The Env glycoprotein of HIV-1, known as gp160, plays a crucial role in facilitating viral entry into host cells by mediating the fusion of the viral envelope with the host cell membrane. This glycoprotein is initially synthesized in the rough endoplasmic reticulum (RER) as a precursor, gp160, which cleaves into two subunits: gp120 and gp41. These subunits assemble into a trimeric spike on the virus surface, which is crucial for infectivity and immune evasion [10, 26]. The synthesis of gp160 involves several complex processes, including glycosylation, disulfide bond formation, and quality-control mechanisms to ensure proper folding. N-linked oligosaccharide modifications are added to proteins during their trafficking through the Golgi, where further processing occurs, contributing to the structural complexity and heterogeneity of the glycoproteins. Once cleaved by furin-like proteases in the Golgi, the mature gp120 and gp41 subunits remain loosely associated, forming functional viral spikes [27, 28].

Notably, the low incorporation of these spikes (approximately 10 per virion) on the virus surface aids immune evasion and limits cytopathicity. HIV-1 utilizes specialized structures known as virological synapses for efficient cell-to-cell transmission, a process in which Env plays a crucial role. These synapses enable the virus to propagate between cells, especially in T cells and macrophages, and help it evade immune detection during transfer [29, 30, 31]. The HIV-1 Env endures a multifaceted trafficking pathway essential to the virus’s infectivity. Synthesized in the endoplasmic reticulum (ER) as a 160-kDa precursor, gp160 forms trimers before being transported to the Golgi apparatus for further processing and maturation. Here, it undergoes proteolytic processing, yielding gp120 and gp41, which form trimeric spikes incorporated into virus particles at the plasma membrane. Env undergoes endocytosis through clathrin adaptors and can either be degraded or recycled back to the plasma membrane [27, 30, 31].

Structurally, gp160 comprises various domains, with gp120 containing variable and constant regions, while gp41 features fusion peptides and transmembrane (TM) domains (TMDs). Critical post-translational modifications include glycosylation and disulfide bonding. During Env’s processing, gp120 undergoes folding through distinct stages before its signal peptide cleaves, stabilizing the protein’s conformation [10, 26]. Key structural components, including the TMD and membrane-proximal external region (MPER), play crucial roles in viral entry by undergoing dynamic conformational changes from a prefusion to a post-fusion state. Cryo-electron microscopy (Cryo-EM) and nuclear magnetic resonance (NMR) studies have revealed these structures at lower resolutions, highlighting their importance in vaccine development and therapeutic targeting. Markedly, Env’s cytoplasmic tail (CT) modifications affect the antigenicity and fusogenic activity, with the CT interacting with Gag proteins during viral assembly [29, 32].

HIV-1 gp120 is integral to the HIV entry mechanism, functioning in tandem with the gp41 protein. It binds to the host CD4 receptor and coreceptors CCR5 or CXCR4, triggering viral fusion with the host cell. Beyond the entry, soluble gp120 induces apoptosis in cells like neurons via disrupted calcium homeostasis and oxidative stress, which leads to neuronal death [18, 27, 33, 34]. Structurally, gp120 consists of an inner and outer domain and a bridging sheet, which houses five conserved regions (C1–C5) and five variable loops (V1–V5). The variable regions contribute to immune evasion by altering glycosylation patterns, which shield gp120 from neutralizing antibodies. High-resolution studies reveal gp120’s conformational changes during CD4 binding, which is critical for coreceptor interaction and subsequent viral fusion [10, 26]. Glycosylation and disulfide bridges are crucial for the stability and function gp120, with N-linked glycans playing a key role in immune evasion. Structural studies also highlight the conserved regions responsible for CD4 binding and the role of the V3 loop in coreceptor specificity [35, 36, 37, 38].

The extensive glycosylation of HIV-1 gp120 serves as both a dynamic immune evasion mechanism and a modulator of viral entry efficiency, with recent studies revealing subtype-specific glycosylation strategies that balance these functions. Approximately 50% of gp120’s molecular mass consists of N-linked glycans, forming a conformationally adaptable shield that masks conserved epitopes while enabling receptor binding plasticity [39, 40]. HIV-1 gp120 utilizes glycan microheterogeneity to create a steric hindrance around CD4-binding and co-receptor interaction sites, enabling breathing motions that transiently expose and conceal vulnerable epitopes and stabilizing alternative Env conformations through glycan-protein interactions. Recent analyses using electron-transfer/higher-energy collisional dissociation (EThcD)-stepped-collision energy HCD (sceHCD)-tandem mass spectrometry (MS/MS) identified 18 conserved N-glycosylation sites and 5 O-glycosylation sites across HIV-1 subtypes, with spatial arrangements facilitating coordinated glycan movement during viral entry. Notably, V1/V2 loop glycans (N156, N160) form mobile “glycan gates” that regulate access to the CD4-binding site, while high-mannose clusters at N332/N392 act as decoys for non-neutralizing antibodies and as anchors for broadly neutralizing antibodies (bnAbs), such as potently germline-targeting antibody 128 (PGT128) [41, 42].

