Failure to detect genomic viral sequences in pancreatic tissues from two children with acute-onset diabetes mellitus

Two cases of fatal, acute-onset, insulin-dependent diabetes mellitus (IDDM) in children were diagnosed. Epidemiologic and serologic studies, as well as histologic analysis of pancreatic tissue in fatal viral infections, support the contention that a viral infection could cause beta cell destruction, leading to IDDM. The presence of nucleic acid sequences from viral agents considered to be potentially diabetogenic, specifically, cytomegalovirus and mumps, rubella, and coxsackie viruses, were investigated in the pancreatic tissues by reverse transcription followed by polymerase chain reaction (RT-PCR) and Southern blot hybridization. Total pancreatic RNAs extracted from five children who died from nondiabetic causes were included as controls. Viral genetic information from any of these four viral agents was not found. This result indicates that the acute IDDM in these cases was not due to a direct infection of pancreatic beta cells by any of the viral agents studied.

Identification of autoantibody epitopes of glutamic acid decarboxylase in stiff-man syndrome patients

Stiff-man syndrome is a neurologic disorder characterized by progressive rigidity of skeletal muscles. Deficiency of the neurotransmitter gamma-aminobutyric acid and autoantibodies to glutamic acid decarboxylase (GAD), the enzyme synthesizing gamma-aminobutyric acid, are closely associated with the disorder, although the relevant antigenic epitopes have not been identified. In the present study, sera from two patients with SMS was used in an immunoblotting assay with recombinant GAD67 (M(r) 67,000) and GAD65 (M(r) 65,000) isoforms to test whether SMS sera can recognize specific epitopes. We found that both SMS sera recognized the GAD65, but not the GAD67, isoform. Using 13 different synthetic GAD peptides to block the autoantibodies, two GAD65 epitopes were identified. One epitope recognized by both patients' sera, was blocked by the peptide representing amino acid residues 354-368. In one patient only, blocking was also observed by a peptide representing residues 390-402, which includes the binding site of the GAD cofactor, pyridoxal 5'-phosphate. A single amino acid substitution in GAD65 at position 401 (leucine to proline) and representing the analogous GAD67 sequence in this region significantly reduced the peptide's inhibitory effect. These findings suggest that SMS GAD autoantibodies share distinct GAD65 linear epitopes and that some SMS patients' autoantibodies may block the active site, explaining SMS GABA deficiency.

Identification of autoantibody epitopes of glutamic acid decarboxylase in stiff-man syndrome patients

Stiff-man syndrome is a neurologic disorder characterized by progressive rigidity of skeletal muscles. Deficiency of the neurotransmitter gamma-aminobutyric acid and autoantibodies to glutamic acid decarboxylase (GAD), the enzyme synthesizing gamma-aminobutyric acid, are closely associated with the disorder, although the relevant antigenic epitopes have not been identified. In the present study, sera from two patients with SMS was used in an immunoblotting assay with recombinant GAD67 (M(r) 67,000) and GAD65 (M(r) 65,000) isoforms to test whether SMS sera can recognize specific epitopes. We found that both SMS sera recognized the GAD65, but not the GAD67, isoform. Using 13 different synthetic GAD peptides to block the autoantibodies, two GAD65 epitopes were identified. One epitope recognized by both patients' sera, was blocked by the peptide representing amino acid residues 354-368. In one patient only, blocking was also observed by a peptide representing residues 390-402, which includes the binding site of the GAD cofactor, pyridoxal 5'-phosphate. A single amino acid substitution in GAD65 at position 401 (leucine to proline) and representing the analogous GAD67 sequence in this region significantly reduced the peptide's inhibitory effect. These findings suggest that SMS GAD autoantibodies share distinct GAD65 linear epitopes and that some SMS patients' autoantibodies may block the active site, explaining SMS GABA deficiency.

