The autonomic nervous system regulates pancreatic β-cell proliferation in adult male rats

Abstract

The pancreatic β-cell responds to changes in the nutrient environment to maintain glucose homeostasis by adapting its function and mass. Nutrients can act directly on the β-cell and also indirectly through the brain via autonomic nerves innervating islets. Despite the importance of the brain-islet axis in insulin secretion, relatively little is known regarding its involvement in β-cell proliferation. We previously demonstrated that prolonged infusions of nutrients in rats provoke a dramatic increase in β-cell proliferation in part because of the direct action of nutrients. Here, we addressed the contribution of the autonomic nervous system. In isolated islets, muscarinic stimulation increased, whereas adrenergic stimulation decreased, glucose-induced β-cell proliferation. Blocking α-adrenergic receptors reversed the effect of epinephrine on glucose + nonesterified fatty acids (NEFA)-induced β-cell proliferation, whereas activation of β-adrenergic receptors was without effect. Infusion of glucose + NEFA toward the brain stimulated β-cell proliferation, and this effect was abrogated following celiac vagotomy. The increase in β-cell proliferation following peripheral infusions of glucose + NEFA was not inhibited by vagotomy or atropine treatment but was blocked by coinfusion of epinephrine. We conclude that β-cell proliferation is stimulated by parasympathetic and inhibited by sympathetic signals. Whereas glucose + NEFA in the brain stimulates β-cell proliferation through the vagus nerve, β-cell proliferation in response to systemic nutrient excess does not involve parasympathetic signals but may be associated with decreased sympathetic tone.

INTRODUCTION

Plasticity in function and mass enables the β-cell to accommodate an increase in insulin demand arising from a changing metabolic environment (25, 29). A variety of stimuli, including growth factors, hormones, neurotransmitters, and nutrients, promote β-cell proliferation, but the precise mechanisms underlying compensatory β-cell mass expansion in response to metabolic challenges have only been partially elucidated.

The central nervous system (CNS) is a key player in the regulation of energy homeostasis (21) and islet hormone secretion (36). Specialized hypothalamic neurons detect local variations in glucose or nonesterified fatty acid (NEFA) levels that then modify their firing rates and more globally the nervous tone (24, 37). The autonomic nervous system (ANS) innervating pancreatic islets comprises parasympathetic and sympathetic nerves that release acetylcholine and catecholamines, respectively. Whereas activation of β-cell M₃-muscarinic receptors by acetylcholine promotes insulin secretion (9), adrenergic receptor activation by epinephrine inhibits insulin secretion (28). Accordingly, short-term exposure to excess nutrients augments insulin secretion by increasing parasympathetic and decreasing sympathetic firing rates (3, 22, 23, 27). Importantly, dysregulation of these pathways is thought to participate in the development of type 2 diabetes (35).

Accumulating evidence suggests that the brain-islet axis also regulates β-cell proliferation. Postnatally, an increase in CNS glucose-sensing neuron activity, associated with the postweaning switch to a carbohydrate-rich diet, promotes β-cell proliferation and mass via parasympathetic signals (34). In the adult rat, celiac vagotomy reduces basal β-cell proliferation (19). Conversely, proliferation is increased following ventromedial hypothalamic lesions that provoke vagal hyperactivity (15) and following infusion of the cholinergic agonist carbachol (41). The increase in β-cell mass in hyperphagic ob/ob mice is compromised following vagotomy (6). Likewise, hepatic insulin resistance in mice triggers β-cell proliferation through a neuronal relay involving the vagus nerve (13, 40). However, the contribution of ANS signals to β-cell proliferation in the context of nutrient surfeit remains to be determined.

We previously developed a rat model of nutrient excess in vivo in which we showed that combined infusion of glucose and NEFA for 72 h leads, in an age-dependent manner, to impaired β-cell function despite a marked increase in β-cell proliferation and mass (7, 10, 42). In a recent study using this model, we observed that the rate of β-cell replication in response to nutrient infusion was of much smaller magnitude when islets were transplanted under the kidney capsule than in the endogenous pancreas, suggesting that noncirculating, perhaps neuronal, signals contributed to the response (26). In the current study, we tested the involvement of the ANS in nutrient-induced β-cell proliferation and, more generally, the contribution of the ANS to the control of β-cell proliferation. Specifically, we asked 1) Do cholinergic and adrenergic agonists directly regulate β-cell proliferation ex vivo? 2) Does nutrient sensing in the brain stimulate β-cell proliferation? 3) Do the parasympathetic and sympathetic branches of the ANS regulate nutrient-induced β-cell proliferation in vivo?

