Molecular Dynamics Simulations of Viral Envelope Protein Conformational Changes: Implications for Antiviral Targeting

Enveloped viruses employ specialized glycoproteins on their surface to mediate fusion with host cell membranes [1]. These envelope proteins, typically classified as class I, class II, or class III fusion proteins, undergo dramatic conformational rearrangements from a metastable prefusion state to a stable postfusion conformation [2]. Capturing the atomic details of these transitions is essential for understanding viral entry mechanisms and for designing targeted inhibitors [3]. Molecular dynamics (MD) simulations have become a principal computational tool for probing these large-scale conformational changes at atomic resolution, providing insights that complement cryo-electron microscopy (cryo-EM) and X-ray crystallography [4, 5]. This article reviews the application of MD simulations to viral envelope protein dynamics, the computational methodologies employed, and the implications for antiviral and vaccine design. Emphasis is placed on class I fusion proteins from viral pathogens relevant to veterinary medicine and comparative virology.

Biophysical Context of Viral Envelope Protein Conformational Changes

Viral envelope proteins are responsible for recognizing specific host cell receptors and catalyzing membrane fusion [1, 3]. For class I fusion proteins, such as influenza hemagglutinin (HA), HIV-1 envelope glycoprotein (Env), Ebola virus glycoprotein (GP), and coronavirus spike (S) protein, the fusion process is triggered by receptor binding and/or low pH exposure within endosomal compartments [2, 4]. The metastable prefusion trimer undergoes a controlled cascade of structural events: the release of the receptor-binding subunit, the extension of a central coiled-coil (the heptad repeat region 1, HR1), and the subsequent insertion of the fusion peptide into the target membrane [5]. Refolding of the HR1 region into a six-helix bundle (6HB) with the heptad repeat region 2 (HR2) draws the viral and host membranes into close apposition, enabling lipid mixing and fusion [1, 2]. These transitions involve substantial movements of individual domains, alterations in secondary structure, and changes in solvation and lipid-protein interactions [6, 7].

Molecular Dynamics Methods for Studying Envelope Proteins

All-atom MD simulations provide a detailed representation of protein, lipid, water, and ion behavior on nanosecond to microsecond timescales [8, 9]. Force fields such as CHARMM36 and AMBER ff14SB parameterize bonded and nonbonded interactions, while explicit solvent models (e.g., TIP3P) capture hydration effects critical for protein stability [5, 10]. Simulations are typically performed on systems that include the envelope protein trimer embedded in a asymmetric lipid bilayer mimicking the viral membrane [9, 10]. The presence of cholesterol, phosphatidylserine, and sphingomyelin influences protein conformation and membrane curvature, as shown for the HIV-1 gp41 transmembrane domain (TMD) [2, 9].

Enhanced sampling techniques are essential for overcoming the energy barriers that separate distinct conformational states [11]. Replica exchange MD (REMD), metadynamics, and umbrella sampling allow the calculation of free energy landscapes (FELs) along specific reaction coordinates [11, 12]. For example, Li et al. applied multiple-replica μs-scale MD simulations to compare the FELs of unliganded and CD4-bound HIV-1 gp120, revealing that CD4 binding increases the number of conformational substates and the conformational entropy of the protein [11]. Coarse-grained (CG) MD, using models such as the Martini force field, permits longer timescale simulations and the study of larger assemblies, such as the clustering of Env proteins on the virion surface [10, 12]. Hybrid approaches that combine CG simulations with backmapping to all-atom resolution provide a multiscale view of protein dynamics [10].

Conformational Dynamics of HIV-1 Envelope Glycoprotein

HIV-1 Env is the prototypical class I fusion protein and the sole immunogenic target on the virion surface [3, 11]. The trimeric complex consists of a surface subunit gp120 and a transmembrane subunit gp41 [11, 12]. MD simulations have been instrumental in characterizing the distinct conformational states of Env, including the unliganded (closed) and CD4-bound (open) forms [11]. Li et al. demonstrated that the unliganded state is structurally and energetically stable, representing a "ground state," while CD4 binding loosens the structural packing and increases flexibility, particularly in the V1/V2 and V3 loops [11].

