IMMUNOPATHOGENESIS OF HUMAN IMMUNODEFICIENCY SYNDROME(HIV) INFECTION[final part]

Sequestration in Lymphoid Organs

Much of the immune damage that is seen in HIV infection probably results from viral replication and its consequences in lymphoid tissue. In early stages of HIV infection, generalized lymphadenopathy is commonly recognized.(154,155) In this condition, nodes are filled with lymphocytes, and the representation of CD4+ and CD8+ T lymphocytes in these sites is generally reflective of what is seen in circulation.(156) In untreated HIV infection, lymph nodes show inflammation with heightened expression of cytokines such as interferon-gamma, IL-1, IL-2, and IL-12.(157-160) The inflammatory condition of the lymphoid tissues likely is a consequence of high-level HIV replication at these sites. These inflammatory lymph nodes are also characterized by heightened expression of molecules such as intercellular adhesion molecules and vascular cell adhesion molecules.(161) This "sticky" and inflammatory state likely results in sequestration of circulating lymphocytes in these sites. As disease advances, there is progressive destruction of lymphoid architecture (162) and ultimately lymphoid tissues are, as is the circulation, depleted of lymphocytes.

Heightened Destruction

As noted above, the immune deficiency of HIV infection is characterized by immune activation, with an increased frequency of circulating lymphocytes that have been activated to enter the cell cycle. Interestingly, this heightened entry to the cell cycle is often aborted (at least after in vitro cultivation) as the activated cells tend to die by mechanisms of programmed cell death (109) and also, in some studies, by necrotic cell death.(163,164) This may be true especially for CD4+ T cells, in which studies of telomere length fail to show evidence of sustained successful division (165) despite evidence of heightened cellular proliferation and turnover in vivo (166,167) or ex vitro.(168,169) In contrast, though CD8+ T cells in HIV infection also die after activation,(170) CD8+ T-cell populations in HIV infection tend to have shortening of the average telomere length reflective of multiple rounds of successful cellular replication (at least among the surviving cells).(165) Importantly, for CD4+ T cells, the proportion of cells that are activated, the proportion of cells that incorporate label into DNA, and the proportion of cells that can be demonstrated to die via apoptosis ex vivo far exceed the proportion of cells that are demonstrably infected by HIV; the same can be said for CD8+ T cells, which are rarely infected in vivo. Although antigenic stimulation in response to peptide antigens of HIV itself may account for some portion of the observed activation, the proportion of CD8 cells activated to express activation markers (either CD38 or HLA-DR) far exceeds the proportion of cells recognized to be HIV reactive;(107) the same appears to be true for CD4+ T cells (171) although screening for CD4 cell reactivity has been less comprehensive than has screening for CD8 cell reactivity. Therefore, cellular activation and cell death in HIV infection appear to be entirely determined neither by direct cytopathic effects of the virus nor by immune activation driven by specific peptide recognition. Alternative explanations, such as dysregulated activation of T cells though mechanisms other than T-cell receptor activation, are yet unproven.

Diminished Production

Whereas HIV infection clearly is characterized by heightened cellular destruction and turnover, there is also evidence that immune cellular production may be impaired, at least at certain stages of infection. With advancing stages of HIV infection, there is evidence of cellular hypoproductivity in bone marrow. Pancytopenia is not uncommon in advanced AIDS and bone marrow biopsies often reveal evidence of hypoplasia. Moreover, CD34+ hematopoietic progenitor cells in bone marrow appear susceptible to infection with HIV (172) and impairment of the function of these cells has been described.(173,174) The mechanisms whereby bone marrow productivity is impaired in HIV disease are incompletely understood and it is likely that concurrent infection with opportunistic pathogens such as cytomegalovirus and Mycobacterium avium complex may contribute to this suppression in some persons. Nonetheless, with administration of suppressive antiretroviral therapies, peripheral blood cytopenias characteristically improve.(175)

T lymphocytes undergo maturation and rearrangement of T-cell receptor genes in the thymus, where T cells with receptors of very high avidity for host HLAs that bind endogenous peptides are deleted (this prevents too much autoimmune reactivity) as are T cells with very low avidity for the host HLAs (this assures that remaining T cells are potentially capable of recognizing host HLAs that bind foreign peptides). The population of antigen-naive T lymphocytes that emerges contains a diverse distribution of T-cell receptors with specificities capable of recognizing a broad array of peptide antigens bound to the host's own cell-surface HLAs. Although thymic activity is greatest during development and childhood, there is evidence of thymic function in adulthood as well.(176,177) The role of the thymus in HIV disease is complex. Thymus size is often preserved in HIV-infected adults (particularly in older persons) (178,179) and there is indication that thymic output often is maintained in infected persons.(180) Moreover, preliminary data indicate that in some persons with HIV infection, thymic size is actually diminished after suppression of HIV replication,(181) suggesting that in some HIV-infected persons, thymic function (or thymic size at least) increases during uncontrolled HIV replication, perhaps in order to keep up with the increased demands of HIV-induced cellular turnover, and that this demand falls as HIV replication and immune cell destruction are diminished by antiretroviral therapies. However, there is also evidence that HIV can infect thymic stromal cells and that HIV strains capable of using the CXCR4 coreceptor (X4 strains) can infect thymocytes as well.(182) Moreover, there is clear evidence of thymic failure among persons who fail to increase circulating CD4 cell numbers with antiretroviral therapy-induced suppression of HIV replication.(183)

IL-7 is a cytokine that may be important in thymopoiesis and in promoting naive T-cell expansion.(184-186) Circulating IL-7 levels are often elevated in HIV infection, particularly as CD4 cell counts fall below 100 cells/µL,(187,188) suggesting that increased levels of this cytokine may play an important role in driving T-cell production and homeostasis. Studies of IL-7 administration in SIV-infected macaques and in humans are ongoing and may help to elucidate a possible role for this agent in the treatment of HIV-associated immune deficiency.

