Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Chronic beryllium disease (CBD) is a granulomatous disorder that develops in 1 to 5% of individuals exposed to beryllium in the aerospace, automotive, ceramics, electronics, and defense industries (1). The lung is the predominant organ involved, although the lymphatic system, skin, liver, and spleen may also be affected by the granulomatous inflammation (4, 5). The diagnosis of CBD requires an exposure history, the presence of noncaseating granulomas and mononuclear cell inflammation in a lung biopsy specimen, and a positive in vitro proliferative response of peripheral blood or bronchoalveolar lavage (BAL) T cells to beryllium salts (6). Evidence suggests that CD4⁺ T cells are important in initiating and perpetuating the immune response to beryllium (1, 7), and that the in vitro T-cell response to beryllium sulfate (BeSO₄) is class II major histocompatibility complex (MHC) antigen restricted (1, 2). The nature of the antigen that interacts with the T-cell receptor (TCR) and MHC is not known.

Pulmonary lesions in sarcoidosis and CBD are histologically indistinguishable (11), but the etiology of sarcoidosis is unknown. These diseases also display similar BAL abnormalities with regard to increased total white-cell and T-cell numbers and CD4⁺ T-cell enrichment. Studies in sarcoidosis have demonstrated CD4⁺ T-cell expansions in the BAL that express particular TCR variable (V) regions. Although the V and V regions expressed by the expanded subset(s) varied among different individual sarcoidosis patients, a trend for increased usage of V8 and V2.3 was reported in some studies (12). Junctional region sequencing showed that the expanded subsets were clonal or oligoclonal in origin, consistent with stimulation by conventional antigen in the lung. We hypothesized that selective T-cell expansions would also characterize BAL CD4⁺ T cells in CBD and that V region expression would be different and more uniform compared with sarcoidosis.

In the present work, we demonstrate skewing of the TCR V repertoire in the BAL compared with the blood of patients with CBD and with healthy controls. Approximately one-third of the T-cell expansions identified in the BAL of the different patients expressed V3, and overall, the repertoire alterations appeared to be distinct from that previously reported in sarcoidosis. Alterations of CD4⁺ TCR expression tended to persist in BAL over time and, in some cases, could be further enhanced by in vitro stimulation with beryllium salts. These findings suggest that beryllium-activated CD4⁺ T cells accumulate and persist in the lungs of patients with CBD, and are likely to be important in the pathogenesis of this disease.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Study Population

The diagnosis of CBD was established in 28 patients (age 50 ± 2.1 yr; 20 men, 8 women; 13 never smokers, 2 current smokers, 13 former smokers) by using previously defined criteria (6, 15), including the presence of granulomatous inflammation in lung biopsies, a beryllium exposure history, and a positive BAL lymphocyte proliferative response to BeSO₄. Exposure to beryllium occurred in the following circumstances: ceramics industry (n = 14), nuclear weapons plant (n = 13), and other (n = 1). The patients with CBD included 19 initially asymptomatic individuals who were diagnosed through screening programs. Three of these individuals subsequently became symptomatic. In addition, nine individuals were initially evaluated after the development of symptoms. The presence of symptoms correlated with more severe functional and radiographic abnormalities, and these 12 patients with CBD were being treated with corticosteroids prior to studying their BAL T-cell repertoire. Seventeen of the 28 patients with CBD were serially studied over a period of 4 yr. A group of 10 healthy individuals (age 43 ± 2.9 yr; 5 men, 5 women; 9 never smokers, 0 current smokers, 1 former smoker) served as controls for the blood and BAL studies. Informed consent was obtained from each patient and volunteer, and the protocol was approved by the Human Subjects Institutional Review Board, National Jewish Medical and Research Center. Table 1 shows the age, gender, race, and smoking status for the patients and healthy controls in this study.

