HIV and SIV infection - the role of cellular restriction and immune responses in viral replication and pathogenesis

Abstract

The human immune deficiency (HIV) and simian immune deficiency (SIV) viruses have long biological history. Both viruses evolved from Africa and reminants of them can be found in the “fossil record” of several species in which they are not endemic. SIV remains endemic in several species of monkeys in Africa where it does not cause immune deficiency. HIV and SIV actively replicate within humans and Asian non-human primates, despite cellular and genetic viral restriction factors and genes, and at times robust innate and adaptive immune responses. While Lentivirus are considered “slow viruses” it is clear in humans and susceptible Asian monkeys that virus production is rapid and highly active resulting in a massive loss of CD4⁺ memory effector T cells early after infection and a continued race between viral evolution, cytotoxic lymphocytes, and failed neutralizing antibody responses. Concurrently, HIV and SIV can infect monocyte/macrophage populations in blood and more importantly in tissues, including the central nervous system, where the virus can remain sequestered and not cleared by antiretroviral therapy, and hide for years. This review will discuss species and cellular barriers to infection, and the role of innate and acquired immunity with infection and pathogenesis of HIV and SIV in select species.

The human immunodeficiency virus (HIV) and the closely related simian immunodeficiency virus (SIV) are obligate parasites that invade host cells and utilize their cellular transcriptional machinery to reproduce. Natural SIV infection of African non-human primates is asymptomatic and usually does not induce significant lymphocyte depletion despite high levels of virus replication. SIVcpz from chimpanzees and SIVsm from sooty mangabeys have crossed species barriers resulting in the generation of HIV-1 and HIV-2, respectively (1–4). Following this cross-species, zoonotic transmission and infection by these viruses, there is a rapid and elevated viral replication in the host, subsequent immunodeficiency, and eventual death as a result of severe immune suppression. In addition, in the experimental model of SIV infection using the non–natural host Asian monkey species results in the development of illness similar to that described in AIDS patients (simian AIDS). It is in these latter situations that HIV and SIV are not stable parasites.

Understanding the benign nature of SIV infection in natural hosts is likely critical to understanding AIDS pathogenesis in humans by HIV and Asian macaques by SIV, and a focus of this concise review. This review will briefly describe the history of HIV and SIV infection, the results of zoonotic transmission, cellular host restriction factors, as well as genetic disposition to infection. Next, we will discuss the immune response to initial acute and then chronic infection, focusing on dynamics of the agents of immune surveillance, and mediators of pathogenesis in the peripheral blood, mucosa, and the central nervous system (CNS).

History of HIV and SIV infection and zoonotic transmission

HIV and SIV belong to the Lentivirus genus and the Retroviridae family. Like all viruses they are obligate parasites, which cannot live outside a living cell or tissue. Classically, the lentivirus is considered a “slow virus” that can infect a broad range of mammalian host cells. It is commonly accepted that SIV, originally from African monkeys, was transmitted to humans, giving rise to HIV and resultant multiple outbreaks of HIV infection and AIDS. The viral strain thought to give rise to HIV-1 clade B responsible for the AIDS epidemic in the US was thought to arise around 1931 (5). All known strains of HIV-1, including the major group M (responsible for the global AIDS epidemic) as well as groups N and O (found only in West-Central Africa), are closely related to SIVcpz strains infecting chimpanzees (1, 3). HIV-2 originates from an independent transmission event where virus was passed from sooty mangabeys to human (2, 4).

Monkey strains of SIV are sexually transmitted and do not usually result in immunodeficiency in their natural hosts, despite the hosts carrying large viral loads. Thus, while SIV is endemic in several strains of African monkeys, it does not cause AIDS. However, horizontal infection of male Asian monkeys by African monkeys via male-to-male biting resulted in lymphoma and AIDS-like illness (6). This transmission paved the way for future experimental SIV infection of Asian macaques and experimental simian AIDS (6). Such work has pointed to cellular restriction factors inhibiting viral infection, the role of genetic backgrounds in disease severity and pathogenesis, as well as innate and acquired immune responses controlling infection.

