Xenin-25 increases cytosolic free calcium levels and acetylcholine release from a subset of myenteric neurons

Abstract

Xenin-25 (Xen) is a 25 amino acid neurotensin-related peptide reportedly produced with glucose-dependent insulinotropic polypeptide (GIP) by a subset of K cells in the proximal gut. We previously showed exogenously administered Xen, with GIP but not alone, increases insulin secretion in humans and mice. In mice, this effect is indirectly mediated via a central nervous system-independent cholinergic relay in the periphery. Xen also delays gastric emptying, reduces food intake, induces gall bladder contractions, and increases gut motility and secretion from the exocrine pancreas, suggesting that some effects of Xen could be mediated by myenteric neurons (MENs). To determine whether Xen activates these neurons, MENs were isolated from guinea pig proximal small intestines. Cells expressed neuronal markers and exhibited typical neuron-like morphology with extensive outgrowths emanating from cell bodies. Cytosolic free Ca²⁺ levels ([Ca²⁺]_i) were measured using Fura-2. ATP/UTP, KCl, and forskolin increased [Ca²⁺]_i in 99.6%, 92%, and 23% of the MENs imaged, respectively, indicating that they are functional and activated by nucleotide receptor signaling, direct depolarization, and cAMP. [Ca²⁺]_i increased in only 12.7% of MENs treated with Xen. This rise was blocked by pretreatment with EGTA, diazoxide, SR48692, and neurotensin. Thus the Xen-mediated increase in [Ca²⁺]_i involves influx of extracellular Ca²⁺ and activation of neurotensin receptor-1 (NTSR1). Xen also increased acetylcholine release from MENs. Amylin, produced by β-and enteroendocrine cells, delays gastric emptying and increased [Ca²⁺]_i almost exclusively in Xen-responsive MENs. Immunohistochemistry demonstrated NTSR1 expression in human duodenal MENs. Thus myenteric rather than central neurons could mediate some effects of Xen and amylin.

xenin-25 (Xen) is a 25-amino-acid neurotensin-related peptide reportedly produced along with glucose-dependent insulinotropic polypeptide (GIP) by a subset of K cells (3, 11). Animal studies from a number of laboratories have demonstrated that exogenously administered Xen delays gastric emptying (29), reduces food intake (1, 8, 9, 28, 36), induces gall bladder contractions (27), increases gut motility (13), and augments secretion from the exocrine pancreas (12, 17, 46). The effects of Xen on gut motility were indirect because they were abolished in cholecystectomized dogs (27).

Our laboratory generated and characterized transgenic mice that lack K cells, and thus K cell products, by expressing an attenuated diphtheria toxin A chain using regulatory elements from the GIP promoter and gene (2, 57). These “GIP/DT” mice exhibit a blunted insulin secretory response to exogenously administered GIP. In the GIP/DT mice, as well as in another mouse model that exhibits a blunted response to GIP, injection of Xen increased the effects of GIP on insulin secretion in vivo even though it had no effect by itself (57). In contrast, Xen failed to increase GIP-mediated insulin release in a variety of in vitro studies, indicating that Xen indirectly increases insulin secretion. We have recently shown that Xen also increases the insulin secretory response to GIP in humans with normal or impaired glucose tolerance, but not type 2 diabetes (56). In mice, activation of M3 muscarinic receptors on the islet β-cell augments insulin release (10, 18–20, 49, 59), and, in contrast to Xen, carbachol synergistically increases GIP-mediated insulin release in vitro (57). Moreover, the in vivo effects of Xen on insulin release in mice are completely blocked by atropine methyl bromide, indicating that a cholinergic relay requiring muscarinic acetylcholine receptors in the periphery indirectly relays the Xen-initiated signal to islet β-cells. Results of immunohistochemical studies examining c-fos expression in the mouse brain suggest that this relay does not utilize parasympathetic neurons (57). Intriguingly, the effects of Xen on gallbladder contraction were mediated by a cholinergic neural pathway that was not abolished by truncal vagotomy (27).

It is well known that myenteric neurons (MENs) play an important role in regulating gastrointestinal function including gut motility, fluid absorption and secretion, local blood flow, and secretion of intestinal hormones (17a). Thus some of the effects of Xen could potentially be mediated by enteric rather than central neurons. In this study, we show that Xen can in fact increase cytosolic free calcium levels ([Ca²⁺]_i) and acetylcholine release from a subset of MENs in the proximal small intestine of adult guinea pigs, suggesting that MENs may play an important role in mediating some of the physiological responses to Xen.

