Zinc fingers poke zebrafish, cause thrombosis!

In this issue of Blood, Liu et al describe the creation of a null mutation for the antithrombin III gene (at3) in zebrafish by using zinc finger nuclease technology.¹ 

In this technology, one injects zebrafish embryos with a pair of RNAs, each coding for both a restriction endonuclease domain (Fok1) to cleave DNA and a DNA-binding domain containing zinc fingers (see figure). These zinc fingers recognize specific sequences flanking a gene’s target coding sequence. This design allows the Fok1 domains to dimerize and cleave the middle portion of the DNA target site. The cell’s endogenous repair system, by the nonhomologous end joining (NHEJ) mechanism, then fixes the cleaved site, causing insertions or deletions. These changes create a different reading frame for the genetic code and typically result in a null mutant phenotype. This zinc finger nuclease technology was used in zebrafish embryos to target several genes.²  Liu et al used this zinc finger technology and elegantly generated null mutations by targeting exon V of zebrafish at3. These AT3–deficient fish exhibited consumptive coagulopathy, or disseminated intravascular coagulation (DIC), as seen by prolongation of time to occlusion (TTO) of the vessel when larvae were subjected to laser injury.

In vertebrates, blood coagulation after injury is initiated by tissue factor (TF) which binds to activated factor VII, thus triggering the coagulation cascade and ultimately generating the enzyme thrombin which cleaves fibrinogen to generate a fibrin clot (see figure). This cascade, required for normal hemostatic function, prevents excessive bleeding when injury occurs. However, a negative feedback reaction to prevent superfluous clotting is also essential to normal hemostasis. Thus, factors such as AT3 subsequently inhibit thrombin and its generation. A lack of AT3 hampers this negative feedback and results in the continuous generation of thrombin, leading to thrombosis.

Human mutations which reduce AT3 levels or its activity have resulted in severe thrombotic conditions, whereas at3 null mutations in mice have resulted in embryonically lethal conditions. Furthermore, at3 null mutations are not found in humans, presumably also due to embryonically lethal conditions. Interestingly, Liu et al reported that the null mutation of zebrafish at3 caused venous thrombosis but not embryonic fatalities. This finding suggests that AT3 is not vital to embryonic development, possibly because thrombosis did not occlude major vessels. Furthermore, if thrombosis were to block minor crucial vessels, embryonic and larval fish are still able to obtain oxygen via epithelial cell diffusion.³ 

After creating the null mutant fish, Liu et al injected either zebrafish or human at3 complementary DNA (cDNA) into the null embryos and were able to rescue the null phenotype, as demonstrated by shorter TTOs. It is well known that inhibition of thrombin by AT3 is enhanced by heparin, purportedly released after injury, which then binds to the heparin-binding site on AT3. Remarkably, when human at3 cDNA containing a mutation in the heparin-binding site sequence was used in these injections, the null fish were still rescued. However, the human at3 cDNA with a mutated reactive center loop sequence was unable to rescue the null phenotype. These results indicate that the heparin-binding site is irrelevant to human AT3 function in zebrafish. Interestingly, previous studies have found that the combination of heparin and zebrafish plasma partially inhibited human thrombin.⁴  This diminished inhibition could be caused by either reduced AT3 levels in zebrafish or species specificity of the AT3/thrombin interaction. The finding of Liu et al that the heparin-binding site is dispensable again emphasizes the species specificity of the AT3/thrombin interaction.

In vitro studies have previously shown that fish blood clots faster than mammalian blood and that the fish anticoagulation system is weaker.^4,5  Although this weaker anticoagulant system may offer an explanation for faster clotting, other possibilities for this disparity include an enhanced coagulation system or a weaker fibrinolytic system. Because zebrafish coagulation occurs faster than mammalian coagulation, one would expect removal of AT3 to result in fatal thrombosis in adult fish. However, Liu et al have demonstrated that adult zebrafish are still able to survive with the null mutation, albeit with a shorter lifespan than their wild-type kin. One hypothesis for the lack of fatal thrombosis in at3 null adult fish is that coagulation activity has already reached its threshold. Thus, although the loss of a weaker anticoagulation system is not fatal to young adults, advancing age, and the enhanced clotting associated with it, results in catastrophe in older fish. Similarly, decreased activity of the fibrinolytic system during aging could also contribute to fatal thrombosis in the null fish. These questions require future investigation.

Although earlier reports documented the first observation of a prothrombotic condition by the knockdown of factor VIIi, because factor VIIi is not present in humans, this knockdown is irrelevant to human thrombosis.⁶  Thus, the work of Liu et al is remarkable as the first gene mutation generated in zebrafish that achieves complete knockout in an anticoagulation factor that is pertinent to human thrombosis. Interestingly, though an easier method for site-specific DNA cleavage is now available, the work of Liu et al is especially laudable given that the more intricate zinc finger nuclease system was used to target at3. The creation of this at3 null model not only provides insight into the comparison of blood clotting between fish and mammals but also provides a model to study the reversal of the AT3 null phenotype. Chemical screens could be used on the AT3 null phenotype to select fish with longer lifespans or fish without DIC. Such screens may provide a novel thrombin suppressor, and although wild-type zebrafish could also be used to detect thrombin inhibitors via a direct thrombin assay, this null fish could also be used to identify thrombin inhibitors and other coagulation factor inhibitors. Furthermore, this phenotype could be used to find gene therapies to correct the mutation. Thus, this model system is innovative and should be useful to the hemostasis and thrombosis community.