In many HIV-infected patients, effective combination antiretroviral therapy drives plasma virus load to undetectable levels within a few months . Analysis of viral decay rates initially suggested that prolonged treatment might lead to eradication . However, HIV can persist in several cellular and anatomical reservoirs [3,4]. These include a very stable reservoir composed of latently infected resting CD4 T cells carrying integrated replication-competent HIV [5–7]. The half-life of this reservoir has been estimated to exceed 43 months , indicating that its eradication would take several decades . To explain this remarkable stability, two non-exclusive hypotheses are commonly forwarded. (i) In patients on highly active antiretroviral therapy (HAART) who have undetectable plasma viral load, evidence of persistent, low-level viral replication has been reported [9–13]. The relationship between the pool of latently infected T cells and residual virus replication is a major unsolved issue. Ongoing replication may lead to continual replenishment of the latent reservoir, thereby explaining its stability . (ii) It has been shown that, among resting CD4 T cells, integrated HIV DNA is present primarily among cells with a memory phenotype [4,14]. These cells may be derived from antigen-activated cells that reverted to a resting memory state. These antigen-specific cells harbouring integrated virus may have long-term survival, consistent with the near life-long memory associated with many infections. This may be applicable to HIV-specific CD4 T cells. Accumulating data point to the presence of these cells in HIV-infected patients at several stages of the disease [15–18]. During active infection, HIV-specific CD4 T cells are continuously exposed to HIV antigens and activated, and may therefore constitute a privileged target for HIV [19,20]. Following the marked reduction in HIV replication on effective and prolonged HAART, some of these infected cells may return to the resting state and therefore contribute to the latent virus reservoir.
Here, to investigate antigen specificities of HIV reservoir CD4 T cells, we examined the capacity of recall antigens, including HIV antigens, to induce virus production by peripheral blood mononuclear cells (PBMC) from patients on prolonged effective HAART.
Patients and methods
We selected five HIV-infected patients who were highly adherent to antiretroviral treatment, on the basis of long-term undetectable plasma viral RNA on unmodified HAART. These patients were previously untreated. HAART consisted of two nucleoside inhibitors and a protease inhibitor, with the exception of patient D, who received a combination of zidovudine and lamivudine and efavirenz (Table 1). All of these patients were cytomegalovirus (CMV)-seropositive.
Detection of HIV double long terminal repeat (LTR) circles
Episomal DNA was extracted from 1 × 106 CD8 cell-depleted PBMC (QIAprep Spin Miniprep Kit, Qiagen, Hilden, Germany) and collected in 50 μl of diethylpyrocarbonate-treated water. Fifteen μl of DNA were used with the following primers to detect HIV double LTR: 5′-AGATCTGGTCTAACCAGAG-3′ located in the R region of the LTR and 5′-TAACTAGAGATCCC TCAGAC-3′ located in the U5 region of the LTR. These primers were first tested on 10 HIV-infected patients with detectable or undetectable plasma virus. HIV double LTR circles were amplified in nine of these 10 patients. The corresponding PCR products of approximately 545 base pairs in pCR2.1 (Invitrogen, Groningen, the Netherlands) were cloned and then sequenced. We found them to include the characteristic U3–U5 junction of double LTR circles . A cloned PCR product was diluted from 2 × 109 to 1 copy per PCR and this serial dilution was used as a positive control for PCR with episomal DNA extracted from patients on prolonged HAART.
