Phosphoenzyme conformational states and nucleotide-binding site hydrophobicity following thiol modification of the Ca2+-ATPase of sarcoplasmic reticulum from skeletal muscle.

Enhanced fluorescence of the ATP analogue 2',3'-O-(2,4,6-trinitrocyclohexyldienylidine)adenosine 5'-triphosphate (TNP-ATP), bound to the Ca2+-ATPase of skeletal muscle sarcoplasmic reticulum, is closely related to phosphoenzyme levels (Bishop, J. E., Johnson, J. D., and Berman, M. C. (1984) J. Biol. Chem. 259, 15163-15171) and has an emission maximum consistent with decreased polarity of the TNP-ATP-binding site. The phosphoenzyme conformation responsible for increased nucleotide-binding site hydrophobicity has been studied by redistribution of phosphoenzyme intermediates following specific thiol group modification. N-Ethylmaleimide, in the presence of 50 microM Ca2+, 1 mM adenyl-5'-yl imidodiphosphate, pH 7.0, at 25 degrees C for 30 min, selectively modified the SH group essential for phosphoenzyme decomposition, which resulted in decreased ATPase activity, Ca2+ uptake, and a decrease in ATP-induced TNP-ATP fluorescence. Phosphorylated (Ca2+, Mg2+)-ATPase levels from [gamma-32P] ATP remained relatively unaffected (3.1 nmol/mg), but the ADP-insensitive fraction decreased from 56 to 15%. Phosphoenzyme levels from 32Pi were also decreased to the same extent as turnover, with equivalent loss of Pi-induced TNP-ATP fluorescence. The E1 to E2 transition, as monitored by the change in intrinsic tryptophan fluorescence, was unaffected. Modification of thiol groups of unknown function did not modify turnover-induced TNP-ATP fluorescence. It is concluded that the ADP-insensitive phosphoenzyme, E2-P, is responsible for enhanced TNP-ATP fluorescence. This suggests that the conformational transition, 2Ca2+outE1 approximately P----2Ca2+inE2-P, is associated with altered properties of the noncatalytic, or regulatory, nucleotide-binding site.

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. compatible with the above is shown in Scheme I.
Catalytic activity of vesicular and purified Ca2+-ATPase shows a complex dependence on substrate concentration. Following saturation of the catalytic site in the micromolar range, ATP at millimolar concentrations further accelerates turnover, an effect postulated to be due to acceleration of the rate-limiting E2 to El transition (2, 3). ATP may also affect other partial reactions of the catalytic cycle, in addition to its role as a true substrate. McIntosh and Boyer (4) have shown that ATP modulates the Pi H"0H exchange reaction, involving steps 6 and 7 of Scheme I. It is uncertain whether this modulation is caused by binding of ATP to an allosteric regulatory site or to the phosphorylated catalytic site following release of ADP from El-P. ADP. ' The ATP analogue TNP-ATP is a useful probe of nucleotide-binding sites and of changes in the environment of these sites during catalysis by the Ca2+-ATPase. TNP-ATP binds to both catalytic and noncatalytic sites ( 5 ) . Binding to nonphosphorylated SR vesicles in the absence of substrate is associated with a moderate increase in fluorescence. However, TNP-ATP fluorescence is increased 10-fold following E-P formation, caused by the addition of either ATP plus Ca2+ or Pi in the absence of Ca2+. The fluorescence level is related quantitatively and dynamically to E-P levels ( 5 , 6). It is uncertain which of the ATP-binding sites is responsible for enhanced TNP-ATP fluorescence. The probe has been suggested to bind to the adenine nucleotide-binding site of the phosphorylated species after ADP has been released in the forward reaction (6). Dupont and Pougeois (7) have also suggested that the increase in fluorescence indicates that H20 molecules are expelled from the nucleotide-binding site as part of the catalytic cycle ( 7 ) . Pi 2 Cs** (in) SCHEME I 'The abbreviations used are: El-P and Ez-P, phosphorylated intermediate forms of El and E, conformations, respectively; TNP-ATP, 2',3'-0-(2,4,6-trinitrocyclohexyldienylidine)adenosine 5'-triphosphate; NEM, N-ethylmaleimide; EGTA, [ethylene bis(oxyethy1-enenitri1o)ltetraacetic acid; E-P, phosphorylated (Ca2+,M%+)-ATPase; AMP-P(NH)P, adenyl-5"yl imidodiphosphate; Mops, 4-morpholinepropanesulfonic acid; SR, sarcoplasmic reticulum; SRV, sarcoplasmic reticulum vesicles. 7041 Bishop et al. (8) suggested that all phosphoenzyme intermediates are related to enhanced TNP-ATP fluorescence based on conditions that result in the accumulation of different E-P species and on TNP-ATP fluorescence lifetime studies. Alternatively, the rate of increase of TNP-ADP fluorescence is closely associated with the rate of formation of the ADP-insensitive phosphoenzyme (E,-P) at low temperatures (9). TNP-ATP fluorescence is sensitive to K+. Increases in the concentration of monovalent cation salts decrease fluorescence in a manner that has similar specificity and affinity to that of the monovalent cation site that accelerates ATP hydrolysis and enhances E-P decomposition (10). Potassium binding enhances E,-P to El-P conversion (11) and accelerates E,-P hydrolysis, the latter step being rate-limiting for catalysis (12). Both of these actions favor a decrease in E,-P with constant levels of total E-P (13). An alternative explanation, suggested by Bishop et al. (S), is that K+ decreases the affinity of the noncatalytic site for trinitrophenylated ATP derivatives.
