Deletions of Any Single Residues in Glu40-Ser48 Loop Connecting A Domain and the First Transmembrane Helix of Sarcoplasmic Reticulum Ca2+-ATPase Result in Almost Complete Inhibition of Conformational Transition and Hydrolysis of Phosphoenzyme Intermediate*

Possible roles of the Glu40-Ser48 loop connecting A domain and the first transmembrane helix (M1) in sarcoplasmic reticulum Ca2+-ATPase (SERCA1a) were explored by mutagenesis. Deletions of any single residues in this loop caused almost complete loss of Ca2+-ATPase activity, while their substitutions had no or only slight effects. Single deletions or substitutions in the adjacent N- and C-terminal regions of the loop (His32-Asn39 and Leu49-Ile54) had no or only slight effects except two specific substitutions of Asn39 found in SERCA2b in Darier's disease pedigrees. All the single deletion mutants for the Glu40-Ser48 loop and the specific Asn39 mutants formed phosphoenzyme intermediate (EP) from ATP, but their isomeric transition from ADP-sensitive EP (E1P) to ADP-insensitive EP (E2P) was almost completely or strongly inhibited. Hydrolysis of E2P formed from Pi was also dramatically slowed in these deletion mutants. On the other hand, the rates of the Ca2+-induced enzyme activation and subsequent E1P formation from ATP were not altered by the deletions and substitutions. The results indicate that the Glu40-Ser48 loop, with its appropriate length (but not with specific residues) and with its appropriate junction to A domain, is a critical element for the E1P to E2P transition and formation of the proper structure of E2P, therefore, most likely for the large rotational movement of A domain and resulting in its association with P and N domains. Results further suggest that the loop functions to coordinate this movement of A domain and the unique motion of M1 during the E1P to E2P transition.

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 and 4). In the catalytic cycle, the enzyme is activated by binding of two Ca 2ϩ ions (E2 to E1Ca 2 , steps 1 and 2) and then autophosphorylated by MgATP to form ADP-sensitive phosphoenzyme (E1P, step 3). Upon formation of this EP, the bound Ca 2ϩ ions are occluded in the transport sites. The subsequent isomeric transition to ADP-insensitive form (E2P, step 4) will result in a reduction in affinity and a change in orientation of the Ca 2ϩ binding sites, and thus a Ca 2ϩ release into lumen (step 5). Finally, hydrolysis takes place and returns the enzyme into an unphosphorylated and Ca 2ϩ -unbound form (E2, step 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 (N, P, and A), which are widely separated in the Ca 2ϩ bound form (E1Ca 2 ) and associated in the Ca 2ϩ -unbound and thapsigargin-bound form (E2(TG)) (5, 6) (Fig. 2). The modeling of tubular crystals formed with decavanadate (E2V) revealed (5) that three cytoplasmic domains gather to form a most compactly organized single headpiece in E2V. With the limited proteolysis experiments, we previously showed (7,8) that E2V is very similar to E2P in domain organization and that E2P is the intermediate having the most compactly organized headpiece in the catalytic cycle. The results further indicated that a large rotation of A domain (by ϳ90° (5)) and its strong association with P and N domains most likely occur during the E1P to E2P transition and suggested that stabilization energy provided by intimate contacts between all three cytoplasmic domains in E2P will provide energy for moving transmembrane helices and release the bound Ca 2ϩ ions.
It is thus crucial to find out structural elements essential for the A domain movement and resulting domain organization and for transmitting these changes to transmembrane helices. We have recently identified the Lys 189 -Lys 205 outermost loop of A domain as to make intimate contact with P domain for formation of the proper structure of Ca 2ϩ -released form of E2P in step 5 (9). P domain was actually documented with the crystal structures E1Ca 2 and E2(TG) to function as a coordinator for transmitting the movements of the transmembrane helices to the cytoplasmic domains (6). In addition, it is also possible that the loops connecting A domain with the transmembrane helices (M1-M3) may play roles in the movement of A domain and in transmitting the movement to transmembrane helices. The likely importance of interactions of M2-and M3-connecting loops with P domain was previously pointed out (6,10), and in fact the proteolytic cut or mutation of the M3-connecting loop at or near the interaction sites was shown to cause almost complete loss of the Ca 2ϩ -ATPase activity due to blockage of the E1P to E2P transition (10 -14). On the other hand, the Glu 40 -Ser 48 loop with an extended structure, the major part of the M1-connecting loop, is well separated from and not having significant interactions with other parts of the molecule in E2V as well as in E1Ca 2 and E2(TG) (Fig. 2), and possible roles of this loop remain unknown. Interestingly, not only A domain, but also M1 connected to this loop, seems to undergo very large and unique structural changes during Ca 2ϩ transport cycle, i.e. up-and-down and horizontal movements and bending near the membrane surface (6).
