Abstract

Research Design and Methods

Results

GPR119 activation and glucose tolerance

Because plasma levels of GLP-1 and GIP increase after GPR119 activation (14), we assessed whether complete loss of incretin action diminishes the glucoregulatory actions of GPR119. The GPR119 agonist AR231453 and oral glucose were administered to WT, Gipr−/−, Glp1r−/−, and DIRKO mice (Fig. 1). AR231453 significantly reduced glucose excursions after oral glucose loading in WT and Gipr−/− mice (Fig. 1, A and B). In contrast, a detectable but nonsignificant reduction in glycemic excursion was observed after oral glucose loading and AR231453 treatment of Glp1r−/− and DIRKO mice (Fig. 1, C and D). To elucidate the importance of enteral vs. parenteral glucose exposure for the glucoregulatory actions of GPR119 agonists, we tested whether AR231453 improved ip glucose tolerance. Oral AR231453 administration failed to control glucose excursions in WT mice after ip glucose challenge (Fig. 1E). Plasma hormones assessed after ip glucose challenge revealed an increase in GLP-1 but no significant changes in circulating GIP, insulin, or glucagon (Supplementary Fig. 1, published on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org). Hence, GPR119 activation requires GLP-1R signaling for optimal control of oral glucose tolerance, and GPR119 activation in the absence of enteral glucose exposure is not sufficient for optimal control of glucose homeostasis.

Open in a separate window

FIG. 1

GPR119 activation and control of oral glucose tolerance in WT and IRKO mice. Age-matched male mice were fasted overnight, and AR231453 (20 mg/kg) or vehicle was administered orally 30 min before an oral glucose tolerance test (OGTT) (A–D) or ip glucose tolerance test (IPGTT) (E) (1.5 g/kg). Blood was collected from the tail vein at 5 min (150 μl) and at 15 min (50 μl) after glucose administration. Blood glucose values and AUC analysis during OGTT for (A) WT, (B) Gipr−/−, (C) Glp1r−/−, and (D) DIRKO mice treated with or without AR231453. Blood glucose values during IPGTT for WT mice treated with or without AR231453 (E). Statistical analysis was assessed by ANOVA. *, P < 0.05; **, P < 0.01. For vehicle-treated mice: n = 12 (A), n = 8 (B), n = 10 (C), and n = 12 (D); for AR231453-treated mice: n = 11 (A), n = 8 (B), n = 10 (C), n = 12 (D), and n = 10–12 (E).

GPR119, incretin and islet hormones and insulin tolerance

To interpret the findings observed with WT and KO mice in Fig. 1, we assessed whether AR231453 regulates circulating levels of incretin and islet hormones to a similar extent in WT and IRKO mice. Plasma levels of total GIP were modestly increased by AR231453 in WT and Gipr−/− mice (Fig. 2A) and significantly increased in Glp1r−/− and DIRKO mice (Fig. 2A). In contrast, plasma levels of total GLP-1 were significantly increased by AR231453 in WT, Gipr−/−, Glp1r−/−, and DIRKO mice (Fig. 2B). Similarly, plasma insulin levels were significantly increased in WT, Gipr−/−, and Glp1r−/− mice after oral glucose and AR231453 administration (Fig. 2C). Although insulin levels were significantly lower in DIRKO compared with WT mice after oral glucose alone, reflecting defective β-cell function in DIRKO islets (19, 25, 26), AR231453 significantly increased plasma insulin levels in DIRKO mice, although to levels lower than those observed in WT or single IRKO mice (Fig. 2C). AR231453 had no effect on plasma glucagon levels in WT mice (Fig. 2D). However, basal glucagon levels were modestly lower in IRKO mice and increased significantly after oral glucose and AR231453 administration in Gipr−/−, Glp1r−/−, and DIRKO mice (Fig. 2D). Insulin sensitivity assessed by insulin tolerance testing was comparable in WT and DIRKO mice in the presence or absence of AR231453 (Fig. 3). Hence, GPR119 activation significantly increases GLP-1 levels in both single IRKO and DIRKO mice, and acute GPR119 activation increases plasma insulin levels despite the complete absence of IR action.

