Taurine Supplementation Enhances Nutrient-Induced Insulin Secretion in Pancreatic Mice Islets

Home , Vanadyl sulfate

DIABETES/METABOLISM RESEARCH AND REVIEWS RESEARCH ARTICLE Diabetes Metab Res Rev 2009; 25: 370–379. Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/dmrr.959

Taurine supplementation enhances nutrient-induced insulin secretion in pancreatic mice islets

Rosane A. Ribeiro1 Abstract Maria L. Bonfleur2 Andressa G. Amaral1 Background Taurine (TAU), a naturally occurring sulfur-containing amino Emerielle C. Vanzela1 acid, is found at high concentrations in plasma and mammalian tissues and regulates osmolarity, ion channel activity, and glucose homeostasis. Silvana A. Rocco3 Several reports have shown that physiological plasma TAU levels seem to be 1 Antonio C. Boschero important for adequate beta (β)-cell function and insulin action, since low 1 Everardo M. Carneiro * concentrations of TAU in the plasma have been reported in the pre-diabetic and diabetic states. 1Departamento de Fisiologia e Biof´ısica, Instituto de Biologia, Methods Glucose tolerance and insulin sensitivity were investigated in mice Universidade Estadual de Campinas supplemented with 2% (w/v) TAU in their drinking water for 30 days, as well (UNICAMP), Campinas, SP, Brazil as the insulin secretion from isolated islets stimulated by glucose or L-leucine. 2Centro de Ciências Biológicas e da Saúde, Universidade Estadual do Results TAU-supplemented mice demonstrated improved glucose tolerance Oeste do Paraná(UNIOESTE), and higher insulin sensitivity, compared to controls (CTL). In addition, their Cascavel, PR, Brazil islets secreted more insulin in response to high concentrations of glucose and 14 3Departamento de Cl´ınica Médica, L-leucine. L-[U- C]leucine oxidation was higher in TAU than in CTL islets, 14 Faculdade de Ciências Médicas whereas D-[U- C]glucose oxidation, ATP levels, glucose transporter (GLUT) Universidade Estadual de Campinas 2 and glucokinase (GCK) protein expressions were similar in both types of 2+ (UNICAMP), Campinas, SP, Brazil islets. The L-type β2 subunit voltage-sensitive Ca channelprotein,aswellas 45Ca uptake, were significantly higher in TAU-supplemented than CTL islets. *Correspondence to: In addition, islets from TAU-supplemented mice secreted more glucagon than Everardo M. Carneiro, CTL islets at low glucose. Departamento de Fisiologia e Biof´ısica, Instituto de Biologia, Conclusions TAU supplementation improves glucose tolerance and insulin Universidade Estadual de Campinas sensitivity in mice, as well as insulin secretion from isolated islets. The latter (UNICAMP), C.P. 6109, CEP effect seems to be, at least in part, dependent on a better Ca2+ handling by 13083-970, Campinas, SP, Brazil. the islets. Copyright  2009 John Wiley & Sons, Ltd. E-mail: [email protected]

Keywords glucose homeostasis; insulin secretion; taurine supplementation

Introduction

Glucose is the main stimulus for insulin secretion from pancreatic β- cells. The stimulus-secretion coupling induced by glucose culminates with increased intracellular Ca2+ concentration ([Ca2+]i), which together with metabolic signals results in a complete biphasic insulin secretion [1]. As with glucose, amino acids exert insulinotropic activity in β-cells. There are three different mechanisms by which amino acids may increase insulin Received: 7 November 2008 secretion: direct depolarization of the plasma membrane by transport of Revised: 21 February 2009 cationic amino acids, such as L-arginine; cotransport with Na+,which Accepted: 2 March 2009 alsoresultsinanincreasein[Ca2+]i; and metabolism of the amino

Copyright  2009 John Wiley & Sons, Ltd. Taurine Supplementation Increases Insulin Secretion 371 acid, resulting in an increase in ATP and in the ATP-to-ADP Animals ratio [2]. TAU (2-aminoethanesulphonic acid), a naturally occur- All experiments were approved by the ethics committee ring sulfur-containing amino acid, which is not incorpo- at UNICAMP. Three-week-old Swiss mice were obtained rated into protein and not used for energy production, from the colony at UNICAMP. The mice were maintained is found at high concentrations in pancreatic islet cells on a 12-h light/darkness cycle (lights on 06 : 00–18 : 00 [3]. Previous studies have shown that this amino acid in h), controlled temperature (22 ± 1 ◦C), and allowed free mammalian tissues regulates several biological processes, access to water and standard laboratory chow (Rodent + including osmolarity [4], Ca2 binding and transport chow; Nutrilab, Colombo, Brazil) ad libitum.At60days, [5–9] and ion channel activity [8,10]. TAU also prevents mice were distributed into two groups: mice that received tissue injury by oxidative stress [11,12], maintains the 2% of TAU in their drinking water for 30 days (TAU structural integrity of the plasma membrane [13], and group, as previously reported [25]) and those in the demonstrates hypoglycemic properties [14–16]. control (CTL) group that received only water. A relationship between low concentrations of TAU in the plasma and the pre-diabetic and diabetic states has also been suggested [17–19]. Thus, restoring the normal General nutritional parameters plasma TAU levels by supplementation may help in the prevention of diabetes mellitus [19,20]. In this regard, the Body weight, food and water intake were measured during maintenance of physiological plasma TAU concentrations the experimental period and expressed by the area under seems to be an important factor for adequate β-cell the curve (AUC) of total body weight and food and water function and insulin action [12,18,20–23]. However, the intake, respectively, during the 30 days of the treatment. physiological role of this amino acid in the endocrine At the end of the supplementation period, fasted and pancreas has not yet been completely elucidated. fed mice were decapitated, their blood collected and the ◦ Conflicting results concerning the effects of TAU on plasma stored at −20 C. Total plasma protein and plasma isolated islets have been found in the literature. It has albumin were measured using standard commercial kits, been shown that TAU does not affect the first phase insulin according to the manufacturer’s instructions (Laborlab, secretion induced by high concentrations of glucose in Guarulhos, SP, Brazil). Plasma glucose was measured isolated rat islets, but instead reduces the rate of insulin using a glucose analyser (Accu-Chek Advantage, Roche secretion during the second phase [3]. In contrast, TAU Diagnostic, Switzerland), insulin was measured by RIA, plus physiological concentrations of glucose significantly and glucagon plasma concentration was quantified by stimulated insulin secretion [17,24]. The ability of TAU commercial kit (Linco Research, St. Charles, MO, USA). to ameliorate insulin secretion, stimulated by glucose and other fuels in fetal and adult isolated islets, was also observed when the plasma TAU of rodents was increased TAU plasma levels or at least restored by TAU supplementation [17,24,25]. In this study, we confirm previous observations that To measure TAU plasma concentrations in fed and fasted TAU supplementation improves glucose tolerance and conditions, 200 µL of plasma were deproteinized by insulin sensitivity in mice [25]. In addition, we observed adding 200 µL of 25% trichloroacetic acid solution and that TAU supplementation improves the islet hormone were then centrifuged at 21 000 g for 10 min. Supernatant secretion, represented by higher insulin release stimulated (100 µL) was collected and mixed with sample loading by glucose or L-leucine and higher glucagon at low buffer (100 µL) (Biochrom 20 reagent kit, Cambridge, glucose levels. These islets also metabolized more L- UK). An aliquot of 25 µL of the mixture was then leucine. Finally TAU-supplemented islets accumulated resolved by liquid chromatography on a Biochrom 20 plus more Ca2+ in the presence of high glucose concentrations, amino acid analyser (Amersham Pharmacia, Piscataway, and showed an increased expression of the L-type β2 NJ, USA). Amino acid standards were analysed first, subunit Ca2+ channel. followed by the samples. Amino acids were quantified using Biochrom 20 control software, version 3.05.

