Activation of α7nAChR Preserves Intestinal Barrier Integrity and Ameliorates Cholestasis Liver Fibrosis in Mice by Enhancing the HO-1 / STAT3 Signaling to Inhibit NF-κB Activation

Background: Activation of alpha-7 nicotinic acetylcholine receptor (α7nAChR) can inhibit the systemic inammatory response and preserve intestinal barrier integrity. This study aimed at elucidating the molecular mechanisms by which α7nAChR activation could inhibit intestinal barrier injury and cholestatic liver brosis in mice induced by bile duct ligation (BDL). Methods: The intestine-specic HO-1 knockout VillinCreHmox1 -/- and control Hmox1 oxp/oxp C57BL/6 mice were subjected to BDL. The therapeutic effects of GST-21, a specic ligand for α7nAChR, on systemic and intestinal inammation, intestinal barrier integrity, liver brosis and injury, HO-1 expression, STAT3, AKT and NF-kBp65 activation were examined in these mice and intestinal epithelial cells co-cultured with macrophages. Results: Compared with BDL mice, α7nAChR activation by GST-21 decreased intestinal and liver injury and brosis in BDL mice, accompanied by reducing serum cytokine responses. In addition, activation of α7nAChR preserved the tight junction protein expression and intestinal epithelial cell barrier integrity in BDL mice and epithelial cells co-cultured with macrophages. The therapeutic effects of α7nAChR activation were mediated by enhancing HO-1 expression, STAT3 phosphorylation, and reducing the NF-kBp65 activation in intestinal tissues and epithelial cells co-cultured with macrophages. Finally, activation of α7nAChR induced HO-1 expression and STAT3 phosphorylation in an interdependent manner, independent of the PI3K/AKT signaling. Conclusion: Activation of α7nAChR enhanced HO-1 expression and STAT3 signaling to inhibit NF-κB activation, preserving the intestinal barrier integrity, and reducing inammation and liver brosis in cholestasis mice. Therefore, targeting α7nAChR may be a promising interventional strategy for primary biliary cholangitis.


Introduction
Primary biliary cholangitis (PBC) is a chronic autoimmune liver disease characterized by non-suppurative small duct and portal in ammation and interface hepatitis [1]. PBC affects the life quality of patients and its progression eventually leads to biliary cirrhosis and liver failure. Unfortunately, there are no speci c and effective therapies for PBC. Pathophysiologically, PBC has been attributed to intestinal barrier injury [2], which is evidenced by commonly clonal T cells homing in the intestine and liver of PBC patients, suggesting that unknown antigens in the intestine stimulate immune responses, leading to the development of PBC [3]. The loss of intestinal barrier integrity also allows commensal organisms and foodborne antigens to enter the liver, exacerbating hepatic in ammation and PBC [4]. Hence, the functional de ciency in the gut-liver axis is a major contributor to PBC development. However, it is unclear whether preservation of intestinal barrier integrity can be a valuable strategy for the intervention of PBC.
Nicotinic acetylcholine receptors (nAChRs) can be activated by its natural ligand of the neurotransmitter acetylcholine (Ach) [5]. Nicotinic acetylcholine receptor α7 (α7nAChR) is crucial for long-term memory and immune regulation [6]. The α7nAChR is mainly expressed in the brain and immune cells, such as macrophages and activation of α7nAChR by Ach in tissue macrophages and other immune cells can inhibit their production of pro-in ammatory cytokines, such as IL-1β, IL-6 and TNFα, known as the "cholinergic anti-in ammatory pathway" (CAP) [7,8]. Actually, activation of the CAP can alleviate acute and chronic in ammatory diseases, such as endotoxemia and in ammatory bowel disease in rodents [9,10]. In addition, activation of α7nAChR by neuronal and non-neuronal Ach can ameliorate intestinal, cardiovascular and lung in ammation, neurodegeneration, sepsis and arthritis [11][12][13][14]. In contrast, chronic nicotine exposure can stimulate biliary cholangiocyte proliferation and induce pro-brotic protein expression by enhancing the Calcium/ERK signaling in healthy adult rats [15]. However, little is known on whether and how activation of the α7nAChR can affect the pathogenic process of bile duct ligation (BDL)-induced in ammation and brosis.
The PI3K/AKT and JAK2/STAT3 pathways are important for the anti-in ammatory effects of α7nAChR activation [16,17]. STAT3 knockout mice can enhance the in ammatory activity of macrophages and neutrophils [18]. Activation of α7nAChR by Ach can enhance the STAT3 signaling that positively feedback up-regulates α7nAChR expression to amplify the anti-in ammatory effect of Ach/ α7nAChR. Furthermore, the activated STAT3 can also enhance nuclear factor erythrocyte-associated factor 2 (NRF2) activation and promote the expression of Heme oxygenase 1(HO-1) and other antioxidant enzymes [19,20]. HO-1 and its metabolic byproducts, such as carbon monoxide (CO), biliverdin, and ferrous iron, have potent antioxidant and anti-in ammatory activities [20]. Activation of the CAP also increases HO-1 expression and enhances the α7nAChR activation-related inhibition of nerve injury, myocardial ischemia-reperfusion injury, and intestinal barrier injury [12,21,22]. However, the exact mechanisms by which activation of α7nAChR inhibits the production of pro-in ammatory factors in gastrointestinal immune cells has not been clari ed although it can alleviate intestinal in ammation and maintain the integrity of intestinal epithelium [23,24]. Up-regulated HO-1 expression can ameliorate endotoxemic in ammatory responses in macrophages [25] and preserve the intestinal barrier integrity in cholestasis rats [26]. Accordingly, we hypothesize that activation of α7nAChR can up-regulate anti-in ammatory HO-1 expression in intestinal epithelial cells to preserve the intestinal barrier integrity in early cholestatic liver injury by modulating the PI3K/AKT and JAK/STAT3 signaling.
In the present study, we tested the therapeutic effect of GTS-21, a speci c α7nAChR agonist, on cholestatic liver injury, intestinal barrier integrity, in ammatory cytokine production in a mouse model of BDL-induced cholestatic liver injury and investigated whether activation of α7nAChR modulated the intestinal epithelial barrier integrity by enhancing HO-1 expression and the PI3K/AKT/STAT3 signaling.

