The Reactivity of Sulfhydryl Groups of Bovine Cardiac Troponin C*

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 264, No. 34, Issue of December 5, pp. 20344-20349,1989 Printed in U.S.A.

The Reactivity of Sulfhydryl Groups of Bovine Cardiac TroponinC* (Received for publication, June 21, 1989)

Franklin FuchsS, Ying-Ming Lious, andZenon Grabarekg From the $Department of Physiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 and the §Department of Muscle Research, Boston Biomedical Research Institute, Boston, Massachusetts 02114

Bovine cardiac troponin C (cTnC) contains 2 cysteine ponin C (sTnC)’ and cardiac troponin C (cTnC) have some residues, Cys-35 located in the nonfunctional Ca2+- notable differences which can be exploited for purposes of binding loop I and Cys-84 in the N-terminalsegment experimental analysis. cTnC (like sTnC) nominally has four of the central helix. We have studied the reactivity of ea2+binding sites, I and I1 located in theN-terminal domain Cys residues in cTnC with 5,5’-dithiobis(2-nitroben- of the polypeptide chain and I11 and IV located in the Czoic acid) (DTNB) and 7-diethylamino-3-(4’-maleimi- terminal domain (2). Some amino acid substitutions at site I dylphenyl)-4-methylcoumarin(CPM). The latter com- of cTnC have rendered that sitenonfunctional (2). Thus, pound fluoresces only when reacted with the protein. cTnC, at saturation,binds three Ca2+ions, as opposed to four The reaction withDTNB followed second order kinet- for sTnC (3, 4). Sites I11 and IV bind Ca2+with high affinity ics with respect DTNB, to the rateconstants being3.37 ( K -2 X lo7 M - ~ ) and also bind Mg2+ ( K -lo3 M-’). Site I1 s” M-’ and 1.82 s” M” in thepresence and absence of binds Ca2+with lower affinity (K -lo5 M-’) but the binding Ca”, respectively. These rates are much slower than is highly specific (3,4). Based on studies relatingmyofibrillar the rateof reaction with Cys-98 of skeletal TnC (sTnC) ATPase activity (4) and skinned fiber force generation (5, 6) or with the urea-denatured cTnC, indicating that bothto free and bound Ca2+,there is general agreement that only Cys residuesare partly buried within the structure of site I1 is regulatory for cross-bridge activation. Under physthe protein. The increase in reactivity was induced by iological conditions, sites 111and IV would always be occupied binding of Ca2+to the singlelow affinity Ca2+binding by Mg2+ or Ca2+. site (site 11). The fluorescence increase upon reaction The three-dimensional structure of cTnC is unknown, but of cTnC with CPM in theabsence of Ca2+could be fitted based on the similarities in sequence homology and physiowith a single exponential equation indicating that both logical function, it seems reasonable to assume that cTnC and cysteine residues are equally available to the reagent. sTnC have a similar structure. In thex-ray diffraction-derived The reaction in the presence of Ca2+ was biphasic. structure of chicken (7) and turkey (8) sTnC, the molecule Analysis of CNBr fragments of cTnC labeled withCPM consists of two globular domains connected by a 29 residue long central helix. The N-terminal domain contains the low under various conditions indicated that in the presence C-terminal domainthe Ca2+of Ca” the reactivityof Cys-84 is increased whilethat affinity “triggering” sites and the of Cys-35 is slightly decreased. This findingis consist- M P sites. In thecrystal only the high affinity sitesare filled ent with the model of Herzberg et d.(Herzberg, O., with Ca2+.The differences in relative disposition of helical Moult, J., and James, M. N. G. (1986)J. Biol. Chem. segments flanking the Ca2+-bindingloops in the N-terminal Ca2+-freedomain as compared with the C-terminal Ca2+-filled 261, 2638-2644) andthedata of Ingrahamand Hodges (Ingraham, R. H., and Hodges, R. S . (1988) sites led Herzberg et al. (9) to propose that the Ca2+-induced conformational transition in the N-terminal domain, which Biochemistry 27, 5891-5898), suggesting that the is believed to be the triggering event, is such that thisdomain Ca2+-induced conformational change in the N-terminal half of TnC involves separationof the helix C fromthe acquires the conformation characteristic for the C-terminal of known central helix, thereby increasing the accessibility of domain and for the otherCa2+-filledsites in proteins Cys-84. The slow overall kinetics, however, indicates three-dimensional structure. Such atransition would require that the structure in the vicinity of Cys residues is a shift in relative positions of two helical segments (helices B relatively compact regardless of Ca2+.We interpret the and C) with respect to the central helix which would make increase in reactivity towardsCPM as consistent with some residues in the central helix more accessible to solvent. Cardiac TnC has2 Cys residues (2). One, Cys-35, is located a Ca2+-induced exposure of a hydrophobic pocket in in inoperative site I and the other, Cys-84, is located in the the vicinityof Cys-84. N-terminal end of the central helix, the region expected to become exposed to solvent when Ca2+ is bound to site 11. Thus, by measurement of the reactivity of Cys residues in cTnC, it should be possible to test thehypothesis of Herzberg et al. (9). The contraction of cardiac muscle, like that of skeletal In thispaper we report onthe effects of Ca2+on the kinetics muscle, is activated by the binding of Ca2+to troponin C (1). Despite similarities in structure and function, skeletal troThe abbreviations used are: sTnC, skeletal troponin C; cTnC, *This work was supported by Grants AR-10551 and R-37-HL05949 from the National Institutes of Health and by the American Heart Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