Isolation and Initial Analysis of Mononuclear Cells from Peripheral Blood and BAL

Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized blood by Ficoll-Hypaque density gradient separation, and the percentage of T cells expressing CD4 and CD8 are shown in Table 1. BAL was performed as previously reported (16), and the yield of white blood cells per milliliter and the percentages of lymphocytes, macrophages, neutrophils, and eosinophils in patients and controls are shown in Table 1. BAL lymphocytes from CBD patients had a significantly larger fraction of CD4⁺ T cells and a smaller fraction of CD8⁺ T cells compared with controls (Table 1). Especially considering the higher white-cell counts and the higher percentages of lymphocytes, CBD patients demonstrated a 9.8-fold increase in the absolute number of CD4⁺ T cells in BAL compared with healthy controls.

Beryllium lymphocyte proliferation tests were performed as previously described (15) in the CBD patients. The mean (± SEM) peak stimulation indices for blood and BAL T cells in response to BeSO₄ exposure were 24.5 ± 5.7 and 59.6 ± 20.2, respectively. In studies at our institution, the mean stimulation index ± 2 SD for healthy controls was 1.9, which was used as a cutoff to indicate an abnormally elevated test.

Human leukocyte antigen (HLA) typing was performed by standard serologic techniques (Immunological Associates of Denver, Denver, CO).

Immunofluorescence Analysis of TCR V and V Expression

Mononuclear cells were analyzed by two-color immunofluorescence for TCR V region gene expression using monoclonal antibodes (mAb) directed to nine different TCR V receptors and two TCR V receptors. Most BAL and blood samples were stained using biotinylated mAb directed against V2 (clone E22E7.2) (17), V3.1 (clone 8F10) (18, 19), V5.1 (clone LC4) (20), V6.7 (clone OT145) (21), V8.1/8.2 (clone MX8) (22), V12 (clone S511) (23), V13.1 (clone H131) (24), V13.2 (clone H132) (24), V17 (clone E175F3) (17), V2.3 (clone F1) (12), and V12.1 (clone 6D6) (25). Anti-V mAb were obtained as follows: V2 and V17 (Immunotech, Marseille, France); V3.1, V5.1, V8.1/8.2, V12, V2.3, and V12.1 (T Cell Sciences, Cambridge, MA). The mAb to V13.1 and V13.2 were generated as described (24) and are available upon request. Streptavidin-phycoerythrin (Fisher Biotech, Pittsburgh, PA) was used as a second-step reagent for TCR staining, and cells were also double-stained with fluorescein isothiocyanate (FITC)-labeled CD3, CD4, and CD8 (all from Becton Dickinson, San Jose, CA). The lymphocyte population was identified using forward and 90-degree light-scatter patterns, and fluorescence intensity was analyzed using an Epics Profile Cytometer (Coulter Electronics, Hialeah, FL) as previously described.

Stimulation of BAL Cells with BeSO₄ and Interleukin-2

Mononuclear cells derived from the BAL of 13 CBD patients were resuspended at a concentration of 2 × 10⁶ cells/ ml in culture media (RPMI 1640 with 10% fetal calf serum, 2 mM L-glutamine, 50 U/ml penicillin, and 50 µg/ml streptomycin) and stimulated with BeSO₄ at a concentration of 1 × 10⁵ M. After 72 h of culture, the stimulated cells were resuspended in fresh culture media with the addition of 25 U/ml recombinant interleukin (IL)-2 (Amgen, Thousand Oaks, CA), and the cultured cells were harvested at Day 7. Cells from four patients were cultured for an additional seven days with the addition of fresh culture media and IL-2 and then analyzed on Day 10. Analysis of TCR V region expression on stimulated BAL cells was performed as described above for the analysis of the unstimulated BAL cells and PBMC with the exception that forward and 90-degree light-scatter patterns were used to gate on lymphocyte blasts rather than the smaller lymphocyte population (23, 26).

