Journal of General Virology (1998), 79, 415–421. Printed in Great Britain ................................................................................................................................................................................................................................................................................... Cytotoxic T lymphocytes in AIDS pathogenesis : lessons to be learned from the macaque model of simian immunodeficiency virus infection Anna Maria Geretti,1, 2 Ellen Hulskotte1 and Albert D. M. E. Osterhaus1 1 2 Institute of Virology, Erasmus University, PO Box 1738, 3000 DR Rotterdam, The Netherlands Department of Virology, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF, UK Introduction General aspects of SIV infection of macaques Increasing evidence indicates a protective role for cytotoxic T lymphocytes (CTL) in the host defence against human immunodeficiency virus (HIV). CTL reactive against HIV antigens have been detected in persons exposed to the virus but lacking evidence of infection. These include a small number of female prostitutes in Africa, sexual partners of infected persons, children born to infected mothers and health care workers exposed to infectious body fluids (reviewed by Rowland-Jones & McMichael, 1995). Based upon the assumption that CTL induction requires endogenous synthesis of viral proteins, these findings suggest that transient HIV infection and virus clearance by CTL are indeed possible. During acute HIV infection, the detection in circulation of specific CTL coincides with the fall in viraemia that follows the initial virus burst (Borrow et al., 1994 ; Koup et al., 1994). In the subsequent stages, persistence of antiviral CTL may be associated with a prolonged asymptomatic period, whereas their decline is usually coincident with disease progression (Klein et al., 1995). Nevertheless, a few patients with advanced disease may maintain relatively strong CTL responses against the core HIV Gag protein (Geretti et al., 1996). Studies on the relationship between virus load and CTL responses in chronically infected patients have also yielded conflicting results, as both direct (Ferbas et al., 1995) and inverse (Klein et al., 1995) correlations have been reported. While further work is needed to clarify such discrepancies, it is clear that antiviral CTL responses can only be interpreted in the framework of the dynamics of virus replication, recognition and clearance of infected cells, and virus attempts to evade immune surveillance. Infection of macaques with simian immunodeficiency virus (SIV) offers a valuable model for studying the complex interaction between lentiviruses and the host immune system. The aim of this review is to outline how this model has contributed to our understanding of the role of CTL in the control of lentiviral infections, and how its full potential may be exploited in the future. SIV of macaques (SIVmac) was first isolated in 1985 from animals held in captivity in the United States. Macaques, however, are not natural hosts of the virus. It is believed that cross-species transmission from sooty mangabeys, which are healthy natural carriers of SIV (SIVsm), was the source of infection. Fighting and biting among animals co-housed in outdoor corrals was the most likely route of horizontal transmission (Gardner, 1996). SIVmac shares with HIV many biological and structural features, including its tropism for the CD4 receptor and CCR5 co-receptor (Chen et al., 1997), a similar genomic organization and extensive genetic homology. The virus establishes persistent infection in macaques and causes an immunodeficiency syndrome closely resembling human AIDS. As in humans, prominent features of the disease include CD4+ cell loss, constitutional symptoms, lymphadenopathy, skin and neurological disorders, opportunistic infections and neoplasias. Although the average time to disease is 1–2 years, the course of infection varies considerably among macaques. Even after infection with the same viral molecular clone, some animals rapidly develop disease and die within months, whereas others may remain asymptomatic for a few years, thus resembling long-term non-progressors with HIV infection. This variability, combined with the opportunity to define parameters of infection such as strain, dose and route of virus inoculation, may prove valuable for clarifying the role of CTL immunity in lentivirus containment. Author for correspondence : Anna Maria Geretti (mail to be sent to address 2). Fax ­44 171 8302854. e-mail geretti!rfhsm.ac.uk 0001-5130 # 1998 SGM Detection of SIV-specific CTL Target cells SIV-specific CTL responses are currently measured against autologous B lymphoblastoid cell lines (B-LCL) immortalized by herpesvirus papio. After infection with recombinant vaccinia virus vectors (rVV) encoding SIV proteins, these cells express endogenously processed antigen in the context of MHC class I molecules. The use of rVV expressing different regions of the SIV genome has shown that SIV-specific CTL target both structural and regulatory viral proteins, although the strength and magnitude of the response vary among animals (Venet et al., 1992 ; Geretti et al., 1997 a). Alternatively, EBF A. M. Geretti, E. Hulskotte and A. D. M. E. Osterhaus