Abstract

Immune reconstitution after antiretroviral therapy in human immunodeficiency virus (HIV)–infected patients may result from the recovery of thymus function, peripheral redistribution, or decreased T cell destruction. This study investigated levels of T cell receptor gene rearrangement excision circles (TRECs) as a measure of recent thymic emigrant cells in peripheral blood lymphocytes of 50 HIV-infected infants and children who were followed-up for 40 months after the start or change of antiretroviral therapy. At baseline, patients exhibited fewer TRECs than did uninfected control subjects. The increase in TRECs after antiretroviral therapy was greater in infants than in older HIV-infected children. Of interest, patients who demonstrated discordant responses (i.e., increased CD4 T cell counts without significant virologic suppression) also had substantial gains in TRECs. Furthermore, TRECs correlated positively with the number of CD4 and naive T cells and negatively with age and virus load. Measurement of TRECs may serve as a useful tool for evaluating immune reconstitution in HIV-infected children receiving antiretroviral therapy

Human immunodeficiency virus (HIV) infection is associated with progressive immune deficiency characterized by a persistent decline in CD4 T cells accompanied by impaired CD4 T cell function [1–5 ]. Treatment with highly active antiretroviral combination therapies often results in significant reduction in plasma HIV RNA levels and increases in CD4 T cell counts [6–11 ]. Several mechanisms have been proposed for the loss of CD4 T cells, including excessive destruction, decreased production, and lymphocyte redistribution [12–16 ]. Excessive destruction has been attributed to HIV-induced cytopathicity, the killing of infected cells by cytotoxic T cells, and bystander cell death of uninfected T cells by apoptosis [5, 13, 17–19 ]. A decrease in progenitor cells and impaired thymus function have been implicated in decreased generation of new T cells [20, 21]. In adults, the recovery of T cells after antiretroviral therapy initially involves cells with the memory phenotype and is followed by increases in cells with the naive phenotype [22, 23]. On the basis of these observations, the initial gain in T cells after therapy has been attributed mainly to the release of cells trapped in lymph nodes [24], and the subsequent increase of T cells is attributed to newly generated cells from the thymus [25–27 ]. As virus replication is controlled, the T cells are spared from HIV-induced cell death [28, 29]

Much attention recently has been placed on the role of the thymus in HIV disease pathogenesis. The work of Smith et al. [30] and McCune et al. [31] demonstrated that thymus size, as determined by computerized tomography, was increased in certain HIV-infected adults. These investigators suggested that the thymus was actively involved in the production of T cells, to counter the excessive T cell destruction due to HIV infection. The development by Douek et al. [32] of a novel method to analyze recent thymic emigrants by quantitation of T cell receptor excision circles (TRECs) in peripheral T cells has provided a specific tool for investigating the role of the thymus in the pathogenesis of HIV [32]. In brief, TRECs are DNA excision circles that are produced during formation of a functional T cell receptor gene by a series of splicings, first of D to J and then of V to the resulting DJ. In this way, V, D, and J segments are selected for use with a constant C region, which gives the receptor its unvarying transmembrane and cytoplasmic domains [33]. Having been cut from the genome, the excised DNA circles can no longer be replicated. Instead, the circles are diluted when the cells undergo mitosis; with each cellular division, a TREC passes to one or the other of the 2 daughter cells

In several studies, an increase in CD4-naive T cells and a decrease in plasma virus load in response to antiretroviral therapy were shown to correlate with TREC levels in HIV-infected individuals [32, 34]. Although there is universal agreement that TRECs are decreased in HIV disease and increase after therapy, the role of the thymus in the immunopathogenesis of HIV disease, vis-à-vis the loss and recovery of T cells, remains controversial. Whether the levels of TRECs accurately represent thymus activity or are predominantly reflecting peripheral events that influence T cells (e.g., proliferation and apoptosis) is under debate. Nevertheless, TREC analysis is considered to be a feasible and practical means of identifying cells of recent thymic origin

In the present study, we evaluated TREC levels in the peripheral blood of HIV-infected children receiving antiretroviral therapy. Children’s response to therapy differs from that of adults, in that early recovery of T cells with the naive phenotype is more common [10]. We investigated the role of the thymus in the reconstitution of T cells by determining TREC levels in peripheral blood lymphocytes. The results were evaluated in the context of immunologic and virologic responses after antiretroviral therapy

Patients and Methods

PatientsWe studied 50 HIV-infected children during regularly scheduled visits to the Pediatric Immunology Clinic at North Shore University Hospital (Manhasset, NY). Patient characteristics (age, virus load, CD4 cell counts, treatment history, and response to antiretroviral therapy) are shown in table 1. Of the 50 children, 16 were receiving antiretroviral therapy for the first time; the remaining 34 children were treatment experienced and had switched therapy at least once. The time of starting or changing therapy was designated as baseline. Median age at baseline was 9.0 years (range, 0.2–19.2 years). Median plasma virus load at baseline was 4.7 log10 HIV RNA copies/mL (range, 3.1–5.9 log10 HIV RNA copies/mL). Median CD4 T cell count at baseline was 439 cells/μL (range, 0–3713 cells/μL)

Table 1

Age, virus load, CD4 cell counts, treatment history, and response to antiretroviral therapy for 50 human immunodeficiency virus (HIV)–infected children

Table 1

Age, virus load, CD4 cell counts, treatment history, and response to antiretroviral therapy for 50 human immunodeficiency virus (HIV)–infected children

Forty-three children were receiving antiretroviral therapy regimens, which included triple or quadruple drug combinations containing ⩾1 classes of drugs: nucleoside or nucleotide reverse-transcriptase inhibitors (RTIs; zidovudine, zalcitabine, didanosine, or abacavir); nonnucleoside RTIs (nevirapine or efavirenz), and protease inhibitors (PIs; nelfinavir, saquinavir, ritonavir, or amprenavir). The other 7 children were receiving 1 or 2 RTIs

