The Reaction of Sulfhydryl Groups of Sodium and Potassium Ion-activated Adenosine Triphosphatase with N-Ethylmaleimide THE RELATIONSHIP BETWEEN LIGAND-DEPENDENT ALTERATIONS OF NUCLEOPHILICITY AND ENZYMATIC CONFORMATIONAL STATES*

The reaction between N-ethylmaleimide and (Na+ + K')-ATPase, performed under ligand conditions which produce each of the kinetic states of the enzyme and their associated conformational forms, was examined through an analysis of the inhibition of enzymatic activity and the incorporation of radiolabeled reagent into the enzyme. The inactivation reactions displayed pseudo-first order kinetics with respect to the concen- tration of active enzyme, indicating that the loss of activity is associated with the alkylation of a unique sulfhydryl group. In the absence of enzyme phosphorylation, the nucleophilicity of this sulfhydryl group is affected primariIy by the nature of the monovalent cation present and does not correlate with the conformational state. A method for determining the actual concentration and specific radioactivity of radiolabeled N-ethylmal- eimide during the reaction with (Na' + K+)-ATPase was developed, allowing the measurement of the total reactive Sulfhydryl groups of native (Na' + K+)-ATP- ase under conditions identical with those of the inactivation studies. The labeling of the enzyme complex is associated almost exclusively with the large polypeptide, which contains four sulfhydryl groups which react with this reagent. One of these residues is presumably the sulfhydryl responsible for inactivation of the enzyme. T w o react stoichiometrically and rapidly with N-ethylmaleimide under all conditions. Lyophilized samples, containing SDS from the previous dialysis, were dissolved in water and made 5% in 2-mercaptoethanol. Three different methods which all gave identical results within experimental error were used for the quantitation of radiolabeled N-ethylmalehide covalently bound to the large polypeptide. The first method utilized preparative gel electrophoresis. A protein sample (200 pg) was applied to a gel which was submitted to electrophoresis and then immediately scanned at 280 nm on a Zeiss PMQII spectro-photometer fitted with a linear transporting chamber. The region containing the a polypeptide was cut out of the gel in 5 to 10 slices (1 mm) and eluted with three changes (2 ml) of 50 nm ammonium bicarbonate buffer, pH 7.8, 0.1% SDS, 25 m~ 2-mercaptoethanol at 37 "C over 24 h. Equivalent portions of blank gels, as controls, were also eluted. These samples were pooled, passed through a Millipore filter (0.45 p), lyophilized, and redissolved in 2001.11 of water. Triplicate aliquots were taken for both liquid scintillation and quantitative amino acid analysis. For the latter, norleucine was added as an internal standard and at least 10 amino acids were used to calculate protein concentration, correcting for the background in the control (~20%). The relative amount of each amino acid was consistent with the published composition of the a polypeptide (35). The molar ratio of incorporation was calculated using the independently determined specific radioactivity of the N-[3H]ethylmaleimide, the specific radio- activity of the a polypeptide, and the value of 121,000 for the molec-ular weight of the a polypeptide (36). This is presented as moles of ESC,' the cysteine adduct of radiolabeled N-ethylmaleimide/mol of a polypeptide. In addition to the direct method of measurement just described, a second method of quantitation was also used. Gels were stained for protein with Coomassie brilliant blue, scanned at 500 nm, large chain of (Na+ + K+)-ATPase which react rapidly with N-ethylmaleimide when the enzyme is poised in the Ez kinetic state; 1.9 f 0.3, in the E*-P state; 1.8 f 0.3, in the El state; and 1.9 f 0.2, in the E1 - P state. The results demonstrate that one additional sulfhydryl group, distinct from the one whose nucleophilicity is sensitive to levels of Na+ and K+ and whose alkylation inactivates the enzyme, is exposed upon stabilization of the enzyme in the EZ kinetic state, and demonstrates a unique conformational form of the enzyme.

The reaction between N-ethylmaleimide and (Na+ + K')-ATPase, performed under ligand conditions which produce each of the kinetic states of the enzyme and their associated conformational forms, was examined through an analysis of the inhibition of enzymatic activity and the incorporation of radiolabeled reagent into the enzyme. The inactivation reactions displayed pseudo-first order kinetics with respect to the concentration of active enzyme, indicating that the loss of activity is associated with the alkylation of a unique sulfhydryl group. In the absence of enzyme phosphorylation, the nucleophilicity of this sulfhydryl group is affected primariIy by the nature of the monovalent cation present and does not correlate with the conformational state.
A method for determining the actual concentration and specific radioactivity of radiolabeled N-ethylmaleimide during the reaction with (Na' + K+)-ATPase was developed, allowing the measurement of the total reactive Sulfhydryl groups of native (Na' + K+)-ATPase under conditions identical with those of the inactivation studies. The labeling of the enzyme complex is associated almost exclusively with the large polypeptide, which contains four sulfhydryl groups which react with this reagent. One of these residues is presumably the sulfhydryl responsible for inactivation of the enzyme. T w o react stoichiometrically and rapidly with N-ethylmaleimide under all conditions. The nucleophilicity of the fourth sulfhydryl group is governed by the conformational state of the enzyme, but the alkylation of this residue does not result in loss of enzymatic activity.
Sodium and potassium ion-activated adenosine triphosphates is the enzyme which catalyzes the hydrolysis of MgATP coupled to the active transport of Na' and K' across cell membranes (1,2). The larger of the two polypeptides which form the enzyme complex, the a polypeptide, is phosphorylated by MgATP during each turnover (3, 4). The formation of this phosphorylated intermediate is dependent on the presence of Na' , but it is hydrolyzed in the presence of K' (5). The differences in the cation sensitivity of the phosphorylated intermediate originally led to a structural description of active transport in terms of an isomerization between two conformations of the enzyme termed El, the inward-facing or Na' form, and Et, the outward-facing or K+ form (6)(7)(8). More recent evidence for the existence of E L and E2 has been provided from observations of the differences between the two cation-enzyme complexes in their affinity for ATP (9)(10)(11)(12), their intrinsic fluorescence spectrum (13, 14), and their susceptibility to proteolysis by trypsin (15-1")-These results have defined the various combinations of ligands whose presence should produce each of the kinetic states and associated conformational forms through which (Na+ + K')-ATPase' passes during its turnover.
