The non-targeted metabolomic proling of liver reveals the molecular perturbations underlying clonorchiasis

Background: Clonorchiasis is a common food-borne parasitic disease worldwide. However, the molecular mechanisms of host-parasite interaction has not been illustrated. Given the parasite could inuence host metabolism, the hepatic metabolic proling will be useful for exploring the pathogenic mechanism of clonorchiasis. Results: The global hepatic metabolic prole of male Wistar rats at 4 and 8 weeks post infection with Clonorchis sinensis were investigated using liquid chromatography-mass spectrometry (LC-MS) metabolomics analysis. Multivariate analysis including principal components analysis and partial least squares discriminant analysis, were applied to clearly separated the metabolites of infected and control rats. These metabolites included lipids, hormones, peptides, alkaloids and amino acids. Metabolic pathways, such as purine and pyrimidine metabolism, primary bile acid biosynthesis and glycerophospholipid metabolism were perturbed during the infection process. Conclusions: These ndings showed signicant systemic metabolic perturbations of the liver in clonorchiasis. It also provided a new insight into the pathogenesis from the metabolic level, which might be helpful for making treatment strategies and prevention of clonorchiasis. by C.sinensis infection compared to controls are presented according to p-values from the pathway enrichment analysis (y-axis) and pathway impact values from pathway topology analysis (x-axis), with the most impacted pathways colored in red color.

alterations and perturbed metabolic pathways in the spleen of infected mice [18]. The metabolomics and associated biochemical pathways could improve the understanding of the pathophysiological mechanism in parasitic diseases. Our previous studies also found C. sinensis infection caused hepatic iron metabolism abnormity [19]. However, there is no study investigating the global hepatic metabolic pro ling or related metabolic pathways during C.sinensis infection. Especially, the function of some small-molecule metabolites in the host-parasite interactions has been poorly explained. Therefore, this present study aimed to identify hepatic metabolic pro ling and detect systemic metabolic pathways in C. sinensis infection animal models by a non-targeted liquid chromatography-mass spectrometry (LC-MS) metabolomics analysis ( Figure 1). Our ndings suggest that the metabolomics could provide a powerful approach for evaluating metabolic pro ling and elucidating the pathophysiological mechanism in clonorchiasis.

Characteristics of C. sinensis infection in rats
The C. sinensis eggs were detected in all infected groups by the Kato-Katz method. The egg detection rate could be detected at 4 weeks post infection 80% (16/20), and get to 100% (20/20) after 8 weeks post infection. The control group did not detected the eggs. Histopathological study of the livers from infected groups were characterized by epithelial hyperplasia, in ammatory in ltration of eosinophil and lymphocyte along the portal areas and hepatocytes hydropic degeneration (Fig. 2). The wall of bile ducts was thickened and the adult C. sinensis were found in some bile ducts. While, all control rats tissues did not present obvious pathological changes.

Metabolite pro ling of liver during C. sinensis infection
The metabolite pro ling in the rats liver during C. sinensis infection was present using by a non-targeted UHPLC-Q-TOF MS system. There were 11, 922 and 11, 296 ions identi ed in each sample pro le in ESI + and ESI − mode, respectively. After removing low-quality ions [relative standard deviation (RSD) > 30%], 11902 and 11289 ions in each sample were recognized in ESI + and ESI − mode, respectively. To assess the reproducibility of our data sets, QC samples were measured, clustered closely to each other and separated from the tested samples. However, the PCA scores plots did not clearly differentiate infected groups from control group in Fig 3 A  Differential metabolites during different infection periods PCA was used to determine the global similarity/differences of samples in the study. As show in Figure 5 (ESI + mode) and Additional le 1: Figure S1 (ESI − mode), the control group was found distinctly to the left, whereas the infected group (8 weeks post infection) cluster slightly to the right without overlapping with the control group, respectively. The parameters R2 and Q2 con rmed the validity of the PCA model in the Additional le 2: Table S1. The clear separation was also constructed in the heatmap in Additional le 3: Figure S2 (ESI + mode) and Additional le 4: Figure S3 (ESI − mode). According to infection time, the metabolites form different infection groups were categorized clearly.
The signi cant differential metabolites involved in clonorchiasis were identi ed, based on VIP > 1 and the P-value. An overview of identi ed altered metabolites detected both in negative and positive analysis mode with up-or down-regulation had be presented in Additional le 5: Table S2. And then a Venn diagram was generated to visualize the common and unique metabolites between the infected groups ( Figure 6). Especially, a total of 35 metabolites, included lipids, amino acids, peptides, antibiotics and alkaloids, were found common in the infected groups (Table 1). For example, choline and L-glutamine were signi cantly upregulated and cytosine, hydroxyproline were signi cantly decreased.

