286

Biochimica et Biophysica Acta, 1038 (1990) 286-290

Elsevier BBAPRO 33627

Characterization of monoclonal antibodies to human platelet myosin that recognize highly conserved epitopes within the 50 kDa fragment of myosin subfragment-1 Varda R. Deutsch 1,2, Shakker Biadsi 1, A m i r a m Eldor 2, Andras Muhlrad 3 and Itzhak K a h a n e 1 J Department of Membrane and Ultrastructure Research, The Hebrew University-Hadassah Medical School, 2 Department of Hematology, Hadassah Unioersity Hospital and 3 Department of Oral Biology, Hebrew Unioersity-Hadassah School of Dental Medicine, Jerusalem (lsrael)

(Received 21 November1989)

Key words: Myosin;Myosinsubfragment; Platelet; Monoclonalantibody

Three monoclonal antibodies directed against human platelet myosin heavy chains (MCH) that recognize homologous sequences contained within the functionally active subfragment-1, in platelet and rabbit skeletal muscle myosin were studied. These antibodies are distinguished by their affinities to different myosins and their differential effect on various ATPase activities. Epitope mapping was accomplished by analyzing antibody binding to proteolytic peptides of myosin head subfragment-I under various experimental conditions. The epitopes recognized by these anti-human platelet MHC monoclonal antibodies reside within a small region of the 50 kDa fragment, beginning 9 kDa from its C-terminus and extending a stretch of 6 kDa towards the N-terminus. These epitopes lie between residues 535-586, and are contained within a highly conserved area of myosin heavy chain.

Introduction Myosin is an ubiquitous mechanochemical protein responsible for biological motility by the coupling of actin and hydrolysis of ATP [1]. Although myosin is found in many cell types it is most abundant in muscle cells and very prevalent in blood platelets where it plays a major role in motile and contractile functions [2,3]. The distinct ATP and actin binding sites reside within a 95 kDa N-terminal head segment designated S-1 (subfragment-l). The amino acid sequence of the various myosin heavy chains in muscle is highly conserved, especially in the head portion of the molecule [1], yet isoforms can be distinguished by their biochemical activity and antigenic properties [4-8]. We have previously reported the presence of two such isoforms in blood platelets which are found in the cytoplasmic or membrane fractions of fresh human

Abbreviations: MHC, myosinheavychain; S-l, subfragment-1;MAb, monoclonal antibody; TBS, Tris-buffered saline; BSA, bovine serum albumin; DTNB, 5,5'-dithio-bis(2-nitrobenzoicacid). Correspondence: V. Deutsch, Department of Hematology,Hadassah University Hospital, Jerusalem 91120, Israel.

platelets [3,9]. Platelet myosin (MHC) heavy chain isoforms can be distinguished by differences in their twodimensional peptide structure [3], by their enzymatic activity and sensitivity to chemical modification, as well as by their reactivity to polyclonal antibodies [9,10]. The difference between cytoplasmic and membrane platelet myosin is likely to reside within the head region, since the two-dimensional peptide maps of the rods are identical and the ATPase activities, actin binding and response to chemical modification of the head of these two myosins differ [9]. In order to further study the structure of the head segment region of platelet MHC we raised monoclonal antibodies to platelet myosin heavy chain and chose to study those that recognized specific epitopes within this region.

Material and Methods Immunization of mice with myosin, which was purified as previously described [3] from the membrane fraction of human platelets, consistently produced antimyosin antibody titers higher than those obtained in mice injected with cytoplasmic myosins. Therefore, monoclonal antibodies were raised, using the method of Kohler and Millstein [11], to myosin that was purified

0167-4838/90/$03.50 © 1990 Elsevier SciencePublishers B.V. (BiomedicalDivision)

287 from platelet membranes. The fusion of immune murine spleenocytes with myeloma Ag8X653 gave rise to 38 clones that produced antibodies against platelet myosin heavy chain, none of which displayed differential recognition of myosin isoforms. Antibodies produced by three clones, designated as A12, H1G12 and F l l G l l cross-reacted with the head region of rabbit skeletal muscle myosin, indicating that they were directed against highly conserved epitopes within the head region. The hybridomas were recloned twice at limiting dilutions to ensure homogeneity of the antibodies produced. These clones were subsequently expanded as ascites in mice, the antibodies produced were purified by protein A Sepharose affinity chromatography. All three antibodies belong to the IgG1 subclass as determined by ELISA, using subclass specific antibodies.