Cellular restriction factors

The observation that experimental SIV strains do not infect human cells in vitro and experimental infection of non-human primates with HIV consistently failed has lead researchers to examine the possibility of cellular host restriction factors that can block or inhibit infection. Two such genetic loci and gene products that have received attention and will be discussed here are the tripartite motif-5alpha isoform (TRIM5alpha) and apolipoprotein beta mRNA-editing enzyme catabolic polypeptide 1-like protein G (APOBEC3G). Using chimeric SIV/HIV (SHIV) studies consistently demonstrated that viral factors required by chimeric SHIV to infect monkeys mapped to the 5’ half of the HIV genome (7, 8). It was found that SHIV that replicated in monkeys must have minimally the SIV parental gag and vif. Subsequent work showed that gag is the target of TRIM5alpha and HIV Vif counteracts APOBEC3G. These interactions are thought to be critical in the understanding the block of HIV infection in non-human primates and SIV infection in humans. It is thought that knowledge of the interactions of these factors with virus will provide insight into anti-viral mechanisms and possibly aid the development of novel vaccine strategies.

TRIM5, found in cells of most primates, is important for anti-viral activity, and is therefore considered part of the innate defense of cells against retroviral infection. TRIM5alpha is a member of the tripartite protein motif family and is an intracellular protein that recognizes and degrades the capsid protein of retroviruses, resulting in a block of viral replication (8). TRIM family members are comprised of an N-terminal RING-like domain, B-box domains and a coiled coil region. The TRIM5alpha protein belongs to a subset of TRIMs that contain a B30.2/SPRY domain. There is a high degree of variation within this domain in primates and it is thought that species-specific restriction to infection can be attributed to this region (9). Variation among TRIM5alpha orthologs accounts for the observed patterns of post-entry blocks to retroviral replication among primate species. Thus, HIV-1 can enter cells of Old World monkeys but encounters a block before reverse transcription can occur. The TRIM5alpha from Old World monkey confers this potent resistance to HIV-1 by acting on the incoming caspid. In contrast, TRIM5alpha from Old World monkey cannot block SIV capsid. (8). In addition, the human ortholog of TRIM5alpha is unable to specifically target HIV. However, the human TRIM5alpha does effectively restrict other retroviruses, such as MLV and EIAV (10). It is possible to assume that HIV and SIV have evolved in their natural hosts to achieve a low level of interaction with TRIM5alpha. The difference in the ability of the human and monkey TRIM5alpha to restrict HIV hinges on a single amino acid. Amino acid 332 in human TRIM5alpha encodes an arginine while the monkey ortholog encodes a proline (11). In vitro experiments, changing the arginine to a proline in the human TRIM5alpha resulted in blocked HIV infection in culture (12). Understanding the species-specific restrictions to HIV infection by monkey TRIM5alpha might allow for the development of an animal model of HIV infection, where monkeys are infected by HIV and not SIV.

In addition to TRIM5alpha, early research pointed to the cellular deaminase, APOBEC3G, as a cellular restriction factor of HIV based on its interactions with the HIV auxiliary protein Vif (virion infectivity factor). Vif is essential for productive HIV infection in cells such as primary peripheral blood lymphocytes and macrophages (13, 14). Early studies demonstrated that a Vif deletion caused decreased infectivity of progeny HIV particles, an effect which was due to a dominant cytoplasmic restriction activity (14, 15). APOBEC3G is packaged with the viral genome of HIV particles. Upon infection of a new cell, and during reverse transcription, APOBEC3G deaminates 2'-deoxycytidines (C) in the DNA minus (−) strand, producing 2'-deoxyuridines (U). Plus-strand synthesis converts the C'U changes into G'A mutations (16–19). The resultant hypermutations can lead to strand breakage, impaired viral initiation of plus viral strand synthesis, or encode stop signals and thus impair subsequent viral functions. In non-permissive cells expressing APOBEC3G, the HIV Vif protein forms a complex with APOBEC3G and is responsible for its polyubiquitation, subsequent degradation and prevention of its encapsidation into virions (20, 21). Thus, APOBEC3G blocks viral replication and HIV Vif works to overcome this block.

