ABSTRACT
Human immunodeficiency virus (HIV-1) entry into cells is mediated by the viral envelope glycoprotein (Env) trimer, which consists of three gp120 exterior glycoproteins and three gp41 transmembrane glycoproteins. When gp120 binds sequentially to the receptors CD4 and CCR5 on the target cell, the metastable Env trimer is triggered to undergo entry-related conformational changes. PF-68742 is a small molecule that inhibits the infection of a subset of HIV-1 strains by interfering with an Env function other than receptor binding. Determinants of HIV-1 resistance to PF-68742 map to the disulfide loop and fusion peptide of gp41. Of the four possible PF-68742 stereoisomers, only one, MF275, inhibited the infection of CD4-positive CCR5-positive cells by some HIV-1 strains. MF275 inhibition of these HIV-1 strains occurred after CD4 binding but before the formation of the gp41 six-helix bundle. Unexpectedly, MF275 activated the infection of CD4-negative CCR5-positive cells by several HIV-1 strains resistant to the inhibitory effects of the compound in CD4-positive target cells. In contrast to CD4 complementation by CD4-mimetic compounds, activation of CD4-independent infection by MF275 did not depend upon the availability of the gp120 Phe 43 cavity. Sensitivity to inhibitors indicates that MF275-activated virus entry requires formation/exposure of the gp41 heptad repeat (HR1) as well as CCR5 binding. MF275 apparently activates a virus entry pathway parallel to that triggered by CD4 and CD4-mimetic compounds. Strain-dependent divergence in Env conformational transitions allows different outcomes, inhibition or activation, in response to MF275. Understanding the mechanisms of MF275 activity should assist efforts to optimize its utility.
IMPORTANCE Envelope glycoprotein (Env) spikes on the surface of human immunodeficiency virus (HIV-1) bind target cell receptors, triggering changes in the shape of Env. We studied a small molecule, MF275, that also induced shape changes in Env. The consequences of MF275 interaction with Env depended on the HIV-1 strain, with infection by some viruses inhibited and infection by other viruses enhanced. These studies reveal the strain-dependent diversity of HIV-1 Envs as they undergo shape changes in proceeding down the entry pathway. Appreciation of this diversity will assist attempts to develop broadly active inhibitors of HIV-1 entry.
INTRODUCTION
The entry of human immunodeficiency virus (HIV-1) into target cells is mediated by the trimeric envelope glycoprotein (Env) complex, which consists of three gp120 exterior envelope glycoproteins and three gp41 transmembrane envelope glycoproteins (1). Binding of gp120 to the receptor CD4 on the target cell surface induces major conformational changes in the pretriggered (state 1) Env (2–5). CD4 binding initially induces the default intermediate Env conformation (state 2), and binding of CD4 to additional Env protomers induces the state 3 (full CD4-bound) Env conformation (6–8). CD4-induced conformational changes allow gp120 to bind the viral coreceptor, either CCR5 or CXCR4 (9–16). CD4 binding also induces the formation of a gp41 prehairpin intermediate, in which three hydrophobic grooves on the surface of a coiled coil formed by the heptad repeat 1 (HR1) region of gp41 are exposed (17–20). These hydrophobic grooves are subsequently occupied by helices from the gp41 heptad repeat 2 (HR2) region during the formation of an energetically stable six-helix bundle that drives the fusion of the viral and target cell membranes (21–26). A hydrophobic sequence called the fusion peptide at the gp41 amino terminus is thought to interact with the target cell membrane during this process (27, 28).
Several inhibitors of HIV-1 entry have been discovered. Two inhibitors, maraviroc and enfuvirtide, are approved for the treatment of HIV-1 infection in humans. Maraviroc binds CCR5 and blocks gp120-coreceptor engagement (29); enfuvirtide is a peptide corresponding to the gp41 HR2 region and acts as a dominant negative inhibitor of six-helix bundle formation (30, 31). A number of small-molecule HIV-1 entry inhibitors are under investigation as potential clinical agents or as probes to understand the mechanism of virus entry. CD4-mimetic compounds bind within the Phe 43 cavity of gp120, near the binding site of CD4 (32–45). CD4-mimetic compounds compete with CD4 and also prematurely activate Env, driving it initially into the CD4-bound conformation and then irreversibly into inactive conformations (32–34, 46). Other HIV-1 entry inhibitors, including BMS-806 and compound 484, block CD4-induced conformational changes in Env, preventing the formation and exposure of the HR1 coiled coil (20, 47–51). Analogues of the prototype BMS-806 are currently in clinical trials in HIV-1-infected humans (52).
PF-68742 was identified in a high-throughput screen for small-molecule inhibitors of cell-cell fusion mediated by the HIV-1JR-FL Env (53). PF-68742 exhibited a 50% inhibitory concentration (IC50) of 0.23 μM in this assay. PF-68742 inhibited the infection of several R5 and X4 HIV-1 strains but was inactive against the simian immunodeficiency virus, SIVmac239. PF-68724 did not significantly interfere with the binding of HIV-1 gp120 to CD4 or CCR5 (53). An HIV-1NL4-3 variant that was resistant to PF-68742 was selected in tissue culture; a single-residue change, G514R in the gp41 fusion peptide, was responsible for the resistant phenotype (53). Of interest, low concentrations of PF-68742 enhanced the infection of the G514R virus. Additional changes in the disulfide loop of the gp41 ectodomain were found to determine the resistance to PF-68742 for some HIV-1 variants (53).
Here, we identify the active stereoisomer of PF-68742 and investigate its ability to inhibit HIV-1 infection of CD4-positive target cells. Unexpectedly, we found that the active stereoisomer can enhance HIV-1 infection of CD4-negative target cells expressing the coreceptor. Characterization of the HIV-1 strain-dependent requirements for inhibition and activation by this compound provides insights into the HIV-1 entry process.
