Critical hydrophobic interactions between phosphorylation and actuator domains of Ca2+-ATPase for hydrolysis of phosphorylated intermediate.

Functional roles of seven hydrophobic residues on the interface between the actuator (A) and phosphorylation (P) domains of sarcoplasmic reticulum Ca2+-ATPase were explored by alanine and serine substitutions. The residues examined were Ile179/Leu180/Ile232 on the A domain, Val705/Val726 on the P domain, and Leu119/Tyr122 on the loop linking the A domain and M2 (the second transmembrane helix). These residues gather to form a hydrophobic cluster around Tyr122 in the crystal structures of Ca2+-ATPase in Ca2+-unbound E2 (unphosphorylated) and E2P (phosphorylated) states but are far apart in those of Ca2+-bound E1 (unphosphorylated) and E1P (phosphorylated) states. The substitution-effects were also compared with those of Ile235 on the A domain/M3 linker and those of T181GE of the A domain, since they are in the immediate vicinity of the Tyr122-cluster. All these substitutions almost completely inhibited ATPase activity without inhibiting Ca2+-activated E1P formation from ATP. Substitutions of Ile235 and T181GE blocked the E1P to E2P transition, whereas those in the Tyr122-cluster blocked the subsequent E2P hydrolysis. Substitutions of Ile235 and Glu183 also blocked EP hydrolysis. Results indicate that the Tyr122-cluster is formed during the E1P to E2P transition to configure the catalytic site and position Glu183 properly for hydrolyzing the acylphosphate. Ile235 on the A domain/M3 linker likely forms hydrophobic interactions with the A domain and thereby allowing the strain of this linker to be utilized for large motions of the A domain during these processes. The Tyr122-cluster, Ile235, and T181GE thus seem to have different roles and are critical in the successive events in processing phosphorylated intermediates to transport Ca2+.

Sarcoplasmic reticulum Ca 2ϩ -ATPase (SERCA1a) 1 is a representative member of P-type ion transporting ATPases and catalyzes Ca 2ϩ transport coupled with ATP hydrolysis (Fig. 1) (Refs. 1 and 2, and for recent reviews, see Refs. [3][4][5][6][7]. In the catalytic cycle, the enzyme is activated by binding of two Ca 2ϩ ions (E2 to E1Ca 2 , steps 1-2) and then autophosphorylated at Asp 351 with MgATP to form ADP-sensitive phosphoenzyme (E1P, step 3), which can react with ADP to regenerate ATP. Upon formation of this EP, the bound Ca 2ϩ ions are occluded in the transport sites. The subsequent isomeric transition to ADPinsensitive form (E2P) will result in a change in orientation of the Ca 2ϩ binding sites and reduction in their affinity and thus the Ca 2ϩ release into lumen (step 4). Finally, hydrolysis takes place and returns the enzyme into an unphosphorylated and Ca 2ϩ -unbound form (E2, steps 5 and 6). E2P can also be formed from P i in the presence of Mg 2ϩ and absence of Ca 2ϩ by reversal of its hydrolysis.
The enzyme has three cytoplasmic domains, nucleotide-binding (N), phosphorylation (P), and actuator (A) domains, and ten transmembrane helices (M1-M10) (Fig. 2). The organization of the cytoplasmic domains markedly differ between the atomic models for the Ca 2ϩ -bound intermediates and those for the Ca 2ϩ -released intermediates, especially in the orientation of the A domain (8 -13). Biochemical studies on the actual intermediates and their analogs indicated (14) that the large rotation of the A domain (by ϳ110°parallel to membrane) and its tight association with the P domain occur during the E1P to E2P transition and Ca 2ϩ release, being consistent with the structural change in E1⅐AlF x ⅐ADP (E1PCa 2 ⅐ADP analog) 3 E2⅐MgF 4 2Ϫ (E2⅐P i analog) shown in Fig. 2 as well as with the change in E1⅐AlF x ⅐ADP 3 E2⅐AlF 4 Ϫ (the analog for the transition state of E2P hydrolysis) (10 -13, 15). For further understanding of the Ca 2ϩ transport mechanism, it is therefore essential to identify structural elements involved in these events and reveal their actual roles in the Ca 2ϩ transport.
