The need to discover alternative crop protection strategies that can be used to improve food safety and security, as well as for maintaining human health, has been the target of many investigations in recent years (Romanazzi et al., 2012; Burketova et al., 2015). In particular, studies have increasingly targeted exogenous molecules that induce defense responses (Walters et al., 2013). In this context, investigations on how non-toxic products can control plant diseases through activation of plant defense responses are fascinating. The effectiveness of compounds that have been described as ‘resistance inducers’ has been tested according to different crop protection strategies. These have shown encouraging results for their use as alternatives to traditional fungicides (Burketova et al., 2015; Oliveira et al., 2016). In particular, the effectiveness of alternative compounds in disease control has been tested in strawberry (Fragaria × ananassa), a perishable small fruit crop of great importance throughout the world, but which easily undergoes fungus-mediated preharvest and postharvest decay (Hemelrijcka et al., 2010).

Often the terms ‘resistance inducer’ or ‘elicitor’ have been used for molecules that can protect plants from diseases through induction of their defense mechanisms (Mandal et al., 2013; Walters et al., 2013). This inducible immunity is based on external recognition of ‘non-self’ signals, and notably for pathways of microbe/pathogen-associated molecular pattern (MAMP/PAMP)-triggered immunity (PTI) and effector-triggered immunity (ETI) (Jones and Dangl, 2006; Dodds and Rathjen, 2010). PTI is initiated in plants when PAMPs are recognized by pattern-recognition receptors. In contrast, ETI is induced by recognition of pathogen avirulence effectors by the host disease-resistance (R) proteins. This can lead to rapid and robust responses that are often associated with programmed cell death via the hypersensitive response, and with systemic acquired resistance (SAR) in the host (Ryals et al., 1994; Dodds and Rathjen, 2010). ETI, basal defense, and PTI act through a common set of signaling components, which include multiple regulatory proteins, reactive oxygen species (ROS), Ca²⁺ signaling, and the phytohormones salicylic acid (SA), ethylene (ET), and jasmonic acid (JA) (Spanu, 2012). These signaling events modulate transcription factor (TF) activities that can lead to massive transcriptional reprogramming. This, in turn, results in accumulation of different enzymes and stress-specific metabolites, such as pathogenesis-related (PR) proteins, hydrolytic enzymes, peroxidases and phytoalexins, and deposition of lignin and callose (Lamb and Dixon, 1997; Oliveira et al., 2016). Induction of resistance in host tissues can also lead to production of beneficial antimicrobial compounds (Romanazzi et al., 2016). The onset of PTI and ETI from infected loci often triggers induced resistance in distal tissues that can confer resistance against a broad spectrum of pathogens (Pieterse et al., 2014). SAR is frequently associated with increased levels of SA and coordinated activation of PR genes, which can promote one or more long-distance signals that enhance the defensive capacity of the plant (Fu and Dong, 2013).

Systemic resistance can also be induced by beneficial microbes that are normally associated with plant rot, which is known as induced systemic resistance (ISR), a process that is usually SA and PR-protein independent. During SAR and ISR responses, plants can obtain systemic resistance against different classes of pathogens for several days (Heil and Bostock, 2002). These compelling features of SAR as a defense response are the basis of the induced resistance concepts that underlie the application of specific elicitors.