Kawakita et al. (14) have characterized discrete classes of thiol groups related to formation (SHF) and decomposition (SHD) of E-P species. These groups may be specifically labeled by treatment of SRV with N-ethylmaleimide (NEM) in the presence or absence of a nonhydrolyzable analogue of ATP and Ca2+. In the absence of the ATP analogue, those thiol groups necessary for E-P formation from ATP (SHF) are also blocked. The presence of an ATP analogue is necessary for protection of the SHF group and to specifically label SHD (14, 15). Modification of SHo thiol groups inhibits the ADPsensitive (E,-P) to ADP-insensitive (E,-P) transition, with a net accumulation of El-P in the forward reaction (14, 15). Nakamura and Tonomura (16) have recently shown that, following SHD modification, Ca2+ is occluded in the ADPsensitive conformation and that E2-P formation, in the forward reaction, is inhibited, suggesting that this is a key event of the translocation process.
The aim of the present experiments was to favor accumulation of El-P by blocking those thiol groups (SHD) that are specifically involved in E-P decomposition and to relate the effects of this blockade to enhanced TNP-ATP fluorescence. In this study, derivatization of SHD thiol groups with NEM was found to inhibit both ATP-and Pi-induced TNP-ATP fluorescence. The significance of this phenomenon is explored with respect to the mechanism of catalysis and energy transduction by the Ca2+-ATPase.
32Pi was purified by the method of de Meis and Tume (17). AMP-P(NH)P was from Boehringer Mannheim. TNP-ATP was synthesized and purified by the method of Hiratsuka (18). Concentrations of the nucleotide were determined by absorbance measurements at 408 nm in 0.1 M Tris-C1, pH 8.0, using c4w = 26,400 M"-cm" (18). SRV were prepared from rabbit hind leg muscle according to the method of Eletr and Inesi (19). Protein determination was by the Lowry method using bovine serum albumin as standard (see Ref. 20). ATPase content of SR preparations was determined by polyacrylamide gel electrophoresis according to the method of Laemmli (21) as modified by McIntosh and Ross (22). The ATPase constituted between 70 and 90% of total protein.
Thiol modification was performed essentially as described by Kawakita et al. (14) except that KC1 was omitted from the incubation medium. SRV (2 mg/ml) were derivatized under the standard conditions of 0.4 mM NEM, 50 p M Ca", 1 mM AMP-P(NH)P, 5 mM MgC12, pH 7.0, at 25 "C. The reaction was started by the addition of NEM and stopped by 10-fold dilution into various assay media. KC1 has previously been shown to decrease phosphoenzyme-induced TNP-ATP fluorescence with [K+]o.S = 50 mM (10). The extent and specificity of NEM modification are not altered in the absence of KC1 (15). The final reaction mixture after dilution contained 0.1 mM AMP-P(NH)P. This nucleotide concentration was shown to decrease ATP-or Pi-induced enhanced TNP-ATP fluorescence by less than 10%. Consequently, 0.1 mM AMP-P(NH)P was included in the medium of all assays for TNP-ATP fluorescence, ATPase activity, and calcium uptake.
