Troponin-C-mediated calcium-sensitive changes in the conformation of troponin I detected by pyrene excimer fluorescence.

Troponin I (TnI) from rabbit white skeletal muscle was labeled at cysteines 48 and 64 with the fluorescent reagent N-(1-pyrene)maleimide. The fluorescence spectra of pyrene-labeled TnI (pyr-TnI) exhibit peaks characteristic of pyrene in its monomeric form and an additional peak resulting from formation of excited dimers (excimers), indicating that the labeled cysteines are close together. Formation of a pyr-TnI-TnC complex in the absence of Ca2+ has little effect on the spectrum, but when Ca2+ is bound to the low-affinity sites of TnC there is a substantial decrease in excimer and a corresponding increase in monomer fluorescence. The involvement of the low-affinity sites in the Ca2+-induced effect is consistent with the fact that Mg2+ has no effect on pyrene fluorescence. On rapid mixing of the pyr-TnI-TnC complex with Ca2+ in a stopped-flow apparatus, most of the excimer decrease is complete within the instrumental dead time, indicating a rate constant k greater than 350 s-1, which is comparable to that of the conformational change in TnC resulting from Ca2+ binding to the low-affinity sites. Rapid mixing of the Mg2-TnC-pyr-TnI complex with Ca2+ yields similar results, suggesting that the type of metal ion present at the high-affinity sites has little, if any, effect on the probe. It has been suggested previously that Cys 48 and 64 are located in a TnT-binding region of TnI (Chong P.C.S. and Hodges, R.S. (1982) J. Biol. Chem. 255, 3757). Our results suggest that a Ca2+-induced structural change in the TnI-binding region of TnC could be transmitted to TnT by affecting the TnT-binding region of TnI as part of the chain of events in the regulation of muscle contraction.

shown to bind to tropomyosin. TnI inhibits actomyosin AT-Pase, but it should not be assumed that with the troponin complex inhibition of activation of myosin ATPase by actin is solely attributable to TnI. TnI binds to actin filaments and to actin-tropomyosin filaments (4,5), inhibiting actin-activated actomyosin ATPase activity. Ca2+ initiates the contraction process by binding to TnC and inducing changes in its secondary and tertiary structure (6)(7)(8)(9)(10)(11). These changes are somehow transmitted from TnC to the other troponin subunits, to tropomyosin, and to actin, resulting in activation of actin-activated ATPase and tension development of muscle (1,12). With a view toward elucidating the molecular mechanisms involved in the transmission of the contraction signal, we have attempted to clarify the role of TnI in thin-filament regulation through the study of conformational changes in TnI in response to Ca2+-binding by TnC.
TnI from rabbit white skeletal muscle is a basic protein consisting of 178 amino acids, M, = 20,700 (13). Although TnI occurs in muscle as a ternary complex with TnC and TnT, it does form binary complexes with TnT (14) and with TnC (15). A fragment of TnI (residues 96-116) produced by cyanogen bromide cleavage binds both to actin, with substantial inhibitory activity, and to TnC (16). Another fragment containing residues 1-20 also binds to TnC (16). Evidence for the functional importance of the cysteine thiol groups of TnI came from the work of Horwitz et al. (14) showing that a biologically active troponin complex can only be formed if the sulfhydryl groups of TnI are kept fully reduced. Subsequently, Chong and Hodges (17) suggested on the basis of sulfhydryl modification studies that Cys 48 and 64 are in a region which is a binding site for TnT, while Cys 133 is exposed to solvent in both binary and ternary complexes. Studies of lysine reactivities also suggest that the portion of TnI consisting of residues 40-98 contains a binding site for TnT (18). The fluorescent reagent N -( 1-pyrene)maleimide has been shown by Lehrer and his colleagues (19-21) to be a useful probe of sulfhydryl proximity and conformational change by virtue of an emission peak corresponding to excited dimers (excimers). Preliminary studies showing disulfide formation in TnI whose Cys 133 had been blocked with iodoacetamide suggested that Cys 48 and 64 are close together and that the pyrene label would be a useful structural probe of TnI. In these studies we have labeled Cys 48 and 64 with the pyrene compound; the spectrum is indicative of a pyrene excimer. The spectrum of labeled TnI complexed with TnC shows Ca2+-sensitive changes. The range of effective Ca2+ concentrations and the lack of a Mg2+ effect on the excimer suggest that the low-affinity Ca2+-binding sites of TnC (22) are responsible for the change.

