Cloning and expression of IDDM-specific human autoantigens.

A DNA cloning approach was taken to identify islet cell protein antigens that are recognized specifically by insulin-dependent diabetes mellitus (IDDM) sera. A human islet cDNA library was generated and screened with diabetic sera. In this article, identification of two clones is described. Proteins expressed by these lambda phages appeared to react specifically with newly diagnosed diabetic sera. Islet cell antibody 12 (ICA12) was tested by Western blotting. ICA512 was not reactive with sera in the Western format but was specifically immunoprecipitated by diabetic sera from an Escherichia coli extract.

Pancreatic islet cell cytoplasmic antibody in diabetes is represented by antibodies to islet cell antigen 512 and glutamic acid decarboxylase.

The presence of serum islet cell cytoplasmic antibodies (ICAs) is a standard autoimmune marker for insulin-dependent diabetes mellitus (IDDM). The antigenic molecule(s) responsible for ICA has not been identified, although antibodies to the 65-kDa isoform of glutamic acid decarboxylase (GAD65) do contribute. We tested 129 IDDM sera for antibodies to ICA512 (anti-ICA512), antibodies to GAD (anti-GAD), and ICAs; we tested for inhibition of ICAs with purified recombinant ICA512 and sheep brain GAD; and we tested for immunofluorescence reactivity on COS7 cells transfected with cDNA clones encoding ICA512 and GAD65. The results were that anti-ICA512 antibodies contribute to ICA reactivity and that these, in combination with anti-GAD antibodies, account for most ICA reactivity in IDDM. Anti-ICA512 antibodies were present at a frequency of 51% in 61 patients with early-onset IDDM (age of onset < or = 20 years) of short duration (< or = 1 month) but only in 9% of 68 patients with an onset age of > 20 years and/or a disease duration of > 1 month. The frequency of anti-GAD antibodies in these sera was similar irrespective of duration or age of onset. Anti-ICA512 and anti-GAD antibodies were demonstrable by indirect immunofluorescence on transfected COS7 cells, and ICA could be inhibited using either recombinant ICA512 or purified brain GAD. We conclude that anti-ICA512 and anti-GAD antibodies contribute to ICA reactivity and that anti-ICA512 antibodies account for the increased frequency of ICA reactivity in early-onset IDDM of short duration.

Human islet glucokinase gene. Isolation and sequence analysis of full-length cDNA.

Pancreatic islet glucokinase (ATP:D-hexose-6-phosphotransferase) cDNAs were isolated from a human islet cDNA library in lambda-gt11. One clone (hlGLK2), 2723 bp plus additional poly(A) residues, appeared to be full length because its size was consistent with a single 2.9-kb glucokinase mRNA on Northern-blot analysis of islet RNA. This cDNA contained an open reading frame of 1395 bp from an ATG codon at position 459, encoding a predicted protein of 465 amino acids (52,000 M(r)). Comparison of the nucleotide sequences of the human islet glucokinase cDNA with that of the recently isolated human liver glucokinase cDNA revealed that the two cDNAs differed completely on their 5'-ends, followed by an identical 2204-bp overlap extending to the 3'-ends. The 5'-ends of islet and liver glucokinase cDNAs predicted proteins that differ by 15 NH2-terminal residues. The overall sequence identity (70%) between the first exons of the human islet and rat islet cDNAs suggested that the islet promoter regions, like the liver promoter regions, have been conserved through evolution. Thus, NH2-terminal differences for human liver and islet enzymes might be explained by use of alternate promoters between the two tissues, analogous to the NH2-terminal differences of the rat liver and rat islet enzymes. If so, this relationship predicts important tissue-specific regulatory functions of these regions. Variations in the glucokinase gene are likely to occur in humans. Isolation of a human islet glucokinase cDNA has provided the sequence necessary to determine whether these variants are important determinants in the genetic predisposition for diabetes mellitus.

Characterization of the LIM/homeodomain gene islet-1 and single nucleotide screening in NIDDM.

