Houston. Texas 77225. Troponin C regulates contraction in striated muscle by alternating ..... cTnC is predicted to occupy the same relative position as Gln-.
THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 268, No. 10, Issue of April 5, pp. 6827-6830, 1993 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A .
Communication Formation ofInter- and Intramolecular Disulfide Bonds Can Activate Cardiac TroponinC* (Received for publication, December 30, 1992, and in revised form, January 27,1993) John A. Putkey$, Darrell G . Dotson, and Pascale Mouawad From the Department of Biochemistry and Molecular Biology, University of Texas Medical School, Houston. Texas 77225
interaction with other muscle regulatory proteins. As a test of this model, Grabarek et al. (1990) and Fujimori et al. (1990) have engineered sTnC such that interactionsbetween helices B/C and D are stabilized and have shown that the activityof the protein is inhibited. Attempts to engineera constitutively active TnC would complementstructure/functionstudiesthatinactivatethe protein.Thisis aseeminglyformidable task; however, a fortuitous observationshowed that dialysis of cTnC toremove dithiothreitol (DTT) from the protein purification buffers conferred Ca2+-independentactivity whenassayed infast skeletal muscle myofibrils and that this activity is reversed upon the addition of DTT (Putkey et al., 1992). This suggested that Ca2+-independent activity can result from the formation of disulfide bonds in cTnC.To determine the molecular basis of this phenomenon, a series of recombinant proteins with substituted Cys residues were generated using recombinant cTnC (cTnC3) as a parent molecule. The functional characteristics of these monocysteine derivatives, and both cTnC3 and tissue-derived cTnC, were determined after dialysis to induce formation of disulfide bonds.The datashow that Ca2+independentforms of cTnCresult from thespontaneous formation of an intramolecular disulfide bond between Cys35 and Cys-84, or dimerizationvia an intermoleculardisulfide bond between Cys-84.
Troponin C regulates contraction in striated muscle by alternating between the Ca2+-bound and apo conformations. We report here that spontaneous formation of an intramolecular disulfide bond between Cys-35 and Cys-84, or dimerization via an intermolecular disulfide bond between Cys-84 in cardiac troponin C, renders the protein Ca2+-independentwhen assayed in fast skeletal muscle myofibrils but to a much lesser extent in cardiac myofibrils. Formation of the intramolecular disulfide bond appears to expose hydrophobic surfaces, as indicated by anincrease in fluorescence from hydrophobic fluorescent dyes, but does not alter the affinity of Ca2+-binding site 11. These disulfide bonds constrain the protein into a conformation that MATERIALS AND METHODS either resembles or can substitute for the Ca2+-bound formof cardiac troponin C in fast skeletal muscle Recombinant Proteins-Generation of an expression plasmid for recombinant cTnC3 and its modified derivatives has been described myofibrils.
previously (Putkey et al., 1989). TheCys residues were converted to Ser by oligonucleotide-primed site-directed mutagenesis using the method of Kunkel (1985).All recombinant proteins were isolated as Troponin C (TnC)’ andcalmodulin are membersof the EF- described previously (Putkey et al., 1989). To induce disulfide bond hand family of Ca2+-binding proteins. In contrast tocalmod- formation, the purified proteinswere simply dialyzed against several changes of 50 mM MOPS, pH 7.0, for at least 48 h. ulin, which has multiple activities, TnC appears tohave the Myofibril ATPase Assay-Isolation of myofibrils from rabbit back singular functionof regulating contraction in striated muscle. or cardiac muscle, preparation of TnC-extracted myofibrils, and the There are two isoforms of TnC; one is found in fast skeletal myofibril ATPase assay were performedas described previously(Putkey et al., 1991). The standard myofibril assay buffer contained 50 muscle (sTnC), and the other is found in slow and cardiac muscle (cTnC). Muscle contraction isregulated by the cyclical mM MOPS, pH 7.0, 2 mM EGTA, and KC1 to achieve a final ionic binding and release of Caz+ at the N-terminal Ca2+-binding strength of 100 mM. Stock assay buffers were prepared without or with CaC12 to achieve pCa 4. The final concentrations of myofibrils sites I and I1 in sTnC and at siteI1 in cTnC (Johnson et al., and TnC proteins were 0.25 mg/ml and 20 rg/ml, respectively. 