A dozen of G-protein-coupled receptors can mediate cell entry of HIV-1 in vitro but only CCR5 and CXCR4 chemokine receptors seem to be actually used by the virus as coreceptors in vivo, in addition to the CD4 receptor.1 CCR5-using viruses are designated R5, CXCR4-using viruses X4, and viruses able to use both coreceptors R5X4.2 Over the course of HIV-1 infection, the coreceptor usage changes from a preference for CCR5 to a preference for CXCR4 in half to one-third of the patients.3 As few as 1-3 amino acid changes in the V3 loop of the HIV-1 gp120 molecule may be sufficient to convert an R5 virus into an X4 variant.4-6 Surprisingly, despite the high mutational rate of HIV-1 genome,7 this R5 to X4 switch usually occurs, if it occurs, only after 5-10 years of infection. It is associated with an accelarated CD4+ T-cell decline and a progression to AIDS and death.8 Thus, whereas R5 virus can be detected at all stages of HIV-1 infection, X4 virus are generally isolated from some patients with late-stage disease. The reasons for the emergence of X4 strains in some patients and not in others after several years of infection remain poorly understood. As the emergence of X4 strains is linked to an increase in disease progression, it is of major importance to understand the host factors that might induce R5 to X4 switch.
One of these host factors might be the level of expression of the HIV-1 coreceptors at the surface of CD4+ T cells. We have previously reported that the mean number of CCR5 molecules at the surface of CD4+ peripheral blood T lymphocytes (CCR5 density) varies from one individual to another, but is stable over time for a given individual.9 We have also shown that this CCR5 density correlates with viral load, disease progression, and response to treatment in HIV-1-infected persons.9-11 Interestingly, this correlation is logarithmic, a small difference in CCR5 expression resulting in a marked difference in HIV-1 RNA plasma level and in CD4+ T-cell loss.9,11 Various authors argue in favour of a competition between R5 and X4 strains, so that a decline in R5 production would result in an increase in X4 production. If this is true, then a low expression of CCR5 should facilitate the expansion of X4 strains. Patients heterozygous for the Δ32 deletion in the CCR5 gene that results in an allele encoding a trunkated protein that is not expressed at the cell surface, present with low CCR5 densities, but the data on their tendency to develop X4 strains are conflicting. Whereas de Roda Husman et al12 reported a delay in R5 to X4 switch in these patients, other authors observed that the X4 phenotype was more frequent in CCR5-Δ32 heterozygotes than in wild-type CCR5 homozygotes.13-15 In vitro, various reports show that blocking CCR5 with a natural ligand,16 an antibody,17 or a synthetic antagonist18 induced the emergence of mutant R5 HIV-1 resistant to the inhibitor rather than virions capable of using X4 as a coreceptor. By contrast, Mosier et al19 provoked the emergence of X4 strains in hu-PBL-SCID mice infected with an R5 strain and treated with a CCR5 antagonist derived from the natural ligand CCL5. Yet, in a primate model, the administration of a CCR5 inhibitor to macaques dually infected with an R5 simian immunodeficiency virus and an X4 Simian HIV (SHIV) did not result in a sustained increase in X4 viral load.20 Thus, these conflicting data do not allow to answer this critical question about the role of the level of CCR5 expression in the emergence of X4 strains, at a moment where CCR5 antagonists are used as anti-HIV therapeutic molecules. Of note, preliminary communications have reported the occasional emergence of X4 strains after monotherapy with CCR5 antagonists.21
In addition to CCR5 density, the level of expression of CXCR4 at the surface of CD4+ T cells might also influence the risk of emergence of X4 strains. We have previously reported an increase in CD4+ T-cell surface CXCR4 density in one third of patients presenting with CD4 counts below 400 cells per microliter. Of note, these patients with higher CXCR4 densities harboured X4 strains more frequently than patients with physiological levels of CXCR4 expression.22 It is therefore tempting to speculate that the overexpression of CXCR4 in the course of the disease of some patients might favour the development of X4 strains.
For these reasons, we tested the hypothesis that a low CCR5 density and/or a high CXCR4 density at the surface of CD4+ T cells could be host factors paving the way for an R5 to X4 switch. To this aim, we compared CCR5 densities between patients harboring or not X4 strains and looked for the effect of CCR5 and CXCR4 densities in an in vitro model of coreceptor commutation.
