To investigate the metabolites profiles during SADS-CoV infection, we first confirmed SADS-CoV replication in porcine ileum epithelial cell line, IPI-FX, at a MOI of 0.1 for a period of 12, 24, and 36 hours ( Figure 1A ). Mock-infected cells were used as controls. The results of indirect immunofluorescence assay (IFA) showed that the virus could be detected at 12 hours post-infection (hpi), and the red fluorescence increased with time, indicating that SADS-CoV had successfully proliferated in the IPI-FX cells. An untargeted metabolomics and lipomics approach based on liquid chromatography-mass spectrometry (LC-MS) were used to identify the metabolic characteristics related to SADS-CoV infection ( Figure 1B ). Infected and mock-infected cells were harvested at 12, 24 and 36 hours, with 5 biological replicates at each time point. Data were collected for subsequent research.

Principal component analysis (PCA) for quality control (QC) samples can be used to investigate the overall distribution of samples in each group and the stability of the whole analysis process. In general, the data revealed consistency and repeatability for both untargeted metabolomics and lipomics, and there was a noticeable distinction between the infected cells and those of controls ( Figures 2A, B ). It is noteworthy that the distinctions at 12 hpi and 24 hpi were more obvious than at 36 hpi, which might indicate that intracellular metabolite differences reduced in the late stage of infection. In total, we obtained 11,375 metabolites from the data of the untargeted metabolomics, but only 2,225 had identification information. The lipidome had much less data, only 868 metabolites but all of them were identified. By classifying the metabolites, it was found that most of them in non-targeted metabolomics belonged to lipids and amino acid, peptide, and analogues, while most of the molecules in lipidome were phosphatidylcholine (PC) and phosphatidylethanolamine (PE) ( Figures 2C, D ). Overall, we have successfully obtained a large amount of available information from metabolomics.

To determine how IPI-FX cells respond to SADS-CoV infection, the identified metabolites in viral infection groups were classified into 8 clusters using the Mfuzz package (v 2.48.0) according to the expression information with the time ( Figure 3 ). The metabolites of untargeted metabolomics classified in clusters 1, 2, 3, 4, and 5 were downregulated, whereas the metabolites in clusters 6, 7, and 8 were upregulation ( Figure 3A ). For lipomics, clusters 1, 3, 4, and 5 showed a downregulation trend, while clusters 2, 6, 7 and 8 showed upregulated expression ( Figure 3B ). Those metabolites clustered in one group are more likely to be considered to belong to the same functional set.

Comparing the metabolomic data between SADS-CoV infected IPI-FX cells and controls, there were significant differences in the level of 1257 metabolites in untargeted metabolomics, while in the lipomics, there were 435 significantly changed metabolites (Fold-Change ≥ 1.2 or ≤ 0.83, q-value < 0.05) ( Figure 4A ). The number of upregulated and downregulated metabolites at each time point was about equal. In order to visually compare common and unique differential metabolites between different comparison groups, Venn diagrams were drawn for all comparison groups with different metabolites ( Figures 4B, C ). The results showed that the metabolites were altered in abundance in different time points compared to those in uninfected cells. Volcano plots were also used to visualize differential metabolites ( Figures 4D–I ). Each point in the volcanic map represents a metabolite. As a whole, we demonstrate that a large number of intracellular metabolites are significantly altered during viral infection.

Moreover, to elucidate the intracellular metabolic characteristics after SADS-CoV infection in IPI-FX cells, we further analyzed the data of the untargeted metabolome. Heatmap analysis depicting hierarchical clustering of the dysregulated metabolite data from a comparison group intuitively showed that the metabolites in the two groups of samples had different dynamic expression profiles ( Figures 5A–C ). To uncover the metabolic pathways potentially involved in the process of SADS-CoV infection, pathway enrichment analysis of differential metabolites based on the Kyoto encyclopedia of genes and genomes (KEGG) database was employed ( Figures 5D–F ). The top 10 metabolic pathways with the smallest p value were drawn as bubble charts. On the whole, we found that a striking feature of SADS-CoV infection was a significant change in amino acid metabolism, including glycine, serine and threonine metabolism, cysteine and methionine metabolism, alanine, aspartate and glutamate metabolism, and arginine biosynthesis.

