Analysis of regulatory T cell associated forkhead box P3 expression in the lungs of patients with sarcoidosis

Abstract

In pulmonary sarcoidosis, the typical T helper 1-mediated immune response in the lungs has been proposed to be co-ordinated by regulatory T cells; however, their exact role needs to be clarified. We used real-time polymerase chain reaction to study genes involved in regulatory T cell functions in CD4⁺ T cells isolated from bronchoalveolar lavage fluid (BALF) of patients (n = 24) and healthy subjects (n = 7). The genes included the transcription factor forkhead box P3 (FoxP3), interleukin (IL)-10, transforming growth factor-β1 and chemokine receptor 2 (CCR2). The same genes were also studied in isolated BALF CD4⁺ T cell receptor AV2S3⁺ and AV2S3^- T cells of patients with lung-restricted AV2S3 T cell expansions (n = 12). Intracellular staining of the FoxP3 protein was performed additionally in 14 patients and nine healthy subjects. mRNA expression of FoxP3, CCR2 and IL-10 was decreased significantly in BALF CD4⁺ T cells of patients. Flow cytometric analysis of CD4⁺ T cells also demonstrated a decreased frequency of FoxP3⁺ cells in the BALF and blood of sarcoidosis patients as well as a reduced intensity (mean fluorescence intensity) of FoxP3 expression in BALF FoxP3⁺ cells of patients. BALF CD4⁺AV2S3⁺ T cells expressed significantly lower levels of FoxP3 and CCR2 mRNA versus BALF CD4⁺AV2S3^- T cells. The main conclusion of our study is that there is a reduced expression of regulatory T cell associated genes in BALF CD4⁺ T cells in sarcoidosis. In addition, our data suggest an effector function of AV2S3⁺ lung-accumulated T cells in sarcoidosis.

Introduction

Sarcoidosis is a T helper 1 (Th1)-mediated inflammatory disease, affecting primarily the lungs and characterized by non-caseating epithelioid cell granulomas consisting of lymphocytes, histocytes and giant multi-nucleated cells [1].

Studies on the T cell repertoires in the bronchoalveolar lavage fluid (BALF) cells revealed that expansions of CD4⁺ T cells expressing the AV2S3 T cell receptor (TCR) gene segment in the lung correlated well with human leucocyte antigen (HLA)-DRB1*0301 and HLA-DRB3*0101 in Scandinavian sarcoidosis patients [2,3]. The expansion of AV2S3⁺ cells at disease onset was found to correlate with a better prognosis, suggesting a protective role for these cells in this disease [4]. In addition, HLA-DRB1*0301 has been associated with a favourable clinical outcome in sarcoidosis patients. DRB1*0301 positive sarcoidosis patients often have Löfgren's syndrome, i.e. an acute onset of the disease with bilateral hilar lymphadenopathy, erythema nodosum and/or bilateral ankle arthritis, with a better prognosis and spontaneous recovery within 2 years [5]. In contrast, patients with non-Löfgren's disease often show an insidious onset with dry cough, low-grade fever, fatigue, shortness of breath, weight loss and more pronounced chest radiographic changes.

More recently, we described a decreased expression of Th1-associated cytokines in HLA-DRB1*0301^pos compared with HLA-DRB1*0301^neg patients, indicating a down-regulated immune response which could be an important feature of patients with resolving disease [6].

Moreover, increased levels of regulatory T cells (T_reg cells) have been reported in sarcoidosis patients [7,8]. These T cells play an important role in maintaining peripheral tolerance [9]. Isolation of pure T_reg cells is not easy to achieve, as their associated surface markers, e.g. the cytotoxic T lymphocyte-associated antigen-4, glucocorticoid-induced tumour necrosis factor (TNF) receptor and interleukin (IL)-2 receptor (CD25) are also expressed on activated T cells [9]. Recently, a member of the forkhead/winged helix family of transcription factors, forkhead box P3 (FoxP3), has been identified as an essential factor for thymic development and function of T_regs[10]. However, the intracellular localization of FoxP3 exerts limitations on isolation of T_regs based on this molecule. Instead, analysis of FoxP3 mRNA and protein levels has provided useful knowledge on regulatory T cells [11,12]. The previously reported increased levels of T_regs in sarcoidosis were based on CD25 expression [7,8]. However, as CD25 is not a unique marker of T_regs, the potential role of regulatory T cells in sarcoidosis must be studied further by analysing more specific markers such as FoxP3.