cardiac troponin C; DTNB, 5,5’-dithiobis(2-nitrobenzoicacid); CNBr, cyanogen bromide; CPM, 7-diethylamino-3-(4’-maleimidylphenyl)-4-methylcoumarin; DTT, dithiothreitol;EGTA,[ethylenebis(oxyethylenenitrilo)]tetraacetic acid; HEPES, 4-(2-hydroxyethyl)1-piperazineethanesulfonic acid; PIPES, 1,4-piperazinediethanesulfonic acid; MOPS, 3-(N-morpholino)propanesulfonicacid MES, 4morpholineethanesulfonic acid; SDS, sodium dodecyl sulfate.

20344

Sulfhydryl Groups of Troponin CCardiac of the reaction of cTnC with two SH-reactive compounds. These are: 1)the standard SHreagent 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) (10) and 2) 7-diethylamino-3-(4'-maleimidylphenyl)-4-methylcoumarin(CPM) (11).CPM is fluorescent only when it is covalently linked to an SH group; thus, thereaction rate can be easily monitored with a spectrofluorimeter. Furthermore, the fluorescent label can be localized on polyacrylamide gels followingelectrophoresis of fragmented cTnC. It hasbeen found that both SH groups of cTnC are relatively inaccessible to these reagents. However, the rate of SH labeling is increased upon the binding of Ca2+to site 11, and this effect seems to be largely accounted for by increased exposure to solvent of Cys-84. MATERIALS AND METHODS

Preparation of Cardiac Troponin C-cTnC was prepared from fresh cow hearts obtained at a nearby slaughterhouse. The purified protein was extracted according to Potter (12) and, in later studies, Szynkiewicz et al. (13). Both preparations gave identical results. Comparative studies were also carried out with sTnC prepared according to Potter (12). Preparation of Reduced cTnC-Freeze-dried cTnC was dissolved in 8 M urea, 20 mM HEPES (pH 7.5), 2 mM EDTA, 10 mMKC1, and 5 mM DTT. After incubation a t room temperature for 1-2 h, the sample was dialyzed in the cold for several hours against a solution containing 100 mM KCI, 10 mM MES (pH 6.2), and 2 mM EDTA. Aliquots of the cTnC solution were then passed through short columns (0.9 X 15 cm) of either Sephadex G-25 or Bio-Gel P-10 equilibrated with 50 mM KCl, 100 mM PIPES (pH7.01, and 2 mM EGTA. For some experiments 100 mM MOPS (pH 7.0) was substituted for PIPES buffer. The nature of the buffer did not influence the experimental results. One-ml fractions were collected from the column, and these were analyzed for both protein concentration and SH content. Protein concentration was determined on the basis of an extinction coefficient (1mg/ml, at 276 nm) of 0.3 in the presence of EGTA (13). The SH content was determined by means of the Ellman reaction (10) with 6 M urea present. In general, for samples of cTnC prepared as described above, the SH content was in the range of 1.4-1.9 mol of SH/mol of cTnC. Reagents and Solutions-DTNB (Sigma) was dissolved in 10 mM phosphate buffer (pH 7.5) at concentrations of2-5mM. Solutions were stored in the refrigerator and prepared fresh every month. CPM (Molecular Probes) was dissolved (concentration 2 mM) in dimethyl formamide and stored in a dark containerat -20 "C. Fresh solutions were prepared weekly. Ca2+concentrations were calculated on the basis of absolute binding constants tabulated by Fabiato and Fabiato (14). Spectrofluorimetry-Fluorescence measurements were made in the ratio mode at a temperature of25 "C with either a Perkin-Elmer MPF-4 or a SPEX Fluorolog 2 spectrofluorimeter. Kinetic Measurements-The rate of labeling of cysteine residues was measured for the reaction of cTnC with DTNB and with CPM. The time course of the absorption changes at 412 nm upon reaction with DTNB was recorded with a Perkin-Elmer Lambda 3A spectrophotometer, and the time course of fluorescence changes upon reaction with CPM was recorded using a Perkin-Elmer MPF-4 spectrofluorimeter. To obtain apparent rate constants, the data were digitized with a Tektronix4600 interactive plotter interfaced with a DEC PDP-11/44 computer andfitted with a mono- or bi-exponential equation using the Marquardt algorithm. Alternatively, when the reaction was known to proceed monoexponentially, the initial slopes were compared as a relative measure of reaction rates (e.g. Fig. 3). Localization of CPM Label-Analysis of specificity of labeling of the 2 Cys residues under various conditions was performed on CNBr digests of labeled cTnC. The high absorption coefficient and the high quantum yield of CPM facilitates localization of labeled peptides on polyacrylamide gel. The large difference in molecular weight of Cyscontaining CNBr peptides facilitates their identification; Cys-35 is in a 44-residue long fragment (residues 2-45) and Cys-84 is in tetrapeptide containing residues 82-85. For complete labeling of cTnC with CPM, a sample of cTnC (0.5 mg) was dialyzed against a solution containing 50 mM NH4 HCOs, 2 mM EDTA, and 2 mM DTT followed by dialysis against a solution containing 50 mMKC1, 100 mM PIPES (pH 7.0), and 2 mM EGTA. CPM in dimethyl formamide was added in four portions at lo-min