RESULTS
Identification of PF-68742 stereoisomers with antiviral activity.Four stereoisomers of PF-68742 exist due to two chiral centers in the molecule (Fig. 1A), but the potential antiviral activity of these stereoisomers was not reported (53). We synthesized and tested the activity of all four PF-68742 stereoisomers against cell-cell fusion mediated by the envelope glycoproteins of HIV-1JR-FL, HIV-1AD8, HIV-1YU2, HIV-2UC1, simian immunodeficiency virus (SIVmac239), and human T-cell leukemia virus (HTLV-I). Of the four stereoisomers, only MF275 efficiently inhibited cell-cell fusion mediated by the HIV-1JR-FL Env, with an IC50 of 9.5 μM (Table 1, Fig. 1B, and data not shown). MF275 did not efficiently inhibit cell-cell fusion mediated by the Envs of HIV-1AD8, HIV-1YU2, HIV-2UC1, SIVmac239, or HTLV-I (Fig. 1B and Table 1). These results indicate that the stereochemistry of PF-68742 is critical for its ability to inhibit the function of some HIV-1 Envs.
Effects of PF-68742 stereoisomers on HIV-1 infection. (A) The four stereoisomers of PF-68742 are shown. (B) The ability of the compounds to inhibit cell-cell fusion mediated by the Envs of the indicated HIV-1 strains is shown. (C) MF275 inhibition of HIV-1 infection of target cells expressing CD4 and CCR5. Recombinant luciferase-expressing HIV-1 containing the indicated Env was incubated at 37°C with various concentrations of MF275 for 30 min. Then, the virus-compound mixtures were added to Cf2Th-CD4/CCR5 cells. After 48 to 72 h of incubation, the cells were lysed and the cell lysates were used for measurement of luciferase activity. The results are reported as the percentages of luciferase activity compared with that seen in the absence of MF275. The means and standard errors from triplicate samples within an experiment are shown. The results of a typical experiment are shown. The experiment was repeated 2 to 9 times with similar results. (D) Time-of-addition experiment. Recombinant luciferase-expressing HIV-1JR-FL was spinoculated onto Cf2Th-CD4/CCR5 cells, and inhibitors were added at different times. After 48 h at 37°C, the cell lysates were used to measure luciferase activity, as described above. (E) MF275 activation of HIV-1 infection of CD4-negative CCR5-expressing target cells is shown. Recombinant luciferase-expressing HIV-1 with the indicated Env was incubated with increasing concentrations of MF275 at 37°C for 30 min. The virus-MF275 mixture was then spinoculated at 1,800 rpm for 30 min at 25°C onto Cf2Th-CCR5 cells, and after 48 h of culture at 37°C, the cells were lysed. Luciferase activity was measured in the cell lysates. The luciferase activity observed in the absence of MF275 was set at 100%, and the level of infection is reported as a percentage of the level seen in the absence of added MF275. Recombinant HIV-1 pseudotyped with AMLV envelope glycoproteins was included as a control virus. The results are reported as the means and standard errors derived from triplicate measurements and are representative of 2 to 6 independent experiments.
Inhibition of Env-mediated cell-cell fusion by PF-68742 stereoisomersa
The four PF-68742 stereoisomers were tested for the ability inhibit the single-round infection of Cf2Th-CD4/CCR5 cells expressing human CD4 and CCR5 by recombinant luciferase-expressing HIV-1 containing different Envs (Fig. 1C and Table 2). Only MF275 inhibited infection by HIV-1JR-FL, consistent with the results of the cell-cell fusion assays described above. MF275 also efficiently inhibited the infection of Cf2Th-CD4/CCR5 cells by HIV-189.6 and HIV-1KB9. The infection of Cf2Th-CD4/CCR5 cells by HIV-1AD8, HIV-1YU2, and several other HIV-1 strains was less sensitive to inhibition by MF275. Infection of Cf2Th-CD4/CCR5 cells by recombinant HIV-1 pseudotyped with the amphotropic murine leukemia virus (A-MLV) Env was not inhibited by MF275. Thus, one PF-68742 stereoisomer, MF275, specifically inhibits infection and cell-cell fusion of CD4-positive CCR5-positive target cells mediated by some HIV-1 Envs.
Inhibition of virus infection by PF-68742 stereoisomersa
We also tested the ability of MF275 and MF276 to inhibit the infection of Cf2Th-CD4/CXCR4 cells expressing CD4 and CXCR4 by R5X4 and X4 HIV-1. MF275, but not MF276, efficiently inhibited the infection of these cells by HIV-1HXBc2 and HIV-189.6 but not HIV-1KB9 (Table 2). In this assay, low concentrations of MF275 stimulated HIV-1KB9 infection, whereas weak inhibition was seen at higher MF275 concentrations. Thus, MF275 can inhibit the infection of cells expressing CD4 and CXCR4 by some strains of HIV-1.
The toxicity of MF275 was evaluated with a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Cf2Th-CD4/CCR5 cells tolerated a 48-h exposure to MF275 well, with a 50% toxic concentration (TC50) of 460 μM (data not shown).
To gain insight into the mechanism of MF275 inhibition of HIV-1JR-FL infection, we performed time-of-addition experiments. After incubation of Cf2Th-CD4/CCR5 cells with HIV-1JR-FL, different entry inhibitors were added at various times and the level of infection was assessed (Fig. 1D). The escape of HIV-1JR-FL from MF275 reproducibly occurred after VRC01 and BNM-III-170 escape and before enfuvirtide escape. The VRC01 antibody and the CD4-mimetic compound, BNM-III-170, both block CD4 binding, whereas enfuvirtide (T20) blocks the formation of the gp41 six-helix bundle. These results indicate that MF275 inhibits HIV-1JR-FL infection after initial CD4 engagement and before six-helix bundle formation.
Activation of CD4-independent HIV-1 infection by MF275.We tested the effect of MF275 on the infection of CD4-negative CCR5-positive Cf2Th-CCR5 cells by HIV-1 variants. Unexpectedly, incubation of HIV-1YU2 and HIV-1AD8 with MF275 resulted in a dramatic increase in the efficiency of infection of these CD4-negative cells (Fig. 1E). The MF275-mediated stimulation of HIV-1YU2 and HIV-1AD8 infection of CD4-negative target cells required CCR5, as no infection was detected in CD4-negative CCR5-negative cells (data not shown). In contrast, little or no activation of infection of the CD4-negative Cf2Th-CCR5 cells by MF275 was observed for HIV-1JR-FL, HIV-189.6, and HIV-1KB9 or for the A-MLV control. Thus, MF275 can activate the CD4-independent infection of some HIV-1 strains.
