The clinical application of the chemotherapeutic agent irinotecan is critically hindered by its low and variable solubility. To provide a fundamental understanding of this issue, we employed molecular dynamics simulations and free energy calculations to detail the solvation thermodynamics of irinotecan. Our analysis reveals that irinotecan's solvation preference is governed by a delicate and often competitive balance between two fundamental physical contributions: the Lennard-Jones term (representing cavity formation and dispersion) and favorable solute-solvent electrostatic interactions. We demonstrate that while polar protic solvents (e.g., water) provide the strongest electrostatic stabilization, their high energetic cost for cavity formation severely limits overall solvation favorability. Conversely, polar aprotic solvents (e.g., pyridine and DMSO) optimize this balance by facilitating easier cavity formation while still providing strong electrostatic interactions, resulting in the most favorable solvation profiles. Notably, irinotecan's unexpectedly high relative solubility in cyclohexane compared to water underscores the critical role of solvent reorganization energy in dictating solution-phase behavior. These molecular-level findings are rigorously validated by structural analyses (connection matrices and radial distribution functions) and a complementary macroscopic solubility parameter analysis (MOSCED framework). This study offers a robust, integrated, and predictive physicochemical framework for understanding and optimizing the formulation of complex, flexible drug molecules.

Salinity threshold for the dewetting of oil droplets from a kerogen surface in NaCl solutions using molecular dynamics simulations and free energy calculations

Membrane fusion is a fundamental process involved in exocytosis, fertilization, or cell entry by enveloped viruses. Membrane fusion is facilitated by fusion proteins, which are anchored in membranes by helical transmembrane domains (TMDs). Previous studies showed that TMD variations may alter the fusion efficiency, suggesting that TMDs are not merely passive anchors; however the mechanism by which TMDs drive fusion is not well understood. We used high-throughput coarse-grained molecular dynamics simulations and free energy calculations to quantify effects of TMDs on the formation of the first fusion intermediate, that is, of a fusion stalk. We analyzed five physiologically relevant TMDs derived from viral fusion proteins and the SNARE complex embedded in various lipid environments. We find that the addition of TMDs favors stalk formation by typically 10 to 30 kJ/mol in a concentration-dependent manner. Using helices with sequences R2LnR2 (n = 6, …, 26), we find that negative hydrophobic mismatch between the TMD and the membrane core strongly promotes fusion. Analysis of the lipid tail order parameters of annular lipids revealed a strong correlation between stalk stabilization and induced lipid disorder. Together, our findings suggest that TMDs actively contribute to membrane fusogenicity by locally perturbing the membrane order.

Ion transport through nanoscale channels enables advanced functionalities, such as ionic current rectification (ICR), with promising applications in neuromorphic computing and biomimetic signal processing. However, the fundamental mechanisms controlling the ion dynamics under nanoconfinement remain poorly understood. Using atomistic molecular dynamics simulations and free energy calculations, we demonstrate that multilayered Janus graphene oxide nanopores exhibit exceptional and tunable ICR performance mediated by interlayer coupling. These structures achieve a rectification ratio enhancement of up to 2 orders of magnitude─from ∼2 in a single layer to over 2000 at 3.5 V/nm in multilayered configurations─and a shift of the peak rectification field from 0.7 to 3.5 V/nm with increasing layer number. Ion distribution analyses reveal distinctive ionic enrichment-depletion behavior unique to multilayered architectures. Thermodynamically, we unveil that synergistic interlayer coupling fundamentally reshapes the free energy landscape, creating a highly asymmetric profile with multiple energy barriers and wells due to entropy─enthalpy competition. Importantly, entropy is identified to play a critical role in stabilizing energy wells and facilitating directional ion transport─a mechanism absent in single-layer systems. These insights provide a mechanistic basis for ion rectification and establish design principles, such as interlayer spacing or number control, for developing high-performance ionic membranes and nanofluidic devices.

Gwt1, an essential acyltransferase in the glycosylphosphatidylinositol (GPI) biosynthesis pathway, is a promising target for the development of high-selectivity antifungal agents. In this study, we combined molecular dynamics (MD) simulations and free energy calculations to characterize the binding mechanism of Gwt1 with its native substrate, palmitoyl-CoA. Our simulations identified key hydrogen-bonding and ionic interactions critical for substrate recognition, particularly involving residues Lys123, Arg181, and Asn432. Potential of mean force (PMF) calculations revealed multiple conformational states of palmitoyl-CoA, including an I-shaped conformation that sterically occludes the GlcN-PI binding site, thereby hindering the acyl transfer step. Leveraging these structural insights, we performed virtual screening targeting the hydrophobic pocket formed by Tyr129, Tyr400, Phe404, and Tyr408, which identified two approved drugs, tivozanib and rosiglitazone, as potential Gwt1 inhibitors. Experimental validation confirmed their antifungal activities against pathogenic fungi, including Cryptococcus neoformans, Candida albicans, and Aspergillus fumigatus. This work provides dynamic mechanistic insights into Gwt1 function and offers a rational strategy for repurposing existing drugs as antifungals targeting the GPI pathway.

