Discussion
Resistance inducers have considerable potential in integrated pest management, but remain rarely used due to concerns including phytotoxicity and limited efficacy (Walters & Fountaine 2009). Our work identifies piperonylic acid (PA) as an effective, broad-spectrum and non-phytotoxic novel resistance inducer. PA treatment significantly reduces the susceptibility of several crop plants to a panel of pests and pathogens that includes organisms with different lifestyles (necrotrophs, (hemi)biotrophs and herbivores), degrees of specialization (generalists and specialists), and host tissues (roots and shoots).
PA is a well-known inhibitor of C4H, the second enzyme of the phenylpropanoid pathway (PPP) (Schalk et al. 1998). Prolonged inhibition of this pathway by growing plants on medium containing PA is detrimental to plant growth and development (Naseer et al. 2012; Van de Wouwer et al. 2016), whereas we show that pulsed PA application has no such effect. The absence of growth defects upon pulse treatment with PA is likely explained by rapid inactivation of PA through conjugation, which we have observed here in tomato and which had previously been reported in A. thaliana (Steenackers et al. 2016).
Foliar PA treatment does not affect growth, but does trigger several IR hallmarks. PA-treated plants experience a transient, local ROS burst which begins within one hour and peaks several hours later, a response consistent with our understanding of ROS as an early messenger in plant immunity (Nanda, Andrio, Marino, Pauly & Dunand 2010; Barna, Fodor, Harrach, Pogány & Király 2012; Segal & Wilson 2018). Further evidence for altered ROS metabolism after PA treatment is provided by induction of peroxidase-encoding genes and higher shoot guaiacol peroxidase activity, the latter of which has also been reported after exposure to the IR inducer COS-OGA (van Aubel et al. 2016). In addition, PA treatment also induces the expression of several IR-associated genes, including PR1a (Schuhegger et al. 2006; Martínez-Medinaet al. 2017), NIMIN2-like (Zwicker et al. 2007) andSAD4 (Shan & He 2018).
The accumulation of (often glycosylated) PPP derivatives is a hallmark of IR, which has been observed with diverse IR stimuli in several plant species (Mhlongo, Steenkamp, Piater, Madala & Dubery 2016; Gamiret al. 2020; Ameye et al. 2020; Huang et al. 2021). While it is known that PPP perturbation is a consequence of IR induction, our research suggests that PPP perturbation, in this case by transient inhibition of C4H, can also be a cause of IR. Based on our results, we propose a model of PA-IR outlined in which transient C4H inhibition by PA leads to an initial PPP perturbation, which results in the accumulation of (unidentified) PPP-derived immune-signaling metabolites and, eventually, IR establishment. The PA-IR state includes primed lignification and phenolic compound accumulation, induction of PR genes and accumulation of (likely PPP-derived) nematistatic metabolites.
In this model, PPP-derived metabolites contribute to PA-IR in multiple ways. First, PPP derivatives are involved in PA-IR signaling, a role consistent with the known role of flavonoids and other PPP-derivatives in plant immune regulation (Mandal, Chakraborty & Dey 2010; Saijo & Loo 2020). Second, PPP derivatives contribute to the PA-IR phenotype through their roles as phytoanticipins/phytoalexins (Yadav et al.2020; Desmedt et al. 2020) and building blocks for cell wall reinforcement (Vogt 2010). Finally, their potent antioxidant activity (Rice-Evans, Miller & Paganga 1997; Agati, Azzarello, Pollastri & Tattini 2012) might help the plant cope with the oxidative stress involved in pathogen attack (Barna et al. 2012; Lehmann, Serrano, L’Haridon, Tjamos & Metraux 2015; Segal & Wilson 2018). With regards to the role of PPP derivates as phytoalexins and phytoanticipins, it is noteworthy that we found that PA-IR induces accumulation of metabolites with a nematistatic effect on M. incognita . A similar accumulation of nematistatic compounds upon IR induction has been reported in oat (Avena sativa) (Soriano, Asenstorfer, Schmidt & Riley 2004). In addition to direct induction of defense metabolites upon PA treatment, we also found evidence that PA primes local lignification and phenolic compound accumulation in roots infected by M. incognita . Local lignification upon pathogen exposure has previously been implicated in both induced and genetic resistance against various pathogens (Yadav et al. 2020), including P. syringae (Leeet al. 2019) and M. incognita (Veronico et al.2018).
