Islet-enriched gene expression and glucose-induced insulin secretion in human and mouse islets.

AIMS/HYPOTHESIS: Our understanding of the transcription factors that control the development and function of rodent islet beta cells is advancing rapidly, yet less is known of the role they play in similar processes in human islets. METHODS: To characterise the abundance and regulation of key proteins involved in glucose-regulated insulin secretion in human islets, we examined the expression of MAFA, MAFB, GLUT2 (also known as SLC2A2), ßGK (also known as GCK) and PDX1 in isolated, highly purified human islets with an intact insulin secretory pattern. We also assessed these features in islets from two different mouse strains (C57BL/6J and FVB). RESULTS: Compared with mouse islets, human islets secreted more insulin at baseline glucose (5.6 mmol/l), but less upon stimulation with high glucose (16.7 mmol/l) or high glucose plus 3-isobutyl-1-methyl-xanthine. Human islets had relatively more MAFB than PDX1 mRNA, while mouse islets had relatively more Pdx1 than Mafb mRNA. However, v-maf musculoaponeurotic fibrosarcoma oncogene homologue (MAF) B protein was found in human islet alpha and beta cells. This is unusual as this regulator is only produced in islet alpha cells in adult mice. The expression of insulin, MAFA, ßGK and PDX1 was not glucose-regulated in human islets with an intact insulin secretory pattern. CONCLUSIONS/INTERPRETATION: Our results suggest that human islets have a distinctive distribution and function of key regulators of the glucose-stimulated insulin secretion pathway, emphasising the urgent need to understand the processes that regulate human islet beta cell function.

Aldosterone decreases glucose-stimulated insulin secretion in vivo in mice and in murine islets.

AIMS/HYPOTHESIS: Aldosterone concentrations increase in obesity and predict the onset of diabetes. We investigated the effects of aldosterone on glucose homeostasis and insulin secretion in vivo and in vitro. METHODS: We assessed insulin sensitivity and insulin secretion in aldosterone synthase-deficient (As [also known as Cyp11b2](-/-)) and wild-type mice using euglycaemic-hyperinsulinaemic and hyperglycaemic clamps, respectively. We also conducted studies during high sodium intake to normalise renin activity and potassium concentration in As (-/-) mice. We subsequently assessed the effect of aldosterone on insulin secretion in vitro in the presence or absence of mineralocorticoid receptor antagonists in isolated C57BL/6J islets and in the MIN6 beta cell line. RESULTS: Fasting glucose concentrations were reduced in As (-/-) mice compared with wild-type. During hyperglycaemic clamps, insulin and C-peptide concentrations increased to a greater extent in As (-/-) than in wild-type mice. This was not attributable to differences in potassium or angiotensin II, as glucose-stimulated insulin secretion was enhanced in As (-/-) mice even during high sodium intake. There was no difference in insulin sensitivity between As (-/-) and wild-type mice in euglycaemic-hyperinsulinaemic clamp studies. In islet and MIN6 beta cell studies, aldosterone inhibited glucose- and isobutylmethylxanthine-stimulated insulin secretion, an effect that was not blocked by mineralocorticoid receptor antagonism, but was prevented by the superoxide dismutase mimetic tempol. CONCLUSIONS/INTERPRETATION: We demonstrated that aldosterone deficiency or excess modulates insulin secretion in vivo and in vitro via reactive oxygen species and in a manner that is independent of mineralocorticoid receptors. These findings provide insight into the mechanism of glucose intolerance in conditions of relative aldosterone excess.

Islet cell autoantigen 69 kD (ICA69). Molecular cloning and characterization of a novel diabetes-associated autoantigen.

