Uncovering salt tolerance mechanisms in pepper plants: a physiological and transcriptomic approach

Pepper is one of the most cultivated crops worldwide, but is sensitive to salinity. This sensitivity is dependent on varieties and our knowledge about how they can face such stress is limited, mainly according to a molecular point of view. This is the main reason why we decided to develop this transcriptomic analysis. Tolerant and sensitive accessions, respectively called A25 and A6, were grown for 14 days under control conditions and irrigated with 70 mM of NaCl. Biomass, different physiological parameters and differentially expressed genes were analysed to give response to differential salinity mechanisms between both accessions. The genetic changes found between the accessions under both control and stress conditions could explain the physiological behaviour in A25 by the decrease of osmotic potential that could be due mainly to an increase in potassium and proline accumulation, improved growth (e.g. expansins), more efficient starch accumulation (e.g. BAM1), ion homeostasis (e.g. CBL9, HAI3, BASS1), photosynthetic protection (e.g. FIB1A, TIL, JAR1) and antioxidant activity (e.g. PSDS3, SnRK2.10). In addition, misregulation of ABA signalling (e.g. HAB1, ERD4, HAI3) and other stress signalling genes (e.g. JAR1) would appear crucial to explain the different sensitivity to NaCl in both accessions. After analysing the physiological behaviour and transcriptomic results, we have concluded that A25 accession utilizes different strategies to cope better salt stress, being ABA-signalling a pivotal point of regulation. However, other strategies, such as the decrease in osmotic potential to preserve water status in leaves seem to be important to explain the defence response to salinity in pepper A25 plants.


Background
Pepper (Capsicum annuum L.) is one of the most important cultivated horticultural species worldwide. Production has increased over the last 20 years from 17 to 36 million tons, and the cultivated area has expanded by about 35% [1]. However, several stresses still significantly affect peppers, which decrease yields and fruit quality. The most important stress is biotic, but peppers are also affected by some abiotic stresses [2]. One of the most relevant ones is salt stress as pepper plants are considered moderately sensitive, sensitive or highly susceptible [3,4]. The source of a high salt concentration that affects plants may be either soil or irrigation water [5]. In pepper plants, dry weight and marketable yield diminished by 46 and 25%, respectively when is irrigated with water at 4.4 dS m − 1 [6].
The root is the first organ affected after the exposure to high Na + and Cl − concentration, since the excess of these ions generates osmotic and ionic stress [4,7]. These ions also move rapidly to photosynthetic organs and cause several negative effects. Indeed salt accumulation in plant tissues provokes changes in the physiological metabolism, such as nutritional imbalances, and generates reactive oxygen species (ROS), among other physiological disorders that lead to reduce biomass and crop production [8,9]. However, some species are able to deal with these negative effects and can be tolerant to salt stress. To reach this condition, a complex network of genes related to salt tolerance is necessary [10], that can modify physiological and biochemical plant responses.
In agricultural species, growers have always tended to select genotypes with increased commercial production, commonly linked to improved tolerance to specific stresses. As a result, it is now possible to find a wide diversity of accessions that differs in terms of grades of tolerance to stresses. In the case of pepper, several authors have demonstrated that the severity of negative effects depends on the variety [11][12][13][14].
This intraspecies variation may be a source of information to find factors like genes, proteins or metabolites related to tolerance, which can be used in, for example, conventional breeding programmes or genetic engineering technologies [10], or to be employed as tolerant rootstocks in grafted plants [15,16].
Several transcriptomic studies revealed the understanding of the genetic mechanisms responsible for the tolerance of pepper plants to various stresses, such as heat stress, chilling or leaf curl virus [17][18][19][20]. In the case of salt stress, specific genetic pathways of tolerance has been addressed [21][22][23] but scarce information is available related to pepper plants.
Consequently, this study compared two pepper accessions previously classified by us as tolerant (A25) and sensitive (A6) to salt stress after analysing a series of physiological and agronomical parameters [12,24]. This study included a series of measurements to evaluate different physiological traits, as well as a transcriptomic analysis, by microarrays, to elucidate the genetic programmes that were expressed and are responsible for tolerance to salt stress. This analysis could reveal the underlying mechanisms in pepper to cope with salinity stress and open up new strategies to improve crop performance under salinity conditions.

Biomass
In order to evaluate whether plants maintained the same growth rate after NaCl treatment, dry biomass was measured in both roots and aerial organs at 14 days after treatment (14DAT). Under the control conditions, both accessions obtained higher values compared to the salt stress conditions (Fig. 1a, b). Nevertheless, growth under the control conditions differed between accessions as A25 obtained higher values in both aerial and root biomass ( Fig. 1a, b). The tolerant A25 accession better maintained both aerial and root dry weight under salt stress conditions compared to A6 accession at the end of the experiment (14DAT).

