The Molecular Mechanism of Targeting the NF-κB Signaling Pathway to Modulate Mycoplasma Ovipneumoniae Infection in Sheep Based on Transcriptome Sequencing


 Background: MicroRNAs (miRNAs) play a key role in Mannan-binding lectin (MBL)-mediated resistance to Mycoplasma ovipneumoniae (MO) pneumonia (MP), by regulating the translation of mRNAs of target genes, thereby regulating the immune response. Sheep with different MBL genotypes have different plasma MBL levels, and show varied resistance to MP. Results: Using previously constructed high-throughput sequencing and transcriptome databases, miRNA-mRNA joint analysis showed that miR-509-5p is differentially expressed. miR-509-5p target genes were predicted using bioinformatic analysis, among which TRAF6 was selected. A TRAF6 dual luciferase reporter vector was co-transfected into 293T cells and sheep respiratory mucosal epithelial cells to detect and analyze changes in luciferase activity. qRT-PCR was used to analyze the effect of miR-509-5p on the expression and regulation of TRAF6 and other genes related to the NF-κB signaling pathway. miR-509-5p mimic and miR-509-5p mimic NC were transfected into susceptible sheep respiratory mucosal epithelial cells, for MO challenge infection experiments. The expression levels of cytokines, TRAF6 and other related genes in the NF-κB pathway were analyzed by ELISA and qRT-PCR. Conclusion: The results proved that miR-509-5p targets TRAF6 to affect the expression of downstream genes and cytokines, which negatively regulates the NF-κB pathway, thereby affecting the inflammatory response.

TargetScan and RNAhybrid2.2 software predicted the target gene of oar-miR-509-5p as TRAF6. We then analyzed binding site of oar-miR-509-5p in the TRAF6 3'-UTR. As shown in Figure 2A, after electrophoresis of the bacterial PCR ampli cation product, the position of the band was consistent with the expected size of the TRAF6 3'-UTR target fragment. The fragment was cloned into psiCHECK -2 and sequenced. A comparison between the sequence of the ligated product and the original sequence showed that they were consistent, indicating that the psiCHECK -2-TRAF6 reporter vector was successfully constructed.
Shanghai Biotech produced the mutated version of the TRAF6 3'-UTR lacking the oar-miR-509-5p binding site. This fragment was cloned into psiCHECK -2 and veri ed by double enzyme digestion with QuickCut TM Not I and QuickCut TM Xho I. Further PCR ampli cation showed that the fragment was in the correct position in psiCHECK -2, indicating that the psiCHECK -2-TRAF6-mut reporter vector was successfully constructed ( Figure 2B) and could be used in the subsequent experimental research.
According to Fig. 2C and D, sequencing of the inserts in both vectors, followed by alignment, further con rmed that reporter vectors psiCHECK -2-TRAF6 containing the oar-miR-509-5p binding site and the psiCHECK -2-TRAF6-mut lacking the miR-509-5p binding site had been successfully constructed.
293T cell culture transfection In this experiment, 293T cells were selected for transfection. The cells were in a good state before transfection, and the cells displayed their characteristic irregular shape (Fig. 3A). The cells transfected with the various constructs detailed in section 2.3 cells maintained their shape, which was similar to that of the cells before transfection, and their growth status was better (Fig. 3B).
Effect of oar-miR-509-5p on the expression of TRAF6 The successfully constructed wild-type psiCHECK -2-3'-UTR-TRAF6 and mutant psiCHECK -2-3'-UTR-TRAF6-M dual luciferase reporter vectors were transfected together with miRNA mimics and negative controls, respectively, into 293T cells. A BioTek microplate reader detected the results of the dual uorescence reporting system. The mimic negative control, psiCHECK -2-3'-UTR-TRAF6, and its mutant co-transformation group were used as negative controls. The results ( Fig. 4) showed that the cells transfected with wild-type plasmid and the miR-509-5p mimic expressed less luciferase activity compared with that of the control group (P < 0.01). However, the cells transfected with the mutant vector expressed similar amounts of luciferase to the control, regardless of whether they were co-transfected the miR-509-5p mimic or its NC. Based on these results, TRAF6 may be the target gene of miR-509-5p.