Animals
The experimental protocols were approved by the Animal Care and Use Committee of Dalian Medical University (Liaoning Province, China; approval ID AEE19011). The VillinCreHmox1 −/− (the intestinespeci c HO-1 knockout) and control Hmox1 oxp/ oxp C57BL/6 mice were obtained from Beijing Viewsolid Biotechnology (Beijing, China). The mice were housed in a speci c pathogen-free room with a consistent temperature (20-22°C) and a cycle of 12/12-h light/dark and allowed to free access to normal rodent chow and water.

Animal model, grouping and treatment
Male mice at 10 weeks of age were injected intraperitoneally with 40 mg/kg pentobarbital (Sigma, USA) and subjected to sham or BDL procedure [32]. Brie y, their cystic and common bile ducts of individual BDL mice were exposed by an incision and ligated using 8 − 0 nylon suture. The sham group of mice received a sham surgery without ligation. After their abdomen was closed, the mice were kept on a heatpad until their awake. The BDL mice were randomized and injected intraperitoneally with vehicle (BDL group) or GTS-21(8 mg/kg) and saline, Stattic (25 mg/kg) or LY294002 (0.5 mg/kg, MCE, USA) every other day for six treatments. The VillinCreHmox1 oxp/ oxp mice received BDL and GTS-21 injection.
Enzyme-linked immunosorbent assay (ELISA) The levels of serum IL-1β, IL-6, IL-10, TNFα, alanine aminotransferase (ALT) and aspartate transaminase (AST) in individual mice were quanti ed in triplicate by ELISA using speci c kits (Elabscience, China) according to the manufacturer's instructions.

Histological examination
Liver and intestinal tissues were xed in 4% paraformaldehyde and para n-embedded. The liver and intestinal tissue sections (4 µm) were routine-stained with hematoxylin and eosin (H&E). The degrees of liver brosis were scored and the levels of intestinal mucosal injuries were scored using the Chiu's scoring system in a blinded manner [26,36]. The liver sections were stained with Picrosirius-red using the Siriusred staining kit (Solarbio, Beijing, China). The amount of collagen deposition was measured using a computerized image-analysis system and expressed as the percentages of the positively stained area in total area.

Immunohistochemistry and Immuno uorescence
The levels of targeted protein expression in intestinal and liver tissue samples were examined by immunohistochemistry and immuno uorescence. Brie y, intestinal and liver tissue sections (4 µm) were dewaxed, rehydrated and treated with 3% H 2 O 2, followed by blocking with 3% bovine serum albumin (BSA, Sigma-Aldrich). The sections were probed overnight with primary antibodies at 4°C, and reacted with horseradish peroxidase (HRP)-conjugated secondary antibodies, followed by visualizing with a DAB substrate kit (Abcam, ab64238). For Immuno uorescence, the tissue sections and cells were permeabilized in 0.3% Triton X 100 and blocked with 5% BSA. After probing with primary antibodies overnight at 4°C, the bound antibodies were detected with uorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG (Santa Cruz Biotech, CA). The negative controls were probed with the vehicle solution.

