Plant-Derived Xanthones against Clostridial Enteric Infections

Intestinal bacterial infections are a major threat to human and animal health. In this study, we found plant-derived antibacterial xanthones, particularly α-mangostin (AMG) from the mangosteen peel, exhibiting extraordinary activities against Clostridium perfringens. Structure–activity relationship analysis showed that prenylation modulated the activity of xanthones. The efficacy of AMG (4, 8, 20 mg/kg body weight) was also demonstrated in the broiler chicken necrotic enteritis model infected with Clostridium perfringens. In the models (n = 6 per group), feed supplementation of AMG maintained the homeostasis of the gut microbiome by reducing the colonization of clostridia and promoting the integrity of intestinal barriers via the upregulation of mucin expression. These results suggest that plant-derived xanthones may be a potential alternative to antibiotics for treating clostridial enteric infections in the clinic.


Introduction
Intestinal infections caused by pathogenic bacteria, particularly Clostridium difficile and Clostridium perfringens, are a serious health problem globally [1][2][3][4]. C. perfringens is the third common foodborne bacterial pathogen causing gas gangrene and food poisoning in humans. In the livestock and poultry industry, C. perfringens induces necrotic enteritis, a major concern which has led to severe economic losses [5]. Antibacterial compounds are still the mainstay of treatment for C. perfringens infections. However, inappropriate and irrational antibiotic treatments not only lead to recurrent infections and antibiotic-associated diarrhea, but also contribute to the emergence and spread of antibiotic resistance [6,7]. Thus, the discovery and development of antibiotic substitutes are urgently needed.
Different antibacterial strategies including fecal microbiota transplant (FMT) and probiotics have been employed to prevent clostridial enteric infections [8][9][10]. However, the risk of FMT and probiotics, regarding their possibility to transmit antibiotic-resistant bacteria, has caused much concern in clinic recently [11,12]. Moreover, the unclear mechanism of probiotics limits its clinical application [13,14]. Compared to these alternatives, plant-derived natural products with antibacterial activity possess robust advances in accessibility, structural diversity, and efficacy. Bioactive molecules extracted from plants have shown great potential for the development of antibiotic substitutes against various bacteria [15][16][17][18].
Plants are abundant in a diverse group of flavonoids with robust activity against infection, inflammation, and tumors. The high hydrophobicity characteristic of flavonoids has indicated the potential for oral medication for intestinal infections to have low bioavailability [19][20][21]. However, it has been shown that flavone and flavonol could be metabolized and degraded by enteric Clostridium orbiscindens [22], paralyzing or abolishing their antibacterial efficacy in the gut ( Figure 1). Notably, one group of flavonoids, xanthones, has been claimed to be resistant to the degradation of gut microbiota, owing to its unique tricyclic Recently, we found that xanthones, especially α-mangostin extracted from mangosteen peels, exhibited robust antibacterial activities against methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE) [18,25]. Given the attractive efficacy, we evaluated the potential of plant-derived xanthones as an alternative strategy for the treatment of enteric infections using the broiler chicken necrotic enteritis model. Our findings provide an alternative strategy for treating clostridial enteric infections in human beings and animals.

Structure-Activity Relationship of Xanthones
We previously found the lead compound α-mangostin (AMG), from five categories of flavonoids, to be a promising antibacterial agent with minimal inhibitory concentrations (MIC50) of 0.5 μg/mL against MRSA and VRE. Herein, we further evaluated the antibacterial activity of AMG and other types of flavonoids against C. perfringens. Compared to mulberrin (flavonoid), glabrol (flavanone), lupalbigenin (isoflavone), and isobavachalcone (chalcone), AMG showed better antibacterial activity against five genotypes of C. perfringens, with MIC values of 0.5 μg/mL (Table 1). Meanwhile, we found that AMG exhibited superior activities to routinely used antibiotics including tylosin, tilmicosin, and lincomycin to treat C. perfringens in livestock and poultry (Table S2). To further test the generality of AMG, we expanded to 79 clinical C. perfringens isolates and found that the MIC50 and MIC90 of AMG were kept at 0.5 μg/mL, indicating much more robust efficacy than tylosin with the MIC50 of 1 μg/mL and MIC90 of 4 μg/mL ( Figure S1A, Table S3). In addition, AMG exhibited the same antibacterial activity under anaerobic conditions against different bacteria including MRSA and VRE (Table S4), suggesting that AMG may be efficient against enteric infections in vivo. Recently, we found that xanthones, especially α-mangostin extracted from mangosteen peels, exhibited robust antibacterial activities against methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE) [18,25]. Given the attractive efficacy, we evaluated the potential of plant-derived xanthones as an alternative strategy for the treatment of enteric infections using the broiler chicken necrotic enteritis model. Our findings provide an alternative strategy for treating clostridial enteric infections in human beings and animals.

