SummaryFactor VIII is a trace plasma glycoprotein involved as a cofactor in the activation of factor X by factor IXa. Inherited deficiency of factor VIII results in the X-linked bleeding disorder hemophilia A which has been documented in humans, horses, sheep and dogs. In this report, the putative proximal promoter, 5’ untranslated region, complete coding sequence and 3’ untranslated region of the canine factor VIII gene have been characterized. When compared to the human gene, the 5’ flanking region shows conservation of transcription factor binding sites in the 5’ untranslated region. Alignment of the amino acid sequence with that of the previously reported human, mouse and pig proteins demonstrates sequence identity of between 77 and 92% for the A1, A2, A3, C1 and C2 domains but an identity of only between 44 and 62% for the central B domain. The three thrombin cleavage sites are conserved in the canine sequence as are the protein C cleavage sites and the von Wille-brand factor binding region. In addition, all six tyrosine residues that are known to undergo sulfation in the human protein are conserved in the dog. The 3’ untranslated region of the canine gene extends 1.5 kilo-bases. The initial 700 basepairs of this sequence are highly GC rich and the sequence terminates with 2 alternative potential polyadenylation sites. The knowledge of this sequence, in combination with a well described canine model of hemophilia A, provides the necessary starting point for studies addressing the long-term evaluation of factor VIII gene therapy using a homologous transgene.

Coagulation factor VIII is an essential cofactor required for normal hemostatic function. A deficiency in factor VIII results in the bleeding disorder hemophilia A. Despite the fact that the factor VIII gene was cloned a decade ago, the mechanisms which control its transcription remain unresolved. In our studies, we have characterized 12 protein binding sites within the factor VIII promoter by DNase I protection assays performed with rat liver nuclear extracts. Three of these elements (sites 1 to 3) are situated within the 5 untranslated region of the gene, while three other sites (sites 4 to 6) lie within the first 100 bp upstream of the transcriptional start site. We have identified an additional site (site 7) ϳ300 bp upstream from site 6, as well as a cluster of five sites in a 250-bp region which terminates ϳ1 kb from the transcriptional start site. Seven of these binding sites (sites 2, 3, 4, 6, 7, 9, and 10) bind members of the C/EBP family of transcription factors. DBP also binds to five of these sites (sites 3, 4, 6, 7, and 9). Utilizing transient transfection studies in HepG2 cells, we have shown that deletion of the factor VIII promoter sequences distal to nucleotide ؊44 results in a significant but small increase in promoter activity. The activity of each of the various 5 deletion constructs is significantly enhanced by cotransfection of C/EBP␣ and D-site-binding protein expression plasmids, while cotransfection of both C/EBP␣ and C/EBP␤ plasmids resulted in a further enhancement of transactivation. These studies also provide evidence of a repressor element located between nucleotides ؊740 and ؊1002. Since the minimal promoter sequence (؊44 to ؉148) maintains the transcriptional activity of the full-length promoter sequence, we proceeded to identify additional factors binding to sites 1 to 4. Competition studies revealed that a ubiquitous transcription factor, NF-Y, binds to site 4, while the liver-enriched transcription factor hepatocyte nuclear factor I (HNF-1) binds to site 1. Mutation analysis of the minimal promoter demonstrated that HNF-1 is critical for activating transcription of the factor VIII gene in vitro. Our results also suggest that the multiple upstream elements that we have identified may act as a backup regulatory region in the event of disruption of the HNF-1 element in the 5 untranslated region.