Complete genomic organization of futb encoding a bovine alpha 3-fucosyltransferase: exons in human orthologous genes emerged from ancestral intronic sequences.

The futb gene, which encodes the first bovine alpha 3-fucosyltransferase described, consists of five exons (a, b, c, d, and e), the first four being located upstream of the coding exon e. Together with the four introns (i1, i2, i3, and i4) they span a DNA genomic sequence of about 10 kb. futb is expressed as four tissue-specific transcripts differing by their 5'-untranslated (5'-UT) regions, but only one transcript includes all exons, while the other three begin at internal sites of exon c. A short sequence of the latter is homologous to distinct 5'-UT exons of FUT6 (alpha 3-fucosylation) and FUT3 (alpha 4-fucosylation), two human genes whose coding sequences are homologous to coding exon e of futb. Upstream and downstream, the exon c intronic regions of the bovine gene are homologous to 5'-UT exons of human FUT3 (exon B) and FUT6 (exons A, B, and C). Thus, exon c appears to be the most ancestral 5'-UT exon known among these alpha 3-fucosyltransferase genes. Interestingly, distribution of short interspersed nuclear elements in the i3 intron adjacent to exon c reveals that two repeat sequences are joined to form a reverse-transcriptase-like encoding sequence highly homologous to an open reading frame located at the 3' end of the bovine gamma globin gene. This organization suggests that duplication events that have generated the primate FUT3-FUT5-FUT6 cluster might have occurred through a long-interspersed-nuclear-element-based mechanism of unequal crossing over, as described for the globin cluster. Complete organization of the bovine futb gene reveals that in addition to duplication events, the lineage leading to primate FUT3, FUT5, and FUT6 genes results from rearrangements of intronic sequences which have created for each new gene specific regulatory 5'-UT exonic sequences.