Chemokine (C-C motif) receptor 5-using envelopes predominate in dual/mixed-tropic HIV from the plasma of drug-naive individuals

Objective: HIV-1 utilizes CD4 and either chemokine (C-C motif) receptor 5 (CCR5) or chemokine (C-X-C motif) receptor 4 (CXCR4) to gain entry into host cells. Small molecule CCR5 antagonists are currently being developed for the treatment of HIV-1 infection. Because HIV-1 may also use CXCR4 for entry, the use of CCR5 entry inhibitors is controversial for patients harboring CCR5-using and CXCR4-using (dual/mixed-tropic) viruses. The goal of the present study was to determine the proportion of CCR5-tropic and CXCR4-tropic viruses in dual/mixed-tropic virus isolates from drug-naïve patients and the phenotypic and genotypic relationships of viruses that use CCR5 or CXCR4 or both.

Design: Fourteen antiretroviral-naive HIV-1-infected patients were identified as having population coreceptor tropism readout of dual/mixed-tropic viruses. Intrapatient comparisons of coreceptor tropism and genotype of env clones were conducted on plasma virus from each patient.

Methods: Population HIV-1 envelope tropism and susceptibility to the CCR5 entry inhibitor, aplaviroc, were performed using the Monogram Biosciences Trofile Assay. Twelve env clones from each patient were analyzed for coreceptor tropism, aplaviroc sensitivity, genotype, and intrapatient phylogenetic relationships.

Results: Viral populations from antiretroviral-naive patients with dual/mixed-tropic virus are composed primarily of CCR5-tropic env clones mixed with those that use both coreceptors (R5X4-tropic) and, occasionally, CXCR4-tropic env clones. Interestingly, the efficiency of CXCR4 use by R5X4-tropic env clones varied with their genetic relationships to CCR5-tropic env clones from the same patient.

Conclusion: These data show that the majority of viruses in these dual/mixed-tropic populations use CCR5 and suggest that antiretroviral-naive patients may benefit from combination therapy that includes CCR5 entry inhibitors.

Introduction

Coreceptor tropism describes the HIV-1 envelope glycoprotein (Env) use of the chemokine receptors, chemokine (C-C motif) receptor 5 (CCR5) (R5-tropic) or chemokine (C-X-C motif) receptor 4 (CXCR4) (X4-tropic), for entry into CD4+ host cells. Natural history studies of HIV-1 infection have shown that most patients harbor R5-tropic virus populations soon after infection and through the asymptomatic phase ^[1]. However, approximately 50% of individuals exhibit tropism expansion to include the use of CXCR4 prior to the decline in CD4+ T-cell numbers and progression to AIDS ^[1,2]. Consistent with natural history studies, cross-sectional epidemiologic studies of virus tropism have shown that approximately 80-85% of HIV-infected antiretroviral therapy (ART)-naive patients exhibit R5-tropic virus prior to receiving ART, with the remainder having a dual/mixed-tropic readout ^[3,4]. Given that CCR5 entry inhibitors do not inhibit CXCR4-using virus, it has become increasingly important to examine the composition of dual/mixed-tropic virus in ART-naive patients in order to understand the potential efficacy of CCR5 entry inhibitors in this patient population.

Methods

Patient population

Fourteen ART-naïve, HIV-1-infected patients with a plasma tropism readout of dual/mixed-tropic virus on two consecutive visits approximately 30 days apart were examined, and the sample from the second visit was tested for HIV Env tropism, aplaviroc susceptibility and clonal analyses. Plasma HIV-1 RNA and absolute CD4+ T-cell counts were determined by Quest Diagnostics (Van Nuys, California, USA and Heston, UK). All work was conducted in accordance with the Declaration of Helsinki. The study was approved by the Institutional Review Boards at each site, and participants signed an informed consent.

