Identification of a 100 kD protein associated with microtubules, intermediate filaments and coated vesicles in cultured cells

Experimental Cell Research 159 (1985) 377-387

Identification of a 100 kD Protein Associated with Microtubules, Intermediate Filaments and Coated Vesicles in Cultured Cells V. I. RODIONOV, E. S. NADEZHDINA, E. V. LEONOVA, E. A. VAISBERG, S. A. KUZNETSOV and V. I. GELFAND Laboratory of Molecular Biology and Bio-organic Chemistry, Moscow State Unioersity, Moscow 119899, USSR

We have obtained several hybridoma clones producing antibodies to microtubule-associated proteins (MAPs) from bovine brain. Interaction of one of these antibodies, named RN 17, with cultured cells was studied by indirect immunofiuorescence and immunoelectron microscopy. RN 17 antibody recognized both high molecular weight (HMW) MAPs, MAP 1 and MAP2, in immunoblotting reaction with brain microtubules. In lysates of cultured cells, it bound to a protein doublet with a molecular weight of 100 kD. By immunofluorescence microscopy we showed that RN 17 antibody stained cytoplasmic fibrils, mitotic spindles and small particles in the cytoplasm of various cultured cells. The cytoplasmic fibrils were identified as both microtubules and intermediate filaments by double fluorescence microscopy and by their response to colcemid and 0.6 M KCI. This identification was confirmed by immunoelectron microscopy which also showed that the particles stained by RN 17 antibody are coated vesicles. Thus, cultured non-neural cells may contain a novel protein that binds to microtubules, intermediate filaments, and coated vesicles. © 1985AcademicPress, Inc.

Microtubules are the ubiquitous cytoskeletal structure composed of tubulin and microtubule-associated proteins (MAPs). Destruction of microtubules by colchicine or vinblastine leads to the inhibition of various intracellular movements, i.e., axoplasmic transport, movement of pigment granules in melanophores, saltatory movement of organelles etc. (see review in [1]). Thus, microtubules are involved in intracellular transport. However, the mechanism of their involvement remains obscure. One approach to this problem is to study microtubule proteins. The major nontubulin components of microtubules, isolated from brain by cycles of polymerization-depolymerization, are high molecular weight (HMW) proteins MAP 1 and MAP 2 [2-4] and a group of r-proteins with a molecular weight (MW) of about 60 kD [5, 6]. Both MAP I and MAP 2 form regular projections on the microtubule surface [7-10]. It is possible that these projections are a key component in a microtubule-dependent intracellular transport. Microtubules have been purified also from a variety of non-neuronal cells. These preparations contain some specific MAPs [I 1-15] as well as MAPs already found in preparations of brain microtubules [16-18]. To study the composition and functions of MAPs, we have obtained several Exp Cell Res 159 (1985)

378

Rodionov et al.

c l o n e s o f h y b r i d o m a cells t h a t p r o d u c e a n t i b o d i e s to b r a i n M A P s . H e r e we d e s c r i b e a n a n t i b o d y p r o d u c e d b y o n e o f the c l o n e s , R N 17. A s d e t e r m i n e d b y i m m u n o b l o t t i n g , this a n t i b o d y r e c o g n i z e s M A P 1 a n d M A P 2 in b r a i n m i c r o t u b u l e p r e p a r a t i o n s , a n d a 100 k D p r o t e i n d o u b l e t in l y s a t e s o f m o u s e e m b r y o fibroblasts. It s t a i n s m i c r o t u b u l e s , i n t e r m e d i a t e f i l a m e n t s a n d c o a t e d vesicles in c u l t u r e d cells. W e h y p o t h e s i z e that the 100 k D p r o t e i n s o f c u l t u r e d cells m a y b e a c o m p o n e n t a s s o c i a t e d w i t h the t r a n s p o r t o f c o a t e d vesicles a l o n g m i c r o t u b u l e s and intermediate filaments. MATERIALS

