Immune reconstitution in human immunodeficiency virus type 1-infected children with different virological responses to anti-retroviral therapy

Abstract

Immune repopulation, despite virological failure, often occurs in children under highly active anti-retroviral therapy (HAART). The aim of this study was to analyse the characteristics of immune repopulation and activation in children with and without virological response to HAART. Fourteen human immunodeficiency virus type 1 (HIV-1)-infected children with suppression of HIV-1 plasma viraemia (virological responders, VR) and 16 virological non-responders (VNR) to therapy were studied at baseline and after approximately 2 years of HAART. During therapy, CD4⁺ T cells increased in both groups, but were higher in the VR than in the VNR group. All CD4⁺ T cell subsets (naive, central memory, effector/memory and CD38⁺) increased significantly in VR children, while there was a significant increase only in naive cells in VNR children. Naive CD8⁺ T cells and T cell receptor rearrangement excision circles (TREC), an indicator of thymic output, increased in both VR and VNR children. Activated CD8⁺CD38⁺ T cells decreased in VR but remained high in VNR children. Levels of circulating lipopolysaccharide (LPS), an indicator of microbial translocation, further increased in VNR children. In conclusion, HAART induced an increase in naive cells in all children, regardless of their virological response. However, the persistence of viraemia resulted in an impaired expansion of memory CD4⁺ T cells susceptible to HIV-1 infection, and together with the microbial translocation sustained the persistence of a high level of immune activation.

Introduction

Mother-to-child transmission of human immunodeficiency virus type 1 (HIV-1) accounts for most paediatric HIV-1 infections. In the absence of anti-retroviral therapy, the establishment of the viral set point in infants occurs slowly, suggesting that induction of anti-viral immunity is not as rapid as in adults [1, 2]. HIV-1 disease is characterized by chronic immune activation, and lymphocyte activation is a hallmark of disease progression [3]. Several studies suggest that, in addition to HIV-1 antigens, other viral and/or microbial antigens, proinflammatory cytokines and chemokines cause immune activation [4–6]. Increased translocation of luminal microbial products may also be involved [7]. Indeed, HIV-1 infection is localized predominantly to the gastrointestinal tract [8]; CD4⁺ T cell depletion and damage to the mucosal barrier lead to translocation of microbes and/or microbial products resulting in systemic immune activation [7].

The administration of highly active anti-retroviral therapy (HAART), which is usually a combination of protease and reverse transcriptase inhibitors, reduces HIV-1 load efficiently to undetectable levels and increases the number of circulating CD4⁺ T cells in children as well as adults [9, 10]. Although HAART does not eradicate HIV-1, due to a persistent low level of viral replication and the longevity of latently infected cells, it allows for substantial recovery of immune function [11, 12]. Immune reconstitution after HAART initiation is different in children. While, in the adult, it is a biphasic process with an initial rapid increase in memory CD4⁺ T cells followed by a slower and smaller increase in naive CD4⁺ T cells [13, 14], in children immune repopulation involves mainly naive cells, due probably to more efficient thymopoiesis [15–18]. Furthermore, some children show a discordant response to HAART characterized by a significant increase in CD4⁺ T cell counts, despite persistent detectable viraemia [15, 16, 19–22]. Although the mechanisms of this discordant response to HAART have not been elucidated fully, it has been suggested that thymic output [15, 16, 19, 23] and an impaired replicative capacity of drug-resistant viruses [20, 24, 25] might be involved.

How different CD4⁺ T cell subsets contribute to immune reconstitution and influence immune activation in virological responder and non-responder children remains to be clarified. The aim of this study was to analyse the characteristics of immune reconstitution and the status of immune activation in children with different virological responses to therapy.