In this regard, we have recently found (16) that Tyr 122 on the loop connecting the A domain and the second transmembrane helix (A domain/M2 linker) is critical for the E2P hydrolysis. In the Ca 2ϩ -unbound structures E2⅐MgF 4 2Ϫ , E2⅐AlF 4 Ϫ , and E2(TG) (E2 fixed with thapsigargin), Tyr 122 is located on the interface of the associated A and P domains. Its aromatic ring is situated near the center of the seven hydrophobic residues clustered from the three regions, Ile 179 /Leu 180 /Ile 232 on the A domain, Val 705 /Val 726 on the P domain, and Leu 119 /Tyr 122 on the A domain/M2 linker, forming hydrophobic interactions (Tyr 122hydrophobic cluster, see Fig. 2 and supplemental Fig. I). The three regions are, however, far apart in the Ca 2ϩ -bound structures E1Ca 2 , E1⅐AMPPCP, and E1⅐AlF x ⅐ADP. It is then possible that the formation of Tyr 122 -hydrophobic cluster and the resulting A-P domain association may have important functions in the Ca 2ϩ transport. Interestingly, Ile 235 on the A domain/M3 linker and the essential T 181 GES loop of the A domain (17,18) are located in the immediate vicinity of Tyr 122 -hydrophobic cluster. It is therefore also of interest to know how these closely located three elements operate to achieve successive events in the transport cycle.
In the present study, we therefore explored by site-directed substitutions and kinetic analyses the possible roles of seven residues in Tyr 122 -hydrophobic cluster and further compared the substitution effects with those of Ile 235 and T 181 GES loop in the same series of experiments. Results revealed that all the seven hydrophobic residues in the Tyr 122 -cluster are critical for the E2P hydrolysis and further indicated that the Tyr 122 -cluster, Ile 235 , and T 181 GES loop have different roles and are critical in the successive events in processing phosphorylated intermediates to transport Ca 2ϩ .

EXPERIMENTAL PROCEDURES
Mutagenesis and Expression-The Stratagene QuikChange TM sitedirected mutagenesis method (Stratagene, La Jolla, CA) was utilized for the substitution of residues in the rabbit SERCA1a cDNA. The ApaI-KpnI or KpnI-SalI restriction fragments with the desired muta-tion were excised from the plasmid and ligated back into the corresponding region in the full-length SERCA1a cDNA in the pMT2 expression vector (19). The pMT2 DNA was transfected into COS-1 cells by the liposome-mediated transfection method. Microsomes were prepared from the cells as described previously (20). The "control microsomes" were prepared from COS-1 cells transfected with the pMT2 vector containing no SERCA1a cDNA. The amount of expressed SERCA1a was quantified by a sandwich enzyme-linked immunosorbent assay as described previously (21). Expression levels of wild-type SERCA1a and the mutants were 2-3% of total microsomal proteins.
Ca 2ϩ -ATPase Activity-The rate of ATP hydrolysis was determined at 25°C in a mixture containing 10 g/ml microsomal protein, 0.1 mM [␥-32 P]ATP, 1 M A23187, 0.1 M KCl, 7 mM MgCl 2 , 0.55 mM CaCl 2 , 0.5 mM EGTA, and 50 mM MOPS/Tris (pH 7.0). The Ca 2ϩ -ATPase activity was obtained by subtracting the Ca 2ϩ -independent ATPase activity, which was determined in the presence of 5 mM EGTA without added CaCl 2 , otherwise as above. The specific ATPase activity/mg of expressed SERCA1a protein was calculated from the amount of expressed SERCA1a and the Ca 2ϩ -ATPase activity of expressed SERCA1a, which was obtained by subtracting the Ca 2ϩ -ATPase activity of the control microsomes from that of the microsomes expressing SERCA1a. This background level with the control microsomes was as low as 3% of the activity of microsomes expressing the wild-type SECRA1a. The specific activity thus obtained with the wild type was 3.36 Ϯ 0.30 mol/min/mg SERCA1a protein (n ϭ 6).
Formation and Hydrolysis of EP-Phosphorylation of SERCA1a in microsomes with [␥-32 P]ATP or 32 P i and dephosphorylation of 32 P-labeled SERCA1a were performed under conditions described in the legends to Figs. 3-7. The reactions were quenched with ice-cold trichloroacetic acid containing P i . Rapid kinetics measurements of phosphorylation and dephosphorylation were performed with a handmade rapid mixing apparatus (22), otherwise as above. The precipitated proteins were separated by 5% SDS-PAGE at pH 6.0 according to Weber and Osborn (23). The radioactivity associated with the separated Ca 2ϩ -ATPase was quantitated by digital autoradiography as described (24). The amount of EP formed with the expressed SERCA1a was obtained by subtracting the background radioactivity with the control microsomes. This background was less than 5% of the radioactivity of EP formed with the expressed wild-type SERCA1a. The amount of EP/mg of SERCA1a protein was calculated from the amount of EP thus obtained and the amount of expressed SERCA1a.