Immunologic and virologic responses to antiretroviral therapy in the study cohort were determined on the basis of changes in plasma virus load and percentage of CD4 T cells. Children who had an increase in percentage of CD4 T cell by ⩾10 from baseline or those who maintained a CD4 cell percentage >25% were designated as immunologic responders (IR+). Children were defined as virologic responders (VR+) if they demonstrated a decline in virus load of ⩾1.5 log10 from baseline or attained undetectable levels (i.e., <400 HIV RNA copies/mL). Children who did not meet the requisite criteria were considered to be immunologic nonresponders (IR) or virologic nonresponders (VR). On the basis of the immunologic and virologic responses, the patients were collectively designated as responders (IR+VR+), nonresponders (IRVR), or discordant (IRVR+ or IR+VR, signifying that one response was present and other was absent). In the discordant group, most patients were in the IR+VR category. Because the IRVR+ type of responses were rare, analysis of discordant patients was restricted to those with the IR+VR pattern. Group designations were made at the times of TREC analysis, after the children had received therapy for ⩾3 months

TREC levels were determined at multiple time points (minimum, 2; maximum, 12; median, 4) over a period of ⩽40 months after the start or change of therapy for 50 patients, resulting in a total of 250 determinations. We longitudinally studied TRECs, including baseline evaluations before the start or change of therapy, for 22 children (including 5 infants)

ImmunophenotypingBlood samples were collected in acid citrate dextrose and were stained for 4-color flow cytometry with fluorochromes (fluorescein isothiocyanate, R-phycoerythrin, peridinin-chlorophyll protein, and allophycocyanin)–labeled monoclonal antibody combinations for CD45RA, CD62L, CD3, CD8, and CD4 (Becton Dickinson). A 100-μL aliquot of whole blood was incubated with the antibody combination for 10 min at room temperature, followed by red cell lysis and fixation, using 1% paraformaldehyde. Cells were stored at 4°C before analysis by flow cytometry (EPICS ELITE ESP; Coulter). A total of 5000 cells was analyzed in a manually set lymphocyte gate. Positive cut-off for fluorescence was set to include <2% of negative control mouse IgG conjugates. CD4 and CD8 T cells simultaneously positive for CD45RA and CD62L antigen were designated as naive T cells

Isolation of peripheral blood mononuclear cells (PBMC) and CD4 and CD8 T cell purificationPBMC were isolated from heparinized venous blood by ficoll-metrizoate (Lymphoprep; Nyegard) density gradient centrifugation. CD4 and CD8 T cells were positively selected by using magnetic beads coated with anti-CD4 and anti-CD8 monoclonal antibodies (Dynal), according to the manufacturer’s instructions

Quantification of TRECsDNA was extracted from PBMC or CD4 or CD8 T cells by proteinase K digestion. TRECs were quantified by real-time polymerase chain reaction (PCR) analysis, using the 5′ nuclease (TaqMan) assay [35] and the ABI Prism 7700 sequence detector system (PE Biosystem). In brief, 0.5 μM primers (CACATCCCTTTCAACCATGCT and GCCAGCTGCAGGGTTTAGG), 0.25 μM TaqMan probe FAM-ACACCTCTGGTTTTTGTAAAGGTGCCCACT-TAMRA, 0.5 U of platinum Taq polymerase, 3.5 mM MgCl2, 0.2 mM dNTPs in TaqMan buffer (PE Biosystem), and 100–250 ng of genomic DNA from PBMC were mixed together in a 25-μL PCR mixture. PCR conditions were 95°C for 5 min, 95°C for 30 s, and 60°C for 1 min for 50 cycles. A standard curve was established with known copies of plasmids containing signal joint TREC fragment (a gift of D. Douek, Division of Infectious Diseases, University of Texas Southwestern Medical Center, Dallas), and TREC values for test samples were calculated with the software provided with the ABI Prism 7700 system. Each sample was run in triplicate, and mean TREC values were used for data analysis. Results were expressed as TRECs/106 cells

During the process of assay standardization, we compared TRECs from cord blood lymphocytes and from PBMC samples from healthy subjects and from patients with primary immunodeficiency diseases, namely severe combined immunodeficiency and DiGeorge syndrome. TREC assay results in a subset of the latter group of patients and healthy control subjects were comparable with results obtained with the same samples by Zhang et al. [26]. In addition, we stimulated TRECs in cord blood and PBMC from healthy subjects for 7 days with phytohemagglutinin (5μg/mL) and interleukin-2 (100 U/mL) and compared them with cells cultured without exogenous stimuli, thereby confirming that TRECs decrease with cell division (data not shown)

Plasma HIV RNA quantitationHIV RNA copies in plasma were estimated by use of the Amplicor HIV monitor kit (Roche Diagnostics), according to the manufacturer’s protocol. Sensitivity of the assay was <400 or <50 HIV RNA copies/mL

Statistical analysisStatistical software (SigmaStat 2.0; Jandel Scientific Software) was used for statistical analysis. To determine the difference between the groups, we used the Wilcoxon signed&amp;rank test. The relationship between different variables was determined by Spearman’s&amp;rank order correlation. The pairs of variables with positive Spearman’s correlation coefficients (rs) and P<.05 tend to increase together. For the pairs with negative rs and P<.05, one variable tends to decrease, whereas the other increases

Results

TRECs in PBMC and CD4 and CD8 T cellsWe first investigated TREC levels in CD4 and CD8 T cells and their relationship with TRECs in PBMC. Levels of TRECs were estimated from DNA isolated from PBMC and magnetic bead–purified CD4 and CD8 T cells from 31 HIV-infected children. As shown in figure 1A TRECs in PBMC correlated strongly with TRECs in purified CD4 (rs=.78; P<.000001) and CD8 (rs=.72; P<.000001) T cells. Collectively, as shown in figure 1B mean (±SD) TREC levels were significantly higher (P<.05) in CD4 (31,195±32,327/106 cells) than in CD8 (12,497±12,528/106 cells) T cells and were intermediate in PBMC (15,351±17,370/106 cells). The ratio between TRECs in CD4 T cells to TRECs in CD8 T cells was significantly higher than the peripheral CD4:CD8 T cell ratio (mean±SD, 2.8±1.51 and 1.1±0.9, respectively, P<.00001; figure 1C)