Fluorescence measurements were performed on a Model SPF 500 Aminco-Bowman spectrofluorimeter at excitation and emission wavelengths of 418 and 525 nm, respectively, with 10-15-nm band passes. NEM-treated SRV were diluted 10-fold into 100 mM Mops/Tris, pH 7.0,50 p M CaC12, 5 mM Mg& and 2 p M TNP-ATP. 100 p M ATP or 10 mM PJTris, pH 7.0, plus 1 mM EGTA was added to achieve enhanced TNP-ATP fluorescence. Absorbances at 418 and 525 nm were measured in the same cuvette at this stage of the assay for estimation of the inner filter effects of NEM binding by the method of Lakowicz (23). Inner filter effect correction amounted to less than 5% of total fluorescence. Intrinsic tryptophan fluorescence studies were performed under the same conditions as for TNP-ATP fluorescence with excitation and emission wavelengths of 284 and 333 nm, respectively, with 4-nm band passes. 0.5 mM EGTA decreased fluorescence by 4%, whereas further addition of 0.51 mM Ca" restored fluorescence to the original levels, as previously shown by Dupont and Leigh (24).
ATPase activities were determined in the presence of 100 mM KC1 by the coupled spectrophotometric assay by following the decrease in absorbance of NADH at 340 nm (25). Ca" transport was measured by the uptake of ''Caz+ into vesicles in 5 mM Tris oxalate with 2 p M TNP-ATP, 250 p M ATP, 5 mM MgCl,, 100 mM Mops/Tris, pH 7.0, at 25 "C. Reactions were stopped by placing SRV on 0.45-pm millipore filters, which were washed with 10 volumes of 5 mM CaC12 in 100 mM imidazole HC1, pH 6.0.
The stoichiometry of NEM derivatization was determined by reacting [3H]NEM (250 pCi/pmol), as above, and the reaction was stopped by diluting SRV into 0.25 M perchloric acid (14), followed by filtration on GF/F glass-fiber filters. Filters were washed with 100 volumes of 0.125 M perchloric acid, and radioactivity was determined in 6 ml of Insta-Gel (Packard Instrument Co.).
Phosphoenzyme levels were measured with [ Y -~~P ] A T P and 32Pi by phosphorylation, acid quenching, and filtration as previously described by Lacaphre et al. (26). 100 p~ [3ZP]ATP was added to 0.4 mg/ml NEM-treated SRV, 50 p M Ca", 5 mM MgC12, 100 mM Mops/ Tris, pH 7.0, at 25 "C for 5 s. The reaction was stopped by the addition of 400 p1 of the reaction mixture to 2 ml of 0.25 M perchloric acid and 15 mM Pi and then filtered on a GF/F glass-fiber filter as described previously. E-P levels from 32Pi (10 mM) were determined in 10% dimethyl sulfoxide essentially as described above in the absence of Ca2+ (in the presence of 1 mM EGTA). The reaction was stopped after 20 s, filtered on GF/F filters, and washed with 25 ml of 0.25 M perchloric acid and 15 mM Pi as described by Lacapere et al. (26). Levels of acid-stable 3ZP-labeled phosphoryl enzyme, after addition of millimolar concentrations of ADP, were measured at 20 ms using a Dionex D-133 Multimix apparatus or manually for times greater than 1 s. The Multimix apparatus was equipped with three mixing syringes and used according to the method previously described by Guillain et al. (27). One volume each from two reactant syringes (A and B), containing SRV and [Y-~'P]ATP (3,OOO-4,000 dpm/nmol), was mixed and aged in a 450-pl coil for 4 s. The third syringe (C), containing 2 volumes (450 pl) of 10 mM MgADP and 1 mM EGTA, was mixed with aged E-32P and approximately 500 pl of ADP-reacted E-32P was collected in the sample collect syringe, containing 500 pl of quench solution. The dead volume between syringe C and the collect syringe was 100 pl, and sample flow rate was 5 ml/s.

RESULTS
In this study, we have examined the effects of modification of sulfhydryl groups of the Ca2+-ATPase on various functional activities of the enzyme . Kawakita et al. (14) have shown that relatively specific derivatization of those thiol groups involved in phosphoenzyme decomposition (SHD) can be achieved in the presence of Ca2+ and of the nonhydrolyzable analogue, AMP-P(NH)P. The latter protects those thiols involved in phosphoenzyme formation following nucleotide binding.
The extent of modification with the reagent NEM depends on the time of the reaction and the concentration of the reagent (Fig. 1   PM NEM for up to 50 min, has been used for further functional studies. NEM modification resulted in a parallel decline of Ca2+ uptake and Ca2+-dependent ATPase activity (Fig. 2), consistent with inhibition of the rate-limiting step for phosphoenzyme decomposition. Derivatization thus had no apparent effect on the ratio of Ca2+ uptake to ATPase activity, which is a measure of coupling of transport to ATP hydrolysis. In previous studies, data on thiol group modification and its effects on transport activity were obtained in the absence of Ca". EGTA, which causes irreversible uncoupling of the Ca2+ pump (28), complicates interpretation of these data.
