CD4-Independent Binding of SIV gp120 to Rhesus CCR5

See allHide authors and affiliations

Science  21 Nov 1997:
Vol. 278, Issue 5342, pp. 1470-1473
DOI: 10.1126/science.278.5342.1470


CCR5 and CD4 are coreceptors for immunodeficiency virus entry into target cells. The gp120 envelope glycoprotein from human immunodeficiency virus strain HIV-1(YU2) bound human CCR5 (CCR5hu) or rhesus macaque CCR5 (CCR5rh) only in the presence of CD4. The gp120 from simian immunodeficiency virus strain SIVmac239 bound CCR5rh without CD4, but CCR5hu remained CD4-dependent. The CD4-independent binding of SIVmac239 gp120 depended on a single amino acid, Asp13, in the CCR5rh amino-terminus. Thus, CCR5-binding moieties on the immunodeficiency virus envelope glycoprotein can be generated by interaction with CD4 or by direct interaction with the CCR5 amino-terminus. These results may have implications for the evolution of receptor use among lentiviruses as well as utility in the development of effective intervention.

HIV-1 entry is mediated by the viral envelope glycoprotein complex consisting of the exterior glycoprotein, gp120, and the transmembrane glycoprotein, gp41 (1). CD4 acts as the primary target cell receptor, binding to the gp120 glycoprotein (2). The gp120-CD4 complex requires an additional target cell coreceptor of the seven-transmembrane-spanning G protein–coupled receptor family (3-5). Formation of this complex is believed to induce a conformational change exposing the gp41 ectodomain, which mediates fusion of the viral and target cell membranes (6).

Binding of CD4 to HIV-1 gp120 exposes structural epitopes detectable with monoclonal antibodies (mAbs) (7). These epitopes include conserved and variable structures on gp120 that are implicated in binding to the coreceptor (4, 5) and suggest a stepwise mechanism by which macrophage-tropic strains of HIV-1 achieve fusion and entry into host cells. Sequential binding of CD4 followed by coreceptor may enable the virus to shield its fusion-sensitive epitopes from the immune system until contact with the target cell surface.

CCR5 cloned from rhesus macaques (CCR5rh) is highly similar to human CCR5 (CCR5hu) (98.3% sequence identity), with only eight amino acid changes (8). Five are conservative substitutions in transmembrane domains; the others are a Lys171 → Arg substitution in the second extracellular loop and two nonconservative substitutions in the NH2-terminus (rhesus → human Thr9 → Ile, Asp13 → Asn) (8). These substitutions do not restrict viral entry, as primary isolates of SIV and HIV efficiently infect CD4+ cells expressing rhesus or human CCR5 (8,9). Rhesus and human CCR5 also support entry by interaction with a wide spectrum of SIV (macrophage- and T cell–tropic) and HIV (macrophage-tropic) envelope glycoproteins expressed on recombinant virions (8, 9). The natural ligands for CCR5, the β chemokines MIP-1α, MIP-1β, and RANTES, inhibit infection by HIV-1, HIV-2, and SIV (10). Here, we examined the interactions of HIV-1 and SIV envelope glycoproteins with rhesus and human CCR5.

The system used involved intact HEK293 cells transiently transfected with expression plasmids for CCR5rh or CCR5hu (11). 125I-labeled MIP-1β bound CCR5hu (12) and CCR5rh with dissociation constants (K d) of 0.4 and 1.2 nM, respectively, comparable to the K d of 0.85 nM observed for endogenous receptors on human primary macrophage cultures (13). 125I-labeled MIP-1α bound with slightly lower affinity to either receptor (K d = 3.5 nM, CCR5rh; K d = 6.4 nM, CCR5hu). Thus, the transfected cell system provides an adequate model of the native conformation of these receptors.

HIV-1(YU2) is a primary macrophage-tropic isolate that uses CCR5 as a coreceptor (4). Recombinant YU2 gp120 envelope glycoprotein bound CCR5rh in transfected HEK293 cells with aK d of 5.2 nM in the presence of 100 nM soluble CD4 (sCD4) (14). In the absence of sCD4, specific binding was not observed (Fig. 1). YU2 gp120 competed with 125I-labeled MIP-1β for binding to rhesus or human CCR5 in the presence of sCD4 (4, 13). Thus, YU2 gp120 binds to CCR5rh with an affinity nearly identical to that observed for CCR5hu. As for CCR5hu(4, 5), CD4 is required for efficient interaction of YU2 gp120 with CCR5rh.

