Candidozyma auris (formerly Candida auris) is an emerging multidrug-resistant fungal pathogen with confirmed cases in over 30 countries. Although whole-genome sequencing (WGS) analysis defined distinct clades during characterization of underlying genetic mechanism behind multidrug resistance, Clade II remains under-evaluated. In this study, a three-level comparative genomic strategy (Global, Clade, Phenotype) was employed by integration of unbiased genome-wide comparative SNP screening (GATK v4.1.9.0), targeted BLAST profiling (BLAST+ v2.17.0), and in silico protein analysis (ColabFold v1.5.5; DynaMut2 v2.0) for systematic evaluation of mechanisms of antifungal resistance in thirty-nine Clade II C. auris clinical isolates and fourteen reference strains. Global and clade-level analyses confirmed that all the clinical isolates belong to Clade II, according to phylogenetic clustering and mating type locus (MTL) conservation. At the phenotype level, a distinct subclade of fluconazole-resistant mutants was identified to have a heterogenous network of mutations in seven key enzymes associated with cell membrane dynamics and the metabolic stress response. Among these, four core mutations (TAC1B, CAN2, NIC96, PMA1) were confirmed as functional drivers based on strict criteria during multitier in silico protein analysis: cross-species conservation, surface exposure, active site proximity, thermodynamic stability, and protein interface interaction. On the other hand, three high-level fluconazole-resistant clinical isolates (≥128 μg/mL) that lacked these functional drivers were subjected to comprehensive subtractive genomic profiling analysis. The absence of coding mutations in validated resistance drivers, yeast orthologs, and convergent variants suggests that there is an alternative novel non-coding or regulatory mechanism behind fluconazole resistance. These findings highlight Clade II’s evolutionary divergence into two distinct trajectories towards the development of a high level of fluconazole resistance: canonical protein alteration versus regulatory modulation.

Iterative polyketide synthases (iPKSs) rely on communication between acyl carrier protein (ACP) and acyltransferase (AT) domains to ensure efficient delivery of starter and extender substrates during biosynthesis. However, the molecular determinants governing the AT:ACP interface remain poorly understood. Here, we use the fungal highly reducing PKS, SimG, a component of the cyclosporin biosynthetic pathway, as a model system to dissect the AT:ACP interface. Using alanine scanning mutagenesis combined with a high-throughput intact protein mass spectrometry assay, we identified interface residues that affect AT:ACP interaction. These experimental constraints were used to guide docking and molecular dynamics simulations to produce a data-driven structural model of the SimG AT:ACP complex in a catalytically competent geometry. We also demonstrate that the SimG AT domain transacylates ACP domains from a range of fungal PKS architectural classes, highlighting significant interface plasticity. These insights advance our fundamental understanding of domain communication in these enigmatic megasynthases and provide a foundation for rational engineering to expand substrate scope towards novel polyketide scaffolds.

Docosahexaenoic acid (DHA), the dominant polyunsaturated fatty acid in photoreceptors, neurons, and synapses, is usually described as a passive structural membrane constituent. We propose a different view: DHA is a quantum-electronically active molecule whose conjugated double-bond system creates an electron-rich matrix that couples with proteins to form quantum “clouds” and high-speed signaling central to recognition, recall, and cognition. Integrating evidence from molecular evolution, biophysics, and neuroscience, we argue that, as the original chromophore, DHA’s unique properties enabled the emergence of the nervous system and continue to provide the electronic substrate for cognition. By suggesting that cognition depends not only on protein-based mechanisms but on DHA-mediated electron dynamics at the membrane–protein interface, this perspective reframes DHA as an active, conserved determinant of brain evolution and function.

Upregulation of serine arginine protein kinase 1 (SRPK1), a protein responsible for phosphorylation of Ser-Arg rich residues aimed at SR proteins, is associated with apoptosis, poor survival, etc. Catalytic sites of the kinase proteins are incompetently preserved, causing difficulty in developing competitive inhibitors for ATP binding sites with broad selectivity; hence, search for inhibitor for the ATP binding pocket of SRPK1 is a necessity for medication against carcinogenesis. Natural product database is explored, and six small molecules are identified; having tolerable pharmacokinetics (low blood brain barrier, moderate clearance rate etc.) and quantum chemical properties are checked. Molecular docking study followed by molecular dynamics give insights into the effective interactions at the ATP pocket. Ligands are screened by MM-GBSA/NMA protocol, followed by estimation of unbinding potential of mean force (PMF) using well-tempered metadynamics. Well-tempered metadynamics confirmed unbinding PMF of -23.71 kcal mol-1 for CNP0199214 and -14.81 kcal mol-1 for MSC1186 (Lig_ref) to a relative difference in PMF of the screened ligand to be ≈7 kcal mol-1. A probable gating mechanism is observed for the reference ligand (Lig_ref) at the protein interface resulting multiple minima in PMF, whereas Lig_4 (CNP0199214) exhibits greater affinity toward the active pocket and therefore choice for a potent compound.