The protein-protein interaction (PPI) between metadherin (MTDH) and Staphylococcal nuclease domain-containing protein 1 (SND1) is a pivotal oncogenic driver in various cancers, yet the atomic-level details of their binding mechanism remain elusive, hindering targeted drug discovery. This study employs integrated computational approaches, including molecular dynamics (MD) simulations, binding free energy calculations, and residue interaction network analysis, to identify hotspot residues at the MTDH-SND1 interface and elucidate the binding mechanism. The results demonstrate that the MTDH-SND1 complex exhibits strong binding affinity, primarily driven by electrostatic and hydrophobic interactions. Structural stability analysis confirmed the complex's integrity during simulations, while dynamic cross-correlation and mutual information analyses revealed a key interaction region (R1) with correlated motions, which was further proved by contact probability analysis. Hydrogen bond analysis identified a stable network involving residues Arg239, Arg243, and Hie263, which were confirmed as hotspot residues by the alanine scanning mutagenesis method. Furthermore, the binding and interaction mechanisms between SND1 and 12 activity-known small molecule inhibitors were investigated and compared with that in the MTDH-SND1 complex. Energy decomposition highlighted that the conserved triad-Arg239, Arg243, and Hie263-is crucial across all systems. This work provides unprecedented atomic-level insights into the MTDH-SND1 interaction and offers a robust structural foundation for the rational design of high-affinity inhibitors against this oncogenic PPI.

Macrocyclic drugs are promising for targeting undruggable proteins, including those in cancer. Our prior work identified BE-43547A2 (BE) as a selective inhibitor of pancreatic cancer stem cells in PANC-1 cultures, but its high lipophilicity limits clinical application. To address this, we designed derivatives retaining BE’s backbone while modifying tail groups to improve its properties. A concise total synthesis enabled a versatile late-stage intermediate (compound 17), serving as a platform for efficient diversification of BE analogs via modular click chemistry. This approach introduced a central triazole ring connected by flexible alkyl spacers. Key properties, including lipophilicity, solubility, and Caco-2 permeability, were experimentally determined. These derivatives exhibited reduced lipophilicity and improved solubility but unexpectedly lost cellular activity. Direct target engagement studies using MicroScale Thermophoresis (MST) revealed compound-dependent deactivation mechanisms: certain derivatives retained binding to eEF1A1 with only modestly reduced affinity (e.g., compound 29), while others showed no detectable binding (e.g., compound 31). Microsecond-scale molecular dynamics simulations and free-energy calculations showed that, for derivatives retaining target affinity, tail modifications disrupted the delicate balance of drug–membrane and drug–solvent interactions, resulting in substantially higher transmembrane free-energy penalties (>5 kcal/mol) compared to active compounds (<2 kcal/mol). These insights emphasize the need to simultaneously preserve both target engagement and optimal permeability when modifying side chains in cell-permeable macrocyclic peptides, positioning compound 17 as a robust scaffold for future lead optimization. This work furnishes a blueprint for balancing drug-like properties with therapeutic potency in macrocyclic therapeutics.

BackgroundEfficient recognition of antigenic peptides bound to major histocompatibility complex (MHC) class I molecules on the surface of cells by immune cells requires sufficiently stable peptide-MHC I complexes. Antigenic peptides of 8–10 amino acids are typically bound in a narrow cleft between two flanking α1 and α2 helices on top of an extended β sheet floor. For some MHC I alleles the efficient loading with high-affinity peptides in the endoplasmic reticulum (ER) requires the transient binding and assistance of the chaperone proteins tapasin and/or TAPBPR. The structures of both chaperones in complex with MHC I molecules have been resolved and indicate similar structural interface elements and also similar structural changes of the bound MHC I molecules which includes a significant shift of an α2–1 helix, a segment of the α2 helix, which partially opens up the binding cleft.MethodsThe role of this α2–1 helix movement for the peptide loading and editing processes is not fully understood. We employed extensive Molecular Dynamics (MD) simulations and free energy calculations to estimate free energy changes associated with the α2–1 helix movement in the absence as well as presence of low- and high-affinity peptides and in complex with tapasin or TAPBPR.ResultsThe α2–1 helix shift with respect to the conformation in a native MHC I peptide complex significantly destabilizes the binding of peptides and can induce partial dissociation in case of low and medium-affinity peptides. Only a bound high-affinity peptide leads to a narrowing of the binding cleft and reduces the interaction of the MHC I with the chaperone molecules.ConclusionsThe simulations indicate that the conformational shifts of the α2–1 helix with respect to the chaperone and the MHC I molecule play a dominant role for destabilizing peptide binding as well as triggering release from the chaperone in case of high-affinity peptide binding. In addition to the role of the α2–1 helix, we also compared the motion of a loop region found near the N-terminus of tapasin and TAPBPR that may also play a role in the chaperone process.

