- Email: [email protected]
Calmodulin-stimulated protein methylation in rat liver cytosol
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 237, No. 2, March, pp. 347-353, 1985
Calmodulin-Stimulated
Protein Methylation
FRANK L. SIEGEL’
in Rat Liver Cytosol’
LYNDA S. WRIGHT
AND
Departments of Pediatrics and Physiological Chemistry Retardation and Human De-velopnent, University
and The Ham-g A. Waisman Center for Mental of Wisconsin, Madison, Wisconsin 53706
Received August 23, 1984, and in revised form October
30, 1984
The in vitro methylation of three liver cytosolic proteins was found to be selectively stimulated by calmodulin. This effect was also seen, although to a much smaller degree, in kidney and lung, but not in testes, brain, or spleen. The three methylated proteins affected by calmodulin have apparent M, = 29,000, 32,000, and 45,000. The stimulation of methylation by calmodulin was greatest for the M, 29,000 protein; there was an equal degree of methylation of the other two proteins. Dialysis of liver cytosolic fractions also stimulated the methylation of these proteins; the methylation of the M, 32,000 and 45,000 proteins was stimulated to a greater extent by dialysis than by calmodulin. The degree of stimulation of methylation of the M, 29,000 protein by calmodulin and dialysis was equivalent, but the addition of calmodulin to dialyzed liver cytosolic fractions gave additive effects on the stimulation of methylation of the M, 29,000 protein, but not of either the M, 32,000 or 45,000 proteins. Troponin C stimulated the methylation of the M, 29,000 protein, but not the M, 32,000 or 45,000 proteins, whereas parvalbumin stimulated methylation of the M, 32,000 protein, but not the M, 29,000 or 45,000 proteins. The effects of calmodulin and dialysis on protein methylation are cation-dependent and substrate-specific; methylation of the M, 29,000 was supported by Mn’+, Ca2+ and Co’+, and to a lesser degree by Mg2’, Ni’+, and Zn’+. Methylation of the M, 3k,OOOprotein was supported only by Mn2+ and Mg2+ and methylation of the n/r, 45,000 protein by Mn2+, Me, Ca’+, Ni2+, and Zn2+, and to a much smaller extent by Fe’+. In extracts of fetal liver, stimulation of protein methylation by calmodulin or dialysis was restricted to the M, 45,000 protein. In regenerating liver, stimulation of the methylation of all three proteins was observed, but the stimulation provided by dialysis plus calmodulin was much less than that observed in preparations from intact adult liver, suggesting a possible negative correlation between the rate of cell division and calmodulin-dependent methylation of these hepatic proteins. These results are consistent with the presence in liver of a minimum of three distinct N-methyltransferases and a dialyzable inhibitor which antagonizes calmodulin-dependent protein methylation. (c 1985 Academic press. Inc. Calmodulin, a ubiquitous calcium-binding protein, has been shown to mediate the calcium-dependent stimulation of a large number of enzymes and physiologi-
cal processes (l-4). Calmodulin is structurally unusual in that it undergoes a highly specific post-translational N-methylation, which converts a single specific lysine residue to trimethyllysine; this takes place in most, but not all, eucaryotic organisms (5-7). We have shown that the enzyme which N-methylates calmodulin is specific for calmodulin as a substrate (8); there are, however, other N-methyltransferases (9-14) and considerable tissue
‘This research was supported by National Institutes of Health Grants NS 11652 and HD 03552. ‘To whom correspondence and reprint requests should be addressed: Neurochemistry Section, Waisman Center, The University of Wisconsin, Madison, Wise. 53706 347
0003-9861/85 $3.00 Copyright All rights
0 1985 by Academic Press, Inc. of reproduction in any form reserved.
348
SIEGEL
AND
specificity in patterns of in vitro methylation (15). While the significance of protein N-methylation is not known, a recent report that N-methylated, but not nonmethylated, calmodulin can activate membrane-bound guanylate cyclase of paramecium (16) suggests that N-methylation may regulate some of the biological activities of calmodulin. Calmodulin is also a substrate for protein carboxylmethyltransferase, an enzyme which catalyzes the methyl esterification of carboxyl groups of aspartic or glutamic acid residues of proteins (17, 18). Carboxylmethylation of calmodulin has been reported to result in the loss of its biological activities (17, 19), presumably by neutralizing negative charges which participate in calcium binding. In this report, we present evidence to indicate that in some tissues the stable methylation of a few specific proteins appears to be calmodulin-dependent and inhibited by a low-molecular-weight compound. Stimulation of protein methylation represents a previously unreported activity of calmodulin. MATERIALS
AND
METHODS
Tissue preparation Holtzman rats (200-300 g) were killed by carbon dioxide asphyxiation and livers were perfused with 0.9% sodium chloride injected into the portal vein. Partial hepatectomies were performed under equithesin (3 ml/kg) anesthesia; median and left lobes (about two-thirds of the liver mass) were removed according to a previously described procedure (20). Fetal rats were surgically removed from timed-pregnancy mothers in our breeding colony. Perfused livers were homogenized pH 7.4, in 5 vol of 0.25 M sucrose, 10 mM Tris-HCl, filtered through cheesecloth, and centrifuged at 100,000~ for 60 min. The resulting high-speed supernatants were designated as cytosolic fractions. In experiments in which dialyzed cytosolic fractions served as the source of enzyme, high-speed supernatants were dialyzed overnight against three changes of 1000 vol Tris-HCl, pH 7.4, and clarified by centrifugation. Protein was assayed by the method of Bradford (21). In vitro methylution. Cytosolic protein (100 fig) was incubated for 60 min at 37°C in 50 ~1 of a solution containing 0.1 M Tris-HCl, pH 8.0, 1 mM DTT: 2 mM EDTA, 2.5 mM MnCl, and 5 &i [methyl-
’ Abbreviations used: DTT, sodium dodecyl sulfate.
