INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sarcoidosis is a chronic, systemic inflammatory disorder of unknown etiology, characterized by the formation of noncaseating granulomata in affected organs, most commonly the lung (1). T cells and macrophages are key components of this process, and these cells can be retrieved for analysis from the deep airways by the technique of bronchoalveolar lavage (BAL). Studies in which this has been done have demonstrated increased numbers of activated CD4⁺ T cells in the lungs of sarcoidosis patients. Investigating the T-cell receptor (TCR) repertoire we have previously shown that Scandinavian sarcoidosis patients often have dramatic, lung-restricted accumulations of CD4⁺ T cells using the V2.3 gene segment (2). These so-called T-cell expansions correlate very strongly with the human leukocyte antigen (HLA) haplotype DR17 (3). Most probably, these T cells were selected by and proliferated in response to a postulated sarcoidosis antigen presented by the DR17 molecule. Of great interest is the finding that Scandinavian patients positive for DR17 have a more acute onset of disease and a much better long-term prognosis than do patients negative for DR17 (3).

T cells are the main regulatory cells of the immune system. This role stems from their capacity to produce cytokines that can be categorized as T helper type 1 (Th1)-derived cytokines, such as interleukin (IL)-2 and interferon (IFN)-, or the Th2-associated cytokines IL-4, IL-5, and IL-13 (4). Th1 cytokines are mainly involved in cellular immune responses and Th2 cytokines are largely involved in providing B-cell help, but both may also counterregulate the effects of the other subset. The balance between Th1 and Th2 cytokines can be of critical importance for the outcome of an immune response, as dramatically demonstrated in Mycobacterium leprae infection, in which a Th1 response is associated with the relatively benign tuberculoid form of disease whereas a Th2 response dominates in the fatal lepromatous form (5). The Th1/Th2 heterogeneity of cytokine production was first shown in CD4⁺ (T-helper cell) clones from mice and humans, but was later shown to apply to other lymphocytes such as CD8⁺ cells and cells (6). In vivo, however, there is a spectrum of cytokine-producing cells between the Th1 and Th2 poles.

In sarcoidosis, lung T cells have been shown to spontaneously release increased levels of IL-2 and IFN- (7), and studies of T-cell clones from BAL (10, 11) and lung parenchyma (12), and of bronchoalveolar lavage fluid (BALF) (13, 14), support the notion of sarcoidosis as a Th1-mediated disease. Also of interest is tumor necrosis factor (TNF)-, since its gene exists in allelic variants that have been correlated with variations in TNF- production (15) and with susceptibility to various diseases (16). In particular, the TNFA2 allele is in linkage disequilibrium with some HLA alleles, notably HLA-DR17 (HLA-DR3 according to the old nomenclature) (17), which we previously showed to predict a good prognosis in Scandinavian sarcoidosis patients (3).

With regard to BAL T cell clones, data on cytokine production are conflicting (11, 12). Since the cloning procedure may skew the cytokine profile, it is preferable to analyze the cells as soon as possible after BAL is performed. This can be achieved by cytokine flow cytometry, in which cells are permeabilized and stained for intracellular cytokines (18, 19). By combining intracellular staining with surface staining, this technique also allows detection of cytokine production in specific cell subsets. CD8⁺ T cells have often been overlooked in discussions of the pathogenesis of sarcoidosis because of their relatively smaller numbers in the lung as compared with CD4⁺ cells, but cytokine flow cytometry allows independent investigation of the cytokine phenotype of both subsets.