Weight loss in rats following intraventricular transplants of pancreatic islets

Because of the body's resistance to permanent weight change, obesity remains a major health problem in modern society. It is hypothesized that the regulatory system defending body weight utilizes pancreatic insulin as an indicator of adiposity to the brain. To take advantage of this negative feedback system, we transplanted neonatal (experiment 1) or adult (experiment 2) pancreatic islets containing insulin-secreting cells into the 3rd ventricle of syngeneic Lewis rats. This resulted in an elevation of the insulin signal within the brain and a significant long-term reduction of body weight. Changes in food intake were consistent with the changes of body weight. The implantation of more islets resulted in a greater reduction of body weight, and changes in weight were inversely correlated with the level of insulin achieved in the cerebrospinal fluid. After the two studies were completed, histological examination revealed the presence of insulin-containing cells within the 3rd ventricle and adjacent hypothalamus. These studies suggest that transplanted insulin-secreting cells may provide a potential therapeutic strategy for maintenance of weight loss.

HLA DR and DQ RFLP analysis in Crohn's disease

A study of 109 Swedish patients and 85 healthy Swedish controls with Crohn's disease (CD) by HLA class II RFLP genotyping was carried out. There was no significant association for any single DR or DQ specificity or phenotypic combination of DR and/or DQ specificities among our study group of Caucasian extraction.

Decreased weight gain in BB rats before the clinical onset of insulin-dependent diabetes

Inbred specific pathogen-free diabetes-prone (DP) and diabetes-resistant (DR) BB rats were crossed to produce F1 and intercrossed to produce F2 rats. Diabetes segregates in these crosses as a recessive trait on rat chromosome 4. The weight gain of genetically diabetes-prone rats born to F1 healthy parents was studied to avoid effects of maternal diabetes. The weight gain of the F2 rats was initially not different from the F1 parents. The F2 rats later developing diabetes grew in parallel with their non-affected siblings up until the last 9 days before onset. During these 9 days they showed a decreased weight gain compared to their healthy litter-mates regardless of age. We conclude that decreased weight gain precedes the abrupt clinical onset of diabetes in BB rats and that it may be due to processes associated with the selective loss of beta cells.

Mapping of glutamic acid decarboxylase (GAD) genes

Glutamic acid decarboxylase (GAD) catalyzes the synthesis of gamma-aminobutyric acid (GABA), which is known as a major inhibitory neurotransmitter in the central nervous system (CNS), but is also present outside the CNS. Recent studies showed that GAD is the major target of autoantibodies associated with the development of insulin-dependent diabetes mellitus and of the rare stiff man syndrome. Studies of GAD expression have demonstrated multiple transcripts, suggesting several isoforms of GAD. In this study, three different genes were mapped by in situ hybridization to both human and mouse chromosomes. The GAD1 gene was mapped to human chromosome 2q31 and to mouse chromosome 2D in a known region of conservation between human and mouse. GAD2, previously mapped to human chromosome 10p11.2-p12, was mapped to mouse chromosome 2A2-B, which identifies a new region of conservation between human and mouse chromosomes. A potential GAD3 transcript was mapped to human chromosome 22q13 and to mouse chromosome 15E in a known region of conservation between human and mouse. It is concluded that the GAD genes may form a family with as many as three related members.

HLA-DQ primarily confers protection and HLA-DR susceptibility in type I (insulin-dependent) diabetes studied in population-based affected families and controls

The association between HLA-DR and -DQ and insulin-dependent diabetes mellitus (IDDM) in a defined high-incidence area was analyzed in a total of 58 population-based patients, representing 77% of IDDM patients with age at onset below 16 years, and in 92 unrelated parents in control families without IDDM. HLA haplotypes were confirmed by analyzing first-degree relatives in both groups. Seven different methods were used to analyze risk: (1) odds ratio, (2) absolute risk, (3) haplotype relative risk, (4) transcomplementation relative risk, (5) relative predisposing effects, (6) stratification analysis, and (7) test of predisposing allele on haplotype. DQB1*0302 indicated somewhat higher risk than did DR4, while DR3 had a higher risk than DQB1*0201; however, the 95% confidence intervals of the risk estimates overlapped. The positive association between IDDM and the DQB1*0201-DQA1*0501-DR3 haplotype seems to be due to DR3 or to an unknown linked gene. More important, DQA1*0301 was present among 93% of the patients, and this allele in various transcomplementation combinations with DQB1 alleles showed closer association to IDDM than did any other alleles. The strong negative association of the DQB1*0602 allele also in the presence of either DR4 or DQB1*0302 or both suggests that, in a high-risk population for IDDM, HLA-DQ primarily confers protection, perhaps by induction of tolerance. Consistent with known functions, HLA-DR may primarily confer susceptibility, perhaps by induction of autoreactive T lymphocytes.