20345

intervals to make the final molar ratio of CPM: protein equal to 2:l. After further incubation at room temperature for 1.5 h, 2 mM DTT was added to inactivate the unreacted CPM, and the sample was passed through a Sephadex G-25 column equilibrated with 50 mM NH,HCO,. The fluorescent protein fraction was collected, lyophilized, and digested by overnight incubation with 0.2 ml of l M CNBr in 70% formic acid. The sample was lyophilized and run on a 20% polyacrylamide gel in the presence of 0.1% SDS. Two major fluorescent bands were cut out and the peptide extracted with 0.1% SDS and passed through a Bio-Gel P-2 column equilibrated with 10 mM NH,HCO,. Peptide fractions were collected and lyophilized. Amino acid analysis of the upper band verified the presence of cTnC fragment 2-45. The sample corresponding to the lower band was further purified on a reverse phase high performance liquid chromatography synchropak C-18 column. Fractions absorbing at 387 nm were analyzed on an amino acid analyzer. A high yield of valine and arginine verified the presence of peptide fragment 82-85. To study the effect of Caz+on the relative incorporation of CPM into Cys-35 and Cys-84, samples of cTnC (8-10 pM) were mixed with 40 p M CPM in a solution containing 50 mM KCl, 100 mM PIPES (pH 7.0), and either 2 mM EGTA (-CaZ') or 2 mM EGTA 2.1 mM CaClz (+Ca2+).The reaction was stopped after 2 min by the addition of excess DTT. Samples were then subjected to CNBr digestion and electrophoretic analysis as described above.

+

RESULTS

Reaction of cTnC andsTnC with DTNB-The general features of the reaction of cTnC with DTNB are illustrated in Fig. 1. Although there aretwo SH groups in cTnC thatcan react with DTNB, the time course is monoexponential with or without Ca2+.At pCa 8.0, with a DTNB concentration of 200 pM, the reaction proceeds quite slowly (t, -25 min). On the other hand, the reaction is complete within 2 min in the presence of 6 M urea. Hence, both SH groups of cTnC must be in a "buried" location where they are notreadily accessible to solvent. Accessibility is enhanced by the addition of sufficient Ca2+to saturate all of the binding sites. The combined data from several such experiments are plotted in Fig. 2. For this figure the initial slopes were normalized to the slopes at pCa 8.0. Reaction rate was increased more than 2-foldby Ca2+,with half-maximal effect being observed at pCa -5.4. Thus, the effect of Ca2+on the DTNB reaction rate is mediated by the binding of Ca2+to site 11. When the reaction rate was measured as afunction of DTNB concentration, cTnC concentration being held constant, the data shown in Fig. 3 were obtained. The apparent first order rate constant was alinearfunction of DTNB concentration up to 1 mM, both in the presence and absence

0.060--

0.040 --

0

5

IO

15

20

25

30

3

Time [rnin]

FIG. 1. Reaction of cardiac troponin C (cTnC) with DTNB at low (pCa 8.0) (0)and high Ca2+ (pCa4.5) (A) concentration, as indicated, and when denatured with 6 M urea (m). Solutions contained 2.2 p M cTnC, 100 p M DTNB, 50 mM KCI, 100 mM PIPES (pH 7.0), 2 mM EGTA.

