SEARCH

SEARCH BY CITATION

Keywords:

  • EDEM2;
  • endothelial protein C receptor;
  • genetic determinants;
  • plasma levels;
  • PROCR;
  • protein C

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Summary. Background: Genetic determinants of plasma levels of protein C (PC) are poorly understood. Recently, we identified a locus on chromosome 20 determining high PC levels in a large Dutch pedigree with unexplained thrombophilia. Candidate genes in the LOD-1 support interval included FOXA2, THBD and PROCR. Objectives: To examine these candidate genes and their influence on plasma levels of PC. Patients/Methods: Exons, promoter and 3′UTR of the candidate genes were sequenced in 12 family members with normal to high PC levels. Four haplotypes of PROCR, two SNPs in the neighboring gene EDEM2 and critical SNPs encountered during resequencing were genotyped in the family and in a large group of healthy individuals (the Leiden Thrombophilia Study (LETS) controls). Soluble endothelial protein C receptor (sEPCR) and soluble thrombomodulin (sTM) plasma levels were measured in the family. Results:PROCR haplotype 3 (H3) and FOXA2 rs1055080 were associated with PC levels in the family but only PROCR H3 was also associated with plasma levels in the healthy individuals. Carriers of both variants had higher PC levels than carriers of only PROCR H3 in the family but not in healthy individuals, suggesting that a second determinant is present. EDEM2 SNPs were associated with PC levels, but their effect was small. PC and sEPCR levels were associated in both studies. sTM was not associated with variations of THBD or PC levels. Conclusions: Chromosome 20 harbors genetic determinants of PC and sEPCR levels and the analysis of candidate genes suggests that the PROCR locus is responsible.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Protein C (PC) is a vitamin K-dependent plasma glycoprotein that circulates at a concentration of ∼40–80 nm and plays a major role in the control of the coagulation cascade. Upon activation by the thrombin-thrombomodulin complex, activated PC (APC) inactivates FVa and FVIIIa, consequently reducing thrombin formation [1]. Binding of PC to the endothelial protein C receptor (EPCR) increases the activation rate of PC 20-fold [2]. Individuals with abnormalities in components of the PC pathway (such as PC or protein S deficiency) have an increased risk of venous thrombosis [3].

Determinants of variation in PC levels are poorly known but, given the high heritability of about 50%, genetic determinants are likely to be important [4,5]. The PC gene (PROC) is located on chromosome (chr) 2 and around 6% of the variability in PC levels has been attributed to polymorphisms in the promoter region of the gene [6,7]. The Genetic Analysis of Idiopathic Thrombosis (GAIT) study estimated that additive effects of genes outside of the PROC structural locus cause approximately half of the phenotypic variation in PC levels [4]. The authors identified a major quantitative trait locus (QTL) for PC levels on chr 16, where NQO1, a gene encoding a quinone reductase involved in the metabolism of vitamin K, is located. Subsequently they found that variations in this gene were associated with PC levels [5].

Variations in the PROCR gene (chr 20, encoding the EPCR) were associated with a moderate increase in levels of PC in the Cardiovascular Health Study (CHS) [8]. While this work was in progress, a genome-wide association scan (GWAS) in the Atherosclerosis Risk in Communities (ARIC) study claimed four loci associated with PC levels, that included PROC, PROCR and three other genes: EDEM2 (chr 20), GCKR (chr 2) and BAZ1B (chr 7) [9].

The GENES study was designed to search for novel hereditary risk factors for venous thrombosis in families with unexplained thrombophilia [10]. Previously, a genome-wide linkage analysis was performed. In one particular family, a QTL influencing PC levels was found on chr 20 with a log-odds (LOD) score of 4.8 at 51 cM [11]. In the 1-LOD support interval (38–64 cM), three candidate genes encoding components potentially influencing PC levels are present, namely forkhead box A2 (FOXA2, previously known as hepatic nuclear factor 3β), thrombomodulin (THBD), and the endothelial protein C receptor (PROCR). While this work was in progress, EDEM2 (located on chr 20 within 50 kb from PROCR) emerged as a fourth candidate gene [9].

FOXA2 encodes a transcription factor for a large number of genes, including PROC. Two binding sites for FOXA2 in the promoter region of PROC have been described and mutations in this region were associated with type I PC deficiency [12].

THBD encodes thrombomodulin (TM), a transmembrane protein that (in complex with thrombin) enhances thrombin-mediated PC activation by more than 1000-fold [1].

PROCR encodes EPCR. A soluble form of EPCR (sEPCR), lacking the transmembrane and cytoplasmic domain, is present in human plasma. sEPCR and EPCR bind PC and APC with similar affinity [13] and the binding of sEPCR to APC inhibits its anticoagulant activity [14]. Four haplotypes of PROCR are present in the European population [15]. Haplotype 3 (H3) is tagged by a missense variation that leads to the Ser219Gly variation (rs867186). The Gly219 variant is associated with increased levels of sEPCR, which can be explained by an increased sensitivity of the protein to sheddases such as metalloprotease ADAM17 [16] and by the expression of an alternatively spliced mRNA that lacks the sequence encoding the transmembrane domain [17]. Levels of sEPCR were associated with PC levels in the Leiden Thrombophilia Study (LETS) [15].