When considering potential metabolites involved in initial IR establishment upon PPP perturbation, it is tempting to look at the defense hormone SA. SA accumulates in PA-treated shoots, and SA accumulation is a known IR marker in dicots (Pieterse et al.2014). However, we found that PA-IR was largely preserved in tomato plants impaired in SA accumulation. SA-independence has also been shown for IR stimuli including COS-OGA and beta-aminobutyric acid in various pathosystems (Cohen, Vaknin & Mauch-Mani 2016; Singh et al.2019).
Metabolites involved in PA-IR signaling might instead be found amongst (glycosylated) flavonoids and phenylpropanoids, as several such metabolites show a very large increase in abundance after PA treatment. Several genes encoding glycosyltransferases acting on flavonoids and phenylpropanoids were induced by PA treatment, including TOMATO WOUND-INDUCED 1 (Twi1). Twi1 encodes a promiscuous glycosyltransferase acting on various benzoic acids, flavonoids and coumarins (Camposet al. 2019). Twi1 silencing perturbs flavonoid metabolism and increases susceptibility to tomato spotted wilt virus (Camposet al. 2019). Although Twi1 is SA-inducible, the continued ability of elicitors to induce Twi1 in tomato plants expressing the NahG construct suggests the existence of SA-independentTwi1 regulation (O’Donnell et al. 1998). Interestingly,Twi1 is already upregulated 6 hours after PA treatment in shoots and roots and remains upregulated by 72 hours in both tissues, indicating that Twi1 induction is an early, systemic and relatively long-lived response to PA treatment. In support of a regulatory role for phenylpropanoid and flavonoid glycosylation in plant immunity, it was recently shown that the A. thalianaglycosyltransferaseUGT73C7 , which glycosylates phenylpropanoids including para- coumaric acid and ferulic acid, plays a major role in A. thaliana immunity against P. syringae and that its overexpression leads to constitutive PR gene induction and SA accumulation (Huang et al. 2021).
A key feature of IR is that it is a systemic as well as a local phenomenon (Vlot et al. 2020). A systems biology view of PA-IR reveals considerable differences between (systemic) roots and (directly treated) shoots. Whereas the root metabolome is much less affected by PA treatment than the shoot metabolome, the transcriptional response (as measured by the number of differentially expressed genes) to PA treatment is comparable in magnitude. This suggests that PA-IR involves systemic transcriptional changes, while direct accumulation of (defense-related) metabolites is mostly confined to local tissues. By contrast, metabolome analysis of tomato plants foliarly treated with oligogalacturonides found greater metabolic changes in distal root tissues than in directly treated leaves (Gamir et al. 2020).
Our data are compatible with at least two - non-exclusive - explanations for the systemic spread of PA-IR. The first is that metabolites with IR signaling functions, most likely PPP-derived, are transported from treated to distal tissues. In tentative support of this hypothesis, it can be noted that the metabolites which accumulate in systemic root tissues during PA-IR form a small subset of those accumulating in directly treated shoot tissues. A second possibility is that the small quantities of PA found in roots after foliar PA treatment are sufficient for establishing PA-IR in roots. Grafting experiments combined with further metabolomic analysis might elucidate the role of metabolite transport in PA-IR.
Although the precise mechanisms behind PA-IR induction and spread remain to be fully elucidated, our research shows that PA-IR is broad-spectrum and that the PA-IR phenotype appears to be conserved amongst flowering plants. While further multi-season and multi-location field trials as well as ecotoxicological assessments will be required before PA can be used in agriculture, our results – which include data from a field trial in naturally nematode-infested soil - indicate the potential of PA as a sustainable crop protection product.