In addition, studies on mice showed that chemokine receptor 2 (CCR2) is expressed on T_reg cells [13,14], and purified CCR2⁺CD4⁺ T cells were able to suppress proliferation of T and B cells in vitro[14]. Mice lacking CCR2 exhibited significantly lower amounts of FoxP3 mRNA than control mice [15]. Studies on cancer patients also revealed a suppressor function for CCR2⁺ cells [16]. Thus, CCR2 is a suitable candidate molecule to be analysed in relation to T_regs in sarcoidosis.

Moreover, the role of T_regs in the immunopathology of sarcoidosis may be strengthened by the fact that IL-10 and transforming growth factor (TGF)-β, two cytokines with anti-inflammatory capacities that are also produced by T_regs, have been suggested to be involved in spontaneous disease remission in sarcoidosis patients [6,17–19].

In the present study we analysed mRNA expression for CCR2, FoxP3, IL-10 and TGF-β1 in CD4⁺ T cells sorted by means of fluorescence activated cell sorter (FACS) flow cytometry from BALF of patients with active sarcoidosis and from healthy controls, as well as in lung-accumulated BALF CD4⁺AV2S3⁺/^- T cells from patients with lung-restricted AV2S3 T cell expansions. FoxP3 expression in CD4⁺ lymphocytes in BALF and blood was also analysed using flow cytometry in sarcoidosis patients and in healthy controls.

Methods

Study subjects

Polymerase chain reaction (PCR) analysis was performed on samples obtained from 33 patients with active sarcoidosis, included consecutively as they were referred to the Division of Respiratory Medicine, Karolinska University Hospital, Stockholm, Sweden. Bronchoalveolar lavage (BAL) samples were obtained usually within 3 months after disease onset. Patients had a clinical picture in accordance with pulmonary sarcoidosis, as determined by symptoms (such as cough, shortness of breath and fatigue), chest radiography and pulmonary function tests, and the diagnosis was established using the criteria by the World Association of Sarcoidosis and other Granulomatous disorders [1]. All patients were non-smokers and no patient was on treatment with immunosuppressive drugs. Patients were divided into two groups: those with Löfgren's syndrome (n = 20) and those without (n = 13).

BALF CD4⁺ T cells were isolated from 24 of the patients: 11 with Löfgren's syndrome and 13 non-Löfgren's patients. FACS-sorted BALF CD4⁺ T cells from seven non-smoking healthy adults were included as controls.

In addition, from 12 of the patients we isolated BALF CD4⁺ T cells expressing either the AV2S3 TCR gene segment (CD4⁺AV2S3⁺) or not (CD4⁺AV2S3^-). All had lung-restricted AV2S3⁺ T cell expansions (≥ 10·5% of CD4⁺ cells in BALF) [3] and all except one had Löfgren's syndrome. In order to obtain a sufficient number of CD4⁺AV2S3⁺/^- T cells, in all but three patients we had to choose to sort either CD4⁺ T cells or CD4⁺AV2S3^+/− T cells.

Additionally, to analyse FoxP3 protein expression, FACS analysis was performed on BALF T cells of 14 patients and nine controls and on blood T cells of seven patients and nine control subjects.

All subjects gave their informed consent to participate in the study, and the local ethics committee approved the study.

Bronchoalveolar lavage

Bronchoalveolar lavage was performed and cells were prepared as described [20].

Flow cytometric analysis and isolation of cells

Bronchoalveolar lavage fluid CD4/CD8 T lymphocyte ratio and TCR AV2S3 expression was analysed by flow cytometry as described previously [6]. For sorting, cells were stained with anti-CD4-phycoerythrin (PE) (Dako Cytomation Norden AB, Solna, Sweden) and anti-human AV2S3 TCR (clone F1)-fluorescein isothiocyanate (Pierce Biotechnology, Rockford, USA). The stained cells were sorted by FACSVantage (BD Biosciences, Montain View, CA, USA). BALF cells were gated on lymphocytes, identified by forward- and side-scatter characteristics, and sorted into three different populations; CD4⁺ T cells from patients and controls, as well as CD4⁺ AV2S3⁺ and CD4⁺AV2S3^- T cells from patients with lung-accumulated T cells expressing the AV2S3 TCR gene segment. The purity of the sorted populations, which was determined by FACS (FACSCanto II; BD Biosciences), was 98% on average.

