Identification and validation of reference genes for quantitative RT-PCR analysis of retinal pigment epithelium cells under hypoxia and/or hyperglycemia

• HPRT1, GUSB and PPIA are the most suitable reference genes for RPE cell gene expression study with hypoxia and/or hyperglycemia.
• The commonly used GAPDH and ACTB were ranked only in the middle place and referred unstable among all the candidate genes.
• Importance of choosing proper reference gene was confirmed by comparing HIF-1α expression normalized to varying reference genes.

Retinal pigment epithelium (RPE) cell-based gene expression studies performed under hypoxia and/or hyperglycemia show huge potential for modeling cell responses in diabetic retinopathy, retinopathy of prematurity and other retinal diseases. However, normalization of gene expression on RPE cells under those conditions has commonly been done using either GAPDH or β-actin as reference genes without any validation of their expression stability. Therefore, we aimed to establish a suitable set of reference genes for studies on RPE cells cultured under both normal culturing glucose and atmospheric oxygen tension (normoxia, 21%), under a low oxygen tension (hypoxia, 1%), under a high glucose growth medium (25 mmol/l) and under the combination of the two changed conditions above for distinct time points taking together from 24 h to 7 days. Quantitative real-time PCR (qRT-PCR) was applied on RNA obtained from a cell line, ARPE-19. Stability of 14 commonly used reference genes was assessed and ranked according to their stability values using the geNorm and NormFinder softwares with the aim to find the most stable expressed gene under all conditions. Our findings confirm that HPRT1, GUSB and PPIA are the most suitable reference genes for RPE cell gene expression experiments subjected to hypoxia and/or hyperglycemia. To emphasize the importance of selecting the most stably expressed reference genes for obtaining reliable results, mRNA expression levels of hypoxia induced factor-1α were analyzed vs the best reference genes, the worst ones and the most commonly used ones. These reference genes gave the most reliable normalization for comparative analyses of gene transcription under those conditions.