Figure 6: Length of the transcriptional memory in trained plants.

The relative levels of mRNA of trained plants after a recovery period under watered conditions for 3 or more days, before and after a subsequent dehydration stress are shown in a–d. Plants were not stressed (Control) or were dehydration stressed 4 times (Trained), with 22-h recovery periods before receiving water in soil for 3, 5 or 7 days (first, second and third columns, respectively). After the indicated days of watering, plants received water (W) or were dehydration stressed (D) for 2 h before the indicated analyses. (a,b) Transcript levels of the trainable (RD29B and RAB18) genes. (c,d) Transcript levels of the non-trainable (RD29A and COR15A) genes measured by reverse transcription–quantitative PCR (qPCR). UBQ10 was used as an internal control. (e) H3K4me3 levels at the RD29B gene measured by ChIP–qPCR, relative to ACT7, shown in f at the regions showing the peak accumulation of the respective marks at each gene (see Fig. 4 for location of the peak regions). (g) Ser5P levels at the RD29B gene measured by ChIP–qPCR, relative to ACT7 shown in h, at the regions showing the peak accumulation of the respective marks at each gene (see Fig. 4 for location of the peak regions). Results from measurements during days 3, 5 or 7 after recovery in soil are shown in the first, second and third columns, respectively. (i) ABA levels before stress (W), after one to four stresses (S1–S4) or one to three recoveries (R1–R3) and before (W) or after a dehydration stress (D) that was initiated after 5 or 7 days of watering of control or trained plants. Experiments in a–c were repeated at least three times, each with three replicates, and the representative experiments shown indicates the mean+s.e.m., n=3 replicates. Experiments in d were performed twice, with two replicates, and the representative experiment shown indicates the mean+range, n=2 replicates.