Report

The cAMP Sensor Epac2 Is a Direct Target of Antidiabetic Sulfonylurea Drugs

See allHide authors and affiliations

Science  31 Jul 2009:
Vol. 325, Issue 5940, pp. 607-610
DOI: 10.1126/science.1172256

Expanding Sulfonylurea Mechanisms

Sulfonylureas are important drugs used for treatment of diabetes that act through adenosine triphosphate–sensitive potassium channels to promote secretion of insulin from the pancreas. Zhang et al. (p. 607) present another mechanism by which the beneficial effects of sulfonylureas may also be obtained. Sulfonylureas were identified in a screen for substances that modify the activity of Epac2, a guanine nucleotide exchange factor for the small guanosine triphosphatase Rap1. Mice lacking Epac2 were less responsive to sulfonylureas, which may suggest that Epac2 would be a useful target for development of drugs for treatment of diabetes.

Abstract

Epac2, a guanine nucleotide exchange factor for the small guanosine triphosphatase Rap1, is activated by adenosine 3′,5′-monophosphate. Fluorescence resonance energy transfer and binding experiments revealed that sulfonylureas, widely used antidiabetic drugs, interact directly with Epac2. Sulfonylureas activated Rap1 specifically through Epac2. Sulfonylurea-stimulated insulin secretion was reduced both in vitro and in vivo in mice lacking Epac2, and the glucose-lowering effect of the sulfonylurea tolbutamide was decreased in these mice. Epac2 thus contributes to the effect of sulfonylureas to promote insulin secretion. Because Epac2 is also required for the action of incretins, gut hormones crucial for potentiating insulin secretion, it may be a promising target for antidiabetic drug development.

Epac is a guanine nucleotide exchange factor (GEF) for the Ras-like small guanosine triphosphatases Rap1 and Rap2 that is activated by the direct binding of adenosine 3′,5′-monophosphate (cAMP); two isoforms of Epac (also referred to as cAMP-GEF), Epac1 (cAMP-GEFI) and Epac2 (cAMP-GEFII), have been identified (14). Epac2 mediates the potentiation of insulin secretion by cAMP in a protein kinase A (PKA)–independent pathway (5, 6). Epac1 undergoes a conformational change upon cAMP binding, and partial and full-length Epac1 sandwiched between cyan fluorescent proteins (CFPs) and yellow fluorescent proteins (YFPs) act as fluorescence resonance energy transfer (FRET) sensors that monitor Epac1 activation in living cells (79). These sensors facilitate investigations of the mechanisms of cAMP signaling such as compartmentalization (8) and oscillation (10).

We generated a FRET sensor (termed C-Epac2-Y) using the full-length Epac2 (fig. S1A) (11, 12). C-Epac2-Y was transiently expressed in simian kidney COS-1 cells, and we examined changes in FRET. Stimulation with 8-bromo-adenosine 3′,5′-monophosphate (8-Br-cAMP), a cAMP analog, decreased FRET in C-Epac2-Y–expressing COS-1 cells (Fig. 1A and fig. S1B). In contrast, treatment of cells with 8-Br-cAMP did not alter FRET in COS-1 cells expressing mutant Epac2 (termed C-MtEpac2-Y), in which both cAMP-binding sites were disrupted (fig. S1, C and D) (11). These results show that cAMP induces a conformational change in Epac2, confirming that C-Epac2-Y functions as a sensor for monitoring Epac2 activation in living cells.

Fig. 1

Sulfonylureas induce changes in FRET in C-Epac2-Y. (A) Emission ratio time courses of C-Epac2-Y treated with 8-Br-cAMP or vehicle [dimethyl sulfoxide (DMSO)] in COS-1 cells. (B) Emission ratio time courses of C-Epac2-Y treated with 500 μM TLB, 100 nM GLB, or vehicle (DMSO) in MIN6 cells. (C) Emission ratio time courses of C-Epac2-Y treated with TLB (100 μM, 250 μM, or 500 μM), 100 nM GLB, 100 nM GLC, or vehicle (DMSO) in COS-1 cells. The YFP/CFP ratio (R) was normalized to R0 to describe FRET efficiency changes (FRET change = R/R0), where R0 is the YFP/CFP ratio at time 0. FRET change was acquired every 5 s. Similar results were obtained in three independent experiments. Data are presented as mean ± SEM (n = 4 to 6 replications per experiment for each point).

