Kathleen Freson, Center for Molecular and Vascular Biology, University of Leuven, Herestraat 49, B-3000 Leuven, Belgium. Tel.: +32 16346017; fax: +32 16345990. E-mail: email@example.com
Summary. Background: The physiological relevance of the collagen glycoprotein VI (GPVI) receptor was known prior to its recognition as a platelet membrane receptor as several patients lacking GPVI as a consequence of autoantibody inhibition presented with a mild bleeding diathesis. Remarkably, patients with a proven GPVI gene mutation have not yet been identified. Results: In the present study, we describe a patient with a lifelong history of bleeding problems, structurally normal platelets but a functional platelet defect. Platelet aggregations are normal except for an absent response to Horm collagen, convulxin and the collagen-related peptide (CRP). ATP dense granule secretion is normal with ADP but absent with Horm collagen. Thrombus formation on a collagen surface in flowing blood is reduced but more single platelets are attached. Remarkably, the platelet function analyzer-100 shows a shortened collagen/ADP closure time. Flow cytometry demonstrates an absent expression of GPVI whereas immunoblot analysis shows strongly reduced levels of GPVI. The patient is compound heterozygous for an out-of-frame 16-bp deletion and a missense mutation S175N in a highly conserved residue of the 2nd Ig-like GPVI domain. The parents without clinical bleeding problems are heterozygous carriers. The mother carries the S175N mutation and presents with a mild functional platelet defect. In vitro studies show a reduced membrane expression and convulxin binding with the mutated S175N compared with the wild-type (WT) GPVI receptor. Conclusions: This study presents the first patient with a proven genetic GPVI defect.
Glycoprotein VI (GPVI) is a 63-kDa transmembrane protein consisting of two immunoglobulin (Ig)-like domains in the extracellular region connected to a highly glycosylated linker, a transmembrane domain and a cytoplasmic tail [1–3]. GPVI is a member of the Ig-like receptors within the leukocyte Ig-like receptor complex (LRC) on human chromosome 19q13.4 . It is the major signaling receptor for collagen on platelets but is also activated by laminin, the collagen-related peptide (CRP) and snake toxins, such as convulxin and alborrhagin [5–8]. The surface expression of GPVI requires the concomitant expression of the γ-subunit of the FcR receptor (FcRγ) and their association is functionally relevant as collagen binding to GPVI results in platelet signaling via the immunoreceptor tyrosine-based activation motif (ITAM) located within the FcRγ subunits.
More than 20 years after the initial report of a patient with a GPVI defect , a total of 11 patients have been described in a recent review . These patients have been described with either an acquired deficiency as a result of the anti-GPVI autoantibodies or a congenital deficiency with normal to strongly reduced or even absent GPVI levels and a defective GPVI-related signaling. However, the possible cases described to date with a congenital GPVI defect have yet to be defined at the molecular level, which renders the interpretation of functional data from these individuals less definitive as a case of GPVI deficiency. GPVI deficiency in humans is usually described as having a mild bleeding disorder but for the acquired condition it can be severe in combination with immune deficiency and the presence of thrombocytopenia . We here describe the clinical and functional GPVI defect in the first patient with proven GP6 gene mutations.
Materials and methods
Platelet electron microscopy, aggregation and secretion
Morphological and functional platelet studies were performed as decribed . Platelets were stimulated with Horm collagen (Nycomed Arzenmittel, Munich, Germany), ADP (Sigma Chemical Co, St Louis, MO, USA), arachidonic acid (Sigma), U46619 (Sigma), convulxin (kind gift from Dr Kenneth Clemetson, Berne, Switzerland) or CRP-XL (kind gift from Dr Richard Farndale, Cambridge, UK). Blood samples from all individuals were obtained after informed consent.
