Abstract
Chronic beryllium disease (CBD) is a hypersensitivity granulomatosis characterized by beryllium hypersensitivity (BH) and mediated by CD4+ T cells. However, all individuals with BH may not develop CBD. To examine the role of the three different human leukocyte antigen (HLA) Class II isotypes in BH with (CBD) and without clinical disease (BHWCD), we performed DNA-based typing of HLA-DPB1, HLA-DQB1, and HLA-DRB1 loci on 55 subjects with BH (25 with established CBD and 30 with BHWCD), and compared this with the results for 82 beryllium-exposed workers with no evidence of BH. The allele distribution was utilized to identify candidate amino acid epitopes that differed between the study groups. HLA-DPB1-E69 was the most important marker for BH, and did not differentiate BHWCD from CBD. A significant association with CBD was observed with HLA-DQB1-G86 (pcorr < 0.04), and HLA-DRB1-S11 was significantly increased in CBD as compared with BHWCD (p < 0.03). These observations suggest that HLA-DPB1-E69 is a marker for susceptibility to BH and not just a progression marker for CBD. In addition, HLA amino acid epitopes on HLA-DRB1 and -DQB1, in concert with or independently of HLA-DPB1-E69, may be associated with progression to CBD.
Keywords: human; lung; human leukocyte antigen molecules; genetics; granuloma
Chronic beryllium disease (CBD) is a hypersensitivity granulomatous disease that predominately affects the lungs (1). In the United States, more than a half-million individuals have been exposed to beryllium and are currently at risk for CBD. The diagnosis of CBD is confirmed by the presence of granulomatous inflammation in the lungs and a positive beryllium-specific proliferative response of bronchoalveolar lymphocytes (2). Although this disorder was thought in the 1960s to have been almost eliminated through the use of environmental controls in the beryllium industry (3), the development of an immunologic test for beryllium hypersensitivity (BH) has allowed screening of individuals for BH (4). Surveys of current and retired workers (4-8) have identified from 1 to 8% with BH. From 42 to 83% of workers who have BH also have evidence of CBD. It is also becoming clear that some individuals without evidence of clinical disease (beryllium hypersensitivity without clinical disease [BHWCD]) may never develop CBD (9). However, this concept still needs to be proven with long-term follow-up studies.
CD4+ T cells are thought to mediate the hypersensitivity to beryllium because of their accumulation at the site of disease activity (1, 10) and their ability to respond to beryllium in vitro (10, 11). This concept was further supported by the observation that the proliferative response of lymphyocytes to beryllium could be inhibited by antibodies to human leukocyte antigen (HLA) Class II molecules but not to HLA Class I molecules (10). Since only a small proportion of beryllium-exposed workers ever developed CBD (3), HLA Class II molecules were proposed as a marker for CBD (12). Molecular typing demonstrated an association of the disease with HLA-DPB1*0201, and E69 on HLA-DPB1 was associated with more than 90% of cases of CBD. The importance of HLA-DPB1-E69 was recently confirmed in additional studies (13, 14). The strong association of E69 with CBD has suggested to some that E69 screening might be a useful test for preemployment determination of susceptibility to CBD (14). The functional importance of these observations has recently been demonstrated by the ability of an anti–HLA-DP monoclonal antibody (mAb) to block the proliferative response of Be-stimulated T cell lines and clones derived from patients with CBD (15, 16). In preliminary experiments, we reproduced these findings with freshly isolated cells, taken not only from patients with CBD, but also from subjects with BHWCD.
In the present study, we investigated the role of HLA Class II molecules not only in patients with CBD, but also in individuals with BHWCD. HLA-DPB1, -DQB1, and -DRB1 alleles were determined with DNA-based methodologies (sequence-specific oligonucleotide probe analysis [SSOP], sequence specific primers–polymerase chain reaction [PCR] [SSP-PCR], and sequence based typing) to correlate the presence of specific HLA alleles and amino acid epitopes with BH, BHWCD, and CBD.
Subjects were referred to the Hospital of the University of Pennsylvania (HUP) for possible CBD. Individuals were considered to have BH if they had a positive blood lymphocyte proliferative test (LPT) to beryllium on at least two occasions or a positive bronchoalveolar cell LPT to beryllium (1). CBD was diagnosed when there was evidence of BH with granulomas on biopsy and/or radiologic changes consistent with a granulomatous process. Control subjects included those evaluated for CBD who did not have BH, or beryllium workers who had LPTs at HUP. Only individuals with negative LPTs were accepted as controls. Ten of the 82 controls had abnormal chest radiographs and had a clinical evaluation at the University of Pennsylvania. All had bronchoscopy and bronchoalveolar lavage (BAL) with negative blood and BAL beryllium-stimulated LPTs. The study protocol was approved by the Committee on Studies Involving Human Beings at the University of Pennsylvania.
LPT were performed according to previously reported procedures (1). The cells were cultured at a concentration of 2.5 × 10 5 cells/well. Stimulants included phytohemagglutinin (PHA) (L-9132; Sigma Chemical Co., St. Louis, MO), concanavalin A (C5257; Sigma), candida (M15; Greer Labs, Lenoir, NC), beryllium sulfate (100 μM, 10 μM, or 1 μM), and beryllium fluoride (100 μM, 10 μM, or 1 μM). The cells were pulsed with 3H-thymidine and harvested, and the uptake of 3H-thymidine was measured as counts per million (cpm). Results were expressed as a stimulation index (SI = mean cpm of test/mean cpm of controls). Bronchoalveolar cells were cultured at a final concentration of 1 × 105 cells/well and harvested after 3 to 5 d, with SI results expressed as just described. A positive test required that a positive response be recorded on two different days or to two different concentrations of beryllium. A positive response was defined as an SI > 3.0 for blood or an SI > 5.0 for BAL cells (1).
DNA was extracted as previously described (17). Locus- and allele-specific amplifications of genomic DNA were performed for HLA-DRB1–, HLA-DQB1– and HLA-DPB1–associated alleles (17). A 270-bp amplified DNA product was verified by electrophoresis in a 2% agarose gel containing ethidium bromide, and was visualized under UV light. Hybridization was performed as described elsewhere (18, 19), using a panel of SSOPs (20).