The Presence of X4 Strains in Patients is not Associated With a Low CCR5 Density
If a low CCR5 density favours the emergence of X4 strains, then subjects harboring CXCR4-using strains should present with a lower CCR5 expression than subjects who do not. To test this eventuality, we measured by quantitative flow cytometry the mean number of CCR5 molecules at the surface of peripheral blood CD4+ T cells of 2 groups of HIV-1-infected individuals in whom X4 strains had (6 women and 18 men, mean age of 37 years; range 25-58 years) or had not (11 women and 32 men, mean age of 41 years; range 21-66 years) been isolated. The mean CD4 counts in each group were 168 [95% confidence interval (CI): 122 to 214] and 174 (95% CI: 144 to 203) cells per microliter (P = 0.89), and viral loads 191,756 (95% CI: 55,727 to 327,786) and 215,690 (95% CI: 130,906 to 300,394) copies per milliliter (P = 0.56), respectively. Figure 1 shows that CCR5 densities were similar in the 2 groups [9035 (95% CI: 7358 to 10713) and 9218 (95% CI: 7947 to 10490) CCR5 molecules per CD4+ T cell, respectively, P = 0.79].
Low CD4+ T-Cell Surface CCR5 Density Does not Favour the Emergence of X4 Strains In Vitro
To further study the putative influence of CCR5 density on the emergence of X4 strains, we established 2 cell lines that differed only by their surface CCR5 density. For this purpose, we transduced a CD4+CXCR4+ Jurkat cell line with HIV vectors delivering the CCR5 gene driven either by a phosphoglycerate kinase (PGK) or an elongation factor 1 (EF1)α promoter. Thus, we obtained 2 sublines expressing the same surface CXCR4 density (Fig. 2A), but 2 different surface CCR5 densities, 24,532 molecules/cell (Jk-CCR5lowcells), and 190,479 molecules/cell (Jk-CCR5high cells), respectively (Fig. 2A). We compared R5 infectibility of these 2 sublines, which differed only in their level of CCR5 expression. For this purpose, we first exposed them to defective virions obtained by cotransfecting 293T cells with a plasmid encoding an R5 envelope together with a plasmid containing the HIV-1 genome deleted in the env gene and a luciferase reporter gene fused to nef. As shown in Figure 2B, single-round R5 infection obtained using these virions was more efficient in Jk-CCR5high cells than in Jk-CCR5lowcells (Student t test = 3.09, P = 0.04). Then, we infected both cell lines with the wild-type R5 strain AD8.23 Here again, AD8 replicated more intensively in the cell subline that expresses the highest cell surface CCR5 density (Fig. 2C). Finally, to study R5 to X4 switch in these 2 cell subpopulations, we exposed them to 175 ng of p24 equivalents of AD8. We used R5 virus issued from a molecular clone to ascertain that the inoculum was homogenous and devoid of X4 virions. To monitor the emergence of X4 strains, we inoculated daily the culture supernatant into a CD4+CCR5−CXCR4+ MT2 cell culture and looked for the appearance of p24 antigen at day 7 and of syncitia at day 14. As shown in Fig. 2D, the R5 to X4 switch occurred at the same time (13 ± 0 days) in both cell cultures. Thus, a low level of surface CCR5 expression does not favour the development of CXCR4-using strains in vitro.
High CD4+ T-Cell Surface CXCR4 Density Favours the Infectibility by X4 Strains In Vitro
We have previously shown that an increase in CXCR4 density occurs at the surface of CD4+ T cells of about one third of HIV-infected subjects with CD4 counts below 400 cells per microliter and that this CXCR4 overexpression is correlated with the presence of X4 strains.22 As X4 production does not induce CXCR4 expression, neither in vitro nor in vivo,22 we wanted to test the hypothesis that CXCR4 overexpression could induce a change in coreceptor usage. First, we questioned whether an increase in CXCR4 density results in an increase in cell susceptibility to be productively infected by an X4 strain. To this aim, we established 2 cell lines expressing different surface CXCR4 densities by transducing CD4+CXCR4+ MT2 cells with feline immunodeficiency virus (FIV) vectors delivering either the CXCR4 gene or the EGFP gene as a negative control. We obtained 2 sublines, MT2-CXCR4high and MT2-CXCR4low that displayed a similar number of surface CD4 molecules but different number of surface CXCR4 molecules (5 × 104 and 11 × 104 molecules/cell, respectively, Fig. 3A). Then, we infected both cell lines with 500 ng of p24 equivalents of the X4 strain NL4.324 and monitored viral production over time. We observed that virus production was earlier and higher in MT2-CXCR4high cells than in MT2-CXCR4low cells (Fig. 3B). Likewise, the TCID50 was 4-fold higher in MT2-CXCR4high cells than in MT2-CXCR4low cells, 4 ng vs 1 ng of p24 equivalents of NL4.3 virus, respectively (data not shown).