Sulfhydllyl Groupsof Cardiac Troponin C

20346 2.400

I

7

8

6 PC0

5

TABLE I Second order rate constants of DTRR r e a c t ~ nwith troponin C from bovine cardiac [c?hC) and rabbit skeletat muscle [sTnC, Conditions: proteinconcentration2 pM, 50 mM KCl, 100 mM PIPES (pH 7.01, 2 mM EGTA, and 2.1 mM CaC12 as indicated, temperature, 25 "C. +EGTA +CaCl, k (s" M") 3.37 cTnC 1.82 40.2 sTnC 442

4

FIG.2. Rate of reaction of DTNB with cTnC as a function of pCa. Rates are expressed as a ratio of initial slope a t a given pCa to the initial slope at pCa 8. Solutions contained 2.0 p M cTnC, 200 FM DTNB, 50 mM KCl, 100 mM PIPES (pH 7.0),2 mM EGTA, and varying Ca2+concentrations. Vertical bars indicate standard error. Each point is mean of three to four determinations. 1.4

'7

" 1.2

8d

0.6

fc

0.4

i

S

0

0.2

0.4 [DTNB]

0.6

0.8

1

(mM)

I

I

I

I

I

440

460

480

500

520

~ a v e ~ e n g t~hn m )

FIG.4. Uncorrected fluorescence emission spectra of CPM-

labeled cTnC. A sample of cTnC was allowed to react with excess of CPM for 24 h and then dialyzed exhaustively. Additions of M P and Ca2+are as indicated. Solution contained 1.3 p M cTnC, 50 mM KCl, 100 mM MOPS (pH 7.0), 2 mM EGTA. Excitation wavelength, 387 nm. Curve 1, M e = 0 mM, pCa = 8; curve 2, Mg2f = 5 mM, pCa = 4; curve 3, Mg2' = 5 mM, pCa = 8.

some secondary monomolecular transitions, it is rather surprising that neither one of the Cys groups was preferentially reaction of DTNB with cTnC and sTnC on the concentration labeled. In search of possible differences in behavior of the 2 of DTNB. Solutions contained 2 FM cTnC, 50 mM KCI, 100 mM cysteines, we have employed a coumarin maleimide derivative PIPES (pH 7.0), and either 2 mM EGTA or 2 mM EGTA + 2.1 mM ~ l CaCI2, as indicated. Reaction proceeded a t 27 'C. Time courses of (CPM), a reagent in which the reaction with s u l ~ ygroups absorption change a t 412 nm were digitized and fitted with mono- is based on a different principle than that of DTNB. The exponential equation to obtain numerical values of the apparent rate fluorescence emission of CPM depends upon the formation constants. 0, sTnC (-Ca*'); 0 , cTnC (-Ca2+); 0,sTnC (+Ca2+);x, of a covalent attachment between an SH group and the cTnC (+Ca2'). maleimide ring (11). Hence, the fluorescence enhancement can be used to monitor the time course of the reaction. An of Ca2+.The reaction rate was increased about 2-fold in the advantage of CPM is that a stable covalent link between the probe and thecysteine side chain enables identification of the presence of Ca". It is of interest to compare these resultswith those obtained labeled residue. In measuring the kinetics of the reaction one with sTnC. The latter has a single Cys residue, located at should remember, however, that the fluorescent signal of the position 98 (15). Earlier work by Potter et al. (16) had first probe is affected by its environment; thus, it may or may not demonstrated that the reactivity of DTNB with Cys-98 was be a quantitative measure of the extent of reaction. On the inhibited by Ca2+.As shown in Fig. 3, not only is reaction rate other hand the environmental sensitivity of the probe may of cTnC. reduced by Ca", but with or without Ca2+, the reaction prove valuable in studying conformational properties proceeds much faster than is observed with cTnC. Cys-98 is Thus, it is important to first determine the effects of metal at the C-terminal end of the centralhelix and obviously more binding on fluorescence emission from the modified cTnC. Fig. 4 shows the fluorescence emission spectra obtained exposed than either Cys-35 or Cys-84 in cTnC. When the second order rate constants were calculated for both cTnC with a sample of cTnC fully reacted with CPM. The excitation and sTnC, theresults shown in Table I were obtained. These wavelength was 387 nm. In theabsence of any added divalent data emphasize the differences between sTnC andcTnC with cations (pCa 8.0, no M$+), there is an emission maximum at respect to SH reactivity. Thus, in the absence of Caz+ the about 470 nm (curve I). The addition of 5 mM M$+, at pCa -8.0, causes an approximately 20% reduction in fluorescence, DTNB reaction proceeds 240 times as fast with sTnC as compared with cTnC, whereas in the presence of Ca2+this with a small red shift (curue 3 ) . Addition of Ca2+to pCa 4 produces a small increase in fluorescence, relative to thatseen difference is reduced to 12-fold. Reaction of cTnC with CPM-Although the second order in the presence of MgZ+ alone, but ingeneral the fluorescence type kinetics of DTNB reaction with cTnC suggests that the was still 5-10% lower than in theabsence of divalent cations rates represent the accessibility of Cys residues rather than (curue 2). A virtually identical curve was obtained upon the

