HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Maechler, P. Articles by Wollheim, C. B. Search for Related Content PubMed PubMed Citation Articles by Maechler, P. Articles by Wollheim, C. B. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 39, 27905-27913, September 24, 1999

Hydrogen Peroxide Alters Mitochondrial Activation and Insulin Secretion in Pancreatic Beta Cells^*

Pierre Maechler, Lan Jornot^§, and Claes B. Wollheim^¶

From the  Division of Clinical Biochemistry and the ^§ Respiratory Division, Department of Internal Medicine, University Medical Center, CH-1211 Geneva, Switzerland

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The effects of a transient exposure to hydrogen peroxide (10 min at 200 µM H₂O₂) on pancreatic beta cell signal transduction and insulin secretion have been evaluated. In rat islets, insulin secretion evoked by glucose (16.7 mM) or by the mitochondrial substrate methyl succinate (5 mM) was markedly blunted following exposure to H₂O₂. In contrast, the secretory response induced by plasma membrane depolarization (20 mM KCl) was not significantly affected. Similar results were obtained in insulinoma INS-1 cells using glucose (12.8 mM) as secretagogue. After H₂O₂ treatment, glucose no longer depolarized the membrane potential () of INS-1 cells or increased cytosolic Ca²⁺. Both and Ca²⁺ responses were still observed with 30 mM KCl despite an elevated baseline of cytosolic Ca²⁺ appearing ~10 min after exposure to H₂O₂. The mitochondrial of INS-1 cells was depolarized by H₂O₂ abolishing the hyperpolarizing action of glucose. These changes correlated with altered mitochondrial morphology; the latter was not preserved by the overexpression of the antiapoptotic protein Bcl-2. Mitochondrial Ca²⁺ was increased following exposure to H₂O₂ up to the micromolar range. No further augmentation occurred after glucose addition, which normally raises this parameter. Nevertheless, KCl was still efficient in enhancing mitochondrial Ca²⁺. Cytosolic ATP was markedly reduced by H₂O₂ treatment, probably explaining the decreased endoplasmic reticulum Ca²⁺. Taken together, these data point to the mitochondria as primary targets for H₂O₂ damage, which will eventually interrupt the transduction of signals normally coupling glucose metabolism to insulin secretion.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The control of insulin secretion in the pancreatic beta cell depends on the precise tuning of glucose metabolism leading to signal transduction (1, 2). Indeed, impaired metabolism secretion coupling results in inappropriate insulin release potentially causing defective blood glucose homeostasis. Dysfunction of the beta cell signal transduction may be of various origin, among which oxidative stress has been proposed to play a critical role.

Type I diabetes, or insulin dependent diabetes mellitus, is an autoimmune disease characterized by altered function and beta cell death subsequent to exposure to inflammation products (3). During insulitis macrophages infiltrate the islets of Langerhans and generate reactive oxygen species such as hydrogen peroxide (H₂O₂), which exert deleterious actions on the beta cells and on mitochondrial oxidative metabolism. It is noteworthy that activated phagocytes can produce as much as 47 nmol of H₂O₂/10⁶ cells within 30 min corresponding to a concentration of 47 µM H₂O₂ in a diluted volume of 1 ml (4). Nitric oxide (NO), another free radical precursor produced by macrophages, suppresses mitochondrial activity leading to a defective insulin release in response to nutrient secretagogues (5). Moreover, it has been shown that NO damages islet cell DNA (6) and mitochondrial DNA in beta cells (7). In general, mitochondrial DNA is more sensitive to oxidative stress than nuclear DNA (8, 9). The mitochondria play a key role in the control of nutrient-induced insulin exocytosis by generating 1) ATP to raise cytosolic Ca²⁺ concentration ([Ca²⁺]_c)¹ through membrane depolarization (1, 2) and 2) additional mitochondrial factor(s) triggering insulin exocytosis (10, 11).

A defective secretory response to nutrients can be encountered in aged patients (12). In normal rats, insulin secretion is maintained throughout life (13), whereas in perinatal malnourished rats aging results in hypoinsulinemia and hyperglycemia (14). Aging has been shown to be associated with the alteration of beta cell function independent of that seen in noninsulin-dependent diabetes mellitus (15). A reduced translation of the mitochondrial genome (16) and of the cytochrome c oxidase activity (17) has been reported in elderly subjects, pointing to an age-related mitochondrial dysfunction in this highly oxidative organelle. Mitochondrial aconitase, a tricarboxylic acid cycle enzyme, is susceptible to oxidative modification during aging in vivo (18). Moreover, it has recently been demonstrated that the mitochondrial adenine-nucleotide translocase is modified oxidatively during aging together with loss of functional activity (19). In cells the mitochondrion is the main source of oxidants. Indeed, imperfect electron transport generates superoxide anions, which are spontaneously dismutated to H₂O₂ (9, 20). Thus, the mitochondria are pivotal in the control of insulin secretion, whereas at the same time generating reactive oxygen species in the cell mostly in the form of H₂O₂.

