Adult Sprague-Dawley (SD) rats (protocol number: 0013219, Qingdao Institute for Drug Control, Shandong, China) weighing 250 ~ 300 g were housed under controlled illumination (12:12-h light-dark cycle starting at 8 a.m.) and temperature (25 ± 2°C) and had free access to laboratory chow pellets and tap water. Animal Guidelines were approved by the Qingdao University Animal Care and Use Committee.

Rats were divided randomly into 2 groups: control group (n = 4) and gastric distension (GD) group (n = 8). The GD was performed as described previously (Gong et al., 2013). All rats were fasted overnight and anesthetized with Inactin (100 mg/kg, i.p.; Sigma-Aldrich Chemical, USA). Following a midline cut on the abdomen, the fundus wall was incised and gastric contents were washed out with warm isotonic saline. A latex balloon (length: 3 cm, diameter: 3 cm, max volume: 12 ml) attached to a polyethylene catheter (PE-50) was inserted into the stomach via the incision and fixed by silk thread. The pylorus was ligated to prevent duodenal reflux into the stomach and change of gastric volume. The abdomen was then closed. The polyethylene catheter was connected to an electronic barostat (Distender Series II, G&J Electronics Inc, Canada). The balloon was dilated to a constant pressure of 60 mm Hg for 20 s with a 4-min inter stimulus interval for 2 h. The control group received no treatment.

Thirty minutes after GD, rats were perfusion-fixed with 4% paraformaldehyde. The brains were moved for post-fixing, dehydrating, embedding, and sectioning. The 15-μm slices through VMH were preincubated for 1 h with 0.01 M PBS (pH 7.4) containing 0.3% Triton X-100, and 5% goat serum, then were incubated overnight at 4°C with primary antibodies: rabbit anti-c-Fos antibody (1:100, sc-8047, Santa Cruz Biotechnology, U.S.A.) and rabbit anti-nesfatin-1 antibody (1:200, H-003-22; Phoenix Pharmaceuticals Inc, U.S.A.). After rinsing, the sections were incubated with goat anti-mouse (1:500, Cy3-conjugated, 115-165-003, Jackson ImmunoResearch, U.S.A.) and goat anti-rabbit (1:50, FITC-conjugated, 111-095-003, Jackson ImmunoResearch, U.S.A.) fluorescent secondary antibody at room temperature for 90 min. After washing off unbound secondary antibodies with PBS, the sections were mounted in Citifluor (Citifluor, London, UK). Photographs of fluorophores were taken under a TCS SP8 two-photon laser scanning confocal microscope (Leica Microsystems AG, Wetzler, Germany). The mean percentage increase of c-Fos positive cells in the VMH was calculated as: (average c-Fos positive cells in stomach-distended rat—average c-Fos positive cells in control rat)/(average c-Fos positive cells in control rat) × 100%.

Forty rats were prepared for GD following the protocol mentioned above, and the electrophysiological recording procedure was based on the previously described method (Li et al., 2013). The anesthetized rats were bound on a stereotaxic frame, and craniotomy was performed around the coordinates: −2.5 mm posterior to Bregma and 0.8 mm lateral. A four-barrel glass microelectrode (5–15 MΩ) was advanced in an increment of 10 μm with the aid of hydraulic micropositioner into the area of VMH (Bregma: P: −1.8 ~−3.2 mm, L (R): 0.5 ~ 1.0 mm, H: 8.5 ~ 9.2 mm) according to the rat brain atlas of Paxinos and Watson (Paxinos, 2007). One barrel was filled with 0.5 M sodium acetate and 2% Pontamine sky blue to record neural discharging, the other three barrels connected with multi-channel pressure injector (PM2000B; Micro Data Instrument Inc., NJ, USA) contained either: nesfatin-1 (Phoenix Pharmaceuticals, Burlingame, CA, USA), SHU9119 (an antagonist of melanocortin-3/4 receptor; Sigma-Aldrich Chemical, MO, USA), or normal saline (NS). The drugs (<1 nl) were superfused on the surface of neurons by a short pulse gas pressure (1500 ms, 5.0 ~ 15.0 psi).