Stability of the MF275-inhibited and MF275-activated states.The activated state of HIV-1 Env induced by soluble CD4 or the CD4-mimetic compounds is short lived, with a half-life of 5 to 7 min at 37°C, followed by an apparently irreversible inactivation (46). We investigated the stability of the inhibited and activated states of Env induced by MF275. HIV-1AD8, HIV-1YU2, and HIV-1JR-FL were incubated with a dimethyl sulfoxide (DMSO) control, MF275, or BNM-III-170 (a CD4-mimetic compound) (41) at 37°C for 30 min. The viruses were then pelleted by centrifugation. After removing the supernatants, the virus pellet was washed once with medium and then resuspended in medium. The virus suspension was added to Cf2Th-CD4/CCR5 or Cf2Th-CCR5 target cells, and 48 h later, the level of infection was measured.
Washout abrogated MF275 inhibition of HIV-1JR-FL infection of Cf2Th-CD4/CCR5 cells (Fig. 2A). In contrast, washout had no significant effect on inhibition of HIV-1JR-FL infection by the CD4-mimetic compound, BNM-III-170 (Fig. 2B). These results demonstrate that, unlike HIV-1 inhibition by CD4-mimetic compounds (46, 54), MF275 inhibition is reversible.
Stability of the MF275-inhibited and MF275-activated states. Recombinant luciferase-expressing HIV-1 with the indicated Env was incubated with a DMSO control, MF275, or BNM-III-170 (a CD4-mimetic compound) at 37°C for 30 min. Half of the virus suspension was then pelleted and washed, before resuspension in medium. The washed (+) and unwashed (−) viruses were incubated with Cf2Th-CD4/CCR5 (A and B) or Cf2Th-CCR5 (C and D) target cells. Forty-eight to seventy-two hours later, the luciferase activity in the cells was measured. The level of infection is reported as a percentage of the level seen in the absence of added compound. The results of a typical experiment are reported as the means and standard errors derived from triplicate measurements. The experiments were repeated with similar results.
The effect of washout on MF275- or BNM-III-170-activated infection of Cf2Th-CCR5 target cells differed in a strain-dependent manner. For example, activation of HIV-1AD8 infection by MF275 was minimally affected by the washout, whereas that of HIV-1YU2 was significantly reduced (Fig. 2C and D). Of interest, after washout, activation of HIV-1JR-FL, HIV-1AD8, and HIV-1YU2 infection by BNM-III-170 occurred at levels 83%, 11%, and less than 1% of that seen without washout, respectively (data not shown). Strain-dependent Env variation influences the ability of the MF275- and BNM-III-170-activated conformations to withstand washing.
MF275 inhibition and activation do not require the gp120 Phe 43 cavity.Like MF275, small-molecule CD4-mimetic compounds can inhibit HIV-1 infection of cells expressing CD4 and CCR5 and can also activate HIV-1 infection of CD4-negative CCR5-positive target cells (33, 34, 54). These activities depend upon the insertion of the CD4-mimetic compounds into the gp120 Phe 43 cavity, which is proximal to the binding site of CD4 (35–45). To evaluate whether MF275 might similarly interact with the Phe 43 cavity, we tested the effect of the S375W substitution, which fills the Phe 43 cavity (55), on the inhibition and activation of HIV-1 infection by MF275. In Cf2Th-CD4/CCR5 target cells, the HIV-1JR-FL S375W mutant was inhibited by MF275 as efficiently as the wild-type HIV-1JR-FL (Fig. 3A). Both wild-type HIV-1YU2 and the corresponding S375W mutant HIV-1YU2 were resistant to MF275 inhibition in these cells (Fig. 3B). The wild-type HIV-1AD8 and HIV-1AD8 S375W viruses exhibited an intermediate level of sensitivity to MF275 inhibition in Cf2Th-CD4/CCR5 cells (Fig. 3C). Infection of the CD4-negative CCR5-expressing Cf2Th-CCR5 cells by the HIV-1JR-FL S375W mutant, like that of the wild-type HIV-1JR-FL, was minimally activated by MF275 (Fig. 3D). In these CD4-negative CCR5-positive cells, HIV-1YU2 S375W and HIV-1AD8 S375W infections were activated to levels within 2-fold of those seen for wild-type HIV-1YU2 and HIV-1AD8, respectively (Fig. 3E and F). These results indicate that neither inhibition nor activation of HIV-1 infection by MF275 depends upon the presence of the gp120 Phe 43 cavity. In this respect, MF275 differs from the CD4-mimetic compounds.
MF275 inhibition and activation do not require the gp120 Phe 43 cavity. The wild-type (WT) or S375W mutant HIV-1JR-FL (A and D), HIV-1YU2 (B and E), or HIV-1AD8 (C and F) viruses were incubated with the indicated concentrations of MF275, and the virus-compound mixtures were added to Cf2Th-CD4/CCR5 cells (A to C) or Cf2Th-CCR5 cells (D to F). After 48 to 72 h at 37°C, the cells were lysed and the cell lysates were assayed for luciferase activity. The results represent the means and standard errors from triplicate measurements and are reported as the levels of luciferase activity relative to that seen in the absence of MF275. The results of a typical experiment are shown and are representative of those obtained in 4 to 6 independent experiments.
Effects of alterations of Env conformation on HIV-1 susceptibility to MF275.Previous studies indicated that changes in particular regions of the HIV-1 Env result in an increased or decreased sampling of Env conformations downstream of the pretriggered (state 1) conformation on the virus entry pathway (7, 51, 56, 57). We tested HIV-1 gp120 inner domain (H66A) and gp41 ectodomain (A582T, L587A) mutants that decrease state 1-to-state 2 transitions (56) for susceptibility to inhibition and activation by MF275. These HIV-1JR-FL mutants exhibited sensitivity to MF275 inhibition in Cf2Th-CD4/CCR5 cells comparable to that of the wild-type HIV-1JR-FL (Fig. 4A). In Cf2Th-CCR5 target cells, activation of HIV-1YU2 infection by MF275 was significantly reduced for these mutants relative to that of the wild-type HIV-1YU2 (Fig. 4B). The observed reduction in activation by MF275 was not related to the basal level of virus infection in Cf2Th-CD4/CCR5 cells. These results are consistent with a model in which MF275 activation involves an induction of Env conformational transitions from state 1 to downstream conformations.