GABA transporter 1 (GAT1), encoded by the SLC6A1 gene, is essential for maintaining inhibitory neurotransmission by mediating the reuptake of GABA from the synaptic cleft. Dysfunction of GAT1 has been linked to several neurological and neurodevelopmental disorders, including epilepsy and Alzheimer’s disease. In this study, we performed a comprehensive computational investigation of reported GAT1 mutations to understand their structural and functional implications. Seven mutations (G63S, Y140C, Q291Δ, F294Δ, N310I, D451G, and G457H) were analyzed using homology modeling, structural validation tools, molecular dynamics (MD) simulation triplicates, and binding free energy calculations via the gmx_MMPBSA method. The wild-type consistently exhibited the most favorable interaction energy (−59.89 kcal/mol), the strongest binding free energy (ΔG = −28.29 kcal/mol), and the most stable hydrogen-bonding network. While all mutants displayed elevated RMSD and energy fluctuations relative to the wild-type, these changes predominantly reflected local conformational disturbances rather than global unfolding, indicating that the overall structural framework of GAT1 remains largely preserved. Among the mutants, G63S exerted the mildest effect on ligand stabilization, whereas Y140C, G457H, Q291Δ, and D451G produced substantial reductions in protein–ligand stability, weakened hydrogen bonding, and increased ligand mobility within the binding pocket. Free-energy analysis further highlighted the pronounced destabilizing influence of N310I, Q291Δ, and G457H on tiagabine binding. These findings provide mechanistic insights into how specific GAT1 mutations may alter transporter stability and function, offering a structural framework for future studies on GABAergic dysfunction and therapeutic development.

The dopamine transporter (DAT) plays a vital role in maintaining dopamine (DA) homeostasis by mediating the reuptake of DA from the synaptic cleft into presynaptic neurons, a process tightly coupled to the cotransport of two sodium ions (Na+) and one chloride ion (Cl-). Although structural studies have revealed key conformational states of DAT, the precise mechanistic contributions of these ions to transporter dynamics and substrate translocation remain incompletely understood. Here, we employed extensive molecular dynamics simulations and free energy calculations to systematically investigate the cooperative roles of Na+ and Cl- ions in human DAT function. Our results demonstrate that Cl- stabilizes extracellular gate closure through coordination with residues in TM2, TM6a, and TM7, while the two Na+ ions reinforce intracellular gate closure via interactions with TM1a and adjacent helices. Sequential binding analysis revealed an energetically favorable and functionally coupled order of binding: Na+ binds first, followed by DA and then Cl-. Mapping of the free energy landscape for DA translocation uncovered four key intermediate states, each stabilized by distinct salt bridges, hydrogen bonds, and π-π stacking interactions that shape the energy barriers along the transport pathway. These findings provide a comprehensive molecular framework for understanding ion-dependent conformational transitions in DAT and offer mechanistic insights for the rational design of therapeutics targeting DAT in neuropsychiatric disorders.

Molecular dynamics investigation of cysteine mutations: Effects on calcium ion affinity and structural stability in the RET cysteine-rich domain.

Lignin is one of the most abundant natural biopolymers and plays a crucial role in the development of safe and sustainable alternatives for healthcare products. In this study, we employed molecular dynamics simulations and free energy calculations to investigate lignin derivatives’ interactions with skin-like membranes. Specifically, we designed a small lignin derivative composed of syringyl and guaiacyl subunits. Our results reveal that molecular size, concentration, and thermal conditions critically influence the insertion, interaction dynamics, and localization behavior of lignin derivatives. Notably, variations in these parameters induce distinct behaviors, including rapid membrane insertion, hydrogen bonding, clustering, and surface adhesion. The findings provide insights into the molecular mechanisms governing lignin derivatives’ interactions with skin-like membranes, with implications for developing bio-based skincare formulations and transdermal delivery systems. Our results highlight the importance of molecular size and concentration in optimizing lignin-derived compounds for dermatological and therapeutic applications.

