Characterization of the actin polymerization-inhibiting protein from chicken gizzard smooth muscle

Transcription

1 Eur. J. Biochem. 170, (1988) 0 FEBS 1988 Characterization of the actin polymerization-inhibiting protein from chicken gizzard smooth muscle Klaus RUHNAU, Elke SCHROER and Albrecht WEGNER Institut fur Physiologische Chemie, Ruhr-Universitat Bochum (Received March 27/July 23, 1987) - EJB An actin polymerization-inhibiting protein, occurring in crude preparations of vinculin from chicken gizzard, has been found to be heterogeneous. The molecular masses of the polymerization-inhibiting peptides have been reported to range from 20 kda to 80 kda [Schroer, E. & Wegner, A (1985) Eur. J. Biochem. 153, , In this paper, a 21-kDa peptide was isolated from the bulk of the other peptides by gel chromatography. The 21-kDa peptide was identified as a polymerization-inhibiting peptide by its ability to retard nucleated actin polymerization and to bind polymeric actin when it was blotted onto nitrocellulose. Antiserum raised to the 21-kDa peptide was found to react with almost all peptides of the blotted heterogeneous polymerization-inhibiting protein. The same peptides which reacted with antiserum cosedimented with polymeric actin. The major peptides of the blotted polymerization-inhibiting protein bound polymeric actin. The largest peptide which reacted with antiserum and cosedimented with polymeric actin had a molecular mass of 85 kda. The results suggest that the preparation of polymerization-inhibiting protein contains mainly polymerization-inhibiting peptides and only some contaminants, and that all the polymerization-inhibiting peptides are proteolytic fragments stemming from a common precursor. In smooth muscle of chicken gizzard an actin polymerization-inhibiting protein has been reported to occur which inhibits polymerization of actin by binding to the barbed ends of actin filaments [l The polymerizationinhibiting protein has the unusual property that it does not nucleate actin polymerization, i. e. actin monomers do not polymerize onto the polymerization-inhibiting protein. The polymerization-inhibiting protein copurifies with vinculin and initially vinculin was thought to be the polymerization-inhibiting protein [4]. Later it was demonstrated that the polymerization inhibition activity can be separated from vinculin by chromatography on carboxymethyl-cellulose or hydroxyapatite [l - 3, 51. Furthermore the polymerization inhibition activity, which is eluted from a DEAE-cellulose column, does not coincide with vinculin [2]. Purification of the polymerization-inhibiting protein is difficult because of the heterogeneity of this protein. SDS/ polyacrylamide gels of preparations of the polymerizationinhibiting proteins reveal numerous peptide bands. Polymerization inhibition activity can be eluted from SDS/polyacrylamide gels in the molecular mass range kda [2]. It is not known which bands of SDS/polyacrylamide gels represent polymerization-inhibiting peptides or contaminants. In this paper we purified a 21-kDa peptide polymerization-inhibiting peptide and we worked out methods for identifying the peptides observed on SDS/polyacrylamide gels as polymerization-inhibiting peptides. Currespundence tu A. Wegner, Institut fur Physiologische Chemie, Postfach , D-4630 Bochum, Federal Republic of Germany Abbreviation. NBD-actin, actin modified with 4-chloro-7-nitro-2- oxa-i,3-diazole. MATERIALS AND METHODS Purification of the polymerization-inhibiting protein The polymerization-inhibiting protein can be separated from its major contaminant vinculin by chromatography on carboxymethyl-cellulose [l, 21. As during this chromatography at ph 5 the polymerization-inhibiting-protein loses activity, we have worked out and describe here a purification procedure in which the carboxymethyl-cellulose chromatography is circumvented. Muscle (300 g) from chicken gizzards was homogenized and extracted according to the method of Feramisco and Burridge [6]. After impurities had been removed by 10 mm MgC12 precipitation, the polymerizationinhibiting protein was collected by (NH&S04 precipitation (20.5 g/100 ml). The precipitate was dissolved and dialyzed overnight against buffer A (5 mm NaCl, 0.25 mm EDTA, 10 mm mercaptoethanol, 5 mm Tris/acetate ph 7.6). The solution was centrifuged for 1 h at x g and applied to a DEAE-cellulose column equilibrated with buffer A. The proteins were eluted by a linear NaCl gradient. The occurrences of the polymerization-inhibiting protein was tested by the polymerization-inhibition assay. Polymerization-inhibiting protein free of vinculin was obtained by pooling the fractions of the first peak which passed through the column without retardation, and the fractions which were eluted before the vinculin peak. In order to concentrate the solution, the polymerization-inhibiting protein was dialyzed against buffer A (two changes), collected on a DEAE-Sepharose CL-6B column (1.5 x 15 cm) and eluted with 200 mm NaCl in buffer A. Purification of a 21-kDa polymerization-inhibiting peptide The polymerization-inhibiting protein was applied to a Sepharose CL-6B column (1.5 x 103 cm) equilibrated with

