GalaxyWater-wKGB: Prediction of Water Positions on Protein Structure Using wKGB Statistical Potential

HydraProt: A New Deep Learning Tool for Fast and Accurate Prediction of Water Molecule Positions for Protein Structures

Water molecules are integral to the structural stability of proteins and vital for facilitating molecular interactions. However, accurately predicting their precise position around protein structures remains a significant challenge, making it a vibrant research area. In this paper, we introduce HydraProt (deep Hydration of Proteins), a novel methodology for predicting precise positions of water molecule oxygen atoms around protein structures, leveraging two interconnected deep learning architectures: a 3D U-net and a Multi-Layer Perceptron (MLP). Our approach starts by introducing a coarse voxel-based representation of the protein, which allows for rapid sampling of candidate water positions via the 3D U-net. These water positions are then assessed by embedding the water-protein relationship in the Euclidean space by means of an MLP. Finally, a postprocessing step is applied to further refine the MLP predictions. HydraProt surpasses existing state-of-the-art approaches in terms of precision and recall and has been validated on large data sets of protein structures. Notably, our method offers rapid inference runtime and should constitute the method of choice for protein structure studies and drug discovery applications. Our pretrained models, data, and the source code required to reproduce these results are accessible at https://github.com/azamanos/HydraProt.

Mining the ZINC database of natural products for specific, testosterone-like, OXER1 antagonists

OXER1, the receptor for the oxidized arachidonic acid metabolite 5-oxo-ETE has been reported to play a significant role in inflammatory responses, being responsible for leucocyte chemotactic responses. Recently, we have identified OXER1 (GPR170) as a membrane receptor for androgens in prostate and breast cancer cells. Testosterone action via OXER1 induces specific Ca2+ release from intracellular organelles, modifies polymerized actin distribution induces apoptosis and decreases cancer cell migration. These actions are antagonized by 5-oxo-ETE. In addition, 5-oxo-ETE through a Gαi protein decreases cAMP, an action antagonized by testosterone. In this work, we mined the ZINC15 database, using QSAR, for natural compounds able to signal through Gαi and Gβγ simultaneously, mimicking testosterone actions, as well as for specific Gβγ interactors, inhibiting 5-oxo-ETE tumor promoting actions. We were able to identify four druggable Gαβγ and seven Gβγ specific OXER1 interactors. We further confirmed by bio-informatic methods their binding to the 5-oxo-ETE/testosterone binding groove of the receptor, their ADME properties and their possible interaction with other receptor and/or enzyme targets. Two compounds, ZINC04017374 (Naphthofluorescein) and ZINC08589130 (Puertogaline A) were purchased, tested in vitro and confirmed their OXER1 Gβγ and Gαβγ activity, respectively. The methodology followed is useful for a better understanding of the mechanism by which OXER1 mediates its actions, it has the potential to provide structural insights, in order to design small molecular specific interactors and ultimately design new anti-inflammatory and anti-cancer agents. Finally, the methodology may also be useful for identifying specific agonists/antagonists of other GPCRs.

HydraMap v.2: Prediction of Hydration Sites and Desolvation Energy with Refined Statistical Potentials

The complex network of water molecules within the binding pocket of a target protein undergoes alterations upon ligand binding, presenting a significant challenge for conventional molecular modeling methods to accurately characterize and compute the associated energy changes. We have previously developed an empirical method, HydraMap (J. Chem. Inf. Model. 2020, 60, 4359-4375), which employs statistical potentials to predict hydration sites and compute desolvation energy, achieving a reasonable balance between accuracy and speed. In this work, we present its improved version, namely, HydraMap v.2. We updated the statistical potentials for protein-water interactions through an analysis of 17 042 crystal protein structures. We also introduced a new feature to evaluate ligand-water interactions by incorporating statistical potentials derived from the solvated structures of 9878 small organic molecules produced by molecular dynamics simulations. By combining these potentials, HydraMap v.2 can predict and compare the hydration sites in a binding pocket before and after ligand binding, identifying key water molecules involved in the binding process, such as those forming bridging hydrogen bonds and unstable ones that can be replaced. We demonstrated the application of HydraMap v.2 in explaining the structure-activity relationship of a panel of MCL-1 inhibitors. The desolvation energies calculated by summing the energy change of each hydration site before and after ligand binding showed good correlation with known ligand binding affinities on six target proteins. In conclusion, HydraMap v.2 offers a cost-effective solution for estimating the desolvation energy during protein-ligand binding and also is practical in guiding lead optimization in structure-based drug discovery.

