ABSTRACT
Macrophage tropism of human immunodeficiency virus type 1 (HIV-1) is distinct from coreceptor specificity of the viral envelope glycoproteins (Env), but the virus-cell interactions that contribute to efficient HIV-1 entry into macrophages, particularly via CXCR4, are not well understood. Here, we characterized a panel of HIV-1 Envs that use CCR5 (n = 14) or CXCR4 (n = 6) to enter monocyte-derived macrophages (MDM) with various degrees of efficiency. Our results show that efficient CCR5-mediated MDM entry by Env-pseudotyped reporter viruses is associated with increased tolerance of several mutations within the CCR5 N terminus. In contrast, efficient CXCR4-mediated MDM entry was associated with reduced tolerance of a large deletion within the CXCR4 N terminus. Env sequence analysis and structural modeling identified amino acid variants at positions 261 and 263 within the gp41-interactive region of gp120 and a variant at position 326 within the gp120 V3 loop that were associated with efficient CXCR4-mediated MDM entry. Mutagenesis studies showed that the gp41 interaction domain variants exert a significant but strain-specific influence on CXCR4-mediated MDM entry, suggesting that the structural integrity of the gp120-gp41 interface is important for efficient CXCR4-mediated MDM entry of certain HIV-1 strains. However, the presence of Ile326 in the gp120 V3 loop stem, which we show by molecular modeling is located at the gp120-coreceptor interface and predicted to interact with the CXCR4 N terminus, was found to be critical for efficient CXCR4-mediated MDM entry of divergent CXCR4-using Envs. Together, the results of our study provide novel insights into alternative mechanisms of Env-coreceptor engagement that are associated with efficient CCR5- and CXCR4-mediated HIV-1 entry into macrophages.
INTRODUCTION
The gp120 glycoproteins of the human immunodeficiency virus type 1 (HIV-1) envelope (Env) initiate contact between the virus and the target cell (46). Viral attachment involves binding of gp120 to cellular CD4 and then to either CCR5 or CXCR4 as a coreceptor (reviewed in references 13 and 14). CD4 binding occurs with high affinity and triggers a conformational change in gp120 that exposes the coreceptor binding site. Current models of gp120 binding to coreceptor, supported recently by analysis of the crystal structure of CXCR4 (79), suggest that the crown of the V3 loop interacts principally with the coreceptor second extracellular loop (ECL2) region while the gp120 bridging sheet and the stem of the V3 loop interact with the coreceptor N terminus (7, 11, 25, 39). The interaction of CD4-bound gp120 with coreceptor induces additional conformational changes in gp120, which leads to a structural rearrangement in gp41 that enables fusion and virus entry (reviewed in reference 75).
The tropism of HIV-1 for particular target cell populations in different tissue compartments is influenced by the coreceptor used by HIV-1 Env for virus entry (reviewed in reference 31). Macrophage (M)-tropic HIV-1 viruses primarily use CCR5 (R5) as a coreceptor (1, 9, 12, 16, 17), whereas T-cell tropic viruses use CXCR4 (X4) (26). Dual-tropic viruses can use both coreceptors (R5X4) (10, 81). Thus, the coreceptor specificity of primary HIV-1 isolates is frequently used to define cellular tropism; for example, R5 viruses are often collectively grouped as M-tropic viral strains (31). However, there is a notable distinction between HIV-1 tropism and coreceptor usage (reviewed in references 30 and 31). Several studies have demonstrated the presence of non-M-tropic R5 viruses, which were replication competent in primary CD4+ T cells but which could not productively infect monocyte-derived macrophages (MDM) (32, 37, 45, 57, 58, 60). Thus, while most M-tropic viruses use CCR5 for HIV-1 entry, not all R5 viruses are M-tropic (reviewed in references 31 and 59). In addition, some highly M-tropic primary HIV-1 strains use CXCR4 for entry into MDM (32, 36). Therefore, the viral determinants that underlie HIV-1 tropism for macrophages are significantly more complex than the coreceptor specificity of the virus.
Most previous studies that have characterized the Env determinants contributing to M-tropism of HIV-1 suggest that alterations in gp120 within or proximal to the CD4 binding site (CD4bs) or which occur in other gp120 regions but exert an influence on the conformation of the CD4bs are important for efficient CCR5-mediated HIV-1 entry into macrophages (18–21, 48–50, 57, 58, 73). These gp120 alterations increase the exposure and/or stabilization of the CD4bs. Since gp120-CD4 binding is short-lived and weak compared with gp120-CD4 complex binding to CCR5 (8), such CD4bs alterations may increase the capacity of M-tropic R5 Envs to interact with relatively low levels of CD4 expressed on the macrophage cell surface. Furthermore, since CCR5 is more mobile in the cell membrane than CD4 (68), the higher-affinity Env-CD4 complexes of M-tropic R5 variants may also permit these complexes to more readily colocalize with CCR5, thus indirectly increasing the efficiency of CCR5 usage.
