Engineered Butyrate-producing Bacillus Subtilis Alleviated Ethanol-induced Intestinal and Liver Damage in Mice

Ethanol-induced intestinal and liver injury are closely associated with intestinal dysbiosis and altered short-chain fatty acid (SCFA) metabolites which is crucial for intestinal health. Bacillus subtilis (BS) strains with biotherapeutic potential can benet the host through maintaining intestinal homeostasis and regulating systematic immunity via producing small molecules, although these molecules do not include butyrate. To combine the advantages of butyrate and BS, we evaluated the bioactivity of an engineered butyrate-producing Bacillus subtilis (BPBS) strain against ethanol exposure in a chronic-binge ethanol feeding mouse model. Our ndings suggested that prophylactic BPBS supplementation restored eubiosis of the gut microbiota and intestinal barrier function, which obviously reduced bacterial translocation of microbial products especially lipopolysaccharide (LPS) to the circulatory system. Additionally, the decrease of serum LPS is responsible for the relief of hepatic inammation via the Toll-like receptor 4 (TLR4) pathway, resulting in improved hepatic structure and function. Collectively, these results demonstrated that engineered BPBS intervention imparted novel hepatoprotective functions by improving intestinal barrier function and reducing systematic inammation under ethanol exposure, as well as paving the way for further exploration of engineered probiotics in improving human health care.


Introduction
Currently, amounting evidence indicate that excessive alcohol consumption contributes to gastrointestinal and liver diseases, which are closely associated with the alterations of gut microbiota abundance and diversity (Boyle et Kimura et al. 2020). Notably, butyrate plays a signi cant role in maintaining intestinal health, and its major mode of action is via mediation of signaling pathways involving nuclear factor kappa B (NF-κB) and inhibition of histone deacetylase (Inan et al. 2000;Hodin 2000). In addition, butyrate as an energy substance provides energy for colonic epithelial cells and even participates in maintaining intestinal homeostasis by facilitating tight junction assembly Page 3/16 and mucus secretion (Guilloteau et al. 2010; J. Chen and Vitetta 2018). Furthermore, butyrate is involved in energy metabolism by stimulating the secretion of insulin via g-protein-coupled receptor 43 (GPR43). In this context, butyrate as a feed additive has been developed and widely used to improve mammal and poultry health Piazzon et al. 2017;Bedford and Gong 2018).
With a deepening understanding of the relationship between the gut microbiota and human health, the gut microbiota has been recognized as a therapeutic target in some diseases, including obesity, enteric disease, ALD, and so on (Sonnenburg and  ingenious method of probiotic bioengineering has been adopted to precisely modulate and restore the effects of gut dysbiosis, such as the designed pyrroloquinoline quinone-secreting probiotic Escherichia coli Nissle 1917 and engineered IL-22-producing Lactobacillus reuteri (Jiang et al. 2017;Singh et al. 2014), but few studies have involved engineered butyrate-producing B. subtilis (BPBS), which has the potential for outstanding bioactivity in maintaining intestinal homeostasis.
In the present study, we investigated the bene cial effects of engineered BPBS that integrated the advantages of the B. subtilis and the central biological effect of butyrate in protecting mice from ethanol exposure, and elaborated the mechanisms of engineered BPBS supplementation in mitigating ethanolinduced intestinal and liver injury by restoring intestinal homeostasis and reducing hepatic in ammation with the assistance of butyrate.