Activation of cells to virus production
Freshly isolated PBMC were first depleted of CD8 T cells (CD8-negative selection, Miltenyi, Bergisch Gladbach, Germany). CD8 cell-depleted PBMC were serially diluted from 1 × 106 to 1 × 103 cells/0.5 ml per well then activated with phytohaemagglutinin (PHA; 0.25 mg/ml; Murex, Chatillon, France) and interleukin (IL)-2 (25 ng/ml; R & D systems, Abington, Oxon, UK), lipopolysaccharide (LPS; 1 μg/ml), recombinant HIV antigens (5 μg/ml p24, Protein Sciences, Meriden, Connecticut, USA; 1 μg/ml Nef, Intracell, London, UK); CMV (diluted 1/50, Biowhittaker, Walkersville, Maryland, USA) or tuberculin purified protein derivative (PPD) (5 μg/ml; Pasteur-Merieux, Marcy l'Etoile, France). The P24 control was a supernatant of insect cells infected with non-recombinant baculovirus and the CMV control was a supernatant of uninfected cultured human embryonic fibroblasts, both used according to the manufacturer's instructions. We obtained purified CD4 T cells by CD4 positive selection (Miltenyi) then CD14 negative selection (Miltenyi). Cells were cultured for 21 days. Cell activation was performed on days 0, 7 and 14. Cell viability was assessed by the Trypan blue dye exclusion method. No significant difference in cell viability was found between the various activation conditions. Supernatants were tested with the Amplicor monitor test (detection limit 20 copies/ml; Roche diagnostics, Neuilly, France). In such a limiting-dilution assay, what is important is the first cell density at which virus production is no longer detected. Therefore, for each cell dilution and activation condition, cultures were performed in duplicate. We first tested one well of each duplicate. If virus RNA was undetectable the second well was tested before recording a negative result.
We selected five HIV-infected patients, who were highly adherent to antiretroviral treatment, on the basis of long-term undetectable plasma viral RNA on unmodified HAART. These patients were previously untreated. On HAART, viral RNA fell to < 20 copies/ml within a median of 2 months, and remained undetectable at each 3-monthly control (Table 1). In all patients tested, viral DNA was detected in freshly isolated PBMC by PCR amplifying total HIV DNA  (data not shown). To detect residual virus replication, we used a PCR procedure that specifically amplifies double LTR episomal viral DNA. A previous study  had demonstrated the in vitro and in vivo instability of HIV double LTR circles, suggesting their utility as markers of residual viral replication, although this has recently been challenged [21,22]. As shown in Fig. 1, HIV double LTR circles were detected in one of the five patients examined, but not in the other four, despite repeated analysis and the use of an optimized PCR procedure that detects a very small number of HIV double LTR circles per PCR (Fig. 1).
To quantify host CD4 T cells infected by replication-competent virus, and to determine their antigen specificities, we used a limiting dilution-based culture assay. Freshly isolated PBMC were first depleted of CD8 T cells to avoid potential cytotoxic reactions. Virus production was examined after activation of CD8 T cell-depleted PBMC at several cell densities (down to 1 × 103 cells/well) by using the common recall antigens PPD and CMV or by using HIV-1 p24 with and without HIV-1 Nef. As shown in Fig. 2, no detectable virus was found in cultures of untreated cells. This indicated latent virus cell infection. Cell activation with the combination of PHA and IL-2 induced detectable HIV production by cells from all five patients. Virus production was detected down to a cell density of 1 × 105/well in one patient and 1 × 104/well in the remaining four patients (Fig. 2). This indicated the persistence of infected cells at a frequency of at least 1 per 105 CD8-depleted PBMC, even in patients with no evidence of HIV double LTR circles.
Specific cell activation with HIV p24 or with the combination of HIV p24 and Nef led, in four of the five patients, to detectable virus production, down to the same cell densities at which PHA and IL-2 still induced virus production. In the case of patient B, HIV antigens led to virus production down to the cell density of 1 × 105 cells/well, while the combination of PHA and IL-2 still induced detectable virus replication at 1 × 104 cells/well (Fig. 2). These results suggested the presence of at least one HIV-specific cell harbouring replication-competent virus among 1 × 104 CD8-depleted PBMC in patients A, C, D and E, and one in 1 × 105 CD8-depleted PBMC in patient B. In all of the patients tested, activation of CD8-depleted PBMC dilutions with PPD or CMV yielded lower apparent frequencies of PPD- and CMV-specific cells harbouring replication-competent virus (one per 1 × 105 to less than one per 1 × 106). Apparent frequencies of replication-competent HIV-infected CD4 T cells with HIV specificity were therefore between fivefold (patient B) and > 100-fold higher (patient C) than cells with CMV and PPD specificities.
