26RFa is a hypothalamic neuropeptide that promotes food intake. 26RFa is upregulated in obese animal models, and its orexigenic activity is accentuated in rodents fed a high-fat diet, suggesting that this neuropeptide might play a role in the development and maintenance of the obese status. As obesity is frequently associated with type 2 diabetes, we investigated whether 26RFa may be involved in the regulation of glucose homeostasis. In the current study, we show a moderate positive correlation between plasma 26RFa levels and plasma insulin in patients with diabetes. Plasma 26RFa concentration also increases in response to an oral glucose tolerance test. In addition, we found that 26RFa and its receptor GPR103 are present in human pancreatic β-cells as well as in the gut. In mice, 26RFa attenuates the hyperglycemia induced by a glucose load, potentiates insulin sensitivity, and increases plasma insulin concentrations. Consistent with these data, 26RFa stimulates insulin production by MIN6 insulinoma cells. Finally, we show, using in vivo and in vitro approaches, that a glucose load induces a massive secretion of 26RFa by the small intestine. Altogether, the present data indicate that 26RFa acts as an incretin to regulate glucose homeostasis.
Introduction
26RFa and its N-extended form 43RFa (also referred to as QRFPs) are RFamide peptides discovered in the brain of various vertebrate species and identified as the cognate ligands of the human orphan G-protein–coupled receptor GPR103 (1–6). Neuroanatomical observations revealed that 26RFa- and GPR103-expressing neurons are primarily localized in hypothalamic nuclei involved in the control of feeding behavior (1,2,5,7). Indeed, intracerebroventricular administration of 26RFa or 43RFa stimulates food intake (1,5,8,9), and the neuropeptide exerts its orexigenic activity by modulating the neuropeptide Y/proopiomelanocortin system in the arcuate nucleus (9). Chronic injection of 43RFa induces a significant increase in mice body weight and fat mass, which is associated with a hyperphagic behavior (8), and the orexigenic activities of 26RFa and 43RFa are more robust in rodents fed a high-fat diet (8,10). Consistent with these observations, expression of prepro26RFa mRNA is upregulated in the hypothalamus of genetically obese ob/ob and db/db mice and rodents subjected to a high-fat diet (5,10). Altogether, these observations support the notion that 26RFa could play a role in the development and maintenance of the obese status.
Obesity is frequently associated with type 2 diabetes, which is characterized by chronic hyperglycemia induced by impaired insulin secretion and increased insulin resistance (11–13). Accumulating evidence supports the peripheral role of neuropeptides controlling feeding behavior such as neuropeptide Y, orexins, ghrelin, corticotropin-releasing factor, or apelin in the regulation of glucose homeostasis (14–18), coining the new concept that hypothalamic neuropeptides may serve as a link between energy and glucose homeostasis and identifying them therefore as potential therapeutic targets for the treatment of diabetes and obesity (19,20).
In the current study, we investigate the possible involvement of 26RFa in the regulation of glucose homeostasis. We notably show an increase in plasma 26RFa levels in patients with diabetes and during an oral glucose tolerance test. We also found that 26RFa attenuates glucose-induced hyperglycemia, potentiates insulin sensitivity, and increases plasma insulin concentrations and that a glucose challenge induces a massive secretion of 26RFa by the gut.
Research Design and Methods
Patients
Subjects (n = 161) were recruited in the Department of Endocrinology, Diabetes and Metabolic Diseases of the University Hospital of Rouen. Patients were divided into four groups according to their weight and glucose tolerance. Obese patients (BMI >30 kg/m2) and patients with type 1 and type 2 diabetes (DT1 and DT2) were defined according to the American Diabetes Association criteria. Clinical examination did not reveal any abnormalities. Healthy subjects (HS) underwent standard endocrine tests to exclude any metabolic abnormalities. All of the subjects included in the study gave written informed consent according to our ethics committee instructions.
Metabolic Parameters in Humans
Plasma 26RFa levels were measured using a specific radioimmunoassay (21). Blood glucose was measured using a glucose oxidase activity test (LX20; Beckman Coulter, Villepinte, France). HbA1c was analyzed by high-performance liquid chromatography (G7 HPLC Analyzer; Tosoh, Lyon, France). Plasma hormone levels were measured using the following assays: plasma insulin (Elecys for automated insulin assay COBAS 6000CE; Roche Diagnostics, Meylan, France), glucose-dependent insulinotropic polypeptide (GIP) (Elisa-assay; Millipore, St. Charles, MO), and glucagon-like peptide 1 (GLP-1) (Elisa Epitope Diagnostics, San Diego, CA). Plasma ghrelin was assayed using a commercial kit (cat. no. EK-031-30; Phoenix, Belmont, CA).
HOMA of insulin resistance (HOMA-IR index) was calculated as the ratio [fasting glucose (mmol · L−1) × fasting insulin (μU · mL−1)]/22.5 (22).
Oral Glucose Tolerance Test in Humans
Nine healthy volunteers recruited for a previous study (23), one patient with gastroparesis, and another with a total gastrectomy underwent a 75-g oral glucose tolerance test (OGTT) after a 12-h fasting period. 26RFa, glucose, insulin, GIP, GLP-1, and ghrelin concentrations were measured in fasting blood samples obtained at 0, 30, 60, 120, 150, and 180 min after oral glucose loading.
Immunohistochemical Procedure
Deparaffinized sections (15 µm thick) of human and mouse tissues were used for immunohistochemistry. All the tissue procurement protocols (for humans) were approved by the relevant institutional committees (University of Rouen) and were undertaken under informed consent of all of the participants. Tissue sections were incubated for 1 h at room temperature with either rabbit polyclonal antibodies against 26RFa (24) diluted 1:400, GPR103 (cat. no. NLS1922; Novus Biologicals, Littletown, CO) diluted 1:100, EM66 (25) diluted 1:200, or mouse monoclonal antibodies against insulin (Sigma-Aldrich, Saint-Quentin Fallavier, France) diluted 1:1,000. The sections were incubated with a streptavidin-biotin-peroxidase complex (Dako Corporation, Carpinteria, CA), and the enzymatic activity was revealed with diaminobenzidine. The slices were then counterstained with hematoxylin. Observations were made under a Nikon E 600 light microscope.
