Membr. and Cell Biol., 1997, Vol. 10 (5),
pp. 503-513
Reprints available directly from the
publisher Photocopying permitted by
license only
© 1997 OPA (Overseas Publishers Association)
Amsterdam B. V. Published in The Netherlands
by Harwood Academic Publishers GmbH
Printed in Malaysia
Polyclonal Antibodies against Human γ-Tubulin Stain
Centrioles in Mammalian Cells from Different Tissues
Yu. A. Komarova, E. V. Ryabov, I. B. Alieva, R. E. Uzbekov, S.
V. Uzbekova, and L. A. Vorobjev
Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow Institute
of Agricultural Biotechnology, Russian Academy of Agricultural Sciences, Moscow
Rabbit polyclonal antibodies were raised against the C-terminal fragment (amino acid residues
318-451) of human γ-tubulin. These antibodies were used to stain cultured cells of various tissues
(epithelium, nervous tissue, fibroblasts) from different animals (human, monkey, pig, rat, kangaroo rat,
mouse, hamster, chicken, triton). The antibodies specifically stained centrioles in the interphase and
mitotic cells of mammals, but not birds (chicken) or amphibians (newt). In the interphase cells,
centrioles were stained as a pair of dots (or as a double dot) in 96-97% of the cells. The distances
between the maternal and filial centrioles varied in different cultures. Procentrioles were stained in
certain cells, but with less intensity than mature centrioles. In mitotic cells, the antibodies revealed
two spots corresponding to two mitotic poles. The spots in mitosis were significantly larger than the
interphase dots, but the staining was more faint. In spontaneous tripolar mitoses, only two poles
were stained. Thus, it was shown that, on the one hand, γ-tubulin is associated with centrioles irrespective of whether or not they serve as the microtubule organizing centres and, on the other hand, γtubulin might not be an essential component of the microtubule organizing centres.
(Received 22 February, 1996)
Gamma-tubulin is a minor protein, which appears to play an essential role in the nucleation of
microtubules on the centrosome, was first described by Oakley [1,2] who studied the
filamentous fungus Aspergillus nidulans. γ-Tubulin was shown subsequently to be a highly
conservative protein whose amino acid sequence remains much the same even in such
distant species as human, fruit fly and Aspergillus nidulans [3] and is virtually identical in
different species of vertebrates [4]. Thus, the homology between human and frogy-tubulins is
as high as 98% [3]. At the same time, the homology of γ-tubulin with α-tubulin and (3tubulin does not exceed 40% [4].
Localization of γ-tubulin in the cell was first described also while studying Aspergillus
nidulans. Antibodies to γ-tubulin stained corpuscles of mitotic spindle, which are microtubule organizing centres in fungi, and in some cases insignificant quantities of microtubules [5]. In subsequent years, γ-tubulin was found in the cell centrosome of human [6],
fruit fly [6, 7], mouse [6, 8], frog [9] and other cell lines [9-11].
Ultrastructural studies showed that γ-tubulin is a component of the pericentriolar material of
centrosome [9] as well as of some other centres of microtubule organization [12]. It is
noteworthy that γ-tubulin is revealed in the centrosome at all stages of the cell cycle [10].
503
504
Yu. A. KOMAROVA et al.
Links of γ-tubulin with microtubules were also investigated. Microtubules were shown to
contain a very low quantity of γ-tubulin [13]. γ-Tubulin is indispensable for the assembly of
microtubules on centrosomes in vivo [5, 8, 12] and possibly in vitro [8]. Furthermore,
association of γ-tubulin was shown to be preceded by polymerization of microtubules on
basal corpuscles in the course of fertilization [14].
Based on the above facts, Oakley proposed a hypothesis of γ-tubulin functions in vivo [3].
According to this hypothesis, by forming rings on the surface of microtubule organizing
centre, the molecules of γ-tubulin bind to those of β-tubulin from α-, β-tubulin dimers, and
namely this "foundation" is used for the subsequent build-up of the microtubule wall.