Once the microelectrode was advanced into the VMH and a stable firing pattern was recorded, GD stimulus was conducted. The change of neuronal firing rate was calculated by the formula: 100% × (firing rate of neurons after treatment−firing rate of neurons before treatment)/(firing rate of neurons before treatment). If the mean firing frequency changed via gastric distension by at least 20% from the mean basal level, the neurons were divided into GD-responsive, GD-excitatory (GD-E), or GD-inhibitory (GD-I), and no-response subcategories. Spike data was recorded and processed using PowerLab data acquisition system (AD Instruments Pty Ltd, Australia).

To verify the correct location of the microelectrode, an iron deposit of pontamine sky blue was formed at the tip of the electrode after a direct current (10 μA, 20 min). Then the rats were perfused with 10% buffered Formalin solution and the brains were frozen. Coronal sections were cut through the regions of the hypothalamus, stained with neutral red, cleared with xyline, and coverslipped. All the recording and microinjection sites were verified under light microscope.

For direct VMH injections, a cannula was implanted. Fasted rats were anesthetized and fixed in the stereotaxic apparatus. After craniotomy, a stainless steel guide cannula (24 gauge, 1.5 mm distance, Plastics One, VA, USA) was implanted into the VMH (Bregma: P: −1.8 ~ −3.2 mm, L (R): 0.5 ~ 1.0 mm, H: 8.5 ~ 9.2 mm). After anchoring the cannula and sealing all skull openings with dental acrylic, a 26-gauge obturator was placed in the cannula. After the recovery period of 10 ~ 14 d, the drugs were injected with a needle (29-gauge) inserted into the cannula and connected to a syringe by a 10 cm piece of polyethylene tubing.

Food intake was calculated according to a previously published method (Chen et al., 2012). Thirty cannula-implanted rats were randomly divided into 3 groups: NS, low-dose nesfatin-1 (10 pmol) and high-dose nesfatin-1 (100 pmol) groups. Rats were fasted at 15:00 on the experiment day and administered NS or nesfatin-1 in the VMH 4 h later. Next, food was placed in the cage, and total food intake in the following 12 h was calculated by measuring the weight of food containers. The outputs were recorded by Data Acquisition software 51,800 (Feed-Drink Monitoring System Ver. 1.31, Ugo Basile, Italy).

The secretion of gastric acid induced by 2-deoxy-D-glucose was studied using an acute gastric fistula method described previously (Xia et al., 2012). Thirty rats with implanted VMH cannulas were randomly divided into 5 groups: NS, nesfatin-1 (100 pmol), 2- DG (200 mg/kg, sc) + NS, 2-DG + nesfatin-1, and 2-DG + nesftin-1 + SHU9119 (1 nmol) groups. Rats were fasted for 16 h and then anesthetized with urethane (1.5 g/kg ip). After incision at the abdominal midline, the esophagus was ligated at the gastroesophageal junction (to preserve the vagal trunks carefully). A double-lumen cannula was inserted into the stomach via the ligated pylorus. Basal acid secretion was established by collection of gastric outputs in 5 ml of perfusate every 15 min for 30 min. The gastric lumen was subsequently rinsed with 5 ml of saline every 15 min for 120 min after drugs were administered in the VMH or subcutaneously. The gastric rinse solution was then titrated to pH 7.0 with 100 mM sodium hydroxide using an automatic titrator (COM-2500; Hiranuma Sangyo Co Ltd, Japan) for titratable acidity.

