T cells containing T cell receptor excision circles are inversely related to HIV replication and are selectively and rapidly released into circulation with antiretroviral treatment

Objective: To examine baseline predictors of T-cell receptor rearrangement excision circle (TREC) levels and their changes during treatment with combined antiretroviral therapy.

Methods: Peripheral blood and lymph node lymphocytes were examined for the presence of TREC by real-time polymerase chain reaction and circulating lymphocyte phenotypes were examined by flow cytometry. Correlates for CD4 and CD8 cell TREC levels at baseline were identified among CD4 and CD8 immunophenotypes, viral load and patient demographics; the significance of TREC changes after initiation of antiretroviral therapy was assessed.

Results: Circulating TREC levels correlated inversely with age, with HIV RNA levels, with activation markers on circulating T cells and with naive CD4 but not CD8 cell frequencies. With initiation of antiretroviral therapy, TREC and naive T cell frequencies increased in peripheral blood during the first 2 weeks of treatment and these changes correlated negatively with TREC frequencies in lymph node aspirates, particularly among CD8 T cells.

Conclusions: These findings suggest that recent thymic emigrants are sequestered in lymphoid tissue during uncontrolled HIV replication and are selectively released into circulation rapidly after initiation of antiretroviral therapies.

Introduction

T cell receptor excision circles (TREC) are episomal fragments generated by rearrangement of T cell receptor genes during thymic maturation [¹]. As these episomes do not replicate with cellular division, their frequency is also diminished by cellular replication.

This study examines TREC levels in peripheral blood and in lymph node aspirate lymphocytes in patients treated with antiretroviral therapy for the first time.

Method

Forty-four antiretroviral treatment-naive patients participating in a randomized trial comparing highly active antiretroviral therapy (HAART) with three doses of atazanavir or nelfinavir plus stavudine and didanosine. TREC levels were measured in peripheral blood and in lymph node aspirate lymphocytes from a subset of subjects. The institutional review board at each study site approved the protocol and patients provided written informed consent before participation. Because immunological responses in the treatment groups were comparable, the data were pooled. The patient group comprised 13 women and 31 men with a median age 36 years [interquartile range (IQR), 14]. Median baseline plasma HIV RNA level was 5.0 log₁₀ copies/ml (IQR, 1) while median CD4 cell count was 280 × 10⁶ cells/l (IQR, 158). Lymphocyte subsets were measured by flow cytometry [²] while TREC cells were measured using real-time polymerase chain reaction [³] in cryopreserved lymphocytes after cells were positively selected by immunomagnetic bead sorting.

Because of skewness in the variables, Spearman rank correlation coefficient was used to identify possible predictors of the proportion of CD4 and CD8 cells containing TREC. These potential predictors included CD4 and CD8 immunophenotypes, plasma viral load and patient demographics (age and sex). Statistical significance of changes in TREC levels between consecutive visits was assessed with the Wilcoxon signed rank test. Since this is a post-hoc analysis, these results must be considered exploratory.

Results

Table 1 shows the patient characteristics.

Table 1

At baseline, both age (Fig. 1a) and blood HIV RNA (Fig. 1b) negatively correlated with TREC-containing CD4 cell proportions (r = -0.34; P = 0.02 for age; r = -0.41, P < 0.01 for HIV RNA). Baseline HIV RNA levels also were negatively correlated with TREC-containing CD8 cell proportions (r = -0.42; P < 0.01; Fig. 1f). TREC proportions correlated with the numbers of naive phenotype CD4 cells (r = 0.43; P < 0.01; Fig. 1c) but not naive CD8 cells (r = -0.01; P = 0.94; Fig. 1g). This was likely a consequence of sustained CD8 cell expansion in HIV disease as absolute TREC-containing CD8 cell numbers did not correlate with absolute CD8 cell numbers (Fig. 1e).

Fig. 1

The proportions of CD4 and CD8 cells expressing CD38 correlated negatively with the proportions of TREC-containing CD4 (r = -0.39; P = 0.01; Fig. 1d) and CD8 (r = -0.50; P < 0.01; Fig. 1h) cells, respectively. When adjusted for their relationships with CD38 expression, the correlation between HIV RNA levels and TREC-containing cell levels decreased (r = -0.41 to -0.29 for CD4 cells, and r = -0.42 to -0.31 for CD8 cells) consistent with a model wherein the effects of HIV replication on TREC levels are at least in part mediated through immune activation.

After 2 weeks of antiretroviral therapy, significant increases were seen in the proportions of TREC-containing CD4 cells (from 0.020 to 0.037; P = 0.01; n = 32) and CD8 cells (from 0.009 to 0.017; P < 0.01; n = 32), and in absolute TREC-containing cells (P < 0.01). Thereafter, both proportions and absolute numbers of TREC-containing cells decreased but not significantly (Fig. 2). Stratification by median CD4 cell counts at baseline revealed that patients with lower CD4 cell counts (< 280 × 10⁶ cells/l) did not experience a significant increase in TREC-containing CD4 cell proportions.

Fig. 2

Lymph nodes were aspirated in a small subset of patients (n = 7) and TREC-containing CD4 and CD8 cells were measured at baseline and after 2 weeks of therapy (Fig. 3). Changes in the TREC proportions between baseline and week 2 in blood correlated negatively with the same changes in lymph nodes, suggesting that TREC-containing cells are selectively released from lymph nodes to the blood. This association was significant for the CD8 compartment (r = -0.929; P < 0.01) but not for CD4 cells (r = -0.214; P = 0.65).

Fig. 3

Discussion

TREC are episomal fragments generated during thymic maturation by rearrangement of T cell receptor genes [¹]. These episomes do not replicate with cellular division and their frequency is diminished by cellular replication. The present study is consistent with previous findings for TREC. First, as others have found, the number of TREC-containing cells correlates inversely with age [^1,4,5] and viral load [⁶]; and correlates positively with CD4 cell counts [⁶] and the proportion of naive T cells [⁷]. Similarly, as has been reported, immune activation decreases the frequency of circulating TREC-containing cells [⁴]. Since early cell increases in patients taking HAART have been thought to represent a redistribution of memory phenotype lymphocytes trapped within lymphoid tissue, while later cellular restoration is thought to reflect increases in thymic production of naive cells [^8,9], prior studies have focused on changes in TREC levels during the second phase of cellular restoration during HAART [^1,5,7]. The results of the present study indicate that TREC-containing recent thymic emigrants are selectively sequestered in lymphoid tissue [¹⁰] perhaps as a consequence of virus-induced inflammation [¹¹]. Importantly, our data also indicate that suppression of HIV replication rapidly (within 2 weeks) results in selective release of these cells into circulation. Consequently, our data indicate that naive T cell restoration in treated HIV disease also has two phases, a rapid first-phase redistribution [²] that selectively releases recent thymic emigrants into circulation and a later progressive increase that may represent restoration of thymic function [¹].

Acknowledgements

The authors thank the patients who participated in this study.

Sponsorship: This work was supported by Bristol-Myers Squibb, UHC Ireland Cancer Center (Statistical Core), and Grants AI36219 and AI38858 from the US National Institutes of Health.