Thus, it may be suggested that γ-tubulin is associated with centrosome only when it
functions as the polymerization centre of microtubules. At the same time, numerous reports
indicate that in cells with the centrosome containing two centrioles, staining revealed two
dots of equal intensity [6,8]. However, the maternal and filial centrioles are known to differ
functionally - the number of microtubules spreading from the maternal centriole is always
considerably higher than that from the filial centriole [15,16]. Moreover, γ-tubulin is also
revealed in basal corpuscles [11] which, as a rule, do not serve the purpose of organization
centres of cytoplasmic microtubules.
This work was aimed at investigating the distribution of γ-tubulin in various centres of
microtubule organization centres in the interphase and mitotic cells.
EXPERIMENTAL
Cell Cultures. Epithelial cells from swine kidney embryo extract (SKEE) were cultured at
37°C and 5% CO2 on medium 199 supplemented with 10% bovine serum and antibiotics
(streptomycin with penicillin, or gentamycin). Cultures of HeLa cells and mouse
L-fibroblasts were grown at 37°C and 5% CO2 on medium 199 supplemented with 10%
embryonal bovine serum and gentamycin. Culture of BiiD-FAF (clone 237) cells from
Chinese hamster was grown at 37°C and 5% CO2 on Eagle's medium supplemented with
10% bovine serum and antibiotics (streptomycin with penicillin, or gentamycin). Vero cells
from green monkey and PtKl cells from kangaroo rat were cultured at 37°C and 5% CO2 on
DMEM/F-12 HAM medium (Sigma, USA) amended with 10% embryonal bovine
serum and gentamycin.
Primary culture of cells from newt lung was kindly supplied by Prof. Yu. A. Chentsov
(Moscow State University). Primary cultures of human, rat and chicken fibroblasts were
kindly supplied by Dr. S. R. Speranskaya (Department of Cytology and Histology, Moscow
State University). Primary culture of rat cerebellum cells on glial sublayer was kindly
donated by Dr. N. K. Isaev (Brain Institute, Russian Academy of Sciences).
Antibodies. Purified (by affinity chromatography) polyclonal antibodies against C-terminal peptide of human γ-tubulin were kindly donated by Dr. R. Kuriyama (Minneapolis,
USA); monoclonal antibodies against acetylated tubulin were the kind gift by Prof. K. Gull
(Manchester, UK). Monoclonal antibodies against a-tubulin (DM-1A) and FITSconjugated secondary antibodies were purchased from Sigma. Texas-Red-conjugated
secondary antibodies were kindly donated by Prof. B. Breton (Rennes, France).
Cloning of a Fragment of Human γ-Tubulin Gene and its Expression in E. coli. To obtain
polyclonal antibodies against γ-tubulin, we used a recombinant protein with 134 amino acid
POLYCLONAL ANTIBODIES AGAINST HUMAN γ-TUBULIN
505
residues corresponding to the C-terminal moiety of human γ-tubulin (amino acid residues
from 318 to 451) (GeneBank M61764 database). The selected segment of γ-tubulin is
immunogenic and its sequence differs from respective segments of α- and β-tubulin
(GeneBank M61764).
The fragment encoding the above C-terminal segment of human γ-tubulin was amplified
by the PCR method according to the Biomaster protocol on a matrix of human rDNA
mRNA isolated from a culture of HeLa cells. Oligonucleotide primers for the amplification
were designed on the basis of data on the mRNA sequence of human γ-tubulin (GeneBank
M61764). The direct primer 5-TCGCATGCATCATCGCCATCCTCAACATCATC-3 includes
the Sphl site fused with the sequence corresponding to positions from 975 to 996 of γtubulin mRNA, the reverse primer 5-ACAAGCTTGCCAACCAGTAAGGCAGATGAG-3
carries the Hindlll site fused with the sequence complementary to positions from 1401 to
1421 of γ-tubulin mRNA. The PCR product was treated with Sphl and Hindlll restrictases
and cloned over the indicated sites into the pQE30 (DIAGEN GmbH, FRG).