In this study, we investigated whether genetic variations in THBD, FOXA2, PROCR and EDEM2 influence the plasma level of PC.

Patients and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Subjects

GENES  The pedigree analyzed in the present study is one of the 22 families originally included in GENES, a study of Dutch families with unexplained thrombophilia [10,11]. Probands with personal and/or family history of venous thrombosis (defined as at least one first-degree or two second-degree relatives with venous thrombosis) but with none of the known inherited thrombophilic defects (i.e. PC-, protein S-, antithrombin deficiency, factor (F) V Leiden or prothrombin G20210A variation) were recruited together with their extended pedigree, including spouses. A standardized history was taken and for most individuals, plasma and DNA samples were obtained for coagulation tests and genotyping [11]. This analysis focuses on the largest pedigree in GENES, consisting of 185 individuals distributed over five generations. In this family, four individuals had a history of thrombosis, of whom two had experienced more than one event. Detailed information about these individuals is given in Table 1.

Table 1.   Characteristics of family members with venous thrombosis
SexFirst episode of VTSecond episode of VTPCPROCRFOXA2
TypeAge (years)Risk factorTypeAge (years)Risk factor(%)Genotypers1055080
  1. PC, protein C; VT, venous thrombosis; PE, pulmonary embolism; DVT, deep vein thrombosis. PC levels are presented as percentage of reference pooled plasma. FOXA2 rs1055080: NM_021784.4:c.*50C>T or NM_153675.2:c.*50C>T.

FemalePE32PregnancyH1H2CC
MaleDVT41Trauma82H2H2CC
MaleDVT4TraumaDVT28Idiopathic114H2H3CT
FemaleDVT40IdiopathicDVT44Idiopathic178H3H3CT

Leiden Thrombophilia Study (LETS)  For replication of the results obtained in the family, relevant DNA variations were genotyped in healthy (i.e. non-thrombotic) individuals, who were the controls in a population-based case–control study for venous thrombosis (LETS). The design of this study has been described before [18]. Briefly, 474 consecutive patients with a first episode of deep vein thrombosis and 474 sex- and age-matched healthy controls were included. All patients were younger than 70 years and had no overt malignancy. Controls were healthy acquaintances and partners brought by the patients, according to pre-established criteria. The mean age for patients and controls was 45 years (range 15–69 for patients and 15–72 for controls). Standardized questionnaire, DNA and plasma samples were obtained from all the participants.

GENES and LETS were approved by the Central Committee on Research Involving Human Subjects (CCMO) and all participants have provided informed consent.

Resequencing of candidate genes

Three genes in the linkage region (LOD-1 support interval) on chr 20 were selected for resequencing.

FOXA2 has two known mRNA splicing variants. The first variant (NM_021784) contains two exons and covers 2422 bases. Alternative splicing at the 5′-end results in a second mRNA variant (NM_153675) in which translation initiation starts six amino acids later than in variant 1. Variant 2 also has an additional (untranslated) 5′-exon, leading to a total mRNA length of 2410 bases. THBD is transcribed from an intron-less gene as a 4109 bases long mRNA (NM_000361). PROCR has four exons with transcription length of 1449 bases (NM_006404).

To investigate genetic variations associated with PC levels in the family, we selected 12 family members based on their PC levels: (i) three with normal levels (72, 75 and 82%); (ii) three with intermediate high levels (114, 116 and 128%); and (iii) six with the highest levels (range: 166 – 212%). Three out of four patients with thrombosis were included in this panel (Table 2). PC plasma level was not available for the fourth individual.

Table 2.   Polymorphisms detected in 12 family members with normal, intermediate and high protein C (PC) levels. Only non-synonymous variations are shown in the table
SamplePC (%)PROCRFOXA2THBD
Haplotypers1055080rs2277764rs1962rs3176123rs1042580rs1042579
  1. *Individuals with thrombosis; [1] homozygous for the common allele, [2] heterozygous and [3] homozygous for the minor allele. PC levels are presented as percentages of the reference pooled plasma.

3604972H2H4111131
2964075H1H2112212
29448*82H2H2111121
29494*114H2H3221111
29599116H1H3112212
29495128H2H3221111
29600166H2H3221212
29680166H1H3222111
29552169H3H3221212
29678170H1H3222111
29529*178H3H3221212
29687212H2H3222111

Exons and their flanking regions, 5′ and 3′ UTRs, and 1000 bp upstream to the initiation codon were resequenced. For FOXA2, the DNA sequence covering both splicing variants and 1000 bp upstream to exon 1 of both isoforms was analyzed. Primers and PCR conditions are available on request.

