Maria Wikén, Department of Medicine Solna, Respiratory Medicine Unit, Lung Research Laboratory L4:01, Karolinska University Hospital Solna, 171 76 Stockholm, Sweden. (fax: +46-8-517 75 451; e-mail: Maria.Wiken@ki.se).
Abstract. Wikén M, Grunewald J, Eklund A, Wahlström J (Respiratory Medicine Unit, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden). Multiparameter phenotyping of T-cell subsets in distinct subgroups of patients with pulmonary sarcoidosis. J Intern Med 2012; 271: 90–103.
Objectives. Sarcoidosis is an inflammatory disorder in which elevated numbers of activated T cells are found in the lung. HLA-DRB1*0301pos (DR3pos) patients are characterized by good prognosis and an accumulation of lung CD4pos T cells expressing the T-cell receptor (TCR) gene segment AV2S3. Our aim was to phenotype lung and blood T-cell subsets in distinct patient groups to better understand the function of these subsets.
Design. Bronchoalveolar lavage (BAL) fluid and whole blood were obtained from a total of 22 patients with sarcoidosis, of whom 11 were DR3pos. Using eight-colour flow cytometry, phenotyping of T cells was performed with regard to CD3, CD4, CD8, CD25, CD27, CD45RO, CD57, CD69, CD103, FOXP3 and TCR AV2S3.
Results. DR3pos patients had fewer FOXP3pos (regulatory) CD45ROpos (memory) BAL T cells than DR3neg patients. Fewer AV2S3pos T cells were FOXP3pos, compared with AV2S3neg cells, thus indicating an effector function and not a regulatory role for this subset. Fewer lung and blood AV2S3pos T cells were CD25pos CD27pos, and more were CD25neg CD27neg and CD69pos, compared with AV2S3neg T cells, indicating a higher degree of differentiation and activation in both compartments.
Conclusion. Our main findings were a lower proportion of regulatory T cells in DR3pos patients, together with the accumulation of AV2S3pos T cells with a highly activated effector phenotype in the lungs of these patients. This may provide for efficient elimination of a harmful antigen in DR3pos patients and could thus help to explain the spontaneous recovery typically seen in these patients.
Sarcoidosis is an inflammatory granulomatous disorder of unknown aetiology. In patients with active disease, the number of lymphocytes, in particular CD4pos T cells, in the lower airways is increased, which results in an enhanced bronchoalveolar lavage (BAL) fluid CD4/CD8 ratio . Sarcoidosis is a heterogeneous disease, with a noteworthy variation in clinical manifestations and outcome. Familial clustering suggests a genetic predisposition , and several studies have demonstrated the importance of the human leukocyte antigen (HLA) class II complex . However, HLA associations with disease risk or disease phenotype vary with ethnicity. We have previously shown that a subgroup of Swedish patients with sarcoidosis, i.e. HLA-DRB1*0301pos (DR3pos) patients who commonly present with acute disease onset, good prognosis and Löfgren′s syndrome (characterized by erythema nodosum and/or ankle arthritis, fever and bilateral hilar lymphadenopathy with or without lung parenchymal infiltration ), have a reduced expression of T helper (Th)1 cytokines in BAL fluid compared with DR3neg patients . In addition, DR3pos patients usually present with an oligoclonal accumulation of lung CD4pos T cells expressing the T-cell receptor (TCR) variable (V) gene segment AV2S3 (TCR AV2S3pos T cells) . The magnitude of the TCR AV2S3pos T-cell expansion correlates with better prognosis, indicating a protective role of these cells . By contrast, DR3neg patients are characterized by a prolonged disease course and are at risk of developing fibrosis .