Transformation of M15[pREP4] E. coli cells by recombinant plasmids, selection of clones,
induction of the recombinant protein expression and its chromatographic purification were
performed according to the DIAGEN protocol.
Immunization of Rabbits. Rabbits of Grey Giant race were initially immunized
intracutaneously at 20 points with the antigen mixed in 1:1 ratio with complete Freund's
adjuvant (to a final volume of 1.5-2 ml). The second and third injections were administered at
one-week intervals, using in both cases incomplete Freund's adjuvant. One third of the
mixture was injected subcutaneously, whereas the remaining quantity was administered as
intramuscular injections into the hind paws. The quantity of antigen introduced as a result
of three immunizations was 200-300-400 µg, respectively. Blood was sampled for one
month after immunization at 7-8-day intervals. Reimmunization was performed three
months after the primary immunization as intracutaneous injections at 20 points with an
antigen mixed with incomplete adjuvant. The maximal liter of antisera (designated further as
R2 and R4) was achieved in 90 days after reimmunization and amounted to 103.
Immunoblotting. Cultured HeLa cells were used for immunoblotting. The total
preparation was obtained in the following manner: cells were removed from glass with
versene, sedimented in a serum-free medium and solubilized in SDS buffer. Proteins were
separated by electrophoresis in 12% PAAG by the Laemmli method [17].
To obtain the cytoskeletal fraction, cultured cells were lyzed in a mixture containing
Triton X-100 under the microtubule-stabilizing conditions (in a buffer containing (mM):
phosphate buffer (pH 6.8), 25; EGTA, 1; EDTA, 0.1; MgCl2, 1, β-mercaptoethanol, 0.1;
PMSF, 1). Then the cells were treated with denaturing 8 M urea solution followed by
dialysis against water with subsequent concentration of material with the help of PEG
(molecular mass, 20 kDa). The concentrated material was treated with SDS buffer.
A mixture of α- and β-tubulins from cattle brain was kindly donated by Dr. E. S.
Nadezhdina (Institute for Protein Research, Russian Academy of Sciences).
Western immunoblotting was conducted using a standard procedure [7]. Binding of
antibodies was revealed by secondary antibodies conjugated with peroxidase (Sigma). oDianisidine was used as a substrate for peroxidase.
Immunofluorescent Studies. The cells intended for immunofluorescent analysis were
lysed before their fixation in a mixture containing 1 % Triton X-100 under the microtubulestabilizing conditions. Glass slips with cells were taken from a Petri dish, washed several
506
Yu.A. KOMAROVA el al.
FIGURE 1. Western immunoblotting with antibodies to γ-tubulin (bands 1-4) and
preimmune serum (5): mixture of isolated α- and β-tubulin, staining with Amide Black; 2,
preparation of proteins from the cytoskeletal fraction of HeLa cells; 3, preparation of total
protein of HeLa cells; 4, mixture of α- and β-tubulin from bovine brain; 5, preparation of
proteins from the cytoskeletal fraction of HeLa cells.
times with saline phosphate (physiological) buffer (pH 7.2) at 37°C and then lysed for 15
min in a solution containing (mM); imidazole, 50 (pH 6.8); MgCl2, 5; EGTA, 1; EDTA,
0.1; 35% glycerol and 1% Triton X-100 (Sigma). Then the cells were rinsed with the same
solution but without glycerol and Triton X-100 and fixed in two different modes: (a) with
4% formaldehyde solution for 30 min; (b) with 1% glutaraldehyde (Merck, FRG) solution in
phosphate buffer for 30 min at room temperature with subsequent treatment with NaBbb,
solution (2 mg/ml, 10 min). The third version consisted in fixation with methanol solution
(for 6 min at -20°C) without preliminary lysis of cells.
After all fixation modes, 1% bovine serum albumin solution in phosphate buffer was
added to prevent background glowing. Then the cells were washed several times with
buffer and subjected to successive immuno-cytochemical staining with antibodies to γtubulin and antibodies to α-tubulin (Sigma).