EXPERIMENTAL PROCEDURES
Protein Preparation-Troponin was prepared using the procedure of Greaser and Gergely (3), followed by chromatography on an Affi-Gel blue column (23). Troponin subunits were isolated as previously described (3). Purified proteins were stored in 6 M urea at -10 "C.
Protein concentrations were determined by adsorption at 280 nm, subtracting the absorption a t 320 nm to correct for light scattering. The following absorbance values, A (l%, 280 nm, 1 cm), were used TnT, 4.58; TnI, 3.97; and TnC, 1.59. The protein concentration of pyrene-labeled TnI was determined using the Bio-Rad protein assay (Bio-Rad Laboratories, Richmond, CA) with unlabeled TnI as the standard.
Protein Labeling-In order to label TnI with pyrene at Cys 48 and 64, it was first necessary to block Cys 133; this was done by reacting the troponin complex with iodoacetamide according to the procedures of Chong and Hodges (17). After quenching the reaction with dithiothreitol, CM-TnI was isolated (3). The sulfhydryl content of the CM-TnI, determined by the Ellman procedure (25), was 2.0 + 0.1 mol/ mol. CM-TnI was rechromatographed on a Sephadex G-25 column equilibrated with 0.5 M KCl, 25 mM MES, pH 6.0. Solid GdmCl was added to a final concentration of 5 M to the pooled fractions containing CM-TnI (1-3 mg/ml), followed by addition of N-(I-pyrene)maleimide dissolved in dimethyl formamide (1 mg/ml) in a 20:l molar ratio. The mixture was stirred for 4 h at room temperature; dithiothreitol was then added to quench unreacted label and the solution was stirred overnight at 4 "C. The reaction mixture (containing pyq-TnI) was chromatographed on a Sephadex G-25 column using 0.5 M KCl, 25 mM MES, pH 6.0, as the eluant, and the labeled protein was dialyzed exhaustively against the same buffer to remove any remaining unreacted label.
Previous work has shown that the spectral properties of pyrenelabeled proteins are influenced by the state of the succinimido ring (19-21, 24). A t pH < 6.0, the ring remains intact (pyrl-TnI), giving rise to characteristic emission peaks. At higher pH, the ring opens by hydrolysis or aminolysis (24) and there is a red shift of about 10 nm in the emission peaks (19-21). Pyrl-TnI was converted to an openring derivative (pyrll-TnI) by addition of solid GdmCl to 5 M and adjusting the pH to 8.5 with bicine buffer. After allowing the solution to stand at room temperature for 24 h, this solution was stored at -10 "C. Aliquots were dialyzed as needed against the desired buffer.
The degree of labeling, determined spectrophotometrically using a value of c = 2.3 X lo' M-' cm-' at 345 nm (20) and assuming the same extinction coefficient for pya-TnI and pyrII-TnI, was 1.9 + 0.1 mol of pyrene/mol of TnI. Other types of modified TnI were prepared for analysis and verification of the locations of the pyrene labels: 1) CM-TnI (carboxamidomethylated at Cys 133) was allowed to react with the chromophore DABMA (cf. Ref. 26) under the conditions used for making pyr-TnI to yield the product CM-DAB-TnI; 2) TnI was labeled in the ternary complex with DABMA using conditions for labeling Cys 133 (17) and will be referred to as DAB-TnI. DABMA, dissolved in dimethyl formamide (1 mg/ml), was added to a 1O:l molar ratio to the troponin complex, and after 4 h a t room temperature, the reaction was quenched with dithiothreitol. Labeled TnI was isolated by the procedures described previously for native TnI (3).
