D. Pipeleers

Implantation of standardized beta-cell grafts in a liver segment of IDDM patients: graft and recipient characteristics in two cases of insulin-independence under maintenance immunosuppression for prior kidney graft

Summary Islet allografts in insulin-dependent diabetic (IDDM) patients exhibit variable survival lengths and low rates of insulin-independence despite treatment with anti-T-cell antibodies and maintenance immunosuppression. Use of poorly characterized freshly isolated preparations makes it difficult to determine whether failures are caused by variations in donor tissue. This study assesses survival of standardized beta-cell allografts in C-peptide negative IDDM patients on maintenance immunosuppression following kidney transplantation and without receiving anti-T-cell antibodies or additional immunosuppression. Human islets were isolated from pancreatic segments after maximal 20 h cold-preservation. During culture, preparations were selected according to quality control tests and combined with grafts with standardized cell composition (≥ 50 % beta cells), viability ( ≥ 90 % ), total beta-cell number (1 to 2 · 106/kg body weight) and insulin-producing capacity (2 to 4 nmol · graft–1· h–1). Grafts were injected in a liver segment through the repermeabilized umbilical vein. After 2 weeks C-peptide positivity, four out of seven recipients became C-peptide negative; two of them were initially GAD65-antibody positive and exhibited a rise in titre during graft destruction. The other three patients remained C-peptide positive for more than 1 year, two of them becoming insulin-independent with near-normal fasting glycaemia and HbA1 c; they remained GAD65- and islet cell antibody negative. The three patients with surviving grafts presented a history of anti-thymocyte globulin therapy at kidney transplantation. Long-term surviving grafts increased C-peptide release following intravenous glucagon or oral glucose but not following intravenous glucose. Thus, cultured human beta-cells can survive for more than 1 year in IDDM patients on maintenance anti-rejection therapy for a prior kidney graft and without the need for an increased immunosuppression at the time of implantation. The use of functionally standardized beta-cell grafts helps to identify recipient and graft factors which influence their survival and metabolic effects. Insulin-independence can be achieved by injection of 1.5 million beta-cells per kg body weight in a liver segment. These beta-cell implants respond well to adenylcyclase activators but poorly to glucose. [Diabetologia (1998) 41: 452–459]

The biosociology of pancreatic B cells.

For almost a century, diabetes research has focused on the endocrine pancreas and its secretory products. This particular orientation started in 1889, after yon Meting and Minkowski had noted the development of diabetes in pancreatectomized dogs [1]. Four years later, Minkowski demonstrated that the antidiabetic activity of the pancreatic organ was not the result of its exocrine secretion [2]. Within the same year, the islets of Langerhans were held responsible for an internal secretion which regulated glucose homeostasis [3, 4]. It then took seven decades before the islet tissue was submitted to functional analysis. The development of techniques for the isolation of pancreatic islets [5-7] finally created opportunities to study the insulin releasing B cells in vitro. Reviews by previous Minkowski lecturers have outlined how this in vitro work has elucidated major events in the process of insulin biosynthesis and release [8-141. The recognition of a cellular heterogeneity in mammalian islets raised, however, the question to which extent the functions of the pancreatic B cells depend on a particular organisation of these cells within their own society or within other (neuro)endocrine systems (Fig. 1). In order to investigate this issue, we developed methods for the purification of islet cells and for their in vitro analysis. It was the purpose to first examine the functional characteristics of B cells which had been isolated from their natural environment, and then to reconstitute, step by step, their in situ configuration and functional activities.

IA-2-autoantibodies complement GAD65-autoantibodies in new-onset IDDM patients and help predict impending diabetes in their siblings

