Increased thymic output in HIV-negative patients after antiretroviral therapy

Objective: To determine the effects of antiretroviral therapy on thymic output independent of HIV infection.

Methods: Thymic output was evaluated by quantifying signal joint T-cell receptor (TCR) recombination excision circles in peripheral blood lymphocytes from HIV-negative patients undergoing prophylactic antiretroviral therapy. Additionally, effects of the HIV protease inhibitor nelfinavir were assessed in vivo on TCR-induced death of murine double-positive thymocytes.

Results: Five out of seven HIV-negative patients undergoing prophylactic antiretroviral therapy exhibited a dramatic increase (1-3 log₁₀) in recent thymic emigrants containing signal joint TCR recombination excision circles while their peripheral T cell compartments remained relatively unaffected. None of the patients developed subsequent HIV infections. Interestingly, nelfinavir did not have significant effects on TCR-induced apoptosis of murine thymocytes in vivo.

Conclusion: Antiretroviral therapy augments thymic output independent of HIV. Furthermore, nelfinavir does not dramatically affect TCR-induced thymocyte death in mice, thus central tolerance remains intact.

Introduction

With the advent of antiretroviral therapy (ART) to treat HIV infections, varying degrees of immune reconstitution have been achieved in patients. Routinely, T cell counts rebound in response to ART, and such a recovery may reflect homeostatic expansion of existing T cells and/or production of new T cells from the thymus. Importantly, complete immune reconstitution requires thymic production of naive T cells with new antigen specificities in order to replenish the HIV-depleted T cell repertoire. Several recent studies of HIV-positive individuals undergoing ART document increases in naive T cell counts as measured by T-cell receptor (TCR) recombination excision circles (TREC) ^[1-4]. However, it is not clear if ART enhances thymic output by reducing viremia, by acting on the thymus directly and independently of HIV, or both.

The extent of immune reconstitution after HIV infection may ultimately be limited by the fact that the virus has devastating effects on the thymus. HIV directly infects double positive thymocytes ^[5-7], single positive thymocytes ^[6-8], and thymic stromal cells ^[9-12]. In addition to directly infecting and killing cells in the thymus, HIV indirectly disturbs thymocyte development by altering the thymic microenvironment. In fact, far more thymocytes die during HIV infection than are directly infected by the virus ^[13]. Consequently, thymic output is dramatically reduced in HIV positive patients ^[14,15]. However, patients who respond favorably to ART tend to have increased thymic mass and produce more naive T cells as measured by phenotypic markers and TREC levels [^2,16,^1-4]. Furthermore, ART has been reported to expand the TCR repertoire in severely immunocompromised HIV patients, which suggests that thymic output is augmented by treatment ^[17]. These observations raise questions as to whether or not antiretroviral drugs and protease inhibitors that comprise ART affect the thymus directly or indirectly by inhibiting viral replication. HIV protease inhibitors block apoptosis in uninfected T cells by antagonizing the mitochochondrial permeability transition pore ^[18-20]. Additionally, it has been observed that individuals who maintain high viral loads after ART may still exhibit increases in naive T cells ^[20]. The issue is controversial, however. Antiretroviral drugs may select for strains of HIV that are less pathogenic to the thymus ^[21,22]. Additional evidence disputing the concept that ART acts directly on the thymus is the observation that the HIV protease inhibitor indinavir failed to boost T cell production in uninfected Scid-hu mice ^[23]. The present study addresses these issues by examining the effects of ART on thymic output in HIV negative patients. In addition, the effects of HIV protease inhibitors on thymocyte survival were examined in an in vivo murine model of TCR-induced apoptosis. Our results suggest that HIV protease inhibitors have great potential to promote immune reconstitution in clinical settings without compromising self-tolerance.