Gas exchange measurements
As photosynthesis is one of the first processes affected after salt stress exposition, so it is crucial to evaluate its parameters and how they progress with time exposure. In this experiment, CO 2 assimilation rate (A N ), stomatal conductance to water vapour (gs), substomatal CO 2 concentration (Ci) and transpiration rate (E) were analysed at 7DAT and 14DAT (Fig. 2). At 7DAT, A25 showed no significant differences in A N and Ci (Fig. 2a, c) between the control and salt stress conditions. Conversely, gs and E decreased in the stressed plants (Fig. 2b, d) but, compared to A6, these parameters in A25 were better maintained as A6 obtained the lowest values of them all. conditions. Measurements were taken at the end of the experiment (14DAT). Data are the mean of 6 replicates and the error bars belong to the standard deviation. Different letters indicate significant differences at P < 0.05 (LSD test) At 14DAT under the control conditions, a better response was observed in stomatal conductance and transpiration with the tolerant accession A25 (Fig. 2b, d), unlike A N and Ci, which remained unchanged (Fig. 2a, c). Different results were found under salt stress conditions, with no significant differences between the two studied accessions for all the gas exchange parameters studied, that scored similar values.

Ion determination
Exposure to high NaCl concentrations disrupts ion homeostasis in plant cells [8]. Thus, the evaluation of the ion concentration in different tissues after exposure to stress was crucial for this experiment. For this purpose, Na + , K + and Cl − concentrations were measured at the end of the experiment (14DAT) in leaves (Fig. 3a, c, e) and roots (Fig. 3b, d, f). Regarding Na + (Fig. 3a, b), the concentration in both leaves and roots increased for the two accessions studied when were subjected to salt stress. It is worth mentioning that the levels in roots were higher than leaves for both accessions and treatments, especially in A25, which showed more Na + accumulation compared to A6.
Regarding the K + concentration (Fig. 3c, d), under high salinity treatment the leakage was higher in leaves compared to roots. Nonetheless, the potassium level remained constant in the roots of the A25 accession in both conditions. In addition, K + levels were higher in A25 under salt stress in both leaves and roots compared to A6.
Besides, a higher Cl − concentration was detected under the salt conditions in all the studied organs and accessions compared to the control (Fig. 3e, f). However, the concentration in the A25 accession under salt stress rose in roots and lowered in leaves compared to A6. Under the control conditions, no significant differences were found in any of the studied organs.

Physiological determinations
The evaluation of the osmotic potential (ψ S ), the starch content as well as the phenols, proline and H 2 O 2 concentration can additionally inform about the capacity of plants to tolerate or not salt stress conditions (Fig. 4).
The ψ S evaluated in the leaves at 14DAT (Fig. 4a) showed that, unlike control conditions, salt stress conditions displayed significant differences between the After starch content analysis we could that the accessions present very different behaviours in both control and stress conditions (Fig. 4b). Concretely, starch content decreased in A6, at 14DAT, in the case of salt stress conditions, lowering to the values of control conditions of A25. On the contrary, A25 had significant higher values in the case of salt stress when compared to its control or A6 under salt stress conditions.
In the case of the total phenol concentration in the leaves (Fig. 4c), we did not find significant differences in the case of control conditions among both accessions at the end of the experiment. However, after salt stress, significant differences were observed, reaching the highest values in the case of A25.
Concerning H 2 O 2 determination (Fig. 4d), differential accumulation in leaves was detected among the tolerant and the sensitive accessions at 14DAT. In the case of control conditions, A6 reached the highest accumulation, decreasing when salt stress was present. On the contrary, A25 significantly increased H 2 O 2 levels if compared to its control conditions or A6 accession under salt stress.
Finally, proline content and relative expression of its putative gene CaP5CS (CA06g06110) in leaves have been determined in the end of the experiment (Fig. 4e-f). In the case of proline content (Fig. 4e), A6 reported the lowest values, without significant differences between control and salt stress conditions; on the contrary, A25 reported more accumulation under salt stress than control conditions or the A6 accession. Similar results were found in the case of the relative expression of the gene CaP5CS (Fig. 4f), whose expression in A25 under salt stress conditions significantly increased 2-fold respect to its control or A6 under salt stress; nevertheless, we have detected as well an increase of CaP5CS expression in the case of A6 under salt stress conditions compared to its control.

