Biosynthesis of platelet glycoprotein V expressed as a single subunit or in association with GPIb-IX Catherine Strassel, Sylvie Moog, Marie-Jeanne Baas, Jean-Pierre Cazenave and Franc¸ois Lanza INSERM U.311, Etablissement Franc¸ ais du Sang-Alsace, Strasbourg, France Glycoprotein (GP) V is noncovalently linked to GPIba, GPIbb and GPIX within the platelet GPIb–V–IX complex, a receptor for von Willebrand factor and thrombin. Two functions have been ascribed to GPV, namely, the modula- tion of thrombin- and collagen-dependent platelet responses. The biosynthesis of this molecule was investigated in pulse– chase metabolic labelling experiments performed in CHO cell lines transfected with GPV, alone or in the p resence o f GPIb–IX. GPV could not be detected at the surface of cells expressing the single subunit but was f ound instead as a soluble f orm i n the culture m edium. In pulse–chase studies, an immature 70 kDa protein was detected in cell l ysates, whereas a fully processed 80–82 kDa form was only observed in the culture supernatants at later chase times. Immature GPV w as N-glycosylated and r etained before the medial Golgi while the secreted molecule contained c omplex sialylated sugars. The mature soluble form of GPV was produced by an enzymatic cleavage wh ich w as not affected by inhibitors of proteasome, calpain or metalloproteinases. When GPV was cotransfected with GPIb–IX, the former was no longer found in the c ulture supernatant but was retained in the cell m embrane as shown by fluorescence- activated cell sorting and confocal microscopy analyses. Surface expressed GPV was processed from an immature 70 kDa form to produce a mature 80 kDa protein, pro- cessing similar t o the intracellular trafficking of GPIb a. These results indicate that correct biosynthesis and surface expression of GPV in platelets requires the presence o f the other s ubunits of the G PIb–V–IX complex. Keywords: biosynthesis; CHO; glycoprotein; glycosylation; platelet. The platelet G PIb–V–IX complex p lays an essential role in the f ormation of the haemostatic plug following vessel w all injury [1]. This cell surface receptor is r equired above a ll under the conditions of high shear stress encountered in arteries and microvessels, where it mediates reversible platelet adhesion by binding to von Willebrand factor (vWF) exposed on the damaged vessel w all [2]. T he functional importance of the GPIb–V–IX complex is highlighted by th e existence of the Bernard–Soulier bleeding syndrome, in which the complex is p resent in strongly decreased amounts or in more rare cases is not functional. GPIb–V–IX is composed of four different type I glycopro- teins: G PIba,GPIbb, GPIX and GPV, al l b elonging t o t he Ôleucine-ric h repeat family Õ [3]. GPIba (135 kDa) is disul- phide-linked to GPIbb (25 kDa) and noncovalently com- plexed with GPIX (20 kDa) and GPV (82 kDa) in a 2 : 2 : 2 : 1 s toichiometry [4,5]. Several f unctions have been proposed for glycoprotein (GP) V . It could act as a negative regulator of thrombin activation [6], or as an a ccessory receptor for collagen dependent platelet adhesion and a ctivation [7]. Although i t has not yet been d emonstrated, GPV could potentially play a role in platelet signalling during adh esion to vWF. Thus, 14–3-3f [8,9] and the calcium dependent regulator calmo- dulin [10] have been shown to b ind to t he cytoplasmic domain o f G PV. A distinctive f eature of this molecule is its extreme sensitivity to cleavage by proteases. After throm- bin-induced platelet activation, cleavage of GPV at a specific site releases a soluble 69 kDa fragment (f1) [11], the physiological significance of which is still unknown. In addition, GPV can be cleaved by elastase to release a 75 kDa fragment and by calpain, at a site near the cell membrane, to generate an 82 k Da fragment [12]. Despite its association with the other subunits of the GPIb–V–IX complex, GPV is more loosely attached [4] and does not appear to be essential for the normal expression o f these subunits. GPIb–IX is expressed e fficiently in transfected cells in the absence of GPV and in the platelets of GPV knock-out mice [13–15]. This is consistent with the obser- vation that al l the reported B ernard–Soulier defects are restricted to the GPIba,GPIbb and GPIX subunits [16]. Nevertheless, s ome studies in transfected cells have indicated that GPV could enhance levels of expression of the complex at the platelet surface [17]. The available biosynthetic studies of receptors restricted to platelets (GPIb, GPIIbIIIa) have essentially relied on their expression in heterologous cells. A pproaches of this type have improved our understanding of the requirements for normal biosynthesis of the GPIb–IX complex and of the consequences of mutations encountered in Bernard–Soulier patients [18]. T hese studies were, however, performed mostly in the absence GPV. On the contrary, the biosynthesis of Correspondence to F. Lanza, INSERM U.311, Etablissement Franc¸ ais du Sang-Alsace, 10 rue Spielmann, BP 36, 67065 Strasbourg Cedex, France. Fax: +33 388 21 25 21, Tel.: +33 388 21 25 25, E-mail: francois.lanza@efs-alsace.fr Abbreviations: Endo-H, endoglycosidase-H; FACS, fluorescence- activated cell sorting; FITC, fluorescein isothiocyanate; GP, glycoprotein; SV40, simian virus 40; vWF, von Willebrand factor. (Received 20 April 2004, revised 21 June 2004, accepted 26 July 2004) Eur. J. Biochem. 271, 3671–3677 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04304.x GPV has never been thoroughly i nvestigated. A better knowledge of the processing of GPV i s likewise necessary to obtain greater ins ight into i ts str uctural and f unctional r ole in platelets. To address t hese questions, w e have developed two heterologous expression cell systems where GPV was transfected in the presence or absence of the other three subunits of the GPIb–V–IX complex, allowing study of its biosynthesis and cellular distributio n. Experimental procedures Materials and cell lines The mAbs V.1 and V.5 ag ainst GPV, ALMA.12 against GPIba, ALMA.16 against GPIX and RAM.1 against GPIbb, were developed in our laboratory [19]. CHO cell lines CHO-K1 and CHO-DUK (deficient in dihydrofolate reductase) were purchased from ATCC (Rockville, MD, USA). The CHO/GPIb–V–IX cell line was derived from a clone expressing GPIbb and GPIX (gift from J. A. Lopez, Baylor College of Medicine, Houston, TX, U SA) by transfection with cDNAs coding for GPIba inserted in the pDX vector [20] and GPV inserted in the pZeoSV vector (Invitrogen, San Diego, CA, USA) and selection in the presence of zeocin. These cells were grown in aMEM (Gibco BRL, Cergy-Pontoise, France) supplemented with 10%(v/v) fetal bovine serum, 200 lgÆmL )1 zeocin and 400 lgÆmL )1 G418 (Boehringer–Mannheim, Germany). The CHO/GPV cell line was obtained by transfection of CHO-DUK cells with GPV cDNA inserted in the pTG2328 vector, selection in the absence of nucleosides and a mplification of GPV expression by growing c ells in the presence of step i ncreased concentrations of methotrexate [21]. This clone was main- tained in suspension in s erum fre e CHO-S-SFMII mediu m (Gibco BRL) supplemented with 1.2 lgÆmL )1 methotrexate (Calbiochem Novabiochem, La Jolla, CA, USA). Expres- sion in pZeoSV and pTG2328 is driven by the simian virus 40 (SV40) early enhancer/promoter. Flow cytometry CHO cells (2 · 10 5 in 100 lL) were incubated for 30 min at 4 °C with purified IgG (10 lgÆmL )1 ) in fluorescence microfluorimetry buffer [RPMI medium, 5% (v/v) normal goat serum, 0.2% (v/v) sodium azide]. After centrifugation at 270 g, the cells were resuspended in buffer containing a 100-fold dilution of fluorescein isothiocyanate-(FITC)-con- jugated g oat a nti-(rat I gG) F( ab¢) 2 or FITC-conjugated goat anti-(mouse IgG) IgG (Jackson Immunoresearch, West Baltimore, PA, USA) for 30 min at 4 °C. Analyses were performed on 10 000 cells in a FACS Calibur flow cytometer (BD Biosciences, Rungis, France). 35 S metabolic labelling CHO cells (2 · 10 8 ÆmL )1 ) were incubated t wice in 50 mL of cysteine- and methionine-free RPMI medium (ICN, Costa Mesa, CA, USA) without serum for 30 min at 3 7 °C. The cells were then pulse-labelled for 15 min a t 37 °C in the same medium containing 1% (v/v) penicillin–strep- tomycin–glutamine (Gibco), 20% (v/v) dialysed f etal bovine serum (Biowest, Nuaille, France) and a mixture of [ 35 S]methionine and [ 35 S]cysteine (Amersham Pharmacia Biotech, Uppsala, Sweden). The labelled c ells were diluted in 10 volumes of ice-cold NaCl/P i , washed once and chased for different times at 37 °C. At each chase time, cell