Bioeffectors as biotechnological tools of the innate immunity: signal transduction pathways involved

Background Unravel the complex functioning of plant immune system is essential and something in which great effort is being made since its performance is not entirely clear yet. Knowing plant immune system allows strengthening it and therefore developing a more efficient and environmentally friendly agriculture, avoiding the massive use of agrochemicals and making plants the main protagonist in the defense against pathogens. The use of beneficial rhizobacteria (bioeffectors) and its derived metabolic elicitors are biotechnological alternatives in plant immune system elicitation. The present work aimed to check the ability of 25 bacterial strains selected from a group of 175, isolated from the rhizosphere of Nicotiana glauca , to trigger the innate immune system of Arabidopsis thaliana seedlings against the pathogen Pseudomonas syringae DC3000. A study of the signal transduction pathways involved in plant response was made. Results The selected 25 strains were chosen because of their biochemical traits and avoiding phylogenetic redundancy. The 5 strains, of the previous 25, more effective in the prevention of pathogen infection were used to elucidate signal transduction pathways involved in the plant immune response, studying the differential expression of Salicylic acid and Jasmonic acid/Ethylene pathway marker genes. Some strains stimulated the two pathways with no inhibitory effects between them, while others stimulated either one or the other. Metabolic elicitors of two strains, chosen for their taxonomic affiliation and for the results obtained in the differential expression of the genes studied, were extracted using n-hexane, ethyl acetate and n-butanol, and their capacity to mimic bacterial effect to trigger the immune system of the plant was studied. N-hexane and ethyl acetate were the most effective fractions against the pathogen in both strains, achieving similar protection rates although gene expression responses were different from that obtained by the bacteria. Conclusions Beneficial rhizobacteria and its metabolic elicitors have great potential as biotechnological tools since they are able to improve plant immune system through the triggering of either Salicylic acid or Jasmonic acid/Ethylene pathway or both pathways simultaneously. These results open a huge amount of biotechnological possibilities to develop biological products for with respect to the total of leaves seedlings Negative control (seedlings inoculated only with nutrient broth and pathogen challenged) was considered as 0% of protection and then data were with positive whose of therefore, significant decreased, which indicate that the metabolic elicitors present in these fraction were only activating the SA mediated transduction pathway, while the bacterial strain activated both. These results show that the elicitors detected by the plant in both cases have to be different, and so would be the PRRs involved in that response All these results show the great number of possibilities offered by elicitors to trigger the immune system of plants, which opens a plethora of biotechnological solutions to different stress situations. total of the plants with 250 mL. Plants were incubated in a culture chamber (Sanyo MLR-350H) with an 8 h light (350 μE s -1 m -2 at 24 °C) and 16 h dark period (20°C) at 70% relative humidity for 72 h, and disease severity was recorded as the number of leaves with disease symptoms relative to the total number of leaves. Results were relativized using the disease

