The role of NuMA in the interphase nucleus

71
Journal of Cell Science 111, 71-79 (1998)
Printed in Great Britain © The Company of Biologists Limited 1998
JCS1506
The role of NuMA in the interphase nucleus
Andreas Merdes1 and Don W. Cleveland1,2,*
1Ludwig
Institute for Cancer Research and 2Division of Cellular and Molecular Medicine, University of California at San Diego,
9500 Gilman Drive, La Jolla, CA 92093-0660, USA
*Author for correspondence
Accepted 17 October 1997; published on WWW 11 December 1997
SUMMARY
NuMA is an essential protein for the formation of spindle
poles in mitosis. During interphase, NuMA is transported
into the nucleus where it resides until prometaphase of the
next mitotic cycle. We tested for a potential function of
NuMA in interphase nuclei that were assembled from
human sperm DNA using frog egg extract immunodepleted
of NuMA. Despite the absence of NuMA, nuclei formed
without visible changes of the chromatin structure,
surrounded by an intact nuclear membrane containing
pores and nuclear lamins. These nuclei were fully
competent to import nuclear substrates and to replicate
their DNA. By screening tissue sections of various organs,
absence of NuMA from the nucleus was observed in a
number of cell types, including sperm, granulocytes in the
blood, and differentiated smooth and skeletal muscle fibers.
Experiments on cultured myoblasts indicated that NuMA
is degraded during muscle cell differentiation. The absence
of NuMA in interphase nuclei of the tissues tested
correlated with a non-spherical, elongated or beaded
nuclear morphology, suggesting that during interphase
NuMA may act as a non-essential nucleoskeletal element.
INTRODUCTION
indeed found to be associated with certain matrix-associated
regions (Luderus et al., 1994). In addition, immunolocalization
studies at the electron microscopic level visualized NuMA in a
filamentous network in the interphase nucleus (Zeng et al.,
1994a), and overexpression of a NuMA mutant that lacked a
functional nuclear localization signal resulted in the formation of
cytoplasmic aggregates with a filamentous substructure (Saredi
et al., 1996). A nucleoskeletal framework provided by NuMA
could contribute to the mechanical stability of the nucleus,
provide attachment sites for the chromatin and nuclear proteins,
serve as tracks along which nuclear components could be
shuttling between the cytoplasm and the nuclear interior (Meier
and Blobel, 1992), or help to maintain the compartmentalization
of nuclear processes as suggested for mRNA splicing (Zeng et
al., 1994b). Also, a system of nuclear filaments might be involved
in tethering the individual chromosomes into a compact mass of
a single nucleus at the end of mitosis.
In this context, several lines of evidence have suggested that
NuMA plays an active role in the reformation of the
postmitotic nucleus by providing cohesion to the segregated
chromatids in anaphase and telophase. Overexpression of
NuMA deletion mutants that were truncated at their amino
terminus or the carboxy terminus resulted in the postmitotic
formation of numerous micronuclei. Also, a mutant hamster
cell line with a defective RCC1 gene showed a similar postmitotic micronuclei formation effect, which could be
suppressed by increased expression of wild-type NuMA
(Compton and Cleveland, 1993).
The nuclear/mitotic apparatus protein (NuMA) is characterized
by its cell cycle-dependent shuttling between the interphase
nucleus and the mitotic spindle (Lydersen and Pettijohn, 1980).
During mitosis, NuMA is an essential component for the
formation and maintenance of mitotic spindle poles (Gaglio et
al., 1995, 1996; Merdes et al., 1996). Spindle assembly
experiments in egg extracts indicated that NuMA acts as a
tethering factor that bundles microtubules and anchors them at
the spindle poles. NuMA is not an integral component of the
centrosome itself, but focuses in a crescent-shaped area
between the centrosome and the spindle equator as the spindle
is being formed. NuMA may achieve its mitotic localization
by the interaction with minus-end directed microtubuledependent motors, as immunoprecipitation experiments have
shown that NuMA associates with the motor protein
cytoplasmic dynein and its activator complex dynactin (Merdes
et al., 1996). During telophase, NuMA releases from the
spindle poles and is transported into the newly formed nucleus
(Yang et al., 1992; Compton et al., 1992).
Little is known about the role of NuMA during interphase.
Biochemically, NuMA represents a protein that is insoluble after
DNase and high salt treatment of the nucleus; it is also resistant
to non-ionic detergents but can be solubilized at high
concentrations of urea (Kallajoki et al., 1991; Harborth et al.,
1995). Thus, NuMA fulfills the biochemical criteria of a nuclear
matrix protein. DNA-binding studies showed that NuMA is
Key words: NuMA, Nucleus, Nuclear matrix, Interphase
72
A. Merdes and D. W. Cleveland
To test for a role of NuMA in the formation of the nucleus
and to examine potential functions of NuMA throughout
interphase, we monitored the formation of nuclei assembled in
frog egg extracts from which NuMA was removed by
immunodepletion. We also tested NuMA-free nuclei for their
ability to assemble a nuclear lamina, to synthesize new DNA,
and to perform nuclear transport. In addition, we tested cells
from various tissues for the presence and localization of
nuclear NuMA.