Methods

Blood specimens from eight healthy adult volunteers who were initiating prophylactic therapy following potential HIV exposure were collected. Post-exposure prophylaxis (PEP) consisted of nelfinavir 1250 mg twice a day, orally and combivir (zidovudine, 300 mg; lamivudine, 150 mg) 1 tablet twice a day for 28 days. Blood was collected prior to the initiation of PEP, during PEP, and 4-6 weeks after PEP was initiated ^[24]. Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation and frozen at -80°C until analysis. This project was reviewed and approved by the Ottawa Hospital Research Ethics Board and the Mayo Foundation Institutional Review Board. DNA was isolated from PBMC with the Easy DNA kit (Invitrogen, Carlsbad, California, USA) and assessed for signal joint (sj) TREC content relative to genomic CCR5 copies by real-time PCR in a spectrofluorometric thermal cycler (ABI PRISM 7700, PE Applied Biosystems, Foster City, California, USA). sjTREC reactions contained 25 pmol primers (forward 5′-CACATCCCTTTCAACCATGCT-3′, reverse 5′-GCCAGCTGCAGGGTTTAGG-3′), 125 nM TaqMan probe (5′-FAM-ACACCTCTGGTTTTTGTAAAGGTGCCCACT-TAMRA-3′), 1 × TaqMan Universal PCR Master Mix (PE Applied Biosystems), and 60 ng template DNA in a total volume of 50 μl. CCR5 reactions contained 25 pmol primers (forward 5′-GTGTCAAGTCCAATCTATGACATCAA-3′, reverse 5′-GCCTGCGATTTGCTTCACA-3′), 125 nM TaqMan probe (5′-FAM-TATTATACATCGGAGCCCTGCCAAAAAATCA-TAMRA-3′), 1 × TaqMan Universal PCR Master Mix, and 60 ng template DNA in a total volume of 50 μl. Thermal cycling conditions consisted of a 2-min incubation at 50°C, and an initial denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 10 min and 60°C for 1 min. For each run, standard curves were generated with 1-100 000 copies of human sjTREC plasmid (provided by D. Douek, NIH, Bethesda, Maryland, USA) or human CCR5 plasmid (cloned by reverse transcription-PCR into pCI vector from Promega, Madison, Wisconsin, USA) to calculate copies of TRECs versus CCR5. TREC values were expressed as TREC copy number per two copies of CCR5 or TRECs/PBL.

Immunophenotyping was performed on whole blood collected at the indicated time points and analyzed by FACS. Fluorochrome-conjugated antibodies specific for CD4, CD8, CD38, CD45RA, CD62L, CD45RO, CD38, and HLA-DR were obtained from BD Pharmingen (San Diego, California, USA). All samples were analyzed on an Epics ELITE flow cytometer ^[24].

Protease inhibitor treatment in mice consisted of 125 mg/kg nelfinavir (Agouron Pharmaceuticals, La Jolla, California, USA) and 13 mg/kg ritonavir (Abbott Pharmaceuticals, Abbott Park, Illinois, USA) in 2% ethanol. The protease inhibitors or vehicle control were administered by oral gavage to 4-6-week-old Balb/c mice every 8 h for 48 h. This low dose of ritonavir inhibits CYP3A metabolism of nelfinavir and effectively increases nelfinavir plasma levels to those achieved in humans receiving a standard dose of nelfinavir ^[25]. Eight-hour nelfinavir trough levels ranged from 1199 to 1258 ng/ml ^[25]. Anti-CD3 (145-2C11) was administered to mice i.p. at 24 h (150 μg) or 40 h (100 μg) after the first dose of protease inhibitors. Forty-eight hours after the first dose of protease inhibitors, thymocytes were harvested and stained with fluorescein isothiocyanate anti-mouse CD4 (H129.19, BD Pharmingen), phycoerythrin anti-mouse CD8a (53-6.7, BD Pharmingen), and 0.5 μg/ml propidium iodide (Molecular Probes, Eugene, Oregon, USA) and then analyzed on a FACScan flow cytometer with CELLQUEST software (Beckton Dickinson, San Jose, California, USA).

Results

To determine the effects of ART on thymic output independent of HIV, we analyzed thymic output in healthcare workers undergoing prophylactic ART after exposure to HIV. Four to six weeks after initiation of treatment with nelfinavir and combivir, five out of seven patients exhibited a dramatic increase (1-3 logs) in the proportion of sjTREC positive cells relative to PBMC (Fig. 1a). The Signed-Rank Test demonstrated statistically significant increases in log values of sjTRECs/PBMC before and after ART (P < 0.023). Control subjects who were not exposed to HIV and received no therapy failed to show significant changes in sjTRECs/PBMC over a 3-week period (Fig. 1a). Interestingly, substantial increases in sjTREC levels after ART were observed in patients of all ages (age range, 24-61 years).

Fig. 1

The observation that sjTREC levels increased in patients undergoing ART suggests that treatment enhances thymic output or survival of naive T cells in the periphery. However, T cell homeostasis appeared to be unaffected or minimally affected by treatment (Fig. 1b, c, d, e), which implicates enhanced thymic output as the primary mechanism responsible for the increases in sjTREC levels observed in patients undergoing ART. Specifically, total numbers of CD4 and CD8 T cells did not significantly change during ART (> 7 days of treatment) or after ART (7-14 days after a 28-day treatment) (Fig. 1b). A small but statistically significant increase in percentages of naive CD4 T cells (CD45RA,CD62L) was observed in patients during ART, which is consistent with the rise in sjTRECs (Fig. 1c). However, such a small increase in naive CD4 T cells could not account for the profound increases in sjTRECs/PBMC observed after ART. In addition, percentages of naive CD8 T cells (CD45RA,CD62L) were unaffected by ART (Fig. 1c). Similarly, proportions of memory (CD45RO) and activated (CD38, HLA-DR) T cells remained relatively stable during and after ART (Fig. 1d and e). These observations suggest that peripheral T cell homeostasis was not dramatically affected by ART, and that the observed increases in sjTRECs/PBMC were most likely attributable to effects of ART on thymopoiesis.

