Genomic organization of core 2 and I branching beta-1,6-N-acetylglucosaminyltransferases. Implication for evolution of the beta-1,6-N-acetylglucosaminyltransferase gene family.

Two human beta-1,6-N-acetylglucosaminyltransferases forming the core 2 O-glycan branch, C2GnT and the I antigen, IGnT, are homologous to each other in three regions of the catalytic domain (A, B, C) and their genes reside at the same locus, chromosome 9, band q21 (Bierhuizen,M.F.A., Mattei, M.-G. and Fukuda,M., Genes Dev., 7, 468-478, 1993). In order to investigate how these two enzymes are related at the genomic level, and how this gene family evolved, we have elucidated their genomic structures. It was found that C2GnT is coded by two exons, of which the second exon encodes the whole translation product. In contrast, the complete coding sequence for IGnT is divided over three exons. Importantly, the highly homologous region B is encoded entirely by exon 2 in the C2GnT gene, while the same region is split between exons 1 and 2 in the IGnT gene. The other highly homologous regions, A and C, are also encoded by exon 2 in the C2GnT gene, while they are encoded by exon 1 and exon 3, respectively, in the IGnT gene. These results strongly suggest that the common ancestral gene was first duplicated and then each duplicated gene evolved into the C2GnT or IGnT gene by intron insertion and divergence following the duplication. The sequences upstream from the transcription initiation sites of the C2GnT and IGnT genes have promoter activity and contain TATA-like sequences.(ABSTRACT TRUNCATED AT 250 WORDS)