key: cord-0727227-bgzyxks1
authors: Aranda-Anzaldo, A.
title: A role for CD8+ T lymphocytes in the pathogenesis of AIDS
date: 1991-12-31
journal: Research in Immunology
DOI: 10.1016/0923-2494(91)90099-5
sha: 5f2cef60af4697ba45c56306e42803db49b773ad
doc_id: 727227
cord_uid: bgzyxks1

Abstract Experimental and clinical evidence is presented which supports the hypothesis that CD8+ T lymphocytes aimed at suppressing HIV replication in infected CD4+ T cells may have an important role in the pathogenesis of AIDS by directly causing a decrease in CD4+ T lymphocyte numbers. Possible models to test this hypothesis are discussed.

The life cycle of a retrovirus is such that the newly infected cell must undergo at least one round of DNA replication and mitosis for the synthesis of new virus to begin (Weiss et al., 1985) . HIV is a retrovirus widely held to be the aetiological agent of AIDS. Several HIVassociated mechanisms have been described to explain the cytopathic effects on CD4* T cells observed in vitro, but so far none of these mechanisms has been observed in vivo (Garry, 1989) . Many HIV isolates obtained from AIDS patients are poorly cytopathic in vitro and noncytopathic HIV isolates have been associated with disease (Levy, !988) . Usually, non-dividing lymphocytes cannot be productively infected by retroviruses like HIV; only the antigen or mitogen-activated ones can. Infection of unactivated T cells leads to incomplete reverse tran-scription of HIV RNA; the truncated DNA copy is extremely unstable, with a half-life of only 24 h (Zack et aL, i990) . Activation of HiVinfected lymphocytes also seems to be necessary to induce the replication of HIV; virus production has been observed only in lymphocyte cultures exposed to antigens, whereas no virus was produced in absence of antigenic stimulation (Fauci, 198~; Ho et al., 1989; Coombs et al., 1989) .

It has been possible to isolate HIV from peripheral blood lymphocytes (PBL) of most HIV-seropositive patients by cocultivation with freshly activated PBL regardless of the clinical stage of infection (Ho et al., 1989; Coombs et al., 1989) . However, plasma viraemia has been found to correlate with the clinical stage of infection, but only 45°7o of patients x~ith plasma viraemia had HIV p24 antigen in either serum Submitted July 3, 1991, accepted July 30, 1991. or plasma, and no correlation was found between the amount of p24 antigen in plasma and the plasma HIV titres (Coombs et al., 1989) . Cohort data show that over time, increasing numbers of patients with preserved high titres of p24 antibody and without detectable titres of p24 antigen develop AIDS. No explanation has been offered for these findings or for the failure to detect viral antigens in 30-50 °7o of patients (Cheingsong-Popov et al., 1991) .

At least one other virus is currently known to have the same major target cell as HIV" the human herpesvirus-6 (HHV6). Infection with HHV6 is quite common in the general population and it is a fact that several HHV6 strains are directly cytopathic for both PBL and purified CD4 ÷ T lymphocytes (Takahashi et al., 1989; Levy et al., 1990; Carrigan et al., 1990) . It has been shown that HHV6 has preferential tropism for activated cells and that addition of cytokines to infected cultures stimulates the production of HHV6 (Lusso et ai., 1988; Carrigan et al., 1990) ; both features are shared by HIV (Fauci, 1988; Garry, 1989 However, the finding that levels of HIV in plasma and PBL are higher than previous estimates has reinforced the idea that HIV direct cytopathic effects alone may be sufficient to explain much of the pathogenesis of AIDS (Ho et al., 1989; Coombs et aL, 1989; Baltimore and Feinberg, 1989) ; but let us consider the case of the common cold which can be caused by a large number of viruses belonging to not less than four different groups (adeno-, rhino-coxsackie and corona viruses). Nevertheless, all such different viruses are able to cause the same clinical syn-drome because all of them are cytopathic viruses that share the same main target cell population. This fact indirectly suggests that in the case of HIV and HHV6 there may be something more than just the killing of their common target cell population to explain why only HIV causes AIDS.

