High GAD65 autoantibody levels in nondiabetic adults are associated with HLA but not with CTLA-4 or INS VNTR

OBJECTIVES

To explore the relationship between genetic background and antibody levels in a nondiabetic population. We evaluated if high levels of autoantibodies against the 65 kDa isoform of glutamic acid decarboxylase (GAD65Ab), were associated with high-risk genes, i.e. HLA, CTLA-4 and INS VNTR genes.

DESIGN AND SUBJECTS

Seventy-five (M/F 39/36) subjects exceeding the 95th percentile of GAD65 autoantibody index and 75 age and sex matched subjects below the 95th percentile, randomly selected amongst participants in the Västerbotten Intervention Programme.

METHODS

The GAD65 Ab were measured in a radioligand-binding assay. HLA class II typing was performed by an oligoblot hybridization method. CTLA-4 repeat length was analysed and divided into short forms and long forms. Class I and class III alleles of INS VNTR were detected. Differences in distribution were tested by Pearson chi-square with Yates correction. Odds ratios (OR) were used to compare groups calculated with Cochran's and Mantel-Haenszel statistics.

RESULTS

The DQB1*0201-DQA1*0501-DRB1*03 haplotype was increased in subjects with high GAD65Ab levels (P = 0.04). This increase seemed to be explained by a difference in haplotype frequencies amongst men (P = 0.01). Calculating OR showed a significant association between the DQB1*0201-DQA1*0501-DRB1*03 haplotype and elevated levels of GAD65Ab in all subjects (OR 2.2, 95% CI 1.02-4.9) as well as in men (OR 4.6, 95% CI 1.3-15.9). There was no association between high levels of GAD65Ab and either INS VNTR or CTLA-4 polymorphisms.

CONCLUSION

Our study suggests that adult males with the DQB1*0201-DQA1*0501-DRB1*03 haplotype tend to develop high GAD65Ab titres. As none of these subjects have developed diabetes these data suggest that HLA may be important in GAD65Ab formation but that additional factors are required for the progression to overt type 1 diabetes.

Animal models of endocrine/organ-specific autoimmune diseases: do they really help us to understand human autoimmunity?

Organ-specific or endocrine autoimmune diseases are complex, polygenic afflictions the penetrance of which is heavily dependent on various environmental influences. Important target tissues are the thyroid, the islets of Langerhans, gastric parietal cells and steroid-producing cells in the adrenal and ovary. The etiology of these diseases remains to be clarified. The pathogenesis is strongly associated with autoimmune phenomena. None of the current treatment approaches provides a cure; rather they represent replacement therapy. An important objective in the treatment of endocrine/organ-specific autoimmune diseases is the detection of individuals at risk for the development of such diseases and the development of interventions to prevent an outbreak of the diseases. This requires an exquisite knowledge of the early etio-pathogenic stages of these diseases. This review concentrates on the usefulness of animal models for a precise understanding of these very early stages. It must be emphasized that studying animal models cannot answer all the problems presented by endocrine/organ-specific autoimmune diseases as seen in the clinic. It must be expected - considering the different etiologies in the different animal models (see below) - that the causes of the diseases in the human and the involvement of various genes and environmental factors may also vary. Yet, particularly in the study of the pre-autoimmune phases of the diseases, there is hardly any alternative to the study of animal models. Only limited series of experiments can be carried out in human subjects at risk to develop such diseases. Moreover, a general semblance (blueprint) of the etio-pathogenesis found in the animal models can lead the way for human studies. Efforts to understand the patho-physiology of the early stages of endocrine/organ-specific autoimmune diseases have mainly involved animal models that "spontaneously" develop such diseases. Of these the bio-breeding diabetes-prone (BB-DP) rat and the non-obese diabetes (NOD) mouse are the most well studied, yet many studies have also been carried out in the obese strain (OS) chicken. Apart from these spontaneous models there are animal models that are induced by environmental perturbations (viruses, toxic substances), by thymectomy procedures or by genetic manipulations, e.g., the RIP-LCMV model and the BDC 2.5 TCR mouse model. A general blueprint has emerged from the studies into the early stages of the pathogenesis of endocrine/organ-specific autoimmune diseases in these animal models: animals at risk to develop endocrine/organ-specific autoimmune diseases show various pre-autoimmune aberrancies in their target glands, T cells, macrophages (Mphi) and dendritic cells (DC). The presumably aberrant target cells, T cells, DC and Mphi need to interact abnormally before autoimmune disease can fully develop. In this abnormal interaction additional aberrancies in other regulatory systems may play a role in a further exacerbation of the self-directed immune response, such as defects in the hypothalamus pituitary adrenal (HPA) axis system. The various aberrancies are partly genetically determined by a variety of separate genes, particularly MHC-related genes, but they may also be environmentally induced (e.g., via viruses, high iodine diet, and other experimental manipulations). Recently evidence has been gathered for pre-autoimmune aberrancies similar to the animal models in the DC/ Mphi compartment and the HPA axis in humans at risk to develop endocrine/organ-specific autoimmune diseases. However, analogous pre-autoimmune abnormalities in human target glands or in T cell function have not yet been found with certainty. We believe that animal models of endocrine/organ-specific autoimmune disease still hold immense promise for the discovery of pathways, genes and environmental factors that determine the development of endocrine/organ-specific autoimmune diseases. Animals affected by such diseases provide a unique opportunity to uncover disease-associated pathways, which are complicated to define in man.

