HLA-DQA1*1 contributes to resistance and A1*3 confers susceptibility to Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects

Diabetologia, Feb 1991

In this study HLA-DQA1 and TNF genes in addition to HLA-DQB1 gene were investigated at DNA level for elucidation of the genetic backgrounds of Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects. DNA, amplified by polymerase chain reaction, was subjected to allele specific oligonucleotide dot blot analysis, restriction fragment length polymorphism analysis or DNA sequencing. Polymorphism of the TNF gene to NcoI did not correlate with Type 1 diabetes in Japanese patients. DQw1.2 had a protective effect against the disease, the DQA1*1 allele was significantly decreased and DQA1*3 allele was significantly increased. Seventeen out of twenty-two Type 1 diabetic patients (77%) were homozygous for DQA1*3 and five out of twenty-two (23%) heterozygous. The DQA1*3 gene of Type 1 diabetic patients had a normal nucleotide sequence. Furthermore, DQA1*3 was found unexpectedly in two patients without DR4 or DR9. These data indicate that DQA1 gene confers susceptibility and resistance to Type 1 diabetes in Japanese subjects.

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://link.springer.com/content/pdf/10.1007%2FBF00500386.pdf

HLA-DQA1*1 contributes to resistance and A1*3 confers susceptibility to Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects

Diabetologia H L A - D Q A I * I contributes to resistance and A l * 3 confers susceptibility to Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects 0 1 The Second Department of Internal Medicine, 2 Department of Nutrition and Physiological Chemistry, Osaka University Medical School, and 3 National Cardiovascular Center, Research Institute, Department of Surgical Research , Osaka , Japan 1 K. Y a m a g a t a 1, T. H a n a f u s a t, H. N a k a j i m a ~, M. S a d a 3, H. A m e m i y a 3, T. N o g u c h i 2, T. T a n a k a 2, N. K o n o t and S. T a r u i 1 2 Dr. K. Yamagata The Second Department of Internal Medicine Osaka University Medical School 1-1-50 Fukushima , Fukushima-ku Osaka 553 Japan Summary. In this study HLA-DQA1 and TNF genes in addition to HLA-DQB1 gene were investigated at DNA level for elucidation of the genetic backgrounds of Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects. DNA, amplified by polymerase chain reaction, was subjected to allele specific oligonucleotide dot blot analysis, restriction fragment length polymorphism analysis or D N A sequencing. Polymorphism of the TNF gene to NcoI did not correlate with Type 1 diabetes in Japanese patients. DQwl.2 had a protective effect against the disease, the D Q A I * I allele was significantly decreased and D Q A I * 3 allele was significantly increased. Seventeen out of twenty-two Type i diabetic patients (77%) were homozygous for D Q A I * 3 and five out of twenty-two (23%) heterozygous. The DQAI*3 gene of Type 1 diabetic patients had a normal nucleotide sequence. Furthermore, D Q A I * 3 was found unexpectedly in two patients without DR4 or DR9. These data indicate that DQA1 gene confers susceptibility and resistance to Type i diabetes in Japanese subjects. Type i (insulin-dependent) diabetes mellitus; HLA-DQA1 gene; HLA-DQB1 gene; tumour necrosis factor; polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) 9 Springer-Verlag 1991 Type 1 (insulin-dependent) diabetes mellitus is considered to be an autoimmune disease. Several genetic factors contribute to the pathogenesis of Type 1 diabetes, and human leucocyte antigen ( H L A ) genes are of great importance in occurrence of the disease. Recent studies at D N A level have shown that aspartic acid (Asp) at position 57 of the H L A - D Q ~ chain protects against Type 1 diabetes and non-aspartic acid (non-Asp) at the same position predisposes to the disease