HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 27, 24442-24452, July 5, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/27/24442    most recent M202011200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Martin, V. Articles by Enouf, J. Search for Related Content PubMed PubMed Citation Articles by Martin, V. Articles by Enouf, J. Social Bookmarking                      What's this?

Three Novel Sarco/endoplasmic Reticulum Ca²⁺-ATPase (SERCA) 3 Isoforms

EXPRESSION, REGULATION, AND FUNCTION OF THE MEMBERS OF THE SERCA3 FAMILY^*

Virginie Martin^§, Raymonde Bredoux, Elisabeth Corvazier, Roosje van Gorp, Tünde Kovàcs^¶, Pascal Gélébart, and Jocelyne Enouf^**

From  INSERM U348, IFR6 Circulation Lariboisière, Hôpital Lariboisière, 8 Rue Guy Patin, 75475 Paris Cedex 10, France and the ^¶ National Medical Center, Institute of Haematology and Immunology, H-1113 Budapest, Hungary

Received for publication, February 28, 2002, and in revised form, April 12, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Sarco/endoplasmic reticulum Ca²⁺-ATPases (SERCAs) pump Ca²⁺ into the endoplasmic reticulum. Recently, three human SERCA3 (h3a-c) proteins and a previously unknown rat SERCA3 (r3b/c) mRNA have been described. Here, we (i) document two novel human SERCA3 splice variants h3d and h3e, (ii) provide data for the expression and mechanisms regulating the expression of all known SERCA3 variants (r3a, r3b/c, and h3a-e), and (iii) show functional characteristics of the SERCA3 isoforms. h3d and h3e are issued from the insertion of an additional penultimate exon 22 resulting in different carboxyl termini for these variants. Distinct distribution patterns of the SERCA3 gene products were observed in a series of cell lines of hematopoietic, epithelial, embryonic origin, and several cancerous types, as well as in panels of rat and human tissues. Hypertension and protein kinase C, calcineurin, or retinoic acid receptor signaling pathways were found to differently control rat and human splice variant expression, respectively. Stable overexpression of each variant was performed in human embryonic kidney 293 cells, and the SERCA3 isoforms were fully characterized. All SERCA3 isoforms were found to pump Ca²⁺ with similar affinities. However, they modulated the cytosolic Ca²⁺ concentration ([Ca²⁺]_c) and the endoplasmic reticulum Ca²⁺ content ([Ca²⁺]_er) in different manners. A newly generated polyclonal antibody and a pan-SERCA3 antibody proved the endogenous expression of the three novel SERCA3 proteins, h3d, h3e, and r3b/c. All these data suggest that the SERCA3 gene products have a more widespread role in cellular Ca²⁺ signaling than previously appreciated.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The cytosolic free Ca²⁺ concentration is a dynamic signal regulating a variety of important cellular functions such as secretion, contraction, metabolism, neuronal plasticity, and gene transcription (1). The free Ca²⁺ concentration is strictly controlled, in space, time, and amplitude. This very tight control is necessary to enable cells to extract relevant information from the Ca²⁺ signal, but also implies that disturbances in the intracellular Ca²⁺ signal can lead to a plethora of consequences.

Ca²⁺ signaling is highly complex and comprises elementary and global oscillatory events involving endoplasmic reticulum (ER)¹ membrane proteins. Among the better known proteins are Ca²⁺ channels, namely inositol 1,4,5-trisphosphate receptors (InsP₃-Rs) (2) and Ca²⁺ pumps or sarco/endoplasmic reticulum Ca²⁺-ATPases (SERCAs) (3-5), which coordinate opposite Ca²⁺ ion fluxes. InsP₃-Rs induce peaks in cytosolic Ca²⁺ concentration as a result of Ca²⁺ release from ER, whereas SERCAs are involved in the falling phase of Ca²⁺ transients because of Ca²⁺ resequestration into the ER Ca²⁺ stores.

Although many data are available on the role of InsP₃-Rs in Ca²⁺ signaling, our knowledge of the role of SERCAs is more restricted. For 35 years research on these enzymes has focused on their biochemical and molecular characterization. The SERCA family includes three gene products, named SERCA1 (ATP2A1), -2 (ATP2A2), and -3 (ATP2A3), which give rise to alternatively spliced mRNA and protein isoforms. The SERCA1 and SERCA2 genes have been known for a while, and have two 3' end splice variants encoding different carboxyl termini isoforms, mainly expressed in adult (SERCA1a) and neonatal (SERCA1b) skeletal muscles, in cardiac muscle (SERCA2a), and in all cell types (SERCA2b). Intriguingly as regards the variety of non-muscle cell functions, compared with muscle cells, until 1998, a unique so-called non-muscle SERCA3a isoform was described (5-7).

In recent years, there have been enormous advances in our understanding of SERCAs, including new insight into their structure (8) and their relevant physiological functions in human diseases (4, 9, 10). SERCA1 gene mutations have been reported in some patients with Brody disease, a muscle disease. Various SERCA2 gene mutations have been reported in Darier-White disease, a rare dermatosis characterized by focal areas of separation between keratinocytes (4). Missense mutations in the SERCA3 gene are described in type II diabetic patients (10). In contrast, knockout of SERCA2 and SERCA3 genes leads to impaired cardiac performance (11), followed by squamous cell tumors (12) and defects in endothelium- and epithelium-dependent relaxation of vascular (13) and tracheal (14) smooth muscles, respectively. These molecular genetic studies of causal genes in humans, coupled with the observation that distinct phenotypes were found in either SERCA2 or SERCA3 knock-out models, point to compensatory phenomena or to many more diseases than those known today caused by an altered function of one of the known or unknown Ca²⁺-transporting proteins (15).

In the meanwhile, SERCA3 genes were suggested to possess a high degree of complexity. We and others showed that the human SERCA3 gene gives rise to two additional 3' end transcripts, SERCA3b and -3c (16-18), that we recently found to encode the predicted proteins in platelets and lymphoid Jurkat T cells (19). Moreover, these splice products would present an uncommonly high degree of species specificity, as mice express slightly different SERCA3b and -3c mRNAs (20) from those of human origin, whereas rats are devoid of similar splice variants, but express a so-called SERCA3b/c mRNA (21). Evidence for the existence of these mouse and rat SERCA3 proteins, however, is still lacking.

Here, we report (i) a growing family of SERCA3 isoforms, including the rat 3b/c and two novel human SERCA3d and SERCA3e proteins; (ii) their plural, diverse, and species-specific distribution patterns; (iii) factors involved in the regulation of their expressions; and (iv) their distinct functional properties determining both cytosolic and endoplasmic reticulum Ca²⁺ concentrations. Taken together, these results should help the understanding of pathophysiological cell Ca²⁺ signaling in a broad variety of cells.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation of Human and Rat Platelets-- Human blood was obtained from different healthy volunteers, and the investigation was performed according to the requirements of the Declaration of Helsinki. The isolation of human platelets was performed as previously described (22). Sixteen-week-old WKY rats and SHR were supplied by the Centre d'Élevage R. Janvier (Le Genest, France). Approximately 10 WKY rats and SHR were used. Animals were anesthetized with ether. Rat blood was diluted in 0.9% NaCl, and the isolation of rat platelets was performed as detailed in Ref. 21. The isolation of rat tissues was as detailed in Ref. 21.

Cell Culture-- The megakaryocytic CHRF-288 11, MEG 01, and Dami as well as adrenal pheochromocytoma PC12 cells were generously given by Dr. Lieberman, Pr. J. Peries, and Pr. Treiman, respectively. The KATO III (gastric cancer), lymphoblastoid Jurkat-T (JurE6-1 clone), promyelocytic HL-60, monocytic U937, HeLa (epithelial-like), NCI-H69 (lung cancer), and human embryonic kidney (HEK)-293 cell lines were obtained from the American Type Culture Collection (Rockville Pike, MD). All cells were grown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum and 2 mM L-glutamine. The PC12 cells were cultured as described in Ref. 23.