In addition to CD4bs modifications, gp120 alterations that affect the exposure of the coreceptor binding domain may also directly influence the efficiency of CCR5-mediated HIV-1 entry into macrophages (34, 71). Using a panel of R5 HIV-1 Envs cloned from primary HIV-1 isolates, we recently showed that efficient CCR5-mediated entry into MDM was associated with Env variants that had an increased ability to scavenge low levels of cell surface CCR5 and which existed in conformations that had greater exposure of CD4-induced (CD4i) epitopes (71). These Envs had reduced sensitivity to the CCR5 inhibitor maraviroc (MVC) and increased dependence on elements within the CCR5 ECL2 region (71). Further support of a direct link between changes in the mechanism and efficiency of CCR5 engagement and M-tropism is illustrated by our recent studies of HIV-1 resistance to MVC (64). Here, we showed that escape from MVC was a result of HIV-1 Env adopting an increased dependence on the CCR5 N terminus and an altered interaction with the CCR5 ECLs, which diminished the efficiency of the interaction between gp120 and CCR5 and ostensibly abolished the M-tropic properties of the Env (64). Thus, certain gp120 conformations that promote an altered and more efficient mechanism of CCR5 engagement appear to be important for efficient CCR5-mediated entry into macrophages.
While efficient CCR5-mediated HIV-1 entry into macrophages appears to be associated with certain Envs that have a more favored interaction with the CCR5 ECL2 region (71), the dependence on the CCR5 N terminus by these Envs is unclear. In addition, little is known about gp120-coreceptor interactions that may contribute to efficient CXCR4-mediated HIV-1 entry into macrophages. Although one previous study identified Ile326 in the gp120 V3 loop as being important for CXCR4-mediated HIV-1 entry into macrophages (28), the mechanism by which this amino acid variant permits macrophage entry is unknown. Here, we identified and characterized alternative mechanisms of coreceptor engagement by HIV-1 Envs that have efficient CCR5- and CXCR4-mediated entry into macrophages. Our results show that efficient CCR5-mediated MDM entry is associated with Envs that have reduced dependence on elements within the CCR5 N terminus, whereas efficient CXCR4-mediated MDM entry is associated with Envs that have increased dependence on an intact CXCR4 N terminus. In addition, we demonstrate that amino acid variants in the gp41-interactive region of gp120 contribute to efficient CXCR4-mediated MDM entry by certain Envs. Furthermore, we confirm the critical role of Ile326 in the gp120 V3 loop for efficient CXCR4-mediated MDM entry by diverse CXCR4-using Envs and provide structural evidence that Ile326 may promote CXCR4-mediated MDM entry by a functional role as a component of the gp120 V3 loop-CXCR4 N terminus interface.
MATERIALS AND METHODS
Cells.293T cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% (vol/vol) fetal calf serum (FCS) and 100 μg of penicillin and streptomycin per ml. JC53 cells are derived from the HeLa cell line and stably express high levels of CD4, CXCR4, and CCR5 on the cell surface (62); the cells were cultured in DMEM supplemented with 10% (vol/vol) FCS and 100 μg of penicillin and streptomycin per ml. U87-CD4 cells (6) were cultured in DMEM supplemented with 10% (vol/vol) FCS, 100 μg of penicillin and streptomycin per ml, and 300 μg of G418 per ml. NP2-CD4 cells (67) were cultured in DMEM supplemented with 10% (vol/vol) FCS, 100 μg of penicillin and streptomycin per ml, and 500 μg of G418 per ml. NP2-CD4/CCR5 and NP2-CD4/CXCR4 cells (67) were cultured in DMEM supplemented with 10% (vol/vol) FCS, 100 μg of penicillin and streptomycin per ml, 500 μg of G418 per ml, and 1 μg of puromycin per ml. Peripheral blood mononuclear cells (PBMC) were purified from the blood of healthy HIV-1-negative donors by density gradient centrifugation. Monocytes were purified from PBMC by plastic adherence and were allowed to differentiate into MDM by culturing for 5 days in Iscove modified Eagle medium (IMEM) supplemented with 10% (vol/vol) pooled AB-positive (AB+) human serum, 100 μg of penicillin and streptomycin per ml, and 12.5 ng of macrophage colony-stimulating factor per ml.
HIV-1 Env plasmids.The R5 HIV-1 Envs used in this study, NB2-C1, NB2-C4, NB6-C3, NB6-C4, NB7-C1, NB7-C2, NB8-C2, NB8-C4, NB23-C2, NB23-C3, NB24-C4, NB25-C2, NB25-C3, and NB27-C2, were cloned from primary R5 HIV-1 isolates (37, 45) and have been described in detail previously (69–71). The X4 HIV-1 Envs SG3, 3.2, Eli, and HXB2 and the R5X4 Envs Macs1-Spln12, and aBL01 have been described previously (29, 36, 53, 54, 56, 66). All Envs are cloned into the pSVIII-Env expression vector (27).
Production of Env-pseudotyped luciferase reporter viruses.Env-pseudotyped luciferase reporter viruses were produced by transfection of 293T cells with pCMVΔP1ΔenvpA, pHIV-1Luc, and pSVIII-Env plasmids using Lipofectamine 2000 (Invitrogen) at a ratio of 1:3:1, as described previously (35, 69, 80). Supernatants were harvested 48 h later, filtered through 0.45-μm-pore-size filters, and stored at −80°C. The 50% tissue culture infective doses (TCID50) of virus stocks were determined by titration in JC53 cells.
