Background

Induction of broadly neutralizing antibodies (NAbs) against diverse strains of Human Immunodeficiency Virus Type 1 (HIV-1) remains an important goal for vaccine development [1–3]. Major obstacles are the remarkable sequence variability of the envelope glycoproteins (Env) and the masking of critical neutralizing epitopes by N-linked glycans and other structural and steric constraints [4–6]. Most HIV-1-infected individuals mount a strong autologous NAb response within the first 6 to 12 months of infection that is highly specific for the subject's transmitted/founder virus. The response generally broadens after several years of infection, where in approximately 10-20 percent of cases the antibodies exhibit considerable breadth of neutralization against diverse strains [7–15].

HIV-1 entry is mediated by binding of trimeric gp120 spikes to CD4 receptor that in turn exposes coreceptor binding sites and facilitates fusion of viral and cell membrane [16]. NAbs bind to exposed epitopes on Env trimers and thereby compromise HIV-1 entry [17, 18, 6, 19]. The discovery of broadly neutralizing monoclonal antibodies (MAbs) from HIV-1-infected patients with the ability to neutralize diverse primary HIV-1 isolates [20–23], suggested that there are indeed vulnerable epitopes on the functional Env trimer [24]. Thus, MAb IgG1b12 binds the CD4-binding site (CD4bs) of gp120 [25] and neutralizes more than 50% of HIV-1 clade B and approximately 30% of non-clade B viruses [26, 27]. Although many neutralization epitopes can be masked by N-linked glycans, one MAb, 2G12 [28, 29], binds to specific glycan residue and neutralizes many clade B isolates but has limited breadth against non-clade B isolates [26, 30, 31]. In addition, highly conserved sequences [32] in the coreceptor binding site (also known as CD4-induced or CD4i region) are potential targets for virus neutralization [33–36]. Thus, antibodies mimicking prototype MAb 17b show significant virus neutralization after triggering gp120 with soluble CD4 (sCD4) [24]. Apart from epitopes in gp120 recognized by broadly neutralizing MAbs, the membrane proximal external region (MPER) in gp41 is vulnerable to NAbs and found to be a target of three well characterized MAbs 2F5, 4E10, and Z13 [37–39]. Antibodies targeting the MPER of gp41 neutralize HIV-1 by blocking viral fusion with the cell membrane and thereby preventing viral entry [40]. 59). Interestingly, these types of antibodies are rarely detected during natural infection [22, 41, 42].

Being highly variable, Env remains a major target of the NAb response in HIV-1-infected individuals; thus, Env-driven antibodies have been shown to neutralize autologous virus variants moderately over time [12, 13, 43, 44], followed by rapid escape from neutralization. Autologous NAbs appear to be directed to variable regions of gp120 and are influenced by the pattern of surface Env glycosylation that varies widely among HIV-1 strains [9, 10, 44–52]. These data indicate that despite a limited role for autologous NAbs in the control of viremia, the antibodies exert selection pressure on Env early in infection. In the case of HIV-1 clade B, the V1, V2 and V3 domains have also been shown to mediate CD4 independence, cellular tropism and receptor utilization in addition to neutralization sensitivity [49, 53–65].

HIV-1 clade C is the dominant genetic subtype circulating in India, Sub-Saharan Africa and China [66–70]. Though much information on autologous NAbs against HIV-1 African clade C is available [9, 10, 42, 49, 50, 52, 71, 72], very limited information is available on the neutralization properties of subtype C HIV-1 in India. Current evidence suggests that sequences for the Indian HIV-1 clade C Env and other genes such as gag and nef form a monophyletic lineage and segregate separately as a sub clade within the more diverse subtype C strains from Africa [69, 73–77]. Recently, Kulkarni et al [27] demonstrated that newly transmitted Indian Envs are antigenically complex despite close genetic similarity. In this paper, we examined the NAb response in subtype C HIV-1-infected individuals in India by using Env clones amplified from uncultured peripheral blood mononuclear cells (PBMC) at the baseline, and plasma at the follow up visits of five recently infected subjects and assessed autologous NAbs at different time points for one year. We found that patient Envs varied considerably in their sensitivities to their autologous plasma antibodies and differed in their susceptibilities to MAbs, indicating distinct mechanisms of autologous neutralization and antibody specificities in these patients.