Blood Glucose and Insulin Measurements in Mice
For intraperitoneal glucose tolerance test and glucose-stimulated insulin secretion assay, mice (male C57BL/6, 20–23 g) were fasted for 16 h with free access to water and then injected intraperitoneally with glucose (2 g/kg) and 26RFa (500 μg/kg) or PBS solution. For measurements of basal glycemia, mice were fed ad libitum and injected intraperitoneally with 26RFa (500 μg/kg) or PBS solution. Blood plasma glucose concentrations were measured from tail vein samplings at various times using an AccuChek Performa glucometer (Roche Diagnostic, Saint-Egreve, France). Plasma insulin concentrations were determined using an ultrasensitive mouse insulin ELISA kit (Mercodia, Uppsala, Sweden). For insulin tolerance test, mice were fasted for 6 h and injected intraperitoneally with 0.75 units/kg body wt human insulin (Eli Lilly, Neuilly-sur-Seine, France) and 26RFa (500 μg/kg) or PBS. For intravenous glucose tolerance test, mice were catheterized in the tail vein under anesthesia 24 h prior to the test. For OGTT and intravenous glucose tolerance test, mice were fasted for 16 h before the test with free access to water. Plasma samples were obtained from decapitation 30 or 120 min after a 2-g/kg oral glucose load or 3 or 30 min after a 1-g/kg i.v. glucose load and assayed for 26RFa.
Insulin Secretion by MIN6 Cells
Mouse insulinoma cells (MIN6), a kind gift from Dr J. Miyazaki, of Osaka University, Osaka, Japan, were grown in DMEM containing 25 mmol/L glucose and supplemented with 15% heat-inactivated FBS in a humidified atmosphere of 5% CO2, 95% air, at 37°C (26). Before the experiments, the culture medium was removed and MIN6 cells (3 × 105 cells/well) were preincubated in a Krebs-Ringer bicarbonate buffer containing 0.2% BSA and 2.8 mmol/L glucose (low glucose [LG]) for 1 h at 37°C (period P1) to evaluate basal insulin secretion. The culture medium was then removed and replaced by the same culture medium (LG) or a culture medium with 16.7 mmol/L glucose (high glucose [HG]) added or not with 26RFa and the GLP-1 analog exenatide (10−6 mol/L each) for 1 h (period P2). Insulin secretion in each well was expressed as the ratio between secretion during the P2 period and secretion during the P1 period.
Quantitative PCR
Total RNA from various tissues of mice and from MIN6 cells was isolated as previously described (27). Relative expression of the 26RFa, GPR103, NPFF2 receptor (NPFF2), insulin receptor (INS-R), GLUT2, and GLUT4 genes was quantified by real-time PCR with appropriate primers (Table 1). GAPDH or β-actin was used as internal control for normalization. PCR was carried out using Gene Expression Master Mix 2X assay (Applied Biosystems, Courtaboeuf, France) in an ABI Prism 7900 HT Fast Real-time PCR System (Applied Biosystems). The purity of the PCR products was assessed by dissociation curves. The amount of target cDNA was calculated by the comparative threshold (Ct) method and expressed by means of the 2-ΔΔCt method.
Sequences of the primers used for the Q-PCR experiments
. | Forward primer . | Reverse primer . |
---|---|---|
Mouse β-actin | AGGTCATCACTATTGGCAACGA | CACAGGATTCCATACCCAAGAAG |
Mouse 26RFa | GAAGGGGACCCACAGACATC | GTCTTGCCTCCCTAGACGGAA |
Mouse GPR103 | CACTGTTGTGACGGAAAT | CCTTCGGGTAGTGTACTGCC |
Mouse INS-R | ATGGGCTTCGGGAGAGGAT | GGATGTCCATACCAGGGCAC |
Mouse GLUT2 | CCCTGTTCCTAACCGGGATG | GGCGAATTTATCCAGCAGCAC |
Mouse GLUT4 | GTGACTGGAACACTGGTCCTA | CCAGCCACGTTGCATTGTAG |
Mouse NPFF2 | ATCTGGAGTGGCAATGATACACA | TCCCACCATGCACAAGACAAA |
Mouse GAPDH | TCCGGACGCACCCTCAT | CGGTTGACCTCCAGGAAATC |
. | Forward primer . | Reverse primer . |
---|---|---|
Mouse β-actin | AGGTCATCACTATTGGCAACGA | CACAGGATTCCATACCCAAGAAG |
Mouse 26RFa | GAAGGGGACCCACAGACATC | GTCTTGCCTCCCTAGACGGAA |
Mouse GPR103 | CACTGTTGTGACGGAAAT | CCTTCGGGTAGTGTACTGCC |
Mouse INS-R | ATGGGCTTCGGGAGAGGAT | GGATGTCCATACCAGGGCAC |
Mouse GLUT2 | CCCTGTTCCTAACCGGGATG | GGCGAATTTATCCAGCAGCAC |
Mouse GLUT4 | GTGACTGGAACACTGGTCCTA | CCAGCCACGTTGCATTGTAG |
Mouse NPFF2 | ATCTGGAGTGGCAATGATACACA | TCCCACCATGCACAAGACAAA |
Mouse GAPDH | TCCGGACGCACCCTCAT | CGGTTGACCTCCAGGAAATC |
Small Interfering RNA
Mouse Qrfpr Silencer Select Pre-designed small interfering (si)RNA (SC-60730) (GPR103 siRNA[m]) and Silencer Negative Control siRNA (SC-37007) were purchased from Santa Cruz Biotechnology (Dallas, TX), and AMAXA cell line nucleofector KIT-V was purchased from Lonza (Basel, Switzerland). MIN6 cells were transfected with 10 μmol/L control or GPR103 siRNAs using AMAXA cell line nucleofector KIT-V, according to the manufacturer's instructions. Forty-eight hours after transfection, cells were subjected to quantitative PCR (Q-PCR) and insulin secretion experiments. Efficiency of transfection was assessed by RT-PCR.