Summary IA-2 has been identified as an autoantigen that is recognized by immunoglobulins from insulin-dependent diabetic (IDDM) patients. Using a liquid phase radiobinding assay, we performed an IA-2-autoantibody (IA-2-Ab) assay in 474 IDDM patients and 482 non-diabetic control subjects aged 0–39 years. IA-2-Ab were detected in 58 % of the patients and 0.8 % of control subjects. Their prevalence in patients was lower than that of islet cell autoantibodies (ICA; 73 %) or glutamic acid decarboxylase (Mr 65 kDa)-autoantibodies (GAD65-Ab; 82 %) but higher than that of insulin autoantibodies (IAA; 42 %). IA-2-Ab were more frequent in patients under age 20 years (70 %) than between 20 and 40 years (45 %; p < 0.001). In the whole IDDM group, 92 % of patients were positive for at least one of the three molecular assays, which is higher than the positivity for the ICA assay (73 %). Only 1 % was negative in the molecular assays and positive in the ICA assay. IA-2-Ab levels were positively correlated with ICA titres (p < 0.001) and HLA DQ A1*0301 – DQ B1*0302 (p < 0.003) by multivariate analysis. In a group of 481 non-diabetic siblings (age 0–39 years) of IDDM patients only 7 were IA-2-Ab positive (1.5 %). All seven were under age 20 years and positive for at least two other autoantibodies and for DQ A1*0301 – DQB1*0302. Four of these seven developed IDDM during the 6–70-month follow-up period. The positive predictive value of IA-2-Ab (57 %) was higher than that of ICA, GAD65-Ab or IAA alone, or in combination (≤ 20 %) but these calculations are restricted by the relatively short observation period and the small number of cases. The only IA-2-Ab-negative case of pre-diabetes was also negative for IAA and GAD65-Ab, while it was strongly positive for ICA. In conclusion, IA-2-Ab show a high diagnostic specificity for IDDM and are predictive markers of impending diabetes in siblings of patients. In combination with other molecular antibody assays they may replace ICA testing in future. Our data also indicate that other autoantibodies than IA-2-Ab, GAD65-Ab and IAA contribute to ICA. [Diabetologia (1997) 40: 95–99]

Physiologic relevance of heterogeneity in the pancreatic beta-cell population

SummaryIn vitro studies on purified rat beta cells have indicated a functional diversity among insulincontaining cells. Intercellular differences were found in the rates of glucose-induced insulin synthesis and release. They are attributed to differences in cellular thresholds for glucose utilization and oxidation, as can be caused by varying activities in rate limiting steps such as glucokinase-dependent phosphorylation. The percent of functionally active beta cells increases dose-dependently with the glucose concentration, making cellular heterogeneity and its regulation by glucose major determinants for the dose-response curves of the total beta-cell population. Beta cells which are already responsive to low glucose concentrations are characterized by a higher content in pale immature granules; their activated biosynthetic and secretory activity accounts for preferential release of newly-formed hormone by the total beta-cell population. At any glucose level, the amplitude of insulin release depends on the percent glucose-activated cells and their cyclic AMP content, an integrator of (neuro)hormonal influences. The in vitro described heterogeneity in beta-cell functions may bear physiological relevance as several of its characteristics are also detectable in intact pancreatic tissue; furthermore, in vitro signs of heterogeneity can be altered by prior in vivo treatment indicating that they express properties of the cells in their in situ configuration. Elevated basal levels of (pro)insulin may reflect the existence of an increased number of beta cells that are activated at low physiologic glucose concentrations. Reductions in stimulated insulin levels can be caused by decreased numbers of beta cells that are activated at the prevailing glucose concentration or by insufficient cyclic AMP levels in beta cells, possibly as a result of inadequate signalling from hormones of local or distal origin. Only few markers are currently available with which to explore these mechanisms in vivo. Additional markers and tests should help assess the possible role of variations in beta-cell heterogeneity in the pathogenesis of diabetes mellitus.

Effect of nutrients, hormones and serum on survival of rat islet beta cells in culture.