ABSTRACT The interactions that occur in the interface of proteins and ligand‐stabilised metal nanoclusters are crucial to understand the adsorption process of biomolecules on the surface of these nanomaterials. Despite the relevance of the adsorption phenomena for biological applications, such as bioimaging, biosensing and targeted drug delivery, efforts to model the interactions observed in the interface of those systems are still scarce in the literature. In this work, a model of the interactions observed in the peptide–Au 38 (p‐MBA) 24 interface was developed, employing clustering analysis, an unsupervised machine learning technique. The accuracy of this model was evaluated by simulating the peptide–Au 38 (p‐MBA) 24 interaction using molecular dynamics simulations and density functional theory calculations. The insights derived from this model can also be applied to the context of protein–AuNC interactions, given that the model was developed to provide a generalisable approach. The developed model was able to predict the amino acids that could interact well or poorly with the gold nanoclusters (AuNC), defining the specific chemical groups responsible for the effect. The results obtained in this study can lead efforts to accelerate discoveries in the fields that rely on the understanding of the interaction observed in the protein–AuNC interface.

Accurate prediction of the set of sequences compatible with a protein-protein interface is an unsolved problem in biology. While supervised sequence-based models trained directly on experimental data can predict variant effects, they often fail to generalize to significantly diverged sequences. We hypothesized that incorporating information from deep learning models of proteins (e.g., ESM, ProteinMPNN) could enhance generalizability. To test this hypothesis, we designed and experimentally screened several deep mutational libraries of the SARS-CoV-2 Spike Receptor Binding Domain (RBD) for binding to the ACE2 receptor. Our large dataset encompasses over 43,000 sequence variants, exhibiting up to 26 substitutions away from the parental RBD sequence, thus exploring a significantly expanded sequence space compared to previous studies. Baseline supervised learning with one-hot encoded sequences achieved high accuracy within training sets but poor performance on unseen libraries. Integrating pre-trained protein model embeddings (ESM2) as a feature showed modest improvement in generalization. To further enhance predictive power, we developed a graph attention network architecture that combines representations of local residue environments using protein structure graphs with long-range inter-residue correlations captured by protein language model (PLM) embeddings (GAN-PLM). By explicitly modeling residue environments, interface geometry, and sequence dependencies, our graph attention model outperformed purely sequence-based models, achieving substantially higher balanced accuracies when predicting functional ACE2-binding variants across the diverse sequence space spanned by our independent libraries. This demonstrates the potential of structure- and sequence-based features into deep learning frameworks to achieve accurate and generalizable predictions of protein interface function, with broad implications for understanding and engineering protein interactions relevant to emerging infectious diseases and therapeutic protein design.

Role of globulins and albumins in oil-water interface and emulsion stabilization properties of pulse proteins

Bladder cancer (BC) is among the ten most common malignant tumors worldwide. Tumor stem cells contribute significantly to postoperative recurrence and disease progression. Understanding stem cell interactions with other tissue cells and developing a prognostic model may improve BC management. Through single-cell RNA sequencing, we identified tumor stem cells in BC tissue. We identified 91 genes specifically upregulated in the stem cell cluster versus other clusters. Of these, 67 whose high expression correlated with poorer patient survival were defined as high-risk stemness genes (HRSGs). Further analysis showed that the MDK-LRP1 axis is the primary communication pathway between stem cells and other cell types, involving 31 of these HRSGs. Based on these 31 HRSGs, patients were stratified into two stemness clusters (ST cluster A and B), with ST cluster B associated with poorer prognosis. We further selected prognosis-related genes from differentially expressed genes between the clusters to construct a risk model. Patients in ST cluster B exhibited higher risk scores, aligning with clinical outcomes. Among the HRSGs, ACTN1 emerged as a key gene, showing elevated expression in patients with poor survival and advanced disease stages. Immunohistochemistry confirmed significantly increased ACTN1 protein levels in BC tissues. Additionally, protein interface analysis indicated that the Cys104 residue of Midkine potentially interacts with both LRP1 and ACTN1 within a 5 Å distance, suggesting a critical interaction site. These findings provide novel insights into stem cell–mediated BC progression and offer potential prognostic and therapeutic targets.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12935-025-04009-0.

