Possible toxic effects of normal and diabetic patient serum on pancreatic B-cells

Diagnosis of Flier's syndrome in a patient with nondiabetic hypoglycemia: a case report and critical appraisal of the literature

PURPOSE

Autoimmune hypoglycemia includes rare syndromes characterized by the presence of either anti-insulin antibodies (IAA) (Hirata's disease) or anti-insulin receptor (anti-ISR) antibodies (Flier's syndrome). Diagnosis is usually based on identification of the specific antibodies, in presence of the Whipple triad. However, most of these cases are classified as idiopathic diseases due to the difficulty to define the pathogenic culprit.

METHODS

Basic research methodologies, including Western Blot and ELISA tests, have been used in this study.

RESULTS

We describe a 21-year-old young woman (PT), non-obese and non-diabetic, with a positive history of autoimmune diseases, admitted to the hospital for recurrent episodes of severe symptomatic hypoglycemia. Counterregulatory response to hypoglycemia was normal as well as the fasting test, so excluding both hormone deficiencies and insulinoma. Since an autoimmune hypoglycemic syndrome was suspected, the hyperactivation of the insulin pathway was experimentally evaluated. At this purpose, human hepatocarcinoma (HepG2) cells were incubated with serum obtained from the patient (PT) and from control individuals. Interestingly, a significant increase of phosphorylation of insulin receptor, Akt, and ERK1/2 was observed in the HepG2 cells incubated with PT serum compared with the controls. ELISA tests revealed significantly increased levels of anti-ISR antibodies in PT serum, while IAA were similar both in PT and in control sera, supporting diagnosis of Flier's syndrome.

CONCLUSIONS

This study emphasizes the importance to identify new strategies for the differential diagnosis of hypoglycemia, not always possible with the routinely used diagnostic tests.

Human amniotic fluid obtained from diabetic women. A potent stimulator of islet cell replication

Recently, human amniotic fluid (HAF) from healthy women was found to stimulate growth and function of pancreatic B-cells. Here, the effect of HAF and serum from healthy probands (HS) was compared with that from probands with gestational (GD), noninsulin-dependent (NIDDM), or insulin-dependent diabetes (IDDM) on islet function and replication. Rat islets were cultured in the presence of either HAF or HS for 7 d. Insulin content and basal insulin release were not different after exposure of the islets to HAF or HS from healthy or diabetic women. In contrast to HS, HAF provoked the islets to deliver significantly more insulin during culture. Additionally, the same islets exhibited a more intense response to a glucose challenge. The degree of HAF-induced insulin release was not influenced by the type of diabetes. HAF and HS from GD and NIDDM women did not influence the islet DNA synthesis in comparison to HAF and HS from healthy pregnant women. However, HAF but not HS from IDDM pregnant women, elicited a significant increase in islet replication. Most effective in stimulating islet cell replication were HAFs from IDDM pregnant women belonging to the White D-type. It was shown that the relatively high concentration of insulin in the HAFs was not directly responsible for the observed increase of the islet DNA synthesis. HAF from women with long-term diabetes is supposed to contain factor(s) that might directly or indirectly enhance islet replication.

Complement activation via the alternative pathway in a patient with insulin-dependent diabetes mellitus

A 41-year old male with insulin-dependent diabetes mellitus previously unsuccessfully treated with a controlled diet and glibenclamide, and subsequently with increasing insulin doses (5 and 20 IU/day) experienced polyuria, glycosuria and loss of weight. On admittance to hospital serum C3 concentrations were found to be depressed. The insulin dose was further increased to 30 IU/day and the patient was also treated with 20 mg nafamostat mesylate given intravenously twice daily for 6 days. On completion of nafamostat mesylate treatment serum C3 concentrations were increased but after 17 days they started to decrease again.

Does protease inhibitor inhibit complement activation caused by the immune complex associated with islet cell surface antibody?

