Deletions or Specific Substitutions of a Few Residues in the NH2-terminal Region (Ala3 to Thr9) of Sarcoplasmic Reticulum Ca2+-ATPase Cause Inactivation and Rapid Degradation of the Enzyme Expressed in COS-1 Cells*

Amino acid residues in the NH2-terminal region (Glu2 – Ala14) of adult fast twitch skeletal muscle sarcoplasmic reticulum Ca2+-ATPase (SERCA1a) were deleted or substituted, and the mutants were expressed in COS-1 cells. Deletion of any single residue in the Ala3–Ser6 region or deletion of two or more consecutive residues in the Ala3–Thr9 region caused strongly reduced expression. Substitution mutants A4K, A4D, and H5K also showed very low expression levels. Deletion of any single residue in the Ala3–Ser6 region caused only a small decrease in the specific Ca2+ transport rate/mg of SERCA1a protein. In contrast, other mutants showing low expression levels had greatly reduced specific Ca2+ transport rates. In vitroexpression experiments indicated that translation, transcription, and integration into the microsomal membranes were not impaired in the mutants that showed very low expression levels in COS-1 cells. Pulse-chase experiments using [35S]methionine/cysteine labeling of transfected COS-1 cells demonstrated that degradation of the mutants showing low expression levels was substantially faster than that of the wild type. Lactacystin, a specific inhibitor of proteasome, inhibited the degradation accelerated by single-residue deletion of Ala3. These results suggest that the NH2-terminal region (Ala3 –Thr9) of SERCA1a is sensitive to the endoplasmic reticulum-mediated quality control and is thus critical for either correct folding of the SERCA1a protein or stabilization of the correctly folded SERCA1a protein or both.

SERCA1a is composed of 10 transmembrane ␣-helices (M 1 to M 10 ) and two main cytoplasmic domains, a small loop (Ala 132 to Asp 237 between M 2 and M 3 ) and a large loop (Asn 330 to Asn 739 between M 4 and M 5 ) (1). In addition, there is a small cytoplasmic NH 2 -terminal domain (Met 1 -Asn 39 ). These cytoplasmic domains are connected by ␣-helical segments (called stalks) to the transmembrane ␣-helices. The large cytoplasmic loop contains the phosphorylation site and the ATP-binding site. Several residues in the small cytoplasmic loop were shown to play essential roles in the conformational transition of the phosphoenzyme intermediate (10 -12). We have recently indicated that Arg 198 in this small loop contributes to the catalytic site (13,14).
The functional role of the NH 2 -terminal domain is less clear, although our recent chemical modification study (15) has suggested that His 5 in this domain is located very close to the catalytic site. It was previously shown (16) that deletion of most of the residues (Glu 2 -His 32 ) in the NH 2 -terminal domain results in greatly reduced expression in COS-1 cells and inactivation of the enzyme. This raises the possibility that the NH 2terminal domain has a region sensitive to the endoplasmic reticulum (ER)-mediated quality control, the machinery of which recognizes and rapidly degrades misfolded proteins (this misfolding can be induced by mutations) and denatured proteins to suppress their cellular expression or accumulation (17,18). However, the ER-mediated quality control of the Ca 2ϩ -ATPase has not yet been reported.
In this study, we have explored the possible roles of much smaller NH 2 -terminal regions (Glu 2 -Ala 14 , especially Ala 3 -Ser 6 ) than the whole NH 2 -terminal domain (Met 1 -Asn 39 ) in cellular expression of SERCA1a and its enzymatic function. We have made 45 mutants of SERCA1a in which residues in the Glu 2 -Ala 14 region have been deleted or substituted, and the mutants have been expressed in COS-1 cells. The results show that deletions or specific substitutions of residues in the Ala 3 -Thr 9 region lead to greatly reduced expression of the mutated SERCA1a proteins and rapid degradation of the expressed SERCA1a proteins. The results further show that residues in the Ala 3 -Ser 6 region are not essential for the Ca 2ϩ transport function. We suggest that the Ala 3 -Thr 9 region is sensitive to the ER-mediated quality control and is thus critical either for correct folding of the SERCA1a protein or stabilization of the correctly folded SERCA1a protein or both.