Single-round HIV-1 entry assays.For single-round entry assays using JC53 cells and U87-CD4 cells expressing wild-type (WT) or mutant coreceptors, 1 × 104 cells cultured in 96-well plates were inoculated with 200 TCID50 of Env-pseudotyped luciferase reporter virus (equating to a multiplicity of infection [MOI] of 0.02) in a volume of 100 μl for 12 h at 37°C. The residual inoculum was then removed and replaced with fresh medium, and the cells were incubated for a further 60 h at 37°C. The CXCR4 mutants tested for coreceptor activity on U87-CD4 cells were CXCR4 carrying a 32-amino-acid deletion of residues 4 to 36 in the N terminus (CXCR4 Δ4–36) and R183A, F199A, and D193A mutants of the CXCR4 ECL2 region (7). The CCR5 mutants tested were CCR5 carrying a 15-amino-acid deletion of residues 2 to 17 in the CCR5 N terminus (CCR5 Δ2–17) and D11A, Y14F, Y15A, and E18A mutants of the CCR5 N terminus (15, 23, 24). For single-round entry assays using MDM, cell monolayers that were approximately 90% confluent in 48-well tissue culture plates were inoculated with 1,500 TCID50 of Env-pseudotyped luciferase reporter virus in a volume of 300 μl for 12 h at 37°C. The residual inoculum was then removed and replaced with fresh culture medium, and the MDM were incubated for a further 96 h at 37°C. In all cell types, the level of HIV-1 entry was measured by luciferase activity in cell lysates (Promega), according to the manufacturer's protocol. Luminescence was measured using a FLUOStar microplate reader (BMG Labtech, GmbH, Germany). Negative controls included mock-infected cells that were incubated with culture medium instead of virus and cells inoculated with luciferase reporter virus pseudotyped with the nonfunctional ΔKS Env (22). Macrophage entry levels were normalized against those attained by the same preparations of reporter virus in JC53 cells to more accurately control for virus infectivity, as described previously (57, 58, 60, 71).
Env structural modeling.The crystal structure of gp120 containing the gp41-interactive domain (Protein Data Bank code 3JWO [55]) was used to model amino acid alterations at positions 261 and 263 within the gp120 C2 region. The crystal structure of CD4-bound YU2 gp120 containing the V3 variable loop and docked with the nuclear magnetic resonance (NMR) structure of an N terminus peptide of CCR5 (kindly provided by P. D. Kwong) (38) was used as a template to introduce the V3 loop sequence of gp120 from MACS1-Spln12 Env using the Mutate Protein protocol, as we have described previously (64). The CCR5 peptide sequence (SPI9Y10DINYY15) was mutated to the CXCR4 N terminus sequence (ISI6Y7TSDNY12) by sequence alignment (the homologous residues are underlined and numbered and the sulfated tyrosine residues are highlighted in bold). Harmonic constraints were applied prior to optimization using the Steepest Decent protocol which incorporates iterative cycles of conjugate gradient energy minimization against a probability density function that includes spatial restraints derived from the template and residue-specific properties (65). The interface between gp120 models and the CXCR4 sulfopeptide model was mapped to atoms predicted to be within 4 Å of the ligand.
Env mutagenesis.All gp120 mutants were synthesized by Geneart (Germany) and subcloned into the pSVIII-Env expression vector (27). The authenticity of the gp120 mutants was verified by full-length sequencing.
RESULTS
Relationship between the efficiency of CCR5-mediated HIV-1 entry into macrophages and dependence on the CCR5 N terminus.To better understand the virus-cell interactions contributing to efficient CCR5-mediated HIV-1 entry into macrophages, in particular, the role of the CCR5 N terminus, we studied a panel of R5 Envs (n = 14) that exhibit diversity in a number of pathophysiological phenotypes including levels of macrophage entry (69–71). The macrophage entry levels by the R5 Envs are broad and range from those that enter macrophages poorly (similar to JR-CSF Env) to those that enter macrophages efficiently (similar to ADA and JR-FL Envs) (71).
Env-pseudotyped luciferase reporter viruses were used in single-round HIV-1 entry assays in MDM or U87-CD4 cells expressing either WT CCR5 or CCR5 with various N terminus mutations shown previously to attenuate coreceptor activity (15, 23, 24). To accurately control for virus infectivity when MDM entry was measured, the level of MDM entry was normalized against that attained in parallel infections of JC53 cells, as we along with others have described previously (57, 58, 60, 71). The level of HIV-1 entry in U87-CD4 cells expressing CCR5 mutants was expressed as a percentage of that attained in cells expressing WT CCR5 and plotted against the level of MDM entry achieved by the same preparations of virus. The CCR5 Δ2–17 deletion mutation abolished the coreceptor activity against all of the Envs (data not shown). However, we observed positive correlations between the level of MDM entry and the ability of Env to tolerate discrete mutations of amino acids D11 (Fig. 1 A), Y14 (Fig. 1B), and E18 (Fig. 1C) in the CCR5 N terminus and a nonsignificant trend toward increased tolerance of mutation of Y15 (data not shown). In addition, we observed a similar trend for the level of tolerance of each of the CCR5 mutants by the individual Envs when they were ranked from the least tolerant to the most tolerant (Fig. 1). Similar levels of WT and mutant coreceptor were expressed on the U87-CD4 cell populations (Fig. 1). These results suggest that efficient CCR5-mediated HIV-1 entry into macrophages is associated with reduced dependence of gp120 on elements within the CCR5 N terminus, in particular, amino acids D11, Y14, and E18, but that an intact N terminus is universally required for CCR5 to maintain coreceptor activity.