in Caucasians [ 1 ]. Recently however, we have proved that about half the Japanese Type i diabetic patients are homozygous for Asp at this position [ 2 ]. This finding indicates that firstly, other genetic markers such as the T N F gene should be explored in Japanese Type 1 diabetic patients and secondly, D Q A 1 gene should also be examined in relation to the antigen presenting function of the D Q (o~-~) molecule. Thus, we have investigated H L A - D Q A 1 , D Q B 1 and T N F genes in Japanese subjects. Subjects and methods Patients and controls Twenty-twounrelated Type i diabetic patients attending the Second Department of Internal Medicine of Osaka University Medical School Hospital were studied. All patients were diagnosed clinically according to the criteria of WHO and were on insulin therapy. Thirty-two control subjects were randomly selected from our staff. Informed consent was obtained from all patients and healthy control subjects. This work was performed in accordance with the principles of the Declaration of Helsinki. Serological HLA typing was done by the NIH microlymphocytotoxicity test. Polymerase chain reaction (PCR ) Genomic DNA was extracted from peripheral lymphocytes and about one microgram was subjected to PCR using 2.5 units of Taq DNA polymerase (AmpliTaq, Perkin-Elmer/Cetus, Norwalk, Ct., USA). Reaction conditions were performed as per the manufacturer's recommendations. Amplification was performedby 30 cycles of denaturation (94 ~ annealing (60 ~ and polymerization (72 ~ Oligonucleotides were synthesized by a DNA synthesizer (Model 381A, Applied Biosystems, Inc., Foster City, Calif., USA). TNFP1 (5'-GCACAGCAGGTGAGGCTCTCC-3') and TNFP2 (5'-GGTGGTGCCACACACCCTI'GG-3') were used as PCR primers for TNF gene amplification. DQAP1 (5'GCCTCTTACGGTGTAAACTTG-Y) and DQAP2 (5'-ATTGGTAGCAGCGGTAGAGTT-Y) were used for HLA-DQA1 gene amplification. HLA-DQB1 was amplified usingDQP1 and DQP2 as reported previously [ 2 ]. P C R - R F L P and PCR-ASO methods Amplified DNA was analysed by restriction fragment length polymorphism (RFLP) or allele specificoligonucleotide (ASO) dot blot methods. Polymorphism of NcoI site in TNF-[~gene has been pre L, NcoI negative; b 8, NcoI positive; ~ p < 0.05 (uncorrected); * and ** indicate p < 0.05 and p < 0.01, respectively. Frequencies of DQB1 allele were calculated as follows: when the result of polymerase chain reaction-allele specific oligonucleotide analysis was X/-, X was counted only once viously reported [ 3 ]. DNA fragment (497 base pairs; bp) including this polymorphic site was amplified and one tenth of the reaction mixture was digested with 10 units of NcoI (Takara shuzo, Kyoto, Japan) for 3 h at 37 ~ The digested sample was subjected to 2% agarose-gel electrophoresis and stained with ethidium bromide. Amplified HLA-DQA1 gene was digested with HaeIII and DdeI. Digested fragments were detected as described above. HLA-D QA1 genotyping (DQAI*I, *2, *3 and *4) was decided by the RFLP patterns as described in detail by Maeda et al. [ 4 ]. In brief, HaeIII digestion of D Q A I * I , *2, *3, *4 yields 44 + 76 + 117bp, 44 + 190bp, 237bp and 44 + 190bp, respectively. DQAI*2 and *4 can be classified by DdeI (DQAI*2; 98 + 136bp, DQAI*4; 107 + 127bp). HLA-DQB1 genotyping was decided using ASO probes (DQI.1, DQ1.2, DQ1.9, DQ1.AZH, DQ3.1, DQ3.2, DQ3.1-26, DQ3.3-26, DQ4 and DQ2 (5'-GCTGGGGCTGCCTGCCG-3')) [ 2 ]. DQwl.12 was identified using DQ1.9 and DQ3.1-26 probes. Hybridization and washing conditions are described elsewhere [ 2 ]. DNA sequencing Amplified DNA fragments were isolated after separation by 8% polyacrylamide gel electrophoresis and subcloned to the HincII site of pUC119 vector. M13 universal primers and PCR primers were used as sequencing primers in dideoxy methods. To exclude the PCR errors, three to six individual clones were sequenced