Abstract
Six allosteric HIV-1 entry inhibitor modulators of the chemokine (C-C motif) receptor 5 (CCR5) receptor are compared for their potency as inhibitors of HIV-1 entry [infection of human osteosarcoma (HOS) cells and peripheral blood mononuclear cells (PBMC)] and antagonists of chemokine (C-C motif) ligand 3-like 1 [CCL3L1]-mediated internalization of CCR5. This latter activity has been identified as a beneficial action of CCL3L1 in prolonging survival after HIV-1 infection (
). The allosteric nature of these modulators was further confirmed with the finding of a 58-fold (HOS cells) and 282-fold (PBMC) difference in relative potency for blockade of CCL3L1-mediated internalization versus HIV-1 entry. For the CCR5 modulators, statistically significant differences in this ratio were found for maraviroc, vicriviroc, aplaviroc, Sch-C, TAK652, and TAK779. For instance, although TAK652 is 13-fold more potent as an HIV-1 inhibitor (over blockade of CCL3L1-mediated CCR5 internalization), this ratio of potency is reversed for Sch-C (22-fold more potent for CCR5-mediated internalization over HIV-1 entry). Quantitative analyses of the insurmountable antagonism of CCR5 internalization by these ligands suggest that all of them reduce the efficacy of CCL3L1 for CCR5 internalization. The relatively small magnitude of dextral displacement accompanying the depression of maximal responses for aplaviroc, maraviroc and vicriviroc suggests that these modulators have minimal effects on CCL3L1 affinity, although possible receptor reserve effects obscure complete interpretation of this effect. These data are discussed in terms of the possible benefits of sparing natural CCR5 chemokine function in HIV-1 entry inhibition treatment for AIDS involving allosteric inhibitors.
HIV-1 has been shown to use the chemokine C receptor CCR5 to gain entry into cells to cause infection (Alkhatib et al., 1996; Choe et al., 1996; Deng et al., 1996; Dragic et al., 1996); therefore, efforts have centered on the development of CCR5 HIV entry inhibitors for potential treatment of AIDS and AIDS prevention. Several molecules have been described previously (Kazmierski et al., 2003, 2005, 2007; Schols, 2006); presently, one of them, maraviroc, has been approved for therapeutic use (Fätkenheuer et al., 2005). Another CCR5 entry inhibitor, aplaviroc (Demarest et al., 2004; Maeda et al., 2004), has been shown to block the receptor in an allosteric manner (Maeda et al., 2004, Watson et al., 2005). Thus, although aplaviroc blocks the binding of the radioactive chemokine 125I-CCL3, it does not affect the affinity for the radioactive chemokine 125I-CCL5 (Maeda et al., 2004; Watson et al., 2005); such probe dependence is a characteristic of allosteric modulators (Jakubík et al., 1997; Christopoulos and Kenakin, 2002) and absolutely precludes a mechanism of orthosteric binding to the chemokine binding pocket.
Binding studies have shown prohibitive mutual binding kinetics between aplaviroc and the CCR5 entry inhibitors TAK779, Sch-C, maraviroc, and vicriviroc (Watson et al., 2005). These and other data support the idea that this new generation of HIV-1 entry inhibitors also are allosteric modulators of the CCR5 receptor (Dragic et al., 2000; Castonguay et al., 2003; Tsamis et al., 2003; Nishikawa et al., 2005; Maeda et al., 2006; Seibert et al., 2006; Westby et al., 2007; Kondru et al., 2008). In light of the fact that allosteric mechanisms are permissive with respect to allowing some receptor functions to be differentially altered more than others (Kenakin, 2005), this study explores the relative effects of CCR5 HIV entry inhibitors on HIV entry and chemokine-induced CCR5 receptor internalization. This latter mechanism has been suggested to be of value in protecting cells from HIV infection in the presence of chemokines. Because it has been reported that the chemokine CCL3L1 is associated with protection from progression to AIDS in patients infected with HIV (Gonzalez et al., 2005), the relative potency of six CCR5 entry inhibitors on HIV entry and CCL3L1-induced CCR5 internalization was studied. We show that there is a wide range of relative activity of these modulators on these two effects. These data are discussed in terms of general statements regarding the therapeutic use of allosteric modulators in AIDS treatment and also the potential for added efficacy for HIV-1 entry blockade through preservation of natural chemokine-induced internalization of CCR5.