Disclosure: M. D'silva: None. B. Malachowska: None. Y. Qiu: None. J. Sepulveda: None. R. Chang: None. S. Sidoli: None. J.L. Nadler: None. I.J. Kurland: Scorpion Therapeutics.An important early hallmark of Type I Diabetes Mellitus (T1D) is beta-cell dysfunction, a critical mechanism of which is cytokine mediated. Our exploratory study defined regulatory programs controlling human islet responses to pro-inflammatory cytokines by mapping abundance proteomics , as well as metabolomics and lipidomics from the same islets exposed to inflammation-inducing cytokines TNF-alpha, IL-1β and IFN-γ over the time course at 1,6 and 24 hrs after cytokine exposure. Methods: Pancreatic islets were purchased from Prodo Labs (https://prodolabs.com/). Islet samples were extracted by the Folch method, and targeted mass spectrometric methods employing a Sciex 6500+ QTRAP was used to identified metabolites, neutral and phospholipids and oxylipins/eicosanoids,from the aqueous and chloroform fractions. The Folch protein interface was collected, redissolved and then digested using S-Trap flters (Protif), 5% SDS. Proteome raw files were searched using Proteome Discoverer software. Results: Approximately 3500 proteins were identified, of which approximately 200 were significantly changed with cytokine vs control treatment that were further identified from the RECON genome-scale metabolic reconstruction as potentially important for metabolic network flux determination. ∼100 of these metabolic proteomic changes were seen at 1 hr , and an additional 85 metabolic proteins were significantly changed at 6 hr after cytokine stimulation along 35 proteins that were still changed from 1hr. Overall, cytokine stimulation resulted in an inhibition of metabolic protein expression at these time points, however, the metabolomic analysis indicated that islet metabolism was re-directed to specific pathways rather than simply suppressed. Levels of nucleotides needed for DNA and RNA synthesis could be seen to be increased, as well as pentose and hexosamine pathway intermediates, which depleted glycolytic and TCA cycle intermediates. The inflammatory islet response was reflected by an increase in platelet-activating factor PAF and lyso PAF at 1 and 6 hours, however the brunt of the oxylipin/eicosanoid response was delayed until 24 hours after cytokine stimulation, and secreted (media) oxylipins were composed mainly of cycloxygenase (COX) products of arachidonic (AA) acid, as well as 12- and 15-lipoxygenase (LOX) products of AA and DHA. At 24 hours after cytokine stimulation, the suppression/elevation of the metabolic responses seen at 1 and 6 hrs were largely resolved, along with elevations in HLA proteins and proteosome subunit expression consistent with immune response activation. Conclusion: Interventions for arresting beta cell dysfunction need to consider the pathways which may be activated sequentially and in coordination with immune activation.Presentation: Monday, July 14, 2025

Molecular glues can drive targeted protein degradation by stabilizing ternary complexes between proteins of interest and E3 ubiquitin ligases, but rational design has lagged due to limited rules for interface recognition and an overreliance on a few ligases (e.g., VHL or Cereblon). We introduce GlueFinder, a systematic, unbiased platform that leverages structural bioinformatics to mine the Protein Data Bank for ligand binding pockets adjacent to the protein interface which are ligandable sites near protein–protein interfaces that can nucleate glue-mediated complex formation. After validating its performance on a benchmark of experimentally solved dimeric structures with known and predicted glues, we applied GlueFinder to three therapeutically important targets, EGFR, HER2, and KRAS, and predicted candidate glues that recruit 24, 111, and 148 distinct E3 ligases to these targets, respectively. We further demonstrate that GlueFinder can promote the formation of non-native EGFR complexes, possibly enabling ternary assemblies that would not form on their own. Together, these results establish a general, computation-guided strategy for molecular glue discovery that decouples design from legacy degrader scaffolds and specific ligase dependencies, expands the usable E3 ligase repertoire, and enables rational targeting of interfacial binding pockets. GlueFinder thus broadens both the scope and precision of targeted protein degradation and moves the field toward mechanism-driven, systematic glue development across diverse therapeutic contexts.

