We have collected samples from 61 European and North American families with PCD. A genome-wide linkage search was performed in 31 multiplex families (169 individuals including 70 affecteds) using 188 evenly-spaced (19 cM average interval) polymorphic markers. Both parametric (recessive model) and non-parametric (IBD allele sharing) linkage analyses were used. No major locus for the majority of the families was identified, although the sample was powerful enough to detect linkage if 40% of the families were linked to one locus. These results strongly suggest extensive locus heterogeneity. Potential genomic regions harboring PCD loci were localized on chromosomes 3p, 4q, 5p, 7p, 8q, 10p, 11q, 13q, 15q, 16p, 17q and 19q. Linkage analyses using PCD families with a dynein arm deficiency provided �suggestive� evidence for linkage to chromosomal regions 8q, and 16pter, while analyses using only PCD families with situs inversus resulted in �suggestive� scores for chromosomes 8q, and 19q.
Our genome wide linkage analysis of 31 multiplex families with ICS using a set of polymorphic markers with an average density of 19 cM provided strong evidence against the existence of a single locus responsible for the majority of families with ICS. Furthermore we failed to detect the existence of 2 major loci, each of which could be responsible for at least 40% of the families with ICS. None of the loci investigated showed significant values for linkage, but several chromosomal regions were identified by both types of analysis (3p, 4q, 5p, 8q, 15q, 16p, 17q, 19q). Additional regions suggested by parametric and/or non-parametric were on 6q, 7p, 10p, 11p.
Among these Loci, the results on chromosomes 4q, 5p, 8q, 16p and 19q support more strongly the possibility of gene implicated in ICS. Our results strongly suggest that ICS a very heterogeneous autosomal recessive phenotype with at least 3 (most likely) genes responsible for this phenotype in different families.
Depending on the family�s subset included in the analyses (SI or DAD), the linkage scores show some significant variations. For example, In the DAD group the scores of the chromosome 8q locus increase significantly for both parametric and non parametric analyses. Chromosome 19q and 16p loci also show important and interresting variations of the linkage scores between all groups. Although it is likely that some of these are false positive results, one should not negate the possiblity of their anchoring genes for the ICS phenotype. Some of the linked loci from our study are known to host a dynein isoform gene (DNAH1, DNAH3 and DNAH12 on 3p21-p14, DNAH7 on 7p, DNAH5 on 5p14-p12 or 8q22-q23, DNAH17 on 17q25). The different dynein genes are excellent candidates for the phenotypes of ICS.