Structure-Activity Relationship of Xanthones
We previously found the lead compound α-mangostin (AMG), from five categories of flavonoids, to be a promising antibacterial agent with minimal inhibitory concentrations (MIC 50 ) of 0.5 µg/mL against MRSA and VRE. Herein, we further evaluated the antibacterial activity of AMG and other types of flavonoids against C. perfringens. Compared to mulberrin (flavonoid), glabrol (flavanone), lupalbigenin (isoflavone), and isobavachalcone (chalcone), AMG showed better antibacterial activity against five genotypes of C. perfringens, with MIC values of 0.5 µg/mL (Table 1). Meanwhile, we found that AMG exhibited superior activities to routinely used antibiotics including tylosin, tilmicosin, and lincomycin to treat C. perfringens in livestock and poultry (Table S2). To further test the generality of AMG, we expanded to 79 clinical C. perfringens isolates and found that the MIC 50 and MIC 90 of AMG were kept at 0.5 µg/mL, indicating much more robust efficacy than tylosin with the MIC 50 of 1 µg/mL and MIC 90 of 4 µg/mL ( Figure S1A, Table S3). In addition, AMG exhibited the same antibacterial activity under anaerobic conditions against different bacteria including MRSA and VRE (Table S4), suggesting that AMG may be efficient against enteric infections in vivo. Subsequently, we examined the structure-activity relationship of AMG against C. perfringens. Generally, we found that the position of prenylation simultaneously at C2 and C8 of the xanthone skeleton, such as α-, β-, and γ-mangostin, is critical for antibacterial activity, consistent with previous studies [26,27]. Meanwhile, the hydroxyl group at C3 on the parent ring structure promoted antibacterial activity, whereas the methoxyl group at C7 inhibited the efficacy ( Figure 2, Table S5). In addition, AMG displays many desirable pharmacological properties, such as low molecular weight, six hydrogen acceptors, and five rotatable bonds (Table S6), according to Lipinski's rules and eNTRy rules [28]. Nevertheless, AMG suffers from high lipophilicity (log P = 4.64) and low aqueous solubility, probably due to the tetracyclic rigid plane composed of an intramolecular hydrogen bond between the hydroxyl group at C1 with the carbonyl group. Taken together, AMG is a potential lead against clostridia, particularly for the treatment of intestinal tract-associated infections. Subsequently, we examined the structure-activity relationship of AMG against C perfringens. Generally, we found that the position of prenylation simultaneously at C2 an C8 of the xanthone skeleton, such as α-, β-, and γ-mangostin, is critical for antibacteria activity, consistent with previous studies [26,27]. Meanwhile, the hydroxyl group at C3 o the parent ring structure promoted antibacterial activity, whereas the methoxyl group a C7 inhibited the efficacy ( Figure 2, Table S5). In addition, AMG displays many desirabl pharmacological properties, such as low molecular weight, six hydrogen acceptors, an five rotatable bonds (Table S6), according to Lipinski's rules and eNTRy rules [28]. Nev ertheless, AMG suffers from high lipophilicity (log P = 4.64) and low aqueous solubility probably due to the tetracyclic rigid plane composed of an intramolecular hydrogen bon between the hydroxyl group at C1 with the carbonyl group. Taken together, AMG is potential lead against clostridia, particularly for the treatment of intestinal tract-associate infections.

