Candidate gene approach in association studies: would the factor V Leiden mutation have been found by this approach?

A re-emerging strategy in the search for disease susceptibility genes is the evaluation of candidate genes, which are thought to play a role in disease pathogenesis. Candidate genes are screened for single nucleotide polymorphisms (SNPs) in a case-control study. The factor V Leiden (FVL) mutation (1691G --> A in the F5 gene) is an important risk factor for venous thrombosis. We asked ourselves whether the FVL mutation would have been found using the candidate gene approach in the absence of prior knowledge of the haplotype structure of the F5 gene. We typed four SNPs in the F5 gene in the Leiden Thrombophilia study, that is, promoter (99930G --> A), exon 13 (55907A --> G), exon 16 (42855A --> G), and intron 19 (37833T --> G). These SNPs were known to have different population frequencies, making their presence in distinct haplotypes likely. None of these SNPs has previously been associated with venous thrombotic risk. Subsequently we derived haplotypes. One haplotype was clearly more frequent in patients than controls (GAAT; 20 versus 9%), suggesting that a polymorphism in or near the F5 gene in this haplotype is associated with an increased thrombotic risk. If we had sequenced the F5 gene in patients homozygous for this haplotype, in order to locate the possible causal polymorphism, we would have found that 16 (76%) patients were homozygous or heterozygous for a missense mutation in exon 10 (1691G --> A), which predicts the replacement of Arg506 by Gln in one of the cleavage sites for activated protein C, a mutation that we now know as the FVL mutation.