Construct a variable-length fragment library for de novo protein structure prediction

Although remarkable achievements, such as AlphaFold2, have been made in end-to-end structure prediction, fragment libraries remain essential for de novo protein structure prediction, which can help explore and understand the protein-folding mechanism. In this work, we developed a variable-length fragment library (VFlib). In VFlib, a master structure database was first constructed from the Protein Data Bank through sequence clustering. The Hidden Markov Model (HMM) profile of each protein in the master structure database was generated by HHsuite, and the secondary structure of each protein was calculated by DSSP. For the query sequence, the HMM-profile was first constructed. Then, variable-length fragments were retrieved from the master structure database through dynamically variable-length profile-profile comparison. A complete method for chopping the query HMM-profile during this process was proposed to obtain fragments with increased diversity. Finally, secondary structure information was used to further screen the retrieved fragments to generate the final fragment library of specific query sequence. The experimental results obtained with a set of 120 nonredundant proteins showed that the global precision and coverage of the fragment library generated by VFlib were 55.04% and 94.95% at the RMSD cutoff of 1.5 Å, respectively. Compared to the benchmark method of NNMake, the global precision of our fragment library had increased by 62.89% with equivalent coverage. Furthermore, the fragments generated by VFlib and NNMake were used to predict structure models through fragment assembly. Controlled experimental results demonstrated that the average TM-score of VFlib was 16.00% higher than that of NNMake.

Efficient Generative Modelling of Protein Structure Fragments using a Deep Markov Model

Fragment libraries are often used in protein structure prediction, simulation and design as a means to significantly reduce the vast conformational search space. Current state-of-the-art methods for fragment library generation do not properly account for aleatory and epistemic uncertainty, respectively due to the dynamic nature of proteins and experimental errors in protein structures. Additionally, they typically rely on information that is not generally or readily available, such as homologous sequences, related protein structures and other complementary information. To address these issues, we developed BIFROST, a novel take on the fragment library problem based on a Deep Markov Model architecture combined with directional statistics for angular degrees of freedom, implemented in the deep probabilistic programming language Pyro. BIFROST is a probabilistic, generative model of the protein backbone dihedral angles conditioned solely on the amino acid sequence. BIFROST generates fragment libraries with a quality on par with current state-of-the-art methods at a fraction of the run-time, while requiring considerably less information and allowing efficient evaluation of probabilities.

Towards chemical validation of Leishmania infantum ribose 5-phosphate isomerase as a drug target

Neglected tropical diseases caused by kinetoplastid parasites (Trypanosoma brucei, Trypanosoma cruzi and Leishmania spp.) place a significant health and economic burden on developing nations worldwide. Current therapies are largely out-dated, inadequate and facing mounting drug resistance from the causative parasites. Thus, there is an urgent need for drug discovery and development. Target-led drug discovery approaches have focused on the identification of parasite enzymes catalysing essential biochemical processes, which significantly differ from equivalent proteins found in humans, thereby providing potentially exploitable therapeutic windows. One such target is ribose 5-phosphate isomerase B (RpiB), an enzyme involved in the non-oxidative branch of the pentose phosphate pathway, which catalyses the inter-conversion of D-ribose 5-phosphate and D-ribulose 5-phosphate. Although protozoan RpiB has been the focus of numerous targeted studies, compounds capable of selectively inhibiting this parasite enzyme have not been identified. Here, we present the results of a fragment library screening against Leishmania infantum RpiB, performed using thermal shift analysis. Hit fragments were shown to be effective inhibitors of LiRpiB in activity assays, and several were capable of selectively inhibiting parasite growth in vitro. These results support the identification of LiRpiB as a validated therapeutic target. The X-ray crystal structure of apo LiRpiB was also solved, permitting docking studies to assess how hit fragments might interact with LiRpiB to inhibit its activity. Overall, this work will guide structure-based development of LiRpiB inhibitors as anti-leishmanial agents.