Materials and Methods
Cells. Human osteosarcoma (HOS) cells expressing hCCR5 and hCD4 (Deng et al., 1996) were acquired from the NIH AIDS Research and Reference Reagent Program (Germantown, MD) and engineered to express an HIV-1 LTR-Luciferase reporter (Jenkinson et al., 2003). The cells were maintained in Dulbecco's modified Eagle's media (DMEM) containing 10% fetal calf serum (FCS), 1× nonessential amino acids, 400 μg/ml G418, 1 μg/ml puromycin, 40 μg/ml mycophenolic acid, 250 μg/ml xanthine, and 13.5 μg/ml hypoxanthine to maintain a selection pressure for cells expressing HIV-1 LTR-luciferase, hCCR5 and hCD4, respectively. Normal donor peripheral blood mononuclear cells (PBMCs) were isolated from random buffy coats (35-40 ml of elutriated whole blood in anticoagulant from HIV-negative donors) received from the American Red Cross, Southeast Division. PBMCs were isolated by density gradient centrifugation over lymphocyte separation medium (Mediatech, Herndon, VA) and stimulated in 150 ml of RPMI 1640 medium containing 20% FCS, 10% interleukin-2 (Zeptometrix, Buffalo, NY), 10 μg/ml gentamicin, and 5 μg/ml phytohemagglutinin (PHA; Sigma, St. Louis, MO) for 24 to 48 h.
Virus Strains. High-titer CCR5-tropic HIV-1 Ba-L was purchased from Advanced Biotechnologies (Columbia, MD) and expanded in PHA-stimulated PBMCs for 2 weeks with a refeeding of the culture after the first week. At the end of the 2-week period, the supernatants were clear-spun to remove cell debris, separated into 0.5-ml aliquots, and frozen at -80°C. An aliquot of the PBMC-expanded stock Ba-L was titrated into assay medium and assayed under standard conditions with test compounds not present. Viral input was determined by selecting a dilution that gave a signal within the linear range of the assay and a signal-to-background ratio of 75 to 100. Virus was diluted in assay medium before use.
HOS HIV Replication Cell Assay. HOS cells were lifted from tissue culture flasks by standard trypsinization techniques, resuspended in DMEM + 2% FCS to a density of 1.2 × 105 cells/ml. Fifty microliters of cells were dispensed into black-walled, clear-bottomed 96-well tissue culture-treated plates and placed in a humidified incubator at 37°C, 5% CO2 overnight. The following day, 50 μl of test compounds, serially diluted in medium, were added to the plates containing the HOS cells, and the plates were returned to the incubator for 1 h. A separate aliquot of diluted compound was added to an equal volume of diluted HIV-1 Ba-L, and 100 μl of this mixture was then added to the HOS/compound mixture. Infection proceeded in a humidified incubator at 37°C, 5% CO2 for 4 days. After this incubation period, supernatants were reduced to a volume of 50 μl, and an equal volume of reconstituted Steady-Glo reagent (Promega, Madison, WI) was added to each well. The plates were read in a TopCount luminometer (PerkinElmer Life and Analytical Sciences, Waltham, MA) at 1 s per well.
PBMC HIV Replication Cell Assay. HIV-1 Ba-L replication in PBMCs was quantitated by measuring reverse transcriptase activity present in the supernatant as described previously (Hazen et al., 2007). PHA-stimulated PBMCs were pelleted at 260g for 15 min, washed once with sterile phosphate-buffered saline, pelleted as before, resuspended to a density of 4 × 106 cells/ml in RPMI 1640 medium (containing 20% FCS, 10% interleukin-2, and 50 μg/ml gentamicin), and 100 μl were distributed to 96-well tissue culture plates. An equal volume of test compounds, serially diluted in medium, was added, and the plates were placed in a humidified incubator at 37°C, 5% CO2 for 1 h. A separate aliquot of diluted compound was added to HIV-1 Ba-L, and 100 μl of this mixture was then added to the PBMC/compound mixture. Infection proceeded in a humidified incubator at 37°C, 5% CO2 for 7 days. Fifty microliters of cell-free culture supernatant was transferred to a new 96-well plate. Ten microliters of RT extraction buffer (500 mM KCl, 50 mM dithiothreitol, and 0.5% Nonidet P-40) was added and mixed, followed by 40 μl of RT assay buffer (1.25 mM EGTA, 125 mM Tris/HCl, 12.5 mM MgCl2, 68 Ci/mmol [methyl-3H]deoxy-TTP, 0.62 O.D. units/ml poly(rA)-poly(dT)12-18). After mixing, room temperature reaction proceeded in a humidified incubator at 37°C, 5% CO2 for 2 h. Radioactivity was captured on 96-well plates (Unifilter DE-81; Whatman, Clifton, NJ) using a Univac vacuum manifold (Whatman). Wells were washed a total of three times: the first wash was 5% Na2HPO4, followed by distilled water, and, finally, 95% ethanol. The plates were read in a TopCount luminometer (PerkinElmer Life and Analytical Sciences) at 10 s/well.