Efficient CCR5-mediated MDM entry is associated with reduced dependence on residues within the CCR5 N terminus. The ability of Env-pseudotyped luciferase reporter viruses to enter U87-CD4 cells expressing CCR5 with mutations at D11 (A), Y14 (B), or E18 (C) was expressed as a percentage of that attained in U87-CD4 cells expressing similar levels of WT CCR5, as shown by the flow cytometry profiles, and plotted against the relative levels of MDM entry achieved by the same preparations of virus, using Prism, version 4.0c (GraphPad Software). Flow cytometry was performed using the 2D7 anti-CCR5 MAb. The middle column shows the ability of the individual Envs to tolerate each of the CCR5 mutations, from the least tolerant to the most tolerant. The data are compiled from independent experiments conducted with three different MDM donors, each tested in triplicate. Error bars represent standard errors of the means. The Spearman correlation coefficient (r) and P values are shown. P values of <0.05 were considered statistically significant. Max, maximum.
Identification and characterization of HIV-1 Envs with highly efficient CXCR4-mediated macrophage entry.Since flexibility in the recognition of CCR5 by gp120 can modulate the efficiency of CCR5-mediated MDM entry, we next reasoned that alterations in the way gp120 interacts with CXCR4 may influence the efficiency of CXCR4-mediated MDM entry. We therefore selected a panel of X4 or R5X4 Envs that, from previous studies (29, 32, 36, 53, 54, 66), were predicted to have either efficient or inefficient CXCR4-mediated MDM entry. These included the X4 SG3, 3.2, Eli, and HXB2 Envs and R5X4 Macs1-Spln12 and aBL01 Envs. To determine how efficiently these Envs enter macrophages, Env-pseudotyped luciferase reporter viruses were used in single-round entry assays in MDM, as described above (Fig. 2A). The results show that 3.2, Eli, and HXB2 Envs have relatively inefficient MDM entry compared to the entry levels of SG3, Macs1-Spln12, and aBL01 Envs. The entry levels of 3.2, Eli, and HXB2 Envs were comparable to those achieved by the non-M-tropic R5 JR-CSF Env, and the entry levels of SG3 and aBL01 Envs were comparable to those achieved by the M-tropic R5 ADA and JR-FL Envs (data not shown). Most notably, Macs1-Spln12 Env displays very efficient MDM entry, consistent with previous studies using this clone (33) and the parental virus isolate (32).
Variation in levels of CXCR4-dependent MDM entry. (A) Monocyte-derived macrophages were inoculated with equivalent amounts of Env-pseudotyped luciferase reporter virus, and entry levels were expressed relative to that attained by the same preparations of reporter virus in JC53 cells, as described previously (57, 58, 60, 71). (B) Monocyte-derived macrophages were inoculated with equivalent amounts of luciferase reporter virus pseudotyped with aBL01 or Macs1-Spln Env after treatment of cells with MVC (1 μM), AMD3100 (1 μM), or both inhibitors. Virus entry levels were expressed as a percentage of that attained in untreated cells. The results shown are means of triplicate wells and are representative of four independent experiments using MDM obtained from different donors. Error bars represent standard errors of the means.
Given that Macs1-Spln12 and aBL01 Envs have an R5X4 phenotype (36, 53) and that certain R5X4 Envs may use either CCR5 or CXCR4 to enter MDM (35, 82, 83), we next determined the coreceptor preference of these Envs in MDM. MDM were either left untreated or treated with the CCR5 inhibitor MVC, the CXCR4 inhibitor AMD3100, or a combination of both inhibitors prior to inoculation with Env-pseudotyped luciferase reporter virus. The level of virus entry in the presence of either or both inhibitors was expressed as a percentage of that attained in untreated cultures (Fig. 2B). aBL01 and Macs1-Spln12 Envs were potently inhibited by AMD3100 but only marginally or slightly inhibited by MVC, respectively. In control R5 virus infections, MVC completely inhibited the entry of ADA Env, but AMD3100 had no inhibitory effect (data not shown). In addition, the X4 SG3, 3.2, HXB2, and Eli Envs were potently inhibited by AMD3100 but unaffected by MVC (data not shown). The combination of both inhibitors completely inhibited the entry of all viruses (Fig. 2B and data not shown). Therefore, we confirmed that the selected X4 and R5X4 Envs use CXCR4 exclusively or predominantly for HIV-1 entry into macrophages.
Relationship between the efficiency of CXCR4-mediated HIV-1 entry into macrophages and dependence on an intact CXCR4 N terminus.To determine the relationship between the efficiency of CXCR4-mediated MDM entry and dependence on an intact CXCR4 N terminus, U87-CD4 cells expressing either WT CXCR4 or CXCR4 containing a 32-amino-acid deletion in the N terminus (Δ4–36) (7) were inoculated with Env-pseudotyped luciferase reporter viruses. We specifically chose this particular CXCR4 mutant as we showed previously that it readily discriminates Envs that have functional differences in the recognition of CXCR4 by gp120 (36). The level of HIV-1 entry in U87-CD4 cells expressing the CXCR4 mutant was expressed as a percentage of that attained in cells expressing WT CXCR4 and plotted against the level of CXCR4-mediated MDM entry (Fig. 3). In these MDM infections, cells were treated with MVC to block any residual CCR5-mediated entry that could occur with the R5X4 Macs1-Spln12 and aBL01 Envs. In contrast to efficient CCR5-mediated MDM entry which was associated with increased tolerance of amino acid substitutions in the CCR5 N terminus (Fig. 1), the efficiency of CXCR4-mediated MDM entry correlated significantly with reduced tolerance of deletion of a large part of the CXCR4 N terminus. Furthermore, we saw no association between the efficiency of CXCR4-mediated MDM entry and tolerance of R183A, D193A, and F199A mutations in the CXCR4 ECL2 region (data not shown). This, again, is in contrast to efficient CCR5-mediated MDM entry in which there is increased dependence on related elements within the CCR5 ECL2 region (71). Similar levels of WT and mutant CXCR4 were expressed on the U87-CD4 cell populations (Fig. 3). Therefore, our results suggest that efficient CXCR4-mediated HIV-1 entry into macrophages is associated with increased dependence of gp120 on an intact CXCR4 N terminus.