Materials and methods

Patients

The study cohort consisted of 30 vertically HIV-1-infected children who were treated at the Pediatric Department of Padova University. The inclusion criterion was their response to HAART: 14 virological responder (VR) children had long-term viral load (VL) suppression (VL ≤ 400 copies/ml) and 16 virological non-responder (VNR) children had persistently detectable VL (VL > 400 copies/ml). All children were naive for protease inhibitor (PI) at HAART initiation. The HAART regimen consisted of a triple-drug combination, including two reverse transcriptase inhibitors (zidovudine, lamivudine, stavudine or nevirapine) and one protease inhibitor (nelfinavir, indinavir or ritonavir). Four VNR children enrolled in this study maintained a detectable VL despite changes in reverse transcriptase inhibitors. CD4⁺ and CD8⁺ T cell counts and plasma HIV-1 RNA levels were followed over time. Peripheral blood mononuclear cells (PBMC) were isolated from peripheral blood by centrifugation on Ficoll-Paque (Pharmacia, Uppsala, Sweden) gradient and then cryopreserved until use. Seventeen age-matched uninfected children, all born to HIV-1-seropositive mothers, were also analysed. The study was approved by the Institutional Ethical Committee; informed consent was obtained for patients and controls from parents or legal guardians.

Flow cytometry quantification of CD4⁺ and CD8⁺ T lymphocyte subpopulations

T lymphocyte subpopulations were stained for four-colour flow cytometry using the following labelled monoclonal antibodies (mAbs): anti-CD3 [fluorescein isothiocyanate (FITC)], anti-CD4 [peridinin chlorophyll protein (PerCP)], anti-CD8 (PerCP), anti-CD27 [phycoerythrin (PE)], anti-CD45RA [allophycocyanin (APC)] and anti-CD38 (PE). Appropriate isotypic controls (mouse IgG1-PE and mouse IgG2b-APC) were used to evaluate non-specific staining. mAbs were obtained from Becton-Dickinson (Becton-Dickinson Biosciences Pharmingen, San Diego, CA, USA). Acquisition was performed in a cytometer (FACSCalibur; Becton-Dickinson) cytometer using the CELLQuest Software (Becton-Dickinson). A total of 50 000 events were collected in the lymphocyte gate using morphological parameters (forward- and side-scatter). The percentage of CD45RA⁺, CD27⁺ and CD38⁺ cells was calculated within the CD3⁺CD4⁺ or CD3⁺CD8⁺ gate. The absolute cell count of different subsets was calculated by multiplying the percentage value by the total number of CD4⁺ or CD8⁺ T cells. CD45RA and CD27 expression was used to identify CD4⁺ and CD8⁺ T cell subsets as follows: CD45RA⁺CD27⁺ (naive), CD45RA⁻CD27⁺ (central memory), CD45RA⁻CD27⁻ (effector/memory CD4⁺ and preterminally differentiated effector CD8⁺ T cells) and CD8⁺CD45RA⁺CD27⁻ (terminally differentiated cytotoxic effector) [26–30].

Quantification of T cell receptor rearrangement excision circles (TREC)

Thymic output was studied in six VR and eight VNR children by measuring TREC in sequential samples collected at HAART entry and during therapy (1, 3, 6, 9, 12, 18 and 24 months). TREC levels were analysed by real-time polymerase chain reaction (PCR), exactly as described previously [15, 16].

Lipopolysaccharide (LPS) levels

Plasma samples were diluted fivefold with endotoxin-free water and then heated to 70°C for 10 min to inactivate plasma proteins. Plasma LPS was quantified with a commercially available assay (Limulus amoebocyte lysate QCL-1000; Cambrex, Milan, Italy) according to the manufacturer's protocol.

Statistical analysis

Differences in characteristics among infants were analysed using non-parametric tests. The Mann–Whitney test was used to compare data between groups, and the Wilcoxon non-parametric test was used to compare data within a group. TREC and CD4⁺CD45RA⁺CD27⁺ T cell count data obtained in sequential samples were analysed by linear regression analysis, and the slopes of TREC/1 × 10⁵ PBMC per day and CD4⁺CD45RA⁺CD27⁺ T cells/μl per day were calculated individually. Correlations between different parameters were analysed using the non-parametric Spearman's test. Analyses were performed with spss statistical software (version 12·0).

