Abstract
We investigated whether an increase in cAMP could normalize glucose-stimulated insulin secretion (GSIS) in uncoupling protein-2 (UCP2) overexpressing (ucp2-OE) β-cells. Indices of β-cell (β-TC-6f7 cells and rodent islets) function were measured after induction of ucp2, in the presence or absence of cAMP-stimulating agents, analogs, or inhibitors. Islets of ob/ob mice had improved glucose-responsiveness in the presence of forskolin. Rat islets overexpressing ucp2 had significantly lower GSIS than controls. Acutely, the protein kinase A (PKA) and epac pathway stimulant forskolin normalized insulin secretion in ucp2-OE rat islets and β-TC-6f7 β-cells, an effect blocked by specific PKA inhibitors but not mimicked by epac agonists. However, there was no effect of ucp2-OE on cAMP concentrations or PKA activity. In ucp2-OE islets, forskolin inhibited ATP-dependent potassium (KATP) channel currents and 86Rb+ efflux, indicative of KATP block. Likewise, forskolin application increased intracellular Ca2+, which could account for its stimulatory effects on insulin secretion. Chronic exposure to forskolin increased ucp2 mRNA and exaggerated basal secretion but not GSIS. In mice deficient in UCP2, there was no augmentation of either cAMP content or cAMP-dependent insulin secretion. Thus, elevating cellular cAMP can reverse the deficiency in GSIS invoked by ucp2-OE, at least partly through PKA-mediated effects on the KATP channel.
Introduction
Previously, uncoupling protein-2 (UCP2) was identified in islets and shown to be a negative regulator of insulin secretion (Chan et al. 1999, 2001, Hong et al. 2001, Zhang et al. 2001). Ucp2 mRNA and protein expressions are increased in obese, diabetic rodents (Zhang et al. 2001) and may contribute to the development of insulin insufficiency. In addition, genetic studies in humans show an association between carriers of −866A/A allele, reduced glucose-stimulated insulin secretion (GSIS) and type-2 diabetes (D’Adamo et al. 2004, Sasahara et al. 2004). A potential strategy for improving insulin secretion in diabetic patients is to directly lower UCP2 activity. However, another strategy might be to find conditions that indirectly negate the effects of UCP2 overexpression.
The enzyme adenylate cyclase catalyzes the conversion of ATP to cAMP, an important second messenger involved in potentiating rather than initiating insulin secretion (Prentki & Matschinsky 1987). Deficient cAMP production has been implicated in streptozotocin diabetes (Dachicourt et al. 1996), while increasing cAMP with forskolin improved GSIS in Goto-Kakizaki (GK) rats (Abdel-Halim et al. 1996). It is unknown whether enhancing cAMP production can ameliorate insulin secretion in ob/ob mice or other models where UCP2 is upregulated.
cAMP is proposed to influence insulin secretion at multiple steps via two distinct pathways, one mediated by protein kinase A (PKA) and the other by cAMP-regulated guanine nucleotide exchange factors (cAMP-GEFs, also known as epac; Holz 2004). Two important cAMP-dependent steps that also require adequate intracellular ATP are exocytosis, including regulation and replenishment of the readily releasable pool of granules (Kashima et al. 2001, Eliasson et al. 2003) and ATP-dependent potassium (KATP) channel inactivation (MacDonald et al. 2002).
We tested the hypothesis that elevation of cAMP subsequent to upregulation of UCP2 expression will normalize insulin secretion because cAMP is an important potentiator of GSIS, acts at many steps in stimulus-secretion pathways of β-cells and has been shown to ameliorate defective insulin secretion in other models.
Materials and Methods
Isolation of rat and mouse islets
All animals used in the present research were cared for in accordance with the Guidelines of the Canadian Council on Animal Care. The protocols were approved by animal use committees from the University of Toronto and the University of Prince Edward Island. Rats were anesthetized with sodium pentobarbital (0.1 ml/100 g). Islets were harvested from 10-week-old male and female lean Zucker rats (Charles River Laboratories, Boston, MA, USA and AVC born) and from 8-week-old lean and ob/ob mice (Jackson Labs, Bar Harbour, ME) and cultured as described (Kibenge & Chan 1995).
Adenoviral infection of rodent islets
Islets were infected with adenovirus encoding the full-length human or rat cDNA for ucp2 (Aducp2) as described (Chan et al. 1999, 2001, Joseph et al. 2004) for 48 h. Control islets received no treatment and were simply cultured in DMEM (Ducbecco’s modified Eagle medium; Invitrogen) for 48 h, because non-specific adenoviral infection had not been shown to affect insulin secretion (Chan et al. 1999). In some experiments, isolated rat islets were cultured overnight in DMEM supplemented with 10 μM forskolin (Sigma). Over-expression of ucp2 was confirmed by quantitative reverse transcriptase (RT)-PCR as described (Joseph et al. 2002).
Perfused mouse pancreas
Control (wild type, WT) and ucp2 knockout (KO) mice were raised in a colony at the University of Toronto. Generation of the KO line has been described previously (Zhang et al. 2001). The pancreases from 4-month-old male ucp2 KO mice were perfused after an overnight fast. The anesthetic was i.p. sodium pentobarbital (80 mg/kg). Details of the perfusion protocol are given elsewhere (Joseph et al. 2002). In these experiments, the glucose infusion was changed from 1.4 to 13.4 mM after 5 min, then returned to 1.4 mM at 25 min. Exendin-4 (1 nM, Bachem, Bubendorf, Switzerland), which acts as an agonist at GLP-1 receptors (Raufman et al. 1992) was added in some experiments.