B-LCL can be sensitized for lysis by incubation with short synthetic viral peptides that bind directly to MHC class I molecules on the cell surface, thus allowing the definition of epitope specificities. Effector cells In vitro restimulation has almost universally been required to amplify SIV-specific CTL responses to detectable levels. As an interesting exception, direct cytotoxicity against the envelope protein (Env), Gag and the regulatory protein Nef has been recently observed with intestinal intraepithelial lymphocytes of two macaques infected with SIVmac #&" (Coudel-Courteille et al., 1997). One of the animals showed signs of advanced disease, including a marked colitis. Although a high proportion of intestinal intraepithelial lymphocytes express the CD8 marker, the exact nature and function of the activated intestinal effector cells, and their contribution to host defence or immunopathogenesis, remain to be determined. The same authors found little or no evidence of direct cytotoxicity in peripheral blood, spleen and lymph nodes. This is in line with the observation that the frequencies of circulating CTL precursors (CTLp) detected in SIV-infected or vaccinated macaques are generally lower than those often measured in asymptomatic HIV-infected adults (Geretti et al., 1997 a). While the requirement for appropriate methods of CTLp restimulation may explain the apparent lack of CTL induction in some vaccine studies, it also highlights the need for standardized assays which would facilitate comparison of different experiments, often including only small numbers of animals. One successful method for the expansion in vitro of HIV- or SIV-specific CTLp is based on stimulation with paraformaldehyde-fixed autologous B-LCL infected with rVV expressing HIV or SIV antigens (Van Baalen et al., 1993). Compared with non-specific methods of stimulation, this approach enhances CTL detection in macaques by inducing selective CTL expansion, with reduced interference from background lysis (Geretti et al., 1997 a). Cell culture under limiting dilution conditions appears to increase further the sensitivity of CTL measurement, as it allows CTLp detection, albeit at low frequencies, in macaques lacking significant responses in standard bulk CTL assays (Geretti et al., 1997 a). Autologous blasts infected with SIV have also been used successfully as stimulator cells (Gallimore et al., 1995). Alternatively, synthetic peptides spanning defined regions of the SIV genome can be used to stimulate the growth of epitope-specific CTL. Should vaccine-induced CTL prevent or limit virus replication? Early vaccine studies in macaques aimed primarily at preventing infection through the induction of virus-neutralizing antibodies. As a result, a protective role of CTL in some of the successful vaccination strategies reported, such as those based upon live attenuated SIV vaccines (reviewed by Ruprecht et al., 1996), can only be hypothesized. More recently, EBG the prevailing view that it may be desirable for a vaccine to prevent disease if not infection, and the appreciation that SIV infection of macaques provides a valuable model for studying HIV pathogenesis, coupled with the development of reliable means of detecting SIV-specific CTL, have drawn considerable attention to antiviral cell-mediated immunity. SIV Env-, Gag- or Nef-specific CTL have been induced in macaques through several immunization strategies, including live attenuated viruses (Cranage et al., 1997), proteins either expressed by live vectors such as rVV (Gallimore et al., 1995 ; Kent et al., 1996), incorporated into iscoms (Hulskotte et al., 1995) or adjuvanted by QS-21 saponin (Newman et al., 1994), non-infectious virus-like particles (Klavinskis et al., 1997), peptides in various formulations (Bourgault et al., 1994 ; Hulskotte et al., 1995 ; Yasutomi et al., 1995) and DNA plasmids (Lu et al., 1996). In those studies that have tested the outcome of subsequent challenge, the detection of vaccineinduced CTL has usually failed to predict complete resistance to infection. In one early report, vaccine-induced CTL against pC11, a well-defined epitope in Gag, failed to protect macaques from intravenous challenge with SIV Macaca nemestrina (SIVmne) (Yasutomi et al., 1995). The induction of relatively low frequencies of CTLp of limited specificity, and the absence of antibodies, were believed to have contributed to the lack of protection. However, in a subsequent report, immunization with Env- and Gag-iscoms and three lipopeptides spanning the central region of Nef induced both virus neutralizing antibodies and Env-, Gag or Nef-specific CTLp, but similarly failed to protect macaques from intravenous challenge with SIVmac H-J (Hulskotte et al., 1995). Furthermore, no cor$# & relation was found between the frequencies of SIV-specific CTLp detected before challenge and the levels of cellassociated virus load measured after infection. It is of note, however, that CTLp frequencies were in most cases below 26}10' peripheral blood mononuclear cells (PBMC). Nevertheless, even the animal with a relatively high frequency of Env-specific CTLp (105}10' PBMC) was not protected from infection. Recently, intravenous, intramuscular and gene gun inoculations of SIV DNA plasmids have produced similar