SummaryThis study quantifies the survival of purified single rat beta cells under different culture conditions. Less than 10% of the cells survive 9 days of culture in Hams F10 medium without supplements. Addition of fetal calf serum (5%) increases cell survival to 54% in the absence and to 78% in the presence of isobutylmethylxanthine (50 μmol/l). The effect of serum is explained, at least partly, by the presence of albumin and of low molecular weight constituents. In serum-free Hams F10 with 50 μmol/l isobutylmethylxanthine, 75% of cells survive after the addition of bovine serum albumin (1%) and of ultroser (0.2%), a commercial serum substitute. Survival of at least 75% of cells is also maintained in Hams F10 with isobutylmethylxanthine plus albumin, and supplemented by metabolizable nutrients or by the peptides glucagon (10−8 mol/l) or growth hormone (1 μg/ml) plus insulin like growth factor-I (50 ng/ml). d-Glucose increases beta-cell survival in a dosedependent manner up to 10 mmol/l; a beneficial effect is also observed with other metabolizable compounds (leucine and glutamine) but not with non-metabolizable monosaccharides. Glucose-induced survival of islet beta cells can be attributed to its dose-dependent recruitment of cells into metabolic activities; however, a 9-day exposure to excessively high nutrient concentrations (> 20 mmol/l glucose) is deleterious to the cells. These results define culture media, with or without serum, wherein at least 75% of single rat islet beta cells can survive for a minimum of 9 days. This will allow for studies on beta-cell toxic conditions and potentially protective agents. The data also serve as basis for developing media with better survival of beta cells in cultured aggregates.

A method for the purification of single A, B and D cells and for the isolation of coupled cells from isolated rat islets.

SummaryA method has been developed for the purification of single A, B and D cells and for the isolation of coupled islet cells. Isolated rat islets were dissociated by repeated pipetting in the presence of trypsin and ethylene glycol bis (β-aminoethyl ether) N, N′ tetra-acetic acid (EGTA), and then filtered through nylon and Percoll; the cell preparation consisted of 70% single and 30% coupled cells. Sizing of the cells led to the recognition of a small-islet cell population (35%; cell volume 200–600 μm3) composed of A and D cells, and a large-islet cell population (65%; cell volume 600–1500 μm3) identified as B cells. Differences in sedimentation velocity formed the basis for the islet cell separation by counterflow elutriation. Single islet cells eluted prior to coupled cells and were distributed over A and D cell-enriched fractions I and II (65% A, 25% B, 10% D) and the B cell-enriched fraction III (93% B). The slightly different densities of A (d=1.068), B (d=1.065) and D (d=1.070) cells allowed a subsequent purification by density gradient centrifugation resulting in a final 10- to 30-fold enrichment in either A, B or D cells. Most coupled islet cells were recovered in fraction IV, occurred mainly as doublets and were composed of 90% B cells and 7% D cells; the multiple pseudopods, which characterize isolated D cells, might contribute to the coupling tendency of the D cells. It is concluded that the purified A, B and D cell fractions and the coupled islet cell preparations offer a direct approach to the study of individual islet cell types and their intercellular communication.

GLUCOSE INDUCES OSCILLATORY CA2+ SIGNALLING AND INSULIN RELEASE IN HUMAN PANCREATIC BETA CELLS

SummaryMechanisms of pulsatile insulin release in man were explored by studying the induction of oscillatory Ca2+ signals in individual beta cells and islets isolated from the human pancreas. Evidence was provided for a glucose-induced closure of ATP-regulated K+ channels, resulting in voltage-dependent entry of Ca2+. The observation of step-wise increases of capacitance in response to depolarizing pulses suggests that an enhanced influx of Ca2+ is an effective means of stimulating the secretory activity of the isolated human beta cell. Activation of muscarinic receptors (1–10 μmol/l carbachol) and of purinergic P2 receptors (0.01–1 μmol/l ATP) resulted in repetitive transients followed by sustained elevation of the cytoplasmic Ca2+ concentration ([Ca2+]i). Periodic mobilisation of intracellular calcium was seen also when injecting 100 μmol/l GTP-γ-S into beta cells hyperpolarized to −70 mV. Individual beta cells responded to glucose and tolbutamide with increases of [Ca2+]i, manifested either as large amplitude oscillations (frequency 0.1–0.5/min) or as a sustained elevation. Glucose regulation was based on sudden transitions between the basal and the two alternative states of raised [Ca2+]i at threshold concentrations of the sugar characteristic for the individual beta cells. The oscillatory characteristics of coupled cells were determined collectively rather than by particular pacemaker cells. In intact pancreatic islets the glucose induction of well-synchronized [Ca2+]i oscillations had its counterpart in 2–5 min pulses of insulin. Each of these pulses could be resolved into regularly occurring short insulin transients. It is concluded that glucose stimulation of insulin release in man is determined by the number of beta cells entering into a state with Ca2+-induced secretory pulses.