Predictors of Immune Deterioration in HIV Infection

The rate of disease progression in untreated HIV infection is highly variable, with some individuals progressing rapidly to experience opportunistic infection and death within months of acquisition of infection and others (ie, long-term nonprogressors) remaining entirely well and maintaining normal CD4 cell counts more than 15 years after infection in the absence of antiretroviral treatment. (Approximately half of persons who acquire HIV infection will develop severe disease--AIDS- -within 10 years if not treated with antiretroviral therapies.) Although long-term nonprogressors represent no more than about 5% of HIV-infected individuals, this variability suggests a need to identify the factors that determine rates of disease progression.

From a clinical perspective, rates of disease progression can be quantified by measuring decreases in circulating CD4 cell numbers over time. This index is highly variable among infected persons. A number of factors that predict the risk of disease progression, measured as rate of CD4 cell decline, progression to opportunistic infection or death, or risk of progression to CD4 cell counts of <200/µL, have been identified. Both viral factors and host factors, and likely their interaction, may predict the risk of HIV disease progression.

The magnitude of HIV replication as reflected in plasma HIV RNA levels is one predictor of the risk for HIV disease progression.(189) The relationship between the extent of HIV replication and disease progression, however, is complex and cannot be conceptualized in terms of a simple linear correlation between plasma HIV RNA level and rate of disease progression across all groups of HIV-infected individuals. For instance, disease progression is seen at significantly lower HIV RNA levels in women than in men.(190) In a limited number of cases, viral heterogeneities may explain differences in rates of disease progression. For example, in a small cohort of individuals who were infected by blood transfusion from a single donor in Australia and subsequently experienced a milder disease course than was expected, the infecting viral isolate was found to have a truncated Nef protein.(191,192) Other studies of HIV-infected persons with divergent disease progression rates have failed to identify plausible sequence or functional differences in the long-term repeat (LTR) and Tat sequences, (193,194) but do not rule out the possibility that heterogeneities in these viral sequences may have an impact on disease course.

Switches in envelope sequences resulting in a phenotype that utilizes the CXCR4 coreceptor are associated with evidence of accelerated HIV disease progression,(195) but the details of how coreceptor use determines disease outcome remain to be established.

The complex interplay between viral fitness and disease progression is potentially significant. Since replicative fitness correlates with plasma HIV RNA levels, more "fit" viruses might be expected to produce faster CD4 cell declines, as has been found to be the case in the context of antiviral drug resistance mutations. In the presence of antiviral drug selection pressure, resistance mutations either to nucleoside reverse transcriptase inhibitors (196,197) or to protease inhibitors (198,199) tend to attenuate the CD4 cell decline induced by wild-type virus. Although this effect may be due in part to diminished replicative capacity of these viruses, there is also reason to believe that decreased plasma HIV RNA levels do not completely explain the effect of fitness on CD4 cell levels and that the intrinsically diminished ability of these viruses to cause immunopathology also may play a role.(200,201)

Host genetic factors further determine the magnitude of HIV replication. For example, persons who are heterozygous for the delta-32 base pair deletion in the CCR5 open reading frame have decreased expression of cell-surface CCR5, lower HIV RNA levels, and slower disease progression.(202) Similarly, the G polymorphism in the -2459 sequence of the CCR5 promoter has been associated with decreased plasma HIV RNA levels and a modest decrease in risk of disease progression.(203) Peripheral blood and Langerhans cells obtained from persons with this genotype show diminished levels of HIV replication in vitro.(15,204) The -28G polymorphism in the promoter of the gene for the CCR5 ligand RANTES supports increased expression of RANTES and is associated with a decreased rate of CD4 decline in infected persons.(205) Another chemokine receptor gene polymorphism, the CCR2 64I allele, is also associated with a decreased risk of HIV disease progression.(203) However, as this receptor is not thought to be important in HIV infection, the mechanism for this effect (possible linkage to another genetic polymorphism) remains to be determined. Persons homozygous for a noncoding sequence in the gene for stromal cell-derived factor 1, the natural ligand for CXCR4, were found to have a delayed risk for progression to AIDS (206) but this observation has not been confirmed in other cohorts.(203)

Factors associated with adaptive immune responses to HIV are also indicators of disease progression risk. For example, certain HLA alleles indicate greater or lesser risks of disease progression.(207,208) In addition, viral mutations that impair binding of specific viral peptides to the HLA of a given patient predict higher levels of HIV replication.(209) Thus, the emergence of viral mutations at loci involved in HLA binding permits escape from protective immune defenses. In addition, homozygosity for HLA alleles is associated with a significantly greater risk of disease progression.(210) Because homozygosity (having fewer distinct HLA molecules) effectively decreases the diversity of peptide-HLA combinations available for recognition, potentially protective immune responses to HIV are limited. Finally, even a single amino acid substitution in an HLA molecule that determines which peptides can be bound and presented by this HLA type can determine a differential risk of HIV disease progression.(211) Thus, some degree of individual variation in HIV replication and ultimately in risk for disease progression is determined by HLA type and diversity.