MATERIALS AND METHODS
Egg extract experiments
Xenopus egg extracts and demembranated sperm were prepared as
described by Murray (1991). Depletion of NuMA from the extracts
was performed as described by Merdes et al. (1996). The depletion
efficiency was monitored by immunoblots of extract aliquots using
anti-Xenopus NuMA antibodies and 125I-Protein A (Merdes et al.,
1996), followed by phosphorimaging. The metaphase-arrested egg
extracts were converted into interphase extracts by addition of 0.4 to
0.8 mM CaCl2. Formation of nuclei in the extracts was monitored by
staining of the DNA with 4′,6′-diamidino-2-phenylindole (DAPI) and
fluorescence microscopy. Immunofluorescence studies were
performed as described earlier (Merdes et al., 1996). Nuclear import
studies were done by adding 0.3 µl of rhodamine-labeled human
serum albumin, coupled to the SV40 large T nucleophilic peptide
sequence (gift from Dr D. J. Forbes’ laboratory, UC San Diego, CA;
Newmeyer and Forbes, 1988), to 10 µl of extract. Nuclear import was
inhibited using wheat germ agglutinin (WGA) at 1 mg/ml. The
amount of imported subtrate was quantified by taking digital images
using a charge-coupled device camera (Princeton Instruments,
Trenton, NJ) and the ‘Toolbar’ and ‘Measure’ functions of the
Metamorph program (Universal Imaging Corp., West Chester, PA).
DNA replication was tested by the incorporation of
bromodeoxyuridine (BrdU) into the nuclei. Initially, 0.5 µl of 2 µM
BrdU was added with the sperm to 15 µl egg extract. At 70 minutes
after CaCl2 addition, nuclei were spun onto coverslips, methanol-fixed
at −20°C, rehydrated, and treated with 1.5 M hydrochloric acid for 30
minutes. For immunofluorescence of the incorporated BrdU, a
monoconal mouse anti-BrdU antibody (Amersham, Arlington
Heights, IL) was used. DNA replication was inhibited using
aphidicolin (Calbiochem, La Jolla, CA) at 200 µg/ml.
Immunofluorescence of the nuclear lamina was performed with the
monoclonal antibody Lo46F7 against Xenopus lamin LIII, donated by
Dr W.W. Franke (German Cancer Research Center, Heidelberg,
Germany; see Benavente et al., 1985). Electron microscopy on nuclei
formed in vitro was done by fixing 20 µl of extract in 500 µl of 1.5%
glutaraldehyde in Sørensen buffer (100 mM sodium phosphate, pH
7.2) for one hour. The nuclei were pelleted in a tabletop
microcentrifuge, washed five times in Sørensen buffer, and postfixed
with 2% osmium tetroxide in Sørensen buffer for one hour. After
washing and dehydration in a graded series of ethanols, the nuclei
were infiltrated with Epon which was subsequently polymerized at
65°C. Ultrathin sections were counterstained with saturated lead
citrate for five minutes and visualized by electron microscopy.
Immunoelectron microscopy using silver-enhanced 1 nm goldconjugated secondary anti-rabbit antibody was performed on nuclei
that were processed by methanol fixation as described for
immunofluorescence of Xenopus NuMA (Merdes et al., 1996). Details
on the secondary reagent and silver enhancement are published by
Yao et al. (1997).
Detection of NuMA in various tissues
Cryosections were cut on a Jung Frigocut 2800 E microtome (Leica
Inc., Deerfield, IL) from various frog tissues that had been immersed
in Tissue-Tek (Sakura Finetek, Torrance, CA) and frozen on dry ice.
The tissue sections were collected on positively charged microscopy
glass slides (Labcraft, Curtin Matheson Scientific, Houston, TX) and
fixed for 10 minutes in methanol at −20°C. Immunofluorescence was
done with a rabbit antibody raised against the proximal portion of
the Xenopus NuMA tail (Merdes et al., 1996). For double
immunofluorescence of NuMA and γ-tubulin, a mouse anti-Xenopus
NuMA tail serum and a polyclonal rabbit anti-γ-tubulin antibody
(gift from Dr Rebecca Heald, EMBL, Heidelberg, Germany) were
used. Secondary antibodies (goat anti-mouse and goat anti-rabbit,
coupled to fluorescein isothiocyanate (FITC) or Texas Red,
respectively) were obtained from Cappel/Organon Teknika (West
Chester, PA). Human and mouse blood, as well as frog sperm were
spotted onto glass coverslips and processed as described above. A
mouse myoblast cell line (C2F3) was obtained from Dr K. Arden
(Ludwig Institute for Cancer Research, San Diego, CA). Cells were
cultured at 37°C and 8% carbon dioxide in Dulbecco’s modified
essential medium (DMEM) containing 20% fetal calf serum.
