Conserved Role of an N-Linked Glycan on the Surface Antigen of Human Immunodeficiency Virus Type 1 Modulating Virus Sensitivity to Broadly Neutralizing Antibodies against the Receptor and Coreceptor Binding Sites.
Product# 1081 HIV-1 gp120 (ADA)
Product# 1011 HIV-1 gp120 (subtype C)
Product# 1031 HIV-1 gp120 (YU2)
HIV-1 establishes persistent infection in part due to its ability to evade host immune responses. Occlusion by glycans contributes to masking conserved sites that are targets for some broadly neutralizing antibodies (bNAbs). Previous work has shown that removal of a highly conserved potential N-linked glycan (PNLG) site at amino acid residue 197 (N7) on the surface antigen gp120 of HIV-1 increases neutralization sensitivity of the mutant virus to CD4 binding site (CD4bs)-directed antibodies compared to its wild-type (WT) counterpart. However, it is not clear if the role of the N7 glycan is conserved among diverse HIV-1 isolates and if other glycans in the conserved regions of HIV-1 Env display similar functions. In this work, we examined the role of PNLGs in the conserved region of HIV-1 Env, particularly the role of the N7 glycan in a panel of HIV-1 strains representing different clades, tissue origins, coreceptor usages, and neutralization sensitivities. We demonstrate that the absence of the N7 glycan increases the sensitivity of diverse HIV-1 isolates to CD4bs- and V3 loop-directed antibodies, indicating that the N7 glycan plays a conserved role masking these conserved epitopes. However, the effect of the N7 glycan on virus sensitivity to neutralizing antibodies directed against the V2 loop epitope is isolate dependent. These findings indicate that the N7 glycan plays an important and conserved role modulating the structure, stability, or accessibility of bNAb epitopes in the CD4bs and coreceptor binding region, thus representing a potential target for the design of immunogens and therapeutics.
IMPORTANCE N-linked glycans on the HIV-1 envelope protein have been postulated to contribute to viral escape from host immune responses. However, the role of specific glycans in the conserved regions of HIV-1 Env in modulating epitope recognition by broadly neutralizing antibodies has not been well defined. We show here that a single N-linked glycan plays a unique and conserved role among conserved glycans on HIV-1 gp120 in modulating the exposure or the stability of the receptor and coreceptor binding site without affecting the integrity of the Env in mediating viral infection or the ability of the mutant gp120 to bind to CD4. The observation that the antigenicity of the receptor and coreceptor binding sites can be modulated by a single glycan indicates that select glycan modification offers a potential strategy for the design of HIV-1 vaccine candidates.
Although the role of neutralizing antibodies has yet to be determined in the only clinical trial of human immunodeficiency virus type 1 (HIV-1) vaccines that has shown a modest degree of protection, it is generally believed that it would be advantageous for a vaccine to elicit broadly neutralizing antibodies (bNAbs) against diverse primary isolates. Passive administration of neutralizing monoclonal antibodies (MAbs) or bNAbs derived from HIV-1-infected patients has been shown to protect macaques from simian-human immunodeficiency virus (SHIV) infection (1,–5). A fraction of HIV-1-infected individuals (∼20 to 30%) generate bNAbs within 2 to 4 years of initial infection (6,–10). However, generation of bNAbs by active immunization has been a challenge, as no HIV-1 vaccine candidate has successfully elicited antibodies with a similar neutralizing breadth (8, 11).
Nevertheless, broadly neutralizing monoclonal antibodies isolated from selected individuals have helped define the targets of bNAbs. These bNAbs are directed against one of five conserved epitopes on HIV-1 envelope glycoprotein (Env); the CD4 binding site (CD4bs), the membrane-proximal ectodomain region (MPER), carbohydrates on the outer domain, a quaternary structure in the V1 and V2 loops, and a newly described epitope present only in cleaved envelope trimers (7, 11,–13). However, HIV-1 has evolved many protective mechanisms to evade host immune responses, including occlusion of potential targets by glycans (14,–20). These glycans account for ∼50% of the molecular mass of the Env surface antigen (gp120) and may mask nearly the entire surface of Env, including the conserved epitopes targeted by some bNAbs, rendering it difficult to elicit antibodies targeting these sites.
We and others previously reported that removal of a single N-linked glycan located at amino acid position N197 (N7) on gp120 resulted in significantly increased sensitivity of HIV-1 to neutralization (21,–23) and enhanced the ability of Env to induce neutralizing antibodies in macaques (21). However, these studies were based on a small number of isolates. Since the N7 glycan is highly conserved across diverse HIV-1 and simian immunodeficiency virus (SIV) isolates (19, 23, 24), it is of interest to determine if the N7 glycan plays a conserved role in the antigenicity, immunogenicity, and function of HIV-1 Env.