FIG.3. Dependence of pseudo-first order rateconstants for

C

Sulfhydryl Groups of Troponin Cardiac addition of saturating concentrations of Ca2+ in the absence of Mg2+.When cTnC was titrated with Ca2+ in the absence of M$+, the fluorescence decreased in the pCa range (8.0-6.5) in which the Ca2+-Mp2+sites were occupied and then increased on further addition of Ca2+ (not shown). The most straightforward interpretation of these results is that occupation of the Ca'+-M$+ sites (by either Ca2+M$+) or causes a decrease in fluorescence, whereas occupation of site I1 causes a small increase influorescence. The fact that Mg" caused asubstantial decrease in fluorescence suggeststhat, at least in solution, there mustbe somedirect interactionbetween the N-terminal andC-terminaldomains of cTnC. A similar conclusion emerges from the recent report of Verin and Gusev (17) on the monomer andexcimer fluorescence of cTnC labeled with N - (1-pyrene)maleimide. Byvarying Ca2+concentration in the presence of 5 mM M$+, the effect of Caz+ on SH reactivity can be measured, since under these conditions the effect of Ca2+ onfluorescence emission itself is quite small. Fig. 5 shows the results of an experiment in which CPM was reacted with cTnC in the absenceof Ca2+ (or M F ) and in the presence of a saturating concentration of Ca'+. In the presence of Ca2+ therewas a biphasic curve characterized by a fast initial rate of reaction which,after 1-2 min, was followed by a much slower second phase. Thus, although the rate in the absence of Ca2+ (pCa 8.0) was initially slower than that in its presence, after about 20 min the fluorescence emission reached a plateau at a value which was roughly twice that seen in thepresence of Ca2+.If at this time the Ca2+ concentration was quickly reduced by addition of 10 mM EGTA (final pCa -7.0) fluorescence increased and converged with the value obtained in the absence of Ca2+. The time course obtained a t pCa 8.0 could be fitted with a monoexponential equation with an apparent rate constant k = 0.070 min". The time course in the presence of Ca'+ (pCa 4.0) required a two exponential fit with kl = 0.65 min" and kz = 0.026 min" and the relative amplitudes of 0.42 and 0.58 for the fast and slow transition, respectively. It would appear that Ca2+ increased the reactivity of one of the SHgroups to CPM butdecreased the reactivity of the other group. No dependence of the rate of reaction on the concentration of CPM could be detected, a n observation suggesting first orderkinetics. The effect on the initial reaction rate of varying the CaZc concentration in the presenceof 5 mM M P is shown in Fig. 6. In the pCa range 8.0-6.5, where Ca2+ would be displacing M e from sites I11 and IV, there was no detectable effect of Caz+ on S H reactivity. However the reaction rate increased

L t

20347

4.0 5.0

5.5 6.0

0 H

6.5

l-

a

t

e.0

0.00000

I

0

'

I

I

I

I

50

100

150

200

Time [secl

FIG.6. Effect of Ca2+oninitial rate of reaction of CPM with cTnC. Solutions contained 0.5 p M cTnC, 20 p M CPM, 50 mM KCl, 100 mM MOPS (pH 7.0), 5 mM MgC12, 2 mM EGTA, and varying additions of Ca2+. Excitation wavelength, 387 nm;emission wavelength, 470 nm.

".i, *..._

'......._ .....,...,

FIG.7. Localization of CPM labeling sites in cTnC. CNBr digests of samples of cTnC labeled with CPM were analyzed using SDS-electrophoresis on 20% polyacrylamide gel. The gel was photographed in UV light to visualize only the bands carrying thefluorescent probe. Both the photograph and the fluorescent scansshown. are L and H indicate the low molecular weight (residues 82-84) and the high molecular weight (residues 2-45) fragments, respectively. Intact cTnC labeled with CPM (tracka). CNBr digest of cTnC labeled with CNBr digest of cTnC CPM in the absence of Ca2+ (track b, -). labeled in the presence of Ca2+ (trackc, . . . .). CNBr digest of cTnC fully labeled with CPM in the presence of 6 M urea (track d, - - -). The large off-scale peak in the scan is the unreactedlabel migrating with the dye front.

6z 1

IY p

-

._ 1'5 TIME ,min

I

20

25

30

FIG.5. Effect of Ca2+on time course of fluorescence enhancement following addition of CPM (20 PM) to cTnC (1 p ~ ) Solutioncontained 50 mM KCl, 100 mM PIPES (pH 7.O), 2 mM EGTA. Excitationwavelength, 387 nm; emission wavelength, 465 nm. Addition of 10 mM EGTA increased pCa from 4.0 to -7.0.

.

markedly as the [Ca"] was increased into the range inwhich site I1 would be titrated. Localization of CPM Label-By combining CNBr digestion with SDS-polyacrylamide electrophoresis it was possible to assess the incorporation of fluorescent label into Cys-35 and Cys-84. Due tolocation of methionine residues, Cys-35would be part of peptide 2-45 and Cys-84 would be part of peptide 82-85 (Val-Arg-Cys-Met). These can be separated easily on 20% polyacrylamide gels. Fig. 7 shows the results of an experiment illustrating the effects of Ca2+ on the fluorescent labeling of the cTnC.Amino acid analyses of the two fluorescent bands areshown inTable 11. The column labeled "high M , band" shows the measured amino acid compositionof the more slowly moving fluorescent band. As shown, there is excellent agreement with the expected amino acidcomposition of peptide 2-45. However,