In contrast, Pieterse et al. 2000 described ISR as an answer triggered by non-pathogen rhizobacteria (bioeffectors). However, different elicitors such as antibiotics, surfactants or chemical inducers (Gozzo and Faoro 2013) are also able to induce ISR. In this case, ISR response was described as dependent on jasmonic acid (JA) and ethylene (ET) signalling pathways and also needs the involvement of NPR1 Van Loon 2004, Pieterse andVan Loon 2007). Plant defensin1 (PDF1) (Berrocal-Lobo et al. 2002, Lorenzo et al. 2003, and MYC2 also play an essential role in this signalling pathway (Pozo el tal. 2008, Pré et al. 2008).
These bioeffectors and some of their elicitors (structural molecules or metabolic molecules released to the medium) induce in plants a physiological alert state prior to stress challenge known as priming (Conrath et al. 2002). Plants in this state are able to develop a faster and/or stronger activation of defensive responses after the attack of pathogens, insects or in response to abiotic stress (Conrath et al. 2006). After bioeffectors or their elicitors are sensed, the SA, JA or ET signalling pathways are activated to trigger plant resistance (Wu et al. 2018). Therefore, the study of these transduction signal pathways is meaningful for understanding the plant immune system and their defences against pathogens. This can contribute to promote the use of bioeffectors and their elicitors as a useful biotechnological strategy to develop a sustainable agriculture without using agrochemicals and pesticides (Wu et al. 2018).
Taking advantage of the well-known ability of the plants to strongly select beneficial bacterial strains in the rhizosphere to survive to adverse conditions (Marilley and Aragno 1999, Lucas Garcia et al. 2001, Berendsen et al. 2012, Stringlis et al. 2018, bacteria isolated from the rhizosphere of Nicotiana glauca, a Solanaceae native to Southern Spain with a strong secondary metabolism (Ramos-Solano et al. 2010), were studied. The effects induced in the plants by the beneficial rhizobacteria depend on molecules (elicitors), so we considered that after the extraction of these elicitors, it would be possible to find out which ones were able to reproduce the effects of the rhizobacteria and therefore were responsible of this effect.
The general objective of this work was to find beneficial rhizobacteria (bioeffectors) from Nicotiana glauca rhizosphere efficient in triggering the innate defence response of Arabidopsis thaliana plants, as well as effective derived metabolic elicitors, trying to elucidate the mechanisms involved in the protection. To achieve this objective the following partial objectives were defined: i) to perform a screening of N. glauca rhizobacteria to select those strains efficient in triggering the innate response of Arabidopsis plants against the pathogen Pseudomonas syringae DC3000, ii) to study the mechanisms involved in plant defence triggered by the most effective bioeffectors against the pathogen P.syringae DC3000, iii) to obtain metabolic elicitors from the most effective bioeffectors and assay their ability to mimic bacterial response.
To reach our goals, ISR experiments were carried out in A.thaliana plants using the bioeffectors and the metabolic elicitors of the chosen strains to protect the plants against P.syringae DC3000; the differential expression of marker genes for the SA and JA/ET transduction pathways were studied on plants inoculated with selected strains and selected metabolic elicitors. In the Gram-negative group, eight genera were found (Serratia, Enterobacter, Pantoea, Erwinia, Cronobacter, Acinetobacter, Pseudomonas and Stenotrophomonas), being Pseudomonas especially diverse in species (5 species identified: P. putida, P. reinekei, P. brassicacearum, P. fragi and P. fluorescens). In the Gram-positive group, only two genera were found, (Bacillus and Brevibacterium).