Cell Treatments-- Lymphocyte activation and induction of HL-60 cell differentiation were performed as previously described (24, 25).

Preparations of Rat and Human Platelet, Recombinant HEK-293 Cell, and Rat Cerebellum Lysates and Membrane Proteins-- Preparation of whole cell lysates from transfected HEK-293 cells (human recombinants) was as in Ref. 24. Isolation of rat and human platelet and recombinant HEK-293 cell enriched intracellular membranes was as described in Refs. 21 and 22. For rat cerebellum membrane isolation, 20 mg of tissue was lysed by adding 1 ml of buffer containing 10 mM KCl, 20 mM Hepes (pH 7.4), 200 mM sucrose, 50 µM EGTA, 50 µM EDTA, 0.5 mM dithiothreitol, 0.2 unit/ml aprotinin, 50 µM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin, 50 µg/ml soybean trypsin inhibitor, 40 µg/ml Bowman-Birk trypsin-chymotrypsin inhibitor, and 5 µg/ml pepstatin A. The lysate was centrifuged at 1200 × g for 15 min at 4 °C. The supernatant was centrifuged at 100,000 × g for 1 h at 4 °C. The pellet was resuspended in a buffer containing 160 mM KCl, 17 mM Hepes (pH 7.4), and 0.2 mM dithiothreitol; aliquoted; frozen; and kept at 80 °C. The protein concentration of the different lysates and membrane fractions was determined using bovine serum albumin as a standard.

Antibodies for Immunoblottings-- For endogenous rat SERCA2b (r2b), and human SERCA2b (h2b), the pan-r and -h2b polyclonal antibody, anti-2b (21), and pan-h2 monoclonal antibody, IID8 (22), were used (BIOMOL, Plymouth Meeting, PA). For r3a and r3b/c transfectants, the pan-r and -h3 polyclonal antibody, the N89 (PA1-910, Affinity Bioreagents, Neshanic Station, NJ), directed against the extreme NH₂-terminal part of the r3a protein (6, 7) and pan-SERCA polyclonal antibody, SERBIO (26), were used. For h3a-e transfectants, the N89 and a pan-h3, monoclonal antibody, PL/IM430, directed against the NH₂-terminal part of SERCA3 proteins, as well as our recently developed isoform-specific (h3a, h3b, and h3c) polyclonal antibodies (19), were used. For PMCAs and InsP₃-Rs, the pan-PMCA monoclonal 5F10, the polyclonal anti-InsP₃-RI (Affinity BioReagents), the polyclonal anti-InsP₃-RII (Santa Cruz Biotechnology, Santa Cruz, CA), and the monoclonal anti-InsP₃-RIII (Transduction Laboratories, Lexington, KY) antibodies (22) were used. Secondary anti-rabbit, anti-guinea pig, and anti-mouse horseradish peroxidase-conjugated antibodies were from Jackson Immunoresearch (West Grove, PA). Antibody binding was detected using enhanced chemiluminescence Western blotting reagents according to the manufacturer's instructions (Amersham Biosciences, Little Chalfont, Bucks, UK). Luminograms were scanned using Adobe Photoshop and, where indicated, quantified by Molecular Analyst, version NIH Image 1. 62b7.

Generation and Characterization of a Novel Isoform (r3b/c)-specific Antibody-- The antibody was generated by immunizing New Zealand rabbits with the bovine serum albumin-conjugated synthetic peptide (see Fig. 6A) as described for isoform-specific anti-h3a and anti-h3c antibodies (19). A portion of the antibody was purified from the final serum on peptide-Sepharose 4B affinity column. The sensitivity and the cross-reactivity of the antibody were analyzed by the Crosslink Laboratory (Budapest, Hungary) by performing both enzyme-linked immunosorbent assay and dot-blot experiments using the corresponding peptide-bovine serum albumin conjugate (data not shown).

Electrophoresis of Proteins and Western Blots-- Electrophoresis was performed on 8% SDS-PAGE, and Western blots were treated as in Refs. 19, 21, and 22, for the N89, PL/IM430, 5F10, anti-InsP₃-Rs, and anti-h3a-c antibodies. For the r3b/c protein, the nitrocellulose membranes were incubated with a 1:250 dilution of the anti-r3b/c antibody in Tris-buffered saline (pH 7.4), 0.1% Tween 20 for 4 h. After washing, the blots were treated with a 1:10,000 dilution of horseradish peroxidase-conjugated anti-rabbit IgG, for 1 h.

RNA Extraction and RT-PCR-- Total RNA extraction from human and rat platelets, human cell lines, and rat tissues was as described (21, 22, 25). For RT-PCR experiments, protocols essentially identical to those described in Refs. 21 and 22 were used. The primers used to amplify r2b and h2b, r3a, r3b/c, h3a, human SERCA3c (h3c), hPMCA1b, hPMCA4b, hInsP₃-R types I, II, or III are detailed in Refs. 21 and 22. The primers used (Genosys Sigma-Aldrich, St. Louis, MO) to amplify human SERCA3b (h3b), SERCA3d (h3d) and SERCA3e (h3e) are in Table I. PCR was initiated by adding 1.25 units of Gold AmpliTaq DNA polymerase (Thermus aquaticus, PerkinElmer, Branchburg, NJ) and Touch Down-PCR was performed for 10 cycles with annealing temperature decrement from 65 to 55 °C. PCR was conducted for different cycles, each consisting of successive periods of denaturation at 95 °C for 1 min, annealing at 55 °C (r2b and h2b), and 58 or 68 °C (Fig. 4b) for 1 min, and extension at 72 °C for 1 min. GAPDH and r2b or h2b amplifications were used as internal RNA controls. PCR products were visualized on ethidium bromide-stained 1.5 and 2% agarose gels or by Southern blotting as in Ref. 22. They were scanned using Adobe Photoshop and, where indicated, quantified by Molecular Analyst, version NIH Image 1.62b7.

Plasmid Constructions-- For expression constructs, the cDNAs encoding the h2b in pcDNA3.1 (Pr. D. H. MacLennan), r3a in pBR322 (Dr. G. Shull), and the inactive h3a in pMT2 (Dr. J. Lytton) were used. The r3a cDNA was NotI/NheI-excised and subcloned in EcoRI-digested and dephosphorylated pcDNA3 mammalian expression vector (Invitrogen, Cergy Pontoise, France) (Dr. A. M. Lompré). 3' end r3b/c DNA was generated by PCR encompassing exons 17-21 (nucleotides 2605-3263). Products were SalI/BspMI-excised after oriented subcloning in pCR2.1 vector. The r3b/c complete cDNA was created by switching a SalI/ApaI fragment of r3a by the specific 3' end r3b/c product. The cDNA of the inactive h3a was EcoRI-excised and subcloned in pUC18. The active h3a was obtained by substituting a PCR-amplified 1149-bp BclI/EcoRV fragment where Ile 817 was replaced by Met. This cDNA was subcloned into EcoRI site of pcDNA3. 3' end variant-specific cDNAs were generated by overlap extension of two variant-specific PCR products covering the region encompassing exon 18-23 (nucleotides 2674-3290). Specific products were subcloned in pCR2.1, and then EcoRV/XbaI-excised after oriented subcloning in pUC18 vector. The h3a-e cDNAs were created by switching the EcoRV/XbaI fragment of the active h3a construct, by the 3' end variant-specific products. Each construct was verified by automated dye terminator sequencing (Applied Biosystems AB1 100 model 377, Genome Express, Grenoble, France) across the junctions and through the modified regions (data not shown).