Results

Genetic properties of clade C envclones

Study subjects are described in Table 1. More than one env clones was obtained from each of five recently infected HIV-1 positive individuals from India at a baseline visit and 6 and 12 months later except for subject IVC5, for whom the last visit was at 24 months (Table 2). Env clones from the baseline visit were obtained from infected PBMC DNA whereas for follow up visits, env was amplified from plasma viral RNA. Phylogenetic analyses of the complete gp160 amino acid sequences revealed that the Env clones were indeed subject specific (Figure 1), with intra-clonal genetic divergences between Env clones obtained from the same subject but at different time points indicated ongoing viral evolution. All Envs possessed low net V3 loop charge, a conserved GPGQ motif (Additional file 1: Figure S1) and were found to be CCR5 tropic (Table 2). Except for patients IVC 3 and IVC 4, no significant variation in total N-linked glycosylation sites (PNLG) was found at the three time points sampled (Figure 2); the number of PNLG varied between 25-31 (Table 2). Median gp160 lengths varied between patients; however they did not differ significantly between clones obtained from the same patient at different times (Figure 3). Although there were no major differences between the variable loops of the patient-specific envelope clones obtained at different time points, Env clones 3-3.J9, 3-5.J25 and 5-4.J49, 5-4.J16 amplified from patients IVC 3 and IVC 5 were found to have shorter V1 and V2 loops compared to the contemporaneous Env clones (Additional file 1: Figure S1).

Table 1 Patient details
Full size table
Table 2 Genetic properties of patient Env clones
Full size table
Figure 1
figure 1

Phylogenetic relationships between inter and intra-patient Env gp160 amino acid sequences used to study virus neutralization as determined by Neighbor-Joining maximum likelihood tree using Mega 4.1. Bootstrapped values indicated that Env sequences were patient specific and indicated monophyletic clustering of intra-patient Env.

Full size image
Figure 2
figure 2

Variations of PNLGs in patient Envs at different time points during the course of infection. The bar represents median values.

Full size image
Figure 3
figure 3

Variations in total gp160 lengths of Env clones obtained at different times in each patient during the course of infection. Each bar represents median gp160 residues. Note that significant differences in median gp160 lengths of Envs between IVC 2 and 4, IVC 2 and 11, IVC 4 and 5 and IVC 5 and 11.

Full size image

Neutralization sensitivity of clonal Envs to autologous plasma varied between study subjects

We next assessed the autologous neutralization of Env clones amplified at three different time points from each of five subjects. All five subjects mounted a moderate NAb response against their early viruses obtained at the baseline except patient IVC2; however this phenotype varied with respect to contemporaneous plasma antibodies (Table 3). Surprisingly, only 1/8 clones from subject IVC-2 was neutralized by the plasma samples obtained at later time points, whereas a few (3/8) Env clones were neutralized by the contemporaneous plasma. Thus, while the autologous NAb response to early Env clones improved over time in four subjects, it diminished over time in one subject. This observation was correlated with a gradual decline in CD4, indicating that autologous NAb possibly has selected the fittest Env variants capable of faster disease progression in this particular patient. The majority of the Envs obtained from follow up visits were resistant to contemporaneous autologous plasma antibodies indicating rapid escape of viral variants. The persistence of a few sensitive Envs such as 3-3.J9/F1, and 4-2.J45 during this period of infection despite mounting humoral immune pressure may indicate that these Env variants had adapted to sustain such immune pressure possibly through certain compensatory changes in Env sequence and retained their sensitivities to autologous neutralizing antibodies.