The key and obvious difference between HIV and HHV6 is that HIV is a retrovirus and as such, it needs host-cell activation for its own replication, while HHV6 is a herpesvirus independent of host-cell activation for its own replication. If HIV replication depends on CD4÷-cell activation, then suppression of HIV replication can be achieved by suppressing CD4÷-cell activation, while in the case of HHV6, its replication can be controlled by other mechanisms which do not target the activation of the host cell.

Viral-induced disease may be due to immunopathological processes and not to viral lysis for their development. In such cases, the incubation period is related to the time for these immunopathological events to result in disease, and not to the period of viral replication. In such situations, the strength of viral replication is not directly related to the severity of clinical illness; the incubation period depends on the mechanism by which the disease is produced, host factors and the virulence and dosage of the infecting agent (Evans, 1989) .

The current picture of HIV infection shows that there are three phases: a short acute phase during which high levels of virus are produced, a chronic phase characterized by very low level of HIV expression, and a crisis, terminal phase where a recrudescence of viral replication occurs (Baltimore and Feinberg, 1989; Clark et 1991). It is important to stress that the early acute phase is clinically characterized by a mild flulike or mononucleosis-like syndrome. The time from infection with HIV to onset of the acute clinical illness is generally 2-4 weeks. The clinical illness usually lasts for 1 to 2 weeks and most of the symptoms are self-limited and do not recur after resolution (unlike symptoms associated with EBV mononucleosis). Following HIV infection, there is an intense period of viral replication demonstrated by the ease of viral isolation and high levels of plasma HIV p24 antigen. The frequency with which HIV can be isolated decreases following seroconversion, and plasma levels of HIV p24 decrease to undetectable levels. Antibodies that inhibit synctium formation in vitro and antibodies that mediate antibodydependent cellular cytotoxicity against virally infected cells develop soon after seroconversion. These findings suggest that the immunological response to primary HIV infection is favourable in the short term (Tindall and Cooper, 1991 ; Clark et al. , 1991) .

However, a CD8 + lymphocytosis with atypical lymphocytes has been observed in primary HIV infection in the second week after the onset of illness (Tindall and Cooper, 1991) . A similar CD8 + lymphocytosis is characteristic of primary cytomegaiovirus (CMV) and Epstein-Barr virus (EBV) infection. In EBV infection, the CD8 ÷ cells have been shown to control infection by killing virus-infected B lymphocyte or suppressing virus replication (Svedmyr and Jondal, 1975) . The increase in CD8 + cell numbers during primary HIV infection occurs concomitant with resolution of clinical symptoms and a sharp reduction in the amount of plasma HIV p24 antigen (Cooper et al., 1988) . It has been reported that during the first weeks of primary HIV infection, there is a decrease in the responsiveness of lymphocytes to antigens and mitogens (Cooper et al., 1988) ; a significant decrease in circulating CD4 ÷ cells at the time of Western blot seroconversion, has also been observed; such a reduction in CD4 + cell numbers was not present during the preceding period when the Western blot was negative. However, prior to Western blot seroconversion, there is an increase in CD8 + cells . Azidothymidine (AZT), an inhibitor of viral reverse transcriptase, can produce a significant reduction (up to 25-fold) in HIV plasma titres when compared with the mean plasma titres of untreated patients with AIDS or AIDS-related complex (Ho et al., 1989) . However, in spite of its clear effect upon HIV replication, AZT produces only a very transient increase in CD4 + cell numbers. It is true that the levels of CD4 ÷ T cells correlate with disease progression, but it is likely that by the time such levels decline, the damage to the ho~t's immune sys~ tem is probably irreversible (Baltimore and Feinberg, 1989) . A profound CD4+lymphocytopenia is not solely characteristic of HIV infection; severe CD4 + cell depletion may occur in patients with visceral leishmaniasis or Pneumocystis carinii pneumonia in the absence of HIV infection (Cozon et al., 1990; Gautier et al., 1991) . It is certain that the HIV titres produced during the chronic phase of infection are so low that they can be excluded as the cause of a meaningful reduction in CD4 + cell numbers.