The objective of the present study was to quantitate the frequency of cells producing IL-2, IL-4, IFN-, and TNF- among peripheral blood lymphocytes (PBL) and in BALF T-cell subsets in sarcoidosis patients. T cells from the PBL population of healthy individuals served as controls. Specifically, we wanted to study CD4⁺ and CD8⁺ T cells separately. We hypothesized that differences exist in the cytokine profiles of HLA-DR17-positive and HLA-DR17-negative patients that may explain why these patient groups have a dramatically different prognosis in sarcoidosis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

A total of 18 patients with untreated sarcoidosis (median age: 39 yr; range: 27 to 59 yr), consisting of eight men and 10 women, participated in the study. Seventeen of the patients had pulmonary sarcoidosis and one had only skin manifestations. Ten had histologically proven disease and the remaining eight had typical clinical findings (symptoms such as dyspnea, fatigue, fever, coughing, and weight loss) and chest radiographic features of sarcoidosis. Specifically, five of these eight patients had classical Löfgren's syndrome, which is pathognomonic of sarcoidosis, and one further patient had uveitis, supporting the diagnosis. Five of the eight patients had a BALF CD4⁺/CD8⁺ cell ratio exceeding 4.0, also supporting the diagnosis (20). The patient with skin manifestations had a lung-restricted expansion of V2.3-positive T cells, which is common in Scandinavian sarcoidosis patients. In the two patients without a positive biopsy, Löfgren's syndrome or uveitis cultures for mycobacterial or fungal infection were negative. Nor were any precipitating antibodies, which might have suggested the diagnosis of extrinsic allergic alveolitis (EAA), detected, although EAA and pulmonary fungal infections are very rare diseases in the region from which the patients in the study were recruited. Two of the patients were smokers, two were ex-smokers, and 14 were never-smokers. Nine patients were HLA-DR17-positive. Disease activity was assessed on the basis of symptoms, chest radiography, and pulmonary function tests, using previously established criteria (21). Thirteen patients had active disease (eight of whom were DR17-positive and five of whom were DR17-negative). Thirteen healthy adults were included as controls. The study was approved by the local ethics committee.

BAL

BAL was performed as described (22). In brief, with patients under local anesthesia, a flexible fiberoptic bronchoscope (OBF Type 1TR; Olympus Optical Co., Tokyo, Japan) was wedged into middle-lobe bronchus and sterile phosphate-buffered saline (PBS) solution at 37° C was instilled in five aliquots of 50 ml each. After each instillation, the fluid was gently aspirated and collected in a siliconized plastic bottle that was kept on ice.

Preparation of Cells

The BALF was strained through a Dacron net (Millipore, Cork, Ireland) and centrifuged at 400 × g for 10 min at 4° C, and the pellet was resuspended in RPMI-1640 medium (Sigma-Aldrich, Irvine, UK). Cells were counted in a Bürker chamber, and total cell viability (median = 95%) was determined by Trypan blue exclusion. Smears for differential counts were prepared by cytocentrifugation (Cytospin 2; Shandon, Runcorn, UK) at 22 × g for 3 min, whereafter cells were stained with May-Grünwald-Giemsa. Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) gradient centrifugation was used to separate peripheral blood mononuclear cells, which were washed twice and diluted in RPMI-1640.

In Vitro Stimulation of PBL and BAL Cells

Cells were pelleted and resuspended in RPMI-1640 medium (Gibco BRL, Paisley, Scotland) supplemented with 100 IU/ml penicillin, 0.1 mg/ml streptomycin (Gibco BRL), and 5% fetal calf serum (JRH Biosciences/Sera Laboratories, Crawley Down, Sussex, UK) at 1 × 10⁶ cells/ml. Cells were stimulated with 1 ng/ml phorbol myristate acetate (PMA) (Sigma Chemical Co., St. Louis, MO) and 0.5 µM ionomycin (Sigma) in the presence of 10 µg/ml Brefeldin A (Sigma). Brefeldin A was used to block intracellular transport mechanisms, thereby producing an accumulation of cytokines in the Golgi apparatus. Cells were cultured in flat-bottom six-well plates (Costar, Corning, NY) for 5 h at 37° C in a humidified atmosphere of 5% CO₂ in air (B5060 incubator; Heraeus, Hannau, Germany). The stimulation time of 5 h was chosen in accordance with previous kinetic studies as optimal for detecting the chosen cytokines (19, 23). Cells were pelleted, resuspended in PBS, washed once more, and resuspended in PBS at 2 × 10⁶ cells/ml. After the addition of 1 volume of 4% formaldehyde (Fluka Chemika, Buchs, Switzerland) in PBS, cells were fixed for 20 min at room temperature (RT). Fixed cells were washed twice in ice-cold PBS and stored in PBS at 4° C in the dark overnight.