MATERIALS AND METHODS

Isolation and culturing of MENs.

All animal procedures were approved by the Washington University Animal Studies Committee. Adult Hartley guinea pigs of both sexes weighing 451–500 g were purchased from Charles River Laboratories (Wilmington, MA) and euthanized by intraperitoneal injection of pentobarbital sodium (150 mg/kg). Because the MENs of interest are located in the duodenum and stomach, only the first 20 cm of the small intestine were used for MEN isolations. MENs were isolated essentially as described (38, 39, 48, 58). Briefly, the proximal small intestine was removed en bloc, flushed with ice-cold Krebs solution, opened longitudinally, and thoroughly cleansed. The longitudinal muscle was removed with the myenteric plexus still attached, minced, and then digested for 1 h at 37°C in collagenase (Type 2, 1.5 mg/ml; Worthington Biochemical, Lakewood, NJ) in oxygenated Krebs Buffer. Tissue was then triturated to disperse cells. After being washed, cells were resuspended in Medium 199 containing 10% FBS, 25 mM glucose, nerve growth factor (100 ng/ml; Invitrogen, Carlsbad, CA), Gentamicin (10 mg/ml; Invitrogen), and a mixture of antibiotic-antifungal drugs (Invitrogen; cat. no. 15240-096). Isolated cells were plated onto glass-bottom culture dishes (MatTek, Ashland, MA) previously coated for 3 h with Poly-L-lysine hydrobromide (20 μg/ml; Sigma Chemical, St. Louis, MO) followed by 1 h with Matrigel (1:10 dilution; BD, Bedford, MA). Cells were cultured at 37°C in an atmosphere of 100% humidity and 5% carbon dioxide/balance air. Starting 24 h later, 10 μM cytosine arabinoside (Sigma Chemical) was included in the culture media to kill proliferating nonneuronal cells. Thereafter, media were changed every 2 days.

Immunohistochemistry.

MENs were isolated, cultured for 7 days, and fixed for 20 min in 4% freshly prepared paraformaldehyde (PGP9.5) or 10 min in ice-cold 100% methanol (neuronal class III β-tubulin). After being blocked in PBS containing 0.3% Triton X-100, cells were incubated for 1 h at room temperature with rabbit antineuronal class III β-tubulin (1:100; Covance, Emeryville, CA) or rabbit anti-PGP9.5 (1:100; EMD Millipore, Billerica, MA). A single section of human duodenum was deparaffinized, subjected to antigen retrieval using EDTA (1 mM, pH 8), blocked, and then incubated for 3 days at 4°C with goat anti NTSR1 (Santa Cruz Biotechnology, Santa Cruz, CA) plus rabbit antivesicular acetylcholine transporter (VAChT; Chemicon/EMD Millipore). Bound primary antibodies were detected using DyLight 549 (red) donkey anti-rabbit and/or DyLight 488 (green) donkey anti-goat (1:500; Jackson ImmunoResearch, West Grove, PA). Nuclei were counterstained with bis-benzimide.

Ratiometric calcium imaging and data analysis.