We next studied if a difference in CXCR4 expression could be responsible for a difference in susceptibility to produce X4 strains in the course of an R5 infection. For this purpose, we derived 2 sublines, HOS-CXCR4high and HOS-CXCR4low, expressing the same surface CCR5 density (Fig. 4A), but 2 different surface CXCR4 densities, 20772 and 4487 molecules/cell, respectively, by transducing a CD4+CCR5+ HOS cell line with HIV vectors delivering the CXCR4-EGFP gene or the EGFP gene alone as a negative control (Fig. 4A). To compare the X4 infectibility of these sublines, we exposed them to defective virions obtained by cotransfecting 293T cells with a plasmid encoding an X4 envelope together with a plasmid containing the HIV-1 genome deleted in the env gene and a luciferase reporter gene fused to nef. As shown in Figure 4B, single X4 life cycles were more efficient in HOS-CXCR4high cells than in HOS-CXCR4low cells. Then, we infected both cell lines with the R5 strain AD8 (1μg p24 equivalents, ie, 100 TCID50), and monitored the emergence of X4 strains. As shown in Figure 4C, the R5 to X4 switch occured earlier in the HOS-CXCR4high cells than in the HOS-CXCR4low cells (12 ± 0 vs 21 ± 0 days, respectively, P = 0.01). The same result was obtained in 3 independent experiments performed by 3 different experimenters and with 2 primary R5 strains (Fig. 4C). Thus, a high level of surface CXCR4 expression favours the development of CXCR4-using strains in vitro.
In this study, we explored the influence of the level of expression of each coreceptor on the emergence of CXCR4-using HIV-1 strains.
We tested the hypothesis that a low CD4+ T-cell surface CCR5 density could favour such an event. This hypothesis is based on the idea that the expansion of R5 strains prevents the expansion of X4 strains, so that the level of CCR5 expression, which determines the level of R5 virus production, could modulate the risk of developing X4 strains. Yet, we found no argument in favour of this hypothesis. HIV-1-infected subjects harboring X4 strains did not express low CD4+ T-cell surface CCR5 densities, and, in vitro, the level of CCR5 expression did not correlate with the delay of appearance of X4 strains after infection with an R5 strain. Actually, the level of CCR5 cell surface expression could have opposite effects on the risk of coreceptor switch. On one hand, a high CCR5 expression could inhibit the expansion of X4 strains. On the other hand, by enhancing R5 replication, a high CCR5 expression could generate more mutant virus, some of them with the capacity to use CXCR4 as a coreceptor. These data are reassuring concerning the risk of inducing the emergence of X4 strains in the context of the usage of CCR5 antagonists.
By contrast, our results are compatible with the hypothesis that a high CD4+ T-cell surface CXCR4 density could facilitate the emergence of X4 strains in individuals infected with R5 strains. First, we observed that the level of expression of CXCR4 at the surface of a cell line determines the infectibility and the productivity of this cell line by an X4 strain. This result fits with the previous reports that the infectibility of T cells25,26 and thymocytes27 by X4 strains is linked to their level of surface CXCR4 expression. It is also in keeping with the observations we have previously reported concerning the link between CCR5 expression and R5 infectibility.28 Second, we observed that the level of expression of CXCR4 at the surface of a cell line determines the delay necessary for X4 strains to expand in the course of an R5 infection. X4 strains are already present in reservoir cells of patients infected by R5 strains2,29 and/or are produced frequently secondary to the random mutations that occur at high rate in the HIV-1 genome. Therefore, the emergence of X4 strains might mainly depend on the capacity of the organism to sustain X4 replication. As CXCR4+ T cells express less surface CXCR4 molecules than CCR5+ T cells express surface CCR5 molecules,22 and as X4 infection of CD4+ T cells is less efficient than R5 infection,30 in the first stages of the infection X4 expansion might be hampered. Yet, below 400 CD4+ T cells per microliter, in some subjects, some factors could induce an increase in CXCR4 expression on CD4+ T allowing then the replication of X4 strains. If this hypothesis is true, it is important to discover the events that induce CXCR4 overexpression in vivo to prevent them. Moreover, the monitoring of CXCR4 density at the surface of peripheral blood CD4+ T cells by routine flow cytometry could identify the subjects at risk for coreceptor switch.
