Currently, through single memory B-cell sorting from HIV-1-infected patients and micro-neutralization screening of B-cell cultures, many new monoclonal antibodies (mAbs) have been isolated and characterized successfully, such as VRC01, PG9, 3BCN117, NIH45–46, PGT Abs, and 10E8 (1–5). Previously, it was a concern that an increase in the antiviral breadth may be accompanied by a loss in neutralizing activity, but new broadly neutralizing antibody (bNAbs) can offer potent and broad neutralization of diverse HIV-1 isolates (6). These discoveries have led to further optimism that effective vaccines against HIV-1 will be discovered, especially those that induce cross-clade neutralizing antibodies (7). However, how to translate the new knowledge about neutralizing epitopes into immunogens that elicit potent and persistent immunity is still a challenge (8).

Since the new-generation antibodies exhibit a sufficient level of breadth in their ability to neutralize genetically diverse HIV-1 strains, as an alternative approach, passive immunization using broadly neutralizing antibodies can be applied for immunization and therapy for HIV-1. Several broadly neutralizing mAbs have been shown to protect against HIV-1 infection in animal models when administered as monotherapy (9–12). However, as a threat, viral resistance is still present in HIV-1 clinical therapeutics, and HIV-1 has been shown to develop resistance to neutralizing antibodies as well. By now, it has been reported that antiviral activity of new-generation neutralizing antibodies in treating an established HIV-1 infection is limited (10, 11, 13–15). Particularly, viral load rebounds quickly in all infected patients treating either high or low doses of neutralizing therapeutic antibody due to the outgrowth of pre-existing or de novo viral escape variants. Thus, it is very difficult for antibodies to eliminate the virus and virus-infected cells in vivo once the resistant mutants rapidly emerge.

On the other hand, there is no cross-resistance with the currently feasible antiretroviral to give their distinct mechanisms of antiviral action. It has been reported that combining several different antiviral mechanism’s neutralizing antibodies would increase the barrier against resistance in vivo than any one antibody alone (10, 11, 13, 16–18). Thus, novel broadly potent neutralizing antibodies are needed to deter the development of HIV-1 resistance.

Parent neutralizing antibodies can be engineered to improve their antigen-binding affinity, effect functions, improve their half life, and apply new delivery methods in vivo. Thanks to advances in antibody-engineering technology (19, 20), such as variable-region modification, modification of the glycosylation pattern, and design of bispecific antibody (bsAb) formats, it has been possible to create novel optimized mAbs that may improve the antiviral activity of bNAbs or target multiple viral epitopes to overcome neutralization escape by potency and breadth improvement. In addition, using viral vectors or non-viral vector-mediated technology to deliver the antibodies into circulation, or even implantation of an HIV-specific bNAb gene into hematopoietic stem/progenitor cells (HSPCs), can provide anti-HIV mucosal immunity by effectively reprogramming the immune system. These new platforms for engineering antibodies make HIV-1-neutralizing antibodies more available as candidate therapeutics against this virus.

HIV-1 entry into cell is triggered by interaction of the viral trimeric envelope (Env) glycoprotein gp120 with CD4 in T cell surface (21, 22). The key steps include (A) binding of gp120 to its receptor CD4 on T helper cells, with the formation and exposure of the coreceptor (CoR)-binding site in gp120; (B) binding of the HIV-1’s gp120 to CoRs (CCR5 or CXCR4); (C) fusion of the viral and host cellular membranes (21, 23). Therefore, HIV-1 infection elicits specific antibody against both gp120, gp41 of the Env and conformational epitopes induced by trimmer. Meanwhile, viral entry makes targets available for new-generation neutralizing antibodies to inhibit HIV-1 infection (Figure 1).

Broad and potent anti-HIV-1 neutralizing mAbs can inhibit entry of the virus into the host cell by various mechanisms. One mechanism is interfering with the attachment of the virus to host cell surface receptors. This can be achieved when the antibody binds to gp120 on HIV-1 spikes, thereby interfering with the binding of the virus’s gp120 to the host cell’s CD4 (24). The same inhibitory effect is achieved by antibodies targeting T cell receptor CD4 or CoRs, such as CCR5 and CXCR4, thereby interfering with virus interaction with cell membrane or blocking the receptor binding sites for the virus unavailable, resulted in against HIV-1 entry (25). Another neutralizing mechanism is post-attachment/pre-fusion neutralization and interference with gp120-required conformational changes and fusion processes at the target cell membrane.