We have identified a novel 69-kD peptide autoantigen (ICA69) associated with insulin-dependent diabetes mellitus (IDDM) by screening a human islet lambda gt11 cDNA expression library with cytoplasmic islet cell antibody positive sera from relatives of IDDM patients who progressed to the overt disease. The deduced open reading frame of the ICA69 cDNA predicts a 483-amino acid protein. ICA69 shows no nucleotide or amino acid sequence relation to any known sequence in GenBank, except for two short regions of similarity with BSA. The ICA69 cDNA probe hybridizes with a 2-kb mRNA in poly(A+) RNA from human pancreas, brain, heart, thyroid, and kidney, but not with skeletal muscle, placenta, spleen, or ovary. Expression of ICA69 was also detected in beta cells and cell lines, as well as in tumoral tissue of islet cell origin. The native ICA69 molecule migrates to 69 kD in SDS-PAGE as detected with specific antibodies. Serum samples from relatives of IDDM patients specifically reacted with affinity-purified recombinant ICA69 on Western blotting. The structural gene for ICA69 was designated ICA1. A homologue in the mouse, designated Ica-1 was mapped to the proximal end of chromosome 6 (within 6 cM of the Met protooncogene). ICA69 adds a novel autoantigen to the family of identified islet target molecules, and by the manner of its identification and characterization large amounts of antigen are available for development of quantitative, convenient predictive assays for autoantibodies and analysis of the role of this molecule in diabetes autoimmunity, as well as its physiologic function.

Coexpression of glucose transporters and glucokinase in Xenopus oocytes indicates that both glucose transport and phosphorylation determine glucose utilization.

A Xenopus oocyte expression system was used to examine how glucose transporters (GLUT 2 and GLUT 3) and glucokinase (GK) activity affect glucose utilization. Uninjected oocytes and low rates of both glucose transport and phosphorylation; expression of GLUT 2 or GLUT 3 increased glucose phosphorylation approximately 20-fold by a low Km, endogenous hexokinase at glucose concentrations < or = 1 mM, but not at higher glucose concentrations. Coexpression of functional GK isoforms with GLUT 2 or 3 increased glucose utilization approximately an additional two- to threefold primarily at the physiologic glucose concentrations of 5-20 mM. The Km for glucose of both the hepatic and beta cell isoforms of GK, determined in situ, was approximately 5-10 mM when coexpressed with either GLUT 2 or GLUT 3. The increase in glucose utilization by coexpression of GLUT 3 and GK was dependent upon glucose phosphorylation since two missense GK mutations linked with maturity-onset diabetes, 182: Val-->Met and 228:Thr-->Met, did not increase glucose utilization despite accumulation of both a similar amount of immunoreactive GK protein and glucose inside the cell. Coexpression of a mutant GK and a normal GK isoform did not interfere with the function of the normal GK enzyme. Since the coexpression of GK and a glucose transporter in oocytes resembles conditions in the hepatocyte and pancreatic beta cell, these results indicate that increases in glucose utilization at glucose concentrations > 1 mM depend upon both a functional glucose transporter and GK.

An encapsulation system for the immunoisolation of pancreatic islets.

Over a thousand combinations of polyanions and polycations were tested to search for new polymer candidates that would be suitable for encapsulation of living cells. The combination of sodium alginate, cellulose sulfate, poly (methylene-co-guanidine) hydrochloride, calcium chloride, and sodium chloride was most promising. In parallel, a novel multiloop chamber reactor was developed to control the time of complex formation and to negate gravitational effects such as pancreatic islet sedimentation and droplet deformation during the encapsulation process. Encapsulated rat islets demonstrated glucose-stimulated insulin secretion in vitro, and reversed diabetes in mice. This new capsule formulation and encapsulation system allows independent adjustments of capsule size, wall thickness, mechanical strength, and permeability, which may offer distinct advantages for immunoisolating cells.

Glucose transport activity and ligand binding (cytochalasin B, IAPS-forskolin) of chimeric constructs of GLUT2 and GLUT4 expressed in COS-7-cells.

