Construction of a GT polymorphism map of human 9q

To construct a framework map of human chromosome 9 consisting of highly informative markers, we identified 36 cosmid clones from chromosome 9 that contained long GT repeat sequences. The cosmids were found to cluster on the long arm of the chromosome, particularly in the q32-34 region. Thirteen highly informative polymorphisms from 9q were identified, with median observed heterozygosity 0.75 and median calculated heterozygosity based upon allele frequencies of 0.75. These new GT repeat polymorphisms (D9S56, D9S58-67), as well as anchor GT polymorphisms for D9S15 (MCT112, 9q13), and ABL and ASS (both 9q34.1) were utilized to construct a linkage map of human 9q by the typing of the Venezuelan Reference Pedigree. Care was taken to avoid errors, including analysis of the data with CHROMLOOK and verification of all double crossover events detected within a 30 cM interval by repetition of the marker analysis. The map was generated using the MAPMAKER program. All positions in the resulting map are favored by odds of greater than 10(4):1. The map has a sex-averaged length of 90 cM (Kosambi function) with a single maximum intermarker recombination fraction of 26%. All other intermarker recombination fractions are less than 15%. As D9S15 is known to be closely linked to markers on proximal 9p, and ASS/ABL are in band 34.1, this set of GT polymorphisms spans the length of 9q and provides a useful panel for linkage analysis of disease genes to this region. The marker order was confirmed by in situ hybridization of the cosmid clones to metaphase spreads of normal human chromosomes, which indicated an excess of recombination in the telomeric region in comparison to centromeric 9q, in agreement with previous chiasmata distribution observations. Two spontaneous new mutations for these GT repeat markers were identified, giving an overall observed spontaneous mutation rate of 0.00045 per locus per gamete. Direct observation of new mutations has not been previously reported for dinucleotide polymorphisms, but the observed rate is consistent with frequencies observed for other VNTR polymorphisms.