Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://cellmotility.genebee.msu.ru/html/articles/alieva2000.pdf Äàòà èçìåíåíèÿ: Tue May 21 17:49:30 2002 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:06:18 2012 Êîäèðîâêà:

Membr. Cell Bio/., 2000, Vol.14 (1), pp. 57-67 Reprints available directly from the publisher Photocopying permitted by license only

© 2000 OPA (Overseas Publishers Association) N.V. Published by license under the Harwood Academic Publishers imprint, part of The Gordon and Breach Publishing Group Printed in Singapore

Interphase Microtubules in Cultured Cells: Long or Short?
I. B. Alieva and I. A. Vorobjev
Belozersky Institute of Physicochemical Biology, Moscow State University, Moscow, 119899 Russia fax: (095) 939 3181; E-mail: alieva@electron.genebee.msu.su

Presently, the question about the length of microtubules in the interphase cell became actual, since the parameters of dynamic instability of the plus end measured in vivo do not allow one to explain the rapid turnover of the long microtubule system. The problem may be solved if one of the following suppositions is assumed: either microtubules undergo rapid depolymerization from the minus end or they are on the average much shorter than it is usually considered. To check the last hypothesis, we have reconstructed microtubules using stereophotography of electron microscopic sections. Microtubules around the cell center in cultures of epithelial cells (kidney of pig embryo (PK) and bovine trachea (FBT)) and fibroblasts (MEF, primary mouse embryo fibroblasts, and L cells), as well as at the periphery of PK cells were studied. All in all, no less than 200 microtubules were found near the centrosome in each cell culture. From 2.5 to 8% microtubules were beyond the studied volume (4.0x5.5x 1.5 µm). Most of microtubules in all studied cell lines were up to 1 µm and about 1/3 of them were 0.2-0.4 µm long. The mean length of microtubules surrounding the centrosome in different cell lines differed insignificantly and equalled 0.4-0.8 µm. In this case, the microtubules attached to the centrosome were on the average slightly shorter than the free ones. Thus, almost all microtubules around the centrosome are short, and the majority of those attached to it do not reach the cell periphery. A similar reconstruction of a part of the PK cell cytoplasm (10x35 µm) has shown that at the periphery, the mean length of microtubules is about 1.6 µm and most of them are 0.5 to 1.5 µm long. Thus, our data confirm the recent hypothesis of Vorobjev et al. (I. A. Vorobjev, T. M. Svitkina, and G. G. Borisy, J. Cell Sci. 110:2635-2645 (1997)) that most of microtubules in the cells are not connected with the centrosomes.
(Received 11 May, 1998)

INTRODUCTION
Microtubules are dynamic structures involved both in organization of mitosis and vital activity of interphase cell. Total amount of microtubules coexisting 57


58

I. . ALIEVA AND I. A. VOROBJEV

in an interphase cell is unknown. Some authors estimate it as several hundreds or even thousands [ 1 -14]. On the other side, it has been shown that the amount of microtubules directly derived from the centrosome is low [5-7], and they may have special properties. In particular, the microtubules directly radiating from the centrosome are depolymerized in the presence of colcemid significantly later as compared with those located in the rest cytoplasm [8]. It is generally recognized that formation of the microtubule network in the cells is connected with the microtubule organizing centers (MTOC), and the centrosome is the main MTOC in multicellular animals [9-12]. It was believed for a long time that all microtubules are formed on the centrosome and then remain bound to it during the whole cell life and are extended from the centrosome to the cell periphery [13]. Experimental data cast some doubt on such an aspect. The specialized cells, like neurons and polarized epithelial cells have in the cytoplasm numerous microtubules not attached to the centrosome [14-16]. It has been shown that after the centrosome removal, nucleation of microtubules is possible in the cytoplast or cell fragment [17-19]. In addition, the results obtained during the last year allow us to state that in normal conditions, noncentrosomal microtubules can be also formed in the cytoplasm of different animal cell cultures [1,20,21], and they can constitute up to 80-90% of the total amount [1]. We have shown previously that the microtubules diverged from the centrosome in different cells are not enough long to reach the cell periphery [6,22], and even substances capable of partial stabilization of microtubules do not cause their noticeable growth [23]. These data are clearly contradictory to the results obtained by light microscopy revealing extended microtubules radiating outward from the centrosome. In this work, we have attempted to solve this contradiction, reconstructed microtubules around the centrosome in cultures of epithelial cells and fibroblasts, and compared them with cytoplasmic microtubules at the cell periphery.
EXPERIMENTAL

