Groux, Torpier, Monté, Mouton, Capron, Ameisen 1 Classification: Biological Sciences. Immunology; or Cell Biology; or Medical Sciences Activation-induced death by apoptosis in CD4 T cells from HIV-infected asymptomatic individuals HERVE GROUX, GERARD TORPIER, DIDIER MONTE, YVES MOUTON’, ANDRE CAPRON, AND JEAN CLAUDE AMEISENT C.1.B.P., INSERM U167-CNRS 624, Institut Pasteur, 59019 Lille, France. * Service des Maladies infectieuses, Hdpital de Tourcoing, France. tT To whom reprint requests should be addressed Key-words: HIV, AIDS pathogenesis, activation-induced CD4+ T-cell death, apoptosis, prevention of T-cell death. Abbreviations: PBMC, peripheral blood mononuclear cells; TCR, T-cell receptor; PHA, phytohemagglutinin A; PWM, pokeweed mitogen; SEB, staphylococcal enterotoxin superantigens; TT, tetanus toxoid recall antigen; Infl, Influenza A hemagglutinin recall antigen; IL, interleukin; mAb, monoclonal antibody. Groux, Torpier, Monté, Mouton, Capron, Ameisen 2 ABSTRACT In immature thymocytes, T-cell receptor (TCR) mobilization leads to an active T-cell suicide process, apoptosis, involved in the selection of the T-cell repertoire. We have proposed that inappropriate induction of such a cell death program in the mature CD4+ T-cell population could account for both early qualitative and late quantitative CD4t+ T lymphocyte defects of human immunodeficiency virus (HIV)-infected individuals. Here, we report that the selective failure of CD4+ T cells from 38 clinically asymptomatic HIV-infected individuals to proliferate jin vitro to major histocompatibility complex class Il (MHC-Il)-dependent TCR mobilization and to pokeweed mitogen is due to an active CD4+ T-cell death process, with biochemical and ultrastructural features of apoptosis, including DNA fragmentation in multiples of 200 base pairs, and chromatin condensation. Activation-induced cell death was not detected in T cells from any of 20 controls, and occurred in purified CD4+ T cells from HIV-infected asymptomatic individuals. Activation- induced CD4t+ T-cell death was prevented by cycloheximide, cyclosporin A, and a CD28 monoclonal antibody. CD28 monoclonal antibody not only prevented apoptosis but also restored T-cell proliferation to stimuli, including pokeweed mitogen, superantigens, and the tetanus and influenza recall antigens. These findings provide new insights into the pathogenesis of AIDS and may represent a basis for the design of specific therapeutic strategies. Groux, Torpier, Monté, Mouton, Capron, Ameisen 3 HIV-infected individuals present early CD4+ T-cell functional defects (1-7) that precede the quantitative reduction in this cell population that leads to AIDS. These functional defects are detected while less than 1/1,000 TH cells are infected (8-10), and are characterized by a selective loss of ability to proliferate in vitro to self MHC-ll-restricted recall antigens and to pokeweed mitogen (PWM) (1-7). CD4* T-cell dysfunction and depletion have been attributed to a wide range of distinct mechanisms. In particular, the early qualitative defects have been related to T-cell suppression (7,11), clonal anergia (7), autoimmune responses (12), inhibitory effects of HIV proteins (13,14), or selective HIV-mediated destruction of memory T cells, leading to the presence of only naive CD4* T cells (15,16). We have proposed the hypothesis (17) that a single mechanism, the inappropriate re-emergence of a T-cell death program in response to activation could account for both early qualitative and late quantitative CD4+ T-cell defects from HIV-infected individuals. Programmed cell death, or activation-induced cell death, or apoptosis, is an active cell suicide mechanism of widespread biological importance (18) that constitutes the physiological response of normal immature thymocytes to activation (18-23); this process is involved in the negative selection of the T-cell repertoire, the deletion of autoreactive T-cell clones, and the establishment of self-tolerance (24). This cell suicide mechanism occurs in the absence of bystander-cell destruction, requires cell activation, Groux, Torpier, Monté, Mouton, Capron, Ameisen 4 initiation of protein synthesis, and involves the activation of an endogenous endonuclease that results in a characteristic regular fragmentation of the entire cellular DNA into multiples of an oligonucleosome unit of 200 base pairs (18-28). In immature thymocytes, apoptosis is not an obligatory response to TCR stimulation, but