results : despite the induction of virus neutralizing antibodies and Env-specific CTL, no protection was induced against intravenous challenge with SIVmac (Lu et al., 1996). #&" Although disappointing, these findings are consistent with the view that CTL may not be able to prevent or control SIV infection, unless stringent qualitative and quantitative requirements are met. This finds indirect support in the observation that macaques immunized with whole inactivated SIVmac H, $# and protected from intravenous challenge with cell-associated virus, share the Mamu-A26 MHC class I allele with the donor of the infected cells (Osterhaus et al., 1992 ; Heeney et al., 1994). Mamu-A26 positive animals also behave as long-term survivors (Bontrop et al., 1996), suggesting that gene products of this allele may be correlated with the ability to mount protective CTL responses against SIV. More direct evidence of Review : CTL in AIDS pathogenesis a protective role of defined CTL subpopulations is provided by a study of macaques immunized with SIV Nef rVV (Gallimore et al., 1995). Although six of seven animals became infected upon intravenous challenge with SIVmac H-J , the frequencies $# & of vaccine-induced Nef-specific CTLp measured before challenge were inversely correlated with the levels of virus load measured after infection. In addition, the animal with the highest CTLp frequency was protected from infection. As these promising findings indicate, the protective role of CTL responses targeting regulatory proteins expressed early during the virus replication cycle deserves further investigation. Recent findings appear to support the concept that protective CTL may limit rather than completely prevent virus replication. Vaccinated macaques lacking both detectable virus and antibody responses after either intravenous or intrarectal challenge with SIVmne showed CTL responses against SIV proteins present in the challenge virus but not in the vaccine (Kent et al., 1996). Similarly, macaques immunized with Enviscoms lacked both detectable virus and anamnestic antibody responses after intravenous challenge with a chimeric simian– human immunodeficiency virus (SHIV), but had CTLp against antigens other than Env, including the non-virion regulatory proteins Rev and Tat (E. G. J. Hulskotte, unpublished). These data indicate that CTL responses provide a highly sensitive marker of transient or low-level virus replication. It is tempting to speculate that similar silent or abortive infections may have remained undetected in some of the early vaccine studies. What is the role of antiviral CTL in mucosal immunity? As most HIV infections are acquired by mucosal routes, current efforts are directed at inducing immune responses that may prevent or limit virus spread after mucosal exposure. The SIV model offers the opportunity to explore issues related to mucosal immunity and infection that are difficult to address in humans. Indeed, evidence that protection from sexually transmitted SIV and, by inference, HIV may be induced by effectively stimulating genital and systemic antiviral CTL is increasing. In female macaques inoculated intravaginally with SIVmac , at least some (1 in 2425 to 1 in 26 686) of the CD8+ #&" cells recovered from the vaginal epithelium are SIV-specific CTLp directed against Env or Gag. In line with the view that progressive CTL compartmentalization occurs during persistent infection, the frequencies of genital CTLp are higher in chronically infected monkeys than in animals with recent infection (Lohman et al., 1995). SIV-specific CTL have also been detected in both peripheral blood and gut-associated lymph nodes of macaques infected intravenously with live attenuated SIVmacC , and resistant to intrarectal challenge with either ) SIVmac H-J or SHIV (Cranage et al., 1997). Furthermore, $# & inoculation of macaques with SIV p27 : Ty virus-like particles by either the rectal-oral and vagino-oral route, or subcutaneous immunization targeting the iliac lymph nodes, has induced specific CTL in the rectal and cervico-vaginal mucosa, as well as in regional lymph nodes, spleen and peripheral blood (Klavinskis et al., 1997). Finally, intra-vaginal inoculations with attenuated SHIV have induced at least partial resistance to intravaginal challenge with pathogenic SIVmac (Miller et al., #$* 1997). Although the presence of SIV-specific CTL in the genital tract was not determined in this study, protected animals had circulating Gag-specific CTL at the time of challenge, with or without specific antibodies in genital secretions. To explain the generation of local and systemic primary immune responses upon vaccine delivery to mucosal sites, antigen uptake by resident dendritic cells is hypothesized, followed by their migration to regional lymph nodes where naive T lymphocytes are stimulated. These would then enter the circulation and migrate to the genital site, where they may be restimulated upon re-exposure to the antigen, thus providing a first line of defence against sexually transmitted viruses. What is the relationship between SIV-specific CTL and virus replication? Kinetics studies have shown that the detection of Gag- or Nef-specific CTL by week 1 or 2 after intravenous SIV infection is coincident with the decrease in virus load