INTRODUCTION

The intestinal tract of humans and animals is a crucial organ for digesting and absorbing nutrients. Maintaining the integrity of intestinal structure and stability of physiological function is essential for the healthy development of organisms [1, 2]. However, in the daily diet of humans and animals, the gut is the most significant barrier to preventing harmful substances from entering the body. Therefore, it is one of the main organs that produce a large amount of reactive oxygen species (ROS) [3]. Excessive ROS production can cause an oxidation-antioxidant imbalance in the body, disrupt the REDOX balance, and lead to oxidative stress injury, seriously affecting the health and performance of the body [4]. This can even result in various intestinal diseases, such as gastroduodenal ulcers, inflammatory bowel disease, and gastrointestinal cancer [5, 6].

Researchers have investigated various indicators, including physiological indicators, gene expression, and signaling pathways, to prevent oxidative stress from causing damage to the gut [7, 8]. One study by Long et al. focused on evaluating the protective effects of proanthocyanidins (PCs) in reducing apoptosis in a mouse intestinal epithelial cell model (MODE-K) exposed to oxidative stress. They measured the levels of superoxide dismutase (SOD), glutathione peroxidase (GSH-PX), glutathione (GSH), as well as the pro-apoptosis-related protein Bax and anti-apoptosis-related protein Bcl-2 [9]. In another study by Wang et al., the aim was to investigate the cytoprotective action of Astragaloside IV (ASIV) and its underlying mechanisms in calf small intestine epithelial cells under hydrogen peroxide (H2O2)-induced oxidative stress. They measured the mRNA and protein levels of nuclear factor erythroid 2-related factor 2 (NFE2L2), heme oxygenase-1 (HMOX1), and NAD(P)H quinone dehydrogenase 1 (NQO1) [10]. Meanwhile, the study by Xiao et al. aimed to investigate the effect of Eucommia ulmoides flavones (EUF) on the Nrf2 pathway and its potential to inhibit oxidative stress in the intestines of piglets. The results revealed that the Nrf2 signaling pathway played a significant role in regulating oxidative stress in the intestines of piglets and is likely influenced by EUF treatment [11]. In a study by Cui et al., it was found that alfalfa saponins can inhibit oxidative stress-induced mitochondrial apoptosis through the MAPK signaling pathway [12].

To date, the effect of oxidative stress on the entire transcriptional expression profile of mouse intestinal epithelial cells has not been reported. In this study, mouse small intestinal epithelial cells were treated with H2O2, and the transcriptome analysis was carried out to evaluate the impact of oxidative stress and elucidate the relevant molecular mechanism, which hopefully identify the corresponding protective measures for preventing and repairing intestinal oxidative damage.

EXPERIMENTAL

Cell culture. The MODE-K cell line (Shanghai GuanDao Biological Engineering Co., Ltd, China) was cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 mg/mL streptomycin. The culture was maintained in a 5% carbon dioxide atmosphere at 37°C [13].

Determination of the optimal H2O2 concentration for oxidative stress. To determine the optimal hydrogen peroxide concentration for oxidative stimulation, MODE-K cells were seeded at a density of 1 × 104 cells/well in 96-well plates. After 24 h of incubation, the cells were treated with varying concentrations of H2O2 (50, 100, 150, 200, and 400 μmol/L) for 4 h. Cell viability was assessed using the CCK-8 assay kit (Beyotime, China) [14]. The experiment was repeated three times.

Measurement of intracellular reactive oxidative species. Intracellular ROS levels were determined using a commercially available kit (Beyotime, China) following the manufacturer’s instructions. Briefly, MODE-K cells were seeded at a density of 1 × 104 cells/well in 96-well plates and cultured for 24 h. The cells were then divided into two groups: a control group and an oxidative stress model group. The control group was cultured under normal conditions, while the oxidative stress model group was treated with 150 μmol/L H2O2. After 4 h of culture, 100 μL of medium containing DCFH-DA fluorescent probe working solution was added to each well and incubated at 37°C for 20 min. The cells were then washed twice with PBS, and intracellular ROS levels were measured at an excitation wavelength of 488 nm and an emission wavelength of 525 nm using a BioTek Synergy H1 microplate reader (BioTek Instruments, Winooski, VT, USA) [15].