Materials and Methods Intraperitoneal glucose (ipGTT) and insulin tolerance test (ipITT) Materials For intraperitoneal glucose tolerance test (ipGTT), blood 14 14 45 D-[U- C]glucose, L-[U- C]leucine and CaCl2 were glucose and insulin levels (time 0) were measured (as purchased from Amersham International (Little Chalfont, previously described) in overnight fasted mice. A glucose Bucks, UK). Routine reagents, L-leucine, L-glutamine, load of 2 g/kg body weight was then administered by 2-aminobicyclo[2,2,1]-heptane-2-carboxylic acid (BCH), ip injection and additional blood samples were collected aminooxyacetic acid (AOA) and ATP were purchased from at 15, 30, 60, 120 and 180 min for measuring plasma Sigma Chemical (St Louis, MO, USA). glucose. At 30 and 60 min of the ipGTT plasma insulin

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr 372 R. A. Ribeiro et al. was also evaluated. For intraperitoneal insulin tolerance 1mol/L was added to the tubes, and the samples were test (ipITT), fed mice were injected with 0.75 U/Kg body stored at −20 ◦C. On the day of the protein assay, 25 µLof weight of human insulin (Biohulin R, Biobras,´ Brazil). HCl 1N was added to the samples, and islet total protein Blood samples were collected before insulin injection, was quantified by the Bradford method [29]. DNA was and at 15, 30, 60 and 90 min for glucose analysis. The measured using fluorimetric probes and a standard curve glucose disappearance rate (Kitt) was calculated using of known DNA concentrations. the formula, 0.693/t1/2.Glucoset1/2 was calculated from the slope of the least-squares analysis of plasma glucose concentrations during the linear decay phase [26]. ATP and ADP levels Islet isolation, static insulin and glucagon secretion To determine ATP and ADP concentrations, groups of 150 islets were pre-incubated for 30 min at 37 ◦Cin Islets were isolated by collagenase digestion of the a KRB solution with 2.8 mmol/L glucose, after which pancreas, as described [27]. For static incubations, the solution was replaced with fresh buffer containing groups of four islets were first incubated for 30 min 22.2 mmol/L glucose (gassed with a mixture of 95% at 37 ◦C in Krebs-Ringer bicarbonate (KRB) buffer with O2/5% CO2) and incubated for 15 min. At the end of 2.8 mmol/L glucose and 3 g of BSA/L, and equilibrated the incubation period, the supernatant was discarded and the islets were frozen in liquid nitrogen. Before with a mixture of 95% O2/5% CO2 to give pH 7.4. This medium was then replaced with fresh buffer and the islets chromatographic analysis, islets were resuspended in incubated for 1 h under the following conditions: glucose 1 mL of a solution containing 50 mmol/L KH2PO4 and 2.8 mmol/L alone or with L-leucine (10 mmol/L), BCH 25 mmol/L citric acid (pH 4.5). The mixture was kept in ◦ (10 mmol/L; a non-metabolized analog of L-leucine), awaterbath(85 C) for 2 min, and the pellet of islets L-glutamine (10 mmol/L) and AOA (10 mmol/L); and was then lysed by mechanical stress and filtered through glucose (22.2 mmol/L) with or without glucagon (43 a0.45 µm Millex filter (Millipore, Milford, MA, USA). pmol/L). At the end of the incubation period, the insulin Subsequently, 200 µL of filtrate was mixed with 20 µL ◦ content of the medium was measured by RIA. For of 2-chloroacetaldehyde solution and heated at 80 C glucagon analyses, groups of 15 islets were incubated for 20 min [30,31]. An aliquot of 25 µL of the reaction for 30 min in 250 µL of KRB buffer with 5.6 mmol/L mixture was then resolved by liquid chromatography. glucose. This medium was then replaced with 250 µL Chromatography: Chromatographic analyses were car- of fresh buffer and the islets were further incubated ried out on a Waters Alliance equipment series 2695 with 0.5 mmol/L glucose for 1 h. Glucagon release (Milford, MA, USA) equipped with a quaternary pump, was measured as previously described and results were a sampler manager, a degasser, and a Waters 2475 flu- expressed as pg/15 islets.h [28]. orescence detector model. The fluorescence of derivative compounds (ATP, ADP) was monitored with excitation and emission wavelengths set at 280 and 420 nm, respec- Dynamic insulin secretion studies tively. Chromatographic separations of the compounds were achieved at room temperature, using a reversed- Groups of 50 freshly isolated islets were placed on phase Cosmosil 5C18-MS column (150 × 4.6mmid.; . µ Millipore SW 1300 filters (8 0 m pore) and perifused in 5 µm particle size), with a Cosmosil guard column a KRB buffer at a flow rate of 1mL/min for 30 min in the (5C18-MS 10 × 4.6 mm) purchased from Phenomenex presence of 2.8 mmol/L glucose (basal conditions). After (Torrance, CA, USA). The column was equilibrated and this period, the islets were perifused with 22.2 mmol/L of eluted under gradient conditions using a flow rate of glucose, or 10 mmol/L of L-leucine, or 10 mmol/L of L- 1 mL/min. The standards and samples were separated leucine and 10 mmol/L of L-glutamine, as indicated in the using a gradient mobile phase consisting of methanol figure legends and Results Section. Perifusion solutions (A) and a solution of 50 mmol/L KH PO and 25 mmol/L were gassed with 95% O /5% CO and maintained at 2 4 2 2 citric acid (pH 4.5) (B), which was prepared immediately 37 ◦C. Insulin release was measured by RIA. before use and filtered through a 0.45 µm filter (Milli- pore, Milford, MA, USA). The gradient conditions were: Islet insulin, total protein and DNA 0–4 min, 2% A; 4–12 min linear gradient, 2–15% A; content 17–18 min reconditioning step of column was 2% A isocratic for 2 min. The chromatographic run time for each Groups of 4 islets were collected and transferred to tubes analysis was 20 min. Aliquots of 25 µL were injected into of 1.5 mL. Deionized water (1 mL) was added to the the HPLC system. System control, data acquisition and samples, followed by sonication of the pancreatic cells processing were performed with a PC-Pentium IV Proces- (3 times, 10-s pulses), and the islet insulin content was sor personal computer from Dell, operated with Microsoft measured by RIA. For protein quantification, groups of 15 Windows XP version 2003, using Waters Empower 2002 islets were transferred to tubes of 0.6 mL, 25 µLofNaOH chromatography software.