We were also able to analyse FoxP3 protein expression in blood and BALF CD4⁺ T cells as well as in BALF CD4⁺AV2S3⁺ and CD4⁺AV2S3^- T cells, using flow cytometry. Anti-FoxP3-PE (clone PCH101-PE) and isotype control (rat IgG2a-PE) and staining kit were purchased from eBioscience (Biosciences, San Diego, CA, USA). Expression of cell-surface markers and intracellular FoxP3 was determined by flow cytometry after gating on CD4⁺ lymphocytes. The data were analysed using FACS Diva software (BD Biosciences).

Quantitative analysis of the gene expression by real-time PCR

Total RNA was extracted and cDNA was synthesized. Gene expression was quantified by real-time PCR using ABI Prism 7700 Sequence Detection System (Applied Biosystems, Foster City, CA, USA), as described previously [6]. RNA specimens were analysed in duplicate using primers and probes for β-actin, IL-10 and TGF-β1 [6]. The assay-on-demand products for FoxP3 (Hs00203958_-m1), CCR2 (Hs00356601_-m1) and universal master mix were purchased commercially (Applied Biosystems). The amount of target gene was normalized to β-actin (housekeeping gene) and the relative expression of a gene in BALF was calculated in relation to the mean value of target gene expression in the healthy control group.

Human leucocyte antigen typing

Human leucocyte antigen class II (HLA-DR) typing was performed on DNA using the PCR and amplification with sequence specific primers [21].

Statistics

Significance levels were calculated according to non-parametric tests, using the Kruskal–Wallis test followed by Dunn's post-test for comparisons between groups or Mann–Whitney U-test for comparison between two groups. Correlations between different parameters were determined with Spearman's rank correlation test. Values of P < 0·05 were regarded as significant. All statistical analyses were performed with Graphpad Prism 4·03 (Graphpad Software Inc., San Diego, CA, USA).

Results

Bronchoalveolar lavage analysis and lung function parameters

Results reflecting the inflammatory intensity in the BALF and pulmonary function tests are presented in Table 1. The results show increased BALF cell concentrations in each patient subgroup and increased percentages of lymphocytes in BALF from non-Löfgren's patients compared with controls, while the percentages of macrophages were decreased in non-Löfgren's patients. There were also significant differences in the BALF cell concentrations and the frequencies of lymphocytes and macrophages when Löfgren's patients were compared with non-Löfgren's patients. The vital capacity and forced expiratory volume in 1 s were lower in the non-Löfgren's patients versus Löfgren's patients (Table 1a). Characteristics of subjects that were included for analysing protein expression of FoxP3, using flow cytometry, are presented in Table 1b.

Table 1.

Bronchoalveolar lavage fluid (BALF) analysis and lung function parameters.

	Löfgren's (n = 20)	Non-Löfgren's (n = 13)	Controls (n = 7)
(a) Samples used for PCR
Sex, male/female	9/11	6/7	3/4
Age, year	36·5 (25–59)**†	51 (34–75)***	27 (21–30)
X-ray stage (0/I/II/III)	0/13/7/0	0/4/6/3	7/0/0/0
BAL analyses
% recovery	73·3 (44–82)	68 (44–80)	72 (61–85)
% viability	95 (85–99·6)	96 (92–98·2)	96 (88–98)
Cell concentration (×106/l)	199·5 (49·1–588·2)*§	257·3 (159–746·1)***	113 (50–167)
Differential cell counts
% macrophages	74·1 (41·6–91)‡	54·8 (36·7–73)***	87 (71–90·5)
% lymphocytes	24·9 (8·5–57·8)‡	44·8 (21·8–61)***	11·2 (9–29)
% neutrophils	1·0 (0–4·8)	1·0 (0–6·0)	1·0 (0·2–4·4)
% eosinophils	0·2 (0–4·6)	0·6 (0–4·0)	1·1 (0–1·6)
CD4/CD8 ratio	8·7 (2·4–28·4)	8·3 (1·6–33)	n.d.
HLA-DRB1*0301	19 of 20	None	n.d.
AV2S3 expansion¶	All	1 of 13	n.d.
Pulmonary function tests
VC (% of ref. value)	90·5 (68–126)†	78 (53–133)	n.d.
FEV1 (% of ref. value)	89 (66–122)†	73 (56–131)	n.d.
Dlco (% of ref. value)	86 (70–126)	78·5 (54–106)	n.d.

