Are increased frequency of macrophage-like and natural killer (NK) cells, together with high levels of NKT and CD4⁺CD25^high T cells balancing activated CD8⁺ T cells, the key to control Chagas’ disease morbidity?

Abstract

The immunological response during early human Trypanosoma cruzi infection is not completely understood, despite its role in driving the development of distinct clinical manifestations of chronic infection. Herein we report the results of a descriptive flow cytometric immunophenotyping investigation of major and minor peripheral blood leucocyte subpopulations in T. cruzi-infected children, characterizing the early stages of the indeterminate clinical form of Chagas’ disease. Our results indicated significant alterations by comparison with uninfected children, including increased values of pre-natural killer (NK)-cells (CD3^– CD16⁺ CD56^–), and higher values of proinflammatory monocytes (CD14⁺ CD16⁺ HLA-DR⁺⁺). The higher values of activated B lymphocytes (CD19⁺ CD23⁺) contrasted with impaired T cell activation, indicated by lower values of CD4⁺ CD38⁺ and CD4⁺ HLA-DR⁺ lymphocytes, a lower frequency of CD8⁺ CD38⁺ and CD8⁺ HLA-DR⁺ cells; a decreased frequency of CD4⁺ CD25^HIGH regulatory T cells was also observed. These findings reinforce the hypothesis that simultaneous activation of innate and adaptive immunity mechanisms in addition to suppression of adaptive cellular immune response occur during early events of Chagas’ disease. Comparative cross-sectional analysis of these immunophenotypes with those exhibited by patients with late chronic indeterminate and cardiac forms of disease suggested that a shift toward high values of macrophage-like cells extended to basal levels of proinflammatory monocytes as well as high values of mature NK cells, NKT and regulatory T cells, may account for limited tissue damage during chronic infection favouring the establishment/maintenance of a lifelong indeterminate clinical form of the disease. On the other hand, development of an adaptive cell-mediated inflammatory immunoprofile characterized by high levels of activated CD8⁺ cells and basal levels of mature NK cells, NKT and CD4⁺ CD25^HIGH cells might lead to late chronic pathologies associated with chagasic heart disease.

Introduction

Chagas’ disease or American trypanosomiasis is a protozoan infection caused by the haemoflagellate protozoan Trypanosoma cruzi. It is one of the most important public health problems in Latin America, affecting 16–18 million people in South and Central America [1].

Human T. cruzi infection evolves from a usually oligosymptomatic acute phase to a chronic disease, where patients can be grouped into distinct categories based on clinical status. The great majority of the patients that progress to the chronic phase remain clinically asymptomatic for many years; this condition characterizes the indeterminate (IND) clinical form of the disease. About 30–40% of patients progress to cardiac (CARD) or digestive symptomatic disease. It is estimated that 30% of all infected individuals will eventually develop heart disease [2].

The factors that underlie and determine the distinct clinical outcomes, mild or severe disease, are not completely understood. However, there is a general consensus that the host immune response plays a pivotal role associated with the pathogenesis as well as the protective events that control chagasic tissue damage [3]. It is also well accepted that T. cruzi induces a strong activation of the immune system during acute infection and that the different immunological mechanisms triggered during the early indeterminate (E-IND) stages of T. cruzi infection may represent an essential component of the immune activity observed during ongoing, clinically distinct chronic infection [4].

In the search to identify differences in the immunological response related to the development/maintenance of distinct chronic disease, we have focused on major and minor peripheral blood leucocyte subsets during E-IND and late chronic IND and CARD Chagas’ disease [5,6]. We have reported previously that an expansion of natural killer (NK) cells before the development of T cell-mediated immunity, in addition to enhancement of circulating activated B cells, are the hallmarks of human immune response during early T. cruzi infection. Moreover, we have reported that an increase of pre-NK cells (CD16⁺ CD56^–), as well as a persistent expansion of activated B cells and down-regulation of CD54 on T cells, are also observed during initial stages of chronic T. cruzi infection [5]. Here, we discuss the hypothesis that T cell-mediated immunity during the early stages of T. cruzi infection may represent a phenomenon restricted to the cardiac and lymph node compartment, and may not be detectable in the peripheral blood.

Despite the T cell-independent nature of the immune response triggered in early Chagas’ disease, we have demonstrated that T cells play an important role in the dynamics of chronic Chagas’ disease [6–8]. Previous reports from our group showed that despite their clinical status, chronic chagasic patients display a high frequency of peripheral blood activated T cells (HLA-DR⁺) as well as lack of CD28 expression on many of their circulating T lymphocytes [7,8]. More recently, ex vivo immunophenotyping demonstrated that IND patients display a higher frequency of both CD4⁺ CD25^HIGH and NKT (CD3⁺ CD16^– CD56⁺) regulatory cells associated with increased levels of circulating ‘cytotoxic’ NK cells (CD3^– CD16⁺ CD56⁺ and CD3^– CD16⁺ CD56^DIM NK cells) [6]. On the other hand, an increased percentage of activated CD8⁺ HLA-DR⁺ T cell subset was associated exclusively with severe clinical forms of Chagas’ disease [6]. We hypothesize that regulatory T cells control the deleterious cytotoxic activity in the indeterminate clinical form, inhibiting the activation of CD8⁺ HLA-DR⁺ T cells. The lack of regulated populations in CARD disease patients could account for exacerbated immune response that culminates in strong cytotoxic activity and tissue damage.