Chimeric constructs of glucose transporters GLUT2 and GLUT4 were transiently expressed in COS-7 cells in order to determine regions of the proteins responsible for their differences in activity and ligand binding. Exchange of the C-terminal tail (aa 479-509) of GLUT4 failed to affect glucose transport activity assayed at 1 mM glucose or ligand binding (cytochalasin B, IAPS-forskolin). In contrast, exchange of the C-terminal half of GLUT4 (aa 222-509) for that of GLUT2 markedly reduced ligand binding (Kd of cytochalasin B binding 1.88 +/- 0.2 microM vs. 0.21 +/- 0.06 in the wild-type GLUT4), and moderately (25%) reduced glucose transport activity. These data support the conclusion that the domains determining differences in ligand binding between GLUT4 and GLUT2 are located in the C-terminal half of the glucose transporters.

Two distinct glutamic acid decarboxylase auto-antibody specificities in IDDM target different epitopes.

Although most individuals with insulin-dependent diabetes mellitus (IDDM) have autoantibodies to glutamic acid decarboxylase (GAD), antibodies to GAD are also present in some individuals with a low risk of developing diabetes. The GAD autoantibodies of IDDM are specific for the GAD65 isoform, do not bind denatured GAD protein, and target epitope(s) dependent on conformation of the protein. However, the IDDM epitopes have been difficult to further define because the antibodies do not bind GAD protein fragments or synthetic peptides. Since the GAD67 isoform is highly homologous to GAD65 but is usually not a target of the GAD autoantibodies in IDDM sera, we created six GAD65/GAD67 chimeric proteins to maintain the overall GAD protein conformation and used these chimeric proteins to map conformation-dependent epitopes of GAD65 targeted by IDDM sera. We find that the GAD binding present in most IDDM sera (n = 11 of 12) is composed of two distinct GAD antibody specificities that target different conformation-dependent regions of the GAD65 protein, one that is located between amino acids 240 and 435 (termed IDDM-E1) and one that is located between amino acids 451 and 570 (termed IDDM-E2). One IDDM serum (n = 1 of 12) bound only the IDDM-E1 region. Identification of epitopes targeted by IDDM sera may allow one to distinguish between GAD antibody-positive individuals at high and low risk of developing IDDM and to determine if differences in the autoimmune repertoire directed at GAD are present. The chimeric GAD65/GAD67 proteins may also be useful in designing GAD assays specific for IDDM.

Insulin degradation by mononuclear cells.

Mononuclear cells from peripheral blood possess insulin receptors that are altered in number or binding affinity in certain metabolic diseases as obesity. The monocyte, and not the lymphocyte, is the cell with the capacity to specifically bind insulin. Furthermore, this binding appears to mirror the receptor status on such insulin target tissues as liver, muscle, and fat. Since liver, muscle, and fat also degrade insulin, mononuclear cells from the blood of normal volunteers were examined for insulin-degrading activity. Intact cells were incubated with 125I-insulin and the amount of degraded insulin was measured by the trichloroacetic acid-precipitation technique. Insulin-degrading activity increased when the number of cells and the time of incubation were increased. Total insulin binding behaved in a similar fashion. Very little degradation was seen at 4 degrees or 15 degrees. The Km for insulin-degrading activity was 7.03 X 10(-8) M. Homogenized mononuclear cells degraded two to five times more insulin than did intact cells and also demonstrated cell concentration, time, and temperature dependence for degradation. The Km of degradation for homogenized mononuclear cells was 2.2 X 10(-8) M. Subcellular fractionation revealed significant degrading activity in the 100,000 X g supernatant, but little activity in the 100,000 X g pellet. A purified lymphocyte preparation did not bind insulin and contained little insulin-degrading activity.

Sodium butyrate increases glucagon and insulin gene expression by recruiting immunocytochemically negative cells to produce hormone.