In the present study, we focused on and explored possible roles of the Glu 40 -Ser 48 loop by site-specific mutagenesis and found that deletions of any single residues in this loop result in almost complete loss of the Ca 2ϩ -ATPase activity, while their substitutions have no or only slight effects. Results further showed that both the E1P to E2P transition and the E2P hydrolysis are almost completely inhibited in all these single deletion mutants. The results indicate that the Glu 40 -Ser 48 loop is critical for formation of the proper structure of E2P and further suggest that the loop may coordinate the motions of A domain and M1 during the E1P to E2P transition and thus possibly contribute to the rearrangement of the transmembrane helices.

EXPERIMENTAL PROCEDURES
Mutagenesis and Expression-Mutations were created by PCR using the QuikChange TM site-directed mutagenesis kit (Stratagene) and plasmid pGEM7-Zf(ϩ) (Promega) containing the ApaI-KpnI fragment of the rabbit SERCA1a cDNA as a template. The ApaI-KpnI fragments were then excised from the PCR products and used to replace the corresponding region in the full-length SERCA1a cDNA in the pMT2 expression vector (15). The pMT2 DNA was transfected into COS-1 cells by the liposome-mediated transfection method. Microsomes were prepared from the cells as described previously (16). 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 (17). The expression levels of all the mutants in the microsomes were comparable with that of the wild type.
ATPase Activity-The rate of ATP hydrolysis was determined at 25°C in a mixture containing 20 g/ml microsomal protein, 0.1 mM [␥-32 P]ATP, 1 M A23187, 7 mM MgCl 2 , 0.1 M KCl, 50 mM MOPS/Tris (pH 7.0), 0.55 mM CaCl 2 , and 0.5 mM EGTA. 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.
Formation and Hydrolysis of EP-Phosphorylation of SERCA1a in microsomes with [␥-32 P]ATP or 32 P i , and dephosphorylation of 32 Plabeled SERCA1a, were performed under conditions described in the legends to figures. 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 (18) or otherwise as above. The precipitated proteins were separated at pH 6.0 by 5% SDS-polyacrylamide gel electrophoresis according to Weber and Osborn (19). The radioactivity associated with the separated Ca 2ϩ -ATPase was quantitated by digital autoradiography as described previously (20). 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 4% 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. (21) 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 (Theoretical Biophysics Group, University of Illinois at Urbana-Champaign).

Effects of Deletions and Substitutions on ATP Hydrolysis-
The specific Ca 2ϩ -ATPase activities of the expressed mutant and wild-type SERCA1a were determined at 25°C (Fig. 3). Deletions of any single residues in the Glu 40 -Ser 48 loop resulted in almost complete loss of the activity, while their nonconservative substitutions caused only partial decrease or rather slight increase in the activity. Single deletions or substitutions of the residues in the adjacent N-terminal (His 32 -Asn 39 ) and C-terminal (Leu 49 -Ile 54 ) regions of the Glu 40 -Ser 48 loop had only slight or moderate effect on the activity, except that the specific substitutions of Asn 39 (N39D and N39T, but not N39A) had a significantly reduced activity. Quadruple alanine substitutions of Asn 39 , Glu 40 , Glu 44 , and Glu 45 were previously shown to cause no loss of function (22), and the present results are in essential agreement.
