Peripheral muscarinic acetylcholine receptors regulate insulin and glucagon release in rodents but their importance for similar roles in humans is unclear. Bethanechol, an acetylcholine analogue that does not cross the blood-brain barrier, was used to examine the role of peripheral muscarinic signaling on glucose homeostasis in humans with normal glucose tolerance (NGT; n = 10), impaired glucose tolerance (IGT; n = 11), and type 2 diabetes mellitus (T2DM; n = 9). Subjects received four liquid meal tolerance tests, each with a different dose of oral bethanechol (0, 50, 100, or 150 mg) given 60 min before a meal containing acetaminophen. Plasma pancreatic polypeptide (PP), glucose-dependent insulinotropic polypeptide (GIP), glucagon-like peptide-1 (GLP-1), glucose, glucagon, C-peptide, and acetaminophen concentrations were measured. Insulin secretion rates (ISRs) were calculated from C-peptide levels. Acetaminophen and PP concentrations were surrogate markers for gastric emptying and cholinergic input to islets. The 150 mg dose of bethanechol increased the PP response 2-fold only in the IGT group, amplified GLP-1 release in the IGT and T2DM groups, and augmented the GIP response only in the NGT group. However, bethanechol did not alter ISRs or plasma glucose, glucagon, or acetaminophen concentrations in any group. Prior studies showed infusion of xenin-25, an intestinal peptide, delays gastric emptying and reduces GLP-1 release but not ISRs when normalized to plasma glucose levels. Analysis of archived plasma samples from this study showed xenin-25 amplified postprandial PP responses ~4-fold in subjects with NGT, IGT, and T2DM. Thus, increasing postprandial cholinergic input to islets augments insulin secretion in mice but not humans.
Trial Registration: ClinicalTrials.gov NCT01434901
Citation: Chowdhury S, Wang S, Dunai J, Kilpatrick R, Oestricker LZ, Wallendorf MJ, et al. (2016) Hormonal Responses to Cholinergic Input Are Different in Humans with and without Type 2 Diabetes Mellitus. PLoS ONE11(6): e0156852. https://doi.org/10.1371/journal.pone.0156852
Editor: Aristidis Veves, Harvard Medical School, UNITED STATES
Received: February 12, 2016; Accepted: May 20, 2016; Published: June 15, 2016
Copyright: © 2016 Chowdhury et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: Portions of this research were supported by funds from the American Diabetes Association (BMW; Grants #1-10-CT-58 and #1-13-CE-46); the Blum Kovler Foundation BMW; National Institutes of Health (BMW; Grant numbers 5RC1DK086163 and 1R01DK088126); the Washington University Diabetes Research Center Immunoassay Core (P60 DK020579); the Washington University Nutrition Obesity Research Center Grant (P30DK056341) from the National Institute of Diabetes and Digestive and Kidney Diseases; the Washington University Digestive Disease Research Core Center (P30 DK52574-16); the Washington University Clinical and Translational Science Award (UL1 TR000448) from the National Center for Advancing Translational Sciences (NCATS); the Biologic Therapy Core Facility of the Alvin J. Siteman Cancer Center at Washington University School of Medicine and Barnes-Jewish Hospital in St. Louis, Mo. (NCI Cancer Center Support Grant P30 CA91842); the NIH National Center for Research Resources (P41 RR00954 and UL1 RR024992). The content is solely the responsibility of the authors and does not necessarily represent the official view of the NIH or any other funding agency. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Transmitters and peptides released from neurons that innervate islets play important roles in regulating insulin and glucagon release [1,2]. In general, parasympathetic and sympathetic neurons that innervate pancreatic islets increase and inhibit insulin release, respectively [1–4]. In rodents, islets are richly innervated by parasympathetic neurons . Previous studies from our laboratory have shown that a cholinergic neural relay amplifies the effects of glucose-dependent insulinotropic polypeptide (GIP) on insulin release in mice . Studies by others using genetically modified mice and/or islets indicate that cholinergic signaling via M3 muscarinic acetylcholine receptors plays an important role in regulating insulin and glucagon release [7–12]. Consistent with mouse experiments, studies with the isolated perfused human pancreas have shown that electrical stimulation of the splanchnic nerve in the presence and absence of selective neural inhibitors increases both cholinergic and sympathetic input to islets which in turn, regulates insulin, glucagon, pancreatic polypeptide (PP), and somatostatin release [13–18]. Further, neurotransmitters regulate insulin release in isolated human islets . In contrast to the in situ and ex vivo studies, physiologic stimuli (e.g. nutrients, stress) would differentially affect parasympathetic versus sympathetic input to islets. Thus, the physiologic relevance of the electrical stimulation and human islet studies is not clear.
There are conflicting reports on the effects of physiologic levels of cholinergic signaling for regulating insulin and glucagon responses in vivo in humans. For example, prior prolonged mild hyperglycemia results in a compensatory increase in C-peptide secretion during intravenous glucose tolerance tests, which is only partially inhibited by atropine . In another study, atropine inhibited the cephalic insulin response to meal ingestion by 20%  Specific anti-psychotic medications that are associated with development of T2DM also exhibit secondary affinity/antagonism to muscarinic M3 receptors . During 50-gram oral glucose tolerance tests, areas under the curve for glucose, glucagon-like peptide-1 (GLP-1), and insulin secretion rates (ISRs) were increased in humans with truncal vagotomy plus pyloroplasty compared to controls . However, these changes are likely indirect because vagotomy also increased the rate of gastric emptying. Conversely, vagotomy for peptide ulcer disease had little effect on plasma glucose levels following intravenous administration of glucose [24,25] and atropine inhibited postprandial PP release but not insulin secretion in Pima Indians . Thus, the importance of cholinergic regulation of insulin and glucagon release in response to a physiologic mixed meal in humans is unclear.
A recent study suggested that in contrast to mice, human islets are poorly innervated by parasympathetic (cholinergic) neurons . If so, a neural cholinergic relay to islets would have little effect on islet physiology. PP is a 36-amino acid peptide produced by a subpopulation of endocrine cells called PP cells. Circulating PP is undetectable in humans after total pancreatectomy indicating it is produced almost exclusively by the pancreas . Although there are species-specific differences , in humans PP cells are mainly localized at the periphery of islets [29–31]. PP is released into the circulation in response to meal ingestion  but not to intravenous infusion of glucose, amino acids, or fat [27,33]. Atropine blocks PP release in response to food intake, insulin-induced hypoglycemia, and intravenous infusion of GIP, bombesin, gastrin releasing peptide, neurotensin, and bethanechol [34–38]. Truncal vagotomy abolishes PP release in most cases studied [34,39,40] but a non-vagal mechanism may also contribute to the regulation of PP release . These collective results suggest that PP secretion is regulated by vagal and non-vagal cholinergic input to islets.