Of particular importance is the high sensitivity of pancreatic beta cells to oxidative stress. Moreover, the diabetic state is associated with increased oxidative stress and free radical damage (20). In fact, the expression of the H₂O₂-inactivating enzymes catalase and glutathione peroxidase in rat pancreatic islets is twenty times lower than in the liver (21). As a consequence the basal catalase activity in islets is very low (1.3 units/mg protein in the rat) and was reported to result in high susceptibility to cytotoxicity during a 16-h exposure to H₂O₂ (10-500 µM) (22). The rat beta cell line INS-1, which was used in the present study, exhibits similar low baseline catalase levels (1.0 units/mg protein) as primary rat islets (22). Taken together, these bindings call for a better understanding of the cellular mechanisms linking oxidative stress to impaired insulin secretion.

In mouse pancreatic beta cells H₂O₂ hyperpolarizes the cell membrane coupled with an increase of cell membrane conductance (23). Moreover, it has recently been shown that H₂O₂ increases intracellular Ca²⁺, decreases the ATP/ADP ratio, and inhibits glucose-stimulated insulin secretion from isolated mouse islets (24). The present work was designed to dissect in more detail oxidative stress-induced cell damage and to parallel mitochondrial parameters with the secretory response in insulin-secreting cells stimulated with glucose after exposure to H₂O₂. We used H₂O₂ as a well established model of a biologically active oxygen-derived intermediate (20) at the concentration of 200 µM. Rat islets and the beta cell line INS-1 have been shown to be sensitive to this H₂O₂ concentration without exhibiting major toxicity (22). Our results show that the defective glucose-induced insulin secretion observed after H₂O₂ treatment correlates with altered mitochondrial activation seen as a loss of mitochondrial membrane potential (_m), decreased ATP generation (measured on-line in living cells), and impaired responses of mitochondrial Ca²⁺ concentration ([Ca²⁺]_m). In addition, the mitochondrial morphology, which depends mainly on _m, was examined. In PC12 cells, the antiapoptotic protein Bcl-2 (for a review see Ref. 25) was reported to prevent the H₂O₂-induced _m loss and the subsequent apoptosis (26). Here we also examined whether overexpression of Bcl-2 could prevent the alteration of mitochondrial morphology during H₂O₂ treatment.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Coelenterazine, Mitotracker, rhodamine-123, and bisoxonol were obtained from Molecular Probes (Eugene, OR); luciferin was from Promega (Madison, WI); bovine serum albumin, doxycycline, methyl succinate, firefly lantern extract, and FCCP were from Sigma; anti-insulin antibody was from Linco (St. Charles, MO); Bcl-2 mouse monoclonal antibody was from Santa Cruz Biotechnology (Santa Cruz, CA); G418 and hygromycin were from Calbiochem.

Cell Culture-- INS-1 cells were cultured in RPMI 1640 medium as described previously (27-29). Stable clones of INS-1 cells expressing the Ca²⁺-sensitive photoprotein aequorin in the cytosol (INS-1/C-29) (28) or targeted either to the mitochondria (INS-1/EK3) (28) or to the endoplasmic reticulum (ER) (INS-1/ER#18) (30) were cultured in the presence of 250 µg/ml G418 for continuous selection of cells expressing the plasmid with the associated neomycin resistance. Clonal INS-1 lines expressing cytosolic luciferase under the control of doxycycline-dependent transcriptional transactivator (INS-r3-LUC7) were used for cytosolic ATP monitoring in living cells (29). Pancreatic islet cells were isolated by collagenase digestion from male Wistar rats weighing ~200 g (31) and were cultured free floating in RPMI 1640 medium (11.1 mM glucose) for 2-4 days.

Insulin Secretion-- For perifusion protocols, islets or trypsinized cells were kept in spinner culture for 2 h in glucose-free RPMI 1640 supplemented with 25 mM HEPES and 1% new born calf serum prior to perifusion in modified Krebs-Ringer bicarbonate HEPES buffer (KRBH) composed of 135 mM NaCl, 3.6 mM KCl, 10 mM HEPES, pH 7.4, 5 mM NaHCO₃, 0.5 mM NaH₂PO₄, 0.5 mM MgCl₂, 1.5 mM CaCl₂, and 2.8 mM glucose. Cells were placed in a thermostatted chamber (35 rat islets or 10⁶ clonal cells/chamber) and perifused at a flow rate of 1 ml/min. For static incubations, INS-1 cells (2 × 10⁵ cells/well in polyornithine-treated 24-well plates) were seeded and cultured for 3-5 days in complete RPMI 1640 medium. Prior to the experiments, cells were maintained for 2 h in the glucose-free culture medium (unless otherwise stated), washed, and preincubated in glucose-free KRBH for 30 min. After further washing, cells were incubated with or without 200 µM H₂O₂ for 10 min, washed again in the presence of 100 units/ml catalase, and finally stimulated for 30 min. Bovine serum albumin (0.1%) was added to buffers as carrier, and insulin secretion and cellular insulin content extracted with acid ethanol were determined by radioimmunoassay using rat insulin as a standard (27).

