Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://cellmotility.genebee.msu.ru/html/articles/alieva97.pdf Äàòà èçìåíåíèÿ: Wed May 8 16:52:00 2002 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:00:25 2012 Êîäèðîâêà:

Membr. Cell Biol.. 1997, Vol. 11 (1), pp. 17-29 Reprints available directly from the publisher Photocopying permitted by license only

© 1997 OPA (Overseas Publishers Association) Amsterdam B. V. Published in The Netherlands by Hanvood Academic Publishers Printed in Malaysia

Stereoscopic Analysis of Microtubule Pattern Around the Centrosome in Interphase PK Cells after Treatment with Taxol and Nocodazole
I. B. Alieva and I. A. Vorobjev
Belosersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University, 119899 Moscow, Russia

In the interphase PK cells, more than 85% of microtubules radiating from the centrosome were not longer than 1.5 µm. A half of microtubules had their proximal ends free. After nocodazole treatment (20 µ), the number of microtubules attached to the centrosome decreased by 20% after 10 min of treatment, remained the same after 20 min of treatment, and increased after 60 min of nocodazole treatment slightly above the control level. After 5 and 60 min of treatment, the number of attached microtubules with the length over 0.7 µm increased twice as compared to the control level. During the first 20 min of nocodazole treatment, the immunofluorescent staining of cells with antibodies to y-tubulin was the same as in the control cells. The number of free microtubules decreased fourfold during the first 5 min, then it decreased slowly (for 20 min) and remained at the same level after 60 min. After 10 min of taxol (12 µ) treatment, the number of attached and free microtubules increased more than two times, whereas the number of attached microtubules with the length over 0.7 µm increased more than tenfold. After 15 min of treatment, the number of attached microtubules was slightly higher, and the number of free microtubules was half of the control level. After 20-60 min of treatment, the number of microtubules of all types decreased. Thus, upon the nocodazole treatment, the microtubules attached to the centrosome were more resistant to depolymerization: however, these microtubules were the most reactive to taxol treatment. The data obtained suggest that (a) in PK cells, the centrosome-attached microtubules occupy not all of the existing templates; (b) during prolonged treatment with inhibitors, the centrosome performed the compensatory reaction - the inhibition of microtubule assembly results in the decrease in the number of active templates on the centrosome; the inhibition of microtubule depolymerization results in the inactivation of hitherto active reserve templates. The microtubules formed on the centrosome within the first minutes disengage from it and then leave the chromosomal region.
(Received 24 July, 1996)

17


18

I. . ALIEVA AND I. A. VOROBJEV

Formation ot a network of microtubules in cells is related to the microtubule organizing centre (MTOC). In multicellular organisms the purpose of MTOC is served by the centrosome [1, 2]. Microtubules are dynamic structures actively involved in both mitosis organization and vital activities of inter-phase cells. As a rule, the lifetime of individual microtubules does not exceed a few minutes [3]. The total number of microtubules simultaneously present in the interphase cell amounts to several hundreds or even thousands [4-6]. On the other hand, the quantity of microtubules radiating from the centrosome is known not to be very large [7-9], and these microtubules are possibly endowed with special properties. Specifically, it has been shown that upon treatment with colcemide the microtubules radially diverging from the centrosome are dismantled considerably later than those in the bulk of cytoplasm [10]. A substantial role in the nucleation of microtubules seems to be played by the recently discovered protein - -tubulin [11] which is a component of pericentriolar material [12] required for the assembly of microtubules in MTOC in vivo [13-15]. In vitro experiments revealed that the conditions of microtubule polymerization on isolated centrosomes differed from those of spontaneous tubulin polymerization [16]. This made it possible to conclude that the centrosome contains special templates facilitating polymerization of microtubules [16]. At present, the molecules of y-tubulin are supposed to perform the role of such templates (or their functional part) [15, 17, 18]. With the growing tubulin concentration in solution the number of microtubules radiating from isolated centrosomes increases to reach a maximum at 20 µM tubulin. This maximum is several times the number of microtubules radiating from the centrosome in intact cells [16]. Thus, if the centrosome does contain certain templates, they seem to be stored in excess - under normal conditions only a small portion of such templates are functional, the rest being in reserve. Therefore, one may suggest that treatment of cells with compounds capable of stabilizing or depolymerizing microtubules would modify the level of involved templates. The aim of the present study was to analyze the pattern and length of microtubules near the centrosome upon application of taxol known to stabilize microtubules and nocodazole stimulating depolymerization of microtubules. EXPERIMENTAL Cell cultures. Cell cultures of pig kidney embryonic epithelium (PK cells) were grown under 5% CO2 at 37°C on medium 199 supplemented with 10% bovine serum and antibiotics (streptomycin with penicillin or gentamycin). Antibodies. Polyclonal antibodies to the C-terminus of human -tubulin (amino acid residues from 318 to 451) were described previously [19]. Mono-