Comparative studies [43, 44] of HIV-1 subtypes reveal evolutionary adaptations in glycan placement, with circulating recombinant form (CRF)07_BC exhibiting dense V4/V5 glycans (N406/N413) conferring enhanced resistance to potently germline-targeting antibody (PG)9/PG16 bnAbs, clade C viruses displaying truncated N301 glycans in the V3 loop to increase CCR5 binding affinity, and clade B viruses incorporating sialylated complex glycans that reduce dendritic cell capture. The CRF07_BC strain exhibits a particular dependence on C2-domain glycans (N197/N289) for maintaining envelope integrity, with their removal resulting in complete loss of infectivity, in contrast to clade B viruses, where these sites are more dispensable. Critical glycosylation sites mediating immune escape include N160 (V2), which shields $\beta$20-$\beta$21 strand epitopes targeted by vaccine research cente (VRC)26-class bnAbs; N276 (C3), which stabilizes the gp120-gp41 interface while blocking antibodies targeting the MPER; and N462 (C4), which protects against CD4-binding site-directed neutralization in transmitted founder viruses.

Metabolic labeling study indicate that oligomannose-rich glycoforms (e.g., the Joel R. Haynes-fast low (JR-FL) strain) enhance dendritic cell capture via mannose receptor binding, facilitating trans-infection while reducing antibody accessibility [45]. In contrast, sialylated variants exhibit reduced immunogenicity but maintain infectivity through improved electrostatic interactions with target cells [43]. The glycan shield further modulates viral entry by optimizing co-receptor binding, maintaining Env conformation for CCR5/CXCR4 engagement, restricting protease accessibility at cleavage sites, such as N88/N230, to limit furin-mediated processing, and regulating membrane fusion through V3 loop glycans (N301/N332) that control the timing of fusion peptide exposure. Notably, selective deglycosylation at N386 (C3) enhances 2G12 bNAb binding by 300%, underscoring the potential for strategic glycan removal to expose conserved epitopes for vaccine targeting [39].

Emerging vaccine strategies leverage glycosylation insights, including glycan-hole engineering via directed removal of non-essential glycans (e.g., N197/N289) to focus antibody responses, cell-line optimization using human embryonic kidney (HEK)293-derived gp120 vaccines to preserve natural glycan heterogeneity [40], and subtype-specific immunogen design tailored to predominant regional glycoforms (e.g., mannose-rich vaccines for clade C-endemic areas). These approaches address the evolutionary paradox in which HIV-1 maintains glycan plasticity for immune evasion while conserving essential glycosylation sites for viral fitness. This is exemplified by cross-clade bnAbs, such as potently germline-targeting antibody DM (PGDM)1400, which targets conserved glycopeptide epitopes, validating this structural vulnerability as a promising therapeutic target [41, 42].

A recent study on the HIV-1_CH040 Env trimer demonstrated that partially open Env trimers facilitate efficient viral entry while conferring resistance to bnAbs targeting gp120’s V1/V2 loops and V3-glycan regions. This structural plasticity enables gp120 to adopt intermediate conformations, optimizing immune evasion and receptor engagement. Notably, CH040 Env resists V1/V2-directed bnAbs (e.g., PG9, PGT145) but remains susceptible to CD4-binding site (CD4bs) bnAbs such as VRC01, suggesting that gp120’s variable loops provide epitope shielding without rigidly stabilizing the trimer in a closed conformation [46].

The viral lipid envelope attaches to host cell membranes primarily through the TM glycoprotein gp41. This 345-amino-acid subunit comprises three key regions: a 21-residue TMD, a 172-residue extracellular domain (ectodomain), and a 142-residue C-terminal segment. Within the ectodomain, crucial fusion elements include a polyproline region, a hydrophobic fusion peptide at the N-terminus, two heptad-repeat regions (HR1 and HR2), and a tryptophan-rich membrane-proximal external region (MPER). These elements assemble into $\alpha$-helical coiled-coil structures, essential for viral fusion and host cell entry. Membrane fusion is facilitated by maintaining the disulfide bridge interaction between HR1 and HR2 [18, 26, 37, 47, 48].