Statistical Analysis

Due to the non-Gaussian distribution of the data, we employed the Wilcoxon rank sum test within the CBD population and between the patients with CBD and normal control subjects (JMP^®, Version 3; SAS Institute Inc., Cary, NC). Due to the large number of comparisons, a P value  0.01 was used to determine statistical significance.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

TCR Expression in Peripheral Blood and BAL T Cells of Patients with CBD and Control Subjects

Freshly isolated BAL and peripheral blood cells were obtained from patients and control subjects, and the percentage of CD4⁺ T cells expressing particular V regions was determined by immunofluorescence staining and cytofluorographic analysis. Figure 1 shows representative examples of TCR repertoires for patients who demonstrated an expansion of at least one V subset in the BAL (defined as a 2-fold increase in the percentage of V-expressing cells in the BAL compared with blood). For example, Patient 1 showed an increased percentage of V8.1⁺ T cells in the BAL compared with the blood CD4 population (13.2% versus 3.8%). In Patient 2, 12% of the CD4⁺ T cells in BAL expressed V3 as compared with 5.6% in the blood. Patient 2 also demonstrated an expansion of V6.7 in the BAL compared with blood (6.5% versus 3.0%). Patient 3 demonstrated a V3 expansion with 10.6% of the CD4⁺ T cells in the BAL expressing V3 compared with 2.0% in the blood, and Patient 4 was noted to have a V13.2 expansion (4.9% versus 0.4%).

View larger version (22K): [in this window] [in a new window]   Figure 1.   TCR V repertoire in BAL compared with blood in four representative CBD patients who demonstrated evidence of an expansion (> 2-fold) in the BAL compared with the blood. Data are expressed as the mean percentage of CD4+ T cells expressing a particular V. The asterisk is used to identify the expanded V in the individual patients.

Overall, using this definition for T-cell subset expansion, 11 of the 28 CBD patients demonstrated expansions. One patient had three expansions and three others had two expansions each. In contrast, only one of the controls demonstrated one V5.1 expansion. Of the 16 T-cell subset expansions in CBD patients, five were V3, three expressed V13.2, and two expressed V6.7 or V8.1/8.2; V5.1, V12, V2.3, and V12.1 were each represented once.

Figure 2 shows mean (± SEM) levels of the TCR V region expression in the BAL and blood of all patients and healthy control subjects. Comparison of BAL versus blood demonstrates remarkable heterogeneity within the BAL repertoire in both patients and controls. Thus, for most V- and V-expressing subsets, the levels in the BAL and blood were similar. The results also indicate that the BAL CD4⁺ repertoire is not just a passive reflection of blood T cells. For example, both controls and patients showed an increase in the levels of V2⁺ cells in blood compared with BAL (P < 0.001 for both controls and patients with CBD). In the control subjects, significant increases in the percentage of V3⁺ and V17⁺ cells were also seen in the blood compared with BAL. However, only CBD patients demonstrated increases in the mean levels of BAL V- expressing subsets compared with blood. We observed a trend for an increase in the expression of V6.7 (4.8 ± 0.5 versus 3.9 ± 0.4%; P = 0.035) and V13.2 (1.5 ± 0.4 versus 0.8 ± 0.2%; P = 0.027) within the BAL as compared with blood of patients with CBD.

View larger version (31K): [in this window] [in a new window]   Figure 2.   TCR V region expressed in the blood and BAL of ten controls (A) and 28 CBD patients (B). Data are expressed as the mean ± SEM percentage of CD4+ T cells expressing a particular V or V. V mAb staining was performed in eight of the 28 CBD patients.

We also compared the mean levels shown in Figure 2 in patients with CBD versus controls. As expected, the percentage of each of the subsets in blood was similar in both groups. However, a significantly greater expression of V3⁺ cells in the BAL of patients with CBD compared with control subjects (5.3 ± 0.7% versus 1.8 ± 0.6%; P = 0.005) was noted, as well as trends for increases in V6.7, V8.1, V13.2, and V17.