Figure 1

HIV-1(YU2) gp120 binding to rhesus CCR5 in transiently transfected HEK293 cells. Cells (106) expressing CCR5rh were incubated with 1.0 nM125I-labeled YU2 gp120 in the presence (▪) or absence (□) of 100 nM sCD4 and the indicated concentrations of unlabeled YU2 gp120. Samples were processed as described (12). Data are expressed as mean cpm ± SEM. Data are representative of six experiments.

Rhesus and human CCR5 efficiently support entry of SIVmac239 (8,9) and other SIV strains (8) in CD4-expressing cells. Direct interactions among SIVmac envelope glycoproteins, CD4, and CCR5 have not been demonstrated. In the presence of 100 nM sCD4, SIVmac239 gp120 competed with 125I-labeled SIVmac239 gp120 (Fig.2A) or 125I-labeled MIP-1α (Fig. 2B) for binding to CCR5rh, with aK d of 8.19 ± 1.06 nM (n = 11) (14). Binding was dependent on the presence of the V3 loop, as a SIVmac239 ΔV3 gp120 mutant lacking this region failed to compete with 125I-labeled SIVmac239 gp120 or125I- labeled MIP-1α at 1000-fold molar excess (Fig.2B). In contrast to CD4-dependent binding of macrophage-tropic HIV-1 glycoproteins, SIVmac239 gp120 bound CCR5rh without sCD4, with a K d of 14.36 ± 3.90 nM (n = 11) (Fig. 2A). The difference in affinities with or without sCD4 was not statistically significant (P = 0.12); however, in the presence of sCD4, 20 to 40% more gp120 was bound. SIVmac239 gp120 also bound CCR5hu in HEK293 cells in the presence of sCD4, with a K d of 2.7 nM (Fig.3C). However, binding to CCR5hu was CD4-dependent (Fig. 2C).

Figure 2

SIVmac239 gp120 binding to rhesus or human CCR5 and CCR5 mutants in transiently transfected HEK293 cells. Cells (106) expressing (A) CCR5rh(squares), (C) CCR5hu (triangles), (D) CCR5rh D13N (circles), or (E) CCR5huN13D (diamonds) were incubated with 1.0 nM 125I-labeled SIVmac239 gp120 and the indicated concentrations of unlabeled SIVmac239 gp120 in the presence (solid symbols) or absence (open symbols) of 100 nM sCD4, as described (15). Data are expressed as mean cpm ± SEM. (B) Cells (106) expressing CCR5rhwere incubated with 0.1 nM 125I-labeled MIP-1α and the indicated concentrations of unlabeled SIVmac239 gp120 in the presence (▪) or absence (□) of 100 nM sCD4; 100 nM SIVmac239 gp120 lacking the V3 loop (ΔV3) was also assayed as a competitor in the presence of 100 nM sCD4 (X). Data are expressed as mean specific binding (cpm ± SEM), where nonspecific binding in the presence of 200 nM unlabeled SIVmac239 gp120 is subtracted from total binding. Data are representative of nine (A), two (B, C, and E), or three (D) independent experiments.

Figure 3

HIV-2 mAb inhibition of125I-labeled SIVmac239 gp120 binding to rhesus CCR5 in transiently transfected HEK293 cells. (A) Cells (106) were incubated with 1.0 nM 125I-labeled SIVmac239 gp120 and 100 nM sCD4 alone, with 300 nM unlabeled SIVmac239 gp120, or with 5 μg of the indicated HIV-2 mAbs [B23, 110C, and 17A react with SIVmac239 as well as other SIV strains (20, 21); 23F and 34G do not react with SIVmac239; 15D reacts weakly and 26G reacts strongly but is non-neutralizing; CG10 is specific for HIV-1 and was included as a negative control]. Data are expressed as mean cpm ± SEM and are representative of three independent experiments. (B) Antibody inhibition of125I-labeled SIVmac239 gp120 binding to rhesus CCR5 in the presence (solid bars) or absence (open bars) of 100 nM sCD4. (Competition by mAb 26C in the absence of CD4 was not tested.) Binding conditions are as described for (A). Data are means ± SEM and are representative of three independent experiments.

Amino acid sequences in several distinct regions of CCR5 have been shown to be essential for coreceptor function (15, 16). Using alanine scanning mutagenesis of CCR5hu, we recently identified a region that included the tyrosine-rich motif Tyr10-Asp-Ile-Asn-Tyr-Tyr (17). In CCR5rh this sequence is Tyr10-Asp-Ile-Asp-Tyr-Tyr, and we hypothesized that this substitution may correlate with the CD4 independence of SIVmac239 binding to CCR5rh.