AND METHODS

Production o f Monoclonal Antibodies Monoclonal antibodies were obtained according to [19] with some modifications. A fraction of brain MAPs was purified by phosphocellulose chromatography [5] from microtubule protein obtained by two cycles of polymerization-depolymerization[20] as described in [21]. BALB/c female mice were immunized intraperitoneally with 1 mg of MAPs mixed with complete Freund adjuvant and were boosted by intravenous injection of the same amount of protein 3 weeks later. After 3 days 108 spleen cells were fused with 10 7 P3X63 Ag8.653 myeloma cells by polyethylene glycol 1500 (Serva, Heidelberg). Cells were plated in 96-well plates on a feeder layer of mouse peritoneal macrophages in Dulbecco's modification of Eagle's medium supplemented with a 15% fetal calf serum (FCS), hypoxanthine, aminopterin and thymidine (HAT medium). Clones producing antibodies to MAPs were selected using an enzyme-linked immunosorbent assay with total MAPs from bovine brain. Positive clones were recloned three times in HAT medium. In order to obtain ascites fluid, hybridoma cells were injected intraperitoneally into pristane-primed mice. Immunoglobulins, precipitated from ascites fluid with ammonium sulfate at 50 % saturation, were used for immunofluorescent staining of cells, immunoelectron microscopy and immunoblotting. RN 17 antibody was of the IgM class.

Fixation, Extraction and Immunofluorescence Microscopy For immunofluorescence staining we used secondary cultures of mouse embryo fibroblasts (MEF), cells of bovine tracheal epithelium (FBT line [22]), and diploid human embryo fibroblasts. The cells were grown in a mixture of Eagle's minimal essential medium (45 %), 0.5 % lactalbumin hydrolysate (45%), bovine serum (9%) and FCS (1%). Cultures were grown on coverslips for 24--48 h, washed three times with phosphate-buffered saline (PBS), then extracted with 0.1% Triton X-100 in a buffer M (50 mM imidazole pH 6.8, 50 mM KCI, 0.5 mM MgCI2, 0.1 mM EDTA, 1 mM EGTA, 1 mM 2mercaptoethanol) supplemented with 4 % polyethylene glycol 40 000, washed with buffer M and fixed with 1% glutaraldehyde buffered with 0.1 M sodium cacodylate, pH 7.2. Fixed cultures were washed thoroughly with PBS, treated with 2 mg/ml sodium borohydride twice for 10 min and then stained with the antibodies by an indirect method. RN 17 antibody against brain MAPs was used at a concentration of about 100 Ixg/ml. All staining of the cells with RN 17 antibody was abolished if the antibody was preincubated with the excess of brain MAPs overnight at 4°C. Monoclonal antibody to tubulin [23] was kindly provided by Drs P. Draber and V. Vicklicky (Institute of Molecular Genetics, Prague); monoclonal antibody to vimentin was a generous gift of Drs A. A. Neyfakh, Jr and I. S. Tint. Polyclonal rabbit antibody to bovine brain tubulin was described earlier [24]. Secondary antibodies were fluorescein- or rhodamine-labelled goat antirabbit immunoglobulins or fluorescein-labelled goat antimouse immunoglobulin(Sigma). These antibodies were used at a dilution of 1 : 50.

Immunoelectron Microscopy Five nm colloidal gold was produced as described in [25] and was conjugated with rabbit monospecific antibody to mouse immunoglobulinsas described in [26].

Exp CellRes 159(1985)

100 kD protein of the cytoskeleton 379 For immunoelectron staining cells were extracted, fixed and stained with primary antibodies as described above. Then cultures were incubated with colloidal gold antibody conjugates, washed with PBS, fixed in 2.5% glutaraldehyde buffered with 0.1 M sodium cacodylate and post-fixed in 1% OsO4. The fixed cultures were embedded in Epon after ethanol-acetone dehydration. Ultrathin sections of these cultures were cut and stained with aqueous uranyl acetate and lead citrate according to Reynolds [27]. The sections were examined and photographed in a Hitachi HU-11B electron microscope operating at 75 kV.