To search for agents that activate Epac2, we screened for effects of various insulin secretagogues on change in FRET in MIN6, mouse clonal pancreatic β cells, transfected with C-Epac2-Y. Tolbutamide (TLB) and glibenclamide (GLB), both of which are sulfonylureas used in treatment of diabetes, decreased FRET (Fig. 1B and fig. S1E). Because Epac2 was identified as a molecule interacting with the sulfonylurea receptor SUR1, a regulatory subunit of adenosine 5′-triphosphate (ATP)–sensitive K+ (KATP) channels (11), we considered the possibility that TLB and GLB might have a secondary effect on FRET by affecting the interaction between exogenously introduced C-Epac2-Y and endogenously expressed SUR1 in the MIN6 cells. To clarify this, we examined the effect of these sulfonylureas on FRET in C-Epac2-Y–transfected COS-1 cells in which no endogenous SUR1 is expressed. Both TLB and GLB still decreased FRET (Fig. 1C), indicating that they might act directly on Epac2. The possibility that sulfonylureas might decrease FRET by elevating intracellular cAMP levels was excluded (fig. S2). Other sulfonylureas, including acetohexamide, glipizide, and chlorpropamide, also decreased FRET to different degrees and with different kinetics (fig. S3A). However, neither nateglinide (NTG) nor repaglinide, glinide-derivatives lacking the sulfonylurea core structure that also stimulate insulin secretion by the closure of KATP channels through acting directly on SUR1 (13), affected FRET (fig. S3B). These results indicate that the core structure of sulfonylurea is required for binding to Epac2. The finding that gliclazide (GLC), a sulfonylurea that has a side chain on the urea group larger than that of other sulfonylureas (fig. S4), did not decrease FRET (Fig. 1C) suggests that the structure or size (or both) of the side chain on the urea group may determine capability of interaction between sulfonylureas and Epac2.

To test whether sulfonylureas bind to Epac2 directly, we performed binding experiments using radiolabeled GLB. GLB exhibited specific binding to the full-length Epac2 or SUR1 expressed in COS-1 cells (Fig. 2A). Binding of [3H]glibenclamide to Epac2 was inhibited by unlabeled TLB or unlabeled GLB in a concentration-dependent manner (Fig. 2B), indicating that TLB probably binds specifically to Epac2 at the same binding site as GLB. The median inhibitory concentration (IC50) values of GLB and TLB for binding Epac2 were 25 nM and 240 μM, respectively, indicating lower affinities for binding Epac2 than for SUR1 (IC50 of GLB was 7.1 nM and of TLB was 140 μM) (14). Specific binding of GLB to Epac2 was not significantly inhibited in the presence of 8-Br-cAMP (Fig. 2B), suggesting that the sulfonylurea-binding site may be distinct from the cAMP-binding sites.

Fig. 2

Sulfonylureas bind specifically to Epac2. (A) Binding of [3H]glibenclamide to SUR1 (top) or Epac2 (bottom). COS-1 cells were transfected with mouse Epac2 or human SUR1 cDNA. Two days after transfection, cells were collected and suspended in a binding assay buffer. [3H]glibenclamide was added (1 to 40 nM) to the cell suspension. Total binding was determined in the absence of unlabeled GLB (open circles). Nonspecific binding in the presence of 100 μM unlabeled GLB was determined at each concentration of [3H]glibenclamide (open triangles). Specific binding (solid circles) was calculated by subtracting nonspecific binding from total binding. Data are presented as mean ± SEM (n = 6 experiments for each point). (B) Displacement curves of [3H]glibenclamide (10 nM) binding to Epac2 by unlabeled GLB, TLB, and 8-Br-cAMP. Data are presented as mean ± SEM (n = 3 experiments for each point).

Because Epac2 shows GEF activity toward Rap1 (4), we tested for possible sulfonylurea-induced activation of endogenous Rap1 in MIN6 cells. TLB activated Rap1 (Fig. 3A). The other sulfonylureas that decreased FRET all activated Rap1 significantly (fig. S5). GLB elicited activation at lower concentrations (10 to 100 nM) but not at higher concentrations (>200 nM). GLC, which did not decrease FRET, also had no effect on Rap1 activity. To determine whether activation of Rap1 by sulfonylureas is mediated through Epac2, we used Epac2-deficient, mouse clonal pancreatic β cells (15). No activation of Rap1 by either TLB or GLB was found in the Epac2-deficient cells, but activation of Rap1 by these sulfonylureas was detected after the adenoviral expression of wild-type (WT) Epac2 (Fig. 3B). Similar results were obtained with chlorpropamide, acetohexamide, and glipizide (fig. S6). NTG did not activate Rap1 even after the introduction of WT Epac2 into the Epac2-deficient cells (Fig. 3B). These results support the conclusion that sulfonylureas activate Rap1 at least in part through Epac2.

Fig. 3

Sulfonylureas activate Rap1 specifically through Epac2 in insulin-secreting cells. (A) Activation of Rap1 in MIN6 cells treated with TLB, GLB, or GLC. 8-Br-cAMP (1 mM) and 12-O-tetradecanoylphorbol-13-acetate (TPA) (1 μM) were used as positive controls. (B) Changes in activation of Rap1 by sulfonylureas through the introduction of Epac2 into Epac2-deficient mouse clonal β cells by means of adenovirus-based gene transfer. A representative blot for each experiment is shown. Similar results were obtained from three to seven independent experiments. Quantification of autoradiography is shown with corresponding bars positioned under the bands. The intensity of the Rap1-GTP signal was normalized by that of total Rap1. Data are presented as mean ± SEM (n = 3 to 7 experiments for each point). Dunnett’s method was used for multiple comparisons with a control group (vehicle was DMSO). *P < 0.05; **P < 0.01; ***P < 0.001. N.S., not significant.