Fast, portable, and reliable detection of chemical and biological compounds is an important challenge in many safety and healthcare applications. Chemical sensors based on photonic crystals that use protein-catalyzed capture (PCC) agents as molecular sensors are promising tools for the portable detection of chemical and biological compounds. We present the development and deployment of a high-throughput virtual screening protocol to computationally identify PCC candidates that maximize the binding sensitivity and selectivity for fentanyl. The approach integrates enhanced sampling molecular dynamics free-energy calculations, Gaussian process regression surrogate models, and Bayesian optimization to efficiently navigate the design space of over 1 million PCC candidates, resolve the sensitivity-selectivity Pareto frontier, and identify the top-performing PCC candidates. We analyze the molecular interactions between our top candidates and fentanyl target to propose design rules for high-performance PCC agents to be used for experimental testing and, ultimately, incorporation into next-generation hydrogel-nanoparticle-based chemical sensing devices.

Development of a fluorescence-based assay for screening of urate transporter 1 inhibitors (II): Optimization of fluorescent substrates and structure-activity relationships analysis.

Serotonin transporter (SERT) regulates serotonergic signals by reuptaking serotonin from the synaptic clefts back into the presynaptic neurons. The recent resolution of the serotonin-SERT complex in multiple conformational states outlined the complete serotonin import cycle. However, a detailed functional appreciation of SERT also involves deciphering the coupling between global structural changes in the transport cycle to the bound chemicals to be transported. By employing molecular dynamics (MD) simulations and free energy calculations in different ligand binding states, here, we reveal how serotonin binding to SERT initiates the global conformational changes essential for serotonin import. Only when serotonin is bound to the central binding site, wedged between transmembrane helices (TMs) 3 and 8, can the system form an interaction network that bridges the two helical domains of the protein, thereby promoting the closure of an extracellular hydrophobic gate and sealing the bound serotonin. To test the role of this hydrophobic gate closure, we designed a series of nonequilibrium MD simulations to steer the outward-facing ↔ occluded transition with different gating configurations. The difference in nonequilibrium work required to fuel the transition indicates that the transition is more likely to happen when the extracellular gate is closed. The transition is not promoted when the gate is open or when 5-HT moves away from TM3 and TM8 toward an alternate pose. Such a local-global coupling is likely shared by other monoamine transporters considering the conservation of all involved structural elements.

The design of small-molecule inhibitors targeting proprotein convertase subtilisin/Kein type 9 (PCSK9) remains a forefront challenge in combating atherosclerosis. While various monoclonal antibodies have achieved clinical success, small-molecule inhibitors are hindered by the unique structural features of the PCSK9 binding interface. In this study, a potential small-molecule inhibitor was identified through virtual screening, followed by molecular dynamics (MD) simulations to explore the binding mechanisms between the inhibitor and the PCSK9 protein. Binding free energies were calculated using molecular mechanics/Generalized Born surface area (MM/GBSA) with the interaction entropy (IE) method, and critical hot-spot residues were identified via alanine scanning analysis. Key residues, including ARG237, ILE369, ARG194 and PHE379, were revealed to form critical interactions with inhibitor and play dominant roles during the inhibitor’s binding. In addition, the polarization effect was shown to significantly influence PCSK9–ligand binding. The identified inhibitor exhibited highly similar binding patterns with two known active compounds, providing valuable insights for the rational design and optimization of small-molecule inhibitors targeting PCSK9. This work contributes to the development of more effective treatments for hyperlipidemia and associated cardiovascular diseases.

The SARS-CoV-2 virus, responsible for the COVID-19 pandemic, has continuously evolved, generating numerous variants with varying degrees of infectivity and transmissibility. The EG.5 subvariant of SARS-CoV-2 emerged globally in mid-2023 as part of the ongoing evolution of the Omicron lineage. Derived from the recombinant XBB.1.9 sublineage, EG.5 has attracted attention due to its enhanced immune escape and sustained transmissibility. As a member of the FLip lineage, EG.5 harbors the convergent F456L mutation in the spike receptor-binding domain (RBD), a key residue for neutralizing antibody recognition. Understanding the molecular mechanisms underlying these variations is crucial for developing effective antiviral strategies. In this study, we employed accelerated molecular dynamics simulations, free-energy calculations, and interaction fingerprint analysis, to investigate the intricate molecular interactions between the spike RBD and the angiotensin-converting enzyme 2 (ACE2) receptor in wild-type SARS-CoV-2 and its variants, specifically Omicron, XBB.1.9.2, and the concerning EG.5 variant. Our findings reveal that electrostatic interactions are the predominant driving force behind the stabilization of the viral spike protein-ACE2 complex. The Omicron, XBB.1.9.2, and EG.5 variants exhibit distinct electrostatic profiles at the spike-ACE2 interface, with mutations at key residues reconfiguring local interactions. These changes enhance ACE2 binding specificity and stabilize the spike-ACE2 complex through intensified electrostatic interactions. The EG.5 variant, with its stronger binding affinity to ACE2, underscores the ongoing threat posed by SARS-CoV-2. The F456L mutation in EG.5 enhances protein stability, further supporting its increased affinity for ACE2. Our research provides valuable insights for designing targeted antiviral therapies, including peptide inhibitors and bioactive compounds. Continuous research is essential to effectively combat COVID-19 and its evolving variants.