CD27, a costimulatory molecule, is expressed on naïve T cells, as well as on subsets of memory T cells . It functions to enhance TCR-induced T-cell expansion of peripheral CD4pos and CD8pos T cells, as well as to promote survival of activated T cells . Antigenic TCR stimulation results in an upregulation of CD27, followed by an irreversible loss with persistent stimulation [11–13]. Hendriks et al. showed that CD27-deficient mice exhibited impaired generation of influenza-specific T cells in lymphoid organs, as well as a reduced number of CD4pos and CD8pos T cells in the lung [10, 14], indicating that CD27 is essential for accumulation of antigen-specific T cells in the target organ. It has further been shown that memory T cells that have lost their CD27 expression produce more cytokines in response to antigenic stimulation as compared with CD27pos memory T cells . In this study, we therefore subdivided T cells based on their expression of CD27 and the memory molecule CD45RO, thereby being able to focus on the most potent effector cells (using CD27neg CD45ROpos as a marker of fully differentiated effector memory cells). Another way of studying differentiation and activation is to use CD27 in combination with the well-known activation marker CD25 (interleukin-2 receptor alpha chain) . In addition, CD4pos T cells expressing the CD57 antigen have been found to be differentiated memory T cells [16, 17].
Regulatory T cells (Tregs) function to balance inflammation and antigen-specific immune responses. To date, the most reliable Treg marker is the FOXP3 transcription factor [18, 19], which is essential for Treg development and suppressive function . Mice or humans with FOXP3 deficiency develop severe multi-organ autoimmune disease [21, 22]. We previously showed reduced mRNA and protein expression of FOXP3 in CD4pos T cells of patients with sarcoidosis compared with healthy controls , whereas others have demonstrated defective Treg function in sarcoidosis, albeit with increased numbers of Treg cells . As such discrepancies could partly be owing to different patient characteristics, in this study we compared subgroups of patients with different prognoses.
Sarcoidosis patients with a genetic variation in the gene encoding the integrin CD103, ITGAE, have an increased number of BAL CD4pos T cells expressing CD103 and exhibit fibrocytic inflammation . This suggests that CD103 is a potential diagnostic marker for disease .
The aim of this study was to perform a phenotypic characterization of CD4pos and CD8pos T cells, using T-cell markers for memory and effector function, regulatory capacity, differentiation and activation, and to compare the proportions of different T-cell subsets from the diseased organ (i.e. the lung) with the proportions in blood. We also investigated whether distinct patient subgroups displayed differences in the proportion of T-cell subsets. In addition, the characteristics of the TCR AV2S3pos T cells accumulating in the lungs of DR3pos patients were assessed.
Materials and methods
Patients with sarcoidosis referred to the Respiratory Medicine Unit (Karolinska University Hospital, Stockholm, Sweden) for diagnostic investigation were included in this study. All patients were diagnosed with pulmonary sarcoidosis as determined by symptoms, chest radiography and pulmonary function tests, and the diagnosis was established using the criteria of the World Association of Sarcoidosis and other Granulomatous Disorders (WASOG) . Written informed consent was obtained from all subjects, and the regional ethical review board approved the study.
A total of 22 patients with sarcoidosis [median age 43 years (mininum-maximum: 23–72); 14 men and eight women] participated in the study. Patients’ clinical characteristics and BAL fluid differential cell counts are shown in Table 1. Eleven patients were HLA-DRB1*0301pos (DR3pos) [median age 35 years (minimum–maximum: 23–50); eight men and three women], all of whom had >10.5% BAL fluid CD4pos TCR AV2S3 [i.e. an AV2S3pos T-cell expansion defined as three times median AV2S3% of CD4pos T cells in peripheral blood of healthy subjects (3 × 3.5%) ], and all except two had Löfgren′s syndrome. Eleven patients were HLA-DRB1*0301neg (DR3neg) [median age 50 years (minimum–maximum: 30–72); six men and five women], of whom only two had Löfgren′s syndrome. Five of the 22 patients were smokers, four were ex-smokers and 13 were nonsmokers. The proportion of current, ex-smokers and nonsmokers was similar in DR3pos and DR3neg patients. All patients were untreated, and all fasted since midnight before BAL was performed in the morning.
Table 1. Characterization of patients with sarcoidosis
HLA-DRB1*0301pos (n = 11)
HLA-DRB1*0301neg (n = 11)
Pulmonary function tests (values show % of predicted): VC, vital capacity (% of reference value); FEV1, forced expiratory volume in 1 s; DLCO, diffusing capacity of the lung for carbon monoxide.
BAL, bronchoalveolar lavage.