Permeabilization and Staining

Cells were pelleted and resuspended in saponin buffer (Earle's balanced salt solution; EBSS) (Gibco BRL) containing 0.1% saponin and 0.01M 4-(2-hydroxyethyl)-1-piperazine-N'-2-ethanesulfonic acid (HEPES) buffer, and the pH was adjusted to 7.40. This produced a reversible permeabilization (24). Aliquots of 50 µl, containing a maximum of 1 × 10⁶ cells, were portioned into a 96-well plate and incubated with 100 µl of unlabeled mouse monoclonal antihuman cytokine antibody (10 µg/ml) in saponin buffer for 20 min at RT. The following monoclonal antibodies (mAbs) (all of isotype IgG₁) were used: anti-IL-2 (BG5; Serotec, Oxford, UK), anti-IL-4 (8D4-8; Pharmingen, La Jolla, CA), anti-IFN- cocktail (7B6 and 1-D1K; Mabtech, Stockholm, Sweden), and anti-TNF- cocktail (Mab1 and Mab11; Pharmingen), with mouse IgG₁ (Dako, Glostrup, Denmark) as a negative control. Cells were washed twice in saponin buffer and stained at RT for 20 min with 100 µl of fluorescein isothiocyanate (FITC)-conjugated goat antimouse IgG₁ antibodies (Caltag, San Fransisco, CA) diluted 1:300 in saponin buffer. For subsequent surface staining, intracellularly stained cells were pelleted, resuspended in EBSS/HEPES without saponin, washed again, and incubated for 10 min in NMS diluted at 1:500 in PBS to block remaining binding sites of the goat antimouse antibodies. Thereafter, cells were incubated with either r-phycoerythrin-(RPE)-labeled anti-CD4 antibodies or RPE-labeled anti-CD8 antibodies (DAKO) for 20 min, washed twice, and resuspended in EBSS/HEPES. For flow-cytometric analysis, data for 10,000 stained cells were collected and analyzed with a fluorescence-activated cell sorter (FACS) (Becton Dickinson, Mountain View, CA). Lymphocytes were gated by forward and side scatter, and the percentages of positively labeled cells in the CD4⁺ and CD8⁺ subsets were scored. Isotype-matched negative control antibodies always stained less than 1% of CD4⁺ and CD8⁺ lymphocytes. Shedding of CD4 did not affect the analysis, since a distinct CD4⁺ cell population was readily identifiable and of the same magnitude as before stimulation and fixation. For IL-2, an additional analysis was performed, since the positioning of the cutoff marker for background fluorescence tended to be more difficult for this cytokine, owing to the low signal intensity. Thus, the quantitative levels of IL-2, expressed as mean fluorescence intensity (MFI), were determined after subtraction of the background fluorescence level calculated from samples labeled with the control antibody.

HLA Typing

HLA class II (HLA-DR) typing was done on DNA through use of the polymerase chain reaction (PCR) and amplification with sequence-specific primers.

Statistics

Results are presented as median values. Significance levels were calculated according to the nonparametric Mann-Whitney U test. A level of p < 0.05 was regarded as significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Total and differential cell counts and CD4⁺/CD8⁺ cell ratios in BALF are given in Table 1. A representative example of intracellular cytokine flow-cytometric analysis of IL-2, IL-4, IFN-, and TNF- expression in CD4⁺ PBL and BALF cells from sarcoidosis patients, and in CD4⁺ PBL cells from healthy controls, is shown in Figure 1.