samples were washed twice in i ce-cold NaCl/P i and l ysed in buffer I [20 m M Tris (pH 7.5), 150 m M NaCl, 5 m M EDTA, 0.2% (w/v) BSA, 1% (v/v) Triton X-100, 1· complete protease inhibitor cocktail, 2 lgÆmL )1 calpain inhibitor (Roche Diagnostics GmbH, Mannheim, Germany)]. Immunoprecipitation and SDS/PAGE Lysates corresponding to 3 · 10 7 cells from suspensions containing 5 · 10 6 cellsÆmL )1 and 2 mL culture superna- tants corresponding to 10 7 cells were clarified by incubating them twice for 1 h at 4 °Cwith100lLofprotein G-Sepharose (Sigma, Saint L ouis, MO, USA). The s amples were then in cubated overnight at 4 °Cwith10lgofthe relevant mAbs and 100 lL of fresh protein G-Sepharose. After centrifugation, the bead pellets were washed once in buffer I, t wice in buff er II [ 0.5% (v/v) T riton X-100, 20 m M Tris (pH 7.5), 150 m M NaCl, 5 m M EDTA, 0.2% (w/v) BSA, 0.1% (w/v) S DS], three times in buffer I II [0.5% (v/v) Triton X-100, 20 m M Tris (pH 7 .5), 500 m M NaCl, 5 m M EDTA, 0.2% (w/v) BSA] and twice in buffer IV [50 m M Tris (pH 7.5)]. The beads were resuspended in Laemmli buffer containing 10 m M dithiothreitol and heated f or 5 m in at 95 °C. Prot eins wer e separated b y 7 .5–15% gradient SDS/PAGE and the gels were fixed and processed for autoradiography. Carbohydrate analyses After the last washing step in buffer IV, the immunopre- cipitates were dissociated from the beads in 30 lLof50m M Tris (pH 7.5) containing 1% (w/v) SDS and 5 m M dithio- threitol and boiled f or 5 min at 95 °C. Sodium citrate (pH 5.5) was added to a final concentration of 0.1 M .The samples were then incubated with endoglycosidase-H (5 mU per sample) for 6 h at 37 °C in t he presence of 1.7 lgÆmL )1 calpain i nhibitor and 1· protease inhibito r cocktail. In neuraminidase treatment, the beads were resuspended directly in 20 lL of 100 m M sodium acetate (pH 5) containing 1 m M CaCl 2 and 150 m M NaCl, before addition of 10 lL of n euraminidase (100 mUÆmL )1 ) (Roche) and incubation for 6 h at 37 °C. Western blotting Supernatants from CHO/GPV cell cultures w ere treated o r not with thrombin (5 UÆmL )1 ) (Sigma) for 5 min at 3 7 °C, before separation by 4.5%)15% SDS/PAGE and immu- noblotting. The blots were probed with the mAb V.5 followed by goat anti-(mouse IgG)–horseradish and p rotein bands were revealed by ECL (Amersham Biosciences, UK). Confocal microscopy CHO cells seeded at 20.10 4 cellsÆmL )1 in NaCl/P i were allowed to adhere to poly L -lysine-coated coverslips in f our- well plates for 2 h at 37 °C. After washing, the cells were fixed with 3% (w/v) paraformaldehyde and permeabilized 3672 C. Strassel et al.(Eur. J. Biochem. 271) Ó FEBS 2004 with 0.05% (w/v) saponin in blocking solution [NaCl/P i containing 0.2% (w/v) BSA and 1% (v/v) goat serum]. CHO/GPIb–V–IX cells were incubated for 45 min at room temperature with the mAb V.1 in blocking solution, washed and incubated with a GaM–Cy5 secondary antibody (Jackson Immunoresearch) for 30 min. These cells were then incubated for a further 30 min with a 1 : 400 dilution of a third antibody (ALMA.12 or ALMA.16) directly coupled to Alexa-488 (Molecular Probes, Eugene, OR, USA). CHO/GPV cells were incubated w ith V.1 coupled directly to Cy3 for 45 min at room temperature. The coverslips were washed in NaCl/P i ,rinsedinwaterand mounted in Mowiol 4-88 (Calbiochem Novabiochem) and the labelled cells were examined under a Zeiss laser scanning microscope (LSM 410 invert) equipped w ith a Planapo oil immersion lens. Results Transfection of GPV in the absence of GPIb–IX results in its inefficient cell surface exposure and release of a soluble form of GPV The CHO/GPV cell line transfected with full length GPV lacked surface expression of GPV as m easured by flow cytometry (Fig. 1A), despite selection of a strongly expressing clone by a series of amplifications in the presence of increasing concentr ations of metho trexate. Analysis of permeabilized cells by confocal microscopy revealed an intracellular pool of GPV with a granular appearance (Fig. 1B). Analysis of culture supernatants by GPV ELISA showed the presence of high levels o f soluble GPV [21], which was identified a s a single 82 kDa band by Western blotting with t he mAb V.5 (Fig. 1C). The identity of t his band was f urther established by its cleavage to a 69 kDa