Oxytocin (OT) has long been deemed the “love hormone” after its important role in social bonding has been documented. In the current study, the relative importance of five OT residues (X2–X5 and X8) was profiled systematically and quantitatively through an integrative approach. In the procedure, eight quantitative sequence-activity models were developed with classical amino acid descriptors and sophisticated random forest (RF) regression based on a panel of 49 known OT analogues. These models were further optimized to achieve both high internal fitting ability and strong external predictive power. The best model, which was built from the combination of VHSE descriptor and RF regression, was then employed to characterize the single position mutation profile of the five investigated OT residues. It is suggested that (i) the polar and charged amino acids presented at positions X2, X3 and X8 are quite unfavorable to OTR–OT binding, and (ii) the positions X4 and X5 have only a moderate effect on the binding as compared to other three positions. Subsequently, three novel OT analogues, i.e. the single-point mutants , and , were designed rationally in terms of the putative mutation profile, and their interaction mechanism with OTR were investigated at molecular level using atomistic molecular dynamics (MD) simulations and molecular mechanics/generalized Born surface area (MM/GBSA) analysis. Further, the agonistic potencies U of OA12 and OA90 were measured to be 0 and 190 IU/mg, respectively, by testing their agonistic capability against rat uterus, which are roughly consistent with the theoretical results arising from computational investigations.
Download Full PDF Version (Non-Commercial Use)