Identification and Testing of Superior Reference Genes for a Starting Pool of Transcript Normalization in Arabidopsis

Genes that are stably expressed during development or in response to environmental changes are essential for accurate normalization in qRT-PCR experiments. To prevent possible misinterpretation caused by the use of unstable housekeeping genes, such as UBQ10, ACT, TUB and EF-1α, as a reference, the use of 20 stably expressed genes identified from microarray analyses was proposed. Furthermore, it was recommended that at least four genes among them be tested to identify suitable reference genes under different experimental conditions. However, testing the 20 potential reference genes under any condition is inefficient. Furthermore, since their stability still varies, there is a need to identify a subset of genes that are more stable than others, which can be used as a starting pool for testing. Here, we validated the expression stability of the potential candidate genes together with the above-mentioned conventional reference genes under six experimental conditions commonly used in plant developmental biology. To increase fidelity, three independent validation experiments were carried out for each experimental condition. A hypothetical normalization factor, which is the geometric mean of genes that were identified as stably expressed genes in each experiment, was used to exclude unstable genes under a given condition. We identified a subset of genes showing higher expression stability under specific experimental conditions. We recommend the use of these genes as a starting pool for the identification of suitable reference genes under given experimental conditions to ensure accurate normalization in qRT-PCR analysis.