John E. Coe, William J. Nelson, F. Hermann Rudenberg and Raymond Garza
Maxwell S. H. Laurans, Michael L. DiLuna, Dana Shin, Faheem Niazi, Jennifer R. Voorhees, Carol Nelson-Williams, Eric W. Johnson, Adrian M. Siegel, Gary K. Steinberg, Michel J. Berg, R. Michael Scott, Gioacchino Tedeschi, T. Peter Enevoldson, John Anson, Guy A. Rouleau, Christopher Ogilvy, Issam A. Awad, Richard P. Lifton and Murat Gunel
Object. A gene contributing to the autosomal-dominant cerebral cavernous malformation (CCM) phenotype, KRIT1 (an acronym for Krev Interaction Trapped 1), has been identified through linkage analysis and mutation screening. The authors collected blood samples from 68 patients with familial CCM and 138 patients with apparently sporadic CCM as well as from their families, in an effort to characterize the prevalence and spectrum of disease-causing sequence variants in the KRIT1 gene.
Methods. The authors used single-strand conformational polymorphism analysis to identify genomic variants in KRIT1, which were sequenced to determine the specific mutation. Among 43 Hispanic-American kindreds who immigrated to the southwestern US from northern Mexico, 31 share an identical founder mutation. This Q455X mutation is found in 18 (86%) of 21 persons with a positive family history and in 13 (59%) of 22 persons with apparently sporadic CCM. This mutation was not found among 13 persons with CCM who were recruited from Mexico. These findings establish the key role of a recent founder mutation in Hispanic persons with CCM who live in the US.
Although nearly all Hispanic families in the US in which there are multiple CCM cases linked to the CCM1 locus, only 13 of 25 non-Hispanic CCM-carrying families have displayed evidence of linkage to the CCM1 locus. Among these 13 families, the authors identified eight independent mutations in nine kindreds. They identified four additional mutations among 22 familial CCM kindreds with no linkage information, bringing the total number of independent mutations to 12. Inherited KRIT1 mutations were not detected among 103 non-Hispanic persons in whom a family history of CCM was rigorously excluded.
Conclusions. All mutations were nonsense mutations, frame-shift mutations predicting premature termination, or splicesite mutations located throughout the KRIT1 gene, suggesting that these are genetic loss-of-function mutations. These genetic findings, in conjunction with the clinical phenotype, are consistent with a two-hit model for the occurrence of CCM.