CNS Sites Controlling the Gastric Pyloric Sphincter: Neuroanatomical and Functional Study in the Rat

Abstract

The pyloric sphincter receives parasympathetic vagal innervation from the dorsal motor nucleus of the vagus (DMV). However, little is known about its higher-order neurons and the nuclei that engage the DMV neurons controlling the pylorus. The purpose of the present study was two-fold. First, to identify neuroanatomical connections between higher-order neurons and the DMV. This was carried out by using the transneuronal pseudorabies virus PRV-152 injected into rat pylorus torus and examining the brains of these animals for PRV labeling. Second, to identify the specific sites within the DMV that functionally control the motility and tone of the pyloric sphincter. For these studies, experiments were performed to assess the effect of DMV stimulation on pylorus activity in urethane-anesthetized male rats. A strain gauge force transducer was sutured onto the pyloric tonus to monitor tone and motility. L-glutamate (500 pmol/30nl) was microinjected unilaterally into the rostral and caudal areas of the DMV. Data from the first study indicated that neurons labeled with PRV occurred in the DMV, hindbrain raphe nuclei, midbrain Edinger-Westphal nucleus, ventral tegmental area, lateral habenula, and arcuate nucleus. Data from the second study indicated that microinjected L-glutamate into the rostral DMV results in contraction of the pylorus blocked by iv administered atropine and ipsilateral vagotomy. L-glutamate injected into the caudal DMV relaxed the pylorus. This response was abolished by ipsilateral vagotomy but not by iv administered atropine or L-NAME. These findings identify the anatomical and functional brain neurocircuitry involved in controlling the pyloric sphincter. Our results also show that site-specific stimulation of the DMV can differentially influence the activity of the pyloric sphincter by separate vagal nerve pathways.

Graphical Abstract

The dorsal motor nucleus of the vagus (DMV) of the brainstem is a crucial region that modulates the activity of gastric function. The pyloric sphincter at the gastroduodenal junction, which regulates the flow of partially digested food from the stomach, is mediated by the nucleus in a site-specific manner. Activation of the DMV in the rostral area contracts the tone and motility of the sphincter, whereas, in the caudal area of the nucleus, it relaxes its tone, whose pathway is yet to be undetermined.

Introduction

The pyloric sphincter at the gastroduodenal junction regulates the flow of gastric chyme. This functional activity is regulated by intrinsic and extrinsic neural connections (Kressel et al., 1994; Lindestrom and Ekblad, 2002). Of particular importance is the extrinsic parasympathetic vagal innervation (Kressel et al., 1994; Lindestrom and Ekblad, 2002; Hayakawa et al., 2003), which originates from cholinergic neurons of the dorsal motor nucleus of the vagus (DMV) in the brainstem (Elfvin and Lindh, 1982; Lindestrom and Ekblad, 2002; Hayakawa et al., 2003). However, little is known about the central neural processes that regulate the pylorus activity, especially related to higher-order nuclei interacting with subnuclei within the DMV, making up the neural pathway.

Direct stimulation of the vagus nerve elicits frequency-dependent excitatory, relaxatory, or biphasic responses of the sphincter. Low-frequency stimulation (0.2-0.5 Hz) evokes excitatory responses (Edin et al., 1980; Allescher et al., 1988a), which are blocked by atropine and hexamethonium, indicating their cholinergic nature (Allescher et al., 1988a). Conversely, high frequency (>0.7 Hz) stimulation of the vagus nerve causes inhibition of pyloric tone and motility (Edin et al., 1980; Allescher et al., 1988a) that is not blocked by atropine, phentolamine, or propranolol, suggesting a non-adrenergic-non-cholinergic (NANC) projection (Allescher et al., 1988a). However, despite these pioneering studies, these results are difficult to interpret because of the heterogeneous composition of vagal fibers and the mode of stimulation (i.e., electrical) wherein both vagal afferents and efferents are recruited.

Gastric emptying (GE) studies have also highlighted the vagus nerve's role in modulating the pylorus activity (Tougas et al., 1992; Malbert et al., 1995; Ishiguchi et al., 2000b; Ishiguchi et al., 2002; Ueno et al., 2005). Electrical stimulation of the vagus decreased pyloric pressure and increased transpyloric flow rate, which was attributed to inhibitory vagal innervation of the pylorus (Malbert et al., 1995). Although, it must be noted that vagal impact on GE is dependent on volume and composition of gastric contents (MacGregor et al., 1977; Prove and Ehrlein, 1982; Houghton et al., 1988b; Houghton et al., 1988a) as emptying of liquids are accelerated by vagotomy, while solids are decelerated by it (Wilbur and Kelly, 1973). In particular, GE of solids is mediated by the coordination of events along the antropyloroduodenal axis (Ishiguchi et al., 2000b; Ishiguchi et al., 2001; Ishiguchi et al., 2002; Ueno et al., 2005), which are disrupted by vagal cooling and vagotomy (Ishiguchi et al., 2000b; Ishiguchi et al., 2001; Ishiguchi et al., 2002; Ueno et al., 2005). Moreover, their inhibition by hexamethonium and the nitric oxide (NO) synthase inhibitor L-NAME (Ishiguchi et al., 2000b; Ishiguchi et al., 2001; Ishiguchi et al., 2002; Ueno et al., 2005) suggests that these events are under the influence of a brainstem derived cholinergic-NO pathway. Altogether, these GE and vagal stimulation studies firmly establish the vagus nerve's importance in controlling pyloric activity; however, they do not specifically address how the activity of the pyloric sphincter is regulated from the DMV. For instance, is the sphincter's activity controlled by separate vagal pathways similar to those controlling the lower esophageal sphincter (LES)? This sphincter is topographically affected by stimulation of LES-projecting DMV neurons (Rossiter et al., 1990; Abrahams et al., 2002; Niedringhaus et al., 2008a). Activation of the rostral area DMV contracts the LES via a cholinergic-cholinergic pathway (Rossiter et al., 1990; Abrahams et al., 2002; Niedringhaus et al., 2008a), while stimulation of the caudal area relaxes it (Barone et al., 1984; Rossiter et al., 1990; Abrahams et al., 2002; Niedringhaus et al., 2008a) through a nitrergic-VIPergic pathway (Niedringhaus et al., 2008a). Moreover, except for the distribution of 1st order neurons in the DMV that project to the pylorus, little is known about higher-order projections from other brain areas. Hence, a series of anatomical and functional studies were undertaken to determine (1) the origin of higher neurons that may influence the activity of the pylorus and directly stimulate (2) the DMV while monitoring pyloric activity. We hypothesized that similar to the LES, pyloric tone and motility are regulated in a site-specific manner from the DMV.

Materials and Methods

Experiments were performed on male Sprague-Dawley rats weighing 300 – 400g (Harlan Laboratories) by the National Institutes of Health guidelines for the use of animals in research and the approval of the Georgetown University Animal Care and Use Committee.

Anatomical Tracing Studies of the Pylorus

Recombinant Bartha strains of pseudorabies virus PRV-152 (conjugated to GFP) and PRV-BaBlu (conjugated to β-Recombinant Bartha strains of pseudorabies 8 pfu/ml were kindly provided by Dr. Lynn Enquist (Princeton University, Princeton, NJ). Viruses were maintained frozen at −80°C in small 250 μl aliquots. Before the first in vivo injection, the virus was centrifuged for 5 min. After that, each subsequent injection was preceded by a 1 min centrifugation of the virus to ensure that the active viral colonies remained suspended for inoculation. Between injections, the virus was kept on ice.