Cell apoptosis analysis by FITC-Annexin-V and propidium iodide. Apoptosis was assessed using the FITC-Annexin-V and propidium iodide (PI) staining kit (Beyotime, China) following the manufacturer’s instructions. Briefly, MODE-K cells were seeded in 6-well plates at a concentration of 3 × 105 cells/well and treated as described above. The cells were then digested with pancreatic enzyme and washed twice with 1× PBS. After centrifugation, the cells were gently resuspended in 195 μL of Annexin V-FITC binding solution, and 5 μL of Annexin V-FITC and 10 μL of PI were added. The cells were then incubated at room temperature in the dark for 15 min. Flow cytometric analysis was performed using a CytoFLEX flow cytometer (Beckman Coulter, USA [16].

RNA extraction and sequencing. Total RNA was isolated and purified from MODE-K cell samples using the TRIzol reagent kit (TaKara, China) according to the manufacturer’s instructions [17]. RNA was extracted from both a normal control group (D) without H2O2 treatment and an oxidative stress model group (S) treated with 150 μmol/L H2O2 for 4 h. Three replicates were performed in each group. Shortly, messenger RNA (mRNA) was isolated using polyA selection with oligo(dT) beads. The isolated mRNA was then fragmented using a fragmentation buffer. Double-stranded cDNA was synthesized using a SuperScript double-stranded cDNA synthesis kit with random hexamer primers. The synthesized cDNA underwent end-repair, phosphorylation, and “A” base addition according to Illumina’s library construction protocol. Libraries were size selected for cDNA fragments of 300 bp and amplified using Phusion DNA polymerase for 15 PCR cycles. The quantified libraries were then sequenced using the NovaSeq 6000 sequencer with 2 × 150 bp read length for paired-end RNA-seq. The RNA samples were sent for sequencing to Shanghai Majorbio Bio-pharm Biotechnology Co., Ltd. (Shanghai, China). The raw data has been deposited in the NCBI Sequence Read Archive with the accession number PRJNA1071369.

Quality control and read mapping. The raw data was processed using fastp (https://github.com/OpenGene/fastp) to remove any sequences containing adapters or low-quality bases [18]. The resulting high-quality reads, or clean reads, were used for subsequent analysis, and parameters such as Q30 and GC content were calculated. The clean reads were then aligned to the mouse reference genome using HISAT2 software (http://ccb. jhu.edu/software/hisat2/index.shtml) [19] with a directional pattern. The sequence of the mouse genome was obtained from the website http://asia.ensembl.org/ Mus_musculus/Info/Index. StringTie software was used to assemble each sample using a referenced method [20].

Differential expression analysis and functional enrichment. The transcriptional abundance was calculated with RSEM software, and gene expression was quantified using FPKM [21]. Differential gene expression analysis was performed using DESeq2, with a fold change (FC) ≥ 2 and a P-adjust < 0.01 considered significant. GOseq R program was utilized to perform Gene Ontology (GO) enrichment analysis on the DEGs to obtain functional annotation. Pathway enrichment analysis was conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG). The GO and KEGG annotation are Blast2go 2.5 and KOBAS 2.1.1 respectively. In addition, functional-enrichment analysis including GO and KEGG were performed to identify which DEGs were significantly enriched in GO terms and metabolic pathways at Bonferroni-corrected P-value ≤0.05 compared with the whole-transcriptome background. GO functional enrichment and KEGG pathway analysis were carried out by Goatools and KOBAS [22], respectively. We excluded genes with TPM differential expression lower than 5 and performed a cluster heatmap analysis for the top 300 DEGs with FC. The DEGs were compared to the STRING database to identify PPI relationships, and the results were visualized using Cytoscape software. Key subnetworks and core genes were identified using the MCODE plug-in (the filter conditions were set as degree cutoff of 2 and k-core of 2 to select the core cluster with the closest relationship in the network), and correlation analysis was performed on the DEGs in the subnetwork [23].

qRT-PCR analysis for the validation of RNA-Seq data. qRT-PCR, performed using an ABI7500 instrument, was used to validate the transcriptome sequencing data. The same total RNA samples used for RNA sequencing (RNA-seq) were used for qRT-PCR. All the primers were confirmed to be effective by Agarose gel electrophoresis (Table S1, see Supplementary Information). Reverse transcription was done using a PrimeScript RT Reagent Kit (Takara, China) with 1 μg total RNA. Individual reactions were prepared using 100 ng cDNA and SYBR Green PCR Master Mix (Vazyme, China) in a final volume of 20 μL. Three biological copies were performed in this experiment. Cycle threshold (Ct) values were normalized to GAPDH, and the relative mRNA quantification was determined using the 2−ΔΔCt method [24].

