Selection of suitable reference genes for gene expression studies in Staphylococcus capitis during growth under erythromycin stress

• Published in: Molecular genetics and genomics : MGG 2016-08, Vol.291 (4), p.1795-1811
• Main Authors: Cui, Bintao, Smooker, Peter M., Rouch, Duncan A., Deighton, Margaret A.
• Format: Article
• Language: English
• Subjects: Algorithms; Animal Genetics and Genomics; Anti-Bacterial Agents - pharmacology; Biochemistry; Biofilms - drug effects; Biomedical and Life Sciences; Erythromycin - pharmacology; Gene Expression; Genes, Bacterial; Genes, Reporter; Genomics; Human Genetics; Life Sciences; Methods Paper; Microbial Genetics and Genomics; Microorganisms; Pathogens; Plant Genetics and Genomics; Software; Staphylococcus - drug effects; Staphylococcus - genetics; Staphylococcus capitis
• ISSN: 1617-4615; 1617-4623
• DOI: 10.1007/s00438-016-1197-9

Accurate and reproducible measurement of gene transcription requires appropriate reference genes, which are stably expressed under different experimental conditions to provide normalization. Staphylococcus capitis is a human pathogen that produces biofilm under stress, such as imposed by antimicrobial agents. In this study, a set of five commonly used staphylococcal reference genes ( gyrB , sodA , recA , tuf and rpoB ) were systematically evaluated in two clinical isolates of Staphylococcus capitis ( S. capitis subspecies urealyticus and capitis , respectively) under erythromycin stress in mid-log and stationary phases. Two public software programs (geNorm and NormFinder) and two manual calculation methods, reference residue normalization (RRN) and relative quantitative (RQ), were applied. The potential reference genes selected by the four algorithms were further validated by comparing the expression of a well-studied biofilm gene ( icaA) with phenotypic biofilm formation in S. capitis under four different experimental conditions. The four methods differed considerably in their ability to predict the most suitable reference gene or gene combination for comparing icaA expression under different conditions. Under the conditions used here, the RQ method provided better selection of reference genes than the other three algorithms; however, this finding needs to be confirmed with a larger number of isolates. This study reinforces the need to assess the stability of reference genes for analysis of target gene expression under different conditions and the use of more than one algorithm in such studies. Although this work was conducted using a specific human pathogen, it emphasizes the importance of selecting suitable reference genes for accurate normalization of gene expression more generally.