EXPERIMENTAL PROCEDURES
Oligonucleotide-directed Mutagenesis and Expression in COS-1 Cells-The methods employed have been described (14). A summary of the methods is as follows. Overlap extension PCR (19) was used to introduce mutations into the rabbit SERCA1a cDNA. The PCR products containing the desired mutation were subcloned into the pT7Blue vector (Novagen, Madison, WI). The mutated fragments were excised and religated back into their original position in the full-length SERCA1a * This work was supported by a grant-in-aid for scientific research from the Ministry of Education, Science, Sports and Culture, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
cDNA that was previously ligated into the EcoRI site of the pMT2 expression vector (20). The plasmid DNA was transfected into COS-1 cells (21) by the liposome-mediated DNA transfection procedure. Microsomal membranes were prepared from the cells as described by Maruyama and MacLennan (22). Microsomal proteins were separated by 7.5% SDS-polyacrylamide gel electrophoresis according to Laemmli (23). The expressed SERCA1a was detected by Western blotting, using VE121G9 monoclonal antibody to the rabbit SERCA1a (Affinity Bioreagents, Golden, CO). After incubation with secondary antibody (sheep anti-mouse IgG horseradish peroxidase-conjugated, Amersham Pharmacia Biotech), the bound proteins were probed using an enhanced chemiluminescence-linked detection system (Amersham Pharmacia Biotech). Immunoreactivity was quantitated by densitometry. In the assay, the standard enzyme used was the deoxycholate-purified rabbit SERCA1a that was prepared by the method of Meissner and Fleischer (24) with slight modifications as described previously (25). In addition to Western blotting, quantitation of SERCA1a expression was also obtained by a sandwich enzyme-linked immunosorbent assay as described below.
Ca 2ϩ Transport Activity-Ca 2ϩ transport activity was assayed at 27°C in a mixture containing 5-10 g/ml microsomal protein, 20 mM MOPS-Tris (pH 7.0), 0.1 M KCl, 7 mM MgCl 2 , 5 mM ATP, 5 mM potassium oxalate, and 0.1 mM 45 CaCl 2 . At different time periods, 0.5-ml samples were filtered through a 0.45-m mixed cellulose ester membrane filter (ADVANTEC Toyo Kaisha, Ltd., Tokyo, Japan) and washed three times with 5 ml of a solution containing 20 mM MOPS-Tris (pH 7.0), 0.1 M KCl, 7 mM MgCl 2 , and 2 mM EGTA. Radioactivity on the filters was measured by liquid scintillation counting. The Ca 2ϩ transport curve in the presence of 0.5 M thapsigargin with the microsomal membranes expressing the wild-type or mutant SERCA1a was not significantly different from that with control microsomal membranes, which were prepared from COS-1 cells transfected with the pMT2 vector containing no SERCA1a cDNA. Therefore, the Ca 2ϩ transport curve of the expressed SERCA1a was obtained by subtracting the amount of Ca 2ϩ transported in the presence of 0.5 M thapsigargin from that in its absence. The Ca 2ϩ transport activity of the expressed SERCA1a was calculated from the initial linear part of the Ca 2ϩ transport curve thus obtained. The specific transport rates/mg of SERCA1a protein were calculated from the thapsigargin-sensitive Ca 2ϩ transport activity and the amount of the expressed SERCA1a, which was quantified by a sandwich enzyme-linked immunosorbent assay as described by Leberer and Pette (26). In this assay, purified sheep anti-rabbit SERCA1a IgG was used to coat the plates, and monoclonal antibody VE121G9 was used for the specific reaction with the bound SERCA1a. After incubation with secondary antibody (sheep anti-mouse IgG horseradish peroxidase-conjugated), staining was performed with tetramethylbenzidine base (TMB-ELISA, Life Technologies, Inc.). The staining reaction was stopped by adding 0.2 N H 2 SO 4 . The absorbance at 450 nm was measured.
