A total of 70 day-3 C. jejuni-free specific pathogen-free (SPF) White Leghorn chickens (Jinan SAIS Poultry Co., Ltd, China) were used in the current study. The C. jejuni (ATCC 33291) strain was obtained from the China Center of Industrial Culture Collection (CICC). Chickens were raised in sterilized isolators with free access to feed and water. Each chicken was orally inoculated with 0.5 mL of C. jejuni solution (1.68 × 10⁸ CFU/mL). All procedures were performed under strict sterilization conditions. At 8 hours post-inoculation with C. jejuni, the venous blood, liver, cecal content, and cecum were collected from each individual, and immediately frozen in liquid nitrogen.

To quantify the C. jejuni levels, 0.1 g cecal content from each chicken was collected and serially diluted with sterile PBS. Subsequently, the cecum content was cultured on Columbia Blood Agar Base plates (Sigma, USA) under microaerophilic conditions (42°C, 85% N₂, 10% CO₂, and 5% O₂) for 48 hours. Chicken with over 1.86×10¹³ CFU/mL C. jejuni in cecal content was classified as the susceptible group (S group), and chicken with below 4.36×10¹⁰ CFU/mL was assigned to the resistant group (R group). All experimental protocols were approved by the Ethics Committee on the Care and Use of Laboratory Animals at Shandong Agricultural University (Approval Number: SDAUA-2019-060).

Approximately 1–2 mm³ liver tissue from each chicken was fixed in a 2.5% glutaraldehyde and washed with 0.1 M phosphate buffer for three times. After post-fixation with 1% osmium tetroxide in 0.1 M phosphate buffer, tissues were dehydrated through a graded ethanol series (30%, 50%, 70%, 90%, and 100%) for 10 minutes per concentration. The tissue was embedded in epoxy resin (Epon 812, Epon, USA), and polymerized at 60°C for 48 hours. Ultrathin sections (60–90 nm) were prepared and examined using a transmission electron microscope (TEM) (Hitachi, Japan) at a magnifications of 80,000x. ImageJ 1.8.0 was used to quantify the mitochondrial length, width, area, and the number of mitochondrial cristae.

To collect the serum, the venous blood was collected from each chicken in the R and S groups, and centrifuged at 3000 × g for 10 minutes at 4°C. IgA (ml002792), IL-6 (ml059839), IL-18 (ml042769), IL-1β (ml002787), IL-17A (ml023404), and TNF-α (ml002790) ELISA kits (MLBIO, Shanghai, China) were used to determine the serum cytokines’ levels according to the manufacturer’s instructions, respectively.

Total RNA was isolated from cecum sample using the Total RNA Kit I (Omega, Hunt Valley, USA) according to the manufacturer’s instructions. The integrity and concentration of RNA were measured using Agilent 2100 Bioanalyzer (Agilent, California, USA) and Nanodrop 2000 (Nanodrop, Wilmington, DE), respectively. Total RNA was purified using the Dynabeads Oligo (dT) (Thermo Fisher Scientific, Massachusetts, USA), and fragmented into approximately 100 nucleotides. The fragmented RNA of each individual was clustered into two libraries: an immunoprecipitation (IP) library and an input library. For the IP library, the fragmented RNA was incubated with an m⁶A-specific antibody in IP buffer (50 mM Tris-HCl, 750 mM NaCl, and 0.5% Igepal CA-630) for 2 hours at 4°C. The IP and input RNA was then reverse-transcribed into cDNA using SuperScript II Reverse Transcriptase (Invitrogen, Waltham, USA), and second-strand synthesis was conducted with NEBNext^® Ultra™ II Directional RNA Library Prep Kit (New England Biolabs, USA). Finally, a total of six libraries in each group (3 replicates × (IP + input)) were constructed, and subjected to paired-end 150 bp sequencing using the Illumina NovaSeq 6000 platform (LC-Bio Technology Co., Ltd., Hangzhou, China).