Stable Transfections in HEK-293 Cells-- cDNA for transfection was purified using the EndoFree plasmid kit (Qiagen, Hilden, Germany). Cells were plated at a density of ~1 × 10⁴ cells/cm² in 100-mm dishes. The following day, cells at ~60% confluence were transfected with 10 µg of the different r3s and h3s and h2b cDNAs using the transfection agent ExGen 500 (Euromedex, Souffelweyersheim, France) according to the instructions of the manufacturer. In short, cell monolayers were exposed to DNA-polymer complexes in a small volume of serum-free RPMI 1640 in an incubator. After 1 h, serum-containing medium was carefully added to monolayers. To establish cell lines that stably express the different SERCA3 isoforms, the neomycin-resistant colonies were selected with 600 µg/ml G418 (Invitrogen) in RPMI 1640. Approximately 2 weeks later, colonies were picked and grown as individual cell lines in the presence of 200 µg/ml G418 for 2 weeks. A mean of 10-12 neomycin-resistant clones were screened for each SERCA transfection efficiencies by Western blotting (data not shown). A mean of three of the highest responsive clones was selected for characterization and Ca²⁺ response studies.

Measurements of Ca²⁺ Pumping Activity-- Measurements of Ca²⁺ pumping activity was essentially performed as described (27). A cell sample of 0.4 ml was diluted into 1.6 ml of calcium-free, Hepes/KCl (pH 7.4) composed of 100 mM KCl, 100 mM sucrose, 20 mM Hepes, 1.4 mM MgCl₂, 1.25 mM NaN₃, 1 µM Fluo-3, 5 µg/ml oligomycin, and 37.5 µg/ml saponin. After 5 min at 37 °C, it was transferred to a constant temperature cuvette holder of a Shimadzu RF-1501 spectrofluorimeter (Shimadzu Europe, Duisburg, Germany). The free Ca²⁺ level was adjusted by stepwise additions from a concentrated CaCl₂ solution, after which 2 mM ATP was added. Fluorescence intensities (F) were continuously recorded at 488-nm excitation and 526-nm emission wavelengths (slits of 10 nm). Calibrations were performed by addition of CaCl₂ or EGTA to obtain F_max and F_min values, respectively. Levels of [Ca²⁺] were calculated from the binding equation [Ca²⁺] = K_d (F  F_min)/(F_max  F).

Measurements of Cytosolic Ca²⁺ Concentration [Ca²⁺]_c and Endoplasmic Reticulum Ca²⁺ Content [Ca²⁺]_er-- Confluent HEK-293 cells were loaded with 1.75 µM Fura-2-AM in culture medium for 45 min at 37 °C, and then incubated with fresh medium for another 15 min. Cells were rinsed with phosphate-buffered saline, detached, collected by centrifugation, and washed twice in 5 ml of Hepes buffer (pH 7.45) containing 136 mM NaCl, 10 mM Hepes, 2.7 mM KCl, 2 mM MgCl₂, 1 mg/ml D-glucose, and 1 mM CaCl₂. Fluorescence measurements were performed with cuvets containing 2 ml of cell suspensions in Hepes buffer, using the same fluorimeter. For calculation of the [Ca²⁺]_c, the ratio of fluorescence at excitation wavelengths of 340 and 380 nm (emission at 510 nm) was calibrated according to Ref. 28.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Alternative Splicing Generates Five Human SERCA3a-e mRNAs: Identification of Two Previously Unknown SERCA3d and SERCA3e Species-- The human and rat SERCA3 genes were recently found to give rise to distinct and species-specific 3' end splice variants, in addition to the h- and rSERCA3a mRNAs (h3a and r3a) (Fig. 1A). The hSERCA3b (h3b) and hSERCA3c (h3c) mRNAs result from the partial or complete insertion of a penultimate exon 21. In rat platelets a rSERCA3b/3c (r3b/c) variant (GenBank^TM accession no. AF458230) is present, in which the 5'-part of exon 21 lacks an ACLYP sequence (black box) and its 3'-part is extended (red box) and terminated using a novel stop codon Sb/c (21). A search for similar 3b/c splice variant in human platelets (Fig. 1B), performing RT-PCR, with sets of primers located in exons 18-22, was unsuccessful. However, additional longer PCR products were detected either on agarose gels (Fig. 1B, I, 7) or by Southern blotting (Fig. 1B, II, 7) when we used primers P3 and P6 amplifying exons 21 and 22. These data pointed to the insertion of a novel exon into the intronic sequence linking exons 21 and 22. Searching for this novel exon, we PCR-amplified and sequenced (data not shown) two products of 144 and 157 bp (Fig. 1B, III, 8).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Identification of two additional h3d and h3e platelet mRNAs. A, representation of the 3' ends of the human (upper) and rat (lower) SERCA3 genes. Boxes represent exons, and lines represent introns. Broken lines show the h3a, h3b, and h3c mRNAs and r3a and r3b/c mRNAs. Sc, Sa, Sb, Sb/c, and pA locate stop codons and polyadenylation signals. The black box represents the human ACLYP sequence absent in rats. The red part of exon 21 represents its extension in r3b/c. B, evidence for previously unknown human platelet SERCA3 mRNAs. RT-PCR demonstrating no DNA contamination (RT), control specific h2b, h3a-c (21) RNAs (I, 1-5), expected products between exons 18 and 21 (I, 6), but additional ones between exons 21 and 22, as visualized on agarose gels (I, 7) or by Southern blotting (22) (II, 7). These products referred to two products differing by 13 nucleotides (III, 8). C, new organization of the 3' end of the human SERCA3 gene. Exons and introns are shown in uppercase and lowercase letters, respectively. Nucleotides are numbered as by Dode et al. (17). The sizes of the penultimate and ultimate introns are indicated. Sd and Se denote the stop codons used in both h3c/h3d and h3a/h3e. D1, D2, and D3 are splice donors for h3b, h3c, and both h3d and h3e, respectively. D, expression of h3d and h3e mRNAs. Upper, representation of the h3d and h3e mRNAs (broken green and blue lines, respectively) and location of the primers used for their PCR amplifications. The yellow box represents the new exon 22. Lower, RT-PCR of h3d and h3e mRNAs in human platelets using the primers indicated, 250 ng of RNA and 35 cycles. In this figure and the following figures, numbers indicate the sizes of PCR products in base pairs (bp).

We analyzed the genomic sequence of the 3' end of hSERCA3 gene (EMBL accession no. Y3021426) (Fig. 1C), as well. We found a 58-bp exon, 188 bp from the beginning of the last intron, presenting canonical exon/intron boundary sites. This optional new exon 22 contains an internal 5' donor splice site (designated D₃). Its complete insertion in h3b and h3c mRNAs gives rise to two novel splice variants, termed hSERCA3d (h3d) (GenBank^TM accession no. AF458228) and hSERCA3e (h3e) (GenBank^TM accession no. AF458229), respectively. An illustration of how these SERCA3 splice variants are generated is shown in the upper part of Fig. 1D.

To prove that the additional PCR products shown in Fig. 1B (II, 7 and III, 8) were relevant to these h3d and h3e mRNAs, we PCR-amplified 3' end SERCA3 products covering exons 18-23 from human platelets (lower part of Fig. 1D). This demonstrated that these mRNAs include exon 18 (lanes 1 and 2), exons 20-22 (lanes 3 and 4), and exon 23 (lanes 5 and 6). Their identities were verified by sequencing (data not shown). Taken together, this splicing mechanism leads to two additional hSERCA3 proteins, which differ in their COOH termini. The h3b-specific (1043 aa) and h3c-specific (1029 aa) COOH termini comprising the last 50 and 36 aa are replaced by a tail of 51 and 59 aa in h3d (1044 aa) and h3e (1052 aa), respectively (sequences in Fig. 6A).