SSP–PCR, using the Amplification Refractory Mutation System (European patent No. 0332435, U.S. patent No. 5595890; under license from Zeneca Limited, Wilmington, DE) technology, was utilized for high-resolution HLA-DRB1 and HLA-DPB1 typing. Appropriate kits (Olerup SSP; Genovision, West Chester, PA) were chosen for subtyping (high-resolution typing) of the HLA-DRB1 locus on the basis of the low-resolution results of HLA-DRB1 typing.
When a specific allele could not be determined by SSOP or SSP–PCR, it was sequenced. The DNA was amplified, using SSOP primers (17), and was purified and cloned with the TA Cloning Kit (Invitrogen, CA). DNA was isolated with the Qiagen plasma mini-kit (Valencia, CA). Sequences were analyzed in both forward and backward directions at the University of Pennsylvania sequencing facility.
Nonpaired t tests were used to compare differences in SI. The frequencies of specific alleles in the patient and control populations were determined and standardized residuals (post hoc cell contributions) were calculated. To determine the significance of a specific amino acid, a chi square value was determined and corrected for multiple comparisons by multiplying the p value by the number of amino acid epitopes tested for each specific Class II isotype, with the product expressed as pcorr. The odds ratio (OR) for subjects versus controls was determined as OR = (Pd [1-Pc])/ ([1-Pd]Pc), where Pd and Pc are the frequencies of individuals positive for a specific amino acid among subjects and controls, respectively.
Fifty-five unrelated subjects with BH and 82 controls who were beryllium-exposed but without BH were typed for HLA Class II genes (HLA-DRB1, -DQB1, and -DPB1) by amplification of genomic DNA through PCR and hybridization with SSOP. Three of the 55 subjects with BH were black and the rest were white (94%). The control population had a similar racial distribution. Eight control subjects had bronchoscopy with lavage and biopsy to rule out beryllium disease.
The 55 subjects with BH were classified as follows: Thirty had evidence of BHWCD. Chest radiographs and pulmonary function tests were normal in all but one, who had evidence of emphysema on chest radiography and of obstruction in pulmonary function tests. Twenty-nine of the 30 subjects with BHWCD underwent high-resolution chest computed tomography (CT). Thirteen had normal studies, and the other 16 had evidence of minimal focal disease consisting of emphysema, bronchiectasis, or small scars. BAL was performed on 25 of these subjects. Beryllium proliferation testing of the BAL cells was negative in 17 subjects and positive in eight. Seven of the eight subjects with positive beryllium proliferation studies of BAL cells had transbronchial biopsies performed. These seven biopsies were normal or showed minimal, nonspecific inflammation. Follow-up chest radiography and pulmonary function tests were performed on six of the eight subjects with positive beryllium proliferation studies of BAL cells. After 4 to 7 yr of follow-up, none of these individuals has developed any chest- radiographic indications of disease or declines in pulmonary function. The one individual with a positive beryllium proliferation test on BAL cells and no biopsy has been followed for 7 yr without any significant physiologic or radiologic changes. The five individuals with BHWCD who did not consent to a bronchoscopy (one was 87 yr old when first seen) have been followed for at least 5 yr without evidence of clinical disease.
Twenty-five subjects had CBD based on evidence of BH and granulomatous disease. In every case, a markedly positive BAL cell proliferative response to beryllium was present. Transbronchial biopsies were positive for typical noncaseating granuloma in 23 of these subjects. The two cases with negative transbronchial biopsy results had typical chest radiographs and high-resolution chest CT scans showing upper-lobe nodular interstitial changes, and restrictive pulmonary function tests consistent with a granulomatous process. Twenty of 25 subjects had abnormal chest radiographs, and 17 of the 25 subjects had restrictive pulmonary function test results.
The results of these subjects' blood and lung LPT are given in Table 1. None of the subjects was receiving corticosteroids at the time of proliferation testing. The LPT for which the results are reported were performed on the day that the subject's blood was saved for HLA testing. Five patients with BHWCD and five patients with CBD had negative blood beryllium proliferation studies on the day on which test results were recorded. All patients with CBD had positive BAL cell proliferation studies, and eight of 25 subjects with BHWCD had positive beryllium proliferation studies of their BAL cells. None of the control samples gave positive results. Although there was no significant difference in the blood beryllium LPT responses between the subjects with BHWCD and those with CBD (p > 0.5), there was a significant difference in the BAL cell beryllium proliferative response between the subjects with BHWCD and the subjects with CBD (p < 0.001).
| Controls | Beryllium Hypersensitivity (all subjects) | Beryllium Hypersensitivity Without Disease | Chronic Beryllium Disease | |||||
|---|---|---|---|---|---|---|---|---|
| Blood | 1.5 ± 0.08 | 15.3 ± 2.5 | 16.8 ± 3.9 | 13.6 ± 3.1* | ||||
| (n = 82) | (n = 55) | (n = 30) | (n = 25) | |||||
| Lung | 1.7 ± 1.2 | 53.0 ± 10.8 | 16.4 ± 7.2 | 89.6 ± 17.7† | ||||
| (n = 8) | (n = 50) | (n = 25) | (n = 25) |
The different HLA-DPB1* alleles identified in the controls and in the BH, BHWCD, and CBD groups are shown in Table 2. Because of uncertainties in the identification of specific alleles, 11 subjects had 21 HLA-DPB1 alleles cloned and sequenced. Post hoc cell contributions were determined to identify alleles that might differ among subjects with BH, BHWCD, or CBD and controls. HLA-DPB1*0101 was decreased in BH and CBD, and HLA-DPB1*0401 was decreased in BH and BHWCD. HLA-DPB1*0402 was increased in CBD, HLA-DPB1*0601 was increased in BH and BHWCD, and HLA-DPB1*1301 was increased in BHWCD. The sequences of the three alleles associated with BH, BHWCD, or CBD were analyzed for polymorphic amino acid epitopes (Table 3) (21). Twenty-three unique epitopes were identified. We determined the frequencies of these epitopes in controls and in subjects with BH, BHWCD, and CBD (Table 4). Three epitopes were significantly different (Bonferroni's corrected n = 23) from those of controls. HLA-DPB1-E69 was significantly associated with BH and with the BHWCD and CBD subgroups. HLA-DPB1-L11 was significantly associated with BH and BHWCD but was not significantly associated with CBD. In contrast, HLA-DPB1-D 55 was significantly associated with BH and had a borderline association with CBD but not with BHWCD. However, when the association of HLA-DPB1-L11 or –D55 was corrected for the association with HLA-DPB1-E69, only the HLA-DPB1 alleles that contained L11 and E69 or D55 and E69 were significantly associated with BH. Among the HLA-DPB1-L11E69-containing alleles and the HLA-DPB1-D55E69-containing alleles, there was no significant difference between BHWCD and CBD (data not shown). Thus, the DPB1-E69 epitope appears to be the most important HLA-DPB1 amino acid epitope for the development of BH and does not predict whether someone with BH will be asymptomatic (BHWCD) or symptomatic (CBD).