Myotube formation was induced by changing the medium and
culturing in 2% horse serum in DMEM for several days.
Immunoblots were performed using anti-Xenopus NuMA tail
antibodies or a human anti-NuMA autoimmune serum
(crossreacting with mouse cells and other species by immunoblot),
provided by Dr D. E. Pettijohn (Denver, CO). Control blots were
stained with anti-myosin II monoclonal antibody (M2-42), provided
by Dr T. Pollard’s laboratory (Salk Institute, La Jolla, CA).
RESULTS
To investigate whether NuMA is essential for any aspect of
nuclear formation, we assembled nuclei in vitro by the
addition of sperm chromatin to frog egg extract that was
depleted of its cytoplasmic pool of NuMA. Up to 98% of
NuMA could be removed by adsorption to antibody-coated
Protein A beads, in the majority of the experiments the
depletion efficiency ranged between 85 and 90% (Fig. 1A). To
assemble nuclei, we used demembranated human sperm or
plasmid DNA, because these DNA-sources had no traces of
endogenous NuMA (Fig. 1B and C). In contrast, Xenopus
sperm contained small amounts of NuMA, as described below.
After addition of the DNA to the mitotically arrested egg
extract and a pre-incubation of 10 to 15 minutes, the cell cycle
state was converted into interphase by the addition of 0.8 mM
calcium chloride which caused a loss of cdc2 activity (not
shown). Within 45 minutes, the hypercondensed sperm
chromatin began to swell until it reached the size and shape
of an interphase nucleus, with an average diameter of 15 µm,
both in NuMA-free extracts and in controls (Fig. 2A to C).
The nuclei that were assembled in NuMA-depleted extract
showed no significant immunostaining for NuMA, as shown
by immunofluorescence (Fig. 2C′) as well as immunoelectron
microscopy (Fig. 3A and B). In order to maintain the
antigenicity of the NuMA epitope, it was necessary to use
relatively mild fixation, which did not fully preserve fine
ultrastructural details of the nuclei. To circumvent this
problem, we also processed parallel samples with
glutaraldehyde and osmium tetroxide fixation, to ensure
optimal preservation of the nuclei, and no differences were
found between NuMA-containing and NuMA-free samples
(Fig. 3C and D). In addition to sperm, bacterial plasmid DNA
was also capable of forming nuclei in depleted and untreated
NuMA in the interphase nucleus
Fig. 1. NuMA depletion from frog egg extracts.
(A) Anti-NuMA immunoblot of untreated egg extracts
(untr.), and extracts after incubation with Protein A beads
coupled to control antibodies (mock), or anti-NuMA
antibodies (depl.). NuMA is almost completely absent in
the depleted extract. (B) Coomassie staining and (C)
anti-NuMA immunoblot with the human-specific
monoclonal antibody 1F1, revealing that NuMA is absent
in human sperm, but present in large amounts in a HeLa
cell extract (identical results were obtained with a
polyclonal anti-NuMA antibody, data not shown). The
migration positions of molecular mass standard proteins
(kDa) are indicated.
egg extract, although the sizes of the various nuclei were
highly variable, between 2 to 50 µm in diameter, and their
nuclear envelopes were sometimes discontinuous (Fig. 3E and
F). In all cases, the nuclear envelopes associated with the DNA
contained nuclear pores (Fig. 3G,H), and nuclear lamins from
the extract were incorporated into sperm nuclei, as revealed
by the staining for the frog-specific lamin LIII (Fig. 3I′ to K′).
To test whether these nuclei were capable of importing
nuclear substrates, the uptake of a peptide containing the
nuclear localization signal from the SV40 large T antigen,
coupled to rhodamine-labeled human serum albumin, was
studied in NuMA-depleted and control extracts. As shown in
Fig. 4A′ and C′, all samples stained positively in this assay.
The degree of nuclear import was comparable between NuMAcontaining and NuMA-free nuclei, as photometric
measurements of the imported substrate revealed an average
pixel intensity of 107±36 (n=11) for controls and 123±24 (n=8)
for NuMA-depleted samples. Nuclear import could be
completely blocked in both cases by the addition of wheat
Fig. 2. Nuclear assembly in NuMA-depleted extracts. (A-C) DNA
staining of nuclei assembled from human sperm in untreated (untr.),
mock-depleted and depleted (depl.) egg extract, using the DNAspecific dye DAPI. (A′-C′) Corresponding NuMA staining, revealing
that no NuMA is present in the nuclei formed in depleted extracts
(C′). Bar, 20 µm.
73
kDa
germ agglutinin, a transport inhibitor known to bind to
glycoproteins at the nuclear pores (Fig. 4B′ and D′). The nuclei
treated with wheat germ agglutinin were on the average smaller
than without the inhibitor (compare Fig. 4A and B, C and D).
In addition, DNA-replication was unaffected by the depletion
of NuMA, because all nuclei were able to incorporate the
nucleotide homologue bromodeoxyuridine (BrdU; Fig. 4F and
G). BrdU incorporation was suppressed in presence of
aphidicolin, a drug that specifically inhibits DNA polymerase
α (Fig. 4H′).