HIV-1 Env tropism and aplaviroc susceptiblity

HIV-1 Env tropism testing was performed on all plasma samples using the Trofile Assay (Monogram Biosciences, South San Francisco, California, USA) ^[5]. Briefly, a replication defective, luciferase-expressing modification of HIV-1 strain, NL4-3, was cotransfected with an expression vector containing the amplified env sequence from the sample of interest into human embryonic kidney 293 (HEK293) cells. The resulting pseudoviruses were then used to infect U87 cells expressing CD4 in combination with either CCR5 or CXCR4 in the presence or absence of CCR5 and CXCR4 entry inhibitors, respectively. An R5X4-tropic reference virus (92HT594) was tested in each assay performed to normalize the R5 and X4 relative light units (RLUs) of the sample virus on CCR5+ and CXCR4+ cells, respectively. In addition, X4-tropic (HXB2) and R5-tropic (JRCSF) HIV-1 controls were tested in each assay. A sample was determined to be X4 or R5 tropic if the luciferase activity (RLUs) from lysed cells expressing the respective coreceptor was above background and if there was clear evidence of RLU inhibition by the specific respective coreceptor inhibitor.

Aplaviroc susceptibility data were also generated using the Trofile Assay. Two measures of susceptibility, the fold change in 50% inhibitory concentration of viral replication relative to a reference virus (FCIC₅₀) and maximum calculated percentage inhibition (MPI), were collected. As previously described ^[6], the percentage inhibition was derived using the equation, [1 - (luciferase activity in the presence of drug/luciferase activity in the absence of drug)] × 100. Inhibition curves defined by the four-parametric sigmoidal function, f(x) 5 a 2[b/(1 1 (x/c)d)], were fit to the data by nonlinear least squares and bootstrapping and were used to calculate the drug concentrations required to inhibit 50% of virus replication (IC₅₀). FCIC₅₀ was determined by comparing the IC₅₀ for the sample virus with the IC₅₀ for the drug-sensitive, R5-tropic control JRCSF. MPI was calculated as the percentage inhibition of virus replication at the highest final drug concentration tested in the assay. As the clinically relevant FCIC₅₀ cut point for aplaviroc resistance has not been determined, a cut point of three-fold, based on the intrinsic assay variability, was used in this analysis. The clinically relevant MPI value for aplaviroc has not been determined. Of note, phenotypic data related to R5X4-tropic clones or dual/mixed-tropic populations or both were obtained on cells that express CCR5 in the absence of CXCR4 and do not account for the ability of the virus(es) to replicate using CXCR4.

Clonal analyses

The specimen from the second visit for each patient was used for a clonal analysis of HIV-1 Env, which was performed at Monogram Biosciences. Forty-eight clones were screened for tropism to determine the relative frequency of HIV Env tropism. Twelve env clones, representing the variety of tropism phenotypes observed in the 48 clone assay, were selected for analysis of full-length HIV env genotype (gp160), aplaviroc sensitivity, and confirmation of tropism. The number of env clones with particular tropism readouts in the 12 clone analysis does not necessarily reflect the relative percentage of these clones in the viral quasispecies, as the numbers of clones are not necessarily chosen in proportion to a larger 48 clone tropism screening. However, selection did represent each of the minority species observed.

Molecular evolutionary analyses

Homologous nucleotide sequence alignments of full-length env clones were generated for each patient using ClustalX version 1.83 with manual refinement in Genedoc version 2.6.0.2. Maximum likelihood-based phylogenetic reconstructions were performed using PAUP 4.0b with the best fitting model of sequence evolution and the corresponding values for the rate matrix, shape of the gamma distribution, and proportion of invariant sites estimated by the Modeltest program. Maximum likelihood bootstrap support values were obtained using PAUP 4.0b with 100 replicates. In addition, 100 replicates of neighbor-joining were performed to corroborate the maximum likelihood findings. Neighbor-joining and maximum likelihood values were highly concordant, and thus, only maximum likelihood trees and bootstrap values are reported. In order to investigate V3 changes in the R5X4 env clones, amino-acid sequences for Env proteins were translated from the nucleotide sequence alignment for each patient using Genedoc version 2.6.0.2 and an analysis of the V3 region of Env was visually performed. An R5X4-tropic env clone was scored as having an amino acid change if the residue at that V3 alignment position was not detected in any of the R5-tropic env clones from the same patient.