Table 3 Neutralization sensitivity of patient envelopes to autologous plasma antibodies
Full size table

Neutralization phenotype of the Envs as assessed with common neutralizers

To test if the Envs obtained from patients at different time points varied in their sensitivities to common broadly neutralizing MAbs, pseudotyped viruses carrying patient Envs were tested in neutralization assays with sCD4 and five MAbs (b6, IgG1b12, 2G12 targeting gp120 and 2F5, and 4E10 targeting gp41). As shown in Table 4 the majority of Env clones were sensitive to sCD4 at concentrations ranging from 0.1 to 6.66 μg/ml. The pseudoviruses that required excess (>6.66 μg/ml) sCD4 for 50% neutralization were considered as resistant in our study. Consistent with the earlier report [27] all Env variants were resistant to 2G12 except those obtained from IVC-3 patient and this resistance was associated with the absence of PNLG at position 295 (HXB2 numbering) at the N-terminal base of V3 loop. The sensitivity of IVC-3 env clones was due to the presence of N295, atypical of clade C. In contrast to clade B and African clade C viruses [10, 26], envelopes from patient IVC 3, 4, 5, 11 were found resistant to IgG1b12. This observation of b12 resistance of the India clade Envs is in line with that reported by Kulkarni et al [27]. As with the MPER-specific MAbs, all the Envs were resistant to 2F5 at the highest concentration tested (Table 4). Interestingly, while 2F5 resistance was found to be associated with the absence of DKW motif in gp41 in most of the Envs, this motif was found to be present in IVC3-3-9F1, IVC3-5-25F2, and all the Envs obtained from IVC-11 and conferred resistance as shown in Additional file 2: Table S1. Our data indicate that residues outside MPER domain possibly modulated 2F5 sensitivity despite the presence of a minimum DKW motif in MPER for 2F5 sensitivity. The ability of 4E10 to neutralize all the env clones was in agreement with the presence of WFXI motif in gp41; however 4 Envs (4-2_NEM.J46b, 4-5_NEM.J5, 5-3_NEM.J4 and 5-3_NEM.J9) despite having WFXI motif (a minimum 4E10 recognition motif), they were found to be moderately resistant to 4E10 up to a concentration of 6.66 μg/ml (Additional file 1: Figure S1 and Additional file 2: Table S1).

Table 4 Neutralization sensitivity to monoclonal antibodies, sCD4 and anti-CD4
Full size table

Envs from one patient (NARI-IVC2) were moderately sensitive to IgG1b12 but were resistant to contemporaneous plasma antibodies

In contrast all others, Envs amplified from a patient (NARI-IVC2) showed reasonable sensitivity to b12 MAb that targets CD4bs in Env. As shown in Figure 4, these Envs were found to provide a 50% reduction in infection in TZM-bl cells at concentrations ranging from 0.2 to 2.23 μg/ml. The extent of b12 sensitivities of Envs obtained from this particular patient were found to be much higher than the two b12-sensitive Indian clade C Envs reported by Kulkarni et al [27]. The degree of b12 sensitivity of IVC Envs, however, did not correlate with their sensitivity to sCD4 and contemporaneous plasma antibodies. Thus, Envs 2-3.J18, 2-5.J3 and 2-5.J11 which showed the highest neutralization sensitivity (IC50 of 0.5, 0.29 and 0.21 μg/ml respectively) to b12 required more sCD4 for 50% neutralization and except for 2-3.J18 showed neutralization resistance to contemporaneous plasma antibodies (Tables 3 and 4). Our data indicated that escape from contemporaneous NAbs in turn mounted structural constraints in Env specifically on CD4 binding site. This feature therefore possibly contributed in reduced sensitivity of NAb resistant IVC2 envelopes to sCD4, although all envelopes in this patient surprisingly retained b12 sensitivity.

Figure 4
figure 4

Sensitivity of Env clones amplified from IVC2 patients to IgG1b12 antibody. Env-pseudotyped viruses were incubated with IgG1b12 at indicated concentrations for 1 hour before TZM-bl cells were added as described in the Methods. The reduction of infectivity of TZM-bl cells was measured as a function of the degree of IgG1b12 mediated neutralization of these Envs.

