Introduction
Stem tunneling by maize stem borers is an important constraint to achieve the potential yield of maize varieties across the world (Malvar, Butron, Ordas & Santiago, 2008). The maize plant protect themselves from borer attack using constitutive and induced defenses but much more attention has focused on studying the first ones although evidence indicates that induced defenses have lower resource allocation costs than constitutive defenses (Beeghly, Coors & Lee, 1997, Cardinal & Lee, 2005, Howe & Jander, 2008, Karban, Agrawal & Mangel, 1997, Klenke, Russell & Guthrie, 1986, Santiago, Butron, Arnason, Reid, Souto & Malvar, 2006, Santiago, Butron, Revilla & Malvar, 2011, Santiago, Souto, Monetti, Ordas, Ordas & Malvar, 2006, Santiago, Souto, Sotelo, Butron & Malvar, 2003). Indeed, maize stem feeding by borers significantly modify antibiosis against stem borer larvae, although antibiosis changes depend on the genotype and the duration of feeding (Cao, ButrĂ³n, Malvar, Figueroa-Garrido & Santiago, 2019, Dafoe, Thomas, Shirk, Legaspi, Vaughan, Huffaker, Teal & Schmelz, 2013). Dafoe et al. (Dafoe et al. , 2013) reported that growth of stem borer larvae was significantly higher when fed with stem tissues preconditioned by 48 h of larval tunneling compared to untreated stem tissues; while Cao et al. (Cao et al. , 2019) stated that the effect of long-term attack by borers on the antibiotic properties of corn stems is genotype-dependent.
Previous studies have already shown that feeding-induced changes in plant metabolites influence the behavior and performance of conspecifics and that influence depends on the time-lag of induction (Poelman, Broekgaarden, Van Loon & Dicke, 2008, Santiago, Cao, Butron, Lopez-Malvar, Rodriguez, Sandoya & Malvar, 2017, Su, Chen, Mescher, Peng, Xie, Wang, Wu, Liu, Li, Wang & Zhang, 2018, Wang, Bezemer, van der Putten & Biere, 2015). According to that, several authors have pointed out to different defense mechanisms involved in the response to long-term feeding by insects compared to those implicated in the early response (Donze-Reiner, Palmer, Scully, Prochaska, Koch, Heng-Moss, Bradshaw, Twigg, Amundsen, Sattler & Sarath, 2017, Gutsche, Heng-Moss, Sarath, Twigg, Xia, Lu & Mornhinweg, 2009, Uawisetwathana, Graham, Kamolsukyunyong, Sukhaket, Klanchui, Toojinda, Vanavichit, Karoonuthaisiri & Elliott, 2015); more specifically, previous studies also suggested important differences between maize stem responses to short and long-term feeding by stem borers (Dafoe, Huffaker, Vaughan, Duehl, Teal & Schmelz, 2011, Rodriguez, Padilla, Malvar, Kallenbach, Santiago & Butron, 2018, Rodriguez, Santiago, Malvar & Butron, 2012). The early stem response to feeding by corn borers was characterized by the activation of signaling mechanisms mediated by phytohormones, whereas these molecules were marginally involved in the long-term response (Dafoe et al. , 2011, Rodriguez et al. , 2012). The stem long-term response was characterized by reorganization of the primary metabolism and a strong redox response mainly mediated by germin-like proteins to produce anti-nutritive and toxic compounds that reduce insect viability (Rodriguez et al. , 2018). However, no studies have simultaneously characterized short and long-term responses to stem attack by stem borers and the current work would be the first attempt to do it. A direct comparison of both responses using several genotypes under the same experimental conditions will shed light on real differences which cannot be disentangled from experimental differences when results from different experiments are compared.
As the components of the metabolome can be viewed as the end products of gene expression, un-targeted metabolomics would be a valuable tool to simultaneously monitor all biological processes operating in the plant response to herbivory and to discover bioactive compounds involved in plant resistance (Jansen, Allwood, Marsden-Edwards, van der Putten, Goodacre & van Dam, 2009, Schauer & Fernie, 2006, Sumner, Mendes & Dixon, 2003). Indeed, metabolomics has been proved as a valuable tool to characterize plant responses to feeding by insects and to identify plant metabolites contributing to increased susceptibility or resistance (Agut, Gamir, Jaques & Flors, 2015, Kang, Yue, Xia, Liu & Zhang, 2019, Liu, Hao, Hu, Zhang, Wan, Zhu, Tang & He, 2010, Marti, Erb, Boccard, Glauser, Doyen, Villard, Robert, Turlings, Rudaz & Wolfender, 2013, Ponzio, Papazian, Albrectsen, Dicke & Gols, 2017, Tzin, Hojo, Strickler, Bartsch, Archer, Ahern, Zhou, Christensen, Galis, Mueller & Jander, 2017, Wang, Qu, Zhang, Hu, Tang & Lu, 2016, Widarto, Van der Meijden, Lefeber, Erkelens, Kim, Choi & Verpoorte, 2006). Herbivory could induce extensive changes to both general and specialized metabolism to prevent the allocation of energy and nutrients to herbivore fitness, rather than to plant fitness because plant metabolites involved in defense are not just toxic, repellent, and/or anti-nutritive molecules, but compounds that could attract enemies of herbivores, participate in nutrient transport and storage to make nutrients less accessible to the insect or involved in phenology shifts that grant herbivore avoidance (Schuman & Baldwin, 2016).
The objectives were (i) to determine changes in the level of antibiosis of the stems induced by feeding of Sesamia nonagrioides(Mediterranean Corn Borer; MCB) larvae for two days (short-term feeding) and nine days (long-term feeding), (ii) to characterize the metabolome of the stem short and long-term responses to borer attack and (iii) to look for metabolic pathways that could modulate plant resistance to MCB