immunolocalization of the multi-sarco/endoplasmic reticulum ca2+atpase system in human platelets

Immunolocalization of the multi-sarco/endoplasmic reticulumCa2þATPase system in human platelets

TUNDE KOVACS,1¹2 GAE TAN BERGER,3 ELISABETH CORVAZIE R,1 KATALIN PASZ TY,2 AN GI E BROWN,4 REGIS BOBE,1

BEL A PAPP,1 FRANK WUYTACK,5 EL ISAB ETH M. CRAMER3

AND JOCELYNE ENO UF1 1U.348 INSERM,

IFR Circulation Lariboisiere, Hopital Lariboisiere, Paris, France; 2National Institute of Haematology and Immunology,Budapest, Hungary, 3U.91 INSERM, Hopital Henri Mondor, Creteil, France, 4Department of Cardiology,King’s College Hospital, London, U.K.; and 5Laboratorium voor Fysiologie, Katholieke Universiteit Leuven,Campus Gasthuisberg, Leuven, Belgium

Received 15 May 1996; accepted for publication 9 December 1996

Summary. We recently identified a multi-SERCA (sarco/endoplasmic reticulum Ca2þ ATPase) system in haemopoie-tic cells comprising the SERCA 2b, SERCA 3 and a newmonoclonal anti-Ca2þ ATPase antibody (PL/IM 430) recog-nizable SERCA isoforms. We have now investigated thesubcellular localization of these enzymes in human plateletsby Western blotting of subcellular membrane fractions andby immunoelectron microscopy. We precisely defined therecognition specificity of the polyclonal anti-SERCA 2b, anti-SERCA 3, anti-SERCA 1 antibodies as well as of themonoclonal antibody PL/IM 430 by testing their recognitionof the tryptic fragments of the SERCA isoforms. The analysisof fragmented membranes enriched in plasma membraneand intracellular membrane components by Western blot-ting showed that the SERCA 2b and the SERCA 3 isoformswere found in both the plasma membrane and theintracellular membrane fractions, whereas the PL/IM 430recognizable SERCA isoform was restricted to membranesassociated with the plasma membrane fraction. Theimmunoelectron microscopical study of the SERCA isoformsin resting platelets showed that: (i) the SERCA 2b isoformwas expressed in membranes associated with the plasma

membrane and open canalicular system, some a-granulesand in unidentified membranes; (ii) the SERCA 3 isoform wasfound associated with plasma and intracellular membranes;and (iii) the PL/IM 430 recognizable SERCA isoform wasobserved only in structures associated with the cytoplasmicface of the plasma membranes, as confirmed by flowcytometry. Finally, since the PL/IM 430 antibody wasraised against intracellular membranes, we looked for apotential membrane redistribution during the isolationprocedure used for the preparation of the immunizingmembranes. Neuraminidase treatment indeed induced atranslocation of the PL/IM 430 recognizable SERCA isoformfrom plasma to intracellular membranes.

Thus, the multi-SERCA system in platelets: (i) is dis-tributed over different platelet membranes, (ii) presents asub-compartmental organization with some overlapping,and (iii) is partly associated with motile membranes,reflecting an unrecognized level of complexity of Ca2þ

stores in these cells.

Keywords: platelets, Ca2þ , SERCA isoforms, Ca2þ pools,immunolocalization.

The role played by Ca2þ ions in various aspects of cellfunction has been known for a long time. However, themechanisms involved in the regulation of the cytosolic Ca2þ

concentration are still under investigation, particularly innon-muscle cells. An increase in cytosolic Ca2þ concentra-tion accompanies cell activation and is due to both Ca2þ

influx from extracellular medium through different types ofCa2þ channels and to the Ca2þ release from intracellularCa2þ pools. A general agreement is growing on the presenceof at least two types of Ca2þ pools, inositol trisphosphate orIP3-sensitive Ca2þ pools, recruited upon phospholipase Cactivation, and IP3-insensitive Ca2þ pools, possibly recruitedby cyclic ADP-ribose. As recently reviewed (Pozzan et al,1994), many questions concerning these intracellular Ca2þ

stores are currently being investigated and debated, such aswhether these Ca2þ pools refer to distinct intracellular Ca2þ

compartments or subsets of intracellular Ca2þ compartments,

British Journal of Haematology, 1997, 97, 192–203

192 q 1997 Blackwell Science Ltd

Correspondence: Dr Jocelyne Enouf, U.348 INSERM, IFR CirculationLariboisiere, Hopital Lariboisiere, 8 rue Guy Patin, 75475 Paris,France.

193Sarco/Endoplasmic Reticulum Ca2þATPases in Platelets

q 1997 Blackwell Science Ltd, British Journal of Haematology 97: 192–203

as well as the possible interrelationship between these Ca2þ

compartments.The Ca2þ content of these intracellular Ca2þ pools is

controlled by intracellular type Ca2þATPases whose functionis to take up Ca2þ ions into the intracellular Ca2þ pools at theexpense of ATP. These Ca2þATPases, also termed sarco/endoplasmic reticulum Ca2þATPases or SERCAs, areencoded by the following multigenic family: SERCA 1,expressed in skeletal muscle (Korczak et al, 1988); SERCA2 which gives rise to the SERCA 2a and SERCA 2b isoforms,mainly expressed in cardiac and smooth muscle cells,respectively (Zarain-Herzberg et al, 1990), and SERCA 3which has been recently found expressed in non-muscle cells(Burk et al, 1989; Lytton et al, 1992; Anger et al, 1993; Bobeet al, 1994; Wuytack et al, 1994). Although these isoformsshow generally large similarities in terms of their biochem-ical characteristics, the possible biological significance of theexpression of different SERCA isoforms may be their associationwith distinct Ca2þ stores needed for a fine regulation ofcytosolic Ca2þ concentration.

Interestingly, we recently described an original multi-SERCA system in human platelets offering the opportunity toinvestigate the SERCA localization and consequently theCa2þ pool organization in haemopoietic cells. Indeed, theintracellular platelet Ca2þ transport system appears to beconstituted by the ubiquitous 100 kD SERCA 2b isoform(Enouf et al, 1992), the 97 kD SERCA 3 isoform (Bobe et al,1994; Wuytack et al, 1994) and a third putative novelSERCA isoform of 97 kD specifically recognized by amonoclonal antibody termed PL/IM 430 (Kovacs et al,1994). Although this last isoform remains to be identified, aclear differentiation between the two 97 kD isoforms, SERCA3 and the PL/IM 430 recognizable SERCA, has beendemonstrated by their different tryptic fragmentation. ThePL/IM 430 recognizable 97 kD SERCA was found to be muchmore sensitive to trypsin when compared to the prolongedtrypsinization required to digest the 97 kD SERCA 3 isoform(Kovacs et al, 1994). In addition, distinct trypsin sites arepresent on the two 97 kD SERCA isoforms because the PL/IM430 recognizable SERCA isoform gives an early 40 kDfragment whereas the SERCA 3 protein gives a late 80 kDfragment. Moreover, this multi-SERCA system is notexpressed in muscle cells and would appear to be typical ofhaemopoietic cells (Papp et al, 1992). Consequently, in orderto further advance in the understanding of the Ca2þ signal inthese cells, we investigated them for the localization of theseSERCA isoforms.