The proportion of CD4 + T cells infected with HIV in asymptomatic subjects is low (1 in 2,500 to 1 in 26,000). Although high proportions of the infected cells contain replicationcompetent provirus copies, this is fully compatible with good health and a high number of CD4 + T cells in the blood even after 4 years of infection with HIV (Brinchmann et al., 1991) . Lymph nodes have been shown to contain an average of only 0,001 or fewer HIV DNA copies per cell, and there is no evidence of sequestration and destruction of large numbers of HIV-infected CD4 + T cells in lymph nodes (Shibata et a!., 1989 ). An important question is whether HIV-infected bone marrow CD34 + stem-progenitor cells serve as a significant reservoir of virus in HIV-infected individuals. However, it has now been shown that infection of bone marrow stem/progenitor cells with HIV occurs rarely, if ever, in vivo. Thus, infected stem cells are not a major source of persistent HIV and do not account for haematopoietic suppression (Davis et al., 1991) . It is difficult to accept that a cytopathic effect by HIV on individual infected cells can kill a sufficient num-ber of CD4 ~ T cells to exceed the capacity of generation of new CD4 + cells if this remains normal (Ludyard and Grossi, 1989) . Recent reports about the virus load in CD4 ÷ T cells from asymptomatic individuals suggest that few of this cells are likely to be lost as a direct consequence of infection by HIV and that additional mechanisms may contribute to the depletion of CD4 ÷ cells (Brinchmann et al., 1991) .

So far, potential treatments for AIDS have been designed with the intention of inhibiting or suppressing HIV replication. However, current evidence suggests that when HIV replicates actively, it produces only a mild mononucleosislike syndrome (Steeper et al., 1988; Tindall and Cooper, 1991) . Thus, perhaps the immune reaction triggered by the CD4 + T-cell activation necessary to permit further HIV replication is the cause that finally leads to immunodeficiency and major pathology.

For some time, it has been realized that suppressor T cells can be a suitable mechanism for limiting immune responses to a foreign agent when the foreign agent is not quickly elimitated, as in chronic infections or parasitic infections, and various types of established organ transplants (Nossal, 1989) . For example, selfreactive T cells that mature in the thymus and enter the periphery could be functionally inhibited by other T cells (suppression/immunoregulation) when these self-reactive T cells encounter the antigen they recognize, thus leading to clonal anergy. It has been shown that clonal anergy can account for antigen-specific tolerance induced in mice (Rammensee et al., 1989) . It has also been shown that CD8 + suppressor T cells can induce clonal anergy in CD4 ÷ T cells responding t~.~ Mycgbaete~ ium leprae antigens in vitro, a fact which suggests that suppressor T cells may operate by interfering with the delivery of costimulatory signals to other antigen-stimulated T cells (Mueller, 1989) .

It has been found that CD8 ÷ T cells from HIV-infected subjects, but not from seronega-tive controls, are able to suppress the replication of HIV in acutely infected CD4 ÷ T cells in vitro, but also in PBL from HIV-infected individuals. This suppression was partly mediated by a diffusible factor, but not by a direct elimination of cells by cytotoxic cell activity; instead, virus production was suppressed. It is noteworthy that complete suppression of virus replication was achieved in vitro even with low numbers of CD8 ÷ T cells from HIV-infected individuals (Walker et al., 1986; Walker and Levy, 1988; Walker et al., 1989) ; this suggests that there is a tendency toward an excess of CD8 ÷ T cells not necessary for efficient HIV suppression. In some cases, T-cell defects observed in AIDS patients lymphocytes can be eliminated by purification of cells before testing for proliferation and T helper functions, sug. gesting that such a restoration of function might be due to the removal of suppressor T cells (Margolick et al., 1985) .