To measure gastric motility, 30 rats with implanted cannulas were randomly divided into 5 groups: NS, nesfatin-1 (100 pmol), SHU9119 (1 nmol), nesfatin-1 + SHU9119, and nesfatin-1 (dorsomedial hypothalamic nucleus, DMN) group. Rats in the last group were implanted cannulas in ventral DMN as a specificity control for VMH. Gastric motility measurement was conducted as previously described (Guo et al., 2015). After rats fasted overnight and anesthetized with Inactin, laparotomy was performed to expose the stomach. For measurement of gastric motility, a strain gauge was sutured onto the serosa of the gastric antrum, 0.5 cm caudal from the pyloric ring. The lead wire of the strain gauge was finally fixed at the nape of the neck with a 2 ~ 3 cm outside end by extending through a subcutaneous tunnel. After recovery for 2 days, the rats were fasted overnight and recorded the gastric motility on a polygraph (3066–23; Chengdu Precision Instruments, Sichuan, China) by connecting the lead wires. The phase III-like contractions were defined as clustered potent contractions with amplitudes of more than 4 g. After achieving stable baseline recordings for at least 20 min, nesfatin-1 or SHU9119 was administered to the VMH. Each animal was recorded for 2 h on 2 different days. The changes in gastric motility were evaluated by the percentage motor index (%MI) of the motor activity in the antrum. Values of the %MI for a 5-min period in the antrum were calculated by 100% × (area under the manometric trace for each 5-min period after nesfatin-1 or vehicle injection)/(area under the manometric trace for the 5-min period immediately before nesfatin-1 or vehicle injection).

Gastric emptying was determined by the phenol red method with a non-nutrient viscous solution as described previously (Guo et al., 2015). Eighteen rats with implanted VMH cannulas were randomly divided into 3 groups: NS, nesfatin-1 (100 pmol), and nesfatin-1 + SHU9119 (1 nmol) groups. In the dark phase, fasted rats were stimulated with nesfatin-1 or the mixture of nesfatin-1 and SHU9119. Immediately, the animals were fed 1.5 mL of a non-nutrient viscous solution by gavage. Twenty minutes later, rats were sacrificed and the gastric contents were collected. Phenol red concentration was fitted to a standard curve.

After anesthetized and mounted on a stereotaxic apparatus, 10 rats were injected 3% Fluorogold (0.2 μl, Sigma-Aldrich Chemical, MO, USA) into the VMH. After a 7-day recovery, the rats were perfusion-fixed with 4% paraformaldehyde. The brains were removed for post-fixing, dehydrating, embedding, and sectioned. Fifteen micrometers sections through NAc were incubated overnight at 4°C with rat anti-nesfatin-1 antibody. After washing, the sections were incubated with goat anti-rat fluorescent secondary antibody at room temperature for 90 min. After washing off unbound secondary antibodies with PBS, the sections were mounted in Citifluor (Citifluor, London, UK). The photographs of fluorophores were taken under a TCS SP8 two-photon laser scanning confocal microscope (Leica Microsystems AG, Wetzler, Germany).

For electrical stimulation, the electrodes were implanted in the NAc shell. Rats were anesthetized and subsequently mounted in a stereotaxic frame. Craniotomy was performed around the coordinates: 1.8 mm posterior to Bregma and 1.7 mm lateral. Electrodes (David Kopf Instruments, Tujunga, CA, USA) were placed in the NAc shell (bregma: P: 0.8 ~ 2.8 mm, L (R): 1.0 ~ 2.5 mm, H: 5.0 ~ 7.0 mm) and secured to the skull with dental cement. After 3-d recovery, the electrode was attached to the cable from a stimulator (Master-9, AMP Instruments LTD, Israel) to stimulate nucleus with a radiofrequency output of square-wave current impulses.

Eighty rats with implanted stimulating electrodes in NAc shell were prepared for GD and electrophysiological recording in the VMH following the process mentioned above. Once a neuron in the VMH was confirmed to be the GD-responsive cell, the direct current stimulation (20 μA, 0.5 ms and 50 Hz) was added to NAc shell without or with the pretreatment of anti-nesfatin-1 antibody. Spike data was recorded and processed using PowerLab data acquisition system.