Cell cultures. The pig embryo kidney epithelium cells ( line) were grown at 37°C and 5% CO2 in the medium 199 supplemented with 10% bovine serum and antibiotics streptomycin with penicillin or gentamycin. L cells (culture of mouse fibroblasts) were grown at 37 °C and 5% CO2 in the medium 199 supplemented with 10% fetal calf serum and gentamycin. Epithelial cells of bovine trachea (FBT) were grown at 37°C and 5% CO2 in Eagle's medium with 10% bovine serum and antibiotics (streptomycin with penicillin or gentamycin. Primary culture of mouse fibroblasts (MEF) were grown at 37°C and 5% CO2 in DMEM/F-12 HAM (Sigma) supplemented with 10% fetal calf serum and gentamycin.


INTERPHASE MICTOTUBULES IN CULTURED CELLS

59

Figure 1. Reconstruction of microtubules around the centrosome in an L (a) and (b) cells. Arrows show microtubules located outside the area under study (positions of their distal ends cannot be detected).

Electron microscopic investigations. For detailed structural analysis of the cell center, before fixing the cells were lyzed in a mixture containing Triton X-100 in conditions favorable for stabilization of microtubules [24]. We have shown in preliminary experiments that the amount and character of distribution of microtubules in the centrosome region are not changed in the lyzed cells as compared with the nonlyzed ones. Further preparation of the cells for electron microscopy was done as described earlier [24]. Analysis of microtubule distribution around the centrosome. Spatial analysis of the centrosome structure and the microtubule distribution inside and around the centrosome was carried out on stereoscopic photographs of serial 0.2-0.25 µm thick sections. Stereoscopic images were obtained by taking


60

I. . ALIEVA AND I. A. VOROBJEV

pictures at an angle of 10° and magnification x 12,000 under an electron microscope H-700 (Hitachi) at the accelerating voltage of 150-175 kV. Total reconstruction of microtubules based on serial microphotographs is shown in Fig. 1. Similar reconstructions were done for all studied cells (seven cells , six cells FBT, five L cells, and four MEF cells). The microtubule arrangement near the cell center was studied using stereoscopic images of serial sections [23]. The analyzed series included all centriole-containing sections, as well as one section above and one below the centriole. The region studied had the square of about 4.0x5.5 µm and thickness of 0.8-1.5 µm, depending on location of centrioles. The microtubules whose proximal end was closer than 100 nm to the nearest electron dense centrosomal structure were considered as attached ones, whereas those with the proximal end at a distance longer than 100 nm from the same structure were considered as nonattached or free microtubules. Besides the above-mentioned microtubules, there were those located also near the centrosome but not directed towards it. They were not counted. Analysis of the microtubule arrangement in the cytoplasm. Since the cell lamella, unlike the centrosomal region, contains a sufficient amount of fibrillar structures often arranged in clusters, microtubules could be confused with different elements of the cytoskeleton. To avoid this, microtubules were marked with particles of colloid gold. The cells were lyzed in the presence of Triton X-100 in conditions favorable for microtubule stabilization [24], then fixed with 1% glutaraldehyde solution in PBS, pH 7.3 for 10 min. Immunoelectron staining was carried out using the first antibodies against tubulin DM-1A (Sigma) and second antibodies conjugated with colloid gold (Sigma). After staining, the preparation was fixed again with 1% glutaraldehyde in PBS for 20-30 min. Then, a standard procedure for electron microscopic preparations was used. Stereoscopic images were obtained by the method similar to the above described one at magnification x7000. The cell lamella was fully photographed on five sequential serial half-thin sections. Reconstruction and data analysis. The microtubule profiles were transferred from photographs to the transparent film. The ends of extended microtubules on different sections were identified using stereoscopic analysis of negatives. The length was measured by curvimeter. To construct the histogram, the Statistika program (StatSoft, Inc., USA) was used.
RESULTS

As was shown previously [24], most microtubules near the centrosome radiate outward from the centrioles, and only single microtubules pass by. Since the area under study was approximately 4.0 x5.5 µm and, depending on position of centrioles, its thickness was from 0.8 tol.5 µm, maximal length of


INTERPHASE MICTOTUBULES IN CULTURED CELLS

61

Figure 2. Histograms of the microtubule length distribution in two variants (on an example of attached microtubules located around the centrosome in cells, (a) only microtubules with distal ends located within the volume under study were considered; (b) all microtubules were considered, including those with distal ends located beyond the volume under study (are included in the last class of the histogram). The ordinate shows the amount of microtubules, the abscissa -- length of mictotubules, µm.