Preparations were placed into 2.5% 1,4-diazobicyclo-[2,2,2]-octane (DABCO) (Sigma)
solution in glycerol, examined under an Opton-3 photomicroscope (FRG) and photographed on aRF-3 film (Tasma, Russia). In parallel, the same cells were photographed in
the phase contrast on a Mikrat-300 film (Tasma).
RESULTS
Immunoblotting. Figure 1 shows the results of Western immunoblotting in which
preparations of total protein, proteins of cytoskeletal fractions of HeLa cells and a mixture of
a- and p-tubulin were used as samples. In the cytoskeletal fraction sample, the sera we had
prepared revealed two closely positioned bands in the region of 50 kDa. In the total
protein sample, these bands were very faint (hardly revealed in photographs). No bands in
this region were revealed in the tubulin sample. Preimmune sera from both rabbits did nor
react with the cytoskeletal fraction preparation.
Influence of different fixation modes on revealing of γ-tubulin in cells. Preliminary
experiments showed that the centrosome in HeLa cells and primary culture of rat
POLYCLONAL ANTIBODIES AGAINST HUMAN γ-TUBULIN
507
FIGURE 2. Immunofluorescent microphotographs of interphase HeLa cells (a, b) and
primary rat fibroblasts (c, d). Scale, 10 µm. a and c. Staining of microtubules with
antibodies to α-tubulin; b and d, staining of centrioles with R4 antibodies to γ-tubulin.
fibroblasts was most selectively and intensely stained by all of the three antibodies. The
antibodies donated by R. Kuriyama were found to stain intensely the nuclei, apparently
owing to a cross-reaction with DNA (data not shown).
Irrespective of the type of fixation agent (methanol, 4% formaldehyde solution, 1%
glutaraldehyde solution), the antibodies R2 and R4 revealed in the cytoplasm a brightly
glowing dot which was double in most cases (Fig. 2b, d; Figs. 3,4b and 56). In subsequent
experiments we used fixation with glutaraldehyde because under this fixation the cell
structure was best preserved and microtubules were not fragmented.
508
Yu. A. KOMAROVA et al.
FIGURE 3. Immunofluorescent microphotograph of interphase cells of Chinese hamster (clone
237). Scale, 10 µm. Staining of centrioles with R4 antibodies against γ-tubulin.
FIGURE 4. Immunofluorescent microphotographs of interphase cells of rat fibroblasts (a, b).
Scale, 10 µm. a. Staining of microtubules and primary cilium with antibodies to
acetylated tubulin; b, staining of centrioles with R4 antibodies to γ-tubulin.
Variability of staining for γ-tubulin in cells from different tissues and species. All
antibodies used in our work were found to stain a single or double dot in the perinuclear
region of the interphase cell. The intensity of staining varied in different culture types. The
strongest staining was observed in cultures of human and monkey cells (HeLa cells (Fig.
2a, b), primary human fibroblasts, Vero) as well as in cell cultures of rodents: rat
(primary fibroblasts, primary culture of cerebellar and glial cells) (Fig. 2c, d), mouse
(L-fibroblasts) and Chinese hamster (clone 237) (Fig. 3). Cultured epithelial cells from
POLYCLONAL ANTIBODIES AGAINST HUMAN γ-TUBULIN
509
FIGURE 5. Immunofluorescent microphotographs of interphase HeLa cells (a, b). Scale,
10 µm. a, Staining of microtubules with antibodies to α-tubulin; b, staining of centrioles
with procentrioles by R4 antibodies to γ-tubulin.
FIGURE 6. Microphotographs of mitotic SKEE cells. Scale, 10 µm. a-c, normal mitosis;
d-f, spontaneous tripolar mitosis, a, d, Immunofluorescence, staining with antibodies to
a-tubulin; b, e, immunofluorescence, staining with R4 antibodies to γ-tubulin; c, f,
phase-contrast microphotographs of the same cells.
swine kidney embryonal extract were stained less intensively (Fig. 6b). Very faint staining
was observed with PtK1 cells of kangaroo rat. No staining by antibodies to γ-tubulin was
revealed with chicken fibroblasts and newt pneumocytes.