is the consequence of incomplete signal transduction related to the nature of the antigen presenting cell and to the absence of certain co-signals (22,23,29,30). A major question in T-cell biology is thus whether TCR mobilization may also lead in certain circumstances to the re-emergence of a functional cell death program in mature T cells. We have investigated whether in vitro activation of T cells from clinically asymptomatic HIV1-infected individuals - including individuals with normal CD4+ T-cell counts- and from controls with polyclonal activators and self-MHC-Il-dependent recall antigens may lead to T-cell death. Since memory T cells specific for a given recall antigens are rare, and might be depleted in HIV-infected individuals, we also investigated the response to the self-MHC-lIl-dependent staphylococcal enterotoxin B superantigens (SEB) (31). These superantigens bind to MHC-II molecules and interact with defined VB TCR molecules expressed by up to 30% of normal human T cells, inducing proliferation in normal mature CD4+ T cells (31) and apoptosis in immature thymocytes (21). Groux, Torpier, Monté, Mouton, Capron, Ameisen 5 MATERIALS AND METHODS Study subjects. Peripheral blood was obtained from 38 HIV-infected individuals in the Service des Maladies Infectieuses, Centre Hospitalier de Tourcoing, France. They were 26 males and 12 females, all clinically asymptomatic (stage Il of the Center for Disease Control (CDC) classification); 25 were CDC stage IIA (no biological abnormalities, CD4>500/mm%, mean 856); 13 were CDC stage IIB (biological abnormalities, CD4<500/mm3, mean 345). HIV infection was related to homosexuality (n=15), heterosexual contact (n=14), intravenous drug use (n=7), or blood transfusion (n=2). Controls were 20 HlV-seronegative donors from the medical staff. Cell preparations. Peripheral blood mononuclear cells (PBMC) were obtained on Ficoll-Hypaque, and cells were cultured as previously described (32). In some experiments, CD4+ or CD8+ T cells were purified by negative selection with magnetic beads coated by anti mouse !gG (Dynal, Biosys, France). Cells (50.106) were plated to plastic Petri dishes in order to harvest adherent cells by scraping. Non adherent cells were incubated with Sug/ml of CD20, CD56, MHC-II, CD4 or CD8 monoclonal antibodies (mAb) in a volume of Smli in RPMI for 30mn. Subsequently, excess antibody was removed by washing twice in RPMI. The cells were then resuspended in 5ml RPMI with magnetic beads (according to the manufacturer's instructions). Groux, Torpier, Monté, Mouton, Capron, Ameisen 6 This mixture was rotated in the cold for 30mn and the cells were passed through a magnetic field twice to remove the cells that had bound to the magnetic beads. Cells were 98% pure as assessed by cytofluorometry. In some experiments, PBMC were depleted either in CD4*+ or CD8* T cells by using the same general method, cytofluorometry analysis revealing less than 2% contaminating cells. Cell proliferation assays. Cell proliferation assays were performed in 96-well culture plates (Nunc) in a final volume of 200u! as previously described (33). PBMC (2.5 10°/ml) were cultured in RPMI/10% fetal calf serum. Mitogens (purchased from Sigma, France) were used respectively at the following final concentrations: phytohemagglutinin (PHA), 10ug/ml; concanavalin A (ConA), 10yg/ml; pokeweed mitogen (PWM), 10ug/mIl; staphylococcal enterotoxin B superantigens (SEB), 1yg/ml; the CD3 mAb was used at tyg/ml. Tetanus toxoid (TT) recall antigen (Biomerieux, France) was used at 10yg/ml; and Influenza A hemagglutinin (Infl) recall antigen (Eurobio, France) at 10yg/mi. After 3 days for mitogens or 6 days for antigens, cultures were pulsed with 1uCi of 3[H]-thymidine (5 Ci/mmol, Amersham, France) during the final 15h of incubation, and harvested. Evaluation of cell death by Trypan Blue exclusion. Cells were incubated in 96-well plates with various stimuli in the same conditions as for proliferation assays. They were harvested by pipetting and diluted 1:2 with 0.1% trypan blue in PBS. The live and dead cells were counted in a hemocytometer. Groux, Torpier, Monté, Mouton, Capron, Ameisen 7 DNA fragmentation assays. DNA fragmentation was determined according to the methods of Wyllie and Morris (28) and Newell et al. (33) with slight modifications. In brief, 107 cells were collected by centrifugation at 200g for 10mn, and lysed in iml hypotonic lysing buffer (SmM Tris Ph 7.4, 5 mM EDTA, 0.5% Triton