Macrophages from high-risk HLA-DQB1*0201/*0302 type 1 diabetes mellitus patients are hypersensitive to lipopolysaccharide stimulation

Levels of nonantigen-induced pro-inflammatory cytokines and prostaglandin in macrophages isolated from human leucocyte antigen (HLA)-matched type 1 diabetes mellitus patients, first-degree relatives and healthy controls were determined. We hypothesize that monocytes isolated from patients are sensitized or preactivated and therefore, have an altered response to in vitro stimulus compared with control groups as measured by levels of pro- and anti-inflammatory mediators. In this study, peripheral blood monocytes were differentiated to macrophages with macrophage-colony stimulating factor (M-CSF) to determine lipopolysaccharide (LPS)-stimulated tumour necrosis factor (TNF)-alpha, interleukin (IL)-1beta, IL-6, IL-12 and prostaglandin E-2 (PGE-2) secretion from hetero- or homozygous HLA DQB1*0201 and *0302 type 1 diabetes mellitus patients, first-degree relatives and homozygous HLA DQB1*0602 healthy controls. LPS-stimulated secretion of TNF-alpha, IL-1beta and IL-6 was immediate and markedly higher in the HLA-DQB1*0201/*0302 type 1 diabetes patients compared with all other groups including HLA-matched healthy first-degree relatives. In DQB1*0201/*0302 diabetes patients PGE-2 secretion was delayed but increased by LPS stimulation compared with HLA-matched healthy relatives. IL-12 was not detected at any condition. These data suggest that macrophages from DQB1*0201/*0302 type 1 diabetes patients are sensitized to secrete both cytokines and PGE-2 following nonantigenic stimulation. Sensitized macrophages may be important to high-risk DQB1*0201/*0302-associated type 1 diabetes.

Preservation of enzyme activity and antigenicity after mutagenesis of the membrane anchoring domain of GAD65

The smaller isoform of glutamic acid decarboxylase, GAD65, is an important autoantigen implicated in the pathogenesis of type 1 diabetes whereas the larger isoform, GAD67 appears to play no major role. The primary difference between the two isoforms resides in the N-terminal part of the molecule including the GAD65 membrane-anchoring domain. The aim of this study was to generate mutants of the membrane targeting domain spanning amino acids 24 to 31 of GAD65 to determine effects on enzyme activity and antibody recognition. Three GAD65 mutants were generated by substituting two, nine or eleven nucleotides coding for the membrane targeting with the corresponding bases of GAD67. SDS-PAGE and Western blotting wildtype (wt) and mutated GAD65 ascertained that they were of similar size and recognized GAD65-specific antibodies. No difference in enzymatic activity was found between the mutants and wt GAD65. GAD65 antibody positive sera from type 1 diabetes patients immunoprecipitated mutated GAD65 whether two, nine or eleven nucleotides were replaced. Mono-or polyclonal antibodies to the N-terminal region demonstrated that the mutated GAD65 with two or nine nucleotides replaced was immunoprecipitated markedly better than wt whereas no difference was detected using antibodies specific for the PLP-binding site in the middle part of GAD65 or the C-terminal region. Taken together, these data suggest that no major conformational changes have been introduced by mutating the membrane-anchoring domain of GAD65.

