High Frequency of Aspartic Acid at Position 57 of HLA-DQ β-Chain in Japanese IDDM Patients and Nondiabetic Subjects

Diabetes, Feb 1990

The HLA-DQ β-chain (DQB1) genes of 72 Japanese patients with insulin-dependent diabetes mellitus (IDDM) and 85 control subjects were studied with polymerase chain-reaction (PCR) amplification and allele-specific oligonucleotide hybridization. DQw4 (DQBBIank) and DQw9 (DQB3.3) were increased in IDDM patients compared with the control subjects, and DQB1.2, DQB1.9, and DQw7 (DQB3.1) were decreased. Thirty-five (48.6%) IDDM patients had both alleles carrying an aspartic acid at position 57 of the DQ β-chain (Asp 57), 35 (48.6%) were Asp 57/non-Asp 57 heterozygous, and 2 (2.8%) had non-Asp 57 alleles only. Of 85 control subjects, the respective values for these three genotypes were 49 (57.6%), 29 (34.1%), and 7 (8.2%), respectively. The high frequency of Asp 57 alleles in both IDDM and control subjects contrasts with data for Whites. Therefore, the Asp 57 hypothesis that the presence of an aspartic acid at position 57 of DQ β-chain provides protection against developing IDDM is not tenable for Japanese IDDM patients. The DRB1 gene, particularly position 57 of the DR β-chain, may contribute to IDDM susceptibility in Japanese.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://diabetes.diabetesjournals.org/content/39/2/266.full.pdf

High Frequency of Aspartic Acid at Position 57 of HLA-DQ β-Chain in Japanese IDDM Patients and Nondiabetic Subjects

TAKUYA AWATA 0 1 TAKESHI KUZUYA 0 1 AYAKO MATSUDA 0 1 YASUHIKO IWAMOTO 0 1 YASUNORI KANAZAWA 0 1 MACHIKO OKUYAMA 0 1 TAKEO JUJI 0 1 0 RESEARCH DESIGN AND METHODS Seventy-two unrelated IDDM patients attending Jichi Medi- cal School Hospital were studied. The diagnosis of IDDM was not merely defined clinically but also confirmed by serum and urine C-peptide assays. Twenty-three patients were men, and 49 patients were women; mean age was 39 yr (range 16-72 yr); mean age at onset of IDDM was 30 yr (range 5-64 yr). HLA typing for the DR locus was done by the National Institutes of Health standard microcytotoxicity test with the 9th International Histocompatibility Workshop serums. This system is able to detect DR antigens - 1 , -2, -3, -4, -w6, -7, -w8, -9, -w10, -w11(5), and -w12(5). Eighty- five HLA-typed control subjects were collected from Blood Transfusion Service, Tokyo University Hospital; all were Jap- anese and were unrelated 1 From the Division of Endocrinology and Metabolism, Jichi Medical School, Tochigi-ken, and the Blood Transfusion Service, Tokyo University Hospital , Tokyo , Japan. Endocrinology and Metabolism, Jichi Medical School , Minamikawachi-machi, Kawachi-gun, Tochigi-ken, 329-04 Japan. Received for publication 20 September 1989 and accepted in revised form 1 November 1989 - Lated with particular HLA antigens. DR3 and DR4 are ike several other autoimmune diseases, insulin-dependent diabetes mellitus (IDDM) is clearly associpositively and DR2 is negatively associated with IDDM in Whites (1,2). HLA class II antigens are heterodimers, and their genes are located on the short arm of chromosome 6. This locus is now divided into three major subregions: DR, DQ, and DP. Several studies have reported that DQ rather than DR might be more closely related to IDDM susceptibility (3-6). Todd et al. (7) and Horn et al. (8) reported that HLA-DQB, particularly amino acid residue 57 of DQ p-chain, was strongly correlated with susceptibility and resistance to IDDM in Whites. The presence of an aspartic