HIV Vif is able to inhibit human APOBEC3G, but is not effective against orthologues in monkeys, such as rhesus macaques and African green monkeys (agm) (22) and the reverse is true; SIVagm-Vif counteracts restriction of monkey APOBEC3G, but not human APOBEC3G. Recent literature suggests the potential for cross species transmission, or lack-there-of, may be determined in part by variations in human and monkey APOBEC3G proteins (23). The difference in species-specific sensitivity between Vif and APOBEC3G has been mapped to a single amino acid change in APOBEC3G (24, 25). The human APOBEC3G encodes an aspartate at amino acid 128 and the African green monkey APOBEC3G encodes a lysine. HIV-1 Vif can inhibit the African green monkey APOBEC3G when the lysine is changed to an aspartate. SIVagm Vif counteracts restriction in when the human APOBEC3G amino acid is changed to a lysine (24–27). This single amino acid may be a barrier that protects humans from infection with SIV strains or a mechanism by which HIV cannot infect monkeys. In addition, HIV and SIV Vif share limited homology (30%) and display species-specific activity (28), leading to speculation that Vif sequences could determinethe block in HIV-1 replication in rhesus monkeys.

Innate intracellular immunity mediated by TRIM5alpha and APOBEC3G may play a particularly crucial role in the defense against lentiviruses. Efforts aimed at enhancing these innate immune factors may ultimately prove to be useful in protecting humans from HIV infection.

HIV and SIV infection–the role of genetic variation

In addition to species barriers to infection described above, genetic differences within species play a role in restricting infection, and perhaps regulating disease severity (see (29)). Thus, certain HLA alleles have been linked to delayed or rapid disease progression in humans infected with HIV and monkeys infected with SIV. The HLA class I system regulates two crucial and related mechanisms for controlling virus. First, HLA I binds viral peptides, affecting the generation of cytotoxic T cell (CTLs) and their response to viral antigens. CTLs destroy infected cells by targeting viral peptides bound to class I molecules. Second, the interaction of class I molecules and the natural killer (NK) cell immunoglobulin-like inhibitor receptors (KIR) play a role in NK-mediated and viral immunity. Both CTL and NK activity promote or impede the initial establishment of infection, confirming the importance of class I molecules, CTLs and KIRs.

A recent paper has described the first use of a whole-genome association strategy to identify human genetic polymorphisms that influence the variation of viral loads in HIV infected patients (30). They reported an association between the HLA-B*5701 allele and viral loads in patients and, in addition, a prominent role for HLA-C. HLA-B*5701 is found among Caucasians and is closely related to the African HLA-B*5703. Both exhibit broad reactivity across a number of gag epitopes (31). There are two other alleles that have shown consistent associations with HIV-1 disease: HLA-B*27 and HLA-B*35 (32–34)The former associates with slow progression and the latter with rapid progression. The possible functional mechanisms of these class I molecules are unknown, however, it is thought that HLA-B*27 may present a conserved immunodominant epitope that is under structural constraint. In addition to specific HLA genotypes, delayed progression to AIDS has been found in those patients with a marked increase in the heterogeneity of the HLA class I loci (32). Furthermore, it has also been reported that homozygosity at class I loci can reduce the CTL repertoire, allowing immune escape to occur rapidly leading to accelerated disease progression (35). Associations between HIV disease progression and MHC class II loci are not as strong as those observed with class I, suggesting that cell-mediated immunity maybe stronger than humoral responses (36).

MHC class I alleles, including Mamu-A*01, Mamu-B*17, and Mamu-B*08 are associated with slow disease progression in SIV-infected macaques (for review see (37)). Expression of two common class I alleles Mamu-A*01 and Mamu-B*17 has been associated with lower viral set points during the chronic phase of infection. Animals with the Mamu-B*08 allele are strongly associated with immune control of SIVmac239 and are elite controllers (38). There is also evidence of a severe reduction and selection of the class I repertoire resulting in susceptibility to disease and a widespread viral infection (39). The use of SIV infected macaques, such as Mamu-A*01, Mamu-B*17, and Mamu-B*08 positive animals, that control viral replication will be a useful tool to understand how humans control HIV infection.

A combination of HLA alleles and certain KIR polymorphisms have been shown to be protective. Qi and colleagues showed KIR3DS1/HLA-B Bw4-80I, which presumably favors NK cell activation, to provide protection against HIV disease and certain opportunistic infections(40). Thus, control of the rapidly evolving pathogen requires host genetic factors that are highly polymorphic and capable of evolving rapidly themselves, characteristics of both the HLA class I and KIR loci.