and p26 antigenaemia that follows the initial virus burst (reviewed by Letvin et al., 1994). In the early phase of infection, virus load and SIV-specific CTL responses show similar kinetics in peripheral blood and lymph nodes. In contrast, during chronic infection, virus sequestration within lymphoid organs (Chakrabarti et al., 1994) is mirrored by CTL compartmentalization to the sites of infection (A. M. Geretti, unpublished). These observations are consistent with CTL being the effector cells of the in vivo immune response against SIV. Further support for a protective role of antiviral CTL comes from the observation that macaques with strong CTL responses against multiple SIV antigens, including the regulatory proteins Nef, Rev and Tat, remained free of disease for at least 2 years after intravenous infection with SIVmac . In contrast either absent, transient or #&" weak CTL responses were observed in animals rapidly progressing to overt disease (Venet et al., 1992). Preliminary findings also indicate that the detection of CTL responses against Rev and Tat in the early phase of infection may be associated with effective virus containment and long-term survival in both HIV-infected humans (Van Baalen et al., 1997) and SIV-infected macaques (A. M. Geretti, unpublished). These data are again in agreement with the hypothesis that CTL targeting early gene products may be effective in virus control before release of progeny virus occurs. Another point of interest is that the same CTL may also recognize latently infected cells that, although not actively producing virus, are known to express multiple spliced viral messenger RNAs encoding the regulatory proteins Nef, Rev and Tat (Seshamma et al., 1992 ; Embretson et al., 1993). EBH A. M. Geretti, E. Hulskotte and A. D. M. E. Osterhaus Fig. 1. Kinetics of circulating SIV-specific CTL precursors (CTLp) in macaques infected intravenously with SIVmac32H-J5. After the initial burst of virus replication, three monkeys (A) lost evidence of either culturable virus or PCR-detectable provirus in peripheral blood, while maintaining low-level virus reservoirs in spleen and lymph nodes ; these animals remained asymptomatic with stable CD4+ cell counts throughout 22 months of observation. Three other monkeys had persistent virus in peripheral blood and high virus load in lymphoid organs : one (B) developed AIDS 18 months after infection, whereas two others (C) remained asymptomatic but over time showed a significant CD4+ cell decline. CTLp frequencies were measured by limiting dilution assays. Autologous B-LCL infected with rVV expressing SIV antigens were used in vitro as stimulator and target cells. The cumulative frequencies of circulating SIV-specific CTLp, after reaching a plateau, declined significantly in the three macaques lacking detectable virus in peripheral blood (A) and in the animal which progressed to overt disease (B), but were maintained in the two asymptomatic macaques with persistent infection in peripheral blood (C). The relatively low frequencies of SIV-specific CTLp usually detected in SIV-infected macaques seem in contrast with the strong CTL responses often detected in asymptomatic HIVinfected persons. Besides the influence of host genetic factors, or the effects of progressive immunodeficiency, relatively low CTLp frequencies may reflect a low degree of antigenic stimulation in vivo, due to rapid down-regulation of SIV replication after infection. Consistent with this view, high CTLp frequencies can be detected occasionally in animals with a high virus load in their PBMC (Geretti et al., 1997 a). Conversely, the frequencies of circulating SIV-specific CTLp may decline significantly over time in asymptomatic SIVinfected macaques lacking detectable virus in PBMC and showing low-level virus reservoirs in lymphoid organs (Fig. 1 ; A. M. Geretti, unpublished). In view of these findings, it is not surprising that the degree of protection conferred by live attenuated SIV vaccines, and the strength of the Env- and Gagspecific CTL responses they induce, appear to be inversely correlated with the level of virus attenuation (Lohman et al., 1994). These data have direct implications for vaccine designs aimed at inducing long-lasting protective CTL responses : a successful vaccine will need to achieve a difficult balance between safety, which is dependent upon a high degree of attenuation, and efficacy, which appears directly related to the ability of the virus to replicate. Are virus-specific CTL deleterious to the host? It has been suggested that in a physiological attempt to eradicate persistent infection, CTL may also induce pathological changes deleterious to the host. This is supported by EBI the finding that HIV-specific CTL are present in both the bronchoalveolar lavage of AIDS patients with lymphocytic alveolitis and in the cerebrospinal fluid (CSF) of patients with AIDS dementia complex (reviewed by Zinkernagel, 1995). The detection of SIV-specific CTL in the skin rash of SIV-infected macaques (Yamamoto et al., 1992) also suggests a role for CTL in mediating tissue damage. SIV-specific CTL have been also detected in the CSF and brain of SIV-infected macaques as early as 1 week after infection and concomitant with the detection of virus (Von Herrath et al., 