Four-domain sCD4 (molecular weight 50 kDa) was expressed in 293F cells after stable transfection. The protein was secreted into the cell culture medium and then purified by using the C-terminal His₆ tag, as previously described [40]. The structure of compound JRC-II-191 (herein referred to as 191) is reported in reference [33]. The CCR5 antagonist ‘Compound A’ [4-nitrobenzyl 1-(3-(N-methylphenylsulfonamido)-3-phenylbutyl)piperidin-4-yl(vinyl)carbamate] [41],[42] was kindly provided by Dr. Martin Springer at Merck Research Laboratories, Rahway, NJ.

The anti-gp120 monoclonal antibody 48d, which recognizes an epitope that overlaps the coreceptor-binding site, was kindly provided by Dr. James Robinson (Tulane University Medical Center) [43]. The monoclonal antibodies IgG1 b12 and 39F recognize the CD4-binding site and the V3 loop of gp120, respectively [44],[45]. The C34-Ig fusion protein consists of the Fc region of human IgG1 linked to the HR2 region of the HXBc2 envelope glycoprotein (amino acid residues 628–661, numbered according to current convention [46]). C34-Ig was produced and purified as previously described [47]. The CD4-Ig fusion protein consists of the first two N-terminal domains of the CD4 molecule and the Fc region of human IgG1. Purification was carried out as described for the C34-Ig molecule [47].

The primary viruses UK7br and UK7br34 were isolated from autopsy brain tissue from an AIDS patient with HIV-1 associated dementia [48]. These envelope glycoproteins were expressed from the pcDNA 3.1 backbone vector. All other envelope glycoproteins described were expressed from the pSVIIIenv vector [49]. The YU2, AD8, JR-FL, 89.6 and KB9 env sequences from Asp 718 (Kpn I) to BamH I were substituted for the corresponding HXBc2 env sequences in the original pSVIIIenv vector. The ΔKS construct, which contains an HIV-1 HXBc2 env gene with a large deletion, was used as a negative control. The YU2(Δct) protein has a truncated cytoplasmic tail of 17 amino acids, with a stop codon introduced after Ala 710 (numbered according to current convention [46]).

The YU2-GS8 construct is a cleavage-defective form of the YU2 HIV-1 envelope glycoproteins that contains an 8-amino acid glycine-serine linker at the gp120/gp41 junction. Starting with the cytoplasmic tail-deleted YU2 envelope glycoproteins, both Arg 508 and Arg 511 near the furin cleavage site were altered to Ser to render the protein cleavage-defective. The 8-amino acid linker, Gly-Gly-Gly-Ser-Gly-Gly-Gly-Ser, was then inserted between Ser 511 (the C-terminal residue of gp120) and Ala 512 (the N-terminal residue of gp41) using the overlapping PCR method.

A sensitive cell-based ELISA with high temporal resolution was developed to measure the binding of antibodies to HIV-1 envelope glycoprotein trimers expressed on cells. COS-1 cells were seeded in 96-well plates (2.4×10⁴ cells per well) and transfected the next day with 0.1 µg of a plasmid expressing the envelope glycoproteins and 0.01 µg of a Tat-expressing plasmid per well using the Effectene transfection reagent (Qiagen). Two days later, cells were washed twice with blocking buffer (35 mg/ml BSA, 10 mg/ml non-fat dry milk, 1.8 mM CaCl₂, 1 mM MgCl₂, 25 mM Tris, pH 7.5 and 140 mM NaCl). For pulse activation experiments, the COS-1 cells were incubated with sCD4 (40 µg/ml, 0.8 µM) or 191 (360 µM) suspended in blocking buffer for three minutes, washed three times with blocking buffer and incubated for different time periods until the C34-Ig or 48d antibodies were added (at 40 µg/ml or 1 µg/ml, respectively, for 30 minutes). To study the temperature dependence of HR1 groove exposure, the sCD4-pulsed cells were incubated at the requisite temperature for different lengths of time; the cells were subsequently returned to room temperature for incubation with C34-Ig. Cells were then washed four times with blocking buffer and four times with washing buffer (140 mM NaCl, 1.8 mM CaCl₂, 1 mM MgCl₂ and 20 mM Tris, pH 7.5). A horseradish peroxidase-conjugated antibody specific for the Fc region of human IgG was then incubated with the samples for 45 minutes at room temperature. Cells were washed 5 times with blocking buffer and 5 times with washing buffer. HRP enzyme activity was determined after the addition of 33 µl per well of a 1∶1 mix of Western Lightning oxidizing and luminol reagents (Perkin Elmer Life Sciences) supplemented with 150 mM NaCl. Light emission was measured with a Mithras LB 940 luminometer (Berthold Technologies). For experiments that measure the rate of decay of C34-Ig or 48d antibody binding, the indicated values for the amount of antibody bound were obtained by subtracting binding measured in the absence of sCD4 from the binding measured at each time point after the sCD4 pulse.