We recruited 67 HIV-1-infected adults presenting with CD4 counts below 400 CD4+ T cells per microliter at the Department of Infectious Diseases of the University Hospital of Montpellier.
CD4+CCR5+ HOS cells are human osteosarcoma cells. They were grown in DMEM medium supplemented with 10% fetal calf serum, 10 mM of glutamax-1,100 U/ml of penicillin and 100 μg/mL of streptomycin (Invitrogen, Cergy pontoise, France) at 37°C and 5% CO2. MT2 cells are a cell line transformed by Human T-cell leukemia virus type 1 (HTLV-1). MT2 cells, Jurkat cells, and peripheral blood mononuclear cells (PBMC) were grown in RPMI-1640 medium supplemented with 10% fetal calf serum, 10 mM of glutamax-1100 U/mL of penicillin and 100 μg/mL of streptomycin (Invitrogen, Cergy pontoise, France) at 37°C and 5% CO2.
Vector production and cell transduction were performed as described elsewhere.28 To generate 2 sublines expressing different cell surface CXCR4 densities, we transduced CD4+CCR5+CXCR4+ HOS cells with an HIV vector delivering the CXCR4 gene fused to the EGFP gene (HOS-CXCR4high cells) or an HIV vector delivering the EGFP gene alone (HOS-CXCR4low cells). To generate 2 sublines expressing different cell surface CCR5 densities, we transduced CD4+CXCR4+ Jurkat cells with HIV vectors delivering the CCR5 gene driven either by a PGK (Jk-CCR5low cells) or an EF1α (Jk-CCR5high cells) promoter. To generate MT2-CXCR4high and MT2-CXCR4low sublines, we transduced CD4+CXCR4+ MT2 cells with FIV vectors delivering the CXCR4 gene or the EGFP gene, respectively.
CXCR4 and CCR5 Phenotyping
CXCR4 and CCR5 densities at the surface of DR−CD4+ T cells was measured by a quantitative flow cytometry assay as described previously.31 Briefly, cells were indirectly labeled with anti-CXCR4 antibody 12G5 (Pharmingen, San Diego, CA) or IgG2b isotype control (Beckman-Coulter, Margency, France), with anti-CCR5 antibody 2D7 (Pharmingen) or IgG2a isotype control (Beckman-Coulter) and fluoresceine isothiocyanate (FITC)-conjugated F(ab')2 fragment goat anti-mouse Ig (H+L) (Jackson, West Grove, PA) and directly labeled with phycoerythrine-conjugated anti-CD4 antibody (Beckman-Coulter). CXCR4 and CCR5 expression were analyzed after gating on DR−CD4+ T cells. CXCR4 and CCR5 densities were calculated by converting FITC fluorescence intensity into antibody-binding capacity, which corresponds, at the saturating concentrations we used, to the number of CXCR4 or CCR5 molecules present on the cell surface. Fluorescence intensity was converted into CXCR4 and CCR5 densities after a calibration curve obtained with standard microbeads (DAKO QIFIKIT, Glostrup, Denmark) precoated by the manufacturer with various densities of MAb and subsequently labeled with the FITC-conjugated anti-Ig probe. Intraassay variability was below 5%.