Expression and Secretion of 26RFa by Intestine Fragments
Three consecutive mice samples of duodenum, proximal jejunum, and ileum were collected. For each intestinal segment, one sample was immediately frozen in liquid nitrogen for Q-PCR experiments and two samples were mounted in Ussing chambers with an exchange surface of 0.07 cm2 as previously described (28). Glucose-free DMEM (1 mL) was applied at the apical and serosal sides of the intestinal segments. At the apical side, DMEM supplemented with either 2.8 mmol/L or 16.7 mmol/L glucose was applied. At the serosal side, only DMEM with 3 mmol/L glucose was applied. After 3 h at 37°C, media from apical and serosal sides were collected and assayed for 26RFa.
Statistical Analysis
Statistical analysis was performed with Statistica (5th version). A Student t test or ANOVA for repeated measures was used for comparisons between two groups. A post hoc comparison using a Tukey honestly significant difference test was applied according to ANOVA results. Statistical significance was set at P < 0.05.
Results
Fasting Plasma 26RFa Levels in Obese Patients and Patients With Diabetes
The phenotypic characteristics of the four groups of subjects studied are reported in Fig. 1A. Circulating levels of 26RFa in the fasting condition were significantly enhanced in obese patients (466 ± 37 pg/mL) and DT2 subjects (488 ± 48 pg/mL) versus HS (338 ± 42 pg/mL; P < 0.05) (Fig. 1B). DT1 patients showed plasma 26RFa levels (494 ± 68 pg/mL) similar to those of DT2 subjects (Fig. 1B). No correlation was found between plasma 26RFa levels and age, BMI, waist circumference, and fasting blood glucose. Conversely, a moderate positive correlation was found between circulating 26RFa and fasting plasma insulin (r = 0.37, P = 0.0015) (Fig. 1C) and the insulin resistance marker HOMA-IR (r = 0.54, P < 0.0001) (Fig. 1D).
Evolution of plasma 26RFa concentrations in obese patients and in patients with diabetes. A: Phenotypes of the patients included in the study. B: Plasma 26RFa levels were measured, using a specific radioimmunoassay for human 26RFa, in the four groups of patients after an overnight fast. Obese patients without or with type 2 diabetes show a significant increase in plasma 26RFa concentrations (P < 0.05) compared with HS. Patients with DT1 also show an increase in circulating 26RFa that does not reach the level of significance (P = 0.057). Data represent means ± SEM of five independent experiments. *P < 0.05. C and D: Correlation analysis revealed that plasma 26RFa is positively correlated with fasting insulin (r = 0.37, P = 0.0015) (C) and the insulin resistance marker HOMA-IR (r = 0.54, P < 0.0001) (D). yr, years.
Evolution of plasma 26RFa concentrations in obese patients and in patients with diabetes. A: Phenotypes of the patients included in the study. B: Plasma 26RFa levels were measured, using a specific radioimmunoassay for human 26RFa, in the four groups of patients after an overnight fast. Obese patients without or with type 2 diabetes show a significant increase in plasma 26RFa concentrations (P < 0.05) compared with HS. Patients with DT1 also show an increase in circulating 26RFa that does not reach the level of significance (P = 0.057). Data represent means ± SEM of five independent experiments. *P < 0.05. C and D: Correlation analysis revealed that plasma 26RFa is positively correlated with fasting insulin (r = 0.37, P = 0.0015) (C) and the insulin resistance marker HOMA-IR (r = 0.54, P < 0.0001) (D). yr, years.
OGTT in Humans
An OGTT with 75 g glucose was performed in nine healthy volunteers. Blood glucose and insulin as well as GLP-1 and GIP showed a rapid elevation 30 min after the glucose load (Fig. 2A and B). By contrast, plasma 26RFa levels were stable during the first 90 min of the test but increased significantly 120 min after the glucose load (P < 0.01) (Fig. 2A and B). The amplitude of the rises was much higher for GLP-1 (423 ± 57%) and GIP (887 ± 229%) than for 26RFa (182 ± 23%) (Fig. 2B). Ghrelin showed a plasma profile totally different from that of 26RFa with a rapid decline at 30 min, and then levels remained stable until the end of the test (Fig. 2B). In the patient with gastroparesis, the 26RFa peak was delayed compared with the HS group (210 min vs. 120 min) (Fig. 2C). Conversely, in the gastrectomized patient, the 26RFa rise occurred earlier than in the healthy volunteers (90 min vs. 120 min) (Fig. 2D).