Immunoblotting Immunoblotting was performed essentially as described in [28]. SDS-PAGE was performed according to Laemmli [29], and proteins were transferred from the gels to nitrocellulose sheets (0.45 I~m, Schleicher & Schiill) by electrophoresis in 10 mM boric acid-NaOH buffer at pH 8.8. After washing with PBS the filters were treated for 1 i2 h in 1% bovine hemoglobulin (Serva), washed with PBS, and were incubated for 2 h in the antibody solution containing 1% hemoglobin and 10% non-immune rabbit serum. Then the filters were washed three times with 0.5 % Triton X-100 in PBS and three times in PBS without Triton. Washed filters were incubated for 2 h with peroxidase-conjugated rabbit antibody to mouse immunoglobulin prepared as in [30] and diluted 1 : 250 with PBS containing 1% hemoglobin and 10% rabbit serum. After washing as described above, the bands were revealed by incubation in 3,3'-dianisidine (50 ~tg/ml) in 10 mM Tris-HC1, pH 7.5, containing 0.0015% H202.

RESULTS

Irnmunofluorescent Staining We obtained several hybridoma clones producing antibodies to microtubuleassociated proteins. The specificity of all these antibodies will be described elsewhere. One of the antibodies, named RN 17, was chosen for examination of cultured cells by immunofluorescence and immunoelectron microscopy. For indirect immunofluorescence staining, the cultures were extracted with 0.1% Triton X-100 and fixed as described in Materials and Methods. The results of the staining of mouse embryo fibroblasts by RN 17 antibody are shown in fig. 1 a. This antibody revealed a cytoplasmic network composed of thin fluorescent fibrils. The network was very dense in the perinuclear areas of the cells. Some fibrils of the network reached the cell margins (fig. 1 b). At higher magnifications one can also see numerous small fluorescent dots (fig. 1 b). Such dots were never observed in cells stained with the tubulin or vimentin antibodies (fig. I c, d). A similar pattern was observed in two other cell lines tested so far: cytoplasmic filaments were decorated with RN 17 antibody in diploid human embryo fibroblasts (fig. 1 e) and epithelial cells of bovine trachea (fig. l f). The RN 17 staining pattern depended strongly on the extraction and fixation procedures. In fixed cells that were not permeabilized before staining, the high background fluorescence masked many of the fluorescent fibrils. However, they could easily be seen in the thin parts of the cells, especially in the lamellae (not shown). Both fibrillar staining and diffuse fluorescence were abolished if RN 17 antibody was preincubated with an excess of H M W brain MAPs. Therefore, both diffuse and fibrillar fluorescence after RN 17 antibody staining was specific and RN 17 antigen was present both on the cytoskeletal filaments and in a soluble form in the cytoplasm. Exp Cell Res 159 (1985)

380

R o d i o n o v et al.

Fig. 1. Immunofluorescence microscopy of cultured cells labelled with RN 17 antibody (a, b, e, J), and

antibodies to tubulin (c) and vimentin (d). (a-d) Mouse embryo fibroblasts; (e) human embryo fibroblasts; (]) FBT cells. Bar, (a, c-f) 20 ~tm; (b) 15 ~tm.

Exp Cell Res 159 (1985)

100 kD protein of the cytoskeleton

381

Fig. 2. Double immunofluorescence staining of mouse embryo fibroblasts with (a) rabbit antitubulin;

(b) RN 17 antibody. Staining of mitotic FBT cells with (c) antitubulin; (d) RN 17 antibody; (e, JO double staining of a mitotic cell with (e) Hoechst 33258 and (f) antivimentin. Bar, (a, b, e, f) 20 ~tm; (c, d) 5 Ixm.

