Structural and Molecular Dynamics of Mycobacterium tuberculosis Malic Enzyme, a Potential Anti-TB Drug Target.

Tuberculosis (TB) is the most lethal bacterial infectious disease worldwide. It is notoriously difficult to treat, requiring a cocktail of antibiotics administered over many months. The dense, waxy outer membrane of the TB-causing agent, Mycobacterium tuberculosis (Mtb), acts as a formidable barrier against uptake of antibiotics. Subsequently, enzymes involved in maintaining the integrity of the Mtb cell wall are promising drug targets. Recently, we demonstrated that Mtb lacking malic enzyme (MEZ) has altered cell wall lipid composition and attenuated uptake by macrophages. These results suggest that MEZ contributes to lipid biosynthesis by providing reductants in the form of NAD(P)H. Here, we present the X-ray crystal structure of MEZ to 3.6 Å. We use biochemical assays to demonstrate MEZ is dimeric in solution and to evaluate the effects of pH and allosteric regulators on its kinetics and thermal stability. To assess the interactions between MEZ and its substrate malate and cofactors, Mn2+ and NAD(P)+, we ran a series of molecular dynamics (MD) simulations. First, the MD analysis corroborates our empirical observations that MEZ is unusually flexible, which persists even with the addition of substrate and cofactors. Second, the MD simulations reveal that dimeric MEZ subunits alternate between open and closed states, and that MEZ can stably bind its NAD(P)+ cofactor in multiple conformations, including an inactive, compact NAD+ form. Together the structure of MEZ and insights from its dynamics can be harnessed to inform the design of MEZ inhibitors that target Mtb and not human malic enzyme homologues.

Structure of a Mycobacterium tuberculosis Heme-Degrading Protein, MhuD, Variant in Complex with Its Product.

Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis, requires iron for survival. In Mtb, MhuD is the cytosolic protein that degrades imported heme. MhuD is distinct, in both sequence and structure, from canonical heme oxygenases (HOs) but homologous with IsdG-type proteins. Canonical HO is found mainly in eukaryotes, while IsdG-type proteins are predominantly found in prokaryotes, including pathogens. While there are several published structures of MhuD and other IsdG-type proteins in complex with the heme substrate, no structures of IsdG-type proteins in complex with a product have been reported, unlike the case for HOs. We recently showed that the Mtb variant MhuD-R26S produces biliverdin IXα (αBV) rather than the wild-type mycobilin isomers. Given that mycobilin and other IsdG-type protein products like staphylobilin are difficult to isolate in quantities sufficient for structure determination, here we use the MhuD-R26S variant and its product αBV as a proxy to study the IsdG-type protein-product complex. First, we show that αBV has a nanomolar affinity for MhuD and the R26S variant. Second, we determined the MhuD-R26S-αBV complex structure to 2.5 Å, which reveals two notable features: (1) two αBV molecules bound per active site and (2) a novel α-helix (α3) that was not observed in previous MhuD-heme structures. Finally, through molecular dynamics simulations, we show that α3 is stable with the proximal αBV alone. MhuD's high affinity for the product and the observed structural and electrostatic changes that accompany substrate turnover suggest that there may be an unidentified class of proteins that are responsible for the extraction of products from MhuD and other IsdG-type proteins.

Enhancing Side Chain Rotamer Sampling Using Nonequilibrium Candidate Monte Carlo.