Figure 3 shows differences in the percentage of V- or V-expressing subsets in the BAL minus blood in each individual studied. A positive value reflects an increased percentage of a particular subset in BAL. If BAL T cells were a passive reflection of those in blood, we would expect the values to cluster around zero. Overall, the scatter was much greater in patients with CBD compared with healthy controls, related to a greater number of high positive values. Consistent with mean values shown above, the percentage of V2⁺ cells was consistently higher in the blood compared with BAL in controls and patients, which is shown as negative values in Figure 3. Similar negative values were apparent for V3, but only in the controls. When we compared the BAL-versus-blood differences in patients and controls, we noted a significant difference for V3⁺ cells (P = 0.005) and a trend for V12.1 (P = 0.03). The largest positive difference (i.e., percentage in BAL greater than blood) in the control subjects was 3.3% (the one V5.1 expansion described above). If we apply this value as a cutoff to define expansions, 17 T-cell subset expansions were identified in the patients with CBD, and, as expected, a strong correlation was apparent with the expansions described above. The results in Figure 3 also suggest that our analysis did not miss large expansions in different patients with CBD that express Vs not covered by the panel of anti-TCR mAb. Thus, for patients with CBD versus controls as a group, a trend for negative values in the subsets covered, caused by reciprocal changes from a large expansion, was not apparent. In addition, no individual patient with CBD showed multiple negative BAL-minus-blood percentages for the subsets analyzed.

View larger version (14K): [in this window] [in a new window]   Figure 3.   Percentage of CD4+ T cells expressing a particular TCR V region in the BAL minus the percentage in the blood of normal controls (A) and CBD patients (B). The solid line represents the mean of each V region analyzed. The dashed line (at 3.3%) represents the largest positive percent difference in the healthy controls.

Presence of V Region Expansions in Relation to Clinical Variables

Of the 11 CBD patients with T-cell subset expansions, six were exposed to beryllium while employed in the ceramics industry and five had worked in a nuclear weapons plant. Three of the five patients with V3⁺ expansions had exposures in the ceramics industry. Thus there appeared to be no predilection for one type of exposure to be associated with the presence of an expansion or for particular V usage. Six of the 19 individuals initially diagnosed through screening programs had detectable V region expansions, compared with five of the nine patients diagnosed after symptom development. Corticosteroid therapy had been initiated in 12 patients with CBD with symptoms prior to enrollment in this study; overall, these individuals demonstrated more severe clinical, physiologic, and radiographic abnormalities. Half of these patients had detectable expansions, whereas six of the 16 asymptomatic patients also had expansions. There was no correlation between the duration of symptoms or corticosteroid treatment and the presence of a detectable expansion (data not shown).

Longitudinal Studies of BAL TCR Repertoire in Patients with CBD

Seventeen of the 28 patients with CBD underwent repeat BAL and TCR repertoire analysis during the study period. Eight of these patients were receiving corticosteroid treatment at the time of the subsequent sampling. Figure 4 shows representative longitudinal results for six patients who demonstrated skewing of V expression in the initial analysis. Subsets of V3, V6.7, and V8.1 are shown because these subsets encompassed most of the abnormalities found in the CBD population. Even though we observed some variability over time, the results indicate that the TCR V expansions in individual patients' BAL persist. Eight of the 11 patients who demonstrated a particular V expansion at the time of the first evaluation underwent repeat analysis, and persistent V expansions (same V subset) were found in six of the eight patients. In addition, none of the patients with CBD who failed to show altered V3 expression in the BAL on first evaluation developed a V3⁺ expansion upon serial analysis. Patient 6 in Figure 4 had a V3 expansion at the first analysis and continued to have an elevated V3 subset in the BAL compared with blood (6.7% versus 3.7%); however, it did not meet our definition of an expansion. Longitudinal changes in patients being treated with corticosteroids did not appear to be different compared with untreated patients (Figure 4). Overall, CBD patients analyzed serially demonstrated similar TCR V region repertoire alterations over time.

View larger version (26K): [in this window] [in a new window]   Figure 4.   Longitudinal studies of BAL TCR repertoire in CBD patients. These patients demonstrated altered TCR V expression (five with elevated V3 and one with elevated V8.1) at the time of the initial evaluation. The square, triangle, and circle represent the percentage of CD4+ T cells expressing V3, V6.7, and V8.1, respectively, at different time points. The arrow represents the initiation of corticosteroids.

