INTERFERENCE WITH THYMIC FUNCTION
Maintenance of a full peripheral CD4+ T cell pool may be hampered not only by HIV-induced peripheral T cell death but also by a lack of de novo T cell generation.33 Because thymopoiesis is generally thought to be the main source of T cell generation, it has been proposed that HIV-induced apoptosis of thymocytes in the presence of slightly increased—or even normal4—peripheral T cell death might explain why CD4+ T cell numbers are gradually lost during asymptomatic HIV infection.3,5,34 Interference with thymic function would also explain why HIV infection is associated with a gradual decline in both naive CD4+ and naive CD8+ T cells.35,36 However, current data do not unequivocally show that thymic output is severely decreased throughout HIV infection, and even if it was, recent data have brought into question the conclusion that this could explain a severe loss of CD4+ T cells in the periphery.
Experiments in mice and data from humans do suggest that HIV is able to infect the thymus. To analyze the effect of HIV infection on thymopoiesis in vivo, McCune and collaborators used the SCID-hu mouse model, in which human fetal liver and thymus are implanted under the kidney capsule of immunodeficient SCID mice.37,38 These mice show T lymphopoiesis for profound periods of time, and the implanted thymus is permissive for infection with HIV.39,40 Upon intrathymic HIV injection, infected cells and increased programmed cell death could be seen throughout the thymus, leading to a loss of CD4+CD8+ double-positive thymocytes, a decrease of the CD4+:CD8+ thymocyte ratio, and an overall reduction of thymocyte cellularity.3 Importantly, HIV infection in the SCID- hu mouse may give rise to more severe effects on the thymus than during natural HIV infection, because HIV in this model system is typically injected directly into the thymus.
Evidence for HIV- induced thymocyte loss is, however, not restricted to the SCID-hu mouse model, as it was also observed in some fetuses aborted from HIV-infected mothers,41 in thymus biopsies from HIV- infected children,42,43 in thymic tissue studied at autopsy from individuals who died of HIV infection,44 and in SIV-infected rhesus monkeys.45 It should be noted, however, that so far, mostly syncytium-inducing (SI) HIV isolates have been shown to interfere with thymopoiesis. When SCID-hu mice were infected with non-SI (NSI) HIV variants, changes in thymus composition occurred much more slowly,46,47 suggesting that loss of thymic output may only play a role in peripheral CD4+ T cell loss relatively late during disease progression. In a rhesus monkey model of HIV infection, loss of thymic output could be observed only at late stages of disease, just before the animals developed AIDS-like symptoms.48 However, because of a lack of direct tools to measure ongoing thymic function in vivo, it remains to be determined how abrogation of thymic function affects peripheral blood T cell numbers.To study more directly whether HIV infection impairs thymic output in humans, an ex vivo marker of recent thymic emigrants would be helpful. Measuring the number of T cells with a naive phenotype does not suffice to measure thymic output, because naive T cells are relatively long- lived and may retain their naive phenotype upon homeostatic proliferation. Recently, several laboratories have tried to quantify thymic output more directly by measuring T cell receptor excision circles (TRECs), the DNA by-products of T cell receptor gene rearrangements.49 As TRECs are extrachromosomal DNA circles, they are not copied during T cell division and are, therefore, only produced upon de novo T cell generation in the thymus. The fact that the average TREC content (i.e., the average number of TRECs per T cell) of CD4+ and CD8+ T cells decreases with age was taken to reflect the age-related involution of the thymus and thereby was taken as evidence that T cell TREC contents can be used as a reliable measure of thymic output.5,50 We showed, however, that TREC data need to be interpreted with great caution, because not only thymic output but also T cell death and proliferation influence the average TREC content of a T cell population.51,52
In HIV-infected individuals, the average TREC content of the naive CD4+ and CD8+ T cell population is typically decreased as compared to healthy controls. Taking the TREC content of the naive T cell population as a direct measure of thymic output, it has been argued that thymic output is severely reduced, even during the asymptomatic phase of HIV infection.
Using a mathematical model, we showed, however, that decreased TREC contents in HIV-infected individuals are much more likely due to increased T cell proliferation than to thymic impairment. In fact, because of the longevity of naive T cells, a decrease in thymic output hardly affects the TREC content of the naive T cell population.51 The TREC contents of peripheral T cells from thymectomized individuals indeed start to decrease only years after thymectomy.53 Because there is ample evidence for increased T cell proliferation during HIV infection, decreased TREC contents in HIV-infected individuals should thus not be taken as evidence for decreased thymic output.It remains to be determined to what extent infection of the thymus contributes to HIV disease progression. In fact, even if HIV infection of the thymus were to decrease the effective thymic output, it is not known to what extent that would affect the peripheral T cell population. In a small group of HIV-infected individuals who were thymectomized for myasthenia gravis before HIV infection, thymectomy did not preclude long-term survival. Increases in CD4+ T cell numbers when those thymectomized individuals were treated with HAART turned out to be similar to those in nonthymectomized HIV-infected individuals on HAART.44
Thymic output in adults was estimated to be low, on the order of 107 to 108 naive T cells per day.29 Because of this, naive T cell reconstitution after T cell depletion of various causes, such as bone marrow/stem cell transplantation, monoclonal antibody treatment,54-56 and HIV infection,25 is very slow, and thymectomy in HIV-negative individuals does not significantly affect peripheral blood T cell numbers.53 It can thus be expected that in adults, HIV-induced abrogation of thymic function will not significantly change peripheral blood T cell numbers. However, in children, in whom thymic function is optimal, HIV infection of the thymus may contribute significantly to HIV-induced naive T cell loss but only if thymic function plays a major role in daily T cell homeostasis.
To study this, we followed changes in total body T cell numbers and TREC numbers with age in a group of healthy children of different ages. We studied total body TREC numbers instead of TREC contents, because total TRECs are not influenced by T cell proliferation and are thus a more direct measure of thymic output than TREC contents.52 If the major source of T cells in young children were the thymus, one would expect the dynamics of T cell and TREC numbers to be similar. Instead, we found large discrepancies between the two. Whereas total body naive T cell numbers increased during the first years of life, total TREC numbers remained stable during this period, suggesting that the increase in naive T cell numbers was largely due to peripheral T cell proliferation.57 Thus, even in children, thymic output seems to be less critical in maintaining or increasing peripheral blood naive T cell numbers, and HIV-related impairment of thymic function may, therefore, not have a significant impact on peripheral blood T cell numbers. Even complete abrogation of thymic output in juvenile macaques that were thymectomized was shown to have little impact on the peripheral T cell compartment, both in healthy and in SIV-infected macaques.58’59Together, these data suggest that inappropriate apoptosis of maturing T cells in the HIV-infected thymus may hamper normal thymic function but that its effect on peripheral blood T cell numbers may not be significant, both in children and in adults.