MICROTUBULE PATTERN AROUND THE CENTROSOME

19

clonal antibodies to -tubulin (DM-1A) and PITC-conjugated secondary antibodies were purchased from Sigma (USA). The Texas-Red-conjugated antibodies were kindly donated by Prof. B. Breton (Rennes, France). Immunofluorescence studies. Prior to their fixation, the cells intended for immunofluorescent analysis were subjected to lysis in a mixture containing 1 % Triton X100 under conditions stabilizing microtubules. Glass cover slips with cells were extracted from a Petri dish, washed with saline phosphate (physiological) buffer (pH 7.2) at 37°C and lysed for 15 min in a solution containing 50 mM imidazole (pH 6.8), 5 mM MgCL, 1 mM EGTA, 0.1 m EDTA, 35% glycerol and 1% Triton X-100 (Sigma). Then the cells were rinsed with the same solution but free of glycerol and Triton X-100 and fixed with 1% glutaric aldehyde (Merck, FRG) in phosphate buffer for 30 min at room temperature with subsequent triple 10-min treatment with NaBH4 (2 mg/ml). To preclude the background luminescence after fixation, 1% bovine serum albumin (BSA) solution in phosphate buffer was added. Then the cells were washed in several buffer replacements and subjected to successive indirect immunochemical staining with antibodies to -tubulin and to -tubulin. Samples were placed into 2.5% 1,4-diazobicyclo[2,2,2]octane (DABCO) (Sigma) solution in glycerol, visualized under an Opton-3 photomicroscope I FRG) and photographed on an RF-3 film (Tasma, Russia). In parallel, the same cells were photographed in the phase contrast using a Mikrat-300 film (Tasma, Russia). Electron microscopy. For thorough studies of the cell centre structure, the cells were lysed before fixation in a Triton X-100-containing mixture under conditions of microtubule stabilization [20]. Preliminary experiments showed that the number and pattern of microtubules near the centrosome underwent no changes in the lysed cells compared to the non-lysed cells. Further preparation of cell samples for electron microscopy was carried out as described earlier [20]. Analysis of microtubule distribution around the centrosome. Spatial analysis of the centrosome structure and the distribution pattern of microtubules inside and around it was performed on stereomicrophotographs of sections 0.2-0.25 µm thick. Stereopairs were obtained by taking shots of sections under an angle of 10° and magnification x 10000 or 12000 using an H-700 electron microscope (Hitachi, Japan) operating under an accelerating voltage of 150-170 kV and equipped with a lanthanum hexaborate cathode and a lateral goniometer. The distribution of microtubules over the cell centre region was studied using stereopairs of serial sections. The analyzed series included all the centriole-containing sections well as a section lying above and one section below the centriole. The area under study was about 33 µm with the width of 0.8 to 1.5 µm (depending on the disposition of centrioles). Division of


20

I. . ALIEVA AND I. A. VOROBJEV

Figure 1. Immunofluorescent microphotograph of interphase PK cells before (a) and after treatment with nocodazole for 60 min (b). Staining of microtubules with antibodies to a-tubulin. Scale, 10 µm. microtubules radiating from the centrosome into classes was performed according to the earlier protocol [20]: class 1, short attached microtubules; class 2, long attached microtubules; class 3, short free microtubules; class 4, long free microtubules. The microtubules are regarded as short when their length does not exceed 0.7 µm; the long microtubules are those with a larger length as well as all the microtubules stretching outside the area under study. The microtubules were regarded as free when their proximal end was at a distance above 100 nm from the nearest electron-dense centrosome structure. Besides all above-mentioned microtubules, near the centrosome there are microtubules which are not directed towards it; their number was not counted.