293T cells cultured in 6-well plates were transfected with the pcDNA3.1-CD4 or ΔKS constructs (1 µg per 1.8×10⁶ cells) using the Effectene reagent (Qiagen). COS-1 cells cultured in 96-well plates were transfected with plasmids expressing the HIV-1 envelope glycoproteins as described above. Two days later, the 293T cells were harvested using 5 mM EDTA in PBS and added at 4°C to the monolayer of COS-1 cells (3×10⁵ 293T cells per well). Cells were subsequently centrifuged at 1,000×g for two minutes to increase contact and incubated at 4°C for 75 minutes. Samples were then washed once with blocking buffer and rapidly equilibrated to 25°C. C34-Ig was then added at different times. Detection of C34-Ig binding was performed as described above.

Flow cytometry was used to detect the binding of C34-Ig to 293T cells that express the cytoplasmic tail-deleted (Δct) HIV-1 envelope glycoproteins [47]. Cells were cultured in 6-well plates (1.5×10⁶ cells per well) and transfected with 0.6 µg of a plasmid expressing the HIV-1 envelope glycoproteins and 0.06 µg of a Tat-expressing plasmid using the Effectene transfection reagent (Qiagen). Two days later, cells were detached using PBS supplemented with 5 mM EDTA and resuspendend in FACS buffer (PBS supplemented with 10 mg/ml BSA and 0.5 mg/ml sodium azide). Cells were then incubated with C34-Ig suspended in FACS medium, with or without sCD4, at the indicated temperature for 45 minutes. Following three washes with FACS buffer, cells were incubated at room temperature for 45 minutes in FACS buffer containing a phycoerythrin-conjugated goat anti-human IgG antibody (Sigma) and an anti-CD4 fluorescein-conjugated antibody (OKT4, eBioscience) to measure C34-Ig and sCD4 binding, respectively. Cells were subsequently washed three times with FACS buffer and analyzed with a FACScan analyzer (Becton Dickinson) using Cellquest software for data acquisition and analysis.

Single-round, recombinant HIV-1 viruses that express the luciferase reporter gene were generated by transfection of 293T cells using the calcium phosphate transfection method (Promega). Cells were seeded in 100-mm tissue culture dishes (approximately 4×10⁶ cells per dish) and transfected the next day with 10 µg of the HIV-1 packaging construct pCMVΔP1ΔenvpA [50], 10 µg of the firefly luciferase-expressing construct pHIvec2.luc and 2.5 µg of the plasmid expressing the HIV-1 Rev and envelope glycoproteins. Sixteen hours following transfection, the medium was changed to culture medium (Dulbecco-modified Eagle medium supplemented with 10% fetal calf serum, DMEM/FCS 10%); 12 hours later, the medium was changed to DMEM/FCS 1%. Virus-containing supernatants were collected 12 hours later, cleared of cell debris by low speed centrifugation and filtered through 0.2 µm filters. Preparations were subsequently loaded onto Float-A-Lyzer dialysis cassettes (molecular weight cutoff 100 kDa, Spectrum Labs) and dialyzed against HS buffer (140 mM NaCl, 10 mM HEPES, pH 7.3) for 24 hours at 4°C. Viral preparations were subsequently aliquoted and stored at −80°C until use.

