Melatonin inhibits insulin secretion and decreases PKA levels without interfering with glucose metabolism in rat pancreatic islets
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
J. Pineal Res. 2002; 33:156–160 Copyright Blackwell Munksgaard, 2002
Journal of Pineal Research
ISSN 0742-3098
Melatonin inhibits insulin secretion and decreases PKA levels
without interfering with glucose metabolism in rat pancreatic islets
Abstract: The effect of melatonin (0.1 lM) on freshly isolated islets from Maria Cecı́lia Picinato, Esther
adult rats was investigated. Melatonin caused a marked decrease of insulin Piltcher Haber, José Cipolla-Neto,
secretion by islets in response to glucose. The mechanism involved was then Rui Curi, Carla Roberta de
examined. Melatonin did not interfere with glucose metabolism as indicated Oliveira Carvalho and Angelo
by the measurement of glucose oxidation. However, the content of the Rafael Carpinelli
protein kinase A (PKA) catalytic a-subunit was significantly decreased in Department of Physiology and Biophysics,
Institute of Biomedical Sciences, University of
islets exposed to melatonin for 1 hr in the presence of 8.3 mM glucose,
São Paulo, São Paulo, Brasil
whereas that of the protein kinase C (PKC) a-subunit remained unchanged.
Melatonin also inhibited forskolin-induced insulin secretion, a well known Key words: glucose metabolism, Islets of
activator of adenylate cyclase (AC) activity. This may explain the low Langerhans, melatonin, pineal, PKA, PKC
content of insulin found in islets incubated in the presence of melatonin for Address reprint requests to Angelo R.
Carpinelli, Departamento de Fisiologia e
3 hr. In fact, 3¢,5¢ -cyclic adenosine monophosphate (cAMP), a product of
Biofı́sica, Instituto de Ciências Biomédicas,
AC activity, stimulates insulin synthesis. These findings led us to postulate Universidade de São Paulo, Avenue Prof
that a down-regulation of the PKA signaling pathway may be the mechanism Lineu Prestes 1524, 05508.900 São Paulo,
Brazil.
involved in the melatonin inhibition of the process of glucose-induced insulin
E-mail: angelo@fisio.icb.usp.br
secretion.
Received January 25, 2002;
accepted April 1, 2002.
of Peschke’s group [7] was performed in islets isolated from
Introduction
the pancreas of neonate rats and after an overnight period
Blood glucose concentration is the most important regula- in culture. The authors did not investigate if melatonin
tory stimulus for B-cell insulin secretion. The process of affects glucose metabolism of the pancreatic islets. Also,
glucose-induced insulin secretion is fully dependent on the effect of melatonin on insulin secretion by islets freshly
metabolism of the sugar in the B-cell. The products of the obtained from adult rats remains to be investigated. These
glucose metabolism, especially adenosine triphosphate points were addressed in the present study.
(ATP), close the ATP potassium channel (KATP), increasing For this purpose, pancreatic islets were isolated from
the intracellular potassium concentration that leads to a adult rats and the effect of melatonin (0.1 lM) was
membrane depolarization [1]. This phenomenon induces the immediately tested for its effect on glucose- and forskolin-
opening of the voltage sensitive calcium channels (VSCC) induced insulin secretion, glucose oxidation, and the
provoking a rapid increase in calcium influx and in the content of PKA and PKC a-subunits.
intracellular calcium concentration [Ca]i that triggers gran-
ule exocytosis [2, 3]. On the other hand, the metabolism of
glucose in the B-cell also activates several enzymes, partic- Materials and methods
ularly phospholipase C (PLC) and adenylate cyclase (AC),
Animals
which induce an increase in protein kinase C (PKC) and
protein kinase A (PKA) activities, respectively [4, 5]. These Male albino rats weighing 150–200 g (45–60 days old) were
kinases have an important role in the mechanism of obtained from the Institute of Biomedical Sciences, Uni-
glucose-induced insulin secretion, including opening of the versity of São Paulo, São Paulo, Brazil. The animals were
VSCC [6]. kept in groups of five at 23C in a room with a light/dark
The insulin secretion, however, is complex and several cycle of 12/12 hr (lights on at 07:00 hr). Ethics approval
other stimuli interfere in the process of glucose-induced was granted for this study by the Institute of Biomedical
insulin release, including autonomic nervous system effects, Sciences, Animal Experimental Committee, University of
circulating metabolites and hormones. Recently, in a series São Paulo, São Paulo, Brazil.
of perfusion experiments with isolated islets, Peschke et al.