1995). Interestingly, different SIV proteins were targeted by CTL recovered from the brain, CSF and peripheral blood, suggesting that these can be separate compartments in SIV infection. As these interesting data indicate, the SIV model provides an excellent system to explore further the role of CTL in HIV-associated neurological disorders. Can SIV variants escape CTL recognition? Selection of mutant viruses resistant to specific CTL may be proposed as one mechanism whereby highly variable viruses such as HIV or SIV escape from immune recognition. Single amino acid substitutions within CTL epitopes can abrogate recognition by affecting either MHC binding or T cell receptor (TCR) interaction (Rothbard et al., 1989), whereas mutations within epitope flanking regions may reduce presentation by affecting peptide processing and transport (Eisenlohr et al., 1992). Recent studies also suggest that mutated epitopes may arise, which still interact with the TCR, but inhibit CTL function by inducing T cell anergy, or by showing partial agonistic or antagonistic activity (reviewed by Klenerman et al., 1996). In this case, even virus variants that do not become Review : CTL in AIDS pathogenesis the predominant viral species may affect immune surveillance, as they can block CTL recognition of viruses that do not contain mutated epitope sequences. Finally, rapid virus evolution and strong antigenic stimulation may also favour immune evasion by causing CTL exhaustion (Von Boehmer, 1993). Although mutant viruses which are resistant to CTL have been detected in persons infected with HIV (reviewed by Koup, 1994 ; Koenig et al., 1995 ; Borrow et al., 1997 ; Goulder et al., 1997), and despite evidence that multiple mechanisms may be operative in HIV attempts to escape recognition (Couillin et al., 1994), the biological relevance of these findings remains somewhat controversial. The fact that HIV-infected persons, or SIV-infected macaques, generally mount CTL responses against multiple viral antigens suggests that despite the loss of CTL recognition of one epitope, complete evasion from immune surveillance should be a rare occurrence. Nevertheless, qualitative aspects of antiviral CTL, such as the nature of the targeted epitope (Moskophidis & Zinkernagel, 1995) and the affinity of effector cell–target cell interactions (Tsomides et al., 1994) may be crucial determinants of effective CTL pressure. It is therefore conceivable that under stringent circumstances, strong CTL responses targeting key epitopes may promote selection of virus mutants which, at least temporarily, can evade immune recognition. The finding that the presence of strong CTL responses against Env is associated with pronounced HIV genetic heterogeneity (Wolinsky et al., 1996) appears to support this view. The effects of CTL pressure on SIV evolution have so far received limited attention. In a study of macaques infected with SIVmac , the presence of strong p11C-specific CTL responses #&" was associated with the detection of amino acid mutations within the CTL epitope (Chen et al., 1992). Although two of four mutated epitope sequences were recognized less efficiently than the prototype sequence, statistical analysis failed to indicate a higher rate of mutations in the epitope compared with other regions of Gag, and selection of the mutated viruses could not be demonstrated. Infection of macaques with molecular clones of SIV may help to clarify the significance of virus escape in pathogenesis and disease progression, as any mutation detected in the viral genome must have originated during the course of the infection. In a long-term non-progressor macaque infected with the molecular clone SIVmac H-J , the detection of ‘ high $# & affinity ’ immunodominant CTL targeting a 9-mer epitope in Gag designated p26A.5 (Geretti et al., 1997 b), coincided with the emergence of a variant virus carrying a mutated epitope sequence (A. M. Geretti, unpublished). Surprisingly, the mutated sequence was detected in the spleen, where it had uniformly replaced the challenge virus, but not in peripheral blood or lymph nodes. One explanation for this finding is that co-existence of viral expression and CTL expansion in splenic white pulp creates a highly favourable microenvironment for immunological pressure (Cheynier et al., 1994). The absence of epitope mutations in animals lacking detectable p26A.5- Fig. 2. Different patterns of epitope-specific CTL recognition. In a longterm non-progressor macaque infected with the molecular clone SIVmac32H-J5, the detection of CTL targeting a 9-mer epitope in Gag designated p26A.5 (amino acids 242–250 of p26) coincided with the emergence in the spleen of a variant virus carrying an aspartic acid to glutamic acid substitution (D ! E) at position 244. Short-term CTL lines generated in vitro by stimulation of PBMC with a peptide representing the prototype p26A.5 epitope were tested for their ability to lyse chromiumlabelled B-LCL pulsed with either the prototype p26A.5 peptide, or a variant peptide carrying the mutated epitope sequence. Four patterns of reactivity were identified : most (38 of 56) CTL lines (e.g. A01) did not recognize the variant peptide even at the highest peptide concentration tested (100 µM) ; eight (e.g. C12) recognized the prototype