Results

Immunological and virological characteristics of HIV-1-infected children

Thirty HIV-1-infected children and 17 uninfected children were studied. All HIV-1-infected children were studied at baseline and after approximately 2 years (median 23 months; range 19–30 months) of HAART. The immunological and virological characteristics at baseline and after HAART are described in Table 1. Before therapy, HIV-1-infected children had a median plasma viraemia of 4·5 log₁₀ copies/ml. Both CD4⁺ T cell percentage and count were significantly lower than those of HIV-1-uninfected controls. In contrast, the CD8⁺ T cell percentage was higher than that of uninfected children. At baseline, CD4⁺ and CD8⁺ T cell counts and percentages were not significantly different in the VR and VNR groups. After 2 years of therapy, there was a significant increase in CD4⁺ T cell percentage and count in both the VR (13–29%, P < 0·001, and 345–686 cells/μl, P < 0·001, respectively) and VNR groups (7–17%, P = 0·001, and 210–433 cells/μl, P = 0·008, respectively). After 2 years of HAART, CD8⁺ T cell percentage decreased in VR children (P = 0·003); however, the absolute CD8⁺ T cell counts did not change significantly in either group, yet tended to increase in VNR children.

Table 1.

Immunological and virological characteristics of the studied population of human immunodeficiency virus type 1 (HIV-1)-infected, highly active anti-retroviral therapy (HAART)-treated children.

				Virological responder			Virological non-responder
Characteristics	Control group*	Total number of HIV-1-infected children before HAART	Pa	Baseline	After HAART	Pb	Baseline	After HAART	Pb
Number of children	17	30		14	14		16	16
Age (years)	8·7	9·3		9·4	11·8		6·4	8·4
	(2·5–17)	(2–14)		(2·8–14·2)	(4·8–16)		(2–14)	(4–17)
Lymphocyte subsets
% CD4+	28·8	10	<0·001	13·3	28·8	<0·001	7·4	17·6	0·001
	(11–49)	(0–34)		(0–34)	(18–49)		(2·5–29)	(6·5–35·7)
CD4+/μl	750	272	<0·001	345	686	<0·001	210	433	0·008
	(369–2338)	(5–916)		(5–571)	(439–1851)		(10–916)	(63–1126)
% CD8+	17·5	44	<0·001	42·6	32·3	0·003	46	51	0·387
	(9·9–29·1)	(6–72)		(25–72)	(19–54)		(20–58)	(23–70)
CD8+/μl	509	739	0·12	702	858	0·391	896	1078	0·332
	(239–1884)	(99–3664)		(225–2046)	(286–1570)		(99–3664)	(284–3451)
CD4+/CD8+	1·9	0·4	<0·001	0·4	1	<0·001	0·3	0·6	0·004
	(1–3·1)	(0·1–1·2)		(0·1–1·2)	(0·4–2)		(0·1–0·9)	(0·1–1·3)
HIV-1 RNA log10 copies/ml	–	4·5	–	4·3	1·6	<0·001	4·9	4·7	0·033
		(2·6–6·1)		(3·2–5·6)	(1·6–2·4)		(2·6–6·1)	(2·9–5·6)

^*Control group: age-matched HIV-1-negative children. Values were expressed as the median (ranges in parenthesis). P^a values between the control group and total HIV-1-infected children at baseline. P^b values between baseline and after HAART in virological responders and virological non-responders.

T cell subsets in uninfected and HIV-1-infected children at baseline

At baseline, compared to the control group, HIV-1-infected children had significantly lower cell counts in all CD4⁺ T cell subsets (naive 157 versus 415 cells/μl, P < 0·0001; central memory 85 versus 290 cells/μl, P < 0·0001; effector/memory 25 versus 77 cells/μl, P = 0·0002) (Fig. 1a). Conversely, HIV-1-infected children had higher cell counts in all CD8⁺ T cell subsets (pre-effectors 131 versus 34 cells/μl, P = 0·002; cytotoxic effector 185 versus 37 cells/μl, P = 0·001; and central memory 204 versus 121 cells/μl, P = 0·038), except naive CD8⁺ T cells, which were significantly lower in HIV-1-infected than uninfected children (191 versus 316 cells/μl, P = 0·006) (Fig. 1b). Interestingly, CD4⁺CD38⁺ T cells were lower (65 versus 209 cells/μl, P < 0·0001), while CD8⁺CD38⁺ T cells were higher in infected versus uninfected children (116 versus 25 cells/μl, P = 0·0001) (Fig. 1b).

Fig. 1.

The phenotypic characteristics of CD4⁺ and CD8⁺ T subsets from 17 human immunodeficiency virus type 1 (HIV-1)-uninfected children and 30 HIV-1-infected children at highly active anti-retroviral therapy initiation are illustrated. CD27, CD45RA and CD38 expression in CD3⁺CD4⁺ T cells (a) and CD3⁺CD8⁺ T cells (b) were analysed by flow cytometry. Boxes and whiskers represent the 25th−75th and 10th−90th percentiles, respectively; the median is the central line in each box. P-values were determined by the Mann–Whitney U-test.