Statistical analysis. The data are presented as means ± standard deviation (SD), and statistical comparisons were performed using ANOVA in SPSS 23.0 (P < 0.05 indicates a significant difference, and P < 0.01 indicates a highly significant difference).

RESULTS

Establishment of the Oxidative Stress Model

To determine the optimal concentration of hydrogen peroxide for inducing oxidative stress, we conducted an experiment using MODE-K cell line. As depicted in Fig. 1a, a low dose of H2O2 (50 μmol/L) significantly increased the survival rate of MODE-K cells (P < 0.05), while a high dose of H2O2 (100‒400 μmol/L) significantly decreased the survival rate of MODE-K cells. Specifically, the survival rates of cells stimulated by 100, 150, 200, and 400 μmol/L H2O2 were 77.6, 55.7, 43.5, and 14.2%, respectively. We selected H2O2 concentration with cell survival rate of about 50% as the optimal working concentration. Therefore, in all subsequent experiments, we induced oxidative stress by exposing the cells to 150 μmol/L H2O2 for 4 h.

Fig. 1.
figure 1

Establishment of the oxidative stress model. (a) The optimal H2O2 concentration for oxidative stress was determined using a cell counting kit 8 (CCK-8) assay, (b) Bar graph summarizing the data on ROS production, (c) Apoptotic cells measured using a CytoFLEX flow cytometer. * p < 0.05 compared with the control group, ** p < 0.01 compared with the control group.

Full size image

Treatment with 150 μmol/L H2O2 increased the intracellular level of ROS in MODE-K cells, and the ROS content was found to be 53.5% higher than that of the control group. Exposure to 150 μmol/L H2O2 for 4 h significantly increased the proportion of apoptosis (P < 0.01) in MODE-K cells (Fig. 1b). Our results demonstrate that H2O2 can effectively promote apoptosis in small intestinal epithelial cells, indicating that the oxidation model was successfully established (Fig. 1c).

Results of RNA-Seq Data Analysis and DEGs Functional Enrichment Analysis

We obtained a total of 49.63 Gb of sequencing data, with each sample yielding more than 7.1 Gb of net data. The GC content of the six samples ranged from 50.07 to 50.93%. The Q30 score for all samples was above 95.14%, and the alignment efficiency was between 95.84 to 96.44% when compared to the mouse reference genome, indicating the reliability of the data for subsequent analysis (Table S2, see Supplementary Information). In the PCA analysis, the control and model groups showed good sample repeatability, with a small distance within these groups. Conversely, the distance between the model group and control group was far, indicating the differences between groups were relatively large (Fig. 2a). We compared gene expression between the control and model groups after the treatment, and we identified a total of 1207 DEGs, of which 859 genes were up-regulated, and 348 genes were down-regulated (Fig. 2b).

Fig. 2.
figure 2

PCA and volcanic maps of transcriptome data. (a) PCA of transcriptome data. (b) Volcano plot of the differentially expressed genes between the control group and model group. S—corresponds to model group, D—corresponds to control group.