species following thrombin treatment. These properties of s oluble recombinant GPV are identical t o those reported for a calpain-derived fragment of platelet GPV [12]. Singly transfected GPV is retained intracellularly as an N-mannose rich 70 kDa form and is released into the cell supernatant as a sialylated 82 kDa polypeptide Pulse–chase and immunoprecipitation experiments were performed on CHO/GPV cell lysates and culture super- natants (Fig. 2A). A broad 70 kDa band corresponding to immature GPV w as present in the cells at early time points and throughout 180 min of chase, w hile shorter chase times revealed that this band was composed of three or four different m olecular mass f orms. Parallel a nalysis of the supernatants revealed a positive signal starting at 60 min of chase a nd having a molecular mass (82 kDa) consistent with a f ully mature form. This b and was not observed in the cell lysates a t a ny chase time a nd conversely the 70 kDa band was absent from the supernatants. Fig. 1. Distribution o f GPV in CHO/GPV cells in the absence of GPIb–IX. (A) G PV expression on the surface of platelets and CHO/G PV cells was analysed by flow cytometry using the mAb V.1 without p ermeabilization of the cells. A positive signal was observed on platelets but n ot on CHO/ GPV cells. (B) The subcellular distribution of GPV was examined in permeabilized CHO/GPV cells by confocal microscopy using Cy3-coupled V.1. An intracellular l abelling was obser ved w ith a granular appearance. (C) Soluble GPV in the supernatants of CHO/GP V cells was analysed by Western blotting with the mAb V.5. An 82 kDa band was detected in untreated samples, which was converted into a 69 kDa band by treatment with 5 UÆmL )1 thrombin. Ó FEBS 2004 Biosynthesis of platelet glycoprotein V (Eur. J. Biochem. 271) 3673 Post-translational s ugar modifications of the cellular and soluble forms of GPV were studied by treatment with endoglycosidase-H (Endo-H) (Fig. 3A) or neuraminidase (Fig. 3 B). The 70 kDa form was sensitive to Endo-H treatment, which reduced it to a narrow 5 0 kDa band, but was not cleaved by n euraminidase. T his i ndicated i ts enrichment in N-mannose sugars added before the medial Golgi compartment. On the other hand, the 82 kDa soluble molecule was resistant to Endo-H an d cleaved by neura- minidase to a 72 kDa form, indicating the p resence o f terminal sialic acid residues added in t he trans-Golgi. These experiments s how that the secreted 82 kDa f orm of GPV is fully processed during transit through the different Golgi compartments b ut is not retained in the p lasma membrane. In the presence of the GPIb–IX complex, GPV is fully processed and targeted mainly to the cell surface Contrary to the singly transfected subunit, GPV cotrans- fected with GPIb–IX was efficiently expressed at the cell surface t ogether with the other subunits of the G PIb–V–IX complex, as demonstrated by flow cytometry (Fig. 4A). Double-labelling confocal m icroscopy with antibodie s against GPV and GPIba or GPIX revealed colocalization of these s ubunits with G PV primarily at the cell m embrane. Hardly any labelling c ould be detected intracellularly, unlike in CHO/GPV cells (Fig. 4B). These results imply that GPIb–IX is required for stable surface expression o f GPV in transfected cells. Very comparable results were obtained when GPV was cotransfected with GPIb–IX i nto leukaemic human K562 cells. This cell line was chosen for biosynthetic studies in order to analyse the processing of GPV in a cell system more closely resembling platelets. Pulse–chase experiments performed on cell extracts alone showed immunoprecipitation at early times of the immature 70 kDa form of GPV, which progressively matured to a cell associated 82 kDa protein. C oncomitantly, GPIba progressed from an immature 85 kDa form to a mature 125 kDa mo le cule. GPIbb and GPIX displayed more modest mass increases of 1–2 kDa and gradually reached their mature sizes of 25 and 20 kDa, respectively, ove r 30 min of chase. This maturation time-course is very similar to that reported p reviously for C HO/GPIb–IX cells [18,22] (Fig. 4 C). Analysis of t he supernatants revealed no secreted forms of GPV (data not s hown), while the m ature GPV in K562/GPIb–V–IX c ells was sensitive to neuraminida se treatment but resistant