Effects of alterations of Env conformation on HIV-1 susceptibility to MF275. The HIV-1JR-FL variants (A, C, and D), HIV-1YU2 variants (B), or HIV-1AD8 variants (E and F) were incubated with the indicated concentrations of MF275, and the virus-compound mixtures were added to Cf2Th-CD4/CCR5 cells (A, C, D, and E) or Cf2Th-CCR5 cells (B and F). After 48 to 72 h at 37°C, the cells were lysed and the cell lysates were assayed for luciferase activity. The results represent the means and standard errors from triplicate measurements and are reported as the levels of luciferase activity relative to that seen in the absence of MF275. The results of typical experiments are shown; the experiments were repeated with similar results. WT, wild type; ΔCT, cytoplasmic tail deleted.
HIV-1 Envs with changes in the gp120 V1/V2 region and the β20-β21 region sample an intermediate conformation (state 2) between state 1 and the full CD4-bound (state 3) conformation (6–8, 51). We tested several of these HIV-1JR-FL mutants for sensitivity to MF275 (Fig. 4C). These mutants were inhibited by MF275 almost as efficiently as the wild-type HIV-1JR-FL. In contrast, these mutants exhibited various degrees of resistance to BMS-806, a blocker of Env conformational change that prefers state 1 (6, 20, 50, 51, 56) (data not shown). These results indicate that MF275 is relatively insensitive to changes in Env that result in greater sampling of state 2/3 conformations.
We compared the MF275 sensitivity of an HIV-1JR-FL Env with a complete deletion of the gp41 cytoplasmic tail to that of the wild-type HIV-1JR-FL. The infection of Cf2Th-CD4/CCR5 cells by both viruses was inhibited by MF275, as previously reported for PF-68742 (53) (Fig. 4D).
We also tested a panel of HIV-1AD8 mutants that exhibit various degrees of CD4 independence (58). Infection of Cf2Th-CD4/CCR5 cells by these mutants was minimally inhibited by MF275, as was observed for wild-type HIV-1AD8 (Fig. 4E). We also examined the infection of CD4-negative Cf2Th-CCR5 cells by this panel of mutants (Fig. 4F). Several of the mutants infected the Cf2Th-CCR5 cells efficiently, as expected. Enhancement of infection of the Cf2Th-CCR5 cells by MF275 was observed for some of the mutants but was not related to the basal level of CD4-independent infection of these cells. The sensitivity of this panel of HIV-1AD8 Env mutants to MF275 appears to be unrelated to the intrinsic ability of the Envs to undergo conformational changes typically induced by CD4 binding (58). Together, the results in Fig. 4 indicate that, depending on the particular HIV-1 strain, MF275 can interact with Envs in state 1 and state 2/3 conformations.
Effect of other HIV-1 entry inhibitors on the sensitivity of HIV-1 to MF275.MF275 and CD4-mimetic compounds both inhibit and activate HIV-1 infection, depending on circumstances (33, 34, 54). The effect of subinhibitory doses of MF275 and a CD4-mimetic compound, BNM-III-170, on the inhibition of infection of Cf2Th-CCR5 cells by the other compound was assessed. The addition of a subinhibitory dose of MF275 before the addition of BNM-III-170 resulted in slightly less efficient inhibition by BNM-III-170 (Fig. 5A). This observation suggests that, under these circumstances, there exists a very weak antagonism between MF275 and BNM-III-170. A subinhibitory dose of BNM-III-170 did not significantly affect MF275 inhibition of HIV-1JR-FL but primed a mild activation of HIV-1AD8 and HIV-1YU2 by MF275 in the Cf2Th-CD4/CCR5 cells (Fig. 5B). We observed no evidence of HIV-1-inhibitory synergy between MF275 and BNM-III-170.
Effects of other HIV-1 Env ligands on HIV-1 sensitivity to MF275. (A to G) The effects of the doses of a single compound (designated on the x axis) on the infection of recombinant luciferase-expressing HIV-1 with the indicated Env are shown (filled symbols and solid lines). The effects of incubating the recombinant viruses with a second agent are shown in open symbols and broken lines. For this purpose, recombinant luciferase-expressing viruses with the indicated Envs were incubated for 30 min at 37°C with the compound to the left of the arrow, either MF275 (30 μM) (A) or the CD4-mimetic compound BNM-III-170 (30 μM) (B and C). (D to G) The HIV-1 vectors were incubated for 30 min at 37°C with the indicated concentrations of MF275. Following incubation with the initial compound, the HIV-1 vectors were incubated with the compounds to the right of the arrows (10 μM BNM-III-170, 30 nM BMS-378806, maraviroc, and enfuvirtide). for 30 min at 37°C. The virus-compound mixtures were then added to Cf2Th-CD4/CCR5 or Cf2Th-CCR5 cells, and 48 to 72 h later, luciferase activity was measured in the cell lysates. The level of viral infection is reported as a percentage of the level seen in the absence of added compound. The results are reported as the means and standard errors from triplicate measurements and are representative of those obtained in two independent experiments.
To investigate the mechanism of MF275 activation of HIV-1 infection of Cf2Th-CCR5 target cells, these infections were carried out in the presence of HIV-1 entry inhibitors (Fig. 5C to G). The CD4-mimetic compound, BNM-III-170, also stimulated the infection of Cf2Th-CCR5 cells by HIV-1YU2, HIV-1AD8, and HIV-1JR-FL, as previously observed (54). The amount of activation of HIV-1YU2 and HIV-1AD8 by MF275 was decreased by prior incubation of the viruses with BNM-III-170 (Fig. 5C). This effect was not seen for HIV-1YU2 when the BNM-III-170 was added 30 min after incubation of the viruses with MF275 (Fig. 5D). These results support a model in which CD4-mimetic compounds and MF275 activate Env through parallel pathways.
We also tested the effect of three other HIV-1 entry inhibitors on MF275-induced activation of infection of Cf2Th-CCR5 cells by HIV-1YU2 and HIV-1AD8. BMS-806 blocks CD4-induced conformational changes in Env (20, 47–49), maraviroc blocks gp120-CCR5 binding (29), and T20 (enfuvirtide) binds the gp41 HR1 coiled coil and blocks six-helix bundle formation (30, 31). All three inhibitors decreased the level of HIV-1YU2 and HIV-1AD8 infection of Cf2Th-CCR5 cells (Fig. 5E to G). These results suggest that MF275 activation of HIV-1 infection of CD4-negative CCR5-positive cells depends upon Env conformational transitions from state 1, coreceptor binding, and additional Env conformational changes leading to the formation/exposure of the gp41 HR1 coiled coil and the gp41 six-helix bundle.
