This study constructed 36 CRF01_AE pseudoviruses which contain full-length env gene of CRF01_AE subtype and these pseudoviruses could be used for neutralization studies and evaluation of vaccines or anti-HIV-1 products in China. The sera of infected or vaccinated individuals were polyclonal antibodies, which can not clearly indicate which epitopes are mainly neutralized by antibodies. Therefore, the use of bnmAbs that can recognize different epitopes can reflect the neutralization effect of pseudovirus on polyclonal antibodies (the sera of vaccinated persons). So we analyzed the the neutralizing activities of 11 bnmAbs against these pseudoviruses. We identified several potential sites that may affect the neutralization activity of bnmAbs in the process.
According to the results of this study, it was hypothesized that bnmAbs recognizing CD4bs(VRC01 and 3BNC117) and MPER(4E10 and 10E8) would have the best neutralizing activity against CRF01_AE pseudoviruses. Although the antibodies were not the same as those we used, Wang’s research showed that bnmAbs target MPER have the best neutralizing, too[12]. Earlier, Chenine et al. reported that bnmAbs recognizing CD4bs and MPER also showed good neutralizing activity against CRF01_AE pseudoviruses from Thailand[13], Japan and Vietnam[14]. Ju et al. first isolated HIV-1 nAbs named F6 from a CRF01_AE-infected donor, it may be more suitable for neutralizing CRF01_AE pseudovirus[15]. We are trying to get F6 now.
VRC01 and VRC03 are two of the bnmAbs which target the CD4 binding site, they are closely related. However, in this study, VRC03 doesn't neutralize most of pseudoviruses, although VRC01 did all. Is the low potency of VRC03 CRF01_AE-specific? Through Xie’s paper, we found that she same situation is true for subtype B virus. As we know that most of the epitopes recognized by vrc03 can also be recognized by VRC01,and VRC01 is able to recognize more epitopes than VRC03. This may be the reason for this results.
The characterization of 36 CRF01_AE pseudovirus variants with 11 bnmAbs identified a number of sites that may affect the neutralization activity of bnmAbs. Antibody 3BNC117 recognizes a discontinuous epitope on gp120, and its neutralization activity is affected by changes at several sites (including P124, T198, V275, A281, R308, S365, Q428, G459, N462, and G471), of which amino acid changes at the G459 site makes the virus more susceptible to neutralization by 3BNC117[16]. Seven of the 36 pseudoviruses used in this study (2019GX001.12, 2019GX001.14, 2019GX001.15, 2019GX001.44, 2019GX001.56, 2019GX001.58 and 2019GX001.59) showed amino acid changes at G459D, while one (2019GX001.15) was resistant to 3BNC117, suggesting that this site may not be the most critical site for neutralizing activity. One pseudovirus, 2019GX014.3, demonstrated a P124Y change and was resistant to 3BNC117, suggesting that this site may have some effect on 3BNC117 neutralizing activity. Pseudoviruses 2019GX001.2, 2019GX001.3, 2019GX001.6, 2019GX001.12, 2019GX001.56 and 2019GX001.59 were obtained from a single patient, and only 2019GX001.2 showed resistance to 3BNC117, with K282R as its specific amino acid change, suggesting that this locus may have an effect on susceptibility to 3BNC117.
Antibody 2G12 is a glycosylation-dependent bnmAb that recognizes glycosylated conformational epitopes, including N295, N332, N386, N392, N397 and N448 [17]; these epitopes have been reported to play an important role in the binding of 2G12 to gp120 [18]. Only two of the 36 pseudoviruses used in this study could be neutralized by 2G12, and the 34 pseudoviruses showing no susceptibility to 2G12 had varying mutations at the above-mentioned loci, suggesting that N295 and N332 play an important role in binding. The two pseudoviruses SZ18S0004.3 and 2019GX030.8 that were highly susceptible to 2G12 carried 339N in addition to 295N and 332N, suggesting that the N339 position may also be important for 2G12 recognition of the virus.
The core sequence of the epitope recognized by the 4E10 bnmAb is located at positions 671–676 with the sequence NWFNIT, with additional important binding related sites. Previous reports have shown that three sites—672W, 673F and 680W—are critical for the neutralization activity of antibodies [19], but the relationship between amino acid mutations in the above sites and neutralization susceptibility in these 36 pseudoviruses showed no regular patterns. In contrast, four strains of 2019GX001, 2019GX001.2, 2019GX001.58, 2019GX001.14, and 2019GX001.15, demonstrated resistance to 4E10, and the simultaneous presence of amino acid changes in G691S in these four strains suggested that this locus may affect the neutralizing activity of 4E10 against the virus.
The epitopes recognized by 10E8 overlap and are more numerous than those recognized by 4E10, and this bnmAb recognizes NWFDITNWLWYIK at positions 671–683 [20], where 671N, 672W, 673F, 676T, 680W and 683K/R are critical for 10E8 binding and neutralization of the virus. It has also been reported that glycosylation at positions 231E and 301 could increase the neutralization efficacy of 10E8, while 693V reduced neutralization by 10E8. Comparison of the sequences revealed that the amino acid changes and neutralizing activity of these 36 pseudoviruses at the above-mentioned loci remained inconsistent, but one pseudovirus, 2019GX001.3 of the 2019GX001 variants, also showed resistance to 10E8, and the only unique mutation site in this variant compared with others was T90A, indicating that this site most likely affected neutralization of the virus by 10E8.
The presence of antibodies in an infected person can drive the virus to mutate, and the antibodies produced in an HIV-1 infected person can stimulate the evolution of the virus that established the infection; even if the patient has not been treated with a bnmAb, it is possible that the same dominant variant in the body will develop quasi-species of escape mutations against bnmAb. For this reason, among others, researchers have tried to use a “cocktail” approach to increase the broad-spectrum and neutralizing activity in bnmAb treatment, which is already available in “3BNC117 + 10-1074” therapy [21]. Among the observed escapes of virus from bnmAbs in this study, the 2019GX001 and 2019GX037 strains both could be neutralized by VRC01, suggesting that the site recognized by VRC01 may be more conserved, and it may be the most suitable bnmAb for the treatment of CRF01_AE. Meanwhile, 10E8 showed the second strongest neutralizing activity. Thida et al. found that VRC01 and 10E8 also had the strongest neutralizing activity against CRF01_AE from Japan and Vietnam[14], suggesting that “VRC01 + 10E8” therapy may be promising for the treatment of CRF01_ AE-infected patients.
Pseudoviruses, as a novel tool with the advantages of safety, high throughput, and the possibility of targeted amino acid mutations, are gradually being used in vaccine evaluation, and their sensitivity may exceed that of viral infection experiments [22].The 36 pseudoviruses from 13 CRF01_AE samples used in this study belong to cluster 1, cluster 2, cluster 4 and cluster 5 respectively. We found that these HIV-1 CRF01_AE pseudoviruses showed higher susceptibility to antibodies against CD4 binding sites and MPER than antibodies against V1V2 and V3 regions. Among them, three bnmAbs—VRC01, 10E8 and 3BNC117—showed the highest neutralizing potency against CRF01_AE variants. For the cluster3, cluster6 and cluster7, we will collect them in the future. Considering the high mutagenicity of the HIV-1 virus, combination bnmAb therapy holds promise for successful treatment of HIV-1 infection.