Internalization of CCR5. U373-MAGI-CCR5-E cells were generated by Dr. Michael Emerman and obtained through the National Institutes of Health AIDS Research and Reagent Program. Culture media was DMEM/Ham's F12 with 10% fetal bovine serum, and 1% GlutaMAX (Invitrogen, Carlsbad, CA). Cells were plated in 96-well view plates (PerkinElmer Life and Analytical Sciences) and harvested from the culture flask using Cell Dissociation Solution from Sigma. They were then plated at 17 × 103 cells/well in 96-well view plates (PerkinElmer Life and Analytical Sciences) and incubated for ≈15 h at 37°C, 5% CO2.
Cell Treatment. Cells were washed twice with 50 μl/well PBS, pH 7.2. Test compounds were prepared as 1:5 serial dilutions in 100:1 dimethyl sulfoxide then diluted 1:100 in minimal essential medium-α + 1% FCS. Cells were incubated with 100 μl of test compound for 35 min at 37°C, 5% CO2. CCL3L1 (20 nM) in α-minimal essential medium + 1% FCS was then added to the cells/compounds, giving a final CCL3L1 of 10 nM (∼IC80). The cells were incubated for 35 min at 37°C, 5% CO2. Cells then were washed twice with PBS, fixed with 3.7% paraformaldehyde, and incubated at room temperature inside a chemical hood for 30 min. The cells were then washed twice with 50 μl/well wash buffer (0.5% Tween 20 in PBS, pH 7.2).
Cell membrane was stained with 4 ng/ml wheat germ agglutinin (reconstituted in physiological buffer solution, pH 7.4)-Alexa Fluor 488 from Invitrogen for 30 min in the dark. Then cells were washed twice with 50 μl/well wash buffer, and 50 μl/well 3.7% paraformaldehyde was added. This mixture was incubated for 5 min in a chemical hood. The cells were then washed twice with 50 μl/well wash buffer before blocking with 50 μl 1% BSA and 1% normal goat serum (Sigma) in PBS, pH 7.2 (block buffer), for 30 min, then permeabilized with PBS, pH 7.2, and 0.2% Triton-X (Sigma T8787) for 15 min in the dark.
The CTC8 (R&D Systems, Minneapolis, MN) mouse anti-human CCR5 primary antibody was reconstituted in PBS, pH 7.2, + 0.5% BSA to a concentration of 500 μg/ml. The Primary Antibody Solution was prepared by mixing block buffer with CTC8 to a 3 μg/ml final concentration. The permeabilization buffer from the previous step was aspirated, and 50 μl/well of primary antibody solution was added. Plates were then incubated at room temperature (in the dark) for 40 min. After the incubation, cells were washed twice with 50 μl/well wash buffer. Fifty microliters of secondary goat anti-mouse Alexa Fluor 568 F′(ab) (Invitrogen), 6 μg/ml in 2 μM Hoeschst 33342 (Lonza Walkersville, Walkersville, MD) in blocking buffers was added, and the plate was incubated at room temperature and protected from light for 40 min. After the incubation, the cells were washed three times with 50 to 100 μl/well wash buffer, then stored in 100 μl/well PBS buffer, pH 7.2. The plates were sealed with TopSeal-A (PerkinElmer Life and Analytical Sciences), and the bottom of the plates were wiped with 70% ethanol to remove any possible water marks. The plates were read using the Cellomics ArrayScan II instrument (Thermo Fisher Scientific Cellular Imaging, Pittsburgh, PA); data were analyzed with the GPCR BioApplication algorithm (Thermo Fisher Scientific Cellular Imaging).
Compounds. All compounds in this work [maraviroc (Perros et al., 2001), aplaviroc (Mitsuya et al., 2002), vicriviroc (Baroudy et al., 2000a,b), Sch-C (Baroudy et al., 2000a,b), TAK779 (Baba et al., 1999), and TAK652 (Shiraishi et al., 2003; Baba et al., 2005)] were synthesized as described in the literature.
Statistical Analysis. For the HOS and PBMC internalization assays, analysis was performed on the pIC50 scale using one-way analysis of variance. Mean and standard deviation were reported for each modulator and assay. S.E.M. was calculated using the analysis of variance estimate. The ratio of IC50 values for internalization versus HIV entry inhibition in HOS cells and PBL (“selectivity”) was estimated as 10 to the power of the difference between the mean pIC50 for HIV-1 entry and mean pIC50 for internalization. 95% confidence limits were calculated using critical values from a t distribution. The selectivity of different modulators was compared on the log-scale using a t test. Degrees of freedom for the t distribution were approximated using Satterthwaite's method (Steel et al., 1997). The correlation between pIC50 values for HIV entry inhibition in HOS cells and PBL was calculated by fitting a total least-squares regression model using maximum likelihood. Observed potency (calc. pKb) of CCR5 modulators was regressed against log concentration of CCL3L1 used in the internalization experiment. A linear regression model with a separate intercept and slope for each modulator was fit using ordinary least squares. For the HOS, PBMC, and U373 assays, normalized response data from duplicate titrations were fit using nonlinear least-squares to an unconstrained four parameter logistic equation, defined as follows: where x is compound concentration, y is normalized response data, Emax is upper bound of response, Emin is lower bound of response, nH is the Hill coefficient, and K is the concentration giving a normalized response halfway between Emin and Emax.