The fusion peptide is initially hidden within the gp120-gp41 complex but becomes exposed after gp120 binds to CD4 and coreceptors. This exposure triggers membrane destabilization and the formation of fusion pores, allowing viral entry. The six-helix bundle formed by HR1 and HR2 is critical for this fusion process. It can be disrupted by peptides derived from HR2, such as enfuvirtide (T-20), used in HIV-1 therapy despite its cost and low oral bioavailability. Neutralizing antibodies like 2F5 and 4E10 target the MPER region [18, 49, 50]. Due to the high conservation, the mutations in the TMD anchor region of gp41 can impact Env-mediated fusion. Models of gp41 suggest it forms a single $\alpha$-helix spanning the membrane, though alternate topologies suggest a multi-pass membrane configuration [18, 26, 48]. A recent study identified antiviral drug series (ADS)-J21 as a small-molecule inhibitor that binds the gp41 N-terminal heptad repeat (NHR) pocket, disrupting six-helix bundle (6-HB) formation by targeting conserved residues (Lys35, Trp32). This mechanism confers broad activity against diverse HIV-1 strains, highlighting gp41’s conserved structural elements as therapeutic targets. Additionally, the membrane-proximal external region (MPER) remains a key target for broadly neutralizing antibodies (bnAbs) such as 10E8 and 4E10, which neutralize even resistant strains like CH040 [51].

The CT of gp41 contains lytic peptide regions (LLP1-3), which regulate fusion efficiency through interaction with host membranes and the viral core. SERINC (serine incorporator) proteins, particularly SERINC5, limit HIV-1 infectivity by modulating the viral Env and membrane fusion processes. The role of gp41’s CT in HIV-1 entry is complex and remains debated, with multiple models suggesting different membrane-spanning configurations. Mutations and palmitoylation in this region impact Env incorporation into virions, viral infectivity, and cell-surface expression [52, 53, 54].

The gp120 and gp41 interaction is central to the HIV entry, driven by noncovalent bonds between these two proteins within the trimeric Env spikes. When gp120 binds to the CD4 receptor on host cells, it undergoes a conformational change that disrupts its interaction with gp41, revealing coreceptor-binding sites. This course triggers structural rearrangements in gp41, transitioning from a prefusion to a post-fusion state, which is essential for forming the fusion pore, allowing viral entry [10]. Various studies have examined the domains involved in gp120-gp41 interactions, identifying the inner domain gp120 and the heptad repeat 1 region of gp41 as key interaction points. Amino acid substitutions, such as W596A and W610A, can disrupt this interaction, and certain broad neutralizing antibodies interfere with these dynamics. Nevertheless, the flexibility of Env trimers can impact the antibody-targeted epitopes exposure [18, 37, 52, 54].

The interaction between gp41 and the HIV-1 matrix (Gag) is also crucial, particularly during viral assembly and maturation. Gag facilitates Env packaging during budding and regulates the Env structure during entry [55]. A study has identified the importance of the gp41 C-terminal tail (CT) in binding to Gag, which prevents premature entry into immature particles. Additionally, mutations in the gp41CT have been associated with resistance to HIV protease inhibitors [56]. The HIV-1 Tat protein interacts with the envelope glycoprotein Env, particularly with gp120, inducing conformational shifts that enhance viral entry by facilitating binding to CD4 and coreceptors. A study on CH040 Env revealed that mutations in the gp41 HR1 and HR2 domains compensate for destabilizing alterations at the gp120-gp41 interface, thereby preserving viral fitness. Substitutions such as Q567R in HR1 and S614G in gp120’s V1V2 domain restore trimer integrity, highlighting convergent evolution in maintaining Env functionality. This underscores the cooperative role of gp120’s apex (V1V2) and gp41’s HR regions in stabilizing the trimer during conformational transitions [57]. Tat complexes with trimeric Env modulate viral entry via alternative pathways. Interactions between Tat and gp120 can alter viral tropism and help the virus evade neutralizing antibodies, adding complexity to the immune response against HIV [10, 50, 58].

CD4, a host cell surface glycoprotein receptor comprising four immunoglobulin-like domains (D1–D4), features an ectodomain that extends approximately 115 Å from the cell surface, with D1 situated at the farthest point from the membrane [59]. CD4 plays a crucial role in activating T cells and in the progression of HIV-1 infection. During T cell activation, it binds to major histocompatibility complex (MHC) class II molecules on antigen-presenting cells, facilitating the recruitment of p56lck and tyrosine kinase, which are key components in activating T helper cells and regulating the adaptive immune response. In HIV-1 infection, CD4 is the virus’s primary receptor, initiating viral attachment and forming the coreceptor binding site on gp120. This interaction aligns gp120 on the viral surface with the coreceptor, a seven-TM protein in the cell membrane, requiring conformational adjustments in CD4, gp120, or both [21].