dithiothreitol;
SDS,
WRIGHT ‘HI-S-adenosyl-L-methionine (80 Ci/mmol). Calmodulin (20 pg) was added to some incubation mixtures, as indicated in figure legends. To verify that the observed methylations represented stable modifications, and not carboxylmethylation, some incubation mixtures were treated with borate prior to electrophoresis to hydrolyze any labile carboxyl methyl esters (22). In this procedure, 25 pl of 0.13 M sodium borate, pH 11, was added to incubation mixtures, which were held at 37°C for 15 min. Reactions were terminated by the addition of 50 ~1 (75 ~1 for boratetreated samples) of 2 X O’Farrell electrophoresis buffer (23) (20% glycerol, 4.6% SDS, and 0.125 M Tris-HCl, pH 6.8) and samples were heated for 2 min in a boiling-water bath prior to electrophoresis. Methylated proteins were fractionated by SDS-polyacrylamide gel electrophoresis in slab gels (12.5% acrylamide, 0.33% bisacrylamide) for 20 h at 75 V/ gel, according to Laemmli (24). Gels were stained 2 h in 0.1% Coomassie blue, 9.25% acetic acid, and 50% methanol, and destained overnight in 12.5% isopropanol, 7.5% acetic acid. Destained gels were washed 30 min in distilled water and incubated 30 min in 1 M sodium salicylate (25), dried in wacuo at 8O’C for 2; h, and placed on Kodak XAR-5 film for 2 weeks at -70°C for fluorography. Molecular weight calibration in electrophoresis experiments was done using the following standards: bovine serum albumin, A4, 66,000; ovalbumin, M, 45,000; aldolase, M, 40,000; trypsinogen, M, 24,000; @-lactoglobulin, M, 18,400; and lysozyme, M, 14,300. Troponin C was a gift of Dr. Marion Greaser; calmodulin was isolated from beef testes as previously described (26). Nonmethylated calmodulin was isolated from cultured Dictyostelium discoideum (7), and a calmodulin isotype from beef testis was isolated by shallow-gradient DEAE-cellulose chromatography (in preparation). This calmodulin activates cyclic nucleotide phosphodiesterase equally as well as does “standard” beef testis calmodulin, but has incompletely characterized structural characteristics which distinguish it from “standard” calmodulin. des(Lys)calmodulin was prepared from beef testis calmodulin by the action of pituitary calmodulin-converting enzyme (31). Rabbit muscle parvalbumin and all reagents were obtained from Sigma Chemical Company. All proteins added to incubation mixtures were homogeneous by SDSpolyacrylamide gel electrophoresis in 15% gels.
RESULTS
Tissue specificity and calmodulin eflects on protein methylation Cytosolic fractions from rat tissues were incubated with [methyl-3H]-S-adenosyl-Lmethionine with and without added calmodulin. Following a 60-min incubation, methylated proteins were fractionated by SDS-polyacrylamide
CALMODULIN-STIMULATED
gel electrophoresis and visualized by fluorography. In preliminary experiments, the protein nature of all radiolabeled bands was verified by incubation with Pronase, RNase, or DNase following the period of methylation. All radiolabeled bands disappeared following treatment with Pronase; none were affected by either RNase or DNase (data not shown). The pattern of methylated peptides shows significant tissue specificity; this is most evident when the methylated proteins in testes are compared with those of other tissues (Fig. 1). In all tissues, peptides of M, 18,000 and 35,000 were the major methylated species; we have previously demonstrated that the M, 18,000 peptide is calmodulin (8). The addition of beef testis calmodulin inhibited the methylation of calmodulin in all tissues, as previously reported for brain (8). This may be due to product inhibition of calmodulin (lysine) N-methyltransferase by methylated calmodulin. Calmodulin also stimulated the methylation of several peptides; in liver, and to a lesser extent in kidney and lung, calmodulin specifically stimulated the methylation of a M,. 29,000 peptide (Fig. 1). A
abcdefghi Liver
Brain
jk Kidney
Testes
Lung
I Spleen
FIG. 1. Fluorogram of SDS-polyacrylamide gel electrophoresis of protein methylation in cytosolic extracts of rat tissues. Cytosolic protein (100 pg) was incubated with [methyl-3H]-S-adenosyl-L-methionine and methylated proteins were resolved by electrophoresis, as described under Materials and Methods. All lanes contain 100 pg of protein; samples in alternate lanes were incubated in the presence of 20 pg of added calmodulin.
PROTEIN
METHYLATION
slight stimulation
of the methylation
349 of a
M, 32,000 peptide was also observed in
these tissues. The only other effect of calmodulin on protein methylation was seen in testes; in this tissue, calmodulin stimulated the methylation of M, 24,000, 47,000 and 48,000 peptides. Spec$city of protein activators of methyylation. To determine if the stimulation
of hepatic protein methylation by calmodulin is specific for calmodulin and if different calmodulins are equally active, a comparison of several calcium-binding proteins was made. Liver cytosolic fractions were incubated under methylating conditions with several calcium-binding proteins, including nonmethylated calmodulin from D. discoideum (27), a calmodulin isotype from beef testes, des(1ys)calmodulin prepared by the action of pituitary calmodulin-converting enzyme (28), parvalbumin, troponin C, and varying amounts of beef testis calmodulin. desmethylcalmodulin was active as beef testis calmodulin in stimulating the methylation of the 29,000-Da peptide (Fig. 2); note that this calmodulin, being nonmethylated, is a good methyl acceptor-the methylated peptide, with an electrophoretie mobility slightly greater than beef calmodulin, comigrated with desmethylcalmodulin on stained gels (data not shown). des(Lys)calmodulin and the calmodulin isotype from beef testes also stimulated methylation of the M, 29,000 protein, as did troponin C, but not parvalbumin (Fig. 2). The extent of methylation of the M, 29,000 peptide increased with increasing amounts of added calmodulin, with an apparent saturation of the effect at about 20 pg of added calmodulin. Stimulation of methylation of the M, 32,000 peptide was observed with beef testis calmodulin (this effect was maximal at 2 pg of added calmodulin), des(lys)calmodulin, and parvalbumin, but not with the calmodulin isotype from beef testes or with troponin C (Fig. 2). Eflects of dialysis. When rat liver cytosolic fractions were dialyzed before incubation, it was found that dialysis alone specifically stimulated the methylation of the Mr 29,000 and 32,000 peptides, as well as a M, 45,000 peptide (Fig. 3). There was
350
SIEGEL
AND
Ko3,
-3s -29
-18
abcdefghijkl FIG. 2. Effects of calcium-binding proteins on protein methylation in rat liver cytolsolic fractions; fluorography of SDS-polyacrylamide gel. Methylation in the presence of added calcium-binding proteins: lane a, no additions; lane b, 10 fig of calmodulin from Dictyoetelium; lane c, 20 pg calmodulin isotype from rat testes; lane d, 20 pg des(lys)calmodulin; lane e, 20 pg parvalbumin; lane f, 20 CL@; troponin C; lanes g-f, added beef testes calmodulin (g, 0.5; h, 2; i, 5; j, 10; k, 20; and 1, 40 pg).
