Biology subjects

Van Lehn, R. C.

Publications and source records attributed to Van Lehn, R. C..

4 recordsLinked to original sources

bioRxiv2026

Scalable Extraction of Information on Protein-Protein Interactions using Topological Data Analysis

Protein-protein interactions (PPIs) govern a wide range of cellular functions. The ability to predict PPI interfaces from protein molecular surfaces is important for understanding protein function and enabling therapeutic discovery. While recent advances in structure-based learning, particularly molecular-surface geometric deep learning frameworks, have demonstrated that protein surfaces encode rich geometric and physicochemical information, such approaches often remain computationally intensive and data-hungry. Alternatively, topological data analysis (TDA) has emerged as a mathematically rigorous framework for extracting robust, multiscale shape information from complex data. In this work, we introduce a scalable TDA framework for extracting information on PPIs directly from localized protein surface patches. Our approach leverages multiscale topological descriptors, evaluated from patch-wise point cloud representations of protein mesh surfaces, combined with supervised machine learning models for interface prediction. On a full dataset of 3,362 proteins, the proposed approach substantially reduced computational cost relative to an established geometric deep learning method, MaSIF-site, decreasing preprocessing time from approximately 27 s/protein to 5-8 s/protein and total training time from approximately 6 h to 1-1.3 h. Importantly, this computational reduction is achieved while maintaining mean test area under the receiver operating characteristic curve (AUC) values of 0.76 and 0.77 for patch radii of 9 [A] and 12 [A], respectively, thus approaching the MaSIF-site test AUC of 0.84. Our results suggest that topology offers a scalable and computationally efficient approach for high-throughput extraction of information from complex biomolecular interfaces.

bioinformatics↗

bioRxiv2026

Revealing interactions between glutathione peroxidase 4 and phosphoinositides

Glutathione peroxidase 4 (GPx4) is the primary enzyme reducing lipid hydroperoxides, preventing membrane oxidative damage and protecting against ferroptosis. GPx4 is known to engage with lipid headgroups through electrostatic interactions, positioning the substrate for reduction. This work reveals and characterizes binding of highly anionic phosphoinositides (PIP lipids) by GPx4. PIPs are vital lipids in human cells and are central to many signaling processes, particularly in cytosolic facing membranes. Lipid overlay assays confirm interactions between GPx4 and phosphorylated PIPs, comparable to known anionic lipid binders. Protein NMR describes the interaction between GPx4 and PIPs within micelles. The greatest resonance shifting occurs with trisphosphorylated PIP, suggesting that higher anionic charge leads to greater binding, a known driver of GPx4 substrate recognition. Preferred anionic interactions were also confirmed with titration and crystallographic structure analysis of inositol phosphate 4 (IP4). A headgroup-binding site on GPx4 is revealed to be proximal to the cationic membrane interaction site. In conjunction with molecular simulations, these results show that PIP lipid interactions allow full engagement of GPx4 with the membrane and positions the headgroup to allow the lipid tail to interact with the catalytic site. Understanding whether GPx4 preferentially interacts with PIPs will allow better understanding of the protective function of this essential enzyme and a mechanism that may protect essential lipid signaling pathways from oxidative damage. SignificanceThis study allows a deeper structural and mechanistic understanding of GPx4, the primary enzyme that reduces lipid hydroperoxides and prevents ferroptosis. Gaining an understanding of phosphoinositide binding to GPx4 reveals a mechanism for potential preservation of these important signaling molecules and for ferroptosis protection. Observation of a specific binding site for headgroup engagement reveals a plausible lipid interaction mode and functional mechanism of this important cytoprotective enzyme.

biophysics↗

bioRxiv2026

Tunable electrostatic interactions of lipid-coated quantum dots with biological membranes

Surface functionalization of inorganic quantum dot nanoparticles is of great interest in the application of these materials toward a wide range of biological applications where membrane interactions are critical. The use of amphiphilic lipids to functionalize the surfaces of quantum dots represents a promising alternative to produce water-soluble and membrane-active materials with facile tuning of the quantum dots surface properties. Here, we demonstrate an experimental approach that yields lipid-coated quantum dots with highly tunable surface charge by controlling the concentration of cationic lipids during preparation. Through fluorescence-activated cell sorting assays, we show that these cationic lipid-coated quantum dots can enhance membrane interactions and increase membrane labeling density in live HEK293 cells. We further employed coarse-grained molecular dynamics simulations to model the lipid self-assembly process using an implicit solvent force field and subsequently model the adsorption of lipid-coated quantum dots to model membranes. Our simulations show that we can control the effective surface charge of lipid-coated quantum dots and influence the strength of adsorption to oppositely charged lipid membranes, a process that is mediated by the release of counterions at the quantum dot-membrane interface. This work supports the future development of biocompatible and water-soluble inorganic nanoparticles with highly tunable surfaces, and provides mechanistic insight into how different lipids can influence nanoparticle-membrane interactions at a molecular scale.

biophysics↗

bioRxiv2025

Decoding protein-membrane binding interfaces from surface-fingerprint-based geometric deep learning and molecular dynamics simulations

Predicting protein-membrane interactions is a formidable challenge due to the subtle physicochemical features that distinguish membrane-binding regions of a protein surface, as well as the scarcity of experimentally resolved membrane-bound protein conformations. Here, we present MaSIF-PMP, a geometric deep learning model that leverages molecular surface fingerprints to predict interfacial binding sites (IBSs) of peripheral membrane proteins (PMPs). MaSIF-PMP integrates geometric and chemical surface features to produce spatially resolved IBS predictions. Compared to existing models, MaSIF-PMP achieves superior performance for IBS classification, while feature ablation studies and transfer learning analyses reveal distinct determinants governing protein-membrane versus protein-protein interactions. We further show that molecular dynamics (MD) simulations can validate model predictions, refine IBS labels, and capture composition-dependent membrane binding patterns. These results establish MaSIF-PMP as an effective framework for IBS prediction and highlight the potential of incorporating conformational dynamics from MD to improve both model accuracy and biological interpretability.

bioinformatics↗