2 584 buffer B (5 mm triethanolamine/hcl ph 7.5, 200 mm NaC1, 200 mg/l NaN3). A 21-kDa peptide was eluted at a volume approximately equal to the bed volume. The 21-kDa peptide could be further purified from minor contaminants by filtration through a PM30 membrane (Amicon). The 21-kDa peptide passed the membrane while higher-molecular-mass contaminants were retained by the membrane. Protein concentrations were determined by the method of Lowry [7]. Preparation of actin Actin was prepared according to the procedure or Rees and Young [8]. Part of the protein was modified with N-ethylmaleimide at cysteine-374 and subsequently with 4-chloro-7-nitro-2-oxa-1,3-diazole at lysine-373 to produce a fluorescently labeled actin (NBD-actin) [9]. The protein was applied to a Sephacryl column (2.5 X 90 cm) equilibrated with buffer (0.5 mm ATp, o.2 mm CaC129 2oo mg/l NaN,, 5 mm triethanolamine/hcl PH 7.5). The concentration of actin was determined photometrically at 290 nm by using an absorption coefficient of M- cm- [lo]. Preparation of antisera to the 21-kDa peptide The 21 -kda peptide, purified by filtration through a membrane, was homogenized in Freund s complete adjuvant and injected subcutaneously into a rabbit at multiple sites (final volume: 50 pg peptide in 0.25 ml plus 0.25 ml Freund s adjuvant). After two initial injections the rabbit was boosted in one-week intervals, each in Freund s complete adjuvant. Every second week the animal was bled from the marginal ear veins. Antibodies were detectable by immunoblot analysis at an antiserum dilution of 1 : 200. Electrophoresis, immunoblotting and staining with polymeric NBD-actin SDS/polyacrylamide gel electrophoresis was performed according to the method of Weber and Osborn [Ill or Laemmli [12]. The gels were stained with Coomassie blue. When immunoreactivity was analyzed or NBD-actin was used for staining, the protein bands were blotted electrophoretically onto nitrocellulose sheets (BA 85, Schleicher & Schull, Dassel, FRG) [13]. The nitrocellulose sheets were stained with NBD-actin by incubation for 1 h with a staining solution containing 20 pm actin (50% NBD-actin, 50% unlabeled actin polymerized by addition of 2 mm MgC12) and 3% bovine serum albumin at room temperature under slow shaking. The nitrocellulose sheets were washed twice for 10 min with 20 mm NaCI, 0.1 % Tween-20 and 5 mm triethanolamine/ HC1 ph 7.5. The location of the protein bands stained with polymeric actin was visualized by irradiation at 284 nm. Photographs were taken using a Polaroid 4 x 5 Land Film T55. Immunoreactivity was analyzed by reaction of the blotted protein bands with rabbit antiserum (diluted 1 : 200) and subsequent incubation with protein A conjugated with horseradish peroxidase (Sigma). The immunoblots were developed with chloronaphthol and Hz02 [14]. Polymerization inhibition assay The ocurrence of the polymerization-inhibiting protein was assayed by retardation of nucleated polymerization. Polymerization-inhibiting protein, 2 pm monomeric and Fig. 1, SDS/po/yacry/amide gel electrophoresis. Lane 1, molecular mass standard proteins: phosphorylase b, 94 kda; bovine serum albumin, 67 kda; ovalbumin, 43 kda; carbonic anhydrase, 30 kda; soybean trypsin inhibitor, 20.1 kda. Lane 2, polymerization-inhibiting protein (50 pg). Lane 3, polymerization-inhibiting protein cosedimented with actin filaments. Lane 4, polymerization-inhibiting protein of the supernatant 0.5 pm polymeric actin were mixed with 2 mm MgClz and buffer C at 37 C. Polymerization was measured by the fold greater fluorescence intensity of polymeric actin compared to monomeric actin. The excitation wavelength was 436 nm, and the fluorescence intensity was measured at 530 nm. The activity of the polymerization-inhibiting protein was quantified by the polymerization retardation factor: r = (sa-si)/xa, where s, is the initial rate of increase of the fluorescence intensity in the absence of polymerization-inhibiting protein and si is the initial rate of increase of the fluorescence intensity in the presence of polymerization-inhibiting protein [2]. RESULTS AND DISCUSSION The 21-kDa peptide An SDS/polyacrylamide gel of a preparation of the polymerization-inhibiting protein is depicted in Fig. 1. The gel reveals numerous peptide bands. By fractionation on a Sepharose CL-6B column, a 21-kDa peptide together with small amounts of a 85-kDa peptide could be separated from the bulk of other