Bacterial strains and cultures
The Bacillus strains used in the study are shown in Table 1. Wild-type B. subtilis sck6 (BS) and engineered butyrate-producing B. subtilis sck6 (BsS-RS06550, BPBS) was engineered by Dr. Liang Bai (Bai et al. 2020). Single colonies of BS and BPBS strains were cultured in LB broth at 37 °C overnight with shaking (200 rpm). Then, the precultures were diluted 1:100 in LB broth and grown at 37 °C with shaking (200 rpm) until the OD600 reached 0.5 ~ 0.6. Then, all the cultures were pelleted by centrifugation, diluted in fresh PBS solution and mixed thoroughly to obtain the appropriate bacterial density (B. subtilis: 5 × 10 8 CFU/ml). all the mice were treated with the control diet for 7 days to acclimatize them to a liquid diet, and then, the ethanol-fed mice received ethanol (5% vol/vol) for 10 days, whereas the pair-fed mice received an isocaloric amount of maltodextrin. To investigate the effects of prophylactic BS intervention, the ethanolfed mice were orally administered BS and engineered BPBS (approximately 1 × 10 8 CFU per mouse) in anaerobic PBS solution or vehicle alone (PBS solution) daily starting from liquid diet acclimatization.
Body weights were measured every other day, and food intake was checked every day. On the last day, the mice received a single dose of ethanol via oral gavage (5 g/kg body weight) and were sacri ced by anesthetizing them with iso urane (4%) after 9 h fasting for excision of tissue samples.

Biochemical analysis
Serum aspartate transaminase (AST) and alanine transaminase (ALT) were measured using the In nity ALT Kit (Thermo Fisher Scienti c). Serum LPS was determined using the Mouse Lipopolysaccharides (LPS) ELISA Kit (Cusabio, Wuhan, China). The serum levels of TNF-α, IL-1β, and IL-6, as well as the hepatic triglyceride (TG) and hepatic lipid peroxidation (MDA) level, were determined using a Mouse ELISA Kit (Solarbio, Beijing, China). All the assays were performed in triplicate according to the manufacturer's instructions.

Intestinal permeability assays
To assay the intestinal permeability of the mice, uorescein isothiocyanate (FITC)-dextran (4 kDa; Sigma-Aldrich) was orally administered (600 mg/kg body weight) 4 h before sacri ce. Blood samples were collected and subsequently centrifuged (4000 rpm, 4℃) for 15 min to isolate serum. Fluorescence was recorded using a spectrophotometer (Tecan) at an excitation wavelength of 485 nm and emission wavelength of 528 nm.

Real-Time qPCR
Total liver and colon RNA were extracted by TRIzol reagent (Invitrogen, USA), total RNA concentration was quanti ed using the NanoPhotometer N50 (Implen, Germany), and reverse transcription was performed using the PrimeScript RT reagent Kit with gDNA Eraser (Takara Bio) on a Mastercycler nexus PCR machine (Eppendorf, Germany). Real-time qPCR was conducted on a LightCycler96 System (Roche, Switzerland) using TB Green Premix Ex Taq II (Tli RNaseH Plus) (Takara Bio) and a cycling program of initial denaturation for 10 min at 95 °C, then 40 cycles of 10 s at 95 °C, 10 s at 62 °C and 10 s at 72 °C, followed by 95 °C for 60 s and a dissociation curve analysis. The primer sequences are listed in Table S1, and the relative gene expression was normalized to 18S and calculated by the 2 −ΔΔCt method (Livak and Schmittgen 2001).

Histopathological observation
For the histological analysis, the liver and colonic tissues were stained with hematoxylin and eosin (H&E). Brie y, the tissues were xed in 10% formalin, and para n-embedded sections were stained with H&E. The para n embedded tissues were sectioned, and then stained with AB-PAS.

Fecal SCFA analysis
Murine feces were collected and quickly frozen in liquid nitrogen, the frozen fecal samples were ground, and the level of SCFA was quanti ed by gas chromatography-mass spectrometry (GC-MS), as previously reported (David et al. 2014 instructions. These raw sequences were processed following the QIIME (v1.9.1) pipeline (Caporaso et al. 2010), and the analysis of gut microbiota diversity and composition of fecal samples was determined.

Statistical analysis
All experimental results were obtained from at least three independent experiments, and the data were expressed as the means ± standard deviation (SD). One-way ANOVA or T-test was used to determine whether the groups were statistically signi cantly different (P < 0.05). GraphPad Prism 8.0 (GraphPad Software) and R were used for all statistical analyses.