After amplification, the PCR product was sequenced using an ABI Prism® 3730 DNA Analyzer (Applied Biosystems, Carlsbad, CA, USA). The results were analyzed using vector NTI® software version 10 (Invitrogen, Paisley, UK). Potentially interesting variations found by sequencing were investigated in all family members and in the healthy individuals using single nucleotide polymorphism (SNP) genotyping assays.

SNP genotyping assays

All SNPs were determined using TaqMan SNP genotyping assays (Applied Biosystems). PCR reactions were performed in 384-well plates using the GeneAmp PCR System 9700 (Applied Biosystems) and fluorescent endpoints were read on a 7900 HT Real-Time PCR System (Applied Biosystems).

Common haplotypes of PROCR were tested in all family members with DNA available. Three haplotype tagging SNPs were chosen: rs2069952 (H1, predesigned assay), rs867186 (H3, predesigned assay) and rs2069951 (H4, custom assay). The presence of the minor allele determines the mentioned haplotype and the presence of three common genotypes determines H2 [15].

Two common variations in FOXA2 (rs1055080 and rs2277764) that were identified during sequencing analysis in individuals with high levels of PC were investigated in all family members and/or in healthy individuals using a custom TaqMan genotyping SNP assay.

Two variations in EDEM2 (rs6120849 and rs3746429) that were associated with PC levels in the ARIC study [9] were investigated in the family and in the healthy individuals by a predesigned TaqMan genotyping SNP assay.

Plasma assays

Protein C.  In both the family and healthy individuals, blood was collected in tubes containing 0.106 m trisodium citrate. Plasma was prepared by centrifugation at 2000 × g for 10 min at room temperature and stored at −70 °C [18]. PC levels were determined using a chromogenic assay (Chromogenix, Mölndal, Sweden). Levels were expressed as percentage of the level in a reference pooled plasma. Measurements in GENES and LETS were performed in different laboratories, several years apart, using different reference pooled plasmas.

Plasma soluble EPCR (sEPCR) levels were determined in the family using the Asserachrom sEPCR ELISA kit (Diagnostica Stago, Asnières, France) according to the manufacturer’s instructions. Samples were tested in duplicate and plasmas were diluted 1/26 prior to the assay.

Plasma soluble TM (sTM) levels were measured in the family using the CD141 ELISA kit (Diaclone, Besançon, France) according to the manufacturer’s instructions. Samples were tested in triplicate in non-diluted plasmas.

Linkage analysis

To assess the influence of the investigated genetic variations and plasma measurements of sEPCR and sTM on the QTL for PC levels on chr 20, linkage analyses were performed using SOLAR [11]. The effects of the genetic variations were assessed by adding them to the marker set or by adding them as covariate to the linkage model (conditional analyses). Effects of sEPCR and sTM levels were assessed by adding them as covariates to the linkage model. In addition, LOD scores for sEPCR and sTM levels were determined on chr 20. Following Lander and Kruglyak [19], and correcting for two phenotypes, we used thresholds of 3.6 for genome-wide significance and 2.2 for suggestive linkage. In addition, linkage disequilibrium (LD, represented as D’ and Cramer’s V) between relevant (multiallelic) SNPs was calculated using the gap package in R [20].

Statistical analysis

Mean and 95% confidence interval were used to compare continuous variables with normal distribution (i.e. PC and sEPCR). One healthy individual from LETS that was using vitamin K antagonists at the time of venepuncture was excluded from the analysis. None of the family members was using vitamin K antagonists at the time of venepuncture. Median and range were used to describe sTM because of the skewed distribution in the family. Linear regression analysis was used to analyze the correlation between sEPCR and PC levels. All calculations were performed using PASW Statistics 17.0 (IBM Corporation, Somers, NY, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Genetic analysis of candidate genes PROCR, FOXA2 and THBD

Resequencing of candidate genes yielded five variations in FOXA2, four variations in THBD and five variations in PROCR. All these variations have been previously reported, and only the non-synonymous variations are summarized in Table 2. For PROCR, DNA variations were in concordance with the known haplotypes and only the haplotypes are shown. One variation in PROCR (rs867186, a tagging SNP for H3) and two variations in FOXA2 (rs1055080, in the 3′-UTR, and rs2277764, in the 5′-UTR) were associated with higher levels of PC. Other SNPs found in FOXA2 were rs1800847, rs1203910 and rs1212275, all leading to synonymous amino acid substitutions (not shown). Relevant variations found in THBD were: rs1042579 (p.Val473Ala), rs1042580, rs3176123 and rs1962 (last three in the 3′-UTR) but none of these was associated with PC plasma levels. This excludes THBD as a likely determinant of PC levels.