Results and Discussion Indication that these antibodies are directed at different sites came from analysis of their relative affinities to the different myosins. Dissociation constants (Kd) of affinity purified antibodies were calculated from experiments in which direct binding to platelet m e m b r a n e myosin at decreasing antibody concentrations was assessed (Table I). ATPase activity is known to reside in the S-1 segment of the myosin heavy chain. We, therefore, tested the effect of these antibodies on Ca 2÷, Mg 2÷, K ÷ E D T A and actin activated ATPase activities of S-1. ATPase activities ( # m o l Pi per mg myosin per min) were calculated from the inorganic phosphate (Pi) liberated by the Fiske Subarrow method and modified by Muhlrad and

TABLE I

Affinities of monoclonal antibodies A12, H1G12 and F11G11 to human platelet membrane myosin Increasing concentrations of purified antibodies A12, H1G12 and F l l G l l were reacted at various dilutions in ELISA assays with immobilized platelet membrane myosin as follows. Dynatech Immulon microtiter plates were coated with 50/~1 of soluble myosin at a concentration of 50/~g/ml in TBS (25 mM Tris in saline, buffered pH 7.0) overnight at 4°C. Plates were washed with TBS, blocked with 0.5% BSA, and incubated with antibodies at decreasing concentrations for 2 h at room temperature. A horseradish peroxidase conjugated anti-mouse IgG (Amersham) and chromogenic substrate ABTS (2,2'-azinobis(3-ethylenebenzthizolinesulfonic acid) substrate were used to detect antibody binding. Reactions were terminated by the addition of 1% SDS, and the amount of antibody bound was assessed by absorbance at 410 nm. Dissociation constants were calculated from the plot of 1/A410 vs. 1/[antibody concentration]. Antibody

Dissociation constant

A12

4 -10 -6

H1G12

0.5.10 -6

FllGll

1.3.10 -7

2O

o

-1

0

1

2

3

4

5

6

7

8

9

10

1/[A12] jug/ml Fig. 1. Inhibition of A12 binding in the presence of ATP. The binding of A12 antibody to platelet myosin was measured by ELISA in the presence (0) or absence (n) of 10 mM ATP and decreasing antibody concentrations. ELISA assays were performed as follows; Dynatech Immulon microtiter plates were coated with 50 pl of soluble myosin at a concentration of 50 ~tg/ml in TBS (25 mM Tris in saline, buffered pH 7.0) overnight at 4 ° C. Plates were washed with TBS, blocked with 0.5% BSA, and incubated with antibodies at various concentrations for at least 2 h at room temperature. A horseradish peroxidase conjugated anti-mouse IgG (Amersham) and chromogenic substrate ABTS (2,2'-azinobis(3-ethylenebenzthizolinesulfonic acid) substrate were used to detect antibody binding. Reactions were terminated by the addition of 1% SDS, and the amount of antibody bound was assessed by absorbance at 410 nm. The effect of ATP on A12 binding is represented as a Lineweaver-Burk plot, bars indicate standard t deviation.

Morales [12]. Antibodies F l l G l l and H1G12 did not significantly effect any of the ATPase activities. Monoclonal antibody A12 had no effect on Ca 2+, Mg 2+ or K + E D T A ATPase activity but consistently inhibited actin activated ATPase by 30-40%. The addition of 10 m M A T P to ELISA assays inhibited the binding of antibody A12 in a noncompetitive manner (Fig. 1), whereas the presence of A T P had no effect on the binding of antibodies F l l G l l or H1G12. D a n - G o o r et al. [13] recently reported that the MAb25 anti-50 k D a monoclonal antibody, whose epitope resides near the N-terminus of the 50 k D a domain, inhibits not only the actin activated but also the other two ATPase activities of myosin. This can be interpreted to mean that while MAb25 directly effects the binding or hydrolysis of ATP, A12 may effect actin activated ATPase activity indirectly by interfering with the binding of S-1 to actin. This notion is supported by the finding that the presence of A T P causes diminished binding of A12 to myosin. It is well known that A T P decreases the affinity between myosin and actin. If the A12 antibody binds in close proximity to the actin binding site of S-l, it is not unexpected that its affinity to S-1 diminishes as well (Fig. 1). In order to relate the functional effects of the antibodies to myosin structure, it was imperative to first