X 100). The lysates were centifuged at 13,000g for i5mn. Supernatants were deproteinized by extraction once in phenol/chloroform and twice in chloroform/isoamy| alcohol (24:1) and precipitated at -20°C in 50% isopropanol, 130 mM NaCl: after electrophoresis on 2% agarose slab gels, DNA was stained by ethidium bromide. Electronmicroscopy. Cells were fixed in 1% glutaraldehyde in 0.1M sodium cacodylate buffer pH 7.4 for 2h at +4°C. Pellets were post fixed in 1% aqueous osmium tetroxide for 1h, en-block stained in 1% aqueous uranyl acetate for 6h and embedded in araldite. Sections were stained with uranyl acetate and lead citrate before examination with a Philips EM 420 electron microscope. Monoclonal antibodies and chemicals. The monoclonal antibodies (mAb) used in this study were CD3 (X35-7, IgG2a), generous gift from Dr. Bourel (France); CD28 (9.3, !gG2a), generous gift from Oncogen corp.; CD28 (CLB 28/1, IgG1), purchased from Jansen; CD20 (IOB20, IgM), CD56 (IOTS56, IgG1), HLA-DR (lOT2a, IgG2b), CD4 (IOT4, IgG2a), CD8 (IOT8, IgG2a), were all purchased from Immunotech, France; CD5 (A50, IgG1), CD44 (P245, IgG2a), generous gifts from Dr. A. Bernard, France. Groux, Torpier, Monté, Mouton, Capron, Ameisen 8 Cyclosporin A was purchased from Sandoz, France; and Cycloheximide from Sigma. Statistical analysis. Statistical significance was assessed by using a Student t test. RESULTS Proliferative response of T cells from _ HIV-infected asymptomatic individuals and from controls. We investigated the jn vitro. proliferation of T cells from 38 HIV-infected asymptomatic individuals and 20 healthy controls in response to the polyclonal activators PHA, ConA, PWM, CD3 mAb, to the tetanus recall antigen, and to the SEB superantigens. T cells from all controls proliferated to all stimuli, while T cells from HIV-infected asymptomatic individuals showed a selective defect in their response to PWM, to SEB and to recall antigen (Fig. 1). Proliferation of T cells from HIV-infected individuals to ConA and to CD3 mAb was only slightly reduced (not shown), as was proliferation to PHA (Fig. 1). This was consistent with the fact that all HIV-infected individuals studied were clinically asymptomatic, with few or no_ biological abnormalities, since defective proliferation to CD3 mAb and to PHA have been reported to be predictive markers of evolution towards AIDS in long term infected individuals or patients (34,35). Cell death in PBMC and purified T cells from HIV-infected Groux, Torpier, Monté, Mouton, Capron, Ameisen 9 asymptomatic individuals and from controls. Addition of PWM or SEB to PBMC from 20 different HIV-infected asymptomatic individuals was followed by cell death after 48h of about 40% and 20% respectively, of PBMC, whereas no cell death was observed at 48h in the unstimulated PBMC from HIV-infected asymptomatic individuals, nor in the unstimulated and stimulated PBMC from 20 different controls (95% viability) (Fig. 2). In order to identify the cell population undergoing cell death after activation, PBMC from a HIV-infected individual and a control were depleted either of CD4* T cells or of CD8* T cells. After addition of PWM or SEB, cell death was only observed in the CD8* T-cell depleted PBMC population from HlV-infected asymptomatic individuals (Fig. 3a), suggesting that neither CD8+t T cells, B cells, natural killer cells nor monocytes underwent activation-induced cell death. Purified CD4+ T cells and CD8+ T cells were also prepared from PBMC from a HIV-infected asymptomatic individual and a control. Activation-induced cell death was only observed in response to PWM or SEB in the CD4+ T lymphocyte population from HIV-infected asymptomatic individuals (Fig. 3b). Mechanism and prevention of T-cell death. Death of T cells from HiV-infected asymptomatic individuals involved 2 features characteristic of apoptosis (18-21,25-28,36). First, gel electrophoresis of the DNA of PBMC from HIV-infected asymptomatic individuals, performed 18h after addition of PWM or SEB, showed a DNA fragmentation pattern in multiples of a 200 base pair oligonucleosome length unit (Fig. 4a). Second, electron Groux, Torpier, Monté, Mouton, Capron, Ameisen 10 microscopy of PBMC from HIV-infected asymptomatic individuals 18h after the addition of PWM revealed nuclear chromatin condensation (Fig. 4c). An essential characteristic of programmed cell death in various cell populations is its dependence on cell activation, gene transcription, and