Evaluation of reference genes for quantitative real-time PCR to investigate seasonal and labor-specific expression profiles of the honey bee abdomen

Owing to the highly developed sociality, division of labor, and passive population management of their colonies, honey bees, Apis mellifera L., have been widely used as model insects to investigate the physiological roles of the genes putatively involved in social development in insects. Quantitative real-time polymerase chain reaction (qRT-PCR) has been broadly applied to investigate the expression patterns of such target genes. However, the selection of suitable reference genes is an essential step to ensure the accurate quantification of target gene transcription levels with this method. Therefore, in this study, we selected seven potential reference genes (rp49, rpL32, rpS18, tbp, tub, gapdh, and ace2) and determined their PCR efficiencies. The seasonal expression efficiencies of five genes (rpL32, rpS18, tbp, gapdh, and ace2) that displayed 90% to 110% amplification efficiencies were then evaluated using three software packages (geNorm, NormFinder, and BestKeeper), and were also applied in the normalization of seasonal expression patterns of a target gene (vg) in the abdomens of forager and nurse honey bee workers. The three software packages resulted in slightly different gene stability ranks, but overall the combination of two genes (rpS18 and gapdh) was determined to be the most optimal for use in the normalization of target gene expression in forager. However, a single gene (either rpL32 or rpS18) should be applied in the nurse. In the comparison between foragers and nurses, either rpL32, rpS18, or gapdh, is suggested to be used as the reference gene for qRT-PCR-based determination of seasonal and labor-specific gene expression profiles.