v (50 g/liter) under constant vigorous stirring.
When the salt is completely dissolved, turn off stirrer, cover beaker, leave in cold
room overnight, and let precipitate gather at bottom of vessel.
The collagen immediately starts to precipitate upon addition of the salt.
6. Collect precipitate by centrifugation for 30 min at 11,000 × g, 4◦ C. Discard supernatant.
7. Redissolve collagen by adding 450 ml of 0.5 M acetic acid to first centrifuge bottle,
transfer liquid to the second bottle, and so on, until collagen is redissolved/redispersed
into ∼900 to 1000 ml in 0.5 M acetic acid.
8. Stir vigorously overnight at 4◦ C until collagen is completely dissolved.
If not dissolved overnight, add more acetic acid and bring volume up to 1600 to 1800 ml.
Dialyze collagen solution
9. Place collagen solution, 300 to 400 ml at a time, in dialysis bags. Dialyze in tank
against 25 liters of 0.5 M acetic acid, then for 3 to 4 days against 50 mM acetic acid.
Change daily and mix content of bags.
See APPENDIX 3C for additional details on dialysis.
10. Dialyze against several changes of 0.02 M Na2 HPO4 in the 25-liter tank over the
next 72 hr.
Precipitation should happen as fast as possible, so change solution frequently in the
beginning and massage bags frequently to facilitate even distribution of reagents. The
collagen precipitates as a thick white gel.
11. Harvest precipitate by centrifugation in 500-ml bottles for 30 min at 11,000 × g,
4◦ C. Redissolve collagen in 0.5 M acetic acid by vigorous stirring overnight at 4◦ C.
12. Dialyze 3 to 4 hr against 0.5 M acetic acid, then overnight against 50 mM acetic
acid, and, finally, overnight against several changes of 5 mM acetic acid.
13. Centrifuge 1 hr at 11,000 × g, 4◦ C. Lyophilize supernatant and store in dessicator
at −80◦ C.
Extracellular
Matrix
10.8.7
Supplement 40
14. Redissolve as follows.
a. Weigh out no more than 150 mg collagen.
b. Cut into 1-cm pieces with sterile scissors.
c. Place collagen pieces into a 125-ml glass Wheaton bottle that has been autoclaved
with a stir bar inside.
d. Add cold 13 mM HCl to make a 3 mg/ml solution and stir briskly at 4◦ C with
occasional shaking for ∼24 hr.
The collagen solution should be slightly opalescent.
15. Centrifuge solution for 20 min at 50,000 × g, 4◦ C, to remove any insoluble material,
if necessary.
Note that the solution remains somewhat opalescent even after centrifugation. This solution may be stored for months at 4◦ C. Freezing should be avoided.
SUPPORT
PROTOCOL 2
LABELING OF COLLAGEN
Rat tail tendon type I collagen may be labeled using [3 H]acetic anhydride as described
in detail in Birkedal-Hansen and Danø (1981) and Birkedal-Hansen (1987), or with
fluorescent reagents. The following fluorescent labeling method was adapted from a
technique devised by the Chen laboratory (G. Ghersi and W.T. Chen, unpub. observ.).
Materials
3 mg/ml rat tail tendon type I collagen originally dissolved in or dialyzed into 13
mM hydrochloric acid (see Support Protocol 1)
Neutralizing buffer (see recipe)
Borate buffer: 0.05 M NaB4 O7 ·10H2 O, pH 9.3, containing 0.04 M NaCl, filter
sterilized
20 to 30 mg tetramethylrhodamine-5-(and 6)-isothiocyanate (TRITC) or
fluorescein isothiocyanate (FITC) stock solutions, dissolved in DMSO
Phosphate-buffered saline (PBS; APPENDIX 2A)
20 mM and 1 M hydrochloric acid, sterile
125-ml glass Wheaton bottle, autoclaved
Platform shaker
1. Mix 8 ml of 3 mg/ml rat tail tendon type 1 collagen with 2 ml neutralizing buffer in
a sterile 125-ml bottle and incubate at 37◦ C overnight to form a gel.
Rat tail tendon type I collagen may be prepared as described in Support Protocol 1, or
purchased from BD Biosciences; bovine skin and human placental type I collagen are
also available from the same supplier.
2. Wash for 1 hr with sterile borate buffer at room temperature by rotating at low speed
on a platform shaker.
3. Remove buffer and replace with 10 ml borate buffer containing 2 to 3 mg/ml TRITC
or FITC (prepared from 20 to 30 mg FITC or TRITC predissolved in a small volume
of DMSO). Incubate at room temperature with gentle shaking for 20 to 30 min or
until the dye diffuses through the gel. Protect from light from this point onward.
4. Wash with multiple changes of PBS at room temperature with rotation on a platform
shaker for several days to remove free dye. Wash out salts with several changes of
water.
At some point, the collagen gel may become detached from the bottle. If so, pipet off
solutions carefully in order to avoid breaking up the collagen.
Matrix
Metalloproteinases
If unbound dye is not throughly washed from the collagen gel, subsequent experiments
may be marred by high background fluorescence.
10.8.8
Supplement 40
5. Add a sterile magnetic stir bar and stir to dissolve the gel in 8 ml pre-chilled 20 mM
hydrochloric acid at 4◦ C. If the collagen is reluctant to dissolve, make sure the pH is
around 2. If necessary adjust the pH by the addition of 1 M hydrochloric acid.
Collagen can also be labeled with Alexa Fluor dyes using protein-labeling kits containing amine-reactive dyes from Molecular Probes/Invitrogen. Neutralize and gel 2.5 mg
collagen as described in step 1, above, and equilibrate with several 10-min washes in
PBS. Add 1 M bicarbonate buffer, pH 8.3 to 0.2 M. Resuspend a vial of Alexa Fluor dye
(premeasured to label 1 mg protein) in a small amount of PBS and add immediately to the
collagen. Incubate for 2 hr at room temperature and then wash extensively and redissolve
as above.
Collagen is labeled in the fibrillar state so that sites important for subsequent alignment
and gelling are not being blocked by the labeling procedure. Consequently, collagen
labeled in this fashion readily dissolves in dilute acid and gels again upon neutralization
and mild heating. Depending on the need, the fluorescently labeled collagen may be
diluted up to 10-fold with unlabeled rat tail tendon collagen and still yield a strong
enough signal for quantification.
GELATIN/CASEIN ZYMOGRAPHY
Zymographic methods are designed to analyze the proteolytic capacity of latent and
active MMPs (Heussen and Dowdle, 1980; Birkedal-Hansen and Taylor, 1982; BirkedalHansen, 1987). This set of techniques is based on a number of unique properties of
MMPs: (1) MMPs retain (or refold to display) catalytic activity after electrophoresis in
SDS-containing buffers as long as heating and reduction are avoided (Birkedal-Hansen
and Taylor, 1982); (2) brief exposure to SDS opens the “cysteine switch” (Springman
et al., 1990; Van Wart and Birkedal-Hansen, 1990) so that both precursor and proteolytically truncated (“activated”) forms of the enzyme display catalytic activity; and
(3) MMP catalytic activity is reversibly inhibited by SDS and readily restored when
SDS is removed by washing with Triton X-100 (Birkedal-Hansen and Taylor, 1982).
It is therefore possible to resolve a heterogenous group of MMPs and non-MMPs in
SDS-containing gels copolymerized with a suitable substrate (gelatin, casein), remove
the SDS, and develop (without distinction) the spontaneous or latent catalytic activity
associated with each electrophoretic band. After appropriate incubation (to allow for
proteolysis), the discrete bands of substrate lysis are made visible by Coomassie blue
staining of the gel (Fig. 10.8.3). SDS opens the “cysteine switch” but instantly inhibits the
switch-open enzyme and blocks autolytic truncation normally associated with activation.
The proenzyme bands therefore migrate at their expected high-molecular weight, but
display proteolytic activity because the switch is unable to again “close” after removal
of the SDS with Triton X-100.
BASIC
PROTOCOL 2
Zymography using gels containing 0.1 to 1.0 mg/ml gelatin are by far the most sensitive.
Gels may either be purchased (Invitrogen) or prepared as described below. Gelatin works
particularly well for MMP-2 and MMP-9, whereas MMP-1, MMP-3, MMP-7, MMP-8,
and MMP-10 are better identified in casein-containing gels.
Materials
Gelatin (bovine skin, Sigma-Aldrich type B6-6269) or casein (Sigma-Aldrich,
technical, C-0376)
2.0 M Tris·Cl, pH 8.8 (APPENDIX 2A)
30/0.8 acrylamide/bisacrylamide (UNIT 6.1)
Glycerol
10% (w/v) SDS (APPENDIX 2A)
TEMED
10% (w/v) ammonium persulfate
Extracellular
Matrix
10.8.9
Supplement 40
MMP preparation of interest (for standards, use 1 to 5 ng purified MMP)
5× electrophoretic sample buffer (see recipe)
Electrophoretic running buffer (see recipe)
Gel washing buffers 1 to 4 (see recipe)
Coomassie blue stain (see recipe)
Gel destaining solution (see recipe)
50-ml centrifuge tubes
57◦ C water bath
Whatman no. 1 filter paper or 0.5-μm syringe filter
Gel washing tray of appropriate size
Additional reagents and equipment for preparing SDS-PAGE gels according to
Laemmli (UNIT 6.1)
NOTE: The following procedure is based on a standard 10% SDS-PAGE according to
Laemmli (Laemmli, 1970; UNIT 6.1) using a 4% stacking gel and a pH 8.3 running buffer.
It is important to avoid heating and/or reduction during sample preparation and running
of the gel.
casein zymography
gelatin zymography
culture medium
A
B
D
+
0
+
5
+
15
+
30
– APMA
30 min
Coomassie
blue
proMMP-9
proMMP-2
MMP-10 MMP-3
00
Q
d-t
E2
yp
e
H1
94
S
t
an
wil
C
m
ut
w
ild
-ty
pe
MMP-2
E
lung extract
Coomassie
blue
Figure 10.8.3 Zymography. (A) Zymography using gelatin-containing polyacrylamide gel. Culture
medium containing proMMP-2 (left) or MMP-2/proMMP-2 and proMMP-9 (right). The proenzymes
display catalytic activity because exposure to SDS during sample preparation opens the cysteine
switch. (B) Detail showing conversion of proMMP-2 to MMP-2 by exposure to aminophenylmercuric
acetate. From Caterina et al. (2000). (C) MMP-2 and MMP-9 activity in extracts of lungs of wildtype mice (left) or mice in which the TIMP-2 gene has been mutated to inactive form (modified from
Caterina et al., 2000). (D) Zymography using casein-containing polyacrylamide gel. (pro)MMP3 and MMP-10 cleave casein embedded in the gel (modified from Windsor et al., 1993). (E)
Casein zymogram of mutant and wild-ype MMP-1. Inactivation of catalytic activity by mutation
of catalytic site glutamic acid (E) to glutamine (Q) that abolishes casein cleavage. A histidine to
serine replacement outside the active site does not. Modified from Windsor et al. (1994).
Matrix
Metalloproteinases
10.8.10
Supplement 40
1. Weigh out appropriate amount of gelatin (for 0.1 to 1.0 mg/ml final concentration)
or casein (for 1.0 mg/ml final concentration) and place in a 50-ml centrifuge tube.
2. For every 10 ml of solution to be prepared, add 4 ml of 2.0 M Tris·Cl, pH 8.8, and
6 ml water. Dissolve by heating in a 57◦ C water bath. Filter through Whatman no. 1
filter paper or syringe filter.
3. Prepare the 10% resolving gel (also see UNIT 6.1) by adding the following to 10 ml
filtered gelatin or casein solution (0.2 to 13 mg/ml in 0.8 M Tris·Cl, pH 8.8; see step 2):
6.6 ml 30/0.8 acrylamide/bisacrylamide
2 g glycerol
0.2 ml 10% (w/v) SDS
13.3 μl TEMED
67 μl 10% (w/v) ammonium persulfate
Pour resolving gel as described in UNIT 6.1.
4. Prepare 4% stacking gel by combining the following (also see UNIT 6.1):
1 ml 30/0.8 acrylamide/bis acrylamide
0.36 ml 2 M Tris·Cl, pH 6.8
75 μl 10% (w/v) SDS
6 ml H2 O
8 μl TEMED
60 μl ammonium persulfate
Pour stacking gel as described in UNIT 6.1.
5. Mix 1 part MMP solution (partially or fully purified MMP, culture medium, concentrated culture medium, or other preparation containing MMP) with 4 parts of 5×
sample buffer (final concentration, 1% w/v SDS). Incubate at room temperature for
10 min, then load 20 to 30 μl into each well of the 15-ml gel prepared in steps 3 and 4.
Alternatively, load 20 to 30 μl per well of an Invitrogen minigel.
6. Run gel at 200 V for 35 to 45 min or until dye front reaches bottom of gel using
electrophoretic running buffer, pH 8.3.
7. Remove gel from electrophoretic apparatus and place in an appropriately sized
container. Wash four times, 20 min each, successively, in washing buffers 1, 2, 3,
and 4 at room temperature. Shake gently throughout.
8. Replace the last wash buffer with fresh washing buffer 4 and incubate 1 to 24 hr at
37◦ C.
A few hours of incubation is usually sufficient to reveal MMP-2 and MMP-9 by gelatin
zymography. Overnight incubation is required to visualize MMP-1, MMP-3, MMP-13,
MMP-7, and MMP-10 by casein zymography.
9. Stain gel with Coomassie blue stain for 30 min and destain with gel destaining
solution for several hours until bands are clear.
Typical results are shown in Figure 10.8.3.
REVERSE ZYMOGRAPHY
Reverse zymography is specifically designed to identify electrophoretic bands which
display MMP-inhibitory activity. The method is based on incorporation of both MMP
activity and gelatin into the running gel. During the ensuing incubation, the SDS-activated
MMP-2 (gelatinase A) cleaves the substrate everywhere in the gel except in and immediately around bands with inhibitory activity such as TIMPs. This method yields well
BASIC
PROTOCOL 3
Extracellular
Matrix
10.8.11
Supplement 40
+/+
+/–
–/–
TIMP-1
TIMP-?
TIMP-2
Mutant TIMP-2
Figure 10.8.4 Reverse zymography. Inhibition of MMP-2 by TIMPs. Skin fibroblast culture
medium obtained from wild-type, hemizygous, or homozygous TIMP-2-deficient mice was resolved by SDS-PAGE in a gel also containing MMP-2 and gelatin. During incubation, MMP-2
cleaves gelatin unless inhibited by electrophoretic bands of TIMPs. The TIMP-2-deficient cells
still express TIMP-1 and unidentified component below TIMP-1, possibly TIMP-3 and a weakly
inhibitory truncated mutant of TIMP-2. Modified from Caterina et al. (2000).
resolved bands of TIMP-1, TIMP-2, TIMP-3, and TIMP-4, as well as mutant forms
of these inhibitors (Fig. 10.8.4). The following protocol is developed by the StetlerStevenson laboratory and used in the authors’ laboratory as well. Quantities are for a
15-ml gel, but can be scaled down as necessary.
Materials
8.7 mg/ml gelatin solution (see recipe)
MMP-2 (Gelatinase A)
2.5% (w/v) Triton X-100
Incubation solution (see recipe)
Additional reagents and equipment for “forward” zymography (see Basic
Protocol 2)
1. Prepare separating gel (17%), copolymerizing gel with gelatin (2.5 mg/ml) and
purified gelatinase A (MMP-2), by mixing the following components (also see
UNIT 6.1):
Matrix
Metalloproteinases
4.2 ml 8.7 mg/ml gelatin solution
0.16 μg/ml (final concentration) gelatinase A (MMP-2)
8.25 ml 30/0.8 acrylamide/bisacrylamide
2.1 ml H2 O
0.29 ml 10% (w/v) SDS
7.3 μl TEMED
73 μl 10% (w/v) ammonium persulfate
10.8.12
Supplement 40
Pour separating gel as described in UNIT 6.1.
Purified MMP-2 may be replaced with culture medium of cells that secrete this enzyme.
The appropriate amount should be determined by trial and error.
2. Prepare 5% stacking gel by combining the following (also see UNIT 6.1):
1.66 ml 30/0.8 acrylamide/bis acrylamide
1.55 ml 2 M Tris·Cl, pH 6.8
125 μl 10% (w/v) SDS
8.2 ml H2 O
10 μl TEMED
200 μl ammonium persulfate
Pour stacking gel as described in UNIT 6.1.
3. Mix samples with 5× sample buffer for reverse zymography. Incubate at room
temperature for ≥10 min, then load 20 to 30 μl into each well of the gel.
4. Run gel at 150 V until buffer front reaches bottom of gel.
5. Remove gel and wash in three changes of 2.5% Triton X-100, each for 2 hr with
gentle shaking.
6. Incubate overnight at 37◦ C in incubation solution.
7. Stain gel with Coomassie blue stain for 20 min and destain in gel destaining solution
for several hours until background is clear.
α 2-MACROGLOBULIN (α 2M) CAPTURE
α2M capture is particularly valuable because it permits assessment of the proteolytic
competence and activity of single bands of MMPs in a mixture of many partially or
fully processed forms. The method was originally devised (Birkedal-Hansen et al., 1976)
for separation of complexes from unreacted forms by molecular sieve chromatography
(Fig. 10.8.5), but it is even more valuable when combined with electrophoretic analysis. The protocol is based on the observation that α2M forms complexes only with
catalytically competent forms of MMPs. Unactivated MMP precursors or forms devoid
of catalytic activity are not captured. The ensuing separation by SDS-PAGE permits
easy identification of bands which have been captured and moved to the top of the gel
because of the large molecular mass of the α2M (Fig. 10.8.5). Bands that escape capture
continue to migrate at their usual position. Complexes formed with α2M are covalent
and therefore not easily dissociated. The ability of α2M to discriminate between latent
and overtly active forms of the enzyme is a result of the α2M inhibition mechanism.
α2M is inert until the attacking proteinase cleaves a peptide bond in the bait region.
This cleavage results in rapid conformational change and liberates a thiol ester which
covalently bonds to and immobilizes the attacking proteinase.
BASIC
PROTOCOL 4
Materials
MMP solution to be tested
2 to 3 mg/ml purified α2M in 50 mM Tris·Cl standard buffer (see recipe for buffer)
100 μg/ml TPCK-treated trypsin (e.g., Sigma) in 50 mM Tris·Cl standard buffer
(see recipe), pH 7.4
1.0 mg/ml soybean trypsin inhibitor in 50 mM Tris·Cl standard buffer (see recipe),
pH 7.4
Extracellular
Matrix
10.8.13
Supplement 40
Antibodies to MMPs of interest
Nitrocellulose paper
Additional reagents and equipment for SDS-PAGE according to Laemmli (UNIT 6.1)
and for immunoblotting (UNIT 6.2)
1. Mix one half of the test solution with a sufficient volume of 1.5 mg/ml α2M to achieve
a ≥10× molar ratio of inhibitor to MMP. Incubate 15 min at room temperature.
2. To compare “activated” and “unactivated” samples, preincubate the other half of the
test sample with 10 μg/ml trypsin (added from 100 μg/ml stock) for 10 min at room
MMP-1
proMMP-1
α 2-macroglobulin
in
+
E2
00 α 2M
Q
E2
00
Q
+
H
19
α
4S 2M
H
19
4S
+
α2
M
ps
Tr
y
in
ps
Tr
y
A
M
AP
AP
M
α
2M
A
+
α2
M
Vo
α 2M-MMP-1
Complex
52K
45K
42K
MMP-1
1
Matrix
Metalloproteinases
2
3
4
5
6
7
8
9
10
Figure 10.8.5 α2-macroglobulin (α2M) capture. The capture technique is based on the property
that proteolytic cleavage of the α2M bait region results in conformational and eventually covalent
capture of the attacking proteinase. Because of the large disparity in molecular weight, captured
and free froms of the proteinase may be separated either by molecular sieve chromatography
(upper panel; Birkedal-Hansen et al., 1976) or by SDS gel electrophoresis (lower panel; Windsor
et al., 1994). Covalently bound proteinase is not released and is readily identified by appropriate
antibody staining. Latent or inactive proteinases are not captured. The method therefore discriminates between enzyme forms with and without catalytic activity at the moment of testing. The
panel shows wild-type and mutant forms of human MMP-1. Samples in lanes 3, 4, and 7 to
10 are pretreated with p-aminophenylmercury acetate (APMA). Samples in lanes 5, and 6 were
preactivated by trypsin. Modified from Windsor et al. (1994).
10.8.14
Supplement 40
temperature, then add 100 μg/ml soybean trypsin inhibitor (added from 1 mg/ml
stock). Incubate separately with α2M as described in step 1.
Commercial sources of α2M are available but should always be checked for activity by
titration with trypsin using a suitable substrate (Sottrup-Jensen and Birkedal-Hansen,
1989). Alternatively, the inhibitor may be prepared by standard techniques as described
by Sottrup-Jensen and Birkedal-Hansen (1989) and Sottrup-Jensen et al. (1983). Activation with trypsin prior to addition of α2M often yields more complete capture than
with organomercurials—e.g., NH2 PheHgAc (APMA)—which seem to gradually inactivate
α2M. Samples preincubated with organomercurials, however, still show partial capture.
3. Mix with 5× electrophoretic sample buffer (final concentration, 1% w/v SDS, 2.5%
v/v 2-ME) without heating, then resolve by by SDS-PAGE using a 10% gel according
to Laemmli (Laemmli, 1970; UNIT 6.1).
4. Transfer to nitrocellulose paper and stain with appropriate MMP antibody using
conventional immunoblotting techniques (UNIT 6.2).
TIMP CAPTURE
Complexes formed with TIMPs are not covalent, although several, but not all, withstand
exposure to low concentrations of SDS, as originally observed by DeClerck et al. (1991),
who first pioneered this technique. This method detects many but not all activated MMPs
that bind TIMPs, including MMP-1 (collagenase-1), MMP-3 (stromelysin-1), MMP7 (matrilysin), MMP-10 (stromelysin-2), and MMP-13. Detection is most conveniently
done by immunoblotting using specific antibodies to the two complex components (MMP
and TIMP; Fig. 10.8.6). The method described below is the authors’ adaptation of the
method of DeClerck (DeClerck et al., 1991). It is based on capture with TIMP-1, but
TIMP-2 capture works just as well.
ALTERNATE
PROTOCOL
Materials
0.1 to 1.0 mg/ml TIMP-1 (Oncogene Research Products, Chemicon International;
also see Bodden et al., 1994) in 50 mM Tris·Cl standard buffer (see recipe),
pH 7.4
10.0 mM NH2 PheHgAc (APMA; Sigma) in Tris·Cl standard buffer (see recipe),
pH 7.4
5× electrophoretic sample buffer (see recipe, but use only 0.5% w/v SDS)
Antibodies to MMPs and TIMP-1 of interest (Calbiochem, Chemicon International)
Additional reagents and equipment for SDS-PAGE (UNIT 6.1) and immunoblotting
(UNIT 6.2)
1. Incubate control and activated samples with 40 to 100 μg/ml TIMP-1 (added from
0.1 to 1.0 mg/ml stock) with and without 1.0 mM NH2 PheHgAc (added from 10.0
mM stock) for 90 min at 37◦ C.
Molecules which are activated by NH2 PheHgAc are captured almost instantly by TIMP-1.
TIMP-1 may be prepared from cultures of fibroblasts or similar cell lines that express fairly
high levels of TIMP-1 activity (Bodden et al., 1994). Concentrations of this compound in
the range of 0.1 to 1.0 mg/liter may be recovered from the culture medium. The purification
scheme is somewhat cumbersome but greatly facilitated by use of antibody-based affinity
chromatography techniques.
2. Mix with 5× electrophoretic sample buffer containing 0.5% SDS. Resolve by SDSPAGE using a 10% gel on ice at 100 V (UNIT 6.1).
Note that the SDS concentration of the sample buffer is reduced to 0.1% (final concentration) in order to avoid dissociation of these entirely noncovalent complexes. This change
is crucial to the success of the technique.
Extracellular
Matrix
10.8.15
Supplement 40
-MMP-1
-TIMP-1
B
A
APMA
MMP-1
TIMP-1
-MMP-10
–
+
–
+
+
–
+
+
+
–
–
+
C
–
+
–
+
+
–
+
+
+
–
–
+
APMA –
MMP-10 +
TIMP-1 –
-TIMP-1
D
+
+
–
+
+
+
–
–
+
Figure 10.8.6 TIMP capture. (A, B) are identical panels stained with antibodies to either human
MMP-1 or TIMP-1. Capture of activated human MMP-1 gives rise to a new band in the 70-kDa
range containing both MMP-1 and TIMP-1 (arrow). (C) proMMP-10 and activated MMP-10 stained
with antibody to human MMP-10. (D) Addition of TIMP-1 to activated MMP-10 results in capture
of the enzyme now migrating in a complex with TIMP-1 in the Mr 70-kDa range. Modified from
Windsor et al. (1993).
3. Transfer to nitrocellulose and stain adjacent lanes with antibodies to TIMP-1 and to
MMP using standard immunoblotting techniques (UNIT 6.2).
BASIC
PROTOCOL 5
FLUORESCENT LABELING OF CRYPTIC CYS-RESIDUE IN MMPs
Most MMP (and ADAM) precursors contain a cryptic thiol group derived from a single,
unpaired cysteine residue in the propeptide. This group is coordinately bonded directly
to the active site Zn (“cysteine switch”) and in this manner plays a significant role
in maintaining the catalytic latency of the proteinase precursors. The protocol below
permits unmasking and detection of this cryptic thiol group (Fig. 10.8.7). The “switch”
opens upon addition of SDS, which allows reaction of the liberated thiol group with a
fluorescent maleimide compound (Yamamoto et al., 1977; Lyons et al., 1991).
Materials
MMP-containing samples
20 μM fluorescent maleimide N-(7-(di-methylamino-4-methyl-3-coumarinyl)
maleimide (DACM) in Tris·Cl standard buffer (see recipe for buffer; prepare
from 1 mM DACM stock in DMSO or ethanol)
2-mercaptoethanol stock in electrophoretic sample buffer (see recipe for buffer):
concentration appropriate to obtain 5% final concentration in reaction mixture
Fluorescent lamp
Photographic equipment
Additional reagents and equipment for SDS-PAGE (UNIT 6.1)
1. Expose companion samples of 50 to 200 μg/ml MMP for 1 hr at room temperature
to 20 μM DACM (final concentration) either in the presence or absence of 1% (w/v)
SDS.
2. Stop reaction by adding 2-mercaptoethanol (as stock solution of appropriate concentration in electrophoretic sample buffer) to a final concentration of 5% (v/v).
Matrix
Metalloproteinases
3. Resolve proteins by SDS-PAGE (UNIT 6.1).
10.8.16
Supplement 40
+S
D
_
_
+S
D
S
S
S
D
+S
_
52K
54K
MMP-1
MMP-3
MMP-10
Figure 10.8.7 Fluorescent labeling of propeptide cryptic thiol residue by fluorescent maleimide.
The cysteine switch is “closed” in the nascent proenzyme and therefore not reactive with a fluorescent maleimide compound (DACM). Exposure to SDS “opens” the switch and renders the cryptic
thiol group reactive with the maleimide resulting in covalent modification of the proenzyme and
generation of a readily detectable fluorescent band. Left panel: MMP-1. Right panel, MMP-3 and
MMP-10. Lower edge of each panel shows Coomassie blue staining of the same bands. Modified
from Windsor et al. (1993).
4. Photograph under long-wavelength UV illumination.
REAGENTS AND SOLUTIONS
Use deionized or distilled water in all recipes and protocol steps. For common stock solutions,
see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.
Coomassie blue stain
0.5% (w/v) Coomassie blue R-250
30% (v/v) methanol
10% (v/v) acetic acid
Store up to 6 months at room temperature
Electrophoretic running buffer, pH 8.3
0.025 M Tris base
0.192 M glycine
0.1% (w/v) SDS
Store up to 1 year at room temperature
Electrophoretic sample buffer, 5×
0.2 M Tris·Cl, pH 6.8 (APPENDIX 2A)
5% (w/v) SDS
20% (w/v) glycerol
continued
Extracellular
Matrix
10.8.17
Supplement 40
0.1% (w/v) bromphenol blue
This is the sample buffer used in Basic Protocol 2.
Gelatin solution, 8.7 mg/ml
Add gelatin (bovine skin, Sigma-Aldrich type B6-6269) to 1 M Tris·Cl, pH 8.8 at
8.7 mg/ml. Dissolve by heating to 57◦ C, then filter through Whatman no. 1 filter
paper.
Gel destaining solution
30% (v/v) methanol
10% (v/v) acetic acid
60% (v/v) H2 O
Gel washing buffers 1 to 4
Buffer 1:
2.5% (v/v) Triton X-100
3 mM NaN3
Buffer 2:
2.5% (v/v) Triton X-100
50 mM Tris·Cl, pH 7.5 (APPENDIX 2A)
3 mM NaN3
Buffer 3:
2.5% (v/v) Triton X-100
3 mM NaN3
5 mM CaCl2
1 μM ZnCl2
Buffer 4:
3 mM NaN3
5 mM CaCl2
1 μM ZnCl2
Buffers may be stored up to 1 year at room temperature.
Incubation solution
0.2 M NaCl
5 mM CaCl2
0.02% (w/v) Brij-35
Store up to 1 year at 4◦ C
Neutralizing buffer (0.2 M naPi )
Matrix
Metalloproteinases
Prepare the following stock solutions:
Solution A: 2.78 g NaH2 PO4 in 100 ml H2 O
Solution B: 5.365 g Na2 HPO4 ·7H2 O in 100 ml H2 O
Prepare working solutions as follows:
15.2 ml Solution A
64.8 ml Solution B
16.6 ml 5 M NaCl
Add 80 ml 0.1 N NaOH
10.8.18
Supplement 40
Tris·Cl standard buffer, pH 7.4
0.2 M NaCl
5 mM CaCl2
COMMENTARY
Background Information
Dissolution of collagen type I
Substrate. Although collapsing the gel by
air drying is advantageous for most purposes,
and the resulting collagen film is more similar
to the density of collagen in interstitial connective tissues (Fig. 10.8.1), it is possible to seed
the cells on top of (or inside) fully hydrated
gels and to monitor the process as the cells
dissolve their way through the collagen gel.
Electron microscopy confirms that hydrated
gels are very loose, with the individual fibrils
spaced far apart. The collagen content is quite
low compared to the liquid phase and accounts
for only 0.03% of the mass and for a similarly
small volume fraction of the gel.
While use of reconstituted type I collagen
fibrils as a substrate offers particular advantages because of its resistance to general proteolysis, it is possible to replace this substrate
with other extracellular matrix components.
Type II collagen does not form fibrils as readily as does type I but might prove useful after additional refinement of the system. Type
III collagen appears to gel adequately for this
purpose and may also be used as a substrate.
Films and gels of type IV collagen may also
be used, as may Matrigel (predominantly composed of laminin), fibrin, and fibronectin. An
important variation using fluorescently labeled
fibronectin was devised by Chen and coworkers (Chen et al., 1984; Chen and Chen, 1987).
Serum. Serum contains a number of factors
expected to either promote or inhibit the proteolytic dissolution of the extracellular matrix
including collagen fibrils. The high concentration of α2M (2 to 3 mg/ml or 3 to 4 × 10−6
M), which effectively blocks most MMPs in
test tube experiments, however, does not inhibit cell-mediated dissolution of the collagen
fibril film. Serum also contains plasminogen
at a concentration of ∼200 μg/ml (2 × 10−6
M). Addition of even low concentrations of
plasminogen (4 μg/ml; 4 × 10−8 M) to serumfree cultures greatly accelerates the rate of dissolution of the collagen fibril film by human
foreskin keratinocytes (or other cells) which
express urokinase-type plasminogen activator
(u-PA) or tissue-type plasminogen activator
(t-PA). The mechanism is not quite well understood but may involve a role for plasminogen in the extracellular activation of certain
proMMP precursors as an essential step in the
dissolution of the substrate.
Cytokines, transcriptional activation. Addition of cytokines, growth factors, and agents
such as TPA, which upregulate or induce expression of MMPs, generally accelerates dissolution of the fibril coating dramatically, but
since these reagents upregulate a wide range
of MMPs, it is not yet possible to determine
whether a single MMP or group of MMPs is
responsible for this effect.
Inhibition. That dissolution of the
collagen fibril coating is mediated by
metalloproteinase-dependent mechanisms is
readily made evident by synthetic inhibitors.
Inclusion of the Zn-chelating agent 1,10phenanthroline completely blocks dissolution,
as do synthetic MMP inhibitors such as BB94,
BB2516 (British Biotech), and Galardin. A
number of synthetic inhibitors currently exist;
some of these may be obtained by directly
contacting the pharmaceutical companies in
question (British Biotech, Roche Diagnostics,
Celltech). Serine proteinase inhibitors such as
α1-antitrypsin (α1AT) and soybean trypsin
inhibitor, as well as cysteine proteinase
inhibitors such as E-64, have no effect on the
rate of dissolution. These findings suggest that
the process(es) that result in dissolution of the
collagen fibrils are absolutely dependent on
MMP activity.
Zymography
Gelatin zymography is a fairly straightforward yet very highly sensitive technique as
long as heating and reduction are avoided during sample preparation. The method yields
discrete, well-resolved, and distinct unstained
bands on a blue background, which are clearly
visible and easy to photograph and document
with transillumination (Fig. 10.8.3). The activity may be quantified by comparison with
standard curves of specific purified MMPs
(Kleiner and Stetler-Stevenson, 1994), but the
rate of lysis varies considerably from MMP to
MMP, and the technique is primarily intended
Extracellular
Matrix
10.8.19
Supplement 40
to provide qualitative information. A variation described by Lyons et al. (1991) permits
monitoring of real-time progress of the reaction under UV light by use of gelatin labeled
by a fluorophore. Although gelatin zymography is highly sensitive, capable of detecting
low picogram quantities of MMPs, the assay
does not reflect the activity of these proteases
present in the sample analyzed. This is because
the addition of SDS to the sample prior to electrophoretic separation results in dissociation of
many enzyme inhibitor complexes. Therefore,
zymography represents an excellent technique
for identification of MMP species present in
a given sample, but overinterpretation of the
results—e.g., assessment of specific activity—
is a common pitfall. Casein zymograms develop more slowly, almost invariably require
overnight incubation, and tend to produce less
sharp bands. (Latent) proenzyme forms also
show up because of the “switch”-opening effect of SDS, but these forms do not necessarily
acquire full catalytic activity. “Activation” by
organomercurials (0.5 to 1.0 aminophenylmercuric acetate in 50 mM Tris·Cl buffer, pH 7.5,
for 20 min to 20 hr) before sample preparation
often results in higher levels of proteolytic activity but also shifts the Mr of the individual
bands because of autolytic cleavage and removal of the propeptide.
α2M capture
Capture techniques permit direct assessment of the ability of various forms of MMPs
to bind to natural inhibitors in a manner that
resists dissolution by exposure to low concentrations of SDS. Zymographic techniques
are not capable of discriminating between latent and catalytically active forms of the enzymes. That, however, can readily be achieved
by α2M capture. TIMP capture (see Alternate
Protocol) on the other hand does not depend on
proteolytic activity and merely requires a correctly folded, but not necessarily catalytically
competent, active site (Windsor et al., 1994).
Matrix
Metalloproteinases
TIMP capture
The method is particularly useful for analysis of the binding capacity of mutants in TIMPs
and in MMPs (Windsor et al., 1994; Caterina
et al., 1997). It is important to recognize that
TIMP binding is not necessarily synonymous
with catalytic competence. Mutants of MMP
which are correctly folded but devoid of catalytic activity, such as the E200Q mutant of
MMP-1 in which the active site glutamate is
replaced with glutamine (and therefore catalytically inactive), still form complexes with
TIMP-1 fully, as well as the native enzyme
(Windsor et al., 1994). TIMP-1 captures both
truncated and full-length forms, as long as the
“switch” is open (by APMA).
Fluorescent labeling of cryptic Cys residue
The nascent closure of the “cysteine
switch” by bonding of the single unpaired
propeptide Cys residue to the active-site Zn2+
converts a catalytic Zn-binding site to a structural Zn-binding site. In order to monitor the
(re)opening of the switch as a preamble to zymogen activation, the authors of this unit reasoned that covalent linkage of a fluorophore to
the free thiol group might render this process
easily visible and potentially quantifiable. That
is indeed the case. The method shows, for instance, that the phenomenon of “switch opening” can be readily visualized in the absence
of propeptide cleavage by exposure either to
SDS or to EDTA.
Critical Parameters and
Troubleshooting
Even for the experienced operator, collagen is not an easy protein to work with. Its
preparation and use require meticulous and
stringent adherence to the rules and conditions
that “work,” often with very little leeway for
shortcuts and modifications. The most important checkpoint comes after the initial gelling.
Unless there is clear and unequivocal evidence
of gelling after 2 hr, efforts should be made to
identify and correct the problem. Since there
is no simple way to measure collagen concentration, the authors have utilized initial dry
powder weight from materials stored in refrigerated dessicator jars as a guide. The concentrations mentioned in this unit refer to powder weight under these conditions. It is absolutely necessary that the solution from which
the collagen is lyophilized be completely saltfree following extensive dialysis against dilute
acetic acid. This problem may be avoided by
purchase of commercial preparations of rat or
bovine type I collagen, but it is necessary to test
the gelling properties of the particular brand in
question at the desired concentration and under the desired conditions. After 2 hr of gelation, the gel should be reasonably firm, i.e., it
should not disintegrate upon gentle flicking of
the plate. If the gel disintegrates during this
test, the problem must first be solved before
proceeding.
The homogeneity of the gel is also very important. This is best checked following the first
10.8.20
Supplement 40
air drying and washing step by staining a newly
prepared film with Coomassie blue. This will
instantly reveal whether the gel is uniform and
homogenous and if it contains particulate matter (which can be removed by centrifugation)
or air bubbles. Both must be avoided, and the
technique must be improved until each gel
is completely uniform and homogenous after
staining. It is also important to ascertain after
the first washing of the first-time dried gel that
salt-crystal deposits (formed during the initial
drying phase) have been completely removed
by washing. This is most easily checked using
the phase-contrast microscope. The gel should
look granular but uniform; any trace of crystal
patterns is a certain indication of inadequate
washing.
Zymography
Gelatin zymography presents few, if any,
technical challenges, hence the popularity and
universal application of this technique. Because of the longer incubation time required
and the lesser sensitivity, casein zymograms
often give less distinct and more diffuse bands.
Although it has not been widely explored, it is
highly likely that a large number of other substrates could be substituted for either gelatin
or casein. Reverse zymography, on the other
hand, is technically challenging and requires
great care and skill as well as considerable
practice and experience. The latter method
is, however, a uniquely powerful technique to
identify discrete MMP-inhibitory bands.
Inhibitor capture
While commercial preparations of α2M are
available, the method is critically dependent on
the native configuration of the inhibitor. Consequently, the authors rely only on freshly isolated inhibitor. Occasionally, methods which
are employed to activate MMPs, such as exposure to organomercurials, adversely affect
the inhibitor and render the capture reaction
partial rather than complete. In some cases
trypsin activation (stopped by soybean trypsin
inhibitor) is preferable, but many mutants are
highly sensitive to trypsin and rapidly degrade
during activation attempts.
The method is fairly straightforward, although care must be taken to exclude any
chemicals from the solutions that interfere
with the Cys-maleimide reaction (e.g., heavy
metals, N-ethylmaleimide, or iodoacetate).
Photographic documentation can be tricky, but
usually works well when using reflected UV
light.
Anticipated Results
Use of 1- to 2-μm films results in complete dissolution within 1 to 4 days. Initially
the cells penetrate the collagen fibril coating
in discrete spots, which eventually coalesce
to form contiguous zones devoid of collagen
fibrils (Fig. 10.8.2). Dissolution of the fibril
coating is strictly limited to the area immediately beneath the cell layer and does not extend beyond the boundaries of the cell colony.
A similar pattern is observed in the presence
of serum or purified plasminogen.
Zymography
When performed correctly, the reverse
zymography–stained gel shows discrete, wellresolved bands of TIMPs on a virtually unstained background, indicating that all of the
gelatin has been degraded except in and around
the TIMP bands (Fig. 10.8.4). While this
method yields important information when
used in qualitative or semiquantitative fashion,
the read-out may be quantified as described by
Kleiner and colleagues (Oliver et al., 1997).
As with direct zymography, reverse zymography is a highly sensitive technique that can
detect as little as 50 to 100 pg of TIMPs in
a given sample (Oliver et al., 1997). However, as with direct zymography, careful interpretation of results is essential. Again, use
of SDS-containing sample buffers and electrophoretic separation of the sample results in
dissociation of some protease-inhibitor complexes. Thus, the levels of TIMPs present may
not accurately reflect the actual free TIMP levels present in the samples analyzed. Alternatively, as described for the TIMP capture assays, not all TIMP-MMP complexes may be
dissociated by SDS, and TIMP-binding to an
MMP active site does not necessarily reflect
proteolytic competency of the enzyme.
α2M capture
Incubation of native α2M with activated
proteinases that cleave the bait region result
in full or partial capture of the attacking proteinase. Complete capture requires a significant molar excess of inhibitor (with the amount
varying from proteinase to proteinase), which
may be determined by titration in preliminary
experiments. Because of the size difference,
captured and uncaptured bands are readily resolved and identified on western blots by staining with anti-MMP antibodies. Latent or catalytically inactive forms are not captured and
remain at their usual migration position in the
gel.
Extracellular
Matrix
10.8.21
Supplement 40
TIMP capture
Remarkably, most TIMP-MMP complexes
survive dilute SDS solutions at room temperature and permit electrophoretic separation
of free and complexed forms. The Mr difference (20 to 30 kDa) is sufficient to fully resolve the bands. As with α2M capture, latent
forms of MMPs (“switch closed”) are not captured, and this method is therefore valuable
in distinguishing “switch-open” and “switchclosed” forms before proteolytic excision of
the propeptide during activation. The active
site, however, does not have to possess catalytic activity, and inactive mutants (if correctly folded) readily form complexes with
TIMPs.
Removal of Zn2+ with EDTA, as expected,
also unmasks the cryptic thiol group. Fully
converted (“activated”) forms of the enzyme
which have lost the entire propeptide no longer
react. Note, however, that the free thiol group
is only a few residues upstream of the ultimate
proteolytic processing site. Partially processed
forms of the proenzymes therefore may still
react with DACM.
Time Considerations
Analysis of the degradation of collagen gels
takes ∼2 days to prepare the gels and 1 to
4 days for the assay itself. It requires ∼2 weeks
to prepare rat tail tendon collagen type I and
3 to 4 days to label the collagen with fluorophore.
Direct zymography takes 2 days to complete, while reverse zymography takes 2 days.
α2M and TIMP capture take 1 to 2 days
depending on the duration of antibody incubation in immunoblotting.
Fluorescent labeling of the cryptic Cys
residue can be completed in a single day.
Literature Cited
Aimes, R.T. and Quigley, J.P. 1995. Matrix
metalloproteinase-2 is an interstitial collagenase. Inhibitor-free enzyme catalyzes the cleavage of collagen fibrils and soluble native type I
collagen generating the specific 3/4-length and
1/4-length fragments. J. Biol. Chem. 270:58725876.
Birkedal-Hansen, H. 1987. Catabolism and
turnover of collagens: Collagenases. Methods
Enzymol. 144:140-171.
Birkedal-Hansen, H. and Danø, K. 1981. A sensitive collagenase assay using [3 H]collagen labeled by reaction with pyridoxal phosphate and
[3 H]borohydride. Anal. Biochem. 115:18-26.
Matrix
Metalloproteinases
Birkedal-Hansen, H. and Taylor, R.E. 1982.
Detergent-activation of latent collagenase
and resolution of its component molecules.
Biochem. Biophys. Res. Commun. 107:11731178.
Birkedal-Hansen, H., Cobb, C.M., Taylor, R.E., and
Fullmer, H.M. 1976. Synthesis and release of
procollagenase by cultured fibroblasts. J. Biol.
Chem. 251:3162-3168.
Birkedal-Hansen, H., Birkedal-Hansen, B.,
Windsor, L.J., Lin, H.Y., Taylor, R.E., and
Moore, W.G.I. 1989. Use of inhibitory (anticatalytic) antibodies to study extracellular
proteolysis. Immunol. Invest. 18:211-224.
Birkedal-Hansen, H., Moore, W.G.I., Bodden,
M.K., Windsor, L.J., Birkedal-Hansen, B.,
DeCarlo, A., and Engler, J.A. 1993. Matrix metalloproteinases: A review. Crit. Rev. Oral Biol.
Med. 4:197-250.
Bodden, M.K., Harber, G.J., Birkedal-Hansen, B.,
Windsor, L.J., Caterina, N.C.M., Engler, J.A.,
and Birkedal-Hansen, H. 1994. Functional domains of human TIMP-1 (tissue inhibitor of
metalloproteinases). J. Biol. Chem. 269:1894318952.
Caterina, N.C.M., Windsor, L.J., Yermovsky, A.E.,
Bodden, M.K., Taylor, K.B., Birkedal-Hansen,
H., and Engler, J.A. 1997. Replacement of conserved cysteines in human tissue inhibitor of
metalloproteinases-1. J. Biol. Chem. 272:3214132149.
Caterina, J.J., Yamada, S., Caterina, N.C.M.,
Longenecker, G., Holmback, K., Shi, J.,
Yermovsky, A.E., Engler, J.A., and BirkedalHansen, H. 2000. Inactivating mutation of the
mouse tissue inhibitor of metalloproteinnases-2
(TIMP-2) gene alters proMMP-2 activation. J.
Biol. Chem. 275:26416-26422.
Chen, J.M. and Chen, W.T. 1987. Fibronectindegrading proteases from the membranes of
transformed cells. Cell 48:193-203.
Chen, W.T., Olden, K., Bernard, B.A., and
Chu, F.-F. 1984. Expression of transformationassociated protease(s) that degrade fibronectin
at cell contact sites. J. Cell Biol. 98:1546-1555.
DeClerck, Y.A., Yean, T.D., Lu, H.S., Ting, J.,
and Langley, K.E. 1991. Inhibition of autoproteolytic activation of interstitial procollagenase by recombinant metalloproteinase inhibitor
MI/TIMP-2. J. Biol. Chem. 266:3893-3899.
Deutsch, D.G. and Mertz, E.T. 1970. Plasminogen: Purification from human plasma by affinity
chromatography. Science 170:1095-1096.
Ghersi, G., Goldstein, L.A., Wang, J.-Y., Yeh, Y.,
Hakkinen, L., Larjava, H., and Chen, W.-T.
2002. Regulation of fibroblast migration on collagenous matrix by novel cell surface protease
complex. J. Biol. Chem. 277:29231-29241.
Havemose-Poulsen, A.P.H., Stolze, K., and
Birkedal-Hansen, H. 1998. Dissolution of type
I collagen fibrils by gingival fibroblasts isolated
from patients of various periodontitis categories.
J. Periodontal Res. 33:280-291.
Heussen, C. and Dowdle, E.B. 1980. Electrophoretic analysis of plasminogen activators in
polyacrylamide gels containing sodium dodecyl
10.8.22
Supplement 40
sulfate and copolymerized substrates. Anal.
Biochem. 102:196-202.
Kleiner, D.E. and Stetler-Stevenson, W.G.
1994. Quantitative zymography: Detection
of picogram quantities of gelatinases. Anal.
Biochem. 218:325-329.
Knäuper, V., Lopez-Otin, C., Smith, B., Knight, G.,
and Murphy, G. 1996. Biochemical characterization of human collagenase-3. J. Biol. Chem.
271:1544-1550.
Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685.
Lin, H.Y., Wells, B.R., Taylor, R.E., and BirkedalHansen, H. 1987. Degradation of type I collagen by rat mucosal keratinocytes. J. Biol. Chem.
262:6823-6831.
Lyons, J.G., Birkedal-Hansen, B., Moore, W.G.I.,
O’Grady, R.L., and Birkedal-Hansen, H. 1991.
Characteristics of a 95-kDa matrix metalloproteinase produced by mammary carcinoma cells.
Biochemistry 30:1450-1456.
Ohuchi, E., Imai, K., Fuji, Y., Sato, H., Seiki, M.,
and Okada, Y. 1997. Membrane type 1 matrix
metalloproteinase digests interstitial collagens
and other extracellular matrix macromolecules.
J. Biol. Chem. 272:2446-2451.
Oliver, G.W., Leferson, J.D., Stetler-Stevenson,
W.G. and Kleiner, D.E. 1997. Quantitative
reverse zymography: Analysis of picogram
amounts of metallopooteinase inhibitors using
gelatinase A and B reverse zymograms. Anal.
Biochem. 244:161-166.
Sottrup-Jensen, L. and Birkedal-Hansen, H. 1989.
Human fibroblast collagenase-α-macroglobulin
interactions. J. Biol. Chem. 264:393-401.
Sottrup-Jensen, L., Stepanik, T.M., Wierzbicki,
D.M., Jones, C.M., Lonblad, P.B., Kristensen,
T., Mortensen, S.B., Petersen, T.E., and
Magnusson, S. 1983. The primary structure of
α-macroglobulin and localization of a factor
XIIIa cross-linking site. Ann. N.Y. Acad. Sci.
421:41-60.
Springman, E.B., Angleton, E.L., Birkedal-Hansen,
H., and Van Wart, H.E. 1990. Multiple modes
of activation of latent human fibroblast collagenase: Evidence for the role of a Cys73 active-site
zinc complex in latency and a “cysteine switch”
mechanism for activation. Proc. Natl. Acad. Sci.
U.S.A. 87:364-368.
Van Wart, H.E. and Birkedal-Hansen, H. 1990.
The cysteine switch: A principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family. Proc. Natl. Acad. Sci.
U.S.A. 87:5578-5582.
Windsor, L.J., Grenett, H., Birkedal-Hansen, B.,
Bodden, M.K., Engler, J.A., and BirkedalHansen, H. 1993. Cell-type-specific regulation
of SL-1 and SL-2 genes. Induction of SL-2, but
not SL-1, in human keratinocytes in response
to cytokines and phorbolesters. J. Biol. Chem.
268:17341-17347.
Windsor, L.J., Bodden, M.K., Birkedal-Hansen, B.,
Engler, J.A., and Birkedal-Hansen, H. 1994.
Mutational analysis of residues in and around
the active site of human fibroblast-type collagenase. J. Biol. Chem. 269:26201-26207.
Yamamoto, K., Sekine, T., and Kanaoka, Y.
1977. Fluorescent thiol reagents. XII. Fluorescent tracer method for protein SH groups using N-(7-dimethylamino-4-methyl coumarinyl)
maleimide. Anal. Biochem. 79:83-94.
Extracellular
Matrix
10.8.23
Supplement 40
Preparation of Extracellular Matrices
Produced by Cultured and Primary
Fibroblasts
UNIT 10.9
Culturing fibroblasts on traditional two-dimensional (2-D) substrates induces an artificial
polarity between lower and upper surfaces of these normally nonpolar cells. Not surprisingly, fibroblast morphology and migration differ once suspended in three-dimensional
(3-D) collagen gels (Friedl and Brocker, 2000). However, the molecular composition of
collagen gels does not mimic the natural fibroblast microenvironment. Fibroblasts secrete
and organize the extracellular matrix (ECM), which provides structural support for their
adhesion, migration, and tissue organization, in addition to regulating cellular functions
such as growth and survival (Buck and Horwitz, 1987; Hay, 1991; Hynes, 1999; Geiger
et al., 2001). Cell-to-matrix interactions are vital for vertebrate development. Disorders
in these processes have been associated with fibrosis, developmental malformations,
cancer, and other diseases.
This unit describes methods for generating tissue culture surfaces coated with a fibroblastderived 3-D ECM produced and deposited by both established and primary fibroblasts.
The matrices closely resemble in vivo mesenchymal matrices and are composed mainly
of fibronectin fibrillar lattices. Utilizing in vivo–like 3-D matrices as substrates allows
the acquisition of information that is physiologically relevant to cell-matrix interactions,
structure, function, and signaling, which differ from data obtained by culturing cells on
conventional 2-D substrates in vitro (Cukierman et al., 2001).
These protocols were initially derived from methods described in UNIT 10.4, which were
modified to obtain fibroblast-derived 3-D matrices and to characterize cellular responses
to them. The basic approach is to allow fibroblasts to produce their own 3-D matrix (see
Basic Protocol). For this purpose, fibroblasts are plated and maintained in culture in a
confluent state. After 5 to 9 days, matrices are denuded of cells, and cellular remnants
are removed. Such extraction results in an intact fibroblast-derived 3-D matrix that is
free of cellular debris and remains attached to the culture surface (see Figure 10.9.1).
The fibroblast-derived 3-D matrices are then washed with PBS and can be stored 2 to
3 weeks at 4◦ C or up to 3 weeks frozen at −80◦ C. Moreover, to analyze the effect of
matrix pliability on cellular behavior, prepared 3-D matrices can be rigidified by chemical
cross-linking (see Support Protocol 1 and UNIT 17.10).
Additionally, to evaluate the quality of the fibroblast-derived 3-D matrices, support
protocols present a variety of procedures for measuring cell responsiveness to the 3-D
matrix microenvironment (see Support Protocols 2 and 3). The rapid cell attachment of
fibroblasts plated within the matrix can be quantified. By plating isolated fibroblasts in
the 3-D matrix, the acquisition of an in vivo–like spindle-shaped morphology can also be
measured. To ascertain whether fibroblasts respond to the 3-D microenvironment when
plated within specific (NIH-3T3) matrices, the phosphorylation level of nonreceptor
focal adhesion kinase (FAK) pY397 can be quantified by immunoblotting (see Support
Protocol 4).
