CD8+ T LYMPHOCYTES IN HIV-1 INFECTION

Many studies have indicated the importance of CD8⁺ T lymphocytes in limiting viral replication. During primary infection, most individuals develop human immunodeficiency virus (HIV)-specific cytotoxic T lymphocytes (CTLs), which are believed to contribute to a significant decrease in viremia.¹ However, disease progression has been associated with either loss of these cells in vivo²-⁵ or impaired function.⁶-⁷ Similarly, subjects with asymptomatic HIV-1 infection maintain CD8⁺ T cells that suppress HIV-1 replication without lysis of the infected cells, but this activity is also apparently lost in the course of disease progression.⁸ Therefore, an understanding of the maintenance of antiviral CD8⁺ T cells is critical to an understanding of HIV-related pathogenesis as well as guiding the design of vaccines and adjuvants to maximize cell-mediated immunity.

Apoptosis exerts tight control over the number of circulating CD8⁺ T lymphocytes throughout life, particularly after the expansion of effector cells during acute viral infections, such as in the dramatic response to Epstein-Barr virus (EBV) infection in adolescents or adults.^9-11 However, the rate of apoptosis of CD8⁺ T lymphocytes remains elevated throughout the course of HIV-1 infection and, therefore, the question arises whether this is abnormal and contributes to the pathogenesis of HIV-associated immunodeficiency.

Expansion and Contraction of CD8⁺ T Lymphocytes

during Responses to Viral Infection

To maintain homeostasis, the increase in cell numbers due to antigen-specific proliferation must be balanced by death of an equal number of cells.

The cytokine milieu during the initial activation is believed to influence the fate of the antigen­specific CD8⁺ T cells. Live bacteria are superior to heat-killed antigen in generating memory CD8⁺ T cells.¹⁸ Similarly, absence of CD4⁺ T cell help via CD40L¹⁹,²⁰ and antigen-presenting cell (APC) interactions via 4-1BB/4-1BBL²¹ leads to impaired generation of memory CD8⁺ T cells, as does an absence of interleukin (IL)-15 signaling.²² Also, the conversion rate of effector memory cells to resting central memory cells is determined during the activation stage, and these latter cells are capable of extensive proliferation and are better able to control viral replication in vivo.²³

The total number of memory CD8⁺ T cells is relatively stable throughout much of adult life.²⁴ Therefore, the viral antigen-specific cells that are retained after the peak of expansion must replace preexisting cells in the total CD8⁺ T lymphocyte population.²⁵ The preexisting cells may be primed for apoptosis by IFN type I,²⁶ and it is also possible that newly generated memory CD8⁺ T cells may outcompete older memory cells for essential survival factors.²⁷

These results suggest that long-term memory CD8⁺ T cell survival is programmed during the initial differentiation, involving helper CD4⁺ T cells and antigen-presenting cells, and the regulation of subsequent apoptosis is central to the maintenance of antiviral CD8⁺ T lympho­cytes.

Role of TNF/TNFR Superfamily Members in Homeostasis

of CD8⁺ T Lymphocytes

Activated effector cells exhibit increased expression of CD95 and tumour necrosis factor receptor 1 (TNFRl) and TNFR2, which are known to mediate induction of apoptosis in vitro.²⁸ However, knockout studies in mice showed that deletion of CD95 or either tumor necrosis factor (TNF)-α or TNFRp55 did not affect apoptosis of effector cells.²⁹-³⁰ Nevertheless, in some reports, a contri­bution of TNF-α was observed in the deletion of activated peripheral CD8 T cells.³¹-³² Loss of Fas/FasL function does not lead to increases in CD8⁺ T cell numbers,³³ and loss of TNF/TNFR (TNF receptor) function reduces inflammation rather than disturbing CD8⁺ T cell homeostasis.²⁸ Also, caspase-8 links CD95 and TNFR ligation to apoptosis,²⁸ but human caspase-8 deficiency leads to immunodeficiency rather than unrestrained lymphoproliferation.³⁴ Therefore, the physio­logical role of apoptosis mediated by these ligands, at least in viral infections, remains unclear.

Role of Bcl-2 Family Members in CD8⁺ T Cell Apoptosis

Many different studies in mice and humans have shown that expanded CD8⁺ T cells during acute viral infection, and antigen-specific cells in particular, reduce intracellular levels of Bcl-2.^9,10,35 The decrease in Bcl-2 in effector cells is associated with a decrease in IL-7 receptor (IL-7R), whereas long-term memory cells regain expression of IL-7R and Bcl-2.^36,37 This situation is similar to early T cell progenitors, for which expression of IL-7R is required for upregulation of Bcl-2 and survival during ontogeny.^38,39

However, it was found in two studies that overexpression of Bcl-2 did not prevent apoptosis of excess effector cells after viral infections in murine models.^40,41 A recent study showed that a knockout dose of proapoptotic Bim, a BH3-only member of the Bcl-2 family that antagonizes the effect of Bcl-2,⁴² prevented apoptosis of effector cells.⁴³ These results suggest that an alternative antiapoptotic member of the Bcl-2 family may be important in effector T cell survival.