Characterization of Labeled Protein-Proteolytic fragments of the DAB-TnI and CM-DAB-TnI preparations were analyzed to verify the locations of the pyrene labels. The labeled TnI preparations were dialyzed against 50 mM NaHC03 and subjected to limited proteolysis by trypsin (~-l-tosylamido-2-phenylethyl-chloromethyl ketonetreated, Worthington Biochemicals, 1:50, w/w) for 4 h at 37 "C. The digestions were stopped by adding a 100-fold molar excess of PMSF and the reaction mixtures were lyophilized. The peptides were dissolved in 0.1% trifluoroacetic acid and analyzed by HPLC (Beckman Instruments) using a C,, pBondapak reversed-phase column (Waters Associates) with a linear gradient of 0-72% acetonitrile over 30 min. The eluant was monitored simultaneously at 230 and 535 nm. DABlabeled peaks were collected, freeze-dried, and subsequently hydrolyzed in 6 N HCI in U~C U O at 110 "C for 20 h. Amino acid analysis was carried out on a Beckman Model 119-CL amino acid analyzer.
Circular dichroism measurements were carried out on labeled and unlabeled TnI (0.2 mg/ml of protein concentration) in solutions containing 0.15 M KC1, 25 mM Pipes buffer (pH 6.8), and 0.5 mM dithiothreitol. Scans were taken between 240 and 200 nm at 25 "C using a path length of 1 mm on a modified Cary Model 60 instrument (Aviv Circular Dichroism Spectropolarimeter, Model 60 DS).
The biological activities of pyr-TnI and reconstituted troponin containing pyr-TnI were measured by determining the actomyosin ATPase activity in the presence of tropomyosin and either Ca2+ or EGTA. One-ml samples containing 0.2 mg of myosin subfragment

RESULTS
Separation and Analysis of the Labeled Peptides-Since we assumed, following Chong and Hodges (17), that the native thiol reactive in the ternary complex is Cys 133, we wished to confirm that the sites of reaction of TnI with a maleimide compound, after blocking the reactive thiol in the ternary complex, are restricted to Cys 48 and Cys 64.
According to the amino acid sequence of TnI (13), Cys 48 and 64 belong to the same tryptic peptide (41-65) while Cys 133 belongs to a different tryptic fragment (residues 132-137). The HPLC elution profile of DAB-TnI tryptic peptides monitored at 535 nm shows a single peak corresponding to 36% acetonitrile concentration (Fig. 1A). In the case of CM-DAB- TnI, there was a group of peaks at 51-56% acetonitrile with no detectable absorption at the 36% acetonitrile concentration, showing that the DAB labels were located on a different peptide (Fig. 1B). According to amino acid analysis, the DAB-TnI peak corresponds to residues 130-137 and the CM-DAB-TnI peak to residues 41-65. These results indicate that when TnI, whose Cys 133 is blocked, is labeled with DABMA, the DAB-labeled fractions contain Cys 48 and 64, suggesting that in the pyrene maleimide labeling procedure the pyrene labels are located at the latter thiols. In order to check whether labeling induces changes in the structure of TnI, we compared the far UV CD spectra of unlabelled and pyr-TnI (see "Experimental Procedures") between 200 and 240 nm. The spectra were similar in shape with values for the ellipticity at 222 nm of -8000 deg cm2 dmol" for unlabeled TnI and -7100 deg cm2 dmol" for pyr-TnI. These numbers are in reasonable agreement with other studies (34). Pyrene-labeled TnI retains its ability to inhibit the hydrolysis of ATP by actomyosin both alone and when it is incorporated into the ternary troponin complex (see Table I). In the latter case, the Ca2+-dependent release of ATPase inhibition is similar to that of unmodified troponin. These results suggest that the presence of the pyrene labels on TnI did not induce changes in the protein that grossly affect its secondary structure or biological activity.