no detectable effect of dialysis on the protein composition of liver cytosolic fractions, as evaluated by comparison of the protein staining electrophoresis patterns of dialyzed and nondialyzed cytosolic fractions (Fig. 3). Dialysis had no detectable effect on protein methylation in cytosolic extracts of rat brain (data not shown). Cation dependence. Preliminary experiments indicated that in the absence of added divalent cations to dialyzed liver cytosol there was no detectable methylation of the liver M, 29,000 and 32,000 proteins, with or without added calmodulin (data not shown). To evaluate the cation requirements of the methylation of thse proteins, divalent cations were added to dialyzed liver cytosolic fractions prior to in vitro methylation. For the methylation of the &f, 29,000 protein, the order of activity of cations was Co2+ > Mn2+ = Ca2+ (Fig. 4). In the presence of 2 mM EDTA plus 2.5 mM added cation, Ca2+, Mn2+, and Co2+ supported high-activity methylation of the iW, 29,000 protein, while methylation of this protein in the presence of 2 mM EDTA and 2.5 mM Mgz+
WRIGHT
was markedly reduced compared with that supported by Me in the absence of EDTA, indicating a requirement for a second divalent cation (data not shown). Only Mn2+ and Mgz+ gave significant stimulation of methylation of the M, 32,000 protein; other cations did not support detectable activity (Fig. 4). Meth&-&m in fetal and regenerating live-r. Cytosolic fractions from livers of 19-day-gestation rat fetuses and regenerating livers 24 h after partial hepatectomy were compared with the corresponding preparations from adult liver with respect to their in vitro methylation of the M, 29,000, 32,000, and 45,000 proteins. No detectable methylation of either the M, 29,000 or 32,000 peptide was seen in either dialyzed or nondialyzed fetal liver cytosolic fractions, in the absence or presence of added calmodulin (Fig. 5); dialysis did not stimulate methylation of the M, 35,000 and 45,000 proteins in this preparation. In regenerating liver cytosolic fractions, no methylation of M, 29,000, 32,000 or
Ko3,
-45
-35 -32 -29
-18
a
b
c
d
e
f
g
FIG. 3. Effects of dialysis and calmodulin on protein methylation in rat liver cytosolic extract; fluorogram of SDS-polyacrylamide gel. Left panel, Coomassie blue stain; right panel, autoradiography. Lanes a, d, and f, no additions to incubation mixture; lanes b, e, and g, 20 fig added calmodulin; cytosolic fractions in lanes f and g were dialyzed before incubation, as described under Materials and Methods. Lane c contained molecular weight standards.
CALMODULIN-STIMULATED
mn Mg ca co
Ni
---m---
Fe
Zn
-46 -36 -32 -29 -24 -18
-+-+-+-+-+ abcdefghi
j
-+-+cat4 klmn
FIG. 4. Effects of cations and calmodulin on protein methylation in rat liver cytosolic fraction; fluorography of SDS-polyacrylamide gel. Divalent cations were added to incubation mixtures as indicated: lanes a and b, 2.5 mM Mnz+; lanes c and d, 2.5 mM Mg2+; lanes e and f, 2.5 mM Ca’+; lanes g and h, 2.5 mM Co’+; lanes i and j, 2.5 mM Niz+; lanes k and 1, 2.5 IIIM Fe’+; and lanes m and n, 2.5 mM Zn*+. Alternate incubation mixtures contained 20 pg of added calmodulin as indicated.
45,000 proteins was seen in the absence or presence of added calmodulin. In dialyzed extracts of regenerating liver, these proteins were methylated, although the stimulation by dialysis was much less than was observed in adult liver (Fig. 5). DISCUSSION
Stimulation of protein methylation by calmodulin represents a previously unreported activity of this calcium-binding
PROTEIN
METHYLATION
protein. The selectivity of this effect, in terms of protein substrates and tissues, indicates that the actions of calmodulin on protein methylation vary from one tissue to another. This could be due to tissue differences in either methyl acceptor proteins or methyltransferase enzymes. The greatest effect of calmodulin is on the methylation of the liver M, 29,000 protein, and we have focused our attention on this aspect of our initial findings. To determine the structural requirements for the stimulation of methylation, four structurally distinct calmodulins were tested for their ability to stimulate methylation of the M, 29,000 and 32,000 peptides in liver cytosolic extracts. All four calmodulins stimulated methylation of the M, 29,000 protein; methylation of the M, 32,000 protein was stimulated by three of the calmodulins, but not the isotype from testes. All four calmodulins stimulated methylation of the M, 29,000 protein to a greater extent than the M, 32,000 substrate. This result indicates that non-Nmethylated calmodulin (from DictyosteZium) is as active as fully methylated calmodulin and that lack of the carboxylterminal lysine residue does not impair the ability of calmodulin to stimulate protein methylation. Since troponin C stimulated methylation of the M, 29,000 protein but not the M, 32,000 protein, while parvalbumin stimulated methylation of the M, 32,000 protein but not the M, 29,000 protein, it appears as if separate methyltransferase enzymes catalyze the methylation of these two substrates. Troponin C is a particulate protein associated ‘Mrx IO3
caPA
-
a
+-+-+-+-+-+-+
bcdefghij
351
k
I
m
n
FIG. 5. Protein methylation in cytosolic extracts from fetal, regenerating, and intact adult rat liver; fluorogram of SDS-polyacrylamide gel. Lanes a and b, fetal liver; lanes c and d, dialyzed fetal liver; lanes e and f, regenerating liver; lanes g and h, dialyzed regenerating liver; lanes i and j, dialyzed adult liver; lanes k and 1, adult liver; lanes m and n, adult liver, borate-treated after methylation. Calmodulin (100 pg), where indicated, was present during methylation. That portion of the gel corresponding to a molecular weight range from M, 20,000 to 50,000 is shown.