RESEARCH DESIGN AND METHODS

Materials and reagents.

Saline (SAL) was obtained from Baxter (Mississauga, ON, Canada), 70% dextrose [glucose (GLU)] was from McKesson (Montreal, QC, Canada), and 20% ClinOleic (CLI, a soybean/olive triglyceride emulsion composed of 85% unsaturated-15% saturated fatty acids) was from Fresenius Kabi (Uppsala, Sweden). Heparin was purchased from Sandoz Canada (Boucherville, QC, Canada). Atropine sulfate, carbachol, yohimbine, and epinephrine bitartrate were obtained from Millipore-Sigma (Oakville, ON, Canada). Isoproterenol hydrochloride and prazosin hydrochloride were purchased from Tocris Bioscience (Bristol, UK). RPMI-1640 and 10% FBS were from Invitrogen (Burlington, VT). Antibodies to Ki-67, insulin, and Nkx6.1 were from Abcam (Ab15580; RRID: AB_443209), DAKO (A0564; RRID: AB_10013624), and the Developmental Studies Hybridoma Bank (F55A12; RRID: AB_532379), respectively.

Animal models.

All procedures were approved by the Institutional Committee for the Protection of Animals at the Centre Hospitalier de l’Université de Montréal. Male Lewis rats weighing 250–350 g (~2 mo old) (Charles River, Saint-Constant, QC, Canada) were individually housed under controlled temperature on a 12:12-h light-dark cycle with access to water and standard laboratory chow ad libitum. For nutrient infusions, animals underwent catheterization of the jugular vein and/or the carotid artery as described (10, 26) and were allowed to recover for 6 days. Vagotomy was performed during the catheterization procedure using the protocol described previously (5). Briefly, the anterior and posterior branches of the vagus nerve that course to the pancreas were exposed and sectioned (celiac vagotomy) or exposed (sham surgery).

Infusions.

Animals were randomized into two groups receiving either 0.9% SAL or 70% GLU plus 20% CLI coinfused with 20 U/ml heparin (GLU + CLI). Blood glucose was measured at 0, 6, 24, 48, and 72 h using a glucometer (Accu-Chek; Roche, Indianapolis, IN). For systemic infusions, nutrients were infused via the jugular catheter, and blood was sampled via the carotid catheter. Glucose infusion rate (GIR) (initially 44 mg·kg⁻¹·min⁻¹) was adjusted to maintain plasma glucose between 13.9 and 19.4 mmol/l throughout the 72-h infusion. CLI was infused at a constant rate of 1.67 ml·kg⁻¹·h⁻¹. The SAL infusion rate matched the highest rate of the GLU + CLI-infused animals. Atropine (5 mg/kg) was injected two times a day, and epinephrine (5 µg·h⁻¹·kg⁻¹) was coinfused with SAL or GLU + CLI throughout the 72-h infusion. All treatment groups were compared with vehicle-treated controls. For infusions toward the brain, nutrients were infused via the carotid artery catheter, and blood was sampled via the tail vein. GLU and CLI were infused at a constant rate of 1 mg·kg⁻¹·min⁻¹ and 42.8 µl·kg⁻¹·h⁻¹, respectively.

Rat islet culture.

Male Lewis rat (2 mo old) islets were isolated as described previously and cultured in RPMI-1640 with 10% FBS (14). Batches of 200 islets were cultured for 72 h in 5.5, 11.1, or 16.7 mmol/l glucose with or without vehicle [0.1 mmol/l BSA + 50% (vol/vol) ethanol] or 0.5 mmol/l NEFA mixture (15% palmitate, 65% oleate, and 20% linoleate) to mimic the composition of CLI. Carbachol (500 μmol/l), epinephrine (1 μmol/l), isoproterenol (1 μmol/l), yohimbine (10 μmol/l), prazosin (10 μmol/l), or vehicle was included throughout the 72-h culture and replaced daily with fresh media. At the end of the culture period, islets were washed one time in PBS and frozen in optimum cutting temperature compound (TissueTek, Torrance, CA) at −80°C for cryosectioning and immunostaining.

