Metabolomics combined with pattern recognition and bioinformatics analysis methods for the development of pharmacodynamic biomarkers on liver fibrosis.

The major obstacle for the development of targeted therapies is the lack of pharmacodynamic (PD) biomarkers to provide an early readout of biological activities. As the modulation of metabolites may reflect the biological changes occurring in the targets, metabolomics is promising to be an efficient way to explore PD biomarkers. In the present study, a liver fibrosis rat model was established by intraperitoneal injection of CCl4 twice weekly for 6 weeks, the treatment of total aglycone extracts of Scutellaria baicalensis (TAES) was begun 4 weeks after the modeling, and gas chromatography-mass spectrometry (GC-MS) based metabolomics combined with pattern recognition and network analysis were carried out for the research on PD biomarkers of TAES on liver fibrosis. After 2 weeks of treatment, TAES shows positive effects on CCl4-induced liver fibrosis. In the metabolomics study, 63 urinary metabolites contributing to liver fibrosis were identified. Six metabolic pathways significantly enriched in metabolomics data were mapped onto a network to determine global patterns of metabolic alterations in liver fibrosis. By topological analysis, 6 metabolites with high centrality in the metabolic sub-network were selected as potential PD biomarkers. Within 24 h of the final administration, the 6 identified urine metabolic biomarkers with response to time variation of TAES were validated as PD biomarkers. This integrative study presents an attractive strategy to explore PD biomarkers, which may give insight into the actual pharmacological effect of target drugs, and the information from PD biomarkers can be combined with pharmacokinetics to select the optimal dose and a schedule of administration for the drugs.