Plasma 26RFa profile during an OGTT in humans. A: Plasma glucose, insulin, and 26RFa levels were measured in nine healthy volunteers during 180 min after a 75-g glucose load, with samplings every 30 min. As expected, a peak of plasma glucose and insulin occurs 30 min after the glucose load. In contrast, plasma 26RFa levels remain stable during the first 90 min of the test but increase dramatically at 120 min and then decrease slowly until the end of the experiment. **P < 0.01. B: Comparison of the plasma 26RFa profile during the OGTT with those of the two incretins GLP-1 and GIP and with that of the other orexigenic neuropeptide, ghrelin. A quick rise of GLP-1 and GIP is observed 30 min after the glucose load, similar to those of glucose and insulin, in contrast to 26RFa, which peaks only at 120 min. Ghrelin levels decreased at 30 min and then remained stable until the end of the test. C: Plasma glucose, insulin, and 26RFa levels were measured every 30 min during 300 min in a patient with gastroparesis. Plasma glucose rises at 30 min, whereas the insulin peak occurs 90 min after the glucose load. Plasma 26RFa levels increase only 210 min after the glucose challenge. D: Plasma glucose, insulin, and 26RFa levels were measured during 180 min, with samplings every 30 min, in a subject who underwent a total gastrectomy. Plasma glucose and insulin levels increase at 60 min after the glucose load. The increase in plasma 26RFa occurs 90 min after the glucose challenge. Data represent the values of a single patient for C and D.
Plasma 26RFa profile during an OGTT in humans. A: Plasma glucose, insulin, and 26RFa levels were measured in nine healthy volunteers during 180 min after a 75-g glucose load, with samplings every 30 min. As expected, a peak of plasma glucose and insulin occurs 30 min after the glucose load. In contrast, plasma 26RFa levels remain stable during the first 90 min of the test but increase dramatically at 120 min and then decrease slowly until the end of the experiment. **P < 0.01. B: Comparison of the plasma 26RFa profile during the OGTT with those of the two incretins GLP-1 and GIP and with that of the other orexigenic neuropeptide, ghrelin. A quick rise of GLP-1 and GIP is observed 30 min after the glucose load, similar to those of glucose and insulin, in contrast to 26RFa, which peaks only at 120 min. Ghrelin levels decreased at 30 min and then remained stable until the end of the test. C: Plasma glucose, insulin, and 26RFa levels were measured every 30 min during 300 min in a patient with gastroparesis. Plasma glucose rises at 30 min, whereas the insulin peak occurs 90 min after the glucose load. Plasma 26RFa levels increase only 210 min after the glucose challenge. D: Plasma glucose, insulin, and 26RFa levels were measured during 180 min, with samplings every 30 min, in a subject who underwent a total gastrectomy. Plasma glucose and insulin levels increase at 60 min after the glucose load. The increase in plasma 26RFa occurs 90 min after the glucose challenge. Data represent the values of a single patient for C and D.
Localization of 26RFa and GPR103 in the Human Pancreas and Gastrointestinal Tract
Treatment of human pancreas sections with the 26RFa and GPR103 antibodies revealed that the neuropeptide and its receptor were specifically localized in the endocrine islets and were present within the same islets (Fig. 3A and B). Treatment of consecutive sections with the 26RFa or GPR103 antibodies and the insulin antibodies revealed that the histological distribution of 26RFa- and GPR103-like immunoreactivity (LI) was similar to that of insulin (Fig. 3C–F).
Immunohistochemical distribution of 26RFa and its receptor, GPR103, in the human pancreas and gastrointestinal tract. A and B: Photomicrographs of pancreas sections showing that 26RFa and its receptor are specifically localized in the endocrine islets and are present within the same islets. C–F: Photomicrographs of pancreas consecutive sections treated with 26RFa or GPR103 and insulin antibodies revealing that the distribution of 26RFa and its receptor is similar to that of insulin. G: Photomicrograph of the antral part of the stomach showing that numerous epithelial cells of the gastric glands are labeled with the 26RFa antibodies. H: Photomicrograph of the duodenum showing that numerous enterocytes and goblet cells of the villosities are labeled with the 26RFa antibodies. I and J: Photomicrographs of ileum (I) and colon (J) slices showing that the distribution of 26RFa is very similar to that observed in the duodenum with the exception that the number of immunoreactive cells is much lower than in the duodenum (arrows). Scale bars: A–F and H, 50 µm; G, 200 µm; I and J, 25 µm.
Immunohistochemical distribution of 26RFa and its receptor, GPR103, in the human pancreas and gastrointestinal tract. A and B: Photomicrographs of pancreas sections showing that 26RFa and its receptor are specifically localized in the endocrine islets and are present within the same islets. C–F: Photomicrographs of pancreas consecutive sections treated with 26RFa or GPR103 and insulin antibodies revealing that the distribution of 26RFa and its receptor is similar to that of insulin. G: Photomicrograph of the antral part of the stomach showing that numerous epithelial cells of the gastric glands are labeled with the 26RFa antibodies. H: Photomicrograph of the duodenum showing that numerous enterocytes and goblet cells of the villosities are labeled with the 26RFa antibodies. I and J: Photomicrographs of ileum (I) and colon (J) slices showing that the distribution of 26RFa is very similar to that observed in the duodenum with the exception that the number of immunoreactive cells is much lower than in the duodenum (arrows). Scale bars: A–F and H, 50 µm; G, 200 µm; I and J, 25 µm.
26RFa-LI was also investigated in the human gut. In the antral part of the stomach, numerous epithelial cells of the gastric glands were labeled with the 26RFa antibodies (Fig. 3G). In the duodenum, 26RFa-LI was detected in most of the enterocytes and goblet cells (Fig. 3H). In the ileum and the colon, the distribution of 26RFa-LI was very similar to that observed in the duodenum, although the number of 26RFa-labeled cells was much lower than in the duodenum (Fig. 3I and J).
Effect of 26RFa on Glucose Homeostasis in Mice
An IPGTT, performed in mice, revealed that 26RFa significantly attenuated (P < 0.01 and P < 0.001) the hyperglycemia induced by an intraperitoneal glucose challenge all throughout the duration of the test (Fig. 4A). Concurrently, during an insulin tolerance test, 26RFa significantly potentiated (P < 0.05 and P < 0.01) insulin-induced hypoglycemia between 30 min and 60 min after the insulin challenge (Fig. 4B). By contrast, 26RFa did not affect basal plasma glucose levels during the 90-min period of the test (Fig. 4C). Finally, an acute glucose-stimulated insulin secretion test revealed that intraperitoneal injection of 26RFa significantly stimulated (P < 0.05 and P < 0.001) glucose-induced insulin production (Fig. 4D).