The hypergammaglobulinaemia seen in many AIDS patients goes against the idea that AIDS is directly related to CD4 ÷ T-cell death, because one would expect a decrease in antibody production with decrease numbers and functions of CD4 ÷ helper T cells. It has been shown that spontaneous immunoglobulin productien in ma et al., 1988) . It has also been observed that a drop in the number of CD4 ÷ T cells occurs only after seroconversion, since it has not been observed in HIV-infected seronegative subjects. Nevertheless, lower CD4 ÷/CD8 ÷ ratios have been observed before seroconversion; however, these are not caused by abnormally low numbers of CD4 ÷ cells, but rather by elevated numbers of CD8 ÷ cells (Imagawa et al., 1989) .

Immunological abnormalities have been observed in HIV-infected homosexual men long before there is a depletion in CD4 ÷ T cells (Marion et al., 1989; Miedma et al., 1989) . Cohort studies of the relationship between CD8 ÷ T-cell counts and progression to AIDS have shown that, contrary to expectations, progression to AIDS is more frequent in those HIV-infected individuals with high early CD8 ÷ Tcell counts (Anderson et aL, 1991) . The presence of spontaneous suppressor cell activity (SSCA) in patients with AIDS and ARC has been found; such cell populations appear to interact in producing normal or augmented downregulation of residual T-lymphocyte function, even in the face of helper/inducer CD4 + T-cell depletion (Katner et al., 1987) . Higher levels of SSCA have been found in patients with early clinical AIDS and advanced terminal disease. Patients surviving clinical AIDS for more than one year had lower levels of SSCA compared to controls, SSCA seems to predict disease progression, since patients with ARC who showed rises in SSCA progressed to AIDS, while those with blunted SSCA did not progress to full disease (Mullin and Mayer, 1991) .

The inhibitory activity of HIV replication by CD8 ÷ T cells seems to be mediated 13y mechanisms other than classical virus-specific MHC class-1-restricted cytotoxicity (Brinchmann et aL, 1990) . In mice, it has been possible to isolate and characterize CD8 ÷ suppressor T-cell clones capable of inhibiting the antigen-induced proliferative responses of CD4 + helper T cells clones, without detectable cytotoxicity for either antigen-presenting cells or responding cells. Such inhibitory activity is neither antigen-specific nor MHC-restricted (Hisatsune et al., 1990) .

Thus, the final recrudescence of HIV replication observed in terminal AIDS could only represent the final exhaustion of the host's immune system due to a CD8 + T-cell-mediated chronic oversuppression which ends by downregulating even CD8 + T-cell activation. Further evidence which suggests that CD8 ÷ suppressor T cells participate in the immunopathogenesis of AIDS has been discussed elsewhere (Laurence et al., 1983; Mullin et al., 1987; Aranda-Anzaldo, 1990; Mullin and Mayer, 1991) . Thus, while normal clonal anergy mediated by noncytotoxic suppressor T cells is an antigen-and cell-specific process, in the case of AIDS we could be facing a suppressor T-cell mediated non-specific pan-anergy which may affect all sorts of CD4 ÷ T-cell populations.

Comparing the proliferative responses from AIDS patient lymphocytes with the responses of lymphocy~tes from HIV-seronegative homosexual controls, it was found that the decreased proliferative response of AIDS lymphocytes cannot be primarily due to a deficient signal via the cell membrane receptors, since direct stimulation of the following intracellular stimulatory pathways did not normalize the response. Furthermore, it was found that AIDS lymphocytes have no major defect in interleukin-2 receptor expression and that such lymphocytes are able to produce interleukin-2 when stimulated directly via the intracellular stimulatory pathways (Hofmann et al., 1989) . Thus, the decreased production of IL2 seen in AIDS lymphocytes must be explained by a defect more central than the inactivation of membrane-dependent stimulatory pathways. It has already been proposed that the decreased proliferative response observed in CD4 + lymphocytes from HIV-infected individuals is linked to the inhibition of the major steps leading to DNA synthesis and cell proliferation (Hofmann et ai., 1989; Aranda-Anzaldo, 1990) .