The gp41 subunit executes the critical membrane fusion step [1, 2]. The fusion peptide inserts into the host membrane, and the TMD anchors the protein in the viral membrane [5, 9]. Zhao et al. performed atomistic simulations of the trimeric TMD in a model asymmetric viral membrane and discovered that water and chloride ions permeate the bilayer to interact with a conserved arginine bundle, (R696)3 [9]. This network of hydrogen bonds and electrostatic interactions stabilizes the TMD and modulates its conformation [9]. The tilting of the TMD trimer causes local membrane thinning and increases the volume of nearby lipids, creating an entropic driving force for conformational changes that precede fusion [9]. These findings underscore the importance of lipid environment and ion permeation in viral fusion [2, 9].

Full-length Env models that include the ectodomain, membrane-proximal external region (MPER), TMD, and cytoplasmic tail (CT) have revealed the interplay between these domains [10, 12]. Majumder et al. combined all-atom and CG simulations of fully glycosylated Env on a model HIV-1 envelope and showed that the CT mediates oligomerization of Env proteins into clusters [10]. Cao and Im found that the ectodomain remains rigid in the prefusion state, while the MPER provides intrinsic flexibility that allows the ectodomain to adopt tilted orientations, potentially facilitating receptor engagement [12]. The centrally positioned R696 residue in the TMD interacts with lipid headgroups, ions, and CT residues, generating conformational variability that perturbs the surrounding membrane and supports fusion [12].

Case Study: Ebola Virus Glycoprotein and the A82V Mutation

The Ebola virus (EBOV) GP is a class I fusion protein that mediates viral entry into host cells [4, 13]. During the 2013-2016 West African epidemic, an alanine-to-valine mutation at position 82 (A82V) in the GP emerged and persisted in most circulating isolates [4, 13]. This mutation increased GP-mediated membrane fusion and altered the dependence of the virus on host factors [4]. Durham et al. used a combination of MD simulations, fluorescence correlation spectroscopy, and single-molecule Förster resonance energy transfer (FRET) imaging to elucidate the molecular basis for this enhanced fusion activity [4, 13].

The MD simulations revealed that the A82V mutation alters the conformational dynamics of GP by tuning an allosteric network of interactions that links the receptor-binding site to the fusion loop [4, 13]. Specifically, A82V increased the mobility of the fusion loop, which is critical for insertion into the host membrane [4]. This enhanced flexibility promoted the formation of a more fusion-competent conformation, explaining the observed increase in viral infectivity in both human and animal cell lines [4, 13]. The study illustrates how a single point mutation can modulate envelope protein dynamics to increase viral fitness, and it highlights the utility of MD simulations in identifying cryptic allosteric sites that could be targeted by antiviral compounds [4].

Methodological Considerations and Force Field Selection

The accuracy of MD simulations depends critically on the force field and simulation conditions [5, 8]. For membrane-embedded envelope proteins, force fields must be balanced with compatible lipid parameters [5, 9]. The CHARMM36 force field has been widely validated for lipid-protein systems and is commonly used in simulations of viral glycoproteins [5, 9, 10]. For the heavily glycosylated Env proteins, carbohydrate parameterization is a significant challenge, but recent developments in Glycam and CHARMM carbohydrate parameters enable realistic modeling of the glycan shield [10, 12, 14].

Enhanced sampling methods are often required to observe the rare events associated with large conformational transitions [11]. Metadynamics and replica exchange MD can accelerate sampling of the free energy landscape [11]. Markov state models (MSMs) constructed from multiple simulation trajectories provide a framework for identifying metastable states and transition pathways [12]. Shehata et al. used microsecond-long all-atom simulations of full-length HIV-1 Env to capture a pronounced tilting motion of the ectodomain relative to the membrane, and they found that N-glycans at positions N88 and N611 play a critical role in modulating this tilt [14].

Implications for Antiviral and Vaccine Design

The detailed conformational dynamics revealed by MD simulations offer multiple avenues for therapeutic intervention [1, 3]. Fusion inhibitors targeting gp41, such as the second-generation inhibitor sifuvirtide (SFT) and its derivative MT-sifuvirtide (MTSFT), have been analyzed using MD simulations [15]. Ancy et al. docked SFT and MTSFT with wild-type and mutant (V10A/A19I/Q24R) HIV-1 gp41 and performed MD simulations to assess binding stability [15]. The simulations showed that the N-heptad repeat (NHR) region of mutant gp41 undergoes a helix-to-loop conformational change that reduces inhibitor binding [15]. MTSFT formed stronger interactions with the NHR than SFT, which explained its enhanced stability and resistance to mutant viruses [15].

