crotubule-associated proteins l and 2, the 240-kDa chain of brain .... Labeled sam- ples were ..... Selden, S. C., and Pollard, T. D. (1986) Ann. N. Y. Acad. Sci.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc
Vol. 262, No. 3, Issue of January 25, pp. 1320-1325,1987 Printed in U.S.A.
Plectin and IFAP-300K Are Homologous Proteins Binding to Microtubule-associated Proteins 1 and 2 and to the240-kilodalton Subunit of Spectrin” (Received for publication, September 4, 1986)
Harald Herrmannzand Gerhard Wiche From the Institute of Biochemistry, University of Vienna, Wahringer Strasse 17, A-1090 Vienna, Austria
Structural and functional characteristics of plectin from intermediate filament preparations of rat glioma C6 cells were compared to those of the intermediate filament-associated protein of M, = 300,000 (IFAP300K) of baby hamster kidney cells (Yang, H A . , Lieska, N., Goldman, A.E., and Goldman, R. D. (1985) J. Cell Biol. 100, 620-631). After radiolabeling and proteolytic digestion under varied conditions, both proteins yielded nearly identical peptide maps. Immunological cross-reactivity, co-migration on one- and two-dimensional high-resolution gels, chromatofocusing, and amino acid analysis demonstrated structural homology as well. In vivo labeling with 32Pishowed that plectin was the target for CAMP-independent protein kinaseswhich phosphorylated 18-kDa domains at the end(s) of the molecule. Previously reported phosphorylation sites for CAMP-dependent and a newly identified site for Ca2+/calmodulin-dependentprotein kinases were located on different domains. In solidphase binding assays, plectin bound to vimentin, microtubule-associated proteins l and 2, the 240-kDa chain of brain fodrin, and a-spectrin from human erythrocytes. Similar characteristics wererevealed for corresponding 300-kDa components of various other cell lines, supporting the concept that plectin is a general cytoskeletal cross-linking element, probably of multiple function.
criteria (2, 3). A 300-kDa band immunoreactive with antiserum to C6 cell plectin has been shown to occur inIF preparations from a variety of cell types and tissues (2-5). In immunofluorescence microscopy studies using a polyclonal antibody preparation, this antigen was visualized throughout the cytoplasm of several tissue cell types (4, 5) and in dense cytoplasmic network arrays of cultured cell lines (2, 3, 5, 6). Morever, in combination with immunoelectron microscopy, plectin was localized at junctional sites of various cell types such as hepatocytes and epithelial cells and at attachment sites of cytoplasmic filaments including Z discs and dense plaques of striated and smooth muscle (4, 5). Recently, a high M , polypeptide, referred to as IF-associated protein of M, = 300,000 (IFAP-300K), has been identified in IF preparations of baby hamster kidney (BHK-21) cells (7, 8). This protein codistributed with BHK-21 IFs in situ as determined using a monoclonal antibody preparation, andcocycled with the IF subunit proteins from BHK-21 cells in rounds of in vitro assembly/disassembly (7). However, because of its selective association with IFs of cultured cells, Goldman and co-workers concluded that IFAP-300K and plectin were distinct proteins(7,8). Therefore, a directcomparison of both polypeptides by chemical as well as immunological methods became important. The data reported here suggest that both proteins, like those of various other fibroblast cell lines, indeed are closely related in primary structure.Furthermore, the identification of several new interaction partners of plectin by solid-phase binding assays supports the concept that this polypeptide is a general cytoplasmic cross-linker.
Plectin, a phosphoprotein of M , = 300,000, originally was EXPERIMENTAL PROCEDURES~ identified as a major component copurifying with intermediate filaments (IFs)’ after high salt/l% Triton X-100 extracRESULTS tion ofC6 cells (1). It is a prominent substrate for CAMPdependent as well as CAMP-independent protein kinases asChemical Comparison of Plectin and IFAP-SOOK-Extracsociated with the cytoskeleton (2). Plectin was first supposed tion of cultured rat glioma C6 cell monolayers with buffers to be related to MAP 2 (l),but was later shown to be a containing 0.77 M NaCl/l% TritonX-100 yielded an insoluble distinct proteinbased on peptide mapping and immunological residue that consisted predominantly of a M , = 300,000 protein, plectin (6), and the IF subunit protein, vimentin (Fig. * This work was supported in part by grants from the Austrian 1).A protein component of electrophoretic mobility identical Research Fund (Osterreichischer Fonds zur Forderung der Wissen- to thatof plectin was prominent also in parallel preparations schaftlichen Forschung). The costs of publication of this article were from BHK-21 cells, together with vimentin and desmin (Fig. defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 1).The 300,000-Da proteins from both cell types were immunoprecipitated to a similar extent using antibodies to C6 U.S.C. Section 1734 solely to indicate this fact. ~~
$ Recipient of a postdoctoral fellowshipfrom Boehringer Ingelheim Funds. Made an oral presentation of the results contained in this paper a t the Second European Congress on Cell Biology, Budapest, Hungary, July 7, 1986. ‘The abbreviations used are: IF,intermediate filament; IFAP300K, IF-associated protein of M, = 300,000; MAP, microtubuleassociated protein; SDS, sodium dodecyl sulfate; PMSF, phenylmethylsulfonyl fluoride; 1-d,one-dimensional; DTT, dithiothreitol; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraaceticacid; NP-40, Nonidet P40.
