Role of Cellular Immunity in Protection against HIV Infection*

ADVANCES IN IMMUNOLOGY,VOL. 6 .5

Role of Cellular Immunity in Protection against HIV Infection SARAH ROWLAND-JONES, RUSUNG TAN, AND ANDREW McMlCHAEL Mokcukr immunobgy Group, lnstihtte of Mdeculor Medicine, John RadcI& Hospital, Wiwfun, Oxford OX3 OW, United Kingdom

1. Introduction

The cellular immune response is mediated by T lymphocytes that release cytokines and lyse target cells expressing foreign antigens. It generally occurs in parallel with the humoral (antibody) response, although the two can be separated in certain circumstances. Infection with viruses usually evokes both arms of the immune response, which broadly differ in their function: The cellular immune response controls the infection and the humoral response prevents further infection with the same agent. Protection of infants by transfer of maternal antibody is an important component of immune protection against infection in children (reviewed in Zinkernagel, 1996). Cytotoxic T lymphocytes (CTLs) are major contributors to the antiviral T cell immune response. This T cell population carries the CD8 cell surface glycoprotein and recognizes peptide antigens presented by class I major histocompatibility (MHC) molecules of the immune system (Zinkernagel and Doherty, 1974, 1975). When these cells make contact with antigen through their specific T cell receptor (TCR),provided this is accompanied by certain important cosignals, the T cell is activated to divide, differentiate, and mediate lysis of infected cells. The Iyhc process is caused both by release of perforin and through fas ligand triggering programmed cell death infas expressing cells (reviewed in Kagi et al., 1995a). CTLs can also release tumor necrosis factor alpha (TNF-a), which can potentiate killing, and interferon-? ( IFN-y), which has activities against a number of pathogens. The CTLs that initially expand on antigen contact can persist as memory cells: The number of memory cells specific for a particular antigen is usually higher than that found in unexposed animals. CD4positive T cells, the T helper (Th) cells, also have antipathogen activities. Th cells can be broadly grouped into three categories depending on the cytokines they are programmed to release: Thl cells produce IFN-.)I and IL-2, Th2 cells secrete IL-4, IL-5, and IL-10, and Tho clones make a mixture of cytokines including IFN-y, IL-2, and IL-4 (Mosmann, 1994; Mosmann et al., 1986). Thl cells play a particular role in defense against pathogens through direct cytotoxicity and by providing help for CTLs, and by means of the cytokines they produce, particularly IFN-y. 277

Copyright 0 1987 by Academic Press All rights of repduction in any form reserved. 0065-2776/97 $25.00

278

SARAH ROWLAND-JONES ET AL,

CTLs recognize small peptides derived from intracellular proteins, including those expressed by intracellular pathogens, bound to class I MHC molecules (Townsend and Bodmer, 1989). Foreign proteins are broken down in the cytosol of infected cells by the proteases of the proteasome complex (Goldberg and Rock, 1992). Peptide fragments are translocated to the endoplasmic reticulum (ER) by a transporter (TAP), where they bind to newly synthesized class I molecules in a groove formed by the Q helices of the aland a2domains (Townsend and Trowsdale, 1993). The MHC molecules, HLA in humans, are extremely polymorphic and most of the polymorphism occurs in the peptide-binding groove. The consequence of this is that different MHC allotypes bind different kinds of peptides. Th cells recognize peptides presented in a similar way by class I1 molecules of the MHC. Class II-associated peptides are normally derived from extracellular proteins taken into the cell and digested in lysosomes, which then meet the class I1 molecules in special endosomal compartments before export of the complexes to the cell surface (reviewed in Germain, 1991). This form of antigen processing normally involves specialized antigen presenting cells, the most potent of which is the dendritic cell (Steinman, 1991). 11. Cellular Immunity in the Control of Other Viruses

There is an extensive body of evidence that MHC class I-restricted CTLs play a central role in the control of intracellular microbial infections. CTLs were first demonstrated to be of importance in virus infections in lymphocyhc choriomeningitis (LCMV) infection in mice (Zinkernagel and Doherty, 1974, 1975). However, it is worth noting that their principal role in this infection is to mediate chronic immunopathology because the virus is not cytopathic (Buchmeier et al., 1980). Subsequently, CTLs were detected in murine influenza virus infection (Zweerink et al., 1977), and it was demonstrated that CTL clones transferred into infected mice had potent antiviral effects, which were largely mediated by killing virusinfected cells (Lin and Askonas, 1981; Lukacher et al., 1984). Similar observationswere made for respiratory syncybal virus (Cannon et al., 1988) and herpes simplex infection in mice (Bonneau and Jennings, 1990). CTLs have been demonstrated in many human virus infections (Bangham and McMichael, 1989). Evidence for a protective role in these infections has been harder to obtain. In influenza, CTL levels correlated with protection from deliberate infection ofvolunteers (McMichael et al., 1983). In immunosuppressed patients following bone marrow transplantation, CTL levels correlate with protection from cytomegalovirus (CMV) infec-

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

279

tion (Reusser et al., 1991) and CMV-specific CTL transfer seems to be an effective immunotherapy in this situtation (Walter et al., 1995a). Epstein-Barr virus (EBV) is an excellent model for human CTL-virus dynamics, providing lessons for the study of HIV. The virus causes acute infectious mononucleosis with high levels of CTL activity and huge expansions of oligoclonal CTLs in the blood (Callan et d., 1996). After clinical recovery, the virus persists in B cells but expresses a very limited range of viral gene products, probably only EBNA-1 in B cells (Rowe et aZ., 1987; Young et al., 1989). The virus has thus evolved a strategy for evading CTL responses and the one remaining gene product inhibits its own proteol p c degradation for presentation to CTLs (Levitskayaet al., 1995). Breakthrough expression of viral genes and transformation of lymphoid cells is controlled by a strong lifelong CTL response. However, immunosuppression, iatrogenic or during HIV infection, can result in the development of lymphomas, transformed by EBV (Beral et al., 1991; Rowe et al., 1991). The role of CTLs in preventing uncontrolled lymphoproliferation has been strongly supported by recent reports that it has been possible to treat some EBV-related lymphomas by cell transfer of EBV-specific CTLs (Rooney et al., 1995). Further support for the importance of CTLs in infections comes from the extreme polymorphism of the HLA class I system, particularly at the HLA-A and -B loci (Bodmer, 1972).The function of HLA class I molecules is to present peptides to T cells (Townsend et al., 1986); viruses are a major source of foreign peptides and it is likely that the polymorphism of the HLA class I system reflects evolutionary selection by intracellular pathogens (Hill, 1992). The best example is the protective effect of HLAB53 against severe life-threatening malaria in children in West Africa (Hill et al., 1991, 1992); the frequency of this allele is very high in this part of the world (0.25 compared with 0.01 in Europe). Conversely, in Southeast Asia HLA-A11 appears to have selected a variant of EBV that has a mutation in the immunodominant epitope that it presents to CTLs (de Campos Lima et al., 1993, 1994). Another indication of the importance of class I MHC-restricted T cell responses in control of virus infections is the finding that several viruses have evolved strategies to evade CTL recognition (Hill and Ploegh, 1995). Thus, adenovirus produces E19 proteins that retain class I MHC molecules in the ER (Andersson et al., 1985);herpes simplex virus expresses ICP47, which interferes with TAP-mediated peptide transport into the ER (Fruh et al., 1995; Hill et al., 1995); EBV EBNA 1 blocks its own proteolytic degradation by the proteasome (Levitskaya et al., 1995); and human CMV recycles nascent class I molecules into the cytosol, where they are degraded (Wiertz et al., 1996).All of these processes decrease class I MHC expression

280

SARAH ROWLAND-JONES ET AL.

on the surface of infected cells and facilitate evasion of CTL recognition. Indeed, these findings imply that persistent viruses have to develop some means of evading the CTL response. 111. CTL Effector Mechanisms

CTLs have two antiviral activities that can be measured in vitro. They can kill virus-infected cells, and this can be demonstrated in vivo as well (Kagi et al., 1994,1995b;Zinkernagel et al., 1986; Klenerman et al., 1996). They can release cytokines (Morris et al., 1982) and chemokines (Cocchi et al., 1995) with antiviral activity. These activities are not mutually exclusive, and all could play a part in anti-HIV activity in viuo. LYSIS OF HIV-INFECTED CELLS A. CTL-MEDIATED The lysis of HIV-infected cells as a means of controlling HIV infection has been assumed but little examined. Klenerman et al. (1996) and Yang et al. (1996) have studied the rate at which CTLs kill virus-infected cells in uitro. Klenerman and colleagues have argued that, by using CTLs taken ex vivo and testing antiviral lytic activity immediately (Walker et al., 1987), the rates could be close to those that are operative in viuo. Thus, at a ratio of peripheral blood mononuclear cells [approximately 1% of which are CTLs (Moss et al., 1995)] to target cells of 50: 1, i.e., 2% infected cells (Pantaleo et al., 1991, 1993), the half-life of infected cells in a patient with a strong ex vivo CTL response was about 12 hr; the range in different patients was found to be from 6 hr to 4 days. The average half-life of HIVinfected cells in vivo is close to 2 days and is independent of CD4+ cell counts (Ho et al., 1995; Wei et al., 1995). Zn vitro, infected cells start producing virus particles around 24-48 hr after infection (Yang et al., 1996). Thus, strong CTL responses could kill many virus-infected cells before production of new virus particles: Even allowing for the lag time of 24 hr before infected cells become targets for CTLs (Yang et al., 1996), a slightly less strong CTL response might kill many infected cells before they had produced their full complement of virions. If CTLs reduce virus production, the effect on the measured half-life (t3)of infected cells would be minimal because what is actually measured after antiretroviral drug therapy is the half-life of plasma virus, which is weighted toward the infected cells that produce the most virus; this could explain why the & of infected cells appears to vary so little over a range of CD4 T cell counts, and presumably also of CTL activities. These estimates assume that the frequency of CTLs in blood is similar to that in lymphoid organs, a reasonable assumption from the limited data available, showing similar levels in blood and lymph nodes (Hadida et al., 1995). According to this model,

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

28 1

the CTL would control the level of virus replication according on the strength of the CTL response; quite small changes in CTL activity could have large effects on virus production and virus load (Klenerman et al., 1996).These estimates, which are based on ex vivo lysis data and reasonable assumptions about the life cycle of infected cells, show that in the aymptomatic mid-phase of HIV infection, CTLs could kill most virus-infected cells, as has been independently suggested by Cheynier and Wain-Hobson (Cheynier et al., 1994; Wain Hobson, 1995). OF HIV REPLICATION BY CD8' CELLS B. SUPPRESSION There is considerable evidence that CD8+ cells play an important role in the control of HIV infection by a direct effect on viral replication. This was first demonstrated by Walker et al. (1986), who showed that HIV could readily be cultured from CD8+-depleted peripheral blood mononuclear cells (PBMCs) taken from healthy seropositive subjects, but that adding back the CD8+ cells suppressed virus production in a dosedependent manner. Further studies demonstrated that this anti-HIV activity is closely correlated with the clinical state and CD4' cell count of the infected individual (Mackewicz et al., 1991). Particularly vigorous activity has been described in a group of long-term nonprogressors (Cao et al., 1995), fuelling speculation that this may be one of the more important mechanisms controlling the length of the asymptomatic period in HIV infection (Levy, 1995). Potent suppression of HIV replication can occur across a semipermeable membrane or by transfer of the supernatant from CD8' cells, suggesting that it is mediated by release of soluble factors. There is no consensus as to whether or not this effect is primarily a property of class I-restricted CTLs, although most of the evidence suggests that these two effector functions are distinct. Although optimal suppression of viral replication was initially observed in HLA-matched CD4' cells, cytotoxicity is not involved: The infected cells are not eliminated from the culture, and removal of the CD8+ cells leads to resumption of HIV replication (Walker et al., 1991).In many studies potent suppression has been observed without any HLA matching (Brinchmann et al., 1990; Levy et al., 1996; Toso et al., 1995),but in other studies MHC class I matching provides maximal antiviral activity and the cells that mediate the suppression have the typical phenotype of CTLs (Tsubota et al., 1989). Some CTL clones have been shown to have potent antiviral activity (Koenig et al. 1995; 0. Yang and B. Walker, unpublished data; D. Nixon, T. Dong, and S. Rowland-Jones, unpublished data; Klenerman et al., 1996). However, the secretion of TNFa by other CTL clones in an antigen-specific manner can actually enhance viral replication in chronically infected T cell lines (Bollinger et al., 1993;

282

SARAH ROWLAND-JONESET AL.

Harrer et al. 1993).At the clonal level, up to 20% of the CD8+ T cell clones generated from an infected donor exhibit the phenomenon (Hsueh et al., 1994).Detailed analysis of a panel of CD8' clones from asymptomatic HIV-infected donors showed that the majority of suppressing clones did not exhibit HIV-specific cytolytic activity, and that some specific CTL clones showed no evidence of viral suppression-although a few clones had both properties (Toso et al., 1995). HIV-suppressing CD8+ cells have been shown to express certain cell surface markers: DR+,CDllb- (Mackewicz and Levy, 1992), CD28+ (Landay et al., 1993), CD29+, CD45RA-, LFA-1' (Tsubota et al., 1989), CD45ROt, and CD38+, but a diversity of other markers has been observed among suppressing clones, suggesting that CD8+ clones with antiviral activity are phenotypically heterogeneous (Toso et al., 1995). Maintenance of the CD8+ antiviral response in HIVinfected people appears to be dependent on the presence of Thl cytokines, particularly IL-2 (Barker et al., 1995). Suppression of many different strains of HIV-1, HIV-2, and SIV has been observed, both laboratory-adapted strains grown in transformed cell lines and patient isolates grown in primary T cells, although the conditions for suppression show some differences between different systems (Levy, 1993).It has been demonstrated that CD8' cells suppress HIV replication at the level of LTR-driven transcription (Levy et al., 1996),and this activity can extend to the LTRs of other viruses, such as HTLV-1 and Rous sarcoma virus (Copeland et al., 1995). Some T cell clones from uninfected people may also have this property (Hsueh et al., 1994), particularly activated T cells such as allostimulated CD8' cells (Bruhl et al., 1996), but these generally suppress HIV replication only in the "endogenous system," using CD8+-depleted cultures from infected donors (Mackewicz and Levy, 1992). The CD8' antiviral factor (CAF) effect could not be assigned to any known cytokine, although some reversal of the activity was seen with antibodies to IL-8 (Mackewicz et al., 1994a). Although IFN-.)I has some viral-suppressive activity (Emilie et al., 1992) and is produced by CTLs after antigen-specific contact, it does not have the properties attributed to CAF. However, recent reports have identified the CC chemokines MIPla, MIP-1/3, and RANTES (Cocchi et al., 1995) as potent suppressive factors produced by CD8' T cells. In addition, IL-16 has some suppressive activity when produced by CD8' cells from SIV-infected African green monkeys, but with much lower potency than the CC chemokines (Baier et al., 1995). The HIV-suppressive activity of human IL-16 has been questioned (Mackewiczet al., 1996), and it is also disputed whether the chemokines account for all of the CAF activity (Levy et aZ., 1996). Initial observations showed that the activity of the chemokines is greatest against

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

283

macrophage-tropic and primary HIV isolates, and they have little effect against laboratory-adapted strains such as HIV-IIIB/LAI (Cocchi et al., 1995).These observations can be explained by the recent identification of members of the chemokine receptor family as coreceptors for HIV. Different isolates of HIV use different coreceptors for cell entry, and coreceptor usage is the principal basis for cellular tropism. HIV-IIIB and other T cell-tropic strains of HIV-1 use the C-X-C chemokine receptor CXCR4 (LESTWfusin) (Feng et al., 1996), the ligand for which (SDF-1) has recently been identified and blocks entry of T-tropic HIV isolates (Bleul et al., 1996;Oberlin et al., 1996).Macrophage-tropicisolates predominantly use CCR-5, a CC chemokine receptor that binds RANTES, MIP-la, and MIP-1P (Alkhatib et al., 1996; Deng et al., 1996; Dragic et al., 1996), which blocks entry of these isolates. Unusually, HIV strains can use other members of the CC chemokine receptor family, namely, CCR-3 (Choe et al., 1996) and CCR2b (Doranz et al., 1996). These findings can account for CD8+ T cell-mediated HIV suppression by the CC chemokines and explain why there has not always been consistency between past studies of HIV suppression using strains of HIV that differ in their tropism and presumably also in their coreceptor usage. However, competition for a viral coreceptor does not account for all the properties of CAF, such as the ability to suppress HIV replication by an effect on LTR-driven transcription (Levy et al., 1996). Such factors, if powerful in vivo, could have interesting implications for CTL surveillance in that infected cells might be pushed into a state of viral latency in which they cease to be targets for CTL recognition. These latently infected cells could then be reactivated later so that virus control is actually not achieved. This process has not been fully explored in these terms. The relative contribution of CTL-mediated killing and the antiviral effect of chemokines and other factors produced by CD8' cells to control HIV infection has yet to be determined. Antiviral factors do not affect the life span of infected cells; if there was no T cell-mediated cytolysis and the virus was cytopathic with a ti of 2 days, the measured half-lives of infected cells would be as described (Ho et al., 1995; Wei et al., 1995). If the only antiviral activity of CD8' T cells was antigen-stimulated chemokine production, viral escape by mutation of CTL epitopes (as described below) would probably be unlikely because mutant and wild-type viruses released from adjacent cells would be equally susceptible to the chemokines. However, mutant virus might stimulate the release of chemokines less effectively so there could be weaker responses in the environment of mutant virions. Similarly, antagonism could inhibit release of chemokines by cells exposed to adjacent wild-type infected cells (Klenermanet al., 1995; P. Klenennan et al., unpublished results) giving benefit to both mutant and wild-type

284

SARAH ROWLAND-JONES ET AL

viruses, although no additional advantage to the mutant. The increasing evidence for viral escape to fixation (see below) implies that lysis is important in control of HIV. On the other hand the chemokine effects are strong and specific; therefore, both probably contribute to control of the virus in vivo. Thus, in real people, there is probably a combination of these activities going on, and the balance between them could determine the efficiency of control. High-level chemokine production might be very effective but may only inhibit entry to T cells and not macrophages (Schmidtmayerova et al., 1996). It is interesting that this antiviral activity should be uniquely available against HIV infection, so far the only virus known to use the chemokine receptors for cell entry, but control still fails. Lysis of infected cells could contribute substantially to the reduction and control of virus load but at the price of killing CD4+T cells and macrophages. The secretion of viral transcription inhibitors would reduce virus load but enhance latency. N. HLA and HN Infection

If CTLs are important in the control of HIV infection, HLA class I type should play a major role in determining disease progression. Selection of epitopes is almost entirely determined by HLA type, and selection of conserved compared to variable epitopes as targets for CTL responses could be a major contributing factor to the rate of disease progression. Furthermore, unlike most other major infectious diseases, HIV has newly arrived as a human pathogen so there could have been no selection over the past millenia to select for resistant HLA haplotypes, as has happened, for example, with malaria in West Africa (Hill et al., 1991). Studies of HLA associations with infectious diseases are, however, often confounded by the extreme polymorphism of the HLA complex. There are currently more than 200 alleles described at the A, B, and C loci and more continue to be identified. Thus, a calculated probability value has to be multiplied by the number of variables (alleles) studied, and few studies survive this statisticalcorrection. This can be countered by performing more than one study, particularly if the second is focused on an HLA type identified as a candidate in the first (Hill et al., 1991).If two independent studies come up with the same result, this can also solve the problem. A related issue is the problem of rarity: Few alleles reach a gene frequency as high as 0.1 so very large studies are needed to show a significant decrease in frequency in patients, i.e., resistance to infection or progression. In the classic study of its kind, Hill et al. (1991) needed to study nearly 2000

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

285

children to show a decrease in antigen frequency from 25 to 15% in order to demonstrate that HLA-B53 offered protection against severe malaria. Against this background, there have been several attempts to demonstrate HLA associations with slow or rapid progression to AIDS in HIV infection. A few HLA types or haplotypes give consistent findings. HLAB35 has been shown in five studies to be associated with rapid progression to AIDS (Cameron et al., 1990; Itescu et al., 1992, 1995; Jeannet et al., 1989; Plum et al., 1990; Sahmoud et al., 1993). However, in a study of apparent resistance to infection in West African prostitutes, Rowland-Jones et al. (1995) showed that HLA-B35 might have some advantages because of its ability to present epitopes that are conserved between HIV-1 and HIV-2. This could be an example in which results in one population cannot be transferred to another not only because of differences in prevalent HLA types but also because of differences in prevalent viruses, i.e., HIV1 or -2 and the different HIV-1 subtypes. The HLA haplotype HLA Al-BB-DR3 has also been found to be associated with rapid progression in four studies (Cameron et al., 1990; Kaplan et al., 1990; Kaslow et al., 1990; Steel et al., 1988). It has been shown that HLA-B8 selects epitopes that vary considerably and that B8 may be particularly susceptible to such epitope variation (Klenerman et al., 1994, 1995; McAdam et al., 1995; Phillips et al., 1991). It is also noticeable that despite the extensive characterization of epitopes made by several groups, no epitope has been described that is presented by HLA-A1, a relatively common HLA ty-pe (McMichael and Walker, 1994). Thus, the A1-B8 combination could be particularly unfavorable. This leads to the prediction that homozygotes should be even more susceptible; there is no information on this point, probably because homozygotes are rare. Recently, the 12th International Histocompatibility Workshop analyzed HLA types in 363 HIV-l-infected patients. The protective effects of HLAB27 were confirmed and similar effects were seen for HLA-A32. Associations with progression were found for HLA-B35, Cw4, B39, and A24 (Thorsby, 1996). Kaslow et al. (1996)have devised a novel way to address the complexities of demonstrating HLA associations with HIV infection. Taking a cohort of over 100 HN-infected men, they ranked them by HLA type and clinical course and calculated an odds ratio for each allele. These were summated in individuals and the results were compared with a second cohort of similar size. Highest ranked of the HLA class I types associated with protection were HLA-B27 and HLA-B57, both of which tend to select conserved epitopes (McMichael and Walker, 1994; Goulder et al., 1997). The TAP2.3 allele in association with HLA-A25, -26, and -32, and B18 also appeared to offer protection. HLA types associated with rapid progres-

286

SARAH ROWLAND-JONES E T A L

sion were B37, B49, and certain combinations of TAP alleles and class I types including B8-TAP2.1 and the B-C combination B35-C4. This method needs to be confirmed in other laboratories, particularly the involvement of TAP polymorphism in contributing to susceptibility or resistance; no other study has been able to identify functional differences between allelic products of TAP (Rowland-Jones et al., 1993b). In a cohort of prostitutes in Nairobi, a small number of women have been identified who show resistance to HIV infection despite repeated exposure (Fowke et al., 1996). In this group, HLA-A"6802 and HLAB18 appear to be more frequent than expected (F. Plummer, personal communication). Some associations of HIV infection with class I1 HLA types have been described, although the published literature is confusing and there are few consistent findings. HLA-DR13 and DR2 were found in one study to reduce transmission from mother to baby (Winchester et al., 1995). HLADR5 has been associated with the sicca-CD8' lymphocytosis syndrome in which progression to AIDS is delayed (Itescu et al., 1989, 1994). Several studies (Cruse et al., 1991; Donald et al., 1992; Fabio et al., 1990; Just et al., 1995) have described more rapid progression associated with the HLADR3-DQ2 haplotype which frequently occurs with HLA-B8 in linkage disequilibrium. In the 12th International Histocompatibility Workshop, protective effects were shown for HLA-DR13 and DQ6 and deleterious effects for DR3-DQ2 (Thorsby, 1996). It has been demonstrated for Epstein-Barr virus that a predominant HLA type in the population can select variants of the virus that fail to elicit strong responses through that HLA molecule (De Campos Lima et al., 1993, 1996). This has not yet been described for HIV, but might be anticipated in populations in which there are predominant HLA types, such as HLA-A2 in most populations or HLA-B35 in West Africa. However, there is currently no suggestion that different HIV clades are in any way selected by prevalent HLA types. Overall there are now sufficient consistent data to be sure that HLA type does play a role in determining susceptiblity to HIV infection and progression to immunodeficiency. This is still a field worth exploring because HLA typing techniques become more sophisticated and applicable to large numbers of DNA samples (Bunce et al., 1995). V. The Nature of HN-Specific CTls

Although most of the HIV-specific CTLs that have been characterized are classical CD8' class I-restricted CTLs; there is some evidence that at least two components exist among CTLs recognizing the envelope protein (Mc-

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

287

Chesneyet al., 1990).Clonalexpansion ofenv-specific CTLs generates CD8' class I-restricted CTLs, but peripheral blood cells can be detected that lyse both matched and mismatched env-bearing targets (Koeniget al., 1988)and that are not blocked by antibodies to CD3 or CD8 (Riviereet al., 1989a).In some cases this may represent antibody-mediated cytotoxicity (Weinholdet al., 1988).In addition, MHC-unrestricted lysis of HIV env-expressing cells has been described by unfractionated PBMC from both infected and uninfected people: This appears to be a property of CD4+cells and is not antigenspecific (Heinkelein, 1996). Characteristically, circulating HIV-specific CTLs also express CD38 and HLA-DR (Ho et al., 1993)and CD45RO and S6F1 (Watret et al., 1993). MHC class 11-restrictedCD4+CTLs have been described in the blood of infected patients (Littauaet al., 1992),but their role in HIV infection is not clear. Studies in envelope vaccine recipients have demonstrated both classical class I-restricted CTLs and CD4+ class 11-restricted CTLs (Hammond et al., 1992),and CD4+ CTLs have also been generated in vitro from the cells of seronegative donors that have been repeatedly stimulatedwithgpl20 (Lanzavecchiaet al., 1988; Orentas et al., 1990; Siliciano et al., 1988).The importance of these findings is that in theory CD4' CTLs have the potential to lyse uninfected, activated CD4' lymphoblaststhat have bound free gpl20 through the high-affinity interaction between gpl20 and the CD4 molecule (Lanzavecchiaet al., 1988;Siliciano et al., 1988).If such a mechanism operated in vivo, then the activation of memory T cells by other pathogens (for example, during opportunistic infections) could lead to CTL-mediated destruction.However,class 11-restrictedenv-specificCTLs have not been demonstrated in fresh tissues of HIV-seropositivedonors. Despite the theoretical possibility of harm, there is no evidence that immunization with recombinant gpl20 or gp160 damages CD4' T cells (Redfieldet al., 1991). W. Measurement of HN-Specific CTLs

Several different methods for assaying HIV-specific CTLs have been used, some unique to this virus. It is worth considering how they interrelate, especially quantitatively, because they do not all measure the same thing. The first assay used was the simplest-direct measurement of CTLs ex vivo without any culture in uitro (Plata et al., 1987; Walker et al., 1987), which is later referred to as CTLe or "fresh" CTL. Nixon et al. (1988) devised a restimulation technique where CTLs are stimulated in vitro by culture with autologous PHA-activated T cells; because some of the latter are infected CD4' T cells, activation should lead to expression of viral antigens. Walker et al. (1988, 1989) have used direct cloning, initiating the culture with anti-CD3 antibody and then cloning by limiting dilution.