Figure 1

Relationship between T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) and TRECs in purified CD4 and CD8 T cells. TRECs were estimated by real-time polymerase chain reaction assay from DNA isolated from PBMC and from purified CD4 and CD8 T cells. A Correlation between TRECs in PBMC with TRECs in CD4 and CD8 T cells, as analyzed by Spearman’s correlation (rs). B Box plots of TRECs present in purified CD4 and CD8 T cells and in PBMC and comparisons between them based on paired t test. C Ratio of TRECs in purified CD4 T cells and TRECs in purified CD8 T cells, compared with peripheral CD4:CD8 T cell ratio by paired t test

Figure 1

Relationship between T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) and TRECs in purified CD4 and CD8 T cells. TRECs were estimated by real-time polymerase chain reaction assay from DNA isolated from PBMC and from purified CD4 and CD8 T cells. A Correlation between TRECs in PBMC with TRECs in CD4 and CD8 T cells, as analyzed by Spearman’s correlation (rs). B Box plots of TRECs present in purified CD4 and CD8 T cells and in PBMC and comparisons between them based on paired t test. C Ratio of TRECs in purified CD4 T cells and TRECs in purified CD8 T cells, compared with peripheral CD4:CD8 T cell ratio by paired t test

Quantitation of PBMC TRECs levels and relationship to virus load and T cell subsetsResults of TRECs from 250 PBMC samples are shown in figure 2. TRECs were 10–233,400 copies/106 PBMC (median, 11,154 copies/106 PBMC). The strength of association of TRECs with the different parameters that were assessed at the same time as TRECs were analyzed by use of the nonparametric Spearman’s&amp;rank order correlation. Levels of TRECs in PBMC correlated negatively with virus load (rs=-.39; P<.00001; figure 2A) and positively with the CD4 T cell percentage (rs=.49; P<.00001; figure 2B) and with CD4 T cell counts (rs=.5; P<.00001; data not shown). A weak negative correlation was noted with the CD8 T cell percentage (rs=-.22; P<.005; data not shown) and the CD8 T cell counts (rs=.2; P=.002; data not shown). The naive CD4 and CD8 T cell percentages, characterized as CD45RA+CD62L+, also correlated strongly with PBMC TREC levels (CD4: rs=.47; P<.00001; CD8: rs=.48; P<.00001; figure 2E2D respectively)

Figure 2

Correlation between T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) with different parameters based on evaluation of 250 samples obtained from 50 human immunodeficiency virus (HIV)–infected children receiving antiretroviral therapy. TRECs were correlated with plasma virus load (A) percentage of CD4 T cells (B) percentage of CD45RA+CD62L+ in CD4 T cells (C) and percentage of CD45RA+CD62L+ in CD8 T cells (D). Data were assessed by Spearman’s&amp;rank order correlation, with rs representing regression coefficient

Figure 2

Correlation between T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) with different parameters based on evaluation of 250 samples obtained from 50 human immunodeficiency virus (HIV)–infected children receiving antiretroviral therapy. TRECs were correlated with plasma virus load (A) percentage of CD4 T cells (B) percentage of CD45RA+CD62L+ in CD4 T cells (C) and percentage of CD45RA+CD62L+ in CD8 T cells (D). Data were assessed by Spearman’s&amp;rank order correlation, with rs representing regression coefficient

PBMC TREC levels in relation to response to antiretroviral therapyCollectively, TRECs had a negative correlation (rs=-.3; P=10-5) with the age of study subjects (figure 3A). Figure 3B shows, in the context of immunologic and virologic responses, the correlation between patient age and TREC copies per 106 PBMC after the start or change of therapy. Patients were classified as responders, nonresponders, or discordant after starting or changing therapy. TREC levels correlated negatively with age in responders (rs=-.39; P=.02) and nonresponders (rs=-.3; P=.03), but TREC levels in nonresponders were lower than those in responders at all ages. No significant correlation of TRECs with age was noted in the discordant group (rs=-.23; P not significant; data not shown)

Figure 3

Correlation between age and T cell receptor gene rearrangement excision circle (TREC) copies per 106 peripheral blood mononuclear cells (PBMC). A TRECs in the entire cohort in relation to age. B TRECs in patients designated as responders and nonresponders in relation to age. The correlation between age and TRECs was analyzed by Spearman’s&amp;rank (rs) order correlation analysis

Figure 3

Correlation between age and T cell receptor gene rearrangement excision circle (TREC) copies per 106 peripheral blood mononuclear cells (PBMC). A TRECs in the entire cohort in relation to age. B TRECs in patients designated as responders and nonresponders in relation to age. The correlation between age and TRECs was analyzed by Spearman’s&amp;rank (rs) order correlation analysis

TREC levels in the 3 response groups also were evaluated in relation to duration of follow-up. Data for TRECs at periods of 6–12 (n=38), 13–24 (n=41), and >24 (n=26) months after the start or change of therapy in the 3 response groups are shown in figure 4. The responder group had significantly higher TRECs than did the nonresponder group at all study points. In the discordant patients, the TRECs were not significantly different from those for responders at any time and were significantly higher than those for nonresponders at 6–12 and 13–24 months after baseline (P<.05)

Figure 4

Box plot of T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) of patients showing different responses to therapy at 6–12, 13–24, and >24 months after starting or changing therapy. Comparative analysis of TREC levels among responders, nonresponders, and discordant patient groups was analyzed by unpaired t test

