An effective vaccine is urgently needed to prevent the human immunodeficiency virus type 1 (HIV-1) epidemic, a disease transmitted to approximately 1.5 million people each year [1]. Currently, lifelong antiretroviral therapy (ART) combinations are the standard of care for the management of HIV in people living with HIV (PLHIV). However, ART does not cure or eradicate the virus from infected cells. Antiretroviral therapy efficiently suppresses HIV viral replication, delays disease progression, restores T-cell immune responses, and improves the quality of life and the survival rate of PLHIV [2,3]. Unfortunately, transmissions still occur either from untreated people living with HIV or from PLHIV on antiretroviral therapy with insufficient virological suppression.

The HIV-1 infection begins with the entry and integration of the viral genome into target cells, which is initiated by the binding of a trimeric envelope (Env) spike on the HIV-1 surface to CD4 receptors and co-receptors CCR5 or CXCR4 on host cells [4]. The binding of Env gp120 to the CD4 binding site on the CD4 T cells induces conformational changes to the gp41 ectodomain which permits membrane fusion and deposition of viral RNA into the host cell cytosol [4,5].

These events result in robust humoral immune responses with the production of non-neutralizing and neutralizing antibodies against the Env trimer to halt the HIV-1 virus from infecting host cells [6,7]. However, HIV-1 possesses several counteracting and subvert mechanisms to evade the antibody effector functions. First, the virion surface displays a few distantly spaced functional spikes, hindering the effective interaction with and activation of B cells [8]. Additionally, the virus produces non-functional uncleaved precursor Env glycoprotein (gp160) and monomeric gp120 and gp41 which act as decoy epitopes that divert the antibody response to non-protective responses [4]. Second, conserved epitopes on the Env trimer are rendered inaccessible to neutralizing antibody responses by an extensive N-glycan shield and conformational masking [8,9]. Third, the Env trimer undergoes rapid mutations that generate enormous sequence variability that prevails over the antibody responses [10,11,12].

During the acute HIV-1 phase, functional non-neutralizing antibodies (nNAbs) against Env are rapidly developed in the infected host [8]. The nNAbs protect against viral spreading through multiple binding events and the involvement of additional immune effector mechanisms. nNAbs are typically strain-specific and exhibit antiviral activity via phagocytosis of infectious virions, antibody-dependent cellular cytotoxicity (ADCC), antibody-dependent complement deposition (ADCD), antibody-dependent virus inhibition (ADCVI), and sequestration of Fc receptor-bearing infected T cells [7,12,13]. Notably, nNAbs produced against HIV-1 primarily target the immunodominant gp41 ectodomain but cannot prevent productive (cis) infection of target immune cells [12].

Generally, in the absence of antiretroviral therapy, PLHIV progresses to AIDS within 8–10 years after infection [14]. In approximately 20% of PLHIV, often referred to as long-term non-progressors (LTNP), disease progression and the development of clinical signs and symptoms are remarkably delayed in the absence of ART [14,15,16,17]. Based on viremic control of HIV-1 viral load, LTNP can be distinguished as viremic controllers and viremic non-controllers [14]. HIV-1 elite-controllers (EC) are a unique subset of approximately 0.3% of LTNP viremic controllers who maintain an undetectable viral load and high CD4 T-cell counts in the absence of ART [17,18]. Elite-controllers exhibit potent broadly neutralizing antibodies (bNAbs) against diverse isolates of HIV-1 which prevent the viral infection of target cells [8]. bNAbs predominately target highly conserved epitope clusters on the Env glycoprotein trimer: CD4 binding site on gp120, V3 loop within the glycan shield, V2 apex on gp120, glycosylated silent face on gp120, gp120/gp41 interface, fusion peptide, and the membrane-proximal external region (MPER) on the transmembrane domain of gp41 [8,10,19,20,21].

Attempts to consistently induce these antibodies to establish either a functional or sterilizing HIV-1 cure through vaccination have been challenging in part due to the unique structural and functional properties of bNAbs. For instance, bNAbs frequently exhibit high levels of somatic hypermutations (SHM) arising from prolonged co-evolution with HIV and are associated with inserts and/or deletions. [22,23] Also, many bNAbs possess exceptionally long heavy chain complementary determining region 3 (CDRH3) or short light chain complementarity determining region 3 (CDRL3), while some exhibit poly-reactivity to host autoantigens. [11,24,25] These immune responses strongly correlate with class I human leukocyte antigen (HLA) polymorphisms on the peptide binding groove which have been associated with the spontaneous control of HIV-1 viral load in elite-controllers. [15] Interestingly, glycan-dependent bNAbs rely on Env trimer architecture and the composition of potential N-glycosylation sites which are considerably conserved among diverse HIV-1 clades and isolates. [26] Alterations in signature glycans which sterically mask conserved epitopes on the Env trimer have been shown to affect viral binding, infectivity, and neutralization [27,28].

