We used male Sprague-Dawley rats (Laboratory Animal Services Center of Capital Medical University, Beijing, China) that ranged in weight from 210 to 230 g. All the animals were housed in animal care facilities at 22 ± 1°C on a 12:12 h light-dark cycle. Food and water were provided ad libitum. Every procedure was approved by the Animal Care and Use Committee of Capital Medical University and was conducted according to the established guidelines of the National Institutes of Health (NIH, USA). All the rats were anesthetized using isoflurane inhalation (2%) and killed by decapitation.

The animal model of Parkinson’s disease was established by bilateral SN microinjection of 6-OHDA (Sigma, USA) as previously reported (Zheng et al., 2011; Zheng et al., 2014). Briefly, all the animals were placed in a stereotaxic instrument. Two holes were drilled in the skull (coordinates: AP, −5.6 mm; ML, ±2.0 mm; DV, −7.5 mm), and 6-OHDA (4 μg in 2 μl of 0.9% saline containing 0.05% ascorbic acid) was bilaterally injected with a 10 μl Hamilton syringe. The control groups were injected with saline containing 0.2% ascorbic acid. The subsequent experiments were performed at 4 or 6 weeks after 6-OHDA administration.

In experiment 1 (Supplementary Figure S1), rats were randomly assigned to the control and 6-OHDA groups. At the end of the 6th week after saline or 6-OHDA injection, fresh brain tissue and gastric corpus tissue (control: n = 8; 6-OHDA: n = 8) were acquired for ELISA, western blot analysis, and ultra-performance liquid chromatography tandem mass spectrometry (UPLC/MS/MS). At the end of the 4th week (control: n = 3; 6-OHDA: n = 3) and 6th week (control: n = 3; 6-OHDA: n = 3) after saline or 6-OHDA injection, gastric tissue for haematoxylin and eosin (HE) staining and whole-mount preparations were obtained through abdominal aorta occlusion and stomach excision. Then, the rats were transcardially perfused, and their brains were postfixed for immunohistochemistry.

In experiment 2 (Supplementary Figure S1), gastric muscularis tissue (control: n = 10; 6-OHDA: n = 10) was obtained at the end of the 6th week after saline or 6-OHDA injection and was processed for ex vivo culture.

In experiment 3 (Supplementary Figure S1), rats were randomly divided into early and late treatment groups. Previous studies have demonstrated that gastric dysmotility is not obvious until 4 weeks after 6-OHDA injection (Song et al., 2014; Toti and Travagli, 2014). Therefore, in the early treatment group, the rats received intraperitoneal injection of PNU-282987 (PNU), a selective agonist of α7nAChR (early-PNU) or vehicle, once a day for 6 weeks on d 1 after 6-OHDA/saline microinjection. PNU were purchased from Abcam (Cambridge, UK). In the late treatment group, the 6-OHDA rats received daily injection of PNU (late-PNU) or vehicle from d 28 to d 56 after the 6-OHDA model was generated. Thus, both groups were subdivided into 4 groups: the control + vehicle (control) group; the control + early- or late-PNU (control + early- or late-PNU) group; the 6-OHDA + vehicle (6-OHDA) group; and the 6-OHDA + early- or late-PNU (6-OHDA + early- or late-PNU) group. The dose of PNU (0.38 mg/kg) was chosen according to published study study (Li et al., 2011).

In experiment 3, 32 rats were divided into 4 subgroup (each subgroup: n = 8) in the early treatment group. Gastric emptying (GE) rate was calculated, and gastric muscularis tissues were obtained for gastric strip contraction, whole mount studies, western blot analysis and real-time PCR. Fresh brain tissue from all rats was acquired for western blot.

In experiment 3, 32 rats were divided into 4 subgroup (each subgroup: n = 8) in the late treatment group. After in vivo gastric motility recording and GE tests, fresh brain and gastric muscularis tissues were obtained for western blot analysis, real-time PCR and electron microscopy.