Efficient CXCR4-mediated MDM entry is associated with increased dependence on an intact CXCR4 N terminus. The ability of Env-pseudotyped luciferase reporter viruses to enter U87-CD4 cells expressing the CXCR4 Δ4–36 deletion mutant was expressed as a percentage of that attained in U87-CD4 cells expressing similar levels of WT CXCR4, as shown by the flow cytometry profiles, and plotted against the relative levels of MDM entry achieved by the same preparations of virus in the presence of 1 μM MVC, using Prism, version 4.0c (GraphPad Software). Flow cytometry was performed using the 12G5 anti-CXCR4 MAb. The data are compiled from independent experiments conducted with three different MDM donors, each tested in triplicate. The Spearman correlation coefficient (r) and P values are shown. P values of <0.05 were considered statistically significant.
Amino acid alterations within the gp41-interactive domain and the V3 loop of gp120 distinguish M-tropic from non-M-tropic CXCR4-using Envs.We next analyzed the amino acid sequences of the six CXCR4-using Envs to better understand Env determinants that may contribute to efficient CXCR4-mediated HIV-1 entry into macrophages. Sequence analysis identified amino acid variations at positions 261 and 263 of the C2 region of gp120 in SG3 (Ser261 and Lys263), Macs1-Spln12 (Lys261 and Lys263), and aBL01 (Arg261 and Glu263) Envs (Fig. 4A). In contrast, HXB2, 3.2, and Eli Envs, all of which enter MDM less efficiently, have consensus Pro261 and Thr263 residues. Molecular models using the recently described 3JWO gp120 crystal structure showed that these alterations occur within the gp41-interactive region of gp120 (55) (Fig. 4B) and span Cys262, which forms a highly conserved disulfide bond with Cys252 (Fig. 4C).
Env sequence alterations associated with efficient CXCR4-mediated MDM entry. (A) The C2 region (residues 251 to 269) and V3 region (residues 297 to 332) of SG3, Macs1-Spln12, aBL01, HXB2, 3.2, and Eli Envs were aligned against the clade B consensus sequence. The relative levels of MDM entry by these Envs when they were pseudotyped onto luciferase reporter virus were scored as described previously (71). Amino acid sequence changes associated with efficient CXCR4-mediated MDM entry are highlighted in gray. (B) The gp120 structure containing the gp41-interactive domain (55) is shown in ribbon representation, with the inner and outer domains shown in light and dark gray, respectively. The positions of Pro261 and Thr263 are highlighted by Corey-Pauling-Koltun (CPK) representation of their α-carbon atoms (red). Residues involved in gp41 interactions are shown in blue stick and blue van der Waals surface representation. (C) Close-up view of the β6 strand of gp120 colored as described for panel B, showing the position of Pro261 and Thr263 relative to the disulfide bond between Cys252 and Cys262, shown in orange stick representation. (D) Three-dimensional model of CD4-bound YU2 gp120 containing the V3 variable loop sequence of Macs1-Spln12 Env and docked with a model of the N terminus peptide of CXCR4. The gp120 protein is colored as described above, the V3 variable loop is shown in blue, and the CXCR4 peptide is shown in orange stick representation with its molecular surface in gray. The position of Ile326 is highlighted by CPK representation of the α-carbon atom (red). (E) Close-up of the binding interface of the gp120 V3 loop and the CXCR4 peptide model, with the potential hydrogen bond formed between Ile326 of gp120 (red stick) and Tyr7 of CXCR4 (yellow stick) shown as a green dotted line.
Env sequence analysis also identified Ile at position 326 within the V3 region of gp120 of the M-tropic SG3, Macs1-Spln12, and aBL01 Envs, whereas the non-M-tropic HXB2, 3.2, and Eli Envs had either Met or a deletion at this position (Fig. 4A). The association between the presence of Ile326 and efficient CXCR4-mediated MDM entry is consistent with the results of a previous study which showed that Ile326 was important for the entry of other CXCR4-using viruses into macrophages (28), but the mechanism by which Ile326 promotes CXCR4-mediated macrophage entry is unknown. To gain insights into how Ile326 may potentiate HIV-1 entry into macrophages via CXCR4, we produced structural models of gp120 containing the V3 loop of Macs1-Spln12 Env docked to a molecular model of the CXCR4 N terminus. These models confirmed that Ile326 is located within the gp120 V3 loop stem (Fig. 4D) and suggest that Ile326 may interact directly with Tyr7 of the CXCR4 N terminus when Tyr7 is modeled as a sulfotyrosine (Fig. 4E), in a similar fashion to how Ile326 of the R5 YU2 Env is known to interact with sulfated Tyr10 of the CCR5 N terminus (25, 39). Similar results were obtained using structural models of gp120 containing the V3 loop of SG3 Env (data not shown). Thus, our models provide evidence that Ile326 of CXCR4-using Envs may be a component of the interface between gp120 and the CXCR4 N terminus that is necessary for HIV-1 entry into macrophages.