352
SIEGEL
AND
with the contractile apparatus (28) and parvalhumin is found in muscle tissue, so that it is unlikely that these proteins influence protein methylation in viva in cytosol from liver or other tissues, but other calcium-binding proteins, such as oncomodulin (29), might act to influence the methylation of specific protein substrates. Since the enzyme which methylates calmodulin is specific for calmodulin as a substrate (S), there appear to be at least four N-methyltransferases in rat liver cytosolic fractions: (i) calmodulin (lysine) N-methyltransferase; (ii) calmodulin/troponin C-dependent N-methyltransferase; (iii) calmodulin/parvalbumindependent N-methyltransferase, and (iv) the enzyme(s) which methylate the M, 35,000 and minor methyl acceptor proteins. Alternatively, one can postulate that calmodulin-dependent protein methylation involves a ternary complex between enzyme, substrate, and calmodulin and that one enzyme may thus mediate the methylation of the M, 29,000,32,000, and 45,000 proteins. This model would indicate that there are at least three distinct N-methyltransferase enzymes. Preliminary data indicate that the N-methyltransferase activity in liver is almost exclusively cytosolic and that there is no calmodulindependent methylation in particulate fractions (data not shown). The finding of selective stimulation of protein methylation by dialysis resembles the report of the effects of a dialyzable peptide inhibitor of the N-methylation of low-molecular-weight compounds (30). This peptide, methinin, has been purified (30) and would appear to be a good candidate for underwriting the effects which we have observed. Our finding that the greatest dialysis-induced stimulation of methylation is that of the M, 45,000 and 32,000 proteins, while the other major substrates of protein methylation, the M, 35,000 and 29,000 proteins and calmodulin, are not affected by dialysis, has two possible explanations. The inhibitor may interact with methyltransferases, and two types of methyltransferase are distinguished by marked differences in their ability to bind the inhibitor. Alternatively,
WRIGHT
the inhibitor may interact with substrates, and selectivity is achieved at this level. In the absence of added cations, there is no detectable calmodulin-dependent or dialysis-dependent protein methylation (data not shown). Mn’+, Ca’+, Co’+, M$+, Ni2+, and Zn2+ all supported the calmodulin-dependent methylation of the M, 29,000 protein, whereas only Mn2+ and Mgz+ stimulated methylation of the M, 32,000 protein. This indicates that either different enzymes methylate the two proteins, or the two protein substrates undergo cation-specific conformational changes which enhance their methyl acceptor activity. The inhibition of methylation of the M, 29,000 protein in the presence of EDTA and excess Me, while no inhibition by EDTA in the presence of excess Ca2+, Mn’+, or Co2+ was found, indicates a requirement for one of the latter three cations. All three of these cations support the activation of cyclic nucleotide phosphodiesterase by calmodulin (32) and compete with one another for binding sites on calmodulin (33). These data thus indicate a calcium requirement for the calmodulin-dependent methylation of the M, 29,000 protein. Although many laboratories have investigated the N-methylation of proteins, no functional significance of this modification has been demonstrated. It was thus of particular interest to find that the calmodulin-dependent methylation of hepatic M, 29,000 and 32,000 proteins was not observed in extracts of fetal or regenerating liver. We are presently attempting to determine the biochemical signals which trigger the calmodulin-dependent methylation of the M, 29,000 liver protein and the specific residues on this substrate which are modified by methylation. REFERENCES 1. KLEE, C. B., AND VANAMAN, T. C. (1983) in Advances in Protein Chemistry (Anfinsen, C. B., Edsall, J. T., and Richards, F. M., eds.), Vol. 35, pp. 213-303, Academic Press, New York. 2. DEMAILLE,J.G.(~~~~)C&~~~C~U Fum 2,114144.
CALMODULIN-STIMULATED 3. CHEUNG, W. Y. (1980) Science ~Washington, D. C.) 207. 19-27. 4. CHAFOULEAS, J. G., AND MEANS, A. R. (1982) Cell 28,41-50. 5. MIYAKE, M., AND KAKIUCHI, S. (1978) Bruin Res. 139.378-380. 6. JACKSON, R. L., DEDMAN, J. R., SCHREIBER, W. E., BHATNAGA, P. K., KNAPP, R. D., AND MEANS, A. R. (1977) B&hem. Biqphys. Res. Commun. 77, 723-729. 7. BAZARI, W., AND CLARKE, M. (1981) J. Biol. Chem. 256, 3598-3603. 8. SITARAMAYYA, A., WRIGHT, L. S., AND SIEGEL, F. L. (1980) J. BioL Chem 25.5, 8894-8900. 9. PAIK, W. K., AND KIM, S. (1970) J. BioL Chem. 245, 6010-6015. 10. WALLWORK, J. C., QUICK, D. P., AND DUERRE, J. A. (1977) J. BioL Chem 252, 5977-5980. 11. WANG, C., LAZARIDES, E., O’CONNER, C. M., AND CLARKE, S. (1982) J. BioZ. Chem 257, 83568362. 12. LISCHIVE, P., AND BUSCH, H. (1982) J. Biol. &em. 257, 14600-14602. 13. PAIK, W. K., FAROOQUI, J., GUPTA, A., SMITH, H. T., AND MILLETT, F. (1983) Eur. J. B&hem 135.259-262. 14. COPPARD, N. J., CLARK, B. F. C., AND CRAMER, F. (1983) FEBS Lett. 164. 330-334. 15. MURTAUGH, T. J., SITARAMAYYA, A., WRIGHT, L. S., GRAVEN, K., ROWE, P. M., AND SIEGEL, F. L. (1980) in Calcium-Binding Proteins: Structure and Function (Siegel, F. L., Carfoli, E., Kretsinger, R. H., MacLennan, D. H., and Wasserman, R. H., eds.) pp. 197-204, Elsevier/ North-Holland, New York. 16. KLUMPP, S., KLEEFELD, G., AND SCHULTZ, J. E. (1983) ,I BioL Chem 258,12455-12459. 17. CAGNON, C., KELLY, S., MANGANIELLO, V., VAUGHN, M., ODYA, C., STRITTMATTER, W.,
PROTEIN
18. 19. 20.
21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
31.
32. 33.
METHYLATION
353
HOFFMAN, A., AND HIRATA, F. (1981) Nature (London) 291, 515-516. JANSON, C. A., AND CLARK, S. (1980) J. Biol Chem. 255, 11640-11643. BILLINGSLEY, M. L., ROTH, R. H., AND DELORENZO, R. J. (1983) .I Biol. Chem. 258. 5352-5357. WAYNFORTH, H. B. (1969) in Techniques in Protein Biosynthesis (Campbell, P. N., and Sargent, J. R., eds.), Vol. 2, pp. 209-249. BRADFORD, M. M. (1976) Anal. B&hem. 72,248254. KIM, S., AND PAIK, W. K. (1970) J. BioL Chem. 245.1806-1813. O’FARRELL, P. H. (1975) J. Biol. Chem 250,40074020. LAEMMLI, U. K. (1970) Nature @ondon) 227,680685. CHAMBERLAIN, J. (1979) Anal. B&hem. 98, 132135. WOLFF, D. J., AND SIEGEL, F. L. (1972) J. Biol. Chem. 247, 4180-4185. BAZARI, W. L., AND CLARKE, M. (1981) J. Biol Chem. 256, 3598-3603. GREASER, M. L., AND GERGELY, J. (1973) .I Biol Chem. 248, 2125-2133. MACMANUS, J. P. (1983) B&him. Biophys. Acta 621, 296-304. LYON, E. S., MCPHIE, P., AND JAKOBY, W. B. (1982) B&hem. Biophys. Res. Commun 108, 846-850. MURTAUGH, T. J., ROWE, P. M., VINCENT, P. L., WRIGHT, L. S., AND SIEGEL, F. L. (1983) in Methods in Enzymology (Means, A. R., and O’Malley, B. W., eds.), Vol. 102, pp. 158-170, Academic Press, New York. TEO, T. S., AND WANG, J. H. (1973) .I BioL Chem. 248.5950-5955. LIN, Y. M., LIU, Y. P., AND CHEUNG, W. Y. (1974) J Biol Chem. 249.4943-4954.