510
Yu. A. KOMAROVA et a/.
The site of γ-tubulin localization in the interphase cell coincides with the cell centre
position. As our antibodies against γ-tubulin stained a single or double dot in the
perinuclear region of interphase cell, it may be suggested that the antigen is localized in the
centrosome region. To check this suggestion, we carried out double staining of cells for αand γ-tubulin. The cells revealed a reticulum of microtubules converging to a single centre
localized near the nucleus. Localization of the convergence centre of microtubules
coincided with the position of the dot of staining for γ-tubulin. Consequently, the antibodies
raised by us against γ-tubulin appear to stain the centriole.
As noted above, staining with antibodies we prepared often revealed a double dot or two
very close dots (Fig. 2b, d; Figs. 3, 4b, 5b). The estimates presented in the table below show
that centrioles were stained as two dots in 86-97% of the interphase cells. Of interest is the
fact that in this case both dots were stained by antibodies with the same intensity (Fig. 2b, d;
Fig. 5b), i.e. γ-tubulin was equally presented in both maternal and filial centrioles. At the same
time, small percentage of cells revealed two double dots stained with different intensity
(Fig. 5b.) The less intensely stained dots, which are very close to either of the two bright dots,
seem to be procentrioles.
According to electron microscopy data [18], centrioles are situated in the centrosome, as a
rule, on the neigbouring ultrathin slices and very rarely one under another. Consequently, at
the light-optical level the centrioles should be in the same optical section. Indeed, in most cells
the two dots revealed by antibodies to γ-tubulin proved to be in the same plane or
neighbouring planes and were also spotted visually. Thus, we could estimate the distance
between centrioles proceeding from the diameter of luminous dots and the clearance
between them. We calculated the distance between two centrioles in the six out of ten
cultures (Table) stained with antibodies to γ-tubulin.
In the interphase cells of HeLa, rat cerebellar and L lines the intercentriole distance in
their pairs was, as a rule, below 1 µm, whereas the markedly spaced centrioles were found
in 3-4% of the cells. In the cells of Vero line and Chinese hamster the proportion of such
centrioles increased to 9-14% and, finally, in the SKEE cells and rat fibroblasts the largely
spaced centrioles were very frequent. In the latter case the distance between centrioles
amounted to 5 microns and more (Table).
Table. Percentage of Interphase Cells Stained by Antibodies to γ-Tubulin as a Function of the
Distance between the Maternal and Filial Centrioles.
Cell line
SKEE
HeLa
Rat cerebellar cells
Rat fibroblasts
L-cells
Chinese hamster cells
Vero
Percentage of stained interphase cells (%) at the following distances
between centrioles, µm
0
<1
1
1-2
2-5
>5
6
11
10
13
14
22
85
81
42
87
84
73
24
3
5
10
6
9
7
16
1
2
4
1
5
4
24
1
17
2
5
2
8
1
14
—
4
3
3
11
POLYCLONAL ANTIBODIES AGAINST HUMAN γ-TUBULIN
511
We carried out double staining of interphase HeLa cells with monoclonal antibodies to
acetylated tubulin and α-tubulin, as well as acetylated tubulin and γ-tubulin. The staining
of HeLa and SKEE cells and rat fibroblasts in the interphase with monoclonal antibodies to
acetylated tubulin showed that the antibodies bind to the centriole and a small portion of
microtubules. Moreover, the antibodies to acetylated tubulin stained brightly the primary
cilium axoneme in cultured cells of SKEE line and rat fibroblasts (Fig. 4a). The antibodies to
γ-tubulin did not reveal the axoneme of primary cilium in cultured cells of the above lines (Fig.
4b). Only basal parts of these cilia proved to be stained - γ-tubulin seems to be localized
exclusively on the centriole.