288 TABLE II

Competitive binding of soluble myosin tryptic fragments to monoclonal antibodies

50-45~

27--

a

b

c

d

Fig. 2. The reaction of anti-platelet myosin monoclonal antibodies with S-1 tryptic peptides produced in the presence of MgATP. Purified S-1 was digested with trypsin (1 : 100), 2 mM MgCI 2 and 2 mM ATP for 20 rain at 25 ° C. This treatment prOduces an additional peptide of 45 kDa which is a derivative of the 50 kDa [15]. The reaction was terminated by boiling the protein in sample buffer for 3 min. The digested S-1 peptides were separated on 7-17% SDS-PAGE gels electrophoretically transferred to nitrocellulose by Western blot and reacted with the antibodies. Lanes containing the separated peptides were cut down the center to enable accurate analysis of antibody specificity. Visualization of bound antibody was done with alkaline phosphatase conjugated second antibody and BCIP/NBT substrate (Protoblot, Promega). The left half of each lane represents the reaction of tryptic peptides with monoclonal antibodies. MAb25 (specific for 50 kDa N-terminus) (a), F l l G l l (b), A12 (c) and H1G12 (d). The right half of each lane was stained with Amido black to visualize total peptides transferred.

locate their epitopes in the primary structure. The location of the epitopes within the S-1 head subfragment was determined by immunoblots on peptides produced by limited trypsinization of S-1. Short treatment with trypsin produces a C-terminal 20 fragment and Nterminal 75 kDa peptide which subsequently breaks down into an N-terminal 27 kDa and a central 50 kDa fragment [14]. In immunoblots all three antibodies reacted with the 75 kDa fragment of the S-1 heavy chain and its 50 kDa derivative. Binding to the 50 kDa peptide is demonstrated in Fig. 2. In competitive ELISA assays soluble 20, 27 and 50 kDa peptides, and whole S-1 were employed as competing antigens for the binding of antibody to immobilized myosin heavy chain (from platelets) or myosin head (rabbit skeletal muscle myosin S-l). The 50 kDa fragment proved to be as potent an inhibitor of antibody binding as the whole S-1 molecule (Table II). The N-terminal 27 kDa fragment of S-l, the C-terminal 20 kDa fragment of S-l, actin and myosin light chain 1 displayed no competition for antibody binding. Our strategy for locating the specific antibody binding epitopes, within the 50 kDa peptide, has been based on producing a series of S-1 peptides by proteolytic cleavage of S-1 under various conditions. These peptides are illustrated as a schematic representation of the

Competitive binding assays were carried out in microtiter plates coated with platelet C-myosin [3] or rabbit skeletal muscle S-1 [22]. Affinity purified antibodies A12, HIG12 or F l l G l l were incubated in the presence of 50 # g / m l soluble S-1 or its tryptic peptides. The isolation of skeletal muscle myosin was done according to the method of Tonomura et al. [21]. S-1 was prepared by the digestion of myosin filaments with a-chymotrypsin [22], and was purified by column chromatography on DE-52 cellulose or Sephadex G-100. S-1 tryptic fragments of 27, 50 and 20 kDa were prepared by digesting rabbit skeletal muscle myosin S-1 with trypsin 1 : 100 (w/w) ratio at 25 o C for 30 rain [14]. These fragments were separated from each other by 6 M GdnHC1 and by ethanol precipitation [12]. Final purification was accomplished by gel filtration chromatography on Sephadex G-100 [12]. The inhibition of antibody binding caused by competition was calculated from the relative absorbance (A) at 410 nm. The data is presented as % inhibition of binding + standard deviation. Immobilized antigen:Platelet myosin heavy chain Competing antigen:S-1 Antibodies: A12 FllGll H1G12

58.1+8.75 72.1+4.17 63.0+0.97

50 kDa

27 kDa

20 kDa

53.6+8.15 68.8+4.55 79.1+3.27

0 0 0

0 0 0

Immobilized antigen:S-1 fragment Competing antigen:S-1 Antibodies: A12 FllGll HIG12

69.0+4.92 78.0+1.68 67.2+3.17

50 kDa

27 kDa

20 kDa

59.5+3.53 75.0+6.18 81.1+1.35

6.55:1.3 0 0

0 0 0

tryptic digestions in Fig. 3. The binding of our antibodies to these peptides was then compared with the binding of another anti-50 kDa antibody (MAb25) whose exact epitope is known to reside near the N-terminus of the 50 kDa [13]. Treatment of S-1 with trypsin in the presence of Mg/ATP produces five major peptides (50, 45, 27, 22 and 20 kDa). The 45 kDa peptide is a 50 kDa derivative which has lost its 5 kDa C-terminal region [15]. There are als0 a number of N-terminal peptide derivatives of

N I

I 27K I I 22K I I 22K I

95K*~

Trypsin, MgATP I 50K*~ I 20KI 45K*~ I I 20KI 35K ~

I

J C

I

Thrombin I 68K*~

I

30K

I

I 20KI

Location of the anti platelet myosin epitopes: I

I°1

I

50K Fig. 3. Schematic representation of myosin S-1 heavy chain and the sites of tryptic cleavage under different conditions. Peptides recognized by anti-platelet myosin antibodies (*) and peptides recognized by anti 50 kDa MAb25 ( - ).