Studies from several laboratories show that PBL from AIDS patients inhibit in vitro T helper responses of PBL from non-infected individuals. Soluble factors that inhibit normal T helper function have been detected in cultures of PBL from HIV-positive patients. Both the inhibitory cells and the suppressive factors blocked self-restricted T helper responses without affecting tbe aUogeneic restricted components of the mixed lymphocyte reaction or responses to Tcell mitogens (Laurence and Mayer, 1984; Hofmann eta!., 1986) . This suggest that selective loss of self-restricted T helper function is due to some soluble factor (or factors) produced by a subset of PBL (Aranda-Anzaldo, 1990, Shearer and Clerici, 1991) .

In AIDS patients, there are thymus histological changes that suggest premature involution and inflammatory changes similar to those seen in graft vs. host disease (GVHD). Immune abnormalities of early HIV infection parallel those seen in what is called immunostimulatory GVH reaction observed in mice, which leads to activation of suppressor ceils that specifically inhibit the murine T helper cells (Shearer et al., 1986) . It has been shown that the major effect of GVHD-induced suppressors in mice is inhibition of DNA synthesis, because such suppressors block the proliferation induced by receptormediated stimulation or exogenously added growth factors and do not significantly inhibit either lymphokine secretion or lymphokine receptor expression (Wall et al., 1989) . These observations bear a striking similarity to those made in PBL from full-blown AIDS patients in which the proliferative response of both CD4 + and CD8 * T cells is defective; such a response may be increased by addition of IL2 but only to subnormal levels (Hofmann et al., 1989) . Also, it is remarkable that in T-cell-induced clonal anergy observed in mice, the unresponsive T cells show a normal density of membrane T-cell receptors as well as the ability to the express IL2 receptor that parallels the situation found in GVHD and in AIDS lymphocytes. The unresponsive T cells in clonal anergy fail to proliferate because they cannot produce IL2, like the AIDS lymphocytes, which can only produce IL2 after direct stimulation of the intracellular pathways (Hofmann et al., 1989) . It has been reported that fulminant viral disease induced by m~,, .... ,~r,,, ~,^~ strains of simian immunodeficiency virus (SIV), in affected monkeys produces pathological changes in the gastrointestinal tract typical of an immunologically mediated reaction like GVHD (Martin, 1990) .

Thus, in an antigen-stimulated host, there will be a persistently activated helper T-lymphocyte population subject to downregulation by an active suppressor T-cell population. The introduction of HIV into this system generates a new population of activated T cells that need to be suppressed to avoid replication and spread of HIV. Suppressor T cells will produce soluble factors able to stop HIV replication by inhibiting DNA replication in the host cell or by blocking the entry of HIV-infected cells into the S phase of the cell cycle. These two mechanisms are not mutually exclusive. The suppressor soluble factors will be able to block DNA replication not only in HIV-infected cells but also in non-infected CD4 + lymphocytes. HIV replication might be contained and controlled for some time by a combination of antibody-mediated mechanisms and CD8 + cell-mediated suppression. This will produce a population of latently (or persistently) HIV-infected CD4 + lymphocytes susceptible to be clonally or polyclonally activated through the years by whichever new antigenic stimuli the host shall be exposed to. This will lead to a new cycle of helper T-lymphocyte suppression mediated by CD8 + cells, thus generating a vicious cycle of activation-suppressionactivation, etc. in which factors with a blocking effect on CD4 + T-cell replication will have a critical role.

The loss of the equilibrium between activation and suppression will be favoured in highrisk groups exposed to continuous antigenic stimulation, leading to a state of immune oversuppression in which certain CD8 + T-cell subsets less susceptible to downregulation by the normal homeostatic controls of the immune network will take a leading role. As a consequence of this, the CD4 + cells will be the first lymphocyte population to decline with a consequent decrease in the production of lymphokines necessary for the activation of macrophages and the maturation to effector function of natural and other constitutive killer cells. Finally, even the CD8 ÷ T-cell numbers will decline due to a lack of the appropriate helper T-lymphocyte stimulation. Nevertheless, the relative number of CD8 + cells will remain elevated in the immunodepressed subjects.