Morphological observation and immunohistochemical staining of mucosal epithelial cells from the sheep respiratory tract
According to the research method, primary cell culture was carried out to obtain two kinds of cells: Sheep airway mucosa epithelial cells and broblasts.
Microscopic examination revealed that the epithelial cells had no xed shape and are arranged like paving stones, which are easy to gather together, whereas the broblasts appear as spindles (Fig. 5A). After 48 h, the puri ed airway mucosal epithelial cells covered the bottom of the culture dish, as shown in Figure   5B. Immunohistochemistry showed that these cells were positive for keratin 11, identifying them as respiratory mucosal epithelial cells (dark brown cytoplasm and blue nuclei). The results showed that the cells cultured using the tissue block adherence method are the cells required for the experiment (Fig. 5C).
Effect of miR-509-5p on the NF-κB pathway related genes To further investigate the regulation of miRNA on the 3' UTR region of the target gene TRAF6 and pathway-related genes, qRT-PCR was used to detect their expression in sheep respiratory mucosa cells treated with the miRNA mimic, the mimic NC, the miRNA inhibitor, and inhibitor NC. Compared with that in the miR-509-5p mimic NC transfection group, TRAF6 expression was signi cantly downregulated in the miR-509-5p mimic transfection group (P < 0.01). The expression levels of other genes also varied, among which TLR4 (toll like receptor 4) and IRAK4 (interleukin 1 receptor associated kinase 4) expression levels were signi cantly upregulated (P < 0.05), TAK1 (TGF-beta-activated kinase 1) expression was signi cantly downregulated (P < 0.05), NFKB (NF-κB) expression was signi cantly downregulated (P < 0.01), and TIRAP (TIR domain containing adaptor protein) and TAB2 (TGF-beta activated kinase 1 (MAP3K7) binding protein 2) expression levels were not signi cantly different (Fig. 6) SNP analysis of MBL sequencing results Cells were isolated from susceptible "BB" type sheep and exon 1 of the MBP gene was sequenced. Analysis of the BB sequence compared with that of the MBP sequence deposited in the NCBI database (accession number fj97762) showed that the 105th base in the "BB" type sequence is T compared with C in the NCBI sequence (Fig. 7).
Detecting the cell death rate by PI staining The results of PI staining to calculate the cell death rate of each group are shown in Table 1. Compared with the control group, the cell death rate of the MO group increased by 32.47% (P < 0.05); while the death rate of the MO + MBL group increased by 6.6% (P > 0.05). The death rate in the MO group increased by 25.87% compared with that in the MO + MBL group (P < 0.05). Detection of IL-12, TNF-α, IL-6, IL-1β levels in cell culture uid When the cells are infected by MO, they produce an in ammatory response. According to Figure 8A, the levels of IL-12 in the two control groups were signi cantly different (P < 0.05), indicating that after MO infection, the expression of cytokine IL-12 increases, suggesting a severe immune reaction. Furthermore, the IL-12 level in blank control group was signi cantly increased compared with that in the miR-509-5p mimic group (P <0.01), and was signi cantly decreased compared with that in the miR-509-5p mimic NC group (P < 0.01), indicating that miR-509-5p was overexpressed. The IL-12 level in the experimental control group was signi cantly increased compared with that in the miR-509-5p mimic group (P < 0.05), and the level in the miR-509-5p mimic group was signi cantly decreased compared with that in the miR-509-5p mimic NC group (P < 0.01). The differences in IL-12 levels between miR-509-5p mimic NC groups were not signi cant. Figure 8B shows that when the cells were infected with MO without any miRNA treatment, the level of TNF-α in the cell culture supernatant was signi cantly increased. The levels of TNF-α in the blank control group and miR-509-5p mimic group were signi cantly different (P < 0.05); and the TNF-α levels in the miR-509-5p mimic group and miR-509-5p mimic NC group were signi cantly different (P < 0.01). There was no signi cant difference among the experimental control group, miR-509-5p mimic NC group, and blank control group. The expression levels of IL-6 in cell culture medium are shown in Figure 8C. The IL-6 levels of the blank control group and the miR-509-5p mimic group were signi cantly different (P < 0.01); the level in the experimental control group was signi cantly decreased compared with that in the miR-509-5p mimic group (P < 0.01). Compared with the IL-6 level in the miR-509-5p mimic NC group, the level in the miR-509-5p mimic group was statistically signi cantly decreased (P < 0.05). The trends of levels of IL-1β in cell the culture medium among the groups were roughly consistent with those of IL-12 and IL-6. Compared the IL-1β levels between the miR-509-5p mimic group and the miR-509-5p mimic NC group, showed a statistically signi cant difference (P < 0.05), while the IL-1β level in the blank group was not signi cantly different from that in the other three groups. The IL-1β level in the experimental control group was not signi cantly different from that in the miR-509-5p mimic group and miR-509-5p mimic NC group. (Fig. 8D)