When highly purified CD4 T cells obtained following CD4-positive and CD14-negative selection of PBMC were treated with HIV p24, no detectable viral RNA was found in the supernatants (Fig. 2). This ruled out the possibility that HIV p24 activated CD4 T cells through a superantigen effect (Fig. 2).
The theoretical involvement of a bystander activation led us to question whether the apparent frequencies of HIV-infected cells of a given specificity (HIV, CMV or PPD) may be lower than their real frequencies. Indeed, it has been demonstrated that pro-inflammatory cytokines are able to induce HIV replication in latently infected CD4 T cells . Here, the addition of LPS, a potent inducer of pro-inflammatory cytokine production by monocytes, to CD8-depleted PBMC did not lead to detectable virus production (Fig. 2), despite the presence of 15–30% of monocytes in CD8-depleted PBMC cultures on the basis of CD14 staining (not shown). Moreover, HIV antigens triggered detectable virus production at higher cell dilutions than CMV or PPD, while it has been reported that the frequency of cytokine-producing CMV-specific CD4 T cells is higher than that of HIV-specific cells in patients on prolonged HAART . Altogether, this argues against significant involvement of a bystander process in HIV production by infected cells induced by HIV antigens.
Our findings suggest that a significant proportion of replication-competent HIV-infected CD4 T cells in these HAART-treated patients with long-term undetectable plasma virus may be memory cells directed against HIV determinants.
In four of these five highly selected patients, no episomal double LTR circles were found in freshly isolated PBMC. These viral forms have been proposed as markers of residual virus replication , although this has recently been challenged [21,22]. Nonetheless, the presence of 2-LTR DNA circles reflects infection of previously uninfected cells. Therefore, their absence in these patients argues against the presence of significant residual virus replication. The absence of virus production following LPS treatment of cultures also points to the absence of infected monocytes, which have been described as potential markers of ongoing viral replication in patients on prolonged HAART . Moreover, no viral RNA was detected in supernatants of untreated cell cultures, pointing to latent viral infection. A significant proportion of the infected cells could therefore correspond to the previously described pool of long-lived, resting, latently infected CD4 T cells . These HIV-specific memory cells may have reverted to a resting state following the marked gradual fall in HIV replication and HIV antigen availability on HAART. Cell infection could have occurred during the transition phase from the activated to the resting state, permitting stable viral integration but not permitting virus replication or host cell destruction. These resting reservoir cells are no longer activated in patients on effective HAART, as their likelihood of re-encountering HIV antigens is very low. Overall, this may result in a pool of reservoir cells with HIV specificity that has a very slow decay rate. A similar cell infection process, with stable viral integration but no virus replication, might also occur with CD4 T cells of other antigen specificities, such as the common environmental pathogen CMV. However, the corresponding memory resting cells are more likely to re-encounter their respective antigens in patients on effective HAART, potentially leading to cell activation, virus replication and subsequent cell destruction. This would mean that, in patients on HAART who have no significant residual viral replication, the pool of reservoir cells with HIV specificity decays more slowly than the pool of cells with common environmental antigen specificities. Alternatively, or in addition to this first hypothesis, the reservoir of HIV-specific cells may initially be larger than other reservoirs, owing to preferential infection of HIV-specific cells that may constitute a predominant proportion of activated cells in patients on active HIV infection.
A third, non-exclusive hypothesis to explain the apparent over-accumulation of reservoir cells with HIV specificity is that selective anergy affects HIV-specific CD4 T cells. Anergy of specific CD4 T cells may occur during chronic viral infections . In the case of HIV, this may explain the lack of in vitro proliferative responses to HIV antigens despite the detection of these cells with other procedures [16,25–27], including tetrameric HLA-DR molecules covalently complexed with peptides . Anergy may occur after the initial steps of cell activation, allowing, in case of simultaneous cell infection, stable virus integration to occur. Anergy may lead cells to return to a resting state and also prevent further activation, resulting in no virus replication. Anergy may be due to various mechanisms, including over-expression of inhibitory molecules such as CTLA4 . An inhibitory role of CD8 T cell subpopulations on HIV-specific CD4 T cell activation cannot be ruled out, as shown in other models of chronic immune activation such as transplantation models [30,31]. In this regard, it should be noted that PBMC were depleted of CD8 T cells in our experiments.