Measurement of cell and intestinal barrier permeability
Caco-2 cells (3×10 5 cells/well) were cultured in the upper chamber of the transwell plates for 10 days to form a monolayer. RAW264.7 cells (2×10 6 cells/well) were cultured in the bottom chamber. Cells are treated as described above. The cell culture was replaced with serum free medium and incubated for another 30 min, and the transepithelial electrical resistance (TEER, Ω/cm 2 ) was evaluated using a resistance measuring instrument (Millipore, USA). Subsequently, the epithelial cells in the upper chamber were treated with 100 µg FITC-dextran − 4 (FD-4, Sigma) and incubated in 37°C for 30 min. After that, 100 µl of medium from the bottom chamber were measured for the uorescence contents using the BioTek Cytation3 with an excitation of 480 nm and emission of 520 nm respectively. In addition, individual mice were orally administrated by gavage with 12 mg FD-4 in 150 µl of sterile water at 13 days post BDL. The next day, their venous blood samples (100 µL each) were collected and the levels of FITC uorescence in individual samples were measured using the BioTek Cytation3.

Real-time quantitative PCR (RT-qPCR)
Total RNA was extracted from the cell and tissue samples using TRIzol (Transgene Biotech) and reversely transcribed into cDNA using the Prime Script™ RT Reagent Kit (Takara). The relative levels of interesting gene mRNA transcripts to the control β-actin were quanti ed by RT-qPCR in the ABI7500 RT-PCR detection system (ABI System) using SYBR® Green RT-PCR Master Mixes (ABI System) and speci c primers (Supplemented Table 1). The PCR reactions were performed in duplicate at 95°C for 10 min and subjected to 40 cycles of 95°C for 15 s and 60°C for 1 min. The data were analyzed by 2 −ΔΔCt .

Western blot analysis
Individual tissue samples were homogenized and cultured cell samples were lysed in lysis buffer. After quanti cation of protein concentrations using the BCA kit (Beyotime, China), the lysate samples (30 µg/lane) were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 12% gels and transferred onto polyvinylidene di uoride (PVDF) membranes. The membranes were blocked and probed with primary rabbit antibodies (Supplemented Table 2) at 4°C overnight and reacted with HRP-conjugated secondary antibodies (Abclonal, China, 1:2000 dilution), followed by visualizing using a chemiluminescent kit (Advasta). The relative levels of targeted proteins to the control β-actin were quanti ed by densitometry analysis using ImageJ software.

Statistical analysis
Data are presented as the mean ± SEM of each group from three or more independent experiments. The difference among groups was analyzed by ANOVA and post hoc Tukey test or Student's T-test using GraphPad Prism version 7 (GraphPad Software, La Jolla, CA). A P-value of < 0.05 was considered statistically signi cant.

