The main reactions of the hepatitis C virus NS3/NS4A serine protease are studied using the second-order Møller-Plesset ab initio method and rather large basis sets to correct the previously reported AM1/CHARMM22 potential energy surfaces. The reaction efficiencies measured for the different substrates are explained in terms of the tetrahedral intermediate formation step (the rate-limiting process). The energies of the barrier and the corresponding intermediate are so close that the possibility of a concerted mechanism is open (especially for the NS5A/5B substrate). This is in contrast to the suggested general reaction mechanism of serine proteases, where a two-step mechanism is postulated.