The stomach was quickly immersed in ice-cold Krebs-Henseleit solution (K-HS) (composition: NaCl, 118.4 mM; KCl, 4.7 mM; MgCl₂ 7H₂O, 1.2 mM; NaHCO₃, 25 mM; KH₂PO₄, 1.2 mM; CaCl₂, 2.5 mM; and glucose, 11 mM). The luminal contents were flushed out with K-HS. The gastric corpus was pinned flat with the mucosal side up in a Sylgard-lined Petri dish containing ice-cold oxygenated K-HS. The mucosa was removed from the underlying tunica muscularis by peeling. The fundus was removed, and the entire gastric corpus was used in the subsequent steps. The tissues were immediately quick-frozen in liquid nitrogen or immediately fixed in 4% buffered formalin for the following experiments. The SN and dorsal medulla were collected on ice from the brains of rats following previously described methods (Zheng et al., 2011; Zheng et al., 2014).

To generate the brain sections, rats were perfused through the left ventricle according to a previous method (Song et al., 2014). The brains were then quickly removed and kept in 4% paraformaldehyde for 24 h post fixation. After dehydration with 15 and 30% sucrose in 0.01 M PBS (pH 7.4), coronal frozen sections including the SN, were cut at a thickness of 20 μm with a cryostat (Leica CM1950, Switzerland). The brain sections were air-dried overnight at room temperature and stored at −80°C.

The longitudinal muscle/myenteric plexus (LMMP) whole-mount preparations were dissected from the gastric corpus as previously reported (Yuan et al., 2001; Xue et al., 2009). Briefly, whole mounts of the gastric muscularis were prepared as previously described. The stomach was quickly excised and opened along the lesser curvature, pinned flat in a Sylgard-coated Petri dish and fixed overnight in 4% paraformaldehyde (pH 7.4) at 4°C. The mucosa and submucosa were removed using fine forceps under a stereomicroscope, and the whole-mount preparations were collected separately in PBS supplemented with 0.1% sodium azide.

The levels of TNF-α, IL-1β, and IL-6 in the gastric muscular tissues were measured by means of ELISA kits (Multi Sciences (Lianke) Biotech Co. Ltd., China). The tissue samples were weighed, homogenized and centrifuged at 11,000 g for 10 min at 4°C. The protein concentration was quantified with the BCA Protein Assay Kit (Thermo Fisher, USA) and the levels of TNF-α, IL-1β, and IL-6 were expressed as picograms per milligram protein.

For Haematoxylin-eosin (HE) staining, the tissues were sliced into 5-μm-thick sections using a cryostat (Leica, CM 1850) after dehydration and then stained with hematoxylin and eosin (Beyotime, China).

For immunohistochemistry, the sections were quenched using 0.3% H₂O₂ to inactivate the endogenous peroxidase and then heated to 95°C-100°C in a beaker containing citrate buffer (0.01 M, pH 6.0) for 15 min for antigen retrieval. After being washed in PBS (3 × 5 min), the sections were incubated with 10% normal goat serum for 1 h at room temperature and incubated with primary antibodies (Table 1) at 4°C overnight. Immunostaining was performed using the PV-9001 or PV-9002 Polymer Detection System with diaminobenzidine according to the manufacturer recommendations, and then, the samples were counterstained with hematoxylin. The sections were photographed with a conventional microscope (Nikon E80i, Japan). For gastric whole-mount preparation, the macrophages in three randomly selected areas of the myenteric plexus in each preparation were counted by Image-Pro plus 6.0.

The details of the immunofluorescence experiments are described in our previous reports (Zheng et al., 2011; Zheng et al., 2014). Briefly, the sections were permeabilized with 0.3% Triton X-100 for 15 min, and then blocked with 10% donkey serum for 30 min at room temperature. The sections were treated with a mixture of two primary antibodies derived from different host species at 4°C overnight (CD163/α7nAChR, CD68/α7nAChR). After washing with PBS, the sections were incubated for 1 h at room temperature with a mixture of the secondary antibodies. DAPI was used to stain the nuclei. Each first and second antibody are listed in Tables 1, 2, respectively. Photomicrographs were taken using a laser scanning confocal microscope (Olympus, FV1000).