Role of gp120 V3 loop and gp41 interaction domain alterations in promoting efficient CXCR4-mediated macrophage entry.To determine whether Ile326 in the gp120 V3 loop is important for efficient CXCR4-mediated MDM entry by the Envs studied here and to determine whether amino acid alterations in the gp41 interaction domain of gp120 may also promote CXCR4-mediated MDM entry, we next constructed a panel of Env mutants using both SG3 and Macs1-Spln12 Envs as templates. The amino acid sequences of the Env mutants relative to their respective wild type (WT) sequences are shown in Fig. 5. SG3-S261P contains a Ser-to-Pro mutation at position 261; SG3-K263T contains a Lys-to-Thr mutation at position 263; SG3-S261P/K263T has both of the preceding mutations present; SG3-I326M contains an Ile-to-Met mutation at position 326 in the V3 loop. An identical set of mutants was made for Macs1-Spln12. All of the Env mutants were shown to be functional and capable of mediating HIV-1 entry into JC53 cells when they were pseudotyped onto luciferase reporter viruses (data not shown). HIV-1 entry assays in NP2-CD4/CCR5 and NP2-CD4/CXCR4 cells confirmed that the introduced mutations did not alter the coreceptor specificity of the Envs (Fig. 5).
Env mutants and their coreceptor usage. The amino acid sequences of the Env mutants were aligned against their respective parental WT Env sequences. The coreceptor usage of the WT and mutant Envs, when pseudotyped onto luciferase reporter viruses and tested in single-round entry assays with NP2-CD4 cells stably expressing CCR5 or CXCR4 (67), is shown on the right.
Equivalent infectious titers of luciferase reporter virus pseudotyped with WT SG3 or Macs1-Spln12 Env or with an alternative Env mutant were used in single-round entry assays in MDM as described above. For SG3, the K263T Env mutant resulted in a relatively modest but significant reduction in macrophage infectivity via CXCR4 (Fig. 6A). This was not observed for the double S261P/K263T Env mutant, suggesting that the influence of amino acid variation at position 263 on macrophage infectivity was counteracted by a nonsignificant trend toward an increase in macrophage infectivity conferred by the S261P Env mutant (Fig. 6A). However, for Macs1-Spln12, macrophage infectivity via CXCR4 was potently abrogated by the K261P Env mutant, the K263T Env mutant, and the double K261P/K263T Env mutant (Fig. 6B). These results indicate that the gp41 interaction domain variants of gp120 identified in Fig. 4 contribute significantly to efficient CXCR4-mediated MDM entry of Macs1-Spln12 but exert a variable influence on the MDM entry of SG3. In contrast, the I326M Env mutants which replace Ile with Met at position 326 in the gp120 V3 loop potently abrogated the macrophage infectivity of both SG3 and Macs1-Spln12 Envs (Fig. 6A and B). Together, our results suggest that the gp41 interaction domain of gp120 exerts a significant but largely strain-specific influence on the efficiency of CXCR4-mediated MDM entry and confirm an important role for Ile326 in the gp120 V3 loop for promoting efficient CXCR4-mediated MDM entry of diverse CXCR4-using Envs.
The effect of mutations in the gp41 interaction domain and the V3 loop of gp120 on CXCR4-mediated MDM entry. Monocyte-derived macrophages were inoculated with luciferase reporter virus pseudotyped with Env mutants derived from SG3 (A) or Macs1-Spln12 (B) Envs in the presence of 1 μM MVC. HIV-1 entry levels were expressed as a percentage of that achieved by equivalent infectious units of reporter virus pseudotyped with the respective WT Envs. The results shown are a compilation of data obtained using MDM from four different donors, each tested in triplicate. Error bars represent standard errors of the means. The combined data sets were analyzed with a Wilcoxon signed ranks test using Prism, version 4.0c (GraphPad Software), and P values of <0.05 were considered statistically significant changes in CXCR4-mediated MDM entry by an Env mutant compared to WT Env. *, P < 0.02; **, P < 0.001.
DISCUSSION
Our results illustrate alternative mechanisms of coreceptor engagement for efficient CCR5- and CXCR4-mediated MDM entry. The reduced dependence on elements within the CCR5 N terminus by R5 Envs with efficient MDM entry extends our recent studies showing that the efficiency of MDM entry by these Envs correlates with increased dependence on H181 and Y184 in the CCR5 ECL2 region and reduced sensitivity to inhibition by MVC (71). Sequence analysis and structural modeling of gp120 revealed that alterations within the V3 region, the gp41 interaction sites, and the CD4 binding site (CD4bs) may directly or indirectly lead to more efficient recognition of CCR5 by gp120 (71). Together, these results suggest that efficient CCR5-mediated MDM entry is dependent on Envs that have a more efficient CCR5 interaction favoring the ECL2 region.