Results

Virus pseudotyped with the env population derived from 14 ART-naive patients identified as having dual/mixed-tropic HIV-1 using the Monogram Biosciences Trofile Assay were selected for clonal analysis. Twelve clones from each patient were selected for tropism confirmation, susceptibility to aplaviroc, and molecular evolutionary analyses. R5-tropic clones were detected in all patients and represented at least 50% of clones in 12 of 14 patients (Fig. 1a). Two patients (DM8 and DM12) primarily harbored (68 and 75%, respectively) R5X4-tropic env clones. Eleven patients had R5-tropic and R5X4-tropic clones, one patient had R5-tropic and X4-tropic clones, and two patients had clones of all three tropism phenotypes.

Fig. 1

In addition to tropism, aplaviroc susceptibility of the R5-tropic component of the virus population from all 14 patients was tested. Aplaviroc IC₅₀ values for the standard HIV-1 controls (strain JRCSF) across all assay runs were within 2.2-fold of each other and comparisons of the X4 and R5 luciferase activity as measured by RLU from interassay controls were within 13-fold of each other. The viral populations from all 14 patients were found to be highly sensitive to aplaviroc inhibition as determined by FCIC₅₀ (Table 1). Additionally, the viral population MPI values for each patient were all more than 96%. Consistent with population level results, the majority of env clones were found to be sensitive to aplaviroc, with FCIC₅₀ and MPI values comparable to the population values for each patient (Fig. 1b). Apart from a single R5X4-tropic clone from patient DM6 with a FCIC₅₀ of 10 and an MPI of 85.74%, the ranges of FCIC₅₀ and MPI values were similar for R5-tropic clones (0.04-1.78 and 87.80-99.96%, respectively) and R5X4-tropic clones (0.02-2.16 and 87.99-99.94%, respectively) (Table 1). With the exception of the aforementioned single clone from patient DM6, no evidence of reduced clonal aplaviroc FCIC₅₀ susceptibility (≥3 FCIC₅₀) was observed.

Table 1

Phylogenetic analyses of full-length env (gp160) sequences within each patient showed that the R5X4-tropic clones fell into two distinct categories. Although all X4-tropic clones segregated (100% bootstrap support) away from intrapatient R5-tropic clones, some R5X4-tropic clones segregated (86-100% bootstrap support) away from intrapatient R5-tropic clones, whereas others clustered with intrapatient R5-tropic clones (Fig. 2). Seven patients (DM1, DM3, DM4, DM5, DM9, DM10, and DM11) had R5X4-tropic clones that grouped separately, three (DM7, DM12, and DM13) had R5X4-tropic clones that intermingled, and three (DM6, DM8, and DM14) had both R5X4-tropic clones that grouped separately and those that intermingled with intrapatient R5-tropic clones. One patient (DM2) had R5X4-tropic clones that grouped separately from the R5-tropic clones but with low (54%) bootstrap support.

Fig. 2

To understand the significance of the phylogenetic categories of R5X4-tropic clones, their relationship with intrapatient R5-tropic clones was assessed by comparing the X4 RLUs, aplaviroc FCIC₅₀ values, and the sequence of V3, a region of Env demonstrated to be a primary determinant of coreceptor use ^[7-9]. Of note, the R5X4-tropic clones that were phylogenetically distinct from intrapatient R5-tropic clones had two or more V3 amino acid changes (Fig. 3a,b), were more sensitive to aplaviroc inhibition (Fig. 3a,c) and produced higher X4 RLUs (Fig. 3b,c). In contrast, the R5X4-tropic clones that intermingled with R5-tropic clones had fewer V3 amino acid changes (one or none) and were less sensitive to aplaviroc inhibition. These R5X4-tropic clones exhibited substantially lower X4 RLUs, often very close to the lower detection limit of the tropism assay.