Because NuMA appeared to be dispensable during nuclear
formation and for any of the nuclear functions tested, we
investigated various tissues of the frog for the presence and
localization of NuMA in the nucleus, to see whether this
protein is ubiquitous in every cell type. Whereas sections of
most tissues, including liver, lung, brain, skin and intestines
revealed a homogeneous staining of NuMA in the nucleus (Fig.
5A′, showing liver cell nuclei), the immunolabeling of the testis
was positive for spermatocytes, but appeared to be negative for
the sickle shaped, elongated frog sperm nuclei (Fig. 5B′).
However, closer inspection of the sperm indicated that there
was a region at the end of each sperm where concentrated
NuMA staining was detectable, but apparently outside the
nucleus itself (Fig. 5C′). To determine whether this area
contained the sperm centrosome, parallel staining of γ-tubulin
was performed (Fig. 5C′′). Indeed, NuMA and γ-tubulin
seemed to colocalize in this area, suggesting an association of
NuMA with centrosomal material. Artificial swelling of the
sperm confirmed the exclusive NuMA localization at the sperm
end but did not reveal any NuMA signal on the DNA (Fig. 5D′),
indicating that the lack of staining was not due to a
permeability problem of the NuMA-antibody into the
hypercondensed chromatin. Also, the antibody reacted
specifically with the 240 kDa-sized NuMA protein band by
immunoblotting of frog sperm (Fig. 5J), assuring that the
immunofluorescence labeling was not the result of antibody
crossreactivity. In contrast to frog sperm, there was no NuMA
at all detected in human sperm, as determined by
immunoblotting (Fig. 1) and immunofluorescence (not shown).
We reasoned that NuMA might be absent from the sperm
chromatin to allow the sperm DNA to undergo
hypercondensation, or to adapt to the non-spherical nuclear
74
A. Merdes and D. W. Cleveland
Fig. 3. A nuclear envelope forms around NuMAfree nuclei. (A-B) Immunoelectron microscopy of
NuMA in extracted nuclei assembled from
human sperm in (A) control and (B) NuMAdepleted extracts. Sections at the periphery of the
nucleus are shown. NuMA was visualized with
silver-enhanced immunogold (indicated by
arrowheads). (C-D) Electron micrographs of
unextracted nuclei formed from human sperm
DNA in (C) undepleted and (D) NuMA-depleted
extracts. (E-F) Electron microscopy of nuclei
formed from bacterial plasmid DNA in (E)
undepleted and (F) NuMA-depleted extracts.
(G) Electron micrograph of a nucleus from sperm
DNA incubated in NuMA-depleted extract,
containing nuclear pores (arrows). (H) Electron
micrograph of a small nucleus assembled from
bacterial plasmid DNA, containing a closed
nuclear membrane with nuclear pores
(arrowheads). (I-K′) Human sperm acquires a
nuclear lamina from frog lamin LIII in undepleted
and NuMA-depleted extracts. (I-K) DNA staining with DAPI of human sperm (I) before incubation in frog egg extract, and after 60 minutes in
calcium chloride treated interphasic extract that was (J) mock-depleted and (K) NuMA-depleted. The corresponding immunofluorescence of
lamin LIII is shown in (I′-K′). Bars: (B,D,F,G,H), 1 µm; (K′), 20 µm. (+) NuMA-containing; (−) NuMA-depleted egg extract.
shape in the case of frogs and to improve sperm motility due
to a less rigid nucleoplasm. To see whether other cell types
with non-spherical nuclei were also devoid of NuMA, we
continued our search in blood cells. One class of leucocytes in
human blood, the granulocytes, represent cells with clearly
non-spherical nuclei. We tested for NuMA using the previously
characterized monoclonal antibody 1F1 (Compton et al.,
1991). Immunofluorescence of eosinophile granulocytes
NuMA in the interphase nucleus
75
Fig. 4. Nuclear import takes
place in NuMA-free nuclei. (AD) DNA staining with DAPI of
sperm nuclei in (A,B) undepleted
and (C,D) NuMA-depleted egg
extracts. (A′-D′) Corresponding
images showing the rhodamine
fluorescence of a labeled nuclear
import substrate, synthesized
from human serum albumin
coupled to the SV40 large T
nuclear signal peptide. The
import is inhibited by wheat
germ agglutinin (WGA), as
shown in B′ and D′. Bar, 20 µm.
revealed that NuMA was absent from their two-lobed nuclei,
whereas a diffuse cytoplasmic NuMA signal was detected (Fig.
5E′). The same result could be obtained in granulocytes from
mouse blood (not shown). In contrast, lymphocytes from the
same human blood preparation contained spherical nuclei and
showed a strong and homogeneous nuclear NuMA staining
(Fig. 5F′). Besides sperm and granulocytes, another cell type
was found to contain NuMA in an unconventional distribution.