radial microtubules, which we managed to determine, was a little longer than 2 µm. Some microtubules with their proximal ends close to the centrosome was beyond the area under study, and we could not trace the arrangement of their distal ends (Fig. 1). However, the percentage of such microtubules in the whole selection was low. It was equal to 6.3% in MEF, 7.9 in L, 6.6 in FBT, and only 2.5% in cells. When counted per cell, mean content of such microtubules made up 6.8 (from 3 to 13) in MEF, 5.4 (from 2 to 11) in L, 4.5 (from 0 to 11) in FBT, and 0.8 (from 0 to 4) in cells. To estimate contribution of microtubules beyond the limits of the volume under study, two variants of histograms showing distribution of microtubules by length were constructed. The first considered only microtubules with


62

I. . ALIEVA AND I.A.VOROBJEV

Fig. 4 Fig.3 Figure 3. Histogram of the microtubule distribution by length around the centrosome in MEF cells. Here and in Figs 4-6: (a) attached microtubules; (b) free microtubules. The ordinate shows the amount of microtubules, %; the abscissa - the microtubule length, µm. Figure 4. Histogram of microtubule distribution by length around the centrosome in L cells.

distal ends within the volume under study. The second was constructed with regard to microtubules whose ends were beyond the volume under study, and these microtubules were considered to be the longest (they were included into the last class of histogram). The histograms obtained for cells seem to be very similar (Fig. 2). Owing to this, in further studies microtubules beyond the limits of the volume under study were excluded from following analysis. Mouse embryo fibroblasts (MEF). The histogram of microtubule distribution by length is shown in Fig. 3. Most of microtubules (over 99% both attached and free ones) radiating from the cell center in MEF cells are no longer than 2 µm, 89% attached and 87% free microtubules are up to 1 µm. Microtubules of 0.2-0.4 µm are prevalent: they make up 35% of free and 29% of attached microtubules. In both case, only about 1% are longer than 1.6 µm. The L line fibroblasts (L cells). Histogram of microtubule distribution by length is shown in Fig. 4. It is seen that in these cells, microtubules are somewhat longer, but the length of most of them (98% attached and 99% free)


INTERPHASE MICTOTUBULES IN CULTURED CELLS

63

Fig. 5 Fig. 6 Figure 5. Histogram of microtubule distribution by length around the centrosome in FBT cells. Figure 6. Histogram of microtubule distribution by length around the centrosome in cells.

radiating outward from the cell center also does not exceed 2 µm. Short microtubules are prevalent: those of 0.2-0.4 µm make up 34% of attached ones. There is no single predominating class among free microtubules, and those of 0.2-0.8 µm are prevalent. The bulk of microtubules are up to 1 µm, 88% and 74%, respectively. However, although only 1% of attached microtubules are longer than 1.6 µm, there are more free microtubules of 1.6-2.2 µm (they make up about 8%). Epithelial cells of bovine trachea (FBT). Histogram of microtubule distribution by length is shown in Fig. 5. The results show that the bulk of microtubules (over 99% both attached and free) radiating outward from the cell center in FBT cells are no longer than 2 µm. In a population of attached microtubules, those of 0.2-0.4 µm are prevalent (35%), whereas microtubules of 0.2-0.8 µm predominate among free ones (46%). In this case the bulk of microtubules are up to 1 µm (87 and 76%, respectively). At the same time, 2% of attached microtubules are longer than 1.6 µm, and about 5% of free ones are also longer than 1.6 µm. Epithelial cells of pig embryo kidney (). Fig. 6 shows a histogram of the microtubule distribution by length in the cell center area of cells. The


64

I. . ALIEVA AND I. A.VOROBJEV

results show that there are no microtubules (both attached and free) radiating outward from the centrosome more than for 2 µm. Microtubules up to 0.2 µm are prevalent: they make up 30% of attached microtubules and 20% of free ones. Respectively, 90 and 83% of attached and free microtubules are up to 1 µm. Histograms of microtubule distribution by length obtained for all cell cultures studied have a distinct peak (modal length value) and are most satisfactorily described by -distribution. The mean length of attached and free microtubules in all studied cell cultures (Table 1) varies insignificantly (all differences between mean values shown in table are statistically uncertain). At the same time, on the average, free microtubules are a little longer than the attached ones.
Table 1. Mean length of microtubules surrounding the centrosome in studied cells. Cell line Mean length of microtubules, µm x±Sx Attached MEF L cells FBT 0.57±0.36 0.53 ±0.41 0.58±0.39 0.43±0.33 Free 0.61+0.31 0.77+0.46 0.76±0.40 0.57±0.38