Current evidence suggests that a multipotential endodermal progenitor cell may give rise to all islet cell phenotypes. We characterized two hormone-producing rat islet (RIN) cell lines derived from a radiation-induced islet tumor by immunocytochemistry, Northern blot analysis, and radioimmunoassay of secreted hormone. Using antisera to glucagon, insulin, and somatostatin, we found that less than 15% of the cells in any of these three islet cell lines contained immunopositive cells. The number of cells staining for the hormone correlated with mRNA levels and immunoreactive secreted hormones. Sodium butyrate, a short-chain aliphatic fatty acid, slowed cell growth and increased dramatically the percentage of cells staining for glucagon and insulin. The increase in immunopositive cells was accompanied by an increase in glucagon and insulin mRNAs and secreted glucagon and insulin. These observations indicate that sodium butyrate increases glucagon and insulin gene expression by recruiting previously immunonegative cells to produce hormone. The relationship of DNA synthesis and hormone production was assessed by pulse-labeling RIN cells with [3H]thymidine, which was followed by autoradiography and immunocytochemistry. [3H]thymidine incorporation was observed in a lower percentage of immunopositive compared with immunonegative cells. Furthermore, sodium butyrate reduced the number of [3H]thymidine-labeled cells and increased the number of immunopositive cells. These observations suggest that sodium butyrate differentiates the islet cells and thereby increases the expression of the glucagon and insulin genes.

Proglucagon processing similar to normal islets in pancreatic alpha-like cell line derived from transgenic mouse tumor.

A pancreatic alpha-like cell line has been established from a glucagonoma arising in transgenic mice expressing a hybrid gene consisting of the rat glucagon-promoter sequence fused to the sequence encoding the SV40 T-antigen oncoprotein. The alpha-tumor cell 1 (alpha TC1) line maintained many characteristics of differentiated alpha-cells for greater than 40 passages in culture and expressed levels of glucagon mRNA 5- to 10-fold higher than those reported previously in rat and hamster islet cell lines. By radioimmunoassay, the cells synthesized considerable amounts of glucagon, glucagonlike peptide I (GLP-I), the major proglucagon fragment, and small amounts of unprocessed proglucagon but no free GLP-II. This distribution of peptides is similar to that found in extracts of rodent pancreases and is distinct from that seen with other islet cell lines, which process proglucagon in patterns more characteristic of intestinal cells. The GLP-I peptide in the alpha TC1 cell line was in the form of GLP-I-(1-37), which is inactive as a stimulator of insulin secretion, and not GLP-I-7-37) or -(7-36)-amide peptides, both of which are potent insulin secretagogues. The alpha TC1 cell line produced glucagon-related peptides in a relatively uniform pattern by immunocytochemistry, and electron microscopy revealed typical alpha-type (glucagon) secretory granules. Although the cell line was derived from an islet tumor producing only glucagon, the alpha TC1 cell line also produced insulin in addition to the glucagon peptides.(ABSTRACT TRUNCATED AT 250 WORDS)

Expression of peptide hormone genes in human islet cell tumors.

The embryogenesis of the pancreas suggests the existence of a common stem cell progenitor of the four islet cell types (insulin, glucagon, somatostatin, and pancreatic polypeptide). We investigated whether neoplastic islet tumors express multiple hormone-specific cellular phenotypes of the islets. By analyses of RNA transcripts and immunoreactive peptides in four human insulinomas and one glucagonoma, we found that the insulin, somatostatin, and glucagon genes were coexpressed in all tumors. The expression of the three hormone genes in a lymph node metastasis of a glucagonoma reduced the possibility that contamination of tumor tissue by normal islets occurred. These observations lend further support to the hypothesis of the multipotentiality of neoplastic islet cells for the expression of genes encoding several different islet hormones.

Identification of glutamic acid decarboxylase autoantibody heterogeneity and epitope regions in type I diabetes.