Formation of EP from ATP-We then performed detailed kinetic analysis with the mutants. EP was formed from ATP at 0°C under conditions otherwise similar to those for the ATPase assay. All the mutants possessed the ability to form EP, and the amount formed was comparable with that of wild type (Fig.  4). In Fig. 5, the fraction of E2P accumulated was determined at steady state (15 s after addition of ATP). In the presence of K ϩ , which strongly accelerates decay of E2P and thus suppresses its accumulation in the wild type (23), the amount of E2P accumulated in all the mutants was very low as in the wild type. In the absence of K ϩ , the fraction of E2P largely increased in the wild type (to 73% of the total amount of EP (E1P plus E2P)), but it remained very low in the single deletion mutants for the Glu 40 -Ser 48 loop and in the mutants N39D and N39T, indicating that almost all EP accumulated was E1P in these mutants. On the other hand, as in the wild type, the fraction of E2P largely increased in the other mutants, in which the residues in the loop were substituted or those in the adjacent Nand C-terminal regions of the loop were deleted or substituted (with the exception of N39D and N39T). The results strongly suggest that the E1P to E2P transition in step 4 is inhibited in the single deletion mutants for the Glu 40 -Ser 48 loop and in the substitution mutants N39D and N39T. We, therefore, examined the decay of E1P accumulated in the presence of K ϩ .
Decay of EP Formed from ATP-Decay of EP formed from ATP in the presence of K ϩ was determined at 0°C by first phosphorylating with [␥- 32  formed was E1P (see Fig. 5) and then terminating phosphorylation by adding excess EGTA to prevent further phosphorylation and thus allow decay of 32 P-labeled EP. The EP decay was well fitted with a single exponential as shown in Fig. 6 with the representative mutants and wild type. The decay rates were thus obtained with all the mutants and summarized in Table I. The decay was almost completely blocked by any single deletions in the Glu 40 -Ser 48 loop, and the rates were only 1-3% of that in the wild type. On the other hand, the rates in the substitution mutants for the Glu 40 -Ser 48 loop and in all the deletion and substitution mutants for the adjacent regions of the Glu 40 -Ser 48 loop were comparable with or only slightly lower than that in the wild type, except the mutants N39D and N39T (but not N39A) that had the significantly reduced rate (dashed lines in Fig. 6).
Hydrolysis of E2P Formed from P i -E2P was formed by 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 (24). In some of the single deletion mutants, including those for the Glu 40 -Ser 48 loop, the amount of EP formed was somewhat reduced (Fig.  7A). Nevertheless the hydrolysis of 32 P-labeled E2P was examined with all the mutants at 0°C by diluting the above phosphorylated samples with a large volume of a solution containing K ϩ and non-radioactive P i . The conditions were thus made otherwise identical to those used for the decay of EP formed from ATP in Fig. 6. The hydrolysis of E2P proceeded with first-order kinetics as shown in Fig. 7, B and C, with the representative mutants and wild type. The hydrolysis rates were thus obtained with all the mutants and summarized in Table I. The hydrolysis was markedly slowed or blocked by any single deletions in the Glu 40 -Ser 48 loop, and the rates were only 0.3-3% of that in the wild type. In contrast, the hydrolysis in the substitution mutants for the Glu 40 -Ser 48 loop and in the deletion and substitution mutants for the adjacent regions of the loop including N39D and N39T (dashed lines in Fig. 7C) was as rapid as, or only slightly slower than, that in the wild type.