[7, 8] showed that melatonin inhibits glucose- and forskolin-
Chemicals
induced insulin secretion. The data indicated that melato-
nin probably acts on cultivated neonate pancreatic islets by Collagenase type V, bovine albumin-fraction V, Tris,
inhibiting AC activity through the pertussis-toxin sensitive phenylmethylsulfonylfluoride (PMSF), aprotinin, dith-
G-protein system via Mel1a receptors [8]. The elegant study iothreitol (DTT) and melatonin were purchased from
156Melatonin and insulin secretion
Sigma Chemical Co. (St Louis, MO, USA). [U-14C]- during the experiment. The 14CO2 produced was exposed
Glucose and 125I-insulin were obtained from Dupont for 2 hr to NaOH solution (1 N) injected outside the tube
NEN products (Boston, MA, USA). Rat insulin standards containing the islets, in the same manner as for HCl.
and antirat insulin antibody were a gift from Dr Leclercq- Biodegradable scintillation liquid (Amersham, Pharmacia,
Mayer, Université Libre de Bruxelles, Belgium. The Upsala, Sweden) was added and radioactivity of the 14CO2
reagents and apparatus for SDS–PAGE and immunoblot- adsorbed by NaOH was counted.
ting were from Bio-Rad (Richmond, CA, USA). Protein A
sepharose 6MB was from Pharmacia (Upsala, Sweden).
Insulin content of the islets
Polyclonal anti-PKA and anti-PKC antibodies were
obtained from Santa Cruz Biotechnology (Santa Cruz, Batches of 30 islets were incubated for 3 hr at 37C in the
CA, USA). KH buffer containing 8.3 mM glucose in the presence or
absence of melatonin 0.1 lM. After incubation, the islets
were washed three times with KH buffer, and resuspended
Insulin secretion during incubation of islets
in 1 mL KH buffer for sonication. Samples of the experi-
Rat pancreatic islets were isolated as described by Lacy and ments were frozen and stored at –20C for insulin deter-
Kostianovsky [9]. Incubations of five islets were carried out mination.
at 37C for 60 min in 0.5 mL of Krebs-Henseleit (KH)
buffer (Na+ 139 mM, K+ 5 mM, Ca2+ 1 mM, Mg2+ 1 mM,
Determination of the content of PKA and PKC
Cl– 124 mM, HCO3– 24 mM) and 0.2% albumin in the
catalytic a-subunit by Western blotting
presence of glucose(5.6; 8.3 or 16.7 mM). The buffer was
equilibrated in a mixture of O2 (95%) and CO2 (5%). When Batches of 100 islets were incubated for 1 hr at 37C in
indicated, melatonin was added to the medium at 0.1 lM the KH buffer containing 8.3 mM glucose in the presence
always. At the end of the experiments, the medium was or absence of melatonin 0.1 lM. After incubation, 800 lL
collected for insulin assay. The effect of forskolin, an of ice-cold buffer A (100 mM Tris, 50 mM Hepes, pH 7.4,
activator of AC, was tested in the presence of 5.6 mM 100 mM sodium pyrophosphate, 10 mM sodium vanadate,
glucose, by adding 1.0, 5.0 or 7.0 lM of the compound to 100 mM sodium fluoride, 1% triton-X-100, 2 mM PMSF
the medium containing or not melatonin. and 0.1 mg/mL aprotinin) were transferred to Eppendorf
tubes containing the islets, sonicated for cell disruption
and centrifuged at 20,000 · g at 4C for 15 min The
Perfusion experiments
protein of the supernatant was measured [11] and used for
The method used was slightly modified from our previous immunoprecipitation with 10 lL of anti PKA catalytic
publications [10]. Islets (n ¼ 100) were preincubated for a-subunit or 10 lL of anti-PKC catalytic a-subunit
60 min at 37C in the presence of 16.7 mM glucose and antibodies and protein A Sepharose 6MB or used as
0.1 lM melatonin. After this period, the islets were washed total extracts followed by 8% SDS–PAGE and immuno-
five times with KH albuminated buffer and perfused in a blot analysis [12].