Measurements of Luminescence-- Luciferase or aequorin-expressing cells were seeded on 13-mm diameter coverslips 3-5 days prior to analysis and maintained in the same medium as above except for the addition of G418 and hygromycin. Prior to luminescence measurements, cells were maintained in glucose- and glutamine-free RPMI 1640 containing 10 mM HEPES for 2-5 h at 37 °C. This period also served to load aequorin-expressing cells with 2.5 µM coelenterazine, the prosthetic group of aequorin (28). For ER#18 cells, the latter procedure was undertaken in Ca²⁺-free buffers to conserve ER aequorin before the measurements as detailed elsewhere (30). Luminescence was measured by placing the coverslip in a 0.5-ml thermostatted chamber at 37 °C approximately 5 mm from the photon detector. We used a home-built photomultiplier apparatus (EMI 9789, Thorn-EMI, United Kingdom), and data were collected each second on a computer photon counting board (EMI C660) prior to calibration for [Ca²⁺]_c and [Ca²⁺]_m as described previously (28). Suspensions of islet cells were perifused with the same buffers as INS-1 cells using a perifusion apparatus (28). Intact cells were perifused with KRBH, and beetle luciferin (10 µM) was added to the buffer for luciferase-expressing cells (29). Luciferase luminescence was used for the monitoring of ATP in living cells as described previously (29).

ATP Measurements in Cell Extracts-- According to the protocol of Stanley and Williams (32), cells were stimulated for 10 min at 37 °C and scraped into 1 ml of 0.4 N HClO₄ to terminate the reaction. Following neutralization with 2 N K₂CO₃, cell extracts were incubated with a luciferin-luciferase mixture in arsenate buffer (0.1 M Na₂HAsO₄·7H₂O, pH 7.4, 20 mM MgSO₄·7H₂O), and the resultant luminescence was measured.

Membrane Potential-- After a 3-5 day culture period, cells were trypsinized (0.025% trypsin, 0.27 mM EDTA), and the cell suspension was maintained for 2 h in a spinner culture with glucose-free RPMI 1640 plus 1% newborn calf serum at 37 °C. For cell membrane potential (_c) measurements, 2 × 10⁶ cells were pelleted, resuspended in 2 ml of KRBH with 100 nM bisoxonol, and transferred to a thermostatted cuvette. Cells were then excited at 540 nm, and emission was recorded at 580 nm. _m was measured as described (10, 33). Briefly, after the spinner culture period, cells were loaded with 10 µg/ml rhodamine-123 for 10 min at 37 °C. After centrifugation, the cells were resuspended and transferred to the fluorimeter cuvette, and fluorescence, excited at 490 nm, was measured at 530 nm. All membrane potential measurements were performed at 37 °C with gentle stirring in an LS-50B fluorimeter (Perkin-Elmer).

Transient Transfection with Human Bcl-2-- Cells were seeded on A-431 extracellular matrix-coated glass coverslips at 2 × 10⁵ cells/2 ml of RPMI 1640 medium in 35-mm dishes. Two to three days later the cells were washed with phosphate-buffered saline. For each well, 20 µl of the polycationic lipid LipofectAMINE (Life Technologies, Inc.) was diluted in 200 µl of RPMI 1640 medium (10 mM HEPES), and 4 µg of plasmid with human Bcl-2 (kindly donated by Dr. Jean-Claude Martinou, Ares Serono, Geneva, Switzerland) was diluted in another 200 µl of the same medium. The liposome and plasmid solutions were then mixed and the volume was adjusted to 500 µl with RPMI 1640-HEPES medium. After a 30-min incubation at room temperature, this transfection mixture was added to the well prior to a 5-h incubation at 37 °C in air/5% CO₂. The cells were then washed twice with RPMI 1640-fetal calf serum (10%) medium and further cultured in the same medium for 48 h before the experiment. This transfection procedure resulted in ~25% of transfected cells as judged by immunofluorescence, using an anti-human Bcl-2 antibody that did not react with the endogenous rat Bcl-2.

Mitochondrial Staining and Immunofluorescence-- Rat islet cells and INS-1 cells were cultured for 2-3 days on A-431 matrix-coated glass coverslips prior to the experiments (11). The cells were washed twice with KRBH, loaded with 100 nM Mitotracker for 25 min at 37 °C, and washed again with KRBH. Then the cells were exposed to KRBH only, with 200 µM H₂O₂ or 1 µM FCCP for 10 min. Treatment was stopped by adding catalase (100 units/ml) and giving fresh KRBH for another 20-min period. The cells were then washed, fixed in 4% paraformaldehyde, and permeabilized with 0.1% Triton X-100 in phosphate-buffered saline, 1% bovine serum albumin. The preparation was then blocked with phosphate-buffered saline bovine serum albumin before staining using mouse monoclonal anti-human Bcl-2 immunoglobulin followed by fluorescein-labeled goat anti-mouse IgG as second antibody. Cells were viewed using a Zeiss laserscan confocal 410 microscope.