Less is known about the innate immune responses that may limit HIV replication. Innate defenses are responsible for the most rapid responses to microbial invasion; they help to control microbial replication and to activate the more specific adaptive immune responses. A preliminary report suggests that the killer immunoglobulin-like receptor (KIR) allele KIR 3DS1, which activates natural killer cells, in the presence of the HLA-B BW4-80ILE allele, protects against disease progression in HIV infection,(212) suggesting that innate immune function is also important in the control of HIV disease. Additional studies indicate that preservation of the function and numbers of plasmacytoid dendritic cells--the major sources of interferon-alfa--is associated with protection of persons with advanced HIV disease from the occurrence of opportunistic infection.(88,89)

Nonspecific immune activation is a clear consequence of HIV infection and is correlated with increased HIV replication. For example, plasma concentrations of beta-2 microglobulin, TNF and its receptors, neopterin, and the soluble IL-2 receptor CD25 each correlate with magnitude of HIV replication and risk of disease progression.(213) Similarly, the expression of the activation marker CD38 on CD8 cells is correlated with HIV levels in plasma,(105) but in some studies independently adds predictive value for disease progression,(214,215) as does the level of CD38 expression on CD4+ T cells.(216) Moreover, levels of CD38 expression seem to exhibit an inverse linear correlation with the extent of CD4+ T-cell restoration in response to antiretroviral therapy.(217) Data from a cross-sectional study support a model wherein HIV replication drives immune activation that drives CD4 cell losses.(218)

Finally, age at the time of HIV infection also appears to be correlated with risk of disease progression. Data from large cooperative cohort studies indicate that the correlation of age with risk of disease progression and HIV-related mortality persists after adjusting for CD4 cell counts and plasma HIV RNA level.(219) Recent data comparing HIV-infected subjects to age-matched healthy controls suggest that the effect of age on the clinical course of HIV infection may be related to depletion of naive T cells, diminished CD28 expression, and reduced thymic volumes in older individuals.(179)

A Model for the Immunopathogenesis of HIV-Induced Immune Deficiency

As this chapter has shown, the interactions between HIV and the host response are complex and only partially characterized. Nonetheless, the existing data permit us to propose a model for the pathogenesis of immune deficiency in HIV infection. In this model, lymphoid sites of HIV replication serve as the major sites of immunopathology. At these inflammatory sites of viral replication, heightened adhesion molecule expression results in increased trapping of circulating lymphocytes, often reflected clinically as generalized lymphadenopathy. Trapped lymphocytes at these sites are exposed to a number of signals. Some of these signals may be driven by toxic viral products such as free envelope glycoprotein and viral regulatory proteins. Others may be mediated by cytokines induced at these sites during the intensive and sustained exposure to viral antigens or viral replication. Still other signals are driven by T-cell receptor engagement though binding of antigenic peptides. As a result of these events, both CD4 and CD8 T cells are activated in a dysregulated fashion, and activation may be both antigen-driven and antigen-independent (ie, not driven through T-cell receptor engagement). In this model, both antigen- and cytokine- or viral product-mediated T-cell activation induces cells to enter the cell cycle. The outcome of these events is not only heightened immune activation but also heightened cell death, because even physiologic cellular activation generally is accompanied by increased T-cell death.

It is proposed that both physiologic and dysregulated activation contribute to the profound immune activation and accelerated cell death that characterizes HIV infection. In early stages of infection, antigen-driven CD8+ T-cell expansion predominates, possibly because activated CD8+ T-cells are less susceptible than CD4+ T cells to productive and cytopathic HIV infection at lymphoid sites. Nonetheless, sustained exposure to this "toxic" milieu results in surviving CD4 and CD8 T-cell populations that are functionally impaired in their ability to mediate effector functions such as cytolysis, and are also impaired in the ability to expand in response to T-cell receptor engagement. The result is a less effective and less adaptable immune response to opportunistic pathogens. With advanced disease, bone marrow, the thymus, or both may fail to keep up with the heightened demand for cellular production, and severe immunodeficiency ensues. This model suggests that sustained exposure to viral replication results in immunologic impairments that are not readily reversible and that immune suppressive strategies may provide adjunctive value to antiviral therapies, and preliminary data suggest that this may indeed be the case.(54,122,220) Consistent with this, clinical responses to treatment with antiviral therapies are often less effective when treatment is begun at more advanced stages of disease.(221)

IMMUNOPATHOGENESIS OF HUMAN IMMUNODEFICIENCY SYNDROME(HIV) INFECTION[contd]

Dendritic Cells

Characterized by large, dendritic cytoplasmic extensions, dendritic cells normally engage in efficient presentation of antigens to T and B lymphocytes in lymph nodes. Epidermal dendritic cells (Langerhans cells), characterized by expression of CD1a and Birbeck granules, may be among the first cells to encounter HIV at mucosal surfaces and, in the course of transporting antigens encountered at epidermal sites, have the capability of transporting HIV to lymphoid tissue. Two populations of dendritic cells can be identified in blood: myeloid dendritic cells (characterized by expression of CD11c) and plasmacytoid dendritic cells (CD123+). Circulating numbers of these cells tend to be diminished in HIV infection.(83-86) Recent studies suggest that there are decreases in both number and function of circulating dendritic cells in HIV-infected persons and that the decreases are not normalized with suppression of HIV replication.(87) The ability of these cells to mature remains incompletely determined. Studies of these cells are hampered by their relative scarcity, as they generally account for <1% of circulating mononuclear cells. The plasmacytoid dendritic cell is a major producer of interferon-alfa and a small but growing body of literature suggests that preservation of these cells in advanced HIV infection is associated with fewer severe opportunistic complications.(88,89)

The follicular dendritic cell found in lymphoid tissue is also a key antigen-presenting cell that traps and maintains intact antigens on its cell surface. In untreated HIV infection, the surface of this cell is often loaded with virus and viral antigen. In the lymph node follicles, this cell provides key signals for the activation of B lymphocytes. The follicular dendritic cell is related to the other dendritic cells discussed above only in terms of its dendritic morphology; the origin of this cell is not well understood.