Stability and Safety Assessment of AMG
Given that many herbal substances or preparations are unstable and toxic for thera peutic purposes [29][30][31], we first investigated the stability of AMG under environmenta conditions. We found that AMG showed a two-to four-fold increase in MICs against an tibiotic susceptible S. aureus and MRSA under light conditions ( Figure 3A). Meanwhile AMG maintained the antibacterial activity under pretreatment with different tempera

Stability and Safety Assessment of AMG
Given that many herbal substances or preparations are unstable and toxic for therapeutic purposes [29][30][31], we first investigated the stability of AMG under environmental conditions. We found that AMG showed a two-to four-fold increase in MICs against antibiotic susceptible S. aureus and MRSA under light conditions ( Figure 3A). Meanwhile, AMG maintained the antibacterial activity under pretreatment with different temperatures ranging from −20 • C to 37 • C for 1 h. However, heat treatments showed reduced trends of antibacterial activity, indicated in Figure 3B. Acid resistance is an important feature of oral drugs. Markedly, AMG exhibited a high tolerance to acid (pH = 1, Figure 3C). Compared to the physiological pH condition, the structure of AMG was consistently maintained after being incubated under physiological pH ranging from 3.8 to 7.4 based on the UV absorption spectrum ( Figure 3D). In addition, the resistance to microbial degradation is a prerequisite for AMG against intestinal infections. The widely distributed Eubacterium spp. and Flavonifractor spp. were dominant species in the cecum (top 30) in 112 chicken and 20 swine health samples ( Figure 3E). Although these microbes can produce flavone reductase (FLR) to specifically catalyze flavones and flavonols, they are not able to degrade xanthones. Correspondingly, AMG displayed comparable antibacterial activity after being incubated in simulated gastric and intestinal fluids (Table S7). Lastly, we assessed the acute toxicity of AMG and observed a median lethal dose (LD 50 ) of more than 1000 mg/kg in either SD rats or ICR mice. It is in agreement with the previous report that the LD 50 value of mangosteen extract is an approximate of 1000 mg/kg in BALB/c mice [32]. Therefore, AMG is a low-toxicity substance according to the acute toxicity-grading standard for oral compounds. Collectively, our findings indicate that AMG is a promising alternative to antibiotics for the treatment of gastrointestinal infections. on the UV absorption spectrum ( Figure 3D). In addition, the resistance to microbial degradation is a prerequisite for AMG against intestinal infections. The widely distributed Eubacterium spp. and Flavonifractor spp. were dominant species in the cecum (top 30) in 112 chicken and 20 swine health samples ( Figure 3E). Although these microbes can produce flavone reductase (FLR) to specifically catalyze flavones and flavonols, they are not able to degrade xanthones. Correspondingly, AMG displayed comparable antibacterial activity after being incubated in simulated gastric and intestinal fluids (Table S7). Lastly, we assessed the acute toxicity of AMG and observed a median lethal dose (LD50) of more than 1000 mg/kg in either SD rats or ICR mice. It is in agreement with the previous report that the LD50 value of mangosteen extract is an approximate of 1000 mg/kg in BALB/c mice [32]. Therefore, AMG is a low-toxicity substance according to the acute toxicity-grading standard for oral compounds. Collectively, our findings indicate that AMG is a promising alternative to antibiotics for the treatment of gastrointestinal infections.

AMG Modulates the Composition of Intestinal Microbial Community
Given the attractive efficacy of AMG, we further investigated its therapeutic potential against intestinal infections in vivo. According to a previous method, the model of necrotizing enteritis was established in Arbor Acres broiler chickens [33]. At 14 days of age, broilers were orally, once per day, given C. perfringens suspension for five days ( Figure  4A). The pathological autopsy of the challenged broilers showed obvious flatulence in the

