All three splice variants of the human sarco/endoplasmic reticulum Ca2+-ATPase 3 gene are translated to proteins: a study of their co-expression in platelets and lymphoid cells.

The molecular cloning of two previously unknown human sarco/endoplasmic reticulum Ca(2+)-ATPase 3 (SERCA3) 3'-end transcripts, 3b and 3c, has been recently published. Data were lacking, however, for the presence of these SERCA3 variants in different tissue or cell types at the protein level. Here we report the co-expression of three human SERCA3 protein isoforms in platelets and T lymphoid Jurkat cells. Isoform-specific polyclonal anti-peptide antibodies have been generated that recognize specifically the SERCA3a, 3b or 3c splice variants at their C-termini, and this has been confirmed by peptide-competition experiments as well. None of these antibodies cross-reacted with the housekeeping SERCA2b isoform co-expressed endogenously with SERCA3 proteins in non-muscle cells. Although all three SERCA3 isoforms could be detected in platelets, the 3a form was the most abundantly expressed species. Its size matched the apparent size of SERCA3a over-expressed in HEK-293 cells. Immunoprecipitation of the SERCA3 variants from platelet membranes using a PL/IM 430-affinity matrix provided evidence that the putative pan-anti-SERCA3 antibody, PL/IM 430, recognizes all SERCA3 protein isoforms. The epitope for the PL/IM 430 antibody could be localized in a 40 kDa N-terminal tryptic fragment common to all three SERCA3 variants. Comparative Western-blot analysis showed that the expression level of the SERCA3a, 3b and 3c isoforms was more than 10 times lower in Jurkat cells than in platelets, whereas expression of the ubiquitous SERCA2b was nearly identical. This work highlights new Ca(2+)-transporting proteins of haematopoietic cells and provides specific antibodies for their detection.

Delayed activation of the plasma membrane calcium pump by a sudden increase in Ca2+: fast pumps reside in fast cells.

There are four genes encoding isoforms of the plasma membrane Ca(2+) pump (PMCA). PMCA variability is increased by the presence of two splicing sites. Functional differences between the variants of PMCA have been described, but little is known about the adaptive advantages of this great diversity of pumps. In this paper we studied how the different isoforms respond to a sudden increase in Ca(2+) concentration. We found that different PMCAs are activated by Ca(2+) at different rates, PMCA 3f and 2a being the fastest, and 4b the slowest. The rate of activation by Ca(2+) depends both on the rate of calmodulin binding and the magnitude of the activation by calmodulin. We found that 2a is located in heart and the stereocilia of inner ear hair cells, 3f in skeletal muscle and 4b was identified in Jurkat cells. Both cardiac and skeletal muscle, and stereocilia recover very rapidly after a cytoplasmic Ca(2+)peak, while in Jurkat cells the recovery takes up to a minute. In stereocilia, 2a is the only method for export of Ca(2+), making the analysis of them unusually straightforward. This indicates that these rates of PMCA activation by Ca(2+) are correlated with the speed of Ca(2+) concentration decay after a Ca2 spike in the cells in which these variants of PMCA are expressed. The results suggest that the type of PMCA expressed will correspond with the speed of Ca(2+) signals in the cell.

Chimaeras reveal the role of the catalytic core in the activation of the plasma membrane Ca2+ pump.

Isoform 2b of the plasma membrane calcium pump differs from the ubiquitous isoform 4b in the following: (a) higher basal activity in the absence of calmodulin; (b) higher affinity for calmodulin; and (c) higher affinity for Ca(2+) in the presence of calmodulin [Elwess, Filoteo, Enyedi and Penniston (1997) J. Biol. Chem. 272, 17981-17986]. To investigate which parts of the molecule determine these kinetic differences, we made four chimaeric constructs in which portions of isoform 2b were grafted into isoform 4b: chimaera I contains only the C-terminal regulatory region of isoform 2b; chimaera II contains the N-terminal moiety of isoform 2b, including both cytoplasmic loops; chimaera III contains the sequence of isoform 2b starting from the N-terminus to after the end of the first (small) cytoplasmic loop; and chimaera IV contains only the second (large) cytoplasmic loop. Surprisingly, chimaera I showed low basal activity in the absence of calmodulin and low affinity for calmodulin, unlike isoform 2b. In contrast, the chimaera containing both loops showed high basal activity, and Ca(2+) activation curves (both in the absence and in the presence of calmodulin) similar to those of isoform 2b. The rates of activation by calmodulin and of inactivation by calmodulin removal were measured, and the apparent K(d) for calmodulin was calculated from the ratio between these rate constants. The order of affinity was: 2b=II>4b=IV>III=I. From these results it is clear that the construct that most closely resembles isoform 2b is chimaera II. This shows that, in order to obtain an enzyme with properties similar to those of isoform 2b, both cytoplasmic loops are needed.