In contrast, our results which show a correlation between reduced tolerance of the CXCR4 Δ4–36 mutation and efficiency of CXCR4-mediated MDM entry (Fig. 3) support the interpretation that M-tropic CXCR4-using Envs have increased dependence on an intact CXCR4 N terminus. However, further studies with single amino acid substitutions within the CXCR4 N terminus are required to determine whether these Envs have increased dependence on the CXCR4 N terminus per se. Our results suggest that compared to the mechanism of coreceptor engagement by M-tropic R5 Envs, different interactions between gp120 and CXCR4 are necessary for CXCR4 expressed on macrophages to support efficient HIV-1 entry. These findings are consistent with the results of a previous study by Yi et al. (84), who showed that the R5X4 89.6 Env, which has relatively inefficient CXCR4-mediated MDM entry compared to CCR5-mediated MDM entry (47), is able to substantially tolerate substitution of the CXCR4 N terminus for that of CXCR2 when it is used in cell-cell fusion assays. In contrast to the reduced sensitivity to inhibition by MVC observed by M-tropic R5 Envs (71), which suggests a more efficient interaction with CCR5, we saw no association between the efficiency of CXCR4-mediated MDM entry and sensitivity to inhibition by AMD3100 (data not shown). These results suggest that M-tropic CXCR4-using Envs use CXCR4 differently but not necessarily more efficiently. In addition, we saw no association between the efficiency of CXCR4-mediated MDM entry and sensitivity to neutralization by the Env MAb IgG1b12 whose gp120 epitope overlaps the CD4bs (data not shown), suggesting that CXCR4-mediated M-tropism by the Envs studied here is probably not influenced by alterations in gp120-CD4 binding.
To gain mechanistic insights into the Env-coreceptor interactions that contribute to efficient HIV-1 entry into macrophages via an increased reliance on an intact CXCR4 N terminus, we undertook Env sequence analysis and structural modeling, which identified amino acid variants at positions 261 and 263 within the gp41-interactive region of gp120 that were associated with efficient CXCR4-mediated MDM entry. Mutagenesis studies confirmed that amino acid variation at either or both of these positions contributed to efficient CXCR4-mediated HIV-1 entry into macrophages by Macs1-Spln12 Env but not consistently for SG3 Env, suggesting a strain-dependent influence of alterations at the gp120-gp41 interface on the efficiency of CXCR4-mediated MDM entry.
Precisely how gp41 interaction domain variants of gp120 potentiate efficient HIV-1 entry into macrophages via CXCR4 is unclear, especially since there is no evidence that these domains of gp120 interact with coreceptor. However, it is noteworthy that several early studies demonstrated the functional importance of the gp120-gp41 interaction for recognition of HIV-1 by neutralizing antibodies directed at the CD4bs (43, 63, 72, 78). Moreover, gp41 interaction domain variants have been observed in certain R5 Envs displaying enhanced MDM entry (71), and alterations within gp41 itself can affect gp120-coreceptor interactions through an as yet unexplained mechanism (2–4). To better understand the potential mechanisms underlying efficient MDM entry promoted by gp41 interaction domain variants, the molecular interactions of the consensus residues at amino acid positions 261 (Pro) and 263 (Thr) of gp120 were modeled on the recently described 3JWO crystal structure of gp120 (55). These residues span Cys262, which forms a highly conserved disulfide bond with Cys252 that is critical for conformational folding of gp120 (44, 77). In addition, the consensus Pro261 residue has been identified by mutagenesis studies to be important for gp120-gp41 interactions (55). Furthermore, Pro261 is located between Asn257 and Asn264, two conserved sites for N-linked glycosylation (Fig. 4A). Thus, mutation of Pro261 is likely to alter the conformation of the preceding polypeptide containing the potentially glycosylated Asn257 and may also affect the positioning of the disulfide bond formed between Cys262 and Cys252, which together may affect the accessibility of gp41 for gp120. Therefore, conformational changes in the gp120-gp41-interactive region influenced by amino acid alterations flanking Cys262 and/or by mutation of Pro261 may alter activation events involved in gp120-gp41 association, which may be required for efficient CXCR4-mediated MDM entry of certain HIV-1 strains.
Sequence analysis also identified the presence of Ile326 in the gp120 V3 loop, which segregated Envs with efficient CXCR4-mediated HIV-1 entry into macrophages from those which enter macrophages poorly. The results of our mutagenesis studies confirmed that Ile326 is critically important for efficient HIV-1 entry into macrophages by both SG3 and Macs1-Spln Envs. Our results support and extend the findings of Ghaffari et al. (28), where, in a reciprocally designed experiment, replacement of Met for Ile at position 326 in two non-M-tropic CXCR4-using Envs enhanced macrophage entry via CXCR4. Together, these studies provide independent evidence that affirms a critical role for Ile326 in promoting efficient CXCR4-mediated HIV-1 entry into macrophages.
Our molecular modeling studies, when coupled with the functional data, provide new mechanistic insights into how Ile326 may promote HIV-1 entry into macrophages via CXCR4. First, we showed that Ile326 is positioned within the stem of the V3 loop, which is a known structural component of the interface between gp120 and the coreceptor N terminus (7, 11, 25, 39). Second, we produced a new structural model of gp120 bound to the CXCR4 N terminus, which showed that Ile326 has the potential to interact directly with Tyr7 of the CXCR4 N terminus when Tyr7 is modeled as a sulfotyrosine, similar to how Ile326 of R5 Envs is known to interact with sulfated Tyr10 of the CCR5 N terminus (25, 39). When coupled with the strong association observed between the efficiency of CXCR4-mediated entry into macrophages by HIV-1 Envs that have Ile326 and increased reliance on an intact CXCR4 N terminus by these Envs, our data support a model whereby Ile326 may potentiate efficient CXCR4-mediated macrophage entry by facilitating an interaction between gp120 and the CXCR4 N terminus that is critical for macrophages to support HIV-1 entry via CXCR4. We acknowledge, however, that at least part of this proposed model is speculative since it relies on predicted structures of gp120 and CXCR4. Crystal structures of gp120 bound to CXCR4 or N-terminal peptides of CXCR4 are required to fully understand the mechanism of gp120 engagement to CXCR4 and the role that alterations in this mechanism have on macrophage entry by HIV-1 via CXCR4.