The gastric samples from the control and 6-OHDA rats were subjected to ultrastructural observation. The pieces of the gastric muscularis region were placed in a fixative made of 2.5% glutaraldehyde in 0.1 M phosphate buffer at pH 7.2 for 2 h at 4°C. The specimens were washed for 3 × 20 min in the same buffer and subsequently postfixed with 1% osmium tetroxide (OsO₄) in the same buffer for 2 h at 4°C. Then, the specimens were rinsed with distilled water, block-stained with a saturated aqueous uranyl acetate solution for 2 h, dehydrated in a graded series of cold alcohols and embedded in SPI-Pon 812 resin. Thin sections (70 nm) were cut using a Leica ultramicrotome (Leica EM UC7, Germany), double-stained with uranyl acetate and lead citrate, and observed with a HITACHI HT7700 (Japan) transmission electron microscope.

The frozen gastric muscular layer (30 mg) was homogenized in 300 μl cold lysis buffer (1% Nonidet P-40, 10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% SDS, 1 mM EDTA, 5 mg/ml leupeptin, 5 mg/ml aprotinin, 1 mM PMSF, 0.5% deoxycholic acid, and 1 mM sodium orthovanadate, all purchased from Sigma-Aldrich) by a sonicator for 10 s until completely dissolved. The homogenates were centrifuged at 12,000 g for 30 min at 4°C. For the studies of NF-κB signaling, the nuclear and cytosolic proteins were extracted using the Nuclear and Cytoplasmic Extraction Kit (Sangon, China). The protein concentration was quantified by the BCA protein assay kit (ThermoFisher Scientific, USA). 50 μg total protein from each sample was loaded into each lane for SDS-PAGE (10% SDS gel). After electrophoresis, the proteins were transferred onto a polyvinylidene fluoride membrane (Merck Millipore, Germany), and the membrane was washed for 10 min with TBST (20 mM Tris-Cl, pH 7.5, containing 0.15 M NaCl and 0.05% Tween-20) and blocked with 5% nonfat milk in TBST for 1 h at room temperature. The membranes were incubated with primary antibodies overnight at 4°C, followed by incubation with the appropriate secondary antibodies (Table 2) for 1 h at room temperature. The western blotting results were semiquantified with ImageJ software (version 1.8.0).

The in vitro organ culture was performed following a previously published report (Kreiss et al., 2004). Briefly, the stomachs were removed, opened, and washed three times with RPMI 1640 medium containing 100 μg/ml streptomycin and 100 U/ml penicillin under sterile conditions. The tissue was pinned on a Sylgard plate, and the muscularis was then further dissected, weighed and placed in individual wells of 24-well polystyrene plates in the absence or presence of α7nAChR agonists (PNU or GTS-21) or antagonist methyllycaconitine citrate (MLA). GTS-21 and MLA were purchased from Abcam (Cambridge, UK). After an incubation period of 2 h in RPMI 1640 at 37°C and 5% CO₂, the cultured tissue was inspected for contamination and cultured for 2 h. The cultured tissue was blotted dry, immediately quick-frozen in liquid nitrogen and stored at –80°C.

Total RNA was extracted from the gastric muscularis using TRIzol Reagent (Invitrogen, USA). First-strand cDNA was synthesized following the protocol of the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, USA). The sequences of the primers and the expected product sizes are listed in Table 3. The qPCR was performed with the TransStart Top Green qPCR SuperMix Kit (TransGen Biotech, China) on a Bio-Rad CFX96 Real-Time System (Bio-Rad Laboratories, C1000 Thermal Cycler, Singapore). The 2^−ΔΔCt method was used to determine the relative amount of mRNA.

