Search for: All records

Molecular dynamics (MD) is the primary computational method by which modern structural biology explores macromolecule structure and function. Boltzmann generators have been proposed as an alternative to MD, by replacing the integration of molecular systems over time with the training of generative neural networks. This neural network approach to MD enables convergence to thermodynamic equilibrium faster than traditional MD; however, critical gaps in the theory and computational feasibility of Boltzmann generators significantly reduce their usability. Here, we develop a mathematical foundation to overcome these barriers; we demonstrate that the Boltzmann generator approach is sufficiently rapid to replace traditional MD for complex macromolecules, such as proteins in specific applications, and we provide a comprehensive toolkit for the exploration of molecular energy landscapes with neural networks.

 

The ability of cells and tissues to respond differentially to mechanical forces applied in distinct directions is mediated by the ability of load-bearing proteins to preferentially maintain physical linkages in certain directions. However, the molecular basis and biological consequences of directional force-sensitive binding remain unclear. Vinculin (Vcn) is a load-bearing linker protein that exhibits directional catch bonding due to interactions between the Vcn tail domain (Vt) and filamentous (F)-actin. We developed a computational approach to predict Vcn residues involved in directional catch bonding and produced a set of associated Vcn variants with unaltered Vt structure, actin binding, or phospholipid interactions. Incorporation of the variants did not affect Vcn activation but reduced Vcn loading and altered exchange dynamics, consistent with the loss of directional catch bonding. Expression of Vcn variants perturbed the coordination of subcellular structures and cell migration, establishing key cellular functions for Vcn directional catch bonding.

 

The stress response protein regulated in development and DNA damage response 1 (REDD1) has been implicated in visual deficits in patients with diabetes. The aim here was to investigate the mechanism responsible for the increase in retinal REDD1 protein content that is observed with diabetes. We found that REDD1 protein expression was increased in the retina of streptozotocin-induced diabetic mice in the absence of a change in REDD1 mRNA abundance or ribosome association. Oral antioxidant supplementation reduced retinal oxidative stress and suppressed REDD1 protein expression in the retina of diabetic mice. In human retinal Müller cell cultures, hyperglycemic conditions increased oxidative stress, enhanced REDD1 expression, and inhibited REDD1 degradation independently of the proteasome. Hyperglycemic conditions promoted a redox-sensitive cross-strand disulfide bond in REDD1 at C150/C157 that was required for reduced REDD1 degradation. Discrete molecular dynamics simulations of REDD1 structure revealed allosteric regulation of a degron upon formation of the disulfide bond that disrupted lysosomal proteolysis of REDD1. REDD1 acetylation at K129 was required for REDD1 recognition by the cytosolic chaperone HSC70 and degradation by chaperone-mediated autophagy. Disruption of REDD1 allostery upon C150/C157 disulfide bond formation prevented the suppressive effect of hyperglycemic conditions on REDD1 degradation and reduced oxidative stress in cells exposed to hyperglycemic conditions. The results reveal redox regulation of REDD1 and demonstrate the role of a REDD1 disulfide switch in development of oxidative stress. 

Virtual screening is a cost- and time-effective alternative to traditional high-throughput screening in the drug discovery process. Both virtual screening approaches, structure-based molecular docking and ligand-based cheminformatics, suffer from computational cost, low accuracy, and/or reliance on prior knowledge of a ligand that binds to a given target. Here, we propose a neural network framework, NeuralDock, which accelerates the process of high-quality computational docking by a factor of 10 6 , and does not require prior knowledge of a ligand that binds to a given target. By approximating both protein-small molecule conformational sampling and energy-based scoring, NeuralDock accurately predicts the binding energy, and affinity of a protein-small molecule pair, based on protein pocket 3D structure and small molecule topology. We use NeuralDock and 25 GPUs to dock 937 million molecules from the ZINC database against superoxide dismutase-1 in 21 h, which we validate with physical docking using MedusaDock. Due to its speed and accuracy, NeuralDock may be useful in brute-force virtual screening of massive chemical libraries and training of generative drug models. 

Small-molecule inhibitors of PD-L1 are postulated to control immune evasion in tumors similar to antibodies that target the PD-L1/PD-1 immune checkpoint axis. However, the identity of targetable PD-L1 inducers is required to develop small-molecule PD-L1 inhibitors. In this study, using chromatin immunoprecipitation (ChIP) assay and siRNA, we demonstrate that vitamin D/VDR regulates PD-L1 expression in acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) cells. We have examined whether a VDR antagonist, MeTC7, can inhibit PD-L1. To ensure that MeTC7 inhibits VDR/PD-L1 without off-target effects, we examined competitive inhibition of VDR by MeTC7, utilizing ligand-dependent dimerization of VDR-RXR, RXR-RXR, and VDR-coactivators in a mammalian 2-hybrid (M2H) assay. MeTC7 inhibits VDR selectively, suppresses PD-L1 expression sparing PD-L2, and inhibits the cell viability, clonogenicity, and xenograft growth of AML cells. MeTC7 blocks AML/mesenchymal stem cells (MSCs) adhesion and increases the efferocytotic efficiency of THP-1 AML cells. Additionally, utilizing a syngeneic colorectal cancer model in which VDR/PD-L1 co-upregulation occurs in vivo under radiation therapy (RT), MeTC7 inhibits PD-L1 and enhances intra-tumoral CD8+T cells expressing lymphoid activation antigen-CD69. Taken together, MeTC7 is a promising small-molecule inhibitor of PD-L1 with clinical potential. 

Protein aggregation results in an array of different size soluble oligomers and larger insoluble fibrils. Insoluble fibrils were originally thought to cause neuronal cell deaths in neurodegenerative diseases due to their prevalence in tissue samples and disease models. Despite recent studies demonstrating the toxicity associated with soluble oligomers, many therapeutic strategies still focus on fibrils or consider all types of aggregates as one group. Oligomers and fibrils require different modeling and therapeutic strategies, targeting the toxic species is crucial for successful study and therapeutic development. Here, we review the role of different‐size aggregates in disease, and how factors contributing to aggregation (mutations, metals, post‐translational modifications, and lipid interactions) may promote oligomers opposed to fibrils. We review two different computational modeling strategies (molecular dynamics and kinetic modeling) and how they are used to model both oligomers and fibrils. Finally, we outline the current therapeutic strategies targeting aggregating proteins and their strengths and weaknesses for targeting oligomers versus fibrils. Altogether, we aim to highlight the importance of distinguishing the difference between oligomers and fibrils and determining which species is toxic when modeling and creating therapeutics for protein aggregation in disease.