Transition from E2 to E1Ca 2 -The mutants and wild type were preincubated in the absence of Ca 2ϩ at pH 6 where equilibrium between E1 and E2 is most shifted to E2 (25) and then phosphorylated at 0°C by simultaneous addition of saturating concentrations of Ca 2ϩ and ATP under conditions otherwise similar to those for the ATPase assay. The time course of EP formation was well described by the first-order kinetics as shown in Fig. 8 with the representative mutants and wild type. The rates were thus obtained with all the mutants and summarized in Table I. The rates of all the mutants, including the single deletion mutants for the Glu 40 -Ser 48 loop, were comparable with that of the wild type and not significantly reduced. When ATP was added to the enzyme preincubated with Ca 2ϩ otherwise as above, the EP formation proceeded at much faster rate (5 s Ϫ1 in the wild type and the comparable rate To realize the states in E2V and E2(TG) from E1Ca 2 , A domain largely rotates by more than ϳ90°and also P and N domains incline as indicated by arrows. Note that the Glu 40 -Ser 48 loop is extended and does not have significant interactions with other parts of the molecule in the three structures and that, in E2V and E2(TG), M3-M5 directly associated with P domain and M6 (with the loop connecting M6 and M7) are inclined and bent toward M1 and the top part of M1 is bent to form an amphipathic helix M1Ј likely at the membrane surface (5, 6). Also note that E2V has the most compactly organized domain structure and E2(TG) has the organized but more relaxed one. in all the mutants). The results show that the Ca 2ϩ -induced E2 to E1Ca 2 transition, which is rate-limiting for the EP formation from E2, is essentially not inhibited in all the mutants.  Loop-In the present study, we explored possible roles of the Glu 40 -Ser 48 loop connecting A domain and M1 helix by mutagenesis and found that deletions of any single residues within this loop (but not their substitutions) almost completely block or strongly inhibit both the E1P to E2P transition and the E2P hydrolysis. Results indicate that the loop with its appropriate length (but not with specific residues) is critical for the rapid isomeric transition and hydrolysis of EP.

Roles of Glu
During the E1P to E2P transition and Ca 2ϩ release into lumen, a large rotation of A domain by ϳ90°and its intimate contact with P and N domains occur to form the most compactly organized cytoplasmic domains in E2P without bound Ca 2ϩ (7,8). In this E2P, the hydrophobic atmosphere (24, 26 -28) is thus realized around the phosphorylation site and a specific water molecule can now attack the acylphosphate bond to hydrolyze. It is therefore likely that the formation of the proper structure of E2P, i.e. the movement of A domain and resulting domain organization, was impaired by the deletions of even single residues (ϳ3.5 Å shortening) in the Glu 40 -Ser 48 loop. The facts found in the structural model E2V, an E2P analogue (7,8), that the Glu 40 -Ser 48 loop is extended and not interacting with other parts of the molecule (Fig. 2), is consistent with the view that no specific residues are involved in the role of this loop but that its length being crucial.
In the detailed and well accepted reaction mechanism for the E1P to E2P transition and Ca 2ϩ release, the process consists of two steps (steps 4 and 5, Fig. 1). The single deletions in the Glu 40 -Ser 48 loop obviously caused blocking of the loss of ADP  Table I. The fraction of E2P determined at each time point (as described in the legend to Fig. 5) was less than 10% of the total amount of EP obtained above (data not shown). sensitivity in step 4 (i.e. the E1P to E2P transition). We have recently found (9) that the mutations in the Lys 189 -Lys 205 outermost loop on A domain do not inhibit the loss of ADP sensitivity but strongly inhibit the subsequent processing of E2P in step 5 and indicated that the final process of gathering of A and P domains is accomplished in step 5 to release the bound Ca 2ϩ by the intimate contact of the Lys 189 -Lys 205 loop with P domain. Together with this study, we thus could identify two structural elements crucial for the successive, but distinct, two steps; the large rotation of A domain and its association with P and N domains (to some extent) in step 4 and the final process for intimate contact of A and P domains in step 5.
Importantly, it was found in comparison of the structures E1Ca 2 and E2(TG) (6) that upon Ca 2ϩ dissociation, M1 connected to the Glu 40 -Ser 48 loop undergoes horizontal and upward movements and a large structural change, i.e. bending at Asp 59 , forming an amphipathic helix M1Ј that has hydrophobic residues on one side and charged residues on the other (Fig. 2). M1Ј is thus likely to be situated at the membrane surface by interactions with lipids and possibly with Arg 63 in M1. This large motion is likely caused by the steric collision of M1 with M3 that inclines together with M4-M6 and P domain toward M1 and A domain (6). The unique structure of M1Ј and M1 are also found in E2V, an analogue of E2P. Then it is possible that the shortening of the Glu 40 -Ser 48 loop by the deletions may inhibit such motion of M1 and thus block the appropriate movements of M3-M6, P domain, and A domain for the E1P to E2P transition. Alternatively, it may also be possible that the appropriate movement of A domain would be inhibited if M1Ј first formed is not flexible enough to move upward from the position at the membrane surface and to compensate the shortening of the Glu 40 -Ser 48 loop by the deletions. In any case, it is likely that, not only the crucial large rotation of A domain and gathering of the cytoplasmic domains, but also the unique movement and structural change of M1, likely contribute to formation of the proper structure of E2P. It is possible that the Glu 40 -Ser 48 loop functions in coordinating these motions of A domain and M1 during the E1P to E2P transition and thus in rendering A domain to accept the inclined P domain at an appropriate position and contributing to cross-talk between the cytoplasmic and transmembrane domains.