Sensitization of some HIV-1 strains to antibody neutralization.The above observations suggest that MF275 binds the Env trimer of some HIV-1 strains in its pretriggered (state 1) conformation and induces changes that promote virus entry into CD4-negative cells expressing CCR5. The binding of CD4-mimetic compounds also produces similar activating events in Env (33, 34, 54) and further sensitizes primary HIV-1 to inhibition by antibodies that are otherwise ineffective at neutralizing the virus (59–63). To test whether MF275 might also sensitize HIV-1 to antibody neutralization, HIV-1AD8, HIV-1YU2, and HIV-1JR-FL were incubated with mixtures of MF275 and various antibodies directed against HIV-1 Env. The virus-compound-antibody mixtures were then used to infect Cf2Th-CD4/CCR5 cells. HIV-1YU2 was inhibited equivalently by the antibodies tested in the absence and presence of MF275 (data not shown). In contrast, MF275 increased the sensitivity of HIV-1JR-FL and HIV-1AD8 to neutralization by the 19b anti-V3 antibody and the 4E10 anti-gp41 antibody (Fig. 6). In addition, HIV-1JR-FL was sensitized by MF275 to the 17b antibody against a CD4-induced gp120 epitope. The CD4-mimetic compound, BNM-III-170, also sensitized HIV-1AD8 and HIV-1JR-FL to neutralization by the 17b, 19b, and 4E10 antibodies. These results indicate that MF275 induces conformational changes in the HIV-1JR-FL and HIV-1AD8 Envs similar to those induced by a CD4-mimetic compound. The induced changes result in Envs that exhibit increased sensitivity to antibodies that recognize state 2/3 Env conformations.
MF275 sensitization of HIV-1 to neutralization by antibodies. HIV-1JR-FL and HIV-1AD8 were incubated with 30 μM MF275 (left) or with 30 μM BNM-III-170 (right) and with the indicated concentrations of antibodies. The virus-compound-antibody mixtures were then added to Cf2Th-CD4/CCR5 cells. After 48 to 72 h at 37°C, the cells were lysed and the cell lysates were assayed for luciferase activity. The mean and standard errors were derived from triplicate samples, and the results are reported as the levels of luciferase activity relative to that seen in the absence of the antibody.
Effect of HIV-1 Env changes on MF275 inhibition and activation of infection.Changes in several residues in the gp120 C terminus (T499A, K500A), the gp41 N-terminal fusion peptide (T529A), and gp41 disulfide-loop region (T605Y, L619Q) have been reported to affect PF-68742 activity (53). We evaluated the sensitivity of these mutants to MF275. The K500A mutants of HIV-1JR-FL, HIV-1AD8, and HIV-1YU2 exhibited phenotypes similar to those of the respective wild-type viruses. The T499A, T529A, T605Y, and L619Q mutants of HIV-1JR-FL were less sensitive to inhibition by MF275 (Fig. 7). These HIV-1JR-FL mutants exhibited near-wild-type sensitivity to BNM-III-170, maraviroc, and enfuvirtide (data not shown). MF275 activation of infection of Cf2Th-CCR5 cells by HIV-1YU2 and HIV-1AD8 was less efficient for the T529A, T605Y, and T499 mutants relative to those of the respective wild-type viruses. These results are consistent with a previous report (53) implicating these residues in MF275 activity.
Effect of HIV-1 Env changes on MF275 inhibition and activation of infection. The HIV-1JR-FL variants, HIV-1YU2 variants, or HIV-1AD8 variants were incubated with the indicated concentrations of MF275, and the virus-compound mixtures were added to Cf2Th-CD4-CCR5 cells or Cf2Th-CCR5 cells. After 48 to 72 h at 37°C, the cells were lysed and the cell lysates were assayed for luciferase activity. The results represent the means and standard errors from triplicate measurements and are reported as the levels of luciferase activity relative to that seen in the absence of MF275. The results of typical experiments are shown; the experiments were repeated with similar results.
To gain insight into a potential binding site of MF275, we examined the location of resistance-associated amino acid residues on the structure of a ligand-free soluble gp140 SOSIP.664 Env trimer (PDB 4ZMJ) (64). The soluble gp140 SOSIP.664 trimers have been shown to assume a state 2-like conformation that differs from the pretriggered state-1 conformation of native membrane Env (65–67). Nonetheless, as MF275 can apparently interact with several different Env conformational states (see above), the structure of soluble gp140 (sgp140) SOSIP.664 Env might be informative. The MF275-resistance-associated residues are localized in the stalk of the Env trimer, near the viral membrane (Fig. 8A). These residues surround a shallow pocket on the Env surface that could potentially accommodate MF275. Two of the top-scoring docking poses for MF275, after energy minimization, are shown in Fig. 8. The putative MF275-binding pocket on the Env surface is well conserved among HIV-1 strains; however, some amino acid residues on the rim of this pocket are very variable in HIV-1 strains. This potentially accounts for the HIV-1 strain-dependent effects observed for MF275. The putative MF275-binding pocket is conserved in the structure of a sgp140 SOSIP.664 Env trimer bound to soluble CD4 and the Fab of a CD4-induced antibody, 17b (PDB 5VN3) (68) (data not shown). Thus, this potential binding site meets the requirement of being present on state 2- and state 3-like conformations of Env. Additional studies are needed to determine whether a comparable binding site exists in the state 1 Env conformation.
Modeling MF275 binding to HIV-1 Env. The two highest-scoring docked poses of MF275 with the ligand-free HIV-1BG505 sgp140 SOSIP.664 Env trimer (PDB 4ZMJ) (64) are shown. (A and B) A Cα worm of the sgp140 SOSIP.664 trimer is shown, with the gp41 sequences of one protomer colored red. The Env trimer apex is at the top of the figure, and the more membrane-proximal stalk at the bottom of the figure. The four Env residues implicated in resistance to MF275 are shown in CPK representation. In panel B, the Env region surrounding the proposed MF275 binding site is shown in detail. (C) A surface representation of the putative MF275 binding site is shown, from a perspective similar to that in panel B. The surface is colored according to the degree of variation found in primate immunodeficiency viruses (red, conserved in all primate immunodeficiency viruses; magenta, amino acid character conserved in all primate immunodeficiency viruses; ochre, conserved in all HIV-1 group M strains; green, variable in HIV-1 strains). For the most variable residues, only the residue number is shown.
