Integrating HDX-MS Solution Data and Computational HDX Modeling to Refine Molecular Dynamics Ensembles
Probing the structural equilibrium that a protein and/or protein complex can adopt in solution is critical in understanding a multitude of its biophysical and biological properties. To that end, a variety of experimental and computational approaches are available to characterize a protein’s structural ensemble, each with their strengths and limitations. One such approach is Hydrogen Deuterium exchange coupled with mass spectrometry (HDX-MS). HDX-MS is a useful solution-based technique for capturing the structure and dynamics of the protein backbone amide groups in ever-increasing complex systems and in a time-resolved manner. However, the data obtained by HDX-MS remains largely limited to the experimentally attainable peptide-level resolution. On the other hand, computational approaches such as molecular dynamics (MD) simulations provide an atomistic resolution of the protein’s structure and dynamics; however, the results may deviate from experimental data due to an inadequate starting high-resolution structure/model, force field inadequacies, or insufficient sampling, often requiring experimental validation. In the Deredge lab, we aim to develop and apply integrative workflows which leverage experimental HDX-MS data to extract representative ensembles from computational MD simulations.
In our development efforts, our lab focuses on:
- The further development of Hydrogen-Deuterium Exchange Ensemble Reweighting (HDXer). HDXer is a maximum entropy reweighting approach that was developed reweigh MD simulations using HDX-MS data in a post-hoc manner. It leverages both techniques to produce a high-resolution set of conformations that best reflect protein dynamics in solution. We are using various model systems to test how different computational, experimental, and theoretical variables impact HDXer’s performance.
- The development of an HDX-MS guided enhanced sampling MD simulation approach using adaptive sampling principles that have been applied with other experimental approaches.
In our application efforts, our lab focuses on demonstrating that such integrative workflows can provide useful, mechanistically informative or hypothesis-generating information in the characterization of various biophysical phenomena. This includes extracting representative ensembles of a proteins in-solution native state ensemble, characterizing the ensemble of protein/protein interactions, protein/nucleic acid interactions and protein/small molecule ligand interactions or modeling missing density from high resolution structures obtained by crystallography or cryo-EM.
Insight into the Structure and Dynamics of the Dengue Nonstructural 5 (NS5) Protein
Flaviviruses are positive-sense, single-stranded RNA viruses which give rise to many of the mosquito-borne and tick-borne viral infections worldwide. Among them, dengue virus is the most prevalent mosquito-borne virus, with up to 400 million people getting infected with the virus annually. Dengue virus has four serotypes (DENV1-4) and humans who have been infected with one serotype, can be re-infected with another serotype leading to the more severe dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS). DHF and DSS are life-threatening conditions, characterized by vascular leakage, thrombocytopenia, and systemic shock, which can be deadly. To date, there are no antiviral drugs approved for the treatment of dengue, and there is no effective vaccine available to protect people from all four serotypes.
The non-structural protein 5 (NS5) is the largest and most conserved protein encoded by flaviviruses. It is a key component of the viral replication complex with multiple enzymatic and biological functions and is a major target for antivirals due to its critical functions. The dengue virus NS5 contains an N-terminal methyltransferase (MTase) domain, responsible for synthesis of the 5’ RNA cap and methylation, and a C-terminal RNA-dependent RNA polymerase (RdRp) domain, responsible for RNA synthesis. Although there are numerous crystal structures of dengue NS5 and the individual domains, there are no RNA-bound structures of the dengue NS5 available to date.
In the Deredge lab, we aim to characterize the different RNA interactions of the dengue NS5 in the context of the full-length, and its individual domains. We also aim to probe its structure and dynamics in its apo state, and RNA bound forms, to identify binding regions critical for its multiple interactions using a combination of biophysical techniques including surface plasmon resonance (SPR), analytical ultracentrifugation (AUC), and HDX-MS, etc. The long-term goal of the project is to expand on the exploration of the conformational landscape of dengue NS5 by employing a combination of enhanced sampling MD simulations and experimental-based ensemble refinement approaches for a computer-aided drug design strategy. Ultimately, it will strengthen drug discovery efforts geared towards NS5 as a therapeutic target against the dengue virus by providing novel structural information for structure-based drug design.
Optimizing protein therapeutics formulation
In Pharmaceutical Sciences, the formulation of therapeutic proteins is about determining the optimal conditions so that a given therapeutic protein is stable, active, often at high concentrations, during long term storage and until delivery to patient. Typically, it involves the tedious testing a combination of buffer conditions and excipients towards the empirical determination of the appropriate conditions to stabilize the native structure, prevent denaturation and protein/protein interaction that may lead to aggregation. This laborious process is often costly. In a joint project with Dr. Alexander MacKerell and Dr. Stephen Hoag’s groups, we are utilizing different computational, and formulation tools as well as HDX-MS to map the protein-protein interactions and protein-excipient interactions for various model protein therapeutics to predict the binding sites of proteins/excipients and offer better formulation condition of protein therapeutics. To that end, the Deredge lab is developing an HDX-MS based approach to quantitatively determine protein excipient interactions and the effect of excipient on local structure and dynamics as a function of excipient concentrations.
Join the Deredge Lab in one of these exciting projects!!!