The methods have been previously described (Song et al., 2014), rats were sacrificed by decapitation, and the gastric corpus was isolated. The luminal contents were immediately removed, and the mucosal and submucosal layers were gently removed. The gastric longitudinal muscle strips (approximately 10 mm in length and 1.5 mm in width) were prepared and suspended vertically under a resting tension of 10 mN in an organ bath containing K-HS that was maintained at 37°C and bubbled with 95% O₂/5% CO₂ (pH 7.4). The responses of the strips stimulated with carbachol (10⁻⁸–10⁻⁴ M) were measured isometrically with a force transducer (MLT0201/RAD; AD Instrument, Australia) on a computer with the PowerLab system (AD Instruments, Australia). The magnitude of the absolute force was normalized to the wet weight of each strip according to a previous study (Tsuchida et al., 2011).

The methods were previously described (Zheng et al., 2011). Briefly, rats were fasted for 24 h with free consumption of water before the GE experiments. For the GE of solid food, the fasted rats had free access to preweighed food pellets for 1 h and then were deprived of food for 4 h. The amount of food consumed was measured by weighing the food before and after feeding. After decapitation, the stomach was exposed via laparotomy and quickly ligated at both the cardium and pylorus. The stomach was removed. The stomach contents were removed by splitting the organ along the greater curvature and then weighed. The GE rate was calculated according to the formula: GE (%) = (1-dry weight of food in the stomach/weight of food consumption) × 100 (Zheng et al., 2011; Zheng et al., 2014).

The methods were previously described (Zheng et al., 2011). The GE of a barium meal was also evaluated by an in vivo X-ray digital imaging system. After fasting for 24 h, rats received 2.5 ml of barium sulfate (2.5 ml, 70% wt/vol)) through oral gavage. Gastric images were obtained using the KODAK in vivo Imaging System. The exposure time was 30 s. Images were then recorded at 30, 60, 90 and 120 min after the barium meal. The GE rate at different time points was calculated by Image-Pro Plus software (version 6.0) according to the following formula: GE (%) = [1- (barium meal at 30, 60, 90 or 120 min)/(barium meal at 0 min)] × 100%.

The methods were previously described (Zheng et al., 2014; Liu et al., 2018). Briefly, after fasting for 24 h, rats were anesthetized. After laparotomy, a strain gauge force transducer (WS100; Xinhang Xingye 349 Tech. Co., Beijing, China) was sutured onto the serosal surface of the gastric antrum (0.5 cm caudal from the pyloric ring) to measure the contractile activity of the longitudinal muscle. The abdomen was then closed, and the wires from the transducer were connected to the recording system (ML112, PowerLab, AD Instruments, Australia). After stabilizing for 2 h, the contraction data were collected by the “Chart & Scope” software (PowerLab, AD Instruments, Australia). The normal migrating motor complexes (MMCs), which include a 4-part cycle comprising phases I, II, III, and IV, were identified in the stomach. The contraction amplitude of phase III was evaluated.

The methods were previously described (Zheng et al., 2011). Briefly, the content of ACh in the muscularis externa of the stomach was measured by the UPLC/MS/MS method. The tissue was accurately weighed and homogenized with 2% aqueous formic acid (10 μl/mg). The homogenate was further ultrasonically dissociated with 2 ml of the reconstitution solvent (acetonitrile/methanol/formic acid, 750/250^/2, V/V^/V) and then centrifuged for 20 min (4°C, 12,000 g). The supernatant was dried with nitrogen, redissolved with 200 μl reconstitution solvent, and centrifuged (4°C, 12,000 g, 10 min). This supernatant was used immediately for analysis (Key Laboratory of Radiopharmaceuticals, Ministry of Education, College of Chemistry, Beijing Normal University).

All data are presented as means ± SD. Statistics and graphs were generated using GraphPad Prism, version 8.0 (GraphPad Software, USA). Differences between two or more groups were analyzed by Student’s t-test or ANOVA followed by Turkey multiple comparisons test. When data were not normally distributed or variances were different, we used the Mann-Whitney U test or the Kruskal-Wallis test followed by Dunn’s multiple-comparisons test. Values of p < 0.05 were considered significant.