to Endo-H (data not shown). Discussion Using stably transfected cell lines, we examined the biosyn- thesis of platelet GPV, a subunit of the GPIb–V–IX Fig. 2. Biosynthesis of GPV in CHO/GPV cells in the absence of GPIb–IX. CHO/GPV cells were pulse-labelled with [ 35 S]Cys and [ 35 S]Met a nd chased for various periods of time. A t different chase times, the c ulture supernatants were collected, the cells were lysed in Triton X-100 buffer and the cell lysates and s upernatants were ana- lysed by immunoprecipitationwiththemAbV.1.AtT 0 ,thecells contained an immature 70 k Da form of GPV which did not appear to progress to a higher molecular mass form during the 180 min of chase. On the o ther hand, an 80 kDa mature form progressively appeared and accumulated in the s upernatants at c hase times o f 60–120 min. Because a t T 0 only th e 7 0 kDa form is present it must be the p rede- cessor of the 80 kDa soluble form. The 70 kDa form th eref ore matures but probably becom es quickly se creted and do es not accum ulate in sufficient amounts to allow detection in t he cell lysate. Fig. 3. Sugar processing of cell associated and sec reted forms of GPV in CHO/GPV cells. Cells were pulse-labelled, c hased and immun opre- cipitated as described in Fig. 2 and th e immunoprecipitates from the cell lysates or supernatants were analysed for Endo-H (A) or neura- minidase (B) sensitivity. (A) The band at  70 kD a corresponding to immature cell associated GPV was converted into a 50 kDa band (the expected siz e of deglycosylated G PV) b y End o-H treatment. In con- trast, the 80 kDa secreted molecule was i nsensitive to Endo-H. (B) The cell associated 70 kDa form o f GPV was resistant to neuraminidase, whereas th e solu ble 80 kDa fo rm was reduced to 70 kD a t hrough loss of terminal s ialic acid res idues. 3674 C. Strassel et al.(Eur. J. Biochem. 271) Ó FEBS 2004 complex, and the influence on its pro cessing and expres- sion of the presence or absence of the other subunits of the complex. In both cases, GPV was processed from an immature mannose-rich intracellular 70 kDa form to a mature sialic acid-rich 82 kDa species. Mature GPV was expressed efficiently at the cell surface in the presence of GPIb–IX, whereas singly t ransfected GPV w as secreted as a soluble m olecule, presumably following enzymatic cleavage. A lack of surface expression of GPV after single-chain transfection was unexpected in view of the presence of a transmembrane region in the construct e ncoding the entire protein and from previous reports of its surface expression in human melanoma cells and mouse L-cells stably trans- fected with GPV alone [17]. Cell membrane expression of GPV has also been observed in CHO cells in transient expression experiments [14]. In our studies, using CHO cells or a human K562 leukaemic cell line, low levels of GPV were detected at the cell membrane 48–72 h aft er transfec- tion but we were never a ble to obtain a stable surface expression. Lack of expression in C HO/GPV cells is probably not related to differences in expression vectors compared to CHO/GPIb–V–IX, as both vectors contain the same SV40 promoter. This was also not due to i nefficient biosynthesis of GPV a s C HO/GPV cells were submitted to extensive gene amplification with methotrexate resulting in a 80-fold increase i n concentrations of soluble GPV. Ampli- fication of GPV expression probably explains part of the intracellular a ccumulation observed i n CHO/GPV cell lines in confocal microscopy and metabolic labelling experi- ments. Although differences with respect to a few studies could be related to the cell types used, i n general the single subunit is not (or only v ery weakly) retained at the cell surface. Consistent with this hypothesis, sign ificant levels of Fig. 4. Surface expression and intracellular processing of GPV in CHO and K562 cells c otransfected with GPV and GPIb-IX. (A) CHO cell s stably transfectedwithGPV,GPIba,GPIbßandGPIXwereanalysedforsurfaceexpression of GPV b y flow cytometry. A positive GPV signal was observed with approximately half the intensity of those of the other subunits. (B) CHO/GPIb–V–IX cells adherent to poly L -lysine were permeabilized, double- labelled with m Abs against GPV (pink) and GPIba (green) (left panel) or GPV ( p ink) and GPIX (green) (right