Portions of this paper (including “Experimental Procedures” and Figs. 5 and 7) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 86M-3044, cite the authors, and include a check or money order for $2.40 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.
1320
Interaction of Plectin with MAPSand Spectrin
1321
C
*
.rl
”
FIG. 1. Identification of plectin in high saltnriton X-100insoluble cell fractions.Rat glioma C6 (lanes 1 and 2) and BHK21 (lanes 3 and 4)cell monolayers were labeledwith t3H]-leucine.IFs
205-
fl
’4
were prepared by procedue B (see “Experimental Procedures”) and boiled after addition of electrophoresissample buffer. Immunoprecipitation from boiled samples was done with antibodies to C6 plectin under radioimmuneprecipitation buffer conditions (28) as described (2). Fluorographs of proteins (lanes 1 and 3) and corresponding immunoprecipitates (lanes2 and 4 ) separated on 7.5%gels are shown. FIG. 3. High-resolution gel electrophoresis of IFAP-SOOR/ The positions of plectin, vimentin, and desmin are indicated by plectin and fingerprint analysis. A, IF preparations from BHKnumbers ( M , X 21 and C6 cells were subjected to gel electrophoresis on 5% polyacrylamide gels (22 cm long). Only the upper part of the gel is shown. W W B and C, the 300,000-Da band ( B ) shown in A and the next lower P$ $ $N band (C) from C6 cells were cut out from the gel, labeled with lZ5I, N and subjectedto tryptic fingerprinting (29). M , X
55 44
36 31
2 50
500
ng V8
FIG. 2. Peptide mappingof plectin and IFAP-SOOK.IF residues prepared by procedureB were labeled by reductivemethylation, and 300,000-Da protein bands were excised from 5% gels after autoradiographic detection and digested during re-electrophoresis with 0.5 pg of V8 protease, and fragments were separated on 12.5%gels. M , X 10-3.
plectin (Fig. 1, lanes 2 and 4). This indicated that the two proteins were immunologically related. We then labeled IF preparations from both cell lines in vitrowith tritium by reductive methylation and isolated the bands of apparent M , = 300,000 corresponding to plectin and IFAP-300K from5% polyacrylamidegels. One-dimensionalpeptide mapping by the limited proteolysis technique (9) using V8 protease produced nearly identical peptide patterns from both proteins (Fig. 2). Analogous results were obtained using chymotrypsin and subtilisin (data not shown). This indicated a highdegree of similarity in primary structure and confirmed the immunological relationship of both proteins. Upon analysis of IF preparations on high-resolution gels, polypeptides of M, between 300,000 and 260,000 became apparent in addition to the co-migrating 300,000-Dabands (Fig. 3). These bands, too,
reacted with antibodies to plectin (data not shown). In accordance with data reported for IFAP-300K (7), one-dimensional peptide mapping (data not shown) as well as tryptic fingerprinting (Fig. 3) revealed similar patterns indicating that these lower bands most likely were derived by endogenous proteolysis of the 300,000-Da polypeptide. Analogousresults were obtained for the high M , polypeptides in IF preparations from other cell lines (data notshown). Lieska et al. (8) reported that IFAP-300K could not be focused under conventional isoelectricfocusing conditions because most of the material did not enter the focusing gel. However, followingthe method of Garrels (lo), we found that a considerablepart of the 300,000-Da polypeptidespresent in C6 and BHK-21 IF preparations entered the first-dimension gel and focused at a pH of 4.7-5.0 (Fig. 4). These results were corroborated by determining the PIof C6 plectin on chromatofocusing columns. Like IFAP-300K (€9,plectin eluted in pH fractions of 5.5-4.9 (Fig. 5). Upon washes with polybuffer of pH 2.4, the remaining plectin came off (last three lanes in Fig. 5). To verify the structural relationship of the 300-kDa protein species eluting with different PI value, 300-kDa bands from fraction at pH 5.3, 4.0, and 3.6 as well as from the sample loaded (Fig.5) were iodinated and subjected to limited digestion with V8 protease. All cleavage patterns obtained were virtually identical (data not shown). Furthermore, the amino acid analysis of C6 plectin and BHK-21 IFAP-BOOK revealed no differen~e.~ Taking these data together, we conclude that the 300-kDa proteins present in IF preparations from BHK21 and various other cultured cell lines are homologous to C6 cell plectin. Localization of Phosphorylation Sites-As previously observed for Chinese hamster ovary cells (2), the major phosphorylated V8 fragments derived from C6 plectin and BHK21 IFAP-300K wereof M , = 14,000-18,000 as revealed by analysis on high percentage (15%) polyacrylamide gels (Fig. 6). Hardly any labeled fragments were seen between this M, range and 300,000, a fact strongly corroborated by analysis on low percentage (6.25%) gels. In contrast, parallel experiG . Weitzer, unpublisheddata.