The HIV-1 Env TMD and MPER represent underutilized drug targets [9, 12]. The conserved arginine bundle (R696)3 and its interactions with water and ions could be disrupted by small molecules that prevent the conformational changes necessary for fusion [9]. Similarly, the allosteric network identified in EBOV GP presents a new target for fusion inhibitors [4, 13]. For veterinary coronaviruses, the design of HRC-derived peptide inhibitors targeting the spike protein has been informed by structural and computational analyses [1]. Iovane et al. characterized a peptide inhibitor of canine coronavirus spike-mediated fusion, demonstrating the translation of computational insights into experimental antivirals [1].

MD simulations also inform the design of vaccine antigens by identifying the conformational states that are most vulnerable to neutralizing antibodies [10, 12]. The accessibility of epitopes on the metastable prefusion trimer is a key consideration for soluble gp140 trimer design [10, 12]. Cao and Im showed how MD trajectories can be used to evaluate the accessibility of antibody epitopes across different regions of Env, including the V1/V2 and V3 loops as well as the MPER [12]. Majumder et al. predicted that broadly neutralizing antibodies targeting these loops efficiently interact with Env clusters on the mature virion surface [10].

The following table summarizes key envelope protein systems and the conformational insights gained from MD simulations.

A Mermaid Diagram: Workflow for MD-Based Antiviral Target Identification

The following diagram represents a typical computational workflow for using MD simulations to identify conformational vulnerabilities in viral envelope proteins and design antiviral inhibitors.

graph TD
    A[Experimental Structure PDB], > B[System Construction: Protein, Membrane, Glycans, Solvent, Ions]
    B, > C[Energy Minimization and Equilibration]
    C, > D[All-Atom MD Simulation (microsecond scale)]
    D, > E[Enhanced Sampling: Replica Exchange, Metadynamics]
    D, > F[Free Energy Landscape Analysis]
    F, > G[Identify Metastable States and Transition Pathways]
    G, > H{Conformational Vulnerability Detected?}
    H, >|Yes| I[Map Allosteric Networks and Cryptic Pockets]
    I, > J[Virtual Screening or Peptide Docking]
    J, > K[Candidate Inhibitor Design and MD Validation]
    K, > L[Experimental Testing: Fusion Assays, Antiviral Activity]
    H, >|No| M[Return to Structure: New Targets or Mutants]

Future Directions and Conclusion

Advances in computing power and algorithmic efficiency continue to extend the timescale and complexity of MD simulations of viral envelope proteins [5, 10]. Multiscale approaches that link all-atom, CG, and continuum models are enabling simulations of entire virion surfaces [10]. Integration with machine learning, such as deep learning-based force fields and generative models for conformational sampling, promises to further accelerate discovery [5]. For veterinary virology, applications to animal coronaviruses, arteriviruses, and filoviruses will benefit from the same computational frameworks developed for human pathogens [1, 4].

Conformational changes in viral envelope proteins represent both a fundamental biophysical phenomenon and a rich target for therapeutic intervention [2, 3]. MD simulations provide the atomic-level resolution necessary to understand these transitions in detail, revealing cryptic binding sites, allosteric networks, and lipid-mediated regulation that are not apparent from static structures [4, 9]. As computational methods mature, their integration with structural biology and experimental virology will continue to drive the rational design of fusion inhibitors, entry blockers, and stabilized vaccine antigens for both human and animal health [1, 12, 14].