acid at position 57 seems to provide protection against the development of IDDM, and Morel et al. (9) showed that almost all White patients with IDDM did not have an aspartic acid at position 57 of the DQ p-chain. In Japan, several studies have reported that DR4 and DR9 are positively and DR2 is negatively associated with IDDM (10-12); DR3 is a rather rare antigen in Japanese, and its association with IDDM is not clear. Little is known about the association of HLA-DQ and IDDM in Japanese. As Aparicio et al. (11) pointed out, it was reported that both haplotypes DR4-DQw4 (DQBBIank [Bl.]) and DR9-DQw9 (DQB3.3), which are associated with IDDM in Japanese, usually have an aspartic acid at the position 57 of the DQ p-chain (7,8). This fact might indicate that the Asp 57 hypothesis stated above is not tenable for Japanese IDDM patients. To answer this question, we analyzed position 57 of the DQ p-chain (DQB1) in Japanese IDDM patients and nondiabetic subjects by the methods of gene amplification and oligonucleotide dot-blot hybridization. was significantly decreased in IDDM {Pc < 0.02, RR = 0.05; Table 1). Frequency of DQB1 alleles. Figure 2 shows an example of dot-blot analyses of 8 individuals with DQB1 allele-specific oligonucleotide probes (Fig. 1). Nine DQB1 alleles were determined in this study: DQB1.1, -1.2, -1.9, -1.AZH, -2, -3.1, -3.2, -3.3, and -Bl. Increased frequencies in IDDM were observed in DQB3.3 (52.8 vs. 36.5%, P < 0.05, RR = 1.93) and DQBBI. (52.8 vs. 25.9%, P < 0.001, RR = 3.15) (Table 2). Decreased frequencies in IDDM were observed in DQB1.2 (1.4 vs. 28.2%, P < 0.0001, RR = 0.05), DQB1.9 (11.1 vs. 27.1%, P < 0.02, RR = 0.35), and DQB3.1 (4.2 vs. 18.8%, P < 0.01, RR = 0.21). When corrected P values were applied, a significant increase was observed in DQBI. {Pc < 0.01), and a significant decrease was observed in DQB1.2 {Pc < 0.001). DQB3.2 was increased in IDDM (RR = 1.63) but was not significant in this study. Of nine IDDM patients carrying only DR4, 5 patients were genotyped as DQBBI. only, and 4 were DQB3.2/BI.; of 13 IDDM patients carrying only DR9, 10 were genotyped as DQB3.3 only, 2 were DQB3.2/3.3, and 1 was DQB1.9/3.3. Therefore, the haplotypes associated with IDDM in Japanese were considered to be DR4-DQw4 (DQBBI.) and DR9-DQw9 (DQB3.3), both rarely observed in Whites. Amino acid at position 57 of DQB1 antigen. DQB1 2, -1.9, -3.1, -3.3, and -Bl. correspond to the alleles carrying an aspartic acid in position 57 (Asp 57 alleles) and -1.1,-1 .AZH, -2, and -3.2 correspond to the non-Asp 57 alleles. We summarized the genotype frequency of DQB1 locus (Table 3). Of 72 IDDM patients, 35 (48.6%) had Asp 57 alleles only, 35 (48.6%) were Asp 57/non-Asp 57 heterozygous, and 2 (2.8%) had non-Asp 57 alleles only. In the control subjects, FIG. 1. Sequences encoding amino acids 52-59. Boxed regions were used as oligonucleotide probes. Designation of DQB1 alleles are same as described in ref. 10. DQB3.1 and DQB3.3 are identical in this region and were distinguished by different probes. Polymerase chain reaction (PCR). Genomic DNA was extracted from peripheral blood by the method that uses guanidine hydrochloride (13). DNA (1 jxg) was subjected to PCR (14) with 2.5 U Taq polymerase (Perkin-Elmer/Cetus, Norwalk, CT) in 100 |xl reaction mixture. Thirty-five rounds of cycling (denaturation, 1 min at 93C; annealing, 1 min at 56C; and extension, 2 min at 70C) with primers 5'-CTCGAATTCGCATGTGCTACTTCACCAACG-3' and 5'-GAGCTGCAGGTAGTTGTGTCTGCACAC-3' were carried out, and a 230-base pair sequence coding the polymorphic outer first domain of HLA-DQ (3-chain was amplified. To evaluate whether the amplification was sufficient, 4 (JLI of product was electrophoresed on 2% SeaKem agarose gels (FMC, Rockland, ME). Oligonucleotide dot-blot analysis. Four microliters of PCRamplified DNA were denatured and dot blotted onto nylon membranes (Hybond N, Amersham, Tokyo). After fixation of DNA by exposure to ultraviolet light, the membranes were prehybridized at 42C for >2 h in a solution of 5 x SSPE (1 x SSPE = 0.15 M NaCI, 1 mM EDTA, and 10 mM sodium phosphate, pH 7.4), 5 x Denhardt's solution, 0.5% sodium dodecyl sulfate (SDS), and 20 jxg/ml denatured salmon sperm DNA. Oligonucleotide probes were end labeled with [32P]ATP by standard procedure with T4 DNA kinase. The membranes were hybridized at 42C for >16 h with the labeled probes in the same solution with prehybridization. Membranes were first washed two times in 2 x SSPE and 0.1% SDS at room temperature for 10 min and then washed in 3 M tetramethylammonium chloride solution (3 M tetramethylammonium chloride; 50 mM Tris-HCI, pH 7.4; 2 mM EDTA; and 0.3% SDS) at 56-60C for 30-60 min (15). We used eight oligonucleotide probes, covering amino acid positions 54-59 or 52-58 of DQ p-chain (Fig. 1) and also used the following six oligonucleotide probes to distinguish DQB3.1 from DQB3.3 and to confirm the data: 5'-CGTGGAGGTGTACCGGGCG-3' (amino acid positions 43-49), 5'-CTGGAGGAGGACCGGGC-3' (positions 68-73), 5'-CGTGCGTTATGTGACCA-3' (positions 23-29), 5'-CGTGCGGGGTGTGACCA-3' (positions 23-29), 5'CGTGCGTCTTGTGACCA-3' (positions 23-29), and 5'CGTGCGTCTTGTAACCA-3' (positions 23-29). Statistical analysis. Differences were assessed by x2-tests. For the 2 x 2 tables, x2-value and relative risk (RR) were calculated by a modification of Haldane's formula (16,17). Pc indicates a P value corrected by multiplying P by the number of alleles studied (11 for DR antigens and 9 for DQB1 alleles). RESULTS Frequency of HLA-DR antigens. Significant differences were observed in DR2 (Pc < 0.001, RR = 0.06) and DR9 (Pc < 0.05, RR = 2.52). DR4 was increased in IDDM (RR = 1.75) but was not significant in this study. DRw11 NS Pc, P value corrected by multiplying P by the number of alleles studied; RR, relative risk calculated by a modification of Haldane's formula (16,17). 1.9 1.AZH 3.1,3 3.2 Blank (3.1,3.3) the respective values were 49 (57.6%), 29 (34.1%), and 7 (8.2%). DISCUSSION In this study, we investigated the HLA-DR antigens and the DQB1 genes of 72 Japanese IDDM patients and 85 control subjects. We observed that DR9 was significantly increased in IDDM compared with control subjects, and DR2 and DRw11 (5) were significantly decreased. DR4 was increased in IDDM, although not significantly in this study. These results, except for that of DRw11(5), are in agreement with previous studies (10-12). The novel negative association of DRw11 (5) with Japanese IDDM requires verification by future studies. With PCR amplification and allele-specific oligonucleotide hybridization, nine different alleles of DQB1 genes were detected. Note that both DQB3.3 and DQBBI., which are positively associated with IDDM, have an aspartic acid at position 57 of DQ p-chain and that the DQB3.3 allele sequence differs from DQB3.2 allele only at position 57 in the first domain (7,8). A high frequency of Asp 57 alleles was detected in both Japanese IDDM patients and control subjects (Table 3). This result contrasts with data in Whites reported by Morel et al. (9); of 27 IDDM patients, none was Asp 57 homozygous, 1 (4%) was Asp 57/non-Asp 57 heterozygous, and 26 (96%) were non-Asp 57 homozygous, whereas those frequencies were 42 (34.1%), 57 (46.3%), and 24 (19.5%), respectively, in 123 control subjects. Therefore, the Asp 57 hypothesis that the presence of an aspartic acid in the position 57 of DQ p-chain provides protection against the development of IDDM does not apply to Japanese IDDM patients. Other genes of the HLA class II antigens or genes in linkage disequilibrium with the class II locus must be sought. With regard to DQ a-chains (DQA1), we obtained preliminary data that all individuals who were either DR4/-, DR9/-, or DR4/9 (32 IDDM patients and 17 control subjects) have the usual DQA1 allele (designated DQA3), known to be associated with DR4 and DR9 by dot-blot analysis with the oligonucleotide probe of 5'-TTCCGCAGATTTAGAAGAT3' (amino acid positions 51-57; 7,8,18). It is possible that the DQA3 allele contributes to IDDM susceptibility in Japanese to some extent. However, it has been reported that the DR4-DQw7 (DQB3.1) haplotype, which is usually associated with the DQA3 allele (7,8), is negatively associated with IDDM in Whites (19). Therefore, DQA3 may not essentially contribute to IDDM susceptibility in Japanese. Another possible and interesting explanation is that DR antigens, particularly position 57 of the DR p-chain, contrib Asp 57/non-Asp 57 non-Asp 57 IDDM 72 35 (48.6%) Nondiabetic 85 49 (57.6%) Asp/non-Asp indicates heterozygous individuals. Asp 57, DQB1.2, -1.9, -3.1, -3.3, and -Blank; non-Asp 57, DQB1.1, -1.AZH, -2, and -3.2. Differences were not significant by x2-test. ute to IDDM susceptibility in Japanese, because this position of DR4-DQw4 and DR9-DQw9 haplotypes is non-Asp, whereas most other haplotypes have an Asp in this position (7,20). The presence of an aspartic acid in position 57 of DR p-chain, not only that of DQp-chain, may be important to protect against IDDM development. However, further studies are needed to examine this possibility and to exclude other possibilities. Of 32 IDDM patients who were either DR4/-, DR9/-, or DR4/9, 1 (3%) patient had both alleles carrying an aspartic acid at position 57 of the DR pchain, 6 (19%) were Asp 57/non-Asp 57 heterozygous of DRB1, and 25 (78%) had non-Asp 57 alleles of DRB1 only by dot-blot analysis. This result is compatible with the possibility that position 57 of the DR p-chain contributes to IDDM susceptibility, although the possibility of DQ a-chain contribution remains. The contribution of DRB or DQA to IDDM susceptibility may be weaker than that of DQB, and this may be one of the reasons why the incidence of IDDM in Japanese is very low. ACKNOWLEDGMENTS This study was supported in part by Grant 01570654 from the Ministry of Education, Science, and Culture, Government of Japan, and a grant from Yamanouchi Foundation for Research on Metabolic Disorders. We thank Dr. S. Ohta, Department of Biochemistry 1, Jichi Medical School, for preparing the oligonucleotides. This paper was presented in part at the 49th annual meeting of the American Diabetes Association, Detroit, Michigan, June 1989.


This is a preview of a remote PDF: https://diabetes.diabetesjournals.org/content/39/2/266.full.pdf

Takuya Awata, Takeshi Kuzuya, Ayako Matsuda, Yasuhiko Iwamoto, Yasunori Kanazawa, Machiko Okuyama, Takeo Juji. High Frequency of Aspartic Acid at Position 57 of HLA-DQ β-Chain in Japanese IDDM Patients and Nondiabetic Subjects, Diabetes, 1990, 266-269, DOI: 10.2337/diab.39.2.266