Results
3.1 Activation of α7nAChR attenuates in ammation and liver functional impairment in BDL mice and modulates the LPS-induced cytokine transcription in macrophages in vitro.
To explore the effect of α7nAChR activation on in ammation and liver injury, we established in a mouse model of cholestatic liver injury by BDL and found that BDL signi cantly decreased body weights at 14 days post-BDL regardless of treatment and HO-1 KO (p < 0.001, for all, Fig. 1a). In comparison with the sham control, BDL did not signi cantly alter the relative levels of α7nAChR expression in intestinal tissues, but treatment with GST-21 signi cantly up-regulated α7nAChR protein expression in intestinal tissues of different groups of BDL mice, regardless of intestinal HO-1 KO, STAT3 or AKT inhibition (p < 0.05, Fig. S3a-b). Furthermore, BDL signi cantly increased the levels of serum IL-1β, IL-6 and TNFα in mice (p < 0.01, p < 0.001), which were signi cantly mitigated in GTS-21 treated mice at 14 days post-BDL (p < 0.05, p < 0.01), but not in the mice received with Stattic treatment and the Villin-CreHmox1-/-mice ( Fig. 1b-d). A similar pattern of serum ALT and AST levels was detected in the different groups of mice at 14 days post-BDL, except for the insigni cant difference in the levels of AST between the GST-21 treated Hmox1 oxp/ oxp and Villin-CreHmox1-/-mice ( Fig. 1e-f). To further understand the effect of α7nAChR activation, we tested how treatment GST-21 could modulate the LPS-stimulated cytokine production in macrophages in vitro. We found that treatment with GST-21 alone did not signi cantly change the relative levels of IL-1β, IL-6, TNFα, IL-22 and IL-10 mRNA transcripts in both U939 and RAW264.7 macrophages, but GST-21 treatment signi cantly mitigated the LPS-increased IL-1β, IL-6 and TNFα mRNA transcripts and further elevated IL-22 and IL-10 mRNA transcripts in these macrophages in vitro (p < 0.001 for all, Fig. 1g-k and l-p). These data indicated that activation of α7nAChR signi cantly alleviated the BDLinduced in ammatory cytokine responses and liver injury in mice, dependent on intestinal HO-1 expression and STAT3 activity, but not PI3K/AKT activity and modulated the LPS-induced cytokine gene transcription in macrophages in vitro.
3.2 Activation of α7nAChR ameliorates intestinal mucosal damages and liver brosis in BDL mice, dependent on intestinal HO-1 and STAT3 activity.
To further explore the protective effect of α7nAChR activation on intestinal mucosal damages and liver brosis, we histologically evaluated intestinal mucosal damages and liver brosis in different groups of mice. Following H&E staining, we observed villous necrosis, in ammatory in ltrates and hemorrhage in the intestinal mucosa of the BDL group were much severer than those in the sham and GST-21 treated mice (Fig. 2a). Similarly, there are many necrotic hepatocytes and in ammatory in ltrates as well as the enlargement of brous tissue surrounding the portal area, a brous septum formation in the liver of BDL mice. Further Sirius Red staining revealed varying levels of collagen deposition in the liver, particularly in the BDL mice without GST-21 treatment. Quantitative analysis indicated Chiu's histological scores of intestinal mucosal damages in individual mice (Table S3) and that Chiu's histological scores in the BDL group were signi cantly higher than that in the sham group, but were signi cantly reduced in the GST-21 treated mice, except for those received Stattic or HO-1 −/− mice (Fig. 2b). A similar pattern of liver brosis grades (detailed pathological grades in Table S4) and collagen deposition contents was detected in the livers of different groups of mice, except for the insigni cant difference in the liver brosis grades between wild-type and HO-1 −/− mice as well as the insigni cant difference in the collagen deposition contents between the GST-21 treated mice and LY294002-treated mice (Fig. 2c-f). These results indicated that activation of α7nAChR ameliorated intestinal mucosal damages and liver brosis in BDL mice, variably dependent on intestinal HO-1 expression, STAT3 and/or PI3K/AKT activities.