Effect of MO on mRNA Expression of TRAF6-Related Genes
As shown in Figure 9, the expression levels of TRAF6 and NF-κB-related genes increased signi cantly after MO infection, while the expression levels of other genes were also signi cantly higher than in the blank control group. This showed that MO infection of respiratory tract mucosal epithelial cells could increase the expression of NF-κB signaling pathway related genes to varying degrees. Compared with that in the miR-509-5p mimic NC group, overexpression of miR-509-5p resulted in decreased expression of downstream genes of the pathway. Among them, the expression levels of TRAF6 and NF-κB genes were signi cantly reduced (P < 0.01), as was the expression level of TAK1 (P < 0.05). Thus, miR-509-5p might target TRAF6 to affect downstream gene expression. analysis on miRNA-mRNA data obtained from the spleens of Yorkshire pigs to obtain miRNAs and genes related to the immune response. The results showed their potential role in the difference in disease resistance between different pig breeds and stages, providing useful information for disease resistance breeding [19]. In humans, the molecular mechanism of male prostate cancer, the second most common cancer, was revealed by combining miRNA-mRNA expression pro les. The identi cation of key genes and valuable biological processes could be used to understand the molecular mechanism of treatments targeting prostate cancer [20]. Furthermore, using multi-omics data, the molecular mechanism of cervical cancer was explored, and key related genes and biological processes were obtained, which provided an important basis for understanding the molecular mechanism of this cancer [21]. The joint analysis of miRNA-mRNA can more accurately screen the key genes that regulate the trait of interest. Currently, joint analysis has become a popular and effective method to study the molecular mechanism of certain traits, however, this technique has not been applied sheep disease resistance breeding traits. Therefore, the present study was based on previously constructed high-throughput sequencing and transcriptome databases. Using miRNA-mRNA joint analysis in these databases, we predicted that TRAF6 might be the target gene of miR-509-5p. Next, we constructed wild-type and mutant TRAF6 3'-UTR region dual luciferase reporter gene vectors to verify the targeting relationship between miR-509-5p and TRAF6. Furthermore, combined with analysis of the expression of related genes in the upstream and downstream pathways of the in ammatory response and the changes of various immune factors, we speculated on the composition of the immunoregulatory pathway of MBL, and veri ed the role of regulatory genes in the mechanism of MBL-mediated anti-infection.