In the present study, we sought to extend earlier observations by examining if the role of the N7 glycan is unique among potential N-linked glycans (PNLGs) in the conserved region of gp120 and determining if the effect of the N7 glycan on Env antigenicity is conserved among diverse isolates of HIV-1. Specifically, we used broadly neutralizing monoclonal antibodies with defined epitope specificities to study the effects of the N7 glycan on the antigenicity of Env from a panel of HIV-1 strains representing diverse clades, tissue origins, coreceptor usages, and neutralization sensitivities. Results from this study indicate that the N7 glycan has a unique and conserved role modulating the exposure or stability of conserved epitopes in the CD4bs and variable loops.
MATERIALS AND METHODS
Construction of N-linked glycosylation site mutants.
Plasmids containing the rev-env genes of individual HIV-1 isolates were used as the template for site-directed mutagenesis using a QuikChange mutagenesis kit (Stratagene). Plasmid encoding QA255.662J Env were kindly gifted by Julie Overbaugh (25). Plasmids encoding 89.6 and ADA were gifted by Joseph Sodroski (26, 27). Plasmids encoding B33, LN40 (28), and JR-FL (29) Envs were gifted by Paul Clapham. A plasmid encoding SF162 was gifted by Leonidas Stamatatos and Cecilia Cheng-Mayer (30), and 1084i was gifted by Ruth Ruprecht (31). PVO.4 was obtained through the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH, from David Montefiori and Feng Gao (32).
To generate glycan mutants of Env, the asparagine (N) residue in the PNLG sequon (N-X-S/T) encoded by most env genes was substituted with glutamine (Q) by site-specific mutagenesis. B33 N197 was replaced with glutamic acid (E) because the N197Q mutant Env did not generate viable virus. The N197 PNLG sequon was restored in QA255 by replacing the proline (P) at position 199 with serine (S) and in JR-FL Env by replacing the aspartic acid (D) at position 197 with N. Conditions for the mutagenesis included a two-step PCR: 95°C for 2 min followed by 30 cycles of annealing and extension at 55°C for 30 s and denaturing at 95°C for 20 s and, finally, annealing at 65°C for 6 min. The mutated rev-env genes were then subcloned into mammalian expression vector pSVIIIΔEnv (33). The results of site-directed mutagenesis were verified by DNA sequencing.
Monoclonal antibodies and polyclonal serum.
The following monoclonal antibodies were obtained from the NIH AIDS Research and Reference Reagent Program: MPER-targeted antibody 4E10 (34) and mannan-targeted antibody 2G12 (35, 36) were from Hermann Katinger, CD4bs-targeted antibody IgG1b12 was from Dennis Burton and Carlos Barbas (37,–40), V3 loop-targeted antibody 447-52D (41,–46) and V2 loop-targeted antibody 697-30D (46,–48) were from Susan Zolla-Pazner, and CD4-induced antibody 17b (49,–54) was from James E. Robinson. Recombinant soluble CD4-183 (sCD4-183), obtained from Pharmacia, Inc., contains the first two domains of human CD4 produced in Escherichia coli (55). CD4bs-targeted antibody VRC01 (56) was obtained from John Mascola, and V1 and V2 loop-targeted antibodies PG9 and PG16 (57) were from Dennis Burton. The CD4-IgG2 was from Progenics Pharmaceuticals, and C5-targeted antibody D7324 was purchased from Aalto BioReagents. The pooled clade B serum was from NABI, and the pooled clade C serum was a gift from Charles Wood.
Infectivity and neutralization assays.
Stocks of Env-pseudotyped viruses were prepared by cotransfection of Env and backbone plasmid in 293T cells (32). pSG3ΔEnv backbone was used to generate clade B Env-pseudotyped virus, and Q23ΔEnv backbone was used to generate clade A and C Env-pseudotyped viruses. All assay stocks were titrated in TZM-bl cells as previously described (32). Infectivity of Env-pseudotyped virus was measured by luciferase reporter gene expression in TZM-bl cells. Neutralization as measured by reduction of luciferase gene expression was performed as previously described (22, 58). Briefly, indicator virus containing 150 50% tissue culture infective doses (TCID50) was incubated with serial 3-fold dilutions of antibody or serum samples in duplicates in a total volume of 150 μl for 1.5 h at 37°C in 96-well flat-bottom culture plates before being added to TZM-bl cells (10,000 cells in 100 μl of growth medium containing DEAE-dextran). One set of control wells received cells plus virus (virus control), and another set received cells only (background control). After a 72-h incubation, 100 μl of cells was transferred to a 96-well white solid plate (Costar) for assays of luciferase activities using BrightGlo substrate solution as described by the supplier (Promega). Neutralization activity was expressed as the concentration of antibodies or reciprocal serum dilution that resulted in 50% reduction (IC50) of relative light units (RLU).