All imaging studies were performed 6 or 7 days after MENs were isolated and carried out at room temperature in a HEPES-buffered salt solution (HBSS) containing, in mM: 140 NaCl, 5.4 KCl, 1 NaH₂PO₄, 1.8 CaCl₂, 1 MgSO₄, 12 HEPES, and 5.5 D-glucose, pH 7.4 ± 0.1. Cells were loaded with Fura-2 by incubation for 60 min with 10 μM acetoxymethyl (AM) ester (Invitrogen) and 0.1% Pluronic F-127 (Invitrogen) in HBSS (pH 7.2) at room temperature, washed with HBSS, and incubated for another 60 min to allow for ester hydrolysis. After being loaded, cells were imaged on an inverted microscope (Nikon Eclipse TE300; Nikon, Melville, NY) equipped with a cooled CCD camera (Cooke, Auburn Hill, MI) using a ×20/0.45 Plan Fluor objective (Nikon). The fluorescence excitation (75 W xenon arc lamp) was provided by band-specific filters (340 and 380 nm; Semrock, Rochester, NY) in combination with a XF73 dichroic beam splitter (Omega Optical, Brattleboro, VA). Pairs of images were collected at 5-s intervals at alternate excitation wavelengths. After the matching background was subtracted, the image intensities were divided by one another to yield ratio values for individual cells. Because a combination of ATP plus UTP (100 μM each) increased [Ca²⁺]_i in >99% of the MENs, cells were stimulated with these nucleotides at the end of each experiment to confirm cell viability. Cells lacking a nucleotide response were excluded from further analysis. [Ca²⁺]_i in individual cells was estimated based on the formula: [Ca²⁺]_i = KD × B × (R − R_min)/(R_max − R), where KD is the indicator's dissociation constant for Ca²⁺ (0.22 μM); R is ratio of fluorescence intensity at two different wavelengths (340/380 nm); R_max and R_min are the ratios of Ca²⁺-free and Ca²⁺-bound Fura-2, respectively; and B is the ratio of the fluorescence intensity of the second excitation wavelength at zero and saturating Ca²⁺ concentrations (23). The calibration constants (R_min, R_max, and B) were determined on the same setup in calibration buffers (Invitrogen) containing either 10 mM EGTA or 39.6 μM free Ca²⁺ using Fura-2/K⁺ salt (Invitrogen). The ratio values were plotted against time, and the presence of significant ratio changes in individual cells was detected using the Grubb's outlier test [α = 0.05 (21, 22)]. The number of cells in which at least one significant [Ca²⁺]_i change occurred after treatment was then compared with the total number of ATP/UTP-responding cells tested, and the fraction of responding cells was calculated for each experimental condition. The outlier analysis was performed using a scientific graphing program SigmaPlot (Systat, Chicago, IL). The proportions of responding vs. nonresponding cells were compared using z-tests. Comparisons of [Ca²⁺]_i levels were performed using two-tailed Student's t-tests.

Acetylcholine release assays.

MENs were isolated and cultured for 7 days in 12-well tissue culture dishes as described above. MENs were incubated for 30 min with [³H]-choline (Perkin-Elmer, Shelton, CT), washed to completely remove extracellular [³H]-choline, treated with the indicated stimuli for 10 min, and [³H]-acetylcholine released into the assay buffer measured by liquid scintillation counting as described (44, 45). Acetylcholine release was normalized to cellular [³H] content.

Activation of Panc1 human exocrine pancreas cells.

Panc1 human exocrine pancreatic carcinoma cells were grown in Dulbecco's modified minimal essential medium containing 10% FBS at 37°C in an atmosphere of 100% humidity and 5% carbon dioxide/balance air (57). For experiments, cells were cultured overnight in serum-free media containing 0.1% BSA. Cells were then treated for 20 min at 37°C with the indicated drug. After being washed with ice-cold PBS, RNA was isolated using Nucleospin RNAII columns (Macherney-Nagel; Mountain View, CA) and cDNA synthesized using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). TaqMan assays (Applied Biosystems) were used to quantitate human c-fos mRNA levels (Assay Hs99999140_m1), which were normalized to human β-actin mRNA levels in the same sample (Assay 4352935E).

Peptides/drugs.

Unless specified otherwise, the following drugs were used at the indicated final concentration. Potassium chloride (50 mM), forskolin (Fk; 2 μM), diazoxide (100 μM), ATP (100 μM), and UTP (100 μM) were from Sigma Chemical. Xenin-25 (1 μM) and human GIP (1 μM) were from Bachem Americas (Torrance, CA). Neurotensin (1 μM) and rat amylin (100 nM) were from American Peptide (Sunnyvale, CA). SR48692 (10 μM) was from Axon Medchem (Groningen, The Netherlands). Xenin-25 with an additional cysteine residue added onto the NH₂ terminus (C-Xen) or COOH terminus [(Xen-C) 1 μM each] were from Biomolecules Midwest (Waterloo, IL). Human albumin (0.1%; Flexbumin, ASD Healthcare, Frisco, TX) was included in all buffers containing peptides to prevent sticking of peptides to plasticware.

RESULTS

Isolated MENs are physiologically functional.