View larger version (43K): [in this window] [in a new window]   Figure 1.   Flow-cytometric dot-plots of intracellular cytokines in CD4+ PBL cells from healthy controls (HC) and in CD4+ PBL and BAL lymphocytes from sarcoidosis patients (PAT). Staining for the respective cytokines (or negative control) is shown on the x axis; staining for CD4+ cells is shown on the y axis. Dot-plots were generated after forward- and side-scatter gating on lymphocytes. The figure in each upper right quadrant denotes the percentage of CD4+ lymphocytes positive for the respective cytokine.

Sarcoidosis Patients

As seen in Figure 1, the greatest differences were observed for IFN- and TNF-, which were produced by substantially greater frequencies of BALF cells. When all subjects' PBL and BALF CD4⁺ and CD8⁺ subsets were compared, there were significantly more cells positive for IFN- among BALF CD4⁺ cells (median: 69.1%) than among PBL CD4⁺ cells (19.0%), and the same was true for BALF CD8⁺ cells (86.8%) as compared with PBL CD8⁺ cells (65.8%) (Table 2, Figure 2). Similarly, there were significantly more cells producing TNF- among BALF CD4⁺ cells (63.9%) than among PBL CD4⁺ cells (32.4%), and among BALF CD8⁺ cells (48.6%) than among PBL CD8⁺ cells (23.0%). In contrast, there were significantly fewer cells positive for IL-4 among BALF CD4⁺ cells (2.0%) than among PBL CD4⁺ cells (4.9%), and among BALF CD8⁺ cells (1.4%) than among PBL CD8⁺ cells (4.4%). For IL-2, there were significantly more positive cells among BALF CD8⁺ cells (15.0%) than among PBL CD8⁺ cells (5.8%). The results for IL-2 should be interpreted with caution, however, since the low staining intensity of this cytokine made the results very sensitive to the exact positioning of the cutoff marker for the negative control. Therefore, an analysis of IL-2 MFIs was also done. This analysis did not reveal any significant differences between the different cell subsets or patient groups with regard to IL-2 MFI values (data not shown).

View larger version (28K): [in this window] [in a new window]   Figure 2.   Percentages of cytokine- positive cells in CD4+ and CD8+ T-lymphocyte subsets in PBL and BALF samples from sarcoidosis patients, as determined by flow cytometry. Note the different scale for IL-4. Values of p are indicated for comparisons of PBL and BAL. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Regarding differences between CD4⁺ and CD8⁺ cell subsets, it was noted that IFN- was produced by significantly more CD8⁺ cells, both in PBL and in BALF. On the other hand, IL-4 had a tendency to be produced by relatively more CD4⁺ cells, in both PBL and BALF. Corresponding differences were also observed in healthy controls.

Investigating differences between HLA-DR17-positive and HLA DR17-negative patients, we detected significantly (p < 0.05) fewer PBL CD4⁺ cells positive for IFN- in DR17-positive patients (Table 2). In DR17-positive patients, there were also tendencies toward a lower expression in BALF CD4⁺ cells of IL-2 and IFN-, and in BALF CD4⁺ and BALF CD8⁺ cells of TNF-, than in DR17-negative patients. Taken together, these results show a tendency toward a less pronounced Th1 response in DR17-positive patients.

Healthy Controls versus Sarcoidosis Patients

Comparing PBL of sarcoidosis patients with those of healthy controls (Table 2), we found significantly more cells positive for IFN- in patients than in controls, both in the CD4⁺ cell subset (19.0% versus 5.8%, respectively) and in the CD8⁺cell subset (65.8% versus 32.1%, respectively). IL-4 was also produced by more cells in patients than in controls, both among CD4⁺ (4.9% versus 2.0%, respectively) and among CD8⁺ cell (4.4% versus 0.9%, respectively) subsets. Likewise, there were more IL-2-positive cells in the CD4⁺ subset of patients' than of controls' PBL (19.2% versus 6.9%, respectively). With regard to TNF-, there were no significant differences between patients and controls.