Statistical Analysis-- Where applicable, values are expressed as mean ± S.E., and significance of difference was calculated by a Student's t test for unpaired data. Traces without S.E. values are representative of at least three independent experiments.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Effect of H₂O₂ on Insulin Secretion in Rat Islets and INS-1 Cells-- Rat pancreatic islets were maintained in culture (11.1 mM glucose) for 2-4 days prior to the experiments. Insulin secretion was stimulated with 16.7 mM glucose for 10 min after basal perifusion at 2.8 mM glucose (Fig. 1A). When 200 µM H₂O₂ was added for 10 min a transient release of insulin was observed. The cells were then washed for 5 min in the presence of catalase to block any remaining extracellular H₂O₂. The subsequent stimulation with 16.7 mM glucose was largely blunted (Fig. 1B), because only 30% of the secretory response was preserved in terms of area under the curve (see Table I). Control representative traces of islets perifused without glucose stimulation are shown in Fig. 1, A and B (thin lines). The tricarboxylic acid cycle intermediate succinate, rendered cell permeant by the ester binding of a methyl group (34), evoked an insulin secretory response (Fig. 1C), which was significantly inhibited by 74% after H₂O₂ treatment (Fig. 1D and Table I). A nonnutrient stimulation of insulin release was induced by KCl, which triggers insulin exocytosis by a simple rise of [Ca²⁺]_c consequent to membrane depolarization (Fig. 1E). In this case a similar secretory response to KCl was observed after H₂O₂ exposure compared with the control stimulation (Fig. 1F). In static incubations at basal 2.8 mM glucose, the insulinoma cells INS-1 released insulin during a 10-min exposure to H₂O₂ in a dose-dependent manner (Fig. 2A). The threshold for the action of H₂O₂ on insulin secretion was 200 µM (+52%, p < 0.02), a concentration used in all following experiments. This 10-min treatment with 200 µM H₂O₂ did not result in apparent cell morphology alteration when observed by phase-contrast microscopy (not shown). Following the 10-min treatment with H₂O₂, the basal secretion (at 2.8 mM glucose) measured during a 30-min incubation was elevated (Fig. 2B). The clonal cells were more sensitive to H₂O₂ than primary cells, because stimulation with 12.8 mM glucose after the oxidative stress period was unable to evoke any secretory response (Fig. 2B). Moreover, because the basal release was elevated, no further effect of 30 mM KCl was observed above the secretory level reached in control preparations. It should be noted that these experiments were performed following a 2-h preincubation period in glucose-free medium, which slightly decreases basal insulin release (Table II). This could possibly sensitize the cells to oxidative stress because of fuel depletion. Therefore, secretion experiments were repeated without a starvation period and showed a similar pattern as in Fig. 2, A and B, in terms of H₂O₂ sensitivity and blunted glucose (12.8 mM) response (Table II).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Effect of H2O2 on insulin secretion in rat islets stimulated by glucose, methyl succinate, or KCl. Rat pancreatic islets were maintained in culture 2-4 days prior to perifusion experiments in KRBH buffer containing basal 2.8 mM glucose. Insulin secretion was stimulated for 10 min with 16.7 mM glucose (A), 5 mM methyl succinate (C), or 20 mM KCl (E). The oxidative stress was imposed for 10 min by perifusing 200 µM H2O2 followed by a 5-min recovery period in basal buffer supplemented with 100 units/ml catalase. Secretagogues were then applied for 10 min: 16.7 mM glucose (B), 5 mM methyl succinate (D), or 20 mM KCl (F). Traces are the mean + S.E. of 4-6 preparations (see Table I for statistics). The thin lines in A and B are representative control traces of three independent experiments done in triplicate without glucose stimulation.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Effect of H2O2 on insulin secretion in INS-1 cells stimulated by glucose or KCl. INS-1 cells were cultured for 3-5 days and then maintained for 2 h in a glucose-free medium prior to preincubation in glucose free KRBH for 30 min. After further washing, cells were incubated with increasing concentrations of H2O2 for 10 min and insulin release was determined (A). In B, cells were incubated without (Control) or with (Oxidative Stress) 200 µM H2O2 for 10 min, washed again in the presence of 100 units/ml catalase, and finally stimulated with 12.8 mM glucose or 30 mM KCl for 30 min. Basal condition refers to KRBH buffer containing 2.8 mM glucose. Values are the mean ± S.E. of quadruplicates from one of three (A) or four (B) experiments. *, p < 0.02; **, p < 0.002; ***, p < 0.0005 versus A, control value without H2O2 (1.75 ± 0.11% content); B, basal conditions in control groups.