Open in a separate window

FIG. 2

GPR119 activation and plasma levels of GIP, GLP-1, insulin, and glucagon in WT and IRKO mice. Age-matched male mice were fasted overnight, and oral AR231453 (20 mg/kg) or vehicle was administered 30 min before an oral glucose tolerance test (OGTT). Plasma was obtained 5 min after oral glucose administration This sample was used to simultaneously measure the levels of (A) total GIP immunoreactivity, (B) total GLP-1 immunoreactivity, (C) and insulin. A second plasma sample was collected from each mouse at 15 min after glucose for the measurement of glucagon levels (D) in WT, Gipr−/−, Glp1r−/−, and DIRKO mice. Statistical analysis was assessed by ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Open in a separate window

FIG. 3

GPR119 activation does not modify glucose profiles in an insulin tolerance test (ITT) in WT and DIRKO mice. WT (n = 10) and DIRKO (n = 8) age-matched mice were fasted for 5 h, and oral AR231453 (20 mg/kg) or vehicle was administered 30 min before ip insulin administration (1.2 U/kg). Blood glucose values during ITT for (A) WT and (B) DIRKO mice.

GPR119 directly increases insulin secretion from isolated islets

We first assessed whether the diminished response of DIRKO mice to AR231453 reflects a selective inability of murine β-cells to respond to direct GPR119 activation and/or a generalized defect in secretagogue-regulated insulin secretion from DIRKO β-cells (19, 25, 26). Acute administration of Arg significantly increased plasma insulin levels in both WT and DIRKO mice (Fig. 4A), demonstrating that DIRKO mice retain the ability to robustly respond to non-GPCR-mediated amino acid insulin secretagogues. One possibility that might explain altered glucoregulatory responses to AR231453 could be a change in the relative level of GPR119 expression in single IR or DIRKO islets. However, no significant differences in the relative levels of GPR119 mRNA transcripts were observed in RNA from WT, Glp1r−/−, and Gipr−/− vs. DIRKO islets (Fig. 4B). As a control for assessment of gene expression in murine islets, we examined Irs2. In keeping with previous findings (22), levels of Irs2 mRNA transcripts were significantly reduced in islet RNA from Glp1r−/− relative to Gipr−/− or WT mice (Fig. 4B). Similarly, consistent with the loss of GLP-1R activity, Irs2 mRNA transcripts were also reduced in RNA from DIRKO islets (Fig. 4B).

Open in a separate window

FIG. 4

GPR119 activation increases insulin secretion from WT and DIRKO islets. A, Arg stimulation increases plasma insulin levels in C57BL/6 and DIRKO mice in vivo. Mice were fasted overnight, plasma was collected to assess fasting insulin levels, and a single ip dose of either Arg (1 mg/g) or vehicle was administered at time zero. Blood was collected from the tail vein 5 min later for determination of plasma insulin levels. B, GPR119 and Irs2 gene expression in WT and IRKO islets. GPR119 and Irs2 mRNA levels were measured by real-time PCR in islets isolated from age-matched WT, Gipr−/−, Glp1r−/−, and DIRKO mice and normalized to levels of 18s RNA in the same samples. C, Insulin secretion from WT and DIRKO islets. Islets were isolated from 10-wk-old C57L/6 and DIRKO mice and incubated under conditions of low glucose (LG) (2.5 mM) for 1 h. Islets were then incubated in low glucose or high glucose (HG) (16.6 mM) and treated for 1 h with either vehicle (DMSO), PACAP, exendin-4, or AR231453. Statistical analysis was assessed by ANOVA. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

We next investigated whether AR231453 directly stimulates insulin secretion from cultured murine islets in vitro. Glucose and PACAP significantly increased insulin secretion in both WT and DIRKO islets (Fig. 4C), whereas the GLP-1R agonist exendin-4 significantly increased insulin secretion in WT but not in DIRKO islets (Fig. 4C). AR231453 significantly increased insulin secretion in a glucose-dependent manner in both WT and DIRKO islets (Fig. 4C). Hence, these observations demonstrate that the diminished glucoregulatory response of DIRKO mice to GPR119 activation in vivo is not simply due to a generalized defect in the responsivity of DIRKO β-cells to insulin secretagogues but likely reflects loss of incretin action in the context of GPR119 activation. Moreover, the inability of AR231453 to reduce glucose excursion during an ip glucose tolerance test (IPGTT) in WT mice is not explained by an inability of murine β-cells to directly respond to this GPR119 agonist.