Tissue transglutaminase autoantibodies and human leucocyte antigen in Down's syndrome patients with coeliac disease

The association between autoantibodies against tissue transglutaminase (tTG) and human leucocyte antigen (HLA)-DQB1 alleles was tested in Down's syndrome (DS) patients with and without coeliac disease (CD). Immunoglobulin A (IgA) and G (IgG) anti-tTG were measured in radioligand binding assays and compared with conventionally analysed IgA antibodies against gliadin (AGA) and IgA autoantibodies against endomysium (EMA) in 48 DS patients. HLA-DQB1 typing was carried out by polymerase chain reaction and hybridization with allele-specific probes in 41/48 patients. Both IgA-tTG and IgG-tTG, as well as EMA, were detected in 7/48 and AGA in 15/48 patients. Intestinal biopsy showed histopathological changes consistent with CD in 9/16 patients. HLA-DQB1 typing, available for 8/9 patients with and for 33/39 without CD, demonstrated that 5/8 with CD had DQB1*02 compared with 7/33 of those without (p = 0.0345). In patients with anti-tTG, 5/6 had the DQB1*02 allele compared with 7/35 of those without (p = 0.0053).

CONCLUSIONS

Anti-tTG are HLA-DQB1*02-associated autoantibodies which together could be useful screening tests for silent CD in DS patients. In patients with gastrointestinal symptoms or clinical signs of malabsorption, anti-tTG should be combined with AGA to detect other forms of enteropathies and CD.

Tissue transglutaminase autoantibodies and human leucocyte antigen in Down's syndrome patients with coeliac disease

The association between autoantibodies against tissue transglutaminase (tTG) and human leucocyte antigen (HLA)-DQB1 alleles was tested in Down's syndrome (DS) patients with and without coeliac disease (CD). Immunoglobulin A (IgA) and G (IgG) anti-tTG were measured in radioligand binding assays and compared with conventionally analysed IgA antibodies against gliadin (AGA) and IgA autoantibodies against endomysium (EMA) in 48 DS patients. HLA-DQB1 typing was carried out by polymerase chain reaction and hybridization with allele-specific probes in 41/48 patients. Both IgA-tTG and IgG-tTG, as well as EMA, were detected in 7/48 and AGA in 15/48 patients. Intestinal biopsy showed histopathological changes consistent with CD in 9/16 patients. HLA-DQB1 typing, available for 8/9 patients with and for 33/39 without CD, demonstrated that 5/8 with CD had DQB1*02 compared with 7/33 of those without (p = 0.0345). In patients with anti-tTG, 5/6 had the DQB1*02 allele compared with 7/35 of those without (p = 0.0053).

CONCLUSIONS

Anti-tTG are HLA-DQB1*02-associated autoantibodies which together could be useful screening tests for silent CD in DS patients. In patients with gastrointestinal symptoms or clinical signs of malabsorption, anti-tTG should be combined with AGA to detect other forms of enteropathies and CD.

Streptozotocin induced diabetes in minipig: a case report of a possible model for type 1 diabetes?

Studies on the pathogenic process in type 1 diabetes are often performed in animal models. Low-dose administration of streptozotocin has been used to induce diabetes with pathological alterations similar to human type 1 diabetes in the animals. Rodent models are frequently used but there is a need of developing new models including larger animals. In this study we wanted to investigate to what extent a minipig was sensitive to low-dose streptozotocin for induction of diabetes with features of human Type I diabetes. A female Göttingen minipig received two low-doses (40 mg/kg) of streptozotocin with an 11-day interval. Serum was analysed for the presence of the enzyme glutamic acid decarboxylase, isoform 65, (GAD65) and autoantibodies against glutamic acid decarboxylase, isoform 65 (GAD65A), isoform 67 (GAD67A), insulinoma antigen 2 (IA-2) and insulin (IAA). Pancreas tissue was fixated in formaldehyde and was sent for pathoanatomical examination. The minipig became hyperglycaemic after the second injection of streptozotocin. The pathoanatomical examination showed atrophy of the beta-cell population, depletion of insulin with preserved glucagon content. There was no sign of insulitis. Both GAD65 and GAD65A were detected while GAD67A and IAA were absent. It is concluded that chronic diabetes developed after low-dose streptozotocin injection in a female minipig with the characteristics of the end stage of type 1 diabetes. This pilot study suggests that minipigs show promise as a model to induce diabetes by injections of low-dose streptozotocin.