Cell culture and transient transfection of βTC-6f7 cells
Cultures of βTC-6f7 (β-TC) cells were grown in RPMI-1640 media (Invitrogen), pH 7.4 as described (Chan et al. 2001). For transient gene transfection, β-TC cells were plated at a density of 2.5 × 105 cells/cm2 and cultured overnight. Plasmids containing either ucp2 or enhanced green fluorescent protein (egfp) cDNA were produced as described (Chan et al. 2001) and transfection carried out as per the instructions of the manufacturers of the transfection agent (Lipofectamine2000, Invitrogen). The degree of ucp2 overexpression (ucp2-OE) induced was assessed by quantitative RT-PCR, as for rat islets (Joseph et al. 2002).
cAMP content
Control and ucp2-OE islets were exposed to 0–10 μM forskolin (Sigma) or exendin-4 (1 nM) in the presence of 2.8 or 11 mM glucose in KRBB–HEPES for 15 min at 37 °C. The medium was then removed and 70% ethanol (0.5 ml/well) was added to the cell pellets. Cell lysates were frozen at −20 °C until assayed by RIA using the acetylation protocol provided in the kit for increased sensitivity (Biomedical Technologies, Inc., Stoughton, MA, USA).
Insulin secretion from cell lines or islets
Control and ucp2-OE β-TC cells or islets were prepared as described for cAMP formation and incubated for 2 h at 37 °C. Where indicated in the Results, additional test compounds were added to islet incubations. The PKA inhibitors, H89 and KT5720 (Sigma), were added to test wells 30 min prior to the addition of forskolin for a further 90 min. The medium was aspirated and placed in clean 1.5 ml microcentrifuge tubes, which were then centrifuged briefly to remove cellular debris. To each cell pellet, 1 ml 3% acetic acid was added to extract the total remaining insulin. All samples were frozen at −20 °C until assayed. Concentrations of insulin were determined by RIA.
Protein kinase A activity
The effects of ucp2-OE on PKA activity were assessed in rat islets using a colorimetric assay (Stressgen, Victoria, BC, Canada). Control and ucp2-OE islets (100 per condition) were incubated with 1 μM forskolin ± 10 μM H89 (Calbiochem, La Jolla, CA, USA) for 30 min, and washed in 0.5 ml PBS (pH 7.4). The islet cytosol was reserved following the cell lysis as recommended in the manufacturer’s instructions. The lysate (60 islet equivalents in 30 μl aliquots) was incubated for 30 min with the ATP substrate. Detection of phosphorylated proteins was facilitated by incubation with phosphoprotein-specific antibody. Although the assay was performed with crude extracts, the ability to inhibit most of the activity with H89 suggests that the majority of kinase activity was due to PKA.
ATP content
ATP and ADP were quantified in ucp2 KO mouse islets by the method of Schultz et al.(1993).
Assessment of KATP channel activity
Efflux of radiolabeled 86Rb+ was quantified to estimate changes in K+channel activity in islets as described previously (Chan & MacPhail 1996). Islets were loaded with 86RbCl (Amersham) for 90 min at 37 °C, during which time the radiolabel was taken up into the islet cells. After washing twice, the islets were resuspended in KRBB–HEPES containing low (2.8 mM) or high (11 mM) glucose to determine what degree of K channel activity was glucose metabolism-dependent and therefore reflective of KATP channels. In some tubes, 1 μM forskolin ± 1 μM H89 were also added. After 20 min, the supernatant buffer and the islet pellets were transferred to clean vials for scintillation counting. Results were expressed as the fractional efflux of 86Rb+/20 min.
Electrophysiological assessment of KATP channel activity was performed using patch clamping of intact islets infected with control (AdEmpty) or Aducp2 virus. Islets were held in the recording chamber with the assistance of a suction pipette. KATP current was measured in the perforated-patch configuration during 200 ms 20 mV incremental step-depolarizations from −140 to −60 mV from a holding potential of −70 mV. Mean outward current at each potential was measured during the last 50 ms of depolarization. The pipette solution contained (mM) K2SO4 28.4, KCl 63.7, NaCl 11.8, MgCl2 1, HEPES 20.8, EGTA 0.5, pH 7.2 with KOH+0.1 mg/ml amphotericin B (Sigma). The bath solution contained (mM) NaCl 115, CaCl2 3, KCl 5, MgCl2 2, glucose 11.1, HEPES 10, pH 7.2 with NaOH. Islets were held in a static bath maintained at 34 °C by an automatic temperature controller (TC-324B, Warner Instrument Corp., Hamden, CT, USA). Current recordings were amplified, digitized, and analyzed using pClamp9.0 software (Axopatch 200B, Axon Instruments, Union City, CA, USA).
Calcium flux measurement
β-TC cells or β-cells from dispersed islets were plated on glass coverslips (about 80% confluency) and transfected with plasmid encoding ucp2 or egfp (control) as described in Chan et al.(2001). Fluorescent measurements were conducted 48 h after transfection. Changes in intracellular Ca2+ concentrations were assessed using Fura 2-AM. The cells were loaded with 5 μM Fura 2-AM for 50 min in a medium containing (in millimolar) 130 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 5 NaHCO3, 10 HEPES, pH 7.4 at 37 °C. The cells were then washed, transferred to an open chamber, placed on an Olympus BX51W1 fluorescent microscope stage and perfused with the same medium at 1 ml/min. All experiments were performed at 36–37 °C using a TC-324B Heater Controller (Warner Instruments). For excitation, a xenon lamp-based DeltaRam high-speed monochromator (Photon Technology International, London, ON, Canada) was used. Cells were excited by dual excitation at 340/380 nm and emission was detected by a 510 nm band pass filter. For control of the monochromator and videocamera, as well as for fluorescent imaging and collecting of data the ImageMaster 3 software (Photon Technology International) was used. The basal intracellular Ca2+concentration ([Ca2+]i) was calculated according to the equation [Ca2+]i = Kd × β × (R−Rmin)/(R−Rmax), where R is the fluorescence ratio at 340 and 380 nm excitations (F340/F380), Rmin and Rmax are the fluorescence ratios at Ca2+- free and Ca2+-saturating conditions, Kd is the dissociation constant ( = 224 nM), and β = F380 (Ca2+-free)/F380 (Ca2+-saturating). Saturating and Ca2+-free conditions were obtained by the addition of 10 μM ionomycin (Sigma) to cells incubated with 5 mM Ca2+, and by addition of 10 mM EGTA to cells incubated in Ca2+-free medium.