Data are shown as median (25th–75th percentile), except for age [which is shown as median (minimum–maximum)]. Mann–Whitney U-test, *P <0.05, ***P <0.001.
X-ray stage (0/I/II/III/IV)
Recovery (% of instilled volume)
Total cell concentration (*106 L−1)
Total cell number (*106)
BAL differential cell counts
Macrophages (*106 L−1)
Lymphocytes (*106 L−1)
Neutrophils (*106 L−1)
Eosinophils (*106 L−1)
BAL CD4/CD8 ratio
BAL AV2S3 (% out of CD4)
Blood AV2S3 (% out of CD4)
Pulmonary function tests
98 (85–111) (n = 10)
91 (65–95) (n = 8)
96 (83–109) (n = 10)
87 (59–108) (n = 9)
88 (85–102) (n = 7)
93 (69–107) (n = 7)
BAL procedure and handling of BAL fluid and whole blood cells
Bronchoalveolar lavage was performed as previously described . Briefly, a flexible fibre optic bronchoscope (Olympus Optical Co., Tokyo, Japan) was passed through the nose, under local anaesthesia, and positioned into a bronchus in the middle lobe. Sterile phosphate-buffered saline (PBS) solution (37 °C) was instilled in five aliquots of 50 mL, after which the fluid was aspirated and collected in a plastic bottle kept on ice. The BAL fluid was strained through a Dacron net (Millipore, Cork, Ireland) and centrifuged at 400 g for 10 min at 4 °C. The pellet was resuspended in PBS, the cells were counted in a Bürker chamber and the viability was determined by trypan blue exclusion.
Bronchoalveolar lavage fluid CD4/CD8 T-cell ratio and TCR AV2S3 expression in BAL fluid cells were determined by flow cytometric analysis (FACSCanto; BD Biosciences, Mountain View, CA, USA) using monoclonal antibodies against CD3 (APC-H7), CD4 (PE-Cy7) and CD8 (PerCP-Cy5.5) (Dako Cytomation Norden AB, Solna, Sweden) and a human TCR AV2S3-specific monoclonal antibody [fluorescein isothiocyanate (FITC), clone F1; Serotec, Oxford, UK] . Whole blood was collected in heparinized tubes.
Immunofluorescent surface staining and flow cytometry
All reagents and antibodies were purchased from BD Biosciences, unless otherwise stated. Expression of cell surface and intracellular markers was determined by flow cytometry, and data were analysed using FACSDiva software (BD Biosciences). On average, 80 000 BAL CD3pos T cells and 40 000 blood CD3pos T cells were evaluated by flow cytometric analysis.
Surface staining Aliquots of fresh BAL cells [106 cells resuspended in 100 μL cell wash (BD Biosciences)] or whole blood (100 μL) were stained for the surface molecules CD3 (Pacific Blue), CD4 (APC-H7), CD8 (AmCyan) and AV2S3 (FITC, only DR3pos patients), in combination with the following antibody panels: (i) CD25 (PE-Cy5), CD27 (APC), CD45RO (Pe-Cy7) and FOXP3 (PE); (ii) CD25 (PE-Cy5), CD27 (APC) and CD69 (Pe-Cy7); (iii) CD57 (Alexa488); (iv) CD103 (FITC), CD25 (PE-Cy5), CD27 (APC) and CD69 (Pe-Cy7); or matched isotype controls, for 20 min in the dark [BAL cells at 4 °C and blood cells at room temperature (RT)]. Red blood cells were thereafter lysed (FACS lysing solution) for 8 min at RT. Remaining white blood cells were centrifuged for 5 min at 400 g at RT and washed twice with RT cell wash, after which 500 μL cell wash was added to tubes ii, iii and iv. After antibody staining, BAL cells were washed twice in cold cell wash, centrifuged for 5 min at 400 g at 4 °C and resuspended in 500 μL cell wash.