chamber containing a special filter (cellulose acetate-
SCWP100, 8 lm pore size). The perfusion was performed
Statistical analysis
with KH albuminated buffer at a constant rate of 1 mL/min
(Fig. 2). The perfusion medium was collected at 1-min Results of perfusion experiments, insulin secretion and CO2
intervals from the 31st until to the 70th minute using a produced are presented as means ± S.E.M. The statistical
fraction collector (ULTRORAC II, LKB, BROMA 2070). treatment was carried out using ANOVA. The mean values
The composition of the perfusion medium was changed as of incremental integrated areas in perfusion experiments
follows: 5.6 mM glucose only from the 31st to the 40th min, was calculated by subtraction of each point after the 40th
16.7 mM glucose without (control) or with 0.1 lM melato- minute, until to the end of the experiment, from the mean
nin (Mel) from the 41st to 60th minutes and again 5.6 mM of the basal values (from the 35th to 40th min). For analysis
glucose only form the 61st to the 70th min. Aliquots of all of the immunoblotting data, Student’s unpaired t-test was
experiments were frozen and stored at –20C for insulin used. The level of significance was set for P < 0.05.
assay.
Results
Measurement of [U-14C]-glucose decarboxylation
Progressive increments of glucose concentrations (5.6, 8.3
Groups of 30 islets were incubated in glass tubes containing and 16.7 mM) proportionally raised the amount of insulin
100 lL of KH albuminated buffer solution, 2 lCi/mL released during 60 min incubation by control islets as
[U-14C]-glucose and 5.6, 8.3 or 16.7 mM glucose in the expected (Fig. 1). Melatonin caused a marked inhibition of
presence and absence of 0.1 lM melatonin, equilibrated insulin secretion in the presence of all glucose concentra-
against a mixture of O2 (95%) and CO2 (5%). The tubes tions. The indole also inhibited insulin secretion by
were transferred into counting vials that were carefully pancreatic islets in the perfusion experiments (Fig. 2).
closed with a rubber stopper and incubated for 120 min in a Under these conditions, insulin release by 16.7 mM glucose
shaker bath at 37C. After this period, 0.1 mL of HCl stimulus was abolished by melatonin (Fig. 2).
(0.5 N) was added, by syringe, to the tubes containing the The total content of insulin was determined in pancreatic
islets to stop the reaction, without loss of the gas produced islets incubated for 3 hr in 8.3 mM glucose in the presence
157Picinato et al.
Fig. 1. Insulin secretion (lU.islet)1.60 min)1) by islets isolated from
adult rats and incubated for 60 min in the presence of 5.6, 8.3 or
16.7 mM glucose in the presence (Mel) or absence (Control) of 0.1 lM Fig. 3. Glucose decarboxylation (pmol.islet)1.120 min)1) by islets
melatonin. The values are presented as means ± S.E.M. of at least isolated from adult rats and incubated in the presence of 5.6, 8.3 or
eight experiments. For details of the experiments see Material and 16.7 mM glucose in the presence (Mel) or absence (Control) of
Methods section. *P < 0.05 Control versus Mel islets. melatonin (0.1 lM). The values are presented as the mean ±
S.E.M. of at least six experiments. For details of the experiments
see Material and Methods section.
Fig. 2. Effect of melatonin on insulin secretion (lU.islet)1.min)1)
Fig. 4. Insulin secretion (lU.islet)1.60 min)1) by islets isolated
by perfused islets. The islets were preincubated in the presence of
from adult rats and incubated for 60 min in the presence of 5.6 mM
16.7 mM glucose and 0.1 lM melatonin. During the first 10 min
glucose in the absence or presence of 0.1 lM melatonin and 1.0, 5.0
(from 31st to 40th min), the two batches of islets were perfused in
or 7.0 lM forskolin. The values are presented as means ± S.E.M.
the presence of 5.6 mM glucose only. Between the 41st to 60th min,
of at least eight experiments. For details of the experiments see
the islets were perfused in the presence of 16.7 mM glucose without
Material and Methods section. *P < 0.05 melatonin versus non-
(Control) or with (Mel) melatonin. Afterwards, the perfusion of the
melatonin treated islets.