Statistical analysis
Data are presented as means ± s.e.m. Results were compared by unpaired t-test, one- or two-way ANOVA, where applicable, and P < 0.05 was considered significant. Prism 3.0 software (GraphPad, San Diego, CA, USA) was used for the statistical analysis.
Results
Glucose- and forskolin-stimulated insulin secretion in ob/ob mouse islets
To determine if forskolin could restore GSIS in a model of insulin insufficiency and documented overexpression of UCP2 (Zhang et al. 2001), islets were obtained from lean and ob/ob mice. In these experiments, ob/ob islets had a ~twofold increase in ucp2 mRNA expression compared with age-matched lean mouse islets (Table 1). In lean islets, glucose alone increased insulin secretion by ~twofold over basal. Forskolin potentiated the response to higher glucose concentrations (Fig. 1A). Basal insulin secretion was elevated in ob/ob mice and there was no response to glucose. Addition of 10 μM forskolin significantly increased the insulin secretion at glucose concentrations from 5.5 to 22.0 mM (Fig. 1B).
Forskolin-stimulated cAMP formation and insulin secretion
Infection of rat islets with Aducp2 induced an 19-fold increase in ucp2 mRNA (Table 1). In these experiments, the ability of forskolin, an adenylate cyclase activator, to stimulate insulin secretion was tested. Both UCP2 expression (P < 0.0001) and forskolin concentration (P < 0.0001) influenced insulin secretion. A significant two-way interaction (P < 0.005) revealed that ucp2-OE cells responded differently compared with control islets to forskolin. Insulin secretion from Aducp2-infected cells was lower than control cells at 0 and 0.1 μM forskolin (P < 0.03) but was not different at 1 and 10 μM forskolin (Fig. 2A). Thus, at higher doses, forskolin normalized insulin secretion in Aducp2-infected rat islets. Similar effects were observed in clonal β-TC cells (not shown; Chan et al. 2001).
To determine if the effects of forskolin on insulin secretion were dependent on PKA activation, islets were preincubated with 10 μM H89 or 1 μM KT5720 for 30 min prior to the addition of forskolin. Figure 2B shows that inhibition of PKA by H89 and KT5720 blocked the stimulatory effects of forskolin on insulin secretion in both ucp2-OE and control islets. That ucp2-OE had no effect on PKA activity in islets is verified in Fig. 2C. Moreover, the increase in PKA activity, as measured by the production of phosphoproteins, was totally negated in the presence of 10 μM H89. cAMP concentrations were measured following exposure to forskolin in rat islets that had normal or elevated ucp2 mRNA expression. In the presence of 11 mM glucose, a similar concentration response to forskolin was seen in control and Aducp2-infected islets (Fig. 2D, P < 0.0001). No differences between control and Aducp2-infected β-cells were observed at any forskolin concentration.
cAMP analogs and insulin secretion
The ability of cell-permeant cAMP analogs 8-Br-cAMP (0.1 mM) and Sp-cAMPs (0.1 mM) to stimulate insulin secretion in ucp2-OE islets was compared with forskolin (1 μM). Both compounds increased insulin secretion in control and ucp2-OE islets (Fig. 2E). To examine PKA-independent effects mediated by epac, 8-pCPT-2-O-Me-cAMP (0.3 mM) was used. Significant effects on control but not ucp2-OE islets were observed (Fig. 2E).
Stimulation of insulin secretion after bypassing KATP channels
The KATP-independent pathway of insulin secretion can be studied by opening KATP channels with diazoxide and depolarizing the β-cells with KCl. This treatment resulted in the stimulation of insulin secretion in both control and ucp2-OE islets (Fig. 2F). Addition of 1 μM forskolin caused additional stimulation of insulin secretion in both groups (P < 0.01).
KATP channel activity
In the presence of 11 mM glucose, β-cells from ucp2-OE islets tended to have greater KATP current through KATP channels than control islets (P = 0.15, n = 5–6; Fig. 3A–C). Increased variability in the ucp2-OE group likely reflects different infection efficiencies between preparations. Because β-cell KATP channels are largely inhibited at 11 mM glucose (Ashcroft & Rorsman 1989, Cook & Ikeuchi 1989) and, thus, the KATP currents are very small, statistical significance was not achieved. Forskolin abrogated this enhancement of KATP current in ucp2-OE islets, diminishing KATP current amplitude to a level similar to that observed in control islets.
Rubidium efflux from β-cells was measured as an index of the effects of forskolin on K+ channel activity. Increasing glucose from 2.8 to 11 mM reduced 86Rb+efflux in control rat islets by ~20% (Fig. 3C), but had no significant effect in ucp2-OE islets (Fig. 3D). Addition of 1 μM forskolin significantly inhibited 86Rb+ efflux from ucp2-OE islets (Fig. 3D). However, in the presence of the PKA inhibitorH89 (10 μM),the inhibitoryeffects of forskolin on 86Rb+efflux in the ucp2-OE islets were reversed.
Calcium flux
Influx of Ca2+ through voltage-dependent Ca2+ channels occurs secondary to the inactivation of KATP channels. Representative traces showing changes in intracellular Ca2+ in β-TC cells transfected with ucp2 plasmid compared with control cells (Fig. 4A and B) and islet β-cells (Fig. 4C) are shown. A summary of the results is presented in Table 2. The basal [Ca2+]i was significantly higher in ucp2-OE than control β-TC cells. Addition of 10 mM glucose invoked an increase in [Ca2+]i that was only 14% of that observed in control cells, consistent with the failure to alter 86Rb+efflux. The increase in [Ca2+]i was identical in ucp2-OE and control β-TC cells when forskolin was applied (~23%) and the magnitude of the response was similar to that seen when the KATP channels were inactivated with the sulphonylurea tolbutamide. Likewise, exposure of the cells to 50 mM KCl evoked a strong increase in [Ca2+]i of > 250 nM in both control and ucp2-OE conditions. In islet β-cells, the increase in [Ca2+]i evoked by glucose was reduced by ~50% in ucp2-OE but the response to forskolin was ~75% of control. The difference probably reflects lower transfection efficiency in the primary cells.