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr Taurine Supplementation Increases Insulin Secretion 373

Glucose and L-leucine oxidation Western blotting was repeated using β-actin (1 : 10 000, Abcam, Cambridge, MA, USA) antibody as an internal Glucose and L-leucine oxidation were measured by control. 14 14 the rate of formation of CO2 from D-[U- C]glucose (20 µCi/mL) or L-[U-14C]leucine (8 µCi/mL) by the islets. Groups of 25 islets were incubated for 2 h at 37 ◦Cin Statistical analysis 30 µLofKRBsupplementedwith22.2mmol/Lofglucose ± or 10 mmol/L of L-leucine. After incubation, oxidation The results are presented as means SEM for the number 14 of determinations (n) indicated. The statistical analyses was stopped with HCl for cell cleavage. CO2 liberated 14 were performed using unpaired Student’s t-test, and the was absorbed by NaOH, and the NaH CO3 obtained was measured by liquid scintillation in a β radiation counter. level of signiﬁcance was set at P < 0.05.

Uptake of 45Ca by isolated islets Results

Groups of 150–200 islets, derived from the same batch Mice features of islets, were pre-incubated for 30 min at 37 ◦CinaKRB buffer containing 2.8 mmol/L of glucose, pH 7.4. The Food and water intakes were similar between the two islets were then incubated for 5 min and 1h in 200 µL groups during the period of TAU supplementation (data 45 of the same medium containing CaCl2 (60 µCi/mL) not shown). At the end of this period, body size and and 22.2 mmol/L glucose. At the end of the incubation weight, as well as various organ weights (liver, spleen, period, 800 µL of ice-cold medium containing 2 mmol/L heart, kidneys, and retroperitoneal and periepididimal of LaCl3 (pH 7.4) was added to stop the reaction. The fats) were not different between TAU-supplemented medium was then removed and an aliquot was saved to and CTL mice (data not shown). Plasma parameters, determine the amount of 45Ca in the solution. The islets such as fasted and fed insulin, fasting glucose, albumin were subsequently washed three times with fresh ice-cold and total protein were also similar between the two + La3 -containing medium, and the islets were then placed groups (Table 1). Glucagon plasma levels from fasted, in a Petri dish and groups of ten islets were transferred to but not fed, TAU-supplemented mice were 35% higher counting vials containing 1 mL of EGTA 0.5 mmol/L. The than CTL (P < 0.005) (Table 1). Table 1 shows that + uptake of 45Ca was expressed as pmol Ca2 per islet per the supplementation methodology applied in our study time of the incubation. efficiently increased plasma TAU concentrations, since TAU plasma levels in fasted and fed supplemented mice were approximately 3 times higher, compared with CTL Western blotting mice (P < 0.05). In addition, the daily TAU intake was 173.5 ± 1.5 mg/day in supplemented mice. Pools of islets were transferred to tubes containing pro- tease inhibitor buffer [32]. Samples were sonicated (3 times, with 10-s pulses) and the protein was determined Glucose homeostasis by the Bradford method using BSA as the standard. For SDS gel electrophoresis and Western blot analysis, At the end of the experimental period, CTL and the samples were homogenized with a loading buffer TAU-supplemented mice were submitted to an ipGTT containing DTT. After heating to 95 ◦Cfor5min,the and ipITT. After glucose loading, plasma glucose proteins were separated by electrophoresis (55 µgpro- concentration reached maximal levels at 30 min in both tein/lane, 10% gels). Following electrophoresis, proteins were transferred to nitrocellulose membranes. The mem- Table 1. Plasma glucose, insulin, glucagon, TAU, albumin and branes were subsequently blotted with specific polyclonal total protein concentrations in fasted and fed CTL- and 2+ TAU-supplemented mice antibodies to the β2 subunit of the voltage-sensitive Ca channel (1 : 1000; Sigma Chemicals, St Louis, MO, USA), CTL TAU GLUT 2 (1 : 1000; Calbiochem, San Diego, CA, USA) or GCK (1 : 500; Santa Cruz Biotechnology, Inc., Santa Glucose (mg/dL) Fasted 83.8 ± 6.586.6 ± 3.6 ± ± Cruz, CA, USA). Visualization of specific protein bands Fed 124.1 4.3 125.3 4.3 Insulin (ng/mL) Fasted 0.34 ± 0.07 0.31 ± 0.05 was performed by incubating the membranes with goat Fed 3.0 ± 0.43.0 ± 0.5 anti-rabbit secondary antibody (1 : 10 000; Zymed Lab- ∗ Glucagon (pg/mL) Fasted 68.6 ± 7.4 105.3 ± 4.3 oratories, Inc., San Francisco, CA, USA) and detecting Fed 95.5 ± 8.3 107.1 ± 10.5 with 1 mL enhanced chemiluminescence reagents (Pierce TAU (µmol/mL) Fasted 2.7 ± 0.48.4 ± 0.8∗ ∗ Biotechnology, Rockford, IL, USA) followed by exposure Fed 3.5 ± 0.59.6 ± 2.0 Albumin (g/dL) 2.4 ± 0.12.5 ± 0.1 to X-ray film (Kodak, AM, Brazil). The band intensities Total proteins (g/dL) 4.9 ± 0.24.8 ± 0.1 were quantified by optical densitometry (Scion, Image, Frederick, MD, USA). After assay of the target proteins, Data are means ± SEM (n = 3–12). ∗p < 0.05 vs CTL.