1979; Robertson et al., 1981; Zot et al., 1982; Putkey et al., FluorescenceMeasurements-Fluorescence measurementswere 1989; Sweeney et al., 1990; Sheng et al., 1990).Herzberg, made in a buffer of 50 mM MOPS, pH 7.0, 150 mM KC1, 1.0 mM Moult, and James (Herzberg et al., 1986) have proposed a EGTA, 15~ L M protein, and30 FM probe. Stock solutionsof bis(ANS) and 9-AC were prepared in 100 mM MOPS, pH 7.0, or water, respecmodel for structural transitions in sTnC, which occur when tively. The excitation and emission wavelengths for bis(ANS) and 9I and 11, that involves theconcerted Ca2+bindstosites were370 and 482 nm, and 362 and 496 nm, respectively. The movement of helices B and C away from helices D, thereby AC resolution of the spectrofluorimeter was set at 5 nm. The concentraexposing a hydrophobic pocket that is thought to be a site of tion of total Ca2+needed to achieve desired free Caz+concentrations was calculated using a computer program of Fabiato (1988). Fluores* This work was supported inpart by grants (to J. A. P.) from the cence measurements were made ina Perkin Elmer LS-5 fluorescence National Institutes of Health (Grant AR-39210) and the Welch spectrometer. The relativefluorescenceemissionvalueswerecorFoundation. The costs of publication of this article were defrayed in rected for nonspecificeffectsof dilution and changes in ionic strength part by the payment of page charges. This article must therefore be by titration of a second sample with equal volumes of KC1 solutions hereby marked “advertisement” in accordance with 18 U.S.C. Section that were of equal ionic strength to the CaC12 solutions. 1734 solelyto indicate this fact. 4 Recipient of a Career Research Development Award from the RESULTS National Institutes of Health. To whom correspondence should be addressed Dept. of Biochemistry and Molecular Biolom. Universitv While characterizing the functional properties of recombiof Texas Medical School, P. 0. Box 20708,6431 Fannin-St., Houston, nant cTnC proteins in extracted fast skeletal muscle myoTX 77225. Tel.: 713-792-5604: Fax: 713-794-4150. fibrils, it was observed that Ca2+-independent activity was The abbreviations usedare: TnC,troponin C; cTnC,cardiac greatly enhancedafter extensivedialysis of theactivator troponin C; sTnC, fast skeletal troponin C; cTnC3, bacterially synproteins against50 mM Tris, pH7.5. Reduction of the dialyzed thesized cTnC(desMl,D2A); 9-AC,9-anthroylcholine;bis(ANS), 1,1’-bi(4-anilino)naphthalene-5,5’-disulfonic acid; DTT, dithiothrei- proteins with 5 mM DTT prior to the assay recovered the tol; MOPS, 4-morpholinepropanesulfonic acid. Ca2+-dependentactivity. These curious data suggested that
6827
Ca*+-independentc T n C
6828
mild oxidation of cTnC induced the formation of a disulfide weight forms that correspond to the three possible disulfide bond that rendered itCa'+-independent. T o study themolec- bonds thatwould yield dimers (data notshown). In all experular basis of this phenomenon,a series of recombinant cTnC iments only a fraction of the total monocysteine derivatives proteins were generated in which one or both of the Cys of cTnC were present in the dimer form. The significance of residues were converted to Ser. Bacterially synthesized cTnC3 the difference in relative molecular mass of the two dimers is has been shown in previous reports tobe functionally identical not clear, but Ca2+-bindingproteins are known to run anomto tissue-derived bovine cTnC (Putkey et al., 1989). Deriva- alously in SDS and urea gels. tives of cTnC3 in which Cys-35 or Cys-84 were changed to Fig. 1B shows that, in the absence of 2-mercaptoethanol, Ser are called