Surgical Preparation and Inoculation Procedure:

Surgery and viral injection were performed in a biosafety level II operating room. Animals were anesthetized with isoflurane (Abbot Laboratories, Cat# B506; 4% induction; 1.5% maintenance; vaporized with 95% oxygen and 5% CO2) via a nose cone. Body temperature was monitored by a rectal thermometer and maintained at 37°C with a surgical heat lamp. Following a surgical level of anesthesia, as indicated by a lack of response to a toe pinch and absence of corneal reflex, an abdominal laparotomy was performed via a midline incision along the linea alba to provide access to the stomach. Next, the gastric antro-pyloro-duodenal region was gently exposed and lifted with a 22-gauge bent gavage needle. Fine-forceps were placed under the pylorus to stretch it and to allow for a saline gauze to be placed in the peritoneal cavity. Recombinant strains of PRV (−152 or −BaBlu) were injected into the pyloric torus, duodenum, antrum, or intraperitoneal (IP) cavity via a 10 μl Hamilton syringe with a 30-gauge needle. In injecting into the pylorus, the syringe needle was inserted diagonally at approximately 1 mm proximal to the torus towards the stomach's lesser or greater curvature. Two injections (each 5 μl) were made into the pylorus. For injections in the antral region, the syringe needle was inserted 5 mm above the pylorus diagonally towards the lesser curvature of the proximal portion of the antrum. A single injection with a total of 1 μl was dispensed into the antrum.

Similarly, to the antrum, one injection (10 μl) was administered to the duodenum. The virus was introduced topically in a single 10 μl volume in control animals into the abdominal cavity. All injections were delivered slowly under visual guidance using a dissecting microscope. After the virus injection, the syringe needle was held in place for ~1 min before being slowly withdrawn on cessation of each injection to minimize virus leakage. The exposed area was washed with sterile saline, and the gauge separating the antro-pyloro-duodenal region was removed. The incision was closed in two layers with 4-0 Vicryl sutures and skin staples. All animals were administered analgesic (buprenorphine 0.03-0.05 mg/kg, SC every 8-12 h) and after that, as needed. A positive PRV infection was based on the DAB staining present at the 72 + hr post-injection mark (positive control).

Histology:

After a survival period of 72h, 96h, and 120h (respectively, 3, 4, and 5 days), the animals were anesthetized with an overdose of pentobarbital (80 mg/kg, IP). After the establishment of anesthesia, animals were placed on an elevated surgical board, and a Y-incision in the thoracic-abdominal region was made to expose the heart. A gavage needle (3 mm ball diameter) attached to a primed perfusion pump (Masterflex, Cole-Parmer Instrument Co) was threaded into the ascending aorta via the left ventricle. The needle was clamped in place, the pump turned on, and the right atrium cut to facilitate perfusion. The perfusion was initiated with a 500 ml heparinized saline wash (0.15M; pH 7.4) that was followed by a 500 ml Paraformaldehyde-Periodate (Sigma Aldrich, Cat# 11448) Lysine (Sigma Aldrich, Cat# L5501; PLP) fixative (4% 0.1 M, and 0.01 M respectively; pH7.4). The brains were then removed and stored overnight in the PLP fixative before being transferred to a 30% sucrose solution (PLP + 30% sucrose). Following cryoprotection, the brains were cut on a cryostat (Reichert-Jung) into sequential 35-40 mm sections and collected in a six-well plate. The tissue was stored in a cryoprotectant (0.1 M Phosphate-Buffered-30% sucrose (Sigma Aldrich, Cat# S0389), 30% ethylene glycol (Sigma Aldrich, Cat# 324558), and 1% Polyvinyl-pyrrolidone (Sigma Aldrich, Cat# PVP40; 40) at −20 °C until immunohistochemical processing. The distance between sequential sections within each well was 210-240 mm.

Immunohistochemistry:

Brain sections were initially washed in phosphate-buffered saline (PBS; 0.01 M; pH7.4) and then incubated in a primary antibody solution for 24 hours at 4°C. Rabbit polyclonal anti-GFP (1:10000; BD Biosciences, Cat# T0005, RRID AB_2839423) and mouse monoclonal anti-b-galactosidase (1:1000; Sigma-Aldrich, Cat# G6282, RRID: AB_259937) were used as primary antibodies against the PRV-152 and the PRV-BaBlu recombinants, respectively. After incubation, the tissue was washed and incubated in a secondary antibody solution for 1 h at room temperature. Biotinylated anti-rabbit IgG (1:500 Vector Laboratories Cat# BA-1000, RRID:AB_2313606) and biotinylated anti-mouse IgM (1:500 Millipore Cat# 20775, RRID: AB_11212339) were used to distinguish between PRV-152 and PRV-BaBlu, respectively. Subsequently, the tissue was washed and then further incubated in the Avidin-Biotin solution (ABC Elite kit; Vector Labs, Cat# SP-2001, RRID: AB_2336231) for 1 h. The tissue was then stained with a 2.5% Nickel 3,3’-Diaminobenzidine (DAB, Cat# Sk-4100) solution, mounted on gelatinized (superfrost) slides, and coverslipped with DPX (Sigma Aldrich, Cat# 06522) for analysis. For cytoarchitectonic analysis, a select number of slides were stained with cresyl violet (Sigma Aldrich, Cat# 255246).

Data Analysis:

To determine neuronal labelling of each coronal section was examined for immunolabeling using a light microscope (Nikon E3600) equipped with bright-field, dark-field, and episcopic fluorescence optics. A neuron was considered labeled if DAB-crystals (dark-brown in color) were present on the neuronal cell bodies. Sections were processed serially in a caudal to rostral orientation for each of the three survival time points 72h, 96h, and 120h (respectively, 3, 4, and 5 days). The PRV-labeled neurons' locations were identified using the atlas of Paxinos and Watson (Paxinos and Watson, 1998).

Functional Studies of the Pyloric Sphincter

Surgical preparation:

Before all experiments, food was withheld overnight, whereas water was provided ad libitum. Animals were anesthetized with an intraperitoneal (IP) injection of urethane (Sigma Aldrich, Cat# U2500; 1.3 mg/kg) dissolved in 0.9% saline. Body temperature was monitored by a rectal thermometer and maintained at 37 ± 1°C with an infrared heat lamp. All animals that underwent neurosurgery were pretreated with dexamethasone (Sigma Aldrich, Cat# D-085; 0.8 mg, SC). After anesthesia was confirmed by lack of pedal and corneal reflexes, rats were intubated via the trachea following a tracheotomy to maintain an open airway and institute artificial respiration when necessary. Next, the carotid artery and the jugular vein were cannulated with polyethylene tubing (PE 50) for monitoring blood pressure and for systemic infusion of drugs, respectively. Blood pressure was monitored by a pressure transducer (ADInstruments, Cat# MLT844) coupled to a PL3516 PowerLab data acquisition system (ADInstruments, Colorado Springs, Co). Both cervical vagi were carefully isolated from each carotid artery on either side and looped with a 5-0 silk thread for later avulsion during the experiment.