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr 374 R. A. Ribeiro et al.

Figure 1. Effect of TAU supplementation on glucose tolerance and insulin sensitivity. Figures A and B represent changes in plasma glucose levels during ipGTT and ipITT, respectively. Data are means ± SEM obtained from 8 mice. ∗P < 0.05 indicates significant difference vs CTL groups. However, the TAU-supplemented mice showed lower glucose values at 15, 30, 60 and 120 min, compared with CTL mice (Figure 1(A)). The AUC of the glucose plasma levels during the ipGTT in TAU was 37% lower Figure 2. Dynamic insulin release induced by 22.2 mmol glucose than the CTL group (P < 0.01). Insulin plasma levels (G) (A); 2.8 mmol/L glucose and 10 mmol/L L-leucine (B); and measured during min 30 and 60 of the ipGTT did by 2.8 mmol/L glucose, 10 mmol/L L-leucine and 10 mmol/L not differ between TAU and CTL mice (0.66 ± 0.05 L-glutamine (C) in islets from TAU-supplemented and CTL ± and 1.18 ± 0.2 vs 0.64 ± 0.12 and 1.21 ± 0.03, at 30 mice. Data are means SEM obtained from 4 independent experiments. ∗P < 0.05 indicates significant difference vs CTL and 60 min, respectively). After insulin administration (ipITT), plasma glucose showed a constant decrease between min 15 and 60 in both mice groups (Figure 1(B)). 0–25). During the stimulatory period with 22.2 mmol/L The glucose decrease was significantly higher in TAU- glucose, the insulin secretion was significantly higher in TAU-supplemented than CTL islets. The AUC for supplemented than in CTL mice with a Kitt of 1.9 ± 0.4 insulin released during min 25–90 of perifusion was and of 0.4 ± 0.2%/min for TAU-supplemented and CTL − ± ± / 1 mice, respectively (P < 0.02). 319 24 and 125 11 ng 50 islets. min , respectively (P < 0.0001). When dynamic insulin release induced by 10 mmol/L of L-leucine was analysed (Figure 2(B)), a Insulin and glucagon secretion similar biphasic secretion was observed in both types of islets. Again, the insulin secretion was significantly Figure 2(A) shows that the increase in glucose concentra- higher in islets from TAU-supplemented compared with tions from 2.8 to 22.2 mmol/L induced a biphasic insulin CTL mice. The sum of insulin released (AUC) during secretion in both types of islets (min 25–90). This biphasic the stimulatory period with L-leucine was 129 ± 20 and − secretion is represented by a first prominent phase, fol- 44 ± 7ng/50 islets. min 1,respectively(P < 0.008). lowed by a period where the secretion was lower than the L-leucine is known to stimulate insulin secretion by two first phase, but still higher than the adaptation period (min mechanisms; firstly, by L-leucine catabolism, and secondly

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr Taurine Supplementation Increases Insulin Secretion 375 by allosteric activation of glutamate dehydrogenase pmol/L glucagon (a concentration close to that observed (GDH), resulting in an increase of the catabolism of in the plasma of fasted mice). Glucagon-induced insulin L-glutamate [33]. As such, to discriminate whether secretion was 43% higher in TAU-supplemented than CTL the higher L-leucine-induced insulin secretion in TAU- islets (16.7 ± 1.6 vs 11.6 ± 1.7 ng/islet.h, respectively; supplemented islets was due to a possible alteration in P < 0.05). GDH activity and/or L-leucine metabolism, dynamic and static insulin release was analysed in the presence of GDH 14 14 activators and L-leucine metabolism inhibitors. D-[U- C]glucose and L-[U- C]leucine The L-leucine-induced dynamic insulin secretion in TAU oxidation, ATP and ADP islet levels, and CTL islets was similar when the perifusion medium GLUT 2 and GCK protein expression contained 10 mmol/L L-glutamine (Figure 2(C)). In static experimental conditions, increased insulin secretion in D-[U-14C]glucose oxidation in response to 22.2 mmol/L response to L-leucine in TAU islets was confirmed. glucose was similar between groups (35.3 ± 4.1; 40.2 ± − Figure 3 shows that at 2.8 mmol/L glucose, the increment 5.9 pmol/islet/2 h 1, for TAU and CTL, respectively), in the insulin secretion induced by 10 mmol/L L-leucine whereas L-[U-14C]leucine oxidation in the presence was significantly higher in TAU than CTL islets (P < of 10 mmol/L of L-leucine was 28% higher in TAU 0.03). Although significantly higher than basal conditions compared with CTL islets (22.9 ± 1.6and17.8 ± − (2.8 mmol/L glucose alone), the insulin secretion induced 0.8 pmol/islet/2 h 1,respectively;P < 0.008). The ATP by L-leucine plus 10 mmol/L AOA (a transaminase concentration and the ATP-to-ADP ratio in the presence inhibitor), 10 mmol/L BCH (a stimulator of GDH), or of 22.2 mmol/L of glucose were similar in TAU and CTL the combination of BCH and 10 mmol/L L-glutamine islets after 15 min of incubation (Table 2). In addition, (that is converted in L-glutamate, a substrate for GDH), the GLUT 2 and GCK protein expressions were similar in was similar in both types of islets. The L-leucine-induced the TAU and CTL groups (Figure 4(A) and (B)). secretion was also similar in both groups when the islets were incubated in the presence of BCH and L-glutamine. Finally, the addition of AOA significantly reduced the Islet insulin and total protein content insulin secretion to similar levels to those obtained when only GDH activity was present. Total insulin content was 48% higher in TAU compared Glucagon release in response to 0.5 mmol/L glucose to CTL islets (130 ± 9; 87 ± 10 ng/islet, respectively; was higher in islets from TAU-supplemented mice, P < 0.005). Protein and DNA content did not differ compared with CTL islets (491 ± 16; 375 ± 34 pg/15 between TAU and CTL islets (0.28 ± 0.01 and 1.14 ± islets.h, respectively; P < 0.01). To verify the stimulatory 0.1 vs 0.25 ± 0.01 µg/islet and 1.02 ± 0.07 ng/islet, effect of glucagon on insulin secretion, islets were respectively), indicating that TAU supplementation did incubated in the presence of 22.2 mmol/L glucose and 43 not alter islet cell mass.