Effects of chronic forskolin exposure on ucp2 expression and insulin secretion
It was previously reported (Lameloise et al. 2001) that culturing clonal β-cells in the presence of high concentrations of forskolin for ~18 h could inhibit the expression of ucp2, possibly representing an alternative mechanism for its ability to improve GSIS. To determine if effects on gene expression could contribute to improved insulin secretion, in addition to its acute stimulatory effects, rat islets were cultured overnight in the presence of 10 μM forskolin (Fig. 5A). Treatment with forskolin increased ucp2 mRNA expression by ~30% (P < 0.05). Basal insulin secretion was increased ~fivefold in islets cultured in forskolin and 16.5 mM glucose failed to elicit any further augmentation of secretion (Fig. 5B). In contrast, GSIS was threefold basal in the control islets.
Glucose and exendin-4 stimulated insulin secretion in UCP2 knockout mice
To determine how null expression of ucp2 affects cAMP-potentiated insulin secretion, ucp2 KO mice were utilized. The results of pancreas perfusion experiments are shown in Fig. 6A (WT) and B (ucp2 KO). In the presence of 1.4 mM glucose (1–5 min from Fig. 6A and B), ucp2 KO mice had threefold higher insulin secretion than WT mice. Higher (13.4 mM) glucose evoked a twofold higher response in ucp2 KO than WT mice (Fig. 6C). In WT mice, 1 nM exendin-4 stimulated insulin secretion by fivefold compared with 13.4 mM glucose alone (Fig. 6A and C). Exendin-4 also enhanced insulin secretion from ucp2 KO mouse islets, but by only 1.8-fold (Fig. 6B,C), such that there was no longer any difference in insulin secretion between ucp2 KO and WT mice.
ATP and cAMP content of UCP2 knockout mouse islets
cAMP, cATP, and cADP contents were measured in freshly isolated ucp2 KO and WT mouse islets. No genotype-dependent difference was observed when cAMP was measured in the presence of glucose alone, 1 nM exendin or 1 μM forskolin (Fig. 6D). However, in the presence of both low and high glucose concentrations, islets from ucp2 KO mice had a higher ATP:ADP ratio than WT mice (Table 3).
Discussion
The findings in the present study support the hypothesis that increasing cellular cAMP can partially ameliorate the effects of low ATP in islets overexpressing UCP2 either endogenously (ob/ob mice) or after genetic manipulation (ucp2-OE). Restoration of the insulin-secretory response to glucose in ucp2-OE islets was largely PKA-dependent and did not appear to involve the non-PKA-dependent cAMP signaling pathway acting through epac.
To study the mechanisms through which cAMP exerts its effects, an abundant source of β-cells was required. Normal rat islets or β-TC cells overexpressing UCP2 (Chan et al. 1999, 2001) were utilized. The quantitative PCR results suggest that ucp2 mRNA induction in both models was approximately 20-fold; however, previous work by Joseph et al.(2004) has shown discordance between transcription and translation levels, such that the actual induction of protein is actually two- to fivefold using these transfection or infection methods.
When β-cell ATP is decreased by ucp2-OE, the ability of glucose to inactivate KATP channels is attenuated (this study and Chan et al. 2001). Similar results were obtained with ucp1-OE in β-cells (Nakazaki et al. 2002). In contrast, in ucp2 KO mouse islets, glucose-stimulated Ca2+influx is enhanced (Joseph et al. 2005), reflecting increased KATP-dependent depolarization. Therefore, treatments that normalize KATP channel activity and Ca2+ influx should overcome the suppressive effects of UCP2 on GSIS. Here, forskolin reduced KATP currents measured by patch clamping and 86Rb+efflux in ucp2-OE islets and normalized glucose-stimulated Ca2+ influx in response to glucose. The effects of forskolin were largely mediated by PKA because the specific inhibitors H89 and KT5720 (Yaekura et al. 1996) blocked both KATP channel inhibition and insulin secretion, whereas the epac pathway-specific agonist 8-pCPT-2′-O-Me-cAMP (Holz 2004) did not enhance insulin secretion in ucp2-OE islets. Even in control islets, the epac agonist 8-pCPT-2-O-Me-cAMP had only modest effects on insulin secretion compared with forskolin, suggesting that this pathway ‘fine tunes’ insulin secretion through mechanisms not sufficient to over-ride other deficits. cAMPexerts its effects at least partially via KATP channels (Light et al. 2002). Using β-cell lines, Light et al.(2002) showed that H89 completely blocked the inhibition of KATP currents by GLP-1. The results presented here support the hypothesis that an increase in cAMP can ameliorate insulin secretion despite low cellular ATP via actions on the KATP channel.
These studies also support the potential effects of cAMP on distal events related to exocytosis (Takahashi et al. 1999). Forskolin enhanced insulin secretion by the KATP-independent pathway in both control and ucp2-OE islets. The magnitude of potentiation in depolarized islets was about 30%, whereas total potentiation in non-depolarized islets was 50–100%, suggesting that a significant fraction of cAMP-mediated effectswere exerted on the exocytotic machinery. Exocytosis of insulin can be positively modulated by both PKA and epac (Holz 2004). Since 8-pCPT-2′-O-Me-cAMP was ineffectual in ucp2-OE islets, we speculate that PKA-dependent steps, for example, replenishment of the readily releasable granule pool were contributing to the enhancement of GSIS under these conditions.
An unexpected finding of these studies was the higher basal [Ca2+]i in ucp2-OE ß-cells. The cause of this observation was not investigated but we speculate that lower cellular ATP may lead to sluggish activity of Ca-ATPases that normally sequester Ca2+ in the endoplasmic reticulum or pump it out of the cell. Interestingly, ucp2 KO mice, which have elevated cellular ATP, did not have lower basal [Ca2+]i but after exposure to palmitic acid, which impairs the ability of WT but not ucp2 KO mouse islets to generate ATP, the [Ca2+]i was lower in the ucp2 KO islets (Joseph et al. 2004).