Figure 4

Box plot of T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) of patients showing different responses to therapy at 6–12, 13–24, and >24 months after starting or changing therapy. Comparative analysis of TREC levels among responders, nonresponders, and discordant patient groups was analyzed by unpaired t test

The relationship of the magnitude of TREC response to that of treatment response in the study cohort was evaluated among 38 children 6–12 months after the start or change of therapy. TREC responses were grouped as <1000, 1000–10,000, or >10,000 TREC copies/106 PBMC (table 2). The highest TREC levels were present more often in responders, and the lowest TREC levels were present more often in nonresponders. Of 18 children with >10,000 TREC copies/106 PBMC (median, 41,606 TREC copies/106 PBMC), 12 were in the responder group and 6 were in the discordant group. Of the 11 children who had 1000–10,000 TRECs (median, 5460 TRECs), 3 were responders, 4 were nonresponders, and the remaining 4 were discordant. Of 9 children who had <1000 TREC copies/106 PBMC (median, 338 copies/106 PBMC), 7 were nonresponders and 2 were discordant

Table 2

Immunologic (IR) and virologic (VR) responses to therapy in human immunodeficiency virus–infected children in relation to T cell receptor gene rearrangement excision circle (TREC) levels 6–12 months after therapy

Table 2

Immunologic (IR) and virologic (VR) responses to therapy in human immunodeficiency virus–infected children in relation to T cell receptor gene rearrangement excision circle (TREC) levels 6–12 months after therapy

Postantiretroviral therapy TREC levels in relation to baseline levelsTREC levels were determined before and after the start or change of antiretroviral therapy for 22 HIV-infected children, 5 of whom were 0.2–0.8 years old (median, 0.26 years) and 17 of whom were 2.6–16.7 years old (median, 9.5 years) at baseline. Figure 5 shows the average TREC values at a median of −1.17 months (range, −6 to 0 months) before and at a median of 8.6 months (range, 6–12 months) after the start or change in therapy. Before a start or change in therapy, TRECs were significantly lower in HIV-infected infants (P<.01) than in HIV-exposed, uninfected infants. In addition, TREC levels in older children were lower than those reported for age-matched, uninfected control subjects

Figure 5

T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) before and after 6–12 months of antiretroviral therapy in infants (A) and older children (B). Human immunodeficiency virus (HIV)–infected children were evaluated at baseline and after starting or changing therapy and were grouped as responders (R), discordant (D), or nonresponders (NR). TREC values of HIV-exposed, uninfected infants (EU) and age-matched controls (C) for infants and older children are shown for comparison. Comparison between TRECs at baseline and after a start or change of therapy was analyzed by paired t test

Figure 5

T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC) before and after 6–12 months of antiretroviral therapy in infants (A) and older children (B). Human immunodeficiency virus (HIV)–infected children were evaluated at baseline and after starting or changing therapy and were grouped as responders (R), discordant (D), or nonresponders (NR). TREC values of HIV-exposed, uninfected infants (EU) and age-matched controls (C) for infants and older children are shown for comparison. Comparison between TRECs at baseline and after a start or change of therapy was analyzed by paired t test

All 5 infants were classified as responders, and, of the 17 older children, 6 were responders, 5 were nonresponders, and the remaining 6 were discordant. Compared with baseline values, mean (±SD) TREC levels for infants and older children in the responder group were significantly increased (78,369±68,308 and 16,128±22,864 copies/106 cells, respectively; P<.01 for each). Compared with TREC increases for older responder children, increases for infants were significantly greater (P<.05), and levels reached those for HIV-exposed, uninfected infants

Nonresponder children had no significant increases in TRECs. Of interest, children with discordant responses who showed increases in CD4 T cell counts without adequate virologic suppression manifested significant increases in TREC levels (P<.05)

A comparative analysis of changes in TREC levels for 22 patients with detectable (n=14) and undetectable (n=8) virus loads revealed that changes in TRECs were significantly greater in patients without detectable virus loads (P=.004; figure 6A). The strength of association between changes in TREC levels and CD4 T cells as a result of antiretroviral therapy in 17 older HIV-infected children was determined by using the nonparametric Spearman’s&amp;rank order correlation. Change in TRECs correlated with changes in CD4 T cell percentages (rs=.5; P<.05; figure 6B) and in CD4 T cell counts (rs=.2; P<.05; data not shown)

Figure 6

Changes in T cell receptor gene rearrangement excision circle (TREC) levels from baseline in relation to virologic response and change in percentage of CD4 T cells. A Changes in TRECs in patients with detectable and undetectable (>400 and ⩽400 human immunodeficiency virus RNA copies/mL, respectively) virus load are shown in the box plots and were analyzed by unpaired t test. B Correlation between changes in TREC levels and changes in percentage of CD4 T cells, as analyzed by Spearman’s&amp;rank order correlation (rs)

Figure 6

Changes in T cell receptor gene rearrangement excision circle (TREC) levels from baseline in relation to virologic response and change in percentage of CD4 T cells. A Changes in TRECs in patients with detectable and undetectable (>400 and ⩽400 human immunodeficiency virus RNA copies/mL, respectively) virus load are shown in the box plots and were analyzed by unpaired t test. B Correlation between changes in TREC levels and changes in percentage of CD4 T cells, as analyzed by Spearman’s&amp;rank order correlation (rs)

Figure 7 shows the levels of TRECs, plasma virus loads, and CD4 T cell percentages sequentially studied in 3 representative children, 1 each from the responder, discordant, and nonresponder categories. In the responder patient, virus load decreased, and CD4 T cell percent and TRECs concurrently increased. A child with a discordant response showed a minimal decrease in virus load but had substantial increases in CD4 T cell percentage and TRECs. In a nonresponder child whose virus load was not controlled, the CD4 T cell percentage and TRECs declined