To examine this possibility, we mutated CCR5rh from Asp13 to Asn13 as it occurs in CCR5hu (18). In transfected HEK293 cells, CCR5rhD13N bound SIVmac239 gp120 only in the presence of sCD4 (K d = 11.6 nM) (Fig. 2D), which suggested that the single charge difference at residue 13 in the NH2-terminus confers CD4-independent binding to rhesus but not human CCR5. To confirm this finding, we made the corresponding mutation (Asn13 → Asp) in CCR5hu(18). CCR5huN13D conferred gain of function for CD4-independent binding of SIVmac239 gp120 (K d = 7.7 nM with sCD4, K d = 0.85 nM without sCD4) (Fig. 2E). HIV-1(YU2) gp120 binding was CD4-dependent for both mutant receptors (19). The other NH2-terminal mutation between CCR5rh and CCR5hu (Thr9 → Ile) has no role in CD4 independence, as CCR5rhT9I retained CD4-independent binding of SIVmac239 gp120 (K d = 11.4 nM with sCD4, K d = 4.04 nM without sCD4) (19).

We tested seven mAbs derived from an HIV-2–infected human for their ability to inhibit SIVmac239 gp120 binding to CCR5rh in HEK293 cells in the presence and absence of sCD4. These mAbs recognize distinct conformation-dependent epitopes common to HIV-2 and multiple strains of SIV (20, 21). The mAbs B23, 110C, or 17A, which react with SIVmac239 (21), efficiently inhibited CD4-independent binding of SIVmac239 gp120 (Fig.3, A and B); mAb 15D, which has been reported to react with SIVmac251 and weakly with SIVmac239 (20), partially inhibited SIVmac239 gp120 binding (Fig. 3A). The mAbs 23F and 34G, which also do not react with SIVmac239, failed to inhibit SIVmac239 gp120 binding to CCR5rh, and mAb 26C, a non-neutralizing mAb that reacts with SIVmac239, does not block CCR5 binding (Fig. 3A). Inhibition of SIVmac239 gp120 binding by neutralizing (but not non-neutralizing) mAbs suggests that these antibodies may interfere with the gp120-CCR5 interaction. Further, antibody inhibition in the absence of CD4 suggests they do not recognize CD4-induced epitopes.

The observation that some lentiviruses use chemokine receptors but not CD4 for entry raises the possibility that primordial lentiviruses used seven-transmembrane-spanning proteins as a sole receptor, and that the use of CD4 evolved with the primate lentiviruses (22). In this view, SIV represents a transitional virus between CD4-independent lentiviruses and HIV-1, for which CD4 use is obligate. This helps to explain the propensity of the related virus, HIV-2, to evolve virus strains that efficiently infect CD4-negative cells by use of chemokine receptors alone (23).

Infection by most SIV strains is significantly enhanced by the presence of CD4 on the target cells. The discrepancy between CD4-independent gp120 binding and CD4 enhancement of SIV infection may be explained by the different context in which these two processes occur. Entry is mediated by a trimeric gp120-gp41 complex on the viral surface (24), whereas the binding assays use gp120 monomers (25). CD4 binding may serve additional functions for intact virus besides the induction of the chemokine receptor binding site.

Conformation-dependent, CD4-independent neutralizing mAbs from HIV-2–infected individuals efficiently inhibited SIVmac239 binding to CCR5rh (Fig. 3, A and B). Alterations in the SIV gp120 V3 and V4 regions have been shown to disrupt a conformation-dependent epitope that is the target of the most potent anti-SIV neutralizing antibodies (26).

The aspartic acid residue that determines CD4-independent SIVgp120 binding resides within a tyrosine-rich motif identified by mutagenesis as critical for the interaction of CCR5 with the HIV-1 and SIV gp120 glycoproteins (17, 27). This motif is conserved in all chemokine receptors used by primary HIV-1 and SIV isolates. Our results suggest that amino acid residues within this region directly contact the gp120 glycoprotein during CCR5 binding. It is likely that the interaction of CCR5 with a previously cryptic site on gp120 near or within the V3 loop triggers subsequent conformational changes in the envelope glycoprotein, leading to virus–cell membrane fusion. In HIV-1, CD4 binding is required to expose or form the CCR5 binding site. In SIV, gp120 moieties, perhaps within the V3 loop, can interact with Asp13 of CCR5rh and initiate high-affinity gp120-CCR5 binding in the absence of CD4. The identification of this single critical amino acid in CCR5 may have utility in the design of assay systems for therapeutics and vaccine candidates.


View Abstract

Stay Connected to Science

Navigate This Article