Relevant findings regarding histopathological alterations in biopsies from chagasic patients showed that tissue CD4⁺ and CD8⁺ T cells increase simultaneously during early infection but not in the chronic phase, supporting the hypothesis of compartmentalized T cell-mediated immune response during early disease and suggesting an immunological imbalance of T cell profile in late chronic Chagas’ disease. In the chronic phase, patients with heart failure present with higher levels of CD8⁺ T cells than CD4⁺ T cells, leading to a lower tissue CD4⁺/CD8⁺ T cell ratio [9,10].

More recently, data have been reported suggesting that monocytes from IND patients display modulatory characteristics related to low HLA-DR and high IL-10 expression, whereas monocytes from CARD patients may be committed to induction of inflammatory responses related to high tumour necrosis factor (TNF)-α expression [11–13].

Increasing numbers of novel cellular parameters and surface markers have been examined as conventional flow cytometry-based investigations, i.e. ‘look and conclude’ analyses, have adopted new gating strategies to analyse immunophenotypes at the single-cell level in a semiquantitative manner. Indeed, flow cytometry has emerged as the methodology of choice for enumerating and characterizing of novel leucocyte subsets using three- and four-colour platform technology. With this technology, several novel phenotypic features of leucocyte subsets are characterized routinely in parallel by their in vitro and in vivo functional properties, such as NKT cells (CD3⁺ CD56⁺) [14], functionally distinct NK subsets (CD3^–CD16^–/+ CD56^–/+) [15,16], regulatory T cells (CD4⁺ CD25^HIGH) [17], macrophage-like monocytes (CD14⁺ CD16⁺) [18] and proinflammatory monocytes (CD14⁺ CD16⁺ HLA-DR⁺⁺) [19].

We have performed a descriptive flow cytometric immunophenotyping investigation based on these new gating strategies to enumerate major and minor leucocyte subpopulations in the peripheral blood of T. cruzi-infected children, characterizing the E-IND stages of Chagas’ disease. Comparative cross-sectional analyses of the predominant immunophenotypes were also performed in those patients exhibiting late chronic IND or CARD disease. Our results suggest that a shift toward high levels of macrophage-like cells (CD14⁺ CD16⁺) and NK cells, besides high frequency of regulatory lymphocytes (NKT and CD4⁺ CD25^HIGH cells), may favour the establishment/maintenance of the lifelong indeterminate clinical form of the disease. On the other hand, maintenance of major cell phenotypic features observed during early infection as well as the development of an adaptive cell-mediated inflammatory immunoprofile characterized by high levels of activated CD8⁺ cells and basal frequency of mature NK cells, NKT and CD4⁺ CD25^HIGH, might lead to a late chronic disease associated with cardiac pathological events.

Patients, materials and methods

Study area

Berilo and José Gonçalves de Minas are located in Jequitinhonha Valley in the north-east of Minas Gerais State, Brazil. Jequitinhonha Valley comprises 970 km². Chagas’ disease was formerly endemic in the area. Together, these two municipalities have 17 632 inhabitants, with 78·43% of these individuals living in rural areas with an economy based on agriculture and cattle ranching [IBGE.Cidades@,http://URL:http://www.ibge.gov.br/cidadesat/default.php (search for Berilo and José Gonçalves de Minas); accessed 4 January 2006].

Bambui is located in the south-west of Minas Gerais State, Brazil. It comprises 1455 km² and is another area in which Chagas’ disease was formerly endemic. It has 22 274 inhabitants, 80% of them in the urban area of the municipality [IBGE.Cidades@,http://www.ibge.gov.br/cidadesat/default.php (search for Bambuí); accessed 4 January 2006].

Study population

School children enrolled in a cross-sectional study performed at 37 communities from Berilo and José Gonçalves de Minas, including 39 school units, participated in a serological screening trial to detect anti-T. cruzi antibodies by enzyme-linked immunosorbent assay (ELISA) using blood eluate from filter paper. The screening immunoassay identified 2·69% of school children with positive results for anti-T. cruzi IgG. Confirmatory immunodiagnosis for Chagas’ disease was performed by ELISA, EIE-Rec-ELISA (Biomanguinhos/FIOCRUZ), indirect immunofluorescence assay (IFA) and haemaglutination (HA) tests. Considering the World Health Organization and Brazilian Health Ministry criteria that recommend the use of at least two serological tests, with distinct principles, to confirm the diagnosis of Chagas’ disease, we confirmed six of the 38 cases first identified with positive results by the screening ELISA, leading to a total prevalence of 0·42%. The seropositive cases included four males and two females, with ages ranging from 9 to 14 years. The clinical and physical examination revealed that all children were asymptomatic, showing normal conventional electrocardiograms (except no. 701, who showed enlargement of Pri−0·22) and unaltered thoracic X-ray (RX) (Table 1). The haemoculture was positive in all children examined (six of six), generally within the first month of blood cultivation in liver infusion tryptose (LIT) media supplemented with 10% fetal calf serum. All seropositive children were treated with benznidazol (Rochagan®, Roche) and are currently under evaluation following the protocol recommended by the Brazilian Health Ministry [20]. Seven non-infected (NI) schoolchildren were included as a control group. The NI-1 group consisted of age-matched schoolchildren with negative serology for anti-T. cruzi IgG immunodiagnosis (ELISA and IFA). The NI-1 group included one male and six females, with ages ranging from 9 to 14 years (mean = 12·4).