CXCR4 and CCR5 were recognized in 1996 as coreceptors essential for HIV-1 entry. Both are chemokine receptors within the G-protein-coupled receptor (GPCR) family, characterized by 7-TM segments. Coreceptor selection plays a key role in determining viral tropism [60]. Typically, CCR5-tropic (R5) viruses drive initial transmission, whereas CXCR4-tropic (X4) viruses or dual-tropic (R5X4) variants often appear in later stages of infection [61, 62]. Each receptor has a seven-TMs helical structure, an N-terminal segment, three extracellular and three intracellular loops, and a C-terminal tail. Various studies have examined modified CXCR4 constructs with stabilizing alterations and T4-lysozyme fusions complexed with different ligands. Additionally, a CCR5 variant fused with rubredoxin has been investigated in complex with maraviroc, an HIV-inhibiting drug, as well as the chemokine CCR5-[5P7] C-C motif chemokine ligand 5 (CCL5) [21, 60, 63].

GPCR structures typically exhibit a topology characterized by a 7-TM helical bundle [64]. According to the two-site model, the N-terminal domain of CXCR4 or CCR5 forms chemokine recognition site 1 (CRS1), which interacts with the globular core domain of chemokines, while the 7TM helices form chemokine recognition site 2 (CRS2), serving as the binding pocket for the N-terminal domain of chemokines [65]. Although these structures provide valuable insights into the architecture of CCR5 and CXCR4, as well as their interactions with various ligands, they do not offer detailed molecular information on their function as HIV-1 coreceptors [63, 66, 67]. Several other cell-surface receptors, such as dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN), the mannose-specific endocytosis receptor, heparan sulfate, $\alpha$4$\beta$7 integrin, and glycolipids, facilitate viral attachment and spread but do not mediate membrane fusion [21, 68].

The interaction between HIV and host integrins, particularly $\alpha$4$\beta$7, plays a multifaceted role in enhancing viral entry and spread, though its mechanisms differ across cell types. The impact of $\alpha$4$\beta$7 on CD4+ T cells and dendritic cells reveals critical pathways for HIV pathogenesis. In CD4+ T cells, $\alpha$4$\beta$7 targets CCR5-high/CD4+ T cells that are metabolically active and highly susceptible to productive HIV infection, forming complexes where $\alpha$4$\beta$7 is physically close to CD4 receptors, streamlining viral binding and entry. Additionally, HIV gp120 binds $\alpha$4$\beta$7, a gut-homing receptor, enabling the virus to exploit lymphocyte migration to gut-associated lymphoid tissue (GALT), facilitating infection in mucosal sites where massive CD4+ T cell depletion occurs during acute infection [69]. Furthermore, $\alpha$4$\beta$7 promotes viral spread via virological synapses by rapidly activating lymphocyte function-associated antigen-1 (LFA-1) upon binding with HIV gp120, stabilizing cell-cell adhesions, and enhancing viral transmission. Dendritic cells in mucosal tissues, which express adhesion molecules like intercellular adhesion molecule 1 (ICAM-1), bind LFA-1 on T cells, allowing HIV virions captured by DCs via $\alpha$4$\beta$7 or other receptors to be transferred to T cells, amplifying viral dissemination [14, 69]. However, conflicting evidence exists regarding $\alpha$4$\beta$7’s role in transmission, show correlations with HIV susceptibility in some populations, such as heterosexual women, while others, including men who have sex with men (MSM) and people who inject drugs (PWID), show no association or even inverse correlations [70]. Notably, the data focus on T cells and DCs, with no direct evidence linking $\alpha$4$\beta$7 to macrophage entry, suggesting that macrophage tropism may instead rely on other integrins, such as LFA-1 or high-affinity CD4 interactions [14].

HIV-1 entry into CD4+ T cells depends on the interaction between the Env and the coreceptors CXCR4/CCR5, which are determined by the virus’s tropism. This process begins when the Env glycoprotein gp120 binds to surface-expressed CD4, causing conformational changes [94] that enable gp120 to interact with coreceptors. This is a crucial step in initiating the fusion process (Fig. 2, Ref. [94], step 1). Following this, the gp41 subunit triggers fusion by embedding its hydrophobic fusion peptide into the lipid membrane after coreceptor binding and conformational alterations [95] (Fig. 2, step 2). The mechanisms of HIV-1 entry and viral membrane fusion have been thoroughly reviewed in other studies [21, 94, 96, 97]. The virus’s attachment to the plasma membrane is modulated by host proteins, such as P-selectin glycoprotein ligand-1 (PSGL-1) and CD43, which play a role in influencing HIV attachment [98]. HIV-1-encoded Vpu and co-clustered Gag proteins facilitate the target cell membrane attachment by downregulating PSGL-1 expression. Furthermore, interferon-induced TM proteins (IFITMs) impede HIV-1 entry by modulating fusion processes with the host membrane [99, 100].

