Dynamic decomposition of transcriptome responses during plant effector-triggered immunity revealed conserved responses in two distinct cell populations

Summary Rapid plant immune responses in the appropriate cells are needed for effective defense against pathogens. Although transcriptome analysis is often used to describe overall immune responses, collecting transcriptome data with sufficient resolution in both space and time is challenging. We reanalyzed public Arabidopsis time-course transcriptome data obtained after a low-dose inoculation of a Pseudomonas syringae strain expressing the effector AvrRpt2, which induces Effector-Triggered Immunity (ETI) in Arabidopsis. Double-peak time-course patterns were prevalent among thousands of upregulated genes. We implemented a multi-compartment modeling approach to decompose the double-peak pattern into two single-peak patterns for each gene. The decomposed peaks revealed an “echoing” pattern: the peak times of the first and second peaks correlated well across most upregulated genes. We demonstrated that two peaks likely represent responses of two distinct cell populations, which respond either cell-autonomously or indirectly to AvrRpt2. Thus, the peak decomposition extracted spatial information from the time-course data. The echoing pattern also indicated a conserved transcriptome response between two cell populations despite different elicitor types. WRKY transcription factors appeared to underlie the conserved transcriptome response. Activation of a WRKY network via different entry-point WRKYs could explain the conserved transcriptome response elicited by different elicitor types.