orotein synthesis in the dying cell (25-27; reviewed in 18). Addition of the protein synthesis inhibitor cycloheximide, or of Cyclosporin A, both of which prevent activation-induced apoptosis in immature thymocytes (20), prevented activation-induced death by apoptosis of T cells from HIV-infected asymptomatic individuals in response to PWM and SEB (Fig. 2, 4a). Restoration of the capacity of CD4+ T cells to proliferate to stimuli. Protein synthesis inhibitors and Cyclosporin A, which prevented apoptosis, also prevented T-cell proliferation in response to stimuli. It has been shown in immature thymocytes that the addition of certain cosignals such as Interleukin (IL)-1, IL-2 and phorbol esters, not only prevent apoptosis, but also allow a proliferative response to stimuli (22,24,29,30). Addition to T cells from WHlV-infected asymptomatic individuals of IL-1, IL-2 or phorbol esters neither prevented apoptosis nor allowed proliferation in response to PWM or SEB (not shown). The CD28 mAb cosignal, that enhances in normal mature activated CD4+ T cells the stability and transcription of several lymphokines mRNA (37,38), has been reported to allow proliferation of normal immature thymocytes to stimuli (39), and to enhance in T cells from HIV-infected individuals the proliferative response Groux, Torpier, Monté, Mouton, Capron, Ameisen 11 to the CD3 antibody (15). As shown in Fig. 5, addition of CD28 mAb to PBMC or purified CD4+ T cells from HIV-infected asymptomatic individuals prevented PWM- and SEB-induced death by apoptosis (Fig. 2, 3b, and 4a) and restored T-cell proliferation in response to PWM and SEB (Fig. 5). CD28 mAb restored proliferation to PWM and SEB of purified CD4+* T cells, but not of purified CD8+ T cells from an HIV-infected asymptomatic individual (not shown). CD28 mAb alone did not induce proliferation of T cells from H|V-infected asymptomatic individuals (Fig. 5), and CD28 mAb did not enhance control T-cell proliferation to PWM, SEB or TT. Two different CD28 mAb (CLB28/1, IgG and 9.3, lgG2a isotype) were used and had the same effect, while control mAbs of the same isotype (CD5, IgG1, and CD44, IgG2) had no effect. CD28 mAb not only restored proliferation of T cells from HIV-infected individuals to PWM and SEB, but also to the tetanus recall antigen (Fig. 5), indicating that tetanus-specific memory T cells were present in the HIV-infected individuals, and suggesting that induction of apoptosis might account for their in vitro unresponsiveness to the recall antigen. Antigen-mediated functional deletion of antigen-specific T cells. Whether antigen may induce selective in vitro deletion of specific memory T cells was further investigated. PBMC from 4 HIV-infected asymptomatic individuals and from 2 controls were first incubated for 10 days with the tetanus antigen in the absence of CD28 mAb. At day 10, cell mortality was 12 to 15% in PBMC from HIV-infected individuals and from Groux, Torpier, Monté, Mouton, Capron, Ameisen 12 controls. Cells were then layered on Ficoll Hypaque, washed twice and incubated for 3 days with PHA or for 6 days with the tetanus or the influenza recall antigens, in the absence or in the presence of the CD28 mAb. Cells from HIV-infected asymptomatic individuals that had been first treated with tetanus recall antigen retained their capacity to proliferate to the influenza recall antigen in the presence of CD28 mAb, but selectively lost their subsequent capacity to proliferate to tetanus in the presence of CD28 mAb, suggesting that tetanus-specific memory T cells had been selectively deleted during the first incubation with the tetanus antigen (Table 1). As also shown in Table 1, preincubation of PBMC from controls with the tetanus antigen did not lead to any subsequent functional impairement. Groux, Torpier, Monté, Mouton, Capron, Ameisen 13 DISCUSSION Our results show that the selective in vitro proliferative defect of CD4+ T - cells from HIV-infected individuals to PWM and to self-MHC-lIl-dependent TCR mobilization by superantigens is related to the induction by these stimuli of an active CD4+ T-cell death process by apoptosis. Activation-induced cell death was not observed in mononuclear cells depleted in CD4*+ T lymphocytes, and occurred in purified CD4* T-cell populations, suggesting that the presence of CD4* T cells was both necessary and sufficient for the induction of this cell death process. Although CD4+ T cells failed to proliferate to TCR mobilization by the