Changes in UCP2 expression did not alter cAMP production in β-cells. The lack of difference is consistent with kinetic analyses of adenylate cyclase, which showed that the Km for ATP is ~350 μM (Dessauer & Gilman 1997). Resting ATP concentrations in β-cell cytosolic and plasma membrane domains have been estimated to be ~1000 μM, approximately doubling after high glucose exposure (Kennedy et al. 1999). Thus, even a 50% reduction in ATP should not affect cAMP generation except under exceptional circumstances. Moreover, as the cAMP content increased after forskolin stimulation, the insulin secretion by ucp2-OE islets was normalized. Because PKA activity was similar in control and ucp2-OE islets, it is unlikely that upregulation of this enzyme contributed to the effect. Results from the ob/ob mouse islets showed that in the presence of forskolin a robust insulin response to glucose was generated. Thus, adequate cAMP levels were able to ‘rescue’ energy-deficient islets. Likewise, cAMP is shown to partially relieve the impairment in GSIS observed in the hyperglycemic GK rat (Abdel-Halim et al. 1996). The gene for ucp2 maps within the quantitative trait locus of chromosome 1 for glucose intolerance in the GK rat (Kaisaki et al. 1998), but whether β-cell ucp2 expression is elevated is unclear. One study showed that the treatment with insulin to normalize glycemia reduced UCP2 protein expression on GK rats fed a high fat diet (Briaud et al. 2002). In neonatal streptozotocin diabetic rats, agents that induce β-cell cAMP also enhance GSIS, although not to control levels (Dachicourt et al. 1996). Our data, together with other published reports, support the development of new treatment modalities for type-2 diabetes mellitus that target cAMP, such as long-acting stimulants of cAMP production or phosphodiesterase inhibitors (Furman et al. 2004). However, caution against chronic amplification of cAMP-mediated effects is warranted in light of the upregulation of ucp2 mRNA observed in this study and the accompanying exaggerated basal insulin secretion. Agonists that increase cAMP also increased UCP2 expression in muscle cells (Nagase et al. 2001) and activation in dopamine-secreting cells (Yamada et al. 2003), presumably via a cAMP response element in the ucp2 promoter (Yoshitomi et al. 1999). Interestingly, a previous report indicated that 18-h forskolin treatment of rat insulinoma cells (INS-1) reduced ucp2 expression (Lameloise et al. 2001). Differences from our study might be due to the experimental model (transformed cells versus rat islets) and the concentration of forskolin used (fivefold higher than the present study).
Previous studies showed that ucp2 KO mice have an increased insulin-secretory response to both basal and stimulatory glucose concentrations (Zhang et al. 2001, Joseph et al. 2004), a finding confirmed here. When WT mouse pancreases were perfused with stimulatory glucose plus exendin-4, a glucagon-like peptide analog that activates adenylate cyclase to elevate cAMP, insulin secretion was increased by ~fourfold. In comparison, although exendin-4 potentiated GSIS in the ucp2 KO mice, the increment was only ~twofold; thus, the difference in insulin-secreting capacity between the WT and ucp2 KO mice was negated. These data suggest that an abundance of ATP does not lead to any exceptional potentiation of GSIS by cAMP. Rather, the situation appears to resemble that found in the Zucker fatty rat, where a decrease in the Km for glucokinase (Chan 1993, Chan et al. 1995) leads to a left shift in the responsiveness to glucose at both the level of the KATP channel (Chan & MacPhail 1996) and insulin secretion (Chan et al. 1995, 1996). The enhanced inhibition of the KATP channel was presumed due to a greater increment in ATP permitted by the increased glucose flux through glucokinase. Thus, these rats do not display any increased sensitivity to cAMP generating incretins, such as GIP (Chan et al. 1984). Similarly in the diabetic sand rat, any increase in response to incretins or forskolin was attributed to an increase in sensitivity to glucose (Nesher et al. 2001). Of interest, however, was the relatively greater enhancement of the first phase than the second phase of insulin secretion by exendin in ucp2 KO mice. The magnitude of the first phase depends upon the number of granules in the readily releasable pool (Barg et al. 2002) and, in the intact, perfused pancreas, the total β-cell mass, as well as KATP channel inactivation. KATP channel closure may be enhanced in ucp2 KO mice because of the relatively higher ATP/ADP at low glucose (Joseph et al. 2004). Also, in ucp2 KO mice, β-cell mass is increased by 50%, but alterations in granule priming and docking in the presence of elevated cAMP cannot be ruled out.
In conclusion, when β-cell ATP is increased or decreased by manipulating UCP2 expression, different effects on ATP-dependent pathways are observed. Whereas the ability of the KATP channel to close in response to glucose is impaired, cAMP formation is unaltered. This supports the theory that various elements of the secretory pathway have differing energy requirements (Detimary et al. 1994). However, when cAMP concentrations in the islet are elevated, insulin secretion in the ATP-deficient state can be normalized, largely through PKA-dependent mechanisms.