Regulation of glutamic acid decarboxylase diabetes autoantigen expression in highly purified isolated islets from Macaca nemestrina

Macaca nemestrina, which may have larger and more numerous pancreatic islets than other species, was used for large scale islet isolation by ductal collagenase perfusion and Ficoll gradient centrifugation. The average yield was 51,000 islet equivalents per pancreas, or 8,750 islets equivalents per g. The average purity was 91%, often exceeding 95%. These are the highest reported size, purity, and yield per g of any nonautomated primate islet series. Perifusion with glucose, arginine, and isobutylmethylxanthine showed appropriate biphasic insulin secretion. Unlike that in the rat, human islet glutamic acid decarboxylase (GAD) isoform expression is restricted. However, glycemic regulation of GAD expression has been shown only in rats. We, therefore, tested hypotheses that M. nemestrina islets also have restricted GAD expression, that GAD expression in primates is stimulated by glucose, and that this stimulation remains restricted to the 64,000 mol wt (GAD65) isoform. Immunoprecipitation of labeled islet extracts showed that GAD65 expression increased 16.7 +/- 0.6-fold during high glucose in vitro culture. After controlling for observed increases in protein synthesis, specific glucose stimulation was still 4.2 +/- 0.2-fold. Specific antisera revealed no GAD67 expression under basal conditions, and isoform restriction was maintained during stimulation. Increased GAD65 synthesis thus accounts for glucose stimulation of 64K expression. These time- and concentration-dependent effects of glucose suggest that hyperglycemia increases autoantigenicity and may accelerate beta-cell destruction in primates, supporting a role for beta-cell rest in insulin-dependent diabetes mellitus prevention.

Beta-cell function in relation to islet cell antibodies during the first 3 yr after clinical diagnosis of diabetes in type II diabetic patients

OBJECTIVE

To determine the effects of islet cell antibodies on beta-cell function during the first 3 yr after diagnosis in type II diabetic patients.

RESEARCH DESIGN AND METHODS

beta-cell function in type II diabetic patients with (n = 11, 50 +/- 5 yr of age) and without (n = 10, 52 +/- 4 yr of age) ICA was followed prospectively and compared with beta-cell function in type I adult diabetic patients (n = 17, 37 +/- 5 yr of age) and in healthy control subjects (n = 34, age 45 +/- 3 yr). beta-cell function was evaluated as fasting C-peptide, 1 + 3 min C-peptide after intravenous glucose, and delta C-peptide after glucagon.

RESULTS

Fasting C-peptide was equal in type II diabetic patients with ICA (0.30 +/- 0.03 nM) and type I diabetic patients (0.24 +/- 0.03 nM) at diagnosis, and decreased (P < 0.05) during 3 yr in these groups but not in type II diabetic patients without ICA. At diagnosis, type II diabetic patients with ICA showed a 1 + 3 min C-peptide (0.92 +/- 0.17 nM) lower (P < 0.001) than control subjects but higher (P < 0.05) than type I diabetic patients (0.53 +/- 0.11 nM). After 1 yr, 1 + 3 min C-peptide in type II diabetic patients with ICA had decreased (P < 0.05) to 0.18 +/- 0.11 nM and was equal to type I diabetic patients (0.38 +/- 0.10 nM). delta C-peptide after glucagon was equally impaired in type II diabetic patients with ICA (0.38 +/- 0.06 nM) and type I diabetic patients (0.35 +/- 0.11 nM) at diagnosis. After 3 yr, type II diabetic patients with ICA had fasting C-peptide of 0.09 +/- 0.04 nM, 1 + 3 min C-peptide of 0.18 +/- 0.10 nM, and delta C-peptide after glucagon of 0.20 +/- 0.09 nM, values equal to type I diabetic patients but lower (P < 0.01) than in type II diabetic patients without ICA, whose values remained unchanged; fasting C-peptide of 0.97 +/- 0.17 nM, 1 + 3 min C-peptide of 2.31 +/- 0.50 nM, and delta C-peptide after glucagon of 1.76 +/- 0.28 nM.

CONCLUSIONS

In patients considered type II diabetic with ICA, beta-cell function progressively decreased after diagnosis, and after 3 yr was similar to type I diabetic patients, whereas beta-cell function in type II diabetic patients without ICA was unchanged.