In strawberry, many non-toxic compounds have been effective in studies carried out with postharvest application, including benzothiadiazole (BTH; also known as acibenzolar-S-methyl) and chitosan (Cao et al., 2011; Romanazzi et al., 2013). This also applies to their spraying before harvest in plastic tunnels (Terry and Joyce, 2000) and in the open field (Feliziani et al., 2015). BTH is a light-insensitive analog of SA that can activate defense responses that lead to SAR, and it is an efficient broad-spectrum resistance inducer against bacterial, fungal, and viral diseases in different monocot and dicot crops (Walters et al., 2013). Chitosan is a deacetylated derivative of chitin that is derived from N-acetylglucosamine units linked by β-1,4-glycosidic bonds; it is produced by chitin deacetylases, and it occurs naturally as a polysaccharide (Bautista-Baños et al., 2006; Romanazzi et al., 2017). Chitosan treatment produces a coating on the surface of fruit that reduces the gas exchange, which slows down respiration and ripening of the fruit (Romanazzi et al., 2009). The ability of BTH and chitosan to induce gene expression and enzyme activity has been tested in several crops. BTH spraying induces SAR genes that encode PR proteins, and the chitinase, glucanase, and ROS scavenger enzymes (Görlach et al., 1996; Ren et al., 2012; Zhu et al., 2016). Moreover, chitosan can induce plant defense enzymes and synthesis of secondary metabolites in several plant species, such as polyphenolic compounds, lignin, flavonoids, and phytoalexins (Coqueiro et al., 2015; Malerba and Cerana, 2016).

In strawberry fruit, BTH and chitosan elicit defenses by increasing enzyme activities and the expression of specific genes (Cao et al., 2011; Landi et al., 2014). However, a global analysis of the transcriptome responses associated with these elicitor compounds has not yet been performed.

RNA sequencing (RNA-Seq) analysis is a powerful tool to study transcriptomes (Trapnell et al., 2010; AbuQamar et al., 2016). This large-scale analytical approach of gene expression can be crucial to determine the effects of elicitors on plant metabolism. Therefore, here, RNA samples from fruits of strawberry plants treated preharvest with BTH and chitosan were analyzed at 6, 12, and 24 h posttreatment (hpt), and transcript abundances were determined to characterize changes in gene expression patterns.

Benzothiadiazole and chitosan have been shown to reduce gray mold and Rhizopus rot following preharvest (Feliziani et al., 2015) and postharvest (Romanazzi et al., 2013) treatments, and to modulate selected defense genes (Landi et al., 2014). To contribute to the understanding of the whole transcriptional changes induced by BTH and chitosan in strawberry fruit, following preharvest application of these elicitors to the plants, RNA-Seq data were generated and validated by RT-qPCR. The differences in the magnitudes of changes observed between the RT-qPCR and RNA-Seq data relate to the differences between these approaches for gene expression analysis.

The RNA-Seq analysis carried out at 6, 12, and 24 hpt revealed that more than 5,000 genes were differentially expressed for each elicitor, as compared to the control. However, at each time, less than 21% of the transcripts were affected in the same way by both elicitors, and within each elicitor response, less than 1% of the transcripts were modulated at all of the times analyzed. This shows the great variability of such gene modulation over time. Enrichment analyses showed the involvement of genes associated with the light phase of photosynthesis, HSPs, storage proteins, and defense signaling. However, these BTH and chitosan treatments modulated the gene transcripts associated to photosynthesis in different ways according to time and, partially, to the chloroplast or nuclear localization of the genes (Rogalski et al., 2015).

Briefly, at 6 hpt, BTH induced significant down-regulation of several photosynthetic nuclear-encoded genes, including the Lhc genes. LHC has an important role in regulation and dissipation of excess energy flow under light stress. The expression of Lhc is coordinately repressed when the energy input through the antenna protein systems exceeds the requirement for CO₂ assimilation (Külheim et al., 2002; Teramoto et al., 2002). Conversely, in the early phase after treatment with chitosan, there was down-regulation of few nuclear-encoded genes as Lhcb7, Psak, and PsbS. The PsbS gene encodes for a crucial pH-sensing protein in the non-photochemical quenching (NPQ) process (Ruban, 2016). NPQ is the thermal energy dissipation process induced by high light protonation that causes pH changes in the thylakoid. The function of NPQ is thus to prevent damage of the photosystems that control the generation of ¹O₂ (Roach and Krieger-Liszkay, 2014; Ruban, 2016). On the other hand, at 6 hpt, a few chloroplast-encoded genes were primarily down-regulated by both elicitors, including the psbA gene. PsbA encodes the D1 protein that with the D2 protein constitutes the core reaction of PSII, and has a role in protecting PSII, which is highly susceptible to photo damage, as a specific sensor of ROS (Huo et al., 2015). Although the NPQ activation depends on the PsbS protein, the pH change of the thylakoid is mainly generated by electron transport, which involves oxidation of water in the oxygen-evolving complex. In this regard, we note the association of BTH treatment with up-regulation of PsbQ, which codes for a protein that belongs to the oxygen-evolving complex in PSII. Previous studies have indicated that PsbQ is the target for effector-protein-triggered immunity in Pseudomonas syringae (Romero-Puertas et al., 2008), and is required for full deployment of ROS associated with plant defenses (Rodríguez-Herva et al., 2012). Therefore, despite the observed differences, both elicitors impact on ROS production.

