|Home | About | Journals | Submit | Contact Us | Français|
To date, only a small number of anti-human immunodeficiency virus type 1 (HIV-1) monoclonal antibodies (MAbs) with relatively broad neutralizing activity have been isolated from infected individuals. Adequate techniques for defining how frequently antibodies of these specificities arise in HIV-infected people have been lacking, although it is generally assumed that such antibodies are rare. In order to create an epitope-specific neutralization assay, we introduced well-characterized HIV-1 epitopes into the heterologous context of simian immunodeficiency virus (SIV). Specifically, epitope recognition sequences for the 2F5, 4E10, and 447-52D anti-HIV-1 neutralizing monoclonal antibodies were introduced into the corresponding regions of SIVmac239 by site-directed mutagenesis. Variants with 2F5 or 4E10 recognition sequences in gp41 retained replication competence and were used for neutralization assays. The parental SIVmac239 and the neutralization-sensitive SIVmac316 were not neutralized by the 2F5 and 4E10 MAbs, nor were they neutralized significantly by any of the 96 HIV-1-positive human plasma samples that were tested. The SIV239-2F5 and SIV239-4E10 variants were specifically neutralized by the 2F5 and 4E10 MAbs, respectively, at concentrations within the range of what has been reported previously for HIV-1 primary isolates (J. M. Binley et al., J. Virol. 78:13232-13252, 2004). The SIV239-2F5 and SIV239-4E10 epitope-engrafted variants were used as biological screens for the presence of neutralizing activity of these specificities. None of the 92 HIV-1-positive human plasma samples that were tested exhibited significant neutralization of SIV239-2F5. One plasma sample exhibited >90% neutralization of SIV239-4E10, but this activity was not competed by a 4E10 target peptide and was not present in concentrated immunoglobulin G (IgG) or IgA fractions. We thus confirm by direct analysis that neutralizing activities of the 2F5 and 4E10 specificities are either rare among HIV-1-positive individuals or, if present, represent only a very small fraction of the total neutralizing activity in any given plasma sample. We further conclude that the structures of gp41 from SIVmac239 and HIV-1 are sufficiently similar such that epitopes engrafted into SIVmac239 can be readily recognized by the cognate anti-HIV-1 monoclonal antibodies.
The presumed rarity of broadly reactive, human immunodeficiency virus type 1 (HIV-1)-specific neutralizing antibodies in the plasma of HIV-positive individuals arises from the observation that very few monoclonal antibodies (MAbs) with such activity have been isolated since HIV-1 was first characterized 20 years ago. Among the small number of well-characterized, broadly neutralizing anti-HIV-1 MAbs, three recognize distinct elements of the gp120 subunit of envelope, including the CD4 binding site (b12), specific glycans on the surface of gp120 (2G12), and the V3 loop (447-52D) (32, 47). In contrast, three other MAbs (2F5, 4E10, and Z13) recognize determinants clustered within a single 30-amino-acid stretch adjacent to the membrane-spanning segment of the viral TM protein (41, 66). While all of these antibodies recognize conformational components of the HIV envelope complex, minimal recognition determinants of four of them (2F5, 4E10, Z13, and 447-52D) have been mapped to short, linear elements within the HIV-1 sequence (13, 18, 41, 66).
The elements required for binding of the 2F5, 4E10, and Z13 antibodies are located in the membrane-proximal external region (MPER) of the HIV-1 gp41 ectodomain (41, 42, 66). An alignment of simian immunodeficiency virus (SIV) sequence with those of several commonly studied HIV-1 strains illustrates the highly conserved nature of this region (Fig. (Fig.1).1). In particular, this region contains the tryptophan-rich motif which is common to both primate and nonprimate lentiviruses (55).
Soluble forms of the gp41 ectodomain or synthetic peptides based on sequences found upstream of the MPER in gp41 form stable, six-helix coiled-coil structures in solution (7, 8, 35, 61-63). This structure is thought to mimic the fusion-active form of gp41 and is highly conserved among fusion proteins of enveloped viruses (60). Synthetic N-terminal and C-terminal gp41 molecules derived from the SIV gp41 sequence also form a coiled-coil structure in solution (35). An SIV C-peptide can interact with HIV-1 N-peptides to form a similar structure, and the SIV C-peptide can inhibit HIV-1 infectivity similar to the autologous peptide (35). Together, these observations indicate that the functional and structural properties of SIVgp41 are very similar to those of HIV-1 gp41. We therefore hypothesized that SIV gp41 can be minimally altered by site-directed mutagenesis to contain the essential recognition elements for MAbs 2F5 and 4E10 yet maintain envelope structure and function.
