Definition of the human genome will require identification of individual genes. Theoretically, coding sequences could be located by virtue of the splicing consensus sequences that border and define their constitutive exons. Unfortunately, exons are very short and are separated by very long introns in vertebrates. Furthermore, splicing consensus sequences are short; and in the case of 3' splice sites not well conserved in vertebrates, making exon detection by sequence comparison difficult. The splicing machinery, however, recognizes exons efficiently and correctly. This proposal describes an experimental strategy to utilize the natural vertebrate splicing system to detect and clone human exons. The approach takes advantage of rules for exon selection in vertebrates that have arisen from our work on vertebrate RNA splicing and polyadenylation. The final desired product will be exon libraries (both with and without adjacent intron sequences). Because this is a DNA-based technology, all genes (exons) will be expected to be represented in equal amounts; regardless of their level of expression in any given cell type. The initial goal will be to isolate exon libraries containing 3' terminal (poly(A) site-containing) exons. Each gene should be represented once (if not alternatively polyadenylated). The average size of vertebrate 3' terminal exons is 635 nucleotides; therefore, a library of 3' exons will be a convenient source for gene-specific probes for in situ chromosome location approaches.
| Krizman, D B; Berget, S M (1993) Efficient selection of 3'-terminal exons from vertebrate DNA. Nucleic Acids Res 21:5198-202 |