Natural Killer Cells

Natural killer cells are large granular lymphocytes with cytolytic capabilities. Lytic activity is greatest against tumor cells and virus-infected cells that have diminished expression of major histocompatibility complex (MHC) class I antigens. Because MHC class I expression is required for peptide presentation to T-cell receptors, natural killer cells comprise a cellular component of the innate host defense system with activity against cells that may escape adaptive host defenses because of failure of MHC class I expression. Lysis by natural killer cells also can be directed against cells recognized by host antibodies through binding of immunoglobulin to fragment constant receptors on the natural killer cell. Thus, natural killer cells contribute to both innate and adaptive immune host defenses. Early studies have demonstrated impairments in natural killer cell activity in persons with AIDS and HIV infection,(90,91) and functional impairments of these cells have been attributed to a failure of the postbinding lytic event.(92) Some of this impairment is correctible in vitro after overnight cultivation in medium or after addition of the helper-cell-derived cytokines IL-2 or interferon-gamma,(90,93) suggesting that "exhaustion" and/or failure of CD4 cell help may underlie this defect.

Gamma-Delta T Cells

These infrequent cells (comprising between 1% and 5% of the T-lymphocyte pool) are cells that may, as do natural killer cells, play roles in both innate and adaptive immune responses. The antigen-binding sites of T-cell receptors of these lymphocytes are comprised of gamma and delta heterodimers as contrasted with the alpha and beta chains of most T lymphocytes. These T cells can recognize microbial antigens directly without processing and presentation on host human leukocyte antigen (HLA) molecules. Although the genes encoding these receptor chains also undergo rearrangement, the diversity of these receptors is more restricted than that of T cells with alpha-beta chain receptors. The cytokinetic and cytolytic functions of cytokinetic are often perturbed, whereas proliferation responses are variably affected in HIV infection.(94-96)

Immune Activation and HIV Infection

It has long been recognized that infection with HIV is characterized not only by development of profound immunodeficiency but also by sustained and dramatic immune activation.(97,98) In fact, a growing body of evidence establishes immune activation as a critical underlying mediator of immune dysfunction and immune deficiency.(99-101) This state of immune activation is manifested both by enhanced expression of phenotypic activation markers on peripheral blood T cells and B cells and by increased plasma levels of inflammatory cytokines; moreover, lymphocytes obtained from HIV-infected persons are more often found in activated phases of the cell cycle.

In HIV-infected subjects, T lymphocytes often express surface markers of immune activation such as HLA class II molecules and CD38, a membrane-bound adenosine 5'-diphosphate (ADP) ribosyl cyclase.(102,103) These markers of activation are elevated in direct proportion to the magnitude of HIV replication (104,105) and some studies have found that the extent of CD38 expression predicts ultimate HIV disease course more accurately than do plasma levels of HIV itself.(104,106) Although a simple explanation for this state of persistent immune activation would be a reflection of HIV-specific T-cell expansion, the frequency of phenotypically activated CD8+ T cells found in HIV-infected subjects is often greater than 80%,(101) substantially exceeding the proportion of cells that can be shown to recognize HIV peptides.(107)

Thus, a significant proportion of this activation may represent a response to other antigens, or may be an indirect (or bystander) effect of HIV replication. In HIV infection, T cells are also more extensively primed to enter the replication phases of the cell cycle. This propensity is evidenced by an increased frequency both of cells expressing the nuclear antigen Ki67 (101,108) and of cells exhibiting increased DNA content and 5-bromo-2'-deoxyuridine (BrdU) incorporation, a reflection of spontaneous progression to the synthesis phase of the cell cycle.(109,110)

Plasma levels of TNF-alpha, IL-1, and IL-6 are often elevated in later stages of HIV infection, and both TNF and IL-6 levels also are directly correlated with plasma HIV RNA levels.(111) Interestingly, in lymphoid tissue, the primary site of HIV replication, levels of TNF-alpha are not generally increased, although expression of IL-1, IL-2, IL-6, IL-12, and interferon-gamma may be elevated.(112)

With administration of antiretroviral therapies, these indices of immune activation tend to fall, indicating that HIV replication induces the state of high-level activation.(113-117) The plausible hypothesis that HIV-induced activation enhances the magnitude of HIV replication by increasing the numbers of cells susceptible to and supportive of productive viral replication remains unproven. In this regard, it should be noted that expression of CD38 may limit the susceptibility of cells to productive HIV replication.(118)

Animal studies support the relationship between immune activation and progressive cellular immune deficiency. A natural host of SIV, the sooty mangabey permits high-level SIV replication but manifests limited evidence of disease.(119) Strikingly, this lack of pathogenicity is accompanied by absence of the extensive immune activation and cellular proliferation that characterizes SIV infection of other primates such as the rhesus macaque, in which immune activation closely mimics the activation seen in HIV-infected humans.(120) Moreover, mangabeys seem to maintain thymic and bone marrow function and do not demonstrate so-called bystander lymphocyte apoptosis,(121) whereby uninfected cells in the vicinity of an infected cell are induced to undergo programmed cell death. Finally, in a preliminary human study, blocking immune activation by administration of the immune suppressant cyclosporine A concomitantly with initiation of combination antiretroviral therapies resulted in more sustained CD4+ T-cell restoration than had been seen with antiviral therapies alone.(122)