The observation showing essentially no effects of the deletions in the Glu 40 -Ser 48 loop on the rate of the E2 to E1Ca 2 transition (and subsequent E1P formation) (steps 1-3, Fig. 8) indicates that the structural importance of the Glu 40 -Ser 48 loop may be already lost in E2. This view is consistent with our previous observation (7,8) that the cytoplasmic domain organization in the more relaxed E2 state is significantly different from that in the most compactly organized E2P state. The view may also be compatible with the notion (6) that E2 is likely in TABLE I Rate constants for partial reaction steps of all the mutants The rate constants for the partial reaction steps were obtained with all the deletion and substitution mutants in the experiments shown in Fig.  6 (decay of E1P formed from ATP), Fig. 7, B and C (hydrolysis of E2P formed from P i ), and Fig. 8  equilibrium with E1 by large scale thermal movements involving both transmembrane and cytoplasmic domains in the absence of ligands that fix the enzyme in certain forms (such as thapsigargin in E2(TG), or vanadate in E2V and covalently bound P i in E2P), hence the possible importance of the unique structure of M1Ј and M1 may be lowered in E2.

Relation to All Other Mutations Found to Inhibit the E1P to E2P Transition and Integrated Picture of Conformational Changes in This Transition and Subsequent
Processing of E2P-In the previous studies, the substitutions of various residues were found to slow or block the loss of ADP sensitivity in step 4. As depicted in Fig. 9 for all these essential or important residues, they are widely distributed in the enzyme, indicating the global conformational changes of the enzyme occurring in the E1P to E2P transition. These include the conserved TGES 184 loop on A domain (29); the conserved residues with their specific substitutions on P domain (Asp 627 3 Glu, Lys 684 3 Arg, Asp 703 3 Asn) (30, 31) and on the loop connecting P and N domains (Asp 601 3 Asn, Pro 603 3 Leu) (30); the residues in the cytoplasmic side of M4 (Pro 312 , Ile 315 , Thr 317 , Leu 319 ) (32, 33); Leu 777 3 Ala and Ala 900 3 Phe at the luminal side of M5 and M8 (34); and the conserved Gly 233 on the Gly 233 -Pro 248 loop connecting A domain and M3 (11). The proteolytic cut within this Gly 233 -Pro 248 loop or at the immediate Glu 231 was also found to block the E1P to E2P transition (10,(12)(13)(14). In addition, the decay of EP formed from ATP was strongly slowed by the substitutions of the Leu 321 -Ser 346 segment from M4 to the phosphorylation site (35), Leu 253 on M3 (36), and Thr 441 , Glu 442 , Lys 492 , and Lys 515 at or near the adenine binding pocket on the N domain (37).