Effect of H₂O₂ on _c in INS-1 Cells-- Bisoxonol fluorescence was used to monitor the _c in a suspension of INS-1 cells. After a baseline at 2.8 mM glucose, 10 mM sugar was added (12.8 mM final), which produced a depolarization of _c, and it was further depolarized by the subsequent addition of 30 mM KCl (Fig. 3A). When 200 µM H₂O₂ was first added to the cuvette, a transient hyperpolarization of _c was followed by a slight, gradual depolarization (Fig. 3B). After a 10-min interval, 100 units/ml catalase was added to neutralize the remaining extracellular H₂O₂. Subsequent exposure to 10 mM glucose caused no significant depolarization of the _c. In contrast, KCl (30 mM) evoked further depolarization (Fig. 3B). When H₂O₂ treatment was applied after glucose stimulation, the glucose-induced cell depolarization was reversed without inhibition of the effect of KCl added subsequently (Fig. 3C).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of H2O2 on c in INS-1 cells. INS-1 cells were kept in suspension for 2 h in a spinner culture with glucose-free medium at 37 °C. For c measurements, 2 × 106 cells were resuspended in 2 ml of KRBH containing the fluorescent dye bisoxonol and transferred to a thermostatted cuvette. In A, the depolarizing effect of 10 mM glucose (12.8 mM final) was tested followed by a control depolarization using 30 mM KCl. In B, 200 µM H2O2 was added for 10 min followed by 100 units/ml catalase before testing the effect of 10 mM glucose (12.8 mM final) and finally KCl (30 mM). In C, 200 µM H2O2 was applied during glucose-induced c depolarization. Each trace is representative of 3-5 independent experiments.

Effect of H₂O₂ on [Ca²⁺]_c in INS-1 Cells-- Attached INS-1 cells stably expressing cytosolic aequorin, a Ca²⁺-sensitive photoprotein, were used to measure [Ca²⁺]_c in a perifusion setup. Glucose (12.8 mM) and KCl (30 mM) raised [Ca²⁺]_c (Fig. 4A). The perifusion of 200 µM H₂O₂ for 10 min resulted in a retarded and sustained increase of [Ca²⁺]_c to ~400 nM (Fig. 4B). When glucose (12.8 mM) was added after H₂O₂ treatment, no further augmentation of [Ca²⁺]_c was observed (Fig. 4C). In contrast, KCl (30 mM) caused a normal rise in [Ca²⁺]_c even after the oxidative stress (Fig. 4D).

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Effect of H2O2 on cytosolic [Ca2+] in INS-1 cells. Adherent INS-1 cells expressing cytosolic aequorin were cultured for 3-5 days before loading with coelenterazine and monitoring of photon emission in a photon counting chamber perifused with KRBH containing basal 2.8 mM glucose. A shows a control trace using cytosolic Ca2+ raising agents, i.e. 12.8 mM glucose and 30 mM KCl. B, 200 µM H2O2 was added for 10 min followed by 100 units/ml catalase (Cat.) (thin line, control trace without H2O2). Effects of 12.8 mM glucose (C) and 30 mM KCl (D) following H2O2 treatment are shown. Each trace is representative of 3-5 independent experiments.

Effect of H₂O₂ on _m in INS-1 Cells-- The _m was measured in a suspension of INS-1 cells by monitoring rhodamine-123 fluorescence. The addition of 10 mM glucose (12.8 mM final) potently hyperpolarized the _m, and 1 µM protonophore FCCP depolarized it (Fig. 5A). H₂O₂ (200 µM) induced an initial moderate and rapid depolarization of _m followed by a slow, progressive further depolarization (Fig. 5B). After catalase treatment, 10 mM glucose (12.8 mM final) failed to hyperpolarize _m and even accelerated the depolarization. The subsequent addition of 1 µM FCCP completed the _m depolarization first initiated by H₂O₂. As a control of enzymatic efficacy, catalase was added before H₂O₂ treatment. In these conditions, H₂O₂ failed to alter the _m, and glucose added thereafter hyperpolarized the _m in a normal manner (Fig. 5B, thin line). When H₂O₂ was applied after glucose stimulation, the hyperpolarization of _m resulting from glucose metabolism was counteracted leading to depolarization of _m (Fig. 5C). H₂O₂ almost completely depolarized the _m because FCCP had only minor effects when added after the oxidative stress.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of H2O2 on m in INS-1 cells. The m was measured in a suspension of 2 × 106 INS-1 cells/2 ml of KRBH using rhodamine-123 fluorescence after a spinner culture period. In A, the glucose-induced (12.8 mM final) hyperpolarization of m was tested followed by the complete depolarization of m using 1 µM of the uncoupler FCCP. In B, 200 µM H2O2 was added for 10 min followed by 100 units/ml catalase before testing the effect of 10 mM glucose (12.8 mM final) and finally FCCP (1 µM). As a control, catalase (Cat.) was also added before H2O2 treatment (B, thin line). In C, 200 µM H2O2 was applied during glucose-induced (12.8 mM final) m hyperpolarization. Each trace is representative of 3-5 independent experiments.

Effect of H₂O₂ on [Ca²⁺]_m in INS-1 Cells-- [Ca²⁺]_m was monitored in perifused INS-1 cells stably expressing mitochondrially targeted aequorin. In control recordings glucose (12.8 mM) raised [Ca²⁺]_m up to 1.3 µM (Fig. 6A). The addition of 200 µM H₂O₂ for 10 min rapidly increased the [Ca²⁺]_m from a baseline of 200 nM to the micromolar range followed by a continuous augmentation up to 1.2 µM (Fig. 6B). After the removal of H₂O₂ and its extracellular depletion by catalase (100 units/ml for 2 min), [Ca²⁺]_m stabilized at a plateau of approximately 1.1 µM. This oxidative stress completely inhibited the glucose response (Fig. 6C). In contrast, H₂O₂ treatment did not affect the [Ca²⁺]_m rise evoked by KCl-induced cell depolarization (Fig. 6D), which also augments [Ca²⁺]_c (Fig. 4D).