This unit will also describe how to mechanically compress the fibroblast-derived 3-D
matrices to obtain 2-D substrate controls (see Support Protocol 5). Moreover, a support
protocol will illustrate how to solubilize the fibroblast-derived 3-D matrices to produce
a matrix-derived protein mixture for additional 2-D coating controls and for subsequent
Extracellular
Matrix
Contributed by Dorothy A. Beacham, Michael D. Amatangelo, and Edna Cukierman
Copyright 10.9.1
Supplement 33
Figure 10.9.1 Fibroblast-derived 3-D matrices before and after extraction process. (A) Culture at day 5 prior to matrix
extraction. (B) The resulting fibroblast-derived 3-D matrix. Panels C and D are magnified insets from A and B, respectively.
Bars represent 50 µm.
biochemical analysis of the matrices (see Support Protocol 6 and Commentary). Lastly,
there is a protocol for isolating primary fibroblasts from fresh tissue samples to produce
additional types of fibroblast-derived 3-D matrices (see Support Protocol 7).
NOTE: All solutions and equipment coming into contact with living cells must be sterile,
and aseptic techniques should be used accordingly.
NOTE: All cell-culture incubations should be performed in a 37◦ C, 10% CO2 humidified
incubator.
BASIC
PROTOCOL
Preparation of
Fibroblast
Extracellular
Matrices
PREPARATION OF EXTRACELLULAR MATRICES PRODUCED BY
CULTURED OR PRIMARY FIBROBLASTS
NIH-3T3 cells (for primary cell lines see Support Protocol 7) must be routinely cultured
in high-glucose Dulbecco’s modified Eagle medium supplemented with 10% calf serum,
100 U/ml penicillin, and 100 µg/ml streptomycin unless otherwise specified. Never allow
cultured NIH-3T3 cells to become completely confluent while maintaining stock cultures.
When cells reach 80% confluence (about once per week), subculture at a 1:20 dilution.
However, prior to plating for matrix deposition, NIH-3T3 cells should be adapted to
grow in 10% fetal bovine serum rather than calf serum for the cells to adopt an optimal
phenotype for matrix production (see Critical Parameters).
10.9.2
Supplement 33
Depending on the laboratory equipment available, and the anticipated uses of the
fibroblast-derived 3-D matrices, a suitable surface on which the matrices will be produced (e.g., glass-bottom dishes, coverslips, or tissue culture dishes) must be selected as
follows:
(1) Disposable glass bottom dishes (MatTek) can be utilized for real-time fluorescent
experiments or for quality assessment assays (e.g., cell attachment and cell shape) using
an inverted fluorescent microscope (see Support Protocols 1 and 2).
(2) Coverslips can be used for immunofluorescence experiments in which samples are
fixed and mounted on microscope slides (see Support Protocol 3), or for mechanical
compression of the fibroblast-derived 3-D matrices to be used as control 2-D surfaces
(see Support Protocol 5).
(3) Regular tissue culture dishes (e.g., 35-mm diameter) can be used for in vivo observations using an inverted microscope, for matrix solubilization and further characterization,
and/or for biochemical analyses (see Support Protocols 5 and 3, respectively). Tissue culture dishes are also used for real-time cell motility analyses (Cukierman, 2005).
Materials
NIH-3T3 cells (ATCC) or primary fibroblasts (see Support Protocol 7)
Confluent medium with fetal bovine serum (FBS; see recipe)
0.25% (w/v) trypsin/0.03% (w/v) EDTA solution (see recipe)
0.2% (w/v) gelatin solution (see recipe)
Ethanol (absolute)
1% (v/v) glutaraldehyde in PBS (see recipe)
1 M ethanolamine (see recipe)
Matrix medium with ascorbic acid (see recipe)
Extraction buffer (see recipe), 37◦ C
10U/ml DNase I (Roche) in PBS+ (see recipe for PBS+ ), optional
Penicillin/streptomycin (Invitrogen)
Fungizone (amphotericin B; Invitrogen)
37◦ C, 10% CO2 humidified incubator
15-cm dishes plus the specific culture vessels for matrix production
Inverted phase-contrast microscope
6-well tissue culture plates or 35-mm dishes (optional)
22-mm circular high-quality coverslips (Carolina; optional)
Bacterial 6-multiwell petri plates for preparing matrices on coverslips
Parafilm strips
Small, sterile fine-pointed tweezers (e.g., Dumont no. 4), optional
Prepare cell cultures
1. Start with a semi-confluent (80% confluent) culture of NIH-3T3 cells cultured in 10
to 12 ml confluent medium containing fetal bovine serum (see Critical Parameters) or
primary fibroblastic cells on a 15-cm culture plate (see Support Protocol 7); aspirate
and discard the culture medium.
2. Rinse the cell layer briefly with 1.5 ml of 0.25% trypsin/0.03% EDTA (trypsin/EDTA)
per 15-cm dish. Then gently aspirate off the solution.
This rinse will remove traces of serum that contains trypsin inhibitors.
3. Add enough trypsin/EDTA solution to cover the cell layer, quickly aspirate excess
liquid, and observe under an inverted microscope at room temperature until the cells
have detached from the culture dish (1 to 3 min).
Extracellular
Matrix
10.9.3
Supplement 33
4. Collect the cells in 10 ml of confluent medium.
5. Add 2 ml of the suspended cells and 10 to 12 ml of confluent medium to a 15-cm
plate and culture for 2 to 3 days (until semi-confluent, up to ∼80% confluence).
As many as five 15-cm culture dishes may be used.
Prepare surfaces for matrix deposition
Although not strictly required, both gelatin coating (steps 6 and 7) and cross-linking of
gelatin (steps 8 through 11) with glutaraldehyde stabilizes the attachment of the matrices
to the culture dish surface and greatly improves final yield. However, the resulting
matrices may be thinner than those obtained without gelatin pre-coating and crosslinking. Therefore, matrix thickness should be determined for each fibroblastic cell type
and a decision to follow the optional steps should be made for each cell type.
6a. For tissue culture dishes: Add 2 ml of 0.2% gelatin solution to a 35-mm tissue culture
dish surface to be used for fibroblast-derived 3-D matrix deposition and incubate for
1 hr at 37◦ C.
Choose 35-, 60-, or 100-mm dishes to be used in this protocol. For 60- or 100-mm dishes,
scale up the volumes of all reagents added from 2 ml to 4 and 8 ml, respectively.
6b. For coverslips: Presterilize by flaming the coverslips after dipping in anhydrous
ethanol (absolute). Then place coverslip in a tissue culture dish and rinse with PBS.
Incubate coverslips in a 0.2% gelatin solution.
7. Aspirate gelatin and add 2 ml PBS.
When using coverslips, after this rinse, transfer coverslips to individual wells of multiwell
bacterial plates.
When preparing coverslips for 3-D matrix deposition, multiwell bacterial petri dishes are
preferred over tissue culture plastic dishes because the fibroblasts do not adhere well to
bacterial petri plastic. Consequently, there is preferential fibroblastic growth on pretreated
glass coverslips instead of on the surface of the petri plastic, conditions conducive
to enhancing matrix production on the coverslip. Placing coverslips on bacterial petri
plastic also facilitates lifting the coverslip off the dish surface with tweezers during matrix
extraction (see step 19) because any cells growing underneath the glass coverslips are
easily dislodged.
8. Aspirate PBS and add 2 ml of 1% glutaraldehyde (prediluted in PBS) to each dish
or well and incubate 30 min at room temperature.
9. Wash coverslips or culture dishes three times for 5 min each with 2 ml of PBS.
10. Add 2 ml of 1 M ethanolamine to each dish and incubate 30 min at room temperature.
11. Repeat the PBS washes (step 9).
This step is a good place to stop if time does not permit cell seeding. Dishes can be left
for 1 to 7 days at room temperature under sterile conditions.
12. Aspirate PBS from dishes and replace with 2 ml matrix medium. If the medium
appears purple, repeat steps 11 and 12 to remove any trace amounts of ethanolamine.
At this point, the surfaces are ready to be seeded with matrix-producing fibroblasts.
Allow cells to deposit matrix
13. Repeat steps 1 to 3 to harvest cells from a semi-confluent dish.
Preparation of
Fibroblast
Extracellular
Matrices
This protocol was developed for NIH-3T3 cells. Nevertheless, other fibroblast cell lines
can be used. For example, the same protocol can be followed from this point on using
human or other primary fibroblastic cells (see Suppport Protocol 7).
10.9.4
Supplement 33
14. Collect cells from each dish in 10 ml of matrix medium, count cells (UNIT 1.1), and
dilute to a final concentration of 2.5 × 105 cells/ml.
15. Aspirate medium (from step 12) and seed 5 ×105 cells in 2 ml of matrix medium
per 35-mm dish and culture for 24 hr.
Use as many dishes as needed; there should be enough cells for ∼100 35-mm plates from
each 15-cm semiconfluent dish. Remember to scale up volumes from 2 ml to 4 or 8 ml for
60- or 100-mm plates, respectively.
16. After 24 hr, carefully aspirate the medium from cells and replace with fresh matrix
medium containing 50 µg/ml of ascorbic acid.
For some primary cell lines, adding a ten-fold higher concentration of ascorbic acid on
this first addition of ascorbic acid after cell plating can increase matrix thickness. Because
the higher ascorbic acid concentration is detrimental for NIH-3T3 cells, whether or not
to use this higher dose should be independently determined for each fibroblastic cell type.
17. Ascorbic acid degrades over time in culture, so change medium with freshly made
matrix medium every 48 hr for a total of 5 to 9 days after step 16.
At this time, the matrix should be sufficiently thick to achieve three-dimensionality (≥10
µm, see below). At this point it is ready to be extracted (see Fig. 10.9.1 A).
Matrices should be extracted after they reach a thickness of at least 10 µm. The time
required for each fibroblast cell type to produce a matrix of this thickness may vary.
Therefore, this time period must be determined empirically for each cell type.
Remove cells from matrix
18. Carefully aspirate the medium and rinse gently with 2 ml PBS by touching the pipet
against the dish wall rather than at the bottom of the dish where the cells are located.
19. Gently add 1 ml of prewarmed (37◦ C) extraction buffer.
If coverslips are being used, gently lift the coverslips with the fine-pointed tweezers (or a
syringe needle) so that extraction buffer reaches under the coverslip. This step will ensure
that the matrix deposited on the coverslip will be separated successfully from the matrix
deposited on the bottom of the culture dish. This will facilitate subsequent handling of the
coverslips without tearing the delicate matrix.
20. Observe the process of cell lysis using an inverted microscope. Incubate at 37o C
until no intact cells are visualized (∼3 to 5 min; see Fig. 10.9.1B).
Remove cellular debris
21. Slowly add 2 to3 ml PBS to dilute the cellular debris. Gently pipet the PBS on the
side of the dish to avoid disturbing the newly formed matrix. Store dishes overnight
at 4◦ C to avoid disturbing the matrix.
The above dilution process should be carried out gently to prevent turbulence that may
cause the freshly denuded matrix-layer to detach from the surface.
22. As cautiously as possible (using a pipet), aspirate the diluted cellular debris, but
without completely aspirating the liquid layer so that the matrix surface remains
hydrated at all times.
Do not attempt to aspirate the whole volume. This will prevent removing the matrix layer.
23. Gently add another 2 ml of PBS and gently aspirate the PBS as described in steps 21
and 22.
24. (Optional) If necessary, minimize DNA by treating the matrix with DNase I. Add 2
ml DNase I and incubate 30 min at 37◦ C.
25. At the end of the incubation, aspirate the enzyme and wash two times with 2 ml PBS.
Extracellular
Matrix
10.9.5
Supplement 33
26. Cover the matrix-coated plates (or coverslips) with at least 3 ml PBS supplemented
with 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml Fungizone. Seal
with Parafilm and store for up to 2 or 3 weeks at 4◦ C.
For signal transduction assays, store the matrix-coated plates in serum-free medium (see
Commentary).
Matrices can also be stored for at least 3 weeks at −80◦ C (longer times have not yet been
tested) without compromising matrix integrity, when compared with matrices from the
same batch stored at 4◦ C. To store matrices at −80◦ C, rinse matrix dishes two times with
sterile, nanopure H2 O and carefully aspirate all the liquid. Then label dishes to indicate
the date of freezing for future reference, seal with Parafilm, and place at −80◦ C. When
needed, thaw matrix dishes at room temperature and rehydrate with PBS prior to use for
cell attachment or replating (see Support Protocols 1 and 5).
27. Confirm the integrity of the matrices directly before use. Examine for matrix integrity
using an inverted phase-contrast microscope.
The matrices should be attached to the culture surface and appear similar to the example
in Figure 10.9.1B.
SUPPORT
PROTOCOL 1
FIXATION OF EXTRACTED MATRICES FOR LACK OF PLIABILITY
ANALYSES
For certain experiments designed to analyze the effect of rigidity or pliability of the
matrix on cell behavior, it is necessary to chemically rigidify the prepared matrices. To
this end, matrices are fixed with 1% glutaraldehyde prior to cell plating and analysis.
Materials
1% (v/v) glutaraldehyde in PBS (see recipe)
Tissue culture dishes or coverslips with matrix
1 M ethanolamine (see recipe)
Penicillin/streptomycin
Fungizone
Parafilm
1. Aspirate PBS, add 2 ml of 1% glutaraldehyde (prediluted in PBS) to each tissue
culture dish or well, and incubate 30 min at room temperature.
2. Wash coverslips or culture dishes three times, for 5 min each, with 2 ml PBS at room
temperature.
3. Add 2 ml of 1 M ethanolamine to each dish and incubate 30 min at room temperature.
4. Repeat the PBS washes (step 2).
5. Cover the matrix-coated plates (or coverslips) with at least 3 ml PBS supplemented
with 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml Fungizone. Seal
with Parafilm. Store up to 2 or 3 weeks at 4◦ C or at −80◦ C as described for NIH-3T3
matrices (see Basic Protocol, step 26).
ASSESSING THE QUALITY OF FIBROBLAST-DERIVED
THREE-DIMENSIONAL MATRICES
Preparation of
Fibroblast
Extracellular
Matrices
The quality of fibroblast-derived 3-D matrices can be tested by one of three assays presented here as support protocols. The first two assays, induction of rapid cell attachment
(see Support Protocol 2) and rapid acquisition of spindle-shape morphology (see Support
Protocol 3), are based on examination of fluorescently labeled cells plated on 3-D matrices. The prelabeling with a fluorescent dye is required to enhance the observation of cells
10.9.6
Supplement 33
within fibroblast-derived 3-D matrices. NIH-3T3 matrix quality can also be assessed in
a third assay to check for down regulation of activated FAKpY397 when normal fibroblasts are plated within prepared NIH-3T3 matrices (see Support Protocol 4). These are
referred to as “re-plated” fibroblasts. To analyze the level of activated [FAKpY397 /total
FAK] of normal fibroblasts re-plated within these matrices, the ratio of FAKpY397 /total
FAK is compared to the level of FAKpY397 /total FAK of the same cells plated on a
traditional fibronectin-coated tissue culture dish, (2-D surface; see Support Protocol 2,
cell attachment assay). Typically, a 1.5- to 4-fold reduction in FAK pY397 /total FAK in
normal fibroblasts replated into NIH-3T3 matrices is observed (Cukierman 2001; see
Support Protocol 4).
Cell Attachment Assay
Human or mouse fibroblasts can be used to evaluate the cell adhesion–promoting activity
of the fibroblast-derived 3-D matrices. It has been reported that these in vivo–like 3-D
matrices (NIH-3T3) are about six-fold more effective than 2-D substrates in mediating
cell adhesion as quantified by a 10-min cell attachment assay (Cukierman et al., 2001).
Briefly, cell nuclei are prelabeled to avoid any background staining from DNA debris on
the 3-D matrix. The live prelabeled cells are rinsed free of excess dye, trypsinized, and
plated on the fibroblast-derived 3-D matrix to be assessed or onto control fibronectincoated surfaces. After 10 min, nonattached cells are washed away, and attached cells are
quantified by counting nuclei.
SUPPORT
PROTOCOL 2
Materials
Semi-confluent fibroblasts (human or mouse) in a 15-cm dish
Confluent medium with fetal bovine serum (see recipe)
Hoechst 33342 stock solution (see recipe)
Phosphate-buffered saline (PBS; APPENDIX 2A), 4◦ C and room temperature
Trypsin/EDTA solution (see recipe)
Glass-bottom no. 1.5 dishes (MatTek Corporation): three containing
fibroblast-derived 3-D matrix (see Basic Protocol) and three with pre-coated 2-D
fibronectin (see recipe)
Fixing solution (see recipe)
15-ml polypropylene conical tubes
Tissue culture centrifuge equipped with rotor suitable for conical 15-ml tubes
Fluorescence inverted microscope equipped with an appropriate camera and set of
filters to visualize Hoechst 33342 (see APPENDIX 1E)
Image analysis software capable of counting objects (optional)
Label cells
1. Start with a semi-confluent 15-cm culture dish containing fibroblasts (mouse or
human); aspirate and discard the culture medium.
2. Add 20 ml of confluent medium containing 40 µl of Hoechst 33342 stock solution
(1:500) to the cells. Incubate 15 min at 37◦ C.
Harvest cells
3. Rinse four times with 10 ml PBS at room temperature, 1 min each rinse.
4. Add enough trypsin/EDTA solution to cover the cell layer, aspirate excess liquid,
and observe under an inverted microscope until cells are detached from the culture
dish (1 to 3 min).
Prepare cell suspension
5. Collect the cells in 10 ml of confluent medium into a 15-ml polypropylene conical
tube and take a sample for counting (UNIT 1.1).
Extracellular
Matrix
10.9.7
Supplement 33
6. Pellet the cells by centrifuging 5 min at 100 × g, room temperature.
7. Discard the supernatant and gently resuspend the cells with confluent medium to a
final concentration of 3.5 × 105 cells/ml.
8. Rotate cells in suspension for 20 min at 37◦ C.
Allow cells to attach
9. Carefully place a 150-µl drop of cell suspension onto the glass-bottom part of the
dishes coated with 3-D matrix or 2-D matrix controls. Incubate 10 min at 37◦ C.
10. Remove from the incubator and tilt the dishes slightly to dislodge the medium droplet
containing unattached cells from the glass portion onto the plastic portion of the dish
and then aspirate.
11. Rinse the dishes by slowly adding (to the plastic portion of the dishes) 3 ml of 4◦ C
PBS.
Fix cells
12. Aspirate PBS carefully and add 2 ml of fixing solution. Incubate 20 min at room
temperature.
13. Aspirate and add 2 ml PBS at room temperature.
Visualize and analyze attached cells
14. Using an inverted fluorescence microscope with appropriate excitation wavelength
and excitation and emission filters (APPENDIX 1E), acquire five random images of the
nuclei from each one of the six dishes utilizing a 10× or 20× objective and count
the nuclei.
Counting of the nuclei can be done automatically utilizing commercially available image
analysis software capable of counting objects (e.g., MetaMorph from Universal Imaging
Corporation). If the counting is done automatically, then images should be acquired with
a 10× objective. However, if the nuclei are to be counted manually, then a 20× objective
is recommended.
The mean number of cells attached to the fibroblast-derived 3-D matrix should be up
to six-fold higher than the number attached to the 2-D matrix control. This result will
confirm the quality of the NIH-3T3-derived 3-D matrix (Cukierman et al., 2001).
SUPPORT
PROTOCOL 3
Determination of Cell Shape
Human or mouse fibroblasts can be used to evaluate induction of spindle-shaped cell
morphology promoted by a good-quality in vivo–like 3-D matrix. A recent report has
established that fibroblasts will acquire an in vivo–like spindle-shaped morphology in
cell-derived 3-D matrices 5 hr after plating (Cukierman et al., 2001). The protocol consists
of prelabeling live fibroblast membranes with a fluorescent dye and incubating the cells on
fibroblast-derived 3-D matrices or controls for a period of 5 hr. After this period of time,
the fibroblast-derived 3-D matrix promotes a spindle-shaped morphology resembling in
vivo fibroblast morphology, thereby confirming the quality of the 3-D matrices.
Materials
Preparation of
Fibroblast
Extracellular
Matrices
Fibroblast-derived 3-D matrix-covered coverslips (see Basic Protocol)
Fibronectin 2-D matrix-coated coverslips (see recipe)
2% (w/v) heat-denatured BSA (see recipe)
Semi-confluent 15-cm dish of fibroblasts
10.9.8
Supplement 33
1,1 -dioctadecyl-3,3,3 ,3 -tetramethylindocarbocyanine perchlorate (DiI) stock
solution (see recipe)
Fixing solution (see recipe)
Prolong Gold mounting medium (Invitrogen)
35-mm tissue culture dishes or 6-well plates
Inverted microscope
15-ml polypropylene conical tubes
Tissue culture centrifuge equipped with rotor suitable for 15-ml conical tubes
Fine-point forceps (e.g., Dumont 4)
Glass microscope slides
Fluorescent microscope equipped with digital camera
Image analysis software capable of measuring elliptical Fourier parameters
Block nonspecific cell binding with BSA
1. Cautiously place fibroblast-derived 3-D matrix and control-coated coverslips (matrix
face up) into 35-mm tissue culture dishes (or 6-well plates).
2. Block nonspecific cell binding by adding 2 ml of 2% heat-denatured BSA and
incubate for 1 hr at 37◦ C.
3. Rinse all blocked coverslips with 2 ml PBS.
At this point, coverslips are ready to be seeded with the prelabeled cells.
Label cell membrane with DiI
4. Start with a semi-confluent 15-cm dish of fibroblasts; aspirate and discard the culture
medium.
5. Rinse the cell layer briefly with 1.5 ml trypsin/EDTA.
This rinse will remove traces of serum that contain trypsin inhibitors.
6. Add enough trypsin/EDTA solution to cover the cell layer, aspirate excess liquid,
and observe under an inverted microscope until cells are detached from the culture
dish (1 to 3 min).
7. Collect the cells in 10 ml of confluent medium containing 4 µg/ml DiI into a 15-ml
polypropylene conical tube.
8. Incubate the cells with the dye in suspension by rotating gently for 30 min at 37◦ C.
9. Pellet the cells by centrifugation 5 min at 100 × g, room temperature.
10. Discard the supernatant by aspiration, and gently resuspend the cells in confluent
medium to a final volume of 10 ml.
11. Repeat steps 9 and 10 four additional times to remove any remaining free dye.
12. Count cells (UNIT 1.1) and dilute (with confluent medium) to a final concentration of
1 × 104 cells/ml.
Plate labeled cells
13. Carefully aspirate PBS from the coverslips in step 3.
14. Add 2 ml of the diluted cell suspension to the dishes containing coverslips and
incubate 5 hr at 37◦ C.
For fast qualitative analysis, cells can be observed and photographed at the end of 5 hr
with an inverted fluorescence microscope (see APPENDIX 1E for wavelength information).
15. Aspirate medium and rinse with 2 ml PBS.
Extracellular
Matrix
10.9.9
Supplement 33
Fix cells
16. Aspirate PBS and fix with 1 ml of fixing solution for 20 min at room temperature.
17. Aspirate fixing solution and rinse with 2 ml PBS.
18. Rinse with 2 ml water to eliminate residual salt.
19. Carefully lift coverslip with fine-point forceps and gently discard excess liquid by
touching the edge of the coverslip onto a paper towel.
20. Mount coverslips (cells face down) on a droplet (∼20 µl) of Prolong Gold mounting
medium placed on a glass microscope slide.
21. Allow mounted samples to dry in the dark for ∼1 hr at room temperature.
At this point, samples are ready for morphometry analysis, or they can be stored overnight
in the dark at room temperature before transferring to ≤4◦ C.
Perform morphometric analysis
22. Acquire fluorescent digital images, slightly over-exposing to visualize the contour
of the cells (for wavelength, see APPENDIX 1E).
Use a magnification that will allow visualization of an entire cell in each image. Randomly
capture images of at least 12 cells per sample and a minimum of 36 cells per substrate.
23. Perform the measurements for both the length (span of the longest cord) and the
breadth (caliper width) of each cell.
24. Calculate the inverse axial ratio by dividing length by breadth.
The inverse axial ratio corresponds to the elliptical form factor (EFF) morphometric
parameter found in the integrated morphometry analysis (IMA) function of MetaMorph
software (Universal Imaging Corporation).
The mean inverse axial ratio induced by a high-quality NIH-3T3-derived 3-D matrix
should be about three-fold greater than that induced by the 2-D fibronectin control
(Cukierman et al., 2001).
SUPPORT
PROTOCOL 4
Lysis of Re-Plated Fibroblasts for Western Blot Analyses
To assure the quality of a batch of NIH-3T3-derived 3-D matrices, the levels of FAK
activity (FAKpY397 ) must be down-regulated at least 1.5 fold (Cukierman et al., 2001)
when compared to classic 2-D cultures. This protocol describes how to lyse normal
human or murine primary fibroblasts after re-plating within 3-D matrices for biochemical
analysis by immunoblotting (see UNIT 6.2). In brief, re-plated normal fibroblasts are lysed
and subjected to immunoblot analysis. Cell lysate extracts can also be stored for later
analyses (see step 11).
Materials
Preparation of
Fibroblast
Extracellular
Matrices
Matrix-coated ≥35-mm dishes (see Support Protocol 2)
Fibronectin-coated ≥35-mm dishes
Cell suspension from confluent cultures of fibroblasts
Lysis buffer (modified RIPA) reagent (see recipe) supplemented with protease and
phosphatase inhibitors (see recipe), ice cold
Normal human or murine fibroblasts re-plated in 3-D matrix dishes (see Basic
Protocol, step 18)
Phosphate-buffered saline without Ca2+ or Mg2+ (CMF-PBS; APPENDIX 2A), ice cold
Dry ice/isopropanol bath
10.9.10
Supplement 33
5× sample buffer supplemented with β-mercaptoethanol
Anti-FAKpY397 and anti-total FAK (see recipes)
Glutaraldehyde-3-phosphate dehydrogenase (GAPDH)
37◦ C, 10% CO2 humidified incubator
Cell scraper (Costar, Fisher Scientific)
1.5-ml microcentrifuge tubes (Eppendorf)
Sonicator (e.g., Branson Sonifier 150)
Scion image software beta version 4.03
Additional reagents and equipment for calculating the amount of 5× sample buffer
supplemented with β-mercaptoethanol (UNIT 6.1) and detecting proteins by
immunoblotting (UNIT 6.2)
Re-plate fibroblasts within 3-D matrices
1. Block nonspecific binding in 3-D matrices with BSA in a 35-mm dish (or 6-well
multiwell dish (see Support Protocol 2, steps 1 and 2).
When using 2-D substrates, precoat 35-mm dishes with 1 ml of 5 µg/ml fibronectin or
other matrix protein (see recipe for pre-coated 2-D fibronectin dishes) for 1 hr at 37◦ C.
Alternatively, use uncoated dishes in PBS for 2-D control substrates.
2. Plate 2 ml of cell suspension at a final concentration of 1 × 105 cells/ml in confluent
medium for each dish and incubate 20 hr in a 37◦ C, 10% CO2 humidified incubator
(see Basic Protocol, steps 1 to 4).
Lyse cells within matrices
3. Supplement 10 ml of ice-cold lysis buffer with protease and phosphatase inhibitors.
4. Carefully aspirate confluent medium from fibroblasts re-plated in 3-D matrix dishes
(see Basic Protocol, step 18).
5. Gently add ice-cold PBS and repeat steps 4 and 5 for a total of two PBS washes.
6. Carefully aspirate again and tip the dishes for 1 min to accumulate the excess PBS
on one side of the dish (∼30◦ to bench top).
It is important to remove all traces of PBS to prevent diluting lysates with PBS for
the purpose of maintaining a uniform volume of lysate for different samples. This will
yield a relatively consistent protein loaded onto SDS-PAGE gels for immunoblotting (see
UNIT 6.2).
7. Carefully aspirate the excess PBS to avoid detaching the matrix layer.
8. Place the dishes (usually a 35-mm dish) on ice and add 200 to 300 µl of ice-cold
lysis buffer.
For larger dishes, add a proportionally larger volume of lysis buffer.
9. Incubate on ice for 5 min with gentle rocking.
Collect the lysate
10. Scrape the cells and matrix from the dish with the cell scraper. Then, tilt the
dish toward one side and collect the lysate mixture into a 1.5-ml microcentrifuge
tube.
11. Sonicate each tube of cell lysate using the remote setting of 3 (medium power) for
30 sec for each tube.
12. Centrifuge the lysates 15 min at 16,100 × g, 4◦ C.
Extracellular
Matrix
10.9.11
Supplement 33
13. Carefully remove the supernatant and transfer to a fresh, labeled tube. If not used
immediately, quick freeze the lysates in a dry ice/isopropanol bath and store for 1 to
2 weeks at −80◦ C.
To quick-freeze cell lysates, prepare a dry ice/isopropanol bath by adding ∼100 ml of
isopropanol to a 400-ml beaker and placing in an ice bucket containing dry ice pellets in
the fume hood. Allow the isopropanol to cool for 30 min. Add 1.5-ml microcentrifuge tubes
containing freshly prepared lysates to a tube-rack and lower the rack into the isopropanol
so that the lysate volume is completely immersed. The lysates should be frozen almost
immediately (smaller aliquots are better). Finally, quickly transfer the tubes to a −80◦ C
freezer. Samples are stable for 1 to 2 weeks. Each individual protein should be tested
since there is some variability.
Analyze lysates
14. To analyze the matrix by immunoblotting, calculate the amount of 5× sample buffer
supplemented with β-mercaptoethanol (UNIT 6.1) that is needed to make a 1× final
concentration after addition to the sample; and add that amount to appropriate lysate
samples.
15. Analyze signaling proteins by immunoblotting and immunodetection (UNIT 6.2) using
antibodies to FAKpY397 and total FAK.
For analysis of phosphoproteins, incubate primary antibodies with TBST with a final
concentration of 5% (w/v) BSA.
For total FAK and other non-phosphorylated protein epitopes, primary antibodies, all
secondary antibodies, and for blocking buffer, use 5% (w/v) nonfat dried milk in TBST as
the diluent.
16. Scan individual protein bands corresponding to FAKpY397 or total FAK and quantify
their optical densities using the public Scion image software beta version 4.03 by
means of the Gelplot2 macro. To adjust for sample loading, quantify glutaraldehyde3-phosphate dehydrogenase (GAPDH, 40 kDa) as a total cellular protein control.
The software can be downloaded from the Scion image software Web site at:
http://www.scioncorp.com/frames/fr download now.htm.
The average protein yield of the matrix and fibroblasitic proteins (lysate) is ∼0.5 to 2 mg
per 35-mm dish.
PREPARATION OF TWO-DIMENSIONAL CONTROLS
Any given cell response induced by in vivo–like fibroblast-derived 3-D matrices could be
due to the three-dimensionality of the matrix, its molecular composition, or a combination
of both. The following two support protocols provide methods for obtaining suitable 2-D
control matrices with the same molecular composition as the 3-D matrices.
SUPPORT
PROTOCOL 5
Preparation of
Fibroblast
Extracellular
Matrices
Mechanical Compression of the Fibroblast-Derived 3-D Matrix
This protocol describes how to apply pressure to the fibroblast-derived 3-D matrix to
collapse the matrix to a flat substrate. Mechanical compression of the 3-D matrix ensures
that all natural components of the 3-D matrix are present, lacking only the element
of three-dimensionality. Briefly, the 3-D sample is compressed using a known weight
applied to a given area. The surface that comes into contact with the matrix is covered
with a Teflon film to prevent sticking and to avoid tearing the flattened matrix as the
weight is retracted.
NOTE: Any other materials fulfilling the same purpose can be substituted.
10.9.12
Supplement 33
Materials
Superglue
Absolute ethanol, optional
Fibroblast-derived 3-D matrix on 22-mm circular coverslips
Phosphate-buffered saline (PBS; see APPENDIX 2A)
Flat platform large enough to rest on the ring (see Fig. 10.9.2)
Suitable spacer smaller in width than the diameter of the ring but longer in height
than the depth of the ring (see Fig. 10.9.2)
12-mm round coverslips (Carolina)
Teflon film: protective overlay composed of: 0.001-in. FEP film, on 0.008-in. vinyl
film, with adhesive back (use to cover laboratory bench-tops, Cole-Parmer
Instrument Company)
Cork borer (12-mm diameter)
Biological hood equipped with UV light
Stand equipped with a horizontal ring
Lifting laboratory jack
Parafilm
Weight (∼158 g)
35-mm dishes
Construct weight holder for matrix compression
1. Glue the flat platform to the spacer in such a way that the spacer will protrude slightly
beyond the bottom of the ring when the platform is placed on the ring (Fig. 10.9.2).
Figure 10.9.2 Diagram showing the components of the mechanical compression device.
(a) Weight. (b) Flat platform. (c) Spacer. (d) 12-mm coverslips. (e) Teflon film. (f) Ring stand.
(g) Fibroblast-derived 3-D matrix to be mechanically compressed. (h) Lifting laboratory jack.
Extracellular
Matrix
10.9.13
Supplement 33
2. Glue four 12-mm round coverslips to the end of the spacer (one on top of the other)
as an extension of the spacer using superglue. Allow enough time for the superglue
to completely dry.
This will facilitate penetration of the coverslip portion into the matrix while avoiding contact between the matrix and the rest of the spacer, and it defines the area of compression.
3. Cut a Teflon circle (12-mm diameter) with the cork borer.
4. Cover the last coverslip with the Teflon film.
5. Sterilize materials by exposing them to a UV light in a biological hood for several
hours with the Teflon film facing the light.
If the compressed matrices are to be in contact with cells for only short periods of time
(e.g., for the 10-min cell attachment assay), rinsing the Teflon film with ethanol and
air-drying should be sufficient to prevent contamination.
6. Place the glued platform with spacer on the ring portion of the stand with the Teflon
facing down.
7. Cover the flat upper surface of the laboratory jack with Parafilm and position the
jack under the ring.
8. Set the weight on the platform and level the ring so that the Teflon film is situated
parallel to the surface of the jack (see Fig. 10.9.2).
Mechanically compress the 3-D matrix
9. Position the fibroblast-derived 3-D matrix-coated coverslip (matrix face up) onto the
jack directly underneath the Teflon film.
10. Slowly raise the laboratory jack until the matrix contacts the Teflon film and the
platform rises above the ring. Wait for 2 min.
At this point, the entire weight should be resting on the matrix, compressing it at a specific
weight per unit area.
11. Slowly lower the jack until the platform rests once again on the ring, and the
compressed matrix is separated from the Teflon film.
12. Place the coverslip with the compressed matrix into a 35-mm dish. Carefully add 2
ml PBS and examine by phase-contrast microscopy to confirm continued integrity
of the compressed matrix.
SUPPORT
PROTOCOL 6
Solubilization of Fibroblast-Derived 3-D Matrix
This protocol describes how to solubilize fibroblast-derived 3-D matrix to generate a
protein mixture that can be used for subsequent coating of surfaces or biochemical
analysis. Briefly, the matrices are treated with a guanidine solution to denature and
solubilize the matrix components, thereby producing a liquid mixture that can be stored
and used for coating surfaces.
Materials
Preparation of
Fibroblast
Extracellular
Matrices
Fibroblast-derived 3-D matrices on 35-mm dishes (see Basic Protocol)
Solubilization reagent (see recipe)
Cell scraper (e.g., rubber policeman, Costar brand, Fisher Scientific )
1.5-ml microcentrifuge tubes
Rotator at 4◦ C
Microcentrifuge
10.9.14
Supplement 33
Prepare dishes
1. Aspirate PBS from fibroblast-derived 3-D matrix-covered dishes.
2. Tip the dishes for 1 min to accumulate the excess PBS on one side of the dish (∼30◦
to bench top).
3. Aspirate the excess PBS carefully to avoid detaching the matrix layer.
Solubilize matrix
4. Place the dishes on ice and add 300 µl of solubilization reagent. Incubate 5 min on
ice.
5. Scrape the dish with a cell scraper toward one side of the dish and pipet the mixture
into a 1.5-ml microcentrifuge tube.
6. Add an additional 200 µl solubilization reagent. Rotate 1 hr at 4◦ C.
Collect solubilized matrix
7. Microcentrifuge 15 min at 12,000 × g, 4◦ C.
8. Transfer the supernatant into a fresh 1.5-ml microcentrifuge tube. Store the
solubilized matrix in solubilization reagent indefinitely at 4◦ C.
The average protein concentration is 1 to 3 mg/ml.
ISOLATION OF PRIMARY FIBROBLASTS FROM FRESH TISSUE SAMPLES
NIH-3T3 cells are particularly well-suited for mesenchymal cell–derived matrix production because they are homogeneous and provide batch-to-batch consistency. When
grown in FBS, their ability to grow at high densities and lack of contact inhibition allows
the NIH-3T3 cells to produce a thicker matrix, usually ≥10 µm, which is optimal for
cell studies of 3-D cultures (see Critical Parameters). However, other primary fibroblasts
and fibroblast cell lines are also suitable for the production of cell-derived matrices. This
protocol describes harvesting of primary fibroblasts from fresh tissue samples.
SUPPORT
PROTOCOL 7
Materials
Fresh tissue samples (murine or human surgical)
Phosphate-buffered saline (PBS; APPENDIX 2A) supplemented with antibiotics (see
recipe), 4◦ C
Confluent medium with fetal bovine serum (FBS; see recipe)
Ciprofloxicin (Invitrogen), optional
60-mm tissue culture dishes
Dissecting scissors, tweezers, and scalpels (Fisher Scientific)
12-well or 6-well tissue culture plates
Sterile laminar flow hood
75-cm2 tissue culture flasks
Prepare tissue samples
1. Rinse tissue samples obtained immediately after surgery (human or murine) three
times in a 60-mm tissue culture dish with pre-cooled (to 4◦ C) PBS supplemented
with antibiotics.
2. Aspirate supplemented PBS, finely chop the tissue sample into 1-mm2 pieces using
a sterile scalpel, assisting with sterile tweezers (see Amatangelo et al., 2005).
3. Using sterile dissecting scissors, make multiple scratches into the plastic surface of
a 12-well or 6-well tissue culture dish in a star-like configuration.
Extracellular
Matrix
10.9.15
Supplement 33
4. Wash the dish two times with 1 ml (12-well plate) or 2 ml (6-well plate) PBS and
gently press tissue pieces into the indentations created by the scratches.
Isolate primary fibroblasts
5. Allow plate to dry for 5 min under the sterile laminar hood.
6. Gently add 1 ml (12-well plate) or 2 ml (6-well plate) confluent medium with FBS to
each of the wells, ensuring that the tissue samples remain attached to the scratched
surfaces. Incubate 2 to 7 weeks. Replace the confluent medium every other day until
primary fibroblasts emerge from the tissue pieces.
This process normally takes 2 to 7 weeks depending on the tissue source.
7. (Optional) As an additional measure to prevent contamination of freshly isolated
surgical samples, culture half of the tissue pieces in confluent medium supplemented
with 250 ng/ml to 2.5 µg/ml Fungizone and 10 µg/ml ciprofloxacin.
8. After fibroblasts are grown to ∼70% confluence in multiwell dishes remove the
tissue pieces.
9. Trypsinize fibroblasts with trypsin/EDTA and passage into a 75-cm2 tissue culture
flask (see Basic Protocol, steps 1 to 4).
10. Once fibroblasts reach confluence in a 75-cm2 flask, harvest and freeze them for
future experimental analysis, and/or use them to produce fibroblast-derived matrices
between passage 2 and 6 (see Basic Protocol).
The authors start counting passages once the fibroblasts are initially transferred into a
15-cm dish. The fibroblasts are stable by morphological and biochemical criteria to at
least passage 6. Morphological criteria include an elongated cell shape and the shape of
the nuclei by Hoechst staining. For tumor-associated fibroblasts, elliptically shaped nuclei
typical of myofibroblastic cells have been observed. They have not yet been genetically
characterized.
Use deionized, distilled water or equivalent in all recipes and protocol steps. For common stock
solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.
Anti-FAKpY397
Make a 1:1000 to 1:2500 dilution of anti-FAKpY397 (Biosource International or
Covance) in 5% (w/v) BSA (Sigma)/TBST (see UNIT 6.2; see recipe for TBST). Store
up to 12 months at −20◦ C.
Anti-total FAK
Use a 1:2500 dilution of anti-total FAK (Upstate Cell Signaling Solutions) in 5%
(w/v) nonfat dried milk (Carnation, Fisher Scientific)/TBST (see UNIT 6.2; see recipe
for TBST). Store up to 12 months at −20◦ C.
Confluent medium with fetal bovine serum
High-glucose Dulbecco’s modified Eagle medium supplemented with:
10% (v/v) fetal bovine serum (APPENDIX 2A)
100 U/ml penicillin
100 µg/ml streptomycin
Store for 1 month at 4◦ C
Preparation of
Fibroblast
Extracellular
Matrices
For surgical or fine needle aspirates tissue samples, 250 ng/ml to 2.5 µg/ml Fungizone can
be added to the cultures.
10.9.16
Supplement 33
Culture medium with calf serum
High-glucose Dulbecco’s modified Eagle medium supplemented with:
10% (v/v) calf serum
100 U/ml penicillin
Store for 1 month at 4◦ C
DiI solution
Dilute 2.5 mg/ml 1,1 -dioctadecyl-3,3,3 ,3 -tetramethylindocarbocyanine perchlorate (DiI; Molecular Probes) stock solution in ethanol to 4 µg/ml with confluent
medium (see recipe) and sterilize by filtration using a 0.22-µm filter. Store up to 12
to 24 months at –20◦ C.
Ethanolamine, 1 M
Prepare a 1 M solution of ethanolamine (Sigma-Aldrich) sterile water by adding
0.062 ml of ethanolamine per milliliter of water. Filter sterilize through a 0.2-µm
filter unit. Prepare fresh.
Extraction buffer
Phosphate-buffered saline (PBS; APPENDIX 2A) containing:
0.5% (v/v) Triton X-100
20 mM NH4 OH
Store up to 1 month at 4◦ C
Fixing solution
Into a 50-ml polypropylene conical tube, add:
2 g sucrose
10 ml 16% (w/v) solution paraformaldehyde (EM-grade from Electron Microscopy
Sciences)
Phosphate-buffered saline (PBS; APPENDIX 2A) to a final volume of 40 ml
Prepare fresh
Gelatin solution, 0.2% (w/v)
Prepare a 0.2% (w/v) gelatin solution in PBS (APPENDIX 2A). Autoclave the solution,
cool, and filter through a 0.2-µm filter. Prepare fresh.
Glutaraldehyde solution in PBS, 1% (v/v)
Dilute a 25% stock of glutaraldehyde (Sigma) to 1% glutaraldehyde in PBS
(APPENDIX 2A). Thaw a 10-ml aliquot of the stock and dilute to a final volume
of 250 ml in PBS. Filter sterilize through a 0.2-µm filter unit and store in 50-ml
aliquots at −20◦ C.
Heat-denatured BSA, 2% (w/v)
Dissolve 2 g BSA fraction V (Sigma) in 100 ml water and filter sterilize using a
low-protein-binding 0.22-µm filter. Store indefinitely at 4◦ C.
Just prior to use, heat the amount needed 5 min at 65◦ C or until the solution starts
to appear slightly opaque (not milky). Cool to room temperature before using for
blocking procedures. Do not store the heat-denatured BSA.
Hoechst 33342 stock solution
Prepare a 2 mM (MW 615.9 g) Hoechst 33342 (bisbenzimide H 33342 fluorochrome, trihydrochloride; Calbiochem) solution in water. Store at 4◦ C, protected
from light.
Extracellular
Matrix
10.9.17
Supplement 33
Lysis buffer (modified RIPA) reagent
50 mM NaCl
1% (w/v) deoxycholic acid, sodium salt (Fisher)
48 mM NaF
1% (w/v) glycerol (Fluka)
1% (w/v) Triton X-100 (Sigma)
Adjust to 100 ml with MilliQ H2 O
Store 3 to 6 months at 4◦ C
Matrix medium with ascorbic acid
Confluent medium (see recipe) containing:
acid sodium salt (Sigma) at a final concentration of 50 µg/ml
Filter sterilize with a 0.2-µm filter
Prepare fresh
L-ascorbic
Ascorbic acid should be freshly prepared just prior to use as a 1000× stock concentration
of 50 mg/ml in PBS to yield a final concentration of 50 µg/ml. Remove a 5- to 10-ml aliquot
of medium, add ascorbic acid, and after filtering, add the ascorbic acid–containing medium
back to the total volume of medium. In cases where a 500 µg/ml final concentration of
ascorbic acid is added on the first day after cell plating, the stock is diluted only 100-fold
instead of 1000-fold (see Basic Protocol, steps 16 and 17).
PBS+
1 mM CaCl2
1 mM MgSO4
Store 6 to 12 months at room temperature
PBS supplemented with antibiotics
100 U/ml penicillin
2.5 µg/ml Fungizone (amphotericin B; Invitrogen)
10 µg/ml ciprofloxicin
Store 1 to 2 weeks at 4◦ C
Pre-coated 2-D fibronectin dishes
In phosphate-buffered saline (APPENDIX 2A), prepare a 5 µg/ml solution of human
plasma fibronectin (see UNIT 10.5 or Sigma). Immediately add 1 ml per 35-mm
tissue culture dish and incubate for 1 hr at 37◦ C. Remove remaining fibronectin
solution and rinse once with PBS. Prepare diluted fibronectin solution fresh for
each experiment.
The above procedure can be used with any desired protein for coating dishes or coverslips.
If solubilized matrix mixture is to be used (see Support Protocol 6), coat with a 30 µg/ml
protein concentration.
Protease and phosphatase inhibitors
Preparation of
Fibroblast
Extracellular
Matrices
1 mM sodium pyrophosphate
1 mM nitrophenol phosphate
5 mM benzamidine
1 mM PMSF
1 mM sodium orthovanadate
continued
10.9.18
Supplement 33
Serine/threonine phosphatase cocktail inhibitor 1 (100 µl/10 ml lysis buffer; Sigma)
Tyrosine phosphatase inhibitor cocktail 2 (100 µl/10 ml lysis buffer; Sigma)
Prepare fresh
Solubilization reagent
5 M guanidine containing:
10 mM dithiothreitol
Store indefinitely at 4◦ C
Tris-buffered saline with Tween (TBST)
150 mM NaCl
0.5% (v/v) Tween-20 detergent (Sigma)
Adjust pH to 8.0 with 12 M HCl
Store up to 2 weeks at 4◦ C
Trypsin/EDTA solution
2.5 g trypsin
0.2 g EDTA
8 g NaCl
0.4 g KCl
1 g glucose
0.35 g NaHCO3
0.01 g phenol red
Bring up to 1 liter with H2 O
Sterilize by filtration with a 0.2-µm filter and store up to 3 months at −20◦ C
COMMENTARY
Extracellular matrix (ECM) was historically regarded as a passive scaffold that stabilizes the physical structure of tissues. With
time, it became evident that the ECM is much
more than a simple physical scaffold. The
ECM is a dynamic structure capable of inducing (and responding to) a large variety of physiological cell responses regulating the growth,
migration, differentiation, survival, and tissue
organization of cells (Buck and Horwitz, 1987;
Hay, 1991; Hynes, 1999). Integrins are receptors for matrix molecules and can mediate
these cell responses by inducing the formation of membrane-associated multi-molecular
complexes. These integrin-based structures
(cell-matrix adhesions) mediate strong cellsubstrate adhesion and transmit information in
a bi-directional manner between ECM and the
cytoplasm. There are three main cell-to-matrix
adhesions. The “focal adhesion” mediates firm
linkage to relatively rigid substrates (Burridge
and Chrzanowska-Wodnicka, 1996). Focal adhesions cooperate with “fibrillar adhesions”
that generate fibrils from pliable fibronectin
(Katz et al., 2000; Pankov et al., 2000). Fibroblasts require culture for several days at
high cell density to generate 3-D matrices and
evolve “three-dimensional-matrix adhesions.”
The requirements for producing 3-D matrix
adhesions include three-dimensionality of the
ECM, integrin α5 β1 , fibronectin, other matrix component(s), and pliability of the matrix (Cukierman et al., 2001). The fibroblastderived matrix provides an in vivo–like 3-D
environment for cultured fibroblasts, thereby
restoring their normally nonpolar surroundings. The fibroblast-derived 3-D matrix can be
used as a suitable in vitro system to investigate
in vivo–like fibroblast-to-matrix interactions,
such as 3-D matrix adhesion signaling.
Critical Parameters
The phenotype of cultured NIH-3T3 fibroblasts as monitored by cell morphology is extremely important for the successful preparation of 3-D matrix-coated dishes. The fibroblasts should be well-spread and flat under
sparse culture conditions. If elongated cells are
commonly observed in the cell population, recloning of the cell line may be necessary to
achieve greater phenotypic homogeneity. The
NIH-3T3 line obtained from ATCC (ATCC#
CRL-1658) has this morphology and produces
Extracellular
Matrix
10.9.19
Supplement 33
Preparation of
Fibroblast
Extracellular
Matrices
an excellent matrix. The NIH-3T3 cells must
be maintained routinely as sub-confluent cultures in a medium containing calf serum to
retain the correct phenotype. However, if the
matrix deposition at confluence is performed
in the presence of calf serum, the resultant
matrices are thicker but less stable and more
likely to detach from the surface than matrices
obtained after culture in fetal bovine serum.
Therefore, NIH-3T3 cells should be changed
to medium containing fetal bovine serum prior
to matrix deposition. While the NIH-3T3 cells
are being adapted for matrix production in
FBS, they take on a more uniform polygonal
morphology and are not as contact inhibited
as those grown in calf serum. NIH-3T3 cells
do not normally take on a very elongated morphology unless they are cultured within 3-D
matrices. To pre-adapt the NIH-3T3 cells to
fetal bovine serum–containing medium, it is
recommended to culture the cells for 15 to 20
passages before plating for matrix deposition.
The Basic Protocol can be modified for
other fibroblastic cell lines capable of secreting
and assembling fibronectin-based matrices. As
described in Support Protocol 7, the authors
have adapted this protocol for the isolation
of primary fibroblasts obtained from human
and/or murine surgical tissue samples. Fibroblasts can be isolated from tissue samples after
∼2 to 7 weeks in culture. In some cases, the
resulting matrix may be too thick or dense to
obtain efficient extraction. In such cases, more
prolonged cell extraction may be needed with
extensive DNase treatment until no cell debris
is detected. The lack of contaminating cellular
debris (in the case of NIH-3T3 cells) in the matrices has been confirmed by immunoblotting
and immunofluorescence staining for cellular
proteins like actin or GAPDH.
Pre-coating surfaces with gelatin promotes
fibronectin binding and results in smooth layers of relatively homogenous matrices that will
not detach from the surface.
The thickness of NIH-3T3-derived 3-D
matrices is measured using a confocal microscope without dehydration of the matrix
(no mounting or fixing). The resultant thickness observed varies between 8 and 20 µm.
Basic molecular characterization of the matrices revealed the presence (among other
molecules) of fibronectin organized in a fibrillar mesh, collagen I and III but not IV,
and small traces of non-organized laminin and
perlecan.
The integrity of these 3-D matrices must
be confirmed prior to every use. This can
be accomplished by using phase-contrast mi-
croscopy and discarding any matrices that are
torn or detached (see Fig. 10.9.1 B). Moreover,
if matrices are to be used for short-term signal transduction assays under serum-depleted
conditions, then freshly made matrices must
be utilized. Matrices that have been stored at
4◦ C or −80◦ C (up to 2 to 3 weeks) should be
used only after assessment of their integrity by
phase-contrast microscopy. Freshly prepared
or stored matrices can be used to test the induction of cellular responses in the presence of
serum such as attachment, morphology, motility, proliferation, and for immunofluorescence
staining. Additionally, biochemical analysis
of the 3-D matrix can be assessed by immunoblotting to test for cell responsiveness to
three-dimensionality by phospho-FAK down
regulation (see Support Protocol 3).
Anticipated Results
The Basic Protocol is based on the ability of densely cultured fibroblasts (start up at
∼2.5 × 105 cells/ml) to coat any available
tissue culture surface by deposition of their
natural matrix, which gradually forms a 3-D
matrix. This intact, naturally produced ECM
is similar in its molecular organization to mesenchymal fibronectin-based extracellular matrices in vivo (Cukierman et al., 2001). The
basic approach is to allow cells to deposit their
own ECM followed by removal of cells, while
avoiding procedures that may alter or denature the native ECM constituents and supramolecular organization.
One NIH-3T3 semi-confluent (80%) cultured 15-cm dish can yield enough cells to
coat 100 35-mm tissue culture dishes.
Time Considerations
The adaptation step after switching NIH3T3 cell medium to fetal bovine serum for
future matrix deposition requires culturing
the cells for 15 to 20 passages. This adaptation process could take between 5 and 22
weeks depending on the rate of NIH-3T3 cell
proliferation and the dilution factor per passage. The rate of NIH-3T3 proliferation is
dependent upon many factors, including the
growth-promoting abilities of the fetal bovine
serum, which unfortunately is largely batchdependent. Therefore, the time required for
adapting NIH-3T3 cells to growth in FBS
should be determined empirically. At this
point, the NIH-3T3 cells are between 20 and
25 passages, and can be cultured in FBS
for at least 20 to 25 additional passages,
resulting in a total of ≥50 passages. After
that, their morphology starts to become more
10.9.20
Supplement 33
spindle-shaped and, therefore, they can no
longer form optimal matrices. When the NIH3T3 cells become too spindle-shaped, they fail
to form uniform monolayers that upon extraction can ultimately produce uneven matrix
coverage. Matrix production will require between 5 and 9 days. About 2 to 7 weeks are
required from the time of tissue isolation to
harvesting primary fibroblasts.