Effects of 26RFa on plasma glucose and insulin levels in mice. A: Effect of administration of 26RFa (500 µg/kg i.p.) during a glucose tolerance test. 26RFa significantly attenuates the hyperglycemia induced by the injection of glucose (2 g/kg i.p.) throughout the duration of the test. B: Effect of administration of 26RFa (500 µg/kg i.p.) during an insulin tolerance test. 26RFa significantly potentiates insulin-induced hypoglycemia between 30 and 60 min after the insulin load (0.75 units/kg i.p.). C: Effect of administration of 26RFa (500 µg/kg i.p.) on basal plasma glucose levels. 26RFa does not alter basal plasma glucose during the 90 min after its injection. D: Effect of administration of 26RFa (500 µg/kg i.p.) during an acute glucose-stimulated insulin secretion test. 26RFa significantly stimulates glucose-induced insulin production 15 and 30 min after the glucose load (2 g/kg). Data represent means ± SEM of three independent experiments (n = 7/experiment). *P < 0.05; **P < 0.01; ***P < 0.001. T0, the beginning of the experiment before injection of the test substance.
Effects of 26RFa on plasma glucose and insulin levels in mice. A: Effect of administration of 26RFa (500 µg/kg i.p.) during a glucose tolerance test. 26RFa significantly attenuates the hyperglycemia induced by the injection of glucose (2 g/kg i.p.) throughout the duration of the test. B: Effect of administration of 26RFa (500 µg/kg i.p.) during an insulin tolerance test. 26RFa significantly potentiates insulin-induced hypoglycemia between 30 and 60 min after the insulin load (0.75 units/kg i.p.). C: Effect of administration of 26RFa (500 µg/kg i.p.) on basal plasma glucose levels. 26RFa does not alter basal plasma glucose during the 90 min after its injection. D: Effect of administration of 26RFa (500 µg/kg i.p.) during an acute glucose-stimulated insulin secretion test. 26RFa significantly stimulates glucose-induced insulin production 15 and 30 min after the glucose load (2 g/kg). Data represent means ± SEM of three independent experiments (n = 7/experiment). *P < 0.05; **P < 0.01; ***P < 0.001. T0, the beginning of the experiment before injection of the test substance.
Effect of 26RFa on Insulin Secretion by MIN6 Cells
We then searched for an eventual direct effect of 26RFa on β-cell insulin secretion. For this, we used the MIN6 cell line. Q-PCR experiments indicated that MIN6 cells expressed both 26RFa and GPR103 mRNA but not NPFF2, the other RFamide receptor that 26RFa can bind to, whatever the glucose concentration applied in the culture medium (Fig. 5A). GPR103, 26RFa, INS-R, and GLUT2 were also expressed in cells transfected with nonsilencing siRNA (Fig. 5B). Transfection of the cells with siRNA to GPR103 resulted in a total loss of GPR103 expression, whereas the expression of the other genes was not altered (Fig. 5B). Both 26RFa and exenatide (10−6 mol/L each) induced a highly significant increase of insulin release by siControl cells cultured in an LG medium. Incubation of the cells in an HG medium also resulted in a significant increase of insulin secretion (Fig. 5C). Addition of 26RFa in this latter condition of cell culture did not alter insulin production, whereas exenatide strongly stimulated insulin secretion (Fig. 5C). No additive effects of 26RFa and exenatide were observed in either the LG or HG condition (Fig. 5C). Invalidation of GPR103 expression resulted in a total loss of 26RFa-induced insulin secretion, whereas the stimulatory effect of exenatide in both LG and HG medium was maintained (Fig. 5D). Addition of 26RFa to exenatide did not significantly alter exenatide-induced insulin secretion (Fig. 5D).
Effect and expression of 26RFa on the MIN6 cell line. A and B: Expression of 26RFa, GPR103, NPFF2, INS-R, and GLUT2 mRNA was determined in native MIN6 cells (A) and in cells transfected with nonsilencing siRNA (siControl) or transfected with siRNA to GPR103 (siGPR103) (B) by Q-PCR and adjusted to the signal intensity of β-actin. Both 26RFa and GPR103 are expressed in native MIN6 cells grown in DMEM plus 25 mmol/L glucose, whereas NPFF2 mRNA is not detected (A). Incubation of the cells in a Krebs-Ringer medium complemented with 2.8 mmol/L glucose for insulin secretion experiments does not modify this expression profile (A). GPR103, 26RFa, INS-R, and GLUT2 are also expressed in cells transfected with nonsilencing siRNA (B). Transfection of the cells with siRNA to GPR103 results in a total loss of GPR103 expression, whereas the expression of the other genes is not altered (B). C and D: Secretion of insulin by siControl or siGPR103 MIN6 cells incubated with an LG or an HG medium in the presence or absence of 26RFa (10−6 mol/L) or the GLP-1 analog exenatide (E) (10−6 mol/L). In LG conditions, both 26RFa and exenatide induce a highly significant increase of insulin release by siControl cells (C). In the HG condition, 26RFa does not alter insulin release, whereas exenatide strongly stimulates insulin secretion (C). No additive effects of 26RFa and exenatide are observed in both LG and HG conditions (C). Invalidation of GPR103 expression results in a total loss of 26RFa-induced insulin secretion, whereas the positive effect of exenatide is not affected (D). Addition of 26RFa to exenatide did not significantly alter exenatide-induced insulin secretion in both LG and HG conditions (D). Insulin mean basal level/well was 88 ± 2 µU/mL. E: Expression of GPR103, INS-R, and GLUT4 mRNA was determined by Q-PCR and adjusted to the signal intensity of GAPDH in the muscle, the liver, and the adipose tissue. GPR103 mRNA is expressed in the three tissues, with higher levels detected in the liver. Expression of INS-R is significantly higher in the liver compared with the muscle and the adipose tissue. Expression of GLUT4 is not detected in the liver and is much lower in the adipose tissue compared with the muscle. Data represent means ± SEM of three independent experiments (n = 5 per experiment). **P < 0.01; ***P < 0.001. ns, nonsignificant.