Host factors such as IFITMs and SERINC5 can restrict the virus’s sensitivity to retroviral envelope glycoproteins [101, 102, 103]. IFITMs serve as critical innate immune restriction factors that inhibit viral membrane fusion, particularly in the context of HIV-1 and other enveloped viruses. These proteins exert their antiviral effects by modulating the biophysical properties of membranes at the sites of viral entry. IFITM3, for instance, induces negative membrane curvature via its amphipathic helix (AH), forming intraluminal vesicles (ILVs) and creating a mechanical barrier that prevents the formation of fusion pores [104]. Concurrently, IFITMs enhance membrane rigidity and lipid order, particularly in lipid-disordered domains, thereby increasing resistance to deformation during viral fusion. Their strategic localization in endolysosomal compartments and the plasma membrane, governed by posttranslational modifications, ensures the proximity-dependent restriction of viral entry [105]. Moreover, IFITMs directly antagonize the HIV-1 Env by impairing its processing, reducing its incorporation into virions, and promoting premature shedding of the surface unit (SU), which collectively diminishes viral infectivity. Additionally, IFITMs trap viral fusion at the hemifusion stage by creating an energetically unfavorable membrane environment, preventing fusion pore formation and promoting viral degradation. Beyond target cells, IFITMs can incorporate into nascent virions, reducing their infectivity independent of specific envelope glycoprotein interactions. Notably, IFITM2 and IFITM3 exhibit stronger antiviral activity than IFITM1, likely due to their predominant localization in the endosome. However, HIV-1 demonstrates adaptability through Env mutations that confer resistance to IFITM-mediated inhibition, highlighting a dynamic evolutionary interplay between host restriction factors and viral evasion strategies. This ongoing co-evolution has profound implications for HIV-1 pathogenesis and transmission, influencing viral fitness and informing potential therapeutic intervention strategies that target IFITM-virus interactions [106].

Upon HIV-1 binding to its receptors and coreceptors, various signaling pathways are triggered and modified (Fig. 3, Ref. [94]). For instance, NACHT, LRR, and PYD domains-containing protein 3 (NLRP3) inhibits F-actin remodeling, which influences susceptibility to HIV-1 infection. The binding of the virus to its receptors through P2Y purinoceptor 2 (P2Y2) signaling triggers the degradation of NLRP3. Without NLRP3, proline-rich tyrosine kinase 2 (PYK2), a protein tyrosine kinase, phosphorylates and activates cytoskeletal rearrangements, facilitating the viral entry (Fig. 3A) [107]. Furthermore, the nuclear factor of activated T cells (NFAT) and NF-$\kappa$B (nuclear factor kappa-light-chain-enhancer of activated B cells) transcription factors, which are essential for T cell activation, are activated by the host-derived p21-activated kinase 2 (PAK2) and the viral protein Nef (Fig. 3B) [108].

In contrast, HIV-1 engagement with its receptor and coreceptors initiates the activation of PLC-$\gamma$ through the myristoylation of Lck at p56 (Fig. 3C). Consequently, phosphatidylinositol (3,4,5)-trisphosphate (PIP3) undergoes hydrolysis, yielding diacylglycerol (DAG) and IP3. IP3 facilitates the opening of Ca²⁺ channels within the endoplasmic reticulum (ER), while DAG stimulates the MAP kinase pathway. Additionally, Vpr facilitates the nuclear import of NFAT and the influx of Ca²⁺ during viral infection. Subsequently, mitogen-activated protein kinase (MAPK) signaling induces the activation of NFAT and extracellular signal-regulated kinase (ERK), thereby enhancing cytokine production through transcriptional regulation of genes and promoting T-cell activation and proliferation [109]. In addition to its functions in modulating cell signaling and evading the immune response, HIV-1 enhances the membrane-bound Fas expression on T-cells and FasL on monocytes, macrophages, and NK cells during infection, triggering apoptosis (Fig. 3D) [110, 111]. Elevated levels of caspase 8 expression result in accelerated apoptosis compared to uninfected cells in vitro. Moreover, apoptosis is further facilitated by virally derived Nef and Tat proteins in the host cytosol, which enhances FasL levels in the plasma membrane and activates caspases 3 and 8 (Fig. 3E) [94, 112]. The binding and fusion of HIV-1 with the host cell activate diverse pathways, spanning multiple cellular processes and ultimately facilitating HIV-1 replication.

In structural biology, remarkable advancements have been made concerning HIV-1 Env and its interactions with cellular receptors. However, constructing a comprehensive molecular model presents challenges, chiefly due to the constraints of the latest structures. These limitations include artificial modifications applied to the Env constructs, which incorporate additional ligands to enhance stability, as well as the absence of membranes. Fig. 4 (Ref. [21]) illustrates a conceptual model depicting the initial stages of HIV-1 entry.