The detailed approaches of electrode implantation, food intake, gastric acid output, gastric motility, and emptying had been described above. Here are the animal groups in every experiment.

In food intake test, 40 rats was randomly divided into 4 groups: control, low (20 μA), high (50 μA) current, and high current + nesfatin-1 antibody groups. During test, the rats were placed in the test chambers and given access to a bowl of pre-weighed chow. The animals were electrically stimulated in the test chamber for 1 h (from 8 to 9 p.m.), and food was weighed at 2, 4, and 6 h after stimulation. In the high current + nesfatin-1 antibody group, nesfatin-1 antibody was pretreated in the VMH before the electrical stimulation.

Forty rats, implanted with electrodes in NAc shell, were used to check the secretion of gastric acid and were randomly divided into 5 groups: basal control, 50 μA, 2-DG (200 mg/kg, sc), 2-DG + 50 μA, and 2-DG + 50 μA + nesfatin-1 antibody groups. Rats in the 2nd, 4th, and 5th groups were stimulated with 50 μA current in the NAc shell, and nesfatin-1 antibody was injected into the VMH of rats in the 5th group.

To record the gastric motility after the electrical stimulation in NAc shell, 30 rats were randomly divided into 3 groups: control, 50, and 50 μA + nesfatin-1 antibody groups. Rats of the 2nd and 3rd groups received the 50 μA stimulation in the NAc shell and SHU9119 was only given to rats of the 3rd group.

In the gastric emptying study, 30 rats were randomly divided into 3 groups: the control group, 50 μA group, and 50 μA + nesfatin-1 antibody groups (n = 10 in each group).

Statistical analyses were processed with SPSS 15.0 statistics software (SPSS Inc, Chicago, IL, USA). The results are presented as the mean ± SEM. The Kolmogorov–Smirnov test was used to assess normal distribution. For data with a normal distribution, two-way repeated-measures ANOVA or Student's t-test was applied, while data that were not normally distributed were assessed for statistical significance using the Mann–Whitney U-test between two groups and Kruskal–Wallis test among more than two groups. A p < 0.05 was considered statistically significant.

To assess activation of VMH nesfatinergic neurons following GD, double c-fos, and NUCB2/nesfatin-1 antibody labeling was performed. GD significantly increased the number of c-fos immunoreactive neurons in VMH to 24.5 ± 4.9 neurons per microscope field, higher than 6.2 ± 0.8 in control group (P < 0.001, Figures 1A,B,G). The results of the nesfatin-1 positive cell counting between normal and GD groups were not statistically significant (P > 0.05, Figures 1C,D,G). Among the c-fos immunoreactive neurons, some cells were double-labeled with the nesfatin-1 antibody (Figures 1E,F). 8.45 ± 0.72 double-labeled neurons per microscope field were observed in the GD group, whereas 2.01 ± 0.24 were observed in the control group (P < 0.05, Figure 1G). Thus, GD activated nesfatinergic neurons in VMH.

In a pre-experiment, we tested the effects of different concentrations of nesfatin-1, including 0.1, 0.5, 1, 5, 10, 50, 100, and 200 nM, on the firing rate of VMH neurons. Our analysis of the dose-response revealed that the half maximal effective concentration (EC 50) of nesfatin-1 on firing rate is approximately 10 nM.

Out of a total of 88 VMH neurons in 40 rats, 64 (72.73%) were identified as GD-responsive neurons based on the change of firing rate elicited by GD. Among GD-responsive neurons, 42 (47.7%) showed an increase in firing rate following GD (GD-E), and 22 (25.0%) showed a decrease in firing rate following GD (GD-I).