The co-inheritance of the rare variations of PROCR (rs817186) and FOXA2 (rs1055080 and rs2277764) in individuals with high PC plasma levels suggests that PROCR and FOXA2 SNPs are inherited as a single haplotype in these individuals. In an attempt to distinguish which gene is actually responsible, we genotyped all members of the family for FOXA2 rs1055080 and for the tagging SNPs of common PROCR haplotypes (Table 3). Genotyping of PROCR showed 52 heterozygotes and two homozygotes for the rs817186 minor allele, whereas for FOXA2 rs1055080, 34 individuals were heterozygotes and no homozygote for the minor allele was present. Except for two individuals, all carriers of the FOXA2 rs1055080 minor allele were also carriers of PROCR H3, again suggesting co-inheritance in the family (D’, 0.913; Cramer’s V, 0.67).

Table 3.   Protein C (PC) and sEPCR plasma levels for PROCR H3 and FOXA2 rs1055080 and EDEM2 rs6120849 and rs3746429 carriers and non-carriers in the family and in the healthy individuals from the Leiden Thrombophilia Study (LETS)
GenotypeFamily (n)LETS (n)Protein CSoluble EPCR
Family (n)Mean (95% CI)LETS (n)Mean (95% CI)Family (n)Mean (95% CI)LETS (n)Mean (95% CI)
  1. n, number of individuals per group; PC levels are presented as percentages of the reference pooled plasma and sEPCR as ng mL−1. *Only one or two measurements were available and therefore, mean and 95% CI were not calculated. Genotyping for FOXA2 rs1055080 (NM_021784.4:c.*50C>T or NM_153675.2:c.*50C>T), EDEM2 rs6120849 (NM_018217.2:c.259–106G>A) and EDEM2 rs3746429 (NM_018217.2:c.1366G>A) was performed in 465 healthy individuals from LETS with DNA available. Genotyping for EDEM2 variations has failed in two individuals from the family. PC measurement was not available for 26 individuals from the family and sEPCR measurement was not available for 16 individuals from the family and for one control from LETS.

PROCR H3
 HxHx10636085106 (102–110)36099 (97–100)96103 (97–108)36094 (91–96)
 H3Hx5210047131 (123–138)100113 (110–117)47261 (240–281)99258 (248–269)
 H3H32102169; 178*10127 (109–144)1337*10439 (399–478)
 Total160470134470144469
FOXA2
 CC126428104109 (105–113)428103 (101–104)114131 (118–144)427138 (129–146)
 CT343530139 (128–150)35100 (93–106)30249 (216–282)35117 (96–139)
 TT2292;123*254;176*
 Total160465134465144464
PROCR H3 + FOXA2
 HxHx + CC10432683106 (102–110)32699 (97–100)94103 (98–109)32694 (92–96)
 HxHx + CT/TT229287;117*2997 (90–104)266;81*2990 (81–98)
 H3Hx/H3H3 + CC2210221120 (111–129)102115 (111–119)20263 (235–291)101279 (265–294)
 H3Hx/H3H3 + CT/TT32828142 (131–153)8111 (101–121)28262 (232–292)8217 (168–265)
 Total160465134465144464
EDEM2 rs6120849
 GG9026370125 (119–132)263105 (103–108)77175 (153–198)263152 (140–165)
 GA5817252105 (100–110)17299 (97–102)56133 (115–152)171116 (108–125)
 AA103010102 (92–112)3096 (89–103)9118 (82–155)30102 (92–111)
 Total158465132465142464
EDEM2 rs3746429
 GG10631086121 (115–127)310104 (102–106)93173 (153–193)310145 (135–156)
 GA4713841106 (100–112)13899 (96–102)44123 (107–139)137117 (107–127)
 AA5175105 (90–120)1793 (83–104)5109 (60–157)17109 (95–122)
 Total158465132465142464

To answer the question of whether PROCR H3 or FOXA2 rs1055080 is responsible for PC level variation, we investigated these SNPs in LETS controls. FOXA2 rs1055080 was determined in 465 healthy individuals, out of whom 37 carried the minor allele, two in a homozygous state (Table 3). Only eight carriers of the minor FOXA2 rs1055080 allele also carried PROCR H3 (D’, 0.18), which suggests that, in this population-based study, PROCR and FOXA2 are not inherited together, reinforcing the idea that the co-inheritance of the rare variations is particular to this family.

FOXA2 rs2277764 was also determined in LETS controls but because of its tight linkage with FOXA2 rs1055080 (D’, 0.94), this variation was not analyzed further.

PROCR H3 is associated with levels of PC in the family and in healthy individuals

Table 3 shows the mean plasma levels of PC and 95% confidence intervals (CIs) for the different genotype groups. Levels of PC cannot be compared directly between the family and healthy individuals because different pooled plasmas were used as a reference. In the healthy individuals, levels of PC were systematically lower than the levels in the family. In the family, mean PC level was higher in PROCR H3 carriers (mean, 131%; 95% CI, 123–138) than in non-carriers (mean, 106%; 95% CI, 102–110). In two individuals from the family who were homozygous for H3, PC levels were 169% and 178%. In healthy individuals, mean PC level was higher in H3 heterozygotes (mean, 113%; 95% CI, 110–117) than in non-carriers of H3 (mean, 99%; 95% CI, 97–100). Mean PC level in homozygotes for H3 was 127% (95% CI, 109–144), which is somewhat higher than the levels in heterozygote carriers, but the confidence intervals overlap.