MAb 2F5 recognizes a determinant within the MPER of HIV-1 gp41 (41). The binding of 2F5 has been characterized by peptide mapping, mutagenesis, circular dichroism, nuclear magnetic resonance, and matrix-assisted laser desorption ionization-mass spectrometry (MALDI-MS) (27, 41, 42, 44, 51, 58, 64, 65), and structures of the MPER and 2F5 have been described previously (1, 3, 43, 56). Based on peptide mapping, the 2F5 epitope was originally identified as the six-residue element ELDKWA (amino acids 662 to 667). This sequence lies just upstream of the membrane-spanning domain of the TM (gp41) protein. Fine mapping with synthetic peptides further identified the central LDKW motif as containing key residues for antibody binding, whereas replacements in the 1st and 6th positions with almost any other amino acid had little or no effect on 2F5 binding (51). Escherichia coli MalE proteins containing inserted ELDKWA motifs are recognized by 2F5 (11), as are insertions of the epitope into the variable regions of HIV-1 gp120 (34) or surface proteins of other viruses (15, 16). However, such chimeric molecules have generally failed as experimental immunogens to induce 2F5-like neutralizing activity.
The 2F5 sequence encodes an unusually long CDR H3 domain. Mutagenesis and structural analyses indicate that the finger-like projection formed by the long CDR H3 domain contributes to the neutralizing activity of the MAb, despite the fact that many of the CDR H3 residues are not involved in making contacts with the core epitope (43, 65). It has been suggested that hydrophobic residues in the finger-like domain interact with the viral lipid membrane and contribute to the overall affinity of the antibody for its target; Kwong and colleagues have further speculated that this domain could act by “scanning” membrane surfaces until the MAb encounters its target in the HIV envelope spike (43).
MAb 4E10 was originally isolated from an HIV-positive individual (4). As with 2F5, a deceptively simple, minimal target element has been defined for 4E10 using overlapping peptides (66). The minimal target sequence for 4E10 (NWFN/DIT) also lies in the MPER domain, C-terminal to the 2F5 epitope and immediately preceding the membrane-spanning domain. While 4E10 will neutralize many strains of HIV-1 containing this exact motif, some strains bearing the sequence are not sensitive to the antibody (66). Surveys of primary HIV-1 isolates have consistently revealed that 4E10 has the broadest activity of any of the well-known HIV-1-specific neutralizing monoclonal antibodies (NAbs) (2, 38, 54, 66).
Monoclonal antibody 447-52D is a human immunoglobulin G3 (IgG3) antibody that was isolated from an HIV-1 clade B-infected patient (18, 19). Recognition of HIV-1 gp120 by MAb 447-52D was mapped to the third hypervariable domain (V3), and peptide mapping defined an epitope including the conserved GPGR motif present at the apex of the V3 loop. Compared to other V3-specific MAbs, neutralization by 447-52D is relatively broad (18, 30, 57). An explanation for why 447-52D can neutralize multiple strains of HIV-1 when most other anti-V3 MAbs cannot was provided by analysis of the crystal structure of a 447-52D Fab in complex with a V3 peptide (57). Both structural and mutagenesis data indicate that 447-52D interacts specifically with three conserved residues at the tip of the V3 loop, giving the MAb its broad specificity. Additional interactions between 447-52D and the length of the V3 peptide occur via main chain atoms, such that overall affinity may not be significantly affected by amino acid side chain variability.
HIV-positive individuals generally produce high-titer, polyclonal antibody responses to viral proteins, including the gp120 and gp41 envelope subunits. Although specificities present in plasma samples can be surveyed using binding assays (such as enzyme-linked immunosorbent assay [ELISA] or Western blot), binding is not predictive of neutralizing ability (5, 6, 17, 22, 39, 40, 46-48, 49). Alternatively, neutralization assays can measure total neutralizing activity present in a given serum or plasma sample, but cannot distinguish the contributions of individual antibody specificities that make up a polyclonal mixture. It is therefore difficult to survey samples to detect neutralizing activity directed at specific epitopes or regions of the envelope oligomer. In order to screen directly for neutralizing activity of defined specificity in plasma samples, we have introduced well-characterized HIV-1 epitopes for monoclonal antibodies 2F5, 4E10, and 447-52D into the heterologous context of SIVmac239. Using the 2F5 and 4E10 epitope-engrafted SIVmac239 derivatives (SIV239-2F5 and SIV239-4E10), we have directly addressed the extent to which neutralizing antibodies of these specificities are present in HIV-positive patients.
Mutations in env were created by site-directed mutagenesis using the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). The following mutagenic primers were used: to create SIV239-2F5 (positions 8614 to 8657), 5′-GAGTTGGATAAGTGGGCTTCGTTTGGCAATTGGTTTGACCTTGC-3′ and (positions 8634 to 8590) 5′-CGAAGCCCACTTATCCAACTCTTGTAATTCATACATGTTCTTCTC-3′; to create SIV239-4E10 and to add the 4E10 epitope to SIV239-2F5 to create SIV239-2F5/4E10 (positions 8638 to 8677), 5′-GGCAATTGGTTTAACATTACTTCTTGGATAAAGTATATAC-3′ and (positions 8667 to 8627) 5′-TATCCAAGAAGTAATGTTAAACCAATTGCCAAACACATC-3′; and to create SIV239-447-52D and to add the 447-52D epitope to SIV239-2F5/4E10 to create SIV239-2F5/4E10/447-52D (positions 7560 to 7604), 5′-TTTACCAGTCCACATTGGGCCTGGACGGGCTTTCCACTCACAACC-3′ and (positions 7599 to 7569) 5′-TGAGTGGAAAGCCCGTCCAGGCCCAATGTGGACTGGTAAAACTGT-3′. The primers were purchased from Sigma-Genosys Biotechnologies, Inc (The Woodlands, TX). In order to generate full-length versions of all of the 3′ mutants, a p239SpX5′ was created by removal of an SphI site and incorporation of an XhoI site in p239SpSp using the following mutagenesis primers: 5′-GGGATCCTCTAGACTCGAGCTGCAGGCTAGCACATTTTAAAGGC-3′ and 5′-GCCTTTAAAATGTGCTAGCCTGCAGCTCGAGTCTAGAGGATCCC-3′. In order to generate the full-length mutants, p239SpX5′ was digested with XhoI and SphI and the corresponding fragment was cloned into each mutant using T4 DNA ligase.