289

B A

95-68--

--5O 30-LCl--

95-68--

--95 --68

30--

--30 --LC1

--27 20 -- LC3 -

Sl

5

I0 2 0

T-Sl

a

b

c

d

e

f

Fig. 4. Reaction of anti-platelet myosin antibodies with S-1 peptides produced by thrombin digestion. S-1 was digested with thrombin at 25 o C for 5-20 min followingtreatment with DTNB. The digests were electrophoresedon a 10-18% SDS-PAGEgel, transferred to nitrocelluloseby Western blot and reacted with antibodies. (A) SDS-PAGELanes: S-1 represents, nondigestedS-l; 5, 10, 20, represent time in minutes that S-1 was digested by thrombin; T-S-1 represents trypsin digested S-1. Vertical numbers represent molecular mass in kilodalton; LCl and LC3,light chains. (B) S-1 digested with thrombin for 5 rain, transferred to nitrocelluloseand stained by antibodies. Lane a, anti-S-l-N-terminus; b, anti-20 kDa; c, FllGll; d, H1G12; e, A12; f, MAb25 anti-50 kDa antibodies; and g, Amido black stain to visualizetotal peptides transferred.

the 50 kDa peptide which can be visualized with an anti-50 kDa N-terminal monoclonal antibody MAb25 [13] (Fig. 2). Antibodies A12, F l l G l l and H1G12 all recognize both the 50 and 45 kDa and are, therefore, not reactant with the extreme C-terminus of 50 kDa. These antibodies, however, do not recognize the second largest N-terminus peptide (35 kDa), and must therefore be directed at epitopes which reside within a 10 kDa region between the 35-45 kDa stretch from the N-terminus of the 50 kDa peptide. Indeed a number of unidentified smaller peptides, which do not contain the N-terminus epitope recognized by MAb25, are detected by A12, H1G12 and F l l G l l . The electrophoretic mobilities of these smaller tryptic peptides seems to be different with the various antibodies (Fig. 2), indicating that their respective epitopes reside on different sites within the 10 kDa stretch. Additional evidence that the epitopes of these antibodies do not lie proximal to the N-terminal of 50 kDa, is that they do not recognize the 29 kDa N-terminal peptide that is produced by trypsinolysis following heat treatment of S-1 (results not shown) [16]. Finer resolution of the region on the primary structure where the epitopes reside was obtained by cleaving the S-1 heavy chain with thrombin. It was shown by Chaussepied et al. [17] that following modification of SH groups within the S-1 heavy chain by DTNB, thrombin cleaves S-1 heavy chain at the carboxyl group of Lys-586 (residue numbering taken from Warrick and Spudich [1]) producing 68 kDa N-terminal and 30 kDa C-terminal peptides. The cleavage site is located 41 kDa from the N-terminus of the 50 kDa fragment. All three

antiplatelet myosin antibodies were found to recognize the 68 kDa N-terminal fragment (Fig. 4), which is also recognized by MAb25 (anti-50 kDa N-terminus) and by another antibody which is directed against the Nterminus of S-1 [18]. These results, together with the finding that the antibodies do not react with the 35 kDa N-terminal peptide, indicate that the epitopes should reside within a 6 kDa fragment between 35 and 41 kDa from the terminus of the 50 kDa fragment. The sequence of this 6 kDa region begins at about Ile-535 and ends at Lys-586 of the rabbit skeletal muscle myosin heavy chain, as published by Warrick and Spudich [1]. When comparing the amino acid sequences of various myosin heavy chains obtained from sources of great evolutionary distances (Acantamoeba, yeast, nematode, chicken and rabbit) to that of rabbit skeletal muscle myosin, it is clear that several highly conserved homologous sequences are contained within this 6 kDa region. The existence of homologous sequences could explain the cross-reaction between monoclonal antibodies that were raised against human platelet myosin and rabbit skeletal muscle myosin. Analysis of antigenic sites within the sequence of this 6 kDa peptide (according to Jameson and Wolf, Ref. 19) showed high levels of antigenicity in the C-terminal region between residues 562-586. This highly antigenic stretch lies near residues 592-599, a region which was shown to participate in the forming of actin-binding site of myosin [20]. Our finding that A12 antibody inhibits the actin-activated ATPase activity without affecting the K + ( E D T A ) and Ca 2+ activated ATPase seems to suggest that the antibody interferes with the binding of