MICROTUBULE PATTERN AROUND THE CENTROSOME RESULTS

21

Immunofluorescence studies. Immunofluorescent analysis included double staining of cells for - and -tubulin. The PK cells revealed a network of microtubules which sometimes converged in a single centre situated near the nucleus (Fig. la). The position of the centre of convergence of cytoplasmic microtubules coincided with the spot stained for -tubulin or two close points corresponding to the centrosome position. After 20 min of treatment with nocodazole, the cytoplasm was found to contain a considerably smaller number of microtubules than in the norm, the centrosome staining with antibodies to -tubulin did not differ visually from that of control. After 60 min of nocodazole treatment, the cytoplasm contained nearly no microtubules, except a small star formed of microtubules diverging radially from the centrosome (Fig. 1b). At this moment the centrosome is stained more strongly with antibodies to -tubulin than the control. After 10 min of treatment with taxol, the staining with antibodies to -tubulin reveals in the cytoplasm a dense network of microtubules. At this moment the staining of centrosome with antibodies to -tubulin proved to be the most intense. Electron microscopy. In preliminary experiments we compared a number of microtubules referred to different classes around the centrosome in the cells fixed using a standard technique and the cells subjected to lysis prior to fixation. Analysis was performed on sections about 0.1 µm thick because of a higher optical density of the cytoplasm in the cells subjected to standard fixation. In both cases the number of microtubules around the centrosome proved to be virtually the same in all of the four classes (data not shown). Therefore, subsequently the distribution of microtubules around the centrosomes of PK cells treated with taxol and nocodazole was compared with the results obtained earlier for PK cells lysed prior to their fixation [18] and with the data of Table 1 making part of this description. Control. In all, 31.2 ± 3.6 microtubules radiate from the cell centre (centrosome) in the PK cells. Most microtubules originate from the active Table 1. Microtubule pattern around the centrosome in cultured PK cells (from [19]).
Structure from which microtubules radiate Number of radiating microtubules, x±S Total 1 Whole centrosome Active centriole Inactive centriole 31.2±3.6 26.8 ±5.4 4.4 ±0.7 14.6±2.0 14.6±2.0 14.6±2.0 Including microtubules of different classes 2 14.6±2.0 14.6±2.0 14.6±2.0 3 14.6±2.0 14.6±2.0 14.6±2.0 4 14.6±2.0 14.6±2.0 14.6±2.0


22

I. . ALIEVA AND I. A. VOROBJEV

Figure 2. Histogram of microtubule distribution around the centrosome alter treatment with nocodazole. Hatched columns, microtubules with free proximal ends radiating from the centrosome; open columns, microtubules with proximal ends fixed on the centrosome. The abscissa, duration of nocodazole action; the ordinate, average number of microtubules per centrosome. centriole. Some microtubules are attached to electron-dense structures: 7.8 ± 0.2 microtubules, to the heads of pericentriolar satellites; 6.4 ± 0.8 and 3.0 ± 0.5 microtubules, to the walls of cylinders of active and inactive centrioles, respectively. Slightly less than a half of all microtubules have free proximal ends directed to the wall of one of the two centrioles or to the heads of centriolar satellites (Table 1). Outside the limits of the area under study, nearly no microtubules (0.4 ± 0.2 per cell, on average) were found to radiate from the centrosome in cultured PK cells. Microtubule pattern around the centrosome in nocodazole-treated cells. In the PK cells treated with nocodazole for 5 min, 19.4 ± 2.9 microtubules radiated from the centrosome. The distribution of microtubules over classes is shown in Table 2 and Fig. 2. Most microtubules are attached to the wall of a centriole or to the heads of centriolar satellites, whereas a much smaller portion (1/6) of microtubules have free proximal ends. In PK cells treated with nocodazole for 10 min, the microtubule distribution pattern varies insignificantly. The distribution of microtubules over classes is shown in Table 2 and Fig. 2. A quarter of microtubules have a free proximal end, all others are attached to MTOC (Table 2). Only 1.5 ± 0.4 microtubules radiate from the centrosome outside the studied area. In cells treated with nocodazole for 20 min, the total number of microtubules and their distribution over classes also change insignificantly (Table 2, Fig. 2). One third of microtubules have free proximal ends. Virtually no microtubules (0.4 ± 0.2) go outside the studied area. In PK cells treated with nocodazole for 60 min, the number of microtubules increases compared to the results of 20-min treatment (Table 2, Fig. 1).