The abilities of the various ligands to stabilize the conformational states of (Na' + K+)-ATPase have also been inferred from apparent alterations in the nucleophilicity of sulfhydryl groups of the enzyme toward the reagent N-ethyhaleimide (18). The reaction between this reagent and the enzyme results in the loss of enzymatic activity (8,19), and the enzyme can be protected from this inactivation by the presence of ATP (20,21). It has not been determined whether this effect is due to the protection of a sulfhydryl group in the adenine nucleotide binding site or a sulfhydryl group at a distant location on the large polypeptide which is affected by ligand-bound stabilization of conformation (21,22). Previous attempts to describe the influence of the binding of ligands to (Na' + K+)-ATPase on the N-ethylmaleimide reaction have often produced contradictory results (20,21,23-25) and have not been based on a systematic and simultaneous comparison between the kinetic inhibition of the enzyme and the total number of alkylated sulfhydryl groups. It is important that this comparison be made in order to distinguish the sulfhydryl groups whose reactivity is affected by conformation state. This would be a logical preliminary to an identification of those portions of the enzyme associated with conformational changes and those in the active site. This report describes experimental evidence demonstrating that there are only four sulfhydryl groups located on the large polypeptide which can be alkylated by N-ethyhdeimide 'The abbreviations used are: (Na' + K+)-ATPase, sodium and potassium ion-activated adenosine triphosphatase; SDS, sodium dodecyl sulfate; CDTA, trans-l,2-diaminocyclohexanetetraacetic acid; DTNB, 5,5'-dithiobis(2-nitrobenzoic acid); ESC, S-(N-ethylsuccinimido)-L-cysteine. under conditions which maintain the native enzyme. The nucleophilicity of one of these is affected by the nature of the ligands present, while that of another is controlled by conformational state. The alkylation of only the former leads to inactivation of the enzyme. It is also shown that conformational states exist in addition to El and EP, since differences occw in the sulfhydryl reactivity between the E:! and E*-P kinetic states.
General Methods-(Na' -t K+)-ATPase activity was assayed as previously described (26) without the added phosphatidyl-L-serine. The Lowry method of protein determination, except in those cases in which protein was determined by amino acid analysis, was used as previously described (26).
Enzyme Preparations-All preparations of (Na+ + K+)-ATPase used in the present experiments were isolated from canine renal medulla. Supernatant enzyme (Step 5) was prepared as described (26) and deoxycholate was removed with XAD4 resin (27). The specific activity of this enzyme was 7 pmol of Pi (mg protein)" rnin". Just prior to reaction with N-ethymaleimide, the enzyme was incubated with 50 mM 2-mercaptoethanol for 5 min at 37 "C and dialyzed at 4 "C under N2 against histidine/sucrose buffer (30 m~ histidinium chloride, pH 7.5,0.25 M sucrose, 2 m~ EDTA). Deoxycholate-treated, zonal gradient microsomes were purified by a modification (28) of the procedure of Jprgensen (29). They had a specific activity of 4 pmol of Pi (mg protein)" mh"' and were also treated with 2-metcaptoethanol as described above.
SDS-purified (Na' + K')-ATPase was prepared by the method of Jprgensen (30) with some modifications. For some experiments, it was necessary to include 10 m~ 2-mercaptoethanol in all buffers used to purify the enzyme. This was removed immediately before the Nethylmaleimide reaction. In a typical large scale preparation, crude microsomes (26) are incubated with recrystallized SDS at a final concentration of 0.7 mg ml" SDS and 1.4 mg ml" protein in a fmal volume of 180 ml, and then layered on a number of identical, discontinuous sucrose gradients (10 ml of 37.3%, 13 ml of 28.876, 10 ml of 15%, 30-ml sample, all in incubation buffer). The gradients are centrifuged for 90 min at 190,000 X g. The cloudy, membrane-containing layer at the 28.8 to 37.3% interface is collected, diluted 2-fold with 50 mM imidazolium chloride, pH 7.5, 2 mM EDTA, 3 mM ATP, 2mercaptoethanol, pH 7.5, sedimented by centrifugation at 140,000 X g for 4 h and resuspended in histidine/sucrose buffer. The specific activity of this preparation was 17 to 20 pmol of Pi (mg protein)" rnin".
Removal of Nucleotides-A combination of methods was used to remove tightly bound adenine nucleotides from SDS-purified enzyme (17). The enzyme preparation (2.5 mg I C ' ) is made 120 m~ in NaCl, 20 m~ in KCl, and 5 m~ in MgClz, sealed under Nz, and incubated for 15 min at 37 "C, to convert tightly bound ATP to ADP, which, in the presence of Mg", has a lower affinity for the enzyme (9,17). It is then dialyzed at 4 "C under NZ against two changes of 30 m~ histidinium chloride, pH 7.5,0.25 M sucrose, 5 mM MgC12 for 16 h, and finally against the same buffer with 5 m~ CDTA replacing the MgCl2.
The enzyme solution is diluted and twice sedimented and resuspended. In some cases, the dialysis step including MgClz was omitted. Usually, the loss of enzyme activity was 10 to 15% for this latter procedure, while up to 30% loss is observed when including the MgClz dialysis step. Since the nucleotide removal procedure also removes 2mercaptoethanol, all buffers are flushed with Nz prior to use.