Beneficial rhizobacteria screening: phylogenetic tree and biochemical tests
Within Bacillus, two species were especially abundant, Bacillus cereus and Bacillus megaterium (See Additional file 1).
Biochemical tests (auxin-like compounds production (Sergeeva et al. 2007), siderophores production (Alexander and Zuberer 1991), phosphate solubilisation (De Freitas et al. 1997), and chitinases production (Frändberg andShnurer 1998, Rodríguez-Kábana et al. 1983)) for identifying putative beneficial rhizobacteria were carried out to the 175 strains. The results of these tests are shown in Table 1. Enterobacter was the only genus across all isolates tested that were capable of producing indole acetic acid (IAA). Siderophore producing isolates were present in all genera. Acinetobacter and Pseudomonas showed the highest percentage of phosphate solubilisers, but also isolates of Enterobacter, Pantoea and Erwinia were able to solubilise phosphate. Finally, all Stenotrophomonas isolates were able to produce chitinases (100%). Isolates able to produce siderophores and also solubilise phosphates belonged to Acinetobacter, Pseudomonas, Enterobacter, Pantoea and Erwinia.
Those able to produce siderophores and also chitinases were present among Stenotrophomonas and Pseudomonas, although less abundant among the latter (2.08%). The unique genus that had isolates with three biochemical traits was Enterobacter. It was able to produce siderophores and IAA and also, to solubilise phosphate.
Within Gram-positive bacteria, none of the isolates produced IAA, however all were able to produce siderophores. Only Bacillus cereus, B. megaterium and Brevibacterium sp. were able to solubilise phosphate. B. cereus and B. subtilis were able to produce chitinases. The isolates that were able to produce siderophores and also solubilise phosphates were B. cereus, Brevibacterium sp. and B. megaterium. The isolates that were able to produce siderophores and also chitinases were B. cereus and B. subtilis. The unique isolate that had three biochemical traits was B. cereus. It was able to produce siderophores and chitinases and also to solubilise phosphate.ISR by beneficial rhizobacteriaAccording to the results obtained from the phylogenetic tree (See Additional File 1) and the biochemical tests (Table 1), twenty-five strains were chosen (fifteen Gram-negative and ten Grampositive) to develop a first protection experiment against the pathogen DC3000. All selected strains had at least two or three biochemical traits, except N 10.7 Serratia odorifera, N 12.34 S. rubidaea and N 11.14 Bacillus endophyticus that only had one activity, but they were able to reduce growth of other strains in plate (data not shown), probably due to the production of antibiotics. The selected strains and their biochemical traits are shown in table 2. Table 3 shows the percentage (%) of protection induced in seedlings of A. thaliana inoculated with the twenty-five selected strains and the percentage of protection of negative and positive control plants. All Gram-negative bacteria significantly protected against the pathogen, except N 8.22,N 10.6,N 10.21,N 15.23 and N 18.10. Protection achieved by N 16.24 was not statistically significant. N 5.12 (P. putida),N 8.17 (S. maltophilia),N 12.34 (S. rubidaea) and N 21.24 (P. fluorescens) were the Gram-negative bacteria that induced the highest protection, even above of that of the positive control. Therefore, these four strains were chosen for assessing differential gene expression of 8 genes, markers of different signal transduction pathways related to plant immune system. Within Gram-positive bacteria, all of them significantly protected against the pathogen, except N 11.14, N 11.22 and N 11.36. Strain N 4.1 (B. cereus) was the Gram-positive bacterium that performed best, so it was selected to assess the differential gene expression of 8 genes, markers of different signal transduction pathways related to plant immune systems. Biochemical traits are Indole Acetic Acid (IAA) production, siderophores production, phosphate solubilisation, chitinases p  with respect to the total of leaves (n=16 seedlings per replicate). Negative control (seedlings inoculated only with nutrient broth and pathogen challenged) was considered as 0% of protection and then data were relativized with respect to it. A positive control (BTH) was also used. Strains in bold are those whose percentage of protection against the pathogen DC3000 exceeded that of the positive control and therefore, those that were selected for further analyses. Asterisks indicate that there were significant statistical differences (p< 0.05) with respect to negative control.
Differential gene expression 6, 12 and 24 hours after pathogen challenge ( The third pattern was a significant increase 24 hapc, followed by strains N 8.17 (Fig. 2)  its metabolic elicitors and to check the capacity of these metabolic elicitors to mimic protective effects of bacteria. They were selected N 12.34 (S. rubidaea), the Gram-negative strain that showed the highest differential expression (Fig. 3) and N 4.1 (B. cereus) as it was the Gram-positive strain with better protection among the Gram-positive and which ranked second among all ( Table 3).
The three fractions extracted from each strain (n-hexane, ethyl acetate and n-butanol), achieved significant protection (  was considered as 0% of protection and then data were relativized with respect to it. A positive control (BTH) was also used. Fractions in bold are those whose percentage of protection against the pathogen DC3000 exceeded that of the positive control and therefore, those that were selected for further analyses. Asterisks indicate that there were significant statistical differences (p< 0.05) with respect to negative control. The two metabolic elicitor fractions from N 12.34 induced the same behaviour in the genes studied: expression of NPR1 and PR2 increased from 6 to 12 hapc, while PDF1 decreased. Both metabolic elicitor fractions from N 4.1 also had the same behaviour: expression of NPR1 and PDF1 decreased from 12 to 24 hapc, while PR3 increased.