Virus was prepared by transient transfection of 293T cells with full-length SIV genomes. Cells were seeded at 1.5 × 106 cells per flask the day before transfection, and each flask was transfected with 3 μg of each plasmid using Transfectin lipid reagent (Bio-Rad, Hercules, CA) following the manufacturer's instructions. Culture supernatants containing virus were harvested 3 days after transfection. To prepare virions for envelope incorporation assays, full-length genomes were transfected using calcium phosphate precipitation (Promega, Madison, WI) according to the manufacturer's recommended protocol.
Infectivity assays and growth curves were performed on an immortalized human CD4+-T-cell line (C8166-45 SIV-SEAP) and rhesus peripheral blood mononuclear cells (PBMCs). C8166-45 SIV-SEAP cells harbor a Tat-inducible secreted alkaline phosphatase (SEAP) reporter construct enabling SIV infection to be measured by SEAP production in the culture supernatant. Aliquots of all virus stocks used in these experiments were subject to serial twofold dilutions, and SEAP activity was measured from the supernatant at day 3 postinfection according to the manufacturer's recommendations, with modifications as described previously (36). Rhesus monkey PBMCs were purified by Ficoll separation of blood taken from specific-pathogen-free animals and were activated with phytohemagglutinin (1 μg/ml) for 2 days before infection. Phytohemagglutinin-activated PBMCs were maintained in R20 medium (RPMI 1640 medium with 20% fetal bovine serum [FBS], l-glutamine, penicillin, and HEPES buffer) supplemented with 20 U of interleukin-2/ml (provided by Maurice Gately, Hoffman-La Roche, Nutley, N.J.)
Virus-containing supernatants were first clarified by two consecutive centrifugation steps for 10 min at 2,600 × g. Virions were then pelleted by centrifugation for 2 h at a high speed (16,000 × g) in a refrigerated microcentrifuge. The viral pellet was washed by resuspension in 1 ml of phosphate-buffered saline (PBS) and pelleted again by centrifugation at a high speed. After this second ultracentrifugation, the viral pellets were resuspended in 50 μl of PBS and the amount of p27 was quantified by antigen capture (Beckman Coulter). Identical quantities of p27 were then mixed with Laemmli buffer and boiled for 4 min. Samples were then electrophoresed through an 8 to 16% polyacrylamide-sodium dodecyl sulfate (SDS) gradient gel. After electrophoresis, the proteins were transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, Mass.). The membranes were blocked with 5% skim milk in PBS-0.05% Tween 20 for 1 h and then incubated with antibodies recognizing the gp120 subunit (rhesus macaque MAb [RhMAb] 3.10A) (12, 53) and p27 (MAb 2F12) (23). Horseradish peroxidase-conjugated anti-rhesus immunoglobulin G was used to detect antibody 3.10A, and horseradish peroxidase-conjugated anti-mouse immunoglobulin G was used to detect monoclonal antibody 2F12. The rhesus monoclonal antibody 3.10A was a gift of J. E. Robinson (Tulane University Medical School). The 2F12 monoclonal antibody was obtained through the NIH AIDS Research and Reference Reagent Program. The membranes were treated with a chemiluminescent substrate (Pierce, Rockford, Ill.). The bands were visualized and analyzed with a Fuji phosphorimager.
Ninety-two plasma samples from 47 HIV-positive clade B-infected donors were obtained from study subjects recruited with a diagnosis of primary (recent) HIV infection as part of the San Diego Acute and Early Infectious Disease Research Program. Serial blood specimens were collected, separated by centrifugation into plasma and cells, and frozen at −70°C. All subjects signed informed consent to protocols approved by the University of California Human Subjects Committee (La Jolla).