Fortebio Octet biolayer interferometry.
Fortebio Octet biolayer interferometry was performed as previously described (59). Briefly, anti-human IgG Fc capture tips (Fortebio) were prewet for 5 min in HBS-EP running buffer (GE Healthcare) immediately prior to use. The microplates used in the Octet were filled with 200 μl of sample or buffer per well and agitated at 1,000 rpm. Tips were then saturated with 10 μg/ml of CD4-IgG2 and then washed with running buffer before being transferred into 2-fold dilution series of gp120, where the association rate constant (ka) (1/Ms) was read. The tips were then transferred into running buffer to measure the dissociation rate constant (kd) (1/s). The tips were regenerated using 0.1 M glycine, pH 1.5 (GE Healthcare), before being used on subsequent samples. Values from a CD4-IgG2-coated tip that was transferred into running buffer alone were subtracted from all test values. Additionally, one uncoated tip was incubated with gp120 alone to ensure that there were no interactions between Env and the capture tips. All interaction analyses were conducted at 25°C.
Viral protein capture assays.
The p24 concentration of a given pseudovirus preparation was measured with an HIV-1 p24 enzyme-linked immunosorbent assay (ELISA; Zeptometrix) and compared to its TCID50. Env capture assays using D7324 were carried out as previously described (43, 60), with slight modifications. In brief, 96-well plates were coated with 10 μg/ml of sheep anti-gp120 capture antibody D7324 (Aalto Bio Reagents Ltd., Dublin, Ireland) overnight at room temperature, blocked with phosphate-buffered saline (PBS)–0.05% Tween–5% milk, washed with PBS–0.05% Tween, and incubated for 2 h at 37°C with wild-type (WT) or N7 mutant Env from one of three sources: gp120 generated by recombinant vaccinia virus-infected cells (61), detergent-lysed pseudovirus, or whole pseudovirus. SF162 or JR-FL Env captured on plates was washed before monoclonal antibody (MAb) 697-30D (at 30 μg/ml or 3 μg/ml, respectively) was added and incubated for 1.5 h at 37°C. After a washing, the bound MAbs were detected using horseradish peroxidase (HRP)-conjugated anti-human IgG (Thermo Fisher) followed by incubation with TMB substrate (Sigma). After 2 M H2SO4 was used to stop the reaction, MAb binding was quantified by optical density readings at 450 nm (OD450). Binding of antibody to each virus was normalized to OD450 readings obtained with HIV-positive polyclonal serum for the same virus. This controlled for possible differences in the number of virions captured. Separate wells included supernatant from parental vaccinia virus-infected cells or ΔV2 gp120 as negative controls (data not shown). Binding of WT virus was included on each plate and set as 100% binding. All experiments were repeated three times in duplicates.
Prism (GraphPad Software, Inc.) was employed for all statistical analyses, all using two-sided tests. The Wilcoxon matched-pairs signed-rank test was used to compare equilibrium dissociation constant (KD), picograms of p24/TCID50, and percent binding values.
The N7 glycan is unique among PNLGs in the conserved regions of gp120 in its ability to modulate the sensitivity of HIV-1 to CD4bs-specific antibodies.
Previous work demonstrated that the N7 glycan modulates sensitivity of HIV-1 to CD4bs-specific neutralizing antibodies (21,–23). To examine if the same effect is exerted by other conserved glycans, we introduced mutations replacing the asparagine (N) residue with glutamine (Q) at each of the PNLG sequons (N-X-T/S) in the conserved regions of surface antigen gp120 of HIV-1 89.6. For this study, we focused on mutant Env proteins that are expressed and processed normally and retain overall functional integrity as indicated by successful rescue of viable virus (data not shown). The N262 glycan mutant did not generate viable virus and thus was not included in further experiments. Neutralization sensitivity of pseudotyped viruses bearing WT or mutant Env was examined in a TZM-bl cell assay against sCD4, CD4bs, and MPER-directed antibodies as well as pooled HIV-positive serum. Results in Fig. 1 show that the N7 glycan is unique among PNLGs in the conserved regions of gp120 in its ability to modulate virus sensitivity to CD4bs-specific neutralizing antibodies. Removal of selected single N-linked glycans modified the sensitivity of virus to certain antibodies; e.g., the N361 mutant was modestly more sensitive to IgG1b12. However, only removal of the N7 glycan increased the sensitivity of 89.6 to all CD4bs-directed antibodies tested as well as HIV-positive pooled serum (21). This indicates that the N7 glycan plays a unique role among conserved glycans on gp120, modulating the stability of or access to the CD4bs in the 89.6 Env.
Increased sensitivity of N7 mutant viruses to CD4bs-specific neutralizing antibodies is conserved among diverse isolates of HIV-1.