Carriers of the FOXA2 rs1055080 minor allele in the family had increased levels of PC (mean, 139%; 95% CI, 128–150) in comparison with non-carriers (mean, 109%; 95% CI, 105–113). This was not seen for heterozygous carriers and non-carriers in the healthy individuals (mean, 100%; 95% CI, 93–106 and mean, 103%; 95% CI, 101–104, respectively). Two individuals were homozygous for the minor allele and PC levels were 92 and 123%. This indicates that the minor allele of FOXA2 rs1055080 is not associated with plasma level of PC in the population.

We further examined whether family members carrying both rare variants in PROCR and FOXA2 had increased levels of PC in comparison to carriers of only one rare variant (Table 3). Mean level of PC was higher in carriers of both PROCR H3 and FOXA2 rs1055080 rare variants (mean, 142%; 95% CI, 131–153) than in carriers of PROCR H3 only (mean, 120%; 95% CI, 111–129). This indicates that, in this family, a second gene variation (other than PROCR H3) may determine PC levels. Because in the healthy individuals no effect of FOXA2 rs1055080 on PC levels was found, this second gene variation is probably not FOXA2 rs1055080.

To verify whether the hypothetical second gene is EDEM2, we analyzed PC levels in carriers and non-carriers of EDEM2 rs6120849 and rs3746429 variations. PC levels were lower in carriers of one or two minor alleles of both variations in the family and in the healthy individuals in comparison to non-carriers, but for rs3746429 the confidence intervals overlap (Table 3).

PROCR H3 is associated with sEPCR levels in the family and in the healthy individuals

sEPCR plasma levels were higher in family members carrying PROCR H3 (mean, 261 ng mL−1; 95% CI, 240–281) in comparison to non-carriers (mean, 103 ng mL−1; 95% CI, 97–108) and the same is true for the healthy individuals from LETS [15] (Table 3). Carriers of the FOXA2 rs1055080 minor allele, however, did show increased sEPCR in the family but not in healthy individuals (Table 3). For EDEM2 rs6120849 and rs3746429, carriers of one or two copies of the minor allele had lower sEPCR levels than carriers of the common allele both in the family and the healthy individuals (Table 3).

sEPCR and PC levels were correlated in both the family (r2 = 0.18) and healthy individuals (r2 = 0.15).

sTM concentration is not different between PROCR H3 carriers and non-carriers

Median sTM plasma level was 1.2 ng mL−1 (range, 0.1–4.0) in the family. Median levels were not different between PROCR H3 carriers (median, 1.3 ng mL−1; range, 0.3–4.0) and non-carriers (median,1.2 ng mL−1; range, 0.1–4.0). sTM was not measured in LETS.

Linkage analysis

Finally, we (re)examined linkage between PC, sEPCR and sTM levels and genetic markers on chr 20, now including PROCR H3 and FOXA2 rs1055080. For PC levels, the addition of new genetic markers did not change the LOD score (Fig. 1A). When the analysis was performed conditional on PROCR H3 and FOXA2, the LOD score for PC went down to <2.0, suggesting that most of the variation of PC levels can be attributed to these genetic markers (Fig. 1B). The fact that the FOXA2 rs1055080 variant was not associated with PC level variation in healthy individuals, however, suggests that PROCR H3 is the functional variant and that the drop in the LOD score observed in the linkage analysis is due to the nearly complete LD between PROCR and FOXA2 in the family (D’, 0.913; Cramer’s V, 0.67).

image

Figure 1.  Log-odds (LOD) scores for protein C levels using additional markers (A) or additional covariates (B), and for sEPCR (C) and sTM (D) levels using additional markers. The horizontal line indicates the genome-wide significance threshold.

Download figure to PowerPoint

For sEPCR, the linkage analysis performed with the initial genetic markers yielded a LOD score of 6.2. Adding FOXA2 rs1055080 to the analysis increased the LOD score to 7.7 and addition of PROCR H3 increased it further to 9.3, reinforcing the idea that this haplotype is largely responsible for sEPCR plasma level variation (Fig. 1C). For sTM, the LOD score remained <1.0 for any model (Fig. 1D).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

In this study, we provide evidence for the association of a common haplotype (H3) of PROCR with high plasma levels of PC. Resequencing of PROCR did not provide proof of additional determinants of plasma PC levels. Furthermore, we excluded that two other genes involved in the PC anticoagulant pathway (i.e. FOXA2 and THBD, both located on chr 20 close to the PROCR gene) determined the plasma level of PC, both at the level of common variations as well as at the level of rare sequence variations. Finally, variations in THBD did not influence the levels of sTM (data not shown) nor were these levels associated with the level of PC. This contrasted with the clear relationship between levels of PC and sEPCR.

