The CpG-dependent plant immune response to self-DNA triggers defence hormone signalling and improves fitness

Abstract

The accumulation of DNA in the cytoplasm or extracellular space is a signal of danger. Plants respond to this signal with a self/non-self-specific activation of early immune signalling events. Here, we asked whether this specificity translates to fitness-relevant resistance to natural enemies. We treated common bean (Phaseolus vulgaris) plants with self-DNA and with non-self-DNA from other plant species. Self-DNA treatment induced jasmonic acid and decreased feeding by a chewing herbivore (Spodoptera frugiperda), while self- and non-self-DNA induced salicylic acid and reduced the population densities of two fungal pathogens (Botrytis cinerea and Sclerotinia sclerotiorum) and of four bacterial pathogens (Enterobacter sp. strain FCB1, Pseudomonas syringae pv. phaseoli and pv. syringae, and Xanthomonas axonopodis pv. phaseoli). Strikingly, a single self-DNA-treatment increased seed production under field conditions in two seasons ca. 1.5-fold and 3.2-fold, while non-self-DNA had lower or no detectable effects. Stronger responses to self- than non-self-DNA seemingly contradict immunological theory. In mammalian immune cells, toll-like receptor 9 (TLR9) is activated by microbial DNA but also by mitochondrial (mt)DNA that is rich in unmethylated CpG motifs. Using bean suspension cells, we observed stronger immunogenic effects of genomic DNA than chloroplast DNA or mtDNA. Moreover, in vitro methylation or cleavage of CpG motifs reduced – without eliminating – the H2O2-inducing properties of DNA. We conclude that the plant immune response to DNA comprises a self/non-self-specific induction of major defence hormones that can be adaptive under natural enemy pressure and that unmethylated CpG motifs contribute to the immunogenic effects of DNA in plants.

ATM and ATR, two central players of the DNA damage response, are involved in the induction of systemic acquired resistance by extracellular DNA, but not the plant wound response

“…The levels of JA peaked at 30 min and those of SA at 24 h ( Figure 2 ; Supplementary Figure S3 ). Subsequent treatments revealed self/nonself-specific effects of DNA on H 2 O 2 and both hormones and in all cases, strongest effects were triggered by self-DNA ( Figure 2 ), similar to earlier observations, although with minor differences between species ( 20 – 22 , 26 , 35 , 36 ). The response times and levels reached by the three signals fall clearly within the standard kinetics of the three signals ( 119 – 121 ).…”

“…Moreover, the authors used the same methyltransferase in combination with restriction enzymes to corroborate the methylation of natural herring DNA and thereby could attribute the lower ROS-inducing activity of herring DNA to a TLR9-like mechanism that requires unmethylated CpG motifs ( 9 ). Similarly, a recent study using the same three enzymes to generate DNA free of unmethylated CpG motifs from naturally methylated DNA of common bean and observed that any enzymatic manipulation decreased the H 2 O 2 -inducing properties as compared to the natural self-DNA, with no detectable differences among fragments with completely methylated CpG motifs and fragments in which all unmethylated CpG motifs had been cleaved enzymatically ( 36 ). These results would be consistent with a mechanism in which DNA triggers a partial induction of H 2 O 2 that is independent of CpG-methylation whereas the full response requires additional activation by unmethylated CpG motifs ( 36 ).…”

“…Treatments with plant DNA, genomic or plasmid bacterial DNA, or synthetic single stranded oligodeoxynucleotides (ssODNs), were shown to trigger early immune responses that are highly conserved between plants and mammals, including Ca 2+ fluxes, membrane depolarization, the formation of reactive oxygen species (ROS) such as hydrogen peroxide (H 2 O 2 ), or the activation of mitogen-activated protein kinases (MAPKs) ( 9 , 20 – 22 , 25 , 27 , 34 , 35 ). The responses include massive transcriptomic reprogramming ( 24 , 25 , 35 ), and studies at the phenotypic level reported not only increased immunity (‘resistance’) to microbial pathogens, but also increased ‘defence’ to plant-specific pests such as leaf and sap-sucking herbivores ( 21 , 23 , 26 , 27 , 29 , 33 – 36 ). Strikingly, stronger responses to self- in comparison to nonself-DNA were reported in most, although not all, of the studies that compared DNA from different sources.…”

“…Even the ‘Mazzoleni-effect’ – a dosage-dependent inhibition of growth by self-DNA, but not nonself-DNA that has been described originally by the group of Stefano Mazzoleni – could subsequently be confirmed in different models, including a tree ( Alnus glutinosa ), freshwater and marine algae and the nematode, Caenorhabditis elegans ( 18 , 21 , 22 , 30 , 31 , 35 , 37 , 38 ). Differential responses to self- versus nonself-DNA are significant even when nonself-DNA from closely related genotypes (species of the same genus or ecotypes of the same species) is used ( 21 , 26 , 31 , 35 , 36 , 38 ).…”

“…However, on the one hand, an induction by DNA has been reported for traits that are typically controlled via WR as well as SAR. Further, more recent support for the PTI model comes from a study that linked resistance-induction by self-DNA in peach ( Prunus persica ) fruits to the brassinosteroid insensitive 1-associated receptor kinase 1 (BAK1) – a cytoplasmatic co-receptor that interacts with various PRRs of DAMPs and PAMPs ( 34 ) - and from two studies that reported self-DNA-induced JA-synthesis in A. thaliana and common bean ( Phaseolus vulgaris ) ( 35 , 36 ). On the other hand, self-/nonself specific effects are mainly reported for early, not enemy-specific immune signals which have also been reported for the DDR, and as far as we know, no PRRs that could act as dsDNA receptors have been reported for plants or could be predicted by in silico approaches ( 4 , 63 – 65 ).…”