Raw data from the PBMC assay were normalized according to the following equation: where PC is the geometric mean of positive controls (mock-infected wells) on each assay plate, and NC is the geometric mean of negative controls (untreated but infected wells) on each assay plate. IC50 values were then derived from the following back calculation using parameter estimates from eq. 1. where Y.50 = log10[(PC + NC)/2] and nH is the Hill coefficient. The molar concentration producing 50% maximal entry inhibition (pIC50) was calculated as -log10(molar IC50).
Raw data from the antiviral assays and U373 internalization assay were normalized by expressing raw values as a percentage of the arithmetic mean of the no-compound controls on each assay plate. IC50 values were estimated using the K parameter estimate from eq. 1. pIC50 was calculated as -log10(molar IC50).
Results
HIV Entry Inhibition. The effects of six CCR5 modulators on HIV entry into HOS cells and human PBMCs were quantified by means of measuring the -log(IC50). As shown in Table 1, potencies of these ligands ranged from 0.4 (TAK652) to 12.6 nM (TAK779) in the HOS assay, the order of potency being TAK652 > maraviroc > aplaviroc > vicriviroc > Sch-C > TAK779. Likewise, in the PBMC assay, potencies ranged from 0.38 (TAK652) to 100 nM (TAK779), the order of potency being TAK652 > aplaviroc > maraviroc > vircriviroc > Sch-C > TAK779. The correlation between anti-HIV activity in HOS and PBMC is shown in Fig. 1 (r = 0.68). The correlation of 0.68 and the graph show that HOS and PBMC seem to be related; however, the fitted line in Fig. 1 is quite different from the line of identity (expected if the two assays gave identical pIC50 values). For the less potent compounds, HOS tends to give higher pIC50 than PBMC, but for compounds that are more potent, the two assays give quite similar pIC50 values. The interpretation of this with respect to differences between results based on PBL versus HOS is not clear.
CCR5 Internalization. The activation of seven transmembrane receptors, such as CCR5, has been shown to cause their internalization into the cell, and this is a common observation for CCR5 receptor activation by a number of chemokines (Amara et al., 1997; Mack et al., 1998). As shown in Fig. 2A, the chemokine CCL3L1 produces concentration-dependent internalization of CCR5 (pEC50 = 8.7) that is prevented by maraviroc (Fig. 2B). Specifically, addition of CCL3L1 (at a concentration producing 80% maximal internalization) to cells pre-equilibrated with a range of concentrations of maraviroc produces a curve for antagonism of CCL3L1-induced internalization. The antagonism is then quantified through the midpoint of this antagonistic curve in the form of a pIC50 (see Fig. 2B).
It was important to determine the system dependence of the estimate of antagonism, because conclusions about molecular mechanisms can better be made with equilibrium dissociation constants for antagonist-receptor complexes (pKB). For a surmountable competitive antagonism, a correction factor needs to be applied to the pIC50 to convert these to pKI values; this is not the case for insurmountable (noncompetitive) antagonism (Kenakin et al., 2006). Therefore, data to assess the mode of action of the CCR5 modulators was obtained by determining concentration-response curves for internalization to CCL3L1 in the absence and presence of a range of concentrations of modulator. As can be seen in Fig. 3, aplaviroc produces an insurmountable antagonism of the internalization effects of CCL3L1 with little displacement of the CCL3L1 curve location parameter (maximal displacement, 2.6-fold; see Table 2). Similar patterns of inhibition were observed for TAK779 (Fig. 4A), maraviroc (Fig. 4B), TAK652 (Fig. 4C), and vicriviroc (Fig. 4D). Table 3 shows a summary of the data for inhibition of CCL3L1-induced CCR5 internalization. It can be seen from Table 3 that the modulators range in potency from 6.6 (TAK779) to 0.76 nM (maraviroc). The relatively small dextral displacement of the curves with depressed maxima is given in Table 2. From these data it can be seen that virtually no displacement was observed for aplaviroc, maraviroc, and vicriviroc, whereas a 7-fold maximal displacement was observed for TAK779 and TAK652. The low level of displacement indicates that pIC50 values are within 3- to 4-fold range.