Full size image

The GO functional annotation analysis of the 1207 DEGs revealed their annotation into 51 different GO classifications, including 22 biological processes, 15 cellular components, and 14 molecular functions (Figs. 3a, 3b and 3c). The category of molecular function is predominantly enriched in the DEGs involved in binding and catalytic activity, with 659 and 253 DEGs (Table S3, see Supplementary Information), respectively. Among biological process category, the DEGs were most annotated in cellular process and biological regulation, with 646 and 577 DEGs (Table S3), respectively. The cellular component category mainly enriched in DEGs for cell part, with 760 DEGs (Table S3). In addition, during the oxidative stress response of mouse small intestinal epithelial cells, other GO classifications were also important, including metabolic process, developmental process, response to stimulus, cellular component organization or biogenesis, organelle, membrane part, organelle part, membrane, protein-containing complex and others, as shown in Fig. 3. For further study, we selected the top 20 enrichment items with more than 100 GO enrichment genes to make the enrichment bubble map (Fig. 3d, Table S4). The DEGs were mainly enriched for the following BP terms: regulation of localization, regulation of developmental process, regulation of signal transduction, regulation of molecular function, regulation of signaling, anatomical structure development, regulation of DNA-templated transcription, response to stress, regulation of multicellular organismal process, positive regulation of nitrogen compound metabolic process, and so on.

Fig. 3.
figure 3

GO annotation of differentially expressed genes between the control group and model group. (a) “Biological process” categories. (b) “Cellular component” categories. (c) “Molecular function” categories. (d) The top 20 of GO enrichment of DEGs.

Full size image

To identify the biological pathways involved in oxidative stress in the mouse small intestine epithelial cells, we mapped the DEGs from both groups to KEGG enrichment analysis. We obtained a total of 287 signaling pathways, among which 63 pathways were significantly enriched, such as MAPK signaling pathway, Glutamatergic synapse, Cortisol synthesis and secretion, TNF signaling pathway, and IL-17 signaling pathway (Fig. 4a, Table S5, see Supplementary Information). These findings suggested that these pathways were activated by H2O2 stimulus and participated in oxidative stress regulation of small intestinal epithelial cells.

Fig. 4.
figure 4

KEGG involved oxidative stress in MODE-K cells. (a) Top 20 KEGG enrichment scatter plot of DEGs. (b) Representative KEGG categories involved in oxidative stress in MODE-K cells.

Full size image

We further analyzed five KEGG categories in this study, namely cellular processes, environmental information processing, genetic information processing, metabolism, and organismal systems. Among them, pathways related to environmental information processing, cellular processes, and organismal systems were significantly enriched. Environmental information processing involved some signal transduction pathways, which played roles in oxidative stress regulation (Table S6). For instance, the MAPK signaling pathway (27 genes), NF- kappa B signaling pathway (12 genes), and PI3K-Akt signaling pathway (25 genes) were found to be involved. In the Cellular Processes, the expression levels of twelve genes related to Apoptosis were altered after acute oxidative stress in mouse small intestinal epithelial cells (Table S6). Additionally, H2O2-induced changes in pathways involved in organismal systems included environmental adaptation, digestion, endocrine, immune, nervous, and sensory functions, all of which should be involved in oxidative stress regulation (Fig. 4b, Table S6).

PPI Reveals Oxidative Stress-Affected Cell Growth, Apoptosis, and so on through the PTGS2-KDM6B-MT-ATP6 Axis

To investigate the differences in gene expression changes in the small epithelial intestine cells of mouse under oxidative stress, we ordered the 1207 DEGs based on their degree of change in expression, from high to low. We then selected the top 300 DEGs (excluding genes with TPM differential expression was lower than 5) to generate a cluster heatmap. The color differences in the heatmap represent the degree of change in gene expression (Fig. 5, Table S7, see Supplementary Information). Among the top 300 DEGs, 174 were up-regulated, and 126 were down-regulated in the model group.

Fig. 5.
figure 5

Hierarchical clustering analysis of the differentially expressed genes between the control and model group.

Full size image

To study the mutual regulation and mode of action among the DEGs, we utilized the STRING database to conduct a PPI analysis on the 300 DEGs to obtain the protein interaction network relationship (Fig. 6a). We visualized this network in Cytoscape and used the MCODE plug-in to screen the two subnetworks with the highest scores (Figs. 6b and 6c). We then used the DAVID Function Annotation tool to annotate the subnetworks. The first subnetwork was found to be involved in the MAPK signaling pathway, PI3K-AKT signaling pathway, and NF-kappa B signaling pathway. The second subnetwork was involved in oxidative phosphorylation. It was discovered that KDM6B linked the two subnetworks. The correlation analysis was conducted for the expression levels of the DEGs that were shown to be related by PPI analysis in the two subnetworks. In the first subnetwork, the correlation between any two genes was positive (P < 0.01), except for CEBPB, which did not exhibit the significant correlation with other genes (Fig. 7a, Table S8, see Supplementary Information). In the second subnetwork, the Trmt10c gene was upregulated, while genes associated with oxidative phosphorylation, such as mt-Nd5, mt-Atp6, mt-Nd6, mt-Nd2, and mt-Co1, were downregulated. TRMT10C was negatively correlated with MT-ND5, MT-ATP6, MT-ND6, MT-ND2, and MT-CO1 (P < 0.01). Except for TRMT10C gene, the other genes were positively correlated each other, with extremely significant results (P < 0.01) (Fig. 7b, Table S9).