Subsequent to vessel cannulation, a laparotomy was performed to expose the stomach. A custom-made half-bridge force transducer (2mm2; 350 ohms, Kyowa, Inc.) was sutured onto the pyloric torus and another affixed to the gastric antrum in an orientation that was aligned with the circular smooth muscle. A small balloon (5x5 mm) was inserted into the stomach (1 cm proximal to the torus) to provide a contractile tone to the pylorus when inflated (~0.3ml). The stomach and strain gauge wires were carefully placed back into the abdomen, and the cavity closed with a 0-0 chromic gut suture. The contact leads of the strain gauge wires were kept outside the cavity for later connection to the bridge amplifiers.

Microinjection Procedure:

To gain access to the dorsal medulla of the rats, they were positioned in a stereotaxic apparatus (David Kopf, Tujunga, CA; Cat# 68801). A limited dorsal craniotomy was performed to expose the medulla, and the underlying dura and pia were cut and reflected. The caudal tip of the area postrema, the calamus scriptorius (CS), was viewed as a reference point for determining the microinjection coordinates. Microinjection was accomplished by a double-barreled glass pipette (ID 9.3mm; OD 35-80 μm; Fredrick Haer, New Brunswick, ME) connected to PE 50 tubing. All microinjections were given into the DMV at a 30° angle from the perpendicular. Each injection was administered within 5–10s in a volume of 30nl by hand-controlled pressure as determined by a calibration tape (Formaline 9006B, Wheeling, IL; Cat# 9006B) affixed to the micropipette. Stereotaxic coordinates for injection into the rostral area of the DMV were: AP = +0.4—0.8 mm to CS; ML= +0.1—0.2 mm lateral to the midline; and DV = −0.4—0.6 mm dorsoventral to the dorsal surface of the medulla. Coordinates for the caudal area of the DMV were: AP = 0.0 to—0.2 mm caudal to CS; ML= +0.3—0.4 mm lateral to the midline; and DV = −0.4—0.6 mm dorsoventral to the dorsal surface of the medulla. Stimulation of the DMV neurons was accomplished by unilateral microinjection of L-glutamate (Sigma Aldrich, Cat# 49621; 500pmol/ 30nl; (Ferreira, 2000)). Before activating DMV neurons, the placement of antral and pyloric strain gages was verified by intravenous (IV) administration of cholecystokinin octapeptide (Bachem, Cat# 4033010; CCK-8). This also served to isolate pyloric responses from those of the antrum. The rationale for using CCK-8 was based on studies that report its excitatory effect on the pylorus and its inhibitory action on the antrum (Lingenfelser et al., 1997; Shi et al., 2003; Adelson et al., 2004). To determine the dose, which produced consistent and repeatable responses while maximizing the differences between the two regions of the stomach, a bolus of CCK-8 in a dose-range of 0.1—3.0 μg/kg was administered IV in a low to a high concentration in 5 animals. A 30 min period was observed between subsequent doses, and each dose was repeated twice. Preceding each experiment, a bolus of saline (IV) was delivered to account for any vehicle effects. Before terminating the experiment, sodium nitroprusside (Sigma Aldrich, Cat# 1614501; SNP) 50μg/kg was administered to determine the drug response's directionality. [Note: a buckling of the strain gage encapsulating material due to muscle contraction can result in a 'negative-going' response, which can be misread as a relaxatory response. Since SNP always produces a decrease in tone via relaxation of the smooth muscle, it serves as a control to test the drug responses veracity.]

Of all the CCK-8 doses tested, the 0.3 μg/kg proved to be the most reliable in that it produced the most consistent and maximal effect on both the pylorus and antrum without any tachyphylaxis. Consequently, the CCK-8 dose of 0.3μg/kg was used to verify strain gage placement in all microinjection experiments.

Both blood pressure and strain gauge force data were acquired using the PowerLab data acquisition system (ADInstruments, Colorado Springs, Co) connected to an Apple G5 computer (Apple, Inc).

Histological Verification of Pipette Tracks:

At the end of each experiment, the rat was euthanized with an overdose of anesthesia. The brain was removed and placed in a fixative-cryoprotectant solution composed of 4% phosphate-buffered paraformaldehyde (Alfa Aesar, Cat# 43368) and 10% sucrose (Sigma Aldrich, Cat# S0389; 0.1M; pH 7.4) for at least 48 hr. The brainstem was dissected and cut on a cryostat into 50mm coronal sections, which were mounted serially onto gelatinized slides and stained with cresyl violet (0.8%). The locations of the microinjection pipette tracks were identified using the atlas of Paxinos and Watson (Paxinos and Watson, 1998). To document microinjection sites, microphotographs and camera lucida drawings were made of each pipette track.

Imaging:

Slides were scanned at 3X magnification using the imaging system that included a camera (JENOPTIK GRYPHAX^®, Jena, Germany) that connected the SMZ1500 stereoscopic zoom microscope (Nikon Instruments, Inc., Melville, New York; Cat# E3600 and SMZ1500). Images were obtained using a bright-field microscope and exported as component TIFF images for analysis in ImageJ software (National Institutes of Health, USA). While we manually traced the DMV's retrograde neuronal labeling for analysis; however, due to the vast number of neurons in the various brain regions labeled the polysynaptically retrogradely PRV-152 virus injection in the pyloric torus, we quantified the label of the neurons in the various nuclei by detects and measures the outline of each neuron via automatic threshold algorithm.

In the DMV, analysis was based on labeling neurons in each brainstem section comprising three zones designated as previous publications (Cruz et al., 2007). This division is based on the position of the DMV relative to the area postrema (AP). It consists of a caudal zone (area of the DMV behind CS), an intermediate zone (area of the DMV that runs parallel to the AP), and a rostral zone (area of the DMV that lies anterior to the AP - the fourth ventricle). In animals, the mean area of neurons was calculated for individual zones, which was applied as the average area across all animals.

In extra-brain structures comprising the nuclei, the labeling neurons for each brain section were calculated as the average area and were further quantified for the average area across all animals.

Data Analysis:

Analysis of all experimental recordings was done 'off-line' using the Chart (ADInstruments) and Prism (GraphPad Software, Inc.) software packages. All experimental recordings were initially filtered using a root mean square (RMS; 2s moving window) algorithm to account for respiration and various other signal artifacts. Strain gage recordings showed both smooth muscle contractions (motility), as well as baseline tone. Baseline values for rostral/excitatory data were taken for a continuous 3 min period before delivering a drug. The lowest points in the experimental tracing determined pyloric tone values, whereas effects on pyloric motility (specifical amplitude) were taken as the highest values. The effect of a microinjected drug was taken as a 5-30s window at the onset of delivery. Within an individual animal, multiple microinjections of an identical drug were averaged to attain a single value for baseline and drug effect. All data representing the excitatory/rostral pyloric responses are presented as means ± standard error of the means. Mean differences in treatment conditions were assessed by a paired t-test. In all cases, p<0.05 was the criterion used to determine statistical significance.

Changes in pyloric tone and motility elicited from the caudal half of the DMV were recorded as percent changes to that induced by intravenous administration of SNP (50μg/kg). This allowed for the degree of inherent relaxation in pyloric tone (due to the presence of anesthesia) to be normalized. A continuous 3-minute baseline (lowest values) was taken before the administration of SNP. After SNP was injected, a continuous 1 min period (representing the lowest values seen 2 minutes after administration) was compared to baseline levels. The difference between the two values was taken as the animal’s maximal (100%) inhibitory response. The effect of a microinjected drug on pyloric tone (5-30s window following injection) in the caudal half of the DMV was determined by first comparing the drug’s effect to baseline (3 min) then comparing to the maximal inhibitory effect in the same animal to the administration of SNP to get the percent maximal inhibitory response. All percentage-based data was assessed using a Shapiro-Wilks normality test. After passing the normality test, its significance (p<0.05) was determined using a paired t-test.