	Löfgren's (n = 7)	Non-Löfgren's (n = 7)	Controls (n = 9)
(b) Samples used for FACS
Sex, male/female	4/3	5/2	6/3
Age, year	43 (25–55)	48 (31–71)*	25 (18–46)
X- ray stage (0/I/II/III/IV)	1/3/3/0/0	0/0/5/1/1	9/0/0/0
BAL analyses
% recovery	75 (55–77)	67 (27–84)	74 (56–79)
% viability	91(79–95)	94 (89–97)	97 (88–99)
cell concentration (*106/l)	135 (93–390)**	203 (85–386)**	73 (54–109)
Differential cell counts
% macrophages	60 (37–88)*	70 (35–85)*	87 (80–93)
% lymphocytes	39 (10–55)*	28 (14–64)*	11 (5·6–19)
% neutrophils	1·7 (0·8–4·5)	1·1 (0·2–2)	2 (0·6–3·8)
% eosinophils	0·2 (0–3·6)	0·3 (0–0·4)	0·2 (0–2·0)
CD4/CD8 ratio	6·2 (2·3–25·7)	6·5 (3·4–14·1)	n.d.
HLA-DRB1*0301	2 of 5 (2nd)	1 of 3 (4 n.d.)	n.d.
AV2S3 expansion	4 of 7	1 of 7	n.d.

^*P < 0·05 versus healthy controls

^**P < 0·01 versus healthy controls

^***P < 0·001 versus healthy controls.

^†P < 0·05 between patient subgroups

^‡P < 0·01 between patient subgroups

^§P = 0·053 between patient subgroups.

^¶Defined as at least three times higher than the corresponding median value in peripheral blood mononuclear cells from healthy controls, i.e. ≥ 10·5% [36]. Data are shown as median (min–max). VC, vital capacity; FEV₁, forced expiratory volume in 1 second; Dlco, diffusing capacity of the lung for carbon monoxide; n.d., not done; PCR: polymerase chain reaction; HLA: human leucocyte antigen; FACS: fluorescence activated cell sorter.

mRNA expression in CD4⁺ T cells

CD25 is not a unique marker of T_reg cells as it is also expressed on activated T cells, while FoxP3 is a crucial transcription factor for the development and function of CD4⁺ T_reg cells that can be used as a reliable marker for functional regulatory T cells. We, therefore, investigated FoxP3 mRNA expression in BALF CD4⁺ T cells from sarcoidosis patients and healthy donors. A significant reduction in FoxP3 mRNA expression was observed in sarcoidosis patients (P = 0·04, Fig. 1a) compared with controls. CCR2 and IL-10 mRNA levels were also decreased in sarcoidosis patients (P = 0·005, P = 0·02 respectively; Fig. 1b and c) compared with controls. In addition, the significantly decreased levels of CCR2 and IL-10 mRNA in patients were found to be confined to non-Löfgren's patients (P < 0·01, P < 0·05 respectively, versus controls; Fig. 2b and c).

Fig. 1.

Relative mRNA expressions of (a) forkhead box P3 (FoxP3), (b) chemokine receptor 2 (CCR2), (c) interleukin (IL)-10 and (d) transforming growth factor (TGF)-β1 in isolated CD4⁺ bronchoalveolar lavage fluid (BALF) cells of patients with sarcoidosis (n = 24) and healthy controls (n = 7). The relative mRNA expression of FoxP3, CCR2 and IL-10 was decreased in BALF CD4⁺ T cells from patients versus healthy controls (P = 0·04, P = 0·005, P = 0·02 respectively). No significant difference in TGF-β1 mRNA expression was detected. Horizontal bars indicate median values. The P-values were calculated using the Mann–Whitney U-test. **P < 0·01, *P < 0·05.