| Alleles | Controls (n = 82) | BH (n = 55) | BH No Disease (n = 30) | CBD (n = 25) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n* | % | n* | % | n* | % | n* | % | |||||||||
| DPB1*0101 | 15 | 9.1 | 1 † | 0.9 | 1 | 1.7 | 0 † | 0.0 | ||||||||
| DPB1*0201 | 26 | 15.6 | 24 | 21.8 | 15 | 25.0 | 9 | 18.0 | ||||||||
| DPB1*0202 | 2 | 1.2 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*0301 | 11 | 6.7 | 10 | 9.1 | 6 | 10.0 | 4 | 8.0 | ||||||||
| DPB1*0401 | 58 | 35.4 | 23 † | 20.9 | 11 † | 18.3 | 12 | 24.0 | ||||||||
| DPB1*0402 | 11 | 6.7 | 11 | 10.0 | 3 | 5.0 | 8 † | 16.0 | ||||||||
| DPB1*0501 | 4 | 2.4 | 4 | 3.6 | 3 | 5.0 | 1 | 2.0 | ||||||||
| DPB1*0601 | 2 | 1.2 | 8 † | 7.2 | 6 † | 10.0 | 2 | 4.0 | ||||||||
| DPB1*0801 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*0901 | 0 | 0.0 | 1 | 0.9 | 1 | 1.7 | 0 | 0.0 | ||||||||
| DPB1*1001 | 4 | 2.4 | 5 | 4.5 | 2 | 3.3 | 3 | 6.0 | ||||||||
| DPB1*1101 | 4 | 2.4 | 4 | 3.6 | 3 | 5.0 | 1 | 2.0 | ||||||||
| DPB1*1301 | 5 | 3.0 | 8 | 7.2 | 6 † | 10.0 | 2 | 4.0 | ||||||||
| DPB1*1401 | 2 | 1.2 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*1501 | 2 | 1.2 | 1 | 0.9 | 1 | 1.7 | 0 | 0.0 | ||||||||
| DPB1*1601 | 3 | 1.8 | 2 | 1.8 | 1 | 1.7 | 1 | 2.0 | ||||||||
| DPB1*1701 | 2 | 1.2 | 2 | 1.8 | 0 | 0.0 | 2 | 4.0 | ||||||||
| DPB1*1801 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DPB1*1901 | 1 | 0.6 | 2 | 1.8 | 1 | 1.7 | 1 | 2.0 | ||||||||
| DPB1*2001 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*2301 | 6 | 3.7 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*2601 | 1 | 0.6 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DPB1*3301 | 2 | 1.2 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*3901 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DPB1*4001 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DPB1*4801 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| Allele | 8 | 9 | 11 | 35 | 36 | 55 | 56 | 57 | 65 | 69 | 76 | 84 | 85 | 86 | 87 | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DPB1*0402 | L | F | G | F | V | D | E | E | I | K | M | G | G | P | M | |||||||||||||||
| DPB1*0601 | V | Y | L | F | V | D | E | D | L | E | M | D | E | A | V | |||||||||||||||
| DRB1*1301 | V | Y | L | Y | A | A | A | E | I | E | I | D | E | A | V |
| HLA Epitope | Controls | BH | BHWCD | CBD | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| % | pcorr | OR | % | pcorr | OR | % | pcorr | OR | ||||||||||||
| DPB1-L11 | 31.7% | 60% | p < 0.02 | OR = 3.2 | 66.7% | p < 0.02 | OR = 4.3 | 52% | p > 0.1 | |||||||||||
| (1.6–6.6) | (1.8–10.5) | |||||||||||||||||||
| DPB1-D55 | 57.3% | 83.6% | p < 0.03 | OR = 3.8 | 80% | p > 0.1 | 88% | p > 0.1* | ||||||||||||
| (1.5–8.8) | ||||||||||||||||||||
| DPB1-E69 | 47.6% | 87.3% | p < 0.002 | OR = 7.6 | 90% | p < 0.002 | OR = 9.9 | 84% | p < 0.03 | OR = 5.8 | ||||||||||
| (3.1–18.7) | (2.8–35.3) | (1.8–18.4) | ||||||||||||||||||
| DQB1-G86 | 6.1% | 18.2% | p > 0.1 | 10% | p > 0.4 | 28% | p < 0.04 | OR = 6.0 | ||||||||||||
| (1.7–21.1) | ||||||||||||||||||||
| DRB1-S11 | 59.8% | 72.7% | p > 0.2 | 60.0% | p >0.8 | 88% | p > 0.1* | |||||||||||||
| DRB1-N37 | 35.4% | 54.5% | p > 0.1 | 46.7% | p > 0.3 | 64% | p > 0.1* | |||||||||||||
| DRB1-E71 | 15.9% | 30.9% | p > 0.1 | 20.0% | p > 0.6 | 44% | p > 0.1* | |||||||||||||
The different HLA-DQB1 alleles identified in the controls and BH, BHWCD, and CBD groups are shown in Table 5. When HLA typing was initiated on these subjects, the HLA-DQB1 alleles *0202 and *0203 were not known. Since there was insufficient DNA to differentiate HLA-DQB1*0201 from HLA-DQB1*0202 or HLA-DQB1*0203, these alleles are expressed as HLA-DQB1*02xx in Table 5. Post hoc cell contributions were again determined to identify the alleles that most likely accounted for the possible differences between controls and subjects with BH, BHWCD, and CBD. HLA-DQB1*0605 was associated with BH and CBD. HLA-DQB1*0604 was associated with CBD, and HLA-DQB1*0608 was associated with BHWCD. We compared the sequences of the three alleles associated with BH, BHWCD, or CBD. The amino acid-polymorphic epitopes from these alleles are illustrated in Table 6. Sixteen different epitopes were identified. The frequencies of these epitopes in controls and in subjects with BH, BHWCD, and CBD were determined (Table 4). Only one amino acid epitope, HLA-DQB1-G86, was significantly (p < 0.04, Bonferroni's corrected n = 16, as compared with controls) associated with CBD. To determine whether this amino acid epitope could distinguish between BHWCD and CBD, we compared the frequency of this amino acid epitope in subjects with BHWCD and CBD. HLA-DQB1-G86 was of borderline significance (p = 0.088). Thus, in individuals with BH, HLA-DQB1-G86 may be a marker for progression to disease (CBD).
| Alleles | Controls (n = 82) | BH (n = 55) | BHWCD (n = 30) | CBD (n = 25) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n* | % | n* | % | n* | % | n* | % | |||||||||
| DQB1*02xx | 37 | 22.6 | 23 | 20.9 | 15 | 25.0 | 8 | 16.0 | ||||||||
| DQB1*0301 | 32 | 19.5 | 19 | 17.3 | 8 | 13.3 | 11 | 22.0 | ||||||||
| DQB1*0302 | 17 | 10.4 | 9 | 8.2 | 7 | 11.7 | 2 | 4.0 | ||||||||
| DQB1*0303 | 9 | 5.5 | 9 | 8.2 | 8 | 13.3 | 1 | 2.0 | ||||||||
| DQB1*0401 | 6 | 3.7 | 2 | 1.8 | 0 | 0.0 | 2 | 4.0 | ||||||||
| DQB1*0402 | 1 | 0.6 | 4 | 3.6 | 2 | 3.3 | 2 | 4.0 | ||||||||
| DQB1*0501 | 19 | 11.6 | 10 | 9.1 | 4 | 6.7 | 6 | 12.0 | ||||||||
| DQB1*0502 | 2 | 1.2 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DQB1*0503 | 3 | 1.8 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DQB1*0504 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DQB1*0601 | 3 | 1.8 | 3 | 2.7 | 1 | 1.7 | 2 | 4.0 | ||||||||
| DQB1*0602 | 21 | 12.8 | 9 | 8.2 | 6 | 10.0 | 3 | 6.0 | ||||||||
| DQB1*0603 | 7 | 4.3 | 9 | 8.2 | 4 | 6.7 | 5 | 10.0 | ||||||||
| DQB1*0604 | 5 | 3.0 | 7 | 6.4 | 1 | 1.7 | 6 † | 12.0 | ||||||||
| DQB1*0605 | 0 | 0.0 | 3 † | 2.7 | 2 † | 3.3 | 1 | 2.0 | ||||||||
| DQB1*0607 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DQB1*0608 | 0 | 0.0 | 2 | 1.8 | 2 † | 3.3 | 0 | 0.0 | ||||||||
The different HLA-DRB1 alleles identified in controls and in subjects with BH, BHWCD, and CBD are shown in Table 7. Because of uncertainties in the identification of specific alleles, 10 subjects had 12 HLA-DRB1 alleles cloned and sequenced. In only one case was insufficient DNA available for sequencing (because the patient had died), preventing the designation of the HLA-DRB1 allele as HLA-DRB1*1301 versus HLA-DRB1*1306 in the presence of HLA-DRB1*1302. In this case, it was elected to include the patient's results and record the allele as HLA-DRB1*1301/6 in Table 7. Post hoc cell contributions were determined to identify the alleles that most likely accounted for the possible differences between controls and subjects with BH, BHWCD, and CBD. HLA-DRB1*1302 was increased in BH and CBD, HLA-DRB1*1308 was increased in BHWCD, HLA-DRB1*0901 was increased in BH and BHWCD, and HLA-DRB1*0302 was increased in CBD. We compared the sequences of the four alleles associated with BH, BHWCD, or CBD. The amino acid-polymorphic epitopes for these alleles are illustrated in Table 8. Twenty-five different epitopes were identified. We determined the frequencies of these epitopes in controls and in subjects with BH, BHWCD, and CBD. Although no specific epitope achieved statistical significance (Bonferroni's corrected n = 31), three epitopes had uncorrected values of p ⩽ 0.01 and were considered to be possibly associated with beryllium-related abnormality. All three associations were with CBD (HLA-DRB1-S11, -N37 and -E71). To determine whether these amino acid epitopes might identify subjects with BH who were more likely to be diseased, we compared the frequency of these amino acid epitopes in BHWCD versus CBD. HLA-DRB1-S11 was significantly associated with CBD (p < 0.02), whereas HLA-DRB1- E71 was of borderline significance (p = 0.055). There was no statistically significant difference in the frequency of HLA-DRB1-N37 in BHWCD versus CBD (p = 0.103). Thus, amino acid epitopes on HLA-DRB1 may be important for CBD.
| Alleles | Controls (n = 82) | BH (n = 55) | BH No Disease (n = 30) | CBD (n = 25) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n* | % | n* | % | n* | % | n* | % | |||||||||
| DRB1*0101 | 18 | 10.0 | 8 | 7.3 | 4 | 6.7 | 4 | 8.0 | ||||||||
| DRB1*0102 | 1 | 0.6 | 2 | 1.8 | 1 | 1.7 | 1 | 2.0 | ||||||||
| DRB1*0103 | 1 | 0.6 | 2 | 1.8 | 0 | 0.0 | 2 | 4.0 | ||||||||
| DRB1*1501 | 13 | 7.9 | 8 | 7.3 | 6 | 10.0 | 2 | 4.0 | ||||||||
| DRB1*1502 | 4 | 2.4 | 2 | 1.8 | 0 | 0.0 | 2 | 4.0 | ||||||||
| DRB1*1601 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*0401 | 10 | 6.1 | 1 † | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*0402 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*0403 | 0 | 0.0 | 2 | 1.8 | 1 | 1.7 | 1 | 2.0 | ||||||||
| DRB1*0404 | 11 | 6.7 | 2 | 1.8 | 2 | 3.3 | 0 | 0.0 | ||||||||
| DRB1*0405 | 1 | 0.6 | 2 | 1.8 | 2 | 3.3 | 0 | 0.0 | ||||||||
| DRB1*0408 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*0413 | 2 | 1.2 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*1101 | 10 | 6.1 | 7 | 6.4 | 4 | 6.7 | 3 | 6.0 | ||||||||
| DRB1*1102 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*1104 | 9 | 5.5 | 4 | 3.6 | 4 | 6.7 | 0 | 0.0 | ||||||||
| DRB1*1105 | 3 | 1.8 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*1201 | 4 | 2.4 | 2 | 1.8 | 1 | 1.7 | 1 | 2.0 | ||||||||
| DRB1*1206 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*1301 | 9 | 5.5 | 7 | 6.4 | 4 | 6.7 | 3 | 6.0 | ||||||||
| DRB1*1302 | 3 | 1.8 | 7 † | 6.4 | 2 | 3.3 | 5 † | 10.0 | ||||||||
| DRB1*1301/6 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*1305 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*1306 | 5 | 3.0 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*1308 | 0 | 0.0 | 2 | 1.8 | 2 † | 3.3 | 0 | 0.0 | ||||||||
| DRB1*1401 | 5 | 3.0 | 2 | 1.8 | 0 | 0.0 | 2 | 4.0 | ||||||||
| DRB1*1402 | 1 | 0.6 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*1405 | 1 | 0.6 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*1406 | 1 | 0.6 | 2 | 1.8 | 0 | 0.0 | 2 | 4.0 | ||||||||
| DRB1*0701 | 24 | 14.6 | 18 | 16.4 | 13 | 21.7 | 5 | 10.0 | ||||||||
| DRB1*0801 | 3 | 1.8 | 4 | 3.6 | 2 | 3.3 | 2 | 4.0 | ||||||||
| DRB1*0802 | 0 | 0.0 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*0803 | 1 | 0.6 | 1 | 0.9 | 0 | 0.0 | 1 | 2.0 | ||||||||
| DRB1*0804 | 3 | 1.8 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||||||||
| DRB1*0901 | 2 | 1.2 | 7 † | 6.4 | 5 † | 8.3 | 2 | 4.0 | ||||||||
| DRB1*0301 | 16 | 9.8 | 10 | 9.1 | 7 | 11.7 | 3 | 6.0 | ||||||||
| DRB1*0302 | 0 | 0.0 | 2 | 1.8 | 0 | 0.0 | 2 † | 4.0 | ||||||||
| Allele | 9 | 11 | 13 | 28 | 30 | 37 | 47 | 57 | 60 | 67 | 70 | 71 | 74 | 78 | 86 | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DRB1*0302 | E | S | S | E | Y | N | Y | D | Y | L | Q | K | R | Y | G | |||||||||||||||
| DRB1*1302 | E | S | S | D | Y | N | F | D | Y | I | D | E | A | Y | G | |||||||||||||||
| DRB1*1308 | E | S | S | D | Y | F | Y | D | Y | I | D | E | A | Y | V | |||||||||||||||
| DRB1*0901 | K | D | F | H | G | N | Y | V | S | F | R | R | E | V | G |
In this study, we compared subjects with BH with and without clinical disease and beryllium-exposed workers without BH for the molecular sequence of HLA-DPB1, -DQB1, and -DRB1 alleles. This is the first study to include a significant number of subjects with BH without clinical disease who were followed for a meaningful period (22, 23). We demonstrated that HLA-DPB1-E69 is a marker for BH and does not discriminate whether an individual will be asymptomatic (BHWCD) or symptomatic (CBD).
HLA-DP, and particularly HLA-DPB1-E69, has been associated with a number of diseases, including anti-Scl antibody expression in systemic sclerosis (24), hard metal lung disease (24, 25), resistance to mixed-cellularity Hodgkin's disease (26), and, in one study (27) but not in another (28), with sarcoidosis. We confirmed prior observations that HLA-DPB1-E69 (12-14) is associated with CBD. These findings are strengthened by the observation that the HLA-DPB1 alleles containing E 69 that were identified in our study differed from the alleles identified in the other studies. In our study, HLA-DPB1*0601 was the allele most strongly associated with BH, whereas in the initial study of HLA associations with beryllium disease (12), HLA-DPB1*0201 was associated with CBD. A second study (13) confirmed the association of HLA-DPB1-E69 with CBD, but the association was strongest with HLA-DPB1*0901, *1001, and *1701. Thus, the common finding in all three studies is an association of HLA-DPB1-E69 with BH. This finding is further supported by the recently reported findings of inhibition of Be-induced proliferation of T-cell lines and clones by anti-HLA-DP (15, 16) in samples from patients with CBD. The studies in which these findings were made defined the functional role of HLA-DP (i.e., presentation of antigenic peptides to CD4+ T cells) in CBD, and demonstrated the importance of HLA-DPB1-E69.
Our study extends the observations of the prior studies by demonstrating that E69-containing HLA-DPB1 alleles confer a risk not only for the development of CBD but also for the development of BH whether or not an individual develops clinical disease. The HLA-DPB1 alleles that contained L11 and D55 also carried a significant risk for the development of BH; however, this was due to the presence of E69 in some of the L11- or D55-containing alleles. Only those alleles of L11 or D55 that contained both L11 and E69 or D55 and E69 were significantly associated with BH. Because the OR for these amino acid-epitope combinations was lower than the OR for E69 alone (data not shown), the associations with L11 and D55 are secondary to the association with E69. In addition, these amino acid-epitope combinations did not associate preferentially with BHWCD as opposed to CBD and were therefore markers only for BH. Wang and coworkers (13) observed that specific E69-containing HLA-DPB1 alleles (i.e., non-*0201) increased the risk for CBD in his population. Although we also observed that the relatively rare E69-containing HLA-DPB1 alleles were increased in our cases of CBD, the risk for CBD if an individual expressed one of these rare E69-containing HLA-DPB1 alleles (OR = 3.28, 95% confidence interval [CI]: 1.28 to 8.43) was not as great as the risk for CBD if an individual expressed any E69-containing HLA-DPB1 allele (OR = 5.8, CI: 1.8 to 18.4). In addition, we did not observe any increase in the risk for CBD among individuals who expressed two E69-containing HLA-DPB1 alleles (data not shown).
The association of BH with an acidic amino acid (E69) in HLA-DP in three separate studies raises the possibility of a direct interaction with the positively charged Be2+ cation. Be+ is the divalent cation with the smallest ionic radius of any divalent cation and may have access to binding sites unavailable to larger divalent cations. When E69 is present in HLA-DPB1, resulting epitope contains three consecutive acidic amino acids, E67, E68, and E69. It is possible that a form of beryllium could have access to and interact directly with these amino acids. By changing the electrostatic interactions of an important peptide ligand binding pocket, beryllium might change the affinity of an HLA-DP molecule for peptide ligands. Amicosante and colleagues (29) have recently demonstrated that invariant chain derived peptide binding to soluble HLA-DP2, but not to HLA-DP2Lys69, is decreased in the presence of beryllium, and that beryllium decreased the binding of the mAb NFLD.M60 to L cells transfected with HLA-DP2 but not to L cells transfected with HLA-DP2Lys69. This demonstrates that beryllium can effect peptide binding to HLA-DP containing E69. The role of HLA-DP for presenting antigen ligands to T cells in CBD has also been demonstrated (15, 16), and we have preliminary data (not shown) that this is also true in BHWCD.
In the studies that evaluated HLA-DPB1-E69 in CBD (12– 14), less than 100% of the patients with CBD tested expressed HLA-DPB1-E69. We have also observed that less than 100% of our subjects with BH were positive for HLA-DPB1-E69. This strongly implies that other HLA molecules must be able to present beryllium-influenced ligands to CD4+ T cells. In the present study, a non-E69-containing HLA-DPB1 allele, HLA-DPB1*0402, was associated with CBD with a post hoc contribution of > 1.96. Of interest was that three of the four patients with CBD who did not express an E69-containing allele had an HLA-DPB1*0402 allele. This suggests that in a minority of CBD patients, a non-E69–containing allele may be important. HLA-DPB1*0402 contains the negatively charged D55E56 and may therefore interact with beryllium in a manner similar to the negatively charged E69.
The finding that certain groups of HLA-DPB1 alleles may be associated with BH, whereas others may be associated with CBD, has important implications. It suggests that peptides with a high affinity for specific HLA-DP proteins in the presence of beryllium may be important for determining the nature of the immune response to this metal. Peptides that have a high affinity for E69-containing HLA-DPB1 alleles in the presence of beryllium may elicit an immune response that could result in either protection or disease, whereas peptides that bind to HLA-DPB1*0402 in the presence of beryllium may preferentially elicit an immune response that results in disease. A somewhat similar situation of HLA selection of peptide response may occur in the immune response to Trichophyton antigen Tri r 2 (30), where a response to peptide 5 is associated with a protective immune reponse. Selective responses to peptides may also be responsible for clinical disease in tuberculosis (31).
In the present study, alleles of HLA-DRB1 and -DQB1 were evaluated for their association with BH. A significant association was noted for HLA-DQB1-G86 with CBD (pcorr < 0.04, OR = 6.0), and a possible association (univariate p < 0.01) with CBD was noted for HLA-DRB1-S11, -N37, and -E71. Three possibilities may account for these associations. First, HLA-DR/DQ may bind directly beryllium-influenced peptides and present them to T cells. In a recent study involving HLA-defined antigen-presenting cells, HLA-DRB1*1501 may have been presenting antigens to CD4+ T cells (15), and it is therefore possible that besides HLA-DP, other HLA Class II molecules may, in specific circumstances, have the ability to present antigenic peptides in BH. In addition, other HLA Class II loci in linkage disequilibrium with the amino acid epitopes found to be possibly associated with CBD may be responsible for molecules presenting antigenic peptides to CD4+ T cells in CBD. This includes not only alleles at the HLA-DRB1 and -DQB1 loci, but also HLA Class II alleles at the DQA1 and DRB3, B4, and B5 loci.
A second possibility is that HLA-DRB1/HLA-DQB1 alleles may contribute peptides that are presented by HLA-DP. A similar mechanism has already been postulated to explain the association of HLA-DRB1 and -DQB1 alleles with rheumatoid arthritis, in which certain HLA-DRB1 epitopes, when present in the context of certain HLA-DQB1 alleles, appear to influence disease progression. In HLA-DQ8–transgenic mice, the HV3 65-79 peptide ligand of the rheumatoid arthritis–nonassociated HLA-DRB1*0402 allele has induced HLA-DQ8– restricted T-cell responses, whereas the peptide ligands from the rheumatoid arthritis–associated HLA-DRB1*0401 or *0404 alleles have not (32). The same peptide ligand from HLA-DRB1*0402 can also prevent collagen-induced arthritis in these HLA-DQ8–transgenic mice (33). Thus, peptide ligands from HLA Class II molecules may influence the immune response by their ability to bind to other HLA Class II molecules.
A further explanation for the possible association of certain HLA-DRB1/HLA-DQB1 alleles with an increased risk of CBD is that the alleles of HLA-DRB1/HLA-DQB1 are in linkage disequilibrium with some other gene that is modulating the immune response and is responsible for the development of clinical disease. Now that the complete sequence and gene map of the human major histocompatibility complex has been published (34), identification of other genes associated with CBD should be feasible.
In the present study, particular care was taken to define the phenotypes of the subjects with BH. Subjects without clinical disease were followed for up to 7 yr, through chest radiography and pulmonary function studies, without showing any signs of clinical disease. Although it is possible that in the future, some of these individuals will develop clinical beryllium disease, we feel that it is unlikely. Most of the subjects with the diagnosis of CBD had obvious clinical disease, with abnormalities on chest radiographs and in pulmonary function tests. In addition, the significant differences between the responses of BAL cells to beryllium stimulation (BH without clinical disease: SI = 16.4 ± 7.2; CBD: SI = 89.6 ± 17.7 [p < 0.001]) strongly suggest that the immunologic response to beryllium has different consequences in these two groups. Surveys of current and retired workers (4-8) suggest that from 42% to 83% of workers who have BH also have evidence of beryllium disease. Thus, although many individuals with evidence of BH will develop disease, not all necessarily do so (9). The concept that BH does not necessarily progress to CBD has been previously addressed (35); however, no data are available on the natural history of BH and risk for the future development of CBD among patients found to have BHWCD. That none of our 30 subjects with BHWCD have developed clinical disease with a follow-up lasting up to 7 yr suggests that the rate of progression to clinical disease from established BHWCD is very low. Since most of these individuals were referred from initial screening programs, the duration of their BH was unknown. However, we have seen anecdotal cases (M. Rossman, personal communication) of the development of rapid onset (within 1 yr) of clinical disease in workers who have recently developed BH. A similar relationship between the presence of specific cellular immune responses and the presence of disease has also been observed for pigeon breeder's disease (36, 37). Although it is highly likely that differences in beryllium exposure and cytokine responsiveness play a role in these differences, the data presented in this paper suggest that differences in HLA alleles, and presumably in peptide binding, may also play a role in the susceptibility to BH and the progression to CBD. These observations may also have important implications in other hypersensitivity/autoimmune reactions.
Genetic screening for susceptibility to beryllium disease on the basis of specific HLA molecules has been suggested (12– 14). We feel that this may be premature, since the interaction of HLA molecules and the manifestations of BH are complex. The recent observation (13) that HLA-DPA1 alleles may also be associated with beryllium disease illustrates another complexity. Until more is known about the risks associated with specific HLA Class II alleles, the mechanisms of beryllium-induced lymphocyte proliferation, and particularly the ways in which beryllium, HLA Class II molecules, and peptides interact, preemployment genetic screening in the beryllium industry will be problematic.
The authors thank Mary McNichol for assistance in preparation of the manuscript.
Supported by NIH grant HL-48210.
| 1. | Rossman MD, Kern JA, Elias JA, Cullen MR, Epstein PE, Preuss OP, Markham TN, Daniele RPProliferative response of bronchoalveolar lymphocytes to beryllium. A test for chronic beryllium disease (published erratum appears in Ann Intern Med 1989;15;110:672). Ann Intern Med1081988687693 |
| 2. | Rossman MD, Jones-Williams W. Immunopathogenesis of chronic beryllium disease. In: Rossman MD, Preuss OP, Powers MB, editors. Beryllium: biomedical and environmental aspects. Baltimore: Williams & Wilkins; 1991. p. 121–132. |
| 3. | Sterner JH, Eisenbud MEpidemiology of beryllium intoxication. Arch Ind Hyg Occup Med41961123151 |
| 4. | Kreiss K, Newman LS, Mroz MM, Campbell PAScreening blood test identifies subclinical beryllium disease. J Occup Med311989603608 |
| 5. | Kreiss K, Mroz MM, Zhen B, Martyny JW, Newman LSEpidemiology of beryllium sensitization and disease in nuclear workers. Am Rev Respir Dis1481993985991 |
| 6. | U.S. Department-of-Energy. Chronic beryllium disease prevention program; final rule. Federal Register 1999;64(235):68854–68914. |
| 7. | Kreiss K, Mroz MM, Newman LS, Martyny J, Zhen BMachining risk of beryllium disease and sensitization with median exposures below 2 micrograms/m3. Am J Ind Med3019961625 |
| 8. | Kreiss K, Mroz MM, Zhen B, Wiedemann H, Barna BRisks of beryllium disease related to work processes at a metal, alloy, and oxide production plant. Occup Environ Med541997605612 |
| 9. | Rossman MD, Regovich J, Atochina O, Liang J, Lee CW, Lee JProgressive lung dysfunction in pre-clinical beryllium disease. Am J Respir Crit Care Med1591999A621 |
| 10. | Saltini C, Winestock K, Kirby M, Pinkston P, Crystal RGMaintenance of alveolitis in patients with chronic beryllium disease by beryllium-specific helper T cells. N Engl J Med320198911031109 |
| 11. | Fontenot AP, Falta MT, Freed BM, Newman LS, Kotzin BLIdentification of pathogenic T cells in patients with beryllium-induced lung disease. J Immunol163199910191026 |