peptides. SDS/polyacrylamide electrophoresis of individual fractions of the gel chromatography are depicted in Fig. 2. Four major protein bands with molecular masses of 85,70,42 and 21 kda were observed in the fractions. The fractions which were eluted at a volume nearly equal to the bed volume contained the 21-kDa peptide together with small amounts of the 85-kDa peptide. The polymerization inhibition activity of the fractions was tested. Activity was detected in the fractions containing the 21-kDa peptide, but also in the fractions containing the 85-kDa, 70-kDa and 42-kDa peptides (Fig. 2A). For preparation of antisera the 21-kDa peptide was purified from the 85-kDa peptide by membrane filtration. In Fig. 3 an SDS/polyacrylamide gel of the highly purified 21-kDa peptide is depicted. The 21-kDa peptide was identified as an actin-binding protein by staining of nitrocellulose sheets onto which the peptide was blotted electrophoretically. The 21-kDa peptide

3 A Elution volume [mll Fig. 2. (A) Elution prqfile of gel filtration on Sepharose CLdB (1.5 x 103 crn) and (B) SDSlpolyacrylamide gel electrophoresis ojjractions eluted. (0) Absorbance at 280 nm (Az8,,);(0) activity r of 0.6 ml polymerization-inhibiting protein added to a total volume of 2.4 ml. (B) The elution volume (ml) of the fractions is given at the bottom of the lanes. Molecular mass standard proteins: 94, 67, 43, 30, 20.1 and 14.4 kda could be stained using polymeric NBD-actin (Fig. 3). The specificity of staining was tested by control proteins. The molecular mass standard proteins (see Fig. 1) were not stained by polymeric NBD-actin. Other actin-binding proteins such as rabbit skeletal muscle tropomyosin, chicken gizzard a-actinin or human platelet gelsolin were not stained by polymeric NBD-actin, suggesting that the 21-kDa peptide is not a proteolytic fragment of these proteins. Also actin blotted onto nitrocellulose sheets could not be stained by this procedure. The 21-kDa peptide was stained only by polymeric NBD-actin but not by monomeric NBD-actin. The effect of the 21-kDa peptide on the elongation of actin filaments was investigated by combining various concentrations of 21-kDa peptide purified by Sepharose CLdB gel chromatography with 2 pm monomeric and 0.5 pm polymeric actin in the presence of 2 mm MgClz at 37 C. The 21 -kda peptide retarded the elongation significantly at substoichiometric ratios of 21-kDa peptide to monomeric actin (Fig. 4). This demonstrates that the 21-kDa peptide does not act as a profilin-like protein, i.e. it does not bind to actin monomers to inhibit polymerization. The substoichiometric ratio sufficient for retardation of polymerization suggests that the 21-kDa peptide binds to the ends of elongating actin filaments and inhibits association of monomers and dis- sociation of filament subunits. Also the staining experiments of nitrocellulose-blotted 21 -kda peptide suggest that only polymeric actin but not monomeric actin binds to the 21-kDa peptide. Characterization of the higher-molecular-muss polymerization-inhibiting peptides Numerous bands are detected on Coomassie-blue-stained SDS/polyacrylamide gels of preparations of the polymerization-inhibiting-protein (Fig. 1). It is not known which of these numerous peptides represents polymerization-inhibiting protein and which bands are contaminants. We, therefore, investigated this open question by using antiserum to the 21-kDa peptide, NBD-actin stain and cosedimentation of the polymerization-inhibiting protein with actin filaments. In Fig. 5 an immunoblot of the polymerization-inhibiting protein is depicted. The gel was stained with antiserum raised to the 21-kDa peptide. Comparison of the immunoblot with the Coomassie-blue-stained gel (Fig. 1) shows that most bands react both with Coomassie blue and with antibodies, suggesting that the majority of the peptide of this preparation contain epitopes common to the 21-kDa peptide. The Coomassie-blue-stained gel reveals a few additional bands,

4 586 Fig. 3. Electrophoresis of the 21-kDapolymerization-inhibiting peptide. Lane 1, molecular mass standard proteins: 94, 67, 43, 30 and 20.1 kda. Lane 2, polymerization-inhibiting protein. Lane 3, SDS/ polyacrylamide gel of the 21-kDa peptide purified by membrane filtration. Lane 4,21-kDa peptide blotted onto nitrocellulose, stained with polymeric NBD-actin and visualized by fluorescence Fig. 5. Immunoblot of the polymerization-inhibiting protein (50 pg) (lane 2). Molecular mass standard proteins (lane 1): 94, 67, 43, 30, 20.1 kda *Cf *-*. : e. g. a 95-kDa peptide (Fig. 1). The polymerization-inhibiting protein may contain a few minor contaminant proteins. We assayed binding of polymeric actin to the individual peptide bands. Nitrocellulose blots of SDS/polyacrylamide gels were incubated with NBD-actin polymerized by 