Xenin-25 is an intestinal peptide reportedly produced by a subset of enteroendocrine cells [42–45]. Effects of xenin-25 are mediated by activation of neurotensin receptor 1 [46–51]. We have shown that in sections of human pancreas, neurotensin receptor 1 is detectable on nerves, but not islet endocrine cells . Further, during graded glucose infusions, administration of xenin-25, alone and more so when co-administered with GIP, profoundly increased PP release in humans . These results strongly suggest that functional cholinergic neurons innervate human as well as rodent islets. Bethanechol is a charged quaternary amine moiety that does not cross the blood brain barrier and can thus be used to study the effects of muscarinic cholinergic signaling in the periphery. The purpose of the present study is to determine if bethanechol and/or xenin-25 increase postprandial cholinergic input to islets and if this signaling is associated with changes in postprandial glucose, GIP, GLP-1, insulin, C-peptide, glucagon, and PP levels and insulin secretion rates (ISRs) in humans with normal glucose tolerance (NGT), impaired glucose tolerance (IGT) and T2DM. Results indicate that increasing postprandial cholinergic signaling to islets by administration of bethanechol or xenin-25 has little net effect on insulin or glucagon release.
Materials and Methods
The study protocol was approved by Washington University’s Human Research Protection Office and the FDA (IND#103,374) and was registered with ClinicalTrials.gov (NCT01434901). Studies were performed by the nursing and medical staff in the Clinical Research Unit of the Institute of Clinical and Translational Sciences of Washington University after obtaining written informed consent. Subject recruitment was initiated on August 15, 2011 and follow-up for the final participant was completed on July 7, 2014. Subjects were recruited through Washington University’s Research Participant Registry as well as from the PI’s database of previous participants. After a phone screen, potential participants underwent a screening visit in the Clinical Research Unit. Subjects were remunerated to encourage completion of the study. Male and female subjects with NGT (n = 10), IGT (n = 11), and mild T2DM (n = 9) were studied. Glucose tolerance was defined by the 2-hour plasma glucose level measured during a 75-gram oral glucose tolerance test (OGTT) using diagnostic criteria of the American Diabetes Association . With respect to T2DM, selection criteria were designed to exclude subjects with advanced beta cell failure: subjects were required to have HbA1c ≤ 9%, could not be using insulin for treatment, had no known history of symptomatic gastroparesis or peripheral neuropathy (e.g., burning or tingling in feet) and were enrolled only if oral anti-diabetic medications could be safely discontinued for 48-hours before each study visit. Women of childbearing potential were required to use birth control. Subjects were excluded if they 1) had a history of chronic pancreatitis and/or risk factors for chronic pancreatitis 2) had a history of significant gastrointestinal disorders, (e.g. peptide-ulcer disease) 3) were taking non-diabetes medications known to affect glucose homeostasis and 4) had any significant chronic illness including heart, renal, liver, inflammatory or malignant disease. The use of choline esters is contraindicated in persons with hyperthyroidism, coronary artery disease, peptic ulcer, asthma, chronic bronchitis, or COPD. Subjects with any of these conditions were excluded. Baseline characteristics for each group were determined during a screening visit following a 12-hour fast and are shown in Table 1. Based on variance and correlation estimates from the current study, 10 subjects would provide 80% power when the true differences are 4500 for PP AUC (sd = 5000, correlation = 0.6) and 29,000 for ISR AUC (sd = 31,623, correlation = 0.6).
This is a crossover study in which each participant underwent a series of meal tolerance tests after a 12-hour overnight fast (Fig 1). Subjects were blinded to treatment. Each visit was separated by at least 2 weeks. Hemoglobin levels were measured before each study visit and anyone with a value <11.2 g/dL had that study delayed. In subjects with T2DM and taking oral diabetes medications, drugs were discontinued for 48 hours before each study visit. An intravenous catheter was placed into a hand vein. This hand was kept warm in a thermostatically controlled box for sampling arterialized venous blood [53,54]. Subjects with a fasting blood sugar ≥ 120 mg/dL were given a bolus of intravenous human insulin (0.01 U/kg) at 30 min intervals as needed to decrease the blood glucose level to 100–120 mg/dL to limit variability of initial glucose levels. Blood glucose concentrations were stable for longer than 20 min before administration of bethanechol.
Bethanechol is an acetylcholine analog that does not cross the blood-brain barrier. Oral dosing was based on a survey of the literature as follows. Measuring the stretch response of the bladder muscle [55,56], the authors reported a progressive increase in mean cystometric pressure with doses of 50, 100, and 200 mg. Of the 20 subjects in this study, sweating, diarrhea, chills, bradycardia and hypotension were noted in only 1–2 patients with the 200 mg dose. The maximal response to oral bethanechol began after ~1 hour and the action of moderate doses (50–100 mg) persisted for 2–3 hours. Bethanechol has also been used to treat humans with xerostomia after radiation therapy (25–50 mg 3 times per day [57,58]) and to relieve side effects due to tricyclic antidepressants (25 mg 3 times per day ). To ensure patient safety in our study, bethanechol was administered in escalating doses of 0, 25, 50, and 100 mg during successive meal tolerance tests. Interim results indicated that the 25 mg dose was ineffective and the 100 mg dose was well-tolerated. Thus, the 25 mg dose was eliminated and the maximal dose was increased to 150 mg. To monitor for symptoms of bethanechol toxicity, subjects were placed on continuous cardiac monitoring with blood pressure and heart rate checked every 15 min. There were no adverse events or unintended effects during this study.
Meal Tolerance Tests.
Boost Plus (Nestle Health Science, Florham Park, New Jersey) is a liquid mixed meal (360 calories, 14 g of fat, 45 g of carbohydrates, and 14 g of protein). Fasted subjects ingested Boost Plus and liquid acetaminophen (ACM; 1.5 g/15 mL; Q-PAP Infants’ Drop; Qualitest Pharmaceuticals, Huntsville, AL) over a 3 min period starting at 0 min. The ACM was included to allow for estimation of the rate of gastric emptying as discussed in our earlier study . Bethanechol (or placebo) was administered orally with 100 mL of water 1 hour before meal ingestion.