Transcriptomic expression results
A microarray experiment analysis was performed to know the transcriptomic changes that could explain the sodium chloride tolerance of the A25 accession.
Under the control conditions when the A25 accession was compared to A6, 196 and 315 genes were up-and down-regulated, respectively (Fig. 5a, b), of which 95 upand 107 down-regulated genes were commonly expressed in control and salt stress conditions. Of all these genes, it is important to highlight the up-regulated genes related to cell wall biosynthesis and expansion (PMEI13, TUB8, EXPA13, XK-1, PME1, CEL5, CSLE1), wax and fatty acid biosynthesis (KASI, LACS2), cell division (CDC2), vitamin transport (BASS1), ABAsignalling (SnRK2.10, TINY2 and ERD4) and photosynthesis (PSBP-1) in A25 vs A6. The genes related to the formation of cellular barriers, such as lignins (PRX71, PRX66) and waxes (WS6D, CER1), were down-regulated in A25 compared to A6. The down-regulation of the  (70 mM) conditions. Measurements were taken at the end of the experiment (14DAT). Data are the mean of 4 replicates, except for proline content, which is the mean of 6 replicates. The error bars belong to the standard deviation. Different letters indicate significant differences at P < 0.05 (LSD test) genes involved in stress protection (CAMTA5, JAR1, CBL9) and photosynthesis (NDHG, CCB3) (Tables 1, 2; Additional file 1) was noteworthy. In accordance with these findings, functional GO analysis displayed related results (Additional file 2, Fig. S1).
We could check the abundance of genes related to "response to stress", "response to abiotic or biotic stimulus", "transport" and "signal transduction". Regarding to cellular components we highlight the significative number of genes related to "chloroplast", "cell wall" or "plastid". KEGG pathways (Additional file 1, Table S1-2), showed significative results in this set of genes. "Biosynthesis of secondary metabolites", "Cutin, suberine and wax biosynthesis", "Fructose and mannose metabolism" and "carbon fixation in photosynthetic organisms" were enriched pathways in downregulated genes.
For the A6 accession, 33 and 25 specific up-and downregulated genes were respectively found after salt stress ( Fig. 5c, d), in which the genes related to cell expansion (EXPA4), photosynthesis (PSBP-1, TROL, PSAE-2) and starch degradation (BAM5) were down-regulated (Table 4; Additional file 1). GO enriched categories of DEGs in A6 after salt stress (Additional file 2, Fig. S4) were similar to A25 in the same conditions but the number of genes were significatively lower. On the other hand, we highlight "Photosynthesis" as a KEGG enriched pathway in specific downregulated DEGs in the comparison A6 in control conditions and A6 after salt stress.

Discussion
In this work, we analysed the gene expression of two pepper accessions under control and salt stress conditions, called A25 and A6, previously classified as tolerant and sensitive to salt stress conditions from an agronomical and physiological point of view [12,24]. In the present study, A25 accession exhibited under control conditions activation of genes related to cell growth and division, as well as cell wall expansion. This evidence, together with the inactivation of starch degradation and defence pathways, provides to A25 an advantage respect to A6 since we observed enhanced biomass. Indeed, several authors have already visualized differences in accessions or varieties of Arabidopsis thaliana, pepper or tomato [13,24,25] under control conditions that may influence the grade of tolerance to salt stress.
On the other side, we detected different strategies to face salt stress in the two studied accessions, what conferred at 14DAT contrasting grades of tolerance. In the several sections below, we have explained the main processes affected by this complex gene regulation network in response to salinity stress.