To minimize the time interval between activation of the virus and attachment to the cells, we magnetically controlled the attachment step. As a magnetically controllable carrier, we used magnetite nanoparticles (50 nm in diameter) that are coated by a starch polymer with phosphate end groups (FluidMag PD, Chemicell). Viruses were preincubated with magnetite nanoparticles (1.2 mg/ml) for 7 minutes at room temperature, followed by incubation with sCD4 (40 µg/ml) for different time periods. Preparations were then added to a confluent monolayer of cells cultured in a 96-well plate, to which was applied a magnetic field using a Nd-Fe-B permanent magnet (OZ Biosciences). Canine thymocyte Cf2Th cells that express CCR5 or both CD4 and CCR5 were used as target cells (approximately 7×10⁴ cells per well). After a one-minute incubation at room temperature, cells were washed twice with DMEM/FCS 10% and cultured for an additional 14 hours. Cells were then detached with trypsin-EDTA and seeded again at a 1∶6 dilution. Forty-eight hours after infection, cells were lysed with passive lysis buffer (Promega) and lysates were assayed for luciferase activity. To each well was added 100 µl of luciferin buffer (15 mM MgSO₄, 15 mM KPO₄ [pH 7.8], 1 mM ATP, and 1 mM dithiothreitol) and 50 µl of 1 mM D-luciferin potassium salt (BD Pharmingen). Luminescence was recorded using a Berthold LB 960 microplate luminometer.

The decay rate of virus infectivity after activation by sCD4 was measured using magnetically-immobilized viruses. This method allows rapid pulsing and washing of the virus before addition of target cells; thus, virus infectivity shortly after activation can be measured. Virus preparations were preincubated with magnetite nanoparticles (1.2 mg/ml, suspended in HS buffer) for 7 minutes and then added to 96-well plates placed over a permanent Ne-Fe-B magnet (OZ Biosciences). After a 4-minute incubation at room temperature, wells were washed twice with culture medium. Immobilized viruses were then pulsed with 40 µg/ml (0.8 µM) sCD4 for 3 minutes, washed twice with culture medium and then incubated at room temperature for different time periods. Cf2Th-CCR5 cells suspended in culture medium (∼1×10⁵ cells) were then added to the wells and centrifuged at 1,000×g for 2 minutes to expedite contact between cells and viruses. Luciferase activity was measured two days later.

The affinity of NBD-556 analogues, including 191, for the HIV-1 gp120 envelope glycoprotein is significantly lower than that of sCD4 [31],[33]. Measured by isothermal titration calorimetry, the K_d of 191 binding to the YU2 gp120 envelope glycoprotein is 760 nM, compared with a K_d of approximately 2 nM for the sCD4-gp120 interaction. At achievable concentrations, 191 induced significantly lower levels of HIV-1 activation than sCD4 (see Results section); levels of infectivity were consequently too low to be monitored for decay. We therefore introduced a number of modifications to the method described above for measuring infectivity decay. Briefly, viruses were incubated with 180 µM 191 at 37°C for 10 minutes. Activation was then halted by a 20-fold dilution of the virus-compound mix in HS buffer equilibrated to 26°C. After dilution, no additional activation of HIV-1 infection was induced by 191 during the time frame of the experiment (data not shown). Samples were then incubated for different time periods, associated with magnetite nanoparticles and magnetically adsorbed to confluent cultures of Cf2Th-CCR5 cells, as described above.