Protein kinase C phosphorylates plasma membrane Ca2+ pump isoform 4a at its calmodulin binding domain.

Phosphorylation by protein kinase C of isoform 4a of the human plasma membrane Ca2+ pump (hPMCA4a) was studied using the COS cell expression system. Phosphorylation of several truncated mutants of hPMCA4a indicated that a single phosphorylation site lies in a region between residues 1113 and 1125. This region is within the calmodulin binding domain and contains a single phosphorylatable residue, serine 1115. Converting this serine to an alanine diminished phosphorylation greatly. Phosphorylation, done in the absence of calmodulin, did not affect subsequent calmodulin binding, but previous binding of calmodulin did inhibit phosphorylation. Moreover, no significant shift in the calmodulin response curve of hPMCA4a was observed when phosphorylation was mimicked by converting serine 1115 to an acidic residue. The calmodulin binding domain of hPMCA4a is much longer than other calmodulin binding domains and has been suggested to consist of two binding lobes interrupted by a short nonbinding region. The findings of this study indicate that serine 1115 is the residue phosphorylated by protein kinase C, and that it lies within the nonbinding region of the calmodulin binding domain.

Expression of hPMCA4b, the major form of the plasma membrane calcium pump in megakaryoblastoid cells is greatly reduced in mature human platelets.

Antibodies 5F10 and JA3 (raised against the erythrocyte Ca2+ pump) were used to identify hPMCA4b as the major form of the plasma membrane Ca2+ pump in human platelets and in three human megakaryoblastoid cell lines, MEG 01, DAMI and CHRF 288-11. 5F10 was used because it has been shown to recognize all known isoforms of the hPMCA and JA3 because it reacts exclusively with hPMCA4b [Caride A.J., Filoteo A.G., Enyedi A., Verma A.K., Penniston J.T. Detection of isoform 4 of the plasma membrane calcium pump in human tissues by using isoform-specific monoclonal antibodies. Biochem J 1996; 316: 353-359]. In addition to hPMCA4b, hPMCA1b was also detected in the megakaryoblastoid cells by using isoform-specific polyclonal antibodies. The apparent size of this isoform, however, was smaller than that seen in HeLa and COS-7 cell membranes indicating the presence of a modified form of hPMCA1b. In platelets, no evidence of the expression of hPMCA1b could be found. The amount of PMCA in these cells was compared with that of the constitutive form of the sarco/endoplasmic reticulum Ca2+ pump in non-muscle cells (SERCA2b) and also with the amount of PMCA in human erythrocytes. A very low level of the plasma membrane Ca2+ pump was found in platelets while in their precursor cells the expression of this Ca2+ pump was much more abundant. Whereas the expression level of PMCA decreased dramatically in mature human platelets, the expression of SERCA2b did not change substantially upon megakaryocytic differentiation.

Protein kinase C phosphorylates the "a" forms of plasma membrane Ca2+ pump isoforms 2 and 3 and prevents binding of calmodulin.

Phosphorylation by protein kinase C of the "a" and "b" variants of plasma membrane Ca2+ pump isoforms 2 and 3 was studied. Full-length versions of these isoforms were assembled and expressed in COS cells. Whereas the "a" forms were phosphorylated easily with PKC, isoform 2b was phosphorylated only a little, and isoform 3b was not phosphorylated at all. Phosphorylation of isoforms 2a and 3a did not affect their basal activity, but prevented the stimulation of their activity by calmodulin and their binding to calmodulin-Sepharose. This indicated that phosphorylation prevented activation of these isoforms by preventing calmodulin binding. Based on these results, phosphorylation of the pump with PKC would be expected to increase free intracellular Ca2+ levels in those cells where isoforms 2a and 3a are expressed.

Immunolocalization of the multi-sarco/endoplasmic reticulum Ca2+ ATPase system in human platelets.