We investigated the role of Epac2 in stimulation of insulin secretion by sulfonylurea. Neither glucose- nor potassium-stimulated insulin secretion was different in pancreatic islets isolated from Epac2−/− mice and those from WT C57BL/6 mice (Fig. 4A). In contrast, both TLB-stimulated and GLB-stimulated insulin secretion from the pancreatic islets of Epac2−/− mice were significantly reduced as compared with those from the islets of control animals (Fig. 4, B and C). Consistent with the findings of FRET response and Rap1 activity, there was no significant difference in insulin secretion in response to GLC (Fig. 4D). To clarify the role of Epac2 in modulating responses to sulfonylurea in vivo, we examined the effects of TLB on serum insulin and blood glucose in animals responding to oral intake of glucose. There were no significant differences in either serum insulin or blood glucose responses to the administration of glucose alone between Epac2−/− mice and WT mice (Fig. 4E). In contrast, the insulin response to concomitant administration of glucose and TLB in Epac2−/− mice was significantly reduced as compared with that in WT mice (Fig. 4F). In addition, the blood glucose levels after concomitant administration of glucose and TLB in Epac2−/− mice were significantly higher than those in WT mice (Fig. 4F). Furthermore, the glucose-lowering effect of TLB was not seen in Epac2−/− mice, and the insulin response to TLB was significantly reduced as compared with that in WT mice (Fig. 4G). These results show that the effects of Epac2 on responses to sulfonylurea operate in vivo.

Fig. 4

Effects of Epac2 on sulfonylurea-induced insulin secretion in vitro and in vivo. (A to D) Insulin secretion from the pancreatic islets of WT mice (open columns) and Epac2−/− mice (solid columns). Effect of glucose alone or potassium (K+, 60 mM) on insulin secretion (A). Effects of (B) TLB (100 μM), (C) GLB (10 nM), and (D) GLC (100 nM) on insulin secretion. Data are presented as mean ± SEM (n = 6 to 8 mice for each point). (E to G) Serum insulin and blood glucose responses to oral administration of glucose alone (E), concomitant administration of glucose and TLB (F), or TLB alone (G). Data are presented as mean ± SEM (n = 6 mice for each point). Unpaired student’s t test was used for evaluation of statistical significance. *P < 0.05; **P < 0.01. Blood glucose level in WT mice decreased significantly 60 min after administration of TLB (0 versus 60 min, P = 0.019, paired test) whereas that in Epac2–/– mice (G) did not (P = 0.36, paired t test).

Sulfonylureas stimulate insulin secretion by closing pancreatic β cell KATP channels through binding to SUR1 (16, 17). Studies of mice lacking the inward rectifier K+ channel member Kir6.2, the pore-forming subunit of KATP channels, and mice lacking SUR1 have shown that closure of the KATP channels is prerequisite for sulfonylureas to stimulate insulin secretion (1820). Although sulfonylurea has also been suggested to act through an intracellular target (2126), the target has not been identified. We identified Epac2 as a direct intracellular target of sulfonylureas. Considered together, in addition to closure of KATP channels, activation of Epac2 is required for sulfonylureas to exert their full effects on insulin secretion (fig. S7). Epac2 has also been implicated in the stimulation of insulin secretion through mechanisms not involving Rap1, such as the modulation of KATP channel activity (27). Because Epac2 also mediates the potentiation of insulin secretion by cAMP-increasing agents such as glucagon-like peptide 1 (28), analogs of which have been developed recently as new hypoglycemic agents (29), Epac2 may be a promising target for the development of drugs to treat diabetes.

Supporting Online Material

www.sciencemag.org/cgi/content/full/325/5940/607/DC1

Materials and Methods

Figs. S1 to S7

References

  • * Present address: Cell Scale Team, Integrated Simulation of Living Matter Group, Computational Science Research Program, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan.

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. We thank L. H. Philipson for critical reading of the manuscript; T. Kataoka for RalGDS-RID cDNA; M. Matsuda, S. Nakajima, and H. Hiroaki for helpful advice; A. Tamamoto and Y. Takahashi for technical assistance; and G. K. Honkawa for assistance with the manuscript. This work was supported by a CREST grant from the Japan Science and Technology Agency and Grant-in-Aid for Scientific Research and by a grant for the Global Centers of Excellence Program “Global Center of Excellence for Education and Research on Signal Transduction Medicine in the Coming Generation” from the Ministry of Education, Culture, Sports, Science and Technology.
View Abstract

Navigate This Article