Recognition patterns for exon-intron junctions in higher organisms as revealed by a computer search.

Most genes of higher eukaryotes are discontinuous. The DNA which codes for a protein is interrupted by introns. In the nucleus, certain machinery is supposed to recognize the exon-intron and intron-exon junctions. Applying computer searching to eighteen genes from various biological species, we examined what kinds of patterns or nucleotide sequences are necessary and sufficient to recognize the splice junctions. We propose that four common patterns of AG/GTA, /GTAAGT, RG/GTGAG and AG/GTXXGT, where R = A or G and X = A, T, G, or C, are often used as signals for exon-intron junctions. This proposal is based on the facts that they are not found in the exons of the genes examined and on the assumption that the recognition machinery scans the mRNA precursor from the 5'-end to the 3'-end.