Methionine, 5-methylthioadenosine (MTA), cystathionine (Cysta), cysteine (Cys), SAM, homocysteine (Hcy), SAH, GSH, 5-methyltetrahydrofolate (5-MTHF), vitamin B6 (VB6), vitamin B9 (VB9), vitamin B6 (VB12), acetonitrile, formic acid, methanol, ascorbic acid, ammonium acetate, tris-(2-carboxyethyl) phosphine, hydrogen peroxide (H₂O₂), 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) were supplied by Sigma–Aldrich (St. Louis, USA). Fetal bovine serum, antibiotics (penicillin and streptomycin), and Dulbecco's modified Eagle's medium mixed with Ham's F-12 were provided by Gibco (Shanghai, China). DL–HMTBA was offered by Macklin (Shanghai, China).

IPEC-J2 cell line was a kind gift from Dr. He Qigai (Huazhong agricultural university, Wuhan, China). Cells were used between passages 70 and 90 and the cell inoculating density was 5,000/well or 3 × 10⁵/well. Each experiment was repeated three times IPEC-J2 cells were cultured in Dulbecco's modified Eagle's medium mixed with Ham's F-12, which was supplemented with 10% fetal bovine serum (v:v) and antibiotics (penicillin and streptomycin) at 37°C in a humidified incubator with 5% CO₂. Cells were cultured without fetal bovine serum for 18 h and then without Met for 6 h, at last, different ratios of L-Met and DL–HMTBA were given for 2 h.

Cells (5,000/well) were seeded in 96-well microplates overnight, then dealt with different concentrations of H₂O₂ for 24 h or different ratios of L-Met and DL–HMTBA for 2 h and then 0.8 mM H₂O₂ for 24 h. Then, 10 μl MTT was added to each well, and the cells continued to incubate for 4 h. The supernatant was discarded, and formazan crystals were resuspended with 150 μl dimethyl sulfoxide. Finally, absorbance was read at 490 nm by a spectrophotometer. Results were presented as a percentage of surviving cells compared with control cells.

Cells (3 × 10⁵/well) were seeded in the upper room of Transwell plates (Corning, USA), then 2.6 ml medium was added into the basal chamber. Millicell-ERS instrument (Millipore, Bedford, USA) was used to measure TEER values. The measured values in the blank wells of uninoculated cells were background TEER. The final TEER values were corrected for background and displayed as Ω cm².

RNA from IPEC-J2 cells was extracted using TRizol reagent (Life Technologies, Merelbeke, Belgium). TaqMan Reverse Transcription Kit (Thermo Fisher, USA) was used to transcribe 2 μg RNA into cDNA. Bio-Rad CFX Connect™ Real-Time PCR Detection System (Bio-Rad, USA) was used to measure the relative mRNA expression. Results were calculated by the 2^−ΔΔCT method. The following primers (Supplementary Table 1) were synthesized by Sangon (Shanghai, China).

Impacts of different ratios of L-Met and DL–HMTBA on Met metabolites and coenzyme was measured by LC–MS/MS, the specific experimental steps were shown as described previously (11). Contents of analytes were calculated per 1 × 10⁶ cells.

Cells were lysed with RIPA buffer containing protease and phosphatase inhibitors (Thermo Scientific, USA) for 30 min, the products were centrifuged for 15 min at 12,000 × g at 4°C, and supernatants were detected for the total protein contents. A total of 30 μg proteins were treated with 10% SDS-PAGE and then transferred to nitrocellulose membranes. They were then incubated with 5% non-fat milk for 2 h and incubated with the primary antibodies: anti-ZO-1 (1:1,000, ABclonal, #A11417), anti-FTO (1:1,000, Santa Cruz, #sc-271713), anti-YTHDF2 (1:1,000, Proteintech, #24744-1-AP), anti-METTL3 (1:1,000, Abcam, #ab195352), and anti-β-actin (1:1,000, ABclonal, #AC026). After washing three times, blots were incubated with the secondary antibody (1: 15,000): anti-rabbit or anti-mouse IgG for another 2 h. Signals were detected with chemiluminescence. Band intensities were measured by Image J (NIH, USA).