In E2V, an E2P analogue, the TGES 184 outermost loop on A domain is situated at the interface of the cytoplasmic three domains and very close to the above listed residues Asp 627 , Asp 703 , Asp 601 , and Pro 603 , which are at or near the outer surface of P domain or between P and N domains and also closely surrounding the phosphorylation site (Asp 351 ) (Fig. 9). Solid and dashed lines show the least squares fit to a single exponential. The rate constants in the hydrolysis were thus obtained with all the deletion and substitution mutants and given in Table I. The extensive hydrogen bond network is actually formed between the TGES 184 loop, the Asp 601 -Pro 603 loop, Asp 627 , the residues in Lys 352 -Asn 359 , and Arg 560 on N domain, and stabilizing the associated state of the domains. Asp 703 and Glu 183 on the TGES 184 loop were predicted by the iron-catalyzed cleavage of Na ϩ /K ϩ -ATPase (38) to participate together in the Mg 2ϩ binding in E2P. These facts, together with the observed blocking of the E1P to E2P transition by the specific substitutions, strongly suggest that the above residues (Asp 601 , Pro 603 , Asp 627 , and Asp 703 ) or at least some of them are involved in the A-P domain interaction together with the TGES 184 loop for the loss of the ADP sensitivity, although information on the threedimensional structure of E2PCa 2 formed in step 4 (the ADPinsensitive EP with bound or occluded (39) Ca 2ϩ ) is yet unavailable. Importantly, Asp 627 , Asp 703 , Lys 684 , Asp 601 , and Pro 603 were all found previously (30,31,40) to be crucial also for the conformation of catalytic site and for the formation and hydrolysis of EP as their substitutions (other than the specific one stated above) diminished the phosphorylation from ATP and from P i . Therefore, they seem to play multiple essential roles and the roles likely alter as the steps proceed in the Ca 2ϩ transport cycle, i.e. the ATP binding and phosphorylation, the isomeric transition of EP (the domain association), and its hydrolysis. Lys 684 seems not to be involved in the domain interaction in E2P (E2V), but likely functions in coordination of the covalently bound phosphate in EP (38,40) and thus in the possible change in the immediate vicinity of the phosphate for the loss of ADP sensitivity.
Obviously the above docking of A domain with P domain to situate the TGES 184 loop at their appropriate interface requires the large rotational movement of A domain during the E1P to E2P transition and also the inclination of P domain toward A domain (see also Fig. 2). The Glu 40 -Ser 48 and Gly 233 -Pro 248 loops are both apart from this interacting face. The Glu 40 -Ser 48 loop is most likely important for the rotational movement of A domain and coordinated motions of M1 to result in the appropriate domain docking, as discussed above. The Gly 233 -Pro 248 loop is interacting with the lower part (helix P6) of bent forward P domain in E2(TG) (and in E2V) but separated in E1Ca 2 , and it was suggested that this loop moves for the E1P to E2P transition from a peripheral to the more central position where it interacts with the bent forward P domain (by the hydrogen bonds between Glu 243 , Lys 712 , Gln 244 , Thr 242 , and Met 239 , the dashed circle in Fig. 9) (10).
Furthermore, the observed long range strong inhibitory effect on the loss of ADP sensitivity of the substitutions of the Pro 312 -Leu 319 segment on M4, Leu 777 on M5, and Ala 900 on M8 (32)(33)(34) also indicate that the transmembrane helices move (likely incline and bend (to some extent)), together with P domain during this transition in step 4, although these movements are not yet enough to open the Ca 2ϩ pathway to lumen. Some residues on M8 -M10, including Ala 900 , may support such movements (as forming a base), because these helices seem not to move at least in E1Ca 2 , E2(TG), and E2V (5, 6), or Leu 777 and Ala 900 may possibly be important for the pivoting function at the closely situated Gly 770 (6).
Taken together, the observations so far made suggest an integrated picture for the conformational changes occurring in the E1P to E2P transition in step 4 and the subsequent processing and hydrolysis of E2P in steps 5 and 6. In step 4, A domain will first largely rotate and then dock and associate with P and N domains by the interactions involving the TGES 184 loop at the three-domain interface and also by the contribution of the Gly 233 -Pro 248 loop for anchoring A domain on P domain that has likely inclined toward A domain together with the motions of M3-M6, M2, and M1. These coordinated movements and the domain interactions are obviously made possible by the conformational energy gained in steps 2 and 3 with the Ca 2ϩ and ATP bindings and accumulated in E1PCa 2 , which has P and N domains with their most closed configuration as realized in the E1Ca 2 ATP complex, P domain thus being distorted (41,42) and