Beryllium-Induced TNF-a Production by CD4 T Cells is Mediated by HLA-DP by

Beryllium (Be) presentation to CD4 T cells from patients with chronic beryllium disease (CBD) results in T cell activation, and these Be-specific CD4 T cells undergo clonal proliferation and Th1-type cytokine production. In exposed workers, genetic susceptibility to this granulomatous disorder is associated with particular HLA-DPB1 alleles. We hypothesized that these HLA-DP molecules could mediate Be-stimulated TNF-a mRNA and protein production. Using intracellular cytokine staining, we found that treatment with an anti-HLA-DP, but not anti-HLA-DR, mAb inhibited Be-stimulated TNF-a expression in lung CD3 CD4 T cells. This mAb also blocked Be-specific T cell proliferation, increased production of TNF-a mature-mRNA transcripts, and increased TNF-a protein production by Be-stimulated CBD peripheral blood mononuclear cells and bronchoalveolar lavage (BAL) cells. The Be-stimulated up-regulation of TNF-a mature-mRNA levels with TNF-a protein production was a unique property of CBD BAL cells, and did not occur in BAL cells from beryllium sensitized patients without CBD, or sarcoidosis BAL cells. This study identifies HLA-DP as a regulatory component in the activation of T cell receptors on Be-specific CD4 T cells from CBD patients resulting in proliferation and pro-inflammatory cytokine production. Abbreviations: Beryllium (Be); Chronic Beryllium Disease (CBD); Beryllium Sensitization (BeS); Major Histocompatibility Complex (MHC); Bronchoalveolar Lavage (BAL); Peripheral Blood Mononuclear Cells (PBMC); Tumor Necrosis Factor-alpha (TNF-a); Beryllium-Lymphocyte Proliferation Test (BeLPT); Introduction Chronic beryllium disease (CBD) is a human granulomatous lung disease for which the causative antigen, beryllium (Be) is known [1]. The development of this disease depends on the presence of Be-specific, effector-memory CD4 T cells in the lungs of CBD patients [2]. Following Be-exposure in culture these bronchoalveolar lavage (BAL) CD4 T cells proliferate and produce large amounts of IL-2, TNF-a and IFN-g [2-5]. High TNF-a levels are associated with disease severity in CBD [6]. Conversely, BAL cells and peripheral blood mononuclear cells (PBMC) from both normal control subjects and beryllium sensitized (BeS) subjects do not up-regulate TNFa protein production after Be-exposure in culture [4,5]. In CBD, the major histocompatibility complex (MHC) class II molecule HLA-DP plays a key role in triggering cytokine production through T cell receptor (TCR) signaling [2]. Genetic susceptibility to both Be sensitization and CBD is governed in part by the presence of a single amino acid polymorphism at position 69 (Glu) of the DPbchain, in certain HLA-DP molecules [7-10]. The mechanism for this genetic susceptibility lies in the ability of certain HLA-DP molecules to bind and present Be to pathogenic CD4 T cells [2,11,12]. These particular HLA-DPB1 alleles drive the differentiation of CD4 T cells bearing TCRs specific for Be-antigen [11]. In the BAL of CBD patients expansions of CD4 T cells bearing related or identical TCRs have been identified suggesting recognition of a Be-antigen complex [11,12]. Studies confirm that anti-MHC class II monoclonal antibody (mAb), in particular anti-HLA-DP mAb blocks proliferation and IFN-g expression in Be-stimulated T cells from CBD patients [2,12,13]. Upon T cell activation TNF-a is among the first cytokines produced [14]. Activation signals trigger TNF-a release by T cells causing inflammation and severe clinical symptoms in humans and animals [15-17]. In lymphocytes, activation signals can drive both the direct splicing of preexisting pre-mRNA transcripts [18] and the translation of preexisting mature-mRNA [19]. Study suggests that Be up-regulates TNF-a production by CBD peripheral blood monocytes in an HLA-independent manner [20]. However, this study [20], and our own past studies [2,4], did not determine if HLA-DP was responsible for the up-regulation of TNF-a mRNA pre-mRNA or mature-mRNA transcripts through Be-specific CD4 T cell activation. We therefore tested the hypothesis that Be exposure could activate the up-regulation TNF-a mRNA and protein expression in CD4 T cells from CBD patients in an HLA-DP dependent manner. We used an antiHLA-DP mAb to block TCR activation and subsequently block the increased production of TNF-a mature-mRNA and the resulting protein production. We found that HLA-DP not only mediates the activation and proliferation of Be-specific CD4 T cells, but it also triggers increased TNF-a mature-mRNA production, only in CBD cells. Methods Subjects. Thirty-eight CBD, 21 BeS, and 8 sarcoidosis subjects were consecutively enrolled from the National Jewish Medical and Research Center’s (NJMRC) Occupational and Environmental Health Clinic. The diagnosis of CBD was established using previously defined criteria, including a history of Be exposure, the presence of granulomatous inflammation and/or mononuclear cell infiltration on lung biopsy, and a positive proliferative response of PBMC and/or BAL T cells to BeSO4 in vitro [21-23]. The diagnosis of BeS was established based on a history of Be exposure, positive proliferative response of PBMCs to BeSO4 in vitro, and the absence of granulomatous inflammation on lung biopsy [21-23]. The diagnosis of sarcoidosis was established based on the presence of granulomatous inflammation on lung biopsy, the absence of a history of Be exposure and a negative proliferative response of PBMC and BAL T cells to BeSO4 in vitro [21-23]. All subjects provided written informed consent according to the protocol approved by the Human Subjects Review Board and completed a modified version of the American Thoracic Society (ATS) respiratory questionnaire [24]. Chemicals and Reagents. A stock solution of 0.2 M BeSO4 (Brush Wellman Inc., Cleveland, OH) was diluted to final concentrations of 100 mM and 10 mM BeSO4 used to treat cells in culture. The amounts of Be used to treat cells were selected based on previous studies that establish optimal stimulatory concentrations [2-6,10-12], as specified in the results. A stock solution of lipopolysaccharide (LPS, E. coli serotype 01111.B4, Sigma Chemical Co., St. Louis, MO) at a final concentration of 1 mg/ml was diluted 1:10 to treat cells in culture. As a positive control for the activation of TNF-a protein production by T cells, Staphylococcal enterotoxin B (SEB, Sigma) was used at a final concentration of 10 ng/ml. Brefeldin A (Golgi Plug, BD Biosciences, San Diego, CA) was used in intracellular TNF-a staining experiments, at a final concentration of 10 mg/ml. The B7.21, anti-HLA-DP monoclonal antibody (mAb) producing hybridoma was a gift of Dr. Ian Trowbridge, Salk Institute, La Jolla, CA. The B7.21 clone was cultivated as described [26], and anti-HLA-DP B7.21 mAb was affinity purified using the ImmunoPure (Protein G) kit (Pierce Chemical Co., Rockford, IL). Anti-HLA-DR LB3.1 mAb (ATCC HB-29B, Manassas, VA) purification was described previously [2]. Cell cultures were incubated in the absence and presence of anti-HLA-DP B7.21 mAb, or anti-HLA-DR LB3.1 mAb, at concentrations of 30 mg/ml, 10 mg/ml and 1 mg/ml. The MOPC 21 IgG1 kappa mAb served as the isotype control for experiments using these mAbs (BD Biosciences). For the identification of cells producing intracellular TNF-a, primary labeled anti-TNF-a, anti-CD3, anti-CD4 antibodies, and their corresponding isotype control antibodies, were purchased from BD Biosciences (San Diego, CA), and used according to the manufacturers instructions. Preparation of peripheral blood and BAL cells and cell culture. At bronchoscopy, BAL was performed as reported [25]. Recovered peripheral blood mononuclear cells (PBMC) and BAL cells were separated using density gradient centrifugation (Ficoll-Hypaque). The yield of BAL white blood cells (WBC) and the percentage of various cell classes are shown in Table 1. Viable PBMC and BAL cells were cultivated as previously described [3,4]. For all experiments PBMC and BAL cell concentrations were adjusted to 1 X 10/ml in RPMI 1640 supplemented with 10% heat-inactivated, iron supplemented, calf serum (Cambrex Bioproducts, Walkersville, MD), 20 mM HEPES, 1 mM sodium pyruvate, 100 U/ml penicillin G, 100 mg/ml streptomycin sulfate and 2 mM L-glutamine (all purchased from Life Technologies, Gaithersburg, MD). 