Intracellular staining of FOXP3 Following surface antibody staining as described earlier, BAL and blood cells of tube i were fixed and permeabilized for 45 min at 4 °C in the dark. The cells were washed twice with 2 mL 1 × permeabilization buffer, centrifuged for 5 min at 400 g at 4 °C and incubated with 2% normal rat serum for 15 min at 4 °C. Next, the cells were stained intracellularly with anti-FOXP3 (clone PCH101-PE) or with anti-IgG2a-PE (both purchased from eBioscience, San Diego, CA, USA) for 30 min at 4 °C, then washed with permeabilization buffer and resuspended in 500 μL cell wash.
Statistical analyses were performed with GraphPad PRISM 4.03 (GraphPad Software Inc., San Diego, CA, USA). The Mann–Whitney U-test was used for comparison between patient subgroups, and the Wilcoxon signed rank test was used for comparisons between BAL fluid and whole blood or between various cell types. P-values <0.05 were considered significant.
Analysis of CD27pos and CD27neg memory T-cell subsets in BAL fluid and blood
Ongoing antigenic stimulation results in a sequential pattern of T-cell differentiation (a schematic model of differentiation is shown in Fig. 1a). In BAL CD4pos T cells, the most common subset was CD27neg memory T cells (CD45ROpos), followed by CD27pos memory T cells, and only a minority was CD27pos naïve T cells (Fig. 1b). In blood, however, the CD27pos naïve T cells and CD27pos memory T cells were the dominant subsets, and only a minority was CD27neg memory T cells (Fig. 1b). This indicates that BAL CD4pos T cells are more differentiated than blood CD4pos T cells. Among BAL CD8pos T cells, the most common subset was CD27pos memory T cells, followed by CD27neg memory T cells, and only a small percentage was CD27pos naïve T cells, whereas blood CD8pos T cells mirrored the pattern of blood CD4pos T cells (Fig. 1c). Thus, BAL CD8pos T cells also showed a higher degree of differentiation and activation than blood CD8pos T cells. There were no differences in the proportion of T-cell subsets in either BAL fluid or blood between patient subgroups.
Fewer FOXP3pos regulatory T cells in HLA-DR3pos patients
The expression of the regulatory T-cell marker FOXP3 was measured in CD27pos and CD27neg memory T cells. Although a high level of expression of CD25 has been shown previously to be a marker for regulatory T cells, it was clear that many FOXP3-expressing cells, especially in BAL fluid, were negative for CD25 (Fig. 2a). Among CD4pos T cells, both in BAL fluid and blood, FOXP3pos cells were primarily confined to the CD27pos memory T-cell subset, but were also detectable at low frequencies in the CD27neg memory subset (Fig. 2b). Some CD8pos T cells also expressed FOXP3, with BAL cells exhibiting the highest percentage (Fig. 2c), possibly indicating the presence of CD8pos regulatory T cells. When comparing patient subgroups, we found that BAL CD27pos and CD27neg memory CD4pos T cells, as well as BAL CD27pos memory CD8pos T cells of DR3pos patients, expressed FOXP3 in a smaller fraction of these subsets compared to DR3neg patients. In blood, there were no differences in FOXP3 expression between patient subgroups. Comparing lung and blood, we found that CD27pos memory cells in the BAL CD4pos T-cell subset expressed substantially more FOXP3 than CD27pos memory cells in blood CD4pos T cells, in both DR3pos and DR3neg patients (Fig. 2d).
Higher proportion of CD25neg CD27pos naïve T cells in BAL fluid of HLA-DR3pos patients
Activation and differentiation of T cells following antigen stimulation can also be characterized by the expression of CD25 and CD27 (a schematic model for activation and differentiation, based on expression of these markers, is shown in Fig. 3a). In BAL CD4pos T cells, the most common subset was CD25neg CD27neg T cells, followed by CD25neg CD27pos naïve T cells, and then CD25pos CD27pos T cells (Fig. 3b). By contrast, CD25neg CD27pos naïve T cells were most common in blood CD4pos T cells, followed by CD25neg CD27neg T cells. However, similar to BAL fluid, only a minority were CD25pos CD27pos T cells in blood (Fig. 3b). Among CD8pos T cells, BAL fluid and blood exhibited similar patterns, with the most common subset being CD25neg CD27pos naïve T cells, followed by CD25neg CD27neg T cells and then CD25pos CD27pos T cells (Fig. 3c). Some differences between patient subgroups were noted. For example, DR3pos patients had a higher proportion of BAL naive CD25neg CD27pos CD4pos T cells (Fig. 3b), and fewer CD25pos CD27pos CD4pos T cells, which demonstrates that a larger fraction of CD4pos T cells in DR3pos patients have not been activated. DR3pos patients also had a higher proportion of blood CD25pos CD27pos CD8pos T cells (Fig. 3c), as well as a lower percentage of blood CD25neg CD27neg CD4pos T cells (Fig. 3b).
