Structure-Based Design and Discovery of a Long-Acting Cocaine Hydrolase Mutant with Improved Binding Affinity to Neonatal Fc Receptor for Treatment of Cocaine Abuse

Despite decades of efforts to develop a pharmacotherapy for cocaine abuse treatment, there is still no FDA-approved treatment of diseases associated with this commonly abused drug. Our previously designed highly efficient cocaine hydrolases (CocHs) and the corresponding Fc-fusion proteins (e.g., CocH3-Fc) are recognized as potentially promising therapeutic enzyme candidates for cocaine abuse treatment, but all with limited biological half-lives. In order to prolong the biological half-life and, thus, decrease the required frequency of the enzyme administration for cocaine abuse treatment, we have modeled the Fc-fusion CocH binding with neonatal Fc receptor (FcRn) in the present study. This approach led to the design and testing of CocH3-Fc(M6), a CocH3-Fc mutant with nearly 100-fold increased binding affinity: from K_d = ~ 4 μM to K_d = 43 nM. As a result, CocH3-Fc(M6) indeed revealed a markedly prolonged biological half-life (t_1/2 = 206 ± 7 h or ~ 9 days) in rats, longer than other known Fc-fusion protein drugs such as abatacept and alefacept (for other therapeutic purposes) in the same species (rats). It has been demonstrated that a single dose of 3 mg/kg CocH3-Fc(M6) effectively blocked 20 mg/kg cocaine-induced hyperactivity on day 18 after CocH3-Fc(M6) administration. This is the first attempt to rationally design long-acting Fc-fusion enzyme mutant based on combined computational modeling and experimental measurement of the Fc-fusion CocH binding with FcRn. The similar structure-based design strategy may be used to prolong the biological half-lives of other Fc-fusion protein drugs.