MICROTUBULE PATTERN AROUND THE CENTROSOME

23

The overwhelming majority of microtubules revealed in these cells are fixed in MTOC, and only 10% of microtubules have free proximal ends. Only 1.9 ± 0.5 microtubules radiate from the centrosome beyond the limits of the area studied. Changes in the number of microtubules radiating from the centrosome after different periods of nocodazole application are summarized in Fig. 2. Under the action of nocodazole on PK cells the number of short and long microtubules unattached to the centrosome decreases over 4 times after the first 5 min of treatment and is further maintained at this level after 10, 20 and 60 min of treatment (Tables 1, 2). During the first 20 min of treatment the number of attached microtubules decreases insignificantly and after 60 min goes on the increase above the normal level. Interestingly, treatment with nocodazole raises considerably the number of microtubules radiating from the inactive centriole, and after 60 min of treatment the difference between the two centrioles is smoothed.
Table 2. Microtubule pattern around the centrosome in nocodazole-treated PK cells. Structure from which microtubules radiate Number of radiating microtubules, x±S Total 1 5-min treatment Whole centrosome Active centriole Inactive centriole 19.4 ±2.9 15.9 ±2.3 3.5 ±0.9 10.8 ±1.6 8.9 ±1.6 1.9 ±0.4 5.2 ±1.4 3.8 ±0.8 0.1 ± 0.1 0.7 ±0.3 0.6 ±0.2 1.4 ±0.9 2.7±0.7 2.6±0.6 0.1±0.1 Including microtubules of different classes 2 3 4

10-min treatment Whole centrosome Active centriole Inactive centriole 17.6 ±3.5 14.2 ±2.9 3.5 ±0.9 9.3 ±1.7 7.2 ±1.5 2.1 ±0.6 3.8 ±0.7 3.3 ±0.6 0.6 ±0.2 1.5 ±0.6 1.1 ±0.4 0.4 ±0.3 3.1 ±0.9 2.7 ±0.7 1.0 ±0.4

20-min treatment Whole centrosome Active centriole Inactive centriole 15.4±2.1 13.0 ±2.0 2.4 ±0.8 10.3 ± 1.7 8.6 ±1.8 1.8 ±0.7 2.9 ±0.7 2.4 ±0.6 0.4 ±0.2 0.7 ±0.2 0.4 ±0.2 0.2 ±0.2 1.6 ±0.4 1.6 ±0.4 0.0 ±0.0

60-min treatment Whole centrosome Active centriole Inactive centriole 25.0 ±2.4 18.6 ±2.3 6.4±1.3 16.7 ±1.7 11.8 ±1.5 4.9±1.0 5.8 ±0.9 4.7 ±0,8 1.1 ±0.4 0.9 ±0.4 0.6 ±0.3 0.3±0.2 1.6 ±0.5 1.5 ±0.5 0.1 ±0.1


