B. Autran

Restoration of the immune system with anti-retroviral therapy

Clinical benefits of highly active anti-retroviral treatments (HAART) are increasingly evidenced by resolving opportunistic infections and malignancies, as well as declining hospitalization and mortality rates [1]. This suggests that potent and sustained suppression of viral replication, at least to some extent, is associated with reconstitution of the immune system even in adult patients treated at advanced stages of the disease. Increased susceptibility to opportunistic infections and tumors mainly results from the loss of memory CD4+ T cell reactivity against recall antigens which is an early event in HIV disease progression. Primary responses of naive CD4+ T cells against new pathogens are suppressed even earlier in the course of HIV disease, and the progressive depletion in naive CD4+ T cells reflects profound alterations in T cell regeneration capacities. Previous studies revealed that monotherapy with ritonavir, a protease inhibitor, resulted in a slight improvement in memory CD4+ T cell responses to recall Ags only when detectable prior to onset of therapy, suggesting that the loss of CD4+ T cell reactivity might be irreversible at advanced stages of the disease [2]. In contrast our group demonstrated more recently that restoration in CD4+ T cell reactivity to specific antigens was feasible when HAART was administered in progressors [3]. Here we address some of the questions raised by immune restoration with HAART when administered at advanced stages of the disease.

Dynamics of HIV-Specific CD8+ T Lymphocytes with Changes in Viral Load

The influence of HIV burden variations on the frequencies of Ag-specific CD8+ T cell responses was evaluated before and during highly active antiretroviral therapy by analyzing the number, diversity, and function of these cells. The frequencies of HLA-A2-restricted CD8+ PBL binding HLA-A2/HIV-epitope tetramers or producing IFN-gamma were below 1%. A panel of 16 CTL epitopes covering 15 HLA class I molecules in 14 patients allowed us to test 3.8 epitopes/patient and to detect 2.2 +/- 1.8 HIV epitope-specific CD8+ subsets per patient with a median frequency of 0.24% (0.11-4. 79%). During the first month of treatment, viral load rapidly decreased and frequencies of HIV-specific CD8 PBL tripled, eight new HIV specificities appeared of 11 undetectable at entry, while CMV-specific CD8+ PBL also appeared. With efficient HIV load control, all HIV specificities decayed involving a reduction of the CD8+CD27+CD11ahigh HIV-specific effector subset. Virus rebounds triggered by scheduled drug interruptions or transient therapeutic failures induced four patterns of epitope-specific CD8+ lymphocyte dynamics, i.e., peaks or disappearance of preexisting specificities, emergence of new specificities, or lack of changes. The HIV load rebounds mobilized both effector/memory HIV- and CMV-specific CD8+ lymphocytes. Therefore, frequencies of virus-specific CD8 T cells appear to be positively correlated to HIV production in most cases during highly active antiretroviral therapy, but an inverse correlation can also be observed with rapid virus changes that might involve redistribution, sequestration, or expansion of these Ag-specific CD8 T cells. Future strategies of therapeutic interruptions should take into account these various HIV-specific cell dynamics during HIV rebounds.The influence of HIV burden variations on the frequencies of Ag-specific CD8+ T cell responses was evaluated before and during highly active antiretroviral therapy by analyzing the number, diversity, and function of these cells. The frequencies of HLA-A2-restricted CD8+ PBL binding HLA-A2/HIV-epitope tetramers or producing IFN-γ were below 1%. A panel of 16 CTL epitopes covering 15 HLA class I molecules in 14 patients allowed us to test 3.8 epitopes/patient and to detect 2.2 ± 1.8 HIV epitope-specific CD8+ subsets per patient with a median frequency of 0.24% (0.11–4.79%). During the first month of treatment, viral load rapidly decreased and frequencies of HIV-specific CD8 PBL tripled, eight new HIV specificities appeared of 11 undetectable at entry, while CMV-specific CD8+ PBL also appeared. With efficient HIV load control, all HIV specificities decayed involving a reduction of the CD8+CD27+CD11ahigh HIV-specific effector subset. Virus rebounds triggered by scheduled drug interruptions or transient therapeutic failures induced four patterns of epitope-specific CD8+ lymphocyte dynamics, i.e., peaks or disappearance of preexisting specificities, emergence of new specificities, or lack of changes. The HIV load rebounds mobilized both effector/memory HIV- and CMV-specific CD8+ lymphocytes. Therefore, frequencies of virus-specific CD8 T cells appear to be positively correlated to HIV production in most cases during highly active antiretroviral therapy, but an inverse correlation can also be observed with rapid virus changes that might involve redistribution, sequestration, or expansion of these Ag-specific CD8 T cells. Future strategies of therapeutic interruptions should take into account these various HIV-specific cell dynamics during HIV rebounds.