Islet-1 (Isl-1) is a unique transcription factor that binds to the enhancer region of the insulin gene. To evaluate this gene in non-insulin-dependent diabetes mellitus (NIDDM), a full-length human Isl-1 cDNA was isolated and the genomic structure was characterized. The cDNA [2,395 bp plus additional poly(A) residues] contained an open reading frame from an initiator methionine at nucleotide 240 to an opal stop codon at nucleotide 1,286 (GenBank accession number UO7559), encoding a predicted protein of 349 amino acids (39 kDa). From their ends, 23 additional clones were sequenced, revealing 15 incomplete cDNAs and 8 intron-containing partially processed precursors. As determined by Northern blotting and reverse transcriptase-polymerase chain reaction analysis, Isl-1 was most abundantly expressed as a 2.4-kb mRNA in human islets, with a restricted pattern of expression in other adult human tissues. Analysis of genomic clones revealed that Isl-1 is encoded by six exons, varying in size from 168 bp (exon 5) to 1,230 bp (exon 6). Exons 2 and 3 each encode a LIM domain, while the homeodomain is completely contained within exon 4.(ABSTRACT TRUNCATED AT 250 WORDS)

ICA512 autoantibody radioassay.

As part of a general program of screening islet expression libraries we have identified a clone from a lambda gt11 human islet expression library that reacts with human diabetic sera and, upon sequencing, was determined to be the neuroendocrine islet autoantigen ICA512 (islet cell antigen 512). In the current communication, we describe the development of a radioassay for autoantibodies to ICA512 (ICA512AA) using in vitro transcribed and translated protein for production of labeled antigen. Our initial results indicate that this radioassay is significantly more sensitive than the enzyme-linked immunosorbant assay, which uses a COOH-terminal fragment of ICA512. The ICA512AA radioassay uses a 96-well format with membrane separation of antibody bound from free antigen and should be readily adaptable to automated large-scale screening. Only 7 microliters of serum is required for triplicate determinations. In order to determine the specificity and sensitivity of this assay and estimate its positive predictive value, we have studied 42 new-onset diabetic patients, 33 first-degree relatives of diabetic patients followed to diabetes, 694 islet cell antibody-negative (ICA-) relatives, and 205 normal control subjects. Thirty-eight percent of new-onset patients and 48% of relatives followed to diabetes express autoantibodies to ICA512 exceeding the 99th percentile of the normal control subjects. In contrast, only 1.4% of ICA- first-degree relatives were positive for ICA512 autoantibodies.(ABSTRACT TRUNCATED AT 250 WORDS)

Cloning and functional expression of the human glucagon-like peptide-1 (GLP-1) receptor.

Truncated forms of glucagon-like peptide-1 are the most potent endogenous stimuli of insulin secretion and have powerful antidiabetogenic effects. To determine the structure and coupling mechanisms of the human GLP-1 receptor we have isolated two pancreatic islet cDNAs, encoding the 463 amino acid receptor and differing mainly in their 3' untranslated regions. The deduced amino acid sequence is 90% homologous with the rat GLP-1 receptor. Northern blot analysis shows expression of a single 2.7 kb transcript in pancreatic tissue. When expressed in COS-7 cells the recombinant receptor conferred specific, high affinity GLP-1(7-37) binding. GLP-1(7-37) increased intracellular cAMP in a concentration dependent manner and caused an increase in the free cytosolic calcium ([Ca2+]i) from an intracellular pool, characteristic of phospholipase C (PLC) activation. Thus, like the structurally related glucagon and parathyroid hormone receptors, the human GLP-1 receptor can activate multiple intracellular signaling pathways including adenylyl cyclase and PLC. Knowledge of the GLP-1 receptor structure will facilitate the development of receptor agonists and elucidation of the important role of GLP-1 in normal physiology and disease states.

An ELISA sandwich capture assay for recombinant fusion proteins containing glutathione-S-transferase.

A sandwich capture ELISA technique is presented for detection of recombinant proteins sharing a common affinity domain or reagents such as antibodies that bind to these proteins. An activated carrier protein (BSA) is modified with reduced glutathione (GT), forming an affinity capture reagent for glutathione-S-transferase (GST) and recombinant fusion proteins bearing the GST moiety. GT-BSA is immobilized on microtiter plates, and a sandwich is formed consisting of the recombinant fusion protein, reactive antibodies, and detection antibodies. An example is given that demonstrates that this format yields equivalent results to a conventional ELISA test with a panel of newly diagnosed diabetic sera reacting with an islet cell autoantigen.