Immune Response to HIV

As noted above, infection with HIV is associated with a brisk immune response to HIV antigens. Although antibody levels are high, neutralizing antibody responses against HIV are not strong, and are followed in rapid sequence by the emergence of viruses resistant to the neutralizing activity of these antibodies.(25) Thus, although these antibodies possess sufficient activity to exert selection pressure, the target epitopes are in regions that can readily sustain mutational escape (123) or can be shielded by mutations altering the numerous glycosylation sites on the viral envelope.(25)

Following initial infection with HIV, the rapid emergence of cytolytic T-cell responses, largely CD8+ T-cell responses, is associated temporally with a decrease in plasma levels of HIV.(21) CD8+ T cells may help control HIV replication in several ways. First, binding of these cells to viral peptides presented by HLAs on the surface of infected cells can trigger a cytolytic response resulting in the destruction of the target cell that is producing virus. This function is largely mediated through the liberation of perforin, which generates a hole in the target cell through which granzymes can enter and destroy the cell before it can produce large numbers of progeny virions. Although most cytolytic activity against viral targets is mediated through this route, CD8+ T cells expressing Fas ligand also can bind to Fas (CD95) on the surface of target cells, thereby inducing apoptotic cell death. Finally, CD8+ T cells can liberate a number of soluble factors with antiviral activity. These include interferon-gamma, which can, via a complex cascade of receptor-mediated binding and activation, render nearby cells relatively resistant to productive viral infection. CD8+ T cells also are sources of the beta chemokines MIP-1a (macrophage inflammatory protein-1 alpha), MIP-1b (macrophage inflammatory protein-1 beta), and RANTES (regulated on activation, normal T expressed and secreted), which bind to CCR5 and, by promoting internalization of this critical HIV coreceptor, decrease the ability of HIV to gain entry into otherwise susceptible cells.(124) A number of other antiviral factors also can be expressed by CD8+ T cells and these may include an incompletely described "cell antiviral factor" (CAF) (125) that blocks HIV replication largely via inhibition of viral transcriptional activation.(126) The precise definition of CAF is not available although CAF activity does appear to be distinguishable from chemokine-mediated (127,128) or defensin-mediated (129) suppression of HIV. CD8-mediated suppression of HIV may be related to disease outcome since this cellular response is augmented by supernatants prepared from cells from long-term nonprogressors (persons with stable CD4 cell counts and sustained low levels of HIV replication in the absence of antiretroviral treatment).(130-133) It is difficult to determine whether these factors and activities actually cause a better disease outcome or whether they are merely reflections of better preservation of host defenses.

Whereas it is clear from both human and animal models that CD8+ T cells are important in control of retroviral replication, it is not entirely clear what, if any, kind of CD8+ T-cell response can confer sustained control of HIV replication. Nor is it clear whether or not the magnitude or the breadth of CD8+ T-cell target recognition reliably predicts disease course.(134,135)

Despite relatively high frequencies of HIV-specific CD8+ T cells in HIV-infected individuals,(134,136-138) sustained suppression of viral replication is rarely achieved. The emergence of viral escape mutations that render virus-infected cells undetectable by host cytotoxic T-lymphocyte assay may help to explain this observation.(139,140) Moreover, there is some evidence that HIV-specific CD8+ T cells may be dysfunctional, as indicated by reduced lytic activity,(141,142) poor proliferation function in vitro,(58) and decreased expression of key signaling molecules (143) that mediate TCR activation. Whether this is a consequence of sustained exposure to high levels of viral antigen or is related to the lack of CD4 help or direct exposure to toxic viral products and the effects of chronic inflammation remains to be determined.

Importantly, CD4+ T-cell responses to HIV antigens are dysregulated in HIV infection, to an extent that exceeds the impairment of responses to other microbial antigens.(103,144,145) Although interferon-gamma expression is readily induced in response to HIV antigens even in advanced disease,(146) CD4 T-cell proliferation is rarely detected in untreated infection except among long-term nonprogressors.(147) CD4+ T-cell proliferation responses to HIV antigens sometimes can be preserved or restored in HIV-infected persons who are treated shortly after acquisition of infection (148) and in a proportion of chronically infected persons in whom viral replication is suppressed by antiretroviral therapies.(149,150) Importantly, restoration of these responses is less common in persons who begin suppressive antiretroviral therapy with moderately advanced or advanced infection.(103,151)

Thus, CD4+ T-cell responses to HIV antigens appear to be selectively impaired during high-level viremia and may be restored when HIV replication is brought under control by therapy.(149,152) Although HIV-reactive CD4+ T cells are preferentially susceptible to HIV infection,(153) it is not likely that this phenomenon is sufficient to explain the impaired proliferation responses seen during uncontrolled HIV replication. For example, the persistence of HIV-specific interferon-gamma responses even in persons with advanced disease (146) suggests that the ability of HIV-specific CD4 cells to expand may be selectively impaired while other HIV-specific immune functions (such as interferon production) may be preserved. Conceivably, HIV-reactive cells potentially capable of proliferation are selectively targeted and destroyed, or their replication capacity is impaired in the setting of viral activity. Sustained and relatively selective infection of HIV-reactive CD4+ cells may underlie the failure to restore CD4+ T-cell proliferation responses when persons with advanced disease initiate treatment with antiretroviral therapy.