In Vitro Stimulation of T Cells with BeSO₄ and IL-2

Previous studies have demonstrated that lymphocytes obtained from the BAL and from the blood of patients with CBD respond to beryllium in culture by proliferation and with the production of IL-2, IL-6, and tumor necrosis factor- (8, 10, 27). BAL lymphocytes from 13 patients with CBD (including four of the 11 with BAL TCR V expansions) were stimulated in culture with BeSO₄, followed by the addition of IL-2 on Day 3. After seven days of culture, the majority of viable cells were blasted, as determined by forward angle and 90-degree side-scatter light patterns. As shown in Figure 5, this short-term culture in BeSO₄ and IL-2 resulted in additional alterations of the TCR V repertoire in some patients. The four CBD patients who showed V expansions on initial evaluation expressed five different expansions. Three of the five TCR V expansions in the patients with CBD who had V expansions at baseline demonstrated additional increases in the percentage of cells expressing that V. For example, Patient 3 demonstrated a further 9.1% increase in the percentage of CD4⁺ T cells expressing V3 (10.6% to 19.7%). However, there were also exceptions to this pattern, with some patients demonstrating no change in the percentage of a subset after stimulation and an occasional patient without an expansion at baseline who developed a particular V expansion after short-term culture. For example, one patient showed an increase from 4.5% to 22.4% in the percentage of CD4⁺ T cells expressing V17 (Figure 5).

View larger version (22K): [in this window] [in a new window]   Figure 5.   TCR repertoire of BAL CD4+ T cells before and after in vitro stimulation with BeSO4 and IL-2. For clarity, only certain Vs are shown. The percentages of CD4+ T cells expressing a particular V are shown for Day 0 and after 7 d of culture and are connected by a solid line. The open circles represent those CBD patients with TCR V expansions at the time of the first evaluation. The open squares show those CBD patients who developed a TCR expansion after short-term culture with BeSO4 and IL-2. An asterisk indicates that this particular patient was evaluated at Days 0 and 14. A dagger marks one patient who developed a V17 expansion after short-term culture.

Previous studies from our laboratories have demonstrated that BAL cells from healthy control subjects do not proliferate or blast in response to BeSO₄ or IL-2 (28). Therefore, BAL cells from controls were not cultured in this study.

HLA Typing of Patients with V Subset Expansions

Differences in the TCR V regions expressed by expanded subsets could be explained at the level of antigen and/or class II MHC-presenting molecule. It seemed possible, therefore, that the V3 expansions identified in this group of patients with CBD were related to a common HLA haplotype. Table 2 compares the results of HLA-DR and -DQ serological typing for the five patients with V3 expansions and one patient with V8.1 expansion. No obvious association between HLA phenotype and alterations in the TCR V expression was observed, although the one patient with CBD with a V2.3 expansion was typed as HLA-DR3. This association between HLA-DR3 and V2.3 expansions has been previously observed by Grunewald and colleagues (12, 14) in sarcoidosis.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

It has been estimated that 800,000 individuals in the United States have current or past beryllium exposure and are at risk for developing CBD (29). Chronic lung disease will develop in about 1 to 5% of these individuals (6), and most cases will not be apparent until years after leaving the workplace in which exposure to beryllium occurred. In addition to exposure, genetic susceptibility has been postulated to be an important variable in the immunopathogenesis of CBD. At least one component of this susceptibility is likely to control the quantity and/or quality of the immune response to beryllium. Considerable data suggest that CD4⁺ T cells are critically involved in the disease process. Thus sensitization is detected by the ability of CD4⁺ T cells to proliferate upon exposure to BeSO₄ in culture (6, 15, 30), and the development of granulomatous inflammation in the lung is associated with the accumulation of CD4⁺ T cells in BAL (1). The present TCR repertoire studies were initiated to provide insight into the nature of these infiltrating T cells, especially to identify subsets with skewed TCR expression and therefore T cells that may have responded to specific antigens. We used immunofluoresence staining to quantitate TCR expression, rather than semiquantitative polymerase chain reaction (PCR) techniques, in an attempt to increase the accuracy of quantitation and to decrease variability in longitudinal studies. Although a number of new anti-TCR mAb have been generated in the last few years (12, 17), our reagents still only covered about 30-35% of the CD4⁺ TCR repertoire. Despite this drawback, we identified at least one BAL CD4⁺ T-cell subset expansion in approximately 40% of the patients with CBD evaluated. It seems unlikely that our analysis missed large expansions that express Vs not covered by the panel of anti-TCR mAb, especially if similar V regions were utilized in different patients. Thus reciprocal decreases in BAL subset percentages compared with blood in the same individual were not apparent. Although our panel of mAb could have missed other relatively small expansions, it seems likely that a more comprehensive panel of anti-TCR mAb would simply reinforce the heterogeneous T-cell response already identified.

