HLA and complement allotypes in Type 1 (insulin-dependent) diabetes

Diabetologia, Mar 1983

Summary A group of patients with Type 1 (insulin-dependent) diabetes mellitus was investigated for HLA-A, B and DR antigens as well as C4 and factor B polymorphism. A significant excess of DR3/DR4 heterozygotes was observed (27% versus 17% by Hardy-Weinberg expectation). The factor B allele BfF1 was present in 13% of patients with Type 1 diabetes (gene frequency of 0.08 versus 0.01 in control subjects). A rare C4 B allele, C4 B2.9, was found in 18% of patients with Type 1 diabetes (n=63) compared with 1.1% of control subjects (n=176). Total C4 deficiency at the C4A locus (C4AQ0,0) was present in 10% of patients with Type 1 diabetes compared with 0% of control subjects. Examination of HLA, C4 and Bf phenotypes in patients with Type 1 diabetes suggested that three high risk supratypes, HLA-A1 B8 BfS C4AQ0 C4 B1 DR3; HLA-B18 BfF1 C4A3 C4BQ0 DR3; HLA-A2 CW3 BW62 BfS C4A3 C4 B2.9 DR4 are markers for susceptibility alleles.

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%2FBF00250155.pdf

HLA and complement allotypes in Type 1 (insulin-dependent) diabetes

Diabetologia HLA and Complement Allotypes in Type 1 (Insulin-Dependent) Diabetes J. McCluskeyI 0 1 V. J. McCann 0 1 P. H. Kay 0 1 P. J. Zilko 0 1 F. T. Christiansen ~ 0 1 G.J. O'NeilP 0 1 R. L. Dawkins I 0 1 0 1Department of Clinical Immunology, Royal Perth Hospital and Queen Elizabeth II Medical Centre, /Diabetic Unit, Royal Perth Hospital , Perth , Western Australia and 3Department of Pathology, Columbia University , New York , USA 1 Professor R. L Dawkins Department of Clinical Immunology Royal Perth Hospital Wellington Street Perth , Western Australia 6000 Summary.A group of patients with Type 1 (insulin-dependent) diabetes mellitus was investigated for HLA-A, B and DR antigens as well as C 4 and factor B polymorphism. A significant excess of D R 3 / D R 4 heterozygotes was observed (27% versus 17% by Hardy-Weinberg expectation). The factor B allele BfF1 was present in 13% of patients with Type 1 diabetes (gene frequency of 0.08 versus 0.01 in control subjects). A rare C4B allele, C4 B2.9, was found in 18% of patients with Type 1 diabetes ( n = 6 3 ) compared with 1.1% of control subjects (n =176). Total C4 deficiency at the C 4 A locus (C4AQ0,0) HLA; complement allotypes; genetics; diabetes - 9 Springer-Verlag 1983 Immunogenetic influences conferring susceptibility to Type 1 (insulin-dependent) diabetes are well documented [ 1 ]. Increased frequencies of HLA-B8, B15, B18 and B40 have been reported. The primary associations appear to be with DR3 and DR4 with a disproportionate number of D R 3 / D R 4 heterozygotes suggesting interaction between alleles on opposing chromosomes. The likelihood that such susceptibility alleles do exist within the major histocompatibility complex is supported by associations with other genetic markers on chromosome 6 including properdin factor BfF1 [ 2, 3 ], the C2 allotype C2 B [ 4 ] and altered phenotype frequencies of the red cell enzyme glyoxalase [ 5 ]. The C4 loci arealso located within HLA and are SUfficiently polymorphic to help in the characterisation of the putative susceptibility loci involved in the pathogenesis of Type 1 diabetes. Recently we began an examination of those patients with potentially important supratypes (or combinations of alleles). The supratype marked by HLA-B15 (62) and DR4 was of particular interest because of its occurrence in rheumatoid arthritis as well as Type 1 diabetes [ 6 ]. More recently, Bertrams et al. [ 7 ] identified a rare C4B allotype, designated C4B4, in 26.5% of 223patients with Type 1 diabetes but only 7% of 385 healthy control subjects. We now report that this allotype is identical to that we have designated C4 B3 [6] and subsequently C4 B2.9 [ 8 ] and confirm that the B15, C4 B2.9, was present in 10% of patients with Type 1 diabetes compared with 