Sera containing islet cell surface antibody were obtained from seven children with insulin-dependent diabetes mellitus soon after the onset of disease. After incubation of 51Cr-labelled rat islet cells with islet cell surface antibody, human AB-type serum with or without nafamostat mesylate was added before further incubation. Radioactivity in the supernatant was measured to determine complement-dependent antibody-mediated cytotoxicity. Cytotoxicity in untreated sera [mean (+/- SD) 19.4 +/- 4.0%] was significantly (P less than 0.001) inhibited by ethyleneglycoltetraacetic acid (EGTA) (7.1 +/- 4.9%), ethylenediaminetetraacetic acid (EDTA) (2.5 +/- 0.9%) and nafamostat mesylate (2.8 +/- 1.8%). Cytotoxicity of nafamostat mesylate-treated serum was significantly (P less than 0.05) lower than that of EGTA-treated serum but not significantly different from that of EDTA-treated serum. There was no difference in cytotoxicity between nafamostat mesylate-treated and untreated, inactivated human serum. The results indicate that the protease inhibitor nafamostat mesylate completely inhibited the complement activation of the immune complex associated with islet cell surface antibody by the classical and alternative pathways.

Immunological aspects of diabetes mellitus: prospects for pharmacological modification

It is now well known that insulin-dependent diabetes is a chronic progressive autoimmune disease. The prolonged prediabetic phase of progressive beta-cell dysfunction is associated with immunological abnormalities. A prediabetic period is suggested by the appearance of islet cell antibodies, anti-insulin antibodies, and anti-insulin receptor antibodies. The existence of activated T lymphocytes and abnormal T cell subsets are also other markers. There is still no concensus about the use of the immunosuppression superimposed upon conventional insulin therapy in early diagnosed IDDM and the follow-up of the relatives of IDDM patients who share the genetic predisposition and serological markers for the risk of future onset of IDDM. Treatment in the prodromal period cannot be justified because a link between the disease and early markers such as ICA has not been established with certainty (Diabetes Research Program NIH, 1983). Many immunopharmacological manipulations were reported to be effective in animal models. However, most of them are not readily applied to human subjects. Moreover, IDDM patients are now believed to be heterogeneous, with a complex genetic background. HLA-DR, and more recently DQ, are closely related to the genetic predisposition to IDDM but those genes are not themselves diabetogenic. The contribution of autoimmunity does not appear to be uniform, and in some cases, the contribution of virus is considered more important. There is a lack of a marker for the future onset of IDDM. ICA and ICSA were found after mumps infection, but the existence of those autoantibodies and even the co-existence of HLA-DR3 do not always indicate the future trend to insulin dependency. More precise markers will be disclosed through the biochemical analysis of the target antigens on pancreatic beta-cell for islet antibodies and effector T cells. Much safer and more effective immunopharmacological treatment will be developed through animal experimentation using rat and mouse models. The recent development and interest in this field will further facilitate the attainment of the goal for the complete prevention of IDDM.

Serum C3 and C4 levels and complement-dependent antibody-mediated cytotoxic activity of islet cell surface antibody in type 1 (insulin-dependent) diabetic children

The role of complement in the pathogenesis of diabetes was studied in 31 Type 1 (insulin-dependent) diabetic children by assaying serum islet cell surface antibody, C3, C4 and serum complement-dependent antibody-mediated cytotoxicity. Nine of 21 islet cell surface antibody-positive children were within 5 months of disease onset and showed significantly lower serum C3 and C4 levels than either 1 year later or the remainder of the islet cell surface antibody-positive children at 6-12 months after disease onset. The overall trend of all islet cell surface antibody-positive diabetic children within 1 year of disease onset was toward increased serum C3 and C4 levels as the disease progressed. Serum C4 concentration and complement-dependent antibody-mediated cytotoxicity which showed an initial negative correlation were uncorrelated 1 year later. Four children who were initially strongly islet cell surface antibody-positive but negative 1 year later also exhibited significantly higher (p less than 0.05) mean serum C4 levels after 1 year. There was a significant decrease in complement-dependent antibody-mediated cytotoxicity when sera from the diabetic children were treated with either ethylene glycol tetra-acetic acid or ethylene diamine tetra-acetic acid. These data strongly suggest that complement-dependent antibody-mediated cytotoxicity induced by the classical complement pathway involving an islet cell surface antibody may play an important role in the pathogenesis of Type 1 diabetes.