The initial attachment of HIV to the target cell surface begins with CD4 binding to the gp120 component of the Env protein. Although a single CD4 molecule can interact with the SOSIP trimer, this interaction does not trigger substantial structural changes, likely due to modifications like potently germline-targeting antibody 145 (PGT145) and donor strain (DS), which stabilize the trimer and prevent it from shifting into the open conformation typically induced by CD4 binding. Whether the Env trimer complex and single CD4 accurately represent a physiologically relevant transitional state is unclear. A single CD4 can induce conformational changes in Env, both in the presence and absence of PGT145 and with a DS mutation. Despite single-molecule (sm) FRET data suggesting that a single CD4 can elicit substantial changes, evidence from native, untriggered conformations of the prefusion SOSIP trimer disputes this [77]. The understanding of CD4-mediated attachment remains uncertain.

Additionally, the effectiveness of CD4-specific antibodies, such as ibalizumab, which target an epitope near the D1–D2 junction on CD4 opposite the gp120 binding site in inhibiting HIV-1 entry has yet to be conclusively proven [59, 113]. Recent studies employing advanced imaging techniques, including spectroscopy and microscopy, have revealed intriguing stoichiometry during membrane fusion among the Env trimer, CD4, and a coreceptor, possibly involving the oligomerization of both receptors [114, 115]. Previous studies indicate that the Env trimer remains functional even if one or more gp120 protomers cannot bind CD4 or the coreceptor, as a single CD4/coreceptor interaction is enough to initiate HIV-1 infection. Given that there are only approximately 14 Env spikes per virion, it is unlikely that an Env trimer engages three CD4s and three coreceptors simultaneously [19, 77, 116, 117].

A single CD4 molecule can induce conformational shifts in the Env trimer, promoting the formation of the coreceptor binding site. However, the exact role of the coreceptor in guiding further structural changes in gp120, which influence gp41’s fusion potential, remains under investigation. Notably, the recent structure of the CD4-gp120-CCR5 complex shows ambiguous allosteric changes in gp120 that might impact gp41. Comparative studies between gp120 bound to both CD4 and CCR5 versus gp120 bound solely to CD4 show no significant differences in gp120’s core [84, 118]. The main structural adjustments involve repositioning the V3 loop and reorientating gp120’s N- and C-termini near the gp41 interface.

Current understanding suggests that upon dissociation of the HIV-1 Env trimer from gp120, the prefusion state of gp41 [74, 119, 120, 121, 122], which forms the “4-helix collar”, undergoes a refolding process involving its four helices ($\alpha$6-9), resulting in an unstable conformation (Fig. 4A) [120]. It is proposed that the partial or complete dissociation of gp120 may serve as a critical trigger for the structural rearrangement of gp41, ultimately facilitating membrane fusion. In the presence of CD4, helix $\alpha$6 is displaced away from the gp120 termini, creating a pocket that accommodates the fusion peptide, which in turn packs against the N terminus of gp120 (Fig. 4A) [74]. Studies have shown that fusion peptides detach from these pockets due to intrinsic conformational dynamics, allowing one side of the gp41 grip to open. Upon binding to CCR5, the N-terminal segment of gp120 bends back to align with the structure at the binding site, even in the absence of a coreceptor. This act suggests that gp120 can dissociate from gp41 spontaneously or through CD4-induced mechanisms [21]. Further investigation is warranted to elucidate the influence of membranes and the organization of Env trimers on these processes.

In the absence of CCR5 binding, the necessity of a coreceptor, despite gp120 dissociation, to activate gp41 independently of CCR5 binding raises a crucial question. Primarily, gp120 dissociation without a coreceptor might prove nonproductive due to the spatial challenge. When the virion binds to the target cell surface with the Env trimer and interacts exclusively with CD4, the gp41 fusion peptide is approximately 160 Å away from the cell membrane surface (Fig. 4A). Premature dissociation of gp120 would prevent the fusion peptide from reaching the target membrane [120]. gp120 must bind to CCR5 to bridge this gap and facilitate membrane fusion, bringing the fusion peptide within a range of approximately 70 Å from the target membrane (Fig. 4A). This distance aligns with the proposed traversal distance of the fusion peptide during gp41 refolding to reach the target receptor [123, 124]. Consequently, CCR5 serves to promote membrane fusion by stabilizing CD4-induced conformational changes. Moreover, the formation of fusion pores likely necessitates the involvement of multiple Env trimers, as demonstrated in other viral fusion proteins [125]. A stable Env-receptor complex would be advantageous for consistent conformational changes across several Env trimers. Thus, to facilitate productive membrane fusion, a coreceptor stabilizes and anchors the conformation of the Env trimer induced by CD4. Therefore, even from a mechanical standpoint, the designation of the HIV-1 primary receptor and coreceptor appears serendipitous.