Prospective evaluation of beta-cell function in insulin autoantibody-positive relatives of insulin-dependent diabetic patients

During the preclinical period of insulin-dependent diabetes mellitus (IDDM), progression to clinical IDDM is characterized by declining beta-cell function. Although the presence of insulin autoantibodies (IAA) improves the ability of islet cell antibodies (ICA) to predict subsequent clinical IDDM, few studies have examined the risk of developing IDDM in subjects positive for IAA but negative for both ICA and antibodies to glutamic acid decarboxylase (64kA). To investigate this question, detailed beta-cell function tests (acute insulin response to glucose [AIRgluc] and slope of glucose potentiation) were performed on eight IAA-positive first-degree relatives of insulin-dependent diabetics. All eight subjects were negative for ICA, and seven were tested for 64kA and were negative. Five subjects were studied prospectively for 22.4 +/- 9.4 months, while three subjects had only initial studies. Initial beta-cell function tests were normal in each subject. AIRgluc was 122.2% +/- 19.0% of the expected normal response, while slope was 168.6% +/- 20.6% of expected normal response. beta-cell function remained normal and remarkably stable in the five subjects followed prospectively. AIRgluc did not significantly change from an initial value of 147.9% +/- 23.1% of expected to 153.2% +/- 22.4% (NS). The slope of glucose potentiation varied little from 165.5% +/- 39.4% initially to 159.5% +/- 27.3% (NS) at the most recent determination. We conclude that among nondiabetic first-degree relatives of IDDM subjects, the presence of IAA in the absence of ICA and 64kA is not usually associated with and therefore does not reliably predict beta-cell dysfunction or progressive deterioration in beta-cell function.

Autoantibodies in IDDM primarily recognize the 65,000-M(r) rather than the 67,000-M(r) isoform of glutamic acid decarboxylase

Glutamic acid decarboxylase autoantibodies may aid in rapid screening strategies predicting IDDM before clinical onset. Rat islets contain GAD65 and GAD67 autoantibody targets, but human islets express only GAD65, now confirmed by direct immunoprecipitation from radiolabeled rat and human islets. Because human IDDM involves beta-cell-specific autoimmunity, we tested 190 new IDDM patients and 51 healthy control subjects for antibodies to recombinant human islet GAD65, rat islet GAD67, or human insulinoma/cerebellum GAD67, each expressed separately in hamster fibroblasts. By using immunoprecipitation, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and densitometric fluorogram scanning, 132 of 190 (70%) of new IDDM patients had GAD65 autoantibodies, whereas only 17 of 190 (9%) had antibodies to rat GAD67 (P < 0.001). Of healthy control subjects, 2 of 51 (3.9%) and 1 of 51 (1.9%) had antibodies to GAD65 and GAD67, respectively. All 17 GAD67 antibody-positive patients also had GAD65 antibodies; 14 of 17 with greater GAD65 than GAD67 index. Control studies showed comparable reactivity between recombinant rat and human GAD67 and between different subcellular preparations of recombinant GAD67 of either species. In conclusion, only GAD65 is expressed in human islets, the autoantibody response is primarily to this isoform, and GAD67 antibodies add little to IDDM detection.

HLA-DR3, DQ2 homozygosity in two patients with insulin-dependent diabetes mellitus superimposed with ulcerative colitis and primary sclerosing cholangitis

Two unrelated young males with the unusual simultaneous presence of insulin-dependent diabetes mellitus, ulcerative colitis and primary sclerosing cholangitis are reported. Both patients manifested homozygosity for the DR3-DQw2 (DQB*0201) HLA genotypes. We believe that homozygosity for this genotype may predispose for this type of multi-organ autoimmune disease.

Protection from BB rat diabetes by the platelet-activating factor inhibitor BN50730