288

SARAH ROWLAND-JONES ET AL.

Others have cloned from bulk cultures initiated by the “Nixon method” and then maintained by addition of IL-2 before cloning by limiting dilution (McAdam et al., 1995; Rowland-Jones et al., 1992). In macaques, Letvin et al. (Tsubota et al., 1989; Yamamoto et al., 1990) have detected CTLs from cultures set up with Con A as the stimulating agent; this has worked in both SIV-infected animals and, surprisingly, vaccinated animals (Shen et al., 1991). In vaccinees and exposed uninfected humans, cultures can be stimulated with autologous PHA blasts infected with SIV or HIV (Gallimore et al., 1995; Gotch et al., 1991). In the study by Gallimore et al. the CTLp frequency measured by limiting dilution analysis correlated with the lysis observed from these “bulk’ cultures. Stimulator cells infected with recombinant vaccinia expressing HIV genes and then inactivated with either paraformaldehyde (van Baalen et al., 1993) or psoralen and ultraviolet light (Lubaki et al., 1994) are also effective in generating specific CTL cultures. Rowland-Jones et al. (1995) have shown that stimulation of PBMC with epitope peptides plus IL-2 is effective; the kinetics of the response were compatible with a secondary in vitro response rather than a primary response. Addition of IL-7 to peptide-stimulated cultures improves CTL generation (Lalvaniet al., submitted): IL-7 is also effective in enhancing the generation of CTLs from HIV- (Carini and Essex, 1994)or vaccinia(Ferrari et al., 1995) stimulated cultures. It is widely assumed that these assays measure the same thing, but in fact there are probably important differences. This is most apparent when responses are quantified (Fig. 1). Estimates of CTL precursor frequencies by limiting dilution assays range from 1in lo4for enu-specific CTL precursors (Hoffenbach et al., 1989) to 1 in 5 X lo3for gag-specific CTLs (Gotch et al., 1990), figures that are only marginally higher than those estimated for other persistent virus infections such as EBV and CMV (Borysiewiczetal., 1988a).Recent exhaustive studies of CTL precursor numbers in HIV-infected people have demonstrated figures of up to 1/2000against gag, with somewhat lower numbers of precursors (up to 1/10,000)against env and pol: These figures were highest in healthy asymptomatic donors with a CD4’ T cell count of more than 400 p1 and appeared to be much lower inpatients with HIV-related disease (Carmichael et al., 1993). Many subsequent studies support these estimates (Klein et al., 1995; Koup et al., 1994; Moore et al., 1994).The frequencies in mid-phase infection are higher than those previouslyreported for EBV (upto 1/10,000), CMV (l/ZO,OOO), andvaricellazoster virus (1/100,000)(Alpet al., 1990;Borysiewicz et al., 1988a,b). However, these values would not be enough to give fresh CTL responses; experiments with CTL clones indicate that an effector :target ratio of between 0.125 : 1and 0.5 : 1is needed to give measurable lysis in a 5-hr chromium release assay (Gotch et al., 1990; S. Rowland-

289

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV -2

10

>10

c10-6 In vilro primary IL-2,IL-7,Th 15-20 divisions

-2

p@ 10

10 -4

-4

Bulk culture 112 10 divisions LDA 112 15 divisions

Fresh response No divisions

FIG.1. Measurement of CTLs. CTL clones probably exist in three developmental stages, indicated as naive CTL (nCTL), memory CTL (mCTL), and effector CTL (eCTL). The figure indicates estimated frequencies of these populations in uninfected individuals (nCTL) and HIV-infected persons (mCTL and eCTL). The number of cell divisions needed to detect each of these populations in vioo together with the essential cofactors, cytokines and Th help, are indicated. Also shown are cell types that are detected by the commonly used assays (bulk culture, limiting dilution assay, and fresh assay).

Jones, unpublished results). This means that there would have to be a frequency of CTLs of around M O O PBMC to account for the lysis observed in blood samples from many HIV-infected donors. Support for a higher CTL frequency comes from measurements of T cell receptor mRNA transcript frequencies of dominant clones by Moss et al. (1995) and Kalams et al. (1994). Using nucleotide probes based on the TCR-j3 chain CDR3 sequence of a dominant clone, frequencies between 1 and 5% were found; active CTLs may express more TCR mRNA than resting T cells so these could be slight overestimates. However, because the actual responses are not normally monoclonal, it is more likely that the actual CTL frequency is underestimated. Altman et al. (1996) used tetrameric peptide-HLA complexes as a direct method of staining CTLs specific for HLA-A2 and a gag or pol epitope. Frequencies of between 0.2 and 1.2%of CD8’ T cells were found in patients, none of whom had detectable fresh CTL assays in a 5-hr chromium release assay. The highest frequency so far obtained by this assay was 1.6%in a sample that did show “fresh” CTL activity (P. Moss and G. Ogg, unpublished data). All these results can be reconciled, given that the limiting dilution assay requires division of CTL precursors. In order to lyse 10% (the usual cutoff

290

SARAH ROWLAND-JONES ET AL

point) of 5 X lo3 target cells, one cloned T cell must grow to at least 1 X lo3cells, i.e., 10 divisions. If the CTLs measured by fresh lytic assays, mRNA transcripts, and direct staining include a substantial population that cannot grow in uitro, the differences would be explained. This would be in line with the finding in acute HIV (Fauci et al., 1996; Pantaleo et al., 1994a) and EBV infection (Callan et al., 1996) of huge expansions of antigen-specific CTLs (over 30% of all CD8’ T cells), most of which die by apoptosis in vitro and probably also in vivo. Following this line of argument, limiting ddution assays measure memory cells and the information gained should be seen as such. Actual CTL activity in patients may be more acurately reflected by assays in which CTLs are directly measured by lysis ex viuo or counted by staining or mRNA transcript quantitation. This difference becomes important in late infection when precursor numbers may be low or may appear to be low because of inefficient CD4’ T cell help, while fresh CTL activity is maintained (Rinaldo et al., 199513). This distinction is also of importance in considering whether CTL killing in vivo is important in controlling HIV and reducing CD4+ T cell numbers (P. Klenerman et al., 1996). MI. Role of HN-Specific CTLs in the Nahrral History of H N Infection

Although immunological abnormalities, particularly of CD4’ cell function, can be detected from the earliest stages of HIV infection, there is nevertheless a vigorous immune response to the virus. Antibodies are generated against all the structural and nonstructural gene products, some of which are able to neutralize heterologous isolates of HIV, whereas others can initiate antibody-dependent cellular cytotoxicity in uitro. A similarly vigorous cellular response against HIV is also observed. The detection of HIV-specific CTLs was first described in 1987 and was remarkable in that HLA-restricted CTLs specific for both gag and env gene products could be readily detected in freshly separated peripheral blood mononuclear cells (Walker et al., 1987) or in alveolar lymphocytes from patients with HIV-related pneumonitis (Plata et al., 1987), without the need for any in vitro culture or restimulation.This level of CTL activitywas unprecedented for virus infections and has subsequently only been described in infection with HTLV-1, another retroviral infection (Jacobson et al., 1990; Parker et al., 1992).Subsequent studies have estimated that between 15 and 88% of HIV-infected subjects have “fresh”or primary HIV-specificCTLs (Grant et al., 1992; Koup et al., 1989; Riviere et al., 198913; Walker et al., 1987). This is clearly a large range and may depend on different assay conditions in different labs-fresh responses are more readily detectable if a longer incubation time (e.g., 16 hr rather than 4 hr) is used (Klenerman et al.,

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

29 1

1996). As indicated previously, this response may be the most relevant to antiviral activity. CD8' class I-restricted CTLs have been demonstrated against most of the HIV gene products, predominantly directed against gag, pol and enu, but also targetting the regulatory proteins such as nef, tat, and vif (McMichael and Walker, 1994). CTLs have usually been studied using peripheral blood lymphocytes, but they have also been isolated from infected organs, such as the lungs (Plata, Autran et al., 1987), lymph nodes (Hadida et ul., 1995; Hoffenbach et al., 1989), spleen (Cheynier et al., 1994), central nervous system ( Jassoy et al., 1992; Kalams and Walker, 1995),and from the vaginal mucosa of simian immunodeficiency virus (S1V)-infectedmacaques (Lohman et al., 1995). Although most of the descriptions of HIV-specific CTLs have been in adults, it is clear that perinatally infected children can also mount an HIV-specific CTL response, even in the first year of life (Luzuriaga et al., 1995). The first identified peptide epitope was an HLAB27-restricted 15-mer in gag (Nixon et al., 1988), now known to be the decamer KRWIILGLNK (Rowland-Jonesand McMichael, 1993),and subsequently a large number of epitope peptides have been identified (reviewed in McMichael and Walker, 1994), which are now recorded in the Los Alamos HIV Molecular Immunology database (available on-line at the WWW site, http://hiv-web,lanl.gov/immuno/) (Korber et al., 1995). A striking feature of HIV infection is that the HIV-specific CTLs of an infected person are directed toward multiple epitopes: This is different from most virus infections previously studied, in which a dominant CTL response has been identified for a given restriction element and most people (or mice) with that MHC type respond through that allele to a single epitope. In some cases, the whole of the virus-specificCTL response is directed against a single peptide epitope-for example, mice of the H2bhaplotype focus their CTL response to vesicular stomatitis virus entirely on a nine amino acid stretch of the nucleocapsid protein (Van Bleek and Nathenson, 1990). People infected with HIV, in the mid-phase of their infection, usually make CTL responses against multiple epitopes through one or several of their HLA molecules, even though one may be the dominant response: For example, at least three peptides from gag are restricted by HLA-B8, and CTLs against all three epitopes can be detected in a single donor (Phillips et al., 1991) and one of these donors also makes a strong CTL response to an A2-restricted peptide in reverse transcriptase (McAdam et al., 1995; Moss et al., 1995).Another healthy donor has been found to make CTL responses to at least six different peptides from an assortment of HIV proteins (T. Harrer et al., 1996a), and we have studied many similar donors (P. Goulder, G. Ogg, and S. Rowland-Jones, unpublished observations). However, we have also found HIV-infected hemophil-

292

SARAH ROWLAND-JONES ETAL.

iacs whose entire CTL response has been directed toward a single epitope in gag over several years (Nixon et al., 1988; Goulder et al., 1997). When the positions of the epitopes are mapped (Fig. 2), it is apparent that there are clusters of epitopes in certain regions of the viral proteins, for instance, in nef (Culmann et al., 1991). The reasons are unclear but may relate to access of the intracellular processing machinery to these regions. On the other hand, epitopes are more evenly distributed through gag. There are several instances of the same epitope being presented by more than one class molecule-for example, the nef peptide 73-82 contains epitopes presented by HLA-AS, - A l l , and -B35 (the latter is an

FIT active site I protease I

Pol

I

I I

A6802 A6802

I

I

I

I

88

835

I I

gag p17 P24 I P15 1 I I I I I I A3 A2 835 cw3 A25 814 88 87IA31 827 88 A11 A'682 ~2 89d3 827 Cw8 862 813 87 852B8A33 A2 A26 835 862 814857

-

I I

A2

A l l A2 A33 835 835

I I

I

A32 835 851 A6802 851 862

I

814

~

%&

net

08 A3A3 B7A2A2A2 A1 1Al1817 852 835 817 851 88837 A24 835 82702 862857 818 862 849 env gp120 I

A3A24

;I:

835

env gp41 I

I

I

I

I

I

A29 835 87 Cw4 A32 A2 851

-

827

844 855

I

I

v3 loop

I

B8 835 A24 B14

111

M A 3 A287 A31 827

FIG.2. Map of HIV epitopes recognized by CTLs. The immunodominant HIV proteins that contribute the majority of the epitopes recognized by CTLs are represented linearly with the amino terminus on the left. Under each epitope the presenting HLA molecule is indicated. Many epitopes are clustered in the same regions and overlap. The details of these epitopes are derived from the Molecular Immunology Database (Korber et al., 1995) and published and unpublished observations from our own and other groups.

I

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

293

octamer, 74-81, but the peptides presented by A3 and A l l are identical) (Culmann et al., 1989; Koenig et al., 1990); another nef peptide, 190-198, is presented by three HLA-A2 subtypes as well as HLA-B52 (Hadida et al., 1995). Although generally donors respond to the epitopes in a predictable manner, indicating the strong selective influence of the HLA type, there are examples in which all donors with a particular class I molecule do not respond to the same epitope: For example, donors with HLA-A"201 usually respond either to an epitope in p17 gag or to one in pol, but rarely to both (McMichael and Walker, 1994; P. Goulder et al., 1997). These epitopes are present in different amounts at the cell surface of HIV-infected A2expressing cells, with the p17 peptide being more abundant (Tsomides et al., 1994), but this does not explain why the pol response is immunodominant for some donors. Responses to a third HLA-A2 epitope, in a highly conserved region of reverse transcriptase (which should therefore be a particularly valuable target for CTLs), are observed only rarely (E. Harrer et al., 1996; P. Goulder and S. Rowland-Jones, unpublished results). In a study of CTL responses to the gag protein, HLA type alone did not always predict the target of the response (Buseyne et al., 1994). It is possible that mutations in the flanking regions of CTL epitopes may lead to different rates of processing, or that immune response genes other than HLA may influence immunodominant responses, but these mechanisms have yet to be demonstrated in humans. These studies underline the extent and complexity of the CTL response to HIV. Much of the evidence for the role of HIV-specific CTLs in controlling HIV infection has come from observations of CTL activity in HIV-infected people at different stages of disease (Fig. 3). Although correlation of CTL activity and disease state provides strong circumstantial evidence of their importance, it remains to be fully proven that CTLs are directly responsible for control of virus load rather than a marker of immunological good health. These studies have also attempted to be quantitative, whereas qualitative differences in CTL activity (e.g., to epitopes that cannot vary without damaging the virus) may be more important. Direct evidence that CTLs are important for control of HIV infection requires uneqivocal proof that escape mutations are selected in uiuo, which is forthcoming (discussed below), demonstration that adoptive transfer of CTLs reduces virus load, and demonstration that vaccine induction of CTLs alone can protect against infection with HIV (or SIV as a surrogate).

A. ACUTEINFECTION The high levels of viremia that characterize primary infection with HIV-1 generally occur 3-6 weeks after HIV exposure and are frequently accompanied by clinical symptoms that include fever, malaise, rash,

294

SARAH ROWLAND-JONES ET AL.

ICTL mCTL

CD4 T

10%

800

10

Virus

-3

1.O% 106

400

200

10

0.1% 10

,

10 4

Virus

10 3

10 0

-5

Time

10 weeks

1

-4

2

3

4

5

6

7

8

9Years

FIG.3. The relationship between CTL response, virus load, and CD4' T cell count during HIV infection. The plots are based on data from several reports and do not represent an actual patient. Shown are virus load as RNA copies per milliliter plasma, CD4+ T cell count as cells/mm3 blood, effector CTL (eCTL) as percentage of CD8 T cells in the peripheral blood and precursor T cell frequencies (pCTL). Data for CTLs are based on published measurements (see text).

lymphadenopathy and, less commonly, neurological problems such as meningoencephalitis and transverse myelitis (Cooper et al., 1985). At this time, plasma viral RNA levels may be as high as 10 million copies per milliliter (Mellors et al., 1995; Piatak et al., 1993),and the CD4 count is low-occasionally it may be sufficiently depressed to allow the development of opportunistic infections (Gupta, 1993). CD4+ T cell function is also markedly abnormal (Pedersen et al., 1990).There is usually a profound CD8' T cell lymphocytosis, with huge (up to 40% of all T cells) oligoclonal expansions that express CD38, CD27, and DR but are CD28 negative (Roos et al., 1994). In culture these CD8+CD28- cells are primed for apoptosis (Brugnoni et al., 1996) and probably represent terminally differentiated effector CTLs (Pantaleo et al., 1994a). Over the next few weeks, the plasma virus load falls by several orders of magnitude, although antibodies with the capacity to neutralize the virus are rarely detected at this stage (Ariyoshi et al., 1992; Koup et al., 1994). Virus-specific CTLp have been described as early as 2 days after clinical presentation and within 3 weeks of the onset of symptoms in four of five patients in one study (Koup et al., 1994); these CTLs had specificity for the env, gag, and pol proteins. In another study HIV-specific CTLs were

CELLULAR IMMUNITY IN PROTECTION AGAINST H N

295

detected in four of five patients as early as 6-8 days after the development of symptoms; CTLs recognizing env, gag, and tat were generated, although the predominant response appeared to be toward gp160 (Borrow et al., 1994). In each of these studies one donor failed to make a detectable CTL response and exhibited a rapidly progressive course of HIV infection without control of virus levels, suggesting that the early generation of a vigorous HIV-specific CTL response may not only be responsible for the initial control of viremia but also influence the subsequent disease course. A third study examining the specificity of HIV-specific CTL responses in acute infection found that seven of nine donors had detectable CTL activity in the first 4 weeks after seroconversion (Lamhamedi-Cherradi et al., 1995). These CTLs were predominantly directed toward env, gag, and pol (particularly the integrase component of pol); only three patients responded to nef and none to rev, vif, or tat. The two patients with weak or undetectable responses in this study reported no clinical symptoms of acute infection and did not show a rapid disease progression. In macaques Letvin et al. (Yasutomi et al., 1993b) have detected CTL precursors as early as 4 days after infection, peaking with the viremia at around 2 weeks; Gallimore et al. (1995) also found the peak of CTLs at 14 days after infection. Studies of the TCR repertoire in primary HIV infection have shown that the CD8' response is represented by large but transient oligoclonal expansions in many patients (Pantaleo et al., 1994a). The expansions are identified by the overrepresentation of CD8' T cells with particular VP chains that have restricted amino acid sequences in the third complementarity determining (CDR3) region. The latter is the most variable part of the /3 chain and the limited variability is typical of an antiviral peptide CTL response (Bowness et al., 1993; Moss et al., 1991), implying that the expanded T cells are antigen-specific, Similar findings have been made in acute SIV infection (Chen et al., 1995,1996).These findings are not unique to HIV and SIV infection because similar huge expansions, up to 40% of all T cells in the blood, have been described in acute infectious mononucleosis (Callan et al., 1996) and may be found in all acute virus infections (Tripp et al., 1995). In the HIV-infected patients, poor prognosis was associated with a particularly narrow repertoire of CD8+ expansions, and it was suggested that a relatively limited CD8+ response may facilitate viral escape from the immune system or lead to more rapid immune exhaustion (Pantaleo et al. 1994a; Safrit and Koup, 1995). More detailed study of the fine specificity of the acute HN-specific CTL responses of two patients mapped them to epitopes in gp41; clones from these patients recognized their autologous virus sequences and continued to do so for up to 15 weeks after presentation (Safrit et al., 1994a). However, in two other patients

296

SARAH ROWLAND-JONES ET A L

with evidence of rapid progression, virus variants with changes in the epitopes recognized by their dominant acute CTL response, sufficient to abrogate CTL recognition, emerged duringthe first few months of infection (Borrow et al., 1997; Price et al., 1996). These last findings (discussed below in the context of immune escape) provide strong evidence for the importance of CTLs in the control of the initial viremia. In addition to cytolytic responses, CD8+ cell-mediated suppression of HIV replication has been described early in HIV infection, before the development of a neutralizing antibody response (Mackewiczet al., 1994b). In these studies, CD8+ suppressive activity was most marked before seroconversion and showed an inverse correlation with plasma viral load in three of seven subjects. Further evidence to support this has come from the simian model of SIVmac infection, in which CD8’ cells capable of inhibiting SIV replication were detected within 13-60 days of experimental infection (Tsubota et al., 1989). Taken together, these studies demonstrate that most people with acute HIV infection develop a broadly reactive HIV-specific CTL response soon after exposure, and that the resolution of acute viremia is in parallel with both the cytotoxic and noncytolytic CD8’ activity. The very high level of CTL response in the acute phase may give the T cells the chance to kill a high proportion of virus-producing cells before they generate large amounts of virus and so bring the infection under a degree of control; in other infections it is complete. There may be a correlation between both the extent of CTL activity and the breadth of the response at the clonal level, and clinical outcome, but this remains to be established in larger studies. It has also been proposed that the dynamics of viremia in acute infection are consistent with a model in which immune responses play no part at all (Phillips, 1996), but the rapid progression in people with weak or no detectable CTL responses and the acquisition of CTL “escape” variants in epitopes recognized by the dominant CTL response early after infection argue strongly against this simple hypothesis. B. ASYMPTOMATIC PERIOD OF HIV INFECTION As indicated previously, in the asymptomatic mid-phase of HIV infection, as many as 1%of peripheral blood mononuclear cells (PBMCs) can be effector CTLs (Gotch et al., 1990), whereas estimates of memory CTLs range from 1 in lo3 to 1 in lo4. The discrepancy between effector and memory CTL numbers is consistent with some degree of terminal differentiation of the effector CTLs, possibly as a result of overstimulation: This could leave the CTLs vulnerable to depletion from clonal exhaustion (Moskophidis et al., 1993).

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

297

This vigorous CTL response against HIV is likely to result from continuous antigen stimulation by a virus that is constantly turning over in multiple sites (Ho et al., 1995). Infection of dendritic cells may also contribute (Cameron et al., 1992; Knight and Macatonia, 1991; Steinman, 1991) CTLes have been detected in long-term nonprogressors with low levels of plasma viremia (T. Harrer et al., 1996b). It has been suggested by Now& and Bangham (1996) that low antigen loads might stimulate strong CTL responses if helper T cell function is good and that the same level of CTLs might require far more antigen when helper T cell functions (or other accessory factors) are impaired. Generally, the anti-HIV CTL response is considered to be stable throughout the symptomatic period. However, a stable total CTL response may conceal an unstable pattern of shifting immunodominant responses in response to variation in dominant virus mutants (Nowak et al., 1995; Phillips et al., 1991). Although this hypothesis is not universally accepted (Miedema and Klein, 1996; Wolinsky et al., 1996), and longitudinal data analyzing immunodominant epitopes and their variation are not abundant, some data from other groups (Autran, et al., 1996b) are consistent with this idea. The implication is that the stability might be an illusion, at least in some patients. It is also likely, based on the arguments presented previously on the role of T cell-mediated killing, that good control of HIV in this phase might be achieved at the cost of a gradual decline in CD4' T cells. C. HIV-SPECIFIC CTL AND DISEASE PROGRESSION There have been several attemps to correlate either the specificity or the magnitude of the HIV-specific CTL response with clinical outcome. Studies in the Amsterdam cohort of HIV-infected donors recruited since 1984 compared the gag-specific CTLp frequency in long-term asymptomatic (LTA) donors and rapid progressors (Klein et al., 1995). All the LTA subjects had detectable CTLp against gag, estimated at between 1/300 and 1/21,000,that were maintained over several years of follow-up, during which CD4' cell numbers and function remained stable and virus load was low. In contrast, one of the six rapid progressors had no detectable gag-specific CTLs, and in four others it was either transient or declined to undetectable levels with disease progression. These studies demonstrated that long-term asymptomatic HIV infection is characterized by sustained gag-specific CTL responses, although the rapid progressors were not protected from disease despite initially high levels of circulating CTLs. Studies in the MACS cohort of homosexual men in Pittsburgh examined the presence of fresh CTLe in healthy HIV-infected subjects and detected CTLes against at least one of gag, pol, and env in 83% of the men during

298

SARAH ROWLAND-JONES ET AL.

the first 8 years after seroconversion (Rinaldo et al., 199513). There was no correlation between the levels of CTLe activity and the CD4+ or CD8+ count, or between the duration of infection or the use of antiretrovirals, nor did the presence or absence of CTLes predict the disease course. These findings contrast with those in a French study, in which fresh effector gag-specific CTLs were elicited in 18 of 38 patients. The risk of progression to CDC stage IV disease was estimated to be 1.89in those without CTLes to gag compared with those with a detectable response (Riviere et al., 1995). There was no significant difference in the risk of progression for those with or without CTLes toward env. These studies may be adversely affected by the insensitivity of the current CTLe assay; detectable lysis at 4 hr needs a CTLe frequency of around 1%;lower levels down to 0.1%, which are still high in conventional terms, may be missed unless the assay is prolonged or a novel method is employed (Altman et al., 1996). Few studies have examined the response to HIV-2, which is endemic in West Africa and is distinguished from HIV-1 by lower rates of transmission and slower disease progression (Markovitz, 1993). A study of 20 HIV2-infected people in The Gambia demonstrated HIV-2-specific CTLs to gag in 90% and to pol in 70%,but to nef in only 25%. An estimate of “total” HIV-specific activity, combining responses against all three proteins, showed an inverse correlation with HIV proviral load: This relationship was strongest for CTLs against gag (Ariyoshi et al., 1995). The determinants of disease progression are thus still poorly defined. Loss of CD4 T cells is likely to be a contributing factor (see below) and the rate of CD4+ cell loss may well be determined by virus load, which is in turn controlled by the CTL response. However, the interrelationship between these parameters is complex (Nowak and Bangham, 1996).Both virus and CTLs destroy HIV-infected cells, but strong CTL responses could substantially reduce virus replication and hence the rate of disease progression. D. LONG-TERM NONPROGRESSORS Other studies have focused on the immune responses of those strictly categorized as “long-term nonprogressors” (LTNPs). This is currently understood to refer to people with at least 8 years of asymptomatic H N infection on no antiretrovirals, whose CD4+ count is more than 500/mm3, and who show no significant “slope” in a plot of their CD4+ cell numbers (Schrager et al., 1994). In general, these people have a broad range of immune responses to the virus, consistent with a largely undamaged immune system, making it hard to determine which is cause and which is effect. Cellular responses in these cohorts include both noncytolytic (Cao et al., 1995) and cytotoxic activity (Pantaleo et al., 1994b). A characteristic

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

299

feature of long-term nonprogressors is a high absolute CD8’ lymphocyte count (Pantaleo et al., 1996) Even in cohorts initially fulfilling the LTNP definition, disease progression can still occur after more than 15years of stable infection. It therefore seems likely that LTNPs constitute a heterogeneous group, and that “absolute nonprogressors” are extremely rare, if they exist at all. More detailed analysis of the subset of subjects with the lowest viral burden and persistent nonprogression from the San Francisco city clinic showed CTL responses were strong, with four of seven subjects demonstratingfresh CTLes against gag and six of seven with restimulated CTLs against several HIV proteins (T. Harrer et al., 1996b). In contrast, neutralizing antibody activity was weak or absent. This study indicates that a detectable plasma viral load, consistentwith active virus replication, is not necessary for the maintenance of circulating activated or fresh CTLes, and implicates CTLs rather than humoral responses as important for long-term nonprogression. More detailed studies of the CTL specificities detected in one donor from this group of LTNPs demonstrated recognition of recombinant vaccinia viruses expressing p17, p24, reverse transcriptase, gp120, gp41, and nef, and characterized six epitopes in detail (T. Harrer et d., 1996a).Interestingly, there was very little variation in the sequences of these epitopes from the donor’s own virus, despite the persistent high levels of CTL activity directed against them. Further studies in the MACS cohort showed that a subset of LTNPs had uniformly low levels of plasma viral RNA, and that this coincided with much higher CTLp frequencies than those found in intermediate or rapid progressors (Rinaldo et al., 1995a). As previously described, there was no correlation with CTLe activity. There was no particular protein targeted by the CTLs of the nonprogressors, but responses to gag, pol, and env were most frequent in the responders. There is little information about whether or not qualitative differences in the responses of progressors and nonprogressors can be found. Another donor from the San Francisco cohort, who has been seropositive since 1978 but remains well with a normal CD4’ count, has been shown to make an immunodominant CTL response to a highly conserved HLAA2-restricted epitope in the active site of reverse transcriptase, with the following sequence: VIYQYMDDL (E. Harrer et al., 1996). The authors speculate that responses to such critical parts of the virus may be particularly valuable in containing virus replication. This epitope is not commonly targeted by donors with HLA-A2, but we have seen responses to it in an exposed but persistently seronegative Nairobi prostitute and to the equivalent HIV-2 sequence in an HIV-2-infected Gambian LTNP (S. Rowland-Jones, unpublished observations).

