Comparative sequence analysis of mammalian factor IX promoters.

Functionally important nucleotides tend to be conserved between species. The advent of PCR has made it possible to rapidly obtain corresponding sequences from numerous species. We have used this approach to obtain sequence data on the factor EX promoter from various mammals, in the hope of determining which nucleotides contribute to factor IX expression in vivo. The results of CAT assays carried out in transiently transfected HepG2 cells, suggest that the promoter lies in the region —98 to +21 (3) (numbering throughout is relative to the most 5' start point of transcription in the human gene, +1 (1, 2)). This region has been shown to contain sequences which can bind NF1 ( -99 to -77) and C/EBP (+1 to +18) in vitro (3). Further evidence that this region is involved in factor EX expression comes from the study of individuals with altered levels of factor EX, and who suffer from a form of haemophilia, known as haemophilia B L êyden. Mutations, in these patients, have been detected at —20, —6 and at +13 (4, 5, 6). Moreover, it has been shown that the A-*G mutation at +13 interferes with the binding of C/EBP to its recognition site (+1 to +18) (3). It is known that the factor EX structural gene is well conserved in mammalian species and cross-species amplification of coding regions has been achieved (7). We took the same approach in an effort to amplify 5' flanking sequences. Two downstream and three upstream primers were prepared. One of the downstream primers, +84 to +106 (A) was homologous to the 5' region of the recently reported mouse cDNA (8), the other was homologous to the human sequence, +71 to +111 (B) (9). The three upstream primers, -133 to -110 (C), -229 to -206 (D) and -323 to -298 (E), were all homologous to the human sequence. Successful amplification of the mouse factor EX 5' region was achieved with primers A and C as well as B and C, whilst the rat factor EX 5' region was amplified only with B and C. In the case of the monkey, all combinations of upstream and downstream primers resulted in successful amplification. The amplified fragments were gel purified and directly sequenced (10). They were also cloned into pBluescript sk+ (Stratagene) and several independent plasmid isolates were sequenced. Difficult regions were checked by reamplifying with one 5' P end-labelled and one unlabelled primer (11) and sequencing the resulting fragment by the Maxam and Gilbert method.