Induction of UCP2 in various experimental models
| Ratio of UCP2 mRNA expression to GAPDH mRNA | Fold increase | |
|---|---|---|
| Expression data are means ± s.e.m. *P < 0.05 and †P < 0.001 compared with respective controls. | ||
| Experimental model | ||
| ob/ob Mouse islets (4) | 0.13 ± 0.01* | 1.86 |
| Lean mouse islets (5) | 0.069 ± 0.020 | |
| β-TC + ucp2 plasmid (3) | 1.82 ± 0.21† | 28.2 |
| β-TC (3) | 0.064 ± 0.007 | |
| Rat islets + AdUCP2 (7) | 0.21 ± 0.08† | 19.4 |
| Rat islets (9) | 0.011 ± 0.007 | |
[Ca2+]i (nM) responses to glucose and other stimuli in UCP2-OE β-TC cells
| Basal [Ca2+]i | Δ[Ca2+]i 10 mM glucose | Δ[Ca2+]i 25 μM forskolin | Δ[Ca2+]i 300 μM tolbutamide | Δ[Ca2+]i 50 mM KCl | |
|---|---|---|---|---|---|
| Data are means ± s.e.m. (n), where (n) is the total number of replicates performed from four independent transfections. Δ[Ca2+]i represents the amplitude of the change (in nM) following the treatment indicated. *P < 0.01 compared with ucp2-OE cells. | |||||
| Control | 73.5 ± 2.18 (14)* | 43.8 ± 9.1 (14) | 22.7 ± 8.0 (4) | 37.7 ± 6.0 (10) | 266 ± 20 (14) |
| UCP2-OE | 106.2 ± 2.38 (20) | 6.0 ± 3.0 (20)* | 23.7 ± 4.2 (9) | 33.4 ± 2.9 (19) | 255 ± 9.5 (20) |
ATP and ADP content of islets isolated from UCP2 KO and WT mice
| Genotype | ATP (pmol/islet) | ADP (pmol/islet) | ATP/ADP | |
|---|---|---|---|---|
| Data are means ± s.e.m. n = 5 for all experiments. *P < 0.05 comparing phenotypes at the same glucose concentrations. | ||||
| Glucose (mM) | ||||
| 2.8 | UCP2 KO | 26.63 ± 2.53 | 2.89 ± 0.46 | 10.27 ± 1.46* |
| WT | 20.8 ± 1.94 | 4.20 ± 0.58 | 5.19 ± 0.49 | |
| 16.7 | UCP2 KO | 32.56 ± 2.52 | 2.30 ± 0.23 | 14.41 ± 0.69* |
| WT | 25.42 ± 1.24 | 2.64 ± 0.26 | 10.23 ± 1.46 | |
Effects of forskolin (□) on glucose (2.8–22.0 mM)-stimulated insulin secretion of islets isolated from (A) lean (n = 14), and (B) ob/ob mice (n = 10). Control islets from each genotype are shown by ▪. *P < 0.05, **P < 0.01 and ***P < 0.001 compared with controls.
Citation: Journal of Endocrinology 190, 3; 10.1677/joe.1.06723
(A) Effects of ucp2-OE (solid bars) compared with controls (open bars) on forskolin-stimulated insulin secretion in the presence of 11 mM glucose in isolated rat islets. n = 8 or greater. *Significant (P < 0.05) effect of ucp2-OE. †Significant (P < 0.05) effect of forskolin. (B) Effects of ucp2-OE (solid bars) on forskolin-stimulated cAMP generation in rat islets exposed to 11 mmol/l glucose. Control islets are represented by open bars. n = 5–8. *Significant effect of forskolin (P < 0.05). (C) Effects of specific PKA inhibitors H89 (10 μM) and KT5720 (1 μM) on forskolin (1.0 μM)-stimulated insulin secretion in ucp2-OE rat islets. *P < 0.05 compared with 11 mM glucose (n = at least 9 for all groups). (D) Effects on insulin secretion of cAMP analogs that activate PKA and epac (0.1 mM 8-Br-cAMP, 0.1 mM Sp-cAMPs or 1 μM forskolin) and an epac-specific analog 8-pCPT-2-O-Me-cAMP (0.3 mM). *P < 0.05 compared with 11 mM glucose. n = 6–10. (E) Activity of PKA in islets overexpressing ucp2 (solid bars, n = 5) compared with controls (open bars, n = 6) under basal conditions (0), stimulated with 1 μmol/l forskolin (Forsk) ± H-89 (10 μM). *P < 0.05 compared with basal PKA activity. (F) Insulin secretion via KATP-independent pathway in control and ucp2-OE islets. Islets were held in a depolarized state by addition of 0.1 mM diazoxide and 20 mM KCl in the presence of 11 mM glucose. Forskolin (1 μM) was added to the depolarized islets. *P < 0.05 compared with 2.8 mM glucose, †P < 0.05 compared with 11 mM glucose. n = 6–10 for all groups.
Citation: Journal of Endocrinology 190, 3; 10.1677/joe.1.06723
KATP channel activity. (A) Representative KATP current recordings, current tracings from control (AdEmpty) and ucp2-OE rat islets in the absence and presence of 1 μM forskolin as the clamp voltage is raised from −140 to −60 mV. Glucose concentration in the bath was 11 mM. (B) I/V curves of grouped KATP current data. Summary of KATP channel current data normalized to cell size (pA/pF) at selected voltages (n = 5–6). (C) Area under the curve (AUC) of current-voltage relationship shown in (B) showing increased KATP current in ucp2-OE islets, and attenuation of this by 1 μmol/l forskolin (n = 5–6). (D) Glucose-induced inhibition of 86Rb+efflux (% per 20 min), as a marker of K+ flux, in control rat islets, indicating closure of KATP channels. Islets were exposed to either 2.8 or 11.0 mmol/l glucose. (E) Effects of ucp2-OE in rat islets on 86Rb+efflux (% per 20 min) in the presence of 2.8 mM glucose (baseline), 11 mM glucose, 11 mM glucose +1 μM forskolin (all n = 15), or 11 mM glucose+1 μM forskolin+10 μM H-89 (n = 8). *P < 0.05 compared with 2.8 mM glucose.
Citation: Journal of Endocrinology 190, 3; 10.1677/joe.1.06723
Representative results of Ca2+ flux measurements in (A) control and (B) ucp2-OE β-TC cells. (C) Ca2+flux measurements in dispersed β-cells from (a) control and (b) ucp2-OE mouse islets. Coverslips with cells or islets loaded with Fura-2 AM were transferred to an open chamber, placed on the microscope stage and perfused at 37 °C at a flow rate of 1 ml/min (see Materials and Methods). Horizontal bars indicate the addition of stimulants as follows: Basal [Ca2+]i was measured in the presence of 1 mM glucose (Glc). The responses to 10 mM glucose and 10 mM glucose+forskolin (Forsk, 25 μM), tolbutamide (Tolb, 300 μM), or KCl (50 mM) were measured sequentially. The vertical bar indicates the magnitude of change in [Ca2+]i. Summary results of replicate experiments are reported in Table 2.
Citation: Journal of Endocrinology 190, 3; 10.1677/joe.1.06723
Effect of culturing rat islets overnight in medium containing 10 μM forskolin on (A) ucp2 expression and (B) insulin secretion stimulated by 2.8 or 16.5 mM glucose. n = 8 for all experiments; *P < 0.05.
