Variation in DNA methylation and response to short-term herbivory in Thlaspi arvense

Plant metabolic pathways and gene networks involved in the response to herbivory are well-established, but the impact of epigenetic factors as modulators of those responses is less understood. Here, we studied the role of DNA cytosine methylation on phenotypic responses after short-term herbivory in Thlaspi arvense plants with two contrasting flowering phenotypes. We investigated the effect of experimental demethylation and herbivory treatments following a 2x3 factorial design. First, half the seeds were submerged in a water solution of the demethylating agent 5-azacytidine and the other half only in water, as controls. Then, we assigned control and demethylated plants to three herbivory categories (i) insect herbivory, (ii) artificial herbivory, and (iii) undamaged plants. The effects of the demethylation and herbivory treatments were assessed by quantifying genome-wide global DNA cytosine methylation, concentration of leaf glucosinolates, final stem biomass, fruit and seed production, and seed size. For most of the plant traits analysed, individuals from the two plant-types responded differently. In late-flowering plants, global DNA methylation did not differ between control and demethylated plants but it was significantly reduced by herbivory. Conversely, in early-flowering plants, demethylation at seed stage was still evident in leaf genomes of reproductive individuals whereas herbivory did not affect their global DNA methylation. In late-flowering plants, artificial herbivory imposed a stronger reduction than insect herbivory in global DNA methylation and final stem biomass, and induced higher concentration of aliphatic glucosinolates. In early-flowering plants, the effects of herbivory were non-significant for the same traits. Finally, the effect of herbivory on reproductive parameters varied with the level of demethylation and the plant flowering-type. The complexity of our results suggests that the genetic background of experimental plants can affect the responses to treatments and it points towards multifaceted genetic-epigenetic interactions in determining herbivory-induced phenotypic plasticity.