Effect of MF275 on gp120 association with the Env trimer.As the potential binding site of MF275 involves both gp120 and gp41, we asked whether MF275 could influence the noncovalent association of gp120 with the Env trimer. Solubilized Env trimers with a carboxy-terminal His6 tag were precipitated with Ni-nitrilotriacetic acid (NTA) beads, in the presence or absence of MF275. We also treated the Env trimers with soluble CD4 in the absence or presence of MF275. In this assay, MF275 stabilized gp120 association with the detergent-solubilized Env trimer for both HIV-1AD8 and HIV-1JR-FL Envs (Fig. 9). The addition of soluble CD4 to the Env trimers resulted in a decrease in the association of gp120 with the Env trimer. MF275 decreased this soluble CD4-induced shedding of gp120. Thus, both in the absence and presence of soluble CD4, MF275 stabilizes the interaction of gp120 and the solubilized Env trimer.
Effect of MF275 on gp120 association with the Env trimer. HOS cells expressing HIV-1AD8 and HIV-1JR-FL Envs with His6 tags at the carboxyl terminus were lysed, and the cell lysates were incubated with the indicated reagents. The input lysates are shown on the left. The lysates were precipitated with Ni-NTA agarose beads, and the precipitates were Western blotted for gp120 and gp41. M, molecular weight markers.
DISCUSSION
Only one of the four possible stereoisomers of PF-68742, herein designated MF275, was found to exhibit antiviral activity against some HIV-1 strains. This suggests that MF275 fits precisely into its target on the HIV-1 Env.
Potent inhibition of virus infection by MF275 was limited to only a few HIV-1 strains, such as HIV-1JR-FL. Time-of-addition experiments suggested that HIV-1JR-FL inhibition by MF275 involves events following initial receptor binding but before six-helix bundle formation. Little or no synergy was observed between MF275 and other entry inhibitors. Little or no inhibitory activity of MF275 survived washing, indicating that inhibition is reversible.
We were surprised to find that infection of CD4-negative CCR5-positive cells by several HIV-1 primary strains was activated by MF275. For these strains, binding of MF275 to the Env trimer can apparently replace the functional requirement for CD4 engagement to trigger virus entry. This property of MF275 is reminiscent of the activity of the CD4-mimetic compounds, which also activate HIV-1 infection of CD4-negative cells expressing coreceptors (33, 34). There are, however, significant differences between HIV-1 Env activation by MF275 and by the CD4-mimetic compounds:
1. CD4-mimetic compounds can potentially activate CD4-independent infection of all HIV-1 strains except the AE recombinants, which have a histidine residue occupying the Phe 43 cavity of gp120. A smaller subset of HIV-1 strains is potentially subject to MF275 activation.
2. CD4-mimetic compounds bind Env in the gp120 Phe 43 cavity (35, 41), whereas the availability of this cavity has only a limited effect on MF275 activation and inhibition.
3. Changes in the gp120 inner domain, V1/V2 region, and β20-β21 region and the gp41 ectodomain can significantly affect Env triggerability or “reactivity,” defined as the propensity of the state 1 Env to undergo conformational changes (7, 51, 56–58). These Env changes dramatically affect HIV-1 sensitivity to CD4-mimetic compounds (7, 51, 54, 56). Although these changes influenced the efficiency with which some HIV-1 strains were activated for CD4-independent infection by MF275, they exerted only minimal impact on virus sensitivity to MF275 inhibition.
4. As the activated state of Env induced by CD4-mimetic compounds is short lived and proceeds to an irreversibly inactivated state, virus inhibition by CD4-mimetic compounds is intimately associated with their ability to induce state 1-to-state 2/3 transitions (46, 54). In contrast, virus inhibition by MF275 apparently acts downstream of state 1-to-state 2/3 Env transitions and is reversible.
MF275 activates CD4-independent infection, sensitizes virus to antibody neutralization, and inhibits virus infection in an HIV-1 strain-dependent manner. These diverse activities suggest that the interaction of MF275 with Env is not limited to a single conformational state. Rather, MF275 must be able to interact with the state 1 conformation, as well as some downstream conformations, of the Env from certain HIV-1 strains. Single-molecule fluorescence resonance energy transfer (smFRET) studies suggest that HIV-1 Envs may spontaneously sample multiple conformations and that Env variation can influence the propensity of primary HIV-1 Envs to make transitions from their dominant state 1 conformation (6–8, 51, 56). In the case of HIV-1YU2, Env changes that limit state 1-to-state 2/3 transitions decreased the susceptibility of the virus to MF275 activation. However, viruses with Envs modified to increase sampling of states 2 and 3 exhibited no consistent changes in MF275 susceptibility. These observations imply that MF275 can interact with Envs in state 1 and downstream conformations.
Our results reveal significant diversity in the responses of HIV-1 strains to MF275, even among relatively closely related HIV-1 strains. For example, three clade B HIV-1 strains (HIV-1JR-FL, HIV-1AD8, and HIV-1YU2) responded differently to MF275. HIV-1JR-FL and HIV-1AD8 were sensitized by MF275 to neutralization by otherwise poorly neutralizing antibodies, and infection of CD4-negative CCR5-expressing cells by HIV-1AD8 and HIV-1YU2 was activated by MF275. The MF275-activated states of HIV-1AD8 and HIV-1YU2 differed in sensitivity to washout, suggesting differences in the stability of the activated states. These observations suggest that the pretriggered state 1 Env conformation of these HIV-1 strains must bind MF275, predisposing Env to conformational changes that differ in a strain-dependent manner (Fig. 10). The infection of CD4-negative CCR5-expressing cells by HIV-1JR-FL is not substantially activated by MF275, possibly because MF275 also acts downstream of receptor binding to inhibit HIV-1JR-FL infection. These downstream events in the HIV-1 entry pathway must also exhibit some strain-dependent diversity to account for the different inhibitory effects of MF275 on these viruses (Fig. 10). HIV-1 sensitivity to gp41 peptides such as enfuvirtide can be influenced by the rate at which Env undergoes conformational transitions (69–77). Thus, while HIV-1 Envs may be conformationally constrained at certain points in the entry pathway (for example, the full CD4-bound [state 3] conformation), considerable strain-dependent variation may be tolerated at other points on the pathway. Such variation will be important to understand when targeting these downstream elements on the entry pathway with small molecules and antibodies.