Differential Distribution of Rat and Human SERCA3 Splice Variants-- To address the question of the physiological significance of such a plurality of SERCA3 variants detected in platelets, their expression patterns were monitored in other rat and human cells and tissues (Fig. 2). We explored the expression of r3b/c by specific RT-PCR (21) in a panel of WKY rat tissues previously shown to co-express r2b and r3a (5, 29, 31). PC12 cells were included because they express two SERCA3 proteins, the 97-kDa r3a and a second one of 115 kDa (23, 32), the predicted molecular mass for r3b/c (Fig. 2A). rSERCA2b (r2b) and GAPDH were used as internal controls to evaluate our semiquantitative RT-PCR. Cerebellum, small and large intestine, pancreas, and spleen were found to co-express the r3b/c variant with r2b and r3a as well as PC12 cells, albeit at very different levels, the highest being found in cerebellum (389 ± 11%), spleen (167.5 ± 9.5%), and pancreas (118 ± 6%) compared with platelets.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Differential cell and tissue distribution of rat and human SERCA3 variants. A, RT-PCR (n = 6) showing differential co-expression of r3a and r3b/c in WKY rat tissues and PC12 cells compared with rat platelets (controls), using 250 ng (r2b, r3a, GAPDH) and 1000 ng (r3b/c) of RNA and 18 and 26 cycles, respectively. B, representation of hematopoiesis. The human cell lines are positioned along with the lineages. C, RT-PCR (n = 6) showing the expressions of h3a-e in these cells by using 250 ng of RNA and 18 (h2b, h3a), 24 (h3b, h3c), and 26 (h3d, h3e) cycles. The primers P1-P11, P2-P9, and P2-P10 were used to amplify h3b, h3d, and h3e, respectively. Inset, expressions of h3a-e in platelets by using 250 ng of RNA and the cycles indicated. D, RT-PCR (n = 6) showing the distribution of h3a-e in non-muscle cells and a fetal tissue using conditions of C. E, PCR (n = 4) showing the expression of h3a-e splice variants in normal human tissues (multiple tissue cDNAs, CLONTECH Laboratories, Inc.) (34) studied by using 1 ng of normalized cDNAs and 20 (h2b, h3a), 30 (h3b, h3c), and 35 (h3d, h3e) cycles. For quantifications, the control values (platelets (A), Jur E6-1 cells (C), NCI-H69 cells (D), and pancreas (E)) were taken as 100%. The expressions are given as percentages of the control values (mean ± S.E.).

For hSERCA3 mRNAs (Fig. 2, B-E), the expression patterns of h3a-e mRNAs were examined by specific RT-PCR using hematopoietic (panels B and C) and non-muscle cell lines (panel D) as well as different tissues (panel E). Cells representing early (MEG 01) and late (CHRF-288 11) stages of megakaryocytic differentiation, Dami, U937, and HL-60 cells were compared with platelets (19). Extension to non-muscle cells was performed using several cancerous cell lines (KATO III, HeLa, NCI-H69) and, furthermore, cell type of embryonic (HEK-293) and fetal origin. A widespread selection of commercial normal tissues was also tested. Again, the housekeeping hSERCA2b (h2b) and GAPDH were used as internal controls. As preliminary experiments showed large variations, cells and tissues expressing all h3a-e variants were used as references to choose variant-specific PCR experimental conditions. Hematopoietic cells were found to co-express the h3a-e variants to increasing degrees, based on the differentiation state of the cells and/or their lineages. This is clear when comparing MEG 01 cells, which express very low levels of h3a-e (7 ± 4, 8 ± 5, 3 ± 2.5, 9 ± 3, and 7 ± 2%, respectively, versus Jur E6-1 cells) with the more differentiated CHRF-288 11 or U937 and Jur E6-1 cells, which co-express the five h3a-e mRNAs at rather high levels, as do mature platelets. However, differences exist as regard their relative expressions, as shown in Fig. 2C (inset), which appear decremental from h3a to h3e in platelets, as well as in other hematopoietic cells (data not shown). SERCA3 gene expression is not restricted to hematopoietic cells, but was also found in other cell types. Different hSERCA3 splice variants were found to be co-expressed with h2b depending on the cell origin. Such a cell-specific distribution of h3a-e mRNAs is exemplified in: (i) KATO III (gastric carcinoma) and HeLa cells, which slightly express h3e (23 ± 13% versus NCI-H69 lung cancer cells); (ii) HEK-293 cells, which slightly express h3d (28 ± 6% versus NCI-H69 cells), and (iii) in fetal liver and NCI-H69 cells, which variably co-express the five h3a-e mRNAs (41 ± 7, 67 ± 7, 52 ± 6, 59 ± 17, and 84 ± 19%, respectively, versus NCI-H69 cells). Study of h3a-e expression in different tissues showed also different patterns with striking variations, the highest expression level being found in pancreas (100%) and lung for h3a-c (66 ± 10, 76 ± 18, and 69 ± 12%, respectively) and h3e (56 ± 4%). In contrast, h3d was rather equally expressed (from 60 to 100%) in the different tissues including cardiac and skeletal muscles, an unexpected finding that points to a housekeeping nature of this new SERCA3 isoform.

Control of the Transcription and Alternative Splicing of the Rat and Human SERCA3 Genes-- To look for mechanisms involved in these processes, we investigated both a pathological model and in vitro-induced differentiation and activation of several cell lines (Fig. 3). All these models were previously studied for modulation of SERCA expression; however, no data have been available for the expression patterns of the SERCA3 splice variants in these models. Neither are data for mechanisms regulating the expression of the SERCA3 gene products known.

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Factors involved in the control of rat and human SERCA3 alternative splicings. A, RT-PCR (n = 4) demonstrating opposite modulations of the expressions of r3a and r3b/c in WKY rats and SHR (conditions of Fig. 2A). B and C, RT-PCR (n = 5) showing differential regulations of the expressions of h3a-e upon PKC (10 nM PMA) (B) and retinoic acid receptor (1 µM ATRA) (C) treatments of HL-60 cells, respectively, by using 18 (h2b, h3a, GAPDH), 25 (h3b, h3c), and 28 (h3d and h3e) cycles. D, RT-PCR (n = 5) showing distinct regulations of same variants upon PMA (10 nM) and ionomycin (0.5 µM) treatments of Jur E6-1 cells by using 18 (h2b, h3a, GAPDH), 22 (h3b, h3c), and 26 (h3d and h3e) cycles. Quantifications are normalized to GAPDH and expressed in arbitrary units taken as 1 for control values: WKY rat platelets (A) and days 0 (B-D) (mean ± S.E.).

Hypertension (Fig. 3A) is an in vivo model accompanied with abnormal blood and vessel cell Ca²⁺ signaling. mRNAs were isolated from the WKY rat and SHR tissues indicated, and the level of splice variant expressions was analyzed as in Fig. 2. Although r3a was found differently up-regulated (× 1.4-2.1), a drastic down-regulation of r3b/c, was found in all the tissues except pancreas (not involved in hypertension), although to different degrees, the highest being found in cerebellum (× 7.3). This suggests that as yet unidentified factors involved in hypertension can directly or indirectly regulate rSERCA3 gene alternative splicing in vivo and that this is manifest as a specific feature in adult animals.

To investigate whether known factors, acting at different levels of the cell to achieve a specific cell function, can modulate SERCA3 splice variants, we studied the in vitro myeloid differentiation of HL-60 cells (Fig. 3, B and C). Treatment of these cells with PMA (panel B) induces their differentiation toward a monocyte/macrophage-like phenotype mediated by PKC activation, whereas ATRA (panel C) induces their terminal neutrophil granulocytic differentiation, mediated by nuclear factors. An opposite regulation of total h2 proteins was reported together with similar up-regulations of total h3 proteins (25). h2b mRNA expression agreed with protein data. For h3 RNAs, lineage-dependent, temporal, and variant-specific up-regulations of their expressions were observed. PMA induced delayed (days 2-3) h3a-e up-regulations as compared with ATRA (day 1). In addition to these distinct timings, they presented monophasic and biphasic patterns, respectively. Moreover, these convergent up-regulations differ in extent according to the splice variants. h3b (× 2.7) and h3c (× 2.0) were mainly concerned in PMA treatment of HL-60 cells, which express exon 21 but not exon 22 whereas ATRA preferentially affected h3d (× 1.4) and h3e (× 1.6), these two expressing both exons 21 and 22. This suggests that factors using different routes act at the selection of splice sites.