MENs were isolated from the myenteric plexus from the proximal guinea pig small intestine. After 1 wk in culture, nearly all cells exhibited typical neuron-like morphology with extensive outgrowths emanating from the cell bodies (Fig. 1A). Cells also expressed neuronal class III β-tubulin (Fig. 1, C, E, and G) and PGP9.5 (Fig. 1, D, F, and H), confirming their neuronal properties. We next examined the ability of the MENs to respond to known excitatory stimuli by monitoring changes in [Ca²⁺]_i using microspectrofluorometry (Table 1 and Fig. 2, A–F). The Fura-2-loaded cells also exhibited neuronal-like morphology (compare Fig. 1, A and B). ATP and UTP (100 μM each), which bind to nucleotide receptors and increase calcium release from intracellular stores (6, 25, 53), increased the Fura-2 ratio in 99.6% of the 1845 MENs imaged (Fig. 2A, Table 1) and augmented average [Ca²⁺]_i from 0.055 ± 0.005 μM to a peak concentration of 0.23 ± 0.014 μM (P < 0.0001). KCl (50 mM) also increased the Fura-2 ratio in 92% of the 74 cells imaged (Table 1 and Fig. 2C) and augmented [Ca²⁺]_i from 0.062 ± 0.004 μM to a peak concentration of 0.26 ± 0.033 μM (P < 0.0001; Fig. 2D). Forskolin (Fk; 2 μM) increased cAMP levels and increased the Fura-2 ratio in 23% of the 67 MENs imaged (Table 1 and Fig. 2E) and augmented [Ca²⁺]_i from 0.055 ± 0.001 μM to a peak level of 0.17 ± 0.025 μM (P < 0.019; Fig. 2F). Thus essentially all Fura-2-labeled MENs were viable and responsive to excitatory stimuli following 1 wk in culture.

Fig. 1.

Primary myenteric neurons (MENs) exhibit typical neuron-like properties. MENs were isolated from adult guinea pig proximal small intestines and cultured for 6–7 days. A and B: phase contrast and Fura-2 fluorescence (excitation at 380 nm) images, respectively, of the same field of Fura-2-loaded cells. Similar morphology has been observed in essentially all MEN preparations. Separate cultures of MENs were fixed and stained for the neuronal markers (Red) neuronal class III β-tubulin (C) or PGP9.5 (D). Nuclei were counterstained blue (E and F). G and H: merged images of C plus E and D plus F, respectively.

Table 1.

Xenin-25 increases cytosolic free calcium levels in myenteric neurons

Treatment	%	Number
ATP/UTP	99.6	1838/1845
ATP/UTP + SR48692	100	186/186
ATP/UTP + EGTA	94	171/182
ATP/UTP + Diazoxide	99.5	186/187
KCl	92	68/74
KCl + EGTA	0	0/64
Fk	23	15/67
Xenin-25	12.7	114/1134
Xenin-25 (30 nM)	9.3	5/54
Xenin-25 + SR48692	0	0/186
Xenin-25 + EGTA	0.5	1/182
Xenin-25 + Diazoxide	1.1	2/187
Cys-Xenin	11.8	12/102
Xenin-Cys	1.6	1/62
Neurotensin	9.3	45/486
Amylin	19.4	14/72
*Xenin-25 after Amylin	93	13/14
GIP	1.7	11/642
Insulin	0	0/104

Data indicate the percentage and number (responsive/imaged) of myenteric neurons (MENs) that exhibited an increase in [Ca²⁺]_i in response to the indicated treatment.

^*Indicates the number of amylin-positive cells that responded to a subsequent treatment with Xenin-25. GIP, glucose-dependent insulinotropic polypeptide.

Fig. 2.

Primary MENs are physiologically functional. MENs were imaged before and after addition of ATP plus UTP (A, B), KCl (C, D), or forskolin (E, F). The average Fura-2 340/380 excitation ratios ± SE are plotted vs. time in A, C, and E. Average ± SE basal and peak cytosolic free calcium concentrations are shown in B, D, and F.

Xenin-25 increases cytosolic free calcium levels in only a subset of MENs.

