Metabolism Clinical and E~~i~~tal VOL. XXVIII, NO. 8

Ia-Type

AUGUST

1979

Alloautigeus and Humoral Autoimmune Responsiveness in Insulin-Dependent Diabetes Mellitus H. Ludwig, G. Schernthaner, and W. R. Mayr

In the search for markers either closely linked to or identical with the hypothetical “diabetogenic major hiitocompatibility gene.” immune region-associated all~ntigens were defined in 80 patients with insulin-dependent diabetes and in 107 controls. A close association between the la-type alloantigen DRw3 and DRw4 and insulin-dependent diabetes was obtained. DRw3 was found in 36% and DRw4 in 32% of the patients compared to 11% and 16%. respectively, of the controls. In addition, a significant influence of DRw3 and humoral anti-islet-cell autoimmunity could be observed, which was found to be due to a high incidence of DRw3 in those patients with islet cell antibody persistence. Isletcell antibodies (ICA) were observed in 60% of the DRwB-positive pstients compared to only 9% of the Dl?w3-negatives with longstanding disease (>5 yrj. These data show a significant association between insulin-dependent diabetes and the la-type alloantigens DRw3 (p uncorr. c: 0.0005) and DRw4 (p uncorr. < 0.025). Furthermore, they provide direct evidence of an association between an la-type alloantigen and persisting humoral autoimmune responsiveness in man.

IN THE

STUDY of the clinical and biologic

importance

of the major

histocompatibility

(MHC) in humans, main interest is focused on the human equivalents of “immune region-associated fIa) antigens.” These antigens are thought to be of major importance in the control of immune responsiveness, susceptibility, and resistance to certain diseases.’ As the linking of a certain disorder to la-type alloantigens is expected to reveal significant clues regarding its underlying ~thomechanism, it seems promising to concentrate on insulin-dependent diabetes mellitus (IDD). This type of diabetes (Type I) is a relatively severe disease and is assumed to be complex

Met&olkn, Vd. 28, No. 8 (August),1979

the result of genetic, viral, and/or autoimmune interactions.2’3’4 Only during the past 3 yr could the inherited susceptibility to develop IDD in certain individuals, at least in part, be delineated to the expression of particular genetic markers. Thus, individuals positive for HLA-B8, B15, Cw3, Dw3, and Dw4 carry an increased risk to develop IDD,s.6 compared to those negative for these genes. By contrast, the HLA-B5 or B7 genes even seem to have some protective effect in the development of IDD.7*sThe clinical significance of these findings was manifold. Besides their role as diabetes-predis~sing factors, HLA genes were found to influence insulin antibody formation’ and islet-ceil antibody (ICA) persistence.” This first evidence of possible interrelations between MHC genes and immune responsiveness in IDD stimulated intensive search for markers playing an even morq important role in the pathophysiology of that disorder. Consequently we decided to use the newly defined immune region-associated antigens to evaluate their possible importance in humoral autoimmune responsiveness in insulin-dependent diabetes. Ia alloantigens are coded for by a locus From the Department of Medicine II, Institute ofBlood Group Serology, University of Vienna, and the Ludwig Bolrzmann Institute for Clinical Endocrinology and Nuclear Medicine, Yienna. Austria. Address reprint requests to Dr. H. Ludwig, Second Department of Medicine, University of Vienna. Garnisongasse 13, A-1090 Wien. Austria. o I979 by Grune & S{[atton, Inc. a1260495/-79/2sos-alIt$OX.~~O 797

798

LUDWIG,

closely linked to HLA-D (and possibly HLA-D itself) and have been designated tentatively as DRw (D related; w workshop indicates a yet provisional definition) antigens by the World Health Organization (WHO) nomenclature committee. ” MATERIALS AND

METHODS

The x2 test was applied for comparison of DRw frequenin patients and controls. However, correction for the number of comparisons made was not performed, since this study was part of an international workshop, which produced similar results in a higher number of patients.” Fisher’s exact test was used for evaluation of the association between DRw3 and ICA; the relative risk was calculated by the formula given by Woolf.” cies

RESULTS

la-type alloantigens, HLA-A, R, and C antigens and islet cell antibodies were studied in 80 patients (36 males, 44 females) with insulin-dependent diabetes (mean age at onset 20.1 years; mean duration of disease 10.8 years) and 107 controls of comparable age and sex. In an additional 141 insulin-dependent diabetics and 114 controls, only HLA-A, B, and C antigens and islet cell antibodies were determined. These individuals were selected for their comparability to the groups examined and described above.