GPR119 controls gut motility through mechanisms independent of the GLP-1R, GLP-2R, or Y2 receptor (Y2R)

Because GLP-1 regulates glucose homeostasis through control of gut motility (27, 28), and GPR119 activation robustly increases circulating levels of GLP-1 (Fig. 2), we examined gastric emptying after administration of AR231453 in WT mice. AR231453 significantly reduced gastric emptying in WT mice (Fig. 5A). To determine whether the effects of AR231453 on gastric emptying were mediated through a GLP-1R-dependent mechanism, we repeated these studies in Glp1r−/− mice. Surprisingly, GPR119 activation continued to reduce gastric emptying in both Glp1r−/− and in DIRKO mice (Fig. 5A), suggesting that AR231453 controls gastric emptying through a mechanism that does not require the GLP-1R or GIP receptor.

Open in a separate window

FIG. 5

Gastric emptying in WT, Glp1r−/−, DIRKO, Gpr119−/−, and Glp2r−/− mice. Age-matched littermate WT and KO male mice were fasted overnight and given a single dose of AR231453 (20 mg/kg) or (A–C) vehicle, or (B) exendin-4 (1ug) before oral administration of a solution of glucose 15% and acetaminophen 1% at a dose of 1.5 g/kg glucose-0.1 g/kg acetaminophen. Gastric emptying rate was determined as described in Research Design and Methods. AUC for plasma acetaminophen levels in (A) WT, Glp1r−/−, and DIRKO mice, (B) Gpr119+/+ and Gpr119−/− mice, and (C) Glp2r+/+ and Glp2r−/− mice. Statistical analysis was assessed by ANOVA and paired t test (as appropriate); *, P < 0.05; **, P < 0.01.

Previous studies of AR231453 demonstrated that this molecule is a highly selective agonist that required the GPR119 receptor for control of glucose tolerance and insulin secretion (7). To assess whether the inhibitory actions of AR231453 on gastric emptying were also mediated through GPR119, or reflected GPR119-independent mechanism(s) of action, we examined gastric emptying in GPR119−/− mice. The GLP-1R agonist exendin-4 potently reduced gastric emptying in both GPR119+/+ and GPR119−/− mice (Fig. 5B). In contrast, AR231453 reduced gastric emptying in GPR119+/+ but not in GPR119−/− mice (Fig. 5B). Hence, GPR119 transduces the actions of AR231453 on gastric emptying.

Because AR231453 likely increases plasma levels of the related L cell-derived peptide GLP-2, and GLP-2R activation may also inhibit the rate of gastric emptying (29), we assessed gastric emptying in Glp2r+/+ and Glp2r−/− mice. AR231452 reduced gastric emptying in both Glp2r+/+ and Glp2r−/− mice (Fig. 5C), demonstrating that the GLP-2R is not the sole dominant mechanism transducing the AR231453-dependent inhibition of gastric emptying (Fig. 5C).

GPR119 activation stimulates local release of PYY from isolated gut mucosa (18), and PYY also inhibits gastric emptying (30); hence, we examined the effect of acute GPR119 activation on plasma PYY. AR231453 administration before oral glucose loading significantly increased plasma PYY in WT mice (Fig. 6A). Accordingly, we assessed whether the actions of AR231453 on gastric emptying were diminished by concomitant administration of the Y2R antagonist, BIIE0246, a reagent previously shown to abrogate the effects of PYY on gut motility (30). AR231453 significantly inhibited gastric emptying despite concomitant administration of BIIE0246 (Fig. 6B), suggesting that the GPR119-mediated control of gut motility is not mediated through the Y2R.

Open in a separate window

FIG. 6

GPR119 activation increases plasma PYY but inhibits gastric emptying independent of the Y2R. Age-matched male mice were fasted overnight and oral AR231453 (20 mg/kg) or vehicle was administered 30 min before an oral glucose tolerance test (OGTT). Blood samples were collected 5 min after glucose for the measurement PYY plasma levels (A). For gastric emptying (B), 10-wk-old C57Bl6 male mice were fasted overnight. At time zero, a single ip dose of either vehicle (V1 = 1.2% DMSO in water) or BIIE0246 (2 mg/kg) was administered 15 min before a single oral dose of either vehicle (V2 = 80% PEG400, 10% Tween 80, and 10% ethanol) or AR231453 (20 mg/Kg). A solution of glucose (15%)-acetaminophen (1%) was administered orally 30 min later at a dose of 1.5 g/kg glucose-0.1 g/kg acetaminophen, and blood samples were collected at 15 and 45 min for assessment of plasma acetaminophen levels as described in Research Design and Methods. Rate of appearance of acetaminophen in plasma as determined by the AUC (B). Statistical analysis was assessed by t test and ANOVA; *, P < 0.05.