Studies were next conducted using MENs treated with 1 μM Xen. As shown in Fig. 3A and Table 1, the first addition of Xen (Xen-1) rapidly and transiently increased the Fura-2 ratio in only 12.7% of 1,134 MENs imaged. Average resting [Ca²⁺]_i increased from 0.034 ± 0.002 μM to a peak concentration of 0.099 ± 0.010 μM (P < 0.002; Fig. 3B). Importantly, the MENs that did not respond to Xen were functionally viable because they responded to a subsequent stimulation with ATP/UTP (at ∼600 s). Interestingly, basal [Ca²⁺]_i was lower in the Xen-positive MENs (0.034 μM; Fig. 3B) compared with the entire population of MENs (0.5–0.6 μM; Fig. 2). Addition of 30 nM Xen activated a similar percentage of MENs [9.3% (5/54 MENs); P = 0.60 vs. 1 μM Xen]. Although Xen and GIP cooperate to increase insulin release in vivo (56, 57), neither GIP nor various doses of glucose had any measurable effect on the Fura-2 ratio when added alone or in all possible combinations with 1 μM Xen (data not shown).

Fig. 3.

Xenin-25 activation of MENs requires neurotensin receptor-1 (NTSR1). MENs were imaged before and after addition of the indicated agent(s). Each color tracing for Fura-2 ratios is for an individual, representative cell. A: Fura-2 ratios are shown for MENs given 2 sequential additions of Xen (Xen-1 and Xen-2, respectively). Cells were washed with HBSS containing human albumin (HSA) between additions of Xen. 2 cells that did not respond to Xen are also shown for comparison. B: estimated cytosolic free calcium levels for the cells in A are shown. Inset: average ± SE basal and peak cytosolic free calcium concentrations following addition of Xen-1. C: MENs were pretreated for 20 min with SR48692, an inhibitor of NTSR1, before addition of Xen. D: MENs were treated with Xen containing an additional cysteine residue at the COOH terminus (Xen-C) or NH₂ terminus (C-Xen) of the native peptide.

Xenin-25 increases cytosolic free calcium levels by binding to neurotensin receptor-1.

Genetic and pharmacological data suggest that the effects of Xen are mediated by neurotensin receptor-1 (NTSR1) (7, 15, 16, 28, 46, 47). The percentage of MENs excited by neurotensin (9.3% of 486) was similar (P = 0.68) to that excited by Xen (Table 1). Next, 10 μM SR48692, a well-characterized nonpeptide NTSR1 antagonist (24, 35, 54), was added to cultured MENs, and 20 min later, neurons were challenged with Xen. Xen did not increase the Fura-2 ratio in any of 186 MENs imaged (Fig. 3C and Table 1; P < 0.001 for Xen vs. Xen plus SR48692). In contrast, SR48692 did not prevent the increase in the Fura-2 ratio when the same cells were subsequently challenged with ATP/UTP (Fig. 3C and Table 1). Once bound by agonist, NTSR1 is internalized and not recycled to the plasma membrane (34, 51, 54). Thus cross-desensitization studies were conducted (Table 2; See Fig. 3A for representative MENs). Only 2.9% of 69 MENs that responded to a first addition of Xen (Xen-1) were reexcited by a subsequent treatment with Xen (Xen-2). Similarly, neurotensin failed to excite MENs that were previously excited by neurotensin. Xen excited only 5.1% of the MENs that were previously excited by neurotensin, indicating that NTSR1 mediates the effects of Xen. In contrast, a small subset (13%) of MENs that were excited by Xen remained responsive to a subsequent addition of neurotensin. Thus we cannot exclude the possibility that an additional neurotensin receptor mediates some of the effects of neurotensin, but not Xen, on a small subset of Xen-responsive MENs.

Table 2.

NTSR1 mediates the effects of Xen on MENs

		Excited/Imaged	Excited/Imaged
Stimulus 1	Stimulus 2	Stimulus 1	Stimulus 2
Xen	Xen	69/665 (10.4%)	2/69 (2.9%)
NT	NT	60/472 (12.7%)	0/60 (0%)
NT	Xen	79/477 (16.6%)	4/79 (5.1%)
Xen	NT	68/508 (13.4%)	9/68 (13.2%)

Data indicate the number (responsive/imaged) and percentage of MENs that exhibited an increase in [Ca²⁺]_i in response to sequential additions of Xen and/or neurotensin (NT). Note that only 3-5% of the MENs that were excited by the first addition of Xen or NT were excited by a subsequent treatment with Xen. All P values comparing each of the 4 sequential treatments are <0.001.

NTSR1, neurotensin receptor 1.