The combination of several polymorphic amino acid residues in the DQalpha and DQbeta chains forms a domain structure pattern and is associated with insulin-dependent diabetes mellitus

IDDM is positively associated with HLA-DQA1*0301-DQB1*0302 (DQ8) and DQA1*0501-DQB1*0201 (DQ2) and negatively associated with DQA1*0102-DQB1*0602 (DQ6). The aim of the present study was to analyze the importance of several polymorphic residues and domains of DQalpha and DQbeta, in addition to residue 52 DQalpha and residue 57 DQbeta, with regard to susceptibility or resistance in new-onset 0- to 15-year-old Swedish children with IDDM (n = 425) and matched controls (n = 367). HLA genotyping identified several polymorphic residues of the DQalpha and DQbeta to be either positively or negatively associated with IDDM, including Arg 52 DQalpha and Asp 57 DQbeta. Leu 69 DQalpha was positively (OR 7.02, P < 0.0001), Ala 69 DQalpha was negatively (OR 0.22, P < 0.0001), Gln 47 DQalpha was positively (OR 5.8, P < 0.0001), Cys 47 DQalpha was positively (OR 2.2, P < 0.0001), Lys 47 DQalpha was negatively (OR 0.47, P < 0.005), and Arg 47 DQalpha was negatively (OR 0.22, P < 0.005) associated with IDDM. Similarly, residues at 11, 18, 45, 48, 50, 53, 55, 61, 64, 66, 76, and 80 were either positively or negatively associated with IDDM. Likewise, for DQbeta, Leu 53 DQbeta was positively (OR 11.01, P < 0.0001), Gln 53 DQbeta was negatively (OR 0.22, P < 0.0005), Arg 70 DQbeta was positively (OR 11.01, P < 0.0001), and Gly 70 DQbeta was negatively (OR 0.19, P < 0.0001) associated like other residues at 71, 74, 84, 85, 86, 89, and 90 DQbeta with IDDM. Certain domains in the DQalpha, RFTIL (at DQalpha positions 52, 61, 64, 66, and 69), were present in 95% of patients compared to 69% of controls (OR 9.01, P(c) < 0.0001), and DQbeta domain GR (at DQbeta positions 45 and 70) was present in 95% of patients and 68% of controls (OR 8.68, P < 0.0001), which correlated better than the individual amino acid residues with IDDM. A combination of the DQalpha and DQbeta chain domains was present in 94% of patients compared to 60% of controls (OR 10.6, P < 0.001). In conclusion, domains in the DQalpha, DQbeta, or both in the DQ molecule explain susceptibility or resistance to IDDM better than individual amino acid residues of DQA1 and DQB1.

GAD65 and IA-2 autoantibodies are common in a subset of siblings of Sardinian Type 2 diabetes families

Type 1 diabetes in Sardinia is very common in children, and we hypothesized that Latent Autoimmune Diabetes of Adult (LADA) might constitute a significant proportion of diabetes in adult Sardinian subjects. Since Type 2 diabetes is a familial disorder, we tested this hypothesis by investigating the prevalence of GAD65 and IA-2 autoantibodies (Ab) in Type 2 diabetes multiplex families of Sardinian ancestry enrolled in the Study Group for the Genetics of Diabetes in Sardinia.

METHODS

A total of 684 individuals were ascertained from 252 Sardinian Type 2 diabetes multiplex families with 2.4 affected siblings per family comprising 190 families with two affected, 37 with three, 15 with four, 7 with five, and 3 with six, in addition to 80 unaffected siblings. Controls were household contacts representing 204 healthy spouses of affected siblings. Diagnosis was at 35-69 years of age and insulin was not given in the first 4 years after diagnosis. GAD65Ab and IA-2Ab were determined in standard radioligand binding assays.

RESULTS

Among affected siblings GAD65Ab were positive in 8.8% of insulin-treated (n = 137; P = 0.0006), in 2.5% of non-insulin-treated (n = 467), and in 1.2% of non-diabetic siblings (n = 80) compared with 0.5% of controls (n = 204). IA-2Ab was positive in 6.6% insulin-treated (P = 0.04), 2.1% non-insulin-treated, and 2.5% non-diabetic siblings compared with 1.5% of controls.

CONCLUSION

A high frequency of GAD65Ab and IA-2Ab as markers of Type 1 diabetes was found among Type 2 diabetes siblings from Sardinian multiplex families despite excluding those who had been treated with insulin during the first 4 years of disease. Our data support the hypothesis that LADA may be common in Sardinian Type 2 diabetes and stress the importance of investigating markers of Type 1 diabetes in studies of Type 2 diabetes.