Relative Inhibition of HIV and Internalization. The curves for inhibition of HIV entry and CCL3L1-induced CCR5 inhibition were compared to yield a ratio of potency for the modulators on these respective processes. Figure 5 shows the percentage inhibition curves for inhibition of HIV entry into HOS cells versus inhibition of internalization for TAK779 and TAK652. It can be seen that although TAK779 is more potent as an inhibitor of internalization than HIV (in this example, pIC50 internalization = 8.64 versus pIC50 HOS = 7.68), this order of relative potency is reversed for TAK652. Specifically, TAK652 is more potent as an inhibitor of HOS than internalization (in this example, pIC50 internalization = 8.5 versus pIC50 HOS = 9.42). This profile can be presented as a single curve by expressing the internalization curve as a function of the pIC50 concentration required to block HIV entry. Specifically, the abscissal axis for TAK779 for the internalization curve can be converted to a ratio of concentrations required for internalization versus HIV entry inhibition ([CCL3L1/IC50 HOS]) to yield a curve that expresses the relative activity of the modulator on internalization as a function of activity in the HOS. Fig. 5, top, shows this conversion for TAK779 and TAK652. From this figure it can be seen that the concentration required to produce 50% inhibition of internalization for TAK779 is 0.09 times the concentration required to block HIV entry into HOS cells (location parameter of the relative activity curve is centered on 0.09). This indicates that TAK779 is 11 times more active as a blocker of internalization than HIV entry. In contrast, the location parameter of the relative activity curve for TAK652 is centered on 9. This indicates that the concentration of TAK652 required to block internalization is 9 times greater than the concentration required to block HIV entry; i.e., TAK652 is 9 times more active as a blocker of HIV entry than CCL3L1-mediated CCR5 internalization.
Figure 6 shows the curves for internalization for all of the modulators expressed as a fraction of the concentration required to block HIV entry into PBMC. It can be seen from this figure that there is considerable texture in the relative effects of these modulators on internalization versus HIV entry. The data describing the specific activity of these modulators on internalization versus HOS cell and PBMC HIV entry is given in Table 4. The selectivities for these modulators on internalization versus HIV entry is shown graphically in Fig. 7. It can be seen from this figure that the modulators fall into loose categories with respect to internalization versus HIV entry. Specifically, for the HOS, compounds cluster into groups comprising TAK652 > (aplaviroc, vicriviroc, maraviroc, TAK779 > SCh-C). For the PBMC, the clusters are slightly different: TAK652 > (aplaviroc, vicriviroc, maraviroc) > (TAK779, Sch-C). To compare the modulators with one another with respect to internalization versus HIV-entry selectivity, t tests were carried out; the summary of these comparisons is shown in Fig. 8. Where modulators do not share a common letter, the differences in their selectivity are statistically different (P ≤ 0.01).
Molecular Mechanism of Internalization Blockade. These entry inhibitors have been shown to be allosteric with respect to binding of HIV and chemokines to CCR5. Under these circumstances, their binding is permissive (Kenakin, 2005) in that there are protein species consisting of simultaneously bound CCL3L1 and modulator and also HIV and modulator. Such systems allow the cobinding ligand (either CCL3L1 or HIV) to influence the effect of the modulator. The most simple model to describe such an allosteric system combines the Ehlert model for allosteric binding (Ehlert, 1988) and the operational model for GPCR function (Black and Leff, 1983); it dictates that the fractional quantity of receptor species bound to either HIV or CCL3L1 in the presence of the modulator is described by the following equation (Kenakin, 2005; Price et al., 2005; Kenakin et al., 2006) where R (response) denotes either CCL3L1-induced internalization or HIV entry. The equilibrium dissociation constants for the cobinding ligand (denoted [A]) and the modulator (denoted [B]) are KA and KB, respectively. τ represents the pharmacological efficacy coupling of the receptor for fusion or separately for internalization, α is the change in the affinity of the cobinding ligand (either HIV or CCL3L1) when the modulator is bound, and β is the change in efficacy of the cobinding ligand with modulator bound [note that in Kenakin (2005) and Kenakin et al. (2006), the efficacy term was denoted ζ; this has been changed to β to be consistent with Price et al., 2005]. It is important to note that, in keeping with the known probe dependence of allosteric modulators (Christopoulos and Kenakin, 2002), the values for α and β for HIV may (and most likely will) be different from those for CCL3L1.