panel) and analysed by confocal microscopy. Co-localization at th e c ell membrane (white) wa s o bserved under both co nditions. (C) K562/GPIb–V–IX cells w ere pulsed-labelled with 35 S and ch ased as described in Fig. 2. At d ifferent chase time s, the cells were lysed in 1% ( v/v) Triton in buffer I and the GPIb–IX complex was immunoprecipitated with ALMA.12 and GPV with V.1. At T 0 , the cells contained a n immature 70 kDa form of GPV which progressed within 60 min to an 80 kD a mature protein . GPIba progressivel y evolved from an early immature 85 kDa form to a mature 125 kDa molecule and this process, first detected at 15 min, was c ompleted within about 30 min. The mo lecular masses of GPIbß and G PIX increased slowly to reach 2 5 and 20 kDa, respectively. Ó FEBS 2004 Biosynthesis of platelet glycoprotein V (Eur. J. Biochem. 271) 3675 soluble GPV were found in the culture supernatants of GPV transfected melanoma cells [17]. A similar phenomenon has been reported for the GPIba subunit by Meyer et al.[23], who u sing methotrexate amplification in CHO cells found inefficient membrane insertion of GPIba transfected as a single-chain and secretion of a glycocalicin-like s oluble form. The exact m echanism l eading to release of GPV into t he culture medium is still unknown. GPV was not derived from membrane fragments or microvesicles as it was not found in a 100 000 g centrifugation pellet (data not shown). T he soluble form probably r esulted from e nzymatic cleavage of GPV above or near the point of membrane insertion. Th is would resemble the reported cleavage of GPV from the platelet surface by calpain an d possibly matrix metallopro- teinases, which releases a soluble 82 kDa fragment. How- ever, attempts to p revent GPV cleavage through incubation of CHO/GPV cells in the presence of a Ca 2+ chelator, impermeable or permeable forms of c alpain inhibitors (calpastatin, lactacystin), were unsuccessful and did not restore surface expression of the mature protein. The similar possibility that GPV is cleaved in platelet precursors in the absence of the other GPIb–V–IX subunits cannot be readily assessed for example in Bernard–Soulier patients. Such studies should be facilitated by the recent and future development o f mouse strains mutated in the GPIb–V–IX complex and the availability of efficient culture systems for megakaryocyte precursors [24,25]. The requirement for the other subunits of GPIb–V–IX for correct surface expression of GPV is illustrated in the present work and has also been documented in other cell types [17,26]. Studies of cells transfected with partial complexes indicated t hat GPIba was a key s ubunit for efficient membrane expression of GPV, a finding which remains t o b e c onfirmed i n p latelets. C onversely, addition of GPV to cells already expressing GPIb–IX did not increase surface expression of the complex [14,15], except in GPV transfected megakaryocytic human erythro-leukemia (HEL) cells [26]. These discrepancies could be due to different levels of GPIba in the various cell systems. The relevance o f t hese findings for platelet biosynthesis is unknown. I t is nevertheless clear that mice deficient in GPV express normal levels of GPIb–IX on the surface of platelets, suggesting t hat in t hese megakaryocytes t he ot her subunits are produced at sufficiently high l evels [6,13]. Metabolic labelling experiments showed similar process- ing of GPV in the p resence or absence of GPIb–IX and in both cases an immature cell associated 70 kDa form was detected at early chase times. In cells expressing only GPV, this form was still present at later chase times and was localized in a granular compartment. Its sensitivity to Endo- H i nd icated the p re sence of h ig h mannose s u gars which are typically added before the cis-Golgi. A 70 kDa band has been reported b y others in cells expressing GPV or GPIb– V–IX [14,22]. These biosynthetic s tudies in cell lysates did not reveal any further sugar ma turation, in agreement with our findings in CHO/GPV cells, w hile their f ailure to detect a mature form of GPV could h ave been due to rapid release of a s oluble protein as reported here. Surprisingly, t hese same cells displayed surface expression of GPV in flow cytometric experiments, suggesting membrane targeting of an incompletely