In this regard, a recent study showed that MCL-1, another antiapoptotic member of the Bcl-2 family, is essential for peripheral T cell survival, is induced by IL-7, and interacts strongly with Bim.⁴⁴ MCL-1 also antagonizes the proapoptotic effect of Bak.⁴⁵ Current data suggest that Bak combines with Bax to mediate lymphocyte apoptosis via the intrinsic mitochondrial pathway.^46,47 This is normally actively prevented by Bcl-2 and MCL-1, which are, in turn, counteracted by apoptotic signals leading to transcriptional or posttranslational changes in the BH3-only proteins Bim, Bid, and Bad.^48,49

Studies of lymphoid tumorigenesis suggest there is a reciprocal interaction between upregula­tion of c-Myc, which promotes both proliferation and apoptosis, and Bcl-2, which promotes quiescence and survival.^48,50 Proliferation of effector T cells, which is regulated by c-Myc,⁵¹ has long been known to involve cytokines that use the common γ-chain as part of their receptors, especially IL-2.⁵² However, IL-2 also seems to regulate Bcl-2 expression and cell survival, in addition to proliferation, as detailed below.

Role of Common Φ-Chain Cytokines in CD8⁺ T Cell Apoptosis

In vitro, IL-2 is a potent growth factor for CD8⁺ T cells, but in vivo, its role as a growth and survival factor^53,54 is much more enigmatic, even though IL-2 is able to induce upregulation of Bcl-2.⁵⁵ IL-2 may actually inhibit the survival of memory CD8⁺ T cells in vivo,⁵⁶ and a knockout dose of IL-2 or IL-2R results in T cell lymphocytosis^57,58 rather than leading to a deficit of peripheral T cells. The inhibitory effects of exposure to IL-2 in vivo may be due to its induction of FasL expression and possible fratricide⁵⁹ as well as its major role in generating regulatory T cells.⁶⁰

However, IL-7 and IL-15 are unequivocally survival factors for mature CD8⁺ T cells in vivo.⁶¹,⁶² In particular, expression of the IL-7Rα chain is closely correlated with Bcl-2 expression and long-term survival.^36,37 Also, IL-15 is required for CD8⁺ memory cell generation and survival,²²’^63-65 and IL-15Rα is required for maintenance of Bcl-2 levels in the periphery.^66,67 IL-15 seems to counteract a proapoptotic effect of IL-2 and increase the survival of memory CD8⁺ T cells.⁵⁶ In addition to direct survival effects, IL-15 induces low-level turnover of memory CD8⁺ T cells⁶⁸ as well as mediates homeostatic prolifer­ation in response to lymphopenia.⁶⁹

Therefore, the common γ-chain cytokines, IL-7, IL-2, and IL-15, seem to be central to CD8⁺ effector and memory homeostasis, controlling proliferation as well as survival via regulation of Bcl-2 and possibly MCL-1.

Role of IFN Type I in CD8⁺ T Cell Apoptosis

Another important factor in CD8 responses is IFN type I, which was originally described to induce nonspecific proliferation in vivo.⁷⁰ Subsequently, this effect of IFN type I was shown to be mediated via induction of IL-15 synthesis.⁶⁸ Production of IFN type I is a hallmark of viral infections and is mainly synthesized in vivo by plasmacytoid dendritic cells.⁷¹ IFN-β was shown to rescue activated T cells from apoptosis,⁷²’⁷³ possibly via activation of protein kinase activity.⁷⁴

IFN type I, at the same time, may lead to attrition of older memory cells.²⁶ IFN-α has also been shown to be involved in inducing apoptosis of B cells via translocation of Daxx to PML bodies⁷⁵ as well as apoptosis of multiple myeloma cells via tumor necrosis factor-related apoptosis­inducing ligand (TRAIL).⁷⁶ Therefore, again, a regulator of apoptosis may have complex, contra­dictory effects, but in the case of viral infections, it should act to inhibit CD8⁺ T effector cell apoptosis through the action of IL-15.