A model of the effects of MF275 on HIV-1 infection. An energy landscape of the HIV-1 Env is depicted, highlighting the local energy wells associated with the unliganded Env conformation (state 1) and the CD4-bound prehairpin intermediate (state 3). The effects of MF275 on activation, sensitization to antibody neutralization, and inhibition of infection of different HIV-1 variants are depicted. The effect of MF275 on each of these processes depends upon the HIV-1 strain from which the Env is derived.
Additional studies are needed to elucidate the interaction of MF275 with Env, particularly with the pretriggered state 1 Env. Nonetheless, Env residues implicated in escape from MF275 inhibition or activation cluster in unliganded (state 2-like) and CD4/17b-bound sgp140 SOSIP.664 structures (64, 68) near a viral membrane-proximal region. Within this region, interactions between gp120 and gp41 occur and are presumably important for the noncovalent association of these Env subunits as well as for receptor-induced triggering of conformational changes in the Env trimer. Consistent with this notion, MF275 stabilized gp120 association with solubilized Env trimers, countering the destabilizing effect of soluble CD4 binding. This activity may contribute to the inhibitory and/or enhancing effects of MF275 on the function of Envs from different HIV-1 strains. Indeed, surrounding the conserved pocket that we propose as a possible binding site for MF275 are Env residues that exhibit considerable variation among HIV-1 strains. A fuller understanding of the function of this region and how viral variation influences the consequences of MF275 binding should assist the development of broader and more potent entry inhibitors targeting this site.
MATERIALS AND METHODS
Compounds.The synthesis of the four stereoisomers of PF-68742 is summarized in Fig. 11 and was achieved beginning with the four commercially available enantiopure N-Boc-4-amino-proline methyl esters. Sulfonamide formation was achieved under basic conditions using 3-fluorobenzenesulfonyl chloride. Further functionalization of the sulfonamide was then achieved using K2CO3 and phenoxy ethyl bromide to provide the disubstituted sulfonamides without interfering with the stereochemical integrity of the molecules. Subsequent removal of the Boc protecting group followed by amide formation utilizing chloroacetyl chloride provided the respective intermediates in good yield. Finally, treatment of the chloroacetamides with 1-(3-aminopropyl)-2-pyrrolidone under basic conditions provided the four desired stereoisomers (MF275, MF276, MF2238, and MF2239) (see Fig. 1A).
Synthesis of PF-68742 stereoisomers. The four stereoisomers of PF-68742 (MF275, MF276, MF2238, and MF2239) were synthesized according to the scheme summarized in the figure. Additional details are available in reference 89.
The compounds were dissolved in dimethyl sulfoxide (DMSO) at a stock concentration of 10 mM, aliquoted, and stored at −20°C. Each compound was then diluted in Dulbecco’s modified Eagle medium (DMEM; Invitrogen) and used for different assays.
Cell lines.293T human embryonic kidney (HEK) cells, COS-1 African green monkey kidney fibroblasts, HOS human osteosarcoma cells, and Cf2Th canine thymocytes (American Type Culture Collection) were propagated at 37°C and 5% CO2 in DMEM with 10% fetal bovine serum (FBS; Sigma) and 100 μg/ml of penicillin-streptomycin (Mediatech, Inc.). Cf2Th cells stably expressing human CD4-CCR5, CCR5, or CD4-CXCR4 were grown in medium supplemented with 200 μg/ml hygromycin (Roche Diagnostics) and 400 μg/ml G418 (Invitrogen).
Measurement of MF275 toxicity.The effect of a 48-h incubation with MF275 on the viability of Cf2Th-CD4/CCR5 cells was evaluated using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay (78). Approximately 1.5 × 104 cells were plated in each well of a 96-well plate. The DMSO control or various concentrations of MF275 were added to duplicate wells. Additional control wells were prepared without cells. The plates were incubated for 48 h at 37°C in 5% CO2, after which, 1/5 volume of the assay solution (Abcam) was added to each well. The plates were incubated for an additional 3 h at 37°C and 5% CO2. The absorbances at 570 nm and 605 nm were measured. R represents the ratio of the absorbance at 570 nm to that at 605 nm. The cell viability was calculated by (Rsample – Rno cells)/(RDMSO – Rno cells) × 100.
Envelope glycoprotein constructs.All Env amino acid residues are numbered by alignment with the prototypic HXBc2 sequence, according to current convention (79). All primers for mutagenesis were designed using the online Integrated DNA Technologies OligoAnalyzer tool. These mutations were introduced by site-directed mutagenesis PCR using PFU Ultra II polymerase (Agilent Technologies) according to the manufacturer’s protocol.
Antibodies.Sensitization of HIV-1 by MF275 to neutralization by the following antibodies was tested: VRC01, PG9, CH58, 4E10, 35O22, VRC34, 2G12, PGT151, 19b, and 17b. The 17b (80) and 19b (81) antibodies were produced by transfection of 293F cells with plasmids expressing the antibody heavy and light chains. 4E10 (82) was purchased from Polymun.
Recombinant viruses expressing luciferase.293T HEK cells were cotransfected with plasmids expressing the pCMVΔP1Δenv HIV-1 Gag-Pol packaging construct, the HIV-1 Envs or control envelope glycoproteins from amphotropic murine leukemia virus (AMLV), human T-lymphotropic virus-1 (HTLV-1), simian immunodeficiency virus (SIVmac239), and human immunodeficiency virus-2 (HIV-2UC1), and the firefly luciferase-expressing vector at a DNA ratio of 1:1:3 μg using the Effectene transfection reagent (Qiagen) (60). The plasmids expressing the HIV-1 Envs and Rev protein were based on pSVIIIenv (60). Cotransfection produced recombinant luciferase-expressing viruses capable of a single round of infection. The virus-containing supernatants were harvested between 36 and 40 h after transfection, cleared of debris by low-speed centrifugation, aliquoted, and frozen at −80°C until further use. The reverse transcriptase (RT) levels of all viruses were measured as previously described (83).