Drugs:

The drugs used were purchased from the following companies: urethane, L-glutamate (Sigma Aldrich, Cat# 49621), L-NAME (Sigma Aldrich, Cat# 11030), sodium nitroprusside (Sigma Aldrich, Cat# 1614501), atropine methyl bromide (Sigma-Aldrich, Cat# M1300000), Cholecystokinin-Octapeptide (Bachem (Torrance, CA), Cat# 4033010).

Results

To determine the distribution of second and higher-order CNS neurons innervating the pylorus, we injected rats with pseudorabies virus (PRV-152) in the ganglia/plexus of the torus of the pylorus. Positive retrograde PRV-DAB immunolabeling was used to analyze several transneuronal labeling neurons in the CNS. In the post-survival period ~72h or more prolonged (n=15), neurons of the hindbrain nuclei and some forebrain nuclei were labeled. In particular, PRV-labeled neurons were evident bilaterally in medullary areas associated with the autonomic function (e.g., area postrema, DMV, NTS, and nucleus ambiguous). Longer post-inoculation survival time (96-120h) allowed for further transneuronal travel and infected the forebrain structures. The study was initially divided into three groups based on post-inoculation survival time points of 72, 96, or 120h. However, the magnitude and distribution of PRV neuronal labeling at the post-survival time point 96h did not differ from that at 72h. Therefore, the two groups were combined, and analysis was performed to form two groups' neuronal labeling, a 72-96 and a 120h PRV post-inoculation period. At post-inoculation periods of 24h (n=3) and 48h (n=5), PRV-152 failed to show any retrograde label in the brain. Similarly, control intraperitoneal (IP) injections of PRV-152 did not infect neurons in the brain at all time points studied (1-5 day survival; n=1 for each time point).

Post-Inoculation Survival 72-96h (3-4 day):

The first observation of labeling occurred at ~72 hours post-injection. In 6 animals, PRV-152 (two 5 μl) injected into the pyloric torus (Figure 1a) retrogradely labeled neurons were found to be located bilaterally throughout the rostral-caudal extent of the DVC (i.e., dorsal motor nucleus of the vagus, nucleus tractus solitarius, and area postrema; Figure 1) and other medullary nuclei (Figure 2 & 3). In particular, labeling was observed in the brainstem nucleus of the DMV (Figure 1). The site where some of the first order large neurons of the efferent vagus nerve reside, namely the DMV, was clearly labeled. Additionally, small neurons interspersed throughout the rostral-caudal extent of the adjoining overlying nuclei of the NTS and area postrema (AP) exhibited significant labeling (Figure 1).

Figure 1:

Neurons labeled by injecting PRV-152 into the pylorus and waiting 72 hours. (a) The location of PRV injections into the pylorus torus is depicted in the diagrams. (b) The representative section of the hindbrain is located 0.5 mm rostral to calamus scriptorius and shows the DVC's nuclei and other hindbrain structures. (c) Representative micrographs of pylorus neurons labeled with PRV in the nuclei of the DVC. [Note: The labeled DMV neurons are premotor to the pylorus.] Five micrographs are labeled in relation to their distance from calamus scriptorius. Abbreviations: AP, area postrema; CC, central canal; DMV, dorsal motor vagal nucleus; NTS, nucleus tractus solitarius. The definition of the subnuclei with the abbreviations in the NTS and other areas can be found in Paxinos and Watson, 1998.

Figure 2:

Photomicrographs of coronal brainstem sections showing transneuronal retrograde labeling in a rat 72h after PRV-152 injection in the pyloric torus. Their stippled boxes of the brainstem sections highlight diffused PRV-labeled neurons and fibers encompassing the RAmb nuclei (a), Amb (b), ECu (c), and ROb (d). [Note: numbers associated with each coronal section refer to its distance from the CS]. Abbreviations: AP, area postrema; Amb, ambiguous; CC, central canal; CS, calamus scriptorius; CVL, caudoventrolateral reticular nucleus; DMV, dorsal motor nucleus of the vagus; ECu, external cuneate; LRt, lateral reticular; NTS, nucleus tractus solitarius; Pa5, paratrigeminal; RPa, raphe pallidus; ROb, raphe obscurus; RAmb, retroambiguous; Sp5, spinal trigeminal tract; TS, tractus solitarius.

Figure 3:

Photomicrographs of brainstem sections showing PRV-152 labeled neurons at +2.0 mm and +2.5 mm rostral to calamus scriptorius. (A) Labeled neurons and fibers in the rostral DVC (stippled box A in the +2.00 mm brainstem section). (B) Shows PRV-labeled neurons in the gigantocellular area (stippled box B in the bottom coronal section). (C) A montage of labeled neurons and fibers traverses the +2.0 mm hindbrain brain section (stippled box C). Abbreviations: AmbC, nucleus ambiguous compact; Gi, gigantocellular reticular; RMg, raphe magnus; ROb, raphe obscurus; RPa, raphe pallidus; RVL, rostroventrolateral reticular.

The distribution and mean areas of the DMV retrogradely labelled neurons from the pylorus is illustrated in Figure 4 along with gastrointestinal (GI) neurons separately infected by the PRV-152 virus in the antrum or duodenum region of the gastroduodenal junction. The representative neurons sampled in the DMV were from the sites illustrated a nucleus caudal or the rostral to CS (same top and left panels of Figure 4a). The mean area (μm²) retrogradely labelled DMV-projections from the pylorus caudal, 206 ± 106, n=134 cells; CS, 243 ± 107, n=139 cells; intermediate, 274 ± 132, n=294 cells; and rostrum, 169 ± 69, n=152 cells for 6 rats. The area for DMV-projection neurons to the antrum or the duodenum were caudal, 166 ± 67, n=67; CS, 201 ± 79, n=69; intermediate, 241 ± 99, n=162; and rostrum, 205 ± 93, n=91 for 3 rats or caudal, 00.00 + 00.00; CS, 264 + 105, n=19; intermediate, 282 + 85, n=27; and, rostrum, 250 + 107, n=17 cells for 1 rat. There was no significant difference between the area size of the DMV-GI neurons.

Figure 4:

Distribution and areas of DMV GI neurons. (a) The representative neurons sampled in the DMV were from the sites that illustrated this nucleus caudal or the rostrum to CS (same top and left panels). A graph shows the means area of the antral-pyloric-duodenal neurons retrogradely by the PRV-152 (5-10 μl) was injected into the various regions of the stomach (antrum, pylorus, or the duodenum). (b). A graph shows the distributions of DMV GI neurons (antrum, pylorus, and duodenum) in the caudolateral DMV sites in the dorsoventral and mediolateral to the central canal. [Abbreviation: See Figure 1.]

To see the topographical distribution of DMV-GI neurons (pylorus, antrum, and duodenum) reported in the rostrocaudal sites of the nucleus (Figure 4a), we obtained the dorsoventral and mediolateral representation of the neurons in reference to the central canal. This topographical distribution is illustrated in Figure 4b.