On the whole, TCR expression in the lungs of the majority of patients with CBD was heterogeneous and similar to that expressed by blood T cells in the same individuals. This was not unexpected. Prior studies in sarcoidosis have suggested that only a subset of T cells are likely to be activated and expanded in the BAL (28). A similar phenomenon seems likely to occur in CBD. This heterogeneous repertoire may reflect the influx of nonspecific T cells to the site of lung inflammation. For example, studies of mice after immunization with myelin basic protein demonstrated little enrichment of the pathogenic T-cell subset in inflammatory brain lesions (31). Instead, T cells at the site of pathology expressed a repertoire similar to lymphoid tissue, with the pathologic subset being obscured by the influx of nonspecific T cells. In prior studies of T cells in patients with sarcoidosis, clonal T-cell populations and skewed TCR expression could be regularly demonstrated only after BAL CD4⁺ T cells were cultured with IL-2 to expand previously activated T cells (28). T-cell responses to numerous different antigens or use of different presenting molecules in the lung could also contribute to a heterogeneous BAL T-cell repertoire in patients, perhaps reflecting the ability of beryllium to complex with multiple different proteins or peptides. The type of beryllium exposure, presence or duration of symptoms, or treatment with corticosteroids appeared to have no effect on either the presence of a BAL T-cell subset expansion or V region expressed.

We defined an expansion as a 2-fold increase in the percentage of a particular V or V subset in BAL compared with blood of the same individual. TCR expression of blood T cells in patients with CBD appeared to be altered little and provided a reasonable basis for comparison with the BAL TCR repertoire. This definition also allowed for known differences in baseline T-cell V region expression among different individuals (18, 21, 23, 28, 32). We also reanalyzed the results using the mean percentage ± 3 SD for the corresponding subset in control BAL. Using this cutoff to define an expansion, 28 expansions were identified in 18 CBD patients compared with none in controls. Although the overlap in identified expansions was extensive, our initial definition for expansion was more stringent. In addition, we compared the percentage of CD4⁺ T cells expressing a particular V or V in the BAL minus blood in both the patients with CBD and healthy control subjects. The largest positive percent difference in the control subjects was 3.3%. In contrast, there were 17 subsets in the patients with CBD with a BAL-minus-blood percentage difference greater than 3.3%. As expected, comparison of expansions defined by this approach showed a strong correlation with those expansions defined by the 2-fold increase in the percentage in BAL versus blood. Overall, these considerations appear to support the approach used to define expansions in the BAL of patients with CBD.

In 11 of 28 patients studied, we identified 16 BAL T-cell subset expansions. These expansions reflect large increases in the absolute number of a particular T-cell subset, considering that the CBD patients had an average 9.8-fold increase in absolute number of CD4⁺ T cells in the BAL compared with the control group. In some patients, we estimated a 30- to 50-fold increase in the number of cells expressing a particular TCR V. About one-third of the expansions expressed V3, which is much greater than expected based on the percentage of T cells in this subset. For example, in past studies, the mean percentage of CD4⁺ T cells expressing V3 was about 5% (26, 28, 32), which is close to the levels found in the blood of patients and control subjects in the present study. The increased percentage of V3⁺ cells in the BAL of patients with CBD was even more remarkable considering that V3⁺ cells appeared to be relatively excluded from the BAL of healthy controls. Indeed, this selective decrease in the BAL of control subjects suggested that the number of patients with CBD with increased percentages of V3⁺ cells was underestimated by our relatively stringent definition of expansion. The relevance of the increased percentages of V3⁺ T cells (or other subsets) was further strengthened by the persistence of most expansions in patients followed longitudinally and by the further expansion of particular V-expressing subsets after in vitro stimulation with BeSO₄. The relatively frequent increase in V3 expression among patients suggests that a similar peptide/beryllium complex or other forms of beryllium complex may be involved in the stimulation of T cells in the lungs of these patients. It is also important to emphasize that our set of anti-TCR mAb covered only a limited portion of the CD4 repertoire (see above), which could have contributed to our not finding expansions in a number of patients. In addition, immunofluorescence staining for the percentage of cells may have missed small but expanded T-cell populations within a particular V or V subset that express the same or similar TCR complementarity determining region 3.