Recent advances in glycoimmunology show that the glycan (glycome) repertoire influences innate and adaptive immune responses. [29] Notably, the glycome in tumors, autoimmune disorders, chronic inflammation, and infectious diseases is characterized by aberrant glycosylation. [30,31] Interestingly, several altered glycan structures mediate immune function and cellular processes by influencing interactions at the cell–cell, and cell-pathogen interfaces [32]. For example, in autoimmune disorders (e.g. IgA nephropathy, rheumatoid arthritis, and systemic lupus erythematosus) the lymphocyte surface proteins typically express altered galactose or N-acetylgalactosamine terminated glycans which impact disease pathogenesis and resolution. [29,30] Additionally, the overexpression of β-galactoside binding lectins (galectins 1/3/9) by several tumors (e.g. melanoma, Hodgkin’s lymphoma, pancreatic carcinoma, and neuroblastoma) may facilitate tumor immune evasion and modulate anti-tumour immune responses. [29,30,33] Conversely, aberrant glycans may be utilized as templates for carbohydrate-based vaccines or diagnostic/therapeutic biomarkers in cancer and infectious diseases [32,34,35].

The HIV glycan shield interacts with the host cell glycan-binding proteins (lectins), sialic acid-binding lectins (siglecs), and galectins. These interactions affect cellular processes and immunological responses during HIV infection. [29,31,36] Siglecs (siglec 6/7/9) and galectins (galectin 1/3/9) interactions have been implicated in HIV immune dysfunction through the induction of T/B-cell exhaustion, apoptosis, and NK cell inhibition. [31] Emerging evidence indicates that the aberrant changes to sialic acid, core fucose, and galactose significantly affect the non-neutralizing antibody responses to HIV infection. For example, in elite controllers, antibody Fc-mediated antiviral effector functions (ADCC/ADCVI) are skewed towards agalactosylated, afucosylated, and asialylated glycoforms. [29,31,37] Also, HIV-infected replication-competent T cells exhibit enhanced surface fucosylation which is essential for T-cell receptor binding, activation, and signaling. Structural analyses of isolated bNAbs have elucidated the dependence on signature glycans within the HIV glycan shield to elicit potent heterologous neutralization. [33] In this review, we provide an overview of the glycan-related obstacles to eliciting bNAbs and describe a reverse vaccinology non-cognate ligand strategy (NCLS) using protein scaffold mimicry to develop HIV-1 vaccine immunogens.

Env-glycans are carbohydrate moieties that are added as post-translational modifications to Env proteins produced by host cells before being displayed on the surface of pathogenic viruses. Glycans on Env are vital for protein folding, replication, infectivity and evading host immune responses. [26,38,39] Glycosylation of HIV-1 Env precursor gp160 occurs in the host cell endoplasmic reticulum. It is further processed in the Golgi, where gp160 is also proteolytically cleaved by a furin-type protease to gp120 and gp41 and eventually trafficked to the cell surface for incorporation as a trimer on budding viral particles. [39,40] The Env glycoprotein is a trimer of non-covalently bound gp41/gp120 heterodimers and the sole glycoprotein on the viral surface coded by the HIV genome [41].

Approximately 50% of the gp120 molecular mass is composed of N-linked oligomannose, hybrid, and complex glycans which form a mannose-patch and are wholly or partially recognized epitopes by some bNAbs. [38,39] Typically, 27–33 potential N-linked glycosylation sites (PNGS) per gp120 protomer exist across HIV-1 clades, which generates enormous Env genetic diversity and contributes to the Env surface variability. [38] HIV utilizes the cell glycosylation pathway to display N-linked glycans, which shield conserved epitopes on the Env from immune recognition. [38,41] In transmitted/founder (T/F) viruses which are responsible for establishing HIV infection after mucosal exposure, distinct epitope patterns with higher levels of mannose, sialylation, and core fucosylation have been reported. [42] The T/F glycosylation profile of Env gp120 frequently differs from chronic infection variants by displaying fewer glycans with higher replication fitness. [38] T/F viruses also have shorter hypervariable loops and contain fewer PNGSs than those from the chronic variants [39,42].