Antibodies to SOX13 (ICA12) are associated with type 1 diabetes.

SOX13 is an islet cell autoantigen (ICA12), identified by antibody screening of an islet cDNA library, using sera from patients with Type 1 diabetes. We ascertained the frequency of antibody reactivity to SOX13 and compared it with other Type 1 diabetes autoantibody reactivities. Antibodies were measured by radioimmunoprecipitation (RIP) using (35) S labelled SOX13 expressed in rabbit reticulocyte lysate. Sera from 109 subjects with Type 1 diabetes, 29 with Type 2 diabetes, 144 with other autoimmune diseases and from 201 controls were tested for anti-SOX13, and results were compared with the frequency of antibodies to glutamic acid decarboxylase (anti-GAD), islet cell antigen 512 (anti-ICA512) and islet cell cytoplasm (ICA). Anti-SOX13 were detected in 20 (18.3%) of 109 subjects with Type 1 diabetes, and more frequently in adults than in children (29% vs 10%). Anti-SOX13 usually occurred with anti-GAD but rarely with anti-ICA512. Seven sera positive for anti-SOX13 did not react with either GAD, ICA512 or islet cell cytoplasm indicating that anti-SOX13 represented a distinct population of antibodies. Reactivity to SOX13 represents a further autoantibody response in adults with Type 1 diabetes and may provide a useful disease marker in subjects in whom other autoantibody tests are negative.

Value of ICA512 antibodies for prediction and diagnosis of type 1 diabetes.

ICA512 was isolated from an islet cDNA expression library and was identified as transmembrane protein closely related to the T-cell tyrosine phosphatase CD45. In order to determine the frequency of antibodies (ab) to ICA512, we tested sera of 124 newly diagnosed type 1 diabetic patients (IDDM) and 30 patients with long standing IDDM, 44 non-diabetic first degree relatives (FDR) with positive ICA or IAA, and 76 healthy control subjects using an ELISA. The mean +/- SD that we obtained in our control population was 4.1 +/- 3.9 U and a cut-off of 16 U was defined as normal range (mean + 3 SD). Of newly diagnosed diabetic patients and patients with long standing IDDM, 32% and 23% respectively had positive ICA512-ab with a mean of 22 +/- 33 U (vs controls p < 0.001) and 14 +/- 14 U (p < 0.01). Of antibody-positive first degree relatives, 36% were found to have elevated ICA512-ab with a mean of 24 +/- 41 U (p < 0.01). In relatives with multiple follow-up samples, ICA512-ab were found to be constantly positive or negative in 86% of cases, whereas fluctuation of ICA512-ab positivity occurred in five relatives in which three developed positive ICA512-ab and two lost ICA512-ab positivity during follow-up. Of ICA512-ab + relatives, 76% progressed to clinical type 1 diabetes within 5 years of follow-up, whereas only 24% developed diabetes in the ICA512-ab negative group (p < 0.01). ICA512-ab were more frequent in newly diagnosed diabetic children below age 15 years (p < 0.02) and in patients with positive ICA (p < 0.001) or positive IAA (p < 0.02). There was, in contrast, no correlation of ICA512-ab with GADA. One patient with newly diagnosed type 1 diabetes exclusively exhibited ICA512-ab. In conclusion, these results suggest that ICA512-ab are related to autoimmune type 1 diabetes and useful as an additional screening marker for the prediction of type 1 diabetes.

Glucagon.glucagon-like peptide I receptor chimeras reveal domains that determine specificity of glucagon binding.