Discussion
Tumor necrosis factor receptor-related factor 6 (TRAF6) is a member of the TRAF family and is widely involved in the in ammatory process. In addition, TRAF6 is closely related to the activation of many in ammatory signaling pathways. Without TRAF6, many in ammatory signaling pathways cannot be activated [22]. Previous studies have shown that miR-146a directly regulates TRAF6 transcripts by binding to multiple sites in its 3' UTR to regulate immune function. Although miR-146a-dependent degradation of TRAF6 mRNA prevents the production of new TRAF6 protein, existing proteins can continue to function. Studies have shown that TRAF6 can affect the in ammatory response as part of the NF-κB signaling pathway [23,24]. miR-146a directly targets TRAF6 in peripheral blood mononuclear cells (PBMCs) of patients with type 2 diabetes, indicating that miR-146a might affect the in ammatory response by regulating TRAF6 [25]. Additionally, there are studies showing that TNFA (TNF-α) is the target gene of miRNA-509. Overexpression of miR-509 can increase apoptosis and inhibit TNF-α expression [26]. Furthermore, it has been reported that iron overload could reduce the expression of miR-146α and activate the expression of TRAF6, thereby promoting the expression of NF-κB-dependent in ammatory factors [27] and augmenting the occurrence of in ammation (Fig.  10). In this study, we used the Dual-Luciferase® Reporter Assay System and transformed cells with reporter vectors containing the wild-type TRAF6-3' UTR and a mutated version lacking the miR-509-5p binding site. Cotransfection of the reporter constructs with miR-509-5p mimics and NC controls showed that in the presence of the wild-type UTR and the miR-509-5p mimic, the luciferase activity was signi cantly lower than that of the NC control group. The luciferase activity in the cells with the mutated UTR reporter vector was similar between the miR-509-5p mimic group and its NC control. This con rmed that oar-miR-509-5p binds to 3' UTR of TRAF6 through base complementation, further con rming the direct targeting relationship between TRAF6 and miR-509-5p. To further study the regulatory effect of miR-509-5p on the target gene TRAF6 and pathway-related genes, qRT-PCR was used to detect mRNA expression levels of target gene TRAF6 and its signal transduction pathway related genes. Overexpression of miR-509-5p signi cantly reduced the mRNA expression level of the target gene TRAF6, whereas miR-509-5p inhibitors signi cantly increased the mRNA expression level of TRAF6. These results suggested that miR-509-5p might play a role in the in ammatory response by targeting TRAF6; however, the regulatory mechanism needs to be further studied to better understand the mechanism resistance to MO infection.
MO is a highly pathogenic microorganism that causes infectious pleura pneumonia in sheep and goats. It is widespread and causes serious economic losses to the sheep industry. MO attaches to and damages ciliary cells, which can lead to dry cough, shortness of breath, wheezing, and progressive weight loss. When these symptoms become irreversible, they can cause death of the infected sheep [27,28]. MBL is an acute-phase protein synthesized and secreted by the liver. It belongs to the protein family of collagen lectins (collectins) in the C-type lectin superfamily and is widely found in liver and blood. MBL plays a leading role in the initial stage of anti-infection by activating the complement system and regulating phagocytosis. MBL is an important natural immune defense molecule and is important for improving the body's immune activity and resisting the invasion of foreign pathogens [29,30]. Tabona  The results showed that a low MBL concentration is related to Mycoplasma ovipneumoniae infection, and the levels of tumor necrosis factor-α, interferon-γ, complement C1 and C3 among the four genotypes were also determined, which showed that the MBL A and B types conferred susceptibility, and the C type conferred resistance [33]. In our previous research, the expression levels of various immune factors in different MBL genotypes infected with MO were detected at the individual level [3].