The adapter sequences, duplicate reads, and low quality reads were filtered out from raw data with the default parameters using fastp (41). The clean reads were mapped to the Gallus gallus reference genome (GRCg7b) using HISAT2 (42). Peak calling analysis was performed with the R package ExomePeak (43). The candidate peak region within the genome was tested by the Poisson distribution model to assess the statistical significance of read enrichment, and the region with P value < 0.05 was considered a peak. The differentially methylated m⁶A peaks (DMPs) between resistant and susceptible groups were identified using Fisher’s test. The distribution of m⁶A peaks across functional elements (5′UTR, start codon, CDS, stop codon, and 3′UTR) was annotated using ANNOVAR (44). Subsequently, the identification and visualization of m⁶A motifs enriched within peak regions were performed using HOMER (45) and the BioSeqUtils package in R (46). The gene expression was quantified using StringTie (47) with default parameters. The differentially expressed genes (DEGs) between R group and S group was identified with DESeq2 package (48). The genes with |log₂Fold change| ≥ 1 and P value < 0.05 were considered DEGs. The DEG harboring at least one DMPs was defined as the differentially methylated gene (DMG). Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis for DMPs, DEGs and DMGs was performed using the LC-Bio OmicStudio platform (https://www.omicstudio.cn/home).

The total RNA from each cecum was denatured at 95°C for 3 minutes and cross-linked to an Immobilon-Ny+ Nylon Membrane Roll (Merck Millipore, Germany). The unbound RNA was washed with Tris-buffered saline containing 0.1% Tween 20 for 5 minutes. After blocking with 5% skimmed milk (BI, Germany) for 1 hour, the membrane was incubated with an anti-m⁶A antibody (1:250 dilution; ab286164, abcam) at 4°C overnight with gentle shaking. Subsequently, the membrane was incubated with an anti-mouse IgG secondary antibody (1:5,000 dilution; ab190475, Abcam) for 1 hour at room temperature. The m⁶A levels were visualized using a chemiluminescent substrate in a chemiluminescence imaging system (Fusion Fx Vilber Lourmat, France) and then quantified using ImageJ 1.8.0.

Following RNA extraction, poly(A)+ RNA was selectively purified using oligo(dT) magnetic beads (Thermo Fisher Scientific, Massachusetts, USA), and fragmented into approximately 100 nts using the Magnesium RNA Fragmentation Kit (New England Biolabs, USA) following the manufacturer’s instructions. The fragmented RNA was then subjected to immunoprecipitation with an m⁶A-specific antibody (Synaptic Systems, Göttingen, Germany) conjugated to Protein A/G magnetic beads (Invitrogen, USA). Both the immunoprecipitated RNA and input RNA were reverse transcribed into cDNA using the PrimeScript RT Reagent Kit (Takara, Dalian, China). qRT-PCR was performed using SYBR Premix Dimer Eraser (Takara, Dalian, China) and specific primers (Sangon, Shanghai, China) on Roche LightCycler^® 96 System ( Supplementary Table S1 ).

The chicken macrophage-like cell line (HD11) was provided by ShanghaiNulen Biotech. (Shanghai, China). HD11 were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Invitrogen, Carlsbad, CA), supplemented with 10% fetal bovine serum (Gibico, Thermo Fisher Scientific, Australia) and 1% penicillin and streptomycin (Servicebio, Beijing, China) at 37 °C in a humidified atmosphere of 5 % CO₂ for amplification. Cells were subcultured when they reached 80% to 90% confluence.