Immunohistochemical staining of pancreata and ex vivo islet preparations.

β-Cell proliferation was assessed by immunohistochemical staining of pancreatic and ex vivo islet cryosections by labeling for Ki-67 and insulin or Nkx6.1 as described previously (42). Briefly, pancreata were weighed and fixed for 3–4 h in 4% paraformaldehyde, cryoprotected overnight in 30% sucrose, and embedded in optimum cutting temperature compound. Cryosections (8 µm) were prepared by cryosectioning (Leica, Wetzlar, Germany). Islet cryosections were fixed with 3% paraformaldehyde before staining. Antigen retrieval was performed using sodium citrate buffer (10 mmol/l; pH = 6). Sections were treated for 1 h at room temperature with PBS-1% Triton-5% donkey serum, and primary antibodies Ki-67 (1:1,000), Nkx6.1 (1:8), and insulin (1:500) were incubated overnight at 4°C. Fluorophore-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, PA) were incubated for 1 h at room temperature, and nuclei were stained with DAPI. Images were acquired with a fluorescence microscope (Zeiss, Thornwood, NY). β-Cell proliferation was expressed as the percentage of double-positive Ki-67⁺ and insulin⁺ (or Nkx6.1⁺) cells to total insulin⁺ (or Nkx6.1⁺) cells. At least 1,500 insulin-positive cells among 10–15 distinct islets on 1–2 sections were manually counted in each animal.

Analytical measurements.

Circulating levels of NEFA (Wako Chemical, Osaka, Japan) and insulin (ELISA; Alpco, Windham, NH) were measured according to the manufacturer’s instructions.

Expression of data and statistical analysis.

Data are expressed as means ± SE. Statistical analyses were performed using Student’s t-test or ANOVA followed by two-by-two comparisons using Tukey’s or Sidak post hoc test (GraphPad Software). P < 0.05 was considered significant.

RESULTS

Carbachol potentiates and epinephrine inhibits β-cell proliferation ex vivo.

We first tested whether cholinergic and adrenergic signals regulate β-cell proliferation. Isolated rat islets were exposed to increasing glucose concentrations for 72 h in the presence or absence of the cholinergic agonist carbachol or the nonselective adrenergic agonist epinephrine. As shown in Fig. 1, A–D, the dose-dependent increase in β-cell proliferation in response to glucose, as measured by the percentage of double Ki-67- and insulin-positive over total insulin-positive cells, was potentiated by carbachol and inhibited by epinephrine. Consistent with the established role of these factors in the control of insulin secretion, insulin levels in the culture media at the end of the treatment were, respectively, increased and decreased by carbachol and epinephrine (Fig. 1E).

Fig. 1.

Regulation of glucose-induced β-cell proliferation in isolated rat islets by carbachol and epinephrine. Isolated rat islets were cultured in the presence of 5.5, 11.1, or 16.7 mmol/l glucose with or without carbachol (carb; 500 µmol/l) or epinephrine (epi; 1 µmol/l) for 72 h. A–C: representative sections of islets exposed to the indicated glucose concentrations and vehicle (A), carbachol (B), or epinephrine (C) immunostained for insulin (green), Ki-67 (red), and nuclei (blue). Arrows show positive nuclei for Ki-67. D: percentage of Ki-67⁺/insulin⁺ cells over total insulin⁺ cells. E: insulin levels in the islet supernatant. Solid line, vehicle; broken line, carbachol; dotted line, epinephrine. Veh, vehicle. Data are means ± SE (n = 6). *P < 0.05, **P < 0.01, and ***P < 0.001 versus vehicle, 2-way ANOVA with Dunnett posttest. Scale bar = 50 µm.

Blockade of α₁- and α₂-adrenergic receptors reverses the inhibition of nutrient-induced β-cell proliferation by epinephrine.