two batches was carried out again in the presence of 5.6 mM glu-
cose only. The values are presented as mean ± S.E.M. of eight
experiments. For details of the experiments see Material and
Methods section. The level of of PKA a-subunit was determined and it was
significantly decreased in islets exposed for 60 min to
8.3 mM glucose plus melatonin (Fig. 5). A similar protocol
or absence of melatonin (0.1 lM). The indole caused a was used to investigate PKC a-subunit levels. In contrast to
significant reduction of insulin content (from 9634 ± 258 what was observed for PKA, the level of PKC a-subunit
to 6665 ± 771 lU per islet; mean ± S.E.M. of 10 sam- was not changed in islets incubated with melatonin (Fig. 5).
ples).
The rates of [U-14C] – glucose decarboxylation were also
Discussion
measured in islets incubated for 120 min in the presence of
increasing glucose concentrations (5.6, 8.3 and 16.7 mM). The carbohydrate metabolism is influenced by melatonin
Melatonin did not cause significant change of [U-14C]- both in humans [13–16] and in rodents [17–19]. Pinealec-
glucose decarboxylation (Fig. 3). This result indicates that tomy induces insulin resistance characterized by decreased
glucose metabolism in pancreatic islets is not affected by glucose tolerance and a reduced glycogenesis in liver and
melatonin and led us to investigate alterations in other muscle of rats [20–22]. Pinealectomy is followed by a
important pathways involved in the process of glucose- reduction in blood sugar levels, and after alloxan treatment,
induced insulin secretion. The effect of melatonin on the blood glucose rise is higher in pinealectomized than in
forskolin-induced insulin secretion was then examined intact rats [23]. Conversely, the infusion of pineal extracts
(Fig. 4). Forskolin per se caused a significant increase of [17] leads to hypoglycemia and increases glucose tolerance
insulin secretion. This effect of forskolin was abolished by and hepatic and muscule glycogenesis after glucose loading.
melatonin, especially at 5 and 7 lM concentrations. The changes in carbohydrate metabolism are, at least
158Melatonin and insulin secretion
evaluated the action of melatonin on [U-14C]-glucose
decarboxylation by incubated islets. Despite the remarkable
inhibitory effect of the indole on glucose induced-insulin
secretion, melatonin did not affect glucose metabolism in
incubated pancreatic islets (Fig. 3). This is a surprising
observation because the process of glucose-induced insulin
secretion is strongly associated with glucose metabolism
[24–26]. The generation of ATP from glucose oxidation
closes KATP channels which are triggered for the complex
process of insulin release in pancreatic islets [1].
Peschke et al. [8] identified Mel1a receptors in pancreatic
islets isolated from neonate rats and showed that melatonin
inhibits glucose- and also forskolin-induced insulin secre-
tion. Forskolin is a well known activator of AC and
increases the production of 3¢,5¢ -cyclic adenosine mono-
phosphate (cAMP) [27]. The mechanisms by which PKA
augments glucose stimulated-insulin release may involve
potentiation of glucose-induced Ca2+ influx by either
cAMP-dependent protein phosphorylation of the voltage-
sensitive Ca2+ channels or by sensitization of the secretory
mechanisms to Ca2+, increasing the efficiency of the
Fig. 5. PKA (A) and PKC (B) a-subunit levels in pancreatic rats
secretory pathway by reversible phosphorylation [28, 29].
incubated in the absence (Control) or presence (Mel) of 0.1 lM Peschke et al. [8] postulated that melatonin probably acts
melatonin. The proteins from the islets were extracted and proc- on overnight-cultivated neonate pancreatic islets by inhib-
essed as described in Materials and Methods. The immunoblotting iting the AC/cAMP-system. This possibility was investi-
was performed using anti-PKA and anti-PKC a-subunit antibod- gated in our experimental model. We incubated isolated
ies. The PKA and PKC protein levels are shown as means islets in the presence of 5.6 mM glucose, in the absence or
± S.E.M. of the scanning densitometric analysis of four distinct
presence of melatonin and melatonin plus forskolin
experiments. *P < 0.05.