300

SARAH ROWLAND-JONES ET AL

The same group has studied the relative efficiency of CTL clones in killing HIV-infected CD4+ cell lines transfected with the appropriate HLA type. They find that maximal lysis is higher with gag- and env-specific clones (similar to peptide-sensitized targets) than with pol-specific clones (Yang et al., 1996). This difference could not be explained by differences in CTL sensitivity for the cognate epitopes and may be due to different levels of expression of gag and pol. Whether this is reflected in lower efficiency of pol-specific CTLs in vivo is not known.

E. DECLINE IN HIV-SPECIFIC CTL ACTIVITYIN LATEDISEASE Most investigators agree that HIV-specific CTL activity becomes progressively harder to detect as disease progresses (Carmichael et d., 1993; McMichael and Walker, 1994; Rinaldo et al., 1995b; Klein et al., 1995; Wolinsky et al., 1996). If CTLs are primarily responsible for keeping virus load at low levels this would lead to escape of virus and ultimately more rapid progression. The reason for the loss of inducible CTLs is not fully known, but there are a number of possible reasons that are discussed. 1 . Is the Decline in CTL Activity Secondary to Loss of CD4’ T Cell Help? The most important potential mechanism is that the loss of CTL activity is secondary to the loss of CD4’ T cell numbers and impairment of their function. It is well established that HIV infection depletes CD4’ T cells (reviewed in Fauci, 1993; Fauci et al., 1996; Pantaleo and Fauci, 1995). If the CTL response is dependent on T helper activity, then the decline in CTL activity is inevitable if the virus is not completely controlled and is likely to accelerate as CD4+ T cell function deteriorates. If, as argued previously, the CTLs are largely responsible for the loss of CD4’ T cells in slowly progressing patients, the CTLs effectively “cut off their own blood supply.” In conventional antiviral CTL assays in vitro, the initial reactivation of human memory CD8+T cells requires antigen and CD4+T cells (Biddison et al., 1981) and is facilitated by addition of IL-2 and IL-7 (Dong et al., 1996; Lalvani et al., 1994). In limiting dilution assays, the benefits of these additions are evident (Carmichael et al., 1993).The long-term maintenance of CTLs in vitro requires, as a minimum, IL-2 and peptide antigen (De Vries and Spits, 1984; McMichael et al., 1986, 1988; Wallace, et al., 1982a,b). Although human CTL clones can be grown in the presence of recombinant IL-2 as the only added cytokine (McMichael et al., 1988), supernatants of activated T cells are better and imply that other cytokines are needed (Wallace et al., 1982b).Although these are most likely to come

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

301

from CD4' T cells, there are other possible sources, including the CTLs themselves, B lymphocytes, and dendritic cells. Although it might be expected that primary CTL responses in vivo would be more dependent on T cell help, some experiments suggest otherwise. Lightman et al. (1987) showed that when mice were depleted of CD4' T cells by infusion of an anti-CD4 antibody and infected with influenza virus, humoral responses were abolished but CTL responses remained, and the mice could clear the acute infection. In LCMV infection in mice, depletion of CD4+ T cells by antibody treatment did not impair the primary CTL response, but the mice failed to maintain CTL memory during persistent infection (Matloubian et al., 1994). However, under more rigorous conditions, Ewing et al. (1994) showed that in mice transgenic for an irrelevant TCR Vfi chain, although depletion of CD4' T cells in vivo greatly impaired the anti-Sendai CTL response, it had less effect on an anti-influenza CTL response. Additional support for the helper T cell independence of the primary CTL response comes from experiments in CD4-'- (Battegay et al., 1994; von Herrath et al., 1996) and MHC II-'- mice (Hou et al., 1995; Rock and Clark, 1996). A different story emerges, however, when epitope peptides are used to prime CTL responses. A consistent finding is that it is very difficult to prime with peptides that are purely class I-restricted epitopes: Epitopes that are recognized by CD4+ T cells have to be added (Sauzet et al., 1995; Shirai et al., 1994,1996). Furthermore, peptide priming of CTL responses in mice can be blocked by anti-CD4 treatment in vivo (Fayolle et al., 1991; Gao et al., 1991). A possible explanation of the contradiction comes from experiments on the roles of antigen dose and of dendritic cells in priming of CTLs. Rock and Clark (1996) primed mice with particulate ovalbumin; MHC class I1 presentation was required at low antigen doses but not at higher doses. Therefore, viruses might appear CD4+ T cell independent because the amount of antigen is usually high, whereas antigens such as minor transplantation antigens (Hurme et al., 1978) and some allo-MHC responses (Lee et al., 1994) might depend more on CD4+ T cell helper for induction of CTL responses. It is probably relevent that dendritic cells, which are capable of presenting particulate antigens by the class I pathway, are sufficient to induce primary CTL responses in vitro without T cell help (Bhardwaj et al., 1994; Young and Steinman, 1990). In summary, the experiments in mice indicate that priming with peptides or low-dose antigen requires T cell help, but that priming by virus infection may not. However, it has been shown repeatedly that the maintenance of CTL memory is dependent on the presence of CD4+ T cells; this could be crucial to our understanding of AIDS pathogenesis.

302

SARAH ROWLAND-JONES ET AL

In humans most of the data on the role of CD4+ T cells in generating and maintaining CTL responses in vivo have come from studies in HIVinfected patients with depletion of CD4+ T cells. Carmichael et al. (1993) showed that the precursor frequencies of HIV-specific CTLs were low in late HIV infection when CD4' T cell counts were <200/pl. EBV-specific CTL numbers were not impaired, so the effect may be antigen specific. However, it is possible that the slightly different culture conditions in the limiting dilution assay for the HIV- and EBV-specific T cells may have contributed to the difference. Klein et al. (1995) also showed a drop in CTL precursor numbers late in HIV infection; similarly, CTL responses in bulk cultures have also been shown to fall as AIDS develops (Rinaldo et al., 1995b). However, in these experiments it is not clear to what extent there is a fall in actual CTL precursor numbers or whether much of the perceived loss is due to impaired CD4' T cell function. Clerici et al. (1990) showed that the weak anti-influenza CTLs in HIV-infected humans could be restored by addition of alloantigenic cells to the culture, except in patients in which the alloreactive T cell response was also lost. This implicates CD4+T cells in the maintenance of the human memory CTL response and suggests that loss of CD4+ T cells in HIV infection is the primary immunopathogenic event. Thus, the failure to find CTLs does not necessarily mean that they are not there; for instance, Rinaldo et al. (1995b) found that fresh blood CTL activity, unlike restimulated CTL responses, did not fall as AIDS developed. However, if CTL precursors are maintained, how does this square with the findings in mice that CTL memory requires continuous CD4+ T cell help? Antigen dose, which is increasingly high in progressive HIV infection and very low in the mouse experiments, may account for the difference. Alternatively, and perhaps more likely,there may be a true decline in CTL numbers as HIV infection progresses, but this is accentuated by the more profound decline in CD4' T cell numbers and function.

2. Is There a Thl to Th2 Switch? A switch in the phenotype of CD4+ cells may also account for loss of helper function. In 1986, Mossman et al. showed that a panel of murine CD4+T cell clones could be classified according to their cytokine secretion profiles. Thl cells were characterized by the production of IL-2, IFN-y, and TNF-P, whereas Th2 cells predominantly secreted IL-4, IL-5, IL-10, and IL-13. These secretion patterns were subsequently termed type 1and type 2 responses (Clerici and Shearer, 1994). Not all T cell clones exhibit such a clear dichotomy of cytokine production, and lymphocytes that produced a mixture of type 1 and type 2 responses are referred to as Tho (Street et al., 1990). Subsequent reports have confirmed the presence of

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

303

T h l and Th2 phenotypes in human cells (but with a strong preponderance of Tho over Th2) and defined their responses to many natural infections (reviewed in Romagnani, 1994). Type 1responses promote cellular immunity, whereas type 2 responses bias the immune response toward antibody production. Infection with HIV provokes both cellular and humoral responses in viva However, neutralizing antibodies appear relatively late in the infection (Moore et al., 1994) and seem to be relatively ineffective because of virus envelope variation. Antibodies to other virus proteins that do not neutralize are made earlier but play an ill-defined antiviral role. As discussed previously, CTLs are probably the key factor in control of disease. Clerici and Shearer (1993) first suggested that a switch from a predominantly type 1 response to a type 2 response, resulting in decreased cellular immunity, may underlie the immune dysfunction in AIDS. Their hypothesis was followed by some evidence that over the course of HIV infection PHAstimulated peripheral blood lymphocytes produced increasing amounts of IL-4 and IL-10 (Clerici et al., 1993a, 1994c; Meyaard et al., 1994). Moreover, the anti-HIV suppressive factor produced by the CD8+cells of longterm survivors as described by Levy et al. (Levy, 1993; Mackewicz et al., 1991) appears to be dependent on a type 1 cytokine environment (Barker et al., 1995). It is not known whether production of C-C chemokines is Thl or Th2 dependent, though it is clear that CD4+ clones can make them (T. Dong et al., unpublished observations). The Thl-2 cytokine switch has not been easily found by other groups. In particular, Graziosi et al. (1994) looked at cytokine mRNA expression in the lymph nodes of HIV-seropositive patients at differing stages of disease and found little change in secretion patterns over the course of disease. Similarly, Maggi et al. were unable to show an increase in type 2 responses in infected versus uninfected patients using PHA or PMA plus anti-CD3-stimulated PBMC (Maggi et al., 1994b).They did, however, find a shift toward IL-4 secretion among CD8' T cell clones derived from the skin (Maggi et al., 1994a), although the relevance of this is unknown. Maggi et al. (1994) and Vyakarnam et al. (1995) have demonstrated that HIV preferentially replicates in the Th2 or Tho cells, whereas T h l cells appear relatively resistant to infection. This correlates with evidence that virus-induced cell death is primarily of Th2 cells, whereas the T h l subset appears protected (Clerici et al., 1994b; Estaquier et al., 1995). This could mean that a switch to an increased number of Th2 cells during infection could result in upregulated virus production in CD4' lymphocytes and subsequent apoptosis. This hypothesis is supported by the finding that HIV-infected patients with high circulating IgE levels, presumably as a result of Th2 activity, appear to have a worse prognosis (Israel Biet et al.,

304

SARAH ROWLAND-JONES ET AL.

1992; Lucey et al., 1990), although in some cases this does not correlate with increased IL-4 levels (Vigano et al., 1995). Thus, patients whose immune systems are activated by atopic conditions or infections that upregulate Th2 responses may cope poorly with HIV infection, although further evidence is needed. A complicating factor is that in vitro experiments do not take into account the complex cellular milieu in vivo. A number of other cells, principally B cells and macrophages, also secrete cytokines that modify lymphocyte responses. Macrophages are an important reservoir of HIV infection and, whereas IL-4 activates T cells and upregulates lymphocyte virus production, the same cytokine has a virostatic effect upon macrophages, as do IL-10 and IL-13 (Montaner and Gordon, 1995). Th2 cytokines therefore might not be wholly bad in HIV infection. If the T h m h O subset of cells is preferentially infected and killed, the loss would result in lower concentrations of IL-4 in the lymphatic microenvironment and the reactivation and secretion of new virus from macrophages to renew the cycle (Montaner and Gordon, 1995). It is currently difficult to interpret whether infection and death of subsets of CD4' lymphocytes is a result of, or the reason for, a T h l to Th2 switch-if indeed such a switch is real. Experiments done to date appear to present conflicting data with respect to a cytokine switch during the course of HIV infection. These disagreements illustrate the complexity of the field, which is highly dependent on the sensitivity and reproducibility of different protocols, on the different sampling times and specimens, and perhaps even on statistical error (reviewed in Romagnani et al., 1994). More convincing is the evidence that HIV replicates preferentially in T h y 0 cells and that the Th2 subset is more susceptible to apoptosis than T h l cells. However, the relevance of this phenomenon in vivo is still unknown.

3. Is There CTL Exhaustion? Another possible reason for the progressive decline in CTL activity is that these activated effector cells become exhausted as a consequence of prolonged high-level stimulation. An analagous situation has been described in an animal model transgenic for an LCMV-specific T cell receptor challenged with high doses of LCMV (Moskophidis et al., 1993a), but this was an acute phase response and there is no evidence that it occurs in humans. We have detected high levels of cells in peripheral blood expressing the same HIV-specific TCR over several years in two hemophiliac donors (Moss et al., 1995) but have not yet analyzed the frequency of these CTLs during progression to AIDS. Recent studies of telomere length as an indication of replicative potential have shown that the CD8+ cells of HIV-infected people have significant telomere shortening (Effros et al.,

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

305

1996): This is in contrast to the CD4’ subset (F. Miedema, personal communication), suggesting that there is much greater CD8+ cell turnover and hence potential for exhaustion. If CTL exhaustion does occur, then failure to generate new CD8’ CTLs could be a result of dendritic cell infection or infection of CD4+ CD8’ CTL precursors in the thymus. However, in a recent trial of adoptive immunotherapy using HIV-specific CTL clones transduced with marker and “suicide” genes, HIV-infected patients with CD4’ counts between 200 and 400/p1 were able to generate CTLs specific for the foreign genes and eliminate the infused CTLs (Riddell et al., 1996); therefore, at this level of immi~nosuppressioneffective CTL responses to new antigens can still be made. It would be instructive to know whether recipients with very low CD4+ T cell numbers would still make such a vigorous immune response. 4. HlV lnfection of CTL

Despite the lack of CD4 expression, there have been occasional reports of HIV infection of CD8’ cells: For example, in long-term SIV-macspecific CTL cultures (Tsubota et al., 1989) and, recently, in as many as 400/million CD8’ cells taken from HIV-infected people in the late stages of disease (Livingstone et al., 1996). It is possible that HIV-specific CTLs become infected at the time of lysis of infected target cells (De Maria et al., 1991). It has also been shown in human fetal thymic explants into SCID mice that CD4’8’ thymocytes can be infected with HIV and are then depleted; because these cells are precursors of CD8+ T cells, it is possible that CTLs could be infected by this route. However, this seems unlikely to be a major mechanism because multiple cell divisions are involved, and infected cells would activate virus expression and be killed by virus cytopathic effects or by the immune response.

5. Suppression of HN-Specific CTL Activity by Other Cells A population of CD8’ HNK1’ CD4- CD16- T cells was identified among the alveolar lymphocytes of AIDS patients (Joly et al., 1989) that were able to inhibit the activity of HIV-specific CTLs as well as CTLs against HLA-alloantigens in a non-MHC-restricted manner. Further studies from this group have shown that these alveolar lymphocytes are characterized by expression of CD57, and that suppression of CTL activity is mediated by a lectin-binding soluble factor (Sadat-Sowtiet al., 1991,1994). “Suppressor” lymphocytes have not yet been convincingly demonstrated in peripheral blood, although a CDS’CD11’ population of PBMC caused some reduction in cytotoxicity mediated by other CD8’ or CD16+ (NK) cells from the same donor (Kundu and Merigan, 1992).

306

SARAH ROWLAND-JONES ET AL

6. Apoptosis of HN-SpeciLfc CTLS The proportion of CD8' cells expressing fas increases during HIV infection, which makes them vulnerable to depletion by fasL-induced apoptosis (Gehri et al., 1996). There is some evidence that CD8' cells taken from people with late-stage HIV infection are prone to apoptosis in the presence of HIV antigens in vitro, which reduces the detection of HIV-specific CTL activity (Chia et al., 1995). This is associated with reduced expression of the Bcl-2 protein, which protects cells from apoptosis (Boudet et al., 1996), and expression of a CD45RO' DR' CD38' CD28- phenotype. Apoptosis has also been seen in the SIV models; Xu et al. (1997) have found that many CD8' T cells in macaques infected with a clone of SIV-mac 251 showed spontaneous apoptosis, probably related to increased expression of fas. This lysis masks the CTL response because the responding CTLs die in culture. It is not yet known what cell mediates the lysis.

7. Speajic Cytokine Defects In addition to the need for IL-2, a bone marrow stroma-derived cytokine, IL-7, has been shown to be important for the generation of CTLs in uitro. Addition of IL-7 to CTL cultures from HIV-infected people can greatly enhance the detection of HIV-specific CTLs (Carini and Essex, 1994): This effect is also seen in generating CTLs from vaccine recipients (Ferrari et al., 1995) and exposed seronegative individuals (Lalvani et al., 1997). However, both CD4' and CD8' cells from some HIV-infected people lack IL-7 receptor expression, and it is difficult to grow CTLs from them, even with the addition of IL-7 (Carini et al., 1994). The basis for the dysregulation of IL-7R expression is not clear. 8. Antagonism by HN Variants A further important possibility, which is discussed in detail below, is that variants of HIV that escape CTL recognition may be generated during the course of infection, and some of these variants may not only evade recognition but may affect CTL responses to the original sequence through TcR antagonism. Thus, as mutant viruses accumulate, the overall CTL response might be impaired, although the CTL precursors are still present. These mechanisms (Sections VII,E,l-VII,E,8) are not mutually exclusive and all could contribute to the repeatedly made observation that the CTL response that can be demonstrated in vitro fails as the CD4' T cell level falls below 2OOlpl. However, the relative contribution of each of these processes is important to understand because the implications for the treatment may be quite different. If the impairment of CTL activity in vitro is due to loss of CTL precursors, perhaps due to clonal exhaustion (Moskophidis et al., 1993b; Pantaleo et al., 1994a; Pantaleo and Fauci,

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

307

1995), an effective treatment might be to replace the CTLs by adoptive transfer (Koenig et al., 1995; Riddell et al., 1996; Riddell and Greenberg, 1995b). On the other hand, if it is secondary to loss of CD4+T cell activity, possibly including a switch from Thl to Th2 responses, treatment would be replacement of CD4+ T cells-assuming that a suitable donor was available (Walker et al., 1993). Some support for this view comes from Greenberg et al., who have treated bone marrow transplant recipients with CMV-specific CD8' CTL clones (Riddell and Greenberg, 1995a,b; Walter et al., 1995a). They found much better survival of the infused clones when there was an ongoing anti-CMV Th response (Walter et al., 1995a).

F. HIV-SPECIFIC CTLs IN EXPOSED PERSISTENT SERONEGATIVES There is now a considerable body of evidence that people exposed to H N who do not appear to be infected have markers of cellular immunity to the virus. The largest studies have demonstrated CD4' cell responsesproliferation and IL-2 secretion-in response to a panel of HIV envelope peptides in up to 75%of seronegative people with a history of HIV exposure by a variety of routes (Clerici et al., 1992, 1993b, 1994a). These responses do not prove that the subject was exposed to replicating virus rather than viral proteins or replication-incompetent HIV, but they do raise the intriguing question of how a cellular response can be generated in the absence of detectable antibody to HIV. The authors of these studies have proposed a model in which low-dose virus exposure preferentially elicits a T h l type of immune response (which could include a CTL response), which may be associated with protection, whereas high-dose exposure stimulates a Th2 response, with antibody production and persistent infection (Clerici and Shearer, 1993). In response to these observations, it was suggested that the detection of MHC class I-restricted CTLs might be a more reliable indication of exposure to infectious virus after HIV exposure (Miedema et al., 1993). HIV-specific CTL responses were first reported in 3 children born to infected mothers (Cheynier et al., 1992); these children lost maternal antibody and remained virus negative by PCR, but HIV-specific CTLs could be detected up to 3 years of age. Subsequently, CTLs to H N were described in 3 more children (Aldhous et al., 1994; Rowland-Jones d al., 1993a) but these responses were transient and observed only in the first year of life: This might suggest that persisting antigen is needed to maintain a CTL response. However, a recent study of 23 uninfected children of HIV-positive mothers, aged 12-50 months, showed that 6 (26%) had fresh HIV-specific CTLe activity (De Maria et al., 1994), implying a high frequency of circulating effector cells (Moss et al., 1995), which is hard to explain in the absence of detectable replicating virus.

308

SARAH ROWLAND-JONES ET AL

In repeatedly HIV-exposed adults, Langlade-Demoyen et al. (1994) described an increased frequency of HLA-A2-restricted HIV-nef-specific CTL precursors (measured by limiting ddution analysis, stimulating the cultures with PHA, and testing on P815-A2 cells transfected’ with HIV genes as targets) in the sexual partners of HIV-l-infected adults: In this study there was a surprisingly high frequency of CTL precursors measured against other HIV products (env and gag) in both exposed and unexposed subjects, which because of the assay conditions could represent crossreactive alloresponses from parous women. We have described the finding of CTLs specific for several HLA-B35restricted HIV peptides (from gag, pol, and nef) in three of six seronegative and apparently uninfected female prostitutes with HLA-B35 in The Gambia, West Africa (Rowland-Jones et al., 1995), where HIV seroprevalence among sex workers is approximately 35%. These CTLs were elicited by stimulation of PBMCs with peptides representing epitopes from HIV-1 and HIV-2, which are recognized by CTLs from infected Gambians with HLA-B35, the most common class I molecule in that population. Initially HIV-2 was the dominant virus in The Gambia but most new infections are with HIV-1; these epitopes are cross-reactive between HIV-1 and HIV2 in that CTLs from a donor infected with one HIV strain recognize the corresponding peptide from the other strain. The CTLs generated from the uninfected women are also cross-reactive and kill target cells infected with recombinant vaccinia virus expressing both HIV-1 and HIV-2 proteins. HIV-specific CTLs have been elicited on four occasions over the past 2 years and have been recently detected in a further two of the original six women-a total of five of six-although these women have been persistently seronegative for at least 7 years. CTLs could not be elicited by the same protocol in a large panel of unexposed controls with HLA-B35. One possible explanation for our findings is that the CTLs were initially primed by exposure to HIV-2, which appears to be both less transmissible (Adjorlolo-Johnson et al., 1994; Kanki et al., 1994) and less pathogenic (Del Mistro et al., 1992; Pepin et al., 1991; Whittle et al., 1994) than HIV1, leading to protection against HIV-1 in women making a CTL response that is cross-reactive between the two viruses. If so, the high frequency of CTL responses in exposed seronegatives with HLA-B35 might be because B35 presents several cross-reactive epitopes, whereas the majority of class I molecules so far studied do not. A larger, prospective study is needed to study the incidence of HIV infection in repeatedly exposed women with and without detectable CTL responses to determine whether these responses are actually associated with protection. We have also looked for HIV-l-specific CTLs in the Nairobi cohort of highly exposed but apparently HIV “resistant” female prostitutes (Fowke

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

309

et al., 1996), using epitope peptides selected for class I molecules, which are common in this group. We find that several of these women also have HIV-specific CTLs that recognize epitopes that are highly conserved between HIV-1 clades, which intuitively may be more likely to be associated with protection (S. Rowland-Jones and Dong, manuscript in preparation). CTLs have also been detected in this group using autologous PHA blasts infected with HIV-IIIB (a clade B virus) (K. Fowke et al, manuscript in preparation), which again suggest that cross-clade CTLs are an important feature of this cohort because virus exposure in Nairobi is mostly to HIV clades A and D. Neither the Gambian prostitutes nor the women in the Nairobi cohort with CTLs have the CCRS receptor defect described in exposed uninfected men in New York (T. Dong and S. Rowland-Jones, manuscript in preparation). A recent report describes the transient appearance of CTLs specific for HIV envelope peptides in 7 of 21 health care workers exposed to HIVinfected blood who did not seroconvert: CTLs were not detected by the same protocol in 20 workers with HIV-negative exposures (Pinto et al., 1995). The rate of actual infection in exposed health care workers is extremely low, but these observations, together with the high rate of HIV-specific proliferative responses observed in this group, suggest that significant exposure to HIV antigens occurs in a majority of people with needle-stick injuries. The transient nature of these responses, like those observed in many of the babies born to infected mothers, suggests that a single exposure does not lead to detectable memory CTLs. This is in contrast to the prostitute cohorts in which exposure may occur on a daily basis and in whom HIV-specific CTLs have been maintained over months and years of study (S. Rowland-Jones, unpublished observations). Thus HIV-specific CTLs without antibody have now been observed in a number of different exposed seronegative groups, but it remains to be established whether or not they are actually associated with protection from future infection rather than simply markers of past exposure. The detection of CTLs in the Nairobi cohort, for whom there is the most convincing evidence of resistance to HIV infection in the face of intense exposures (Fowke et al., 1996), may point toward a protective role. G. POTENTIAL ADVERSEEFFECTS OF HIV-SPECIFIC CTL ACTIVITY HIV-specific CTLs have been isolated from the lungs of infected people and their activity correlates with measured abnormalities in pulmonary gas exchange and alterations in the alveolar-capillary barrier (Autran et al., 1990). It is possible that CTL activity against HIV-infected alveolar macrophages leads to injury to the pulmonary epithelium by some as yet undefined mechanism. Similarly,the presence of HIV-specific CTLs in the CSF

3 10

SARAH ROWLAND-JONES ET AL

of patients with neurological disease (Sethi et al., 1988) led to speculation that they may be contributing to neuropathology. More vigorous HIVspecific CTL activity may be found in the CSF of some patients with AIDSrelated dementia rather than in their peripheral blood, which suggests that the recruitment of CTLs into the CSF could play a part in the pathogenesis of dementia (Jassoy et al., 1992). The CD8+lymphocytotosis syndrome is characterized by increased numbers of circulating CD8' cells that infiltrate salivary glands, lung, gut, and kidneys of a few HIV-infected people with particular HLA types (as described previously). Although the infiltrating cells have not been definitely shown to be HIV-specific CTLs, they have the phenotype of antigendriven expansions. Surprisingly, people with this syndrome have a relatively good prognosis, which implies that an excess of CD8' antigen-specific cells is beneficial rather than harmful (Itescu et al., 1993). Some HIV-specific CTLs release cytokines on contact with HIV-infected cells such as TNFs (Jassoy et al., 1993), which have the potential to upregulate HIV replication (Harrer et al., 1993). TNF-a secretion from some CTL clones produced in response to rgpl20 vaccination could be shown to increase HIV production from chronically infected T cell lines (Bollinger et al., 1993). Lysis of uninfected activated CD4' lymphoblasts by CD8+ cells has been observed in humans (but not chimpanzees) infected with HIV (Zarling et al., 1990). Further studies have shown that these CTLs recognize an activation-dependent, nonpolymorphic molecule on uninfected CD4' lymphocytes (Bienzle et al., 1996); however, HIV-specific CTLs do not have this activity. If such a mechanism operates widely, then this could contribute to the decline in CD4+cell numbers during the course of HIV infection. Similarly, if CD4' env-specific CTLs are really operating in vivo, then lysis of uninfected CD4+ cells that have adsorbed gpl20 onto their surfaces could contribute to immune depletion (Lanzavecchia et al., 1988; Siliciano et al., 1988). It remains to be established if these actually are important mechanisms for immunopathology. The role of CTLs in destroying HIV-infected CD4+ cells has been discussed previously. Zinkernagel (1995) has argued that if HIV is not cytopathic in vivo, there will be a critical balance between virus dynamics and the immune response: If there are few HIV-infected cells and an effective immune response, then the virus will be controlled or even eliminated. However, if the balance is in favor of the virus, then CTLs could be responsible for immunopathology. This model predicts that individuals with high viral loads but a poor CTL response would do well clinically; recent results in sooty mangabeys, which are frequently infected with SIV but do not get sick, suggest that this situation is present because the virus

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

311

is not cytopathic in this species (M. Feinberg and P. Johnson, personal communication). In humans the evidence is compelling that HIV is cytopathic in vivo (Ho et al., 1995; Wei et al., 1995). In this situation HIV infection in the absence of an immune response is rapidly progressive and fatal (Pantaleo and Fauci, 1995); CTLs can slow this process but will gradually erode CD4' cell numbers. WII. Does H N Escape from the Cn Response?