self-MHC-ll-dependent tetanus and influenza recall antigens, activation- induced CD4+ T-cell death in response to these antigens could not be detected. This may be related to the fact that memory T cells specific for a given recall antigen are rare and that activation-induced cell death spares bystander cells (18), or alternately, as previously suggested, that the antigen-specific memory T cells have been already depleted jn vivo (15,16). We observed that addition of a CD28 mAb co-signal that prevented apoptosis and restored T-cell proliferation in response to PWM and to superantigens, also restored T-cell proliferation to the tetanus and influenza _ recall antigens, indicating thus that the specific memory T cells were present in HIV-infected asymptomatic individuals. Preincubation of T cells with the tetanus antigen in the absence of CD28 mAb led to a subsequent selective Groux, Torpier, Monté, Mouton, Capron, Ameisen 14 loss of their capacity to proliferate to this antigen in the presence of the antibody, while the T-cell proliferative response to the influenza antigen in the presence of CD28 was not impaired. This suggested that antigen-specific activation-induced CD4+* T-cell death was the mechanism most likely to account for the failure of the memory T cells to proliferate to these recall antigens. In HiV-infected asymptomatic individuals, less than 0.1% of peripheral blood CD4t T cells are infected (8-10). Since in vitro activation with PWM resulted in death of around 40% of the CD4* T cells, the possibility that apoptosis occurred only in HIV-infected CD4+ T cells could be excluded. Recent observations of apoptosis in mature murine CD4t+t T cells (33,40) suggest at least two indirect mechanisms that may account for the re-emergence in the CD4+ T-cells of an activation-induced death program. First, CD4+ T cells from HIV-infected asymptomatic individuals may be primed in vivo for apoptosis upon further activation. Pretreatment of mature murine CD4t+ T cells with CD4 antibody has been shown to prime them for apoptosis upon selective mobilization of their a8 TCR but not of their CD3 complex (33), a response ressembling that of CD4+ T cells from HIV-infected individuals. Obvious candidates for such an in vivo priming include the binding to CD4 of secreted HIV-gp120 envelope protein released in serum or lymph (41), gp120-anti-gp120 antibody immune complexes, or anti-CD4 autoantibodies. However, preliminary results obtained in our laboratory suggest that preincubation of normal mature human CD4+ T cells Groux, Torpier, Monté, Mouton, Capron, Ameisen 15 with CD4 antibody or gp120, whether cross-linked or not, does not lead to apoptosis upon further simulation. A second possibility is that CD4+ T cells from HIV-infected individuals have no intrinsic abnormalities, but that defective antigen presenting cell function accounts for induction of T-cell apoptosis. Whether antigenpresenting cell from HIV-infected individuals are unable to provide relevant co-signals required for a proliferative response to PWM, or to MHC-Il-dependent TCR mobilization, is currently under investigation. It has been shown that restimulation of a mature murine CD4+* T-cell clone by cross-linked CD3 antibodies in the absence of antigen presenting cell results in an active cell death process that involves interferon y, and is prevented by anti-interferon y antibody (40). Our preliminary data, however, indicate that anti-interferon y antibody does not prevent activation-induced death of CD4* T cells from HliV-infected individuals. A third possible interpretation of our findings, that cannot be completely excluded, is that apoptosis is the consequence of a CD4t T-cell/CD4* T-cell killing process. Induction of apoptosis in their target cells is one of the means by which cytotoxic T lymphocytes (CTL) kill their targets (18,42). Altough cytotoxic properties have been mainly ascribed to a sub-population of CD8* T cells, CD4+ CTL clones have been described (42). Since cyclosporin A, which does not prevent CD8*+ CTL- or CD4+ CTL-mediated apoptosis of target cells (42), prevented apoptosis of CD4* T cells from HIV-infected individuals, we think that an activation-induced CD4t T-cell Groux, Torpier, Monté, Mouton, Capron, Ameisen 16 suicide process, in the absence of any participation of CD4* killer T cells, represents at this stage the simplest explanation for our observations. Our findings suggest that activation-induced T-cell death may occur