Fig. 2.

The relative RNA transcript for (a) forkhead box P3 (FoxP3), (b) chemokine receptor 2 (CCR2), (c) interleukin (IL)-10 and (d) transforming growth factor (TGF)-β1 in bronchoalveolar lavage fluid CD4⁺ T cells of patient subgroups [Löfgren's patients (n = 11) and non-Löfgren's patients (n = 13) and healthy controls (n = 7)] is shown. The relative CCR2 and IL-10 mRNA levels were decreased significantly in non-Löfgren's patients compared with controls. Horizontal bars indicate median values. The P-values were calculated using the Kruskal–Wallis test followed by Dunn's post-test. **P < 0·01, *P < 0·05.

We found no significant differences for TGF-β1 mRNA, which was expressed in BALF CD4⁺ T cells of all individuals (Fig. 1d). No significant differences in gene expression were found between patient subgroups (Fig. 2), but a tendency to decreased CCR2 mRNA level in non-Löfgren's patients compared with Löfgren's patients (Fig. 2b).

Correlation between gene expressions and BALF parameters

When investigating associations between gene expression and BALF cellular parameters in the whole patient group, it was found that CCR2 mRNA levels correlated negatively with the percentages of BALF lymphocytes (r = −0·48, P = 0·02; Fig. 3a). In contrast, the levels of CCR2 were correlated positively with the frequency of BAL lymphocytes in healthy controls (r = 0·66, P = 0·16; Fig. 3b).

Fig. 3.

Correlation between chemokine receptor 2 (CCR2) mRNA expression and bronchoalveolar lavage fluid (BALF) percentages of lymphocytes. The relative CCR2 mRNA expression in BALF CD4⁺ T cells correlated negatively with the frequency of BALF lymphocytes in (a) patients with sarcoidosis (r =−0·47; P = 0·02), while it correlated positively with the frequency of BALF lymphocytes in (b) healthy subjects (r = 0·66, P = 0·16). The correlations were analysed using Spearman's rank correlation test. •: Löfgren's patients; ○: non-Löfgren's patients.

A significant correlation between FoxP3 and CCR2 mRNA levels was detected in isolated CD4⁺ T cells from Löfgren's patients (r = 0·65, P = 0·03; Fig. 4a), and a similar tendency was seen in healthy subjects (r = 0·75, P = 0·066). A significant negative correlation was observed between FoxP3 mRNA expression and the frequency of BALF lymphocytes (r =−0·84, P = 0·001; Fig. 4b) in Löfgren's patients.

Fig. 4.

The levels of forkhead box P3 (FoxP3) and chemokine receptor 2 (CCR2) mRNA were measured in Löfgren's patients. (a) The levels of FoxP3 and CCR2 mRNA correlated positively in Löfgren's patients (n = 11; r = 0·65, P = 0·03). (b) The mRNA expression of FoxP3 was correlated negatively with the frequency of bronchoalveolar lavage fluid lymphocyte in Löfgren's patients (r = −0·84, P = 0·0013). The correlation was analysed using Spearman's rank correlation test.

In addition, both FoxP3 and IL-10 mRNA levels showed significant negative correlations with the frequency of BALF CD4⁺AV2S3⁺ T cells (r =−0·64, P = 0·05; r = −0·65, P = 0·04 respectively, data not shown) in Löfgren's patients.

CD4⁺AV2S3⁺ T cells versus CD4⁺AV2S3^- T cells

The mRNA expression of FoxP3, CCR2, IL-10 and TGF-β1 was evaluated and compared between isolated BALF CD4⁺ AV2S3⁺ and BALF CD4⁺AV2S3^- T cells of patients with lung-restricted AV2S3 T cell expansions (n = 12). The mRNA levels for FoxP3 and CCR2 were decreased significantly in AV2S3⁺versus AV2S3^- T cells (P = 0·0002 and P = 0·002 respectively, Fig. 5a, b). BALF CD4⁺AV2S3⁺ T cells also showed a tendency to a reduced IL-10 mRNA expression (P = 0·09, Fig. 5c), while no difference was detected in mRNA expression for TGF-β1 (Fig. 5d).

Fig. 5.