Our results regarding the role of chr 20 in determining PC levels are in agreement with previous reports [8,9]. Most notably, a recent study by Tang et al. [9] that was published while this work was in progress, came to very similar findings using a radically different approach. In that study, two neighboring genes (i.e. PROCR and EDEM2) were found to be determinants of variations in PC levels. In the family, carriers of PROCR H3 who also carried the minor allele of FOXA2 rs1055080 had higher PC levels than carriers of H3 alone. The fact that this effect was absent in the healthy individuals from LETS indicates that a second determinant is present between the FOXA2 and PROCR genes, which could well be EDEM2.

When we genotyped the family and the healthy individuals from LETS for EDEM2 rs6120849 and rs3746429 and analyzed the association with PC plasma levels, PC levels were lower in carriers of one or two minor alleles in both the family and healthy controls, which is in accordance with the data from Tang et al. [9]. In addition, we observed that sEPCR levels were also lower in carriers of the minor alleles of EDEM2, which suggests that, also in this case, PC levels are influenced through EPCR. This is supported by the fact that we could not see any relationship between the minor alleles and levels of other coagulation proteins (FII, FV, FVII, FVIII, FIX, FX or FXI; data not shown).

The precise mechanisms underlying the association of PROCR and PC levels are not known, and there might be an important role for sEPCR. It is known that PC (and APC) has a comparable affinity (Kd≈30 nm) for membrane-bound EPCR and sEPCR [13]. Therefore, complex formation between PC and sEPCR in plasma is to be expected and thus, high levels of sEPCR might drive high levels of PC. The problem with this explanation is that, in general, PC levels are much higher than sEPCR levels, even in the presence of PROCR H3. PC levels in carriers of H3 in the family are estimated to range from 63 to 163 nm, whereas sEPCR levels in the same individuals are estimated to range from 4 to 18 nm. Based on these estimations it is difficult to simply explain the relationship between PC and sEPCR levels. Perhaps, higher sEPCR levels are associated with low levels of EPCR on the endothelial membrane, thus leading to a redistribution of PC between the membrane-bound compartment and a soluble compartment. There is indeed evidence that the H3 haplotype-associated Gly219 involves increased EPCR shedding from the endothelium [16], but evidence that this leads to lower density on the endothelial membrane is not available. In preliminary studies we have analyzed blood-originated endothelial cells (BOECs) from carriers and non-carriers of PROCR H3 (three individuals of each) by flow cytometry, but the data suggest that the expression of EPCR on the membrane is not different between the groups (unpublished data). It has also been hypothesized that the local concentration of sEPCR at the endothelial surface is higher than the concentrations measured in the plasma, possibly exceeding the Kd of PC interaction [21], but this also does not readily explain the increased plasma PC levels. Future research might solve this problem.

There are claims that H3 not only influences PC levels but also the level of other coagulation proteins (e.g. FVII) [22,23]. In the present GENES study or in the healthy individuals from LETS, this association was not confirmed, neither was H3 associated with levels of FII, FV, FVIII, FIX or FXI (data not shown).

Because low levels of PC, as in individuals with inherited PC deficiency, are a risk factor for venous thrombosis, it is tempting to assume that high levels of PC are protective, but there is no evidence that the latter is indeed the case. Thus, it is also reasonable to assume that the H3 haplotype, from the perspective of PC levels, would protect against venous thrombosis. This does not seem to be the case: in the family, two of the H3 carriers who had high levels of PC experienced recurrent venous thrombosis, but this family study obviously does not have sufficient power to determine the relationship between H3 and venous thrombosis. Population-based case–control studies have not been conclusive. Some authors report increased risk of venous thrombosis in H3 carriers [21,24,25], and others claim no association [15,26,27]. The risk of arterial thrombosis in H3 carriers is also not clear, varying from a protective effect to increased risk according to the population studied [8,28,29]. It seems fair to conclude though, that the markedly elevated levels of PC associated with H3 do not protect against thrombotic disease.

In conclusion, our data provide new evidence for the association of PROCR H3 and sEPCR with plasma levels of PC and suggest that FOXA2 and THBD, two other genes on chr 20, are not involved. Further studies are necessary to elucidate the mechanisms underlying this association.