Hormonal Signalling
Hormone signalling and biosynthesis have been considered an essential point of the regulation of plant tolerance or susceptibility to stress [25]. Accordingly, our results uncover several genes involved in jasmonates (JAs) and abscisic acid (ABA) synthesis, degradation or signalling that could explain the behaviour of the analysed accessions.
Jasmonates are key elements in the regulation of a wide range of processes when different abiotic stresses are present [26][27][28]. However, they need to be conjugated with a series of compounds to be active [29]. We found that gene jasmonate resistant 1 (JAR1), responsible of the creation of an active jasmonyl-isoleucine (JA-Ile) conjugate, was up-regulated when salt stress and control were compared in the A25 accession, but it was absent in A6. Several authors have demonstrated by external applications that JAs improve the activity of different antioxidant enzymes, growth and development, photosynthetic activity and Na + homeostasis [26,30,31].
ABA is a well-known hormone that plays a central role in tolerance to different abiotic stresses as it performs a Table 1 Summary of the specific differentially expressed genes after 14DAT in the comparison A25/A6 in plants subjected to control conditions. It is represented both the up (FC > 1) and down-regulated genes (FC < 1), as well as the fold change (FC) and the adjusted P-value obtained for each gene (significant differences were considered when P < 0.05). Genes without abbreviation are represented with "-"  Table 2 Summary of the common differentially expressed genes after 14DAT in the comparison A25/A6 in plants subjected to control and salt stress conditions. It is represented both the up (FC > 1) and down-regulated genes (FC < 1), as well as the fold change (FC) and the adjusted P-value obtained for each gene (significant differences were considered when P < 0.05). Genes without abbreviation are represented with "-"  Table 3 Summary of the common differentially expressed genes after 14DAT in the comparison NaCl/Control in the accession A25. It is represented both the up (FC > 1) and down-regulated genes (FC < 1), as well as the fold change (FC) and the adjusted P-value obtained for each gene (significant differences were considered when P < 0.05) wide variety of functions in plant growth and development, it regulates plant water balance by stomata opening, and it plays a crucial role in osmotic stress tolerance [32]. Increasing ABA concentration and signalling are wide responses of the tolerance described by several authors, which favours stomata closure and, thus, avoids excess transpiration. However, this fact also compromises plant growth as it diminishes photosynthetic activity [25,33]. In our experiment, we found several DEGs in A25 described as regulators of ABA, or are regulated by ABA signalling (HAB1, ERD4, CAMTA5, Tiny2, CBL9, Snrk2.5, Snrk2.10, HAI3) and, thus, play a central role in controlling tolerance. Of all the ABA-related DEGs found in A25, one of the most relevant was the up-regulated gene hypersensitive to ABA1 (HAB1). HAB1 encodes a functional type 2C protein phosphatase (PP2C) and has been reported as a positive or negative regulator of ABA signalling, depending on the splice variant [34,35]. Overexpression of this gene has been reported, in fact, that leads to a minor or major ABA sensitivity, modifying stomata opening and gene expression [35,36].
A family of transcription factors, which has been reported to be regulated by ABA and plays an important role in stress tolerance, is the Calmodulin-binding transcription activators family (CAMTA) [37]. It has been demonstrated that the CAMTA family can bind to the promoters of different members of the dehydrationresponsive-element-binding (DREB) transcription factors family and modulate the stress response [38]. In our case, we found the up-regulation of CAMTA5 genes and DREB member TINY2 in the A25 accession, which may indicate that both genes enhanced the response to salt stress by improving growth, development, the expression of stress-responsive genes or ABA-mediated stomatal closure [39][40][41][42].

Biomass and cell growth
Salt stress negatively affects cell growth and plant biomass. However, greater biomass conservation is considered a sign of tolerance [43][44][45][46]. In this study, at 14DAT Table 4 Summary of the common differentially expressed genes after 14DAT in the comparison NaCl/Control in the accession A6. It is represented both the up (FC > 1) and down-regulated genes (FC < 1), as well as the fold change (FC) and the adjusted P-value obtained for each gene (significant differences were considered when P < 0.05)  Table 5 Summary of the specific differentially expressed genes after 14DAT in the comparison A25/A6 in plants subjected to salt stress conditions. It is represented both the up (FC > 1) and down-regulated genes (FC < 1), as well as the fold change (FC) and the adjusted P-value obtained for each gene (significant differences were considered when P < 0.05) a better maintenance of root and aerial biomass was found in A25 compared to the A6 accession under the salt stress conditions. Biomass preservation is usually associated with the differential expression of a wide variety of genes related to cell growth and division, some of which were identified in this experiment. One of these genes is an ABA-related gene called ERD4 (early-responsive to dehydration 4), which was up-regulated in the A25/A6 comparison under both the control and salt stress conditions. This gene has been described in the bibliography as being overexpressed in tolerant transgenic A. thaliana plants when salt is added [47]. One of the keys to improve plant growth is defined by the ability of plants to maintain water status. After salt addition, plants undergo a reduction in the content of water in cells; in order to avoid it, plants set different mechanisms. Among them, the accumulation of a wide range of compatible osmolytes, such as proline, is crucial to help lower the Ψ s [48]. Herein we demonstrated that the tolerant accession maintained at lower levels the osmotic potential and accumulated more proline content in leaves under salt stress, what can be related to the preservation of water status of plants, as has been previously reported [49]. Indeed, we detected as well a positive correlation among the up-regulation of the gene CaP5CS and the proline content (R 2 = 0.85), what has been described as a signal of tolerance [4]. Despite these results, we do not discard the possibility that other compounds described in bibliography, such as sugars, glycinebetaine or non-compatible osmolytes such as ions were all participating to reduce the osmotic potential [48].
We also detected up-regulation of a series of genes in the A25 accession related to cell division and expansion. One of these genes was the cell division control 2 (CDC2), which regulates the G1/S and G2/M transitions in mitosis [50]. It has been demonstrated that abiotic stresses, such as drought, can negatively affect CDC2 activity [51]. As the expression in the A25 accession was 6.91-fold higher in salt stress, cell division rhythm improved. Additionally, we found an increased expression of the expansin genes in the A25 accession under salt stress compared to the control or the A6 accession, responsible for the nonenzymatically loosening and extension of plant cell walls [52]. This finding suggests that A25 improved cell wall expansion and turgor, which may lead to better growth and development, as other authors have already demonstrated [53,54].