Interaction of Plectin with MAPS and Spectrin
1322
-
c6
BHK-21 c
akl P'
-
*
" P
A
4.8
IEF
A
5.3
A
4.8
- eP
A G E
A
5.3
FIG.4. Two-dimensional gel electrophoresis of plectin. C6 or BHK-21 cells were labeled in vivo with 3zPi,and IFs prepared by procedure B were taken up in sample solution as described by Garrels (10). Isoelectric focusing (ZEF) was performed on 14-cm-long 3.5% polyacrylamide gels containing 2% Serva pH 3-10 and 2% Serva pH 4-6 ampholytes. The basic end (2 cm) of the focusing gel was discarded in order to prevent contamination with nonfocused proteins during pre-equilibration with electrophoresis sample buffer. Probably because less was loaded, C6 plectin focused somewhat sharper than BHK-21 plectin. For measuring the pH gradient, gelsfocused in parallel were cut into5-mm slices, each slice was incubated with 1ml was determined on a pH of 9.2 M urea in degassed H20, and the pH meter (30). P, plectin; V ,vimentin; D, desmin. PAGE, polyacrylamide gel electrophoresis.
ments with 3H- (Fig. 2) and '251-labeledplectin4 showed a multitude of fragments in the M , range from approximately 100,000 to 300,000. Thus, theremoval of low M, phosphopeptides by V8 eliminated all of the label from the molecule and left behind unlabeled fragments of high Mr. These fragmentation patternswere observed over a wide concentration range of protease (1 ng to 10 pg). These data strongly suggest, therefore, that plectin is phosphorylated at the end(s)of the polypeptide chain. In addition tothis bulk label, we observed labeled V8 fragments of -30 kDa that were only generated at medium proteaseconcentration. By incubating permeabilized cells with [-p3'P]ATP in the presence or absence of Ca'+/calmodulin, these fragmentswere shown to be derived from a domain of plectin phosphorylated exclusively by a Ca2+/calmodulindependent protein kinase (Fig. 7). Thus, thephosphorylation sites affected by CAMP-independent and Ca2+/calmodulindependent kinasesclearly were located on differentmolecular domains. Solid-phaseBinding Assays-In order to detect possible binding partners of plectin, blots of proteins to be assayed were overlaid with 32P-or lZ5I-labeledC6 plectin. 32P-labeled plectin bound to vimentin from IF preparations4 as well as to chromatographically purified vimentin (Fig. 8A, lune I), confirming previous copolymerization experiments (6): Furthermore, lZ5I-plectinbound to MAP 1 and MAP 2 of two-timescycled hogbrain microtubule preparations, but not totubulin (lane 2). In agreement with plectin's binding to a prominent 240-kDa protein of brain membrane preparations (data not shown), lZ5I-plectinbound to the 240-kDa subunit of fodrin purified from brain (lane 3). Moreover, lZ5I-plectinbound to
B
A 1 213243
32P
f
I
T;
k urn
w
-330 -' -300
-
C D I
300
-4 4
240
-
FIG.8. Binding of purified plectinto proteins transblotted to nitrocellulose. A, overlays with 32P(lune 1 ) - or lZ5I(lanes 2-4)labeled plectin. 3ZP-Labeledplectin (1.2-1.6 X lo3 cpmlpg) was purified (see text) from cells radiolabeled in vivo with 32Pi(2) and used at a concentration of 0.1-1 pg/ml. lZ5I-labeledplectin (0.2-2.0 X lo6 cpm/pg) was prepared by labeling chromatographically purified plectin with 9-labeled Bolton-Hunter reagent in 100 mM borate (pH 8.5) for 30 min on ice and was used at 0.05-1.0 pg/ml. Gel blots of 55 the following purified samples (see "Experimental Procedures") are shown: lane I, C6 cell vimentin; lune 2, hog brain microtubule proteins; lane 3, hog brain fodrin; lane 4, human erythrocyte spectrin. Blots from full-length gels are shown. Lane 1 was from a 6.25% gel, and lanes 2-4 were from 5% gels. Autoradiography was for 16 h to 4 6.25% 15% days. B , overlays with unlabeled plectin. Protein blots of chromatoFIG.6. Proteolytic fragmentation patterns of 