45Ca uptake and Ca2+ channel protein expression

45Ca uptake by TAU-supplemented and CTL islets at 22.2 mmol/L glucose after 5 min incubation was similar (1.54 ± 0.21 vs 1.32 ± 0.12 pmol 45Ca/islet.5 min, n = 37–33; respectively). However, after 1 h of incubation (representing the steady state), 45Ca uptake by TAU islets was signiﬁcantly higher than CTL islets (Figure 5, P < 0.009). Western blotting analysis revealed that the expression of the β2 subunit of the L-type voltage- sensitive Ca2+ channel was signiﬁcantly higher in TAU- supplemented, compared with CTL islets (P < 0.04)

Table 2. ATP, ADP concentrations and ATP-to-ADP ratio in islets from CTL- and TAU-supplemented mice. Islets were incubated with 22.2 mmol/L glucose for 15 min

CTL TAU Figure 3. Comparative effects of the GDH activity and L-leucine metabolism on insulin secretion in response to L-leucine. Islets ATP (pmol/islet) 1.40 ± 0.17 1.54 ± 0.16 were incubated with 2.8 mmol/L glucose (G) and/or stimulated ADP (pmol/islet) 2.22 ± 0.17 2.77 ± 0.25 with a combination of L-leucine, AOA, BCH and L-glutamine ATP-to-ADP ratio 0.59 ± 0.04 0.58 ± 0.02 (10 mmol/L for all stimuli). Data are means ± SEM (n = 14 for each group). ∗P < 0.05 indicates signiﬁcant difference vs CTL Data are means ± SEM of 10 values from three independents experiments.

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr 376 R. A. Ribeiro et al.

Figure 5. Islet 45Ca uptake after 1 h of incubation in the presence of 22.2 mmol/l glucose. Data are means ± SEM of 25–21 observations for TAU-supplemented and CTL islets, respectively. ∗P < 0.05, signiﬁcant difference vs CTL

[19,25,34–36], our study also demonstrates that TAU supplementation enhances glucose tolerance and insulin sensitivity in vivo. Hypoglycemic properties of TAU have been suggested to be mediated by the interaction of TAU with the insulin receptor (IR) [15]. It has also been reported that this amino acid programs glucose metabolism in mice and amplifies insulin action, since in insulin-resistant rodents TAU supplementation normalizes plasma insulin and glucose levels, insulin sensitivity, blood pressure, and improves hypercholesterolemia and energy expenditure in obesity [18,19,34–37]. In accordance with these studies, previous data from our laboratory showed that enhanced glucose homeostasis in TAU-supplemented mice is due to a higher liver and muscle IR phosphorylation in response to insulin [25]. This effect is probably due to a TAU interaction with the IR in peripheral tissues, because acute TAU ip injection to CTL mice increased IR phosphorylation in these tissues to levels similar to those found when insulin was administered [25]. Such a process may be responsible for the higher glucose tolerance and insulin sensitivity in TAU-supplemented mice because insulin plasma levels, in basal conditions (in fasted and fed mice) and during the ipGTT, were similar from those observed in CTL mice. TAU supplementation in our study increased fasting plasma glucagon levels, a hormone released during Figure 4. (A) GLUT 2, (B) GCK, (C) β2 subunit of volt- 2+ hypoglycemia that induces hepatic glucose output age-dependent Ca channel (Cavβ2)andβ-actin protein expressions. ∗P < 0.05 indicates significant difference vs CTL [38,39]. Thus, by enhancing glucagon levels, TAU exerts a possible dual effect on glucose homeostasis; firstly in the fed state, by increasing glucose uptake, utilization, (Figure 4(C)). β-actin protein expression, measured in and/or storage, and secondly under fasting conditions, parallel, did not differ. where it increases hepatic glucose mobilization, avoiding undesirable hypoglycemia. High glucose tolerance in supplemented mice may Discussion account for a possible alteration in glucose-induced insulin secretion by pancreatic islets. Despite the fact that The current work highlights the importance of TAU supplementation did not alter insulin plasma levels, TAU supplementation for pancreatic islet functional- in vitro insulin secretion showed that supplementation ity. In agreement with other experimental models improved β-cell sensitivity to fuel secretagogues.

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr Taurine Supplementation Increases Insulin Secretion 377