Kinetics of N-Ethylmaleimide Inactivation of (Nu' + K+)-ATPase Activity-The reaction buffer was histidine/sucrose buffer. The final protein concentration was 0.3 to 0.5 mg ml" in a reaction volume of 180 to 200 pl. In circumstances requiring the absence of Na' or K' , the Tris salt of all weak acids was used. An ATP-regenerating system, using pyruvate kinase and phosphoenolpyruvate as described else-where (31), was included in some experiments. The enzyme was allowed to equilibrate with ligands for 5 min at 37 "C, while stining under NP. Fresh stock solutions of N-ethylmaleimide were prepared in histidine/sucrose buffer for each experiment. The reaction is initiated by the addition of 0.06 volume of 90 mM N-ethylmaleimide to give a nominal final concentration of 5.0 mM, which was checked by measuring absorbance at 302 nm and using the molar extinction coefficient of 620 (32). The loss of N-ethylmaleimide, measured by this method, is less than 5% after 30 min at 37 "C. An aliquot (25 pl) was withdrawn immediately after initiation of the reaction and subsequent aliquots were withdrawn at 5-min intervals. The reaction was quenched by the addition of the aliquots to an equal volume of cold reaction buffer containing 50 m~ 2-mercaptoethanol. The dilution of the ligands remaining from the inactivation experiment during the assay is at least 100-fold, and therefore, their contribution to the assay is negligible. The logarithm of the fraction of remaining activity was plotted against time. All plots were linear which permitted calculation of a pseudo-first order rate constant from the half-time of inactivation. Insignificant differences in the inactivation rates were observed when enzyme from which nucleotide had been removed by CDTA dialysis was compared to enzyme from which it had not. Larger differences, however, were observed between stripped and unstripped enzyme, namely a more rapid loss of activity in stripped enzyme, when nucleotide removal involved the MgCb dialysis step. These rates were, however, always identical in a relative sense throughout all the ligand conditions when a given enzyme preparation was compared to another preparation. Therefore, only relative values of the rate constants under various ligand conditions were tabulated. Most of the reactions were carried out for 30 min since several of the conditions had rates such that more than two half-lives were covered during this time period. The usual loss of enzymatic activity in controls due to unspecified inactivation over 30 min at 37 "C was 5%. All values were nevertheless corrected for control inactivation.

ide and N-[14C]Ethylmaleimide-N-[3H]Ethylmaleimide or N-['4C]
ethylmaleimide provided by the manufacturer in pentane, was diluted with pentane, aliquoted, and stored at 4 "C in sealed ampules under Ar. To prepare a stock aqueous solution, a sample was added to hiitidine/sucrose buffer and the pentane removed by evaporation with a stream of N2. On many occasions, a polymeric film formed during the transfer step, and this could explain much of the loss in material. Also, some loss must result from sublimation into the Nz stream. The usual recovery of reagent was 30 to 50% based on the ESC assay, and the specific radioactivity agreed with the supplier's value. A greater and more consistent yield, up to 65%, was obtained when the transfer was made into 95% ethanol. This procedure could not be employed, however, in the experiments described here because the ethanol necessary to transfer the reagent inhibited the enzyme. The concentration and specific radioactivity of radiolabeled Nethylmaleimide was determined by its reaction with L-cysteine (33). Two 5-4 aliquots are withdrawn from any sample and added to 100 pl of a freshly prepared solution of 5 mM L-cysteine dissolved in 0.2 M sodium citrate, pH 2.2. The reaction is initiated by raising the pH to 5.8 with 1 M NaOH. The mixture is incubated under Nz at 37 "C for 30 min, quenched by the addition of 0.5 ml of the citrate buffer, and submitted to quantitative amino acid analysis on a Beckman model 118C amino acid analyzer. Elution with 0.2 M sodium citrate, pH 3.4, 1% thiodiglycol, resolves the two ninhydrin-reactive diastereomers of S-(N-ethy1succinimido)-L-cysteine with retention times of 40 and 51 min, respectively. Crystalline adduct, synthesized independently (33), serves as a standard for analyzer calibration. An equal volume of the same sample is then applied and eluted in identical fashion with the exception that the effluent is collected without reacting it with ninhydrin and the radioactivity associated with each diastereomer is pooled. For each sample the specific radioactivity of each diastereomer was determined and these were averaged.
Reaction of Radiolabeled N-Ethylmaleimide with (Nu' + K+)-ATPase-Supernatant enzyme or SDS-purified enzyme (1 to 3 mg ml-'), prepared in the presence or absence of 2-mercaptoethanol, and suspended in 50 or 125 pl of the desired reaction buffer, was labeled by the addition of 10 or 25 pl of the stock radiolabeled N-ethylmaleimide solution, respectively. Two 5-pl aliquots were immediately withdrawn, each added to 100 pl of 0.2 M sodium citrate, pH 2.2, and the specific radioactivity and concentration of N-ethylmaleimide determined. In the eight experiments of this type, the molar concentration of radiolabeled N-ethylmaleimide was 3 to 4 mM while the specific radioactivity varied in an unpredictable fashion from 1 to 3 X IO4 cpm nmol" for N-[3H]ethylmaleimide and from 3 to 8 x lo3 cpm nmol" for N- [14C]ethylmaleimide. An aliquot of 5 p1 was also immediately removed, as were subsequent aliquots taken at various times, and quenched by the addition to 50 p1 of iced reaction buffer containing 25 nm 2-mercaptoethanol. These samples were later assayed for enzymatic activity in order to calculate the fractional loss.