To look for transcription factors underlying differences in the expression of the splice variants, the model of lymphocyte activation (Fig. 3D) mimicked by the effect of a combination of PMA and ionomycin on Jurkat T cells was used. The resulting PKC activation and Ca²⁺ mobilization led to activation of calcineurin, induced signal transduction to the nucleus by transcription factor NF-AT, and altered gene expression. A down-regulation of total h3 proteins and an up-regulation of total h2 proteins were observed (24). Here, working at RNA level, minor up-regulation of h2b was observed coupled to different down-regulations of the h3a-e variants. Major down-modulations of the expression of h3a (86.5 ± 8.0%), h3c (58.5 ± 4.5%), and h3e (46.0 ± 7.5%) were detected, whereas the expression of h3b and h3d did not change such a significantly. Considering that the differences between the structures of h3b and h3d and those of h3c and h3e concerned the partial or complete insertion of exon 21, respectively, this may underline the role of the calcineurin-signaling pathway in the regulation of the h3b/h3c alternative splice site.

Stable Expression of Recombinant Rat and Human SERCA3 Isoforms-- To study the new isoforms, rSERCA3 (3a and 3b/c) and hSERCA3 (3a-e) cDNAs were constructed and stably expressed in HEK-293 cells. Although 10-12 clones were obtained for r3a and h3a-e recombinant proteins, only one, r3b/c, was established, which turned to be unstable. Varying conditions for r3b/c did not result in more efficient transfection. No explanation is yet available for such a discrepancy between this recombinant and the others.

Fig. 4 characterizes one clone of each SERCA3 stable transfectant at RNA (panels A and B) and protein levels (panels C and D). Moreover, because the overexpression of one type of Ca²⁺-ATPase has been reported to possibly influence the expression of the other structures involved in Ca²⁺ signaling (33), the expression of h2b and ubiquitous PMCA isoforms, hPMCA1b and hPMCA4b, as well as that of the hInsP₃-RI, II and III (22) were studied. In all cases, HEK-293 cells transfected with empty vector (pcDNA3) were used as controls.

View larger version (40K): [in this window] [in a new window]   Fig. 4.   HEK-293 cells stably transfected with r3a, r3b/c, and h3a-e constructs overexpress the corresponding recombinant proteins. Control HEK-293, HEK-293 cells stably transfected with empty vector (pcDNA3), and recombinants selected for their highest SERCA3 protein expressions (data not shown) have been characterized for r3 and h3 recombinants and endogenous h2b, hPMCAs, and hInsP3-Rs. A, RNA study of rat transfectants. RT-PCR (n = 5) was performed using 250 ng of RNA and 18 (r3a, r3b/c, h2b, GAPDH), 22 (hPMCA1b), 23 (hInsP3-RI, II, III), and 24 (hPMCA4b) cycles. B, RNA study of human transfectants (n = 3) by using 250 ng of RNA and 17 (h2b, h3a, GAPDH), 22 (hPMCA1b), 23 (h3b, h3c, hInsP3-Rs), 24 (hPMCA4b), and 25 (h3d, h3e) cycles. C, protein study of rat recombinants. Panel shows blots (n = 3) of 30 µg of rat platelet and recombinant membrane proteins (SERCAs), 5 µg of rat platelet, and 10 µg of recombinant membrane proteins (hPMCAs and hInsP3-Rs). D, protein study of human recombinants. Panel shows blots (n = 5) of 30 µg of human platelet membrane proteins (SERCAs, hPMCAs, and hInsP3-Rs), 10 µg of human whole cell lysate proteins of human recombinants (SERCAs and hPMCAs), and 50 µg of recombinant membrane proteins (hInsP3-Rs).

RNA studies showed for rat transfectants (panel A) (i) no modulation of endogenous h2b, (ii) the expected increase in r3a and r3b/c, (iii) slight up-regulations of hPMCAs, and (iv) slight variations in the expression of major hInsP₃-RI and minor hInsP₃-RII and RIII. Analyses of human transfectants showed (panel B) (i) no modulation of h2b except for h3e where it appeared to be down-regulated, (ii) the expected significant and specific up-regulations of h3a-e; (iii) no regulation of hPMCA1b, but a slight down-regulation of hPMCA4b in h3c and h3e, and (iv) slight regulations of hInsP₃-Rs.

Similarly, recombinants were studied for protein expressions by SDS-PAGE followed by Western blotting. In agreement with their predicted mass in kDa, resolution of r3 and h3 proteins by 8% SDS-PAGE permitted us to easily separate r3a (109.3) from r3b/c (116.4). For h3a-e, h3a (109.2) was separated from h3c (112.4), whereas h3b (113.9) co-migrated with h3d (114.1) and h3e (114.9) in our hands. For Western blotting, the same pan-SERCA3 antibodies (N89, SERBIO, PL/IM430) and rat and human platelet membranes were used to estimate transfection efficiencies. Isoform-specific h3a-c antibodies were used as further controls. Pan or subtype-specific antibodies were used for hPMCAs and hInsP₃-R proteins (panels C and D).

r3 (panel C) and h3 (panel D) recombinant proteins showed high transfection efficiencies compared with HEK-293 cells transfected with pcDNA3. The overexpression levels reached those of rat or human platelet membranes (100,000 × g fractions), as judged by higher or similar recognitions by the pan-SERCA3 antibodies. We could estimate that the increase in SERCA3 expression in transfected HEK-293 cells was ~150-200-fold over their endogenous level in HEK-293 cells and ~3- and 1.5-fold over the endogenous rat and platelet SERCA3a, respectively. As expected, h3a, h3b, and h3c were recognized by the isoform-specific anti-h3a-c antibodies (19), respectively. The anti-h3a antibody also recognized h3e. The peptide initially used to generate the anti-h3a antibody was later found to present partial sequence homology with h3e (see Fig. 6A). The electrophoretic migrations of the r3 and h3 recombinant proteins were in agreement with their predicted molecular weights, regardless of which antibody was used (NH₂-terminal- or COOH-terminal-specific). This ensured the intact nature of recombinant proteins. As concerns endogenous h2b; hPMCA1b and -4b; and hInsP₃-RI, -RII, and -RIII proteins, the results supported the slight variations of mRNAs except for hInsP₃-RIII in r3a and h3e, and hInsP₃-Rs in h3c. The apparent lower expression levels in r3 and h3 recombinants (compared with HEK-293 cells transfected with pcDNA3) are because of the high levels of overexpressed proteins that account in estimation of total proteins loaded onto the gels. As concerns the exceptions of hInsP₃-Rs, a possible explanation is their distinct membrane location and recovery in mixed membrane fractions.

Functional Characteristics of the New Rat and Human SERCA3 Recombinant Proteins in Cellular Ca²⁺ Responses-- Until now, no data have been published on intracellular Ca²⁺ responses in cells stably transfected with SERCAs. However, recent results, obtained using cells transiently transfected with h2b, showed a 2-fold increase in ER Ca²⁺ content ([Ca²⁺]_er) (34, 35). In our work on intact cells, the Ca²⁺ pumping properties of the r3 and h3 recombinant proteins were studied, using two techniques (Fig. 5). The first one involves saponin permeabilization of transfected HEK-293 cells in the presence of the non-cell permeant Ca²⁺ indicator Fluo-3, followed by ATP addition, upon which ATP-dependent removal of Ca²⁺ from the cytosolic compartment was monitored. This method eliminates PMCA activities and enables us to determine the specific extent and Ca²⁺ dependence of SERCA activity. The second technique involves the loading of transfected cells with the cell-permeant Ca²⁺ indicator Fura-2, to determine the resting cytosolic Ca²⁺ concentration ([Ca²⁺]_c) and that obtained after treatment with the Ca²⁺ ionophore, ionomycin. This Ca²⁺ response induced by 5 µM ionomycin in the presence of EGTA to prevent Ca²⁺ influx from extracellular space was used as an estimate of the Ca²⁺ content of the endoplasmic reticulum ([Ca²⁺]_er).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Recombinant rat and human SERCA3 proteins have distinct Ca2+ pumping properties and differently modulate [Ca2+]c and [Ca2+]er. A, by plotting different levels of free [Ca2+] versus the corresponding SERCA activity over the first 100 s after ATP addition (nanomoles of Ca2+/mg of protein/min), a free [Ca2+] corresponding to zero SERCA activity, the apparent Ca2+ affinity, could be calculated (typical experiment using a h3e-transfected cell line). B, apparent Ca2+ affinity for the different SERCA3 isoforms after correction for background level of SERCA activity in HEK-293 cells transfected with pcDNA3. C, SERCA activity determined at a free [Ca2+] of ~1100 nM (mean ± S.E. of n = 6-12) and corrected for background levels observed in HEK-293 cells transfected with pcDNA3. D, the [Ca2+]c was recorded in the presence of 2 mM EGTA, whereas the magnitude of the Ca2+ response evoked by a saturating dose of ionomycin (5 µM) was taken as an estimate of [Ca2+]er (typical experiment using a h3d-transfected cell line). E, quantitative comparison of the [Ca2+]c of the recombinants (mean ± S.D. of n = 9-12). F, quantitative comparison of the [Ca2+]er of the same recombinants (mean ± S.D. of n = 9-11).