Islet Cell Surface Antibodies from Insulin-dependent Diabetics Bind Specifically to Pancreatic B Cells

Viable rat islet cells were used to detect islet cell surface antibodies (ICSA) in the sera of diabetic and control patients. ICSA were present in almost all recent-onset insulin-dependent diabetics younger than 30 yr (15/16); their incidence in other diabetics (6/22) was also higher than in normal controls (1/18) or in patients with autoimmune thyroiditis (1/12). The varying specificity of the ICSA for the different islet cell types led to the recognition of class I sera, whose ICSA bind exclusively to B cells, class II sera, binding only to A and pancreatic polypeptide (PP) cells and class III sera, reacting with the three islet cell types but not with D cells. Most recent-onset insulin-dependent diabetics younger than 30 contained class I-ICSA, which is consistent with an autoimmune basis of their disease and with an involvement of surface antibodies in the B cell destruction. The presence of class II ICSA in three older diabetics and in one normal control raises the question whether autoimmune reactions against A and PP cells exist and are associated with a distinct entity in islet disease. It is concluded that the autoimmune form of diabetes mellitus represents a heterogeneous group, in which ICSA-positive patients can be distinguished on the basis of their ICSA-binding to one or more islet cell types. Three techniques can be used for the further identification of circulating ICSA, namely binding experiments with purified A or B cells, electron microscopical analysis of ICSA-binding islet cells purified by fluorescence-activated cell sorting, and the immunocytochemical characterization of ICSA-positive cells.

Sensitivity of rat pancreatic A and B cells to somatostatin

SummaryIslet A and B cells were purified from the rat pancreas and examined for their respective sensitivity to somatostatin. Both somatostatin-14 (S14) and -28 (S28) inhibited glucagon and insulin release through direct interactions with the corresponding cell types. A dose-dependent suppression of the secretory activities was paralleled by a reduction in cellular cyclic AMP formation with similar ED50 values for both actions. The somatostatin effects on pancreatic hormone release may thus be mediated via an inhibition of adenylate cyclase activity. In pancreatic A cells, S14 and S28 were equally potent inhibitors with ED50 values ranging from 2 × 10−12 to 2 × 10−11 mol/l. Pancreatic B cells exhibited a similar sensitivity to S28 as the A cells (ED50 of 2 to 5 × 10−11 mol/l), but not to S14 (ED50 of 2 × 10−9 mol/l). Extrapolation of these in vitro sensitivities of islet A and B cells to the in vivo situation suggests that both cell types can respond to circulating S28 levels and that A cells are sensitive to both locally and distally released S14. Islet B cells appear insensitive to the normal peripheral S14 levels but could respond to locally released somatostatin. The marked difference in the sensitivities of islet A and B cells to S14 suggests that these cell types are equipped with different somatostatin receptors. This notion was further supported by the cell-selective actions of the synthetic S14 analogues [D-Trp8, D-Cys14]S14 and desAsn5[D-Trp8, D-Ser13]Sl4.

Nicotinamide protects human beta cells against chemically-induced necrosis, but not against cytokine-induced apoptosis

Summary Nicotinamide intervention trials are presently undertaken to prevent Type I (insulin-dependent) diabetes in high risk subjects. They are based on studies in rodents reporting nicotinamide protection against beta-cell injury in vitro and in vivo. This study examines whether nicotinamide can protect human beta cells in vitro. At concentrations (2 and 5 mmol/l) to protect rat beta cells against necrosis by streptozotocin or hydrogen peroxide, nicotinamide prevents hydrogen peroxide-induced necrosis of human beta cells. As with rat beta cells, nicotinamide fails to protect human beta cells against apoptosis induced by a combination of the cytokines interleukin-1β, interferon-γ and tumour necrosis factor-α. In rat beta cells, nicotinamide (2 to 20 mmol/l) was also found to induce apoptosis, in particular during the days following its protection against necrosis; this cytotoxic effect was not observed with human beta cells. These data demonstrate that nicotinamide can protect human beta cells against radical-induced necrosis, but not against cytokine-induced apoptosis. This effect is not associated with a delayed apoptosis as in rat beta cells. [Diabetologia (1999) 42: 55–59]