For this purpose, a series of different SERCA isoform-specific antibodies were used. Firstly, we re-explored theimmunoreactivities of the different antibodies, by studyingtheir recognition of the different SERCA isoforms on intactand proteolysed human platelet membranes. Indeed,because initial characterizations were performed usingheterologous systems which over-express a recombinantSERCA isoform, we re-investigated these characterizations inintact platelets which co-express a multi-SERCA system.Secondly, we investigated the subcellular expression patternof the different SERCA isoforms by platelet fractionation, andby immunogold labelling of ultrathin cryosections of intact

cells (the latter technique having been well documented toprecisely locate different platelet antigens (Berger et al, 1993;Rendu et al, 1993; Harrison & Martin-Cramer, 1993).

The results show a heterogenous distribution of the plateletmulti-SERCA system over different compartments with adifferentiation between the PL/IM 430 recognizable 97 kDSERCA, the SERCA 2b and SERCA 3 isoforms and suggestthat platelet Ca2þ pools should be regarded as complex,interacting and spatiotemporally highly organized systems.

MATERIALS AND METHODS

Materials. ATP (sodium salt), trypsin (type XIII), soybeantrypsin inhibitor, Bowman-Birk trypsin–chymotrypsin inhi-bitor, aprotinin, phenylmethylsulphonyl fluoride, bovine serumalbumin, anti-mouse-IgG horseradish–peroxidase conjug-ate, neuraminidase (type X) were purchased from SigmaChemical Co., St Louis, U.S.A. Anti-rabbit- and anti-guinea-pig IgG horseradish-peroxidase conjugates were obtainedfrom Jackson ImmunoResearch, West Grove, Pa., U.S.A.GP1b/CD 42b–IgG phycoerythrin conjugate was from Dako,Glostrup, Denmark. Protein A-gold was obtained from theDepartment of Cell Biology, Utrecht, The Netherlands. Goatanti-mouse-IgG gold conjugate was purchased from BioCell,Cardiff, U.K. [g-32P]ATP (110 TBq/mmol, 370 MBq/ml),Enhanced Chemiluminescence (ECL) Western blottingreagents and Rainbow coloured protein molecular massmarkers were from Amersham International, U.K. Glutar-aldehyde was from Ladd Research Industries, Burlington,U.K. The electrophoresis reagents were from Bio-Rad,Richmond, Calif., U.S.A. The nitrocellulose membraneswere from Schleicher and Schuell, Germany. All otherreagents were of analytical grade.

Preparation and subfractionation of platelet membranes.Normal human blood was obtained from healthy adultvolunteers and the investigation was performed according tothe requirements of the Declaration of Helsinki. Plateletmembrane vesicles were prepared as previously described(Le Peuch et al, 1983; Enouf et al, 1984). Briefly, freshplatelets were isolated and washed in modified Tyrode’sbuffer (36 mM citric acid, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2,103 mM NaCl, 5 mM glucose, pH 7.4). The cells were thendisrupted by controlled ultrasonication and centrifuged at19 000 g to eliminate unlysed platelets, mitochondria andgranules. The supernatant was centrifuged at 105 g and thepellet was used as a source of mixed membrane fractions.

In some experiments these membranes were furthertreated for the isolation of plasma and intracellularmembranes by layering the 105 g fraction over a 40%sucrose solution followed by centrifugation at 105 g for120 min. The two membrane fractions obtained at thesucrose interface and in the pellet were characterized bydifferent parameters (Enouf et al, 1984). Briefly, concana-valin A labelling, the phosphodiesterase assay, and thecholesterol/phospholipid ratio have been compared. Therecognition of plasma membrane antigens has been followedby antibodies raised against glycoprotein IIb–IIIa. Finally, thetwo membrane fractions were analysed by electrophoresis oftheir proteins. The differences observed led us to conclude

that the two membrane fractions refer to plasma andintracellular membranes.

Trypsin proteolysis of the membrane vesicles. Trypsintreatment of 105 g membrane vesicles was carried out inproteolysis medium containing 160 mM KCl, 17 mM K-Hepes,pH 7.0, 0.1 mM DTT, 0.05 mM CaCl2, and 400 mg ofmembrane protein/ml. Tryptic digestion was conductedusing 10 mg/ml trypsin at 48C for the times indicated.Digestion was stopped with 10-fold excess (w/v) of soybeantrypsin inhibitor. Digested samples were then used formembrane phosphorylation and Western blotting.

Phosphoenzyme intermediate formation of intact and proteo-lysed Ca2þATPases. Phosphorylation of Ca2þATPases with [g-32P]ATP (0.05 mM final concentration) was carried out for1 min at 48C in the proteolysis medium, as described earlier(Papp et al, 1991). The reaction was stopped by addition of asolution containing 6% trichloroacetic acid, 10 mM phos-phoric acid and 1 mM ATP. Precipitates were then washedthree times with the same solution and finally dissolved inthe electrophoresis sample buffer. The phosphorylatedproteins (50 mg/well) were submitted to 7.5% acidic SDS-polyacrylamide gel electrophoresis, electroblotted ontonitrocellulose membrane, and autoradiographed usingKODAK X-Omat AR films at ¹808C.

Antibody preparation and characterization. (i) Anti-SERCA 2bantibody. Antiserum against the SERCA 2b isoform wasraised against a peptide specific for SERCA 2b. Althoughthe first 993 amino acids are common to the SERCA 2aand SERCA 2b isoforms, they differ because the last four C-terminal amino acids in the SERCA 2a isoform are replacedin SERCA 2b by an extended sequence of 49 amino acids.The peptide used to prepare the antibody was specific forSERCA 2b as it encompassed the last 12 C-terminal aminoacids of this extended sequence. The preparation andspecificity of this antibody have both previously beendescribed (Wuytack et al, 1989).

(ii) PL/IM 430 monoclonal antibody. The PL/IM 430monoclonal antibody was raised using highly purifiedplatelet intracellular membranes as immunogen (Hack et al,1988b) and purified from culture supernatant of thecorresponding hybridoma cell line by protein A affinitychromatography at high ionic strength (3 M NaCl) asdescribed (Harlow & Lane, 1988). This antibody inhibitedCa2þ uptake into platelet membrane vesicles (Hack et al,1988b; Papp et al, 1993b) and blocked Ca2þ sequestrationinto intracellular Ca2þ pools in saponin-permeabilized platelets(Hack et al, 1988a), with no effect on Ca2þATPase activity oron phosphoenzyme intermediate formation of the plateletCa2þATPases. It specifically recognized the 97 kD SERCAisoform in platelets and in different human cell lines (Pappet al, 1992) but showed no cross-reactivity with the PMCA oferythrocyte plasma membrane (Papp et al, 1991).