Sections of skeletal muscle of the frog hind leg revealed an
exclusively granular or punctate NuMA staining in the
elongated muscle cell nuclei (Fig. 5G′). To see whether muscle
fibers other than from skeletal muscle showed a similar NuMA
staining pattern, sections of the heart and the stomach (the
latter containing smooth muscle cells) were investigated. As in
skeletal muscle, these samples also revealed a weak punctate
immunofluorescence of NuMA in all nuclei of muscle fibers
(Fig. 5H′ and I′).
Immunoblotting of several tissues revealed that NuMA was
present in liver and lung at a slightly lower molecular mass
than in frog sperm and eggs (Fig. 5J), perhaps indicating a
difference in the phosphorylation state of NuMA due to their
different cell cycle state or an alternatively spliced isoform
(Tang et al., 1993; Cleveland, 1995). Muscle tissue, however,
showed no detectable amounts of full length NuMA by
immunoblotting, but revealed the presence of reactive protein
bands at ~100 kDa and below (Fig. 5J). This indicated that the
granular NuMA signal seen inside muscle fiber nuclei most
likely represents NuMA degradation products.
Finally, to test whether concentration of NuMA in nuclear
foci is a general feature of muscle fibers, we cultured mouse
myoblasts and induced the formation of myotubes in vitro
by serum starvation. Fig. 6B shows that once multinucleate
myotubes are formed, NuMA is no longer homogeneously
distributed over the nucleoplasm, but a major proportion is
concentrated in focal areas inside the nucleus (Fig. 6B′′).
These NuMA foci also colocalize with regions of high
chromatin condensation, as revealed by parallel staining of
the DNA with DAPI (Fig. 6B′). Thus, the particular NuMA
staining seen in muscle tissue is recapitulated during
differentiation in vitro. Immunoblotting of cells after two
days of differentiation revealed that NuMA is only present
in a lower molecular mass form (approximately 200 kDa)
and at a significantly reduced amount. Continued culturing
of the myotubes resulted in a complete loss of the NuMA
signal (Fig. 6C). We therefore conclude that NuMA is
degraded in maturing muscle fiber cells and that the degraded
form is accumulated in the nuclear foci seen by
immunofluorescence.
76
A. Merdes and D. W. Cleveland
Fig. 5. NuMA localization in
cells from various tissues. (AI) DNA staining with DAPI.
(A′-I′) Immunofluorescence of
NuMA. Fluorescence
micrographs of tissue sections
from (A) frog liver, (B) frog
testis, (G) frog hindleg
muscle, (H) frog heart muscle,
(I) frog stomach, as well as
preparations of (C) untreated
and (D) swollen frog sperm,
(E) an eosinophile granulocyte
from human blood and (F) a
human lymphocyte. (C′′) The
immunofluorescence of γtubulin in frog sperm,
corresponding with the
NuMA-labeling in C′ and the
DNA-staining in C. (J) A
NuMA-immunoblot of frog
egg extract, frog liver, frog
lung, frog sperm and frog leg
muscle. The positions of
molecular mass standards are
indicated. Bars: in F′ (for
A,A′,B,B′,D,D′,E,E′,F,F′) and
I′ (for C,C′,C′′,G,G′,H,H′,I,I′),
20 µm.
DISCUSSION
Previous work has shown that NuMA is a fairly abundant
protein with 2×105 copies per cell (Compton et al., 1992), and
that it localizes exclusively to the nucleoplasm in interphase
cells in culture, as well as in cells from adult tissues, such as
skin, liver, brain (Kallajoki et al., 1992), esophagus, cornea,
and epidermis (Tang et al., 1993). Surprisingly, in this report
we have found that several cell types in the adult organism,
including sperm cells, granulocytes in mammalian blood and
muscle fiber cells all lack regularly distributed NuMA in the
nucleus. Common to all these cell types is the non-spherical
NuMA in the interphase nucleus
77
Fig. 6. NuMA is degraded during muscle fiber maturation and enriched in intranuclear granules. (A,B) Phase contrast light micrographs of
myoblasts and fused myotubes, respectively. (A′,B′) Corresponding DNA-staining with DAPI. (A′′,B′′) NuMA-immunofluorescence in the
nuclei of (A′′) myoblasts and (B′′) fused, multinuclear myotubes. (A′′,B′′) Identically exposed. (C) Immunoblots of myoblasts, myotubes after
two and four days in culture, and HeLa cell extract. Top: a human autoimmune antibody against NuMA, crossreacting with NuMA in mouse
myoblasts, was used. Bottom: immunoblot using an anti-myosin II antibody, indicating that comparable amounts of protein was loaded in all
samples. Myosin II overexpression was not yet visible at the indicated stages of cell culture. Bar, 20 µm.