Metabolic pathways affected by C.sinensis infection
The biological pathways and roles involved in clonorchiasis were identi ed by MetaboAnalyst. As shown in Additional le 6: Table S3 and Additional le 7: Figure S4, Glycerophospholipid metabolism was found the most perturbed metabolic pathways among the infection groups, compared to the control group. While, other metabolic pathways included valine, leucine and isoleucine biosynthesis, Primary bile acid biosynthesis at 4 weeks post-infection group; linoleic acid metabolism and pyrimidine metabolism were identi ed at 8 weeks post-infection group. In addition, Arginine and proline metabolism, valine, leucine and isoleucine biosynthesis, purine and pyrimidine metabolism, alpha-Linolenic acid metabolism and glycerophospholipid metabolism involved in clonorchiasis were constructed a net-work in gure 7.

Discussion
Clonorchiasis is an important health problem worldwide [20]. However, the hepatic pathological mechanism of clonorchasis remained unclear. Metabolomics provide a powerful method to explore hostparasite interactions and discover the pathogenic mechanism at the biochemical level [21]. Recently, some studies had discovered endogenous metabolites and perturbed metabolic pathways in response to parasite infection [22]. However, it is still unclear whether C.sinensis infection cause metabolic change and dysregulation of biochemical pathways in the liver. In this study, a non-targeted UHPLC-Q-TOF MS method was applied to analysis the hepatic metabolite pro ling and related metabolic pathways during C. sinensis infection. Theses results will be useful for exploring the pathophysiology mechanism, making the diagnosis and prevention therapies for clonorchiasis.
Firstly, this present study indicated that C.sinensis infection can induce systemic metabolic perturbations in the liver. Theses metabolites belonged to lipids, fatty acids, amino acids, products of amino acid metabolim, nucleotides. These metabolites might be a direct signal of C. sinensis activity or the consequence of the host response to the parasite. In addition, a total of 35 features/metabolites were shared in the infected groups. As shown in Table 1, the levels of choline, L-glutamine, sn-glycerol-3phosphoethanolamine increased, and cytosine, glycocholic acid, hydroxyproline were found decreased. These dynamic metabolites is not only the interplay between the host and C.sinensis, but also could perturb the biochemical pro les of both the parasite and host [23].
The identi cation and quanti cation of amino acids had provided signi cant insights into the pathogenesis of various diseases, such as cancer and arthritis [24]. In this study, we found signi cant changes of amino acids and eicosanoid metabolites during C.sinensis infection. These metabolites might been required by the parasites to support their growth [25,26], and affect the host cell recruitment and proliferation during infection [27]. For example, leucine, as the branched-chain amino acids (BCAA), could provide energy when persistently energy consumption [28]. The up-regulate of leucine indicated C.sinensis need more energy to grow up. As an essential precursor for gluconeogenesis in the liver [29], the elevated level of alanine might indicate the impaired hepatic gluconeogenesis in clonorchiasis. While, S-Methyl-5'-thioadenosine participate the cysteine and methionine metabolism, as important participator in the tricarboxylic acid (TCA) cycle [30]. The decreased of S-Methyl-5'-thioadenosine indicated decreasing metabolic TCA cycle activity by the parasite. Glutamine is catabolized via TCA cycle and generatesγ-aminobutyric acid, which are essential for the growth of C.sinensis [31].