Role of Molecules Produced by Effector T Cells in CD8⁺ T Cell Apoptosis

Recently, it was shown that disruption of the perforin gene leads to impaired contraction of primary CD8⁺ T cell expansion⁷⁷ and unrestrained secondary CD8⁺ T cell responses.⁷⁸ Similarly, a knockout dose of the IFN-γ gene delays contraction of CD8⁷⁷ and CD4 responses.⁷⁹

An emerging concept is that many products of an activated T cell, which are believed to mediate important antiviral effector functions as well as immune-activating functions, also have a feedback role in homeostasis. In addition to perforin and IFN-γ, these products include TNF-α, FasL, IL-2, and, possibly, CD70 (via CD27).⁸⁰

activation, proliferation, and apoptosis OF CD8⁺ t lymphocytes during HIV infection

HIV-1 infection typically results in a long-term doubling of the number of circulating CD8⁺ T lymphocytes⁸¹ as well as an increased number of effector phenotype CD8⁺ T cells in lymphoid tissue.⁵ Paradoxically, as detailed below, there is an increased rate of apoptosis of these cells in vivo, and in cultured cells ex vivo, compared with CD8⁺ T cells from healthy uninfected individuals.

Dynamics of CD8⁺ T Lymphocytes in Chronic HIV-1 Infection

HIV-1 infection is characterized by an elevated rate of proliferation of CD8⁺ T cells, approximately five- to eightfold, as determined by ²H-glucose⁸² or BrdU⁸³ incorporation into DNA or expression of the cell cycle antigen Ki-67.⁸⁴ CD8⁺ T cells also exhibit elevated expression of activation antigens, especially CD38 and HLA-DR.⁸⁵ Both activation and proliferation are decreased after initiation of potent antiretroviral therapy,⁸³’^86-89 demonstrating their dependence on viral replication. In particular, the level of expression of CD38 is highly correlated with plasma HIV RNA viral load^90,91 in both treated and untreated individuals and is closely aligned with the rate of decline of CD4⁺ T cells.⁹²

The observed discrepancy between the several-fold elevation in activation and proliferation, but only a net doubling in cell number, suggests that the steady state is balanced by increased apoptosis of CD8⁺ T lymphocytes. Furthermore, mathematical analysis of BrdU incorporation suggests that HIV-1 or SIV infection enlarges a pool of high-turnover CD8⁺ T cells that are both produced and disappear at an increased rate, compared with CD8⁺ T cells in uninfected controls.⁸⁹’⁹³ It is usually assumed that decay of labeled cells equates to cell death, but this was not found to be significantly elevated in one study,⁸⁹ and an alternative explanation may be trafficking of labeled cells¹⁴ to tissues that are not sampled.

The cause of the increased activation and proliferation of CD8⁺ T cells is not clear but may be critical to the later survival of the cells. Studies in mouse models of acute viral infection would suggest that a large proportion of the activated, proliferating CD8⁺ T cells are antigen specific.^12,13 Also, in the case of acute EBV infection, up to 50% of the CD8⁺ T cells have been shown to be antigen specific.⁹⁴ HIV-1 antigen-specific CD8⁺ T cells have an activated CD38⁺ phenotype^95,96 but numerically seem to be a small subset of the population of activated CD8⁺ T cells, even when comprehensive overlapping peptide sets for all HIV-1 proteins are used to stimulate peripheral blood mononuclear cells (PBMCs).⁹⁷’⁹⁸

It is possible that antigen-specific cells may be underestimated due to a number of factors: rapid divergence of autologous virus epitopes from consensus sequences^99,100; responses to cryptic, frame­shift epitopes¹⁰¹; or antigen-specific cells’ functions other than IFN-γ secretion.^102,103 It is plausible that the increasing heterogeneity of viral quasispecies during the asymptomatic stage of HIV-1 infection^104,105 continually drives constant recruitment of new antigen-specific CD8⁺ T cells.

Alternatively, nonspecific effects of cytokines may drive CD8 activation and proliferation. Previously, it was shown that IFN type I nonspecifically induced CD8⁺ T cell turnover in mice, and this was dependent on IL-15 production.⁶⁸ IFN type I production is typically elevated in viral infections, and there is a reported increase of mRNA for IFN type I target genes in lymph nodes during SIV infection,¹⁰⁶ consistent with recruitment of activated plasmacytoid dendritic cells⁷¹ to these sites.¹⁰⁷ There are two reports that the plasma level of IL-15 is increased in HIV-1 infec­tion,^108,109 but other reports have suggested that production of IL-15 may actually be decreased.^110,111 Consistent with the latter possibility, it has been found that the number of circulating, IFN type I-producing plasmacytoid dendritic cells is decreased in progressive HIV-1 infection.¹¹²

Overall, the increased activation and proliferation of CD8⁺ T cells during chronic HIV-1 infection may be a composite of antigen- and cytokine-driven processes.