A domain with the M3-and M2-connecting loops being repositioned from E1Ca 2 (and thus ready for its subsequent large rotation) as revealed by the complete protection of P domain and these loops against proteinase K and V8 (but not trypsin at Arg 198 ) in our systematic proteolysis (7,8) and also in the very recent mutation and proteolysis study (40). It is also important to note that P and N domains should be opened in step 4 from this most closed configuration to some extent (likely by the reverse movement of N domain and/or further inclination of P domain toward A domain), and hence A domain can rotate in and associate with P and N domains to cause the loss of ADP sensitivity (i.e. the ␤-phosphate of ADP cannot reach or access any more the covalently bound phosphate at Asp 351 ) (8). (In E2V, the A-N domain association involves the residues near the adenine binding pocket (Glu 439 and Glu 486 , the dashed circle in Fig. 9) besides Arg 560 at the three-domain interface.) In the subsequent step 5, the final process of gathering of A and P domains is accomplished by the formation of strong interaction between the polar residues surrounding Val 200 (Arg 198 , Asn 201 , Gln 202 , and Asp 203 ) on the FIG. 9. Residues previously found to be essential for the E1P to E2P transition on the structural model E2V. The residues of which mutations were previously shown to inhibit the loss of ADP sensitivity (step 4) are represented with Asp 351 (phosphorylation site) and Arg 560 as a CPK model in cyan (carbon atom), red (oxygen atom), and dark blue (nitrogen atom) in E2V (Protein Data Bank accession code 1FQU (5)). The dashed circles show the regions of the hydrogen bonding interactions between the Gly 233 -Pro 248 loop and helix P6 on P domain and between A and N domains. The Lys 189 -Lys 205 loop is also shown. For details, see "Discussion." Lys 189 -Lys 205 loop (another outermost loop of A domain besides the TGES 184 loop) and those on P domain (Arg 678 and Glu 680 on the basis of E2V) to produce the most compactly organized single headpiece (9). This intimate contact of A and P domains will likely provide the conformational energy enough to further distort P domain and rearrange more the transmembrane helices and thus to open the luminal gate and release the bound Ca 2ϩ . The intimate contact of the domains also produces the hydrophobic atmosphere (and the fine-tuning) around the phosphorylation site, so that the specific water molecule can attack the acylphosphate bond and the essential residues can participate in the hydrolysis. Upon the E2P hydrolysis in step 6, the domain organization changes to the more relaxed E2 state in which the interactions of the Lys 189 -Lys 205 loop with P domain are likely lost (9), but A domain has not yet rotated back to the state found in E1Ca 2 and still associated with P and N domains (6,8) by the possible interactions between A and N domains and between P domain and the loops connecting A domain to M2 and M3 as found in E2(TG) (6).
Mutants in Darier's Disease-Darier's disease, a human autosomal dominant skin disorder, was shown to be caused by mutations in the SERCA2b protein (43,44), and deletions and substitutions ⌬41, ⌬42, N39D, and N39T were found in the Darier's disease pedigrees (45,46). It was indicated very recently (47) that these mutants have no Ca 2ϩ transport activity. Because these residues are conserved in SERCA1a, our study further shows that these mutations reduce completely or strongly the Ca 2ϩ -ATPase activity of SERCA2b due to inhibition of the E1P to E2P transition (in all ⌬41, ⌬42, N39D, and N39T) and the E2P hydrolysis (in ⌬41 and ⌬42) ( Fig. 3 and Table I) and therefore cause disruptions of calcium homeostasis and the disease. It should be noted that the mutant N39A has the normal activity and kinetic properties, in contrast to the Darier's disease mutants N39D and N39T. It is possible that these threonine and aspartate substitutions may cause unfavorable interactions with other residues, because Asn 39 forms a hydrogen bond network within A domain and at its junction with the Glu 40 -Ser 48 loop (Fig. 9), and the interacting residues likely change during the catalytic cycle as in fact so in E1Ca 2 (Ala 142 , Pro 160 ), E2V (Thr 226 ), and E2(TG) (Ala 142 , Arg 143 , Gly 227 ). Asn 39 seems to adjust the proper movement of A domain at this junction and assist the role of the Glu 40 -Ser 48 loop by changing the interaction network during the E1P to E2P transition. Interestingly, in the mutants N39D and N39T, the E2P hydrolysis proceeded with the normal rate in contrast to the retarded E1P to E2P transition ( Fig. 7C and Table I). It is likely that the structural defect caused by the Asn 39 substitutions would be overcome when the intimate contact of the cytoplasmic three domains is once formed in the subsequent step 5 by the contribution of the Lys 189 -Lys 205 loop.