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.

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.

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.