Discussion

Studies of the biological activity of GPR119 using synthetic receptor agonists initially focused on activity of this receptor in islet cells, demonstrating potent stimulation of insulin release from β-cells in vitro, and in rodents in vivo (7). The complexity of GPR119 biology was further expanded to encompass gut peptide action, because pharmacological GPR119 activation also increased levels of both GIP and GLP-1 in mice and rats (11, 12, 14, 17). Furthermore, endogenous GPR119 appears essential for maximal GLP-1 secretion, because GPR119−/− mice exhibited reduced levels of circulating GLP-1 after nutrient or glucose ingestion (17). These findings raised the possibility that one or both incretin peptides contribute to the glucose lowering actions of GPR119 in the postprandial state.

We have now examined the relative importance of three distinct components of GPR119 action that might contribute to the glucoregulatory actions of GPR119 agonists, namely 1) incretin peptides, 2) a direct insulinotropic role for GPR119 in islets, and 3) the control of gastric emptying. Our data are consistent with a model invoking all three mechanisms as contributing to the totality of GPR119 action in vivo. We demonstrate that although the glucoregulatory actions of AR231453 are diminished in DIRKO mice, AR231453 continues to significantly increase plasma insulin levels concomitant with oral glucose loading in DIRKO mice. Hence, an intact incretin axis is not required for the insulinotropic actions of GPR119 agonists in vivo. The failure of AR231453 to lower glycemic excursion in DIRKO mice despite increasing insulin levels may be explained in part by the lack of GIP and GLP-1 action together with increased plasma levels of glucagon observed in the same experiments.

DIRKO mice exhibit modest defects in β-cell proliferation and impaired up-regulation of insulin gene expression and insulin secretion in response to high fat feeding (19); hence, one explanation for the relatively reduced insulinotropic response to AR231453 in DIRKO mice invokes a potential defect rendering DIRKO islets unable to respond adequately to GPR119 or other insulin secretagogues (19). Several lines of evidence argue against this interpretation of the data. First, GPR119 expression was comparable in RNA isolated from WT, single IRKO, and DIRKO islets. Furthermore, Arg briskly increased plasma insulin levels in both WT and DIRKO mice in vivo. Moreover, DIRKO islets exhibited a significant insulin secretory response to 1) AR231453 and 2) PACAP in vitro. Hence, the available data clearly demonstrate that GPR119 activation likely promotes insulin secretion via two complementary mechanisms: a direct β-cell effect and by activation of the incretin axis.

Interpretation of data from IRKO mice must be also tempered by the realization that these mice exhibit compensatory adaptations in levels of circulating incretins and in incretin responsivity (31, 32). Indeed, we observed a more robust induction of plasma GIP levels with AR231453 in the absence of a functional GLP-1R and increased plasma levels of GLP-1 in Glp1r−/− and DIRKO mice. These findings are consistent with previous studies suggesting that enteroendocrine cells regulate peptide hormone secretion in response to ambient levels of circulating gut hormones (33). Nevertheless, as previous (34) and more recent (35) studies have demonstrated that classical antagonists used to disrupt GLP-1 action, such as exendin (9-39), may not be completely specific for the GLP-1R, the use of mice with genetic disruption of IR genes provides a valuable complementary model for assessing the importance of IR signaling for glucoregulatory mechanisms activated by engagement of the GPR119 receptor.

Previous studies of GPR119 action have not addressed whether glucagon contributes to the glucoregulatory properties observed with GPR119 agonists. GPR119 has been predominantly localized to β-cells, and in one report to PP cells (36), but not α-cells. Although AR231453 had no effect on levels of plasma glucagon in WT mice, circulating glucagon levels were surprisingly increased after AR231453 administration in single IRKO and DIRKO mice. Basal GLP-1R signaling is known to tonically inhibit glucagon secretion, and even transient blockade of the GLP-1R using the antagonist exendin (9-39) increases levels of plasma glucagon after oral glucose challenge in normal and diabetic subjects (37, 38). Hence, it is possible that loss of the suppressive component of GLP-1 action on α-cells in Glp1r−/− and DIRKO mice unmasks a previously undetected action of GPR119, directly or indirectly, on the α-cell.

Nevertheless, there is very little data demonstrating that endogenous GIP action is critical for control of glucagon secretion, and it remains unclear how loss of the Gipr also leads to enhanced glucagon secretion after GPR119 activation in the Gipr−/− mouse. Hence, further exploration of the connection between GPR119 signal transduction and mechanisms regulating α-cell secretion appears warranted.

