Background: Serine protease inhibitors (SERPINs) are a superfamily of inhibitors, which act as molecular mousetraps. SERPINs only interact with activated serine proteases. This enables physiological protease activity, but limits it to prevent pathology. Interestingly, the specificity of SERPINs can be altered by changing their reactive center loop (RCL), which has substantial therapeutic potential. Recently, an engineered α1-antitrypsin (α1AT) variant (³⁵⁷KRK³⁵⁹) that is specific for activated protein C (APC), showed great promise for the treatment of hemophilia B (Polderdijk et al, 2017). Currently, there is limited insight into how RCL variation affects SERPIN specificity, mostly because design is based on synthetic tripeptide substrate studies. This hampers effective redesign of therapeutic SERPINs.

Aims: To develop a SERPIN specificity prediction platform and to apply this in the design of novel therapeutic SERPINs.

Methods: Insight into RCL specificity was obtained by creating eight libraries for each of the key positions within the RCL (P4-P4′; Figure 1) based upon α1AT-Pittsburgh backbone (P1=arginine; M358R). Mutants were screened against 7 coagulation enzymes and marked for their inhibitory capacity.

Results: The inhibitory capacity from the eight libraries (160 variants) were combined into a linear prediction model. We subsequently tested its predictive power in separate test-cohorts (P3 and P2 mutations only) where it showed strong correlation against 4 out of 7 targeted enzymes. In comparison, no correlation was observed against the predictive value of peptide-substrate libraries. To validate our platform, we designed novel APC-specific inhibitors for therapeutic use in hemophilia. Compared to the published ³⁵⁷KRK³⁵⁹ variant, our SERPINs showed a 3 to 5-fold increase in APC inhibition, while maintaining the required specificity. Furthermore, in the presence of thrombomodulin, our SERPINs restored the coagulation potential at 10 to 20-fold lower concentrations compared to the ³⁵⁷KRK³⁵⁹ variant.

Conclusions: The library based approach offers unique insight into SERPIN RCL-specificity, allowing for the improved design of therapeutic SERPINs.

[α1-antitrypsin reactive center loop libraries. Libraries were created based upon the α1AT-Pittsburgh reactive center loop (RCL). ]

« Back to ISTH 2020 Congress

ISTH Congress Abstracts - https://abstracts.isth.org/abstract/a-novel-prediction-platform-to-enhance-the-design-of-therapeutic-serpins/