MICROTUBULE PATTERN AROUND THE CENTROSOME

23

The overwhelming majority of microtubules revealed in these cells are fixed in MTOC, and only 10% of microtubules have free proximal ends. Only 1.9 ± 0.5 microtubules radiate from the centrosome beyond the limits of the area studied. Changes in the number of microtubules radiating from the centrosome after different periods of nocodazole application are summarized in Fig. 2. Under the action of nocodazole on PK cells the number of short and long microtubules unattached to the centrosome decreases over 4 times after the first 5 min of treatment and is further maintained at this level after 10, 20 and 60 min of treatment (Tables 1, 2). During the first 20 min of treatment the number of attached microtubules decreases insignificantly and after 60 min goes on the increase above the normal level. Interestingly, treatment with nocodazole raises considerably the number of microtubules radiating from the inactive centriole, and after 60 min of treatment the difference between the two centrioles is smoothed.
Table 2. Microtubule pattern around the centrosome in nocodazole-treated PK cells. Structure from which microtubules radiate Number of radiating microtubules, x±S Total 1 5-min treatment Whole centrosome Active centriole Inactive centriole 19.4 ±2.9 15.9 ±2.3 3.5 ±0.9 10.8 ±1.6 8.9 ±1.6 1.9 ±0.4 5.2 ±1.4 3.8 ±0.8 0.1 ± 0.1 0.7 ±0.3 0.6 ±0.2 1.4 ±0.9 2.7±0.7 2.6±0.6 0.1±0.1 Including microtubules of different classes 2 3 4

10-min treatment Whole centrosome Active centriole Inactive centriole 17.6 ±3.5 14.2 ±2.9 3.5 ±0.9 9.3 ±1.7 7.2 ±1.5 2.1 ±0.6 3.8 ±0.7 3.3 ±0.6 0.6 ±0.2 1.5 ±0.6 1.1 ±0.4 0.4 ±0.3 3.1 ±0.9 2.7 ±0.7 1.0 ±0.4

20-min treatment Whole centrosome Active centriole Inactive centriole 15.4±2.1 13.0 ±2.0 2.4 ±0.8 10.3 ± 1.7 8.6 ±1.8 1.8 ±0.7 2.9 ±0.7 2.4 ±0.6 0.4 ±0.2 0.7 ±0.2 0.4 ±0.2 0.2 ±0.2 1.6 ±0.4 1.6 ±0.4 0.0 ±0.0

60-min treatment Whole centrosome Active centriole Inactive centriole 25.0 ±2.4 18.6 ±2.3 6.4±1.3 16.7 ±1.7 11.8 ±1.5 4.9±1.0 5.8 ±0.9 4.7 ±0,8 1.1 ±0.4 0.9 ±0.4 0.6 ±0.3 0.3±0.2 1.6 ±0.5 1.5 ±0.5 0.1 ±0.1


MICROTUBULE PATTERN AROUND THE CENTROSOME

25

Figure 3. Histogram of microtubule distribution around the centrosome after treatment with taxol. Hatched columns, microtubules with free proximal ends radiating from the centrosome; open columns, microtubules with proximal ends fixed on the centrosome. The abscissa, duration of nocodazole action; the ordinate, average number of microtubules per centrosome.

change (Table 3, Fig. 3). Most of the microtubules are attached to MTOC, whereas less than 1/5 have free proximal ends. As few as 1.1 ± 0.6 microtubules radiate from the centrosome outside the studied area. The dynamics of changes in the number of microtubules radiating from the centrosome after different periods of taxol action are shown in Fig. 3. During the first 10 min of taxol action on PK cells a considerable rise in the number of microtubules is observed in all of the four classes (Tables 1, 3). This increase is provided for by the active centriole, whereas the number of microtubules radiating from the inactive centriole does not change. After 15 min and later the number of microtubules in all classes decreases dramatically (Tables 1, 3). Despite this decrease, the number of attached microtubules is still above normal, whereas the number of free microtubules becomes twice as low as in the norm. The dynamics of these changes is preserved with time; however, if the number of attached microtubules goes down to the norm after 20 min and then remains the same, the number of free microtubules becomes 4 times as small as in the norm after 60 min (Tables 1, 3).