Activation of α7nAChR preserves the expression of tight junction proteins and intestinal epithelial barrier integrity in vivo and in vitro
Given that tight junction proteins are crucial for intestinal barrier integrity, we investigated the levels of ZO-1, Occludin and Claudin-1 expression in intestinal tissues from different groups of mice by Western blot and RT-qPCR assays. We found that compared to the sham group, BDL signi cantly reduced the relative levels of ZO-1, Occludin and Claudin-1 protein expression in intestinal tissues of mice, which was partially rescued by GST-21 treatment (Fig. 3a-d). However, the preservation of tight junction protein expression by GST-21 was attenuated in intestinal HO-1 −/− mice and by treatment with Stattic in mice. A similar pattern of ZO-1, Occludin and Claudin-1 mRNA transcripts was detected in the intestinal tissues of different groups of mice, except for signi cantly reduced levels of ZO-1 and Claudin-1 mRNA transcripts in the LY294002-treated mice ( Fig. 3e-g). Functionally, we found that treatment with GST-21 signi cantly mitigated the BDL-increased serum levels of FITC-dextran, and the therapeutic effect of GST-21 on preserving intestinal barrier integrity was partially reduced in the Stattie-treated mice or intestinal HO-1-/mice (Fig. 3h). These demonstrated that activation of α7nAChR mitigated the BDL-increased intestinal barrier permeability by preserving tight junction protein expression in the intestinal tissues of mice.
Next, we validated the results in a cellular model. Given that the α7nAChR is mainly expressed in immune cells, such as macrophages in the periphery, we measured how GST-21 treatment modulated tight junction protein expression and monolayer epithelial permeability in a co-culture system. Following culture alone or co-culture with macrophages, we found that treatment with GST-21 did not signi cantly change the LPS-decreased ZO-1, Occludin and Claudin-1 expression in cultured intestinal epithelial Caco2 or HT29 cells alone, but GST-21 treatment completely blocked the LPS-decreased ZO-1, Occludin and Claudin-1 expression when these epithelial cells co-cultured with macrophages ( Fig. 3i-l and m-p).
Moreover, a similar pattern of TEER was detected in the different groups of cells, indicating that GST-21 treatment restored the LPS-decreased TEER of monolayer epithelial cells when co-culture with macrophages (Fig. 3q). Additionally, GST-21 treatment also signi cantly reduced the LPS-increased FITCdextran contents in the supernatants of co-cultured cells (Fig. 3r-s), suggesting that activation of α7nAChR in macrophages by GST21 preserved the intestinal epithelial monolayer integrity in the coculture system. Immuno uorescence revealed that compared with the control Caco2 cells alone, LPS treatment decreased ZO-1 expression, which was not affected by GST-21 (Fig. 3t). In contrast, GST-21 treatment signi cantly mitigated the LPS-decreased ZO-1 expression and cytoskeletal structural damages in the Caco2 cells when co-cultured with macrophages in vitro. Together, these data demonstrated that activation of α7nAChR in immune cells, such as macrophages, signi cantly preserved tight junction protein expression and intestinal epithelium integrity in vivo, dependent on HO-1 expression and STAT3 activity and in the vitro co-culture system.
3.4 Activation of α7nAChR in macrophages enhances the LPS-stimulated HO-1 and IL-10 expression in intestinal epithelial cells and tissues.
Given that HO-1 and IL-10 are important for intestinal barrier function we tested whether treatment with GST-21 to activate α7nAChR in macrophages could modulate HO-1 and IL-10 expression in the monolayer of intestinal epithelial cells [27]. We found that treatment with LPS, but not GST-21, obviously increased the levels of HO-1 expression in Caco2 and HT29 cells while treatment with GST-21 failed to alter HO-1 expression in the LPS-treated cells (Fig. 4a-c and d-e). In contrast, treatment with GST-21 further enhanced the LPS-increased HO-1 and IL-10 expression in both Caco2 and HT29 cells when co-cultured with macrophages ( Fig. 4a-f). Furthermore, the enhanced HO-1 expression by GST-21 was signi cantly reduced by HO-1 silencing or treatment with Stattic, but not with LY294002 in both Caco2 when cocultured with macrophages ( Fig. 4g-i). Similarly, BDL signi cantly increased the levels of HO-1 expression in intestinal tissues while GST-21 treatment further enhanced the BDL-increased HO-1 expression in intestinal tissues, which was failed in HO-1 −/− mice, and signi cantly mitigated in the Stattic-treated mice, but not in the LY294002-treated mice (Fig. 4j-l). Moreover, GST-21 treatment also further increased the LPS-enhanced IL-10 expression in Caco2 cells when co-cultured with macrophages, which was no affected by HO-1 silencing and Ly294002 treatment (Fig. 4m). In contrast, treatment with Stattic to inhibit STAT3 activity signi cantly attenuated the GST-21-enhanced HO-1 expression in Caco2 cells that were cocultured with macrophages. Additionally, GST-21 treatment enhanced the BDL-up-regulated IL-10 expression in intestinal tissues, which was not affected by HO-1 knockout, but signi cantly mitigated by treatment with Stattic or ly294002 in intestinal tissues of mice (Fig. 4n) These results indicated that activation of α7nAChR in macrophages enhanced the LPS or BDL-up-regulated HO-1 and IL-10 expression in intestinal epithelial cells, which may be partially dependent on the STAT3 signaling.
3.5 Activation of α7nAChR enhances the HO-1/STAT3 and PI3K/AKT signaling to attenuate NF-κB activation in vitro and in vivo.
Finally, we explored how activation of α7nAChR enhanced HO-1 expression and modulated the STAT3, PI3K/AKT and NF-kB signaling in intestinal epithelial cells and intestinal tissues of BDL mice. After knockdown of HO-1 expression in Caco2 cells, we found that compared with the control cells, LPS upregulated HO-1 expression and treatment with GST-21 further signi cantly elevated the relative levels of HO-1 expression in Caco2 cells when co-cultured with macrophages, which were dramatically reduced by HO-1 silencing and partially mitigated by Stattic treatment (Fig. 5a-e). Furthermore, LPS treatment also enhanced STAT3, AKT and NF-kBp65 phosphorylation and treatment with GST-21 signi cantly increased STAT3 and AKT phosphorylation, but decreased NF-kBp65 phosphorylation in Caco2 cells. The therapeutic effect of GST-21 on STAT3 and AKT phosphorylation was abrogated by HO-1 silencing, Stattic or LY294002 treatment in Caco2 cells. Moreover, HO-1 silencing, Stattic and LY294002 treatment also abrogated the GST-21-decreased NF-kBp65 phosphorylation in LPS-treated Caco2 cells. These data suggest that activation α7nAChR would enhance HO-1 expression by enhancing the STAT3 and PI3K/AKT signaling to attenuate the NF-kB activation in intestinal epithelial cells when co-cultured with macrophages, preserving the intestinal barrier integrity. Similarly, GST-21 treatment further enhanced the BDL-up-regulated HO-1 expression, STAT3 and AKT phosphorylation but dramatically mitigated the BDLinduced NF-kBp65 phosphorylation in intestinal tissues of the different groups of mice (Fig. 5f-j). In addition, HO-1 knockout, Stattic or LY294002 treatment signi cantly mitigated or abrogated the GST-21enhanced HO-1 expression, STAT3 and AKT activation, but partially rescued NF-kBp65 activation in intestinal tissues of different groups of mice, except that LY294002 treatment failed to signi cantly modulate the GST-21-enhanced HO-1 expression. These data further indicated that activation of α7nAChR enhanced HO-1 expression by enhancing the STAT3 and PI3K/AKT signaling to attenuate the NF-kB activation in intestinal tissues, contributing to inhibition of cholestatic liver brosis in mice.
Immunohistochemistry revealed that compared with the control, the anti-p-STAT3 signals increased in intestinal tissues of the BDL group and were further stronger in the GST-21 or LY294002 group (Fig. 6a).
In contrast, the anti-pSTAT3 signals were much weaker in the HO-1 −/− or Stattic group than in the GST-21 group. A similar pattern of anti-HO-1 staining signals was observed in intestinal tissues of different groups of mice, consistent with the results from the Western blot assay (Fig. 6b). In contrast, the strongest anti-pNF-kBp65 signals were observed in intestinal tissues of the BDL group, which were dramatically reduced in the GST-21 group (Fig. 6c). The anti-pNF-kBp65 signals were completely or partially rescued in the HO-1 −/− , Stattic or LY294002 group of mice. These three lines of evidence demonstrated that activation of α7nAChR up-regulated HO-1 expression by enhancing the STAT3 signaling to attenuate the NF-kB activation in intestinal tissues, preserving the intestinal barrier integrity and inhibiting cholestatic liver brosis in mice.