Our studies also provided a description of the TCR repertoire of BAL T cells in healthy control subjects, which may reflect T cells that remain after responses to respiratory stimuli or perhaps T cells that comprise a resident lung population. In general, we found that the repertoire of BAL CD4⁺ T cells in the healthy controls was similar to that in the circulating population. We did note a relatively consistent exclusion of V2⁺ and V3⁺ T cells from the BAL. These findings are different compared with one previous report in which semiquantitative PCR techniques were used to quantitate TCR expression of BAL T cells in healthy control subjects (33). The previous study did not separate CD4 and CD8 populations. The difference in results and the reason for the relative exclusion of certain T-cell subsets from BAL are unexplained at this time.

T cells from patients with CBD respond to BeSO₄ in a class II MHC-restricted manner (1), and we hypothesized that the patients with V3⁺ T-cell expansions might share a common class II MHC haplotype and therefore share a similar presenting MHC/antigen complex. However, HLA-DR and -DQ types varied considerably in the patients with CBD with altered V3 expression, and common HLA molecules were not identified. It is possible that molecular typing might disclose some similarities among the different DR or DQ molecules expressed by these patients. More importantly, a previous study suggested that nearly all patients with CBD may express a particular HLA-DP (DPB1*0201) molecule compared with the general population (34). We did not examine DP expression in our patients, and HLA-DP may not be in linkage disequilibrium with HLA-DR or -DQ. The best demonstration of the potential importance of HLA in granulomatous lung disease is in sarcoidosis. Grunewald and associates (12, 14) found that patients carrying HLA-DR3 had increased percentages of CD4⁺ T cells expressing V2.3 selectively in the BAL during periods of active disease. We found one patient who had a V2.3 expansion, and this patient expressed a haplotype encoding HLA-DR3.

Several previous studies have examined the TCR repertoire of BAL T cells in patients with sarcoidosis, a granulomatous lung disease of unknown etiology (12, 26, 28, 33). Similar to our findings in CBD, patients demonstrated very heterogeneous repertoires and some were noted to have V skewing and clonal expansions. In a number of patients, skewing and clonal expansions were not apparent until BAL T cells were cultured in IL-2. Among the expanded V subsets and clones in sarcoidosis patients, TCR expression was remarkably diverse, although there was a suggestion for increased usage of V8 and V2.3 (in HLA-DR3 individuals) in certain studies (12). In the present study, we identified two patients with CBD with V8 expansions and only one expressing V2.3, and we found an increased number of patients with CBD with V3⁺ T-cell expansions. These results suggest that the TCRs expressed by expansions in CBD and therefore the antigens recognized in the lungs are distinct in these two diseases with similar lung pathology.

Together, our results suggest that there is a selective expansion or selective accumulation of certain CD4⁺ T-cell subsets in the lungs of patients with CBD, and it seems reasonable to speculate that this is related to local stimulation by an antigen that involves beryllium. The nature of such an antigen is unknown. It has been hypothesized that beryllium may bind to different self peptide(s), which are recognized as foreign when presented by class II MHC molecules. Our findings showing selective expansions that express different Vs in different patients appear to be most consistent with conventional antigen (peptide) recognition. The variability of TCR V region usage among different patients with CBD makes stimulation by the same superantigen(s) unlikely. However, prior to sequencing the junctional regions of the TCR expressed by expansions and determining whether clonal T-cell expansions are present, we cannot exclude the possibility that beryllium combines with self peptide and class II MHC in some way to stimulate T cells as a superantigen. The T-cell expansions identified in the present studies should provide the tools necessary to understand how beryllium causes T-cell stimulation and T-cell-mediated disease.