Flow cytometric analysis
The following antibodies were used: FITC-conjugated anti-CD41a (HIP8), PE-conjugated CD62P (AC1.2), FITC-conjugated anti-CD42a (Beb1), FITC-conjugated CD49b and FITC-conjugated anti-CD61 (RUU-PL7F12) (BD Biosciences Pharmingen, Heidelberg, Germany). The monoclonal GPVI antibody HY101 (epitope in the extracellular GPVI domain) was kindly obtained from Dr Mark Kahn (Philadelphia, PA, USA) . We used Cell Quest software for two-color immunofluorescence acquisition on a FACSCalibur flow cytometer (BD Biosciences).
Flow perfusion studies
Platelet interaction with immobilized Horm collagen (200 μg mL−1) was studied under flow conditions producing a shear rate of 600 s−1 as described . Platelet adhesion was quantified with a light microscope (Leica DM RBE; Leica Microsystems, Wetzlar, Germany) equipped with a CCD camera. The coverage was analysed by the Java image processing program imagej 1.34g (National Institutes of Health, Bethesda, MD, USA).
Assessment of platelet function by whole blood platelet function analyzer (PFA100)
The PFA100 device (Dade-Behring, Marburg, Germany) was used to measure platelet function at high shear conditions on whole citrate blood. The method determines the time to occlusion of an aperture in a membrane coated with collagen and adenosine diphosphate (closure time CT/ADP) or epinephrine (closure time CT/EPI).
Platelet immunoblot analysis
The preparation of platelet and HEK293 extracts, quantification of the total protein concentration and the immunoblot analysis were done as previously described . Blots were revealed with a rabbit polyclonal GPVI antibody (H-139; Santa Cruz Biotechnology, Santa Cruz, CA, USA; epitope amino acid 201–339), a mouse monoclonal FcRγ antibody IV.3 (HB217, American Type Culture Collection, Rockville, MD, USA) or the mouse monoclonal XLαs antibody made in our laboratory . The secondary antibody was conjugated with HRP and staining was performed with the Western blotting ECL detection reagent (GE Healthcare Life Sciences, Uppsala, Sweden).
Molecular analysis of the FCER1G and GP6 genes
Total RNA was extracted from platelets from the patient and mother using TRIzol reagent (Invitrogen Corp., Carlsbad, CA, USA). RT-PCR generated GP6 and FCER1G fragments covering the complete coding sequence of both genes were cloned in the TOPO-TA vector (Invitrogen) and sequenced on an ABI310 (Perkin-Elmer, Norwalk, CT, USA) sequencer. PCR and sequencing of genomic DNA from leukocytes from patient, mother and father were performed with intronic primers for GP6 exon 3 and 4. All primers are listed in Table S1. PCR conditions are available upon request.
Recombinant GPVI expression in HEK293 cells
The complete coding sequence of the wild-type (WT) and S175N GP6 gene (Genbank accession number NM_016363 from nucleotide 29 to 1048) was amplified from platelet cDNA from the mother and each cloned in the pcDNA3.1 vector (Invitrogen). Primer sequences are listed in Table S1 and constructs were confirmed by sequencing. HEK293 cells were transfected with pcDNA/GP6 and pcDNA/S175N-GP6 using lipofectamin (Invitrogen). The expression of GPVI in transfected HEK293 cells was analyzed via flow cytometry and immunoblot analysis as described above. All experiments were repeated three times with similar results.