Especially interesting are attached microtubules which are beyond the volume under study and are candidates for the role of radial microtubules visible in the light microscope. Their total amount in cells made up 3 microtubules per 7 analyzed cells. In FBT there are 11 microtubules per 6 cells, in MEF -12 microtubules per 4 cells, and in L cells - 12 microtubules per 5 cells. Thus on the average no more than 2-3 microtubules radiate from the centrioles to the cell periphery. However, the above-described data on the microtubule distribution may be incomplete. A small area used for reconstruction does not allow revealing of potentially long microtubules, although it is evident that they should be sufficiently rare. It cannot be also excluded that the microtubule dynamics in the immediate vicinity of the centrosome for some reason differs from that in other parts of the cell. For comparison, all microtubules in a relatively big region of the cell were studied. Histogram in Fig. 7 showing the length distribution of microtubules in the cell lamella in general is similar to that for microtubules around the centrosome. However, on the average microtubules were longer, 1.65+1.01 µm. More than 90% microtubules are up to 3 µm in length. Those of 1-2 µm make 35%, shorter than 1 µm comprise 50%. About 11% microtubules are of 2-3 µm and only about 4% are longer than 3 µm. Single micro-


INTERPHASE MICTOTUBULES IN CULTURED CELLS

65

Figure 7. Histogram of microtubule distribution by length in the cell lamella. The ordinate shows the amount of microtubules, %; the abscissa the microtubule length, µm.

tubules (less than 1%) are longer than 5 µm. No microtubules longer than 6 µm were found.
DISCUSSION

These observations have confirmed the previous conclusion that the majority of microtubules radiating outward from the centrosome are very short (less than 1 µm) and do not reach the cell periphery. At least for cells this conclusion can be extended: most of microtubules in the cytoplasm are also short (less than 3 µm) and are not connected with the centrosome. The following fact deserves a special discussion. In MEF, FBT, and L cells on the average from 4.5 to 6.8 microtubules are beyond the volume under study, whereas there are fewer such microtubules in cells, i.e., potentially 4-7 microtubules in MEF, FBT, and L cells can be long and reach the cell periphery. However, the light optics in some cases makes possible to see tens of long microtubules radiating from the centrosome region in cultured cells. Thus, it is quite possible that microtubules usually considered as radiating outward from the centrosome are not such in reality. Their proximal ends may be separated by the distance up to 1 µm from centrosomal structures, which cannot be seen in light microscope, since the microtubule density around centrioles is very low. There are data in the literature on separation of extended microtubules from the centrosome in living cells [20]. Probably, minus ends of these microtubules are located near the centrosome, but not attached to it. At first sight, the presence of a large amount of short microtubules in an inerphase cell is contradictory to the dynamic instability concept suggesting that in the stationary condition, the amount of microtubules should become lower and they should become longer [25]. However, it should be taken into consideration that in a living cell, microtubules do not grow spontaneously,


66

I. . ALIEVA AND I. A. VOROBJEV

but rather from the gammasomes, special primers containing gamma-tubulun [26, 27]. Suppose, there is an excess of primers, then the amount of simultaneously existing microtubules should be high and not decrease with time. As is known, the main part of primers for the microtubule growth are located in the centrosome. We believe that these primers provide for arising of numerous microtubules around centrioles, but they cannot provide for their elongation, an due to this, the bulk of microtubules remain short. Probably, formation of long microtubules in the cell is a random and rare event regulated only by the plus-end dynamics. In our previous investigations [6, 24], we have shown qualitatively single microtubules radiating from the centrosome to the periphery of the cell culture cells, whereas most microtubules were short. The results differ from obtained in vitro, according to which numerous long up to 10-20 µm and longer appear around the isolated centrosomes [25, 28, 29]. Thus, although specific growth of microtubules from the centrosome was repeatedly demonstrated in vitro, these data can be hardly applied to the living cells. The main contradiction is the fact that such intensive intracellular growth of microtubules is observed very rarely. We did not find in the literature data of quantitative analysis of the number and distribution pattern of microtubules around the centrosome, but some authors noticed relatively little microtubules radiating from the centrosome [30, 31]. Then, the question arises what is the origin of numerous microtubules observed in the cell cytoplasm? Our observations show that formation of peripheral microtubule system is not due to the selective anchoring of some of the microtubules on the centrosome as was believed previously [28]. On the opposite, there is reason to believe that the bulk of peripheral microtubules are formed independently of the centrosome. Such a view comes against an evident difficulty in explanation of radial distribution of quite long microtubules that is clearly visible in many cells upon light microscopy. Presently, we have no data to answer this question. However, one can suggest that the radial system of microtubules in fact is a system of relatively short ones overlapped like the fire-escape segments. Such a situation seems to be most obvious for cells, where on the average less than one microtubule protrudes beyond the limits of the centrosome nearest neighborhood. The next report will deal with checking this hypothesis.