32P-labeled graphically purified hog brain fodrin preparations, separated on 5% plectin. Rat glioma C6 and BHK-21 cells were labeled in the loga- gels (lane 1 , Coomassie Brilliant Blue stain), were overlaid with 0.4 rithmic growth phase with 32Piusing 0.5-0.1 mCi/ml for 3-12 h as pg of chromatographically purified plectin/ml of PBS (lune 2 ) or with described (2) except that no depletion was performed. Labeled sam- PBS alone (lane 3 ) , followed by antiserum to C6 plectin. Bound ples were then run on 5% gels, and plectin bands detected by auto- immunoglobulins in lanes 2 and 3 were detected by the Proto-Blot radiography wereexcised. Plectincontainedinthese bands was color reaction (see text). digested during re-electophoresis with 0.5 pg of V8 on 6.25 or 15% 'H. Herrmann and G. Wiche, unpublished data. gels, as indicated. M,X 100
-
-
I'
-L
Interaction of Plectin with MAPs and Spectrin
1323
we had ruled out any immunological relationship between plectin and MAPs; furthermore, peptide mapping showed that plectin and MAP 2 are not homologous (2, 3). Differences existin cellular associationsreported forIFAP-300K and plectin. Whereas IFAP-300K was reported to be selectively associated with IFs of BHK-21 cells (7, 8), plectin has been shown t o be widespread both with regard to cell types and organelles (2-6, 11).Caution is necessary, however, in comparing both setsof data because a single monoclonalantibody preparation was used in the studies on IFAP-300K,whereas all studies on plectin antigenswere carried out with conventional affinity-purified rabbit antibodies. Thelikelihood that accessibility of epitopes for antibodies becomes a limiting factorthatultimately may lead toanapparentrestricted distribution of a particular antigen is much higher in studies with a single monoclonal antibody compared to those with conventional rabbit antibodieswhich recognize several different epitopes along one polypeptide chain. In fact, following the observations suggest the association of the 300-kDa protein from BHK-21 cells with cellular components other than IFs. First, like C6 plectin, the protein specifically bound toseveral polypeptides asidefromvimentin in solid-phasebindings assays. Second, analogous to C6 plectin, themajor part of the DISCUSSION protein was found in thehigh salt/Triton X-100-soluble fracprotein^.^ In this report,we show that plectin,originally identified in tions in contrast to IF subunit In an extension of previous studies, the bulk of plectin’s high salt/detergent-resistantcytoskeletal residues from rat glioma C6 cells, andIFAP-300KfromBHK-21 cells are phosphorylation sites, acted on by a CAMP-independent kihomologous proteins. Homology was demonstrated by im- nase, have been located on domains less than 20 kDa from munological cross-reactivity, peptide maps, co-migration on the end(s) of the molecule. The phosphorylationof plectin a t high-resolution gels, identical PI ontwo-dimensional gels and the molecular end domains is similar to that of other cytochromatofocusing columns, similar aminoacid composition,3 skeletal proteins of high M,, e.g. P-spectrin ( E ) , myosin (13), and comparable characteristicsof phosphorylation by endog- the neurofilament protein H (14), fibronectin (15), and, to a enous protein kinases. The two proteins also show an analo- certain degree, MAP 2 (16). In this study, plectin has been gous distributiontovarious cell fractions (2): Moreover, identified additionally as a substrate for Ca2+/calmodulindependent kinases. The site(s)of this phosphorylationresides 300,000-Da proteinsfoundinallother cell linestested (Chinese hamster ovary, 3T3, HeLa, A-431, and WI-38) were on a V8-generated 30-kDa fragment of unknown molecular immunologically related and hardly distinguishable from