Kazushige Ejiri

Production of monoclonal antibodies to islet cell surface antigens using hybridization of spleen lymphocytes from non-obese diabetic mice.

SummaryNon-obese diabetic mice display a syndrome with dramatic clinical and pathological features similar to those of Type 1 (insulin-dependent) diabetes in man. Circulating autoantibodies to the surface of islet cells were demonstrated in some of these mice by a protein A radioligand assay. To produce monoclonal antibodies to islet cell surface antigens, therefore, we took the spleens of non-obese diabetic mice, transferred the spleen cells into non-immunized recipient mice, which were made immunologically incompetent by a large dose of X-irradiation, and then fused their lymphocytes with FO mouse myeloma cells. After screening the resultant hybrids, one stable hybridoma (3A4) that produced a monoclonal antibody (IgG1) specifically bound to the surface of islet cells was obtained. The purified monoclonal antibody was bound to the surface of transplantable Syrian golden hamster insulinoma cells sevenfold more than control antibody. Adsorption of the antibody on mouse spleen lymphocytes or thymocytes resulted in only a slight decrease in 125I-protein A binding to insulinoma cells. This antibody also reacted with the surface of mouse and rat islet cells, but not with that of rat spleen cells or hepatocytes. A spectrophotometric assay for peroxidase activity demonstrated that six times more peroxidase bound to insulinoma cells incubated with the antibody than to cells treated with control antibody. Furthermore, this antibody could be visually detected in the immunoenzymatic labelling of the surface of insulinoma cells. In summary, we have developed a novel method of producing monoclonal antibodies to the surface of islet cells for probing into the pathogenesis of Type 1 diabetes.

Possible involvement of cholinergic nicotinic receptor in insulin release from isolated rat islets

Insulin release is influenced by the autonomic nervous system. Regarding parasympathetic control, previous reports have shown that regulation of insulin release is executed exclusively through muscarinic receptors in the pancreatic islets. In the present study, however, we examined the effect on insulin release at the islet level of various agents affecting the parasympathetic nervous system, especially nicotinic receptor blockers. Pancreatic islets isolated from adult Wistar male rats were incubated with these agents and insulin release in the media was measured. Acetylcholine chloride (10(-5) M), as well as distigmine bromide (10(-6), 10(-5) M), both of which are cholinesterase inhibitors, stimulated insulin release, whereas atropine (5 x 10(-6), 5 x 10(-5) M) suppressed it. On the other hand, serum and IgG from myasthenia gravis patients, containing anti-acetylcholine receptor antibodies, affected insulin release, and alpha-bungarotoxin (10(-9)-10(-7) M), a nicotinic receptor blocker, stimulated insulin release dose-dependently. The present observations suggest that insulin release is influenced by the parasympathetic nervous system, mediated via not only muscarinic but also nicotinic receptors.

Participation of nicotinic receptor in hormone release from isolated rat islets of Langerhans

Pancreatic hormone release is generally thought to be regulated through adrenergic as well as muscarinic receptors. We have previously observed possible nicotinic involvement in insulin release. In the present study, we incubated isolated rat islets for 60 min with various concentrations of atropine (a muscarinic receptor blocker), alpha-bungarotoxin (alpha-Btx, a nicotinic receptor blocker), and anti-acetylcholine receptor antibody (IgG) (anti-Ach.R.Ab) obtained from a patient with myasthenia gravis. Atropine suppressed insulin release, and alpha-Btx and anti-Ach.R.Ab potentiated it; atropine did not suppress glucagon release, while alpha-Btx and anti-Ach.R.Ab raised it. None of these agents influenced somatostatin release. These observations suggest that muscarinic as well as nicotinic receptors influence insulin release, as nicotinic receptors do glucagon release. Neither nicotinic nor muscarinic receptors seem to regulate somatostatin release.

Complements in diabetes mellitus: activation of complement system evidenced by C3d elevation in IDDM

To characterize insulin-dependent diabetes mellitus (IDDM) and non-insulin-dependent diabetes mellitus (NIDDM) in terms of the complement system, some components of the system as well as the related substances and indices were studied. CH50, C3, C4 and C3bINA significantly increased in both IDDM and NIDDM compared with non-diabetic healthy controls. ACH50 was also elevated in NIDDM, whereas it was similar in IDDM and controls. Besides, the serum concentration of C3d, a breakdown product of C3, was higher in IDDM than in NIDDM and healthy controls, but that in NIDDM did not differ significantly from the control. B1Hg1 was not different among IDDM, NIDDM and non-diabetic controls. These observations suggested that there is a high level of complements in both types of diabetes mellitus, but the complement activation seems to be much enhanced in IDDM compared with NIDDM.

Sophisticated mesh filtration technique of a large-scale isolation of islets and their function

A large-scale isolation of islets is required for islet transplantation. We improved our conventional method, and could obtain about three times more islets than by the conventional methods. Pancreata of adult Wistar rats were inflated by injection of buffer with (A) or without 1.3 mg/ml collagenase (B). The rats were bled from the inferior vena cava and the aorta in (A) simultaneously with the inflation. They were further digested with collagenase and filtered through two different meshes (pore size: 1190 and 590 microns) (A1) or three different meshes (pore size: 1190, 590 120 microns) (A2) in order. Insulin released from islets isolated in this manner was determined by 1-h incubation with 3.3 and 16.7 mM glucose. Besides, 600 islets each were transplanted into the liver of streptozotocin-induced diabetic Wistar rats and their fasting plasma glucose was measured at weekly intervals. (1) With these methods more numerous islets were harvested by A1 (mean: 554) and A2 (mean: 746) than B (mean: 224). (2) Insulin released at both glucose concentrations was similar among islets obtained by A1, A2 and B. (3) The plasma glucose-lowering effect was similar among the islets obtained by these methods. (4) A more selected range of islet sizes was obtained by A2 than A1. These observations indicate that the present techniques (A1 and A2) are less time-consuming and simpler for a large-scale isolation of islets.