In the forebrain, DAB-positive neurons were absent at 72 h post-injection, except in the paraventricular nucleus of the hypothalamus (PVN). The brainstem nuclei consistently identified by PRV labeling in all six animals are summarized in Table 1 based on their locations are in coronal sections of the atlas of the rat (Paxinos and Watson, 1998).

Table 1:

Brain nuclei labeled with the retrograde polysynaptic PRV-152 tracer were injected into the pylorus after 72h (3-day) and 120h (5-day) survival time. [Note: the monosynaptic retrograde tracer only label parasympathetic neurons in the dorsal motor nucleus (DMV). However, the PRV-152, a polysynaptically tracer, not only the DMV neurons retrogradely labeling but also labels other neurons in the brainstem and other brain nuclei that transynaptically coupled with projecting pyloric DMV neurons.]

72h Survival Time n=6	120h Survival Time n=6
Area Postrema	Arcuate
Caudoventrolateral Reticular Nucleus	Basomedial Amygdala
Dorsal Motor Nucleus	Central Amygdala
External Cuneate	Central Medial Thalamus
Gigantocellular Reticular	Dorsomedial Hypothalamic
Lateral Reticular	Edinger-Westphal
Nucleus Ambiguous, Compact	Lateral Habenula
Nucleus Tractus Solitarius	Lateral Hypothalamic Area
Paratrigeminal	Parabrachial Pigmented
Paraventricular Hypo.	Paraventricular Thalamic
Raphe Magnus	Periventricular Hypothalamic
Raphe Obscurus	Supramammillary
Raphe Pallidus	Ventromedial Hypothalamic
Retroambiguous	Ventral Tegmental Area
Rostroventrolateral Reticular
Spinal Trigeminal
Tractus Solitarius

To determine if transneuronal identification of the brainstem nuclei from the pylorus was similar to that of other gastric regions, PRV-152 injections were made in the antrum (n=1) of the stomach. After an identical post-inoculation survival period (72 h), PRV labeling was found bilaterally throughout the brainstem and PVN. No distinct differences were apparent between the distribution of neurons in the nuclei retrogradely labeled from the antrum or the fundus. However, when labeling in these nuclei was compared to that of the pylorus, the medullary raphe nuclei (obscurus, magnus, and pallidus) were densely labeled in pylorus-injected animals, whereas it was inconsistent and sparse (especially in the raphe magnus) in PRV injected antrum and fundus animals. A representative coronal section from a pylorus-injected animal encompasses the raphe nuclei (see Figure 3).

Post-Inoculation Survival 120h (5 day):

PRV-152 was injected into the pyloric torus of 6 animals. Transneuronal labeling was found bilaterally throughout the brainstem, as well as the forebrain. Labeling was found in the forebrain structures amongst others in the hypothalamic nuclei, amygdaloid area, mediodorsal thalamic nucleus and Edinger-Westphal nucleus, (Figure 5), which are essential components of the viscerosensory circuitry controlling emotion, energy, thermal, hormonal, and cognitive regulation. The brain nuclei labeled by PRV-152 in all six animals are summarized in Table 1, and their location is based on coronal sections adapted from the atlas of Paxinos and Watson (Paxinos and Watson, 1998). In addition, using imaging analysis software (ImageJ, NIH) to quantify the label of the neurons in the various nuclei, as illustrated in Figure 6, each neuron was defined by a threshold algorithm that detects the outline of the label cell and measures its area. We specifically looked at all the nuclei for which the neurons were measured, which have a distinct density of the transneuronal labeling (see Figure 6a, b). The mean area (μm²) of the neurons included were: ARC, 146 ± 104, n=149 cells; PVN, 125 ± 82, cell=340 ; LH, 182 ± 102, n=172; AD, 157 + 102, n301; MD, 74 ± 50, n=32; EW, 356 ± 91, n=80 (Figure 6c). A mixed-effect analysis showed that the neurons differed significantly amongst the different nuclei (F _{(4.000, 854.4)} = 86.23). Moreover, neuronal labeling in the area of the EW nucleus is significantly greater than for any neurons in other nuclei that we analyzed (ACR, 210.4; PVN, 231.2; LH, 173.9; AD, 199.3; MD, 282.5).

Figure 5:

A montage of forebrain and midbrain regions showing PRV-152 labeled neurons after a post-inoculation period of 120h. The regions labeled are (a) lateral habenula, (b) mediodorsal thalamic nucleus, (c) amygdala, (d) paraventricular nucleus of the hypothalamus (PVN), (e) lateral hypothalamus, (f) Edinger-Westphal, (g) arcuate nucleus, and (h) posterior hypothalamus. [Note: the arrow in the left panels indicated the sagittal location from which the panels' left and right coronal images were obtained.]

Figure 6:

Mean areas of neurons in the brain regions that were polysynaptically retrogradely are PRV-152 (5-10 μl) that were injected by the pylorus sphincter. (a & b). The illustrated brain regions (a) for the bright-field representative microscope images (b) for sample analysis. (c) We analyzed the sites of the neurons that had distinct densities in the various nuclei. A graph shows the mean areas of these neurons of the nuclei. [Abbreviation: 3V, third ventricle; Aq, aqueduct; IC, internal capsule; ARC, arcuate hypothalamic nucleus; PVN, paraventricular hypothalamic nucleus; LH, lateral hypothalamic area; AD, amygdaloid area; MD, mediodorsal thalamic nucleus; EW, Edinger-Westphal nucleus.

To assess if DMV has a site-specific effect of differential responses in the pylorus, similar to that seen with the LES (Rossiter et al., 1990; Niedringhaus et al., 2008a), and to discern the nature of the neurotransmitter(s) associated with these responses, we used L-glutamate to stimulate this nucleus.

L-glutamate microinjection into the rostral half of the DMV increases pyloric motility and tone:

To monitor this activity of the pylorus torus, miniature strain gauge transducers (Figure 7a) were attached to this and the adjacent gastric antrum. The antrum and the pylorus synchronicity associate with each other in a phase-response and in a time-dependent manner (Ishiguchi et al., 2002). We used drug-induced responses for verification of strain gauge placement of this area of the stomach. Intravenously administered cholecystokinin octapeptide (CCK-8; 0.3μg/kg IV) increases pyloric contraction and decreases the stomach's antral activation (Figure 7e). Moreover, the stomach's gauges' attachment is the strain stress that by the opposites of the intended directional, we use sodium nitroprusside (SNP; 50μg/kg IV) at the end of each experiment to determine if the blood pressure and pyloric activity displayed the same inhibitory direction (Figure 7f).

Figure 7:

A method for the recording of the pylorus sphincter. (a) A photograph of our custom-made half-bridge strain gauge transducer (2mm x 2mm, see stippled white box), each attached to the pylorus torus and antrum, for recording tone and motility. (b & c) Microinjection of in DMV as seen in the photograph. (d) The illustration depicts how an amplitude (left panel) or tone (right panel) was analyzed. The value of the peak-to-peak average is calculated based on the derivation of a sum of crest values of a wave within a given period divided by the number of phasic events. (e & f) CCK-8 and SNP induced responses used for verification of strain gage placement and directionality. [Note: differential effects of intravenous administration of 0.3μg/kg cholecystokinin octapeptide (CCK-8; IV) on pyloric and antral strain gage recordings. Please note the directionality of 50μg/kg sodium nitroprusside (SNP; IV) on pyloric tone and motility.]