appearance of their nuclei. In line with the mounting evidence
which indicates that NuMA constitutes a nucleoskeletal
element (Zeng et al., 1994a; Saredi et al., 1996, 1997), our
findings suggest that NuMA may also be involved in defining
the nuclear shape in interphase cells and that the absence of
NuMA may allow the cell to modulate the nuclear
architecture to adapt to specific functions upon
differentiation. In frog sperm, the sickle-shaped nucleus may
need to retain a certain degree of elasticity in order to allow
sperm motility. If a nucleoskeletal element such as NuMA
were present in sperm, it would probably increase the
mechanical rigidity of the cell, placing such sperm at a
potentially serious disadvantage during fertilization. An
alternative, but not mutually exclusive possibility, is that
NuMA could have been excluded from the sperm nucleus to
allow dense packing of the hypercondensed DNA. Consistent
with these roles, frog sperm contained some NuMA protein,
but this was outside the nucleus, most likely associated with
centrosomal material. This small amount of NuMA could be
supporting the organization of the sperm-induced
microtubule aster after fertilization of the frog egg, and
ensure the anchoring of those microtubules to the center of
the aster that have disconnected from the centrosomal
nucleation origin (Merdes and Cleveland, 1997).
NuMA in muscle cells appears to be degraded during muscle
fiber maturation. On one hand, it is possible that degeneration
of nuclear material is a general property of the mature muscle
fibers. Consistent with this, muscle fiber cells are unable to
proliferate and cannot regenerate themselves, but are replaced
by satellite cells induced to become myocytes (Snow, 1977).
On the other hand, NuMA degradation could be part of an
active process to adapt the nucleus to the mechanical
requirements of the muscle fiber cell. It has been documented
that in smooth muscle fibers, the nucleus retains an elongated
shape with a smooth contour as long as the fiber is relaxed. As
the muscle fiber contracts, the nucleus becomes shortened in
length and shows deep clefts and convolutions of its surface
(Lane, 1965), apparently due to the deforming forces of the
contracting cell. Similarly, as in the sperm cell, the presence of
rigid nuclear skeletal elements such as NuMA (Zeng et al.,
1994a) would most likely interfere with such nuclear plasticity.
Release of NuMA to allow modulation of the nuclear shape
may also play a role in the differentiation of myeloid stem cells
into granulocytes, which display a very unusual nuclear
78
A. Merdes and D. W. Cleveland
morphology characterized by two large lobes interconnected
by a small nucleoplasmic bridge in eosinophiles. However, the
reason for remodeling the granulocyte nucleus is not known.
In our examples of NuMA-free nuclei, the absence of NuMA
could be due to a lack of expression, specific posttranslational
modifications, or by proteolytic degradation, as shown in
muscle fiber cells. The machinery that leads to the proteolysis
of NuMA may involve some of the recently described ‘death
proteases’ or ‘caspases’ that are involved in apoptotic nuclear
degradation (Earnshaw, 1995). There is evidence that apoptosis
does involve proteolysis of NuMA (Weaver et al., 1996; Hsu
and Yeh, 1996) and that the apoptotic degradation products are
of similar size as the NuMA forms found in myotubes. Perhaps
degradation of nuclear proteins in terminally differentiated
cells such as granulocytes or muscle fibers reflects a step that
precedes the execution of apoptotic cell death upon further
specific signals. This would explain the rapid turnover that
granulocytes undergo upon parasite infection or muscle fibers
upon muscle injury. The absence of NuMA from the nucleus,
however, is not always linked to nuclear degradation or future
cell death, as NuMA-free sperm cells give rise to a complete
new organism following fertilization of the egg. Taken
together, our results presented here, as well as earlier reports
document that NuMA is normally present in regular interphase
nuclei, but not in all cell types. Thus, if NuMA plays a role as
a structural nuclear component in interphase, its function must
be non-essential and might be redundant with other structural
elements, such as the nuclear lamins, chromatin binding
proteins, or other, yet unidentified components (Georgatos,
1994). This idea is consistent with our finding that nuclei of
normal spherical shape are formed in the absence of NuMA in
frog egg extracts which allow nuclear formation in the test tube
without any mechanical constraints of the cytoskeleton, plasma
membranes, or surrounding tissue. Furthermore, these
experiments show that neither nuclear transport nor
incorporation of proteins such as lamin LIII into the nuclear
scaffold require NuMA. Although the depletion of the
structural nuclear envelope protein lamin LIII inhibits DNA
replication (Newport et al., 1990; Meier et al., 1991), our
immunodepletion data demonstrate that NuMA is not essential
for this process.