Interestingly, the R5 Env clone NB24-C4 had moderate levels of macrophage infectivity but was the most sensitive of the clones tested to alteration of D11, Y14, and E18 of the CCR5 N terminus (Fig. 1). This clone was also highly sensitive to Y15 of the CCR5 N terminus but could tolerate mutation of H181 and Y184 of the CCR5 ECL2 region reasonably well when it was used in cell-cell fusion assays (71), suggesting a specific increased reliance on several elements with the N-terminal region of CCR5 by this Env. This pattern of increased reliance on different N-terminal CCR5 residues is reminiscent of that which we (64) along with others (5, 51, 52, 61, 76) have described as a mechanism involved in HIV-1 escape from CCR5 antagonists. This raises the possibility that NB24-C4 may have some basal resistance to CCR5 antagonists, similar to that which has been described previously for a virus with basal resistance to aplaviroc (74). Further studies which characterize the plateaus of inhibition of NB24-C4 by CCR5 antagonists are required to determine if this is the case.
Our study has some limitations and points requiring clarification. First, the majority of the Envs used were cloned from cultured primary HIV-1 isolates and, thus, may carry mutations that are the result of virus culture which may not be relevant for replication in vivo. Further studies of Envs amplified directly from blood or tissues are required to fully understand the mechanisms of coreceptor engagement that are required for CCR5- and CXCR4-mediated M-tropism in vivo. Second, while the results shown in Fig. 1 demonstrate a spread in the tolerance of R5 Envs for mutations at D11 and Y14 in the CCR5 N terminus when they are pseudotyped onto luciferase reporter viruses, these mutations abolished the coreceptor activity of CCR5 when these Envs were used in a cell-cell fusion assay (71). These discrepant findings suggest that gp120-coreceptor interactions may differ for cell-cell spread of HIV-1 and cell-free HIV-1 infection. Third, the results of our study differ from those of Peters et al. (57, 58) and Isaacman-Beck et al. (42), who did not observe correlations between CCR5-mediated M-tropism and sensitivity to inhibition by CCR5 antagonists or alterations in the mechanism of CCR5 engagement, respectively. The R5 Envs used by Peters et al. (57, 58) have well-defined mutations which alter the conformation of the CD4bs that are important for M-tropism, whereas the R5 Envs in the present study do not. Moreover, the Envs studied by Isaacman-Beck et al. (42) are subtype C, whereas the Envs studied here are subtype B (69, 70), suggesting that gp120-coreceptor interactions which contribute to CCR5-mediated M-tropism may not be common to HIV-1 of different subtypes.
In summary, our results provide new mechanistic insights into the virus-cell interactions involved in efficient CCR5- and CXCR4-mediated HIV-1 entry into macrophages. In addition, since tissue macrophages are considered to be a significant reservoir of both CCR5- and CXCR4-using HIV-1 strains that fuel high plasma viral loads at late stages of infection after virtually all CD4+ T cells are depleted (40, 41), our results have potentially important implications for HIV-1 pathogenesis and treatment. In particular, our results may help guide the design and development of novel HIV-1 coreceptor inhibitors with increased activity against M-tropic HIV-1 variants, which may assist in reducing the tissue macrophage burden of HIV-1.
ACKNOWLEDGMENTS
We thank J. Sodroski and D. Gabuzda for providing Env plasmids, J. Sodroski for providing pCMVΔP1ΔenvpA and pHIV-1Luc plasmids, D. Kabat for providing JC53 cells, J. Sodroski, R. Doms, and A. Brelot for providing CCR5 and CXCR4 mutants, N. Shimizu and H. Hoshino for permission to use NP2-CD4 cells and coreceptor-expressing derivatives, and D. Mosier and R. Nedellec for supplying the NP2-based cell lines. We also thank M. Gouillou and T. Mota for advice with statistics. Maraviroc was provided by Pfizer. The following reagent was obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: Bicyclam JM-2987 (hydrobromide salt of AMD3100).
This study was supported, in part, by a grant from the Australian Center for HIV and Hepatitis Virology Research (ACH2) to P.R.G. and M.J.C. M.R. is supported by a Monash University Postgraduate Research Scholarship. P.R.G. is the recipient of an Australian National Health and Medical Research Council (NHMRC) Level 2 Biomedical Career Development Award. P.A.R. is the recipient of an Australian NHMRC R. Douglas Wright Biomedical Career Development Award. The authors gratefully acknowledge the contribution to this work of the Victorian Operational Infrastructure Support Program received by the Burnet Institute.
FOOTNOTES
- Received 29 June 2011.
- Accepted 1 August 2011.
- Accepted manuscript posted online 10 August 2011.
- Copyright © 2011, American Society for Microbiology. All Rights Reserved.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