Structural and molecular dynamics insights into the competitive inhibition of the platelet-activating factor receptor by acyl-PAF

Positive allosteric modulators (PAMs) of the μ opioid receptor (MOR) offer a promising path toward safer opioid therapeutics, yet their mechanisms of action remain poorly understood. Here, we uncover the structural and mechanistic basis of BMS-986187, a chemically distinct MOR PAM with in vivo efficacy, using an integrated approach combining cryogenic electron microscopy (cryo-EM), molecular dynamics (MD) simulations, signaling assays, and site-directed mutagenesis. We identify a previously uncharacterized allosteric site for BMS-986187, a lipid-facing pocket formed by MOR transmembrane helices 2, 3, and 4, distinct from sites occupied by other known MOR PAMs or negative allosteric modulators. BMS-986187 engages both receptor residues and a neighboring cholesterol molecule, suggesting a cooperative ligand–lipid mechanism. Our studies pinpoint residues essential for allosteric modulation, while information-theory analysis of MD trajectories uncovers specific allosteric communication pathways linking the PAM site to both the orthosteric agonist DAMGO and the G protein interface. Together, these findings redefine the landscape of MOR allosteric modulation by revealing a novel binding site, a potentially lipid-sensitive allosteric mechanism, and the molecular wiring of long-range communication within MOR. This work provides a new molecular framework for the rational design of PAMs targeting opioid receptors with improved precision and possible therapeutic potential.

Advances of computational protein design: Principles, strategies and applications in nutrition and health.

Cytokine-mediated immunotherapy has rapidly emerged as an effective alternative approach for cancer treatment by modulating the anti-tumor response. Interleukin-18 (IL-18) is considered as a promising cancer therapeutic agent due to the ability of cytokines to inhibit cancer by enhancing natural killer (NK) cell and cytotoxic T cell responses. Since the activity of IL-18 is required for the specific binding to IL-18 receptors, the modification of binding residue at the protein interface is an attractive strategy for IL-18 activity enhancement. The aim of this study was to design and predict mutations increasing the activity of IL-18 through computational structure-based energy calculation and molecular dynamic simulations. Four candidate mutations, E6M, E6M+N111S+R131G, E6M+K129M+R131G, and E6M+N111S+K129M+R131G, could affect/facilitate the receptor binding and stability compared to the wild-type via electrostatic interaction. MD simulations demonstrated that the predicted mutation on IL-18 had no influence on the overall conformation stability, but increased flexibility in the β8–β9 hairpin loop. Furthermore, the dynamic behavior suggested that these candidates could be an alternative for the improvement of IL-18 biological activity, though the full simulation of the IL-18 complex remains necessary. In summary, this study offered a computer-aided design strategy which was of beneficial use in the design and development of IL-18 to increase its cytokine potency and efficiency.

Chemical inducers of proximity (CIPs) stabilize biomolecular interactions, often causing an emergent rewiring of cellular biochemistry. While rational design strategies can expedite the discovery of heterobifunctional CIPs, monovalent, molecular glue-like CIPs have relied predominantly on serendipity. Envisioning a prospective approach to discover molecular glues for a pre-selected target, we hypothesized that pre-existing ligands could be systematically decorated with chemical modifications to empirically discover protein-ligand surfaces that are tuned to cooperatively engage another protein interface. Here, we used sulfur(VI)-fluoride exchange (SuFEx)-based high-throughput chemistry (HTC) to install 3,163 structurally diverse chemical building blocks onto ENL and BRD4 ligands and then screened the crude products for degrader activity. This revealed dHTC1, a potent, selective, and stereochemistry-dependent degrader of ENL. It recruits CRL4 CRBN to ENL through an extended interface of protein-protein and protein-ligand contacts, but only after pre-forming the ENL:dHTC1 complex. We also characterized two structurally distinct BRD4 degraders, including dHTC3, a molecular glue that selectively dimerizes the first bromodomain of BRD4 to SCF FBXO3 , an E3 ligase not previously accessible for chemical rewiring. Altogether, this study introduces HTC as a facile tool to discover new CIPs and actionable cellular effectors of proximity pharmacology.

The continual emergence of new severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants, as well as unwanted side effects, limited worldwide availability, and risk of drug resistance from approved antivirals, continue to limit the effectiveness of coronavirus disease 2019 treatment. To continue to uncover natural products (NPs) that can inhibit validated biochemical targets of coronaviruses, we focused on a subset of brown algae extracts from Puerto Rico that demonstrated potential anti-SARS-CoV-2 activity through selective disruption of the viral Spike-receptor binding domain (RBD) to the host angiotensin-converting enzyme-2 (ACE2) cellular entry receptor, which in turn led to further evaluation of a Lobophora variegata organic extract. Employing bioassay-guided fractionation and modern NPs dereplication tools, we quickly identified fucoxanthin as an active constituent with selective inhibitory activity against spike-RBD/ACE-2 binding while also exhibiting moderate potency in vitro toward pseudo-virus-infected VERO cells. Preliminary in silico docking demonstrated that fucoxanthin binds stably at the RBD-spike protein interface and has provided a hypothesized binding site for further chemical and biological studies.