following the WHO Nomenclature Committee recommendation. Statistical analysis The statistical differences were assessed by Fisher's exact test and the calculation was performed using SAS statistical analysis programme. TNF allele L/La L/Sb S/S K.Yamagata et al.:DQA1 gene and Japanese Type 1diabetes mellitus unexpected D Q A genotypes were also observed in a few control subjects (e. g. DR1/8-DQA1/3) on one allele, such unusual association to D Q A I * 3 was not found on both alleles in control subjects. Seventeen out of 22 (77%) Type 1 diabetic patients were homozygous for the D Q A I * 3 allele and 5 out of 22 (23%) were heterozygous. In contrast, only 11 of 32 (34%) control subjects were homozygous for the D Q A I * 3 allele. All Type i diabetic patients h a d D Q A l * 3 at least on one allele. Frequency of H L A - D Q A 1 alleles in Japanese Type i diabetic patients and control subjects are shown in the left lower part of Table 1. A1"3 was significantly increased (p = 0.000"*) and A1"1 was significantly decreased (p = 0.001"*) in Type 1 diabetes mellitus. Decreased frequency of A1"4 was observed in the patients, but was not significant. To further investigate whether the unique sequence variation in D Q A I * 3 gene could be observed or not in the Type 1 diabetic patients, amplified D N A fragments from three DR4/DR4, one D R l l / 1 3 (lane 5, Fig. 1B) and one DR8/14 (lane 6, Fig. 1B) Type 1 diabetic patients were sequenced but the unique sequence was not detected. Frequencies of DQB1 alleles are also shown on Table 1. When the result of PCR-ASO analysis was X/-, X was counted only once. Increased frequencies were observed in DQwl.1, DQw8 (DQw3.2), DQw9 (DQw3.3) and DQw4, but they were not significant in this study. On the other hand, D Q w l . 2 was significantly decreased (p = 0.040*) in Type i diabetes. Discussion TNF gene is located in H L A region (class III) and polymorphism of NcoI site in TNF[3 (not TNFcz) has been established [ 3 ]. Badenhoop et al. pointed out the increase of the heterozygosity of this site in Caucasian patients [ 5 ]. However, as shown in Table 1, the frequency did not differ between the control subjects and Type 1 patients in our study. Thus, this polymorphism itself seems to have little influence on manifestation of the disease in Japanese. Besides, amplified D N A sequences of the NcoI site negative type in two patients and one control subject were identical to the sequence which was previously published [ 3 ] and thus it was not unique to Type I diabetes. The participation of H L A - D Q A 1 to Type 1 diabetes has been suggested in Black [ 6 ], Japanese [ 7 ] and Caucasian subjects [ 8 ]. Particularly in Japanese subjects the increase of D Q A I * 3 allele and the decrease of D Q A I * 4 have been reported previously with P C R - A S O methods [ 7 ]. P C R - R F L P methods could type the D Q A 1 allele easily and rapidly with appropriate restriction endonucleases [ 4 ]. Moreover, whether the individual is homozygous or heterozygous can be decided by the patterns of digestion. Using this system, our results demonstrated the increase of D Q A I * 3 allele in Type i diabetic patients. In addition, this study has clearly shown for the first time in Japanese Type i diabetic patients, that the D Q A I * I allele was significantly decreased. The frequency of the D Q A I * 4 allele did not differ significantly between Type 1 diabetic patients and control subjects. So D Q A I * I seems to have a protective effect against Type 1 diabetes in Japanese subjects. The D Q A I * I gene is in linkage disequilibrium with DR1, 2, w6 and w8 [ 4 ]. The decrease of the D Q A I * 1 allele in patients may reflect the decrease of D R 2 in Type 1 diabetic patients. Sequence analysis revealed that D Q A I * 3 allelic sequence of Japanese Type 1 diabetic patients is identical