Discussion
In the present study, the efficiency of bioeffectors and derived metabolic elicitors to trigger the immune system of A. thaliana conferring protection against P.syringae DC3000 has been shown.
The 175 strains were isolated in 2010 (Ramos- Solano et al. 2010) from the rhizosphere of wild populations of N. glauca. This plant species was chosen as it was hypothesized that its very active secondary metabolism would select a good group of bacteria to ensure plant fitness.
The rationale of plant's selection capacity has been widely demonstrated, and also the use of the rhizosphere as a source of highly specialized strains (Anwar et. al 2016, Aarab et al. 2015, Lucas et al. 2013, Ramos Solano et al. 2006, Barriuso et al. 2005, since it is one of the most complex and diverse ecosystems on earth. This suggests a definite role of plant-derived metabolites in the microbiome assemblage in the rhizosphere (Hacquard et al. 2017). According to previous results, the common culturable bacterial genera in the rhizosphere of N. glauca includes Bacillus sp., Pseudomonas sp., Enterobacter sp., Acinetobacter sp., Burkholderia sp., Arthrobacter sp., and Paenibacillus sp. (Ramos-Solano et al. 2010).
In the present study, almost 100% of the strains produced siderophores. Siderophore production is related to iron limiting nutrient (Lucas et al. 2013, Raymond et al. 1984, Jin et al. 2006), but also has been related to biocontrol and/or systemic induction of secondary metabolism, and therefore, siderophore-producing strains may have the ability to protect plants against pathogens through complex and inducible secondary metabolism, which is probably related to defence (Sinclair et al. 2004, Barriuso et al. 2008).
Regarding the production of auxins and the ability to solubilise insoluble phosphorus, only one genus of those of our study was capable of producing auxins (Enterobacter sp). However, the solubilisation of phosphates was a very abundant activity among the strains studied. Our results support that N.
glauca selects rhizobacteria related to nutrition or biocontrol activities (phosphate solubilisation and siderophore production) rather than those able to affect plant growth regulator balance (auxins production).
The production of chitinases was well represented within the Gram-positive group, but among the Gram-negatives, only the Stenotrophomonas genus was able to produce them, consistent with Ramos Solano et al. (2010). Many species of rhizosphere microorganisms produce chitinolytic enzymes to protect themselves against fungi, since chitin is a major structural component of most fungal cell walls. Therefore, these microorganisms have an excellent potential as biocontrol agents (Lorito et al. 1993, Sid et al. 2003, Adesina et al. 2007).
The strains that were selected for ISR experiment were able to produce siderophores, and they had also some other complementary capacities, mainly the production of chitinases. This selection criterion has already been used by other authors with the aim of finding bacteria capable of inducing systemic resistance in plants ( Ramos-Solano et al. 2010, Van Loon et al. 1998, Ramamoorthy et al. 2001. N 16.15 (Enterobacter sp.) was the only non-siderophore producing isolate, but it was one of the two strains that produced auxins, and was chosen for this reason. Some authors have shown that auxins are related to the induction of systemic resistance (Akram et al. 2016, Petti et al. 2012). Three strains, N 10.7 (S. odofirera), N 12.34 (S. rubidaea) and N 11.14 (B. enterophyticus) were chosen with only one biochemical trait, because of their capacity to reduce growth of other strains in plate (data not shown), probably due to the production of antibiotics. This working scheme has proved to be very effective, since 16 out of the 25 strains chosen induced systemic resistance against the pathogen DC3000 (Table 3).
To determine signal transduction pathways triggered by the five outstanding strains, from the 25 previously selected, the differential expression of marker genes of the SA and JA/ET signalling pathways was studied. For this experiment, the criterion followed for the bioeffector selection was the highest protection against P. syringae DC 3000 infection within both bacterial groups (Gram-positive and Gram-negative). To date, most bioeffectors studied for their ability to trigger ISR mechanisms belong to the group of Gram-negative bacteria, especially bacteria of the genus Pseudomonas.
However, Gram-positive bacteria, and among them, those of the genus Bacillus, have gained much importance in the last decade because of the great potential to trigger resistance mechanisms against a wide range of pathogens (Kannojia et al. 2018, Gutierrez Albanchez et al. 2018. Three types of defensive responses were detected, according to the time needed to increase gene expression: rapid, intermediate and slow. The rapid response (6 hapc) was generated by strains N.5.12 (P. putida) (Fig. 1) and N 21.24 (P. fluorescens) (Fig. 4). N 5.12, induced a strong differential expression of NPR1, a marker of SA pathway, PDF1 and PR3, markers of the JA/ET pathway.
Interestingly, N 21.24 induced a strong differential expression of ICS and LOX2 involved in SA and JA synthesis, respectively. The intermediate response (12 hapc) was produced by N 12.34 (S. rubidaea) ( Fig. 3), which induced a strong differential expression of markers of SA pathway (NPR1 and PR2), and markers of the JA/ET pathway (PDF1 and PR3). The different behaviour generated by these three strains is also reflected in their defensive capacity. Although the three induced resistance above the positive control (BTH), N 5.12 and N 12.34 induced a lower protection than N 21.24, which was the most effective of all the tested. Contrary to Caarls et al. (2015), we observed a simultaneous high expression of NPR1 and PDF1 6 hapc for N 5.12 and 12 hapc for N 12.34, suggesting that SA is not suppressing the expression of PDF1 as these authors indicated. This may be related to the monomerisation process of NPR1 protein, present in the cytoplasm (which has not been determined in this work) as well as with the location of this protein ( Although metabolic elicitors of the two fractions studied protected to the same extent as the bacteria, the expression of the analysed genes has different behaviours. The strain N 12.34 induces gene expression levels more intensely (up to 140 times. Figure 3) than metabolic elicitors ( Fig. 6a and b).
The different intensity could be due to either the abundance of elicitors when the bacteria is delivered alive, holding all determinants, as compared to a subset of the same elicitors delivered on fractions, or because the plant is more sensitive to elicitors not present in the hexane and ethyl acetate fractions. The large difference in the levels of genetic expression indicates a level of priming also different. It is known that the priming can modify the distribution of energetic resources compromising plant growth in favour of a more production of metabolites involved in defensive response (Lucas et al. 2014, Van Hulten et al. 2006. Therefore, in this case the use of metabolic elicitors may have advantages over bioeffectors. Interestingly, metabolic elicitors in both fractions from Serratia N 12.34 were able to activate the SA pathway, increasing the expression of NPR1 and PR2 ( Fig. 6a and b). In both fractions, PDF1 expression (marker of the JA/ET pathway) decreased, which indicate that the metabolic elicitors present in these fraction were only activating the SA mediated transduction pathway, while the bacterial strain activated both. These results show that the elicitors detected by the plant in both cases have to be different, and so would be the PRRs involved in that response (Tang et al. 2017).
Regarding the Bacillus strain N 4.1, the two metabolic elicitor fractions ( Fig. 6c and d) did not match the bacterium except for PR3, a marker of the JA/ET pathway. These results suggest a lower diversity of effective metabolic elicitors, pointing out a more relevant role of structural elicitors triggering the SA mediated pathway observed with bacterium strain.
All these results show the great number of possibilities offered by elicitors to trigger the immune system of plants, which opens a plethora of biotechnological solutions to different stress situations.
Application of elicitors has many advantages from the agronomic point of view because it is more economical and profitable to conserve a molecule than an alive bacterium, which has nutritional and environmental requirements. In addition, the use of elicitors also implies less environmental aware for possible cases of ecological niches competition between edaphic species and also avoids problems of infectious pathogenesis and alterations of the rhizosphere (Timmusk et al. 2017, Rosier et al. 2018).

