[Vascular hereditary dementia CADASIL type in Colombia. III. Linkage analysis to notch3 gene]

To perform linkage analysis between the Short Tandem Repeats (STR) microsatellite markers D19S923, D19S929, D19S22, which are in strong genetic linkage to Notch3 gene in order to contrast the hypothesis that the vascular hereditary dementia phenotype described in a multigenerational extended pedigree from Colombia correspond to CADASIL (Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy). Even we know that using techniques as the Single Strand Conformational Polymorphisms (SSCP) could determine mutations in Notch3, the rationality of this approach is that intronic variations could not be defined and that we are interested in determine if some forms of the clinical presentation and its phenotypic variability make part of CADASIL.

The CADASIL phenotype is caused by mutations in the Notch3 gene. Clinical features of CADASIL are: 1. Recurrent cerebra-vascular episodes; 2. Migraine history; 3. History of transitory ischemic attack and, 4. Behavior changes and dementia.

By using SIMLINK we showed that the extended genealogy had the enough power to detect significant LOD (logarithm of oods) score values when Notch3 was considered the disorder cause. Linkage analysis was carried out by using parametric and non parametrical methods. The Elston-Stewart general method was used as the parametrical analysis and the sib pair method as the non-parametrical one. We perform simulations changing the affection status codification by including as affected or not including those individuals with migraine. Furthermore, in order to detect the stability of the results, we changed the penetrance values, the genetic frequencies on both, the marker loci and the affection locus.

The maximum pair-wise LOD score was 2.04 which was detected at the marker D19S23 with q= 0.11cM. This distance correspond exactly with the Notch3 location. That is 100 times more probable that there is linkage that there is not. In other words this probability could be explained as if the phenotype correspond to CADASIL than to other vascular dementia. The non parametric results were compatibles with the parametric ones. When the migraine symptom was considered as a part of the affected status, the LOD score values showed not linkage.

The results of the linkage analysis to these STR microsatellite markers suggest that the vascular hereditary dementia phenotype described in this family correspond to CADASIL caused by a polymorphism on the Notch3 gene. On the contrary, these same results suggest that the migraine phenotype is not a part of the progressive dementia.