Given the importance of CTLs in controlling HIV infection and the variability of the virus it must be virtually inevitable that escape mutation will occur. The virus genome is lo4 nucleotides, the mutation rate is estimated to be per generation (Coffin, 1995), and lo9 or 10" virions are generated every day (Ho et al., 1995; Perelson et al., 1996). Thus, there are 10' or lo9 point mutations made each day, with every possible point mutation occurring multiple times and in many combinations, although the vast majority of mutations result in defective virus. Recent evidence on the sequences of the proteases supports the view that the virus quasispecies contains vast numbers of mutants; escape mutations are there before any drug is given (Kozalet al., 1996). However, the database of HIV amino acid sequences shows that there are clear consensus sequences (Korber et al., 1995) and nearly all patients examined have predominant virus sequences that are close to the consensus sequence (Holmes et al., 1992; Meyerhans et al., 1991; Phillips et al., 1991). This implies that conserved sequences are selected at or around the time of transmission in the absence of any immune response, presumably for their transmission characteristics. There are different consensus sequences in different geographical areas that constitute the different HIV-1 clades (Louwagie et al., 1993). Examination of envelope amino acid sequences at seroconversion demonstrates conservation even in the most variable part of the molecule, the V3 loop (Zhang et al., 1993; Zhu et al., 1993; Wolfs et al., 1990). Once the immune response begins to control the virus there is a diversification in the envelope sequence that must result largely if not entirely from selection by neutralizing antibody (McKeating et al., 1989; Wolfs et al., 1990). Evidence for this comes from measurement of nonsynonymous to synonymous or dN/dS (coding to noncoding) nucleotide changes in envelope sequences in virus in macaques infected with molecular clones of SIV (Bums and Desrosiers, 1994; Shaper and Mullins, 1993). Certain amino acid changes were clearly selected. Although antibody is almost certainly responsible, it has been suggested that there could be another force acting on the envelope (Weiss, 1996). The envelope and probably

312

SARAH ROWLAND-JONESET AL.

the V3 region binds to a coreceptor as well as CD4 on T cells, CCR5 (or more unusually CCR2b or CCR3), or CXCR4 (Alkhatib et al., 1996; Choe et al., 1996; Doranz et al., 1996; Dragic et al., 1996; Feng et al., 1996), and the virus could change its envelope to bind to different coreceptors and so change its tropism. It is possible that an increase in target cell range could also act as a selecting force on the envelope sequence. However, it is very likely that the high titer of neutralizing antibody that is often present has a selective effect. Both antibody selection and alteration in tropism could combine as a potent force in the pathogenesis of HIV infection. The initial viremia is controlled by the CTL response, as described previously: The appearance of the latter occurs just as the viremia falls (Koup et al., 1994). The high levels of expanded CTLs seen (Pantaleo et al., 1994a; Pantaleo and Fauci, 1995) are almost certainly sufficient to kill most virus-infected cells before new virions are produced (Klenerman et al., 1997). Selection of escape mutants can occur at this time. Borrow et al. (1997) and Price et al. (1996) have both described acutely infected patients in whom there were strong CTL responses to single dominant epitopes. In both, escape mutants were selected and completely dominated the virus quasispecies within 8-10 weeks. For both examples, the mutant epitopes, one in env and one in nef, failed to bind to the presenting HLA molecule, HLA B44 and B8, respectively. The apparent association between oligoclonalityin the primary response and poor outcome (Pantaleo and Fauci, 1995) could reflect selection of such escape variants and/or overstimulation and exhaustion of responding T cell clones. Unlike escape from antibody, in which the escape mutant is dominant and penetrates the barrier of neutralizing antibody to infect target cells, a mutant within an infected cell is recessive if there are multiple virus genomes within the cell: When the cell is killed all virus variants within that cell die. However, an HIV-infected cell only integrates a single copy of HIV cDNA and the mutations arise preintegration so that all viral products in the cell have the same mutations. In these circumstances selection by CTLs can occur. In the previous examples of escape from CTLs, there was a CTL response that was directed toward one immunodominant epitope. This could be crucial to the ability of HIV to escape from CTLs. In the original description of virus escape from CTLs, the infected mice were transgenic for an antiviral CTL TCR so that there was a monospecific selection force acting on the virus, LCMV (Aebischer et al., 1991;Pircher et al., 1990). Similarly, Koenig et al. (1995) described a patient who was treated by infusion of very large numbers of a single CTL clone specific for a nef epitope presented by HLA-A3. This resulted in the appearance of virus with deletions in nef that eliminated the epitope. It is likely that the transfer of such a large

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

313

number of a single clone made it the dominant selection force. In these circumstances a single mutation can evade the whole CTL response. In established infection with HIV, the picture appears more complex. In a study of HIV-infected patients with HLA-B8, it was found that they tended to make immunodominant CTL responses to an epitope in gag p17, 24-31, which is variable and can escape from CTL recognition (McAdam et al., 1995; Phillips et al., 1991). However, the same patients also respond to a second epitope in gag, p24 258-267, which can also vary. Nowak et al. (1995) showed that in one of these patients the response was unstable over time and that the dominant CTLs switched from one epitope to the other. At the same time, there was a fluctuation in the predominant virus sequence with escape variants for each epitope increasing in frequency when the CTL response was strong and decreasing when it was weak. In a theoretical model, Nowak et al. (1995; Nowak and McMichael, 1995) argued that as escape mutations were selected in the immunodominant epitope, the T cell response switched to a second epitope in an immunodominance hierarchy and this changed the selective force acting on the virus so that different mutants were selected. Thus, there were shifts in immunodominant epitopes and in the predominant virus species. A mathematical model was based on the simple premises that immunodominant CTL clones outgrew other reacting clones (not only to the same epitope but also to other epitopes in the virus) and that once the antigen changed they would decline, to be replaced by rapidly reacting clones to the next epitope. This model appeared to describe what was seen in the patients with HLA-B8 when studied over several months. Many patients have been described who have concurrent CTL responses to multiple epitopes (Autran et al., 1996a; Autran and Letvin, 1991; Chenciner et al., 1989; Hadida et al., 1992; Harrer et al., 1994, 1996; T. Harrer et al., 199613; Johnson et al., 1991, 1992, 1993; McMichael and Walker, 1994; Safrit et al., 1994a,b; Tsomides et al., 1994; Venet and Walker, 1993; Walker et al., 1987,1988,1989).When examined, almost all have mutations in these epitopes in provirus that would affect the CTL response (Couillin et al., 1994, 1995; Klenerman et al., 1994, 1995; Meier et al., 1995; Meyerhans et al., 1989, 1991; Nowak et al., 1995; Phillips et al., 1991; RowlandJones et al., 1992). The argument has been made that it would not be possible for virus to escape from CTLs because simultaneous muations would have to occur for the escape to be possible (Miedema and Klein, 1996); even at the rate of accumulation of mutations observed (Coffin, 1995), simultaneous mutation at more than two epitope sites would be extremely rare. There are three counters to this argument. One is that the CTL responses are rarely equal so that one may account for as much as 90% of the CTL activity (Goulder et al., 1997); the selective pressure

3 14

SARAH ROWLAND-JONES ET AL

would be much greater at this epitope. Careful quantitation of the strengths of the component CTL responses may be needed to identify where the selection forces are strongest. Second, even if some responses are equally strong, a small reduction in the total CTL response may give some mutant virus variants sufficient edge to gain advantage. Third, and most likely, the multiple CTL responses seen in mid-infection may have resulted from progressive escape of the type described previously (Nowak et al., 1995; Nowak and McMichael, 1995).The CTL response may start as monospecific and then diversify as epitope escape mutants are selected (Borrow et al., 1997; Price et al., 1996).Thus, the heterogeneous CTL responses that seem typical of HIV-infected individuals may reflect escape and selection that has already occurred. In a number of studies, analysis of mutations in major epitopes and comparison with nonepitope regions suggests escape and selection. This was clearly shown by Price et al. (1996) in the acute anti-nef response in an acutely infected patient, where dN/dS ratios increased over time at the epitope compared to adjacent regions. Phillips et al. (1991) showed increased variability in the p17 24-32 region of gag in patients with HLAB8, and similar observations have been made by Couillin et al. (1994, 1995) on an HLA-All-restricted epitope in gag and by Wolinsky et al. (1996) on an HLA-B7-restricted epitope in env. These are strong arguments for prior selection even where the changes in CTL response and virus were not directly observed. Sometimes the CTL response remains specific for a small number of epitopes over several years and in some cases only 1. Patients with HLAB27 all respond to an immunodominant epitope in gag 263-272 that is highly conserved in all clades (Phillips et al., 1991; Goulder et al., 1997). In three of six patients this was the only response consistently obtained. In two patients followed over 6 years the response was strong and stable, but as CD4+ counts fell to very low levels both individuals selected a mutant virus with a lysine instead of arginine at the second anchor position. HLA-B27 has a near absolute requirement for arginine at position 2 in the epitope peptide (Jardetzky et al., 1991; Rammensee et al., 1995). If this is replaced with lysine the peptide binds only poorly to HLA-B27 and has a greatly increased off-rate (Colbert et al., 1994; Goulder et al., 1997). Thus, CTLs fail to recognize cells infected with this mutant virus. It is not clear whether the change in virus preceded or followed the accelerated fall in CD4+ counts to very low levels in these patients. It is also puzzhng that the escape should have happened so late, after 12 years in both cases. The mutant virus must have been present earlier, although it is possible that the lysine mutation has to be compensated for by one or more other mutations in the gag p24 so that a viable escape virus would be very rare.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

315

Other immune responses may have contributed to the control of HIV in these patients so that the virus could only escape from the CTLs when these responses failed; a contribution from Thl T cells that failed when the CD4' count fell to very low levels might also explain the findings. It is possible that a very strong CTL response might control the lysine escape virus if the epitope is briefly exposed at the surface of infected cells so that the virus would only escape when the CTL response fell below a threshold. There are several ways by which a mutant epitope might escape recognition by CTLs. These can be mapped according to the step in the class antigen processing pathway that is affected (Koup, 1994). Mutation in the regions flanking epitopes might interfere with selection of the epitope by the cytoplasmic proteases (Couillin et al., 1994). Currently, there is little understanding of what amino acid sequences might be involved; the proteases select the C-terminal residue of the peptide (reviewed in Elliott et al., 1993) so alterations here could affect their action, as well as affecting HLA binding if the peptide is generated. Cerundolo et al. (1991) showed that presentation of an influenza virus epitope in NP by HLA A"6801 was abrogated by changes outside the epitope. Couillin et al. (1994) have described a patient with HLA-A11 who failed to respond to the expected dominant epitope in HIV nef; the epitope sequence itself was unaltered but there were differences in the flanking regions, which might have interfered with processing. This type of escape is hard to identifybut could be common. Some mutant peptides might not be transported into the ER; for instance, prolines at the second and third positions of the transported peptide seem to interfere with transport by TAP (Neefjes et al., 1993). The clearest escape mutations are those where the altered peptide fails to bind to the presenting HLA molecule; there are now several examples in which complete escape to fixation occurs (Borrow et al., 1997; Nowak et al., 1995; Price et al., 1996; Goulder et al., 1997); for nef in particular, deletions that remove the whole epitope are described (Koeniget al., 1995; Price et al., 1996), leaving the virus apparently stdl viable and virulent. Mutations that affect interactions with the TCR are the most common in a given epitope (Reid et al., 1996). Such changes, which need not necessarilybe in residues in direct contract with the TCR, are likely to affect only some clones (McAdam et al., 1995). However, the T cell response is often oligoclonal (Kalams et al., 1994; Moss et al., 1991, 1995), so it may be susceptible to this type of change; one possible consequence might be a diversification of the T cell response. Apart from nonrecognition, antagonism is also possible (Klenerman et al., 1994, 1995; Meier et al., 1995). The altered peptide ligand has a partial interaction with the TCR and prevents the CTLs from lysing cells infected with wild-type virus.

316

SARAH ROWLAND-JONES ETAL.

Cytokine or chemokine release may also be affected so that CTLs fail to inhibit virus replication (P. Klenerman, unpublished results). Meier et al. (1995) have shown that such effects can occur when the antagonist and agonist are on different cells. However, inhibition of CTL lysis of cells infected with wild-type virus was seen only when the antagonist cells were in excess. Thus, mutant virus may gain an advantage at a focal site until they are in local excess, then the response to the wild-type virus is also impaired so that the mutant may not reach fixation. Also, the wild-type virus may act as an antagonist for the new CTL response to the mutant virus, impairing CTL responses to the new variant. The descriptions of escape mutants underline the importance of CTLs in controlling the virus. They further suggest that lysis is an important in vivo mechanism for control of thls virus infection; escape is very simple if the alternative is death. CD8+ T cells that release chemokines could also select escape mutants, but if a mutant virus-infected cell was next to a cell infected with wild-type virus, virus released from both might be equally inhibited from infecting new cells. However, this could be counteracted by antagonism. Now the most important issue to resolve is the contribution of escape mutation to the progression of HIV infection to AIDS. In one model (Nowak and McMichael, 1995), escape as seen in the two acutely infected patients described previously would be typical and would be the first of a continuing series of escape mutations changing the response to different epitopes in a hierarchy of imunodominance. If the CTL responses faded to eliminate virus variants completely, the result would be a broadening of CTL specificity with shifting patterns of immunodominance constantly altering predominant virus variants (Nowak et al., 1995). In a few patients the HLA type may fortuitously select parts of the virus that do not vary very much because the sequence cannot change significantly without impairing the structure and function of the protein. In the previous example, HLA-B27 dominantly selects a conserved epitope in which escape occurs late, for reasons that are as yet unclear. In this model the CTL response gradually becomes more heterogeneous in response to successive rounds of virus escape and this results in a gradually increasing virus load. If the immunodominance hierarchy is inversely related to the amount of virus antigen needed to maintain a virus-suppressive response, this would be inevitable. Higher virus loads would hasten the decline of CD4+ T cells. A point might be reached when the impaired CD4+ T cell activity fails to maintain CTL function and immune control would collapse (Nowak and McMichael, 1995). The alternative view is that, although escape cannot be denied, it is rare and not a major contributor to the development of AIDS (Meyerhans et

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

317

al., 1991). Thus, in the patients with HLA-B27, AIDS had already developed before escape occurred and so had little to do with the decline in

CD4' T cells. However, the strongest proponent of this view argues that the CTLs are actually responsible for the loss of CD4+ T cells (Cheynier et al., 1994). If this is so then there must be pressure on the virus to escape. The jury is still out in this case and further longitudinal studies on the relationship between immunodominant CTL responses and the escape variants should resolve the argument. IX. Therapeutic Implications of the Importance of HN-Specific CTLs

A. ADOPTIVEIMMUNOTHERAPY The adoptive transfer of autologous class I MHC-restricted CTL clones or lines to patients infected with HIV has been based on the success of this therapy in the control of other viral infections in humans (Riddell and Greenberg, 1995b). Apart from the obvious intention to treat patients, this type of study directly addresses the role of CTLs in HIV infection. In uncontrolled trials, donor-derived CMV-specific CTL clone infusions protected bone marrow recipients from CMV disease until reconstitution of immune responses (Riddell et al., 1992; Walter et al., 1995b). These studies also demonstrated the relative safety of lymphocyte infusions, the ability of transferred cells to survive up to 12 weeks in most patients, and the apparent dependence of infused CTLs on CD4+ lymphocytes for survival and proliferation. Infusion of donor leukocytes or donor-derived CTL lines also appears to be effective for the treatment of EBV proliferative disorders in bone marrow recipients (Papadopoulos et al., 1994; Rooney et al., 1995) and confirms the relative safety of infusions and the relative long-lived survival of infused cells. Early experiments with adoptive transfer of lymphocytes for HIV infection have been discouraging. Koenig et al. (1995) transferred large numbers (around 10" cells) of a nef-specific clone to an HIV-infected patient (CD4 count approximately400/pl) twice over a period of 14 months. As expected, the number of CD8' cells was increased posttransfusion but surprisingly a matching rise in CTL-specific lysis was not seen. This is probably because the patient still possessed high baseline activity against the nef epitope or, less likely, because the infused cells rapidly distributed to the lymphatic system with the displacement of non-nef-specific cells into the periphery. Unexpectedly, there were transient increases in virus load following both infusions and an apparent concomitant decline in CD4' cell counts. The first infusion was given with a massive dose of IL-2 that probably contributed to the virus replication by activating CD4' T cells in viva However, a similar rise in virus load was observed after the second infusion without

318

SARAH ROWLAND-JONESET AL.

IL-2. There was one direct effect of the infusion: Sequencing of viral clones from the patient revealed a selection for viruses deleting part or all of the targeted nef epitope. This implies that the infused CTLs were active against the virus but-possibly because they were monoclonal-they were ultimately ineffective. The patient subsequently died of AIDS. Riddell et al. (1996) have transferred gag-specific clones to six patients with HIV. As a safety measure cells were genetically modified to carry both the hygromycin phosphotransferase gene and the Herpes simplex virus thymidine kinase (HSV-TK)gene, which could, if required, efficiently phosphorylate gancyclovir and eliminate the transfused cells. Unfortunately, the infused cells appeared to express HSV-TK resulting in a class I HLA-restricted CTL response and elimination of the foreign cells following the second infusion given 2 weeks after the first. At least the study showed that patients with CD4+counts of around 2OOlp1 can make primary CTL responses to a novel antigen. Data on gag-specific CTL activity or virus loads after the first infusion were not published. From the limited data currently available, therefore, adoptive transfer of CTLs does not appear to provide significant benefit. Administration of two or more clones may decrease the likelihood that viral mutants arise and large doses of simultaneous IL-2 may be harmful. This raises the problem of survival of the infused CTLs in the recipient; data from CMVspecific CTL transfer shows that they last longer in the presence of Th activity. This is unlikely to be present in recipients with low CD4+ T cell counts, the group with low CTL activity who might benefit from the transfer. The success or failure of CTL infusions will depend on the critical mechanisms of viral control. Comparison of adoptive transfer between HIV and other non-HIV viral infections may not be valid. In the case of CMV or EBV, cellular immunity, which normally controls latent virus, is absent because of bone marrow ablation and CTL infusions replace this function until host immunity is restored. In the case of HIV, host immunity is rarely if ever restored and the primary pathology is of the immune system. Thus, long-lasting effects will be needed, making the role of accessory cells and factors critical. Other mechanisms of viral control may also render CTL infusion therapy ineffective. For example, if a population of CD8+CD28- cells are primarily selected by ex vim expansion methods, these cells when infused may be susceptible to apoptosis in vivo or fail to function adequately (Levine et al., 1996). Moreover, there may remain a pool of latently infected cells that are sequestered from CTL lysis (Chun et al., 1995) or that express viral antigens defectively (Tsomides et al., 1994). In these cases CTL infusions are unlikely to stem the course of disease. Finally, the mechanism by which CTLs inhibit viral replication may represent a combination of

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

319

antigen-triggered CTL lysis as well as antigen-triggered chemokine secretion. Therefore, it may be necessary to determine whether antigen-specific clones are those that produce inhibitory chemokines. Immune reconstitution of HIV-infected patients by adoptive transfer of syngeneic whole lymphocytes or fractionated class I1 MHC-restricted CD4' lymphocytes has also been attempted. In a report from Bex et al. (1994), an uninfected twin was first vaccinated with recombinant envelope protein from HIV and, followingdemonstration of envelope-specific CTLs, his peripheral blood lymphocyte population was obtained by leukopheresis and transfused into the infected twin who had an undetectable CD4+count. No changes in the clinical, immunological, or lymphocyte parameters of the recipient were noted following the first infusion. A second infusion was performed using lymphocytes preincubated with HIV gp160. After the second transfer there was a increase in total lymphocyte counts (both CD4' and CD8+ cells) and an improvement in functional proliferation assays. However, there was also an increase in viral load following both transfer. The infected twin subsequently died of unrelated causes. In a similar infusion of lymphocytes obtained by leukopheresis from an uninfected identical twin to his HIV-infected sibling,we have demonstrated partial changes in the TCR repertoire of CD8' cells and a temporary in vivo expansion of CD4+and CD8+ cells up to 4 weeks after transfer (R. Tan and S. Rowland-Jones, unpublished observations). In a study from Lane et al. (1990), 16 HIV seropositive twins were treated with zidovudine combined with six infusions of peripheral blood lymphocytes obtained from the seronegative twin and bone marrow transplantation at the end of the study. After lymphocyte infusions, there was an increase in the percentage of CD4+cells present and an increase in the number of patients with delayed-type hypersensitivity reaction to tetanus toxoid. However, there was no change in the clinical status of the patients and the immunologic changes were transient. A second study that has recently begun is an ongoing phase I/II trial designed to examine the safety and efficacy of repeated infusions of activated, ex vivo expanded syngeneicT lymphocytes in HIV-infected twins. Some data on five patients have been collected. In one patient there was a marked rise in plasma virus coincident with each infusion. No further data are available. Restoration of T cell repertoire may be possible but this has yet to be demonstrated. It seems likely that transfer of activated CD4' lymphocytes would result in the infection of infused cells and a subsequent rise in viral load. Direct transfer of lymphocytes without IL-2 or anti-CD3 activation may represent a better alternative. Overall, immunotherapy, either of anti-HIV CTL clones or more general attempts to restore the immune system, is currently unproven, expensive, and laborious to carry out. Currently, it cannot compete with the promising

320

SARAH HOWLAND-JONES ET AL

reductions in virus load achieved by the newer antiretroviral drugs. However, it is quite likely that, although the drugs may control the virus in the short or medium term, there could be a place for immunotherapy in attempting to repair greatly damaged immune systems to facilitate longer term management of patients.

B. VACCINESTHAT INDUCE A CELLULAR RESPONSE The search for an effective vaccine for HIV has progressed on an empirical basis because the protective immunological correlates of infection remain unknown. It is assumed, however, that host immune response to infection (in addition to host genetic and viral factors) is at least partly responsible for the varying rates of progression to AIDS and the phenomena of both exposed and uninfected patients and long-term nonprogressors (Haynes et a!., 1996). Neutralizing antibodies are readily produced following infection and most appear to be directed toward portions of the envelope protein (Barin et al., 1985; Putney et al., 1986; Steimer et al., 1991; Von Gegerfelt et al., 1991; Weiss et al., 1985). Furthermore, antibodies to gp120 have been correlated with a decreased incidence of maternofetal transmission of HIV in some studies (Devash et al., 1990; Goedert et al., 1989). Unfortunately, the expression of new HIV variants during the long time course of natural infection results in relatively rapid loss of neutralizing activity (Albert et al., 1990; Groopman et al., 1987; Montefiori et al., 1991; Moore et al., 1994; Nara et al., 1990; Reitz et al., 1988). Moreover, a strong correlation between neutrahzing antibody activity and progression of disease has not been demonstrated. Similarly, vaccine-induced antibody responses alone do not appear sufficient to protect monkeys or humans from HIV infection, and in some cases antibody responses may even enhance virus replication (Mascola et al., 1993). It therefore appears unreasonable to assume that even a large and broad antibody response to the envelope region could provide the sterilizing immunity that is thought to be required to prevent host infection and progression to disease (Graham and Wright, 1995; Hilleman, 1995). However, a recent study in which macaques were protected from SIV challenge by a recombinant vaccinia expressing SIV env or gag/ pol showed that the protected animals developed CTL to SIV antigens in the challenge but not the vaccine preparation, suggesting that sufficient infection had occurred on challenge to elicit these new responses (Kent et al., 1996).Thus, sterilizing immunity may not be essential for protection. Given the increasing evidence that the very strong CTL response in the primary phase of infection plays a role in reducing the very high initial viremia, it is reasonable to ask whether prior induction of this response by vaccines could terminate an HIV infection or contain it more effectively.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

32 1

Recent vaccine attempts have therefore focused on generating CTL responses. Even if there are low levels of circulating effector CTLs when HIV exposure occurs, good levels of memory cells should ensure a stronger and more rapid CTL response. This may not necessarily produce sterilizing immunity but could result in lower virus loads and better long-term survival. In vivo virushost systems for investigating HIV vaccines have included higher primates (SIV/cynomolgusand rhesus macaques and HIVkhimpanzees) and humans (live viral vector or subunit preparations). Most primate studies have been hampered by small numbers of vaccinees and inconsistent results. In addition, a potentially important difference between these models is the relative pathogenicity of each virus in a given system (Hoth et al., 1994). SIV produces relatively rapid disease in monkeys and HIV generally produces an indolent, nonfatal infection in chimpanzees (Fultzet d,1986);the pathogenic course of HIV in humans falls between these two models. The best evidence to date that a retroviral vaccine could provide protective immunity comes from the use of an attenuated strain of SIV. Macaques immunized with a nef-deficient virus were able to resist subsequent challenge with large doses of wild-type (heterologous) virus (Almondet al., 1995;Danieletal., 1992).The basis ofprotection has been shown not to involve antibodies ( J. Stott, personal communication), so cellular immune responses, which are vigorous in animals infected with the attenuated virus (Gallimore et al., 1995),are candidates, but by default. The attenuatedvirus itselfis an unlikely vaccine candidate because nef-deficient SIV appears to be highly pathogenic to neonatal rhesus monkeys despite its apparent safety in adults (Baba et al., 1995),raising doubts as to its suitability for human trials. Although the mechanism has not been defined, it suggeststhat an immature or incomplete immune system may be at risk from attenuated viruses. Attenuated HIV vaccines are unlikely to be tested in humans because of these concerns. A natural experiment exists, however, in the case of a nef-deficient virus that was isolated from a long-survivingAustralian cohort of patients and that appeared to lack pathogenicity (Deacon et al., 1995),although it has been argued that this may not have been the case with one of the patients who died from Pneumocystis carinii infection (Ruprecht et al., 1996). In an earlier series of experiments inactivated whole SIV also conferred protection to macaques (Carlson et nl., 1990; Desrosiers et al., 1989; Murphey-Corb et al., 1989; Stott et al., 1990)but only under highly specific conditions in which both infectious challenge and vaccine virus were grown in human rather than monkey cell lines. The mechanism of this protection now appears to be mediated through neutralizing antibodies generated against class I and class I1 molecules that were incorporated into the viral envelope during virus culture (Stott et al., 1991). A correlation between CTL responses and protection from SIV has been lacking in most studies. However, Gallimore et al. (1995) have shown a

322

SARAH ROWLAND-JONES ET AL

link between high precursor levels of vaccinia nef-induced CTLs and protection from intravenous challenge of SIV. Gallimore et al. made the point that to achieve protection, vaccines had to induce high levels of memory CTLs, >1 in lo4PBMC, which is not often attained by currently available vaccine candidates. Therefore, more studies of this type are needed. Subunit vaccines have also had limited success, restricted primarily to protecting monkeys from homologous challenge with SIV (Hu et al., 1992). One problem with recombinant vaccines is that the uptake of peptide or protein by endocytosis biases cells toward a class 11-mediated response with the production of CD4' CTLs (Orentas et al., 1990). The role of these cells in natural protection is unknown. However, evidence that a subunit vaccine could produce a CD8+restricted CTL response in humans was presented by Hammond et al. (1992), who vaccinated seronegative volunteers with a vaccinia-env construct followed by a booster consisting of soluble env protein. Class I-restricted envelope-specific lysis was subsequently detected in two patients. Other modified live-virus vector vaccines in trial include vaccinia-env constructs with or without gag p24 (Hu et al., 1993; Perales et al., 1995),canary avipox vectors (Pialow et al., 1995) expressingenv and gag proteins, and similar constructs in BCG (Yasutomiet al., 1993a).In a phase 1study of recombinant canary pox vectors expressing gp160, 7 of 18 vaccinated volunteers developed some envelope-specific CTLs, the majority of which exhibited a CD3'CDB' phenotype (Pialow et al., 1995). Recently, SIV-specific memory CTL responses were found in monkeys vaccinated with a DNA vaccine encoding SIV env and gag (Johnson et al., 1992; Yasutomi et al., 1996),but, as with previous studies, these responses were not associated with protection from disease (Yasutomi et al., 1996).These viral vector and DNA vaccines may provide the balance required between the potential efficacy of attenuated vaccines and their potential danger. X. Conclusions

The importance of the CTL response in controlling HIV infection rests on six lines of evidence. The temporal inverse relationship between CTL responses and virus loads is compelling but indirect. Similarly, the demonstration of very strong CTL responses in infected patients and their presence at the sites of infection (Cheynier et al., 1994) is very suggestive of an important role. The evidence that CTLs can inhibit virus replication in vitro is solid and rests on both cytotoxic mechanisms and the potent effects of the chemokines and other secreted factors, but the studies are in vitro or at best ex uivo. Selection of escape mutants in instances in

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

323

which the CTL response focuses on a single epitope and the subsequent fixation of those variants is very strong evidence, but how far can it be generalized? Our own view is that this is a very important component of the whole pathogenesis of HIV infection but this still needs more proof. Vaccine induction of strong CTL responses is now possible, e.g., by DNA vaccines, and this should allow direct testing of the proposal that CTLs alone can control SIV in macaques and, by implication, HIV in humans. Further studies in the highly exposed resistant cohorts, a few of whom have already revealed the importance of the CCR5 receptor in primary infection, could cement the hypothesis that at least some of them are protected by their cellular immune response. Together these arguments make a strong case for the CTL response being one of the major elements in understanding HIV infection and AIDS pathogenesis. There are already initiatives to develop vaccines that elicit CTL responses and some early human trials are in progress. There remains, however, a requirement to probe and test the hypotheses proposed in this review. There could still be some surprises.