Fig. 3

Discussion

The present study provides the first clonal characterization of the tropism, aplaviroc susceptibility, and phylogenetic relationships of individual full-length HIV-1 envelope glycoproteins from the plasma of dual/mixed-tropic ART-naive patients. We found that the majority of these patients primarily harbor R5-tropic env clones. Similar to observations recently made by Toma et al. ^[10], an examination of the R5X4-tropic clones in the present study revealed that R5X4-tropic clones with a reduced efficiency of CXCR4 use grouped phylogenetically with intrapatient R5-tropic clones, whereas those with a greater efficiency of CXCR4 use segregated away from intrapatient R5-tropic clones. In addition, as noted by Toma et al., we observed that clones with a lower efficiency of CXCR4 use had V3 sequences that were more similar to R5 topic clones. These two categories of R5X4-tropic clones could indicate that low X4 RLU clones are transitional intermediates evolving toward more efficient use of CXCR4. Alternatively, the closeness of the low X4 RLUs to the sensitivity threshold of the tropism assay may have resulted in the misclassification of these clones as using CXCR4. Additional studies are needed to examine the ability of these clones to infect primary cells and to assess the potential clinical significance of differences in efficiency of coreceptor use.

Currently, the only commercially available assay used to determine HIV coreceptor tropism categorizes virus quasispecies as R5 tropic (≥90% of the viral population use CCR5), X4 tropic (≥90% use CXCR4), or dual/mixed tropic ^[5]. For dual/mixed-tropic specimens, the assay does not distinguish between populations composed of R5X4-tropic viruses or a mixture of R5-tropic, R5X4-tropic, or X4-tropic viruses or all. Clonal analyses of virus populations from ART-naive patients classified as dual/mixed tropic in this study revealed that these patients harbor various combinations of R5-tropic clones with either R5X4-tropic or X4-tropic clones or both. Importantly, R5-tropic clones predominated in the plasma of most patients. In addition, all but one clone analyzed was highly sensitive to aplaviroc inhibition as measured by FCIC₅₀ when tested on cells expressing CCR5 in the absence of CXCR4.

In addition to an increased FCIC₅₀, the inability of a compound to completely inhibit virus replication at very high concentrations (reduction in MPI) has also been proposed as a CCR5 entry inhibitor resistance mechanism ^[11-13]. The clonal aplaviroc MPI values in this study ranged from 85.74 to 99.96% and were similar for R5 and R5X4-tropic clones. Although the clinical significance of MPI values remains to be determined, this range is similar to what was observed in drug-naive patients prior to treatment in two dose-ranging studies of aplaviroc (CCR100136 and CCR102881, unpublished observations).

Further research is needed to determine the clinical utility of tropism testing, in particular for patients whose plasma virus produces a dual/mixed-tropic readout. Studies are also needed to examine the composition of dual/mixed-tropic viral populations in ART-experienced patients. In addition, larger cross-sectional and longitudinal studies are needed to understand virus characteristics at the population and clonal levels in different HIV-1-infected populations and the relationship between tropism change and disease progression in the context of ART; specifically whether or not the detection of CXCR4-using virus in this setting has the same clinical prognosis as seen in natural history studies.

The cost and time associated with performing routine, clonal analyses of HIV-1 coreceptor tropism at the clonal level precludes the widespread use of this assay for clinical management. As such, emerging data examining HIV-1 coreceptor tropism at the population level in both ART-naive and ART-experienced patients should be interpreted with caution, in particular for instances of a dual/mixed tropism readout. The ultimate proof of the utility of a CCR5 entry inhibitor for patients with dual/mixed-tropic virus will be determined by the clinical outcome of treating these patients with a regimen including a CCR5 entry inhibitor. The predominance of CCR5-using viruses in ART-naive patients who tested as dual/mixed tropic at the population level in the present study suggests that CCR5 entry inhibitors may prove effective as part of an appropriate combination regimen in these patients.

Acknowledgements

We are grateful to the patients who participated in this study and to Monogram Biosciences for scientific discussion.

Authors' contributions: D.M.I. contributed to the conception and design of this study, data management and analysis, interpretation of results, and drafting, editing/completion of the paper. H.A.-M. contributed to the analysis of data, interpretation of results, editing/completion of the paper. K.M.K. contributed to the conception and design of this study, data management and analysis, interpretation of results, and editing/completion of the paper. C.C.L. contributed to the analysis of data, interpretation of results, and editing/completion of the paper. J.F.D. contributed to the conception and design of this study, analysis of data, interpretation of results, and editing/completion of the paper.