The small interfering RNA (si-YTHDC2) for YTHDC2 (Sense: 5’-CAGCUUUAAUUGUGAGAAATT-3’; Anti-sense: 3’-UUUCUCACAAUUAAAGCUGTT-5’) and negative control si-NC were obtained from Sangon Biotech (Shanghai, China). C. jejuni lipopolysacharide (LPS) was obtained from FUJIFILM Wako (Cat. No. 128-05671, Japan). The HD11 cells were seeded in 6-well plates at a density of 1 × 10⁶ cells per well and cultured for 24 h. The si-YTHDC2 or si-NC was transfected using Lipofectamine™ 3000 (Thermo Scientific, Invitrogen, US) in serum-free Opti-MEM^®I Medium (Gibco, Invitrogen, Carlsbad, CA), and incubated at 37 °C in a humidified atmosphere of 5 % CO₂ for 8 h. Following 24 hours transfection, the cells were incubated with LPS challenge (5 μg/mL) for 8 h, and the expression of YTHDC2 was detected using RT-qPCR ( Supplementary Figure S1 ).

Apoptosis assay was performed using the Annexin V-FITC/propidium iodide Cell Apoptosis Detection Kit (Servicebio, Wuhan, China) followed by flow cytometry analysis. Briefly, the cells were rinsed and resuspended with 1× binding buffer to a concentration of 5 × 10⁶ cells/mL. A 100 μL cell suspension was incubated with Annexin-FITC (5 μL) and propidium iodide (5 μL) for 10 min at room temperature in the dark. Data were analyzed using Flowjo software (Version: 10.9.1).

Total RNA from cecum or HD11 was extracted using TRIzol Reagent (Thermo Scientific, Invitrogen, US) following the manufactures’ instructions. One μg total RNA from each sample used for RNA sequencing was reverse-transcribed into cDNA with PrimeScript™ RT Reagent Kit (Takara, Japan). Quantitative real-time PCR (qRT-PCR) was performed using SYBR Premix Dimer Eraser (Takara, Japan) and gene-specific primers (Sangon, Shanghai, China) ( Supplementary Table S1 ). The relative gene expression level was calculated using the 2^-ΔΔCt method.

In the current study, the statistical analysis was performed with SPSS 26.0 software. T-tests and One-way analysis of variance (ANOVA) followed by Tukey’s post hoc test were employed to ascertain differences among groups. Among these, T-tests were used for two groups and One-way ANOVA with Tukey’s multiple comparisons were used for multiple groups (n = 4). The data were presented as mean ± SEM. Results with P value < 0.05 were considered statistically significant.

The C. jejuni colonization level in chicken cecum of the R group was significantly lower than that of the S group (P < 0.01) ( Figure 1A ). The level of proinflammatory cytokine (IL-8, IL-1β, IL-18, TNF-α, and IL-17A) in the S group was significantly higher compared to the R group. Whereas, the level IL-6 was significantly decreased in the S group (P < 0.05) ( Figure 1B ). The mitochondria in the R group were closely arranged, clearly visible, and evenly distributed, whereas those in the S group appeared damaged and swollen ( Figure 1C ). Additionally, the average cross-sectional area and the inter-membrane space distance of mitochondria in the R group were significantly smaller than those in the S group (P < 0.05). Whereas, the number of mitochondria per μm² and the cristae density per mitochondria in the R group were significantly higher than those in the S group ( Figure 1D ).

To elucidate the mechanism underlying chicken responding to C. jejuni inoculation, we evaluated the global level of m⁶A in chicken cecum in the R and S groups using dot blot method. As shown in Figure 2A , the global level of m⁶A modification decreased substantially in S group (P < 0.01). Further analysis of MeRIP-seq and RNA-seq indicated that 40–50 million raw reads were obtained from each sample. After removing the low quality reads, more than 39 million clean reads with Q20 values above 98% were generated from each sample ( Supplementary Table S2 ), and more than 86.86% clean reads could be uniquely aligned to the chicken genome. Moreover, more than 91.82% clean reads could be mapped to the exon region ( Supplementary Figure S2 ). A total of 30,427 and 30,367 m⁶A peaks, corresponding to 13,969 and 13,875 genes, were identified in the R and S groups, respectively ( Figure 2B ). Most genes contained just 1–2 peaks ( Figure 2C ). The m⁶A peaks in the R and S groups were predominantly enriched in the 3′UTR, CDS region, and stop codon, followed by the start codon and 5′UTR ( Figure 2D ). Moreover, m⁶A modifications were widely distributed throughout the chicken genome, with the highest number of peaks located on chromosome 1, accounting for 14.1% in each group ( Figure 2E ). Additionally, enriched motifs in both groups matched the canonical m⁶A motifs ‘RRACH’ and ‘DRACH’ (R = A or G; D = A, G or U; H = A, C or U), while multiple motifs corresponding to non-canonical m⁶A sites including ‘CUACG’ and ‘CGACG’ were also identified ( Supplementary Figure S3 ).