In dynamic studies, the islets from TAU-supplemented voltage-dependent channels, in parallel with an increase mice showed an increase in the first phase and in the in Ca2+ handling by the islets, in accordance with recent total insulin secretion in response to glucose (Figure 2). observations [25]. These alterations are not due to modifications in GLUT 2 Furthermore, these data also provide new information and GCK expression, glucose oxidation and ATP content, regarding TAU effects upon islet hormone secretion. nor the ATP-to-ADP ratio; however, they may be due We observed that, at low glucose concentration, TAU- to high Ca2+ uptake induced by glucose. This enhanced supplemented islets secreted more glucagon than CTL. 2+ Ca influx probably results from the high β2 subunit It is known that the control of glucagon secretion is expression of the voltage-sensitive Ca2+ channel protein multifactorial and involves direct effects of nutrients, in the plasma membrane of the TAU islets. Among the hormones and neurotransmitters. In α cells, the closure of 2+ several proteins and messengers that regulated Ca KATP-channels, due to an increased ATP-to-ADP ratio, handling, the β subunit plays multiple roles in L-type leads to membrane depolarization and activation of channels activity [40]. voltage-dependent Na+ and N-type Ca2+ channels that In β-cell, Ca2+ entry through the voltage-sensitive Ca2+ contribute to the generation of action potentials, increased channel directly stimulates secretory granule trafficking [Ca2+]i,andfinallyglucagonsecretion[39].Inadditionto and triggers insulin exocytosis [41]. In L-type Ca2+ the glucagon action in hepatic glucose output, a possible channel the α1 subunit forms the pore, whereas β subunit paracrine effect of this hormone in β-cell increasing is entirely cytosolic. β subunits are believed to play insulin release is suggested [47,48]. This hypothesis is a key role in the assembly/expression of the channel supported by studies that demonstrate the presence of complex and in the modulation of Ca2+ currents [40,41]. the glucagon receptor in β-cells [49,50], and a decreased β2 and β3 subunits are expressed in RINm5F cells as well glucose-stimulated insulin secretion in human islets in as in rodent pancreatic islets [42–45]. Although the β the presence of a glucagon receptor antagonist [51]. subunit function in β-cells is not completely understood, Thus, the observations that TAU-supplemented islets some evidences support their importance for channel release more glucagon at low glucose and that these activity and insulin secretion. Islets from Otsuka Long- islets secreted more insulin when incubated at high Evans Tokushima Fatty diabetic rats showed decreased glucose in the presence of glucagon suggests that TAU insulin secretion in response to glucose together with supplementation improves whole endocrine hormone 2+ a significant reduction in Ca channel activity and β2 sensing to nutrients and possible glucagon paracrine and β3 subunits gene expression [44]. When α1 subunits effects on insulin secretion. from RINm5F cells was transfected to Chinese hamster In β-cells, glucagon acts via a G protein-coupled 2+ ovary cells it only showed Ca current when β2 subunits receptor, increasing intracellular cAMP levels that were co-expressed, suggesting that these subunits are culminate with protein kinase A (PKA) activation [49]. important to voltage-sensitive Ca2+ channel activity in β- PKA contributes to enhance insulin secretion activating cells [46]. β subunits can also act as non-channel proteins the exocytotic machinery and Ca2+ flux [52–54]. It is 2+ in β-cell. This is based on the fact that β-cells lacking β3 known that PKA phosphorylates Ca channel subunits subunits showed enhanced [Ca2+]i and insulin secretion and may increase open channel probability [55,56]. in the presence of glucose due to high inositol-1,4,5- As such, a combination between increased glucagon triphosphate production and Ca2+ mobilization from paracrine action in the β-cell, PKA activation, and intracellular stores [45]. Thus, the increased insulin increased Ca2+ channel protein expression may contribute secretion in TAU-supplemented islets could be due to a to enhance Ca2+ uptake and insulin secretion in the better Ca2+ handling in combination with elevated insulin presence of high glucose in TAU-supplemented islets. content in these islets. L-leucine is the most potent amino acid secretagogue The relationship between TAU and Ca2+ regulation has of insulin and this is due to its ability to activate been documented. In many tissues, TAU markedly affects GDH, associated with the generation of ATP by its own the kinetics of Ca2+ movement across both cellular and metabolism [33]. We observed that TAU islets secreted subcellular membranes. In cardiac cells, TAU is an agonist more insulin than CTL when stimulated by L-leucine, of the Ca2+ current and delayed rectified K+ current at low and that this alteration is probably due to high L- [Ca2+]i, whereas at high [Ca2+]i, TAU inhibits both types leucine catabolism in the TAU group. This hypothesis of currents resulting in the prolongation of the duration of is supported by the fact that insulin secretion was not the action potential [5]. In isolated rat liver mitochondria, different between groups when carried out in the presence TAU-enhanced Ca2+ uptake via the uniport system [6]. In of a non-metabolizable analog of L-leucine, BCH [57], or rat pancreatic β-cells, TAU increased insulin secretion and with the combination of BCH with L-glutamine. This 2+ Ca oscillation via inhibition of the KATP channels due is also supported by static and dynamic studies when to its interaction with a benzamido-binding site on SUR1 islets were perifused or incubated with L-glutamine plus [8]. TAU also restores impaired glucose-induced insulin L-leucine with or without BCH, a condition in which secretion in uncoupling protein 2-overexpressing β-cells L-leucine-induced insulin release occurs predominantly by inducing Ca2+ sequestration into the mitochondrial via GDH alosteric activity, since L-glutamine decreases matrix [9]. Our work is the first to show that this amino L-leucine oxidation [58]. These results together with 2+ acid increases the expression of the β2 subunit of the Ca the similar insulin release between groups in the

