Human monoclonal antibodies against the HIV-1 envelope glycoproteins are useful tools in the structural and functional analysis of the viral envelope and have crucial roles in guiding the design of prophylactic anti-HIV vaccines. Despite a huge expansion of HIV-1 subtype-C worldwide, the clade-C viruses remain to be one of the least studied subtypes especially in terms of HIV-1 neutralizing antibodies. Using the rationale from previous studies showing that viruses with GPGQ residues at the tip of the V3 crown of the HIV-1 envelope induce potent and cross reactive NAbs as compared to viruses with GPGR motif, we generated here three anti-V3 mAbs from Indian donors presumably infected with subtype-C viruses bearing GPGQ at the V3 crown . The functional analysis of the Abs generated reveals a potent neutralization potential with tier 1 viruses while such activity was limited with the tier 2 viruses tested.
The anti-V3 Abs were selected from EBV-transformed B-cell cultures of 33 HIV-1 infected antiretroviral drug naïve patients using V3C-CTB fusion protein . The advantage of using a conformationally constrained instead of a linear V3 peptide for selection of mAbs from cultures or as animal immunogens has been previously demonstrated [36, 41, 48–50]. We found 1-25% (mean = 6%) of the transformed wells positive for binding with V3C-CTB in the first screening. The characteristic nature of the B-cells from HIV-1 infected subjects and the conditions used to immortalize them apparently affects the number and type of Ab-producing cell lines that grow out . The overall positive percentage of Ab secreting culture wells was relatively good and could be attributed to the high titers of anti-V3 Ab reactivity of the corresponding plasma . In contrast to a high percentage of positive secretors in the initial screening, we were able to stabilize only three (277, 903 and 904) anti-V3 Ab producing B-cell clones. This loss could be in part due to the outgrowth of the non-secretor B cells over the secretor B cells in subsequent steps of secondary screening, cell fusion and dilution cloning process. Moreover, the B cells from HIV-1 infected patients are mostly dysfunctional and polyclonally activated , and such properties have been associated with a low persistence of EBV infectivity [51, 53, 54].
The amino acid sequence variation of V3 across the various HIV-1 subtypes is often related to a differential immune response focused to V3 which is expected to originate due to the subtype specific conformational differences in the V3 region [55, 56]. For instance, the HIV-1 V3 crown residues GPGQ in non-clade B and GPGR in clade B viruses respectively are the major determinants of Ab binding and neutralization [19, 40]. Epitope mapping of the anti-V3 Abs with overlapping V3 peptides reveal that their core epitopes lie in the crown region only. Indeed, the recent immunological and structural studies of anti-V3 mAbs have observed similar pattern of binding, wherein essentially all the anti-V3 mAbs bind to ~14 residues in the V3 crown [57–59]. Two of our anti-V3 mAbs (277, 903) showed binding to subtype-A or C but not to subtype-B derived proteins and peptides while mAb 904 displayed cross reactive binding with subtype-B as well. The binding pattern (in context of clade specific or cross reactive V3 antibodies) of the two anti-V3 mAbs 903 and 904 was similar to binding of polyclonal anti-V3 plasma antibodies from the respective patients, however it was different for mAb 277 in the context of the plasma Abs of this patient (Figure 1A-C). The finding highlights the importance of pre-screening of plasma for binding to peptides from different viral subtypes prior to isolation of mAbs. One plausible reason for the non-binding of mAb 277 to the V3B peptide in contrast to its corresponding polyclonal plasma may be attributed to the higher number of subtype-C specific B-cell clones in the B cell repertoire of this patient, as indicated by its very high binding to V3C (Figure 1C). It may also partly be ascribed to biased selection with a CTB construct containing only V3C sequence, which might allow it to preferentially pick up the B cell clone with clade-C V3 specificity. Interestingly, the anti-V3 mAbs 277 and 903, which show clade-A or C (both having a common GPGQ crown motif) specific binding use the same variable heavy chain gene (VH3-30) whereas the cross reactive mAb 904 uses a different VH gene (Table 1). The finding suggests a possible association of antibody gene usage with epitope specificity, however the number of the mAbs generated in this study is too small for comparison. Remarkably, a recent analysis of anti-HIV Abs has pointed out a preferential usage of VH5-51 gene usage of anti-V3 Abs , and such a preference was later shown to define a conserved antigenic structure within the V3 .
The Ab accessibility on the HIV-1 native virus is often challenged by the glycosylation pattern and epitope masking [39, 62, 63]. This effect has been particularly recognized for the V3 region wherein the neighboring regions including V1/V2 shield the epitopes recognized by anti-V3 Abs . Although studies have suggested that the HIV-1 V3 loop remains accessible on most of the viruses , however this information is limited to subtype-B viruses and remains to be explored for other subtypes. Consistent with the binding to gp120 proteins (Du156.12 and JRFL), the three anti-V3 mAbs were able to bind intact native virions with a similar binding pattern (Figure 2A). The results suggest that V3 epitopes are well exposed over the intact trimeric viruses (Du156.12 and JRFL), and these findings are highly supported by previous work in the literature [36, 43]. The rationale of using same proteins (gp120) and its corresponding viruses (intact virion) for the binding assays was to minimize the effect of both, the V3 loop sequence and the neighboring regions, on the local and global orientation of V3 and on the subsequent binding of Abs.
The anti-V3 mAbs showed potent neutralizing activity against subtype-A and C tier 1 viruses, however this activity was restricted for tier 2, especially the subtype-B viruses. The finding was intriguing given the ability of the anti-V3 mAbs to bind to two representative intact virions of subtype-B (JRFL) and subtype-C (Du156.12), and yet failing to reach IC50 neutralization titers. However, it should be noted that mAbs 903 and 904 which display a better affinity than 277, showed neutralization of up to 19% at 30 μg/ml, though not reaching IC50 neutralization titers, with these viruses (data not shown). Overall, the data suggest that higher concentrations of these mAbs may be effective in neutralization, however that needs to be confirmed in detail. Together, these results suggest that effective concentrations for binding and neutralization may vary substantially, and the high affinity binding by mAb might be critical for neutralization. The data are highly supported by various studies conducted previously [15, 36].