Identification of valid reference genes for circadian gene-expression studies in human mammary epithelial cells

The circadian clock controls most of the physiological processes in the body throughout days and nights' alternation. Its dysregulation has a negative impact on many aspects of human health, such as obesity, lipid disorders, diabetes, skin regeneration, hematopoiesis and cancer. To date, poor is known on the molecular mechanisms that links mammary gland homeostasis to the circadian clock but recent reports highlight the importance of loss of circadian genes for mammary gland development and during tumour progression in breast cancer. Gene expression studies are then required to clarify how the circadian clock can modulates the human mammary gland development during ontology and its behaviour in physiological and oncogenic context. For this, in addition to genome-wide studies, real-time quantitative RT-PCR (qPCR) is a powerful and pertinent technique to quantify the expression of a reduced set of genes of interest in many different samples. Relative quantification of qPCR data requires the use of reference genes for normalisation. For circadian studies, reference genes expression must not oscillate in mirror of the circadian clock and must not be affected by the synchronisation protocols required in vitro to reset the circadian clock. Inappropriate selection of reference genes can consequently affect the amplitude of gene expression oscillation and bias data interpretation. Currently, no standard reference genes have been validated regarding these criteria for human mammary epithelial cells and the purpose of this study was to fill this gap. For this, we used the RefFinder tool, which combines four different algorithms, on 9 candidate reference genes. We compared reference genes stability using three different synchronisation protocols applied on four different mammary epithelial cell lines. This allowed us to define a set of reference genes in human mammary epithelial cells whose expression remains stable despite synchronisation protocols. We observed that the synchronisation of cells by serum shock was the most suitable procedure for maintaining the amplitude of oscillation of clock genes over time and we identified RPL4, RPLP0, HSPCB and TBP as an optimal combination of reference genes for the normalisation of the oscillatory expression of clock genes in human mammary epithelial cells.