AMG Modulates the Composition of Intestinal Microbial Community
Given the attractive efficacy of AMG, we further investigated its therapeutic potential against intestinal infections in vivo. According to a previous method, the model of necrotizing enteritis was established in Arbor Acres broiler chickens [33]. At 14 days of age, broilers were orally, once per day, given C. perfringens suspension for five days ( Figure 4A). The pathological autopsy of the challenged broilers showed obvious flatulence in the intestinal tract, especially in the posterior segment of the small intestine. Meanwhile, the bacterial load of C. perfringens increased significantly in the cecum of infected broilers ( Figure S2A). These results indicated that the necrotizing enteritis model was successfully established.
biotics 2023, 12, x FOR PEER REVIEW 5 of ( Figure S2A). These results indicated that the necrotizing enteritis model was successfu established. The gut microbial community after antibiotic treatment may provide a favorable vironment for C. perfringens, and cause antibiotic-associated diarrhea [34][35][36]. To ver the potential effect of AMG on intestinal microbiota, we supplemented the diet with AM for three consecutive days at 24 h after infection. Cecal contents were subsequently c The gut microbial community after antibiotic treatment may provide a favorable environment for C. perfringens, and cause antibiotic-associated diarrhea [34][35][36]. To verify the Antibiotics 2023, 12, 232 6 of 14 potential effect of AMG on intestinal microbiota, we supplemented the diet with AMG for three consecutive days at 24 h after infection. Cecal contents were subsequently collected, and microbiota was analyzed based on 16S rRNA gene sequencing. Metagenomics analysis indicated that Firmicutes were the predominant phylum in the cecum ( Figure S1B). The increased number of beneficial bacteria, such as Lactobacillus spp., and reduction of pathogenic bacteria, such as Brucella spp., was demonstrated in AMG treatment groups ( Figure 4B). In addition, the α-diversity analysis showed that species abundance and diversity were maintained after AMG treatment, compared to the uninfected groups ( Figure 4C and Figure S1C). These findings are consistent with the operational taxonomic units (OTUs) result ( Figure 4D). In addition, the β-diversity analysis showed that the composition of the intestinal microbial community after AMG treatment is similar to that in uninfected groups ( Figure 4E and Figure S1D). Last, based on LEfSe analysis, we found that Bacillus spp. dominate the cecum flora in the AMG treatment group, whereas Clostridium spp. dramatically increased in infections. Similarly, the LDA scores indicated that the relative abundances of Blautia and Lachnospiraceae were significantly higher in the challenged group than the AMG treatment group ( Figure 4F and Figure S1E). AMG sharply inhibited pathogenic bacteria and reversed the dominant bacteria as the control. These results suggest that AMG reduces the abundance of pathogenic bacteria and potentiates the growth of probiotics, especially Lactobacillus, thereby efficiently ameliorating the dysbiosis in necrotizing enteritis.

AMG Effectively Controls Clostridial Enteric Infections
Modulation of the host immune response is also an effective method to control necrotic enteritis with C. Perfringens. To further evaluate the effect of AMG on the host, we first assessed the expression levels of inflammatory factors in mammalian cells after AMG treatment based on quantitative PCR (qPCR) and ELISA tests. We found that AMG inhibited IL-1β expression, whereas it showed no effects on TNF-α ( Figure 5A and Figure S2B,D). Interestingly, AMG at the low level (1.2 µmol/L) enhanced the phagocytic activity of macrophages ( Figure 5B). Therefore, AMG exhibits anti-inflammatory and immunomodulatory potential.
The therapeutic effect of AMG in vivo was subsequently evaluated using the necrotizing enteritis model. AMG had a significant inhibitory effect on C. perfringens in a dose-dependent manner ( Figure 5C). Given that C. perfringens usually colonizes in the cecum, we further evaluated the distribution of AMG and found that AMG was predominantly found in the cecum as well ( Figure 5D and Figure S2C). Moreover, the intestinal villi structure of broilers in the infected groups disintegrated entirely based on pathological analysis. Inspiringly, AMG treatment almost maintained the intestinal villi structure of broilers indicating that AMG effectively alleviates the pathological lesion of intestinal tissue caused by necrotizing enteritis ( Figure 5E). Enteric clostridial infection has been reported to reduce the expression of intestinal tight junction genes, and promotes the hydrolysis of tight junction proteins [37]. We found that AMG significantly up-regulated the expression of mucin protein MUC2 through qPCR analysis ( Figure 5F). This result was also confirmed by immunofluorescence colocalization of MUC2 ( Figure 5G). Thus, AMG is effective against necrotizing enteritis in chickens, suggesting that plant-derived molecules are prospective sources for developing antibiotic alternatives to prevent gastrointestinal infections.
crotic enteritis with C. Perfringens. To further evaluate the effect of AMG on the host, we first assessed the expression levels of inflammatory factors in mammalian cells after AMG treatment based on quantitative PCR (qPCR) and ELISA tests. We found that AMG inhibited IL-1β expression, whereas it showed no effects on TNF-α ( Figures 5A and S2B,D). Interestingly, AMG at the low level (1.2 μmol/L) enhanced the phagocytic activity of macrophages ( Figure 5B). Therefore, AMG exhibits anti-inflammatory and immunomodulatory potential.

Discussion
Anaerobic Clostridium spp., especially Clostridium difficile and Clostrdium perfringens, ubiquitously distribute in the abiotic environment and biotic intestinal tracts. These pathogens are the primary causative agents of intestinal diseases such as necrotizing enteritis and antibiotic-associated diarrhea [38,39]. Antibiotics are routinely used to control broiler necrotic enteritis such as lincomycin, tylosin, penicillin, and tilmicosin. However, antibiotic resistance and antibiotic-associated diarrhea become a serious problem in the treatment of C. perfringens [40]. Plant-derived compounds have drawn much attention as a promising source of antibiotic substitutes [41][42][43] due to their abundance, versatile chemodiversity, and high accessibility [44]. Moreover, the unique membrane targeting mechanism without antibiotic resistance indicates the antibacterial potential [18]. To find a novel strategy to control necrotic enteritis infected with C. perfringens, we detected the antibacterial activity of plant-derived flavonoids against C. perfringens. The lead compound AMG of xanthones exhibits robust bactericidal activity against C. perfringens, with MIC 50 of 0.5 µg/mL and MIC 90 of 0.5 µg/mL. AMG could not only maintain the homeostasis of the gut microbiome by reducing the colonization of clostridia, but could also promote the integrity of intestinal barriers via the upregulation of mucin expression. Thus, plant-derived xanthones provide an alternative strategy to treat clostridial enteric infections.
The therapeutic efficacy frequently suffers from pharmacokinetic behaviors. First, pH stability is the unique advantage of gastrointestinal drugs. Inspiringly, AMG, the candidate of xanthones, maintained the structure and the antibacterial activity after being incubated at pH = 1 for 1 h. Moreover, previous studies have demonstrated the poor bioavailability of AMG, suggesting its low gastrointestinal absorption [45,46]. We found AMG was mainly distributed in the cecum ( Figure 5D), supporting the idea that AMG was hardly absorbed in the gastrointestinal tract. Most importantly, it has been confirmed AMG could not be metabolized and degraded by enteric C. orbiscindens, owing to the unique tricyclic privileged structures [22] (Figure 1). Taken together, AMG can exert its efficacy fully in the intestine, reducing the dosage and administration times and decreasing the risk of drug resistance in a superior manner to other antibiotics.
Bioactive substances with dual functions of antibacterial and immune regulation are potential choices for antibacterial therapy [47]. Encouragingly, AMG exhibits antiinflammatory and immunomodulatory potential ( Figure 5A,B) and effectively alleviate the pathological lesions of intestinal tissue ( Figure 5E,F). Bacteria and host-oriented dualfunctional antibacterial xanthones, including AMG, provide an effective strategy for the treatment of intestinal infection with C. perfringens ( Figure 6). Bioactive substances with dual functions of antibacterial and immune regulation are potential choices for antibacterial therapy [47]. Encouragingly, AMG exhibits anti-inflammatory and immunomodulatory potential ( Figure 5A,B) and effectively alleviate the pathological lesions of intestinal tissue ( Figure 5E,F). Bacteria and host-oriented dualfunctional antibacterial xanthones, including AMG, provide an effective strategy for the treatment of intestinal infection with C. perfringens ( Figure 6).

Chemicals and Bacterial Strains
Standard compounds of xanthone were purchased from Chengdu Biopurify Phytochemicals Ltd. with a purity of ≥98%. Before starting the experiment, compound stock

Chemicals and Bacterial Strains
Standard compounds of xanthone were purchased from Chengdu Biopurify Phytochemicals Ltd. with a purity of ≥98%. Before starting the experiment, compound stock solutions were prepared in dimethylsulfoxide (DMSO, Aldrich, Taufkirchen, Germany) at a concentration of 5120 µg/mL. S. aureus ATCC 29,213 was obtained from the American Type Culture Collection (ATCC); CVCC2030 and CVCC60102 were purchased from China Veterinary Microbial Species Preservation and Management Center, China Veterinary Drug Control Institute. Other C. perfringens clinical isolates were obtained from ileum contents, cloacal swabs, or feces of chickens, pigs, and cattle from large-scale livestock and poultry farms during 2019 in China, then identified and preserved by the National Center for Veterinary Drug Safety Evaluation at China Agricultural University, as noted in Table 1.

Antibacterial Tests
Minimum inhibitory concentrations (MICs) were determined by the broth microdilution method following the Clinical and Laboratory Standards Institute (CLSI) 2022 guidelines. The range of concentrations typically employed for each experiment was 128-0.25 µg/mL (2.56% final DMSO concentration). Tylosin was used as the positive control in this experiment. All C. perfringens bacterial strains were sub-cultured on Brain Heart Infusion agar (BHA, Bridge Technology Co., Ltd., Beijing, China) with 5% sheep blood and incubated overnight in an anaerobic jar containing 10% H 2 , 10% CO 2, and 80% N 2 at 37 • C. The test medium for C. perfringens species was Fluid Thioglycollate Medium (FTG, Bridge Technology Co., Ltd., Beijing, China), and bacterial suspensions were diluted to obtain a final inoculum of 1.5 × 10 6 CFU/mL for broth microdilution experiments. The lowest concentrations of compounds with no visible growth of bacteria were the MICs after incubation at 37 • C for 24 h.

Stability of AMG
To determine the light stability, AMG was pretreated with sunlight exposure for 3 h, and the residual antimicrobial activity was evaluated against S. aureus. To determine the thermal stability and pH stability, AMG was treated at a temperature ranging from −20 to 60 • C, and at pH ranging from 1.0 to 6.0 at 37 • C for 1 h, respectively. Subsequently, all samples were readjusted to pH 7.0 to determine the residual antimicrobial activity against S. aureus.

Antibacterial Activity of AMG in the Simulated Gastrointestinal Model
Simulated gastric fluid (SGF) and simulated intestinal fluid (SIF) were prepared using a commercial kit (Phygene, Fuzhou, China). AMG was first mixed with SGF. After 2 h incubation at 37 • C with continuous shaking at 100 rpm, the SGF was adjusted to neutral pH with 2 mol/L NaOH and then mixed with an equal volume of 2 × SIF. The mixtures were two-fold diluted in FTG Medium and mixed with C. perfringens suspended in BHA at a final concentration of 1 × 10 6 CFUs mL −1 in 96-well plates. Subsequently, the mixtures were incubated anaerobically at 37 • C for 5 h to evaluate their antibacterial activity.

In vitro RT-qPCR and ELISA Assays
MH-S cells were seeded at 1 × 10 6 cells per well in 12-well culture plates (Corning Incorporated Co., New York, NY, USA) and incubated at 37 • C overnight to form monolayers. MH-S cells were treated with 10 µmol/L AMG, and 1 µg/mL LPS was added and co-incubated for 4 h after half an hour. In addition, Dexamethasone (DXM) was used as a controlled drug. After that, supernatants of cell cultures were acquired by centrifugation at 1000× g for 5 min at 4 • C. The levels of TNF-α, IL-6, and IL-1β in the supernatants were detected by commercial ELISA kits (BioLegend, Inc., San Diego, CA, USA) following the manufacturer's instructions. Meanwhile, Total RNA was extracted using a HiPure Total RNA Plus Mini Kit (Magen Biotechnology Co., Guangzhou, China), and 1 µg of extracted RNA was reverse transcribed with a PrimeScript™ RT reagent Kit (Takara, Dalian, China) following the manufacturer's protocol. The messenger RNA levels of TNF-α, IL-6, and IL-1β relative to the control genes β-actin in MH-S cells were determined by real-time PCR tests using SYBR Green Master Mix. The primer sequences are listed in Supplementary  Table S8.

Phagocytosis Assay
MH-S cells were seeded into a 12-well plate at a concentration of 3 × 10 5 cells per well at 37 • C for 8 h. Cells were treated with 1.2 µmol/L AMG for 24 h, and the medium of the phagocytosis inhibition positive control group was replaced with DMEM containing 1 µmol/L Cytochalasin D (CCD) 1 h before the incubation process was completed. The spent culture medium at the end of the incubation process was removed and the MH-S cells washed with PBS once. We suspended the FITC-labeled particles into DMEM (1:1000 dilution) and added them to the plates (500 µL/well), this proceeded for half an hour, and then we removed the medium and washed with PBS again. We trypsinized cells and resuspended them with cold PBS to stop phagocytosis. Finally, the cell suspension was filtered through a 300-mesh copper grid and run in a flow cytometer to detect FITC fluorescence.

Animals and Ethics Statement
Commercial 1-day-old Arbor Acres (AA) broiler chickens were purchased from Beijing Arbor Acres Poultry Breeding Co., Ltd. (Beijing, China). All animal studies were carried out at China Agricultural University, and all the animals were randomized to cages for each experiment and had free access to food and water. The animals were maintained in strict accordance with the regulations of the Administration of Affairs Concerning Experimental Animals, approved by the State Council of the People's Republic of China (14 November 1988) and by the China Agricultural University (CAU) Institutional Animal Care and Use Committee guidelines (ID: SKLAB-B-2010-003) approved by the Animal Welfare Committee of CAU.

In Vivo Necrotic Enteritis (NE) Disease Model
The NE disease induction model was performed as previously described [33,48]. Briefly, a total of 36 commercial 1-day-old Arbor Acres broiler chickens (Beijing Arbor Acres Poultry Breeding Co. Ltd., Beijing, China) were divided into four experimental groups, including a non-challenged control, a challenged control group, and two challenged groups supplemented with either AMG (4 mg/kg, 8 mg/kg or 20 mg/kg body weight) or amoxicillin (1 mg/kg body weight). From day 0 to day 11, all broiler chickens received a control diet with no antibiotics. On day 12 the feed was changed to a mixture of 50% wheat-based and 50% fish meals. On the evening of the 13th day, the feed was withdrawn, and each bird was orally challenged with 1.0 mL of C. perfringens culture (10 9 CFU/mL). From day 18 to day 20, birds were treated with AMG for three days. On day 21, chickens were euthanized with inhaled carbon dioxide gas, and their small intestines (duodenum to ileum) were examined for gross necrotic lesions. C. perfringens was recovered from the intestines of killed birds by culturing on Tryptose Sulfite Cycloserine Agar Base (TSC, Bridge Technology Co., Ltd., Beijing, China).

Microbiota Analysis
Total genome DNA from about 150 mg of cecum content of broiler chickens treated with AMG was extracted using the CTAB/SDS method. The 16S rRNA gene V3-V4 region was amplified using specific primers F341 (CCTACGGGRSGCAGCAG) and R806 (GGAC-TACVVGGGTATCTAATC). The chimera sequences were filtered using USEARCH, and then the remaining sequences were assigned to generate the same operational taxonomic units (OTUs) at 97% similarity. The alpha analysis based on USEARCH alpha_div was conducted to reveal the diversity indices. The beta diversity analysis was performed using USEARCH beta_div. Linear discriminant analysis was performed with LefSe (version 1.0.8) with default parameters (LDA score ≥ 2). Anosim analysis was used, Vegan Package(Version 2.4.2). The data were analyzed using R version 3.4.3 and GraphPad PRISM 8 (GraphPad Software, San Diego, CA, USA). All data are expressed as the mean ± SD.

Tissue Distribution of AMG
The distributions of AMG in the cecum were conducted according to a previously reported method [45]. About 0.2 g of cecum contents was taken from each chicken and homogenized with 1 mL sterile water at 5000 rpm for 10 min. AMG was extracted by adding 2 mL ethyl acetate, followed by 5 min vortex and 5 min ultrasonic extraction. After centrifuging at 9000× g rpm at 4 • C for 10 min, the supernatant was collected for nitrogen blow drying at 40 • C and redissolved in 500 µL 80% acetonitrile. Finally, samples were centrifuged at 14,000× g rpm for 15 min, and the supernatant was collected into filter vials and stored until LC-MS/MS analysis.

Determination of AMG
The analytical method for measuring AMG was modified from a previous report [49]. The sample and standard compound were dissolved with acetonitrile and completed with 0.22 µm pore filter membranes. Subsequently, 5 µL of the supernatant was directly injected into a reversed-phase C18 column (Cadenza CD-C18, 2 mm × 75 mm i.d., 3-µm particle size; ImtaktKyoto, Japan) and kept at 4 • C until injection. The mass transitions for AMG were m/z 411.10/355.10 (collision energy, 21.0 eV), in the multiple reaction monitoring (MRM) mode with positive ionization. These analytes were separated on the column with a gradient elution consisting of 0.1% formic acid in acetonitrile (A phase) and 0.1% formic acid in water (B phase) at a flow rate of 0.30 mL/min. The gradient conditions were as follows: 0.00-1.00 min, 80% A; 1.00-4.00 min, 80-100% A; 4.00-5.00 min, 100% A; 5.00-5.50 min, 100-80% A; 5.50-6.50 min, 80% A for 1 min. All analytical data were processed using the Shimadzu lab solutions (Version 5.106).

Cecal Epithelial Integrity Assays
The RNA expression levels of cecal tight junction proteins Claudin-1, Occludin, and ZO-1 and mucin protein MUC2 were measured in each group. Briefly, about 1 cm of cecal tissue was cut into a precool centrifuge tube containing 1 mL Buffer RL on ice. We homogenized the tissue until no pieces could be seen, then centrifuged the homogenate at 14,000× g rpm at 4 • C for 5 min. After that, total RNA was extracted using a commercial RNA extraction kit.

Immunofluorescence Staining
Briefly, tissues were cleared with xylene, dehydrated in an ethanol series, then embedded in paraffin. Then, the paraffin section slide was dewaxed, and antigen retrieval was conducted by transferring the slide to a container with unmasking solution EDTA (pH = 9.0) and microwaving. After washing, the samples were incubated with Autofluorescence Quenching Kit for 5 min and rinsed with running water for 10 min. Subsequently, blocking solution BSA was added for 30 min. Then, samples were incubated with primary antibodies overnight at 4 • C in the humid box. After repeated washing, the samples were incubated with secondary antibodies at room temperature for 50 min protecting them from light. After repeated washing, DAPI was added for 10 min to visualize DNA. Images of the samples were obtained after the mounting step using confocal scanning microscopy. The fluorescence of DAPI (blue) was detected at the excitation wavelength of 380 nm and emission wavelength of 420 nm. The fluorescence of MUC2 (red) was detected at an excitation wavelength of 510~560 nm and an emission wavelength of 590 nm.

Statistical Analysis
Experimental results were analyzed for statistical significance using GraphPad Prism 8.0.1 (GraphPad Software Inc.). Differences were analyzed by Student's t-test and one-way ANOVA. The values reported were shown as mean ± standard deviation (s.d.). p values were indicated in the figures, and not significant (p > 0.05) was marked with n.s.

Conclusions
Our results demonstrate the potential of xanthones against clostridial enteric infections, providing an alternative strategy for the treatment of clostridial pathogens-associated infections. Plant-derived xanthones may be a promising source for the discovery and development of novel antibiotics to treat clostridial enteric infections in clinic.

Supplementary Materials:
The following supporting information is available online at https:// www.mdpi.com/article/10.3390/antibiotics12020232/s1, Table S1: Bacteria strains were used in this study, Table S2: Antibacterial activities of clinically applied antibiotics against clostridia, Table S3: Antibacterial activities of AMG against 79 clinical C. perfringens isolates, Table S4: Antimicrobial activities of AMG in aerobic and anaerobic conditions, Table S5: Structure-activity relationships for antibiotic activity 10 xanthones, Table S6: Physiochemical Properties of AMG, Table S7: Stability test of AMG in the gastroenteric fluid, Table S8: Primer sequences used in the study; Figure S1: AMG modulates the composition of intestinal microbial community, Figure S2: AMG is efficacious in the necrotizing enteritis model of clostridial enteric infections.