Previous work examined the role of the N7 glycan in only a limited number of isolates (21,–23). Because the sequon for the N7 glycan is present in more than 95% of available HIV and SIV sequences (19, 23, 24), we examined if the N7 glycan plays a conserved role in diverse isolates of HIV-1. To test this, we modified the N7 glycan site on the Env from a panel of HIV-1 strains representing different clades, tissue origins, coreceptor usages, and neutralization sensitivities (Table 1), including two viruses, QA255 and JR-FL (25, 29), that lack the N7 glycan site in the Env of WT viruses. To determine the effects of the N7 glycan on the panel of Env, we used site-directed mutagenesis to remove the N7 glycan site of most Env proteins by replacing the asparagine (N) with a glutamine (Q), or to restore the N7 glycan site for those Env missing this site in their WT sequences (see Materials and Methods). Plasmids expressing WT or N7-modified Envs were then used to generate pseudoviruses. The neutralization sensitivity of these pseudoviruses was first tested using antibodies that recognize CD4bs or CD4 induced (CD4i) epitopes (Table 2).
a Input virus was normalized by their infectivity at 150 TCID50 in TZM-bl cells. The column subheadings indicate the respective monoclonal antibodies and epitopes they recognize. N.D., not determined. Results represent the averages of two or more experiments, each performed in duplicates.
b The presence (+) or absence (−) of the N7 glycan is indicated. Unless otherwise indicated, the N residue in the PNLG sequon was replaced with Q. “N197S” indicates an N-to-S substitution at the same (197) position.
c Fold change was calculated as the ratio between the IC50s with N7− and N7+ viruses. Only values greater than or equal to 3 are shown. Increased neutralization sensitivity greater than 4-fold is highlighted in red; decreased sensitivity greater than 4-fold is highlighted in green.
d A subneutralizing concentration of sCD4 was used for these assays. Because the viruses exhibit different sensitivities to sCD4, 20% of the IC50 was used for each individual virus.
All blood and lymph node isolates tested demonstrated increased sensitivity to sCD4 and neutralizing antibodies directed against the CD4bs when the N7 glycan was absent. This includes VRC01, which has a small footprint (56, 62), indicating that the N7 glycan affects the access to or stability of the core residues in the CD4bs.
While the absence of the N7 glycan increased the sensitivity of blood- and lymph node-derived HIV-1 to sCD4, such an effect was not observed in central nervous system (CNS)-derived viruses (Table 2). However, the absence of the N7 glycan did increase the sensitivity of JR-FL and B33 to VRC01. The same effect was observed for all CNS isolates tested against IgG1b12, indicating that the N7 glycan controls the access to or the stability of the CD4bs in CNS isolates as well as blood and lymph node isolates.
To confirm that these effects are due to the absence of the N7 glycan rather than the specific amino acid present at the PNLG site, we introduced different amino acid substitutions at the N7 sequon of a blood isolate, ADA, and a lymph node isolate, LN40. Replacement of N197 with serine in LN40 (LN40 N197S) resulted in an increase in sensitivity to CD4bs antibodies similar to that found with the N197Q mutation, indicating that the effect is not specific to the amino acid change. However, the same N197S substitution in ADA Env resulted in increased neutralization resistance to b12 and no change in sensitivity to VRC01. Examination of the Env sequence of ADA showed that the serine substitution at N197 created an upstream PNLG site at N195 (Fig. 2). Thus, the restoration of neutralization resistance resulting from this upstream PNLG is consistent with the role of a glycan near the N7 site capable of modulating CD4bs exposure.
It has been reported that glycan modification in conserved regions of HIV-1 gp120 reduced viral infectivity (21, 23, 63). To determine if the increased sensitivity of N7 mutants to CD4bs antibodies is associated with reduced infectivity of the mutant viruses, we examined the infectivity of the N7 glycan mutant viruses that showed the greatest increase in neutralization sensitivity to CD4bs antibodies compared to their WT versions. We measured the TCID50 of these pseudotyped viruses normalized to their p24 content. Results in Fig. 3 show that the removal of the N7 glycan did not impact the infectivity of these viruses, supporting the notion that the increased neutralization sensitivity of N7 glycan mutants to CD4bs antibodies was not due to their decreased infectivity.
Removal of the N7 glycan does not alter the kinetics of binding of gp120 to CD4.
