Genome Wide Scan of Complete Blood Count (CBC) Measures Suggests Strong Linkage of Red Blood Cell (RBC) Count to Chromosome 4q25.

Previous reports suggest that peripheral blood counts are strongly influenced by environmental and genetic determinants; however few of the genetic factors that regulate these quantitative traits have been discovered. We analyzed CBC data from 395 samples collected from an 854-member Amish pedigree with von Willebrand disease. 71 individuals of the pedigree are heterozygous for a missense mutation at position 4120, represented by a single base substitution (C>T) that predicts an arginine to cysteine change at position 1374 (R1374C) in the A1 domain of the mature von Willebrand factor molecule. The detection of genetic signals is likely to be enhanced in groups that live in a more homogeneous environment like the Amish. Linear and quadratic age effect accounted for varying proportions of the gender-specific variation in the CBC measures (from 1% to 70%). The variance component associated with additive polygenic effects was estimated for each CBC phenotype using MENDEL to obtain estimates of heritability. Significant heritability was found for platelet (h2= 0.518, p <0.0005), white blood cell (WBC) (h2=0.395, p <0.0001), RBC counts (h2= 0.358, p <0.0025) and mean corpuscular hemoglobin concentration (MCHC) (h2= 0.547, p <0.0005). Lower heritability was found for red cell distribution width (RDW) (h2= 0.217, p <0.001) and mean corpuscular hemoglobin (MCH) (h2= 0.231, p < 0.02) hematocrit (Hct) (h2= 0.126, p <0.10) and hemoglobin (Hb) (h2= 0.055, p > 0.2). Interestingly since significant heritability for Hct and Hb was reported in other studies, and to rule out the significant effect of bleeding due to the VWF mutation, we estimated heritability of Hb excluding all members of the pedigree that exhibited the 4120 C>T mutation, h2= 0. 21 (p < 0.05). A primary genetic screen at a 10 cM average interval was performed on the entire Amish pedigree in collaboration with the Marshfield genotyping center. Standard Screening Set 16 with 400 short tandem repeat polymorphic (STRP) markers was utilized generating a total of approximately 160,000 genotypes on the 395 samples. Due to data complexity, the Markov Chain Monte Carlo (MCMC)-based program LOKI was used to conduct multipoint linkage analysis for eight CBC measures that were adjusted for age, sex and mutation status to control for potential confounding. The outcome of interest from the MCMC analysis is represented by the Bayes factor (BF) which examines the probability of linkage in complex pedigrees at every centimorgan (cM). A complete analysis of the genome scan with 1,000,000 iterations was performed. A quantitative trait locus (QTL) suggesting strong linkage was identified for the RBC measure at position 4q25 with a BF of 58.17 at 114.5 cM. This result was confirmed by splitting the pedigree into several smaller subsets and conducting traditional linkage analysis using MERLIN (all individuals in the pedigree were included in at least one subset) with a LOD score of 0.66 (p < 0.05). Interestingly, a previous twin study found evidence for linkage for RBC to 4q32. Similar to animal studies, no evidence of linkage was observed in chromosomal regions known to contain the genes that encode for the hemoglobin chains, erythropoietin or erythropoietin receptor. This analysis in a very large pedigree identified a region of strong linkage for RBC that will be analyzed at higher resolution for the presence of novel modifying genes and alleles that may potentially be important for our understanding of the control of erythropoiesis.