to that reported previously [ 4 ]. Since D Q A I * 3 is in linkage disequilibrium with DR4 and DR9 [ 4 ], the high frequency of D Q A I * 3 may be accounted for by the linkage with D R 4 and DR9. However, we noticed two patients who have neither D R 4 nor DR9 but were homozygous for D Q A I * 3 (Fig. 1 B). Their D Q A types were reconfirmed by D N A sequencing. Though the observed patient number is only two and the unexpected association between D Q A I * 3 and non-DR4 or DR9 was detected on one allele of a few control subjects, our finding suggests the principal role of D Q A I * 3 gene in susceptibility to the disease. The decreased frequency of the D Q A I * I allele and increased frequency of the D Q A I * 3 allele strongly suggest that H L A - D Q A I * I contributes to resistance and A1"3 confers susceptibility to Type1 diabetes in Japanese subjects. DQB1 alleles were also investigated in a larger number of subjects than our previous report [ 2 ]. Increased frequencies were observed in DQwl.1 (non-Asp), DQw8 (non-Asp), DQw9 (Asp) and DQw4 (Asp) in Type 1 diabetes, although they were not significant. DQwl.2 (Asp) was significantly decreased (p = 0.040*) in Japanese patients as Awata et al. reported recently [ 9 ]. This protective effect of DQwl.2 was not observed by Todd et al. [ 7 ], who studied Japanese patients who developed the disease under the age of 16 years. The difference in the age of onset may be a possible explanation for the different incidence of DQwl.2, since most of our patients developed the disease at over the age of 16 years. In this study we have shown the importance of D Q A I * 3 in Japanese Type i diabetic patients. However, whether the D Q A I * 3 gene itself or an unknown gene in linkage with D Q A 1 is responsible remains to be resolved. Recently, a study with I-A transgenic N O D mice [ 10 ] indicated the significance of I-Acz (corresponding to human D Q a ) chain in the development of diabetes. This result supports that D Q A 1 gene itself is important for genetic susceptibility to Type i diabetes. Recently D Q molecules consisting of Arg 52 residuebearing ~ chain and non-Asp 57 residue-bearing [3chain were suggested as being susceptible in Caucasian patients [ 8 ]. In Japanese Type i patients, a combination of D Q A I * 3 (Arg52) c~chain and DQw4 and/or DQw9 (Asp 57) 13chain is predominant in our study. Although the combination of responsible c~-[3dimer is different, these data indicate the importance of considering the whole D Q molecule. Given the fact that class II M H C molecules can present antigens to T cells, three-dimensional configuration of a class II M H C c~-[3dimer is important for its interaction with antigens as well as to T cell receptors. In this context, future research should investigate the function of a class II M H C cz-[3dimer to clarify the pathogenic mechanism of Type I diabetes mellitus. Acknowledgements. We thank Dr. H.Inoko, Mr. M.Inoue, Dr. T.Yamasaki, Dr. M.Takenaka, Mr. K. Yamada for their advice and R e f e r e n c e s Mr. T.Tanaka for preparing the oligonucleotides. This study was supported by the Scientific Research Fund from the Ministry of Education, Science and Culture of Japan and the grant from Insulin Study Group. K. Yamagata et al.: DQA1 gene and Japanese Type 1 diabetes mellitus A n n o u n c e m e n t s Do It Diabetes Care Optimization Through Information Technology Study Group of the EASD This initiation and first annual meeting will be held on April 24-26, 1991 in Gubbio, Perugia, Italy. Those interested in participating the study group or the initiationworkshop are requested to submit ashort position statement (200 words) on their field of interest, previous work in the field and on their potential contribution to the workshop. For further information please contact." Prof. M. Massi-Benedetti, University of