Changes in T cell subsets during HAART in VR and VNR children

At baseline, CD4⁺ and CD8⁺ T cell subsets of VR children did not differ significantly from those of the VNR group, even if pre-effector (CD45RA⁻CD27⁻), effector (CD45RA⁺CD27⁻) and CD38⁺ CD8⁺ T cell counts were more variable in VNR than VR children (Fig. 2a,b). After 2 years of HAART, CD4⁺ T cell counts increased in both groups, but in a different manner. In VR children, the peripheral CD4⁺ T cell increase was significant in naive (168–462 cells/μl, P = 0·001), central memory (97–213 cells/μl, P = 0·002) and effector/memory cell subsets (32–59 cells/μl, P = 0·035) and reached values similar to those observed in the uninfected children (Fig. 3a). Conversely, in VNR children, only naive CD4⁺ T cells increased significantly (113–263 cells/μl, P = 0·015), while values of all other CD4⁺ T cell subsets remained significantly lower than those observed in uninfected children (Fig. 3b). CD4⁺CD38⁺ T cells increased in both VR and VNR groups (66–142 cells/μl, P = 0·002, and 65–124 cells/μl, P = 0·045, respectively; Fig. 3a,b). A positive correlation between CD4⁺CD38⁺ and CD4⁺CD45RA⁺CD27⁺ T cell counts was observed at baseline and after HAART in both VR (at baseline r_s = 0·84, P = 0·0002; and after HAART r_s = 0·55, P = 0·05) and VNR children (r_s = 0·91, P < 0·0001; and r_s = 0·80, P = 0·0002, respectively). There was also a positive correlation between the increase of naive and CD38⁺ CD4⁺ T cell counts in both VR (r_s = 0·51, P = 0·04) and VNR groups (r_s = 0·96, P < 0·0001). Although the total CD8⁺ T cell count did not change during HAART (Table 1), a rise in naive CD8⁺ T cell subsets was observed in both VR (from 195 to 358 cells/μl, P = 0·006) and VNR children (from 165 to 288 cells/μl, P = 0·004). No significant changes were observed in the other CD8⁺ T cell subsets (central memory, pre-effector and cytotoxic effector cells) in either group, although the absolute cell numbers tended to increase in VNR children. CD8⁺CD38⁺ T cells decreased significantly only in VR children (107–60 cells/μl, P = 0·003), while these remained high in VNR children (Fig. 3c,d).

Fig. 2.

T cell subsets are illustrated in virological responder (VR) and virological non-responder (VNR) children at highly active anti-retroviral therapy initiation. CD27, CD45RA and CD38 expression in CD3⁺CD4⁺ T cells (a) and CD3⁺CD8⁺ T cells (b) were analysed in 14 VR and 16 VNR children by flow cytometry. Boxes and whiskers represent the 25th−75th and 10th−90th percentiles, respectively; the median is the central line in each box. P-values were determined by the Mann–Whitney U-test.

Fig. 3.

T cell repopulation in virological responder (VR) and virological non-responder (VNR) children is illustrated. CD27, CD45RA and CD38 expression in CD3⁺CD4⁺ T cells (a and b) and CD3⁺CD8⁺ T cells (c and d) were analysed in 14 VR children (a and c) and in 16 VNR children (b and d) at baseline and after 2 years of highly active anti-retroviral therapy. Data were compared with those obtained from 17 uninfected children. Boxes and whiskers represent the 25th−75th and 10th−90th percentiles, respectively; the median is the central line in each box. P-values were determined by the Wilcoxon test.