Highly signi cantly, with the gradual decrease of in ammation, the level of immune factors might be related to the severity of the disease, which is consistent with the results of the present study. After MO infection of susceptible respiratory mucosal epithelial cells, the levels of IL-12, IL-6, IL-1β, and TNF-α increased to varying degrees compared with those in the blank control group. Studies have shown that diethylnitrosamine (DEN) exposure promotes the production of IL-6 in Kupffer cells (KCs) in a manner dependent on the Toll-like receptor adaptor protein, MyD88, and its ablation also protects male mice from DEN-induced hepatocarcinogenesis [34]. Yang et al. Found that IL-1β can induce the production of in ammatory mediators in HeLa epithelial cell lines, thereby regulating the MAPK signaling pathway [35]. It has been reported that IL-8 plays a key role in the pathogenesis of upper respiratory tract asthma caused by allergies. In addition, IL-8 is expressed in bronchial epithelial cells. Research also indicated that MBL activation of extracellular regulated kinase (ERK) might be one of the mechanisms leading to increased production of IL-8 [36]. Many studies have shown that the promotion cytokines expression and the release of in ammatory factors is an important part of activating the body's in ammatory response [37]. Overexpression and inhibition of miRNA function showed that miRNA-146a could negatively regulate the production of in ammatory factors IL-8 and RANTES [37]. Rossato [38] identi ed miR-187 as an IL-10-dependent miRNA, which plays a role in IL-10-mediated inhibition of TNF-α, IL-6, and IL-12 p40 subunit (IL-12p40). When miR-187 expression was silenced, LPS-induced TNF-α, IL-6, and IL-12p40 production increased signi cantly. The results of these studies are consistent with the ndings of the present study. MO infection of respiratory mucosal epithelial cells overexpressing miR-509-5p suppressed the expression of cytokines IL-12, TNF-α, IL-6, and IL-1β. Thus, we hypothesized that miR-509-5p might regulate the disease resistance to Mycoplasma ovipneumoniae by targeting the NF-κB signaling pathway. Based on the above test results, a regulatory pathway diagram was drawn ( Figure 11).
However, this study only studied from mRNA level of the genes, and did not analyze or discuss the mechanism at the protein level. In addition, in the study of the molecular mechanism of different resistance genotypes of MBL-mediated MO resistance, hepatocytes secreting the MBL protein were not selected for veri cation experiments; however, the regulatory mechanism in respiratory tract mucosal epithelial cells of the susceptible type (MBL) was discussed. The main reason for choosing these cells is that the respiratory tract mucosal epithelial cells are the host cells infected by Mycoplasma ovipneumoniae. It is not clear whether MBL regulates the molecular mechanism of disease resistance through miR-509-5p. MBL is an innate immunomodulator and the immune regulation mechanism is extremely complex; therefore, further experimental research is needed to determine the molecular regulatory mechanism of MBLmediated disease resistance.
The results of our study revealed that there is a true targeting relationship between TRAF6 and miR-509-5p. In addition, when MO infected susceptible respiratory mucosal epithelial cells, the expression levels of genes and related cytokines in the NF-κB signaling pathway increased to varying degrees. Furthermore, overexpression of miR-509-5p resulted the suppression of the expression of downstream genes of the TRAF6 and NF-κB signaling pathways and reduced cytokine IL-12 levels. This suggested that miR-509-5p targeting of TRAF6 affects the expression of downstream genes, thereby achieving disease resistance by targeting the NF-κB signaling pathway to regulate Mycoplasma ovipneumoniae infection, and provides a basis for further study of the molecular regulation mechanism of MBL-mediated disease resistance.