0% of control subjects. Examination of HLA, C4 and Bf phenotypes in patients with Type 1 diabetes suggested that three high risk supratypes, HLA-A1 B8 BfS C4AQ0 C4B1 DR3; HLA-B18 BfF1 C4A3 C4BQ0 DR3; HLA-A2 CW3 BW62 BfS C 4 A 3 C4B2.9 DR4 are markers for susceptibility alleles. DR4 supratype is increased in frequency in Type 1 diabetes as well as rheumatoid arthritis. Furthermore, we have identified two other high risk supratypes which include B8 and B18. Materials and Methods A group of 82 unrelated patients with Type 1 diabetes mellitus, randomly selected from the diabetic clinic of the Royal Perth Hospital, Western Australia was studied. Type t diabetes was defined by a tendency to ketosis, non-obesity and absolute insulin requirement within at least 2 years of onset. The onset of diabetes was before the age of 40years in 64patients and after the age of 40years in the other 18. Healthy unrelated control subjects were selected from the West Australian town of Busselton. Both patients and control subjects were of European ancestry. HLA typing was performed using sera well characterised against cells typed in previous International Workshops. HLA-ABC typing was performed using the standard microlymphocytotoxicity method. DR typing was performed by a microcytotoxicity method using a nylon wool separated B cell enriched population [ 9 ]. Three complement components (properdin factor B (Bf), C2 and C4) are encoded by genes in the HLA complex. Current mapping data place these genes as a cluster between HLA-B and D/DR. Bf is controlled by codominant genes at a single locus, with two common alleles, F and S, and two rare alleles, F1 and $1. C4 is controlled by codominant genes at two distinct loci, Chido (C4 B) and Rodgers (C4 A). At the two loci many structural alleles and two quite frequent null alleles (C4AQ0 and C4BQ0) have been found. The polymorphisms of the complement components have been identified using immunoelectrophoretic techniques. Workshop for the Genetics of Complement held at Boston July 1982 this variant was distinguished from the C4 B3 variant described by Awdeh et al. and associated with 21 hydroxylase deficiency [ 13 ]. Allele frequencies in the disease groups were compared to those in the control groups using the Z 2test with Yates correction. In general it was not possible to assign genotypes and haplotypes as in family studies. The term 'supratype' is used to describe a particular combination of alleles at different loci. In previous studies we have shown that these supratypes are generally, if not always, inherited from one parent but we prefer to use 'haplotype' only when a family study has given unequivocal results. For convenience the three supratypes referred to here are abbreviated as follows: HLA-A2, CW3 BW62, BfS, C4A3, C4B2.9 DR4 (62/4-S-3-2.9); A1, CW7, B8, BfS, C4AQ0, C4B1, DR3 (8/3-S-0-1); B18, BfF1, C4A3, C4BQ0 DR3 (18/3-F1-3-0). These supratypes were assigned whenever all of the relevant complement alleles were represented in the phenotype together with at least one of the g or DR alleles. The only difficulty arose in the case of null alleles at the C4 loci. If there is a complete blank at one locus, homozygous deficiency may be assumed. If two allotypes are identified it can be assumed that null alleles cannot be present. If only one allotype (a) is demonstrated the genotype could be a, a; a, o or a, x where x is an unidentified allele. Distinction between a,a and a,o can often be achieved by comparing the density of the C 4 A and C4 B bands. Results As previously reported, significant increases in HLAB15, B18 and B40 in Type1 diabetes were found. H L A - D R antigens were examined in 64patients with Type 1 diabetes and the results are shown in Table 1. The previously d o c u m e n t e d increased frequencies o f HLA-DR3 and DR4 and decreased frequency o f HLADR2 in Type i diabetes compared with control subjects are again noted. When D R