Expression in a Cell-free Transcription/Translation System-PCR mutagenesis was used to insert a HindIII site immediately before the initiator methionine and a SacI site immediately after the stop codon of the SERCA1a using the full-length SERCA1a cDNA as template. The PCR product was subcloned into the pT7Blue vector, and then a coding region for the SERCA1a was ligated as a 5Ј HindIII to 3Ј SacI fragment into pSP64 poly(A) vector (Promega, Madison, WI). PCR mutagenesis was employed to make mutants using this plasmid as template and mutagenic 5Ј-flanking primers. The PCR product carrying a HindIII site 5Ј to the SERCA1a cDNA was subcloned into the pT7Blue vector, and then the approximately 640-base HindIII-KpnI fragments were excised from the vector. The restriction fragments carrying the mutations were religated back into the original position in the SERCA1a cDNA that was ligated into the pSP64 poly(A) vector. The pSP64 poly(A) vector harboring the wild-type or mutant SERCA1a cDNA was added into a reaction mixture containing the rabbit reticulocyte lysate in vitro transcription and translation mix and canine pancreatic microsomal membranes (both from Promega) in the presence of [ 35 S]methionine (Redivue TM , Amersham Pharmacia Biotech) and was incubated according to the manufacturer's instructions. Membrane fractions were then isolated from the reaction mixture by centrifugation. The samples were either untreated or treated with 1 M KCl or 100 mM Na 2 CO 3 (pH 11.5) for 10 min on ice and recovered by centrifugation. The pellets were separated by 7.5% SDS-polyacrylamide gel electrophoresis according to Laemmli (23) and subjected to Western blotting or to digital autoradiography of the dried gel using Bio-Imaging Analyzer BAS2000 (Fuji Photo Film Co., Ltd., Tokyo, Japan).
Pulse-Chase Experiments and Immunoprecipitations-COS-1 cells were transfected with the pMT2 vector containing the wild-type or mutated SERCA1a cDNA and cultured for 22 h, as described previously (14). The cells were starved for 2 h at 37°C in methionine/cysteine-free Dulbecco's modified Eagle's medium and pulse-labeled with [ 35 S]methionine/cysteine (80 Ci/ml) (Redivue TM PRO-MIX TM , Amersham Pharmacia Biotech) in the medium for 2 h at 37°C. The cells were then chased in minimum essential medium (Life Technologies, Inc.) at 37°C. At different times after the start of the chase, the cells were washed three times with phosphate-buffered saline and harvested. The cells were then lysed for 1 h at 4°C in the lysis buffer (1% (v/v) IGEPAL CA-630 (Sigma), 15 mM Tris-HCl (pH 7.5), 0.15 M NaCl, and 1 mM EDTA) containing 1 mM phenylmethylsulfonyl fluoride (Sigma) and 100 units/ml aprotinin (Sigma). After insoluble material was removed by centrifugation at 18,000 ϫ g for 30 min at 4°C, incorporation of 35 S into the total protein pool was determined by spotting an aliquot of lysate onto the mixed cellulose ester membrane filter and boiling the filter for 10 min in 5% (w/v) trichloroacetic acid containing 5 mM methionine and 5 mM cysteine. The radioactivity on the spot was quantitated by digital autoradiography of the dried filter using Bio-Imaging Analyzer BAS2000. Lysate volumes for immunoprecipitation were normalized by trichloroacetic acid-precipitable radioactivity. Immunoprecipitation of the SERCA1a was performed by incubating the lysates overnight at 4°C with purified sheep anti-rabbit SERCA1a IgG and protein G-Sepharose (Amersham Pharmacia Biotech) in the lysis buffer. The beads were washed five times with the lysis buffer, resuspended in Laemmli sample buffer (23), and centrifuged. The supernatants were subjected to 7.5% SDS-polyacrylamide gel electrophoresis according to Laemmli (23). The radioactivity associated with the separated SERCA1a was quantitated by the digital autoradiography of the dried gels as above. When effects of lactacystin were investigated, the pulse labeling and chase were performed in the absence and presence of 10 M lactacystin; other conditions were as described above.
Miscellaneous Methods-Protein concentrations were determined by the method of Lowry et al. (27) with bovine serum albumin as a standard. Dideoxy sequencing (28) was carried out to ensure fidelity of the PCR amplification step and to confirm the presence of the correct mutations. Polyclonal anti-rabbit SERCA1a antibody was prepared by injecting a sheep with the deoxycholate-purified rabbit SERCA1a (25) as described by Leberer and Pette (26). Anti-rabbit SERCA1a IgG was purified through DEAE Affi-Gel Blue (Bio-Rad) column chromatography of an ammonium sulfate-precipitated fraction (0 -33%) of antirabbit SERCA1a antiserum, according to standard procedures.