26

I. . ALIEVA AND I. A. VOROBJEV

DISCUSSION

Although a specific character of microtubule growth from the centrosome was repeatedly demonstrated (see [21 ] for discussion), these data are however hardly applicable to living cells. The basic contradiction consists in the fact that in vitro numerous long (several microns long) microtubules grow out around the centrosome [16, 22], whereas in living cells such a situation is very rare. In thorough studies of cell cultures reported earlier, only single microtubules stretched from the centrosome to the cell periphery. Our observations confirmed the two earlier inferences: (a) most microtubules diverging radially from the centrosome are short (about 1 um) and do not reach the cell periphery; (b) about 50% of microtubules radiating from the centrosome have free proximal ends. Thus, formation of the peripheral system of microtubules is not related to any selective anchoring of some microtubules on the centrosome as suggested previously [21]. Our studies have revealed that the microtubules attached to the centrosome and those with the free proximal ends display a different behaviour. When the PK cells are treated with nocodazole, the number of free microtubules decreases dramatically during the first 5 min and then remains invariable during the next 60 min. The number of attached microtubules exhibits a gradual and insignificant decrease during 20 min. When the cells are treated with taxol, the number of all microtubules, except the short free ones, radiating from the centrosome increases drastically during the first 10 min. Thus, under the impact of nocodazole, i.e. under depolymerization conditions, the attached microtubules proved to be more "stable". On the other hand, these microtubules responded more actively to the application of taxol, i.e. they proved to be more "labile" under stabilization conditions. The short free microtubules appear to form an intermediate population - their number decreases dramatically immediately after the application of taxol as well as nocodazole. We presume that they are elongated under the action of taxol (i.e. they pass to the class of long microtubules), whereas nocodazole induces their depolymerization as a result of catastrophic disassembly. On the contrary, the population of short attached microtubules is less affected in the course of both impacts than all the other microtubules. Nowadays, it is generally recognized that the centrosome contains templates initiating the growth of individual microtubules [16]. In interphase cells, the total number of templates on the centrosome is relatively small, not exceeding several tens [23], whereas the cell is the site of simultaneous growth of up to a thousand microtubules [6]. In order to explain the discrepancy between the number of microtubules attached to the centrosome and the total number of microtubules in a cell, a conveyor hypothesis was proposed for the assembly of microtubules [1, 7]. According to this hypothesis, upon reaching a certain length, the microtubules become capable


MICROTUBULE PATTERN AROUND THE CENTROSOME

27

of disengaging from the centrosome and pass to the cytoplasm [1, 7]. As suggested in [ 18], the templates are protein complexes whose core is -tubulin - "gamma-somes". Thus, tubulin is supposed to be the marker of the templates on which growth of microtubules is initiated. We assume that in the PK cells the microtubules are able to detach themselves from the templates (gamma-somes) and migrate rapidly to the cytoplasm after achieving a relatively small length (about 1 µm). This may happen, for instance, owing to the operation of a minus-end motor (dynein, certain kinesines). There is a certain probability that a vacated template becomes the site of a new microtubule growth. Evidently, near the centrosome free microtubules will be more labile than the microtubules attached to it. The share of actuated templates at each moment reflects the probability of microtubule growth initiation. In the course of incubation of cells with taxol as well as with nocodazole the dynamics of microtubule populations around the centrosome does not correspond to that dictated by the "dynamic instability" and the concept of stable templates. The first 60 min of treatment induce opposite changes in the number of microtubules in virtually all the classes. During 5-10 min of nocodazole action the number of attached microtubules decreases first and then goes on the increase. During 10 min of treatment with taxol the number of attached microtubules rises first to decrease subsequently. The result is that in the course of 60-min treatment with nocodazole the number of microtubules attached to the centrosome increases and even becomes 1.5 times as high as their normal level. On the contrary, after 20 min of treatment with taxol the number of microtubules near the centrosome proves to be below the initial level. This suggests that the number of microtubules radiating from the centrosome is regulated by the centrosome itself depending on conditions of tubulin polymerization in the cytoplasm. It is important to note that changes in the number of microtubules correlate with the intensity of centrosome staining with antibodies to -tubulin. Upon nocodazole application the intensity of centrosome staining corresponded to that of control during the first 20 min and became much stronger after 60 min. In contrast, in the case of taxol it corresponded to the staining of control during the first 10 min and then became fainter. We presume that a prolonged action of the inhibitor of microtubule polymerization activates additional templates in the centrosome, and the cell brings about a compensatory reaction - a larger number of microtubules than in the norm radiate from the cell centre against the background of microtubule disappearance in the cytoplasm. Under the impact of taxol the cell displays the opposite behaviour: the initially triggered reserve templates are inactivated, whereas the microtubules formed detach themselves and leave the centrosome. Thus, in cultured PK cells the conditions of polymerization of microtubules on the centrosome differ from those of their polymerization at the cell