| 12. | Richeldi L, Sorrentino R, Saltini CHLA-DPB1 glutamate 69: a genetic marker of beryllium disease. Science2621993242244 |
| 13. | Wang Z, White P, Petrovic M, Tatus O, Newman L, Maier L, Marrone BDifferential susceptibilities to chronic beryllium disease contributed by different glu-69 HLA-DPB1 and -DPA1 alleles. J Immunol163199916471653 |
| 14. | Richeldi L, Kreiss K, Mroz MM, Zhen B, Tartoni P, Saltini CInteraction of genetic and exposure factors in the prevalence of berylliosis. Am J Ind Med321997337340 |
| 15. | Fontenot AP, Torres M, Marshall WH, Newman LS, Kotzin BLBeryllium presentation to CD4+ T cells underlies disease-susceptibility HLA-DP alleles in chronic beryllium disease. Proc Natl Acad Sci USA9720001271712722 |
| 16. | Lombardi G, Germain C, Uren J, Fiorillo MT, duBois RM, Jones-Williams W, Saltini C, Sorrentino R, Lechler RHLA-DP Allele-specific T cell responses to beryllium account for DP-associated susceptibility to chronic beryllium disease. J Immunol166200135493555 |
| 17. | Vaughan RW, Lanchbury JS, Marsh SG, Hall MA, Bodmer JG, Welsh KIThe application of oligonucleotide probes to HLA Class II typing of the DRB sub-region. Tissue Antigens361990149155 |
| 18. | Obata F, Ito K, Kaneko T, Yang Y-G, Onda K, Ito I, Yabe N, Watanabe K, Kashiwagi NHLA-DR gene frequencies in the Japanese population obtained by oligonucleotide genotyping. Tissue Antigens381991124132 |
| 19. | Bugawan TL, Begovich AB, Erlich HARapid HLA-DPB typing using enzymatically amplified DNA and nonradioactive sequence-specific oligonucleotide probes. Immunogenetics321990231241 |
| 20. | Ballas M, Oechsle G, Eiermann TH, Müller C, Wöpl A, Goldmann SF. A DRB oligonucleotide typing system defining 50 of 56 DRB alleles. In: Tsuji K, Aizawa M, Sasazuki T, editors. Proceedings of the Eleventh International Histocompatibility Workshop and Conference. New York: Oxford University Press; 1992. p. 311–314. |
| 21. | Chicz RM, Graziano DF, Trucco M, Strominger JL, Gorga JCHLA-DP2: Self peptide sequences and binding properties. J Immunol159199749354942 |
| 22. | Saltini C, Richeldi L, Losi M, Amicosante M, Voorter C, van den Berg-Loonen E, Dweik RA, Wiedemann HP, Deubner DC, Tinelli CMajor histocompatibiity locus genetic markers of beryllium sensitization and disease. Eur Respir J18200118 |
| 23. | Wang Z, Farris GM, Newman LS, Shou Y, Maier LA, Smith HN, Marrone BLBeryllium sensitivity is linked to HLA-DP genotype. Toxicology16520012738 |
| 24. | Rihs HP, Conrad K, Mehlhorn J, May-Taube K, Welticke B, Frank KH, Baur XMolecular analysis of HLA-DPB1 alleles in idiopathic systemic sclerosis patients and uranium miners with systemic sclerosis. Int Arch Allergy Immunol1091996216222 |
| 25. | Potolicchio I, Mosconi G, Forni A, Nemery B, Seghizzi P, Sorrentino RSusceptibility to hard metal lung disease is strongly associated with the presence of glutamate 69 in HLA-DPβ chain. Eur J Immunol27199727412743 |
| 26. | Taylor GM, Gokhale DA, Crowther D, Woll PJ, Harris M, Ryder D, Ayres M, Radford JAFurther investigation of the role of HLA-DPB1 in adult Hodgkin's disease (HD) suggests an influence on susceptibility to different HD subtypes. Br J Cancer80199914051411 |
| 27. | Lympany PA, Petrek M, Southcott AM, Newman Taylor AJ, Welsh KI, Du Bois RM. HLA-DPB polymorphisms: glu 69 association with sarcoidosis. Eur J Immunogenet 1996;23:353–359. |
| 28. | Maliarik M, Chen K, Major M, Sheffer R, Popovich JJ, Rybicki R, Iannuzzi MAnalysis of HLA-DPB1 polymorphisms in African-Americans with sarcoidosis. Am J Respir Crit Care Med1581998111114 |
| 29. | Amicosante M, Sanarico N, Berretta F, Arroyo J, Lombardi G, Lechler R, Colizzi V, Saltini CBeryllium binding to HLA-DP molecule carrying the marker of susceptibility to berylliosis glutamate β69. Hum Immunol622001686693 |
| 30. | Woodfolk JA, Sung SJ, Benjamin DC, Lee JK, Platts-Mills TAEDistinct human T cell repertoires mediate immediate and delayed-type hypersensitivity to the Trichophyton antigen, Tri r 2. J Immunol165200043794387 |
| 31. | Vordermeier HM, Harris DP, Friscia G, Roman E, Surcel HM, Moreno C, Pasvol G, Ivanyi JT cell repertoire in tuberculosis: selective anergy to an immunodominant epitope of the 38-kDa antigen in patients with active disease. Eur J Immunol22199226312637 |
| 32. | Zanelli E, Krco CJ, Baischi JM, Cheng S, David CSImmune response of HLA-DQ8 transgenic mice to peptides from the third hypervariable region of HLA-DRB1 correlates with predisposition to rheumatoid arthritis. Proc Natl Acad Sci USA93199618141819 |
| 33. | Das P, Bradlley DS, Geluk A, Griffiths MM, Luthra HS, David CSAn HLA-DRB1*0402 derived peptide (HV3 65-79) prevents collagen-induced arthritis in HLA-DQ8 transgenic mice. Hum Immunol601999575582 |
| 34. | MHC-Sequencing-Consortium. Complete sequence and gene map of a human major histocompatibility complex. Nature 1999;401:829–938. |
| 35. | Newman LS, Lloyd J, Daniloff E. The natural history of beryllium sensitization and chronic beryllium disease. Environ Health Perspect 1996; 104(Suppl. 5):937–943. |
| 36. | Bouic PJ, Nel NC, de Beer PM, Joubert JRIn vitro reactivities of blood lymphocytes from symptomatic and asymptomatic pigeon breeders to antigen and mitogens. Int Arch Allergy Appl Immunol891989222228 |
| 37. | Mendoza F, Melendro EI, Baltazares M, Banales JL, Ximenez C, Chapela R, Selman MCellular immune response to fractionated avian antigens by peripheral blood mononuclear cells from patients with pigeon breeder's disease. J Lab Clin Med12719962328 |