The enzyme-labeling reaction was incubated at 37 "C under N2 and Three different methods which all gave identical results within experimental error were used for the quantitation of radiolabeled Nethylmalehide covalently bound to the large polypeptide. The first method utilized preparative gel electrophoresis. A protein sample (200 pg) was applied to a gel which was submitted to electrophoresis and then immediately scanned at 280 nm on a Zeiss PMQII spectrophotometer fitted with a linear transporting chamber. The region containing the a polypeptide was cut out of the gel in 5 to 10 slices (1 mm) and eluted with three changes (2 m l ) of 50 nm ammonium bicarbonate buffer, pH 7.8, 0.1% SDS, 25 m~ 2-mercaptoethanol at 37 "C over 24 h. Equivalent portions of blank gels, as controls, were also eluted. These samples were pooled, passed through a Millipore filter (0.45 p), lyophilized, and redissolved in 2001.11 of water. Triplicate aliquots were taken for both liquid scintillation and quantitative amino acid analysis. For the latter, norleucine was added as an internal standard and at least 10 amino acids were used to calculate protein concentration, correcting for the background in the control (~20%). The relative amount of each amino acid was consistent with the published composition of the a polypeptide (35). The molar ratio of incorporation was calculated using the independently determined specific radioactivity of the N-[3H]ethylmaleimide, the specific radioactivity of the a polypeptide, and the value of 121,000 for the molecular weight of the a polypeptide (36). This is presented as moles of ESC,' the cysteine adduct of radiolabeled N-ethylmaleimide/mol of a polypeptide. In addition to the direct method of measurement just described, a second method of quantitation was also used. Gels were stained for protein with Coomassie brilliant blue, scanned at 500 nm, and absorbance areas of a polypeptide were calculated. From one set of gels containing duplicate samples of enzyme labeled under each ligand condition, the a polypeptide region was sliced out and prepared for hydrolysis and quantitative amino acid analysis as described (36). From this data, an extinction coefficient for the large chain area measurements was determined, and with this extinction coefficient, the amount of a polypeptide present in all of the samples could be calculated. The radioactivity associated with a polypeptide was assessed by slicing the gel into I-mm slices and placing each slice in 5 ml of a scintillation fluid consisting of 50 ml of Protosol, 50 ml of Liquifluor and 1.0 liter of toluene, for 12 h at 50 "C. The recovery of radioactivity was greater than 90% and the quenching measured with blank slices, proved to be negligible. Finally, the third method used the values determined by the two other, more direct, methods. Three ligand conditions, whose incorporation levels were always the most consistent, were chosen as standards. At least one sample of each standard was included in each experimental set. The gels of each sample run in triplicate, were scanned at 500 nm, the areas associated with the a polypeptide calculated, and the corresponding regions * I t will be assumed that the products of the alkylation of the enzyme are ethylsuccinimidocysteine residues within the polypeptide chain.
sliced from each gel and radioactivity counted. The total counts per min were normalized to the scanned areas, and the molar ratios of Nethylmaleimide incorporated into each a polypeptide, for all experimental conditions, were calculated from the normalized values obtained for the standards and their known molar ratios. The relative values among the three standards in each experiment always agreed closely with expectation.

RESULTS
Inhibition of (Na+ + K+)-ATPase by N-EthyZmaleimide-Incubation of SDS-purified and nucleotide-free (Na+ + K+)-ATPase with 5 m N-ethylmaleimide at pH 7.5,37 "C, and in the presence of various combinations of ligands progressively inhibits its enzymatic activity. The combination and concentration of the ligands chosen here represent conditions in which saturation of a given ligand site or sites is ensured and which, as a result, produce a kinetic state and poise a majority of the enzyme in a discrete equilibrium conformation. Kinetic examinations of this inactivation, over periods of 30 to 60 min, were performed. The enzyme was preincubated in the presence or absence of ligands, and the reaction then was initiated by the addition of N-ethylmaleimide. Aliquots were withdrawn at 5-min intervals, the reaction quenched with 2-mercaptoethanol, and the samples were assayed for (Na' + K+)-ATPase activity. In all cases, except for that in which oligomycin was present, the inactivation reactions of the various conformational forms of the enzyme displayed pseudo-fiist order kinetics with respect to the concentration of active enzyme, indicating that the loss of activity is associated with the alkylation of a unique sulfhydryl group.
A comparison of the respective rate constants for the loss of enzymatic activity, therefore, permits this sulfhydryl group to take the role of a reporter of the conformation. Each of the pseudo-fist order rate constants was calculated, and they are presented in Table I in units relative to the slowest reaction.
These results demonstrate that, in the absence of enzyme phosphorylation, the nucleophilicity of the sulfhydryl group involved in the loss of eazymatic activity is governed mainly by the nature of the cation present and does not correlate with the conformational changes associated with each kinetic state of the enzyme. In the presence of K+, the rate of the alkylation of this sulfhydryl is about 2.5 to 5 times the rate in the presence of Na+ under the same conditions. Control experiments, involving alkylation in the presence of Na+, K+, or Tris at ionic strengths equivalent to that of 100 rn Na+, proceed at rates identical with those presented in Table I for the other concentrations of these cations. This observation indicates that the differences in covalent inhibition between the two cation substrates cannot be attributed to an ionic strength effect. The addition of M P to an otherwise equivalent solution also seems to decrease the rate by about 30%. It can also be seen that the binding of ATP in the absence of these cations, decreases the inactivation rate to values characteristic of alkylation in the presence of Na' .
The phosphorylation of the enzyme in the presence of either Na+, M e , and ATP, or Mg+, Pi, and strophanthidin, results in the E2-P kinetic state (4, 7, 4 4 , 45). Under either of these conditions, the enzyme activity is the least susceptible to alkylation, with 80% usually remaining after 30 min. If Mg2+ is omitted from the incubation with strophanthidin and Pi, the loss of activity is substantial, suggesting that actual phosphorylation is necessary for this protection of the critical When a less purified preparation of (Na' + KC)-ATPase, deoxycholate-treated, zonal gradient microsomes (28, 29) is incubated with N-ethylmaleimide in the presence of an ATP regeneration system under conditions required for Na+-dependent phosphorylation, the same slow rate of inactivation sulfhydryl group. Relatiue rates of the inactivation of (Nu+ + K+)-ATPase with N-ethylmaleimide (Na' + K+)-ATPase was added to several 0.2-ml solutions such that the final concentrations were 0.25 to 0.5 mg d-' protein, 100 mM NaC1, 10 m~ CDTA, 20 m~ KC1, 5 m~ MgClz, 10 n" ATP, 3 pM strophanthidin, 20 pg ml" oligomycin, 10 m~ P,, or an ATP-regenerating system consisting of 50 m?vf phosphoenolpyruvate, 50 mM MgC12, and 0.2 mg rn" pyruvate kinase, in the combinations noted. All samples contained 0.25 M sucrose, 30 m~ histidinium chloride, pH 7.5. The reaction was initiated by the addition of N-ethylmaleimide, present at a final concentration of 5 m~, and the samples were incubated at 37 O C for 30 min. An aliquot was immediately withdrawn after the addition of N-ethylmaleimide, with subsequent aliquots withdrawn at 5-min intervals, and the reaction was quenched. The samples were assayed for (Na+ + K')-ATPase activity and pseudofirst order rate constants (k) were calculated from hear plots of the logarithm of the per cent of remaining activity against time. They are presented in units relative to the slowest reaction (kA. The actual pseudo-first order rate constants varied between Pi + strophanthidin

5.0
Enzyme alone (-CDTA) EZ  These conditions represent experimental controls; the enzyme conformation has not been determined in the presence of these ligands.