An additional proposal for NuMA-function in the nucleus is
derived from previous experiments showing that overexpression
of head- or tail-less NuMA mutants in cultured mitotic cells
causes the formation of micronuclei when the cell exits from
mitosis (Compton and Cleveland, 1993). These data suggest
that NuMA plays an important role in the postmitotic
reformation of the nucleus. In contrast, the data presented in this
paper demonstrate that nuclear formation is independent of the
presence of NuMA, and NuMA-free nuclei show no detectable
alterations in the ultrastructure of both the chromatin and the
nuclear envelope. Certainly, the system used to assay nuclear
formation in our current work, i.e. the formation of a nucleus
from a compact mass of sperm chromatin, may be different
from the postmitotic nuclear reformation around multiple
individual chromosomes. Yet, our data are fully consistent with
reports by Yang and Snyder (1992) and Kallajoki et al. (1993),
who both observed the formation of normal daughter nuclei
after mitosis, despite the apparently complete sequestration of
NuMA in the cytoplasm or at the centrosomes, caused by
antibodies microinjected during anaphase. In addition,
immunofluorescence of NuMA in telophase cells indicates that
the formation of daughter cell nuclei precedes nuclear import
of NuMA (Yang et al., 1992). Therefore, the most likely
explanation of the post-mitotic micronuclei formation after
expression of NuMA mutants (Compton and Cleveland, 1993)
or after NuMA-antibody injection into PtK2 cells prior to
anaphase (Kallajoki et al., 1993) is a disruptive effect on the
mitotic spindle resulting in inefficient chromosome tethering to
the pole late in anaphase and/or telophase, rather than a direct
effect on the formation of the nucleus. Consistent with this, it
has been clearly demonstrated that NuMA is essential during
the formation of mitotic spindles by tethering microtubules at
the spindle poles (Merdes et al., 1996; Gaglio et al., 1995,
1996). Finally, the question why NuMA is localized to the
interphase nucleus in most cells still remains to be answered.
Probably, incorporation of NuMA into the nucleus after
completion of mitosis may allow the cell to segregate this
protein to avoid interference of NuMA with the building of a
functional interphase microtubule network. Overexpression
studies of wild-type NuMA in cultured cells have revealed that
excess NuMA that is not imported into the interphase nucleus
can indeed bind to microtubules in the pericentrosomal area and
in some instances reduce the density of the microtubule network
(Gueth-Hallonet et al., 1996). In this view, the interphase
nucleus provides a storage compartment from which NuMA is
released upon mitotic phosphorylation and nuclear envelope
breakdown in prometaphase.
We thank Drs K. Arden, D. J. Forbes, W. W. Franke, R. Heald, D.
E. Pettijohn, and T. Pollard for their generous donations of antibodies
and cultured cells. We thank our colleagues in the laboratory for their
help and stimulating discussions. This work was supported by NIH
grant R37 GM 27036 to D.W.C. A.M. was supported, in part, by an
EMBO long-term fellowship. Salary support for D.W.C. is provided
by the Ludwig Institute for Cancer Research.
REFERENCES
Benavente, R., Krohne, G. and Franke, W. W. (1985). Cell type-specific
expression of nuclear lamina proteins during development of Xenopus laevis.
Cell 41, 177-190.
Cleveland, D. W. (1995). NuMA: a protein involved in nuclear structure,
spindle assembly, and nuclear re-formation. Trends Cell Biol. 5, 60-64.
Compton, D. A., Yen, T. J. and Cleveland, D. W. (1991). Identification of
novel centromere/kinetochore-associated proteins using monoclonal
antibodies generated against human mitotic chromosome scaffolds. J. Cell
Biol. 112, 1083-1097.
Compton, D. A., Szilak, I. and Cleveland, D. W. (1992). Primary structure of
NuMA, an intranuclear protein that defines a novel pathway for segregation
of proteins at mitosis. J. Cell Biol. 116, 1395-1408.
Compton, D. A. and Cleveland, D. W. (1993). NuMA is required for the
proper completion of mitosis. J. Cell Biol. 120, 947-957.
Earnshaw, W. C. (1995). Nuclear changes in apoptosis. Curr. Opin. Cell Biol.
7, 337-343.
Gaglio, T., Saredi, A. and Compton, D. A. (1995). NuMA is required for the
organization of microtubules into aster-like mitotic arrays. J. Cell Biol. 131,
693-708.
Gaglio, T., Saredi, A., Bingham, J. B., Hasbani, M. J., Gill, S. R., Schroer, T.
A. and Compton, D. A. (1996). Opposing motor activities are required for
the organization of the mammalian mitotic spindle pole. J. Cell Biol. 135,
399-414.
Georgatos, S. D. (1994). Towards an understanding of nuclear morphogenesis.
J. Cell Biochem. 55, 69-76.
Gueth-Hallonet, C., Weber, K. and Osborn, M. (1996). NuMA: a bipartite
nuclear location signal and other functional properties of the tail domain.
Exp. Cell Res. 225, 207-218.
NuMA in the interphase nucleus
Harborth, J., Weber, K. and Osborn, M. (1995). Epitope mapping and direct
visualization of the parallel, in-register arrangement of the double-stranded
coiled-coil in the NuMA protein. EMBO J. 14, 2447-2460.
Hsu, H. L. and Yeh, N. H. (1996). Dynamic changes of NuMA during the cell
cycle and possible appearance of a truncated form of NuMA during
apoptosis. J. Cell Sci. 109, 277-288.
Kallajoki, M., Weber, K. and Osborn, M. (1991). A 210 kDa nuclear matrix
protein is a functional part of the mitotic spindle; a microinjection study
using SPN monoclonal antibodies. EMBO J. 10, 3351-3362.