Starch degradation
Abiotic stresses may also affect starch accumulation and degradation, as it may be remobilised to release energy, sugars, carbon and derived metabolites when photosynthesis is limited [55]. In general terms, under salt stress conditions a decrease of the starch content has been described, although an improved accumulation has been observed in tolerant plants [56], as we have noticed in this experiment in A25. However, a better starch degradation into soluble sugars has been also linked with tolerance to stress, since they may interact with hormones, genes and proteins, regulating diverse pathways as well as growth and development [57]. In line with this, βamylase 1 (BAM1) was found to be up-regulated only under the salt stress conditions in the A25 accession, what would respond to transitory starch degradation in guard and mesophyll cells of mature leaves, as other authors has already observed under osmotic and salt stress conditions [58,59].

Ion homeostasis
When plants come into contact with salt, it is crucial to maintain ion homeostasis to avoid toxic accumulation. Plants cope with this situation by different mechanisms that can contribute to salt tolerance, some of which are very well documented in the bibliography [9,60].
One of the most important and abundant cations in plants cells is K + , which decreases under salt stress conditions because of replacement with Na + . The enhanced K + homeostasis in the A25 accession in both organs indicated that K + played an pivotal role to contribute to the salt stress tolerance, as other authors have already demonstrated [9,16]. In this line, it has been previously described that this cation, together with other compatible osmolytes, can contribute to the decrease in osmotic potential of plants [61,62], what could indicate similar functions herein since the accumulation was more evident in A25 accession. In this experiment, additionally, we have detected that A25 may possess some mechanism to keep K + inside cells by the evaluation of the DEGs; one possible candidate that could explain it is AKT1, a passive transporter that specifically introduces K + into root and mesophyll cells [63,64]. Thus we detected the up-regulation of the negative regulator of ABA signalling CBL9 (calcineurin B-like protein 9) and the down-regulation of positive regulator HAI3 (Highly ABA-Induced 3) in the A25 accession under salt stress conditions [65]. These genes play opposite roles in the regulation of AKT1 as CBL9 is a positive regulator [66], and HAI3 could be a repressor as this gene presents a high homology to HAI2 [67,68].
The accumulation of Cl − ions and especially Na + in pepper plant tissues, performs diverse physiological functions [15,16]. When Na + reaches toxic levels, plants may decrease the influx into cells and improve efflux and compartmentalisation in other organelles where ions are not toxic [7]. In our experiment, we found that Na + was accumulated in the roots of both accessions after 14DAT. As this accumulation was especially pronounced in A25, and root biomass had improved compared to the A6 seedlings, this effect could be associated with compartmentalisation in vacuoles or other organelles, as other authors have already demonstrated [4,69]. Despite the negative effect on plant growth consequence of its toxic effect, accumulation of ions under salinity can help to maintain the turgor pressure of plants [15,70]. The adjustment of the osmotic potential through inorganic ion uptake implies a much lower energy cost than that conferred by the organic molecules synthesised in cells [71].
In leaves, Na + was equally accumulated in both accessions, but biomass improved only in A25, a cue that Na + management was diverse in both pepper accessions. In line with this, we noticed that ion transport in leaves was closely linked to the protection of chloroplasts in A25 as we found some related genes. One of these genes was BASS1 (bile acid/sodium symporter 1), which was up-regulated in the control and salt stress treatments in the A25/A6 comparison. This gene, which encodes a symporter of Na + and pantoate, a precursor of Vitamin B5, could play a double role in A25: on the one hand, it conferred protection from Na + toxicity in chloroplasts to conserve photosynthesis responses; on the other hand, the pantothenate cycle was promoted [72,73]. We also found the gene TIL (temperature-induced lipocalin), which was up-regulated in A25 compared to A6 under salinity stress, which can avoid excess Na + and Cl − accumulation in chloroplasts by protecting chlorophyll b degradation in this way [74].
An up-regulated gene found in the A25/A6 comparison under the salt stress conditions was OCT4 (organic cation/carnitine transporter 4), which decreases the concentration of toxic Na + in the cytoplasm by accumulating in vacuoles. This family of genes is responsible for the symport of Na + and organic molecules like carnitine [75,76]. So it would play an important role in osmotic balance through ionic homeostasis in our tolerant accession.