There were 514 DMPs identified between the R group and the S group including 270 hyper-methylated and 244 hypo-methylated peaks (P < 0.05, |log₂ Fold change| ≥ 1) ( Figure 3A , Supplementary Table S3 ), which mainly located in the stop codon (34.7%) ( Figure 3B ). Among which, the 270 hyper-methylated peaks were widely distributed on 36 chromosomes, while the 244 hypo-methylated peaks were widely distributed on 32 chromosomes ( Figure 3C ). Moreover, there were 38 hyper-methylated and 46 hypo-methylated peaks mainly located on chromosome 1. These DMPs were annotated to 514 genes, approximately 98% genes contained one m⁶A peak ( Figure 3D ). GO enrichment analysis indicated that the DMPs were significantly enriched in 324 terms (212 biological process (BP), 76 molecular function (MF), 36 cellular component (CC)) (P < 0.05) ( Supplementary Table S4 ). In terms of BP, the major immune-related pathways including regulation of mitochondrion organization, regulation of canonical NF-kappaB signal transduction, apoptotic signaling pathway, MyD88-dependent toll-like receptor signaling pathway, and lipopolysaccharide-mediated signaling pathway were significantly enriched ( Figure 3E ). Notably, PIDD1, ZFAND6, and CAPN3 were significantly enriched in the regulation of canonical NF-kappaB signal transduction, and apoptotic signaling pathway. PTAFR and CD180 were significantly enriched in the lipopolysaccharide-mediated signaling pathway. KEGG pathway enrichment results revealed that the DMPs were significantly enriched in Wnt signaling pathway, mTOR signaling pathway, Toll-like receptor signaling pathway, calcium signaling pathway, VEGF signaling pathway, autophagy, and tight junction (P < 0.05) ( Figure 3F ). Notably, WNT6, PRKCB, WNT4, and SLC38A9 were demonstrated significantly associated with the mTOR signaling pathway.

To elucidate the functional consequences of the gene expression modified by m⁶A methylation, the global transcriptomic landscape of chicken cecum in the R and S groups were performed using RNA-seq. Totally, 365 DEGs were identified including 166 upregulated genes and 199 downregulated genes (P < 0.05, |log₂Fold change| ≥ 1) ( Figure 4A , Supplementary Table S5 ). Notably, the expression of m⁶A modification related gene YTHDC2 were upregulated in the R group, but no significance was observed in other m⁶A modification related genes such as METTL14, FTO, ALKBH5 ( Supplementary Figure S4 ). Further GO and KEGG analysis for DEGs identified 193 significantly enriched terms (128 BP, 49 MF, 16 CC) (P < 0.05) ( Supplementary Table S6 ). In terms of BP, the major immune-related terms including defense response to bacterium, regulation of polarized epithelial cell differentiation, negative regulation of T-helper 1 type immune response, positive regulation of B cell apoptotic process, and negative regulation of cytokine activity were significantly enriched ( Figure 4B ). Among these, IL-10, DEFB4A, and AvBD1 were significantly enriched in defense response to bacterium. KEGG enrichment results showed that the DEGs were significantly enriched in eight pathways including PPAR signaling pathway, apoptosis, p53 signaling pathway, tyrosine metabolism (P < 0.05) ( Figure 4C ).