To confirm that 6-OHDA induced degeneration in the vagal brain-gut axis, the expression of choline acetyltransferase (ChAT), the rate-limiting enzyme for ACh synthesis, was examined in the dorsal medulla and gastric corpus after bilateral 6-OHDA microinjection. As previously observed, the 6-OHDA rats showed a marked reduction in the number of TH-immunoreactive (TH- IR) neurons (Figure 1A a and b) and the protein expression of TH in the SN (Figure 1B). Immunohistochemistry and western blot revealed decreases in the number of ChAT-IR neurons (Figure 1A c and d) and ChAT protein levels in the dorsal medulla (Figure 1B) in 6-OHDA rats. Furthermore, the protein level of ChAT (Figure 1C) and ACh content (Figure 1D), as determined through UPLC/MS/MS analysis, were markedly reduced in the gastric corporal muscularis. These data indicate that 6-OHDA-induced degeneration of the SN altered cholinergic tone in the vagal brain-gut axis.

To confirm the occurrence of gastric muscularis inflammation in 6-OHDA rats, ELISA and HE staining were performed. The ELISA results showed that the gastric mucosa of 6-OHDA rats showed no significant change in the levels of TNF-α, IL-1β, and IL-6 (Figure 2A left). However, TNF-α and IL-1β levels in the gastric muscularis were significantly increased (Figure 2A right). HE staining showed infiltration of inflammatory cells were of small to medium size and had large nuclei with scanty cytoplasm at 4 weeks (Figures 2Bb,b’) and 6 weeks (Figures 2Bc,c’) in 6-OHDA rats. Although HE staining did not show obvious difference of infiltration of inflammatory cells between 4 and 6 weeks, the disarranged muscle fibers and increase of intercellular space were observed at 6 weeks (Figures 2Bc,c’), but not at 4 weeks (Figures 2Bb,b’). The protein expression of α7nAChR in the gastric muscularis was also significantly increased in 6-OHDA rats (Figure 2C). The results of double immunofluorescence staining showed that CD163-positive resident macrophages (Figure 2Da) were detectable in control rats, while there were few CD68-positive proinflammatory macrophages (Figure 2Dc) and α7nAChR-positive cells (Figure 2Da’). However, α7nAChR-positive cells were common, exhibited strongly immunoreactivity and were double-stained with CD163 (Figure 2Db’’) and CD68 (Figure 2Dd’’) in the gastric muscularis of 6-OHDA rats.

To investigate the anti-inflammatory effects of the α7nAChR, we isolated gastric muscularis tissues from control and 6-OHDA rats. We first evaluated the effects of the α7nAChR agonists PNU and GTS-21 and the antagonist MLA on the phosphorylated NF-κB p65 (p-p65) protein level and monocyte chemotactic protein-1 (MCP-1) mRNA level. As shown in Figure 3A, increased expression of p-p65 protein was detected in the gastric muscularis of 6-OHDA rats, which was significantly reduced after treatment with PNU or GTS-21 at a dose of 100 μmol L⁻¹. However, the α7nAChR antagonist MLA at a dose of 100 μmol L⁻¹ significantly abolished the inhibitory effects of both PNU and GTS-21 on the expression of p-p65 (Figure 3A left). The trend of MCP-1 mRNA expression in response to PNU, GTS-21, and MLA was similar to that of p-p65 protein expression (Figure 3A right). Furthermore, we measured the levels of inflammatory factors in the gastric muscularis after activating of α7nAChR with PNU in vivo. MCP-1, TNF-α, IL-1β and IL-6 levels were significantly increased in the gastric muscularis of 6-OHDA rats. After 4-weeks of PNU intraperitoneal injection, which was initiated 28 days after 6-OHDA treatment (late-PNU), there were significant reductions in all the above mentioned cytokines (Figure 3B right). Similar results were also observed in 6-OHDA rats administered PNU for 6 weeks, started from d 1 after 6-OHDA microinjection (early-PNU). The levels of MCP-1, IL-1β and IL-6 were significantly decreased, except for the TNF-α level in the early-PNU group (Figure 3B left).

In whole-mount gastric corpus preparation, CD163-positive resident macrophages were regularly distributed, but CD68-positive infiltrating macrophages were rare in the gastric muscular layer in control rats. However, the number of both CD163-and CD68-positive macrophages was significantly increased in 6-OHDA rats, and this increase was significantly reversed by early-PNU (Figure 3C) and late-PNU treatment (Figure 3D).