plec- location. Furthermore, CAMP-dependent phosphorylation occurs on a distinct 25-kDa fragment whose location is also tin by peptide mapping. Thus, plectin’s primary structure unknown (2). Therefore, these three kinases seem to phosseems to havebeen highly conserved in evolution. Yang et ul. (7) pointed out that BHK-21 cell IFAP-300K phorylate plectin on distinct domains of the molecule. The finding that three independently regulated kinase systems did not show cross-reactivity with MAPs; and therefore the operate on the plectin molecule opens up ample possibilities protein seemed to be distinct from plectin, which originally we had claimed to be structurally related to MAPs(1).How- for sophisticated regulation at themolecular level. ever, in later studies with well-defined antibody preparations, Solid-phase binding assays, commonly used to demonstrate specific protein interactions (17-20), led to the identification of several interaction partners of plectin including vimentin, plasma membrane MAP 1, and MAP2. The in vitro interaction of plectin with I C6 microtubules,assembled by temperature-dependentor microfilaments - spectrin taxol-induced polymerization, has previously been shown usI ingimmunoblottingandimmunoelectron microscopy (21). Copolymerization experiments also indicated plectin’s inter- plectinaction with vimentin filaments(6, 8).3Plectin’s ability to bind I to vimentin and high M , MAPs, as demonstrated here, sugIF gests that one of its possible cellular functions is the crossmicrotubules I \ linking of cytoplasmic microtubules with IFs. The physical MAP1,2 - plectin connection of these two systems has been proposed based on / microfilaments I their cellular codistribution (22) and the collapse of IFs into IF perinuclear coils after disruption of microtubules with colcemid (23, 24). Other binding partners of plectin were the 240,000-Da chain of hog brain fodrin and the a-spectrin chain nuclear membrane of human erythrocytes. Both proteins have been shown to possess similar primary structure,whereas P-spectrin and the FIG. 9. Schematic model of plectin’s integration into cyto235,000-Da chain of fodrin are each unique (25). Thus, plectin skeleton. Taking into account plectin’s binding to vimentin, MAP 1, MAP 2, and the 240-kDa spectrin-type molecules as shown in Fig. probably binds to acommon domain of these 240,000-Da spectrins. 9, the protein could be engaged in the construction of the threedimensional lattice between plasma membrane and nucleus. If the binding partnersidentified in vitro are also interacta-spectrin from human erythrocytes (lune 4 ) . Overlays with heat-denatured lZ5I-plectinshowed no binding to any of the blotted proteins (data not shown), indicating that a native conformation of plectinisneeded forbinding. Infurther studies, the bindingof plectin to fodrin was shown to depend on the plectin concentration, and unlabeled plectin was found to competewith the labeled protein for binding (data not shown). The binding of plectin to the 240-kDa polypeptide chain of fodrin, a-spectrin, MAP 1, and MAP 2 was further demonstrated by similar overlays usingunlabeledplectin whose binding was detected by incubation with antibodies to plectin followed by secondary antibody coupled to alkaline phosphatase, as shown for fodrin in Fig. 8B. Control blots that were not incubated with plectinshowed no reaction with plectin antibodies, eliminating thepossibility of fodrin crossreactingwithantibodiestoplectin (Fig. 8B, lune 3 ) . The various binding patterns of plectin were not influenced by 1 mM ATP or by Ca2+/calmodulin. Binding of plectin to rabbit muscle actin or myosin was not observed (data not shown). Comparableresults were obtainedinparallelexperiments carriedoutwithplectinfromChinesehamster ovary and BHK-21 cells.