ELISA and binding assays using recombinant GST-tagged GPVI
The complete coding sequence of the WT and S175N GP6 gene was also cloned in the pGEX-4T2 vector (Promega, Madison, WI, USA) for the production of glutathione S-transferase (GST) fusion proteins. WT (GST-GPVI WT) and mutant (GST-GPVI Mutant) GST fusion proteins were produced and purified as described . The binding of GST-GPVI WT or mutant with convulxin was analyzed via ELISA. Microtitration wells were coated overnight with convulxin (2 nm) or CRP (0.5 μg mL−1) in phosphate-buffered saline (PBS), and blocked with 0.2% bovine serum albumin (BSA) for 2 h. GST-GPVI WT or mutant (in PBS, 0.1% BSA, 0.1% Tween 20) were added to the wells and incubated for 2 h. After washing with PBS, 0.1% BSA, 0.1% Tween 20, bound recombinant GPVI was detected using an anti-GST monoclonal antibody made in our laboratory  and an HRP-coupled goat anti-mouse antibody. In addition, the interaction of GST-GPVI WT or mutant with antibodies against GST or GPVI (H-139) was also analyzed. Microtitration wells were coated overnight with the GST-GPVI fusion proteins in PBS, and blocked with 0.2% BSA for 2 h. Anti-GST or anti-GPVI (in PBS, 0.1% BSA, 0.1% Tween 20) were added to the wells and incubated for 2 h. After washing with PBS, 0.1% BSA, 0.1% Tween 20, the bound primary antibody was detected using a HRP-coupled goat anti-mouse or anti-rabbit antibody.
We used graphpad instat 3.01 software (Graphpad Software Inc., San Diego, CA, USA). Continuous variables with small to mild skewness were summarized as mean ± SD and compared using the Student’s t-test for unpaired data. Two-sided P-values below 0.05 were considered significant.
Patient description and functional platelet studies
We here describe a 31-year-old female with ecchymoses, epistaxis, several posttraumatic and postsurgery bleeding complications since her childhood and menorraghia. Coagulation and von Willebrand disease (VWD) studies revealed no abnormalities. Blood counts were normal, including a normal platelet count (208 × 109 L−1) and mean platelet volume (8.5 fL). Electron microscopy showed platelets with a normal amount of alpha and dense granules. The parents of the proposita have no clinical bleeding problems. The father accepted DNA sampling but preferred not to participate in other studies.
The patient’s platelets had normal aggregation responses to all standard agonists but showed an absent response to Horm collagen, convulxin and CRP-XL (Fig. 1A and Fig. S1). Platelets from her mother have a mildly reduced response to Horm collagen, convulxin and CRP-XL. The release of ATP from the dense granules in platelets from the patient was absent for collagen 5 μg mL−1 but was normal for 10 μm ADP (1.57 μm ATP; normal range 1–3.5 μm).
Perfusion of whole blood on Horm collagen at intermediate shear stress (600 s−1) showed a strongly reduced aggregate formation for the patient but also a significantly increased adhesion of single platelets (Fig. 1B). Consistent with these findings, it was previously shown that inhibition of GPVI via the monoclonal antibody 9O12 reduced the platelet aggregate formation under flow conditions but still permits the initial platelet adhesion via integrin α2ß1 with the adherence of numerous single platelets . The inhibition of platelet activation via anti-αIIbß3 antibodies or RGD-containing peptides also resulted in increased platelet adhesion to collagen, as platelets that are not incorporated in the aggregates are now available to adhere . Remarkably, the PFA100 closure time for collagen/ADP was shortened for the patient; 62 and 64 s with a reference value of CT/ADP = 71–118 s. The closure time for collagen/epinephrine was normal; 113 s with a reference value of CT/EPI = 94–193 s. Her mother had normal PFA100 closure times. This suggests that the adherence of more single platelets in the presence of additional ADP for platelet activation via ADP receptors, result in an enhanced closure.
Genetic analysis of GP6 and FCER1G
The sequence of the coding region of the genes for FcRγ (FCER1G) and GPVI (GP6) genes were studied using platelet RNA from the patient. No abnormalities were found in FCER1G but several genetic variants were detected in GP6. TOPO-TA subcloning of the GP6 RT-PCR fragments showed the presence of one allele with a 16-bp deletion in exon 3 and five previously described non-synonymous GP6 single nucleotide polymorphisms (nsSNPs) (rs1613662, rs1654416, rs2304167, rs1654413, rs1671152) [17,18] and a second allele with a missense mutation p.Ser175Asn (p.S175N). The out-of-frame 16-bp deletion generated a hypothetical short GPVI protein of only 56 amino acids as a result of an early stop codon. All these GP6 genetic variants were confirmed at gDNA from leukocytes of the proposita (Fig. S2A). The nsSNPs and 16-bp deletion were inherited from her father whereas her mother is a heterozygous carrier of the p.S175N mutation. The p.S175N mutation is not present in the recently described high resolution SNP map of the GP6 gene . Codon 175 is a highly conserved residue in the 2nd Ig-like GPVI domain as shown via alignment with GPVI of other species and even other LRC proteins (Fig. S2B) . As the mother has no other GPVI genetic abnormalities, the p.S175N mutation probably accounts for her mild GPVI platelet defect but a heterozygous GP6 defect is not sufficient to cause a clinical phenotype.