peptide with higher efficiency than the variant peptide ; seven (e.g. A05) recognized both the prototype and the variant peptide with a similar efficiency ; and three (e.g. F09) recognized the variant peptide with higher efficiency than the prototype peptide. specific CTL favours immunological pressure over simply growth advantage as the cause of virus mutation. The variant virus escaped recognition by most, but not all p26A.5-specific CTL (Fig. 2). The latter observation suggests a certain redundancy of CTL responses targeting the same epitope, also described for pC11-specific CTL (Shen et al., 1994 ; Chen et al., 1996), which may represent a first-line safeguard mechanism against emerging virus variants. Remarkably, CTL subpopulations targeting the variant sequence also localized preferentially in the spleen. The low virus load detected in this compartment suggests effective virus containment. However, the idea that a variant virus with even a slight advantage might have eventually replaced the challenge virus also in other compartments and gradually affected the control of the infection remains a valid hypothesis. Conclusions Over the past decade, the macaque model of SIV infection has provided significant new information on the host–virus interplay taking place during the course of lentiviral infections. Increasing evidence indicates a role for antiviral CTL in the containment of SIV replication, and supports the view that the quality of antiviral CTL is as important as their quantity in EBJ A. M. Geretti, E. Hulskotte and A. D. M. E. Osterhaus determining both the outcome of infection and the course of the disease. It is now believed that the ability to induce antiviral CTL that would limit, if not completely prevent virus spread, particularly after mucosal exposure, is an important prerequisite of an effective HIV vaccine. The concept that the design of such a vaccine will require improved understanding of what may constitute a protective CTL response is also now widely accepted. The SIV model offers valuable opportunities to better understand the mechanisms of lentivirus-induced disease and to clarify the basis for protective immunity. Hopefully, it will fulfil its promise – helping the development of a successful vaccination strategy against HIV. References Bontrop, R. E., Otting, N., Niphuis, H., Teeuwsen, V. & Heeney, J. L. (1996). The role of histocompatibility complex polymorphism on SIV infection in rhesus macaques. Immunology Letters 51, 35–38. Borrow, P., Lewicki, H., Hahn, B. H., Shaw, G. M. & Oldstone, M. B. (1994). Virus-specific CD8­ cytotoxic T-lymphocyte activity associ- ated with control of viraemia in primary human immunodeficiency virus type 1 infection. Journal of Virology 68, 6103–6110. Borrow, P., Lewicki, H., Wei, X., Horwitz, M. S., Peffer, N., Meyers, H., Nelson, J. A., Gairin, J. E., Hahn, B. H., Oldstone, M. B. A. & Shaw, G. M. (1997). Antiviral pressure exerted by HIV-1-specific cytotoxic T lymphocytes (CTL) during primary infection demonstrated by rapid selection of CTL escape virus. Nature Medicine 3, 205–211. Bourgault, I., Chirat, F., Tartar, A., Levy, J. P., Guillet, J. G. & Venet, A. (1994). Simian immunodeficiency virus as a model for vaccination against HIV. Induction in rhesus macaques of Gag- or Nef-specific cytotoxic T lymphocytes by lipopeptides. Journal of Immunology 152, 2530–2537. Chakrabarti, L., Cumont, M. C., Montagnier, L. & Hurtrel, B. (1994). Variable course of primary simian immunodeficiency virus infection in lymph nodes : relation to disease progression. Journal of Virology 68, 6634–6642. Chen, Z. W., Shen, L., Miller, M. D., Ghim, S. H., Hughes, A. L. & Letvin, N. L. (1992). Cytotoxic T lymphocytes do not appear to select for mutations in an immunodominant epitope of simian immunodeficiency virus Gag. Journal of Immunology 149, 4060–4066. Chen, Z. W., Shen, L., Regan, J. D., Kou, Z., Ghim, S. H. & Letvin, N. L. (1996). The T cell receptor gene usage by simian immunodeficiency virus Gag-specific cytotoxic T lymphocytes in rhesus macaques. Journal of Immunology 156, 1469–1475. Chen, Z., Zhou, P., Ho, D. D. and others (1997). Genetically divergent strains of simian immunodeficiency virus use CCR5 as a coreceptor for entry. Journal of Virology 71, 2705–2714. Cheynier, R., Henrichwark, S., Hadida, F., Pelletier, E., Oksenhendler, E., Autran, B. & Wain-Hobson, S. (1994). HIV and T cell expansion in splenic white pulps is accompanied by infiltration of HIV-specific cytotoxic T lymphocytes. Cell 78, 373–387. variations in the main immunogenic region of the HIV-1 Nef protein. Journal of Experimental Medicine 180, 1129–1134. Cranage, M. P., Whatmore, A. M., Sharpe, S. A., Cook, N., Polyanskaya, N., Leech, S., Smith, J. D., Rud, E. W., Dennis, M. J. & Hall, G. A. (1997). Macaques infected with live attenuated SIVmac are protected against superinfection via the rectal mucosa. Virology 229, 143–154. Eisenlohr, L. C., Yewdell, J. W. & Bennink, J. R. (1992). Flanking sequences influence the presentation of an endogenously synthesized peptide to cytotoxic T lymphocytes. Journal of Experimental Medicine 175, 481–487. Embretson, J., Zupanic, M., Ribas, J. L., Burke, A., Racz, P., TennerRacz, K. & Haase, A. T. (1993). Massive covert infection of helper T lymphocytes and macrophages