Calmodulin-Stimulated
Protein Methylation
FRANK L. SIEGEL’
in Rat Liver Cytosol’
LYNDA S. WRIGHT
AND
Departments of Pediatrics and Physiological Chemistry Retardation and Human De-velopnent, University
and The Ham-g A. Waisman Center for Mental of Wisconsin, Madison, Wisconsin 53706
Received August 23, 1984, and in revised form October
30, 1984
The in vitro methylation of three liver cytosolic proteins was found to be selectively stimulated by calmodulin. This effect was also seen, although to a much smaller degree, in kidney and lung, but not in testes, brain, or spleen. The three methylated proteins affected by calmodulin have apparent M, = 29,000, 32,000, and 45,000. The stimulation of methylation by calmodulin was greatest for the M, 29,000 protein; there was an equal degree of methylation of the other two proteins. Dialysis of liver cytosolic fractions also stimulated the methylation of these proteins; the methylation of the M, 32,000 and 45,000 proteins was stimulated to a greater extent by dialysis than by calmodulin. The degree of stimulation of methylation of the M, 29,000 protein by calmodulin and dialysis was equivalent, but the addition of calmodulin to dialyzed liver cytosolic fractions gave additive effects on the stimulation of methylation of the M, 29,000 protein, but not of either the M, 32,000 or 45,000 proteins. Troponin C stimulated the methylation of the M, 29,000 protein, but not the M, 32,000 or 45,000 proteins, whereas parvalbumin stimulated methylation of the M, 32,000 protein, but not the M, 29,000 or 45,000 proteins. The effects of calmodulin and dialysis on protein methylation are cation-dependent and substrate-specific; methylation of the M, 29,000 was supported by Mn’+, Ca2+ and Co’+, and to a lesser degree by Mg2’, Ni’+, and Zn’+. Methylation of the M, 3k,OOOprotein was supported only by Mn2+ and Mg2+ and methylation of the n/r, 45,000 protein by Mn2+, Me, Ca’+, Ni2+, and Zn2+, and to a much smaller extent by Fe’+. In extracts of fetal liver, stimulation of protein methylation by calmodulin or dialysis was restricted to the M, 45,000 protein. In regenerating liver, stimulation of the methylation of all three proteins was observed, but the stimulation provided by dialysis plus calmodulin was much less than that observed in preparations from intact adult liver, suggesting a possible negative correlation between the rate of cell division and calmodulin-dependent methylation of these hepatic proteins. These results are consistent with the presence in liver of a minimum of three distinct N-methyltransferases and a dialyzable inhibitor which antagonizes calmodulin-dependent protein methylation. (c 1985 Academic press. Inc. Calmodulin, a ubiquitous calcium-binding protein, has been shown to mediate the calcium-dependent stimulation of a large number of enzymes and physiologi-
cal processes (l-4). Calmodulin is structurally unusual in that it undergoes a highly specific post-translational N-methylation, which converts a single specific lysine residue to trimethyllysine; this takes place in most, but not all, eucaryotic organisms (5-7). We have shown that the enzyme which N-methylates calmodulin is specific for calmodulin as a substrate (8); there are, however, other N-methyltransferases (9-14) and considerable tissue
‘This research was supported by National Institutes of Health Grants NS 11652 and HD 03552. ‘To whom correspondence and reprint requests should be addressed: Neurochemistry Section, Waisman Center, The University of Wisconsin, Madison, Wise. 53706 347
0003-9861/85 $3.00 Copyright All rights
0 1985 by Academic Press, Inc. of reproduction in any form reserved.
348
SIEGEL
AND
specificity in patterns of in vitro methylation (15). While the significance of protein N-methylation is not known, a recent report that N-methylated, but not nonmethylated, calmodulin can activate membrane-bound guanylate cyclase of paramecium (16) suggests that N-methylation may regulate some of the biological activities of calmodulin. Calmodulin is also a substrate for protein carboxylmethyltransferase, an enzyme which catalyzes the methyl esterification of carboxyl groups of aspartic or glutamic acid residues of proteins (17, 18). Carboxylmethylation of calmodulin has been reported to result in the loss of its biological activities (17, 19), presumably by neutralizing negative charges which participate in calcium binding. In this report, we present evidence to indicate that in some tissues the stable methylation of a few specific proteins appears to be calmodulin-dependent and inhibited by a low-molecular-weight compound. Stimulation of protein methylation represents a previously unreported activity of calmodulin. MATERIALS
AND
METHODS
Tissue preparation Holtzman rats (200-300 g) were killed by carbon dioxide asphyxiation and livers were perfused with 0.9% sodium chloride injected into the portal vein. Partial hepatectomies were performed under equithesin (3 ml/kg) anesthesia; median and left lobes (about two-thirds of the liver mass) were removed according to a previously described procedure (20). Fetal rats were surgically removed from timed-pregnancy mothers in our breeding colony. Perfused livers were homogenized pH 7.4, in 5 vol of 0.25 M sucrose, 10 mM Tris-HCl, filtered through cheesecloth, and centrifuged at 100,000~ for 60 min. The resulting high-speed supernatants were designated as cytosolic fractions. In experiments in which dialyzed cytosolic fractions served as the source of enzyme, high-speed supernatants were dialyzed overnight against three changes of 1000 vol Tris-HCl, pH 7.4, and clarified by centrifugation. Protein was assayed by the method of Bradford (21). In vitro methylution. Cytosolic protein (100 fig) was incubated for 60 min at 37°C in 50 ~1 of a solution containing 0.1 M Tris-HCl, pH 8.0, 1 mM DTT: 2 mM EDTA, 2.5 mM MnCl, and 5 &i [methyl-
’ Abbreviations used: DTT, sodium dodecyl sulfate.
