The Selection and Validation of Reference Genes for mRNA and microRNA Expression Studies in Human Liver Slices Using RT-qPCR.
Tomas ZarybnickyPetra MatouškováMartin AmbrožZdeněk ŠubrtLenka SkálováIva BoušováPublished in: Genes (2019)
The selection of a suitable combination of reference genes (RGs) for data normalization is a crucial step for obtaining reliable and reproducible results from transcriptional response analysis using a reverse transcription-quantitative polymerase chain reaction. This is especially so if a three-dimensional multicellular model prepared from liver tissues originating from biologically diverse human individuals is used. The mRNA and miRNA RGs stability were studied in thirty-five human liver tissue samples and twelve precision-cut human liver slices (PCLS) treated for 24 h with dimethyl sulfoxide (controls) and PCLS treated with β-naphthoflavone (10 µM) or rifampicin (10 µM) as cytochrome P450 (CYP) inducers. Validation of RGs was performed by an expression analysis of CYP3A4 and CYP1A2 on rifampicin and β-naphthoflavone induction, respectively. Regarding mRNA, the best combination of RGs for the controls was YWHAZ and B2M, while YWHAZ and ACTB were selected for the liver samples and treated PCLS. Stability of all candidate miRNA RGs was comparable or better than that of generally used short non-coding RNA U6. The best combination for the control PCLS was miR-16-5p and miR-152-3p, in contrast to the miR-16-5b and miR-23b-3p selected for the treated PCLS. Our results showed that the candidate RGs were rather stable, especially for miRNA in human PCLS.
Keyphrases
- endothelial cells
- binding protein
- poor prognosis
- gene expression
- genome wide
- induced pluripotent stem cells
- transcription factor
- magnetic resonance
- newly diagnosed
- dna methylation
- magnetic resonance imaging
- oxidative stress
- pulmonary tuberculosis
- long non coding rna
- big data
- bioinformatics analysis
- deep learning
- artificial intelligence
- contrast enhanced
- nucleic acid