REFERENCES Adjorlolo-Johnson, G., De Cock, K., Ekpini, E., Vetter, K.,Sibailly,T., Brattegard, K., Yavo, D., Doorly, R., Whitaker, J. P., Ketsens, L., Ou, C.-Y., George, J. R., and Gayle, H. (1994). Prospective comparison of mother-to-child transmission of HN-1 and HN-2 in Abidjan, Ivory Coast. 1.Am. Med. Assoc. 272, 462-466. Aehischer, T., Moskophidis, D., Rohrer, U. H., Zinkernagel, R. M., and Hengartner, H. (1991). In vitro selection of lymphocyhc chorio-meningitisvirus escape mutants by cytotoxic T lymphocytes. Proc. Nutl. Acad. Sci. USA 88, 11047-11051. Albert, J., Abrahamsson, B., Nagy, K., Aurelius, E., Gaines, H., Krook, A., NystriSm, G., and Fenyo, E.-M. (1990). Rapid development of isolate-specificneutralizing antibodies after primary HIV-1 infection and consequent emergence of virus variants which resist neutralization by autologous sera. AIDS 4, 107. Aldhous, M. C., Watret, K. C., Mok, J. Y., Bird, A. G., and Froebel, K. S. (1994).Cytotoxic T lymphocyte activity and CD8 suhpopulations in children at risk of HIV infection. Clin. Exp. Immunol. 97(1), 61-67. Alkhatib, G., Combadiere, C., Broder, C. C., Feng, Y., Kennedy, P. E., Murphy, P. M., and Berger, E. A. (1996).CC CKR5: A RANTES, MIP-18, MIP-lfl receptor as a fusion cofactor for macrophage-tropic HIV-1. Science 272, 1955-1958. Almond, N., Kent, K., Cranage, M., Rud, E., Clarke, B., and Stott, E. J. (1995). Protection by attenuated simian immunodeficiency virus in macaques against challenge with virusinfected cells. Lancet 345(8961), 1342-1344. Alp, N., Sissons, J. G. P., and Borysiewicz, L. K. (1990). Automation of limiting dilution cytotoxicity assays. J. Immunol. Methods 129, 269. Altman, J., Moss, P. A. H., Goulder, P., Barouch, D., McHeyzer-Williams, M., Bell, J. I., McMichael, A. J., and Davis, M. M. (1996). Phenotypic analysis of antigen-specific T lymphocytes. Science 274,94-96. Anderson, M., Paabo, S., Nilsson, T., and Peterson, P. A. (1985). Impaired intracehlar transport of class I MHC antigens as a possible means for adenovirus to evade immune surveillance. Cell 43, 215.

324

SARAH ROWLAND-JONES ET AL.

Ariyoshi, K., Hanvood, E., Chiengsong-Popov,R., andweber, J. (1992). Is clearance of HIV1viraemia at seroconversionmediated by neutralising antibodies? Lancet 340,1257-1258. Ariyoshi, K., Cham, F., Berry, N., Jaffar, S., Sabally, S., Corrah, T., and Whittle, H. (1995). HIV-2-specific cytotoxic T-lymphocyte activity is inversely related to proviral load. AIDS (Philadelphia) 9(6), 555-559. Autran, B., and Letvin, N. L. (1991). HIV epitopes recognized by cytotoxic Tlymphocytes. AIDS 5, S145-Sl50. Autran, B., Plata, F., Guillon, J. M., Joly, P., Mayaud, C., and Debre, P. (1990). HIVspecific cytotoxic T lymphocytes directed against alveolar macrophages in HIV-infected patients. Res. Virol. 141(2), 131-136. Autran, B., Hadda, F., and Haas, G . (1996a). Evolution and plasticity of CTL responses against HIV. Curt-. Opin. Zmmunol., 8(4), 546-553. Baba, T. W., Jeong, Y. S., Pennick, D., Bronson, R., Greene, M. F., and Ruprecht, R. M. (1995).Pathogenicityof live, attenuated SIV after mucosal infection of neonatal macaques. Science 267(5205), 1820-1825. Baier, M., Werner, A., Bannert, N., Metzner, K., and Kurth, R. (1995). HIV suppression by interleukin-16. Nature 378(6557). Bangham, C . R. M., and McMichael, A. J. (1989). T-cell immunity to viruses. In “T Cells.” Wiley, New York. Barin, F., McLane, M. F., Allan, J. S., Lee, T. H., Groopman, J. E., and Essex, M. (1985). Virus envelope protein of HTLV-I11 represents major target antigen for antibodies in AIDS patients. Science 228(4703), 1094-1096. Barker, E., Mackewicz, C. E., and Levy, J. A. (1995). Effects of Thl and Th2 cytokines on CD8+ cell response against human immunodeficiency virus: Implications for long-term sunival. Proc. Natl. Acad. Sci. USA 92, 11135-11139. Battegay, M., Moskophidis, D., Rahemtulla, A., Hengartner, H., Mak, T. W., and Zinkernagel, R. M. (1994). Enhanced establishment of a virus camer state in adult CD4+ T-celldeficient mice. ]. Virol. 68(7), 4700-4704. Beral, V., Peterman, T., Berkelman, R., and Jaffe, H. (1991). AIDS-associated non-Hodgkin lymphoma. Lancet 337,805-809. Bex, F., Hermans, P., Sprecher, S., Achour, A., Badjou, R., Desgranges, C., Cogniaux, J., Franchioli, P., Vanhulle, C., Lachgar, A., et 41. (1994). Syngeneic adoptive transfer of anti-human immunodeficiencyvirus (HIV-1)-primed lymphocytes from a vaccinated HIVseronegative individual to his HIV-1-infected identical twin. Blood 84,3317-3326. Bhardwaj, N., Bender, A., Gonzalez, N., Bui, L. K., Garrett, M. C., and Steinman, R. M. (1994). Influenza virus-infected dendritic cells stimulate strong proliferative and cytolytic responses from human CD8+ T cells.]. Clin. Znuest. 94(2), 797-807. and Shearer, G. M. (1981). T cell subpopulations required Biddison, W. E., Sharrow, S. 0.. for the human cytotoxic T lymphocyte response to influenza virus: Evidence for T cell help. I. Zmmunol. 127, 487-491. Bienzle, D., Smaill, F. M., and Rosenthal, K. L. (1996). Cytotoxic T-lymphocytes from HIV-infected individuals recognize an activation-dependent, non-polymorphic molecule on uninfected CD4+ lymphocytes. AZDS (London) 10(3),247-254. Bled, C. C., Farzan, M., Choe, H., Parolin, C., Clark-Lewis, I., Sodroski, J., and Springer, T. (1996).The lymphocyte chemoattractant SDF-1 is a ligand for LESTFUfusin and blocks HIV-1 entry. Nature 382, 829-833. Bodmer, W. F. (1972). Evolutionary significance of the HLA system. Nature 237,139-143. Bollinger, R. C., Quinn, T. C., Liu, A. Y., Stanhope, P. E., Hammond, S. A., Viveen, R., Clements, M. L., and Siliciano, R. F. (1993). Cytokines from vaccine-induced HIV-1 specific cytotoxicT lymphocytes: Effects on viral replication. AIDS Res. Hum. Retroviruses 9(11), 1067-1077.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

325

Bonneau, R. H., and Jennings, S. R. (1990). Herpes simplex virus-specific cytolyhc T lymphocytes restricted to a normally low responder H-2 allele are protective in vivo. Virology 174(2),599-604. Borrow, P., Lewicki, H., Hahn, B. H., Shaw, G. M., and Oldstone, M. B. (1994). Virusspecific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection. 1.Virol. 68(9), 6103-6110. Borrow, P., Lewicki, X, Wei, H., Honvitz, M. S., Peffer, N., Meyers, H., Nelson, J. A,, Gairin, J. E., Hahn, €3. H., Oldstone, M. B. A., and Shaw, G. M. (1997). Antiviral pressure exerted by HIV-1 specific CTL during primary infection demonstrated by rapid selection of CTL escape virus. Nature Med. 3, 205-211. Borysiewicz, L. K., Graham, S., Hickling, J. K., Mason, P. D., and Sissons, J. G. P. (1988a). Human cytomegalovirus-specific cytotoxic T cells: Their precursor frequency and stage specificity. Eur. J. Immunol. 18, 269. Borysiewicz, L. K., Hickling, J. K., Graham, S., Sinclair, J., Grange, M. P., Smith, G. L., and Sissons, J. G. P. (1988b). HCMV-specific cytotoxic T cells. Relative frequency of stage-specific CTL recognising the 72 kD immediate early protein and glycoprotein B expressed by recombinant vaccinia virus. J. Erp. Med. 168, 919. Boudet, F., Lecoeur, H., and Gougeon, M.-L. (1996). Apoptosis associated with ex vivo downregulation of Bcl-2 and upregulation of Fas in potential cytotxicCD8+ T lymphocytes during HIV infection. /. lmmunol. 156, 2282-2293. Bowness, P., Moss, P. A. H., Rowland-Jones, S. L., Bell, J. I., and McMichael, A. J. (1993). Conservation of T cell receptor usage by HLA B27-restricted influenza-specific CTL suggests a general pattern for antigen-specific MHC class I-restricted responses. Eur. J. lmmunol. 23, 1417-1421. Brinchmann, J. E., Gaudernack, G., and Vartdal, F. (1990). CD8+ T cells inhibit HIV replication in naturally infected CD4+ T cells. Evidence for a soluble inhibitor.]. Immunol. 144,2961-2966. Brugnoni, D., Prati, E., Malacame, F., Gorla, R., Airo, P., and Cattaneo, R. (1996). The primary response to HIV infection is characterised by an expansion of activated CD8+ CD28- cells. AIDS 10, 104-106. Bruhl, P., Kerschbaum, A., Zimmerman, K., Eibl, M. M., and Mannhalter, J. W. (1996). Allostimulated lymphocytes inhibit replication of HIV-1. AIDS Res. Hum. Retroviruses 12, 31-37. Buchmeier, M. J., Welsh, R. M., Dutko, F. J., and Oldstone, M. B. A. (1980). The virology and immunologyof lymphocytic choriomeningitisvirus infection. Ado. Immunol. 30,275-331. Bunce, M., ONeill, C. M., Bamardo, M. C., Krausa, P., Browning, M. J., Moms, P. J., and Welsh, K. I. (1995). Phototyping: Comprehensive DNA typing for HLA-A, B, C, DRB1, DRB3, DRB4, DRB5 & DQBl by PCR with 144 primer mixes utilizing sequencespecific primers. Tissue Antigens 46, 355-367. Bums, D. P., and Desrosiers, R. C. (1994). Envelope sequence variation, neutralizing antibodies, and primate lentivirus persistence. Cum. Topics Microbiol. lmmunol. 188, 185-219. Buseyne, F., Janvier, G., Fleury, B., Schmidt, D., and Riviere, Y. (1994). Multispecific and heterogeneous recognition of the gag protein by cytotoxic T lymphocytes (CTL) from HIV-infected patients: Factors other than the MHC control the epitopic specificities. Clin. Exp. lmmunol. 97(3),353-360. Callan, M. F. C., Steven, N., Krausa, P., Wilson, J. D. K., Moss, P. A. H., Gillespie, G. M., Bell, J. I., Rikinson, A. B., and McMichael, A. J. (1996). Large clonal expansions of CD8+ T cells in acute infectious mononucleosis. Nature Med., 2(8), 906-911. Cameron, P. U., Mallal, S. A., French, M. A., and Dawkins, R. L. (1990). Major histocompatibility complex genes influence the outcome of HIV infection. Ancestral haplotypes with C4 null alleles explain diverse HLA associations. Hum. Immunol. 29(4), 282-295.

326

SARAH ROWLAND-JONES ETAL.

Cameron, P. U., Freudenthal, P. S., Barker, J. M., Gezelter, S., Inaba, K., and Steinman, R. M. (1992). Dendritic cells exposed to HIV-1 transmit a vigorous cytopathic infection to CD4+ T cells. Science 257, 383-387. Cannon, M. J.. Openshaw, P. J. M., and Askonas, B. (1988). Cytotoxic cells clear virus but augment lung pathologv in mice infected with respiratory syncytial virus. ]. Exp. Med. 168, 1163-1168. Cao, Y.,Qin, L., Zhang, L., Safrit, J., and Ho, D. D. (1995). Virological and immunological characterisation of long-term survivors of HIV-1 infection. N . Engl.]. Med. 332,201-208. Carini, C., and Essex, M. (1994). Interleukin %independent interleukin 7 activity enhances cytotoxic immune response of HIV-1-infected individuals. AZDS Res. Hum. Retroviruses 10(2), 121-130. Carini, C., McLane, M. F., Mayer, K. H., and Essex, M. (1994).Dysregulation ofinterleukin7 receptor may generate loss of cytotoxic T cell response in human immunodeficiency virus type 1 infection. Eur. 1. Immunol. 24( 12), 2927-2934. Carlson, J. R., McGraw, T. P., Keddie, E., Yee, J. L., Rosenthal, A., Langlois, A. J., Dickover, R., Donovan, R., Luciw, P. A., Jennings, M. B., et al. (1990). Vaccine protection of rhesus macaques against simian immunodeficiencyvirus infection. AIDS Res. Hum. Retroviruses 6(11), 1239-1246. Carmichael, A., Jin, X., Sissons, P., and Borysiewicz, L. (1993). Quantitative analysis of the human immunodeficiency virus type 1 (HIV-1)-specific cytotoxic T lymphocyte (CTL) response at different stages of HIV-1 infection: Differential CTL responses to HIV-1 and Epstein-Barr virus in late disease. ]. Exp. Med. 177(2), 249-256. Cerundolo, V., Tse, A. G. D., Salter, R. D., Parham, P., and Townsend, A. (1991). CD8 independence and specificity of cytotoxic T lymphocytes restricted by HLA Aw68.1. Proc. R. SOC. London B 244, 169-177. Chen, Z. W., Kou, Z. C., Lekutis, C., Shen, L., Zhou, D., Hdoran, M., Li, J., Sodroski, J., Lee Pamtz, D., and Letvin, N. L. (1995). T cell receptor V beta repertoire in an acute infection of rhesus monkeys with simian immunodeficiency viruses and a chimeric simian-human immunodeficiency virus. ]. Exp. Med. 182(l), 21-31. Chen, Z. W., Shen, L., Regan, J. D., Kou, Z., Ghim, S. H., and Letvin, N. L. (1996). The T cell receptor gene usage by simian immunodeficiency virus gag-specific cytotoxic T lymphocytes in rhesus monkeys. ]. Immunol. 156(4), 1469-1475. Chenciner, N., Michel, F., Dadaglio, G., Langlade-Demoyen, P., Hoffenbach, A., GarciaPons, F., Rautmann, G., Guy, B., Guillon, J.-M., Mayaud, C., Girard, M., Autran, B., Kieny, M.-P., and Plata, F. (1989). Multiple subsets of HIV-spepcific cytotoxic T lymphocytes in humans and in mice. Eur. J. Immunol. 19,1537-1544. Cheynier, R., Langlade-Demoyen, P., Marescot, M.-R., Blanche, S., Blondin, G., WainHobson, s., Griscelli, C., Vilmer, E., and Plata, F. (1992). Cytotoxic T lymphocyte responses in the peripheral blood of children born to HIV-I-infected mothers. Eur. ]. Immunol. 22,2211-2217. Cheynier, R., Henrichwark, S., Hadida, F., Pelletier, E., Oksenhendler, E., Autran, B., and Hobson, S. (1994). HIV and T cell expansion in splenic white pulps is accompanied by infiltration of HIV-specific cytotoxic T lymphocytes. Cell 78(3), 373-387. Chia, W. K., Freedman, J., Li, X.,Salit, I., Kardish, M., and Read, S. E. (1995).Programmed cell death induced by HIV type 1 antigen stimulation is associated with a decrease in cytotoxic T lymphocyte activity in advanced HIV type 1 infection. AIDS Res. Hum. Retroviruses 11(2), 249-256. Choe, H., Farzan, M., Sun, Y.,Sullivan, N., Rollins, B., Ponath, P. D., Wu, L., Mackay, C. R., LaRosa, G., Newman, W., Gerard, N., Gerard, C., and Sodroski, J. (1996). The fl-chemokine receptors CCRS and CCRS facilitate infection by primary HIV-1 isolates. Cell 85, 1135-1148.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

327

Chun, T. W., Finzi, D., Margolick, J., Chadwick, K., Schwartz, D., and Siliciano, R. F. (1995). In vivo fate of HIV-1-infected T cells: Quantitative analysis of the transition to stable latency. Nature Med. 1(12), 1284-1290. Clerici, M. and Shearer, G. (1993). A Thl+ Th2 switch is a critical step in the aetiology of HIV infection. Immunol. Today 14(3), 107-111. Clerici, M., and Shearer, G. (1994).The Thl-Th2 hypothesis of HIV infection: New insights. Immunol. Today 15,575-581. Clerici, M., Stocks, N. I., Zajac, R. A., Boswell, R. N., Via, C. S., and Shearer, G. M. (1990). Circumvention of defective CD4 T helper cell function in HIV-infected individuals by stimulation with HLA alloantigens./. Immunol. 144(9), 3266-3271. Clerici, M., Giorgi, J. V., Chou, C. C., Gudeman, V. K., Zack, J. A., Gupta, P., Ho, H. N., Nishanian, P. G., Berzofsky, J. A., and Shearer, G. M. (1992). Cell-mediated immune response to human immunodeficiency virus (HIV) type 1 in seronegative homosexual men with recent sexual exposure to HIV-1.1. Infect. Dis. 165(6), 1012-1019. Clerici, M., Hakim, F., Venzon, D., Blatt, S., Hendrix, C., Wynn, T., and Shearer, G. M. (1993a). Changes in IL-2 and IL-4 production in asymptomatic HIV-seropositive individuals.1.Clin. Invest. 91, 759. Clerici, M., Sison, A. V., Berzofsky, J. A., Rakusan, T. A., Brandt, C. D., Ellaurie, M., Villa, M., Colie, C., Venzon, D. J., Sever, J. L., et al. (1993b). Cellular immune factors associated with mother-to-infant transmission of HIV. AIDS 7(l l ) , 1427-1433. Clerici, M., Levin, J. M., Kessler, H. A., Harris, A., Berzofsky, J. A., Landay, A. L., and Shearer, G. M. (1994a). HIV-specific T-helper activity in seronegative health care workers exposed to contaminated blood. 1.Am. Med. Assoc. 271(l),42-46. Clerici, M., Sarin, A., Coffman, R. L., Wynn, T. A., Blatt, S. P., Hendrix, C. W., Wolf, S. F., Shearer, G. M., and Henkart, P. A. (1994b). Type l/type 2 cytokine modulation of T-cell programmed cell death as a model for human immunodeficiencyvirus pathogenesis. Proc. Nutl. Acud. Sci. USA 91(25), 11811-11815. Clerici, M., Stocks, N . I., Zajac, R. A., Boswell, R. N., Via, C. S., and Shearer, G. M. (1990). Circumvention of defective CD4 T helper cell function in HIV-infected individuals by stimulation with HLA alloantigens./. Immunol. 144(9), 3266-3271. Clerici, M., Wynn, T. A., Berzofsky, J. A,, Blatt, S. P., Hendrix, C. W., Sher, A,, Coffman, R. L., and Shearer, G. M. (1994~).Role of interleukin-10 in T helper cell dysfunction in asymptomatic individuals infected with the human immunodeficiency virus. 1. Clin. Inuest. 93(2), 768-785. Cocchi, F., DeVico, A. L., Ganino, D. A., Arya, S. K., Gallo, R. C., and Lusso, P. (1995). Identification of RANTES, MIP-1 alpha, and MIP-1 beta as the major HIV-suppressive factors produced by CD8+ T cells. Science 270(5243), 1811-1815. Coffin, J. M. (1995). HIV population dynamics in vivo: Implications for genetic variation, pathogenesis, and therapy. Science 267(5197), 483-489. Colbert, R. A., Rowland-Jones,S. L., McMichael, A. J., and Frelinger, J. A. (1994). Differences in peptide presentation between B27 subtypes: The importance of the P1 side chain in maintaining high affmity peptide binding to B'2703. Immunity 1, 121-130. Cooper, D., Gold, J.. Maclean, P., Donovan, B., Finlayson, R., Barnes, T., Michelmore, H., Brooke, P., and Penny, R. (1985). Acute AIDS retrovirus infection. Definition of a clinical illness associated with seromnversion. Lancet 1, 537-540. Copeland, K., McKay, P. J., and Rosenthal, K. L. (1995). Suppression of activation of the HIV LTR by CD8+ cells is not lentivirus specific. AIDS Res. Hum. Retroviruses 11, 1321-1325. Couillin, I., Culmann, P. B., Comard, E., Choppin, J., Levy, J. P., Guillet, J. G., and Saragosti, S. (1994). Impaired cytotoxic T lymphocyte recognition due to genetic variations

328

SARAH ROWLAND-JONES ET AL.

in the main immunogenic region of the human immunodeficiency virus 1 NEF protein [see comments].1.Exp. Med. 180(3), 1129-1134. Couillin, I., Connan, F., Culmann, P. B., Gomard, E., Guillet, J. G., and Choppin, J. (1995). HLA-dependent variations in human immunodeficiencyvirus Nef protein alter peptide/ HLA binding. Eur. J. Immunol. 25(3), 728-732. Cruse, J. M., Brackin, M. N., Lewis, R. E., Meek, W., Nolan, R., and Brackin, B. (1991). HLA disease association and protection in HIV infection among African Americans and Caucasians. Pathobiology 59(5), 324-328. Culmann, B., Gomard, E., Kieny, M.-P., Guy, B., Dreyfus, F., Saimot, A.-G., Sereni, D., and Levy, J.-P. (1989).An antigenicpeptide of HIV-1 NEF protein recognized by cytotoxic T lymphocytes of seropositiveindividuals in association with different HLA-B molecules. Eur. J. Immunol. 19,2383-2386. Culmann, B., Gomard, E., and Kieny, M.-P. (1991). Six epitopes reacting with human cytotoxic CD8+ T cells in the central region of the HIV-1 nef protein. J. Immunol. 146, 1560-1565. Daniel, M. D., Kirchhoff, F., Czajak, S. C., Sehgal, P. K., and Desrosiers, R. C. (1992). Protective effects of a live attenuated SIV vaccine with a deletion in the nef gene. Science 258, 1938-1941. Deacon, N. J., Tsykin, A., Smith, S. K., Ludford-Menting, M., Hooker, D. J., McPhee, D. A,, Greenway, A. L., Ellett, A., Chatfield, C., Lawson, V. A., Crowe, S., Maerz, A., Sonza, S., Learmont, J., Subvan, J. S . , Cunningham, A,, Dwyer, D., Dowton, D., and Mills, J. (1995).Genomic structure of an attenuated quasi species of HIV-1 from a blood transfusion donor and recipients. Science 270, 988-991. De Campos Lima, P., Gavioli, R., Zhang, Q. J., Wallace, L. E., Dolcetti, R., Rowe, M., Rickinson, A. B., and Masucci, M. G. (1993).HLA-All epitope loss isolates of EpsteinBarr virus from a highly A l l + population. Science 260(5104), 98-100. De Campos-Lima, P. O., Levitsky, V., Lee, S. P., Hu, L. F., Rickinson, A. B., and Masucci, M. G. (1994). T cell responses and virus evolution: Loss of HLA All-restricted CTL epitopes in EBV isolates from highly All-positive populations by selective mutation of anchor residues. I. Exp. Med. 179, 1297-1305. De Campos Lima, P. O., Levitsky, V., Imreh, M., Gavioli, R., and Masucci, M. G. (1996). Epitope-dependent selection of highly restricted or diverse T-cell repertoires in response to persistent infection by Epstein Barr virus. Submitted for publication. Del Mistro, A., Chotard, J., Hall, A. J., Whittle, H., De, R. A,, and Chieco, B. L. (1992). HIV-1 and HIV-2 seroprevalence rates in mother-child pairs living in The Gambia (west Africa).I. Acquir. Immune Defc. Syndr. 5(l),19-24. De Maria, A., Pantaleo, G., Schnittman, S. M., Greenhouse, J, J., Baseler, M., Orenstein, J. M., and Fauci, A. S. (1991).Infection of CD8+ T lymphocytes with HIV. Requirement for interaction with infected CD4+ cells and induction of infectious virus from chronically infected CD8+ cells. I. Immunol. 146(7),2220-2226. De Maria, A., Cirillo, C., and Moretta, L. (1994).Occurrence of HIV-specific CTL activity in apparently uninfected children born to HIV-1-infected mothers. I. Infect. Dis. 170, 1296-1299. Deng, H., Liu, R., Ellmeier, W., Choe, S., Unutmaz, D., Burkhart, M., Di Manio, P., Marmon, S., Sutton, R. E., Hill, C. M., Davis, C. B., Peiper, S. C., Schall, T. J., Littman, D. R., and Landau, N. R. (1996).Identification of a major co-recptor for primary isolates of HIV-1. Nature 381, 661-666. Desrosiers, R. C., Wyand, M. S., Kodama, T., Ringler, D. J., Arthur, L. O., Sehgal, P. K., Letvin, N. L., King, N. W., and Daniel, M. D. (1989).Vaccine protection against simian immunodeficiency virus infection. Proc. Natl. Acad. Sci. USA 86(16), 6353-6357.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