antigen combinations are studied the observed frequency of DR3, DR4 heterozygotes in Type 1 diabetes is more than predicted for Hardy-Weinberg equilibrium (27% versus 17%) and nearly five times that observed in the control population (27% versus 6%, p < 0.001). Properdin factor B allotype frequencies in Type I diabetes are shown in Table 2. The unusual allotype BfF1 was present at a gene frequency o f 8.0% in Type 1 diabetes compared with 1% in control subjects and the frequency of BfF1 was similar in those with onset greater than and less than 40 years (Table 2). Two BfF1 homozygous patients were identified both of onset less than 40 years. Among the 11 patients with a BfF1 phenotype, nine were HLA-B18 positive and blanks were present in the remaining two patients. HLA-DR3 was present in all nine o f the BfF1 positive patients who were D R typed. These findings are consistent with the presence of a 'diabetogenic' supratype m a r k e d by B18, BfF1 and DR3. 3lae frequencies o f C 4 A and C4B phenotypes in Type 1 diabetes versus control subjects are shown in Table 3. The most salient finding is the presence o f the rare C4B allotype designated by us as C4B2.9 and found in 18% of Type I diabetic patients versus 1.1% of control subjects (p < 0.001). The variant C4B2.9 migrates electrophoretically slightly cathodal to the C4 B3 found in association with 21 hydroxylase deficiency and HLA-BW47. 1@ 79 24 80 30 18a Properdin factor B typing was performed on serum samples subjected to high voltage electrophoresis on agarose gel medium followed by immunofixation with antiserum to factor B [ 10 ]. C4 typing was performed using the immunofixation electrophoresis system of Awdeh and Alper [ 11 ] and the glycine-barbital buffer system described by O'Neill et al. [ 12 ]. Briefly EDTA plasma samples were frozen to - 70 ~ within 1-2 h of collection. Before electrophoresis, samples were treated overnight at 4 ~ with neuraminidase from Clostridium perfringens (Sigma, St. Louis, Missouri, USA). The treated plasma was subjected to high voltage electrophoresis until an HbS1 marker migrated 7 cm. The gel was then overlaid with anti-C4 (Atlantic/ Scarborough, Maine, USA), pressed, washed and stained with Coomassie brilliant blue. Characteristic bands were identified corresponding to the gene products of the two loci C 4 A and C4B. Standards were included on each run and assignments made by two independent observers. On occasions the bands of the C 4 A and C4B loci may overlap. The C4B locus products have greater haemolytic activity than those of the C4 A locus. Therefore a haemolytic overlay technique was used to differentiate overlapping bands. Allotypes were numberically designated C4 A1-7 and C4 B1 7 using the notation of Awdeh and Alper [ 11 ]. An unusual C4 B variant, designated by us as C4 B2.9, was identified. Exchange of sera has confirmed that this variant is identical to the C4 B4 variant described by Bertrams et al. [ 7 ]. At the Fourth International 62/4-S-3-2.9 Supratype When HLA and Bf were examined in the 11 C4B2.9positive diabetic patients, all were found to carry BfS. Eight were B15 (BW62) positive and seven were DR4 positive with D R blanks in some of the remaining cases. Accordingly 10 of these 11 patients had the 62/4-S-3-2.9 supratype as defined above. 8/3-S-0-1 Supratype From Table 3, it can also be seen that there is a significant increase in the frequency of apparent C4AQ0 homozygotes in Type l diabetes (10% versus 0%, p < 0.001). HLA phenotypes of all six C4 AQ0,0 patients with Type 1 diabetes included HLA-A1, B8, C4 B1, BfS, DR3 and three of the six appeared to be homozygous at all loci. These findings suggest the HLA-A1, B8, DR3 supratype found in Type I diabetes can now be further defined by including C4AQ0, C4 B1 and BfS. 18/3-F1-3-0Supratype Examination of two presumed BfF1 homozygous patients with Type 1 diabetes disclosed the phenotype B18, C4A3, C4BQ0, BfF1, DR3 in