Miscellaneous-Protein concentrations were determined by the method of Lowry et al. (25) with bovine serum albumin as a standard.   Data were analyzed by nonlinear regression using the program Origin (Microcal Software, Inc., Northampton, MA). Three-dimensional models of the enzyme were reproduced by the program VMD (26).

RESULTS
Ca 2ϩ -ATPase Activity and Formation of EP from ATP-The specific Ca 2ϩ -ATPase activity of the expressed SERCA1a was determined at 25°C. The alanine and serine substitutions of each of all the seven residues in the Tyr 122 -hydrophobic cluster (Leu 119 , Ile 179 , Leu 180 , Ile 232 , Val 705 , Val 726 , and Tyr 122 ) resulted in the very large to almost complete loss of the activity (with larger loss by the serine substitution in each residue) (Fig. 3A). The conservative phenylalanine substitution of Tyr 122 showed only moderate reduction as we observed previ-ously (16). The substitutions of Thr 181 , Gly 182 , and Glu 183 in the T 181 GES loop and of Ile 235 caused almost complete loss of the activity, whereas the substitution of Ser 184 showed only moderate reduction. The results observed with the T 181 GES loop essentially agree with the previous finding by Clarke et al. (17) and Clausen et al. (18).
All these mutants formed EP from ATP in the presence of saturating 100 M Ca 2ϩ (see the legend to Fig. 3 for the EP levels measured at steady state at 0°C), and thus all the mutants yielded sufficient levels of EP for kinetic characterization. The affinities for Ca 2ϩ determined with the EP formation in all the mutants were close to that in the wild type (Table I, K 0.5 ). The rates of the E2 to E1Ca 2 transition in steps 1 and 2 were then determined at pH 6 where the equilibrium between E1 and E2 is most shifted to E2 (27). In this experiment, the enzyme was first preincubated in the absence of Ca 2ϩ and then phosphorylated by simultaneous addition of saturating Ca 2ϩ and ATP. The time courses of EP formation were well described by the first order kinetics (data not shown), and the rates obtained were summarized in Table I. When ATP was added to the enzyme preincubated with Ca 2ϩ (E1Ca 2 ) otherwise as above, the EP formation was much faster, therefore the rates obtained above actually reflect the rate-limiting E2 to E1Ca 2 transition. The substitutions of the hydrophobic residues Leu 119 , Ile 179 , Val 705 , Val 726 , and Tyr 122 somewhat reduced the rate, but the extents of reduction were much less than those in the ATPase activity. The substitutions of Thr 181 and Glu 183 on the T 181 GES loop moderately reduced the rate, whereas those of Gly 182 and Ser 184 did not affect much. The effect of substitution of Glu 183 actually agreed with the previous observation by Clausen et al. (18). The substitutions of Ile 235 on the A domain/M3 linker had essentially no effect on the rate of E2 to E1Ca 2 transition.
The E1P to E2P Transition and Decay of EP Formed from ATP-We then analyzed the E1P to E2P transition at 0°C. In Fig. 3B and C, the amount of ADP-insensitive EP (E2P) accumulated was determined at 15 s after addition of ATP (nearly the steady state). In the presence of K ϩ , which strongly accelerates the hydrolysis of E2P and thus suppresses its accumulation in the wild type (28), the fraction of E2P accumulated in the mutants for Ile 235 and the T 181 GES loop was very low or even lower than in the wild type. On the other hand, the mutants for the Tyr 122 -hydrophobic cluster (especially the serine substitution mutants) accumulated a substantial amount of E2P. In the absence of K ϩ , the E2P accumulation largely increased in the wild type and even to a much higher level in the mutants for the Tyr 122 -hydrophobic cluster. By contrast, the amount of E2P accumulated in the mutants for Ile 235 and the T 181 GES loop was still very low in the absence of K ϩ , although E183A accumulated to some extent. These results suggest that the E1P to E2P transition is not inhibited in the mutants for the Tyr 122 -hydrophobic cluster but strongly inhibited in the mutants for Ile 235 and T 181 GES loop (in E183A to some extent).