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr 378 R. A. Ribeiro et al. presence of the stimulus: L-leucine/AOA or with the 5. Satoh H. Cardiac actions of taurine as a modulator of the ion combination of BCH/L-glutamine/L-leucine plus AOA, channels. Adv Exp Med Biol 1998; 442: 121–128. 6. Palmi M, Youmbi GT, Fusi F, et al. Potentiation of mitochondrial when L-leucine metabolism was abolished, indicate that Ca2+ sequestration by taurine. Biochem Pharmacol 1999; 58: TAU supplementation may enhance L-leucine catabolism 1123–1131. in islets. In this way, our data suggest that the 7. Foos TM, Wu JY. The role of taurine in the central nervous system and the modulation of intracellular calcium homeostasis. production of α-ketoisocaproate (KIC) from L-leucine Neurochem Res 2002; 27: 21–26. was enhanced in TAU-supplemented islets. In addition, 8. Park EJ, Bae JH, Kim SY, et al. Inhibition of ATP-sensitive K+ channels by taurine through a benzamido-binding site the higher L-glutamine levels present in pancreatic on sulfonylurea receptor 1. Biochem Pharmacol 2004; 67: islets raises L-glutamate content and α-ketoglutarate 1089–1096. production by GDH and/or aspartate aminotransferase 9. Lee SH, Lee HY, Kim SY, Lee IK, Song DK. Enhancing effect of taurine on glucose response in UCP2-overexpressing beta cells. (AAT), pathways. Under elevated acetyl-CoA content Diabetes Res Clin Pract 2004; 66(Suppl 1): S69–S74. (from KIC production) in TAU-supplemented islets, α- 10. Lim JG, Lee HY, Yun JE, et al. Taurine block of cloned ATP- + ketoglutarate might favour the tricarboxylic acid cycle sensitive K channels with different sulfonylurea receptor subunits expressed in Xenopus laevis oocytes. Biochem expansion via oxaloacetate production, and consequently, Pharmacol 2004; 68: 901–910. mitochondrial energy production for β-cell depolarization 11. Huxtable RJ. Physiological actions of taurine. Physiol Rev 1992; and sequential events such Ca2+ influx that directly 72: 101–163. 12.TasS,SarandolE,AyvalikSZ,SerdarZ,DiricanM.Vanadyl stimulates insulin secretion. sulfate, taurine, and combined vanadyl sulfate and taurine In conclusion, we demonstrate that increasing plasma treatments in diabetic rats: effects on the oxidative and TAU levels by supplementation modifies β-cell sensitivity antioxidative systems. Arch Med Res 2007; 38: 276–283. 13. Moran J, Salazar P, Pasantes-Morales H. Effect of tocopherol to glucose and L-leucine. In glucose-induced insulin and taurine on membrane fluidity of retinal rod outer segments. + secretion, TAU supplementation may alter Ca2 handling. Exp Eye Res 1987; 45: 769–776. However, the increase in the islet catabolism of L-leucine 14. Kulakowski EC, Maturo J. Hypoglycemic properties of taurine: not mediated by enhanced insulin release. Biochem Pharmacol amino acid may be the main alteration provoked by TAU 1984; 33: 2835–2838. regarding L-leucine-induced release. In parallel, glucagon 15. Maturo J, Kulakowski EC. Taurine binding to the purified insulin release was increased in TAU-supplemented islets and receptor. Biochem Pharmacol 1988; 37: 3755–3760. 16. Kaplan B, Karabay G, Zagyapan RD, et al. Effects of taurine in may also account for the increased insulin secretion from glucose and taurine administration. Amino Acids 2004; 27: β-cells. These data support the presumable role of TAU 327–333. in improving β-cell function in type 2 diabetes and pre- 17. Cherif H, Reusens B, Ahn MT, Hoet JJ, Remacle C. Effects of taurine on the insulin secretion of rat fetal islets from dams fed diabetic states that present a decreased ability to release alow-proteindiet.J Endocrinol 1998; 159: 341–348. insulin in response to glucose and L-leucine. 18. Nandhini AT, Thirunavukkarasu V, Ravichandran MK, Anu- radha CV. Effect of taurine on biomarkers of oxidative stress in tissues of fructose-fed insulin-resistant rats. Singapore Med J 2005; 46: 82–87. 19. Tsuboyama-Kasaoka N, Shozawa C, Sano K, et al.Taurine(2- Acknowledgements aminoethanesulfonic acid) deficiency creates a vicious circle promoting obesity. Endocrinology 2006; 147: 3276–3284. This study was supported by grants from the Funda¸cao˜ de 20. Arany E, Strutt B, Romanus P, et al. Taurine supplement in early life altered islet morphology, decreased insulitis and delayed the Amparo aPesquisadoEstadodeS` ao˜ Paulo (FAPESP; 05/59707- onset of diabetes in non-obese diabetic mice. Diabetologia 2004; 0), Conselho Nacional para o Desenvolvimento Cient´ıfico e 47: 1831–1837. Tecnologico´ (CNPq) and Coordena¸cao˜ de Aperfei¸coamento de 21. Boujendar S, Reusens B, Merezak S, et al. Taurine supplemen- Pessoal de N´ıvel Superior (CAPES). We are grateful to Léscio tation to a low protein diet during foetal and early postnatal life restores a normal proliferation and apoptosis of rat pancreatic D. Teixeira for animal care and technical assistance, and Nicola islets. Diabetologia 2002; 45: 856–866. Conran for editing the English. 22.KaniukNA,KiralyM,BatesH,et al. Ubiquitinated-protein aggregates form in pancreatic beta-cells during diabetes-induced oxidative stress and are regulated by autophagy. Diabetes 2007; 56: 930–939. Conflict of interest 23. Xiao C, Giacca A, Lewis GF. Oral taurine but not N-acetylcysteine ameliorates NEFA-induced impairment in insulin sensitivity and The authors have no conflicts of interest. beta cell function in obese and overweight, non-diabetic men. Diabetologia 2008; 51: 139–146. 24. Cherif H, Reusens B, Dahri S, Remacle C, Hoet JJ. Stimulatory effects of taurine on insulin secretion by fetal rat islets cultured References in vitro. J Endocrinol 1996; 151: 501–506. 25. Carneiro EM, Latorraca MQ, Araujo E, et al.Taurine supplementation modulates glucose homeostasis and islet 1. Straub SG, Sharp GW. Glucose-stimulated signaling pathways function. JNutrBiochem2008, in press. in biphasic insulin secretion. Diabetes Metab Res Rev 2002; 18: 26. Bonora E, Manicardi V, Zavaroni I, Coscelli C, Butturini U. 451–463. Relationships between insulin secretion, insulin metabolism and 2. Newsholme P, Bender K, Kiely A, Brennan L. Amino acid insulin resistance in mild glucose intolerance. Diabete Metab metabolism, insulin secretion and diabetes. Biochem Soc Trans 1987; 13: 116–121. 2007; 35: 1180–1186. 27. Bordin S, Boschero AC, Carneiro EM, Atwater I. Ionic 3. Bustamante J, Lobo MV, Alonso FJ, et al. An osmotic-sensitive mechanisms involved in the regulation of insulin secretion by taurine pool is localized in rat pancreatic islet cells containing muscarinic agonists. JMembrBiol1995; 148: 177–184. glucagon and somatostatin. Am J Physiol Endocrinol Metab 2001; 28. Hoy M, Bokvist K, Xiao-Gang W, et al. Phentolamine inhibits 281: E1275–E1285. exocytosis of glucagon by Gi2 protein-dependent activation of 4. Schaffer S, Takahashi K, Azuma J. Role of osmoregulation in the calcineurin in rat pancreatic alpha cells. JBiolChem2001; 276: actions of taurine. Amino Acids 2000; 19: 527–546. 924–930.