Table 1.

Patient characteristics.

Group	No. of individuals	Age range (years)	Sex (male/ female)
Non-infected children (NI-1)	7	9–14	6/1
Non-infected adults (NI-2)	12	20–59	3/9
Early indeterminate (E-IND)	6	9–14	4/2
Chronic indeterminate (IND)	8	44–67	3/5
Chronic cardiac (CARD)	13	50–70	5/8

Cross-sectional analyses of late chronic chagasic patients compared with uninfected adult controls were carried out to evaluate major immunophenotypic features. All late chronic infected individuals as well as the uninfected adult controls were from Bambuí, Minas Gerais State, Brazil, and participated in serological examination to confirm the positive or negative diagnosis for T. cruzi infection, respectively. The diagnosis was based on standard serological tests, including IFA and HA tests. In this study, we used 21 samples from chagasic patients with late chronic disease. According to their clinical records, the late chronic chagasic patients were divided into two categories, namely IND and CARD clinical forms. Patients presenting asymptomatic T. cruzi infection, classified as indeterminate (n = 8), had no clinical manifestations of the disease other than their positive serology. Patients with cardiac dysfunction, CARD (n = 13), presented dilated cardiomyopathy and were diagnosed by a detailed clinical examination, including eletrocardiography (ECG), 24-h Holter examination and chest X-ray. Twelve seronegative adults were included in this study as negative controls 2 (NI-2 = 12). All were living in an area endemic for Chagas’ disease (Table 1).

Informed written consent was obtained from all participants or through their parents or legal guardians in the case of the school children. This work complied with resolution number 196/1996 from the National Health Council for research involving humans and was approved by the Ethical Committee at Centro de Pesquisas René Rachou (CPqRR/FIOCRUZ protocol 11/2004), Belo Horizonte, Minas Gerais, Brazil.

Blood samples

A 5-ml sample of peripheral blood was collected from each subject using ethylenediamine tetraacetic acid (EDTA) as the anticoagulant. The samples were collected by trained professionals in an ambulatory hospital. After the collection, the whole peripheral blood was analysed by flow cytometry.

Specific monoclonal antibodies used for immunophenotyping

Mouse anti-human monoclonal antibodies (mAbs), conjugated with fluorescein isothiocyanate (FITC), phycoerythrin (PE) or tri-colour (TC) and specific for cell-surface markers were used simultaneously for two- or three-colour flow cytometric assay. In this study, we used anti-human FITC-conjugated mAbs including anti-CD3 (UCHT1), anti-CD4 (RPA-T4), anti-CD5 (L17F12), anti-CD8 (B9·11), anti-CD16 (3G8), anti-CD18 (YF118·3), anti-CD54 (15·2), anti-CD62L (DREG-56) and mouse IgG1 as the isotypic control (679·1Mc7). The following second-colour reagents were used: anti-human PE-conjugated mAbs anti-CD3 (UCHT1), anti-CD4 (RPA-T4), anti-CD19 (4G7), anti-CD23 (M-L233), anti-CD25 (3G10), anti-CD28 (15E8), anti-CD38 (AT13/5), anti-CD56 (B159), anti-HLA-DR (TÜ36) and mouse IgG2a as the isotypic control (UCTH-1). All antibodies were purchased from Becton-Dickinson (Mountain View, CA, USA). The third colour parameter was evaluated using TC-conjugated mAbs and included anti-CD8 (M-L233), anti-CD14 (TüK4), anti-CD16 (3G8) and anti-CD19 (4G7), all purchased from Caltag Laboratories (Burlingame, CA, USA).

Flow cytometric analysis of peripheral blood

White blood cell phenotypes were analysed following an immunofluorescence procedure recommended by Becton-Dickinson, modified as follows: 100 µl peripheral blood which had been collected in Vacutainer tubes containing EDTA (Becton Dickinson) was mixed in 12 × 75 mm tubes with 5 µl undiluted mAbs specific for several cell surface markers; the tubes were incubated in the dark for 30 min at room temperature. Following the incubation, erythrocytes were lysed with 2 ml FACS lysing solution (Becton Dickinson Biosciences Pharmigen, San Diego, CA, USA). The remaining cells were then washed twice with 2 ml phosphate-buffered saline containing 0·01% sodium azide. Cell preparations were fixed in 200 µl FACS Fix solution (10 g/l paraformaldehyde, 1% sodium-cacodylate, 6·65 g/l sodium-chloride, 0·01% sodium azide). Cytofluorimetric data acquisition was performed with a Becton-Dickinson FACScalibur instrument. cellquest™ software provided by the manufacturer was used for data acquisition and analysis.

Statistical analysis

Differences between groups were first evaluated by minitab software (release 13·20) to evaluate the independence, normality and variance of data sets. Those data sets meeting the three criteria were considered parametric and were compared further by analysis of variance (anova) followed by the Tukey test, using the prism 3·0 program. Non-parametric data were analysed by the Kruskal–Wallis test followed by Dunn’s test. Correlation analysis was performed by Pearson’s and Spearman’s tests, respectively. Significance was defined in both cases at P < 0·05.

