Selection of reference genes for gene expression studies in ultraviolet B-irradiated human skin fibroblasts using quantitative real-time PCR

Quantitative real-time PCR followed by analysis with the NormFinder and geNorm software programmes was performed. The initial screening of the expression patterns demonstrated that the expression of VIM was suppressed after UVB irradiation at doses ≥25 mJ/cm2 and that the expression of TUBA1A was significantly reduced by UVB doses ≥75 mJ/cm2 in cultured human dermal fibroblasts. The analysis of the experimental data revealed ACTB to be the most stably expressed gene, followed by GAPDH (aglyceraldehyde-3-phosphate dehydrogenase), under these experimental conditions. By contrast, VIM was found to be the least stable gene. The combination of ACTB and TUBB1 was revealed to be the gene pair that introduced the least systematic error into the data normalisation.The data herein provide evidence that ACTB and TUBB1 are suitable reference genes in human skin fibroblasts irradiated by UVB, whereas VIM and TUBA1A are not and should therefore be excluded as reference genes in any gene expression studies involving UVB-irradiated human skin fibroblasts.Ultraviolet B (UVB) radiation is the most active, albeit minor, constituent of solar light. UVB has both direct and indirect adverse biological effects that may result in photo-aging and photo-carcinogenesis. The DNA damage caused by UVB irradiation is considered to be responsible for basal cell carcinoma and squamous cell carcinoma [1,2]. UVB is also suspected of lowering the immune defence system of the skin [3]. Given these effects, the gene expression of dermal fibroblasts after UVB irradiation has become a significant area of study, with a total of 384 manuscripts found in PubMed using the keyword 'UVB' just within the last year.Quantitative real-time PCR (qPCR) is the most powerful method used to quantify gene expression. Similar to other expression study methods, the sample data are usually required to be normalised against either another data set or particular references to correct for any differences in the amount of start