Expression of platelet membrane receptors
Flow cytometric analysis of the different platelet glycoprotein receptors revealed a normal expression of integrin αIIbß3 (CD61/CD41), integrin α2ß1 (CD49/CD29) and GPIX (CD42) (Table S2). In contrast, the membrane expression of GPVI was almost absent in the patient but normal in her mother (Fig. 2A and Table S2). P-selectin (CD62P) expression was absent after stimulation by 2 μg mL−1 CRP-XL in the patient and normal in the mother (Table S2). Immunoblot analysis of platelet lysates revealed a reduced but not absent GPVI expression in the patient and a normal expression in the mother compared with three unrelated normal controls (Fig. 2B,C). The expression of the FcRγ receptor in the patient’s platelets was comparable with the normal controls. The 50% reduction in GPVI expression for the patient is compatible with the out-of-frame 16-bp deletion generating a short GPVI protein of 56 amino acids. In contrast to the flow cytometric detection of membrane localized GPVI, the GPVI receptor with the p.S175N mutation is present in de platelet lysates from the patient shown using immunoblot analysis.
In vitro characterization of the GPVI S175N mutation
To further understand the functional consequences of the p.S175N GPVI mutation on GPVI expression, membrane localization and convulxin binding, the WT and mutant (S175N) GPVI receptor were cloned in different plasmids. First, HEK293 cells were transiently transfected with an empty vector, GPVI WT, GPVI mutant or an equal amount of both (GPVI WT/mutant). Flow cytometric analysis of these receptors revealed an absent membrane expression of the GPVI mutant (Fig. 3A) though its total cellular expression was comparable with the GPVI WT (Fig. 4B,C). The combined expression of both receptors, showed a normal GPVI membrane expression. This could explain why the patient’s mother, being heterozygous for GPVI WT/mutant, has a normal GPVI expression as determined by flow cytometry. To study GPVI receptor shedding, GPVI expression was studied using immunoblot analysis in the supernatants of the HEK293 cells and in patient and control plasma samples but no differences were observed (data not shown). It is not yet clear whether S175N GPVI is retained in the cell or not recognized by the monoclonal anti-GPVI antibody (HY101). Flow cytometric analysis using another anti-GPVI antibody (H-139) with an epitope in the intracellular GPVI domain could not detect GPVI in normal platelets.
In addition, the GPVI WT and mutant receptor were expressed as GST fusion proteins. Binding studies with these fusion proteins show a comparable interaction with the anti-GST and anti-GPVI (H-139) antibodies (Fig. 4A) but the binding with convulxin and CRP-XL were significantly reduced for the GPVI mutant receptor compared with the WT (Fig. 4B). The interaction between the GST fusion proteins and the anti-GPVI antibody (HY101) showed no differences in binding affinity but the interaction was very weak, possibly because of hindrance by the large N-terminal GST tag near the extracellular epitope domain of this GPVI antibody.