by HIV during the incubation period of AIDS. Nature 362, 359–362. Ferbas, J., Kaplan, A. H., Hausner, M. A., Hultin, L. E., Matud, J. L., Liu, Z., Panicali, D. L., Nerng-Ho, H., Detels, R. & Giorgi, J. V. (1995). Virus burden in long-term survivors of human immunodeficiency virus (HIV) infection is a determinant of anti-HIV CD8+ lymphocyte activity. Journal of Infectious Diseases 172, 329–339. Gallimore, A., Cranage, M., Cook, N., Almond, N., Bootman, J., Rud, E., Silvera, P., Dennis, M., Corcoran, T., Stott, J., McMichael, A. & Gotch, F. (1995). Early suppression of SIV replication by CD8­ Nef-specific cytotoxic T cells in vaccinated macaques. Nature Medicine 1, 1167–1173. Gardner, M. B. (1996). The history of simian AIDS. Journal of Medical Primatology 25, 148–157. Geretti, A. M., Dings, M. E. M., van Els, C. A. C. M., van Baalen, C. A., Wijnholds, F. J., Borleffs, J. C. C. & Osterhaus, A. D. M. E. (1996). HIV- 1 and EBV specific cytotoxic T lymphocyte precursors exhibit different kinetics in HIV-1 infected persons. Journal of Infectious Diseases 174, 34–45. Geretti, A. M., Hulskotte, E. G. J., Dings, M. E. M., van Baalen, C. A., van Amerongen, G. & Osterhaus, A. D. M. E. (1997 a). Simian immuno- deficiency virus (SIV)-specific CD8­ cytotoxic T lymphocyte responses of naive and vaccinated cynomolgus macaques infected with SIVmac32H(J5) : quantitative analysis by in vitro antigenic stimulation. Vaccine 15, 1269–1275. Geretti, A. M., Hulskotte, E. G. J., Dings, M. E. M., van Baalen, C. A., van Amerongen, G. & Osterhaus, A. D. M. E. (1997 b). CD8+ cytotoxic T lymphocytes of a cynomolgus macaque infected with simian immunodeficiency virus (SIV)mac32H-J5 recognize a nine amino acid epitope in SIV Gag p26. Journal of General Virology 78, 821–824. Goulder, P. J. R., Phillips, R. E., Colbert, R. A., McAdam, S., Ogg, G., Nowak, M. A., Giangrande, P., Luzzi, G., Morgan, B., Edwards, A., McMichael, A. J. & Rowland-Jones, S. (1997). Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nature Medicine 3, 212–217. Heeney, J. L., Van Els, C., De Varies, P., Ten Haaft, P., Otting, N., Koornstra, W., Boes, J., Dubbes, R., Niphuis, H., Dings, M., Cranage, M., Norley, S., Jonker, M., Bontrop, R. E. & Osterhaus, A. (1994). Major histocompatibility complex class I-associated vaccine protection from simian immunodeficiency virus-infected peripheral blood cells. Journal of Experimental Medicine 180, 769–774. Coudel-Courteille, A., Le Grand, R., Tulliez, M., Guillet, J. G. & Venet, A. (1997). Direct ex vivo simian immunodeficiency virus (SIV)-specific Hulskotte, E. G. J., Geretti, A. M., Siebelink, K. H., van Amerongen, G., Cranage, M. P., Rud, E. W., Norley, S. G., de Vries, P. & Osterhaus, A. D. M. E. (1995). Vaccine-induced virus-neutralizing antibodies and cytotoxic activity detected from small intestine intraepithelial lymphocytes of SIV-infected macaques at an advanced stage of infection. Journal of Virology 71, 1052–1057. cytotoxic T cells do not protect macaques from experimental infection with simian immunodeficiency virus SIVmac 32H (J5). Journal of Virology 69, 6289–6296. Couillin, I., Culmann-Penciolelli, B., Gomard, E., Levy, J.-P., Guillet, J. G. & Saragosti, S. (1994). Impaired CTL recognition due to genetic Kent, S. J., Hu, S. L., Corey, L., Morton, W. R. & Greenberg, P. D. (1996). Detection of simian immunodeficiency virus (SIV)-specific ECA Review : CTL in AIDS pathogenesis CD8­ T cells in macaques protected from SIV challenge by prior subunit vaccination. Journal of Virology 70, 4941–4947. Rothbard, J., Pemberton, R., Bodmer, H., Askonas, B. A. & Taylor, W. R. (1989). Identification of residues necessary for clonal specific Klavinskis, L. S., Bermeier, L. A., Gao, L., Mitchell, E., Ward, R. G., Layton, G., Brookes, R., Meyers, N. J. & Lehner, T. (1997). Mucosal or recognition of a cytotoxic T cell determinant. EMBO Journal 8, 2321–2328. Rowland-Jones, S. & McMichael, A. (1995). Immune responses in HIVexposed seronegatives ; have they repelled the virus ? Current Opinion in Immunology 7, 448–455. targeted lymph node immunisation of macaques with a particulate SIV p27 protein elicits virus-specific CTL in the genito-rectal mucosa and draining lymph nodes. Journal of Immunology 157, 2521–2527. Klein, M. R., van Baalen, C. A., Holwerda, A. M., Kerkhof Garde, S. R., Bende, R. J., Keet, I. M. P., Eeftinck-Schattenkerk, J. M., Osterhaus, A. D. M. E. & Miedema, F. (1995). Kinetics of Gag-specific CTL re- sponses during the clinical course of HIV-1 infection : different for rapid progressors and long-term asymptomatics. Journal of Experimental Medicine 181, 1365–1372. Klenerman, P., Phillips, R. & McMichael, A. (1996). Cytotoxic T-cell antagonism in HIV-1. Seminars in Virology 7, 31–39. Koenig, S., Conley, A. J., Brewah, Y. A., Jones, G. M., Leath, S., Boots, L. J., Davey, V., Pantaleo, G., Demarest, J. F., Carter, C., Wannebo, C., Yannelli, J. R., Rosenberg, S. A. & Lane, H. C. (1995). Transfer of HIV- 1-specific cytotoxic T lymphocytes to an AIDS patient leads to selection for mutant HIV variants and subsequent disease progression. Nature Medicine 1, 330–336. Koup, R. A. (1994). Virus escape from CTL recognition. Journal of Experimental Medicine 180, 779–782. Koup, R. A., Safrit, J. T., Cao, Y., Andrews, C. A., McLeod, G., Borkowsky, W., Farthing, C. & Ho, D. D. (1994). Temporal