Figure 7

Longitudinal measurement of levels of T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC), percentage of CD4 T cells, and plasma virus load in 3 representative human immunodeficiency virus (HIV)–infected children from responder, discordant, and nonresponder categories. X-axis, time in months under investigation; Y1-axis, percentage of CD4 T cells; Y2-axis, TREC levels and plasma virus load

Figure 7

Longitudinal measurement of levels of T cell receptor gene rearrangement excision circles (TRECs) in peripheral blood mononuclear cells (PBMC), percentage of CD4 T cells, and plasma virus load in 3 representative human immunodeficiency virus (HIV)–infected children from responder, discordant, and nonresponder categories. X-axis, time in months under investigation; Y1-axis, percentage of CD4 T cells; Y2-axis, TREC levels and plasma virus load

Discussion

HIV infection leads to a qualitative and quantitative loss in T cells, and several different mechanisms have been implicated in this loss [1, 12, 16, 17, 36–39 ]. The damage caused by HIV recently was shown to include destruction not only of CD4 T cells but also of the intricate microenvironments required for lymphocyte proliferation peripherally and thymopoiesis centrally [40–43 ]. The relative contributions of the deleterious effects of HIV on the thymus, vis-à-vis those on the peripheral lymphocyte pool in the ensuing CD4 T cell deficiency, remain unclear. In the present study, we investigated the levels of TRECs in peripheral T cells as a measure of thymus function in HIV-infected children starting or changing antiretroviral therapy. TREC levels were evaluated in the context of immunologic and virologic response to antiretroviral therapy

Our findings show that patients who respond immunologically and virologically have the highest levels of TRECs and that the gain in TRECs is greater in infants than in older children. In patients manifesting discordant responses (i.e., those with increases in CD4 cell counts without appreciable or durable virologic suppression), the CD4 cell count increase also was associated with increases in TRECs. These findings, together with the observation that the CD4:CD8 ratio of TREC-containing T cells was normal, suggest that thymic output of T cells has an important role in the immunologic reconstitution resulting from antiretroviral therapy in HIV-infected children

Several studies have identified effects of HIV within the thymus [42, 44, 45]. For example, there have been reports of destruction of thymic architecture and limited thymopoietic potential in patients with later-stage HIV disease, compared with a relatively low frequency of productively infected cells and relatively preserved histology in early infection [14, 46, 47]. Studies in the SCID-Hu mouse model have indicated that HIV infects and depletes thymocytes double-positive for CD4 and CD8 molecules, and studies in the rhesus monkey–simian immunodeficiency virus model have indicated that infection results in inhibition of thymic precursors [48]. In agreement with other reports, the levels of TREC-containing T cells in our study population correlated inversely with age and plasma virus load and positively with CD4 T cells and naive CD4 and CD8 T cell subpopulations [25, 26, 32]. These findings provide indirect evidence for the involvement of the thymus in HIV pathogenesis and suggest that thymus output has a role in the decrease of CD4 T cells, as well as in their recovery after treatment

A recent study [34] refuted the contention that TRECs are reflective of thymus activity in patients with HIV infection and provided alternative explanations for the observed changes in TRECs. The authors contended that the decline in TRECs observed in HIV infection and their rapid increase after potent antiretroviral therapy are better explained by changes in the rate of peripheral T cell division and not by HIV-associated thymic impairment and recovery after antiretroviral therapy. These investigations suggest that the loss of TRECs in HIV infection is merely a reflection of a dilution of TREC-containing cells that occurs because of increased cell division resulting from persistent hyperactivation of the immune system by HIV. Increase in TRECs after potent antiretroviral therapy was attributed to a decrease in T cell division

This explanation does not adequately explain many of the observations made in our study and those of others. Although there is agreement that T cell activation and proliferation is an important component of HIV pathogenesis, this feature by itself does not explain the observed findings. The proportion of TREC-containing CD4 and CD8 T cells in peripheral circulation was estimated in the present study by measuring the TREC levels in magnetically purified CD4 and CD8 T cells. The mean ratio of CD4 TRECs and CD8 TRECs was 2.9, which is similar to or greater than that expected in peripheral T cells of healthy HIV-uninfected individuals. If T cell proliferation was the sole factor influencing TREC levels, then there should have been fewer CD8 TRECs, since it is well known that T cell turnover in CD8 T cells is far greater than that in CD4 T cells and would have led to further dilution of CD8 T cell TRECs

The potential role of the thymus in modulating TREC levels in HIV infection also is supported by the observed changes in TREC levels in patients who were discordant in their responses to therapy (i.e., those with increases in CD4 T cell counts without appreciable virologic suppression). TREC responses in these patients were similar to those in patients who showed good virologic suppression. Thus, despite ongoing virus replication, their TRECs were not affected by the presumed concurrent immune activation. Although this observation needs to be substantiated by analysis of lymphocyte proliferation and activation, it supports the likelihood that thymus output contributed to the CD4 T cell count increases in these patients. In addition, others have shown that, immediately after potent antiretroviral therapy, there is increased proliferation of T cells in association with an increase of naive and memory T cell pool [31, 49, 50]. Such activity would be expected to lead to a decrease in TRECs by dilution rather than to an increase

These findings collectively argue in favor of a role of the thymus in contributing, at least partly, to the changes in TREC levels in peripheral T cells. The positive correlation between rate of change in TREC levels with the rate of change in peripheral CD4 T cell count over the same time period also provides strong supportive evidence for thymus output as a cause for increases in CD4 T cell counts

TRECs in HIV-infected infants were decreased, compared with those in HIV-exposed, uninfected infants, but after therapy they increased to levels present in HIV-exposed, uninfected infants. Compared with older children, infants manifested greater gains in TRECs after therapy. These findings are in agreement with the concept of a greater thymus reserve in infancy and preservation of the immune system with institution of therapy early in the course of infection. In older children as well, the disparity in TREC levels in responder and nonresponder children was noted >24 months after a start or change of treatment. In all instances, TREC levels in children showing discordant increases in CD4 T cell counts were similar to those observed in responders who showed both immunologic and virologic responses, and they were significantly higher than levels in nonresponders for 12–24 months after therapy