Literature Cited
Amatangelo, M.D., Bassi, D.E., Klein-Szanto, A.J.,
and Cukierman, E. 2005. Stroma-derived threedimensional matrices are necessary and sufficient to promote desmoplastic differentiation
of normal fibroblasts. Am. J. Pathol. 167:475488.
Buck, C.A., and Horwitz, A.F. 1987. Cell surface
receptors for extracellular matrix molecules.
Annu. Rev. Cell Biol. 3:179-205.
Burridge, K., and Chrzanowska-Wodnicka, M.
1996. Focal adhesions, contractility, and signaling. Annu. Rev. Cell Dev. Biol. 12:463-518.
Cukierman, E. 2005. Cell migration analyses within
fibroblast-derived 3-D matrices. Methods Mol.
Biol. 294:79-93.
Cukierman, E., Pankov, R., Stevens, D.R., and
Yamada, K.M. 2001. Taking cell-matrix adhesions to the third dimension. Science 294:17081712.
Friedl, P., and Brocker, E.B. 2000. The biology of
cell locomotion within three-dimensional extracellular matrix. Cell Mol. Life Sci. 57:41-64.
Geiger, B., Bershadsky, A., Pankov, R., and
Yamada, K.M. 2001. Transmembrane crosstalk
between the extracellular matrix and the cytoskeleton. Nat. Rev. Mol. Cell Biol. 2:793-805.
Hay, E.D. 1991. Cell biology of extracellular matrix, 2nd ed. Plenum Press, New York.
Hynes, R.O. 1999. Cell adhesion: Old and new
questions. Trends Cell Biol. 9:M33-M77.
Katz, B.Z., Zamir, E., Bershadsky, A., Kam, Z.,
Yamada, K.M., and Geiger, B. 2000. Physical
state of the extracellular matrix regulates the
structure and molecular composition of cellmatrix adhesions. Mol. Biol. Cell 11:1047-1060.
Pankov, R., Cukierman, E., Katz, B.Z., Matsumoto,
K., Lin, D.C., Lin, S., Hahn, C., and Yamada,
K.M. 2000. Integrin dynamics and matrix
assembly: Tensin-dependent translocation of
alpha(5)beta(1) integrins promotes early fibronectin fibrillogenesis. J. Cell Biol. 148:10751090.
Key References
Cukierman et al., (2001). See above.
The source for procedures and materials described
in this unit.
Contributed by Dorothy A. Beacham,
Michael D. Amatangelo, and Edna
Cukierman
Fox Chase Cancer Center
Philadelphia, Pennsylvania
Extracellular
Matrix
10.9.21
Supplement 33
Purification and Analysis of Thrombospondin-1
UNIT 10.10
Thrombosponding-1 (TSP-1) is a trimeric matricellular protein that is expressed by many
cells. It contains several different domains that allow it to participate in cell adhesion, cell
migration, and cell signaling. Recently, TSP-1 has been shown to activate transforming
growth factor-β (TGF-β) and to inhibit both angiogenesis and tumor growth. This unit
describes two protocols: the purification of TSP-1 from platelet-rich plasma (see Basic
Protocol 1) and the purification of TSP-1 proteolytic fragments (see Basic Protocol 2).
ISOLATION OF THROMBOSPONDIN-1 FROM HUMAN PLATELETS
TSP-1 is released from platelet α-granules in response to thrombin and can therefore be
readily purified from the supernatant of thrombin-treated platelets. Human platelets can
be obtained from the Red Cross or from hospital blood banks. Outdated pheresis units of
platelet-rich plasma are a good source of TSP-1. Platelets are separated from plasma and
other blood components by a series of centrifugation steps. The isolated platelets are
washed repeatedly to remove plasma proteins and the washed platelets are then activated
by exposure to thrombin. Next the TSP-1-containing supernatant is passed over a
heparin-Sepharose column. Lower-affinity heparin-binding proteins are washed away and
the TSP-1 is eluted under conditions of high salt. The TSP-1-containing fractions are
pooled, precipitated, and loaded onto a 10% to 20% continuous sucrose gradient and
subjected to ultracentrifugation. The gradient is divided into fractions and the protein
concentrations are determined by measuring optical density. The level of purity is
normally >95% as determined by SDS-PAGE (UNIT 6.1).
BASIC
PROTOCOL 1
Materials
Platelet-rich plasma
Baenziger A buffer (see recipe)
Baenziger B buffer (see recipe)
1 M CaCl2 (APPENDIX 2A)
1 N NaOH (optional)
Thrombin
Diisopropyl fluorophosphate (DFP)
Heparin-Sepharose CL-6B (Amersham Pharmacia Biotech)
0.15, 0.25, 0.55, and 2.0 M heparin-Sepharose column buffers (see recipe)
Anti-vitronectin immunoaffinity column: prepare in advance according to
manufacturer’s instructions using an Affi-Gel Hz Immunoaffinity kit (Bio-Rad)
and anti–human vitronectin antibody (e.g., GIBCO/BRL)
Ammonium sulfate
10% and 20% (w/v) sucrose gradient solutions (see recipe)
15- and 50-ml centrifuge tubes (conical bottom preferred)
Preparative centrifuge (Sorvall RC-B3 or equivalent) and rotor (H4000 or
equivalent)
40-ml Oak Ridge centrifuge tubes
High-speed centrifuge (Beckman J2-MC or equivalent) and rotor (JA-20 or
equivalent)
1 × 12–cm chromatography column
Fraction collector and appropriate tubes
Spectrophotometer set at 280 nm
Gradient maker
14-ml ultracentrifuge tubes
Ultracentrifuge (Beckman LM-80 or equivalent) and rotor (SW 41Ti or equivalent)
Contributed by Karen O Yee, Mark Duquette, Anna Ludlow, and Jack Lawler
Extracellular
Matrix
10.10.1
Supplement 17
NOTE: Platelets are temperature sensitive and activated by untreated glass surfaces;
therefore, they should be handled at room temperature in plasticware, and centrifuges and
buffers should be warmed to room temperature before use.
Prepare platelets
1. Transfer platelet-rich plasma to 50-ml centrifuge tubes (conical bottom preferred)
and centrifuge in a Sorvall RC-B3 preparative centrifuge 20 min at 1400 × g (2800
rpm in an H4000 rotor), 20°C.
Pheresis units are preferable, but random donor units of platelet-rich plasma also work well.
2. Carefully pour off the supernatant. Gently resuspend the cell pellet in Baenziger A
buffer at a ratio of 15 ml buffer per 2 ml packed cells.
3. Transfer the platelet suspension to 15-ml centrifuge tubes and centrifuge 8 min at 120 ×
g (800 rpm in an H4000 rotor), room temperature.
Most of the platelets will remain in suspension following this centrifugation, while
erythrocytes and leukocytes will pellet.
4. Leaving behind the red cell pellet, carefully transfer the platelet suspension to 50-ml
centrifuge tubes (∼22 ml per tube).
5. Add Baenziger A buffer to a final volume of 50 ml. Mix by inverting the tube several
times and centrifuge 20 min at 1400 × g (2800 rpm in an H4000 rotor), 20°C.
Wash platelet pellet
6. Carefully pour off the supernatant. Resuspend each cell pellet in 15 ml Baenziger A
buffer and then add buffer to a final volume of 50 ml. Invert the tube to mix and
centrifuge 20 min at 1400 × g (2800 rpm in an H4000), room temperature. Repeat
once.
7. Remove the supernatant and resuspend the pellet in 15 ml Baenziger B buffer. Add
sufficient Baenziger B to achieve a ratio of 50 ml buffer per 2 to 3 ml packed cells.
Mix the tube by inversion.
8. Add 100 µl of 1 M CaCl2 per 50 ml suspension.
From this point on, 2 mM calcium must be present at all times to maintain the conformational integrity of the thrombospondin molecule.
9. Check the pH of the suspension using pH paper. Adjust to pH 7.6 by adding 1 N
NaOH as necessary.
Activate platelets
10. Optional: If the platelets are from outdated units, enhance their response to thrombin
by incubating 5 min in a 37°C water bath.
11. Add 50 U thrombin per 50 ml platelet suspension and immediately mix by gentle
inversion. Continue mixing 2 to 3 min at room temperature, then place on ice.
Platelet aggregation should be evident upon examination of the suspension. The platelets
will form large clumps and settle to the bottom of the tube, causing the supernatant to
appear somewhat clear after 2 to 3 min. Outdated platelets respond more slowly than fresh
ones. Outdated units should therefore be mixed for an additional 2 to 3 min.
Purification and
Analysis of
Thrombospondin-1
12. Remove the cellular debris by centrifuging the tubes 5 min at 1400 × g (2800 rpm in
an H4000 rotor), 4°C. Transfer supernatant to a 40-ml Oak Ridge centrifuge tube.
10.10.2
Supplement 17
From this point on the TSP-1-containing supernatant must be kept on ice and all subsequent
steps must be performed at 4°C.
13. Add sufficient DFP to achieve a final concentration of 1 mM (i.e., 0.181 µl/ml).
CAUTION: DFP is a powerful serum protease inhibitor and is highly toxic. Great care
should be taken in its use. DFP is volatile and should be used in a fume hood.
Isolate TSP-1 supernatant
14. Centrifuge 20 min in a Beckman J2-MC high-speed centrifuge at 34,957 × g (17,000
rpm in a JA-20 rotor), 4°C.
15. Transfer the supernatant to a clean 50-ml tube. Place the sample on ice and leave
overnight at 4°C.
This incubation step is necessary to allow formation of fibrin fibrils, which are then
removed by centrifugation (step 17). If the supernatants are applied to the heparinSepharose column without performing this procedure, the fibrin fibrils will form on the top
of the column and the flow rate will be decreased significantly.
Isolate TSP-1
16. Prepare and pour enough heparin-Sepharose CL-6B, according to the manufacturer’s
instructions, to produce a 5-ml bed volume in a 1 × 12–cm chromatography column.
Equilibrate the column with 50 ml of 0.15 M heparin-Sepharose column buffer.
17. Following the overnight incubation (step 15), centrifuge the supernatant 20 min at
1400 × g (2800 rpm in an H4000 rotor), 4°C. Transfer the supernatant to a new tube.
18. Load the supernatant onto the equilibrated heparin-Sepharose column at a flow rate
of ∼3 ml/min.
The TSP-1 will be immobilized on the column following this step. If necessary, the protocol
may be paused at this point; however, the column should be washed extensively with 0.15
M heparin-Sepharose column buffer before pausing. TSP-1 is stable on the column for 3
to 4 days.
19. Connect the column to a fraction collector with appropriate tubes and elute the
column with 40 ml of 0.15 M heparin-Sepharose column buffer at a flow rate of ∼3
ml/min, collecting twenty 2-ml fractions. Repeat with 0.25 M heparin-Sepharose
column buffer.
Little or no TSP-1 will be present in these first two elutions.
20. Elute TSP-1 by applying 40 ml of 0.55 M heparin-Sepharose column buffer and
collect in 2-ml fractions. Determine which fractions contain protein by measuring
their absorbance at 280 nm. Calculate the total amount of protein in milligrams using
the following formula: total protein = OD280 × 1.08 × volume.
After elution, >80% of total protein is TSP-1.
21. Strip the heparin-Sepharose column by applying 100 ml of 2.0 M heparin-Sepharose
column buffer. Equilibrate and store the column in 0.15 M heparin-Sepharose column
buffer at 4°C.
The column can be used repeatedly if treated in this manner.
22. Pool the protein-containing fractions and apply to an anti-vitronectin immunoaffinity
column.
Although vitronectin is present in only trace amounts in the TSP-1-containing fraction
(<1%), cells adhere strongly to vitronectin, which can pose a problem in certain applications involving purified TSP-1. It is therefore advisable to remove it. The authors are unable
to detect vitronectin in the immunoaffinity flowthrough by immunoblotting.
Extracellular
Matrix
10.10.3
Supplement 17
For storage and reuse of the immunoaffinity column, refer to the manufacturer’s instructions.
23. Transfer the flowthrough to a 40-ml Oak Ridge tube. Precipitate the protein by adding
ammonium sulfate to 40% (w/v). Mix by inverting the tube until the solid is dissolved.
24. Centrifuge the sample 20 min at 34,957 × g (17,000 rpm in a JA-20 rotor), 4°C.
The precipitated protein should form a milky-white pellet following centrifugation. It may
be useful to note the orientation of the tube in the rotor to aid in locating the pellet.
25. Carefully pour off the supernatant and briefly leave the tube inverted to drain away
all remaining liquid.
26. Resuspend the pellet in sufficient 0.15 M heparin-Sepharose buffer so that the protein
concentration is ∼1 mg/ml.
The concentration does not change significantly from the value determined in step 20.
27. Using a gradient maker, prepare 12 ml of a 10% to 20% continuous sucrose gradient
in a 14-ml ultracentrifuge tube.
28. Carefully load the protein-containing sample onto the gradient, using no more than
2 mg protein on each gradient in order to achieve good resolution.
Drawing up the solution into a pipet tip and slowly discharging it by turning the volume
adjustment wheel on the pipettor works well. It is important not to disturb the gradient.
See UNIT 5.3 for more information concerning sucrose gradients.
29. In a Beckman LM-80 ultracentrifuge, centrifuge the gradients 18 hr at 247,605 × g
(38,000 rpm in an SW 41Ti rotor), 4°C.
30. Fractionate the sucrose gradients into 0.5-ml aliquots. Read the absorbance of each
fraction at 280 nm.
The concentration of TSP-1 in each fraction is determined using the equation c = A280/εl,
where ε is the molar extinction coefficient of TSP-1 (1.08 in M–1cm–1) and l is the pathlength
of the cuvette.
TSP-1 is the major peak located in the middle third of the gradient. The peak usually
appears in the eighth fraction from the bottom and continues over approximately eight
fractions. There is also a minor peak located higher in the gradient that is composed mainly
of β-thromboglobulin. The purified TSP-1 can be frozen directly in the sucrose gradient
solution and stored 3 to 5 years at −70°C. If necessary for specific applications, sucrose
can be removed by dialysis (APPENDIX 3C).
BASIC
PROTOCOL 2
Purification and
Analysis of
Thrombospondin-1
ISOLATION OF PROTEOLYTIC FRAGMENTS OF TSP-1
The identification of functional sites within larger proteins can be accomplished by
producing individual domains for functional studies. The proteolytic digestion of native
molecules or the expression of individual domains by recombinant approaches is typically
used for this purpose. While the authors were developing the procedure for the purification
of TSP-1, it was observed that molecules that lacked the 25,000-Da N-terminal domain
were not retained by the heparin-Sepharose column. This observation led to the early
identification of the N-terminal domain as a high-affinity heparin-binding site and to the
development of a procedure to purify this domain. During the development of the isolation
protocol, it was also found that TSP-1 purified in the presence of calcium was distinct
from that purified in the presence of EDTA. The removal of calcium from the protein
renders some regions much more labile to proteolysis (Lawler and Hynes, 1986).
Subsequent sequencing studies revealed that the type 3 repeats are a contiguous set of
calcium-binding sites. Removal of calcium causes the type 3 repeats and the adjacent
10.10.4
Supplement 17
C-terminal domain to unfold and become more labile to proteolysis. Thus, it is difficult
to design a protocol that utilizes proteolysis in order to isolate the type 3 repeats and the
C terminus of TSP-1. These domains are either resistant to digestion in the presence of
calcium or are readily digested in its absence. In the absence of calcium, the central
70,000-Da core of the protein can be produced by chymotryptic digestion. This structure
contains the intrachain disulfide bonds and hence is trimeric with a molecular weight of
210,000 Da. A procedure for preparing the N-terminal heparin-binding domain and the
central core region is provided below.
Materials
TSP-1 (see Basic Protocol 1)
TBS (see recipe) containing 2 mM CaCl2
0.5 M EDTA (APPENDIX 2A)
Chymotrypsin
Diisopropyl fluorophosphate (DFP)
0.8 × 3–cm column of immobilized soybean trypsin inhibitor (Pierce)
0.8 × 3–cm column of heparin-Sepharose CL-6B (Amersham Pharmacia Biotech)
0.15, 0.25, and 0.55 M heparin-Sepharose column buffers (see recipe)
Centriplus centrifugal filter device (3000-Da cutoff; Millipore)
1.3 × 30–cm column of Sephadex G-200
Fraction collector
Spectrophotometer set at 280 nm
1. Begin with 5 mg purified TSP-1 in 3 ml TBS containing 2 mM CaCl2 on ice. Add 0.5
M EDTA to a final concentration of 5 mM.
2. Dissolve chymotrypsin in TBS to a final concentration of 1 mg/ml and add 50 µl to
the sample.
This gives an enzyme-to-substrate ratio of 1:100 (w/w).
3. Digest 20 hr in a covered ice bucket in a 4°C cold room.
This approach produces a very reproducible digestion pattern.
4. Terminate the digestion by adding DFP to 1 mM (0.181 µl/ml) and incubating an
additional 2 hr on ice.
5. Pass the sample over a 0.8 × 3–cm column of immobilized soybean trypsin inhibitor
to remove the chymotrypsin.
6. Collect the flowthrough from the column and apply to a 0.8 × 3–cm heparinSepharose column connected to a fraction collector with appropriate tubes.
7. Collect 1-ml fractions while the sample is flowing onto the column and throughout
the elution. Elute the column in a stepwise fashion with 20 ml (each) of 0.15, 0.25,
and then 0.55 M heparin-Sepharose buffer.
The 25,000-Da fragment elutes in the 0.55 M heparin-Sepharose buffer. The protein is
∼95% pure and can be used directly or concentrated further. Some applications may require
dialysis to reduce the salt concentration.
The 210,000-Da fragment is in the initial flowthrough fractions. It should be dialyzed and
purified on a Sephadex G-200 column as described in the following steps.
8. Pool the flowthrough fractions and concentrate to 2 ml using a Centriplus centrifugal
filter device.
9. Apply this sample to a 1.3 × 30–cm Sephadex G-200 column equilibrated with TBS
at 4°C.
Extracellular
Matrix
10.10.5
Supplement 17
Sephadex G-200 is fragile and care should be exercised to follow the manufacturer’s
directions for removing fines and for pouring and eluting the column.
10. Apply TBS and collect fifty 1-ml fractions. Measure the OD280 of each fraction.
The 210,000-Da fragment is eluted in the molecular weight peak near the void volume of
the column. This material can be concentrated using the Centriplus centrifugal filter device
(step 8) and should be >95% pure.
Use deionized or distilled water in all recipes and protocol steps. For common stock solutions, see
APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.
Baenziger A buffer
0.102 M NaCl
0.0039 M K2HPO4
0.0039 M Na2HPO4
0.022 M NaH2PO4
0.0055 M glucose
Store up to 2 weeks at 4°C
Baenziger B buffer
0.015 M Tris⋅Cl, pH 7.6 (APPENDIX 2A)
0.14 M NaCl
0.005 M glucose
Heparin-Sepharose column buffers, 0.15, 0.25, 0.55, and 2.0 M
0.002 M CaCl2
0.15, 0.25, 0.55, or 2.0 M NaCl
The molarity of the buffer refers to the concentration of the NaCl.
Sucrose gradient solutions, 10% and 20% (w/v)
0.14 M NaCl
0.002 M CaCl2
10% or 20% (w/v) sucrose
Store up to 1 week at 4°C
TBS (Tris-buffered saline)
0.015 M NaCl
COMMENTARY
Purification and
Analysis of
Thrombospondin-1
The thrombospondins are a family of extracellular matrix proteins currently consisting
of five members, thrombospondins 1 to 4 and
cartilage oligomeric matrix protein (COMP).
For comprehensive reviews, see Adams (2001)
and Chen et al. (2000). These proteins are
synthesized by many tissues with patterns of
expression that are temporally and spatially
regulated. All thrombospondin family members are composed of a series of multidomain
structures and have the ability to bind large
numbers of calcium ions. Calcium binds to the
thrombospondins through a cooperative
mechanism that involves a significant conformational change in the protein. Through interactions with molecules on the cell surface and
components of the extracellular matrix, the
10.10.6
Supplement 17
thrombospondins play a role in cell adhesion,
migration, differentiation, and proliferation.
Thrombospondin-1 (TSP-1) was the first
member of the gene family to be identified and
has been the most extensively characterized.
TSP-1 is a large multifunctional glycoprotein
with a molecular weight of 420,000 Da, and is
a trimer composed of identical subunits each
with a molecular weight of 142,000 Da. TSP-1
is expressed by both normal and tumor cells
and has a number of domains that allow it to
interact with cells and other proteins. These
include (1) a heparin-binding domain that interacts with proteoglycans, integrin α3β1, and
cell-surface glycosaminoglycans (Clezardin et
al., 1997; Merle et al., 1997); (2) three type 1
repeats that interact with CD36, matrix metalloproteinases, fibronectin, and heparan sulfate
proteoglycans, and also activate latent TGF-β
(Bornstein, 1995; Schultz-Cherry et al., 1995;
Crawford et al., 1998); (3) an RGDA sequence
within the last type 3 repeat, which interacts
with integrin αvβ3; and (4) a C-terminal cellbinding domain that contains a recognition sequence for the integrin-associated protein
CD47 (Gao et al., 1996). In this unit, the authors
focus on the activities of TSP-1 that involve the
type 1 repeats and the interaction of TSP-1 with
integrins (Fig. 10.10.1). The interaction of
TSP-1 with proteoglycans is discussed in detail
in a recent review by Chen et al. (2000).
NH2
procollagen
FQGVLQNVRFVF
type 1
TSP-1 and transforming growth factor-â
Recently, TSP-1 has been shown to activate
transforming growth factor-β (TGF-β) by binding to the latency-associated protein and altering the conformation of TGF-β to make it
accessible to its receptor (Schultz-Cherry et al.,
1995; Crawford et al., 1998). The region of
TSP-1 responsible for TGF-β activation is the
amino acid sequence KRFK, which is found at
the start of the second type I repeat (SchultzCherry et al., 1995; Crawford et al., 1998; Fig.
10.10.1). TGF-β is a 25-kDa homodimeric cytokine and a known tumor suppressor (Markowitz and Roberts, 1996). It is secreted in a
latent complex consisting of mature TGF-β, the
latency-associated protein, and sometimes an
additional latent TGF-β-binding protein. The
latent TGF-β-binding protein is thought to target latent TGF-β to sites in the extracellular
matrix where it is sequestered until activated.
Activation of TGF-β has been demonstrated in
vitro by activators such as acids, plasmin, or
cathepsin D (Munger et al., 1997). TSP-1 and
the αvβ6 integrin have been shown to activate
TGF-β in vivo (Crawford et al., 1998; Munger
et al., 1999). Activation of TGF-β by TSP-1 was
demonstrated in vivo when TSP-1-deficient
mice were injected with a peptide containing
the sequence KRFK. The lungs of the injected
mice became morphologically more similar to
wild-type mice and active TGF-β was detected
in the bronchial epithelial cells (Crawford et al.,
1998). In some contexts, however, TSP-1 does
not appear to be a good activator of TGF-β
type 2
KRFK
CD36 CD36
α3β1
type 3
COOH
RFYVVMWK
RGD
αvβ3 CD47
ss
heparinbinding
domain
25,000 Da
70,000 Da proteolytic fragment
Figure 10.10.1 Representative model of TSP-1 identifying the different structural and functional
domains. The binding sites for the various integrins, CD36, and CD47 are indicated below the model.
Amino acid sequences that mediate receptor binding and activation of TGF-β are indicated above
the model. The proteolytic fragments isolated in the protocol are shown at the bottom. The
FQGVLQNVRFVF sequence is a GAG-independent cell binding site and the RFYVMWK sequence
is an integrin-associated protein (CD47) binding site.
Extracellular
Matrix
10.10.7
Supplement 17
(Abdelouahed et al., 2000; Grainger and Frow,
2000). These data indicate that post-translational modification or other factors may regulate the ability of TSP-1 to activate TGF-β.
Thus, co-expression of TSP-1 and TGF-β does
not necessarily mean that TSP-1 will activate
latent TGF-β in that tissue.
Purification and
Analysis of
Thrombospondin-1
The role of TSP-1 in angiogenesis and
cancer
TSP-1 has been shown to be an effective
inhibitor of angiogenesis, tumor progression,
and metastasis (Chen et al., 2000; Lawler,
2002). While TSP-1 levels are very low in many
tumor cells, expression of TSP-1 is high in the
tumor stroma (Brown et al., 1999). Overexpression of TSP-1 in MDA-MB-435 human breast
carcinoma cells decreased tumorigenesis and
metastasis in vivo (Weinstat-Saslow et al.,
1994). Furthermore, the tumors derived from
cells formed by a fusion of low-TSP-1-expressing human breast cancer cells and high-TSP-1expressing normal breast epithelial cells were
smaller in nude mice as compared to the tumors
formed from the breast cancer cells alone (Zajchowski et al., 1990). Lastly, one group has
shown that plasma TSP-1 secreted from primary HT1080 fibrosarcomas in nude mice inhibited growth of experimental metastases
(Volpert et al., 1998). Moreover, if the implanted fibrosarcoma cells were transfected
with an antisense TSP-1 construct prior to implantation, melanoma cell invasion of the lung
was not inhibited.
Recently, the authors have shown that recombinant proteins comprising the second type
1 repeat of TSP-1 and containing the TGF-β
activating sequence KRFK inhibited B16F10
tumor growth in mice (Miao et al., 2001). Furthermore, it was observed that treatment with
a TGF-β antibody or soluble TGF-β receptor
reversed this inhibition, suggesting that TSP-1
activation of TGF-β is part of the inhibitory
pathway. By contrast, an effect of TGF-β was
not observed with Lewis lung carcinoma because these cells have acquired mutations that
have rendered them unresponsive. Vascular
density was decreased in both B16F10 and
Lewis lung carcinoma tumors treated with the
recombinant proteins through a TGF-β-independent mechanism.
In another study, Streit et al. (1999) overexpressed full-length TSP-1 in A431 human carcinoma cells and implanted these cells in the
flanks of nude mice. Decreased tumor growth
and angiogenesis were observed in tumors expressing TSP-1. Recent work has demonstrated
that the KRFK sequence in the second type 1
repeat of TSP-1 is partly responsible for this
growth inhibition and the decrease in tumor
angiogenesis (K. Yee, unpub. observ.). In another recent study, TSP-1 null mice were
crossed with c-neu transgenic mice to create a
mouse that develops breast tumors and does not
express TSP-1. These mice developed tumors
that were larger and more vascular than the
tumors of mice overexpressing TSP-1 (Rodrídguez-Manzaneque et al., 2001). The
authors also determined that the absence of
TSP-1 in these tumors resulted in an increase
in the amount of active matrix metalloproteinase 9 (MMP-9).
The effects of TSP-1 on endothelial cell
migration and angiogenesis have been previously observed by several groups (Tolsma et
al., 1993; Dawson et al., 1997; Qian et al., 1997;
Iruela-Arispe et al., 1999; Jiménez et al., 2000;
Nör et al., 2000). These studies demonstrate
that TSP-1 is able to prevent tumor progression
in several in vivo cancer models and that one
of the ways TSP-1 inhibits tumor growth may
be through decreasing tumor angiogenesis. In
a different avenue of thinking, many groups
have examined MMP-2 and MMP-9 with regards to breast cancer progression (Benaud et
al., 1998; Martorana et al., 1998; Remacle et
al., 1998; Rudolph-Owen et al., 1998; Lee et
al., 2001). MMP-2 and -9 are gelatinases that
degrade collagen types IV, V, VII, and X, as well
as denatured collagen and gelatin (Dollery et
al., 1995). Recently, TSP-1 has been shown to
interact with MMP-2 and -9 and inhibit their
activation (Bein and Simons, 2000; Rodrídguez-Manzaneque et al., 2001). This interaction is mediated by the type 1 repeats of
TSP-1. Therefore, one of the mechanisms
through which TSP-1 inhibits both tumor progression and tumor angiogenesis may be due
to its ability to inhibit MMP activation and
prevent growth factor and cell mobilization.
Angiogenesis is a complex process that involves multiple cell types. TSP-1 does have
possible effects on the recruitment of immune
cells and on the proliferation and migration of
vascular smooth muscle cells. In some assays,
these effects can predominate, leading to the
conclusion that TSP-1 supports angiogenesis.
The preponderance of in vivo data indicates that
the anti-angiogenic effects predominate in tumors.
TSP-1 and CD36
CD36 is an integral membrane glycoprotein,
a member of the class B scavenger receptor
10.10.8
Supplement 17
family, and is located within the caveolae of the
cell membrane. It is expressed in many cells
including microvascular endothelium, adipocytes, skeletal muscle, dendritic cells, and hematopoietic cells including platelets and
macrophages (Febbraio et al., 2001). CD36 is
also a receptor for TSP-1 and binds to the
specific sequence CSVTCG in the second and
third type 1 repeats of TSP-1, while TSP-1 type
1 repeats bind the CD36 LIMP-II Emp sequence homology (CLESH) region of CD36
(Crombie and Silverstein, 1998). This binding
initiates a signal that involves the nonreceptor
tyrosine kinases fyn, lyn, and yes as well as
p38MAPK (Huang et al., 1991). One of the
endpoints of this cascade is activation of
caspase 3 and endothelial cell apoptosis (Guo
et al., 1997; Jiménez et al., 2000; Nör et al.,
2000). CD36 signaling is one of the mechanisms by which TSP-1 inhibits angiogenesis
and tumor progression (Dawson et al., 1997;
Simantov et al., 2001).
The initial work on exploring the anti-angiogenic effect of TSP-1 through CD36 utilized
peptides containing the CSVTCG sequence.
These peptides inhibited endothelial cell migration and angiogenesis (Iruela-Arispe et al.,
1991; Tolsma et al., 1993; Dawson et al., 1999).
Antibodies to CD36 also inhibited endothelial
cell migration (Dawson et al., 1997) and, in
CD36-null mice, TSP-1 did not inhibit angiogenesis in a cornea pocket assay (Jiménez et al.,
2000). Therefore, binding of TSP-1 to CD36
on endothelial cells inhibits angiogenesis and
tumor progression.
TSP-1 and integrins
Integrins are a family of cell surface receptors composed of both an α and a β subunit
(Hynes, 1992). TSP-1, in both soluble and matrix-bound forms, can interact with β1 and β3
integrins; however, the physiological consequences of binding are dependent upon the
integrin engaged, the cell type, and in some
cases the involvement of accessory proteins.
TSP-1 and â1 integrins
In breast carcinoma cells, α3β1 is essential
for chemotaxis towards TSP-1 and cell spreading on an immobilized TSP-1 matrix (Chandrasekaran et al., 1999). This interaction is
mediated through binding of the integrin to
residues 190 to 201 of the N-terminal region of
TSP-1 (Krutzsch et al., 1999). In the presence
of a β1-activating antibody, the adhesive properties of the carcinoma cells on TSP-1 are
enhanced. This is characterized by rearrange-
ment of F actin filaments into filopodia that
terminate at points that are rich in β1 and are
in contact with TSP-1. Signaling through the
insulin-like growth factor-I receptor (IGF-IR)
can also potentiate this adhesion. Recent evidence suggests that IGF-IR signaling activates
α3β1 by promoting association with the mitochondrial molecule heat shock protein 60
(Barazi et al., 2002).
Small-cell lung carcinoma cells also bind
residues 190 to 201 of TSP-1 through α3β1
(Guo et al., 2000). This interaction stimulates
the cells to extend neurite-like processes and
differentiate along a neuronal pathway. When
epidermal growth factor is added to these cultures, binding to TSP-1 through this receptor
also suppresses cell proliferation. This mechanism may be important for the antitumorigenic
effects of TSP-1.
In response to loss of cell-cell contact, endothelial cells engage immobilized TSP-1
through α3β1 and are stimulated to adhere to
TSP-1 and proliferate (Chandrasekaran et al.,
2000). This effect can be induced through disruption of cell contacts through wounding or
by inhibiting vascular endothelial (VE) cadherin, indicating a role for TSP-1 in supporting
repair of wounded endothelium. However, classically, TSP-1 is known for inhibiting endothelial cell proliferation and angiogenesis (Good
et al., 1990). Indeed, endothelial cells exposed
to a soluble TSP-1 peptide that recognizes
α3β1 have decreased proliferation and motility
(Chandrasekaran et al., 2000). These opposing
effects on endothelial cells suggest that tight
regulation of TSP-1/α3β1 interaction and signaling exists. Recent studies using melanoma
cells demonstrated that the ability of TSP-1 to
bind α3β1 is altered when TSP-1 is bound to
fibronectin (Rodrigues et al., 2001). Conformational regulation of TSP-1 may represent one
mechanism by which integrin-mediated cellular responses are controlled.
Activated T-lymphocytes can adhere to intact TSP-1 through α4β1 and α5β1 integrins
(Yabkowitz et al., 1993). This may have implications for mediating T cell activation, as
stimulation of the ERK pathway by TSP-1 in
these cells can be inhibited using anti-β1 function-blocking antibodies (Wilson et al., 1999).
A role for TSP-1 in modulating the inflammatory response would not be surprising since
TSP-1-deficient mice suffer from inflammatory disease (Lawler et al., 1998).
Extracellular
Matrix
10.10.9
Supplement 17
Purification and
Analysis of
Thrombospondin-1
TSP-1 and â3 integrins
In platelets, it was originally discovered that
αvβ3 and, to a lesser extent, αIIbβ3 (GPIIbIIIa)
function as adhesion receptors for TSP-1. The
recognition site for these integrins is the RGD
motif located in the type 3 repeats of TSP-1.
TSP-1 can influence integrin function directly and indirectly through its interaction with
nonintegrin receptors. In platelets, binding of
the C terminus of TSP-1 to the transmembrane
receptor integrin-associated protein (IAP or
CD47) leads to assembly of a TSP-1/IAPαIIbβ3 complex on the platelet surface. This complex can further activate αIIbβ3 and cause
phosphorylation of focal adhesion kinase, resulting in both augmentation of platelet aggregation and attachment to fibrinogen (Chung et
al., 1997). A necessity for G-protein signaling
has since been added to this cascade of events
(Frazier et al., 1999).
TSP-1/IAPαvβ3 complexes are also important in other cell types. On vitronectin substrates, C32 human melanoma cells are stimulated to spread in response to complex formation (Gao et al., 1996). More recently, an
increase in latent TGF-β activation, induced by
tamoxifen treatment of breast carcinoma cells,
has been shown to be dependent on localization
of TSP-1 to the cell surface by this mechanism
(Harpel et al., 2001).
Another example of TSP-1 affecting integrin function through cooperation with other
receptors occurs in the clearance of apoptotic
neutrophils. Here, TSP-1 associates with CD36
on the macrophage surface and αvβ3 associates
on the neutrophils where it forms a bridge,
allowing the recognition of neutrophils for ingestion (Savill et al., 1992). This process can
be modulated on a second exposure of macrophages to neutrophils by ligation of αvβ3,
α6β1, and α1β2 (Erwig et al., 1999).
αvβ3 is also expressed on endothelial cells.
In sickle cell anemia patients, both αvβ3 in the
endothelium (Solovey et al., 1999) and TSP-1
plasma levels are elevated. These proteins have
been implicated in recurring vaso-occlusion
problems in sickle cell patients caused by exaggerated adhesion of the sickle cell red blood
cells (SS-RBCs) to the endothelium. Indeed, it
has been demonstrated that TSP-1 enhances
adhesion of SS-RBCs to cultured endothelial
cells and that antibodies to αvβ3 can block this
event (Kaul et al., 2000). It is as yet unknown
if this is a direct consequence of TSP-1/αvβ3
association.
Critical Parameters
The response of the platelets to thrombin is
a critical factor contributing to the success of
the purification procedure. Since platelets become less responsive during storage, the platelet-rich plasma should be processed as soon as
possible after collection. Since platelets are
temperature sensitive, buffers and centrifuges
used in the purification procedure should be
warmed to room temperature before beginning
the procedure. The platelets should also be
handled gently during the resuspension steps to
prevent mechanical activation. Moreover, since
platelets are activated by untreated glass surfaces, all transfer pipets and tubes should be
plastic.
TSP-1 is susceptible to proteolysis following its secretion into the supernatant. It is important to work quickly following the activation
step to minimize exposure to proteases secreted
from the platelets and the thrombin used for the
activation. The supernatant should be treated
immediately with DFP following the debrisclearing centrifugation step in order to inactivate these proteases. The supernatant should be
kept on ice at all times during the remaining
purification steps.
The association of TSP-1 with calcium
maintains the confirmation of the molecule. It
is therefore essential that calcium be present in
all solutions during and subsequent to thrombin
treatment. A concentration of 2 mM is recommended.
Troubleshooting
The problem most likely to be encountered
in the purification procedure is unresponsive
platelets. To remedy this situation the procedure can be performed on a small scale using
fresh platelets. This will provide a sense of how
the aggregated platelets should appear following thrombin treatment. Another method for
assaying platelet responsiveness is to perform
electrophoresis on the supernatant from the
thrombin-treated platelets. TSP-1 is a major
component of the platelet α-granule and should
appear as a prominent band running at an apparent molecular weight of 185,000 Da on
discontinuous Laemmli SDS gels (UNIT 6.1).
This anomalously high value for the molecular
weight of the subunit is probably due to a
decrease in the amount of SDS bound to the
large number of negatively charged residues in
the type 3 repeats.
10.10.10
Supplement 17
Anticipated Results
The purification procedure should result in
producing ∼200 µg TSP-1 per 100 ml outdated
platelet-rich plasma, which is most often >95%
pure as determined by SDS-PAGE. There is
evidence that some preparations of TSP-1 produced according to this method may contain
trace amounts of active TGF-β bound to the
TSP-1. It is possible to remove this contaminant
by adjusting the pH of the sucrose gradient
solutions to pH 11, as TGF-β will dissociate
from TSP-1 under alkaline conditions (Murphy-Ullrich et al., 1992; Schultz-Cherry et al.,
1994). The pH of the TSP-1-containing fractions should be returned to pH 7.6 immediately
following centrifugation.
Whereas the protocol for purifying TSP-1
proteolytic fragments does not require many
steps and is reasonably efficient, it is important
to bear in mind that the N-terminal domain only
represents ∼18% of the total mass of the protein.
Thus, if one starts with 5 mg total protein, a
yield of 400 to 500 µg is appropriate. Since the
210,000-Da fragment represents about onehalf of the protein, yields of 1 to 1.5 mg can be
expected.
Time Considerations
The purification procedure is extended over
a period of 3 days. The amount of time required
to perform this procedure will depend in part
on the amount of material to be processed.
Approximately 3 to 4 hr should be allowed to
isolate the TSP-1-containing supernatant (steps
1 to 15). Purification of TSP-1 (steps 17 to 28)
will require another 3 to 4 hr. It is possible to
leave the TSP-1 bound to the heparinSepharose column for a number of days prior
to continuing the elution process.
The purification of proteolytic fragments
also takes ∼3 days. The limited tryptic digestion
is done overnight. Elution of the heparinSepharose column can be done in ∼1 day and
the elution of the G-200 column requires another day.
LITERATURE CITED
Abdelouahed, M., Ludlow, A., Brunner, G. and
Lawler, J. 2000. Activation of platelet-transforming growth factor β-1 in the absence of
thrombospondin-1. J. Biol. Chem. 275:1793317936.
Adams, J. 2001. Thrombospondins: Multifunctional
regulators of cell interactions. Annu. Rev. Cell
Dev. Biol. 17:25-51.
Barazi, H.O., Zhou, L., Smyth Templeton, N.,
Krutzsch, H.C., and Roberts, D.D. 2002. Identification of heat shock protein 60 as a molecular
mediator of α3β1 integrin activation. Cancer
Res. 62:1541-1548.
Bein, K. and Simons, M. 2000. Thrombospondin-1
type 1 repeats interact with matrix metalloproteinase 2: Regulation of metalloproteinase activity.
J. Biol. Chem. 275:32167-73.
Benaud, C., Dickson, R.B. and Thompson, E.W.
1998. Roles of the matrix metalloproteinases in
mammary gland development and cancer. Breast
Cancer Res. Treatment 50:97-116.
Bornstein, P. 1995. Diversity of function is inherent
in matricellular proteins: An appraisal of thrombospondin 1. J. Cell Biol. 130:503-506.
Brown, L.F., Guidi, A.J., Schnitt, S.J., Water, L.V.D.,
Iruela-Arispe, M.L., Yeo, T.-K., Tognazzi, K.,
and Dvorak, H.F. 1999. Vascular stroma formation in carcinoma in situ, invasive carcinoma and
metastatic carcinoma of the breast. Clin. Cancer
Res. 5:1041-1056.
Chandrasekaran, S., Guo, N.-H., Rodrigues, R.G., Kaiser, J., and Roberts, D.D. 1999. Pro-adhesive and
chemotactic activities of thrombospondin-1 for
breast carcinoma cells are mediated by a3b1 integrin and regulated by insulin-like growth factor 1
and CD98. J. Biol. Chem. 274:11408-11416.
Chandrasekaran, L., He, C.H., Al-Barazi, H.,
Krutzsch, H.C., Iruela-Arispe, M.L., and
Roberts, D.D. 2000. Cell-contact-dependent activation of α3β1 integin modulates endothelial
cell responses to thrombospondin-1. Mol. Biol.
Cell 11:2885-2900.
Chen, H., Herndon, M.E., and Lawler, J. 2000. The
cell biology of thrombospondin-1. Matrix Biol.
19:597-614.
Chung, J., Gao, A., and Frazier, W.A. 1997. Thrombospondin acts via integrin associated protein to
activate the platelet integrin αIIbβ3. J. Biol.
Chem. 272:14740-14746.
Clezardin, P., Lawler, J., Amiral, J., Quentin, G., and
Delmas, P. 1997. Identification of cell adhesive
active sites in the N-terminal domain of thrombospondin-1. Biochem. J. 321:819-827.
Crawford, S.E., Stellmach, V., Murphy-Ullrich, J.E.,
Ribeiro, S.M.F., Lawler, J., Hynes, R.O., Boivin,
G.P. and Bouck, N. 1998. Thrombospondin-1 is
a major activator of TGF-β1 in vivo. Cell
93:1159-1170.
Crombie, R. and Silverstein, R. 1998. Lysomsomal
integral membrane protein II binds thrombospondin-1. J. Biol. Chem. 273:4855-4863.
Dawson, DW., Pearce, S.F.A., Zhong, R., Silverstein, R.L., Frazier, W.A., and Bouck, N.P. 1997.
CD36 mediates the in vitro inhibitory effects of
thrombospondin-1 on endothelial cells. J. Cell
Biol. 138: 707-717.
Dawson, D.W., Volpert, O.V., Pearce, S.F.A.,
Schneider, A.J., Silverstein, R.L., Henkin, J., and
Bouck, N. 1999. Three distinct d-amino acid
substitutions confer potent antiangiogenic activity on an inactive peptide derived from a thrombospondin-1 type 1 repeat. Molec. Pharmacol.
55:332-338.
Extracellular
Matrix
10.10.11
Supplement 17
Dollery, C.M., McEwan, J.R., and Henney, A.M.
1995. Matrix metalloproteinases and cardiovascular disease. Circ. Res. 77:863-868.
Erwig, L.P., Gordon, S., Walsh, G.M., and Rees, A.J.
1999. Previous uptake of apoptotic neutrophils
or ligation of integrin receptors downmodulates
the ability of macrophages to ingest apoptotic
neutrophils. Blood 93:1406-1412.
Febbraio, M., Hajjar, D.P., and Silverstein, R.L.
2001. CD36: A class B scavenger receptor involved in angiogenesis, atherosclerosis, inflammation, and lipid metabolism. J. Clin. Invest.
108:785-791.
Frazier, W.A., Gao, A., Dimitry, J., Chung, J.,
Brown, E.J., Lindberg, F.P., and Linder, M.E.
1999. The thrombospondin receptor integrin-associated protein (CD47) functionally couples to
heterotrimeric Gi. J. Biol. Chem. 274:85548560.
Gao, A.-G., Lindberg, F.P., Dimitry, J.M., Brown,
E.J., and Frazier, W.A. 1996. Thrombospondin
modulates αvβ3 function through integrin-associated protein. J. Cell Biol. 135:533-544.
Iruela-Arispe, M.L., Lombardo, B., Krutzsch, H.C.,
Lawler, J., and Roberts, D.D. 1999. Inhibition of
angiogenesis by thrombospondin-1 is mediated
by 2 independent regions within the type 1 repeats. Circulation 100:1423-1431.
Jiménez, B., Volpert, O.V., Crawford, S.E., Febbraio, M., Silverstein, R.L., and Bouck, N. 2000.
Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin1. Nature Med. 6:41-48.
Kaul, D.K., Tsai, H.M., Liu, X.D., Nakada, M.T.,
Nagel, R.L., and Coller, B.S. 2000. Monoclonal
antibodies to αvβ3 (7E3 and LM609) inhibit
sickle red blood cell-endothelium interactions
induced by platelet-activating factor. Blood
95:368-374.
Krutzsch, H.C., Choe, B.J., Sipes, J.M., Guo, N.-H.,
and Roberts, D.D. 1999. Identification of an
α3β1 integrin recognition sequence in thrombospondin-1. J. Biol. Chem. 274:24080-24086.
Lawler, J. 2002. Thrombospondin-1 as an endogenous inhibitor of angiogenesis and tumor
growth. J. Cell. Mol. Med. 6:1-12.
Good, D.J., Polverini, P.J., Rastinejad, F., Le Beau,
M.M., Lemons, R.S., Frazier, W.A., and Bouck,
N. 1990. A tumor suppressor-dependent inhibitor of angiogenesis is immunologically and functionally indistinguishable from a fragment of
thrombospondin. Proc. Natl. Acad. Sci. U.S.A.
87:6624-6628.
Lawler, J. and Hynes, R.O. 1986. The structure of
human thrombospondin, an adhesive glycoprotein with multiple calcium-binding sites and homologies with several different proteins. J. Cell
Biol. 103:1635-1648.
Grainger, D.J. and Frow, E.K. 2000. Thrombospondin-1 does not activate transforming growth factor β1 in a chemically defined system or in
smooth muscle cell cultures. Biochem J.
350:291-298.
Lawler, J., Sunday, M., Thibert, V., Duquette, M.,
George, E.L., Rayburn, H., and Hynes, R.O.
1998. Thrombospondin-1 is required for normal
pulmonary homeostasis and its absence causes
pneumonia. J. Clin. Invest. 101:982-992.
Guo, N.-H., Krutzsch, H.C., Inman, J.K., and
Roberts, D.D. 1997. Thrombospondin-1 and
type 1 repeat peptides of thrombospondin-1 specifically induce apoptosis of endothelial cells.
Cancer Res. 57:1735-1742.
Lee, J., Weber, M., Mejia, S., Bone, E., Watson, P.,
and Orr, W. 2001. A matrix metalloproteinase
inhibitor, batimastat, retards the development of
osteolytic bone metastases by MDA-MB-231
human breast cancer cells in BalbC nu/nu mice.
Eur. J. Cancer 37:106-113.
Guo, N.-H., Smyth Templeton, N., Al-Barazi, H.,
Cashel, J., Sipes, J.M., Krutzsch, H.C., and
Roberts, D.D. 2000. Thrombospondin-1 promotes α3β1 integrin-mediated adhesion and
neurite-like outgrowth and inhibits proliferation
of small cell lung carcinoma cells. Cancer Res.
60:457-466.
Harpel, J.G., Shultz-Cherry, S., Murphy-Ullrich,
J.E., and Rifkin, D.B. 2001. Tamoxifen and estrogen effects on TGF-βformation: Role of
thrombospondin-1, αvβ3, and integrin-associated protein. Biochem. Biophys. Res. Comm.
284:11-14.
Huang, M.-M., Bolen, J.B., Barnwell, J.W., Shattil,
S., and Brugge, J.S. 1991. Membrane glycoprotein IV (CD36) is physically associated with the
Fyn, Lyn and Yes protein-tyrosine kinases in
human platelets. Proc. Natl. Acad. Sci. U.S.A.
88:7844-7848.
Hynes, R.O. 1992. Integrins: Versatility, modulation
and signaling in cell adhesion. Cell 69:11-25.
Purification and
Analysis of
Thrombospondin-1
on cord formation by endothelial cells in vitro.
Proc. Natl. Acad. Sci. U.S.A. 88:5026-5030.
Iruela-Arispe, L., Bornstein, P., and Sage, H. 1991.
Thrombospondin exerts an antiangiogenic effect
Markowitz, S.D. and Roberts, A.B. 1996. Tumor
suppressor activity of the TGF-βpathway in human cancers. Cytokine Growth Factor Rev. 7:93102.
Martorana, A.M., Zheng, G., Crowe, T.C., O’Grady,
R.L., and Lyons, J.G. 1998. Epithelial cells upregulate matrix metalloproteinases in cells
within the same mammary carcinoma that have
undergone an epithelial-mesenchymal transition. Cancer Res. 58:4970-4979.
Merle, B., Malaval, L., Lawler, J., Delmas, P., and
Clezardin, P. 1997. Decorin inhibits cell attachment to thrombospondin-1 by binding to a
KKTR-dependent cell adhesive site present
within the N-terminal domain of thrombospondin-1. J. Cell. Biochem. 67:75-83.
Miao, W.-M., Seng, W.L., Duquette, M., Lawler, P.,
Laus, C., and Lawler, J. 2001. Thrombospondin1 type 1 repeat recombinant proteins inhibit tumor growth through transforming growth factor
β dependent and independent mechanisms. Cancer Res. 61:7830-7839.
10.10.12
Supplement 17
Munger, J.S., Harpel, J.G., Gleizes, P.-E., Mazzieri,
R., Nunes, I., and Rifkin, D.B. 1997. Latent
transforming growth factor-β: Structural features and mechanisms of activation. Kidney Int.
51:1376-1382.
Munger, J.S., Huang, X., Kawakatsu, H., Griffiths,
M.J.D., Dalton, S.L., Wu, J., Pittet, J.-F., Kaminski, N., Garat, C., Matthay, M.A. et al. 1999. The
integrin αvβ6 binds and activates latent TGFβ-1:
A mechanism for regulating pulmonary inflammation and fibrosis. Cell 96:319-328.
Murphy-Ullrich, J.E., Schultz-Cherry, S., and Höök,
M. 1992. Transforming growth factor-βcomplexes with thrombospondin. Mol. Biol. Cell
3:181-188.
Nielsen, B.S., Sehested, M., Kjeldsen, L., Borregaard, N., Rygaard, J., and Danø, K. 1997. Expression of matrix metalloprotease-9 in vascular
pericytes in human breast cancer. Lab. Invest.
77:345-355.
Nör, J.E., Mitra, R.S., Sutorik, M.M., Mooney, D.J.,
Castle, V.P., and Polverini, P.J. 2000. Thrombospondin-1 induces endothelial cell apoptosis
and inhibits angiogenesis by activating the
caspase death pathway. J. Vasc. Res. 37:209-218.
Qian, X., Wang, T.N., Rothman, V. L., Nicosia, R.F.,
and Tuszynski, G.P. 1997. Thrombospondin-1
modulates angiogenesis in vitroby up-regulation
of matrix metalloproteinase-9 in endothelial
cells. Exper. Cell Res. 235:403-412.
Remacle, A.G., Noël, A., Duggan, C., McDermott,
E., O’Higgins, N., Foidart, J.M., and Duffy, M.J.
1998. Assay of matrix metalloproteinases types
1, 2, 3 and 9 in breast cancer. Br. J. Cancer
77:926-931.
Rodrigues, R.G., Guo, N.-H., Zhou, L., Sipes, J.M.,
Williams, S.B., Smyth Templeton, N., Gralnick,
H.R., and Roberts, D.D. 2001. Conformational
regulation of the fibronectin binding and α3β1
integrin-mediated adhesive activities of thrombospondin-1. J. Biol. Chem. 276:27913-27922.
Rodrídguez-Manzaneque, J.C., Lane, T.F., Ortega,
M.A., Hynes, R.O., Lawler, J., and IruelaArispe, M.L. 2001. Thrombospondin-1 suppresses spontaneous tumor growth and inhibits
activation of matrix metalloproteinase-9 and
moblization of vascular endothelial growth factor. Proc. Natl. Acad. Sci. U.S.A. 98:1248512490.
Rudolph-Owen, L.A., Chan, R., Muller, W.J., and
Matrisian, L.M. 1998. The matrix metalloproteinase matrilysin influences early-stage mammary tumorigenesis. Cancer Res. 58:5500-5506.
Savill, J., Hogg, N., Ren, Y., and Haslett, C. 1992.
Thrombospondin cooperates with CD36 and the
vitronectin receptor in macrophage recognition
of neutrophils undergoing apoptosis. J. Clin. Invest. 90:1513-1522.
Schultz-Cherry, S., Ribeiro, S., Gentry, L., and Murphy-Ullrich, J.E. 1994. Thrombospondin binds
and activates the small and large forms of latent
transforming growth factor-βin a chemically defined system. J. Biol. Chem. 269:26775-26782.
Schultz-Cherry, S., Chen, H., Mosher, D.F., Misenheimer, T.M., Krutzsch, H.C., Roberts, D.D., and
Murphy-Ullrich, J.E. 1995. Regulation of transforming growth factor-βactivation by discrete
sequences of thrombospondin-1. J. Biol. Chem.