Fluorescence Properties of the TnI Derivatives-Pyrl-TnI, produced at pH 6.0, exhibits the fluorescence peaks at 376, 396, and 416 nm, characteristic of the pyrene monomer, and an additional broad peak at 475 nm, characteristic of the pyrene excimer (Fig. 2). The fluorescence maxima of the pyrlr- TnI monomer, formed at pH 8.5, are shifted to 386, 406, and 426 nm and that of the excimer, to 487 nm. The excimer:monomer ratio for pyrII derivative is considerably higher than for pyrl-TnI, as is the case with the corresponding pyrene derivatives of tropomyosin (19). In what follows, data are shown only for pyrII-TnI; similar results were obtained with pyrI-TnI. Effects of TnC and TnT on pyr-TnI Fluorescence-Addition of TnC in the absence of Ca2+ to pyr-TnI had little effect on monomer and excimer peaks. Addition of Ca2+ to this complex caused a 25% decrease and a 5-nm blue shift of the excimer peak with a corresponding increase in the monomer peak (Fig.  3). Addition of MgZ+ in the absence of Ca2+ had no effect on excimer fluorescence, although there was a slight (5%) decrease in monomer fluorescence (not shown). Ca2+ titration of the pyr-TnI-TnC complex in the presence of an EGTA/ NTA buffer system shows that the decrease in excimer occurs in the Ca2+ concentration range corresponding to the dissociation constant of Ca2+ from the low affinity sites of TnC (pCa = 6.22, Fig. 4). These results are confirmed by the stoichiometric titration which shows that very little change occurs on adding 2 Ca*+/mol of TnC and that most of the fluorescence change occurs between 2 and 4 mol of Ca2+/mo1 of TnC (Fig. 5).
Stopped-flow experiments were performed to study the kinetics of the response of pyrene-TnI excimer to Ca2+-binding by TnC. When pyr-TnI-TnC was rapidly mixed with  buffer containing Ca2+, a biphasic decrease in excimer fluorescence was observed (Fig. 6). About 80% of the total fluorescence change occurred within the mixing time of the instrument ( 2 ms) indicating a rate constant k, > 350 s". The remaining change was a slower process with a rate constant k2 = 11 s-'. When pyr-TnI-TnC in Mg2' buffer was mixed with Ca2+ buffer, a similar biphasic change was found (kl > 300 s-', k2 = 9 s-', data not shown). Addition of TnT to pyr-TnI (10 mol/mol) had no effect on pyrene fluorescence. However, addition of TnT to the pyr-TnI-TnC complex, produced a 20% drop in monomer fluorescence, while having little change in excimer. This suggests that the binding of TnT to the complex in some way alters the local milieu of one or both of the pyrene fluorophors without significantly affecting the distance between them or their positions relative to one another. Changes in excimer fluorescence of the ternary complex induced by Ca2' were similar to those of the pyr-TnI-TnC complex.

DISCUSSION
The presence of a n excimer component in the fluorescence spectrum of TnI labeled at Cys 48 and 64 shows that these residues are close to each other. The increased yield of excimer with pyrIr-TnI compared with the case of pyrl-TnI probably wr in the TnI-TnC Complex 369 results from opening of the succinimido ring in the former (21,24). The opening of the ring would reduce the constraints that might interfere with stacking interactions of the pyrene moieties (cf. Ref. 21). Binding of Ca2+ by the low-affinity sites of the TnC induces a strong decrease in excimer fluorescence, suggesting that the pyrene molecules are pulled further apart by a conformational change affecting the region of TnI that contains Cys 48 and 64. Little, if any, effect is seen upon M$+or Ca2+-binding to the high-affinity sites. The structural changes observed here are consistent with earlier studies showing that inhibition of actomyosin ATPase in reconstituted myofibrils is reversed as Ca2+ is bound to the low-affinity sites of TnC (12).