The neutralization sensitivity of each virus was tested using a secreted, engineered alkaline phosphatase (SEAP) reporter cell assay as previously described (36). Virus equivalent to 1 ng of p27 capsid protein for SIV239 and -SIV239-4E10, 5 ng of p27 for SIV316 and -SIV239-2F5, and 2 ng of p24 HIV-1 capsid for NL4-3 were chosen as the lowest levels of viral input sufficient to give a clear SEAP signal within the linear range for each viral strain. SEAP activity was measured when levels were sufficiently over background to give reliable measurements (at least 10-fold). To perform neutralization assays, 96-well plates were set up as follows: to the first three columns, 25 μl of medium (RPMI 1640, 10% FBS) was added; to each of the other columns (no. 4 through 12), 25-μl aliquots of successive twofold dilutions of test antibody or plasma in RPMI 1640-10% FBS were added. All plasmas were heat inactivated at 56°C for 30 min before use in neutralization assays. Each virus in a total volume of 75 μl was then added to each well in columns 3 through 12. Virus-free medium was added to columns 1 and 2 (mock). The plate was incubated for 1 h at 37°C. After incubation, 5,000 target cells (C8166-45 SIV-SEAP) in a volume of 100 μl were added to each well. The plate was then placed into a humidified chamber within a CO2 incubator at 37°C. SEAP activity was measured using the chemiluminescent Phosphalight secreted alkaline phosphatase assay system (Applied Biosystems) according to the manufacturer's recommendations, with modifications as described previously (36). Neutralization activity for all antibodies and plasma samples was measured in triplicate using a Victor V multilabel counter (Perkin-Elmer) and reported as the percentage of SEAP activity ± standard deviation.
The production and characterization of RhMAbs recognizing SIV envelope glycoproteins have been described previously (12). In the present study, the following RhMAbs were used: 1.9C, 1.10A, and 3.11H. The murine KK41 MAb used in these experiments was obtained from the AIDS Reagent Repository; production and characterization of KK41 reagent were described previously (28, 29).
Neutralization sensitivity of each virus to a panel of 92 plasma samples from 47 patients was analyzed using the C8166-45 SIV-SEAP reporter cell line as described above. For each plasma sample, the percentage of SEAP activity was determined in triplicate at two dilutions: 1/20 and 1/200.
In order to evaluate the participation of 4E10-specific antibodies in neutralization, the neutralization assays were adapted. The 94-1 peptide (60 μg/ml) and MAb 4E10 (or plasma) were preincubated for 30 min at 37°C. The peptide-antibody solution was added (1:1 by volume) to 239-4E10 virus (1 ng of p27 per well), and the mixture was incubated for 1 h at 37°C. The virus-peptide-antibody mixture was added (1:1 by volume) to 5,000 C8166-45 SIV-SEAP cells. The plate was then placed into a humidified chamber within a CO2 incubator at 37°C for 4 days. SEAP activity was measured from the supernatant according to the manufacturer's recommendations, with modifications as described previously (36). A competition was observed when the addition of peptide resulted in a decrease in the neutralizing capacity of the antibody (or plasma).
IgG was purified from human serum LTNP2, using protein A affinity chromatography (Amersham), according to the manufacturer's instructions. Following elution of IgG from the column (0.1 M citric acid, pH 3.0), IgG-containing fractions were neutralized and pooled. To purify serum IgA, the serum was first depleted of IgG using protein A as described above, and the depleted serum was mixed 1:1 (vol) with PBS loaded onto a Jacalin affinity chromatography column (Pierce), according to the manufacturer's instructions. The Jacalin column was washed with 10 to 15 column volumes of PBS and IgA eluted with 0.1 M melibiose in PBS. Before use in neutralization assays, purified IgG and IgA were dialyzed extensively against PBS, pH 7.4, and verified for concentration and purity using relative absorbance at 280 nm and SDS-polyacrylamide gel electrophoresis (PAGE), respectively.
Recognition elements for three NAbs were introduced into the SIVmac239 env sequence by site-directed mutagenesis. NAbs 2F5, 4E10, and 447-52D were chosen because minimal binding determinants for all three have been mapped to simple, linear sequence elements in the HIV-1 env gene (13, 18, 41, 66). The analogous sites in the SIVmac239 gp120 and gp41 proteins were readily identified by sequence alignment, and the corresponding recognition sequences were created by site-directed mutagenesis (Fig. (Fig.1).1). Full-length proviral constructs were generated as described in Materials and Methods, and virion stocks were produced by transient transfection of 293T cells. For all subsequent experiments, virion stocks were normalized by amount of p27 as determined by antigen capture (Beckman Coulter).
C8166-45 SIV-SEAP indicator cells were used to quantify the infectivity of each of the viruses under conditions that approximated a single cycle of infection. Cells were infected with normalized amounts of the parental SIV239, the neutralization-sensitive derivative SIV316, and five SIVmac239 variants bearing HIV epitopes (SIV239-2F5, SIV239-4E10, SIV239-447-52D, SIV239-2F5/4E10, and SIV239-2F5/4E10/447-52D) (Fig. (Fig.1).1). The results of a representative experiment are shown in Fig. Fig.2.2. All of the epitope-engrafted variants had decreased infectivity compared with the SIV239 parent; only the variant bearing all three HIV epitopes (SIV239-2F5/4E10/447-52D) was not detectably infectious. SIV239-4E10 had only a moderate (10-fold) decrease in infectivity. For SIV239-2F5, SIV239-447-52D, and SIV239-2F5/4E10, the effect was more severe (300-fold, 600-fold, and 500-fold, respectively). The infectivity and replicative capacity of SIV239-447-52D and SIV239-2F5/4E10 were too low to use for neutralization assays.