dithiothreitol;
SDS,
WRIGHT ‘HI-S-adenosyl-L-methionine (80 Ci/mmol). Calmodulin (20 pg) was added to some incubation mixtures, as indicated in figure legends. To verify that the observed methylations represented stable modifications, and not carboxylmethylation, some incubation mixtures were treated with borate prior to electrophoresis to hydrolyze any labile carboxyl methyl esters (22). In this procedure, 25 pl of 0.13 M sodium borate, pH 11, was added to incubation mixtures, which were held at 37°C for 15 min. Reactions were terminated by the addition of 50 ~1 (75 ~1 for boratetreated samples) of 2 X O’Farrell electrophoresis buffer (23) (20% glycerol, 4.6% SDS, and 0.125 M Tris-HCl, pH 6.8) and samples were heated for 2 min in a boiling-water bath prior to electrophoresis. Methylated proteins were fractionated by SDS-polyacrylamide gel electrophoresis in slab gels (12.5% acrylamide, 0.33% bisacrylamide) for 20 h at 75 V/ gel, according to Laemmli (24). Gels were stained 2 h in 0.1% Coomassie blue, 9.25% acetic acid, and 50% methanol, and destained overnight in 12.5% isopropanol, 7.5% acetic acid. Destained gels were washed 30 min in distilled water and incubated 30 min in 1 M sodium salicylate (25), dried in wacuo at 8O’C for 2; h, and placed on Kodak XAR-5 film for 2 weeks at -70°C for fluorography. Molecular weight calibration in electrophoresis experiments was done using the following standards: bovine serum albumin, A4, 66,000; ovalbumin, M, 45,000; aldolase, M, 40,000; trypsinogen, M, 24,000; @-lactoglobulin, M, 18,400; and lysozyme, M, 14,300. Troponin C was a gift of Dr. Marion Greaser; calmodulin was isolated from beef testes as previously described (26). Nonmethylated calmodulin was isolated from cultured Dictyostelium discoideum (7), and a calmodulin isotype from beef testis was isolated by shallow-gradient DEAE-cellulose chromatography (in preparation). This calmodulin activates cyclic nucleotide phosphodiesterase equally as well as does “standard” beef testis calmodulin, but has incompletely characterized structural characteristics which distinguish it from “standard” calmodulin. des(Lys)calmodulin was prepared from beef testis calmodulin by the action of pituitary calmodulin-converting enzyme (31). Rabbit muscle parvalbumin and all reagents were obtained from Sigma Chemical Company. All proteins added to incubation mixtures were homogeneous by SDSpolyacrylamide gel electrophoresis in 15% gels.
RESULTS
Tissue specificity and calmodulin eflects on protein methylation Cytosolic fractions from rat tissues were incubated with [methyl-3H]-S-adenosyl-Lmethionine with and without added calmodulin. Following a 60-min incubation, methylated proteins were fractionated by SDS-polyacrylamide
CALMODULIN-STIMULATED
gel electrophoresis and visualized by fluorography. In preliminary experiments, the protein nature of all radiolabeled bands was verified by incubation with Pronase, RNase, or DNase following the period of methylation. All radiolabeled bands disappeared following treatment with Pronase; none were affected by either RNase or DNase (data not shown). The pattern of methylated peptides shows significant tissue specificity; this is most evident when the methylated proteins in testes are compared with those of other tissues (Fig. 1). In all tissues, peptides of M, 18,000 and 35,000 were the major methylated species; we have previously demonstrated that the M, 18,000 peptide is calmodulin (8). The addition of beef testis calmodulin inhibited the methylation of calmodulin in all tissues, as previously reported for brain (8). This may be due to product inhibition of calmodulin (lysine) N-methyltransferase by methylated calmodulin. Calmodulin also stimulated the methylation of several peptides; in liver, and to a lesser extent in kidney and lung, calmodulin specifically stimulated the methylation of a M,. 29,000 peptide (Fig. 1). A
abcdefghi Liver
Brain
jk Kidney
Testes
Lung
I Spleen
FIG. 1. Fluorogram of SDS-polyacrylamide gel electrophoresis of protein methylation in cytosolic extracts of rat tissues. Cytosolic protein (100 pg) was incubated with [methyl-3H]-S-adenosyl-L-methionine and methylated proteins were resolved by electrophoresis, as described under Materials and Methods. All lanes contain 100 pg of protein; samples in alternate lanes were incubated in the presence of 20 pg of added calmodulin.
PROTEIN
METHYLATION
slight stimulation
of the methylation
349 of a
M, 32,000 peptide was also observed in
these tissues. The only other effect of calmodulin on protein methylation was seen in testes; in this tissue, calmodulin stimulated the methylation of M, 24,000, 47,000 and 48,000 peptides. Spec$city of protein activators of methyylation. To determine if the stimulation
of hepatic protein methylation by calmodulin is specific for calmodulin and if different calmodulins are equally active, a comparison of several calcium-binding proteins was made. Liver cytosolic fractions were incubated under methylating conditions with several calcium-binding proteins, including nonmethylated calmodulin from D. discoideum (27), a calmodulin isotype from beef testes, des(1ys)calmodulin prepared by the action of pituitary calmodulin-converting enzyme (28), parvalbumin, troponin C, and varying amounts of beef testis calmodulin. desmethylcalmodulin was active as beef testis calmodulin in stimulating the methylation of the 29,000-Da peptide (Fig. 2); note that this calmodulin, being nonmethylated, is a good methyl acceptor-the methylated peptide, with an electrophoretie mobility slightly greater than beef calmodulin, comigrated with desmethylcalmodulin on stained gels (data not shown). des(Lys)calmodulin and the calmodulin isotype from beef testes also stimulated methylation of the M, 29,000 protein, as did troponin C, but not parvalbumin (Fig. 2). The extent of methylation of the M, 29,000 peptide increased with increasing amounts of added calmodulin, with an apparent saturation of the effect at about 20 pg of added calmodulin. Stimulation of methylation of the M, 32,000 peptide was observed with beef testis calmodulin (this effect was maximal at 2 pg of added calmodulin), des(lys)calmodulin, and parvalbumin, but not with the calmodulin isotype from beef testes or with troponin C (Fig. 2). Eflects of dialysis. When rat liver cytosolic fractions were dialyzed before incubation, it was found that dialysis alone specifically stimulated the methylation of the Mr 29,000 and 32,000 peptides, as well as a M, 45,000 peptide (Fig. 3). There was
350
SIEGEL
AND
Ko3,
-3s -29
-18
abcdefghijkl FIG. 2. Effects of calcium-binding proteins on protein methylation in rat liver cytolsolic fractions; fluorography of SDS-polyacrylamide gel. Methylation in the presence of added calcium-binding proteins: lane a, no additions; lane b, 10 fig of calmodulin from Dictyoetelium; lane c, 20 pg calmodulin isotype from rat testes; lane d, 20 pg des(lys)calmodulin; lane e, 20 pg parvalbumin; lane f, 20 CL@; troponin C; lanes g-f, added beef testes calmodulin (g, 0.5; h, 2; i, 5; j, 10; k, 20; and 1, 40 pg).