Insight into Functional Membrane Proteins by Solution NMR: The Human Bcl-2 Protein—A Promising Cancer Drug Target

Evasion from programmed cell death (apoptosis) is the main hallmark of cancer and a major cause of resistance to therapy. Many tumors simply ensure survival by over-expressing the cell-protecting (anti-apoptotic) Bcl-2 membrane protein involved in apoptotic regulation. However, the molecular mechanism by which Bcl-2 protein in its mitochondrial outer membrane location protects cells remains elusive due to the absence of structural insight; and current strategies to therapeutically interfere with these Bcl-2 sensitive cancers are limited. Here, we present an NMR-based approach to enable structural insight into Bcl-2 function; an approach also ideal as a fragment-based drug discovery platform for further identification and development of promising molecular Bcl-2 inhibitors. By using solution NMR spectroscopy on fully functional intact human Bcl-2 protein in a membrane-mimicking micellar environment, and constructs with specific functions remaining, we present a strategy for structure determination and specific drug screening of functional subunits of the Bcl-2 protein as targets. Using 19F NMR and a specific fragment library (Bionet) with fluorinated compounds we can successfully identify various binders and validate our strategy in the hunt for novel Bcl-2 selective cancer drug strategies to treat currently incurable Bcl-2 sensitive tumors.

Pyrimidone inhibitors targeting Chikungunya Virus nsP3 macrodomain by fragment-based drug design

The macrodomain of nsP3 (nsP3MD) is highly conserved among the alphaviruses and ADP-ribosylhydrolase activity of Chikungunya Virus (CHIKV) nsP3MD is critical for CHIKV viral replication and virulence. No small molecule drugs targeting CHIKV nsP3 have been identified to date. Here we report small fragments that bind to nsP3MD which were discovered by virtually screening a fragment library and X-ray crystallography. These identified fragments share a similar scaffold, 2-pyrimidone-4-carboxylic acid, and are specifically bound to the ADP-ribose binding site of nsP3MD. Among the fragments, 2-oxo-5,6-benzopyrimidine-4-carboxylic acid showed anti-CHIKV activity with an IC50 of 23 μM. Our fragment-based drug discovery approach provides valuable information to further develop a specific and potent nsP3 inhibitor of CHIKV viral replication based on the 2-pyrimidone-4-carboxylic acid scaffold. In silico studies suggest this pyrimidone scaffold could also bind to the macrodomains of other alphaviruses and coronaviruses and thus, have potential pan-antiviral activity.

Applications of Solution NMR in Drug Discovery

During the past decades, solution nuclear magnetic resonance (NMR) spectroscopy has demonstrated itself as a promising tool in drug discovery. Especially, fragment-based drug discovery (FBDD) has benefited a lot from the NMR development. Multiple candidate compounds and FDA-approved drugs derived from FBDD have been developed with the assistance of NMR techniques. NMR has broad applications in different stages of the FBDD process, which includes fragment library construction, hit generation and validation, hit-to-lead optimization and working mechanism elucidation, etc. In this manuscript, we reviewed the current progresses of NMR applications in fragment-based drug discovery, which were illustrated by multiple reported cases. Moreover, the NMR applications in protein-protein interaction (PPI) modulators development and the progress of in-cell NMR for drug discovery were also briefly summarized.

Discovery of small molecules targeting the tandem tudor domain of the epigenetic factor UHRF1 using fragment-based ligand discovery

Despite the established roles of the epigenetic factor UHRF1 in oncogenesis, no UHRF1-targeting therapeutics have been reported to date. In this study, we use fragment-based ligand discovery to identify novel scaffolds for targeting the isolated UHRF1 tandem Tudor domain (TTD), which recognizes the heterochromatin-associated histone mark H3K9me3 and supports intramolecular contacts with other regions of UHRF1. Using both binding-based and function-based screens of a ~ 2300-fragment library in parallel, we identified 2,4-lutidine as a hit for follow-up NMR and X-ray crystallography studies. Unlike previous reported ligands, 2,4-lutidine binds to two binding pockets that are in close proximity on TTD and so has the potential to be evolved into more potent inhibitors using a fragment-linking strategy. Our study provides a useful starting point for developing potent chemical probes against UHRF1.

Systematic optimization of a fragment-based force field against experimental pure-liquid properties considering large compound families: Application to oxygen and nitrogen compounds.

The CombiFF approach is a workflow for the automated refinement of force-field parameters against experimental condensed-phase data, considering entire classes of organic molecules constructed using a fragment library via combinatorial...