Full size image

Sensitivity of Envs to contemporaneous autologous NAbs correlated positively with increased sensitivity to sCD4 and inversely with anti-CD4 antibody

To assess whether the increased sensitivity of patient envelopes to autologous NAbs could be due to greater flexibilities of gp120 interactions with CD4, we next compared the sensitivities of patient Envs to autologous plasmas, sCD4 and an anti-CD4 monoclonal antibody (SIM.2) (hybridoma supernatant) that blocks gp120-CD4 binding. Interestingly, Envs that were resistant to contemporaneous plasmas were less sensitive to sCD4 and required less anti-CD4 antibody (SIM.2) for 50% inhibition. Thus, as shown in Figure 5, a positive association was seen between Env sensitivity to contemporaneous autologous plasma and an increased sensitivity to sCD4 and inverse correlation between Env sensitivity to autologous NAb anti-CD4 antibody, suggesting that Envs with increased sensitivities to sCD4 exhibited greater exposure of epitopes than are targeted by autologous antibodies. The reduced sensitivity of Envs to SIM.2 suggests that Envs with more exposed epitopes for sCD4 require more anti-CD4 antibody for optimum inhibition to entry. Overall, the sensitivities of Envs to sCD4 varied and inversely correlated with their inhibition by SIM.2.

Figure 5
figure 5

Correlations between autologous neutralization sensitivities of patient Envs with their relative susceptibilities to sCD4 and anti-CD4 antibody (SIM.2). Note that Envs that required sCD4 more than 6.66 μg/ml were given a value of 7 μg/ml for the benefit of calculation. A strong correlation was observed between autologous neutralization and Env sensitivity to sCD4 (P < 0.0001) and SIM.2 (P = 0.0004) and between Env susceptibilities to sCD4 and anti-CD4 (P =< 0.0001).

Full size image

Increased sensitivity of patient Envs to contemporaneous NAb and sCD4 correlated with reduced CD4 dependence

We next investigated if Envs with increased sensitivity to autologous antibodies and sCD4 exhibited greater binding to cell surface CD4. Thus, HeLa cells expressing low CD4 but high CCR5 (RC49 cell line) were infected with Env-pseudotyped viruses and the degree of infection was obtained by measuring the intracellular p24. As shown in Figure 6, Envs with increased sensitivity to autologous NAbs (such as 2-3.J18, 3-3.J9, 4.J2, 4-2.J45, 5-4.J22 and 5-4.J49) showed reduced CD4 dependence. However, this phenomenon was found to be independent of the patients and the follow up times examined here (Additional file 3: Figure S2). As expected, we found that increased sensitivity of Envs to autologous NAbs was correlated with reduced CD4 dependence (P < 0.0155) and increased susceptibility to sCD4 (P < 0.0001) (Figure 7). Collectively, our data showed an inverse association of autologous neutralization sensitivity of patient Envs with CD4 dependence.

Figure 6
figure 6

Variation in CD4-dependence of pseudoviruses carrying patient Envs. Pseudoviruses carrying distinct patient Envs were used to infect HeLa cells (RC49 cell line) and the infectivity expressed as percentage infection of these pseudoviruses that infected HeLa cells expressing high CD4 and high CCR5 (JC53 cell line).

Full size image
Figure 7
figure 7

Correlation between CD4 dependence of patient Envs with their sensitivities to autologous plasma antibodies and sCD4. Association of CD4 usage of Env-pseudotyped viruses with autologous plasma antibodies and (P < 0.0155) and sCD4 (P < 0.0001) indicated that Env-pseudotyped viruses with low CD4 dependence tend to be more susceptible to autologous NAb in the patients tested here.