Monocyte-derived dendritic cells have a phenotype comparable to that of dermal dendritic cells and display ultrastructural granules distinct from Birbeck granules.

Most monocyte‐derived dendritic cells (DC) display CD1a, like Langerhans cells (LC) and some dermal DC, but their relationship with these skin DC remains unclear. To address this issue, we studied the expression of different antigens characteristic of skin DC and of monocyte/macrophages in CD1a+ and CD1a– monocyte‐derived DC. Their phenotype indicated that they may be related to dermal DC rather than to LC, i.e., they were all CD11b‐positive, and 72% were Factor XIIIa‐positive, but they did not express E‐cadherin nor VLA‐6. It is interesting that CD1a+ and CD1a– cells showed intracytoplasmic granules that were different from LC Birbeck granules. These pheno‐typical and ultrastructural features are comparable to those of CD14‐derived DC obtained from cord blood precursors [C. Caux et al. J. Exp. Med. 184, 695–706]. These results show a close relationship between these two in vitro models, which are both related to dermal DC. J. Leukoc. Biol. 64: 484–493; 1998.

Status of long-term asymptomatic HIV-1 infection correlates with viral load but not with virus replication properties and cell tropism

Qualitative and quantitative virological parameters were investigated in 68 long‐term nonprogressor (LTNP) HIV‐1‐infected patients and 9 slow‐progressor controls. LTNP status was defined as an asymptomatic HIV infection for at least 8 years, a stability of CD4+ cell counts ≥600 cells/mm3 and no antiretroviral therapy. LTNP subjects exhibited a lower median plasma RNA load than controls (6,000 vs 40,000 RNA copies/ml) despite a wide range of values in both groups. When compared to the control group, LTNP subjects also exhibited a lower virus isolation rate (65% vs 100%) and cell‐associated viremia (0.75 vs. 56.8 number of infectious unit/million cells) when CD8‐depleted CD4+ cells were tested. By contrast, no major differences in virus replication properties or cell tropism were observed. After 1 year of follow‐up, no major overall changes in the virological parameters was observed in the 50 LTNP subjects evaluated at this time. However, nine patients had started antiretroviral therapy, and six others had increased viral loads. Despite the progression observed during the first year of follow‐up, the hypothesis that there is a specific subgroup of LTNP patients who will not develop disease cannot be ruled out as yet. J. Med. Virol. 58:256–263, 1999.

Evolution and plasticity of CTL responses against HIV

Exceptionally potent cytotoxic T lymphocyte responses are generated after HIV invasion and probably control the primary infection as well as the asymptomatic phase of HIV infection. The chronic phase appears as a quasi-equilibrium between waves of new HIV variants and variant-specific CTLs, thus sustaining continuous CTL activation which eventually fails to eradicate HIV disease progression and the reascension of viral replication. Meanwhile, both the host and the virus develop various strategies either to stop or to evade this potentially deleterious permanent CTL activity. The transient effectiveness of CTLs opens perspectives for understanding disease progression generally as well as for immune therapeutic strategies.

Up‐regulation of adhesion and MHC molecules on splenic monocyte/macrophages in adult haemophagocytic syndrome

Haemophagocytic syndrome (HPS) has been associated with the abnormal activation of mono/macrophages and increased cytokine production. However, neither the phenotype of haemophagocytic monomacrophages nor the cellular origin of cytokine production have been described. We studied splenic monomacrophages and lymphocytes from five patients with HPS (two HIV− and three HIV+) and from controls without HPS (three normal HIV− and two pathological HIV+). Using flow‐cytometry, we observed a marked increase in the expression of MHC class I and II, M‐CSF‐receptor and adhesion molecules LFA‐1, LFA‐3, ICAM‐1 (P < 0.05) on HPS+ splenic monomacrophages compared to HPS−, which was independent of their HIV status. A high percentage of CD8+ lymphocytes from 4/5 HPS+ patients produced TNFα and IFNγ, but no IL‐6 upon in‐vitro activation. In a fifth patient CD4+ but not CD8+ lymphocytes produced these cytokines. Although other cytokines might be involved in the pathophysiology of HPS as suggested by the high expression of M‐CSF‐receptor, these results suggest that TNFα and IFNγ secretion by T cells might play a role in the up‐regulation of adhesion and MHC molecules on monomacrophages from HPS.