Induction of donor-specific tolerance to islet allografts in nonhuman primates

Pancreatic tissue grafting is by far the most physiological therapeutic solution to the insulin deficiency of diabetes. Recent clinical trials have indicated somewhat successful use of nonsteroidal immunosuppressive regimens and a successful nonhuman primate trial using CD154 for costimulation blockade was reported. However, these protocols need to be replaced with safe and efficacious ones in which long-term allotolerance would make these treatments routine in a clinical setting. With the specific objective of testing whether peripheral infusions of stem cells or stem cell fractions in conjunction with islet allografting would induce allograft tolerance, we have established a macaque diabetic model. The macaques were rendered diabetic by streptozotocin and required daily doses of insulin to maintain lower blood glucose levels. The diabetic macaques then received islets and stem cells from unrelated and MHC-mismatched donors without any immunosuppression. In our initial analysis, 5 of 7 macaques that received stem cell infusions at the time of islet allografting have shown allograft survival longer than the group of macaques that received islets without the stem cell infusion. One of these five macaques has been normoglycemic for 10 months, with no exogenous insulin. This macaque received stem cell population enriched for CD34+ cells with depletion of CD18 cells, which have shown low or no allostimulatory potential in mixed lymphocyte cultures. Increased levels in insulin and C-peptide levels were shown in the macaques after islet transplantation.

Induction of islet allotolerance in nonhuman primates

Recent clinical trials have pioneered the successful use of a nonsteroidal immunosuppressive regimen and established a basis for application in a routine clinical setting. In this study, a single islet transplant was not sufficient to regulate blood glucose levels, and a second transplant became necessary. A similar observation was made in our macaque islet transplant study, where animals after the second transplantation have shown trends towards normoglycemia in the presence of mycophenolate mofetil. All five animals that received the second transplant have shown an initial rise in C peptide levels, which rapidly decreased as we tapered the MMF dose from 20 mg/kg BID to 5 mg/kg SID. Two animals of the five that were preconditioned with MMF one week prior to transplantation have shown significantly higher C peptide levels. We believe that it is very important to understand the relationship between the first graft failure and subsequent islet allograft success. Since graft success did not correlate with number of transplanted islets, the correction of blood glucose levels toward normoglycemia after the second transplantation suggests a mechanism by which the allotolerance to second transplant is facilitated by the first islet transplantation. These initial observations suggest approaches to "tolerize" the recipient to accept the second-transplant islets (a) through preconditioning the animal to improve the rate of success for the first transplant or (b) through tolerization to islets in the first transplant to facilitate better engraftment of the second-transplant islets.

Treatment outcome for IDDM patients in relation to glutamic acid decarboxylase autoantibodies and serum IgG to periodontal pathogens

BACKGROUND

Patients with insulin-dependent diabetes mellitus (IDDM) have elevated risk for periodontitis (PD) relative to subjects without diabetes. Whether refractory PD in IDDM patients is related to autoimmunity as indicated by serum glutamic acid decarboxylase autoantibody GAD Ab levels or to host bacterial immunity as reflected by serum antibody titers to periodontal pathogens is unknown.

AIMS

To determine if non-surgical periodontal treatment outcome differs between GAD Ab-seropositive and -seronegative IDDM patients by assessing the following parameters: (1) pretreatment serum levels of GAD Ab, (2) pretreatment serum IgG titers to key periodontal pathogens, and (3) changes in periodontal pocket probing depth (PDC) after treatment.

METHODS

Before and two months after periodontal treatment of 11 GAD Ab-seronegative and 7 -seropositive subjects, PDC was assessed and serum GAD Ab and IgG to Porphyromonas gingivalis (Pg), Bacteroides forsythus (BJ), and Actinobacillus actinomycetemcomitans (Aa) were studied using established radioligand precipitation and enzyme-linked immunosorbent assays, respectively.

RESULTS

The PDC decrease was significantly better for GAD Ab-seronegative subjects than for seropositive subjects (median 1.4 mm+/-0.5 s.d. versus 0.5 mm+/-0.3 s.d., p<0.03, Mann-Whitney). GAD Ab levels and PDC were positively correlated (r=+0.71, p<0.05) for sero-positive subjects but were neutral (r=-0.07) for seronegative subjects. Serum IgG to Pg and GAD Ab levels were positively associated (r2=0.42) in seropositive subjects. Logistic regression analysis confirmed that GAD Ab status was the primary discriminator for PDC (p<0.04).