Conclusion
The enormous biotechnological potential of the rhizosphere as a source of bacterial strains capable of establishing a beneficial relationship with plants and of modifying their defensive metabolism, improving their ability to defend themselves from pathogen attacks, has been evidenced.
In addition, triggering SA and/or JA/ET defensive pathways by bacteria seem to be more complex than current description in the literature and the concept of simultaneous elicitation of different pathways of plant immune system has been reinforced.
Each bacterium had a different effect in the genes studied, even within the same bacterial genus. In addition, the metabolic elicitors of the two studied strains had different effects to that produced by the bacteria, confirming the presence of many different bacterial molecules able to trigger plant metabolism. This is very interesting since it opens a huge amount of biotechnological possibilities to develop biological products for agriculture in different situations and plant species.

Methods
A screening of 175 isolates was carried out. Firstly, biochemical tests for putative beneficial rhizobacteria traits were carried out to all isolates. The 16S rRNA partial sequencing of all isolates was analysed and a phylogenetic tree was performed with these sequences. Twenty-five strains selected based on their biochemical traits and avoiding phylogenetic redundancy were assayed to determine their ability to trigger plant protection (ISR). The most effective strains (5) were studied to understand the mechanisms involved in protection. Finally, metabolic elicitors (molecules released to the medium) were obtained from the two most effective to demonstrate their ability to mimic the protective response triggered by the strain.

Origin of bacteria
Bacteria used in this work were isolated from the rhizosphere of wild populations of Nicotiana glauca Graham in three different soils and physiological stages of the plant. A total of 960 isolates were obtained and 50% were tested for their putative beneficial rhizobacteria traits, as explained in the work of Ramos- Solano et al. (2010). In the present study, a subset of 175 strains from the nonassayed group of bacteria were used. These isolates and the pathogen P. syringae DC3000 were maintained in 20% glycerol, frozen at -80ºC and plated to check viability.