To determine if N7 glycan removal alters the affinity of Env protein to its receptor, we examined the kinetics of binding of CD4-IgG2 to gp120 proteins with or without the N7 glycan. Monomeric gp120 proteins from ADA, JR-FL, and PVO.4 were purified as previously described (61), and their kinetics of binding to CD4-IgG2 were compared by biolayer interferometry (Table 3). Different gp120 proteins showed different KD values, in agreement with observations from previous studies (64,–66). However, the presence or absence of the N7 glycan did not affect the KD values of gp120, indicating that N7 glycan mutations did not alter the properties of binding of monomeric gp120 to its receptor, consistent with the lack of effect of the N7 glycan on viral infectivity.
|Env||N7 glycana||KD (M)||ka (1/Ms)||kd (1/s)||P valueb|
|ADA||+||4.74E−09||5.13E + 04||2.46E−04||0.25|
|−||6.85E−09||4.21E + 04||2.91E−04|
|JR−FL||+||3.64E−09||7.19E + 04||2.61E−04||0.25|
|−||2.72E−09||6.69E + 04||1.81E−04|
|PVO.4||+||4.02E−08||4.59E + 04||1.80E−03||>0.99|
|−||4.01E−08||5.52E + 04||2.20E−03|
Effect of the N7 glycan on virus sensitivity to CD4-induced neutralizing antibodies.
It has been demonstrated that removing the N7 glycan increases the sensitivity of HIV-1 not only to CD4bs-directed antibodies but also to CD4-induced (CD4i) neutralizing antibodies, potentially by enhancing the stability of or access to epitopes recognized by CD4i antibodies (22). Furthermore, CD4-independent viruses, which presumably have Env in an open conformation upon CD4 binding, are significantly more sensitive to neutralization by sCD4, similar to our finding with N7-deglycosylated viruses (21, 22). Therefore, we examined the sensitivity of the panel of HIV-1 strains to CD4i antibody 17b in the presence or absence of subneutralizing concentrations of sCD4 (Table 2). While the N7 glycan modulated sensitivity of virus to CD4bs antibodies, its effect on viral sensitivity to CD4i antibodies appeared to be isolate dependent. Increased neutralization sensitivity to 17b was observed only in N7 mutants of 89.6, ADA, and LN40 N197S. The N7 mutant of a tier 3 isolate, PVO.4 (67), was only neutralized by 17b, with IC50s close to 20 μg/ml, which was the upper limit of the assay, making it difficult to determine if the N7 glycan has any effect on neutralization sensitivity to 17b. All other blood and lymph node viruses tested, either with or without the N7 glycan, were resistant to neutralization by 17b, in either the presence or absence of sCD4.
For the few CNS isolates tested, N7 glycan removal did not modulate the neutralization sensitivity to 17b; however, we also observed an isolate-dependent effect of the N7 glycan on virus sensitivity to neutralization by 17b. The presence of subneutralizing concentrations of sCD4 augmented the neutralizing sensitivity of B33 to 17b, but no change was seen in SF162 or JR-FL. JR-FL was resistant to neutralization by the highest concentration of 17b used in the presence or absence of sCD4.
We also investigated if the increased sensitivity to CD4i antibodies correlated with the ability of the mutant viruses to infect CD4-negative cells. Infectivity of pseudotyped viruses bearing WT or mutant Env on CD4-negative cells (NNP2-CCR5 for R5-tropic viruses or U87-CXCR4 for 89.6 ) was determined using inocula normalized for their infectivity in isogenic CD4-expressing cells (NP2-CCR5-CD4 or U87-CXCR4-CD4, respectively) (Fig. 4). Increased infectivity in CD4-negative cells was observed only in ADA N197S and 89.6 N7 glycan mutants. Therefore, the ability of N7 mutant Env to mediate CD4-independent viral entry appears to be isolate specific and dependent on the specific amino acid substitutions at the PNLG site.
Effect of the N7 glycan on neutralization sensitivity to V3-specific monoclonal antibodies.
The increased susceptibility of the N7 mutants to CD4i antibodies indicates the possibility of enhanced exposure or stability of the bridging sheet. Indeed, previous work has shown that removing the N7 glycan results in increased sensitivity of HIV-1 to antibodies targeting the V3 loop and MPER (21, 22). Additionally, the crystal structure of soluble, cleaved trimers indicates that the crown of the V3 loop is buried under the N7 glycan of an adjacent protomer (68, 69). We examined if the removal of the N7 glycan correlated with increased exposure or stability of neutralizing epitopes in the V3 loop, which is a major determinant of coreceptor usage. Neutralization sensitivity to 447-52D, an antibody that recognizes the tip of the V3 loop (42), was determined for pseudotyped viruses with or without the N7 glycan (Table 4). Interestingly, removal of the N7 glycan increased the susceptibility of all viruses neutralized by 447-52D, including the CNS-derived isolates. Even PVO.4 and JR-FL, two isolates that are resistant to 50 μg/ml of 447-52D when the N7 glycan is present, were readily neutralized by the same antibody when the N7 glycan was removed. This indicates that the N7 glycan alters the accessibility or stability not only of the CD4bs but also of the coreceptor binding site.
Effect of the N7 glycan on neutralization sensitivity to monoclonal antibodies directed to quaternary neutralizing epitopes in V2.