no detectable effect of dialysis on the protein composition of liver cytosolic fractions, as evaluated by comparison of the protein staining electrophoresis patterns of dialyzed and nondialyzed cytosolic fractions (Fig. 3). Dialysis had no detectable effect on protein methylation in cytosolic extracts of rat brain (data not shown). Cation dependence. Preliminary experiments indicated that in the absence of added divalent cations to dialyzed liver cytosol there was no detectable methylation of the liver M, 29,000 and 32,000 proteins, with or without added calmodulin (data not shown). To evaluate the cation requirements of the methylation of thse proteins, divalent cations were added to dialyzed liver cytosolic fractions prior to in vitro methylation. For the methylation of the &f, 29,000 protein, the order of activity of cations was Co2+ > Mn2+ = Ca2+ (Fig. 4). In the presence of 2 mM EDTA plus 2.5 mM added cation, Ca2+, Mn2+, and Co2+ supported high-activity methylation of the iW, 29,000 protein, while methylation of this protein in the presence of 2 mM EDTA and 2.5 mM Mgz+
WRIGHT
was markedly reduced compared with that supported by Me in the absence of EDTA, indicating a requirement for a second divalent cation (data not shown). Only Mn2+ and Mgz+ gave significant stimulation of methylation of the M, 32,000 protein; other cations did not support detectable activity (Fig. 4). Meth&-&m in fetal and regenerating live-r. Cytosolic fractions from livers of 19-day-gestation rat fetuses and regenerating livers 24 h after partial hepatectomy were compared with the corresponding preparations from adult liver with respect to their in vitro methylation of the M, 29,000, 32,000, and 45,000 proteins. No detectable methylation of either the M, 29,000 or 32,000 peptide was seen in either dialyzed or nondialyzed fetal liver cytosolic fractions, in the absence or presence of added calmodulin (Fig. 5); dialysis did not stimulate methylation of the M, 35,000 and 45,000 proteins in this preparation. In regenerating liver cytosolic fractions, no methylation of M, 29,000, 32,000 or
Ko3,
-45
-35 -32 -29
-18
a
b
c
d
e
f
g
FIG. 3. Effects of dialysis and calmodulin on protein methylation in rat liver cytosolic extract; fluorogram of SDS-polyacrylamide gel. Left panel, Coomassie blue stain; right panel, autoradiography. Lanes a, d, and f, no additions to incubation mixture; lanes b, e, and g, 20 fig added calmodulin; cytosolic fractions in lanes f and g were dialyzed before incubation, as described under Materials and Methods. Lane c contained molecular weight standards.
CALMODULIN-STIMULATED
mn Mg ca co
Ni
---m---
Fe
Zn
-46 -36 -32 -29 -24 -18
-+-+-+-+-+ abcdefghi
j
-+-+cat4 klmn
FIG. 4. Effects of cations and calmodulin on protein methylation in rat liver cytosolic fraction; fluorography of SDS-polyacrylamide gel. Divalent cations were added to incubation mixtures as indicated: lanes a and b, 2.5 mM Mnz+; lanes c and d, 2.5 mM Mg2+; lanes e and f, 2.5 mM Ca’+; lanes g and h, 2.5 mM Co’+; lanes i and j, 2.5 mM Niz+; lanes k and 1, 2.5 IIIM Fe’+; and lanes m and n, 2.5 mM Zn*+. Alternate incubation mixtures contained 20 pg of added calmodulin as indicated.
45,000 proteins was seen in the absence or presence of added calmodulin. In dialyzed extracts of regenerating liver, these proteins were methylated, although the stimulation by dialysis was much less than was observed in adult liver (Fig. 5). DISCUSSION
Stimulation of protein methylation by calmodulin represents a previously unreported activity of this calcium-binding
PROTEIN
METHYLATION
protein. The selectivity of this effect, in terms of protein substrates and tissues, indicates that the actions of calmodulin on protein methylation vary from one tissue to another. This could be due to tissue differences in either methyl acceptor proteins or methyltransferase enzymes. The greatest effect of calmodulin is on the methylation of the liver M, 29,000 protein, and we have focused our attention on this aspect of our initial findings. To determine the structural requirements for the stimulation of methylation, four structurally distinct calmodulins were tested for their ability to stimulate methylation of the M, 29,000 and 32,000 peptides in liver cytosolic extracts. All four calmodulins stimulated methylation of the M, 29,000 protein; methylation of the M, 32,000 protein was stimulated by three of the calmodulins, but not the isotype from testes. All four calmodulins stimulated methylation of the M, 29,000 protein to a greater extent than the M, 32,000 substrate. This result indicates that non-Nmethylated calmodulin (from DictyosteZium) is as active as fully methylated calmodulin and that lack of the carboxylterminal lysine residue does not impair the ability of calmodulin to stimulate protein methylation. Since troponin C stimulated methylation of the M, 29,000 protein but not the M, 32,000 protein, while parvalbumin stimulated methylation of the M, 32,000 protein but not the M, 29,000 protein, it appears as if separate methyltransferase enzymes catalyze the methylation of these two substrates. Troponin C is a particulate protein associated ‘Mrx IO3
caPA
-
a
+-+-+-+-+-+-+
bcdefghij
351
k
I
m
n
FIG. 5. Protein methylation in cytosolic extracts from fetal, regenerating, and intact adult rat liver; fluorogram of SDS-polyacrylamide gel. Lanes a and b, fetal liver; lanes c and d, dialyzed fetal liver; lanes e and f, regenerating liver; lanes g and h, dialyzed regenerating liver; lanes i and j, dialyzed adult liver; lanes k and 1, adult liver; lanes m and n, adult liver, borate-treated after methylation. Calmodulin (100 pg), where indicated, was present during methylation. That portion of the gel corresponding to a molecular weight range from M, 20,000 to 50,000 is shown.