Pathogenesis of Immune Deficiency in HIV Infection

The characteristic depletion of CD4+ T lymphocytes in HIV disease appears to result from factors other than the direct cytopathic effect of HIV itself. Cellular destruction, diminished cellular production, and cellular sequestration all appear to contribute to decreases in numbers of circulating CD4+ T cells.

IMMUNOPATHOGENESIS OF HUMAN IMMUNODEFICIENCY SYNDROME(HIV) INFECTION

HIV and the Intricate Relationship between Viral Pathogenesis and Immune Defenses

As intracellular parasites, all viruses must be intimately familiar with host cellular machinery and capable of suborning it to support their replication cycle. For HIV, this relationship is particularly complex and intimate because HIV targets, infects, and incapacitates cells central to antimicrobial defenses. Thus, host immune defenses and HIV pathogenesis are inextricably linked. Whereas this parasitic relationship may contribute to the persistence and progression of HIV infection, careful study of the relationship between HIV and the immune system has also yielded important insights into mechanisms of immune homeostasis and host defenses in general. This chapter will examine briefly the proposed mechanisms whereby HIV infects host immune cells, the mechanisms whereby host defenses are mobilized to attenuate HIV replication, the strategies HIV uses to evade host immune responses, and finally, the mechanisms whereby HIV induces immune deficiency that places persons at risk for the opportunistic infections and malignancies that define AIDS.

Acquisition of HIV Infection

In most infectious diseases, a number of factors contribute to the risk of acquisition of infection and to the occurrence of illness after exposure to a pathogenic organism. These include the nature of the exposure (eg, the route, the size of the microbial inoculum), the "virulence" of the microbe, and the nature of the host susceptibility to infection. The nature of the exposure clearly determines the risks of infection. Parenteral exposure to blood infected with HIV carries a substantial risk of infection. Among individuals transfused with blood of HIV-infected persons before screening of blood donors was practiced, the risk of infection approached 100%.(1) Transmucosal infection risks vary according to the site of exposure, with risks of transmission through rectal exposure exceeding the risks of transmission through vaginal exposure and both of the above exceeding the risks of transmission across oral mucosa. Mucosal inflammatory disease tends to enhance the risk of transmission particularly if associated with ulceration. Though firm data are lacking, epidemiologic evidence of seroconversion after accidental needlestick injuries (2) or sexual contact with infected persons with different levels of plasma HIV RNA suggest that the magnitude of the inoculum also contributes to the risk of infection.(3,4) Similarly, mother-to-infant transmission of HIV is enhanced among women with high levels of plasma HIV RNA, even after taking into account other known predictors of transmission,(5-7) and the intensity of exposure to contaminated antihemophilic factor concentrates has been shown to predict the risk of HIV infection among hemophiliacs.(8) Insight gained from persons at high risk for infection yet who persistently remain seronegative indicates that certain genetic loci can dramatically affect risk for acquisition of HIV infection. Specifically, persons homozygous for a 32-base-pair deletion (the so-called delta-32 mutation) in the C-C motif chemokine receptor 5 (CCR5) open reading frame that results in failure of surface expression of this key viral coreceptor are protected from acquisition of HIV infection.(9-11) In the rare instances when such persons have been found to be infected, they appear to acquire infection with viruses that may be capable of entry using the CXC motif chemokine receptor 4 (CXCR4) coreceptor.(12-14) In addition, persons with the -2459G polymorphism in the CCR5 promoter that may result in diminished CCR5 expression also may have a somewhat lower risk of infection than do persons with the alternative -2459A nucleotide at this site.(15) One other rare polymorphism in the CCR5 gene, characterized by a point mutation at position 303, introduces a premature stop codon in the elongating product chain and prevents the expression of a functional CCR5 coreceptor when associated with the delta-32 deletion, also conferring virtually complete resistance to CCR5-using viruses.(16)

As these studies have been performed among groups at risk for infection by both parenteral and mucosal routes, these observations suggest that acquisition of infection is highly dependent upon expression of the HIV coreceptor CCR5. The location of this critical "bottleneck" that requires CCR5 expression remains to be determined. One model proposes that CCR5-receptor availability is critical at the level of the mucosal dendritic (Langerhans) cells, which express CCR5 but much less CXCR4 or other C-type lectin receptors to which HIV may bind to facilitate cellular entry. On the other hand, a nonmucosal location for this bottleneck is suggested by the high prevalence of the CCR5 delta-32 homozygous state among seronegative hemophiliacs who otherwise appear to be at high risk for parenteral acquisition of infection.(17) Importantly however, among HIV seronegative cohorts at very high risk of either parenteral or transmucosal infection, only a minority (eg, 16% in a group of hemophiliacs at >95% risk of infection according to treatment history) are homozygous for the delta-32 mutation,(17) indicating that other mechanisms determine risks for and protection from HIV infection in these settings. Of note, members of several high-risk, HIV-seronegative cohorts have demonstrated immunologic "memory" of HIV exposure. Specifically, mucosal immunoglobulin A (IgA) capable of cross-clade HIV binding and neutralization has been found in genital secretions of some high-risk uninfected persons,(18) and low levels of CD8+ T cells reactive to HIV peptides have been found in circulation in other groups of high-risk seronegative individuals.(19,20) It is not yet clear whether these immune defenses are actually responsible for protection against infection or, alternatively, are a reflection only of exposure while protection is mediated by some other mechanisms that remain to be defined.