Since the N7 glycan is located at the C-terminal stem of the V2 loop, we examined if the removal of the N7 glycan would affect virus sensitivity to monoclonal antibodies PG9 and PG16, which recognize a glycan-dependent, quaternary neutralizing epitope in V2 (70). Most of the isolates in the test panel lack the N160 glycan recognized by PG9 and PG16 and are resistant to neutralization by these antibodies (Fig. 2). Therefore, we focused on isolates that are sensitive to PG9 and PG16. Results in Table 4 show that the N7 glycan also affected the sensitivity of HIV-1 to these monoclonal antibodies, but in an isolate-dependent manner. PVO.4 showed increased sensitivity to PG9 and PG16 when the N7 glycan was absent, whereas the other N7 mutant viruses became more resistant. Interestingly, the introduction of a PNLG site upstream of the N7 site in ADA N197S restored sensitivity to PG9 and partially so to PG16, again indicating that a glycan at or around the N7 glycan site may play a role maintaining the quaternary structure recognized by PG9 and PG16. This isolate-dependent effect of the N7 glycan on neutralization sensitivity to PG9 and PG16 is in contrast to its effect on the sensitivity to CD4bs and V3-directed neutralizing antibodies, where an increased sensitivity was observed in all isolates tested. In this context, it is of interest to note that the panel of isolates tested have various V2 loop sequences, amino acid lengths, and numbers of glycosylation sites, including the presence or absence of N160 (Fig. 2), which is a part of the epitope targeted by PG9 and PG16.
In order to determine if the effect of the N7 glycan on sensitivity to V2-specific antibodies is limited to those isolates that maintain the N160 PNLG, we examined WT and N7 mutant Env from SF162 and JR-FL for reactivity to monoclonal antibody 697-30D, which also recognizes a conformation-dependent epitope in V2 but does not rely upon the quaternary form of Env or the presence of the N160 glycan (Fig. 5) (47). The binding of 697-30D to WT and mutant Env was determined by antigen capture assays using three forms of the Env that differ in their conformational and oligomeric structures: monomeric gp120s and proteins from lysed pseudovirus and whole pseudovirus, with the last exhibiting the most sterically constrained forms of Env (71,–75). Binding of 2G12 and HIV-positive serum was detected with all three forms of Env tested (data not shown) and was therefore used as positive controls. Vaccinia virus-infected cell lysate was included as a negative control. All values described were normalized to binding by HIV-positive serum and were expressed as percentages of binding to the WT Env.
N7-deglycosylated SF162 Env from lysed virions and monomeric gp120s showed significantly reduced binding to 697-30D compared to that of their glycosylated counterparts (P ≤ 0.01) (Fig. 5A). However, there was no difference in 697-30D binding between glycosylated and deglycosylated Env when whole SF162 virions were used, indicating that the N7 glycan may alter the conformation of the V2 loop in the less constrained gp120 but not in the trimeric form on whole virions where there is reduced local flexibility. However, no differences were observed between N7-glycosylated or deglycosylated forms of JR-FL Env, as presented in gp120, whole virions, or lysed virions (Fig. 5B), possibly as a result of the more constrained nature of the V2 loop in JR-FL compared to that in SF162. Again, these findings support the notion that the N7 glycan affects the antigenicity of the V2 loop in an isolate-specific manner.
Effect of the N7 glycan on neutralization sensitivity to monoclonal antibodies directed to epitopes in the mannan cluster and the MPER.
To examine the effect of the N7 glycan on the antigenicity of distal parts of the Env, we tested the sensitivity of WT and N7 mutant viruses to mannan-dependent and MPER-specific antibodies, 2G12 and 4E10, respectively (34,–36). There was little to no change in 2G12 or 4E10 sensitivity when the N7 glycan was removed (Table 4). While QA255, 89.6, ADA, and LN40 showed ∼3-fold sensitivity difference to 2G12 and PVO.4 and SF162 exhibited a modest (∼4-fold) increase in sensitivity to 4E10 when the N7 glycan was removed, other viruses demonstrated no change. These data indicate that the N7 glycan plays less of a role in these distal epitopes than in the CD4bs and V2 and V3 loops.
The N7 glycan modulates virus sensitivity to antibodies from HIV-positive individuals.
The highly conserved nature of the N7 glycan indicates an important role for virus survival. To demonstrate this potential role, we examined the sensitivity of the WT and N7 mutant viruses in the panel of HIV-1 isolates against pooled serum samples from HIV-positive individuals. As expected, there was a wide spectrum of neutralization sensitivities among the viruses tested against the clade B and clade C serum pools (Table 4). However, most of the viruses tested, regardless of their neutralization sensitivities, showed 3- to 60-fold-increased sensitivity to either the clade B or clade C HIV-positive pooled sera when the N7 glycan was absent, consistent with a conserved role of the N7 glycan in the evasion of host neutralizing antibody responses. The only exception was a clade A isolate, QA255, for which the N7-glycosylated form was highly resistant to the clade B serum used (1:20), making it difficult to determine any increase in sensitivity for the N7-deglycosylated virus.
