## Abstract

Abstract 2139

Poster Board II-116

The natural anticoagulant protein C (PC) circulates in blood at a concentration of about 60 nM. Inter-individual variations in the levels of PC are in part genetically determined, but which loci in the genome are involved is only partially known. In a recent study we identified a locus on chromosome 20 which was associated with high PC levels in a large pedigree from the GENES study (LOD score >5 at 55 cMorgan). Candidate genes related to the PC pathway under the LOD-1 region encoded FOXA2 (previously known as HNF3 beta, a nuclear factor regulating protein C gene transcription), thrombomodulin (THBD,which is key to activation of PC), and the endothelial protein C receptor (PROCR). Here we present data that pinpoint a SNP in PROCRas being responsible for the observed segregation of high PC levels. The pedigree has 218 members and was ascertained through a proband with a family history of venous thrombosis (VT). Classical genetic risk factors for thrombosis (i.e. PC-, PS-, antithrombin deficiency, factor V Leiden and prothrombin G20201A) were not present. Complete medical data, plasma measurements and DNA was available for 161 family members. The mean age was 47±15 (range 15-87) years. The mean PC plasma level was 116±25% (range 72-212). Four family members had experienced VT and 2 had had recurrence. These symptomatic members had normal to high PC levels (66, 82, 114 and 178%).Haplotypes (and genotypes) for PROCR were determined in the family members by TaqMan assay using tag SNPs (single nucleotide polymorphisms) and PROCR H3 was associated with the levels of PC in the family. Furthermore, the promoter, exons, and 3UTR of the 3 candidate genes were sequenced in 13 individuals, 9 with high and 4 with normal plasma PC levels. Critical SNPs that were encountered during sequencing were genotyped in all family members, namely FOXA2 rs1055080 (3UTR) and rs2277764 (promoter region). As those 2 SNPs were inherited together in the set of 13 patients and also in the LETS (data not shown), our further analysis used only rs1055080. Plasma soluble endothelial protein C receptor (sEPCR) and soluble thrombomodulin (sTM) levels were measured with an ELISA assay. PC and sEPCR and levels were compared between PROCR H3 and FOXA2 rs1055080 carriers and non carriers by Student's t-test. sTM was analyzed by Mann-Whitney test. Association between PC levels and sEPCR/sTM levels were evaluated using linear regression analysis. Afterwards associations were adjusted for the PROCR H3 and FOXA2 rs1055080 SNP separately to detect their possible confounding effect. DNA sequencing only yielded previously reported SNPs in FOXA2, THBD and PROCR. Only the above mentioned SNPs were associated with PC plasma levels. Linkage analysis for PC levels using the original markers (from Marshifield) and adding the new PROCR and FOXA2 SNPs did not change the LOD score. When the analysis was adjusted for the mentioned markers, the LOD score dropped below 2. sEPCR has a bimodal distribution; mean ± SD was 103±27 ng/ml for the first mode and 262±70 ng/ml for the second mode. Median (range) sTM was 1.2 ng/ml (0.1-4). Linkage analysis for sEPCR levels yielded a high LOD score (above 6) that was accentuated to above 8 when PROCR H3 was included as a marker. For sTM, the LOD score was low with every combination of markers. PC, sEPCR and sTM levels were compared between PROCR H3 carriers and non-carriers and both PC levels and sEPCR levels were influenced by this PROCR haplotype, but not sTM. In conclusion, chromosome 20 harbors a locus which influences PC levels and also the levels of sEPCR, but not the levels of sTM. A detailed analysis with SNPs in PROCR, THBD and FOXA2suggests that the so-called PROCR H3 is directly responsible for the increased PC and sEPCR levels in this family. PROCR H3 is known to represent a g.A6936G substitution leading to a p.Ser219Gly replacement in the transmembrane domain of EPCR. The Gly219 isoform is more sensitive to sheddases (such as the ADAM17 metalloprotease) and is associated with generation of truncated mRNA lacking the transmembrane domain. However, the exact mechanism by which EPCR and sEPCR levels influence the level of PC remains to be determined

Disclosures:

No relevant conflicts of interest to declare.

## Author notes

*

Asterisk with author names denotes non-ASH members.