Cf2Th cells cultured in 6-well plates (6×10⁵ cells per well) were transfected with different amounts of the pcDNA3.1-CCR5 plasmid, which expresses human CCR5, using the Effectene transfection reagent. Two days later, cells were detached using 5 mM EDTA in PBS and re-seeded in 96-well plates (7×10⁵ cells per well). Viruses were then magnetically adsorbed to the transfected cells in the presence of 20 µg/ml (400 nM) sCD4 in the culture medium and infectivity was measured two days later.

Soluble CD4 mimics (SCMs) can exert varied effects on HIV-1 infection of CD4⁺ and CD4⁻ target cells. Although soluble CD4 (sCD4) typically inhibits HIV-1 infection of CD4⁺CCR5⁺ cells, enhancement of HIV-1 infection of CCR5⁺ cells lacking CD4 is sometimes seen after sCD4 treatment [14]. Sensitivity to these effects varies among different HIV-1 strains [49]. We examined the sensitivity to sCD4 of recombinant HIV-1 viruses (designated HIV-1(AD8) and HIV-1(YU2), respectively) pseudotyped with the envelope glycoproteins of the primary HIV-1 isolates AD8 and YU2, which have similar binding affinities for sCD4 [51],[52]. In CD4⁺CCR5⁺ target cells, sCD4 inhibited the infectivity of HIV-1(AD8) and HIV-1(YU2) with IC₅₀ values of 80 nM (4 µg/ml) and 12 nM (0.6 µg/ml), respectively (Figure 1A). In CD4⁻CCR5⁺ cells, sCD4 activated infection of HIV-1(AD8) much more efficiently than that of HIV-1(YU2).

A similar pattern of HIV-1 activation and inactivation was observed for the NBD-556 analogue, JRC-II-191, herein referred to as 191. The affinity of 191 for the HIV-1_YU2 gp120 envelope glycoprotein (K_d = 760 nM) is significantly lower than that of sCD4 (K_d = ∼2 nM). Nevertheless, at proportionately higher concentrations, 191 resembled sCD4, inhibiting HIV-1 infection of CD4⁺CCR5⁺ cells and enhancing infection of CD4⁻CCR5⁺ cells (Figure 1B). For both sCD4 and 191, the enhancement of infection of cells lacking CD4 became progressively less efficient at higher concentrations; in some cases, decreases in viral infectivity were associated with escalating doses. These characteristics of HIV-1 envelope glycoproteins differ from those observed with several SIV envelope glycoproteins, which demonstrate increasingly efficient infection of CD4⁻CCR5⁺ cells with increasing sCD4 concentrations (data not shown, and ref. [15]). These observations suggested the existence of a concentration-dependent, virus-inhibitory process superimposed on the HIV-1 activation process.

Because virus attachment to cells in culture progresses in a slow, diffusion-dependent manner [53],[54], significant time intervals can elapse between virus binding by SCMs and the association between virus and cell. Thus, the diminished capacity of the SCMs to enhance HIV-1 infection of CD4⁻CCR5⁺ cells at higher concentrations might reflect the involvement of time-dependent as well as concentration-dependent processes that abrogate infectivity after activation occurs. To test this possibility, we examined the effect of neutralizing concentrations of sCD4 on HIV-1(YU2) infection when the time interval between sCD4-virus binding and virus-cell attachment was reduced. For this purpose, we applied a system that magnetically controls virus motion to eliminate viral dependence on passive diffusion for cell attachment; this system allows near-complete transfer of the virus inoculum to the cell-bound state within one minute [53],[55]. When HIV-1(YU2) was briefly incubated with a high concentration of sCD4 and then added to CD4⁻CCR5⁺ cells, infection occurred only when the virus was magnetically adsorbed to the cells (Figure 1C). These results suggested the possibility of a time-dependent process that decreases HIV-1 infectivity after activation by sCD4, and prompted us to measure the stability of the sCD4-activated intermediate form of the HIV-1 envelope glycoproteins.