Epinephrine acts on both α- and β-adrenergic receptors. In islets, the α-adrenergic branch is predominant, and epinephrine inhibits insulin secretion (28). To confirm that epinephrine inhibits nutrient-induced β-cell proliferation, isolated rat islets were exposed to 16.7 mmol/l glucose and an oleate-based NEFA mixture that we have shown previously to potently stimulate β-cell proliferation (26). Epinephrine dose dependently inhibited β-cell proliferation (Fig. 2, A and B) and decreased insulin accumulation in the media (Fig. 2C). To assess the role of α-adrenergic receptors, islets were cotreated with epinephrine and yohimbine, a specific α₂-antagonist, or prazosin, a specific α₁-antagonist, in the presence of 5.5 mmol/l glucose, 16.7 mmol/l glucose, or 16.7 mmol/l glucose + NEFA (Fig. 2, D–I). Yohimbine reversed the inhibitory effect of epinephrine on β-cell proliferation induced by 16.7 mmol/l glucose + NEFA, but not 5.5 or 16.7 mmol/l glucose alone (Fig. 2, D and E), and restored insulin secretion (Fig. 2F). Similarly, the inhibitory effect of epinephrine on β-cell proliferation and insulin secretion in response to nutrients was reversed when islets were exposed to prazosin (Fig. 2, G–I).

Fig. 2.

Regulation of glucose- and nonesterified fatty acid (NEFA)-induced β-cell proliferation in isolated rat islets by α-adrenergic signaling. Isolated rat islets were cultured in the presence of 16.7 mmol/l glucose + NEFA mix (0.5 mmol/l; oleate, linoleate, palmitate; NEFA) with or without epinephrine (epi; 10⁻¹⁰ to 10⁻⁵ mmol/l) (A–C) or in the presence of 5.5 mmol/l glucose, 16.7 mmol/l glucose, or 16.7 mmol/l glucose + NEFA with all conditions exposed to epinephrine with or without yohimbine (yoh; 10 µmol/l) or prazosin (pra; 10 µmol/l) (D–I) for 72 h. A, D, and G: representative sections of islets immunostained for insulin (green), Ki-67 (red), and nuclei (blue). Arrows show positive nuclei for Ki-67. B, E, and H: percentage of Ki-67⁺/insulin⁺ cells over total insulin⁺ cells. C, F, and I: insulin levels in the islet supernatant. Open circles represent individual data points. White bar, 5.5 mmol/l glucose; gray bar, 16.7 mmol/l glucose; black bar, 16.7 mmol/l glucose + NEFA. Veh, vehicle. Data are individual values and means ± SE (n = 4–10). *P < 0.05, **P < 0.01, and ***P < 0.001, 1-way ANOVA with Dunnett posttest. Scale bar = 50 µm.

Activation of β-adrenergic receptors does not affect nutrient-induced β-cell proliferation.

To determine the potential impact of the β-adrenergic branch, we treated isolated rat islets with 16.7 mmol/l glucose + NEFA and the selective β-adrenergic agonist isoproterenol. Isoproterenol did not stimulate β-cell proliferation (Fig. 3, A and B) or insulin accumulation in the media (Fig. 3C) at any of the concentrations tested. In a second series of experiments, we assessed the effects of isoproterenol in the presence of 5.5 mmol/l glucose, 16.7 mmol/l glucose, or 16.7 mmol/l glucose + NEFA. Here again, isoproterenol did not increase β-cell proliferation (Fig. 3, D and E) or insulin accumulation in the media (Fig. 3F) in any of the experimental conditions.

Fig. 3.