290 actin to myosin. This interference can easily be explained by steric hindrance caused by the attachment of the bulky antibody to its epitope which is proximate to the actin binding site. Based on the combined results of binding affinities, competitive E L I S A assays and i m m u n o b l o t m a p p i n g we conclude that these antibodies recognize different epitopes which reside within a 6 k D a region of the myosin molecule located near the C-terminus of the 50 k D a tryptic fragment, but at least 9 k D a away from it. The identification of the low molecular mass peptides to which these antibodies bind, and the exact location of each of their epitopes, is currently under investigation. At the present time it is difficult to speculate the role of these highly conserved sequences. We hope that these antibodies will be used as high resolution tools for studying the structure of these sequences within the myosin head and in comparing different types of myosin isoforms. Acknowledgements This work was supported by a Basic Research F o u n d a t i o n grant of the Israel A c a d e m y of Sciences and Humanities, and post-doctoral fellowships awarded by the L a d y Fellowship Trust and the Israel Cancer Research F u n d to V.R.D. The authors are indebted to Dr. K. Sutoh for the generous gift of anti-S-l-Nterminus antibody. References 1 Warrick, H.M. and Spudich, J.A. (1987) Annu. Rev. Cell. Biol. 3, 379-421.

2 Pollard, T.D., Thomas, S.M. and Niederman, R. (1974) Anal. Biochem. 60, 258-266. 3 Peleg, I., Kahane, I., Eldor, A., Groschel-Stewart, U., Mestan, J. and Muhlrad, A. (1983) J. Biol. Chem. 258, 9290-9295. 4 Peltz, G., Spudich, J.A. and Parham, P. (1985) J. Cell Biol. 100, 1016-1023. 5 Flicker, P.F., Peltz, G., Sheetz, M.P., Parham, P. and Spudich, J.A. (1985) J. Cell Biol. 100, 1024-1039. 6 Sweeney, H.L., Kushmerick, M.J., Mabuchi, K., Streter, F.A. and Gergely, J. (1988) J. Biol. Chem. 263, 9034-9039. 7 Mabuchi, K., Pinter, K., Mabuchi, Y., Streter, F.A. and Gergely, J. (1984) Muscle Nerve 7, 431-438. 8 Staron, R.S. and Pette, D. (1986) Histochemistry 86, 19-23. 9 Peleg, I., Muhlrad, A., Eldor, A., Groschel-Stewart, U. and Kahane, I. (1984) Arch. Biochem. Biophys. 234, 442-453. 10 Peleg, I., Eldor, A., Muhlrad, A., Oroschel-Stewart, U. and Kahane, I. (1985) Thromb. Res. 38, 567-577. 11 Kohler, G. and Millstein, C. (1975) Nature 256, 495-497. 12 Muhlrad, A. and Morales, M.F. (1984) Proc. Natl. Acad. Sci. USA 81, 1003-1007. 13 Dan-Goor, M., Kessel, M., Silberstein, L. and Muhlrad, A. (1988) J. Muscle Res. Cell Motil. 9, 75-76. 14 Mornet, D., Pantel, P., Audermard, E. and Kassab, R. (1979) Biochem. Biophys. Res. Commun. 98, 928-932. 15 Hozumi, T. (1983) Biochemistry 22, 779-784. 16 Setton, A. and Muhlrad, A. (1984) Arch. Biochem. Biophys. 235, 411-417. 17 Chaussepied, P., Morner, D., Audermard, E., Derancourt, J. and Kassab, R. (1986) Biochemistry 25, 1134-1140. 18 Sutoh, K., Tokunaga, M. and Wakayabashi, T. (1987) J. Mol. Biol. 195, 953-956. 19 Jameson, B. and Wolf, F. (1988) Comp. Appl. Biosci. (CABIOS) 4, 181-186. 20 Sutoh, K. (1983) Biochemistry 22, 1579-1585. 21 Tonomura, Y., Appel, P. and Morales, M.F. (1966) Biochemistry 5, 515-521. 22 Weeds, A.G. and Taylor, R.S. (1975) Nature 257, 54-56.

Characterization of monoclonal antibodies to human platelet myosin that recognize highly conserved epitopes within the 50 kDa fragment of myosin subfragment-1.

Three monoclonal antibodies directed against human platelet myosin heavy chains (MCH) that recognize homologous sequences contained within the functio...
659KB Sizes 0 Downloads 0 Views