Since both IFN- and IL-4 were produced by greater proportions of cells of sarcoidosis patients than those of controls, we calculated IFN-/IL-4 ratios for different cell subsets. There were no significant differences in the IFN-/IL-4 ratio when either patient PBL CD4⁺ cells (IFN-/IL-4 ratio = 4.0) were compared with healthy control CD4⁺ cells (IFN-/IL-4 ratio = 3.8) or when patient PBL CD8⁺ cells (IFN-/IL-4 ratio = 16.0) were compared with control CD8⁺ cells (IFN-/IL-4 ratio = 28.6). In contrast, in comparing lung and blood samples of patients, we found the IFN-/IL-4 ratio to be much higher in BALF CD4⁺ cells (IFN-/IL-4 = 43.0) than in PBL CD4⁺ cells (IFN-/IL-4 ratio = 4.0, p < 0.0001), and in BALF CD8⁺ cells (IFN-/IL-4 ratio = 48.6) than in PBL CD8⁺ cells (IFN-/IL-4 ratio = 16.0, p < 0.01).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The cytokine profile of T-cell subsets in PBL and BALF from sarcoidosis patients was determined flow cytometrically. We studied CD4⁺ and CD8⁺ cells separately, through the combination of intracellular staining for selected cytokines (IL-2, IL-4, IFN-, and TNF-) and extracellular staining for surface markers. To our knowledge, this was the first time that this experimental approach was used to study lung and blood T-cell subsets in sarcoidosis. This method was chosen so as to obtain a better reflection of in vivo cytokine patterns than can be achieved with cloning techniques. A predominantly Th1 profile was detected for BALF cells as compared with PBL, since in both CD4⁺ and CD8⁺ cell subsets there were significantly greater frequencies of T cells producing IFN- and TNF-, and significantly fewer cells positive for IL-4. The highest IFN-/IL-4 ratio was detected in BALF CD8⁺ cells. HLA-DR17-positive patients, previously shown to have a very good prognosis (3), had significantly fewer IFN--positive cells in the PBL CD4⁺ subset, and a tendency toward a less pronounced Th1 response in BALF. As compared with healthy controls, sarcoidosis patients had more peripheral blood T cells producing IFN-, IL-4, and IL-2.

Detection of intracellular cytokines by flow cytometry has several advantages over other methods for evaluating inflammatory cell function, such as determination of cytokine secretion in bulk culture or studies of T-cell clones. It provides a direct means of determining cell function at the single-cell level, and allows specific cell subsets, identifiable by their expression of certain cell-surface markers, to be studied in complex inflammatory cell mixtures (19). Cytokine flow cytometry has previously been used successfully for the study of BALF cells (23). However, it is not a direct measure of the in vivo level of cell activation, since the cells to be studied have to be stimulated in vitro to achieve detectable levels of cytokines. A study including six sarcoidosis patients found that without such stimulation, only BALF cells positive for IFN- were detectable (25). PMA/ionomycin has been considered the most reliable stimulus in most studies, and there is evidence that the cytokine patterns observed after PMA/ionomycin stimulation correspond to the in vivo cellular potential for cytokine production. For example, although PMA/ionomycin is a powerful stimulus, it does not indiscriminately activate all cytokine genes (26). Furthermore, T-cell clones have been shown to display the same cytokine profile whether stimulated with antigen or PMA/ionomycin (27).