Molecular simulations are a valuable tool for studying biomolecular motions and thermodynamics. However, such motions can be slow compared to simulation time scales, yet critical. Specifically, adequate sampling of side chain motions in protein binding pockets is crucial for obtaining accurate estimates of ligand binding free energies from molecular simulations. The time scale of side chain rotamer flips can range from a few ps to several hundred ns or longer, particularly in crowded environments like the interior of proteins. Here, we apply a mixed nonequilibrium candidate Monte Carlo (NCMC)/molecular dynamics (MD) method to enhance sampling of side chain rotamers. The NCMC portion of our method applies a switching protocol wherein the steric and electrostatic interactions between target side chain atoms and the surrounding environment are cycled off and then back on during the course of a move proposal. Between NCMC move proposals, simulation of the system continues via traditional molecular dynamics. Here, we first validate this approach on a simple, solvated valine-alanine dipeptide system and then apply it to a well-studied model ligand binding site in T4 lysozyme L99A. We compute the rate of rotamer transitions for a valine side chain using our approach and compare it to that of traditional molecular dynamics simulations. Here, we show that our NCMC/MD method substantially enhances side chain sampling, especially in systems where the torsional barrier to rotation is high (≥10 kcal/mol). These barriers can be intrinsic torsional barriers or steric barriers imposed by the environment. Overall, this may provide a promising strategy to selectively improve side chain sampling in molecular simulations.

Introducing the new bacterial branch of the RNase A superfamily.

Bovine pancreatic ribonuclease (RNase A) is the founding member of the RNase A superfamily. Members of this superfamily of ribonucleases have high sequence diversity, but possess a similar structural fold, together with a conserved His-Lys-His catalytic triad and structural disulfide bonds. Until recently, RNase A proteins had exclusively been identified in eukaryotes within vertebrae. Here, we discuss the discovery by Batot et al. of a bacterial RNase A superfamily member, CdiA-CTYkris: a toxin that belongs to an inter-bacterial competition system from Yersinia kristensenii. CdiA-CTYkris exhibits the same structural fold as conventional RNase A family members and displays in vitro and in vivo ribonuclease activity. However, CdiA-CTYkris shares little to no sequence similarity with RNase A, and lacks the conserved disulfide bonds and catalytic triad of RNase A. Interestingly, the CdiA-CTYkris active site more closely resembles the active site composition of various eukaryotic endonucleases. Despite lacking sequence similarity to eukaryotic RNase A family members, CdiA-CTYkris does share high sequence similarity with numerous Gram-negative and Gram-positive bacterial proteins/domains. Nearly all of these bacterial homologs are toxins associated with virulence and bacterial competition, suggesting that the RNase A superfamily has a distinct bacterial subfamily branch, which likely arose by way of convergent evolution. Finally, RNase A interacts directly with the immunity protein of CdiA-CTYkris, thus the cognate immunity protein for the bacterial RNase A member could be engineered as a new eukaryotic RNase A inhibitor.

Blind prediction of cyclohexane-water distribution coefficients from the SAMPL5 challenge.

In the recent SAMPL5 challenge, participants submitted predictions for cyclohexane/water distribution coefficients for a set of 53 small molecules. Distribution coefficients (log D) replace the hydration free energies that were a central part of the past five SAMPL challenges. A wide variety of computational methods were represented by the 76 submissions from 18 participating groups. Here, we analyze submissions by a variety of error metrics and provide details for a number of reference calculations we performed. As in the SAMPL4 challenge, we assessed the ability of participants to evaluate not just their statistical uncertainty, but their model uncertainty-how well they can predict the magnitude of their model or force field error for specific predictions. Unfortunately, this remains an area where prediction and analysis need improvement. In SAMPL4 the top performing submissions achieved a root-mean-squared error (RMSE) around 1.5 kcal/mol. If we anticipate accuracy in log D predictions to be similar to the hydration free energy predictions in SAMPL4, the expected error here would be around 1.54 log units. Only a few submissions had an RMSE below 2.5 log units in their predicted log D values. However, distribution coefficients introduced complexities not present in past SAMPL challenges, including tautomer enumeration, that are likely to be important in predicting biomolecular properties of interest to drug discovery, therefore some decrease in accuracy would be expected. Overall, the SAMPL5 distribution coefficient challenge provided great insight into the importance of modeling a variety of physical effects. We believe these types of measurements will be a promising source of data for future blind challenges, especially in view of the relatively straightforward nature of the experiments and the level of insight provided.