Founder Effect in Protein S-Deficient Families Sharing a Hot Spot Mutation in PROS1

To the Editor:

PROS1 mutations causing deficiency of the anticoagulant protein S are associated with recurrent venous thromboembolism.1Recently, an Arg(⁠$C̲$GA)410Ter(⁠$T̲$GA) mutation has been reported in five French, three Danish, and one Austrian family.2-4 Whether this Arg410Ter mutation has arisen de novo in all nine families, or whether the unusually high number of families carrying this mutation is due to a founder effect is currently unclear. CpG dinucleotides are known to be mutational hotspots, and specifically C to T transitions occur frequently.5 

Four of the five French families share the same sequence variation located in the 3′ untranslated region,2 but a possible French founder has not been confirmed yet. We want to present data demonstrating that the three Danish families (F, G, and J) have a common ancestor who has no familial relation to the Austrian family (Fam-10) (Table 1).

Families F, G, and J and Fam-10 were haplotyped by analysis of 12 microsatellite markers (D3S3571, D3S3545, D3S1296, D3S3633, D3S1276, D3S1595, D3S3634, D3S3556, D3S1603, D3S3716, D3S1271, and D3S35746 7) flanking PROS1 and spanning 30.6 cM of the centromeric region of chromosome 3. Polymerase chain reactions (PCRs) were performed in 20 μL volumes containing 0.05 mmol/L dNTP (Boehringer Mannheim, Mannheim, Germany), 1.5 mmol/L MgCl₂, Taq DNA Polymerase (Promega Corp, Madison, WI), and 4 pmol of each primer. The forward primers were labeled with fluorescent dyes (6-FAM, TET, or HEX; Perkin-Elmer, Norwalk, CT) and the PCR products were separated on an ABI PRISM377 sequencer (Perkin-Elmer), followed by analysis with GeneScan Software 2.02 (Perkin-Elmer).

The markers D3S1296, D3S3633, D3S1276, D3S1595, D3S3634, D3S3556, D3S1603, and D3S3716, spanning 18.6 cM of the 30.6 cM flanking the PROS1 gene on chromosome 3, were found to segregate with the Arg410Ter mutation in families F, G, and J but not in family Fam-10. From these results, a common founder in the Danish patients can be assumed, because the likelihood of the occurrence of this haplotype in unrelated individuals carrying the Arg410Ter mutation is 9.0 × 10⁻⁸ (Table 1). In the Austrian family, the mutation arose independently.

Our study shows that haplotype analysis in reportedly unrelated families with the same mutation at a mutational hotspot is very valuable for clarification if the sequence change is due to the existence of a common ancestor or the occurrence of independent mutational events. Missing a founder effect will result in an overestimation of the mutation rate.