Phosphoenzyme formation, either in the forward direction from ATP plus Ca2+ or from Pi in the presence of EGTA, enhances fluorescence of bound TNP-ATP (29). Original studies, performed at pH 8.0 in 20% (v/v) glycerol, showed Pi-induced fluorescence to be half of that induced by ATP (5). In the present study (Fig. 3), Pi-induced fluorescence was assayed at pH 7.0 in 10% (v/v) dimethyl sulfoxide, which maximizes E-P levels. Under these conditions, ATPand Pi-indl xed fluorescence are equivalent. NEM modification diminished fluorescence enhancement by either ATP or Pi to approximately similar extents (compare Fig. 3, A and B, a t 16  and 40 min with Fig. 3C). NEM modification did not alter the relatively low TNP-ATP fluorescence from the nonphosphorylated enzyme prior to the addition of ATP or Pi (Fig. 3,  A and B).
Fluorescence enhancement was accompanied by a blue shift in the emission spectrum of bound TNP-ATP (545 to 530 nm) (Fig. 4, truces b and c), consistent with previous suggestions of increased hydrophobicity of the TNP-ATP-binding site during turnover (6). NEM modifications that decreased fluorescence had no effect on the X , , , of emission (Fig. 4,  traces d and e ) . The alteration in fluorescence does not appear to be due to a change in light scattering as a result of flocculation of vesicles since fluorescence emission in the ranges 450-475 and 650-700 nm was unaltered. NEM modification also had little effect on the absorbance at 410 and 530 nm, causing negligible inner filter effects (data not shown). Enhanced TNP-ATP fluorescence was unchanged under conditions (40 p~ NEM) that have been reported to result in the modification of those thiol groups of unknown function (SHN) (Fig. 3C).
TNP-ATP fluorescence has been related to E-P levels (5, 6) under static and dynamic conditions. Kawakita et al. (14) have reported that inclusion of AMP-P(NH)P protects a group of sulfhydryls (SH,) that are related to E-P formation; and thus, NEM modification had no effect on ATP-dependent E-P levels. These findings are confirmed in Fig. 5. However, Pi-induced E-P levels were diminished by approximately 50% following derivatization. This suggests that the thiol groups that are modified (SH,) are involved in both the hydrolysis of E-P to Pi and, in the reverse reaction, of the formation of covalent E -P from Pi, with the exclusion of water. Thus, under these conditions, the decline in Pi-dependent TNP-ATP fluorescence is readily explained by decreased E -P formation. However, the decrease in TNP-ATP fluorescence from ATP plus Ca2+ occurs when total E -P levels are unaltered.
Total E-P levels include both El-P and E2-P intermediates.
Pi-induced E-P, in the absence of Ca2+, is assumed to be predominantly E,-P; while during turnover, both El-P and  l (30, 31).
The relative proportions of ADP-sensitive and ADP-insensitive E-P species have been determined following rapid quenching by millimolar concentrations of ADP and EGTA (Fig. 6A). This shows a rapid decay in E-P from 3.1 nmol/ mg, complete within the mixing time, which was followed by a slower decay in the ensuing 5 s. The data yield a value of the ADP-sensitive fraction of 44%, which is consistent with previous data (13). Similar experiments performed on NEMmodified vesicles showed that the ADP-sensitive E-P species was increased to 85% (Fig. 6B) (16, 32, 33). Exponential fits of the slow phases of E-P decay (Fig. 6, A and B ) following addition of ADP gave similar values of 0.71 and 0.60 s" in the control and NEM-modified vesicles, respectively.
A possible explanation of the inhibition of Pi-induced enzyme phosphorylation by modification of SH, is that in addition to its effects on the transition of El-P to E2-P, there is a concomitant block in interconversion of the free enzyme species, El and Ez. However, the effects of Ca2+ and EGTA on intrinsic tryptophan fluorescence, which has been shown to report the El e E2 conformational transition (24), were unaffected by NEM modification, (data not shown).  (6, 29). The latter site shows a relatively higher hydrophobicity that results in a 7-10-fold increase in bound TNP-ATP fluorescence. TNP-AMP competitively inhibits stimulation of catalysis by millimolar concentrations of ATP, implicating this site as the regulatory site (35). However, the precise nature of the phosphoenzyme species responsible for enhanced fluorescence of bound TNP-nucleotides is controversial. It has been proposed that the E,-P intermediate alone (7,9) or, alternatively, that all phosphorylated intermediates are involved (8). We present evidence in this study that the E,-P conformation is the most likely subspecies that results B, the assay was repeated using NEM-modified SRV (0) (as described for Fig. 3). Nonlinear least-squares fits for phosphoenzyme decay showed rate constants of 0.71 and 0.60 s-l, whereas extrapolation of the fitted curve intercepted the rapid decay phase in E-P at 1.17 and 0.65 nmol/mg before and after NEM modification, respectively. in increased site hydrophobicity.