Acute Infection

Acute infection with HIV is often associated with a febrile illness and clinical evidence of systemic dissemination of virus to lymphoid tissue, the central nervous system, and other sites. High-level viral replication is reflected in high concentrations of virus in plasma and in lymphoid tissue. Viral replication characteristically peaks and then falls concurrently with the appearance in circulation of virus-specific CD8+ cytotoxic T cells.(21,22) As is the case in numerous other viral infections, these cytotoxic T lymphocytes are able to lyse infected host cells and likely attenuate the magnitude of HIV replication. Although animal models have established the importance of CD8+ cells in control of replication with the related simian immunodeficiency virus (SIV),(23,24) it has been very difficult to establish with certainty the nature of the CD8+ T-cell response that determines optimal control of HIV replication (see "Immune Response to HIV" below). Within several months after acquisition of infection, and in the absence of antiviral therapy, a "steady-state" level of HIV replication is established. This level tends to remain relatively stable for many years in a given individual but can vary enormously from person to person. A number of factors may determine steady-state HIV replication levels and these likely include the nature of host adaptive immune defenses, heterogeneities in viral replicative capacity, and heterogeneities in intrinsic host factors that may affect the magnitude of viral propagation.

Antibodies reactive with HIV antigens appear in circulation within a few weeks of infection but generally are first detectable after viral levels have begun to fall to the steady-state level. Although these antibodies often have strong neutralizing activity against the infecting virus, rapid viral escape from neutralization is characteristic, reflecting the enormous adaptability of the viral envelope, including its ability to revise its glycosylation sites, resulting in altered 3-dimensional configuration sufficient to escape antibody-mediated neutralization.(25)

Viral Reservoirs

Whereas most HIV replication is thought to take place in activated CD4+ T lymphocytes in lymphoid tissue, other cell populations may become infected and may play important roles in the persistence of HIV infection. Resting T cells constitute a significant reservoir of latent HIV that may be activated to complete the replication cycle upon activation of the host cell. At one end of the spectrum, in the activated T cell, multiple cellular factors and the viral Tat protein upregulate HIV transcription, resulting in viral production and ultimately destruction of the host cell.(26) At the other end of the spectrum, fully quiescent T cells, ie, those in the G0 phase of the cell cycle, are incapable of sustaining productive HIV replication, due to blocks in reverse transcription (27) as well as inability to enter the nucleus of the resting cell. Recent evidence indicates that, between those extremes, quiescent cells can be induced by exposure to certain cytokines to move far enough along the cell cycle (ie, to the G1 phase) to remove barriers to reverse transcription. Such cells are therefore susceptible to infection by HIV, but do not undergo full activation and cell cycling.(28) It is thought that these cells subsequently return to the fully quiescent state, in which they are protected from the cytopathic effects of massive viral replication. Infection of quiescent cells thus may establish a repository of infected cells capable of maintaining HIV for many years. How this takes place is not entirely clear but recent studies implicate the role of the HIV Nef protein as inducing a chain of events that renders resting CD4+ T cells susceptible to HIV infection.(29)

It has been shown in vivo that HIV can infect T cells that are not fully activated. During the course of HIV infection, integrated and infection-competent provirus can be found in a population of resting memory CD4+ T cells (30,31) and the frequency of these cells tends to remain stable for years, decreasing only minimally with the administration of combination antiretroviral therapies.(32,33)

Other proposed potential reservoirs of infection include sites within the genitourinary tract (34,35) and certain populations of monocytes and tissue macrophages,(36-38) particularly those in the central nervous system (39) and possibly the kidney.(40) In the resting memory cell compartment, sequence analysis has provided evidence of some replenishment of this reservoir over time. The relative stability and long half-life of these cells indicates that current treatment strategies likely will not be capable of eradication of infection in this compartment.(41) Neither intensive and prolonged administration of antiretroviral therapies (33,41) nor the design of strategies to activate expression of virus from these reservoirs by activating T cells through the T-cell receptor (TCR) or with IL-2 has been able to eradicate virus in infected persons, although the frequency with which virus can be found in resting memory cells has been modestly diminished by these therapies.(42,43) Moreover, mathematical modeling and limited experimental evidence suggest that pharmacologically induced, high-level T-cell stimulation not only is unlikely to eliminate the latent reservoir, but also could potentially lead to T-cell depletion and disrupt CD4/CD8 T-cell homeostasis.(44) Several excellent reviews of the role of cellular reservoirs in the pathogenesis of HIV infection have been published recently.(45-48)

Immune Deficiency and Chronic Infection

Although the precise mechanisms of immune dysfunction remain incompletely understood, virtually every arm of the immune response may be affected by HIV infection.

CD4+ T Cells

Progressive depletion in numbers of circulating CD4+ T cells occurs in almost all cases of untreated HIV infection. The number of circulating CD4+ T cells is widely used as a measure of global "immune competence" and provides a predictor of the immediate risk for opportunistic illnesses.(49) Earlier in the course of infection, many HIV-infected persons have a syndrome of generalized lymphadenopathy characterized by accumulation of lymphocytes within inflammatory lymph nodes and upregulation of adhesion molecule expression. Early in the course of infection, memory CD4+ T cells are selectively depleted from circulation; as disease advances, CD4+ T cells of both the naive and memory phenotype are lost from circulation.(50) In advanced disease, all CD4 cell populations are depleted from circulation and from lymphoid tissue sites.

Functional abnormalities of CD4+ T cells are also characteristic of progressive HIV infection. Failure of CD4+ lymphocytes to undergo cell division, for example, has been demonstrated following stimulation of T cells from infected individuals with antigens or mitogens in vitro. A sequential loss of immune responsiveness to recall antigens, followed by alloantigens and then mitogens has been described.(51) Diminished expression of IL-2 is readily demonstrable (51,52) in cells from HIV-infected individuals and may be related to the proliferation defects. In contrast, expression of interferon-gamma by these cells is often unimpaired,(52) suggesting that the defective responsiveness is not a consequence of depletion of antigen-reactive cells but rather a selective impairment in the ability of these cells to respond after engagement of TCRs. The function of CD4+ T cells that specifically recognize antigens from HIV itself appears to be selectively impaired early in the course of HIV infection (see "Immune Response to HIV" below).

Using anti-TCR antibody stimulation to characterize proliferation defects in CD4+ T cells indicates that proliferation defects in HIV disease are associated with early G1-phase cell-cycle arrest (53) and are more commonly observed in persons who have experienced sustained CD4 cell losses.(54) As a key role of CD4+ T cells is to facilitate immune responses though production of immunomodulatory cytokines, the loss of these cells and the failure of remaining cells to function properly constitutes a critical impairment in immune capability. Specific CD4+ T-cell responses to HIV antigens appear to be selectively and lastingly impaired during early HIV infection (see "Immune Response to HIV" below).

CD8+ T Cells

In early HIV infection, CD8+ T-cell numbers tend to increase, reflecting expansion of memory CD8+ T cells, particularly HIV-reactive cells. CD8 cell expansions persist until far advanced stages of HIV disease, when all T-cell numbers tend to fall.(55) In contrast to memory CD8 cell expansions, proportions of naive CD8 cells tend to fall in early infection, but absolute numbers of these cells do not fall until HIV disease progresses.(50) For example, in earlier disease CD8+ T cells that recognize cytomegalovirus are present in large numbers, but in advanced disease the cytolytic function of CD8+ T cells directed against opportunistic pathogens is demonstrably impaired.(56) It is not entirely clear whether the CD8+ cells present in early disease are functionally "normal," as the maturation phenotype of CD8+ T cells recognizing pathogen-derived peptides has been found to be variably perturbed.(57) Whether this is the cause or the consequence (or the interaction of both) of greater exposure to opportunistic pathogen-derived antigens in HIV-infected immunosuppressed persons is difficult to sort out.

As is seen with CD4+ T cells in HIV infection, CD8+ T cells obtained from HIV-infected persons may fail to proliferate in response to TCR activation in vitro.(58) In this setting, however, it is not clear whether the failure to proliferate is a consequence of failure of CD4+ T-cell help (via provision of IL-2 that is essential for CD8+ T-cell proliferation), a reflection of an intrinsic failure of CD8+ T-cell function, or a consequence of CD8+ T-cell maturation to a predominantly effector phenotype.

B Lymphocytes and Antibody Production

As with cellular immune responses, the humoral immune system in HIV infection is characterized by paradoxical hyperactivation and hyporesponsiveness. Hyperactivation is reflected in dramatic polyclonal hyperglobulinemia, only a portion of which is directed against HIV antigens;(59) bone marrow plasmacytosis;(60) heightened expression of activation molecules on circulating B lymphocytes;(61,62) the presence of autoreactive antibodies in plasma;(59,63) and instances of clinical autoimmunelike disease. B-cell hyperreactivity may contribute to the increased risk of B-cell lymphomas in HIV-infected persons, but no causal link has been clearly established.(64) Neither is the etiology of hyperglobulinemia well understood. Elevated plasma levels of the endogenous B-lymphocyte stimulator have been found in HIV-infected persons (65,66) and this may contribute to the B-lymphocyte activation of HIV infection and AIDS. At the same time, diminished B-lymphocyte responsiveness to antigenic stimulation in vitro is characteristic of HIV-infected persons,(62,67-69) who often fail to develop protective antibody responses after immunization with protein or with polysaccharide vaccines.(70-73) The characterization of antibody responses to polysaccharides as "T-cell independent" is only partially correct. Although antibodies can be induced to polysaccharides in the absence of linked peptides that induce cognate help by proximate CD4+ T cells, these responses are not optimal. Moreover, B-lymphocyte responses to pure sugars still require some degree of T-helper support. Lack of CD4 help may therefore underlie the poor antibody responses to polysaccharides that are seen in HIV infection.

Monocytes and Macrophages

Tissue macrophages are often infected with HIV in vivo (74-77) and, because they are generally not killed by the virus, may serve as reservoirs for viral replication. In tissue sites, infected macrophages may be the source not only of viral proteins but also of inflammatory mediators of pathology such as proinflammatory cytokines, tumor necrosis factor (TNF), interleukin-1 (IL-1), IL-6, and IL-10, as well as chemotactic chemokines. Circulating monocytes on the other hand seldom have been shown to harbor infectious HIV; moreover, in vitro studies of blood monocytes largely have failed to show substantial HIV-induced impairments in key functions such as antigen presentation and differentiation.(78,79) On the other hand, in vitro infection of monocyte-derived macrophages with HIV dramatically impairs the ability of these cells to ingest and kill foreign microbes (80,81) as well as to present antigen to T cells.(82) These findings suggest that impaired function of these infected cells in vivo may well contribute to the overall immune dysfunction of HIV infection.