(Fig. 4). Confirming Peschke’s observations, melatonin
inhibited forskolin-induced insulin secretion. To better
clarify this point we then carried out experiments to
partially, caused by the effect of melatonin on insulin levels determine the content of PKA a-subunit in islets incubated
[18]. Lima et al. [19] have shown that pinealectomized rats in the presence of melatonin. The content of PKA was
exhibit a moderate decrease of plasma glucose and an significantly decreased in islets exposed for 60 min to
increase in circulating insulin. Although melatonin clearly melatonin in the presence of 5.6 mM glucose with no
controls glucose homeostasis and insulin levels, the action alteration in the content of PKC a-subunit (Fig. 5). A
of this indole on glucose-induced insulin secretion is not decreased content of cAMP may explain the lower content
fully understood. of insulin observed in pancreatic islets after 3 hr incubation.
In the present study, melatonin caused a remarkable In fact, cAMP has been shown to stimulate insulin mRNA
decrease of glucose-induced insulin secretion (Fig. 1) transcription rate (30, 3129).
during a 60-min incubation in the presence of rising glucose The results presented herein led us to conclude that
concentrations (5.6, 8.3 and 16.7 mM). The inhibitory effect 0.1 lM melatonin acutely inhibits insulin secretion by fresh
of melatonin on insulin secretion became more pronounced islets isolated from adult rats as reported by Peschke’s
with higher glucose concentrations in the medium. To group for overnight cultivation of neonate islets. This effect
verify in a dynamic and more detailed way the inhibition is not associated with alteration in B-cell glucose metabo-
observed under static conditions, islets were preincubated lism but with a decrease of PKA content. It remains to be
in the presence of 16.7 mM glucose and with 0.1 lM examined, however, if other pathways in the process of
melatonin being perfused afterwards (Fig. 2). When mela- insulin release are also affected by melatonin.
tonin and 16.7 mM glucose were concomitantly added to
the medium, insulin secretion was abolished (from the 41st
to the 60th min perfusion) in comparison with the islets
Acknowledgments
exposed to 16.7 mM glucose only. These findings confirm This research has been supported by FAPESP, CAPES,
the previous observations of the Peschke’s group. These CNPq and PRONEX. The authors are grateful to the
authors showed that melatonin decreases insulin secretion constant technical assistance of Marlene S. da Rocha.
induced by pulses of glucose (50 mM) or KCl (100 mM).
Their experiments, however, were carried out in islets
isolated from neonate rats after being cultivated overnight.
References
Considering these results, the following question was 1. SMITH PA, ASHCROFT FM, RORSMAN P. Simultaneous
addressed: does melatonin interfere with glucose metabo- recordings of glucose dependent electrical activity and ATP-
lism? In the Peschke study [7, 8], glucose metabolism of the regulated K(+) -currents in isolated mouse pancreatic beta-
islets was not examined. To answer this question, we cells. FEBS Lett 1990; 261:187–190.
159Picinato et al.
2. NIKI I. Ca2+ signaling and the insulin secretory cascade in the 17. MILCU SM, MILCU I, NANU L. Le role de la glande pinéale
pancreatic beta-cell. Jpn J Pharmacol 1999; 80:191–197. dans le metabolisme des glicides. Ann Endocrinol (Paris) 1963;
3. SATIN LS. Localized calcium influx in pancreatic beta-cells. its 24:233.
significance for Ca2+ dependent insulin secretion from the 18. DIAZ B, BLAZQUEZ E. Effect of pinealectomy on plasma glu-
islets of Langerhans. Endocrine 2000; 13:251–262. cose, insulin and glucagon levels in the rat. Horm Metab Res
4. HUGHES SJ, ASHCROFT SJH. Cyclic AMP, protein phospho- 1986; 18:225–229.
rylation and insulin secretion. In: Nutrient Regulation of 19. LIMA FB, MACHADO UF, BARTOL I et al. Pinealectomy causes
Insulin Secretion. FLATT PR, ed. Portland Press, London, glucose intolerance and decreases adipose cell responsiveness
1992; pp. 271–288. to insulin in rats. Am J Physiol 1998; 275:E934–E941.
5. HOWELL SL, JONES PM, PERSAUD SJ. Regulation of insulin 20. LA FLEUR SE. A suprachiasmatic nucleus generated rhythm in
secretion: the role of second messengers. Diabetologia 1994; basal glucose concentrations. J Neuroendocrinol 1999; 11:643–
37:S30–S35. 652.