ABSTRACTβ‐lactoglobulin (BLG) proteins from bovine milk have been widely studied as effective natural stabilizers due to their amphiphilic nature that helps to reduce surface tension. The present work investigates BLG‐stabilized oil‐in‐water emulsions (BLG‐emulsions [BLGem]) at pH 7, focusing on the effects of protein, oil, and salt concentrations under three different conditions (fresh, aged, and aged at elevated temperature). Alongside methods to study emulsions such as optical microscopy, rheology, and dynamic light scattering, we employ synchrotron‐based small‐angle x‐ray scattering (SAXS) and reveal the structural changes in the emulsions over time. We show that higher protein (2–5 wt.%) and higher oil concentrations (40–75 wt.%) improve emulsion stability by reducing the rate of phase separation and coalescence compared to the emulsions with lower protein and oil concentrations and obtain information about how the protein conformation and interface layer thickness varies with the conditions used. Ageing from Days 1 to 90 induce structural changes, including decreased scattering intensity and droplet coalescence. Our studies prove SAXS as an effective tool for studying concentrated systems, giving insights into the structure and behavior of food grade emulsions.Practical Applications: The findings from this study deepen our understanding of protein‐stabilized emulsion systems and the effects of ageing and temperature. The systems studied were designed to replicate real food formulations, characterized by high protein and oil concentrations, aligning with industrial needs for real food products such as creams, sauces, and nutritional beverages. Understanding protein‐stabilized emulsions serves as an important first step toward exploring more complex systems, such as Pickering emulsions.

Fatty acids of specific chain lengths offer precursors for high-value renewable energy and fine chemicals industries. In plants and algae, the fatty acid chain length is determined by thioesterase-mediated hydrolysis of fatty acids from acyl carrier proteins through a hitherto unclear mechanism. Herein, a 2.50 Å resolution X-ray crystallography structure and an AlphaFold Multimer-generated model were used to identify active-site, substrate-binding, and protein-binding features contributing to catalysis. Coupled with mutational studies to determine impacts on product formation, we propose a catalytic mechanism involving water as a general base with surface residues specific to coordinating acyl carrier protein alignment. Binding tunnel restructuring altered substrate specificity of the thioesterase, and introduction of a non-native thioesterase with matching protein interface gave 95% hydrolysis of C12 fatty acids, offering new approaches for algae fatty acid biosynthetic design.

Mutant RAS proteins are among the most prevalent drivers of human cancer, and the glycine to aspartic acid mutation at codon 12 (G12D) is the most common variant. Mutation-selective covalent inhibitors spare RAS in healthy tissue and enable extended pharmacodynamic effect, but covalent targeting of RASG12D is hindered by low nucleophilicity and high proteomic abundance of carboxylic acids. We overcame these challenges with compounds that bind cyclophilin A (CYPA) to create a neomorphic protein-protein interface between CYPA and active RAS that enables selective, enzyme-like rate enhancement of the covalent reaction between D12 and electrophilic warheads with exceptionally low intrinsic reactivity. This approach yielded orally bioavailable compounds with marked antitumor activity in multiple preclinical models of KRASG12D cancers, including the investigational agent zoldonrasib (RMC-9805) currently undergoing clinical evaluation (NCT06040541).

The phosphorylation of the microtubule-associated tau protein plays a key role in the regulation of its physiological function. In particular, tau hyperphosphorylation affects its binding to the tubulin surface, destabilizing the tau-microtubule interface and leading to the accumulation of fibrillar aggregates in the brain. In this work, we performed classical molecular dynamics simulations for the tau-R2/tubulin assembly with various phosphorylation states of serines 285, 289, and 293. We analyze the resulting trajectories to obtain a detailed view of the protein interface in the complex and the impact of tau phosphorylations on the stability of this assembly and on the mobility of the tubulin disordered C-terminal tails (CTTs). We show how the tubulin CTTs help maintain the tau-R2 fragment on the tubulin surface despite the destabilizing effect induced by phosphorylations. Conversely, tau phosphorylation affects the CTTs' flexibility and their potential activity as MAP-recruiting hooks. Furthermore, counterion-mediated bridges between the phosphate groups and tubulin glutamates also contribute to the binding of tau-R2 on the MT. Overall, the complex dynamics of this fuzzy phosphorylated assembly shed new light on the importance of the cytoplasmic environment in neurons in the context of Alzheimer's disease.