Thus far, no data have been presented to show that HIV protease inhibitors influence thymocyte survival independent of HIV. Therefore, we sought to examine more specifically the effects of nelfinavir on TCR-induced death of murine thymocytes. When mice are injected with anti-CD3 (2C11), double positive thymocytes die rapidly due to direct TCR stimulation and indirect stimulation from activated cells in the periphery. This model is used to mimic negative selection and provided a system in which to analyze the effects of nelfinavir on central tolerance. We found that nelfinavir does not inhibit double positive thymocyte death after administration of 2C11 nor is it toxic to thymocytes in vivo at doses that approximate a therapeutic dose in humans ^[25] (Fig. 2). Thus, negative selection appeared to be unaffected by nelfinavir treatment.

Fig. 2

Discussion

It is of great potential value to develop novel therapies aimed at replenishing the T cell compartment. To that end, we have examined the effects of ART on thymic output independent of HIV infection. Of the seven healthcare workers undergoing prophylactic ART after exposure to HIV, five exhibited a dramatic increase (1-3 log₁₀) in sjTRECs/PBMC. The present study also examines potential mechanisms of ART responsible for increasing numbers of sjTREC-positive cells in the periphery. If ART promotes thymocyte survival, more thymocytes will survive to exit into the periphery as TREC-positive recent thymic emigrants. Because the maximal size of the peripheral T cell compartment is fixed, recent thymic emigrants will displace existing T cells, leading to an increase in measured values of sjTRECs/PBMC. Alternatively, if ART preferentially enhances survival of peripheral T cells, the total peripheral T cell compartment will expand. However, global changes in numbers or percentages of T cell subsets in the periphery were not observed in patients that had received ART. Consequently, we conclude that the observed increases in sjTRECs/PBMC can be explained by effects of ART on thymopoiesis not T cell homeostasis. Together, the data indicate that the net effect of ART is an increase in thymic output and a rejuvenation of the T cell compartment.

ART may augment thymic output by promoting the survival of developing thymocytes. In fact, a precedent for anti-apoptotic effects of HIV protease inhibitors is well documented in the literature. In contrast, reverse transcriptase inhibitors are not thought to possess anti-apoptotic properties. A variety of data now indicate that HIV protease inhibitors, including nelfinavir, are intrinsically anti-apoptotic in a number of in vitro settings (reviewed in ^[20]). In addition, it has recently been demonstrated that nelfinavir confers anti-apoptotic effects in an in vivo model of mouse sepsis, where treatment with nelfinavir protects against lymphoid apoptosis and consequently enhances survival.

The proposal that HIV protease inhibitors protect thymocytes from apoptosis raises concerns that ART may undermine central tolerance. However, our results indicate that nelfinavir does not hamper thymocyte death under conditions resembling negative selection. The implication is that central tolerance is not dramatically compromised after treatment. Biochemical analyses suggest that nelfinavir inhibits apoptosis at the level of the mitochondria. Specifically, nelfinavir inhibits cytochrome c release and consequently impacts downstream caspase activation ^[18-20]. Although incompletely understood, negative selection can occur independent of the classical mitochodrial pathway ^[26,27]. Therefore, it is not surprising that nelfinavir did not influence negative selection.

Furthermore, experience in humans confirms that ART does not compromise tolerance, as autoimmunity is a rare phenomenon in patients undergoing ART ^[28]. Even if subtle effects on central tolerance are observed, peripheral tolerance will remain in place to counteract autoimmunity. Mechanisms of peripheral tolerance such as anergy, negative costimulation, and suppression by regulatory T cells are not dependent on apoptosis and should not be compromised by antiapoptotic effects of ART.

The fact that sjTREC levels rose after ART in HIV-negative patients while the peripheral T cell compartment remained unchanged, suggests that treatment promotes thymocyte survival. We have specifically examined TCR-induced death of double positive thymocytes and found that nelfinavir does not significantly affect negative selection. Additional work will be required to determine if HIV protease inhibitors prevent thymocyte death at alternative stages of thymocyte development or after death stimuli not associated with negative selection (glucocorticoids, death receptors, or deprivation of survival factors). Overall, our findings suggest that administration of HIV protease inhibitors may promote immune reconstitution in various clinical settings without compromising self-tolerance.

Acknowledgements

We thank Nathan R. Foster (Mayo Clinic Cancer Center Statistics) for statistical analyses.

Sponsorship: Supported by A144959 (DJM) and Burrroughs Wellcome Translational Research Award (ADB).