In each T-cell subset, the expression of the activation marker CD69 was measured (illustrative histograms are shown in Fig. S1a). Our data showed that BAL T cells (Fig. S1b) were much more activated than blood T cells (Fig. S1c), but overall only minor differences between patient subgroups were detected with regard to CD69 expression.
With a few minor exceptions, the above differences between DR3pos and DR3neg patients, with regard to CD27/CD45RO and CD25/CD27 as well as FOXP3 expressions, were also seen when patients were subdivided into those with and without Löfgren′s syndrome (data not shown).
No differences with regard to CD57 expression between patient subgroups
The expression of the memory T-cell marker CD57 was analysed on CD4pos and CD8pos T cells in BAL fluid and blood of DR3pos and DR3neg patients (representative plots are shown in Fig. S2a). There were no differences between patient subgroups with regard to CD57 expression on CD4pos or CD8pos T cells (Fig. S2b). (The analysis of CD57 in blood CD8pos T cells was omitted because of difficulties in defining a distinctly CD57-positive cell population). BAL CD4pos T cells expressed CD57 to a much higher degree than blood CD4pos T cells.
Among the 11 DR3neg patients, we were able to obtain clinical follow-up data for seven individuals 2 years after BAL was performed. Disease resolution (judged by clinical investigation, such as X-ray and lung function tests) had occurred in three of these patients, while four were found to still have active disease. No obvious differences could be detected between these patient subgroups with regard to T-cell differentiation, T cell activation or regulatory phenotype.
Summary of differences between patient subgroups
The main findings were that in the BAL CD4pos T cell subset, DR3pos patients exhibited fewer FOXP3pos cells among CD27pos CD45ROpos T cells, more naïve CD25neg CD27pos T cells, and fewer activated CD25pos CD27pos T cells, compared to DR3neg patients.
TCR AV2S3pos CD4pos T cells of HLA-DR3pos patients are effector cells and not regulatory cells
All DR3pos patients included in our study had an accumulation of BAL TCR AV2S3pos CD4pos T cells (median proportion of TCR AV2S3pos cells was 27% of total CD4pos T cells in BAL fluid and 3.7% in blood). To characterize these cells in comparison to other lung CD4pos T cells, the phenotypic analysis was performed on AV2S3pos and AV2S3neg cells separately (Fig. 4a).
CD27neg memory T cells was the most common subset among both BAL AV2S3pos and AV2S3neg T cells, followed by CD27pos memory T cells, and only a small proportion were CD27pos naïve T cells (Fig. 4b). Compared with AV2S3neg T cells, AV2S3pos T cells in BAL fluid had a higher proportion of CD27neg memory T cells and a lower proportion of naïve T cells, but a similar level of CD27pos memory T cells.
A very prominent difference between AV2S3pos and AV2S3neg cells was found when analysing the regulatory T-cell marker FOXP3. The CD27pos memory T cells of the AV2S3pos T-cell subset were positive for FOXP3 to a much lower degree than the CD27pos memory T cells of the AV2S3neg cell subset (median 2.5% vs. 19%, P =0.001) (Fig. 4c). Although there were few CD27neg FOXP3pos cells, we attempted to further characterize regulatory cell subsets by analysing the proportion of AV2S3pos cells within the CD27pos and CD27neg FOXP3pos memory cells. We found an accumulation of AV2S3pos T cells within the CD27neg memory T-cell subset expressing FOXP3 (Fig. 4d), suggesting the existence of antigen-specific Tregs.
