Modified nucleotides are ubiquitous with RNAs, also in contact with drugs that target the ribosome. Whether this represents a stabilization of the drug-ribosome complex, thus affecting the drug\'s affinity and possibly also intrinsic efficacy, remains an open question, however. The challenge of answering this question has been taken here with the only human-ribosome-targeting small-molecule currently in clinical use, the antitumor plant alkaloid homoharringtonine (HHT). The approach consisted in dissecting HHT-nucleotide interaction energies from QM-MM simulations in explicit water. What emerged is a network of mostly weak interactions of the large, branched HHT with standard nucleotides and a single modified nucleotide, out of the four ones present at PCT\'s A site. This is unlike the case of the small, compact marine antitumor alkaloid agelastatin A, which displays only a few, albeit strong, interactions with site-A ribosome nucleotides. This should aid tailoring drugs targeting the ribosome.

This chapter aims to present some basic multiscale approaches available for enzyme simulations, and to point out practical details and pitfalls that are not often discussed in the literature, but can greatly influence the outcome of any in silico enzyme study. We cover principle methodological steps of multiscale studies of general enzyme reactions. This includes choice of starting structures, boundary conditions, potential energy surfaces, reaction coordinates, simulation methods, as well as the choice of method for the treatment of nuclear quantum effects. Together, these and additional steps are crucial for the success of enzyme-modeling projects and should be considered prior to embarking on multiscale modeling.

Biologically relevant interactions of piano-stool ruthenium(II) complexes with ds-DNA are studied in this article by hybrid quantum mechanics-molecular mechanics (QM/MM) computational technique. The whole reaction mechanism is divided into three phases: (i) hydration of the [Ru(II) (η(6) -benzene)(en)Cl](+) complex, (ii) monoadduct formation between the resulting aqua-Ru(II) complex and N7 position of one of the guanines in the ds-DNA oligomer, and (iii) formation of the intrastrand Ru(II) bridge (cross-link) between two adjacent guanines. Free energy profiles of all the reactions are explored by QM/MM MD umbrella sampling approach where the Ru(II) complex and two guanines represent a quantum core, which is described by density functional theory methods. The combined QM/MM scheme is realized by our own software, which was developed to couple several quantum chemical programs (in this study Gaussian 09) and Amber 11 package. Calculated free energy barriers of the both ruthenium hydration and Ru(II)-N7(G) DNA binding process are in good agreement with experimentally measured rate constants. Then, this method was used to study the possibility of cross-link formation. One feasible pathway leading to Ru(II) guanine-guanine cross-link with synchronous releasing of the benzene ligand is predicted. The cross-linking is an exergonic process with the energy barrier lower than for the monoadduct reaction of Ru(II) complex with ds-DNA.