Regulation of glucose- and nonesterified fatty acid (NEFA)-induced β-cell proliferation in isolated rat islets by isoproterenol. Isolated rat islets were cultured in the presence of 16.7 mmol/l glucose + NEFA mix (0.5 mmol/l; oleate, linoleate, palmitate; NEFA) with or without isoproterenol (iso; 10⁻⁸ to 10⁻⁴ mol/l) (A–C) or 5.5 mmol/l glucose, 16.7 mmol/l glucose, or 16.7 mmol/l glucose + NEFA with or without isoproterenol (1 µmol/l) (D–F) for 72 h. A and D: representative sections of islets immunostained for insulin (green), Ki-67 (red), and nuclei (blue). Arrows show positive nuclei for Ki-67. B and E: percentage of Ki-67⁺/insulin⁺ cells over total insulin⁺ cells. C and F: insulin levels in the islet supernatant. Open circles represent individual data points. White bar, 5.5 mmol/l glucose; gray bar, 16.7 mmol/l glucose; black bar, 16.7 mmol/l glucose + NEFA. Data are individual values and means ± SE (n = 6). ***P < 0.001 versus vehicle, 1-way ANOVA with Sidak posttest. Scale bar = 50 µm.

Overall, these results indicate that cholinergic stimulation promotes, whereas adrenergic stimulation via α-adrenergic receptors inhibits, β-cell proliferation ex vivo.

Nutrient infusion toward the brain increases β-cell proliferation.

We then asked whether local sensing of nutrients in the brain can stimulate β-cell proliferation and, if so, whether the response involves vagal efferents. Rats were infused with GLU + CLI in the carotid artery toward the brain at concentrations that do not alter systemic blood glucose levels (Fig. 4A). Plasma insulin (Fig. 4B) and NEFA (Fig. 4C) transiently decreased and increased, respectively, at the 6-h time point but then returned to control levels. GLU + CLI infusion also decreased caloric intake (Fig. 4D). Intracarotid GLU + CLI infusion led to a significant increase in β-cell proliferation, as assessed by the percentage of double-positive cells for Ki-67 and insulin compared with SAL-infused rats (1.7 ± 0.3-fold increase, n = 7, P < 0.01; Fig. 4, E and F). To confirm that proliferating cells were indeed β-cells, we used a second β-cell marker, the transcription factor Nkx6.1. The percentage of Ki-67- and Nkx6.1-positive cells in islets from GLU + CLI-infused rats was 2.3 ± 0.3-fold greater than in the SAL-infused group [n = 3, P < 0.05; Supplemental Fig. S1 (https://doi.org/10.6084/m9.figshare.8024102)].

Fig. 4.

Metabolic parameters and β-cell proliferation following central infusion of glucose (GLU) + ClinOleic (CLI). Normal (A–F) and sham-operated (sh) or vagotomized (vgx; G–L) rats were infused with saline (SAL) or GLU + CLI via the carotid artery toward the brain for 72 h. Blood glucose (A and G), plasma insulin (B and H), plasma nonesterified fatty acid (NEFA, C and I), and caloric intake (D and J) during the infusion. E and K: percentage of Ki-67⁺/insulin⁺ cells over total insulin⁺ cells. F and L: representative pancreatic sections immunostained for insulin (green), Ki-67 (red), and nuclei (blue) after 72-h infusions. Arrows show positive nuclei for Ki-67. A–D: black circle, SAL; white circle, GLU + CLI. J: black circle, SAL-sh; white circle, GLU + CLI-sh; black square, SAL-vgx; white square, GLU + CLI-vgx. Open circles represent individual data points. Data are individual values and means ± SE (n = 5–10). *P < 0.05 and **P < 0.01 versus SAL, 1-way ANOVA with Sidak posttest. Scale bar = 50 µm.

To determine whether the vagus nerve mediates the proliferative response, in a second series of animals we performed celiac vagotomy or sham surgery before the infusion. Circulating glucose, insulin, NEFA, and caloric intake were unaltered in vagotomized rats compared with sham-operated animals (Fig. 4, G–J). Celiac vagotomy had no effect on the basal levels of β-cell proliferation in SAL-infused animals (Fig. 4, K and L). GLU + CLI infusions increased β-cell proliferation in sham-operated animals, similar to the results shown in Fig. 4, E and F, an effect that was prevented by celiac vagotomy (Fig. 4, K and L).

These data demonstrate that a local increase in glucose and NEFA levels in the brain can stimulate β-cell proliferation through the vagus nerve.

Blockade of pancreatic vagal innervation or muscarinic signaling does not prevent β-cell proliferation in response to systemic nutrient infusion.