The Env crystal structure shows five variable regions (V1-V5) fused with five conserved regions (C1-C5) that are densely glycosylated with N- and O-linked glycans which are prime targets for neutralizing antibodies, and the focus of vaccine development. [23,43,44] Env glycosylation is usually clustered as unprocessed oligomannose N-linked glycans, distinctively divergent from typical mammalian cell glycosylation and concentrated at the intrinsic mannose patch on the outer domain of the gp120 subunit and the trimer-associated mannose patch at the trimer apex. [27,40] These glycan clusters are unique targets for bNAbs and are conserved across different HIV-1 group M clades (Fig. 1). [45] In this section, we describe Env trimer epitopes and highlight conserved glycans relevant to HIV-1 vaccine design.

HIV-1 Env trimer structure and the conserved epitopes targeted by broadly neutralizing antibodies (bNAbs). The Env model is based on the cryo-electron microscopy structural model of the BG505 SOSIP.664 trimer available on RSCB PDB under Acc. No. 4ZMJ

This glycan patch is a relatively well conserved region of virus vulnerability on the intrinsic mannose patch with multiple bNAbs heavy and light chain complementary determining region interactions, thus representing a supersite for antibody neutralization. [46,47] N322-dependent bNAbs (e.g. BG18, PGT121, PGT128, PGT130, PGT131, PGT133, PGT136, and PGT137) directly bind to the glycan shield and the GDIR peptide motif at the base of V3 loop inducing conformational changes, which restricts access to this epitope. Notably, 2G12 and PGT135 are GDIR peptide motif independent. [48] The N332 supersite bNAbs possess unique characteristics: first, the molecular orientation of the N332 site is more accessible to various angles of approach which allows diverse binding by glycan-dependent bNAbs to this conserved epitope; second, N332 glycan-directed bNAbs develop after a relatively short period of infection and show lower levels of somatic hypermutation; third, N332-dependent bNAbs exhibit broad interactions with surrounding glycans increasing surface contact with the Env glycoprotein [47,49].

Extensive neutralization breadth has been shown for glycan-specific bNAbs across HIV-1 viral panels, and the absence of vital PNGS on the variable loops may lead to the development of resistance with loss of infectivity. [50,51] During HIV infection, PNGS deletions contribute to immune escape from strain-specific neutralizing antibodies but can also lead to the formation of bNAb epitopes [52,53].

This domain on gp120 is located at the apex of the trimer where it assumes at least three conformational states (β-strand, α-helix, 310 helix) required for Env trimer stabilization, viral entry, neutralization resistance, and integration into the host cell. [54,55,56] Typically, V1/V2 bNAbs (PG9, PG16, PGT145, and CAP256.09) are elicited by the β-strand conformation and have long CDRH3. [54,57] These bNAbs exhibit 10–20% mutation levels in the variable region of immunoglobulin heavy chain (VH) that target the quaternary epitope on V2 with varying neutralization potency. [58] Interactions between bNAbs and the conserved epitopes on the V1/V2 domain are achieved by the interaction between acidic residues (aspartate and sulphated tyrosine) at the tip of CDRH3 bNAbs and basic lysine-rich residues on the Env spike which provides access to the N160 high mannose glycans, and sialic acid-containing hybrid glycans at positions N156 and N173 (HXB2 numbering). [59] The antibody responses to the V2 apex appear early in natural infection and are developed in several elite-controllers making it a favorable immunogen target albeit difficult to achieve due to the plasticity of Env conformational states during HIV infection. [54,59] In non-human primate, simian-human immunodeficiency viruses (SHIV) challenge experiments, a V2-directed bNAb PGDM1400 demonstrated modest potency and was subsequently transitioned into human clinical trials. Unfortunately, results from clinical trials (NCT03205917) investigating PGDM1400 alone, or in combination with other classes of bNAbs, showed only a temporary reduction in HIV-1 viral load [60].

In the RV144 trial (NCT00223080) where vaccine efficacy was 31.2%, the generation of IgG to V1/V2 loops was associated with a reduction in the risk of acquiring HIV. [55,61] Analysis of the monomeric gp120 immunogens from the RV144 trial also revealed that signature glycans (N135, N141, N156, and N160) on this epitope are vital for inducing antibody responses. [62] Modifications to the heterologous prime-boost approach utilized in the RV144 have demonstrated that heterologous DNA proteins induce sustained humoral responses against V1/V2 loop with broadly binding functional antibodies confirming the durability of this epitope for vaccine development. However, subsequent clinical trials (HVTN705 (NCT03060629) and HVTN702 (NCT02968849)) failed to show similar vaccine efficacy [61].

The V3 loop is obscured by high-mannose and complex N-glycans and is composed of three parts: a crown, stem, and base which contains relatively conserved domains that are vital for HIV-1 entry into target cells via co-receptor CCR5 or CXCR4. [11,63,