Gene evolution and regulation in the chorion complex of Bombyx mori. Hybridization and sequence analysis of multiple developmentally middle A/B chorion gene pairs.

Twenty-two pairs of chorion genes belonging to the A and B multigene families have been characterized and mapped within two segments of a 320 kb (1 kb = 10(3) bases or base-pairs) chromosomal walk in the domesticated silkmoth Bombyx mori. Eighteen of the gene pairs belong to two groups that are typified by the previously characterized A/B.L12 and A/B.L11 chorion gene pairs, and are defined by two respective types of short (approx. 280 base-pairs) bidirectional promoter sequences. In the chromosome, the L12-like and L11-like pairs are interspersed with each other and with the remaining four gene pairs, which have unrelated promoter sequences. We have sequenced the promoter regions and adjacent small exons of all L12-like and L11-like A and B genes in the walk. The L12-like promoters are highly conserved, whereas L11-like promoters are somewhat more variable. Reconsideration of previous data on RNA accumulation and disappearance during choriogenesis, in the light of the sequences, indicates that L12-like genes are developmentally early-middle, while L11-like genes correspond to two developmental subgroups, middle I and middle II. Sequence comparisons of all these promoters, as well as the previously characterized promoters of the developmentally late HcA and HcB genes, identify short elements of possible regulatory significance. The sequences, as well as extensive cross-hybridization analysis with short probes derived from the reference A/B.L12 gene pair, under carefully controlled conditions of stringency, indicate the occurrence of sequence transfers among A or B genes. These sequence transfers, which could result from gene conversions or unequal crossovers, are less abundant than in the HcA and HcB families, but do result in a patchwork of similarities and differences in the A and B genes. The transfers appear to be least frequent between the moderately divergent A genes that belong to different temporal classes, while the L12-like and L11-like B genes appear to be extensively homogenized in sequence.