Involvement of ethylene receptors in the salt tolerance response of Cucurbita pepo

Abiotic stresses have a negative effect on crop production, affecting both vegetative and reproductive development. Ethylene plays a relevant role in plant response to environmental stresses, but the specific contribution of ethylene biosynthesis and signalling components in the salt stress response differs between Arabidopsis and rice, the two most studied model plants. In this paper, we study the effect of three gain-of-function mutations affecting the ethylene receptors CpETR1B, CpETR1A, and CpETR2B of Cucurbita pepo on salt stress response during germination, seedling establishment, and subsequent vegetative growth of plants. The mutations all reduced ethylene sensitivity, but enhanced salt tolerance, during both germination and vegetative growth, demonstrating that the three ethylene receptors play a positive role in salt tolerance. Under salt stress, etr1b, etr1a, and etr2b germinate earlier than WT, and the root and shoot growth rates of both seedlings and plants were less affected in mutant than in WT. The enhanced salt tolerance response of the etr2b plants was associated with a reduced accumulation of Na+ in shoots and leaves, as well as with a higher accumulation of compatible solutes, including proline and total carbohydrates, and antioxidant compounds, such as anthocyanin. Many membrane monovalent cation transporters, including Na+/H+ and K+/H+ exchangers (NHXs), K+ efflux antiporters (KEAs), high-affinity K+ transporters (HKTs), and K+ uptake transporters (KUPs) were also highly upregulated by salt in etr2b in comparison with WT. In aggregate, these data indicate that the enhanced salt tolerance of the mutant is led by the induction of genes that exclude Na+ in photosynthetic organs, while maintaining K+/Na+ homoeostasis and osmotic adjustment. If the salt response of etr mutants occurs via the ethylene signalling pathway, our data show that ethylene is a negative regulator of salt tolerance during germination and vegetative growth. Nevertheless, the higher upregulation of genes involved in Ca2+ signalling (CpCRCK2A and CpCRCK2B) and ABA biosynthesis (CpNCED3A and CpNCED3B) in etr2b leaves under salt stress likely indicates that the function of ethylene receptors in salt stress response in C. pepo can be mediated by Ca2+ and ABA signalling pathways.


Introduction
One of the great challenges facing agriculture today is the development of production systems that mitigate the deleterious effects of climate change, including drought and salinity 1 . In arid and semi-arid areas, soil and water salinity constitute two of the most important abiotic stresses that limit crop production. At present, >1 billion hectares worldwide are affected by soil salinity 2 .
Crop development and performance is severely affected by salinity. The primary effects of salinity are very similar to those caused by drought. A high concentration of salt in the soil reduces the plant's ability to absorb water, known as the osmotic effect due to salinity. This not only leads to reduced absorption of essential elements, such as K + , Ca 2+ and NO 3 − , but also a toxic accumulation of Na + and Cl − in aerial parts of the plant 3 . The accumulation of salt in leaf cells inhibits cell expansion and photosynthetic activity, which ultimately leads to a reduction in crop yield 4 .
The entrance and the perception of Na + in roots are little known processes. Sodium can enter the root through non-selective cation channels (NSCCs) 5 , although extracellular cation receptors, such as MONOCATION INDUCED [Ca 2+ ] and INCREASES 1 (MOCA1), have been detected, which are capable of sensing sodium and other cations, as well as promoting the influx of Ca 2+ into the cell 6,7 . The perception of the stress signal triggers a secondary signalling by reactive oxygen species (ROS) and abscisic acid (ABA), which also regulate the intracellular level of Ca 2+ . The cytosolic calcium activates phosphorylation cascades of Ca 2+ -dependent proteins or calcium sensors, including calmodulins (CaM), CaM-like (CML) and calcineurin B-like proteins (CBL), which leads to regulation of stress response genes 8,9 . To deal with salinity, plants have implemented three general mechanisms that improve plant tolerance to salt stress: (i) restoration of ion homoeostasis (Na + /K + homoeostasis); (ii) restoration of osmotic homoeostasis; and (iii) prevention and repair of cell damage. Ionic homoeostasis mediated by membrane ion transporters constitutes the main response mechanism against salt stress. Plasma membrane Na + /H + antiporters, such SALT OVERLY SENSITIVE 1 (SOS1/AtNHX7) and AtNHX8, extrude Na + into the extracellular medium in response to an increase in intracellular Ca 2+ . HIGH-AFFINITY K + TRANSPORTER1-like (HTK1-like) also has a strong affinity for Na +9 , which excludes the translocation of Na +10, 11 . Tonoplast Na + /H + antiporters, such as Na + /H + EXCHANGER 1-4 of Arabidopsis (NHX1-4), are also activated by Ca 2+ , transporting Na + (and K + ) into the vacuole [12][13][14] , reducing toxic Na + in the cytoplasm, and decreasing the osmotic potential of the cell. The overexpression of both plasma membrane and tonoplast antiporters results in a greater tolerance to salinity in a wide range of plant species [15][16][17] . K + transporters, including the high-affinity transporter family HAK/KT/ KUP, the HKT family of high-affinity K + transporters and the KEA family of K + efflux antiporters, are also involved in salt tolerance by maintaining K + /Na + homoeostasis 10,17 . To restore osmotic homoeostasis and cell volume and turgor, salt also activates the production of compatible solutes or osmolytes, including proline, sugar alcohols, sorbitol and anthocyanins, among others 18,19 . These osmolytes also function as protectors of membranes and proteins by reducing oxidative damage 20,21 .