Samlpes and cells
This study was approved by the Ethics Committee of Animal Experiments from the Animal Science and Technology College at Shihezi University. All samples were collected in strict accordance with the committee's guidelines. Ninety-day sheep fetus tracheas were obtained from a live animal slaughtering center in Shihezi, Xinjiang province. The specimens were collected by trained personnel and specialized veterinarians following the animal welfare protocols of the First A liated Hospital of the Medical College of Shihezi University and College of animal science and technology. During the experiment, every effort was made to minimize animal suffering.

miRNA-mRNA joint analysis and prediction of target genes
This analysis was based on a high-throughput sequencing database and transcriptome sequencing database established earlier in this project (Unpublished). First, according to the fold-change, P-value, and Q-Value of the high-throughput sequencing database, it was determined that mir-509-5p was differentially expressed. miRNA-mRNA joint analysis also determined that mir-509-5p was a differentially expressed miRNA. TargetScan and RNAhybrid (USA) were used to predict the target genes of the differentially expressed oar-miR-509-5p (Ovis aries miR-509-5p), and the intersection of the results of the two analyses was used as the candidate target genes. Then, combined with sequencing data, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and reference literature for target gene screening, the candidate target genes were identi ed for subsequent experimental studies.

Identi cation and construction of target sequence wild-type and mutant psiCHECK TM -2 vector
Construction of target sequence wild-type psiCHECK TM -2 vector The 3'-untranslated region (UTR) of one of the candidate target genes, TRAF6 (encoding TNF receptor associated factor 6) was predicted and used to design primers to amplify the wild-type sequence. The primers were designed using Primer 6.0, and added sites for the Not I and Xho I restriction endonucleases to the ends of the amplicon. The 5 end bases were protected and the primers were synthesized by Xinjiang Kuntai Rui Company ( Table 2). The wild-type 3'-UTR sequence was then cloned into vector psiCHECK -2 (Promega, USA). Construction of mutant psiCHECK TM -2 vector Next, the miRNA binding site in the TRAF6 3'-UTR sequence was mutated. The following sequence was designed and cloned into the psiCHECK -2 vector, which was completed by Shanghai Biotech (China):

CTCGAGCTTTCACTTTTCACTTTCATGCATTGGAGAAGGAAATGGCAACCCACTCCAGTGTTCTTGGCTGGAGAATCCCAGGGACGGGGGAGCCTGGTGGGCTGCCATC
The sequence in bold is the mutated miRNA binding site in the TRAF6 3'-UTR sequence.

Double Digestion of psiCHECK TM -2 Vector
The psiCHECK -2 vector contains two reporter genes, the re y luciferase gene (hluc+) and the Renilla luciferase gene (hluc). In practical use, the Renilla luciferase value is used as an internal reference, and the re y luciferase activity is used as the reporter. 293T cell resuscitation, passage, and transfection The human embryonic kidney cell line 293T was donated by the College of Animal Science and Technology of Shihezi University. The frozen cells were thawed in a 37 °C water bath. Then, after sterilizing with 75% ethanol, the supernatant liquid was discarded, and the appropriate complete medium (Dulbecco's modi ed Eagle's medium (DMEM): Serum: double antibody=90%:10%:1%) was added to the cells in the cryovial tube. The cells were cultured at 37 °C in 5% CO 2 . When the cells reached 70-80% density, the cell culture uid is discarded and the cells were digested with a 0.25% pancreatin solution (Takara, Dalian, China) for 30-60 s. The state of the cells was observed under a microscope. When most of the cells had shed and dispersed, complete culture solution was added to stop digestion. The cells were transferred to a 15 mL centrifuge tube and centrifugedg. The supernatant was discarded and the cells were resuspended in complete medium. Finally, the cells were added to a 6-well culture dish.
To obtain an effective transfection e ciency, the optimal ratio of transfection was determined as 1:2 via pre-experimental plasmid transfection. As observed under an inverted uorescence microscope, transfection was performed when the cell density reached 70-80%. Then, according to the miRNA product manual (Ruibo Biotechnology Co., Ltd., China), miRNA mimics and inhibitors were formulated as a 20 µM stock solution for transfection experiments. The constructs were transfected into ve groups of cells: 1. Control group; 2. Wild-type psiCHECK -2 recombinant plasmid + miRNA mimic; 3. Wild-type psiCHECK -2 recombinant plasmid + miRNA Negative Control (NC); 4. mutant psiCHECK -2 recombinant plasmid + miRNA mimic; 5. mutant psiCHECK -2 recombinant plasmid + NC miRNA. Three parallel wells were set for each group and transfection was performed according to the instructions of the Lipo2000 reagent.
Luciferase activity assay The Dual-Luciferase® Reporter Assay System kit (Tiangen, China) was used to assess luciferase activity. At 24 h after transfection, the culture solution was discarded and the cells were rinsed two to three times with pre-warmed 1 × phosphate-buffered saline (PBS). Then, cell lysis solution (Takara) was added to fully lyse the cells. The lysed cells were transferred to a clean Eppendorf tube, centrifuged, and 20 µL of the cell lysate supernatant was added to a Lockwell maxisorp detection plate, followed by 100 µL of Luciferase Reaction Reagent. After shaking and mixing, the plate was placed immediately into the chemiluminescence instrument (Zeiss, Oberkochen, Germany) to detect the re y luciferase activity value. After detecting the re y luciferase activity value, 100 µL of Luciferase Reaction Reagent II was added, and after shaking and mixing, the plate was placed into the chemiluminescence instrument to measure the activity of Renilla luciferase.
Primary culture and puri cation of sheep respiratory epithelial cells Ninety-day-old fetuses were obtained from beef and sheep abattoirs in Shihezi City (Xinjiang Uygur Autonomous Region, China) and their tracheas were collected rapidly. The tracheas were rinsed with PBS buffer two to three times, cut into small pieces and spread on the bottom of a Petri dish. Subsequently, culture medium (DMEM: Serum: double antibody=90%:10%:1%) was added and the plates were placed at 37 °C in 5% CO 2 for culture; the medium was changed every 48 h. After 3-4 days of culture, the mucosal epithelial cells of sheep respiratory tract were puri ed. First, the tissue block was removed from the culture medium and rinsed with PBS two to three times. The PBS was discarded and 2 mL of trypsin (0.25%) without EDTA was added to each petri dish, and the tissue was digested at 37 °C, 5% CO 2 for 2 min. Culture solution was added to stop the reaction. At this point, most of the broblasts had adhered to the wall, while the respiratory mucosal epithelial cells did not adhere. By repeating this method multiple times, puri ed airway mucosal epithelial cells could be obtained.