Engagement of CD4 induces a major structural rearrangement of the HIV-1 envelope glycoproteins [2]. Two functionally important regions of the HIV-1 envelope glycoproteins are formed and exposed as a result of CD4 binding: i) the coreceptor-binding site on the gp120 envelope glycoprotein [3]–[7],[56]; and ii) the trimeric coiled coil composed of the N-terminal heptad repeat (HR1) regions of the three gp41 subunits [8]–[10],[47]. To investigate the conformational stability of the sCD4-activated envelope glycoprotein intermediate, we monitored changes over time in the exposure of the coreceptor-binding site on gp120 and the HR1 coiled coil on gp41 after pulse activation by sCD4.

To measure the progression of conformational changes that immediately follow HIV-1 envelope glycoprotein activation, we developed a novel enzyme-linked immunosorbent assay (ELISA)-based system that utilizes live cells as a platform for expression of membrane-bound trimeric envelope glycoprotein complexes The system allowed sensitive detection of ligand binding to full-length envelope glycoproteins at high temporal resolution. COS-1 cells that express the HIV-1_YU2 envelope glycoproteins were pulse-activated with sCD4 at 37°C for 3 minutes and incubated at 37°C for different time periods. The following probes were then added: i) monoclonal antibody 48d, which binds to a gp120 epitope that overlaps the coreceptor-binding site [43]; or ii) the C34-Ig protein, which is composed of the C34 peptide corresponding to the gp41 HR2 region linked to the Fc constant region of human IgG1 [47]. C34-Ig binds to the hydrophobic groove formed at the interface of the HR1 coiled-coil helices [10],[47]. The sCD4 pulse significantly enhanced the binding of the C34-Ig protein and the 48d antibody to the HIV-1 envelope glycoproteins (Figure 2A and 2B). The time interval between the sCD4 pulse and addition of the 48d antibody did not affect the level of 48d binding observed (Figure 2B). Thus, after the sCD4-induced conformational change, the epitope of the 48d antibody remains stably exposed over the time period examined.

In contrast to the above results, the exposure of the C34-Ig binding site after the sCD4 pulse was transient at 37°C (Figure 2C). Stability of the HR1-exposed state was highly dependent on the temperature of incubation (Figure 3A). Although sCD4 efficiently bound to the envelope glycoprotein complex and induced the formation/exposure of the HR1 groove at 4, 25 and 37°C, this state was stable only at lower temperatures (Figure 3B). The loss of exposure of the HR1 groove was not reversible by repeated pulsing of the envelope glycoprotein complexes with saturating concentrations of sCD4 (data not shown). At the different temperatures examined, in the absence of sCD4, C34-Ig binding was similar to that measured for the negative control cells transfected with the ΔKS plasmid. No correlation was observed between the temperature of incubation and the binding of C34-Ig in the absence of sCD4.

Given the constant level of 48d antibody binding to the sCD4-pulsed envelope glycoprotein-expressing cells, the observed time-dependent decrease in C34-Ig binding cannot be attributed to changes in cell-surface levels of the envelope glycoproteins, as might occur due to gp120 shedding [17]. Moreover, we observed that sCD4 binding to the cells remained constant over the time period examined. We performed a kinetic experiment to measure the decay of HR1 exposure using FACS analysis. The binding of sCD4 and C34-Ig to the envelope glycoprotein-expressing cells was measured simultaneously at different time points after the sCD4 pulse, using phycoerythrin- and fluorescein-conjugated secondary antibodies. We observed a gradually decreasing capacity of C34-Ig to bind despite constant levels of bound sCD4 (Figure S1).

In summary, after engagement of sCD4, the HIV-1 envelope glycoprotein complex undergoes conformational changes that result in formation/exposure of the gp120 coreceptor-binding site and the gp41 HR1 groove. However, despite continued engagement of sCD4 and stable exposure of the 48d epitope on gp120, the induced envelope glycoprotein intermediate undergoes a spontaneous temperature-dependent and apparently irreversible change of conformation.

