The carboxyl-terminal region of human coagulation factor X as a novel naturally-occuring linker for fusion strategies

Background: Fusion with human serum albumin (HSA) represents a well-established technique to extend half-life of therapeutic proteins used to treat bleeding disorders. A key issue to be addressed is the fusion strategy, which mainly involves intervening linker sequences ensuring proper spatial separation and biological activity. Here, we explored direct fusion to coagulation factor X (FX) carboxyl-terminal region, previously demonstrated by us to be dispensable for secretion/function, as a natural linker suitable for fusion with HSA. We also produced, as comparators, fusion proteins exploiting flexible (glycine/serine; FX-GS-HSA) or cleavable (incorporating the FX activation site; FX-CL-HSA) linkers, mimicking the long-acting recombinant activated factor FVII (rFVIIa)-HSA and factor IX (FIX)-HSA, respectively. Methods: Proteins were stably expressed in human embryonic kidney 293 cells and characterized for secretion (ELISA), activation (Western blotting, time-course amidolytic assays) and activity (thrombin generation, PT-based assays). Concentrated proteins were tail-vein injected into wild-type C57/BL6 mice to evaluate half-life. Results: All chimeras were efficiently secreted and possessed remarkable activity, with FX-HSA and FX-CL-HSA displaying comparable functional features both in thrombin generation and PT-based assays. The two chimeras were able to shorten coagulation times in a dose-dependent manner, with virtually normal pro-coagulant activity for FX-HSA (88.7±6.0% of FX) and FX-CL-HSA (98.0±16.2%). Conversely, the FX-GS-HSA variant showed a reduced efficiency in both assays, with a coagulant activity of 55.8±5.4% of FX. Upon incubation with activators, FX-HSA and FX-CL-HSA displayed a correct activation profile while the FX-GS-HSA activation was slightly defective. In fluorogenic-based assays, the FX-HSA showed normal activity over time and a specific amidolytic activity (1.0±0.12) comparable to that of FX. Moreover, FX-HSA showed prolonged plasma persistence in mice (the fusion protein was detectable by ELISA at 36 hours post-injection whereas FX alone disappeared at 16 hours). Overall, the FX-HSA features indicate that the FX carboxyl-terminus represents an intrinsic sequence allowing tandem fusion. Conclusion/summary: Our results provide the first experimental evidence for i) a coagulation factor fusion protein with features independent from artificial linkers, ii) the suitability of FX carboxyl-terminal region as a novel naturally-occurring linker for fusion purposes, and iii) the suitability of FX as a versatile platform for protein engineering, with potential therapeutic meaning for patients. Interestingly, we can try to speculate that the absence of an artificial sequence joining FX with HSA, as well as with other potential fusion partners, could have a low association with the risk of triggering an immune response.