No influence of the VAMP8 rs1010 single nucleotide polymorphism on platelet functions in vitro.

Dear Editor:

Platelet activation plays an important role in the pathogenesis of arterial thrombosis [1], as illustrated by the preventive effect of antiplatelet agents on cardiovascular events such as acute coronary syndrome and myocardial infarction (MI). Platelet granule secretion is a key step in sustained platelet aggregation and clot formation. The main secretory organelles in platelets are dense granules and a-granules. Dense granules contain small molecules such as ATP, serotonin, calcium and ADP which, once secreted, contribute to the recruitment of other platelets. The α-granules contain large proteins such as fibrinogen, von Willebrand factor and growth factors, and their membranes express a specific receptor, P-selectin, that is used as a marker of degranulation. During platelet secretion, heteromeric complexes are formed between membrane proteins called soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) present on vesicles or granules (v-SNAREs) or on the target membrane (t-SNAREs) [2]. VAMP3 and VAMP8 (v-SNAREs) bind to syntaxin 4 (t-SNAREs) in platelets [3]. Recently, three independent studies comprising about 1820 cases and 1100 controls linked a VAMP8 single nucleotide polymorphism (SNP: rs1010 A/G) to the risk of MI [4]: the rs1010 G VAMP8allele was associated with early-onset MI (P= 0.025, [OR] 1.75; CI 1.17–2.62).

Here we sought a functional effect of the rs1010 SNP on platelet functions ex vivo, independently of known cardiovascular risk factors and hormone status, in a homogeneous population of 100 healthy male Caucasian volunteers aged 18–35 years. They were recruited as described elsewhere [5] and were extensively phenotyped in terms of platelet function.

Venous blood was collected from each volunteer on two occasions one week apart, between 8:00 and 10:00 a.m. after an overnight fast, in 0.105 mol/l sodium citrate (BD Vacutainer Becton Dickinson, Le Pont de Claix, France) with a 19-gauge needle. Platelet-rich plasma (PRP) was obtained by centrifugation at 150 ×g for 10 min. at room temperature. Autologous platelet-poor plasma was used to adjust the platelet count of PRP to 250 × 109/l. Aggregation studies were performed within 2 hrs after blood collection, with the following agonists: arachidonic acid 1 mmol/l (Helena Biosciences Europe, Saint-Leu la Forêt, France), ADP 2 μmol/l (Sigma Aldrich, Saint Quentin Fallavier, France), the thromboxane receptor agonist U46619 1 μmol/l (Calbiochem, Merck Eurolab, Fontenay-sous-Bois, France), and Horm collagen 1 μ/ml (Nycomed Pharma, Paris, France). Maximal aggregation was recorded during a 5-min. period, and the results for the paired samples were averaged. We also determined the concentration of the thrombin receptor activating peptide (SFLLRN, Diagnostica Stago, Asnières, France), causing double-wave aggregation in the presence of 100 μmol/l amastatin. This concentration reflects the minimal SFLLRN concentration necessary to provoke granule secretion and sustained platelet aggregation. To evaluate maximal platelet secretion capacity, collagen (100 μg/ml), U46619 (10 μmol/l) or SFLLRN (100 μmol/l) was added to PRP in the presence of epti-fibatide (Integrilin®, 4 (μg/ml) to prevent aggregation. The expression of P-selectin (CD62P or GMP140) on the platelet surface was then quantified by flow cytometry as described elsewhere [5]. Genomic DNA was isolated from peripheral blood mononuclear cells by using the Qiamp Maxi Kit® (Qiagen, Courtaboeuf, France) according to the manufacturer's instructions. Rs1010 SNP status was determined by amplifying the 3’UTR region of the VAMP8 gene, using 5′-CCGGGGGACCAAGGTACCTTCTGGGGCATACAcC-3′ and 5′-CTGGGTCACTCACTCTGCC-3′ as upstream and downstream primers, respectively. The upstream primer is a mutagenic primer that bears a sequence change at the penultimate position (lower case), introducing an Nco I restriction site into the amplicons when the A allele is amplified. Twenty microlitres of amplification mixture was subjected to Nco I(20 units) digestion at 37ｰC overnight, and the restriction products were resolved on 2% agarose gel. Digestion of the A allele yielded two products, of 318 and 32 bp, while the amplicon of the G allele remained undigested (350 bp).