Copyright  2009 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2009; 25: 370–379. DOI: 10.1002/dmrr Taurine Supplementation Increases Insulin Secretion 379

29. Bradford MM. A rapid and sensitive method for the quantitation 44. Iwashima Y, Abiko A, Ushikubi F, et al.Downregulationof of microgram quantities of protein utilizing the principle of the voltage-dependent calcium channel (VDCC) beta-subunit protein-dye binding. Anal Biochem 1976; 72: 248–254. mRNAs in pancreatic islets of type 2 diabetic rats. Biochem 30. Kawamoto Y, Shinozuka K, Kunitomo M, Haginaka J. Determi- Biophys Res Commun 2001; 280: 923–932. nation of ATP and its metabolites released from rat caudal 45. Berggren PO, Yang SN, Murakami M, et al. Removal of Ca2+ artery by isocratic ion-pair reversed-phase high-performance channel beta3 subunit enhances Ca2+ oscillation frequency and liquid chromatography. Anal Biochem 1998; 262: 33–38. insulin exocytosis. Cell 2004; 119: 273–284. 31. Katayama M, Matsuda Y, Shimokawa K, et al.Simultaneous 46. Ihara Y, Yamada Y, Fujii Y, et al.Moleculardiversityand determination of six adenyl purines in human plasma by functional characterization of voltage-dependent calcium high-performance liquid chromatography with fluorescence channels (CACN4) expressed in pancreatic beta-cells. Mol derivatization. J Chromatogr B Biomed Sci Appl 2001; 760: Endocrinol 1995; 9: 121–130. 159–163. 47. Kawai K, Yokota C, Ohashi S, Watanabe Y, Yamashita K. 32. Amaral ME, Ueno M, Carvalheira JB, et al. Prolactin-signal Evidence that glucagon stimulates insulin secretion through transduction in neonatal rat pancreatic islets and interaction its own receptor in rats. Diabetologia 1995; 38: 274–276. with the insulin-signaling pathway. Horm Metab Res 2003; 35: 48. Cabrera O, Berman DM, Kenyon NS, et al. The unique 282–289. cytoarchitecture of human pancreatic islets has implications 33. Sener A, Malaisse WJ. L-leucine and a nonmetabolized analogue for islet cell function. Proc Natl Acad Sci U S A 2006; 103: activate pancreatic islet glutamate dehydrogenase. Nature 1980; 2334–2339. 288: 187–189. 49. Moens K, Heimberg H, Flamez D, et al. Expression and 34. Loizzo A, Carta S, Bennardini F, et al. Neonatal taurine functional activity of glucagon, glucagon-like peptide I, and administration modifies metabolic programming in male mice. glucose-dependent insulinotropic peptide receptors in rat Early Hum Dev 2007; 83: 693–696. pancreatic islet cells. Diabetes 1996; 45: 257–261. 35. Anuradha CV, Balakrishnan SD. Taurine attenuates hyperten- 50. Moens K, Flamez D, Van Schravendijk C, et al.Dualglucagon sion and improves insulin sensitivity in the fructose-fed rat, an recognition by pancreatic beta-cells via glucagon and glucagon- animal model of insulin resistance. Can J Physiol Pharmacol like peptide 1 receptors. Diabetes 1998; 47: 66–72. 1999; 77: 749–754. 51. Huypens P, Ling Z, Pipeleers D, Schuit F. Glucagon receptors on 36. Nakaya Y, Minami A, Harada N, et al. Taurine improves insulin human islet cells contribute to glucose competence of insulin sensitivity in the Otsuka Long-Evans Tokushima Fatty rat, a release. Diabetologia 2000; 43: 1012–1019. model of spontaneous type 2 diabetes. Am J Clin Nutr 2000; 71: 52. Wan QF, Dong Y, Yang H, et al. Protein kinase activation 54–58. increases insulin secretion by sensitizing the secretory machinery 37. Nandhini AT, Anuradha CV. Taurine modulates kallikrein to Ca2+. JGenPhysiol2004; 124: 653–662. activity and glucose metabolism in insulin resistant rats. Amino 53. Hatakeyama H, Kishimoto T, Nemoto T, Kasai H, Takahashi N. Acids 2002; 22: 27–38. Rapid glucose sensing by protein kinase A for insulin exocytosis 38. Barg S. Mechanisms of exocytosis in insulin-secreting B-cells and in mouse pancreatic islets. JPhysiol2006; 570: 271–282. glucagon-secreting A-cells. Pharmacol Toxicol 2003; 92: 3–13. 54. Dyachok O, Gylfe E. Ca2+-induced Ca2+ release via inositol 39. Gromada J, Franklin I, Wollheim CB. Alpha-cells of the 1,4,5-trisphosphate receptors is amplified by protein kinase A endocrine pancreas: 35 years of research but the enigma and triggers exocytosis in pancreatic beta-cells. JBiolChem remains. Endocr Rev 2007; 28: 84–116. 2004; 279: 45 455–45 461. 40. Chien AJ, Zhao X, Shirokov RE, et al. Roles of a membrane- 55. Gao T, Yatani A, Dell’Acqua ML, et al.cAMP-dependent localized beta subunit in the formation and targeting of regulation of cardiac L-type Ca2+ channels requires membrane functional L-type Ca2+ channels. JBiolChem1995; 270: targeting of PKA and phosphorylation of channel subunits. 30 036–30 044. Neuron 1997; 19: 185–196. 41. Yang SN, Berggren PO. The role of voltage-gated calcium 56. Felix R. Molecular regulation of voltage-gated Ca2+ channels. channels in pancreatic beta-cell physiology and pathophysiology. Journal of Receptors and Signal Transduction 2005; 25: 57–71. Endocr Rev 2006; 27: 621–676. 57. Liu YJ, Cheng H, Drought H, et al. Activation of the KATP 42. Iwashima Y, Pugh W, Depaoli AM, et al. Expression of calcium channel-independent signaling pathway by the nonhydrolyzable channel mRNAs in rat pancreatic islets and downregulation after analog of leucine, BCH. Am J Physiol Endocrinol Metab 2003; glucose infusion. Diabetes 1993; 42: 948–955. 285: E380–E389. 43. Safayhi H, Haase H, Kramer U, et al. L-type calcium channels 58. Macdonald MJ, Hasan NM, Longacre MJ. Studies with leucine, in insulin-secreting cells: biochemical characterization and beta-hydroxybutyrate and ATP citrate lyase-deficient beta cells phosphorylation in RINm5F cells. Mol Endocrinol 1997; 11: support the acetoacetate pathway of insulin secretion. Biochim 619–629. Biophys Acta 2008; 1780: 966–972.