e Protection is observed to the same degree as Na+ + ATP + Mg", however, the inactivation does not exhibit pseudo-first order kinetics.
is observed. If components of the regeneration system are omitted, however, substantial loss of enzymatic activity occurs. Since this more impure enzyme preparation contains substantial Mp-stimulated and strophanthidin-insensitive ATPase activity, ATP, when present alone, is rapidly hydrolyzed and the enzyme cannot maintain the phosphorylated conformation. This is supported further by the observation that phosphorylation by ATP of SDS-purified enzyme, a preparation containing very little Mg'-stimulated ATPase activity, gives identical results regardless of the presence or absence of the ATP regeneration system. With this preparation of enzyme, only about 15% of the ATP is hydrolvzed during the incubation under these conditions. It is likely that earlier studies of the alkylation (18, 20, 23, 46) of the ATPdependent phosphorylated state, utilizing impure preparations of enzyme, were complicated by losses of ATP similar to those described here.
High Na+ concentration (0.5 M) in addition to Mg2+ and ATP, conditions which produce the El -P kinetic state (39, 40, 47, 48), produces an enzyme form only slightly more nucleophilic than the E2-P. Oligomycin also is thought to stabilize the El -P form of the enzyme in the presence of Na+, M F , and ATP (8, 41, 42). Following alkylation in the presence of oligomycin, the (Na' + K')-ATPase activity remaining after 30 min was the same as that found in the E2-P form, but the kinetics was not pseudo-first order. Therefore, a quantitative comparison of the rate constants cannot be made in this case. Nevertheless, the sulfhydryl residue whose alkylation leads to inactivation of the enzyme displays almost equivalent nucleophilicity in both of the phosphorylated forms of the enzyme, El-P and E2-P, and therefore can detect no differences between them.
The experimental results obtained with other reporters of ligand-stabilized conformations are usually interpreted in terms of two basic forms of the enzyme, E1 and Ez, with the assumption that E, and El-P are conformationally equivalent, as well as Ez and Ez-P. The results reported here, however, demonstrate that the conformational responses detected by this activity-associated sulfhydryl group depends on ligand conditions which do not strictly correlate with the stabilization of either the El or E 2 forms.
Determination of Concentration and Specific Radioactiuity of Radiolabeled N-Ethylmaleimide-The amount of Nethylmaleimide incorporated by (Na' + K+)-ATPase while the enzyme is maintained in different ligand-stabilized kinetic states is a measure of the total accessible sulfhydryl groups. A successful evaluation of this quantity potentially could indicate structural changes affecting sulfhydryl groups other than the one associated with the loss of activity which was just described. Indeed, earlier experiments (18), which followed the incorporation of N-ethylmaleimide into (Na+ + K')-ATPase under various ligand conditions, suggested that this might be the case. At the outset of the present experiments, however, it became clear that there are serious difficulties involved in the estimation of the concentration and specific radioactivity of N-ethylmaleimide due to the tendency of the reagent to polymerize, hydrolyze, or evaporate. Therefore, an accurate method for determining the actual concentration and specific radioactivity of active, monomeric, radiolabeled N-ethylmaleimide during the reaction with (Na' + K+)-ATPase was necessary. The method developed is based on the reaction of N-ethylmaleimide with L-cysteine to form ethylsuccinimidocysteine (33). The two diastereoisomeric products are ninhydrin-positive and can be resolved by cation exchange chromatography on an amino acid analyzer. The molar concentration of ESC is quantitatively determined by the ninhydrin reaction and the effluent is collected to measure the associated radioactivity. From these two quantities the actual specific radioactivity of N-ethylmaleimide can be determined. Furthermore, since the reaction converts the N-ethylmaleimide quantitatively into ESC? the true molar concentration of Nethylmaleimide in the original sample is also established.
This assay was used to monitor the fate of the N-ethylmaleimide during various steps of the enzyme-labeling procedure (Table 11). These manipulations include transfer of the radio- Losses and changes in specific radioactivity of N-ethylmaleirnide resulting from preparation of alkylation mixtures A stock solution was prepared by transferring radioactive N-ethylmaleimide from the pentane solution supplied by the manufacturer to histidine/sucrose buffer. To this stock solution (Na+ + K+)-ATPase (1 to 3 mg m l -l ) was added to initiate an alkylation experiment. Samples were removed from the stock solutions or from the alkylation mixtures immediately following the addition of enzyme for direct determinations of the molar concentration and specific radioactivity of the N-ethylmaleimide in each of these solutions by the reaction of radiolabeled N-ethylmaleimide with L-cysteine. The values tabulated are calculated from these assays, the specifications listed by the manufacturer, or the actual mass of crystalline reagent used in the experiments. Step in procedure  labeled N-ethylmaleimide from pentane to reaction buffer, removal of the pentane, the addition of concentrated, nonradioactive N-ethylmaleimide to achieve the desired final concentration of the stock reagent solution, and the addition of enzyme to the stock reagent solution to initiate the alkylating reaction. As can be seen (Table 11), there is a loss of both radioactive and nonradioactive N-ethylmaleimide during the preparation of the stock solution as well as when the stock solution is mixed with the enzyme. This variability in recovery illustrates the complete impossibility of calculating the specific radioactivity indirectly and emphasizes the absolute necessity for a quantitative assay in studies which rely upon the specific radioactivity during sulfhydryl determinations with radiolabeled N-ethylmaleimide.