Full size image

Discussion

In contrast to the HIV-1 neutralization properties of African clade C, there is only one report on the neutralization properties of HIV-1 clade C Env clones amplified from co-cultured PBMCs of acutely infected Indian patients [27]. One of the disadvantages in obtaining Env clones from co-culture is that it would potentially select for virus variants that become adapted for favorable replication in the absence of any immune pressure in vitro. This process would therefore fail to select viruses growing in vivo which are responsible for the pathogenesis in the natural course of infection. In the present study, we characterized for the first time the autologous NAb response in subtype C HIV-1 infected Indian patients using multiple molecular Env clones amplified without culture from each study subject. We found that while moderate NAb responses developed in three subjects (IVC 3, 4 and 11), no significant NAb response was detected at all three time points against contemporaneous autologous virus in the remaining two subjects (IVC 5 and IVC 11). In agreement with previous reports, as with both subtype B and African subtype C Envs, we found that in four patients (IVC3, 4, 5 and 11), Envs obtained at baseline and earlier time points were neutralized by plasma antibodies obtained at later time points, indicating repeated cycles of escape [45, 52]. Of potential interest, Env clones obtained at all time points from IVC2 patient were moderately sensitive to IgG1b12, whereas Env clones from the remaining subjects were resistant to this MAb. Surprisingly NAb response in this patient waned over the period of time as plasma from later time points failed to neutralize many contemporaneous as well as earlier envelopes. Intriguingly, no correlation was observed between b12 sensitivity and sCD4 sensitivity as the b12 epitope overlaps CD4 binding site. One plausible explanation for this observation could be that this patient did not develop b12 like antibodies and possibly the absence of selective pressure on the b12 binding site caused the high sensitivity of these envelopes from IVC-2 towards b12. It was also possible that due to lack of co- evolution of b12 and other CD4 binding sites in Env, we did not find any association between b12 and sCD4 sensitivities to Env clones obtained from this particular patient. These observations indicate the presence of compensatory amino acid residues in the IVC-2 Env clones positioned either in the CD4bs or in the proximity that favored enhanced neutralization by b12 MAb. It would be important to further investigate the Env sequence that modulated b12 sensitivity in this patient.

Although we found repeated cycles of escape from autologous NAbs in all the patients, one Env variant (4-2_NEM.J45) obtained from patient NARI-IVC4 at the first follow up retained unusually high sensitivity to contemporaneous and earlier and follow-up plasmas with a mean ID50 of greater than 1: 3000. The persistence of this sensitive Env against which high titer of NAb was developed for at least 6 months makes this envelope interesting; in particular retention of neutralizing epitopes under immense humoral immune pressure probably indicates that this envelope might be more fit in terms of CTL pressure or increased infectivity to compensate for increased sensitivity to NAbs as previously described by Moore et al [45, 52]. When tested against common HIV-1 neutralizing MAbs, most Envs obtained at different time points from all the five participants were resistant to IgGb6, IgG1b12, 2G12 and 2F5 and sensitive to 4E10 only. Intriguingly, two Env variants each from subjects IVC4 and IVC5 despite containing the minimum WFXI motif in gp41 MPER domain for 4E10 recognition, were found to require 4E10 antibody in excess (>6.66 μg/ml) of that required to provide 50% neutralization compared to all other Envs. Nakamura et al [78] recently showed that while F673N and W680G confers 4E10 resistance of HIV-1 envelopes, W680R showed variable 4E10 resistance. In all cases, IC50 values were reported to be in the range of greater than 50-100 μg/ml. In our study, we did not find any of these substitutions in these four Envs, suggesting that the relative resistance of these Envs over others tested here are probably due to changes outside the MPER. Nonetheless, these 4 Envs showed 30-40% sensitivity to 4E10 at a concentration of 6.66 μg/ml, indicating these Envs required excess 4E10 for 50% neutralization but certainly not as much as that would require for W680G or F673N as shown by Nakamura et al [78]. One Env variant each from subjects IVC2 and IVC 3 obtained at first follow up visits that showed unusual sensitivity to 17b, indicating exposed CD4i epitopes. These two Env variants in contrast to the majority of the Env clones were also found to be efficient at infecting HeLa cells expressing low levels of CD4 thereby indicating the presence of exposed CD4i epitopes on Env that enabled them to productively infect HeLa cells expressing low CD4. Nonetheless, two Env variants (5.4.J22 and 5.4.J49) obtained from IVC 5 patient at 2 years showed increased infectivity to HeLa cells expressing low CD4 but were resistant to 17b, indicating that these Envs evolved to conceal their coreceptor binding region on gp120 without compromising low CD4 dependence in the same way that most circulating variants do.