Theoretically, a single Env trimer can participate with three CD4 molecules and three CCR5 coreceptors, collectively facilitating trimer activation. This configuration would necessitate binding three CD4s and three coreceptors to a single Env trimer, bringing it close to the cell membrane and enabling productive fusion. HIV-1 employs alternative entry mechanisms that facilitate CD4-independent infection, significantly impacting viral persistence and reservoir establishment [126]. These pathways involve structural adaptations in the viral Env protein and the exploitation of host receptors, such as DC-SIGN, enabling viral dissemination while evading immune detection. Structurally, CD4-independent HIV-1 variants exhibit an enhanced intrinsic reactivity of Env trimers, requiring lower activation energy to transition between conformational states. This allows direct exposure of the coreceptor-binding site (e.g., CCR5/CXCR4) without CD4 priming, thereby facilitating premature gp41 coiled-coil formation and bypassing CD4-induced activation steps. However, this structural flexibility also increases susceptibility to neutralization due to the exposure of vulnerable epitopes, rendering such variants rare in vivo but highlighting HIV-1’s adaptability in selecting entry routes [127]. Additionally, dendritic cells contribute to viral persistence through the C-type lectin receptor DC-SIGN, which mediates HIV-1 capture and transmission via two distinct mechanisms. First, DC-SIGN binds to gp120 through carbohydrate-independent interactions, allowing virions to remain infectious within endosomal compartments for prolonged periods [128, 129].

Second, X4-tropic HIV undergoes low-level replication in DC-SIGN+ cells, generating new viral progeny for delayed transmission [124]. This enables dendritic cells to serve as long-term viral reservoirs, with subsequent trans-infection of CD4+ T cells occurring via synaptic transfer of intact virions or delivery of replication-competent viruses from intracellular compartments [129]. Furthermore, CD4-bypass mechanisms contribute to the formation of a latent reservoir by enabling HIV-1 entry into resting memory T cells, including CD4low central memory (TCM) and transitional memory (TTM) subsets, through DC-SIGN-mediated transfer [130]. These pathways facilitate immune evasion by reducing immune activation via endosomal entry routes that circumvent pathogen recognition receptors [131]. Additionally, latently infected memory T cells persist through IL-7-driven homeostatic proliferation and antigen-independent survival mechanisms [130]. Recent studies indicate that DC-SIGN+ dendritic cells in mucosal tissues preferentially transfer HIV-1 to long-lived memory T cells, thereby establishing stable viral reservoirs that are resistant to antiretroviral therapy [132, 133]. This alternative transmission route may underlie the rapid reseeding of viral reservoirs following treatment interruption, emphasizing the need for targeted therapeutic strategies to disrupt CD4-independent viral dissemination.

Moreover, the configuration of the CD4 molecule within the CCR5 complex, which aligns almost parallel to the TM helices of CCR5 and potentially parallel to the bilayer plane, may have implications for membrane function. If the CD4 ectodomain exhibits sufficient rigidity, the CD4-gp120-CCR5 complex formation could bring a local bend in the membrane. Investigating whether this potential membrane bending phenomenon can account for the observed blockage of CD4 and CCR5 colocalization on cell surfaces induced by gp120 when cholesterol depletion occurs, a condition known to promote membrane curvature, would be essential [134]. Additionally, further research is warranted to explore the potential roles of phospholipids during HIV-1 entry.

The fusion of two bilayer membranes is thermodynamically favorable, yet it encounters a formidable kinetic barrier [135, 136]. Fusion agents, including viral fusion proteins, effectively lower this barrier by harnessing the released free energy during the conformational changes of proteins (Fig. 4B, Fig. 5 (Ref. [137])). The merging of two bilayers into a single monolayer typically follows a general pathway, likely culminating in an intermediate state known as ‘hemifusion’, wherein the proximal leaflets of both bilayers have fused while their distal ones remain separate. This hemifusion intermediate is thought to possess a stalk-like structure (Fig. 4B), and in studies of viral protein-mediated fusion, it appears to play a crucial intermediate role [135, 138]. The kinetic barriers associated with forming and transitioning from this intermediate state are likely considerable.