Microinjection of L-glutamate (500pmol/30nl; n=20) into the DMV produced an immediate (13 ± 3 s) excitatory effect that was observed as a single pyloric contraction with an amplitude change of 2.6 ± 0.4 g compared to baseline (t=5.490, df=19; p<0.0001; Figure 8a). Changes in pyloric tone and frequency were not significantly affected (most probably due to the low baseline 0.20 ± 0.3 g; 0.21 ± 0.13 Hz). In all animals, L-glutamate re-challenge induced pyloric excitation within 10 min of the previous injection, showing no significant differences between first and second responses (2.10 ± 0.2 g; p>0.05). To control for vehicle effects, saline (0.9%) was microinjected into the DMV 10 min after L-glutamate (n=9). The saline vehicle failed to produce any significant changes in pyloric activity (t=3.537, df=8; n=9; Figure 8c), and microinjection of L-glutamate into the adjacent hypoglossal nucleus also had no significant effect (0.10 ± 0.3 g; 0.25 ± 0.09 Hz from baseline; n=4; p>0.05). All microinjection sites were histologically verified, and a representative coronal section with the location of pipette tracks for all DMV-rostral experiments appear Figure 8b.

Figure 8:

Microinjection of L-glutamate into the rostral half of the DMV causes a pylorus contraction. (a) Representative tracing showing the repeatability of L-glutamate-induced responses from the pyloric sphincter. (b) Camera lucida drawing depicting L-glutamate unilateral microinjection sites before and after saline-vehicle treatment, ipsilateral vagotomy, atropine methyl bromide, or site-specificity studies outside the DMV. Insert in (b) is a phase-contrast photomicrograph of a representative pipette track (red) within the DMV (arrow). (c) Microinjection of L-glutamate (left trace) or vehicle (right trace) into the DMV. (d) L-Glutamate elicited effects on pyloric activity (left trace) blockade ipsilateral vagotomy (right trace). (e) The increase in pyloric phasic contraction amplitude induced by L-glutamate's microinjection into the DMV is blocked by intravenous (IV) administration of atropine methyl bromide. (f) A bar graph showing the change in pyloric tone elicited from the DMV by microinjection of L-glutamate before and after IV administration of atropine. Abbreviations: 4V, fourth ventricle; 12N, hypoglossal nucleus; AP, area postrema; CC, central canal; DMV, dorsal motor nucleus of the vagus; NTS, nucleus tractus solitarius; TS, solitary tract.

To determine if the excitatory pyloric response was evoked from the DMV and not from the overlying nucleus of the solitary tract (NTS), an ipsilateral cervical vagotomy was performed in twenty animals. This procedure functionally separates responses elicited from the DMV from those of the nearby NTS; DMV-origin responses are blocked by ipsilateral vagotomy, whereas those elicited from the mNTS require bilateral vagotomy for their inhibition (Ferreira, 2000; Ferreira et al., 2002; Cruz et al., 2007). L-glutamate injected into the DMV produced an excitatory response of 2.60 ± 0.4 g that was blocked by ipsilateral vagotomy (0.10 ± 0.3 g; t=5.490, df=20; p<0.0001; Figure 8d), thus establishing it to be site-specific to the DMV.

Based on the previous vagal stimulation studies (Allescher et al., 1988a), we wanted to investigate if acetylcholine (ACh) was the neurotransmitter phenotype of this vagal efferent pathway of the pylorus. Hence, L-glutamate (500pmol/30nl) was microinjected into the DMV in the absence or presence of atropine methyl bromide (1mg/kg; IV), a muscarinic antagonist that due to its quaternary structure does not cross the blood-brain barrier. L-glutamate produced a significant increase in the pyloric phasic contraction amplitude (1.6 ± 0.6 g; p<0.05; n=13), which was blocked following administration of atropine (−0.4 ± 0.4 g; t=4.988, df=22; p<0.0001; Figure 8e-f). However, after atropine administration, an increase in pyloric tone was seen in 5 of the 13 animals (0.4 ± 0.1 g; time to onset 2.6 ± 0.2 min), which never returned to baseline for the duration of the experiment (~4hr). [Note: Before terminating the experiment, SNP 50 μg/kg was administered IV to determine the drug response directionally (see Methods).]

Effects of L-glutamate microinjection into the caudal half of the DMV on pyloric decreases in contractions and tone:

L-glutamate microinjection into the caudal half of the DMV decreases pyloric motility and tone (−0.2 ± 0.1 g; n=13) to an extent 34 ± 9% of that induced by SNP (−0.5 ± 0.1 g; t=4.56, df=24; p<0.001; 50μg/kg; IV; Figure 9a). The decrease in tone occurred immediately after L-glutamate's microinjection (time to nadir 35 ± 10 s). A complete return to baseline values was seen within 1 minute of the microinjection. These L-glutamate-induced decreases in pyloric tone were repeatable after a 10 min interval (−0.2 ± 0.0 g; 42% of SNP response; p<0.05; Figure 9c). Changes in pyloric phasic contraction frequency could not be assessed as the decrease in pyloric tone from L-glutamate microinjections depressed all pyloric activity for the drug effect duration. Saline vehicle effects were assessed in all 13 animals, and no significant difference in tone from baseline was evident (0.1 ± 0.1 g; −13 ± 5% of the SNP response; Figure 9a). Statistically significant differences were present between L-glutamate-induced responses and those of saline vehicle microinjected into the DMV (t=4.56, df=24; p<0.001). Site-specific control injections into brain regions surrounding the DMV were done in 18 animals. In every animal, L-glutamate microinjected into these sites did not significantly affect pyloric tone (−0.02 ± 0.04; 10% of SNP response). All microinjection sites were histologically verified and are represented on a coronal section shown in Figure 9b.

Figure 9:

Microinjection of L-glutamate into the caudal half of the DMV causes pylorus inhibition. (A) Representative tracing showing the pyloric sphincter's responses of microinjection of L-glutamate (left trace) or vehicle (right trace) into the DMV. (B) Camera lucida drawing depicting L-glutamate unilateral microinjection sites before and after saline-vehicle treatment, ipsilateral vagotomy, atropine methyl bromide, or site-specificity studies outside the DMV. Insert in B is a photomicrograph of a representative pipette track (red shading) within the DMV border. (C) Tracing showing the repeatability of L-glutamate-induced inhibition responses from the pyloric sphincter (left trace) that is not blocked ipsilateral vagotomy (right trace). (D) The decrease in pyloric phasic contraction amplitude induced by L-glutamate's microinjection into the DMV is blockade by intravenous (IV) administration of atropine methyl bromide, and L-NG-Nitroarginine methyl ester (L-NAME). Abbreviations: CC, central canal; DMV, dorsal motor nucleus of the vagus; NTS, nucleus tractus solitarius.