The decay profile of sCD4-induced HR1 groove exposure was determined for a panel of envelope glycoproteins from primary and laboratory-adapted HIV-1 strains. Measurements were conducted at room temperature (24–26°C) to slow the rate of decay, allowing greater accuracy. Interestingly, a wide range of stabilities of the HR1 groove-exposed state was observed in the different HIV-1 envelope glycoproteins (Figure 4A). For example, the YU2 and AD8 envelope glycoproteins exhibited half-lives of 5 and 100 minutes, respectively. The different decay rates of HR1 groove exposure could not be explained by differences in the affinity of the trimeric envelope glycoprotein complexes for sCD4 (Figure S2A), or in kinetic differences in CD4-Ig binding to these complexes (Figure S2B). We also found that the deletion of the gp41 cytoplasmic tail had no effect on the stability of the CD4-activated intermediate (Figure S2C), even though the cell-surface level of envelope glycoprotein expression was increased, as expected [57].

Envelope glycoproteins from HIV-1_89.6 and HIV-1_KB9 exhibited highly stable sCD4-induced intermediates that did not appreciably decay at room temperature for up to 1 hour after the sCD4 pulse. For these envelope glycoproteins, the half-lives of the HR1-exposed state measured at 37°C were 16 and 17 minutes, respectively (Figure S2D).

Does the sCD4-induced conformational instability in the HIV-1 envelope glycoproteins have consequences for virus infectivity? To examine this question, we devised a strategy to measure changes over time in the capacity of sCD4-activated virus to infect CD4⁻CCR5⁺ cells. First, recombinant, luciferase-expressing viruses were associated with magnetite nanoparticles and immobilized on tissue-culture plates by constant application of a magnetic field. Next, the immobilized viruses were pulsed with sCD4 for 3 minutes and incubated at 25°C for different time periods. Then, CD4⁻CCR5⁺ cells were added and immediately pelleted onto the immobilized viruses. Infectivity was assessed by measuring luciferase activity in the cells 48 hours later. The levels of initial infectivity for viruses pseudotyped with the different HIV-1 envelope glycoproteins were significantly increased by incubation with sCD4 (data not shown). However, infectivity subsequently declined at different rates for viruses with different envelope glycoproteins (Figure 4B). Similar to the stability of HR1 groove exposure, the rate of decay of infectivity was highly dependent upon temperature (Figure S3). At room temperature, the half-life of infectivity exhibited by the sCD4-treated viruses pseudotyped with different HIV-1 envelope glycoproteins correlated well with the half-life of the HR1 groove-exposed state following incubation with sCD4 (R² = 0.9374, Figure 4C).

Assuming that the linear regression model generated using the data obtained at 25°C applies to the situation at 37°C, we first estimated the half-lives of HIV-1 infectivity by fitting the half-lives of HR1 groove exposure measured at 37°C to the model. For the most stable envelope glycoproteins (derived from HIV-1_89.6 and HIV-1_KB9), we estimated a maximal half-life of infectivity of 5–7 minutes at 37°C. In accordance with the predicted values, the measured half-lives of infectivity at 37°C for viruses containing the KB9 and 89.6 envelope glycoproteins were 7 and 8 minutes, respectively, after the sCD4 pulse (Figure S2E).

We compared the infectious half-lives of viruses with different HIV-1 envelope glycoproteins following incubation with sCD4 (using CD4⁻CCR5⁺ cells, as described above) and without sCD4 treatment (using CD4⁺CCR5⁺ cells). Incubation of the viruses at 37°C for increasing time periods in the absence of sCD4, followed by infection of CD4⁺CCR5⁺ cells, revealed a gradual reduction of infectivity, with half-lives ranging from 6.9 to 12.1 hours (Figure S4). However, no correlation between the functional stability of the native and sCD4-bound states of each of the envelope glycoproteins was observed.