(iii) Anti-SERCA 1 antibody. The polyclonal anti-SERCA 1antibody was elicited from guinea-pig against rabbit skeletalmuscle Ca2þATPase (SERCA 1 isoform) (Enouf et al, 1988)and purified by immunoaffinity chromatography on anantigen (SERCA1) column by SERBIO, Gennevilliers, France.This polyclonal antibody reacted with SERCA 1 in a rabbitskeletal sarcoplasmic reticulum membrane preparation and

recognized the surface-exposed large cytoplasmic loop(s) ofthe antigen. No cross-reactivity was observed between thisantibody and the erythrocyte or cardiac sarcolemmalCa2þATPases (PMCAs). Although the crude, non-purifiedantiserum slightly reacted with SERCA 2a in cardiac muscleand with SERCA 2b in smooth muscle (Magnier et al, 1992),the purified IgG did not show immunoreaction with eitherSERCA 2a or SERCA 2b (unpublished results).

(iv) Anti-SERCA 3 antibody. A polyclonal anti-SERCA 3antibody was elicited in rabbits against a synthetic peptidespecific for SERCA 3. This peptide was selected to be distinctfrom the homologous stretches of amino acids in SERCA 1 orSERCA 2. It encompassed amino acid residues 29–39 in theN-terminal part of the rat SERCA 3 protein. The anti-SERCA3 antibody recognized rat SERCA 3 protein expressed in COS-1cells, but did not react with SERCA 1, 2a and SERCA 2bproteins (Wuytack et al, 1994).

Immunodetection of the different SERCAs by Western blotting.Membrane proteins were submitted to either 7.5% acidicSDS–polyacrylamide gel electrophoresis or to 11% Laemmli-type SDS-PAGE and were electroblotted onto nitrocellulosemembranes. Membranes were blocked overnight at roomtemperature with a solution containing 10 mM Tris-HCl(pH 7.4), 150 mM NaCl, 0.1% Tween 20, and 5% non-fat drymilk. The membranes were then incubated with either a1 : 250 dilution of the anti-SERCA 2b serum, a 1 mg/mlsolution of the purified monoclonal antibody PL/IM 430, a1 : 1000 dilution of the purified anti-SERCA 1 antibody, or a1 : 1000 dilution of the N 89 anti-SERCA 3 antibody, for 2 hat room temperature. Thereafter, the nitrocellulose sheetswere extensively washed and further incubated for 2 h with a1 : 50 000 dilution of the anti-rabbit-IgG horseradish–peroxidase conjugate for immunostaining of the SERCA 2bisoform, a 1 : 1000 dilution of the anti-mouse-IgG horse-radish–peroxidase conjugate, and a 1 : 100 000 dilution ofthe anti-guinea pig-IgG horseradish–peroxidase conjugatefor the immunostaining of the PL/IM 430 recognizableSERCA isoform and a 1 : 100 000 dilution of the anti-rabbit-IgG horseradish–peroxidase conjugate for the immunostain-ing of the SERCA 3 isoform. After washing steps, antibodybinding was revealed by using ECL Western blotting reagentsaccording to the instructions of the manufacturer.

Immunoelectron microscopy. Blood samples were harvestedby venepuncture into plastic tubes containing ACD-C buffer(6.8 mM citric acid, 11.2 mM trisodium citrate, 24 mM glucose,pH 4.2). The platelet-rich plasma (PRP) was obtained bycentrifugation for 10 min at 180 g and 228C. Platelets wereisolated from PRP by centrifugation for 10 min at 1100 g and228C and washed three times with Tyrode buffer containing3.5 mg/ml bovine serum albumin. The washed platelets wereresuspended and fixed with 1% glutaraldehyde in 0.1 M

phosphate buffer, pH 7.4, for 1 h at 228C, washed three timeswith the same buffer, embedded in sucrose, and frozen inliquid nitrogen. Then, the immunochemical reactions wereperformed on thin sections collected on copper gridsaccording to the method of Slot et al (1988). Briefly, thesections were labelled by a first incubation with the primaryantibodies (1 : 10 dilution of the purified PL/IM 430monoclonal antibody, 1 : 30 dilution of the anti-SERCA 2b


194 Tunde Kovacs et al



antibody, 1 : 30 dilution of the anti-SERCA 1 antibody or1 : 30 dilution of the anti-SERCA 3 antibody) diluted inPBS containing 1% of bovine serum albumin for 20 min at228C, washed and then incubated with a 1 : 80 dilution ofprotein A–gold (10 nm) for 20 min at room temperaturefor immunolabelling of the polyclonal antibodies. Forimmunolabelling of the PL/IM 430 monoclonal antibody,protein A–gold was replaced by a 1 : 80 dilution of goat anti-mouse-IgG–gold conjugate. The sections were counterstainedusing 2% of uranyl acetate pH 7.0 and methyl celluloseuranyl. Samples were observed on a Philips 450 CM 10electron microscope.

Fluorescent staining. Washed platelets were isolated accord-ing to Radomski & Moncada (1983). Platelets were gentlymixed with a 1 : 100 dilution of PL/IM 430 monoclonalantibody followed by goat anti-mouse-IgG fluoresceinisothiocyanate (FITC) conjugate and incubated for 5 min atroom temperature. A second sample of platelets wasincubated with 0.05% of saponin for 10 min prior to theaddition of PL/IM 430 monoclonal antibody and the goatanti-mouse-IgG FITC conjugate. As a negative control a finalsample of platelets was mixed with a 1 : 100 dilution of goatanti-mouse-IgG FITC conjugate alone. Each of the plateletsamples was then double stained by incubating 5 ml of theplatelet suspension with 2 ml of the anti-GP 1b/CD 42bphycoerythrin conjugate for 5 min at room temperature.Subsequently, 250 ml ice-cold Tyrode buffer were added andthe platelets were immediately analysed using a FACScanflow cytometer (Becton Dickinson, Mountain View, Calif.,U.S.A.) connected to a consort 32 computer system.

Cytofluorimetric analysis. Data were recorded and analysedusing the LYSIS II software program (Becton Dickinson). Theinstrument was set for the measurements of forward and sidescatter, and for the green fluorescence and red fluorescence ofeach sample (the fluidics were set at a low flow rate toprevent platelet activation and to optimize resolution). Thered fluorescence threshold was set to exclude backgroundnoise and debris and include only cells specifically stained forglycoprotein 1b so that only platelets were in the analysisgate (Shattil et al, 1987). The acquisition gate was setaccording to the side and forward scatter properties of thecells. In each sample data from 104 platelets were recordedin list mode. Following data acquisition all samples wereanalysed together. A two-parameter display of the greenfluorescence versus forward scatter (a parameter related tocell size) was used to set the analysis gate. The greenfluorescence analysis gate was set above the level of non-specific fluorescence (derived from the analysis of cellsstained in the absence of antibody) so that any plateletsfalling above this line were specifically stained for the PL/IM430 recognizable SERCA. The FACScan was calibrated dailywith microbeads standards (Becton Dickinson).

