The Medical Ethics Review Committee of Harbin Medical University reviewed and ethically approved this study. All procedures with animals were carried out based on the National Guidelines for Experimental Animal Welfare (Ministry of Science and Technology of the People’s Republic of China, 2006) and were approved by the Animal Welfare Committee of Harbin Medical University. Additionally, significant efforts to reduce animal suffering and the number of animals were made.

Metacercariae of C. sinensis were obtained from infected Pseudorasbora parva caught from the Songhuajiang River of Heilongjiang Province. The method of collecting and purifying metacercariae are showed as below. We placed the fish in an ice box at 0°C and then transported to the laboratory. Next, we washed the fish with running water, crushed the tissue with a Waring Blender, and then digested with artificial gastric juice (0.5% pepsin in 1% HCl) at 37°C for more than 12 h. At last, the digested mixture was filtered three times, passed through sieves of 1000, 300, and 106 µm. Metacercariae were isolated by centrifugation, collected and preserved at 0.1 M phosphate buffered saline (PBS, pH = 7.4) at 4°C. A total of 12 wistar rats (specific pathogen free, SPF), aged about 6-8 weeks were bought from the Harbin Medical University Laboratory Animal Center. All rats had free access to standard chow and water to acclimate to their environment for 1 week before modeling. All rats were randomly divided into two groups, infected group (n=6) was individually infected orally with 100 C. sinensis metacercariae, control group (n=6) was administered with 100 µL of sterile normal solution as controls. The blood samples were collected from tail vein at 4 and 8 weeks post infection (wpi) with C. sinensis, respectively. All animals were weighed and the food intake of rats was monitored at 4 wpi and 8 wpi.

The Kato-Katz method was used to observe whether the rats had successfully infected with C. sinensis. A total of 41.7 mg of feces were collected. The number of eggs per smear was converted to the eggs per gram of feces with multiplying by 24 (Hong et al., 2003). The rats were sacrificed at 8 wpi, C. sinensis adults were collected from bile ducts of infected rats and flushed with saline. Worm recovery rate was determined by dividing the recovered worm number by number of metacercariae infected.

The contents of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) in serum were measured with commercial reagent kits (Nanjing Jiancheng Bioengineering Institute, Jiangsu, China). The experimental procedure was processed following manufacturer’s instructions. Additionally, the level of inflammatory factors (IL-1β, IL-6, IL-10 and TNF-α) were detected by Luminex 200 liquid suspension microarray and Bio-Plex Pro Rat Cytokine 1 10PLX EXP Kit. The experimental procedure briefly described as follows: 50 µL diluted mixed micro-beads was added into wells, then 50 µL standard products and samples were added to each well (diluted twice with assay buffer in advance). The plate was incubated with shaking at room temperature for 60 min and washed three times. After that, re-incubated with 25 µL of biotin-labeled detection antibody for 30 min and washed three times. After washing, the samples were re-incubated in 50 µL diluted temperature for 10 min at room temperature. Following another washed, the magnetic beads were re-suspended with 125 µL Assay Buffer. Each sample was measured in duplicate. Data were analyzed and statistics calculated with GraphPad Prism 4 software (San Diego, CA, USA).

The blood samples obtained at 4 wpi and 8 wpi were centrifuged at 3,000 g for 10 min at 4°C. The serum samples were frozen at −80°C until used for metabolite extraction. The frozen serum samples (50 μL) were placed in an EP tube, and added 200 μL extract solution (acetonitrile: methanol = 1: 1) including isotopically-labelled internal standard mixture. The mixture was vortexed for 30 s, and sonicated in ice water bath for 10 min. The mixture was kept at -40 °C for 1 h, followed by centrifugation at 12000 rpm for 15 min at 4 °C. The supernatant (200 μL) was transferred to a fresh tube and then dried in a vacuum concentrator at 37 °C. The samples were dissolved in 100 μL of 50% acetonitrile, vortexed for 30 s, and sonicated on ice for 10 min. After centrifugation at 13000 rpm for 15 min at 4°C, 75 μL of supernatant was placed in a fresh glass vial for LC/MS analysis. The quality control (QC) sample was collected by Mixing an equal aliquot of the supernatants from all samples.