Altogether, our results show that HIV antigens can induce marked HIV replication by latently infected cells from patients on prolonged HAART. In most cases, HIV antigens were as efficient in inducing virus production as was the potent combination of PHA with high concentrations of recombinant IL-2.
This may provide a rationale for the use of recombinant HIV antigens to activate latently infected cells and thereby to reduce this reservoir, as it is likely that infected cells die following virus activation, while HAART inhibits new cell infection. Specific immunotherapy based on recombinant HIV antigens might have fewer adverse effects than cytokine therapy. However, several issues remain to be clarified, including the immunologic mechanisms of reservoir cell formation, and those involved in the residual HIV replication observed in some patients.
We thank S. Wain-Hobson for critical reading of the manuscript.
1. Hammer SM, Squires KE, Hughes MD. et al
. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team.
N Engl J Med 1997, 337: 725–733.
2. Perelson AS, Essunger P, Cao Y. et al
. Decay characteristics of HIV-1-infected compartments during combination therapy. Nature 1997, 387: 188–191.
3. Schrager LK, D'Souza MP. Cellular and anatomical reservoirs of HIV-1 in patients receiving potent antiretroviral combination therapy. JAMA 1998, 280: 67–71.
4. Pierson T, McArthur J, Siliciano RF. Reservoirs for HIV-1: mechanisms for viral persistence in the presence of antiviral immune responses and antiretroviral therapy. Annu Rev Immunol 2000, 18: 665–708.
5. Wong JK, Hezareh M, Gunthard HF. et al
. Recovery of replication-competent HIV despite prolonged suppression of plasma viremia. Science 1997, 278: 1291–1295.
6. Finzi D, Hermankova M, Pierson T. et al
. Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy. Science 1997, 278: 1295–300.
7. Chun TW, Stuyver L, Mizell SB. Presence of an inducible HIV-1 latent reservoir during highly active antiretroviral therapy. Proc Natl Acad Sci USA 1997, 94: 13193–13197.
8. Finzi D, Blankson J, Siliciano JD. et al
. Latent infection of CD4+ T cells provides a mechanism for lifelong persistence of HIV-1, even in patients on effective combination therapy. Nature Med 1999, 5: 512–517.
9. Dornadula G, Zhang H, VanUitert B. et al
. Residual HIV-1 RNA in blood plasma of patients taking suppressive highly active antiretroviral therapy. JAMA 1999, 282: 1627–1632.
10. Sharkey ME, Teo I, Greenough T. et al
. Persistence of episomal HIV-1 infection intermediates in patients on highly active anti-retroviral therapy. Nature Med 2000, 6: 76–81.
11. Furtado MR, Callaway DS, Phair JP. et al
. Persistence of HIV-1 transcription in peripheral blood mononuclear cells in patients receiving potent antiretroviral therapy. N Engl J Med 1999, 340: 1614–1622.
12. Zhang L, Ramratnam B, Tenner-Racz K. et al
. Quantifying residual HIV-1 replication in patients receiving combination antiretroviral therapy. N Engl J Med 1999, 340: 1605–1613.
13. Lambotte O, Taoufik Y, de Goer MG, Wallon C, Goujard C, Delfraissy JF. Detection of infectious HIV in circulating monocytes from patients on prolonged highly active antiretroviral therapy. J Acquir Immune Defic Syndr 2000, 23: 114–119.
14. Chun TW, Carruth L, Finzi D. et al
. Quantification of latent tissue reservoirs and total body viral load in HIV-1 infection. Nature 1997, 387: 183–188.
15. Rosenberg ES, Billingsley JM, Caliendo AM. et al
. Vigorous HIV-1-specific CD4+ T cell responses associated with control of viremia. Science 1997, 278: 1447–1450.