Results

Low values of CD3⁺ T lymphocytes, mainly CD8⁺ T lymphocytes and impaired T cell activation, are the hallmark of the early indeterminate clinical form of Chagas’ disease

The percentage of T cell populations and the major subsets CD4⁺ and CD8⁺ are shown in Fig. 1. Statistical analysis demonstrated a lower percentage of circulating T lymphocytes (CD3⁺) in children with E-IND Chagas’ disease in comparison to non-infected children (NI-1) (Fig. 1a). Further analysis revealed that the decrease in CD3⁺ T cells was correlated with a significant decrease in the CD8⁺ T lymphocyte subset (r = 1, P = 0·0167) (Fig. 1c). No significant differences were found in the mean values of the circulating CD4⁺ T cell subset (Fig. 1b).

Fig. 1.

Analysis of T and B cell subsets in the peripheral blood of early indeterminate Trypanosoma cruzi infected children (E-IND, •) and non-infected children (NI-1, ○). Phenotypic studies were performed by a double-labelling protocol using anti-CD3 fluorescein isothiocyanate (FITC) and anti-CD4 phycoerythrin (PE) or anti-CD8-PE for T cell analysis. Anti-CD19 FITC and anti-CD5-PE or anti-CD23-PE were used to identify B cell subsets. The results are expressed as scattering of individual values and mean percentage of total T cells (a), T cell subsets (b, c), and total B cells (d) within gated lymphocytes. B lymphocyte subsets, including B-1 cells (e), conventional B cells (f) and activated B lymphocytes (g) were analysed within gated CD19⁺ lymphocytes. Significant differences (connecting lines) and P-values are shown on figure.

Analysis of activated T cells revealed a lower ratio of CD4⁺ HLA-DR⁺ and CD4⁺ CD38⁺ T cells with no changes in CD8⁺ HLA-DR⁺ and CD8⁺ CD38⁺ T cells, parallel to an unaltered profile of CD28 expression within CD4⁺ and CD8⁺ T cells (Table 2).

Table 2.

Frequency of activation marker and adhesion molecule expression by peripheral blood CD4⁺ and CD8⁺ T cell subsets from early indeterminate Trypanosoma cruzi-infected children early indeterminate (E-IND) and non-infected children (NI-1).

	T cell subsets
	CD4+		CD8+
Phenotypea	NI-1	E-IND	NI-1	E-IND
HLA-DR+	3·4 ± 0·9	1·3 ± 1·8*	5·3 ± 2·6	4·0 ± 2·8
CD28+	94·2 ± 5·1	97·6 ± 0·8	54·8 ± 14·7	66·1 ± 11·3
CD38+	61·3 ± 17·4	45·2 ± 10·2*	54·7 ± 8·2	56·2 ± 9·9
CD62L+	78·8 ± 10·3	68·1 ± 7·4*	52·4 ± 13·0	50·3 ± 12·0
CD18+	20·8 ± 9·7	27·8 ± 12·7	63·4 ± 14·5	55·2 ± 13·1
CD54+	7·8 ± 10·0	2·9 ± 4·9	27·4 ± 15·2	70·6 ± 15·0*

^aThe results are expressed as proportion within a given T cell subset, e.g. ratio of CD4⁺ HLA-DR⁺ within the CD4⁺ population, allowing the normalization of data when percentage of a given subset may differ.

^*Statistically significant differences (P < 0·05) in comparison to NI-1.

Higher values of activated B lymphocytes is observed of early indeterminate Chagas’ disease

Phenotypic analysis at the single-cell level was used to analyse the frequency of B cell subsets (conventional B lymphocytes/CD19⁺ CD5^– and B1/CD19⁺ CD5⁺) as well as their activation status, using anti-CD23 PE and anti-CD19 FITC in a dual-platform to identify activated B cells (CD19⁺ CD23⁺) (Fig. 1). Our findings showed no significant differences in the mean percentage of B cells and their major subsets between E-IND and NI-1 children (Fig. 1d,e,f). Interestingly, analysis of CD19⁺ B cells co-expressing the CD23 cell-surface activation marker showed an increased mean ratio of double-positive B lymphocytes within CD19⁺ cells in E-IND in comparison to NI-1 children (Fig. 1g).

Despite the lack of activation phenotypes among circulating CD8⁺ T cells, early indeterminate chagasic children displayed a high migratory potential of cytotoxic T cells

In order to quantify the frequency of circulating T cell subsets co-expressing surface selectin (CD62L) and integrins (CD18 and CD54), a three-colour flow cytometry analysis was carried out using a cocktail of monoclonal antibodies, including anti-CD62L, anti-CD18 or anti-CD54 FITC, plus anti-CD4 PE and anti-CD8 TC. Our results indicated CD4⁺ T cell activation, demonstrated by the lower ratio of circulating CD62L⁺ cells, despite unaltered levels of CD18⁺ and CD54⁺ cells, among CD4⁺ T lymphocytes (Table 2).

On the other hand, regardless of no phenotypic changes signalling the activation of CD8⁺ T cells (i.e. expression of HLA-DR, CD28, CD38, CD62L and CD18), our data demonstrated that increased levels of circulating CD8⁺ CD54⁺ T cells would be an immunological event that suggest the increased migratory potential of this cytotoxic population, which is important in controlling tissue parasitism (Table 2).

Low levels of macrophage-like (CD14⁺ CD16⁺) cells and expansion of CD14⁺ CD16⁺ HLA-DR⁺⁺ proinflammatory monocytes were observed in the early chronic T. cruzi-infected children

Ziegler-Heitbrock [18] suggested that, in humans, the expression of CD14 and CD16 by monocytes can be used to define at least two subsets of monocytes with distinct functional properties. In this context, CD14⁺ CD16^– cells are considered to be classical monocytes whereas CD14⁺ CD16⁺ cells are typically macrophage-like cells. Herein we have focused our analysis on major and minor circulating monocyte subpopulations. Our data demonstrated that E-IND samples displayed low levels of macrophage-like cells compared to NI-1 samples (Fig. 2a).