Relative mRNA expressions of (a) forkhead box P3 (FoxP3), (b) chemokine receptor 2 (CCR2), (c) interleukin (IL)-10 and (d) transforming growth factor (TGF)-β1 in isolated CD4⁺ AV2S3⁺ and CD4⁺ AV2S3^- bronchoalveolar lavage fluid (BALF) cells from patients with lung-restricted T cell receptor AV2S3⁺ T cell expansions (n = 12). The relative mRNA expression of FoxP3 and CCR2 was decreased in BALF CD4⁺ AV2S3⁺versus CD4⁺ AV2S3^- T cells (P = 0·0002, P = 0·002 respectively). No significant difference in IL-10 and TGF-β1 mRNA expression was detected. The P-values were calculated using the Mann–Whitney U-test. The lines indicate T cell subpopulations from the same patient. **P < 0·01, ***P < 0·001.

Intracellular staining of FoxP3

Intracellular staining showed a significantly decreased frequency of FoxP3-expressing BALF CD4⁺ T cells in patients (n = 14) compared with controls (n = 9) (P < 0·0001, Fig. 6a). Also the frequency of FoxP3-expressing blood CD4⁺ T cells was decreased in patients (n = 7) versus controls (n = 9) (P = 0·008, Fig. 6b). The difference in the frequency of CD4⁺ cells expressing FoxP3 in blood and BALF from patients was significant (P = 0·02, Fig. 6b), and in healthy controls we found a highly significant increased frequency of FoxP3-expressing CD4⁺ cells in BALF compared with blood (P = 0·0002, Fig. 6b).

Fig. 6.

Flow cytometric analysis of forkhead box P3 (FoxP3) protein expression in CD4⁺ T cells in bronchoalveolar lavage fluid (BALF) and blood as well as in BALF CD4⁺AV2S3⁺/^- T cells. (a) The frequency of FoxP3-expressing cells in BALF CD4⁺ T cells was decreased in patients with sarcoidosis (n = 14) versus controls (n = 9) (P < 0·0001). (b) The frequency of FoxP3-expressing cells was also decreased in blood CD4⁺ T cells of sarcoidosis patients (n = 7) compared with controls (n = 9) (P = 0·008). In healthy controls (n = 9) and patients (n = 7) the percentage of FoxP3-expressing CD4⁺ T cells was higher in BALF versus blood (P = 0·0002, P = 0·02 respectively). The lines indicate T cell subpopulations from the same individual. (c) The level of FoxP3 expression is presented as the mean fluorescence intensity (MFI) of FoxP3⁺ cells. The MFI of FoxP3 expression was decreased in BALF cells of patients compared with controls (P = 0·0002). In healthy controls, the intensity of FoxP3 expression was higher in BALF T regulatory cells versus in blood (P < 0·0001). (d) The frequency of FoxP3-expressing cells is lower in CD4⁺AV2S3⁺ T cells versus CD4⁺AV2S3^- T cells; one representative dot-plot of three is shown, demonstrating FoxP3 expression in 1·7% of CD4⁺AV2S3⁺versus 6·2% of CD4⁺AV2S3^- T cells. Horizontal bars indicate median values. The P-values were calculated using the Mann–Whitney U-test. **P < 0·01, ***P < 0·001.

The mean fluorescence intensity (MFI) of FoxP3 after deduction of background intensity of CD4⁺FoxP3^- cells was also reduced significantly in CD4⁺FoxP3⁺ BALF T cells from patients compared with controls (P = 0·0002, Fig. 6c). In contrast, there was no difference in the intensity of FoxP3 expression in CD4⁺FoxP3⁺ blood cells from patients versus controls. We found no difference in the level of FoxP3 expression in blood and BALF from patients, while a significantly increased MFI of FoxP3 expression in BALF versus blood in healthy controls was detected (P < 0·0001, Fig. 6c).

In addition, in three patients analysed, the percentage of cells that expressed FoxP3 was also decreased in CD4⁺AV2S3⁺ T cells versus CD4⁺AV2S3^- T cells (1·7% versus 6·2%, 2·3% versus 4·0% and 1·1% versus 3·6%, Fig. 6d).