CONCLUSION

Detection of elevated GAD Ab levels in combination with elevated IgG titers to Pg before treatment is indicative of IDDM patients with refractory PD.

Combinations of beta cell specific autoantibodies at diagnosis of diabetes in young adults reflects different courses of beta cell damage

To explore the natural course of beta cell function in recent onset diabetes, a subgroup (n=157) of all incident cases (n=879) 15-34 years old, 1992-1993 in Sweden, and with positivity for at least one autoantibody of islet cell antibodies (ICA), glutamic acid decarboxylase antibodies (GADA) or tyrosine phosphatase antibodies (IA-2A) were followed prospectively for the first four years with annual analysis of C-peptide. The aim was to relate the course of beta cell function, measured as C-peptide, in early diabetes with the presence of different islet autoantibodies at diagnosis. We found that patients positive for ICA alone (n=11) had significantly higher C-peptide levels both at diagnosis and during the first three years compared with the other patients (n=146; p=0.022, p<0.001, p=0.004 and p=0.0022). Patients positive for GADA alone or in combination with other antibodies (n=125) had significantly lower C-peptide during the first three years after diagnosis compared with the other patients (n=32, p<0.001, p=0.0011 and p=0.0136). Patients with two or three autoantibodies had C-peptide levels similar to levels found in patients positive only for GADA. However, after four years, there were no significant differences between any of the groups of different autoantibody combinations. At diagnosis, 55% (86/157) of the patients had C-peptide levels above the lower normal range of 0.25 nmol/l, but the frequency of patients with beta cell function above this level decreased after two years to 41% (65/157; p=0.035) and after four years to 22% (35/157; p=0.0041). It is concluded that young adult diabetic patients positive only for ICA at diagnosis have a better preserved beta cell function with higher levels of C-peptide during the first three years compared with patients positive for GADA alone or in combinations with other autoantibodies.

Comparison of the prevalence of glutamic acid decarboxylase (GAD65) and gliadin antibodies (AGA) in a randomly selected adult estonian population

We aimed to test the hypothesis that gluten might be associated with the development of islet cell autoimmunity. A random sample of 200 persons (87 males, mean age 42.4 years) from Estonia including one patient with type I diabetes mellitus was studied. IgG-type glutamic acid decarboxylase (GAD65) antibodies were determined using radioligand-binding assay and IgG/IgA-type gliadin antibodies (AGA) by enzyme-linked immunosorbent assay. Generic HLA-DRB1* alleles were analyzed using a polymerase chain reaction. Although our results revealed the highest GAD65Ab index and a high IgA-type AGA in a person with diabetes, no correlation between GAD65Ab and AGA values was revealed among the other 199 persons (p > 0.05). There were also no differences between test values among persons with and without different HLA-DRB1* alleles (p > 0.05). In the GAD65Ab assay, one person (0.5 %; 95 % CI: 0 - 1.5) out of 199 exceeded the 99(th) centile of the GAD65Ab index. In summary, the present study does not confirm the possibility that there is a relationship between the immune reactivity against GAD65 and gliadin, at least in persons without type I DM.

Prediction of silent celiac disease at diagnosis of childhood type 1 diabetes by tissue transglutaminase autoantibodies and HLA

AIMS

The aims were to estimate the diagnostic sensitivity and specificity of autoantibodies to tissue transglutaminase (IgA- and IgG-tTG), gliadin (AGA) and endomysium (EMA) in relation to human leukocyte antigen (HLA)-DQB1 alleles to identify silent celiac disease at diagnosis of type 1 diabetes.

METHODS

IgA- and IgG-tTG were measured in radioligand binding assays in 165 type 1 diabetic patients. Data on HLA-DQB1 were available for 148 patients and on both AGA and EMA for 164 patients. For patients considered positive for AGA or EMA, or both, an intestinal biopsy was suggested. HLA-DQB1 typing was carried out by polymerase chain reaction and hybridization with allele specific probes.

