APPROPRIATE T CELL DEATH: CHRONIC IMMUNE ACTIVATION IN HIV PATHOGENESIS

Because the moderate increase in CD4⁺ and CD8⁺ T cell turnover in HIV infection is not likely to be a homeostatic response to compensate for CD4⁺ T cell loss (earlier referred to as the “pull” hypothesis), the alternative “push” hypothesis has been put forward.

One of the strongest arguments for the idea that HIV induces increased T cell proliferation is that shortly after initiation of HAART, when viral reproduction is already prevented but CD4⁺ T cell numbers are still low, T cell proliferation decreases dramatically.³⁰ Moreover, increases in the expression of the proliferation marker Ki67 during HIV-1 infection are not confined to the CD4⁺ T cell pool and are, in fact, even more profound in the CD8⁺ T cell pool, which becomes depleted only in the final stages of HIV infection.²⁶’²⁷’²⁹’³⁰ In line with this observation, HIV-1-induced telomere shortening,²² T cell apoptosis,¹⁷’²⁰’²¹’⁶¹ and TREC dilution⁵¹ have been found to be most pronounced in the CD8⁺ T cell population.

If chronic immune activation is a key player in HIV-1 pathogenesis, one would expect the level of immune activation post-seroconversion to be associated with disease progression. High propor­tions of proliferating Ki67⁺CD4⁺ and Ki67⁺CD8⁺ T cells post-seroconversion and high levels of the activation markers CD70, CD38, and HLA-DR expressed by CD4⁺ and CD8⁺ T cells were found to be associated with rapid disease progression.^62-64 In fact, the level of immune activation turned out to be an even better predictor for disease progression than HIV viral load.^63,64 Interest­ingly, CD8⁺ T cell activation was found to correlate positively with HIV viral load,^63,65,66 which is strongly associated with the rate of disease progression.

Interestingly, we found that increased levels of immune activation before seroconversion are also predictive for disease progression after an individual becomes infected with HIV. In a study of homosexual men, we showed that low CD4⁺ T cell counts and high CD70 expression by CD4⁺ T cells before seroconversion were associated with fast disease progression upon HIV infection.⁶⁴ Similarly, HIV-1-infected injecting drug users with low CD4⁺ TREC contents pre-seroconversion showed a steeper decline in CD4⁺ T cell count throughout infection than those with high pre­seroconversion CD4⁺ TREC contents.⁶⁷ Thus, the level of immune activation before HIV-1 infection seems to affect disease progression.

Also, during HAART, the level of immune activation was shown to be correlated with changes in CD4⁺ T cell counts.^66,68 HAART causes a rapid decrease in plasma virus load, thereby alleviating chronic immune activation, leading to an immediate decrease in T cell proliferation and apoptosis levels. High residual levels of immune activation during HAART turned out to be associated with poor immune reconstitution, again underscoring the central role of immune activation in HIV infection.⁶⁶

Based on these data, the alternative paradigm emerged that hyperactivation of the immune system is the key factor in HIV pathogenesis and that increased levels of T cell apoptosis are a natural result of this hyperactivated state of the immune system.^7,60 High levels of apoptosis were not only described for HIV infection but also, for example, during the acute phase of immune response in mice infected with acute lymphocytic choriomeningitis virus (LCMV)⁶⁹ and humans infected with Epstein-Barr virus^70,71 or cytomegalovirus.⁷² Meanwhile, many additional data have been found to support this view.

In Vivo Labeling Studies

Recent T cell labeling studies using BrdU, deuterated glucose, and deuterated water have confirmed that T cell turnover is increased in both the CD4⁺ and CD8⁺ T cell population of HIV-infected individuals.²⁸,^73-76 Upon half-an-hour in vivo pulse-labeling of T cells from HIV-infected individuals with BrdU, Kovacs et al. identified at least two distinct T cell subpopulations, one characterized by a slow and the other by a faster rate of T cell turnover.⁷⁵ HIV infection turned out to increase the average turnover of CD4⁺ and CD8⁺ T lymphocytes by activating resting cells and driving these cells into the rapidly proliferating subpopulation. Interestingly, however, HIV did not affect the turnover of either of the two subpopulations, and HAART decreased the proportion of labeled cells in the rapidly proliferating pool but did not affect the decay rate of labeled cells.⁷⁵