Dynamics of CD8⁺ T Lymphocytes in Nonpathogenic Chronic HIV-1 and SIV Infection

Long-term nonprogressors¹¹³’¹¹⁴ and subjects infected with less pathogenic HIV-2¹¹⁵ exhibit reduced activation of CD8⁺ T cells. Simlarly, sooty mangabeys and chimpanzees, which are relatively unaffected by simian immunodeficiency virus (SIV) and HIV infection, respectively, have less activation of CD8⁺ T cells than rhesus macaques, which rapidly lose their CD4⁺ T cells when infected with SIV.^116-118 Importantly, in the case of sooty mangabeys, the lower levels of CD8 activation and stable CD4 cell counts are found despite relatively high plasma viral loads.^117,119 Altogether, these observations suggest that CD8 T cell activation may be an important part of the pathogenesis of HIV-1 infection.

Dynamics of CD8⁺ T Lymphocytes in Primary HIV-1 Infection

After exposure to HIV-1 and an incubation period of 2 to 3 weeks, the acute onset of symptoms is associated with a peak of viral load and commonly a brief lymphopenia at presentation,¹²⁰ involving both CD4⁺ and CD8⁺ T cells. The lymphopenia may result from shutdown of efferent lymph,¹²¹ because there is a rapid increase in the lymphocyte count in the days afterward.¹²⁰’¹²² The CD4⁺ T cell count exhibits a nadir coinciding with the peak of viremia and remains depressed in approximately 50% of subjects, but the CD8⁺ T cell count progressively increases to a peak approximately 30 days after the onset of symptoms.¹²² Thereafter, the CD8⁺ T cell count remains elevated, resulting in the typical long-term inversion of the CD4: CD8 ratio, although the CD8⁺ T cell count decreases just before the onset of AIDS.^123,81 In nonpathogenic primate models, despite intense viral replication during the acute phase, there are no major long-term changes in CD4⁺ or CD8⁺ T cell counts in blood or lymph nodes,^117,124 compared with pathogenic primate models.¹²⁵

Symptomatic acute EBV infection in adolescents and adults also results in a very large increase in the CD8⁺ T cell count. Unlike HIV-1 infection, the activation of CD8⁺ T cells returns to normal in the vast majority of patients for the next 16 weeks of follow-up.¹²⁶

In both of these acute infections, the increase in CD8⁺ T cells is due to the appearance of highly activated, proliferating memory effector phenotype cells.^11,127,128 However, in acute HIV-1 infection, only a small proportion of the increased CD8⁺ T cells seems to be antigen specific in most subjects.^129,130 Whether the activation and proliferation are antigen specific or cytokine driven is unclear, similar to the situation in chronic infection, as described above.

Dysfunction of CD8⁺ T Cells in HIV-1 Infection

The impairment of HIV-specific CD4⁺ T cell helper cells late in acute infection¹³¹ probably affects the state of differentiation of antigen-specific CD8⁺ T cells. The evidence suggests that CD8⁺ effector T cells control high levels of viremia during resolution of acute HIV¹³²’¹³³ and SIV¹³⁴ infections, but they are typically unable to completely prevent ongoing replication.

It has been suggested that HIV-specific CD8⁺ CTLs are functionally impaired in chronic HIV infection, with low expression of perforin and possible alterations in T cell antigen receptor (TCR) signaling⁶’⁷ as well as a state of differentiation of HIV-specific CD8⁺ T cells that is quite different to comparable cytomegalovirus (CMV)- and EBV-specific cells.¹³⁵’¹³⁶ The production of IL-2 by CD4⁺ T cells is characteristically reduced in untreated HIV-1 infection,¹³⁷ particularly in antigen­specific CD4⁺ T cells,¹³⁸’¹³⁹ and lack of IL-2 may be the cause of impaired CTL differentiation¹³⁵’¹³⁶ and perforin expression.¹⁴⁰

Consistent with this possibility, long-term nonprogressors maintain antigen-specific CD4⁺ T cell responses,¹⁴¹ and their CD8⁺ T cells proliferate and express perforin, in response to HIV antigen, unlike cells from progressing subjects.¹⁴² However, treatment of HIV-infected subjects with IL-2 does not lead to an increase in circulating CD8⁺ T lymphocytes or an increase in HIV antigen­specific responses.¹⁴³’¹⁴⁴

In the absence of CD4⁺ T cell help, short-term CD8⁺ effector cells may be constantly generated in HIV-1 infection, but these cells may be preprogrammed to be senescent.¹⁴⁵’¹⁴⁶ In fact, early HIV- 1 infection is characterized by an increased proportion of CD8⁺ T cells that are CD28-negative CD57⁺,¹⁴⁷ but such cells do not proliferate in vitro and are particularly prone to apoptosis.¹⁴⁸ Therefore, apoptosis as a result of incomplete differentiation during acute infection may be respon­sible for inadequate antiviral CD8⁺ T cell activity during HIV and SIV infections.

increased RATE OF CD8⁺ t cell apoptosis