Stable and dynamic gene expression patterns over diurnal and developmental timescales in Arabidopsis thaliana

Abstract

Introduction

Due to their sessile nature within a cyclical environment, plants have evolved an internal timekeeper called the circadian clock. It consists of a core interlocking loop of transcriptional regulators, whose components are mostly conserved across embryophytes (Petersen et al., 2022; Wang et al., 2024). Circadian clocks confer fitness to plants by allowing them to coordinate responses to photoperiod, light quality, temperature, and other environmental cues (Dodd et al., 2005; Atamian et al., 2016; Rubin et al., 2017; Xu et al., 2022). This coordination takes place through widespread transcriptional control of the plant transcriptome, among other methods of regulation (Nagel et al., 2015; Ezer et al., 2017; Hayama et al., 2017; Romanowski et al., 2020; Xiong et al., 2022). Estimates of the proportion of the Arabidopsis and wheat transcriptomes that are controlled by the clock range from 30% to 50% (Covington et al., 2008; Romanowski et al., 2020; Rees et al., 2022).

Large-scale circadian RNA sequencing (RNA-seq) experiments are typically performed in seedlings or juvenile plants, due to the ease of performing timeseries experiments under different entrainment and free-running conditions. Yet, many of the clock genes have fundamental roles in the timing of developmental processes (Inoue et al., 2018; Wang et al., 2024). A key example of this is the evening complex (EC), a tripartite protein complex containing EARLY FLOWERING 3 and 4 (ELF3 and ELF4) and LUX ARRHYTHMO (LUX) (Nusinow et al., 2011; Herrero et al., 2012). The EC, in combination with GIGANTEA (GI), acts upstream of the photoperiodic flowering pathway in Arabidopsis and rice (Fowler et al., 1999; Park et al., 1999; Sawa & Kay, 2011; Andrade et al., 2022). Leaf senescence occurs concurrently with the vegetative-to-reproductive transition (Redmond et al., 2024) and is also regulated by the circadian clock. Mutations in clock genes, including ELF3, PSEUDO-RESPONSE REGULATOR9 (PRR9), and CIRCADIAN-CLOCK ASSOCIATED1 (CCA1) have all been shown to affect the onset of leaf senescence (Sakuraba et al., 2014; Kim et al., 2018; Song et al., 2018). This emphasises the need to study circadian rhythms in adult plants when the relevant developmental transitions are occurring.

Moreover, developmentally associated processes exhibit diel or diurnal patterns, meaning that they have an oscillating expression over a 24-h period. These patterns begin at the earliest stages of plant development and extend into all developmental stages. Germination responds to diurnally fluctuating temperatures in many plant species (Thompson et al., 1977). Hypocotyl development is mediated through auxin- and temperature-related processes via the clock-controlled PHYTOCHROME-INTERACTING FACTORS 4 and 5 (PIF4 and PIF5) (Nozue et al., 2007; Seaton et al., 2015). Many of the central integrators that control flowering time are diurnally expressed, such as FLOWERING LOCUS T (FT), SUPPRESSOR OF OVEREXPRESSION OF CO1 (SOC1), and LEAFY (Wendell et al., 2017). Additionally, many of the transcription factors that regulate the synthesis of key plant hormones involved in development, like auxin and abscisic acid (ABA), are diurnally expressed (Balcerowicz et al., 2021).

Tissue specificity also plays a role in the link between the clock and development. For instance, the vascular clock plays a dominant role over the epidermal clock in leaves. Moreover, these tissue-specific clocks influence two distinct developmental processes, flowering and hypocotyl development (Endo et al., 2014). Vong et al. (2024) suggested that each cell type's transcriptional activity varied across diel and developmental timescales, but it is unclear how the daily oscillations in cell-type activity vary over developmental timescales.

There remains a large gap in our knowledge of how diel genes vary over development and how developmental genes vary across the day. Here, we address this gap by measuring diel gene expression over one of the most crucial developmental transitions an annual plant experiences: the vegetative-to-reproductive transition. We measure gene expression over two timescales: every 24 h (the diel timescale) and over c. 2 wk (the developmental timescale). As individual plants experience developmental asynchrony in their floral transition (Klingenberg, 2019; Redmond et al., 2024), we use a photoperiod shift from short days (SD; 8 h light d⁻¹) to long days (LD; 16 h light d⁻¹) to induce synchronised vegetative-to-reproductive transitions. One day of LD conditions is sufficient to induce the maximum gene expression response of FT (Corbesier et al., 1996; Krzymuski et al., 2015). We therefore measure diel gene expression in ageing plants after the inductive SD to LD signal.

Here, we determine the extent of transcriptional changes over both developmental and diel timescales. First, we find that gene expression changes most dramatically over the developmental timescale and that core clock genes have broadly stable phases and amplitudes of expression per day. Second, we observe that the expression dynamics of targets of a key Arabidopsis energy sensor vary across both scales. Third, we show that transcriptional targets of core clock genes exhibit differing changes in amplitude over development. Fourth, we identify tissue-specific changes in rhythmic processes in ageing leaves. Finally, one of the most important applications of our work is in identifying sets of genes that are stable across both timescales, as these could serve as important controls for reverse transcription quantitative polymerase chain reaction. We suggest a filtered set of housekeepers that we would encourage the community to use for studies that aim to study gene expression of clock-controlled processes in ageing plants.