both cases. In addition the other BfF1 positive patients with Type I diabetes were B18, C4A3, DR3 and apparently C4BQ0. This suggests that a third supratype 18/3-FI-3-0, is also important. Discussion Previously recognised HLA associations with Type 1 diabetes are again confirmed. In particular, the frequency of DR3, DR4 heterozygotes in Type 1 diabetic patients is nearly five times that found in control subjects (p < 0.01) and significantly increased even for HardyWeinberg expectations within Type 1 diabetes. This agrees with other reports and suggests that susceptibility to Type 1 diabetes involves interaction of major histocompatibility region loci on opposing chromosomes. By helping to map this region, complement markers may help determine whether any such interaction involves identical or differing alleles on opposing chromosomes. Complement allotyping of the C4 and Bf loci has allowed the definition of three supratypes associated with Type 1 diabetes, namely 8/3-S-0-1, 62/4-S-3-2.9 and 18/3-F1-3-0. It seems probable that these supratypes are in linkage with a putative disease locus. Current genetic models o f the inheritance of Type 1 diabetes usually assume that susceptibility alleles at the putative disease locus are identical. However, there is evidence that supratypes may have differing genetic effects in relation to disease pattern. The 18/3-F1-3-0 supratype may have an association with younger age of onset [ 14, 15 ] though this is not found in all studies [4] J.McCluskey et al.: C4, Bland HLA in Type I Diabetes and is not obvious from our data. On the other hand, the supratype 62/4-S-3-2.9 may be associated with high titre anti-insulin antibodies [ 16 ] and low prevalence of islet cell antibodies [ 17 ] while 8/3-S-0-1 probably correlates with low titre anti-insulin [ 16 ] and persistent antiislet cell antibodies [ 17 ]. Similarly associations with other diseases and each supratype are distinct. Systematic lupus erythematosus is associated with the supratype 8/3-S-0-1 [ 18 ] while idiopathic membranous nephropathy is associated with 18/3-F1-3-0 [ 19 ]. In a recent study of a large group o f patients with rheumatoid arthritis we have demonstrated the 62/4-S-3-2.9 supratype in approximately 10% o f patients compared with 1% of control subjects [ 6, 8 ]. Family studies in rheumatoid arthritis indicate 62/4-S-3-2.9 is a haplotype [20]. These findings may explain the observation of an increased prevalence of rheumatoid arthritis in relatives of patients with Type 1 diabetes and vice versa [ 21 ]. Thus, apparently differing clinical and immunological associations of the supratypes found in Type 1 diabetes and their overlap with separate immunopathic diseases suggests that linkage with identical disease susceptibility alleles cannot be assumed. Furthermore, examination of the full phenotypes of our patients reveals one or more of the defined supratypes in two-thirds of cases. Indeed, the observed excess of DR3, DR4 heterozygotes largely reflects assortment of these supratypes. Identical susceptibility alleles on all supratypes would not predict an excess of DR3, DR4 heterozygotes over DR3 or DR4 homozygosity as has been documented [ 22 ]. Thus the data are most consistent with interaction of different diabetogenic susceptibility alleles associated with discrete supratypes of the major histocompatibility complex. Acknowledgements. This work was supported by grants from the Western Australian Arthritis and Rheumatism Foundation and the National Health and Medical Research Council of Australia. Publication 8226 of the Departments of Clinical Immunology, Royal Perth Hospital and The Queen Elizabeth II Medical Centre, Perth, Western Australia. 