LC-MS/MS analyses were carried out using a ExionLC Infinity series UHPLC System (AB Sciex). Compounds were separated by HPLC on a Waters ACQUITY UPLC BEH Amide (2.1 * 100 mm, 1.7 μm, Waters). The mobile phase consisted of solvent A (25 mmol/L ammonium acetate and 25 mmol/L ammonia hydroxide in water, pH = 9.75) and solvent B (acetonitrile). The column temperature was maintained at 25°C, and the auto-sampler temperature was 4°C. The gradient elution procedure was run as below: 0~0.5 min, 95% B; 0.5~7.0 min, 95%~65% B; 7.0~8.0 min, 65%~40% B; 8.0~9.0 min, 40% B; 9.0~9.1 min, 40%~95% B; 9.1~12.0 min, 95% B. The injection volume was 2 μL (pos) or 2 μL (neg), respectively, and the flow rate was set at 0.5 mL/min.

The TripleTOF 5600 mass spectrometry (AB Sciex) was applied to acquire MS/MS spectra based on information-dependent acquisition (IDA) in LC/MS analyses. The data acquisition software (Analyst TF 1.7, AB Sciex) automatically the complete scan survey MS data as it collects and triggers the acquisition of MS/MS spectra according to preselected criteria.

For each cycle, 12 precursor ions with the strongest intensity and greater than 100 were selected for MS/MS. The collision energy (CE) was 30 eV, and the cycle time was 0.56 s. ESI source parameters were set as below: Gas 1: 60 psi, Gas 2: 60 psi, Curtain Gas: 35 psi, Source Temperature: 600 °C, Declustering potential: 60 V, Ion Spray Voltage Floating (ISVF): 5000 V in positive modes or -4000 V in negative modes.

The raw data (.wiff) files generated by MS were converted into mzXML format using ProteoWizard. Following conversion, data was processed with R package XCMS(version 3.2) to obtain peak deconvolution, alignment and integration. Minfrac and cut off were set as 0.5 and 0.3 respectively. Metabolites were identified using an internal MS2 database (m/z and retention). Before multivariate statistical analyses, the data were normalized by total area normalization, and the raw peak area for each metabolite was divided by the total peak area of all metabolites to calculate the relative abundance. Then the data were mean-centered and scaled using UV scaling or pareto scaling. After that, the data was imported to SIMCA-P (version 13.0, Umetrics, Umea, Sweden) for multivariate statistical analyses, including principal component analysis (PCA) and orthogonal partial least-square discriminant analysis (OPLS-DA). PCA and OPLS-DA were performed to distinguish infected from control group. The quality of the models was assessed based on the value of R2 and Q2. The variable importance in the projection (VIP) value of OPLA-DA mode was used to illustrate its contribution to the classification.

Student’s t test was used to test the statistical significance differences, and P-values (q-value) < 0.05 was considered as statistically significant. In order to observe the distinctions in metabolic state between infected group and control group, log₂ transformation was performed for cluster analysis. Log2 fold change (FC) was used to assess the changes in the abundance of metabolites between different groups. In this study, based on other studies (Zhou et al., 2017; Liu et al., 2019; Zheng et al., 2019), VIP > 1, P <0.05 and FC ≥ 1.2 and < 0.833 were used to select differential metabolites. Heatmaps were used to depict the relatively disorder and unbalanced metabolic profile of C. sinensis-infected rats compared to control rats. On the basic of abundance of differentially expressed metabolites data (log2-scaled), MultiExperiment Viewer (MeV) v. 4.9 software (http://mev.tm4.org/) was used to generate heatmaps. Metabolite pathways of the differentially expressed metabolites in both ion modes were analyzed using KEGG database (http://www.genome.jp/kegg/) and MetaboAnalyst 3.0. The pathways influenced after infection were obtained, based on the p-values from the pathway enrichment analysis (y-axis) and pathway impact values from pathway topology analysis (x-axis).

