Evidence on Chromosome 11 for a Quantitative Trait Locus Influencing Levels of Activated Protein C-Protein C Inhibitor Complex.

Earlier we showed high heritability for the variation in levels of activated protein C-protein C inhibitor complex in a large kindred with type I protein C deficiency (Vossen et al. 2004). To identify genomic regions, which might harbour a gene with a mutation that accounts for a difference in the levels of activated protein C-protein C inhibitor complex, we performed a variance component linkage analysis. Family members were genotyped for 375 autosomal markers at the Marshfield Medical Research Foundation with an average marker spacing of 9.4 cM (range 0–18 cM) and an average marker heterozygosity of 75% (range 42–89%). Levels of activated protein C-protein C inhibitor complex were measured using commercial paired antibody sets from Affinity Biologicals Inc. (Ancaster, Ontario, Canada). We estimated the probability of Identity By Descent (IBD) using the multipoint IBD method in Simwalk2 (Sobel & Lange 1996), in which the proportion of alleles shared identical by descent at marker loci is used to estimate IBD sharing at arbitrary points along the chromosome for each relative pair. Then, variance component linkage analysis was performed using SOLAR (Almasy & Blangero 1998) to test whether a proportion of the genetic variance in the levels of activated protein C-protein C inhibitor complex could be attributed to specific genomic locations. The levels of activated protein C-protein C inhibitor were log-transformed to reduce skewness (from 3.0 to 0.2) and kurtosis (from 11.6 to 0.4) and were assumed to distribute as a multivariate normal density with correlation =h2K + c2H + q2B + e2I, where matrix K contains the kinship coefficients, H contains 1 for pairs from the same household and 0 otherwise, B contains the IBD probabilities, and I represents the identity matrix. The parameters for heritability (h2), household effect (c2), and heritability contributed to a specific genomic location (q2) as well as the effects of the covariates were estimated simultaneously using maximum likelihood analysis. Lod scores were computed as the log10 likelihood for q2 estimated to q2=0. A lod score of 3.3 was used as cut-off point for statistical evidence for significant linkage (as suggested by Lander and Kruglyak 1995), and a lod score above 1.9 but below 3.3 for suggestive linkage evidence. We found suggestive linkage evidence for levels of activated protein C in complex with protein C inhibitor in 121 tested family members (without a history of venous thrombosis) for a region on chromosome 11 (marker D11S969, 146 cM) with a LOD-score of 2.6, with age added to the model. As no obvious candidate genes were available under the peak, and because the linkage region is too wide to be certain of linkage, we are currently finemapping the peak by adding microsatellite markers to the genome scan to narrow the region on chromosome 11.