Engineered BPBS intervention ameliorates ethanol-induced injury in mice
To explore the effects of BPBS intervention on ethanol-induced injury, we constructed a chronic-binge ethanol feeding mouse model based on a Lieber-DeCarli diet, as described previously (Bertola et al. 2013); the diagrammatic representation of the whole experiment is shown in Figure 1a. During the liquid diet acclimatization period, prophylactic intervention with BS or BPBS was administered by gavage daily until the end of the experiment, and then, the ethanol-containing diet was fed for 10 days. As expected, ethanol consumption signi cantly lowered the murine body weight after 5 days (Figure 1b) and notably increased the liver/body weight ratio (Figure 1c). However, supplementation with wild-type BS and BPBS signi cantly alleviated the decline in body weight gain and partially reduced the liver/body weight ratio, and the e ciency of BPBS intervention was greater than that of BS (Figure 1b, 1c). Additionally, the ethanol exposure stimulated the high expression of IL-6 in serum (Figure 1d), however, the serum IL-6 was signi cantly alleviated via the intervention of BS or BPBS. Collectively, prophylactic BPBS rather than BS supplementation is e cient to alleviate ethanol-induced injury in mice. In line with prior studies, the in uence of ethanol feeding on gut microbiota was con rmed, and ethanol exposure remarkably reduced the gut microbiota abundance and diversity compared with that of the pair-fed group (Figure 2a-c). Meanwhile, the wild-type BS and BPBS intervention signi cantly increased the microbial diversity ( Figure 2a) and obviously restored the microbial composition in the β-diversity analysis (Figure 2c). Interestingly, BPBS supplementation is likely to work more e ciently in a murine ALD model than wild-type BS.

BPBS intervention restores the ethanol-induced gut microbiota disorders in mice
Next, we explored the taxonomic shifts in the bacterial community. At the phylum level, Firmicutes and Bacteroidetes were dominant in the fecal microbiota of the pair-fed group, whereas Firmicutes, Proteobacteria and Bacteroidetes were dominant in the ethanol-fed group (Figure 3a). Interestingly, the proportion of the family Enterobacteriaceae was signi cantly increased in the ethanol-fed group compared with the pair-fed group (Figure 3c), and more signi cant alterations were detected in the BS or BPBS intervention groups than the ethanol-fed group (Figure 3b). After the administration of BS or BPBS, the genera Bacillus and Ruminococcaceae were signi cantly increased (Figure 3d). Speci c enrichment of the families Lachnospiraceae and Prevotellaceae was observed in the gut microbiota of the BPBS-fed group (Figure 3c). We further analysed the altered gut microbiota and found that gram-negative bacteria were enriched along with obvious potentially pathogenic phenotypes in the ethanol-fed group (Figure 3e, 3f). In contrast, the administration of BPBS aborted these potentially pathogenic phenotypes and improved the gut microbiota dysbiosis induced by ethanol consumption.

BPBS intervention improves ethanol-disrupted intestinal barrier in mice
Gut microbiota dysbiosis directly in uences the physiological status of the intestine, and the improved gut microbiota contributes to facilitate host defence against hazardous substances or unfriend environments. We further investigated how prophylactic BPBS intervention mitigated ethanol-induced intestinal injury. Ethanol exposure notably disrupted intestinal barrier integrity with a high FITC concentration in the EtOH-fed group (Figure 4a). However, signi cantly reduced serum FITC levels were observed in the mice with BPBS supplementation compared to the EtOH-fed mice, suggesting a recovery of intestinal integrity. In accordance with the intestinal permeability, ethanol feeding resulted in a signi cant increase in serum LPS compared with that of the pair-fed group, while the BS and BPBS intervention signi cantly decreased serum LPS levels compared with those of EtOH-fed mice (Figure 4b), and BPBS supplementation worked better than wild-type BS. Overall, these results showed that the administration of BPBS mitigated the translocation of endotoxin from the intestinal lumen to circulatory system.
Intestinal barrier integrity is largely dependent on tight junctions, of which Occludin is the major component (Feldman et al. 2005). The protein expression of Occludin in the colon was determined, and ethanol feeding obviously decreased its expression ( Figure 4d); however, the administration of BS and BPBS partially restored the protein Occludin expression compared with that of the EtOH-fed group. In line with the protein results, we observed the same restoration of Occludin gene expression in the BPBS group, along with the increased ZO-1 expression (Figure 4c), suggesting that BPBS administration could stimulate the expression of tight junction genes. Furthermore, histological analysis of the colon showed a damaged and thin mucosal layer after ethanol feeding in comparison with those of the pair-fed mice (Figure 4e), whereas supplementation with BPBS signi cantly restored Muc2 gene expression (Figure 4c) and increased the secretion of mucins in intestine (Figure 4f). Moreover, ethanol feeding signi cantly cut down the butyrate yield of the gut microbiota and activated the in ammatory reaction in the colon, with increased IL-6, IL-1β and TNF-α gene expression compared with the pair-fed group (Figure 4g,4h); however, the administration of BPBS dramatically restored the butyrate contents in the intestine and decreased the excretion of in ammatory cytokines (Figure 4g,4h). Altogether, our results suggested that BPBS intervention replenished the butyrate yield of the gut microbiota and alleviated the in ammatory reaction in colon, as well as rebuilding intestinal barrier function through restoring the tight junction and mucin components.