Fig. 6.
figure 6

Protein−protein interaction (PPI) network among the Top 300 differentially expressed genes in the model group compared with the control group. (a) The PPI network by String software. (b, c) The most significant two subnetworks identified by MCOD-E cells.

Full size image
Fig. 7.
figure 7

Heatmaps for correlation analysis of the DEGs mRNA level in subnetworks. (a) Heatmap for correlation analysis of the DEGs mRNA level in the first subnetwork showed in Fig. 6b. (b) Heatmap for correlation analysis of the DEGs mRNA level in the second subnetwork showed in Fig. 6c.

Full size image

Validation of RNA Sequencing Results by qRT-PCR

To validate the RNA sequencing analysis, we selected DEGs, whose roles were critical under oxidative stress, to make qRT-PCR detection. TNF and GADD45A belonged to the MAPK signaling pathway, NF-kappa B signaling pathway, and apoptosis signaling pathway. JUN and DDIT3 belonged to the MAPK signaling pathway and apoptosis signaling pathway. PTGS2 was involved in the NF-kappa B signaling pathway, HSPA1B in the MAPK signaling pathway, and MDM2 in the PI3K-AKT signaling pathway. KDM6B were associated with cellular response to hydrogen peroxide. MT-ND6 and MT-ND2 were associated with oxidative phosphorylation. After oxidative stress, the qRT-PCR results of the selected DEGs were consistent with the results of RNA-seq (Figs. 8a, 8b), and there was a high correlation between them (R2 = 0.9968). Our RNA sequencing results are, therefore, reliable.

Fig. 8.
figure 8

Verification of the selected ten differentially expressed genes by qRT-PCR. (a) Comparison of RNA-seq data with qRT-PCR data. (b) Correlation between RNA-Seq and qRT-PCR gene expression.

Full size image

DISCUSSION

Oxidative stress can have an impact on animal reproduction and growth. Intestinal epithelial cells have a dual role of absorbing nutrients and functioning as a physical barrier between the body and intestinal flora. Oxidative damage to the intestinal epithelium causes the breakdown of this barrier, negatively affecting human and animal health [25], which means that it is important to identify the key genes involved. RNA-Seq is considered an effective method in transcriptome analysis. Using this deep sequencing technology, the whole transcriptome changes can be obtained in a high-throughput way. This study subjected mouse small intestinal epithelial cells to oxidative stress experiments and performed transcriptome sequencing after oxidative stress. Additionally, according to the identified changes in various pathways and gene expression levels, we analyzed the molecular mechanisms involved in regulating oxidative stress in the mouse small intestine epithelial cells.