RESULTS

Characterization of the reactivity of the different anti-SERCAantibodies on intact and trypsinolysed platelet Ca2þATPasesAs a necessary prerequisite for this study concerning the

immunorecognition of different SERCAs in platelets, we firstreviewed the question of the specificity of the anti-SERCAantibodies. Some of the antibodies used in the present studyhave been characterized previously (Wuytack et al, 1994,1989; Enouf et al, 1988), but these characterizations wereperformed either on intact Ca2þATPase systems or by testingtheir specificity using over-expression models for the differentSERCAs. In contrast, we tested the comparative immuno-reactivities of the different antibodies on intact humanplatelet membranes as controls and on the trypsinolysedCa2þATPases, since this provides the only available way todissociate the 97 kD SERCA 3 isoform from the 97 kD PL/IM430 recognizable isoform (see Introduction). Fig 1 shows thetryptic effect on the platelet Ca2þATPases which were tracedby means of their specific autophosphorylation (E , P) andthe recognition of the tryptic fragments by the anti-SERCA 2b, PL/IM 430, anti-SERCA 1 and anti-SERCA 3antibodies in parts A, B, C and D, respectively. The anti-SERCA 2b antibody recognized the 100 kD SERCA 2bisoform and its C-terminal 57 kD fragment (part A). ThePL/IM 430 monoclonal antibody recognized specifically the97 kD SERCA, its 73/68 kD doublet and the 40 kD fragment(part B). The anti-SERCA 1 and the anti-SERCA 3 antibodiesimmunostained both the 73/68 kD and 40 kD trypticfragments coming from the 97 kD PL/IM 430 recognizableisoform, and the 25 kD N-terminal fragment of the 97 kDSERCA 3 protein (parts C and D). However, the anti-SERCA 1antibody showed a stronger reaction with the 73/68 kD and40 kD fragments than with the SERCA 3 derived 25 kDfragment (part C) and the converse was true for the anti-SERCA 3 antibody (part D). These observations corroboratedthe higher recognition of the PL/IM 430 recognizable 97 kDisoform by the anti-SERCA 1 antibody (Kovacs et al, 1994)and an expected major recognition of the SERCA 3 isoformby the anti-SERCA3 antibody.

Hence, three antibodies were available: the anti-SERCA 3,the anti-SERCA 2b and the PL/IM 430 antibodies whichrecognized respectively the SERCA 3, the SERCA 2b and thenew 97 kD SERCA in human platelets. In addition, the anti-SERCA 1 antibody could be used to confirm the visualizationof this latter isoform; however, in this case some slight cross-reactivity with the SERCA 3 isoform must be taken intoaccount.

Comparative immunoblots of plasma and intracellular enrichedplatelet membranesIn an initial investigation of the subcellular localization ofthe different SERCAs in platelets, we performed membranefractionation to roughly dissociate between plasma mem-branes and intracellular membranes. Mixed platelet mem-branes were isolated (105 g fraction) and centrifuged through a40% sucrose solution. The enriched plasma membrane (PM)and intracellular membrane (IM) fractions obtained wereprobed for their different SERCA isoform contents (Fig 2) byWestern blotting using the anti-SERCA 2b (A), the PL/IM430 (B), the anti-SERCA 1 (C) and the anti-SERCA 3 (D)antibodies. The PL/IM 430 and anti-SERCA 1 recognizable97 kD isoform was found exclusively in the plasma-membrane-enriched fraction, whereas the 100 kD SERCA



Fig 1. Comparison of immunoreactivities of different anti-SERCA antibodies on tryptic fragments of the SERCA 2b, PL/IM 430 recognizableSERCA and the SERCA 3 isoforms in platelet membranes. Mixed human platelet microsomal membrane vesicles (105 g fraction) were proteolysedusing 10 mg/ml trypsin at 48C for 5 or 20 min. Then, the intact and trypsin-proteolysed Ca2þATPases were labelled with 0.05 mM [g-32P]ATP,resolved by electrophoresis on 7.5% acidic SDS-polyacrylamide gel (50 mg of membrane proteins/lane), blotted onto nitrocellulose membranesand either autoradiographed (panel E , P) or immunostained using the anti-SERCA 2b polyclonal antibody (panel A), the purified PL/IM 430monoclonal antibody (panel B), the anti-SERCA1 polyclonal antibody (panel C) and the anti-SERCA 3 polyclonal antibody (panel D) as describedunder Materials and Methods. Lanes 0, intact platelet Ca2þ ATPases. Lanes 5, 5 min trypsin-proteolysed platelet Ca2þ ATPases. Lanes 20, 20 mintrypsin-proteolysed platelet Ca2þATPases. The apparent molecular masses of the intact enzymes and their tryptic fragments were estimatedusing rainbow-coloured protein molecular mass markers. The filled arrows on the left side of this figure mark the E , P-forming species; the openarrows on the right side mark the immunostained species. The figure is typical of at least three different experiments.

Fig 2. Comparative Western blottings of platelet membrane fractionsenriched in plasma and intracellular membranes using the anti-SERCA 2b, the PL/IM 430, the anti-SERCA 1 and the anti-SERCA 3antibodies. The 105 g mixed platelet membrane preparation wasfurther fractionated by differential centrifugation on a 40% sucrosesolution as previously published (Enouf et al, 1984) and describedunder Materials and Methods. The resulting membrane fractionsenriched in plasma and intracellular membranes were submitted to11% SDS-PAGE (50 mg of membrane proteins/lane), electroblottedand immunostained using the anti-SERCA 2b antibody (A), the PL/IM430 antibody (B), the anti-SERCA 1 antibody (C) and the anti-SERCA 3 antibody (D) as explained under Materials and Methods.The apparent molecular masses of the native 100 kD SERCA 2b, the97 kD PL/IM 430 recognizable and SERCA 3 isoforms were estimatedusing rainbow-coloured protein molecular mass markers. Lanes PM,plasma membrane fractions. Lanes IM, intracellular membranefractions. The figure is typical of three different experiments.



Fig 3. Immunolocalization of the 97 kD PL/IM 430 recognizable SERCA isoform. (A) Resting normal platelets were immunolabelled with the PL/IM 430monoclonal antibody by the post-embedding immunogold staining technique as described under Materials and Methods. The gold particles (PL/IM 430recognizable SERCA epitope; arrowheads) appear mainly associated with the plasma membrane (pm) whereas only rare labelling is present inintracellular compartments. a: a-granule. (Original magnification ×72 000). (B) Cryosectioned resting normal platelets were immunolabelledwith the anti-SERCA 1 antibody. Immunogold particles (arrowheads) appear to be associated with similar structures as in part A, e.g. mainlyplasma membrane (pm) and some rare a-granules (a). (Original magnification ×45 000). (C) Resting normal platelets were treated for thedetection of the PL/IM 430 recognizable SERCA epitope by flow cytometry. Results express platelet number as a function of fluorescence intensityobtained using FITC goat anti-mouse-IgG in the absence (grey peak) or in the presence (white peak) of the PL/IM 430 antibody, the latter beingperformed before (a) and after (b) saponin permeabilization. In unpermeabilized platelets (a) the labelling for PL/IM 430 was close to backgroundstaining (grey peak) whereas permeabilization with saponin led to a significant increase in fluorescence intensity (b).

2b and 97 kD SERCA 3 isoforms appeared both in thefractions enriched in plasma and in intracellular membranes.So, whether or not the membrane fractions were totallypure, this already suggested that the PL/IM 430 recognizable97 kD isoform would present a subcellular localizationdistinct from the SERCA 2b and SERCA 3 isoforms.