cTnC3(C35S) and cTnC3(C84S),respectively. dialyzed cTnC3 migrates as a predominantly single molecular Although this standard nomenclature is descriptive of the weight specieswith an apparent molecular weight that is nature of the mutation, it must be kept in mind that, for slightly smaller than cTnC3(A-Cys). This difference in apexample, the Cys residue that is retained in cTnC3(C35S) parent molecular weight is more clearly seenin panel C. and is available for formation of disulfide bonds is Cys-84. Quantitation of Cys residues by exchange of free sulfhydryl For convenience, cTnC3 in which both Cys are changed to groups with 5,5'-dithiobis(nitrobenzoic acid) gave 0.2-0.3 mol Ser is called cTnC3(A-Cys). of Cys/mol of dialyzed cTnC3 .After reductionof cTnC3 with Fig. 1 summarizes numerous experiments that tested the 10 mM DTT followed bydesalting, exchangeof free sulfhydryl effect of reducing reagentsonthe relative electrophoretic groups with 5,5'-dithiobis(nitrobenzoic acid) showed 1.6-2.0 migration rates in SDS-polyacrylamide gels of tissue-derived mol of Cys/mol of protein. The data support a model in which and recombinant cTnC proteins that had been dialyzed to dialysis of cTnC3 allows formation of an intramolecular diinduce the formation of disulfide bonds. Panel A shows that sulfide bond that causes a more compact protein, which miall dialyzed proteins co-migrate as a single molecular species grates faster in anSDS-gel than thereduced protein. Dialysis in the presence or absence of EGTA appeared to when applied to the gel in a sample buffer that contains 2mercaptoethanol. Panel B shows that in the absence of 2- have no significant effect on the extent of formation of the mercaptoethanol, cTnC3(A-Cys) migrates as asingle band intramolecular disulfide bond in cTnC3. No Ca2+was bound with no higher molecular weight forms, but that the mono- to the cTnC3 afterdialysis in the presence of 0.5 mM EGTA cysteine derivatives show two molecular weight species. The and trace amounts of 45Ca2+.Ca2+that was bound to cTnC3 lower molecular weight forms co-migrate with cTnC3(A-Cys) that had been dialyzed with trace amounts of 4sCa2+,but in and correspond to the monomers. The single higher molecular the absence of EGTA, could be removed by addition of 0.5 weight forms are dimers that result from an intermolecular mM EGTA followed by desalting. This, togetherwith the data disulfide bond between Cys-35 in cTnC3(C84S) andbetween shown in Fig. 3, shows that formation of the intramolecular Cys-84 cTnC3(C35S). in Dialysis of a mixture of disulfide bond does not induce unusual Ca"-binding propercTnC3(C35S) and cTnC(C84S)yields three higher molecular ties in siteI1 of the free protein. Fig. 2A shows the results of a typical experiment, which measured the activity of dialyzed cTnC3,cTnC3(C35S), A 1 2 3 4 -66k cTnC3(C84S), and cTnC3(A-Cys) in fast skeletal myofibrils. - 45 Dialysis of cTnC3 induceda substantial amount of Ca2+-31 independent activity that was eliminated by treatment of the -21.5 dialyzed proteins with DTT. Dialysis or pretreatment with - 14.4 DTT had little or no effect on the activities of cTnC3(C84S) -6.5 or cTnC3(A-Cys). However, dialysis of cTnC3(C35S), which B 1 2 3 4 has a single Cys residue a t position 84, confers significant Ca2+-independentactivity that is eliminatedby treatment of the dialyzed protein with DTT. The extentof Ca2+-independent activity for cTnC3(C35S) is less than for cTnC3, but Fig. 1 shows that only a fraction of the total protein is in the dimer form after dialysis. Fig. 2B shows the activities of dialyzed cTnC3 and its C modified derivatives in cardiac muscle myofibrils. Dialyzed 1 2 3 .. - 4 _ ~ -200k , cTnC3 showed some DTT-sensitive Ca2+-independentactiv-116 ity but to much lesser extent than that seen in fast skeletal -- "66.2 muscle myofibrils. Potential Ca2+-independentactivity of ex-45 - -31 ogenous cTnC3 may be minimized by residual unextracted cTnCinthe myofibrils. Inourhands, we are unable to - -21.5 completely extract cTnCfrom cardiac myofibrils as evidenced "-14.4 "-6.5 by Ca2+-dependent