Interaction of Plectin with MAPS and Spectrin
1324
14. Geisler, N., Fischer, S., Vanderkerckhove, J., Van Damme, J., Plessmann, V., and Weber, K. (1985)EMBO J. 4,57-63 15. Hynes, R. (1985)Annu. Reu. Cell Biol. 1,67-90 16. Bums, R. G., and Islam, K. (1984)Eur. J. Biochem. 141, 599608 17. Davies, J., and Bennett, V. (1983)J. Bwl. Chem. 258,7757-7766 18.Heimann, R., Shelanski, M. L., and Liem, R. K. H. (1985)J. Bwl. Chem. 260,12160-12166 19. Leiser, M., Rubin, C. S., and Erlichman, J. (1986)J. Bwl. Chem. 261,1904-1908 20. Glenney, J. R., Jr., and Weber, K. (1983)Methods Enzymol. 102, 204-210 21. Koszka, C., Leichtfried, F. E., and Wiche, G. (1985)Eur. J. Cell Biol. 38,149-156 22. Singer, S. J., Ball, E. H., Geiger, B., and Chen, W.-T. (1982)Cold Spring Harbor Symp. Quunt. Bwl. 46,303-316 23. Ishikawa, H., Bischoff, R., and Holtzer, H. (1969)J. Cell Biol. 38,538-555 Acknowledgments-We thank John Wyatt and J. Mitchell Dalton 24. Goldman, R. D. (1971)J. Cell Biol. 51,752-762 for expert technical assistance and Benjamin Feldman for comments 25. Glenney, J. R., Jr., Glenney, P., and Weber, K. (1982)Proc. Natl. on the manuscript. Acad. Sci. U. S. A. 79,4002-4005 26. Selden, S. C., and Pollard, T. D. (1986)Ann. N . Y. Acad. Sci. 466,803-812, and references therein REFERENCES 27. Brenner, S. L., and Korn, E. D. (1979)J. Biol. Chem. 254,86201. Pytela, R., and Wiche, G. (1980)Proc. Natl. Acad. Sci. U. S. A. 8627 77,4808-4812 28. Gilead, Z., Jeng, Y.-H., Wold, W. S. M., Sugawara, K., Rho, H. 2. Herrmann, H., and Wiche, G. (1983)J. Bwl. Chem. 258,14610M., Harter, M. L., and Green, M. (1976)Nature 264,264-266 14618 29. Elder, J. H., Pickett, R.H., 11, Hampton, J., and Lerner, R.A. 3. Wiche, G., and Baker, M. (1982)Exp. Cell Res. 138, 15-29 (1977)J. Bid. Chem. 252,6510-6515 4. Wiche, G., Krepler, R., Artlieb, U., Pytela, R., and Denk, H. 30. OFarrell, P. H. (1975)J. Biol. Chem. 250,4007-4021 (1983)J. Cell Bwl. 97,887-991 31. Starger, J. M., Brown, W. E., Goldman, A. E., and Goldman, R. 5. Wiche, G., Krepler, R., Artlieb, U., and Aberer, W. (1984)Exp. D. (1978)J. Cell Biol. 78,93-109 Cell Res. 155,43-49 32. Herrmann, H., Pytela, R., Dalton, J. M., and Wiche, G. (1984)J. 6. Wiche, G., Herrmann, H., Leichtfried, F., and Pytela, R. (1982) Biol. Chem. 259,612-617 Cold Spring Harbor Symp. Quant. Bwl. 46,475-482 33. Laemmli, U. K. (1970)Nature 227,680-685 7. Yang, H.-Y., Lieska, N., Goldman, A. E., and Goldman, R. D. 34. Herrmann, H., Dalton, J. M., and Wiche, G. (1985)J. Biol. Chem. (1985)J. Cell Biol. 100, 620-631 260,5797-5803 8. Lieska, N., Yang, H.-Y., and Goldman, R. D. (1985)J. Cell. Biol. 35. Morrow, J. S., Speicher, D. W., Knowles, W. J., Hsu, C. J., and 101,802-813 Marchesi, V. T. (1980)Proc. Natl. Acad. Sci. U. S. A. 77,65929.Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli, 6596 U.K. (1977)J. Biol. Chern. 252,1102-1106 36. Nelson, W. J., and Traub, P. (1982)J. Bwl. Chem. 257,553610.Garrels, J. I. (1979)J. Biol. Chem. 254,7961-7977 5543 37. Blose, S. H., and Meltzer, D. I. (1981)Exp. Cell Res. 135, 29911. Zernig, G., and Wiche, G. (1985)Eur. J. Cell Biol. 38,113-122 309 12. Marchesi, V. T. (1985)Annu. Rev. Cell Bwl. 1,531-561 13. Trotter, J. A., Nixon, C. S., and Johnson, M. A. (1985)J. Bwl. 38. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951)J. Bid. Chem. 193,265-275 Chem. 260,14374-14378
ing with plectin in uiuo, plectin may cross-link IFs with microtubules and connect these filament systems via spectrin to the plasma membrane according to the scheme shown in Fig. 9. Furthermore, plectin may cross-link IFs with microfilaments via MAP 2 (26) and/or spectrin (27). In conclusion, our data demonstrate that plectin hasa conserved primary structure and is ubiquitously distributed in cultured fibroblasts as well as othercell types. Its structural conservation, including strategically located phosphorylation sites, suggests thatthis protein is involved in important cellular functions. As a potential multifunctional cross-linking element of the cytoskeleton, plectin molecules are expected to possess several specific binding domains whose characterization is in progress.