Fig. 2.

Analysis of monocyte and natural killer (NK) cell subsets in the peripheral blood of early indeterminate Trypanosoma cruzi-infected early indeterminate (E-IND, •) children and non-infected children (NI-1, ○). Monocyte subpopulation analysis was performed by a triple-labelling platform using anti-CD14 TC, anti-CD16 fluorescein isothiocyanate (FITC) and anti-HLA-DR phycoerythrin (PE) to identify macrophage-like cells (CD14⁺ CD16⁺) (a), proinflammatory monocytes (CD14⁺ CD16⁺ HLA-DR⁺⁺) (b). Natural killer (NK) phenotypic studies were performed by a triple-labelling protocol using anti-CD3 FITC, anti-CD56 PE and anti-CD16 tri-colour (TC) to identify total NK cells CD3^– CD16^–/+ CD56^–/+ (c), pre-NK cells CD3^– CD16⁺ CD56^–/CD3^– CD16^–/+ CD56^–/+ (b), mature NK cells CD3^– CD16⁺ CD56⁺/CD3^– CD16^–/+ CD56^–/+ (e). Data of monocyte subsets are expressed as scattering of individual values and mean percentage of cells within gated monocytes. The results of total NK cells were calculated within gated lymphocytes, whereas the frequency of NK cell subsets were reported within gated NK cells. Significant differences (connecting lines) and P-values are shown on figure.

Within the CD14⁺ CD16⁺ macrophage-like cells, two monocyte populations can be distinguished: classical HLA-DR⁺ monocytes and proinflammatory HLA-DR⁺⁺ monocytes [19]. Quantification of the CD14⁺ CD16⁺ HLA-DR⁺⁺ proinflammatory monocytes was carried out by first gating on the monocyte population identified on dot-plots based on their morphometric and immunophenotypic features, such as SSC^intermediateCD14⁺, followed by the selection of CD14⁺ CD16⁺ cells (macrophage-like cells) and further enumeration of those CD14⁺ CD16⁺ cells with high expression of HLA-DR, as proposed by Belge et al. [19]. Our results showed a higher value of CD14⁺ CD16⁺ HLA-DR⁺⁺ proinflammatory monocytes within CD14⁺ CD16⁺ monocytes in E-IND samples compared with NI-1 samples (Fig. 2b).

Pre-NK cells (CD3^– CD16⁺ CD56^–) are expanded in peripheral blood in early indeterminate Chagas’ disease

As proposed by Gaddy and Broxmeyer [21], distinct NK cell subsets can be identified based on the differential expression of two major NK cell markers: CD16 and CD56. In order to quantify the frequency of major NK cell subsets, pre-NK cells (CD3^– CD16⁺ CD56^–) and mature NK cells (CD3^–CD16⁺ CD56⁺) were quantified on a three-colour platform using anti-CD16 TC, anti-CD56 PE and CD3 FITC to exclude NKT cells. Our data showed that despite absence of statistically significant differences in the percentage of total NK cells (CD3^– CD16^–/+ CD56^–/+) (Fig. 2c) or of mature NK cells (CD3^– CD16⁺ CD56⁺) (Fig. 2e), E-IND children showed a higher percentage of pre-NK cells (CD3^– CD16⁺ CD56^–) by comparison with NI-1 children (Fig. 2d).

Decreased percentages of NKT cells (CD3⁺ CD16^– CD56⁺) are associated with the early indeterminate clinical form of Chagas’ disease

NKT cells are a unique T lymphocyte subpopulation, distinct from conventional T cells, because they express surface markers of both T cell and NK cell subsets. They can provide protection against infectious diseases by rapidly producing cytokines, through their cytolytic activity or via stimulation of other cell populations [22]. In order to quantify the frequency of circulating NKT cells, we used the same three-colour flow cytometry platform described for enumeration of NK cell subpopulations. Data analysis was performed by classifying NKT cells as NKT1 (CD3⁺ CD16⁺ CD56^–), NKT2 (CD3⁺ CD16^– CD56⁺) or NKT3 (CD3⁺ CD16⁺ CD56⁺), as proposed by Vitelli-Avelar et al. [6]. Our results revealed a significantly lower frequency of the NKT2 (CD3⁺ CD16^– CD56⁺) subset in E-IND samples by comparison with NI-1 samples (Fig. 3b). No differences were observed when the values of NKT1 and NKT3 subsets were evaluated (Fig. 3a,c).

Fig. 3.

Analysis of regulatory T cells [natural killer (NK) T and CD4⁺CD25^HIGH] in the peripheral blood of early indeterminate Trypanosoma cruzi-infected children (E-IND, •) and non-infected children (NI-1, ○). NKT phenotypic studies were performed by a triple-labelling protocol using anti-CD3 fluorescein isothiocyanate (FITC), anti-CD56 phycoerythrin (PE) and anti-CD16 TC to identify NKT subsets including NKT1 cells CD3⁺ CD16⁺ CD56^–/CD3⁺ (a), NKT2 cells CD3⁺ CD16^– CD56⁺/CD3⁺ (b) and NKT3 cells CD3⁺ CD16⁺ CD56⁺/CD3⁺ (c) analysed within gated CD3⁺ lymphocytes. Regulatory T cells were identified through a double staining procedure with anti-CD4 FITC and anti-CD25 PE monoclonal antibodies to identify regulatory CD4⁺ CD25^HIGH T cells (d) within gated lymphocytes. Data are expressed as scattering of individual values and mean percentage of cells. Significant differences (connecting lines) and P-values are shown on the figure.