Low infection frequency of macrophages in the spleens of HIV+ patients

Macrophages are often considered as a reservoir of latent infection in HIV+ patients, and their infection may indeed be very important functionally. However, some quantitative studies did not find high infection frequencies in peripheral blood monocytes. Since lymphoid organs are the major site of infection, macrophage infection was tested in spleens removed from HIV+ patients for treatment of different syndromes. Ten replicates of limiting dilutions from different cell populations were submitted to a nested PCR specific to conserved regions of HIV1 env DNA. On an average, 1/2,300 adherent cells carried HIV1 DNA (n = 7; range: 1/55,000 to 1/660). These adherent cells, obtained after two days of culture, comprised the whole macrophage population, with no biases introduced by surface molecule selection, but were not pure (41-78% macrophages). Only 1/37,000 CD14+ monocyte/macrophages were positive (n = 6; range: 1/130,000 to 1/22,000). Therefore, the infection frequency of the isolated splenic monocytes/macrophages from these patients could be estimated at between 1/37,000 and 1/2,300. In contrast, 1/60 CD4+ T lymphocytes were positive (n = 7; range: 1/190 to 1/17). Within the experimental limits, such as cell isolation, required for accurate quantification, this study in the spleen indicates, as have other studies on peripheral blood, that macrophages do not quantitatively constitute an important reservoir of HIV when compared to CD4+ T lymphocytes.

HIV-specific cytotoxic T lymphocytes directed against alveolar macrophages in HIV-infected patients

A CD8+ lymphocytic infiltration of the lungs is frequently observed in HIV-infected patients, even prior to the onset of opportunistic infections. In such patients, we could demonstrate that most of these CD8+ alveolar T lymphocytes displayed the D44 marker and were functional cytolytic T lymphocytes directed against autologous HIV-infected alveolar macrophages. This primary cytolytic activity was HLA-restricted and, at least partially, specific for the HIV envelope protein, since HLA-A2 alveolar T lymphocytes could specifically lyse cell lines expressing both the HLA-A2 and Env antigens. In contrast to data obtained in peripheral blood, no ADCC activity was observed against the Env antigen. HIV-specific alveolar T-lymphocyte cytolytic activity decreased with progression towards AIDS as shown by studies of a series of 40 patients. Functional abnormalities of the lung epithelium could be associated with the specific lysis of alveolar macrophages, suggesting that local tissue injury could result from the in vivo immune conflict between alveolar HIV-specific CTL and HIV-infected macrophages.

HIV-specific cytotoxic T lymphocytes against alveolar macrophages: specificities and downregulation.

To analyse the evolution of alveolar-lymphocyte-mediated cytotoxic activity directed against autologous alveolar macrophages (AM), cytotoxic assays against various HIV+ target cells were performed in a cohort of 75 patients with HIV-associated lymphoid interstitial pneumonitis (LIP) studied at distinct stages of HIV infection. Our data confirm that alveolar HIV-specific cytotoxic T lymphocytes (CTL) against AM were detectable before AIDS in patients with CD8+ LIP. Mild CD8+ lymphocytic alveolitis occurs silently in 62% of stage II and III patients with no respiratory symptoms. In these cases, the lack of spontaneous alveolar-lymphocyte-mediated cytotoxic activity against autologous AM may contrast with the detection of primary alveolar CTL specific for HIV proteins such as nef. In AIDS patients, the alveolar CTL lytic efficiency against both AM- and HIV-antigen-expressing cells can be inhibited by a suppressor factor produced by alveolar CD8+ CD57+ cells. Therefore, spontaneous CTL lysis of AM may be (1) limited to a subgroup of patients with active LIP and (2) controlled by distinct mechanisms, including suppressor phenomenons, and HIV replication levels in AM.

Dynamics of HIV variants and specific cytotoxic T-cell recognition in nonprogressors and progressors

Infection with the human immunodeficiency virus (HIV) results in a disease characterized by a rapid viral replication, immunodeficiency and chronic immune activation. The vigorous polyspecific cytotoxic T-cell (CTL) response directed against multiple HIV epitopes reduces HIV-infected cell numbers, although unable to eradicate the virus. The plasticity of the specific CTL repertoire ensures adaptation to the high rate of viral variation that can be found in CTL epitopes of several HIV-1 proteins. However, viral persistence occurs despite continuous CTL recognition and although functional importance of conserved sites in the different HIV proteins may impose constraints to viral variation. In the reverse transcriptase (RT) which is a major target for antiretroviral therapy, the impact of the continuous pressure of drug therapy is more obvious than that of the CTLs. Shifts in immunodominant RT regions seem to allow the maintenance of the HIV-1 RT CTL recognition with disease progression and antiretroviral therapy. In respect to new highly active drug combinations, understanding the capacity of virus-specific CTLs to control residual viral variants seems very important and may allow development of efficient immunotherapies to prevent drug-induced viral resistance.