Equation 4 can be used to deduce selective effects of modulators on affinity (α) and efficacy (β) of CCL3L1 for CCR5. The change in the location parameter of CCL3L1 curves in the presence of modulator (expressed as a dose ratio of EC50 values where DR = [EC50 in presence of modulator]/[EC50 in absence of modulator]) is given by: Because these modulators completely suppress CCL3L1-induced internalization, β = 0, reducing the relation to: It can be seen from eq. 6 that if the modulator reduces the affinity of CCL3L1 for CCR5 (α < 1), then DR will necessarily be > 1; i.e., CCL3L1 curves will shift to the right with increasing concentrations of modulators. However, if the modulator has no effect on affinity (α = 1) or potentiates CCL3L1 (α > 1), then DR values will be low (little shift to the right) for systems of low receptor reserve (low τ values). Dextral displacement still can be observed, however, in systems of high receptor reserve (high τ values). The fact that there was little dextral displacement observed in Figs. 3 and 4 for aplaviroc, maraviroc, and vicriviroc (full curves were not obtained for Sch-C) is consistent with the notion that these modulators produce very little effect on CCL3L1 affinity. Thus, their main mechanism of action for inhibition of internalization is through blockade of CCL3L1 efficacy for the induction of internalization. Some dextral displacement was observed for TAK779 and TAK652, indicating that these modulators reduce the affinity of CCL3L1 as well as the efficacy.
Discussion
Numerous lines of evidence from structural and mutagenesis studies (Dragic et al., 2000; Castonguay et al., 2003; Tsamis et al., 2003; Nishikawa et al., 2005; Maeda et al., 2006; Seibert et al., 2006; Westby et al., 2007; Kondru et al., 2008) and biochemical studies (Watson et al., 2005) indicate that these small-molecule inhibitors of CCR5-mediated HIV entry are allosteric modulators. The data describing the allosteric modulation of CCR5 receptors reported here illustrate some further characteristic allosteric properties of these ligands. Specifically, allosteric modulators have three features that distinguish them from orthosteric antagonists. Although the former simply preclude access of the endogenous ligand to the receptor, allosteric modulators cobind with the natural ligand to the receptor and can modify the natural ligands' subsequent actions. This leads to the following three behaviors: 1) modulator effects are saturable (receptor function can be modified but need not be completely blocked), 2) modulators are probe-dependent (different endogenous ligands can be differentially affected), and 3) modulators can affect the affinity and efficacy of ligands to a different extent. This study illustrates two of these three properties: probe dependence and differential alteration of affinity and efficacy.
The allosteric probe dependence demonstrated in these studies is the differential relative sensitivity of the HIV entry and internalization process, both mediated by CCR5. The highly statistically significant difference in the ratio of IC50 values for HIV entry versus internalization within the group of CCR5 modulators is inconsistent with a simple orthosteric binding of these ligands to the receptor. If the interaction were orthosteric, then the antagonism for all pathways mediated by the single receptor would be uniform. This does not necessarily mean that the potency of antagonism for internalization and HIV entry should be the same, because the two assays may be run at differing stimulus inputs. However, any differences seen between the pIC50 values for the two processes in an orthosteric environment would be constant for all antagonists, and no texture in relative activity would be seen. The fact that there are statistically significant differences in the relative potencies for the blockade of internalization versus HIV entry precludes orthosteric interaction as a mechanism and is consistent with the known allosteric nature of these molecules.
The second allosteric property exhibited by these molecules interacting with CCR5 is the separate modulation of affinity and efficacy. This has already been demonstrated in a striking manner for aplaviroc, a modulator that completely suppresses the efficacy of RANTES for CCR5 but has insignificant effects on the affinity of RANTES for the receptor. Thus, aplaviroc does not affect RANTES binding but does affect the ability of RANTES to mediate calcium response (Maeda et al., 2004; Watson et al., 2005). In the present studies, the concentration-response curves to CCL3L1 observed in the presence of increasing concentrations of the modulators indicated that three of then (maraviroc, aplaviroc, vicriviroc) produce their blockade of CCL3L1-induced CCR5 internalization through blockade of efficacy (the mechanism of receptor activation leading to internalization) without affecting the receptor affinity for CCL3L1. In contrast, some reduction in the affinity of CCL3L1 was produced by TAK779 and TAK652.
The separate modulation of affinity and efficacy by allosteric ligands is a potentially therapeutically valuable property. For example, if an allosteric antagonist were to increase the affinity of the receptor for the endogenous agonist, then a unique increase in antagonism with increasing agonist stimulus would result. This is because allosteric effects are reciprocal, meaning that if the allosteric modulator increases the affinity of the endogenous agonist, then the endogenous agonist also will increase the affinity of the receptor for the antagonist. Therefore, the higher the level of agonism in the system, the more potent will be the antagonist (assuming it blocks the effect through a suppression of efficacy). Such a stimulus-linked antagonism has been shown for the modulator ifenprodil for N-methyl-d-aspartate receptors (Kew et al.,1996) and for the cannabinoid CB1 receptor allosteric modulator Org27569 (Price et al., 2005). In the present case, this would mean that the modulator would increase its ability to block HIV in patients with higher levels of virus.