Effects of Xen on gut motility are blocked by SR48692 as well as by elimination or amidation of the COOH-terminal leucine (15). Thus an intact COOH-terminal leucine is required for Xen activation of NTSR1. We have previously demonstrated that Panc1 cells are activated by Xen (57). Using these cells, we confirmed that Xen with a cysteine added to the NH₂ terminus (N-Xen), but not to the COOH terminus (Xen-C), increased NTSR1 signaling (data not shown). As shown in Fig. 3D and Table 1, Xen-C increased the Fura-2 ratio in only 1.6% of the 62 MENs imaged. In contrast, C-Xen increased the Fura-2 ratio in 11.8% of 102 MENs, which included MENs that previously failed to respond to Xen-C [Fig. 3D; (P = 0.04 for C-Xen vs. Xen-C)]. Taken together, these results indicate that Xen excites MENs by binding to NTSR1.

The increase in [Ca²⁺]_i involves influx of extracellular calcium.

MENs were incubated in HBSS without calcium containing 0.5 mM EGTA before addition of KCl, Xen, or ATP/UTP (Table 1). EGTA treatment blocked the increase in the Fura-2 ratio following addition of KCl in all of the 64 MENs imaged, indicating that the extracellular calcium was effectively chelated. EGTA also prevented calcium mobilization in Xen-treated MENs because only 0.5% of the 182 MENs imaged still exhibited an increase in the Fura-2 ratio (P = 0.002 for Xen vs. Xen plus EGTA). Importantly, subsequent challenge of the EGTA/Xen-treated neurons with ATP/UTP increased the Fura-2 ratio in 94% of the 182 MENs imaged, indicating that 1) the MENs could still mobilize calcium from intracellular stores, 2) EGTA and/or low extracellular calcium levels were not toxic, and 3) the MENs remained physiologically functional. Thus influx of extracellular calcium is involved in the Xen-mediated [Ca²⁺]_i increase.

Xenin-25-responsive MENs have ATP-regulated potassium channels.

Subsets of MENs contain ATP-regulated potassium channels (37). Following pretreatment of neurons with 100 μM diazoxide to open these channels, the Xen-mediated increase in the Fura-2 ratio was noted in only 1.1% of the 187 MENs imaged (Table 1), which is much lower than the 12.7% noted in the absence of the drug (P < 0.001). When the MENs were subsequently treated with ATP/UTP, the Fura-2 ratio increased in >99% of the 187 MENs imaged. Thus the Xen-responsive MENs have ATP-regulated potassium channels.

Xenin-25 increases acetylcholine release from MENs.

MENs were next isolated, cultured, and then assayed for acetylcholine release (Fig. 4). Compared with vehicle alone (albumin), ATP/UTP augmented acetylcholine release 1.75-fold (P < 0.05), consistent with the fact that [Ca²⁺]_i increased in ∼100% of the MENs treated with these nucleotides. Xen increased [Ca²⁺]_i in a smaller percentage of MENs but still increased acetylcholine release 2.5-fold (P < 0.001).

Fig. 4.

Xenin-25 increases acetylcholine release from myenteric neurons. Acetylcholine released into the culture media was measured after addition of the indicated excitatory stimuli. Values represent average ± SE (n = 6). *P < 0.05, **P < 0.01, and ***P < 0.001.

Amylin increases cytosolic free calcium levels in MENs.

Amylin is produced by islet β-cells as well as several different subtypes of enteroendocrine cells located in the proximal gut (40, 50). Like Xen, amylin delays gastric emptying (40, 50). As shown in Fig. 5, A and B, and Table 1, amylin also increased the Fura-2 ratio in 19% of the 72 MENs imaged. This increase corresponds to a 2.9-fold (P < 0.0001) increase in [Ca²⁺]_i from a basal level of 0.025 ± 0.001 μM to a peak concentration of 0.074 ± 0.004 μM. Interestingly, this basal [Ca²⁺]_i was similar to that in Xen-responsive neurons (Fig. 3A), which is lower than the average basal [Ca²⁺]_i in the general population of MENs (Fig. 5, C and D). In contrast to Xen, amylin treatment resulted in a sustained increase in [Ca²⁺]_i. Of the 14 MENs that responded to amylin, 13 also responded to a subsequent challenge with Xen (Fig. 5A, Table 1). Similarly, 72 MENs that responded to Xen were also stimulated by a subsequent treatment with amylin. Thus Xen and amylin increased [Ca²⁺]_i in the same subset of MENs. Although insulin and amylin are coproduced by islet β-cells, insulin did not increase the Fura-2 ratio in any of 104 MENs imaged (Table 1).