In Fig. 4, the time course of accumulation of E2P from E1Ca 2 and ATP was determined in the absence and presence of K ϩ with the mutants for the Tyr 122 -hydrophobic cluster, which were shown above to accumulate the large amount of E2P. The total amount of EP reached its maximum level very rapidly (within ϳ1 s) and remained unchanged during the period of observation (data not shown) and thus the time course actually reflects the accumulation of E2P from E1P. The E2P accumulation apparently proceeded with first order kinetics, and the rates obtained are summarized in Table I. In the absence of K ϩ (Fig. 4, A and B), the wild type accumulated a fair amount of E2P and therefore the time courses of the mutants could be compared with that of the wild type. The rates of the mutants for the Tyr 122 -hydrophobic cluster in the absence of K ϩ were comparable to or even slightly faster than that of the wild type, although it was reduced to some extent in the specific mutants such as L119A, V705S, and V726S. In each of the mutants for the Tyr 122 -cluster, the apparent rate observed in the presence Solid lines show the least squares fit to a single exponential, and the apparent rates to reach the steady-state E2P level thus obtained are given in Table I.   TABLE I Kinetic parameters determined for partial reaction steps The affinity of the transport sites for Ca 2ϩ (K 0.5 ) and the Hill coefficient (n) were determined by the EP formation from ATP in the presence of various concentrations of Ca 2ϩ under the conditions otherwise as described in the legend to Fig. 3B and by the least square fit to the Hill equation. The rates for the E1PCa 2 formation from E2 and ATP (i.e. the rate-limiting E2 to E1Ca 2 transition in steps 1 and 2) were determined at the saturating 100 M Ca 2ϩ at 0°C. The rates for the partial reaction steps were obtained at 0°C in the experiments in Fig. 4 (loss of ADP sensitivity, i.e. accumulation of E2P from E1P in step 4), Fig. 5 (decay of EP formed from ATP (EP ATP )), and Fig. 6 (hydrolysis of E2P formed from P i (E2P Pi ) in steps 5 and 6). The rates for the decay of EP ATP and for the hydrolysis of E2P Pi were also determined at 25°C in Fig. 7 and are shown in the last two columns. In parenthesis, the values obtained with the wild type are normalized to 100%.   . 3B). d Almost all of EP accumulated was E1P (see Fig. 3C). e The rate most likely reflects the rate-limiting E1P to E2P transition in step 4.
of K ϩ (Fig. 4, D and E, Table I) was similar to or somewhat slower than that in the absence of K ϩ . Results nevertheless indicate that the substitutions of the residues of Tyr 122 -hydrophobic cluster do not inhibit the E1P to E2P transition, and therefore the large accumulation of E2P in these mutants may be due to possible inhibition of the subsequent hydrolysis of E2P. This kind of analysis was not possible with the mutants for Ile 235 and the T 181 GES loop, because they do not accumulate E2P (see Fig. 3, B and C) except the mutant E183A in the absence of K ϩ as it accumulated to some extent with the reduced rate (Fig. 4C and Table I). For these mutants, therefore, the decay of E1P was determined in the presence of K ϩ instead of the E2P accumulation from E1P (Fig. 5C). The decay time courses were well fitted with a single exponential and the rates obtained were summarized in Table I. This EP decay most likely reflects the rate-limiting E1P to E2P transition (as in the wild type (28)). In fact almost all of EP present at each of the time points in the EP decay in these mutants was E1P (data not shown) as well as at the start of the decay reaction (see Fig.  3C). In all the mutants for Ile 235 and the T 181 GES loop, the decay rate was markedly reduced to 2-4% (or 10% in S184A) of that of the wild type. The results indicate that Ile 235 on the A domain/M3 linker and the residues in T 181 GES loop are critical for the E1P to E2P transition. The conclusion for Thr 181 , Gly 182 , and Glu 183 essentially agrees with the original finding (17). We also found the markedly slowed E1P to E2P transition with S184A at 0°C despite its fairly high ATPase activity at 25°C (in which the rate-limiting step is likely the E1P to E2P transition). It is possible that thermal motion of the T 181 GES loop affects the contribution of Ser 184 to the E1P to E2P transition.
The same set of experiments was done with the mutants for the Tyr 122 -hydrophobic cluster but in this case was done to examine rather the possible inhibition of decay of EP after the E1P to E2P transition (Fig. 5, A and B, and Table I). Time courses were apparently fitted very well with a single exponential although a substantial fraction of the total amount of EP was E2P in these mutants at the start of decay reaction (see Figs. 3B and 4, D and E). In all the mutants for Tyr 122 -hydrophobic cluster, the EP decay was strongly or almost completely inhibited (with the stronger inhibition by the substitution with serine than with alanine in each residue). The EP decay was much slower than the E1P to E2P transition in each of the mutants (compare with Fig. 4, note the difference in time scale). Thus, all the hydrophobic residues forming the Tyr 122cluster are critically important for the decay of EP after the E1P to E2P transition.  Table I.