Additionally, our results also revealed that purine and pyrimidine metabolism were involved in clonorchiasis. Hydroxyproline origin from the post-translational hydroxylation of proline in proteins (primarily collagen) [31]. Both proline and hydroxylation have important metabolic, physiological and structural functions in collagens [32]. The decreased levels of hydroxyproline suggested that C.sinensis might facilitate its release from the liver cell of host.
The other identi ed signi cantly differential metabolites were lipids, including PC(6:0/6:0), alpha-linolenic acid (ALA), linolenic acid and glycerophosphocholine. Except the primary components of biological membranes, lipids also participate in signal recognition and transmission [33]. C. sinensis could utilize lipids for energy or physical requirements to survive longer, by up-regulating expression and enzymatic activity of CsACAT (an enzyme involved in fatty acid metabolism). The higher survival rate of adult worms was also reported in an environment with a higher level of lipids [34]. Therefore, the dysregulation of lipid metabolism suggested that the parasite modulated host cell environment to survival and used lipids to complete membrane biogenesis [35].
Because C.sinensis lack the oxygen-dependent pathways to synthesize fatty acids, it has to depend on its host for complex lipids [36]. For example, alpha-linolenic acid (ALA), an essential omega-3 fatty acid, is commonly found in seeds, nuts, and vegetable oils [37]. However, because ALA is an essential fatty acid (i.e., cannot be synthesized by the rat), its increased level must been derived from the rat diet. As to arachidonic acid (the main polyunsaturated fatty acid), it is released and metabolized to a series of eicosanoids, including the in ammatory leukotrienes and prostaglandins [38]. The up-regulation of arachidonic acid and its in ammatory metabolites mediated the regulation of cell survival, chemotaxis, migration, apoptosis and mitogenesis [39].
With regard to the metabolic pathway, linoleic acid metabolism and glycerophospholipid metabolism were found involved in clonorchiasis. Especially in glycerophospholipid metabolism, glycerophospholipids compose the main lipid category of mammalian cell membranes [40], which play a vital role in regulating transport processes, protein function and signal transduction [41]. Additionally, glycerophospholipids are also essential components of lipoproteins and in uence their function and metabolism [33]. The perturbed in these metabolites pathway play biological roles relevant to hostparasite interaction.
This study applies highly sophisticated technology to understand the impact of a neglected parasite on metabolism of the host. However this study has several limitations. Firstly, the main limitation was that samples were not collected from the control group over time. Thus, a larger samples with time-matched control groups should be assessed in future studies. Secondly, we only investigated metabolic changes in the liver. The serum or other tissue (spleen) might also be useful to address the pathogenesis of clonochiasis. Therefore, we will investigate the other tissue using metabolomics method, in order to explore the parasite-host interaction. Finally, the identi ed differential metabolites may likely participate in the pathogenesis of clonorchiasis, but the underlying molecular mechanisms still needs further investigations to reveal them.

Conclusion
Based on a non-targeted LC-MS/MS metabolomics method, we provide a new insight into the pathogenesis of clonorchiasis from the metabolic level. These differential metabolites mainly involved lipids (glycerophosphocholine) and amino acid (leucin and L-glutamine). Several critical metabolic pathways might play a great biological roles on the host-parasite interaction, including glycerophospholipid metabolism, purine and pyrimidine metabolism and primary bile acid biosynthesis. Theses results will be helpful for understanding the pathogenesis of clonorchiasis. In future, supplemental assays should be done to further investigate the metabolic liver alterations.