Lamin Filament Assembly Derived from the Atomic Structure of the Antiparallel Four-Helix Bundle

The nucleoskeletal protein lamin is primarily responsible for the mechanical stability of the nucleus. The lamin assembly process requires the A11, A22, and ACN binding modes of the coiled-coil dimers. Although X-ray crystallography and chemical cross-linking analysis of lamin A/C have provided snapshots of A11 and ACN binding modes, the assembly mechanism of the entire filament remains to be explained. Here, we report a crystal structure of a coil 2 fragment, revealing the A22 interaction at the atomic resolution. The structure showed detailed structural features, indicating that two coiled-coil dimers of the coil 2 subdomain are separated and then re-organized into the antiparallel-four-helix bundle. Furthermore, our findings suggest that the ACN binding mode between coil 1a and the C-terminal part of coil 2 when the A11 tetramers are arranged by the A22 interactions. We propose a full assembly model of lamin A/C with the curvature around the linkers, reconciling the discrepancy between the in situ and in vitro observations. Our model accounts for the balanced elasticity and stiffness of the nuclear envelopes, which is essential in protecting the cellular nucleus from external pressure.

GalaxyDock2-HEME: Protein-ligand docking for heme proteins

Prediction of protein-ligand binding poses is an essential component for understanding protein-ligand interactions and computer-aided drug design. Various proteins involve prosthetic groups such as heme for their functions, and adequate consideration of the prosthetic groups is vital for protein-ligand docking. Here, we extend the GalaxyDock2 protein-ligand docking algorithm to handle ligand docking to heme proteins. Docking to heme proteins involves increased complexity because the interaction of heme iron and ligand has covalent nature. GalaxyDock2-HEME, a new protein-ligand docking program for heme proteins, has been developed based on GalaxyDock2 by adding an orientation-dependent scoring term to describe heme iron-ligand coordination interaction. This new docking program performs better than other noncommercial docking programs such as EADock with MMBP, AutoDock Vina, PLANTS, LeDock, and GalaxyDock2 on a heme protein-ligand docking benchmark set in which ligands are known to bind iron. In addition, docking results on two other sets of heme protein-ligand complexes in which ligands do not bind iron show that GalaxyDock2-HEME does not have a high bias toward iron binding compared to other docking programs. This implies that the new docking program can distinguish iron binders from noniron binders for heme proteins.

Systematic Improvement of the Performance of Machine Learning Scoring Functions by Incorporating Features of Protein-Bound Water Molecules

Water molecules at the ligand-protein interfaces play crucial roles in the binding of the ligands, but the behavior of protein-bound water is largely ignored in many currently used machine learning (ML)-based scoring functions (SFs). In an attempt to improve the prediction performance of existing ML-based SFs, we estimated the water distribution with a HydraMap (HM) method and then incorporated the features extracted from protein-bound waters obtained in this way into three ML-based SFs: RF-Score, ECIF, and PLEC. It was found that a combination of HM-based features can consistently improve the performance of all three SFs, including their scoring, ranking, and docking power. HydraMap-based features show consistently good performance with both crystal structures and docked structures, demonstrating their robustness for SFs. Overall, HM-based features, which are a statistical representation of hydration sites at protein-ligand interfaces, are expected to improve the prediction performance for diverse SFs.