Ribeiro et al. reported that after a 7-day infusion of deuterated glucose, an increased transition of CD4⁺ T cells from the resting to the activated compartment could be measured in HIV-infected individuals compared to healthy controls.⁷⁶ In contrast, Ribeiro et al. found an HIV-related increase in proliferation and death rates of activated CD4⁺ T cells.⁷⁵ This discrepancy between the two studies may be due to the different labeling techniques. Whereas cessation of label administration leads to an immediate decay of deuterium-labeled DNA, BrdU-labeled cells show much slower decay kinetics. Because BrdU measures the proportion of positive cells rather than label-positive DNA strands (as in the case of deuterium studies), cell division in the absence of label infusion can actually increase the amount of BrdU⁺ cells, because the label is divided equally over the daughter cells, rendering them BrdU⁺. Thus, the net decay rate of BrdU⁺ cells in the presence of continuous cell division will be lower than the actual death rate.⁷⁷

A general problem of in vivo labeling techniques such as those described here is that they are biased toward measuring the dynamics of the T cell population with the highest turnover.

Despite the above-mentioned shortcomings of these labeling techniques, these studies, which are the first to show in vivo T cell kinetics, confirmed the hypothesis that T cells that recently divided and, hence, picked up the label tend to have a relatively short life span.^7,60 The fact that recently divided cells in HIV-infected patients are prone to die argues against a homeostatic response to CD4⁺ T cell depletion and is consistent with the idea that T cell division and death are driven by immune activation.

HIV-Infected Individuals with Discordant Responses to Antiviral Therapy

Another argument in favor of the idea that immune activation is the key player in HIV infection comes from HIV-infected individuals who show discordant responses to antiviral therapy. Most patients treated with antiretroviral therapy show a significant decline in viral load concomitant with an increase in peripheral blood T cell numbers. However, some patients are only virologically responsive to antiviral therapy but do not show immunologic improvement (“virologic responders”), whereas others are responsive only in immunologic terms, showing increases in peripheral blood T cell numbers despite detectable viremia (“immunologic responders”).

HIV-2

Differences in the natural course of HIV-1 and HIV-2 infection have also been taken as evidence for a central role of immune activation in HIV infection. In HIV-2-infected patients, the asymp­tomatic phase takes much longer compared to HIV-1 infection, because CD4⁺ T cells decline much more slowly. HIV-2-infected patients have lower levels of T cell activation, proliferation, and apoptosis and a lower plasma viral load than HIV-1-infected individuals. When HIV-1- and HIV- 2-infected individuals were stratified on the basis of CD4⁺ T cell counts, the levels of T cell activation and proliferation were shown to be similar despite large differences in plasma viral load.^65,84 Sousa et al. interpreted this close association between immune activation and the CD4⁺ T cell count as evidence that the level of immune activation, and not viral load, determines disease outcome.⁶⁵ One could, however, also reason the other way around, namely, that the rate at which CD4⁺ T cells are lost, which is known to be higher in HIV-1 compared to HIV-2 infection, should apparently be due to differences in HIV viral load, because it is the only apparent difference between HIV-1- and HIV-2-infected individuals with similar CD4⁺ T cell counts.

SIV

If in HIV infection high levels of immune activation are associated with a bad prognosis, one would also expect the reverse, namely, that low levels of immune activation is associated with a good prognosis. The latter situation seems to be the case in some SIV-infected primates. In sooty mangabeys, one of the natural hosts of SIV, CD4⁺ T cells are typically not depleted, despite high viral loads. Both immune activation and T cell apoptosis were shown to be unaffected in these animals when compared to noninfected sooty mangabeys. Sooty mangabeys may have been selected in the presence of SIV to give only a limited immune response against the virus. Transmission of SIV to other primate species that are not natural hosts to SIV, such as macaques, is known to induce CD4⁺ T cell decline and progression to AIDS. SIV infection of rhesus macaques is characterized by increased levels of immune activation.¹³ In vivo labeling of dividing T cells using BrdU confirmed these findings by showing that in SIV-infected macaques, the turnover of both CD4⁺ and CD8⁺ T cells was increased compared with uninfected macaques.⁷⁷ These data support the view that low levels of T cell activation are associated with a good prognosis, regardless of the presence of high levels of virus load, again underscoring the central role of chronic immune activation in HIV and SIV pathogenesis.

Collectively, these data changed our viewpoint on HIV infection drastically. Where once it was thought that “inappropriate apoptosis” was the main problem in HIV infection and increased T cell proliferation was a response to cope with this loss of cells, much research is now based on the opposite idea that chronic T cell activation is driving T cell loss through a natural process of “appropriate apoptosis” of activated T cell clones.