Immunolocalization of the PL/IM 430 recognizable SERCA, theSERCA 2b and SERCA 3 isoforms in resting plateletsIn order to confirm these biochemical observations and torefine the subcellular localization of these SERCA isoforms,we further performed immunoelectron microscopy of intacthuman platelets using the same antibodies.



Fig 4. Immunolocalization of the SERCA 3 andSERCA 2b isoforms. Resting normal plateletswere treated for immunolabelling of the SERCA3 and SERCA 2b isoforms by the post-embedding immunogold staining technique asin Figs 3A and 3B. The labelling (arrowheads)for SERCA 3 isoform (A) shows randomdistribution, including an association with theplasma membrane (pm) and some a-granulemembranes (a) (original magnification×38 000), whereas the distribution of SERCA2b immunolabelling (B) appears moreextended and present on the plasmamembrane (pm), open canalicular systemmembrane (ocs), a-granule membrane (a) andunidentified intracytoplasmic vesicles (v).(Original magnification ×42 000.)



Immunolocalization of the PL/IM 430 recognizable SERCA.The immunogold labelling of the PL/IM 430 recognizableisoform appeared to be mainly restricted to the plasmamembrane (pm) of resting platelets (Fig 3A, centre cell).Indeed, whereas some gold particles appeared in the cyto-plasm, and a few at the lower left end of the cross-sectionmay be localized to channels of the dense tubular system, anumber of gold particles were spread over the surface

membrane. This result was confirmed by using the anti-SERCA 1 antibody, though there were additional goldparticles associated with intracellular compartments, includ-ing some a-granules (Fig 3B). To further specify the PL/IM430 epitope with respect to the side of the plasma membrane,we performed immunoflow cytometry experiments (Fig 3C).Indeed, as the size of immunoglobulins was larger than thethickness of the plasma membrane, one had to verify

Fig 5. Translocation of the PL/IM 430 recognizable SERCA isoform upon neuraminidase treatment. Fresh human platelets were eitherneuraminidase-treated or not (control) according to Menashi et al (1981). Platelet samples were thus either used for mixed microsomalmembranes isolation as described under Materials and Methods and further fractionated as in Fig 2, or for immunoelectron microscopy.

(A) Cell fractionation. The resulting plasma membrane (PM) and intracellular membrane (IM) fractions were electrophoresed (50 mg ofmembrane proteins/lane), blotted onto nitrocellulose membranes and treated with the PL/IM 430 antibody as in Figs 1 and 2. The molecular mass of thePL/IM 430 recognizable SERCA isoform was estimated using sainbow-coloured protein molecular mass markers. Lanes ¹: control platelets; lanesþ: neuraminidase treated platelets. The Figure is typical of three different experiments.

(B) Immunoelectron microscopy. After neuraminidase treatment, platelets display some morphologic signs of activation with the appearanceof pseudopods and a central bundle of filaments. Few a-granules remain in the cytoplasm. Immunogold labelling for PL/IM 430 has decreased onthe plasma membrane (pm) and has been redistributed toward the cell centre (arrow heads). a, a-granule membrane; ocs, open canalicularsystem. (Original magnification ×43 200.)

whether the immunolabelling was due either to a bona fideplasma-membrane localization of the antigen or to a sub-plasma-membrane localization in intracellular membranesadjacent to the plasma membrane. Intact whole plateletswere immunolabelled with PL/IM 430 in the absence (a) orin the presence (b) of prior permeabilization with 0.05%saponin. In unpermeabilized platelets (a) the PL/IM 430labelling was close to background (grey peak) whereaspermeabilization with saponin led to a significant increase influorescence intensity (b), demonstrating that the antibodyhas to enter the cell to recognize its antigen and that theepitope recognized by PL/IM 430 is located intracellularlybut close to the plasma membrane.

Immunolocalization of the SERCA 3 and SERCA 2b isoforms.Immunolocalization of the SERCA 3 isoform is shown inFig 4A. The gold particles showed a random distribution inthe cell, including association with plasma membranes andsome secretory a-granules (a). Hence, the SERCA 3 epitopesclearly showed a much more pronounced intracellularlocalization than was the case for the PL/IM 430 epitope.The SERCA 2b isoform (Fig 4B) showed a more diffuseimmunolabelling in the same platelets and was foundassociated with different plasma membranes and intracellularmembranes. Indeed, although SERCA 2b immunolabellingwas associated with both plasma membranes and theirinvaginations, i.e. the so-called open canalicular system(ocs), some gold particles were also found bound to themembranes of a-granules (a), as well as to unidentifiedintracytoplasmic vesicles (v).

Effect of neuraminidase treatment on the PL/IM 430 recogniz-able SERCA localizationThe plasma membrane associated localization of the PL/IM430 epitope was surprising, since this antibody wasraised against highly purified intracellular membranesisolated after neuraminidase treatment of intact platelets.A possible explanation of this discrepancy could be that theneuraminidase treatment results in some platelet activationand subsequent reorganization of intracellular structures.

To test this possibility, we treated intact platelets withneuraminidase, performed the same fractionation of plasma(PM) and intracellular (IM) enriched membranes as thoseused in Fig 2 and compared the immunoreactivities of thePL/IM 430 antibody with the plasma and intracellularmembrane fractions obtained under these experimentalconditions. Fig 5A shows that neuraminidase treatmentresulted in a net decrease in the PL/IM 430 recognition ofthe plasma membrane fraction, which was compensated by ahigher recognition of the intracellular membranes (comparelanes þ with lanes ¹). This translocation of the PL/IM 430recognizable SERCA detected by membrane fractionationwas further confirmed by electron microscopy (Fig 5B).Indeed, immunogold labelling for the PL/IM 430 epitope wasdecreased on the plasma membrane (pm) and redistributedtoward the cell centre. Thus, the most plausible explanationfor these results is that the PL/IM 430 recognizable isoform islocalized in intracellular membranes associated with plasmamembranes in resting platelets, which redistribute upon cellactivation.