GalaxyWater-CNN: Prediction of Water Positions on the Protein Structure by a 3D-Convolutional Neural Network

Proteins interact with numerous water molecules to perform their physiological functions in biological organisms. Most water molecules act as solvent media; hence, their roles may be considered implicitly in theoretical treatments of protein structure and function. However, some water molecules interact intimately with proteins and require explicit treatment to understand their effects. Most physics-based computational methods are limited in their ability to accurately locate water molecules on protein surfaces because of inaccurate energy functions. Instead of relying on an energy function, this study attempts to learn the locations of water molecules from structural data. GalaxyWater-convolutional neural network (CNN) predicts water positions on protein chains, protein-protein interfaces, and protein-compound binding sites using a 3D-CNN model that is trained to generate a water score map on a given protein structure. The training data are compiled from high-resolution protein crystal structures resolved together with water molecules. GalaxyWater-CNN shows improved water prediction performance both in the coverage of crystal water molecules and in the accuracy of the predicted water positions when compared with previous energy-based methods. This method shows a superior performance in predicting water molecules that form hydrogen-bond networks precisely. The web service and the source code of this water prediction method are freely available at https://galaxy.seoklab.org/gwcnn and https://github.com/seoklab/GalaxyWater-CNN, respectively.

Prediction of Water Distributions and Displacement at Protein-Ligand Interfaces

The retention and displacement of water molecules during formation of ligand-protein interfaces play a major role in determining ligand binding. Understanding these effects requires a method for positioning of water molecules in the bound and unbound proteins and for defining water displacement upon ligand binding. We describe an algorithm for water placement and a calculation of ligand-driven water displacement in >9000 protein-ligand complexes. The algorithm predicts approximately 38% of experimental water positions within 1.0 Å and about 83% within 1.5 Å. We further show that the predicted water molecules can complete water networks not detected in crystallographic structures of the protein-ligand complexes. The algorithm was also applied to solvation of the corresponding unbound proteins, and this allowed calculation of water displacement upon ligand binding based on differences in the water network between the bound and unbound structures. We illustrate use of this approach through comparison of water displacement by structurally related ligands at the same binding site. This method for evaluation of water displacement upon ligand binding may be of value for prediction of the effects of ligand modification in drug design.

New Antagonists of the Membrane Androgen Receptor OXER1 from the ZINC Natural Product Database

OXER1 (oxoeicosanoid receptor 1) was deorphanized in 1993 and found to be the specific receptor for the arachidonic acid metabolite 5-oxo-ETE. Recently, we have reported that androgen binds to this receptor also, being a membrane androgen receptor, triggering a number of its membrane-mediated actions (cell migration, apoptosis, cell proliferation, Ca²⁺ movements). In addition, our previous work suggested that a number of natural monomeric and oligomeric polyphenols interact with OXER1, acting similar to testosterone. Here, we interrogated the natural product chemical space and identified nine polyphenolic molecules with interesting in silico pharmacological activities as putative OXER1 antagonists. The molecule with the best pharmacokinetic-pharmacodynamic properties (ZINC15959779) was purchased and tested on OXER1, in prostate cancer cell cultures. It showed that it has actions similar to those of testosterone in inhibiting cAMP, while it had no action in intracellular Ca²⁺ mobilization or actin cytoskeleton rearrangement/migration. These results are discussed under the prism of structure-activity relationships and in silico models of the OXER1 binding groove. We suggest that these compounds, together with the previously reported (poly)phenolic compounds, can be lead structures for the exploration of the anti-inflammatory and antiproliferative effects of OXER1 antagonists.

Accurate Prediction of Hydration Sites of Proteins Using Energy Model With Atom Embedding

We propose a method based on neural networks to accurately predict hydration sites in proteins. In our approach, high-quality data of protein structures are used to parametrize our neural network model, which is a differentiable score function that can evaluate an arbitrary position in 3D structures on proteins and predict the nearest water molecule that is not present. The score function is further integrated into our water placement algorithm to generate explicit hydration sites. In experiments on the OppA protein dataset used in previous studies and our selection of protein structures, our method achieves the highest model quality in terms of F1 score, compared to several previous studies.