Several specific sites on gp120 have been shown to be targeted by bNAbs. (A) The CD4-binding site (CD4bs) on gp120 is recognized by a number of broadly neutralizing mAbs. These Abs include VRC01, NIH45–46, 12A12, and 3BNC117 (1, 4, 8, 26–28), which neutralize 80–90% of circulating HIV-1 isolates with high potency in vitro, as well as CH103 and 8ANC131 with relatively less potency; (B) the V1V2 site has been proven to be another site of the spike that is vulnerable to neutralizing antibodies. PG9, PG16, and PGDM1004 are representatives of V1V2-directed neutralizing mAbs that generally compete with one another for antigen (29–31); (C) N-linked glycan-containing epitopes form a site of spike vulnerability, including the gp120 V3 loop, and several nearby glycans, 2G12, PGT128, and PGT121/10-1074 represent the bNAbs that recognize this region (2, 32, 33); (D) there are also bNAbs that target the coreceptor-binding site (CoRbs) on gp120. The exposure of this site, also known as the CD4-induced (CD4i) site, is (as its name suggests) induced by conformational changes that occur as a result of the binding of CD4 to gp120. Antibodies targeting this region are called CD4-induced antibodies, and include 17b, X5, m36, 48d, and E51 (34–36). Based on the length of the IgG heavy chain CDR, most full-size CD4i Abs do not have high antiviral potency because full-size Abs targeting CD4i-specific epitope is sterically restricted during viral entry (37).

The membrane-proximal external region (MPER) of gp41 contains highly conserved residues that also play a critical role in the virus fusion process. Only 671–683 AA as a linear epitopes of this MPER are recognized by potent neutralizing mAbs (38, 39). These bNAbs include 4E10, 2F5, Z13, and 10E8 (3, 40, 41). Although previous studies show that combination of 2F5 and 10E8 protects rhesus against SIV infection, 10E8 still exhibits most potent and broad neutralizing activity compare with other MERP active antibody targeting gp41 of HIV-1. It can neutralize about 92% of circulating HIV-1 isolates in vitro.

Ibalizumab (iMab) is a mAb that has broad and potent activity against HIV-1; it inhibits HIV-1 by binding mainly to domain 2 (D2) of CD4 on host target cells, inhibiting post-CD4 binding events required to infect cells (42). PRO-140 is a monoclonal antibody directed against the CCR5 CoR (CCR5 antagonists), and it potently inhibits CCR5-mediated HIV-1 entry without blocking the natural activity of CCR5 in vitro (43).

Almost all of the potent neutralizing antibodies have some unusual features. The CD4-specific antibodies exhibit unusual maturation with somatic maturation frequencies of 20–32%. The V1V2-specific antibodies display unusually long CDRH3 of 24–32 amino acid. The V3 glycan-specific antibodies show a relatively high somatic mutation CDR H3s with deletions or insertions. The gp41’s MPER-specific antibodies display relatively high somatic mutation. Understanding the targeting sites, inhibition activities, and unusual feature of above antibodies will contribute to generation of more potent neutralizers by antibody engineering.

As we know, the Env-targeted antibodies comprise mainly of binding antibodies without neutralizing activity or strain-specific antibodies after HIV-1 infection or HIV-1 vaccine immunization. The hurdles of neutralizing antibody generation against HIV-1 are mainly caused by conformational mask and diversity of HIV-1 strains. The new-generation bNAbs, being applied for therapy, also faced the limitation of antiviral breadth and potency. Thus, antibody engineering might provide a potential approach to overcome the limitation. The primary role of the antibody’s variable region is to bind to the specific antigen or epitope, resulting in neutralization against HIV-1. The affinity and the specificity of the mAb are two important antigen-binding properties, and engineering these properties has been extensively studied recently to improve the effectiveness of HIV-1 neutralizing antibodies. Meanwhile, new strategies are currently focused on adding more functionality to existing mAbs. That is, there are a variety of strategies being undertaken to enhance the antiviral properties of mAbs.