The evidence accumulated to date strongly suggests that viral fusion proteins facilitate the reduction of the kinetic barriers inherent in membrane fusion. More significantly, they act as catalysts, driving and facilitating the process of membrane fusion, as follows:

Step 1: The initial stage involves the formation of protein bridges between bilayers as the fusion proteins undergo conformational changes (Figs. 4B,5). Across all viral fusion proteins examined thus far, two essential membrane-interacting elements are present: a TM anchor at the C-terminus, which secures the protein within the viral membrane, and a hydrophobic patch that binds to the target membrane. Moreover, these proteins typically adopt a trimeric configuration in their fusion-active state. The fusion reaction commences when the fusion protein binds to a ligand, often protons, in response to low pH in an endosome [96, 139]. Still, cellular or viral proteins sometimes trigger a conformational change that propels each subunit toward contact with the target membrane (Figs. 4B,5). Many fusion proteins are fragments of larger precursors, such as hemagglutinin subunit 2 (HA2) from the influenza virus hemagglutinin or gp41 from the HIV envelope protein. They must shed their N-terminal fragments to initiate the fusion process, which typically contain receptor-binding domains (e.g., HA1 or gp120 in the influenza virus hemagglutinin). While evidence for an extended intermediate is compelling, it remains indirect. The subsequent step involves collapsing into a folded-back conformation, often referred to as a ‘pre-hairpin intermediate’. Depending on the intermediate, the half-life could range from several minutes for HIV-1 gp41 to just a few seconds in other cases. Step 2: Following the bridge’s collapse, the fusion peptide or loop in the target membrane converges with the TM anchor (Figs. 4B,5). This collapse distorts the two bilayers, forming a nipple-like structure around a relatively confined area [140]. The fusion peptide, which perturbs the bilayer and lowers the distortion energy, may further potentiate this membrane distortion. Step 3: The distortion of the individual membranes reduces the energy barrier between separated and hemifused bilayers (not necessarily symmetrically, due to differing anchoring of the fusion protein at the two ends), leading to the formation of a hemifusion stalk between them (Figs. 4B,5). Step 4: A transient fusion pore is generated when the hemifusion stalk opens. During protein refolding, a final conformational change renders the open state irreversible (Figs. 4B,5). While some fusion proteins cause the pores to flicker open and close, others do not. Flickering may depend on the rapidity with which the conformational change is completed. At least two fusion-protein trimers are likely required to execute steps 3 and 4. Nonetheless, the precise number of participating trimers and their interaction dynamics remain topics of ongoing discussion [96, 139].

Central to this process is the formation of a stable six-helix bundle within the gp41 subunit, which provides the necessary free energy to drive membrane fusion. Structurally, the Env complex consists of the receptor-binding gp120 and the fusion-mediating gp41 subunits. The latter contains two functionally significant heptad repeat regions—HR1 (proximal to the N-terminal fusion peptide) and HR2 (adjacent to the transmembrane domain)—which undergo substantial rearrangements during membrane fusion [141]. In its native state, gp41 exists in a metastable prefusion conformation within the Env trimer. Upon CD4 receptor binding and subsequent coreceptor engagement, gp41 transitions through a series of intermediate states before forming the thermodynamically stable six-helix bundle, which juxtaposes viral and host membranes, facilitating fusion [142]. The postfusion six-helix bundle consists of a central trimeric HR1 coiled coil, into which HR2 segments pack in an antiparallel orientation, generating a trimer-of-hairpins configuration essential for overcoming energetic barriers to fusion [141, 142]. Structural studies have identified key intermediate states, including an asymmetric fusion-intermediate conformation stabilized by a MPER-specific neutralizing antibody, suggesting potential sites for therapeutic intervention [143, 144]. Molecular dynamics simulations indicate that the final postfusion conformation is stabilized by hydrophobic interactions between fusion peptides and transmembrane domains, forming an energetically favorable core [143]. The transition to the six-helix bundle is an exergonic process that releases sufficient free energy (estimated at 40–60 kT) to overcome kinetic constraints associated with membrane fusion [142, 143]. Experimental findings reveal that bundle formation is directly coupled to membrane merger, as evidenced by studies employing lysophosphatidylcholine (LPC) to inhibit hemifusion. LPC-mediated arrest of membrane fusion prevented six-helix bundle formation, an effect reversed upon LPC removal, confirming that the free energy released during the transition is a primary driver of pore formation rather than the bundle itself [142, 145]. These insights into gp41 structural transitions have profound implications for HIV therapeutics, particularly in the development of fusion inhibitors and vaccines. Notably, MPER-targeting broadly neutralizing antibodies can stabilize late-stage fusion intermediates, highlighting the potential for interventions that disrupt membrane fusion even after initiation [143].

To exemplify these generalizations, we consider the fusion proteins of three viruses, each with established 3D structures for pre-fusion and post-fusion states: influenza, dengue, and vesicular stomatitis. While traditionally classified as class I, class II, and class III viral fusion proteins [146, 147]. This classification can now be somewhat ambiguous and may obscure more than it elucidates; thus, it’s advisable to avoid it. Viral fusion proteins from paramyxoviruses and alphaviruses also exhibit structures in both conformational states [148, 149, 150, 151].