Citation: Journal of Endocrinology 190, 3; 10.1677/joe.1.06723
Effects on insulin secretion of glucose (▪) versus glucose+1 nM exendin-4 (□) in isolated pancreases from WT (A, n = 10 each) and ucp2 KO mice (B, n = 10 each). Pancreases were perfused with 1.4 mM glucose for 1–5 min, 13.4 mM glucose ± exendin-4 for 6–25 min (indicated by the horizontal bar), and 1.4 mM glucose for 26–40 min. (C) Mean insulin secretion following glucose (Glc) or glucose+exendin-4 (Glc+EX) treatment during 12–20 min (WT open bars, ucp2 KO solid bars) from the data in A and B. This time frame was chosen to cover the second phase of insulin secretion. (D) cAMP content in isolated islets from WT (open bars, n = 8) and ucp2 KO mice (solid bars, n = 10) incubated with 13.4 mM glucose alone or with 1 nM exendin-4 or 1 μM forskolin. *P < 0.05 compared with WT, †P < 0.05 compared with glucose alone, within genotype.
Citation: Journal of Endocrinology 190, 3; 10.1677/joe.1.06723
The authors are grateful to the Canadian Diabetes Association and the Canadian Institutes of Health Research for research grants to C B C and M B W and thank Simon C Lee for technical assistance. T S M was the recipient of a Natural Sciences and Engineering Research Council Postgraduate Studentship Award. J W J was the recipient of a Canadian Institutes of Health Research Studentship Award and J D M of the Geoff Hogan Memorial Award from the University of Prince Edward Island. J E M-F is supported by a postdoctoral fellowship from the Canadian Institutes of Health Research and the Edward Christie Stevens Fellowship from the Faculty of Medicine, University of Toronto. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.
References
Abdel-Halim SM, Guenifi A, Khan A, Larsson O, Berggren PO, Ostenson CG & Efendic S 1996 Impaired coupling of glucose signal to the exocytotic machinery in diabetic GK rats: a defect ameliorated by cAMP. Diabetes 45 934–940.
Ashcroft FM & Rorsman P 1989 Electrophysiology of the pancreatic beta-cell. Progress in Biophysics and Molecular Biology 54 87–143.
Barg S, Eliasson L, Renstrom E & Rorsman P 2002 A subset of 50 secretory granules in close contact with L-type Ca2+ channels accounts for first-phase insulin secretion in mouse ß-cells. Diabetes 51 S74–S82.
Briaud I, Kelpe CL, Johnson LM, Tran PO & Poitout V 2002 Differential effects of hyperlipidemia on insulin secretion in islets of langerhans from hyperglycemic versus normoglycemic rats. Diabetes 51 662–668.
Chan CB 1993 Glucokinase activity in isolated islets from obese fa/fa Zucker rats. Biochemical Journal 295 673–677.
Chan CB & MacPhail RM 1996 KATP channel-dependent and -independent pathways of insulin secretion in isolated islets from fa/fa Zucker rats. Biochemistry and Cell Biology 74 403–410.
Chan CB, Pederson RA, Buchan AM, Tubesing KB & Brown JC 1984 Gastric inhibitory polypeptide (GIP) and insulin release in the obese Zucker rat. Diabetes 33 536–542.
Chan CB, MacPhail RM, Kibenge MT & Russell JC 1995 Increased glucose phosphorylating activity correlates with insulin secretory capacity of male JCR:LA-corpulent rat islets. Canadian Journal of Physiology and Pharmacology 73 501–508.
Chan CB, Lowe JM & Debertin WJ 1996 Modulation by glucose of insulin secretion and glucose phosphorylating activity in cultured pancreatic islets from obese (fa/fa) Zucker rats. International Journal of Obesity and Related Metabolic Disorders 20 175–184.
Chan CB, MacDonald PE, Saleh MC, Johns DC, Marban E & Wheeler MB 1999 Overexpression of uncoupling protein 2 inhibits glucose-stimulated insulin secretion from rat islets. Diabetes 48 1482–1486.
Chan CB, De Leo D, Joseph JW, McQuaid TS, Ha XF, Xu F, Tsushima RG, Pennefather PS, Salapatek AM & Wheeler MB 2001 Increased uncoupling protein-2 levels in beta-cells are associated with impaired glucose-stimulated insulin secretion: mechanism of action. Diabetes 50 1302–1310.
Cook DL & Ikeuchi M 1989 Tolbutamide as mimic of glucose on beta-cell electrical activity. ATP-sensitive K+channels as common pathway for both stimuli. Diabetes 38 416–421.
D’Adamo M, Perego L, Cardellini M, Marini MA, Frontoni S, Andreozzi F, Sciacqua A, Lauro D, Sbraccia P, Federici M et al.2004 The −866A/A genotype in the promoter of the human uncoupling protein 2 gene is associated with insulin resistance and increased risk of type 2 diabetes. Diabetes 53 1905–1910.
Dachicourt N, Serradas P, Giroix MH, Gangnerau MN & Portha B 1996 Decreased glucose-induced cAMP and insulin release in islets of diabetic rats: reversal by IBMX, glucagon, GIP. American Journal of Physiology 271 E725–E732.
Dessauer CW & Gilman AG 1997 The catalytic mechanism of mammalian adenylyl cyclase. Equilibrium binding and kinetic analysis of P-site inhibition. Journal of Biological Chemistry 272 27787–27795.
Detimary P, Gilon P, Nenquin M & Henquin JC 1994 Two sites of glucose control of insulin release with distinct dependence on the energy state in pancreatic B-cells. Biochemical Journal 297 455–461.
Eliasson L, Ma X, Renstrom E, Barg S, Berggren P-O, Galvanovskis J, Gromada J, Jing X, Lundquist I, Salehi A et al.2003 SUR1 regulates PKA-independent cAMP-induced granule priming in mouse pancreatic B-cells. Journal of General Physiology 121 181–197.
Furman B, Pyne N, Flatt P & O’Harte F 2004 Targeting beta-cell cyclic 3′ 5′ adenosine monophosphate for the development of novel drugs for treating type 2 diabetes mellitus. A review. Journal of Pharmacy and Pharmacology 56 1477–1492.