Effects of Deletions and Substitutions of Residues in the NH 2 -terminal Region of SERCA1a on Expression in COS-1
Cells-Amino acid residues in the NH 2 -terminal Glu 2 -Ala 14 region of SERCA1a were deleted or substituted, and the mutants were expressed in COS-1 cells. A typical example of Western blots of the deletion mutants expressed in microsomal membranes is shown in Fig. 1. Visual inspection reveals that expression was only slightly reduced by deletion of 3 residues from Glu 2 to Ala 4 but greatly reduced by deletions of 4 -13 consecutive residues from Glu 2 to Ala 14 . Expression was also greatly reduced by deletions of 2-4 consecutive residues from Ser 6 to Thr 9 .
In addition to the above deletion mutants, 33 mutants were made in which residues in the Glu 2 -Thr 9 region were deleted or substituted. The expression levels of the mutants were determined by quantitative densitometry of the proteins visualized with enhanced chemiluminescence and normalized to that of the wild type (Fig. 2). Expression was only partially reduced in the deletion mutants ⌬2, ⌬2-3, and ⌬2-4 but greatly reduced in the mutants with deletions of 4 -13 consecutive residues from Glu 2 to Ala 14 , as expected from inspection of Fig. 1. Expression of the mutants with deletions of 2-4 consecutive residues in the Ala 3 -Thr 9 region was also greatly reduced. Strikingly, expression of the mutants with deletions of any single residue in the Ala 3 -Ser 6 region was markedly reduced. Expression was greatly reduced in the substitution mutants A4K, A4D, and H5K but not significantly or only partially reduced in other substitution mutants tested. These results indicate that deletions of 1 or more residues in the Ala 3 -Thr 9 region or specific substitutions of Ala 4 with lysine and aspartic acid or His 5 with lysine result in strongly reduced expression in COS-1 cells.

Effects of Deletions and Substitutions of Residues in the NH 2 -terminal Region of SERCA1a on the Ca 2ϩ Transport
Rate-The specific Ca 2ϩ transport rates/mg of SERCA1a protein were determined with the mutants in which residues in the NH 2 -terminal Glu 2 -Ser 8 region were deleted or substituted, and the rates were normalized to that of the wild-type SERCA1a (Fig. 3). Deletion of any single residue in the Ala 3 -Ser 6 region caused only a small decrease in the specific Ca 2ϩ transport rate. This indicates that these single-residue dele-tions have only small effects on the enzyme structure essential for Ca 2ϩ transport function. The specific Ca 2ϩ transport rates in the substitution mutants A3D, A4L, H5D, and S6L were not significantly different from that of the wild type. These results show that the amino acid residues in the Ala 3 -Ser 6 region are not essential for Ca 2ϩ transport function. Previously, Skerjanc et al. (16) reported that there is no indication from site-directed mutagenesis that specific residues in the Glu 2 -His 32 region are crucial for enzymatic activity. This is consistent with our above conclusion.
On the other hand, when 2-4 consecutive residues in the Glu 2 -Ser 8 region were deleted (⌬2-5, ⌬3-4, ⌬4 -5, ⌬5-6, and ⌬6 -8), the specific Ca 2ϩ transport rates were greatly reduced. The transport rates were also greatly reduced in the substitution mutants A4K, A4D, and H5K. These results indicate that deletions of 2 or more residues in the Glu 2 -Ser 8 region, or specific substitutions of Ala 4 with lysine and aspartic acid or His 5 with lysine, induce structural changes that lead to inactivation of the enzyme. This is in harmony with our previous results from chemical modification of His 5 (15) suggesting that His 5 is located very close to the catalytic site.
In Vitro Expression of Mutant SERCA1a in Pancreatic Microsomal Membranes-In vitro expression experiments were performed with a cell-free transcription/translation system containing pancreatic microsomal membranes in the presence of [ 35 S]methionine (Fig. 4). The wild-type SERCA1a or five mutants (⌬2-5, ⌬4 -5, ⌬5-6, ⌬6 -8, and A4K) showing very low expression levels in COS-1 cells (see Fig. 2) were expressed in this system. The membrane fractions isolated were either untreated or treated with salt or base under conditions

FIG. 2. Expression levels of various deletion and substitution mutants of SERCA1a in microsomal membranes from COS-1 cells.