The binding of glucagon to its hepatic receptor triggers a G-protein-mediated signal that ultimately leads to an increase in hepatic glucose production (gluconeogenesis) and glycogen breakdown (glycogenolysis). In order to elucidate the structural domain(s) of the human glucagon receptor (hGR) involved in the selective binding of glucagon, a series of chimeras was constructed in which various domains of the hGR were replaced by homologous regions from the receptor for the glucoincretin hormone, glucagon-like peptide I (GLP-IR). hGR and GLP-IR are quite similar (47% amino acid identify) yet have readily distinguishable ligand binding characteristics; glucagon binds to the recombinant hGR expressed in COS-7 cells with a Kd that is 1000-fold lower than the Kd for glucagon binding to GLP-IR. In the present study, chimeric receptors were transiently expressed in COS-7 cells and analyzed for glucagon binding. Expression of each receptor chimera was confirmed by immunofluorescence staining using a hGR-specific monoclonal antibody. This report identifies several non-contiguous domains of the hGR that are important for high affinity glucagon binding. Most notable are the membrane-proximal half of the amino-terminal extension, the first extracellular loop, and the third, fourth, and sixth transmembrane domains.

ICA 512, an autoantigen of type I diabetes, is an intrinsic membrane protein of neurosecretory granules.

Islet cell autoantigen (ICA) 512 is a novel autoantigen of insulin-dependent diabetes mellitus (IDDM) which is homologous to receptor-type protein tyrosine phosphatases (++PTPases). We show that ICA 512 is an intrinsic membrane protein of secretory granules expressed in insulin-producing pancreatic beta-cells as well as in virtually all other peptide-secreting endocrine cells and neurons containing neurosecretory granules. ICA 512 is cleaved at its luminal domain and, following exposure at the cell surface, recycles to the Golgi complex region and is sorted into newly formed secretory granules. By immunoprecipitation, anti-ICA 512 autoantibodies were detected in 15/17 (88%) newly diagnosed IDDM patients, but not in 10/10 healthy subjects. These results suggest that tyrosine phosphorylation participates in some aspect of secretory granule function common to all neuroendocrine cells and that a subset of autoantibodies in IDDM is directed against an integral membrane protein of insulin-containing granules.

Islet cell antigen 512 is a diabetes-specific islet autoantigen related to protein tyrosine phosphatases.

Islet cell Ag 512 (ICA512) is a recombinant human Ag that was isolated from an islet cDNA expression library by screening with human insulin-dependent diabetes mellitus sera. Specificity of reaction with diabetic sera was demonstrated initially by immunoprecipitation with a small number of diabetic and normal serum samples. To permit quantitative and rapid serum testing, ICA512 was purified and adapted to an ELISA format. In this way, a sensitivity of 48% with newly diagnosed diabetic sera has been measured with a panel of 80 sera. DNA sequencing of ICA512-3, a cDNA that contains a 1644 bp open reading frame, suggests that it codes for a transmembrane protein having a single membrane-spanning segment and a cytoplasmic domain that is closely related to the first intracellular (catalytic) domain of the T cell protein tyrosine phosphatase, CD45. Northern blot analysis of poly(A)+ RNAs from several human tissues indicates that ICA512 mRNA is expressed in brain and pancreas.

Insulin signaling in mice expressing reduced levels of Syp.

Syp is a protein tyrosine phosphatase implicated in insulin and growth factor signaling. To evaluate the role of syp in insulin's regulation of plasma glucose, we generated knockout mice. Homozygous knockout mice die prior to day 10.5 of embryonic development. Hemizygous mice express half the levels of syp protein compared with their wild type littermates but do not display any gross morphological changes. Total body weight (age 2-10 weeks) and plasma insulin and glucose levels both in fasting and glucose-challenged states were comparable in the wild type and the hemizygous mice. No differences were observed in insulin-induced glucose uptake in soleus muscle and epididymal fat; insulin inhibition of lipolysis was also similar. We injected insulin into the portal vein of the mice to examine upstream events of the insulin signaling cascade. Tyrosine phosphorylation of insulin receptor and insulin receptor substrate-1 (IRS-1) from hemizygous tissue was similar to that of wild type tissue. Association of the p85 subunit of phosphatidylinositol 3-kinase to IRS-1 increased an average of 2-fold in both groups. We did not observe an increase of IRS-1/syp association after insulin administration, but we did note a significant basal association in both wild type and hemizygous tissue. Our results do not support a major role for syp in the acute in vivo metabolic actions of insulin.