We identified 1207 DEGs, with 859 up-regulated and 348 down-regulated genes, which suggested that the genes participated in the oxidative stress response of small intestinal epithelial cells. KEGG pathway and KEGG category analyses revealed changes in some signaling pathways such as the MAPK pathway, NF- kappa B pathway, and PI3K-Akt pathway following oxidative stress in small intestinal epithelial cells. During oxidative stress, the MAPK pathways can be activated, leading to apoptosis, inflammation, and cell proliferation [26]. The NF-κB pathway is another important signaling pathway involved in the regulation of cellular immune response, apoptosis, and inflammation. Oxidative stress can activate the NF-κB pathway, leading to apoptosis and inflammation [27]. The signal transduction of PI3K-Akt is closely related to many biological processes, such as cell survival, proliferation, and differentiation, which is beneficial to antioxidant process [28]. The MAPK pathway is a critical signaling pathway that regulates various cellular processes, including proliferation, differentiation, apoptosis, and stress responses, through cascades. H2O2 treatment of cells can activate the MAPK signaling pathway, affecting the cell cycle, altering the circadian rhythm, and causing cell apoptosis and other diseases. Recent studies have shown that inhibiting the MAPK signaling pathway contributes to antioxidant stress. For example, Ma et al. found that Baicalin inhibited the MAPK signaling pathway, contributing to antioxidant stress and anti-inflammatory effects [29]. Guo and Dong discovered that inhibiting the MAPK signaling pathway by targeting the MEK2 protein could reduce hydrogen peroxide-induced apoptosis and alleviated oxidative stress-induced damage to lens epithelial cells [30]. Tang et al. demonstrated that berberine exhibited a significant reduction in the expression levels of ERK, JNK, and NF-κB genes and proteins in the jejunum and ileum, which were induced by oxidative stress. Berberine was found to regulate the oxidative stress by modulating the signaling pathways of MAPK, and ultimately preserve the intestinal health of piglets [31]. These findings suggested that H2O2 could activate the MAPK signaling pathway, leading to cell apoptosis, which was consistent with our KEGG analysis, indicating that the MAPK signaling pathway has been identified to play a critical role in promoting oxidative stress in the intestine. Therefore, inhibiting the MAPK signaling pathway by targeting its key proteins might be an effective way to improve the antioxidant stress ability of small intestinal cells.

Ten DEGs (JUN, GADD45A, DDIT3, BCL2L11, CDKN1A, MDM2, GDF15, PTGS2, TNFAIP3, and HSPA1A) in the first subnetwork of PPI network were closely linked and significantly correlated. The proteins were annotated to the MAPK pathway, PI3K-AKT pathway, and NF-kappa B pathway by DAVID, with most genes closely related to the MAPK pathway. This was consistent with the KEGG analysis results of DEGs. JUN is a proto-oncogene and has been identified as one of the key genes involved in oxidative stress damage in the cerebellum of patients [32]. GADD45A is a growth retardation and DNA damage-inducing protein. Its overexpression is associated with the activation of the p38 MAPK signaling pathway, and its knockdown can alleviate oxidative stress [33]. DDIT3 (DNA damage-inducible transcript 3) plays a dual role in balancing glycolysis and oxidative phosphorylation, establishing a connection between glycolysis and mitochondrial respiration. DDIT3 promotes glycolysis and ATP production by inhibiting TIGAR, a negative regulator of glycolysis. Additionally, a fraction of DDIT3 goes to mitochondria and inhibits oxidative phosphorylation through LONP1-mediated downregulation of COQ9 and COX4 [34]. BCL2L11 is a pro-apoptotic protein and a member of the BCL2 family. Increased expression of the BCL2L11 gene promotes apoptosis, and targeted downregulation of BCL2L11 expression can inhibit apoptosis in adult rat astrocytes [35]. CDKN1A is a cyclin-dependent kinase inhibitor, and its mRNA levels increase after oxidative stress in various mouse and human tissues, making it a marker of cellular senescence [36]. The MDM2 gene product is a zinc finger protein that binds to the p53 protein and inactivates its transcriptional regulatory function. Phosphorylation of Mdm2 serine residue 183 by Akt increases nuclear Mdm2 stability, reduces p53 levels, and prevents senescence in primary cells exposed to reactive oxidative species [37]. GDF15, a member of the S100 calcium-binding protein family, is a growth differentiation factor and a circulating molecule associated with oxidative stress and inflammation. It serves as a novel biomarker for oxidative stress and inflammation [38]. PTGS2, also known as prostaglandin-endoperoxide synthase 2, is involved in the inflammatory response of various diseases. PTGS2 silencing can reverse the promotion effect of LPS on inflammation [39]. TNFAIP3, or tumor necrosis factor alpha-inducible protein 3, plays an important role in the pathogenesis of autoimmune diseases and chronic inflammatory diseases. TNFAIP3 protein can reduce pain in lumbar disc herniation by regulating the NF-κB pathway [40]. HSPA1A is a protective protein whose expression is elevated after stress. He et al. found that HSPA1A could inhibit H2O2-induced apoptosis and improved apoptosis after spinal cord injury [41]. The PPI network analysis revealed that these genes in the first subnetwork served as central hubs, which confirmed their key roles in the oxidative stress regulation of mouse small intestine cells, some being like the other types of cells or tissues according to the previous reports. Furthermore, the regulatory relationships among these genes uncovered the underlying mechanisms, which provided a reference for future studies on gene regulation.