The filaments were also clearly stained in the cells extracted with 0.5 % Brij 35. However, RN 17 antibody failed to bind to the cells extracted with 1% Triton X100 in buffer M. The best visualization of the cytoplasmic filaments with RN 17 antibody was achieved after extraction with 0.1% Triton X-100 and glutaraldehyde fixation. This procedure (see Materials and Methods) was routinely used in the present work. The distribution of the filaments stained by RN 17 antibody in the interphase cells was very similar to the distribution of microtubules, as visualized after labelling with the tubulin antibody. Both systems of filaments were most dense in the perinuclear regions of the cells. Like the microtubules, the RN 17-binding filaments reached the cell margins (fig. 1 a, b, c). In contrast, the intermediate filaments of mouse embryo fibroblasts were usually absent from the leading lamella (fig. I d). Double staining with the polyclonal antibody to tubulin and RN 17 antibody confirmed that most of the microtubules bound RN 17 antibody (fig. 2 a, b). Exp Cell Res 159 (1985)

382 R o d i o n o v et al.

Fig. 3. Distribution of (a, d) tubulin; (b, e) vimentin; (c, J0 RN 17 antigen in mouse embryo fibroblasts

treated with 1 vM colcemid for (a-c) 24 h, or (d-D 0.6 M KCI after Triton extraction. Bar, 29 [xm.

In mitotic FBT cells, both RN 17 and the tubulin antibody stained mitotic spindles (fig. 2 c, d). Both antibodies decorated half-spindles and interzone regions. However, the monoclonal antibody to an intermediate filament protein, vimentin, did not stain spindles in mitotic FBT ceils identified by Hoechst 33258 staining (fig. 2 e, f). The results presented above clearly show that RN 17 antibody stains microtubules both in interphase and mitotic detergent-extracted cells. However, the difference in distribution of RN 17 antigen and tubulin was evident in the cells that were treated to disrupt microtubules. Fig. 3 a shows that 24 h treatment of mouse embryo fibroblasts with 0.5 ~tg/ml colcemid led to the complete destruction of cytoplasmic microtubules. In such cells the vimentin antibody revealed a system of intermediate filaments collapsed around nuclei (fig. 3 c). A similar collapsed system was stained by RN 17 antibody in the colcemid-treated cells (fig. 3 b). Complete dissolution of cytoplasmic microtubules could also easily be achieved if Triton-extracted cells were treated for 30 min with 0.6 M KC1 in buffer M (fig. 3 d). As expected, this treatment had little effect on intermediate filaments (fig. 33'). In KCl-treated cells RN 17 antibody revealed fibrils that spread from the perinuclear zone, but like the intermediate filaments, these did not reach the leading edge (fig. 3e). Therefore, in cells with depolymerized microtubules, the distribution of RN 17 antigen coincides with the distribution of Exp Cell Res 159 (1985)

100 kD protein of the cytoskeleton

383

Fig. 4. Localization of RN 17 antigen in mouse embryo fibroblasts by colloidal gold antibody staining.

(a--c) RN 17 antibody; (at) tubulin antibody. (a, c, d) 0.2 ~tm; (b) 0.1 ~tm.

intermediate filaments. We cannot exclude the possibility that RN 17 antigen associated with intermediate filaments after microtubule disassembly. However, a more probable explanation of this result is that RN 17 antigen in the cells is bound not only to microtubules but also to intermediate filaments, and that microtubule disruption simply improves the visualization of RN 17 antigen bound to intermediate filaments. To verify this explanation and to study RN 17 antigen distribution in greater detail, we used immunoelectron microscopy.

Immunoelectron Microscopy For immunoelectron microscopy we used indirect immuno-gold staining. Triton-extracted cells were stained with RN 17 or tubulin antibodies and then were secondarily labelled with the monospecific rabbit antibody against mouse immunoglobulins coupled to 5 nm colloidal gold. The results of the staining of mouse embryo fibroblasts with tubulin antibody are shown in fig. 4 d. Gold particles were observed to be homogeneously distributed along microtubules in the section. A few particles were seen randomly scattered in the cytoplasm. After RN 17 antibody staining, both microtubules and intermediate filaments were labelled (fig. 4 a, c). The gold particles on the surface of the microtubules in RN 17-stained 25-858338

Exp Cell Res 159 (1985)

384 Rodionov et al. 1

2

3

4

5

6

7

!e

iO

iO' Fig. 5. [mmunoblot analysis with l, 2, RN 17 antibody of

I

microtubule proteins; 3, 5, proteins of mouse embryo fibroblasts; 4, 6, FBT cells. 1, 3, 4, Electrophoresis; 2, 5, 6, RN 17 antibody staining; 7, MW markers (from top to bottom: myosin, 205 000; fl-galactosidase, 116 000; phosphorylase b, 97400; bovine serum albumin, 66000; ovalbumin, 45 000; carbonic anhydrase, 29000).