Discussion
High concentrations of bile acids, especially more toxic hydrophobic bile acids, will damage cholangiocytes and hepatocytes, leading to cholestasis, which progressively causes liver brosis and even cirrhosis [37]. The intestinal barrier damages have been associated with the development of cholestatic liver diseases. However, it is still unclear whether intestinal barrier dysfunction is a cause or consequence of cholestatic liver disease, such as PBC. Nonetheless, impairment of intestinal barrier integrity is often observed in patients with many diseases [38], and intestinal barrier integrity has been thought to be a therapeutic target.
In this study, we employed a mouse model of BDL-induced cholestatic liver brosis and an intestinal epithelium-macrophage co-culture system to investigate the therapeutic effects of a7nAChR activation on the intestinal barrier integrity and to explore its possible mechanism. The available data highlighted that GTS-21 treatment to activate a7nAChR attenuated systemic in ammation, intestinal and liver injury and brosis by preserving the intestinal barrier integrity in BDL mice. Evidently, GST-21 treatment increased serum IL-10, but decreased serum TNF-a, IL-6 and IL-1β, ameliorating the imbalance of pro-in ammatory and anti-in ammatory cytokine responses in BDL mice. Furthermore, activation of a7nAChR preserved the expression of ZO-1, Occludin and Claudin-1 tight junction proteins in intestinal tissues to reduce the intestinal barrier permeability in BDL mice. Mechanistically, we found that activation of a7nAChR enhanced HO-1 expression and STAT3 activation, but attenuated the NF-κB signaling in intestinal tissues of BDL mice. Similar data were obtained from in vitro epithelial cells co-cultured with macrophages. More importantly, the therapeutic effects of α7nAChR activation were abrogated by HO-1 de ciency and signi cantly mitigated by inhibition of STAT3 activity in intestinal epithelial cells in vivo and in vitro. These novel ndings may uncover the therapeutic mechanisms underlying the action of α7nAChR activation in preserving the intestinal barrier integrity and inhibiting cholestatic liver brosis. Therefore, α7nAChR may be a therapeutic target for the intervention of cholestatic liver brosis, such as PBC.
A previous study has shown that in ammation, but not the direct toxicity of bile acids is a dominant factor of cholestatic liver brosis [39]. In the present study, we found that BDL signi cantly increased serum levels of TNFα, IL-1β, and IL-6 in mice, which may directly damage cholangiocytes and reduce the expression of bile transporters in cholangiocytes, increasing the accumulation of bile acids in the liver and deteriorating liver injury and brosis [40]. Actually, a reduction in pro-in ammatory cytokine production and inhibiting the in ammatory pathway can drastically mitigate the liver injury, retarding the process of cirrhosis in BDL rodents [41]. We found that activation of a7nAChR signi cantly decreased serum levels of TNFα, IL-1β and IL-6 in BDL mice, which were associated with liver injury in BDL mice. Our data support the notion that in ammation is crucial for the development and progression of cholestatic liver injury. Conceivably, control of in ammation may be a valuable strategy for the intervention of cholestatic liver diseases, such as PBC.
The dysfunctional gut-liver axis is closely related to the pathogenesis of chronic liver disease because the liver is the most vulnerable organ for gut microbial infection [42]. Bacteria and their metabolites can through the disrupted intestinal barrier enter the bile duct and liver to enhance in ammation that deteriorates cholestatic liver injury [43]. It is well known that ZO-1, Occludin, and claudin-1 and others are important for intestinal barrier integrity [44]. We found that BDL dramatically decreased these tight junction proteins expression in intestinal tissues, accompanied by increased intestinal barrier permeability in mice, consistent with the ndings from other models [45]. Similarly, LPS treatment also damaged the monolayer of the intestinal epithelial cell barrier and increased its permeability in a coculture system, similar to a previous observation [46]. In contrast, activation of α7nAchR preserved the tight junction protein expression and mitigated the BDL or LPS-increased intestinal barrier permeability in vivo and in vitro. These ndings support the concepts that disrupted intestinal barrier integrity and bacterial infection participate in the pathogenesis of cholestatic liver injury. Preserving and restoring intestinal barrier integrity may inhibit the progression of cholestatic liver injury. We are interested in further investigating how BDL or cholestasis disrupts the intestinal barrier integrity.
Activation of the CAP, particularly for the α7nAChR, can inhibit the in ammatory pathogenesis of several diseases, including acute lung in ammation and acute pancreatitis [47,48]. In this study, we found that activation of α7nAChR not only decreased pro-in ammatory IL-1β, IL6 and TNFα production but also enhanced IL-10 and IL-22 production in macrophages under a LPS stimulation. IL-10 is a potent antiin ammatory cytokine and can inhibit different types of immune and in ammatory responses. IL-22 can be secreted by different types of cells, including macrophages and stimulate other types of cells to produce antimicrobial peptides and proteins to defense bacterial infection, particularly in the intestine [49].