Considerable previous experimentation supports the hypothesis that oleoylethanolamide (OEA), an endogenous fatty acid derivative, functions as an endogenous agonist at the GPR119 receptor (8, 11). Intriguingly, GPR119 expression has been detected in the central nervous system, and OEA reduces food intake (8), raising the possibility that GPR119 functions as a component of a satiety circuit controlling body weight. Similarly, a small molecule GPR119 agonist, PSN632408, also suppressed food intake and reduced body weight in rodents (8). More recent findings illustrate that the satiety effects of OEA are complex and require cluster of differentiation 36 and per-oxisome proliferator-activated receptor α (39). Moreover, OEA retains its anorectic actions in Gpr119−/− mice (17). In contrast, we found no effect of GPR119 activation on food intake using a selective GPR119 agonist in WT, single IRKO, or DIRKO mice (data not shown). These findings are consistent with the normal body weight previously reported for Gpr119−/− mice (7, 17) and suggest that pharmacological activation or genetic disruption of GPR119 does not produce a phenotype linked to disordered control of energy homeostasis.

Because GPR119 activation significantly increases proglucagon-derived peptide secretion from the gut L cell and GLP-1, oxyntomodulin and GLP-2 all inhibit gastric emptying (2), we hypothesized that enhanced L-cell secretion might be associated with reduced gastric emptying. Surprisingly, however, there is little previous data examining whether selective GPR119 activation controls gastric emptying. We found that AR231453 significantly inhibited gastric emptying in WT mice. Unexpectedly, AR231453 also inhibited gastric emptying in Glp1r−/−, DIRKO, and Glp2r−/− mice, demonstrating that the inhibitory effect of GPR119 agonists on gut motility is not strictly dependent on the GLP-1, GIP, or GLP-2 receptors. Nevertheless, the ability of AR231453 to reduce gastric emptying was lost in Gpr119−/− mice, emphasizing that the inhibition of gastric emptying does not reflect an “off target” mechanisms of action of AR231453 and clearly requires a functional GPR119 signaling system. Consistent with our findings, the GPR119 ligand OEA also inhibits gastric emptying in mice, through poorly understood mechanisms independent of peroxisome proliferator-activated receptor α or cannabinoid receptors (40). Moreover, consistent with our findings using AR231453 in Glp1r−/− mice, the effects of OEA on gut motility were not diminished by coadministration of the GLP-1R antagonist exendin(9-39) (40).

More recent experimentation has delineated a role for PYY as a downstream target for GPR119 regulation in the gastrointestinal mucosa. GPR119 activation using the agonist PSN632408 inhibited epithelial electrolyte secretion in a PYY- dependent and Y1 receptor-dependent manner (18). Unexpectedly, the glucoregulatory and insulinotropic actions of PSN632408 were also attenuated in Pyy−/− mice. Our data extend these findings by demonstrating that AR231453 increased plasma levels of PYY in association with enteral glucose loading in vivo. Nevertheless, although exogenous PYY is known to inhibit gastric emptying (41), the inhibitory effects of AR231453 on gastric emptying were not diminished by coadministration of the Y2R antagonist BIIE0246. Because multiple gut peptides inhibit gastric emptying, and many of these (GLP-1, GLP-2, oxyntomodulin, PYY, etc.) are downstream targets of GPR119, simply inhibiting the action of a single peptide is unlikely to completely reverse the inhibition of gastric emptying observed with GPR119 agonists. Hence, the precise mediators and mechanisms coupling GPR119 activation to control of gut motility require further study.

In summary our data illustrate that GPR119 controls enteral glucose tolerance through at least three distinct yet complementary mechanisms. First, activation of the β-cell GPR119 directly enhances insulin secretion in isolated murine islets independent of IRs. Moreover, we identify a role for GPR119 in the control of gastric emptying that is not completely dependent on the GLP-1R, GLP-2R, or Y2R. Finally, we demonstrate that functional IRs are required for maximal control of oral glucose tolerance after GPR119 activation. Taken together, it seems likely that multiple enteroendocrine cell-derived peptides simultaneously contribute to the glucoregulatory signals and control of gastric emptying emanating from GPR119 activation in the gut. These findings may have implications for optimization of therapeutic strategies and the safety of employing GPR119 agonists for the treatment of type 2 diabetes.

Supplementary Material

Acknowledgments

Abbreviations

Footnotes

References