cells were usually clustered. Such clusters were never observed in the cells stained with tubulin antibody. RN 17 antibody also stained the surfaces of vesicles about 0.1 ~tm in diameter (fig. 4 b). These vesicles are most probably coated vesicles, as in most sections a 15 nm thick coat covering the vesicles could be seen. The vesicles stained by RN 17 and colloidal gold most likely corresponded to the tiny dots seen after immunofluorescence staining (figs 1 b, 3 b). The vesicles were not stained by the antibody to tubulin (fig. 4 d). Therefore, the results of immunoelectron staining are in close agreement with the immunofluorescent data. These results show that the antigen, recognized by RN 17 antibody, is associated with microtubules, intermediate filaments and coated vesicle detergent-extracted cells.

Identification of Antigen Recognized by RN17 Antibody The antigenic components recognized by RN 17 antibody were identified by immunoblotting (fig. 5). In brain microtubule preparations RN 17 antibody bound to two polypeptides with the same electrophoretic mobilities as MAP 1 and MAP2 (fig. 5, 1, 2). It is known, however, that HMW MAPs are very sensitive to proteolysis, and proteolytic degradation of MAP 1 can produce polypeptides with electrophoretic mobilities similar to those of MAP2 [31]. To prove that the smaller of the two RN 17-reactive polypeptides in the microtubule preparation was in fact MAP2, we heated brain microtubule proteins for 5 min at 100°C in buffer M supplemented with 1 M KCI. This treatment precipitated MAP 1, while the heat-stable MAP2 remained soluble. Immunoblotting with this preparation Exp Cell Res 159 (1985)

100 kD protein of the cytoskeleton

385

revealed one RN 17-reactive polypeptide co-migrating during electrophoresis with MAP 2 (not shown). Therefore, RN 17 antibody recognized both MAP 1 and MAP 2 in brain microtubule preparations. We have also examined the identity of the proteins in cultured cells that are recognized by RN 17 antibody. This antibody bound to a 100 kD protein doublet in the lysate of mouse embryo fibroblasts (fig. 5, 3, 5). The same doublet was recognized in the fibroblasts extracted with 0.1% Triton X-100 in buffer M (the procedure used for preparing cells for immunofluorescence staining). The I00 kD doublet was also recognized by RN 17 antibody in FBT epithelial ceils (fig. 5, 4, 6). In these cells RN 17 antibody labelled a polypeptide with a MW of 210 kD. This protein had the same electrophoretic as the 210 kD component of neurofilaments from bovine brain, but purified neurofilament proteins did not react with RN 17 antibody as determined by immunoblotting (not shown). It should be noted that RN 17 antibody did not bind to any protein of cultured cells with a MW close to that of MAP 1 or MAP 2. This result is probably not due to proteolysis of HMW MAPs, since we analysed the samples immediately after dissolving cells in SDS solution and boiling the lysate. DISCUSSION The results presented above show that antibody RN 17 recognizes a 100 kD MW protein(s) in a number of cultured cells. This protein: (a) has common antigenic determinants with major brain microtubule-associated proteins MAP 1 and MAP2; (b) is associated with microctubules, intermediate filaments and coated vesicles in detergent-extracted cells. The polypeptide composition of microtubules from non-neuronal cells has been described in a number of papers [11-18, 32--42]. It is clear that the major microtubule-associated proteins in many cultured cells are different from MAP 1 and MAP2 [I1-15], although both MAP1 [37, 38] and MAP2 [17, 18 39-42] have also been observed to be components of the microtubule network in non-neuronal cells (see, e.g. [18, 37--42]). One might anticipate that antibody RN 17, which binds to MAP 1 and MAP 2 in brain microtubule preparations, would recognize the same two polypeptides in cultured cells. However, immunoblotting clearly showed that this antibody binds to a 100 kD protein in fibroblasts and to both 100 and 210 kD proteins in FBT epithelial cells. It is noteworthy that FBT cells contain MAP 2 [40]. MAP 2 from these cells reacts with a polyclonal antibody to MAP 2 and is a component of cytoplasmic and mitotic microtubules. Probably, both MAP 1 and MAP2 in cultured non-neuronal cells are slightly different from their brain counterparts and our monoclonal antibody is directed to the epitope presented only in brain proteins. Another probable explanation of the inability of RN 17 antibody to react with MAP 2 on immunoblots of cultured cells is that the cells of non-neural origin contain too small amounts of MAP 2 to be determined by this technique. Exp Cell Res 159 (1985)