Derivatization of thiol groups on the Ca2+-ATPase has provided a useful method for the modification of the distribution of phosphorylated intermediates. The decrease of ATPase activity and Ca2+ uptake, under conditions that do not alter total E-P levels, are essentially in agreement with a number of previous studies involving selective SHD modification (14, 15). These have described the effect of SHD modification as being the result of inhibition of the conversion of ADP-sensitive to ADP-insensitive E-P (16, 36). Nakamura and Tonomura (16) have cited this evidence as the basis for their conclusion that Ca2+ ions are occluded by the El-P species.
The finding that the SHD-modified enzyme cannot be phosphorylated in the reverse direction by Pi suggests two possibilities. Either the E,-P conformation cannot be attained from both forward and reverse phosphorylation processes or the transition of E, to E,, whether phosphorylated or not, is inhibited. The latter argument raises the possibility that the E , conformation can exist in the absence of bound Ca2+. This is unlikely since intrinsic tryptophan fluorescence was unaffected.
SHD modification has previously been described as Ca2+sensitive (14). The protocol may, in fact, be introducing a complication of uncoupling, brought about by incubation of SR vesicles in low [Ca"] (<lo-' M), as described by McIntosh and Berman (28). This would also complicate the interpretation of Ca'+-modulated conformational sensitivity toward SHD modification. Hence, the protocol, as described by Kawakita et al. (14), using high Ca" only was employed in the present study. Whatever the mechanism involved in the modification of conformational transitions, it appears that the E,-P conformation is inhibited by SHD modification, even in the presence of Ca2+.
The aim of the present study was to determine the nature of the phosphorylated intermediate species that result in increased hydrophobicity of the noncatalytic binding site on the Ca2+-ATPase, to which TNP-nucleotides bind and show enhanced fluorescence. Under conditions of NEM modification that did not alter total E-P levels during catalysis, TNP-ATP fluorescence was inhibited from both ATP plus Ca2+ and from Pi in the absence of Ca2+. This would indicate that El-P, the predominant species under these conditions, is unaltered by SHD modification. Furthermore, this species does not appear to be responsible for enhanced fluorescence. Two possible mechanisms were considered. The first is that E,-P formation and its level determine enhanced TNP-ATP fluorescence. The decreased fluorescence during turnover (approximately 50% following SHD modification) correlates well with the decrease of ADP-insensitive E-P from 1.3 to 0.7 nmol/mg. A second mechanism that requires consideration is related to the possibility that catalytic and regulatory nucleotide sites may, in fact, be nonidentical, in which case events at the catalytic site would be transmitted to the noncatalytic site by a conformational transition. Fluorescence enhancement may be related to energy coupling since EGTA pretreatment that uncouples transport causes a parallel decrease in fluorescence enhancement. Berman (37,38) has shown dissociation of enhanced TNP-ATP fluorescence from E-P formation under conditions that uncouple transport from catalytic activity and E-P levels and has proposed that intermediate conformational events of a strictly ordered type couple catalytic intermediates to the transport cycle. SHo modification would then block conformational changes originating at the catalytic site from being transmitted to the regulatory site, thus preventing enhanced fluorescence of the El-P conformation. This mechanism would appear unlikely in view of the parallel block in E2-P formation and of TNP-ATP fluorescence from the forward and reverse directions. The E,-P species is therefore the most probable intermediate responsible for enhanced TNP-ATP fluorescence, in support of a recent study showing a close correlation between enhanced TNP-ATP fluorescence and the ADP-insensitive phosphoenzyme at varying [KC11 and pH (39).
The significance of the increased hydrophobicity of the noncatalytic, or regulatory, site in the E,-P conformation may be to prevent access of water to the acyl phosphate, prior to Ca2+ release to the lumen of the vesicle, in compliance with rules for coupling, formulated by Pickart and Jencks (40). These require that the 2Ca2+.E2-P conformation not be dephosphorylated in the Ca2+-bound conformation.