Cellular immunohistochemistry
The cell slides were xed in 4% paraformaldehyde solution, dried, incubated with 3% H 2 O 2 at room temperature for 10 min, and washed three times with PBS.
The slides were then incubated with 0.3% Triton X-100 for 20 min, and then with PBS-diluted goat serum blocking solution for 40 min. Primary antibodies recognizing keratin 11 (Anti-pan Cytokeratin antibody [C-11], Abcam, Cambridge, MA, USA) were added, and the slides were incubated at 37 °C for 2 h, and washed three times with PBS. Then, an equal amount of horseradish peroxidase-labeled secondary antibody was added, incubated at 37 °C for 60 min, and washed three times with PBS. Finally, the cells on the slides were stained using a 3,3'-Diaminobenzidine (DAB) staining kit. The cells were then counterstained with hematoxylin for 30 s to 1 min at room temperature for 1 min, and washed with PBS for 10 s. The slides were sealed with gum and examined under a microscope.
Measuring target genes and pathway genes mRNA expression Total RNA was extracted from cells and isolated using the TRIzol reagent (Invitrogen). One microgram of total RNA was used to perform reverse transcription using a PrimeScript RT Reagent Kit with gDNA Eraser (Takara) according to the manufacturer's instructions. The primers for all target genes are shown in  The measured bacterial count was colony forming units (CFU) = 10 6 . The instructions of the PI staining test kit were followed to complete the staining process. In the dark, 20 μL of the cell-dye mixture was dropped onto the loading chip, and the chip was inserted into the Adam-mc counter for detection and calculation of cell mortality.

Statistical analysis
Each experiment was performed 3 times. The data were expressed as the mean ± the standard error and all statistical analysis were performed with SPSS software version 19.0 (IBM Corp., Armonk, NY, USA). Differences between means were accepted as statistically signi cant at P < 0.05 and as a trend at 0.05 < P < 0.10. Availability of data and materials The data used and/or analyzed in the current study are available from the corresponding author on reasonable request.

Competing Interests
We certify that there is no competing of interest with any nancial organization regarding the material discussed in the manuscript.

Funding
This research was supported by grants from the National Natural Science Foundation of China (No: 31460587). All authors participated in writing the nal draft of the manuscript and agreed on the nal format.
Authors` Contributions MZ, HZ and ZZ designed the study. MZ and YL performed the animal inoculation trial. MZ performed the laboratory procedures and bioinformatics analysis.
MZ, PW, and YN analyzed and interpreted the data. MZ drafted the manuscript. MZ, HZ and ZZ critically revised the manuscript. All author(s) have read and approved the manuscript. Results of the comparison of PCR product sequencing