References

[1] Iovane V, Giugliano R, Gentile A, et al. Design and Characterization of an HRC-Derived Peptide Inhibitor of Canine Coronavirus Spike-Mediated Fusion. Pathogens. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41901768/

[2] Zhao J, Jiang Y, Zhao L, et al. Cholesterol Mediates Gp41 Linear Aggregation and Enhanced Membrane Curvature during HIV-1 Infection. J Phys Chem B. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41701904/

[3] Unarta IC, Crotzer S, Gnanakaran S. Tracking the protein conformational motions driving HIV-1 membrane fusion. Sci Rep. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/41495110/

[4] Durham ND, Jain A, Howard A, et al. Molecular basis for the increased fusion activity of the Ebola virus glycoprotein epidemic variant A82V: Insights from simulations and experiments. Cell Rep. 2025. URL: https://pubmed.ncbi.nlm.nih.gov/40186866/

[5] Zhao M, Lopes LJS, Sahni H, et al. Insertion and Anchoring of the HIV-1 Fusion Peptide into a Complex Membrane Mimicking the Human T-Cell. J Phys Chem B. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/39670799/

[6] Zhao M, Lopes LJS, Sahni H, et al. Insertion and Anchoring of HIV-1 Fusion Peptide into Complex Membrane Mimicking Human T-cell. bioRxiv. 2024. URL: https://pubmed.ncbi.nlm.nih.gov/39131401/

[7] Risueño C, Abrescia NGA, Coluzza I. Insights into Hepatitis C Virus E2(core) Interactions with Human Cellular Receptor CD81 at Different pHs from Molecular Simulations. J Phys Chem B. 2022. URL: https://pubmed.ncbi.nlm.nih.gov/36255318/

[8] Alaofi AL. The Glu143 Residue Might Play a Significant Role in T20 Peptide Binding to HIV-1 Receptor gp41: An In Silico Study. Molecules. 2022. URL: https://pubmed.ncbi.nlm.nih.gov/35745059/

[9] Hollingsworth LR, Lemkul JA, Bevan D, et al. HIV-1 Env gp41 Transmembrane Domain Dynamics are Modulated by Lipid, Water, and Ion Interactions. bioRxiv. 2018. URL: https://www.semanticscholar.org/paper/298ebc8f529d1641bf491be87a684015c0e3ded2

[10] Majumder A, Dutta M, Cherek L, et al. Structure and Dynamics of the HIV-1 Envelope Protein on the Virion Envelope. bioRxiv. 2026. URL: https://www.semanticscholar.org/paper/e4fb9a351c1336fdef4c4c40808bacd96b8d7733

[11] Li Y, Deng L, Liang J, et al. Molecular dynamics simulations reveal distinct differences in conformational dynamics and thermodynamics between the unliganded and CD4-bound states of HIV-1 gp120. Phys Chem Chem Phys. 2020. URL: https://www.semanticscholar.org/paper/614c7d149536e4a97c44e4060aaeeee8dee1f98c

[12] Cao Y, Im W. Conformational Variability of HIV-1 Env Trimer and Viral Vulnerability. bioRxiv. 2025. URL: https://www.semanticscholar.org/paper/e56fdde22fc3cab6082a9a668de5d8aff1600559

[13] Durham ND, Jain A, Howard A, et al. Molecular basis for the increased membrane fusion activity of the Ebola virus glycoprotein A82V variant from the 2013-2016 epidemic: insights from simulations and experiments. bioRxiv. 2024. URL: https://www.semanticscholar.org/paper/0a1f603caff943c78a22deed5e5bfcc99f37cf6f

[14] Shehata M, Casalino L, Duquette M, et al. N-Glycans Modulate HIV-1 Env Conformational Plasticity. bioRxiv. 2025. URL: https://www.semanticscholar.org/paper/6f69d91009385a510b6d8cc41b3dca442bf7888b

[15] Ancy I, Sivanandam M, Kalaivani R, et al. Insights of inhibition mechanism of sifuvirtide and MT-sifuvirtide against wild and mutant HIV-1 envelope glycoprotein41: a molecular dynamics simulation and binding free energy study. 2020. URL: https://www.semanticscholar.org/paper/6d980864de92dea8809c492d20b90e776e91815f

[16] Pan Y, Shlyakhtenko L, Lyubchenko Y. High-Speed AFM directly visualizes conformational dynamics of the HIV-Vif protein complex. bioRxiv. 2020. URL: https://www.semanticscholar.org/paper/5334ad4950b57489c0b716215f5641971d6b8b61 *** Disclaimer: This article is for educational and informational purposes only. It is not intended to substitute for professional veterinary advice, diagnosis, treatment, or regulatory guidance. Always consult a licensed veterinarian or qualified specialist regarding animal health, disease diagnosis, and therapeutic decisions.