ATPase activity in the extracted myofibrils that is 25-3096 of the maximal activity of the reconstiFIG.1. Effect of reducing reagents on the electrophoretic tuted myofibrils. Alternatively, differences in the strength of mobility of cTnC proteins. Proteins a t a concentration of 1-2 mg/ association of cTnC with cardiac myofibrils as compared to ml were dialyzed for 48 h against 50 mM MOPS, pH 7.0. Aliquots of skeletal myofibrils may favor breakage of the disulfide bond each protein were denatured in SDS sample buffer with or without 15 mM P-mercaptoethanol. The proteins were then applied toa 15% in the cardiacsystem. If the intermolecular disulfide bond is polyacrylamide gel (Laemmli, 1970). Panel A shows proteins dena- preserved whencTnC associates with cardiacmyofibrils, then tured in sample buffer with 8-mercaptoethanol. L a n e s 1-4 contain the data suggest a difference in the specificity of the Ca2+cTnC3,cTnC3(C84S),cTnC3(C35S),and cTnCS(A-Cys),respecdependent conformational change in the N-terminaldomain buffer without 8- of TnC thatis required for the regulation of ATPase activity tively. Panel B shows proteins denatured in sample mercaptoethanol. Lanes 1-4 contain cTnC3, cTnC3(C84S), or contraction in fast skeletal andcardiac muscle. cTnC3(C35S), and cTnC3(A-Cys), respectively. In panel C, lanes I Since Ca2+-dependent exposure of hydrophobic regions in respecand 2 contain cTnC3 with or without P-mercaptoethanol, 8- TnC andcalmodulin is thought tobe important for activation, tively. Lanes 3 and 4 contain cTnC3(A-Cys) without and with the fluorescent dyes 9-anthroylcholine (9-AC) and l,l'-bi(4mercaptoethanol, respectively.
-
-
~
. ~
6829
Ca2+-independentc T n C A. Fast Skel.tal Muscle Myoflbrlts
Dye: Protein:
-
Dithiothreitol
+
Activator
none
p
I
+ -
+ -
+ -
+ -
cTnC3
cTnC3 (C35S)
cTnC3 (C84S)
cTnC3 (A-Cys)
E. Cardlac Muscle
Myoflbrllr
b#s(ANS) cTnC3
bis(ANS) cTnC3(A-Cys)
9-AC cTnC3
9-AC cTnC3(A-Cys)
I 8
6
7
5
PCa
FIG.3. Effect of DTT on the interaction of fluorescent dyes with cTnC proteins. Panel A shows the effect of DTT on the interaction of 9-AC and his(ANS) with cTnC3 and cTnC3(A-Cys). The initial fluorescencewas set to zero with buffer alone in the cuvette. Sequential additions of probe, protein, DTT, and Ca2+ to pCa 4.5 were then made, and the relative fluorescence units were recorded. The results represent the average valuesfrom 4 experiments with bis(ANS) and2 experiments with 9-AC. Error burs indicate the + + + + + Dithiothreilol standard deviation. Panel B shows the effect of varying concentraActivator none cTnC3 cTnC3 cTnC3 cTnC3 tions of Ca2+ on the fluorescence from bis(ANS) in the presence of (WS) (A"@ (C3W cTnC3with ( 0 ) orwithout (0)DTTandcTnC3(A-Cys)inthe FIG.2. Functional characteristics of cTnC proteins in fast presence of DTT (A).The data represent the average of 3 separate skeletal muscle myofibrils (panel A ) and cardiac myofibrils titrations. The average initial fluorescence values at pCa 9 were 155, (panel B ) . Dialyzed cTnCproteins were incubatedwith (+) or 394, and 49 for cTnC3 + DTT, cTnC3 - DTT, and cTnCJ(A-Cys) without (-) 5 mM D T T at 22-25 'C for at least 20 min just prior to DTT, respectively. The data were fit to the following equation, F = the assay. Ca'+-independent activity (black burs) is superimposed on [Fmax(lj/(l + 1 0 n 1 ( i o K e a ( 1 l - l o 6 d )] + [F,,,czi/(l + 10n2('oKKCs(zi"o~a' 112 theactivityinthepresence of Caz+ (open burs). All assays were where F ,., n, and kCa define, respectively, the maximal change in error burs represent the standard devia- fluorescence, the Hill coefficient,and the Ca2+ concentration performed in triplicate. The which at tion for the triplicate determinations. the fluorescence change is half-maximal for the two components of the response. All fits gave a percenterror of less that 0.05. The low (PCBSoL)affinity anilino)naphthalene-5,5'-disulfonic acid (bis(ANS)) were calculated $aSO values for the high (pCaBOH) used to probe hydrophobic regions in cTnC3 and cTnC3(A- Ca2+-binding sites areshown in the lower left corner.