The urea cycle is responsible for amino acid metabolism, especially closely related to arginine and glutamate metabolism ( Figure 5G ). Metabolomic data showed that arginine levels were downregulated and glutamine levels were upregulated in SADS-CoV infected cells, suggesting that these metabolites might have a certain function in the process of virus replication. Notably, amino acid metabolism is always related to energy metabolism. The citrate cycle (TCA cycle), an important energy metabolism, was active at 12 hpi ( Figures 5D, G ). We hypothesized that the virus might disrupt the TCA cycle to rapidly obtain the ATP needed for its own replication during the early stage of infection. In addition, obvious changes in lipid response and inflammatory pathways were observed at 36 hpi, which may be associated with the aggravation of cytopathological characteristics in the late stage of infection ( Figure 5F ). Taken together, SADS-CoV infection led to disorder of a large number of cellular metabolic pathways, especially amino acid metabolism.

Similarly, we further analyzed the lipomics data to explore important lipid metabolites or metabolic pathways during SADS-CoV infection. Cluster analysis of differential metabolite expression level also demonstrated that the metabolites between infected and mock-infected cells had different dynamic expression profiles, speculating that the virus might organize its own metabolic networks to serve self-replication ( Figures 6A–C ). The expression patterns of differential metabolites at different time points were analyzed by Bubble chart ( Figures 6D–F ). In particular, we found that triglyceride (TG) and monoglycosylceramide (CerG1) showed an up-regulated trend at all three time points of infection. The former belongs to glycerolipids, while the latter is a kind of saccharolipids. This observation suggested that these two metabolites might be involved in the process of viral replication. Furthermore, to investigate the differential metabolic pathway related to SADS-CoV infection, we also conducted metabolic pathway enrichment analysis in lipomics ( Figure 6G ). Unexpectedly, only ferroptosis was observed to be enriched at 12 hpi. We suspect that this may be because the functions of many lipid molecules have not been clearly studied and there is less annotation information available for analysis.

Ferroptosis is a novel pattern of cell death characterized by excessive accumulation of iron and overwhelming lipid peroxidation (Tang et al., 2021). In order to study the specific details of ferroptosis in lipomics, we performed metabolic pathway enrichment analysis network plot. The results revealed that in the early stage of SADS-CoV infection, ferroptosis was suppressed and the expression of phosphatidylethanolamine (PE) (18:0/20:4) was downregulated ( Figures 7A, B ). Besides, we detected the expression levels of some markers of ferroptosis in SADS-CoV infected cells, and found that positive regulators expression (Acyl-CoA synthetase long chain family member 4, ACSL4; Cytochrome c oxidase subunit II, COX2; NADPH oxidase 4, NOX4) were down-regulated, while negative regulators expression (Glutathione peroxidase 4, GPX4; Solute carrier family 7 member 11, SLC7A11; Ferritin heavy chain 1, FYH1) were up-regulated ( Figures 7C–H ). Taken together, we preliminarily demonstrated that the intracellular ferroptosis was inhibited during the early stage of SADS-CoV infection in IPI-FX cells.

IPI-FX (porcine intestinal epithelial cells) were kindly provided by Prof. Shaobo Xiao (Huazhong Agricultural University) (Wang et al., 2019), which were grown in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, New York NY, USA) containing 10% fetal bovine serum (FBS, Gibco, New York NY, USA), 100 U/mL penicillin, and 100 mg/mL streptomycin. SADS-CoV strain GDS04 (NCBI accession number: MF167434) was isolated and propagated in our laboratory.

IPI-FX cells were infected with the SADS-CoV at an MOI of 0.1 for 12, 24, or 36 hours. After collection, cells were fixed with pre-cooling methanol for 15 min and then washed with phosphate-buffered saline (PBS). Cells were then incubated with Triton X-100 (0.5%) for 15 min and following PBS washing. After blocked with 1% bovine serum albumin (BSA, Sangon, Shanghai, China) for 1 hour, cells were incubated with primary antibodies against SADS-CoV N protein (prepared and preserved in our laboratory) for 1 hour at room temperature. After PBS washing, cells were incubated with a 1:500 dilution of Cy3-conjugated goat anti-mouse IgG secondary antibodies (Abcam, Boston MA, USA) for 1 hour, following by 4’,6-diamidino-2-phenylindole dihydrochloride (DAPI, Invitrogen, Carlsbad CA, USA) staining for another 5 min. The cells were finally washed with PBS and observed using a fluorescence microscope (Nikon, Tokyo, Japan).

