Selection of reference genes for real-time quantitative PCR studies of kumquat in various tissues and under abiotic stress

• Sixteen candidate reference genes of kumquat was evaluated for normalization in real-time quantitative PCR studies, including two groups of homologous genes and six other genes also commonly used as reference genes in previous studies.
• Popular statistical methods in this field, geNorm, Normfinder, and Bestkeeper, has been used flexibly aimed to get reliable results in this study.
• Each ACT gene or UBQ gene performed differently in this study and homologous genes should be taken into account in other similar studies.
• We confirmed the best reference gene and the best reference gene combinations for normalization under different experimental conditions.

The proper selection of stable reference genes for real-time quantitative PCR (RT-qPCR) can help to reliably assess gene expression results under specific experimental conditions. The present study evaluated the stability of 16 candidate reference genes of kumquat using the methods of geNorm, Normfinder, and Bestkeeper in different tissues of kumquat (Fortunella crassifolia Swingle) under some type of abiotic stress. The results showed that homologous genes should be taken into account for normalization in RT-qPCR studies. For the three specific types of stresses, i.e., salt stress, drought stress, and heavy metal stress, ACT6, ACT8, and ACT7 should be selected as internal controls, respectively. For low-temperature stress, the use of multiple reference genes for normalization may be optimal, such as ACT7+ACT9 (according to geNorm) or ACT7+TUB4 (according to Normfinder). We suggest that ACT7 alone or the combination of ACT7+ACT1+EF1-α or ACT1+ACT8 can meet the requirements for the normalization of RT-qPCR studies under the experimental conditions tested here. Although some commonly used reference genes (TUA1, TUB1, TUB4, CYP, GAPDH, etc.) showed much higher variability, some are still useful in combinations under specific conditions.