A total of 25 amino acids in serum samples were determined by targeted UPLC-MS/MS method, due to their significant differential abundance from the untargeted metabolomics analysis. Almost 50 μL serum sample was employed for metabolite extraction, then 250 μL acetonitrile/methanol/formic acid (74.9:24.9:0.2,v/v/v) was added to the serum, which contained additional stable isotope labeled internal standards of valine-d8 and phenylalanine-d8. The mixture was vortexed for 1 min and kept at room temperature for 10 min, and then centrifuged at 14,000 g for 10 min at 4 °C. After that, the supernatant was placed into the sampling vial pending UPLC-MS/MS analysis.

The targeted analysis was performed on a Waters ACQUITY UPLC system (Waters Corporation, Milford, MA, USA) coupled with a Waters Xevo TQD Mass Spectrometer (Waters Corporation, Manchester, UK). The separation was performed on ACQUITY UPLCTM HILIC column (100 mm × 2.1 mm i.d., 1.7 μm; Waters Corporation, Milford, MA, USA). The UHPLC/MS-MS program was performed as follows. Briefly, the sample injection volume was 2 μL, and the flow rate was 300 μL/min. Substances were recovered from the column by gradient elution (Liu et al., 2015). ESI and Selected Reaction Monitoring (SRM) scans in the positive ion mode were used for mass spectrometric detection. In each transition, cone voltage and collision energies were optimized, with the ion spray voltage 3.2 kV, and the source temperature 150 °C. Internal standard peak areas (valine-d8 and phenylalanine-d8) were monitored for quality control and individual samples with peak areas differing from the group mean by more than two standard deviations were reanalyzed (Additional Files: Table S1 ). MarkerLynx Application Manager software (Version 4.1; Waters Corporation, Milford, MA, USA) was used for automated peak integration. The metabolite peaks were manually checked for the quality of integration.

For phosphatidylcholines analysis, 10 μL serum sample was mixed with 190 μL water and 480 μL extract solution (MTBE: MeOH = 5: 1) containing internal standards. The mixture was vortexed for 60s, and sonicated for 10 min. Then the mixtures were centrifuged at 3000 rpm for 15 min at 4°C. A total of 250 μL supernatant was transferred to a fresh tube. The remaining sample was mixed with 250 μL MTBE, followed by vortex, sonication and centrifugation. After that, another 250 μL supernatant was taken out and repeated the above steps once. Subsequently, supernatants were collected and combined and dried in a vacuum concentrator at 37 °C. The dried samples were reconstituted with 100 μL of resuspension buffer (DCM: MeOH: H2O = 60: 30: 4.5), then vortexed (30 s), sonicated (10 min), and centrifuged (12000 rpm for 15 min at 4°C). Finally, 30 μL supernatant was taken out for LC/MS analysis using a UHPLC system (1290, Agilent Technologies) with a Phenomen Kinetex C18 column (2.1 × 100 mm, 1.7 μm) coupled to TripleTOF 6600 mass spectrometry (AB Sciex). The quality control (QC) sample was prepared by mixing the same volume of supernatants from all samples.

Targeted serum lipidomics analysis was performed by a SCIEX ExionLC series UHPLC System (ACQUITY UPLC HSS T3, 1.8 μm, 2.1 × 100 mm) coupled with AB Sciex Trap 6500+ mass spectrometer (AB Sciex). The mobile phase consisted of solvent A (40% water, 60% acetonitrile, and 10 mmol/L ammonium formate) and solvent B (10% acetonitrile, 90% isopropanol, and 10 mmol/L ammonium formate). The column temperature was maintained at 45°C, the auto-sampler temperature was 6°C, and the injection volume was 2 μL. Ion source parameters were set as below: ISVF: +5500/-4500 V, Curtain Gas: 40 psi, Temperature: 350°C, Ion Source Gas1: 50 psi, Ion Source Gas2: 50 psi, DP: ± 80V. The target compounds were quantitatively analyzed by Biobud-v2.1.4.1 software. Based on the peak area and the actual concentration of the same lipid internal standard (IS), the absolute concentration of lipids was calculated.

