Research Ethics

This study was approved by the Institutional Review Board of Jinshan Hospital, Fudan University (No. 2017-01). All procedures were in accordance with the guidelines of the Institutional Animal Care and Use Committee of Fudan University.

Laboratory animals and models

Twenty-four male Sprague-Dawley rats (weighing 150-200 g) were divided into three subgroups: the MHE rats (n=8), the ZN rats (MHE treated by zinc sulfate, n=8), and the CN rats (sham operation controls, n=8). MHE rats were modeled by partially ligating the portal vein with the surgical procedures described in a previous report [10]. The ZN rats were provided zinc sulfate with a dose of 30 mg/kg/day elemental zinc (150 mg of zinc sulfate dissolved in 500 mL of deionized water) after partially portal vein ligation [7]. The dissolved zinc sulfate doses were adjusted weekly according to the rat’s body weight and water consumption. The SN rats were given deionized water without any treatment. All rats were placed in metabolic cages and housed in a well-ventilated room with temperature (25±2 ℃), air humidity (50±10%) and a light/dark cycle of 12 h. All rats were allowed to drink and eat freely for 56 days. Ain-93 experimental animal standard rat purified feed were used to fed all the rats.

Cognitive Assessment

Morris water maze (MWM) were used for cognitive assessment on the day 57-60. A black round pool with 180 cm diameter and 50 cm height were filled with water (30 cm from the bottom). The water temperature was adjusted to 24±1 ℃. The pool was divided into four quadrants (left top, right top, left bottom, and right bottom). A circular transparent platform of 12 cm was placed in the right top with 2 cm below the water surface. The rats were randomly placed from one quadrant into the water. Each rat had 60 s to find the platform. the time of the rats climbed the platform was record as the latency. If the rats were unable to reach the platform within 60 s, they would be guided to the platform. All the rats on the platform were allowed to rest for 20 s. All the rats were trained for 4 consecutive days. The tracks of the rats were recorded by a computer equipped with a camera.

 Sample Preparation, Blood Ammonia and Brain Mn Content Measurements

After the MWM test, the rats were killed by intravenous injection of KCl (1-2 mg/kg) under deep anesthesia by isoflurane. The brain tissues were removed, cleaned by cold saline, and died by filter papers. The striatums were separated on a frozen plate according to the anatomical map of the rat brain. The samples were stored in liquid nitrogen and then transferred to a -80℃ refrigerator for further analysis.

1H-NMR Metabolomics

Samples were weighed (10-20 mg) and added with 0.6 mL ice-cooled extraction solution (CH3OH: H2O=2:1). After vortexing for 1 min, the sediments were broken by a tissue crusher (30 Hz, 90 s). The samples were then centrifuged (4 °C, 10,000×g, 10 min), rotary evaporated, and freeze-dried for 24 h. The solid residues were added with 600 μL buffer solution (0.1 M NaH2PO4/K2HPO4, pH: 7.4). D2O (99.96% deuteriumenriched containing 0.05% TSP) was used as a chemical shift reference. Then samples were then transferred into 5 mm NMR tubes.

The 1H-NMR was performed by a 600 MHz NMR spectrometer (AVANCE III, Bruker, German) with Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence and water suppression. The basic parameters were as follows: pulse=90 °; water presaturation during the relaxation delay=2.5 s; width=10 ms; sampling points=32 K; spectral width=20 ppm; sampling time=1.36 s; scanning times=64; empty scanning times=; temperature=298 K. 1H-NMR free induction decays data were preprocessed in a previous reported pipeline [11]. Briefly, group delay correction, solvent suppression, apodization, fourier transform, zero-order phase correction, internal referencing, baseline correction, negative values zeroing, warping, window selection, bucketing, water region removal, zone aggregation and normalization were performed in sequences. The identification and quantification of the metabolites were performed by using a spectroscopy of 191 pure metabolites as a reference. Metabolites within 2 standard deviations of the background noise were not included in the further analysis.

Supervised analysis (sparse projection to latent structures discriminant analysis, sPLS-DA with 3-fold cross-validation) was used for pattern recognition analysis of the significantly affected metabolites in MHE vs. ZN rats and MHE vs. CN rats. The metabolites with variable influence on projection (VIP) score > 1 and P < 0.05 were defined as the key metabolites in MHE vs. ZN rats, and in MHE vs. CN rats, respectively.

A diffusion algorithm was used for enrichment analysis of the key metabolites in MHE vs. ZN rats, and in MHE vs. CN rats. The top-ranked significant different biological pathways were also plotted based on the KEGG database.

 Statistical Analysis

All the statistical analyses were performed in R (Version 4.0.2; http://www.r-project.org/). Data met normality and variance homogeneity were measured by ANOVA followed by false discovery rate (FDR) correction. If not met normality or variance homogeneity, the data were measured by Mann-Whitney U test followed by FDR. The "mixOmics" package was used for sPLS-DA analysis and VIP score calculation; the "FELLA" package was used for enrichment and pathway analysis. P<0.05 was considered statistically significant.