Changes in CD4 cell count and viral RNA burden are common markers for HIV disease progression. However, evidence has existed for several years that many patients with HIV disease experience a broad loss of leukocyte subsets without an apparent preference for depleting CD4 T cells 123. Effects on cell types other than CD4 T cells were documented in macaques after showing substantial B cell loss during acute SIV infection 4, and in humans by showing depletion of γδ T cells 5 that are CD4 negative. Many other examples confirmed that the profound impact of HIV on "non-CD4" leukocyte populations depends on indirect mechanisms, as opposed to direct cell killing that occurs when HIV or SIV infects and destroys susceptible CD4+ cells. Surely, uninfected CD4+ cells can also be destroyed by indirect mechanisms if they escape direct infection. Since both direct and indirect mechanisms are driven by viral burden, it has been difficult to distinguish their contributions to CD4+ T cell depletion and progressing disease. This technical obstacle has blocked efforts to explore new therapies that target indirect mechanisms for cell depletion in HIV disease.
Besides depleting immune cells, HIV infection is characterized by defects in cell effector functions. Low cytotoxicity among circulating HIV-specific CD8+ T cells has been attributed to loss of CD3-complex zeta chains 6 and T cell exhaustion was linked to high levels of PD-1 or CTLA-4 on the surface of T cells 78. During chronic HIV infection, immune dysfunction can also result from interactions with regulatory T cells (Tregs) 910 that are exported to the periphery during high thymic turnover 11. Ultimately, Tregs contribute to immune suppression and to cell death via the Fas/FasL apoptotic pathways (Fig. 1).
Mechanisms of indirect cell depletion in HIV infection
Despite overwhelming evidence for indirect cell depletion, little is known about how it occurs in vivo. Cell loss is accelerated during elevated viremia and a broad recovery of leukocytes attends highly active antiretroviral therapy (HAART) 12. Some of these changes may be related to release of sequestered lymphocytes from secondary lymphoid tissues 13 as viral antigen declines during therapy. HIV effects on bone marrow, thymus, and hematopoietic stem cells were also proposed as causes for general leukocyte loss 14. While HIV does indeed infect and alter bone marrow and thymus, the kinetics for cell loss during acute infection, the rates for recovery during HAART and the durable repertoire defects in T and B cell subsets after HAART 1516 argue that these dynamic changes occur within mature populations without an overwhelming impact from declining stem cell output. With large expansions and contractions of mature cell populations, declining stem cells may have a lesser impact on leukocyte depletion in AIDS.
A process described as "chronic immune activation" or "hyperactivation" occurs during HIV disease and is accompanied by higher expression of TNF superfamily ligands and their receptors, e. g. Fas/FasL and TRAIL-DR5 17181920. Chronic activation coincides with "activation induced cell death" (AICD), a term coined initially to describe cell loss occuring when lymphocytes are activated by viral antigen in the absence of appropriate costimulation. AICD connects viral gene expression to general lymphocyte destruction, ultimately resulting in reduced immune protection from HIV. Now it is known that both viral and host products can activate lymphocytes and induce death receptor expression. Secreted HIV Tat protein up-regulates FasL and TRAIL in T cells or macrpophages, which in turn, induce apoptosis in bystander cells thereby providing a mechanism for cell death among uninfected cells 212223. In HIV disease, both antigen-specific and polyclonal activation have been observed 172425. This pan-activation sensitizes lymphocytes to apoptotic cell death that occurs when the death ligand, usually cell-associated, contacts a cell that is expressing death receptors.
Early investigators 26 proposed apoptotic death as an important contributor to CD4 T cell depletion. They observed that triggering through the T cell receptor failed to induce proliferation of PBMC from HIV+ asymptomatic donors; instead, CD4 cells in these cultures had features consistent with Fas-mediated apoptosis 27. We confirmed this result in SIV-infected macaques 28. Alternatively, signaling through the T cell receptor 29 can activate the mitochondrial pathway for apoptosis, distinct from the Fas-triggered caspase-8 death pathway 30. Both pathways are generally invoked in AICD (Figure 1). Since antigen-specific memory T cells are more likely to express Fas, cell killing appears tied to antigen specificity. Both caspase 8 and mitochondrial apoptosis mechanisms tend to delete immune cells that respond to antigen. Preferential apoptotic cell death among antigen-activated T cells could account for the lack of immune control over opportunistic pathogens and an insufficiency of viral immunity that fails to prevent viral persistence or progressing disease. Once the host environment is "set," and after viremia triggers higher expression of FasL, every encounter with antigen has the potential to drive cell depletion among all lymphocyte compartments.