There exist several models of virus-induced immunodeficiency in mice, cats, cattle and monkeys (Weiss, 1989) . Although none of the clinical syndromes obse~ ved in these animal models is completely equivalent to human AIDS, such animal models remain the only possible source of in vivo experimental data. Animal models are suitable for testing the suggested hypothesis that reduction or elimination of CD8 ÷ suppressor T cells may lead to clinical improvement or abrogation of the immunodeficient state.

For example, severe experimental immunodeficiency can be induced in mice by a murine leukaemia virus (MuLV), which is replication-defective. Although in this model, there is no evidence that the virus infects murine helper T cells, the B-cell hyperplasia, hyperglobulinaemia and immunosuppression observed seem to be dependent essentially on the presence of T lymphocytes, since the defective virus affects nude mice much less severely (Aziz et al., 1989) .

Current evidence shows that there is an extremely low level of HIV provirus present in the circulating PBL from HIV-infected persons, and that more than 99 °7o of the lymphocytes which contain provirus do not express the virus (HO et al. , 1989; Schnittman et aL , 1989; Simmonds et al., 1990; Brinchmann etaL, 1991) . Thus, we may be facing a situation in which the observed final increase in the percentage of infected CD4 + cells in AIDS patients is a consequence of clonal growth and not of virus replication and spread. It has been shown that viral replication per se is not necessary to induce the murine acquired immunodeficiency syndrome, because strictly helper-free defective MuLV is able to produce such a syndrome without any evidence of viral replication (Huang et al., 1989) . A striking observation is that the number of infected cells is at least 1000-fold higher in diseased mice than in the newly infected ones. It was concluded that the expansion of the infected cell population occurs from cell division and not from reinfection, since these mice were found to harbour no virus able to replicate on mouse cells. The data indicate that the expanded cell population is clonal or oligoclonal, and as such, it originated from a single (or a few) infected cells which transformed or gained a growth advantage after infection by the defective retrovirus. Thus, it was suggested that interaction of such virus-modified cells with other cells of the immune system could trigger the immunodeficiency (Huang et ai., 1989) .

There have been attempts to use some immunosuppressive drugs like cyclosporin in the treatment of AIDS, and partial beneficial effects have been observed (Andrieu et al., 1988) . It is known that precursors of CD8+ T cells are readily inactivated by agents that destroy or block rapidly dividing cells (such as cyclophosphamide and cyclosporin). Mature CD8 + T cells are resistant to these agents. The so-called Ts-I cells able to produce suppressor soluble factors in response to antigens are particularly short-lived. Hence, in systems where antibody production and cell-mediated immunity are strongly suppressed because the antigen elicits more CD8 + than CD4 + T cells, elimination of the blockage induced by Ts-1 cells (either by low doses of cyclosporin or cyclophosphamide) at about the time the antigen is administered can result in enhancement of the immune response, but if the agent is given after the antigen has elicited enough CD8 + T cells, then no permanent enhancement will result (Davis et al., i980).

Thus, it is especially noteworthy that the murine acquired immunodeficiency syndrome (MAIDS) associated with proliferation of target cells that have been infected by a replicationdefective strain of MuLV can be prevented, and normal T-cell function can be restored as well in these mice by treatment with cyclophosphamide, but not with other antineoplastic drugs like 5-aza-deoxycytidine or arabinosylcytosine, r+.r,....~+tr~d~,A 4.h,.+,.l-,,.,,.,.+-.l,-~,~l.-,*~e.*-,.l~,',,.~:rl~ " ~.4..~'~'~.~-~r] [../11LJ v lSd~.,~lt tllOt v~., yK, IKYI./IIK.13I-/IIt]I.IIIIIKIIK; [S aHILIIIII;~tK:I K;U from the early days after the inoculation of mice with defective MuLV (Simard and Jolicoeur, 1991) . Such a paradoxical effect of an otherwise immunosuppressive drug may be explained, at least in part, by the fact that low doses of cyclophosphamide are known to preferentially eliminate the suppressor T cells from the lymphocyte population (Shand, 1979) .