Fig. 5.

Amylin and Xenin-25 activate the same subset of MENs. MENs were treated sequentially with amylin (Am), Xen, and then nucleotides. A: tracing represents the average Fura-2 ratio ± SE for 14 MENs that responded to amylin. B: average ± SE basal and peak cytosolic free calcium concentrations following addition of amylin. C: basal Fura-2 ratios were plotted for MENs that were amylin positive (Am+) or negative (Am-) that were imaged simultaneously. D: average basal [Ca²⁺]_i was calculated for Am+ vs. Am- MENs. ***P < 0.001.

NTSR1 is expressed on duodenal MENs in humans.

Paraffin-embedded sections of human duodenum were incubated with antibodies directed against VAChT (red), a marker of cholinergic neurons, plus antibodies recognizing human NTSR1 (green). As shown in Fig. 6, both proteins are expressed in cell bodies throughout a single ganglion. However, all VAChT-positive neurons do not express NTSR1. In addition, some nerve fibers within the longitudinal muscle are positive for NTSR1, but not VAChT, whereas others are positive for VAChT, but not NTSR1. Thus MENs that should be able to respond to neurotensin are present in the human duodenum, and only a subset of these MENs is cholinergic.

Fig. 6.

NTSR1 is expressed in human MENs. A single paraffin-embedded section of human duodenum was stained for NTSR1 (green) plus vesicular acetylcholine transporter (VAChT) (red). Nuclei were counterstained blue. Original photos were taken with an ×80 objective. D is a merged image of A, B, and C. Note that both antigens are expressed in/on cell bodies within a single ganglion. Area D1 shows extensive staining for both NTSR1 and VAChT. Area D2 shows adjacent cell bodies with weak cytoplasmic staining for NTSR1 and only 1 cell body positive for VAChT. Areas D3 and D4 show nerve fibers that express only NTSR1 or VAChT, respectively.

DISCUSSION

Exogenously administered Xen delays gastric emptying (29), reduces food intake (1, 8, 9, 28, 36), induces gall bladder contractions (27), increases gut motility (13), augments secretion from the exocrine pancreas (12, 17, 46), and amplifies the effects of GIP on insulin and glucagon release (41, 56, 57). In mice, the effects of exogenously administered Xen on insulin release are indirectly mediated by a cholinergic relay that does not appear to utilize parasympathetic neurons (57). The effects of Xen on gallbladder contraction are also mediated by a cholinergic neural pathway that was not abolished by truncal vagotomy, indicating that parasympathetic neurons do not relay this cholinergic response (27). Kirchgessner and Gershon (33) have shown that 1) injection of a retrograde tracer into the pancreas labels neurons in the myenteric plexus of the stomach and proximal duodenum, 2) injection of an anterograde tracer into a myenteric ganglia labeled axon terminals in the pancreas, and 3) activation of sodium channels in the duodenal mucosa excited neurons in myenteric and pancreatic ganglia and activated pancreatic acinar and islet cells. Thus myenteric neurons could potentially mediate some of the effects of exogenously administered, circulating Xen. However, it is unknown whether MENs in this region of the gut respond to Xen. The present study utilized primary cultured guinea pig MENs as a first step to show that Xen 1) increases [Ca²⁺]_i in only a small subset of MENs, 2) increases acetylcholine release from MENs, 3) requires NTSR1 to excite MENs, and 4) involves influx of extracellular calcium. It will be important to determine whether the Xen-mediated increase in [Ca²⁺]_i in the MENs is mediated by voltage-gated calcium channels and/or Ca²⁺-induced Ca²⁺ release. Because diazoxide blocked the ability of Xen to increase [Ca²⁺]_i, the Xen-responsive MENs have ATP-regulated potassium channels, which may also be required for Xen activity. It should also be noted that, in guinea pigs, primary afferent neurons that relay glucose and pressure signals from the duodenal villi to the pancreas are submucosal rather than myenteric (32). Thus it is not known whether these or other submucosal neurons are also excited by Xen. We recently showed that Xen increases the effects of GIP on insulin release in humans without T2DM (56) and delays gastric emptying in humans with and without T2DM (manuscript in preparation). Because NTSR1 is expressed on myenteric neurons in the human proximal small intestine (Fig. 6), MENs could also mediate some effects of Xen in humans.