Decrease of circulating CD4⁺ CD25^HIGH T cells highlights impaired immunoregulation in T. cruzi-infected children

In humans, it has been proposed that only the CD4⁺ CD25^HIGH population, comprising ∼1–2% of circulating CD4⁺ T cells, exhibits regulatory functions [17]. Enumeration of CD4⁺ CD25^HIGH regulatory T cells was carried out by first gating on lymphocytes based on their morphometric features on forward- versus side-scatter dot plots, followed by the selection of CD4⁺ cells presenting high expression of CD25 [17]. Our results demonstrated that lower values of CD4⁺ CD25^HIGH regulatory T cells are observed in E-IND samples than in NI-1 samples (Fig. 3d).

Using the same gating strategy described previously to evaluate E-IND and NI-1 samples, we performed a parallel investigation of major peripheral blood leucocyte phenotypes of IND, CARD and NI-2 subjects, including: CD14⁺ CD16⁺ (macrophage-like), CD14⁺ CD16⁺ HLA-DR⁺⁺ (proinflammatory monocytes), CD4⁺ CD25^HIGH (regulatory T cells), CD8⁺ HLA-DR⁺ (activated CD8⁺ T cells), CD3^– CD16⁺ CD56^– (pre-NK cells), CD3^– CD16⁺ CD56⁺ (mature NK cells) and CD3⁺ CD16^– CD56⁺ (NKT2 cells) (Fig. 4). Our data demonstrated a higher value of circulating macrophage-like, regulatory T cells, mature NK cells and NKT cells in IND than in NI-2 samples (Fig. 4a,c,f,g, respectively). Interestingly, the value of regulatory T cells observed in IND samples was also significantly higher than that observed in samples from CARD patients (Fig. 4c). Basal values (reference average of cells observed in healthy individuals) of proinflammatory monocytes and low levels of pre-NK cells were also observed in IND samples compared to NI-2 samples (Fig. 4b,e). It was remarkable to note that the IND group presented divergent data regarding the value of activated CD8⁺ T cells, with some individuals presenting low basal levels of activated CD8⁺ T cells (median value = 3·9%) and others displaying extremely high levels of CD8⁺ HLA-DR⁺ cells/CD8⁺ T cells (median value = 74·5%), suggesting the existence of distinct subgroups of individuals (Fig. 4d, dotted rectangles). Which immunological feature could compensate the high levels of cellular immune response in these individuals in order to maintain the asymptomatic disease? To answer this question, we assessed these phenotypic features at an individual level, which pointed out that all IND patients who presented with a high value of CD8⁺ HLA-DR⁺ cells also had low values of regulatory T cells (Fig. 5, left panel, top graph). Moreover, these individuals also presented immunophenotypes that suggest a more active role of innate cellular response, because they also displayed higher values of proinflammatory monocytes and mature NK cells (Fig. 5, left panel, top graph). Confirmatory analysis was carried out by correlation studies that further validated these findings, showing a negative correlation between CD8⁺ HLA-DR⁺ and CD4⁺ CD25^HIGH cells and a positive association between CD8⁺ HLA-DR⁺ and CD14⁺ CD16⁺ HLA-DR⁺⁺ cells, with the latter also correlated with the frequency of CD3^– CD16⁺ CD56⁺ cells (Fig. 5, left panels, bottom graphs).

Fig. 4.

Analysis of major discriminatory immunophenotypes among individuals with indeterminate disease (IND,) or cardiac disease (CARD, ) and uninfected adults (NI-2, ). Phenotypic studies were performed using a double or triple-labelling protocol to identify macrophage-like CD14⁺ CD16⁺ cells (a), proinflammatory monocytes (CD14⁺ CD16⁺ HLA-DR⁺⁺, b), regulatory T cells (CD4⁺ CD25^HIGH, c), activated CD8⁺ HLA-DR⁺ T cells (d), pre-natural killer (NK) cells (CD3^– CD16⁺ CD56^–/CD3^– CD16^–/+ CD56^–/+, e), mature NK cells (CD3^– CD16⁺ CD56⁺/CD3^– CD16^–/+ CD56^–/+, f) and NKT2 cells (CD3⁺ CD16^– CD56⁺, g). The results are expressed in box-plot format. The box stretches from the lower hinge (defined as the 25th percentile) to the upper hinge (the 75th percentile) and therefore contains the middle half of the scores in the distribution. The median is shown as a line across the box. Therefore 1/4 of the distribution is between this line and the top of the box and 1/4 of the distribution is between this line and the bottom of the box. Significant differences compared with NI-2 and CARD are indicated by letters a and c, respectively, at P < 0·05.

Fig. 5.