The expression levels of various deletion and substitution mutants of SERCA1a in the microsomal membranes were determined by Western blotting as described under "Experimental Procedures" and normalized to that of the wild-type SERCA1a (100%). In the mutant S6K7S8/AAA, all residues in the sequence of Ser 6 -Lys 7 -Ser 8 were substituted with alanine. The values presented are the mean Ϯ S.D. of four independent transfections. in which all but integral proteins are usually removed (29), and the samples were subjected to SDS-polyacrylamide gel electrophoresis.
The digital autoradiogram of the gel showed a single major band for each sample (Fig. 4, left panel) at the position of the rabbit SERCA1a (Fig. 4, arrows). These major bands were identified as the expressed SERCA1a protein by Western blotting with the monoclonal anti-rabbit SERCA1a antibody (Fig.  4, right panel). The digital autoradiogram and Western blot showed that the expression levels of the mutants were similar to or even higher than those of the wild type and that all the mutants tested and the wild type were integrated into the membrane in a fashion resistant to extraction with salt or base. These results indicate that transcription, translation, and integration into the microsomal membranes are not impaired in these mutants. This raised the possibility that these mutants are degraded rapidly in COS-1 cells, although the interference of the mutations with the membrane assembly of the expressed SERCA1a in COS-1 cells is also possible because such interference was previously demonstrated by Zhang et al. (30) with the mutations in the SERCA1a segment from the phosphorylation site (Asp 351 ) to the transmembrane helix M 4 . Thus, we examined the degradation of the mutants in COS-1 cells by pulsechase experiments.
Degradation of Mutant SERCA1a in COS-1 Cells-COS-1 cells expressing the wild-type or mutant SERCA1a were pulselabeled with [ 35 S]methionine/cysteine, chased, and then lysed. The SERCA1a in the lysate was immunoprecipitated. Lysate volumes for immunoprecipitation were normalized by trichloroacetic acid-precipitable radioactivity. The radioactivity of the FIG. 3. The rates of Ca 2؉ transport catalyzed by deletion and substitution mutants of SERCA1a in microsomal membranes from COS-1 cells. The specific Ca 2ϩ transport rates/mg of SERCA1a protein of various deletion and substitution mutants of SERCA1a in the microsomal membranes were obtained as described under "Experimental Procedures" and normalized to that of the wildtype SERCA1a (100%). The values presented are the mean Ϯ S.D. of five independent transfections. The specific Ca 2ϩ transport rate of the wild-type SERCA1a was 8 -11 mol of Ca 2ϩ /min/mg of SERCA1a protein. SERCA1a thus obtained was quantitated by SDS-polyacrylamide gel electrophoresis and digital autoradiography (Fig. 5). Relative amounts of the radiolabeled wild-type SERCA1a increased during the chase period. This shows that degradation of the wild-type SERCA1a was slower than the decrease in the total trichloroacetic acid-precipitable radioactivity.
Degradation was not affected by the H5D substitution, which had no effect on the expression level in COS-1 cells (see Fig. 2) and on the specific Ca 2ϩ transport rate (see Fig. 3). In contrast, degradation was accelerated moderately in the single-residue deletion mutants ⌬3, ⌬4, ⌬5, and ⌬6, in which expression in COS-1 cells was reduced substantially (see Fig. 2) but the specific Ca 2ϩ transport rate was reduced only partially (see Fig. 3). Degradation was more strongly accelerated in the 2-residue deletion mutants ⌬3-4, ⌬4 -5, and ⌬5-6 and the substitution mutants in which both the expression levels in COS-1 cells (see Fig. 2) and the specific Ca 2ϩ transport rates (see Fig.  3) were reduced greatly. Degradation was most strongly accelerated in the 4-residue deletion mutant ⌬2-5, in which expression in COS-1 cells was at the lowest level (see Fig. 2) and the specific Ca 2ϩ transport rate was abolished almost completely (see Fig. 3). These results indicate that the reduced expression of these mutants in COS-1 cells was due to accelerated intracellular degradation of the mutants. The results further suggest that substantial acceleration of degradation and strong suppression of cellular expression of the mutants probably can be induced even by small structural changes that have only small effects on the specific Ca 2ϩ transport rate as shown with the single-residue deletion mutants.