Functionally, IL-22 can through its receptors of IL-10R2/IL-22R1 promote epithelial cell proliferation, survival and repair to enhance the intestinal barrier integrity although it can synergistically work with IL-17A or TNFα to enhance pro-in ammatory responses in some pathological conditions [50,51]. The decreased pro-in ammatory cytokines and increased IL-10 and IL-22 by GST-21 treatment clearly demonstrate that α7nAChR activation can alleviate the imbalance of pro-in ammatory and antiin ammatory responses, particularly in an intestinal bacterial infection, to preserve the intestinal barrier integrity. Therefore, the intestinal-intrinsic α7nAChR signaling may serve as an endogenous protective modality for intestinal barrier integrity to inhibit cholestatic liver injury.
HO-1, an anti-in ammatory and anti-oxidant protein, can inhibit oxidative stress-related in ammation to protect from different types of damage during the pathogenic process of intestinal diseases, liver injury and pulmonary in ammation [52,53]. Actually, elevated HO-1 expression can increase the expression of tight junction proteins to preserve the intestinal barrier integrity in mice with cholestatic liver injury and CCl4 injected mice [26,52]. We found that α7nAChR activation up-regulated HO-1 expression in intestinal tissues in BDL mice and epithelial cells co-cultured with macrophages, consistent with previous observations [54]. Moreover, α7nAChR activation increased IL-10 expression in macrophages, independent of HO-1 expression. It is possible that the enhanced IL-10 expression in macrophages by α7nAChR activation may promote HO-1 expression in Caco2 cells in the co-culture system, consistent with a previous study [55]. Notably, α7nAChR activation increased IL-10 expression, dependent on STAT3 activity, in intestinal epithelial cells. Given that GST-21 treatment signi cantly up-regulated IL-22 expression the enhanced IL-10 expression by α7nAChR activation in intestinal epithelial cells is likely mediated by IL-22related STAT3 activation or IL-10 autocrine-enhancement. It was unlike from IL-6-mediated SAT3 activation because α7nAChR activation dramatically down-regulated the LPS-induced IL-6 production in epithelial cells and macrophages. Our data disagree with the fact that inhibition of STAT3 phosphorylation can reduce in ammatory responses [56,57]. The discrepancy may stem from varying experimental conditions between us and those of others. Nevertheless, the increased anti-in ammatory cytokines and HO-1 expression, together with decreased pro-in ammatory cytokines, preserved the intestinal barrier integrity and inhibited cholestatic liver injury.
Our previous RNA-seq data revealed the differentially expressed genes in LPS-induced Caco-2 cells were enriched in the STAT3 signaling ( Figure S4). In the present study, activation of α7nAChR enhanced STAT3 and AKT phosphorylation but attenuated NF-kBp65 activation in intestinal tissues and epithelial cells in a HO-1 dependent manner because HO-1 silencing or HO-1 knockout dramatically mitigated or abrogated the GST-21-enhanced STAT3 and AKT phosphorylation and GST-21-21-decreased NF-kBp65 activation in intestinal tissues and epithelial cells, consistent with a previous observation [58]. The HO-1 dependence may stem from its antioxidant activity to attenuate the NF-kB activation or enhanced IL-10 and IL-22 expression to promote STAT3 activation. In addition, activated STAT3 can promote ZO-1 and Occludin expression and reduce the intestinal barrier permeability [59,60]. Thus, our ndings extended previous observations and may provide new insights into the regulation of a7nAChR on HO-1 expression and the STAT3 signaling in intestinal tissues during the pathogenic process of cholestatic liver injury.
It was notable that inhibition of the PI3K/AKT signaling did not signi cantly modulate the HO-1 expression, but did signi cantly decreased the GST-21-enhanced STAT3 phosphorylation, and increased the NF-kBp65 activation in the LPS-treated intestinal epithelial cells. In contrast, inhibition of the PI3K/AKT signaling did slightly reduce HO-1 expression and did not affect STAT3 phosphorylation although it did signi cantly enhance the NF-kBp65 activation in intestinal tissues of BDL mice in a HO-1 dependent manner. The difference may stem from a strong stimulation in vitro experimental system, which caused the minor cellular injury. Apparently, the activated STAT3 signaling may cross-talk with the PI3K/AKT signaling to promote the survival of intestinal epithelial cells. Therefore, the HO-1/STAT3 signaling may be critical for preserving the intestinal barrier integrity while the PI3K/AKT may have moderate activity during the pathogenic process of cholestatic liver injury.
We recognized our studies had limitations. First, GTS-21, one of the most speci c α7nAChR agonists, may be a concern for its clinical application so that discovery of newly safe agonists is urgently needed.