Relationship between TREC and naive CD4⁺ T cells in VR and VNR children

TREC levels were quantified in six VR and eight VNR children for whom sequential samples during therapy were available. At baseline, TREC levels in PBMC and CD4⁺CD45RA⁺CD27⁺ T cell counts in the VR children did not differ significantly from those in the VNR group (Fig. 4a,b). After 2 years of HAART, TREC levels increased significantly in both VR (1503–5018, P = 0·04) and VNR children (595–2117, P = 0·03, Fig. 4a). In this subset of children the increase in CD4⁺CD45RA⁺CD27⁺ T cells occurred in both groups (Fig. 4b), and reflected the increase observed in the entire VR and VNR study groups (Fig. 3a,b). A positive correlation between TREC levels and CD4⁺CD45RA⁺CD27⁺ T cells was observed in both groups at baseline (VR, r_s = 0·94, P = 0·005; VNR, r_s = 0·83, P = 0·01) and after HAART (VR, r_s = 0·88, P = 0·0019; VNR, r_s = 0·81, P = 0·015). Although the slope of the increase in TREC did not differ significantly in VR and VNR children, the slope of the increase in CD4⁺ naive T cells was lower in VNR than VR children (Fig. 4c,d). Moreover, the slopes of the TREC were correlated positively with the slopes of the CD4⁺ naive T cells in VR (r_s = 0·90, P = 0·037), but not in VNR children (r_s = 0·48, P = 0·33).

Fig. 4.

The dynamics of thymic output and CD4⁺CD45RA⁺CD27⁺ T cells in six virological responder (VR) and eight virological non-responder (VNR) children is illustrated. T cell receptor rearrangement excision circles (TREC) (a) and CD4⁺CD45RA⁺CD27⁺ T cell levels (b) were analysed at baseline and after 2 years of highly active anti-retroviral therapy. Slopes of TREC (c) and CD4⁺CD45RA⁺CD27⁺ T cells (d) in VR and VNR children. Boxes and whiskers represent the 25th−75th and 10th−90th percentiles, respectively; the central line in each box indicates the median. P-values were determined by the Wilcoxon (a and b) and the Mann–Whitney U-tests (c and d).

LPS levels in HIV-1-infected children

Plasma LPS levels were quantified in uninfected and HIV-1-infected children at baseline, before initiation of HAART. The patients showed significantly higher values than the control group (P = 0·0001; Fig. 5a); however, baseline LPS levels were not considerably different in the VR and VNR groups. After 2 years of HAART (Fig. 5b), LPS levels were unchanged in VR children but were increased still further in VNR children (56–67 pg/ml, P = 0·035).

Fig. 5.

Microbial translocation in human immunodeficiency virus type 1 (HIV-1)-infected children is illustrated. Plasma lipopolysaccharide levels were determined in HIV-1-uninfected and HIV-1-infected children before highly active anti-retroviral therapy (HAART) initiation (a) and in virological responder (VR) and virological non-responder (VNR) children at baseline and after HAART (b). Boxes and whiskers represent the 25th−75th and 10th−90th percentiles, respectively; the central line in each box indicates the median. *P < 0·001 in comparison of the control group versus HIV-1-infected children.

Discussion

HAART has dramatically modified the course of HIV-1 infection, resulting in a drastic decrease of viral load to undetectable levels and a significant increase in the peripheral CD4⁺ T cell repopulation. In HAART-treated children, immune repopulation may occur despite persistent viraemia [15, 16, 19–22]. To add to the scarce information available on immune reconstitution in HIV-1-infected children, we investigated peripheral immune repopulation in children with and without a virological response to therapy.

Our study demonstrated that immune reconstitution was quite different in VR and VNR children with regard to both CD4⁺ and CD8⁺ T cell compartments. Moreover, immune activation decreased in VR children, yet persisted and even increased in VNR children.

Prior to HAART initiation, VR and VNR patients had similar levels of plasma viraemia and did not show significant differences in any T cell subsets and TREC level. When compared to age-matched HIV-1-uninfected children, both VR and VNR children showed cell depletion in all CD4⁺ T cell subsets, but a higher cell count in all CD8⁺ T cell subsets except naive cells. This finding, together with the high level of activated CD8⁺CD38⁺ cells, is in agreement with the concept that HIV-1 infection drives a broad activation and differentiation of CD8⁺ T cells, pushing them from naive to late-stage differentiation [31].