6. LEVITAN IB. Phosphorylation of ion channels. J Membr Biol 21. LA FLEUR SE. Polysynaptic neural pathways between the
1985; 87:177–190. hypothalamus, including the suprachiasmatic nucleus, and the
7. PESCHKE E, PESCHKE D, HAMMER T et al. Influence of mela- liver. Brain Res 2000; 871:50–56.
tonin and serotonin on glucose-stimulated release from 22. SHIMAZU T, OGASAWARA S. Effects of hypothalamic stimula-
perifused rat pancreatic islets in vitro. J Pineal Res 1997; tion on gluconeogenesis and glycolysis in rat liver. Am J Physiol
23:156–163. 1975; 228:1787–1793.
8. PESCHKE E, FAUTECK JD, MUBHOFF U et al. Evidence for a 23. CSABA G, BARATH P. Are Langerhan’s islets influenced by the
melatonin receptor within pancreatic islets of neonate rats: pineal body? Experientia 1971; 27:962.
functional, autoradiographic, and molecular investigations. 24. MALAISSE WJ, SENNER A, HERCHUELZ A et al. Insulin release:
J Pineal Res 2000; 28:156–164. the fuel hypothesis. Metabolism 1979; 28:373–386.
9. LACY PE, KOSTIANOVSKY M. Method for the isolation of the 25. MATSCHINSKY FM. A lesson in metabolic regulation inspired
intact islets of Langerhans from the rat pancreas. Diabetes by the glucokinase glucose sensor paradigm. Diabetes 1996;
1967; 16:35–39. 45:223–241.
10. PICINATO MC, CURI R, MACHADO UF et al. Soybean and olive 26. HENQUIN JC. Triggering and amplifying pathways of regula-
oils enriched diets increase insulin secretion to glucose stimulus tion of insulin secretion by glucose. Diabetes 2000; 49:1751–
in isolated pancreatic rat islets. Physiol Behav 1998; 2:289–294. 1760.
11. LOWRY OH, ROSEMBROUGH NJ, FARR AL et al. Protein 27. VANECEK J. Cellular mechanisms of melatonin action. Am
measurements with the Folin phenol reagent. J Biol Chem Physiol Soc 1998; 78:687–721.
1951; 193:265–275. 28. AMMALA C, ELIASSON L, BOKVIST K et al. Activation of protein
12. VELLOSO LA, CARNEIRO EM, CREPALDI SC et al. Glucose and kinases and inhibition of protein phosphatases play a central
insulin-induced phosphorylation of insulin receptor and its role in the regulation of exocytosis in mouse pancreatic b cells.
primary substrates IRS-1, IRS-2 in rat pancreatic islets. FEBS Proc Natl Acad Sci USA 1994; 91:4343–4347.
Lett 1995; 377:353–357. 29. LESTER LB, FAUX MC, NAUERT JB et al. Targeted protein
13. ORTEGA-CORONA BG, ESPARZA-AVALOS N, BENITEZ-KING G kinase A and PP-2B regulate insulin secretion through revers-
et al. Effect of melatonin administration on plasma glucose ible phosphorylation. Endocrinology 2001; 142:1218–1227.
levels. Proc West Pharmacol Soc 1991; 34:75–77. 30. HUSSAIN MA, DANIEL PB, HABENER JF. Glucagon stimulates
14. VAN CAUTER E, BLACKMAN JD, ROLAND D et al. Modulation expression of the inducible cAMP early repressor and
of glucose regulation and insulin secretion by circadian suppresses insulin gene expression in pancreatic beta-cells.
rhythmicity and sleep. J Clin Invest 1991; 88:934–942. Diabetes 2000; 49:1681–1690.
15. VAN CAUTER E, SHAPIRO ET, THILIL H et al. Circadian modu- 31. SKOGLUND G, HUSSAIN MA, HOLZ GG. Glucagon-like peptide
lation of glucose and insulin responses to meals: relationship to 1 stimulates insulin gene promoter activity by protein kinase
cortisol rhythm. Am J Physiol 1992; 262:E467–E475. A-independent activation of the rat insulin I gene cAMP
16. VAN CAUTER E. Putative roles of melatonin in glucose regu- response element. Diabetes 2000; 49:1156–1164.
lation. Therapie 1998; 53:467–472.
160You can also read