Alkylation of (Na' + F)-ATPase with Radiolabeled N-Ethyhaleimide-Two preparations of (Na+ + K')-ATPase, toethanol, denatured with SDS, and the polypeptides separated by electrophoresis on 5% SDS-polyacrylamide gels. The gels were scanned at 280 nm, and the areas corresponding to the (Y and / 3 chains were sliced and counted. Alternatively, the gels were stained, scanned at 500 nm, and sliced and counted. 23 Typical results are shown in Fig. 1. It can be seen that the great majority of the radioactivity is associated with the (Y Aliquots were removed from this reaction mixture immediately following the addition of enzyme for direct determinations of the molar concentration and specific radioactivity of N-ethylmaleimide, and to assay for (Na+ + K+)-ATPase activity. The samples were incubated for 2 h at 37 "C while aliquots were removed, the reaction quenched, and large chain purified. In some experiments, these aliquots were also assayed for (Na+ + K+)-ATPase activity and the per cent of remaining activity was calculated. The specific radioactivity of the purified large chain, representing the incorporation at each time interval, was determined by quantitative amino acid analysis and converted to moles of ESC for every 121,000 g of protein (mol of ESC (mol a)").  Table 1 1 1 show that after 30 min, an average of 4.0 mol of ESC (mol a)-' is found regardless of the preparation of enzyme, range of specific radioactivity, or absolute concentration of N-ethylmaleimide. This substantiates the reproducibility of the conformational state as well as the reliability of the specific radioactivity determinations. Simultaneously, enzyme activity had declined by 70% over this time period consistent with the results from the kinetic inactivations.
A more complicated alternative to the earlier conclusion that only one sulfhydryl is involved in the inactivation of the enzyme would be that a rate-limiting reaction with one sulfhydryl group could be followed by the rapid reaction of several others, some or all of which caused partial or complete activity loss. The only additional incorporation of N-ethylmaleimide, however, during the next 90 min, 0.2 to 0.3 mol of ESC (mol a)-', appears to correlate fairly closely with the loss of the remaining 20 to 30% of the activity (Table 111). This indicates that 3 to 3.5 sulfhydryl groups have become rapidly and stoichiometrically alkylated by 30 min and do not react further, while only one other, the one responsible for inactivation continues to react after 30 min. Certainly, the most likely conclusion at this time is that only one sulfhydryl participates in the inactivation of the enzyme.
It has been shown that, during the isolation of membranes, oxidation of sulfhydryl groups can occur. These can be regenerated with 2-mercaptoethanol leading to increased reactivity with N-ethylmaleimide (49). To examine this possible source of confusion, (Na' + K')-ATPase, purified in the presence of 10 mM 2-mercaptoethanol from the initial kidney dissection until a final dialysis just prior to the labeling reaction, was included in these time course studies. No difference in the   TABLE IV Alkylation of large chain of (Na' + h?)-ATPase with N-[3HJethylmaleimide (Na+ + K+)-ATPase was added to several 0.15-ml solutions such gether. The samples were incubated at 37 "C for 30 mi n, the reaction that the final concentrations were 2 mg ml" protein, 100 m~ NaC1, quenched, and large chain purified. Its specific radioactivity was 10 m~ CDTA, 20 RMI KC1,5 RMI MgC12,lO mM ATP, 3 p~ strophandetermined by quantitative amino acid analysis and converted to thidin, 20 pg ml" oligomycin, 10 m~ Pi, or an ATP-regenerating moles of ESC for every 121,000 g of protein (mol ESC (mol a)-'). The system consisting of 50 m~ phosphoenolpyruvate, 50 mM MgC12,0.2 moles of "alkylated active site-SH" were calculated from inactivation mg ml" pyruvate kinase, in the Combinations noted. All samples values assuming that the inactivation was due to the alkylation of contained 0.25 M sucrose, 30 m~ histidinium chloride, pH 7.5. N-[3H] only one of the enzyme's sulfhydryl groups. The difference between Ethylmaleimide was also present at 3 to 4 mM and its specific total alkylation and active site alkylation is presented as "alkylated radioactivity was determined immediately after mixing reagents toexternal-SH."

Moles ESF
No 1.9 f 0.2 final level of incorporation was noted when these precautions were taken (Table 111).
Since the ligand conditions used in these experiments produce the greatest levels of incorporation found in the studies now to be described, it follows that fluctuations in any of the variables examined in these initial experiments, namely Nethylmaleimide concentration, method of preparation, or the presence of a reducing reagent, would not produce significant changes in the levels of N-ethylmaleimide incorporation under any of the ligand conditions examined in these experiments. Nevertheless, enzyme was routinely purified by the SDS procedure, including 10 m~ 2-mercaptoethanol in all of the purification steps, and N-ethylmaleimide concentrations were kept as close to 3 to 4 m~ as possible. Finally, three different methods for determining the protein concentrations of a chain were employed, and these yielded equivalent values. If it is assumed that the loss of enzyme activity is due to the alkylation of a single sulfhydryl group, then the molar fraction of that sulfhydryl which has been alkylated is equal to the fraction of the activity which has been lost. This amount, subtracted from the total incorporation, yields values for those reactive sulfhydryl groups, in addition to the activity-associated sulfhydryl, available for alkylation. These residues are referred to as external sulfhydryl groups. There are 2.8 f 0.3 external sulfhydryl groups on the large chain of (Na+ + K+)-ATPase which react rapidly with N-ethylmaleimide when the enzyme is poised in the Ez kinetic state; 1.9 f 0.3, in the E*-P state; 1.8 f 0.3, in the E l state; and 1.9 f 0.2, in the E1 -P state. The results demonstrate that one additional sulfhydryl group, distinct from the one whose nucleophilicity is sensitive to levels of Na+ and K+ and whose alkylation inactivates the enzyme, is exposed upon stabilization of the enzyme in the EZ kinetic state, and demonstrates a unique conformational form of the enzyme.