Both techniques were validated by studying HEK-293 cells, stably transfected with an h2b construct, as compared with HEK-293 cells transfected with pcDNA3. No differences in their Ca²⁺ pumping activities were observed. Similarly, no major difference was found in their resting [Ca²⁺]_c, whereas an increase in [Ca²⁺]_er was observed, which reached a factor of 3 (data not shown).

For r3 and h3 recombinants, two to three clones were studied including those characterized in Fig. 4. Ca²⁺ pumping activities were measured at free [Ca²⁺] ranging from 500 to 2000 nM (Fig. 5A). The apparent Ca²⁺ affinities of the transfected proteins were estimated after extrapolation of the data to a free [Ca²⁺] at which the Ca²⁺ pumping activities are deduced to be zero. The values were corrected for background SERCA activity in HEK-293 cells transfected with pcDNA3 (88 ± 13 nmol/mg of protein/min). All recombinant proteins were found to function as Ca²⁺ pumps, exhibiting similar apparent affinities toward Ca²⁺ of ~800 nM (Fig. 5B). In addition, the data were used to estimate relative pumping activities at ~1100 nM free [Ca²⁺] (Fig. 5C), i.e. close to the theoretical affinity of rat and human SERCA3a protein deduced from Ca²⁺ transport studies (18). Recombinant proteins were found to possess different Ca²⁺ pumping activities (Fig. 5C), which vary from 49 ± 21 to 136 ± 47 nmol/mg of protein/min, over that of pcDNA3. The highest and lowest activities were found for r3a and h3c, respectively, whereas h3a-h3b-h3d and h3e gave intermediate values.

Resting [Ca²⁺]_c (Fig. 5, panels D (for typical results) and E) were found to vary from 35.8 ± 6.0 nM to 82.6 ± 6.7 nM, the highest [Ca²⁺]_c being obtained in h3c- and h3e-transfected cells. The [Ca²⁺]_c results from the combined activity of InsP₃-Rs, PMCAs, and SERCAs (Fig. 4). However, in our studies, high Ca²⁺ pumping activities and low resting [Ca²⁺]_c were found to coincide, thereby suggesting that it is the SERCA activity that determines [Ca²⁺]_c and that the involvement of PMCA activity in the present situation is minimal.

The [Ca²⁺]_er (Fig. 5, panels D (for typical trace) and F) of cells transfected with h3a, h3c, and h3e constructs were found to be similar or slightly lower than those for cells transfected by pcDNA3 (201 ± 25 nM). This slight decrease in releasable [Ca²⁺]_er might be explained by the formation of SERCA dimers, thereby creating Ca²⁺ channels as reported (34). h3b-, h3d-, and r3a-transfected cells were found to present considerably higher ionomycin-induced Ca²⁺ responses, reaching 358 ± 21, 348 ± 23, and 335 ± 21 nM, respectively. Among the alternatives to explain these differences in the increase in [Ca²⁺]_er is the localization of these SERCA-proteins in ER regions exposed to local increased [Ca²⁺] and/or their association with resident ER regulatory molecules, as has previously been shown for h2b (36, 37).

Taken together, our results indicate that all SERCA3 isoforms pump Ca²⁺ ions toward the ER and differently modulate both [Ca²⁺]_c and [Ca²⁺]_er, demonstrating for the first time the functional roles of SERCA3 gene products in living cells.

Existence of the Endogenous r3b/c, h3d, and h3e Proteins-- To prove the existence of native proteins, studies were performed on tissue or cell membrane proteins by Western blotting using different antibodies (Fig. 6). Fig. 6A shows the carboxyl termini of rat and human SERCA3 variants and the epitopes of the isoform-specific antibodies used. For the r3b/c protein (Fig. 6B), we used the N89 and SERBIO antibodies and generated a novel isoform-specific anti-r3b/c antibody. WKY rat and SHR cerebellum was studied because of the expected high expression of r3b/c in WKY rats and its down-regulation in hypertension (Fig. 2A). Western blot was first performed to check the specificity of the anti-r3b/c antibody (upper left). When using lysates from r3a-transfected HEK-293 cells, no protein band was detected, whereas a positive signal was obtained with the N89 antibody. By testing membranes fractionated from WKY rat cerebellum, with the N89, SERBIO, and anti-r3b/c antibodies in the absence or in the presence of the peptide used for immunization (upper middle), we detected the r3b/c protein. r3a and r3b/c proteins appeared as two dissociated protein bands with the N89 and SERBIO antibodies. The r3b/c protein (1061 aa) migrated at the expected molecular weight, i.e. at a lower position than PMCAs and a higher position than r3a (999 aa) and r2b (1042 aa). It co-migrated exactly with the upper bands recognized by the N89 and SERBIO antibodies. The r3b/c was recognized by the anti-r3b/c antibody, and the signal was abolished in the presence of 10 µM synthetic peptide used for immunization. This identity of the r3b/c protein was confirmed by comparing proteins isolated from WKY rat and SHR cerebellum (lower). As expected, in SHR membranes, r3a was up-regulated as revealed by the N89 and SERBIO antibodies whereas the r3b/c protein was drastically down-regulated, as shown by the N89, SERBIO, and anti-r3b/c antibodies. r2b was not modulated (21).

View larger version (32K): [in this window] [in a new window]   Fig. 6.   The three novel r3b/c, h3d, and h3e isoforms exist as endogenous proteins. A, carboxyl termini of r3b/c and hSERCA3 proteins and epitopes of the isoform-specific polyclonal anti-r3b/c and -h3a-c antibodies (19). Slashes mark the splice sites. The underlined sequences represent the amino acid stretches of the peptides used for immunization (r3b/c and h3a-c) and the partial sequence homology of h3e with h3a. B, evidence for r3b/c protein. Upper, specificity of the anti-r3b/c antibody. Left, blots of 20 µg of whole cell lysate proteins of the r3a recombinant. Right, blots of 10 µg of WKY rat cerebellum membrane proteins in the absence or presence of 10 µM peptide used for immunization (anti-r3b/c antibody). Lower, expressions of r3a and r3b/c in hypertension. Blots of 10 µg of cerebellum membrane proteins isolated from WKY rat (W) and SHR (S) (n = 3). C, evidence for h3d and h3e proteins. Left, PCR-amplified h3a-h3e RNAs in HEK-293 cells, same cells transfected with pcDNA3 or r3a cDNA by using 250 ng of total RNA and 28 cycles (n = 3). Right, blots of 2 µg of membrane proteins (mb) isolated from h3a and h3e recombinants, 100 µg of membrane proteins isolated from untreated HEK-293 cells, HEK-293 cells transfected with pcDNA3, or HEK-293 cells transfected with r3a (n = 3).