Next we assessed the effect of nesfatin-1 (10 nM, ~1 nL) via electrode ejection on the surface of the GD-responsive neuron. In 42 GD-E neurons, nesfatin-1 excited 61.9% (26/42) neurons, inhibited 16.7% (7/42), and did not affected 21.4% (9/42). Overall, nesfatin-1 increased the firing rate of GD-E neurons by 54.48 ± 15.39% (from 4.13 ± 1.21 to 6.38 ± 1.48 Hz, P < 0.05, Figures 2A,C). Sixteen of Twenty-two (72.7%) GD-I neurons were inhibited by nesfatin-1, resulting in a significant decrease in the firing rate from 3.84 ± 1.10 to 2.31 ± 0.77 Hz (39.84 ± 9.18%, P < 0.05, Figures 2B,C). Further, pretreatment with the melanocortin 3/4 receptor antagonist SHU9119 partially blocked the effect of nesfatin-1 on the GD-responsive neurons (Figure 2). SHU9119 alone had no effect on these neurons.

Nesfatin-1 was microinjected into the VMH of freely moving rats via cannula (Figure 3A). Low (10 pmol) and high (100 pmol) dose nesfatin-1 reduced cumulative food intake by 15.21 ± 2.34 and 28.63 ± 5.38%, respectively (P < 0.05, vs. NS group) over a 2 h period. Cumulative food intake over 4 h was reduced in the high dose group (P < 0.05, 80.47 ± 14.62%), but not in the low dose group, compared with the NS group. No statistically significant effect of nesfatin-1 on cumulative food intake was found 6 h after drug delivery, illustrating the acute nature of nesfatin-1's effect.

To assess the role for VMH nesfatin-1 in gastric function, we first examined gastric acid secretion. VMH injection of nesfatin-1 (100 pmol) had no effect on basal gastric acid secretion (Figure 3B). However, gastric acid output induced by 2-DG was inhibited by VMH injection of nesfatin-1 (Figure 3B). The peak acid production was decreased by 48.63 ± 10.21% (P < 0.05, vs. 2-DG group). In the 2-DG + nesfatin-1 + SHU9119 group, the gastric acid output was not different than the 2-DG alone group, suggesting that nesfatin-1's mechanism of action in the VMH involve the melanocortin-3/4 receptor.

Next, we examined gastric motility via an implanted strain gauge in freely moving rats. Rats receiving nesfatin-1 in VMH showed a change in percent motor index (MI%), a measure of gastric motility, starting 4.1 ± 0.8 min after nesfaitn-1 delivery with the peak change in MI% 15.2 ± 3.1 min following nesfatin-1. The MI% 10–15 min after nesfatin-1 injection was 35.2 ± 8.9% lower than that of NS group at the same time point [Figures 3Ca,b,D, P < 0.05]. Pre-treatment with 1 nmol SHU9119 blocked the effect of nesfatin-1 (P < 0.05, vs. nesfatin-1 group), while SHU9119 alone did not change gastric motility [Figures 3Cc,d,D]. To investigate the site specificity of VMH for these effects, we unilaterally injected nesfatin-1 into ventral DMN, a nucleus close to VMH. The results showed that there was no significant change for the gastric motility after nesfatin-1 was administrated into DMN [P > 0.05 vs. NS group, Figures 3Ce,D].

Lastly, we examined the effect of VMH nesfatin-1 on gastric emptying in awake rats. VMH injection of nesfatin-1 (100 pmol) inhibited the gastric emptying rate of a non-caloric viscous solution to 50.5 ± 15.6% [P < 0.05, vs. NS group (94.4 ± 13.9%), Figure 3E]. As with gastric acid secretion and gastric motility, pre-treatment of 1 nmol SHU9119 blocked the effect of nesfatin-1 on gastric emptying (87.2 ± 9.2%, P < 0.05, vs. nesfatin-1 group).