The respective frequencies of the A allele (wild-type) and the G allele (mutated risk allele) were 0.63 and 0.37, in line with published data [4]. The genotype was in Hardy-Weinberg equilibrium. As shown in Table 1, the rs1010 variant was not associated with maximal platelet aggregation, whatever the agonist, even with a low ADP concentration (used to observe interindividual variations), nor with the lag time for collagen aggregation. We also analysed the aggregation surface, that is likely to better represent the reversibility of aggregation, but no difference was found according to the genotype (data not shown). Finally, we determined the minimal SFLLRN concentration necessary to provoke platelet secretion (and thus a biphasic aggregation profile) and found it to be similar in the two genotype groups (median values 9, 10 and 9 μmol/l for AA, AG and GG subjects, respectively).

1

Ex vivo platelet functions according to the rs1010 genotype

	Platelet maximal aggregation (%)						Number of P-selectin copies on platelets
Genotype	Arachidonic acid, 1 m mol/l	ADP 2 μmol/l	U 46619 1 μmol/l	Collagen 1 μg/ml	Collagen lag time (sec)	Basal level	Collagen 100 μg/ml	U46619 10 μmol/l	SFLLRN 100 μmol/l
AA (n= 39)	75.9 (69.0–79.1)	40.4 (26.9–67.4)	70.1 (48.8–75.9)	72.4 (67.6–76.7)	1.0 (0.8–1.2)	376 (287–552)	5607 (4534–7213)	10396(7887–11229)	11356 (10211–12551)
AG (n= 49)	77.4 (72.9–79.8)	37.0 (27.7–67.9)	73.3 (67.0–79.4)	73.7 (68.9–80–7)	1.0 (0.8–1.2)	506 (288–621)	5443 (3209–7540)	9858 (8212–11708)	11309 (9579–12776)
GG (n= 12)	75.4 (71.9–78.0)	56.9 (24.5–73.1)	71.4 (65.9–77.9)	72.3 (59.4–75.2)	1.0 (0.9–1.2)	510 (300–577)	6579 (3913–7217)	9921 (5803–11969)	10268 (9259–12765)
P	0.38	0.83	0.30	0.45	0.90	0.34	0.74	0.9	0.62

Left: maximal platelet aggregation (%) and collagen aggregation lag time.

Right: P-selectin (CD62P) expression level on platelets at baseline and after activation, as determined by quantitative flow cytometry.

Owing to the skewed distribution of the continuous variables, trends across genotype groups were tested with the non parametric Kruskal–Wallis test. Data are expressed as medians and interquartile ranges.

P-selectin expression levels at the platelet surface upon stimulation with various agents are shown in the table. As expected, maximal P-selectin expression was achieved with the thrombin-receptor-activating peptide SFLLRN. We preliminary controlled that, in our experimental conditions, i.e. in the absence of calcium (citrated platelet-rich-plasma), such activation did not induce microparticule generation, that could have major influence on the analysis (microparticules were found below 1%, data not shown). As in the aggregation tests, no difference in P-selectin expression was observed between the genotype groups, whatever the agonist.

Thus, despite the use of weak (ADP) and strong (SFLLRN) platelet agonists and an accurate method to quantify α-granule secretion, the VAMP8 rs1010 SNP had no observable impact on platelet functions ex vivo. Based on the allelic frequency of the VAMP8 SNP and on the mean and standard deviation of the CD62P level after SFLLRN stimulation, this study had 80% power to detect a difference of 10% or more between carriers and non carriers of the mutated allele. A smaller effect of this SNP cannot therefore be ruled out.

This study has two limitations: first, we focused on a-granule secretion: we did not directly measure dense granule secretion, and the platelets were not subjected to shear stress. Second, we only tested platelets from healthy men with normal platelet function. It is conceivable that the rs1010 SNP might affect functions of circulating activated platelets in patients with cardiovascular disease. Interaction with atherosclerotic plaque activates platelets and induces P-selectin expression in vivo, creating a reactive surface for leucocyte recruitment. It would be particular interest to measure the number of platelet-leucocyte aggregates [6, 7] in a population of genotyped patients with cardiovascular disease.

Despite these limitations, this study has the merit of being the first to test a putative association between a VAMP8 gene variant and platelet functions ex vivo. Our negative results are in line with those of a recent study that failed to confirm the link between the VAMP8 SNP and coronary heart disease in a population of 2145 patients with familial hypercholesterolaemia [8].