386 Rodionoo et al. At present, we do not know whether 100 kD protein recognized by RN 17 antibody is really related to the HMW microtubule-associated proteins MAP 1 and MAP2, or whether this antibody merely recognizes an epitope shared by both the brain microtubule-associated proteins and the 100 kD protein. The answer to this question will require the direct biochemical analysis of purified components. The immunofluorescence data show that RN 17 antibody labels microtubules, intermediate filaments and small particles in the detergent-extracted cultured cells. The same staining pattern was observed in the thin lamellar cytoplasm of fixed non-extracted cells. However, in the thick central parts of the cells its visualization was obscured by background fluorescence. This background seems to be due to the presence of Triton-soluble RN 17 antigen in the cytoplasm because both fibrillar staining and diffuse fluorescence were abolished if RN 17 antibody was neutralized with an excess of HMW brain MAPs. It is possible that the fibrillar staining we observed is due to the redistribution of antigen following Triton extraction, but this possibility seems unlikely, since similar patterns of fibrillar staining are also observed in the thin regions of unextracted cells. Our observation of immunofluorescence staining patterns is in close agreement with the data obtained by immuno-gold staining. Immunoelectron microscopy showed that RN 17 antibody decorates clusters of particulate material that seem to be randomly distributed on microtubules, and intermediate filaments. The most unexpected result was the staining of coated vesicles by RN 17 antibody. Most probably, the tiny dots seen after immunofluorescence staining correspond to the coated vesicles revealed by immuno-gold labelling and electron microscopY. On the basis of the immunological analysis presented above we suggest that 100 kD protein is specifically associated with microtubules, intermediate filaments and coated vesicles. This result may be related to previous b;ochemical data, describing the interaction between microtubules and coated vesicles [42-44]. If this is the case, and 100 kD protein is really a component of microtubules, intermediate filaments and coated vesicles, it seems possible that this protein may mediate a form of intracellular transport of coated vesicles that is dependent on the cytoskeleton. We are planning to use RN 17 antibody as a means of studying this possibility. We are grateful to Professor J. M. Vasiliev who provided his laboratory facilities for producing monoclonal antibodies. We would also like to thank Professors A. S. Spirin and Yu. S. Chentsov for their support and Drs A. A. Neyfakh, Jr and D. B. Murphy, for their comments on the manuscript.

REFERENCES 1. Schliwa, M W, Cell and muscle motility, vol. 5, p. 1 (1984). 2. Dentler, W L, Granet, S & Rosenbaum, J L, J cell biol 65 (1975) 237. 3. Murphy, D B & Borisy, G G, Proc natl acad sci US 72 (1975) 2696. Exp Cell Res 159 (1985)