Complete metabolic profiles and plasma concentrations of glucose, insulin, C-peptide, glucagon, PP, ACM, total GIP, active GLP-1, and HbA1c were determined as previously described [33,60,61]. The PP assay involves an extraction step to remove interfering compounds . A cholinergic symptom survey was administered before, during, and immediately after each study visit. Any diarrheal episodes during the study were recorded by the nurses and post-visit episodes determined by telephone follow-up. Heart rate and systolic and diastolic blood pressure and were measured bedside throughout each study visit.
Data Analysis and Statistics.
Basal glucose and hormone levels and ISRs were calculated for each individual by averaging values for the -90, -75, and -60 time points for all study visits. Means, SDs, and SEMs were then calculated for each group. ISRs were derived by stochastic deconvolution of the peripheral C-peptide concentrations as in earlier studies [60,61] using population-based estimates of C-peptide clearance kinetics [62–64]. Differences in baseline characteristics and/or placebo treatments between groups were evaluated by one-way ANOVA for continuous variables and fisher exact test for categorical variables. Areas under the curve (AUCs) were calculated using the trapezoid method and incremental AUCs (iAUCs) were determined by subtracting baseline AUC from the AUC. Data for AUCs and iAUCs were analyzed using mixed effects models with subject as a random effect and bethanechol as a fixed effect using SAS v9.4. Baseline values were used as a covariate for the analysis of the AUCs but not iAUCs. Paired comparisons were limited to evaluating the effects of 50, 100, and 150 mg bethanechol compared to placebo. Outcome measures through time were analyzed using the mixed random effects repeated measures model with covariance structure estimated by a spatial model (SAS 9.4). Subject and subject by drug interaction were random effects.
Xenin-25 Study with Archived Samples
The study protocol was approved by Washington University’s Human Research Protection Office and the FDA (IND#103,374) and was registered with ClinicalTrials.gov (NCT00949663). Studies were performed in the Clinical Research Unit of the Institute of Clinical and Translational Sciences of Washington University after obtaining written informed consent. Study design (Fig 2), detailed procedures and an initial set of results have been published . Subject recruitment was initiated on January 1, 2010 and follow-up for the final participant was completed on March 20, 2012. Subjects had been recruited through Washington University’s Volunteers for Health Office and were remunerated to encourage completion of the study. No new subjects were enrolled or studied for the present report and only a subset of archived plasma samples from the prior study were analyzed (Table 1; n = 10 per group). Patients gave written informed consent for future analyses of archived samples. There were no adverse events or unintended effects during this study.
Briefly, subjects were screened, assigned to groups with NGT, IGT, and T2DM, and then studied as described above. However, instead of oral bethanechol administration at minus 60 min, a primed-constant intravenous infusion of xenin-25 (4 or 12 pmoles/kg/min) or placebo [albumin alone (Alb)] was initiated along with Boost Plus ingestion at time zero and peptide infusion was continued until the 300-min time point. PP assays and data analysis are as described above. Group characteristics in the xenin-25 study  were very similar to those in the bethanechol study (Table 1).
For the bethanechol study, values for two-hour glucose, fasting glucose and HbA1c were generally in the order of NGT<IGT<T2DM (Table 1). Body mass index, age, and gender were not statistically different between groups. Six of the 9 subjects with T2DM were treated with metformin. Of subjects with T2DM, two required insulin before one visit and two required insulin before three visits. Eight patients were withdrawn from the study because: one had difficulty swallowing the ACM, two had low baseline blood pressure measurements, two were lost to follow-up, one had anemia, one experienced hypertension during the recovery time and one had a normal OGTT but was taking metformin. Basal values for each parameter were typically similar for each individual at each of their visits but a post hoc analysis of data revealed that subject #179 (T2DM) had not fasted before the 0 mg visit. Because this subject had received the 25 mg dose of bethanechol, data for this visit were used in place of the placebo. No studies were aborted due to bradycardia (<50 beats per min), symptomatic hypotension or lightheadedness, nausea, vomiting or diarrhea. However, with the 100 mg dose of bethanechol, 2 subjects (1 each with IGT and T2DM) experienced an asymptomatic drop in mean arterial pressure and 1 subject (with IGT) experienced an asymptomatic increase in heart rate. These 3 subjects were not administered the 150 mg dose of bethanechol.
The postprandial PP response is similar in humans with and without T2DM
In the bethanechol study, fasting PP levels were not different between groups (Table 1). As shown in Fig 3A–3C, administration of the placebo (with water) caused small changes in PP levels over the next 60 min (i.e. prior to meal ingestion). With meal ingestion, levels rapidly increased in all 3 groups, peaked by 30 min and then slowly returned to baseline values by ~180 min. Group differences in the PP AUCs (Fig 4A–4C) and iAUCs (Fig 5A–5C) from 0 to 300 min in response to placebo did not reach statistical significance (p = 0.30 and p = 0.34, respectively, by 1-way ANOVA).
Subjects with NGT, IGT, and T2DM were administered separate meal tolerance tests with placebo (blue dots) or bethanechol at a dose of 50 mg (green squares), 100 mg (yellow triangles), or 150 mg (inverted red triangles). Plasma levels of PP (Panels A-C), intact GLP-1 (Panels D-F), and total GIP (Panels G-I) were measured at the indicated times before and after meal ingestion. Values represent group means ± SEMs for subjects with NGT (Panels A, D, G), IGT (Panels B, E, H), and T2DM (Panels C, F, I). The number of subjects receiving the 0, 50, 100, and 150 mg dose of bethanechol is indicated for each group. Differences in subject number within each group are because several subjects did not receive the 150 mg dose. GLP-1 and GIP levels were only measured in samples from individual subjects receiving both the 0 mg and 150 mg doses of bethanechol. P values for the bethanechol effect (B) and for bethanechol-time interaction (B*T) are indicated in each panel. Statistically significant P values for individual time points are shown if the bethanechol or bethanechol-time interaction was significant.
Areas under the curve (AUC) were calculated for each individual at the indicated dose of bethanechol from data shown in Fig 3. Group means ± SEM are shown. P values for a bethanechol effect were determined using the mixed effects model and are shown in each panel. Statistically significant p values for each dose of bethanechol versus placebo are shown only if the bethanechol effect is significant.