Effect and expression of 26RFa on the MIN6 cell line. A and B: Expression of 26RFa, GPR103, NPFF2, INS-R, and GLUT2 mRNA was determined in native MIN6 cells (A) and in cells transfected with nonsilencing siRNA (siControl) or transfected with siRNA to GPR103 (siGPR103) (B) by Q-PCR and adjusted to the signal intensity of β-actin. Both 26RFa and GPR103 are expressed in native MIN6 cells grown in DMEM plus 25 mmol/L glucose, whereas NPFF2 mRNA is not detected (A). Incubation of the cells in a Krebs-Ringer medium complemented with 2.8 mmol/L glucose for insulin secretion experiments does not modify this expression profile (A). GPR103, 26RFa, INS-R, and GLUT2 are also expressed in cells transfected with nonsilencing siRNA (B). Transfection of the cells with siRNA to GPR103 results in a total loss of GPR103 expression, whereas the expression of the other genes is not altered (B). C and D: Secretion of insulin by siControl or siGPR103 MIN6 cells incubated with an LG or an HG medium in the presence or absence of 26RFa (10−6 mol/L) or the GLP-1 analog exenatide (E) (10−6 mol/L). In LG conditions, both 26RFa and exenatide induce a highly significant increase of insulin release by siControl cells (C). In the HG condition, 26RFa does not alter insulin release, whereas exenatide strongly stimulates insulin secretion (C). No additive effects of 26RFa and exenatide are observed in both LG and HG conditions (C). Invalidation of GPR103 expression results in a total loss of 26RFa-induced insulin secretion, whereas the positive effect of exenatide is not affected (D). Addition of 26RFa to exenatide did not significantly alter exenatide-induced insulin secretion in both LG and HG conditions (D). Insulin mean basal level/well was 88 ± 2 µU/mL. E: Expression of GPR103, INS-R, and GLUT4 mRNA was determined by Q-PCR and adjusted to the signal intensity of GAPDH in the muscle, the liver, and the adipose tissue. GPR103 mRNA is expressed in the three tissues, with higher levels detected in the liver. Expression of INS-R is significantly higher in the liver compared with the muscle and the adipose tissue. Expression of GLUT4 is not detected in the liver and is much lower in the adipose tissue compared with the muscle. Data represent means ± SEM of three independent experiments (n = 5 per experiment). **P < 0.01; ***P < 0.001. ns, nonsignificant.
Expression of GPR103 in Mouse Insulin-Sensitive Tissues
Q-PCR realized in insulin target tissues revealed that GPR103 mRNA was expressed in the striated muscle, the liver, and the adipose tissue, with higher levels in the liver (P < 0.001) (Fig. 5E). Expression of INS-R and GLUT4 was also investigated in the same tissues. Expression of INS-R was significantly higher (P < 0.001) in the liver compared with the muscle and the adipose tissue (Fig. 5E). Expression of GLUT4 was not detected in the liver and was much lower (P < 0.001) in the adipose tissue compared with the muscle (Fig. 5E).
Effect of Glucose on 26RFa Secretion by the Mouse Gastrointestinal Tract
We finally investigated whether glucose could induce 26RFa release from the gut in mice. As illustrated in Fig. 6A, 26RFa mRNA was actually expressed in the duodenum, jejunum, and ileum of the mouse, with higher levels in the duodenum (P < 0.05) (Fig. 6A). Concurrently, immunohistochemical experiments revealed the presence of intensely 26RFa-labeled cells in the mouse upper and lower intestine (Fig. 6B). In the duodenum, scattered positive cells were found in the epithelium (Fig. 6B, 1). In the ileum, the localization of clusters of 26RFa-positive cells suggested that these cells may correspond to Paneth cells (Fig. 6B, 2). Treatment of consecutive jejunum sections with either the 26RFa antibodies or the EM66 (a fragment of the chromogranin, secretogranin II) antibodies revealed that the two antibodies label the same cells (Fig. 6B, 3 and 4), indicating that the 26RFa-containing cells probably correspond to neuroendocrine cells. Oral administration of glucose induced a highly significant increase (P < 0.001) in plasma 26RFa levels 30 min after the glucose challenge that corresponded to the plasma glucose peak (Fig. 6C). By contrast, intravenous administration of glucose did not significantly alter plasma 26RFa levels all along the test (Fig. 6D). We also used a model of Ussing chambers in which we exposed the luminal (apical) side of intestine fragments to LG or HG concentrations, and we measured 26RFa levels in both the apical and basal sides of the intestine fragments. Duodenal fragments exposed to HG released massive amounts (P < 0.01) of 26RFa in the basal chamber compared with duodenum slices exposed to LG (Fig. 6E). In contrast, we did not observe any alteration of 26RFa flows in jejunum and ileum fragments treated with LG versus HG (Fig. 6E).