X4 phenotype was determined using the MT2 assay.32 Briefly, PBMC were depleted in CD8+ T cells using anti-CD8 antibody-coated magnetic beads (Dynal Biotech, S.A., Compiègne, France), cocultured with phytohemagglutinin-stimulated healthy donor PBMC. CD4+CCR5−CXCR4+ MT2 cells cultivated at 0,25 × 106 cells per milliliter were inoculated with 150 μL of coculture supernatant containing 400 pg/mL of HIV-1 p24 gag antigen, and monitored during 14 days for the presence of syncitia. Moreover, the presence of CXCR4-using virions was confirmed by measuring HIV-1 p24 gag antigen in MT2 culture supernatant at day 7. To avoid false positive result due to virus carryover, MT2 cells were trypsinized and extensively washed 24 hour after exposure to the cell supernatants. The X4 strain NL4.3 and the R5 strain AD8 were used as positive and negative controls in each experiment. We have verified each time by sequencing that the X4 virions emerging in R5-infected cultures did not correspond to contaminating NL4.3 but to mutated R5 strains.
Cells were plated in 24-well plates (0,25 × 106 cells per well for HOS and Jurkat cells, 0,5 × 106 cells per wells for MT2 cells), exposed for 24 hours to the virus, washed extensively, and cultured at the initial concentration. Virus production was monitored by measuring p24 concentration in the cell supernatant by using a commercial enzyme-linked immunosorbent assay kit (Beckman-Coulter).
One-Round Infection Assay
To produce replication-defective HIV virions, 293T cells were cotransfected with the pNL4.3env-luc+ (AIDS Reagent Program) plasmid on one hand, and with the pCMV. AD8-Env plasmid encoding the R5 envelope of HIV-1 prototype AD8 or the pCMV NL4.3-Env plasmid encoding the X4 envelope of HIV-1 prototype NL4.3. For the 1-round infection assay, 8 × 104 HOS cells were plated per well in a 24-well plate and exposed to AD8-pseudotyped or NL4.3-pseudotyped defective virions for 48 hours. Luciferase activity was measured by luminometry at the end of the culture, by using the Luciferase Assay System (Promega, Charbonnières-les-Bains, France).
CXCR4 and CCR5 expression on CD4+ peripheral blood T cells from infected individuals harboring or not X4 strains was compared by the Mann-Whitney U test. Differences in the infectivity and in the kinetics of R5 to X4 switch were analysed using the Student t test.
HOS-CD4+CCR5+ and Jurkat-CD4+CCR5+ cells were provided by the EU programme EVA/MRC centralised Facility for AIDS Reagents, NIBSC, United Kingdom (grant number QLK2-CT-1999-00609 and GP828102).
1. Berger EA. HIV entry and tropism: the chemokine receptor connection. AIDS Res Hum Retroviruses
2. Moore JP, Kitchen SG, Pugach P, et al. The CCR5 and CXCR4 coreceptors-central to understanding the transmission and pathogenesis of human immunodeficiency virus type 1 infection. AIDS Res Hum Retroviruses
3. Regoes RR, Bonhoeffer S. The HIV coreceptor swich: a population dynamical perspective. Trends Microbiol
4. De Jong J, De Ronde A, Keulen W, et al. Minimal requirements for the human immunodeficiency virus type 1 V3 domain to support the syncytium-inducing phenotype: analysis by single amino acid substitution. J Virol
5. Fouchier R, Groenink M, Kootstra NA, et al. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gp120 molecule. J Virol
6. Shioda T, Levy JA, Cheng-Mayer C. Small amino acid changes in the V3 hypervariable region of gp120 can affect the T-cell-line and macrophage tropism of human immunodeficiency virus type 1. Proc Natl Acad Sci U S A
7. Coffin JM. HIV population dynamics in vivo: implications for genetic variation, pathogenesis, and therapy. Science
8. Berger EA, Murphy PM, Farber JM. Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu Rev Immunol
9. Reynes J, Portales P, Segondy M, et al. CD4 T cell surface CCR5 density as a host factor in HIV-1 disease progression. AIDS
10. Gervaix A, Nicolas J, Portales P, et al. Response to treatment and disease progression linked to CD4+ T cell surface CCR5 density in HIV-1 vertical infection. J Infect Dis