For the detection of the h3d and h3e proteins (Fig. 6C), we used the pan-anti-h3, PL/IM430 antibody. To look for cells expressing h3d and h3e proteins, we re-investigated HEK-293 cells. By modifying the conditions of PCR used in Fig. 4 (28 versus 18 cycles), and studying h3a-e RNAs expression in untreated HEK-293 cells and HEK-293 cells transfected either with pcDNA3 or with r3a, we (i) detected h3a in all three HEK-293 cell lines; (ii) showed the absence of h3b and h3c; (iii) confirmed the expression of h3d both in untreated HEK-293 cells and HEK-293 cells transfected with r3a, and furthermore its drastic down-regulation in HEK-293 cells transfected with pcDNA3; and (iv) found a slight expression of h3e in all three models of HEK-293 cells (Fig. 6C, left panel). The amounts of h3a, h3d, and h3e mRNAs, however, remained very low. To circumvent the problem of the predicted low expression level of proteins as well, we probed high amounts of HEK-293 cell membrane proteins.

Western blots (Fig. 6C, right panel) showed that two protein bands could be detected in untreated HEK-293 cells or in HEK-293 cells transfected with pcDNA3 or r3a, when using the PL/IM430 antibody. The positions of these two protein bands corresponded well to the molecular weights expected for h3a and h3d or h3e proteins and appeared to be similar to those of the h3a and h3e proteins, respectively (see transfectants, left panel). Similar relative amounts of h3a and h3d or h3e were observed in HEK-293 cells and the r3a-transfected ones (the apparent lower expressions in r3a transfectant being the result of the high level of overexpressed r3a protein, which accounts in calculating the quantity of total membrane proteins loaded onto the gel). As h3e mRNA is expressed at lower level than h3d mRNA, we concluded that the protein band at 114.1 kDa, immunostained with PL/IM430 mainly corresponded to the h3d isoform. The down-regulation of the PL/IM430-colored 114.1-kDa protein band in HEK-293 cells transfected with pcDNA3 further supports the expression of h3d in HEK-293 cells and cells transfected with r3a cDNA. Moreover, based on the expected very low expression of h3d in HEK-293 cells transfected with pcDNA3, we suggest that the remaining faint protein band at ~114.1 kDa refers to h3e. To try to establish the presence of h3e, we used the anti-h3a antibody that recognizes the h3e recombinant protein (Fig. 4D). However, no specific protein band could be detected with this antibody. It can confirm a very low expression of the h3e protein and/or the lack of sensitivity of this antibody under these conditions.

Taken together, these data demonstrate and/or strongly suggest the endogenous expression of the three novel SERCA3 isoforms, r3b/c, h3d, and h3e.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Here, we report that alternative splicings of the recently described SERCA3 genes result in multiple and species-specific variants, presenting diverse co-expressions in a variety of rat and human cells and tissues. Their splicing pattern is modulated as exemplified upon hematopoietic differentiation and in hypertension and is affected by various signalization pathways. Recombinant SERCA3 proteins exhibit distinct calcium pumping properties and differently modulate both basal [Ca²⁺]_c and the [Ca²⁺]_er. Western blot analysis performed using newly generated isoform-specific polyclonal anti-SERCA3 and a pan-SERCA3 monoclonal antibody demonstrated the existence of endogenous SERCA3 proteins. Altogether, the data show that alternative splicings of the SERCA3 genes generate a plurality of functionally distinct calcium pump isoforms in a broad variety of cells.

Multiple and Species-specific Co-expression of SERCA3 Isoforms-- The first five members of the SERCA family (1a, 1b, 2a, 2b, and 3a), described 10-15 years ago, were found by using techniques such as cloning or ribonuclease protection assay. One can say that a dogma has been accepted in the literature, consisting of (i) the strict muscle-specific expression of the 1a, 1b, and 2a variants; (ii) the ubiquitous nature of 2b; and (iii) the non-muscle-specific expression of the 3a isoform. Because of the low apparent Ca²⁺ affinity, its functional role in Ca²⁺ signaling has been questioned. Furthermore, no species specificity of these five SERCA proteins has been reported. Approximately 10 years were needed to discover two additional SERCA3b and -3c mRNAs by performing cloning and RT-PCR, a fact that initiated the revision of SERCA pumps and resulted in the finding and characterization of new SERCA3 gene products.

We demonstrated the existence of five additional SERCA3 proteins: h3b (19); h3c (19); and h3d, h3e, and r3b/c (present work) in several cell types and/or tissues. These findings double the known members of the SERCA family. In addition, the existence of two similar but distinct mouse 3b and 3c proteins are expected, as well. Such an increasing plurality of SERCA proteins may provide additional possibilities for various cell types to tightly regulate their elementary and/or global Ca²⁺ signals. There is variable and modulated redundancy of SERCA3 isoform expressions, according to cell and tissue functions. This would sustain the idea that the ER must be viewed as a modulable cellular compartment depending upon the relative distribution of SERCAs, which permit the governing of cell-specific Ca²⁺ signaling (38).

A complex species specificity of the expression of the SERCA3 isoforms is more and more evident, which resides both in the nature of the isoforms as mentioned above, and in their cell- or tissue-specific expression patterns. Rat and human platelets express two and five SERCA isoforms, respectively. Similar relative expressions of 3a and 3b variants are observed in human and mouse kidney (17, 20). Different relative expressions of 3a isoforms are found in rat and human brain, compared with pancreas or lung (Ref. 21 and present work). This particularity of the SERCA3 family might suggest that one isoform can be involved in different pathophysiological processes according to the species. Although this might explain some differences in humans and mice, it also complicates the use of animal models in seeking to understand the biological role of the hSERCA3 isoforms.

Our present data document that at least one member of the SERCA3 family, h3d, is equally expressed in all the tissues examined, including cardiac and skeletal muscles, abolishing the idea of the non-muscle cell-specific expression of the SERCA3 gene products.

Control of SERCA3 Alternative Splicings-- The expression of the SERCA3 variants can vary as a function of pathophysiological situations, as exemplified here upon hematopoietic cell differentiation and in hypertension. This manipulation of alternative splicing presents a unique way of regulating Ca²⁺ signaling processes. Many pathways are expected to regulate SERCA3 expression when considering the multiple sites for transcription factors present in the 5'-untranslated promoter region of the human SERCA3 gene (17). The present work points to some of them. Calcineurin is now attracting wide attention as a regulator of gene transcription involved in T-lymphocyte activation, morphogenesis of heart valves, or myocardial hypertrophy (39), all processes involving SERCA modulation. Here, we observed that the calcineurin pathway regulates SERCA3 gene expression in a splice variant-specific manner. This adds SERCAs to the other membrane proteins involved in Ca²⁺ signaling (40-42) recently shown to be targets of calcineurin regulation of transcription.

PKC signaling is involved in a variety of cell functions. Previously, we found major rapid and sustained induction of h3b and transitory h3c expression upon PMA-induced megakaryocytopoiesis (22). Here, we show that PKC-induced monocyte/macrophage differentiation is also accompanied by convergent transcriptional up-regulations of h3a-e, which mainly concern h3b and h3c, but they are sustained and delayed.

Retinoic acid receptors regulate specific genes coding for proteins involved in differentiation and/or growth arrest (43). This signaling pathway is also of clinical importance because differentiation therapy using retinoids is a very successful method in the treatment of acute promyelocytic leukemia. Our previous in vitro observations of up-regulations of SERCA3 proteins during myeloid differentiation were also found in vivo upon differentiation-induction therapy in patients with acute promyelocytic leukemia (25). ATRA-induced granulocytic differentiation may mainly involve up-regulations of h3d and h3e, according to our present data.

Taken together, this means that the expression of h3a-e variants can be tightly regulated by effectors using different routes and is manifest in a time- and species-dependent manner. This would allow fine and spatiotemporal variations in Ca²⁺ signals, a possible dynamic change in response to specific functional requirements, to fulfill optimally the demands for the fine-tuning of the intracellular effects of Ca²⁺ signaling.