LC-MS-level methanol, acetonitrile and formic acid were purchased from Thermo Fisher Scientific (Waltham MA, USA); ammonia formate was purchased from Honeywell Fluka (Sandy UT, USA). Ultrapure water was obtained from a Millipore system (Billerica MA, USA).

d₃-Leucine, ¹³C₉-Phenylalanine, d₅-Tryptophan, ¹³C₃-Progesterone were used as internal standard (#1) for untargeted metabolomics. MG 18:1(d7), 2 μg/mL; PS 15:0-18:1(d7), PE 15:0-18:1(d7), LPE 18:1(d7), 5 μg/mL each sample; PA 15:0-18:1(d7), 7 μg/mL; PI 15:0-18:1(d7), DG 15:0-18:1(d7), 10μg/mL each sample; LPC 18:1(d7), 25 μg/mL; PG 15:0-18:1(d7), SM d18:1-18:1(d9), 30 μg/mL each sample; TG 15:0-18:1(d7)-15:0, 55 μg/mL; cholesterol(d7), 100 μg/mL; PC 15:0-18:1(d7), 160 μg/mL; CE 18:1(d7), 350 μg/mL were used as internal standard (#2) for lipomics.

25 mg samples were weighed and added into 2 mL thickened centrifuge tubes, simultaneously, 2 magnetic beads were added as well. Thereafter, 10 μL of the prepared internal standard (#1) was added to each sample at the same time. Then 800 µL of precooled extraction reagent (methanol: acetonitrile: water (2:2:1, v/v/v)) was added, samples were grinded with TissueLyser at 50 Hz for 5 min, then were placed at -20°C for 2 hours. Samples were centrifuged at 25,000 × g for 15 min at 4°C and 600 µL of each sample were added in split-new EP tubes and frozen dry. 120 μL of 50% methanol were added to the dried samples and mixture was shaken until completely dissolved. And then samples were centrifuged at 25,000 × g for 15 min at 4°C and 600 µL of each sample were added in split-new EP tubes. 10 μL of each sample was taken and mixed as QC samples.

25 mg samples were weighed and added into 2 mL thickened centrifuge tubes, simultaneously, 2 magnetic beads were added as well. Two steel balls and 800 µL of pre-chilled dichloromethane/methanol (3:1, v/v) precipitant were added into each sample, as well as 10 µL of the prepared internal standard (#2). Then samples were implemented ice bath ultrasound for 10 minutes and refrigerated overnight at -20°C. Samples were centrifuged at 25,000 × g for 15 minutes at 4°C and 600 µL supernatant 600 was taken and drained in freezer dryer. 120µL lipid reconstituted solution (isopropanol: acetonitrile: water = 2:1:1) was reconstituted and shaken for 10 min. Then samples were implemented ice bath ultrasound for 10 minutes and centrifuged at 25,000 × g for 15 minutes at 4°C. 10 μL of each sample was taken and mixed as QC samples.

Waters UPLC I-Class Plus (Waters, Milford MA, USA) and tandom Q Exactive high resolution mass spectrometer (Thermo Fisher Scientific, Waltham MA, USA) was employed for separation and detection of metabolites.

Chromatographic separation was performed on an ACQUITY UPLC BEH C18 column (Waters, USA) and the column temperature was maintained at 45°C. In the positive mode, the mobile phase consisted of 0.1% formic acid (A) and acetonitrile (B). The mobile phase consisted of 10 mM ammonium formate (A) and acetonitrile (B) in the negative mode. The gradient conditions were as follows: 0-1 min, 2% B; 1-9 min, 2-98% B; 9-12 min, 98% B; 12-12.1 min, 98% B to 2% B; and 12.1-15 min, 2% B. The flow rate was 0.35 mL/min and the injection volume was 5 μL.

Q Exactive perform was used for primary and secondary mass spectrometry data acquisition. The full scan range was 70-1050 m/z with a resolution of 70,000, and the automatic gain control (AGC) target for MS acquisitions was set to 3e6 with a maximum ion injection time of 100 ms. Top 3 precursors were selected for subsequent MSMS fragmentation with a maximum ion injection time of 50 ms and resolution of 17,500, the AGC was 1e5. The stepped normalized collision energy was set to 20, 40 and 60 eV. ESI parameters were set as follows: Sheath gas flow rate: 40; Aux gas flow rate: 10; positive-ion mode Spray voltage(|KV|): 3.80; negative-ion mode Spray voltage(|KV|): 3.20; Capillary temperature: 320°C; Aux gas heater temperature: 350°C.

Equipments for separation and detection of metabolites were the same as that for untargeted metabolomics above.