11. Reynes J, Portales P, Segondy M, et al. CD4+ T cell surface CCR5 density as a determining factor of viral load in HIV-1-infected individuals. J Infect Dis
12. de Roda Husman AM, Koot M, Cornelissen M, et al. Association between CCR5 genotype and the clinical course of HIV-1 infection. Ann Intern Med
13. Brumme ZL, Goodrich J, Mayer HB, et al. Molecular and clinical epidemiology of CXCR4-using HIV-1 in a large population of antiretroviral-naïve individuals. J Infect Dis
14. D'aquila RT, Sutton L, Savars A, et al. CCR5/delta(ccr5) heterozygosity: a selective pressure for the syncitium-inducing human immunodeficiency virus type 1 phenotype. J Infect Dis
15. Zhang L, He T, Huang Y, et al. Chemokine receptor usage by diverse primary isolates of human immunodeficiency virus type 1. J Virol
16. Maeda Y, Foda M, Matsushita S, et al. Involvement of both the V2 and V3 regions of the CCR5-tropic human immunodeficiency virus type 1 envelope in reduced sensitivity to macrophage inflammatory protein 1alpha. J Virol
17. Aarons EJ, Beddows S, Willingham T, et al. Adaptation to blockade of human immunodeficiency virus type 1 entry imposed by the anti-CCR5 monoclonal antibody 2D7. Virology
18. Trkola A, Kuhmann SE, Strizki JM, et al. HIV-1 escape from a small molecule, CCR5-specific entry inhibitor does not involve CXCR4 use. Proc Natl Acad Sci U S A
19. Mosier DE, Picchio GR, Gulizia RJ, et al. Highly potent RANTES analogues either prevent CCR5-using HIV-1 infection in vivo or rapidly select for CXCR4-using variants. J Virol
20. Wolinsky SM, Veazy RS, Kunstman KJ, et al. Effect of a CCR5 inhibitor on viral loads in macaques dual-infected with R5 and X4 primate immunodeficiency viruses. Virology
21. Lewis ME, Van der Ryst E, Youle M, et al. Phylogenetic analysis and co-receptor tropism of HIV-1 envelope sequences from patients with emergence of CXCR4 using virus following treatment with the CCR5 antagonist UK-427,857. Presented at: 44th Interscience Conference on Antimicrobial Agents and Chemotherapy; October 30-November 2, 2004; Washington, DC. Late-Breaker Abstract H-548b. 2004.
22. Lin Y-L, Portales P, Baillat V, et al. CXCR4 overexpression during the course of HIV-1 infection correlates with the emergence of X4 strains. J Acquir Immune Defic Syndr
23. Theodore TS, Englund G, Buckler-White A, et al. Construction and characterization of a stable full-length macrophage-tropic HIV type 1 molecular clone that directs the production of high titers of progeny virions. AIDS Res Hum Retroviruses
24. Adachi AG, Gendelman HE, Koenig S, et al. Production of acquired immunodeficiency syndrome-associated retrovirus in human and non-human cells transfected with an infectious clone. J Virol
25. Bermejo M, Martin-Serrano J, Oberlin E, et al. Activation of blood T lymphocytes down-regulates CXCR4 expression and interferes with propagation of X4 strains. Eur J Immunol
26. Moonis M, Lee B, Bailer RT, et al. CCR5 and CXCR4 expression correlated with X4 and R5 HIV-1 infection yet sustained replication in TH1 and TH2 cells. AIDS Res Hum Retroviruses
27. Zaitseva M, Lee S, Rabin RL, et al. CXCR4 and CCR5 on human thymocytes: biological function and role in HIV-1 infection. J Immunol
28. Lin YL, Mettling C, Portales P, et al. Cell surface CCR5 density determines the postentry efficiency of R5 HIV-1 infection. Proc Natl Acad Sci U S A
29. Shankarappa R, Margolick JB, Gange SJ, et al. Consistent viral evolutionary changes associated with the progression of human immunodeficiency virus type 1 infection. J Virol
30. Vicenzi E, Panina Bordignon P, Biswas P, et al. Envelope-dependent restriction of Human Immunodeficiency Virus type 1 spreading in CD4+ T lymphocytes: R5 but not X4 viruses replicate in the absence of T-cell receptor restimulation. J Virol
31. Reynes J, Portales P, Clot J, et al. CD4+ T cell surface CCR5 density and virus load in persons infected with Human Immunodeficiency Virus type 1. J Infect Dis
32. Koot M, Keet RP, Voos AH, et al. Pronostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to AIDS. Ann Intern Med