After the test, the culture solution was discarded, the cells in the Transwell chamber were washed with D-Hank's solution preheated at 37°Cthen 1.5 ml of fluorescent yellow solution (100 μg/ml) was added into the upper room. Next, the Transwell plates were placed in a cell culture incubator for 2 h. After treatment, the liquid in the lower chamber was collected for testing. The absorbance was measured with a fluorescence spectrophotometer, the excitation and emission wavelength of which were 427 and 536 nm, respectively. The ratio of fluorescent yellow contents in the lower and upper room was the fluorescent yellow flux rate (%).

Cells were lysed with RIPA buffer containing protease and phosphatase inhibitors (Thermo Scientific, USA) for 30 min at 4°C, the products were centrifuged for 15 min at 12,000 × g, 4°C, and then the supernatants were collected to determine total protein content. Concentrations of H₂S in the remaining supernatant were determined using ELISA Kit (Camilo, Nanjing, China). Briefly, samples were incubated in the microplates, which were pre-coated with monoclonal antibodies, then a specific enzyme-linked antibody for H₂S was added. After washing for 5 times, the substrate solution was added to develop color. Results were measured using a spectrophotometer at a 450 nm wavelength.

Total DNAs from IPEC-J2 cells were extracted by using a DNA Extraction kit (Genenode, Beijing, China). MethylFlash Methylated DNA Quantification Kit (Epigentek, NY) was used for the detection of DNA methylation. Briefly, genomic DNA was denatured and then incubated with capture and detection antibodies for 5-methylcytosine (m⁵C) and results were measured using a spectrophotometer at a 450 nm wavelength. Results were displayed as the percentage of m⁵C (%) in the total cell DNA.

Effects of L-Met and DL–HMTBA on m⁶A modification in IPEC-J2 cells were measured by LC–MS/MS. The specific experimental steps were shown as described previously (19). The calculated m⁶A/A was determined for the m⁶A level.

An m⁶A immunoprecipitation (MeRIP) was applied to detect the specific m⁶A modification levels of the target genes, the specific experimental steps were shown in the previous study (19). The primers used for this detection were shown in Supplementary Table 2.

Data were shown as the mean ± SEM and analyzed using GLM one-way ANOVA in SAS 8.2 software, and Ducan multiple comparisons. GraphPad Prism version 8.0 was used for correlation analysis. p < 0.05 means that the dates are statistically significant and p < 0.01 shows extreme significance.

In this study, IPEC-J2 cells were dealt with different ratios of L-Met and DL–HMTBA (100 and 0%, 85 and 15%, 70 and 30%, 40 and 60%, 0%, and 100%) with normal culture or H₂O₂ added. When DL–HMTBA was used as the Met source, mRNA expression levels of ZO-1, claudin, and occludin were significantly upregulated, and the expression level of ZO-1 gradually increased with the increase in DL—HMTBA proportion (p < 0.05) (Supplementary Figure 1).

Then, the influence of pre-treatment of IPEC-J2 cells with different ratios of L-Met and DL–HMTBA on the permeability of monolayer cells after H₂O₂ injury was investigated. According to the results of the MTT assay, treatment with 0.8 mm H₂O₂ for 24 h was selected to build the oxidative stress model (Supplementary Figure 2). Compared with the 100% L-Met group in the presence of H₂O₂, when the ratio of L-Met to DL–HMTBA was ≤ 40%:60%, the cell viability was significantly higher; when the ratio was ≤ 70%:30%, the cell TEER value was significantly higher; when the ratio was ≤ 85%:15%, the fluorescent yellow permeability was significantly lower (Figures 1A–C) (p < 0.01). Further studies on the gene and protein expression levels of tight junctions have shown that when only DL–HMTBA is present, mRNA levels of ZO-1, claudin, and occludin were significantly higher (Figures 1D–F), and when the ratio was ≤ 40%:60%, the protein level of ZO-1 was significantly higher (Figure 1G).

Next, metabolic changes of Met with different ratios of L-Met and DL–HMTBA were investigated. Compared with the 100% L-Met group, when the proportion of L-Met to DL–HMTBA was ≤ 85%:15%, the contents of SAM, Cysta, MTA, and the proportion of SAM to SAH were significantly lower; when the ratio was ≤ 40%:60%, the contents of Met and SAH were significantly lower; when only DL–HMTBA was present, the intracellular Hcy content was significantly lower (p < 0.05). Besides, the contents of Cys, GSH, and H₂S and the cofactors VB6, VB9, and VB12 involved in metabolism in each treatment group were not significantly different (Figure 2).