16. Pitcher CJ, Quittner C, Peterson DM. et al
. HIV-1-specific CD4+ T cells are detectable in most individuals with active HIV-1 infection, but decline with prolonged viral suppression. Nature Med 1999, 5: 518–525.
17. Oxenius A, Price DA, Easterbrook PJ. et al
. Early highly active antiretroviral therapy for acute HIV-1 infection preserves immune function of CD8+ and CD4+ T lymphocytes. Proc Natl Acad Sci USA 2000, 97: 3382–3387.
18. Blankson JN, Gallant JE, Siliciano RF. Proliferative responses to human immunodeficiency virus type 1 (HIV-1) antigens in HIV-1-infected patients with immune reconstitution. J Infect Dis 2001, 183: 657–661.
19. Veazey RS, Tham IC, Mansfield KG. et al
. Identifying the target cell in primary simian immunodeficiency virus (SIV) infection: highly activated memory CD4(+) T cells are rapidly eliminated in early SIV infection in vivo. J Virol 2000, 74: 57–64.
20. Schnittman SM, Lane HC, Greenhouse J, Justement JS, Baseler M, Fauci AS. Preferential infection of CD4+ memory T cells by human immunodeficiency virus type 1: evidence for a role in the selective T-cell functional defects observed in infected individuals. Proc Natl Acad Sci USA 1990, 87: 6058–6062.
21. Pierson TC, Kieffer TL, Ruff CT, Buck C, Gange SJ, Siliciano RF. Intrinsic stability of episomal circles formed during human immunodeficiency virus type 1 replication. J Virol 2002, 76: 4138–4144.
22. Butler SL, Johnson EP, Bushman FD. Human immunodeficiency virus cDNA metabolism: notable stability of two-long terminal repeat circles. J Virol 2002, 76: 3739–3747.
23. Chun TW, Engel D, Mizell SB, Ehler LA, Fauci AS. Induction of HIV-1 replication in latently infected CD4+ T cells using a combination of cytokines. J Exp Med 1998, 188: 83–91.
24. Zajac AJ, Blattman JN, Murali-Krishna K. et al
. Viral immune evasion due to persistence of activated T cells without effector function. J Exp Med 1998, 188: 2205–2213.
25. Picker LJ, Maino VC. The CD4(+) T cell response to HIV-1. Curr Opin Immunol 2000, 12: 381–386.
26. Wilson JD, Imami N, Watkins A. et al
. Loss of CD4+ T cell proliferative ability but not loss of human immunodeficiency virus type 1 specificity
equates with progression to disease. J Infect Dis 2000, 182: 792–798.
27. Dybul M., Mercier G, Belson M. et al
. CD40 ligand trimer and IL-12 enhance peripheral blood mononuclear cells and CD4+ T cell proliferation and production of IFN-gamma in response to p24 antigen in HIV-infected individuals: potential contribution of anergy to HIV-specific unresponsiveness. J Immunol 2000, 165: 1685–1691.
28. Yassine Diab B, Younes S, Breton G, et al
. Identification of HIV-specific CD4+ T cells in peripheral blood of HIV-1 infected individuals by tetrameric HLA-DR molecules covalently complexed with peptides. First IAS Conference on HIV Pathogenesis and AIDS
. Buenos Aires, July 2001 [abstract 82:11].
29. Leng Q, Bentwich Z, Magen E, Kalinkovich A, Borkow G. CTLA-4 upregulation during HIV infection: association with anergy and possible target for immune reconstitution. AIDS 2002, 16: 519 –29 .
30. Colovai AI, Liu Z, Ciubotariu R, Lederman S, Cortesini R, Suciu-Foca N. Induction of xenoreactive CD4+ T cell anergy by suppressor CD8+ CD28- T cells. Transplantation 2001, 69: 1304–1310.
31. Ciubotariu R, Vasilescu R, Ho E. et al
. Detection of T suppressor cells in patients with organ allografts. Hum Immunol 2001, 62: 15–20.