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Effect of H2O2 on mitochondrial [Ca2+] in INS-1 cells. Adherent INS-1 cells expressing mitochondrial aequorin were cultured for 3-5 days before loading with coelenterazine and monitoring photon emission in a photon counting chamber perifused with KRBH containing basal 2.8 mM glucose. A shows the effect of 10 mM glucose (12.8 mM final) on mitochondrial Ca2+. B shows the effect of a 10-min treatment with 200 µM H2O2 followed by the addition of 100 units/ml catalase. C and D show the effects of 12.8 mM glucose (C) and 30 mM KCl (D) following H2O2 treatment. Each trace is representative of 3-5 independent experiments.

Effect of H₂O₂ on ATP Generation in INS-1 Cells-- Cellular ATP was measured after appropriate incubation in cell extracts (static) and monitored in living INS-1 cells expressing cytosolic luciferase (perifusion). When measured from static incubations, glucose (12.8 mM) increased the ATP content by 34% compared with nonstimulated cells. A 10-min treatment with 200 µM H₂O₂ reduced the cellular ATP by 57%. When the cells were stimulated with 12.8 mM glucose after the oxidative stress, the ATP content was still 40% lower compared with control cells despite a significant increase relative to cells exposed to H₂O₂ (Table III). When monitored in perifusion, glucose (12.8 mM) increased ATP production approximately 40% above baseline (Fig. 7A). When oxidative stress was applied during glucose stimulation (12.8 mM), the glucose-induced increase in cytosolic ATP was returned to basal levels (Fig. 7B). At 2.8 mM glucose, the addition of 200 µM H₂O₂ provoked a continuous decrease of the cytosolic ATP concentration until the removal of H₂O₂ at the end of the 10-min period of oxidative stress (Fig. 7C). After a recovery phase corresponding to the catalase treatment, the cells were exposed to 12.8 mM glucose, which slightly increased cytosolic ATP. Because the inhibition of ATP generation has been shown to result in a marked decrease of ER Ca²⁺ (24), we have also monitored the latter parameter in INS-1 cells expressing aequorin in the ER compartment. The levels of ER Ca²⁺ were lowered by exposure of the cells to 200 µM H₂O₂ (Fig. 7D). The lag time for the decrease of ER Ca²⁺ was 46.8 ± 3.9 s and the amplitude was 58.3 ± 10.8 µM at 3 min after the addition of H₂O₂ (n = 4).

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Effect of H2O2 on cytosolic ATP and ER Ca2+ in INS-1 cells. Adherent INS-1 cells expressing cytosolic luciferase or ER aequorin were cultured for 3-5 days before recording in a photon counting chamber perifused with KRBH containing basal 2.8 mM glucose. A shows the changes of cytosolic ATP generated by glucose (12.8 mM final). B shows the effect of 200 µM H2O2 during glucose stimulation (12.8 mM) while monitoring cytosolic ATP. C shows the effect of 200 µM H2O2 at basal glucose (2.8 mM) followed by the addition of 100 units/ml of catalase for 5 min before testing the effect of glucose (12.8 mM final). D shows ER Ca2+ changes in control cells (thin line) and cells exposed to 200 µM H2O2 (thick line) at basal glucose (2.8 mM). Each trace is representative of 3-5 independent experiments.

Effect of H₂O₂ on Mitochondrial Morphology in Cells Overexpressing Bcl-2-- INS-1 cells were cultured on glass coverslips transiently transfected with human Bcl-2, and mitochondria were stained with Mitotracker prior to treatment with 200 µM H₂O₂ or 1 µM FCCP for 10 min. Control cells exhibited normal filament-like mitochondria (Fig. 8A). Exposure to H₂O₂ modified the mitochondrial morphology into a more condensed, globular pattern (Fig. 8B). Treatment with the uncoupling protonophore FCCP produced similar effects on the mitochondria (Fig. 8C). Cells overexpressing the antiapoptotic protein Bcl-2 were identified by immunostaining using anti-human Bcl-2 antibody (Fig. 8, E, G, and I). This antibody did not react with the endogenous rat Bcl-2, revealing only the transgene. In control cells, human Bcl-2 co-localized with mitochondria. In addition, there was a diffuse staining, because Bcl-2 is located predominantly, albeit not exclusively, in the outer mitochondrial membrane but also on the endoplasmic reticulum and in the nuclear membrane (Fig. 8E). Overexpression of Bcl-2 did not protect against the alteration of mitochondrial morphology induced by H₂O₂ or by the depolarization evoked by FCCP (Fig. 8, F and H, respectively).