Localization of γ-Tubulin in Mitotic Cells. The results of double staining of mitotic cells of
HeLa and SKEE lines with antibodies to α- and γ-tubulin have shown that the site of its
localization coincides with the position of the poles of mitotic spindle in the dividing cell
(Fig. 6a-c). Thus, staining for γ-tubulin revealed in mitotic cells two bright spots
corresponding to two mitotic poles. Each pole was stained as a single bright spot whose size
was 1.5-2 times that of dots stained in the interphase. Furthermore, the spindle
microtubules near the mitotic pole were also observed to be slightly stained (Fig. 6b). The
residual body in the HeLa and SKEE cells was not stained by antibodies to γ-tubulin.
In samples of HeLa and SKEE cells we revealed spontaneous multipolar mitotic figures
characterized by the presence of a γ-shaped metaphase platelet. In tripolar mitoses,
microtubules diverged from all three poles (Fig. 6b, e) but only two poles were intensely
stained for γ-tubulin (Fig. 6d-f).
DISCUSSION
The rabbit polyclonal antibodies against the C-terminal fragment of human γ-tubulin,
which were obtained by us, stained cultured cells of different tissues (epithelium, nerve
tissue, fibroblasts), various animal species (human, monkey, pig, rat, kangaroo rat, mouse,
hamster, chicken, triton). These antibodies were found to bind specifically to centrioles of
both interphase and mitotic cells of mammals but not birds and amphibians. These results are
in agreement with the literature data [3] indicating that the amino acid sequences of γtubulin in different mammalian species differ only by single replacements. On the other
hand, the amino acid sequences of γ-tubulin in other animal species (fruit fly, yeast) differ
more substantially from those of γ-tubulin in mammals [2,4,6,9,19].
Contrary to Joshi and other authors who used for immunization the N-terminal
17-member peptide of γ-tubulin [8, 20], we prepared antibodies to the C-terminal moiety of
this molecule. Nonetheless, the staining pattern we obtained with R2 and R4 antibodies fully
corresponds to the localization of γ-tubulin reported earlier [9,20].
The antibodies against two peptides contained in the C-terminal segment of γ-tubulin
used in our study were prepared earlier and found to stain not only the centrosome but also
the residual body in mitosis [21]. Since we observed no staining in any of the cell cultures
studied, the described antibodies against the above peptides [21] may be suggested to cross
react with an unknown component of the residual body.
Recently, γ-tubulin was suggested to form around centrioles complexes acting as primers
for microtubules that may be triggered or switched off in the cell cycle [22]. Our
observations are in agreement with this suggestion. In the interphase cell, one centriole is
512
Yu. A. KOMAROVA et al.
known to be maternal with most microtubules starting from it, whereas the second filial
centrosome is virtually devoid of microtubules [23, 24]. The overwhelming majority of
interphase cells revealed staining of two centrioles. The number of such cells was too large
for all of them to be in the G2-period when two pairs of centrioles can diverge after
replication [15].
It is noteworthy that we did not observe any differences in the staining pattern of two
diverged centrioles in the interphase cells. Thus, according to the model proposed by
Oakley, it should be assumed that γ-tubulin is stored on the filial centriole long before it
becomes the major polymerization centre of microtubules, and subsequently its quantity
does not change dramatically. Upon transition from the interphase to mitosis the quantity of
microtubules diverging from the centrosome increases many times [25] with the
respective increase in the number of primers for their polymerization [26]. However, we
failed to observe any significant enhancement of fluorescence in mitotic cells - the staining of
mitotic poles was less intensive than that of centrioles in the interphase; however, the
diameter of the luminous spot was somewhat larger in mitosis. Therefore, it may be
suggested that transition from the interphase to mitosis induces activation of γ-tubulincontaining complexes [22] and their dispersion over a larger volume (corresponding to a
mitotic halo).
While studying the poles of tripolar mitosis we found that two centriole-carrying poles
are stained with antibodies to γ-tubulin with the same intensity as the poles of normal
mitosis, whereas the third pole is stained very faintly. Taking into account that in tripolar
mitoses the typical distribution of centrioles over poles is 2:2:0 [27,28], one may suggest
that the unstained pole is devoid of centrioles and have to accept that high concentration of γtubulin is not obligatory for the centriole-devoid centres of microtubule organization.
