ABSTRACT: Real-time quantitative PCR (RT–qPCR) is the favoured method for gene expression analysis in molecular biology due to its sensitivity, specificity, cost-effectiveness, and reproducibility. To obtain the accuracy and reliability of RT-qPCR, the use of reliable reference genes is inevitable. There were many reports about the physiological response of giant reed (Arundo donax L.) to abiotic stresses. However, there is little use in the validation of reference genes under different treatments. It still belongs to the blank that the research about selecting reference genes under salt and alkali. In this study, the expression stability of twenty-three candidate reference genes in leaves and roots were assessed under salt, drought, and alkali stresses using geNorm, NormFinder, BestKeeper, and Delta Ct algorithms. Our results showed that no one gene had an invariant expression under different conditions. For example, under drought stress, UPL3, UBC2, and APT1 were better reference genes in leaves, RPL5 and FPS2 were better in roots. Under alkali stress, GAPDH, APT1, and RPS5 were better reference genes in leaves; UPL3, ACT2, and SAMDC2 were better in roots. In addition, the expression of MSD1 was used to further confirm the validated reference genes under salt, drought, and alkali stresses. It was proved that the use of inappropriate reference genes in giant reed significantly altered the relative expression of target genes and even reversed the results. Consequently, our results provided guidelines for reference gene selection under salt, drought, and alkali stresses and a foundation for more accurate and widespread use of RT-qPCR in the giant reed.