processed form. Immunoprecipitation o f surface l abelled proteins would however, be required to f ully confirm this hypothesis. In the presence of GPIb–IX, t he cell associated immature 70 kDa protein progressed to a more mature s ialylated 82 kDa species with kinetics comparable to those of the full maturation of GPIbab and GPIX. This mole cule was able to reach a nd remai n at the cell membrane as d emonstrated by surface b iotinylation studies (data not shown). Both immature and mature GPV appeared as broad bands on SDS/PAGE gels, which is also a characteristic of platelet GPV and an indication of heterogeneity of the sugar content at the eight consensus N-glycosylation sites and putative O-glycosylation sites [11]. The appearance of the cell attached and soluble mature forms of GPV with comparable kinetics would suggest a normal progression of the latter t hrough the Golgi a pparatus followed by i ts rapid cleavage. SDS/PAGE and glycosidase analyses indicated that singly expressed GPV had similar properties to the complex a ssociated form in CHO cells and p latelets. Despite these similarities, the nature of the s ugars could differ in CHO and platelet GPV, as observed previously for the GPIba subunit. Availability of a recombinant form of soluble G PV with biochemical and functional properties [7] similar to those o f the native platelet species will be important in the perspec- tive of determination of its 3-D s tructure, the mapping of ligand domains such as the sites interacting with collagen and the mechanisms of cleavage by proteases. Acknowledgements We thank Juliette Mulvihill for r evie wing the English of th e manuscript, Martine Santer for her help in the development of the CHO cell lines and Sylve tte Chasserot for help in c onfocal microscopy studies at Plateforme d’Imagerie In vitro IFR des Ne urosciences (Strasbourg). Catherine Strassel is supported by a grant f or ARMESA (Association d’Recherche e t de De ´ veloppement en Me ´ decine et e n Sante ´ Publique) and from ARC (Association de Recherche contre le Cancer). References 1. Nurden,A.T.&Caen,J.P.(1975)Specificrolesforplateletsurface glycoproteins in platelet function. Na ture 255, 7 20–722. 2. Baumgartner, H.R., T schopp, T.B. & Weiss, H.J. (1978) Defective adhesion of platelets to subendothelium in von Willebrand’s disease and Bernard–Soulier syndrome. Thromb. Haemost. 39, 782–783. 3. Berndt, M.C., Shen, Y., Dopheide, S.M., G ardiner, E.E. & Andrews, R.K. (2001) The vascular biology of the g lycoprotein Ib–IX–V c omplex. Thromb. Haemost. 86, 178–188. 4. Modderman, P.W., A dmiraal, L.G., von Sonnenberg, A. & dem Borne, A.E. (1992) Glycoproteins V and Ib-IX form a non- covalent complex in t he platelet membrane. J. Biol. Chem. 267, 364–369. 5. Phillips, D.R. & Agin, P.P. (1977) Platelet plasma membrane glycoproteins. Evidence for the presence of nonequivalent dis- ulfide bonds using nonreduced-reduced two-dimensional gel elec- trophoresis. J. Biol. Chem. 252, 212 1–2126. 6. Ramakrishnan, V., Reeves, P.S., DeGuzman, F., Deshpande, U., Ministri-Madrid, K., DuBridge, R.B. & Phillips, D.R. (1999) Increased thrombin resp onsiveness i n p latelets from mice lacking glycoprotein V. Proc. Natl Acad. Sci. USA. 96, 13336– 13341. 3676 C. Strassel et al.(Eur. J. Biochem. 271) Ó FEBS 2004 7. Moog, S., Mangin, P., Lenain, N., Stras sel, C., Ravanat, C., Schuhler, S ., Freund, M., Santer, M., Kahn, M., Nieswandt, B ., Gachet, C ., Cazenave, J.P. & Lanz a, F. (2001 ) Platelet g lycopro- tein V binds to collagen a nd participates in platelet adhesion and aggregation. Blood 98, 1038–1046. 8. Calverley, D.C., Kavanagh, T.J. & Roth, G .J. (1998) Human signaling protein 14-3-3zeta interacts with platelet glycoprotein Ib subunits Ibalpha and I bbeta. Blood 91, 1295–1303. 9. Andrews,R.K.,Harris,S.J.,McNally,T.&Berndt,M.C.(1998) Binding of purified 14-3-3 zet a sign aling p ro tein to discrete amino acid sequen ces w ithin the cytoplasmic domain of the platelet membrane glycoprotein Ib–IX–V complex. Bioc hem istry 37, 638–647. 10. Andrews, R.K., Munday, A.D., Mitchell, C.A. & Berndt, M.C. (2001) Interaction of c almodulin with the cytoplasmic domain of the p latelet membrane glycoprote in Ib–IX–V com plex. Blood 98 , 681–687. 