Perugia, Via Enrico dal Pozzo, 1-06100 Perugia, Italy. Correspondenceforthe Study Group:Dr. Dr. Klaus Piwernetz, Diabe tescenter Bogenhausen, 3.Med.Abt., Klinikum Miinchen-Bogenhausen, Englschalkingerstr.77, W-8000 Manchen 81, FRG. IDF Satellite Symposium: Controversies in Diabetic Neuropathy This symposium will be held on June 29-Juli 3, 1991 in New York. Chairman: Dr. J.D. Ward, Royal Hallamshire Hospital, Sheffield, Post EASD Symposium: Biochemistry and Biophysics of Insulin Secretion This symposium will be held on September 15-18, 1991 at the University of Ulcester, Coleraine, UK. Topics include: Metabolism, ion channels, calcium, phospholipids, nucleotides, protein kinases, biosynthesis, cell interactions, glucose toxicity, diabetic Beta-cell, actions of nutrients, peptides, neurotransmitters and drugs. For further details please contact: Prof. Peter R.Flatt, Biomedical Sciences Research Centre, University of Ulster, Coleraine, Northern Ireland, BT52 1SA, UK. Tel: ( + 44) 0265-44141, Fax: ( + 44) 026540906. 1. Todd JA , Bell JI , McDevitt HO ( 1987 ) HLA-DQ[3 gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus . Nature 329 : 599 ~ 04 2. Yamagata K , Nakajima H , Hanafusa T , Noguchi T , Miyazaki A , Miyagawa J , Sada M , Amemiya H , Tanaka T , Kono N , Tarui S ( 1989 ) Aspartic acid at position 57 of D QI3chain does not protect against Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects . Diabetologia 32 : 762 - 764 3. Nedospasov SA , Shakhov AN , Turetskaya RL , Mett VA , Azizov MM , Georgiev GR Korobko VG , Dobrynin VN , Filippov SA , Bystrov NS , Boldyreva EF , Chuvpilo SA , Chumak o v A M , ShingarovaLN , OvchinnikovYA ( 1986 ) Tandem arrangement of genes coding for tumor necrosis factor (TNF-a) and lymphotoxin (TNF-[3) in the human genome . Cold Spring Harbor Symposia on Quantitative Biology , Vol LI: 611 - 624 4. Maeda M , Murayama N , Ishii H , Uryu N , Ota M , Tsuji K , Inoko H ( 1990 ) A simple and rapid method for HLA-DQA1 genotyping by digestion of PCR-amplified D N A with allele specific restriction endonuclease . Tissue Antigens 34 : 290 - 298 5. Badenhoop K , Schwarz G , Trowsdale J , Lewis V , Usadel KH , Gale EAM , Bottazzo GF ( 1989 ) TNF-c~ gene polymorphisms in Type I (insulin-dependent) diabetes mellitus . Diabetologia 32 : 445 - 448 6. Todd JA , Mijovic C , Fletcher J , Jenkins D , Bradwell AR , Barnett AH ( 1989 ) Identification of susceptibilityloci for insulin-dependent diabetes mellitus by trans-racial gene mapping . Nature 338 : 587 - 589 7. Todd JA , Fukui Y , Kitagawa T , Sasazuki T ( 1990 ) The A3 allele of the HLA-DQA1 locus is associated with susceptibility to type 1 diabetes in Japanese . Proc Natl Acad Sci USA 87 : 1094 - 1098 8. Khalil I , D'Auriol L , Gobet M , Morin L , Lepage V , Deschamps I , Park MS , Degos L , Galibert F , Hors J ( 1990 ) A combination of HLA-DQ[3 Asp57-negative and HLA-DQc~ Arg52 confers susceptibility to insulin-dependent diabetes mellitus . J Clin Invest 85 : 1315 - 1319 9. Awata T , Kuzuya T , Matsuda A , Iwamoto Y , Kanazawa Y , Okuyama M , Juji T ( 1990 ) High frequency of aspartic acid at position 57 of HLA-DQl3-chain in Japanese IDDM patients and nondiabetic subjects . Diabetes 39 : 266 - 269 10. Miyazaki T , Uno M , Uehira M , Kikutani H , Kishimoto T , Kimoto M , Nishimoto H , Miyazaki J , Yamamura K ( 1990 ) Direct evidence for the contribution of the unique I-AN~ to the development of insulitis in non-obese diabetic mice . Nature 345 : 722 - 724


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2FBF00500386.pdf

K. Yamagata, T. Hanafusa, H. Nakajima, M. Sada, H. Amemiya, T. Noguchi, T. Tanaka, N. Kono, S. Tarui. HLA-DQA1*1 contributes to resistance and A1*3 confers susceptibility to Type 1 (insulin-dependent) diabetes mellitus in Japanese subjects, Diabetologia, 1991, 133-136, DOI: 10.1007/BF00500386