Genetic differences in genes that encode chemokines or their receptors are also associated with susceptibility to HIV infection. One of the most studied and well known is the Δ32 mutation in the HIV-1 co-receptor CCR5, which generates a non-functional receptor that does not support virus attachment and entry by R5 or dual tropic HIV viruses (41). Cells isolated from individuals that are homozygous Δ32/Δ32 are resistant to R5 tropic viruses (41).Infected individuals who are heterozygous comprise a lower frequency than in the general population and display a slower progression to AIDS (41, 42). Additionally, polymorphisms and copy numbers of the ligands of CCR5 and the other common co-receptor, CXCR4, have been implicated as factor in disease progression. Perhaps the most impressive and recent example of the importance of CCR5 in viral infection was the report of an HIV infected patient that received a bone marrow transplant from a CCR5Δ32 donor (43). The patient has remained without viral rebound for 20 months post-transplantation and discontinued anti-retroviral therapy.

In addition to the several host factors associated with delayed progression, in a few individuals this slow progression is related to virus genotype. Several defects in the viral genome of HIV-1 strains infecting long-term non-progressors have been reported, such as mutations in rev, vif, vpr, vpu and nef. A deletion close to the 5' end of Nef that impaired Nef function was found in a long-term non-progressor with HIV infection (44). Patients infected with nef-deletion strains have a significant longer survival time than patients with wild-type virus. However, nef-deletion viruses can lead to immune deficiency and a decline in CD4 counts even when viral levels in the plasma are low (45).

HIV and SIV infection of target cells – the role of cellular restriction factors

HIV and SIV infect CD4+ T cells and monocyte/macrophages, and in some instances dendritic cell (DC) populations using combinations of CD4 and chemokine receptors, the most common being CCR5 and CXCR4 (46). R5 viruses are characterized as ones that use CCR5 for infection and X4 use CXCR4. Chemokine receptor usage is critical in determining tissues targeted by infection.

In addition to the chemokine receptor usage, activation state of the CD4+ T cell dictate viral infection. Newly emergent lymphocytes from the thymus are considered “naïve” resting cells. These cells, which are abundant in blood and organized lymphoid tissues, are refractory to infection by HIV and SIV. Naïve CD4 T lymphocytes do not express significant levels of CCR5 and use the cellular restriction factor, APOBEC3 (47)(13, 48–50). However, long-lived effector memory CD4 T lymphocytes are significant targets of HIV and SIV. Viral infection of memory effector population results in cell lyses and death and thus has an early critical deleterious effect on future anti-viral responses.

Macrophages and some subsets of dendritic cells, which express CD4 and CCR5, are targets of HIV and SIV in vivo (51–53). On the other hand, non-activated blood monocytes are not permissive to HIV infection, though SIV infection of monocytes has been reported (53, 54). Monocytes traffic from the bone marrow through the blood and into tissues where they become macrophages. Monocytes are often thought as the vehicles that carry virus, as integrated provirus, to tissues but, it maybe that monocytes acquire the ability to replicate virus only once they have differentiated into macrophages (55–58). Current evidence suggests that restriction of HIV replication in monocytes is independent of reverse transcription precursors, but dependent on differentiation dependent cellular cofactors (59). Monocytes and DC populations have higher levels of the anti-HIV protein APOBEC3G, and the recently reported APOBEC3A, compared to macrophages. The inhibition of infection maybe attributed to presence of these factors (60). Although macrophages harbor potent antiviral restriction factors, primate lentiviruses have evolved the accessory protein Vpx, which counters this restriction via its interaction with the Vpr interacting protein, damaged DNA binding protein 1 (DDB1) (61, 62). Unlike infected CD4 T cells, the productive infection of monocytes and macrophages does not result in cell lyses and in fact produces a long-lived cellular reservoir for virus replication (63). What causes infected macrophages to live longer than non infected cells, or why they are not lysed by viral infection is unknown. Stevenson and authors showed that potential apoptotic killing of HIV infected macrophages is subverted by the envelope glycoprotein (env) which down regulates the death ligand tumor necrosis factor related apoptosis induced ligand (TRAIL-R1/D4) and up regulates anti-apoptotic genes Bfl-1 andMcl-1 (61). Similar to monocytes, where dendritic cells become infected depends upon there maturation (64). In contrast to monocytes, which are relatively refractory to HIV infection, a small subset of immature dendritic cells (iDC) can by infected by HIV. Recent reports underscore the role of APOBEC3G/3F in regulating such infection. Interestingly, as iDCs mature their cellular levels of APOBEC3G/3F increase correlating with increased resistance to infection (60).

The resistance of myeloid progenitor cells to HIV infection demonstrates an especially interesting story of the biology of endogenous anti-viral factors within cells. It would seem critical from the hosts point of few that hematopoeitc stems cells are resistant to infection, because such infection early could wipe out critical cells of the immune system. The hematopoietic stem cells (HSC) are one of the only cell types that resist HIV infection despite the presence of HIV receptors, CD4 (expressed at low levels), CCR5 and CXCR4. It has been shown that the block to HIV infection of HSCs occurs at viral fusion and entry. This block to infection can be over-ridden experimentally, using pseudotyped virus containing the envelope of vesicular stomatis virus (VSV) G type underscoring the role of envelop fusion and entry as a restrictive factor (65). In addition to fusion and viral entry, P21 within the HSC has been shown to play a role inhibiting infection. Silencing p21, using siRNA constructs, results in low-level HIV replication in HSC (66). This result is specific to HIV since SIV infection of HSCs was not affected by depletion of p21. p21 functions in normal cell physiology as an inhibitor of cyclin-dependent kinase in the G1 phase of cell cycle. In HSCs, p21 appears to act by blocking cell-cycle progression as well as interacting with DNA repair proteins. Controlled infection of HSC results in the formation of excessive circular forms of HIV from a lack of proper integration. On the other hand, in p21 depleted HSCs integrated provirus can be seen. Cells are less susceptible to infection if they are quiescent or undergoing DNA repair pathways, in these ways p21 maybe acting in inhibiting HIV infection in HSCs. Thus, some CD4 T cells, monocytes/macrophages and dendritic cells have anti-viral factors or maturation and activation states, which can inhibit viral infection. Despite this, critical elements of the immune system are infected early and later in the course of HIV and SIV, resulting in immune and histological pathogenesis. This is discussed below, following the time course of infection and critical target cells.

HIV and SIV infection early, asymptomatic and chronic events of infection

The most common route of HIV infection is across mucosal barriers a result of sexual exposure. The first infected cell detected in humans and monkeys are resting memory CD4 T lymphocytes, which are found in the mucosa three to four days after infection. Within a week virus is detected in the mucosal lymph nodes where it interacts with the tissue DCs, the Langerhans cells, which carry virus on their surface to lymph nodes. Virus production is increased in draining lymph nodes where DC’s and perhaps monocytes interact with viral antigen specific CD4 T cells resulting in amplification of infection through CD4 T cells. Interestingly, this interaction can also result in maturation and activation of monocytes (67, 68). At this time in the gut, virus spreads to gut-associated lymphoid tissues (GALT) resulting in an increased infection of CD4+ memory effector cells resulting in their rapid depletion (69–72). With loss of CD4 T cells in the gut, a considerable translocation of bacteria occurs across this barrier, which is thought to activate innate immune responses and monocyte/macrophages likely via toll-like receptors (TLRs) (73, 74). The dramatic loss of CD4 T cells in the gut and bacterial translocation points to a rather quick and severe effect of this “lentivirus,” which was once considered a “slow virus” when considering the gut mucosa. The dramatic loss of CD4 effector T cells in the gut is huge in contrast to the small change in the effector T cells in the blood where they are a minor population of total lymphocytes (70).

The rapid depletion of CD4⁺ effector T cells in the gut coincides with peak plasma viral production around 21 days post infection. In the majority of patients with the exception of rapid progressors, peak plasma virus drops shortly thereafter, which maybe due to the depletion of the target CD4 T population, the development of CCR5 memory T cells, or the presence of viral antigen specific CD8 T cells (75). Viral set point is then established in patients or animals not on anti-retroviral therapy. Using the SIV-infected monkey model, one can readily detect virus throughout the body in the first day after intravenous infection. Virus can be found immediately in all lymphoid tissues and mucosal tissues, and as early as 14 days within the brain. Within lymphoid tissues evidence of productive infection is first found in the paracortex of the lymph nodes, the periarterioloar sheaths in the spleen, and the thymic medulla, which underlies the thymic cortex and contains mature lymphocytes. Early in the lymphoid tissues, memory CD4+ CCR5+ T cells are again the primary targets of infection, similar to the mucosal tissues (76, 77). Between 2 to 3 weeks after infection, macrophage infection occurs in lymphoid tissues, due to the evolution of virus and a switch in co-receptor usage from CXCR4 to CCR5-tropic viruses.

Diffuse labeling of viral RNA and proteins is observed over germinal centers due to antibody trapping of virus and their contact on follicular dentritic cells (78). Infection of the thymus occurs in the face of high CD4 T cell regeneration, resulting in thymic involution. Interestingly, in contrast to the widely held belief that this involution resulted in loss of thymus function and CD4 T cell depletion, T cell progenitors are released at a higher rate from the thymus despite high T cell death in the periphery (79). It has been suggested that if antiretroviral therapy is instituted during this stage of thymic involution, successful anti-viral immunity utilizing CD4 helper cells can be achieved (69).

Innate and adaptive immune responses to HIV and SIV

In addition to the innate anti-viral response of CD4 T cells, monocytes and DC, the immune system generates innate and adaptive immune responses targeting HIV and SIV. The early innate responses, which clearly do not do the job, consist of natural killer (2–3 fold increase in blood) cells (NK), activated DC’s responses (80, 81), and elevated anti-viral cytokines including interferon(IFN). Adaptive immune responses consist of CD8 T cells and antibodies. Evidence using the SIV infected rhesus macaque model where a chimeric, humanized anti-CD8 depleting antibody was infused early in infection, demonstrated that when CD8 T cells were depleted, SIV viral RNA in the plasma was elevated. Viral levels decreased significantly following the return of the CD8 T cells in blood, thus establishing the role of virus specific CD8 T cell immunity in controlling plasma virus (75). In humans, CD8 T cell numbers increase by 10% with infection and correlate nicely with virus reduction (82, 83). With increase in CD8 T cell activity there is a heightened CD4 T cell response, which is somewhat compromised by HIV infection, but likely represents a form of CD4 help (84). Although it is generally thought that early infection by HIV and SIV triggers strong immune responses, it is clear that the magnitude and quality of such immune responses decrease over time. Thus, CD8 killing is diminished over time as are CD4 T cell responses to recall antigens (85, 86). So, the development of AIDS occurs potentially over a long period of time despite memory CD4 T cells being depleted relatively early in infection. This points to the continued and apparently sustained increase of CD4 T cells maintained and turnover throughout infection that maintains threshold levels of mucosal CD4 T cells. Likely, the ongoing destruction of memory CD4 T cells is balanced by proliferation of these cells to maintain a steady level. This is in contrast to the idea the viral latency, where it was thought that infection becomes dormant. Ho et al., used antiretroviral therapy to measure viral turnover and found high sustained viral replication even in asymptomatic patients (70, 87). Once viral set point is established, CD8 T cells continue to hold viral levels low. This is best evidenced by the ongoing generation of viral escape mutants, a response by which the virus mutates from continued selective immune pressure (88–90). Both CD4and CD8 T cell responses evolve with viral infection and different immunodominant patterns occur with acute and chronic infection, likely a result of the generation of viral escape mutants, viral fitness as well as new T cell responses to viral mutation.

After primary infection, neutralizing antibody responses are detected, but seem to lag behind viral evolution, questioning the efficiency of neutralizing antibodies controlling virus. However, neutralizing antibodies in specific human individuals has demonstrated the ability to stop viral infection directly, or block CD4 entry of virus in target cells (91–93). The continued replication of virus in patients and monkeys, naïve to highly active antiretroviral therapy (HAART), eventually leads to opportunistic infections, and the development of AIDS. In a small percentage of the populations, there exist unique individuals who have low to non-detectable plasma virus and do not develop AIDS. These individuals fall into two groups, one termed long-term non-progressor who have relatively low viral loads, and so called elite controllers with levels of < 50 RNA copies/mL. Genetic and immunologic studies have demonstrated members of these groups have CD8 T cell responses that target Gag over other proteins with diverse HLA class I alleles. The elite controllers have significantly more CD4 and CD8 T cells that secrete INFgamma and IL-2, and lower levels of HIV neutralizing antibodies (94). Some have found a strong association of HLA-B57/*5801 that select for rare GAG variants that are associated with strong control (31). In addition, robust mucosal responses have been found in the elite controller group (95).

Although it receives less attention in the literature, macrophage infection in tissues is significant, found in the bone marrow, gut, lung, lymph node and brain, and likely contributes to cellular and tissue reservoirs of HIV and SIV over time.