TCR AV2S3pos CD4pos T cells of HLA-DR3pos patients are more differentiated and activated than TCR AV2S3neg T cells
CD25neg CD27neg T cells were most common in both BAL AV2S3pos and AV2S3neg T cells, followed by CD25neg CD27pos T cells and then CD25pos CD27pos T cells (Fig. 4e). AV2S3pos BAL T cells comprised of more CD25neg CD27neg T cells, but fewer CD25pos CD27pos T cells, compared with AV2S3neg T cells, indicating a more differentiated phenotype. It is interesting that there were fewer CD25neg CD27pos T cells among blood AV2S3pos T cells and more CD25neg CD27neg T cells compared with AV2S3neg T cells (Fig. 4e), indicating that circulating AV2S3pos cells in peripheral blood also display signs of antigenic stimulation. In addition, the AV2S3pos BAL T cells were more activated, compared to AV2S3neg cells, as judged by their expression of CD69 (Fig. S3a). Similarly, a higher proportion of blood AV2S3pos cells were positive for CD69 (Fig. S3a). However, comparing AV2S3pos cells in BAL fluid and blood, these cells were markedly more differentiated and activated in the lung.
TCR AV2S3pos CD4pos BAL T cells of HLA-DR3pos patients express CD57 to a higher degree than TCR AV2S3neg T cells
Bronchoalveolar lavage AV2S3pos T cells expressed CD57 to a significantly higher degree than BAL AV2S3neg T cells, suggesting a more differentiated phenotype. In blood, there was no difference in CD57 expression in AV2S3pos and AV2S3neg T cells (Fig. S3b).
More CD25neg CD27neg terminally differentiated T cells among CD103pos BAL or blood T cells
As a detailed analysis of TCR AV2S3pos T cells only was of interest in DR3pos patients, we used an anti-CD103 antibody (CD103 is a marker of intra-epithelial T cells) to analyse intra-epithelial T cells in DR3neg patients (representative plots are shown in Fig. 5a). Our data showed that there was a higher expression of CD103 in BAL fluid compared to blood on both CD4pos and CD8pos T cells (Fig. 5b).
The expressions of activation marker CD25 and differentiation marker CD27 were analysed in CD103pos and CD103neg CD4pos T cells in BAL fluid and blood. Among CD4pos T cells in BAL fluid, there were fewer CD25negCD27pos naïve and CD25posCD27pos T cells, but more CD25neg CD27neg T cells (terminally differentiated cells), in the CD103pos T cells compared with the CD103neg T cells (Fig. 5c). The phenotypic difference between CD103pos and CD103neg cells in blood mirrored that of BAL fluid but was even more pronounced in blood. In addition, the subsets of CD103pos BAL T cells expressed CD69 to a higher degree than CD103neg T cells (Fig. S4) indicating a more activated state.
In this study, we characterized CD4pos and CD8pos T cells in BAL fluid and peripheral blood of two groups of patients with sarcoidosis, characterized by different clinical manifestations. We found that the most common types of BAL CD4pos T cells were CD27neg CD45ROpos (‘fully differentiated effector memory’) cells followed by CD27pos CD45ROpos memory T cells, whereas the opposite was seen in blood. Among BAL memory T cells, DR3pos patients, typically with Löfgren′s syndrome and good prognosis, had fewer FOXP3pos cells (i.e. regulatory T cells) compared with DR3neg patients, who are prone to develop chronic disease. TCR AV2S3pos T cells, which accumulate in large numbers in the lungs of DR3pos patients, were more differentiated and more activated than corresponding AV2S3neg T cells, both in BAL fluid and blood. Furthermore, in general, AV2S3pos T cells were found to be effector cells rather than regulatory cells.
We found that the majority of memory (i.e. CD45ROpos) CD4pos T cells in BAL fluid had lost the expression of CD27 in both DR3pos and DR3neg patients, whereas the majority of blood CD4pos memory T cells retained CD27 expression. This is in line with a previous study by Mack et al.  in patients with chronic beryllium disease (CBD), a granulomatous disorder with many similarities to sarcoidosis. Thus, T cells exhibit a more differentiated phenotype in the affected organ. In addition, there was a correlation between loss of CD27 expression on lung T cells and severity of lung inflammation in patients with CBD . Furthermore, the expression of the memory T-cell marker CD57, which is associated with potent effector function [32–34], was investigated. Our results demonstrated more CD57pos CD4pos T cells in the lungs, compared with blood. This is in line with what is seen in patients with CBD , once again indicating an accumulation of late memory T cells in the affected organ.