These observations suggest that HIV-induced immunologic damage is reversible, particularly in infants. The mechanism(s) whereby CD4 T cell increase and presumed thymus regeneration occur in patients who do not show a virologic response to therapy is unclear. It is possible that treatment that is inefficient in controlling virus replication can nevertheless prevent adverse effects on the immune system, including the thymus, by other mechanisms. Among the possible causes for the prevention of immunologic injury are those related to altered viral fitness or pathogenicity [51, 52]. Alternatively, ⩾1 antiretroviral drugs may have beneficial effects on the immune system [53, 54]

One factor that might impact TREC levels that has received insufficient attention is the role of T cell apoptosis in modulating levels of TREC. The report of Hazenberg et al. [34] did not take into consideration the role that increased T cell activation has on induction of apoptosis. It is possible that TREC levels may be influenced either by apoptosis of peripheral T cells or by apoptosis of TREC-containing T cells themselves. In the former case, TREC levels would be expected to increase because of enrichment, whereas in the latter situation, TRECs would be expected to decrease. We and others have shown that HIV infection is associated with increased peripheral T cell apoptosis, which potentially contributes to the decrease in peripheral T cell pool [13, 17, 18, 28, 55]. Controlling the virus load by potent antiretroviral therapy reduced apoptosis in the peripheral T cell compartment, and the decrease in T cell apoptosis was associated with a quantitative increase of CD4 T cell counts [28, 55]. Recent findings from our laboratory have revealed a strong correlation between levels of TRECs and apoptosis in peripheral T cells (rs=.87; P<.0001; authors’ unpublished data)

Overall, the studies reported here indirectly suggest that increases in TRECs after the initiation of potent antiretroviral therapy are reflective of production of new T cells from the thymus. These findings, however, do not refute the possibility that peripheral expansion of T cells is also operative in maintaining CD4+ T cell homeostasis. Further studies that can directly measure TREC-containing T cells for proliferation and apoptosis and evaluation of the half-life of TRECs are needed to clarify the significance of using TRECs as a marker for thymus output in HIV disease

Acknowledgments

We thank Maria Marecki and Heriberto Borrero (North Shore University Hospital, Manhasset, New York) for excellent technical assistance, D. Douek (Division of Infectious Diseases, University of Texas Southwestern Medical Center, Dallas) for the generous gift of T cell receptor gene rearrangement excision circles (TRECs) containing plasmids, and L. Zhang (Aaron Diamond Institute, Rockefeller University) for TREC data of control subjects and immunodeficient patients

References

1
Picker
LJ
Maino
VC
The CD4+ T cell response to HIV-1
Curr Opin Immunol
 , 
2000
, vol. 
12
 (pg. 
381
-
6
)
2
Perelson
AS
Essunger
P
Ho
DD
Dynamics of HIV-1 and CD4+ lymphocytes in vivo
AIDS
 , 
1997
, vol. 
11
 
Suppl A
(pg. 
S17
-
S24
)
3
Anderson
RW
Ascher
MS
Sheppard
HW
Direct HIV cytopathicity cannot account for CD4 decline in AIDS in the presence of homeostasis: a worst-case dynamic analysis
J Acquir Immune Defic Syndr Hum Retrovirol
 , 
1998
, vol. 
17
 (pg. 
245
-
52
)
4
Beltz
L
Thymic involution and HIV progression [letter]
Immunol Today
 , 
1999
, vol. 
20
 pg. 
429
 
5
Rosenberg
YJ
Anderson
AO
Pabst
R
HIV-induced decline in blood CD4/CD8 ratios: viral killing or altered lymphocyte trafficking
Immunol Today
 , 
1998
, vol. 
19
 (pg. 
10
-
7
)
6
Autran
B
Carcelaint
G
Li
TS
, et al.  . 
Restoration of the immune system with anti-retroviral therapy
Immunol Lett
 , 
1999
, vol. 
66
 (pg. 
207
-
11
)
7
Clark
DR
de Boer
RJ
Wolthers
KC
Miedema
F
T cell dynamics in HIV-1 infection
Adv Immunol
 , 
1999
, vol. 
73
 (pg. 
301
-
27
)
8
Gibb
DM
Newberry
A
Klein
N
de Rossi
A
Grosch-Woerner
I
Immune repopulation after HAART in previously untreated HIV-1–infected children. Paediatric European Network for Treatment of AIDS (PENTA) Steering Committee [letter
Lancet
 , 
2000
, vol. 
355
 (pg. 
1331
-
2
)
9
Oxenius
A
Price
DA
Easterbrook
PJ
, et al.  . 
Early highly active antiretroviral therapy for acute HIV-1 infection preserves immune function of CD8+ and CD4+ T lymphocytes
Proc Natl Acad Sci USA
 , 
2000
, vol. 
97
 (pg. 
3382
-
7
)
10
Essajee
SM
Kim
M
Gonzalez
C
, et al.  . 
Immunologic and virologic responses to HAART in severely immunocompromised HIV-1–infected children
AIDS
 , 
1999
, vol. 
13
 (pg. 
2523
-
32
)
11
Franco
JM
Leon-Leal
JA
Leal
M
, et al.  . 
CD4+ and CD8+ T lymphocyte regeneration after anti-retroviral therapy in HIV-1–infected children and adult patients
Clin Exp Immunol
 , 
2000
, vol. 
119
 (pg. 
493
-
8
)
12
Li
TS
Tubiana
R
Katlama
C
Calvez
V
Ait
MH
Long-lasting recovery in CD4 T-cell function and viral-load reduction after highly active antiretroviral therapy in advanced HIV-1 disease
Lancet
 , 
1998
, vol. 
351
 (pg. 
1682
-
6
)
13
Rowland-Jones
S
HIV infection: where have all the T cells gone
Lancet
 , 
1999
, vol. 
354
 (pg. 
5
-
7
)
14
Goldberg
B
Stricker
RB
Apoptosis and HIV infection: T-cells fiddle while the immune system burns
Immunol Lett
 , 
1999
, vol. 
70
 (pg. 
5
-
8
)
15
Gougeon
ML
Ledru
E
Lecoeur
H
Garcia
S
T cell apoptosis in HIV infection: mechanisms and relevance for AIDS pathogenesis
Results Probl Cell Differ
 , 
1998
, vol. 
24
 (pg. 
233
-
48
)
16
Frost
SD
Michie
CA
Lymphocyte dynamics, apoptosis and HIV infection
Trends Microbiol
 , 
1996
, vol. 
4
 (pg. 
77
-
82
)
17
Kabelitz
D
Pohl
T
Oberg
HH
Pechhold
K
Dobmeyer
T
Apoptosis of mature T lymphocytes: putative role in the regulation of cellular immune responses and in the pathogenesis of HIV infection
Prog Mol Subcell Biol
 , 
1996
, vol. 
16
 (pg. 
58
-
71
)
18
Kirschner
DE
Mehr
R
Perelson
AS
Role of the thymus in pediatric HIV-1 infection
J Acquir Immune Defic Syndr Hum Retrovirol
 , 
1998
, vol. 
18
 (pg. 
95
-
109
)
19
Koka
PS
Jamieson
BD
Brooks
DG
Zack
JA
Human immunodeficiency virus type 1–induced hematopoietic inhibition is independent of productive infection of progenitor cells in vivo
J Virol
 , 
1999
, vol. 
73
 (pg. 
9089
-
97
)
20
Mitsuyasu
R
HIV protease inhibitors: immunological insights
AIDS
 , 
1999
, vol. 
13
 