We compared replication of epitope-engrafted variants SIV239-2F5 and 239-4E10 with the parental SIV239 and the neutralization-sensitive variant SIV316 in activated, IL-2-stimulated PBMCs taken from two uninfected rhesus monkeys (Fig. (Fig.3).3). Normalized amounts of mutant and parental virus stocks produced from transient transfection of 293T cells were used to infect PBMCs from both rhesus monkeys (see Materials and Methods). Cell supernatants were harvested on the indicated days, and viral replication was monitored by measuring the production of p27 in the cell culture supernatant. Both recombinant viruses were replication competent in PBMCs from both monkeys. SIV239-4E10 replicated with kinetics very similar to those of the parental SIV239, whereas replication of SIV239-2F5 was notably delayed.
Mutations within the HIV-1 MPER domain have been described that affect incorporation of envelope into newly formed virions (55). Creation of four of the epitope-engrafted viruses described in this report involved site-directed mutagenesis of the analogous domain in the SIV239 gp41 protein (SIV239-2F5, SIV239-4E10, SIV239-2F5/4E10, and SIV239-2F5/4E10/447-52D). All four variants also displayed reduced infectivity compared to the parental SIVmac239 (Fig. (Fig.2).2). To ascertain whether this reduced infectivity might be related to altered levels of Env incorporation into virions, stocks of each variant were lysed, separated by SDS-PAGE, and analyzed by immunoblotting (Fig. (Fig.4).4). The ratio of virion-associated envelope protein (gp120) to capsid protein (p27) was determined by densitometric analysis, and the ratio for each variant was then compared to the parental SIVmac239 (Fig. (Fig.4B).4B). The levels of incorporation relative to SIVmac239 were 29% (SIV316), 12% (SIV239-2F5), 17% (SIV239-4E10), 52% (SIV239-447-D), 7% (SIV239-2F5/4E10), and 9% (SIV239-2F5/4E10/447-D). Thus, the most profound effects on incorporation occurred in the variants with changes in the MPER domain. However, decreased incorporation did not correlate absolutely with decreased infectivity. For example, relative incorporation levels for SIV239-2F5 and SIV239-4E10 were similar (12% and 17%), but infectivity of the two viruses differed dramatically (300-fold and 10-fold, respectively). Env incorporation of the SIV239-447-D variant was the least affected (52% of wild-type incorporation), and yet the infectivity of this variant was down 600-fold. Thus, incorporation levels are clearly not sufficient to explain the full range of decreased infectivity displayed by the epitope-engrafted SIVmac239 variants.
We next measured the neutralization sensitivity of 239-2F5 and 239-4E10 and the corresponding parental clone (SIV239) by the 2F5 and 4E10 monoclonal antibodies. Neither antibody was able to neutralize SIV239, whereas each MAb specifically neutralized the virus engrafted with the corresponding epitope (Fig. (Fig.5).5). SIV239-2F5 displayed considerable neutralization sensitivity to MAb-2F5, achieving 50% neutralization at a concentration of 0.05 μg/ml (Fig. (Fig.5B).5B). This degree of sensitivity to neutralization is similar to the highest sensitivity previously reported for MAb 2F5 when tested against 93 different strains of HIV-1 (2). Fifty percent neutralization of SIV239-4E10 by MAb 4E10 was achieved at 5 μg/ml (Fig. (Fig.5A),5A), a level of potency similar to the median observed against a panel of HIV-1 primary isolates (2).
In order to further test the specificity of the recognition of SIV239-2F5 and SIV239-4E10, we measured the neutralization sensitivity of these viruses and the parental clone (SIV239) to Fab Z13e1, an affinity-enhanced variant of Fab Z13 (M.B.Z., manuscript in preparation). Fab Z13 was cloned by phage display using a library generated from an HIV-1-seropositive donor and recognizes a predominantly linear and relatively conserved epitope immediately C terminal of the 2F5 epitope (66). Although Fab Z13 competes with MAb 4E10 for binding to a synthetic peptide derived from gp41, the epitopes recognized by 4E10 and Z13 are not identical (66). Fab Z13e1 was unable to neutralize 239-2F5, 239-4E10, or the parental SIV239 at any of the concentrations tested (data not shown). This observation supports the strong specificity of the recognition of the epitope-engrafted mutants for the corresponding antibodies.
Despite the introduction of different HIV-specific epitopes, the engrafted SIV239 variants displayed very similar susceptibilities to neutralization by pooled plasma from 12 SIV-infected rhesus macaques compared to the parental SIV239 (Fig. (Fig.6).6). Fifty percent neutralization of SIV239 was achieved at a 1/50 dilution. Similarly, 50% neutralization of SIV239-2F5 and SIV239-4E10 was achieved at dilutions of 1/50 and 1/100, respectively.