270:7304-7310.
Simantov, R., Febbraio, M., Crombie, R., Asch,
A.S., Nachman, R.L., and Silverstein, R.L. 2001.
Histidine-rich glycoprotein inhibits the antiangiogenic effect of thrombospondin-1. J. Clin.
Invest. 107:45-52.
Solovey, A., Gui, L., Ramakrishnan, S., and Hebbel,
R.P. 1999. Sickle cell anemia as a possible state
of enhanced anti-apoptotic tone: Survival effect
of vascular endothelial growth factor on circulation and unanchored endothelial cells. Blood
93:3824-3830.
Streit, M., Velasco, P., Brown, L.F., Skobe, M.,
Richard, L., Riccardi, L., Lawler, J., and Detmar,
M. 1999. Overexpression of thrombospondin-1
decreases angiogenesis and inhibits the growth
of human cutaneous squamous cell carcinomas.
Am. J. Pathol. 155:441-452.
Tolsma, S.S., Volpert, O.V., Good, D.J., Frazier,
W.A., Polverini, P.J., and Bouck, N. 1993. Peptides derived from two separate domains of the
matrix protein thrombospondin-1 have antiangiogenic activity. J. Cell Biol. 122:497-511.
Volpert, O.V., Lawler, J., and Bouck, N.P. 1998. A
human fibrosarcoma inhibits systemic angiogenesis and the growth of experimental metastases via thrombospondin-1. Proc. Natl. Acad. Sci.
U.S.A. 95:6343-6348.
Weinstat-Saslow, D.L., Zabrenetzky, V.S., VanHoutte, K., Frazier, W.A., Roberts, D.D., and
Steeg, P.S. 1994. Transfection of thrombospondin 1 complementary DNA into a human breast
carcinoma cell line reduces primary tumor
growth, metastatic potential, and angiogenesis.
Cancer Res. 54:6504-6511.
Wilson, K.E., Li, Z., Kara, M., Gardner, K.L., and
Roberts, D.D. 1999. β1 integrin- and proteoglycanmediated stimulation of T lymphoma cell adhesion
and mitogen-activated protein kinase signaling by
thrombospondin-1 and thrombospondin-1 peptides. J. Immunol. 163:3621-3628.
Yabkowitz, R., Dixit, V.M., Guo, N., Roberts, D.D.,
and Shimizu, Y. 1993. Activated T-cell adhesion
to thrombospondin is mediated by the α4β1
(VLA-4) and α5β1 (VLA-5) integrins. J. Immunol. 151:149-158.
Zajchowski, D.A., Band, V., Trask, D.K., Kling, D.,
Connolly, J.L., and Sager, R. 1990. Suppression
of tumor-forming ability and related traits in
MCF-7 human breast cancer cells by fusion with
immortal mammary epithelial cells. Proc. Natl.
Acad. Sci. U.S.A. 87:2314-2318.
Contributed by Karen O Yee, Mark Duquette,
Anna Ludlow, and Jack Lawler
Beth Israel Deaconess Medical Center
Boston, Massachusetts
Extracellular
Matrix
10.10.13
Supplement 17
Purification of SPARC/Osteonectin
UNIT 10.11
SPARC (secreted protein acidic and rich in cysteine) is a founding member of the
matricellular group of proteins that have been shown to mediate interactions between cells
and the extracellular matrix (ECM; Bornstein and Sage, 2002). Other proteins within this
family include thrombospondins 1 and 2, osteopontin, tenascins C and X, and Cyr61.
Over the last several years, a wealth of data, largely from mice with targeted disruptions
of the respective genes, has emerged identifying various targets of the matricellular
proteins that influence cell behavior—e.g., growth factors, cell-cycle regulatory proteins,
ECM components, adhesion proteins and/or their receptors, cell survival, collagen
fibrillogenesis, and immune cell function. In vivo, these effects can be translated into
abnormalities in blood vessel morphogenesis and connective tissues, wound healing, bone
formation, and responses to various types of injury. Therefore, study of one or more of
the matricellular proteins affords insight from a somewhat unusual and underexplored
perspective: the interface between the cell surface and the extracellular milieu.
SPARC belongs to a family of several genes, only one other of which, SC1/hevin, has been
characterized beyond a limited degree (Brekken and Sage, 2000). SPARC-null mice
exhibit many phenotypic abnormalities that follow logically from the effects of SPARC
on cultured cells (i.e., de-adhesion, antiproliferation, interaction with growth factors and
ECM, and regulation of collagen production). These characteristics include (1) accelerated dermal wound healing and fibrovascular invasion of sponge implants, (2) reduced
foreign body response, (3) thin skin with decreased collagen, which is deposited as
small-diameter fibrils, (4) excessive accumulation of adipose tissue, (5) osteopenia, and
(6) cataract formation (Bornstein and Sage, 2002). Providing a mechanistic explanation
for any one of these phenotypes requires experiments, largely in vitro, with active purified
protein in clearly defined assays with quantitative endpoints. This unit presents several
protocols for the purification of SPARC (see Basic Protocol and Alternate Protocols 1, 2,
and 3), and for the measurement of its biological activity and conformation (see Support
Protocols 1 and 2). Since the end product—i.e., natural SPARC or recombinant
(rSPARC)—differs according to the source, guidelines for the choice of each protocol,
and its advantages and limitations, have been included with the Basic Protocol (purification of SPARC from cultured cells), Alternate Protocol 1 (rSPARC from E. coli), Alternate
Protocol 2 (rSPARC from insect cells), and Alternate Protocol 3 (SPARC from blood
platelets). A method for determining endotoxin levels is presented in Support Protocol 3.
NOTE: To prevent denaturation of SPARC due to adsorption to surfaces, only
polypropylene or siliconized glass should be used.
NOTE: All solutions and equiptment coming into contact with live cells should be sterile
and a septic technique should be used accordingly
PURIFICATION OF SPARC FROM PYS-2 CELLS
This protocol describes the purification of SPARC from cultured PYS-2 cells. This cell
line, originally derived from a mouse parietal yolk sac carcinoma, has been a consistent
reproducible source of biologically active SPARC for nearly two decades (Sage and
Bornstein, 1995). The following procedure can be applied to most cell culture supernatants and involves essentially three steps: (1) precipitation of culture medium, (2)
ion-exchange chromatography, and (3) molecular-sieve chromatography. Advantages of
the PYS-2 cell line are its immortality, its high rate of growth, its copious production
(secretion) of SPARC, and the presence of few other secreted products in the culture
medium. It is also possible to radiolabel SPARC metabolically if desired. A commercial
Contributed by E. Helene Sage
BASIC
PROTOCOL
Data Processing
and Analysis
10.11.1
Supplement 17
source of SPARC, isolated according to this protocol and of ∼80% purity, is available
from Sigma-Aldrich.
Materials
50% to 70% confluent PYS-2 cells (see recipe)
DMEM (serum-free; APPENDIX 2A)
1100 Ci/mmol (12.5 Ci/ml) [trans-35S]methionine/cysteine (ICN; optional)
DMEM minus methionine and cysteine (optional)
0.2 M PMSF stock solution (see recipe)
N-Ethylmaleimide (NEM)
Ammonium sulfate, ultrapure
DEAE buffer, 4°C (see recipe)
NaCl
∼2 × 20–cm DEAE column (see recipe)
S-200 buffer (see recipe)
Scintillation fluid (optional)
Sephacryl molecular-sieve column (see recipe)
0.05 M acetic acid
Plastic pipets
50-ml polycarbonate high-speed centrifuge tubes
Low-speed GPKR (Beckman) centrifuge with swinging bucket rotor
High-speed refrigerated centrifuge with GSA (Sorvall) or JA-17 rotors (Beckman)
or equivalent
12,000- to 14,000-MWCO dialysis tubing (Spectrapor) or equivalent, prewashed
with DEAE buffer
Dialysis clips (optional)
Standard gradient maker (e.g., Amersham Biosciences)
Peristaltic pump
Fraction collector
Lyophilizer
50 or 250 ml centrifuge tubes
Additional reagents and equipment for SDS-PAGE (UNIT 6.1) with autoradiography
(UNIT 6.3), if appropriate, and determination of protein concentration by
spectroscopy (APPENDIX 3B)
CAUTION: When working with radioactivity, take appropriate precautions to avoid
contamination of the experimenter and surroundings. Carry out the experiments and
dispose of wastes in appropriately designated area, following guidelines provided by the
local radiation safety officer (also see APPENDIX 1D).
Collect and precipitate tissue culture medium containing secreted SPARC
1. Replace medium in 20 to 30 dishes or flasks of PYS-2 cells (grown to 50% to 70%
confluency) with 12 to 13 ml serum-free DMEM and preincubate 15 min at 37°C.
Replace with fresh medium and then incubate 18 to 24 hr.
If desired, purification can be monitored by adding 500 ìCi of 1100 Ci/mmol [35S]methionine to one dish and processing the medium in parallel with nonlabeled medium from the
other dishes. Alternatively, if radiolabeled SPARC of high specific activity is required for
experimental purposes, [35S]methionine/cysteine can be added to all dishes. When using
label, incubate cells in serum-free DMEM lacking methionine and cysteine.
Purification of
SPARC/Osteonectin
2. Collect the medium from the cell layer by gentle aspiration via plastic pipet and
transfer to centrifuge tubes. Remove cellular debris by centrifuging in a clinical (i.e.,
10.11.2
Supplement 17
tissue-culture) centrifuge 5 min at 1,000 × g, room temperature, or in GPKR
centrifuge at 1000 × g, 4°C.
3. Pool all supernatants in a siliconized flask. Add 0.2 M PMSF drop-wise with stirring
to a final concentration of 0.2 mM, and NEM to a final concentration of 10 mM. Stir
on ice until medium reaches 4°C.
For 100 ml medium, add 0.1 ml PMSF stock solution and 62.5 mg NEM. Take care not to
lyse cells in any of these procedures.
4. Add solid ultrapure ammonium sulfate to the medium in an amount equivalent to 50%
(w/v) of the starting volume over a period of several hours. Stir 12 to 24 hr at 4°C.
For 100 ml medium, add 50 g ammonium sulfate, in very small increments (e.g., 1 to 2 g)
over several hours (e.g., 3 to 5). This detail is important for maintenance of neutral pH and
for efficient precipitation of protein, which consists mainly of laminin 1, type IV collagen,
bovine serum albumin (BSA), and SPARC.
Do not allow the solution to foam by stirring too rapidly, as this indicates the proteins are
denaturing.
5. Transfer medium to 50-ml polycarbonate high-speed centrifuge tubes and centrifuge
in a high-speed refrigerated centrifuge with JA-17 rotor 30 min at 40,000 × g, 4°C.
Discard the supernatant. Keep tubes containing pellets on ice or store up to 1 to 2
months at –70°C.
6. Thaw, if necessary, and dissolve each pellet by gentle vortexing in 2 to 5 ml DEAE
buffer, 4°C. Pool these solutions and transfer to 12,000- to 14,000-MWCO dialysis
tubing, prewashed with DEAE buffer and closed on one end. Rinse each centrifuge
tube with 1 ml buffer and add this solution to the bag.
7. Close the open end of the dialysis bag with double knots or dialysis clips, leaving 1
to 2 in. (2.5 to 5 cm) extra space to allow for change in volume. Immerse the bag
(containing ∼40 ml) in a 500-ml graduated cylinder containing 500 ml DEAE buffer,
4°C. Dialyze with stirring overnight (or 4 to 6 hr), and change the dialysis buffer
twice (2 to 3 hr each) for an additional 4 to 6 hr dialysis.
Wear gloves when handling dialysis tubing to minimize exposure to radioactivity as well
as to protect the sample from contamination. Mix bag contents several times by inversion.
8. Remove dialysis tubing, cut tip off carefully (if knotted) or remove clips, and empty
contents into one or two 50-ml centrifuge tubes. Clarify the solution by centrifuging
in a JA-17 rotor 20 min at 10,000 × g, 4°C. If appropriate, retain 10 to 25 µl for
scintillation counting and for SDS-PAGE (UNIT 6.1) with autoradiography (UNIT 6.3),
as assessment of starting material.
The sample is now ready for ion-exchange chromatography.
Chromatograph on DEAE cellulose
9. Prepare gradient buffer B by adding 2.336 g NaCl to 200 ml DEAE buffer (200 mM
NaCl final). Fill the front chamber of a standard gradient maker (containing a stir bar
or paddle) with 200 ml DEAE buffer (gradient buffer A) and the second chamber
with 200 ml gradient buffer B.
Ensure that the narrow opening between the two chambers is filled with gradient buffer A
before adding gradient buffer B. An air block will inhibit flow of B into A.
10. Use a peristaltic pump to add the entire sample onto an ∼2 × 20–cm DEAE column,
and follow with one to two column volumes DEAE buffer. Discard this eluate, which
contains unbound protein.
11. If phenol red (from DMEM) is seen to bind to the resin, wash the column until it is
no longer visible, or until the A280 of the flowthrough is at baseline.
Data Processing
and Analysis
10.11.3
Supplement 17
Phenol red will interfere with the monitoring of the column effluent at 280 nm.
12. Connect the gradient maker to the peristaltic pump for delivery to the column bed.
Connect a fraction collector to the column and set to collect 3-ml fractions of eluate
in polypropylene or siliconized glass tubes. Elute bound proteins with a linear
gradient of 0% to 100% buffer B over ~300 ml.
All chromatographic procedures must be carried out at 4°C.
A less complicated alternative to the continuous gradient is the use of two stepwise elutions,
the first consisting of 100 ml of 75 mM NaCl in DEAE buffer, followed by 100 ml of 175
mM NaCl in DEAE buffer. SPARC will elute in the second buffer.
13. For radiolabeled SPARC (step 2), monitor the effluent by scintillation counting 20µl aliquots from alternate fractions suspended in 3 ml scintillation fluid. For nonradiolabeled SPARC, monitor alternate fractions by absorbance at 280 nm.
SPARC is eluted at 150 to 175 mM NaCl. See Sage et al. (1989) for an example of the
elution profile. If the location of the peak containing SPARC is in doubt, individual fractions
can be analyzed by SDS-PAGE (UNIT 6.1).
14. Pool fractions containing SPARC, and dialyze the pooled sample (∼20 ml) against
four changes of 4 liters (each) water over 24 to 48 hr, 4°C (see steps 6 to 8).
After 24 to 48 hr, a precipitate containing SPARC, together with laminin and traces of BSA,
should appear in the dialysis bag. Depending on the concentration of protein and/or the
water used (pH 5.5 is optimal), precipitation may fail to occur. In this case, lyophilize the
protein (step 16b), redissolve in DEAE buffer at 25% of the original volume, and repeat
dialysis and precipitation (steps 14 and 15).
If the column will be reused, it should be regenerated as described (see Reagents and
Solutions).
15. Decant the entire contents of the bag into a centrifuge tube and centrifuge 30 min at
48,000 × g, 4°C. Discard the supernatant.
16a. For immediate use: Dissolve pellet in 2 ml S-200 buffer, clarify by microcentrifugation for 1 min at top speed or 10,000 × g, and proceed to molecular-sieve chromatography (step 18).
16b. For storage before chromatography: Resuspend pellet in 2 to 4 ml water, shell-freeze
by twirling the tube in dry ice/ethanol to effect freezing of the solution on the sides
of the vessel, and then lyophilize. Store up to 1 to 2 months at –70°C. Before use,
resuspend in 1 to 2 ml S-200 buffer, stir 4 to 6 hr at 4°C, and clarify the solution by
microcentrifugation at top speed for 1 min.
Shell-freezing increases the efficiency of lyophilization and improves solubility of the
protein after storage.
Pellets from several preparations can be pooled prior to molecular-sieve chromatography.
Purify SPARC by molecular-sieve chromatography
17. Remove buffer from the top of a Sephacryl molecular-sieve column and apply the
sample gently onto the resin. Allow the sample (optimally 1 to 2 ml) to flow into the
bed. Add 2 to 4 ml S-200 buffer to the top of the column, reconnect the buffer reservoir,
and allow effluent to flow by gravity at 8 to 10 ml/hr (0.17 ml/min) by adjustment of
the pressure head (i.e., the reservoir containing S-200 buffer above the column).
It is important not to disturb the column bed during sample loading, as the precision of
elution can be affected.
Purification of
SPARC/Osteonectin
10.11.4
Supplement 17
In some cases it may be necessary to use a peristaltic pump, pulling buffer from the bottom
of the column, at ∼10 ml/hr. If the flow rate is too high, the column will pack too tightly
and will cease to flow.
18. Collect 80 fractions of 1 to 1.5 ml each and monitor effluent by absorbance at 280
nm and/or by counting 10 to 25-µl aliquots in 3 ml scintillation fluid.
The exact position of elution of SPARC will vary with chromatographic parameters (e.g.,
column size, sample size, flow rate). It is therefore advisable to monitor the column effluent
and, if necessary, to check 10 to 25 ìl of each fraction by SDS-PAGE (see below). The initial
peak (at Vo) contains laminin, whereas the leading shoulder of the peak corresponding to
the elution position of SPARC contains most of the BSA.
19. Pool peak fractions corresponding to SPARC (approximately ten fractions, corresponding to 55 to 65 ml total column effluent). Dialyze this pool against four changes
of 4 liters of 0.05 M acetic acid each, 4°C, and lyophilize.
Alternatively, the sample can be stored at –70°C in S-200 buffer without dialysis or
lyophilization, or it can be dialyzed directly into another buffer as desired.
20. Determine the concentration of SPARC by absorbance at 280 nm, using the extinction
coefficient (ε) 0.838 mg ml−1 cm−1 (APPENDIX 3B).
21. Analyze the purified protein by SDS-PAGE (UNIT 6.1) with autoradiography (UNIT 6.3).
When heating samples at 95°C, use reducing (i.e., 50 mM DTT) and nonreducing
conditions.
For detection using Coomassie blue, from 1 to 5 ìg SPARC is recommended; for detection
by autoradiography, ∼104 cpm is recommended. A single broad band, or occasionally a
doublet, should be obtained with an apparent Mr of 39,000 (with DDT) or 43,000 (without
DDT), the latter co-migrating with an ovalbumin molecular weight standard. The yield of
purified SPARC is ∼500 ìg per 30 maxiplates (150-mm diameter) of PYS-2 cells (2 to 3 ×
108 cells).
PURIFICATION OF rSPARC FROM E. COLI
The preceding procedure (see Basic Protocol) allows for the purification of murine
SPARC from cultured (tumor) cells. Limitations of a mammalian cell culture system as
a protein source are its cost, potential contamination of the product by serum and cellular
proteins/proteinases, and the low yield of product. To circumvent these problems, Bassuk
et al. (1996a) expressed human rSPARC with a C-terminal histidine tag in E. coli. A
soluble (monomeric) form and an insoluble (aggregated) form of SPARC were recovered,
the latter sequestered in inclusion bodies within the host. Soluble (monomeric) SPARC
from E. coli is biologically active and can be purified in relatively large quantities with
minimal contamination by endotoxin or bacterial proteins. Isolation of the soluble form
is accomplished by anion-exchange, nickel-chelate affinity, and gel-filtration chromatographies. Anion-exchange chromatography on DEAE-Sepharose is used as an initial
isolation step. Metal-chelate affinity chromatography provides an efficient purification
of rSPARC that has been expressed with a (His)6 sequence. Gel-filtration chromatrography separates monomers of SPARC from dimers, trimers, and higher oligomers. This
procedure is outlined below. It assumes that a competent strain of E. coli—e.g.,
BL21(DE3)—has been transformed with a SPARC expression plasmid—e.g., pSPARC
wt (human)—with a hexahistidine (His)6 sequence at the 3′ end (Bassuk et al., 1996a)
and has been propagated and frozen as a glycerol stock.
Additionally, the aggregated form can be unfolded by urea treatment, purified by nickelchelate affinity chromatography, and renatured by gradual removal of the denaturant.
After disulfide bond isomerization, the disaggregated monomers are further purified by
ALTERNATE
PROTOCOL 1
Data Processing
and Analysis
10.11.5
Supplement 17
high-resolution gel-filtration chromatography (Bassuk et al., 1996b). As the disaggregation/renaturation procedure is complicated and time consuming, the reader is referred to
Bassuk et al. (1996b) for this additional protocol.
Additional Materials (also see Basic Protocol)
LB medium with appropriate selective reagents (APPENDIX 2A)
E. coli strain transfected with SPARC expression vector (Bassuk et al., 1996a)
Inducing agent (e.g., IPTG; APPENDIX 3A)
10 mM sodium phosphate, pH 7.0 (APPENDIX 2A)/10% (v/v) glycerol
90 mM sodium phosphate buffer (pH 7.8)/10% (v/v) glycerol/0.2 mM AEBSF,
4°C (see recipe), with and without 0.5 M NaCl
DEAE-Sepharose Fast Flow anion-exchange resin (Amersham Biosciences):
equilibrate in 90 mM sodium phosphate buffer (pH 7.8)/10% (v/v) glycerol/0.2
mM AEBSF and allow to settle
5 M NaCl (APPENDIX 2A)
0.2 M AEBSF stock solution (see recipe)
Nickel/nitrilotriacetic acid (Ni-NTA) metal-chelate affinity resin (Qiagen)
50 mM sodium phosphate (pH 5.3, 6.0, and 7.8)/0.5 M NaCl/10% (v/v) glycerol
(see recipe)
1.6 × 60–cm Superdex 70 column (see recipe)
50 mM Tris⋅Cl (pH 8.0)/0.15 M NaCl (see recipe)
1× PBS (APPENDIX 2A) containing 1 to 4 mM Ca2+ (optional)
French press
2 × 20– and 1 × 10–cm chromatography columns
Flow cell coupled to a UV monitor set at 280 nm
Chart recorder
Conductivity meter (optional)
Disposable 10-ml gel-filtration column, sterile (optional)
Additional reagents and equipment for transfecting SPARC expression vector
(APPENDIX 3A) and for SDS-PAGE on minigels (UNIT 6.1)
Extract E. coli
1. Inoculate 1.3 liters LB medium containing appropriate selective reagents with a
suitable E. coli strain transfected with SPARC expression vector using standard
techniques (APPENDIX 3A). Grow to midexponential phase (OD600 ∼0.5) and induce
with the appropriate agent.
Induction of rSPARC in midexponential phase cells is necessary for high levels of expression. The procedure and chemical(s) used depend on the E. coli strain and the vector into
which SPARC cDNA is cloned. For example, IPTG was used at a final concentration of 1
mM for SPARC cloned into pET22b vector and transfected into strain BL21(DE3) (Bassuk
et al., 1996a).
2. After the cells have been induced, grow an additional 1 to 4 hr.
3. Recover the cells by centrifuging 20 min at 7000 × g, room temperature. Discard the
supernatant and resuspend the pellet in 20 ml of 10 mM sodium phosphate, pH 7.0,
containing 10% (v/v) glycerol. Disrupt by performing two cycles in a French press
at 20,000 psi.
Cells can alternatively be broken open by sonication on ice.
Purification of
SPARC/Osteonectin
4. Separate soluble from insoluble material by centrifuging 30 min at 10,000 × g, 4°C.
Decant soluble extract (supernatant) into a separate tube.
10.11.6
Supplement 17
Soluble extracts and insoluble pellets at this stage can be stored up to 1 month at −80°C.
Refer to Bassuk et al. (1996b) for details on processing pellets for aggregated SPARC.
Perform initial chromatography on DEAE-Sepharose
5. If necessary, thaw the soluble extracts on ice. Dilute to 100 ml with ice-cold 90 mM
sodium phosphate buffer (pH 7.8)/10% (v/v) glycerol/0.2 mM AEBSF, 4°C.
6. Add 50 ml settled DEAE-Sepharose Fast Flow anion-exchange resin. Stir gently 12
to 18 hr at 4°C.
7. Pour slurry into a 2 × 20–cm chromatography column, allow to settle, and wash with
∼250 ml of 90 mM sodium phosphate buffer (pH 7.8)/10% (v/v) glycerol/0.2 mM
AEBSF until the absorbance at 280 nm is <0.01.
8. Assemble a linear gradient (see Basic Protocol 1, step 9) by adding 250 ml of 90 mM
sodium phosphate buffer pH 7.8)/10% (v/v) glycerol/0.2 mM AEBSF to the front
compartment of a gradient maker, and 250 ml of 90 mM sodium phosphate buffer
(pH 7.8)/10% (v/v) glycerol/0.2 mM AEBSF with 0.5 M NaCl (7.3 g) to the other
compartment.
9. At 4°C pump gradient onto the column at 3 ml/min. Collect 8-ml fractions and
monitor the column eluate with a flow cell coupled to a UV monitor set at 280 nm
and a chart recorder set at full scale equal to 1 OD unit.
It is advisable also to monitor the eluate by conductivity; read every fourth fraction in a
conductivity meter (clean probe after each reading). rSPARC elutes at a concentration of
0.10 to 0.25 M NaCl (conductivity of 14 to 20 mmho).
10. Analyze 50-µl aliquots of fractions by SDS-PAGE on minigels (see Basic Protocol
and UNIT 6.1). Pool fractions containing rSPARC, and adjust solution to 0.5 M NaCl
by adding 5 M NaCl.
In the absence of post-translational modification, rSPARC migrates on an SDS-polyacrylamide gel with an apparent Mr of 34,000 to 38,000 Da after reduction. Adjustment of ionic
strength can be monitored easily by conductivity measurement; the final conductivity of
the pooled fractions containing SPARC should be equivalent to that of 50 mM sodium
phosphate (pH 7.8)/0.5 M NaCl/10% (v/v) glycerol. At this point, fractions can be stored
up to 1 month at –80°C.
Perform metal-chelate affinity chromatography
11. Add 0.2 M AEBSF stock solution to the pooled sample to a final concentration of
0.2 mM.
12. Mix sample with a slurry of Ni-NTA metal-chelate affinity resin, using 3 to 5 ml resin
per liter original bacterial culture. Adjust pH to 7.8 with 1 N NaOH or 1 N HCl, and
stir gently for 1 hr at 4°C.
13. Pour slurry into a chromatography column (e.g., 1 × 10 cm), allow to settle, and wash
with ∼60 ml of 50 mM sodium phosphate (pH 7.8)/0.5 M NaCl/10% (v/v) glycerol
at a flow rate of 0.5 ml/min until the absorbance at 280 nm is <0.01.
14. Pass 15 column volumes of 50 mM sodium phosphate (pH 6.0)/0.5 M NaCl/10%
(v/v) glycerol through the column to remove nonspecifically bound proteins (i.e.,
until A280 <0.01).
15. Elute rSPARC from the column with 20 ml of 50 mM sodium phosphate (pH 5.3)/0.5
M NaCl/10% (v/v) glycerol.
16. Store up to 1 month in ∼1-ml aliquots at –80°C.
Data Processing
and Analysis
10.11.7
Supplement 17
Perform gel-filtration chromatography
17. Apply ∼1 ml rSPARC solution onto a 1.6 × 60–cm Superdex 70 column and elute by
gravity using 50 mM Tris⋅Cl (pH 8.0)/0.15 M NaCl buffer at a flow rate of 0.1 ml/min.
Collect fractions of 1.25 ml each, monitored at 280 nm.
Over an elution range of 100 ml, rSPARC monomer elutes between 46 and 50 ml. This
should be verified by SDS-PAGE, as should the removal of oligomers (compare migration
with and without 50 mM DTT).
18. Optional: Perform buffer exchange as needed for experimental design using sterile
disposable 10-ml gel-filtration columns. Elute in the buffer of choice (e.g., 1× PBS
containing 1 to 4 mM Ca2+ for in vitro studies).
19. Store samples up to 3 months at –80°C.
ALTERNATE
PROTOCOL 2
PURIFICATION OF rSPARC FROM INSECT (Sf9) CELLS
This protocol describes the purification of human rSPARC produced in a baculovirus
expression system using insect (Sf9) cells (Bradshaw et al., 2000). Advantages of this
system over those described above (see Basic Protocol and Alternate Protocol 1) are
higher yield of rSPARC, production of protein in a nonbacterial system to minimize
contamination by endotoxin (E. coli) or serum proteins (mammalian cells), and the
potential for post-translational modifications of protein similar to those in mammalian
cells.
This protocol assumes that the starting materials are human (or other species) SPARC
cDNA (minus the signal sequence) subcloned into a baculovirus expression vector
(Pharmingen), Spodoptera frugiperda 9 (Sf9) insect cells cotransfected with the SPARC
expression vector and linearized baculovirus, and high-titer stocks of recombinant virus
generated for subsequent infections of Sf9 cells grown in suspension in serum-free media
(Invitrogen). Information on the latest versions of the Sf9/baculovirus expression system
is readily available from the Pharmingen instruction manual, Baculovirus Expression
Vector System, and the Invitrogen manual, Growth and Maintenance of Insect Cell Lines
for Expression of Recombinant Proteins using the Baculovirus Expression System.
Materials
Sf9 cells (Invitrogen) infected with baculoviral SPARC expression vector, grown
in serum-free Sf-900 II medium (Invitrogen)
200 mM MOPS, pH 6.5 (see recipe)
10 N NaOH (APPENDIX 2A)
Q-Sepharose Fast Flow column (see recipe)
200 and 400 mM LiCl/20 mM MOPS, pH 6.5 (see recipe)
0.1 N acetic acid, 4°C: 0.6 ml glacial acetic acid in 100 ml H2O
Hanks’ buffered saline solution (see recipe)
50-ml conical tube
0.22-µm filter bottle
AktaPrime automated liquid chromatography system (Amersham Biosciences) or
equivalent conventional model
10,000-NMWL Ultrafree-15 (Millipore) or Centricon Plus-80 (Amicon)
centrifugal filter device
0.22-µm sterile syringe-driven filter
Purification of
SPARC/Osteonectin
10.11.8
Supplement 17
Prepare Sf9 conditioned medium for chromatography
1. Transfer Sf9 cells infected at 2 to 4 × 105 cells/ml with baculoviral SPARC expression
vector, grown 4 to 5 days in serum-free Sf-900 II medium, to 50-ml conical tubes.
Centrifuge 45 min at 6000 × g, 4°C. Transfer the supernatant to a 0.22-µm filter bottle
and discard the cell pellets.
This system optimizes for the efficient secretion of recombinant protein. It is important to
avoid lysis of the cells (and contamination of the medium) during this step.
2. Sterile-filter the supernatant and measure the volume. Add 1⁄10 vol 200 mM MOPS,
pH 6.5, and adjust the pH to 6.5 with 6 N NaOH.
Purify rSPARC by anion-exchange chromatography
3. Pump the sample onto a Q-Sepharose Fast Flow column at a flow rate of 5 ml/min.
4. Using either an AktaPrime automated liquid chromatography system or equivalent
conventional model, assemble a continuous linear salt gradient from 200 to 400 mM.
Use 200 ml of 200 mM LiCl/20 mM MOPS, pH 6.5, supplied to the buffer valve
(AktaPrime System) or front chamber (conventional gradient maker) and 200 ml of
400 mM LiCl/20 mM MOPS, pH 6.5, to the buffer switch valve or rear chamber.
5. Start the gradient pumping at a rate of 5 ml/min and collect 3.5-ml fractions over 300
ml, monitoring the column effluent at 280 nm.
The fractions can also be checked by SDS-PAGE (see Basic Protocol and Alternate Protocol
1; UNIT 6.1). Human rSPARC produced by Sf9 cells migrates at ∼38,000 to 40,000 Da after
reduction, and the doublet shifts to a single band of ∼36,000 Da in the absence of reducing
agent. The doublet is the result of heterogeneous glycosylation (Bradshaw et al., 2000).
Dialyze sample
For dialysis using acetic acid
6a. Pool the fractions containing rSPARC in 12,000- to 14,000-MWCO dialysis tubing
and dialyze against three 1- to 2-liter changes (each) of 0.1 N acetic acid, 4°C.
7a. Aliquot the samples according to use. Snap-freeze on dry ice or in liquid nitrogen,
lyophilize, and store at –70°C.
Acetic acid is used when the sample is to be lyophilized and concentrated.
The above procedure results in rSPARC of ∼80% purity by SDS-PAGE. An additional
purification step (entailing molecular-sieve chromatography) can be performed after the
sample has been lyophilized. Follow the procedure described above (see Basic Protocol, steps
17 to 21). Significant losses of rSPARC (Sf9) should be expected with this procedure, however.
For dialysis using saline solution
6b. Pool the fractions containing rSPARC in 12,000- to 14,000-MWCO dialysis tubing
and dialyze against three 4-liter changes (each) of 1× Hanks buffered saline solution
(HBSS) containing 1 µM CaCl2. Concentrate in a 10,000-NMWL Ultrafree-15 or
Centricon Plus-80 centrifugal filter device to 1 or 2 ml. Filter sterilize using 0.22-µm
sterile syringe-driven filter.
Saline solution is used when the sample is to be used for cell culture.
7b. Aliquot according to use. Snap freeze on dry ice or in liquid nitrogen and store at –70°C.
Data Processing
and Analysis
10.11.9
Supplement 17
Analyze SPARC
8. Determine protein content by UV spectroscopy at 280 nm (APPENDIX 3B), using the
extinction coefficient (ε) 0.838 mg ml−1 cm−1. Estimate the purity of SPARC by
SDS-PAGE (5 µg/sample lane; UNIT 6.1).
The yield is 2 to 4 mg SPARC (∼80% purity) per 400 ml of initial Sf9 cell culture suspension
(at 2 to 4 × 106 cells/ml).
An updated version of the Sf9/baculovirus expression system is now available from
Invitrogen. Termed “InsectSelect,” it is a virus-free system that relies on expression of
protein from a single nonlytic, integrative plasmid transfected into Sf9 or other insect cells,
and is claimed to be optimal for secreted proteins.
ALTERNATE
PROTOCOL 3
Purification of
SPARC/Osteonectin
PURIFICATION OF SPARC/OSTEONECTIN FROM TISSUES
SPARC was originally isolated from fetal bovine mineralized bone matrix, of which it is
a major noncollagenous component, and was termed osteonectin (Termine et al., 1981).
Two other significant sources of SPARC are platelets (osteonectin; Kelm and Mann, 1991)
and the Engelbreth-Holm-Swarm (EHS) sarcoma, a murine basement membrane–producing tumor (termed BM-40; Sasaki et al., 1999, and references therein). SPARC,
osteonectin, and BM-40 are now recognized as the same protein. Many of the functional
properties of SPARC were deduced from biochemical/biophysical studies of the tissuederived protein, which can be isolated in significantly greater quantities compared to
yields typically described from in vitro sources. In this protocol, purification of SPARC
from human platelets is described, based on an original report by Kelm and Mann (1990).
There are several advantages to using platelets as a source of SPARC: (1) human blood
is a readily available source for human SPARC; (2) bovine blood is an excellent source
of SPARC and requires neither screening for pathogens nor the rigorous safety procedures
associated with the use of human material; and (3) denaturing conditions are not involved
(the extraction of osteonectin from bone matrix or EHS tumor includes the use of EDTA
and, in some cases, guanidinium⋅HCl; Termine et al., 1981; Kelm and Mann, 1990; Sasaki
et al., 1999). It is important to note, however, that differences have been reported between
bone and platelet osteonectin from the same species, notably in the specificity of collagen
binding that was attributed to differences in glycosylation (Kelm and Mann, 1991).
Investigators interested in tissue-specific modifications of SPARC and their functional
implications are encouraged to consult the references cited above. Bovine bone and human
platelet osteonectin are available commercially from Calbiochem, although their method
of purification is not specified. Haematologic Technologies sells human platelet
osteonectin isolated by affinity chromatography on an anti-osteonectin monoclonal
antibody column, as well as bovine bone osteonectin isolated from 0.5 M EDTA extracts
of demineralized bone. All commercial preparations should be tested for activity in one
or more of the assays described (see Support Protocols 1 to 3).
Materials
Platelet-rich plasma or platelet suspension, or informed, nonsmoking, aspirin-free,
consenting adult blood donors
0.156 M citrate containing 0.1 M dextrose and 5.0 µM prostaglandin E1 (Sigma;
optional)
0.02 M Tris⋅Cl, pH 7.6/0.15 and 1.0 M NaCl (see recipe)
Thrombin
Sepharose 4B–AON IgG column (see recipe)
3.0 M NaSCN/0.02 M Tris⋅Cl (pH 7.6)/0.15 M NaCl (see recipe)
0.05 M NH4HCO3
10.11.10
Supplement 17
19-G butterfly needles
50 or 250-ml plastic centrifuge bottles with caps
12,000 to 14,000-MWCO dialysis tubing
Lyophilizer
Additional reagents and equipment for thrombin activation of platelets (Kelm and
Mann, 1990) and SDS-PAGE (UNIT 6.1; also see Basic Protocol, step 22)
Prepare activated platelet supernatant
1a. For predrawn plasma: Purchase platelet-rich plasma or platelet suspensions from a
local blood bank.
1b. For in-house drawn plasma: Draw 480 ml fresh blood from informed, nonsmoking,
aspirin-free, consenting adults via 19-G butterfly needles into 0.156 M citrate containing
0.1 M dextrose and 5.0 µM prostaglandin E1 (to prevent platelet activation). Remove red
cells and leukocytes by centrifuging 30 min at 1000 × g, room temperature.
CAUTION: Appropriate biosafety practices must be followed when working with human
blood or blood products. Human blood must be screened for HIV and other infectious
viruses. In addition, safety glasses, a double layer of gloves, and protective laboratory
clothing should be worn at all times. Use double containment (e.g., place a tube or bag
containing blood in a beaker prior to any manipulation) and ensure that all containers
including centrifuge bottles are tightly capped.
2. Centrifuge platelet-rich suspension 30 min at 27,000 × g, 4°C, to pellet the platelets.
Decant supernatant.
3. Wash platelets with 200 ml of 0.02 M Tris⋅Cl (pH 7.6)/0.15 M NaCl, centrifuge as in
step 2. Resuspend platelets in 50 ml of the same buffer. Count with a hemacytometer
and suspend 4.5 × 1010 cells in 50 ml of 0.02 M Tris⋅Cl (pH 7.6)/0.15 M NaCl. Activate
by adding 2.5 U/ml thrombin as described by Kelm and Mann (1990).
4. Transfer to plastic, capped centrifuge bottles and isolate activated platelets by
centrifuging 30 min at 25,000 × g, room temperature. Discard the pellet using
appropriate containment and retain the supernatant.
CAUTION: Autoclave human products prior to disposal.
Isolate SPARC/osteonectin by immunoaffinity chromatography
5. Apply platelet supernatant to a Sepharose 4B-AON IgG column (∼2 × 20–cm). After
the applied solution has permeated the resin, clamp off the column and allow the
sample to remain within the column bed for 16 hr at 4°C.
6. Wash the column with 0.02 M Tris⋅Cl (pH 7.6)/0.15 M NaCl until the effluent shows
an A280 of 0.01.
7. Wash the column with 0.02 M Tris⋅Cl (pH 7.6)/1.0 M NaCl until a baseline absorbance is achieved.
8. Elute SPARC/osteonectin with 3.0 M NaSCN/0.02 M Tris⋅Cl (pH 7.6)/0.15 M NaCl
and collect in a single tube or in fractions.
9. Transfer the effluent to 12,000 to 14,000-MWCO dialysis tubing, dialyze against two
changes of 2 liters of 0.05 M NH4HCO3, and then lyophilize.
Peak fractions of the eluted SPARC can also be dialyzed against 0.02 M Tris⋅Cl (pH
7.6)/0.15 M NaCl, PBS, or DMEM (minus phenol red), as dictated by the intended use of
the purified SPARC, and stored up to 1 month at –80°C. If the NH4HCO3 fails to lyophilize
completely, redissolve the powder in water, dialyze against 0.01 N acetic acid, and repeat
the lyophilization.
It is important to minimize the exposure of SPARC/osteonectin to NaSCN, and to keep all
reagents at 4°C during affinity chromatography and dialysis.
Data Processing
and Analysis
10.11.11
Supplement 17
10. Determine the concentration of SPARC/platelet osteonectin at A280 using the extinction coefficient (ε) 0.838 mg ml−1 cm−1.
11. Monitor the purity of SPARC by SDS-PAGE (UNIT 6.1; also see Basic Protocol, step 22).
Platelet SPARC/osteonectin should be >80% pure by SDS-PAGE using a Coomassie blue
stain. It exhibits an apparent Mr of ∼3000 greater than that of bone osteonectin purified
according to the same protocol (shown to be due to differences in glycosylation), but is
comparable to that reported for SPARC isolated from PYS-2 cell culture media.
Immunoaffinity chromatography typically produces somewhat low recoveries of the protein
antigen, albeit in a high state of purity given the minimal steps used in the isolation protocol.
The total amount of SPARC/osteonectin in human platelets (prior to affinity purification)
was reported by Kelm and Mann (1990) to range from 0.65 to 2.2 ìg/108 platelets.
ASSAYS FOR THE EVALUATION OF SPARC ACTIVITY
All proteins need to be evaluated, not only for their extent of purity, but also for their
activity and conformational integrity. The latter is especially critical in the case of
recombinant proteins, which are produced either in biologically “inappropriate” hosts
(e.g., SPARC in a prokaryotic system) or at levels that preclude proper processing, folding,
and/or editing. Moreover, the importance of post-translational modification to the functions of many proteins is poorly understood. In the case of SPARC, N-linked glycosylation
(one site) appears not to be critical for activity, at least in the assays that have been used;
however, rSPARC (see Alternate Protocols 1 and 2) has consistently displayed less activity
(up to 50%) than SPARC purified from PYS-2 cells (see Basic Protocol; Yost et al., 1994;
Bassuk et al., 1996a; Bradshaw et al., 2000). These protocols describe biological assays
that test two major effects of SPARC on cultured cells: de-adhesion and inhibition of
proliferation. Other assays based on biochemical measurements (e.g., circular dichroism,
binding assays) are standard procedures and are discussed elsewhere (see Commentary).
SUPPORT
PROTOCOL 1
Proliferation Assay
Endotoxin will inhibit cell proliferation, and endothelial (especially BAE) cells are
particularly sensitive. At 10 ng endotoxin/mg SPARC, there should be <10% inhibition
of [3H]thymidine incorporation in BAE cells exposed to 60 µg SPARC/ml. To determine
the effect of endotoxin on other types of cells, treat the cells with a titration of CSE and
measure proliferation.
Bovine aortic endothelial (BAE) cells
DMEM/0% and 10% (w/v) FBS (APPENDIX 2A)
Purified SPARC (see Basic Protocol or Alternate Protocol 1 to 3) and appropriate
control buffer
6.71 Ci/mmol (1 mCi/ml) [methyl-3H]thymidine (PerkinElmer)
10% (w/v) trichloroacetic acid (TCA), ice cold
95% (v/v) ethanol
0.4 N NaOH
Glacial acetic acid
Scintillation fluid
24-well tissue culture plate
15-ml conical tube
Radioactivity warning tape
Purification of
SPARC/Osteonectin
Additional reagents and equipment for trypsinizing cells (UNIT 1.1)
10.11.12
Supplement 17
CAUTION: When working with radioactivity, take appropriate precautions to avoid
contamination of the experimenter and surroundings. Carry out the experiments and
dispose of wastes in appropriately designated area, following guidelines provided by the
local radiation safety officer (also see APPENDIX 1D).
1. Starve a 100-cm dish of confluent bovine aortic endothelial (BAE) cells in serum-free
DMEM for 3 to 4 days.
2. Trypsinize cells (UNIT 1.1), resuspend in 10 ml DMEM/10% FBS, and centrifuge
briefly (i.e., 5 min at 1000 × g, room temperature) to pellet.
3. Rinse cells twice with 5 ml serum-free DMEM.
4. In a 24-well tissue culture plate, plate triplicate wells containing 5 × 104 cells in 500
µl (final 1 × 105 cells/ml) of the following solutions, using the same volume for
SPARC and buffer:
Serum-free DMEM (control)
DMEM/2%FBS containing 5 µg/ml SPARC dissolved in DMEM
DMEM/2% FBS containing 20 µg/ml SPARC dissolved in DMEM
DMEM/2% FBS containing buffer alone (control).
Other buffers compatible with cell culture may be used but not acetic acid.
5. Incubate 16 to 18 hr at 37°C.
By this time, cells will have begun to synthesize DNA (S phase).
6. Prepare label in a 15-ml conical tube by adding 20 µl (20 µCi) of 6.71 Ci/mmol
[methyl-3H]thymidine/ml DMEM. Place 55 µl (1.1 µCi) of this mixture into each
well. Swirl plate gently to mix. Label plate with radioactivity warning tape, and
incubate 4 hr at 37°C.
CAUTION: Perform this step in a laminar-flow hood with absorbent bench pad and
radioactive waste receptacle.
7. Wash each well twice with 500 µl ice-cold 10% TCA, and drain completely.
CAUTION: Collect radioactive media and washes for safe disposal.
8. Wash with 500 µl of 95% ethanol. Remove ethanol and add 500 µl of 0.4 N NaOH
per well. Incubate 30 min at room temperature with shaking.
9. Add 100 µl glacial acetic acid to neutralize the solution.
Extremes of pH can result in precipitation of scintillation cocktail and/or quenching.
10. Place the contents of each well into a collection vial containing 3 ml scintillation
fluid. Cap, mix by inversion, and measure cpm in a scintillation counter.
There should be no precipitate in the vials; check pH if this occurs and adjust to neutrality.
For rSPARC, expect >70% inhibition of [3H]thymidine incorporation at 50 ìg SPARC/ml.
For SPARC purified from PYS-2 cells, the effective dose at which 50% inhibition of
[3H]thymidine incorporation occurs (ED50) is 20 ìg SPARC/ml.
De-adhesion Assay
This protocol is presented as a rapid, inexpensive, and diagnostic assay for the de-adhesive
activity of SPARC on nontransformed cells in vitro. The activity is based on the
diminishment of focal adhesions produced by cultured cells. These structures can be
distinguished by immunofluorescence staining of vinculin in wedge-shaped structures at
the periphery of the cell, which are diagnostic for focal adhesion complexes.
SUPPORT
PROTOCOL 2
Data Processing
and Analysis
10.11.13
Supplement 17
One 100-mm dish of nearly-confluent bovine aortic endothelial (BAE) cells,
passaged not greater than ten times, grown in DMEM/10% FBS containing
appropriate antibiotics
DMEM/2% and 10% FBS (APPENDIX 2A)
Purified SPARC (see Basic Protocol and Alternate Protocols 1 to 3) and
appropriate control buffer
12-well tissue culture dishes
Phase-contrast microscope (UNIT 4.1)
Additional reagents and equipment for trypsinizing cells (UNIT 1.1)
1. Trypsinize (UNIT 1.1) a 100-mm dish containing a nearly confluent monolayer of
bovine aortic endothelial (BAE) cells, passaged greater than ten times, and grown in
DMEM/10% FBS containing appropriate antibiotics.
The size of the dishes is optional and can be adjusted according to the availability of
SPARC. Scale the volume of medium as appropriate for size of dish or well.
2. Transfer trypsinized cells to appropriate centrifuge tubes, pellet in a clinical centrifuge 5 min at 1000 × g, room temperature, and resuspend in an appropriate volume
of DMEM/2% FBS. Plate 5 to 7.5 × 104 cells in triplicate wells of a 12-well tissue
culture dish.
3. Add the following solutions to cells in triplicate, using the same volume for SPARC
and buffer:
No addition
20 µg/ml SPARC
40 µg/ml SPARC
Appropriate control buffer.
4. Mix gently and incubate 1 hr at 37°C.
5. Check the plate carefully under a phase-contrast microscope. Examine several
representative fields and count the number of cells in the following groups:
Fully spread cells (group a)
Partially spread cells (group b)
Rounded cells (group c).
Cells to which SPARC has not been added should be attached and beginning to spread.
Cells to which SPARC has been added should be less spread (i.e., rounded). If control cells
(i.e., no SPARC) have not spread, wait an additional 1 to 2 hr. There should be no toxicity
or cell death.
6. Quantify the activity of SPARC according to the rounding index (RI):
RI = [(1 × a) + (2 × b) + (3 × c)]/(a + b + c)
An RI = 1 represents a culture with only spread cells, whereas a culture with
increasing numbers of round cells would approach the maximum, RI = 3.
A titration curve can be generated using different concentrations of SPARC. Anticipate that
different types of cells will show differential sensitivity to SPARC. Cell lines (e.g., 3T3,
NRK) and transformed cells typically do not respond to SPARC.
Purification of
SPARC/Osteonectin
10.11.14
Supplement 17
Endotoxin Assay
Endotoxin is derived from gram-negative bacteria (e.g., E. coli) and is a commonly
encountered contaminant of buffers, columns, and glassware. In addition, soluble
(monomeric) SPARC purified from E. coli may contain endotoxin. Endotoxin interferes
with bioassays for SPARC, so it is necessary to assess samples for the presence of
endotoxin.
SUPPORT
PROTOCOL 3
Materials
Purified SPARC (see Basic Protocol and Alternate Protocols 1 to 3) and
appropriate buffer
Limulus Amoebocyte Lysate (LAL) Pyrochrome kit (Associates of Cape Cod) for
the Detection and Quantification of Gram-Negative Bacterial Endotoxin:
Pyrochrome LAL reagent
Pyrochrome Reconstitution buffer
Control Standard Endotoxin (CSE)
50% (v/v) glacial acetic acid
Nonpyrogenic 96-well tissue culture plate
Microtiter plate reader
1. Prepare SPARC titrated at 0.2, 1, and 5 µg/ml in DMEM or HBSS. Pipet 50 µl of
each into triplicate wells of a nonpyrogenic 96-well tissue culture plate, leaving 16
empty wells for standards and controls (step 5).
2. Tap vial containing Pyrochrome LAL reagent. Remove and discard stopper.
3. Add 3.2 ml Pyrochrome Reconstitution buffer to the LAL reagent. Mix gently but
thoroughly. Cover with Parafilm and place for 3 to 5 min on ice. Store on ice up to 3 hr.
4. Prepare standards by adding 2.0 ml water to the vial containing the Control Standard
Endotoxin (CSE) to yield 1.0 endotoxin units (EU)/ml endotoxin. Vortex and store on ice.
5. Place 200 µl water in each of five wells in a fresh 96-well plate. Make a five-step
serial dilution using a ratio of 1:1 at each step, by adding 200 µl of 1.0 EU/ml
endotoxin to the first well, mixing, transferring 200 µl to the next well, and repeating
until the series is complete. Pipet 50 µl of each dilution into triplicate wells of the
SPARC-containing plate (step 1), and include a 50-µl water-only control.
The serial dilutions above will result in final concentrations of 0.5, 0.25, 0.125, 0.0625,
and 0.0313 EU/ml endotoxin, respectively.
6. Pipet 50 µl of reconstituted Pyrochrome LAL reagent (step 3) into each well, shake
on a microtiter plate shaker for 30 sec, and incubate at 37°C for 30 min.
7. Stop reaction by adding 25 µl of 50% glacial acetic acid per well.
8. Measure OD405 in a microtiter plate reader. Determine the concentration of endotoxin
in the sample by comparison to the curve generated from the standards.
The expected concentration of endotoxin is <10 ng/mg SPARC, at an estimated level of 10
EU/ng endotoxin. Endotoxin levels range from 5 to 15 EU/ng. The level of endotoxin that
affects cells depends on the cell type.
Data Processing
and Analysis
10.11.15
Supplement 17
Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see
AEBSF (aminoethylbenzenesulfonyl fluoride) stock solution, 0.2 M
Dissolve 4.794 g AEBSF (Calbiochem) in 100 ml water. Make fresh.
DEAE buffer
500 ml 8 M urea stock solution (see recipe)
50 ml 1 M Tris⋅Cl, pH 7.5 (APPENDIX 2A)
448 ml H2O
1 ml 0.2 M PMSF stock solution (see recipe)
625 mg N-ethylmaleimide (NEM)
Adjust pH to 8.0 with 10 N NaOH
Chill to 4°C
Make fresh buffer for each column run
Final concentrations are 50 mM Tris⋅Cl, pH 8.0, 0.2 mM PMSF, 10 mM NEM, and 4 M
urea.
DEAE column
Pack an ∼2 × 20–cm column (e.g., Amersham Biosciences) at 4°C with a 20% slurry
of DE-52 cellulose (Whatman) equilibrated in DEAE buffer (see recipe). Equilibrate
with several column volumes (∼70 ml each) DEAE buffer, delivered via a peristaltic
pump connected from a reservoir to the bottom of the column (pumping upward
ensures more efficient utilization of theoretical plates for ion exchange). The column
can be stored at 4°C and reused for several months. After storage, flush the column
with several volumes of fresh DEAE buffer immediately before use. To regenerate
a DEAE column, pump one column volume of DEAE buffer containing 500 mM
NaCl (29.2 g/liter) followed by several column volumes of DEAE buffer until the
absorption and conductivity of the elution buffer is restored to baseline (see
Commentary).
Columns manufactured by Amersham Biosciences work well, as they are thick walled and
are equipped with high-quality fittings that can withstand the pressures delivered by a
peristaltic pump.