329

Devash, Y., Calvelli, T. A., Wood, D. G., Reagan, K. J., and Rubinstein, A. (1990). Vertical transmission of human immunodeficiency virus is correlated with the absence of highaffinity/aviditymaternal antibodies to the gp120 principal neutralizing domain. Proc. Nutl. Acad. Sci. USA 87(9), 3445-3449. De Vries, J. E., and Spits, H. (1984). Cloned human cytotoxic T lymphocyte (CTL) lines reactive with autologous melanoma cells. I. In vitro generation, isolation, and analysis to phenotype and specificity. J. Immunol. 132(l),510-519. Donald, J. A., Rudman, K., Cooper, D. W., Baumgart, K. W., Garsia, R. J., Gatenby, P. A., and Rickard, K. A. (1992). Prozression of HIV-related disease is associated with HLA DQ and DR alleles defined by restzction fragment length polymorphisms. Tissue Antigens . 39(5).241-248. Dong, T., Boyd, D., Rosenberg, W., Alp, N., Takiguchi, M., McMichael, A. J., and RowlandJones, S. L. (1996). Increased affinity for HLA-B35 of an influenza A epitope modified at an “anchor” residue results in an antagonist peptide. Eur. J. Immunol. 26, 335-339. Doranz, B. J., Rucker, J., Yi, Y., Smyth, R. J., Samson, M., Peiper, S. C., Parmentier, M., Collman, R. G., and Doms, R. W. (1996). A dual-tropic primary HIV-1 isolate that uses Fusin and the P-chemokine receptors CKR-5, CKR-3 and CKR-2b as fusion cofactors. Cell 85, 1149-1158. Dragic, T., Litwin, V., aaway, G. P., Martin, S. R., Huang, Y., Nagashima, K. A., Cayanan, C., Maddon, P. J., Koup, R. A., Moore, J. P., and Pawton, W. A. (1996). HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature 381,667-673. Effros, R. B., Allsopp, R., Chiu, C.-P., Hausner, M. A., Hiji, K., Wang, L., Harley, C. B., Villeponteau, B., West, M. D., and Giorgi, J. V. (1996). Shortened telomeres in the expanded CD28- CD8+ subset in HIV disease implicate replicative sensecence in HIV pathogenesis. AIDS 10, F17-F22. Elliott, T., Smith, M., Driscoll, P., and McMichael, A. (1993). Peptide selection by class I molecules of the MHC. Cum. +in. Zmmunol. 3(12), 854-866. Emilie, D., Maillot, M. D., Nicolas, J. F., Fior, R., and Galanaud, P. (1992). Antagonistic effect of interferon-gamma on tat-induced transactivation of HIV long terminal repeat. J. B i d . Chem. 267(29), 20565-20570. Estaquier, J., Idziorek, T., Zou, W., Emilie, D., Farber, C. M., Bourez, J. M., and Ameisen, J. C. (1995).T helper type l/T helper type 2 cytokines and T cell death: Preventive effect of interleukin 12 on activation-induced and CD95 (FAS/APO-1)-mediated apoptosis of CD4+ T cells from human immunodeficiency virus-infected persons. J. Exp. Med. 182(6),1759-1767. Ewing, C., Allan, W., Daly, K., Hou, S., Cole, G. A,, Doherty, P. C., and Blackman, M. A. (1994). Virus-specific CD8+ T-cell responses in mice transgenic for a T-cell receptor beta chain selected at random. J. Virol. 68(5),3065-3070. Fabio, G., Smeraldi, R. S., Gringeri, A., Marchini, M., Bonara, P., and Mannucci, P. M. (1990). Susceptibility to HIV infection and AIDS in Italian haemophiliacs is HLA associated. Br. J. Haemutol. 75(4), 531-536. Fauci, A. S. (1993). Immunopathogenesis of HIV infection. J. AIDS 6, 655-662. Fauci, A. S., Pantaleo, G., Stanley, S., and Weissman, D. (1996). Immunopathogenic mechanisms of HIV infection. Ann. Internal Med. 124(7), 654-663. Fayolle, C., Deriaud, E., and Leclerc, C. (1991).In vivo induction of cytotoxicT cell response by a free synthetic peptide requires CD4+ T cell help. J. Immunol. 147(12), 4069-4073. Feng, Y., Broder, C. C., Kennedy, P. E., and Berger, E. A. (1996). HIV-1 entry cofactor: Functional cDNA cloning of a seven-transmembrane G protein-coupled receptor. Science 272, 872-877. Ferrari, G., King, K., Rathbun, K., Place, C. A., Packard, M. V., Bartlett, J. A,, Bolognesi, D. P., and Weinhold, K. J. (1995).IL-7 enhancement of antigen-driven activatiodexpansion of

330

SARAH ROWLAND-JONES ET AL.

HIV-1-specific cytotoxic T lymphocyte precursors (CTLp). Clin. Exp. Immunol. 101(2), 239-248. Fowke, K., Nagelkerke, N., Kimani, N., Simonsen, N., Anzala, A. O., Bwayo, J., Oyugi, J., MacDonald, K., Ngugi, E., and Plummer, F. (1996). Resistance to HIV-1 infection amongst prostitutes. Lancet 348, 1347-1351. Fruh, K., Ahn, K., Djabdah, H., Sempe, P., van Endert, P. M., Tampe, R., Peterson, P. A., and Yang, Y.(1995).A viral inhibitor ofpeptide transporters for antigen presentation. Nature 375(6530), 415-418. Fultz, P. N., McClure, H. M., Swenson, R. B., et. d. (1986).Persistent infection of chimpanzees with human T-lymphotrophic virus type IWimmunodeficiency syndrome. J. Virol. 58, 116-124. Gallimore, A., Cranage, M., Cook, N., Almond, N., Bootman, J., Rud, E., Silvera, P., Dennis, M., Corcoran, T., Stott, J., andetal. (1995). Early suppression of SIV replication byCD8+ nef-specific cytotoxic T cells in vaccinated macaques. Nature Med. I(l l ) , 1167-1173. Gao, X. M., Zheng, B., Liew, F. Y., Brett, S., and Tite, J. (1991). Priming of influenza virus-specific cytotoxic T lymphocytes in vivo by short synthetic peptides. J. Immunol. 147(lo), 3268-3273. Gehri, R., Hahn, S., Rothen, M., Steuerwald, M., Nuesch, R., and Erb, P. (1996). The Fas receptor in HIV infection: Expression on peripheral blood lymphocytes and role in the depletion of T cells. AIDS (Philadelphia) 10(1), 9-16. Germain, R. N. (1991).Antigen presentation: The second class story. Nature 353,605-606. Goedert, J. J., Mendez, H., Drummond, J. E., Robert, G. M., Minkoff, H. L., Holman, S., Stevens, R., Rubinstein, A., Blattner, W. A., Willoughby, A., et al. (1989). Mother-toinfant transmission of human immunodeficiencyvirus type 1:Associationwith prematurity or low anti-gpl20. Lancet 2(8676), 1351-1354. Goldberg, A. L., and Rock, K. L. (1992). Proteolysis, proteasomes and antigen presentation. Nature 357, 375-378. Gotch, F. M., Nixon, D. F., Alp, N., McMichael, A. J.. and Borysiewicz, L. K. (1990). High frequency of memory and effector gag specificcytotoxicT lymphocytes in HIV seropositive individuals. Int. Immunol. 2, 707. Gotch, F. M., Hovell, R., Delchambre, M., Silvera, P., and McMichael,A. J. (1991).Cytotoxic T-cell response to simian immunodeficiency virus by cynomolgus macaque monkeys immunized with recombinant vaccinia virus. AIDS 5(3),317-320. Goulder, P. J. R., Phillips, R. E., Colbert, R. A., McAdam, S., Ogg G., Nowak, M. A., Giongrande, P., Luzzi, G., Morgan, B., Edwards, A., McMichael, A. J., and RowlandJones, S. (1997). Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nature Med. 3, 212-217. Goulder, P. J. R., Sewell, A. K., Lalloo, D. G., Price, D. A,, Whelan, J. W., Evans, J., Taylor, G. P., Luzzi, G., Giangrande, P., Phillips, R. E., and McMichael, A. J. (1997). Patterns of immunodominance in HIV-1 specific cytotoxic T-lymphocyte responses in two HLA identical siblings with HLA-A"0201 are influenced by epitope mutations. J. Exp. Med. 185, 1-11. Graham, B. S., and Wright, P. F. (1995). Candidate AIDS vaccines. N . Engl. J. Med. 333(20), 1331- 1339. Grant, M. D., Smaill, F. M., Singal, D. P., and Rosenthal, K. L. (1992). The influence of lymphocyte counts and disease progression on circulating and inducible anti-HIV-1 cytotoxic T-cell activity in HIV-1-infected subjects. AIDS 6(lo), 1085-1094. Graziosi, C . , Pantaleo, G., Gantt, K. R., Fortin, J.-P., Demarest, J. F., Cohen, 0.J., Sekaly, R. P., and Fauci, A. S. (1994). Lack of evidence for the dichotomy of T h l and Th2 predominance in HIV-infected individuals. Science 265, 248-252.

CELLULAR IMMUNITY IN PROTECTION AGAINST H N

331

Groopman, J. E., Benz, P. M., Ferriani, R., Mayer, K., Allan, J. D., and Weymouth, L. A. (1987). Characterization of serum neutralization response to the human immunodeficiency virus (HIV). AIDS Res. Hum. Retrouimses 3(1), 71-85. Gupta, K. K. (1993). Acute immunosuppression with HIV seroconversion. N. Engl. J. Med. 328, 228-229. Hadida, F., Parrot, A., Kieny, M. P., Sadat, S. B., Mayaud, C., Debre, P., and Autran, B. (1992). Carboxyl-terminal and central regions of human immunodeficiency virus-1 NEF recognized by cytotoxic T lymphocytes from lymphoid organs. An in vitro limiting dilution analysis. J. Clin. Inuest. 89(1), 53-60. Hadida, F., Haas, G., Zimmerman, N., Hosmalin, A., Spohn, R., Samri, A., Jung, G., Debre, P., and Autran, B. (1995). CTLs from lymphoid organs remgnise an optimal HLA-A2restricted and HLA-B52-restricted nonapeptide and several epitopes in the C-terminal region of HIV-1 nef. J. Immunol. 154,4174-4186. Hammond, S. A., Bollinger, R. C., Stanhope, P. E., Quinn, T. C., Schwartz, D., Clements, M. L., and Siliciano, R. F. (1992).Comparative clonal analysisof human immunodeficiency virus type 1 (HIV-1)-specific CD4+ and CD8+ cytolytic T lymphocytes isolated from seronegative humans immunized with candidate HIV-1 vaccines. J. Exp. Med. 176(6), 1531-1542. Harrer, E., Harrer, T., Buchbinder, S., Mann, D. L., Feinberg, M., Yilma, T., Johnson, R. P., and Walker, B. D. (1994). HIV-1-specific cytotoxic T lymphocyte response in healthy, long-term nonprogressing seropositive persons. AIDS Res. Hum. Retmimses 10(2), 1889-2229. Harrer, E., Harrer, T., Barbosa, P., Feinberg, M., Johnson, R. P., Buchbinder, S., and Walker, B. D. (1996). Recognition of the highly conserved YMDD region in the human immunodeficiency virus type 1 reverse transcriptase by HLA-A2-restricted cytotoxic T lymphocytes from an asymptomatic long-term nonprogressor. J. Infect. Dk. 173(2), 476-479. Harrer, T., Jassoy, C., Harrer, E., Johnson, R. P., and Walker, B. D. (1993). Induction of HIV-1 replication in a chronically infected T-cell line by cytotoxic T lymphocytes. J. Acquir. Immune Defc. Syndr. 6(8),865-871. Harrer, T., Harrer, E., Kalams, S., Barbosa, P., Trocha, A., Johnson, R. P., Elbeik, T., Feinberg, M. B., Buchbinder, S. P., and Walker, B. D. (1996a). Cytotoxic T lymphocytes in asymptomaticlong-term non-progressing HIV-1 infection.]. Immunol. 156,2616-2623. Harrer, T., Harrer, E., Kalams, S., Trocha, A., Johnson, R. P., Elbeik, T., Feinberg, M. B., Cao, H., Ho, D. D., Buchbinder, S. P., and Walker, B. D. (1996b). Strong cytotoxic T cells and weak neutralising antibodies in long-term non-progressing HIV-1 infection. AIDS Res. Hum. Retrovinrses 12, 585-592. Haynes, B. F., Pantaleo, G., and Fauci, A. S. (1996). Toward an understanding of the correlates of protective immunity to HIV infection. Science 271(5247), 324-328. Heinkelein, M., Eder-Konig, I., Klinker, H., Ruckle-Lanz, H., and Jassey, C. (1996). Lysis of HIV-1 antigen-expressing cells by CD4 and CD8 T cells ex uiuo. J. In$ Dk. 174, 209-213. Hill, A., and Ploegh, H. (1995).Getting the inside out: The transporter associated with antigen processing (TAP) and the presentation of viral antigen. Proc. Nutl. Acud. Sci. USA 92(2), 341-343. Hill, A., Jugovic, P., York, I., Russ, G., Bennink, J., Yewdell, J., Ploegh, H., and Johnson, D. (1995). Herpes simplex virus tums off the TAP to evade host immunity. Nature 375(6530), 411-415. Hill, A. V. S. (1992). HLA and infection. J. R. Colloq. Phys. 26, 11-16. Hill, A. V. S., Allsopp, C. E. M., Kwiatkowski, D., Anstey, N. M., Twumasi, P., Rowe, P. A,, Bennett, S., Brewster, D., McMichael,A. J., and Greenwood, B. M. (1991). Common

332

SARAH ROWLAND-JONES ET AL

West African HLA antigens are associated with protection from severe malaria. Nature 352,595-600. Hill, A. V. S., Elvin, J. E., Willis, A. C., Aidoo, M., Allsopp, C. E. M., Gotch, F. M., Gao, X. M., Takiguchi, M., Greenwood, B. M., Townsend, A. R. M., McMichael, A. J., and Whittle, H. C. (1992). Molecular analysis of the association of HLA-B53 and resistance to severe malaria. Nature 360, 434-439. Hilleman, M. (1995).Whether and when an AIDS vaccine? Nature Med. 1(11),1126-1129. Ho, D. D., Neumann, A. U., Perelson, A. S., Chen, W., Leonard, J. M., and Markowitz, M. (1995). Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 373(6510), 123-126. Ho, H. N., Hultin, L. E., Mitsuyasu, R. T., Matud, J. L., Hausner, M. A., Bockstoce, D., Chou, C. C., O'Rourke, S., Taylor, J. M., and Giorgi, J. V. (1993). Circulating HIVspecific CD8+ cytotoxic T cells express CD38 and HLA-DR antigens. 1. lmmunol. 150(7),3070-3079. Hoffenbach, A., Langlade-Demoyen, P., Dadaglio, G., Vilmer, E., Michel, F., Mayaud, C., Autran, B., and Plata, F. (1989). Unusually high frequencies of HIV-specific cytotoxic T lymphocytes in humans. 1.lmmunol. 142,452-462. Holmes, E. C., Zhang, L. Q., Simmonds, P., Ludlam, C. A., and Leigh-Brown, A. J. (1992). Convergent and divergent sequence evolution in the surface envelope glycoprotein of human immunodeficiency virus type I within a single infected patient. Proc. Natl. Acad. Sci. USA, 89,4835-4839. Hoth, D. F., Bolognesi, D. P., Corey, L., and Vermund, S. H. (1994). NIH conference. HIV vaccine development: A progress report. Ann. Intern. Med. 121(8), 603-611. Hou, S., Mo, X. Y., Hyland, L., and Doherty, P. C. (1995). Host response to Sendai virus in mice lacking class I1 major histocompatibility complex glycoproteins. J. Virol. 69(3), 1429-1434. Hsueh, F. W., Walker, C. M., Blackbourn, D. J.. and Levy, J. A. (1994). Suppression of HIV replication by CD8+ cell clones derived from HIV-infected and uninfected individuals. Cell. Immunol. 159, 271-279. Hu, S. L., Abrams, K., Barber, G. N., Moran, P., Zarling, J. M., Langlois, A. J., Kuller, L., Morton, W. R., and Benveniste, R. E. (1992).Protection of macaques against SIVinfection by subunit vaccines of SIV envelope glycoprotein gp160. Science 255(5043), 456-459. Hu, S. L., Stallard, V., Abrams, K., Barber, G. N., Kuller, L., Langlois, A. J., Morton, W. R., and Benveniste, R. E. (1993). Protection of vaccinia-primed macaques against SIVmne infection by combination immunization with recombinant vaccinia virus and SIVmne gp160.J. Med. Pn'matol. 22(2-3), 92-99. Hurme, M., Hetherington, C. M., Chandler, P. R., and Simpson, E. (1978). Cytotoxic Tcell responses to H-Y: Mapping of the Ir genes. J. Exp. Med. 147(3), 758-767. Israel Biet, D., Labrousse, F., Tourani, J. M., Sors, H., Andrieu, J. M., and Even, P. (1992). Elevation of IgE in HIV-infected subjects: A marker of poor prognosis. 1.Allergy Clin. lmmunol. 89(Pt. l),68-75. Itescu, S., Brancato, L. J., and Winchester. R. (1989). A sicca syndrome in HIV infection: Association with HLA-DR5 and CD8 lymphocytosis [see comments]. Lancet 2(8661), 466-468. Itescu, S., Mathur, W. U., Skovron, M. L., Brancato, L. J., Marmor, M., Zeleniuch, J. A., and Winchester, R. (1992). HLA-B35 is associated with accelerated progression to AIDS. 1.Acquir. lmmune Dejic. Syndr. 5(1), 37-45. Itescu, S., Dalton, J. , Zhang, H. Z., and Winchester, R. (1993). Tissue infiltration in a CD8 lymphocytosis syndrome associated with human immunodeficiency virus-1 infection has the phenotypic appearance of an antigenicallydriven response.]. Clin. lnoest. 91(5),22162225.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

333

Itescu, S., Rose, S., Dwyer, E., and Winchester, R. (1994). Certain HLA-DR5 and -DR6 major histocompatibility complex class I1 alleles are associated with a CD8 lymphocytic host response to human immunodeficiency virus type 1 characterized by low lymphocyte viral strain heterogeneity and slow disease progression. Proc. Nutl. Acud. Sci. USA 91(24), 11472- 11476. Itescu, S., Rose, S., Dwyer, E., and Winchester, R. (1995). Grouping HLA-B locus serologic specificitiesaccording to shared structural motifs suggests that different peptide-anchoring pockets may have contrasting influences on the course of HIV-1 infection. Hum. Immunol. 42(1), 81-89. Jacobson, S., Shida, H., McFarlin, D. E., Fauci, A., and Koenig, S . (1990). Circulating CD8+ cytotoxic T lymphocytes specific for HTLV-1 pX in patients with HTLV-1 associated neurological disease. Nature 348, 245-248. Jardetzky, T. S., Lane, W. S., Robinson, R. A., Madden, D. R., and Wiley. D. C. (1991). Identification of self peptides bound to purified HLA-B27. Nature 353(6342), 326-329. Jassoy, C., Johnson, R. P., Navia, B. A., Worth, J., and Walker, B. D. (1992). Detection of a vigorous HIV-1-specific cytotoxic T lymphocyte response in cerebrospinal fluid from infected persons with AIDS dementia complex. J. Immunol. 149(9), 3113-3119. Jassoy, C., Harrer, T., Rosenthal, T., Navia, B. A., Worth, J., Johnson, R. P., and Walker, B. D. (1993). Human immunodeficiency virus type 1-specific cytotoxic T lymphocytes release gamma interferon, tumor necrosis factor alpha (TNF-alpha), and TNF-beta when they encounter their target antigens. J Virol. 67(5), 2844-2852. Jeannet, M., Sztajzel, R., Carpentier, N., Hirschel, B., and Tiercy, J. M. (1989). HLA antigens are risk factors for development of AIDS. J. Acquir. Immune Defc. Syndr. 2( 1). 28-32. Johnson, P. R., Montefiori, D. C., Goldstein, S., Hamm, T. E., Zhou, J., Kitov, S., Haigwood, N. L., Misher, L., London, W. T., Gerin J. L., et al. (1992). Inactivated whole SIV vaccine in macaques: evaluation of protective efficacy against challenge with cell-free virus or infected cells. AZDS Res. Hum. Retrooiruses 8(8), 1501-1505. Johnson, R. P., Trocha, A., Yang, L., Mazzara, G. P., Panicali, D. L., Buchanan, T. M., and Walker, B. D. (1991). HIV-1 gag-specific cytotoxic T lymphocytes recognize multiple highly conserved epitopes. Fine specificity of the gag-specific response defined by using unstimulated peripheral blood mononuclear cells and cloned effector cells. J. Immunol. 147(5), 1512- 1521. Johnson, R. P., Trocha, A,, Buchanan, T. M., and Walker, B. D. (1992). Identification of overlapping HLA class I-restricted cytotoxic T cell epitopes in a conserved region of the human immunodeficiency virus type 1 envelope glycoprotein: Definition of minimum epitopes and analysis of the effects of sequence variation. J. Exp. Med. 175(4), 961-971. Johnson, R. P., Trocha, A,, Buchanan, T. M., and Walker, B. D. (1993). Recognition of a highly conserved region of human immunodeficiency virus type 1 gp120 by an HLACw4-restricted cytotoxic T-lymphocyte clone. J. Virol. 67(l), 438-445. Joly, P., Guillon, J.-M., Mayaud, C., Plata, F., Theodorou, I., Denis, M., Debre, P., and Autran, B. (1989). Cell-mediated suppression of HIV-specific cytotoxic T lymphocytes. J. Zmrnunol. 143, 2193-2201. Just, J. J., Abrams, E., Louie, L. G., Urbano, R., Wara, D.,Nicholas, S. W., Stein, Z., and King, M. C. (1995). Influence of host genotype on progression to acquired immunodeficiency syndrome among children infected with human immunodeficiency virus type 1.J. Pediutr. 127(4), 544-549. Kagi, D., Ledermann, B., Burki, K., Seiler, P., Odermatt, B., Olsen, K. J., Podack, E. R., Zinkernagel, R. M., and Hengartner, H. (1994). Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature 369, 31-37.