We then assessed whether vagal innervation also contributes to β-cell proliferation in response to a systemic increase in glucose and NEFA levels. Rats were infused for 72 h with SAL or GLU + CLI following celiac vagotomy or sham surgery. GLU + CLI infusions increased blood glucose (Fig. 5A), plasma insulin (Fig. 5B), and NEFA (Fig. 5C) levels. Food intake was decreased to compensate for the calories provided by the GLU + CLI infusion, such that total caloric intake was not different between the groups (data not shown). In agreement with our previous report (26), GLU + CLI infusions in sham-operated animals led to a marked increase in β-cell proliferation compared with SAL-infused rats (23 ± 5-fold increase, n = 6, P < 0.01; Fig. 5, D and E). With the use of Nkx6.1 instead of insulin as a β-cell marker, the percentage of Ki-67- and Nkx6.1-positive cells in islets from GLU + CLI-infused rats was 38 ± 5-fold higher than in the SAL-infused group [n = 3, P < 0.01; Supplemental Fig. S1 (https://doi.org/10.6084/m9.figshare.8024102)].

Fig. 5.

Metabolic parameters and β-cell proliferation following systemic infusion of glucose (GLU) + ClinOleic (CLI) and effects of celiac vagotomy and atropine. Sham-operated (sh) or vagotomized (vgx) (A–E) or vehicle (veh) or atropine-treated (atr) (F–J) rats were infused with saline (SAL) or GLU + CLI via the jugular vein for 72 h. Blood glucose (A and F), plasma insulin (B and G), and plasma nonesterified fatty acid (NEFA, C and H) during the infusion. D and I: percentage of Ki-67⁺/insulin⁺ cells over total insulin⁺ cells after the 72-h infusion. E and J: representative pancreatic sections immunostained for insulin (green), Ki-67 (red), and nuclei (blue). Arrows show positive nuclei for Ki-67. Open circles represent individual data points. Data are individual values and means ± SE (n = 6). *P < 0.05 and ***P < 0.001 versus SAL, 1-way ANOVA with Sidak posttest. Scale bar = 50 µm.

Blood glucose (Fig. 5A), plasma insulin (Fig. 5B), and NEFA (Fig. 5C) levels were similar between vagotomized and control rats in both SAL- and GLU + CLI-infused groups as was the GIR in the GLU + CLI-infused groups (data not shown). Neither basal nor GLU + CLI-induced β-cell proliferation was prevented by vagotomy (Fig. 5, D and E).

To confirm the lack of involvement of the parasympathetic nervous system in GLU + CLI-induced proliferation, we coinjected the muscarinic antagonist atropine two times a day during the 72-h GLU + CLI infusion. Similar to our findings in vagotomized rats, blood glucose (Fig. 5F), plasma insulin (Fig. 5G), and NEFA (Fig. 5H) levels were not affected by atropine. However, atropine reduced caloric intake (data not shown) as shown previously in rats (38, 39). Consistent with our findings in vagotomized rats (Fig. 5), neither basal nor GLU + CLI-induced β-cell proliferation was inhibited by atropine (Fig. 5, I and J).

Altogether, these data indicate that vagal efferent signals are implicated in the β-cell proliferative response to GLU + CLI infusion in the brain but that the vagus nerve does not appear to mediate the large increase in β-cell proliferation in response to peripheral nutrient infusion under these experimental conditions.

Epinephrine blunts nutrient-induced β-cell proliferation in vivo.

In light of the potent inhibitory effect of epinephrine on β-cell proliferation in isolated islets (Figs. 1 and 2), we asked whether epinephrine also inhibits β-cell proliferation in response to GLU + CLI infusions in vivo (Fig. 6). Epinephrine was coinfused with GLU + CLI or SAL for 72 h. SAL-infused epinephrine-treated rats had higher blood glucose levels during the infusion (Fig. 6A). In GLU + CLI-infused animals, the GIR had to be reduced in the epinephrine-treated group (Fig. 6B) to maintain blood glucose levels in the target range (Fig. 6A). Plasma insulin levels were reduced by epinephrine treatment in GLU + CLI but not SAL-infused rats (Fig. 6C), whereas plasma NEFA levels were not affected (Fig. 6D). Despite similar levels of circulating glucose and NEFA, nutrient-induced β-cell proliferation was completely abrogated by epinephrine (Fig. 6, E and F).