FIG. 6. Hydrolysis of E2P formed from P i without Ca 2؉ . A-C,
Microsomes expressing the wild type or mutant were phosphorylated with 32 P i at 25°C for 10 min in 10 l of a mixture containing 2 g of microsomal protein, 0.1 mM 32 P i , 5 mM EGTA, 7 mM MgCl 2 , 50 mM MOPS/Tris (pH 7.0), and 35% (v/v) Me 2 SO. The mixture was then cooled and diluted at 0°C by addition of 0.19 ml of a mixture containing 2.1 mM non-radioactive P i , 105 mM KCl, 5 mM EGTA, 7 mM MgCl 2 , and 50 mM MOPS/Tris (pH 7.0). At different times after the dilution, hydrolysis was quenched with trichloroacetic acid. The time courses of the wild type (WT) and mutants are indicated with different symbols in each panel of the figure. The amounts of EP formed with 32 P i at zero time are normalized to 100%. Solid lines show the least squares fit to a single exponential, and the rates thus obtained are given in Table I.

Hydrolysis of E2P Formed from P i -The hydrolysis of E2P
was therefore directly examined by first phosphorylating the enzyme with 32 P i in the absence of Ca 2ϩ and K ϩ and presence of 35% (v/v) Me 2 SO, which extremely favors E2P formation (29), and then by diluting the phosphorylated samples at 0°C with a large volume of solution containing non-radioactive P i and K ϩ without Ca 2ϩ (Fig. 6). The conditions were thus made otherwise the same as those used for the E2P accumulation from ATP in Fig. 4 and for the decay of EP formed from ATP with Ca 2ϩ in Fig. 5. Hydrolysis of 32 P-labeled E2P proceeded with first order kinetics, and the rates obtained were summarized in Table I. The hydrolysis was almost completely inhibited by the substitutions of the residues of Tyr 122 -hydrophobic cluster, Ile 235 on the A domain/M3 linker, and Glu 183 in the T 181 GES loop. Thus all these residues are critical for the rapid E2P hydrolysis. The result on Glu 183 is consistent with that by Clausen et al. (18) who clearly demonstrated the essential role of Glu 183 in the hydrolysis and predicted that this residue coordinates the attacking water molecule, as in fact depicted on the structure E2⅐MgF 4 2Ϫ (12) and E2⅐AlF 4 Ϫ (13). We found further that the hydrolysis rate was, in contrast, significantly increased by the substitutions of Thr 181 and Ser 184 and not much altered by that of Gly 182 . Thus, the essential catalytic role of the T 181 GES loop in the E2P hydrolysis is confined to Glu 183 .
In Fig. 7, we performed the kinetic analysis also at 25°C with the mutants for the Tyr 122 -hydrophobic cluster and observed that both the hydrolysis of E2P formed from P i and the decay of EP formed from ATP were strongly slowed by the substitutions in the Tyr 122 -cluster and thus much slower than the E1P to E2P transition that occurred within ϳ1 s. In the mutants L119S, I179S, and I232A, the decay of EP formed from ATP determined at 25°C was even much slower than the E1P to E2P transition at 0°C (Table I). We also noticed at 25°C in Fig. 7 that the decay of EP formed from ATP was substantially slower than the hydrolysis of E2P formed from P i in each of the mutants. This is typically shown with L119S, I179S, L180A, Y122A, and the mutants for Val 705 and Val 726 . Thus it may be possible that in the EP decay of the mutants for the Tyr 122hydrophobic cluster, there is a rate-limiting step after the E1P to E2P transition but before the E2P hydrolysis (see the last section under "Discussion"). DISCUSSION In this study, we explored functional roles of the seven hydrophobic residues clustered on Tyr 122 at the A-P domain interface in the Ca 2ϩ -unbound structures of Ca 2ϩ -ATPase (Tyr 122 -hydrophobic cluster, Figs. 2 and 8, and see supplemental Fig. I for details of their hydrophobic interactions by LIGPLOT (44)). The effects of substitutions were also compared with those of the juxtaposed T 181 GES loop and Ile 235 on the A domain/M3 linker. We found that all the substitutions in Tyr 122 -cluster almost completely inhibited the decay of EP after the E1P to E2P transition, and those of Ile 235 and the T 181 GES loop blocked the E1P to E2P transition (Figs. 3-5, Table I). The hydrolysis of E2P was in fact markedly slowed by the substitutions in Tyr 122 -cluster, and it was blocked also by those of Glu 183 and Ile 235 (Fig. 6, Table I). The results indicate that the closely located elements Tyr 122 -hydrophobic cluster, T 181 GES loop, and Ile 235 play crucial and respective roles for these catalytic processes to transport Ca 2ϩ . As discussed in the following sections, they seem to participate in successive events that involve large motions of the A and P domains, changes in interactions between these domains, and formation of the proper configuration of the catalytic site with the phosphatase function.