Methods
Parasites, animals and infection C.sinensis metacercariae were collected from Pseudorasbora parva captured in the Songhuajiang River of Heilongjiang Province. The process of collection and preparation is performed as described previously [42]. In brief, the sh were placed in an ice box at 0°C and transported to the laboratory. Next, the sh were washed with tap water, disrupted in a Waring Blender and digested with a pepsin-HCl (0.6%) arti cial gastric juice, at 37°C for 12 h. Then, the digested mixture was passed through three sieves with the mesh size of 1000 μm, 300 μm and 106 μm. After centrifugation, mass pure metacercariae were collected and kept in 0.1 M phosphate-buffered saline (PBS, pH 7.4) at 4°C until used.
Wistar rats (150g, 5-6 weeks old, male) were purchased from the Harbin Medical University Laboratory Animal Center. There was a rat in a single animal cage. The rats were housed with an inverse 12 hours day-night cycle with lights on at 8:30pm in a temperature (20±1℃) and humidity (55±5%) controlled room, and fed on standard laboratory food (CS-101; Liaoning Changsheng biotechnology co., Ltd) and water.
All the rats were randomly divided into 3 groups, 4 weeks post infection (wpi), 8 wpi and control groups. Animal model experiment were carried out in the afternoon. In the infection groups, the rats were randomly selected and individually infected orally with 50 metacercariae (n=10 per group). The Kato-Katz method was used to detect C. sinensis eggs from the rat feces [43]. Brie y, KK smears were prepared using 41.7 mg of feces, and the EPG (eggs per gram of feces) was determined by multiplying the number of eggs per smear by 24. Control group rats (n=10) were fed with 50 μl of sterile normal solutio.
At 4 wpi and 8 wpi, the rats were sacri ced under ether anesthesia by cervical dislocation and livers were quickly removed and rinsed with saline solution (0.9% NaCl w/v), and stored at −80°C for analysis. And then livers from control rats were collected at 8 wpi. Inhalation anesthesia (ether anesthesia) reduced animal suffering and increased the success rate of anesthesia. Brie y, every rat was put into a glass anesthesia box containing several ether-soaked cotton balls (2-4ml). After slow breathing, decreased limb tensioning and dull corneal response, the rats were sacri ced by cervical dislocation.
For histology, liver samples were xed by 4% paraformaldehyde, then dehydrated in a graded series of ethanol and embedded in para n wax. Sections were made (5μm) using a microtome, then stained with hematoxylin-eosin (H&E) and examined under an optical microscope (Olympus, Tokyo, Japan).

UHPLC-Q-TOF MS analysis
Analyses were performed using an UHPLC (1290 In nity LC, Agilent Technologies) with a quadrupole time-of-ight mass spectroscopy (AB Sciex Triple TOF 5600) in Shanghai Applied Protein Technology Co., Ltd.
In the HILIC separation process, we utilized a 2.1 mm × 100 mm ACQUIY UPLC BEH 1.7 µm column (waters, Ireland) to analysis the samples. The mobile phase contained A=25 mM ammonium acetate and 25 mM ammonium hydroxide in water and B= acetonitrile. The gradient was set at 85% B (1 min), reduced to 65% in 11 min, reduced to 40% in 0.1 min (hold on 4 min), increased to 85% in 0.1 min, and then re-balanced in 5 min.
The ESI source situation were used as below: Curtain gas (CUR) as 30, source temperature: 600℃, Ion Source Gas1 (Gas1) as 60, Ion Source Gas2 (Gas2) as 60, IonSpray Voltage Floating (ISVF) ± 5500 V. The instrument was set over the m/z range 60-1000 Da, with the accumulation time of TOF MS scan (0.20 s/spectra) in MS analysis. Over the m/z range 25-1000 Da, product ion scan was set at 0.05 s/spectra in auto MS/MS acquisition. With the advantages of high sensitivity mode, information dependent acquisition (IDA) was applied to acquire the product ion scan. The condition was set as below: the collision energy (CE) was set at 35 ± 15 eV; Declustering Potential (DP), 60 V (+) and 60 V (−); exclude isotopes within 4 Da, candidate ions to monitor per cycle: 10.
During mass spectra collection, samples were placed in automatic sampler at 4℃. To monitor the stability and repeatability of analytical system, quality control (QC) samples were prepared by pooling 10 μL of each sample and injected prior to analysis. And then blank and QC samples were injected every 5 samples injections throughout the analytical run.

