Abstract
Abstract: :
Purpose: To test for genetic linkage to previously identified loci for highly penetrant autosomal dominant high myopia in a cohort of 51 UK families. Methods: The families comprised 306 individuals, 174 were classed as affected according to the criteria employed by Young et al (AJHG 1998, 63, 109-119) and DNA was available for 244. Subjects were typed for microsatellite markers spanning a ~40cM region on chromosome 18p (AJHG 1998, 63, 109-119), 12q (AJHG 1998, 63, 1419-1424) and 17q (AJHG 2001, 66, S313), together with markers flanking COL2A1, COL11A1 and FBN1. Allele frequencies were estimated from founders and married-in individuals. We used the same disease gene segregation model as was employed in the original publications. Two-point linkage analysis was performed using MLINK, and non-parametric and multipoint analyses using Genehunter, with additional maximization over the parameter α, the proportion of linked families. Estimates of power and empirical p-values were determined from SLINK simulations. Results: Maximum two-point and multipoint results were: 18p - Two-point HLOD = 0.593, multipoint HLOD = 0.278 and NPL = 0.784 for D18S59. 12q - Two-point Zmax = 2.614 (asymptotic p = 0.0003) and multipoint HLOD = 1.405, α = 0.26 (empirical p = 0.009) for D12S332. NPL = 1.821 (p = 0.036) for D12S1607. 17q - Two-point HLOD = 0.597 and multipoint HLOD = 0.015 for D17S1868. NPL = 1.266 (p = 0.1034) for D17S942. As there was weak evidence of excess allele sharing, but no evidence for a dominant locus by parametric analysis, the analyses were repeated for a recessive model: Two-point HLOD = 1.240, α = 0.30 for D17S1795, multipoint HLOD = 1.261, α = 0.17 (empirical p = 0.013) between D17S956 and D17S956. Connective tissue genes - No evidence for linkage or excess allele sharing. Whilst low LOD scores do not preclude linkage in a proportion of families, power simulations show that such results are unlikely if a locus is causative in 50% or more of the families. Conclusions: This study provides nominally significant evidence for replication of linkage to the 12q locus, no evidence of a major contribution from the 18p locus and equivocal evidence for involvement of the 17q locus. When corrected for multiple testing the maximum multipoint HLOD for 12q is borderline significant. Considering the low power in the presence of heterogeneity and statistical factors involved in replicating linkage, these results support the conclusion that a gene on chromosome 12q is responsible for high myopia in excess of 25% of UK families showing apparent fully penetrant autosomal dominant transmission.
Keywords: myopia • linkage analysis • genetics