The CA microsatellite repeat marker, D7S522, is located at the center of a approximately 1000 kb smallest common deleted region that is lost in many forms of human cancer. It has been proposed that a putative tumor suppressor gene lies in close proximity to D7S522, within this smallest common deleted region. However, the genes located in proximity to D7S522 have remained elusive. Recently, we identified five independent BAC clones (approximately 100-200 kb) containing D7S522 and the human genes encoding caveolins 1 and 2. Here, we present the detailed organization of the caveolin locus and its relationship to D7S522, as deduced using a shot-gun sequencing approach. We derived two adjacent contigs for a total coverage of approximately 250 kb. Analysis of these contigs reveals that D7S522 is located approximately 67 kb upstream of the caveolin-2 gene and that the caveolin-2 gene is located approximately 19 kb upstream of the caveolin-1 gene, providing for the first time a detailed genetic map of this region. Further sequence analysis reveals many interesting features of the caveolin genes; these include the intron-exon boundaries and several previously unrecognized CA repeats that lie within or in close proximity to the caveolin genes. The first and second exons of both caveolin genes are embedded within CpG islands. These results suggest that regulation of caveolin gene expression may be controlled, in part, by methylation of these CpG regions. In support of this notion, we show here that the CGs in the 5' promoter region of the caveolin-1 gene are functionally methylated in two human breast cancer cell lines (MCF7 and T-47D) that fail to express the caveolin-1 protein. In contrast, the same CGs in cultured normal human mammary epithelial cells (NHMECs) are non-methylated and these cells express high levels of the caveolin-1 protein. Comparison of the human locus with the same locus in the pufferfish Fugu rubripes reveals that the overall organization of the caveolin-1/-2 locus is conserved from pufferfish to man. In conclusion, our current studies provide a systematic basis for diagnostically evaluating the potential deletion, mutation, or methylation of the caveolin genes in a variety of human tumors.