Tissue Typing The sera distributed and the lymphocytotoxicity assay as recommended for the seventh histocompatibility workshop” were used for Ia typing on B-lymphocyte enriched cell suspensions, which were obtained by resetting peripheral blood lymphocytes with neuraminidase-treated sheep red blood cells. HLA-A, B, and C locus typing was performed by means of the standard NIH microlymph~ytotoxicity technique.

Islet-Cell Antibodies The indirect immunofluorescence technique was used to demonstrate islet cell antibodies. Several human and monkey pancreas glands were tested for optimal suitability as antigens. Finally, a human blood group 0 pancreas taken from a 4-yr-old kidney donor was selected and used in all further experiments. All sera were tested undiluted on unfixed 5 p tissue sections as described previously.‘* A TRITC-labeled F(ab)z anti-IgC conjugate, which fubilled optimal specificity and sensitivity criteria, kindly donated by Dr. Baudner (Behringwerke; Marburg/Lahn), was used. All specimens were screened independently by 2 investigators in a Leitz Ortholux II microscope (equipped with a Pleompaque II). In case of equivocal results, serum specimens were retested and judged according to a scoring system. Table 1. Phenotype

‘NS,

AND MAYR

Statistical Analysis

Patients

IeTvpa Altoantigen

SCHERNTHANER,

Frequencies IDD fn - 80)

DRwl

5 (6%)

DRw2

9 112%)

of la-Type

Allwntigens

COWidS tn - 107)

Frequencies of la-type alloantigens observed in insulin-dependent diabetics and controls are listed in Table 1. A significantly increased frequency of DRw3 (p uncorr. < 0.~5) and DRw4 (p uncorr. < 0.025) was found in the patients. Individuals positive for DRw3 carry a 4.5-fold risk and those positive for DRw4 a 2.9-fold increased risk to develop IDD. Islet-cell antibodies were found in 30 (38%) of the 80 ia-typed patients and in only 2 of the 107 controls. These antibodies were detected to be closely associated with DRw3 positivity in insulin-de~ndent diabetes: 59% of the DRw3positive patients were ICA-positive, compared to 25% of the DRw3-negatives (p < 0.005). The evaluation of the distribution of ICA according to the prevalence of DRw3 and the duration of disease (Fig. -1) revealed a high incidence of ICA (60%) in DRw3-positive and low frequency of these antibodies (9%) in DRw3negative patients with long duration of disease (>5 yr; p < 0.001). In the same 55 DR typed patients with long-lasting disease, the respective distribution of ICA was 45% in the HLA-38positive and 17% in the HLA-B&negative patients. Likewise, in the 133 patients with longlasting diabetes, of the total 221 HLA-A, B, and C locus typed patients, the prevalence of ICA was 37% in the HLA-B8-positive and 22% in the HLA-B&negative patients. However, in Insulin-Dependent

X2

Diabetics (IDD) and Controls

P wlcan)

Relative Risk

4 (4%)

0.63

NS

1.72

15 (14%)

0.30

NS

0.73

16.76

to.OoD5

4.50

DRw3

29 (36%)

12(11%)

DRw4

25 (31%)

17 (16%)

9.36

to.025

2.95

DRw5

10 (13%)

25 VZ3%)

3.65

NS

0.47

DRw6

2 (3%)

9 (8%)

2.89

NS

0.28

DRw7

14 (18%1

22 (21%)

0.28

NS

0.82

not significant.

la-TYF’E ALLOANTIGENS

IN IDD

Fig. 1. Incidence of ICA in DRw3-positive and DRw3-negative insulin-dependent diabetes analyzed irrespective of duration of disease (n = W and in patients with short fn = 251 and longlasting diabetes fn = 55).