E2P Hydrolysis and Tyr 122 -Hydrophobic Cluster-In E2⅐MgF 4 2Ϫ , E2(TG), and E2⅐AlF 4 Ϫ , Tyr 122 -hydrophobic cluster is formed from the seven residues in three regions at the A-P domain interface (see Figs. 2 and 8) and thus likely produces the tight association between these domains through the hydrophobic interactions in the cluster. In the actual E2P, the A and P domains are tightly associated as in E2⅐MgF 4 2Ϫ and E2⅐AlF 4 Ϫ (14,30). The observed block of E2P hydrolysis by the substitutions in Tyr 122 -cluster therefore indicates that the formation of the Tyr 122 -cluster and the resulting A-P domain association are essential for formation of the catalytic site of E2P as the phosphatase. Because the substitutions of any single residue in Tyr 122 -cluster blocked the E2P hydrolysis, even the small perturbation of Tyr 122 -cluster seriously affects the proper configuration of catalytic site.
The catalytic site of E2P requires a precisely positioned water molecule that attacks the acylphosphate in the in-line associative mechanism (31) and a catalytically essential Mg 2ϩ . They are probably coordinated by Glu 183 on the T 181 GES loop (the water molecule) (18) and by the DGV 705 ND loop (Mg 2ϩ ) (32)(33)(34), as depicted in E2⅐MgF 4 2Ϫ (12) and E2⅐AlF 4 Ϫ (13) (see supplemental Fig. II). The T 181 GES loop is directly connected to Ile 179 /Leu 180 in the Tyr 122 -cluster, and Val 705 in the DGV 705 ND loop is involved in the cluster. It is therefore likely that the formation of Tyr 122hydrophobic cluster fixes Glu 183 and the DGV 705 ND loop at their precise positions, and thus the substitution-induced perturbation of the Tyr 122 -cluster disrupted such precise configuration of the catalytic site and inhibited the E2P hydrolysis. The consequence of substitutions of the Tyr 122 -cluster is consistent with the previous finding (35) that the proteolytic cleavage at Leu 119 (which is in the Tyr 122 -cluster) blocks the E2P hydrolysis but does not inhibit the E1P to E2P transition.
The E1P to E2P Transition and T 181 GES Loop-In contrast to the observed block of E2P hydrolysis, none of the substitutions in the Tyr 122 -hydrophobic cluster inhibited the E1P to E2P transition. The E1P to E2P transition was, on the other hand, blocked by the substitutions in the T 181 GES loop in the immediate vicinity of the Tyr 122 -cluster. This loop is located at FIG. 7. Comparison at 25°C between the decay of EP formed from ATP with Ca 2؉ and the hydrolysis of E2P formed from P i without Ca 2؉ . A-D, For all the mutants of the Tyr 122 -hydrophobic cluster, the decay of EP formed from ATP with Ca 2ϩ (open symbols) and the hydrolysis of E2P formed from P i without Ca 2ϩ (closed symbols) were determined at 25°C otherwise as described in the legends to Figs. 5 and 6, respectively. The two time courses are thus determined under otherwise the same conditions for each of the mutants. The amounts of EP formed at zero time are normalized to 100%. Solid lines show the least squares fit to a single exponential, and the rates obtained are given in Table I. In the wild type, the decay of EP formed from ATP and the hydrolysis of E2P formed from P i were completed within 1 s (data not shown), being consistent with the rapid turnover of the wild type at 25°C (16.9 s Ϫ1 as calculated from the ATPase activity and the amount of EP (described in the legend to Fig. 3)).