Glutamic acid decarboxylase (GAD) is an autoantigen of the islet cell antibodies (ICAs) present in type I diabetes. GAD autoantibodies are also found in patients with stiffman syndrome and in certain ICA-positive individuals who rarely develop diabetes on long-term follow-up. This latter subset of ICA has been termed restricted or beta-cell-specific ICA because the antibodies react with only the beta-cells of the islet. By immunoprecipitation of recombinant GAD65 and GAD67 protein and protein fragments, 83% of sera from individuals with new-onset diabetes or prediabetes (n = 30) had GAD65 autoantibodies, but only 26% had GAD67 autoantibodies. In contrast, all restricted ICA sera (n = 6) had both GAD65 and GAD67 autoantibodies. In both types of sera, the binding of GAD67 autoantibodies could be blocked by preincubation of the serum with GAD65 and GAD67, but the binding of GAD65 autoantibodies could not be blocked by preincubation with GAD67. The titer of GAD65 autoantibodies was much higher in the restricted ICA sera (titer > 1:1,000) than in the sera from individuals with new-onset diabetes or prediabetes (titer < 1:100) and was reflected by the greater amount of GAD65 protein immunoprecipitated by restricted ICA sera (2.61 +/- 1.39 U) compared with sera from individuals with new-onset diabetes (0.51 +/- 0.34 U). The restricted ICA sera immunoprecipitated equimolar amounts of GAD65 protein fragments, suggesting a non-conformational or linear epitope; epitope mapping localized the major epitope region to amino acids 361-442 and a second minor epitope region to amino acids 1-195.(ABSTRACT TRUNCATED AT 250 WORDS)

Enhancer and promoter element interactions dictate cyclic adenosine monophosphate mediated and cell-specific expression of the glycoprotein hormone alpha-gene.

The glycoprotein hormone alpha-gene is preferentially expressed in placental cell lines, but it is also expressed in several other cell lines indicating that the differential activity of the alpha-gene regulatory elements in various cell types is more quantitative than qualitative. The 5'-flanking region of the alpha-gene contains several distinct DNA regulatory sequences including an upstream regulatory element [(URE) -181 to -150 base pairs (bp)] that stimulates basal expression and an 18 bp twice-repeated cAMP-responsive element [(CRE) -146 to -111 bp]. We constructed an array of fusion genes containing the URE and/or the CRE linked to different truncated promoters [alpha-gene, somatostatin (SRIF), glucagon, Simian Virus 40]. These constructions were transiently expressed in placental, fibroblast, or islet cell lines to identify regulatory sequences involved in cell-specific expression as well as interactions between the URE, the CRE, and different promoter elements. The URE, CRE, and alpha-promoter elements contribute approximately 3-, 6-, and 5-fold, respectively, to preferential expression in JEG-3 cells. In JEG-3 cells, the URE is strictly dependent on the CRE for activity, but it functions in a promoter-independent manner. In contrast, the CRE is markedly promoter dependent. When linked to heterologous enhancers, the alpha-promoter is more active in JEG-3 cells than in other cell lines, thereby contributing substantially to preferential expression in placental cells. Although the CREs derived from the alpha and SRIF genes both activate expression of the alpha promoter, only the alpha CRE activates the SRIF promoter in JEG-3 cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Phospholipid and glutamic acid decarboxylase autoantibodies in diabetic neuropathy.

OBJECTIVE: To determine the prevalence and significance of phospholipid autoantibodies (PLAs) and glutamic acid decarboxylase (GAD) autoantibodies in the circulation of normal patients and diabetic patients with and without neuropathy. RESEARCH DESIGN AND METHODS: We measured PLAs in a total of 78 patients (a diabetic group with somatic or autonomic neuropathy [n = 40] another group without neuropathy [n = 38]), and GAD autoantibodies in a subset of 22 patients. RESULTS: PLAs are found in 2% of the general population. We found PLAs in 32% of the diabetic population without neuropathy, in 88% of those with neuropathy, in 55% of those with retinopathy, and in 25% of those with established nephropathy. The frequencies of immunoglobulins in the neuropathic group were: IgG = 78%, IgM = 33%, and IgA = 23%. There was no correlation between PLAs and microalbuminuria, macrovascular disease, fibrinogen, duration of diabetes, or neuropathy, but there was a strong correlation with total neuropathy score. Sera with high PLA IgG titers bound to the surface of neuroblastoma cells and inhibited cell growth. Antibodies to GAD65 were present in 32% and to GAD67 in 0% of patients. No titers of GAD65, GAD67, or the GAD65 ratio were associated with the degree of neuropathy of the presence of PLAs. CONCLUSIONS: PLAs occur frequently in the sera of patients with diabetes and correlate with the extent of neuropathy, suggesting a role for PLAs in the etiology thereof. The measurement of PLAs may constitute a marker for ongoing damage to nerves.