Conclusions
Our data highlighted that α7nAChR activation inhibited intestinal and systemic in ammation, intestinal and liver injury to preserve the intestinal barrier integrity in BDL mice and intestinal epithelial cells cocultured with macrophages. Mechanistically, α7nAChR activation enhanced the HO-1 expression, STAT3 and AKT phosphorylation, but decreased NF-kBp65 activation in intestinal tissues of BDL mice and LPStreated epithelial cells. The enhanced HO-1 expression and STAT3 activation were interdependent in intestinal tissues of BDL mice and LPS-treated epithelial cells. Accordingly, the HO-1/SATA3 signaling may be crucial for α7nAChR activation to preserve intestinal barrier integrity and inhibit the pathogenic process of cholestatic liver injury in mice. Therefore, our ndings may provide new mechanisms underlying the action of α7nAChR activation in inhibiting cholestatic liver injury and uncover that α7nAChR may be a therapeutic target for the intervention of PBC. and Yanling Jin was responsible for the H&E staining experiment. Zhijun Duan reviewed the nal manuscript. All authors read and approved the nal manuscript.

Funding Sources
No funding was received for this study.

Availability of data and materials
Not applicable Declarations Ethics approval and consent to participate The study was approved by the Ethics Committee of Dalian Medical University, Huazhong University of Science and Technology (approval ID AEE19011).

Consent for publication
Not applicable.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.