We then detected changes in the mRNA levels of pivotal enzymes associated with the metabolism of Met. Compared with the 100% L-Met group, when the proportion of L-Met to DL–HMTBA was ≤ 85%:15%, the mRNA level of methionine adenosyltransferase 2A (MAT2A) was significantly higher; when the ratio was ≤ 70%:30%, the mRNA level of Cysta-beta-synthase (CBS) was significantly higher; when the ratio was ≤ 40%:60%, the mRNA levels of S-adenosylhomocysteine hydrolase (AHcy) and Cystagamma-lyase (CTH) were significantly higher (p < 0.01), in addition, only in 100% DL–HMTBA group, the mRNA levels of 5-methyltetrahydrofolate-Hcy methyltransferase (MTR) and methylenetetrahydrofolate reductase (MTHFR) were significantly higher (Figure 3) (p < 0.05).

The aforementioned studies showed that DL–HMTBA significantly affects Met metabolism and gene expression in IPEC-J2 cells. Then, is there a specific correlation between changes in metabolites and changes in gene expression levels? Therefore, we next analyzed the correlation between metabolites of Met and the gene expression of Met metabolic enzymes and tight junction proteins. As shown in Figure 4, intracellular contents of Met, SAM, SAH, Hcy, Cysta, and the proportion of SAM to SAH, were significantly negatively correlated with the mRNA levels of most genes (p < 0.05). Notably, a significant negative correlation was between the proportion of SAM to SAH and each gene expression level (p < 0.01), and the correlation coefficient with the mRNA level of MAT2A was 0.884.

We considered that the difference in mRNA expression levels caused by DL–HMTBA and L-Met might be due to the difference in the metabolism between them, which leads to the difference in the contents of SAM and SAH and then changes in intracellular methylation modification. Therefore, we next tested the contents of m⁵C and m⁶A, the major methylation modifications of DNA and mRNA, respectively. As shown in Figure 5A, L-Met and DL–HMTBA had no significant effect on m⁵C content in the cell DNA, whereas DL–HMTBA significantly reduced m⁶A mRNA levels (p < 0.01).

Therefore, we examined the mRNA levels of the major RNA MTRs, demethylases, and recognition proteins and further detected the protein levels of the genes with significant changes. As shown in Figures 5C–G, when the ratio was ≤ 70%:30%, the mRNA levels of MTR-like 3 (METTL3), fat mass and obesity-associated protein (FTO), and YT521-B homology domain family 2 (YTHDF2) were evidently higher than those in 100% L-Met group (p < 0.01). Among those rations, different ratios had no obvious impact on the protein level of METTL3 or YTHDF2. However, when the ratio was ≤ 40%:60%, the protein level of FTO was obviously higher than that in the 100% L-Met group (p < 0.01).

L-Met and DL–HMTBA had significantly different impacts on the mRNA levels of MAT2A, CBS, and ZO-1. The aforementioned findings also showed that DL–HMTBA could reduce m⁶A modification of mRNA in IPEC-J2 cells. DL–HMTBA may affect the metabolic fate of mRNA in IPEC-J2 cells, for example, affecting the stability of mRNA. Therefore, actinomycin was used to investigate the influence of L-Met and DL–HMTBA on the mRNA stability of these genes. The results showed that the mRNA stability of MAT2A after 2 h and ZO-1 after 1 h of actinomycin treatment was significantly improved by DL–HMTBA (Figures 6A,C) (p < 0.05).

Next, the m⁶A modification levels of these specific differentially expressed genes, MAT2A and ZO-1, were studied. As shown in Figures 6D–M, DL-HMTBA significantly reduced the m⁶A modification levels of MAT2A and ZO-1 compared with L-Met. This finding indicated that DL-HMTBA promoted the mRNA expression of MAT2A and ZO-1, which was related to the reduction of the m⁶A modification of these two genes and the improvement in the stability of the corresponding mRNA in IPEC-J2 cells.