Effects of tris(hydroxymethyl)aminomethane on biosynthesis and release of insulin in the pancreatic Langerhans islets

Tris(hydroxymethyl)aminomethane (Tris) has been shown to inhibit selectively the Golgi apparatus and Golgi-endoplasmic reticulum-lysosomal system (GERL system) of several kinds of cells including pancreatic B cells. This study was designed to assess the effect of Tris on insulin, glucagon and somatostatin release and insulin synthesis in pancreatic B cells by using isolated rat pancreatic islets. Tris suppressed glucose-induced insulin release, whereas it did not affect the glucagon and somatostatin release. Furthermore, the incorporation of [3H]leucine into the insulin fraction was suppressed by 10 mM Tris, but the sum of the radioactivity of both proinsulin and insulin fraction were not influenced. The present study suggests that the Golgi apparatus and GERL system may play a role in insulin secretion and biosynthesis in pancreatic B cells.

Vacor inhibits insulin release from islets in vitro

It has been reported that Vacor, a rodenticide containing N-3-pyridylmethyl-N’-p-nitrophenyl urea, causes insulin-dependent diabetes mellitus. The pathomechanism of Vacor-induced diabetes mellitus has not been clarified yet. The effect of Vacor, therefore, was studied in terms of insulin release from isolated rat pancreatic islets. Vacor suppressed glucose-stimulated insulin release, but did not affect the insulin release induced by theophylline or 12-o-tetra-decanoylphorbol 13-acetate. It is suspected that the suppression of insulin release from pancreatic islets by Vacor may contribute to the pathogenesis of Vacor-induced diabetes mellitus and that this suppression might not be related to cAMP and C-kinase.

Complements in non-insulin-dependent diabetes mellitus with complications

A relation of the complement system to the development of complications in non-insulin-dependent diabetes mellitus (NIDDM) was evaluated by measuring some components of the complement system. CH50, C3, C4 and C3bINA were significantly elevated in subjects with NIDDM as compared with healthy non-diabetic controls. However, CH50 and C3 did not differ between diabetics with and without complications. C4 was higher in diabetics with retinopathy as well as with retinopathy and neuropathy than in diabetics without these complications. ACH50, beta 1Hg1 and C3d were similar in subjects with NIDDM and non-diabetics, and not associated with complications of NIDDM. C3d/C3 in NIDDM without complications was lower than in healthy subjects, but did not significantly differ between the types of complications. These results suggest that the high level of complements in NIDDM might be due to enhanced production of complements and the development of diabetic complications would be related to the elevated level of complements.

Effect of cooling rate on insulin release from frozen-thawed dispersed rat islet cells

Rat islet cells, dissociated with EDTA-Dispase, were immersed in 10% dimethyl sulfoxide at 20 degrees C for 15 min and frozen to -40 degrees C at a cooling rate of 0.5 or 1.0 degree C/min and subsequently further to -80 degrees C at 3 degrees C/min by a programmable freezer. After being maintained at -80 degrees C for 10 min, they were rapidly thawed in a water bath at 37 degrees C. They were cultured for 12 h and preincubated in 3.3 mM glucose-containing Krebs-Henseleit bicarbonate buffer (KHBB) for 1 h. Groups of 10(4) cells were then incubated in 3.3 or 16.7 mM glucose-containing KHBB for another hour. As a control, non-frozen-thawed cultured islet cells were incubated similarly. The non-frozen rat islet cells released 1.29 pg insulin/cell.60 min in the presence of 3.3 mM glucose and this release level was significantly elevated to 1.64 pg insulin/cell.60 min in the presence of 16.7 mM glucose. The cells frozen at 0.5 degree C/min releasing 1.55 pg insulin/cell.60 min in the presence of 3.3 mM glucose also responded to 16.7 mM glucose and released the significantly high level of 1.87 pg insulin/cell.60 min. However, the islet cells frozen at a cooling rate of 1 degree C/min secreted 1.74 and 1.92 pg insulin/cell.60 min in the presence of 3.3 mM and 16.7 mM glucose respectively. There was no significant difference between these levels. These results indicate that cryopreservation at a cooling rate of 0.5 degree C/min may be adequate for the preservation of dispersed pancreatic endocrine cells.

Production of anti-insulin monoclonal antibody and its application to immunoassay of insulin and immunohistochemistry

An unlimited supply of suitable antisera is wanted for immunoassays, analysis of antigenic determinants and precise localization of antigens in biological systems. Therefore, we produced a monoclonal antiporcine insulin antibody by the hybridoma technology and assessed it in comparison with polyclonal antibody. The spleen cells of BALB/c mice immunized against porcine insulin were hybridized with mouse myeloma cells (P3-X63-Ag8-U1). The monoclonal antibody thus generated was shown to have high binding capacity and specificity to porcine insulin in radioimmunoassay. It reacted with human insulin as well, but did not crossreact with other polypeptide hormones produced in the pancreatic islets such as glucagon, somatostatin and pancreatic polypeptide. In immunohistochemistry human and dog islets were stained by this monoclonal antibody. Rat islets were not stained, although they reacted with polyclonal anti-insulin antibody. The insulin of human serum samples measured using the monoclonal antibody was tightly correlated with that using the polyclonal antibody. These observations indicate that our hybridoma-derived monoclonal antibody is useful for immunoassay as well as localization of insulin.