It is well know that it is possible to isolate different HIV strains, each of them having different biological properties, from the same HIV-infected person (Saag et al., 1988 ; Chiodi et al., 1989) . Perhaps some strains (quasi-species) of viruses like MuLV or HIV are either replication-deficient or replication-defective such that they able to trigger and perpetuate an am-biguous immune response in which suppressor T cells aimed at curtailing the expansion of the infected cell population have a dominant role (Meyerhans et al., 1989) . It is likely that suppressor factors produced by the suppressor cells will be able to stop HIV replication by inhibiting DNA replication in the host cell or by blocking the entry of the HIV-infected cell into the S phase of the cell cycle. However, it is also possible that such suppressor factors would not be able to distinguish between ceils that are latently infected, cells that actively express the virus and cells that are expanding in numbers due to the growth advantage conferred by a replicationdetective quasispecies of the virus in question. The net result would be downregulation and inhibition of the whole range of the target cell population independently of infection status. It may be speculated that fully replicationcompetent strains of HIV would produce a balanced immune response like the one seen in classical viral infections in which cell-mediated immunity is able to restrict the spread of both the virus and the infected-cell population without causing further harm to the host.

Un r61e pour les lymphocytes T CD8 + dans la pathogen~se du SIDA Des preuves cliniques et exp6rimentales sont avan-c~es pour corroborer l'hypoth~se que les lymphocytes T CD8 + visant/~ la suppression de la r6plic~,tion du VIH dans les cellules T CI34 + pourraient jouer un rfle important dans la pathogen~se du SIDA en provoquant directement une diminution du nombre des lymphocytes T CD4 + . Des modules en vue de tester cette hypoth~se sont discut6s.

Mots-clds: SIDA, HIV, R6plication, Lymphocyte T, Immunosuppression; Anergie clonale, Maladie du greffon contre l'hfte, Immunod6ficience expgrimentale; Mise au point.

Since submission of this article, it has been reported that cyclosporine A protects mice from the immunodeficiency disease induced by replication-defective MuLV (Cerny et ai. (1991 ), Eur. J. Immunol. 21, 1747 -1750 .

CD8* T lymphocytes and progression to AIDS in HIV-infected men: some observations

Effects of cyclosporin on T-cell subsets in human immunodeficiency virus disease

Is immune oversuppression the direct cause of AIDS ?

Severe immunodeficiency disease induced by a defective murine leukaemia virus

HIV revealed.New Engi

+ T cells in~.0it HIV replication in naturally infected CD4 + T cells

Few infected CD4 + T cells but a high proportion of replication-competent provirus copies in asymptomatic human immunodeficiency virus type 1 infection

Suppression of human immunodeficiency virus type 1 replication by human herpesvirus-6

Relation between humoral responses to HIV gag and env proteins at seroconversion and clinical outcome of HIV infection

Biological characterization of paired human immunodeficiency virus type 1 isolates from blood and cerebrospinal fluid

High titers of cytopathic virus in plasma of patients with symptomatic primary HIV-I infection

Plasma viremia in human immunodeficiency virus infection

Characterization of T-lymphocyte responses during primary HIV infection

Profound CD4 + T cells lymphocytopenia in the absence of HIV infection in a patient with visceral leishmaniasis

Absent or rare human immunodeficiency virus infection of bone marrow stem/progenitor cells in vivo

Does HIV cause AIDS? an historical perspective

The human immunodeficiency virusinfectivity and mechanisms of pathogenesis

Potential mechanisms for the cytopathic properties of HIV

Unexplained CD4 + T-cell deficiency in non-HIV patients presenting as a Pneumocystis carinii pneumonia