~ u ~ f h yGroups d ~ l of Cardiac T r o ~ ~ n C in

20348 TABLEI1

the kinetics and Ca” dependence of SH labeling. An important point which emerges from this study is that both SH groups of cTnC are well “hidden” from the solvent, a property which might be expected from the crystal structure. By measuring the SH reactivity of cTnC and sTnC it is possible to make a direct comparison of the surface accessibility of cys-98, located at the C-terminal end of the sTnC central helix, with the surface accessibility of Cys-35 and CysAmino acid High M, bandcTnC (2-45) Low M,bandcTnC (82-85) 84 in cTnC. As first shown by Potter et ab. (16), and confirmed mo1/100 mol here, Cys-98 of sTnC is highly reactive. This reactivity is ASP 12.0 11.9 6.2 inhibited by the binding of Ca2+(or M e ) to thehigh affinity Thr 2.4 4.5 4.8 sites (16). In the case of cTnC, the SH groups are relatively Ser 8.2 2.9 2.4 nonreactive, but reactivity is stimulated by the binding of Glu 7.018.8 19.0 Gly” 10.7 7.1 Ca” to the single low affinity site, Although both CPW and Ala 10.9 11.9 4.1 DTNB exhibited a Ca2+-dependentincrease of reaction rate CYS 0.4 O.Ob 3.4 O.Ob (Figs. 2 and 6), in the case of CPM it was possible to show Val 7.3 7.1 16.1 50.0 that the binding of Caz+to site I1 caused a selective increase Met 0.4 o.oc 0.0“ in Cys-84 labeling. Ile 5.9 7.1 1.4 Cys-35 is located in Ca2+binding site I (residues 28-40). Leu 3.1 7.3 7.1 Aminoacid substitutions in this domain have rendered it Tyr 0.9 2.4 Phe 6.4 7.1 0.3 nonfunctional as a Ca’+ binding site andone would not expect His 0.2 0.0 3.4 Ca2+to have a direct effect on the reactivity of Cys-35 or on LYS 11.9 10.7 the behavior of fluorescence probes attached to Cys-35. Cys‘4% 0.6 0.0 14.5 50.0 84, however, occupies an important position at theN-terminal The low M, sample had a high level of GIy, a ont tam in ant from end of the centra1 helix. On the one hand, it is close to the the Tris-Gly electrophoretic buffer. Gly was not taken into account physiological regulatory site (site 11). On the other hand it is in the calculation of amino acid composition of this sample. adjacent to the stretch of amino acid residues (89-loo), which, Expected to be in form of CPM derivative. Not analyzed. Expected to be converted to homoserine upon CNBr digestion. at least in sTnC, is a region of contact between troponin C Not analyzed. and troponin I (19-23). Thus, it is very likely that conforCNBr cleavage produces several short peptides which would mational changes taking place in this region are closely conbe expectedto comigrate with peptide 82-85. Thus, thefaster nected to the regulatory function of cTnC. It is of interest moving fluorescent band had measureable amounts of most that in the work of Ingraham and Hodges (18)Ca’+ stimulated of the amino acids. Nevertheless, the high contents of valine the iodoacetamide labeling of Cys-84 both in isolated cTnC and arginine are consistent with the assumption that the and in cTnCwhich was part of the troponin complex. Although our present data are in a qualitative agreement fluorescent species in this band was peptide 82-85. In the presence of a saturating concentration of Ca2+,there was an with the model of Herzberg et al. (9) and with the data of approximately 50% increase in labeling of Cys-84 and a small Ingraham and Hodges (X+),several new features emerge that decrease in labeling of Cys-35, This patternis consistent with require thorough consideration. We have used two different the biphasic reaction kinetics illustrated in Figs. 5 and 6 and labels, and there is an apparent difference in the kinetics of identifies Cys-84 as the residue labeled during the initial fast the reaction depending on the probe. We will address separately the implications of each set of data. phase of reaction. All the time courses obtained in the reaction of cTnC with DTNB couldbe fitted with a single exponential equation. DISCUSSION Since there are two SH groups in cTnC, this observation may One ofthe outstanding problems of muscle function which indicate that the Cys residues are equally accessible to the still awaits solution is the nature of the molecular “switch” reagent. Similar results could be obtained if the labeling was which links the binding of Ca2+to troponin C to theactivation cooperative, i e . the reaction of the second Cys residue in the of cross-bridge cycling, A key event must be a conformational molecuIewasmuch faster after the first one was already change in troponin C which initiates a signal that is trans- labeled. At present we cannot d i s t i n ~ i s hbetween these two mitted to the other members of the regulatory protein com- models. We also did not attempt to analyze whether any of plex. The recent solution of the crystal structure of sTnC (7, the residues is preferentially labeled with DTNB. Since the 2 8) constitutes a major step forward in our understanding of cysteines can form a disulfide bridge and can be cross-linked Ca2+activation, since it is now possible to formulate structural with dimaleimides,’ it is likely that a possible disulfide exmodels whichare amenable to experimental testing. Herzberg change reaction occurring after the initial labeling of any of et al. (9) have proposed a model of Ca2+activation based on the SH groups would result in a random distribution of the the assumption that the binding of Ca2+to the N-terminal label. The linear dependence of the pseudo-first order rate domain causes its structureto resemble that of the C-terminal constant on the concentration of DTNB indicates that the domain with divalent cations bound at sites I11 and IV. This reaction with DTNB follows second order kinetics. Thus, the alteration in the structure of the N-terminal domain would numerical value of the second order rateconstantsmust entail separation of helix C from the central helix and would represent the relative accessibility of Cys residues for an expose Gln-85 (in turkey sTnC) to the solvent. This model effective collision with DTNB. The approximately %fold incan be tested by monitoring the reactivity of cysteine in cTnC, crease in the second order rate constantupon Ca2+binding to since Cys-84 is the residue on bovine cTnC which corresponds site 11, although in qualitative agreement with the prediction in position to Gln-85. The results obtained in this study, of Herzberg et al. (9), is much smaller than one might expect taken together with the recent work of Ingraham and Hodges from the difference in the relative solvent accessibility of Gln(18), provide support for this model. Our study adds to that F. Fuchs, Y-M. Liou, and Z. Grabarek, unpublished observations. of Ingraham and Hodges in that further data are provided on Amino acid analysis of CNBr fragments of cTnC carrying the fluorescent probe CPh4 The columns labeled “high M, band” and “low M, band” show measured amino acid composition of slow-moving and fast-moving fluorescent bands, respectively. The columns “cTnC (2-45)” and “cTnC (82-85)” show expected amino acid composition of peptides 2-45 and 82-85, respectively.