The platelet-activating factor inhibitor BN50730, a hetrazepine, was injected intraperitoneally daily from 30 days of age into diabetes-prone BB rats. While 96% (22/23) Tween 80 injected control rats developed diabetes, 0.05 mg/kg BN50730 decreased the frequency to 72% (17/24; n.s.) and 0.5 mg/kg to 56% (14/25; p < 0.01). Mean onset age in controls was 81 +/- 9 days (mean +/- SD), but BN50730 delayed onset to 87 +/- 15 days in the low and 93 +/- 12. days (p < 0.01) in high dose rats. The relative degree of insulitis was reduced in both low (p < 0.01) and high (p < 0.05) dose treated groups. Serum insulin in young prediabetic controls decreased from 84 +/- 34 microU/ml to 38 +/- 20 in the 22 rats developing diabetes (p < 0.001). Serum insulin in BN50730-protected compared to unprotected rats was 114 +/- 49 and 32 +/- 22 (p < 0.001) in the low, and 91 +/- 46 and 21 +/- 15 (p < 0.001) microU/ml in the high dose group, respectively. Increased serum insulin correlated with preserved islet beta cells and decreased insulitis. Treatment did not affect thyroiditis. Thus, platelet-activating factor may be involved in insulitis pathogenesis and platelet-activating factor inhibitors may decrease autoimmune beta cell destruction.

Quantitative assay using recombinant human islet glutamic acid decarboxylase (GAD65) shows that 64K autoantibody positivity at onset predicts diabetes type

At and before onset, most insulin-dependent diabetics (IDDM) have islet GAD65 autoantibodies (GAD65Ab). Since IDDM also occurs in older patients where non-insulin-dependent diabetes is common, we studied GAD65Ab at onset to classify diabetes type. Our quantitative immunoprecipitation assay uses recombinant human islet GAD65 stably expressed in hamster fibroblasts. Electrophoretic mobility was identical to native islet GAD65. Like native antigen, recombinant GAD65 migrated as two bands during electrophoresis, but converted to one under stronger reduction. Immunoprecipitation was linear with respect to antibody or antigen concentration. In 120 population-based diabetic patients of all ages grouped by treatment at onset and after 18 mo, GAD65Ab were present in 70% on insulin (n = 37), 10% on oral agent (n = 62, P < 0.0001), 69% changing from oral agent to insulin (n = 16, P < 0.001), and 1 of 33 controls. 65% with GAD65Ab, versus 8% without, changed from oral agent to insulin (P < 0.01). The GAD65Ab quantitative index was remarkably stable, and only 2 of 32 patients changed antibody status during follow-up. Concordance between GAD65Ab and islet cell antibodies was 93%. Quantitative correlation was approximate but significant. This highly sensitive, quantitative, high capacity assay for GAD65Ab reveals treatment requirements better than clinical criteria, perhaps guiding immunomodulatory therapy.

Improved specificity of ICA assays in the Fourth International Immunology of Diabetes Serum Exchange Workshop

The goal of the Fourth International Workshop for Standardization of ICA Measurements was to determine the specificity of ICA assays and their ability to distinguish between control sera (n = 57) and sera from IDDM-related individuals--representing relatives of IDDM patients (n = 21), healthy individuals who later developed IDDM (n = 8), or newly diagnosed IDDM patients (n = 23). Results from 28 laboratories were analyzed. The mean specificity (percentage of control sera reported as negative) among 27 laboratories was 91%, including 6 laboratories with 100% specificity. Nevertheless, 78% of laboratories found at least one control sample > 0 JDF U. Among samples from first-degree relatives, the mean concordance was 86%, including three sera found negative (0 JDF U) by all laboratories. Among individuals who later developed diabetes, the mean concordance was 93%, with two sera found positive by 100% of laboratories. In sera from newly diagnosed IDDM patients, the mean concordance was 82%. Three sera were found positive and one serum negative by all laboratories. The JDF U of the sera considered to be positive were significantly greater than each laboratory's average for the controls. In conclusion, the results from laboratories participating in the Fourth International ICA Workshop demonstrated excellent specificity, good concordance, and an ability to separate control sera from defined, IDDM-related subjects.

Predictive value of islet cell and insulin autoantibodies for type 1 (insulin-dependent) diabetes mellitus in a population-based study of newly-diagnosed diabetic and matched control children