It is well documented that the single-residue deletion of phenylalanine (⌬F508) from a cytoplasmic portion of cystic fibrosis transmembrane conductance regulator (CFTR), which is a 1480-residue protein containing 12 putative transmembrane segments, induces its rapid ER quality control-mediated degradation and leads to greatly reduced plasma membrane expression but does not severely impair the function of CFTR (31). This situation closely resembles our present results with the single-residue deletion mutants. This prompted us to explore the possibility that structural changes induced by deletions or substitutions in the Glu 2 -Ser 6 region of SERCA1a are The radioactivity associated with SERCA1a at zero time (i.e. immediately before the start of the chase) is normalized to 100%. Symbols refer to the wild-type SERCA1a (E) and the ⌬3 (‚), ⌬4 (OE), ⌬5 (f), ⌬6 (), ⌬3-4 (ࡗ), ⌬4 -5 (Ⅺ), ⌬5-6 (ƒ), ⌬2-5 (छ), A4D (q), H5D (•), and H5K (") mutants. recognized by the ER quality control machinery.
The effect of lactacystin, a specific proteasome inhibitor, on degradation of the mutants was examined (Fig. 6). The presence of 10 M lactacystin in the medium throughout the pulse and chase periods resulted in a substantially reduced rate of degradation of the single-residue deletion mutant ⌬3. This indicates that proteasome is involved in the ⌬3 deletion-induced acceleration of degradation and suggests that the mutant ⌬3 is degraded by the ER quality control machinery as the mutant ⌬F508 of CFTR. However, lactacystin did not affect the degradation of the mutants ⌬2-5, ⌬4 -5, ⌬5-6, and H5K, whose degradation was much more rapid than that of the singleresidue deletion mutants (see Fig. 5). This finding suggests that a lactacystin-insensitive protease(s), rather than proteasome, is involved in the very rapid degradation of these mutants. Unexpectedly, degradation of the wild type was accelerated by addition of lactacystin. The reason for this acceleration remains obscure.
No degradation intermediates were detected by immunoprecipitation analysis using polyclonal anti-rabbit SERCA1a antibody in the pulse-chase experiments (data not shown), in agreement with the previously reported findings that no degradation intermediates were detected in the ER quality control of the enzyme 3-hydroxy-3-methylglutaryl-CoA reductase (32) and the mutated ATP-binding cassette transporter Pdr5 (33). This suggests that the rate-limiting step in the degradation of the mutants occurs before proteolysis by proteasome or other proteases. It is likely that this rate-limiting step (possibly unfolding of the protein) is more strongly accelerated by deletions of 2 or more residues in the Glu 2 -Ser 6 region or by the A4D or H5K substitution than by single-residue deletions in the Ala 3 -Ser 6 region.
The present results indicate that the NH 2 -terminal region (Ala 3 -Thr 9 ) of SERCA1a is very sensitive to the ER quality control, the machinery of which recognizes misfolded or denatured proteins and rapidly degrades these abnormal proteins (17,18). Therefore, it is very likely that this region is critical for either correct folding of the SERCA1a protein or stabilization of the correctly folded SERCA1a protein or both. Single mutations in other sequence segments in the SERCA1a were previously reported to have effects similar to those reported in this study. Zhang et al. (30) showed that mutation of Ala 331 to Arg yields very low protein levels in COS-1 cells, whereas transcription is normal. Yu et al. (34) reported that mutation of Phe 256 to Glu reduces expression greatly but not enzyme activity.
The NH 2 -terminal region (Ala 3 -Thr 9 ) of SERCA1a shares virtually no homology with the NH 2 -terminal domains of plasma membrane Ca 2ϩ -ATPase (35), Na ϩ ,K ϩ -ATPase (36,37), or H ϩ ,K ϩ -ATPase (38,39). It was previously shown that deletion of the NH 2 -terminal 18 -75 residues from the plasma membrane Ca 2ϩ -ATPase (40) or deletion of the NH 2 -terminal 1-32 residues from the Na ϩ ,K ϩ -ATPase (41) does not inhibit cellular expression of the protein. It is possible that the role of the NH 2 -terminal region described in this paper is specific to the family of sarco(endo)plasmic reticulum Ca 2ϩ -ATPases, be-cause the NH 2 -terminal domains of the Ca 2ϩ -ATPases in this family have considerably high homology to each other (42).