29.8

Sulfhydryl Groups of Cardiac Troponin C 85 in turkey sTnC. The residue in a homologous position in cTnC is Cys-84. The solvent-accessible surface of Gln-85 should, according to Herzberg et al. (9), increase 500-fold upon Ca2+binding, from 0.1 to 50 A'. One possible explanation forthe difference between the experimental resultand the theoretical prediction may be that the DTNBmolecule is relatively large as compared with H20; thus, theaccessibility for DTNB and thesolvent accessibility may not be the same. However, Ingraham and Hodges (18)used the much smaller iodoacetamide and obtained an approximately 2-fold increase in the extent of reaction, a result consistent with our data rather thanwith the theoretical prediction. Another possibility is that some structural differences between the bovine cTnC used in our experiments and theturkey sTnC make the reactivity of Cys residues in cTnC less Ca2+-sensitive. Another important aspect of our results is that aside from Ca2+sensitivity the rates of cTnC reaction with DTNB are very slow. The reaction with Cys-98 of sTnC is 240 and 12 times fasterthan thatwith Cys-35 and Cys-84 of cTnC in the absence and presence ofCa", respectively. Since the Cterminal domain, and in particular the vicinity of Cys-98 in sTnC, is essentially unfolded in the absence of Ca'+ (24), the 442 s-' M" rate constantobtained here is most likely close to the maximal reaction rate, under the given experimental conditions, with a fully exposed Cys residue. Based on our results we conclude that Cys residues in cTnC are relatively inaccessible to DTNBwith or without Ca2+. The kinetics of reaction of cTnC with CPM is different from that of DTNB andreflects differences in the mechanism of reaction between the two reagents. The time course for the reaction in the absence of Ca2+could be fitted with a single exponential equationand the apparent rate constants did not depend on the concentration of CPM in the range of 2-40 PM. On the otherhand,all the time courses obtainedin the presence of Ca2+required a two exponential fit. Neither of the two rate constantsdepended, however, on the concentration of CPM. We conclude that the reaction with CPM is essentially first order with respect to CPM. The faster initial rate in the presence of Ca2+most probably results from an increase in noncovalent binding of CPM to cTnC preceding the covalent attachment of the SH groups to the maleimide ring of CPM. Presumably, it is that second step that results in the increase of fluorescence. Since we have identified Cys84 as the residue preferentially labeled by CPM in the presence of Ca", our data suggest that with Ca2+bound to site I1 there is a hydrophobic pocket adjacent to Cys-84 that can accommodate the hydrophobic coumarin moiety. The formation of such a hydrophobic pocket has been suggested previously (25) and is also consistent with the model of Herzberg et al. (9) and with the known requirement for hydrophobic residues in peptides interacting with troponin C (26,27). Ca'+ binding to site I1 has very little effect on the fluorescence of CPM-labeled cTnC. Much larger fluorescence enhancement was noted by Johnson et al. (28) in the case of cTnC labeled with 2-(4'-iodoacetamidoani1ino)naphthalene6-sulfonic acid. In contrast, the binding ofMg'+ in the Cterminal domain induces a fairly substantial reduction (-20%) in CPM fluorescence. This observation provides additional evidence that theC-terminal and N-terminal ends of cTnC can interact in solution (17, 29). Both x-ray scattering (30) and resonance energy transfer measurements (31) suggest