SUPPLEMENT TO PLECTIN AND IFAP-300X ARE HOMOLOGOUS PROTEINS BINDING TO MICROTUBULE-ASSOCIATED PROTEINS
AND 2 AND TO THE 240 XDA
1
SUBUNIT OF SPECTRIN BY HARALD H E W N AND G W H A R D WICHE
EXPmIMENTAL
PROCEDURES
Materials - The following materials were purchased from companies g i v e n in parentheses: Radiochemicals IAmerahanl; IODO-GEN (Pierce); Sepharose CL-48, pzotein A-sepharoae CL-4B. Polybuffer Exchanger 94 and Polybuffer 7 4 (Pharnacial; chemicals for~ o l ~ a c r ~ l a n iqdeel s . Servalvtes. anrotinin andurea ISeTYa. Heidelberg); ~di..l~...yl,-...p..ti. and bisacril&ide for 2-d gels (BOH and Bio-Rad, respectively); gel electrophoreeie M standards: Myosin ( I I = 205 0001 -galactosidase ( M - 116.000) phosphory1:se b (M bovin; s e r k albumin ( M = 65.6DOi; PMSF, C ~ I o I a m i n eT, calmodufin=, ~ 7 i : o ~ ~ t ~ ~ ! gents and buffer 8UbstasCeS (Siva); BA 85 nitrocellulose paper (Schleicher 6 Schuelll; blotting device, Eco)no-col~mns and biogel P-10 (Bio-Rad); ProtoBlotTn Imnunoscreening System employing alkaline pho-phatase coupled to goat anti-rabbit IgG (Prmega Biotec, Madison, WI); Staphylococcue aureus Y8 protease In, = 2 1 . 0 0 0 ) IMiles Laboratories); a 1 1 Other nucleotides and enzymes = 44 0001 glyceraldehyde-3-phosphate including phoephoglycerate kinase ( M dehydrogenase I n = 36 000) DNase Ir(M ;1 O d O l and RNase A (I4 13,1001 IBoehringer Mann~ei~l;'sDs.'glycine a n 6 a 1 1 Lasic chemicals of highest purity available IMerck. Darmatadt); and cell culture reagents (Flow).
-
6,-
-
Antibodies Polyclonal antibodies to rat glioma C6 plectin were raised in rabbits u s i n g electrophoretically purified immunogen from IF preparations Iprocedure A, see belovl. Detaile o f the imuniration and antibody characterization have been g i v e n ( 3 ) .
-
Cell culture Tissue Culture c e l l s were from the American Type Culture Collection. Human epidermoidcarcinoma c e l l s A-431 were kindly provided byDr. W.W. Franke. C e l l a were cultured according to the instructions Of the suppliers. preparative quantities Of cells Yere groun at 37.C in glass roller bottlee fluehed with carbon dioxide.
-
Purification Of plectin Prior to column chromatography, IF preparationsrere Suspended in PBS supplemented with 0.6 M potassium iodide, 0.11 Triton x-100 and 1 mM PMSF, vigorously YDrteXed for5 min with intermittant cooling and pelleted ina n Eppendorf centrifuge for nin 10 at 4-C. This washing step was repeated once or twice. Washed preparationswere dissolved in 0.6 m l Of 9n urea, 2 mM DTT, 0 . 1 1 Triton X-100, 1 mM PUSF, 10 mM bistrispropane. pH 1.5 (column buffer) by homogenization at room temperature followed by centrifugation in a Beckman type 65 rotor at 40,000 rpol f o r 1 h without refrigeration. The supernatant was applied to a column 10.7 X 30 cm) at an elution rate Of one drop p e r 30 Bet. Elution was with column buffer atroom temoerature collectini fractions of drops. 25 Peak fractionswere assessed by fastsosPOlyacrylaalde gel runs I 3 0 mi") and quick stainingldestaining procedures. For radiolabeling or overlay experiments, fractions of plectin were trannfered into the desired buffer by chromatoqraphy on P-10 columns.
.
~~~~
Chromatofocusinq - Peak fractlonr of plectin from molecular s i e ~ i n gcolumns were transfered into 25 r n Tris'HC1, ~ pH 7.4, 1 mM EDTA, 1 m u DTT, 8M urea and loaded onto I 0.1 x 10 m column of Polybuffer Exchanger 94. Chronatofocksing vas performed by elution with Polybuffer 1 4 (diluted 1:8 in the above buffer at pH 4) as described ( 8 ) . A final elution was performed at p~ 2.4. sodium 3H- borohydride Radiolabelin of roteins - Labelingwith done a s descrqbed p32). Iodination of calnodul was performed with Ncl l Y t ? and I O W - G E N 132). Plectin was iodinated with tRI-BOltOn and Hunter reagent in 100m M Sodium borate, pH8.5, using 5 VCi perP 9 of protein. The reaction was Stopped by desalting into 100 on borate, pH 8.5. 3 situ phosphorylation was done l a described ( 2 ) . 3 radiolabeling was asdescribed (2) except that no depletion was performed.
