Abstract

Familial polyposis coli (FPC) is caused by an autosomal dominant gene on chromosome 5, and it has been proposed that colorectal cancer in the general population arises from loss or inactivation of the FPC gene, analogous to recessive tumor genes in retinoblastoma and Wilms' tumor. Since allelic loss can be erroneously scored in nonhomogeneous samples, tumor cell populations were first microdissected from 24 colorectal carcinomas, an additional nine cancers were engrafted in nude mice, and nuclei were flow-sorted from an additional two. Of 31 cancers informative for chromosome 5 markers, only 6 (19%) showed loss of heterozygosity of chromosome 5 alleles, compared to 19 of 34 (56%) on chromosome 17, and 17 of 33 (52%) on chromosome 18. Therefore, it appears that (i) FPC is a true dominant for adenomatosis but not a common recessive gene for colon cancer; and (ii) simple Mendelian models involving loss of alleles at a single locus may be inappropriate for understanding common human solid tumors.

Related Content