To examine projections to VMH we performed fluorogold retrograde labeling in VMH. 7 d after fluorogold injection, we observed abundant retrograde-labeled neurons in the shell of NAc (Figure 4A). In the same tissue sections, nesfatin-1 immunostaining revealed several nesfatin-positive neurons in the NAc shell (Figure 4B). Several flurogold/nesfatin-1 double positive neurons were observed in the NAc shell (Figure 4C). Quantitative analysis indicated approximately 43.2% fluorogold-labeled neurons expressed nesfatin-1, and about 28.3% nesfatin-1 positive neurons contained fluorogold. These results indicate that nesfatinergic neurons in the NAc shell send axonal projections to the VMH. Figure 4D illustrated the position of immunohistochemical figures in the NAc.

Next, we examined the physiological effects of NAc shell modulation of VMH GD-repsonsive neurons. We applied a 20 μA current to NAc shell while simultaneously recording single unit discharges from GD-responsive neurons in the VMH. One hundred and three GD-responsive neurons (54 GD-E and 49 GD-I) in the VMH of 80 rats were found. NAc shell electrical stimulation alone further excited 32 of 54 GD-E neurons (3.73 ± 0.97 to 5.53 ± 0.97 Hz, P < 0.05, Figures 5A,C). Notably, pretreatment with nesfatin-1 antibody in the VMH decreased the firing rate to 4.18 ± 1.02 Hz (P < 0.05).

When GD-I neurons were examined, 26 of 49 (53.06%) were excited by NAc shell stimulation (2.74 ± 0.71 to 5.08 ± 1.15 Hz, p < 0.05, Figures 5B,C). Unlike GD-E neurons, the firing rate of GD-I neurons was increased from 5.08 ± 1.15 to 6.81 ± 1.13 Hz by the nesfatin-1 antibody (P < 0.05). Notably, injection of the nesfatin-1 antibody, alone, to VMH did not significantly modify the firing rate of the GD-responsive neurons (GD-E neurons: 3.34 ± 0.95 vs. 3.73 ± 0.97 Hz; GD-I neurons: 2.86 ± 0.91 vs. 2.74 ± 0.71 Hz, p > 0.05, Figures 5A,B).

Next, we examined the effect of Nac stimulation on food intake and gastric functions, in awake, behaving rats. Low (20 μA) current stimulation to NAc shell had no statistically significant effect on cumulative food intake (P > 0.05, vs. control group). High (50 μA) current stimulation decreased cumulative food intake at 2, 4, and 6 h to 55.88 ± 6.11, 63.73 ± 6.76, and 72.19 ± 7.56% of control, respectively (P < 0.05, vs. control group, Figure 6A). Interestingly, pretreatment with an anti-nesfatin-1 antibody in the VMH reversed the decrease of food intake induced by NAc stimulation (P < 0.05, vs. high current group, Figure 6A).

Fifty microampere current stimulation to NAc shell inhibited 2-DG-induced gastric acid output, reducing gastric acid output 15 min following 2-DG from 9.16 ± 1.53 to 5.52 ± 0.88 μEq (P < 0.05). This effect was partially blocked by administration of a nesfatin-1 antibody to VMH (7.68 ± 0.98 μEq at 15 min, P < 0.05 vs. 2-DG + 50 μA group, Figure 6B).

Examining the effect of NAc stimulation on gastric motility, MI% was reduced 72.25 ± 9.82% by 50 μA NAc shell stimulation (P < 0.05, vs. control group). This effect was partially blocked by a nesfatin-1 antibody, where MI% in the 50 μA + nesfatin-1 antibody group was only 94.32 ± 12.61% of control (P < 0.05, vs. 50 μA group, Figure 6C).

Gastric emptying was also affected by 50 μA NAc shell stimulation, where NAc stimulation reduced emptying of the non-caloric viscous solution by 68.36 ± 8.79% (P < 0.05, vs. control group). Nesfatin-1 antibody pretreatment partially blocked this effect, as gastric emptying in the 50 μA + nesfatin-1 antibody group was 90.91 ± 14.02% control (P < 0.05, vs. 50 μA current group, Figure 6D). These results illustrate that involvement of nesfatin-1 on NAc modulation of gastric functions via projections to the VMH.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.