Interaction ofMAPS Plectin with
and Spectrin
-
Polyacrylamide qel electrophoresis and PeeLtide maDcdnq 1-d SDS-polyacrylamlde g e l electrophoresis was done according to Laenmli ( 3 3 ) as previously described 12). In several exnerinents theDH of the runnin- buffer v a s adjusted to pH8.9. 2-6 gel elec'trophoresis wa; done essentiallj as described by ~ a r r e l s(10) with 2% ampholyte* pH 4 - 6 plus 2% anpholytes PA 3 10. 1- and 2-d peptide mapping was performed as previously described (9.29.32).
Ca2'/CaM
-
1325
--
+ +
f *.*? b
0
-
Protein blottinq and overlav experiments POI overlay experiments proteins of various samples were separated on 5% polyacrylamide gels and blotted nitroto c e l l u l o S e 120 h. 300 nA) in 2 5 n M borate. p H 8.8, 2 mM EDTA, 1 m M 2-mereaptoethanol, 20% methanol. The paperswere then blocked (Ih, 37'C) in PBS containing 5% defatted milk powder (191, 1 mM EGTA, 1 mM 2-mercaptoethanol, 0.005% NP-40 and 0.5 mM MgCl and briefly washed with PES. Overlays with samples of plectin were perfo%d for 3 h at 37-C with constant agitationin the Same Solution used for blocking but without milk powder. The concentration of plectin in the overlay solutions was between 0.05 and 1 vghl. Radiolabeled plectin was homogeneous a s judged from autoradiographes of electrophoresis gels. Incubations were stopped by washing 8 times with PBS ( 5 to 10 min eachl. When radiolabeled plectin samples were used, the nitrocellulose papers were blotted dry and autoradiographed. Unlabeled plectin vas detected by incubation with rabbit antibodies to plectin followed by secondary antibodies conjugated to alkaline phosphatase and color development (Proto-Blot I m u n o screening system) according to instructions provided by the manufacturers.
55
31
-
Other procedures Microtubule proteins from hog brain and brain membranes were prepared a s described (32, 3 4 and 171. DEAE-column purified hog brain fodrin was prepared essentially as described by Davies and Bennett (17); the identity of fodrin was confirmed by Ca2*-dependent binding of radiolabeled callnodvlin lunovblinhed data1.Soectrinwas oreoared bv extraction Of human red cell ghosts at 37-C in low ionic strength buffers ( 3 5 ) . Vimentin was purified from C6 IF preparations by molecular Sieve chromatography followedby chromatography on DEAE-Sepharose CL-16 136). Immunoprecipitation of SDS-boiled samples was done by the method ofBloae and Meltrer (371 a s described ( 2 ) . Protein was quantitated using the method Of Lowry al. et 138) with bovine serum albumin as standard. Ouantitation Of stained protein profiles was done in a Beckman DU 88 spectrophotometer.
. .
18
. .
14
"
2.5
z5 05 "
pg V8 Figure 7. 1-d peptide mapping of cytoskeletal plectin labeled ~ith[y-'~PlATP rn CHO cell monolayers were extracted With 0.15% Trlton X-100 as de-01 and cvtoskeletona remainlnoon the dish Were labeled with 10 uM [y"'P]ATP.in the-absence or presence &f 1 n#4 CaC1, and 5 pg/ml Of calmodvlin for 20 min on ice. Cytoskeletons were dissolved I" electrophoresis sample buffer and boiled. P l e c t m was isolated by inmunoprecipitation, electrophoresis Of the inmunocomplexeson 5% gels and excision ofbands. This materlal was then digested duringre-electrophoresis on 15% gels u s m g increasing amounts Of V8 DrOteaSe as indicated. Full lenoth oel is shorn. Arrowheads indicate phosphorylated fragments produced from plgctin labeled in the presence of Numbers. Mrx10-'. Ca"/calmodulin.
&.
Figure 5. Chromtofocusing Of C6 plectin. Plecrin purified by column chroall eluted fraomatography was subjected to chromatofocusing and aliquots of measurement; numbers, pH of t i m S were analyzed by gel electrophoresLs and pH fractions: 5 , sample loaded. Most Of the plectin eluted at pH 5.3 to 4.9. El"tion with Polybuffer74adjusted to pH 2.4 set free some more plectin (last 3 fraEtiOnS, falllng PHI. Lines indicate Mr Standards of 300,000, 240,000. 205,000, 116,000, 97,400, 66,000. 44.000 Ifrom top to bottom).