For HIV-1 neutralization, there are several reasons why it is difficult to find targets for bNAbs (8, 70, 71), including (A) complex genetic diversity of HIV-1 clades, circulating recombinants and mutants; (B) that glycans of gp120 cover the neutralizing epitopes, which makes them inaccessible to antibodies; and (C) that HIV-1 all strains bind to the CD4 molecule, but the CD4-binding site of gp120 is located in a pocket that is difficult for an antibody to access. Thus, optimized activities by HIV-1 neutralizing antibody engineering display one of the potential approaches to solve the above difficulties.

In general, virus neutralization is considered to happen when a neutralizing antibody occupies the adequate epitopes on the viral surface. This “occupancy” model, also known as the “multi-hit model,” indicates that, when neutralizing antibodies achieve enough binding density on a virion, it will result in inhibition of attachment to cellular receptors or fusion processes (72). Here, engineering the antigen-binding properties of neutralizing antibodies provide better inhibitory activity with the “critical binding site” model compare with “multi-hit” model. High affinity/specificity modification, the potent neutralization is determined by its targeting of vital binding sites and be less dependent on achieving high antibody densities against HIV-1. Moreover, engineering antibodies may enable these optimized potent antibodies to be delivered at a lower dose with less frequency in vivo.

In terms of synergistic antiviral activity of bispecific antibodies, binding with higher affinity usually contributes to neutralizing activity of viruses expressing high densities of spikes (73). However, HIV-1 has a limited number of spikes and low density of viral trimers will cause neutralizing antibody to be less efficient for inhibition through interfering with bivalent binding to the HIV-1. Mature HIV-1 particles express only 10–15 randomly distributed viral spikes, and the spikes are, thus, spaced too far apart to be bridged by a bivalent antibody (73, 74), suggesting that only the special geometry of a bsAb can overcome the low density of spikes on HIV-1. Thus, bivalent binding needs rational design to generate heterotypic bivalent (bispecific) binding for HIV-1, such as pro140/10E8 CrossMab. In a word, dual conserved targets of bispecific antibodies diminish the importance of steric hindrance to access the tight crypt during viral entry into cells. The bispecific antibodies also mediate the cross-linking of virions that, when combined with greater activity, results in inhibiting HIV-1 entry into host cells.

For variable region modification, the targeted mutagenesis approach utilizes site-directed mutagenesis against the functional contact region to generate limited collections of the specific variants of the parent antibody. Screening based on divergent HIV-1 Envs could reduce the cross-reactivity to other antigens, broaden the specificity to HIV-1 cognate antigens, overcome genetic diversity of HIV-1, and substantially enhance its inhibitory activity. Meanwhile, the effect of glycan has been studied, and it was found that Fc bearing fucosylated and sialylated glycans have reduced affinities for FcRs (75). It has been reported that an N-linked carbohydrate specific introduction into IgG’s variable region caused optimized solubility (76, 77). Optimized placement of glycan in the antibody variable region, using the steric hindrance against viral entry strategy, markedly improves the antibody’s activity. Indeed, the precise and proper glycan-addition approach may be applied to enhance other bNAbs’ functional activity against HIV-1. As a potential alternative approach, HIV-1 antibody gene delivery technology provides the possibility of high density that antibodies bind to virion through relative stable and persistent expression, resulting in protection against HIV-1 mucosal infection.

HIV-1 presents special hurdles to producing neutralizing antibodies with high potency and breadth. In order to successfully defend against HIV-1, extensive strategies may be required to generate “the better” antibody against virus by applying multiple novel antibody engineering technologies. Variable region technologies, bsAb generation, and antibody delivery would appear to be essential technologies for developing next-generation antibody therapeutics to enhance the magic bullet with radiation against HIV-1.

MS and YL were involved in the review design and writing of the manuscript. HZ performed figure drawing of the manuscript. YS was involved in the manuscript drafting and review for contribution of knowledge and correction, as well as the final draft approval.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.