SERCA3 Isoforms Modulate Ca²⁺ Signaling in Intact Cells-- Until recently, possibly because of the belief in their restricted and cell-specific expression, no major attention was paid to the role of SERCAs in Ca²⁺ signaling, except for SERCA2a (44). However, mathematical modeling of experimental data and studies using SERCA inhibitors show that ER Ca²⁺ pumps are involved in modifying Ca²⁺ transient profiles and oscillatory patterns (45). Recently, a direct and substantial refilling of ER by SERCAs occurring continuously during agonist-induced signaling was demonstrated (46). Overexpression of SERCA1 was found to cause an increase in the frequency of Ca²⁺ waves (47), and the temporal patterns of Ca²⁺ signals have been found to depend critically on the SERCA2a and -2b isoforms (36). Here, we show that SERCA3 splice variants encode proteins that present similar affinities toward Ca²⁺ but distinct effects on both [Ca²⁺]_c and [Ca²⁺]_er. Their similarities may suggest some interchangeability without replacing the original settings as shown for SERCA1 and SERCA2 isoforms (48, 49). Their differences in modulating both [Ca²⁺]_c and [Ca²⁺]_er should be of importance in a variety of pathophysiological situations. If [Ca²⁺]_c controls cytoplasmic enzyme activities, [Ca²⁺]_er is involved in fundamental biochemical processes such as the folding and processing of newly synthesized membranes and secretory proteins. Although, over the last decades, interest has been focused on disturbances of [Ca²⁺]_c, recently, the ER has attracted attention as a subcellular compartment in which [Ca²⁺] disturbances may contribute to pathological processes culminating in neuronal injury (30).

Because Ca²⁺ signaling accompanies many cell functions, the challenge now is to understand it and the mechanisms involved in the decoding of oscillatory calcium signals. A plurality of Ca²⁺ signaling may be expected to turn the activation of a tremendous variety of receptor-dependent transduction signaling into highly specific biological responses. Now, at a time that the field of subcellular Ca²⁺ homeostasis is evolving rapidly, the present findings open up areas of investigation seeking to understand normal and abnormal cell Ca²⁺ signaling.

	ACKNOWLEDGEMENTS

We thank Neville Crawford (Royal College of Surgeons, London, UK) and Frank Wuytack (Laboratorium voor Fysiologie, K.U., Leuven, Belgium) for the PL/IM430 and anti-2b antibodies; Marek Treiman (Department of Medical Physiology, Division of Renal and Cardiovascular Physiology, University of Copenhagen, The Panum Institute, Copenhagen, Denmark) for the PC12 cells; and Gary Shull (Department of Molecular Genetics, Biochemistry and Microbiology, University of Cincinnati College of Medicine, Cincinnati, OH), Anne-Marie Lompré (INSERM U446, Faculté de Pharmacie, Universite Paris Sud, Châtenay-Malabry, France), and David MacLennan (Banting and Best Department of Medical Research, University of Toronto, Toronto, Ontario, Canada) and Jonathan Lytton (Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Alberta, Canada) for the r3a, h2b, and h3a cDNAs. We thank Mickaël Leguet and Jérémy Odillard for blood and tissue collection and Véronique Briquet-Laugier for technical advice.

	FOOTNOTES

^* This work was supported in part by INSERM, the Association pour la Recherche sur le Cancer (France), and Hungarian Academy of Sciences Grant OTKA T032766.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EBI Data Bank with accession number(s) AF458228-AF458230.

^§ Recipient of fellowships from the Groupe d'Etudes sur l'Hémostase et la Thrombose (Amersham Biosciences) and the Société Française d'Hématologie.

Recipient of fellowships from Nestlé France and the Agence pour la Recherche et l'Information des Fruits et Légumes Frais.

^** To whom correspondence should be addressed. Tel.: 33-1-53-20- 37-91; Fax: 33-1-49-95-85-79; E-mail: jocelyne.enouf@inserm.lrb.ap-hop-paris.fr.

Published, JBC Papers in Press, April 15, 2002, DOI 10.1074/jbc.M202011200

	ABBREVIATIONS

The abbreviations used are: ER, endoplasmic reticulum; SERCA, sarco/endoplasmic reticulum Ca²⁺-ATPase; PMCA, plasma membrane Ca²⁺-ATPase; RT, reverse transcriptase; InsP₃-R, inositol 1,4,5-trisphosphate receptor; PMA, 4-phorbol 12-myristate 13-acetate; ATRA, all-trans-retinoic acid; WKY, Wistar Kyoto; SHR, spontaneously hypertensive rat(s); PKC, protein kinase C; h3, human SERCA3; h3a-e, human SERCA3a, -3b, -3c, -3d, and -3e isoforms; r3a, rat SERCA3a; r3b/c, rat SERCA3b/c; aa, amino acid(s); HEK, human embryonic kidney; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				P. C. Chandrasekera, M. E. Kargacin, J. P. Deans, and J. Lytton Determination of apparent calcium affinity for endogenously expressed human sarco(endo)plasmic reticulum calcium-ATPase isoform SERCA3 Am J Physiol Cell Physiol, May 1, 2009; 296(5): C1105 - C1114. [Abstract] [Full Text] [PDF]

				P. Nurden, N. Debili, W. Vainchenker, R. Bobe, R. Bredoux, E. Corvazier, R. Combrie, E. Fressinaud, D. Meyer, A. T. Nurden, et al. Impaired megakaryocytopoiesis in type 2B von Willebrand disease with severe thrombocytopenia Blood, October 15, 2006; 108(8): 2587 - 2595. [Abstract] [Full Text] [PDF]

				R. Rizzuto and T. Pozzan Microdomains of Intracellular Ca2+: Molecular Determinants and Functional Consequences Physiol Rev, January 1, 2006; 86(1): 369 - 408. [Abstract] [Full Text] [PDF]

				R. Bobe, R. Bredoux, E. Corvazier, J. P. Andersen, J. D. Clausen, L. Dode, T. Kovacs, and J. Enouf Identification, Expression, Function, and Localization of a Novel (Sixth) Isoform of the Human Sarco/Endoplasmic Reticulum Ca2+ATPase 3 Gene J. Biol. Chem., June 4, 2004; 279(23): 24297 - 24306. [Abstract] [Full Text] [PDF]

				L. Dode, J. P. Andersen, N. Leslie, J. Dhitavat, B. Vilsen, and A. Hovnanian Dissection of the Functional Differences between Sarco(endo)plasmic Reticulum Ca2+-ATPase (SERCA) 1 and 2 Isoforms and Characterization of Darier Disease (SERCA2) Mutants by Steady-state and Transient Kinetic Analyses J. Biol. Chem., November 28, 2003; 278(48): 47877 - 47889. [Abstract] [Full Text] [PDF]

				R. J. Fairclough, L. Dode, J. Vanoevelen, J. P. Andersen, L. Missiaen, L. Raeymaekers, F. Wuytack, and A. Hovnanian Effect of Hailey-Hailey Disease Mutations on the Function of a New Variant of Human Secretory Pathway Ca2+/Mn2+-ATPase (hSPCA1) J. Biol. Chem., June 27, 2003; 278(27): 24721 - 24730. [Abstract] [Full Text] [PDF]

				C. P. Chandrasekera and J. Lytton Inhibition of Human SERCA3 by PL/IM430. MOLECULAR ANALYSIS OF THE INTERACTION J. Biol. Chem., March 28, 2003; 278(14): 12482 - 12488. [Abstract] [Full Text] [PDF]

				L. Dode, B. Vilsen, K. Van Baelen, F. Wuytack, J. D. Clausen, and J. P. Andersen Dissection of the Functional Differences between Sarco(endo)plasmic Reticulum Ca2+-ATPase (SERCA) 1 and 3 Isoforms by Steady-state and Transient Kinetic Analyses J. Biol. Chem., November 15, 2002; 277(47): 45579 - 45591. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/27/24442    most recent M202011200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Martin, V. Articles by Enouf, J. Search for Related Content PubMed PubMed Citation Articles by Martin, V. Articles by Enouf, J. Social Bookmarking                      What's this?