At 12 hpt, a reversal of the trend occurred in the strawberry fruit. In particular, chitosan induced up-regulation of the main chloroplast genes, which were initially down-regulated, while at 24 hpt, the nuclear genes were mainly up-regulated by both elicitors. The expression of the photosynthetic genes changes according to the high or low input of light, with this being down-regulated or up-regulated, respectively (Häusler et al., 2014). Previous studies have frequently shown down-regulation of photosynthesis associated with pathogen attacks (Berger et al., 2007), environmental stress (Song et al., 2014; Kosová et al., 2015) and toxic effects of fungicides (Petit et al., 2012). We have shown that the resistance inducers tested here that are effective in disease control, also affect the photosynthetic processes without destroying them. Indeed, the systems were regenerated and an overexpression of the genes involved was observed at the later time, which appears to be induced to restore the equilibrium and for the increased protein metabolism that is useful in plant defense. Proteomic and biochemical analyses of resistant and susceptible plants have shown that the ability to maintain active photosynthesis during an infection is a crucial element in plant defense (Zhang et al., 2013). In this regard, the role of photosynthesis in non-photosynthetic fruit tissue and on its metabolism is not clear. For the strawberry fruits analyzed in the present study, green tissues were associated only with achenes (Meyerhoff and Pfündel, 2008). However, a study on tomato has indicated that ripe fruits are unlikely to be net assimilators of CO₂, despite the high levels of expression of the photosynthetic genes, which suggests that these have a role in the improvement of fruit quality (Cocaliadis et al., 2014). The present data suggest that these elicitors influenced chloroplast functionality in different ways, with effects on the network transcriptome responses. The localization of the genes that encode the chloroplast proteins implies that there are molecular and physiological mechanisms that coordinate nuclear and plastid gene expression (Jarvis and López-Juez, 2013). This led to the development of the concept of retrograde signaling from chloroplast to nucleus (Koussevitzky et al., 2008; Kleine and Leister, 2013). However, the signaling between chloroplasts and the nucleus is bidirectional. In anterograde regulation, for adequate plastid development (Lefebvre-Legendre et al., 2015), nuclear-encoded regulators can modify the expression of both chloroplast and nuclear genes, even at the post-transcriptional level (Kleine and Leister, 2013). These signaling networks are dependent on plastid developmental and functional stage, intracellular distance, time scale, and target place of action (Szechyñska-Hebda and Karpiñski, 2013). In the present study, not all of the photosynthetic genes were affected by BTH and chitosan, and some of them showed different trends. This might be linked to the evolution of transduction signals that can be independent (e.g., PsbS, Lhc; Engelken et al., 2010), unique (e.g., psbE, psbF; Kiss et al., 2012), or associated with different protein characteristics (e.g., Lhcb7; Klimmek et al., 2006).