Fig. 6.

Effects of epinephrine treatment on metabolic parameters and β-cell proliferation in response to systemic infusion of glucose (GLU) + ClinOleic (CLI). Rats were infused with saline (SAL) or GLU + CLI for 72 h and either vehicle (veh) or epinephrine (epi). Blood glucose (A), glucose infusion rate (GIR, B), plasma insulin (C), and plasma nonesterified fatty acid (NEFA, D) during the infusion. E: percentage of Ki-67⁺/insulin⁺ cells over total insulin⁺ cells. F: representative pancreatic sections immunostained for insulin (green), Ki-67 (red), and nuclei (blue). Arrows show nuclei positive for Ki-67. Open circles represent individual data points. Data are individual values and means ± SE (n = 4–6). **P < 0.01 and ***P < 0.001 versus SAL. Scale bar = 50 µm.

DISCUSSION

In this study, we investigated the contribution of the ANS to the β-cell proliferative response to nutrients in rats. First, ex vivo studies in islets showed that β-cell proliferation is increased by muscarinic and decreased by adrenergic agonists. Blockade of α-adrenergic receptors restored nutrient-induced β-cell proliferation in the presence of epinephrine, whereas selective β-adrenergic stimulation with isoproterenol did not affect β-cell proliferation. Second, local infusion of nutrients toward the brain stimulated β-cell proliferation, and this effect was abrogated by surgical elimination of vagal signals. Third, neither celiac vagotomy nor muscarinic inhibition blocked the increase in β-cell proliferation following systemic nutrient infusion, indicating that parasympathetic nervous signals do not contribute to the proliferative response to peripheral nutrients. In contrast, adrenergic stimulation with epinephrine completely blunted nutrient-induced β-cell proliferation in vivo.

The regulation of β-cell proliferation by cholinergic and adrenergic agonists in isolated islets was accompanied by parallel changes in insulin secretion. Therefore, we cannot rule out the possibility that autocrine and/or paracrine regulation of β-cell proliferation by insulin might contribute to the observed effects. However, several lines of evidence do not support this possibility. First, intracarotid GLU + CLI infusions increase β-cell proliferation (Fig. 4E) without concomitant changes in circulating insulin (Fig. 4B), suggesting an insulin-independent effect. Second, in isolated islets, the level of β-cell proliferation in response to nutrients does not always correlate with the accumulation of insulin in the media (for example, compare the effects of glucose in Fig. 4A versus Fig. 4I and Fig. 4E versus Fig. 4F). Third, we found that the insulin receptor antagonist, S961, does not affect nutrient-induced β-cell proliferation in isolated rat islets (data not shown), which is consistent with studies in mouse islets (32). Finally, in our previous study using similar experimental conditions ex vivo, we directly showed that exogenous insulin is unable to stimulate β-cell proliferation (26). Overall, these data argue against a major contribution of the autocrine/paracrine action of insulin to the proliferative effect of glucose and NEFA in β-cells.

Previous studies indicate that epinephrine acting via α₂-adrenergic receptors, which are highly expressed in β-cells, blunts insulin secretion by activating inhibitory G proteins to lower cAMP levels (33). Accordingly, adra2a knockout mice show reduced circulating glucose levels and increased insulin levels (30). In partially pancreatectomized rats, the decrease in β-cell α₂-adrenergic receptor function stimulates insulin secretion and β-cell proliferation in weanling rats during pancreatic regeneration (4). In addition, Berger et al. (1) showed that, whereas loss of adra2a results in an increase in β-cell proliferation in neonatal mice, α₂-agonists inhibit β-cell proliferation in adult mouse islets ex vivo as was suggested in earlier studies (31). Likewise, N’Guyen et al. (27) observed a reduction of insulin hypersecretion in response to glucose infusion following adrenergic stimulation, suggesting a role for the sympathetic tone in regulation of β-cell functional compensation. Consistent with these finding, we showed that epinephrine potently inhibits nutrient-induced β-cell proliferation via α-adrenergic receptors. These data suggest that alleviation of tonic sympathetic inhibition should stimulate β-cell proliferation. Further studies will be required to test this hypothesis.