The second theme of this article relates to the therapeutic ramifications of the allosteric mechanism for these ligands in the blockade of HIV entry. Unlike orthosteric antagonists, which preclude receptor probe (HIV, chemokines) binding to the receptor, allosteric ligands produce permissive effects (Kenakin, 2005) because the modulator and probe bind simultaneously to the receptor. This opens the possibility of a probe dependence that would enable a natural chemokine functional activity for the receptor but preclude use by HIV (see Fig. 9, schematic). It is not completely clear whether sparing natural chemokine function for CCR5 systems would be beneficial or detrimental therapeutically, although there is considerable evidence to allow informed speculation.
The apparent normal health of patients with the Δ32 allele [which prevents functional cellular expression of CCR5 to the cell surface (Dean et al., 1996; Huang et al., 1996)] supports the notion that removal of CCR5 function has no negative effect on normal health. Under these circumstances, there would seem to be no impetus to preserve chemokine function, at least within the context of preventing deleterious effects of removing CCR5 function. However, a closer examination of the Δ32 population reveals a higher incidence of health abnormalities, such as increased severity of symptoms in West Nile Virus disease with concomitant increased risk of death (Glass et al., 2006) and, notably, abnormality in hepatic function. For example, patients with the Δ32 allele have a greater incidence of liver disease and sclerosing cholangitis (Eri et al., 2004). Furthermore, hepatic pathology is found in patients with the Δ32 allele, who have increased vulnerability to liver disease (Welniak et al., 2004) and a greater risk of biliary lesions and mortality after liver transplantation (Moench et al., 2004). In addition, a mild form of immunodeficiency has been associated with loss of CCR5 function (de Silva and Stumpf, 2004; Lederman et al., 2006). These abnormalities in liver function coupled with the observed acute liver toxicity of the CCR5 HIV entry inhibitor aplaviroc (Crabb, 2006) suggest that CCR5 may play a role in normal liver function or protection against liver insult. However, a strong counter to this idea is the observation that no liver toxicity is seen in long-term maraviroc treatment of patients with AIDS in clinical trials. On the strength of this finding alone, it may be premature to seriously consider the loss of CCR5 function as directly deleterious to normal health.
The other possible benefit of sparing normal CCR5 chemokine function, in addition to its role in normal initiation of adaptive immune responses and trafficking of effector cells to sites of infection and inflammation (Lederman et al., 2006), is the potential protective effect of this system against HIV. This possibility is rooted in the knowledge that chemokines block HIV entry in vitro (Simmons et al., 1997; Mack et al., 1998). Likewise, high levels of chemokines have been associated with delay of AIDS progression (Ullum et al., 1998; Garzino-Demo et al., 1999; Shieh et al., 2001; Heredia et al., 2003; Rogez et al., 2003; Xiang et al., 2004). A particularly interesting study of 1064 patients infected with HIV in 57 populations around the world showed a strong inverse correlation between the gene copy number for CCL3L1 (this is variable in humans) and progression to AIDS. Thus, patients with high CCL3L1 gene copy number showed a highly statistically significantly greater survival rate compared with patients with low CCL3L1 gene copy number.
Although some HIV inhibition by chemokines may be due to steric inhibition of HIV binding (Pastore et al., 2003), the most prevalent data indicate that chemokines protect against HIV infection through internalization of the receptor (Alkhatib et al., 1997; Amara et al., 1997; Mack et al., 1998; Pastore et al., 2003). In fact, some modified chemokines, such as PSC-RANTES (Hartley et al., 2004), have been engineered to prevent HIV entry through CCR5 internalization. Therefore, an obvious possible reason for the HIV protective effect of high gene copy number for CCL3L1 is that the patients producing higher levels of CCL3L1 may have lower cell surface levels of CCR5 because of CCL3L1-induced internalization of CCR5. It is noteworthy that an inverse link between CCL3L1 and CCR5 receptor levels has been reported (Ketas et al., 2007). The corollary to this idea is the known faster progression to AIDS of patients infected with HIV who have allelic variants of the CCR5 promoter that cause elevated levels of CCR5 on the cell surface (Martin et al., 1998; McDermott et al., 1998).
As a baseline, a CCR5 HIV treatment that blocks HIV entry and chemokine function should be effective. However, if a functioning chemokine system assists in preventing HIV entry and subsequent progression to AIDS, then a CCR5 HIV-blocking drug that allows chemokines to function normally (Fig. 9) may offer a significant improvement in efficacy. Under these circumstances, the modulator-mediated HIV entry blockade would be assisted by a healthy functioning natural chemokine system (much like the effect seen with CCL3L1 gene copy number).