Bethanechol increases the PP response in humans with IGT, but not NGT or T2DM
To determine the effects of bethanechol within each group, the meal tolerance tests were repeated with escalating doses of bethanechol, with each dose administered during a subsequent visit. Compared to placebo, ingestion of 50, 100, and 150 mg bethanechol had no statistically significant effect on PP levels before meal ingestion (Fig 3A–3C; p = 0.76, p = 0.36, and p = 0.35 in groups with NGT, IGT, and T2DM, respectively). However, after meal ingestion, there were progressive and bethanechol-dose-dependent increases in the PP levels (Fig 3B; p<0.004), AUCs (Fig 4B; p = 0.014), and iAUCs (Fig 5B; p = 0.01) in the subjects with IGT, but not NGT or T2DM (Figs 3A–3C, 4A–4C and 5A–5C). The dose-response in AUCs was linear in the IGT group (p = 0.0014). With the 150 mg dose versus the placebo, the AUC and iAUC in the IGT group increased 1.45-fold (p = 0.0037) and 2-fold (p = 0.0026), respectively.
Bethanechol increases the GLP-1 response in humans with IGT and T2DM, but not NGT
Next, intact GLP-1 levels were measured in samples from the 0 mg and 150 mg study visits. As shown in Fig 3D–3F, plasma levels of intact GLP-1 increased rapidly after meal ingestion. With placebo, levels peaked at 15–20 min, rapidly declined until 60 min, and then slowly decreased until the 300 min time point. Bethanechol slightly increased the early GLP-1 responses in the groups with NGT (Figs 3D and 6D) and IGT (Figs 3E and 6E) but this response was statistically significant only in the IGT group (p = 0.02 for a bethanechol-time interaction). In contrast, bethanechol increased the postprandial GLP-1 response in the group with T2DM (Figs 3F and 6F; p = 0.037 for bethanechol effect). The 300-min AUC for GLP-1 was also significantly increased by the 150 mg dose of bethanechol (p = 0.029; Fig 4F).
Bethanechol increases early GIP release in humans with NGT, but not IGT or T2DM
As shown in Figs 3G–3I and 6D–4F, plasma levels of total GIP increased rapidly after meal ingestion, peaked 30 to 60 min later and then continually declined until the 300 min time point. With the placebo, postprandial peak GIP levels were similar in all 3 groups (Fig 3G–3I). However, bethanechol administration increased peak GIP levels in the group with NGT, but not IGT or T2DM. A repeated measures 2-way ANOVA revealed that the interaction between time and bethanechol was highly significant in the group with NGT (p = 0.02 for a bethanechol-time interaction), but not IGT (p = 1.0) or T2DM (p = 0.44). The 300-min GIP AUC (Fig 4G–4I) and iAUC (Fig 5G–5I) were not significantly altered by 150 mg bethanechol in any group.
Bethanechol has no effect on glucose homeostasis in humans with NGT, IGT, and T2DM
Consistent with progressively worsening glucose tolerance, fasting and postprandial plasma glucose levels (Fig 7A–7C) and ISRs (Fig 7D–7F) as well as their respective AUCs (Fig 8A–8F) and iAUCS (Fig 9A–9F) with the placebo were in the order of T2DM>IGT>NGT (p<0.0001 and p = 0.04 for respective glucose and ISR AUCs). Moreover, there were rapid and transient increases in the postprandial glucagon responses in the order of T2DM>IGT>NGT (Fig 7G–7I) after which glucagon levels decreased (p = 0.05 and p = 0.46 for respective 0–60 and 0–300 minute AUCs). Unlike PP, GLP-1, and GIP responses, there were no statistically significant bethanechol or bethanechol-time interactions affecting glucose, ISRs, or glucagon levels (Fig 7A–7I), AUCs (Fig 8A–8I), and iAUCs (Fig 9A–9I) within any group. Similarly, bethanechol did not affect the levels, times to peak value, AUCs and iAUCs for plasma ACM (Figs 7J–7L, 8J–8L and 9J–9L) indicating that it did not affect the rate of gastric emptying.
Plasma glucose (panels A-C), glucagon (panels G-I), and ACM (panels J-L) levels and insulin secretion rates (panels D-F) were determined during meal tolerances as described in Fig 3. Values represent group means ± SEMs for subjects with NGT (Panels A, D, G, J), IGT (Panels B, E, H, K), and T2DM (Panels C, F, I, L). Symbols are the same as in Fig 4. P values for bethanechol and for bethanechol-time interaction are indicated in each panel. Data for the 100 mg dose of bethanechol for one subject with T2DM was excluded from the analysis because baseline values were 4.5 standard deviations from the mean.
Areas under the curves (AUCs) were calculated from data shown in Fig 7. Group means ± SEM are shown. There were no statistically significant differences in any response within each group.
Symptomatic effects of bethanechol
Subject surveys indicated that there were no dose-dependent effects of bethanechol on diarrhea, nausea, vomiting, chest pains, dizziness, heart palpitations, shortness of breath, fever, chills, blurred vision, or changes in salivation, sweating, or frequency of urination. Meal ingestion transiently reduced the group mean arterial blood pressure (7–10 mmHg) and increased resting heart rate (~8 beats per min) in each group but bethanechol had little additional dose-dependent effect on either outcome (Fig 10). The absence of significant effects on sweating, blood pressure, and heart rate are consistent with previous studies by others administering similar doses of oral bethanechol [55,56].
Mean arterial blood pressure (MAP) was calculated from systolic (SP) and diastolic (DP) blood pressure readings using the formula MAP = DP+1/3(SP-DP). Values represent group means ± SEMs for subjects with NGT (Panels A, D), IGT (Panels B, E), and T2DM (Panels C, F). Symbols and statistics for bethanechol-time interaction are as described in Fig 3.
Xenin-25 infusion increases the PP response in humans with NGT, IGT, and T2DM
We previously showed that intravenous administration of an intestinal peptide called xenin-25 during intravenous graded glucose infusions increases the PP response in humans with NGT, IGT, and T2DM . This PP response is completely inhibited by atropine sulfate (manuscript in preparation) indicating it is mediated by increased cholinergic input to islets. Thus, we determined if xenin-25 also amplifies postprandial PP responses and if these responses are larger than those measured after administration of 150 mg bethanechol. Archived plasma samples from our previous xenin study were used for these measurements . A detailed examination of the PP response in humans with NGT showed that basal PP concentrations and postprandial levels, temporal profiles, and 300-min AUCs were similar to those measured during placebo administration in the bethanechol study (Not Shown; 300-min PP AUCs = 15,747 ± 2800 vs 17,705 ± 4476, respectively). Hence, timing and route of drug administration does not affect the PP response. Further, the PP response was increased by xenin-25 infusion in a dose-dependent fashion (Fig 11A and 11B). Because PP levels peak ~30 min after meal ingestion, PP levels were measured in additional samples from subjects with IGT and T2DM prepared 30 min after boost plus ingestion and compared to those from the bethanechol study (150 mg dose). As shown in Fig 11C–11E, infusion of xenin-25 increased PP levels to a much greater extent than the highest dose of bethanechol in subjects with NGT, IGT, and T2DM (Fig 3A–3C). However, infusion of xenin had no effect on postprandial ISRs in any group when normalized to plasma glucose levels .