association of cellular immune response with the initial control of viraemia in primary HIV-1 syndrome. Journal of Virology 68, 4650–4655. Letvin, N. L., Reimann, K. A., Yasutomi, Y., Ringler, D. J. & Yamamoto, Y. (1994). The SIVmac specific cytotoxic T lymphocyte response in the acutely infected rhesus monkey. Current Topics in Microbiology and Immunology 188, 175–184. Lohman, B. L., McChesney, M. B., Miller, C. J., McGowan, E., Joye, S. M., Van Rompay, K. K., Reay, E., Antipa, L., Pedersen, N. C. & Marthas, M. L. (1994). A partially attenuated simian immunodeficiency virus induces host immunity that correlates with resistance to pathogenic virus challenge. Journal of Virology 68, 7021–7029. Lohman, B. L., Miller, C. J. & McChesney, M. B. (1995). Antiviral cytotoxic T lymphocytes in vaginal mucosa of simian immunodeficiency virus-infected rhesus macaques. Journal of Immunology 155, 5855–5860. Lu, S., Arthos, J., Montefiori, D. C., Yasutomi, Y., Manson, K., Mustafa, F., Johnson, E., Santoro, J. C., Wissink, J., Mullins, J. I., Haynes, J. R., Letvin, N. L., Wyand, M. & Robinson, H. L. (1996). Simian immuno- deficiency virus DNA vaccine trial in macaques. Journal of Virology 70, 3978–3991. Miller, C. J., McChesney, M. B., Lu, X., Dailey, P. J., Chutkowski, C., Lu, D., Brosio, P., Roberts, B. & Lu, Y. (1997). Rhesus macaques previously infected with simian}human immunodeficiency virus are protected from vaginal challenge with pathogenic SIVmac239. Journal of Virology 71, 1911–1921. Moskophidis, D. & Zinkernagel, R. M. (1995). Immunobiology of cytotoxic T-cell escape mutants of lymphocytic choriomeningitis virus. Journal of Virology 69, 2187–2193. Ruprecht, R. M., Baba, T. W., Li, A., Ayehunie, S., Hu, Y., Liska, V., Rasmussen, R. & Sharma, P. L. (1996). Live attenuated HIV as a vaccine for AIDS : pros and cons. Seminars in Virology 7, 147–155. Seshamma, T., Bagasra, O., Trono, D., Baltimore, D. & Pomerantz, R. J. (1992). Blocked early-stage latency in the peripheral blood cells of certain individuals infected with human immunodeficiency virus type 1. Proceedings of the National Academy of Sciences, USA 89, 10663–10667. Shen, L., Chen, Z. W. & Letvin, N. L. (1994). The repertoire of cytotoxic T lymphocytes in the recognition of mutant simian immunodeficiency virus variants. Journal of Immunology 153, 5849–5854. Tsomides, T., Aldovini, A., Johnoson, R. P., Walker, B. D., Young, R. A. & Eisen, H. N. (1994). Naturally processed viral peptides recognized by cytotoxic T lymphocytes on cells chronically infected by human immunodeficiency virus type 1. Journal of Experimental Medicine 180, 1283–1293. Van Baalen, C. A., Klein, M. R., Geretti, A. M., Keet, I. M. P., Miedema, F., van Els, C. A. C. M. & Osterhaus, A. D. M. E. (1993). Selective in vitro expansion of HLA class I-restricted HIV-1 Gag-specific T cells : cytotoxic T-lymphocyte epitopes and precursor frequencies. AIDS 7, 781–786. Van Baalen, C. A., Pontesilli, O., Huisman, R. C., Geretti, A. M., Klein, M. R., De Wolf, F., Miedema, F., Gruters, R. A. & Osterhaus, A. D. M. E. (1997). HIV-1 Rev and Tat specific cytotoxic T lym- phocyte frequencies inversely correlate with rapid progression to AIDS. Journal of General Virology 78, 1913–1918. Venet, A., Bourgault, I., Aubertin, A. M., Kieny, M. P. & Levy, J. P. (1992). Cytotoxic T-lymphocyte response against multiple simian immunodeficiency virus (SIV) proteins in SIV-infected macaques. Journal of Immunology 148, 2899–2908. Von Boehmer, H. (1993). Immunology. Tolerance by exhaustion. Nature 362, 696. Von Herrath, M., Oldstone, M. B. & Fox, H. S. (1995). Simian immunodeficiency virus (SIV)-specific CTL in cerebrospinal fluid and brains of SIV-infected rhesus macaques. Journal of Immunology 154, 5582–5589. Wolinsky, S. M., Korber, B. T., Neumann, A. U., Daniels, M., Kuntsman, K. J., Whetsell, A. J., Furtado, M. R., Cao, Y., Ho, D. D. & Safrit, J. T. (1996). Adaptive evolution of human immunodeficiency virus -type 1 during the natural course of infection. Science 272, 537–542. Yamamoto, H., Ringler, D. J., Miller, M. D., Yasutomi, Y., Hasunum, T. & Letvin, N. L. (1992). Simian immunodeficiency virus-specific cytotoxic T lymphocytes are present in the AIDS-associated skin rash in rhesus monkeys. Journal of Immunology 149, 728–734. Newman, M. J., Munroe, K. J., Anderson, C. A., Murphy, C. I., Panicali, D. L., Seals, J. R., Wu, J.-Y., Wyand, M. S. & Kensil, C. R. (1994). Yasutomi, Y., Koenig, S., Woods, R. M., Madsen, J., Wassef, N. M., Alving, C. R., Klein, H. J., Nolan, T. E., Boots, L. J., Kessler, J. A., Emini, E. A., Conley, A. J. & Letvin, N. L. (1995). A vaccine-elicited, single viral Induction of antigen-specific killer T lymphocyte responses using subunit SIVmac251 gag and env vaccines containing QS-21 saponin adjuvant. AIDS Research and Human Retroviruses 10, 853–861. Osterhaus, A. D. M. E., De Vries, P. & Heeney, J. (1992). AIDS vaccine development. Nature 355, 684–685. epitope-specific cytotoxic T lymphocyte response does not protect against cell-free simian immunodeficiency virus challenge. Journal of Virology 69, 2279–2284. Zinkernagel, R. M. (1995). Are HIV-specific CTL responses salutary or pathogenic? Current Opinions in Immunology 7, 462–470. ECB