Role of Fas/FasL mediated cell death in HIV/SIV infections
Apoptosis is observed consistently among uninfected cells in SIV+ macaques or HIV+ humans 31. Chronic immune activation drives cells into apoptosis 32 possibly involving Fas/FasL interactions 33, and reflects an exaggeration of the normal processes for homeostatic cell regulation during HIV disease 34. The roles of assassin and victim are not always clear in these interactions. Activated CD4+ T cells can express FasL and become the effectors of cell death including the destruction of resting B cells 35, even as they also become targets for cell killing and display high rates for apoptosis during disease.
Many studies have explored the mechanisms connecting HIV/SIV antigen expression and apoptotic cell death. Human macrophages express FasL after exposure to HIV 36, creating a link between antigen presentation and Fas/FasL-mediated apoptosis. HIV up-regulates FasL in CD4 T cells 3738 after they are exposed to soluble Tat, gp120 22 or Nef proteins 39. Higher levels of FasL, both cell-associated and in plasma, and Fas were observed in specimens from HIV+ patients whose PBMC were especially susceptible to Fas-mediated cell death in vitro 40. Cross-linking of CD4 by gp120 complexes or viral particles increased the susceptibility to apoptosis triggered by FasL or TNF-α 4142.
The patterns of CD4 T cell depletion appear to be antigen-specific, as perturbation of the CD4 receptor repertoire is significantly associated with higher plasma viremia 43. This can be explained by an AICD mechanism or by virus infection of antigen-activated cells as we proposed 44. Less data are available to show whether CD8+ T cell or B cell loss during HIV infection is antigen-specific, although altered receptor repertoire have been reported in both cases 454647. The γδ T cell population shows a specific pattern of depletion in HIV disease 5, losing the critical Vγ2-Jγ1.2+ subset that is required for pathogen and tumor cell responses but few cells express CD4 and γδ T cells do not support HIV replication. To the extent that antigen stimulation is related to cell depletion, AICD may be invoked as a mechanism for γδ T cell killing. Thus, AICD is a mechanism that potentially links antigen stimulation with expression of death ligand receptors, leading to specific cell depletion.
Neuronal cells that are depleted during chronic HIV infection comprise an important example of bystander depletion since they are not infected by HIV. Neuronal cell loss has been attributed to interaction with viral proteins such as gp120, vpr, nef, and tat, and to soluble neurotoxic factors released by infected macrophages 48. The primary mechanism for neuronal depletion is apoptosis via extrinsic Fas-related death receptors or intrinsic mitochondrial pathways 48. The fate of neuronal cells during AIDS is reminiscent of the cell culture studies mentioned previously that documented apoptotic destruction of uninfected cells upon exposure to viral proteins 2024383949.
Broadly acting immune stimulation during HIV or SIV infection especially implicates the Fas death pathway since activated T or B cells express Fas. Acute SIV infection triggers rapid increases in Fas and FasL expression in peripheral blood 50 as well as in thymus and other lymphoid tissues 51. During acute SIV infection, most T cells express Fas 52 and there is abundant local expression of FasL among intestinal lymphocytes 53. Ablation of intestinal lamina propria CD4+ cells during SIV infection can be attributed in part, to the Fas/FasL mediated apoptotic pathway but it is controversial whether apoptotic death of uninfected cells exceeds virus-mediated killing of infected cells. Detailed studies of viral burden and the rising proportion of SIV+ intestinal T cells argued that direct depletion of infected cells accounts for the rapid cell loss 52 without the need to invoke apoptotic killing of uninfected cells. These investigators stated that up to 60% of all memory T cells were infected but that those cells were lost within 4 days. They defined memory CD4+ cells as those that were CD45RA- or CD45RA+CD95+, meaning that Fas+ (CD95+) cells that are highly susceptible to apoptosis are included. Circulating lymphocytes and lymphocytes from lymph nodes and mucosa were sampled, but lymphocytes were not sampled from the spleen or liver, for example, so it is impossible to rule out margination as an explanation for lymphocyte loss. There is no way to determine from the experimental design, whether infected cells are being depleted faster than uninfected cells.
Another recent publication from Mattapallil's laboratory notes that early mucosal HIV/SIV infection (2-4 days after inoculation) demonstrates high levels of infected CD4+ Th-17 T cells that are depleted during the course of infection 54. Th-17 cells are proinflammatory effectors of antiviral immunity and are commonly suppressed by Treg, most likely being depleted via the Fas pathway. Once again it is not clear whether depletion is due to cell death in the category of direct depletion of virus-infected cells or indirect depletion of uninfected bystander cells. It is highly improbable that any one mechanism will explain all events in AIDS pathogenesis, so we expect both mechanisms to be operating during infection.
Several important studies have implicated apoptosis among uninfected cells as a major mechanism for leukocyte depletion. Fas ligation was a probable cause for apoptosis in T cells from SIV infected macaques 55. However, caspase-independent pathways for T cell apoptosis were thought to drive cell death in other SIV infection studies 56. Cell loss, especially in gut-associated lymphoid tissues, likely occurs by multiple mechanisms and we would expect depletion of both CD4+ and CD4- cell subsets at these loci of intense viral replication.