Infection by single-round luciferase viruses.Cf2Th-CD4/CCR5, Cf2Th-CCR5, or Cf2Th-CD4/CXCR4 target cells were seeded at a density of 6 × 103 cells/well in 96-well luminometer-compatible tissue culture plates (PerkinElmer) 24 h before infection. On the day of infection, increasing concentrations of MF275, MF276, BNM-III-170, BMS-378806, maraviroc, enfuvirtide, or antibody were incubated with recombinant viruses at 37°C for 30 min. In the case of washout assays, increasing concentrations of compound were incubated with recombinant viruses at 37°C for 30 min; the virus was pelleted by centrifugation at 21,000 × g for 15 to 30 min at room temperature, the supernatant was discarded, and the virus pellet was washed once with medium before resuspension in medium. In the case of order-of-addition assays, one or more concentrations of compound were incubated with virus at 37°C for 30 min; then, one or multiple concentrations of a second compound were added to the virus-compound mixture and incubated at 37°C for an additional 30 min. In the case of sensitization assays, a constant concentration of compound was incubated with virus at 37°C for 30 min; then, increasing concentrations of antibody were added to the virus-compound mixture and incubated at 37°C for an additional 30 min. In all the above-described cases, the mixtures were then added to the target cells. At this point, in assays involving Cf2Th-CCR5 target cells or recombinant viruses with poor entry (<105 relative light units [RLU]/20 μl), the virus-compound mixtures were spinoculated onto target cells by centrifugation at 1,800 rpm for 30 min at 21°C (84). In the case of time-of-addition assays, recombinant viruses were spinoculated onto Cf2Th-CD4/CCR5 target cells at 1,800 rpm for 30 min at 21°C before being incubated at 37°C. A constant concentration of compound was then added to the virus-target cell mixture after increasing time intervals, and the virus-compound-target cell mixture was returned to a 37°C incubator for a total of 3 to 5 h. In all cases, the virus-compound-target cell mixtures were then diluted 1:4 in medium and incubated for 48 to 72 h at 37°C; after this time, the medium was removed from each well and the cells were lysed by the addition of 30 μl of passive lysis buffer (Promega) and three freeze-thaw cycles. An EG&G Berthold LB 96V microplate luminometer was used to measure the luciferase activity of each well after the addition of 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 firefly d-luciferin free acid, 99% (Prolume).
Measurement of Env-mediated cell-cell fusion.COS-1 cells were seeded at a density of 2.2 × 104 cells/well in 96-well luminometer-compatible tissue culture plates (PerkinElmer). The COS-1 cells were cotransfected with plasmids expressing the α fragment of β-galactosidase (85), HIV-1 Envs or control envelope glycoproteins, including amphotropic murine leukemia virus (AMLV), human T-lymphotropic virus-1 (HTLV-1), simian immunodeficiency virus (SIVmac239), and human immunodeficiency virus-2 (HIV-2UC1), and the Tat protein at a DNA ratio of 10:10:1 μg using the Effectene transfection reagent (Qiagen) (86). Concurrently, Cf2Th-CD4-CCR5, Cf2Th-CCR5, or Cf2Th-CD4-CXCR4 target cells (4 × 106) were transfected with a plasmid expressing the ω fragment of β-galactosidase (85) using the Effectene transfection reagent (Qiagen). Forty-eight hours posttransfection, target cells were removed from culture dishes with trypsin-EDTA and adjusted to 2 × 105 cells/ml in medium. The medium was removed from the donor cells and replaced with target cells followed by increasing concentrations of MF275. The virus-compound-target cell mixtures were then incubated for 6 to 8 h at 37°C; after this time, the medium was removed from each well and the cells were lysed by the addition of 20 μl of Galacto-Star lysis buffer (Applied Biosystems) and three freeze-thaw cycles. Galacto-Star substrate (diluted 1:50 in 100 μl/well Galacto-Star buffer diluent) was then added, and plates were incubated at room temperature before an EG&G Berthold LB 96V microplate luminometer was used to measure the β-galactosidase activity of each well.
Modeling MF275 binding to HIV-1 Env.Maestro (87) was used to prepare the Env docking target, which was based on the ligand-free sgp140 SOSIP.664 trimer (PDB 4ZMJ) (64). Hydrogens were added, termini were capped, and Cys 605 was reverted to the wild-type threonine residue, followed by a restrained energy minimization. AutoDock 4.2 (88) was used to generate 50 local energy-minimal conformations of MF275 with nonpolar hydrogens merged with heavy atoms to make united atoms. Docking was performed within a rectilinear box (50 by 50 by 50 Å3) centered on the four Env residues implicated in resistance to MF275. Docked poses were selected after 2.5 million rounds of energy evaluations. The two top-scoring poses are shown in Fig. 8.
Association of gp120 with solubilized Env trimers.HOS cells were transfected with plasmids expressing wild-type HIV-1JR-FL and HIV-1AD8 Envs with carboxy-terminal His6 tags. After 48 h, the cells were lysed in a buffer containing 1.5% Cymal-5 in the absence or presence of MF275 (60 μM) and/or soluble CD4 (200 nM). The clarified cell lysates were incubated with Ni-NTA agarose beads for 1.5 h at 4°C. After washing, the bound proteins were solubilized by boiling in LDS sample buffer (Invitrogen) and Western blotted. The Western blots were developed with a polyclonal goat anti-HIV-1 gp120 antibody (Invitrogen) and the 4E10 anti-gp41 antibody.
ACKNOWLEDGMENTS
We thank Elizabeth Carpelan for manuscript preparation. We also thank Irwin Chaiken and Wayne Hendrickson for valuable discussions and input.
This study was supported by the National Institutes of Health (GM56550/AI150471 and AI24755) and the late William F. McCarty-Cooper. N.M. was supported by amfAR grant 107431-45-RFNT, NIH AI90682, and a Ragon Institute Innovation Award. A.H. was supported by the phase II amfAR research grant 109285-58-RKVA.
FOOTNOTES
- Received 18 July 2019.
- Accepted 3 August 2019.
- Accepted manuscript posted online 7 August 2019.
- Copyright © 2019 American Society for Microbiology.
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.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