There was no change in the low level of incorporation into the /3 subunit regardless of the ligand conditions, indicating that any conformational change occurring in the a chain that would affect the / 3 chain is not detected by sulfhydryl groups of the /3 subunit.

DISCUSSION
Four different sulfhydryl residues on the large chain of (Na+ + K+)-ATPase have been identified in the present experiments. The alkylation of one of them inactivates the enzyme. This one will be referred to as the activity-associated sulfhydryl. Two sulfhydryl groups are alkylated under all circumstances examined and will be referred to as the invariant sulfhydryl groups. Finally, one reacts only when the enzyme is alkylated while poised in the Ez state and this one will be referred to as the Ez sulfhydryl. These assignments serve to stress the fact that these residues are independent functionalities whose nucleophilicities respond to different changes in the state of the enzyme. Although the results presented here are interpreted to identify four sulfhydryl groups of the (Y polypeptide that react with N-ethylmaleimide, only the isolation of individual modified peptides from N-ethylmaleimide-labeled a polypeptide would confirm the rhetorical equality drawn here between the incorporation stoichiometry and the actual presence of four reactive sulfhydryl groups. An analogy of this situation, however, has been encountered in the case of the reaction of Nethylmaleimide with the sarcoplasmic reticulum Ca2+-ATPase, an enzyme whose single polypeptide closely resembles the a chain of the (Na+ + K')-ATPase (50, 51). 4 The analysis of peptides from Ca2'-ATPase (52) labeled with radioactive N-ethylmaleimide and digested with pepsin suggests that there is an equivalence between the absolute amount of Nethylmaleimide incorporated and the number of unique, reactive sulfhydryl groups in a protein closely related to (Na' The experiments described here utilize ligand combinations and concentrations which were recently demonstrated, in studies of other conformational probes, to stabilize exclu-sively5 a particular kinetic state of purified renal (Na+ + K+)-ATPase. It is under these conditions that the enzyme was reacted with an excess of either radiolabeled or nonradioactive N-ethylmaleimide. The pseudo-first order rate constants which describe the loss of enzymatic activity were measured and compared with the actual incorporation of radiolabeled N-ethylmaleimide under identical conditions. The results of these experiments demonstrate that the apparent nucleophilicity of the sulfhydryl residues of (Na+ + K+)-ATPase responds to two different signals. The rate of alkylation of the activity-associated sulfhydryl by N-ethylmaleimide is dependent on the presence of Na+ or K' or the phosphorylation of the enzyme (Table I). If the contribution of this group is subtracted from the total incorporation, it can further be shown that the reaction of the Ez sulfhydryl is controlled only by conformational state. Its alkylation, however, does not inactivate the enzyme.
The results of previous experiments which have followed ligand-associated structural changes of the a polypeptide by monitoring various reporters (Table V) have been interpreted in terms of a two-state isomerization of the enzyme between an El and an E2 conformation. It has been implied in these earlier reports that the Ez and phosphorylated E2-P kinetic states of the enzyme are conformationally equivalent and distinct from the El and El-P states (7, 13, 15, 16, 53). The results presented here, however, illustrate clear changes when the E2-P state becomes the Ez state of the enzyme. In particular, the nucleophilicity of both the Ez sulfhydryl (Table IV) and, less dramatically (Table I), the activity-associated sulfhydryl increase. Clearly, the transformation which produces a structural change in these regions of the catalytic subunit is not monitored by some of the other probes which have been employed. Furthermore, enzyme poised in the El or El-P state displays the same unreactivity at the Ez sulfhydryl as the Ez-P state even though structural differences between EZ-P and El or El-P at other locations of the protein have been observed by other methods (11, 13, 15, 16). Table V suggests that the evidence for only two exclusive conformational states has never been 5The conditions described in Tables I, IV, (15)(16)(17) phan fluorescence ity" sulfhydryl (9, 10, 12) groups very convincing. In particular, there are striking differences in ouabain affinity between E 2 and E2-P as well as E1 and El-P.

A close examination of
In fact, ouabain may not even bind to any unphosphorylated form of the enzyme (31,54,56-58). This observation has been discussed previously as evidence for the existence of conformational states other than the El and E 2 forms (56). Although there is evidence for differences in the intrinsic tryptophan fluorescence of the enzyme between E1 and El-P (40), all changes in intrinsic tryptophan fluorescence observed in the presence of ATP have been stated to be artifacts (59), further complicating interpretations of these results. In addition, direct comparisons between El-P and any other form very rarely have been made. The conclusion which can be reached from these considerations, as well as the results of the present report (Table IV), is that there are at least four distinct conformations through which the enzyme passes during turnover, each displaying a different set of properties. These are El, El-P, Ez-P, and Ez. If this is the case, the question of when in the kinetic sequence the enzyme changes from an outward-facing form to an inward-facing form, and vice uersa (N), remains an open one.
The possibility that even a fiih conformational state exists is raised by a comparison of the sulfhydryl reactivity in the presence of saturating Na' and ATP on the one hand and K' and ATP on the other (Table I). Although structural differences between the enzyme in the presence of Na' and K+ can be readily detected by changes in affinity for ATP (9-12, 37, 38), upon saturation with ATP, in the absence of M e , the conformation adopted when either Na+ or K' are present has been assigned as E1 (13, 15). In addition, major structural differences between the E2 state and that existing in the presence of saturating ATP and K' have been observed in sulfhydryl cross-linking studies (61, 62) and tryptic cleavage patterns (15, 17) which have also been interpreted as indicating that the enzyme is in the El state under these circumstances. Yet the experimental results presented here, as well as in earlier studies (20), show that even when ATP is present at saturation, alkylation of the activity-associated sulfhydryl is dramatically increased by the addition of K+ while that of the other accessible sulfhydryl groups remains unchanged (Table IV). This effect of K+ presumably also results from a structural difference in the enzyme which alters the environment about this sulfhydryl group, and which i s not detected by the other methods. This additional conformation of the enzyme, stabilized by saturating concentrations of K' and ATP, could possibly represent the contribution of a K+-occluded state (38).