Addendum

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

M. C. Pintao and S. Roshani performed the genetic and laboratory analysis, analyzed the data and wrote the manuscript. These authors equally contributed to the paper. M. C. H. de Visser analyzed the genetic data and critically reviewed the paper. C. Tieken performed part of the laboratory analysis during his internship as a graduate student. M. W. T. Tanck performed the linkage analysis and critically reviewed the paper. I. M. Wichers was responsible for the collection of patient samples and critically reviewed the paper. J. C. M. Meijers was responsible for plasma coagulation tests in the GENES and critically reviewed the paper. F. R. Rosendaal is responsible for the Leiden Thrombophilia Study (LETS) and critically reviewed the paper. S. Middeldorp designed and coordinated the GENES study and critically reviewed the paper. P. H. Reitsma coordinated these investigations, discussed and analyzed the data, and contributed to writing of the manuscript.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

This study was supported by the Netherlands Heart Foundation grant numbers 2005B248 and 2006B160. The LETS was supported by the Netherland Heart Foundation grant number 89063. The Mammalian Genotyping Service of the Center of Medical Genetics, Marshfield, WI, USA, is greatly acknowledged for genotyping our study population. We also would like to acknowledge the help of T. Bovill in establishing our relationship with Marshfield. S. Middeldorp is Clinical Established Investigator of the Netherlands Heart Foundation (2008T056). M.C.H. de Visser is a Postdoctoral Fellow of the Netherlands Heart Foundation (2005T055). We are thankful to C. Koch and K. Los for their field work collecting samples and interviewing families and to the Department of Experimental Vascular Medicine of the Academic Medical Center for their excellent technical support. We are also grateful to I. van der Linden for her technical support with sEPCR and sTM plasma measurements.

Disclosure of Conflict of Interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