Non-selective inhibition of insulin and glucagon release by xenogeneic islet cell surface antibodies

The influence of xenogeneic islet cell surface antibodies (ICSA) on the hormonal secretion of non-B islet cells has not been completely elucidated. Accordingly, we investigated the influence of xenogeneic antiserum on glucagon release from A cells, as representative of non-B islet cells, together with other characteristics of the antiserum. Anti-islet cell sera were produced in rabbits by xenogeneic immunization with dispersed hamster islet cells. Rabbit anti-hamster islet cell surface antibodies were detected both qualitatively by indirect immunofluorescence analyses and quantitatively by 125I-protein A radioligand assay. However, antiserum did not induce cell surface immunofluorescence on rat or mouse islet cells. As a result of evaluation of the specific cytotoxicity using 51Cr release assay, antiserum was observed to induce a significantly higher release of 51Cr compared with that of normal rabbit serum in complement-dependent antibody-mediated cytotoxicity. Both glucose-stimulated insulin and arginine-stimulated glucagon release were suppressed by xenogeneic antiserum not only in the presence but also in the absence of complement. It is concluded, therefore, that xenogeneic antiserum has a relative species specificity and non-selectively binds to islet cells in contrast with the non-species specificity and preferential binding to pancreatic B cells of human ICSA, although heterogeneity in ICSA-positive sera has been suggested.

Selective complement fixation by pancreatic B-cells after binding to islet cell surface antibodies

A two wavelength immunofluorescence system was used to demonstrate islet cell surface antibodies (ICSA) from juvenile diabetics and complement on the same viable rat islet cell. When incubated with rat islet cells at 4 C and at 37 C, ICSA specific for B-cells proved capable of binding complement. In contrast, ICSA specific for both B and non-B cells were able to fix complement on all cell types at 4 C but only on B-cells at 37 C. This study provides further evidence for the B-cell specific cytotoxic effect of ICSA in the presence of complement at physiological temperature. In addition, the ultrastructural events leading to B-cell destruction after ICSA and complement binding were studied by transmission electron microscopy (TEM).

On insulin secretion

Aspects of insulin secretory mechanisms and models of diabetogenic B cell damage are discussed. Measurements of fluxes of 3H-labelled triphenylmethylphosphonium ion, 86Rb+, 42K+, 22Na+, and 45Ca2+ in isolated islets indicate that the triggering of insulin release depends on alterations in the interaction of ions with the B cells. One difficulty in the detailed analysis of these alterations are uncertainties which arise when macroscopic concepts for homogenous phases are applied to microscopic and heterogenous compartments, as exemplified by the meaning of pH in insulin secretory granules and of membrane electric potential. Nonetheless, the importance of an apparent decreased K+ permeability in mediating the insulin-releasing action of glucose, and of an apparent increased Na+ permeability in mediating the potentiating action of acetylcholine is emphasized. Fluorescent probing of Ca2+ by chlorotetracycline revealed effects of glucose alone as well as glucose-dependent and atropine-sensitive effects of acetylcholine. Although acetylcholine, sulfonylureas, and certain thiol-blocking agents may stimulate insulin release by direct effects on the B cell plasma membrane, a high capacity for D-glucose transmembrane transport has probably evolved in order that the interior of the B cells can always sense the circulating glucose concentration. A signal to secretion is thought to be transmitted from glucose metabolism to altered ion fluxes by intervention of reduced pyridine nucleotides and hypothetical redox protein for which thioredoxin may be a model. The insulin secretory defect in hereditary diabetic C57BL/KsJ-db/db-mice is apparently linked to a decreased basal permeability for K+ and a failure of the B cells to decrease further this permeability in response to glucose. Functioning B cells are acutely damaged when exposed to heterologous serum or alloxan in vitro; cytotoxic activation of complement by the alternative pathway could perhaps occur during islet inflammation. Protection experiments with free-radical scavengers in vitro and in vivo support the theory that hydroxyl radicals are instrumental in the production of alloxan diabetes. Rapid reduction of alloxan by thioredoxin in the presence of molecular oxygen and NADPH leads to strong chemiluminescence from luminol indicative of an intense radical protection. The sensitivity of B cells to alloxan may be due to physiological specializations of their plasma membranes, involving the highly effective glucose carrier or the hypothetical oxidation/reduction systems or both.