According to the data from the present study, the gene expression fluctuations of the affected chloroplast genes occurred from 6 to 12 hpt, while for the nuclear genes, this was mostly from 6 to 24 hpt, which stresses the importance of their intracellular localisation. Several candidates of retrograde signaling pathways have been proposed, including metabolite abundance, ROS pathways, and photosynthetic pigments (Singh et al., 2015). However, it has been recently suggested that all of these can be traced back to photosynthetic electron transfer energy balance/imbalance as the initiator of the chloroplast retrograde signaling cascade (Foyer et al., 2012; Gollan et al., 2015). Chloroplasts coordinate cellular activities and functions under stress responses, to promote survival against environmental perturbations (Gururani et al., 2015). This behavior can be used as a defense strategy, which affects carbohydrate metabolism in the stressed plant tissues. As an adaptive response to biotic and abiotic stresses, this will allow the plant to invest resources in the suddenly required defense responses, without this being debilitating to the plant. In this context, the present data highlight that the elicitors affect the chloroplasts through their actions on the electron flow, and consequently the redox/ROS status. This induces a transcriptome reprogramming in the nuclear genome, which includes up-regulation of defense-related genes (Driedonks et al., 2015; Gollan et al., 2015).

The production of specific ROS might be strictly involved in the elicitor responses linked to the chloroplast. However, up-regulation of the ROS scavenger was not observed in the early phases after these treatments. Whether ROS act by damaging the cell protective or signaling factors depends on the delicate equilibrium between ROS production and scavenging. Hence, ROS can damage cells as well as initiate responses, such as gene expression modulation and cellular immune responses (Gill et al., 2013). In terms of the role of ROS in the signaling networks, the PRX ROS-scavenger genes affected in the present study showed the same fluctuations in their expression that were observed for the photosynthetic nuclear-encoded genes, which suggests similar regulatory networks. PRX acts as an intracellular redox sensor and it transmits information related to the cellular levels of ROS (Awad et al., 2015). A role for PRX as a retrograde signal during oxidative stress under light conditions has also been suggested (Dietz, 2011).

Up-regulation of the TRX genes at 12 hpt with chitosan might have a role in the detoxification mechanisms (Hanschmann et al., 2013). The TRX detoxificant proteins are known to regulate numerous photosynthetic enzymes, and cross-talk between the plastid and the TRX system has been suggested to mediate light-dependent activation of primary photosynthetic reactions in plant chloroplasts, through reduction of disulphide bridges in redox-regulated enzymes (Meyer et al., 2012; Nikkanen et al., 2016). Based on the present data, we note that chitosan induced the simultaneous over-expression of chloroplast-encoded genes and TRXs at 12 hpt. However, other analytical approaches will be needed to clarify which molecules are actually involved in the signaling network.

At 12 hpt, KEGG analysis showed different involvement of the ‘glutathione metabolism pathway,’ which was here represented by the GST genes that were up-regulated by BTH and down-regulated by chitosan. GST is known as a cell detoxification system, and it is involved in the NPR1-independent SA-mediated pathway (Ghanta et al., 2011). In the same way, GRX-C9 was induced only by BTH, and this is SA dependent (Herrera-Vásquez et al., 2015), in agreement with the SAR responses known to be induced by BTH (Görlach et al., 1996).

The present data suggest that the different initial impacts of these two elicitors on chloroplast functionality are pivotal keys that steer the cascade signaling pathways. We have highlighted here the predominant influence of BTH (an analog of SA) on strawberry genes linked to LHC components, and subsequently genes of other photosynthetic complexes that are mainly involved in the nuclear genes network. Previous studies have shown that the application of high concentrations of SA to Arabidopsis leaves rapidly induces stomatal closure and reduces the electron transport rate (Janda et al., 2012). Typically, high light-induced changes in stomatal conductance causes rapid repression of LHC genes and results in a photo-respiratory oxidative burst that helps to inform the cell of the redox changes in the PET chain, and to influence expression of photosynthetic genes (Foyer et al., 2012).

