Transcriptional analysis on biomarkers of liver injury induced by anti-tuberculosis drugs

Objective The study utilized human transcriptome microarray to explore biomarkers for diagnosing drug-induced liver injury （DILI） caused by anti-tuberculosis drugs.

Methods A 6-month follow-up study was conducted on 152 patients treated with anti-tuberculosis drugs in designated hospitals in Shanghai. The blood samples were collected at the 0， 2， 4， 8， 12 and 24 weeks after treatment. According to the clinical biochemical indicators， the research subjects were divided into DILI cases （34 cases） and Control cases （118 cases）. Single factor analysis was conducted on the influencing factors between the two groups. In a 1∶1 matched DILI-control study， RNA samples of 13 pairs of cases were sequenced by the whole transcript expression mRNA array. Differentially expressed genes （DEGs） were screened by Hotelling's T² value sequencing and the expression trend analysis of genes by STEM （short-time series expression miner）， and the functional enrichment and pathway analysis of DEGs were carried out.

Results In total 152 clinical cases， weight of patients was a risk factor for the occurrence of hepatotoxicity caused by anti-tuberculous drugs. Based on the analysis results of mRNA array， 513 DEGs were screened by Hotelling's T² value sequencing method， which were enriched in 32 annotations of GO （Gene Ontology） analysis and 10 pathways of KEGG （Kyoto encyclopedia of genes and genomes） analysis. One differential expression pattern was screened by STEM， which was enriched in 2 biological process notes of GO. Among them， the key genes AIM2， CD86， CXCL10 and non-coding RNAs SCARNA10， SNHG10 and SNORD105 are potential biomarkers of DILI caused by anti-tuberculosis drugs.

Conclusion In this research for biomarkers conducted on cases with liver injury caused by anti-tuberculosis drugs， biological pathways associated with hepatotoxicity are identified and a series of key genes related with drug-induced liver injury are found， which provides the basis for mechanism study and searching for earlier and more sensitive biomarkers.