We next measured neutralization of the three viruses by selected monoclonal antibodies. We chose three anti-gp120 rhesus monoclonal antibodies from different competition groups (1.9C, 1.10A, and 3.11H) (12) and an anti-gp41 murine monoclonal antibody (KK41) (28, 29). We have previously reported that all of these MAbs have significant neutralizing activity against the neutralization-sensitive SIV239 derivatives SIV316 and SIVΔV1V2, and against lab-adapted SIVmac251 (26). As shown previously, SIV239 was not effectively neutralized by any of the monoclonal antibodies tested (Fig. (Fig.6).6). Likewise, SIV239-2F5 and SIV239-4E10 were resistant to neutralization by MAbs 1.9C, 3.11H, and KK41. Interestingly, RhMAb 1.10A had some neutralizing activity against SIV239-2F5 (Fig. (Fig.66).
We next sought to use the epitope-engrafted SIV239 variants SIV239-2F5 and SIV239-4E10 to survey a panel of HIV-positive plasmas for 2F5-like and 4E10-like neutralizing activity. Our panel included samples taken from 47 HIV-1-seropositive individuals; for 45 individuals, samples from two time points were analyzed. Time points for individuals were anywhere from 1 to 8 years apart; plasma for the first time point in most cases was obtained within the first 12 to 18 months of infection. To control for nonspecific neutralizing activity or for activity cross-reacting with SIV, plasma samples were also screened against the parental SIV239 and its neutralization-sensitive derivative, SIV316. All samples were also screened against HIV-1 NL4-3 as a control for the presence of HIV-1-specific neutralizing antibodies. Each virus/plasma combination was determined in triplicate at two dilutions (1/20 and 1/200). Table Table11 summarizes the neutralization results. None of the plasmas achieved 90% neutralization of any strain at a 1/200 dilution, although four of the plasma samples (from LTNP2, VI2992, i447, and d381) exhibited 50% or greater neutralization of SIV239-4E10 at a 1:20 dilution in this initial screen (Table (Table1).1). The most notable neutralizing activity detected in this survey was the 95% neutralization of SIV239-4E10 by a 1:20 dilution of plasma from LTNP2 (sample 1/20/04). A 1:20 dilution of this plasma sample consistently yielded 85 to 95% neutralization in three independent tests (Table (Table1,1, Fig. Fig.7,7, and Fig. Fig.8).8). The same sample also appeared to have some weaker neutralizing activity against the other three viruses in the screen (Table (Table1).1). LTNP2 has never been prescribed or taken antiviral drugs, ruling out an effect of antiviral drugs present in the plasma sample.
The 92 plasma samples were obtained from HIV-1-seropositive individuals. We examined the ability of these samples to neutralize HIV-1 strain NL4-3. Plasma samples from 39 of the 47 individuals were able to neutralize NL4-3 at one or both time points (Table (Table1).1). Among the cases in which only one of two time points from a single individual neutralized HIV-1 NL4-3, there was a strong tendency for the later time point to be the positive one (30 out of 32 individuals). The range of neutralizing antibody titers against HIV-1 NL4-3 obtained in this assay was similar to, or slightly less than, values previously reported using a pseudotype assay (52).
To examine a possible correlation between neutralization by LTNP2 plasma and the presence of antibodies specific for the 4E10 epitope, we performed neutralization experiments in the presence of the competitor peptide 94-1. Peptide 94-1 was previously identified as a peptide that bound the 4E10 antibody and could inhibit its neutralizing activity (F. M. Brunel, M. B. Zwick, R. M. F. Cardoso, J. D. Nelson, I. A. Wilson, D. R. Burton, and P. E. Dawson, submitted for publication). Serial dilutions of LTNP2 plasma were preincubated with 94-1 peptide (60 μg/ml). The peptide-plasma solution was incubated with 239-4E10 virus, and the mixture was added to the reporter cells. Peptide 94-1 did not block the neutralization of 239-4E10 virus by LTNP2 plasma (Fig. (Fig.7)7) but did compete neutralization of SIV239-4E10 by MAb 4E10. The correlations between 4E10-specific antibodies and neutralization in the other three plasmas with some apparent neutralizing activity against SIV239-4E10 were similarly investigated. Preincubation with 94-1 peptide (60 μg/ml) similarly failed to block the observed neutralization of SIV239-4E10 virus by the three plasmas tested (Table (Table22).
Heat-inactivated plasma from patient LTNP2 (1/20/04) reduced infectivity of SIV239-4E10 by approximately 50% at a 1:40 dilution (Fig. (Fig.8A).8A). In order to investigate whether the observed neutralization activity was antibody mediated, IgG and IgA fractions were purified from the plasma of patient LTNP2 (sample 1/20/04). SIV239-4E10 was preincubated with dilutions of purified, concentrated IgG and IgA and used to inoculate C8166-45 SIV-SEAP indicator cells. At day 4 postinfection, SEAP activity was measured. Results are depicted in Fig. Fig.8B.8B. Purified IgG isolated from patient LTNP2 failed to neutralize SIV239-4E10, and purified IgA showed very modest neutralization activity against SIV239-4E10 (50% inhibitory concentration [IC50] = 250 μg/ml). Therefore, although IgA may contribute slightly to the observed neutralization by LTNP2 plasma, a majority of the neutralizing activity does not appear to be antibody mediated.