DISCUSSION

Although the regulation of cytosolic Ca2þ concentration iswell documented in platelets (Siess, 1989; Rink & Sage,1990; Sargeant & Sage, 1994; Heemskerk & Sage, 1994),much less is known concerning the roles of the differentSERCAs-associated intracellular Ca2þ pools (Dean, 1989;Authi, 1993; Haynes, 1993). Accordingly, the first conclu-sion that can be drawn from this study is that in platelets theSERCAs are directly or indirectly associated with differentmembranes such as the plasma membrane, the opencanalicular system, a-granules and unspecified intracellularmembranes. This is in agreement with earlier studies using aless-specified anti-SERCA 1 antibody (Herbener & Dean,1988). Secondly, it suggests a particular subcompartmentaldistribution of the different SERCA isoforms. Evidence ispresented for the subplasmalemmal localization of the PL/IM430 recognizable SERCA, both by cell fractionation andimmunoelectron microscopy, using either the specific PL/IM430 monoclonal antibody or the polyclonal anti-SERCA 1antibody which also binds to the PL/IM 430 recognizable97 kD SERCA. In contrast, by taking into account the non-specific recognition of the PL/IM 430 recognizable 97 kDSERCA isoform by the anti-SERCA 3 antibody demonstratedin Fig 1, both the SERCA 2b and SERCA 3 isoforms areprobably dispersed inside the cell. Such a subcompartmentalorganization would agree with: (i) the extensive smoothendoplasmic reticulum described in mature platelets as thedense tubular system (Crawford & Scrutton, 1994); (ii) theevidence for distinct Ca2þ pools in platelets, as shown eitherby studies on isolated microsomes or on intact single cellsusing the comparative effects of thapsigargin and/orphysiological effectors to deplete distinct Ca2þ pools (Brune& Ullrich, 1991; Authi et al, 1993; Papp et al, 1993b;Heemskerk et al, 1993b; Cavallini et al, 1995; Engelenderet al, 1995); and (iii) the similarity between platelets andsome non-muscle cells, such as chromaffin cells whichpresent partial overlap of the IP3-sensitive and insensitiveCa2þ stores (Robinson & Burgoyne, 1991), in contrast withPC 12 cells in which total overlapping would exist (Zacchettiet al, 1991). Thirdly, the platelet Ca2þ store equipmentappears as a dynamic structure as has been recentlydescribed for endoplasmic reticulum in other cell types (Lee& Chen, 1988; Sitia & Meldolesi, 1992; Stendahl et al,1994). Indeed, at least the PL/IM 430 recognizable SERCAisoform-associated Ca2þ store appears linked to theplasma membranes, but after platelet activation followingneuraminidase treatment it was found migrating inside thecells as observed by cell fractionation as well as by immuno-electron microscopy. This controlled dynamic association ofthe PL/IM 430 recognizable SERCA with plasma membranesmay cause technical problems during the isolation ofmembrane fractions, especially from cells such as plateletswhich can be easily activated (Bokkala et al, 1995). This alsomight explain the paradox of why the PL/IM 430 mono-clonal antibody, which was obtained by using intracellularplatelet membranes isolated after free-flow electrophoresis,gave an apparent plasma membrane distribution. Finally, theplatelet model with its co-expression of distinct SERCA





isoforms, and corresponding Ca2þ pool organization intodistinct regions of the cell can function as a paradigm forother cells.

The characteristic distinct localization of the SERCA 2b,SERCA 3 and the PL/IM 430 recognizable SERCA invitesspeculation on the putative distinct roles of these Ca2þ pools.For the SERCA 2b-associated Ca2þ pool, we suggest that itrefers to the thapsigargin-sensitive Ca2þ pool and that it isresponsible for the basal Ca2þ response. This view is basedon: (i) the fact that significant Ca2þ release has beenattributed to the thapsigargin-sensitive Ca2þ pool; and (ii)our previous observation that the intracellular Ca2þATPaseinhibitor thapsigargin has a significant effect on the SERCA2b isoform (Papp et al, 1991). In agreement with thishypothesis is the fact that rat platelets which have beenfound to be deficient in intracellular Ca2þ release (Heemskerket al, 1993a) express comparatively less SERCA 2b isoformcompared with human platelets (Papp et al, 1993a). TheSERCA 3 isoform, which we found up-regulated underpathological conditions associated with an increase incytosolic Ca2þ concentration (Papp et al, 1993a), could befound in a Ca2þ pool particularly sensitive to a high cytosolicCa2þ concentration. This would confine the SERCA 3 proteinto some regions of the endoplasmic reticulum exhibiting ahigh Ca2þ concentration and it would explain the puzzling,but very typical, low affinity towards Ca2þ of the SERCA 3protein compared to the other members of the SERCA family(Lytton et al, 1992). The PL/IM 430 recognizable SERCAisoform-associated Ca2þ pool could correspond to the IP3-sensitive Ca2þ pool based on the observations that: (i) thefilling state of the PL/IM 430 recognizable SERCA isoform-associated Ca2þ pool is a prerequisite for IP3-induced Ca2þ

release (Papp et al, 1993a); and (ii) the localization of thePL/IM 430 recognizable SERCA corresponds to that of theIP3 receptor which has been found to be preferentiallylocated at the periphery of the platelet plasma membrane(Bourguignon et al, 1993). In addition, the platelet IP3

receptor (O’Rourke et al, 1995) was found associated withintracellular membranes obtained after cell fractionation(Dean & Quinton, 1995). Again, as shown here for the PL/IM430 recognizable isoform, this can mean some artefactuallocalization due to endoplasmic reticulum re-organization,although the presence of different IP3 receptor isoforms indifferent membrane fractions cannot be eliminated.

A localization of the IP3-sensitive Ca2þ pool in a subplasmamembrane region can be easily justified for an agonistreleasable Ca2þ pool. Indeed, the recruitment of this Ca2þ

pool would specifically be involved in the burst of Ca2þ

occurring as a response of the binding of platelet activators,phospholipase C activation and subsequent IP3 formation.The further question concerns the mechanism of replenish-ment of such a Ca2þ pool, particularly as the PL/IM 430recognizable SERCA isoform is difficult to autophosphorylateunder standard conditions. One possibility is that thisoccurs through the effect of a regulatory protein. TheRap1 protein has been suggested by different authors toregulate platelet Ca2þATPases. Interestingly, immuno-localization of Rap1 proteins in platelets showed anassociation of the Rap1 protein with plasma membranes

(Berger et al, 1994). In addition, our recent investigationsallowed us to propose that the Ca2þATPase target ofthis regulation should be the PL/IM 430 recognizableSERCA isoform, providing further evidence for thesubplasmalemmal localization of this putative Rap1-PL/IM430 recognizable SERCA complex-associated Ca2þ pool.Another alternative is that this Ca2þ pool uses its link tothe plasma membrane to allow Ca2þ influx from theextracellular medium, making the PL/IM 430 recognizableSERCA-associated Ca2þ pool a good candidate in thecapacitative model of Putney (1990). These hypotheses arecurrently under investigation.

ACKNOWLEDGMENTS

We are grateful to Professor N. Crawford, London, forgenerously supplying the PL/IM 430 hybridoma cells, toDr A.-M. Lompre, Orsay, as well as to the laboratory ofSERBIO, Gennevilliers, for preparing the polyclonal anti-SERCA 1 antibody and to Dr J.-M. Masse, U.91 INSERM,Creteil, for photographic work and technical help.

This work was supported mainly by Institut National dela Sante et de la Recherche Medicale U 348 and ReseauEst-Ouest No. 93 EO 01 and by grants from the OMFB(Hungarian–French Intergovernmental S&T CooperationProgramme, Ref. 6, NP-889/1994-F), the OTKA F 017669, the Comite de Paris de la Ligue Nationale contre leCancer, the Ministere des Affaires Etrangeres (Balaton Ref.94 008), the Fondation pour la Recherche Medicale and theMinistere de l’Education Nationale, de l’EnseignementSuperieur, de la Recherche et de l’Insertion Professionnelle(ACC SV No. 9).

REFERENCES

Anger, M., Samuel, J.L., Marotte, F., Wuytack, F., Rappaport, L. &Lompre, A.M. (1993) The sarco(endo)plasmic reticulumCa2þATPase mRNA isoform, SERCA3, is expressed in endothelialand epithelial cells in various organs. FEBS Letters, 334, 45–48.