Photoprotection
When plants come into contact with salt stress, one of the primary affected processes is photosynthesis. The photosynthetic parameters herein analysed reflected that only A25 maintained them at 7DAT compared to the control conditions, although both accessions were equally affected at the end of the experiment. These reasons suggest that A25 kept the plant's photosynthetic capacity levels high for longer times [77]. In addition, some genes involved in the protection of photosynthesis were differentially expressed in both accessions. In line with this, we found the up-regulation of ABA-related gene fibrillin 1A (FIB1A) in the A25 NaCl/control, which suggests that fibrillin was accumulated in chloroplasts and, consequently, could improve protection and efficiency of PSII [78]. Together with FIB1A, other previously explained genes contributed to photoprotection, such as TIL, BASS1 or JAR1.

ROS scavenging
When photosynthesis is disturbed by salt stress, a series of secondary effects is detected, such as oxidative stress, which may lead reactive oxygen species (ROS) to toxic levels [79]. The ability to reduce the quantity of all these molecules by efficient ROS-scavenging mechanisms is vital for acquiring tolerance. In our specific case, we detected that improving the accumulation of phenolic compounds was enhanced in the A25 accession under salt stress, what has been widely described to improve the antioxidant capacity of plants [80]. Additionally, the accumulation of H 2 O 2 in plants, as in the case of A25, has been identified by several authors as a signal of ROS damage. Nonetheless, it has been proposed in the last decades to play also a role as a secondary messenger to activate plant antioxidant processes related with abiotic stress acclimation and consequently mediate adaptative responses to abiotic stress [16], so similar functions are proposed. Accumulation of other molecules in A25 tolerant accession, such as proline, could also be implied in ROS detoxification and salt protection, as other authors have already described [81]. Regarding gene expression, we have found up-regulation of the gene SPDS3 (spermidine synthase 3) in A25 under salt stress conditions, which catalyses the formation of spermidine, a polyamine that improves multiple processes in plants, such as ROS scavenging, the K + /Na + ratio and PSII efficiency by protecting thylakoid membranes and chlorophyll content [82][83][84]. We also found another upregulated gene in A25 under salt stress conditions, called sucrose non-fermenting 1-related protein kinase 2-10 (SnRK2.10), which regulates the gene expression, protein level and/or enzymatic activity of several ROS-related enzymes, and is also involved in H 2 O 2 accumulation and ascorbate cycle regulation in A. thaliana [85].

Conclusions
After analysing the physiological parameters and DEGs of both accessions, we conclude that different tolerance strategies simultaneously took place in the A25 tolerant accession after exposure to salt stress, with ABAsignalling being a pivotal point of regulation, and an important network was established between different genes and physiological traits to reveal the complex response induced by salinity. These results provide valuable results about salt stress mechanisms of an important crop like pepper. It is noteworthy that we also found several genes that probably contributed to tolerance, but their functions have not yet been discovered.

Plant material
Based on previous studies [12,24], two accessions of C. annuum were selected depending on their grade of tolerance to salt stress: code A6 (Pasilla bajio, Mexico) was sensitive and code A25 (Numex big Jim, Nuevo Mexico) was tolerant. All the accessions used for the present study belong to the germplasm bank placed in the Institute for Conservation and Improvement of Valencian Agrodiversity "COMAV" (Universitat Politècnica de València, Valencia, Spain). Maria José Diez, director of COMAV, verify A6 and A25 with the deposition numbers BGV013994 and BGV014452 respectively.
Seeds were sown in 104-hole seed trays filled with enriched substrate for germination. When plants had 6-8 real leaves, they were placed in 5-l polyethylene pots covered with aluminium sheet (roots were previously cleaned of substrate). Pots were filled with a nutrient solution containing (in mmol L − 1 ) 12 , which was artificially aerated. The electrical conductivity (EC) and pH of this nutrient solution were 1.7 dS m − 1 and 6.5, respectively. The nutrient solution was added daily to compensate for absorption. After 14 days of plant acclimation, salt stress was induced by the addition of NaCl 70 mM by replacing the plant pot solution to obtain an EC of 8.5 dS m − 1 and a pH of 6.1. The layout design was completely randomised with 10 plants per accession and treatment.
During the culture and experiment, plants were grown in a greenhouse at the Polytechnic University of Valencia (UPV, Valencia, Spain) under natural light conditions (800-1000 μmol m − 2 s − 1 ), with a temperature range of 18-25°C and 50-70% relative humidity (RH).
All the parameters were measured 14 days after stress induction, except in the photosynthetic parameters, where measurements were taken after 7 days (7DAT) and 14 days (14DAT) of treatment.

Biomass determination
Six replications per accession and treatment were harvested at 14DAT for the biomass parameters. Aerial organs and roots were separated and weighed (FW). Immediately afterwards, they were dried by placing them in an oven at 65°C for 72 h. After this time, everything was weighed again to determine dry weight (DW).