To elucidate the regulatory role of m⁶A modification, the correlation between m⁶A modification and gene expression was performed. In general, we found that the expression of gene containing m⁶A peak was significantly higher than that of genes without m⁶A modifications in both the R and S groups (P < 0.05) ( Figure 5A ). Further analysis revealed that 58 DMGs were identified in R group including 26 hyper-methylated DMGs (16 mRNAs upregulated and 10 mRNAs downregulated), and 32 hypo-methylated DMGs (23 mRNAs upregulated and 9 mRNAs downregulated) ( Figure 5B , Supplementary Table S7 ). The m⁶A peaks of the DMGs predominantly located in CDS region (n=27) and 3’UTR (n=18) ( Figure 5C ). The GO enrichment results showed that these DMGs were mainly enriched in 11 terms including autophagy, toll-like receptor 9 signaling pathway, Notch signaling pathway, intraciliary transport involved in cilium assembly (P < 0.05) ( Figure 5D ). KEGG enrichment results showed that these DMGs were significantly enriched in 7 pathways including C-type lectin receptor signaling pathway, Calcium signaling pathway, toll-like receptor signaling pathway, apoptosis, MAPK signaling pathway, and mTOR signaling pathway (P < 0.05) ( Figure 5E ). Interestingly, the hypo-methylated DMGs including SUSD5 and IFT74 were significantly enriched in Notch signaling pathway, but the expression at mRNA level was increased. The regulation of autophagy pathway was enriched by two hyper-methylated DMGs WDR41 and EPG5, which have higher mRNA expression. Moreover, the DMG FOS with hypo-methylated level and lower expression at mRNA level was significantly enriched in apoptosis pathway. The hypo-methylated DMG STAB2 was mainly associated with defense response to bacteria ( Figures 5F, G ).

In general, the qRT-PCR results for seven randomly selected genes were highly correlated with the sequencing results (R² = 0.933, P < 0.0001) ( Figure 6A ). Additionally, the m⁶A modification related genes, just the expression level of YTHDC2 in the S group was significantly lower than that in the R group, which is consistent with the RNA-seq results (P < 0.05) ( Figure 6B ). The expression level of immune related DMGs EPG5, IFT74, SUSD5, STAB2 and WDR41 in the S group were significantly lower than those in the R group, whereas the expression level of FOS in the S group was significantly higher than that in the R group (P < 0.05) ( Figure 6C ). Using MeRIP-qPCR, we found that the m⁶A modification level of EPG5, STAB2 and WDR41 in R group was increased, but the m⁶A modification levels of SUSD5, IFT74 and FOS in R group were decreased compared to S group (P < 0.05) ( Figure 6D ).

To further evaluate the function of YTHDC2, the YHTDC2 knockdown model in chicken HD11 cell line was constructed. The apoptosis rate of HD11 in LPS group was higher than that in NC group, whereas the apoptosis rate of HD11 in si-YTHDC2 group was lower than that in NC group (P < 0.05) ( Figure 7A ). The qRT-PCR results of apoptosis related gene CASP3 and anti-apoptosis related gene BCL2L1 further supported the above results (P < 0.05) ( Figure 7B ). Additionally, YTHDC2 knockdown could increase the expression of the autophagy related genes ATG5 and ULK1, whereas LPS stimulation could decrease the expression of the ATG5 and ULK1 (P < 0.05) ( Figure 7B ). Moreover, YTHDC2 knockdown significantly increased the expression of DMGs: EPG5, WDR41, SUSD5, STAB2, FOS, and IFT74 (P < 0.05). Compared to NC group, the expression of EPG5 and IFT74 was increased, whereas the expression of STAB2 was decreased in the LPS group (P < 0.05) ( Figure 7B ). These results indicated that YTHDC2 could be involved in regulating the apoptosis and autophagy process of HD11 cells through altering the expression of DMGs including IFT74, SUSD5, STAB2, EPG5 and FOS.