Kallajoki, M., Weber, K. and Osborn, M. (1992). Ability to organize
microtubules in taxol-treated mitotic PtK2 cells goes with the SPN antigen
and not with the centrosome. J. Cell Sci. 102, 91-102.
Kallajoki, M., Harborth, J., Weber, K. and Osborn, M. (1993).
Microinjection of a monoclonal antibody against SPN antigen, now
identified by peptide sequences as the NuMA protein, induces micronuclei in
PtK2 cells. J. Cell Sci. 104, 139-150.
Lane, B. P. (1965). Alterations in the cytologic detail of intestinal smooth
muscle cells in various stages of contraction. J. Cell Biol. 27, 199-213.
Luderus, M. E., den Blaauwen, J. L., de Smit, O. J., Compton, D. A. and
van Driel, R. (1994). Binding of matrix attachment regions to lamin
polymers involves single-stranded regions and the minor groove. Mol. Cell.
Biol. 14, 6297-6305.
Lydersen, B. K. and Pettijohn, D. E. (1980). Human-specific nuclear protein
that associates with the polar region of the mitotic apparatus: distribution in a
human/hamster hybrid cell. Cell 22, 489-499.
Meier, J., Campbell, K. H. S., Ford, C. C., Stick, R. and Hutchison, C. J.
(1991). The role of lamin LIII in nuclear assembly and DNA replication, in
cell-free extracts of Xenopus eggs. J. Cell Sci. 98, 271-279.
Meier, U. T. and Blobel, G. (1992). Nopp140 shuttles on tracks between
nucleolus and cytoplasm. Cell 70, 127-138.
Merdes, A., Ramyar, K., Vechio, J. D. and Cleveland, D. W. (1996). A
complex of NuMA and cytoplasmic dynein is essential for mitotic spindle
assembly. Cell 87, 447-458.
Merdes, A. and Cleveland, D. W. (1997). Pathways of spindle pole formation:
different mechanisms; conserved components. J. Cell Biol. 138, 953-956.
Murray, A. W. (1991). Cell cycle extracts. In Methods in Cell Biology, vol. 36
(ed. B. K. Kay and H. B. Peng), pp. 581-605. Academic Press, San Diego,
California.
79
Newport, J. W., Wilson, K. L. and Dunphy, W. G. (1990). A laminindependent pathway for nuclear envelope assembly. J. Cell Biol. 111, 22472259.
Newmeyer, D. D. and Forbes, D. J. (1988). Nuclear import can be separated
into distinct steps in vitro: nuclear pore binding and translocation. Cell 52,
641-653.
Saredi, A., Howard, L. and Compton, D. A. (1996). NuMA assembles into an
extensive filamentous structure when expressed in the cell cytoplasm. J. Cell
Sci. 109, 619-630.
Saredi, A., Howard, L. and Compton, D. A. (1997). Phosphorylation
regulates the assembly of NuMA in a mammalian mitotic extract. J. Cell Sci.
110, 1287-1297.
Snow, M. H. (1977). Myogenic cell formation in regenerating rat skeletal
muscle injured by mincing II. An autoradiographic study. Anat. Rec. 188,
201-218.
Tang, T. K., Tang, C. C., Chen, Y. L. and Wu, C. W. (1993). Nuclear proteins
of the bovine esophageal epithelium II. The NuMA gene gives rise to
multiple mRNAs and gene products reactive with monoclonal antibody W1.
J. Cell Sci. 104, 249-260.
Weaver, V. M, Carson, C. E., Walker, P. R., Chaly, N., Lach, B., Raymond,
Y., Brown, D. L. and Sikorska, M. (1996). Degradation of nuclear matrix
and DNA cleavage in apoptotic thymocytes. J. Cell Sci. 109, 45-56.
Yao, X., Anderson, K. L. and Cleveland, D. W. (1997). The microtubuledependent motor centromere-associated protein E (CENP-E) is an integral
component of kinetochore corona fibers that link centromeres to spindle
microtubules. J. Cell Biol. 139, 435-447.
Yang, C. H. and Snyder, M. (1992). The nuclear-mitotic apparatus protein is
important in the establishment and maintenance of the bipolar mitotic spindle
apparatus. Mol. Biol. Cell. 3, 1259-1267.
Yang, C. H., Lambie, E. J. and Snyder, M. (1992). NuMA: an unusually long
coiled-coil related protein in the mammalian nucleus. J. Cell Biol. 116, 13031317.
Zeng, C., He, D. and Brinkley, B. R. (1994a). Localization of NuMA protein
isoforms in the nuclear matrix of mammalian cells. Cell Motil. Cytoskel. 29,
167-176.
Zeng, C., He, D., Berget, S. M. and Brinkley, B. R. (1994b). Nuclear-mitotic
apparatus protein: A structural protein interface between the nucleoskeleton
and RNA splicing. Proc. Nat. Acad. Sci. USA 91, 1505-1509.