By studying CD4⁺ and CD8⁺ cells separately, we detected a predominantly Th1 profile in BALF cells belonging to both of these subsets. The highest percentage of IFN--positive cells, and the lowest frequency of IL-4⁺ cells, were found among BALF CD8⁺ cells, highlighting the potentially important contribution of CD8⁺ T cells to the Th1-type inflammatory process, although CD4⁺ cells are more numerous. The prominent skewing of BALF CD8⁺ cells toward a Th1 type is consistent with our recent demonstration of an increased expression of CD26, a proposed Th1 marker, not only on BALF CD4⁺ but also on BALF CD8⁺ cells in sarcoidosis (28). The study in which this was demonstrated also showed high expression of several activation markers in both the BALF CD4⁺ and CD8⁺ T-cell subsets, in some instances particularly so in the CD8⁺ subset. In addition, the pattern of IL-12 receptor expression by CD4⁺ and CD8⁺ BALF T cells in sarcoidosis implies that both subsets are committed to Th1-type cytokine production (29). The relative functional roles of CD4⁺ and CD8⁺ T cells in our sarcoid inflammation remains to be investigated, but was outside the scope of our study. However, the potential for CD8⁺ cells to play an important role in such inflammation can be inferred from the demonstration that in some patients, sarcoid alveolitis is dominated by CD8⁺ T cells (30). Previous investigations, using other techniques, also support the concept of sarcoidosis as a Th1-mediated disease, reporting increased spontaneous release of IL-2 and IFN- from BALF T cells (7- 9), increased IFN- production by T-cell clones from BALF (10, 11) and the lung parenchyma (12) of sarcoidosis patients, and increased levels of Th1 cytokines in BALF (13, 14). The detection of Th1 cytokine messenger RNA dominance and high levels of Th1 cytokine proteins in sarcoid lymph nodes was also interpreted as evidence for a Th1-mediated process at the site of granuloma formation (31, 32). Besides the increased numbers of IFN--positive cells in BALF in sarcoidosis, a Th1 shift in BALF cells is also clear from our finding of reduced frequencies of IL-4-producing cells in the pulmonary compartment. Previous studies, of IL-4 levels in BALF (13, 14) or in supernatants of BAL-derived T-cell clones from sarcoidosis patients (11), generally failed to detect any differences from the BALF of healthy controls, or from PBL-derived clones, respectively, although in one investigation, a reduced level of IL-4 in BALF was reported (33). With regard to IL-2, our results must be interpreted with caution for reasons outlined in the RESULTS section, but our data agree with a previous study of T-cell clones showing that IFN- production was much more common than IL-2 production in BALF T-cell clones (11). When interpreting cytokine data, it should be noted that PBL in healthy individuals contain several times more IFN--producing than IL-4-producing cells, as shown by our data and previously reported by others (34). With regard to the situation in the healthy lung, cytokine flow cytometry is difficult to perform because of the low numbers of lymphocytes retrievable in the absence of alveolitis, although analysis of cloned T cells indicates an increased proportion of IFN--positive cells in BALF as compared with PBL (35).

Interestingly, we found that the frequencies of CD4⁺ and CD8⁺ T cells positive for IFN-, in particular, but also for IL-4, were higher in the peripheral blood of patients than that of controls. This was also true for IL-2 in the CD4⁺ subset. The higher numbers of cytokine-producing cells detected in the blood of sarcoidosis patients may reflect the systemic nature of sarcoidosis. It also agrees with our previous demonstration (28) of higher numbers of PBL T cells positive for various activation markers in sarcoidosis patients than have previously been reported in the blood of healthy individuals. However, the PBL IFN-/IL-4 ratio (reflecting the Th1/Th2 balance) did not differ significantly between sarcoidosis patients and controls. In contrast, there was a very clear Th1 bias both of CD4⁺ and of CD8⁺ cells in the lung as compared with the blood of sarcoidosis patients (i.e., a bias localized to the primary site of disease).

Despite the general agreement on a predominantly Th1-type infiltrate in the lungs of sarcoidosis patients, some controversy has surrounded data relating to BALF cells, which have shown shifts toward both the Th1 and Th2 ends of the cytokine spectrum (11, 12). Our study, however, clearly shows that the overwhelming majority of BALF T cells are of the Th1 type. The contrasting results of other studies, using cloning techniques, may be due to skewing of the cytokine profile during cloning processes. In addition, the cloning efficiency for BALF cells is lower than that for PBL (11). Our data also agree with findings of increased levels of the Th1-inducing cytokine IL-12 in BALF cells in sarcoidosis (14).