FIG . 8. Motions and associations of the A and P domains involving Tyr 122 -hydrophobic cluster, Ile 235 , and T 181 GES in the Ca 2؉ transport cycle. The coordinates for the structures E1Ca 2 , E1⅐AlF x ⅐ADP (E1PCa 2 ⅐ADP analog), E2⅐MgF 4 2Ϫ (E2⅐P i analog), and E2(TG) (E2 fixed with thapsigargin) were obtained from the Protein Data Bank (accession code 1SU4, 1WPE, 1WPG, and 1IWO, respectively (8, 9, 11, 12)) and the outermost surface of the enzyme in E1⅐AlF x ⅐ADP but docked on the phosphorylation site in the P domain forming a hydrogen-bonding network in E2⅐MgF 4 2Ϫ and E2⅐AlF 4 Ϫ (Refs. 10 -13, see also supplemental Fig. II). It was predicted that the docked T 181 GES loop sterically inhibits the access of ␤-phosphate of ADP to the covalently bound phosphate at the phosphorylation site, causing the loss of ADP sensitivity (12). The observed block of the E1P to E2P transition in the mutants for the T 181 GES loop is consistent with this view. Thus during the E1P to E2P transition, the A domain largely rotates to bring the T 181 GES loop to the phosphorylation site for their docking (Fig. 8). Importantly, our results further revealed that the formation of Tyr 122 -hydrophobic cluster in the immediate vicinity of the T 181 GES loop is not crucial for these events but will become essential after the E1P to E2P transition for producing the catalytic site of phosphatase as discussed above. 2 With respect to Glu 183 , we observed here that both the E1P to E2P transition in the presence of K ϩ and the hydrolysis of E2P formed from P i were almost completely inhibited by its substitution (Figs. 3B and 5C, Table I). Our results indicate that Glu 183 likely contributes first to the proper docking of the T 181 GES loop on the phosphorylation site for the E1P to E2P transition and then performs its essential function for the E2P hydrolysis by coordinating an attacking water molecule (18). 3 For the E2P hydrolysis, Glu 183 is likely fixed at the precise position by the formation of Tyr 122 -hydrophobic cluster as discussed above. It is also noteworthy that the E2P hydrolysis was accelerated by the substitution of Thr 181 and Ser 184 (Table I), therefore these residues are not essential for hydrolysis but may be stabilizing E2P in its ground state of the hydrolysis rather than the transition state.  Fig. I). Ile 235 may therefore be contributing to the processes of EP with these likely hydrophobic interactions. Importantly, the A domain/M3 linker is strained in E1P and in E2P, as noted in the crystal structure (12) and as indicated by the complete resistance of this linker against proteinase K attack in these states (14,15,30). It has been predicted that the strain of the linker is a major driving force for the large rotation of the A domain parallel to membrane in the E1P to E2P transition (12). The E2P hydrolysis to E2 also involves significant motion of the A domain (12-14, 30, 35, 36), in this case, its tilting perpendicular to membrane (by ϳ30°in E2⅐MgF 4 2Ϫ 3 E2(TG), see Fig. 8a) (12,13). It is then possible that the hydrophobic interactions between Ile 235 on the A domain/M3 linker and the A domain function to sustain the strain of this linker in E1P and in E2P, thereby allowing the strain of this linker to be utilized for large motions of the A domain during these processes. The removal of hydrophobic side chain of Ile 235 by the substitutions therefore disrupted such interactions and inhibited the strain-induced motions of the A domain essential for the E1P to E2P transition and for the E2P hydrolysis.
This view is consistent with the previous observation (37) that proteolytic cleavage of the A domain/M3 linker at Thr 242 (see Fig. 2) blocks both the E1P to E2P transition and the E2P formation from P i (reversal of E2P hydrolysis). The view is also compatible with the observed inhibition of these processes by the substitution of Gly 233 at the N terminus of this linker (38). Gly 233 may be important for the proper configuration of this linker at its junction with the A domain. For the E2P hydrolysis, it is also possible that the substitutions of Ile 235 disrupted its interactions with Val 705 /Ile 179 /Leu 180 /Ile 232 in the Tyr 122cluster and thus disrupted the catalytic site with the precisely positioned DGV 705 ND loop and Glu 183 . In Na ϩ ,K ϩ -ATPase, the E1P-E2P equilibrium was recently shown to be shifted in favor of E1P by the substitution of Ile 265 (corresponding Ile 235 of Ca 2ϩ -ATPase), and also the importance of interaction of Ile 265 with Val 714 /Asn 715 in E2P (corresponding Val 705 /Asn 706 in Ca 2ϩ -ATPase) was suggested (39).