RESULTS

Three patients, left out from further study of antibodies, but not from HLA-DQB1 analysis, had treated celiac disease at diagnosis. Out of the other 162 type 1 diabetic patients tested, nine had IgA-tTG, six IgG-tTG, eight EMA, and 11 AGA. Biopsy was suggested for nine patients, of whom six showed villous atrophy, one did not and two refused to participate. Thus, silent celiac disease was probable in 8/162 and biopsy-verified in 6/162, where five patients were AGA-positive and six either EMA-, IgA-tTG- or IgG-tTG-positive. Of the 11 patients with celiac disease (three with treated and eight with silent celiac disease), 10 were HLA-DQB1-typed, of whom 65% (13/20) had the DQB1*02 allele, compared with 36% (100/276; p = 0.011) of those without celiac disease. IgA-tTG levels were higher in patients having either *02 or *0302 (0.6; -1.3-112.4 RU) compared with those not having these alleles (0.4; -0.7-3.4 RU; p = 0.023).

CONCLUSION

IgA-tTG are HLA-DQB1*02-associated autoantibodies with high sensitivity and specificity for silent celiac disease at diagnosis of type 1 diabetes.

Low serum conditions for in vitro generation of human macrophages with macrophage colony stimulating factor

Animal serum is often used to generate human macrophages in vitro. Since fetal calf serum (FCS) may complicate antigen uptake, processing and presentation on HLA molecules, we tested the ability of M-CSF to generate macrophages at low fetal calf serum conditions. Peripheral blood monocytes from 12 individuals were cultured 1-4 days with 0-100 ng/ml macrophage colony stimulating factor (M-CSF) at either 1 (low) or 5% (v/v) FCS. Regardless of number of days in culture, maximal (50-100 ng/ml) M-CSF stimulation and low FCS induced 65+/-5% esterase positive cells in all individuals compared to 52+/-7% without M-CSF (P<0.001). M-CSF increased the mean proportion of esterase positive cells after 24 or 96 h by 13% (P<0.005) and 13% (P<0.005), respectively, in 1% FCS, and 8% (P<0.05) and 2% (NS), respectively, in 5% FCS, indicating a slight negative interaction between 5% FCS and M-CSF (P<0.05). All cells were positive for CD14 and HLA class II, but cell number did not increase, confirming that M-CSF promote macrophage differentiation also at low FCS. M-CSF increased the average cell size after 24 or 96 h by 5.9+/-1.0 (P<0.05) and 8.6+/-0.5 (P<0.001) microm, respectively, without an increase in 5% FCS, further demonstrating the efficiency of M-CSF to promote macrophage generation at low FCS. The culture supernatants were negative for IL-1beta and TNF-alpha, which demonstrates that M-CSF did not activate the macrophages. The generation of human macrophages by M-CSF at low FCS should prove useful in studies where higher FCS concentrations may interfere with the assay.

Evaluation of a novel radioimmunoassay using 125I-labelled human recombinant GAD65 for the determination of glutamic acid decarboxylase (GAD65) autoantibodies

Autoantibodies to the islet cell 65-kilodalton isoform of glutamic acid decarboxylase (GAD65) are present in most patients with type 1 diabetes mellitus years before the clinical manifestation of the disease. GAD65 autoantibodies are also present in a subset of patients with type 2 diabetes who frequently become insulin dependent. In the present study, we evaluated a new, commercially available radioimmunoprecipitation assay for measuring GAD65 autoantibodies using 125I-labelled human recombinant GAD65. Results obtained with this assay were compared with those obtained by a reference assay based on 35S-labelled recombinant GAD65. Analyses were performed on 67 patients with type I diabetes, 350 with type 2 diabetes, and 150 apparently healthy individuals. An excellent agreement was found between the results obtained by the 125I-GAD65 assay and those obtained by the reference method. The receiver operating characteristic (ROC) curve analysis was used to evaluate the sensitivity and specificity of the two assays. The sensitivity of each assay was determined from the results of the 67 type 1 patients, while the specificity was based on the 150 healthy individuals. Based on the ROC curves, the two assays appeared identical, with a sensitivity of 84% and a clinical specificity of 98%. In conclusion, based on our results, this simple, one-step centrifugation, high-capacity 125I-GAD65 assay has the same sensitivity and specificity as the reference assay and is highly suitable to detect GAD65 autoantibodies in human samples.

A novel monoclonal antibody specific for the N-terminal end of GAD65

We describe a novel monoclonal antibody raised towards the N-terminus of the 65 kDa isoform of glutamate decarboxylase (GAD65). This N-GAD65 mAb is highly specific for GAD65 and can be used in a wide range of applications such as Western blot analysis, immunoprecipitation and immunocytochemistry. Full-length cDNAs coding for N-GAD65 mAb heavy and light chains were cloned and characterized by nucleotide sequence analysis. Both heavy and light chain-specific cDNAs are functional and are members of mouse heavy chain subgroup IgG1 and kappa chain group 2, respectively. Comparing the N-GAD65 mAb gene sequences with GenBank shows that the cDNAs were not reported previously.