Hanks’ buffered saline solution
0.14 g/l CaCl2 (1.26 mM final)
40 g/l KCl (5.33 mM final)
0.6 g/l potassium phosphate, monobasic (0.44 mM final)
0.1 g/l magnesium chloride, hexahydrate (0.50 mM final)
0.1 g/l magnesium sulfate, heptahydrate (0.41 mM final)
0.35 g/l sodium bicarbonate (4.00 mM final)
0.048 g/l sodium phosphate, dibasic (0.30 mM final)
Store up to 3 months at 4°C
LiCl, 200 mM/20 mM MOPS, pH 6.5
200 ml of 200 mM MOPS, pH 6.5 (see recipe)
80 ml 5 M LiCl
Add H2O to 1700 ml
Add H2O to 2000 ml
Store up to 1 year at 4°C
Purification of
SPARC/Osteonectin
10.11.16
Supplement 17
LiCl, 400 mM/20 mM MOPS, pH 6.5
50 ml 200 mM MOPS, pH 6.5 (see recipe)
40 ml 5 M LiCl
Add H2O to 480 ml
Adjust pH with 6 N NaOH
Add H2O to 500 ml
LiCl, 2 M/20 mM MOPS, pH 6.5
50 ml 200 mM MOPS, pH 6.5 (see recipe)
200 ml 5 M LiCl
Add H2O to 480 ml
Adjust pH with 6 N NaOH
Add H2O to 500 ml
Store up to 1 year at 4 °C
Molecular-weight standards
Molecular-weight standards include a marker for the excluded (outer; Vo) and
included (inner; Vi) volume of the column. For Vo, use 500 µl of 0.1% (w/v) blue
dextran in S-200 buffer (see recipe). Clarify by centrifugation before applying to
the column. For Vi, use 25,000 to 100,000 cpm [35S]methionine or [3H]proline
(which can be detected by scintillation counting), or any small protein (Mr <10,000)
or peptide that is minimally hydrophobic and nonglycosylated.
MOPS, 200 mM, pH 6.5
41.86 g 3-(N-morpholino)propanesulfonic acid (MOPS)
Add H2O to ∼800 ml
Add H2O to 1000 ml
NaSCN, 3.0 M/0.02 M Tris⋅Cl (pH 7.6)/0.15 M NaCl
243.24 g sodium isothiocyanate (NaSCN)
20 ml 1 M 0.22-µm-filter-sterilized Tris⋅Cl, pH 7.5 (APPENDIX 2A)
8.766 g NaCl
Add H2O to 1000 ml
Store up to 1 year at 4 °C
PMSF (phenylmethylsulfonyl fluoride) stock solution, 0.2 M
Dissolve 3.48 g PMSF in 100 ml isopropanol. Store up to several years at 4°C or
room temperature.
Add this reagent to aqueous solutions drop-wise while vortexing, or it will precipitate.
PYS-2 cells, 50% to 70% confluent
Using 150-mm tissue-culture dishes or equivalent plastic flasks, grow PYS-2 cells
(ATCC CRL-2745) to between 50% and 70% confluence (7–10 × 106 cells per plate)
in DMEM/10% (v/v) FBS (APPENDIX 2A).
Cells undergo >1 population doubling in 24 hr under these conditions.
Data Processing
and Analysis
10.11.17
Supplement 17
Q-Sepharose Fast Flow column
Pour a 1.7 × 20–cm column of Q-Sepharose Fast Flow resin (Amersham Biosciences) equilibrated in 200 mM LiCl/20 mM MOPS, pH 6.5 (see recipe). Equilibrate by running two to three column volumes of 200 mM LiCl/20 mM MOPS, pH
6.5 (see recipe), through the resin. After use, strip the column with 100 ml of 2 M
LiCl/20 mM MOPS, pH 6.5 (see recipe), and equilibrate with 60 ml of 200 mM
LiCl/20 mM MOPS, pH 6.5. Store up to 1 year at room temperature.
S-200 buffer
999 ml Hanks’ balanced salt solution (HBSS; see recipe) with Ca2+ and Mg2+
(Life Technologies and APPENDIX 2A)
1 ml 0.2 M PMSF stock solution (see recipe)
Filter sterilize with a 0.22-µm filter
Prepare fresh for each run and keep at 4°C
Calcium is present in HBSS as 0.14 g/liter CaCl2, and magnesium is present as 0.1 g/liter
MgCl2⋅6H2O and 0.1 g/liter MgSO4⋅7H2O.
Sephacryl molecular-sieve column
At 4°C, pour an ∼1 × 100–cm column (e.g., Bio-Rad) of Sephacryl S-200 (Amersham Biosciences) in a slurry of cold, sterile S-200 buffer (see recipe). Allow column
bed to pack slowly but steadily, with controlled elution from the bottom port of the
column at ∼10 ml/hr (0.17 ml/min), to a bed height of ~95 cm. Run several column
volumes (∼80 ml each) of S-200 buffer through the packed bed and then calibrate
using molecular-weight standards (see recipe). Store in S-200 buffer at 4°C for up
to several days prior to use. For longer storage and reuse (up to several months),
store in S-200 buffer containing 0.1% (w/v) sodium azide at 4°C. Periodically clean
(i.e., remove sample debris from the top of the column up) and flush with fresh S-200
buffer containing 0.1% sodium azide.
Azide must be flushed out completely (monitor at 280 nm) prior to chromatography of
SPARC, as azide is toxic to cells and may also interfere with the properties of SPARC.
Sepharose 4B-AON IgG column
Following manufacturer’s instructions, pour anti-osteonectin (AON-5031; 20 mg)
monoclonal IgG1 antibody (Haematologic Technologies) coupled to CNBr-activated Sepharose 4B (Amersham Biosciences) into a 5- to 10-ml column, at 4°C.
Equilibrate in several column volumes of 50 mM Tris⋅Cl (pH 8.0)/0.15 M NaCl (see
recipe), by gravity flow at 0.1 to 0.5 ml/min. After use reequilibrate column in 0.02
M Tris⋅Cl, pH 7.6/0.15 M NaCl. Store up to 1 month in that same buffer at 4°C.
Sodium phosphate, 50 mM (pH 5.3, 6.0, or 7.8)/0.5 M NaCl/10% (v/v) glycerol
29.2 g NaCl
5.75 g sodium phosphate dibasic
1.37 g sodium phosphate monobasic
100 ml glycerol
Adjust pH to 5.3 or 6.0 with 6 N HCl, or to 7.8 with 6 N NaOH
Add H2O to 1000 ml
Store up to 1 to 2 days at 4°C
Purification of
SPARC/Osteonectin
10.11.18
Supplement 17
Sodium phosphate, 90 mM (pH 7.8)/10% (v/v) glycerol/0.2 mM AEBSF
10.35 g sodium phosphate dibasic
2.466 g sodium phosphate monobasic
100 ml glycerol
1 ml 0.2 M AEBSF stock solution (see recipe)
Add H2O to 1000 ml
Store up to 1 to 2 days at 4°C
Superdex 70 column
In the cold (i.e., 4°C), pour a 1.6 × 60–cm column of Superdex 70 gel-filtration resin
(Amersham Biosciences) equilibrated in 50 mM Tris⋅Cl (pH 8.0)/0.15 M NaCl (see
recipe). Calibrate using the molecular-weight standards (see recipe) blue dextran
(Vo) and [3H]proline (Vi) as described.
Tris⋅Cl, 0.02 M (pH 7.6)/0.15 and 1.0 M NaCl
20 ml 1 M 0.22-µm-filter-sterilized Tris⋅Cl, pH 7.5 (APPENDIX 2A)
8.77 or 58.44 g NaCl
Add H2O to 1000 ml
Store up to 1 month at 4°C
Tris⋅Cl, 50 mM (pH 8.0)/0.15 M NaCl
8.76 g NaCl
4.44 g Tris⋅Cl
2.65 g Tris base
Add H2O to 1000 ml
Store up to 1 month at 4°C
Urea stock solution, 8 M
Add 1920 g ultra pure urea (Life Technologies) in 2 liters water by dissolving ∼200
g at a time. After all urea has dissolved, add water to 4 liters. Filter through Whatman
no. 3 paper and store up to 1 month at 4°C.
COMMENTARY
The abundance of SPARC in many tissues,
and its high levels of secretion by most cells in
vitro, belie the difficulty of its recovery as an
intact, active protein after purification. SPARC
(as osteonectin) was found to be a major noncollagenous component of fetal and adult bone
(Termine et al., 1981). In situ hybridization of
SPARC by numerous investigators has shown
that the mRNA is abundant in most fetal tissues,
presumably associated with morphogenesis,
growth, and angiogenesis but is somewhat limited in the corresponding adult tissues (for reviews, see Lane and Sage, 1994; Brekken and
Sage, 2000; Bradshaw and Sage, 2001).
SPARC mRNA and protein are found in relatively high amounts in adult tissues that exhibit
continuous turnover (gut epithelium) and remodeling (bone), and are produced in response
to injury (wound healing) and certain types of
pathologies (tumors, scleroderma). In more
quiescent and/or established tissues, however,
levels of SPARC are low. Since SPARC affects
both the adhesion and proliferation of most
normal cells, its association with angiogenesis
and other processes requiring cell migration,
differentiation, and synthesis of extracellular
matrix (ECM) is not surprising.
Data Processing
and Analysis
10.11.19
Supplement 17
Purification of
SPARC/Osteonectin
There are several structural features of
SPARC that should be considered in the context
of a purification protocol.
1. SPARC is typically a secreted protein with
two post-translational modifications that can be
troublesome. There are fourteen cysteines, all
of which are disulfide-bonded, and the folding
and correct formation of disulfide bridges are not
trivial in recombinant proteins produced at high
levels, especially in yeast and bacteria. This has
certainly been the case for SPARC (Yost et al.,
1994; Bassuk et al., 1996b). Additionally, secreted SPARC contains a single complex-type
carbohydrate chain (N-linked) which is not produced in nonmammalian systems. Interestingly,
the carbohydrate has been shown to be variable
in mammalian SPARCs—i.e., the carbohydrate
from platelet SPARC is different from that from
bone. In addition, cultured cells can assemble and
process the oligosaccharide side-chain structures
differently (Lane and Sage, 1994). It is important
to remember that purification of SPARC from
tissues such as bone will result in the recovery of
nonsecreted SPARC that has unprocessed highmannose-type oligosaccharide.
2. SPARC binds other proteins, including
growth factors. The association of SPARC with
albumin (probably through adventitious disulfide interchange) has been troublesome, but can
be avoided by the use of serum-free culture
(e.g., Sf9 cells, E. coli, or a serum-independent
mammalian cell line). Anticipate that isolation
of SPARC from tissues (including platelets)
can result in contamination from plasma and
tissue fluid components (e.g., albumin) as well
as ECM proteins to which SPARC binds (collagen types I, III, IV, V, and thrombospondin
1). Moreover, SPARC also interacts with platelet-derived growth factor (PDGF) AB and BB
and vascular endothelial growth factor (VEGF)
with a Kd ≅ 10−9 M. If possible, it is best to avoid
these proteins when choosing a source of
SPARC, as additional purification steps to remove the contaminants will invariably result in
lower yields and loss of activity.
3. SPARC binds to several cations (Cu2+, Fe2+)
and has an absolute requirement for Ca2+. The
disulfide-bonded EF-hand, a Ca2+-binding
loop at the C terminus, is reasonably stable,
with a Kd for Ca2+ ≅ 10−7 M, and is thought to
serve a structural function. The N terminus,
however, contains from five to eight low-affinity (Kd ≅ 10−3 to 10−5 M) Ca+2-binding sites
(glutamic acids). Association of Ca2+ with this
region of SPARC serves to neutralize its excessive negative charge and confers α-helicity to
this domain. It is therefore critical that SPARC
is not exposed to EDTA or other chelating
agents during purification, and that the protein
is stored in the presence of 1 to 4 mM Ca2+. One
of the assays for native structure of SPARC,
circular dichroism (see below), depends on
α-helicity as a function of Ca2+ binding within
this low-affinity site.
Three protocols have been discussed that
maximize both the yield and the purity/native
structure of either natural or rSPARC. Most
cultured cells secrete reasonably high levels of
SPARC into the culture medium, an environment in which SPARC is stable over several
days at 37°C. Proteolytic degradation of
SPARC has rarely been a problem, especially
with the judicious use of protease inhibitors, as
described in the protocols. Since both human
and murine tumor cells can also secrete high
levels of SPARC in vitro and in general are
more tolerant of low serum (or, preferably, the
absence of serum), they are a logical choice for
the isolation of nonrecombinant SPARC, especially if they exhibit high rates of growth and
secretion (see Basic Protocol).
Advantages of a recombinant protein expression system include the (theoretically) substantially higher yields of protein, as well as the
potential of producing mutated versions of the
protein. Both the E. coli and Sf9 cell systems
can achieve these goals with respect to SPARC
(see Alternate Protocol 1 and 2). Additionally,
SPARC from any species for which the sequence is known can be engineered by the
polymerase chain reaction (APPENDIX 3F) into a
suitable expression vector. Disadvantages include potential problems with folding and posttranslational modification of rSPARC; however, assessment of purity and activity of the
SPARC produced in both E. coli and Sf9 cells
has shown that these are both viable routes for
the production of SPARC. Although the activity
of rSPARC appears to be ∼50% of that of the
PYS-2-derived protein, the substantially
greater yields may offset this limitation.
Any modification of the primary structure
of SPARC must be considered as potentially
deleterious to its conformation and/or activity.
The (His)6 sequence, tagged onto the C terminus of SPARC to facilitate its purification by
metal-affinity chromatography, could affect
one or more properties of SPARC (e.g., nuclear
translocation, de-adhesion) and should be controlled for in subsequent experiments. As discussed in preceding paragraphs, post-translational differences need to be considered as
well—i.e., the lack of carbohydrate in E. coli
rSPARC (see Alternate Protocol 1), and a dif-
10.11.20
Supplement 17
ferent or additional type of glycosylation conferred by Sf9 cells (see Alternate Protocol 2).
There may be situations in which the proper
SPARC for study will be that isolated from a
given tissue (e.g., bone). References have been
included (see Alternate Protocol 3) for the extraction of SPARC from this tissue. The use of
denaturants and EDTA could be problematic,
although renaturation is always an option.
Since both platelet and bone SPARC are available commercially (see Alternate Protocol 3),
it is advisable to purchase a small amount and
to test it according to the parameters required.
Troubleshooting
Many of the caveats at various stages of
purification of SPARC have been detailed
within each protocol. The principal problems
are low recovery and poor bioactivity.
Recovery of SPARC depends on several
factors, not the least important of which is the
output of SPARC in vitro. Despite claims of
immortality, transformed or tumor cells do not
live forever in culture. Successive passages and
cycling of cells on and off serum (or growth in
the absence of serum) can affect their eventual
viability. Therefore, it is important to monitor
the secretion of SPARC over time (this also
applies to the production of rSPARC). SPARC
is produced optimally by subconfluent cells; at
confluence or near-confluence, SPARC is secreted at a reduced rate, and will associate with
the cell surface or ECM. Presented below is a
list of other possible causes of recovery loss, as
well as potential solutions; however, the reader
should bear in mind that some losses are indeed
unavoidable.
1. Failure of SPARC to redissolve completely
in the various buffers used for purification or
assay. Clarification of solutions is always recommended.
2. Precipitation of SPARC during freezing or
thawing. Snap-freezing on dry ice, and quickthawing at room temperature, are recommended.
3. Incomplete precipitation during dialysis
against water, which can be checked by SDSPAGE (UNIT 6.1) of a small aliquot of the supernatant.
4. Irreversible binding and/or denaturation of
SPARC on membrane-type centrifugal concentrators (e.g., Centricons). Losses should be determined if the investigator chooses to concentrate purified SPARC in this manner. There are
always new products on the market that claim
to minimize this problem.
5. Degradation due to proteolysis by intrinsic
proteinases or to bacterial contamination. Protease inhibitors should always be used during
purification of SPARC, as described, and bacterial contamination should be minimal if sterile buffers or buffers containing sodium azide
(NaN3) are used.
6. Recovery can be compromised by the use of
untreated glass vessels; only polypropylene or
siliconized-glass containers should be used.
Surface denaturation of SPARC occurs readily,
either from adsorption to surfaces or from rapid
stirring or overzealous mixing.
Denaturation of SPARC can be minimized
with careful handling and attention to a few
details.
1. The protein should be stored at –70° or
–80°C, not at 4°C and especially not at −20°C.
2. 1 to 4 mM Ca2+ should be present in buffers
containing SPARC.
3. Stirring of solutions should be steady but
not rapid.
4. Purification of the protein should be conducted at 4°C whenever possible.
5. Only reagents (e.g., urea) of the highest
purity should be used.
6. Reducing/oxidizing conditions, which can
result in the scrambling of disulfide bonds,
should be avoided.
Assays for SPARC bioactivity have been
described elsewhere (see Support Protocols 1
to 3) and need not be repeated here. However,
an important criterion for the correct folding of
SPARC is the circular dichroism spectra obtained in the presence and absence of Ca2+.
These spectra are relatively easy to perform and
interpret. Examples for SPARC purified from
PYS-2 cells, E. coli, and Sf9 cells have been
published (Sage et al., 1989; Bassuk et al.,
1996a; Bradshaw et al., 2000). The method
relies on a characteristic increase of the mean
residue ellipticity (θ) at 220 nm as a function
of increasing concentrations of Ca2+, indicative
of a shift toward α-helicity. SPARC preparations that do not exhibit this transition are likely
to be contaminated by other components and/or
denatured.
For the use of SPARC in proliferation (i.e.,
[3H]thymidine incorporation) assays, it is important to measure levels (if any) of contaminating growth factors that could affect the results. Both PDGF and VEGF bind to SPARC
(see Background Information) and are anticipated to stimulate the proliferation of smooth
muscle cells, fibroblasts (PDGF), and endothelial cells (VEGF). Kits based on ELISA are now
available for the detection of these factors;
Data Processing
and Analysis
10.11.21
Supplement 17
alternatively, detection could be accomplished
by immunoblot analysis after SDS-PAGE of
SPARC under reducing conditions (UNIT 6.2),
although the former method allows for greater
sensitivity.
Anticipated Results
Isolation from PYS-2 cells (see Basic Protocol) should yield ∼500 µg per 30 maxiplates
(150-mm diameter) PYS-2 cells (∼107
cells/plate). The protein is of high purity (>90%
by SDS-PAGE) and retains maximal biological
activity. For example, an ED50 of 20 µg/ml (0.6
µM) has been defined as an effective concentration for the induction of cell rounding by
SPARC.
Yields of rSPARC from E. coli and Sf9 cells
are greater than those from PYS-2 cells (see
Alternate Protocols 1 and 2), but are in large
part dependent on the efficiency of the expression system (i.e., the particular expression vector, the host and its growth properties, and
whether the rSPARC is secreted or retained
within the cell). Using a first-generation
Sf9/baculovirus expression system, the authors’
laboratory typically recovers 2 to 4 mg human
rSPARC (of ∼80% purity) from an initial suspension of ∼109 cells. The InsectSelect system,
which eliminates the need for viral infection, is
likely to be an improvement over the earlier
version. rSPARC can be purified to ≥80% and
displays biological activity in cell rounding and
proliferation assays.
The immunoaffinity-based chromatographic purification of SPARC from platelets
will theoretically produce a highly purified protein, in reasonable yields, although the amount
of SPARC in the starting material (α-granules
of platelets) is low, from 0.7 to 2.2 µg/108 cells.
One limiting factor is the availability of the
monoclonal antibody used for the purification.
This reagent must not only bind soluble SPARC
with relatively and selectively high affinity, but
must also release SPARC readily into the elution buffer without compromise of the SPARC
or the antibody itself. Moreover, the antibody
must function while coupled to an affinity resin.
It is therefore important to ensure that a sufficient supply of the antibody is commercially
available, as the column will have to be repacked periodically with new affinity-coupled
resin. An alternative is to purchase a hybridoma
cell line secreting a suitable anti-SPARC IgG
that can be propagated in the laboratory.
Purification of
SPARC/Osteonectin
Time Considerations
The Basic Protocol and Alternate Protocols
1 to 2 each require ∼1 week from the time of
medium (PYS-2 and Sf9 cells) or cell (E. coli)
collection until the final lyophilization (or buffer exchange) step. Allow 1 to 2 days for the
preparation of buffers and columns, and for the
washing of columns. PYS-2 cells are usually
ready for beginning the collection of medium
24 hr after plating, and medium is removed
from the cells 18 to 24 hr later. Similar time
frames apply to E. coli (grown overnight, diluted to an appropriate density in log phase, and
induced) and to Sf9 cells (grown in flasks over
3 to 4 days to generate conditioned medium
containing rSPARC).
In all the protocols, convenient stopping
points have been noted. There is temporal flexibility in the purification process, especially
during the dialysis steps.
Literature Cited
Bassuk, J.A., Baneyx, F., Vernon, R.B., Funk, S.E.,
and Sage, E.H. 1996a. Expression of biologically active human SPARC in E. coli. Arch.
Biochem. Biophys. 325:8-19.
Bassuk, J.A., Braun, L.P., Motamed, K., Baneyx, F.,
and Sage, E.H. 1996b. Renaturation of secreted
protein acidic and rich in cysteine (SPARC) expressed in Escherichia coli requires isomerization of disulfide bonds for recovery of biological
activity. Intl. J. Biochem. Cell Biol. 28:10311043.
Bornstein, P. and Sage, E.H. 2002. Matricellular
proteins: Extracellular modulators of cell function. Curr. Opin. Cell Biol. 64:608-616.
Bradshaw, A.D. and Sage, E.H. 2001. SPARC, a
matricellular protein that functions in cellular
differentiation and tissue response to injury. J.
Clin. Invest. 107:1049-1054.
Bradshaw, A.D., Bassuk, J.A., Francki, A., and Sage,
E.H. 2000. Expression and purification of recombinant human SPARC produced by baculovirus. Mol. Cell Biol. Res. Comm. 3:345-351.
Brekken, R.A. and Sage, E.H. 2000. SPARC, a
matricellular protein: At the crossroads of cellmatrix communication. Matrix Biol. 19:569580.
Kelm, R.J. and Mann, K.G. 1990. Human platelet
osteonectin: Release, surface expression, and
partial characterization. Blood 75:1105-1113.
Kelm, R.J. and Mann, K.G. 1991. The collagen
binding specificity of bone and platelet
osteonectin is related to differences in glycosylation. J. Biol. Chem. 266:9632-9639.
Lane, T.F. and Sage, E.H. 1994. The biology of
SPARC, a protein that modulates cell-matrix interactions. FASEB J. 8:163-173.
Sage, E.H., Vernon, R.B., Funk, S.E., Everitt, E.A.,
and Angello, J. 1989. SPARC, a secreted protein
10.11.22
Supplement 17
associated with cellular proliferation, inhibits
cell spreading in vitro and exhibits Ca+2 dependent binding to the extracellular matrix. J. Cell.
Biol. 109:341-356.
Sage, E.H. and Bornstein, P. 1995. Matrix components produced by endothelial cells: Type VIII
collagen, SPARC, and thrombospondin. In Extracellular Matrix: A Practical Approach. (M.A.
Haralson and J. R. Hassell, eds.) pp. 131-160.
Oxford University Press, Oxford.
Key References
Lane and Sage, 1994. See above.
This review of SPARC provides useful summaries of
its location/abundance in tissues, sequence homologies, and physical characteristics.
Brekken and Sage, 2000. See above.
An up-to-date review of the structure and biology of
SPARC.
Sasaki, T., Miosge, N., and Timpl, R. 1999. Immunochemical and tissue analysis of protease-generated neoepitopes of BM-40 (osteonectin,
SPARC) which are correlated to a higher affinity
binding to collagens. Matrix Biology 18:499508.
Reed, M., Puolakkainen, P.A., Lane, T.F., Dickerson, D., Bornstein, P. and Sage, E.H. 1993. Differential expression of SPARC and thrombospondin-1 in wound repair: Immunolocalization and in situ hybridization. J. Histochem.
Cytochem. 41:1467-1477.
Termine, J.D., Kleinman, H.K., Whitson, S.W.,
Conn, K.M., McGarvey, M.L., and Martin, G.R.
1981. Osteonectin, a bone-specific protein linking mineral to collagen. Cell 26:99-105.
A useful reference for immunostaining and in situ
hybridization protocols for the detection of SPARC.
Yost, J.C., Bell, A., Seale, R., and Sage, E.H. 1994.
Purification of biologically active SPARC expressed in Saccharomyces cerevisiae. Arch. Biochem. Biophys. 314:50-63.
This manuscript has been prepared with the
assistance of Sarah E. Funk and Gail Workman.
Acknowledgement
Contributed by E. Helene Sage
The Hope Heart Institute
Seattle, Washington
Data Processing
and Analysis
10.11.23
Supplement 17
Analysis of Fibronectin Matrix Assembly
UNIT 10.12
Fibronectin (FN) is one of the most ubiquitous components of the extracellular matrix
(ECM). It plays a critical role in organizing ECM structure and influences cell behavior
through interactions with cell surface receptors. Many types of cells secrete cellular FN
and assemble it into a fibrillar network. Assembly proceeds via a step-wise process in
which FN is initially organized into fine cell-associated fibrils and, through continued
accumulation of FN, these fibrils are converted into a dense network of detergent-insoluble
fibrils. Differential solubility in the detergent deoxycholate (DOC) is the principle for
biochemical analysis of FN matrix (DOC-solubility assay).
In this unit, basic methods of detection, quantification, and visualization of the fibrillar
FN matrix are described. The Basic Protocol for analysis of the matrix assembly process
is based on the DOC-solubility assay and describes isolation and analysis of a FN matrix
from cultured cells. Alternate protocols are also provided for analyzing matrix assembly
using exogenous FN (see Alternate Protocols 1 and 2) or by metabolic labeling (see
Alternate Protocol 3). In addition to biochemical analysis of matrix assembly, Alternate
Protocol 4 and Alternate Protocol 5 describe visualization of matrix organization directly by incorporation of fluorescently labeled FN and by indirect immunofluorescence
staining, respectively.
Protocols described in this section require cell culture (UNIT 1.1), purification of plasma
FN (UNIT 10.5), metabolic labeling of cells (UNIT 7.1), immunoblotting (UNIT 6.2), immunoprecipitation (UNIT 7.2), and immunofluorescence staining (UNIT 4.3).
NOTE: All tissue culture incubations are performed in a humidified 37◦ C, 5% CO2
incubator. Some media, e.g., DMEM, require increased levels of CO2 to maintain the
medium at pH 7.4.
NOTE: All solutions and equipment coming into contact with cells must be sterile, and
proper aseptic technique must be used.
ANALYSIS OF MATRIX ASSEMBLY USING A DOC-SOLUBILITY ASSAY
Fibroblasts growing on tissue culture surfaces synthesize FN and assemble it into a fibrillar
matrix. The assay is based on the insolubility of stable FN matrix in 2% DOC detergent
(McKeown-Longo and Mosher, 1983). Cells are lysed in DOC lysis buffer and centrifuged
to separate DOC-insoluble matrix from DOC-soluble material containing cell-associated
and intracellular FN. The DOC-insoluble FN is solubilized in a buffer containing 1%
SDS. The DOC-soluble and -insoluble fractions are resolved by SDS-PAGE, transferred
to nitrocellulose, and analyzed by immunoblotting.
BASIC
PROTOCOL
Materials
Sub-confluent (80% confluent) fibroblasts in a 10-cm tissue culture dish
PBS (APPENDIX 2A)
Trypsin/EDTA solution (GIBCO, Invitrogen)
Culture medium containing 10% FN-depleted serum (see UNIT 10.5 for FN-depletion)
DOC lysis buffer (see recipe)
SDS-solubilization buffer (see recipe)
BCA protein assay kit (Pierce Chemical)
2× SDS sample buffer (see recipe)
Extracellular
Matrix
Contributed by Iwona Wierzbicka-Patynowski, Yong Mao, and Jean E. Schwarzbauer
Copyright 10.12.1
Supplement 25
15-ml screw-cap tube
24-well tissue culture plate
Rubber policeman
1-ml syringe and 26-G, 3/8-in. needle
Additional reagents and equipment for cell culture (UNIT 1.1), gel electrophoresis
(UNIT 6.1), and immunoblotting (UNIT 6.2)
Prepare cell culture
1. Aspirate the medium from a sub-confluent culture of fibroblasts growing on a 10-cm
tissue culture dish.
2. Rinse cells with 5 ml PBS to remove any residual serum.
3. Add 2 ml of trypsin/EDTA solution to cell layer and incubate for 1 to 5 min at room
temperature.
4. When cells detach from the dish, add 2 ml of culture medium and transfer all cells
to a 15-ml screw-cap tube. Count the cell number (UNIT 1.1) and centrifuge 5 min at
100 × g, room temperature.
5. Resuspend fibroblasts in culture medium containing FN-depleted serum at 2.5 ×
105 /ml (1 ml/well). Plate cells onto a 24-well tissue culture plate and incubate up to
8 hr in an incubator.
The cell densities for matrix assembly vary for different cell lines and incubation times.
The cells should be plated at subconfluency for overnight incubations. For incubations of
several hours, cells should be plated almost confluent or touching. Usually, fewer cells are
required for fibroblasts than other cell lines for a given surface area. If experiments are
done in different-sized tissue culture dishes or wells, the amount of reagents and number
of cells need to be scaled up or down accordingly. To avoid the introduction of exogenous
FN from serum, FN-depleted serum should be used to supplement the medium instead.
Perform DOC-solubility assay
6. At desired time, aspirate medium from the wells and gently wash the cells with cold
PBS.
The desired time is determined by the purpose of the experiment and cell type used
for FN matrix assembly. For cells expressing endogenous FN, it can take 4 to 6 hr to
accumulate amounts of DOC-insoluble FN detectable by immunoblotting. DOC-insoluble
matrix should be easily detectable in cells cultured overnight.
7. Add 200 µl of DOC lysis buffer to each well and scrape cells off of the dish using a
rubber policeman. Collect cell lysate with a 1-ml syringe attached to a 26-G, 3/8-in.
needle. To reduce viscosity, pass the cell lysate through the needle five times, transfer
to a 1.5-ml microcentrifuge tube labeled “DOC-insoluble,” and keep tube on ice.
The amount of DOC lysis buffer should be adjusted accordingly for different-sized dishes
(e.g., 2 ml for 10-cm dish, 1 ml for 6-cm dish, and 0.5 ml for 35-mm dish). Volumes
can be adjusted to achieve the desired total protein concentration. Cell lysates should
be thoroughly scraped off of the tissue culture surface. Lysates are passed (usually five
passes) through a small-gauge needle to shear genomic DNA and reduce the viscosity. This
procedure should be carried out without generating air bubbles. Alternatively, viscosity
can be reduced by treating the samples with Triton X-100 followed by DNase I as described
by Quade and McDonald (1988).
8. Microcentrifuge the lysates in 1.5-ml microcentrifuge tubes 15 min at 14,000 rpm,
4◦ C.
Analysis of
Fibronectin
Matrix Assembly
10.12.2
Supplement 25
9. Carefully remove supernatant into a new 1.5-ml microcentrifuge tube labeled “DOCsoluble” and keep on ice.
In some cases, the pellet of insoluble material is not very obvious, therefore, always
mark the side where the pellet will reside after centrifugation. Remove supernatant as
completely as possible and keep the pipet tip away from the pellet.
10. Add 25 µl of SDS-solubilization buffer to the insoluble pellet and mix thoroughly.
It is important to thoroughly dissolve the pellet and to wash the walls of the tube. Do this
by pipetting the SDS-solubilization buffer up and down and by vortexing. Scale up or down
the volume of SDS-solubilization buffer for different sample sizes (e.g., 62.5 µl for 35-mm
dishes). The amount of SDS-solubilization buffer or the concentration of SDS in the buffer
can be increased if cells are plated on a protein-rich substrate (such as Matrigel, gelatin,
or a 3-D matrix prepared from cultured fibroblasts).
Determine total protein concentrations
11. Estimate protein concentrations for DOC-soluble fractions using a BCA protein assay
kit. Follow the manufacturer’s instructions.
12. Normalize samples for the same amount of protein by adjusting volume with 2×
SDS sample buffer and boil 2 min.
Protein concentration in the DOC-soluble fraction is proportional to the number of cells
in the culture and is used to adjust gel sample volumes on a per-cell basis. In a typical
experiment, the total protein concentration ranges from 300 to 800 µg/ml from one well
of a 24-well plate. The maximum amount of protein should be electrophoresed to ensure
detection of FN (usually 3 to 10 µg/lane). DOC-insoluble sample volume for SDS-PAGE
is based on protein concentration in the corresponding DOC-soluble fraction. To detect
monomeric FN, samples should be reduced with 0.1 M DTT in the SDS sample buffer.
Analyze samples by immunoblotting
13. Resolve protein samples using a 5% polyacrylamide-SDS gel and transfer proteins to
nitrocellulose. Perform electrophoresis and immunoblotting according to protocols
described in UNITS 6.1 & 6.2, respectively.
The amount of FN in both fractions can be detected using anti-FN antibodies, followed
by secondary antibodies and ECL reagents.
QUANTIFICATION OF MATRIX ASSEMBLY USING 125 I-LABELED
PROTEIN A
ALTERNATE
PROTOCOL 1
Using 125 I-labeled protein A to detect FN in immunoblots allows the amount of assembled
FN matrix to be quantified. After DOC-soluble and DOC-insoluble fractions are separated
by SDS-PAGE and transferred to a nitrocellulose membrane, FN is detected with an antiFN antibody and secondary antibody followed by radiolabeled protein A. The intensity
of the protein band is then measured using a phosphorimager scanner as described below.
CAUTION: Experiments involving radioactive material handling have to be performed
by trained personnel and in a designated area to avoid contamination. See APPENDIX 1D for
safe use of radioisotopes.
Materials
Samples of DOC-soluble and -insoluble FN from cultures (see Basic Protocol,
steps 1 to 10)
5% (w/v) BSA in TBS buffer (see APPENDIX 2A for TBS)
Primary anti-FN antibody (e.g., HFN7.1, ATCC)
Rabbit secondary antibody (e.g., unconjugated rabbit anti-mouse IgG, Pierce
Chemical)
Extracellular
Matrix
10.12.3
Supplement 25
I-labeled protein A (10 µCi/µg,specific activity; MP Biomedicals)
Buffer A (see recipe)
125
Plastic wrap
Phosphorimager screen (cassette) and scanner
ImageQuant software
Additional reagents and equipment for gel electrophoresis (UNIT 6.1) and
immunoblotting (UNIT 6.2)
1. Perform electrophoresis of samples of DOC-soluble and -incoluble FN from cultures
and transfer to nitrocellulose according to protocols described in UNITS 6.1 and 6.2.
2. Block nitrocellulose with 5% BSA in TBS buffer overnight at 4◦ C.
3. Dilute primary antibody in 10 ml of 5% BSA in TBS and incubate with nitrocellulose
filter for 1 hr at room temperature. Wash three times with 10 ml of 5% BSA in TBS
buffer, 10 min each wash. Incubate with rabbit secondary antibody diluted to 1 µg/ml
in 10 ml of 5% BSA in TBS 1 hr at room temperature. Wash three times with 10 ml
TBS buffer, 10 min each wash.
For optimal binding of protein A and to amplify the signal from the primary anti-FN
antibody, rabbit secondary antibody should be used. The authors usually use unconjugated
rabbit anti-mouse IgG (H+L) (Pierce Chemical) to detect monoclonal anti-FN antibodies.
4. Incubate with ∼6 µCi of 125 I-protein A in 5% BSA in TBS buffer. Wash three times
with 10 ml TBS, 10 min each wash.
5. Wrap the nitrocellulose in plastic wrap and place in phosphorimager cassette. Expose
to phosphor screen for desired amount of time.
The time of exposure is empirically determined. Bands can usually be detected after an
overnight exposure but weaker signals may require exposure for ≥1 week.
6. Read the screen on phosphorimager scanner. Determine the number of counts associated with each band using ImageQuant software.
ALTERNATE
PROTOCOL 2
ANALYSIS OF ASSEMBLY OF EXOGENOUS FN
Some cell lines (e.g., CHO and many tumor cell lines) do not produce significant levels
of FN. To assess their matrix assembly capability and to study regulation of the assembly
process, the addition of exogenous FN is required. For quantitation purposes, 125 I-labeled
FN can be included in the exogenous FN.
Materials
Purified plasma FN (UNIT 10.5)
I-labeled FN (∼1 µCi/µg; MP Biomedicals; optional)
Additional reagents and equipment for trypsinization and collection of cells, and
isolation and analysis of DOC-insoluble and DOC-soluble FN (see Basic
Protocol)
125
1. Prepare purified FN from blood plasma using the protocol described in UNIT 10.5.
2. Trypsinize and collect cells following Basic Protocol, steps 1 to 5.
3. Allow cells to attach and spread, usually 60 min at 37◦ C
Analysis of
Fibronectin
Matrix Assembly
10.12.4
Supplement 25
4. Add 25 to 50 µg/ml of exogenous FN and incubate for desired amount of time.
The amount of exogenous FN can be varied and the optimal amount for assembly should
be determined empirically. To quantify the amount of FN in the matrix, 125 I-labeled FN
can be added together with unlabeled FN. Alternatively, cells can be allowed to assemble
exogenous FN for a period of time and then 125 I-labeled FN can be added for a shorter
period. Additional reagents such as activators or inhibitors of matrix assembly can be
added along with exogenous FN or at any time during the incubation.
5. Isolate and analyze DOC-insoluble and DOC-soluble FN (see Basic Protocol, steps
6 to 13).
When 125 I-labeled FN is included, the SDS-polyacrylamide gel is dried and directly exposed to a phosphorimager screen (for gel drying, see Alternate Protocol 3).
ANALYSIS OF METABOLICALLY LABELED FN
Using metabolically labeled cells for matrix assembly studies allows one to determine the
incorporation of endogenous FN over specific time periods and also provides radiolabeled
material for quantification. 35 S-labeled FN in DOC-soluble and -insoluble fractions are
isolated by immunoprecipitation and analyzed using a phosphorimager after resolution
by SDS-PAGE.
ALTERNATE
PROTOCOL 3
Materials
Cell cultures for labeling
Culture medium containing FN-depleted serum (see UNIT 10.5 for FN-depletion)
Labeling medium (see recipe)
35
S-methionine (>1000 Ci/mmol)
IP buffer (see recipe)
Protein A–Sepharose beads
35-mm tissue culture dish or 6-well plate
Phosphorimager screen and scanner
ImageQuant software
Additional reagents and equipment for cell preparation (see Basic Protocol), IP
protocol (UNIT 7.2)
1. Prepare cells according to Basic Protocol, steps 1 to 5.
2. Plate 1 × 106 cells in culture medium containing FN-depleted serum in 35-mm tissue
culture dishes or 6-well plates. Let cells attach and spread.
Alternatively, cells can be plated and allowed to grow until 80% to 90% confluent.
3. Aspirate medium, rinse cells with 2 ml labeling medium minus methionine, and
replace with 1 ml labeling medium. Add 25 µCi of 35 S-methionine per milliliter of
labeling medium and mix well. Incubate cells for desired amount of time.
The optimal concentration of 35 S-methionine depends on cell type and length of labeling.
A 24-hr labeling period with 25 µCi/ml of 35 S-methionine is typically used to determine
assembly competence and FN expression by cells. Shorter labeling times with increased
amounts of 35 S-methionine can be used (e.g., 50 µCi/ml for 6 to 8 hr or 100 µCi/ml for
2 hr).
4. Remove medium and save for detection of FN in the medium (see UNIT 10.5).
5. Wash cells with 2 ml ice-cold PBS.
The waste PBS should be disposed of in a properly labeled radioactive waste container.
6. Prepare DOC-soluble and -insoluble fractions as described in Basic Protocol, steps
6 and 7, using 500 µl DOC-lysis buffer and 62.5 µl SDS-buffer.
Extracellular
Matrix
10.12.5
Supplement 25
7. Determine protein concentration. Normalize the samples to equal protein concentration in ∼125 µl of DOC-soluble sample and 50 µl of DOC-insoluble sample.
8. Adjust volume to 500 µl using stock solutions to give composition of IP buffer.
9. Immunoprecipitate FN from the soluble and insoluble fractions using anti-FN antibody. Follow the IP protocol described in UNIT 7.2.
10. Run immunoprecipitated samples on 5% SDS-PAGE. Dry the gel and expose to
phosphorimager screen. Scan the intensity of FN bands using a phosphorimager
scanner and analyze using ImageQuant software.
Electrophoresis of one-third of the immunoprecipitate is usually sufficient to detect FN
in fibroblast matrix. To dry gel, fix in 50% methanol/10% acetic acid solution 30 min at
room temperature, rehydrate with several changes of water, place on a sheet of Whatman
paper, place on gel dryer, cover with plastic wrap, and dry at 80◦ C under vacuum.
ALTERNATE
PROTOCOL 4
DIRECT DETECTION OF MATRIX ASSEMBLY BY INCORPORATION OF
FLUORESCENTLY LABELED FIBRONECTIN
In addition to the biochemical methods, FN matrix assembly can be monitored using
fluorescence techniques. Fibrillar matrix can be detected by indirect immunofluorescence
staining using antibodies with fluorescent tags (see Alternate Protocol 5) or matrix can
be labeled directly by incorporation of fluorescently tagged FN.
Materials
Purified FN (UNIT 10.5)
50 mM sodium bicarbonate, pH 8.
Sulfo-NHS-rhodamine or fluorescein (Pierce Chemical)
CAPS-NaCl solution
Cell cultures (see Basic Protocol, steps 1 to 4)
Culture medium containing FN-depleted serum (see UNIT 10.5 for FN-depletion)
PBS/Mg (PBS containing 0.5 mM MgCl2 )
3.7% (v/v) formaldehyde in PBS/Mg
0.5% NP-40 (v/v) in PBS/Mg
FluoroGuard (Bio-Rad)
Nail polish
Spectrophotometer
12-mm circular coverslips
24-well plate
Fine-tip forceps
Beakers
Paper towels
Kimwipes
Glass microscope slides
Prepare fluorescently labeled FN
1. Dialyze 1 mg/ml of purified FN in 50 mM sodium bicarbonate, pH 8.5, overnight at
4◦ C.
2. Immediately prior to use, make 1 mg/ml of sulfo-NHS-rhodamine in distilled water.
Add 40 µg of sulfo-NHS-rhodamine per 1 mg of dialyzed FN.
3. Incubate 2 hr on ice in dark.
Analysis of
Fibronectin
Matrix Assembly
4. Dialyze the reaction mixture against CAPS-NaCl solution at 4◦ C. Use two changes
of at least a 100-fold excess volume each.
10.12.6
Supplement 25
5. Determine protein concentration by reading A280 using a spectrophotometer. Store
100-µl protein aliquots for 6 months at −80◦ C.
Prepare cultures
6. Place 12-mm circular coverslips in the wells of a 24-well plate.
Sterilize coverslips prior to use by autoclaving.
7. Prepare cells in culture medium containing FN-depleted serum according to Basic
Protocol, steps 1 to 4.
8. Plate cells on coverslips in wells of 24-well plates and let them attach and spread for
30 to 60 min.
Make sure that coverslips are at the bottom of the wells rather than floating in the medium.
9. Add 25 µg/ml of rhodamine-labeled FN and incubate cells at 37◦ C for desired amount
of time.
10. Aspirate medium and gently wash cells with 1 ml PBS/Mg. Fix with 1 ml of 3.7%
formaldehyde in PBS/Mg 15 min at room temperature. Aspirate fixing solution and
wash three times with 1 ml PBS/Mg.
11. Carefully remove coverslips from wells using fine-tip forceps. Set up three beakers
containing 100 ml PBS/Mg. Wash coverslips by dipping several times in each beaker.
Do a final wash in 100 ml water. Drain coverslips on a dry paper towel and dry the
clean (non-cellular) face of coverslip with a Kimwipe.
Alternatively, cells can be removed from the well prior to fixation. Place in a humidified
chamber (e.g., a Petri dish lined with a moistened paper towel). Gently pipet 25 to 50 µl
of fixing solution on top of cells and incubate 15 min at room temperature.
Visualize incorporated labeled FN
12. Place a drop (2 to 4 µl) of FluoroGuard on a glass microscope slide. Carefully place
the coverslip with cells face down on top of the FluoroGuard.
13. Seal periphery of the coverslip with nail polish and let dry.
Properly sealed slides can be stored several months at −20◦ C.
14. Examine slides with a fluorescence microscope equipped with rhodamine filters.
DETECTION OF MATRIX ASSEMBLY BY INDIRECT
IMMUNOFLUORESCENCE STAINING
ALTERNATE
PROTOCOL 5
Fibrillar matrix can be detected by indirect immunofluorescence staining using primary
anti-FN antibody and secondary antibodies with fluorescent tags.
Materials
Cell cultures
Primary anti-FN antibody
2% (w/v) ovalbumin in PBS/Mg solution
Fluorescein-conjugated or rhodamine-conjugated goat anti-mouse IgG (or
anti-rabbit IgG)
Petri dishes
Fluorescence microscope
Additional reagents and equipment for detection of FN matrix (see Alternate
Protocol 4)
Extracellular
Matrix
10.12.7
Supplement 25
1. Follow Alternate Protocol 4, steps 6 to 10, except do not add fluorescently labeled
FN.
Exogenous FN can be added to cells if the cell type does not produce FN.
For detection of intracellular proteins, fixed and washed cells can be permeabilized with
1 ml 0.5% NP40 in PBS/Mg, 15 min at room temperature followed by three washes with
1 ml PBS/Mg.
2. Carefully remove coverslips from wells using fine-tip forceps. Place in a humidified
chamber (e.g., a Petri dish lined with a moistened paper towel).
3. Add 25 to 50 µl of diluted primary anti-FN antibodies in 2% ovalbumin/PBS/Mg
solution and incubate for 30 min at 37◦ C.
The optimal antibody dilution should be determined by the individual user. Typical dilutions are: 1:50 to 1:250 for hybridoma culture supernatant or polyclonal antisera and
1:500 to 1:2000 for ascites fluid or concentrated hybridoma supernatant. Antibody incubations can be done in a 37◦ C incubator (without CO2 ).
4. Wash coverslips by dipping in 100 ml PBS/Mg three times.
5. Incubate cells with 25 to 50 µl of the secondary antibody (e.g., fluorescein- or
rhodamine-conjugated goat anti-mouse or rabbit IgG) in 2% ovalbumin in PBS/Mg
solution 30 min at 37◦ C.
6. Rinse and mount the coverslips according to Alternate Protocol 4, steps 12 to 13.
7. Examine coverslips using a fluorescence microscope with appropriate filters.
Buffer A
25 mM Tris·Cl, pH 7.5
150 mM NaCl
0.1% (v/v) Tween-20
Store up to 6 months at 4◦ C
CAPS-NaCl solution
10 mM 3-[cyclohexylamino]-1-propanesulfonic acid (CAPS)
150 mM NaCl
Adjust to pH 11 with 5 N NaOH
Store up to 6 months at 4◦ C
Cell labeling medium
To cell culture medium without methionine, add non-radioactive methionine (tissue
culture–grade) to a final concentration that is 0.1× the amount in normal medium.
Store up to 3 months at 4◦ C.
Analysis of
Fibronectin
Matrix Assembly
DOC lysis buffer
2% (w/v) sodium deoxycholate (from 10% DOC stock solution in dH2 O, kept for
months at −20◦ C)
20 mM Tris·Cl, pH 8.8 (from 1 M Tris·Cl stock solution, pH 8.8, stored at room
temperature)
continued
10.12.8
Supplement 25
2 mM phenylmethysulfonyl fluoride (PMSF; 0.2 M PMSF in ethanol and stored at
−20◦ C)
2 mM EDTA (from 0.5 M EDTA stock solution, stored at room temperature)
2 mM iodoacetic acid (freshly prepared 100 mM stock in H2 O)
2 mM N-ethylmaleimide (freshly prepared 100 mM stock in H2 O)
Prepare fresh
Dissolve any precipitated PMSF in the stock solution by vortexing.
IP buffer
2.5 mM EDTA
2.5 mg/ml BSA
0.5% (v/v) NP-40
0.5% (w/v) DOC
0.1% (w/v) SDS
Prepare fresh
SDS sample buffer, 2×
10% (v/v) glycerol
2 mM EDTA
4% (w/v) SDS
0.05% (w/v) bromphenol blue
SDS solubilization buffer
1% (w/v) SDS (from 20% SDS stock solution, kept at room temperature)
2 mM PMSF
2 mM EDTA
2 mM iodoacetic acid (freshly prepared in H2 O)
2 mM N-ethylmaleimide (freshly prepared in H2 O)
Prepare fresh
COMMENTARY
The extracellular matrix (ECM) is a protein
network that acts as a framework for tissue
architecture and dynamically regulates many
cellular functions such as adhesion, migration,
growth, and differentiation. A major component of most matrices, fibronectin (FN) is a
multifunctional glycoprotein synthesized by
many cell types including fibroblasts, endothelial cells, myoblasts, and astrocytes (Hynes,
1990; Pankov, 2002). In addition to cellular
FN produced by cells in tissues, there is a
considerable amount of FN in blood plasma.
This plasma FN is made by hepatocytes and
differs from cellular FN by alternative splicing (Schwarzbauer, 1991). FN is synthesized
and secreted as a disulfide-bonded dimer. It is
assembled by cells into a fibrillar matrix via
a regulated, step-wise process (Schwarzbauer
and Sechler, 1999; Wierzbicka-Patynowski
and Schwarzbauer, 2003). Initiation of assembly depends on FN binding to cell surface integrin receptors. Once immobilized, dimeric
FN is active to participate in FN-FN interactions leading to formation of fibrils. As
the process proceeds, short fine fibrils become longer and denser. Initially the integrinassociated fibrils are soluble in deoxycholate
(DOC) detergent. As additional FN dimers
are assembled, the DOC-soluble pool is converted into DOC-insoluble matrix in which
the FN dimers are quite tightly associated into
detergent-stable high-molecular weight multimers. (McKeown-Longo and Mosher, 1983).
Extracellular
Matrix
10.12.9
Supplement 25
Thus, conversion from cell-associated fine fibrils to stable matrix can be monitored by analysis of the amounts of FN in these two pools.
FN fibril organization can be examined by immunofluorescence staining.
Biochemical and microscopic analyses provide distinct types of information about FN
matrix assembly. Immunoblotting of detergent
lysates allows quantification of the amounts of
FN in DOC-soluble and -insoluble fractions as
well as determination of whether matrix FN is
intact or has been proteolyzed. Immunofluorescence analyses allow one to follow the progression of FN assembly from fine fibrils to
dense, stable matrix as well as to determine
the overall distribution of FN in the cell layer.
staining usually shows a bright fibrillar pattern at times corresponding to detectable FN
by immunoblotting.
Critical Parameters
McKeown-Longo, P.J. and Mosher, D.F. 1983.
Binding of plasma fibronectin to cell layers of
human skin fibroblasts. J. Cell Biol. 97:466-472.
At the time of analysis, cells should be sufficiently dense to ensure optimal conditions for
fibril formation between adjacent cells but not
so crowded that they are approaching quiescence. Critical steps in the DOC-solubility assay include reducing the viscosity of genomic
DNA, gently removing DOC-soluble from the
DOC-insoluble fraction after centrifugation,
and completely dissolving the DOC-insoluble
pellet.
Anticipated Results
Fibroblasts and other cells that synthesize
significant levels of FN usually yield a visible amount of DOC-insoluble material within
several hours after plating. Cell types that produce very little FN, such as many tumor cell
lines, can require incubations as long as 24 hr
for isolation of detectable DOC-insoluble FN.
To increase the amount of FN available for
assembly, cell cultures can be supplemented
with exogenous FN allowing detection of FN
matrix after much shorter incubation periods.
The sensitivity of the DOC-solubility assay is
in the nanogram range. Immunofluorescence
Time Considerations
Time after plating cells is variable depending on the experiment. Performance time for
the DOC-solubility assay depends on the number of samples but should be easily completed
in 1 to 2 hr. SDS-PAGE and immunoblotting
are described in UNITS 6.1 & 6.2. Similarly, antibody staining requires ∼2 hr for fixation and
incubations.
Literature Cited
Hynes, R.O. 1990. Fibronectins. Springer-Verlag,
New York.
Pankov, R. and Yamada, K.M. 2002. Fibronectin at
a glance. J. Cell Sci. 115:3861-3863.
Quade, B.J. and McDonald, J.A. 1988. Fibronectin’s amino-terminal matrix assembly site
is located within the 29-kDa amino-terminal
domain containing five type I repeats. J. Biol.
Chem. 263:19602-19609.
Schwarzbauer, J.E. 1991. Alternative splicing of
fibronectin: Three variants, three functions.
BioEssays 13:527-533.
Schwarzbauer, J.E. and Sechler, J.L. 1999. Fibronectin fibrillogenesis: A paradigm for extracellular matrix assembly. Curr. Opin. Cell Biol.
11:622-627.
Wierzbicka-Patynowski, I. and Schwarzbauer, J.E.
2003. The ins and outs of fibronectin matrix assembly. J. Cell Sci. 116:3269-3276.
Contributed by Iwona
Wierzbicka-Patynowski, Yong Mao,
and Jean E. Schwarzbauer
Princeton University
Princeton, New Jersey
Analysis of
Fibronectin
Matrix Assembly
10.12.10
Supplement 25
Non-Radioactive Quantification of
Fibronectin Matrix Assembly
UNIT 10.13
Fibronectin (FN) matrix assembly is a cell-dependent process that converts soluble FN
molecules into elaborate extracellular fibrillar matrices. This process relies on activated integrins, cellular contractility, and unmasking of cryptic fibronectin assembly sites for generation of insoluble fibrils (Geiger et al., 2001). The signaling pathways involved in matrix
assembly have just begun to be elucidated (Wierzbicka-Patynowski and Schwarzbauer,
2002), and further studies will require simple and reliable assays for quantification of matrix assembly associated with parallel determinations of the activity of various signaling
molecules.