The differences in the nucleophilicity of the activity-associated sulfhydryl in the presence of ATP and either K' and Na' also offers insight into the location of this sulfhydryl on the catalytic subunit. Earlier reports have implied that the protection of enzymatic activity observed in the presence of ATP was due to the steric hindrance of a critical sulfhydryl group located in the ATP binding site (22,25). It has not been, however, decided whether the protection observed in the presence of ATP is a steric or an allosteric effect. There is now strong evidence which demonstrates that (Na' + K+)-ATPase has a single, exclusive ATP binding site which is within the active site but which can be converted between high and low affinity forms depending on the presence of Na' or K+, respectively (12, 63). In the case of the inhibition of enzymatic activity caused by N-ethylmaleimide in the presence of K' and ATP (Table I), the ATP concentration utilized was saturating even considering the low affinity due to the simultaneous presence of K' . Therefore, the ATP binding site was as occupied as it was in the presence of Na' and ATP or ATP alone. It follows that the protection observed when Na' is exchanged for K' (Table I) cannot be due to any steric protection, afforded by the binding of ATP, of a sulfhydryl group located in the ATP binding site. An alternative location for the activity-associated sulfhydryl could be near the cation binding site or in an area around the ATP binding site which responds allosterically to cation binding. It is not clear which step of the kinetic cycle of the enzymatic reaction is blocked as a result of the alkylation of the activityassociated sulfhydryl by N-ethylmaleimide. The most consistent observation concerning this question is that N-ethylmaleimide inhibits the ATPase activity more rapidly than the Na+-dependent phosphorylation of the enzyme and that the properties of the phosphorylated intermediate change so that it can phosphorylate ADP in the presence of Na' while becoming less sensitive to K'-stimulated hydrolysis (8,23, 24,  42,55,64,65). These results have been interpreted as evidence for the existence of two phorphorylated states of the enzyme, El-P and E,-P, that are connected in the kinetic cycle by a conformational isomerization. It is this isomerization which is thought to be inhibited as a result of the alkylation of Nethylmaleimide (8, 42, 65). Various degrees of inhibition of Na+-dependent phosphorylation, however, have been observed in these studies, complicating this simple interpretation of the effects of N-ethylmaleimide (21, 23, 24, 55). When purified enzyme is incubated with N-ethylmaleimide in the presence of K' and ATP, the modifcation which occurs produces only a slight inhibition of Nac-dependent phosphorylation (10% loss after 30 m i n ) and substantial loss of ATPase activity (60% loss after 30 min) (24). In contrast to these results, it was observed that purified enzyme, at identical Nethylmaleimide concentration and pH but in the absence of ligands, lost the ability to become phosphorylated to the same extent as the inhibition of ATPase activity (80% loss after 30 min) (21).
It is possible to explain these observations in terms of the results presented here in Tables I and IV. If only the E2 sulfhydryl is alkylated, the enzyme remains fully active (Table   IV) and if only the activity-associated sulfhydryl is alkylated, as is the case when K' and ATP are present, the transition between El-P and E2-P is blocked, inactivating the enzyme (Table I) but not affecting Na+-dependent phosphorylation. If, and only if, however, both sulfhydryl groups are alkylated, as is the case when no ligands are present, Na+-dependent phosphorylation and overall turnover might be blocked coincidentally.
A coincident inhibition of ATPase activity and the ability of the enzyme to be phosphorylated is also produced by ATP analogues which alkylate sulfhydryl groups as well as the sulfhydryl oxidant DTNB (22, 26). Since the enzyme can be protected from both of these inhibitors by ATP, these results have been interpreted as evidence for the existence of a sulfhydryl group located within the ATP binding site. There is, in fact, a cysteine residue located in the sequence of the polypeptide very close to the aspartate residue that is phosphorylated (50,67). This cysteine residue, however, could not be alkylated when the enzyme was treated with a low concentration of N-ethylmaleimide (1 mM) at 0 "C, in the absence of ligands, a treatment which eliminated ATPase activity but not the ability of the enzyme to be phosphorylated (42). All of these observations taken together suggest that there exists within the ATP binding site a single sulfhydryl group whose alkylation results simultaneously in the loss of both Na+dependent phosphorylation and ATPase activity. Furthermore, this sulfhydryl group, necessarily unique from the 4 residues discussed here, is apparently less nucleophilic toward N-ethylmaleimide under most conditions than even the activity-associated sulfhydryl and requires an affinity reagent for its alkylation.
The reaction of sulfhydryl groups of the enzyme with reagents other than N-ethylmaleimide, also, results in inhibition of enzymatic activity (66,68-70). These reagents include mercurials (69-72), maleimide derivatives (68), and sulfhydryl oxidants (66). Although it is likely that some of the sulfhydryl groups which react with the members of this collection of reagents also participate in the reaction with N-ethylmaleimide, and in both cases may react stoichiometrically, it is not certain that this is the case. The different rates of inhibition of ATPase activity and Na+-dependent phosphorylation observed among these reagents, and again between each of these reagents and N-ethylmaleimide, do in fact indicate a substantial variation in the nucleophilicity of the sulfhydryl groups of the enzyme. It should be noted that, while there are 20 to 25 sulfhydryl groups on the (Y subunit (35), only a small percentage of them react with any one of these reagents. Finally, owing to the difference in sue, solubility, and nature of the various electrophiles, it is not clear if the reaction of one of them with a certain sulfhydryl will produce the same effect as the reaction of N-ethylmaleimide with that same sulfhydryl. Sorting out these various entanglements may be quite complicated.