Thus, our studies have shown that γ-tubulin is bound to centrioles irrespective of their
functional activity, i.e. independently of whether or not they are centres of microtubule
organization.
This work was supported by the Russian Foundation for Basic Research (grants
96-04-50935 and 95-04-12703) and the International Science Foundation (grant MRJ300).
REFERENCES
1. C. F. Weil, C. E. Oakley, and B. R. Oakley, Mol. Biol. Cell 6:2963-2968 (1986)
2. C. E. Oakley and B. R. Oakley, Nature 338:662-664 (1989).
3. B. R. Oakley, y-Tubulin. Microtubules, J. S. Hyams, ed. (New York: Wiley-Liss, 1994).
4. Y. Zheng, C. E. Oakley, and B. R. Oakley, J. Cell Biol. 115:382a (1991).
5. B. R. Oakley, C. E. Oakley, Y. Yoon, and M. K. Jung, Cell 61:1289-1301 (1990).
6. Y. Zheng, M. K. Jung, and B. R. Oakley, Cell 65:817-823 (1991).
7. J. M. Raff, D. R. Kellogg, and B. M. Alberts, J. Cell Biol. 121:823-835 (1993).
8. H. C. Joshi, M. J. Palacios, L. McNamara, and D. W. Cleveland, Nature 356:80-83 (1992).
9. T. Steams, L. Evans, and M. Kirschner, Cell 65:825-S36 (1991).
10. V. Muresan, H. C. Joshi, and J. C. Besharse, J. Cell Sci. 104:1229-1237 (1993).
11. H. C. Joshi, Curr. Opin. Cell Biol. 6:55-62 (1994).
12. T. Horio, S. Uzawa, M. K. Jung, B. R. Oakley, K. Tanaka, and M. Yanagida, J. Cell Sci.
99:693-700(1991).
13. P. W. Baas andH. C. Joshi,. J. Cell Biol. 119:171-178 (1992).
14. T. Steams and M. Kirschner, Cell 76:623-637 (1994).
POLYCLONAL ANTIBODIES AGAINST HUMAN γ-TUBULIN
513
15.1. A. Vorobjev and E. S. Nadezhdina, Int. Rev. Cytology, 106:227-293 (1987). 16.1.
B. Alieva and I. A. Vorobjev, CellBiol. Intern. 19:103-112 (1995). 17. U. K. Laemmli,
Nature 227:680-685 (1970).
18.1. B. Alieva and I. A. Vorobjev, Nauch. Dokl. Vysshei Shkoly. Biol. Nauki 6:45-50 (1989) (in
Russian).
19. R. G. Burns, J. CellSci. 108:2123-2130 (1995).
20. T. P. Dibbayawan, J. D. J. Harper, J. E. Elliott, B. Gunning, and J. Marc, CellBiol. Intern.
19:559-567 (1995).
21. M. Julian, Y. Tollon, I. Lajoie-Mazenc, A. Moisand, Mazarguil, A. Puget, and M. Wright, J. Cell Sci.
105:145-156 (1993).
22. B. R. Oakley, Nature 318:555-556 (1995).
23.1. B. Alieva, I. A. Vorobjev, and Yu. S. Chentsov, Dokl. Akad. NaukSSSR 305:1232-1234 (1989)
(in Russian). 24.1. B. Alieva, E. A. Vaisberg, E. S. Nadezhdina, and I. A. Vorobjev, The
Centrosome, V. C.
Kalnins, ed. (New York: Acad. Press, 1992): 103-129.
25. D. R. Kellogg, M. Mortiz, and В. М. Alberts, Anna. Rev. Biochem. 63:639-674 (1994).
26. R. Kuryama and G. G. Borisy, J. CellBiol. 91:822-826 (1981).
27. G. Keryer, H. Ris, and G. G. Borisy, J. CellBiol. 98:2222-2229 (1984). 28.1.
B. Alieva and I. A. Vorobjev, Chromosoma 100:532-542 (1991).