200 ml aliquots of cells were cultured in triplicate wells per treatment condition, in 96 well round bottom plates (costar #3799, Corning Inc., Corning, NY) and either unstimulated (none), or stimulated, as specified in the results. Immunofluorescence analysis for intracellular TNF-a expression. Methods for identification of CD3 CD4 CBD BAL T cells and intracellular TNF-a staining after beryllium-exposure are described [2,27-29]. A subset of CBD BAL cells (n = 5) were incubated with medium alone, 10 ng/ml SEB, or 10 mM BeSO4 for 6h, with 10 mg/ml of brefeldin A added after the first hour of stimulation [27-29]. After stimulation, cells were washed and stained with mAb’s directed against CD3 and CD4 (BD Biosciences, San Diego, CA)[2,29]. After staining for external surface marker expression, the cells were washed with PBS + 1% BSA and fixed (Caltag Laboratories, Inc., Burlingame, CA) for 15 min at room temperature. The cells were washed, permeabilized (Caltag Laboratories Inc.) and stained with mAb directed against TNF-a for 30 min at 4C. The lymphocyte population was identified using forward and 90 light scatter patterns, and gating was performed on CD3-expressing lymphocytes, in order to differentiate lymphocytes from macrophage sub-populations that might express low levels of CD4. BAL macrophages were identified using forward and 90 light scatter patterns and gating was performed to exclude CD3-expressing lymphocytes. Immunofluorescence intensity was analyzed using a FACSCalibur cytometer (BD Biosciences)[2]. In a second set of cultures CBD BAL cells (n = 5) were unstimulated, or exposed to 10 mM BeSO4 in the presence of 30 mg/ml, 10 mg/ml or 1 mg/ml mAbs directed against either HLA-DP (B7.21) or HLA-DR (LB3.1) at the beginning of culture as previously described [2]. Similar experiments with anti-HLA-DQ mAb were not performed since this MHC class II molecule has not been implicated in CBD pathogenesis [2]. The data for this experiment are expressed as the percent (%) of inhibition of TNF-a expression in comparison to the Be-stimulation control, as previously described [2]. Beryllium-induced lymphocyte proliferation. For clinical evaluation of Be sensitization the blood and BAL beryllium lymphocyte proliferation tests (BeLPT) were performed according to the clinical assay described by Mroz et al. [22]. PBMC and BAL cell concentrations were adjusted to 1 X 10/ml of complete culture medium and 200 ml aliquots were then cultured in triplicate, in the absence or presence of 10 mM BeSO4. Untreated and treated cells were cultured for 4, 5 and 6 days. During the last 4 hr of culture, cells were pulsed with DNA specific precursor tritiated thymidine deoxyriboside (HTdR, S.A. 5 Ci/mM, Amersham, Arlington Heights, IL) H-DNA was harvested onto glass filters, and counted in a Packard TopCount NTX scintillation counter (Packard Instruments Co, Meriden, CT). Thymidine up-take for the unstimulated controls on days 4, 5 and 6 are normally in the range of 150 cpm to 500 cpm. In some experiments, 30 mg/ml of anti-HLA-DP B7.21 mAb was added at the beginning of the culture as previously described [11,12]. For the clinical evaluation of PBMC and BAL T cell proliferation in response to Be stimulation shown in Table 1, we report the mean + standard error of the mean (SEM) peak stimulation index (SI) as the ratio of the test sample counts per minute (cpm) to the cpm in the unstimulated (medium alone) control [22]. In a separate experiment, CBD (n = 8) PBMC cell concentration was adjusted to 1 X 10/ml of complete culture medium and 200 ml aliquots were then cultured in triplicate per treatment. PBMC were unstimulated (medium alone), or treated with 30 mg/ml of anti-HLA-DP B7.21 mAb alone, or 30 mg/ml of MOPC 21 isotype control mAb (BDBiosciences). A positive stimulation control was performed in which PBMC were simultaneously stimulated with 10 mM BeSO4 in the presence of 30 mg/ml of isotype control mAb. Test PBMC were treated with 10 mM BeSO4 in the presence of 30 mg/ml, 10 mg/ml and 1 mg/ml of anti-HLA-DP B7.21 mAb. Untreated and treated PBMC were cultured for 4, 5 and 6 days. During the last 4 hr of culture, cells were pulsed with HTdR (S.A. 5 Ci/mM, Amersham, Arlington Heights, IL), H-DNA was harvested onto glass filters, and counted in a Packard TopCount NTX scintillation counter (Packard Instruments Co, Meriden, CT). For this experiment (represented in Fig. 2) the peak cpm for each set of untreated and treated triplicate cultures were expressed as the mean + SEM cpm. TNF-a determinations. To determine culture supernatant TNF-a protein levels, PBMC concentrations from CBD patients (n = 7) were adjusted to 1 X 10/ml of complete culture medium and 200 ml aliquots were then cultured in triplicate, per treatment. PBMC were unstimulated (medium alone), stimulated with 100 mM BeSO4 or 10 mM BeSO4 in the presence of 30 mg/ml, 10 mg/ml or 1 mg/ml of anti-HLA-DP B7.21 mAb. As controls, PBMC were stimulated with MOPC 21 isotype control mAb + 10 mM BeSO4 or with 1 mg/ml of LPS. In a second experiment, BAL cells from CBD subjects (n = 10), BeS subjects (n = 19) and sarcoidosis subjects (n = 8) were adjusted to a concentration of 1 X 10/ml of complete culture medium and 200 ml aliquots were cultured in triplicate in the absence and presence of 10 mM BeSO4. For all experiments, culture supernatants were harvested 72h after Be exposure and TNF-a levels were determined ELISA kits (R&D; Systems, Minneapolis, MN)[4,5] with a sensitivity of 7.8 pg/ml. For reverse transcription (RT)-PCR analysis, PBMC and BAL cell concentration was adjusted to 1 X 10/ml of complete culture medium and 200 ml aliquots were then cultured in five wells per treatment. Cell cultures were unstimulated or exposed to 10 mM BeSO4 in the absence and presence of increasing amounts of anti-HLA-DP B7.31 mAb, as above. Cells were harvested for mRNA analysis after 72h of in vitro Be exposure. Total RNA was isolated from cell pellets using the RNeasy kit (Qiagen Inc., Valencia, CA). One mg of total RNA was reverse transcribed into cDNA using the RNA PCR Core Kit (Applied Biosystems, Foster City, CA) and 5 ml of the cDNA was used for PCR reactions. RNA samples were randomly selected for treatment with RQ1 RNase-Free DNase (Promega, Madison, WI) for 30 min at 37C and run against the untreated samples