A'

20349

that in solution at neutral pH the C-terminal and N-terminal domains of sTnC are closer together than they are in the crystalline state. Acknowledgments-One of us (F. F.) thanks Dr. JohnGergely for the hospitality of his laboratory and for his advice and encouragement. We thank Yasuko Mabuchi for excellent technical assistance and Anna Wongfor amino acid analysis of peptide samples. REFERENCES 1. Leavis, P. C., and Gergely, J . (1984) CRC Crit. Rev. Biochem. 16, 235-305 K. (1976) Biochemistry 1 5 , 2. Van Eerd,J.-P.,andTakahashi, 1171-1180 3. Leavis, P. C., and Kraft, E. L. (1978) Arch. Biochem. Biophys. 186,411-415 4. Holroyde, M. J., Robertson, S. P., Johnson, J. D., Solaro, R. J., and Potter, J. D. (1980) J. Biol. Chem. 2 5 5 , 11688-11693 5. Pan, B.-S., and Solaro, R. J. (1987) J. Biol. Chem. 262, 78397849 6. Hofmann, P. A., and Fuchs,F. (1987) Am. J. Physiol. 2 5 3 , C541C546 7. Sundaralingam, M., Bergstrom, R., Strasburg, G., Rao, S. T., Roychowdhury, P., Greaser, M.,and Wang,B. C. (1985) Science 227,945-948 8. Herzberg, O., and James, M. N. G. (1985) Nature 313,653-659 9. Herzberg, O., Moult, J., and James, M. N. G. (1986) J. Biol. Chem. 261,2638-2644 10. Ellman, G. L. (1959) Arch. Biochem. Biophys. 8 2 , 70-77 11. Sippel, T. 0.(1981) J. Histochem. Cytochem. 29, 1377-1381 12. Potter, J. D. (1982) Methods Enzymol. 8 5 , 241-263 13. Szynkiewicz, J., Stepkowski, D., Brzeska, H., and Drabikowski, W. (1985) FEBS Lett. 181,281-285 14. Fabiato, A., and Fabiato, F. (1979) J. Physiol. (Paris) 75, 463505 15. Collins, J. H. (1974) Biochem. Bwphys. Res. Commun. 58, 301308 16. Potter, J. D., Seidel, J. C., Leavis, P., Lehrer, S. S., and Gergely, J. (1976) J. Biol. Chem. 2 5 1 , 7551-7556 17. Verin, A.D., and Gusev, N. B. (1988) Biochim. Biophys. Acta 9 5 6 , 197-208 18. Ingraham, R. H., and Hodges, R. S. (1988) Biochemistry 2 7 , 5891-5898 19. Grabarek, Z., Drabikowski, W., Leavis, P. C., Rosenfeld, S. S., and Gergely, J. (1981) J. Biol. Chem. 2 5 6 , 13121-13127 20. Weeks, R. A., and Perry, S. V. (1978) Biochem. J. 173,449-457 21 Chong, P. C. S., and Hodges, R. S. (1981) J . Biol. Chem. 2 5 6 , 5071-5076 22. Dalgarno, D. C., Grand, R. J. A., Levine, B.A., Moir, A. J . G., Scott, G . M. M., and Perry, S.V. (1982) FEBS Lett. 150, 5458 23. Leszyk, J., Collins, J. H., Leavis, P. C., andTao, T. (1987) Biochemistry 26,7042-7047 24. Nagy, B., Potter, J. D., and Gergely, J. (1978) J. Biol. Chem. 2 5 3 , 5971-5974 25. Dalgarno, D. C., Klevit, R. E., Levine, B. A., Scott, G. M. M., Williams, R. J . P., Gergely, J., Grabarek, Z., Leavis, P. C., Grand, R. J. A., and Drabikowski, W. (1984) Biochim. Biophys. Acta 7 9 1 , 164-172 26. Cachia, P. J., Van Eyk, J. E., Ingraham, R. H., McCubbin, W. P., Kay, C. M., and Hodges, R. S. (1986) Biochemistry 2 5 , 3553-3562 27. Van Eyk, J. E., and Hodges, R. S. (1987) Biochem. Cell. Biol. 65, 982-988 28. Johnson, J. D., Collins, J. H., Robertson, S. P., and Potter, J. D. (1980) J. Biol. Chem. 255,9635-9640 29. Wang, C.-L. A., Leavis, P. C., and Gergely, J. (1983) J. Biol. Chem. 258,9175-9177 30. Heidorn, D. B., and Trewhella, J. (1988) Biochemistry 2 7 , 909915 31. Wang, C.-L. A., Zhan, Q., Tao, T., andGergely, J. (1987) J . Biol. Chem. 262,9636-9640