View larger version (57K): [in this window] [in a new window]   Fig. 8.   Effect of H2O2 on mitochondrial morphology in INS-1 cells. Cells were cultured on glass coverslips and (D-I) transfected with Bcl-2. Mitochondria were stained with Mitotracker prior to the appropriate treatment for 10 min followed by fixation and visualization under a laserscan confocal microscope. A, control cells; B, cells treated with 200 µM H2O2; C, cells treated with 1 µM of the uncoupler FCCP. In D-I, cells were transiently transfected with human Bcl-2. Cells expressing the transgene were identified by immunostaining using anti-human Bcl-2 antibody (E, G, and I). D, F, and H show the mitochondrial staining of untransfected cells and Bcl-2-transfected cells (arrow) for control, H2O2, and FCCP conditions, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mitochondria play a key role in the control of nutrient-induced insulin secretion by coupling glycolysis to distal events leading to exocytosis (1, 2, 35). However, this highly aerobic organelle also produces H₂O₂ from dismutated superoxide anions (9) generating an intracellular oxidative stress that may damage neighboring molecules. Among these, mitochondrial enzymes such as aconitase (18) and adenine-nucleotide translocase (19) are susceptible to oxidative modification. Mitochondrial DNA is also highly sensitive to oxidative stress (8, 9). In the present study, we used H₂O₂ as a recognized biologically important oxidant (20). The cells were exposed to 200 µM H₂O₂, a level seen during activation of phagocytes in vitro (4).

Exposure of rat islets to H₂O₂ resulted in a transient increase in insulin release at basal nonstimulatory glucose concentration and impaired the glucose-induced secretory response tested subsequent to the oxidative stress. A similar pattern was obtained using the mitochondrial substrate methyl succinate as a secretagogue, suggesting that pathways downstream of glycolysis are susceptible to oxidative alterations. On the contrary, the KCl-evoked cell depolarization was still able to promote insulin exocytosis following exposure to H₂O₂ in the islet perifusion experiments. This points to the mitochondrion rather than to the exocytotic process as the sensitive site. These results are in accordance with a recent report from our laboratory (36) using the diabetogenic compound alloxan known to generate H₂O₂ and free radicals (20). Exposure of INS-1 cells to alloxan for 5 min resulted in elevated basal insulin release; thereafter glucose failed to stimulate oxidative metabolism and insulin secretion, whereas the secretory response to KCl was largely preserved (36). Taken together, these data suggest that oxidative stress primarily alters nutrient-evoked insulin secretion and that the elevated basal insulin release is the consequence of an increase in [Ca²⁺]_c. Indeed, treatment of INS-1 cells with either H₂O₂ (present study) or alloxan (36) results in the elevation of [Ca²⁺]_c. The steady state [Ca²⁺]_c attained under these conditions was similar to that measured with KCl on the plateau phase, which could explain why the effects of H₂O₂ and KCl on insulin secretion in static incubation were not additive.

In the mitochondria, exposure to H₂O₂ caused a rapid elevation of Ca²⁺, even at early time points when [Ca²⁺]_c was still at basal level, excluding a simple transfer between these two compartments. Rather, the first phase of [Ca²⁺]_m rise might be explained by a direct inhibition by H₂O₂ of the Na⁺/Ca²⁺ antiporter that controls the efflux of Ca²⁺ from the mitochondrial matrix (37). Transfer of reducing equivalents to the electron transport chain increases _m, which enhances the driving force for mitochondrial Ca²⁺ uptake mediated by a low affinity uniporter (38). The failure of glucose to hyperpolarize _m following oxidative stress undoubtedly explains the lack of the normal [Ca²⁺]_m increase during glucose stimulation (28). These mitochondrial alterations led to the diminished ATP synthesis and probably to an insufficient generation of additional mitochondrial factors required for nutrient-evoked insulin exocytosis (10). The depolarized _m explains the altered mitochondrial morphology, which was not prevented by overexpression of Bcl-2 (Fig. 8). Both alterations are also observed in mitochondrial DNA-deficient INS-1 cells associated with impaired nutrient-induced insulin secretion (39). It is noteworthy that a mutation in the human mitochondrial tRNA^Leu(UUR) gene, which is associated with mitochondrial diabetes, also yields a similar mitochondrial morphology (40).