to control for the presence of DNA contamination in RNA samples. TNF-a primers for the PCR reactions were: 5’GAACCCCGAGTGACAAGCCTG that spans the +1530 to +1551 region of the TNF-a gene and 3’ GGCGGTTCAGCCACTGGAGCT that spans the +1889 to +2010 region of the TNF-a gene. TNF-a cycling conditions were: 94° 30s, 59° 30s, 72° 40s for 28 cycles. Composition of the final PCR reaction mixture included 3 mM of each primer, 1.5 mM MgCl2, 200 mM dTNPs, 2.5 units of AmphliTaq (Applied Biosystems) and 5 ml of 10X PCR Buffer II (Applied Biosystems), to a total volume of 50 ml. The primers were designed such that PCR amplification would result in two distinct bands. One band consists of 380 base pairs corresponding to TNF-a un-spliced, pre-mRNA that spans a region of the TNF-a gene from exon 3, through intron 3, to a region of exon 4. A second band consists of 90 base pairs corresponding to spliced TNFa mature mRNA that spans the region of the TNF-a gene from exon 3 to exon 4, with the intron 3 region removed. RT-PCR controls employed quantification of b-Actin expression. b-Actin primers were: 5’ATCGGCACCACACCTTCTACAATGCGCTGCG and 3’CGTCATACTCCT GCTTGCTGATCCACATGTGC. The final PCR reaction mixture included 3 mM of each primer, 1.5 mM MgCl2, 200 mM dTNPs, 2.5 units of AmpliTaq (Applied Biosystems) and 5 ml of 10X PCR Buffer II (Applied Biosystems) to a total volume of 50 ml. b-Actin cycling conditions were: 94° 45s, 60° 45s, 72° 2min for 30 cycles. This yields a PCR fragment consisting of approximately 800 base pairs. PCR samples were separated on a 1% Agarose-TAE gel by electrophoresis for 1 hr at 110V and the gels stained with Vistra Green (Amersham). The stained bands were scanned using a FMBIO II Phosphoimager (Hitachi Inc., Alameda, CA) and fluorescent bands were quantified (counts) using FMBIO Analysis 8.0 software. The RT-PCR data are presented as the ratio of either TNF-a pre-mRNA counts or TNF-a mature-mRNA counts to b-Actin mRNA counts. TNFa mRNA detection was performed on a subset of samples by real time PCR using the Applied Biosystems 7700 Sequence Detection System. The RT step was performed separately in a Perkin Elmer 9700 thermocycler, using one mg of RNA with the RNA PCR Core Kit (Applied Biosystems). 5 ml of cDNA and 3 mM of each primer were added to Applied Biosystems SYBR Green Master Mix to at total volume of 50 ml. Real-time primers for TNF-a pre-mRNA were: 5’ CTTAGTGGGATACTCAGAACG spanning the +1446 to +1468 region of the TNF-a gene and 3’ GGCGGTTCAGCCACTGAGCT spanning the +1534 to +1556 region of the TNF-a gene. PCR amplification yields a 161 base pair product. The 5’ primer was designed to overlap both the 3 and 4 exons of the TNF-a gene to control for erroneous amplification of TNF-a pre-mRNA or DNA. b -Actin primers were: 5’ GATGACCCAGATCATGTTTGA and 3’ ATGAGGTAGTCAGTCAGGTCC resulting in amplification of a 200 base pair product. Real-time PCR analysis of known TNF-a and b-Actin copy number standards allowed for quantification of test samples. Cycling conditions for real time PCRs were: 95 15 sec, 60 1 min, for 40 cycles. A dissociation curve was run and analyzed for all samples to confirm specificity of amplification. Total mRNA was separated on a 1% agarose-MOPS gel by electrophoresis, transferred by northern blot to a positively charged nylon membrane (Roche Diagnostics, Indianapolis, IN), incubated overnight at 24C and detected using the DIG Luminescent Detection Kit (Roche Diagnostics). The DIG-labeled northern analysis probe was: 5’TGCTGCACTTTGGAGTGATCGGCCCCCAGAGGGAAGAGTTCCCCAGGGACCT CTCTCTAATCAGCCCTCTGGCCCAGGCAGTCAGATCATCTTCTCGAACCCCGA GTGACAAGCCTGTAGCCCATGTTGTAGCAAACCCTCAAGC-3’. Statistics. Repeated Measures ANOVA was used to determine the effect of treatments while adjusting for the variability of subjects. In cases where there was also a time variable, a doubly repeated measures model was used. Individual contrasts were calculated to compare treatment means of interest. Normalizing transformations were made in cases where the data were non-Gaussian. When data transformations were unsuccessful, suitable nonparametric tests were substituted for parametric tests. Results Demographics. Clinical characteristics are shown in Table 1. Overall, no significant difference in age was observed between the study populations. Due to the hiring practices of the beryllium industry, most of the BeS and CBD subjects were male [30]. Eight percent of the CBD subjects and 20% of the BeS subjects enrolled in this study were current smokers. Although exposure to cigarette smoke has been shown to decrease antigen-presenting capacity of alveolar macrophages [31,32] we observed no significant difference in the clinical BeLPT proliferative response in blood with regard to current smoking status. Fifty percent of the sarcoidosis subjects enrolled in this study were also current smokers. Despite this, these sarcoidosis subjects continue to have a lymphocytic alveolitis, but their blood and BAL cells did not proliferate in response to Be stimulation in the clinical BeLPT. The BAL consists of a complex, mixed cell population. On a per ml basis, the BeS BAL consisted of 22.8 + 3 X 10 white blood cells/ml (mean + SEM) represented by 87% + 2% macrophages that serve as potential beryllium-antigen presenting cells (APCs) and 11% + 2% lymphocytes. The BeS BAL cellularity and differential of BeS subjects are approximately equal to that observed in normal control, non-Be exposed control subjects [3,4]. In contrast, the BAL cells from both CBD and sarcoidosis subjects are representative of a chronic T cell alveolitis. The BAL cellularity showed a significant elevation in the number of white blood cells per ml of BAL fluid from CBD (p = 0.024) and sarcoidosis (p = 0.032) subjects compared to the BAL of BeS subjects. As previously reported [21,30] we also observed an increase in the absolute number of lymphocytes per ml in the BAL cells from CBD. The BeS BAL contained 0.25 + 0.04 X 10 lymphocytes/ml (mean + SEM) whereas the CBD BAL contained 1.12 + 0.13 X 10 lymphocytes/ml (p = 0.0001 versus BeS), and the sarcoidosis BAL contained 1.3 + 0.35 X 10/ml (p = 0.0029 versus BeS). The CBD BAL contains approximately 2.2 X 10 macrophages/ml and the BeS BAL contains approximately 2 X 10 macrophages/ml. Consistent with the presence of lung disease in the CBD population and its absence in BeS subjects, the CBD BAL contains more than 4 times as many lymphocytes/ml as compared to the BeS BAL. The peak stimulation index (SI) for PBMC from CBD patients was 43 + 10 and 15 + 3 for BeS PBMC indicating a significant level of proliferation in response to Be exposure in vitro consistent with the presence of Be-specific blood T cells in both BeS and CBD [22]. In contrast, the peak SI for Be-exposed sarcoidosis PBMC was 1.4 + 0.2. In comparison, only BAL cells from CBD patients proliferated in response to Beexposure in vitro. The peak SI for CBD BAL cells was 64 + 17 in comparison to a peak SI of 2.4 + 0.3 for BeS BAL cells and 1.2 + 0.05 for sarcoidosis BAL cells. Thus, Bespecific T cells were only present in the lungs of CBD patients, and not in the BAL of either BeS or sarcoidosis patients. Inhibition of Be-induced TNF-a with an anti-HLA-DP mAb. We used a blocking mAb against HLA-DP, B7.21 and anti-HLA-DR, LB3.1, to determine the role of these MHC class II molecules in Be-induced TNF-a production in CD3 CD4expressing T cells. By intracellular cytokine staining, minimal background TNF-a production was observed in cells treated with medium alone, whereas SEB (10 ng/ml) stimulation induced TNF-a expression in 44% of the BAL CD4 T cells. In comparison to medium alone (0.06% CD4 T cells expressing TNF-a), 7.5% of the CD4 T cells from this CBD patient expressed TNF-a following Be-stimulation (10 mM BeSO4) (p = 0.008). Since alveolar macrophages are a potential source of TNF-a [1,4,17], we compared the fold-change in TNF-a expression in CBD (n = 5) BAL macrophages and CD3 CD4-expressing T cells, after BeSO4 exposure in culture. The percentage of CD4 T cells expressing TNF-a after BeSO4 exposure increased 89-fold compared to an increase of only 1.2-fold for alveolar macrophages. Thus, despite the slight ability of BAL macrophages to produce TNF-a after BeSO4 exposure, the predominant TNF-a producing cell population in the CBD lung, in response to BeSO4, was the CD3 CD4expressing T cell. The addition of anti-HLA-DP mAb inhibited Be-stimulated TNF-a production. For example, 30 mg/ml, 10 mg/ml, and 1 mg/ml of anti-HLA-DP mAb inhibited 91%, 86%, and 75% of Be-induced TNF-a production by CD4 T cells, respectively (Fig. 1a second panel). On the other hand, essentially no inhibition of Be-stimulated TNF-a expression was seen after the addition of anti-HLA-DR mAb (Fig. 1a third panel). Using BAL cells from 5 CBD patients, the addition of 30 mg/ml of anti-HLA-DP mAb resulted in a 96 ± 1.5% (range, 91-100%) inhibition of Be-stimulated TNF-a expression (Fig. 1b). Similarly, 10 mg/ml and 1 mg/ml of anti-HLA-DP mAb inhibited 94 ± 2.7% (range, 84-100%) and 82 ± 6.5% (range, 78-91%) of TNF-a expression. Much less inhibition (14 ± 5%) of TNF-a expression in CD4 T cells was seen with 30 mg/ml of the anti-HLA-DR mAb, LB3.1 (Fig. 1b). Overall, the predominant MHC class II molecule involved in the presentation of Be to CD4 T cells was HLA-DP. Due to this critical role of HLA-DP in T cell activation, proliferation and pro-inflammatory cytokine expression [2,11,12], the remainder of this study focused on HLA-DP. Inhibition of Be-induced PBMC proliferation, TNF-a protein, and TNF-a mRNA production with an anti-HLA-DP mAb. PBMCs from CBD patients (n = 8) enrolled in this study were examined for their proliferative responses in the presence of 10 mM BeSO4 with and without the addition of anti-HLA-DP B7.21 mAb. BeSO4 induced a proliferative response in the PBMC from these CBD patients with a mean + SEM thymidine incorporation of 1784 + 750 cpm as compared to medium alone (558 + 98 cpm, p = 0.006, contrasts from repeated measures ANOVA) (Fig 2). Increasing concentrations of anti-HLA-DP B7.21 mAb completely inhibited this proliferative response. For example, the addition of 30 mg/ml of B7.21 mAb decreased Be-induced proliferation to a mean thymidine incorporation of 449 + 131 cpm (p = 0.0003 versus Bestimulation), which approximates the background proliferation of PBMC (p = 0.29 versus none). The addition of an anti-IgG1 isotype control mAb alone did not induce CBD PBMC proliferation, 332 + 84 cpm (p n.s. versus none), and had no effect on Bestimulated proliferation of CBD PBMC, 1386 + 5,15 cpm (p n.s. versus the Be-stimulated alone, and p < 0.05 versus none). Exposure of CBD PBMC to 10 mM BeSO4 up-regulated TNF-a protein production to a median of 717 pg/ml (range; minimum = 472-maximum = 1417; p = 0.0005 versus none, contrasts for repeated measures ANOVA) whereas unstimulated (none = median 332 pg/ml, range 16-684) and anti-HLA-DP mAb treated control CBD PBMC (median 430 pg/ml, range 115-585, p > 0.05 versus none) produced constitutive levels of TNF-a (Fig. 3a). Anti-HLA-DP mAb inhibited beryllium-stimulated TNF-a production in a dose-dependent manner. For example, 30 mg/ml, 10 mg/ml and 1 mg/ml of anti-HLA-DP mAb resulted in a 47% (p = 0.0031), 53% (p = 0.0435) and 10% inhibition of Be-stimulated TNF-a production, respectively. Treatment with 30 mg/ml of either isotype control mAb or B7.21 mAb alone did not significantly increase TNF-a protein levels: median = 313 pg/ml (range 152-583) and median = 462 pg/ml (range 115-1342) respectively. CBD PBMC treated with 30 mg/ml of isotype mAb + 10 mM BeSO4 upregulated TNF-a protein production to a median of 838 pg/ml (range 368-4914, p < 0.05 versus none) as did control PBMC cultures treated with LPS (median = 9927 pg/ml, range 164525695, p < 0.05 versus none). We used RT-PCR to determine whether anti-HLA-DP B7.21 mAb could also inhibit TNF-a mature-mRNA production (Fig. 3b). We observed no alterations in the levels of constitutively expressed TNF-a pre-mRNA in the PBMCs of CBD patients under any experimental condition (data not shown). Be-exposure (10 mM BeSO4) upregulated the production of TNF-a mature-mRNA by CBD PBMC (median ratio 0.38, min. 0.05; max. 1.24, versus unstimulated control median ratio 0.09; min. 0.01. max. 0.82, p = 0.0428). In comparison, anti-HLA-DP mAb inhibited the increased production of Be-stimulated CBD PBMC TNF-a mature-mRNA in a dose-dependent manner. For example, 30 mg/ml, 10 mg/ml and 1 mg/ml of anti-HLA-DP mAb inhibited the production of TNF-a mature-mRNA by 56% (p < 0.05), 32% (p > 0.05) and 24% respectively. Treatment with 30 mg/ml of either isotype control mAb or anti-HLA-DP B7.21 mAb alone, did not significantly increase TNF-a mature-mRNA levels, and 30 mg/ml of isotype control mAb + 10 mM BeSO4 treatment did not significantly inhibit Be-stimulated increase in TNF-a mature-mRNA production in CBD PBMC. Northern analysis confirmed the ability of anti-HLA-DP B7.21 mAb to block the Be-induced increase in TNF-a mature-mRNA levels in CBD PBMCs (Fig. 3c). Be-stimulated TNF-a protein production by BAL cells from CBD patients. We tested the hypothesis that the Be-stimulated up-regulation of TNF-a protein production might be a unique property of CBD BAL cells. We determined the peak TNFa levels in the culture supernatants of unstimulated and Be-stimulated (100 mM BeSO4) CBD (n = 10), BeS (n = 10) and sarcoidosis control (n = 8) BAL cells (Fig. 4). Bestimulated CBD BAL cells significantly up-regulated the production of TNF-a protein to 1704 + 240 pg/ml (mean + SEM) in comparison to the unstimulated CBD BAL cell controls (133 + 90 pg/ml, p < 0.05, Wilcoxon Rank Sum Test). Be-stimulated BeS BAL cells produced 244 + 33 pg/ml of TNF-a compared to 32 + 22 pg/ml TNF-a for unstimulated BeS BAL cells (p < 0.05). However, Be-stimulated BeS BAL cell TNF-a levels were not significantly increased versus Be-stimulated CBD BAL cells or unstimulated BAL cells from CBD subjects. Following Be exposure of BAL cells from sarcoidosis patients, 435 + 301 pg/ml of TNF-a was produced which was similar to the amount of TNF-a produced by unstimulated sarcoidosis BAL cells (503 + 342 pg/ml). Be-stimulated TNF-a mRNA production by BeS and CBD BAL cells. Based on the CBD PBMC RT-PCR analysis in our first study, an increase in the levels of TNFa mature-mRNA was taken to indicate that TNF-a pre-mRNA had been spliced into mature-mRNA [18,19]. We determined the kinetics of TNF-a mRNA production in unstimulated and Be-stimulated BeS BAL (n = 18) and CBD BAL (n = 18) cells. The levels of TNF-a pre-mRNA and mature-mRNA following either exposure to medium alone or 100 mM BeSO4 remained unchanged at 0h, 6h and 24h in BeS BAL cells (Fig 5). In comparison, Be-stimulated CBD BAL cells significantly increased the levels of TNFa mature-mRNA at 6h and 24h, while their levels of TNF-a pre-mRNA remained relatively constant over the entire time course (Fig 5). As a positive experimental control, we compared the ability of 100 mM BeSO4 versus LPS (1 mg/ml E. coli serotype 0111.B4, Sigma) to up-regulate TNF-a maturemRNA levels in a subset of CBD (n = 11) and BeS (n = 5) BAL cells (Fig. 6a). After 24h, CBD BAL cells up-regulated the levels of TNF-a mature mRNA in response to Bestimulation (median ratio 0.60; min. 0.32; max. 1.40; p < 0.05 versus the unstimulated control = median ratio 0.25; min. 0.09; max. 0.78, Wilcoxon Rank Sum Test), and to LPS-stimulation (median ratio 0.73; min. 0.19; max. 0.1.88; p < 0.05 versus the unstimulated control). In comparison, after 24h, the levels of Be-stimulated BeS BAL cell TNF-a mature-mRNA (median ratio 0.25; min. 0.12; max. 0.76) were comparable to the levels present in the unstimulated controls (median ratio 0.16; min. 0.13; max. 0.63), but LPS-stimulated cells had significantly increased TNF-a mature-mRNA levels (median ratio 1.7; min. 0.40; max. 2.82; p > 0.05 versus the unstimulated control) (Fig. 6b). Not shown, both LPSand Be-stimulation up-regulated TNF-a mature-mRNA expression by 6h only in CBD BAL cells. To confirm the relevance of our observations in blood cells compared to cells from the target organ (e.g. lung), we used real-time PCR analysis to determine the levels of Be-stimulated (100 mM BeSO4) TNF-a mature-mRNA in a subset of our CBD BAL (n = 5) and CBD PBMC (n = 5) cell samples (Fig 7). Real-time PCR analysis generates a TNF-a mature-mRNA:b actin ratio in a different instrument scale from that generated by RT-PCR analysis. Hence the ratios expressed on the Y-axis of Fig 7, are not directly comparable, in other than a qualitative sense, to the Y-axis ratios in Fig. 3b, Fig 5, and Fig 6 where reverse transcription PCR analysis was used. After 24h, Be-stimulated CBD BAL cells (Fig 7a) significantly increased the levels of TNF-a mature-mRNA (median ratio 0.66; min. < 0.001; max. 0.06, p > 0.05, Wilcoxon Rank Sum Test) versus the unstimulated control (median ratio < 0.001; min. < 0.001; max. 0.06). Treatment with 30 mg/ml of anti-HLA-DP B7.21 mAb alone did not increase the levels of TNF-a maturemRNA (median ratio < 0.001; min. < 0.001; max. 0.11) whereas 30 mg/ml of ant-HLADP B7.21 mAb + 10 mM BeSO4 significantly blocked the Be-induced increase in TNF-a mature-mRNA levels (median ratio 0.01; min. < 0.01; max. p < 0.05 versus Bestimulated CBD BAL cells). Not shown, treatment with 30 mg/ml of isotype control mAb did not increase the levels of TNF-a mature-mRNA in comparison to the unstimulated control; and treatment with 30 mg/ml of isotype control mAb + 10 mM BeSO4 did not inhibit the Be-induced increase in TNF-a mature-mRNA levels. Virtually identical results were obtained by real-time PCR analysis of TNF-a mature-mRNA levels using CBD PBMC (Fig 7b). Discussion In human chronic granulomatous lung diseases such as sarcoidosis and hypersensitivity pneumonitis, the investigation of cytokine gene activation is hampered by lack of an etiologic agent. Beryllium, the etiologic agent of CBD, has been shown to up-regulate IL2, IFN-g and TNF-a gene expression [2-6], permitting us to study the effects of a specific antigen on specific cytokine gene and protein expression. In the present study, we show that anti-HLA-DP mAb, but not anti-HLA-DR mAb, inhibited antigen specific, Beinduced TNF-a expression by CD4 BAL T cells from CBD patients in a dose-dependent manner. Blocking HLA class II mediated beryllium-antigen recognition by these T cells also inhibits the up-regulation of TNF-a pre-mRNA splicing into TNF-a mature mRNA transcripts, and the production of TNF-a protein in both PBMC and BAL cells from CBD subjects. A recent study showed that a glutamic acid at residue 69 (Glu) in the HLADPB1 gene is a marker of beryllium sensitization and that HLA-DPB1 Glu homozygosity acts as a functional marker, associated with CBD severity [33]. Studies have also shown that HLA-DP is the predominant MHC class II molecule involved in the presentation of Be to either blood or BAL Be-specific CD4 T cells, with HLA-DR playing a minor role [1,11,12]. Fontenot et al. [2] showed that 30 mg/ml of anti-HLA-DP mAb inhibited 88% + 4% Be-stimulated IFN-g production by CD4 BAL T cells from CBD subjects, with much less inhibition of IFN-g expression seen with saturating amounts of anti-HLA-DR mAb. Based on the predominant role of HLA-DP in the development of beryllium hypersensitivity and its importance in T cell activation, proliferation and IFN-g protein expression, we focused our study on HLA-DP. Our data lend further support for the importance of beryllium-antigen signaling through the HLADP/Be-antigen/TCR complex in triggering T cell cytokine expression. Notably, we have identified the antigen-specific T cell as a principal source of TNF-a in a human granulomatous lung disorder. These data suggest an important link between innate and adoptive immune pathways in granulomatous disease. Fontenot et al. [11] and Lombardi et al. [12] showed that the proliferation of blood and BAL T cells from CBD subjects recognize Be in the context of HLA-DP. We therefore tested whether HLA-DP might also regulate TNF-a gene activation. The hypothesis that Be-induced TNF-a production is dominated by the activation of Bespecific CD4 T cells centered on the observation that LPS-stimulation up-regulated TNF-a protein production by both BeS BAL cells and CBD BAL cells, but only CBD BAL cells up-regulated TNF-a protein levels and the levels of TNF-a mature-mRNA in response to Be stimulation. LPS-stimulated TNF-a is most likely due to the activation of alveolar macrophages present in the mixed BAL cell population of both BeS and CBD subjects. Our data show that only the CBD BAL mixed cell compartment, not the BeS BAL cells or the sarcoidosis BAL cells, contains Be-specific CD4 T cells capable of upregulating TNF-a gene expression in response to Be exposure. Both unstimulated and Be-exposed BAL cells from sarcoidosis subjects produced high TNF-a protein levels. Bost et al. [39] noted that isolated CBD and sarcoidosis BAL macrophages, but not BeS BAL macrophages, expressed increased TNF-a mRNA transcripts suggesting that sarcoidosis BAL cells up-regulate TNF-a mRNA and protein production in vitro, but not in response to Be stimulation. The peripheral blood and BAL compartments are composed of two distinct mixed cell populations that contain cells capable of presenting the Be-antigen to Be-specific T cells via the HLA-DP molecule. It is important to acknowledge that due to the mixed cell nature of these compartments, other cells besides Be-specific CD4 T cells could be activated by Be exposure for the production of TNF-a. For example, Amicosante et al. [20] found that anti-HLA-DP treatment inhibited Be-stimulated CBD PBMC proliferation and IFN-g production, but not TNF-a production, suggesting that Bestimulated TNF-a gene activation might be regulated in mononuclear phagocytes, independent of Be-specific CD4 T cell activation. Sawyer et al. [34] and Hamada et al. [35] showed that H36.12j cells, a mouse hybrid macrophage cell line could up-regulate TNF-a protein production independent of transcription factor up-regulation. Galbraith et al. [36] found that Be-stimulated THP-1 cells, a human monocyte cell line, produce and accumulate TNF-a and IL-1b mRNA with the cells, but do not release exogenous cytokine protein. Thus, our study does not exclude other cellular sources of TNF-a, especially in the mixed BAL cell population. However, the results of our blocking experiments, plus the 89-fold increase in expression of TNF-a by CD3 CD4-expressing BAL T cells from CBD subjects, after BeSO4 exposure, compared to the 1.2-fold increase by alveolar macrophages strongly suggests that the CD4 T cell is the predominant cell responsible for the production of TNF-a following BeSO4 exposure in culture. The CBD subjects enrolled in this study showed the typical features of this disease, including the presence of non-caseating granulomas in their lungs, an increased BAL lymphocytosis and positive BeLPT proliferation responses in both their PBMC and BAL cell compartments [21,22]. In comparison, the BeS subjects had normal lung histology, normal BAL cell constituents and Be-induced proliferation that was limited to blood, not BAL. Thus, the major difference between subjects with BeS and CBD is the presence of Be-specific CD4 T cells and granulomatous inflammation in the lungs of CBD patients [2,21,22]. Although approximately 10% of BeS patients progress to CBD per year [37], the factors involved in the progression from sensitization to disease and the resultant migration of Be-specific T cells to the lung remain poorly characterized. Using both proliferation assays and intracellular cytokine staining, we have been unable to identify Be-induced lymphocyte proliferation or Th1-type cytokine production by BAL cells from BeS subjects [4] or normal, non-Be exposed control subjects [3,5]. Thus, within the lungs of BeS subjects, Be-specific CD4 T cells are either not present or are present in numbers below our level of detection. Cigarette smoke contains over 5000 chemicals [38], some of which have been shown to possess immunosuppressive properties and to inhibit antigen presenting cell activity for alveolar macrophages [31,32]. Thus, it is possible that cigarette smoke exposure hypothetically could have decreased the proliferative capacity of T cells in a small subset of smoking subjects in our study (3 of 38 CBD subjects and 4 of 21 BeS subjects). However, we found no significant association between smoking status and Beinduced thymidine uptake. To serve as a negative control, 8 sarcoidosis subjects were enrolled into this study and fifty percent of these subjects were active smokers. Despite continued smoke exposure, these subjects continued to exhibit a BAL lymphocytosis, suggesting an on-going sarcoidosis-antigen presenting activity, corresponding clonal expansion of sarcoidosis-antigen responsive lung T cells, and active pulmonary disease. In addition, these subjects did not have occupational Be exposure and their blood and BAL should therefore not contain Be-specific T cells. In addition to the activation of TCR on Be-specific CD4 T cells, we envisage several hypothetical mechanisms by which beryllium could also up-regulate TNF-a expression in CBD macrophages. Beryllium induces the HLA-DP-dependent expression of IFN-g by CD4 BAL T cells from CBD patients and the production of IFN-g mRNA and protein [2,3]. Beryllium-induced IFN-g serving as a macrophage activation factor could induce CBD macrophage TNF-a production in CBD BAL mixed cell cultures. It is also possible that Be could stimulate macrophage TNF-a production via activationinduced mRNA splicing [18]. Although the receptor for such activation is unknown, this might explain in part how beryllium up-regulates TNF-a in certain macrophage cell lines [34-36]. We speculate that ligation of Be-specific TCR by the HLA-DP-Be-antigen complex might signal TNF-a genes in CBD macrophages. Little is known about a possible role for HLA-DP in the up-regulation of cytokine synthesis. Support for this notion comes from studies using B cells as the APCs for T cell activation where MHC class II molecules can play a role in transducing signals resulting from the formation of the TCR-antigen-B cell receptor (BCR) complex [40-43]. In this model, class II MHCmediated signal transduction does not require that the MHC molecule have a cytoplasmic signaling domain [40] since antigen stimulation of resting B cells induces the translocation of immunoglobulin (Ig)-a/Ig-b heterodimers from the BCR to class II MHC molecules [44] thereby providing the domain for tyrosine kinase activation and subsequent signal transduction. No structures similar to the BCR-Ig-a/Ig-b complex have been identified in mononuclear phagocytes. Although unlikely that formation of the HLA-DP-Be-antigen-TCR complex would serve to directly signal the APC’s TNF-a gene via HLA-DP, it may be that subsequent ligation of co-stimulatory molecules on APCs, and T cells, could serve to up-regulate cytokine gene expression by both cells. Together, our findings suggest that Be stimulates four important HLA-DPmediated events, predominantly, but perhaps not exclusively, in CD4 CBD T cells. HLA-DP directs: 1) the accumulation of the Be-specific CD4 CBD T cells in the lung [2], 2) the proliferation of these CD4 T cells [2,11,12], 3) an increase in the levels of TNF-a mature-mRNA and 4) the up-regulation of TNF-a protein production. We envisage that in CBD, HLA-DP drives the clonal expansion of these Be-specific CD4 T cells into an effector-memory T cell population that regulates the synthesis and production of TNF-a through activation of their Be-specific TCRs [2]. Our study identifies HLA-DP, for the first time, as an essential regulatory component in the formation of Be-antigen stimulated TNF-a mRNA and protein in CBD. The current study does not address the molecular mechanism of how Be induces TNF-a gene activation. We hypothesize that Be acts at one of three regulation points 1) TNF-a gene transcription, 2) processing of the TNF-a pre-mRNA transcript, 3) stability of the TNF-a mature mRNA transcript. Should TNF-a mature-mRNA be stabilized, we would expect to see a linear accumulation of mature-mRNA over time and this was not observed. An increased transcription could yield higher levels of mature-mRNA which we did observe, suggesting that transcription inhibitors [45] should block the splicing of preinto mature mRNA and reduce the levels of these transcripts after Be exposure. If splicing rates alone were affected, the pre-mRNA levels would be expected to show a significant decrease, however, we show that pre-mRNA transcripts remain at comparably constant levels in both BeS and CBD BAL cells over time. It is possible that Be-stimulation could by-pass transcriptional regulation of the TNF-a gene and directly activates the splicing of TNF-a pre-mRNA in CD4 CBD BAL T cells [18]. Inhibitors of mRNA splicing [45] should reduce mature-mRNA transcript levels without altering the levels of pre-mRNA. While these studies, currently underway, will definitively show how TNF-a mRNA transcript levels are controlled in CBD, based on our present observations we hypothesize that transcriptional regulation of the TNF-a gene will emerge as a key regulatory point in the synthesis of Be-stimulated TNF-a in Be-specific CD4 T cells. AcknowledgmentsThe authors wish to thank Mary Solida, RN and Linda Staehler, RN for theirpatient care, Dr. Jay Westcott and Michelle Bausch for their technical assistance. Wethank Dr. John Cambier and his colleagues for their helpful suggestions. We are indebtedto those patients who make this, and other, beryllium-related research possible. References1. Sawyer, R.T., L.A. Maier, L.A. Kittle, and L.S. Newman. 2002.Chronic berylliumdisease: a model interaction between innate and acquired immunity. Internatl.Immunopharm. 2:249-261.2. Fontenot, A.P., S.J. Canavera, L. Gaharavi, L.S. Newman, and B.L. Kotzin. 2002.Target organ localization of memory CD4 T cells in patients with chronicberyllium disease. J. Clin. Invest. 110:1473-1482.3. Tinkle, S.S., L.A. Kittle, B.A. Schumacher, and L.S. Newman. 1997. Berylliuminduces IL-2 and IFN-g in berylliosis. J. Immunol. 158:518-526. 4. Tinkle, S.S., and L.S. Newman. 1997. Beryllium-stimulated release of tumornecrosis factor-a, interleukin-6, and their soluble receptors in chronic beryllium disease. Am. J. Respir. Crit. Care Med. 156:1884-1891.5. Maier, L.A., R.T. Sawyer, S.S. Tinkle, L.A. Kittle, E.A. Barker, R. Balkissoon,C.S. Rose, and L.S. Newman. 2001. IL-4 fails to regulate in vitro beryllium-induced cytokines in berylliosis. Eur. Respir. J. 17: 403-415.6. Maier, L.A., R.T. Sawyer, R.A. Bauer, L.A. Kittle, P. Lympany, D. McGrath, R.DuBois, E. Daniloff, C.S. Rose, and L.S. Newman. 2001. High beryllium-stimulated TNF-a is associated with the –308 TNF-a promoter polymorphism and with clinical severity in chronic beryllium disease. Am. J. Respir. Crit. CareMed. 164:1192-1199.7. Richeldi, L., R. Sorrentino, and C. Saltini. 1999. HLA-DPB1 glutamate 69: Agenetic marker of beryllium disease. Science 262:242-244. 8. Wang, Z., S.P. White, M. Petrovic, O.L. Tatum, L.S. Newman, L.A. Maier, andM.L. Marrone. 1999. Differential susceptibilities to chronic beryllium diseasecontributed by different glu69 HLA-DPB1 and –DPA1 alleles. J. Immunol.163:1647-1653.9. Rossman, M.D., J. Stubbs, C.W. Lee, E. Argyris, E. Magira, and D. Monos. 2002.Human leukocyte antigen class II amino acid epitopes. Susceptibility andprogression markers for beryllium hypersensitivity. Am. J. Respir. Crit. CareMed. 165:788-794.10. Fontenot, A.P., L.S. Newman, and B.L. Kotzin. 2001. Chronic beryllium disease:T cell recognition of a metal presented by HLA-DP. Clin. Immunol. 100:4-14.11. Fontenot, A.P., M. Torres, W.H. Marshall, L.S. Newman, and B.L. Kotzin. 2000.Beryllium presentation to CD4 T cells underlies disease-susceptibility HLA-DPalleles in chronic beryllium disease. Proc. Nat. Acad. Sci. USA. 97:12717-12722.12. Lombardi, G., C. Germain, J. Uren, R. Fiorillo, R.M. DuBois, W. Jones-Williams,C. Saltini, R. Sorrentino, and R. Lechler. 2001. HLA-DP allele-specific T cellresponses to beryllium account for DP-associated susceptibility to chronicberyllium disease. J. Immunol. 166:3549-3555.13. Saltini, C., K. Winestock, M. Kirby, P. Pinkston, and R.G. Crystal. 1989.Maintenance of alveolitis in patients with chronic beryllium disease by beryllium-specific helper T cells. N. Engl. J. Med. 320:1103-1109.14. Litton, M.J., B. Sander, M. Murphy, A. O’Garra, and J.S. Abrams. 1994. Earlyexpression of cytokines in lymph nodes after treatment in vivo withStaphylococcus enterotoxin B. J. Immunol. Meth. 175:47-58. 15. Alegre, M., P. Vandenabeele, V. Flamand, M. Moser, O. Leo, D. Abramowicz, J.Urbain, W. Fiers, and M. Goldman. 1990. Hypothermia and hypoglycemiainduced by anti-CD3 monoclonal antibody in mice: role of tumor necrosis factor.Eur. J. Immunol. 20:707-710.16. Ferran, C., K. Sheehan, M. Dy, R. Schreiber, S. Merite, P. Landais, L.H. Noel, G.Grau, J. Bluestone, and J.F. Bach. 1990. Cytokine-related syndrome followinginjection of anti-CD3 monoclonal antibody: Further evidence for transient in vivoT cell activation. Eur. J. Immunol. 20:509-515.17. Tracey, K.J., and A. Cerami. 1993. Tumor necrosis factor, other cytokines anddisease. Annu. Rev. Cell Biol. 9:317-343.18. Yang, Y., J.F. Chang, J.R. Parnes, and C.G. Fathman. 1998. T cell receptor (TCR)engagement leads to activation-induced splicing of tumor necrosis factor (TNF)nuclear pre-mRNA. J. Exp. Med. 188:246-254.19. Swanson, B.J., M. Murakami, T.C. Mitchell, J. Kappler, and P. Marrack. 2002.RANTES production by memory phenotype T cells is controlled by aposttranscriptional, TCR-dependent process. Immunity 17:605-615.20. Amicosante, M., F. Berretta, A. Franchi, P. Rogliani, C. Gotti, M. Losi, R. Dweik,and C. Saltini. 2002. HLA-DP-unrestricted TNF-a release by berylliumstimulated peripheral blood mononuclear cells. Eur. Respir. J. 20:1174-1178.21. Newman. L.S., K. Kreiss, T.R. King Jr, S. Seay, and P.A. Campbell. 1989.Pathologic and immunologic alterations in early stages of beryllium disease: Re-examination of disease definition and natural history. Am. Rev. Respir. Dis.139:1479-1486. 22. Mroz, M.M., K. Kreiss, D.C. Lezotte, P.A. Campbell, and L.S. Newman. 1991.Re-examination of the blood lymphocyte transformation test in the diagnosis ofchronic beryllium disease. J. Allergy Clin. Immunol. 88:54-60.23. Hunninghake, G.W., U. Costabel, M. Ando, R. Baughman, J.F. Cordier, R. DuBois, A. Eklund, M. Kitaichi, J. Lynch, G. Rizzato, C. Rose, R. Selroos, G.Semenzato, O.P. Sharma. 1999. Statement on sarcoidosis. Am. J. Respir. Crit.Care Med. 160:736-755.24. Ferris, B.G. 1978. Epidemiology standardization project. Am. Rev. Respir. Dis.118:1-120.25. Watters, L.C., M.I. Schwartz, and R.M. Cherniak. 1997. Idiopathic pulmonaryfibrosis: Pretreatment bronchoalveolar lavage cellular constituents and theirrelationship with lung histopathology and clinical response to therapy. Am. Rev.Respir. Dis. 135:696-704.26. Watson, A.J., R. DeMars, I.S. Trowbridge, and F.H. Bach. 1983. Detection of anovel human class II HLA antigen. Nature 304:358-361.27. Waldrop, S.L., C.J. Pitcher, D.N. Peterson, V.C. Maino, and J.L. Picker. 1997.Determination of antigen-specific memory/effector CD4 T cell frequencies byflow cytometery; evidence for a novel, antigen-specific homeostatic mechanismin HIV-associated immunodeficiency. J. Clin. Invest. 99:1739-1750.28. Pitcher, C.J., C. Quittner, D.M. Peterson, M. Connors, R.A. Koup, V.C. Maino,and L.J. Picker. 1999. HIV-1-specivic CD4 T cells are detectable in mostindividuals with active HIV-1 infection, but decline with prolonged viralsuppression. Nat. Med. 5:518-525. 29. Wahlstrom, J., K. Katchar, H. Wigzell, O. Olerup, A. Eklund, J. Grunewald.2001. Analysis of intracellular cytokines in CD4 and CD8 lung and blood Tcells in sarcoidosis. Am. J. Respir. Crit. Care Med. 163:115-121.30. Kreiss, K., M.M. Mroz, B. Zhen, J.W. Martyny, and L.S. Newman. 1993.Epidemiology of beryllium sensitization and disease in nuclear workers. Am.Rev. Respir. Dis. 148:985-991.31. Hance A.J., F. basset, G. Saumon, C. Danel, D. Valeyre, J.P. Battesti, J. Chretien,and R. Georges. 2003. Smoking and interstitial lung disease. The effect ofcigarette smoking on the incidence of pulmonary histiocytosis X and sarcoidosis.Annl. N.Y. Acad. Sci. 465:643-656.32. Murin, S., K.S. Bilello, and R. Matthay. 2000. Other smoking-affected pulmonarydiseases. Clinics in Chest. Med. 21:121-137ix.33. Maier, L.A., D.S. McGrath, H. Sato, P. Lympany, K. Walsh, R. duBois, L.Silveira, A.P. Fontenot, R.T. Sawyer, E. Wilcox, and L.S. Newman. 2003.Influence of MHC class II susceptibility to beryllium sensitization and chronicberyllium disease. J. Immunol. 