CNS infection by HIV and SIV, the role of monocyte/macrophages

The central nervous system by HIV and SIV provides and example of long-term infection and pathogenesis which can occur over years, and perhaps an example of the virus functioning as a good parasite, that is long lived. Virus enters the CNS early after infection in humans and non-human primates, as early as 7–21 days post infection, concurrent with peak plasma virus (69, 96). At this point, viral DNA is detected within perivascular cuffs, as well as in the choroid plexus and meninges. During the asymptomatic phase of infection, it is difficult, if not impossible, to detect productive viral infection (viral RNA and protein). Productive infection is robust with the development of AIDS, where perivascular macrophages, multi-nucleated giant cells (MNGC) and in some cases parenchymal microglia are infected. Factors controlling the development of encephalitis and giant cell formation in the CNS are not well defined, but clearly immune suppression plays a role. This is best exemplified in the rhesus macaque model of neuroAIDS. In monkeys, similar to humans pre-HAART, only approximately 30–40% of infected animals develop AIDS and SIV encephalitis with MNGC, virus in the CNS and infected macrophages (97). And this occurs usually over the course of 2–3 years post infection. CD8 lymphocyte depletion of animals during the time of infection results in elevated plasma virus during the entire time of infection, and greater than 95% of the persistently CD8 lymphocyte depleted (more than 28 days CD8 depleted) animals develop AIDS and SIV encephalitis, and this occurs rapidly within 3–5 months (98).

The mechanisms of viral entry into the CNS is most likely via precursor cells to perivascular macrophages, that come from bone marrow, through the blood and enter the CNS on their way to becoming perivascular macrophages (96, 99). We and others have demonstrated that the immune phenotype of CNS perivascular macrophages is very similar to populations of blood monocytes, by flow cytometry (100–103). Moreover, there is an expansion of activated monocytes in the blood that correlated with the incidence of HIV encephalitis (HIVE), and CNS disease severity in monkeys (103). The factors that determine the release of these cells from the bone marrow to traffic to the CNS and other tissues is not define, but clearly microbial translocation from the CD4 T cell depleted gut likely plays a role (74). Kuroda et al. (unpublished data) have also made the interesting observation that BrdU+ monocytes, leaving the bone marrow 24 hours post BrdU administration in the periphery, increases with animals that subsequently develop immune deficiency and is a better predictor of animals that will develop immune deficiency. It appears that these cells are leaving the bone marrow to traffic in a response to replace apoptotic cells in lymphoid tissues (unpublished). Thus, viral infection of macrophages in tissues, which might lead to macrophage apoptosis, can potentially also result in accumulation of macrophages and virus in tissues. Related to this, we have consistently found proliferating cellular nuclear antigen (PCNA) expression in macrophages within the bone marrow and brain of SIVE animals, were it is associated with macrophages that are infected and not undergoing significant proliferation (104). It is very likely these cells have recently taken residence in the brain and are undergoing DNA repair, and effect of active viral replication (104). One last observation in terms of CNS infection is that viral strains in the CNS appear to be compartmentalized, and evolve in the CNS in a manner that is distinct from other target organs (105). Salemi has recently demonstrated using phylogenetic analysis and molecular clock that the CNS meninges are a constant source of virus first entering the CNS, from which newer replicated species distribute in different brain regions. These observations underscore the blood monocyte/macrophages and perivascular macrophages as a source of new virus seeding the CNS, but also suggest that virus can continue to evolve within brain macrophages at different sites, independently (105). The role of viral recombination within superinfected macrophages (more than one viral particle per cell) will likely be important for future studies of brain infection. Such studies might point to when important viral strains enter the brain, and whether such virus is carried by monocyte/macrophages.

Summary

HIV and SIV are pathogens with a very long history and evolution. Several species have developed endogenous blocks to infection, which are not fully functional in humans and new world monkeys. In addition to cellular restriction factors to infection, innate and acquired immune responses are operational. HIV and SIV infection of susceptible hosts results in rapid and consistent depletion of CD4 memory effector cells, and macrophage populations. Infection of mucosal sites including the gut is rapid and on going, while infection of tissue macrophages in sites including the CNS can occur over years and represents a cellular and tissue reservoir. Future study of restriction and genetic factors as well as host response to infection are critical to future therapies to control viral replication and clear viral reservoirs.