After 2 years of HAART, immune reconstitution in VR children occurred in all cell subsets, while in the VNR children only naive CD4⁺ T cells increased. Of interest, both VR and VNR children showed an increase in TREC levels. This result is in agreement with previous findings [15, 16, 19, 21] and indicates that the persistence of viraemia during HAART does not impair the increase in thymic function and the subsequent output of naive cells. The finding that naive CD8⁺ T cells increased in both VR and VNR groups confirms that thymic function was recovered. However, the increase in TREC correlated with the increase in naive CD4⁺ T cells in VR, but not in VNR children, in agreement with a previous study [15]. In addition, for a similar TREC increase, the naive CD4⁺ T cell increase was lower in VNR than VR children. It is possible that thymic emigrants in the context of ongoing viraemia are activated from circulating viral antigen, proliferate and differentiate into memory cell subsets [32], which are more susceptible to HIV-1 infection than naive cells. Although the mechanisms underlying partial recovery of immune function despite virological failure remain to be elucidated, it is likely that HAART prevents HIV-1 replication in T cell progenitors [33–35]. In addition, the impaired replicative capacity of drug-resistant viruses and the shift of viral co-receptor usage from CXCR4 expressed on naive cells to CCR5 expressed on memory CD4⁺ T cells [20, 33, 36] probably both contribute to peripheral immune repopulation by naive CD4⁺ T cells.

Along with the rise in CD4⁺, CD4⁺CD38⁺ cells increased in both VR and VNR groups and these were correlated positively with increases in naive cells. As opposed to what occurs in adults, these results suggest that CD38 is a marker of cell immaturity rather than cell activation in CD4⁺ T cells in children [37, 38]. Finding that its expression is high in naive CD4⁺ T cells [39], that it decreases over time [40] and that it increases in CD4⁺ T cells during immune reconstitution [37, 38] all support this concept. Besides representing a marker of cell immaturity, CD38 in CD4⁺ T cells is likely to have multiple functional activities [37]. Further studies are required to clarify in depth the role of CD38 in CD4⁺ T cells in children.

None the less, our study demonstrates that persistence of viraemia greatly impairs the expansion of memory and effector CD4⁺ T cell subsets. In VR children, these cell subsets increased and reached levels similar to those observed in uninfected controls. The peripheral CD4⁺ T cell subset repopulation may be due to an expansion of memory and effector CD4⁺ T cells, cellular redistribution from tissue/lymph nodes to the periphery and reduced apoptosis [41]. In VNR children, these cell subsets remained significantly lower. This may be due to the persistence of circulating virions in VNR children and also to the greater susceptibility of these cells to HIV-1 infection compared to naive cells [33, 42]. The peripheral depletion may also be induced by a decrease in cell proliferation and an increase in cell death [41]. Further studies with markers of cell apoptosis and survival, not feasible in the present analysis due to the small numbers of cells available, will be very useful to clarify this aspect.

With regard to the CD8⁺ T cell compartment, the percentage of CD8⁺ T cells decreased in VR children; this percentage decrease was significant not only because of the CD8⁺ lowering effect, but also because CD4⁺ increased notably. No significant changes occurred in memory and effector cell subsets, yet absolute cell counts tended to further increase in VNR children. This finding suggests that HIV-1-driven chronic stimulation and activation continued to persist in VNR children. This was also supported by the finding that activated CD8⁺CD38⁺ T cells decreased significantly in VR children, yet remained persistently high in VNR children. Reduction of CD38 expression in the VR group may reflect a reduction of antigenic stimulation due to the undetectable plasma viraemia [37, 43]. Immune activation is driven not only by HIV-1, but also by other pathogens. HIV-1 infection occurs in the gastrointestinal tract, damaging the immune component of the gastrointestinal mucosa, with subsequent translocation of microbial products such as LPS, a potent immunostimulatory molecule that affects systemic immune activation. At baseline, circulating levels of LPS were higher in HIV-1-infected children than in controls. While a reduction in LPS levels has been reported in HIV-1-infected adults during therapy [7], LPS levels remained roughly stable in VR children, but tended to increase even further in VNR children during HAART. This finding strongly supports the concept that persistence of immune activation in children might be caused not only by persistence of HIV-1 antigens but also by other pathogens and microbial products.

In summary, this study has demonstrated that HAART induces immune repopulation of different CD4⁺ T cell subpopulations only in children with a virological response to therapy. Moreover, persistence of viraemia did not affect thymic output and the increase in naive cells, but greatly impaired CD4⁺ T cell differentiation. The persistence of HIV-1 virions/antigens, together with circulating microbial products, sustained high levels of activated CD8⁺CD38⁺ T cells. The persistence of systemic immune activation might ultimately determine the progression to AIDS.