The tea plant, Camellia sinensis (L.) O. Kuntze, is an economically important, perennial woody plant rich in catechins. Although catechins have been reported to play an important role in plant defences against microbes, their roles in the defence of tea plants against herbivores remain unknown. In this study, we allowed the larvae of Ectropis grisescens, a leaf‐feeding pest, to feed on the plants, and alternatively, we wounded the plants and then treated them with E. grisescens oral secretions (WOS). Both approaches triggered jasmonic acid‐, ethylene‐ and auxin‐mediated signalling pathways; as a result, plants accumulated three catechin compounds: (+)‐catechin, epicatechin and epigallocatechin. Not only was the mass of E. grisescens larvae fed on plants previously infested with E. grisescens or treated with WOS significantly lower than that of larvae fed on controls, but also artificial diet supplemented with epicatechin, (+)‐catechin or epigallocatechin gallate reduced larval growth rates. In addition, the exogenous application of jasmonic acid, ethylene or auxin induced the biosynthesis of the three catechins, which, in turn, enhanced the resistance of tea plants to E. grisescens, leading to the coordination of the three signalling pathways. Our results suggest that the three catechins play an important role in the defences of tea plants against E. grisescens.

The selection of reliable reference genes (RGs) for normalization under given experimental conditions is necessary to develop an accurate qRT-PCR assay. To the best of our knowledge, only a small number of RGs have been rigorously identified and used in tea plants (Camellia sinensis (L.) O. Kuntze) under abiotic stresses, but no critical RG identification has been performed for tea plants under any biotic stresses till now. In the present study, we measured the mRNA transcriptional levels of ten candidate RGs under five experimental conditions; these genes have been identified as stable RGs in tea plants. By using the ΔCt method, geNorm, NormFinder and BestKeeper, CLATHRIN1 and UBC1, TUA1 and SAND1, or SAND1 and UBC1 were identified as the best combination for normalizing diurnal gene expression in leaves, stems and roots individually; CLATHRIN1 and GAPDH1 were identified as the best combination for jasmonic acid treatment; ACTIN1 and UBC1 were identified as the best combination for Toxoptera aurantii-infested leaves; UBC1 and GAPDH1 were identified as the best combination for Empoasca onukii-infested leaves; and SAND1 and TBP1 were identified as the best combination for Ectropis obliqua regurgitant-treated leaves. Furthermore, our results suggest that if the processing time of the treatment was long, the best RGs for normalization should be recommended according to the stability of the proposed RGs in different time intervals when intragroup differences were compared, which would strongly increase the accuracy and sensitivity of target gene expression in tea plants under biotic stresses. However, when the differences of intergroup were compared, the RGs for normalization should keep consistent across different time points. The results of this study provide a technical guidance for further study of the molecular mechanisms of tea plants under different biotic stresses.With the increasing popularity of gene expression analysis in biological research, quantitative real-time polymerase chain reaction (qRT-PCR) has become a critical and powerful tool for rapid and reliable quantification of mRNA transcriptional expression levels of target genes due to its high-throughput screening, sensitivity, simplicity, specificity and accuracy 1,2 . Relative quantification of target gene expression under certain stresses has been widely studied since the beginning of this century 3 . An accurate assay of gene expression through qRT-PCR relies on every step of sample preparation and processing, e.g., the integrity of purified RNA, the efficiency of reverse transcription, and the overall transcriptional activity of the tissues or cells analysed 4 ; each step needs to be accurately normalized by stably expressed reference genes (RGs) 5,6 . Therefore, the selection of reliable RGs for normalization under given experimental conditions is a requirement for developing an accurate qPCR assay.Housekeeping genes, such as the glyceraldehyde 3-phosphate (GAPDH), the actin gene (ACTIN), translation elongation factor EF-1 alpha (EF-1α), 18 s r...