Ursodeoxycholic acid improves liver function via phenylalanine/tyrosine pathway and microbiome remodelling in patients with liver dysfunction

Ursodeoxycholic acid (UDCA) is a metabolic by-product of intestinal bacteria, showing hepatoprotective effects. However, its underlying molecular mechanisms remain unclear. The purpose of this study was to elucidate the action mechanisms underlying the protective effects of UDCA and vitamin E against liver dysfunction using metabolomics and metagenomic analysis. In this study, we analysed blood and urine samples from patients with obesity and liver dysfunction. Nine patients were randomly assigned to receive UDCA (300 mg twice daily), and 10 subjects received vitamin E (400 IU twice daily) for 8 weeks. UDCA significantly improved the liver function scores after 4 weeks of treatment and effectively reduced hepatic deoxycholic acid and serum microRNA-122 levels. To better understand its protective mechanism, a global metabolomics study was conducted, and we found that UDCA regulated uremic toxins (hippuric acid, p-cresol sulphate, and indole-derived metabolites), antioxidants (ascorbate sulphate and N-acetyl-L-cysteine), and the phenylalanine/tyrosine pathway. Furthermore, microbiome involvement, particularly of Lactobacillus and Bifidobacterium, was demonstrated through metagenomic analysis of bacteria-derived extracellular vesicles. Meanwhile, vitamin E treatment did not result in such alterations, except that it reduced uremic toxins and liver dysfunction. Our findings suggested that both treatments were effective in improving liver function, albeit via different mechanisms.


Bile acid profiling
Plasma samples were prepared according to the manufacturer's instructions. For liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis, 10 μL of a diluted filtrate was injected into an Agilent 1200 series high-performance liquid chromatography (HPLC) system (Agilent Technologies, Santa Clara, CA, USA), consisting of a binary pump (G1312A) and an autosampler (G1367B), with a thermostat (G1330B) set at 10 °C. Separation was achieved on an analytical column using the Biocrates® bile acids (BAs) kit (Biocrates Life Science AG, Innsbruck, Austria), equipped with an AJ0-4287 SecurityGuard™ ULTRA cartridge for a C18 HPLC column (Phenomenex, Torrance, CA, USA). Mobile phase A comprised 10 mM ammonium acetate (NH4Ac) and 0.015% formic acid in Milli-Q® water, and mobile phase B comprised 10 mM NH4Ac and 0.015% formic acid in acetonitrile:methanol: v/v/v). An API 4000 QTRAP (Applied Biosystems/MDS Sciex, Foster City, CA, USA), equipped with an electrospray ionisation source, was used for MS analysis. Seventeen BAs were identified and quantified using the LC-MS/MS system (scheduled multiple reaction monitoring). The abundance of each BA was calculated from the area under the curve by normalisation to its respective isotope-labelled internal standard using the Analyst® 1.5.2 software (Applied Biosystems/MDS Sciex). Calibration curves, quality controls, and samples were evaluated using the MetIDQ™ software (Biocrates). Outliers were removedif samples from more than 50% of the subjects showed values below the limit of detection. A heatmap view was generated using MetaboAnalyst; each concentration was transformed into a log scale for visualisation.

Extracellular vesicle isolation and DNA extraction from human urine samples
Extracellular vesicles (EVs) from human urine samples were isolated using differential centrifugation at 10,000 × g for 10 min at 4 °C . After centrifugation, bacteria and foreign particles were thoroughly eliminated by sterilising the supernatant by filtration through a 0.22-μm filter. To extract DNA, EVs were boiled for 40 min at 100 °C . To eliminate remaining floating particles and debris, the supernatant was collected after 30 min of centrifugation at 13,000 rpm at 4 °C, and EV DNA was extracted using a PowerSoil DNA isolation kit (MO BIO Laboratories, Inc., Carlsbad, CA, USA) following the standard protocol. DNA was quantified using a QIAxpert system (Qiagen, Germany).

Bacterial metagenomic analysis using EV DNA from human urine samples
primers, which are specific for the V3-V4 hypervariable regions of the 16S rRNA gene. Libraries were constructed using polymerase chain reaction products according to the MiSeq system guide (Illumina, San Diego, CA, USA) and quantified using the QIAxpert system (Qiagen). Amplicons were quantified, pooled at an equimolar ratio, and then sequenced using the MiSeq system (Illumina) according to the manufacturer's recommendations.

Analysis of bacterial composition of the microbiota
Raw pyrosequencing reads were filtered according to the barcode and primer sequences using the MiSeq system (Illumina). Taxonomic assignment was performed using the profiling program MDx-Pro ver. 1 (MD Healthcare, Korea). This program allows the selection of high-quality sequencing reads after examining the read length ( 300 bp) and the quality score (average Phred score  20). Operational taxonomic units were clustered using the sequence clustering algorithm CD-HIT. Subsequently, taxonomic assignment was carried out using UCLUST and QIIME against the 16S rDNA sequence database in GreenGenes 8.15.13. Based on sequence similarities, all 16S rDNA sequences were assigned at several taxonomic levels. The bacterial composition at each level was plotted as a stack bar.
If clusters could not be assigned at the genus level because of the lack of sequences or redundant sequences in the database, the taxon was assigned at a higher level, which is indicated in parenthesis. Plots of the relative taxon abundance for samples collected before and after UDCA or vitamin E treatment, summarised at the genus level. Individual samples are presented along the horizontal axis, and relative taxon frequencies is denoted by the vertical axis.