Perspective on the Structural Basis for Human Aldo-Keto Reductase 1B10 Inhibition

Human aldo-keto reductase 1B10 (AKR1B10) is overexpressed in many cancer types and is involved in chemoresistance. This makes AKR1B10 to be an interesting drug target and thus many enzyme inhibitors have been investigated. High-resolution crystallographic structures of AKR1B10 with various reversible inhibitors were deeply analyzed and compared to those of analogous complexes with aldose reductase (AR). In both enzymes, the active site included an anion-binding pocket and, in some cases, inhibitor binding caused the opening of a transient specificity pocket. Different structural conformers were revealed upon inhibitor binding, emphasizing the importance of the highly variable loops, which participate in the transient opening of additional binding subpockets. Two key differences between AKR1B10 and AR were observed regarding the role of external loops in inhibitor binding. The first corresponded to the alternative conformation of Trp112 (Trp111 in AR). The second difference dealt with loop A mobility, which defined a larger and more loosely packed subpocket in AKR1B10. From this analysis, the general features that a selective AKR1B10 inhibitor should comply with are the following: an anchoring moiety to the anion-binding pocket, keeping Trp112 in its native conformation (AKR1B10-like), and not opening the specificity pocket in AR.

Flexible pivoting of dynamin PH-domain catalyzes fission: Insights into molecular degrees of freedom

The neuronal dynamin1 functions in the release of synaptic vesicles by orchestrating the process of GTPase dependent membrane fission. Dynamin1 associates with the plasma membrane-localized phosphatidylinositol-4,5-bisphosphate (PIP2) through the centrally-located pleckstrin homology domain (PHD). The PHD is dispensable as fission (in model membranes) can be managed, even when the PHD-PIP2 interaction is replaced by a generic polyhistidine- or polylysine-lipid interaction. However, the absence of the PHD renders a dramatic dampening of the rate of fission. These observations suggest that the PHD-PIP2 containing membrane interaction could have evolved to expedite fission to fulfill the requirement of rapid kinetics of synaptic vesicle recycling. Here, we use a suite of multiscale modeling approaches to explore PHD-membrane interactions. Our results reveal that (a) the binding of PHD to PIP2-containing membranes modulates the lipids towards fission-favoring conformations and softens the membrane, and (b) PHD associates with membrane in multiple orientations using variable loops as pivots. We identify a new loop (VL4), which acts as an auxiliary pivot and modulates the orientation flexibility of PHD on the membrane — a mechanism we believe may be important for high fidelity dynamin collar assembly. Together, these insights provide a molecular-level understanding of the catalytic role of PHD in dynamin-mediated membrane fission. [Media: see text] [Media: see text] [Media: see text] [Media: see text]

Geometric potentials from deep learning improve prediction of CDR H3 loop structures

Motivation Antibody structure is largely conserved, except for a complementarity-determining region featuring six variable loops. Five of these loops adopt canonical folds which can typically be predicted with existing methods, while the remaining loop (CDR H3) remains a challenge due to its highly diverse set of observed conformations. In recent years, deep neural networks have proven to be effective at capturing the complex patterns of protein structure. This work proposes DeepH3, a deep residual neural network that learns to predict inter-residue distances and orientations from antibody heavy and light chain sequence. The output of DeepH3 is a set of probability distributions over distances and orientation angles between pairs of residues. These distributions are converted to geometric potentials and used to discriminate between decoy structures produced by RosettaAntibody and predict new CDR H3 loop structures de novo. Results When evaluated on the Rosetta antibody benchmark dataset of 49 targets, DeepH3-predicted potentials identified better, same and worse structures [measured by root-mean-squared distance (RMSD) from the experimental CDR H3 loop structure] than the standard Rosetta energy function for 33, 6 and 10 targets, respectively, and improved the average RMSD of predictions by 32.1% (1.4 Å). Analysis of individual geometric potentials revealed that inter-residue orientations were more effective than inter-residue distances for discriminating near-native CDR H3 loops. When applied to de novo prediction of CDR H3 loop structures, DeepH3 achieves an average RMSD of 2.2 ± 1.1 Å on the Rosetta antibody benchmark. Availability and Implementation DeepH3 source code and pre-trained model parameters are freely available at https://github.com/Graylab/deepH3-distances-orientations. Supplementary information Supplementary data are available at Bioinformatics online.