For the data obtained by targeted analysis, the results were expressed as means ± standard deviation (SD), SPSS Statistics software version 22 (IBM, Armonk, NY, USA) was used for statistical analysis, and Prism version 8.0.1 (GraphPad Software, San Diego, CA) was used to generate box-plots (*P <0.05, **P <0.01). To identify the important metabolites of C. sinensis infection, the area under the curve was assessed in receiver operating characteristic analyses.

Infection was identified by the detection of C. sinensis eggs in rat feces from all infected rats ( Figure S1 ). Rats were sacrificed at 8 wpi, the mean worm recovery rate in the infected group was over 50%. No worms were recovered from the control group. At 4 wpi, the food intake of rats in infection group was 8.64 ± 0.39 g, the control group was 8.99 ± 0.48 g; at 8 wpi, the food intake of rats in infection group was 24.02 ± 0.36 g; the control group was 25.30 ± 0.95 g. There was no significant difference between two groups at the two time points, respectively. Additionally, the mean body weight did not significantly differ between infected (4 wpi 294.64 ± 12.09 g and 8 wpi 395.67 ± 13.83 g) and control rats (4 con 298.5 ± 11.95 g and 8 con 403.03 ± 13.56 g), respectively. As shown in Figure S2 , ALT and AST increased both at 4 wpi and 8 wpi in the infected group (P <0.05). The level of IL-6, IL-10 and TNF-α were significantly increased both at 4 wpi and 8 wpi (P <0.05), while there was no significant change in IL-1β at two time points (P > 0.05).

We characterized serum metabolomic changes in C. sinensis-infected and control groups using UHPLC-QTOF-MS. A total of 10530 and 6560 ions were identified in the positive electrospray ionization (ESI+) mode and negative electrospray ionization (ESI−) mode, respectively (Additional Files: Table S2 and S3 ). Next, PCA score plots were first employed to depict the clustering behavior of serum metabolic profiles in different infection times. Although the PCA score plots showed a clear separation at 8 wpi, it could not clearly separate the 4 wpi group from the control ( Figure S3 ). Next, the OPLS-DA analysis was performed and showed good discrimination among different infected rats groups in both ESI+ and ESI− modes ( Figure 1 ). Additionally, the PCA and OPLS-DA model parameters are present in Table S4 . These results suggest that the metabolic patterns of C. sinensis-infected rats are different from control rats.

Based on the VIP > 1, P <0.05 and FC ≥ 1.2 and < 0.833, 522 and 3190 metabolites of serum were significantly altered at 4 and 8 wpi in the ESI+ mode, respectively. In addition, a total of 364 and 2332 metabolites were altered at 4 and 8 wpi in the ESI− mode, respectively. The volcano and heat map plots of these metabolites are present in Figure 2 . As shown in heat map, the serum metalites of C.sinensis-infected rats significantly deviated from corresponding controls in ESI+ and ESI− mode. Meanwhile, the concentration of differential metabolites from 8 wpi group changed more significantly than those from 4 wpi group.

After removing the unannotated metabolites, differential metabolites in each group are shown in Table S5 . The common and unique differential metabolites among four groups are shown by Venn diagrams ( Figure S4 ). A total of 10 different metabolites were shared in the two infected groups in ESI+ mode, while only 5 metabolites were shared in ESI− mode. Both in the ESI+ mode and ESI− mode, the number of differential metabolites identified between the 8 wpi and 8 con group was larger. These results showed that the number of differential metabolites increased during the infection progresses.