In the second subnetwork, the core genes MT-ND5, MT-ATP6, MT-ND6, MT-ND2, MT-CO1, and TRMT10C were annotated in the oxidative phosphorylation signaling pathway through DAVID annotation, indicating their involvement in the response of mouse small intestinal epithelial cells to oxidative stress by regulating energy metabolism [42]. TRMT10C is tRNA methyltransferase 10C, the mutation of which is the cause of mitochondrial disease, and its importance for RNA processing is keeping the correct function of mitochondria [43]. TRMT10C may act as a post-transcriptional regulator of genes by controlling N1-methyladenosine (m1A) levels in MT-ND5, MT-ATP6, MT-ND6, MT-ND2, MT-CO1 and other genes [44]. These genes are all from mitochondria, the central organelle of energy metabolism that is exposed to ROS. Excessive ROS production leads to the decreased ATP production, inhibition of intracellular electron transport chains, and DNA damage in mitochondria [45]. If this continues, mitochondrial bioenergetics and homeostasis can be affected, leading to cell death. Mitochondria are known to be the primary targets of oxidative stress, but the underlying mechanism promoting oxidative stress remains incompletely understood. The interconnected mitochondrial genes identified in this paper provided a foundation for further elucidating this mechanism.

We attempted to find a direct link between the two subnetworks through the PPI network and discovered that KDM6B linked the two subnetworks. KDM6B is a histone demethylase containing the Jumonji C domain. It regulates the expression of important genes in the mitochondria and is indispensable for ensuring mitochondrial function, which is consistent with our finding that KDM6B played a key role in mitochondrial dysfunction caused by oxidative stress. Jia et al. found that PTGS 2/MAPK signaling played a key role in inhibiting inflammatory response and oxidative stress [46]. Our results further outlined the possible mechanisms: PTGS2 gene was regulated by other core genes through MAPK and PI3K/AKT/NF-kappa B pathways in the first subnetwork, and then PTGS2 regulated the expression of KDM6B, KDM6B, in turn, down-regulated MT-ATP6 and subsequently affected the other mitochondrial genes associated with oxidative phosphorylation in the second subnetwork, which finally led to mitochondrial dysfunction, energy metabolism imbalance and cell apoptosis that indicated the damage of oxidative stress. These results suggested that the PTGS2-KDM6B-MT-ATP6 axis played a key role in coordinating nucleo-mitochondrial genome interactions and maintaining cell homeostasis. All in all, through this deep sequencing technology and bioinformatics analysis, we obtained essential information for revealing the mechanism of the small intestinal epithelial cells in response to oxidative stress at the transcriptional level, which opened a new avenue for further in-depth study of the regulatory mechanism at the protein level and provided a reference for preventing and treating oxidative damage of intestine in the future.

CONCLUSIONS

We discovered that oxidative stress affected the growth, metabolism, and apoptosis of mouse small intestinal epithelial cells through the MAPK and PI3K/AKT/NF-kappa B signaling pathways, and the mitochondria-related genes, both of which were interconnected through PTGS2-KDM6B-MT-ATP6 axis. The possible regulatory mechanism was that KDM6B was regulated by the hub genes (JUN, GADD45A, DDIT3, BCL2L11, CDKN1A, MDM2, GDF15, TNFAIP3, HSPA1A, and PTGS2) of the signaling pathways and then regulated the mitochondria-related genes (MT-ATP6, MT-ND2, MT-ND5, MT-ND6, MT-CO1, and TRMT10C), which led to mitochondria dysfunction, metabolism imbalance, and cell apoptosis. These findings provide a theoretical basis for understanding the molecular mechanism of oxidative stress and further proposing the corresponding measures to safeguard intestine from oxidative injury, such as regulating MAPK and PI3K/AKT/NF-kappa B signaling pathways, and enhancing the expressions of genes related to mitochondrial biogenesis and function.