Analysis of major discriminatory immunophenotypes of indeterminate (IND, left panels) and cardiac patients (CARD, right panels). Analysis of individual data from IND demonstrates an association (dotted rectangles and lines) between high frequency of CD8⁺ HLA-DR⁺ cells in IND (○) with low frequency of regulatory T cells CD4⁺ CD25^HIGH, high frequency of proinflammatory monocytes (CD14⁺ CD16⁺ HLA-DR⁺⁺) and high levels of mature natural killer (NK) cells (CD3^– CD16⁺ CD56⁺). Analysis of individual data from CARD demonstrates an association (dotted rectangles and lines) between the high frequency of CD8⁺ HLA-DR⁺ cells (•) with low levels of mature NK cells (CD3^– CD16⁺ CD56⁺) and NKT cells (CD3⁺ CD16^– CD56⁺). Confirmatory correlation analysis validates the negative association between CD8⁺ HLA-DR⁺ and CD4⁺ CD25^HIGH cells and a positive association between CD8⁺ HLA-DR⁺ and CD14⁺ CD16⁺ HLA-DR⁺⁺ cells as well as CD14⁺ CD16⁺ HLA-DR⁺⁺ cells and mature NK cells (CD3^– CD16⁺ CD56⁺) in IND. Analysis performed with data obtained from CARD validates the negative association between CD8⁺ HLA-DR⁺ with mature NK cells (CD3^– CD16⁺ CD56⁺) as well as NKT2 cells (CD3⁺ CD16^– CD56⁺). Correlation analysis (r and P-values) are shown in the figure.

Increased percentage of activated CD8⁺ T cells and basal values of NK, NKT and regulatory T cells are major phenotypes related to late cardiac Chagas’ disease

Analysis of major cardiac Chagas’ disease discriminatory immunophenotypic features revealed that high levels of circulating CD8⁺ HLA-DR⁺ cells is the hallmark of the CARD group by comparison with the NI-2 group (Fig. 4d). Additional analysis at an individual level further demonstrated that in the CARD group, individuals displaying higher levels of CD8⁺ HLA-DR⁺ cells (higher than median value = 27·5%) also presented lower levels of mature NK cells, and were confined within a subgroup showing a low value of NKT cells (Fig. 5, right panel, top graph). Confirmatory analysis by correlation studies validated these findings, showing a negative correlation between CD8⁺ HLA-DR⁺ cells and both mature NK and NKT cell subpopulations (Fig. 5, right panel, bottom graphs).

Discussion

Understanding the role of immune responses to T. cruzi and the mechanisms of injury in Chagas’ disease has been a major challenge. T. cruzi infection simultaneously triggers multiple compartments of the innate and adaptive immune system. It is possible that the strong immune stimulation and the intense inflammatory process elicited during early infection by T. cruzi [3] could be not only a major determinant of the immunopathology of the late disease, but could also be a crucial factor in confining the aetiological agent to an intracellular site, controlling the consequences of life-long infection and preventing tissue damage [4,23–25]. However, the early stage of T. cruzi infection has been studied mainly in experimental mouse models, and the precise mechanism underlying the immunological events in humans is poorly understood [5,24].

The present studies involved a cross-sectional investigation of major and minor changes in peripheral blood leucocyte subpopulations during early and late phases of Chagas’ disease. The subjects included T. cruzi-infected children in the early stages of the IND clinical form of disease as well as chagasic adults typifying late chronic Chagas’ disease.

Our findings showed a lower value of T cells, due mainly to a drop in the value of CD8⁺ T cells, in addition to an increase in activated B cells and impaired T cell activation are hallmarks of early indeterminate Chagas’ disease (Fig. 1 and Table 2). The data presented here are consistent with our previous results from phenotypic characterization of peripheral blood leucocytes from early T. cruzi infection in Bolivian children [5]. This mixed activated/modulated immunological status can be explained partially by the action of distinct T. cruzi surface molecules that induce activation of B cells with non-specific Ig secretion [3,26,27] as well as suppression of T cell activation [28–30]. Consistent with this hypothesis, several studies have hypothesized that T. cruzi-derived glyco-inositol-phospholipids (GIPLs) and B cell activation could play a role in the conspicuous Ig production observed during infection, whereas membrane glycosyl-phosphatidyl-inositol (GPI)-anchored molecules are able to trigger suppression of human T cell response [29,30]. It has been suggested that the polyclonal activation of B cells and the T cell anergy may represent the mechanism of parasite evasion, i.e. misleading the immunological system and preventing the development of a strong adaptive immune response, thereby favouring disease onset and immunopathology [31].

The importance of NK cells in resistance to acute Chagas’ disease is illustrated by studies showing that neutralization of endogenous interleukin (IL)-12 or interferon (IFN)-γ as well as depletion of NK cells renders animals more susceptible to infection with T. cruzi [32,33]. Thus, NK cells are an important source of IFN-γ, before development of T cell-mediated immunity. Besides cytotoxic activity and cytokine secretion, NK cells can control B cell Ig secretion independent of T cell induction [27,34]. The higher frequency of pre-NK cells (Fig. 2b) reported here might be related to the early activation of B cells (Fig. 1g), contrasting with cell phenotypes pointing to impaired T cell activation (Table 2).

CD16⁺ CD56^– pre-NK cells have been considered to be precursors of functional and phenotypically distinct mature CD16⁺ CD56⁺ NK cells [21,35]. Pre-NK cells have higher proliferative capacity and are better sources of cytokines, whereas mature NK cells display mainly cytotoxic activities [35]. Our data suggest that the expansion of pre-NK cells might be related to important mechanisms of macrophage activation during early indeterminate Chagas’ disease. Macrophages are efficiently activated by NK derived IFN-γ, which invokes nitric oxide production and controls parasite replication during the early stages of T. cruzi infection [36–39]. Despite the low levels of circulating macrophage-like cells, our results demonstrated an increased frequency of CD14⁺ CD16⁺ HLA-DR⁺⁺ proinflammatory monocytes [19] among circulating CD14⁺ CD16⁺ cells in infected children (Fig. 2a,b).