The authors state that they have no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  • 1
    Dahlbäck B, Villoutreix BO. The anticoagulant protein C pathway. FEBS Lett 2005; 579: 33106.
  • 2
    Stearns-Kurosawa DJ, Kurosawa S, Mollica JS, Ferrell GL, Esmon CT. The endothelial cell protein C receptor augments protein C activation by the thrombin-thrombomodulin complex. Proc Natl Acad Sci U S A 1996; 93: 102126.
  • 3
    Rosendaal FR, Reitsma PH. Genetics of venous thrombosis. J Thromb Haemost 2009; 7(Suppl. 1): 3014.
  • 4
    Souto JC, Almasy L, Borrell M, Gari M, Martinez E, Mateo J, Stone WH, Blangero J, Fontcuberta J. Genetic determinants of hemostasis phenotypes in Spanish families. Circulation 2000; 101: 154651.
  • 5
    Buil A, Soria JM, Souto JC, Almasy L, Lathrop M, Blangero J, Fontcuberta J. Protein C levels are regulated by a quantitative trait locus on chromosome 16: results from the Genetic Analysis of Idiopathic Thrombophilia (GAIT) Project. Arterioscler Thromb Vasc Biol 2004; 24: 13215.
  • 6
    Spek CA, Koster T, Rosendaal FR, Bertina RM, Reitsma PH. Genotypic variation in the promoter region of the protein C gene is associated with plasma protein C levels and thrombotic risk. Arterioscler Thromb Vasc Biol 1995; 15: 2148.
  • 7
    Aiach M, Nicaud V, Alhenc-Gelas M, Gandrille S, Arnaud E, Amiral J, Guize L, Fiessinger JN, Emmerich J. Complex association of protein C gene promoter polymorphism with circulating protein C levels and thrombotic risk. Arterioscler Thromb Vasc Biol 1999; 19: 15736.
  • 8
    Reiner AP, Carty CL, Jenny NS, Nievergelt C, Cushman M, Stearns-Kurosawa DJ, Kurosawa S, Kuller LH, Lange LA. PROC, PROCR and PROS1 polymorphisms, plasma anticoagulant phenotypes, and risk of cardiovascular disease and mortality in older adults: the Cardiovascular Health Study. J Thromb Haemost 2008; 6: 162532.
  • 9
    Tang W, Basu S, Kong X, Pankow JS, Aleksic N, Tan A, Cushman M, Boerwinkle E, Folsom AR. Genome-wide association study identifies novel loci for plasma levels of protein C: the ARIC study. Blood 2010; 116: 50326.
  • 10
    Wichers IM, Tanck MW, Meijers JC, Lisman T, Reitsma PH, Rosendaal FR, Büller HR, Middeldorp S. Assessment of coagulation and fibrinolysis in families with unexplained thrombophilia. Thromb Haemost 2009; 101: 46570.
  • 11
    Tanck MW, Wichers IM, Meijers JC, Büller HR, Reitsma PH, Middeldorp S. Quantitative trait locus for protein C in a family with thrombophilia. Thromb Haemost 2011; 105: 199201.
  • 12
    Spek CA, Greengard JS, Griffin JH, Bertina RM, Reitsma PH. Two mutations in the promoter region of the human protein C gene both cause type I protein C deficiency by disruption of two HNF-3 binding sites. J Biol Chem 1995; 270: 2421621.
  • 13
    Kurosawa S, Stearns-Kurosawa DJ, Hidari N, Esmon CT. Identification of functional endothelial protein C receptor in human plasma. J Clin Invest 1997; 100: 4118.
  • 14
    Liaw PC, Neuenschwander PF, Smirnov MD, Esmon CT. Mechanisms by which soluble endothelial cell protein C receptor modulates protein C and activated protein C function. J Biol Chem 2000; 275: 544752.
  • 15
    Uitte de Willige S, van Marion V, Rosendaal FR, Vos HL, de Visser MCH, Bertina RM. Haplotypes of the EPCR gene, plasma sEPCR levels and the risk of deep venous thrombosis. J Thromb Haemost 2004; 2: 130510.
  • 16
    Qu D, Wang Y, Song Y, Esmon NL, Esmon CT. The Ser219-->Gly dimorphism of the endothelial protein C receptor contributes to the higher soluble protein levels observed in individuals with the A3 haplotype. J Thromb Haemost 2006; 4: 22935.
  • 17
    Saposnik B, Lesteven E, Lokajczyk A, Esmon CT, Aiach M, Gandrille S. Alternative mRNA is favored by the A3 haplotype of the EPCR gene PROCR and generates a novel soluble form of EPCR in plasma. Blood 2008; 111: 344251.
  • 18
    Koster T, Rosendaal FR, Briët E, van der Meer FJ, Colly LP, Trienekens PH, Poort SR, Reitsma PH, Vandenbroucke JP. Protein C deficiency in a controlled series of unselected outpatients: an infrequent but clear risk factor for venous thrombosis (Leiden Thrombophilia Study). Blood 1995; 85: 275661.
  • 19
    Lander E, Kruglyak L. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat Genet 1995; 11: 2417.
  • 20
    Zhao JH. 2LD, GENECOUNTING and HAP: Computer programs for linkage disequilibrium analysis. Bioinformatics 2004; 20: 13256.
  • 21
    Saposnik B, Reny JL, Gaussem P, Emmerich J, Aiach M, Gandrille S. A haplotype of the EPCR gene is associated with increased plasma levels of sEPCR and is a candidate risk factor for thrombosis. Blood 2004; 103: 13118.
  • 22
    Ireland HA, Cooper JA, Drenos F, Acharya J, Mitchell JP, Bauer KA, Morrissey JH, Esnouf MP, Humphries SE. FVII, FVIIa, and downstream markers of extrinsic pathway activation differ by EPCR Ser219Gly variant in healthy men. Arterioscler Thromb Vasc Biol 2009; 29: 196874.
  • 23
    Smith NL, Chen MH, Dehghan A, Strachan DP, Basu S, Soranzo N, Hayward C, Rudan I, Sabater-Lleal M, Bis JC, de Maat MP, Rumley A, Kong X, Yang Q, Williams FM, Vitart V, Campbell H, Malarstig A, Wiggins KL, van Duijn CM et al. Novel associations of multiple genetic loci with plasma levels of factor VII, factor VIII, and von Willebrand factor: the CHARGE (Cohorts for Heart and Aging Research in Genome Epidemiology) Consortium. Circulation 2010; 121: 138292.
  • 24
    Navarro S, Medina P, Mira Y, Estellés A, Villa P, Ferrando F, Vayá A, Bertina RM, España F. Haplotypes of the EPCR gene, prothrombin levels, and the risk of venous thrombosis in carriers of the prothrombin G20210A mutation. Haematologica 2008; 93: 88591.
  • 25
    Galanaud JP, Cochery-Nouvellon E, Alonso S, Chauleur C, Mercier E, Lissalde-Lavigne G, Fabbro-Peray P, Reny JL, Mares P, Dauzat M, Quere I, Gris JC. Paternal endothelial protein C receptor 219Gly variant as a mild and limited risk factor for deep vein thrombosis during pregnancy. J Thromb Haemost 2010; 8: 70713.
  • 26
    Medina P, Navarro S, Estellés A, Vayá A, Woodhams B, Mira Y, Villa P, Migaud-Fressart M, Ferrando F, Aznar J, Bertina RM, España F. Contribution of polymorphisms in the endothelial protein C receptor gene to soluble endothelial protein C receptor and circulating activated protein C levels, and thrombotic risk. Thromb Haemost 2004; 91: 90511.
  • 27
    Yamagishi K, Cushman M, Heckbert SR, Tsai MY, Folsom AR. Lack of association of soluble endothelial protein C receptor and PROCR 6936A/G polymorphism with the risk of venous thromboembolism in a prospective study. Br J Haematol 2009; 145: 2216.
  • 28
    Ireland H, Konstantoulas CJ, Cooper JA, Hawe E, Humphries SE, Mather H, Goodall AH, Hogwood J, Juhan-Vague I, Yudkin JS, di Minno G, Margaglione M, Hamsten A, Miller GJ, Bauer KA, Kim YT, Stearns-Kurosawa DJ, Kurosawa S. EPCR Ser219Gly: elevated sEPCR, prothrombin F1+2, risk for coronary heart disease, and increased sEPCR shedding in vitro. Atherosclerosis 2005; 183: 28392.
  • 29
    Medina P, Navarro S, Corral J, Zorio E, Roldán V, Estellés A, Santamaría A, Marín F, Rueda J, Bertina RM, España F. Endothelial protein C receptor polymorphisms and risk of myocardial infarction. Haematologica 2008; 93: 135863.