(A) PP levels were measured at the indicated times in subjects with NGT during meal tolerance tests with albumin alone (Alb; blue dots) or xenin-25 at a dose of 4 pmol/kg/min (Lo-Xen; yellow squares) or 12 pmol/kg/min (Hi-Xen; red triangles). As in the bethanechol study, the liquid meal was ingested from 0–3 min. (B) 0–300 min AUCs for data in panel A are shown. (C-E) PP levels were measured in the samples collected before (0 min) and 30 min after meal ingestion in humans with NGT (Panel C), IGT (Panel D), and T2DM (Panel E) during a primed-continuous infusion of xenin-25 (12 pmoles/kg/min).
Bethanechol does not cross the blood-brain barrier and thus, can be used to study the effects of muscarinic cholinergic signaling in the periphery on glucose homeostasis. The PP response was used as a surrogate measure for cholinergic input to islets. Results showed that PP levels, AUCs, and iAUCs were not different between groups during administration of the placebo. Bethanechol had relatively little effect on PP levels before meal ingestion but significantly amplified postprandial PP levels, AUCs (1.45-fold), and iAUCs (2-fold) in humans with IGT in a dose-dependent fashion. In contrast, bethanechol did not increase the PP response in the groups with NGT and T2DM. Thus, PP cells in humans with IGT, but not NGT or T2DM, develop hypersensitivity to cholinergic input. Even though the highest dose of bethanechol increased the PP response in the IGT group, there was no corresponding effect on plasma glucose or glucagon concentrations, ISRs, or gastric emptying. In contrast to bethanechol, infusion of xenin-25 at 12 pmoles/kg/min increased the PP response nearly 4-fold in all 3 groups. As we previously reported, this dose of xenin-25 delayed gastric emptying but did not affect ISRs when normalized to plasma glucose levels . These collective results suggest that increasing cholinergic input to islets plays only a minor role in regulating postprandial insulin and glucagon secretion in humans.
Consistent with our earlier  and present (Figs 3, 4 and 5) studies, others have also reported that PP levels and responses are not increased in humans with T2DM [65,66]. In contrast, some studies concluded that PP levels and responses are increased in T2DM [67,68]. However, PP levels and responses are known to dramatically increase with age [65,69] and protein and fat elicit much larger PP responses compared to oral glucose . Additionally, in the present study an extraction step was incorporated to remove contaminants that artifactually increase PP levels in human plasma samples . Further, a dual-antibody sandwich ELISA that does not cross react with highly related NPY and PYY or with other gut peptides was used for our PP measurements. This type of ELISA exhibits greater antigen specificity than that obtained using a single antibody RIA. Hence, the reason(s) for the discrepancies in PP responses in humans with versus without T2DM is likely due to differences in assay procedures and specificities but also raises the intriguing possibility that propancreatic polypeptide may be differentially processed to peptides with unique bioactivities in humans with NGT, IGT, and T2DM.
An unexpected finding was that bethanechol had complex and differential effects on GLP-1 and GIP release in humans with NGT, IGT, and T2DM during mixed meal tolerance tests. Specifically, bethanechol increased the GLP-1 response in the groups with IGT and T2DM (T2DM > IGT) but increased GIP release only in the group with NGT. Our study was not designed to determine if these were direct or indirect effects of bethanechol action on intestinal K and/or L cells. In spite of this, changes in the patterns for GIP and GLP-1 release were not accompanied by alterations in profiles for ISRs, rate of gastric emptying, plasma glucagon levels, and glucose concentrations within any group. This is consistent with previous results from our laboratory showing that postprandial, endogenously released, circulating GLP-1 plays little role in regulating postprandial insulin secretion in humans . That bethanechol altered GLP-1 release but not insulin secretion in the group with T2DM is also consistent with the well-known observation that the response to endogenous incretins is blunted in T2DM . Thus, strategies to increase release of endogenous GLP-1 in humans with T2DM may not represent an effective intervention for treating this disease. It is also important to note that the GLP-1 response is as rapid as that for GIP even though most GLP-1 producing cells are located in the distal intestine whereas GIP-producing cells reside predominantly in the proximal gut. This suggests that early GLP-1 release is mainly under neural rather than nutritional control. Consistent with this idea, a host of neurotransmitters and peptides increase GLP-1 release in the vascularly perfused rat ileum .
Several limitations to the current study should be addressed. First, bethanechol had only modest effects on the PP, GIP, and GLP-1 responses within each effected group. Although the doses used in our study are known to affect the stretch response of the bladder muscle without eliciting hypotension or bradycardia [55,56], it is possible that higher doses would have exerted greater effects and thus, possibly altered glucose levels, ISRs, glucagon concentrations, gastric emptying, or other factors that regulate glucose homeostasis. However, the bethanechol doses used in the current study revealed that physiologically relevant changes in endogenous PP, GIP, and GLP-1 release do not affect postprandial glucose homeostasis. This result could have potentially been masked by a greater degree of cholinergic agonism. However, that xenin-25 profoundly increased PP release but not ISRs argues against this. A second limitation is that bethanechol activates all muscarinic receptors in the periphery and thus, the current study assessed the effects of stimulating multiple superimposed cholinergic signaling pathways and it is possible that numerous positive and negative responses exactly offset each other. If this is the case, it would suggest that cholinergic signaling could possibly act to maintain a pre-established “set-point” for glucose homeostasis without increasing insulin secretion. It should be noted that of the 5 known muscarinic acetylcholine receptors, it is the M3 subtype that increases insulin release from beta cells [7–12]. Thus, even though a xenin-25-mediated increase in cholinergic input to islets did not amplify ISRs, M3 subtype-specific agonists may still represent a therapeutic strategy for increasing ISRs in T2DM. Third, the lower doses of bethanechol may not have remained active for the duration of the study though it is clear that the 150 mg dose had effects in all groups and the PP response in the group with IGT was bethanechol dose-dependent. Finally, our results compared the effects of cholinergic signaling on postprandial ISRs and glucagon levels in humans with NGT, IGT, and T2DM. However, our results may not be applicable to other metabolic or pathophysiologic conditions.