Most studies evaluating immune markers for prediction of Type 1 (insulin-dependent) diabetes mellitus have focused on first degree relatives, although only 10% of newly-diagnosed patients have an affected first degree relative. The Swedish Childhood Diabetes Register identifies 99% of all diabetic children at diagnosis. In this population-based study, islet cell antibodies and insulin autoantibodies in 0-14-year-old Swedish consecutively-diagnosed patients and control subjects were analysed to define their sensitivity and specificity. Over 16 months (1986-1987), 515 Swedish children developed diabetes. Plasma samples were obtained from 494 (96%) patients, and 420 matched control children. Among patients, the frequency of islet cell antibodies was 84% (415 of 494), insulin autoantibodies 43% (145 of 334); 40% (135 of 334) were positive for both and 88% (294 of 334) were positive for one or both. Among control children, 3% (14 of 420) had islet cell antibodies, 1% (4 of 390) insulin autoantibodies, and 4% (16 of 390) had either autoantibody marker. The predictive value of finding a patient with the disease was only 7% since 4% of the control children were antibody-positive and the cumulative incidence rate up to 15 years of age is 0.38%. None of the autoantibody-positive (n = 21) or negative control children developed diabetes during 3 to 5 years of follow-up. Longitudinal investigations of islet cell or insulin-autoantibody-positive healthy children are necessary to accurately determine the conversion rate from marker positivity to disease onset.

Pancreatic beta-cell function evaluated by intravenous glucose and glucagon stimulation. A comparison between insulin and C-peptide to measure insulin secretion

Insulin and C-peptide responses to 0.5 g kg-1 intravenous glucose and 1.0 mg glucagon were studied in 34 healthy subjects (age 19-78 years, mean 45). Fasting blood glucose (r = 0.59; p < 0.001) and glycosylated haemoglobin (r = 0.61; p < 0.001) increased with age, but not the initial C-peptide and insulin responses to the glucose infusion. However, the C-peptide response at 70 min (r = 0.36; p < 0.05), 80 min (r = 0.41; p < 0.05), and 90 min (r = 0.46; p < 0.01) after the glucose infusion correlated with age as well as both insulin (r = 0.42; p < 0.05) and C-peptide (r = 0.45; p < 0.05) responses to the glucagon injection. Reproducibility of insulin and C-peptide responses was evaluated by duplicate tests, separated 2-143 days in time, in 10 healthy subjects (age 19-48 years, mean 32 years) showing no significant differences in median within-subject variation between the initial (1 + 3 min) or overall (0-90 min area under curve) insulin (24% and 17%, respectively) and C-peptide (15% and 14%, respectively) responses to glucose, while the within-subject variation for the fasting values and the response to glucagon was higher (p < 0.05) for insulin (47% and 32%, respectively) than C-peptide (13% and 14%, respectively). Between-subject variation was also lower (p < 0.001) for C-peptide than for insulin. Thus, C-peptide measurements in healthy subjects are more reproducible than insulin measurements in determination of beta-cell function.

Recombinant glutamic acid decarboxylase (representing the single isoform expressed in human islets) detects IDDM-associated 64,000-M(r) autoantibodies

GAD is an autoantigen in IDDM. Molecular cloning and specific antibodies allowed us to demonstrate that only the lower M(r) GAD64 isoform is expressed in human islets, in contrast to human brain, rat islets, and rat brain, all of which express both GAD64 and GAD67. Expression of the human islet GAD64 isoform in COS-7 and BHK cells resulted in an enzymatically active rGAD64, which is immunoreactive with diabetic sera comparable with that of the islet 64,000-M(r) autoantigen. Immunoprecipitation analyses showed that 21/28 (75%) IDDM sera had rGA D64 antibodies compared with only 1/59 (1.7%) of the healthy control sera. In immunoblot analyses, an SMS serum--but only 1/10 randomly selected IDDM sera--recognized the blotted rGAD64 without relation to immunoprecipitation titers. In conclusion, only the GA D64 isoform is expressed in human islets, in contrast to rat islets, which also express the GAD67 isoform. The immunological properties of human rGAD64 are comparable with the native 64,000-M(r) islet autoantigen, allowing further studies of the immunopathogenesis of IDDM.

Genetic dissection of autoimmune type I diabetes in the BB rat

The BB rat is among the best models of insulin-dependent diabetes mellitus--with onset and pathogenesis closely resembling the human disease. One unusual feature is a severe T-cell lymphopenia, which appears to be inherited as a recessive trait controlled by a single gene, Lyp. Based on genetic analysis of several crosses, we show that development of diabetes involves at least three genes: Lyp, which is tightly linked to the neuropeptide Y (Npy) gene on chromosome 4, a gene linked to the major histocompatibility complex (MHC) on chromosome 20, and a third unmapped gene for which the Fischer rat strain carries an allele conferring resistance.