Suppl 1
(pg. 
S19
-
27
)
21
Hellerstein
MK
McCune
JM
T cell turnover in HIV-1 disease
Immunity
 , 
1997
, vol. 
7
 (pg. 
583
-
9
)
22
Mezzaroma
I
Carlesimo
M
Pinter
E
, et al.  . 
Long-term evaluation of T-cell subsets and T-cell function after HAART in advanced stage HIV-1 disease
AIDS
 , 
1999
, vol. 
13
 (pg. 
1187
-
93
)
23
Fleury
S
Pantaleo
G
T cell regeneration in HIV-infected subjects under highly active antiretroviral therapy
Int J Mol Med
 , 
1999
, vol. 
4
 (pg. 
91
-
7
)
24
Douek
DC
Koup
RA
McFarland
RD
Sullivan
JL
Luzuriaga
K
Effect of HIV on thymic function before and after antiretroviral therapy in children
J Infect Dis
 , 
2000
, vol. 
181
 (pg. 
1479
-
82
)
25
Jaworowski
A
Crowe
SM
Does HIV cause depletion of CD4+ T cells in vivo by the induction of apoptosis
Immunol Cell Biol
 , 
1999
, vol. 
77
 (pg. 
90
-
8
)
26
Zhang
L
Lewin
SR
Markowitz
M
, et al.  . 
Measuring recent thymic emigrants in blood of normal and HIV-1–infected individuals before and after effective therapy
J Exp Med
 , 
1999
, vol. 
190
 (pg. 
725
-
32
)
27
Kitchen
SG
Killian
S
Giorgi
JV
Zack
JA
Functional reconstitution of thymopoiesis after human immunodeficiency virus infection
J Virol
 , 
2000
, vol. 
74
 (pg. 
2943
-
8
)
28
Chavan
SJ
Tamma
SL
Kaplan
M
Gersten
M
Pahwa
SG
Reduction in T cell apoptosis in patients with HIV disease following antiretroviral therapy
Clin Immunol
 , 
1999
, vol. 
93
 (pg. 
24
-
33
)
29
Ensoli
F
Fiorelli
V
Alario
C
, et al.  . 
Decreased T cell apoptosis and T cell recovery during highly active antiretroviral therapy (HAART)
Clin Immunol
 , 
2000
, vol. 
97
 (pg. 
9
-
20
)
30
Smith
KY
Valdez
H
Landay
A
, et al.  . 
Thymic size and lymphocyte restoration in patients with human immunodeficiency virus infection after 48 weeks of zidovudine, lamivudine, and ritonavir therapy
J Infect Dis
 , 
2000
, vol. 
181
 (pg. 
141
-
7
)
31
McCune
JM
Hanley
MB
Cesar
D
, et al.  . 
Factors influencing T-cell turnover in HIV-1–seropositive patients
J Clin Invest
 , 
2000
, vol. 
105
 (pg. 
R1
-
R8
)
32
Douek
DC
McFarland
RD
Keiser
PH
, et al.  . 
Changes in thymic function with age and during the treatment of HIV infection
Nature
 , 
1998
, vol. 
396
 (pg. 
690
-
5
)
33
Nemazee
D
Receptor selection in B and T lymphocytes
Annu Rev Immunol
 , 
2000
, vol. 
18
 (pg. 
19
-
51
)
34
Hazenberg
MD
Otto
SA
Stuart
JW
, et al.  . 
Increased cell division but not thymic dysfunction rapidly affects the T-cell receptor excision circle content of the naive T cell population in HIV-1 infection
Nat Med
 , 
2000
, vol. 
6
 (pg. 
1036
-
42
)
35
Livak
KJ
Flood
SJ
Marmaro
J
Giusti
W
Deetz
K
Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization
PCR Methods Appl
 , 
1995
, vol. 
4
 (pg. 
357
-
62
)
36
Meyaard
L
Miedema
F
Immune dysregulation and CD4+ T cell loss in HIV-1 infection
Springer Semin Immunopathol
 , 
1997
, vol. 
18
 (pg. 
285
-
303
)
37
Rizza
P
Santini
SM
Logozzi
MA
, et al.  . 
T-cell dysfunctions in hu-PBL-SCID mice infected with human immunodeficiency virus (HIV) shortly after reconstitution: in vivo effects of HIV on highly activated human immune cells
J Virol
 , 
1996
, vol. 
70
 (pg. 
7958
-
64
)
38
Shearer
GM
HIV-induced immunopathogenesis
Immunity
 , 
1998
, vol. 
9
 (pg. 
587
-
93
)
39
Sheppard
HW
Ascher
MS
The relationship between AIDS and immunologic tolerance
J Acquir Immune Defic Syndr
 , 
1992
, vol. 
5
 (pg. 
143
-
7
)
40
Iida
T
Ichimura
H
Shimada
T
, et al.  . 
Role of apoptosis induction in both peripheral lymph nodes and thymus in progressive loss of CD4+ cells in SHIV-infected macaques