Possible Contribution of Tyr 122 -Hydrophobic Cluster to Ca 2ϩrelease Process-The Ca 2ϩ release into lumen from the trans-2 It is also noteworthy that the substitution of Gly 233 connected directly to Ile 232 in the Tyr 122 -cluster was previously found to block the E1P to E2P transition (38), thus being in contrast to the substitutions of Ile 232 (Tyr 122 -cluster) that do not inhibit the E1P to E2P transition. Since Gly 233 is located on the A domain/M3 linker (at its N terminus), Gly 233 may be critical for the possible structural role of A domain/M3 linker (see the section for "Ile 235 on A Domain/M3 Linker" under "Discussion"). 3 It should be noted that Glu 183 was originally found by Clarke et al. (17) to be critical for the E1P to E2P transition, whereas it was not the case in the recent study by Clausen et al. (18) who nevertheless clearly demonstrated the essential function of this residue in the E2P hydrolysis. We observed in this study that the E1P to E2P transition was strongly inhibited in E183A in the presence of K ϩ but not so seriously in its absence (as this mutant accumulated E2P from E1P in the absence of K ϩ although with a significantly reduced rate, see Fig. 4C). Thus, we concluded that in the presence of K ϩ , Glu 183 functions also for the E1P to E2P transition. The essential contribution of Glu 183 to the E1P to E2P transition, i.e. the docking of the T 181 GES loop on the P domain in the presence of K ϩ may possibly be related to the recent prediction (43) that K ϩ bound specifically at the P domain largely affects the associated state of the A and P domains. placed according to the reaction cycle in Fig. 1. For simplicity, only the A (green) and P (orange) domains are depicted. a, the side view of the structure as in Fig. 2; b, the top view from the cytoplasmic side. The junction of the P domain with the N domain is indicated by "To N" and "From N". Large arrows indicate the motions of the A and P domains to realize the next structural state. The phosphorylation site (Asp 351 ), Mg 2ϩ ion (dark green sphere) bound at the phosphorylation site, and the P i analogs AlF 3 (11). In E1⅐AlF x ⅐ADP 3 E2⅐MgF 4 2Ϫ , which probably designates essential changes for the Ca 2ϩ release into lumen (12), the A domain largely rotates parallel to the membrane plane (by ϳ110°) and associates with the inclined P domain. Thus in E2⅐MgF 4 2Ϫ , the Tyr 122 -hydrophobic cluster is formed (cyan circle, see supplemental , the A domain tilts perpendicular to membrane (by ϳ30°) and the Tyr 122 -hydrophobic cluster works as a platform for the tilting. The tilting of the A domain during the E2P hydrolysis to E2 is probably coupled with the closure of lumenal Ca 2ϩ gate to prevent Ca 2ϩ leakage (12). port sites of EP is thought to be induced by the large motions of the A and P domains and their tight association and by the transmission of these changes to the transmembrane domain through M1/M2/M3 connected to the A domain and M4/M5 connected to the P domain to open the lumenal gate (12,14). E2P formed form P i without Ca 2ϩ has the opened lumenal gate (15,40) and represents probably the state immediately after the Ca 2ϩ release. Obviously the docking of T 181 GES loop on the phosphorylation site upon the large rotation of the A domain during the E1P to E2P transition is a necessary (first) event for the Ca 2ϩ release (Fig. 1), but we wondered whether this event alone is sufficient for the Ca 2ϩ release, since the formation of Tyr 122 -hydrophobic cluster and the resulting A-P domain association were indicated to be critical for the proper structure of E2P. The Tyr 122 -hydrophobic cluster is seen in all the Ca 2ϩunbound structures but not in the Ca 2ϩ -bound structures. Then an interesting question is whether the formation of Tyr 122 -hydrophobic cluster is involved in the Ca 2ϩ -release process, more specifically, whether the cluster is formed during the hydrolysis of E2P from its ground state to transition state (i.e. the cluster is formed after the Ca 2ϩ release) or formed already in E2P of its ground state (in this case, formation of the cluster may be involved in the Ca 2ϩ -release process).
In this regard, it is of interest to note the EP decay kinetics at 25°C with the mutants for the Tyr 122 -cluster; the decay of EP formed from ATP with Ca 2ϩ was much slower than the E1P to E2P transition (the loss of ADP sensitivity) and also substantially slower than the hydrolysis of E2P formed form P i without Ca 2ϩ (Fig. 7 and Table I). The results suggest a kinetic limitation in a step after the loss of ADP sensitivity, but before the E2P hydrolysis in these mutants, and can be accounted for by the previously proposed reaction sequence: E1PCa 2 3 E2PCa 2 3 E2P ϩ 2Ca 2ϩ , then E2P ϩ H 2 O 3 E2 ϩ P i (41). In this sequence, the Ca 2ϩ -releasing step E2PCa 2 3 E2P ϩ 2Ca 2ϩ after the loss of ADP sensitivity (E1PCa 2 3 E2PCa 2 ) came to be rate-limiting by the substitution-induced inhibition (although the E2P hydrolysis was also strongly slowed). We therefore suggest that the formation of Tyr 122 -hydrophobic cluster is important not only for the proper configuration of E2P as the phosphatase but also for the Ca 2ϩ release.
Importantly, the associated A and P domains in E2⅐MgF 4

2Ϫ
and E2⅐AlF 4 Ϫ are interacting also at Val 200 on which polar residues of the two domains gather to form a hydrogen-bonding and ionic interaction network (12) (see Fig. 8 (42) are essentially the same as those of the substitutions in the Tyr 122 -cluster observed here. It is therefore tempting to speculate that the tight associations of the A and P domains produced at the two regions (Tyr 122 -cluster and Val 200 region) in addition to the docking of T 181 GES loop on the P domain accomplish the Ca 2ϩ release into lumen and produce the catalytic site with the phosphatase function.