INTERPHASE MICTOTUBULES IN CULTURED CELLS REFERENCES

67

1. Vorobjev, LA., Svitkina, T.M., and Borisy, G.G., 1997, J. Cell Sci. 110,2635. 2. Soltys, B.J. and Borisy, G.G., 1985, J. Cell EM. 100,1682. 3. Schulze, E. and Kirschner, M., 1986, J. CellBiol. 102,1020. 4. Schulze, E. and Kirschner, M., 1987, J. Cell Biol. 104, 277. 5. Vorobjev, LA. and Chentsov, Yu.S., 1982, Tsitologiya. 24. 1286. 6. Alieva, I.B. and Vorobjev, I.A., 1991, Tsitologiya. 33,18. 7. Yu, W., Centonze, V.E., Ahmad, F.J., and Baas, P., 1993, J. CellBiol. 122, 349. 8. Vorobjev, LA. and Chentsov, Yu.S., 1985, Tsitologiya. 27,1101. 9. Mclntosh, J.D., 1983, Modem CellBiol 2,115. 10. Vorobjev, LA. and Nadezhdina, E.S., 1987, Intern. Review ofCytol. 106,227. 11. Bershadsky, A.D. and Gelfand, V.I., 1981, Proc. Natl. Acad. Sci. USA. 78, 3610. 12. Kellog, D.R., Moritz, M., and Alberts, B.M., 1994, Annu. Rev. Biochem. 63, 639. 13. Alberts, B.M., Bray, D., Lewis, J., Raff, M., Roberts, K., and Watson, J.D., Molecular Biology of the Cell, Garland Publishing, New York, London, 1993. 14. Chalfie, M. and Thomson, J.N., 1979, J. Cell Biol. 82,278. 15. Bray, D. and Bunge, M.B., 1981, J. Neurocytol. 10,589. 16. Mogensen, M.M., Mackie, J.B., Doxey, S.J., Stearns, ., and Tucker, J.B., 1997, CellMotyl. Cytoskel. 36, 276. 17. Karsenti, E., Kobayashi, S., Mitchison, ., and Kirschner, M., 1984, J. Cell Biol. 98,1763. 18. Maniotis, A. and Schliwa, M., 1991, Cell. 67,495. 19. Rodionov, V.I. and Borisy, G.G., 1997, Science. 275, 215. 20.Keating. T.J., Peloquin, J.G., Rodionov, V.I., Momcilovic, D., and Borisy, G.G., 1997, Proc. Natl. Acad. Sci. USA. 94, 5078. 21. Yvon, A.C. and Wadsworth, P., 1977, J. Cell Sci. 110, 2391. 22. Alieva, I.B., Vaisberg, E.A., Nadezhdina, E.S., and Vorobjev, LA., The Centrosome, Acad. Press, N.Y, 1992, pp. 103-129. 23. Alieva, I.B. and Vorobjev, LA., 1997, Membr. and CellBiol. 14, 629. 24. Alieva, I.B., Vorobjev, LA., and Chentsov, Yu.S., 1989, Dokl. Akad. Nauk. 305, 1232. 25. Mitchison, T. and Kirshner, M.W., 1984, Nature. 312, 232. 26. Oakley, B.R., Gamma-Tubulin. Microtubules, Wiley-Liss, New York, 1994. 27. Oakley, B.R., 1995, Nature.378,555. 28. Kirschner, M.W. and Mitchison, ., 1986, Cell. 45, 329. 29. Belmont, L.D., Hyman, A.A., Sawin, K.E., and Mitchison, T.J., 1990, Cell. 62, 579. 30. Gudima, G.O., Vorobjev, LA., and Chentsov, Yu.S., 1988, J. Cell. Sci. 89,225. 31. Bre, M.-H., Kreis, .., and Karsenti, E., 1987, J. CellBiol. 105,1283.