Expression and secretion of 26RFa by the gastrointestinal tract in mice. A: Expression of 26RFa mRNA was determined by Q-PCR and adjusted to the signal intensity of GAPDH in duodenal, jejunal, and ileal segments. 26RFa mRNA is expressed in the three regions of the mouse gastrointestinal tract with higher levels in the duodenum (n = 6). B: Immunohistochemical distribution of 26RFa in the mouse intestine. Intensely immunoreactive cells are found in the epithelium of the duodenum (1, arrows) and the ileum (2). In the ileum, 26RFa-positive cells are mostly found at the basis of the villosities and might thus correspond to Paneth cells (2, arrows). Treatment of consecutive jejunal sections with either the 26RFa (3) or the EM66 (4) antibodies, which are markers of neuroendocrine cells, indicates that the same cells are labeled by the two antibodies (3, 4, arrows). C and D: Evolution of plasma 26RFa levels during an OGTT (C) and an intravenous glucose tolerance test (D). A highly significant increase in plasma 26RFa concentration is observed 30 min after the oral administration of glucose that corresponds to the plasma glucose peak (C). Conversely, no significant alteration of plasma 26RFa levels is detected after intravenous administration of glucose (D) (n = 10 for each time point). E: Secretion of 26RFa from duodenal, jejunal, and ileal segments was evaluated in LG (2.8 mmol/L) or HG (16.7 mmol/L) conditions using a Ussing chambers model (n = 8). Duodenal fragments exposed to HG concentrations at their apical pole release massive amounts of 26RFa in the basal chamber compared with duodenal slices exposed to LG concentrations. In contrast, the 26RFa flow toward the basal compartment is not altered in the jejunum and the ileum in HG conditions versus LG conditions. Data represent means ± SEM of three independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001. ns, nonsignificant.
Expression and secretion of 26RFa by the gastrointestinal tract in mice. A: Expression of 26RFa mRNA was determined by Q-PCR and adjusted to the signal intensity of GAPDH in duodenal, jejunal, and ileal segments. 26RFa mRNA is expressed in the three regions of the mouse gastrointestinal tract with higher levels in the duodenum (n = 6). B: Immunohistochemical distribution of 26RFa in the mouse intestine. Intensely immunoreactive cells are found in the epithelium of the duodenum (1, arrows) and the ileum (2). In the ileum, 26RFa-positive cells are mostly found at the basis of the villosities and might thus correspond to Paneth cells (2, arrows). Treatment of consecutive jejunal sections with either the 26RFa (3) or the EM66 (4) antibodies, which are markers of neuroendocrine cells, indicates that the same cells are labeled by the two antibodies (3, 4, arrows). C and D: Evolution of plasma 26RFa levels during an OGTT (C) and an intravenous glucose tolerance test (D). A highly significant increase in plasma 26RFa concentration is observed 30 min after the oral administration of glucose that corresponds to the plasma glucose peak (C). Conversely, no significant alteration of plasma 26RFa levels is detected after intravenous administration of glucose (D) (n = 10 for each time point). E: Secretion of 26RFa from duodenal, jejunal, and ileal segments was evaluated in LG (2.8 mmol/L) or HG (16.7 mmol/L) conditions using a Ussing chambers model (n = 8). Duodenal fragments exposed to HG concentrations at their apical pole release massive amounts of 26RFa in the basal chamber compared with duodenal slices exposed to LG concentrations. In contrast, the 26RFa flow toward the basal compartment is not altered in the jejunum and the ileum in HG conditions versus LG conditions. Data represent means ± SEM of three independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001. ns, nonsignificant.
Discussion
Incretins are peptides released by the gut after ingestion of glucose or nutrients that act directly on pancreatic β-cells to stimulate insulin secretion and lower plasmaglucose. Until now, two incretins had been identified: GIP and GLP-1 (29). Here, we provide evidence that 26RFa, a neuropeptide of 26 amino acids, initially discovered as an orexigenic molecule in the hypothalamus (1), is produced in abundance in the gut and acts as an incretin hormone.
We first investigated plasma 26RFa levels in obese patients with or without type 2 diabetes. The two groups show higher concentrations of circulating 26RFa compared with healthy subjects. This finding is consistent with previous data obtained in obese animal models showing an upregulation of the 26RFa system in genetically obese mice and rats fed a high-fat diet (5,8,10) and suggests that, in humans as in rodents, upregulation of the 26RFa system may play a role in the development and maintenance of the obese status. However, high plasma 26RFa levels were also detected in patients with DT1 who were not obese. Correlation analysis revealed a moderate, but significant, positive correlation between plasma 26RFa and plasma insulin and the insulin resistance marker HOMA-IR. Altogether, these observations suggested a possible link between the 26RFa system and glucose homeostasis. We thus examined the evolution of plasma 26RFa levels during an OGTT in healthy subjects. Our data reveal a significant increase in plasma 26RFa during the test, indicating that an oral glucose load influences the concentrations of circulating 26RFa. Comparison of the plasma 26RFa profile with those of GIP and GLP-1 during the OGTT reveals that the 26RFa response to glucose challenge is delayed compared with those of the two incretins and that the amplitude of the 26RFa peak is much lower than those of GIP and GLP-1. This suggests that the mechanism of action of 26RFa to regulate glucose homeostasis is different from those of GIP and GLP-1 in humans.
Immunohistochemical examination of human pancreatic slices indicates that 26RFa and its receptor are present in endocrine islets but virtually absent in the exocrine pancreas. Our data also show that 26RFa and GPR103 are expressed by the insulin-producing β-cells, suggesting an effect of the neuropeptide on β-cell activity. Consistent with this hypothesis, a recent study reports that 26RFa and GPR103 are expressed in the pancreatic islets and that 26RFa and its N-extended form, 43RFa, prevent β-cell death and apoptosis (30). The current study shows that 26RFa and its receptor are present in the same pancreatic islets, indicating that 26RFa may regulate β-cell activity via an autocrine mechanism.