Geometric Potentials from Deep Learning Improve Prediction of CDR H3 Loop Structures

Antibody structure is largely conserved, except for a complementarity-determining region featuring six variable loops. Five of these loops adopt canonical folds which can typically be predicted with existing methods, while the remaining loop (CDR H3) remains a challenge due to its highly diverse set of observed conformations. In recent years, deep neural networks have proven to be effective at capturing the complex patterns of protein structure. This work proposes DeepH3, a deep residual neural network that learns to predict inter-residue distances and orientations from antibody heavy and light chain sequence. The output of DeepH3 is a set of probability distributions over distances and orientation angles between pairs of residues. These distributions are converted to geometric potentials and used to discriminate between decoy structures produced by RosettaAntibody. When evaluated on the Rosetta Antibody Benchmark dataset of 49 targets, DeepH3-predicted potentials identified better, same, and worse structures (measured by root-mean-squared distance [RMSD] from the experimental CDR H3 loop structure) than the standard Rosetta energy function for 30, 13, and 6 targets, respectively, and improved the average RMSD of predictions by 21.3% (0.48 Å). Analysis of individual geometric potentials revealed that inter-residue orientations were more effective than inter-residue distances for discriminating near-native CDR H3 loop structures.

Charge and hydrophobicity are key features in sequence-trained machine learning models for predicting the biophysical properties of clinical-stage antibodies

Improved understanding of properties that mediate protein solubility and resistance to aggregation are important for developing biopharmaceuticals, and more generally in biotechnology and synthetic biology. Recent acquisition of large datasets for antibody biophysical properties enables the search for predictive models. In this report, machine learning methods are used to derive models for 12 biophysical properties. A physicochemical perspective is maintained in analysing the models, leading to the observation that models cluster largely according to charge (cross-interaction measurements) and hydrophobicity (self-interaction methods). These two properties also overlap in some cases, for example in a new interpretation of variation in hydrophobic interaction chromatography. Since the models are developed from differences of antibody variable loops, the next stage is to extend models to more diverse protein sets. Availability The web application for the sequence-based algorithms are available on the protein-sol webserver, at https://protein-sol.manchester.ac.uk/abpred, with models and virtualisation software available at https://protein-sol.manchester.ac.uk/software.

The structural basis for RNA selectivity by the IMP family of RNA-binding proteins

The IGF2 mRNA-binding proteins (ZBP1/IMP1, IMP2, IMP3) are highly conserved post-transcriptional regulators of RNA stability, localization and translation. They play important roles in cell migration, neural development, metabolism and cancer cell survival. The knockout phenotypes of individual IMP proteins suggest that each family member regulates a unique pool of RNAs, yet evidence and an underlying mechanism for this is lacking. Here, we combine systematic evolution of ligands by exponential enrichment (SELEX) and NMR spectroscopy to demonstrate that the major RNA-binding domains of the two most distantly related IMPs (ZBP1 and IMP2) bind to different consensus sequences and regulate targets consistent with their knockout phenotypes and roles in disease. We find that the targeting specificity of each IMP is determined by few amino acids in their variable loops. As variable loops often differ amongst KH domain paralogs, we hypothesize that this is a general mechanism for evolving specificity and regulation of the transcriptome.

Charge and hydrophobicity are key features in sequence-trained machine learning models for predicting the biophysical properties of clinical-stage antibodies

Improved understanding of properties that mediate protein solubility and resistance to aggregation are important for developing biopharmaceuticals, and more generally in biotechnology and synthetic biology. Recent acquisition of large datasets for antibody biophysical properties enables the search for predictive models. In this report, machine learning methods are used to derive models for 12 biophysical properties. A physicochemical perspective is maintained in analysing the models, leading to the observation that models cluster largely according to charge (cross-interaction measurements) and hydrophobicity (self-interaction methods). These two properties also overlap in some cases, for example in a new interpretation of variation in hydrophobic interaction chromatography. Since the models are developed from differences of antibody variable loops, the next stage is to extend models to more diverse protein sets.AvailabilityThe web application for the sequence based algorithms are available on the protein-sol webserver, at https://protein-sol.manchester.ac.uk/abpred, with models and virtualisation software available at https://protein-sol.manchester.ac.uk/software.

The structural basis for RNA selectivity by the IMP family of RNA binding proteins

The Igf2 mRNA binding proteins (ZBP1/IMP1, IMP2, IMP3) are highly conserved post-transcriptional regulators of RNA stability, localization and translation. They play important roles in cell migration, neural development, metabolism and cancer cell survival. The knockout phenotypes of individual IMP proteins suggest that each family member regulates a unique pool of RNAs, yet evidence and an underlying mechanism for this is lacking. Here, we combine SELEX and NMR spectroscopy to demonstrate that the major RNA binding domains of the two most distantly related IMPs (ZBP1 and IMP2) bind to different consensus sequences and regulate targets consistent with their knockout phenotypes and roles in disease. We find that the targeting specificity of each IMP is determined by few amino acids in their variable loops. As variable loops often differ amongst KH domain paralogs, we hypothesize that this is a general mechanism for evolving specificity and regulation of the transcriptome.