BPBS intervention attenuates ethanol-induced hepatic injury
To determine how BPBS protects the liver against ethanol exposure, the biochemical and pathologic changes were further carried out. The development of hepatic injury induced by ethanol feeding was con rmed by obviously increased serum ALT, AST, hepatic triglycerides and MDA levels compared with those of the pair-fed mice (Figure 5a-5d). Daily administration of BPBS remarkedly reduced the serum levels of ALT, AST and hepatic triglycerides (Figure 5a-5c). Moreover, obvious neutrophil in ltration in liver tissues after ethanol feeding was observed through H&E staining and was signi cantly alleviated by dietary BPBS supplementation (Figure 5e). Next, we quanti ed hepatic gene expression related to steatosis. Notably, the major liver functions of triglyceride synthesis and fatty acid uptake were disordered, with a signi cant increase in the expression of peroxisome proliferator activated receptor-γ (PPAR-γ) and transporter CD36 for fatty acids (Figure 5f). Additionally, the decreased expression of Fas, SCD1 and Srebp-1c induced by ethanol feeding were improved via dietary BPBS supplementation, which suggested that BPBS administration likely accelerated fatty acid synthesis in the liver and attenuated the hepatic function injury induced by ethanol consumption.

BPBS intervention ameliorates ethanol-induced liver in ammation
Ethanol consumption seriously damaged the intestinal gut integrity and thus accelerated gut bacterial translocation (especially LPS) in the bloodstream, which greatly triggered hepatic in ammation and contributed to the development of ethanol-induced liver diseases. Additionally, mounting evidence have shown that LPS induces organic in ammation based on a TLR4-dependent mechanism (Kayagaki et al. 2013;Hagar et al. 2013;Park et al. 2009). Interestingly, the increased serum LPS levels signi cantly stimulated the protein expression of TLR4 and nuclear factor-κB (NF-κB) in the liver (Figure 6a), along with an increased release of critical proin ammatory cytokines in serum, such as tumor necrosis factor alpha (TNF-α) and IL-1β (Figure 6c,6d). Accordingly, hepatic gene expression related to in ammatory cytokines, including TNF-α, IL-1β, NF-κB and monocyte chemoattractant protein-1 (MCP-1), was greatly enhanced in the ethanol-fed group compared to the pair-fed group (Figure 6b). However, after the administration of wild-type BS and BPBS, the expression of hepatic TLR4 protein was remarkably decreased, along with concomitant signi cant decreases in the TLR4-regulated gene expression of NF-κB, TNF-α, IL-1β, and MCP-1 in the treated group compared to the ethanol-fed group. Notably, BPBS intervention seemed to function better in ameliorating ethanol-induced hepatic in ammatory status via the LPS/TLR4 pathway.

Declarations
Ethics approval and consent to participate All the animal procedures were monitored by the Animal Care and Research Ethics Committee of the Nankai University.
Availability of data and materials Data will be shared whenever it is required. Funding