Ion determination
Six replications per accession and treatment of leaves and roots were collected and dried at 65°C for 72 h at the end of the experiment (14DAT). Dried samples were ground with a mortar and used for the ionic analysis.
With Na + and K + , samples (0.2 g for leaves, 0.1 g for roots) were incinerated in a muffle furnace for 12 h at 550°C. Ions were extracted with 2% nitric acid in an ultrasonic bath for 30 min at 40°C. The concentration of such ions was determined by an ICP emission spectrometry (iCAP 6000, Thermo Scientific. Cambridge, England, UK).
Regarding chloride concentration (Cl − ), dry plant material (0.125 g of roots and leaves) was extracted with 0.1 N HNO 3 in 10% (v/v) acetic acid and was quantified by potentiometric titration with AgNO 3 in a chloride analyser (Sherwood, MKII 926).

Osmotic potential
Four replications per accession and treatment were analysed. Firstly, leaves samples were frozen in liquid nitrogen and stored at − 80°C. After that, samples were introduced in a 1.5 mL tube and centrifuged for 10 min at 9000 x g. Leaf sap was measured with an osmometer (Digital osmometer, Wescor, Logan, UT, United States). Osmolyte content (mmol kg − 1 ) was converted into MPa by the Van't Hoff equation.

Starch content
Starch determination was analysed according to [86] with modifications, using for that four replications per genotype and treatment. Fresh samples were frozen and ground in liquid nitrogen and stored at − 80°C. Samples (0.3 g of leaves) were mixed with heated ethanol 80% (v/ v) and boiled at 85°C for 10 min three times. After that, samples were centrifuged at 10,000 x g. The precipitated was resuspended in perchloric acid 35% (v/v) and left for 24 h at room temperature. After that, samples were diluted with water and then filtered. The filtrate was then mixed with anthrone acid solution and placed in boiling water for 7.5 min. After cooling the samples, absorbance was measured at 630 nm. D-glucose was used as standard.

Total phenol content
Total phenol content was measured according to [87] with modifications. Four replications per genotype and treatment were frozen, ground in liquid nitrogen and stored at − 80°C. Samples (0.1 g of leaves) were mixed with 1.5 mL of 80% methanol (v/v) and extracted in an ultrasound bath at room temperature for 30 min. After that, leaf extract was diluted in extraction solution (dilution 1:4). When samples were diluted, 0.7 mL of Folin-Ciocalteu solution (Sigma-Aldrich®; 1:10 dilution), and 0.7 mL of 6% (w/v) Na 2 CO 3 were added to samples and were incubated at room temperature in dark conditions for 1 h before measuring the absorbance at 765 nm. Standard curve was determined by the content of gallic acid.

Hydrogen peroxide content
The H 2 O 2 content was quantified followwing [88,89], with modifications. Four replications per genotype and treatment were frozen and ground in liquid nitrogen and then conserved at − 80°C. Samples (0.25 g of leaves) were mixed with 2 mL of 0.1% (w/v) trichloroacetic acid (TCA) and centrifuged at 10,000 x g at 4°C for 8 min. A volume of 0.4 mL of sample was diluted with 0.6 mL of 0.1% (w/v) TCA. Afterwards, 0.5 mL of 100 mM potassium phosphate buffer (pH = 7) and 2 mL of 1 M KI were added and incubated for 1 h at room temperature under dark conditions. Absorbance was measured at 390 nm. Standard curve was determined by known concentrations of H 2 O 2 .

Proline determination and CaP5CS gene expression
Proline content was determined according to [90]. Six replications per genotype and treatment were used for its determinations. Leaves were dried at 65°C for 72 h and then ground with a mortar. Samples (0.02 g) were mixed with 3% sulfosalicylic acid and then the homogenate was centrifuged at 12,000 x g for 5 min. Afterwards, glacial acetic acid and ninhydrin reagent were added to the samples. and boiled at 100°C for 1 h. After cooling the samples, absorbance was measured at 520 nm. Proline content was calculated using a known standard curve.
Additionally, the putative gene Delta-1-pyrroline-5carboxylate synthase of pepper (CaP5CS, CA06g06110), implied in the synthesis of proline, was measured in leaves at 14DAT following the methodology described below in the section "Validation of Microarrays Analysis by Real-Time Quantitative PCR (RT-qPCR)". The primers used for its determination were (5′-3′): TTTA GTGATGGGTTCCGCTTTG (Forward) and CAATCC CTCGAVCTCCAACTC (Reverse).
Extraction and quality measurement of Total RNA Three replications of leaves per treatment and accession were frozen in liquid nitrogen immediately after harvest and conserved at − 80°C at 14DAT. At the time of RNA extraction, samples were ground to a fine powder with a mortar and liquid nitrogen. Total RNA was extracted using the MACHEREY-NAGEL NucleoSpin® RNA kit. Approximately 0.1 g was weighed, and RNA was obtained following the protocol "RNA purification from cultured cells and tissue" by the producer; DNase treatment was used to remove DNA from samples and was acquired from the same producer. Total RNA was eluted in 50 μL of RNAse-free water and was immediately aliquoted and conserved at − 80°C. The total RNA samples with 260/280 and 260/230 ratios > 2 (measured by a NanoDrop ND1000) and RNA integrity (RIN) value > 7.0 (measured by the Agilent 5067-1511 Bioanalyzer 2100 System) were used for microarray hybridisation.