334

SARAH ROWLAND-JONES ET AL

Kagi, D., Ledermann, B., Burki, K., Zinkernagel, R. M., and Hengartner, H. (1995). Lymphocyte-mediated cytotoxicity in vitro and in vivo: Mechanisms and significance. Immunol. Rev. 146,95-115. Kagi, D., Seiler, P., Pavlovic, J., Ledermann, B., Burki, K., Zinkernagel, R. M., and Hengarh e r , H. (1995).The roles of perforin- and Fas-dependent cytotoxicityin protection against cytopathic and noncytopathic viruses. Eur. J. Immunol. 25(12), 3256-3262. Kalams, S. A. and Walker, B. D. (1995). Cytotoxic T lymphocytes and HIV-1 related neurologic disorders. Curr. Topics Microbiol. Immunol. 202(79), 79-88. Kalams, S. A,, Johnson, R. P., Trocha, A. K., Dynan, M. J., Ngo, S., D’Aquila, R. T., Kurnick, J. T., and Walker, B. D. (1994). Longitudinal analysis of TCR gene usage by HIV-1 envelope-specificCTL clones reveals a limited TCR repertoire. /. Exp. Med. 179, 12611271. Kanki, P., Travers, K., M’Boup, S., Hsieh, C. C., Marlink, R., Gueye-NDiaye, A., Siby, T., Thior, I., Hernandez-Avila, M., Sankale, J. L., Doye, I., and Essex, M. E. (1994). Slower heterosexual spread of HIV-2 than HIV-1. Lancet 343,943-946. Kaplan, C., M d e r , J. Y.,Doinel, C., Lefrere, J. J., Paquez, F., Rouger, P., Salmon, D., and Salmon, C. (1990). HLA-associated susceptibility to acquired immune deficiency syndrome in HIV-1-seropositive subjects. Hum Heredity 40(5),290-298. Kaslow, R. A., Duquesnoy, R., VanRaden, M., Kingsley, L., Marrari, M., Friedman, H., Su, S., Saah, A. J., Detels, R., Phair, J.. et al. (1990). Al, Cw7, B8, DR3 HLA antigen combination associated with rapid decline of T-helper lymphocytes in HIV-1 infection. A report from the Multicenter AIDS Cohort Study [see comments]. Lancet 335(8695), 927-930. Kaslow, R. A,, Carrington, M., Apple, R., Park, L., Munoz, A., Saah, A. J., Goedert, J. J.,Winkler, C., O’Brien, S. J., Rinaldo, C., Detels, R., Blattner, W., Phair, J., Erlich, H., and Mann, D. L. (1996). Influence of combinations of human major histocompatibility complex genes on the course of HIV-1 infection. Nature Med. 2(4), 405-411. Kent, S. J., Hu, S. L., Corey, L., Morton, W. R., and Greenberg, P. D. (1996). Detection of simian immunodeficiency virus (SIV)-specific CD8+ T cells in macaques protected from SIV challenge by prior SIV subunit vaccination. 1.Viml. 70(8), 4941-4947. Klein, M. R., Van, B. C. A., Holwerda, A. M., Garde, S. R. K., Bende, R. J., Keet, I. P. M., Eeftinck, S. J. K. M., Osterhaus, A. D. M. E., Schuitemaker, H., and Miedema, F. (1995). Kinetics of Gag-specific cytotoxic T lymphocyte responses during the clinical course of HIV-1 infection: A longitudinal analysis of rapid progressors and long-term asymptomatics. J. Exp. Med. 181(4), 1365-1372. Klenerman, P., Rowland-Jones, S., McAdam, S., Edwards, J., Daenke, S., Lalloo, D., Koppe, B., Edwards, A., Phillips, R. E., and McMichael, A. J. (1994). Naturally ocurring HIV-1 gag variants antagonise cytotoxic T-cell activity. Nature 369,403-406. Klenerman, P., Meier, U. C., Phillips, R. E., and McMichael, A. J. (1995). The effects of natural altered peptide ligands on the whole blood cytotoxic T lymphocyte response to human immunodeficiency virus. Eur. J. Immunol. 25(7), 1927-1931. Klenerman, P., Phillips, R. E., Rinaldo, C. R., Wahl, L. M., Ogg,G., May, R. M., McMichael, A. J., and Nowak, M. A. (1996). Cytotoxic T lymphocytes and viral turnover in HIV-1 infection. Proc. Natl. Acad. Sci. USA 93, 15323-15328. Knight, S. C. and Macatonia, S. E. (1991).Effect of HIV on antigen presentation by dendritic cells and macrophages. Res. Virol. 142(2-), 123-128. Koenig, S., Earl, P., Powell, D., Pantaleo, G., Merli, S., Moss, B., and Fauci, A. S . (1988). Croup-specific, major histocompatibilitycomplex class-I-restricted cytotoxic responses to human immunodeficiency virus 1 envelope proteins by cloned peripheral blood T cells from an HIV-1 infected individual. Proc. Nutl. Acad. Sci. USA. 85, 8638-8642.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

335

Koenig, S., Fuerst, T. R., Wood, L. V., Woods, R. M., Suzich, J. A., Jones, G. M., de la Cruz, V. F., Davey, R. J., Venkatesan, S., and Moss, B. (1990). Mapping the fine specificity of a cytolytic T cell response to HIV-1 nef protein. J. Zmmunol. 145(1), 127-235. 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., and 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 Med. 1(4),330-336. Korber, B., Koup, R., Walker, B., Haynes, B., Moore, J., and Myers, G. (1995). HZV Mol. Immunol. Database. Theoretical Biology and Biophysics Group T-10, Los Alamos National Laboratory. Koup, R. A. (1994). Virus escape from CTL recognition . J. Exp. Med. 180(3),779-782. Koup, R. A,, Sullivan, J. L., Levine, P. H., Brettler, D., Mahr, A,, Mazzara, G., McKenzie, S., and Panicali, D. (1989). Detection of major histocompatibility complex class Irestricted, HIV-specific cytotoxic T lymphocytes in the blood of infected hemophiliacs. Blood 73(7), 1909-1914. Koup, R. A,, Safrit, J. T., Cao, Y., Andrews, C. A., McLeod, G., Borkowsky, W., Farthing, C., and Ho, D. D. (1994). Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Virol. 68(7), 4650-4655. Kozal, M. J., Shah, N., Shen, N., Yang, R., Fucini, R., Merigan, T. C., Richman, D. D., Morris, D., Hubbell, E., Chee, M., and Gingeras, T. R. (1996). Extensive polymorphisms observed in HIV-1 clade B protease gene using high-density oligonucleotide arrays. Nature Med. 2(7), 753-759. Kundu, S. K., and Merigan, T. C. (1992). CD8+ CDl1+ suppressor cells in HIV-infected asymptomatic patients: Effect on HIV-specific cytotoxicity. Viral Zmmunol. 5( l), 15-25. Lalvani, A., Aidoo, M., Allsopp, C. E., Plebanski, M., Whittle, H. C., and Hill, A. V. (1994). An HLA-based approach to the design of a CTL-inducing vaccine against Plasmodium falciparum. Res. Zmmunol. 145(6), 461-468. Lamhamedi-Cherradi, S., Culmann-Penciolelli, B., Guy, B., Ly, T. D., Goujard, C., Guillet, J. G., and Comard, E. (1995). Different patterns of HIV-1-specificcytotoxic T-lymphocyte activity after primary infection. AIDS (Philadelphia)9(5), 421-426. Landay, A. L., Mackewicz,C. E., and Levy. J. A. (1993).An activatedCD8+ Tcell phenotype correlates with anti-HIV activityand asymptomaticclinical status. Clin. Immunol. Zmmunopathol. 69, 106-116. Lane, H. C., Zunich, K. M., Wilson, W., Cefali, F., Easter, M., Kovacs, J. A., Masur, H., Leitman, S. F., Klein, H. G., Steis, R. G., etal. (1990). Syngeneicbone marrowtransplantation and adoptive transfer of peripheral blood lymphocytes combined with zidovudine in human immunodeficiency virus (HIV) infection. Ann Znternal Med. 113(7), 512-519. Langlade-Demoyen, P., Ngo-Giang-Huong, N., Ferchal, F., and Oksenhendler, E. (1994). HIV nef-specific cytotoxic T lymphocytes in noninfected heterosexual contact of HIVinfected patients. J. Clin. Znoest. 93, 1293-1297. Lanzavecchia, A., Roosnek, E., Gregory, T., Berman, P., and Abrignani, S. (1988). T cells can present antigens such as HIV gp120 targeted to their own surface molecules. Nature 334(6182), 530-532. Lee, R. S., Grusby, M. J., Glimcher, L. H. , Winn, H. J., and Auchincloss, H., Jr. (1994). Indirect recognition by helper cells can induce donor-specific cytotoxic T lymphocytes in vivo. /. Exp. Med. 179(3), 865-872. Levine, B. L., Mosca, J. D., Riley, J. L., Carroll, R. G., Vahey, M. T., Jagodzinski, L. L., Wagner, K. F., Mayers, D. L., Burke, D. S., Weislow, 0. S., Louis, D. C. S., and June,

336

SARAH ROWLAND-JONES ET AL.

C. H. (1996).Antiviraleffect and exvivo CD4+ T cell proliferation in HIV-positive patients as a result of CD28 costimulation. Science (Wmhington D C ) 272(5270), 1939-1943. Levitskaya, J,, Coram, M., Levitsky, V., Imreh, S., Steigenvald Mullen, P. M., Klein, G., Kurilla, M. G., and Masucci, M. G . (1995).Inhibition of antigen processing by the internal repeat region of the Epstein-Barr virus nuclear antigen-1. Nature 375(6533), 685-688. Levy, J. A. (1993). HIV pathogenesis and long-term survival. AlDS 7, 1401-1410. Levy, J. A. (1995). HIV research: A need to focus on the right target. Lancet 345,1619-1621. Levy, J. A., Mackewicz, C. E., and Barker, E. (1996). Controlling HIV pathogenesis: The role of noncytotoxic anti-HIV response of CD8+ T cells. Zmmunol. Today 17(5),217-224. Lightman, S., Cobbold, S., Waldmann, H., and Askonas, B. A. (1987). Do L3T4+ T cells act as effector cells in protection against influenza virus infection. Immunology 62(l ) , 139-144. Lin, Y.,and Askonas, B. A. (1981). Biological properties of an influenza A virus specific killer T cell clone. J. Exp. Med. 154, 225-234. Littaua, R. A,, Oldstone, M., Takeda, A., and Ennis, F. A. (1992). A CD4+ CTL clone to a conserved epitope on HIV-1 p24: Cytotoxic activity and secretion of IL-2 and IL-6. J. Virol. 66(1), 608-611. Livingstone, W. J., Moore, M., Innes, D., Bell, J. E., Simmonds, P., and Group. (1996). Frequent infection of peripheral blood CD8+ T-lymphocytes with HIV-1. Lancet 348,649-654. Lohman, B. L., Miller, C. J., and McChesney, M. B. (1995).Antiviral cytotoxic T lymphocytes in vaginal mucosa of Simian immunodeficiencyvirus-infected rhesus macaques. J. Immunology. 155(12), 5855-5860. Louwagie, J., McCutchan, F. E., Peeters, M., Brennan, T. P., Sanders Buell, E., Eddy, G . A,, van der Groen, G., Fransen, K., Gershy Damet, G. M., Deleys, R., et al. (1993). Phylogenetic analysisof gag genes from 70 international HIV-1 isolates provides evidence for multiple genotypes. AZDS 7(6), 769-780. Lubaki, M. N., Egan, M. A,, Siliciano, R. F., Weinhold, K. J., and Bollinger, R. C. (1994). A novel method for detection and ex vivo expansion of HIV type 1-specific cytolyhc T lymphocytes. AIDS Res. Hum. Retroviruses 10(ll),1427-1431. Lucey, D. R., Zajac, R. A., Melcher, G . P., Butzin, C. A., and Boswell, R. N. (1990). Serum IgE levels in 622 persons with human immunodeficiency virus infection: IgE elevation with marked depletion of CD4+ T-cells. AIDS Res. Hum. Retroviruses 6(4), 427-429. Lukacher, A. E., Braciale, V. L., and Braciale, T. J. (1984). In vivo effector function of influenza virus-specific cytotoxic T lymphocyte clones is highly specific. J. Exp. Med. 160(3),814-826. Luzuriaga, K., Holmes, D., Hereema, A,, Wong, J., Panic&, D. L., and Sullivan, J. L. (1995).HIV-1-specificcytotoxicT lymphocyte responses in the first year oflife.]. lmmunol. 154(l), 433-443. Mackewicz, C., and Levy, J. A. (1992). CD8+ cell anti-HIV activity: Nonlyhc suppression of virus replication. AIDS Res. Hum. Retroviruses 8(6), 1039-1050. Mackewicz, C. E., Ortega, H. W., and Levy, J. A. (1991). CD8+ cell anti-HIV activity correlates with the clinical state of the infected individual.]. Clin. lnoest. 87(4),1462-1466. Mackewicz, C. E., Ortega, H., and Levy, J. (1994a). Effect of cytokines on HIV replication: Lack of identity with the CD8+ antiviral factor. Cell. lmmunol. 153,329-343. Mackewicz, C. E., Yang, L. C., Lifson, J. D., and Levy, J. A. (1994b). Non-cytolyhc CD8 T-cell anti-HIV responses in primary HIV-1 infection. Lancet. 344, 1671-1674. Mackewicz, C. E., J. A. Levy, W. W. Cruikshank, H. Komfeld and D. M. Center. (1996). “Role of IL-16 in HIV replication.” Nature 383, 488-489.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

337

Maggi, E., Giudizi, M. G., Biagiotti, R., Annunziato, F., Manetti, R., Piccinni, M. P., Parronchi, P., Sampognaro, S., Giannarini, L.. Zuccati, G., et al. (1994a).The-like CD8+ T cells showing B cell helper function and reduced cytolytic activity in human immunodeficiency virus type 1 infection. J. Erp. Med. 180(2),489-495. Maggi, E., Mazzetti, M., Ravina, A., Annunziato, F., de Carli, M., Piccinni, M. P., Manetti, R., Carbonari, M., Pesce, A. M., del Prete, G., and et al. (1994b). Ability of HIV to promote a TH1 to THO shift and to replicate preferentially in TH2 and THO cells [see comments]. Science 265(5169), 244-248. Markovitz, D. M. (1993). Infection with the human immunodeficiency virus type 2. Ann. Internal. Med. 118(3), 211-218. Mascola, J. R., Mathieson, B. J., Zack, P. M., Walker, M. C., Halstead, S. B., and Burke, D. S. (1993). Summary report: Workshop on the potential risks of antibody-dependent enhancement in human HIV vaccine trials. AZDS Res. Hum. Retroviruses 9(12), 11751184. Matloubian, M., Concepcion, R. J., and Ahmed, R. (1994). CD4+ T cells are required to sustain CD8+ cytotoxic T-cell responses during chronic viral infection. J. Virol. 68(12), 8056-8063. McAdam, S. N., Klenerman, P., Tussey, L., Rowland-Jones, S. L., M o o , D., Phillips, R. E., Leigh Brown, A., Gotch, F. M., and McMichael, A. J. (1995). Immunogenic HIV variants that bind to HLA-B8 but fail to stimulate CTL responses. J. Zmmunol. 155, 2729-2736. McCbesney, M., Tanneau, F., Regnault, A,, Sansonetti, P., Montagnier, L., Kieny, M. y., and Riviere, Y. (1990). Detection of primary cytotoxic T lymphocytes specific for the envelope glycoprotein of HIV-1 by deletion of the env amino-terminal signal sequence. Eur. /. Zmmunol. 20( 1).215-220. McKeating, J. A,, Gow, J., Goudsmit, J., Pearl, L. H., Mulder, C., and Weiss, R. A. (1989). Characterization of HIV-1 neutralization escape mutants. AZDS 3, 777-784. McMichael, A. J., and Walker, B. D. (1994). Cytotoxic T lymphocyte epitopes: Implications for HIV vaccines. AZDS 8(suppl. l),S155-Sl73. McMichael, A. J., Gotch, F. M., Noble, G. R., and Beare, P. A. S. (1983). Cytotoxic T-cell immunity to influenza. N . Eng. /. Med. 309, 13-17. McMichael, A. J., Gotch, F., and Rothbard, J. (1986). HLA B37 determines an influenza A virus nucleoprotein epitope recognized by cytotoxic T lymphocytes. J. Exp. Med. 164(1397), 1397-1406. McMichael, A. J., Gotch, F. M., Santos-Aguado, J., and Strominger, J. L. (1988). Effect of mutations and variations of HLA-A2 on recognition of a virus peptide epitope by cytotoxic T lymphocytes. Proc. Nad. Acad. Sci. USA 85(9194), 9194-1006. Meier, U.-C., Klenerman, P., Griffin, P., James, W., Koppe, B., Larder, B., McMichael, A. J., and Phillips, R. (1995).T cell receptor antagonism by variable human immunodeficiency viruses. Science 270, 1360-1362. Mellors, J. W., Kingsley, L. A,, Rinaldo, C., Todd, J. , Hoo, V., Kokka, R., and Gupta, P. (1995). Quantitation of HIV-1 RNA in plamsa predicts outcome after seroconversion.” Ann. Intern. Med. 122, 573-579. Meyaard, L., Otto, S. A,, Keet, I. P. M., Van Lier, R. A. W., and Miedema, F. (1994). Changes in cytokine secretion patterns of CD4+ T-cell clones in HIV infection. Blood 12, 4262-4268. Meyerhans, A., Cheynier, R., Albert, J., Seth, M., Kwok, S., Sninsky, J., L., MorfeldtManson, Asjo, B., and Wain-Hobson, S. (1989).Temporal fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolations. Cell 58, 901-910.

338

SARAH ROWLAND-JONES ET AL.

Meyerhans, A., Dadaglio, G., Vartanian, J. P., Langlade-Demoyen, P., Frank, R., Asjo, B., Plata, F., and Wain-Hobson, S. (1991). In vivo persistence of a HIV-1-encoded HLAB27-restricted cytotoxic T lymphocyte epitope despite specific in vitro reactivity. Eur. J. Immunol. 21(10), 2637-2640. Miedema, F., and Klein, M. (1996).AIDS pathogenesis: A finite immune response to blame? Science 272,505. Miedema, F., Meyaard, L., and Klein, M. R. (1993). Protection from HIV or AIDS? Science 262,1074-1075. Montaner, L., and Gordon, S. (1995).Th2 down-regulationof macrophage HIV-1 replication. Science 267,538-539. Montefiori, D. C., Zhou, I. Y., Barnes, B., Lake, D., Hersh, E. M., Masuho, Y., and Lefkowitz, L. B., Jr. (1991).Homotypic antibody responses to fresh clinical isolates of human immunodeficiency virus. Virology 182(2), 635-643. Moore, J. P., Cao, Y., Ho, D. D., and Koup, R. A. (1994). Development of the anti-gpl20 antibody response during seroconversion to human immunodeficiency virus type 1. J. Virol. 68(8), 5142-5155. Moms, A. G., Lin, Y.-L., and Askonas, B. A. (1982).Immune interferon release when a cloned cytotoxic T cell meets its correct influenza-infected targtet cell. Nature 295, 150-152. Moskophidis, D., Laine, E., and Zinkernagel, R. M. (1993a). Peripheral clonal deletion of antiviral memory CD8+ T cells. Eur. J. Immunol. 23(12), 3306-3311. Moskophidis,D., Lechner, F., Pircher, H., andzinkernagel, R. M. (1993b). Virus persistence ,in acutely infected immunocompetent mice by exhaustion of antiviral cytotoxic effector T cells. Nature 362(6422), 758-761. Mosmann, T. R. (1994). Cytokine patterns during the progression to AIDS. Science 265, 193- 194. Mosmann, T. R., Chenvinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L. (1986). Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins. J. Immunol. 136(7), 2348-2357. Moss, P. A. H., Moots, R. J., Rosenberg, W. M. C., Rowland-Jones, S. J., McMichael, A. J., and Bell, J. I. (1991). Extensive conservation of alpha and beta chains of the human T cell antigen receptor recognizing HLA-A2 and influenza matrix peptide. Proc. Natl. Acad. Sci. USA 88,8987-8991. Moss, P. A. H., Rowland-Jones, S. L., Frodsbam, P. M., McAdam, S., Giangrande, P., McMichael, A. J., and Bell, J. I. (1995). Persistent high frequency of human immunodeficiency virus-specific cytotoxic T cells in peripheral blood of infected donors. Proc. Nutl. Acud. Sci. USA 92(13), 5773-5777. Murphey-Corb, M., Martin, L. N., Davison-Fairburn, B., Montelaro, C., Miller, M., West, M., Ohkawa, S., Baskin, G. B., Zhang, J.-Y., Putney, S. D., Allison, A. C., and Eppstein, D. A. (1989). A formalin-inactivated whole SIV vaccine confers protection in macaques. Science 246, 1293-1297. Nara, P. L., Smit, L., Dunlop, N., Hatch, W., Merges, M., Waters, D., Kelliher, J., Gallo, R. C., Fischinger, P. J., and Goudsmit, J. (1990). Emergence of viruses resistant to neutralization by V3-specific antibodies in experimental human immunodeficiency virus type 1 IIIB infection of chimpanzees. J. Virol. 64(8),3779-3791. Neefjes, J. J., Momburg, F., and Hammerling, G. J. (1993). Selective and ATP-dependent translocation of peptides by the mHC-encoded transporter. Science 261,769-771. Nixon, D. F., Townsend, A. R. M., Elvin, J. G., Rizza, C. R., Gallwey, J., and McMichael, A. J. (1988).HIV-1 gag-specificcytotoxicT lymphocytesdefined with recombinant vaccinia virus and synthetic peptides. Nature 336,484-487.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

339

Nowak, M. A., and Bangham, C. R. M. (1996). Population dynamics of immune responses to persistent viruses. Science 272, 74-79. Nowak, M. A., and McMichael, A. J. (1995). How HIV defeats the immune system. Sci. Am. 273(2), 58-65. Nowak, M. A., May, R. M., Phillips, R. E., Rowland-Jones, S., Lalloo, D. G., McAdam, S., Klenerman, P., Koeppe, B., Sigmund, K., Bangham, C. R. M., and McMichael, A. J. (1995).Antigenic oscillations and shifting immunodominance in HIV-1 infections. Nature 375(6532). 606-61 1. Oberlin, E., Amara, A., Bachelerie, F., Bessia, C., Virelizier, J.-L., Arenzana-Seisdedos. F., Schwartz, O., Heard, J.-M., Clark-Lewis, I., Legler, D., Loetscher, M., Baggiolini, M., and Moser, B. (1996). The CXC chemokine SDF-1 is the ligand for LESTWfusin and prevents infection by T-cell-line adapted HIV-1. Nature 382, 833-835. Orentas, R. J., Hildreth, J. E., Obah, E., Polydefkis, M., Smith, G. E., Clements, M. L., and Siliciano, R. F. (1990). Induction of CD4+ human cytolytic T cells specific for HIVinfected cells by a gp160 subunit vaccine. Science 248(4960), 1234-1237. Pantaleo, G., and Fauci, A. S . (1995). New concepts in the immunopathogenesis of HIV infection. Annu. Rev. Immunol. 13, 487-512. Pantaleo, G., Graziosi, C., Butini, L., Pizzo, P. A., Schnittman, S. M., Kotler, D. P., and Fauci, A. S. (1991). Lymphoid organs function as major reservoirs for HIV. Proc. Natl. Acad. Sci. USA 88,9838-9842. Pantaleo, G., Graziosi, C., Demarest, J. F., Butini, L., Montroni, M., Fox, C. H., Orenstein, J. M., Kotler, D. P., and Fauci. A. S . (1993). HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. 362, 355-358. Pantaleo, G., Demarest, J. F., Soudeyns, H., Graziosi, C., F., Denis, J. W., Adelsberger, P., Borrow, Saag, M. S., Shaw, G. M., Sekaly, R. P., and Fauci, A. S. (1994). Major expansion of CD8+ T cells with a predominant Vb usage during the primary immune response to HIV. Nature 370, 463-467. Pantaleo, G., Menzo, S., Vaccarezza, M., Graziosi, C., Cohen, O., Demarest, J., Montefiori, D., Orenstein, J., Fox, C., Schrager, L., Margolick, J., Buchbinder, S., Ciorgi, J., and Fauci, A. S. (1994). Studies in subjects with long-term nonprogressive HIV infection. N . Engl. J. Med. 332,209-216. Pantaleo, G., Vaccarezza, M., Graziosi, C., Cohen, 0. J., and Fauci, A. S . (1996). Antiviral immunity in HIV-1 infected long-term non-progressors. Sem. Virol. 7 , 131-138. Papadopoulos, E. B., Ladanyi, M., Emanuel, D., Mackinnon, S., Boulad, F., Carabisi, M. H., Castro-Malaspina, H., Childs, B. H., Gillio, A. P., Small, T. N., Young, J. W., Kernan, N. A,, and O’Reilly, R. J. (1994). Infusions of donor leukocytes to treat EBVassociated lymphoproliferative disorders after allogeneic bone marrow transplantation. N . Engl. J. Med. 330, 1185-1191. Parker, C. E., Daenke, S., Nightingale, S., and Bangham, C. R. (1992). Activated, HTLVl-specific cytotoxic T-lymphocytes are found in healthy seropositives as well as in patients with tropical spastic paraparesis. Virology. 188(2),628-636. Pedersen, C., Dickmeiss,E., Gaub, J., Ryder, L. P., Platz,P., Lindhardt, B. O.,and Lundgren, J. D. (1990). T-cell subset alterations and lymphocyte responsiveness to mitogens and antigen during severe primary infection with HIV: A case series of seven consecutive HIV seroconverters. AIDS 4(6), 523-526. Pepin, J., Morgan, G., Dunn, D., Gevao, S., Mendy, M., Gaye, I., Scollen, N., Tedder, R., and Whittle, H. (1991). HIV-2-induced immunosuppression among asymptomatic West African prostitutes: Evidence that HIV-2 is pathogenic, but less so than HIV-1. AIDS 5(10), 1165-1172.

340

SARAH ROWLAND-JONESET AL.

Perales, M. A., Schwartz, D. H., Fabry, J. A,, and Lieberman, J. (1995). A vaccinia-gpl60based vaccine but not a gp160 protein vaccine elicits anti-gpl60 cytotoxic T lymphocytes in some HIV-1 seronegative vaccinees.J. Acquir. Immune D$c. Syndr. Hum. Retrovirol. 10(l), 27-35. Perelson, A. S., Neumann, A. U., Markowitz, M., Leonard, J. M., and Ha, D. D. (1996). HIV-1 dynamics in viva: Virion clearance rate, infected cell life-span, and viral generation time. Science 271(5255), 1582-1586. Phillips, A. N. (1996).Reduction of HIV concentration during acute infection: Independence from a specific immune response. Science 271(5248), 497-499. Phillips, R. E., Rowland-Jones, S. L., Nixon, D. F., Gotch, F. M., Edwards, J. P., Ogunlesi, A. 0. , Elvin, J. G., Rothbard, J. A., Bangham, C. R., Rizza, C. R., and McMichael, A. J. (1991). Human immunodeficiency virus genetic variation that can escape cytotoxic T cell recognition. Nature 354(6353), 453-459. Pialoux, G., Excler, J. L., Riviere, Y., Gonzalez Canali, G., Feuillie, V., Coulaud, P., Gluckman, J. C., Matthews, T. J., Meipier, B., Kieny, M. P., et al. (1995). A prime-boost approach to HIV preventive vaccine using a recombinant canarypox virus expressing glycoprotein 160 (MN) followed by a recombinant glycoprotein 160 ( M N U I ) .The AGIS Group, and 1'Agence Nationale de Recherche sur le SIDA. AIDS Res. Hum. Retrouinises 11(3),373-381. Piatak, M., Sag, M., and Yang, L. (1993). High levels of HIV-1 in plasma during all stages of infection determined by competitive PCR. Science 259, 1749-1754. Pinto, L. A., Sullivan, J., Berzofsky, J. A., Clerici, M., Kessler, H. A., Landay, A. L., and Shearer. G. M. (1995).ENV-specificcytotoxicTlymphocyte responses in HIVseronegative health care workers occupationally exposed to HIV-contaminated body fluids. J. Clin. Invest. 96(2),867-876. Pircher, H., Moskphidis, A., Rohrer, U., Burki, K., Hengartner, H., and Zinkernagel, R. M. (1990).Viral escape by selection of cytotoxic T cell-resistant variants in viva. Nature 346,629-633. Plata, F., Autran, B., Martins, L. P., Wain-Hobson, S., Raphael, M., Mayaud, C., Denis, M., Guillon, J. M., and Debre, P. (1987). AIDS virus specific cytotoxic T lymphocytes in lung disorders. Nature 328, 348-351. Plum, G., Siebel, E., Bendick, C., Rasokat, H., Braun, M., Krickeberg, H., and Mauff, G. (1990). Major histocompatibility complex class I to 111 allotypes in patients with AIDSrelated compleflalter-Reed 5, disseminated Kaposi's sarcoma and in normal controls. The ARC-IVIG Study Group. Vox Sang 1, 15-20. Price, D. A., Goulder, P. J. R., Klenerman, P., Sewell, A. K., Easterbrook, P. J.. Troop, M., Bangham, C. R. M., and Phillips, R. E. (1997).Cytotoxic T lymphocyte escape variants during primary infection. Proc. Natl. Acud. Sci. USA 94, 1890-1895. Putney, S. D., Matthews, T. J., Robey, W. G . , Lynn, D. L., Robert Guroff, M., Mueller, W. T., Langlois, A. J., Ghrayeb, J.. Petteway, S. R., Jr., Weinhold, K. J., et al. (1986). HTLV-III/LAV-neutralizingantibodies to an E. coli-produced fragment of the virus envelope. Science 234(4782), 1392-1395. Rammensee, H. G., Friede, T., and Stevanoviic, S. (1995).MHC ligands and peptide motifs: First listing. lmmztnogenetics 41(4): 178-228. Redfield, R. R., Bin, D. L., Ketter, N., Tramont, Polonis, V., Davis, C., Brundage, J. F., Smith, G., Johnson, S., Fowler, A,, et al. (1991). A phase I evaluation of the safety and immunogenicity of vaccination with recombinant a 1 6 0 in patients with early human immunodeficiency virus infection. Military Medical Consortium for Applied Retroviral Research [see comments]. N . Engl. J. Med. 324(24), 1677-1684.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