Over the past 20 years, only a small collection of monoclonal antibodies with relatively broad neutralizing activity against HIV-1 isolates has been generated (2, 14, 46). The epitopes recognized by such MAbs, including both gp120-specific antibodies (b12, 2G12, and 447-52D) and gp41-specific antibodies (2F5, 4E10, and Z13), form a short list of potentially important targets for antibody-mediated neutralization. It is frequently speculated that broadly active, anti-HIV-neutralizing antibodies such as these must be rare among infected individuals. However, this has been a difficult hypothesis to test directly. Assays for antibody specificity are typically assays for binding, using recombinant proteins, peptides, or lysed virions as target antigens; while such assays can be used to catalogue the various epitopes recognized by a polyclonal sample, they do not specify which among the many targets are relevant to neutralization. Moreover, binding to recombinant envelope proteins does not predict neutralizing activity (5, 6, 17, 22, 39, 40, 45, 47-49). Neutralization assays have the opposite problem: while commonly used assays can readily detect and quantify antibody-mediated neutralizing activity in a given plasma sample, such assays do not specify which epitope or combination of epitopes forms the target of neutralization. In some cases, it is possible to approach the problem indirectly: for example, by first depleting a sample with monomeric gp120 and asking whether neutralizing activity remains in the depleted sample or by using a specific protein or peptide as a competitor.
In this study, we reasoned that we could use a heterologous virus engineered to contain limited patches of HIV-1-derived sequence to survey patient samples directly for epitope-specific, neutralizing activity. Toward this end, well-defined HIV-1 epitopes were introduced into the related lentivirus SIVmac239 by site-directed mutagenesis. The epitope sequences for three MAbs (2F5, 4E10, and 447-D) were chosen based on the following two observations. (i) For all three epitopes, the homologous region in the SIV sequence could be readily located due to the proximity of conserved residues (Fig. (Fig.1).1). (ii) Despite the likelihood that these epitopes have conformational components, the minimal elements required for binding of all three MAbs are clearly defined, linear sequences (Fig. (Fig.1).1). The engrafting of complex, nonlinear epitopes will require more detailed structural comparisons between the HIV and SIV envelope complexes.
SIVmac239, the parental clone we used for construction of the epitope-engrafted variants, is largely insensitive to antibody-mediated neutralization (24-26, 36, 37, 50). To our knowledge, monoclonal antibodies with potent neutralizing activity against SIVmac239 have not been identified. In contrast, several MAbs with potent activity against a broad range of HIV-1 isolates have been isolated (e.g., b12, 2G12, 2F5, and 4E10). These observations raise the possibility that SIVmac239 is an extreme outlier with regards to neutralization sensitivity/resistance. This is not a trivial issue, given that the SIVmac239-infected rhesus macaque is a popular model for AIDS vaccine research. However, based on our results, we argue that SIVmac239 is not unusually resistant to neutralization. Introduction of the 2F5 and 4E10 epitopes into SIVmac239 resulted in variants that were readily neutralized by the cognate MAb and at concentrations similar to those seen against primary isolates of HIV-1 (2). Moreover, sensitivity of the engrafted variants was not the result of a global increase in neutralization-sensitivity (Fig. (Fig.6).6). The lack of SIV-specific monoclonal antibodies with neutralizing activity against SIVmac239 is probably not due to an atypical neutralization-resistant phenotype but may indicate that the SIVmac239 Env spike has very low immunogenicity. Plasma from SIVmac239-infected monkeys often displays low-titer neutralizing activity against SIVmac239 (see Fig. Fig.6),6), a situation apparently analogous to a portion of primary HIV-1 infections in humans (52).
Several lines of evidence indicate that the SIV and HIV-1 envelope oligomers are functionally and structurally similar. At the sequence level, conserved features are readily aligned (Fig. (Fig.11 and www.hiv.lanl.gov/content/hiv-db/), crystal structures of the respective gp120 subunits display striking similarities (10, 31), structures of the gp41 ectodomains are nearly superimposable (7-9, 35, 60-63), and peptides derived from the SIV gp41 ectodomain can inhibit both SIV and HIV-1 fusion (35). Finally, our data indicate that sufficient similarity exists between the structures of the SIV and HIV-1 TM proteins such that HIV-derived epitopes are readily recognized by cognate anti-HIV-1 NAbs when introduced into the SIV gp41 context. Given these results, along with published functional and structural similarities between the envelope proteins of HIV-1 and SIV, it is reasonable to assume that the strategies by which SIVmac239 resists antibody-mediated neutralization are largely similar to those displayed by primary isolates of HIV-1.
Creating these epitopes in SIV resulted in reductions in infectivity relative to the parental SIVmac239. Although all of the introduced changes resulted in diminished incorporation of Env oligomers, the extent of reduced incorporation was not dramatic, and there was no correlation between incorporation levels and infectivity. This suggests that the observed decreases in incorporation are a secondary effect and that the substitutions in Env primarily affected the structure/conformation of the complex or its subunits in a manner that resulted in impaired function of the oligomeric spike. The SIV239-447-52D variant displayed only slightly reduced Env incorporation (less than twofold relative to the wild type) but severely diminished infectivity. Given the evidence in the literature for a role of V3 in coreceptor recognition and entry (20), alterations of the SIV239 V3 loop to create the 447-52D epitope may have affected binding to coreceptor and the entry process.