This unit provides a protocol (see Basic Protocol) for non-radioactive determination of
the rate of incorporation of biotinylated fibronectin into the insoluble matrix organized
by cultured cells. This protocol provides a simple method for quantifying changes in
matrix assembly that result from different experimental treatments or conditions with
concomitant determinations of the activation state of various signaling molecules that
may be involved in the process of matrix assembly. This unit also provides a protocol
(see Support Protocol) for biotinylation of purified fibronectin.
QUANTIFICATION OF MATRIX ASSEMBLY USING BIOTINYLATED
FIBRONECTIN
BASIC
PROTOCOL
This protocol can be applied to nearly all cultured cell lines with little or no modifications. It describes labeling of the matrix assembled by cultured cells with biotinylated fibronectin, followed by isolation of detergent-insoluble fibronectin matrices (see
UNIT 10.12). Quantification of the incorporated biotinylated FN is performed by electrophoresis (UNIT 6.1), electroblotting (UNIT 6.2), and detection with peroxidase-conjugated
streptavidin. The quantities of intermediate filament proteins present in the detergentinsoluble fractions are determined by immunoblotting for use as the internal controls
for isolation efficiency of the detergent-insoluble matrix in each fraction. The detergentsoluble fractions are used to monitor the ability of cells to bind fibronectin under the
conditions being tested and to determine simultaneously the activation state of the signaling molecules of interest.
Materials
Fibroblasts (or any cell line of interest)
Dulbecco’s modified Eagle medium supplemented with 10% (v/v) fetal bovine
serum (DMEM/10% FBS; APPENDIX 2A)
Biotinylated fibronectin (see Support Protocol)
PBS (APPENDIX 2A), ice cold
DOC extraction buffer (see recipe)
2× SDS sample buffer (APPENDIX 2A)
1 M NaF
0.1 M sodium orthovanadate solution (APPENDIX 1B)
10 mM leupeptin (APPENDIX 1B)
25 mM pepstatin A (APPENDIX 1B)
0.2 M phenylmethanesulfonyl fluoride (PMSF; APPENDIX 1B)
2× SDS sample buffer (APPENDIX 2A)
Extracellular
Matrix
Contributed by Roumen Pankov and Kenneth M. Yamada
Copyright 10.13.1
Supplement 25
8% (w/v) polyacrylamide separating gels with 4% (w/v) stacking gels (UNIT 6.1) or
commercially available pre-cast 4% to 12% gradient gels (e.g., Novex) for SDS
gel electrophoresis
Prestained protein molecular size standards (e.g., Novex)
Transfer buffer (UNIT 6.2)
Ponceau S solution (UNIT 6.2)
Tris-buffered saline with 0.1% (v/v) Tween 20 (TTBS; APPENDIX 2A)
Blocking solution: TTBS containing 5% (w/v) dry nonfat milk (TTBS/milk)
Streptavidin, horseradish peroxidase (HRP)-conjugated (e.g., Jackson
ImmunoResearch)
Enhanced chemiluminescence (ECL) detection reagent (UNIT 14.2)
Primary antibody: monoclonal anti-vimentin (e.g., Sigma)
Secondary antibody: horseradish peroxidase (HRP)-conjugated anti-rabbit or
anti-mouse antibodies (e.g., Amersham Bioscience)
35-mm tissue culture dishes
Plastic cell scraper (rubber policeman)
23-G needle and 1-ml syringe
Micropipettors
Porous electrotransfer pads
Whatman 3MM filter papers cut to gel size
Two nitrocellulose membranes cut to gel size
SDS-PAGE/transfer apparatus (e.g., Bio-Rad, Novex)
Constant-voltage/current power supply (e.g., Bio-Rad)
Flat containers
Rocking shaker
Heat-sealable plastic bags and sealer
Sonicator/ultrasonic processor
Plastic wrap
X-ray film (e.g., Hyperfilm; Amersham Bioscience)
Tube heater (e.g., Thermomixer; Eppendorf) or boiling water bath
Film cassette for X-ray film
X-ray film developer
Additional reagents and equipment for dialysis (APPENDIX 3C), tissue culture
(UNIT 1.1), SDS-PAGE (UNIT 6.1), and immunoblotting (UNIT 6.2)
NOTE: All reagents and equipment coming into contact with living cells must be sterile
and aseptic technique should be used accordingly.
NOTE: All tissue culture incubations should be performed in a 37◦ C, 10% CO2 humidified
incubator. Use pre-warmed cell culture medium for all treatments.
Prepare and treat cells
1. Plate cells in 35-mm dishes so that after spreading they will be ∼90% to 95%
confluent, and culture overnight in DMEM/10% FBS.
Depending on the size of fibroblasts used, the desired confluency can be obtained by
plating between 0.25 × 106 cells (e.g., primary human fibroblasts) and 0.5 × 106 cells
(e.g., NIH 3T3 cells).
2. After overnight incubation, wash cells with 1.0 ml DMEM/10% FBS and add
1 ml/plate of the same medium containing 20 µg/ml biotinylated FN.
Quantification of
Fibronectin
Matrix Assembly
Treatment with reagents of interest can be incorporated in this step, which can include
chemical compounds, peptides, antibodies, etc. Add an appropriate volume of stock
10.13.2
Supplement 25
solution containing the reagent to one plate and the same volume of the solvent (e.g.,
DMSO) to another plate that will serve as a control. Label each plate.
If the presence of serum in the medium interferes with the action of the tested reagent (e.g.,
growth factors), replace serum with 1% (w/v) bovine serum albumin during the treatment
period.
3. Incubate plates in a tissue culture incubator for 4 hr.
Depending on the ability of the cells to form fibronectin matrix, this time period can
be varied. For example, a 3-hr incubation is sufficient for primary human fibroblasts to
incorporate readily detectable quantities of biotinylated fibronectin into the detergentinsoluble fraction, while 4 to 6 hr are necessary for β 1 null GD 25 cells to polymerize
enough labeled FN for reliable detection.
4. Aspirate medium and wash cell monolayers three times with 2 ml ice-cold PBS each.
5. Lyse cells in 0.5 ml DOC extraction buffer, scrape plates with plastic scraper, pass
the lysate five times through a 23-G needle attached to a 1-ml syringe, and transfer
into labeled 1.5-ml microcentrifuge tubes. Keep lysates on ice.
Shearing DNA by passing lysates through a thin needle is necessary to reduce viscosity
and to allow sedimentation of small, insoluble matrix aggregates during centrifugation.
Check that the needle is firmly attached to the syringe and gently aspirate and expel lysate
from the syringe while avoiding the formation of bubbles.
6. Centrifuge lysates 20 min at 20,000 × g, 4◦ C.
Prepare detergent-insoluble matrix
7. Transfer a 100-µl aliquot of the supernatant into new 1.5-ml microcentrifuge tubes
labeled DOC soluble, mix with 100 µl of 2× SDS sample buffer, and leave on ice.
8. Carefully remove the rest of the supernatant, leaving the DOC-insoluble pellet intact.
9. Wash the pellet by resuspending it in 100 µl DOC extraction buffer, pipetting up and
down five times with a micropipettor.
10. Centrifuge 10 min at 20,000 × g, 4◦ C.
11. Carefully remove the supernatant and dissolve the pellet in 50 µl of 2× SDS sample
buffer.
Attention should be paid to avoid losing the pellet, which is usually very small and
sometimes difficult to visualize.
Analyze samples by SDS-PAGE
12. Boil samples collected at steps 7 and 11 in a water bath for 3 min or heat 5 min on a
95◦ C heating block.
The samples can be stored sealed and frozen at least 1 month at −20◦ C.
13. Cast an 8% polyacrylamide separating gel with a 4% stacking gel (UNIT 6.1) or use a
commercially available pre-cast 4% to 12% gradient gel.
14. Load 25 µl of each sample/gel lane and a separate lane containing prestained protein
standards on the gel.
Load the set of DOC-insoluble samples first, followed by the set of DOC-soluble samples.
Divide the two sets with marker proteins or an empty well, so that after the transfer the two
portions of the membrane can be separated. Alternatively, load the two sets of samples
on two separate gels.
15. Electrophorese the gel(s) at 150 V until the bromophenol blue dye reaches the bottom
of the gel (see UNIT 6.1).
Extracellular
Matrix
10.13.3
Supplement 25
Transfer separated proteins from gel to membrane
16. When electrophoresis is complete, remove gel(s) from gel plates, cut off the stacking
gel(s), and incubate the separating portion of gel(s) in 50 ml transfer buffer for 15 min.
Use gloves to handle gels and membranes, since oil from hands can interfere with the
transfer.
17. Assemble the transfer sandwich consisting of porous electrotransfer pad, Whatman
3MM filter paper, nitrocellulose membrane, equilibrated acrylamide gel, second
Whatman 3MM filter paper, and second pad (Fig. 6.2.1).
All pads, filter papers, and nitrocellulose membranes should be handled using gloves
and pre-wetted with transfer buffer. The transfer cassette should be assembled submerged
under the transfer buffer to avoid trapping air bubbles. Keep the orientation of the gel
(judged by the position of the prestained protein standards) such that it will ensure the
correct order of the samples after transfer onto the nitrocellulose membrane.
18. Place the transfer sandwich into the electroblotting apparatus filled with transfer
buffer with the nitrocellulose membrane on the cathode side of the gel. Connect
the apparatus to the power supply and transfer proteins for 1 hr at 100 V (constant
voltage) with cooling (UNIT 6.2).
Transfer time depends on the size of the proteins, acrylamide percentage, and thickness
of gel. The completeness of transfer can be easily judged by the extent of transfer of the
prestained protein standards.
19. After completing the electrotransfer, turn off the power supply and disassemble the
apparatus and the transfer cassette. Remove the nitrocellulose membranes and stain
with 50 ml Ponceau S solution in a flat container for 5 min. Destain membranes with
several rinses of distilled water.
If the two sets of samples were run on the same gel, cut membrane along the well that
separates the two sets (see step 14).
Two membranes can be incubated in the same container by orienting them back to back.
Staining with Ponceau S does not interfere with subsequent reactions and provides a good
estimate of protein loading, separation, and quality of transfer. The amounts of protein
in the DOC-insoluble samples will be several-fold lower than the proteins in the DOCsoluble samples, because most cellular proteins are soluble in DOC, whereas only FN
and a few other proteins are insoluble.
The Ponceau S solution can be reused several times.
Probe membranes with streptavidin-HRP and antibodies
20. Rinse membranes one time with TTBS and incubate in 50 ml blocking solution for
30 min at room temperature with gentle shaking.
Milk proteins in the blocking buffer are used to saturate free protein-binding sites and
to prevent nonspecific binding. Do not incubate the membrane >1 hr in blocking buffer,
since it has a slight stripping effect and may cause detachment of transferred proteins.
21. Dilute streptavidin-HRP as recommended by the manufacturer in a final volume of
10 ml TTBS containing 3% dry nonfat milk. Place membranes in a heat-sealable
plastic bag, add diluted streptavidin-HRP, and seal the bag. Incubate membranes 1
hr at room temperature with gentle shaking.
Two membranes can be incubated in the same bag by orienting them back-to-back. Remove
all air bubbles from the bag before sealing.
Quantification of
Fibronectin
Matrix Assembly
22. Remove membranes from the bag and wash them three times, 15 min each, with
50 ml TTBS in a flat container with vigorous shaking.
10.13.4
Supplement 25
23. Use the ECL immunodetection protocol (UNIT 6.2) to detect biotinylated fibronectin.
Incubate membranes with ECL solution for 1 min, remove excess fluid by touching
the edge of the membrane held vertically to a horizontal piece of filter paper, wrap
membranes in plastic wrap, and expose to X-ray film.
Do not allow membranes to dry after the exposure.
24. Rinse membranes with 10 ml TTBS.
Probe membranes for control proteins
25. Incubate the membrane containing the DOC-insoluble samples with anti-vimentin
antibody, and the membrane containing DOC-soluble samples with anti-actin antibody. Dilute the antibodies according to the manufacturer’s recommendations in a
final volume of 10 ml TTBS containing 3% dry nonfat milk. Place membrane in a
heat-sealable plastic bag, add diluted antibodies, and seal the bag. Incubate membranes for 1 hr at room temperature (or overnight at 4◦ C) with gentle shaking.
This second reaction is used as an internal control for the efficiency of matrix isolation
and loading of the gels. If primary cells are used in the experiment, they may not express
detectable amounts of vimentin. In such cases, a different intermediate filament protein
can be used as a marker. The choice of appropriate marker will depend on the origin of
the primary cells (see Coulombe et al., 2001).
26. Repeat step 22.
27. Dilute the secondary antibody in a final volume of 10 ml TTBS containing 3% dry
nonfat milk according to manufacturer’s instructions. Place membranes in a new
heat-sealable bag, add diluted antibody, and seal. Incubate 30 to 45 min at room
temperature with gentle shaking.
Either HRP- or alkaline phosphatase–conjugated secondary antibody can be used. HRPconjugated secondary antibodies can be combined with the high-sensitivity ECL detection
system.
This system allows detection of signals from weak antibodies, although attention should
be paid to linearity if accurate quantification of the signal is necessary (see Commentary).
28. Repeat step 22 and detect the secondary antibody using the ECL procedure described
in step 23.
The membrane containing resolved proteins from DOC-soluble fractions can be re-probed
again with antibodies recognizing the phosphorylated (activated) forms of different signaling molecules of interest. This step is possible if the molecular masses of the signaling
molecules are different from those of fibronectin (250 kD) and actin (45 kD) and if the
new signals on the membrane will not interfere with any previous band detected. Alternatively, the remainder of DOC-soluble samples can be used for additional immunoblotting
experiments with other antibodies.
Quantify gels
29. Measure the optical density (absorbance) of the signals from biotinylated FN and the
antibodies for each sample using densitometry (UNIT 6.3) or image processing software
(e.g., NIH Image).
30. Normalize the densitometry values from biotinylated FN to the readings for vimentin
in the DOC-insoluble fraction and the densitometry readings from biotinylated FN to
the readings for actin in the DOC-soluble fraction. Calculate fold or percent changes
relative to the control.
Extracellular
Matrix
10.13.5
Supplement 25
SUPPORT
PROTOCOL
BIOTINYLATION OF PLASMA FIBRONECTIN
Biotin is a vitamin that binds with high affinity to avidin and streptavidin. Because of its
small size (244 Da), it can be used to label proteins without significant risk of affecting
their function.
This protocol describes labeling of fibronectin with sulfo-NHS-biotin, followed by a
dialysis step to remove unconjugated biotin.
Fibronectin (e.g., Sigma or purified as described in UNIT 10.5)
Bicarbonate buffer (see recipe)
Sulfo-NHS-biotin (e.g., Pierce)
1. Dialyze 0.5 mg fibronectin against 1 liter of bicarbonate buffer overnight at 4◦ C or 2
hr at room temperature, change bicarbonate buffer once, and dialyze for an additional
1 hr at room temperature.
For a 25-sample preparation, 0.5 mg of fibronectin in 0.5 to 1.0 ml will be sufficient.
2. Immediately prior to use, dissolve 0.5 mg sulfo-NHS-biotin in 0.5 ml deionized
water.
3. Immediately add 40 µl sulfo-NHS-biotin solution to 0.5 ml of fibronectin solution
and incubate 30 min at room temperature on a gently rocking shaker.
Avoid harsh mixing and foaming, because fibronectin tends to denature and precipitate at
the liquid/air interface.
4. Dialyze the biotinylated fibronectin against 1 liter of TBS overnight at 4◦ C or 2 hr
at room temperature, change TBS buffer once, and dialyze for an additional 1 hr at
room temperature.
5. Centrifuge fibronectin solution in a microcentrifuge 15 min at maximum speed at
room temperature and save the supernatant.
This step will remove possible precipitates from the solution.
6. Determine protein concentration using BCA assay (APPENDIX
aliquots indefinitely at −70◦ C. Avoid repeated freeze-thawing.
3H).
Store 200-µl
Quantification of
Fibronectin
Matrix Assembly
DOC extraction buffer
1% (w/v) sodium deoxycholate
2 mM N-ethylmaleimide, add fresh
2 mM iodacetic acid, add fresh
2 mM EDTA (APPENDIX 2A)
50 µM leupeptin (APPENDIX 1B), add fresh
50 µM pepstatin (APPENDIX 1B), add fresh
1 mM PMSF (APPENDIX 1B), add fresh
1 mM sodium vanadate (APPENDIX 1B), add fresh
50 mM NaF (APPENDIX 1B), add fresh
Prepare fresh
This DOC extraction buffer is based on the work of McKeown-Longo and Mosher (1983).
10.13.6
Supplement 25
Bicarbonate buffer
50 mM NaHCO3
100 mM NaCl
Adjust pH to 8.5 with 1 M NaOH if necessary
Store up to 2 weeks at 4◦ C
COMMENTARY
Fibronectin matrices not only provide substrates for cell attachment and tissue organization, but they also regulate migration, cell
growth, and differentiation. These matrices
are organized by cells from secreted cellular fibronectin and soluble plasma FN from
blood, where this glycoprotein is present at
high concentrations (300 µg/ml). In vitro,
cells polymerize exogenous fibronectin from
serum-supplemented culture medium together
with the secreted FN (McKeown-Longo and
Mosher, 1983). Even cells that do not produce
fibronectin are able to form matrices when this
molecule is provided exogenously (Sottile and
Hocking 2002). This property permits the use
of labeled exogenous FN as a tracer during the
process of matrix assembly.
Experiments with iodinated fibronectin
have shown that shortly (2 to 10 min) after
addition to the culture medium, it binds to
the cell culture and becomes resistant to simple rinses with buffers that preserve cell viability (McKeown-Longo and Mosher, 1983).
Prolonged incubation leads to increased binding and formation of two different fibronectin
pools that can be distinguished by their solubility in 1% deoxycholate (DOC). The DOCsoluble pool represents FN bound by cellular
receptors and preexisting matrix fibrils, while
the DOC-insoluble pool is believed to include
bound fibronectin that is incorporated in the
matrix through detergent-resistant interactions
such as disulfide bonding. Based on the similarities between the incorporation of exogenous and endogenous FN, the quantities of labeled FN in different fractions appear to be
proportional to the total amounts of fibronectin
present in these fractions. Thus, the ability of
cells to bind and organize fibronectin matrix
under different conditions where a variety of
signaling pathways are affected can be studied by following the relative distributions and
quantifying the amounts of labeled (tracer) fibronectin between these two fractions.
Purification of the DOC-insoluble fraction
from relatively small amounts of cultured cells
very often poses technical difficulties in handling such small and often invisible pellets.
Possible losses of part of the DOC fractions
will lead to erroneous interpretation of the results. This serious problem can be avoided by
a parallel determination of the amount of an
intermediate filament protein present in this
fraction. Due to their high insolubility, these
proteins resist DOC extraction, and their quantities can be used as internal controls for the efficiency of isolation and recovery of the DOCinsoluble fraction.
Radioactive isotopes are widely used for
quantification purposes, but this method demands special training, equipment, and disposal of reagents. Substitution of biotinylated
fibronectin for radioiodinated fibronectin simplifies the technique while still preserving the
necessary level of sensitivity. Moreover, covalently linking N-hydroxysulfosuccinimide
(NHS)-coupled biotin to fibronectin is a routine and easy procedure. It permits detection
of biotinylated FN with avidin through the
strongest known noncovalent recognition reaction (Ka = 1015 M–1 ).
Addition of phosphatase inhibitors to the
DOC extraction buffer permits the use of the
DOC-soluble fractions for determination of
the activity of different phosphorylated signaling molecules by using simple immunoblotting techniques and phosphospecific antibodies, making this method more versatile than the
classical DOC solubility assay.
Troubleshooting
Several parameters play critical roles for
success in the quantification of fibronectin
matrix assembly. Incorporation of readily detectable levels of biotinylated fibronectin into
the DOC-insoluble fraction is necessary for accurate quantification. This step depends on the
ability of the cells being studied to organize
matrix, and it can be achieved by optimizing
the duration of the labeling period.
Obtaining a high signal-to-noise ratio after
immunoblotting is also essential for successful quantification of the amounts of biotinylated fibronectin and changes in signaling pathways. This goal can be achieved by: (1) loading
Extracellular
Matrix
10.13.7
Supplement 25
Figure 10.13.1 Determination of matrix assembly by the methods described in this unit. (A)
Primary human fibroblasts were cultured overnight in normal medium, washed with medium without
serum containing 1% BSA, and incubated in the same medium supplemented with 20 µg/ml
biotinylated fibronectin without additional agents (lane 1), with 10 µM ROCK inhibitor Y27632
(lane 2), or with 2 µM lysophosphatidic acid (LPA; lane 3) for 4 hr. Deoxycholate (DOC)-insoluble
and -soluble fractions were resolved on 4% to 12% gradient polyacrylamide gels, transferred to
nitrocellulose membranes, and probed with HRP-conjugated streptavidin for determination of the
amount of the incorporated FN (biotinylated fibronectin). The same membranes were re-probed
with antibodies against vimentin and actin for determination of the efficiency of purification and gel
loading. (B) Samples as in (A) from the DOC-soluble set were assayed by immunoblotting with
antibodies against the activated form of focal adhesion kinase (phospho-FAK), activated form of
mitogen-activated protein kinase (phospho-MAPK) and total MAPK for determination of the effect
of Y27632 and LPA on the these signaling molecules.
Quantification of
Fibronectin
Matrix Assembly
sufficient amounts of proteins from both fractions to ensure trouble-free detection of the biotinylated FN; (2) use of freshly added phosphatase and protease inhibitors to the DOC extraction buffer; (3) use of sufficiently specific
antibodies that recognize the phosphorylated
but not the unphosphorylated forms of the signaling molecules of interest; and (4) following proper techniques for SDS-PAGE and immunoblotting (see UNITS 6.1 & 6.2).
Accurate comparison and quantification by
densitometry of the amounts of biotinylated
FN as well as the phosphorylation levels of the
signaling molecules studied in different samples can be achieved if the detection system
is kept in a linear range. That is, loading two
times the amount of sample should be reflected
in a doubling of signal. The enhanced chemiluminescence (ECL) system should be optimized to obtain linearity by adjustments of the
amount of protein loaded on the gel, concentrations of primary and secondary antibody, and
X-ray film exposure time. If the weakest signal is detectable and the strongest signal is still
within the linear range of the film (e.g., not sat-
urated), then the rest of the samples are also in
the linear range and the results can be used for
quantification.
Anticipated Results
Typical results expected after performing
the Basic Protocol are presented in Figure
10.13.1. Easily detectable amounts of biotinylated FN should be present in the control lanes
(Fig. 10.13.1 A, lane 1) in both DOC-insoluble
and DOC-soluble sets of samples. Different
treatments may have different effects on matrix assembly. In the example presented, blocking cellular contractility with Y27632 strongly
decreased both binding of FN to the cells
(Fig. 10.13.1 A, DOC soluble, lane 2) and
its incorporation into the matrix (Fig. 10.13.1
A, DOC insoluble, lane 2). The opposite effect was observed after stimulation of cellular
contractility with lysophosphatidic acid (LPA;
Fig. 10.13.1 A, lane 3). Calculation of the
differences observed (see Basic Protocol) revealed an 11.5-fold reduction in the incorporation of labeled FN into the DOC-insoluble matrix after treatment with Y27632 and a 4.9-fold
10.13.8
Supplement 25
increase after stimulation with LPA. The reduction in the DOC-soluble fraction after treatment with Y27632 was 5-fold, suggesting that
this agent may affect not only formation of the
matrix, but also initial FN binding to the cell
surface.
The decrease in matrix assembly after inhibition of cellular contractility was accompanied by a reduction in the activation of FAK,
but not of MAPK (Fig. 10.13.1 B, lane 2), suggesting that in this particular experimental setting, the activity of FAK may be important for
matrix assembly. While the effects of agents
such as Y27632 and LPA on matrix formation
are clear, drawing unambiguous conclusions
about the relevant signaling events demands
a number of additional experiments. Nevertheless, such initial correlative data between
signaling and matrix assembly provide a good
starting point.
Time Considerations
The procedure described in the Basic Protocol can be completed in 3 days.
The first day includes the time for cell
attachment after plating (overnight); labeling
with biotinylated FN (4 hr) and isolation of
DOC-soluble and DOC-insoluble fractions
(2 hr). The second day comprises SDS-PAGE
and electrotransfer (3.5 hr for mini gels);
probing with streptavidin-HRP and ECL
reaction (2.5 hr); and overnight incubation
with the primary antibody. The third day is
for completion of the immunoreactions and
ECL processing (4 hr).
There are a number of points where the procedure can be interrupted: (1) after preparation
of the SDS-PAGE samples; (2) after the electrotransfer (membranes can be stored wet or
dry in resealable plastic bags at 4◦ C); and (3)
after completion of the first ECL development
(membranes can be stored wet in resealable
plastic bags at 4◦ C).
The procedure described in the Support
Protocol can be completed in 1 day if the 2hr dialysis period is employed, or 36 hr if the
overnight dialysis is used.
Literature Cited
Coulombe, P.A., Ma, L., Yamada, S., and Wawersik,
M. 2001. Intermediate filaments at a glance.
J. Cell Sci. 114:4345-4347.
Geiger, B., Bershadsky, A., Pankov, R., and Yamada, K.M. 2001. Transmembrane extracellular matrix–cytoskeleton crosstalk. Nat. Rev. Mol.
Cell Biol. 2:793-805.
McKeown-Longo, P.J. and Mosher, D.F. 1983.
Binding of plasma fibronectin to cell layers of
human skin fibroblasts. J. Cell Biol. 97:466-472.
Wierzbicka-Patynowski, I. and Schwarzbauer, J.E.
2002. Regulatory role for SRC and phosphatidylinositol 3-kinase in initiation of fibronectin matrix assembly. J. Biol. Chem.
277:19703-19708.
Contributed by Roumen Pankov and
Kenneth M. Yamada
National Institute of Dental and
Craniofacial Research,
National Institutes of Health
Bethesda, Maryland
Extracellular
Matrix
10.13.9
Supplement 25
Use of Hyaluronan-Derived Hydrogels for
Three-Dimensional Cell Culture and
Tumor Xenografts
UNIT 10.14
Monica A. Serban,1 Anna Scott,2 and Glenn D. Prestwich1
1
Department of Medicinal Chemistry and Center for Therapeutic Biomaterials,
The University of Utah, Salt Lake City, Utah
2
Glycosan BioSystems, Salt Lake City, Utah
ABSTRACT
The practice of in vitro three-dimensional (3-D) cell culture has lagged behind the
realization that classical two-dimensional (2-D) culture on plastic surfaces fails to mirror
normal cell biology. Biologically, a complex network of proteins and proteoglycans that
constitute the extracellular matrix (ECM) surrounds every cell. To recapitulate the normal
cellular behavior, scaffolds (ECM analogs) that reconstitute the essential biological cues
are required. This unit describes the 3-D cell culture and tumor engineering applications
of Extracel, a novel semisynthetic ECM (sECM), based on cross-linked derivatives of
hyaluronan and gelatin. A simplified cell encapsulation and pseudo-3-D culturing (on
top of hydrogels) protocol is provided. In addition, the use of this sECM as a vehicle to
obtain tumor xenografts with improved take rates and tumor growth is presented. These
engineered tumors can be used to evaluate anticancer therapies under physiologically
C 2008 by John Wiley
relevant conditions. Curr. Protoc. Cell Biol. 40:10.14.1-10.14.21. & Sons, Inc.
Keywords: hyaluronan r semisynthetic extracellular matrix r hydrogel r
biodegradable scaffold
INTRODUCTION
This unit describes the use of chemically modified, cross-linkable derivatives of hyaluronan (HA) hydrogels for more physiologically significant in vitro cell culturing and in
vivo tumor and tissue engineering applications. Traditional two-dimensional (2-D) culturing conditions lead to aberrant cell behavior that may have limited relevance to in vivo
conditions (Roskelley et al., 1994; Weaver et al., 1997; Wang et al., 1998; Cukierman
et al., 2001). Mammalian cells do not grow in a physiologically realistic manner on plastic. In vivo, an extracellular matrix (ECM) surrounds the cells in all tissues. The ECM
is a complex network of proteins and glycosaminoglycans (GAGs), which form a 3-D
microenvironment that plays an integral part in signaling cells to proliferate, migrate,
differentiate or invade (Galbraith et al., 1998; Geiger et al., 2001; Lutolf and Hubbell,
2005; Holmbeck and Szabova, 2006).
HA is a major constituent of the ECM and is the only nonsulfated GAG present (Knudson
and Knudson, 2001). It is biocompatible and biodegradable, and it performs important
biological functions such as stabilizing and organizing the ECM (Fraser et al., 1997;
Dowthwaite et al., 1998), regulating cell adhesion and motility (Dowthwaite et al., 1998;
Cheung et al., 1999), and mediating cell proliferation and differentiation (Entwistle et al.,
1996).
The HA-derived hydrogel (Extracel) discussed in this unit is composed of chemically modified HA containing reactive thiol groups (known as CMHA-S, Glycosil, or
Carbylan-S) and chemically modified gelatin containing reactive thiol groups (known as
Current Protocols in Cell Biology 10.14.1-10.14.21, September 2008
Published online September 2008 in Wiley Interscience (www.interscience.wiley.com).
DOI: 10.1002/0471143030.cb1014s40
Copyright Extracellular
Matrix
10.14.1
Supplement 40
Gtn-DTPH or Gelin-S), which are co-cross-linked with polyethylene glycol diacrylate
(PEGDA or Extralink) to form a semisynthetic ECM (sECM; Shu et al., 2004, 2006;
Prestwich, 2007, 2008). For clarity and consistency in this unit, we will use the names
of the commercially available materials.
All three components of the hydrogel are available as lyophilized solids. Once reconstituted, the solutions can be easily pipetted and transferred into multiple formats (including
any well size or tissue culture insert), or it can be injected into an animal model. The
hydrogel is formed by mixing the chemical cross-linker, Extralink, with either Glycosil
only or Glycosil mixed with Gelin-S. Once the cross-linker is added, the mixture will
become more and more viscous until a solid hydrogel is formed. The gelation time can
be controlled by the user depending upon requirements.
The HA component Glycosil can be cross-linked alone with Extralink, but most mature
cell types do not adhere to HA-only hydrogels. Some cancer cells and stem cells will
grow and proliferate in HA-only hydrogels, but usually some attachment factor (e.g.,
gelatin, an RGD peptide, collagen, laminin, or fibronectin) needs to be mixed with the
Glycosil prior to cross-linking with PEGDA. The hydrogel retains proteins greater than
70 kDa in size, so even though the ECM-derived proteins are not covalently attached
to the hydrogel, they are entrapped and will only be able to diffuse out as the sECM
degrades. Growth factors are also retained within the hydrogel in a similar fashion (Cai
et al., 2005; Pike et al., 2006; Riley et al., 2006).
Using three basic components to make an sECM simplifies the biological ECM to
a consistent, fully defined, experimentally controllable material for research. Because
ECM proteins and growth factors can be incorporated into these hydrogels, it is possible
to make a fully defined mimic of specific ECMs found in mammalian tissues if the target
tissue ECM composition is known. Additionally, since the basic hydrogel can be formed
with only Glycosil and Extralink, animal-free hydrogels can also be made.
STRATEGIC PLANNING
For successful use of these protocols, the researcher must be familiar with how to culture
the cells of interest in a 3-D environment or be prepared to conduct several experiments
to determine the optimal conditions. Cells cultured in 3-D behave differently than those
cultured 2-D on tissue culture–treated plastic. At a minimum, the cell morphology and
gene expression patterns can change (Bissell et al., 2003). Because cells receive signals
from the matrix on which they are grown (even if this matrix is plastic), the composition
and stiffness (compliance) of this matrix help determine the growth and functional
characteristics of the cells (Yeung et al., 2005; Engler et al., 2006). In the case of naı̈ve
mesenchymal stem cells, the matrix stiffness can cause lineage restriction. For fibroblasts,
it changes the amount and arrangement of actin stress fibers (Ghosh et al., 2007). For
many cell types, the differences when plating on stiff versus compliant surfaces is not
yet characterized. Finally, cells respond differently when encapsulated within a hydrogel
or when plated on the surface.
Use of
hyaluronanderived hydrogels
for 3-D culture
For in vitro cell growth, the culture medium and cell seeding density are very important.
It is possible to use the optimal tissue culture–plastic culture conditions as a starting
point for the hydrogel experiments. However, it is likely that some modification to these
conditions will be required. For the in vivo tumor xenografts, the cell density, injection
volume, and hydrogel dilution are critical for the experimental outcome (Liu et al.,
2007a). The method of making the hydrogel affects its final properties. There are several
variations to this general protocol which are discussed in subsequent sections. Prior to
using the hydrogels, the following questions need to be addressed:
10.14.2
Supplement 40
1. What gelation time is required?
2. Will cells be encapsulated in the hydrogel?
3. Will ECM proteins be incorporated into the hydrogel?
4. Will growth factors be incorporated into the hydrogel?
5. What hydrogel compliance is required?
Based on the answers to these questions, additional steps may be required in the hydrogel
preparation. If it is your first time using these HA-derived hydrogels, performing a simple
gelation test before starting the first experiment will greatly improve the chances of
success. The test takes ∼1 hour and will allow you to understand fundamentally how the
materials work. A protocol for performing this test is given in Basic Protocol 1.
This unit contains six protocols which detail how to make the HA-derived hydrogels
(Basic Protocol 1), vary their compliance (Basic Protocol 2) and composition (Basic
Protocol 3), use them for cell growth in vitro (Basic Protocols 4 and 5), and implant them
in mice for in vivo experimentation (Basic Protocol 6).
NOTE: All solutions and equipment coming into contact with cells must be sterile, and
proper aseptic technique should be used accordingly.
NOTE: All culture incubations should be performed in a humidified 37◦ C, 5% CO2
incubator unless otherwise specified.
STANDARD HA-DERIVED HYDROGEL PREPARATION
The basic hydrogel, Extracel, is the foundation tool for all the protocols discussed in
this unit. Its preparation is required for 3-D cell and pseudo-3-D culture (encapsulation
and surface growth) and tumor xenograft experiments. Extracel is composed of Glycosil
(thiol-modified HA), Gelin-S (thiol-modified gelatin), Extralink (PEGDA), and degassed,
deionized water (DG Water). Glycosil, Gelin-S, and Extralink are available as lyophilized
solids. They must be reconstituted using DG Water prior to forming the hydrogel. When
reconstituted, they form low-viscosity solutions in phosphate-buffered saline (PBS),
pH ∼7.4. The hydrogel is formed by mixing all three components together. The gelation
time is highly dependent upon the pH of the Extracel solution: the higher the pH, the
faster the gelation time. Additionally, depending upon the amount of Extralink used and
the concentration of the Glycosil and Gelin-S solutions, gelation will occur in 10 min to
>2 hr. Once the Extralink is added there is a time limit on using the hydrogel because it
becomes impossible to pipet after the gelation point is reached.
BASIC
PROTOCOL 1
Materials
7.5-ml Extracel Hydrogel Kit (Glycosan BioSystems) containing:
Glycosil (three 1-ml vials)
Gelin-S (three 1-ml vials)
Extralink (three 0.5-ml vials)
DG Water (one 10-ml vial)
Serum-free cell culture medium
37◦ C water bath
1-ml syringes with long-tip 20-G × 11/2 -in. needles, sterile
37◦ C shaking or rocking incubator
4-ml glass vials
Extracellular
Matrix
10.14.3
Supplement 40
Prepare the gel
1. Remove Glycosil, Gelin-S and Extralink vials from the −20◦ C freezer and heat them
to 37◦ C (∼30 min).
2. Remove the DG Water from the −20◦ C freezer and thaw in a 37◦ C water bath
(∼15 min).
3. Under aseptic conditions and using a syringe with the exact amount of liquid, add
1.0 ml DG Water to the Glycosil vial. Repeat for the Gelin-S vial.
4. Incubate both vials horizontally at 37◦ C, with shaking (for maximum mixing).
NOTE: Vigorous shaking will speed up dissolving time.
It will take <30 min for the solids to fully dissolve. Solutions will be clear and slightly
viscous.
0.5 ml DG Water to the Extralink vial. Invert several times to dissolve.
6. As soon as possible and within 4 hr of making the solutions, mix equal volumes of
Glycosil and Gelin-S in a sterile container. Mix by pipetting up and down gently or
inverting the vial.
7. To form the hydrogel, add Extralink to the Glycosil + Gelin-S mix in a 1:4 volume
ratio (0.25 ml of Extralink to 1.0 ml Glycosil + Gelin-S).
Perform gelation tests with Extracel
8. Follow steps 1 to 5 (above) for standard hydrogel reagent preparation.
9. Add 0.25 ml Glycosil and 0.25 ml Gelin-S to a small glass vial. Pipet up and down
to mix.
10. Add 0.125 ml Extralink to the vial and pipet up and down to mix. Record the time.
The initial solution of Glycosil + Gelin-S + Extralink will be low viscosity (similar to
medium).
11. Every few minutes, invert the vial. Record the time at which the hydrogel no longer
flows when the vial is inverted.
As the hydrogel forms, the liquid will become more viscous.
The gelation time is the difference between the two recorded times. This establishes the
maximum length of time you will have to use Extracel after the Extralink is added.
12. Repeat steps 8 to 11, but in addition, add 0.5 ml PBS to the vial of Glycosil + Gelin-S
from step 9.
This gelation time will be substantially longer due to the dilution of the Glycosil and
Gelin-S.
13. Repeat steps 8 to 11, but in addition, add 50 μl cell culture medium (no serum or
additives) to the vial of Glycosil + Gelin-S in step 9. Pipet up and down to mix.
This gelation time will be about the same as with the first trial (step 11). This simulates
the addition of cells in medium into the hydrogel prior to cross-linking.
Use of
for 3-D culture
These gelation tests will help the user determine the time constraints for working with
Extracel, once the cross-linker Extralink is added to the Glycosil + Gelin-S. They will
also help familiarize the user with how the hydrogel is formed, prior to working with it in
an experiment.
10.14.4
Supplement 40
HA-DERIVED HYDROGEL STIFFNESS VARIATION
As discussed above, hydrogel stiffness can have a dramatic effect on how cells behave
in culture. Using the Extracel Hydrogel Kit as per the standard instructions results in a
hydrogel compliance of ∼100 Pa (J. Vanderhooft, unpub. observ.).
BASIC
PROTOCOL 2
For the HA-derived hydrogels, compliance variation can be achieved in two different
ways: (1) varying the concentration of the cross-linker used and (2) varying the concentrations of the Glycosil and Gelin-S solutions.
By increasing the concentration of Extralink, the compliance can be increased to ∼500
Pa (Ghosh et al., 2007). Alternatively, diluting the Extracel solutions can decrease it to
below the threshold of detection (∼20 Pa). Glycosil-only hydrogels cross-linked with
Extralink are ∼300 Pa. Changing the concentration of Extralink significantly alters the
gelation time, as does diluting the Glycosil and Gelin-S solutions. Doubling the Extralink
concentration will decrease the gelation time by ∼50%. A 2-fold volume dilution will
more than double the time for the hydrogel to form.
Materials
Glycosil (three 1-ml vials)
Gelin-S (three 1-ml vials)
Extralink (three 0.5-ml vials; purchase additional vials separately, if required)
DG Water (one 10-ml vial)
Serum-free cell culture medium
Phosphate-buffered saline (PBS; APPENDIX 2A), pH ∼7.4 and ∼7.6
37◦ C water bath
Method 1: vary cross-linker concentration
1a. Prepare Glycosil and Gelin-S as in Basic Protocol 1, steps 1 to 4.
2a. Under aseptic conditions and using a syringe with the exact amount of liquid, add the
appropriate amount of DG Water to the Extralink vial based on the desired hydrogel
stiffness (see Table 10.14.1). Invert several times to mix.
3a. As soon as possible and within 4 hr of making the solutions, mix equal volumes of
Glycosil and Gelin-S by pipetting up and down.
4a. To form the hydrogel, add Extralink to the Glycosil + Gelin-S mix in a 1:4 volume
ratio (e.g., 0.5 ml Extralink to 2.0 ml Glycosil + Gelin-S).
The gelation time will decrease with the higher Extralink concentration.
For the stiffest hydrogel, there is insufficient Extralink in the standard Extracel 7.5-ml
Hydrogel Kit to use all of the Glycosil and Gelin-S. Individual Extralink vials can be
purchased, if required.
Table 10.14.1 Amounts of PBS Added to Extralink When Preparing Hydrogels of Different Stiffnesses by Cross-linker Concentration Variation
Condition
Volume PBS (ml)
Notes
A
0.25
Stiffest
B
0.5
Standard
C
1.0
Softest
Extracellular
Matrix
10.14.5
Supplement 40
Table 10.14.2 Amounts (ml) of PBS Added to Hydrogel Reagent Solutions When Preparing
Hydrogels of Different Stiffnesses by Hydrogel Component Dilution
Stiffest ————————————————————→ Softest
Standard
A
B
C
D
Glycosil
0.00
0.25
0.50
0.75
1.00
Gelin-S
0.00
0.25
0.50
0.75
1.00
Extralink
0.00
0.13
0.25
0.38
0.50
Method 2: dilute hydrogel solutions
1b. Prepare the hydrogel kit reagents as in Basic Protocol 1, steps 1 to 5 (standard
hydrogel reagent preparation).
2b. Based on the desired stiffness of hydrogel, aseptically add (using a syringe) varying volumes of sterile PBS to the prepared 1-ml Glycosil and Gelin-S vials (see
Table 10.14.2). Invert to mix.
3b. Also add varying amounts of sterile PBS to the prepared 0.5-ml Extralink vial (see
Table 10.14.2). Invert several times to mix.
4b. As soon as possible and within 4 hr of making the solutions, mix equal volumes of
Glycosil and Gelin-S by pipetting up and down.
5b. To form the hydrogel, add Extralink to the Glycosil + Gelin-S mix in a 1:4 volume
ratio (e.g., 0.5 ml Extralink to 2.0 ml Glycosil + Gelin-S).
The gelation time will increase with the solution dilution.
BASIC
PROTOCOL 3
ECM COMPONENT INCORPORATION IN HYDROGELS
Gelin-S provides basic cell attachment sites for cell lines and some primary cell types
(Shu et al., 2006; Prestwich et al., 2007). However, several cell types are dependent upon
specific ECM components for growth and differentiation. For specific cell performance,
matricellular and extracellular proteins (e.g., laminin, collagen, fibronectin, vitronectin,
aggrecan, decorin) may be added to Glycosil-only hydrogels by the user (Mehra et al.,
2006). These proteins are easily incorporated noncovalently into the hydrogel prior to
gel formation and retained there after gel formation because of their size. The following
protocol describes how to prepare Glycosil-only hydrogels mixed with a laminin isoform
from a particular animal source.
Materials
1-ml vial of Glycosil (Glycosan BioSystems)
0.5-ml vial of Extralink (Glycosan BioSystems)
DG Water (Glycosan BioSystems)
500 μg/ml commercial (e.g., Sigma) or laboratory-prepared laminin stock solution
(or other sterile, cellular matrix protein in aqueous solution): prepared according
to the manufacturer’s instructions, if commercially obtained
37◦ C water bath
Use of
for 3-D culture
1. Remove the Glycosil and Extralink vials from the −20◦ C freezer and heat them
to 37◦ C (∼30 min). Thaw the laminin solution (for commercial product, per the
manufacturer’s instructions).
For example, it is necessary to thaw Sigma laminin L6274 overnight.
10.14.6
Supplement 40
(∼15 min).
1.0 ml of DG Water to the Glycosil vial.
4. Place the vial horizontally at 37◦ C, with shaking (for maximum mixing).
It will take <30 min for the solids to fully dissolve. Solution will be clear and slightly
viscous.
0.5 ml of DG Water to the Extralink vial. Invert several times to dissolve.
6. Add 125 μl of commercially obtained or laboratory-prepared laminin to the 1 ml of
Glycosil solution. Mix thoroughly.
7. To form the hydrogel, add Extralink to the Glycosil + laminin mix in a 1:4 volume
ratio (0.25 ml Extralink to 1.0 ml Glycosil + 0.125 ml laminin).
8. Vary the composition of the hydrogel, as desired, as follows:
a. Increase or decrease the amount of laminin.
b. Vary the source of the laminin.
c. Use other ECM proteins (e.g., a specific type of collagen, fibronectin, vitronectin,
decorin) in place of or in conjunction with laminin.
CELL GROWTH ON HA-DERIVED HYDROGEL SURFACE
This protocol describes how to make Extracel hydrogels in a 24-well plate format for
cell growth on the surface. The protocol can easily be adapted for use with 6-, 12-, 48and 96-well plates.
BASIC
PROTOCOL 4
Materials
7.5-ml Extracel Hydrogel kit (Glycosan BioSystems)
Phosphate-buffered saline (PBS; APPENDIX 2A), sterile
1–2 × 104 cells/ml medium suspension of cultured cells of interest: prepared
according to standard procedures (e.g., see UNIT 1.1)
Cell culture medium with serum
0.05% trypsin EDTA (VWR)
10× collagenase/hyaluronidase (StemCell Technologies)
37◦ C water bath
15-ml sterile, conical tubes
24-well tissue culture plates
Sterile plate-sealing film (e.g., Axy Seal, VWR) and roller
Light microscope (10× magnification)
Coat plates
1. Prepare the hydrogel kit reagents as in Basic Protocol 1, steps 1 to 5 (standard
hydrogel reagent preparation).
an additional 1.0 ml of sterile PBS to both the Glycosil and the Gelin-S vials. Shake
to mix.
Extracellular
Matrix
10.14.7
Supplement 40
an additional 0.5 ml of sterile PBS to the Extralink vial. Shake to mix.
4. Transfer Glycosil and Gelin-S solutions into a sterile 15-ml conical tube. Mix for at
least 3 min with a 25-ml pipet by pipetting up and down.
If the Extracel solutions are not well mixed, the hydrogel surface may not be uniform.
This can cause variation in how the cells attach and grow on the hydrogel.
5. Remove a 24-well plate from the packaging.
6. Just before use, add the 1.0 ml Extralink to the tube containing Glycosil + Gelin-S.
Mix at least 2 min by pipetting with a 25-ml pipet.
7. Pipet 500 μl into each of ten wells. Rock the plate by hand to ensure that the surface
of the plate is equally coated.
8. Remove 300 μl from each well using a pipet, leaving 200 μl of hydrogel in each
well. Repeat steps 7 and 8 until all wells are coated.
9. Cover each plate with a sterile film. Seal with a roller so that each well is isolated.
Since these are thin coatings they will dehydrate very easily to form films if they are not
completely sealed.
10. Allow gelation to occur on the bench top.
It will take >2 hr for gelation to occur.
11. Store up to 4 months at 4◦ C until ready for use. Do not freeze.
Culture cells on the HA-derived hydrogel surface
12. Remove a precoated 24-well plate from storage at 4◦ C.
13. Allow it to warm to room temperature or place in the incubator to increase the
temperature to 37◦ C prior to plating.
14. Add 500 μl of the cell suspension in medium to each well on top of the hydrogel.
NOTE: Cells should be cultured in the same medium as when they are grown on tissue
culture–treated plastic. This medium may or may not contain serum, depending upon the
cell type.
Cell seeding density depends upon the experiment and the cell type. As a rough guideline,
follow the cell seeding density used for seeding a tissue culture–treated plastic 24-well
plate.
15. Incubate at least 1 hr in a 37◦ C, 5% CO2 incubator.
16. Verify cell attachment under the microscope. Once confirmed, add the appropriate
amount of medium (0.5 to 1.5 ml) to each well and return the plate to the incubator.
17. Change the medium as required (based on changes in the medium’s phenol red
indicator) by carefully aspirating off the medium.
The hydrogel can easily be removed by vacuum aspiration as well, so this must be done
gently and carefully.
18. Pipet 1 to 2 ml medium into each well. Try to avoid disrupting the gel.
19. Return the plate to the incubator.
Use of
for 3-D culture
Recover cells from hydrogel surface
20. Aspirate the medium and wash the hydrogel surface with 1 to 2 ml PBS per well.
10.14.8
Supplement 40
21. Add 0.5 ml trypsin solution to the hydrogel surface.
Other products (e.g., Accutase, Detachin, TrypLE) that are gentler than trypsin and are
better tolerated by cells can also be used. However, they may also degrade the hydrogel
so that recovered cells carry some hydrogel particles with them. If this occurs, then use a
10× collagenase/hyaluronidase solution to digest the remaining hydrogel.
22. Incubate at 37◦ C until the cells begin to detach (∼15 min).
23. Gently tap the plate to loosen the cells.
24. Add 2 ml medium with serum to the hydrogel surface and pipet up and down to get
a uniform cell suspension.
25. Transfer the cells to a 15-ml culture tube. Add 8 ml medium with serum (10 ml final
volume). Centrifuge the cells 5 min at 120 × g, room temperature.
26. Remove the supernatant and replace with 1 to 2 ml fresh medium.
Cell viability will be similar to that of cells grown on plastic and detached with trypsin.
CELL ENCAPSULATION IN HA-DERIVED HYDROGELS
Encapsulating cells in HA-derived hydrogels and growing them in tissue culture inserts
is the best way (in the absence of a bioreactor) to mimic in vivo conditions in vitro.
This protocol describes how to make Extracel hydrogels in a 24-well plate format, using
tissue culture inserts. Other insert formats also work, but the amount of hydrogel used
per insert should be varied based on the insert volume.
BASIC
PROTOCOL 5
It is not always necessary to recover cells from the hydrogels. Cells cultured by encapsulating them in tissue culture inserts can be treated like tissue. The hydrogel can be
removed from the insert, embedded in paraffin, sectioned, and stained as per standard
protocols. Note that small molecule dyes and stains that are less than 70 kDa in size will
freely diffuse into the gel.
It is not possible to perform direct antibody staining of cells encapsulated in HA-derived
hydrogels because the antibodies are too large to permeate the gel. If embedding, sectioning, and staining is not desirable, then the cells must be recovered from the hydrogel.
Materials
10× collagenase/hyaluronidase (StemCell Technologies)
Sterile phosphate-buffered saline (PBS)
∼0.4–2 × 104 cells/ml medium suspension of cultured cells of interest: prepared
according to standard procedures (e.g., see UNIT 1.1)
Cell culture medium with and without serum
37◦ C water bath
24-well plate with tissue culture inserts (e.g., 6.5-mm Costar tissue culture–treated
polycarbonate membrane polystyrene plates, Corning; 8.0-μm pore size
Millicel, Millipore
35-mm sterile petri dishes
Surgical scalpel
15-ml conical centrifuge tube
Extracellular
Matrix
10.14.9
Supplement 40
Encapsulate cells
1. Prepare hydrogel kit reagents as in Basic Protocol 1, steps 1 to 5 (standard hydrogel
reagent preparation).
2. Determine the volume of suspension required to obtain the desired seeding density
in 2.5 ml of the hydrogel.
Seeding density varies with cell type, but a typical range is 10,000 to 50,000 cells per
insert.
3. Prepare two 24-well plates with tissue culture inserts by removing them from their
sterile packaging.
4. Mix 1.0 ml of Glycosil and 1.0 ml of Gelin-S.
5. Centrifuge the volume determined in step 2 for 5 min at 120 × g, room temperature,
and discard the supernatant. Resuspend the cell pellet in the 2 ml of Glycosil +
Gelin-S.
6. Just before pouring the hydrogels, add 0.5 ml of Extralink to Glycosil + Gelin-S
with cells. Mix completely by pipetting up and down.
Once the Extralink is added you have <20 min before the hydrogel forms.
7. Quickly pipet 100 μl of Extracel mix into each insert.
Do not add medium at this point because this will dilute the hydrogel and prevent it from
gelling.
8. Incubate the plates for ∼1 hr in a 37◦ C, 5% CO2 incubator to allow the Extracel to
gel.
9. Remove the plates from incubator and verify that the hydrogel is solid. If so, add
1.8 ml medium (with serum, if required) to each well. Incubate in a 37◦ C, 5% CO2
incubator.
10. Change the medium as required:
a. Move each tissue culture insert to an adjacent empty well.
b. Aspirate the medium.
c. Tap each insert carefully to remove the medium above the hydrogel in the insert.
Aspiration can be used to remove the medium, but, the gel can also easily be removed by
vacuum aspiration, so this must be done gently and carefully.
d. Replace the insert into its original well.
e. Slowly and carefully pipet 1.8 ml medium into each well.
Try to avoid disrupting the gel.
f. Return the plate to the 37◦ C, 5% CO2 incubator.
Cells behave differently when cultured in 3-D than when grown on the surface of a
hydrogel or tissue culture–treated plastic. The cells will grow at different rates (typically
slower) and have different morphologies (depending upon the hydrogel stiffness and
composition). Additionally, the cells are not passaged in the traditional manner. Since the
volume of the hydrogel provides a large volume for growth, the cultures can be maintained
for many days, even weeks, before the cells become confluent.
Use of
for 3-D culture
Recover encapsulated cells
11. Dilute the 10× collagenase/hyaluronidase 10-fold in the cell culture medium (without serum) used to cultivate the cells.
Do not use undiluted enzyme because this results in low cell viability.
10.14.10
Supplement 40
If using medium that contains serum for culture, make sure to wash the hydrogels with
serum-free medium or PBS before starting the digestion process because the serum will
inactivate the enzymes. At a minimum, wash hydrogels twice for 1 hr to clear serum).
12. Remove the tissue culture insert from the 24-well culture plate. Place upside down
in a petri dish.