Other granulomatous diseases, such as berylliosis, tuberculosis, and tuberculoid leprosy are also characterized by a Th1 response. In these disorders as well as in sarcoidosis, IFN- may contribute to granuloma formation through its effect on macrophages (36).

The Th1/Th2 balance in the lung could potentially determine the prognosis in sarcoidosis. Our hypothesis that HLA-DR17-positive and -DR17-negative patients, who were previously shown to differ sharply with respect to prognosis (3), would also differ in their cytokine profiles could not be substantiated by the data of the present study, since there were no significant differences in frequencies of cytokine-positive BALF cells between these two patient groups. However, considering the major role of the Th1 cytokine IFN- in driving granulomatous inflammation, our finding of a tendency toward a less pronounced Th1 response in DR17-positive patients may help explain the good prognosis in this group. Our data in this respect agree with those of a previous study (33) demonstrating that patients with spontaneous disease resolution had higher levels of IgE and IgG₄ in their BALF, indicating a local, relative Th2 response. Since our most striking finding in DR17- positive patients was the very large accumulation in BALF of V2.3-positive CD4⁺ T cells (2), it is of particular interest to characterize these specific cells with respect to their cytokine profile and phenotype, something that we are presently pursuing.

TNF- may be classified as a Th1-type cytokine on the basis of its macrophage-stimulating capacity. In sarcoidosis, an increased TNF- release from alveolar macrophages (AM) has been associated with an increased risk of disease progression (37). TNF- is of particular interest because of the polymorphism in the promoter region of its gene (15). An increased prevalence of the TNFA2 allele in the clinically benign Löfgren's syndrome has been reported (38), and this allele is in linkage disequilibrium with several HLA genes, including HLA-DR17, which is associated with a good prognosis (3, 17). It could be speculated that this linkage might contribute to the benevolent prognosis for DR17- positive patients if the TNFA2 allele were associated with altered TNF- production by T cells and/or macrophages. However, we did not detect any significant differences in this limited number of DR17-positive and -negative patients in T-cell TNF- production. Therefore, our data do not support the hypothesis that HLA-DR17 is associated with altered TNF- production (via linkage disequilibrium with the TNFA2 allele). It appears more likely that the inherent properties of the DR17 molecule as a presenter of antigenic peptides may explain the spontaneous disease resolution in DR17-positive patients.

Intracellular cytokine staining is a valuable tool for characterizing immune responses, and has previously been used by others to analyze cytokine profiles of T cells from healthy individuals, to take but one example. The frequencies of cytokine-producing cells that we found in healthy controls were in the same range as those reported by other groups using similar protocols (39), thus validating the methodology employed. In addition, other techniques may help elucidate possible differences in cytokine production among patient subgroups. It is also of interest to study a broader range of cytokines; the immunosuppressive transforming growth factor-, for example, has been shown to be released preferentially from BALF cells of patients with active sarcoidosis who show spontaneous remission (42). As previously mentioned, studies of more specific T-cell subsets are of interest, in particular the V2.3-positive cells found to accumulate in the lungs of DR17-positive patients.

In conclusion, by using intracellular cytokine flow cytometry, we demonstrated significantly more cells producing IFN- and TNF- in CD4⁺ as well as in CD8⁺ BALF T cell subsets than in the same subsets of PBL from sarcoidosis patients, and significantly fewer cells positive for IL-4. Thus, both major T-cell subsets in BALF display a clear Th1 cytokine profile, suggesting that not only CD4⁺ cells but also the previously somewhat overlooked CD8⁺ T-cell subset may have important roles in the inflammatory process in the lungs of sarcoidosis patients. The increased numbers of cytokine-producing cells in the peripheral blood of sarcoidosis patients as compared with healthy controls may mirror the systemic nature of the disease. Studies of more specific T-cell subsets should help to further elucidate pathogenic mechanisms in sarcoidosis.