Microarray hybridisation
The RNA extracted from the leaf samples was prepared for microarray hybridisation at the Genomic Service of the IBMCP Institute (Instituto de Biología Molecular y Celular de Plantas) in Valencia (Spain) by Agilent technologies. cDNA synthesis and labelling on Agilent Tomato microarrays were carried out using the Agilent One Colour RNA Spike-in Kit and the Agilent Low Input Quick Amp Labeling Kit. Microarray hybridisation and washing were next performed with the Agilent Gene Expression Hybridization kit and Gene Expression Wash Buffers. Agilent microarray 4*44 k (Agilent G2519F) was selected for hybridisation (reference AMADID 22270 Tomato). Microarray scanning was done with a GenePix 4000B (Axon Molecular Devices, Sunnyvale, USA) and data were extracted by the Agilent Feature Extraction software, version 9.5.1.

Microarray data analysis
The obtained spot intensity values were analysed on the Babelomics 5 platform [91]. Firstly, raw data were normalised, which consisted in a background correction, rescaling all the microarrays to a unique final distribution and reshaping data to a suitable distribution. At this point, data were transformed from tomato probes to pepper and Arabidopsis thaliana genes by the Bioinformatics service at the IBMCP Institute in Valencia (Spain) to then take the average among all the probes of the same pepper gene. Raw data were then separated into categories (accession and treatment) and analysed by a class comparison test. All the differentially expressed genes (DEGs) of the class comparison, both up-and down-regulated, were described as their orthologue of A. thaliana by the database of Araport 11. Using the resultant DEGs, GO classification has been developed using the platform Bio-Analytic Resource for Plant Biology (http://bar.utoronto.ca/#) and classified in "Biological Process", "Molecular Function" and "Cellular Component". Additionally, DEGs were also subjected to a KEGG enrichment analysis, performed with DAVID Functional Annotation Tool [92].

Validation of microarrays analysis by real-time quantitative PCR (RT-qPCR)
The RNA of leaves used for microarray analyses was used as well for validation of the obtained results from the class comparisons. For that purpose, firstly the RNA samples were retrotranscribed by the use of PrimeScript RT reagent kit (Takara Bio) in a total volume of 20 μL. After that, all the samples were fifty-fold diluted to perform RT-qPCR. A volume of 2 μL of diluted sample was used per well (total volume of 20 μL) and SYBR premix Ex Taq (Tli RNseH plus, Takara Bio) was used to conduct the reaction. Three technical replicates were evaluated per every biological replicate. Additionally, a relative standard curve was designed to obtain the mean relative expression to get the final results. RT-qPCR was carried out on a StepOnePlus Real-Time PCR System (Life Technologies), and the sequence of reactions was firstly an incubation at 95°C for 10 min, followed by 40 cycles of 95°C and 60°C for 15 s and 1 min, respectively. Amplification specificity was estimated if a unique peak was found in the melting curve of each gene analysed. EF1α and β-TUB were used as reference genes, described by [93,94], respectively, since they concluded that they were suitable for the studies of salt stress in pepper plants. Primers sequences and results obtained have been detailed in Additional file 3.

Statistical analysis
The experiment layout was a completely randomised design. The data from the biomass, gas exchange measurements, ion concentration analyses, osmotic potential, starch, total phenols, H 2 O 2 and proline content and gene relative expression were subjected to a two-way ANOVA (Statgraphics Centurion XVI for Windows, Statistical Graphics Corp.), where both accession and treatment were considered to be the factors of the analysis. With the photosynthesis parameters, 7DAT and 14DAT were analysed independently. As the interaction between both factors was significant, a one-way ANOVA was performed by joining both factors of the two-way ANOVA. Ulterior comparisons were made using Fisher's least significance difference (LSD) test at P < 0.05 with the same software.
Class comparison analyses of the microarrays were done using the Babelomics platform. Different treatments of the same accession (Salt/Control) and different accessions of the same treatment (A25/A6) were compared by a Limma test to compare genes, and a Benjamini and Hochberg test was run to reduce the false discovery rate. The adjusted P-value was selected at 0.05.