2 mm MgC12. In Fig. 6 the fluorescence of bound NBD-actin is displayed. Major bands were observed corresponding to molecular masses of 85, 70, 42 and 21 kda. Diffuse staining was detected between kda (Fig. 6). Polymeric actin binds to the major bands of the preparation of the polymerization-inhibiting protein. Staining of the numerous minor bands was not observed probably because of the low sensitivity of the method. The NBD-actin stain indicates that at least the major bands of the preparation of the polymerization-inhibiting protein represents actin-binding protein. As mentioned in the preceding section, the NBDactin stain is specific. Even actin and some tested actin-binding proteins can not be stained by this procedure. Polymerization-inhibiting protein (1 ml, 1 mg/ml) was mixed with actin (1 ml, 2 mglml). Polymerization was initiated by 2 mm MgC12. After a 30-min incubation at 37T, the mixture of polymerization-inhibiting protein and polymeric actin was centrifuged for 1 h at x g. The sedimented pellet and the supernatant were applied to SDS/polyacrylamide gels (Fig. 1). As estimated from the intensity of the Coomassie blue stain about 25% of the polymerization-inhibiting protein was cosedimented with actin and 75% remained in the supernatant. The volume of the pellet was 2 pl, while the volume of the supernatant was 2000 pl. Thus, the polymerization-inhibiting protein was concentrated in the pellet 300-fold. Considerable amounts of the polymerizationinhibiting protein bound to actin filaments. In a control experiment phosphorylase 6, bovine serum albumin, ovalbumin, carbonic anhydrase, soybean trypsin inhibitor and a-lactalbumin were shown not to be cosedimented with actin filaments (data not shown). The cosedimented peptides and the

5 587 antiserum-stained peptides revealed striking similarities. The 95-kDa peptide occurring in the polymerization-inhibiting protein preparation and in the supernatant (Fig. 1) was missing both in the cosedimented peptides and in the antiserumstained peptides (Figs 1 and 5). This findig supports the conclusion that the 95-kDa peptide is a contaminant. No protein above 85 kda or below 21 kda cosedimented with actin filaments, in good agreement with the previously reported result that only in the kDa molecular mass range is polymerization inhibition activity eluted from SDS/polyacrylamide gels [2]. Below 21 kda, two bands which did not cosediment were stained by antiserum. Perhaps these two peptides are small inactive proteolytic fragments of the polymerizationinhibiting protein. In this paper we provided evidence that a 85-kDa peptide of our preparation is the largest peptide which inhibits actin polymerization. The reaction with antibodies suggests that the numerous smaller peptides which are observed on immunoblots, have homologous amino acid sequences. The peptides are likely to be proteolytic fragments of a common precursor which may be the 85-kDa peptide. Wilkins et al. [5] have prepared antibodies to polymerization-inhibiting peptides. This preparation comprises peptides in the kda molecular mass range. The antibodies have been found to cross-react with 200-kDa and 150-kDa polypeptides of cardiac muscle tissue. Wilkins et al. [5] have proposed that these 200-kDa and 150-kDa proteins are the precursors of the polymerization-inhibiting peptides. We could not find peptides above 85 kda which react with antiserum to the 21 -kda peptide although on Coomassie-blue-stained gels a number of higher-molecular-mass proteins can be detected. Direct comparison of our results with the results of Wilkins et al. [5] is difficult because the antisera were raised to different peptides. Clearly, more biochemical work on single pure peptides is necessary before final conclusions about the precursor of the polymerization-inhibiting peptides can be drawn. We thank Mrs E. Werres for excellent technical assistance. This study was supported by the Deutsche Fors~hungsgemein.~chuft (SFB 168). REFERENCES 1. Evans, R. R., Robson, R. M. & Stromer, M. H. (1984) J. Biol. Chem. 259, Schroer, E. & Wegner, A. (1985) Eur. J. Biochem. 153, Wilkins, J. A. & Lin, S. (1986) J. Cell. Biol. 102, Wilkins, J. A. & Lin, S. (1982) Cell 28, Wilkins, J. A,, Risinger, M. A. & Lin, S. (1986) J. Cell Biol. 103, Feramisco, J. R. & Burridge, K. (1980) J. Biol. Chem. 255, Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. (1951) J. Biol. Chem. 193, Rees, M. K. &Young, M. (1967) J. Biol. Chern. 242, Detmers, P., Weber, A., Elzinga, M. & Stephens, R. E. (1981) J. Biol. Chem. 256, Wegner, A. (1976) J. Mol. Biol. 108, Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, Laemmli, U. K. (1970) Nature (Lond.) 227, Towbin, H., Staehelin, T. & Gordon, J. (1970) Proc. Nut1 Acad. Sci. USA 76, Hawkes, R., Niday, E. & Gordon, J. (1982) Anal. Biochem. 119,