Different functional domains of GLUT2 glucose transporter are required for glucose affinity and substrate specificity.

GLUT2 is the major glucose transporter in pancreatic beta-cells and hepatocytes. It plays an important role in insulin secretion from beta-cells and glucose metabolism in hepatocytes. To better understand the molecular determinants for GLUT2's distinctive glucose affinity and its ability to transport fructose, we constructed a series of chimeric GLUT2/GLUT3 proteins and analyzed them in both Xenopus oocytes and mammalian cells. The results showed the following. 1) GLUT3/GLUT2 chimera containing a region from transmembrane segment 9 to part of the COOH-terminus of GLUT2 had Km values for 3-O-methylglucose similar to those of wild-type GLUT2. Further narrowing of the GLUT2 component in the chimeric GLUTs lowered the Km values to those of wild-type GLUT3. 2) GLUT3/GLUT2 chimera containing a region from transmembrane segment 7 to part of the COOH-terminus of GLUT2 retained the ability to transport fructose. Further narrowing of this region in the chimeric GLUTs resulted in a complete loss of the fructose transport ability. 3) Chimeric GLUTs with the NH2-terminal portion of GLUT2 were unable to express glucose transporter proteins in either Xenopus oocytes or mammalian RIN 1046-38 cells. These results indicate that amino acid sequences in transmembrane segments 9-12 are primarily responsible for GLUT2's distinctive glucose affinity, whereas amino acid sequences in transmembrane segments 7-8 enable GLUT2 to transport fructose. In addition, certain region(s) of the amino-terminus of GLUT2 impose strict structural requirements on the carboxy-terminus of the glucose transporter protein. Interactions between these regions and the carboxy-terminus of GLUT2 are essential for GLUT2 expression.

Characterization of monoclonal antibody 3G5 and utilization of this antibody to immobilize pancreatic islet cell gangliosides in a solid phase radioassay.

Monoclonal antibody 3G5, which was initially produced by immunization of mice with fetal rat brain, reacts specifically by indirect immunofluorescence with all cells of the pancreatic islets of human, rat, mouse, and bovine pancreas. This antibody reacts with the cell surface of isolated islet cells as well as the rat (RIN5F) insulinoma cell line. Antibody 3G5 reacts with islets, thyroid follicular cells, pituitary, and the adrenal medulla of a pattern similar to but distinct from those of antibody A2B5 and tetanus toxin, both of which react with complex gangliosides (sialic acid-containing glycosphingolipids). The antigen with which antibody 3G5 reacts also has the properties of ganglioside (neuraminidase sensitive, extracted into chloroform-methanol, partitioned into a methanol-water phase, soluble in water, and nondialyzable). Antibody 3G5, adsorbed to polyvinyl plates, can immobilize islet ganglioside micelles to which 125I-labeled 3G5, A2B5, and tetanus toxin all bind. The ability to immobilize micelles containing several complex gangliosides has led to a solid phase radioassay to detect antiganglioside antibodies. Monoclonal antibody 3G5 joins antibody A2B5 and tetanus toxin as markers for distinct complex gangliosides found on pancreatic islets and neurons.

Two regions of GLUT 2 glucose transporter protein are responsible for its distinctive affinity for glucose.

The glucose transporter in the hepatocyte and pancreatic beta-cell (GLUT 2) has a lower affinity for glucose than other members of the glucose transporter family. To investigate the molecular mechanism for the distinctive affinity of GLUT 2 for glucose, we expressed chimeric GLUT 2 and GLUT 4 proteins in Xenopus oocytes and measured 3-O-methyl-D-glucose transport. In the oocyte system, GLUT 2 had a Km of 31.8 +/- 2.8 mM for 3-O-methyl-D-glucose, whereas GLUT 4 had a Km of 7.2 +/- 2.4 mM under equilibrium exchange conditions. GLUT 4/GLUT 2 chimera that contained the intracellular loop and transmembrane domains 7-12 of GLUT 2 (amino acids 239-497) had a Km similar to that of wild-type GLUT 2. A GLUT 4/GLUT 2 chimera in which the COOH-terminal 30 amino acids of GLUT 4 were replaced with the corresponding region of GLUT 2 had a 2-fold higher Km than GLUT 4, but still had a much lower Km than GLUT 2. These results indicate that both transmembrane domains 7-12 and the COOH-terminus of the protein are responsible for the distinctive glucose affinity of GLUT 2.