352
SIEGEL
AND
with the contractile apparatus (28) and parvalhumin is found in muscle tissue, so that it is unlikely that these proteins influence protein methylation in viva in cytosol from liver or other tissues, but other calcium-binding proteins, such as oncomodulin (29), might act to influence the methylation of specific protein substrates. Since the enzyme which methylates calmodulin is specific for calmodulin as a substrate (S), there appear to be at least four N-methyltransferases in rat liver cytosolic fractions: (i) calmodulin (lysine) N-methyltransferase; (ii) calmodulin/troponin C-dependent N-methyltransferase; (iii) calmodulin/parvalbumindependent N-methyltransferase, and (iv) the enzyme(s) which methylate the M, 35,000 and minor methyl acceptor proteins. Alternatively, one can postulate that calmodulin-dependent protein methylation involves a ternary complex between enzyme, substrate, and calmodulin and that one enzyme may thus mediate the methylation of the M, 29,000,32,000, and 45,000 proteins. This model would indicate that there are at least three distinct N-methyltransferase enzymes. Preliminary data indicate that the N-methyltransferase activity in liver is almost exclusively cytosolic and that there is no calmodulindependent methylation in particulate fractions (data not shown). The finding of selective stimulation of protein methylation by dialysis resembles the report of the effects of a dialyzable peptide inhibitor of the N-methylation of low-molecular-weight compounds (30). This peptide, methinin, has been purified (30) and would appear to be a good candidate for underwriting the effects which we have observed. Our finding that the greatest dialysis-induced stimulation of methylation is that of the M, 45,000 and 32,000 proteins, while the other major substrates of protein methylation, the M, 35,000 and 29,000 proteins and calmodulin, are not affected by dialysis, has two possible explanations. The inhibitor may interact with methyltransferases, and two types of methyltransferase are distinguished by marked differences in their ability to bind the inhibitor. Alternatively,
WRIGHT
the inhibitor may interact with substrates, and selectivity is achieved at this level. In the absence of added cations, there is no detectable calmodulin-dependent or dialysis-dependent protein methylation (data not shown). Mn’+, Ca’+, Co’+, M$+, Ni2+, and Zn2+ all supported the calmodulin-dependent methylation of the M, 29,000 protein, whereas only Mn2+ and Mgz+ stimulated methylation of the M, 32,000 protein. This indicates that either different enzymes methylate the two proteins, or the two protein substrates undergo cation-specific conformational changes which enhance their methyl acceptor activity. The inhibition of methylation of the M, 29,000 protein in the presence of EDTA and excess Me, while no inhibition by EDTA in the presence of excess Ca2+, Mn’+, or Co2+ was found, indicates a requirement for one of the latter three cations. All three of these cations support the activation of cyclic nucleotide phosphodiesterase by calmodulin (32) and compete with one another for binding sites on calmodulin (33). These data thus indicate a calcium requirement for the calmodulin-dependent methylation of the M, 29,000 protein. Although many laboratories have investigated the N-methylation of proteins, no functional significance of this modification has been demonstrated. It was thus of particular interest to find that the calmodulin-dependent methylation of hepatic M, 29,000 and 32,000 proteins was not observed in extracts of fetal or regenerating liver. We are presently attempting to determine the biochemical signals which trigger the calmodulin-dependent methylation of the M, 29,000 liver protein and the specific residues on this substrate which are modified by methylation. REFERENCES 1. KLEE, C. B., AND VANAMAN, T. C. (1983) in Advances in Protein Chemistry (Anfinsen, C. B., Edsall, J. T., and Richards, F. M., eds.), Vol. 35, pp. 213-303, Academic Press, New York. 2. DEMAILLE,J.G.(~~~~)C&~~~C~U Fum 2,114144.
CALMODULIN-STIMULATED 3. CHEUNG, W. Y. (1980) Science ~Washington, D. C.) 207. 19-27. 4. CHAFOULEAS, J. G., AND MEANS, A. R. (1982) Cell 28,41-50. 5. MIYAKE, M., AND KAKIUCHI, S. (1978) Bruin Res. 139.378-380. 6. JACKSON, R. L., DEDMAN, J. R., SCHREIBER, W. E., BHATNAGA, P. K., KNAPP, R. D., AND MEANS, A. R. (1977) B&hem. Biqphys. Res. Commun. 77, 723-729. 7. BAZARI, W., AND CLARKE, M. (1981) J. Biol. Chem. 256, 3598-3603. 8. SITARAMAYYA, A., WRIGHT, L. S., AND SIEGEL, F. L. (1980) J. BioL Chem 25.5, 8894-8900. 9. PAIK, W. K., AND KIM, S. (1970) J. BioL Chem. 245, 6010-6015. 10. WALLWORK, J. C., QUICK, D. P., AND DUERRE, J. A. (1977) J. BioL Chem 252, 5977-5980. 11. WANG, C., LAZARIDES, E., O’CONNER, C. M., AND CLARKE, S. (1982) J. BioZ. Chem 257, 83568362. 12. LISCHIVE, P., AND BUSCH, H. (1982) J. Biol. &em. 257, 14600-14602. 13. PAIK, W. K., FAROOQUI, J., GUPTA, A., SMITH, H. T., AND MILLETT, F. (1983) Eur. J. B&hem 135.259-262. 14. COPPARD, N. J., CLARK, B. F. C., AND CRAMER, F. (1983) FEBS Lett. 164. 330-334. 15. MURTAUGH, T. J., SITARAMAYYA, A., WRIGHT, L. S., GRAVEN, K., ROWE, P. M., AND SIEGEL, F. L. (1980) in Calcium-Binding Proteins: Structure and Function (Siegel, F. L., Carfoli, E., Kretsinger, R. H., MacLennan, D. H., and Wasserman, R. H., eds.) pp. 197-204, Elsevier/ North-Holland, New York. 16. KLUMPP, S., KLEEFELD, G., AND SCHULTZ, J. E. (1983) ,I BioL Chem 258,12455-12459. 17. CAGNON, C., KELLY, S., MANGANIELLO, V., VAUGHN, M., ODYA, C., STRITTMATTER, W.,
PROTEIN
18. 19. 20.
21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
31.
32. 33.
METHYLATION
353
HOFFMAN, A., AND HIRATA, F. (1981) Nature (London) 291, 515-516. JANSON, C. A., AND CLARK, S. (1980) J. Biol Chem. 255, 11640-11643. BILLINGSLEY, M. L., ROTH, R. H., AND DELORENZO, R. J. (1983) .I Biol. Chem. 258. 5352-5357. WAYNFORTH, H. B. (1969) in Techniques in Protein Biosynthesis (Campbell, P. N., and Sargent, J. R., eds.), Vol. 2, pp. 209-249. BRADFORD, M. M. (1976) Anal. B&hem. 72,248254. KIM, S., AND PAIK, W. K. (1970) J. BioL Chem. 245.1806-1813. O’FARRELL, P. H. (1975) J. Biol. Chem 250,40074020. LAEMMLI, U. K. (1970) Nature @ondon) 227,680685. CHAMBERLAIN, J. (1979) Anal. B&hem. 98, 132135. WOLFF, D. J., AND SIEGEL, F. L. (1972) J. Biol. Chem. 247, 4180-4185. BAZARI, W. L., AND CLARKE, M. (1981) J. Biol Chem. 256, 3598-3603. GREASER, M. L., AND GERGELY, J. (1973) .I Biol Chem. 248, 2125-2133. MACMANUS, J. P. (1983) B&him. Biophys. Acta 621, 296-304. LYON, E. S., MCPHIE, P., AND JAKOBY, W. B. (1982) B&hem. Biophys. Res. Commun 108, 846-850. MURTAUGH, T. J., ROWE, P. M., VINCENT, P. L., WRIGHT, L. S., AND SIEGEL, F. L. (1983) in Methods in Enzymology (Means, A. R., and O’Malley, B. W., eds.), Vol. 102, pp. 158-170, Academic Press, New York. TEO, T. S., AND WANG, J. H. (1973) .I BioL Chem. 248.5950-5955. LIN, Y. M., LIU, Y. P., AND CHEUNG, W. Y. (1974) J Biol Chem. 249.4943-4954.