Although different patients have been described with an expected genetic GPVI defect , the associated molecular defects have not yet been found. Recently, it was even suggested that the GPVI defect was inherited as an autosomal dominant disease . Here we describe an autosomal recessive mode of inheritance for a GPVI defect in a patient with a moderate to mild bleeding phenotype and for the first time a genetically proven compound heterozygous mutation. The patient carries a paternally inherited out-of-frame 16-bp deletion and a maternally inherited missense mutation p.S175N in a highly conserved GPVI residue. This deletion results in a short truncated GPVI protein of only 56 out of the 339 amino acids. The recombinant p.S175N GPVI receptor shows a strongly reduced membrane expression and decreased interaction with convulxin and CRP-XL. This mutation would disrupt a characteristic structural feature of the KIR-family proteins, the tryptophan-containing ß-bulge, WSXXS, in which both serines have been observed to participate in hydrogen bounds necessary for structural integrity of the Ig-fold .
The mother of the patient has no clinical bleeding phenotype, a mildly reduced response towards Horm collagen, CRP and convulxin and is a heterozygous carrier of the p.S175N GPVI mutation. Flow cytometric analysis and immunoblot analysis show a normal GPVI receptor expression in her platelets. Expression studies in HEK293 cells using an equal combination of WT and mutant GPVI constructs resulted in a normal expression of GPVI using flow cytometry with the HY101 anti-GPVI antibody. If the p.S175N mutation would disrupt GPVI recognition by HY101, a reduced and not a normal GPVI membrane expression would be expected using flow cytometry in her platelets and the dual transfected HEK293 cells. These results suggest that the p.S175N mutation affects the membrane expression of GPVI. More recent studies have shown the importance of GPVI dimerization  or oligomerizaton  for proper GPVI signaling. As a result of the lack of sufficient platelet extracts, we could not study these characteristics in the patient or her mother but the p.S175N mutation could affect the GPVI association and therefore result in a very mild phenotype in the mother.
The role for GPVI in platelet adhesion to a collagen surface under flow conditions has been studied in great detail . The patient’s platelets did not respond to collagen in aggregation studies but an enhanced adhesion of single platelets to collagen was found under intermediate shear conditions (600 s−1). Platelets from the patient express normal integrin α2ß1 levels, which probably accounts for the adherence of the single platelets. Others have already shown that integrin α2ß1 regulates platelet adhesion to collagen when GPVI is blocked by the monoclonal antibody 9O12 that interacts with the first Ig domain or via inhibition of the two P2Y receptors . Under these conditions, a reduced platelet aggregate formation was observed but also an adherence of numerous single platelets. Blockade of the P2Y receptors resulted in an increased surface area coverage compared with the non-treated control platelets. It is not yet clear why our patient has a shortened PFA100 closure time for collagen/ADP. The patient’s platelets have a normal aggregation response to ADP. Further studies are needed to elucidate this duality. A similar duality seems to be present in the literature as subjects homozygous for the minor GPVI haplotype (PEALN) show a 3-fold reduced response to CRP-XL in platelet aggregation with a reduced thrombus formation on a collagen type III surface  and still in three studies this minor haplotype was shown as a risk factor for myocardial infarction instead of a protective factor [17,25,26]. The GP6 SNP (re1613662) was also recently identified as a gene variant significantly associated with deep vein thrombosis . This duality could also contribute to the fact that complete loss of GPVI activity results in a relatively mild clinical bleeding phenotype.
In conclusion, in the present study we described the first genetic proven GPVI defect in a patient with mild clinical bleeding problems.
K.F. participated in the experimental design and analysis, manuscript preparation; C.T. and C.W. performed experiments and analyzed the data; C.H. and C.VG. studied the patient and participated in manuscript preparation.
K.F. holds a postdoctoral research mandate and C.V.G. is holder of a clinical-fundamental research mandate of the Fund for Scientific Research-Flanders (F.W.O.-Vlaanderen, Belgium). Supported by the ‘Excellentie financiering KULeuven’ (EF/05/013), by research grants G.0453.05 and G.0124.02 from the FWO-Vlaanderen (Belgium) and by GOA/2004/09 from the Research Council of the University of Leuven (Onderzoeksraad K.U. Leuven‘ Belgium).
Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.