[Ca²⁺]_c was elevated following H₂O₂ treatment, and glucose added thereafter failed to elicit a [Ca²⁺]_c response. This was not the case for the KCl-evoked depolarization of _c, which resulted in normal [Ca²⁺]_c rise, suggesting that the voltage sensitive L-type Ca²⁺ channels were still functional. Similar effects of H₂O₂ have been reported recently in the CRI-G1 insulin secreting cell line (41) and in mouse islets (24). Krippeit-Drews et al. (24) postulated that the first phase of intracellular Ca²⁺ rise is because of Ca²⁺ mobilization from the mitochondria, whereas the second phase would reflect Ca²⁺ influx through pathways distinct from L-type Ca²⁺ channels. Herson et al. (41) suggested that the first phase is caused by mobilization of intracellular Ca²⁺ stores insensitive to thapsigargin and therefore proposed the mitochondria as the source of the initial [Ca²⁺]_c rise. Regarding the second phase, these authors (41) postulated that H₂O₂ opens a nonselective cation channel in the plasma membrane, which was described in a previous study as a novel Ca²⁺ influx pathway activated by oxidative stress (42). We did not observe the second phase [Ca²⁺]_c rise probably because of the 50 times lower concentration of H₂O₂ used in the present study. The measurements of Ca²⁺ changes both in the mitochondrial and ER compartments permitted us to conclude that the first phase of [Ca²⁺]_c elevation is probably because of diminished pumping of Ca²⁺ into the ER as a consequence of a decrease in cytosolic ATP (30, 43). Indeed, ER Ca²⁺ was lowered by H₂O₂ treatment secondary to the marked decrease in cytosolic ATP and before the elevation of [Ca²⁺]_c. These results substantiate the critical role of cytosolic ATP in determining the filling state of the ER Ca²⁺ stores.

The following sequence of events for the actions of H₂O₂ can be proposed. We suggest that hydrogen peroxide or derivative products such as free radicals directly inhibits the mitochondrial Na⁺/Ca²⁺ antiporter, as previously observed (37), blocking the efflux of Ca²⁺ from the mitochondria. This inhibition would lead to the rapid rise in [Ca²⁺]_m independent of any changes in the cytosolic compartment, an effect recently reported in endothelial cells (37). Meanwhile, oxidative stress inhibits tricarboxylic acid cycle enzymes such as aconitase (18), attenuating the generation of reducing equivalents, in particular NADH. This would lead to inactivation of the electron transport chain with the concomitant loss of _m and the decline in ATP generation. Moreover, oxidative modification of the adenine-nucleotide translocase would inhibit the translocation of ATP to the cytosol (19). The depolarization of _m is unlikely to be because of the opening of the transition pore, at least on this short time scale, because overexpression of Bcl-2 did not prevent the mitochondrial effects evoked by H₂O₂ treatment.

The retarded elevation in [Ca²⁺]_c following exposure to H₂O₂ may well be explained by the observed decrease in cytosolic ATP. Indeed, lowering of cytosolic ATP by the mitochondrial uncoupler FCCP results in a marked decrease of ER Ca²⁺ as monitored in INS-1 cells expressing aequorin in this cellular compartment (30). Reduction of cytosolic ATP results in impaired pumping of Ca²⁺ into the ER, an effect mimicked by the sarco/endoplasmic reticulum Ca²⁺-ATPase inhibitors. The latter effect has been demonstrated in insulin-secreting cell lines (30, 41) and has also been shown to lead to apoptosis (44, 45). Thus, the lowering of ER Ca²⁺ by H₂O₂ treatment is another demonstration of the close functional coupling between mitochondria and ER Ca²⁺ stores (45, 46). This effect on ER Ca²⁺ certainly accounts for the rise in [Ca²⁺]_c observed shortly after H₂O₂ treatment. Ca²⁺ influx from the extracellular space was also reported to occur following exposure to H₂O₂ in mouse islets (24) or to alloxan (generating free radicals) in INS-1 cells (36). It has been postulated that a nonselective cation channel mediates this retarded Ca²⁺ influx, which is promoted by oxidative stress (41). Whatever the mechanism, it can be surmised that both the inhibition of Ca²⁺ pumping by the sarco/endoplasmic reticulum Ca²⁺-ATPase into the ER and the Ca²⁺ influx from the extracellular space contribute to the elevation of [Ca²⁺]_c.

Dysfunction of the beta cell by oxidative stress is implicated in type I diabetes and in aging. The present study provides evidence for the involvement of the mitochondria in impaired signal transduction, coupling glucose metabolism to insulin secretion following an oxidative stress. Moreover, a model is suggested for the sequence of events leading to the elevation of [Ca²⁺]_c. Further work will address the identification of factors normally required for nutrient-induced insulin secretion but are missing in the damaged beta cell.

	ACKNOWLEDGEMENT

We are indebted to C. Bartley, G. Chaffard, and H. Petersen for expert technical assistance.

	FOOTNOTES

^* This work was supported by the Swiss National Science Foundation Grants 32-49755.96 (to C. B. W.) and 31.37799.93 (to L. J.) and by a grant from the Silva-Casa Foundation attributed through the AETAS Foundation for Research on Aging (Geneva) (to C. B. W.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Division de Biochimie Clinique, Centre Médical Universitaire, 1 rue Michel-Servet, CH-1211 Geneva 4, Switzerland; Tel.: 41-22-702-55-48; Fax: 41-22-702-55-43; E-mail: claes.wollheim@medecine.unige.ch.

	ABBREVIATIONS

The abbreviations used are: [Ca²⁺]_c, cytosolic calcium concentration; [Ca²⁺]_m, mitochondrial calcium concentration; KRBH, Krebs-Ringer bicarbonate HEPES buffer; _c, cell membrane potential; _m, mitochondrial membrane potential; NO, nitric oxide; ER, endoplasmic reticulum; FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES