Preventive Antibiotic Treatment Increases Neonatal Chicks Susceptibility to Salmonella Infection Via Disrupting Gut Microbiota and Linoleic Acid Metabolism

Background: Antimicrobial agents have been widely used in animal farms to prevent and treat animal diseases. However, antimicrobial agents may change the bacterial community and increase susceptibility to the pathogenic bacteria infection. Here, we used metagenomic and metabonomic approach to investigate the effects of orfenicol (FFC) pre-treatment on colonization of Salmonella enterica serovar Enteritidis (S. Enteritidis) in intestines of neonatal chicks through analysis of host responses, microbiota and metabolic changes. Results: We observed that FFC pre-treatment signicantly increases the level of S. Enteritidis in the cecal contents, spleen and liver and also induces changes to the cecal microbiota and metabolism. Prior to S. Enteritidis infection, FFC signicantly reduced the content of Lactobacillus, and signicantly affected the linoleic acid metabolism pathway, including signicantly reducing the levels of conjugated linoleic acid (CLA), and signicantly increasing the abundance of 12,13-EpOME and 12,13-diHOME in cecum. After infection with S. Enteritidis, the abundance of Proteobacteria were signicantly increased and the Salmonella-induced intestinal inammatory responses and intestinal barrier damage were exacerbated. Supplementation with CLA could maintain intestinal integrity, reduce intestinal inammation, and directly inhibit Salmonella growth to effectively reduce the Salmonella colonization, whereas the 12,13-diHOME through promoting intestinal inammation and destroying the intestinal barrier function to support the Salmonella infection. Conclusions: Overall, FFC can decrease levels of Lactobacillus and CLA, and elevate cecal 12,13-diHOME concentrations in neonatal chicks and thereby increases susceptibility to Salmonella infection. This study revealed a potential health impact of antibiotics and disturbed gut microbiota and linoleic acid metabolism might be an intestinal health-impairing attribute and may contribute to Salmonella colonization. was signicantly decreased. Moreover, the inammatory cytokines IL-6, and IFN-γ) were also signicantly increased in the FT group. These results indicate that the FFC exacerbated the Salmonella-induced inammatory response. It is possible that antibiotic-treated causes Gram-negative bacterium releases LPS, which promotes intestinal inammation, and this can be conrmed by serum LPS levels. Moreover, FFC may also shifts the gut microbiota and metabolic proling of neonatal chicks, causing microbiotic and metabolic disorders and support Salmonella colonization. metabolism including signicantly reducing the levels of CLA, and signicantly increasing the abundance of 12,13-EpOME and 12,13-diHOME. And we found that the CLA can maintain intestinal integrity, reduce intestinal inammation, and directly inhibit Salmonella growth to effectively reduce the Salmonella colonization in chicks. Whereas the 12,13-diHOME though promoting intestinal inammation and destroying the intestinal barrier function to support the Salmonella colonization. These ndings suggest that the decreased levels of Lactobacillus and CLA, and elevated cecal 12,13-diHOME concentrations in FFC pre-treated neonatal chicks might be an intestinal health-impairing attribute and may contribute to Salmonella

(CLA), and signi cantly increasing the abundance of 12,13-EpOME and 12,13-diHOME in cecum. After infection with S. Enteritidis, the abundance of Proteobacteria were signi cantly increased and the Salmonella-induced intestinal in ammatory responses and intestinal barrier damage were exacerbated. Supplementation with CLA could maintain intestinal integrity, reduce intestinal in ammation, and directly inhibit Salmonella growth to effectively reduce the Salmonella colonization, whereas the 12,13-diHOME through promoting intestinal in ammation and destroying the intestinal barrier function to support the Salmonella infection.
Conclusions: Overall, FFC can decrease levels of Lactobacillus and CLA, and elevate cecal 12,13-diHOME concentrations in neonatal chicks and thereby increases susceptibility to Salmonella infection. This study revealed a potential health impact of antibiotics and disturbed gut microbiota and linoleic acid metabolism might be an intestinal health-impairing attribute and may contribute to Salmonella colonization.

Background
The global human population will reach 8.5 billion by 2030, which rising high demand for international livestock products [1]. Chickens are the most abundant livestock in the world, which more than 60 billion chickens are produced annually, with production predicted to increase further in the next 20 years [2,3].
The health of chickens is essential to ensure the safety of poultry products and effective control and management of disease. Salmonella enterica is a major medical and economic problem worldwide, and causes signi cant morbidity and mortality in poultry [4,5]. The Salmonella colonization in chickens is of particular importance because the association of this pathogen with poultry products is considered to be a signi cant source of infection to humans, is the leading cause of foodborne disease outbreaks worldwide [6][7][8][9].
Microbiota-mediated mechanisms of colonization resistance is critical to prevent pathogens invasion in some niches within animal hosts [10]. Particularly, the intestine is best site to characterized the mechanisms of colonization resistance because the intestinal lumen contains trillions of commensal microbes and those microbes are interact closely with their hosts as well as each other in intricate microbial networks [11,12]. The microbial community of the chicken intestine is highly diverse with more than 1000 species of bacteria and the population density can be up to about 10 11 cells/g intestinal contents, which can protect chickens from the infection [13,14]. The gut microbiota helps protect chickens from colonization by zoonotic pathogens, but this can be weakened through antibiotics administration that disturb the gut bacterial community and metabolism [15]. The large-scale intensive rearing systems usually containing ocks of 20,000 birds or more, which depend on antibiotics to prevent and treat animal disease [16]. In China, the veterinary antibiotics accounted for 84.3% (pig: 52.2%, chicken: 19.6%, and other animals: 12.5%), whereas the human antibiotics only shared 15.6% [17]. And it is estimated that the global consumption of antibiotics used for chickens, pigs, and cattle will increase by 67%, from 63,151 tons in 2010 to 105,596 tons in 2030 [18]. Consequently, disruptions of the microbiota composition induced by antibiotics lead to disorder in the ecological balance between microbes and host, dramatically reduces the colonization resistance and enhances the susceptibility to infection by various pathogens [19][20][21]. In oral administration infection models in mice, antibiotic pretreatment allows e cient colonization of the cecum and colon by S. enterica serovar Typhimurium, and enhances Salmonella expansion and fecal shedding [22,23].
Furthermore, the neonatal chicks exhibit highly susceptibility to infections with Salmonella serovars because of their immature gut microbiota, and the complexity of the neonate gut microbiota gradually increases from day 1 to day 19 of life [24,25]. Although the chickens are coprophagic and transfer of cecal microbiota from adult chickens to neonatal chicks increases resistance to Salmonella infection [26,27]. Unlike other farm animals, the neonatal chicks are hatched in the clean environment of a hatchery without any contact with adult chickens and their colonization resistance is only dependent on the environment [28]. If a pathogen appears in the environment, the immature gut microbiota of the newly hatched chick enabling such a pathogen essentially unrestricted multiplication. Therefore, the large-scale intensive rearing systems usually depend on antibiotics to prevent and control the disease outbreaks, which makes the gut microbiota of chicks disrupted and may lead to the chicks are more susceptible to Salmonella infection.
We hypothesize that the newly hatched chicks are pre-treated with an antibiotic accompany changes with the disorders of the gut microbiota and metabolic pro le, which lead to the chicks are more susceptible to Salmonella infection. Therefore, it is necessary to better understand how antibiotics affect the mechanism of colonization resistance. The objectives of the study were to investigate the in uence of a 7-day treatment course of orfenicol, a most commonly used broad-spectrum antibiotic in poultry in many countries [17,18,29], on the composition of the microbiota community and metabolic pro le of the neonatal chicks, and nd the potential factors enable support S. Enteritidis growth in the cecum of neonatal chicks.

Bacterial strains
The Salmonella enterica serovar Enteritidis (S. Enteritidis, ATCC 13076) oR mutant strain was used as the challenge strain. The S. Enteritidis oR mutant was constructed using the plasmid homologous recombination integration method as previously described [30]. Brie y, the sequence of oR gene (NG_047860.1) was synthesized and cloned into E. coli Cloning vector. Then the upstream and downstream homologous recombinant arms were ampli ed by PCR from S. Enteritidis genome use ultradelity DNA polymerase, and oR sequence was ampli ed by PCR from template vector. The fragments were assembled by the Fusion PCR to construct the gene targeting fragment, and the pCVD442 suicide plasmid was digested with restriction endonuclease to construct gene target plasmid pCVD442-oR, which was transformed into E. coli SY327λpir. The plasmid pCVD442-oR was conjugated into S. Enteritidis using E. coli SY327λpir as a donor strain and plated on orfenicol (FFC) resistant chromogenic xylose lysine tergitol 4 (XLT4) agar plate to select for positive clones that had integrated the suicide plasmid. Finally, a colony that was FFC resistant veri ed by PCR and gene sequencing. Prior to inoculation, S. Enteritidis was grown overnight in Luria-Bertani broth at 37 °C with shaking at 200 rpm.

Chicken, orfenicol intervention and infection
Leghorn layer chicks (1-day-old) were hatched from the same batch eggs of speci c-pathogen-free (SPF) birds (Beijing Boehringer Ingelheim Vital Biotechnology Co., Ltd., China), and each assigned group was reared in an individual GJ-1 SPF isolator (Suzhou Fengshi Laboratory Animal Equipment Co., Ltd., China). Animals were received non-medicated chick feed and water ad libitum and raised under controlled environmental conditions with a 16-h lighting cycle and a temperature of 32°C at day 1 which was gradually reduced and maintained at 24°C on day 10.
Animal protocol 1: Effect of orfenicol pre-treatment on intestinal Salmonella colonization. Eighty-eight newly hatched chicks were assigned at random to four groups, each group included 22 chicks in three time points (n = 7 to 8 chicks in each time point) and treated with a 7-day antibiotic treatment (30 mg/kg b. w.) of orfenicol (FFC) or infected with ~10 8 cfu of the challenge strain S. Enteritidis by oral gavage. The treatments were as follows: (1) NT, control group neither FFC-treated nor S. Enteritidis-infected; (2) FT, FFC-treated group; (3) ST, S. Enteritidis-infected group; (4) and FST, FFC-pre-treated and S. Enteritidisinfected group. On days 11, 18 and 25, the chicks were euthanized for analysis.
After chicks were euthanized, the cecal contents and internal organs were aseptically collected and homogenized using the PBS. For enumerating Salmonella loads, an aliquot (100 μl) of appropriate dilutions was spread onto XLT4 agar plates (50 ug/ml orfenicol), which Salmonella appearing typical black colonies after incubation at 37 °C for 24 h.
Histopathology and microscopic analysis of the intestine Parts of ileal tissue were perfusion-xed with formalin for 24 h. After gradient dehydration with ethanol, specimens were embedded in para n. Subsequently, 5 μm sections were rehydrated and stained with Alcian blue. Representative images were obtained by a BA400 digital microscope (Motic Group CO., LTD., China). The mean density was calculated by the integral optical density and area of the positive Alcian blue staining using the Image-Pro ® Plus v6.0 analysis system (Media Cybernetics, USA), and signi cance of differences was determined by a one-way ANOVA. A P-value of less than 0.05 was considered statistically signi cant. To determine the degree of lesion, pathological score was monitored as previously described [32]. And the scanning electron microscope (SEM) (Inspect TM , FEI Ltd., USA) of the intestinal villi was conducted as previously described [33].
DNA extraction, 16S rRNA gene sequencing and data analysis Seven or eight chicks per treatment were randomly chosen at three different time points, 11, 18, and 25 days of age, and euthanized by carotid artery bleeding. The cecal contents were collected within 5 min of euthanasia, immediately placed in pre-cooling cryogenic vials, and stored at −80 °C until DNA extraction. Total genomic DNA was extracted from cecal contents using the QIAamp DNA Stool Mini Kit (Qiagen, Germany) according to manufacturer's protocols, and stored at −20 °C prior to further analysis. The concentration and quality of extracted DNA samples were measured by Nanodrop 2000 (Thermo Fisher Scienti c, Waltham, MA, United States) and agarose gel electrophoresis, respectively.
Using the isolated genomic DNA as the template, the V3-V4 hypervariable regions of the bacterial 16S rRNA genes were PCR-ampli ed with primers 338F (5′-ACTCCTACGGGAGGCAGCA-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) following the method previously described [34]. Amplicons were then sequenced on the Illumina MiSeq platform (Illumina Inc., USA) using 2 × 250 bp cycles. These sequence data are deposited in the NCBI database within the Bioproject PRJNA655362 under the SRA study SRP277009. The QIIME was employed to process the sequencing data. Briefly, raw sequencing reads with exact matches to the barcodes were assigned to respective samples and identified as valid sequences. The low-quality sequences were ltered through following criteria [35,36]: sequences that had a length of <150 bp, sequences that had average Phred scores of <20, sequences that contained ambiguous bases, and sequences that contained mononucleotide repeats of >8 bp. Paired-end reads were assembled using FLASH [37]. After chimera detection, the remaining high-quality sequences were clustered into operational taxonomic units (OTUs) at 97% sequence identity by UCLUST [38]. A representative sequence was selected from each OTU using default parameters. OTU taxonomic classi cation was conducted by BLAST searching the representative sequences set against the Greengenes Database [39] using the best hit [40]. An OTU table was further generated to record the abundance of each OTU in each sample and the taxonomy of these OTUs. OTUs containing less than 0.001% of total sequences across all samples were discarded. To minimize the difference of sequencing depth across samples, an averaged, rounded rare ed OTU table was generated by averaging 100 evenly resampled OTU subsets under the 90% of the minimum sequencing depth for further analysis.
Sequence data analyses were mainly performed using QIIME and R packages (v3.2.0) [41]. OTU-level alpha diversity indices, such as Shannon indices, species abundance and Pielou indices were calculated using the OTU table in QIIME. Beta diversity analysis was performed to investigate the structural variation of microbial communities across samples using Bray-Curtis distances metrics and visualized via principal coordinate analysis (PCoA). The taxonomy compositions and abundances were visualized using MEGAN [42] and GraPhlAn [43]. Linear discriminant analysis effect size (LEfSe) was performed to detect differentially abundant taxa across groups using the default parameters [44].
Quantitative PCR for microbiota analysis Bacterial composition of the microbiota was measured by qPCR as previously described [45][46][47][48]. All qPCR reactions were performed using the Bio-Rad real-time PCR detection system (Bio-Rad CFX Maestro 1.1, 3.0, USA) and SsoFast EvaGreen Supermix (Bio-Rad Inc., USA) and tested according to the manufacturers' instructions. The extracted genomic DNA in the cecal content was used as a template for qPCR using the main group-speci c primers (Supplementary Table 1): All eubacteria, Lactobacillus, Bacteroidetes, Enterobacteriaceae, Clostridium butyricum and Faecalibacterium prausnitzii. Serial dilutions of plasmids containing the target gene cloned into the pMD-19 T cloning vector (TaKaRa, Dalian, China) were analyzed to generate a standard curve and calculate absolute counts of the target gene.

Metabolomics for chicken cecal content
Untargeted metabolomics. Chickens were sacri ced, and the cecum was resected. The cecal contents were obtained and stored at -80 °C until use for the metabolomics analysis. The method of sample preparation for LC/MS detection are previously described [49]. Brie y, 50 mg freeze-dried sample, 800 μl methanol, and 5 μl DL-o-Chlorophenylalanine (internal standard) were added to a 1.5 mL Eppendorf tube. And all the samples were grinded to ne powder using grinding mill at 65 HZ for 90 s followed by being vortexed for 30 s, and centrifuged at 12,000 rpm, 4°C for 15 min. Then 200μL of supernatant was transferred to a new vial for LC-MS analysis. A total of 10 μl of the sample solution at 4 °C were injected to the LC-MS system (Thermo, Ultimate 3000LC, Exactive Orbitrap) with an Agilent C18 column (Hypergod C18, 100 x 2.1 mm 1.9 μm) and the column temperature was maintained at 40 °C. The mobile phase consisted of solution A and B [A was 0.1% formic acid/5% acetonitrile/water (v/v/v) and B was 0.1% formic acid/acetonitrile (v/v)] and the ow rate was 350 μl/min. The gradient was set as follows: 0% B at 0 min, 20% B at 1.5 min, 100% B at 9.5 min to 14.5min, and 0% B at 14.6 min to 18min. Samples were analyzed in positive and negative ion modes using 300 °C of heater temperature, 350 °C of capillary temperature, and 3.0 KV of spray voltage. The ow rates of sheath gas, aux gas, and sweep gas were 45, 15, and 1 arb, respectively. The peaks were aligned according to the m/z value and normalized migration time. The peak areas were calculated by normalizing against the internal standards, and the metabolites were identi ed searching against the database based on the m/z value and normalized migration time. Compound Discoverer Software (Thermo) was used to process the Thermo RAW les, and the data after editing were performed Multivariate Analysis using SIMCA-P 14.0 software (Umetrics AB, Umea, Sweden). Metabolites selected as biomarker candidates for further statistical analysis were identi ed on the basis of variable importance in the projection (VIP) threshold of 1 from the sevenfold cross-validated OPLS-DA model, which was validated at a univariate level with adjusted P < 0.05. And the MetaboAnalyst (version 3.0) was used for the identi cation of metabolic pathways [50].
Targeted metabolomics. 50 mg dried cecal contents and 800 μl methanol were added to a 1.5 mL Eppendorf tube. And the sample was grinded to ne powder using grinding mill at 65 HZ for 90 s followed by being vortexed for 30 sec, and centrifuged at 12,000 rpm, 4 °C for 15 min. Next, the 200 μl of supernatant was taken for detection.
For quantitative detection of linoleic acid and CLA, 1 μl of each sample was injected onto a DB-5 column (60 m x 0.25 mm 0.25 μm) using Thermo Trace 1300 GC (Thermo Fisher Scienti c, USA) system online with mass spectrometer (ISQ7000, Thermo Fisher Scienti c, USA) (GC-MS). The temperature program was as follows: the initial oven temperature 140 °C was hold for 5 min, then programmed to increase with 10 °C /min to 180 °C, with 4 °C/min to 210 °C, to reach nally with 10 °C /min 260 °C and hold for 20 min.
Helium (99.999% purity) was used as carrier gas with a ow rate of 1.5 ml/min. The MS inlet line and the ion source temperatures were maintained at 260 and 230 °C, respectively, and the MS ionization energy was 70 eV. A full scan mode set from 5 min to 20 min, monitoring m/z range from 33 to 550 Da, was used for the identi cation of possible interferences from the matrix extract.
For quantitative detection of 12,13-EpOME and 12,13-diHOME, 4 μl of each sample was injected onto an Acquity UPLC BEH C18 column (100 mm x 2.1 mm x 1.7 μm) using an Acquity UPLC (Waters Corporation, USA) coupled with triple quadrupole mass spectrometry (API5500, AB SCIEX LLC., USA) (UPLC-QqQ-MS). The mobile phase consisted of solution A and B (A was water and B was acetonitrile) and the ow rate was 300 μl/min. The gradient was set as follows: 10% B at 0 min, 10% B at 1.0 min, 90% B at 1.5 min, 90% B at 5.0 min, 10% B at 6.0 min, 10% B at 7.0 min. Samples were analyzed in negative ion modes using 550 °C of atomizing temperature, -4.5 KV of spray voltage, and MRM reaction monitoring of scanning method. The ow rates of curtain gas, collision gas, GS1 (atomizing gas) and GS2 (auxiliary gas) were 35,9,55, and 55 arb, respectively.
RNA isolation and RT-qPCR of the cytokines from tissue for expression analysis.
Total RNA from the ileum and liver tissue was extracted by using Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA) according to the manufacturer's instructions. The quality and concentration of RNA were measured using the Nanodrop 2000 spectrophotometer. 1 μg of total RNA from each sample was reverse transcribed into cDNA using the SuperScript II (Invitrogen Life Technologies, Carlsbad, USA). The primers used for the reverse transcription were oligo (dT) primer and random hexamers. The quantitative PCR reaction was performed with the SsoFast EvaGreen Supermix using a Bio-Rad CFX realtime PCR detection system following the manufacturer's protocols. Primers used in this study were listed in Supplementary Table 2 [51,52]. Relative mRNA expression levels of each target gene (ZO-1, Occludin, Claudin 3, MUC2, TFF2, IL-22, IL-17A, INF-α, CYP1A2, EPHX2 and GAPDH) were calculated using the log2 of the fold change method. The sample was run in triplicate parallel reactions in all samples.

Data and statistical analysis
The heatmap of the interrelationship between the differential ora and the metabolites was drawn using R (3.6.1) pheatmap package. The correlation coe cient calculated by heat map (R < 0.5) was used to exclude the metabolites and ora with weak correlation and no correlation, and Cytoscape (3.7.1) software was used to draw the correlation network diagram, which the ora and metabolites were used to form points, and the line segment represented the correlation size. The distribution of bacterial communities and their potential correlations with differential metabolites were determined using canonical correspondence analysis (CCA), using the R (3.6.1) vegan package. Statistical analyses were conducted with SPSS 20.0 (SPSS Inc., USA). The data collected are presented geometric median or mean ± standard deviation. Statistical signi cance was determined by Mann-Whitney test or One-way ANOVA. Mann-Whitney test was used for comparing two groups. One-way ANOVA with a Dunnett's multiple comparison test was used for pair-wise comparison of means from more than two groups in relation to the control group. The p values of less than 0.05 were considered statistically signi cant (*p < 0.05; **p < 0.01; ***p < 0.001).

Florfenicol exposure increases susceptibility to S. enteritidis infection
We established a study design (Fig. 1a) in which SPF chicks were fed either the FFC-treated (FT), the S. Enteritidis-infected (ST) or simultaneous treatment of S. Enteritidis and FFC (FST). Under the SPF environment, all chicks were cultured negative for Salmonella spp. until experimental infection with S. enteritidis and control group remained culture negative for Salmonella spp. throughout the study. The colonization and translocation of S. enteritidis in the intestinal tract of chicks directly determines the survival and pathogenicity of S. enteritidis. Therefore, we examined S. enteritidis levels in the caecum, spleen and liver and found that the antibiotic (FFC) could promote S. enteritidis colonization and translocation in the intestines of chicks. The number of S. enteritidis (log10 CFU/g tissue) signi cantly increased by 25.49% (Cecal contents, P < 0.01), 23.04% (Spleen, P < 0.01) and 21.33% (Liver, P < 0.01), respectively, in the FST group compared with those in the ST group at 3 days post-infection (dpi). Similar results were observed at day 18 (10 dpi) and day 25 (17 dpi), although their Salmonella loads are less than the 3 dpi (Fig. 1b). The above results indicate a robust in uence of antibiotic upon susceptibility to oral S. enteritidis infection in neonatal chicks.

Florfenicol administration aggravates S. enteritidis-induced morphology and intestinal barrier function injury
The FFC intervention made the chicks more susceptible to Salmonella infection. It is possible that antibiotics disrupted the immature intestinal barrier homeostasis of the chicks, thus changing the intestinal permeability and making the chicks carrying more Salmonella in their internal organs. Therefore, we investigated the effects of FFC administration on S. enteritidis-induced intestinal morphology injury. H&E staining showed that the NT group exhibited an intact structure of the ileal mucosa, neat intestinal villi, deep crypts, and a clear and complete gland structure, as also observed in the FT group (Additional le 1: Figure S1a, b). The ST group showed that the structure of the ileal mucosa was incomplete, villi had a shorter length and sparse distribution, and the crypts were shallow (Additional le 1: Figure S1c). However, the FST group increased loss of mucosal structures, and atrophic crypts as well as lamina propria bowel edema can be observed (Additional le 1: Figure S1d). The histological injury score (Additional le 1: Figure S1e) was assessed based on the H&E staining images, and the score showed quanti able results of tissue damage. The score of the chicks in the ST group (7.13 ± 0.44) was signi cantly higher than normal chicks (0.63 ± 0.18). Compared with the ST group, the FFC preadministration (10.75 ± 0.45) signi cantly increased the injury score of the ileum. We also utilized SEM to examine the intestinal structure in different groups. The results showed that the NT group had complete ileal villi, which formed full and closely arranged structures (Fig. 2e), the FT group also had intact ileal villi, but the arrangement structure was relatively loose (Fig. 2f). As expected, the ileal villi in the ST group were damaged (Fig. 2g), whereas those in FST group showed more severely (Fig. 2h). These results suggest that, although FFC has less effect on intestinal morphology, it can aggravate the intestinal morphological damage in the presence of Salmonella invasion.
The effect of FFC on intestinal barrier function changes in ileum after S. enteritidis infection were also examined. We found that FFC can exacerbate the S. enteritidis-induced Ileum permeability increase ( Fig.  3a-d). The serum DAO and LPS levels in the FT group were signi cantly (P < 0.001 and 0.05 respectively) higher than those in the NT group (Fig. 3c, d). In the case of Salmonella infection, serum D-lactate, DAO and LPS levels in both ST and FST group were signi cantly (P < 0.001) increased compared with that in NT group. However, FFC treatment signi cantly (P < 0.001) increased the serum D-lactate, DAO and LPS contents exposed to Salmonella infection ( Fig. 3b-d). Alcian blue staining indicated that FFC signi cantly (P < 0.05) decreased the acidic mucin of ileum compared with NT group, as evident by the quantitative evaluation of positive Alcian blue staining (Fig. 2a, b) using the integral optical density measurement (Fig. 3a). Similarly, FFC treatment also signi cantly decreased (P < 0.01) mean density of acidic mucin ( Fig. 2d; Fig. 3a) exposed to Salmonella infection. Transcriptional analysis of a range of relevant intestinal barrier genes was used to determine changes between these groups (Fig. 5). FFC treated signi cantly altered the gene transcription (Caludin1, IL-17A, IFN-α) in FT group. However, in the case of Salmonella infection, the FFC signi cantly reduced the expression of ZO-1, Occludin, Caludin1, MUC2 and TFF2, and signi cantly increased the expression of IL-17A, IL-22 and IFN-α. Furthermore, treatment with FFC signi cantly (P < 0.01) decreased SIgA secretion, but had no effect on serum IgG (Fig. 3e, f). Nevertheless, FFC treated reduced the SIgA secretion more seriously (P < 0.001) after Salmonella infection (Fig. 3f). These results indicate that FFC intervention, to some extent, increased the intestinal mucosal permeability of chicks, reduced mucosal immunity and signi cantly increased the degree of damage to the intestinal mucosal barrier after Salmonella infection.
The levels of Salmonella colonization are tightly interconnected with the trigger of intestinal in ammation [53]. This suggested that the increased colonization levels might be linked to increased mucosal in ammation. Indeed, FFC treated chicks featured higher levels of gut in ammation after Salmonella infection (Fig. 4). This was veri ed by quantitative analysis of proin ammatory cytokines and antiin ammatory cytokines in ileum tissue. Salmonella infection after FFC-pretreated signi cantly increased the level of the cytokines IL-1β (Fig. 4a), IL-6 ( Fig. 4b), IL-8 (Fig. 4c), TNF-α (Fig. 4e), and IFN-γ (Fig. 4f), whereas IL-10 ( Fig. 4d) was signi cantly decreased. Moreover, the in ammatory cytokines (IL-1β, IL-6, TNF-α and IFN-γ) were also signi cantly increased in the FT group. These results indicate that the FFC exacerbated the Salmonella-induced in ammatory response. It is possible that antibiotic-treated causes Gram-negative bacterium releases LPS, which promotes intestinal in ammation, and this can be con rmed by serum LPS levels. Moreover, FFC may also shifts the gut microbiota and metabolic pro ling of neonatal chicks, causing microbiotic and metabolic disorders and support Salmonella colonization.

Florfenicol administration alters the gut microbiota
The composition and density of the gut microbiota play an important role in combating Salmonella invasion, and oral pretreatment with antibiotics decreases colonization resistance and leads to an obviously post-antibiotic expansion of the Salmonella loading in the gut [53]. Thus, we hypothesized that the more Salmonella population observed in FFC-treated group might be linked to a disorder of the microbiota composition and density. To this end, comparative microbiota analysis of the cecal content of different groups at three different stages of infection (Day 3, 10 and 17 post-infection) was performed by 16S rRNA gene sequencing. Additional le 2: Figure S2 shows estimates of the diversity of the microbiota, presented as plots of the Shannon index, Observed Species and Pielou index measure of αdiversity. The α-diversity of the cecal microbiotas from chicks at 3 dpi was not neither affected by FFC treatment nor S. Enteritidis infection (Additional le 2: Figure S2a). However, a signi cantly decrease in alpha diversity was observed in FST group at 10 dpi (Additional le 2: Figure S2b). And Additional le 2: Figure S2c shows that the Shannon and Pielou index of the cecal microbial communities are signi cantly increase in the FST group at 17 dpi. These results indicated that the α-diversity of gut microbiota from chicks are not signi cant affected by a single FFC treatment or Salmonella challenge. However, the infection of Salmonella after pretreatment with antibiotics signi cantly disturbed the alpha diversity of chicks.
The results of phylum and genus distributions of microbial composition are shown in Additional le 3: Figure S3 and Additional le 4: Figure S4, respectively. Firmicutes (71.40 -99.62%) dominated the chicks gut microbiota in the four groups at three different stages of infection (Additional le 3: Figure S3). At 3-, 10-and 17-days post infection, the FST group had the highest relative abundance of Proteobacteria (2.68%, 1.30%, and 1.32%, respectively) compared with other three groups (Additional le 3: Figure S3). And at 17-days post infection, compared with the NT group (27.40%), the FFC (7.76%) signi cantly reduced the relative abundance of Bacteroidetes, and Salmonella infection (22.27%) has less effect on Bacteroidetes. However, the infection of Salmonella after pretreatment with FFC almost limits the growth of Bacteroidetes (0.01%) (Additional le 4: Figure S4). We further applied the LEfSe method to identify speci cally abundant bacterial taxa among these groups (only those taxa that obtained a log linear discriminant analysis [LDA] scores > 3 were ultimately considered). A cladogram from phylum to genus level abundance is shown in Fig. 6. In total, 21, 21, and 28 differentially abundant bacterial taxa were identi ed at three different stages of infection, respectively (Fig. 6). In the non-treated chicks, LEfSe highlights the greater differential abundance of Lactobacillus at 3 and 10 dpi, and Bacteroides at 17 dpi. Notably, the relative abundance of Enterobacteriaceae was signi cantly higher in the FST group compared with other three groups at all three different points. However, the other taxa were changed irregularly at different times in different groups. Moreover, the relative abundance of these biomarkers was showed in Additional le 5: Figure S5, and consistent results are obtained. We also established taxonomic cladogram at 11 day (3 dpi), and the relative abundance of taxa node in each group was showed in the form of a pie chart (only those taxa that the relative abundance > 0.1% were ultimately considered) (Fig. 7a). Similarly, the abundance ratio of Lactobacillus in the control group was signi cantly higher than the other three groups. Additionally, the abundance ratio of Enterobacteriaceae in the FST group was dominated among these four groups. Furthermore, at the genus level, Salmonella was only found in the challenged groups, and the abundance ratio of Salmonella in the FFC pretreatment group was signi cantly higher than that in the unpretreated group (Fig. 7a). Additionally, we determined the cecal loads of these biomarkers and two intestinal protective bacteria by quantitative PCR (qPCR) (Fig. 7b). At 11 day (3 dpi), the FFC pre-treatment signi cantly reduced the densities of total bacteria, Lactobacillus, clostridium butyricum and faecalibacterium prausnitzii. Although single Salmonella infection had no effect on densities in the cecal contents, Salmonella infection after pretreatment with FFC group harbored much higher densities of Enterobacteriaceae, and lower densities of Lactobacillus, Bacteroides, clostridium butyricum and faecalibacterium prausnitzii than the control group. At 25 day (17 dpi), the clostridium butyricum and faecalibacterium prausnitzii were present at equivalent densities in the cecal contents of four groups. However, the signi cant differences in the bacterial densities of total bacteria, Lactobacillus, Bacteroides and Enterobacteriaceae were still apparent between NT and FST group or ST and FST group (Fig. 7b). The Lactobacillus and Bacteroides are generally considered as the bene cial bacteria that provide protection for the gut, whereas the Enterobacteriaceae was known to be the potential pathogens of poultry and/or humans. These observations suggest that FFC exposure signi cantly decreased the abundance of Lactobacillus in chicks, and this inhibitory effect may provide a growth advantage for Enterobacteriaceae, especially Salmonella, in the gut of chicks.
The similarity of microbial communities (β-diversity) was visualized through PCoA of Bray-Curtis distances. At 3 dpi. The PCoA plots showed that microbial communities from Salmonella or FFC treated chicks clearly separate from those of the non-treated chicks. The rst axis of the PCoA explained 19.0% of the variation in bacterial diversity while the second axis explained 13.0% (Fig. 7C). The rst axis can roughly distinguish the antibiotic pre-treated chicks and non-pretreated chicks, and second axis can roughly distinguish the Salmonella infected birds and non-infected birds. The PCoA at 10 dpi showed that the microbiota composition was very similar between the NT and FT chicks, whereas the ST and FST group are still obviously distinguish from the NT group (Additional le 6: Figure S6a). Intriguingly, at 25 day (17 dpi), the PCoA demonstrated that both the microbiota composition of ST and FT group is tended to the NT group, whereas microbiota composition of FST group is still a striking divergence from the NT group (Additional le 6: Figure S6b). These ndings suggest that a single FFC or Salmonella treatment can cause some changes in microbiota composition of chicks, and they would generally recover after two weeks. Whereas FFC pretreatment hindered the recovery from microbiota composition of chicks for Salmonella infection.

Florfenicol administration alters the metabolic pro ling
We hypothesized that differences in key metabolites may be crucial to the effect of Salmonella colonization on chicks. Therefore, we conducted metabolomic analysis by LC-MS to determine the differential levels of metabolites on day 11 (3 dpi) in cecal contents of different groups. The principalcoordinate analysis (PCA) score plot showed that the metabolome of NT group and ST group was signi cantly separated among four groups, whereas there was no clear distinction in cecal metabolites between the FT and FST group (Fig. 8). For further analysis, orthogonal projections to latent structuresdiscriminate analysis (OPLS-DA) and permutation test plot of OPLS-DA was carried out to explore the differences between these groups. As shown in Additional le 7: Figure S7, the OPLS-DA showed that the cecal metabolites of the NT group were clearly distinguished from those of the FT group (Additional le 7: Figure S7a), ST group (Additional le 7: Figure S7c), and FST group (Additional le 7: Figure S7e). In addition, there was also a clear separation between the FST group and ST group in cecal metabolites (Additional le 7: Figure S7g).
From the OPLS-DA models, we identi ed 72 differential metabolites between NT and FT group, 42 differential metabolites between NT and ST group, 69 differential metabolites between NT and FST group, and 57 differential metabolites between FST and ST group according to the threshold (VIP > 1, and p < 0.05; Welch's t test). The signi cantly differential metabolites of these groups are shown in Supplementary Table 3. We next performed the pathway enrichment analysis based on these differential metabolites to comprehensively understand the effect of FFC on metabolism of chicks (Fig. 9). Linoleic acid metabolism, aminoacyl-tRNA biosynthesis, lysine biosynthesis, phenylalanine metabolism and lysine degradation were enriched after FFC treated (Fig. 9a). Arginine and proline metabolism, lysine biosynthesis, lysine degradation and D-glutamine and D-glutamate metabolism were enriched after Salmonella infected (Fig. 9b). Linoleic acid metabolism, aminoacyl-tRNA biosynthesis, lysine biosynthesis, butanoate metabolism and phenylalanine metabolism were enriched in Salmonella infection after FFC pretreatment group (Fig. 9c). Linoleic acid metabolism was enriched between the FST and ST group (Fig. 9d). The above results indicate that the linoleic acid metabolism is the most remarkable metabolism pathway in FFC treated group with or without Salmonella challenge. We next mapped the metabolic pathway of linoleic acid based on identi ed differential metabolites, and the relative amount (mean ± SD) of these metabolites in four groups was also showed (Fig. 9e). The metabolites that affect the metabolic pathways of linoleic acid mainly include linoleic acid, 12,13-EpOME and 12,13-diHOME, and the relative amount of these metabolites in the FT and FST group was signi cantly higher than the NT and ST group. Notably, the relative levels of 12,13-EpOME and 12,13-diHOME were signi cantly high in the FFC-pretreated group, but almost none in the non-pretreated group (Fig. 9e).

Correlation between the differential gut microbiota and metabolites
After nding marked differences in the content of metabolites as well as the microbial composition after FFC-pretreated, we analyzed whether there were any speci c correlations between the microbial taxa and key metabolites. The Spearman correlation analysis revealed an association between four bacterial genera with nine discriminant metabolites in FFC-pretreated chicks (Fig. 10a). Enterobacteriaceae is a taxon with strong correlation, particularly with linoleic acid, 12,13-EpOME, 12,13-diHOME and L-tyrosine (positive correlations), while only L-ascorbic acid negatively correlated. Furthermore, the Clostridium positively correlated with L-palmitoylcarnitine, linoleic acid, 12,13-diHOME and L-tyrosine, while the taxon negatively correlated with L-ascorbic acid, anandamide and 4-pyridoxic acid. The Lactobacillus genus negatively correlated with L-palmitoylcarnitine, linoleic acid, 12,13-EpOME, 12,13-diHOME and L-tyrosine, and positively correlated with L-ascorbic acid. Lastly, a less strong positive correlation was detected between the Ruminococcus and 4-pyridoxic acid and gamma-aminobutyric acid. The CCA test showed that the Enterobacteriaceae was the most important bacterial factor in uencing the linoleic acid metabolism (including linoleic acid, 12,13-EpOME, 12,13-diHOME) after FFC-pretreated (Fig. 10b).
Additionally, the correlation network between differential bacterial taxa and metabolites is consists of 13 nodes and 22 edges. And the results also showed that the metabolic pathway of linoleic acid has a strong positive correlation with Enterobacteriaceae, while Lactobacillus negatively correlated (Fig. 10c).
Since linoleic acid can be produced by Lactobacillus into CLA [54], and in our study, there is a signi cant negative correlation between the linoleic acid and Lactobacillus. Therefore, we hypothesized that the non-FFC pretreated chicks (more abundance of Lactobacillus) may have more CLA contents. However, CLA is an isomer of linoleic acid, and the use of untargeted metabolomics detection cannot distinguish these substances, so we using the targeted LC-MS to detect these substances including linoleic acid, 9c,11t-CLA, 10t,11c-CLA, 12,13-EpOME and 12,13-diHOME (Fig. 10d). In line with results of the metabolic pro ling, the contents of linoleic acid, 12,13-EpOME and 12,13-diHOME were higher in the FFC-pretreated groups. Moreover, we con rmed more CLAs concentrations in the cecal contents of non-FFC pretreated chicks, and the 9c,11t-CLA level is signi cantly higher than the10t,11c-CLA (Fig. 10d). Furthermore, the Spearman correlation analysis showed a strong association between the abundance of lactobacillus and CLA concentrations (Additional le 8: Figure S8). Collectively, these ndings hinted that 12,13-EpOME and 12,13-diHOME may be the key metabolites for prolonging gut colonization of Salmonella, whereas the CLA may limit the Salmonella growth during infection.
Conjugated linoleic acid Attenuates, yet 12,13-diHOME promotes, the S. enteritidis Colonization To address if CLA and 12,13-diHOME affect the Salmonella colonization more directly, we preadministered these compounds to newly hatched chicks before infected with S. Enteritidis (Fig. 11a). By day 3 post-infection, the Salmonella loads in the caecum, spleen and liver were signi cantly reduced in the chicks pretreated with CLA, whereas those were signi cantly increased in the chicks pretreated with 12,13-diHOME (Fig. 11b). Consistent with the fecal Salmonella loads, the pre-treatment of CLA signi cantly reduced, whereas 12,13-diHOME signi cantly increased the enteropathy of chicks by 3 dpi (Additional le 10: Figure S10). Furthermore, CLA-pretreated chicks also exhibited decreases intestinal permeability (serum D-lactate, DAO and LPS levels), and decreases pro-in ammatory levels (IL-1β, IL-6, IL-8, TNF-α and IFN-γ), as well as a signi cant increase in IL-10 levels. Similarly, the 12,13-diHOMEpretreated chicks gets the opposite results (Fig. 12). We also compared the effect of these two metabolites on the expression of intestinal barrier function genes after Salmonella infection (Fig. 13). The results showed that the CLA signi cantly increased the expression of ZO-1 and Occludin, whereas the 12,13-diHOME signi cantly reduced the expression ZO-1, Occludin, Caludin1 and MUC2, and signi cantly increased the expression of IL-17A (Fig. 13). To evaluate whether orally administration CLA and 12,13-diHOME reach to the gut lumen, we quanti ed the concentrations of these substances in the cecal contents, and observed a signi cant increase in cecal contents levels compared with non-treated chicks (Additional le 11: Figure S11). Together, these results demonstrate that pretreated CLA can attenuates, yet 12,13-diHOME promotes the Salmonella colonization in the gut of neonatal chicks.

Discussion
The antibiotics administration can perturb the gut bacterial community, resulting in weaken the gut colonization resistant to the pathogens [22,45,53,55]. Yet, the mechanisms that promotes Salmonella outgrowth after antibiotic pretreatment in chicks remain poorly described. In this study, we investigated the effect of antibiotic (FFC) pre-administration on the intestinal Salmonella colonization of chicks and its mechanism through microbiome and metabolomics. Similar to the reported papers [45,55], our results indicated that FFC can signi cantly increase the loads and prolongs gut colonization of S. Enteritidis. And the abundance of Salmonella also signi cantly increased in the endogenous organs (liver, spleen) exposed to FFC pretreatment. Salmonella depends on the two pathogenicity islands (SPI1and SPI2) to enter the intestine and adheres to the surfaces of intestinal epithelial cells to subsequently arrive in the subepithelial tissue via a series of invasive pathological pathways [56]. In the current study, we found that FFC-pretreatment could exacerbate Salmonella-induced morphology and intestinal barrier function injury and increased the intestinal barrier permeability. The animal assay revealed that FFC could directly decrease SIgA concentration, mucous layer density and increase the concentrations of serum DAO and LPS. And the real-time PCR demonstrated that FFC directly signi cantly decrease the expression of claudin 1, and increase the expression of IL-17A and IFN-α. The SIgA re ects the intestinal immunity state, and is able to stabilize intestinal colonization by symbiotic microorganisms and confer resistance to future invasion by exogenous pathogens [57][58][59]. Studies have shown that the gut microbiota is the most important source of immune microbial stimulation, use of antibiotics can disrupt the delicate ecosystem of the neonatal microbiome, which may cause an impaired stimulation of SIgA, and a low IgA response in turn lead to a reduced mucosal barrier function [60-62]. Furthermore, FFC can also aggravates Salmonella-induced in ammation in ileum, including upregulated the secretion of IL-1β, IL-6, IL-8, INF-γ, TNF-α, and decreased the IL-10 concentration. Previous study reported that the intestinal in ammation provides a growth advantage for Salmonella [53,[63][64][65]. And our results showed that the FFC could directly increase the concentrations of these proin ammatory cytokines. Taken together, these ndings imply that FFC pretreatment impaired intestinal immunity, increased intestinal permeability and in ammation, as well as aggravated Salmonella-induced intestinal barrier function damage, which promoted Salmonella colonization in neonatal chicks.
Based on the gut microbiota plays an important role in combating Salmonella invasion and maintaining intestinal immunity [53,66], we explore the intestinal ora of neonatal chicks in different treated groups.
The present study showed that the Firmicutes dominated the gut microbiota of neonatal chicks at day 11 and 18, and the mature microbial communities of chickens (at day 25) were dominated by Firmicutes and Bacteroidetes. This nding was consistent to previous studies [67, 68]. However, FFC administration signi cantly decreased the abundance of Lactobacillus at day 11 and 18 (4 and 11 days post treated), and signi cantly decreased the Bacteroides at day 25 (18 days post treated). The Lactobacillus spp. are considered a probiotic and have been used in feed processing for decades because of their bene cial effects on immunity, growth, and intestinal colonization resistance of livestock [69][70][71]. For example, the Lactobacillus rhamnosus reduced the colonization of pathogenic Salmonella, Clostridium, and E. coli strains to the intestinal mucus of pig [72]. The Lactobacillus acidophilus can bind to cultured human intestinal cell lines and inhibit the cell invasion by enterovirulent bacteria including Salmonella Typhimurium [73]. And another study showed that the Lactobacillus plantarum exerts an antagonistic effect on pathogenic bacteria by increasing the content of SIgA [74]. In our study, the FFC treatment signi cantly decreased the abundance of Lactobacillus in chicks, which indicates that this genus may be the main target bacteria for FFC antimicrobial effects, and the similar observations have been found by others using FFC therapy on intestinal microbiota in chickens [75]. And this reduction may be responsible for the promotion of Salmonella colonization after FFC pre-treatment. Bacteroidetes is the dominant phylum in the mature microbiota of chickens [76], and also have some inhibit effects on the gut colonization of Salmonella. Miki et al. nd that Bacteroides spp. can accelerate Salmonella Typhimurium elimination from the intestinal lumen of mice by producing vitamin B6 [45]. And another study demonstrates that the Bacteroides species confer colonization resistance to Salmonella Typhimurium infection by producing the propionate, which directly limits Salmonella growth by disrupting intracellular pH homeostasis [77]. And our results showed that at day 25, FFC pretreatment signi cantly reduced the abundance of Bacteroidetes and had more Salmonella loads in the cecum compared with the nonpretreated group, suggesting that FFC may have delayed the maturation of chicken intestinal ora and hinder clearance of Salmonella. Furthermore, the Salmonella infection after FFC pre-treated chicks had the had the highest relative abundance of Proteobacteria, which was known to be the potential pathogens of poultry and/or humans. The recent study showed that the orfenicol preventive treatment of calves showed a 10-fold increase in facultative anaerobic Escherichia spp, which a signature of imbalanced microbiota [78]. And Sáenz et al. oral administration of orfenicol to the sh observed a shift in the gut microbiome towards well-known putative pathogens such as Salmonella, Plesiomonas, and Citrobacter [79]. Combined with our results, it was showed that the FFC administration could changes the overall structure of the gut microbiota and promotes the growth of Proteobacteria, especially Salmonella. Moreover, although the microbial community of chicken is complex and relative stable, and the restoration of the microbiota after antibiotics withdrawal could be expected [80, 81], our results indicate that using antibiotics at an early age of chickens could have a profound effect on microbial composition (Additional le 5: Figure S5). And the study also demonstrated that the maturation of intestinal microbiota was signi cantly retarded and eventually delayed by antibiotics intervention at early ages of chicks [82].
We next use metabolomics to analyze how may FFC affect Salmonella gut colonization. Our data reported here suggest that the linoleic acid metabolism is the most remarkable metabolism pathway for FFC pre-administration. And we identi ed that the linoleic acid, 12,13-EpOME and 12,13-diHOME are the most important substances affecting linoleic acid metabolic pathway, which contents of these metabolites are signi cantly higher after FFC pre-treatment. It's worth noting that the concentrations of 12,13-EpOME and 12,13-diHOME were signi cantly high in the FFC-pretreated group, but almost none in the non-pretreated group. Linoleic acid is rstly metabolized to 12,13-EpOME by cytochrome P450 (CYP) epoxygenases, and followed by hydrolysis catalyzed by soluble epoxide hydrolases (sEHs) to form the diols 12,13-diHOME [83]. They have multiple pathological features, such as decreasing post-ischemic cardiac recovery, participating in the vascular cognitive impairment, increasing skeletal muscle fatty acid uptake and impeding immune tolerance in asthma child [31,[84][85][86]. And the 12,13-diHOME produced by sEH hydrolysis of 12,13-EpOME showed stronger cytotoxicity [83, 87]. Our analysis of metabolic enzymes in the pathway found that FFC can signi cantly increase the expression of CYP1A2, whereas had no signi cant effect on sEH (Additional le 9: Figure S9). Besides sEH produced by the liver, a variety of gut bacteria can also produce [31]. Correlation analysis results showed that the concentration of 12,13-diHOME was signi cantly positively correlated with Enterobacteriaceae and Clostridium, so we suspected that the sEH may be produced by these bacteria in the gut. A recent study showed that the sEH and sEHproduced lipid metabolites induces intestinal barrier dysfunction, bacterial translocation and colonic in ammation in mice [88]. Therefore, we suggest that 12,13-diHOME may promotes the intestinal colonization of Salmonella. Then, we pretreated with 12,13-diHOME for neonatal chicks and found that it signi cantly increased Salmonella colonization. Our results also showed that the 12,13-diHOME pretreatment signi cantly increase the Salmonella-induced expression of intestinal proin ammatory cytokines, exacerbate morphology and intestinal barrier function injury and increased the intestinal barrier permeability. The intestinal in ammation, particularly that due to proin ammatory cytokines, can disrupt barrier function and lead to intestinal permeability, and promote the pathogens colonization [63,89,90].
Previous studies showed that the diHOMEs has the pro-in ammatory effect on vascular endothelial cells [91], lung [31] and peripheral nervous tissue [92], and our study indicate that the 12,13-diHOME also has a pro-in ammatory effect on intestinal epithelial cells. Moreover, the diHOMEs have also been shown to disrupt mitochondrial function, eliciting the mitochondrial permeability transition and causing cellular apoptosis [93,94], and this may be why 12,13-diHOME exacerbates intestinal barrier damage. Therefore, it suggests that 12,13-diHOME could contribute to Salmonella colonization in the intestinal of chicks, in partly by promoting intestinal in ammation and destroying the intestinal barrier function.
CLA is the second factor affecting the Salmonella gut colonization after FFC pre-administration. Our correlation analysis combined with targeted metabolomic found that the Lactobacillus and CLA showed a signi cant positive correlation, and FFC pretreatment both signi cantly reduced the abundance of Lactobacillus and CLA in the gut lumen. CLA can be formed from linoleic acid by Lactobacillus and can inhibit the growth of the pathogenic bacteria [95]. Therefore, we think the CLA may be another factor that affects Salmonella colonization after FFC pretreatment. We pretreated with CLA for neonatal chicks and found that it effectively reduced Salmonella colonization, accompanied by the increased the expression of tight junction proteins (ZO-1 and occludin), alleviated the Salmonella-induced intestinal in ammation, and intestinal barrier injury. And we believe that CLA reduces the Salmonella intestinal colonization including several aspects. Firstly, CLA treatments could signi cantly upregulated the concentration of tight junction proteins (ZO-1, occludin, E-cadherin 1 and claudin-3) and ameliorated epithelial apoptosis  [102]. And Tabashsum et at. also showed that the CLA produced by Lactobacillus inhibited the growth and survival of Salmonella by altering the relative expression of genes related to Salmonella virulence [103]. Thus, our results suggest that CLA can maintain intestinal integrity, reduce intestinal in ammation, and inhibit Salmonella growth to effectively reduce the gut colonization of Salmonella in chicks. In view of the fact that CLA can be produced by Lactobacillus, and our research shows that FFC treatment dramatically reduce the content of Lactobacillus and CLA. Therefore, FFC may reduce the production of CLA by inhibiting the Lactobacillus growth, thereby reducing the colonization resistance of neonatal chicks to Salmonella infection.

Conclusion
Taken together, this study indicates that FFC pre-treatment signi cantly increases gut susceptibility to S. Enteritidis, and signi cantly increases the Salmonella-induced in ammatory responses and intestinal barrier damage in neonatal chicks. The metagenomic and metabonomic analysis revealed that FFC pretreatment signi cantly reduced the content of Lactobacillus, and signi cantly affected the linoleic acid metabolism pathway, including signi cantly reducing the levels of CLA, and signi cantly increasing the abundance of 12,13-EpOME and 12,13-diHOME. And we found that the CLA can maintain intestinal integrity, reduce intestinal in ammation, and directly inhibit Salmonella growth to effectively reduce the Salmonella colonization in chicks. Whereas the 12,13-diHOME though promoting intestinal in ammation and destroying the intestinal barrier function to support the Salmonella colonization. These ndings suggest that the decreased levels of Lactobacillus and CLA, and elevated cecal 12,13-diHOME concentrations in FFC pre-treated neonatal chicks might be an intestinal health-impairing attribute and may contribute to Salmonella colonization.

Ethics approval
All experiments in this study were reviewed and approved by the Institutional Animal Care and Use Committee at the Sichuan University.

Competing interests
The authors declare that they have no competing interests.     The gene expression pro le in response to FFC pretreatment or S. Enteritidis infection by qRT-PCR at 3dpi. Represented as log2 of the fold change between the treatment group and the control group (NT).
Statistical analysis was conducted using one-way ANOVA and Dunnett's multiple comparison test and denoted as follows: *P < 0.05, **P < 0.01, ***P < 0.     Correlation between key bacterial taxon and differential metabolites, and their absolute abundance. a Spearman correlation between differential metabolites and bacterial taxon was calculated for FT and NT group at 3 dpi (*P < 0.05, **P < 0.01, ***P < 0.001). Positive correlation was labelled in red and negative correlation was labelled in blue. b CCA plot of the differential metabolites and bacterial taxon for FT and NT group at 3 dpi. CCA ordination plot shows the correlations between bacterial community structures and metabolite factors. The correlations between the metabolite factors and key bacterial taxon are represented by the length and angle of the arrows. c Correlation network analysis of the key bacterial taxon and differential metabolites for FT and NT group at 3 dpi.  Effect of CLA or 12,13-diHOME pretreatment on S. Enteritidis infection in neonatal chicks. a Animal experiment design. Newly hatched chicks (n = 10) were randomly divided into four groups (NT, ST, CLA and 12,13-diHOME), which treated with a 7-day treatment of CLA or 12,13-diHOME and following infected with ~108 cfu of the challenge strain S. Enteritidis by oral gavage. Animals were euthanized, and the S.
Enteritidis loads in the cecal contents and organs were determined by plate counting method. b At 3 days Page 38/40 post infection, chicks were sacri ced and S. Enteritidis loads in the cecal content, spleen, and liver were determined. Bar indicates median. ns, not signi cant (p ≥ 0.05); *p < 0.05; **p < 0.01; ***p < 0.001; Mann-Whitney U test.

Figure 13
The gene expression pro le in response to CLA and 12,13-diHOME pretreatment and following S.
Enteritidis infection by qRT-PCR at 3dpi. Represented as log2 of the fold change between the treatment group and the control group (NT). Statistical analysis was conducted using one-way ANOVA and Dunnett's multiple comparison test and denoted as follows: *P < 0.05, **P < 0.01, ***P < 0.001; ns, not signi cant. NT: Control group. ST: S. Enteritidis infected group. CLA: CLA pretreated and following S. Enteritidis infection group. 12,13-diHOME: 12,13-diHOME pretreated and following S. Enteritidis infection group.

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