The previous discussion illustrates the possible advantages of a function-sparing CCR5 allosteric modulator in terms of HIV entry, raising the possibility that there may be some deleterious effects of a functioning CCR5 chemokine system in AIDS as well. This raises the alternative possibility that there might be beneficial aspects of the blockade of CCR5 function by CCR5 entry inhibitor modulators. Specifically, the role of CCR5 in chemokine-induced activation and migration of leukocytes (Beck et al., 1997), the pro-inflammatory effects of CCR5 activation by chemokines (Nardese et al., 2001) and possible role in HIV replication through CCR5 receptor activation, either by chemokines (Kinter et al., 1998; Lin et al., 2006) or HIV envelope (Weissman et al., 1997), could theoretically contribute to complications in the treatment of AIDS. With this in mind, the blockade of the inflammatory effects of CCR5 activation could actually be beneficial. However, these arguments are usually considered when CCR5 agonists such as PSC-RANTES, through their known active CCR5 internalizing properties, are proposed as treatments. Under these circumstances, CCR5 agonism would be a prevalent and constant effect. The scenario with function-sparing CCR5 modulators is quite different in that they would only allow chemokine stimulation due to normal physiological function and would not produce constant agonism. Accordingly, it could be argued that an intermittent CCR5 activation through a naturally operating chemokine system would not produce undue inflammatory agonism. On balance, the idea of preserving natural chemokine function may be so dissimilar to a scenario of powerful synthetically induced CCR5 agonism as to not be relevant to this present discussion.
In summary, it presently is not clear whether CCR5 function-sparing HIV entry inhibition would be useful in the treatment and prevention of AIDS, although the data linking CCL3L1 function to survival is strongly suggestive. It is clear that as more allosteric modulators are introduced into therapy, texture in HIV versus chemokine inhibition, an expected property of allosteric modulation, may be observed. It will be very interesting to compare therapeutic profiles of these modulators in the future to determine, through translational medicine, whether such function-sparing effects truly provide a therapeutic advantage.
Footnotes
-
ABBREVIATIONS: CCL, chemokine (C-C motif) ligand; CCR, chemokine (C-C motif) receptor; Sch-C, (Z)-(4-bromophenyl){1′-[(2,4-dimethyl-1-oxido-3-pyridinyl)carbonyl]-4′-methyl-1,4′-bipiperidin-4-yl}methanone O-ethyloxime; maraviroc (UK 427,857), 4,4-difluoro-N-((1S)-3-{(3-endo)-3-[3-methyl-5-(1-methylethyl)-4H-1,2,4-triazol-4-yl]-8-azabicyclo[3.2.1]oct-8-yl}-1-phenylpropyl)cyclohexanecarboxamide; TAK779, N,N-dimethyl-N-[4-[[[2-(4-methylphenyl)-6,7-dihydro-5H-benzocyclohepten-8-yl]carbonyl]amino]benzyl]tetrahydro-2H-pyran-4-aminium chloride; TAK652, (S)-8-[4-(2-butoxyethoxy)phenyl]-1-isobutyl-N-(4-{[(1-propyl-1H-imidazol-5-yl)methyl]sulfinyl}phenyl)-1,2,3,4-tetrahydro-1-benzazocine-5-carboxamide monomethanesulfonate; vicriviroc (Sch-D), 4,6-dimethyl-5-{[4-methyl-4-((3S)-3-methyl-4-{(1R)-2-(methyloxy)-1-[4-(trifluoromethyl)phenyl]ethyl}-1-piperazinyl)-1-piperidinyl]carbonyl}pyrimidine; aplaviroc 873140, 4-{[4-({(3R)-1-butyl-3-[(R)-cyclohexyl(hydroxy)methyl]-2,5-dioxo-1,4,9-triazaspiro[5.5]undec-9-yl}methyl)phenyl]oxy}benzoic acid hydrochloride; MIP-1α (CCL3), macrophage inflammatory protein 1-α; MIP-1β (CCL4), macrophage inflammatory protein 1-β; RANTES (CCL5), regulated on activation normal T-cell expressed and secreted; HOS, human osteosarcoma; CCL3L1 (LD78b), APLAADTPTACCFSYTSRQIPQNFIADYFETSSQCSKPSVIFLTKRGRQVCADPSEEWVQKYVSDLELSA (MIP-1α, LD78-β isoform, human); Org 27569-0, 5-chloro-3-ethyl-1H-indole-2-carboxylic acid [2-(4-piperidin-1-yl-phenyl)-ethyl]-amide; PBMC, peripheral blood mononuclear cells; PBL, peripheral blood lymphocytes; DMEM, Dulbecco's modified Eagle's media; FCS, fetal calf serum; PHA, phytohemagglutinin; PBS, phosphate-buffered saline.
- Received October 8, 2008.
- Accepted November 24, 2008.
- The American Society for Pharmacology and Experimental Therapeutics