Authi, K.S. (1993) Ca2þ homeostasis and intracellular pools inhuman platelets. Mechanisms of Platelet Activation and Control(ed. by K. S. Authi, S. P. Watson and V. V. Kakkar), pp. 83–98. NewYork: Plenum Press.

Authi, K.S., Bokkala, S., Patel, Y., Kakkar, V.V. & Munkonge, F.(1993) Ca2þ release from platelet intracellular stores bythapsigargin and 2,5-di-(t-butyl)-1,4-benzohydroquinone:relationship to Ca2þ pools and relevance in platelet activation.Biochemical Journal, 294, 119–126.

Berger, G., Martin-Cramer, E., Caen, J.P., Berndt, M.C. & Tanza, D.(1993) Ultrastructural demonstration of CD36 in the alpha-granule membrane of human platelets and megakaryocytes.Blood, 82, 3034–3044.

Berger, G., Quarck, R., Tenza, D., Levy-Toledano, S., de Gunzburg, J.& Cramer, E. (1994) Ultrastructural localization of the small GTP-binding protein rap1 in human platelets and megakaryocytes.British Journal of Haematology, 88, 372–382.

Bobe, R., Bredoux, R., Wuytack, F., Quarck, R., Kovacs, T., Papp, B.,Corvazier, E., Magnier, C. & Enouf, J. (1994) The rat plateletCa2þATPase isoform is the sarco/endoplasmic reticulumCa2þATPase 3 protein. Journal of Biological Chemistry, 269,1417–1424.

Bokkala, S., El-Daher, S., Kakkar, V.V., Wuytack, F. & Authi, K.S.(1995) Localization and identification of Ca2þATPases in highlypurified human platelet plasma and intracellular membranes:evidence that the monoclonal antibody PL/IM430 recognizes theSERCA 3 Ca2þATPase in human platelets. Biochemical Journal,306, 837–842.

Bourguignon, L.Y.W., Iida, N. & Jin, H. (1993) The involvement ofthe cytoskeleton in regulating IP3 receptor-mediated internalCa2þ release in human blood platelets. Cell Biology InternationalReports, 17, 751–758.

Brune, B. & Ullrich, V. (1991) Different calcium pools in humanplatelets and their role in thromboxane A2 formation. Journal ofBiological Chemistry, 266, 19232–19237.

Burk, S.E., Lytton, J., MacLennan, D.H. & Shull, G.E. (1989) cDNAcloning, functional expression, and mRNA distribution of a thirdorganellar Ca2þ pump. Journal of Biological Chemistry, 264,18561–18568.

Cavallini, L., Coassin, M. & Alexandre, A. (1995) Two classes ofagonist-sensitive Ca2þ stores in platelets, as identified by theirdifferential sensitivity to 2,5-di-(tert-butyl)-1,4-benzohydroqui-none and thapsigargin. Biochemical Journal, 310, 449–452.

Crawford, N. & Scrutton, M.C. (1994) Biochemistry of the bloodplatelet. Haemostasis and Thrombosis (ed. by A. L. Bloom, C. D.Forbes, D. P. Thomas and E. G. D. Tuddenham), pp. 89–114.Churchill Livingstone, London.

Dean, W.L. (1989) Structure, function and subcellular localization ofa human platelet Ca2þATPase. Cell Calcium, 10, 289–297.

Dean, W.L. & Quinton, T.M. (1995) Distribution of plasmamembrane Ca2þATPase and inositol 1,4,5-trisphosphate receptorin human platelet membranes. Cell Calcium, 17, 65–70.

Engelender, S., Wolosker, H. & de Meis, L. (1995) The Ca2þATPaseisoforms of platelets are located in distinct functional Ca2þ pools andare uncoupled by a mechanism different from that of skeletal muscleCa2þATPase. Journal of Biological Chemistry, 270, 21050–21055.

Enouf, J., Bredoux, R., Boizard, B., Wautier, J.L., Chap, H., Thomas, J.,de Metz, M. & Levy-Toledano, S. (1984) Simultaneous isolation oftwo membrane fractions: biochemical, immunological and func-tional characterization. Biochemical and Biophysical ResearchCommunications, 123, 50–58.

Enouf, J., Bredoux, R., Papp, B., Djaffar, I., Lompre, A.M., Kieffer, N.,Gayet, O., Clemetson, K., Wuytack, F. & Rosa, J.P. (1992) Humanplatelets express the SERCA2-b isoform of Ca2þ-transport ATPase.Biochemical Journal, 286, 135–140.

Enouf, J., Lompre, A.M., Bredoux, R., Bourdeau, N., de la Bastie, D. &Levy-Toledano, S. (1988) Different sensitivity to trypsin of thehuman platelet plasma and intracellular membrane Ca2þ pumps.Journal of Biological Chemistry, 263, 13922–13929.

Hack, N., Authi, K.S. & Crawford, N. (1988a) Introduction ofantibody (PL/IM430) to a 100 kDa protein into permeabilisedplatelets inhibits intracellular sequestration of Ca2þ. BioscienceReports, 8, 379–388.

Hack, N., Wilkinson, J.M. & Crawford, N. (1988b) A monoclonalantibody (PL/IM430) to human platelet intracellular membraneswhich inhibits the uptake of Ca2þ without affecting theCa2þ þ Mg2þ-ATPase. Biochemical Journal, 250, 355–361.

Harlow, E. & Lane, D. (1988) In: Antibodies, a Laboratory Manual,pp. 283–318. Cold Spring Harbor Laboratory, New York.

Harrison, P. & Martin-Cramer, E. (1993) Platelet alpha-granules.Blood Reviews, 7, 52–62.

Haynes, D.H. (1993) Effects of cyclic nucleotides and protein kinaseson platelet calcium homeostasis and mobilization. Platelets, 4,231–242.

Heemskerk, J.W.M., Feijge, M.A.H., Reitman, E. & Hornstra, G.(1993a) Rat platelets are deficient in internal Ca2þ release and

require Ca2þ influx of extracellular Ca2þ for activation. FEBSLetters, 284, 223–226.

Heemskerk, J.W.M. & Sage, S.O. (1994) Aspects of calcium signallingin platelets and other cells. Platelets, 5, 295–316.

Heemskerk, J.W.M., Vis, P., Feijge, M.A.H., Hoyland, J., Mason, W.T. &Sage, S.O. (1993b) Roles of phospholipase C and Ca2þATPase incalcium responses of single, fibrinogen-bound platelets. Journal ofBiological Chemistry, 268, 356–363.

Herbener, G.H. & Dean, W.L. (1988) Immunocytochemical localizationof the Ca2þ-ATPase polypeptide in human platelets. Biochemical andBiophysical Research Communications, 153, 848–853.

Korczak, B., Zarain-Herzberg, A., Brandl, C.J., Ingles, C.J.,Green, N.M. & MacLennan, D.H. (1988) Structure of rabbit fast-twitch skeletal muscle Ca2þATPase gene. Journal of BiologicalChemistry, 263, 4813–4819.