-
+
Cys), which exist predominantly assingle molecular forms in the presence and absenceof reducing reagents. Fig. 3A shows that the steady state fluorescent intensity from both 9-AC and bis(ANS) in the presence of dialyzed cTnC3 is significantly greater than for the reduced form of the protein. The presence or absence of D T T had no effect on the interaction between either probe and cTnCB(A-Cys). Although the addition of Ca2+ to thereduced forms of both proteins caused an increase influorescence, whichis consistentwith the exposure of hydrophobic surfaces, the magnitude of this increase is much less than the 24-fold increase in fluorescence from 9AC that is induced by the binding of Ca2+ to calmodulin (LaPorte et al., 1980). Fig. 3A shows that the magnitude of the steady state fluorescence frombis(ANS)associated with cTnC3(A-Cys) is much less than for cTnC3. This might be expected since the relative partition coefficients for Ser and Cys show Ser to be more hydrophilic. However,it may also reflect more pervasive structural changes that could affect Caz+-binding properties of the high and low affinity Ca2+-binding sites. To assess this, solutions of bis(ANS) and dialyzed cTnC3 or cTnC3(A-Cys) were titrated with Ca2+ in thepresence and absence of DTT. Fig. 3B shows that despite the difference in absolute magni-
tude of the steady state fluorescence frombis(ANS) associated with cTnC3 and cTnCS(A-Cys) at pCa 9, titration with Ca2+ in the presence of D T T resulted in anincrease in fluorescence of similar magnitude. Moreover, conversion of both Cys residues to Ser had little or noeffect on the &a,, values for the high and low affinity Ca2+-binding sites.DTT had no effect ontheCa2+titration curve for cTnC3(A-Cys)(datanot shown). Titrationof cTnC3 with Ca2+in the absence of D T T to preserve theintramolecular disulfide bondresultsin a decrease in fluorescence from bis(ANS) rather than the increase that is seen when the reduced protein is titrated with Ca2+.However, the &aBOvalues for the high and low affinity Ca2+-binding sites areunaffected by the presence orabsence of the intramolecular disulfide bond. DISCUSSION
In this study, we have presented evidence that intra- and intermolecular disulfide bonds can form between Cys-35 and Cys-84 of cardiac troponin C under mildly oxidizing conditions, and that these bonds confer specific functional peculiarities to the protein. Since seems it likely that the intracellular reducing potential would prevent formation of an intra-
6830
Ca2+-independent cTnC
molecular disulfide bond in cTnC in vivo, the significance of sTnC (Herzberg and James,1985; Herzberg et al., 1986) and these observations are their implications on the structure and cTnC (Krudy et al., 1992) suggests that dimerization via a molecular mechanism of action of TnC. bond in theregion of Gln-82 in sTnC or Cys-84 in cTnCcould The formationof an intramoleculardisulfide bond in cTnC occur only if helices B and C move away from helix D to in vitro is not necessarily surprising, since Fuchset al. (1989) expose these residues. The resulting structurewould resemble and Ingraham et al. (1988) have reported that the sulfhydryl two cupped hands clasped together at a 90" angle, in which groups of Cys-35 and Cys-84 are relativelyinaccessible to the fingers representthe B and C helices and the palms labeling reagents andmay be inclose proximity to each other represent thehydrophobic pocket. In