We recently identified a multi-SERCA (sarco/endoplasmic reticulum Ca2+ ATPase) system in haemopoietic cells comprising the SERCA 2b, SERCA 3 and a new monoclonal anti-Ca2+ ATPase antibody (PL/IM 430) recognizable SERCA isoforms. We have now investigated the subcellular localization of these enzymes in human platelets by Western blotting of subcellular membrane fractions and by immunoelectron microscopy. We precisely defined the recognition specificity of the polyclonal anti-SERCA 2b, anti-SERCA 3, anti-SERCA 1 antibodies as well as of the monoclonal antibody PL/IM 430 by testing their recognition of the tryptic fragments of the SERCA isoforms. The analysis of fragmented membranes enriched in plasma membrane and intracellular membrane components by Western blotting showed that the SERCA 2b and the SERCA 3 isoforms were found in both the plasma membrane and the intracellular membrane fractions, whereas the PL/IM 430 recognizable SERCA isoform was restricted to membranes associated with the plasma membrane fraction. The immunoelectron microscopical study of the SERCA isoforms in resting platelets showed that: (i) the SERCA 2b isoform was expressed in membranes associated with the plasma membrane and open canalicular system, some alpha-granules and in unidentified membranes; (ii) the SERCA 3 isoform was found associated with plasma and intracellular membranes; and (iii) the PL/IM 430 recognizable SERCA isoform was observed only in structures associated with the cytoplasmic face of the plasma membranes, as confirmed by flow cytometry. Finally, since the PL/IM 430 antibody was raised against intracellular membranes, we looked for a potential membrane redistribution during the isolation procedure used for the preparation of the immunizing membranes. Neuraminidase treatment indeed induced a translocation of the PL/IM 430 recognizable SERCA isoform from plasma to intracellular membranes. Thus, the multi-SERCA system in platelets: (i) is distributed over different platelet membranes, (ii) presents a sub-compartmental organization with some overlapping, and (iii) is partly associated with motile membranes, reflecting an unrecognized level of complexity of Ca2+ stores in these cells.

Characterization of the inositol trisphosphate-sensitive and insensitive calcium stores by selective inhibition of the endoplasmic reticulum-type calcium pump isoforms in isolated platelet membrane vesicles.

In mixed platelet membrane vesicles the presence of two distinct endoplasmic reticulum-type calcium pump enzymes of 100 and 97 kD molecular mass has been demonstrated. We have previously shown that both calcium pumps were recognized by polyclonal anti-sarcoplasmic reticulum calcium pump antisera [11]. In the present work we studied the effects of several calcium pump inhibitors on active calcium transport and inositol trisphosphate-induced calcium release in these vesicles in an attempt to assign the two calcium pump isoenzymes to specific calcium pools. The effect of the PL/IM 430 inhibitory anti-calcium pump antibody was compared to that of other calcium pump inhibitors acting predominantly on the 100 and the 97 kD calcium pump isoforms, respectively. The PL/IM 430 antibody, which recognized the 97 kD pump on Western blots and 2,5-di-(tert-butyl)-1,4-benzohydroquinone, which inhibited phosphoenzyme formation of the same pump isoform, inhibited calcium accumulation predominantly into an inositol trisphosphate-releasable calcium pool. On the other hand, low concentration of thapsigargin, which inhibited phosphoenzyme formation mainly of the 100 kD pump isozyme, had a more pronounced effect on calcium uptake into an inositol trisphosphate-resistant pool. These data suggest that in platelets the 97 kD calcium pump isoform is likely to be associated with the inositol trisphosphate-sensitive calcium storage organelle.

Simultaneous presence of two distinct endoplasmic-reticulum-type calcium-pump isoforms in human cells. Characterization by radio-immunoblotting and inhibition by 2,5-di-(t-butyl)-1,4-benzohydroquinone.

Phosphorylation, immunoblotting, limited proteolysis and drug-sensitivity analysis were used to characterize the sarcoendoplasmic-reticulum Ca2+ ATPases in a variety of human cell types. In platelets, several megakaryoblastoid and lymphoblastoid cell lines two distinct autophosphorylated forms of these ATPases with molecular mass of 100 and 97 kDa could be observed, whereas in several other cell types the 97 kDa form was absent. On immunoblots the 97 kDa species was specifically recognized by an inhibitory monoclonal antibody raised against the Ca2+ pump of platelet internal membranes, yielded on trypsinolysis a major fragment of 80 kDa, exhibited a distinct electrophoretic migration pattern as compared with the skeletal-, cardiac- and smooth-muscle Ca2+ pumps, and its autophosphorylation was strongly inhibited by the Ca(2+)-mobilizing agent 2,5-di-(t-butyl)-1,4-benzohydroquinone (tBHQ). The 100 kDa species reacted with an antibody specific for the cardiac- and smooth-muscle Ca2+ pumps, yielded on trypsinolysis fragments of 55 and 35 kDa, and its autophosphorylation was much less sensitive to tBHQ inhibition. These findings indicate the simultaneous presence of two different endoplasmic-reticulum Ca2+ pumps in a variety of human cell types, and may explain the previously observed differences in the Ca(2+)-handling characteristics of different intracellular Ca2+ pools and cell types.