How NAbs drive the Env evolution that impacts on CD4 affinity, tropism and sensitivity to NAbs is not very clear in early HIV-1 clade C infection although two groups using HIV-1 clade B Envs showed association of R5 macrophage tropism with increased CD4 affinity consistent with increased resistance to anti-CD4 monoclonal antibodies [79, 80]. Although in general, the majority of the Envs obtained from all the patients were moderately sensitive to sCD4, we found a few Envs (5.J41, 4-5.J5, 5-4.J16, 11.J25 and 11.J28) that showed autologous antibody resistance but were moderately sensitive to sCD4 indicating that these Envs evolved strategies in escaping autologous neutralization however they retained a very high affinity for the CD4 receptor. The CD4 binding site (CD4bs) on Env experiences most selective pressure as potent NAbs are directed to this domain as documented earlier [15, 49]. Under this selective pressure exerted by humoral immunity, CD4bs is compelled to acquire changes in Env sequences to escape from NAbs that in turn would restrict Env binding efficiently to CD4 receptors [81]. In our study we found that all the Envs that were sensitive to autologous plasma antibodies were moderately susceptible to sCD4 indicating in this scenario, autologous NAbs were mostly directed towards the CD4 binding domain and escape from NAbs possibly had compromised Env binding with CD4. When tested for the extent of CD4 exposure of gp120, Envs that were sensitive to autologous antibodies as well as to sCD4 were found to require less cell surface CD4 for efficient entry, indicating an inverse correlation between Env sensitivity to autologous NAbs and CD4 dependence. The relationship between sensitivity of Envs to sCD4 and anti-CD4 antibodies with variable dependence to cell surface CD4 were described previously by different investigators. Gorry et al [82] showed that a neurotropic Env obtained from brain tissue with higher affinity to CD4 was found to be increasingly sensitive to CD4 mimetic, CD4-IgG2. Later, Dunfee et al [83] showed that Envs with N283 substitution could productively infect cells expressing low cell surface CD4 and show greater affinity to sCD4. Similar observations were found by Vermeire et al [81], where they showed that a NL4-3 variant that evolved to infect cells expressing low CD4 in presence of the small molecule CADA was found to be highly susceptible to heterologous sera and was concordant with increased sensitivity and resistance to sCD4 and anti-CD4 respectively. In addition, Peters et al [79, 84] demonstrated that patient-derived Envs that were able to exploit low CD4 on cell surface were proportionately resistant and sensitive to anti-CD4 antibody and sCD4 respectively.

In conclusion, in the present study, we have shown for the first time the neutralization properties of HIV-1 India clade C Env clones obtained from patients followed up with recent infection over time to their autologous antibodies during the natural course of infection and investigated their genetic relatedness with sensitivity to known broadly neutralizing monoclonal antibodies and degree of exposure to CD4 for efficient entry. While variations in autologous neutralization of viruses are expected, all available data on the mechanisms of resistance and sensitivity to neutralizing antibodies of geographically diversified HIV-1 clade C that contributes to major global HIV-1 pandemic will help designing strategies fostering vaccine discovery.

Methods

Patient details, PBMC and plasma samples

All five recently infected study subjects acquired HIV-1 through heterosexual contacts and were ART naïve at the time of blood collection. The mean CD4 counts ranged from 328-606 cells per cubic millimeter (mm3). Based on detuned ELISA results [85–87] and history of exposure within the last 6 to 8 months, these patients were selected as recently infected patients for further characterization. Plasmas used for HIV-1 envelope amplification and tested for antibody assays were obtained at baseline, 6 and 12 months respectively.