P-p65 is an important mediator of NF-κB activation. As shown in the left panel of Figure 4A, the 6-OHDA rats exhibited an increase in the protein expression of p-p65 in the gastric muscularis, which was significantly reduced by early-PNU treatments. Moreover, the cytosolic protein expression of inhibitory IκB (IκB), an important inhibitory protein of nuclear p-p65, was significantly decreased in the 6-OHDA rats, while early-PNU treatment completely restored the level of IκB that had been decreased (Figure 4A left). The protein expression levels of p-p65 and IκB in the late-PNU treatment group were similar to those observed in the early-PNU treatment group (Figure 4A right).

Since proinflammatory macrophages can dampen GI motility through upregulation of iNOS and COX-2 expression (Hori et al., 2001), the effects of PNU on the expression of iNOS and COX-2 were also determined. The results showed that the increased expression of iNOS in the gastric muscularis of 6-OHDA rats was prevented by early-PNU treatment (Figure 4B left). Similarly, the increase in the level of COX-2 in the 6-OHDA rats was also significantly decreased by late-PNU treatment (Figure 4B right).

The inflammatory muscular microenvironment may downregulate gastric SMC contractile protein expression in 6-OHDA rats (Xiu et al., 2020). To confirm the action of α7nAChR activation on gastric SMC contractile proteins, we examined the expression of alpha-smooth muscle actin (α-SMA) and myosin light chain 20 (MLC20) in 6-OHDA rats treated with or without PNU. We found that the levels of both α-SMA and MLC20 were significantly decreased in 6-OHDA rats, but eary-PNU treatment restored the level of both α-SMA and MLC20 (Figure 4C).

Carbachol, a cholinergic agonist, dose-dependently increased the contraction of gastric muscle strips in vitro, while this action was much weaker in the 6-OHDA rats (Figure 5A left). Early-PNU treatment significantly rescued the decreased contractions in 6-OHDA rats (Figure 5A left). Furthermore, the GE rate of solid food in 6-OHDA rats, which had been significantly reduced, was partially restored by early-PNU treatment (Figure 5A right).

Gastric motility and the GE rate of barium meal in vivo were evaluated by a strain gauge force transducer and a digital X-ray imaging system, respectively. The results showed that the maximal amplitude of the phase III-like contractions of MMCs was decreased in 6-OHDA rats, whereas late-PNU treatment slightly reversed this reduction (Figure 5B). X-ray imaging was performed for 120 min after a barium meal (Figure 5C), and the 6-OHDA rats exhibited a decreased GE rate at 30, 60, 90 and 120 min, while late-PNU treatment significantly improved the GE rate of 6-OHDA rats at 60 and 90 min (Figure 5C).

Furthermore, the results of transmission electron microscopy showed SMC hypertrophy and collagen deposition in the extracellular matrix in the gastric corpus muscularis of 6-OHDA rats (Figure 5Db). There was a decrease in the number of dense plaques in SMCs and an increase in the number of slender cytoplasmic prolongations from SMCs in 6-OHDA rats compared with control rats, in which SMCs were differentiating (Figures 5Da,b). This suggests that the SMCs in 6-OHDA rats might undergo dedifferentiation. Furthermore, the alterations in ultrastructural features were greatly ameliorated by late-PNU treatment (Figure 5Dc).

It has been reported that nicotine can alleviate toxin-induced nigrostriatal damage by activating α7nAChR in the brain (Baladi et al., 2016). To determine whether the PNU treatment-induced improvement in gastric motility in 6-OHDA rats was associated with the neuroprotective effect of α7nAChR activation in the brain, we examined tyrosine hydroxylase (TH) expression in the SN and ChAT expression in the dorsal medulla. Unlike central application of nicotine, both early-PNU treatment and late-PNU treatment via intraperitoneal injection did not prevent the decreases in TH expression in the SN and ChAT expression in the dorsal medulla (Figure 6).