Site-directed mutagenesis of K396R of the 65 kDa glutamic acid decarboxylase active site obliterates enzyme activity but not antibody binding

The role of K396 in the enzymatic catalysis and the antigenicity of the 65 kDa isoform of glutamate decarboxylase (GAD65) was analyzed using the K396R GAD65 mutant. GAD65 is a major autoantigen in Type 1 diabetes and autoantibodies directed to GAD65 are widely used markers for this disease. We found that (1) recombinant human GAD65 is fully enzymatically active; (2) the K396R mutation abolished GAD65 activity; and (3) the K396R mutant retained full antigenicity to GAD65 autoantibodies in serum from Type 1 diabetes patients, but not to polyclonal antibodies raised to the catalytic domain.

Recognition of glutamic acid decarboxylase (GAD) by autoantibodies from different GAD antibody-positive phenotypes

Autoantibodies against the smaller isoform of glutamic acid decarboxylase (GAD) are markers for Type 1 diabetes. GAD65 autoantibody (GAD65Ab)-positive individuals in the general population are, however, mostly at low risk of developing Type 1 diabetes, suggesting that GAD65Ab phenotypes may be associated with different underlying pathogenic processes. The aim of this study was to test the hypothesis that Type 1 diabetes patients (n = 243; group I), GAD65Ab-positive healthy individuals (n = 28; group II), and healthy first-degree relatives of Type 1 diabetes patients (n = 41; group III) have antibody phenotypes that recognize different GAD65 epitopes. Sera from groups I-III were tested for their binding to GAD65 and GAD67, as well as six different GAD65/67 fusion proteins. Regardless of group, sera reactive to both GAD65 and GAD67 showed broader epitope reactivity than GAD65-specific sera. Furthermore, Type 1 diabetes patients showed a more restricted epitope binding than healthy individuals and first-degree relatives, demonstrating significantly less binding to the N-terminal part of GAD65 and to GAD67. Our analysis demonstrates that the N-terminal part is essential for full antibody binding to GAD65, in particular, to the middle epitope. It is suggested that Type 1 diabetes is associated with restricted GAD65Ab epitope specificity.

Prognostic factors for the course of beta cell function in autoimmune diabetes

This study presents a 2-yr follow-up of 281 patients, aged 15-34 yr, diagnosed with diabetes between 1992 and 1993. At diagnosis, 224 (80%) patients were positive for at least one of the following autoantibodies: islet cell antibodies (ICAs), glutamic acid decarboxylase antibodies (GADAs), or tyrosine phosphatase antibodies (IA-2As); the remaining 57 (20%) patients were negative for all three autoantibodies. At diagnosis, C-peptide levels were lower (0. 27; 0.16-0.40 nmol/L) in autoantibody-positive patients compared with autoantibody-negative patients (0.51; 0.28-0.78 nmol/L; P: < 0. 001). After 2 yr, C-peptide levels had decreased significantly in patients with autoimmune diabetes (0.20; 0.10-0.37 nmol/L; P: = 0. 0018), but not in autoantibody-negative patients. In patients with autoimmune diabetes, a low initial level of C-peptide (odds ratio, 2. 6; 95% confidence interval, 1.7-4.0) and a high level of GADAs (odds ratio, 2.5; 95% confidence interval, 1.1-5.7) were risk factors for a C-peptide level below the reference level of 0.25 nmol/L 2 yr after diagnosis. Body mass index had a significant effect in the multivariate analysis only when initial C-peptide was not considered. Factors such as age, gender, levels of ICA or IA-2A or insulin autoantibodies (analyzed in a subset of 180 patients) had no effect on the decrease in beta-cell function. It is concluded that the absence of pancreatic islet autoantibodies at diagnosis were highly predictive for a maintained beta-cell function during the 2 yr after diagnosis, whereas high levels of GADA indicated a course of decreased beta-cell function with low levels of C-peptide. In autoimmune diabetes, an initial low level of C-peptide was a strong risk factor for a decrease in beta-cell function and conversely high C-peptide levels were protective. Other factors such as age, gender, body mass index, levels of ICA, IA-2A or IAA had no prognostic importance.