In spite of the limitations, our results indicate that bethanechol has different effects on PP, GIP, and GLP-1 release in humans with NGT versus IGT versus T2DM. Even with these differences, bethanechol had no measureable effect on glucose homeostasis in any group. As shown in this as well as our earlier study , infusion of xenin-25 in humans with NGT increased the postprandial PP response nearly 4-fold and reduced the GLP-1 response 6-fold but had no effect on insulin or glucagon responses. These results suggest that cholinergic signaling and circulating GLP-1 play only minor roles in regulating glucose homeostasis in humans. Because islet responses to endogenously released circulating incretins  but not to exogenously infused GIP  or GLP-1 [72,73] are blunted in T2DM, it is critical to determine if and how GLP-1 and GIP released from respective intestinal L and K cells regulate glucose homeostasis in humans. Moreover, our data suggest that cholinergic mechanisms that regulate insulin secretion are different in humans and mice and thus, extreme caution must be exercised when extrapolating results from animals to humans.
S1 Appendix. Original IRB approved protocol for bethanechol study.
The authors would like to thank the nurses of the Clinical Research Unit at Washington University School of Medicine for administering the meal tolerance tests and Dr. Yan Yan of the Department of Surgery for helpful discussions.
Conceived and designed the experiments: BMW. Performed the experiments: SC SW JD RK LO DR. Analyzed the data: SC SW BWP MJW BMW. Contributed reagents/materials/analysis tools: BWP MJW BMW. Wrote the paper: BMW MJW BWP.
- 1. Ahren B (2000) Autonomic regulation of islet hormone secretion—implications for health and disease. Diabetologia 43: 393–410. pmid:10819232
- 2. Ahren B (2004) Neuropeptides and Insulin Secretion. In: DeFronzo RA, Ferrannini E, Keen H, Zimmet P, editors. International Textbook of Diabetes Mellitus. Hoboken, New Jersey: John Wiley & Sons, Ltd. pp. 153–163.
- 3. Ahren B (2009) Islet G protein-coupled receptors as potential targets for treatment of type 2 diabetes. Nat Rev Drug Discov 8: 369–385. nrd2782 [pii]; pmid:19365392
- 4. Ahren B (2012) Islet nerves in focus—defining their neurobiological and clinical role. Diabetologia 55: 3152–3154. pmid:23001378
- 5. Rodriguez-Diaz R, Abdulreda MH, Formoso AL, Gans I, Ricordi C, Berggren PO et al. (2011) Innervation patterns of autonomic axons in the human endocrine pancreas. Cell Metab 14: 45–54. pmid:21723503
- 6. Wice BM, Wang S, Crimmins DL, Diggs-Andrews KA, Althage MC, Ford EL et al. (2010) Xenin-25 potentiates glucose-dependent insulinotropic polypeptide action via a novel cholinergic relay mechanism. J Biol Chem 285: 19842–19853. pmid:20421298
- 7. Duttaroy A, Zimliki CL, Gautam D, Cui Y, Mears D, Wess J (2004) Muscarinic stimulation of pancreatic insulin and glucagon release is abolished in m3 muscarinic acetylcholine receptor-deficient mice. Diabetes 53: 1714–1720. pmid:15220195
- 8. Gautam D, Gavrilova O, Jeon J, Pack S, Jou W, Cui Y et al. (2006) Beneficial metabolic effects of M3 muscarinic acetylcholine receptor deficiency. Cell Metab 4: 363–375. pmid:17084710
- 9. Gautam D, Han SJ, Hamdan FF, Jeon J, Li B, Li JH et al. (2006) A critical role for beta cell M3 muscarinic acetylcholine receptors in regulating insulin release and blood glucose homeostasis in vivo. Cell Metab 3: 449–461. pmid:16753580
- 10. Gautam D, Han SJ, Duttaroy A, Mears D, Hamdan FF, Li JH et al. (2007) Role of the M3 muscarinic acetylcholine receptor in beta-cell function and glucose homeostasis. Diabetes Obes Metab 9 Suppl 2: 158–169. pmid:17919190
- 11. Renuka TR, Robinson R, Paulose CS (2006) Increased insulin secretion by muscarinic M1 and M3 receptor function from rat pancreatic islets in vitro. Neurochem Res 31: 313–320. pmid:16733808
- 12. Zawalich WS, Zawalich KC, Tesz GJ, Taketo MM, Sterpka J, Philbrick W et al. (2004) Effects of muscarinic receptor type 3 knockout on mouse islet secretory responses. Biochem Biophys Res Commun 315: 872–876. pmid:14985093
- 13. Brunicardi FC, Druck P, Seymour NE, Sun YS, Elahi D, Andersen DK (1990) Selective neurohormonal interactions in islet cell secretion in the isolated perfused human pancreas. J Surg Res 48: 273–278. 0022-4804(90)90058-A [pii]. pmid:1971030
- 14. Brunicardi FC, Sun YS, Druck P, Goulet RJ, Elahi D, Andersen DK (1987) Splanchnic neural regulation of insulin and glucagon secretion in the isolated perfused human pancreas. Am J Surg 153: 34–40. pmid:3541657
- 15. Brunicardi FC, Druck P, Sun YS, Elahi D, Gingerich RL, Andersen DK (1988) Regulation of pancreatic polypeptide secretion in the isolated perfused human pancreas. Am J Surg 155: 63–69. pmid:3277467
- 16. Brunicardi FC, Druck P, Seymour NE, Sun YS, Gingerich RL, Elahi D et al. (1989) Splanchnic neural regulation of pancreatic polypeptide release in the isolated perfused human pancreas. Am J Surg 157: 50–57. pmid:2910127
- 17. Brunicardi FC, Shavelle DM, Andersen DK (1995) Neural regulation of the endocrine pancreas. Int J Pancreatol 18: 177–195. pmid:8708389
- 18. Brunicardi FC, Goulet RJ, Sun YS, berlin SA, Elahi D, Andersen DK (1984) Neural and Hormonal Regulation of Insulin Release in the Isolated Perfused Human Pancreas. Surgical Forum 35: 214–217.