Strawberry plants treated with chitosan underwent partial photo-inhibition, as seen by down-regulation of specific genes in the strawberry fruit. The early responses to chitosan were down-regulation of genes encoding HSPs that are normally induced by abiotic and biotic stress. In particular, several genes encoding HSP20 proteins were affected, which prevent thermal aggregation of proteins (Park and Seo, 2015). This response might be associated with the properties of chitosan. In contrast to BTH, when chitosan is applied to plant tissues, it forms a physical barrier that results in decreased transpiration (Romanazzi et al., 2009), which affects the sensitivity of both heat stress and light signaling, as underlined by the results of the enrichment analysis. Heat stress commonly leads to inhibition of photosynthesis in higher plants (Song et al., 2014). However, the early down-regulation of the HSP genes might have a negative effect on protein refolding, and thus affect ROS production (Zeeshan et al., 2016). This is indicated as the ‘protein processing in endoplasmic reticulum’ pathway that is associated with HSP down-regulation following chitosan spraying. This trend changed at 12 hpt, when up-regulation of the HSP genes was observed with both elicitors, which highlights again the elicitor-induced modulation and recovery of gene responses. Therefore, the impact of the elicitors on the strawberry tissues was like an induction of stress, which, in the early phase from treatment for BTH, was linked with photosynthetic process, while, for chitosan, it was mostly linked with a heat response. Then, cross-talk between the heat response and the photosystem processes can be suggested (Kindgren et al., 2012; Song et al., 2014; Zhou et al., 2014). The present study showed up-regulation of specific TFs, including the NAC family, which is involved in heat stress and photosynthesis responses (Huang et al., 2015; Shao et al., 2015), associated to light responses, with FAR1 and HY5 (Bae and Choi, 2008; Kindgren et al., 2012), and involved in ROS homeostasis signaling (Wang and Wang, 2015). In addition, the HY5 genes are associated with retrograde signaling (Kindgren et al., 2012). This suggests that these TFs function in the regulation of multiple stress tolerance through modulation of transcriptome reprogramming responses.

One of the earliest signaling events after recognition of a pathogen is seen for the ion fluxes across the plasma membrane, which includes influx of Ca²⁺ into the cytosol (Lecourieuxn et al., 2006). The present data showed the involvement of Ca²⁺ signaling with both elicitors. Differences among the elicitors were observed through the analysis of the activation of the hormone-signaling molecules that are associated with the inducible immune responses. The DEG analysis showed that SA signaling was up-regulated with BTH, while ET metabolism (synthesis/degradation) was modulated by chitosan, with both occurring in the early phase after their application. However, hormonal cross-talk involving the WRKY TFs appears to have a major role in the induced hormonal changes that modulate disease and resistance (Spoel and Dong, 2008; Huot et al., 2014). WRKY50, WRKY51, and WRKY53 TFs are known repressors of JA signaling and are involved in hormonal cross-talk (Pieterse et al., 2012), and in the present study they were induced by both elicitors. Up-regulation of several transcripts of the WRKY70 and WRKY40 genes was associated exclusively with BTH treatment, and these might have pivotal roles in determining the balance between SA-dependent and JA-dependent defense pathways (Kazan and Lyons, 2014).

Salicylic acid has been indicated as the principal signal in SAR (Fu and Dong, 2013; Coqueiro et al., 2015). SAR activation results in the coordinated production of PR proteins (Pieterse et al., 2012). Here, the GO analysis at 6 hpt indicated that the cell-wall and extracellular-matrix terms, and the xyloglucan transferase activity, were induced by BTH. This suggests that remodeling of the cell-wall architecture is important to enhance disease resistance, as well as the involvement of structural proteins with PR-protein functions.