100 kD protein of the cytoskeleton

387

4. Keats, R A B & Hall, R H, Nature 257 (1975) 418. 5. Weingarten, D B, Lockwood, A, Hwo, S-Y & Kirschner, M W, Proc natl acad sci US 72 (1975) 1858. 6. Cleveland, D W, Hwo, S-Y & Kirschner, M W, J mol biol 116 (1977) 207. 7. Zingsheim, H-P, Herzog, W & Weber, K, Eur j cell biol 19 (1979) 175. 8. Kim, H, Binder, L I & Rosenbaum, J L, J cell biol 80 (1979) 266. 9. Vallee, R B & Davis, S, Proc natl acad sci US 80 (1983) 1342. 10. Kuznetsov, S A, Rodionov, V I, Gelfand, V I & Rosenblat, V A, Biochem biophys res commun 119 (1984) 173. 11. Doenges, K H, Nagle, B W & Bryan, J, Biochemistry 16 (1977) 3455. 12. Nagle, B W, Doenges, K H & Bryan, J, Cell 12 (1977) 573. 13. Bulinski, J C & Borisy, G G, Proc natl acad sci US 76 (1979) 293. 14. Wiche, G, Lawrence, S, Honig, L S & Cole, R D, J cell biol 80 (1979) 553. 15. Weatherbee, J A, Luftig, R B & Weihing, R R, Biochemistry 19 (1980) 4116. 16. Cleveland, D W, Spiegelman, B M & Kirschner, M W, J biol chem 254 (1979) 12670. 17. Weatherbee, J A, Sherline, P, Mascardo, R N, Isant, J G, Luftig, R B & Weihing, R R, J cell biol 92 (1982) 155. 18. Valdivia, M M, Avila, J, Coil, J, Calaco, C & Sandoval, I V, Biochem biophys res commun 105 (1982) 1241. 19. Galfre, G, Howe, S C, Milstain, C, Butcher, G W & Howard, J C, Nature 266 (1977) 550. 20. Shelanski, M L, Gaskin, F & Cantor, C R, Proc natl acad sci US 70 (1973) 765. 21. Kuznetsov, S A, Rodionov, V I, Gelfand, V I & Rosenblat, V A, FEBS lett 95 (1978) 339. 22. Machatkova, M & Pospisil, Z, Folia biologica (Praha) 21 (1975) 117. 23. Viklicky, V, Draber, P, Hasek, J & Bartek, J, Cell biol int rep 6 (1982) 725. 24. Bershadsky, A D, Tint, I S, Gelfand, V I, Rosenblat, V A, Vasiliev, J M & Gelfand, I M, Cell biol int rep 4 (1978) 345. 25. Mihlpfordt, H, Experientia 38 (1982) 27. 26. Horrisberger, M, Rosset, J & Bauer, H, Experientia 31 (1975) 1147. 27. Reynolds, E S, J cell biol 17 (1963) 208. 28. Towbin, H, Staehelin, T & Gordon, J, Proc natl acad sci US 76 (1979) 4350. 29. Laemmly, U K, Nature 227 (1970) 680. 30. Wilson, B & Nakane, P K, Immunofluorescence and related staining techniques (ed W Knapp) p. 215. North-Holland Biomedical Press, Elsevier, Amsterdam (1978). 31. Hill, A M, Maunoury, R & Pantalony, D, Biol cell 41 (1981) 43. 32. Wiche, G & Cole, R D, Proc natl acad sci US 73 (1976) 1227. 33. Weatherbee, J A, Luftig, R B & Weihing, R R, J cell biol 78 (1978) 47. 34. Solomon, F, Magendantz, A & Salzman, A, Cell 18 (1979) 431. 35. Bulinski, J C & Borisy, G G, J biol chem 255 (1980) 11570. 36. Duerr, A, Pallas, D & Solomon, F, Cell 24 (1981) 203. 37. Bloom, G S, Luca, F C & Vallee, R B, J cell biol 98 (1984) 331. 38. Wiche, G, Briones, E, Koszka, G, Artlieb, U A & Krepler, R, EMBO j 5 (1984) 991. 39. Sheterline, P, Exp cell res 115 (1978) 460. 40. Kuznetsov, S A, Rodionov, V I, Gelfand, V I & Rosenblat, V A, Cell biol int rep 4 (1980) 1017. 41. Sloboda, R D & Dickersin, K, J cell biol 87 (1980) 170. 42. Imnof, B A, Marti, U, Boiler, K, Frank, H & Berchmeier, W, Exp cell res 145 (1983) 199. 43. Wiedenmann, B & Mimms, L F, Biochem biophys res commun 115 (1983) 303. 44. Kelly, W G, Passaniti, A, Woods, J W, Daiss, J L & Roth, T F, J cell biol 97 (1983) 1191. Received December 7, 1984 Revised version received March 21, 1985

Exp Cell Res 159 (1985)