Since activation of neurons in the NTS by L-glutamate induces relaxation in gastric tone (see, e.g., Ferreira et al., 2002; Cruz et al., 2007; Herman et al., 2009), ipsilateral vagotomy was performed to assess the extent to which activation of this nucleus contributed to the relaxatory response. Microinjection of L-glutamate (n=13) in the DMV, before ipsilateral vagotomy, produced a 34 ± 9% drop in pyloric tone compared to SNP (−0.2 ± 0.1 g), which was significantly attenuated after ipsilateral vagotomy to 8.5 ± 7.2% of SNP (−0.04 ± 0.04; t=2.21, df=24; p=0.037; Figure 9c). Attempts to identity the post-ganglionic neurotransmitter(s) involved in the above pyloric response were examined in a series of studies. Previous studies evaluating electrical vagal stimulation (Allescher et al., 1988a) determined that the vagal efferent fibers innervating post-ganglionic neurons on the pylorus were non-adrenergic non-cholinergic (NANC) in origin, with likely transmitter candidates being nitric oxide (NO) and adenosine triphosphate (ATP) (Soediono and Burnstock, 1994; Ishiguchi et al., 2000b). However, before the role of these neurotransmitters were investigated, it was imperative to establish that the pylorus's mediated relaxation was not via the established cholinergic-cholinergic pathway. To this, atropine methyl bromide was administered (1mg/kg; IV). In 5 animals, L-glutamate produced an initial change in pyloric tone of 41 ± 10% of the SNP response (−0.2 ± 0.1 g; p<0.05) that was unaffected after administration of atropine (−0.2 ± 0.2 g or 33 ± 12% of SNP response; Figure 9d). Similar was that reported above, in three out of the five animals, atropine caused an increase in pyloric tone (0.3 ± 0.1 g; p<0.05).

To determine the role of NO in eliciting the decrease in pyloric tone, L-glutamate was microinjected into the DMV before and after administering the nitric oxide synthase inhibitor, L-NG-Nitroarginine methyl ester (L-NAME; 10mg/kg; IV). L-glutamate was injected at two different time points, 4- and 14-min post-L-NAME administration. The rationale for a 2-time point assessment on the effect of L-NAME to attenuate L-glutamate is based on the ability of L-NAME to cross the blood-brain barrier. At 4 min post-administration, L-NAME has a dominant peripheral action. By 14 min, L-NAME can penetrate the blood-brain barrier and thus mediate its effects on the nitrergic neurons of the CNS (Ferreira et al., 2002; Rogers et al., 2003; Ferreira et al., 2005). Nitrergic neurons that reside in the NTS can have functional effects on the stomach by modulating NTS 2nd order projecting neurons (Ferreira et al., 2002). In 5 animals, at both 4 and 14 min post-administration of L-NAME, L-glutamate mediated decrease in pyloric tone was not significantly affected (12 ± 41% of SNP response and 29 ± 6% of SNP, respectively; p<0.05) compared to prior L-glutamate injection (35 ± 6% of SNP response p<0.05; Figure 9e). Also, L-NAME's intravenous administration caused a significant increase in pyloric tone (0.23 ± 0.08 g; p<0.05; n=5).

Discussion

The study's main findings regarding brain control of the pylorus were two-fold: brain nuclei that comprise the neural circuity of the pyloric sphincter were identified, and the neurotransmission of the premotor neuron in the DMV regulating its function was revealed.

Transneuronal retrograde studies label higher-order pre-vagal' neurons in nuclei of the hindbrain, midbrain, and forebrain, whose distribution differs from those reported for other areas of the stomach. Brain nuclei with distinct labels were found in the hindbrain raphe nuclei, midbrain Edinger-Westphal nucleus, ventral tegmental area, lateral habenula, and arcuate.

The neural connections from the brain to the pylorus have been studied in detail by several investigators (Elfvin and Lindh, 1982; Hayakawa et al., 2003; Zalecki, 2012). However, due to their use of monosynaptic retrograde tracers (e.g., HRP, CTb, WGA, FG, etc.), these studies were only able to identify the distribution of first-order neurons. In the brain, these neurons were found to be topographically distributed exclusively in the medial half of the DMV, with the most significant density found at the level of the area postrema (Hayakawa et al., 2003). Our PRV-152 transneuronal studies corroborated their observations.

To address the location of higher-order neurons that regulate the pylorus, we used the transneuronal PRV-152 virus. This virus has been successfully used to label brain nuclei when injected into the different areas of the stomach (Rinaman et al., 1993; Yang et al., 1999; Rinaman et al., 2000; Rinaman and Schwartz, 2004; Balcita-Pedicino and Rinaman, 2007; Niedringhaus et al., 2008b) and other peripheral areas (Billig et al., 2000; Billig et al., 2003). Injection of PRV-152 into the pylorus, similar to the studies of in the stomach (e.g., the antrum in Rinaman et al., 1993; Rinaman and Schwartz, 2004) labeled brain nuclei (e.g., paraventricular, dorsomedial and lateral hypothalamic nuclei, amygdaloid complex nuclei, etc.) along the hindbrain-forebrain axis that is known to be involved in the viscerosensory function. Additionally, the virus labeled other nuclei that include the midline raphe, Edinger-Westphal, ventral tegmental area, lateral habenula, and arcuate.

By differentially projecting to the pylorus, these distinct nuclei could selectively influence it compared to the other regions of the stomach, including the antrum and the fundus. Although electrophysiological studies (Browning and Travagli, 1999) suggest that gastric-projection DMV neurons have a higher frequency of action potential firing to depolarizing current when with compared than intestinal-projection DMV neurons (e.g., duodenum), the pylorus comprises tissue that has characteristics of both intestinal and gastric-origin (Schulze-Delrieu and Shirazi, 1983; Daniel et al., 1992; Kressel et al., 1994).

Interestingly, another nucleus labeled from PRV-152 injection into the pylorus, which has not been reported for other areas of the stomach, was the Edinger-Westphal nucleus (EW). This midbrain nucleus comprises two distinct cell phenotypes: cholinergic neurons that are part of the oculomotor complex and neuropeptidergic neurons (e.g., urocortin-1, NPY, etc.) that are thought to play a role in stress, weight, and sympathetic based functions (Vaughan et al., 1995; Koylu et al., 1998; Vetter et al., 2002; Gaszner et al., 2004). This neuropeptidergic neurons project throughout the brain (Douglass and Daoud, 1996; Koylu et al., 1998; Bittencourt et al., 1999; Kozicz et al., 2011). Due to its complexity and the confusion in its literature, it has been proposed by some that the EW should be divided into two nuclei, a preganglionic (EWpg) and an (EWcp) central projecting. The present transneuronal tracer findings show distinctly labeled neurons forming a chain that is oriented dorsoventrally within the EW nucleus (see Figure 5f and Figure 6b). Therefore, these DAB positive neurons are more than likely of EWcp origin, which would be a logical assumption that the pylorus, a gastric region, has not been reported to be involved with oculomotor responses.

PRV injection of the pylorus also showed distinct labeling within the VTA. This area comprises pathways associated with reward and is implicated in addiction. These reward-centered nuclei could influence the pylorus as a critical mediator of gastric emptying as a means to control satiety and satiation (Janssen et al., 2011). A reward reinforcement of energy-dense foods via activation of dopaminergic neurons of VTA (Figlewicz et al., 2007) could provide positive/negative feedback to the stomach influencing the continuation or cessation of its consumption.

In summary, PRV labeling from the pylorus produces distinct labeling in multiple nuclei that have been implicated, amongst others, in the regulation of energy balance and hedonic behavior (e.g., arcuate, parabrachial, PVN, habenula, ventral tegmental area). In particular, hedonic behavior as it relates to drug and food craving has of late received a lot of attention (for review, see Reece, 2011), as it is thought to involve a complex interplay of anorexigenic (melanocortins) and addiction (e.g., opioids) signaling pathways. Thus, it is legitimate to speculate that one way by which these higher-order nuclei may influence appetitive behaviors is through modulating pyloric activity, and hence gastric emptying.