13. Remove the membrane by using a surgical scalpel to cut it loose from the insert.
The membrane will stay attached to the insert, but usually flips up out of the way.
14. Turn the insert right side up and using the back of a 10-μl pipet tip punch the hydrogel
out of the insert into the petri dish.
15. Place the hydrogel in a 15-ml conical tube and add 5 ml diluted collagenase/hyaluronidase solution to the hydrogel for each 100 μl of hydrogel.
16. Incubate overnight at 37◦ C, with gentle shaking.
At the end of the incubation there will still be some hydrogel left in the tube.
If the 10-fold dilution of 10× collagenase/hyaluronidase is not satisfactory, try a 5-fold
dilution with digestion overnight.
Be cautious about mechanically breaking up the hydrogel prior to digestion because this
can lower cell viability significantly.
17. Centrifuge in the conical tube 5 min at 120 × g, room temperature.
18. Aspirate and discard the supernatant.
Wash the cells
19. Add 5 ml PBS to wash the cell pellet.
20. Repeat steps 17 and 18.
21. Resuspend the cell pellet in 5 ml PBS.
NOTE: In the PBS you can see any remaining hydrogel.
22. Centrifuge cell suspension 15 min at 120 × g, room temperature.
23. Aspirate and discard the supernatant.
24. Add 5 ml medium and repeat the centrifugation.
25. Aspirate off all medium but ∼0.5 ml and resuspend pellet (in the remaining 0.5 ml)
in the desired volume of cell culture medium.
HA-DERIVED HYDROGELS FOR TUMOR XENOGRAFTS
Clinically relevant cancer models are necessary to improve our ability to successfully
treat the disease. Anticancer drug discovery efforts require models that can predictably
translate preclinical results to efficacy in human patients. Most commonly used are the
human tumor xenograft models, where human cancer cells are injected into immune
compromised mice. Typically these cells are injected in serum-free medium or buffer or
Matrigel (a tumor-derived basement membrane extract). Poor ‘‘take’’ is often a problem,
and many cell lines or patient-derived cells will not form tumors by injection in buffer
or medium.
HA-derived hydrogels can be used for the delivery and growth of cancer cells in vivo for
the growth of orthotopic and subcutaneous tumors. Using a hydrogel to deliver cancer
cells can offer several advantages (Liu et al., 2007a):
BASIC
PROTOCOL 6
Extracellular
Matrix
10.14.11
Supplement 40
The incidence of cancer formation is increased and variability in tumor size is
reduced.
The growth of organ-specific cancers is enhanced with improved tumor-tissue
integration.
Vascularization is increased and necrosis is reduced in tumors.
Cancer seeding on adjacent tissues or organs is minimized.
The general animal health is improved, leading to better data with fewer animals.
Use the pilot study described below to determine the optimal:
Extracel dilution factor
Cell density
Hydrogel injection volume
Coordination of surgical or injection manipulations with hydrogel handling.
The protocol provided below is based on nine mice with two subcutaneous injections
each (see Table 10.14.3), where each experimental condition (cell density and injection
volume) has an “n = 3” (see Table 10.14.4). Please adjust this protocol (cell density
and injection volume, especially) as required, based on experimental requirements and
experience.
NOTE: These guidelines describe how to prepare Extracel hydrogels for encapsulation
of cancer cells and injection of this suspension into experimental animals for research
purposes only.
NOTE: Researchers are responsible for obtaining a valid Institutional Animal Care and
Use Committee (IACUC) protocol prior to initiation of any experiments (if applicable).
The guidelines below only pertain to the operational use of the Extracel product in order
to assist in preparing an IACUC protocol.
Table 10.14.3 Composition of Injections Mixes (μl)
Glycosil Gelin-S
Cells +
Extralink medium
Total
volume
Six 100-μl injections
Injection 1: 90% Extracel + 10% cell
suspension
250
250
125
63
688
suspension
130
130
65
325
650
suspension
120
120
60
30
330
suspension
70
70
35
175
350
Three 200-μl injections
Table 10.14.4 Pilot Study Conditions
Use of
for 3-D culture
Cells/ml
Injection volume (μl)
Mice 1-3
5 × 106
100
Mice 4-6
5 × 10
100
Mice 7-9
5 × 10
200
7
7
10.14.12
Supplement 40
NOTE: We recommend conducting a benchtop study with Extracel to confirm the Extracel
characteristics prior to initiating animal experiments and gain familiarity with handling
and timing of use. The gelation time and final hydrogel properties are highly dependent
upon the medium used, extent of hydrogel dilution, and final hydrogel pH (see Basic
Protocol 1, steps 8 to 13).
NOTE: We strongly urge researchers to conduct pilot animal studies to optimize experimental conditions and familiarize the researcher with the handling of Extracel prior to
doing large-scale animal testing. The pilot study will provide important information on
the time course for tumor growth from a given cell line or primary tumor source, optimal
injection size, cell concentration, and Extracel dilution.
Materials
Extracel Hydrogel kit (Glycosan BioSystems) containing:
Glycosil
Gelin-S
Extralink
DG Water
Tumor cells
Cell culture medium (without serum)
Research animals
Iodine and 70% (v/v) ethanol and sterile swabs
37◦ C water bath
Prepare hydrogels
1. Remove Glycosil, Gelin-S, and Extralink vials from the −20◦ C freezer and heat
them to 37◦ C (∼30 min).
(∼15 min).
1.0 ml of DG Water to the Glycosil vial. Repeat for the Gelin-S vial.
4. Place both vials horizontally at 37◦ C, with shaking (for maximum mixing).
It will take <30 min for the solids to fully dissolve. Solutions will be clear and slightly
viscous.
0.5 ml DG Water to the Extralink vial. Invert several times to dissolve.
Prepare cells
6. Prepare cells for encapsulation by resuspending them in the relevant sterile cell
culture medium (without serum) to the appropriate cell density and volume (5 ×
107 cells/ml for 100-μl and 200-μl injections and 5 × 106 cells/ml for 100-μl
injections).
This protocol assumes that a suspension of 100 μl or 200 μl of Extracel + cells will be
injected into nine research animals.
The cell loading and amount of injected Extracel hydrogel used depends upon the application. The amounts discussed in these guidelines are based on published tumor xenograft
experiments (Liu et al., 2007a; Prestwich et al., 2007; Prestwich 2008), where a cell concentration of 5 × 107 cells/ml was employed. Lower concentrations may also be effective;
however, they will require longer tumor formation times.
Extracellular
Matrix
10.14.13
Supplement 40
Prepare animals and inject cell suspensions
7. Prepare the research animals for surgery as dictated by an approved IACUC protocol
and sterilize the sites for surgery with iodine and alcohol swabs.
For subcutaneous injections, the hydrogel with cells is injected under the skin. It is possible
to perform two injections per animal, one on each side.
For orthotopic (surgically implanted) injections, the animal is opened surgically and the
hydrogel with cells injected into the desired location (e.g., onto the pancreas).
8. Add the appropriate volume of cell suspension to the appropriate amount of Glycosil
+ Gelin-S (see Tables 10.14.4). Mix the resulting suspension by gently vortexing or
pipetting.
The exact time for the hydrogel to become viscous and gel depends on the dilution factor
of Extracel and the pH value of the hydrogel solution.
The pH of medium used to dilute the Extracel and the dilution factor can profoundly
affect the gelation time. As provided by the manufacturer, the gelation time is ∼20 min
at ambient temperature. However, the greater the dilution factor, the longer the gelation
time.
The pH of Extracel as provided by the manufacturer is controlled to be approximately
7.4 prior to cell encapsulation and further dilution. However, the cell culture medium
used can increase or decrease the pH and change the gelation time. For Extracel, a higher
pH results in a faster gelation time. For multiple injections, many researchers desire a
slower gelation time of 60 or more min. Lowering the pH by 0.1 or 0.2 units, to pH 7.3
or 7.2, combined with dilution with medium, allows researchers to identify an optimal
pH/dilution condition for their specific operational needs.
If stiffer hydrogels are required, increase the concentration of Extralink used or decrease
the subsequent dilution factor (or resuspend the initial lyophilized Extracel components
in half of the indicated water amounts).
Extracel hydrogels form by the reaction of thiols in Glycosil and Gelin-S with the acrylate
groups of the cross-linker Extralink. Both Glycosil and Gelin-S can form hydrogels in the
absence of Extralink via disulfide bond formation; however, this reaction is normally very
slow (hours instead of minutes).
9. When the animals are ready for injection of the hydrogel, add the appropriate amount
of Extralink to the cells + Glycosil + Gelin-S. Mix the resulting suspension by gently
vortexing or pipetting.
Once the Extralink is added to the Glycosil + Gelin-S + cells, you have between 20 min
and 2 hr before the hydrogel forms. Prepare accordingly. If you cannot inject all the
animals within this amount of time, consider dispensing aliquots of the cells + Glycosil +
Gelin-S into individual injection amounts and adding the Extralink just prior to injection
into each animal.
10. Draw the Extracel + cells into a sterile 1-ml syringe with a 20-G needle.
11. Inject the required amount of hydrogel into the research animal at the desired location.
12. After injection, properly care for the research animals and monitor for tumor
formation.
COMMENTARY
Use of
for 3-D culture
Mammalian tissues are composed of a conglomerate of interconnected cells that perform
similar functions within an organism. Cells can
interact with each other directly or indirectly,
and their activity is modulated by autocrine
and paracrine regulatory mechanisms. In epithelial tissues, cells are in close contact with
each other. The majority of other tissue types
are comprised of cells that are surrounded by a
complex network of macromolecules and proteins referred to as the ECM.
10.14.14
Supplement 40
Cell culture is a vital tool for basic research
in cell biology, drug discovery, drug evaluation
processes, and protein biotechnology. Classical tissue culturing techniques were recently
proven to be poor mimics of the physiological cellular environment (Bissel et al., 2003,
2005). Currently, two types of 3-D culturing
methods are commonly used. One is referred
to as 3-D “embedded” cell culture, while the
other is known as 3-D “on-top” (Lee et al.,
2007). Both methods require an extracellular
matrix (ECM) equivalent as the 3-D culturing
microenvironment.
At present, the leading ECM equivalent employed for 3-D culture is Matrigel. This is a
natural, murine sarcoma–derived product. Its
composition includes laminin, collagen, entactin, and growth factors. Matrigel was tested
in numerous 3-D cell culturing applications,
invasion assays, and tumor xenografts and
yielded satisfactory results (Kleinman et al.,
1986; UNIT 12.2). Nonetheless, Matrigel has
drawbacks, the most serious of which pertain
to difficulty of use, lack of experimental control of composition, batch-to-batch variability,
and lack of utility in translational research for
cell therapy (Prestwich, 2007).
A different natural ECM analog, PureCol
(consisting of purified type I collagen;
Nutracon, http://www.purecol.nu), is widely
used in cell culture and tissue engineering
and as a coating material for medical devices (Elsdale and Bard, 1972; Emerman and
Pitelka, 1977; Bell et al., 1979; Schor et al.,
1982; Weinberg and Bell, 1986). PureCol has
long gelation times (45 to 60 min at 37◦ C)
that make this material unsuitable as a vehicle
for 3-D applications. For 3-D encapsulation,
the gelation time of PureCol is such that the
cells will settle by gravity prior to gelation, so
they are not suspended throughout the hydrogel. However, this material is easy to use, has a
very long history in cell culture, and is suitable
for pseudo-3-D plate coating.
Although naturally derived ECM extracts
provide biological recognition and meet key
requirements such as presentation of receptor
binding ligands and cell-induced proteolytic
degradations, they are far from ideal. Issues
of limited availability, batch-to-batch variability, pathogen transmission, immunogenicity,
technical challenges in handling, and the inability to customize composition and compliance opened the door for a new generation of
semisynthetic ECM equivalents.
One such commercially available material
is PuraMatrix, a synthetic self-assembling
peptide-based material that forms fibrous
scaffolds which can be used for 3-D cell
embedding or surface plating (Zhang et al.,
1995; Holmes et al., 2000; Semino et al.,
2003; Bokhari et al., 2005; Yamaoka et al.,
2006). This nonanimal-derived material is
nonimmunogenic and is suitable for in vivo
studies. A major weakness of this material is its
preparation protocol; the pH of the initial
reagent is 3.0, which strictly limits the
time of cell exposure to this environment.
Furthermore, the gelation procedure for this
material requires extensive handling. For
example, the medium needs to be changed
three times in 30 min. Increased handling
increases the risk of cell culture contamination
and thus limits use to small-scale experimental
protocols.
In this unit, we introduced an sECM
known commercially as Extracel, a hydrogel based on chemically modified hyaluronan
(Glycosil) and gelatin (Gelin-S) that are cocross-linked with polyethylene glycol diacrylate (Extralink). A generic synthetic scheme
for this scaffold is presented in Figure 10.14.1.
This biomaterial sustains cell growth and proliferation, while eliminating many of the issues posed by other biomaterials. Its preparation protocol is very user friendly and cell
friendly and is suitable for large-scale experimental protocols. The gelation times can
be adjusted by varying pH or temperature,
and the compliance (stiffness) can be altered by adjusting the degree of cross-linking
(Ghosh et al., 2007). In addition, its nature overcomes the issue of immunogenicity in in vivo applications (Liu et al., 2004,
2006a,b, 2007a,b; Duflo et al., 2006; Shu
et al., 2006; Orlandi et al., 2007; Prestwich
et al., 2007). The biological performance of the
four aforementioned ECM equivalents both
in pseudo-3-D and 3-D cell cultures were recently compared and contrasted (Serban et al.,
2008).
Critical Parameters
The critical parameters required for experimental success were mentioned in each of the
protocols. Below, we briefly summarize four
key factors that can affect the experimental
results:
Extracellular
Matrix
10.14.15
Supplement 40
10.14.16
Supplement 40
Figure 10.14.1 Generic synthetic scheme for Extracel. Extracel is composed of CMHA-S [thiol-modified hyaluronic acid (HA), trade name of Glycosil], Gtn-DTPH
(thiol-modified gelatin, trade name of Gelin-S) and PEGDA (polyethylene glycol diacrylate, trade name of Extralink). In this schematic, “linker” refers to the PEGDA
molecule. When mixed together, PEGDA chemically cross-links CMHA-S and Gtn-DTPH to form a hydrogel. NOTE: each CMHA-S and Gtn-DTPH molecule has multiple
modification sites so that the covalent bonds are formed many times on each HA and gelatin molecule. Reprinted from Methods, Vol. 45, Serban, M.A. and Prestwich,
G.D., Modular extracellular matrices: Solutions for the puzzle, Copyright 2008 with permission from Elsevier.
Table 10.14.5 Troubleshooting Guide to Working with Hydrogels
Problem
Possible cause
Solution
Hydrogel sets too
quickly
High solution pH
Adjust solution pH to ∼7.4
Extensive handling time
Dilute solutions
High solution concentration Aliquot gel components and cross-link
near time of use
Hydrogel sets too
slowly
Low solution pH
Low solution concentration
Reconstitute the lyophilized
compounds with less water
Encapsulated cells
settle to bottom
High solution dilution
Reconstitute the lyophilized
compounds with less water
Improper cross-linker-to-gel Optimize cross-linker-to-gel
components ratio
components ratio
Add cell suspension to the hydrogel
only when mix is becoming viscous
Tumor formation not
optimal
Improper solution pH
Improper solution
concentration
Adjust solution concentrations
Improper cross-linker-to-gel Run a pilot, bench-top experiment to
components ratio
determine optimal hydrogel formulation
based on experimental needs
Setup time
Solution pH
Solution dilution factor
Cell seeding density.
The last three factors mentioned can be customized to fit experimental requirements.
The duration of material handling is dictated by the chosen properties of the Extracel
components (i.e., higher solution pH leads to
faster gelation or lower dilution factor causes
faster gelation). Although the protocols provided here are intended to serve as a general
guide for experimental setup, it is important
to recognize that individual cell types and
lines might require optimization. For instance,
human tracheal scar fibroblasts were found
to prefer a gelatin-rich formulation of Extracel (Serban et al., 2008). Based on individual experimental needs, benchtop studies
should be conducted to customize the protocols in order to fit the researcher’s needs.
These trials should only take a short time
(a few hours) and can ensure experimental
success.
The cell seeding density should be adjusted
accordingly, especially when cell will be 3D encapsulated. It is important to differentiate between surface (2-D) versus embedded
(3-D) culturing. To extrapolate an initial 3-D
cell seeding density if the 2-D seeding number
is known, simply tripling the cell number is
a good starting point. Then, work from this
cell density to optimize the cell density for a
particular experiment. Using classical analytical methods, cell proliferation or viability for
both pseudo-3-D or 3-D culturing conditions
can easily be determined. Colorimetric (MTS)
assays and staining procedures such as fluorescein diacetate/propidium iodide (FDA/PI)
or hematoxylin and eosin (H&E), are perfectly
compatible with Extracel.
Troubleshooting
See Table 10.14.5 for troubleshooting hints
for these protocols.
Anticipated Results
For cells grown on the surface of Extracel
hydrogels (Basic Protocol 4), you should notice cell attachment in ∼2 hr. Cells will elicit
a morphology consistent with the hydrogel
on which they are grown (Fig. 10.14.2). Cell
viability should be similar to the classical 2-D
culturing conditions.
Cells that are encapsulated in Extracel hydrogels should be homogeneously distributed
in the hydrogel in a 3-D environment (you can
Extracellular
Matrix
10.14.17
Supplement 40
Figure 10.14.2 T31 human tracheal scar AM/Ethidium fibroblasts grown on Extracel. Reprinted
from Acta Biomater., Vol. 4, Serban, M.A., Liu, Y. and Prestwich, G.D., Effects of extracellular matrix
analogues on primary human fibroblast behavior, pp. 67-75, Copyright 2008 with permission from
Elsevier.
Figure 10.14.3 Calcein-homodimer-1 staining of Extracel-embedded T31 fibroblasts (M.A.
Serban, Y. Lue, and G.D. Prestwich unpub. observ.) For color version of this figure see
http://www.currentprotocols.com.
Use of
for 3-D culture
monitor this microscopically by changing the
focal planes; see Fig. 10.14.3). Cell viability
should be similar to the classical 2-D culturing
conditions.
For tumor xenografts (Basic Protocol 6),
both subcutaneous and orthotopic (surgically
implanted) injections should result in well
localized, vascularized, and differentiated tumors (Fig. 10.14.4 and Fig. 10.14.5). The
use of Extracel as a delivery vehicle for
tumor engineering leads to increased incidence of cancer formation, reduced variability in tumor size, enhanced growth of
organ-specific cancers, improved vascularization, and lower occurrence of core necrosis and adjacent cancer seeding (Liu et al.,
2007a).
Time Considerations
The time considerations for hydrogel handling for each of the six protocols were discussed during the process description. Gelling
and incubation times are specific to the
applications.
Conflict of Interest Statements
Glenn, D. Prestwich is Chief Scientific
Officer and equity holder as cofounder for Glycosan BioSystems, Inc., and Senior Scientific
Advisor and equity holder as cofounder for
Carbylan BioSurgery, Inc., and Sentrx Animal
Care, Inc.
Anna Scott is the Director of Operations
and equity holder as cofounder for Glycosan
BioSystems, Inc.
10.14.18
Supplement 40
Figure 10.14.4 Gross view of breast tumors 4 weeks after subcutaneous injection of breast
cancer cells in Extracel (reprinted from Liu et al., 2007a).
Figure 10.14.5 Gross view of colon tumors 4 weeks after subserous injection of colon cancer
cells in Extracel (reprinted from Liu et al., 2007a).
Literature Cited
Bell, E., Ivarsson, B., and Merrill, C. 1979. Production of a tissue-like structure by contraction
of collagen lattices by human fibroblasts of different proliferative potential in vitro. Proc. Natl.
Acad. Sci. U.S.A. 76:1274-1278.
Bissell, M.J, Rizki, A., and Mian, I.S. 2003. Tissue
architecture: The ultimate regulator of breast epithelial function. Curr. Opin. Cell Biol. 15:753762.
Bissell, M.J, Kenny, P.A., and Radisky, D.C. 2005.
Microenvironmental regulators of tissue structure and function also regulate tumor induction
and progression: The role of the extracellular
matrix and its degrading enzymes. Cold Spring
Harb. Symp. Quant. Biol. 70:343-356.
Bokhari, M.A., Akay, G., Zhang, S., and Birch,
M.A. 2005. The enhancement of osteoblast
growth and differentiation in vitro on a peptide hydrogel-polyHIPE polymer hybrid material. Biomaterials 26:5198-208.
Cai, S., Liu, Y., Shu, X.Z., and Prestwich, G.D.
2005. Injectable glycosaminoglycan hydrogels
for controlled release of human basic fibroblast growth factor. Biomaterials 26;60546067.
Cheung, W.F., Crue, T.F., and Turley, E.A. 1999.
Receptor for hyaluronan mediated motility
Extracellular
Matrix
10.14.19
Supplement 40
(RHAMM), a hyaladherin that regulates cell responses to growth factors. Biochem. Soc. Trans.
27:135-142.
Cukierman, E., Pankov, R., Stevens, D.R., and
Yamada, K.M. 2001. Taking cell-matrix adhesions to the third dimension. Science 294:17081712.
Dowthwaite, G.P., Edwards, J.C., and Pitsillides,
A.A. 1998. An essential role for the interaction
between hyaluronan and hyaluronan-binding
proteins during joint development. J. Histochem.
Cytochem. 46:641-651.
Duflo, S., Thibeault, S.L., Li, W., Shu, X.Z., and
Prestwich, G.D. 2006. Vocal fold tissue repair
in vivo using a synthetic extracellular matrix.
Tissue Eng. 12:2171-2180.
Elsdale, T. and Bard, J. 1972. Collagen substrata
for studies on cell behavior. J. Cell Biol. 54:626637.
Emerman, J.T. and Pitelka, D.R. 1977. Maintenance
and induction of morphological differentiation
in dissociated mammary epithelium on floating
collagen membranes. In Vitro 13:316-328.
Engler, A.J., Sen, S., Sweeney, H.L., and Discher,
D.E.,2006. Matrix elasticity directs stem cell lineage specification. Cell 126:677-689.
Entwistle, J., Hall, C.L., and Turley, E.A. 1996.
Hyaluronan receptors: Regulators of signalling
to the cytoskeleton. J. Cell. Biochem. 61:569577.
Fraser, J.R., Laurent, T.C., and Laurent, U.B. 1997.
Hyaluronan: Its nature, distribution, functions
and turnover.. J. Intern. Med. 242:27-33.
Galbraith, C.G. and Sheetz, M.P. 1998. Forces
on adhesive contacts affect cell function. Curr.
Opin. Cell Biol. 10:566-571.
Geiger, B., Bershadsky, A., Pankov, R., and
Yamada, K.M. 2001. Transmembrane crosstalk
between the extracellular matrix-cytoskeleton
crosstalk. Nat. Rev. Mol. Cell Biol. 2:793-805.
Ghosh, K., Pan, Z., Guan, E., Ge, S., Liu, Y.,
Nakamura, T., Ren, X., Rafailovich, M., and
Clark, R. 2007. Cell adaptation to a physiologically relevant ECM mimic with different viscoelastic properties. Biomaterials 28:671-679.
Holmbeck, K. and Szabova, L. 2006. Aspects of
extracellular matrix remodeling in development
and disease. Birth Defects Res. C Embryo Today
78:11-23.
Holmes, T.C., de Lacalle, S., Su, X., Liu, G., Rich,
A., and Zhang, S. 2000. Extensive neurite outgrowth and active synapse formation on selfassembling peptide scaffolds. Proc. Natl. Acad.
Sci. U.S.A. 97:6728-6733.
Kleinman, H.G., McGarvey, M.L., Hassel, J.R.,
Star, V.L., Cannon, F.B., Laurie, G.W. and
Martin, G.R. 1986. Basement membrane complexes with biological activity. Biochemistry
25:312-318.
Use of
for 3-D culture
Knudson, C.B. and Knudson, W. 2001. Cartilage
proteoglycans. Semin. Cell Dev. Biol. 12:6978.
Lee, G.Y., Kenny, P.A., Lee, E.H., and Bissell, M.J.
2007. Three-dimensional culture models of normal and malignant breast epithelial cells. Nat.
Methods. 4:359-365.
Liu, Y., Shu, X. Z., Gray, S. D., and Prestwich,
G. D. 2004. Disulfide-cross-linked hyaluronangelatin sponge: Growth of fibrous tissue in vivo.
J. Biomed. Mater. Res. A 68:142-149.
Liu, Y., Ahmad, S., Shu, X.Z., Sanders, R.K.,
Kopesec, S.A., and Prestwich, G.D. 2006a. Accelerated repair of cortical bone defects using a synthetic extracellular matrix to deliver
human demineralized bone matrix. J. Orthop.
Res. 24:1454-1462.
Liu, Y., Shu, X.Z., and Prestwich, G.D. 2006b. Osteochondral defect repair with autologous bone
marrow-derived mesenchymal stem cells in an
injectable, in situ, cross-linked synthetic extracellular matrix. Tissue Eng. 12;3405-3416.
Liu, Y., Shu, X.Z., and Prestwich, G.D. 2007a. Tumor engineering: Orthotopic cancer models in
mice using cell-loaded, injectable, cross-linked
hyaluronan-derived hydrogels. Tissue Engineering. 13;1091-1110.
Liu, Y., Shu, X.Z., and Prestwich, G.D. 2007b.
Reduced post-operative intra-abdominal adhesions using Carbylan-SX, a semisynthetic glycosaminoglycan hydrogel. Fertil. Steril. 87:940948.
Lutolf, M.P. and Hubbell, J.A. 2005. Synthetic biomaterials as instructive extracellular microenvironments for morphogenesis in tissue engineering. Nat. Biotechnol. 23:47-55.
Mehra, T., Ghosh, K., Shu, X.A., Prestwich, G.D.
and Clark, R.F. 2006. Molecular stenting with
a cross-linked hyaluronan derivative inhibits
collagen gel contraction. J. Invest. Dermatol.
126:2202-2209.
Orlandi, R. R., Shu, X. Z., McGill, L., Petersen, E.,
and Prestwich, G. D. 2007. Structural variations
in a single hyaluronan derivative significantly alter wound-healing effects in the rabbit maxillary
sinus. Laryngoscope 17:1288-1295.
Pike, D.B., Cai, S., Pomraning, K.R., Firpo, M.,
Fisher, R.J., Shu, X.Z., Prestwich, G.D., and
Peattie, R. 2006. Heparin-regulated release of
growth factors in vitro and angiogenic response in vivo to implanted hyaluronan hydrogels containing VEGF and bFGF. Biomaterials.
27:5242-5251.
Prestwich, G.D. 2007. Simplifying the extracellular
matrix for 3-D cell culture and tissue engineering: A pragmatic approach. J. Cell. Biochem.
101:1370-1383.
Prestwich, G.D. 2008. Evaluating drug efficacy and
toxicology in three dimensions: Using synthetic
extracellular matrices in drug discovery. Acc.
Chem. Res. 41:139-48
Prestwich, G.D., Liu, Y., Yu, B., Shu, X.Z., and
Scott, A. 2007. 3-D culture in synthetic extracellular matrices: New tissue models for drug
toxicology and cancer drug discovery. Adv. Enzyme Regul. 47:196-207.
10.14.20
Supplement 40
Riley, C.M., Fuegy, P.W., Firpo, M.A., Shu, X.Z.,
Prestwich, G.D., and Peattie, R.A. 2006. Stimulation of in vivo angiogenesis using dual growth
factor-loaded cross-linking glycosaminoglycan
hydrogels. Biomaterials. 27:5935-5943.
Roskelley, C.D., Desprez, P.Y., and Bissell, M.J.
1994. Extracellular matrix-dependent tissuespecific gene expression in mammary epithelial
cells requires both physical and biochemical signal transduction. Proc. Natl. Acad. Sci. U.S.A.
91:12378-12382.
Schor, S.L., Schor, A.M., Winn, B., and Rushton,
G. 1982. The use of three-dimensional collagen gels for the study of tumour cell invasion
in vitro: Experimental parameters influencing
cell migration into the gel matrix. Int. J. Cancer
29:57-62.
Semino, C.E., Merok, J.R., Crane, G.G.,
Panagiotakos, G., and Zhang, S. 2003. Functional differentiation of hepatocyte-like spheroid
structures from putative liver progenitor cells in
three-dimensional peptide scaffolds. Differentiation 71:262-270.
Serban, M.A., Liu, Y., and Prestwich, G.D. 2008.
Effects of extracellular matrix analogues on primary human fibroblast behavior. Acta Biomater.
4:67-75.
Shu,X.Z., Liu, Y., Palumbo, F., Luo, Y., and
Prestwich, G.D. 2004. In situ cross-linkable
hyaluronan hydrogels for tissue engineering.
Biomaterials 25:1339-1348.
Shu, X.Z., Ahmad, S., Liu, Y., and Prestwich, G.D.
2006. Synthesis and evaluation of injectable, in
situ cross-linkable synthetic extracellular matrices for tissue engineering. J. Biomed. Mater. Res.
A 79:902-912.
Wang, F., Weaver, V.M., Petersen, O.W., Larabell,
C.A., Dedhar, S., Briand, P., Lupu, R., and
Bissell, M.J. 1998. Reciprocal interactions between beta 1-integrin and epidermal growth
factor receptor in three-dimensional basement
membrane breast cultures: A different perspective in epithelial biology. Proc. Natl. Acad. Sci.
U.S.A. 95:14821-14826.
Weaver, V.M., Petersen, O.W., Wang, F., Larabell,
C.A., Briand, P., Damsky, C., and Bissell, M.J.
1997. Reversion of the malignant phenotype of
human breast cells in three-dimensional culture
and in vivo by integrin blocking antibodies. J.
Cell. Biol. 137:231-245.
Weinberg, C.B. and Bell, E. 1986. A blood vessel
model constructed from collagen and cultured
vascular cells. Science 231:397-400.
Yamaoka, H., Asato, H., Ogasawara, T., Nishizawa,
S., Takahashi, T., Nakatsuka, T., Koshima, I.,
Nakamura, K., Kawaguchi, H., Chung, U.I.,
Takato, T., and Hoshi, K. 2006. Cartilage tissue engineering using human auricular chondrocytes embedded in different hydrogel materials.
J. Biomed. Mater. Res. A 78:1-11.
Yeung, T., Georges, P.C., Flanagan, L.A., Marg,
B., Ortiz, M., Funaki, M., Zahir, N., Ming,
W., Weaver, V., and Janmey, P.A. 2005. Effects
of substrate stiffness on cell morphology, cytoskeletal structure and adhesion. Cell Motil.
Cytoskel. 60:24-34.
Zhang, S., Holmes,
R.O., Su, X.,
complementary
port mammalian
16:1385-1393.
T.C., DiPersio, C.M., Hynes,
and Rich, A. 1995. Selfoligopeptide matrices supcell attachment. Biomaterials
Extracellular
Matrix
10.14.21
Supplement 40
Generation of Micropatterned Substrates
Using Micro Photopatterning
UNIT 10.15
Andrew D. Doyle1
1
National Institute of Dental and Craniofacial Research, National Institutes of Health,
Bethesda, Maryland
ABSTRACT
Micro photopatterning (μPP) has been developed to rapidly test and generate different
patterns for extracellular matrix adsorption without being hindered by the process of
making physical stamps through nanolithography techniques. It uses two-photon excitation guided through a point-scanning confocal microscope to locally photoablate
poly(vinyl) alcohol (PVA) thin Þlms in user-deÞned computer-controlled patterns. PVA
thin Þlms are ideal for surface blocking, being hydrophilic substrates that deter protein
adsorption and cell attachment. Because gold substrates are not used during μPP, all
live-cell ßuorescent-imaging techniques including total internal reßection ßuorescence
microscopy of GFP–linked proteins can be performed with minimal loss of ßuorescence
signal. Furthermore, because μPP does not require physical stamps for pattern generation, multiple ECMs or other proteins can be localized within microns of each other. This
unit details the setup of μPP. It also provides troubleshooting techniques. Curr. Protoc.
Cell Biol. 45:10.15.1-10.15.35. Keywords: micro photopatterning r micropatterning r extracellular matrix r
two-photon confocal microscopy r photoablation r polyvinyl alcohol r thin Þlm
INTRODUCTION
This unit describes the generation of micropatterned substrates using a direct-writing
method known as micro photopatterning or μPP. Micropatterning of extracellular matrix
(ECM) components, where ECM proteins are applied to a two-dimensional surface in a
speciÞed pattern, has been an important technique for understanding how cells respond
and react to their physical surroundings. The most common manner in which to generate
micropatterns is through the process known as microcontact printing or μCP, where
a physical stamp is used to “ink” ECM patterns onto a gold-coated coverslip (Lehnert
et al., 2004). More details about this process can be found in the Background Information
section of the Commentary.
While μCP has been reÞned and can now generate sub micron-sized patterns (Lehnert
et al., 2004), it has several limitations. One major limitation is that the patterns of a master
cannot be readily changed. This requires a new master for each new pattern or stamp. A
second limitation is the use of gold for alkanethiol attachment. Being an electron-dense
metal, gold strongly quenches green ßuorescent protein (GFP) ßuorescence, leaving
live-cell ßuorescence imaging nearly impossible.
Here, we describe in detail how to generate ECM micropatterned glass-bottomed dishes
or coverslips using μPP that bypasses many of the issues associated with other micropatterning techniques (Doyle et al., 2009). This technique utilizes high-powered two-photon
(TP) laser excitation channeled through a point-scanning confocal microscope in order
to physically ablate or remove a thin Þlm of poly(vinyl) alcohol (PVA). This exposes
the underlying glass surface to which ECM proteins can directly attach. PVA’s high
hydrophilicity and relative inertness make it an optimal candidate for deterring protein
Current Protocols in Cell Biology 10.15.1-10.15.35, December 2009
Published online December 2009 in Wiley Interscience (www.interscience.wiley.com).
DOI: 10.1002/0471143030.cb1015s45
Copyright Extracellular
Matrix
10.15.1
Supplement 45
adsorption to non-ablated regions of the thin Þlm. Furthermore, PVA’s high refractive
index (∼1.5) allows for all ßuorescence techniques, including total internal reßection
ßuorescence (UNIT 4.12), even through nonablated regions of the Þlm. In addition, by
performing several rounds of μPP in series, multiple proteins can be deposited locally
within microns of each other.
STRATEGIC PLANNING
While μPP is a fairly straightforward methodology, it does call for experience and
knowledge using a confocal microscope. While the day-to-day processing requires little
expense, two important pieces of equipment are necessary: (1) A spincoater to evenly and
thinly distribute PVA over the glass surface, and (2) a two-photon laser point-scanning
confocal microscope to locally ablate patterns in the PVA thin Þlm. While each piece
of equipment is an added expense (spincoater: ∼$6,000 to $12,000; Zeiss 510 LSM
NLO system: >$500,000) both can be found on most university campuses. Check with
Material Science or Bioengineering departments for spincoating devices, and Biology
or Physics departments for availability of two-photon confocal systems. Throughout this
unit, we refer to the Zeiss 510 LSM NLO system (NLO stands for NonLinear Optics) as
the confocal microscope and the AIM software version 4.2, which runs the microscope.
It is assumed that the user has basic knowledge using the software in expert mode.
Although not described here, it is our view that with slight alterations to these protocols,
other two-photon confocal microscope types (i.e., Olympus) could be used to attain local
ablation of the PVA thin Þlm.
The process of μPP can be broken down into Þve stages: (1) glass surface activation, (2)
PVA thin Þlm deposition, (3) photoablation, (4) surface quenching, and (5) ECM adsorption. Each part of the process can be a stopping point, with the steps that follow occurring
up to several days or even weeks apart. For example, following glass surface activation,
dishes/coverslips can be kept desiccated for over month at 4◦ C, which allows you to
generate 30 or more dishes at a time, but other steps can be performed in smaller batches
more frequently. Table 10.15.1 shows the longevity of samples after their given stage of
processing. The following sections will describe in detail each of the protocols associated
with the stages, as well as provide background information and troubleshooting tips.
Table 10.15.1 Longevity of Samples After the Given Stage of
Processing
Processing stage
BASIC
PROTOCOL 1
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
Stability time (days)
Glass surface activation
30+
PVA thin Þlm deposition
7-14
Photoablation
30+
Quenching
30+
ECM adsorption
1-3
ACTIVATION OF THE GLASS SURFACE
This protocol describes how to prepare the glass surface for direct conjugation to the PVA
thin Þlm. The Þrst order of business is to clean the glass surface of any organic residues
through acid washing. This is crucial so the silanes are uniformly distributed over the glass
to provide covalent attachment of the PVA thin Þlm. The term “activated glass” or “activation” here refers to adding a reactive aldehyde group conjugated directly to the glass
surface, through silanes. The aldehydes can react directly with hydroxyl groups found on
the PVA polymer, and hence the thin Þlm generated in later steps is covalently attached
10.15.2
Supplement 45
APTMS + GA
TESBA
TESUDA
O
H
O
H
GA
O
H
H
O
H
+
O
H2 N
N
H3C
H3C
O
O
Si
O
CH3
H3C
O
O
Si
O
O
O
CH3
H3C
Si
O
CH3
O
Si
O
O
CH3
glass surface
Figure 10.15.1 Schematics of APTMS, TESBA, and TESUDA conjugated to a glass surface. The addition of
glutaraldehyde (GA) to APTMS-coated glass results in “activation” by attachment to the free amino group, leaving a
free aldehyde to react with PVA during spin coating.
to the glass. The silanes in question are (3-aminopropl)trimethoxysilane (APTMS), triethoxysilylbutraldehyde (TESBA), and 11-(triethoxysilyl)undecanal (TESUDA). The
Þrst is an amino-terminated silane, which requires glutaraldehyde for “activation,” while
the latter two are both aldehyde terminated (Fig. 10.15.1). Protocols for using both types
are detailed below. Caution should be used when handling silanes; they can cause severe
burns and damage many surfaces.
NOTE: After the completion of each process (i.e., acid washing, silanization, etc.) it
is important to visually inspect all dishes under a microscope with a 10× objective to
observe if there are any imperfections in the surface. If there are, it is best to discard the
affected dishes before any further steps.
Materials
50% (v/v) nitric acid
Deionized or distilled water
200 mM NaOH solution (see recipe)
(3-aminopropyl)trimethoxysilane (APTMS: 97% or higher; Gelest)
50% (v/v) glutaraldehyde (Electron Microscopy Sciences)
Aldex (Waste and Compliance Management)
Drierite (W.A. Hammond Drierite Company)
R-3603 Tygon tubing (Norton Performance Plastics)
1000-μl barrier Þlter pipet tips
Low-pressure air jet
Fume hood
Thirty MatTek glass-bottomed dishes (P35G-1.5-10-CMatTek)
Carrying tray
Pasteur pipets
Automatic pipettor
Extracellular
Matrix
10.15.3
Supplement 45
2000-ml beaker
Scale
Scintillation vials or other small glass containers
Drying oven
500-ml screw-top container (e.g., Nalgene) or other similar desiccated containers
for storage
Prepare the glass surface for silanization
1. Before starting the acid washing, create a compressed air blower using 3 ft of Tygon
tubing and a 1000-μl barrier Þlter pipet tip. Connect one end of the tubing to a
low-pressure air jet (commonly found on most laboratory benches). Insert the pipet
tip into the other end, tip out.
Alternatively, this can be attached to a compressed air or nitrogen tank. This will used
throughout the process to dry the dishes. Use low pressure for drying, 1/4 to 1/3 of the way
open or under 10 psi.
2. In a fume hood, arrange thirty MatTek dishes on a carrying tray.
3. Using a Pasteur pipet Þt into an automatic pipettor, add a small amount of 50% (v/v)
nitric acid to each dish, just enough to cover the glass area, usually between 300 and
500 μl. Incubate for 25 min at room temperature.
4. After the incubation period has elapsed, place dishes in a large 2000-ml beaker and
rinse with deionized or distilled water, under continuous ßow for a minimum of 4 hr
or overnight.
Be sure that dishes are not ßoating.
5. Remove dishes from water and aspirate remaining water from dishes.
6. Arrange dishes on a carrying tray again, add 300 to 500 μl of 200 mM NaOH to
each dish, and incubate for 15 min at room temperature.
This step helps to exchange H+ residues associating with the glass from the acid washing
and neutralizing or replacing them with OH− , which is more conducive for the binding
of the methoxy or ethoxy portion of the silane.
7. Rinse dishes two times, each time with 3 ml deionized or distilled water, then dry
under compressed air (device created in step 1).
Silanize glass surfaces
8. Place a scintillation vial on a scale, tare, and weigh out 1% (w/v) APTMS.
CAUTION: Silanes corrode metals, plastics, and all organics. Only use glass for transferring and wear chemical-resistant gloves when handling.
9. Insert a glass Pasteur pipet into the APTMS and tilt container to 45◦ . Allow capillary
action to bring the APTMS into the pipet.
10. Place your gloved Þnger on the top open-end of the pipet, transfer, and release a total
of 100 mg into the tared vial. Add 9900 μl distilled water to the scintillation vial.
Replace vial cap, mix gently, and incubate for 1 min at room temperature.
11. Add ∼300 to 500 μl of the 1% APTMS solution to each dish, only covering the glass
portion. Incubate 5 min at room temperature.
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
12. Aspirate APTMS and dispose of properly (contact your chemical safety ofÞcer for
the proper disposal).
13. Rinse two times, each time with 3 ml distilled water over 10 min.
14. Aspirate the water and dry the dishes with compressed air.
10.15.4
Supplement 45
15. Replace the dish lids and incubate at 65◦ C in a drying oven for a minimum of 3 hr.
This step cures the silanes. They can also be left desiccated at room temperature overnight
and achieve the same level of curing.
Incubation above 65◦ C leads to warping of the dishes.
Activate amino-terminated silanes with glutaraldehyde
16. Mix 100 μl of 50% glutaraldehyde with 9900 μl distilled water to make a 0.5%
glutaraldehyde solution. Add ∼300 to 500 μl of the solution to only the glass surface
of the silanized dishes. Incubate for 30 min at room temperature.
To convert this amino-terminated silane (APTMS) into an active aldehyde, glutaraldehyde
(a bi-functional aldehyde) is added to react directly with the amino group.
17. Remove the glutaraldehyde solution by aspiration and discard properly in Aldex.
Aldex is used to inactivate the reactive aldehydes allowing for proper disposal.
18. Rinse dishes three times, each time in 3 ml distilled water over 20 min.
19. Aspirate the water and blow-dry the surface again.
20. Store up to 1 month at 4◦ C in a desiccated storage container (add Drierite to the
container bottom or to a small vial placed inside it).
USING ALDEHYDE-TERMINATED SILANES FOR SURFACE ACTIVATION
In order to bypass the glutaraldehyde-activation steps of the Basic Protocol, TESBA or
TESUDA, which are already terminated in a reactive aldehyde group, can be used for
silanization. Both are triethoxysilanes, which renders them water insoluble and requires
ethanol as the solvent.
ALTERNATE
PROTOCOL 1
Additional Materials (also see Basic Protocol 1)
Absolute ethanol (200 proof)
Triethoxysilylbutraldehyde (TESBA) or 11-(Triethoxysilyl)undecanal (TESUDA;
both from Gelest)
Prepare the dishes
1. Follow steps 1 through 7 of Basic Protocol 1 for acid washing of glass-bottomed
dishes.
Silanize the glass surface
2. Place a scintillation vial on a scale, tare, and weigh out 1% (w/v) of either silane.
3. Insert a glass Pasteur pipet into the silanes and tilt container to 45◦ . Use capillary
action to bring the silanes into the pipet.
4. Place your gloved Þnger on the top open-end of the pipet, transfer and release a total
of 100 mg into the tared vial.
5. Add 9900 μl of absolute ethanol to the scintillation vial. Mix gently with the cap on
and incubate for 5 min at room temperature.
6. Add ∼300 to 500 μl of the 1% silane solution to each dish, only covering the glass
portion. Incubate for 5 min at room temperature.
7. Dispose of the silanes properly (contact your chemical safety ofÞcer for the proper
disposal).
Extracellular
Matrix
10.15.5
Supplement 45
Rinse and dry the silanized dish
8. Flooding the dish, rinse twice, each time with 3 ml absolute ethanol followed by a
single rinse with 3 ml distilled water for 2 min.
9. Aspirate the distilled water and dry dishes with compressed air.
10. Replace the dish lids and incubate at 65◦ C in a drying oven for a minimum of 3 hr.
This step cures the silanes. They can also be left desiccated at room temperature overnight
to achieve the same level of curing.
Curing at temperatures above 65◦ C leads to warping of the dishes.
11. Store up to 1 month at 4◦ C in a desiccated storage container (add Drierite to the
container bottom or to a small vial placed inside it).
BASIC
PROTOCOL 2
GENERATING POLYVINYLALCOHOL (PVA) THIN FILMS
Here, we describe the mixing of the PVA solution and spincoating it into a thin Þlm
on the activated glass dishes. Spincoating is a process by which a solution (in this case
PVA) is thinned evenly across a surface using centripetal force. This is often used in
the semiconductor industry to apply agents to silicon wafers. A ßat glass disc or dish is
attached via vacuum suction to a central “chuck” (Fig. 10.15.2). Chucks come in many
sizes to Þt the size of the disc/dish. The vacuum secures the disc from moving during
spinning. Most Spincoaters require a vacuum source (in our case a gel pump) and a
source of high-pressure air, (a normal compressed air cylinder). The latter is to keep
liquids and other materials from coating the rotor.
Materials
Distilled water
Poly(vinyl) alcohol powder (mol. wt. between 13,000 and 100,000; 98%
hydrolyzed minimum; Sigma)
2 N HCl
5 M NaCl
400-ml glass beaker
Stirrer/hot plate
50-ml conical tubes
Scale
Flea Micro magnetic stir bar (VWR)
50-ml Sterißip (0.2-μm pore size)
Gel vacuum pump
High-pressure compressed air source (e.g., compressed air cylinder)
Vortex
5 to 10 MatTek dishes with activated glass surface (Basic Protocol 1 or Alternate
Protocol 1)
Pipettor
Spincoater with a chuck capable of accepting 50-mm or smaller items (A
WS-400B-6NPP/LITE spincoater from Laurell Technologies shown in
Fig. 4.15.2 is used here)
Scintillation vial or 35-mm tissue culture dish
Nalgene 500-ml screw top or other similar containers for storage
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
Make the PVA solution
The PVA solution can be used repeatedly over a period of 1 week. We recommend making
it fresh weekly.
10.15.6
Supplement 45
Figure 10.15.2 Laurell Technologies Spincoater with an “activated” MatTek dish in position before
spincoating PVA/HCl solution. The chuck (arrow) can be changed according to the size required
for the dish/disc being spun.
1. Add ∼270 ml of water to a 400-ml glass beaker and begin heating on a stirring hot
plate at medium-high heat.
2. Weigh out 2.835 g of PVA into a 50-ml conical tube.
3. With the tube still on the scale, bring the weight/volume up to 50 g with distilled
water. Add a stir bar to the conical tube and cap.
4. Loosen the cap of the tube slightly and place the PVA-containing tube into the beaker
of hot, near boiling water. Be sure the water line of the beaker is between the 35 to
40 ml lines on the conical tube.
PVA will not go into solution until it reaches ∼90◦ C.
5. Turn on the stirring mechanism to medium and continue heating until PVA goes into
solution, ∼10 to 15 min. Have the 50-ml Sterißip ready.
6. Inspect the PVA solution and be sure no PVA crystals/powder remains. When sure,
immediately Þlter the PVA solution into the Sterißip (0.2 μm) using vacuum.
The PVA must be hot (above 90◦ C) to Þlter. Once cooled, the viscosity of the solution
increases and makes Þltering nearly impossible.
7. Allow the PVA solution to cool to room temperature.
8. Once cooled, transfer 8876 μl to a new 50-ml conical tube. To this, add 1124 μl of
2 N HCl and mix by vortexing.
This acidiÞed PVA solution can now be used for spincoating.
Extracellular
Matrix
10.15.7
Supplement 45
Spincoat thin Þlms of PVA
9. Retrieve 5 or 6 of the “activated” MatTek dishes that were stored desiccated at 4◦ C
(Basic Protocol 1 or Alternate Protocol 1).
10. Add ∼300 to 500 μl of the PVA/HCl solution with a pipettor to the glass surfaces of
the dishes and incubate for 5 min at room temperature. Be sure to completely cover
the glass area.
11. Center a single dish on the chuck of the spincoater. Follow the manufacturer’s
instructions for operation. Pull a vacuum and turn on your compressed air. Initiate
your spin. When Þnished, repeat for each dish.
For spincoating, we have found that a relatively short spin of 40 sec at a high velocity
(7000 rpm) works best for generating a thin coating of PVA, with few imperfections.
Acceleration of the spincoater is also important: 550 rpm gets the dish up to speed within
18 sec. However, you may need to adjust all variables (time, velocity, and acceleration)
to determine which best suits your needs.
12. Add 2 to 3 ml of 5 M NaCl to either a scintillation (no cap) vial or a 35-mm tissue
culture dish and place in the storage container.
13. Place dishes within the storage container and incubate a minimum of 2 hr before
ablating. Store dishes within the container at 4◦ C for up to two weeks prior to
photoablation.
High molarity or super-saturated salt solutions are effective in regulating humidity within
a closed environment with each salt type maintaining a different relative humidity. For
NaCl, it is ∼50% to 60%, which is optimal for reducing issues with PVA crystallization,
and loss of surface hydrophilicity.
At this point in the processing, the activated glass surface is now coated with a submicron
(between ∼100 to 200 nm) thin Þlm of PVA, which will deter protein adsorption and cell
attachment (Fig. 10.15. 3).
glass surface
Figure 10.15.3 Representation of the PVA thin film generated on dishes following processing through Basic
Protocol 2 using APTMS together with glutaraldehyde as the cross-linker.
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
10.15.8
Supplement 45
PHOTOABLATION WITH TWO-PHOTON CONFOCAL MICROSCOPY
In this section, the actual process of photoablation is described. It is divided into several
sections for clarity. First, the conÞguration settings for the confocal microscope are
addressed, followed by the pre-ablation setup and how to generate pattern templates. We
then detail how to photoablate single Þelds of view (FOV), and Þnally how to automate
the process using the Multi Time macro and tiling functions. Throughout this section
and the ones that follow, we will often be referring to dialog boxes and buttons within
the AIM software. The title of each dialog/button will be bolded for easier referencing.
For all intents and purposes, the actual photoablation process merely utilizes the intrinsic
capabilities of a point-scanning confocal microscope. However, the major difÞculty is in
setting the proper conÞgurations to elicit localized ablation efÞciently. Once conÞguring
is complete, the process can be performed with relative ease.
BASIC
PROTOCOL 3
Materials
Glass cleaner
Immersion oil
Zeiss 510 LSM NLO confocal microscope or later model with 1.5-W minimum
tunable two-photon titanium:Sapphire laser, and a 633-nm HeNe2 laser (5 mW
power output)
AIM software (Zeiss MicroImaging)
63× oil immersion objective with numerical aperture of 1.3 or higher capable of
NLO transmission
PVA thin Þlm-coated MatTek dishes (Basic Protocol 2)
Set up confocal microscope conÞguration
The following steps are meant to guide you through conÞguration setup and scan settings
that are required for the ablation process. Throughout this section are screen shots of the
AIM software (version 4.2) to help in understanding how and where to change settings
(indicated by bolded numbers in Þgures). It is important to note that the direct light path
from the TP laser to the confocal scan head should be aligned at least once a month.
Ablation efÞciency is greatly reduced when mirrors are misaligned. Have a qualiÞed
individual align the mirrors (microscope facility director, Zeiss service representative,
etc.) at 755 nm before beginning the ablation process.
1. Turn on the confocal microscope and boot the computer. Be sure to turn the twophoton laser from the standby position to the on position.
There is no need to ignite the mercury arc lamp since it is not used to Þnd the thin Þlm.
2. Open the AIM software in expert mode.
3. Go into the Acquire menu and open the Laser window (Fig. 10.15.4, 1; i.e., the
inserted numeral 1 in Fig 10.15.4); turn on the 633-nm laser and tune the two-photon
to 755 nm.
4. In the same Acquire menu, select ConÞguration (Fig. 10.15.4, 2) to open up the
conÞguration window (Fig. 10.15.5).
5. In Channel Mode, select Singletrack or Multitrack.
6. For the primary dichroic (Fig. 10.15.5A), select the HFT KP 700/488 (1).
This allows the near-infrared (NIR) light from the two-photon to be reßected to your
sample. While the 633-nm light is not properly matched for the dichroic, it provides
back reßection of the glass/thin Þlm interface, similar to backscatter, helping you Þnd the
appropriate z-plane for ablation.
Extracellular
Matrix
10.15.9
Supplement 45
Figure 10.15.4 The LSM 510 AIM software Expert Mode window with the Acquire menu open. All screen
shots of the AIM software are courtesy of Carl Zeiss MicroImaging.
Figure 10.15.5 Configuration Control window screen shot for Zeiss’ AIM software (version 4.2). (A) Red line depicts the light path from the lasers to the PVA-coated dish, while the green line illustrates the reflected light path
from the dish to the channel 3 photomultiplier tube (Ch3). Numbers show the primary dichroic, (1) mirror (2), and
filters (3 and 4) required for obtaining a reflected light image of the thin film surface. (B) The laser excitation panel
with 633-nm and 755-nm TP settings. Courtesy of Carl Zeiss MicroImaging. For color version of this figure go to
http://www.currentprotocols.com/protocol/cb1015.