A recent publication described gene expression profiles at three stages of HIV infection: acute, chronic, and AIDS 57. The acute stage had high levels of FasL mRNA expression that were diminished during the chronic stage. This observation coincides with earlier published studies showing a high level of PBMC-susceptibility to AICD during the acute stage of SIV infection, and this susceptibility subsides during the chronic stage 58.
Curiously, infected cells are more resistant to apoptosis than uninfected cells 2359. The apoptosis resistance in persistently infected lymphoid and monocytic cells was shown to be independent of active viral production and involved a modulation of the mitochondrial pathway 60. A consequence is that indirect cell killing through apoptotic mechanisms like Fas/FasL, will destroy activated, uninfected cells while sparing the fraction of infected cells. Such a process will tend to increase the proportion of infected cells and diminish the proportion of uninfected cells in a tissue heavily burdened by SIV or HIV. Perhaps this mechanism contributes to the very high proportions of infected cells noted in macaque intestinal tissues after SIV infection 52 and may help to reconcile apparent differences between direct and indirect cell depletion models.
Substantial data have been accumulated on HIV induction of Fas or FasL, the susceptibility of PBMC from HIV patients to apoptotic cell death and the reversal of these conditions by HAART. When viremia was suppressed by effective HAART, CD4 cells in PBMC had significantly reduced apoptosis levels that correlated with increasing CD4 counts in blood even though lymphoid tissue FasL levels were unchanged 6162. Similar findings have been reported for SIV infections. However, the problem remains that susceptibility to apoptosis, measured in vitro, rises and falls with changes in viremia, making it difficult to separate direct from indirect killing mechanisms in terms of their relative contributions to disease progression. In murine systems, these problems are addressed readily by the use of knock-out mutations eliminating Fas, FasL or both molecules. For AIDS-related questions, the initial studies are done most appropriately in nonhuman primates using SIV or SHIV to establish persistent infection and then applying interventions to modulate the Fas/FasL pathway.
Using a recombinant humanized anti-FasL monoclonal antibody 63 we developed a protocol for treating rhesus macaques to interrupt the Fas/FasL system. Animals received a total of 5 injections of anti-FasL given once per week beginning 2 weeks before SIV-inoculation and finishing 2 weeks after virus inoculation. The pilot study showed no effect of anti-FasL on plasma viremia, but found increased virus-specific immunity and delayed disease among treated animals 64. A larger study using immunized macaques indicated that anti-FasL treatment preserved memory T cells, antigen responses after SIV infection, and was associated with decreased levels of Treg cells 65. In the two interventional studies, anti-FasL delivered before and during the acute infection had a durable effect on immune status and disease many months after treatment stopped. These changes were not reflected in vRNA levels that were similar among treatment and control groups, but were detected in the composition and activity of other T cell populations. This means that uninfected effector cells must have been preserved by the treatment.
Another indication of the importance for indirect cell killing comes from studies of "naturally-infected" macaques i.e., sooty mangabeys infected with SIVsmm that maintain plasma viremia, do not show high susceptibility to in vitro apoptosis among PBMC and remain disease-free 66. Here, apoptosis and immune activation were low and animals did not develop disease despite chronic viremia. The observation of preserved CD4 T lymphocytes with regenerative capacity in spite of high viremia, suggests that virus alone cannot explain the massive loss of CD4 lymphocytes that occurs in pathogenic SIV and HIV infections 67 and indirect cell killing mechanisms may be important.
We reported 68 that half of 56 macaques tested showed high levels of MHC-unrestricted cytolysis prior to SIV inoculation and that animals from this group were all among the rapid progressors. Subsequently, we found that MHC-unrestricted cytolysis involves the Fas/FasL pathway 64 and rapid progressors had higher baseline levels of cell killing through this pathway. In corroboration of these findings, high levels of lymph node cell apoptosis during acute SIV infection also predict that animals will become rapid progressors 69.
Reports showing lower levels of indirect cell killing/apoptosis among naturally-infected macaques despite similar viremia 646770, the correlation between apoptosis and rapid progression 69 and intervention studies showing disease-sparing after brief treatment with monoclonal antibody against FasL 6465 attest to strong relationships between apoptotic killing of uninfected cells and pathogenesis. The in vitro and tissue studies on HIV agree with these findings. Disagreements remain about the mechanisms for cell death, the roles for viral proteins and the relative importance of Fas versus non-Fas death pathways. In addition to direct infection and cell destruction by viral cytopathic effects or antigen-specific cytotoxicity, indirect cell killing mechanisms have broad impact on host immune capacity and are important in the pathogenesis of HIV/AIDS.