We conclude that after engagement of sCD4, HIV-1 undergoes a transient phase of activation, during which attachment to cells that express CCR5 will allow entry to occur. However, this activated state is short-lived and is followed by apparently irreversible structural rearrangements and loss of infectivity. Thus, at any given time point after exposure to sCD4, HIV-1 infectivity represents the combined result of an activation process and progression through the activated intermediate state to inactivation (Figure S5).

CD4-mimetic compounds exemplified by NBD-556 and 191 induce conformational changes in the HIV-1 envelope glycoproteins similar to those induced by sCD4 [31],[33]. These changes include formation/exposure of both the coreceptor-binding site on gp120 and the HR1 groove on gp41 (Figure 5). The similarity between the effects of 191 and sCD4 on HIV-1 infection of CD4⁺ and CD4⁻ cells (Figure 1) suggested that 191 may also induce transiently activated intermediate states in the HIV-1 envelope glycoproteins. To examine this possibility, we compared the decay profiles associated with sCD4 and 191. We examined the YU2 and AD8 envelope glycoproteins, which exhibit dramatic differences in the decay rate of the sCD4-induced intermediate. For each of these envelope glycoproteins, sCD4 and 191 induced identical decay profiles, both for HR1 groove exposure and for infectivity (Figure 6). These results suggest that sCD4 and 191 induce a similar, transient intermediate state in the HIV-1 envelope glycoproteins. Moreover, the differences in decay rates between the AD8 and YU2 envelope glycoproteins were evident for both sCD4 and 191, suggesting that the stability of the induced state is determined by intrinsic properties of the envelope glycoproteins rather than by the activating agent.

To examine the longevity of the activated intermediate induced during the normal HIV-1 entry process, we examined the stability of HR1 groove exposure after the binding of the envelope glycoproteins to cell-surface CD4. For this purpose, 293T cells that express CD4 (but not CCR5) were used to activate the HIV-1 envelope glycoproteins expressed on COS-1 cells. Pilot experiments demonstrated that the observed increases in C34-Ig binding to the envelope glycoproteins-expressing cells depended on the presence of CD4 on the activating cell, and were not observed for a mutant envelope glycoprotein (YU2-GS8), which binds CD4 efficiently but does not expose the HR1 groove in response (see Figure S6 and Materials and Methods). Relative to the effects of treatment with sCD4 or 191, the HR1 groove exposure consequent to activation by cell-surface CD4 was surprisingly long-lived (Figure 7A). The AD8 and YU2 envelope glycoproteins, which exhibited significant differences in the half-lives of HR1 groove exposure after activation by sCD4, both formed stable intermediates following activation by cell-surface CD4.

During HIV-1 entry, CCR5 binding to the CD4-induced envelope glycoprotein intermediate promotes progression along the path to virus entry [58]. Differences in the longevity of the CD4-induced state would be predicted to manifest themselves as an altered dependency on the density of target cell CCR5. To test this, the CCR5 available on CD4-expressing cells for interaction with the HIV-1 envelope glycoproteins was varied in two ways: 1) the addition of increasing amounts of Compound A, a CCR5 inhibitory compound [41],[42],[59], to cells expressing both CD4 and CCR5; and 2) transfection of different amounts of a CCR5-expressing plasmid into cells expressing CD4. The latter method is based on the direct relationship between the amount of CCR5 DNA transfected and the cell-surface density of the expressed CCR5 molecule (Figure S7). The cells were then exposed to HIV-1(AD8) or HIV-1(YU2) and the efficiency of infection measured. HIV-1(AD8) and HIV-1(YU2) behaved indistinguishably in response to variations in the CCR5 available on target cells expressing CD4 (Figure 7B and 7C). In contrast, when sCD4 was used to activate infection of Cf2Th cells expressing different levels of CCR5, entry by HIV-1(AD8) was significantly better than that of HIV-1(YU2) (Figure 7D). These results are consistent with the existence of a relatively long-lived envelope glycoprotein intermediate being induced by CD4 expressed on the target membrane.