CD8 + suppressor T cell clone capable of inhibiting the antigen and anti-T cell receptor-induced proliferation of Th clones without cytolytic activity

Quantitation of human immuno=deficiency virus type 1 in the blood of infected persons

Immunological studies in the acquired immunodeficiency syndrome-II. Active suppression or intrinsic defect investigated by mixing AIDS cells with HLA-DR identical normal cells

Stimulation of AIDS lymphocytes with calcium ionophore (A23!87) and phorbol ester (PMA): studies of cytoplasmic free Ca 2 *, IL-2 receptor expression, IL-2 production and proliferation

lmmunodeficiency and clonal growth of target cells induced by helper-free defective retrovirus

Human immunodeficiency virus type 1 infection in homosexual men who remain seronegative for prolonged periods

Spontaneous suppressor cell activity in patients with AIDS and associated conditions

Soluble suppressor factors in patient with acquired immunodeficiency syndrome and its prodrome" elaboration in vitro by T-lymphocyte-adherent cell interactions

Immunoregulatory lymphokines of T hybridomas from AIDS patients" constitutive and inducible suppressor factors

Mysteries of HiV" challenges for therapy and prevention

Human herpesvirus 6 inhibits human immunodeficiency virus type 1 replication in cell culture

In vitro cellular tropism of human B-lymphotropic virus (human herpesvirus-6)

Clonal analysis of T lymphocytes in AIDS. Evidence for an abnormality affecting individual helper and suppressor T cells

Evidence that prior immune dysfunction predisposes to HIV infection in homosexual men

Fast-acting slow viruses

Temporal fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolations

Immunological abnormalities in human immunodeficiency virus (H IV) infected asymptomatic homosexual men. HIV affects the immune system before

Bcell activation in HIV infection: relationship of spontaneous immunoglobulin secretion to various immunological parameters

Do tolerant T cells exist?

Infection with HTLV-I and HTLV-III in T8 lymphoproliferative disease

Correlation of spontaneo,~ suppressor cell activity with progression of human immunodeficiency virus infection

Immunologic tolerance: collaboration between antigen and lymphokines

Clonal anergy induced in mature V6" T lymphocytes on immunizing MLs-I b mice with MLs-I a expressing cells

Extensive variation of human immunodeficiency virus type-! in vivo

The reservoir for HIV-I in human peripheral blood is a T cell that maintains expressionof CD4

The immunopharmacology of cyclophosphamide

A model for the selective loss of major histocompatibility complex self-restricted T cell immune responses during the development of AIDS

Early T-helper cell defects in HIV infection

Human immunodeficiency viral DNA is readily found in lymph node biopsies from seropositive individuals. Analysis of fixed tissue using PCR

The effect of antineoplastic drugs on murine acquired immunodeficiency syndrome

Human immunodeficiency virus-infected individuals contain provirus in small numbers of peripheral mononuclear cells and at low copy numbers

Heterophil-negative mononucleosis-like illness with atypical lymphocytosis in patients undergoing seroconversions to the human immunodeficiency virus

Cytotoxic effector cells specific for B-cell lines transformed by Epstein-Barr virus are present in patients with infectious mononucleosis

Predominant CD4 ÷ T-lymphocyte tropism of human herpesvirus 6-related virus

Primary HIV infection: host responses and intervention strategies

CD8 ÷ lymphocytes can control HIV infection in vitro by suppressing virus replication

A diffusible lymphokine produced by CD8 ÷ lymphocytes suppresses HIV replication

CD8 ÷ T lymphocyte control of HIV replicationin cultured CD4 ÷ cells varies among infected individuals

Immunodeficiency in graft-versus-host disease (GVHD) II: effects of GVHD-induced suppressor cells on CD4 ÷ T cell clones

RNA tumor viruses

Human immunodeficiency virus type 1 (HIV-I) infection a median of 18 months before a diagnostic Western blot. Evidence from a cohort of homosexual men

HIV-1 entry into quiescent primary lymphocytes: molecular analysis re~eals a labile latent viral structure