Perhaps more interesting are the two variants with changes in the MPER domain, SIV239-2F5 and SIV239-4E10. Despite similar levels of envelope incorporation, they differed dramatically in infectivity and replication kinetics. Based on an extensive structural analysis of the HIV-1 2F5 epitope, Ofek et al. have suggested that the 2F5 epitope probably interacts with another region of the gp41 ectodomain prior to activation and fusion (43). There is further, anecdotal evidence for intermolecular interactions involving the MPER; for example, we have previously reported the accumulation of compensatory changes in the MPER region during passage of a SIVmac239 variant with a 100-amino-acid deletion of V1-V2 in gp120 (25). We have also described several unrelated alterations in SIV gp120 that give rise to increased sensitivity to a gp41-specific MAb (26), and multiple reports describe substitutions in HIV-1 gp120 or gp41 that point toward interactions of the two subunits (e.g., see references 21, 33, and 59). Perhaps reduced infectivity of SIV239-2F5, and to a lesser extent SIV239-4E10, is due to a weakened or destabilized interaction between the MPER and another region of the oligomer. “Rescuing” infectivity of these variants, either by replacing more of the SIV env sequence with HIV sequence or by repeated passage in cultured cell lines and selection of compensatory changes, could identify interesting interactions within the oligomeric structure.
Replication-competent SIVmac239 derivatives containing recognition elements for the 2F5 and 4E10 neutralizing antibodies were used to screen 92 plasma samples from 47 HIV-positive individuals, several of whom have well-documented neutralizing antibody titers (52), using a quantitative neutralization assay. None of the 92 samples contained measurable levels of neutralizing antibodies against either SIV239-2F5 or SIV239-4E10. A preliminary survey of an additional 25 plasma samples also failed to detect any 2F5-like or 4E10-like neutralizing activity (data not shown). These results are consistent with the hypothesis that MPER-specific neutralizing activity, as typified by MAbs like 2F5 and 4E10, is not common among HIV-positive individuals. Alternatively, MPER-specific neutralizing activity may be present but accounts for only a small percentage of the total, HIV-specific neutralizing activity in a sample and is therefore below the limit of detection. In light of this second possibility, it should be noted that NAb 4E10 has relatively low potency (average IC50 = 7.4 μg/ml against virions pseudotyped with a panel of clade B envelopes) (2). Fifty percent neutralization of the SIV239-4E10 virus was achieved at a concentration of 5 μg/ml of MAb 4E10, which can be used as a rough approximation of the sensitivity of the assay; plasma samples with 4E10-like activity below this value would not be readily detected in our survey. In a survey of a large panel of HIV-1 strains, all of the HIV-specific MAbs tested, including 2F5 and 4E10, were clearly most potent against laboratory strains of the virus (2). Thus, one promising modification to increase sensitivity of the epitope-engrafted virus screen would be to place HIV-derived epitopes into a neutralization-sensitive SIV context, such as SIVmac316 or lab-adapted SIVmac251 (26).
Larger sample sizes will be required to confirm the absence of such activity from HIV-positive cohorts or, if such specificities do arise, to more accurately determine prevalence among HIV-positive individuals. The samples used in this screen were limited to patients infected with clade B viruses, and future comprehensive screens should be expanded to include samples drawn from individuals infected with viruses of other clades. Our results do not address the possibility that HIV-positive patients generate MPER-specific antibodies directed at epitopes not represented by the 2F5 or 4E10 sequences. For example, SIV239-4E10 was not neutralized by Fab Z13, most likely due to inclusion of an asparagine instead of an aspartic acid in the fourth position of the N-W-F-N/D-I-T epitope (66). Thus, it may also be possible to expand the repertoire of epitope-engrafted SIV strains to represent a broader array of HIV-1-derived sequences. SIV variants engrafted with other elements of the HIV-1 envelope could also be used to identify as yet unknown, neutralizing specificities in individual patient samples, for screening and cloning of MAbs from such individuals once they are identified, for confirming specificity of plasma induced in mice or guinea pigs inoculated with recombinant immunogens, and for tracking the presence of passively administered NAbs in HIV-positive patients undergoing antibody therapy.
This work was supported by US Public Health Service grants R01 AI057039 and AI062524 (W.E.J.); by grants AI50421 (R.C.D), AI 33292 (D.R.B.), AI058725 (M.B.Z.), and RR00168 (NEPRC); and by awards from the International AIDS Vaccine Initiative to the Neutralizing Antibody Consortium (D.R.B. and R.C.D.). D.D.R. was supported by grants AI27670, AI043638, AI29164, and AI047745; a grant to the UCSD Center for AIDS Research (AI 36214); and by the Research Center for AIDS and HIV Infection of the San Diego Veterans Affairs Healthcare System.
We thank Susan Zolla-Pazner for the gift of MAb 447-52D, Florence Brunel and Phil Dawson for the 94-1 peptide, and Ann Hessell for IgG/IgA purification.