when all A, B, and C locus typed patients (n = 221) irrespective of duration of disease were analysed, only a weak tendency for increased incidence of ICA in HLA-B8 positives was found. In the 25 diabetics with short duration of disease (t5 years), the incidence of ICA neither differed significantly between DRw3-positive (56%) and negative (62%) nor between HLABS-positive (77%) and HLA-B&negative (56%) diabetics. In the corresponding group of the 221 IDDs, no significant association between B8 and ICA was observed, Furthermore, humoral antiislet-cell immunity was not associated with DRw4 or any of the other HLA genes tested for. Five of the 221 A, B, and C typed controls were ICA-positive, and 2 of them were HLA-B8positive. DISCUSSION

The present investigation demonstrates increased frequencies of the Ia-type alloantigens DRw3 (p uncorr. < 0.0005) and DRw4 (p uncorr. i: 0.025) in insulin-dependent diabetes. From these data, a 4.5-fold increased risk to develop IDD can be calculated in DRw3positive individuals compared to those subjects who do not carry this antigen. This figure is higher than the magnitude of the relative risk calculated for the other IDD-associated HLA-B and C locus genes B8, B15, and Cw3, but is similar to that reported for the D locus gene Dw3.16 However, following the mode of analysis suggested by Svejgaard and RyderiS and testing for the gene primarily associated with the hypothetical diabetes-predisposing gene(s), only a slight predominance of DRw3 over B8 was found. Furthermore, the definite relationship

between the serologically defined Ia alloantigens and the HLA-D determinants, which previously have been reported to be also very closely associated with IDD,u’ remains to be determined. This includes the possibility that DRw3 is not only in close linkage disequilibrium with the D locus determinant Dw3 but is supposedly identical with it. In whatever way, the HLA antigens associated with IDD do not seem to predispose by themselves to diabetes but presumably are in linkage dis~uilibrium with diabetogenic and/or immune response genes. Further analysis of our data revealed a significant association between DRw3 and humoral anti-islet-cell autoimmunity that proved to be due to the high incidence of ICA in DRw3 positive patients with long duration of disease (>5 yr). In evaluating the interrelation between HLA-B8 and ICA-positivity in these patients a weaker association was noted. In the group of the 221 HLA-A, B, and C typed IDD patients, a positive association between incidence of ICA and B8 was also found in those patients with long duration of disease (~5 yr). However, no such association could be obtained by analyzing these 221 patients without regarding the interval between onset of IDD and determination of ICA. These findings indicate a positive influence of HLA-genes on islet-cell antibody persistence, which is in accordance with data already reported,” but beyond that points to DRw3 as the strongest associated factor in determining longlasting antiislet cell autoimmunity. The incidence of ICA in the controls studied is too low to allow any conclusions regarding the role of DRw3 (and B8) as denominators for increased humoral autoimmune responsiveness in healthy individuals. However, on the basis of

LUDWIG,

800

the observations in diabetics, a correlation between this gene(s) and ICA is to be expected even in subjects otherwise asymptomatic. Since a su~tantial influence of the immune region-associated gene DRw3 on islet cell antibody formation seems apparent, further elucidation of the clinical and pathogenic significance of these antibodies is of interest. Up to now, it remains unknown whether ICA are the cause or rather the result of Langerhans islet tissue destruction, However, careful investigations of the ability of serum factors and lymphocytes to lyse cultured human insulinoma cells revealed in only a very few cases a tissue-damaging activity of serum components in anti~y-de~ndent cell-mediated cytotoxicity.” Further proof against the direct pathogenic effect of islet cell antibodies comes from studies of twins and sporadic observations of newborns of ICA-positive diabetic mothers.18 IDD was neither found regulariy in the ICA-positive cotwin nor in children with ICA transferred transplacentally. However, ICA-positive patients on oral hypogly-

SCHERNT~ANER,

AND MAYR

cemic agents or ICA-positive patients with autoimmune polyendocrine disorders after having been observed for a considerable length of time are reported to develop IDD.” Thus, ICA might conceivably be a marker for asymptomatic and latent diabetes and might predict the occurence of diabetes in the future.” In our opinion, these antibodies might prove to be a hallmark for ongoing inflammation of the islets of Langerhans, which becomes of clinical significance only when destruction of more than 90% of beta-cells occurs. In conclusion, a significant association between insulin-dependent diabetes and the Iatype alloantigen DRw3 (p uncorr. < 0.0005) and DRw4 (p uncorr. < 0.025) is reported in the present investigation. Furthermore, DRw3 was found to be significantly increased in those patients with islet antibody persistance and long duration of disease (>5 yr). These data seem to provide direct evidence of an association between Ia-type alloantigens and persisting humoral autoimmune responsiveness in man.