It is important to point out that strong, uncontrolled activation of NK cells as well as proinflammatory monocytes may also lead to tissue damage leading to the development of cardiomyopathy and/or digestive megas [40,41]. Thus, the establishment of immunoregulatory mechanisms seems to be an important key to controlling immune activity and preventing deleterious effects of excessive stimulation of the immune system that may lead to fatality. Current and previous reports have suggested that, in human liver, NKT cells may play an important role in eliminating autologous cytotoxic T cells via apoptosis of activated CD8⁺ T cells [42,43]. Moreover, it has been also proposed that IFN-γ and perforin production as well as NK and CD8⁺ T cell cytotoxicity are efficiently regulated by CD4⁺ CD25⁺ regulatory T cells [44]. We have described here a lower value of NK T cells, as well as CD4⁺ CD25^HIGH regulatory T cells, in infected children (Fig. 3b,d), consistent with a higher levels of NK cells and proinflammatory monocytes (Fig. 2d,b). These findings suggest that the inability to shift the immune response toward higher levels of CD4⁺ CD25^HIGH may contribute to the development of cardiac tissue damage.

The low values of NKT and regulatory CD4⁺ CD25^HIGH cells during early Chagas’ disease raise the question of why no phenotypic features related to T cell activation can be observed in the peripheral blood of E-IND patients. We believe that T cell-mediated immunity during the early indeterminate clinical form of Chagas’ disease may represent a phenomenon restricted to the inflammatory sites, not detectable in the peripheral blood, considering previous reports describing the presence of these cells in the cardiac inflammatory infiltrate during early human Chagas’ disease [10]. This hypothesis is supported by our findings of a higher percentage of CD8⁺ T lymphocytes carried by CD54, an important adhesion molecule involved in migration pathways from the bloodstream to tissue inflammatory sites. We hypothesize here that the increased levels of circulating CD8⁺ CD54⁺ T cells reflect incipient immunological events during early chronic Chagas’ disease, suggesting the enhanced migratory potential of this cytotoxic population to control tissue parasitism.

Once we determined the immunophenotypic profile of circulating leucocytes during early Chagas’ disease, we then investigated the major discriminatory phenotypes during late chronic indeterminate and cardiac Chagas’ disease. Comparative cross-sectional analysis of major immunophenotypes exhibited by late chronic chagasic patients with those exhibited by patients bearing early indeterminate disease suggested that a shift towards high values of macrophage-like cells, together with basal values of proinflammatory monocytes, regulatory CD4⁺ CD25^HIGH T cells and high levels of mature NK cells and NKT cells, would be responsible for development of late chronic asymptomatic disease (Fig. 4). On the other hand, the development of a cell-mediated inflammatory immunoprofile characterized by high levels of activated CD8⁺ HLA-DR⁺ T cells in the presence of basal levels of mature NK cells, NKT cells and regulatory CD4⁺ CD25^HIGH cells would account for the development of late chronic cardiac disease (Fig. 4).

It is important to observe that unlike the CARD patients (Fig. 5, right panel, top graph), the IND patients that present high levels of activated CD8⁺ HLA-DR⁺ T cells (Fig. 5, left panel, top graph) also count with high levels of mature NK cells (CD3^– CD16⁺ CD56⁺ cells) that may contribute to the establishment/maintenance of their asymptomatic clinical status. We have reported previously that blood samples from patients with the late indeterminate clinical form of Chagas’ disease display a higher value of CD4⁺ CD25^HIGH and NKT (CD3⁺ CD16^– CD56⁺) regulatory cells, as well as increased levels of circulating NK cells. In the present study, we have also shown a correlation between the high levels of CD4⁺ CD25^HIGH T cells and the low frequency of activated CD8⁺ T cells (Fig. 5, left panel, bottom graphs). We have also documented previously the existence of an increased frequency of activated CD8⁺ HLA-DR⁺ T cells and low levels of CD4⁺ CD25^HIGH in patients with severe clinical forms of Chagas’ disease [6]. In the present study, we have addressed this issue further, demonstrating that patients bearing cardiac Chagas’ disease display, in addition to the high levels of activated CD8⁺ T cells, an opposite immunological profile of low values of NK and NKT cells (Fig. 5, right panel, top graph).

Taken together, our findings suggest that the expansion of proinflammatory monocytes CD14⁺ CD16⁺ HLA-DR⁺⁺ as well as high values of pre-NK cells, in a microenviroment deficient in NKT cell and CD4⁺ CD25^HIGH cell populations, represent an important immunological profile that controls parasite load in E-IND Chagas’ disease. However, the persistence of this immunophenotypic pattern parallels the establishment of a strong adaptive CD8⁺ T cell activation that could lead to late chronic disease associated with cardiac damage. On the other hand, the shift of this immunological pattern towards high values of macrophage-like cells, together with enhanced frequency of mature NK cells, NKT cells and regulatory CD4⁺ CD225^HIGH T cells, could be beneficial, limiting tissue damage and leading to lifelong persistence of the indeterminate form of Chagas’ disease.