this hypothetical model, in thehydrophobic core of the N-terminalglobular domain in the hydrophobic pockets would be occluded and unavailable cTnC. Although the crystal structure of cTnC has not been for interaction with troponin I. If the N-terminal hydrophobic pocket in TnC does not resolved, Cys-35 and Cys-84 are predicted tobe 18.2 and 21.2 A apart in the Ca2+-bound and apo forms, respectively, based provide a critical siteof interaction with TnI, itwould explain on a model that proved accurate in predicting the nuclear the behavior of severalcalmodulin antagonists on striated Overhauser effect (NOE) patterns for ring Phe protons Cin muscle contraction (Solaro et al., 1986; Kurebayashi et al., terminal domain of cTnC3 (Brito et al., 1991). Refinements 1988; El-Saleh et al., 1987). For example, calmidazolium is a to themodel that would allow for an intramoleculardisulfide potent calmodulin antagonist that also binds to sTnC. Hyside chains bond would most likely involve the repositioning of Cys-35 drophobic portions of the drug make contact with such that the hydrogen bonding patterns reported by Krudy of hydrophobic amino acids (Reid et al., 1990) that have et al. (1992) between the @-strandsof the active Ca2+-binding recentlybeenshown to be important for the binding of a myosin light chain kinase peptide tocalmodulin (Ikura et al., site I1 and the inactive siteI are allowed. A high degree of conformational flexibility in the N-termi- 1992). Thus calmidazolium most likely functions by binding nal domain of cTnC may allow the intramolecular disulfide to thehydrophobic pockets of calmodulin thereby preventing bond to form. Fig. 1 shows that both sulfhydryl groups in a productive interaction with target enzymes. Such a drug cTnC3 have sufficient conformational flexibility to interact mightbe expected to inhibit the ability of calmodulin to with each other in the same molecule, presumably in the regulate contraction in TnC-extracted skinnedmuscle fibers. interior core of cTnC, and at the surface of two different However, calmidazolium potentiates rather than antagonizes cTnC molecules. As noted by Krudy et al. (1992), there may the regulation of fast skeletal and cardiacmuscle fibers concalmodulin (El-Saleh et al., be differences in the relative degree and nature of conforma- taining either sTnC, cTnC, or 1987). tional flexibility intheN-terminaldomains of cTnC and sTnC. An analysis of the down field shifted amide protons in the NMR spectra of Ca2+-bound cTnC3 showed a singlestrong Acknowledgment-We thank Dr. Paul Rosevear for helpful discushydrogen bond between opposing @-strandsin the N-terminal sions. domain, which is absent in the apoform. In contrast, Tsuda REFERENCES et al. (1988)haveshown the presence of ahydrogen bond J. A,, Strynadka, N. C. J., James, M. N.G., and Brito, R. M. M., Putkey, between @-strands in the N-terminal domain of the apo form Rosevear, P. R. (1991) Biochemistry 30,10236-10245 El-Saleh, S. C., and Solaro, R. J. (1987) J. Biol. Chem. 262, 17240-17246 of sTnC but not in the Ca2+-bound form. Fabiato, A. (1988) Methods Enzymol. 167,378-417 The effect of theintramolecular disulfide bondonthe Fuchs, F., Liou, Y.-M., and Grabarek, Z. (1989) J. Biol. Chem. 264, 203443n.1ACl interaction of fluorescent dyes with cTnC3 is consistent with Fu'imori, K., Sorenson, M., Herzberg, O., Moult, J., and Reinach, F. C. (1990) anincreasein exposedhydrophobicsurfaces.However, hature 346,182-184 whether or not there is a causal relationship between this Graharek, Z., Tan, R.