Genome Editing of Factor IX Paralogs in Zebrafish Demonstrates Subfunctionalization of Procoagulant Function

Hemophilia is an inherited monogenic bleeding disorder affecting ~400,000 individuals worldwide, with 20% of cases due to deficiency of factor IX (FIX, hemophilia B), affecting 1 in 30,000 male births. Although there are many exciting new treatments for this disorder, there are still unknown features of FIX. This includes potential distribution from plasma to tissue, which may affect some therapies. The zebrafish is a unique model to study coagulation disorders due to its high fecundity, external development, and optical transparency, as well as a high degree of genetic and functional conservation with the mammalian hemostatic system. Due to an ancestral gene duplication event in zebrafish, 30-40% of the zebrafish genome is duplicated, including the FIX gene (f9), resulting in the paralogs f9a and f9b. Such paralogs may arise due to subfunctionalization of the ancestral gene, raising the possibility that this could identify roles for mammalian FIX that have not previously been discovered. Here we utilize CRISPR/Cas9 genome editing to produce large deletions in f9a and f9b using guide RNAs that delete the intervening regions between exons 2 and 8, uncovering evidence for subfunctionalization.

Assessment of hemostasis in 3 day post fertilization (dpf) mutant larvae using laser-mediated endothelial injury unexpectedly revealed that single and double f9a and f9b homozygous mutants exhibited normal thrombus formation. Upon crossing these mutants into an antithrombin (at3) deficient background (which exhibits spontaneous venous thrombosis), loss of F9b (f9b^-/-;at3^-/- mutants) prevented intravascular thrombosis development. f9a^-/-;at3^-/- double mutants were indistinguishable from f9a^+/+;at3^-/- siblings, both showing spontaneous thrombosis. These results demonstrated subfunctionalization of F9 activities, which could be due to differential functions or tissue-specific expression. The latter was investigated using whole mount in situ hybridization in wild-type larvae, revealing that f9a and f9b are both expressed in the liver. To evaluate function, microinjection of zebrafish f9a and f9b cDNAs regulated by a ubiquitous promoter was performed in one-cell stage f9b^-/-;at3^-/- zebrafish embryos. Analysis at 3 dpf found that only F9b could restore spontaneous thrombosis, confirming functional differences between F9a and F9b.

Alignment of human FIX to F9a and F9b indicated conservation of functional domain sequences (signal sequence, propeptide, Gla, Egf1, Egf2, activation peptide, and protease). These domains were individually swapped between the f9a and f9b plasmids. Except for Egf1, all individual domains of F9a exchanged into F9b were able to restore thrombus formation in f9b^-/-;at3^-/- mutants. Replacement of F9b Egf1 with a F9a^Q59G modified Egf1 domain (F9b has a glycine at this orthologous position), resulted in normal function.

In summary, these data show clear evidence for subfunctionalization of F9 in zebrafish larvae, with F9b retaining procoagulant function. F9a possesses an active protease domain, but variation in Egf1 prevents it from contributing to canonical circulating FIX activity. We hypothesize that the alternative role for F9a might represent a previously unknown activity or tissue localization of human FIX. This putative role would also be deficient in hemophilia B patients and/or may be lacking in current therapies. An additional possibility is that this represents another means for regulation of FIX. In both cases, further study could lead to novel therapeutic ideas for patients with hemophilia B.