341

Reid, S., et al. (19%). Antagonistic HIV-1 gag peptides induce structural changes in HLAB8.J. Exp. Med., in press. Reitz, M. S., Wilson, C. , Naugle, C., Gdlo, R. C., and Robert-Guroff, M. (1988). Generation of a nutralization-resistant variant of HIV-1 is due to selection for a point mutation in the envelope gene. Cell 54, 57-63. Reusser, P., Riddell, S. R., Meyers, J. D., and Greenberg, P. D. (1991). Cytotoxic Tlymphocyte response to cytomegalovirus after human allogeneic bone marrow transplantation: Pattern of recovery and correlation with cytomegalovirus infection and disease. Blood 78(5), 1373-1380. Riddell, S. R., and Greenberg, P. D. (1995a). Cellular adoptive immunotherapy after bone marrow transplantation. Cancer Treatment Res. 76, 337-369. Riddell, S. R., and Greenberg, P. D. (1995b). Principles for adoptive T cell therapy of human viral diseases. Annu. Reu. Immunol. 13,545-586. Riddell, S. R., Watanabe, K. S., Goodrich, J. M., Li,C. R., Agha, M. E., and Greenberg, P. D. (1992). Restoration of viral immunity in immunodeficient humans by the adoptive transfer of T cell clones. Science 257, 238-241. Riddell, S. R., Elliott, M., Lewinsohn, D. A,, Gilbert, M. J.. Wilson, L. , Manley, S. A,, Lupton, S. D., Overell, R. W., Reynolds, T. C., Corey, L., and Greenberg, P. D. (1996). T-cell mediated rejection of gene-modified HIV-specific cytotoxic T lymphocytes in HIVinfected patients. Nature Med. 2(2), 216-223. Rinaldo, C., Huang, X.-L., Fan, Z., Ding, M., Beltz, L., Logar, A., Panicali, D., Mazzara, G., Liebmann, J., Cottrill, M., and Gupta, P. (1995a). High levels of anti-HIV-1 memory CTL activity and low viral load are associated with lack of disease in HIV-1-infected longterm non-progressors. ]. Virol. 69, 5838-5842. Rinaldo, C. J., Beltz, L. A., Huang, X. L., Gupta, P., Fan, Z., and Torpey, D. R. (1995b). Anti-HIV type 1 cytotoxic T lymphocyte effector activity and disease progression in the first 8 years of HIV type 1 infection of homosexual men. AIDS Res. Hum. Retroviruses 11(4),481-489. Riviere, Y., Tanneau, S. F., Regnault, A,, Lopez, O., Sansonetti, P., Guy, B., Kieny, M. P., Fournel, J. J., and Montagnier, L. (1989a). Multiple cytotoxic effector cells are induced by infection with the human immunodeficiency virus. Res. Immunol. 140(1),110-115. Riviere, Y., Tanneau-Sdvadori, F., Regnault, A., Lopez, O., Sansonetti, P., Guy, B., Kieny, M.-P., Fournel, J.-J., and Montagnier, L. (1989b). Human immunodeficiencyvirus-specific cytotoxic responses of seropositive individuals: Distinct types of effector cells mediate killing of targets expressing gag and env products. J. Virol. 63, 2270-2277. Riviere, Y., McChesney, M. B., Porrot, F.,Tanneau, S. F., Sansonetti, P., Lopez, O., Pidow, G., Feuillie, V., Mollereau, M., Chamaret, S., et al. (1995). Gag-specificcytotoxicresponses to HIV type 1are associated with a decreased risk of progression to AIDS-related complex or AIDS. AIDS Res. Hum. Retroviruses ll(8): 903-907. Rock, K. L., and Clark, K. (1996). Analysis of the role of MHC class I1 presentation in the stimulation of cytotoxic T lymphocytes by antigens targeted into the exogenous antigenMHC class I presentation pathway. J. Immunol. 156(lo), 3721-3726. Romagnani, S. (1994). Lymphokine production by human T cells in disease states. Annu. Rev. Immunol. 12, 227-257. Romagnani, S., Del Prete, G., Manetti, R., Ravina, A., Annunziato, F., De Cadi, M., Mazzetti, M., Piccinni, M.-P., D’Elios, M., Parronchi, P., Sampognaro, S., and Ma@, E., (1994). Role of ThlRh2 cytokines in HIV infection. Immunol. Rev. 140, 73-92. Rooney, C. M., Smith, C. A., Ng, C. Y. C., Loftin, S., Li, C., Krance, R. A., Brenner, M. K., and Heslop, H. E. (1995). Use of gene-modified virus-specific T lymphocytes to control EBV-related lymphoproliferation. Lancet 345, 9-13.

342

SARAH ROWLAND-JONES ET AL.

Roos, M., De Leeuw, N., Claessen, F., Huisman, H., Kootstra, N., Meyaard, L., Schellekens, P., Schuitemaker, H., and Miedema, F. (1994). Viro-immunological studies in acute HIV1 infection. AZDS 8, 1533-1538. Rowe, M., Rowe, D. T., Gregory, C. D., Young, L. S., Farrelli, P. J., Rupani, H., and Rickinson, A. B. (1987). Differences in B cell growth phenotype reflect novel patterns of Epstein-Barr virus latent gene expression in Burkitt's lymphoma. EMBO J. 6, 2743. Rowe, M., Young, L. S., Crocker, J., Stokes, H., Henderson, S., and Rickinson, A. B. (1991). Epstein-Ban virus (EBV)-associated lymphoproliferative disease in the SCID mouse model: Implications for the pathogenesis of EBV-positive lymphomas in man. J. Exp. Med. 173, 147-158. Rowland-Jones, S., and McMichael, A. (1993). Cytotoxic T lymphocytes in HIV infection. Sem. Virol. 4, 83-94. Rowland-Jones, S. L., Phillips, R. E., Nixon, D. F., Gotch, F. M, Edwards, J. P., Ogunlesi, A. O., Elvin, J. G., Rothbard, J. A., Bangham, C. R., Rizza, C. R., et al. (1992). Human immunodeficiencyvirus variants that escape cytotoxic T-cell recognition. AZDS Res. Hum. Retrooiruses 8(8), 1353-1354. Rowland-Jones, S. L., Nixon, D. F., Aldhous, M. C., Gotch, F., Ariyoshi, K., Hallam, N., Kroll, J. S., Froebel, K., and McMichael, A. J. (1993a). HIV-specific CTL activity in an HIV-exposed but uninfected infant. Lancet 341, 860-861. Rowland-Jones, S. L., Powis, S. H., Sutton, J., Mockridge, I., Gotch, F. M., Murray, N., Hill, A. B., Rosenberg, W. M., Trowsdale, J., and McMichael, A. J. (1993b). An antigen processing polymorphism revealed by HLA-B8-restricted cytotoxic T lymphocytes which does not correlate with TAP gene polymorphism. Eur. J. Zmmunol. 23(8), 1999-2004. Rowland-Jones, S. L., Sutton, J., Ariyoshi, K., Dong, T., Gotch, F. M., McAdam, S., Whitby, D., Sabally, S., Gallimore, A., Corrah, T., Takiguchi, M., Schultz, T., McMichael, A. J., and Whittle, H. (1995). HIV-specific cytotoxic T cells in HIV-exposed but uninfected Gambian women. Nature Med. 1(1), 59-64. Ruprecht, R. M., Baba, T. W., and Liska, V. (1996). Attenuated HIV vaccine: Caveats. Science 271(5257), 1790-1792. Sadat-Sowti, Debre, P., Idziorek, T., Guillon, J.-M., Hadida, F., Okzenhendler, E., Katlama, C., Mayaud, C., and Autran, B. (1991). A lectin-binding soluble factor released by CD8+ CD57+ lymphocytes from AIDS patients inhibits T cell cytotoxicity. Eur. J. Zmmunol. 21, 737-741. Sadat-Sowti, B., Parrot, A., Quint, L., Mayaud, C., Debre, P., and Autran, B. (1994).Alveolar CD8+ CD57+ lymphocytes in HIV infection produce an inhibitor of cytotoxic functions. Am. J. Respir. Crit. Care Med. 14,Q 972-980. Safrit, J. T., and Koup, R. A. (1995). The immunology of primary HIV infection: Which immune responses control HIV replication? C u m @in. Zmmunol. 7, 456-461. Safrit, J. T., Andrews, C. A., Zhu, T., Ho, D. D., and Koup, R. A. (1994a). Characterization of human immunodeficiency virus type 1-specificcytotoxic T lymphocyte clones isolated during acute seroconversion:Recognitionof autologousvirus sequences within a conserved immunodominant epitope. J. Exp. Med. 179(2),: 463-472. Safrit, J. T., Lee, A. Y., Andrews, C. A,, and Koup, R. A. (1994b). A region of the third variable loop of HIV-1 a 1 2 0 is recognized by HLA-B7-restricted CTLs from two acute seroconversion patients. 1.Immunol. 153(8), 3822-3830. Sahmoud, T., Laurian, Y., Gazengel, C., Sultan, Y., Gautreau, C., and Costagliola. D. (1993). Progression to AIDS in French haemophiliacs: Association with HLA-B35. AIDS 741, 497-500. Sauzet, J. P., Deprez, B., Martinon, F., Guillet, J. G., Gras Masse, H., and Gomard, E., (1995). Long-lasting anti-viral cytotoxic T lymphocytes induced in vivo with chimericmultirestricted lipopeptides. Vaccine 13(14), 1339-1345.

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

343

Schmidtmayerova, H., Sherry, B., and Bukrinsky, M. (1996).Chemokines and H N replication. Nature 382, 767. Schrager, L. K., Young, J. M., Glenn Fowler, M., Mathieson, B. J., and Vermund, S. H., (1994). Long-term sunivofi of HIV-1 infection: Definitions and research challenges. AIDS ~ ( S U P Pl), ~ . S95-SlO8. Sethi, K. K., Naher, H., and Stroehmann,I. (1988).Phenotypic heterogeneity of cerebrospinal fluid-derived HIV-specific and HLA-restricted cytotoxic T-cell clones. Nature 335(6186), 178-181. Shaper, E. G., and Mullins, J. I. (1993).Rates of amino acid change in the envelope protein correlate with pathogenicity of primate lentiviruses.J. Mol. Eool. 37(l),57-65. Shen, L.. Chen, Z. W., Miller, M. D., Stallard, V., Mazzara, G. P., Panicali, D. L., and Letvin, N. L. (1991). Recombinant virus vaccine-induced SIV-specific CD8+ cytotoxic T lymphocytes. Science 252(5004), 440-443. Shirai, M., Pendleton, C. D., Ahlers, J., Takeshita, T., Newman, M., and Berzofsky, J. A. (1994). Helper-cytotoxic T lymphocyte (CTL) determinant linkage required for priming of anti-HIV CD8+ CTL in vivo with peptide vaccine constructs. /. Immunol. 152(2), 549-556. Shirai, M., Chen, M., Arichi, T., Masaki, T., Nishioka, M., Newman, M., Nakazawa, T., Feinstone, S. M. and Berzofsky, J. A. (1996).Use of intrinsic and extrinsic helper epitopes for in vivo induction of anti-hepatitis C virus cytotoxic T lymphocytes (CTL) with CTL epitope peptide vaccines. J. Infect. Dis. 173(1),24-31. Siliciano, R. F., Lawton, T., Knall, C., Karr, R. W., Berman, P., Gregory, T., and Reinhen, E. L. (1988). Analysis of host-virus interactions in AIDS with anti-gpl20 T cell clones: Effect of HIV sequence variation and a mechanism for CD4+ cell depletion. Cell. 54,561-575. Steel, C. M., Ludlam, C. A., Beatson, D., Peutherer, J. F., Cuthbert, R. J., Simmonds, P., Morrison, H., and Jones, M. (1988).HLA haplotype A1 B8 DR3 as a risk factor for HIVrelated disease. Lancet 1(8596), 1185-1188. Steimer, K. S., Scandella, C. J,, Skiles, P. V., and Haigwood, N. L. (1991). Neutralization of divergent HIV-1 isolates by conformation-dependent human antibodies to Gp120. Science 254(5028), 105-108. Steinman, R. M. (1991). The dendritic cell system and its role in immunogenicity.Annu. Rev. lmmunol. 9(271), 271-296. Stott, E. J,, Chan, W. L., Mills, K. H., Page, M., Taffs, F., Cranage, M., Greenaway, P. and Kitchin., P. (1990). Preliminary report: Protection of cynomolgus macaques against simian immunodeficiency virus by fixed infected-cell vaccine [see comments]. Lancet 336(8730). 1538- 1541. Stott, J., Kitchin, P., Page, M., Flanagan, B., Taffs, L., Chan, W., Mills, K., Silvera, P., and Rodgers, A. (1991). Anti-cell antibody in macaques. Nature 353, 393. Street, N. E., Schumacher, J. H., Fong, T. A. T., Bass, H., Fiorentino, D. F., Leverah, J. A,, and Mosmann, T. R. (1990). Heterogeneity of mouse helper T cells: Evidence from bulk cultures and limiting dilution cloning for precursors of Thl annd Th2 cells. J. lmmunol. 144, 1629. Thorsby, E. (1996).HLA associated diseases, a summary of the workshop component. Proc. 12th Int. Histocompatibility Workshop, in press. Toso, J. F., Chen, C.-H., Mohr, J. R., Piglia, L., Oei, C., Ferrari, G., Greenberg, M. L., and Weinhold, K. J. (1995).Oligoclond CD8 lymphocytes from persons with asymptomatic HIV-1 infection inhibit HIV-1 replication. ]ID 172, 964-973. Townsend, A., and Bodmer, H. (1989).Antigen recognition by class I -restricted cytotoxic T lymphocytes. Annu. Rev. Immunol. 7 , 601-624.

344

SARAH ROWLAND-JONES ET AL.

Townsend, A., and Trowsdale, J. (1993). The transporters associated with antigen presentation. Sem. Cell. Biol. 4(1), 53-61. Townsend, A,, Rothbard, J., Gotch, F., Bahadur, B., Wraith, D., and McMichael, A. (1986). The epitopes of influenza nucleoprotein recognized by cytotoxic T lymphocytes can be defined with short synthetic peptides. Cell 44(959), 959-968. Tripp, R. A., Hou, S., McMickle, A., Houston, J., and Doherty, P. C. (1995). Recruitment and proliferation of CD8+ T cells in respiratory virus infections. 1.Immunol. 154(l l ) , 6013-6021. Tsomides, T. J., Aldovini, A,, Johnson, R. P., Walker, B. D., Young, R. A,, and Eisen, H. N. (1994). Naturally processed viral peptides recognized by cytotoxic T lymphocytes on cells chronically infected by human immunodeficiency virus type 1. I. Exp. Med. 180(4), 1283-1293. Tsubota, H. C., Lord, C. I., Watkins, D. I., Morimoto, C., and Letvin, N. L. (1989). A cytotoxic T lymphocyte inhibits acquired immunodeficiency syndrome virus replication in peripheral blood lymphocytes. 1.Exp. Med. 169, 1421-1434. van Baalen, C. A,, Klein, M. R., Geretti, A. M., Keet, R. I., Miedema, F., van Els, C. A., and Osterhaus, A. D. (1993). Selective in vitro expansion of HLA class I-restricted HIV1Gag-specificCD8+ T cells: CytotoxicT-lymphocyte epitopes and precursor frequencies. AIDS 7(6),781-786. Van Bleek, G. M., and Nathenson, S. M. (1990). Isolation of an endogenously processed immunodominant viral peptide from the class I H2Kb molecule. Nature 348, 213. Venet, A,, and Walker, B. D. (1993). Cytotoxic T-cell epitopes in HIV/SIV infection. AIDS. 7, Suppl, S117-Sl26. Vigano, A., Principi, N., Crupi, L., Onorato, J., Vincenzo, Z . G., and Salvaggio, A. (1995). Elevation of IgE in HIV-infected children and its correlation with the progression of disease. J. Allergy. Clin. Immunol. 95(2), 627-632. Van Gegerfelt, A., Albert, J., Morfeldt Manson, L., Broliden, K., and Fenyo, E. M. (1991). Isolate-specific neutralizing antibodies in patients with progressive HIV-1-related disease. Virology 185(l),162-168. van Herrath, M. G., Yokoyama, M., Dockter, J., Oldstone, M. B., and Whitton, J. L. (1996).CDCdeficient mice have reduced levels of memory cytotoxic T lymphocytes after immunization and show diminished resistance to subsequent virus challenge. 1. Virol. 70(2), 1072- 1079. Vyakarnam, A., Matear, P. M., Martin, S. J., and Wagstaff, M. (1995). Thl cells specific for HIV-1 gag p24 are less efficient than Tho cells in supporting HIV replication, and inhibit virus replication in Tho cells. ImmunoZogy 86, 85-96. Wain Hobson, S. (1995). AIDS. Virological mayhem. Nature 373(6510), 102. Walker, B. D., Chakrabarti, S., Moss, B., Paradis, T. J., Flynn, T., Duma, A. G., Blumberg, R. S., Kaplan, J. C., Hirsch, M. S., and Schooley, R. T. (1987). HIV-specific cytotoxic T lymphocytes in seropositive individuals. Nature 328(6128), 345-348. Walker, B. D., Flexner, C., Paradis, T. J., Fuller, T. C., Hirsch, M. S., Schooley, R. T., and Moss, B. (1988). HIV-1 reverse transcriptase is a target for cytotoxic T lymphocytes in infected individuals. Science 240(64-66). Walker, B. D., Flexner, C., Birch, L. K., Fisher, L., Paradis, T. J., Aldovini, A,, Young, R., Moss, B., and Schooley, R. T. (1989). Long-term culture and fine specificity of human cytotoxic T-lymphocyte clones reactive with human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 86(23), 9514-9518. Walker, C. M., Moody, D. J., Stites, D. P., and Levy,J. A. (1986). CD8+ lymphocytes can control HIV infection in vitro by suppressing virus replication. Science 234, 1563-1566. Walker, C. M., Erickson, A. L., Hsueh, F. C., and Levy, J. A. (1991). Inhibition of human

CELLULAR IMMUNITY IN PROTECTION AGAINST HIV

345

immunodeficiencyvirus replication in acutely infected CD4+ cells by CD8+ cells involves a noncytotoxic mechanism. /. Virol. 65(ll),5921-5927. Walker, R., Blaese, R. M., Carter, C. S., Chang, L., Klein, H., Lane, H. C., Leitman, S. F., Mullen, C. A., and Larson, M. (1993). A study of the safety and survival of the adoptive transfer of genetically marked syngeneic lymphocytes in HIV-infected identical twins. Hum. Gene. Ther. 4(5), 659-680. Wallace, L. E., Rickinson, A. B., Rowe, M., and Epstein, M. A. (1982a). Epstein-Barr virus-specific cytotoxic T-cell clones restricted through a single HLA antigen. Nature 297(5865),413-415. Wallace, L. E., Rowe, M., Gaston, J. S., Rickinson, A. B., and Epstein, M. A. (198213). Cytotoxic T cell recognition of Epstein-Barr virus-infected B cells. 111. Establishment of HLArestricted cytotoxic T cell lines using interleukin 2. Eur. /. lmmunol. 12(12), 1012-1018. Walter, E. A., Greenberg, P. D., Gilbert, M. J., Finch, R. J., Watanabe, K. S., Thomas, E. D., and Riddell, S. R. (1995a).Reconstitution of cellular immunity against cytomegalovirus in recipients of allogeneic bone marrow by transfer of T-cell clones from the donor. N . Engl. /. Med. 333(16), 1038-1044. Watret, K. C., Whitelaw, J. A., Froebel, K. S., and Bird, A. G. (1993).Phenotypic characterization of CD8+ T cell populations in HIV disease and in anti-HIV immunity. Clin. Exp. lmmunol. 92, 93-99. Wei, X., Ghosh, S. K., Taylor, M. E., Johnson, V. A., Emini, E. A., Deutsch, P., Lifson, J. D., Bonhoeffer, S., Nowak, M. A., Hahn, B. H., et al. (1995). Viral dynamics in human immunodeficiency virus type 1 infection. Nature 373(6510), 117-122. Weinhold, K. J., Lyerly, H. K., Matthews, T. J., Tyler, D. S., Ahearne, P. M., Stine, K. C., Langlois, A. J., Durack, D. T., and Bolognesi, D. P. (1988). Cellular anti-GPl2O cytolytic reactivities in HIV-1 seropositive individuals. Lancet 1(8591), 902-905. Weiss, R. A. (1996). HIV receptors and the pathogenesis of AIDS. Science 272,1885-1886. Weiss, R. A,, Clapham, P. R., Cheingsong Popov, R., Dalgleish, A. G., Came, C. A., Weller, I. V., and Tedder, R. S. (1985). Neutralization of human T-lymphotropic virus type I11 by sera of AIDS and AIDS-risk patients. Nature 316(6023), 69-72. Whittle, H., Morris, J., Todd, J., Corrah, T., Sabally, S., Bang&, J., Ngom, P., Rolfe, M., and Wilkins, A. (1994). HIV-2-infected patients survive longer than HIV-1-infected patients. AlDS 8, 1617-1620. Wiertz, E. J. H. H., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J., and Ploegh, H. L. (1996). The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84, 769-779. Winchester, R., Chen, Y.,Rose, S., Selby, J., and Borkowsky, W. (1995). Major histocompatibility complex class I1 DR alleles DRB1'1501 and those encoding HLA-DR13 are preferentially associated with a diminution in maternally transmitted human immunodeficiency virus 1 infection in different ethnic groups: Determination by an automated sequencebased typing method. Proc. Natl. Acad. Sci. USA 92(26), 12374-12378. Wolfs, T. F., de Jong, J. J., Van den Berg, H., Tijnagel, J. M., Krone, W. J., and Goudsmit. J. (1990). Evolution of sequences encoding the principal neutralization epitope of human immunodeficiency virus 1 is host dependent, rapid, and continuous. Proc. Natl. Acad. Sci. USA 87(24), 9938-9942. Wolinsb, S. M., Korber, B. T. M., Neumann, A. U., Daniels, M., Kunstman, K. J., Whetsell, A. J., Furtado, M. R., Cao, Y., Ho, D. D., Safrit, J. T., and Koup, R. A. (1996). Adaptive evolution of human immunodeficiency virus-type 1during the natural course of Infection. Science 272, 537-540. Xu, X.-N., Screaton, G. R., Gotch, F. M., Dong, T., Tan, R., Almond, N., Stebbings. R., Kent, K., Nagata, S., Stott, J. E., and McMichael, A. J. (1997). Evasion of CTL responses by net-dependent induction of Fas ligand expression on SIV-infected cells. Submitted.

346

SARAH ROWLAND-JONES ET AL.

Yamamoto, H., Miller, M. D., Tsubota, H., Watkins, D. I., Mazzara, G. P., Stallard, V., Panicali, D. L., Aldovini, A., Young, R. A., and Letvin, N. L. (1990). Studies of cloned simian immunodeficiency virus-specific T lymphocytes. gag-Specific cytotoxic T lymphocytes exhibit a restricted epitope specificity. ] Immunol. 144(9),3385-3391. Yang, O., Kalams, S., Rosenweig, M., Trocha, A., Jones, N., Koziel, M., Walker, B. D., and Johnson, R. P. (1996).Efficient lysis of HIV-1-infected cells by cytotoxic T lymphocytes. J. Vtrol. 70, 5799-5806. Yasutomi, Y., Koenig, S.,Haun, S. S., Stover, C. K., Jackson, R. K., Conard, P., Conley, A. J.,Emini, E. A., Fuerst, T. R., and Letvin, N. L. (1993a).Immunization with recombinant BCG-SIV elicits SIV-specific cytotoxic T lymphocytes in rhesus monkeys. 1. Immunol. 150(7), 3101-3107. Yasutomi, Y.,Reimann, K. A., Lord, C. I., Miller, M. D., and Letvin, N. L. (1993b).Simian immunodeficiency virus-specific CD8+ lymphocyte response in acutely infected rhesus monkeys. I. Virol. 67(3),1707-1711. Yasutomi, Y., Robinson, H. L., Lu, S., Mustafa, F., Lekutis, C., Arthos, J., Mullins, J. I., Voss, G., Manson, K., Wyand, M., and Letvin, N. L. (1996). Simian immunodeficiency virus-specific cytotoxic T-lymphocyte induction through DNA vaccination of rhesus monkeys. 1.Virol. 70(l),678-681. Young, J. W., and Steinman, R. M. (1990).Dendritic cells stimulate primary human cytolytic lymphocyteresponses in the absence of CD4+ helperT cells.]. Exp.Med. 171(4),1315-1332. Young, L., Alfieri, C., Henessey, K., Evans, H., OHara, C., Anderson, K. C., Ritz, J., Shapiro, R. S., Rickinson, A.. Kieff, E., and Cohen, J. I. (1989).Expression of Epstein-Barr virus transformation associated genes in tissues of patients with EBV lymphoproliferative disease. N . Engl. I. Med. 321, 1080. Zarling, J. M., Ledbetter, J. A., Sias, J., Fultz, P., Eichberg, J., Gjerset, G., and Moran, P. A. (1990). HIV-infected humans, but not chimpanzees, have circulating cytotoxic T lymphocytes that lyse uninfected CD4+ cells. 1.Immunol. 144(8), 2992-2998. Zhang, L. Q., MacKenzie, P., Cleland, A., Holmes, E. C., Leigh Brown, A. J., and Simmonds, P. (1993). Selection for specific sequences in the external envelope protein of HIV-1 upon primary infection.1.Virol. 67, 3345-3356. Zhu, T., Mo, H., Wang, N., Nam, D. S., Cao, Y., Koup, R. A., and Ho, D. D. (1993). Genotypic and phenotypic characterization of HIV-1 in patients with primary infection. Science. 261, 1179-1181. Zinkemagel, R. (1995). Are HIV-specific CTL responses salutary or pathogenic? Curt-. Opin. Immunol. 7,462-470. Zinkemagel, R. M. (1996).Immunology taught by viruses. Science 271(5246), 173-178. Zinkemagel, R. M., and Doherty, P. C. (1974). Restriction of an in vitro T cell-mediated cytotoxicity in lymphocytic choriomeningitiswithin asyngeneic or semiallogeneicsystem. Nature 248,701-702. Zinkemagel, R. M., and Doherty, P. C. (1975). H-2 compatibility requirement for T-cell mediated lysis of target cells infected with lymphocytic choriomeningitis virus. 1. Exp. Med. 141, 1427-1436. Zinkemagel, R. M., Haenseler, E., Leist, T., Cemy, A., Hengartner, H., and Althage, A. (1986). T cell-mediated hepatitis in mice infected with lymphocytic choriomeningitis virus. Liver cell destruction by H-2 class I-restricted virus-specific cytotoxic T cells as a physiological correlate of the 51Cr-release assay?]. Exp. Med. 164(4), 1075-1092. Zweerink, H. J.. Coutneidge, S. A., Skehel, J. J., Crumpton, M. J., and Askonas, B. A. (1977).Cytotoxic T cells kill influenza virus infected cells but do not distinguish between serologically distinct type A viruses. Nature 267, 354-356. This article was accepted for publication on 31 October 1996.