1. Cudworth AG ( 1978 ) Type 1 diabetes mellitus . Diabetologia 14 : 281 - 291 2. Raum D , Alper CA , Stein R , Gabbay KH ( 1979 ) Genetic marker for insulin-dependentdiabetes mellitus . Lancet 1 : 1208 - 1210 3, Dornan J , Allan P , Noel PE , Larson B , Farid NR ( 1980 ) Properdin factor B (Bf) allele BfF1 specifies an HLA-B18 diabetogenic haplotype . Diabetes 29 : 423 - 427 4. Wolf E , Cudworth AG , Markwick JR , Wells L , Bodansky HJ , Gorsuch AN , Spencer KM , Lister J ( 1981 ) Bf and C2 complement factors in Type I diabetes . Diabetologia 20 : 667 (Abstract) 5. McCann VJ , Davis RE , Welborn TA , Constable IJ , Beale DG ( 1981 ) Glyoxalase phenotypes in patients with diabetes mellitus . Aust N Z J Med 11 : 380 - 382 6. Kay PH , McCluskey J , Christiansen FT , Feeney D , McCann VJ , Zilko PJ , Dawkins RL , O'Nefl GJ (t982) Complement allotyping reveals new genetic markers in rheumatoid arthritis and diabetes mellitus . Tissue Antigens (in press) 7. Bertrams J , Hintzen U , Schlicht V , Schoeps S ( 1982 ) C4: another marker for Type I diabetes . Lancet 1 : 41 8. O 'Neill GJ , Nerl C , Kay PH , Christiansen FT , McCluskey J , Dawkins RL ( 1982 ) Complement C4 is a marker for adult rheumatoid arthritis . Lancet 2 : 214 9. Danilovs JA , Ayoub G , Terasaki PI ( 1980 ) B lymphocyte isolation by thrombin - nylon wool . In: Terasaki PI (ed) Histocompatibility testing 1980 . UCLA Tissue Typing Laboratory, Los Angeles, California, pp 287 - 288 10. Alper CA , Boenisch T , Watson L ( 1972 ) Genetic polymorphism in human glycine-rich beta-glycoprotein . J Exp Med 135 : 68 - 80 11. Awdeh ZL , Alper C ( 1980 ) Inherited structural polymorphism of the fourth component of human complement . Proc Natl Acad Sci USA . 77 : 3576 - 3580 12. O 'Neill GJ , Yang SY , Dupont B ( 1978 ) Two HLA-linkedloci controlling the fourth component of human complement . Proc Natl Acad Sci USA . 75 : 5165 - 5169 13. Awdeh ZL , Raum D , Fleischnick E , Crigler JF , Gerald PS , Alper CA ( 1981 ) MHC-linked complement haplotypes (complotypes) in congenital adrenal hyperplasia . Clin Res 29 : 287 (Abstract) 14. Kirk RL , Sergeantson SW , Theophilus J , Zimmet P , Whitehouse S , Court JM ( 1979 ) Age relationship between insulin-dependent diabetes mellitus and rare alleles of properdin factor B . Lancet 2 : 537 15. Bertrams J , Baur MP , Grfineklee D , Giles FA ( 1981 ) Age related association of insulin-dependentdiabetes mellitus with BfF1 and the HLA-B18, BfF1 haplotype . Diabetologia 21 : 47 - 49 16. Bertrams J , Jansen FK , Grfineklee D , Reis HE , Drost H , Beyer J , Gries FA , Kuwert E ( 1976 ) HLA antigens and immunoresponsiveness to insulin in insulin-dependentdiabetes mellitus . Tissue Antigens 8 : 13 - 19 17. Lendrum R , Walker G , Cudworth AG , Woodrow JC , Gamble DR ( 1976 ) HLA-linked genes and islet cell antibodies in diabetes mellitus . Br Med J 1 ; 1565 - 1567 18. Christiansen FF , Uko G , Dawkins RL , McCluskey J , Zilko PJ ( 1982 ) Complement allotyping in systemic lupus erythematosus: association with C4A null . In: Dawkins RL , Christiansen FT , Zilko PJ (eds) Immunogenetics in rheumatology . Excerpta Medica , Amsterdam pp 229 - 234 19. Rittner Ch , Bertrams J ( 1980 ) On the significance of C2, C4 and factor B polymorphisms in disease . Hum Genet 56 : 235 - 247 20. McCluskey J , Kay P , Christiansen FT , Zilko PJ , Dawkins RL ( 1982 ) Complement C4 allotypes and rheumatoid arthritis . In: Dawkins RL , Christiansen FT , Zilko PJ (eds) Immunogenetics in rheumatology . Excerpta Medica , Amsterdem pp 121 - 126 21. Thomas D , Young A , Bottazzo GF , Cudworth AG ( 1982 ) Type 1 (insulin-dependent)diabetes and rheumatoid arthritis: A common link ? Ann Rheum Dis (in press) 22. Barbosa J , Bach FH , Rich SS ( 1982 ) Genetic heterogeneity of diabetes and HLA . Clin Genet 21 : 25 - 32 Received: 7 June 1982 and in revised form: 15 November 1982


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2FBF00250155.pdf

J. McCluskey, V. J. McCann, P. H. Kay, P. J. Zilko, F. T. Christiansen, G. J. O'Neill, R. L. Dawkins. HLA and complement allotypes in Type 1 (insulin-dependent) diabetes, Diabetologia, 1983, 162-165, DOI: 10.1007/BF00250155