- 19. Rorsman P, Braun M (2013) Regulation of insulin secretion in human pancreatic islets. Annu Rev Physiol 75: 155–179. pmid:22974438
- 20. Teff KL, Townsend RR (2004) Prolonged mild hyperglycemia induces vagally mediated compensatory increase in C-Peptide secretion in humans. J Clin Endocrinol Metab 89: 5606–5613. pmid:15531518
- 21. Ahren B, Holst JJ (2001) The cephalic insulin response to meal ingestion in humans is dependent on both cholinergic and noncholinergic mechanisms and is important for postprandial glycemia. Diabetes 50: 1030–1038. pmid:11334405
- 22. Silvestre JS, Prous J (2005) Research on adverse drug events. I. Muscarinic M3 receptor binding affinity could predict the risk of antipsychotics to induce type 2 diabetes. Methods Find Exp Clin Pharmacol 27: 289–304. pmid:16082416
- 23. Plamboeck A, Veedfald S, Deacon CF, Hartmann B, Wettergren A, Svendsen LB et al. (2013) Characterisation of oral and i.v. glucose handling in truncally vagotomised subjects with pyloroplasty. Eur J Endocrinol 169: 187–201. pmid:23704713
- 24. ADAMS GF (1951) Intravenous glucose tolerance before and after vagotomy for peptic ulcer. Gastroenterology 17: 63–71. pmid:14802682
- 25. Kesterson JE, Powell LC, McCleery RS (1951) The effect of vagotomy on the glucose tolerance of patients with peptic ulceration. Gastroenterology 18: 230–234. pmid:14840608
- 26. Vozarova de Court , Weyer C, Stefan N, Horton M, DelParigi A, Havel P et al. (2004) Parasympathetic blockade attenuates augmented pancreatic polypeptide but not insulin secretion in Pima Indians. Diabetes 53: 663–671. pmid:14988250
- 27. Adrian TE, Besterman HS, Cooke TJ, Bloom SR, Barnes AJ, Russell RC (1977) Mechanism of pancreatic polypeptide release in man. Lancet 1: 161–163. pmid:64696
- 28. Larsson LI, Sundler F, Hakanson R (1976) Pancreatic polypeptide—a postulated new hormone: identification of its cellular storage site by light and electron microscopic immunocytochemistry. Diabetologia 12: 211–226. pmid:782992
- 29. Larsson LI, Sundler F, Hakanson R (1975) Immunohistochemical localization of human pancreatic polypeptide (HPP) to a population of islet cells. Cell Tissue Res 156: 167–171. pmid:1091353
- 30. Bergstrom BH, Loo S, Hirsch HJ, Schutzengel D, Gabbay KH (1977) Ultrastructural localization of pancreatic polypeptide in human pancreas. J Clin Endocrinol Metab 44: 795–798. pmid:849991
- 31. Baetens D, De MJ, Gepts W (1977) Immunohistochemical and ultrastructural identification of the pancreatic polypeptide-producing cell (PP-cell) in the human pancreas. Cell Tissue Res 185: 239–246. pmid:340043
- 32. Lonovics J, Devitt P, Watson LC, Rayford PL, Thompson JC (1981) Pancreatic polypeptide. A review. Arch Surg 116: 1256–1264. pmid:7025798
- 33. Chowdhury S, Wang S, Patterson BW, Reeds DR, Wice BM (2013) The combination of GIP plus xenin-25 indirectly increases pancreatic polypeptide release in humans with and without type 2 diabetes mellitus. Regulatory Peptides 187: 42–50. pmid:24183983
- 34. Schwartz TW, Holst JJ, Fahrenkrug J, Jensen SL, Nielsen OV, Rehfeld JF et al. (1978) Vagal, cholinergic regulation of pancreatic polypeptide secretion. J Clin Invest 61: 781–789. pmid:641155
- 35. Fletcher DR, Shulkes A, Bladin PH, Booth D, Hardy KJ (1983) Cholinergic inhibition of meal stimulated plasma neurotensin like immunoreactivity in man. Life Sci 33: 863–869. pmid:6888194
- 36. Fletcher DR, Shulkes A, Bladin PH, Hardy KJ (1983) The effect of atropine on bombesin and gastrin releasing peptide stimulated gastrin, pancreatic polypeptide and neurotensin release in man. Regul Pept 7: 31–40. pmid:6647889
- 37. Amland PF, Jorde R, Burhol PG, Giercksky KE (1985) Effects of atropine on GIP-induced insulin and pancreatic polypeptide release in man. Scand J Gastroenterol 20: 321–324. pmid:3890140
- 38. Layer P, Chan AT, Go VL, Zinsmeister AR, DiMagno EP (1993) Cholinergic regulation of phase II interdigestive pancreatic secretion in humans. Pancreas 8: 181–188. pmid:8096337
- 39. Schwartz TW, Rehfeld JF, Stadil F, Larson LI, Chance RE, Moon N (1976) Pancreatic-polypeptide response to food in duodenal-ulcer patients before and after vagotomy. Lancet 1: 1102–1105. S0140-6736(76)90065-9 [pii]. pmid:57509
- 40. Stern AI, Hansky J, Korman MG, Coupland G, Waugh J (1980) Pancreatic polypeptide. Release following surgery for duodenal ulcer disease. Dig Dis Sci 25: 485–488. pmid:7389536
- 41. Veedfald S, Plamboeck A, Hartmann B, Svendsen LB, Vilsboll T, Knop FK et al. (2015) Pancreatic polypeptide responses to isoglycemic oral and intravenous glucose in humans with and without intact vagal innervation. Peptides 71: 229–231. S0196-9781(15)00216-8 [pii]; pmid:26218807
- 42. Feurle GE, Hamscher G, Kusiek R, Meyer HE, Metzger JW (1992) Identification of xenin, a xenopsin-related peptide, in the human gastric mucosa and its effect on exocrine pancreatic secretion. J Biol Chem 267: 22305–22309. pmid:1429581
- 43. Hamscher G, Meyer HE, Metzger JW, Feurle GE (1995) Distribution, formation, and molecular forms of the peptide xenin in various mammals. Peptides 16: 791–797. pmid:7479317
- 44. Feurle GE (1998) Xenin—a review. Peptides 19: 609–615. pmid:9533652
- 45. Anlauf M, Weihe E, Hartschuh W, Hamscher G, Feurle GE (2000) Localization of xenin-immunoreactive cells in the duodenal mucosa of humans and various mammals. J Histochem Cytochem 48: 1617–1626. pmid:11101630
- 46. Kamiyama Y, Aihara R, Nakabayashi T, Mochiki E, Asao T, Kuwano H (2007) The peptide hormone xenin induces gallbladder contractions in conscious dogs. Neurogastroenterol Motil 19: 233–240. pmid:17300294
- 47. Leckstrom A, Kim ER, Wong D, Mizuno TM (2009) Xenin, a gastrointestinal peptide, regulates feeding independent of the melanocortin signaling pathway. Diabetes 58: 87–94. pmid:18984739
- 48. Cline MA, Nandar W, Rogers JO (2007) Xenin reduces feed intake by activating the ventromedial hypothalamus and influences gastrointestinal transit rate in chicks. Behav Brain Res 179: 28–32. pmid:17270287
- 49. Cooke JH, Patterson M, Patel SR, Smith KL, Ghatei MA, Bloom SR et al. (2009) Peripheral and central administration of xenin and neurotensin suppress food intake in rodents. Obesity (Silver Spring) 17: 1135–1143.