Holz GG 2004 Epac: a new cAMP-binding protein in support of glucagon-like peptide-1 receptor-mediated signal transduction in the pancreatic beta-cell. Diabetes 53 5–13.
Hong Y, Fink BD, Dillon JS & Sivitz WI 2001 Effects of Adenoviral Overexpression of uncoupling protein-2 and -3 on mitochondrial respiration in insulinoma cells. Endocrinology 142 249–256.
Joseph JW, Koshkin V, Zhang CY, Wang J, Lowell BB, Chan CB & Wheeler MB 2002 Uncoupling protein 2 knockout mice have enhanced insulin secretory capacity after a high-fat diet. Diabetes 51 3211–3219.
Joseph JW, Koshkin V, Saleh MC, Sivitz WI, Zhang CY, Lowell BB, Chan CB & Wheeler MB 2004 Free fatty acid-induced beta-cell defects are dependent on uncoupling protein 2 expression. Journal of Biological Chemistry 279 51049–51056.
Joseph JW, Chan CB & Wheeler MB 2005 UCP2 knockout mouse islets have lower consumption and faster oscillations of beta cell oxygen. Canadian Journal of Diabetes 29 19–26.
Kaisaki PJ, Woon PY, Wallis RH, Monaco AP, Lathrop M & Gauguier D 1998 Localization of tub and uncoupling proteins (Ucp) 2 and 3 to a region of rat chromosome 1 linked to glucose intolerance and adiposity in the Goto-Kakizaki (GK) type 2 diabetic rat. Mammalian Genome 9 910–912.
Kashima Y, Miki T, Shibasaki T, Ozaki N, Miyazaki M, Yano H & Seino S 2001 Critical role of cAMP-GEFII-Rim2 complex in incretin-potentiated insulin secretion. Journal of Biological Chemistry 276 46046–46053.
Kennedy HJ, Rafiq I, Pouli AE & Rutter GA 1999 Glucose enhances insulin promoter activity in MIN6 beta-cells independently of changes in intracellular Ca2+ concentration and insulin secretion. Biochemical Journal 342 275–280.
Kibenge MT & Chan CB 1995 Identification of biochemical defects in pancreatic islets of fa/fa rats: a developmental study. Obesity Research 3 171–178.
Lameloise N, Muzzin P, Prentki M & Assimacopoulos-Jeannet F 2001 Uncoupling protein 2: a possible link between fatty acid excess and impaired glucose-induced insulin secretion? Diabetes 50 803–809.
Light PE, Manning Fox JE, Riedel MJ & Wheeler MB 2002 Glucagon-like peptide-1 inhibits pancreatic ATP-sensitive potassium channels via a protein kinase A- and ADP-dependent mechanism. Molecular Endocrinology 16 2135–2144.
MacDonald PE, El-kholy W, Riedel MJ, Salapatek AMF, Light PE & Wheeler MB 2002 The multiple actions of GLP-1 on the process of glucose-stimulated insulin secretion. Diabetes 51 S434–S442.
Nagase I, Yoshida T & Saito M 2001 Up-regulation of uncoupling proteins by ß-adrenergic stimulation in L6 myotubes. FEBS Letters 494 175.
Nakazaki M, Kakei M, Ishihara H, Koriyama N, Hashiguchi H, Aso K, Fukudome M, Oka Y, Yada T & Tei C 2002 Association of upregulated activity of KATP channels with impaired insulin secretion in UCP1-expressing insulinoma cells. Journal of Physiology 540 781–789.
Nesher R, Warwar N, Khan A, Efendic S, Cerasi E & Kaiser N 2001 Defective Stimulus-secretion coupling in islets of Psammomys obesus, an animal model for type 2 diabetes. Diabetes 50 308–314.
Prentki M & Matschinsky FM 1987 Ca2+, cAMP, and phospholipid-derived messengers in coupling mechanisms of insulin secretion. Physiological Reviews 67 1185–1248.
Raufman JP, Singh L, Singh G & Eng J 1992 Truncated glucagon-like peptide-1 interacts with exendin receptors on dispersed acini from guinea pig pancreas. Identification of a mammalian analogue of the reptilian peptide exendin-4. Journal of Biological Chemistry 267 21432–21437.
Sasahara M, Nishi M, Kawashima H, Ueda K, Sakagashira S, Furuta H, Matsumoto E, Hanabusa T, Sasaki H & Nanjo K 2004 Uncoupling protein 2 promoter polymorphism -866G/A affects its expression in ß-Cells and modulates clinical profiles of Japanese type 2 diabetic patients. Diabetes 53 482–485.
Schultz V, Sussman I, Bokvist K & Tornheim K 1993 Bioluminometric assay of ADP and ATP at high ATP/ADP ratios: assay of ADP after enzymatic removal of ATP. Analytical Biochemistry 215 302.
Takahashi N, Kadowaki T, Yazaki Y, Ellis-Davies GCR, Miyashita Y & Kasai H 1999 Post-priming actions of ATP on Ca2+-dependent exocytosis in pancreatic beta cells. PNAS 96 760–765.
Yaekura K, Kakei M & Yada T 1996 cAMP-signaling pathway acts in selective synergism with glucose or tolbutamide to increase cytosolic Ca2+ in rat pancreatic beta-cells. Diabetes 45 295–301.
Yamada S, Isojima Y, Yamatodani A & Nagai K 2003 Uncoupling protein 2 influences dopamine secretion in PC12h cells. Journal of Neurochemistry 87 461–469.
Yoshitomi H, Yamazaki K & Tanaka I 1999 Mechanism of ubiquitous expression of mouse uncoupling protein 2 mRNA: control by cis-acting DNA element in 5′-flanking region. Biochemical Journal 340 397–404.
Zhang CY, Baffy G, Perret P, Krauss S, Peroni O, Grujic D, Hagen T, Vidal-Puig AJ, Boss O, Kim YB et al.2001 Uncoupling protein-2 negatively regulates insulin secretion and is a major link between obesity, beta cell dysfunction, and type 2 diabetes. Cell 105 745–755.