in vivo and account for the progressive depletion of CD4+ T cells that leads to AIDS. In vitro proliferation assays of CD4*+ T cells in response to various recall antigens in the presence of CD28 mAb, should allow one to assess at any given time the extent of the memory CD4+ T-cell repertoire that is remaining in vivo. In particular, it will be possible to test whether CD4+t T cells specific for pathogens continuously present in HIV-infected individuals, such as HIV itself, herpes virus, or cytomegalovirus, are deleted earlier in vivo than CD4+ memory T cells specific for pathogens that have be been rarely (influenza) or never (tetanus) encountered subsequent to HIV infection. Over the years, CD4+ T-cell depletion may extend beyond the T-cell repertoire specific for antigens present in the patient. Infection by pathogens that produce superantigens, such as staphylococci or streptococci (31), could induce the deletion of a large number of CD4+ T cells expressing the matching Vp molecules, independently of their antigen-specificity. Whatever the stage of the disease, the in vivo rate of CD4+ T-cell depletion should directly depend on the percentage of activated T cells and not on the percentage of HlV-infected T cells. In fact, since CD4*+ T-cell activation is required for H|IV-provirus integration (43), and since activation will result in a rapid cell deletion through apoptosis, this T-cell suicide process could provide an Groux, Torpier, Monté, Mouton, Capron, Ameisen 17 explanation for the very low percentage of HIV-infected T cells in HiV-infected individuals (8-10). Finally, our observation that cyclosporin A and the CD28 mAb prevent activation-induced death of CD4* T cells from HiV-infected asymptomatic individuals, and that CD28 mAb also restores their proliferative response to stimuli, might have implications for the design of specific therapeutic strategies aimed at the prevention of AIDS. Animal models of AIDS-related diseases should allow testing the potential beneficial effect of early in vivo correction of CD4+ T-cell apoptosis on the further evolution of CD4+ T-cell counts, and on the course of the disease. ACKNOLEDGEMENTS We thank A. Bernard, B. Axelsson, Oncogen Corp., and R. vanLier, for providing mAbs, W. Greene for IL-2R cDNA probes; B. Plouvier for expert technical assistance; J.M. Bourez _ for clinical work; M. Houache and C. Sartiaux for flow cytofluorometry analysis; F. Ameisen for helpful discussions. This work was supported by INSERM, CNRS, ANRS, European Community (TS 2005F); and by a FERS fellowship to H.G. Groux, Torpier, Monté, Mouton, Capron, Ameisen 18 REFERENCES 1. Lane, H.C., Depper, J.M., Greene, W.C., Whalen, G., Waldmann, T.A., & Fauci, A.S. (1985) New Engl. J. Med. 313, 79-84 2. Shearer, G.M., Bernstein, D.C., Tung, K.S.K., Via, C.S., Redfield, R, Salahudain, S.Z., & Gallo, R.C. (1986) JU. Immunol. 137, 2514-2521 3. Hofmann, B., Jakobsen, K.D., Odum, N., et al. (1989) J. Immunol. 142, 1874-1880 4. Miedema, F., Petit, A.J.C., Terpestra, F.G., Eeftinck Schattenkerk, J.K.M., DeWolf, F., Al, B.J.M., Roos, M., Lange, J.M.A., Danner, S.A., Goudsmit, J., & Schellekens, P.T.A. (1988) J. Clin. Invest, 82, 1908-1914 5. Clerici, M., Stocks, N.1., Zajac, R.A., Boswell, R.N., Lucey, D.R., Via, C.S., & Shearer, G.M. (1989) UJ. Clin. Invest. 84, 1892-1899 6. Clerici, M., Stocks, N.I., Zajac, R.A., et al. (1989) Nature, 339, 383-385 7. Fauci, A.S. (1988) Science 239, 617-622 8. Harper, M., Marselle, L., Gallo, R., & Wong-Staal, F. (1986) Proc. Natl Acad. Sci. USA 83, 772-776 9. Schnittman, S.M., Psallidopoulos, M.C., Lane, H.C., et al. (1989) Science 245, 305-308 10. Brinchmann, J.E., Albert, J., & Vartdal, F. (1991) JU. Virol. 65,2019-2023 11. Hofmann, B., Odum, N., Jakobsen, B.K., Platz, P., Ryder, L.P., Nielsen, J.O., Gerstoft, J., & Svejgaard, A. (1986) Scand. J. Immunol. 23, 669-678 12. Golding, H., Shearer, G.M., Hillman, K., et al. (1989) J. Clin. Invest. 83, 1430-1435 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24, 25. 26. Groux, Torpier, Monté, Mouton, Capron, Ameisen 19 Mann, D.L., Lasane, F., Popovic, M., Arthur, L.O., Robey, W.G., Blattner, W.A., & Newman, N.J. (1987) J. /mmunol. 138, 2640-2644 Viscidi, R.P., Mayur, K., Lederman, H.M., & Frankel, A.D. (1989) Science 246, 1606-1608 Schnittman, S.M., Lane, H.C., Greenhouse, J., et a/., (1990)Proc. Nat! Acad. Sci. USA 87, 6058- Van Noesel, C.J.M., Gruters, R.A., Terpstra, F.G., Schellekens, P.T.A., van Lier, R.A.W., & Miedema, F. (1990) JU. Clin.Invest. 