REFERENCES 1. Bach FH, van Rood JJ: The major hist~ompatibility complex. N Engl J Med 295806-8 13, 1976 2. Irvine WJ: Classification of idiopathic diabetes. Lancet 1:638-641,1977 3. Nerup J, Andersen 00, Christy P, et al: HLA, autoimmunity, virus and the pathogenesis of juvenile diabetes mellitus. Acta Endocr (Suppl Kbh) 205167%176,1976 4. Maugh TH If: Diabetes: epidemiology suggests a viral connection. Science 188:347-351.1975 5. Nerup J, Platz P, Andersen 00, et ai: HLA and diabetes meltitus. Lancet 2:864-865, 1974 6. Schernthaner G, Mayr WR, Pacher M, et al: HLA-AS, W15 and T3 in juvenile onset diabetes mellitus. Horm Metab Res 7:521-522,197s 7. Ludwig H, Schernthaner G, Mayr WR: The importance of HLA genes to sus~ptibility in the development of juvenile onset diabetes mellitus. A study of 93 patients and 68 first degree blood relations. Diabete et Metabolisme 3:43-48, 1977 8. Mayr WR, Ludwig H, Schernthaner G, et al: Ia serology, la antigens in insulin-dependent diabetes mellitus and HLA-ABC serology: a cooperative study of the BaselWien region. Hist~mpatibiiity Testing. Munksgaard, Copenhagen, 1977, p 39 I 9. Schernthaner G, Ludwig H, Mayr WR, et al: Genetic factors on insulin antibodies in juvenile onset diabetes. N Engl J Med 295:622,1976 10. Irvine WJ, McCallum CJ, Gray RS, et al: Pancreatic islet-cell antibodies in diabetes mellitus correlated with the

duration and type of diabetes, coexistent autoimmune diseaseand HLA type. Diabetes 19:138-147, 1977 1I. Bodmer WF, et al: Histocompatibility Testing 1977 (eds): Copenhagen, 1978, p 14 12. Ludwig H, Schernthaner G, Mayr WR: Inselzellantik&per beim juvenilen Diabetes mellitus. Dtsch Med Wschr 102:1221-1226.1977 13. Batchelor SR. Morris PJ: HLA and disease, in Histocompatibility Testing 1977. Bodmer WF, Batchelor JR, Bodmer JG, et ai (eds): Copenhagen, 1978, p 205 14. Woolf B: On estimating the relation between blood groups and disease. Ann Hum Genet 19:251-259, 1955 15. Svejgaard A, Ryder LP: Associations between HLA and disease, in Dausset J, Svejgaard A (eds): HLA and Disease. Munksgaard, Copenhagen, 1977 16. Thomsen M, Platz P, Anderson 00, et al: MLC typing in juvenile diabetics and idiopathic Addison’s disease. Transplant Rev 22: 125-l 47, 1975 17. Huang SW, MacLaren NK: Insulin-dependent diabetes: a disease of autoaggression. Science 192:64-66, 1976 18. Gamien TR. Aynsley-Green A, Irvine WJ, et al: Immunological studies in the neonate of a mother with Addison’s disease and diabetes mellitus. Clin Exp Immunol 28:192-195, 1977 19. Irvine WJ, Gray RS, McCallum CJ: Pancreatic islet cell antibodies as a marker for asymptomatic and latent diabetes and praediabetes. Lancet 2:1097-l 102.1976

Ia-Type alloantigens and humoral autoimmune responsiveness in insulin-dependent diabetes mellitus.

Metabolism Clinical and E~~i~~tal VOL. XXVIII, NO. 8 Ia-Type AUGUST 1979 Alloautigeus and Humoral Autoimmune Responsiveness in Insulin-Dependent D...
526KB Sizes 0 Downloads 0 Views