Increased insulin-like growth factor II production and consequent suppression of growth hormone secretion: a dual mechanism for tumor-induced hypoglycemia.

We investigated the pathophysiology of fasting hypoglycemia associated with large tumors of mesenchymal origin. We studied two patients with symptomatic fasting hypoglycemia (plasma glucose, 1.92 and 2.03 mmol/L) and a large mesenchymal neoplasm. Before therapy, the plasma insulin-like growth factor II (IGF-II) level measured by RIA was elevated (1879 and 1084 micrograms/L; normal range, 358-854 micrograms/L), the serum GH response to hypoglycemia was impaired, and the plasma IGF-I level was low in both patients. After treatment of the tumor, all of these abnormalities resolved in both patients. Northern blot analysis of tumor RNA revealed extremely high levels of IGF-II mRNA (greater than 100-fold higher than those in normal adult liver). Tumor fragments released IGF-II into tissue culture medium (2.1 and 7.2 micrograms IGF-II/g tissue.24 h). These findings indicate that secretion of IGF-II into the circulation by the tumor was the likely source of the elevated plasma IGF-II levels. We suggest that the primary event in tumor-induced hypoglycemia is overproduction of IGF-II by the tumor, which gives rise to hypoglycemia by a dual mechanism: increased glucose utilization mediated by the insulin-like actions of IGF-II and inhibition of GH secretion.

Human B cells secreting immunoglobulin G to glutamic acid decarboxylase-65 from a nondiabetic patient with multiple autoantibodies and Graves' disease: a comparison with those present in type 1 diabetes.

Antibodies to glutamic acid decarboxylase-65 (GAD65) are present in a number of autoimmune disorders, such as insulin-dependent (type 1) diabetes mellitus (IDDM), stiff man syndrome, and polyendocrine autoimmune disease. Antibodies to GAD in IDDM patients usually recognize conformation-dependent regions on GAD65 and rarely bind to the second isoform, glutamic acid decarboxylase-67 (GAD67). In contrast, those present in stiff man syndrome and polyendocrine disease commonly target the second isoform (GAD67) and include antibodies that are less dependent on the conformation of the molecule. By immortalizing peripheral blood B cells with Epstein-Barr virus, we have generated three human IgG autoantibodies, termed b35, b78, and b96, to GAD65 from one patient with multiple autoantibodies to endocrine organs and Graves' disease. All three autoantibodies are of the IgG1 isotype, with islet cell activity, and do not react with GAD67. The regions on GAD65 recognized by the three autoantibodies have been investigated by immunoprecipitation with a series of chimeras, by binding to denatured and reduced antigens, and using protein footprinting techniques. Using chimeric GAD proteins, we have shown that b35 targets the IDDM-E1 region of GAD65 (amino acids 240-435) whereas both b78 and b96 target the IDDM-E2 region of GAD65 (amino acids 451-570). Furthermore, examination of binding to recombinant GAD65 and GAD67 by Western blotting revealed some differences in epitope recognition, where only b78 bound denatured and reduced GAD65. However, b35, b78, and b96 autoantibodies had different footprinting patterns after trypsin treatment of immune complexes with GAD65, again indicating different epitope recognition. Our results indicate that antibodies to GAD65 present in nondiabetic patients with multiple autoantibodies to endocrine organs show similarities to those in IDDM (by targeting IDDM-E1 and IDDM-E2 regions of GAD65) as well as subtle differences in epitope recognition (such as binding to denatured and reduced GAD65 and by protein footprinting). Thus, the GAD65 epitopes recognized by autoantibodies in different autoimmune diseases may overlap and be more heterogeneous than previously recognized.