Only 12.7% of the MENs were excited by Xen, which is consistent with the fact that individual MENs respond to distinct sets of ligands with respect to calcium mobilization as well as acetylcholine release (4, 30, 31, 44, 45, 52, 58, 60). Additional studies will be required to determine whether the cholinergic neurons are directly activated by Xen or whether other neurotransmitters (e.g., tachykinin) released from Xen-responsive MENs indirectly excite cholinergic MENs. It is interesting to speculate that Xen-responsive MENs from adult guinea pigs with T2DM may exhibit properties that are distinct from those in the healthy guinea pigs reported here.

In contrast to the peripheral effects on GIP-mediated insulin release (57), the anorexic effects of Xen are associated with increased activation of central neurons (8, 36). However, a 500-fold higher dose of Xen was administered in the feeding studies compared with our insulin release experiments. As reviewed, amylin is a 37-amino-acid polypeptide released predominantly along with insulin by islet β-cells (40, 50). Amylin is also produced by several different subtypes of enteroendocrine cells located in the proximal gut, many of which are released into the bloodstream immediately after meal ingestion (43). The amylin receptor is composed of a calcitonin receptor plus receptor activity modifying protein-1 or -3, and stimulation increases cAMP (42). Activation of NTSR1 also increases cAMP as well as MAPK activity (34, 54). Like Xen, amylin delays gastric emptying, inhibits food intake, and activates central neural circuits. We now show that amylin and Xen excite the same subset of MENs, indicating that some effects of both peptides could be mediated by myenteric rather than central neurons. Thus Xen and amylin are components of a combinatorial system that could allow MENs to integrate and appropriately respond to changes in nutritionally regulated signals. Neurotensin is a 13-amino acid peptide produced by a variety of cells including a subset of enteroendocrine N cells in the distal intestine and is released into the bloodstream mainly in response to ingestion of fat (5, 55). Because neurotensin and Xen activate NTSR1, neurotensin could also be a player in this system. Our data also indicate that the subset of MENs that respond to amylin and Xen have lower basal [Ca²⁺]_i than nonresponders. Thus the Xen responsive MENs may 1) be hard wired differently from nonresponders, 2) express a different repertoire of calcium channels on the plasma membrane, 3) have different set points for calcium uptake into the endoplasmic reticulum and mitochondria, and 4) remain very quiescent until they are excited. This last point is particularity important for a nutrient-regulated pathway. It will be important to determine whether Xen and amylin activate the same or distinct signaling pathways in MENs and whether additional gut-derived peptides also activate this same subset of MENs.

Ironically, although Xen was reported to represent a physiological peptide present in human plasma (14), we have not been able to detect it in either mouse or human plasma prepared under fasting, fed, or postprandial conditions (56, 57). Despite this, studies examining the pharmacological effects of exogenously administered Xen are providing important insights into the potential role of a specific subset of MENs for regulating gastrointestinal function and potentially glucose homeostasis. Additional studies will be required to determine whether Xen-responsive MENs project to the pancreas and regulate islet function.

GRANTS

Portions of this research were supported by funds from NIH grant number 1R01 DK088126, The Blum Kovler Foundation, and the Live Cell Imaging Facility of the Center for Investigation of Membrane Excitability Diseases (CIMED) at Washington University.

DISCLOSURES

Washington University is pursuing a patent related to the use of Xenin-25 to treat type 2 diabetes mellitus. In the future, this could lead to personal financial benefit to BMW and the University.

AUTHOR CONTRIBUTIONS

Author contributions: S.Z., K.H., and S.W. performed experiments; S.Z., K.H., and B.M.W. analyzed data; S.Z., K.H., and B.M.W. interpreted results of experiments; S.Z., K.H., S.W., and B.M.W. edited and revised manuscript; S.Z., K.H., S.W., and B.M.W. approved final version of manuscript; B.M.W. conceived and designed research; B.M.W. prepared Figs.; B.M.W. drafted manuscript.