-Y., Wang, J., Tao, T., and Gergely, J. (1990) Nature 345,132-135 phenomenon and the Ca2+-independent activity is notclear, Herzberg, O., and James,M. N. (1985) Nature 313,653-659 O., Moult, J., and James, M. N. (1986) J. Biol. Chem. 261, 2638since conversion of both Cys residues to Ser also causes a Herzberg, 3644 significant change in fluorescence from bound dyes but with Ikiia,"., Clore, G. M., Gronenhorn, A. M., Zhu, G., Klee, C. B., and Bax, A. (1992) Science 266,632-638 no effect on function. It is interesting that the intramolecular Ingraham, R. H., and Hodges, R. S. (1988) Biochemistry 27,5891-5898 disulfide bond in cTnC3 has little if any effect on the Ca2+- Johnson, J. D., Charlton, S. C., and Potter, J. D. (1979) J. Biol. Chem. 254, 3497-3502 binding propertiesof site 11. If Ca2+-independent activity were Krudy, G. A,, Brito, R. M. M., Putkey, J. A., andRosevear,P. R. (1992) Biochemistry 31, 1595-1602 due to a conformation that resembled the Ca2+-bound form Kunkel, T.A. (1985) Proc. N ~ t lAcad. . Sci. U. S. A. 8 2 , 488-492 of cTnC3, then itmight be anticipated that the Ca2+-binding Kurebayashi, N., and Ogawa, Y. (1988) J. Physiol. 403,407-424 affinity of site I1 would be increased by formation of the Laemmli, U. K. (1970) Nature 227,680-685 LaPorte, D. C., Wierman, B. M., and Storm, D. R. (1980) Biochemistry 19, disulfide bond. This prediction is consistentwith the data of ~ Q L I ""_" ~ aI Pearlstone et al. (1992), who have shown that the affinityof Pearlstone, J. R., Borgford, T., Chandra, M., Oikawa, K., Kay, C. M., Herzherg, O., Moult, J., Herklotz,A,, Reinach, F. C., and Smillie,L. B. (1992) Biochemsites I and I1 in sTnC is increasedby 0.2 to 0.42 pCa units by istry 31,6545-6553 J. A., Sweeney, H. L., and Campbell, S. T.(1989) J. Biol. Chem. 264, changing hydrophobic residues in the N-terminal domain to Putkey, 12.170-1 2378 Thr or Gln, which presumably forces the helices B and C Putkey, J. A,, Lui, W., and Sweeney, H. L. (1991) J. Biol. Chem. 266, 1488114884 awayfromhelix D to attain a conformation that is more Putkey, J. A,, Dotson, D. G., and Mouawad, P. (1992) FASEB J. 6, A281 similar to the Ca2+-bound state. Reid, D. G., MacLachlan, L. K., Gajjar, K., Voyle, M., King, R. J., and England, P. J. (1990) J. E d . Chem. 265,9744-9753 Perhaps the most intriguing aspectof these results is that Robertson, S. P., Johnson, J. D., and Potter, J. D. (1981) Biophys. J. 34,5595G4 dimerization of cTnC via a disulfide bond between Cys-84 "" Z., Strauss, W. L., Francois, J. M., and Potter, J. D. (1990) J. B i d . residues confers significant Ca2+-independent activity when Sheng, Chem. 265,21554-21560 assayed infastskeletal muscle myofibrils. Cysteine-84 in Solaro, R. J., Bousquet, P., and Johnson,J. D. (1986) J . Phrmacol. Exp. Ther. 238,502-507 cTnC is predicted occupy to the samerelative position as Gln- Sweeney, H. L., Brito, R. M. M., Rosevear, P. R., and Putkey,J. A. (1990) Proc. Natl. Acad. Sci. U. S. A . 87,9538-9542 82 in helix D at the entrance to the hydrophobic pocket in S., Hasegawa, Y., Yoshida, M., Yagi, K., and Hikichi, K. (1988) BiosTnC. An inspection of the crystal structuresfor the apo and Tsuda, chemistry 27,4120-4126 Zot, H. G., and Potter, J. D. (1982) J . Biol. Chem. 267, 7678-7683 predictedCa2+-boundforms of theN-terminaldomain of
"_"
uyI-