The significant findings of functional studies of the pyloric sphincter are: (1) activation of the rostral half of the DMV increases the phasic contractility of the pylorus that is vagally mediated by a postganglionic cholinergic pathway, and (2) stimulation of the caudal half of the DMV inhibits pyloric activity, the neurotransmitter(s) of which is yet to be determined.

Based on a retrograde tracer study (Hayakawa et al., 2003), vagal projection from DMV to pylorus arises from neurons that are medially located in the nucleus. In particular, their density is most pronounced at ~0.5-0.7 mm rostral to the CS. Hence, in our in vivo microinjection studies, we initially targeted this part of the medial portion of the DMV to determine its influence on pyloric activity. Stimulation with L-glutamate of this area of the DMV invariably caused a large amplitude single-event phasic contraction of the pylorus. Ipsilateral vagotomy in the same animals blocked this excitatory effect confirming that it was of DMV origin. Incidentally, ipsilateral vagotomy also affected pyloric tone, but this was not consistent, which corroborates the observations of others (Schulze-Delrieu and Shirazi, 1983; Mearin et al., 1986; Allescher et al., 1988a). We found that it could produce an increase in any given animal, a decrease, or no effect at all. It is possible that this may be due to the internal state of each of the animals, including the plane of anesthesia, metabolic state, etc.

Alternatively, a change in pyloric tone may depend on underlying vagal anatomy, i.e., conditional on which vagus is cut, the left or the right. However, this is unlikely, as studies in humans have shown that the state of the pyloric tone is independent of which vagus is cut (Schulze-Delrieu and Shirazi, 1983; Mearin et al., 1986).

The cholinergic nature of the excitatory vagal pathway to the pylorus was revealed by cholinergic receptor blockade with atropine methyl bromide (a quaternary structure that does not cross the blood-brain-barrier). Intravenous administration of atropine blocked L-glutamate-induced excitatory pyloric response from the DMV. Interestingly, following atropine treatment, an increase in pyloric tone was generally observed. This was unexpected as inhibition of ACh-mediated effects typically results in the inhibition of motility with no change in tone as seen in intragastric pressure recordings (Ferreira et al., 2002; Cruz et al., 2007). It is conceivable that this atropine-mediated effect on pyloric tone may be due to the deactivation of a local antropyloric circuit. The observed that short periods of pyloric inhibition are preceded by antral contractions (Allescher et al., 1988a).

Moreover, antral distension inhibits the pylorus (Daniel et al., 1992; Ishiguchi et al., 2001). The likely neurotransmitter mediating this response may be NO, which is dependent on the presence of interstitial cells of Cajal (ICC; Burns et al., 1996). Since ACh is known to affect these cells via muscarinic receptors, there could be a potential interaction of ACh on ICC to enhance NO transmission, the level of which would depend on the basal pyloric activity. Atropine would disrupt this interaction and mitigate basal NO transmission to the pylorus, thereby increasing the tone.

These data add to the already established view that the rostral half of the DMV contains gastric-projecting neurons that excite the smooth muscle of the stomach via cholinergic-cholinergic neurotransmission (Rossiter et al., 1990; Ferreira et al., 2002; Ferreira et al., 2005; Cruz et al., 2007; Niedringhaus et al., 2008a). The vagal stimulation studies (Allescher et al., 1988b) implied the existence of an excitatory cholinergic pathway that innervated the pylorus.

Microinjection of L-glutamate into the caudal area of the DMV induced relaxation of the pyloric sphincter. This response was blocked by ipsilateral vagotomy, thus, establishing the response to be of efferent vagal motoneuron origin, similar to that generated from the rostral half of the nucleus. These observations are similar to those seen with DMV-induced relaxation of the LES (Rossiter et al., 1990; Abrahams et al., 2002; Niedringhaus et al., 2008a).

To determine the neurotransmitter pathway responsible for pyloric relaxation activated from the DMV, nitric oxide (NO) was chosen as a likely candidate. This neurotransmitter was based on several observations. First, NO is generally known to relax smooth muscle tone (Oliveira et al., 1992; Bayguinov and Sanders, 1993; Soediono and Burnstock, 1994; Ishiguchi et al., 2000b; Toda and Herman, 2005). Second, there are descending inhibitory projections from antral myenteric ganglia to the pylorus that are positive for NO/VIP/Galanin (Deloof and Rousseau, 1985; Allescher et al., 1988b; Allescher and Daniel, 1994; Ishiguchi et al., 2001; Lindestrom and Ekblad, 2002; Toda and Herman, 2005; Cruz et al., 2007). Third, it has been shown in multiple animal models, including humans, that nNOS positive neurons innervate the myenteric and submucosal ganglia of the pylorus (Allescher et al., 1992; Kressel et al., 1994; Altdorfer et al., 1996; Tomita et al., 1999; Toda and Herman, 2005). Additionally, loss of nitrergic neurons results in hypertrophy and dysfunction of the sphincter (Watkins et al., 2000; Sivarao et al., 2008).

To assess the role of NO in the caudal DMV mediated inhibitory responses, we microinjected L-glutamate before and after intravenous administration of the nNOS antagonist L-NAME. Two time-points were examined to separate the nitrergic peripheral effects from centrally at 4 min and 14 min L-NAME. Since nitrergic neurons are present in both the NTS and DMV and modulate gastric activity (Panico et al., 1995; Lingenfelser et al., 1997; Ferreira et al., 2002; Rogers et al., 2003; Ferreira et al., 2005), the earlier time-point was used to establish peripheral based effects, while the latter was used to determine effects that were of central origin. Interestingly, L-NAME at the earlier time point seemed to attenuate the pyloric response, but not significantly. At the latter time-point and beyond, we did not see any effect of the L-glutamate-induced relaxation of the pylorus from the DMV. A possible explanation for this is that NO release at the pyloric smooth muscle could be responsible for this vagal efferent response; however, it may be possible could be overcome by inhibiting nitrergic interneurons of the NTS, which could mediate effects on other gastric regions (Ferreira et al., 2002) and therefore, through local projections influence the pylorus. L-NAME did significantly increase pyloric tone, as well as increase the amplitude of phasic contractions. This suggests that NO can antagonize the release of ACh and directly oppose its action at the smooth muscle. It also provides evidence of a NO-mediated basal tone on pyloric activity.

Although we did not assess its role, we cannot rule out adenosine triphosphate (ATP) and vasoactive intestinal polypeptide (VIP) as a possible postganglionic neurotransmitters that may be involved in the inhibitory response induced from the DMV. Both in vitro (Soediono and Burnstock, 1994; Ishiguchi et al., 2000a) and in vivo (Ishiguchi et al., 2000a) studies have shown a role for ATP in mediating relaxation of pyloric smooth muscle. Furthermore, there is evidence that either P2x or P2y purinoceptor activation affects smooth muscle relaxation (Ishiguchi et al., 2000a). Additionally, VIP is present in the pylorus (Lindestrom and Ekblad, 2002) and is responsible for the slow-phase relaxation of the LES that is elicited from the DMV (Niedringhaus et al., 2008a).

In summary, similar to the LES, site-specific stimulation of the DMV induces different functional responses in the pylorus mediated by divergent neural pathways. Excitation of the rostral half of the nucleus contracts the pylorus via a cholinergic-cholinergic pathway. In contrast, activation of its caudal half relaxes the sphincter but by a different pathway (s) whose identity is yet to be established.