Discussion

In the present study, using quantitative PCR and flow cytometry we surveyed regulatory T cells in BALF and blood of sarcoidoisis patients, particularly during the acute onset of the disease, and healthy controls. These cells were investigated through the study of the regulatory T cell-specific transcription factor, FoxP3 and of regulatory T-associated cytokines, TGF-β1, IL-10 as well as of CCR2 mRNA expression in CD4⁺ T cells isolated from the lungs of sarcoidosis patients and healthy controls.

Our data demonstrate a down-regulation of FoxP3 mRNA and a reduced frequency of FoxP3-expressing BALF CD4⁺ T cells in sarcoidosis patients. This reduced FoxP3 expression may reflect a lack of suppressive function of regulatory T cells, resulting in an augmented Th1 immune response. In healthy controls, we found a significantly higher frequency of FoxP3-expressing CD4⁺ T cells in BALF versus blood, in line with results from Hartl et al. [22]. In addition, we detected a higher intensity of FoxP3 expression in the lung T_reg cells of healthy controls. These data suggest a higher T_reg activity in the BALF compartment compared with blood in healthy individuals, which is absent in patients with sarcoidosis. Such a high T_reg activity in healthy lungs may be of importance for the maintenance of immune homeostasis. Earlier reports based on CD25 expression, in contrast, indicated that the number of T_regs in BALF of sarcoidosis patients is increased [7,8]. Because CD25 is also expressed by activated T cells it is, however, difficult to use CD25^bright expression as a marker for T_reg cells. In the present study we, therefore, analysed FoxP3 expression, which is a more reliable marker for T_reg cells. Although there is a major overlap between CD25^bright and FoxP3-expressing cells, they do not delineate exactly the same T cell populations [23]. In addition, the expression of FoxP3 in CD4⁺CD25^- T cells is reported to result in the achievement of a T_reg-like phenotype [10], which is in line with the demonstration that FoxP3 levels can determine the degree to which a T cell expresses a regulatory phenotype [24]. This is the first time that FoxP3 expression in BALF from both sarcoidosis patients and healthy control subjects has been compared. Two previous studies focused instead upon CD25^bright expression to define T_reg cells, and found increased numbers of T_reg cells in sarcoidosis [7,8]. Besides focusing on CD25 instead of FoxP3, these studies also differed from our study in other aspects. The characteristics of the patient populations differed between studies; for instance, Miyara et al. did not include Löfgren's patients, and furthermore the patients in the Miyara study were defined using different criteria. Moreover, the timing for obtaining BALF may differ in the two studies, and we compared our BALF data between patients and healthy controls, while Miyara et al. did not investigate healthy controls.

In the study by Miyara et al. the isolated BALF CD25^bright T cells from sarcoidosis patients exhibited anti-proliferative activity, but did not inhibit completely TNF-α and interferon (IFN)-γ production [7]. Such a limited functional capacity of T_regs to suppress TNF-α and IFN-γ secretion has also been reported in patients with rheumatoid arthritis [25]. Of interest, we observed that FoxP3 intensity (MFI) in the CD4⁺FoxP3⁺ BALF T cells of patients was decreased significantly, which may indicate lower suppressive activity in T_reg cells from patients than that in controls. Such a reduced capacity of T_regs expressing lower levels of FoxP3 has been demonstrated in other diseases [26–29]. Recently, evidence supporting a mutual counter-regulation between T effector cells and regulatory T cells was presented. For example, a suppressive function of TNF-α on regulatory T cells in patients with active rheumatoid arthritis was described and associated with down-modulation of FoxP3 [30]. Interestingly, high levels of TNF-α have been described in the lungs of patients with sarcoidosis [31,32]. The reduced FoxP3 expression in BALF CD4⁺ T cells may have significant effects on regulation of the sarcoidosis inflammation.

However, it may be speculated that the functional role of T_regs differs between patient subgroups, e.g. whether there is an offending antigen that can be eradicated or not may determine whether T_reg activity is detrimental or beneficial.