16S rRNA partial sequencing phylogenetic analysis
Bacteria were identified by 16S rRNA partial sequencing phylogenetic analysis. They were grown in PCA (Plate Count Agar (Conda)

Phylogenetic tree
An unrooted tree was performed with MEGA v4.0.2. with aligned sequences in MAFFT v6. The evolutionary distances were inferred using the neighbour-joining method. The bootstrap consensus tree inferred from 1000 replicates was taken to represent the evolutionary history of the taxa analysed. The percentage of replicate trees in which the associated taxa clustered together in more than 50% of the 1000 replicates of the bootstrap test are shown next to the branches. All positions containing gaps and missing data were eliminated from the data set (complete deletion option).

First ISR experiment. Screening for isolates able to induce systemic resistance
Based on phylogenetic analysis and putative beneficial rhizobacteria traits, twenty-five strains were selected for a first induced systemic resistance (ISR) assay. These bacteria (bioeffectors) were inoculated in A. thaliana plants at root level and challenged with the pathogen to evaluate their ability to protect plants.
Arabidopsis thaliana wild type Columbia ecotype 0 seeds (provided by the Nottingham Arabidopsis Stock Centre (NASC)) were germinated in quartz sand and two-week-old seedlings were then individually transplanted to 100 mL pots filled with 12:5 (vol/vol) peat/sand mixture (60 g/pot). Fortyeight plants per treatment (strains and controls) were used; plants were arranged in three replicates, with sixteen repetitions each. Plants were watered with 5 mL of tap water once a week and with 5 mL of half-strength Hoagland solution per plant once a week. Strains were inoculated twice by soil drench with 3 mL of a suspension of bacterial cells, grown for 24 h in nutrient broth (Conda) at 28 ºC, and adjusted to a density of 10 8 cfu mL -1 , in the first and the second week after transplant. Negative control plants were mock-inoculated by soil drench with 3 mL of sterile nutrient broth and positive control plants were inoculated by soil drench with 10 mL of BTH (Benzothiadiazole) 0.5 mM (Sumayo et al. 2013). Four days after the second bacterial inoculation, plants were pathogen challenged with P.
syringae DC3000. One day before pathogen challenge, plants were maintained with 99% relative humidity to ensure stomata opening in order to allow disease progress. P. syringae DC3000 was centrifuged (10 min at 4000 rpm) and cells were resuspended in 10 mM MgSO 4 to achieve 10 8 cfu mL -severity of leaves inoculated with P. syringae DC3000 (negative control) as 0% protection. All the ISR experimental design is represented as a timeline in Additional File 2.

Second ISR experiment. Study of the signal transduction pathway involved in plant protection
Based on results obtained in the first ISR experiment, the most protective strains (5) were selected to perform a second experiment to analyse the signal transduction pathways involved in plant protection triggered by bacteria. The expression of some marker genes after pathogen challenge were assessed by qPCR. Genes analysed were NPR1 (Nonexpressor of Pathogenesis Related Genes1), PR1 (Pathogenesis-Related Gene 1) and ICS (Isochorismate Synthase 1) as markers of the SA signalling pathway (Pieterse et al. 2014, Ding et al. 2018, Caarls et al. 2015, Kazan 2018, Vlot et al. 2009, Seyfferth and Tsuda 2014, Niu et al. 2011, Nie et al. 2017, Wildermuth et al. 2002, PDF1 (Plant Defensin 1), LOX2 (Lipoxygenase 2) and the transcriptional factor MYC2 as markers of the JA-ET signaling pathway (Caarls et al. 2015, Niu et al. 2011, Nie et al. 2017, Pangesti et al. 2014, Lorenzo and Solano 2005, Liu et al. 2016, Du et al. 2017, and two pathogenesis-related proteins genes, PR2 (encoding b-1,3-glucanase ) and PR3 (encoding chitinase), as SA and JA/ET markers, respectively (Wu et al. 2018, Jiang et al. 2016, Lemarié et al. 2015, Van Loon and Van Strien 1999, Spoel and Dong 2012, Jeandet et al. 2013, Schenk and Schikora 2013, Silva et al. 2018. thaliana was handled as described in the first ISR assay (See Additional File 2). Instead of recording disease severity 72 h after pathogen challenge (hapc), all the leaves of sixteen plants (treated with each bacteria (5)) were harvested at 6, 12 and 24 hapc, powdered in liquid nitrogen and stored at performed.
The retrotranscription was performed using iScript tm cDNA Synthesis Kit (Bio-Rad and 30 s at 72ºC, followed by melting curve to check results. To describe the expression obtained in the analysis, cycle threshold (Ct) was used. Standard curves were calculated for each gene, and the efficiency values ranged between 90 and 110%. Results for gene expression were expressed as differential expression by the 2 -ΔΔCt method. Sand gene (AT2G28390) was used as reference gen (Remans et al. 2008). Gene primers used are shown in Table 5. Table 5. Primers forward and reverse used in qPCR analysis.