- 50. Kim ER, Mizuno TM (2010) Role of neurotensin receptor 1 in the regulation of food intake by neuromedins and neuromedin-related peptides. Neurosci Lett 468: 64–67. pmid:19857548
- 51. Zhang S, Hyrc K, Wang S, Wice BM (2012) Xenin-25 Increases Cytosolic Free Calcium Levels and Acetylcholine Release from a Subset of Myenteric Neurons. Am J Physiol Gastrointest Liver Physiol 303: G1347–G1355. pmid:23086920
- 52. None (1997) Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 20: 1183–1197. pmid:9203460
- 53. McGuire EA, Helderman JH, Tobin JD, Andres R, Berman M (1976) Effects of arterial versus venous sampling on analysis of glucose kinetics in man. J Appl Physiol 41: 565–573. pmid:985402
- 54. Brooks DC, Black PR, Arcangeli MA, Aoki TT, Wilmore DW (1989) The heated dorsal hand vein: an alternative arterial sampling site. JPEN J Parenter Enteral Nutr 13: 102–105. pmid:2926973
- 55. Diokno AC, Lapides J (1977) Action of oral and parenteral bethanechol on decompensated bladder. Urology 10: 23–24. pmid:327652
- 56. Lapides J, Friend CR, Jemian EP, Sonda LP (1963) Comparison of Action of Oral and Parenteral Bethanechol Chloride upon the Urinary Bladder. Invest Urol 1: 94–97. pmid:14076313
- 57. Gorsky M, Epstein JB, Parry J, Epstein MS, Le ND, Silverman S Jr (2004) The efficacy of pilocarpine and bethanechol upon saliva production in cancer patients with hyposalivation following radiation therapy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 97: 190–195. pmid:14970777
- 58. Jham BC, Teixeira IV, Aboud CG, Carvalho AL, Coelho MM, Freire AR (2007) A randomized phase III prospective trial of bethanechol to prevent radiotherapy-induced salivary gland damage in patients with head and neck cancer. Oral Oncol 43: 137–142. pmid:16798061
- 59. Everett HC (1975) The use of bethanechol chloride with tricyclic antidepressants. Am J Psychiatry 132: 1202–1204. pmid:1166898
- 60. Chowdhury S, Reeds DN, Crimmins DL, Patterson BW, Laciny E, Wang S et al. (2014) Xenin-25 delays gastric emptying and reduces postprandial glucose levels in humans with and without type 2 diabetes. Am J Physiol Gastrointest Liver Physiol 306: G301–G309. pmid:24356886
- 61. Wice BM, Reeds DR, Tran H, Crimmins DL, Patterson BW, Dunai J et al. (2012) Xenin-25 Amplifies GIP-Mediated Insulin Secretion in Humans with Normal and Impaired Glucose Tolerance, but not Type 2 Diabetes. Diabetes 61: 1793–1800. pmid:22522617
- 62. Van Cauter E, Mestrez F, Sturis J, Polonsky KS (1992) Estimation of insulin secretion rates from C-peptide levels. Comparison of individual and standard kinetic parameters for C-peptide clearance. Diabetes 41: 368–377. pmid:1551497
- 63. Sparacino G, Pillonetto G, Capello M, De Nicolao G., Cobelli C (2002) WINSTODEC: a stochastic deconvolution interactive program for physiological and pharmacokinetic systems. Comput Methods Programs Biomed 67: 67–77. pmid:11750948
- 64. Sparacino G, Cobelli C (1996) A stochastic deconvolution method to reconstruct insulin secretion rate after a glucose stimulus. IEEE Trans Biomed Eng 43: 512–529. pmid:8849464
- 65. Adrian TE (1978) Pancreatic polypeptide. J Clin Pathol Suppl (Assoc Clin Pathol) 8: 43–50.
- 66. Amland PF, Jorde R, Aanderud S, Burhol PG, Giercksky KE (1985) Effects of intravenously infused porcine GIP on serum insulin, plasma C-peptide, and pancreatic polypeptide in non-insulin-dependent diabetes in the fasting state. Scand J Gastroenterol 20: 315–320. pmid:3890139
- 67. Chia CW, Odetunde JO, Kim W, Carlson OD, Ferrucci L, Egan JM (2014) GIP contributes to islet trihormonal abnormalities in type 2 diabetes. J Clin Endocrinol Metab 99: 2477–2485. pmid:24712564
- 68. Tsuda K, Seino Y, Mori K, Seino S, Takemura J, Kuzuya H et al. (1983) Hyperfunction of the entero-PP axis in non-insulin dependent diabetes mellitus. Horm Metab Res 15: 581–585. pmid:6363240
- 69. Brimnes DM, Rasmussen BK, Hilsted L, Jensen R, Hilsted J (1997) Basal serum pancreatic polypeptide is dependent on age and gender in an adult population. Scand J Clin Lab Invest 57: 695–702. pmid:9458492
- 70. Nauck M, Stockmann F, Ebert R, Creutzfeldt W (1986) Reduced incretin effect in type 2 (non-insulin-dependent) diabetes. Diabetologia 29: 46–52. pmid:3514343
- 71. Dumoulin V, Dakka T, Plaisancie P, Chayvialle JA, Cuber JC (1995) Regulation of glucagon-like peptide-1-(7–36) amide, peptide YY, and neurotensin secretion by neurotransmitters and gut hormones in the isolated vascularly perfused rat ileum. Endocrinology 136: 5182–5188. pmid:7588257
- 72. Campbell JE, Drucker DJ (2013) Pharmacology, physiology, and mechanisms of incretin hormone action. Cell Metab 17: 819–837. pmid:23684623
- 73. Drucker DJ (2013) Incretin action in the pancreas: potential promise, possible perils, and pathological pitfalls. Diabetes 62: 3316–3323. db13-0822 [pii]; pmid:23818527