AIDS Res Hum Retroviruses
 , 
2000
, vol. 
16
 (pg. 
9
-
18
)
41
Kaneshima
H
Shih
CC
Namikawa
R
, et al.  . 
Human immunodeficiency virus infection of human lymph nodes in the SCID-hu mouse
Proc Natl Acad Sci USA
 , 
1991
, vol. 
88
 (pg. 
4523
-
7
)
42
Fauci
AS
Pantaleo
G
Stanley
S
Weissman
D
Immunopathogenic mechanisms of HIV infection
Ann Intern Med
 , 
1996
, vol. 
124
 (pg. 
654
-
63
)
43
Schnittman
SM
Singer
KH
Greenhouse
JJ
, et al.  . 
Thymic microenvironment induces HIV expression: physiologic secretion of IL-6 by thymic epithelial cells up-regulates virus expression in chronically infected cells
J Immunol
 , 
1991
, vol. 
147
 (pg. 
2553
-
8
(erratum: J Immunol 1992; 148:981)
44
Quijano
G
Siminovich
M
Drut
R
Histopathologic findings in the lymphoid and reticuloendothelial system in pediatric HIV infection: a postmortem study
Pediatr Pathol Lab Med
 , 
1997
, vol. 
17
 (pg. 
845
-
56
)
45
Rosenzweig
M
Clark
DP
Gaulton
GN
Selective thymocyte depletion in neonatal HIV-1 thymic infection
AIDS
 , 
1993
, vol. 
7
 (pg. 
1601
-
5
)
46
Neben
K
Heidbreder
M
Muller
J
, et al.  . 
Impaired thymopoietic potential of immature CD3CD4+CD8 T cell precursors from SIV-infected rhesus monkeys
Int Immunol
 , 
1999
, vol. 
11
 (pg. 
1509
-
18
)
47
Haynes
BF
Markert
ML
Sempowski
GD
Patel
DD
Hale
LP
The role of the thymus in immune reconstitution in aging, bone marrow transplantation, and HIV-1 infection
Annu Rev Immunol
 , 
2000
, vol. 
18
 (pg. 
529
-
60
)
48
Goldstein
H
Pettoello-Mantovani
M
Katopodis
NF
Kim
A
Yurasov
S
A model for studying disseminated HIV infection
Semin Immunol
 , 
1996
, vol. 
8
 (pg. 
223
-
31
)
49
Fleury
S
Rizzardi
GP
Chapuis
A
, et al.  . 
Long-term kinetics of T cell production in HIV-infected subjects treated with highly active antiretroviral therapy
Proc Natl Acad Sci USA
 , 
2000
, vol. 
97
 (pg. 
5393
-
8
)
50
Hellerstein
M
Hanley
MB
Cesar
D
, et al.  . 
Directly measured kinetics of circulating T lymphocytes in normal and HIV-1–infected humans
Nat Med
 , 
1999
, vol. 
5
 (pg. 
83
-
9
)
51
Martinez-Picado
J
Savara
AV
Sutton
L
D’Aquila RT
Aquila
RT
Replicative fitness of protease inhibitor–resistant mutants of human immunodeficiency virus type 1
J Virol
 , 
1999
, vol. 
73
 (pg. 
3744
-
52
)
52
Miyaura
M
Yoshida
A
Sakurai
A
Fujita
M
Koyama
AH
Mutational analysis of HIV-1 gag proteins
Int J Mol Med
 , 
2000
, vol. 
6
 (pg. 
265
-
9
)
53
Andre
P
Groettrup
M
Klenerman
P
, et al.  . 
An inhibitor of HIV-1 protease modulates proteasome activity, antigen presentation, and T cell responses
Proc Natl Acad Sci USA
 , 
1998
, vol. 
95
 (pg. 
13120
-
4
)
54
Weichold
FF
Bryant
JL
Pati
S
Barabitskaya
O
Gallo
RC
HIV-1 protease inhibitor ritonavir modulates susceptibility to apoptosis of uninfected T cells
J Hum Virol
 , 
1999
, vol. 
2
 (pg. 
261
-
9
)
55
Bohler
T
Debatin
KM
Linde
R
Sensitivity of CD4+ peripheral blood T cells toward spontaneous and CD95 (APO-1/Fas)–induced apoptosis in pediatric human immunodeficiency virus infection [letter]
Blood
 , 
1999
, vol. 
94
 (pg. 
1829
-
33
)
Presented in part: 7th Conference on Retroviruses and Opportunistic Infections, San Francisco, 30 January to 2 February 2000 (abstract 323); Immunology 2000, Seattle, 12–16 May 2000 (abstract 3932)
Financial support: National Institutes of Health (AI-28281, AI-48857, and DA-05161)
Informed consent was obtained from patients or their guardians, and human experimentation guidelines of the North Shore University Hospital were followed in the conduct of clinical research
Present affiliation: Brooklyn Hospital Center, Brooklyn, New York