experiment.
Based on data from qPCRs and protection from the first ISR experiment, two strains were chosen to isolate their metabolic elicitors and check their capacity to mimic bacterial protection: N 12.34 because it was the one with best differential expression results and N 4.1 because it was the Grampositive one with best protection against disease results.
Metabolic elicitors were extracted according to Sumayo et al. (2013) protocol until obtaining nhexane, ethyl acetate and n-butanol fractions. Briefly, strains were grown in nutrient broth (Conda) on a rotary shaker (180 rpm) at 28 ºC for 24 h. Cells were eliminated by centrifugation at 8,000 g for 15 min. Five hundred mL of the obtained supernatant was filtrated by 0,2 mm. This filtrate was used to extract metabolic elicitors. First, a double extraction 1:1 (v/v) with n-hexane was made. The remaining aqueous phase was extracted twice with ethyl acetate (1:1 v/v), and finally, the aqueous phase was extracted twice with n-butanol (1:1, v/v). The organic phases (n-hexane, ethyl acetate and n-butanol) were pooled and evaporated to dryness in a rotary evaporator at 50 ºC. The dry residues obtained were dissolved in 25 mL 10 % Dimethyl sulfoxide (DMSO).
A third ISR assay on A. thaliana plants to evaluate the ability of three metabolic elicitor fractions from N 12.34 and N 4.1 was carried out. Four treatments per strain were defined: a) metabolic elicitors in the n-hexane fraction, b) metabolic elicitors in the ethyl acetate fraction, c) metabolic elicitors in the n-butanol fraction, and e) positive control (BTH (Sumayo et al. 2013)). An additional control (negative control) with DSMO was included to ensure that elicitor effects were due to bacterial components and not to the chemical. All were pathogen challenged.
thaliana was handled as described in the first ISR assay (See Additional File 2). Treatments were delivered to seedlings by soil drench (50 mL). Negative control was treated with 50 mL of DMSO. The pathogen was also inoculated as described in the first ISR assay. Seventy-two hours after pathogen inoculation, disease severity was recorded and relativized as in the first ISR experiment.

RT-qPCR analysis of the genes triggered by metabolic elicitor fractions (fourth ISR experiment)
Based on data from the third ISR experiment, another ISR assay was carried out using the protocol explained above. and controls with n-hexane and ethyl acetate. Sterile nutrient broth was used to obtain control nhexane and control ethyl-acetate fractions. Plants were inoculated by soil drench (50 mL) and challenge inoculation with DC3000 was performed as explained above.

Statistical analysis
One-way ANOVA with replicates was used to check the statistical differences in all data obtained. Prior to ANOVA analysis, homoscedasticity and normality of the variance was checked with Statgraphics plus 5.1 for Windows, meeting requirements for analysis. When significant differences appeared (P < 0.05) a Fisher test was used (Sokal and Rohlf 1980

Declarations
Ethics approval and consent to participate: "Not applicable" Consent for publication: "Not applicable" Availability of data and materials: The data that support the findings of this study are available from the corresponding author upon reasonable request.
Competing interests: The author(s) declare no competing interests.
Funding came from state research funds, limited to freeing up the money without intervening in the study design, collection, analysis, and interpretation of data, or manuscript writing.  Differential gene expression (seedlings inoculated with N 4.1 (Bacillus cereus) vs negative control) 6 (n=16), 12 (n=16) and 24 (n=16) h after pathogen challenge; a) NPR1, ICS, PR1