11. Lanza, F., Morales, M., de La Salle, C., Cazenave, J .P., Clemet- son, K.J., Shimomura, T. & Phillips, D.R. (1993) Cloning and characterization of the gene encoding the hu man platelet glyco- protein V. A member of the leucine-rich glycoprotein family cleaved d uring t h rombin-induc ed platelet ac tivation. J. Biol. Chem. 268, 20801–20807. 12. Wicki, A.N . & Clemetson, K .J. (1985) Structure and function of platelet membrane glyco proteins Ib a nd V. Eff ects of leukocyte elastase and other proteases on platelets response to von Will- ebrand factor and t hrombin. Eur. J. Biochem. 15 3, 1–11. 13. Kahn, M.L., Diacovo, T.G., Bainton, D.F., Lanza, F., Tre jo, J. & Coughlin, S.R. (1999) Glycoprotein V-deficient platelets have undiminished thrombin responsiveness and d o not e xhibit a Bernard–Soulier p henotype. Blood 94, 4112–4121. 14. Li,C.Q.,Dong,J.F.,Lanza,F.,Sanan,D.A.,Sae-Tung,G.& Lopez, J.A. (1995) E xpression of platelet g lycoprotein (GP) V i n heterologous cells and evidence for its association with GP Ib alphainformingaGPIb–IX–Vcomplexonthecellsurface. J. Biol. Chem. 270 , 16302–16307. 15. Cranmer, S.L., Ulsemer, P., Cooke, B.M., Salem, H.H., de la Salle, C., Lanza, F. & Jackson, S.P. (1999) Glycoprotein (GP) Ib-IX-transfected cells roll on a v on Willebrand factor matrix under flow. Importance of the GPIb/actin-binding protein (AB P- 280) interaction in maintaining adhesion under high shear. J. Biol. Chem. 274, 6097–6106. 16. Lopez, J.A., Andrews, R.K., Afshar-Khargh an, V. & Berndt, M.C. (1998) Bernard–Soulier s yndrome. Blood 91 , 4397–4418. 17. Meyer, S.C. & F ox, J.E. (1995) Interaction of platelet glycoprotein V with g lycoprotein I b-IX regulates expression of the glycopro- teins and binding of von Willebrand factor to g lycoprotein Ib-IX in transfected cells. J. Biol. C hem. 270, 1 4693–14699. 18. Ulsemer, P., Strassel, C., Baas, M.J., Salamero, J., Chasserot- Golaz, S., Cazenave, J.P., De La Salle, C. & Lanza, F. (2001) Biosynthesis and intracellular post-translational processing of normal and mutant platelet glycoprotein GPIb–IX. Biochem. J. 358, 295–303. 19. Perrault, C., Moog, S., Rubinstein, E., Santer, M., Baas, M.J., de la Sal le, C., Ravanat, C ., Dambach, J., F reund, M., Santoso, S., Cazenave, J .P. & Lanza, F. (2001) A novel monoclonal antibody against the extracellular domain of GPIbbeta modulates vWF mediated platelet adhesion. Thromb. Haemost. 86, 1 238–1248. 20. Lopez, J.A., Leung, B., Reynolds, C.C., Li, C.Q. & Fox, J.E. (1992) Efficient plasma membrane expression of a functional platelet glycoprotein Ib–IX complex requires the presence of its three subunits. J. Bio l. Chem. 267, 12851–12859. 21. Azorsa, D.O., Moog, S., Ravanat, C., Schuhler, S., Follea, G., Cazenave, J.P. & Lanza, F. (1999) Measurement of GPV released by activated platelets using a sensitive immunocapture ELISA – its use to f ollow platelet storage in transfusion. Thromb. H aemost. 81, 131–138. 22. Dong, J.F., Gao, S. & Lopez, J.A. (1998) Synthesis, assembly, and intracellular transport of the platelet glycoprotein Ib–IX–V complex . J. Biol. Chem. 273, 31449–31454. 23. Meyer, S., Kresbach, G., Haring, P., Schumpp-Vonach, B., Clemetson, K.J., Hadvary, P. & Steiner, B. (1993) E xpression and characterization of functionally ac tive fr agme nts of th e plate let glycoprotein (GP) Ib–IX complex i n mammalian c ells. Incorpor- ation of GP Ib alpha into the cell surface membrane. J. Biol. Chem. 268, 20555–20562. 24. Nakano, T., Kodama, H. & Honjo, T. (1994) Generation of lymphohematopoietic c ells from embryonic stem cells in culture. Science 265, 1098–1101. 25. Fujimoto, T .T., Kohata, S., Suzuki, H., Miyazaki, H. & Fujimura, K. (2003) Production of functional platelets by differentiated embryonic stem ( E S) cells in vitro . Blood 102, 4044–4051. 26. Calverley, D.C., Yagi, M., Stray, S.M. & Roth, G.J. (1995) Human platelet glycoprotein V: its role in enhancing expression of the glycoprotein Ib r ecep tor. Blood 86, 1 361–1367. Ó FEBS 2004 Biosynthesis of platelet glycoprotein V (Eur. J. Biochem. 271) 3677 . Biosynthesis of platelet glycoprotein V expressed as a single subunit or in association with GPIb-IX Catherine Strassel, Sylvie Moog, Marie-Jeanne Baas,. form of GPV was produced by an enzymatic cleavage wh ich w as not a ected by inhibitors of proteasome, calpain or metalloproteinases. When GPV was cotransfected