171:6910-6918.34. Sawyer R.T., H. Hamada, L.A. Kittle, and L.S. Newman. 2000. Beryllium-stimulated tumor necrosis factor-a production by a mouse hybrid macrophage cellline. Toxicology 143:235-247.35. Hamada H., R.T. Sawyer, L.A. Kittle, and L.S. Newman. 2000. Berylliumstimulation does not activate transcription factors in a mouse hybrid macrophagecell line. Toxicology 143:249-261. 36. Galbraith, G.M., J.P. Pandey, M.G. Schmidt, P. Arnaud, and J.M. Goust. 1996.Tumor necrosis factor alpha gene expression in human monocytic THP-1 cellsexposed to beryllium. Arch. Environ. Health. 51:29-33.37. Kreiss, K., M.M. Mroz, B. Zhen, H. Wiedmann, and B. Barna. 1997. Risks ofberyllium disease related to work processes at a metal, alloy, and oxideproduction plant. Occup. Environ. Med. 54:605-612.38. Stedman, R.L. 1968. The chemical composition of tobacco and tobacco smoke.Chem. Rev. 68:153-207.39. Bost, T.W., D.W.G. Riches, B. Schumacher, P.C. Carre, T.Z. Kahn, J.A.Martinez, and L.S. Newman. 1994. Alveolar macrophages from patients withberyllium disease and sarcoidosis express increased levels of mRNA for tumornecrosis factor-a and interleukin-6 but not interleukin 1b. Am. J. Respir. Cell. Mol. Biol. 10:506-513.40. Mooney, N.A., C. Grillot-Courvalin, C. Hivroz, L.Y. Ju, and D. Charron. 1990.Early biochemical events after MHC class II-mediated signaling on human Blymphocytes. J. Immunol. 145:2070-2076.41. Lane, P.J.L., F.M. McConnell, G.L. Schieven, E.A. Clark, and J.A. Ledbetter.1990. The role of class II molecules in human B cell activation. J. Immunol.144:3684-3692.42. Cambier, J.C., D.C. Morrison, M.M. Chien, and K.R. Lehmann. 1991. Modelingof T cell contact-dependent B cell activation. J. Immunol. 146:2075-2082. 43. Andre, P., J.C. Cambier, T.K. Wade, T. Raetz, and F.W. Wade. 1994. Distinctstructural compartmentalization of the signal transduction functions of majorhistocompatibility complex class II (Ia) molecules. J. Exp. Med. 179:763-768.44. Lang, P., J.C. Stolpa, B.A. Freibert, F. Crawford, J. Kappler, A. Kupfer, and J.C.Cambier. 2001. TCR-induced transmembrane signaling by peptide/MHC class IIvia associated Ig-a/b dimmers. Science 291:1537-1540. 45. Jarrous, N., F. Osman, and R. Kaempfer. 1996. 2-Aminopurine selectivelyinhibits splicing of tumor necrosis factor alpha mRNA. Mol. Cell. Biol. 16:2814-2822. Table 1. Clinical Characteristics of the Study SubjectsaBeS(n = 21)CBD(n = 38)Sarcoidosis(n = 8)Age (years)54 ± 8.657 ± 244 ± 6.6Gender (M/F)16/533/58/0Smoking Status (CS/FS/NS)b4/7/93/16/194/1/3Race (C/AF/H)c8/1/231/1/34/1/1Corticosteroid Treatment(Yes/No)3 (inhaled)/18 5 (oral) 2(inhaled)/311 (oral)/7 BAL CellsWBC Count (x104)22.8 ± 3 34 ± 4d 50 ± 15dMacrophages (%)87 ± 265 ± 472 ± 8Lymphocytes (%)11 ± 233 ± 4d26 ± 7dPeak Blood BeLPT StimulationIndex (SI)15 ± 3 d 43 ± 10d 1.4 ± 0.2 Peak BAL BeLPT StimulationIndex (SI)2.4 ± 0.3 64 ± 17d 1.2 ± 0.05 aData are expressed as mean ± SEM.bC = Caucasian; AF = African-American; H = Hispanic.cCS = Current smoker; FS = Former smoker; NS = Never smoker.dp < 0.05 List of FiguresFigure 1. (a) Intracellular expression of TNF-a after stimulation of CBD BALCD4 T cells. Staining patterns are shown for the BAL cells from CBD patient 1 aftershort-term culture with medium alone or 10 mMBeSO4. The percentage of CD4 CBDBAL T cells expressing intracellular TNF-a is shown in the upper right quadrant of each density plot. The second and third sets of panels show that the expression ofBeSO4-induced TNF-a is inhibited in CD4 CBD BAL T cells by treatment with increasing amounts of anti-HLA-DP mAb, but not increasing amounts of anti-HLA-DR mAb. (b)Beryllium-stimulated TNF-a expression by CBD BAL cells (n = 5) is inhibited (% inhibition) by increasing amounts of anti-HLA-DP (B7.21) mAb but not anti-HLA-DR(LB3.1) mAb.Figure 2. The peak HTdR incorporation (mean + SEM counts per minute) forCBD PBMC (n = 8) that were unstimulated (none) or 10 mM BeSO4-stimulated in theabsence and presence of increasing amounts of anti-HLA-DP (B7.21) mAb +BeSO4. (* p< 0.05, contrasts from repeated measures ANOVA, for BeSO4 versus the unstimulatedcontrol, and in the presence of B7.21 mAb versus the BeSO4 stimulated cells.Figure 3. (a) The levels of TNF-a (pg/ml)(n = 7) and (b), the levels of TNF-amature-mRNA (TNF-a mature-mRNA : b Actin mRNA, n = 12) in CBD PBMC thatwere unstimulated (none) or 10 mMBeSO4-stimulated in the absence or presence ofincreasing amounts of anti-HLA-DP mAb (B7.21). Shown on each graph, the lower andupper bars are the minimum and maximum values, the lower and upper box bars are theinterquartile range (IQR) at the twenty-fifth and seventy-fifth (25%, 75%) percentiles, thecenter bar in the box is the median and each dot represents the result for the individual patient (for graphs (a) and (b); * p < 0.05 versus the BeSO4-stimulated group, contrastsfor repeated measures ANOVA). (c) A representative northern blot of total TNF-a mature-mRNA isolated from the CBD PBMC of a single subject that were unstimulatedor 10 mMBeSO4-stimulated in the absence and presence of increasing amounts of B7.21anti-HLA-DP mAb. Controls were stimulated with 30 mg/ml B7.21 mAb, 30 mg/mlMOPC 21 isotype control mAb, or with 30 mg/ml MOPC 21 isotype control mAb + 10mMBeSO4.Figure 4. TNF-a production (pg/ml) by unstimulated (none) or 100 mM BeSO4-stimulated BAL cells from a subset of (a) CBD (n = 10) (b) BeS (n = 19) or (c)sarcoidosis (n = 8) study subjects. * p < 0.05; the difference between Be-stimulated andunstimulated (none) TNF-a levels were compared using Wilcoxon Rank Sum test for each group.Figure 5. Reverse transcription PCR analysis of the ratio of the 380 base pairTNF-a mature-mRNA PCR product to the 800 base pair b actin mRNA PCR product andthe ratio of the 90 base pair TNF-a pre-mRNA product:b-Actin mRNA fromunstimulated (black box) and 100 mMBeSO4-stimulated (open diamond) BeS BAL (n =18) and CBD BAL (n = 18) cells at 0h, 6h and 24h. * p < 0.05, doubly repeated measuresmodel.Figure 6. Reverse transcription PCR analysis of the ratio of the 380 base pairTNF-a mature-mRNA PCR product to the 800 base pair b-Actin mRNA PCR productfrom unstimulated (none), 100 mMBeSO4-stimulated and 1 mg/ml LPS-stimulated (a)CBD BAL (n = 5) and (b) BeS BAL (n = 5) cells at 24 hr. * p < 0.05 for Be-stimulated CBD BAL cells versus the unstimulated control (none), and for LPS-stimulated BeS andCBD BAL cells versus the unstimulated control (none), Wilcoxon Rank Sum Test.Figure 7. Real time PCR analysis of the ratio of the 161 base pair TNF-a mature-mRNA PCR product to the 200 base pair b-Actin mRNA PCR product at 24 hr fromunstimulated (none), 10 mMBeSO4-stimulated, 30 mg/ml anti-HLA-DP B7.21 mAbstimulated and 30 mg/ml anti-HLA-DP B7.21 + 10 mM BeSO4-stimulated (a) CBD BALcells (n = 6) and (b) CBD PBMC (n = 5). * p < 0.05 versus the correspondingunstimulated control, Wilcoxon Rank Sum Test. Medium AloneBeSO4a CD4TNF-aCD4TNF-a BeSO4 +30 mg/ml a-DPBeSO4 +10 mg/ml a-DPBeSO4 +1 mg/ml a-DP CD4TNF-aBeSO4 +30 mg/ml a-DRBeSO4 +10 mg/ml a-DRBeSO4 +1 mg/ml a-DR1.1%0.06%7.5% 7.2%8.1%8.9%2.1%0.7% 020406080100 0102030anti-HLA DP mAb anti-mAb (mg/ml)% Inhibitionanti-HLA DR mAbb 050010001500200025003000 BeSO4NONE 10 mM 10 mM 10 mM 10 mMB7.21 mAb NONE NONE 1 mg/ml 10 mg/ml 30 mg/mlp = 0.006 p = 0.0003p = 0.0088p = 0.847Sawyer Fig. 2