Glycans on HIV-1 envelope proteins have been shown or postulated to play an important and multifaceted role in viral replication, including structural integrity of the virus, coreceptor usage, evasion of host immune responses, and as targets for bNAbs (21, 23, 57, 70, 76). Here, we report a conserved role for the N-linked glycan at amino acid position 197 (N7) in modulating the antigenicity of the CD4bs and epitopes in the V3 loop, a major determinant for coreceptor usage. This role is conserved among a panel of viruses tested, including isolates of different clades, tissue origins, coreceptor usages, and neutralization sensitivities (Tables 2 and and4).4). The N7 glycan is also unique among PNLGs in the conserved regions of HIV-1 in modulating the sensitivity to CD4bs-directed neutralizing antibodies while maintaining the integrity of distal epitopes (2G12 and MPER). The increased sensitivity of N7 mutants was not due to a decrease in viral infectivity (Fig. 3) or to any defect in the ability of the N7 mutant gp120 to bind the CD4 receptor (Table 3). The observation that N7 mutants also showed increased sensitivity to pooled HIV-positive serum is consistent with a role for the N7 glycan in viral evasion from the host immune responses in vivo (Table 4).
It is of interest that the absence of the N7 glycan in CNS versus blood or lymph node isolates resulted in differential sensitivities to sCD4. While the blood and lymph node isolates demonstrated increased sensitivity to all CD4bs-directed antibodies when the N7 glycan was absent, the same mutation in CNS isolates showed no change in sensitivity to sCD4 (Table 2). This differential sensitivity may be derived from the immune-privileged environment from which CNS isolates were obtained; i.e., there is little selective pressure for CNS isolates to evade antibodies against the CD4bs (77, 78). CNS isolates typically display a higher avidity for CD4, requiring lower levels to infect cells, and are more resistant to blocking by anti-CD4 antibodies than spleen-derived envelopes (79). Interestingly, the V1-V2, V3, and C2 regions, which include the N7 glycan site, determine the low CD4 dependence and high avidity for CD4 in CNS isolates (80). These factors may contribute to the unchanging sensitivity of CNS isolates to sCD4 when the N7 glycan is removed. However, removal of the N7 glycan did result in increased sensitivity of CNS viruses to other CD4bs-directed antibodies, e.g., VRC01 or IgG1b12, and to V3-directed antibody 447-52D (Tables 2 and and4),4), thus conforming to the conserved role of the N7 glycan masking these conserved epitopes.
While the role we ascribe to the N7 glycan is through inference by site-specific mutation with a conserved amino acid substitution (N to Q) at the PNLG sequon, several observations support our notion that the effect of the N7 mutant is due not to the amino acid substitution per se but rather to the absence of the glycan at this sequon. As shown in Table 2, the N197Q and N197S mutants resulted in similar changes in the sensitivity of LN40 to sCD4- and CD4bs-specific antibodies VRC01 and IgG1b12. Interestingly, the same amino acid substitutions in ADA resulted in divergent phenotypes, with the N197S mutant exhibiting a sensitivity to VRC01 comparable to that of the WT virus and increased resistance to IgG1b12, while the N197Q mutant showed increased sensitivity to these antibodies. It is noteworthy that the replacement at amino acid position 197 with S resulted in the creation of a new upstream PNLG sequon at N195 (Fig. 2). Thus, we interpret the divergent phenotype of the N197Q and N197S mutants as evidence supporting a conserved role for an N-linked glycan at the base of the V2 loop: either the N7 glycan in the case of the WT ADA or the newly created glycan at N195 in the case of N197S. The differences in the apparent molecular weights of WT and N7 mutant Env expressed in recombinant vaccinia viruses also support the occupancy of the N7 PNLG in all the isolates tested (data not shown).
Although the N7 glycan modulates the receptor and coreceptor binding site in a conserved manner, its effect on the ability of the Env to mediate CD4-independent entry is specific to the isolate tested and the specific amino acid change introduced at the PNLG site. Data from this and previous studies showed ADA N197S and 89.6 as the only two isolates that display enhanced ability to mediate CD4-independent entry when the N7 glycan is absent (Fig. 4) (21, 22). In these cases, the enhanced ability to mediate CD4-independent entry correlated with increased sensitivity to CD4i antibody 17b (Table 2) (22). However, for other isolates tested, the presence or absence of the N7 glycan does not affect 17b sensitivity, even though increased susceptibility to a V3 antibody, 447-52D, was observed (Table 4). Furthermore, as in the case of ADA, the creation of an upstream PNLG sequon in the N197S mutant resulted in the acquisition of CD4-independent phenotype, whereas the absence of the N7 glycan in N197Q retained CD4 dependence (Fig. 4). Thus, even though the N7 glycan modulates the access or stability of epitopes in the CD4bs and coreceptor binding sites, neither the N7 glycan nor susceptibility to 17b per se is predictive of CD4 independence in an isolate-dependent or sequence-specific manner.