Chromatographic separation was performed on CSH C18 column (Waters, USA). At positive ion mode with mobile phase A consisting 60% acetonitrile in water + 10 mM ammonium formate + 0.1% formic acid and mobile phase B consisting 90% isopropanol + 10% acetonitrile + 10 mM ammonium formate + 0.1% formic acid. At positive ion mode, with mobile phase A consisting 60% acetonitrile in water + 10 mM ammonium formate and mobile phase B consisting 90% isopropanol + 10% acetonitrile + 10 mM ammonium formate. The column temperature was maintained at 55°C. The gradient conditions were as follows: 40-43% B over 0~2 min, 43-50% B over 2-2.1 min, 50-54% B over 2.1-7 min, 54-70% B over 7-7.1 min, 70-99% B over 7.1-13 min, 99-40% B over 13-13.1 min, held constant at 99-40% B over 13.1~15 min and washed with 40% B over 13.1-15 min. The flow rate was 0.4 mL/min and the injection volume was 5 μL.

Q Exactive perform was used for primary and secondary mass spectrometry data acquisition. The stepped normalized collision energy was set to 15, 30 and 45 eV. All other parameters were set the same as that above.

After importing the off-line data of mass spectrometry into compound discoverer 3.2 (Thermo Fisher Scientific, USA, https://mycompounddiscoverer.com/) software and analyzing the mass spectrometry data in combination with bmdb (BGI metabolome database), mzcloud database and chemspider online database, a data matrix containing information such as metabolite peak area and identification results will be obtained. Paramenters were set as: parent ion mass deviation: < 5 ppm; mass deviation of fragment ions: < 10 ppm; retention time deviation: < 0.2 min.

After importing the off-line data of mass spectrometry into lipidsearch (v 0.4.1) (Thermo Fisher Scientific, USA, https://www.thermofisher.cn/order/catalog/product/OPTON-30880) software for mass spectrometry data analysis, the identification results and results of lipid containing molecules will be obtained, the data matrix of quantitative results and other information, and then the table will be further analyzed and processed. Paramenters were set as: identification type: product; search quality deviation of parent ion and daughter ion: 5 ppm; response threshold: Relative response deviation of sub ions5.0%; peak lifting quality deviation: 5ppm; m-score: 5.0; c-score: 2.0; identification level selection: “A”, “B”, “C”, “D”; positive ion mode adduct form: [M+H]⁺, [M+NH4]⁺, [M+Na]⁺; negative ion mode adduct form: [M-H]^-, [M-2H]^-, [M+HCOO]^-; retention time deviation: 0.1 min.

Metabolites with more than 50% missing values in QC samples and more than 80% missing values in experimental samples were removed, and missing values were filled with k-Nearest Neighbor Algorithm (KNN). Data were normalized to obtain relative peak areas by Probabilistic Quotient Normalization, PQN (Di Guida et al., 2016). Quality control-based robust LOESS (Dunn et al., 2011) signal correction was used to correct Batch effect, then metabolites with a coefficient of variation larger than 30% on their Relative peak area in QC Samples were removed.

Principal component analysis (PCA) was performed with R language (https://www.R-project.org/). R package Mfuzz (v 2.48.0) was used to cluster metabolite expression levels from samples at different time point (Kumar and Futschik, 2007). R package UpsetR (v 1.4.0) was employed to visualize intersecting sets (Lex et al., 2014). R package ggplot2 (v 3.3.6) was used to visualize the bar, bubble and volcano plots (Villanueva and Chen, 2019). For metabolic network mapping, the compound names were uploaded into the a web tool Chemical Translation Service (http://cts.fiehnlab.ucdavis.edu/batch) (Wohlgemuth et al., 2010) to obtain the PubChem Compound Identifiers (CID) and the Kyoto Encyclopedia of Genes and Genomes identifiers (KEGG ID). Metabolite enrichment analysis was performed with the web-based software MBROLE (v 2.0) (Lopez-Ibanez et al., 2016). Significantly changed metabolites matching the KEGG database were imputed, following overrepresentation analysis against the KEGG pathway module. Finally, annotations with q-values lower than 0.05 were considered significantly enriched.

To analyze mRNA expression, 0.1 μg of total RNA was used to reverse transcribe in order to construct cDNA using the First Strand cDNA Synthesis Kit (TOYOBO, Tokyo, Japan) according the manufacturer’s protocol. Quantitative real-time PCR was performed using the SYBR Green Real-time PCR Master Mix (Yeasen, Shanghai, China) and a LightCycler480 II system (Roche, Basel, Switzerland). Amplification was performed as follows: 50°C for 2 min, 95°C for 10 min, followed by 40 cycles at 95°C for 15 s, 60°C for 15 s, and 72°C for 30 s. The relative expression values of mRNAs were normalized to that of GAPDH in each sample using the 2^−ΔΔCt method. Primers used for detecting ferroptosis markers were shown in Table 1 .