Malaria remains a critical global health challenge, especially in Sub-Saharan Africa, with drug-resistant strains heightening the need for new treatment strategies. Plasmepsin II, a key enzyme in the life cycle of malaria presents a promising target for novel antimalarial drugs. This study investigates the interaction of luteolin, apigenin and their glycoside derivatives from Allophylus africanus with PMII target using molecular docking, molecular dynamics simulation and free energy calculations. Luteolin derivatives, particularly luteolin-7-O-glucoside and luteolin-3’,7-di-O-glucoside showed strong binding with PMII at −9.1 and −9.5 kcal/mol, respectively, while in apigenin derivatives apigenin-6,8-di-C-hexoside exhibited the most significant binding energy (−10.2 kcal/mol). The free energy calculations further confirmed the strong binding affinity with the apigenin-8-C-hexoside, demonstrating the best binding free energy (−86.646 kJ/mol). The study highlights the potential of these compounds as promising candidates for antimalarial drug development, although further experimental validation is needed.

Polycystic ovarian syndrome (PCOS) is a hormonal disorder caused by excessive secretion of male sex hormones in females. Herbal remedies for PCOS are lightning up as they bypass the adverse effects and are profoundly safe on prolonged usage. The present study included a selection of 34 herbs pursuing biological effects on the uterus, and their major chemical constituents were subjected to a series of in silico techniques using different software. The proteins contributing majorly to the hormonal functions like human cytochrome P450 CYP17A1 (3RUK), progesterone (1E3K), and estrogen receptor (1X7R) were selected for the study. Molecular docking studies were performed using AutoDock 1.5.7. The pharmacokinetic properties were predicted using the SwissADME online tool, while toxicity parameters were assessed with OSIRIS toxicity explorer and pkCSM. Molecular dynamics simulations and free energy calculations were performed using the Schrödinger suite. Constituents with a basic steroidal nucleus demonstrated high binding energy values. An analysis of all the in silico techniques showed that Sarsasapogenin from Asparagus racemosus exhibited strong binding energies of -10.88 kcal/mol, -10.51 kcal/mol, and -9.79 kcal/mol with the selected specific proteins. In molecular dynamics simulations, Sarsasapogenin displayed ideal stability, with RMSD fluctuations below 3 Å and RMSF slightly higher than the corresponding peak of apoprotein. Additionally, it showed a favorable druglikeness profile and non-toxic effects across all screened parameters. From the list of the selected constituents, Sarsasapogenin was found to be ideal, and further research on it for targeting PCOS is expected to yield promising results.

Tyrosinases are responsible for the biosynthesis of melanin pigments, which are involved in human skin disorder pathways such as hyperpigmentation. To prevent and treat these disorders, tyrosinase inhibitors are used pharmacologically and cosmetically as skin lightning agents, although the existing agents manifest unsatisfactory potency, absorption, and safety profiles. We herein design an in silico pipeline based on docking, molecular dynamics simulation, active learning, and free energy calculation to enable the discovery of a next-generation tyrosinase inhibitor. We propose the ActiHerb framework, which integrates high-throughput virtual screening of the NPASS natural product library with Bayesian active learning by disagreement uncertainty quantification, achieving a 97% reduction in computational cost while maintaining discovery efficacy. Through iterative refinement using the ActiHerb framework and alanine scanning-interaction entropy validation, we identified two promising lead compounds: NPC28366 from jujube root (Ziziphus jujuba) and NPC272853 from elecampane (Inula helenium). Our work demonstrates that the combination of deep learning and computational chemistry can efficiently discover natural product tyrosinase inhibitors with desirable potency and favorable safety profiles for topical applications, establishing a generalizable framework for computational-driven discovery of cosmetic actives.