The N7 glycan also modulates the sensitivity of virus to neutralizing antibodies recognizing a quaternary structure in V1-V2. Interestingly, this effect appears to be isolate dependent (Table 4) (81). PVO.4 was the only isolate that demonstrated increased sensitivity to PG9 and PG16 antibodies when the N7 glycan was removed, while the same glycan mutant in every other virus tested showed increased resistance. Since PVO.4 has the longest V2 loop length and the highest number of potential N-linked glycan sites, we speculate that PVO.4 may have a higher degree of structural constraint in the V2 loop than other isolates tested. If so, the conformation-dependent epitope in V2 recognized by PG9 and PG16 may be better preserved in PVO.4 than in the other isolates. Thus, when the N7 glycan is removed, the better-preserved PG9 and PG16 epitope in PVO.4 Env may be better exposed (and thus the higher sensitivity to these MAb), whereas the greater flexibility in other isolates may result in the loss of this epitope (and thus greater resistance). Furthermore, we note that the isolate-dependent effect of the N7 glycan on the epitopes recognized by V2-specific MAb is not dependent on the presence or the nature of the N160 glycan in the isolates tested. This is because the N7 glycan also affects Env binding to a V2-specific, conformation-dependent antibody, 697-30D, the epitope of which is not dependent on the N160 glycan (Fig. 5). Thus, the isolate-dependent effect of the N7 glycan on V2 antigenicity is likely mediated through its effect on the positioning or the conformation of the V1-V2 loop.
Currently, we do not have a clear understanding of the basis of the effect of the N7 glycan reported in this study. As shown in Table 3, the presence or absence of the N7 glycan does not affect the affinity of gp120 to CD4-IgG2. This observation is consistent with previous publications demonstrating that the CD4 binding site on gp120 is not modified by carbohydrates (49, 82). Others have demonstrated that the N7 glycan may help to stabilize the native Env by occluding the V3 loop, shielding V3 epitopes, and inhibiting its premature release prior to CD4 binding (68, 69). These findings, together with our data reported here, support the possibility that the biological property of the N7 glycan may only be discernible in the trimeric form of Env. Furthermore, we have little information on the nature of the glycans on the N7 site and how that may affect its role in different isolates. Further studies will be needed to provide a better understanding of the structural and biophysical basis of our observations.
We report here the conserved phenotype of an N-linked glycan in masking epitopes in the CD4bs and coreceptor binding sites, epitopes targeted by some of the most potent and broadly reactive neutralizing antibodies. Because glycans on HIV-1 play an important role in the structure and function of Env (63, 83, 84) and because multiple studies have shown that carbohydrates modulate Env antigenicity (21, 23, 57, 70, 76, 85,–90), glycan modification may represent a rational approach for immunogen design. Although multiple studies have shown little or negative impact of glycan modifications on Env immunogenicity (91,–95), we previously demonstrated that removal of specific glycans may enhance the ability of HIV Env to elicit neutralizing antibody responses (21). Similar results were also obtained by the removal of specific N-linked glycans proximal to the CD4 binding region (23). Mutant Env with specific glycans removed also binds to the unmutated common ancestors of broadly neutralizing MAbs better than glycosylated versions (88, 96). Furthermore, glycan modification may play a role in modifying the immunogenicity of other vaccine candidates, including those for hemorrhagic fever (97) and feline immunodeficiency virus (98). Taken together, our finding of the conserved nature of the N7 glycan in masking the receptor and coreceptor binding sites of HIV-1 Env may inform the design of immunogens more capable of eliciting bNAbs against these important targets.
We thank Wenjin Guo, Patricia Firpo, Thaddeus Davenport, and Miklos Guttman for expert advice; Taryn Urion and Jennifer McKenna for technical assistance; Cecilia Cheng-Mayer, Paul Clapham, Julie Overbaugh, James Robinson, Ruth Ruprecht, Joseph Sodroski, Leonidas Stamatatos, Charles Wood, Susan Zolla-Pazner, and the NIH AIDS Research and Reference Reagent Program for providing monoclonal antibodies and reagents; and Shan Lu and Kelly Lee for helpful discussions.
This study was supported in part by NIH grants R01 AI076170, P01 AI082274, and P51 OD010425 and Bill and Melinda Gates Foundation award OPP1033102.
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)