Interestingly, although several studies have shown that β₂-adrenergic signaling contributes to postnatal islet development (2) and potentiates glucose-stimulated insulin secretion (8, 12, 18), our data using isoproterenol suggest that β₂-activation does not potentiate β-cell proliferation induced by either glucose or NEFA in rat islets ex vivo. We expected that, at low concentrations, isoproterenol might increase β-cell proliferation and insulin secretion in vitro as shown in rodent and human cell lines (11, 20). This discrepancy may be because of differential responsiveness between human and rodent islets; whereas human β-cells are sensitive to β₂-agonists, in rat islets glucagon-secreting α-cells, but not β-cells, respond to β₂-activation (16, 17).

A contribution of the parasympathetic nervous system to the regulation of β-cell proliferation by nutrients was suggested by several studies. First, nutrient infusion increases parasympathetic activity in rats and humans (3, 23, 27). Accordingly, we found here that nutrient infusion locally in the brain was sufficient to promote an increase in β-cell proliferation through the vagus nerve. Second, in agreement with the proproliferative effect of cholinergic stimulation in mouse islets (40), we showed that carbachol potentiates glucose-induced proliferation in rat islets ex vivo. In contrast, parasympathetic signals were not required for the large increase in β-cell proliferation observed in response to systemic nutrient infusion. These findings are not in agreement with studies performed by Imai et al. (13) and Yamamoto et al. (40), who showed that pancreatic vagotomy impairs compensatory increases in islet insulin content and β-cell mass in response to hepatic insulin resistance in mice. Likewise, celiac vagotomy in Sprague-Dawley rats lowers basal β-cell proliferation rates (19), and β-cell proliferation induced by ventromedial hypothalamic lesions is blunted by subdiaphragmatic vagotomy in rats (15). Besides possible species- or strain-related differences, there are several potential explanations to these discordant observations. First, variations in the surgical protocol may lead to different effects of the pancreatic vagotomy (5). We are confident, however, that our surgical approach effectively eliminated pancreatic vagal innervation, since it abrogated the proliferative response to GLU + CLI infusions toward the brain. Second, the degree of insulin resistance is likely an important factor that must be considered. In a previous study (26) we showed that the β-cell proliferative response to GLU + CLI results from a combination of nutrient-induced insulin resistance and direct effects of glucose and NEFA on the β-cell. On the other hand, in the present study, we observed that infusion of GLU + CLI toward the brain is capable of stimulating β-cell proliferation through vagal efferents without changes in peripheral insulin sensitivity. Therefore, we surmise that the contribution of vagal inputs is minimal, and clearly insufficient, to mount the strong β-cell compensatory response to peripheral nutrient infusions observed under our experimental conditions. Finally, we cannot exclude the possibility that a greater contribution of parasympathetic signals might have been detected at earlier time points during the infusion or under conditions of milder insulin resistance.

Overall, our results demonstrate that β-cell proliferation is stimulated by parasympathetic signals and inhibited by sympathetic signals ex vivo and in vivo. Whereas nutrients infused directly in the brain can stimulate β-cell proliferation through the vagus nerve, β-cell proliferation in response to systemic nutrient excess does not appear to involve the parasympathetic nervous system but may be associated with decreased sympathetic tone.

GRANTS

This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant R01-DK-58096 to V. Poitout and Canadian Institutes of Health Research Grant MOP 77686 to V. Poitout. V. S. Moullé was supported by a Postdoctoral Fellowship from the CRCHUM. T. Alquier was supported by a salary award from Fonds de Recherche Québec-Santé.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

V.S.M., T.A., J.G., and V.P. conceived and designed research; V.S.M., C.T., A.-L.C., K.V., M.E., and G.F. performed experiments; V.S.M., C.T., A.-L.C., K.V., T.A., J.G., and V.P. analyzed data; V.S.M., T.A., J.G., and V.P. interpreted results of experiments; V.S.M. prepared figures; V.S.M. drafted manuscript; V.S.M., T.A., J.G., and V.P. approved final version of manuscript; T.A., J.G., and V.P. edited and revised manuscript.