We next combined CD27 with the activation marker CD25. Among BAL CD4pos T cells, DR3pos patients had a higher proportion of CD25neg CD27pos naïve cells, but a lower percentage of CD25pos CD27pos cells, compared with DR3neg patients, suggesting that the T cells are less activated among DR3pos patients. This may be related to our previous finding, in a study to investigate total BAL cell mRNA expression and BAL fluid cytokine content, of a less pronounced Th1 response in the lungs of DR3pos patients . This is also in agreement with the findings of a recent study by Heron et al.,  even though they used different markers and categorized patients according to X-ray stage.
As CD4pos TCR AV2S3pos T cells accumulate in large numbers in the lungs of DR3pos patients, and an increased number of these cells are correlated with a better prognosis, we were interested to determine whether they are effector or regulatory T cells. We found that a much smaller percentage of the AV2S3pos T cells expressed FOXP3 than the AV2S3neg T cells, thus clearly demonstrating what had previously been shown at the protein level by Idali et al.  in a few individuals. That report from our laboratory also showed that the mRNA level of FOXP3 was significantly lower in sorted TCR AV2S3pos T cells as compared with sorted AV2S3neg T cells . Thus, studies at both the mRNA and protein levels show that AV2S3pos T cells are predominantly effector cells and not regulatory T cells.
In the present study, we further characterized the AV2S3 cells into different stages of differentiation and found that AV2S3pos T cells had a more differentiated phenotype, thus extending the previous findings from our laboratory . It is noteworthy that this applies not only to the lung AV2S3pos T cells, as the findings of the present study show that the more differentiated and activated phenotype is also a property of AV2S3pos T cells in peripheral blood. Together, the results of the present and previous studies indicate that AV2S3pos T cells are predominantly highly activated Th1 effector cells and not regulatory cells. With regard to total FOXP3pos regulatory T cells, we found that there are fewer in DR3pos patients.
CD103 may be a marker of prolonged fibrogenic inflammation in the lung, as Lohmeyer et al.  previously demonstrated that sarcoidosis patients with higher X-ray stage had a higher proportion of CD103pos CD4pos T cells in BAL fluid. This is also supported by the results of Heron et al. . We analysed the proportion of CD103pos T cells in DR3neg patients, although there were too few patients to classify them according to X-ray stage. As expected, there were more CD103pos cells in BAL fluid than in blood. Furthermore, using the markers CD25 and CD27, we found CD103pos T cells to be more activated and differentiated than CD103neg T cells. This is in line with the findings of Braun et al. , showing that CD103pos T cells are terminally differentiated effector cells involved in prolonged inflammation.
In conclusion, our data show that DR3pos patients with sarcoidosis have a lower proportion of BAL memory T cells expressing FOXP3 (i.e. regulatory T cells), compared with DR3neg patients. Furthermore, TCR AV2S3pos T cells of DR3pos patients were found to be effector cells, rather than regulatory cells, and were also more activated than TCR AV2S3neg T cells. These findings suggest that efficient T-cell help for antigen removal by macrophages may be a mechanism behind the spontaneous disease remission typically seen in DR3pos patients with sarcoidosis. However, further studies aimed at delineating the functional capacities of the various T-cell subsets are warranted.
Conflict of interest statement
The authors have no conflicts of interest to declare.
The authors thank Heléne Blomqvist, Margitha Dahl, Benita Dahlberg, Benita Engvall, Gunnel de Forest and Lotta Pousette for their excellent technical assistance.
This study was supported by the Swedish Heart-Lung Foundation, the King Oscar II Jubilee Foundation, the Swedish Research Council, the Mats Kleberg Foundation, the Söderberg Foundation, the Stockholm County Council and Karolinska Institutet.