Kovacs, T., Corvazier, E., Papp, B., Magnier, C., Bredoux, R.,Enyedi, A., Sarkadi, B. & Enouf, J. (1994) Controlled proteolysisof Ca2þATPases in human platelet and non-muscle cell membranevesicles. Journal of Biological Chemistry, 269, 6177–6184.

Lee, H.C. & Chen, L.B. (1988) Dynamic behavior of endoplasmicreticulum in living cells. Cell, 54, 37–46.

Le Peuch, C.J., Le Peuch, D.A.M., Katz, S., Demaille, J.G.,Hinckle, M.T., Bredoux, R., Enouf, J., Levy-Toledano, S. & Caen, J.(1983) Regulation of calcium accumulation and efflux fromplatelet vesicles: possible role for cyclic-AMP-dependent phosphor-ylation and calmodulin. Biochimica et Biophysica Acta, 731, 456–464.

Lytton, J., Westlin, M., Burk, S.E., Shull, G.E. & MacLennan, D.H.(1992) Functional comparisons between isoforms of the sarco-plasmic or endoplasmic reticulum family of calcium pumps.Journal of Biological Chemistry, 267, 14483–14489.

Magnier, C., Papp, B., Corvazier, E., Bredoux, R., Wuytack, F.,Eggermont, J., Maclouf, J. & Enouf, J. (1992) Regulation of sarco-endoplasmic reticulum Ca2þ-ATPases during platelet-derivedgrowth factor-induced smooth muscle cell proliferation. Journalof Biological Chemistry, 267, 15808–15815.

Menashi, S., Weintroub, H. & Crawford, N. (1981) Characterizationof human platelet surface and intracellular membranes isolated byfree flow electrophoresis. Journal of Biological Chemistry, 256,4095–4101.

O’Rourke, F., Matthews, E. & Feinstein, M.B. (1995) Purification andcharacterization of the human type 1 Ins(1,4,5) P3 receptor fromplatelets and comparison with receptor subtypes in other normaland transformed blood cells. Biochemical Journal, 312, 499–503.

Papp, B., Corvazier, E., Magnier, C., Kovacs, T., Bourdeau, N.,Levy-Toledano, S., Bredoux, R., Levy, B., Poitevin, P., Lompre, A.M.,Wuytack, F. & Enouf, J. (1993a) Spontaneously hypertensive ratsand platelet Ca2þATPases: specific up-regulation of the 97 kDaisoform. Biochemical Journal, 295, 685–690.

Papp, B., Enyedi, A., Kovacs, T., Sarkadi, B., Wuytack, F.,Thastrup, O., Gardos, G., Bredoux, R., Levy-Toledano, S. &Enouf, J. (1991) Demonstration of two forms of calcium pumpsby thapsigargin inhibition and radioimmunoblotting in plateletmembrane vesicles. Journal of Biological Chemistry, 266, 14593–14596.

Papp, B., Enyedi, A., Paszty, K., Kovacs, T., Sarkadi, B., Gardos, G.,Magnier, C., Wuytack, F. & Enouf, J. (1992) Simultaneous presenceof two distinct endoplasmic-reticulum-type calcium-pump isoformsin human cells. Biochemical Journal, 288, 297–302.

Papp, B., Paszty, K., Kovacs, T., Sarkadi, B., Gardos, G., Enouf, J. &Enyedi, A. (1993b) Characterization of the inositol trisphosphate-sensitive and insensitive calcium stores by selective inhibition ofthe endoplasmic reticulum-type isoforms in isolated plateletmembrane vesicles. Cell Calcium, 14, 531–538.





Pozzan, T., Rizzuto, R., Volpe, P. & Meldolesi, J. (1994) Molecular andcellular physiology of intracellular calcium stores. PhysiologicalReviews, 74, 595–636.

Putney, J.W. (1990) Capacitative calcium entry revisited. CellCalcium, 11, 611–624.

Radomski, M.W. & Moncada, S. (1983) An improved method forwashing of human platelets with prostacylin. Thrombosis Research,30, 383–389.

Rendu, F., Martin-Cramer, E., Tenza, D. & Fischer, S. (1993)Redistribution of granulophysin and SRC protein in normal andgray platelets after activation. Platelets, 4, 167–174.

Rink, T.J. & Sage, S.O. (1990) Calcium signalling in human platelets.Annual Review of Physiology, 52, 431–449.

Robinson, I.M. & Burgoyne, R.D. (1991) Characterization of distinctinositol 1,4,5-trisphosphate-sensitive- and caffeine-sensitive cal-cium stores in digitonin-permeabilized adrenal chromaffin cells.Journal of Neurochemistry, 56, 1587–1593.

Sargeant, P. & Sage, S.O. (1994) Calcium signalling in platelets andother nonexcitable cells. Pharmacology and Therapeutics, 64, 395–443.

Shattil, S.J., Cunningham, S.M. & Hoxie, J.A. (1987) Detection ofactivated platelets in whole blood using activation dependentmonoclonal antibodies and flow cytometry. Blood, 70, 307–315.

Siess, W. (1989) Molecular mechanisms of platelet activation.Physiological Reviews, 69, 58–78.

Sitia, R. & Meldolesi, J. (1992) The endoplasmic reticulum: adynamic patchwork of specialised subregions. Molecular Biology ofthe Cell, 3, 1067–1072.

Slot, J.W., Geuze, J.J. & Weerkamp, A.J. (1988) Localization ofmacromolecular components by application of the immunogoldtechnique on cryosectioned bacteria. Methods in Microbiology, 20,236–246.

Stendahl, O., Krause, K.-H., Krischer, J., Jerstrom, P., Theler, J.-M.,Clark, R.O., Carpentier, J.-L. & Lew, D.P. (1994) Redistribution ofintracellular Ca2þ stores during phagocytosis in humanneutrophils. Science, 265, 1439–1441.

Wuytack, F., Eggermont, J., Raeymaekers, L., Plessers, L. &Casteels, R. (1989) Antibodies against the non-muscle isoformof the endoplasmic reticulum Ca2þ transport ATPase. BiochemicalJournal, 264, 765–769.

Wuytack, F., Papp, B., Verboomen, H., Raeymaekers, L., Dode, L.,Bobe, R., Enouf, J., Bokkala, S., Authi, K.S. & Casteels, R. (1994) ASERCA 3-type Ca2þpump is expressed in platelets, in lymphoidcells and in mast cells. Journal of Biological Chemistry, 269, 1410–1416.

Zacchetti, D., Clementi, E., Fasolato, C., Lorenzon, P., Zottini, M.,Grohovaz, F., Fumagalli, G., Pozzan, T. & Meldolesi, J. (1991)Intracellular Ca2þ pools in PC 12 cells: a unique, rapidly-exchanging pool is sensitive to both inositol 1,4,5-trisphosphateand caffeine-ryanodine. Journal of Biological Chemistry, 266,20152–20158.

Zarain-Herzberg, A., MacLennan, D.H. & Periasamy, M. (1990)Characterization of rabbit cardiac sarco/endoplasmic reticulumCa2þ ATPase gene. Journal of Biological Chemistry, 265, 4670–4677.

immunolocalization of the multi-sarco/endoplasmic reticulum ca2+atpase system in human platelets

Documents