7. Select photomultiplier tube or channel 3 (Ch3) and set the following Þlters along the
light path illustrated below: Mirror (2), NFT 545 (3), and LP 560 (4).
8. Set the 633-nm laser to 1.5% power and the 755-nm TP to ∼90% by selecting
Excitation, this opens the Excitation window (Fig. 10.15.5B).
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
The percent power is a reßection of the total amount allowed through the AOTF or AOM
for the 633-nm HeNe2 and the 755-nm TP, respectively. Using the TP at 90% does not
affect the lifetime of the laser.
10.15.10
Supplement 45
Figure 10.15.6 Scan Control window illustrating the Channels dialog panel. Courtesy of Carl
Zeiss MicroImaging.
9. Click on the ConÞg button on the middle right-hand side of the window. Save the
conÞguration at this point.
You may want the title to be basic at this point. Later on, this basic version can be changed
and saved to reßect your zoom and scan settings (discussed in the next few steps).
10. Now that you have created and saved the proper basic conÞguration, open the Scan
window and select the Channels menu (Fig. 10.15.6). Set the detector gain to
approximately halfway up, in the low 500s (Fig. 10.15.6, 1).
11. Be sure that frame scan is selected (Fig. 10.15.6, 2) and the pinhole is set to 1 Airy
for a 63× objective.
12. Select the Mode window (Fig. 10.15.7).
The Mode window is divided vertically into four separate control boxes: (1) objective,
line stepping, and frame control, (2) scan speed, (3) pixel depth, scan direction, and
averaging, and (4) zoom, rotation, and offset. Because of the complexity of this window,
each control box will be addressed in separate steps below.
13. Objective, line stepping, and frame control (Fig. 10.15.8): For μPP, this box in
the Mode window is important for controlling the pixel size, which will impact the
Extracellular
Matrix
10.15.11
Supplement 45
Figure 10.15.7 Scan Control window illustrating the Mode dialog panel. The four boxes within
the panel are discussed in detail in the step numbered on the left side of the panel with enlarged
images. Courtesy of Carl Zeiss MicroImaging.
Figure 10.15.8 Objective, line stepping, and frame control box in the Scan control window of the
AIM software. Courtesy of Carl Zeiss MicroImaging.
region of interest (ROI) template, how detailed or deÞned the ablation patterns are,
as well as the total scan time (the larger the pixel array the longer the scan time).
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
We have found that using a 512 × 512 array is a good intermediate (Fig. 10.15.8, 1): fast
scanning with a reasonably high level of detail. Line step should be 1 (Fig. 10.15.8, 2).
Objective can be varied but should have a high numerical aperture (N.A), minimum
of 1.0.
10.15.12
Supplement 45
Figure 10.15.9 Scan speed box in the Scan control window of the AIM software. Courtesy of
Carl Zeiss MicroImaging.
Figure 10.15.10 Pixel depth, scan direction, and averaging box in the Scan control window of
the AIM software. Courtesy of Carl Zeiss MicroImaging.
14. Scan speed (Fig. 10.15. 9): This sets the relative rate of speed the galvanometric
mirrors move and scan the FOV or the ablation area. As scan speed increases
(Fig. 10.15. 9, 1), both pixel time and Scan time (Fig. 10.15.9, 2 and 3, respectively)
decrease. The Pixel Time (or dwell time) is the amount of time the laser dwells on any
given pixel in the FOV. Since efÞciency of the ablation process is directly dependent
on the total amount of light energy (in μjoules/μm2 ), the dwell time is dependent on
the total power output of the TP laser (in our case ∼1200 mW at 755 nm). A scan
speed of 4 using a 63× provides efÞcient ablation with our setup.
Scan speed will need to be decreased when the pixel size changes, for instance, when a
digital zoom is used to generate smaller patterns with the same objective lens. This will
be discussed in later sections.
15. Pixel depth, scan direction, and averaging (Fig. 10.15.10): Scan direction
and scan average are equally important here. Choose to reverse scan direction
(Figure 10.15.10, 1; reverse arrow). This decreases the scan rate by half when compared to the single direction. When choosing a scan, the correction dialog box is
opened. This helps to align the scanning properly (Fig. 10.15.10, 2; see your Zeiss
representative or manual for more on how to do this). Scan average (Fig. 10.15.10, 3)
provides the same basic function it normally does when imaging: repeated scanning
of the same line or frame in the FOV. However, instead of decreasing background
noise it adds a second (or more) pass over the area being ablated. For example, increasing this number from 1 to 2 overall doubles the laser dwell time by performing
a second pass. It will also double the scan time (from step 14). Leave the scanning
in line mode. The method does not matter since you are not saving the image Þles.
16. Zoom, rotation, and offset (Fig. 10.15.11): In this box, you can set the appropriate
zoom (Fig. 10.15.11, 1). This is helpful if you want to decrease the size of an entire
ROI template. For example, you can generate a circle pattern with a diameter of 10 μm
simply by zooming in by 2× with a ROI template containing a circle with a 20-μm
diameter. Here it is set to 1.6, which, when using a 63× objective, is equivalent to
100×. This is not an optical or a true digital zoom: the same pixel array is scanned but
Extracellular
Matrix
10.15.13
Supplement 45
Figure 10.15.11 Zoom, Rotation, & Offset box in the Scan control window of the AIM software.
Courtesy of Carl Zeiss MicroImaging.
simply in a smaller region on the galvanometric mirrors. Rotation (Fig. 10.15.11, 2)
is important only when generating multiple patterns (same or different) next to each
other, and comes into play in later sections when using the automated tiling function.
In many confocal microscopes, the galvometric mirrors are slightly offset from the
true horizontal or vertical plane of the stage. By correcting the rotation for this offset,
larger patterns (i.e., long lanes or lines) can be ßawlessly connected and generated.
Setting the proper rotation offset is discussed later in Support Protocol 1. FOV offset
is normally not changed.
The Zoom function can also go below 1 to 0.7 giving you a larger FOV. However, the
rotation dialog will be reset to 0 and cannot be changed.
17. At this point, the confocal conÞgurations are properly set. Save the ConÞguration
again in the ConÞguration control panel, as you did in step 3 above.
Pre-ablation set up
While the previous section led you through the proper confocal conÞgurations required
for efÞcient ablation of the PVA thin Þlm, this section will guide you through how to Þnd
the proper z-plane to ablate the thin Þlm, how to align the z-plane, and how to generate
ROI templates for use in ablation. After these steps, ablation can be performed.
18. Bring your PVA thin Þlm dishes in their container to room temperature. Place the
dishes you will be patterning near or on the microscope to allow them to acclimate
to the temperature of the confocal microscope since differences in oil, objective, and
dish temperature will cause focus drift.
19. Boot the system as before in step 1, turning on the appropriate lasers.
20. Clean the bottom of the dish with glass cleaner thoroughly and dry. Add a small drop
of oil directly to the bottom of the dish and place it in the single dish holder.
21. Before proceeding, make sure the stage insert is clean and properly Þt into the stage
with all adjustment screws up (not contacting the stage plate).
22. After loading the software, load the conÞguration for ablation saved in previous
steps.
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
23. Bring the 63× objective up to your sample using the course focus knobs until the oil
hits the glass.
10.15.14
Supplement 45
Figure 10.15.12 Scan Control window with Channels panel displayed. Box indicates an
unchecked 755-nm laser line. Courtesy of Carl Zeiss MicroImaging.
24. Before using Fast XY to Þnd your sample, be sure to inactivate the 755-nm laser line
in the Scan control window under the Channels panel (Fig. 10.15.12, 1 black box).
This is imperative since this amount of light will ablate the surface as soon as you reach
the focal plane.
25. Using the Fast XY function, begin scanning for the PVA thin Þlm surface by rotating
the Þne focusing knob (up: clockwise on the right-hand side of the microscope). Do
Extracellular
Matrix
10.15.15
Supplement 45
Figure 10.15.13 Reflected light images while trying to find the PVA thin film. (A,B) Images were
taken while scanning in the z-plane. The bright linear areas indicate passing by the glass/PVA thin
film (arrows). (C) Once found, the field of view (FOV) should appear equally (uniformly) bright if the
surface is level.
this at a moderate to fast pace moving 300 to 400 μm in ∼10 sec time. While doing
this, keep your eye on the monitor. When you reach or pass the focal plane it will appear as a bright line or set of lines on the screen in Fast XY mode (Fig. 10.15.13A,B).
26. Adjust the objective Z position until nearly the entire FOV is in focus
(Fig. 10.15.13C).
Once in focus, you may need to adjust the PMT gain up or down depending on the
brightness. The brightest z-plane, which is the glass interface, should be near pixel
saturation (∼255 for an 8-bit image).
27. More often than not, the FOV’s brightness is unevenly distributed, such as in
Figure 10.15.14A, and requires adjustment since the TP light is maximally absorbed
only at the focal plane. To adjust properly you Þrst need to know which corner or
edge is low. Adjust the focus so the focal plane is below the glass. Slowly raise the
objective until you start to see brightness in the FOV: whichever area appears bright
Þrst is low and needs to be raised using the adjustment screws in the stage insert
plate (Fig. 10.15.14B). In the example in Figure 10.15.10A, both the upper-right and
lower-right adjustment screws need to be turned clockwise, or screwed in to raise the
stage up. Perform several half turns. The image should go black, indicating the stage
has been adjusted. Refocus with the Þne focusing knobs. Repeat this process until
the FOV demonstrates even illumination, as in Figure 10.15.13C. This adjustment
process should be performed before ablation of any dish. When Þnished with each
dish return adjustments screws to their neutral positions.
28. Wait 5 min for the focus to adjust due to temperature variations, and then refocus to
the brightest FOV. Open the Stage and Focus control window (Fig. 10.15.15). In
the stage control window, set the Z Focus step (Fig. 10.15.15, 1) to 0.25 μm.
29. Using the focusing arrows (Fig. 10.15.15, 2), raise the focal plane to 0.75-μm above
the brightest FOV. This is the proper height to perform ablations. The FOV should
be slightly dimmer. Set this point to zero (Z: Fig. 10.15.15, 3).
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
30. In the stage position box, select zero (Fig. 10.15.15, 4). Now you will know the
position of your Þrst ablation site.
31. Now that the dish has been leveled, ablation can commence once you have generated
a ROI template. In essence, this is the same way you would generate ROIs for
FRAPing (Fluorescence Recovery After Photobleaching, UNIT 21.1) a sample.
10.15.16
Supplement 45
Figure 10.15.14 Dish leveling. (A) Reflected light image of a dish that is slightly tilted down on the right, as
shown by the increase in brightness as you focus up to the glass surface. Below the image is a cartoon side
view representing the glass (gray) with respect to the horizon (black line). (B) Image of the confocal stage. Inset
shows a magnified view (white box) of one of four adjustment screws in the stage plate. The other adjustment
screws are indicated by asterisks. (C) Simplified schematic of the tilted dish in (A) and how to correct the tilt by
screwing in both right-side adjustment screws, until an in-focus evenly illuminated FOV is attained (right-side
of arrow).
32. To generate ROIs, select the Edit ROI window (Fig. 10.15.16A,C). The window
allows you to generate and save hundreds of different ROIs in a single FOV as
templates.
33. To begin, in the Scan Control window select New to generate a new image window.
34. From the bottom dialog boxes in the Edit ROI window (Fig. 10.15.16A, 1 and black
box), select a shape (i.e., circles, polygons rectangles, etc.). Then draw the shape onto
the new image window (Fig. 10.15.16B). Once drawn, the shapes size and location
information will appear in the Edit ROI window, checked (Fig. 10.15.16C, 4). From
here, you can resize or reposition the shape or uncheck it and remove it.
Several macros are available free from Zeiss that allow you to repeat a single ROI multiple
times on the same window, which is helpful for generating a dot-based matrix, etc. It is
also helpful to know the size of a single pixel with the objective and zoom you are using.
This can be found by selecting the Info button in the image window (Fig. 10.15.16B, 2).
35. The X and Y Scaling (Fig. 10.15.16B, 3) can then be used for pattern spacing and
sizing, as well as knowing the size of the FOV, important for moving the stage
Extracellular
Matrix
10.15.17
Supplement 45
Figure 10.15.15 Stage and Focus Control window in the Zeiss AIM software. 1 indicates the
focus step set at 0.25 μm. 2 indicates toggle arrows for focus position. 3 indicates the focus zeroing
button. 4 indicates the Stage position zeroing button. Courtesy of Carl Zeiss MicroImaging.
horizontally or vertically when generating larger repeated patterns. Save and name
the template when completed.
Photoablate the thin Þlm
36. With the above steps completed, you can now proceed with photoablation. First,
check that you are still focused ∼0.75-μm above the brightest focal plane and
readjust if necessary.
37. In the Scan control window under the Channels panel, check the 755-nm TP to on.
38. Select the ROI template of your choice from the Edit ROI window. In the Scan
control window, select the ROI button (2nd row from top 2nd from the right, see
Fig. 10.15.12).
39. Select Single scan button. The ROIs in the template should be slowly scanned from
top to bottom, taking ∼15 sec with the parameters set and discussed earlier.
Because of the intense level of light hitting the sample, the ROIs will appear saturated
with light (255 on an 8-bit scale).
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
40. Once the Scan has completed, uncheck the TP 755-nm line in the scan window,
deselect the ROI tab, and Fast XY scan the FOV to observe the post ablation result
as shown in Figure 10.15.17.
10.15.18
Supplement 45
Figure 10.15.16 Edit ROI window and pattern generation. (A) Edit ROI window with the shape dialog
boxes (1 and black box). (B) Example of generating a single circle once the circle dialog button is selected
in the Edit ROI window in (A). 2 indicates the information button, which shows on the on the left-hand side
of the image window. 3 (upper left) indicates where pixel scaling information can be found. (C) Edit ROI with
the ROI definitions (4 and black box) for the circle shown in panel (B). Courtesy of Carl Zeiss MicroImaging.
Figure 10.15.17 From template to ablated pattern. The three images representing the ROI
template (left), what is observed during the ablation (middle), and the post-ablation pattern (right).
41. To generate multiple FOVs of the same pattern use the Stage and Focus control
window to move the stage over by setting the xy step to exactly one FOV (use the
calibrated xy information multiplied by your scanning window size, assuming x and
y are equal).
This works best for dot or separated patterns such as the example illustrated in
Figure 10.15.17. However, if using a linear pattern, which needs to be seamlessly continued, we recommend reducing the xy step by 1 or 2 μm to provide an appropriate
overlap.
Automate μPP with macro functions
To this point, you should be able to efÞciently photoablate PVA thin Þlms for a single FOV.
Once the conÞgurations have been set correctly and the ROI templates are generated,
there is no true need to sit at the scope moving the stage from place to place if the
process can be automated. This can be achieved through the Multitime macro in the
AIM software. Originally intended for time-lapse imaging, Multitime allows the user to
choose multiple locations within the dish. Instead, here we use Multitime to automate
the μPP process. An additional feature is the ability to “tile” around a speciÞed point;
that is to image an array of images around a single point in a tile or grid-like fashion. The
Extracellular
Matrix
10.15.19
Supplement 45
Figure 10.15.18 Using the Stage and Focus Control window to mark multiple tiling positions. The
black box indicates four points that were selected for tiling. Courtesy of Carl Zeiss MicroImaging.
next several steps will help you to achieve this automation process. While the Multitime
macro has many functions, we are only going through those that directly pertain to
automating μPP.
42. After you have found the proper z-plane and zeroed your position as in steps 28 to
30, go to the Stage and Focus control window (Fig. 10.15.18). In the Stage position
box (center), select Mark Pos. (Fig. 10.15.18, 1). Keep this window open since you
will be referring to it later.
The example in Figure 10.15.18 shows how this was repeated four times 750-μm apart.
Each mark is listed in the dropdown menu below. These marked positions will be used for
tiling in the Multitime macro.
It is helpful to move the stage while in Fast XY mode. Once to the correct xy position,
readjust the z-plane to be ∼0.75-μm above the brightest FOV then mark the position, but
do not zero.
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
43. Prior to starting, create a new database (File>New File) and save it in an appropriate
place.
10.15.20
Supplement 45
Figure 10.15.19 Multitime macro window (A) and the Options window (B). The bold numbers
indicate the order each step is to be performed. Courtesy of Carl Zeiss MicroImaging.
44. In the main AIM window, select the Macros menu. If the Multitime macro does
not appear in the window, load it (if unfamiliar with Macros ask your Microscope
facility director or your Zeiss representative for help with installation).
45. Open the Multitime window. At the bottom of the window (Fig. 10.15.19A, bottom,
1), select the Image DB dialog button to choose the database you generated earlier.
46. From the window buttons on the right-hand side, select Options (Fig. 10.15.19A,
2 and 10.15.19B). In the Option window, Þrst create a temporary image database
(Fig. 10.15.19B, 1).
This is where the software saves your ablation images.
47. Next, check the dialog box that says “Delete Temporary Þles after Þnal experiment”
(Fig. 10.15.19B, 2). Close window.
48. From the top of the window, select Multiple locations (Fig. 10.15.19A, 3), which
will allow you to choose more than one point to scan.
Steps 44 to 48 in Multitime only need to be performed once. After performing these actions
once, Multitime will remember the settings.
49. Next, choose Edit location (Fig. 10.15.19A, 4). The edit location window
(Fig. 10.15.20A) will allow you to tile around the marked positions from step 42.
50. Under the Grid tab, change the x and y Grid numbers to the appropriate number
of FOVs to tile in each plane (Fig. 10.15.20A, 1), and then select Create Grid
Locations (Fig. 10.15.20A, 2).
This generates a list of points in the Multiple locations dropdown menu.
Extracellular
Matrix
10.15.21
Supplement 45
Figure 10.15.20 Tiling with Edit Locations window. (A) The Edit Locations window grid tab showing
the Grid Numbers dialog (1) and the Create Grid Locations button (2) for grid generation. (B) The
Edit locations window tile tab with Tile Numbers (1), configuration (2), XY correction (3), and Create
Tile Locations (4) highlighted. (C) Schematic representation of the tiled grid generated based on the
parameters shown in (B). The asterisk represents the marked positions chosen in the Stage and Focus
control window. Courtesy of Carl Zeiss MicroImaging.
51. Next, select the tile tab (Fig. 10.15.20B). From top to bottom (1), change the tile
numbers x and y (Fig. 10.15.20B, 1), (2) select the conÞguration to be used from the
dropdown menu, then load it by hitting Load Conf. (Fig. 10.15.20B, 2), and (3) set
your X and Y correction (Fig. 10.15.20B, 3; used for overlapping or spacing FOVs
apart).
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
10.15.22
Supplement 45
52. In the Stage and Focus control window, select a position from the dropdown menu,
and select Move To.
53. Then, back in the Edit Location window select Create Tile Locations
(Fig. 10.15.20B, 4). If more than one location is to be tiled, move to its position
then select Add Tile Location (Fig. 10.15.20B, 5; after your initial locations all
other locations are added in this manner).
54. Go to the Scan Control window and select ROI.
IMPORTANT NOTE: This is very important. If not selected, the Multitime Macro will
photoablate the entire FOV.
55. Select the No Tile Mode (Fig. 10.15.19, 5).
This stops the tiling of the images collected into a larger single image.
56. Select Start in the Multitime window.
You should start seeing Multitime running. At the bottom of the Multitime window, information should begin appearing telling you the position, scan number, and other information.
It is helpful to time the process for a single FOV plus the stage movement to estimate your
Þnishing time.
57. Once this Multitime has Þnished, check several FOVs to be sure the ROI template
has been repeated. Close the Multitime Macro window.
58. If more patterning is required in the same dish (the same or a different pattern),
simply move to a new nonablated region being sure to know your current position
with respect to all previously ablated areas.
59. Repeat Þnding the z-plane, reset zero for x, y, and z.
The original zero point (0, 0, 0 for x, y, and z, respectively) should now be listed in the
Stage and Focus control window as something different. For example, if you moved over
2000 μm in x, 0 μm in y, and 1.0 μm in z, the original position should read: x = −2000.00
y = 0.00 z = −1.00.
DISH QUENCHING
Following the photoablation of the PVA thin Þlms, the next step is quenching any
unreacted aldehydes. Quenching the thin Þlm is important for three reasons: (1) it reduces
any reacted aldehydes leading to a stronger covalent attachment of the thin Þlm to the
glass surface; (2) the reduction of the aldehydes, especially if glutaraldehyde is used,
decreases autoßuoresence in the blue to green wavelengths (490 to 540 nm); and (3) it
acts to block any free radicals that may be produced in the thin Þlm during photoablation.
Because sodium borohydride is very hygroscopic, we suggest desiccating it at room
temperature in small aliquots in microcentrifuge tubes. Use a single tube 4 to 5 times
and discard the remainder. The reaction that occurs when NaBH4 comes in contact with
water is temperature dependent, being more vigorous as temperature increases.
BASIC
PROTOCOL 4
Materials
μPP-patterned dishes (Basic Protocols 1 through 3)
200 mM ethanolamine buffer (see recipe)
Sodium borohydride solution (NaBH4 ; see recipe)
1 M NaOH solution (see recipe)
Phosphate-buffered saline (PBS; Hyclone, cat. no. SH30264.02)
Phosphate-buffered saline (PBS) with penicillin/streptomycin and fungizone (see
recipe)
Storage containers
Scale
1. To each photoablated dish, add 2 ml of ethanolamine buffer after photoablation. Store
dishes containing ethanolamine buffer up to 1 month at 4◦ C in an airtight container.
The ethanolamine buffer should be added on the same day (preferably within an hour) of
photoablation.
Extracellular
Matrix
10.15.23
Supplement 45
2. If photopatterned dishes have been stored at 4◦ C, allow them to warm at room
temperature for 5 min.
3. Weigh out 40 mg of NaBH4 into a single 1.5-ml microcentrifuge tube. Add 1 ml of
1 M NaOH. Triturate several times.
4. To each dish, add 20 μl of the NaBH4 solution mixed in step 3. Replace the dish lid,
swirl several times to mix, and incubate up to 8 min at room temperature.
You should start to see bubbling after 1 to 2 min, which shows the reaction is occurring.
5. Aspirate NaBH4 solution and rinse two times, each time with 3 ml PBS. Add 1
to 2 ml of PBS with penicillin/streptomycin and fungizone. Store up to 1 month
at 4◦ C.
BASIC
PROTOCOL 5
ADSORBING EXTRACELLULAR MATRIX AND PLATING CELLS
The Þnal step of μPP is adsorption of an extracellular matrix (ECM) molecule to the
photoablated patterns. Here, we describe the attachment of Þbronectin to photoablated
dishes; however, any other ECM molecule or even growth factors can be absorbed in this
fashion. Once the ECM is adsorbed to the surface, it is important to block attachment of
other molecules (other ECMs, growth factors, etc.) found in serum using heat-denatured
bovine serum albumin (BSA). For ßuorescence microscopy techniques where the patterns
are invisible, prior direct conjugation of a ßuorescent dye to the ECM molecule of choice
is helpful for pattern visualization. A general ßuorescent dye labeling protocol for Nhydroxy succimidyl ester-based dyes can be found in Support Protocol 2. Pluronic F-127
is a nonionic detergent/surfactant, which is used here to help with blocking nonspeciÞc
protein adsorption to nonablated surfaces.
Materials
Fibronectin at 2 mg/ml concentration in PBS or other suitable buffer
Phosphate-buffered saline (PBS) with 0.1% (v/v) pluronic F-127 (see recipe)
μPP patterned dishes (Basic Protocols 1 through 4)
Lyophilized bovine serum albumin (BSA)
Phosphate-buffered saline (PBS; Hyclone, cat. no. SH30264.02)
2 M NaCl solution
Phosphate-buffered saline (PBS) with penicillin/streptomycin and fungizone (see
recipe)
NIH/3T3 cells (ATCC) grown to 60% to 70% conßuency in a 100-mm diameter
dish in 10% CO2 incubator
Hanks balanced salt solution (HBSS; Invitrogen)
0.5% (w/v) trypsin/EDTA solution (Invitrogen)
NIH/3T3 Þbroblast culture medium (see recipe)
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
Tissue culture hood
37◦ C waterbath
400-ml beaker
Stirrer/hot plate
Scale
50-ml conical tubes
Glass test tube capable of holding 30 to 50 ml
Flea Micro magnetic stir bar (VWR)
Digital thermometer
Ice in an ice bucket
Vacuum aspirator
Airtight storage container
10.15.24
Supplement 45
Benchtop swinging-bucket rotor centrifuge with adapters for 50-ml conical tubes
Inverted microscope equipped with a 10× phase contrast objective
Prepare coating solution
1. Calculate the total amount of Þbronectin solution needed for all dishes.
Because the glass surface only needs to be covered, we use ∼100 μl per dish.
2. Next, calculate the volume of Þbronectin (at 2 mg/ml) needed to attain the proper
concentration (10 μg/ml).
For example, for four dishes, add 2 μl of 2 mg/ml Þbronectin to a microcentrifuge tube
followed by 398 μl of PBS with 0.1% pluronic F-127 buffer for a 10 μg/ml solution.
3. Prewarm the concentrated Þbronectin solution to ∼37◦ C prior to mixing the solution.
Fibronectin is an active molecule, which over time will lose activity when left at 4◦ C
for extended periods. We suggest keeping small aliquots (∼20 μl) frozen at −80◦ C
and defrosting and using an aliquot for only 1 week stored at 4◦ C. Once diluted, the
Þbronectin solution must be used promptly and cannot be stored.
Coat the dish in the pattern
4. In a tissue culture hood, add 100 μl of the 10 μg/ml Þbronectin to each dish, cover,
and incubate for 1 hr at 37◦ C.
5. While waiting, heat 250 ml of water in a 400-ml beaker on the stirrer/hot plate to
∼85◦ C.
Heat-denature BSA
6. Weigh out 0.3 g of BSA in a 50-ml conical tube. Bring volume/weight up to 30 g
with PBS.
7. Incubate in a 37◦ C waterbath until BSA goes into solution, ∼15 min.
This BSA solution should be made fresh and can be used for a maximum of only 1 day.
8. Once the BSA has gone into solution, transfer the 1% BSA solution to a glass test
tube and add the stir bar. Next place the test tube in the ∼85◦ C water. Turn on the
stirplate to medium.
9. Directly measure the temperature of the 1% BSA solution with the digital thermometer. Set a timer for 3 min. Wait until the solutions temperature reads 83◦ C and then
start the timer. Monitor the temperature over the next 3 min.
If the solutions temperature goes above 85◦ C, remove the test tube from the water bath
and cool brießy, keeping the temperature a minimum of 83◦ C.
10. After 3 min, remove 1% BSA solution from the beaker of water and cool in an ice
bath until the solution’s temperature is below 37◦ C.
11. After the 1-hr incubation of the Þbronectin on the μPP dishes, bring the dishes to the
tissue culture hood, aspirate Þbronectin, and rinse three times, each time with 3 ml
PBS.
Rinse and block
12. Add 2 ml of 2 M NaCl solution to each dish and incubate 5 min at room temperature
to reduce nonspeciÞc protein binding.
13. Aspirate NaCl solution, and rinse three times, each time with 3 ml PBS.
14. Add 2 ml of heat-denatured 1% BSA solution (from step 10) to each μPP dish and
again incubate for 1 hr at 37◦ C.
Extracellular
Matrix
10.15.25
Supplement 45
15. Rinse dishes three times, each time with 3 ml PBS, and store in an airtight container
in 2 ml of PBS plus penicillin/streptomycin and fungizone until use. Use within 3
days of this last step.
Attach NIH/3T3 Þbroblasts to μPP patterns
16. Detach 60% to 70% conßuent NIH/3T3 Þbroblasts from 100-mm tissue culture dish
by rinsing twice, each time with 6 ml of 37◦ C HBSS.
17. Add 5 ml of 37◦ C trypsin/EDTA solution and wait 30 to 60 sec.
18. Aspirate excess solution and incubate for 2 to 3 min at 37◦ C.
19. Add 10 ml of 37◦ C NIH/3T3 Þbroblast culture medium to the dish, triturate, and
transfer to a 50-ml conical tube.
20. Centrifuge the cells in the swinging-bucket centrifuge 4 min at 1000 × g, room
temperature.
21. After centrifugation, aspirate excess medium leaving the cell pellet undisturbed.
22. Tap the tube Þrmly on the tissue culture hood working area. Add 10 ml of NIH/3T3
Þbroblast culture medium to tube, and triturate several times to loosen cell clumps.
23. Add 1 to 2 ml of cells to each μPP dish. Incubate for 10 to 15 min at 37◦ C, and then
check cell attachment to patterns using a 10× phase contrast objective on an inverted
microscope.
When cells begin to attach they should appear phase dense when compared to nonadherent cells.
24. If the proper number of cells is not attached, check dishes every 5 min.
25. When the proper number of cells has attached, gently aspirate the excess and add 1.5
to 2 ml of fresh NIH/3T3 Þbroblast culture medium. Incubate for a further 30 min at
37◦ C before imaging.
SUPPORT
PROTOCOL 1
SETTING THE CONFOCAL SCAN HEADS’ ROTATION OFFSET
As mentioned earlier in Basic Protocol 3, the x and y galvanometric scan head mirrors can
be slightly misaligned or offset from a true horizontal or vertical plane. This becomes an
issue when using the Multitime Macro for tiling: improper scanner alignment will result
in the pattern being askew from FOV to FOV. For example, if generating a lined pattern
that is meant to be continuous, lines may be offset. The following protocol alleviates this
rotation alignment issue. This can be done simply in one of two ways: Þrst, is to use a
grid slide supplied by Zeiss for the alignment, and second is to generate a grid of your
own using μPP. Both require the same steps. Any differences between the methods are
discussed.
Materials
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
Zeiss 510 LSM NLO confocal microscope or later model with 1.5 W minimum
tunable two-photon titanium:sapphire laser, and a 633-nm HeNe2 laser (5-mW
power output), and a 543-nm HeNe1 laser (1-mW power output)
PVA thin Þlm–coated MatTek dishes (Basic Protocol 2; for option 1)
Arc lamp
Grid slide, Objektträger (for option 2; Zeiss, cat. no. 474028)
Fluorescent highlighter, any color (for option 2)
Kimwipes
10.15.26
Supplement 45
For option 1
1a. Turn on the microscope, the AIM software, and the appropriate lasers for performing
μPP.
2a. Load your μPP conÞguration.
3a. Place a PVA thin Þlm–coated dish on the stage and follow the pre-ablation setup in
Basic Protocol 2, steps 18 through 30 in order to Þnd the proper z-plane for μPP.
4a. Create a new ROI template in a pattern similar to Figure 10.15.21.
The grid and circle pattern help to determine the rotational offset. The spacing of the
grid is not important, just that horizontal and vertical lines, as well as curved lines are
incorporated into the pattern.
5a. Being sure to have your zoom set for 1×, photoablate the pattern in the PVA thin
Þlm. Save the conÞguration with the TP 755-nm line unchecked.
6a. In the Focus and Stage control window, mark the position.
For option 2
1b. Turn on the arc lamp followed by the microscope, the AIM software, and the 543-nm
HeNe1 laser.
Figure 10.15.21 Microphotopatterned grid used for rotational alignment of the XY galvanometric
scanning mirrors in the confocal head with the stage.
Extracellular
Matrix
10.15.27
Supplement 45
Figure 10.15.22
MicroImaging.
Tile Scan Rotation dialog window in the AIM software. Courtesy of Carl Zeiss
2b. Select a conÞguration to image Cy3, Alexa Fluor 543, or Rhodamine dyes.
If needed, ask your microscope facility manager for help.
3b. On the grid slide, locate the grid in the center. Mark the grid with the ßuorescent
highlighter. Gently wipe off excess with a Kimwipe.
4b. Add a drop of oil over the grid and position the slide properly in the stage holder.
Use epißuorescence to Þnd the grid using a 63× 1.4 NA objective. Once the grid
is found, switch to LSM mode, and adjust the z-plane while in Fast XY. Adjust the
settings (PMT gain, laser output, etc.) and save as a new conÞguration.
5b. Scan the grid pattern until you Þnd a region that contains both the vertical and
horizontal lines, as well as part of a curved arc.
The grid pattern usually has a small and large circle pattern.
6b. In the Focus and Stage control window, mark the position.
For options 1 and 2
7. Open the Macros menu and select the Multime Macro. Open the Edit Locations
window and select the tile tab (Macros>Multime>Edit Locations> Tile tab).
8. Select the conÞguration you saved in step 5a or 4b for either option and select load
ConÞg.
9. Select the Find Rotation button.
A new window should appear similar to the one shown in Figure 10.15.22.
10. Choose Calibrate (Fig. 10.15.22, 1). This should take ∼20 sec before a number will
appear in the Rotation [◦ ] dialog on the left, which represents the rotational offset
(Fig. 10.15.22, 2).
11. Go to the Zoom, Rotation and Offset box in the Mode panel of the Scan control
window. Use the number found in the Rotation [◦ ] dialog for your rotational offset
in your μPP conÞgurations. Save the conÞgurations once complete.
SUPPORT
PROTOCOL 2
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
DIRECT FLUORESCENT LABELING OF FIBRONECTIN
For any type of micropatterning including μPP, it is important to know whether the ECM
is being adsorbed to the patterns and not nonspeciÞcally. While antibodies can help with
this determination, the direct approach is often best, especially when conducting live-cell
ßuorescence imaging. Below, we detail one method of directly labeling Þbronectin with
N-hydroxy succimidyl (NHS) ester-based ßuorescent dyes. Before proceeding, several
items should be noted: (1) NHS ester reactions are pH and temperature dependent, (2)
the reactions are hygroscopic and once in contact with aqueous solutions begin reacting
immediately, and (3) the ratio of protein to dye is important, with the reaction time
being based on this and the parameter (1) above. For every 1 mg of protein, 5 to 10 μg
of dye should be used (200:1 or 100:1 ratio). This is slightly below a 10-molar excess
10.15.28
Supplement 45
normally suggested for dye-to-protein conjugations. Over-labeling ECM proteins can
negatively affect cell attachment and/or migration. If using different starting protein
amounts, recalculate the amount of dye to match these ratios. More information about
these other types of ßuorescent conjugations can be found in “Bioconjugate Techniques”
written by Greg Hermanson (Hermanson, 1996).
Materials
NHS-ester-based ßuorescent dye of choice (several are available from Invitrogen
and Pierce)
Dimethyl sulfoxide (DMSO)
500 to 1000 μl of Þbronectin at 2 mg/ml concentration or 2 mg of lyophilized
Þbronectin
100 mM borate buffer, pH 9.0 (see recipe)
Slide-A-Lyzer (Pierce)
Aluminum foil
End-over-end rotating mixer, e.g., Labquake rotating mixer (sometimes termed a
rotisserie shaker)
Desalting spin column or dye-removal columns capable of ∼1 ml volumes (Pierce)
Centrifuge capable of 10,000 × g with 15-ml conical tube holders
1. Keep NHS-ester-based ßuorescent dyes, which are hygroscopic, in DMSO until use.
Dilute the lyophilized dye with DMSO to a concentration of 1 mg/ml. Split into
aliquots of ∼25 μl. Store at −20◦ C until use.
2. If starting from lyophilized Þbronectin, add 1 ml of 100 mM borate buffer (pH 9.0)
to make a 2 mg/ml concentration. If starting from Þbronectin in PBS or other buffers
(∼pH 7.4), dialyze in borate buffer (see APPENDIX 3C).
3. Warm 2 mg/ml Þbronectin to room temperature prior to reacting with NHS ester
dyes.
4. Defrost NHS-ester dye (1 mg/ml concentration) prior to opening the tube since it
will absorb moisture from the air.
5. Add 1 ml of Þbronectin (2 mg/ml, in borate buffer) to a 1.5-ml microcentrifuge tube.
6. Add 10 to 20 μl of the concentrated dye (1 mg/ml) to the Þbronectin solution. Close
the tube and wrap with aluminum foil.
7. Incubate 1 hr at room temperature on an end-over-end mixer at ∼8 rpm.
For other ECM molecules, it is recommended that the reaction be incubated at 4◦ C for
2 hr.
8. After 1 hr, remove unreacted dye using either a desalt spin column or dye removal
column and follow the manufacturer’s protocol.
Alternatively, gel Þltration or dialysis of the unreacted dye can be performed.
USING MULTIPLE ECM PROTEINS WITH μPP
One advantage of μPP over other patterning techniques is the ability to repeat the
process after an initial photoablation, quenching, ECM adsorption, and blocking. This
allows placement of different ECMs within microns of each other at the subcellular levels
(Fig. 10.15.23). It is crucial here to use 0.1% pluronic F-127 for protein dilution, as well
as for rinsing steps to deter nonspeciÞc protein adsorption. The following protocol details
the process.
SUPPORT
PROTOCOL 3
Extracellular
Matrix
10.15.29
Supplement 45
Figure 10.15.23 Dual ECM patterns by performing μPP twice in series. A first ablation was
performed followed by quenching, ECM adsorption, and blocking. A second round of ablation was
done is the presence of the second ECM. Green dots are fibrinogen and red lines are vitronectin.
Dots are spaced 5-μm apart. For color version of this figure go to http://www.currentprotocols.com/
protocol.cb1015.
Additional Materials (also see Basic Protocol 3)
Two different, ßuorescently labeled ECM molecules/growth factors at the proper
Þnal concentration (user deÞned)
1% (w/v) heat-denatured BSA solution (prepare fresh and keep <1 day)
Phosphate-buffered saline (PBS) with 0.1% pluronic F-127 (see recipe)
PVA thin Þlm–coated MatTek dishes (Basic Protocol 2)
Permanent marker
1. Prior to photoablation (Basic Protocol 3), mark a single side of a PVA thin Þlm–
coated dish along the edge of the attached coverslip.
2. Align the marked edge with the front of the microscope stage.
3. Proceed with rest of the photoablation procedure.
This part can be automated.
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
4. Continue with the full coating, blocking, and cell plating process through Basic
Protocol 5, substituting a ßuorescently labeled protein during the adsorption step
(step 4 of Basic Protocol 5).
We suggest that this Þrst labeled protein have a ßuorophore in the visible range of emission,
between 510 and 610 nm. Far red dyes such as Cy5, Alexa Fluor 633 and 647, or Dylight
649 are not recommended for this Þrst stage.
10.15.30
Supplement 45
5. After blocking the ßuorescently labeled protein with 1% heat denatured BSA
(step 14, Basic Protocol 5), perform a second photoablation, if desired.
Here there are several options that may depend on the Þrst adsorbed protein: (1) the
surface of the Þlm can be dried using compressed air immediately before photoablation,
(2) the surface can be left in PBS, or (3) the second ßuorescently labeled protein can be
added to the surface. The photoablation process can still occur in solution (options 2 and
3); however, due to the presence of water, its efÞciency can be reduced.
6. Align the marked side of the photoablated dish with the front of the stage. Use
epißuorescence to scan the area(s) of the dish for the Þrst photoablation site.
7. Once found, realign the dish to the best of your ability by hand. Fast XY scan the
FOV using the most suitable conÞguration for the ßuorophore used.
8. While scanning, make the Þne alignment adjustments using the rotation dialog box
in the Mode panel of the Scan control window. To help with this, open the Edit ROI
window and choose the ROI template that was used to generate the Þrst pattern.
Leave the Edit ROI window open without selecting the ROI tab in the Scan control
window, and continue scanning.
This keeps the ROI template visible while scanning a full FOV.
9. Manually or using the Stage and Focus control window, align the ROI template
over the ßuorescent patterns. If using the same pattern, offset the template from the
original position in either the x or y planes, or both.
The amount of offset will depend on the original ROI.
10. Load the μPP photoablation conÞguration into the AIM software. Photoablate single
FOVs at a time.
We recommend this step not be automated.
11. Once Þnished, follow steps 6 through 8 in Basic Protocol 5, substituting PBS with
0.1% pluronic F-127 for PBS.
12. Following each protein added to the photoablated dish, block the surface with 1%
heat-denatured BSA to prevent nonspeciÞc protein attachment.
Quenching with sodium borohydride is only required after the initial photoablation.
Requenching will reduce ßuorophore ßuorescence and is not recommended.
13. Repeat the process (steps 3 through 5), if needed.
14. Plate cells on the surface as in steps 16 through 25 of Basic Protocol 5.
Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see
Borate buffer, 100 mM, pH 9.0
3.092 g boric acid (powder 99.5%; Sigma)
Add distilled water to 400 ml
Add several solid NaOH pellets at a time while mixing until the pH is ∼9.0
Filter sterilize using a 0.2-μm Þlter
Extracellular
Matrix
10.15.31
Supplement 45
Ethanolamine buffer, 200 mM
4.88 g ethanolamine hydrochloride (EtOHNH3 , crystalline form, 99%; Sigma)
Add sodium phosphate buffer (see recipe) to 250 ml
Filter sterilize using a 0.2-μm Þlter
NIH/3T3 Þbroblast culture medium
440 ml Dulbecco’s modiÞed Eagle’s medium (DMEM, high-glucose modiÞed;
Hyclone)
5 ml penicillin/streptomycin (10,000 U/μg per ml each, respectively; Invitrogen)
50 ml bovine calf serum (BCS; Hyclone)
Sterile Þlter using a 0.2-μm Þlter
Store up to 1 month at 4◦ C
Phosphate-buffered saline (PBS) with 0.1% pluronic F-127
199 ml phosphate-buffered saline (PBS; Hylcone)
1 ml of 20% pluronic F-127 in DMSO (Invitrogen)
Mix well on a stirplate
Sterile Þlter
Warming to ∼37◦ C during mixing will help mix the solutions.
Phosphate-buffered saline (PBS) with penicillin/streptomycin and fungizone
490 ml DPBS/modiÞed containing calcium and magnesium (Hyclone)
5 ml Amphotericin B (250 μg/ml; Invitrogen)
5 ml penicillin/streptomycin (10,000 U/μg per ml each, respectively; Invitrogen)
Mix and store up to 6 months at 4◦ C
Sodium borohydride solution
40 mg sodium borohydride (NaBH4 , hygroscopic powder; Sigma)
1 ml of 1 M NaOH (see recipe)
Mix well
Prepare fresh each time.
Sodium hydroxide, 1 M
20 g NaOH pellets
Distilled water to 500 ml
Mix well
Filter sterilize and store up to 6 months at room temperature
Sodium hydroxide solution, 200 mM
50 ml of 1 M NaOH solution (see recipe)
200 ml of distilled water
Mix well
Filter sterilize
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
Sodium phosphate buffer, 100 mM (pH 8.0)
6.90 g sodium phosphate monobasic (NaH2 PO4 )
Add several solid NaOH pellets at a time while mixing until pH is ∼8.0
Filter sterilize using a0.2-μm pore-size Þlter
10.15.32
Supplement 45
COMMENTARY
Micropatterning using self-assembled
monolayers
Micropatterning of ECM molecules originated by using self-assembled monolayers
(SAMs) of alkanethiolates attached to goldcoated surfaces (Singhvi et al., 1994; Mrksich
et al., 1996). The alkanethiol molecules consist of a sulfhydryl end terminal, a middle
spacer, which is normally an ethylene-glycol
backbone, and a head group that differs from
the end terminal. Sulfhydryls or thiols have a
high afÞnity for electron-dense gold and they
bind, leaving the head groups pointing upward
away from the gold surface. By changing the
head group of the alkanethiol to a hydrophobic
methyl (CH3 ) or a hydrophilic hydroxyl (OH)
group, the surface chemistry is altered; this
will promote or deter ECM protein adsorption,
respectively. Traditionally, in order to physically isolate hydrophobic from hydrophilic regions on a two-dimensional surface, a “rubber stamp” is generated that can physically
ink the hydrophobic alkanethiol onto a gold
surface. The remaining regions are backÞlled
with a hydrophilic alkanethiol, and Þnally an
ECM protein can be added, which will only
attach to the patterned hydrophobic regions
of the surface. This process, known as microcontact printing (μCP), relies mostly on nanolithography techniques to generate a silicon
“master” mold from which the polydemetylsiloxyane (PDMS) stamp is created (Singhvi
et al., 1994).
Poly(vinyl) alcohol properties
As alluded to earlier, PVA is a highly hydrophilic polymer. It consists of a carbon backbone and hydroxyl groups located on every
other carbon. PVA comes in varying molecular
weights (mol. wt.), from as low as 6000 to
>100,000 Da. PVA is generated from the hydrolysis of poly(vinyl) acetate. The percent
hydrolysis that is listed with most PVAs deÞnes the total amount of poly(vinyl) acetate
hydrolyzed to PVA. The percent hydrolyzed
should be as high as possible and is related
to its hydrophilicity, with 98% to 99% being
ideal for this application. With regards to the
mol. wt., the larger the PVA monomer, the
thicker the thin Þlm becomes. We have found
that using any of the molecular weights between 13,000 and 100,000 in a 5% solution
can be used for μPP. Interestingly, after ablation of a high mol. wt. PVA Þlm, the patterns remain visible via phase contrast and DIC
imaging after submersion in buffer. This is not
the case with 13,000 mol. wt. PVA, although
labeling with ßuorescent ECM proteins conÞrms proper local ablation (A.D.D., unpub. observ.). Because of this, low-molecular-weight
PVA thin Þlms are generally better for higher
resolution ßuorescence microscopy, and high
mol. wt. PVA is helpful for visualizing the
ECM patterns at lower magniÞcations.
Photoablation with two-photon microscopy
The process of photoablation is based on
the ability of the PVA polymer to absorb light
in the ultraviolet (UV) range (100 to 380 nm;
Matsumoto et al., 1958). Other large polymers that have the ability to form a hydrogel
such as polyacrylamide and polyethylene glycol can undergo photolytic degradation (Chen
et al., 2003; Yamato et al., 2003). Two-photon
femto-second pulse lasers mimic UV wavelengths using 720- to 760-nm light, and can
excite UV-based ßuorophores such as DAPI,
coumarin, and Hoechst. For more information
on properties and the process of two-photon
excitation and confocal microscopy, we suggest reviewing UNITS 4.5 & 4.11 on confocal
and two-photon excitation microscopy, respectively. Absorption of UV light can initially
result in polymerization of many polymer solutions (Du, 2007). However, continued exposure can disrupt primary bonds; in PVA’s case
the –OH bond to the carbon backbone. Further
exposure results in a breakdown of the carbon
backbone itself. With μPP, we use this property of PVA to locally breakdown or ablate
the thin Þlm, exposing the underlying glass to
which ECM proteins can later be adsorbed.
Troubleshooting
The key elements during the multi-step μPP
process requiring consideration are: (1) the
PVA conjugation to the glass surface, (2) how
the PVA thin Þlm is stored, and (3) several
parameters associated with the two-photon
laser for proper photoablation. Many other
troubleshooting tips are found throughout the
text, where they are directly pertinent to the
protocols.
Improper cleaning and/or activation of the
glass surface can result in thin Þlm detachment. Checking dishes for debris during each
phase of the processing is important and
should not be overlooked.
The PVA thin Þlms need to be hydrated.
PVA hydrogels can undergo crystallization
Extracellular
Matrix
10.15.33
Supplement 45
Generation of
Micropatterned
Substrates
Using Micro
Photopatterning
when dehydrated (Peppas and Merrill, 1977).
While this process greatly increases their
tensile strength, it greatly reduces their hydrophilicity from removal of hydroxyl groups
and, hence, PVA’s ability to deter protein adsorption and cell attachment is compromised.
This will have no effect on the photoablation
process, however. Critical times where dehydration can greatly affect the thin Þlm are initially after spincoating, when the macromolecular monolayer may still be reacting with the
activated glass surface. Short-term exposure to
dry conditions such as during the photoablation on the microscope is tolerated.
Many issues stem from proper maintenance
of the two-photon laser. We found that the
amount of ßuorescently labeled Þbronectin attached to a given area is highly dependent
on the total amount of light energy reaching
the PVA thin Þlm (Doyle et al., 2009). Small
alignment issues of the two-photon source
with the confocal scan head will greatly reduce the light throughput to your sample, and
can result in improper photoablation within the
whole Þeld of view or just a part of it. If the
two-photon system is heavily used by multiple users, more frequent mirror alignments
should be performed. Other issues with uneven photoablation can arise from focus drift,
bubbles in the immersion oil, and an uneven
FOV. Another issue to factor into proper photoablation is the tuning of the TP laser. Technically speaking, the two-photon absorption
of a given wavelength should be the same
between different two-photon sources. However, there can be variation in the best or most
suitable wavelength to maximize wavelength
absorption. As it is recommended by most experts, you should test several wavelengths until
the best one is found for your particular twophoton source to illuminate, or in this case
photoablate, your sample. Starting with 755
nm, tune the Ti:Sapphire laser up or down ten
nanometers at a time. Photoablate a simple pattern such as a Þeld of same-sized dots at the
particular wavelength, and then document the
time taken and whether the ablation was efÞcient (complete removal of the PVA thin Þlm)
or not (only partial removal). Partial photoablation of patterns, where the PVA thin Þlm is
not completely removed from the glass surface, can lead to an inability of ECM protein
adsorption and, hence, can affect cell attachment and/or migration.
urations and ROI templates, you should be
able to generate micropatterns to which ECM
or other proteins of interest can readily adsorb. Once patterns have been produced, you
should Þnd that cells should readily attach
to patterns, especially linear structures (lines
and lanes). There should be limited autoßuorescence from the dish surface and all ßuorescence microscopy techniques, from TIRF
to spinning-disk confocal and two-photon
confocal microscopy, should be effortlessly
performed.
Time Considerations
As described in the Strategic Planning section, the many different parts of μPP can
be performed not only on separate days, but
weeks, if not months, apart between glass activation (Basic Protocol 1) and ECM adsorption/cell attachment (Basic Protocol 5). It is
best to plan accordingly. For example, activating 30 dishes or more depending on your
usage should be enough for 4 weeks. However,
it is not prudent to continue all 30 dishes
through the thin Þlm deposition stage (Basic
Protocol 2), unless you can process 30 dishes
in a single week. Furthermore, how many of
the 30 dishes can be used in a 1 to 3 day period
after ECM adsorption needs also to be considered. Preplanning for dish need and usage
will decrease your issues at key steps. One important time consideration is between formation of the PVA thin Þlm through spincoating
and the addition of the ethanolamine buffer
after photoablation. Although as rare as this
would occur, a short time (<2 hr) between
these two steps could cause release of the thin
Þlm from the glass surface due to a lack of
covalent attachment. Hence, the ethanolamine
buffer should be added as soon as practical
after spincoating.
Literature Cited
Chen, S., Kancharla, V.V., and Lu, Y. 2003. Laserbased microscale patterning of biodegradable
polymers for biomedical applications. Int. J. of
Material & Product Technol. 18:457-468.
Doyle, A.D., Wang, F.W., Matsumoto, K., and
Yamada, K.M. 2009. One-dimensional topography underlies three-dimensional Þbrillar cell
migration. J. Cell. Biol. 184:481-490.
Anticipated Results
Du, J.Z., Sun, T.M., Weng, S.Q., Chen, X.S.,
and Wang, J. 2007. Synthesis and characterization of photo-cross-linked hydrogels
based on biodegradable polyphosphoesters and
poly(ethylene glycol) copolymers. Biomacromolecules 8:3375-3381.
It is expected that after generating the PVA
thin Þlm dishes and creation of the conÞg-
Hermanson, G.T. 1996. Bioconjugate Techniques.
1st ed. Academic Press, San Diego, Calif.
10.15.34
Supplement 45
Lehnert, D., Wehrle-Haller, B., David, C., Weiland,
U., Ballestrem, C., Imhof, B., and Bastmeyer,
M. 2004. Cell behavior on micropatterned substrata: Limits of extracellular matrix geometry
for spreading and adhesion. J. Cell Sci. 117:4152.
Matsumoto, M., Imai, K., and Kazusa, Y. 1958. Ultraviolet spectra of polyvinyl alcohol. J. Polymer
Sci. 117:426-428.
Mrksich, M., Chen, C.S., Xia, Y., Dike, L.E., Ingber,
D.E., and Whitesides, G.M. 1996. Controlling
cell attachment on contoured surfaces with selfassembled monolayers of alkanethiolates on
gold. Proc. Natl. Acad. Sci. U.S.A. 93:1077510778.
Peppas, N.A. and Merrill, E.W. 1977. Development
of semicrystalline poly(vinyl alcohol) hydrogels
for biomedical applications. J. Biomed. Mater.
Res. 11:423-434.
Singhvi, R., Kumar, A., Lopez, G.P.,
Stephanopoulos, G.N., Wang, D.I., Whitesides,
G.M., and Ingber, D.E. 1994. Engineering cell
shape and function. Science 264:696-698.
Yamato, M., Konno, C., Koike, S., Isoi, Y., Shimizu,
T., Kikuchi, A., Makino, K., and Okano, T. 2003.
Nanofabrication for micropatterned cell arrays
by combining electron beam-irradiated polymer
grafting and localized laser ablation. J. Biomed.
Mater. Res. A 67:1065-1075.
Extracellular
Matrix
10.15.35
Supplement 45

"Extracellular Matrix". In: Current Protocols in Cell Biology

Transcription

Similar documents

Lotus Speculos

AF8000E AF10000E manual

Exam link

babies wild - Marina Cano

my portfolio in PDF - Wouter van Faassen | Urbanism

Chrismon Booklet - Gatlinburg Presbyterian Church

Now It`s Bipartisan: Jeanine Pirro and Chris Owens

Vinyl Windows and Doors

Homeless teen murdered for porridge

27 February 2015, PLATINUM WEEKLY, 11 Marais Street, Tel