26RFa is a neuropeptide initially known to be produced by hypothalamic neurons (1,4,5) that cannot be responsible for the robust concentrations of the peptide detected in the blood (present data and 22). The observation that oral glucose load results in an increase of plasma 26RFa, as observed for incretins, suggested that the gastrointestinal tract may produce 26RFa. Indeed, our immunohistochemical investigations reveal that 26RFa is produced in abundance in the stomach and small intestine and in lower amounts in the colon, indicating that the gut is probably an important source of circulating 26RFa.
We thus carried out several experiments in mice, which revealed first that intraperitoneal administration of 26RFa during a glucose tolerance test significantly attenuates glucose-induced hyperglycemia. In contrast, the neuropeptide does not affect glucose basal levels, suggesting that 26RFa exerts an antihyperglycemic effect rather than a hypoglycemic effect. Concurrently, we show that 26RFa enhances insulin sensitivity during an insulin tolerance test, and increases insulin production during an acute glucose-stimulated insulin secretion test, indicating that the antihyperglycemic action of the neuropeptide is probably the result of both increased insulin sensitivity in target tissues and stimulation of insulin production. In agreement with this latter hypothesis, we found that 26RFa stimulates insulin secretion by the MIN6 mouse insulinoma cell line, which expresses GPR103, and that invalidation of GPR103 totally abolishes 26RFa-induced insulin release. In addition, we show that NPFF2, the other RFamide receptor that 26RFa and 43RFa can bind to (6), is not expressed by MIN6 cells. Altogether, these findings indicate that 26RFa is able to stimulate insulin release by insulin-secreting cells via the activation of GPR103. Consistent with this finding, it has been recently reported that 43RFa, the N-extended form of 26RFa, stimulates insulin secretion by human pancreatic islets and INS-1E β-cells and that this effect is exerted via GPR103 (30). However, in the same study, the authors report that 26RFa inhibits insulin secretion via a signaling pathway distinct from GPR103, an observation previously made in rat perfused pancreas (31), and suggest that NPFF2 may mediate the insulinostatic activity of 26RFa. Here, we show that MIN6 cells do not express NPFF2, which might explain the discrepancy between our data and those found in INS-1 cells. However, at present, it is not known whether INS-1 cells or human pancreatic islets express NPFF2, and the effects of activation of NPFF2 on insulin secretion remain unknown as well.
The increased insulin sensitivity induced by 26RFa led us to investigate the expression of the 26RFa receptor in insulin target tissues. We found that GPR103 is coexpressed with the insulin receptor and GLUT4 in the muscle, liver, and adipose tissue, with a higher expression in the liver, suggesting that 26RFa may play a role in glucose uptake, a hypothesis that awaits further study. Interestingly, in a previous study, GPR103 was also detected in epididymal fat pads and shown to be elevated by 16-fold in high-fat mice (32). In addition, these authors showed, using 3T3-L1 adipocyte cells, that 26RFa elicits a dose-dependent increase in fatty acid uptake and increases the expression of genes involved in lipid uptake as well as triglyceride accumulation (32).
One main feature of incretins is release by the gut under an oral glucose challenge (33). We thus investigated the possibility that 26RFa may be released by the gut after glucose ingestion in mice. We confirmed the presence of 26RFa in the gut, with a higher expression of the neuropeptide in the duodenum, as previously observed in humans (present study). We found that an oral glucose load results in a massive increase in plasma 26RFa levels 30 min after glucose ingestion and that, conversely, intravenous administration of glucose does not affect plasma 26RFa. Consistent with this finding, we show that application of high glucose concentrations at the apical side of duodenal fragments induces an important release of 26RFa at the serosal side of the duodenal slices, indicating that glucose ingestion induces an important secretion of 26RFa from the duodenum to the blood. Supporting this view, we show that, in a human patient with delayed gastric emptying, the rise in plasma 26RFa levels during an OGTT is delayed by 90 min compared with healthy subjects. Conversely, in a patient who underwent a total gastrectomy, the 26RFa peak occurs earlier at 90 min.
The current study provides evidence that 26RFa meets the criteria of an incretin hormone. However, 26RFa also shows differences with GIP and GLP-1. For instance, during an OGTT in humans, the 26RFa response to the oral glucose load is delayed compared with the GIP and GLP-1 peaks that overlap with those of glucose and insulin. At present, the physiological relevance of this delayed 26RFa response in humans remains unclear. One hypothesis could be that the aim of the delayed 26RFa increase is to sustain the GIP and GLP-1 insulinotropic action; that, however, deserves further investigation.
In conclusion, we report for the first time that obesity and diabetes in humans are associated with elevated plasma levels of the orexigenic neuropeptide 26RFa. We also show that 26RFa produced by the gut is released in the blood after a glucose challenge and reduces glucose-induced hyperglycemia via, at least in part, a direct stimulatory effect on insulin secretion by pancreatic β-cells. Taken together, these findings strongly suggest that 26RFa is a novel incretin that regulates glucose homeostasis.
Article Information
Acknowledgments. The authors thank H. Lemonnier, of INSERM U982, University of Rouen, Rouen, France, for technical assistance.
Funding. This work was supported by INSERM (U982), the University of Rouen, the Institute for Research and Innovation in Biomedicine, Plate-Forme de Recherche en Imagerie Cellulaire de Haute-Normandie, and Conseil Régional de Haute-Normandie.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. G.P., H.L., Y.A., and N.C. contributed to the study design and interpretation and wrote the manuscript. L.J., M.E.O., and D.A. contributed to the Q-PCR experiments. L.J., A.A., M.C., and J.L. contributed to the in vivo experiments on mice. L.J., M.E.O., M.M., C.B., and J.K.-C. contributed to the in vitro experiments on cell lines. M.P. and F.G. contributed to the immunohistochemical experiments. G.P., H.B., and H.L. contributed to the recruitment of the human cohort and to the studies on this cohort. G.P., P.D., F.C., and H.L. critically revised the manuscript and helped with the analysis of data and with the discussion. N.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.