Data processing
The raw MS data (Wiff format les) were converted to mzXML format using ProteoWizard (http://proteowizard.sourceforge.net/) and processed using R package XCMS (version 3.2). For peak picking, the parameters were set: centWave m/z = 25 ppm, peak width = c (10, 60), pre lter = c (10, 100). For peak grouping, bw = 5, mzwid = 0.025, minfrac = 0.5 were used. R i836 3.1.1 was used to run XCMS and Collection of Algorithms of Metabolite Pro le Annotation (CAMERA, 1.42.0). The code was used to perform data processing showed in supporting information. The parameters was loaded on supporting information.The variables with more than 50% of the nonzero measurement values in at least one group were identi ed. Compound identi cation of metabolites was performed by comparing the accuracy of m/z values (< 25 ppm), and MS/MS spectra were interpreted with an in house database (Shanghai Applied Protein Technology Co., Ltd.) established with authentic standards.

Statistical analysis
After normalized to total peak intensity, the processed data were uploaded into SIMCA-P (version 14.1, Umetrics, Umea, Sweden). The data were mean-centered and Pareto-scaled before multivariate statistical analysis. PCA, an unsupervised multivariate statistical method, could provide an overview of all observations in data tables, such as groupings, trends, and outliers of the different groups of samples. Then, the supervised partial least-squares discriminant analysis (PLS-DA) were applied to discriminate the different variables between groups. A default seven-fold cross-validation and testing with 200 random permutations were used to avoid the over-tting of supervised PLS-DA models. The goodness-of-t parameters for the PLS model, R2 and Q2, were calculated which varied from 0 to 1. The values of R2 and Q2 parameters were used to verify the tness and predictive ability of the model. The variable importance in the projection (VIP) value was applied to clarify its contribution to the classi cation in the PLS-DA model. Fold change (FC) was set to be ≥ 1.2 or ≤ 0.8333. Log2 FC based on metabolite abundance was used to demonstrate how the differential abundance of liver metabolites varied at different stages of infection groups.
Heatmaps were employed to describe the unbalanced metabolic pro ling among C.sinensis infected and control rats. Euclidean distance algorithm for similarity measure and average linkage clustering algorithm (clustering uses the centroids of the observations) for clustering were selected when performing hierarchical clustering. Based on the deferentially expressed metabolite data (log2-scaled), heatmaps were structured by the MultiExperiment Viewer (MeV) v. 4.9 software (http:// mev.tm4.org/).
The Pathway analysis of the differentially expressed metabolites was depended on KEGG database (http://www.genome.jp/kegg/) and MetaboAnalyst 3.0 (http://www.MetaboAnalyst.ca/). According to our research background and the pathway enrichment results, the corresponding KEGG pathways were extracted. The analyses were applied based on the sher' exact test, considering the metabolites as background data sets. And only pathways with FDR-values under a threshold of 0.05 were considered as signi cant. The identi ed pathways in uenced by C.sinensis infection compared to controls are presented according to p-values from the pathway enrichment analysis (y-axis) and pathway impact values from pathway topology analysis (x-axis), with the most impacted pathways colored in red color. In the study, we made great efforts to reduce animal suffering and the number of animals.

Consent for publication
All authors have given their permission to publish data presented in this manuscript.

Availability of data and materials
All the data supporting our ndings is contained within the manuscript.

Competing interests
The authors declare that they have no competing interests.  Experimental ow chart.     Venn diagram displaying the common and unique metabolites between two group. There were 35 signi cantly differential metabolites were shared in infected groups. R4, 4 weeks-post infection, R8, 8 weeks-post infection, C0, healthy control.