28

I. . ALIEVA AND I. A. VOROBJEV

periphery. The microtubules do not grow simultaneously out of all the templates existing on the centrosome, and as a rule, they are detached from the centrosome before reaching a length of 1.5 µm. Apparently, under normal conditions there is a regulatory mechanism ensuring a relatively low level of microtubule growth initiation - by analogy with the cell-based mechanism for stabilization of microtubules through their post-translational modification [24]. Summing up the above considerations, we may conclude that the dynamics (i.e. length and distribution pattern) of microtubules around the centrosome, studied with the help of compounds stabilizing or depolymerizing microtubules, is determined at the initial stages by the instability of microtubules proper and subsequently by the operation of the centrosome as a carrier of templates. The microtubules associated with the centrosome are short and prove to be more labile than the cytoplasmic microtubules. At the same time, as shown earlier [25-28] and confirmed in the present work, the population of the centrosome-associated microtubules is, on the whole, more stable than the cytoplasmic network of microtubules. This work was supported by the Russian Foundation for Basic Research (grants 93-04-06523, 96-04-50935 and 95-04-12703) and the International Science Foundation (grant MRI300).

REFERENCES 1. I. A. Vorobjev and E. S. Nadezhdina, Intern. Review Cytol. 106:227-293 (1987). 2. D. R. Kellog, M. Moritz, and . . Alberts. Annu. Rev. Biochem. 63:639-674 (1994). 3. M. D. Saxton, D. L. Stemple, R. J. Leslie, E. D. Salmon, M. Zavortink, and J. R, Mclntosh, J. CellBiol. 99:2175-2186(1984). 4. B.J.SoltysandG.G.Borisy, J. Cell. Biol. 100:1682-1689(1985). 5. E. Schuize and M. Kirschner, J. Cel. lBiol. 102:1020-1031 (1986). 6. E. Schuize and M. Kirschner, J. Cell. Biol. 104:277-288 (1987). 7. I. A. Vorobjev and Yu. S. Chentsov, Tsitologiya 24:1286-1289 (1982) (in Russian). 8. I. B. Alieva and I. A. Vorobjev, Tsitologiya 33:18-26 (1991) (in Russian). 9. Yu. W. Centonze, F. J. Ahmad, and P. Baas, J. Cell. Biol. 122:349-359 (1993). 10. I. A. Vorobjev and Yu. S. Chentsov, Tsitologiya 27:1101-1105 (1985) (in Russian). 11. . E. Oakley and . R. Oakley, Nature 338:662-664 (1989). 12. . Steams, L. Evans, and M. Kirschner, Cell 65:825-836 (1991). 13. . R. Oakley, . E. Oakley, Y. Yoon, and M. K. Yung, Cell 61:1289-1301 (1990). 14. T. Horio, S. Uzawa, . . Jung, . R. Oakley, K. Tanaka, and M. Yanagida, J. Cell Science 99:693-700 (1991). 15. . . Joshi, M. J. Palacios, L. McNamara, and D. W. Cleveland, Nature 356:80-83(1992). 16. T. Mitchison and M. W. Kirschner, Nature 312:232-237 (1984).