PR1 is a useful molecular marker for the SAR response (Fu and Dong, 2013). BTH is known to mimic the defense-associated effects of SA (Görlach et al., 1996), and for this reason, up-regulation of the PR1 gene was expected. Nevertheless, PR1 was up-regulated by BTH only at 24 hpt. Other central regulators in the SAR response were up-regulated by BTH, including NPR3 (a paralog of NPR1; Fu et al., 2012; Kuai et al., 2015) and the NIMIN proteins that are active at high SA concentrations (Fu and Dong, 2013) as regulators of late SAR genes (e.g., PR1), for prevention of their premature activation (Glocova et al., 2005; Weigel et al., 2005; Hermann et al., 2013). This is in agreement with our previous studies in strawberry fruit that showed that BTH did not affect PR1 gene expression (Landi et al., 2014). The NIMIN–NPR connection might constitute a molecular device to monitor SA levels in diseased plants, which would allow the plant to translate gradually to an increasing gradient of the defense hormone SA in two clear decision steps, as early and late SAR gene expression. These mechanisms of regulation of SAR were not activated by chitosan. However, chitosan caused the induction of genes associated to PR proteins including PR1.

Starting from 12 hpt, the photosystem process recovered, with up-regulation of photosynthetic genes induced by both elicitors. This had strong consequences on protein metabolism. Indeed, the elicitors affected the protein network, which included ribosomal and storage proteins, as well as disease-resistance proteins. Plant adjustments to an altered environment require high numbers of novel proteins to be synthesized, as well as for proteins to be degraded (Kosová et al., 2015). The elicitors strongly affected modulation of the ribosomal genes, which were mainly up-regulated at 12 hpt and down-regulated at 6 and 24 hpt. The present results suggest that ribosomal proteins change in a stress-specific manner, as part of the adaptation to elicitor stress, which will probably have biased protein translation (Wang et al., 2013). Several genes that encoded resistance proteins were induced by the elicitors, and up-regulated genes were predominant at 6 hpt and again at 24 hpt, which underlines the modulation of defense-response genes according to time.

Massive induction was recorded for genes that encode storage proteins (e.g., legumins, 12S storage proteins, vicilins, patatins) and genes involved in lipid metabolism (e.g., oleosins). In particular, strong up-regulation was observed at 24 hpt for both elicitors. This suggests that the expression of genes that encode storage proteins correlates with over-expression of genes involved in the photosynthetic process. In addition to having essential roles for plant survival, the storage proteins have roles in defense mechanisms (Cândido et al., 2011) through their insecticidal and antimicrobial properties, as has been observed for vicilin and patatin (Banerji and Flieger, 2004).

Our data confirm that plant responses to elicitors are dynamic processes that induce deep changes in the kind, quantity and timing of the genes involved. This establishes novel homeostasis between plants and their environment that can enhance plant defense mechanisms against pathogens. The crucial impact of BTH and chitosan on the photosynthetic process generally begins with down-regulation, followed by over-expression of fundamental photosynthetic genes. This helps to maintain the imbalance/balance of ROS/redox signaling, and attributes a key role to the chloroplasts as the sensors of environmental changes, which allows them to protect the photosynthetic apparatus from stresses. However, the specific characteristics of these resistance inducers, such as the formation a protective film on plant tissues by chitosan and the analogous actions of BTH compared to SA, drive the response network in the early phase after their application. The typical SA signaling during plant immunity was found to be associated with BTH. However, the involvement of PR proteins with both elicitors, and in particular of PR1, which is one of the SAR response markers, suggests SAR induction also for chitosan.

We have here highlighted that the resistance inducers BTH and chitosan that are effective in the control of postharvest diseases of strawberry (Feliziani et al., 2015) deeply modulate the cellular metabolism. The genes identified in the present study can represent markers to better elucidate plant/pathogen/resistance-inducer interactions and to design novel sustainable disease-management strategies.

LL performed the experiments, analyzed the data, and wrote the article; RMDMA performed most of the experiments, analyzed the data, and supervised and complemented the writing; SP supervised the writing; EF performed part of the experiments; FF designed the experiments, and supervised and complemented the writing; GR designed the experiments, supervised and complemented the writing, and coordinated the collaboration of the authors.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer EA and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.