In order to validate the untargeted metabolomics analysis, 25 amino acids in serum samples were further quantitatively analyzed by targeted UPLC-MS/MS. Among them, 14 amino acids were significantly changed during C. sinensis infection ( Figure 3 ). For example, aminobutyric acid, dimethylglycine, glutamate, and 4-Hydroxy-L-proline decreased at 4 wpi and 8 wpi. The branched-chain amino acids (BCAAs, including isoleucine, leucine and valine), citrulline and creatine increased at 4 wpi and decreased at 8 wpi. We compared the results between targeted and untargeted amino acids in Table S6 . It was shown that glutamate and glutamine decreased at 8 wpi, and threonine increased at 8 wpi in both methods. Then metabolic pathway analysis of these amino acid metabolites was subsequently performed (Additional Files: Table S7 ). Based on P value < 0.05 and impact value > 0.1 (Wang et al., 2021), the main enrichment pathways included glycine, serine and threonine metabolism, arginine biosynthesis, D-glutamine and D-glutamate, glyoxylate and dicarboxylate metabolism, glutathione metabolism, and alanine, aspartate and glutamate metabolism. Additionally, the ROC analysis showed the AUC values for glutamate, dimethylglycine, γ-aminobutyric acid and glutamine of 0.938, 0.813, 0.715 and 0.694, respectively ( Figure 4A ).

Some lipids were also identified as the differential metabolites in untargeted metabolomics. Thus, we further analyzed the changes in the concentration of phosphatidylcholines (PCs) in serum. A total of 40 PCs were identified and 17 PCs were found differentially expressed at 4 wpi and 8 wpi ( Figure 5 ). The levels of PC(16:0/20:1), PC(16:0/22:4), PC(16:0/22:5), PC(18:0/16:1), PC(18:0/20:3), PC(18:0/22:4), PC(18:0/22:5), PC(18:1/20:3), and PC(18:2/20:1) showed increased at two time points, whereas the levels of PC(16:0/18:3), PC(18:0/18:2) and PC(18:2/18:2) were increased at 4 wpi, but decreased at 8 wpi. Moreover, PC(16:0/20:2), PC(16:0/20:3), PC(18:0/20:2), and PC(18:1/18:2) showed increased only at 4 wpi, while PC(18:0/20:1) increased only at 8 wpi. According to AUC analysis, PC(18:2/20:1), PC(18:0/22:5), PC(16:0/20:1), PC(16:0/20:3), PC(18:0/16:1) and PC(18:0/20:3) had high AUC values (AUC > 0.7), which indicated that these PCs had good predictive ability for disease ( Figure 4B ).

The biological metabolic pathways of the differentially expressed metabolites were determined by enrichment analysis using KEGG annotation and MetaboAnalyst. Based on P value < 0.05 and impact value > 0.1, the most significant differential metabolic pathways at 4 wpi were glycerophospholipid metabolism and taurine and hypotaurine metabolism; while, glycerophospholipid metabolism, histidine metabolism, arginine biosynthesis, alanine, aspartate and glutamate metabolism, and pyrimidine metabolism were significantly changed at 8 wpi, as shown in Figure 6 and Table S8 .

We further made the pathway analysis of up-regulated and down-regulated metabolites (Additional Files: Table S9 ), respectively. At 4 wpi, up-regulated metabolites were enriched in glycerophospholipid metabolism and taurine and hypotaurine metabolism, while at 8 wpi, up-regulated metabolites were enriched in fructose and mannose metabolism. Down-regulated metabolites were enriched in glycerophospholipid metabolism at 4 wpi, and down-regulated metabolites were enriched in arginine metabolism, histidine metabolism, D-glutamine and D-glutamate metabolism, nicotinate and nicotinamide metabolism, pyrimidine metabolism, alanine, aspartate and glutamate metabolism, glycerophospholipid metabolism at 8 wpi. It seems that both up-regulated and down-regulated metabolites are enriched in glycerophospholipid metabolism. Finally, alanine, aspartate and glutamate metabolism, histidine metabolism, glycerophospholipid metabolism, and pyrimidine metabolism were used to construct integrated metabolic networks ( Figure 7 ).