86, 293-299 Ameisen, J.C., & Capron, A. (1991) Immunol. Today , 4, 103-105 Duvall, E., & Wyllie, A.H. (1986) Immunol. Today 7, 115-19 Smith, C.A., Williams, G.T., Kingston, R., Jenkinson, E.J., & Owen, J.J.T. (1989) Nature 337, 181-184 Shi, Y., Sahai, B.M., & Green, D.R. (1989) Nature 339, 625-626 Jenkinson, E.J., Kingston, R., Smith, C.A., Williams, G.T., & Owen, J.J.T. (1989) Eur. J. Immunol. 19, 2175-2177 McConkey, D.J., Hartzell, P., Amador-Perez, J.F., Orrenius, S., & Jondal, M. (1989) J. Immunol. 143, 1801-1806 Nieto, M.A., Gonzalez, A., Lopez-Rivas, A., Diaz-Espada, F., & Gambon, F. (1990) J. Immunol. 145, 1364-1368 Blackman, M., Kappler, J., & Marrack, P. (1990) Science 248, 1335-1341 Williams, G.T., Smith, C.A., Spooncer, E., Dexter, M., & Taylor, D.R. (1990) Nature 343, 76-79 Ucker, D.S., Ashwell, J.D., & Nickas, G. (1989) J. Immunol. 143, 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Groux, Torpier, Monté, Mouton, Capron, Ameisen 20 3461-3469 Odaka, C., Kizaki, H.. & Tadakuma, T. (1990) J. Immunol. 144, 2096-2101 Wyllie, A.H., & Morris, R.G. (1982) Am. J. Pathol. 109, 78-87 McConkey, D.J., Hartzell, P., Chow, S.C., Orrenius, S., & Jondal, M. (1990) J. Immunol., 145, 1227-1230 McConkey, D.J., Orrenius, S., & Jondal, M. (1990) /mmunol. Today, 11, 120-121 Marrack, P., & Kappler, J. (1990) Science 248, 705-711 Groux H., Huet, S., Valentin, H., Pham, D., & Bernard, A. (1989) Nature 339, 152-154 Newell, M.K., Haughn, L.J., Maroun, C.R., & Julius, M.H. (1990) Nature 347, 286-289 Miedema, F., Tersmette, M., & VanLier, R.A.W. (1990) Immunol. Today, 11, 293-297 Schellekens, P.T.A., Roos, M.T.L., DeWolf, F., Lange, J.M.A., & Miedema, F. (1990) J. Clin. Immunol., 16, 121-127 Wyllie, A.H., Kerr, J.F.R., & Currie, A.R. (1980) Int. Rev. Cytol. 68, 251-306 Lindsten, T., June, C.H., Ledbetter, J.A., Stella, G., & Thompson, C.B. (1989) Science 244, 339-343 Fraser, J.D., Irving, B.A., Crabtree, G.R., & Weiss, A. (1991) Science. 251, 313- Huet, S., Groux, H., Caillou, B., Valentin, H., Prieur, A.M., & Bernard, A. Groux, Torpier, Monté, Mouton, Capron, Ameisen 21 (1989) J. Immunol. 143, 798-801 40. Liu, Y., & Janeway C. (1990) J. Exp. Med. 172, 1735-1739 41. Pinching, A.J., & Nye, K.E. (1990) Immunol. Today 11, 256-259 42. Ju, S. (1991) J. Immunol. 146, 812- 43. Gowda, S.D., Stein, B., Mohagheghpour, N., Benike, C.J., & Engleman, E.G. (1989) J. Immunol. 142, 773-780 Groux, Torpier, Monté, Mouton, Capron, Ameisen 22 LEGENDS FIG. 1. Proliferative response of T cells from HIV-infected asymptomatic individuals and from controls. Histograms represent the mean + S.D. of triplicate measurement of 3[H]-thymidine incorporation of T cells from 38 HIV-infected asymptomatic individuals ( ) and from 20 HlV-seronegative controls ( ) incubated with medium alone, phytohemagglutinin (PHA) (10ug/ml), pokeweed mitogen (PWM) (10yug/ml), superantigens (SEB) (1ug/ml), and the tetanus toxoid (TT) recall antigen (10ug/ml). * = p<0.00004. Groux, Torpier, Monté, Mouton, Capron, Ameisen 23 FIG. 2. Cell death in peripheral blood mononuclear cells from HIV-infected asymptomatic individuals and from controls after jin vitro culture with PWM and SEB. Histograms represent the mean + §S.D. of triplicate measurements of the percentage of cell mortality in PBMC from 20 different HIV-infected asymptomatic individuals (1) and 20 controls (2), 48hr after incubation with medium alone, PWM (10ug/ml), or SEB (1ug/mi), in the absence or in the presence of cycloheximide (50ug/ml), cyclosporin A (Cyclosporin A) (200ng/ml), or CD28 mAb (10ug/ml). Cell death was assessed by trypan blue permeability. “ = p<0.001 Groux, Torpier, Monté, Mouton, Capron, Ameisen 24 FIG. 3.Cell death in purified T-cell populations from HIV-infected asymptomatic individuals and from controls after jin vitro culture with PWM and SEB. a) Histograms represent the mean + S.D. of triplicate measurement of cell mortality in CD8* T cell- or CD4+ T cell-depleted PBMC from a HIV-infected asymptomatic individual (1) and a control (2). T-cell proliferation was also measured by 3[H]-thymidine incorporation after 3 days; the CD8+ T cell-depleted PBMC from a control proliferated to PWM (5,250+651cpm) and to SEB (21,642+2,317cpm), whereas neither the CD4+ T-cell-depleted PBMC from the control, nor the CD8+ or CD4* T-cell-depleted PBMC from the HIV-infected asymptomatic individual proliferated to either stimuli. b) Histograms represent the mean + S.D. of triplicate measurement of cell mortality in purified CD4+ and CD8+ T cells from a HIV-infected asymptomatic individuals (1) and a control (2). T-cell proliferation was also measured by °%[H]-thymidine incorporation after 3 days; purified CD4+ and CD8t+t T cells from the HIV-infected individual and from the control proliferated normally to PHA. CD4+ T cells from the control proliferated in response to PWM (4,817 +351 cpm) and to SEB (21,892 + 1,628 cpm), whereas CD8+ T cells from the control and CD4+ and CD8+ T cells from the HIV-infected individual did not proliferate to either stimuli. Groux, Torpier, Monté, Mouton, Capron, Ameisen 25 FIG. 4. Apoptosis of T cells from HIV-infected asymptomatic individuals in response to PWM or SEB. a) DNA fragmentation in PBMC from HIV-infected asymptomatic individuals and controls after overnight incubation with medium alone, PWM (5yg/ml) or SEB (1ug/ml) in the absence or presence of Cyclosporin A (200ng/ml) or CD28 mAb. DNA weight markers (lane 1); control PBMC with PWM (lane 2): PBMC from 8 HIV-patients with: medium (lanes 3, 12), PWM (lanes 4-9, 13) or SEB (lanes 10,11). Lanes 3-4 and lanes 12-14: cells from the same patients. Lanes 4-6 and 8-11 and 13 show a clear ladder of degraded DNA bands which are multiple of 200 base pairs, characteristic for apoptosis (10-13, 17-19). DNA fragmentation was suppressed when Cyclosporin A (lane 7, cells from same patient as lane 6) or CD28 mAb (lane 14, cells from same patient as lane 13) were added to PWM. b and c) Electron micrographs of PBMC from a HlV-patient after 24h incubation with b: medium or c: PWM (5ng/ml). Cells showing various stages of chromatin condensation, characteristic for apoptosis (10, 17, 19, 20), can be seen in © (arrows). Bars= 5um. Groux, Torpier, Monté, Mouton, Capron, Ameisen 26 FIG. 5. Effect of CD28 mAb on the proliferative response of T cells from HIV-infected asymptomatic individuals. Circles represent the mean + S.D. of triplicate measurement of 3[H]-thymidine incorporation of T cells from each of 19 HIV-infected asymptomatic individuals incubated with PWM, SEB, TT, or medium alone, in the absence ( ) or presence ( ) of CD28 mAb. (Certain circles representing similar values are superposed). * = p<0.00004 Groux, Torpier, Monté, Mouton, Capron, Ameisen 27 TABLE 1. Antigen-mediated functional deletion of antigen-specific memory T cells from HIV-infected asymptomatic individuals. Results represent mean + SD of triplicate measurements of 3[H]-thymidine incorporation of PBMC from 4 HIV-infected asymptomatic individuals and from 2 controls. Results showing functional deletion are underlined. Preincubation with TT: PBMC (2.5 10®/ml were cultured for 10 days with the tetanus recall antigen (TT) (10ug/ml). Cells were then incubated with different stimuli, and proliferation measured by 3[H]-thymidine incorporation after 3 days for PHA or PWM and 6 days for TT or Influenza A recall antigens (Infl) (10yg/ml), in the presence or absence of CD28 mAb (10ug/ml). 30000 7 & 2. 20000 4 o 10000 _ 0 MEDIUM PHA EA HIV-patients FAcontrols TT 2 MEDIUM 4 medium F] cycloheximide cyclosporin A [1 CD28 mAb Qo oO °o A _- 5074 40 7 30 7 Ayyeyiow |ja9 % mortality cell % mortality % cell 40 7 30 4 20 5 10 - 40 - 30 4 20 7 10 - a [] PBMC ME CD8 depleted CD4 depleted MEDIUM PWM SEB PBMC CD4 cells + CD28 mAb CD8 cells + CD28 mAb SE] NM O 2 “1 2) MEDIUM PWM SEB —— — TS cpm 25000 7 20000 7 15000 10000 - 5000 ~ Gunung MEDIUM PWM SEB .Thymidine incorporation (cpm) Preincubation followed by stimulation with medium CD28 TT TT+CD28 Infl Infl+CD28 PHA No preincubation 121452 314487 236478 26174321 341458 32174206 186214817 HIvt 1 with TT 614487 258494 3181497 415487 721475 67314345 171824751 No preincubation 447434 313457 317427 57364187 151457 38424137 179804243 HIV+ 2 with TT 221467 318461 176450 118+21 345+46 53514142 164414419 No preincubation 324438 249464 437431 37814108 438+51 62424107 281434243 HIVt 3 with TT 479461 408437 459+61 438+71 386448 26744157 245974267 No preincubation 3504217 4054104 1574117 15494817 108+64 15604258 2823246712 HIV+ 4 with TT 103437 207483 4834121 3424106 4354265 21704598 2331544612 No preincubation 6804217 354470 39304881 32904721 23144567 32544432 3121742642 Control 1 with TT 5694258 208462 70914932 43824792 20904589 18504678 2721844812 No preincubation 477481 387473 48304523 40874238 57214267 42394540 205874267 Control 2 with TT 297454 579469 57614287 59834345 35494138 48204534 315894354 TABLE 1