Stimulatory effect of serum from diabetic patients on insulin release from mouse pancreatic islets maintained in tissue culture

Islets of Langerhans from NMRI-mice were kept for one week in tissue culture in medium supplemented with human serum obtained from either normal healthy subjects or newly diagnosed juvenile diabetic patients before insulin treatment. Islets cultured in diabetic serum released more insulin than islets cultured in normal serum, whether tissue culture medium 199 with 5.5-8.3 mmol/l glucose and 10% serum, or culture medium RPMI 1640 with 11 mmol/l glucose and 0.5% serum were used. Islets kept for one week in culture with diabetic serum did not show any decrease in DNA content or glucose induced insulin secretion and biosynthesis. It is concluded that serum from newly diagnosed insulin-dependent diabetic patients stimulates insulin release from isolated mouse islets kept in tissue culture. The underlying mechanism is unknown.

A new mutation (db3J) at the diabetes locus in strain 129/J mice. II. Studies of pancreatic alpha cell function in culture

Monolayer cell cultures from pancreatic islets of aging 129/J strain diabetes (db3J/db3J) and lean littermate control mice were tested for differences in glucagon and insulin secretion in either serum-free Eagle's minimal essential medium (MEM) or Dulbecco's modified minimal essential medium (DMEM). There was a highly significant (p less than 0.0001) main effect of genotype and type of culture medium on glucagon secretion with time. Thus, although numbers of A-cells were not demonstrably increased in db3J/db3J cultures in DMEM, mean medium glucagon levels increased 2.7-, 18-, and 32-fold above littermate normal culture levels at days 4, 6, and 8 respectively. In MEM, the two populations could not be discriminated on the basis of glucagon secretion. By contrast, insulin secretion over culture days showed a highly significant (p less than 0.0001) dependence on genotype, but not type of medium, with the B-cell enriched db3J/db3J preparations secreting between 20 and 30 times as much insulin as controls in both medida. Analysis revealed that the heightened secretory responsiveness of mutatn A-cells in DMEM as compared to MEM was primarily elicited by the elevated DMEM amino acid concentration and specifically lysine (0.8 mmol/l in DMEM versus 0.4 mmol/l in MEM). In pulse-chase experiments using 14 day db3J/db3J cultures, incorporation of 3H-tryptophan into protein that eluted from Biogel P-10 columns in the native glucagon peak indicates that DMEM stimulated glucagon biosynthesis as well as secretion. This study reveals an augmented sensitivity of db3J/db3J A-cells to stimulation by basic amino acids in long-term culture.

The influence of the major histocompatibility complex (H-2) on experimental diabetes in mice

Mice with different histocompatibility loci on an identical background genome (congenic resistant lines of mice) were used to study the possible influence of the histocompatibility complex on experimental diabetes. The major histocompatibility complex (H-2) was not found to influence the diabetogenic effect of encephalomyocarditis (EMC) virus. In contrast the glucose intolerance following heterologous and homologous immunization with pancreatic antigens appeared H-2 influenced. Antibodies against cell surface components on viable B-cells were present in serum from mice with glucose intolerance induced by homologous immunization. The results suggest that the susceptibility to experimental autoimmune diabetes in mice is influenced by the H-2 complex.

Islet-cell-surface antibodies in juvenile diabetes mellitus

Using an indirect immunofluorescence test on suspensions of viable, insulin-producing islet cells from rats, we found that 32 per cent (28/88) of insulin-treated patients with juvenile diabetes have islet-cell-surface antibodies in their circulation. These antibodies also occurred in four of nine children with glucose intolerance, in one of 24 healthy children and in nondiabetic children with thyroid disorders. In the diabetic children, the immunofluorescent reaction was inhibited by preadsorption of serum to islet cells but was little affected by preadsorption to rat hepatocytes or erythrocytes or to acetone powders of various rat tissues, including pancreas. These results show that organ-specific, nonspecies-specific antibodies reactive with the cell surface of the islet cells can be present in serum from diabetic children, and provide an approach to investigation of immunopathological aspects of diabetes mellitus.