Addition of TnT to pyr-TnI had no effect on the fluorescence spectrum. It is possible that TnT could not bind to TnI because the two very bulky, hydrophobic pyrene groups dlrectly block the TnT binding site or because the presence of the labels induces a substantially reduced affinity of TnI for TnT.
Our kinetic studies of the response of pyr-TnI-TnC to Ca2+binding indicate a biphasic change in conformation. The large, rapid conformational change in pyr-TnI induced by Ca2+ binding is nearly complete within the dead time of the instrument ( k > 350 s-'), as is the structural change in TnC resulting from binding of Ca2+ to the low-affinity sites (31). This induced structural change in TnI is sufficiently rapid to be involved in the regulatory mechanism. The basis for the slow change ( k = 11 s-') is not clear. The slow change could result from Ca2+ binding to the high-affinity sites, although no change in excimer was observed in the ca2+ titration experiments in the pCa region corresponding to the high-affinity sites (Fig. 3). We cannot, however, exclude the possibility that the high-affinity sites are responsible for some structural change when Ca2+ is bound to the low-affinity sites. Alternatively, the excimer change results only from binding of Ca2+ to the low-affinity sites, but the ensuing conformational change itself is biphasic.
Robertson et al. (30) have shown in modeling studies that complete Mg2+-Ca2+ exchange at the high-affinity sites of TnC, at which M$+ is bound in resting muscle, would be too slow to occur within the time that the peak of muscle twitch tension is reached. Hence, they suggest that Ca2+ binding at the low-affinity sites is the key event for activation. Our stopped-flow experiments showing rapid changes in TnI upon Ca2+ binding to the low-affinity sites of TnC and detecting no difference in excimer fluorescence between the Mgz-Cap-TnC-TnI and Car-TnC-TnI suggest that the events reflected in the induced pyrene excimer changes are associated with the process that in vivo leads to activation.
In this study TnI has been specifically labeled, and a conformational change is induced in the labeled region of TnI by Ca2+ binding to TnC. Previous work has suggested other Ca2+ sensitive conformational changes in TnI. Johnson et al.
(31), in stopped-flow studies on the interaction of IAANSlabeled TnI in the ternary complex with Ca2+, found a biphasic Ca2+ induced change with k1 = 110 s-' and ks = 3 s-'. They report that the IAANS label is primarily on Cys 48 (31, 32); however, it is not clear from their report how the TnI was labeled, nor is the specificity of the labeling stated, and hence it is difficult to reconcile their results with ours. Nishio and Iio (33) reacted troponin with IAANS under conditions in which Cys 133 should have been labeled. They observed a single rate constant of k > 630 s" for the IAANS fluorescence change upon Ca2+ binding to TnC. Thus, their results also show a rapid conformational change in TnI, although they cannot be directly compared with our results since their label probes a different region of TnI.
The domain of TnI containing residues 48 and 64 has been suggested as a region interacting with T n T based on studies of amino acid reactivities (17,18). Cys 48 and 64 are accessible to reaction with iodoacetamide in purified TnI, but are unreactive in TnI-TnT and whole troponin complexes (17). The lysine modification studies of Hitchcock show that Lys 40 and 65 have reduced reactivities in TnI-TnT and Tn complexes as compared with TnI. Furthermore, the reactivities of Lys 40 and 65 in the ternary complexes are Ca2+-sensitive, suggesting that a Ca2+-induced structural change occurs in the TnT-binding region of TnI. Our results using the pyrene probe clearly show that a rapid conformational change occurs in the putative T n T binding site of TnI in response to Ca2+ binding to the low-affinity sites of TnC. This change could be transmitted to TnT and thence to tropomyosin as part of the regulatory mechanism of the troponin-tropomyosin complex. Experiments to study this possibility are in progress.