Estimation of the Circadian Phase by Oscillatory Analysis of the Transcriptome in Plants

Plants have a circadian clock, a biological oscillator with an approximately 24 h period, which is known to dominate various physiologically active rhythms (Harmer et al., 2000). Photoperiodicity induction of flowering, stomatal opening, and concentration of nutrients (e.g. ascorbic acid) are also regulated by the circadian clock (Kotchoni et al., 2008). Recent research has clarified that the internal body time of the circadian clock (the phase of the circadian clock) is also related to pest resistance and metabolism of reactive oxygen species (Lai et al., 2012; Goodspeed et al., 2013). Therefore, knowledge of the phase of circadian clock is thought to be important in control of growth and crop quality. The circadian clock works via clock genes, whose expression varies periodically. For the model plant Arabidopsis thaliana, a set of genes with periodic variation in expression level, called clock genes or clock-related genes, such as LHY (LATE ELONGATED HYPOCOTYL), CCA1 (CIRCADIAN CLOCK ASSOCIATED 1), and TOC1 (TIMING OF CAB EXPRESSION 1) and PRRs (PSEUDORESPONSE REGULATORs) are known factors related to the central oscillator of the clock (Nakamichi et al., 2012; Nagel et al., 2015; Kamioka et al., 2016; Liu et al., 2016; Ezer et al., 2017). Such genes generate a self-sustained circadian rhythm by negative feedback loops (Nohales and Kay, 2016). On the other hand, based on global gene expression analysis through microarray analysis, about 20% of plant genes are under the control of the circadian clock (Li et al., 2017). Recently, for crop species such as rice, tomato and lettuce, the circadian rhythm of the transcriptome in the field and in the plant factory has been clarified (Nagano et al., 2012; Matsuzaki et al., 2015; Tanigaki et al., 2015; Higashi et al., 2016a; 2016b). It is also possible to estimate the phase of the circadian clock by analyzing the periodicity of the transcriptome (Matsushika et al., 2000; Ueda et al., 2004). A method for estimating the phase of the circadian clock, called the molecular timetable method, has been constructed for mammals (Ueda et al., 2004). It is possible to estimate body time from transcriptome data with high accuracy within about 2 h. In the standard application of this method, the period of the circadian rhythm for each expressed gene is regarded as a constant value, 24 h. In addition, the waveform of the circadian rhythm is assumed to be a simple form, that is, a cosine curve. However, in plants, the period and the waveform of the circadian rhythm varies depending on species, environment of cultivation, and other factors (Ninomiya, 1984; Higashi et al., 2014). In this study, the model plant A. thaliana, and lettuce