Treatment with infliximab, a chimeric anti-TNF-α monoclonal antibody that specifically inhibits TNF-α, has been associated with an improvement in selected cases with chronic sarcoidosis [33]. In addition, Valencia et al. showed that treatment with infliximab up-regulated FoxP3 mRNA and protein expression in CD4⁺CD25⁺ regulatory T cells, which recovered their suppressive function in patients with acute rheumatoid arthritis [30]. Glucocorticoid-induced up-regulation of FoxP3 expression in patients with asthma was also reported [34]. This effect of anti-TNF-α and cortisone treatment on T_reg cell recovery needs to be studied further in sarcoidosis; for example, does the response to therapy correlate with the capacity to up-regulate FoxP3?

A reduction of regulatory T cell functions in sarcoidosis is supported further by the observation (present data) that both FoxP3 and CCR2 mRNA correlate negatively with the frequency of BAL lymphocytes in sarcoidosis patients, implying that reduced regulatory T cell activity may lead to more intensive alveolitis. Alternatively, high levels of Th1 lymphocytes and their associated cytokines (e.g. TNF-α and IFN-γ) in the lungs of sarcoidosis patients may play a role in down-regulating FoxP3 and CCR2 expression in BALF CD4⁺ T cells. The situation in sarcoidosis patients is in contrast to that in healthy controls, where there is no strong Th1 bias in the lungs, and where the level of CCR2 mRNA is correlated positively with the frequency of BAL lymphocytes.

Furthermore, the positive correlation between FoxP3 and CCR2 mRNA expression in Löfgren's patients as well as in healthy controls suggests that FoxP3⁺ T_reg cells in the lungs normally express CCR2, consistent with the findings from other groups, who described a link between the presence of regulatory T cells and the expression level of CCR2 [14,15].

In a previous study, we reported an increase in IL-10 mRNA from unseparated BALF cells of sarcoidosis patients [6]. However, in the present study, where purified CD4⁺ T cells were analysed for expression of IL-10 mRNA, decreased levels were observed, which may suggest alveolar macrophages as an important source of IL-10 production in the lungs of sarcoidosis patients. In addition the decreased levels of CCR2 and IL-10 (Th2-related genes) were most pronounced in non-Löfgren's patients, in line with the previously observed exaggerated Th1 immune response, especially in non-Löfgren's patients [6].

T cell receptor AV2S3⁺ T cells have been associated with resolving disease. One possible explanation for this association could be that the AV2S3⁺ cells are regulatory cells, down-modulating the inflammation. Alternatively, they could be effector cells that are particularly efficient in eradicating an offending antigen. The observation that FoxP3 mRNA expression is reduced in TCR AV2S3⁺ T cells in BALF is in line with our previous finding that both CD25 and CD27 expressions were reduced on AV2S3⁺ compared with AV2S3^- BALF CD4⁺ T cells [35], while these cells expressed higher levels of effector–function-associated molecules such as CD28 on their surface [35]. Our mRNA data on FoxP3 expression in CD4⁺AV2S3⁺ T cells were also confirmed by flow cytometric evaluation, which showed a lower frequency of CD4⁺AV2S3⁺ T cells expressing FoxP3 compared with CD4⁺AV2S3^- T cells. These results are consistent with the suggestion that expanded AV2S3⁺ T cells in the lungs of Löfgren's patients do not act as regulatory but as effector T cells, which may be important for eradicating a postulated sarcoidosis antigen.

A deficiency in FoxP3, and an associated reduced capacity of T_reg cells, may contribute to a general susceptibility to develop sarcoidosis. Such a deficiency could either be a constitutive property of regulatory T cells in these individuals or may reflect an enhanced capacity to down-modulate FoxP3 expression in response to, for example, TNF stimulation.

In summary, we have demonstrated reduced FoxP3 mRNA levels, as well as reduced CCR2 and IL-10 mRNA levels, in BALF CD4⁺ T cells of sarcoidosis patients, which may indicate a significant defect in regulatory T cell functions in sarcoidosis. We also found a reduced frequency of FoxP3⁺ cells in BALF of patients compared with controls, and moreover that T_regs present in the lungs of patients had reduced levels of FoxP3 expression. Finally, our data suggest that the CD4⁺ TCR AV2S3⁺ T cells that accumulate in the lungs of patients with Löfgren's syndrome are not regulatory, but effector T cells.

Up-regulation of FoxP3 expression to reconstitute regulatory T cell activities may be considered as a future subject for research towards establishing new treatment strategies for sarcoidosis cases with chronic inflammation.