Amplification and cloning of gp160

gp160 amplification from peripheral blood mononuclear cell (PBMC) DNA and from reverse-transcribed plasma viral RNA was carried out by nested PCR using 5'-TAGAGCCCTGGAAGCATCCAGGAAG-3' as forward and 5'-TTGCTACTTGTGATTGCTCCATGT-3' as reverse primer in the first round and 5'-CACCGGCTTAGGCATCTCCTATGGCAGGAAGAA-3' as forward and 5'-TATCGGTACCAGTCTTGAGACGCTGCTCCTACTC-3' as reverse primer in the second round by using Platinum Taq proof reading polymerase (Invitrogen Inc.). Plasma viral RNA was purified by using a nucleic acid isolation kit as described by the manufacturer (Roche Inc.). cDNA from diluted viral RNA was prepared using Superscript III first strand synthesis kit (Invitrogen Inc.). gp160 was amplified by two rounds of nested PCR gp160 amplicons were purified and ligated into either pcDNA 3.1/V5-His-TOPO (Invitrogen Inc) or pSVIIIenv [84].

DNA sequencing and phylogenetic analysis

Sequence analysis was performed using cycle sequencing and big dye terminator methods by automated genetic analyzer (Applied Biosystems, Inc; Model 3730XL) as described earlier [88]. Nucleotide and deduced amino acid sequences were aligned using MEGA software and phylogenetic tree was constructed by the neighbor-joining method [88].

Pseudovirion preparation and measurement of virus titer

Pseudotyped viruses carrying patient Envelope were produced by cotransfection of env+ pSVIIIenv or env+ pcDNA 3.1/V5-His-TOPO with env-defective HIV-1 backbone vector (pSG3ΔEnv) [44, 89], into 293T cells during log growth phase in 6-well tissue culture trays (Corning Inc) using calcium phosphate (Promega Inc) following manufacturer's protocol. Cell supernatants carrying progeny pseudotyped viruses were harvested at 48 hours post-transfection, and stored at -152°C until further usage. The infectivity assays were done in TZM-bl cells in 96-well microtiter plate and infectivity titers determined by measuring the luciferase activity respectively as described elsewhere [90].

Neutralization Assay

Patient plasma samples were evaluated for NAb activity against Env pseudotyped viruses using a single round reporter assay in TZM-bl cells as described previously with few modification [90]. Briefly, 200 TCID50 of pseudovirus was incubated with serial 3 fold dilutions of plasma sample in duplicates in a total volume of 150 μl for 1 hr at 37°C in 96-well flat-bottom culture plates. Freshly trypsinized cells (10,000 cells in 100 μl of growth medium containing 25 ug/ml DEAE Dextran) were added to each well. One set of control wells received cells plus pseudovirus (virus control) and another set received cells only (background control). After 48 hours of incubation, luciferase activity was measured by using the Bright-Glo Luciferase Assay System (Promega Inc.). The 50% inhibitory dose (ID50) was defined as either the plasma dilution or sample concentration (in the case of sCD4 and MAbs) that caused a 50% reduction in relative luminescence units (RLU) compared to virus control wells after subtraction of background RLU.

p24 antigen immunostaining

Immunostaining of HeLa cells infected with pseudoviruses was carried out as described earlier [84]. p24 positive cells were regarded as foci of infection, and virus infectivity was estimated as focus-forming units (FFU) per milliliter.

Nucleotide sequence accession numbers

All env sequences have been submitted to GenBank (accession numbers: [GenBank:EU908214] to [GenBank:EU908221], [GenBank:EU908224] to [GenBank:EU908225], and [GenBank:GU945306] to [GenBank:GU945333]).

Statistical analyses

Correlations between NAb response and magnitude of envelope binding to sCD4, RC49 cells and anti-CD4 antibody (SIM.2) were assessed by calculating Spearman's non-parametric 2-tailed correlation co-efficient with 95% confidence level using GraphPad Prism software. The percent infectivity of Env clones in HeLa cells expressing low CD4 (RC49) were plotted and compared by Mann-Whitney and two-way ANNOVA tests using GraphPad Prism software. Correlations were considered significant with P values less than 0.05. To avoid digression of correlation, one Env clone (4.2J45) was not included during assessing the correlation between Env sensitivity to NAb and sCD4 (Figure 5) and between NAb and HeLa cell (RC49) (Figure 6B) infection as the sensitivity of this Env clone to it autologous plasma was exceptionally higher (ID50 greater than 6000; see Table 3)