Peter Poliak

This tutorial describes the practical use of some recent methodological advances implemented in the GROMOS software for biomolecular simulations. It is envisioned as a living document, with additional tutorials being added in the course of time. Currently, it consists of six distinct tutorials. The first tutorial describes the use of time-averaged restraints to enforce agreement with order parameters derived from NMR experiments. The second tutorial describes the use of extended thermodynamic integration in the double-decoupling method to compute the affinity of a small molecule to a protein. The molecule involved bears a negative charge, necessitating the application of post-simulation corrections. The third tutorial is based on the same molecular system, but computes the binding free energy from a path-sampling method with distance-field distance restraints and Hamiltonian replica exchange simulations. The fourth tutorial describes the use of Gaussian accelerated MD, an enhanced sampling technique. The fifth tutorial is about the use of Accelerated Enveloping Distribution Sampling (AEDS) an enhanced version of EDS. The sixth tutorial illustrates how to run QM/MM simulations with the Buffer Region Neural Network (BuRNN) method. The tutorials are written for users with some experience in the application of molecular dynamics simulations.

The integration of quantum mechanics and molecular mechanics (QM/MM) within molecular dynamics simulations is crucial to accurately model complex biochemical systems. Here, we present an enhanced implementation of the QM/MM interface in the GROMOS simulation package, introducing significant improvements in functionality and user control. We present new features, including the link atom scheme, which allows the modeling of QM regions as a part of bigger molecules. Benchmark tests on various systems, including QM water in water, amino acids in water, and tripeptides validate the reliability of the new functionalities. Performance evaluations demonstrate that the updated implementation is efficient, with the primary computational burden attributed to the QM program rather than the QM/MM interface or the MD program itself. The improved QM/MM interface enables more advanced investigations into biomolecular reactivity, enzyme catalysis, and other phenomena requiring detailed quantum mechanical treatment within classical simulations. This work represents a significant advancement in the capabilities of GROMOS, providing enhanced tools to explore complex molecular systems.

Lytic polysaccharide monooxygenases (LPMOs) are widely distributed in fungi, and catalyze the oxidative degradation of polysaccharides such as cellulose. Despite their name, LPMOs possess a dominant peroxygenase activity that is reflected in high turnover numbers but also causes deactivation. We report on the influence of small molecules and ions on the activity and stability of LPMO during catalysis. Turbidimetric and photometric assays were used to identify LPMO inhibitors and measure their inhibitory effect. Selected inhibitors were employed to study LPMO activity and stability during cellulose depolymerization by HPLC and turbidimetry. It was found that the fungal metabolic products oxalic acid and citric acid strongly reduce LPMO activity, but also protect the enzyme from deactivation. QM calculations showed that the copper atom in the catalytic site could be ligated by bi- or tridentate chelating compounds, which replace two water molecules. MD simulations and QM calculations show that the most likely inhibition pattern is the competition between the inhibitor and reducing agent in the oxidized Cu(II) state. A correlation between the complexation energy and the IC₅₀ values demonstrates that small, bidentate molecules interact strongest with the catalytic site copper and could be used by the fungus as physiological effectors to regulate LPMO activity.

A series of polynitroxide amides possessing 2,2,5,5-tetramethyl-1-pyrrolidinyloxy (PROXYL) and/or 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) units connected through various bridges were synthesized and their properties were analyzed. EPR spectroscopy provided detailed insight into their paramagnetic character and related properties. A thorough examination of the EPR spectra of dinitroxides in organic solvents provided valuable information on the intramolecular motions, thermodynamics, and spin-exchange mechanisms. Analysis of low-temperature X- and Q-band EPR spectra of the dissolved dinitroxides provided spin–spin distances that were comparable with the theoretical values obtained by DFT. Cyclic voltammetry investigations revealed (quasi)reversible electrochemical behavior for PROXYL-derived biradicals, whereas significant loss of the reversibility was found for TEMPO-containing bi- and polyradicals. The inhibitory activities of the nitroxides against model bacteria, yeasts, and filamentous fungi were assessed.