The phytohormones ABA and ethylene play key roles in the defensive response of plants against abiotic stresses 7,22 . ABA is a positive regulator of plant defensive response. Under both salinity and water deficit, plants induce the production of ABA biosynthesis genes, such as NINE-CIS-EPOXYCAROTENOID DIOXYGENASES (NCEDs) and ABA DEFICIENTS (ABAs). ABA is then perceived by the ABA receptors PYRABACTIN RESISTANCE/PYR-ABACTIN RESISTANCE LIKE (PYR/PYL), which induce phosphorylation activity of the ABA-dependent SUCROSE NON-FERMENTING RELATED PROTEIN KINASES (SnRKs) family, and the activation of the ABA-dependent transcriptional network involved in ionic and osmotic adjustments in response to salt stress 7,23 .
The function of ethylene in salt stress response is, however, more controversial 24 . It is generally presumed that ethylene improves the response of plants to salt stress 25 . However, other authors supported a negative role of ethylene during salt stress, at least in certain growth stages in which its induction can activate oxidative stress and leaf senescence 26 . In Arabidopsis, ethylene positively regulates salinity response, and both ethylene biosynthesis and signalling genes are required for salt tolerance 27,28 . The biosynthesis ACS and ACO genes in Arabidopsis are induced under salinity conditions, but certain individual members can play a negative role in salt tolerance [29][30][31] . The ethylene signalling elements also participate in the response of plant to salt stress, but their functions are also unclear 24 . In Arabidopsis, the positive elements of the ethylene response are generally upregulated in response to salt and are positive regulators of salt tolerance; whereas, negative elements are downregulated by salt and are considered to be negative regulators of salt tolerance 24 . In contrast, orthologous ethylene positive signalling genes, including MHZ7/OsEIN2, MHZ6/OSEIL1 and OsEIL2, have an opposite function in rice, since their suppression produces salinity tolerance, while their individual overexpression enhances salt sensitivity 32 .
The five ethylene receptors of Arabidopsis, ETR1, ERS1, ETR2, ERS2 and EIN4, are negative regulators of the ethylene signal pathway, but play a contrasting role in salt tolerance. They possess highly similar amino acid sequences and domain structures. The ethylene binding property of all of the receptors resides in three or four Nterminal transmembrane helices that are located within the membrane of the endoplasmic reticulum 33,34 . These N domains are connected by a GAF domain to a C-terminal His protein kinase domain that is positioned in the cytoplasm 33,34 . ETR1, ETR2, and EIN4 have an additional C-terminal receiver domain 34 . Both gain-of-function and loss-of-function mutants have been described for the five Arabidopsis ethylene receptor genes. Dominant gain-offunction mutations in a single receptor gene lead to ethylene insensitivity; whereas, recessive loss-of-function mutations confer little or no phenotype, but the combination of two or three loss-of-function ethylene receptor mutations confers constitutive ethylene responses 35 . The function of the five Arabidopsis ethylene receptor genes in salt tolerance has been investigated in loss-of-function mutants during germination, finding that ETR1 and EIN4 inhibit, while ETR2 stimulates and ERS1 and ERS2 have no effect on, seed germination under salt stress [36][37][38] . These contrasting roles do not appear to require an ethylene canonical signalling pathway, but occur by regulating ABA signal transduction 24,25,36,39 . Silencing of alfalfa MsETR2 abolishes ethylene-triggered tolerance to salt stress, indicating that this ethylene receptor is a positive regulator of salt tolerance in alfalfa 40 .
Recently, García et al. 41 isolated four Cucurbita pepo mutants, etr1a, etr1a-1, etr1b and etr2b, all exhibiting a reduced response to ethylene, as well as concomitant changes in developmental traits regulated by ethylene [41][42][43] . The four mutations affected sex determination in this monoecious species, as well as female fertility. They convert female into male or female-sterile hermaphrodite flowers, which prevents self-fertilisation of homozygous mutant plants, and forces the maintenance of mutations in segregating populations 42,43 . The duplicated genome of C. pepo 44 contains six ethylene receptor genes, two paralogs for either ETR1 (CpETR1A and CpETR1B), ERS1 (CpERS1A and CpERS1B) and ETR2 (CpETR2A and CpETR2B), and the identified mutations affect three of the ethylene receptor genes. etr1a-1 and etr1a are A95V and P36L amino acid exchanges in the first and third transmembrane helix of CpETR1A, respectively, etr1b is a T94I amino acid exchange in the third transmembrane helix of CpETR1B, and etr2b is an E340K amino acid exchange in the coiledcoil domain between the GAF and histidine-kinase domains of CpETR2B 42,43 .
In this paper, we investigated the response of etr1b, etr1a, and etr2b gain-of-function mutants to salt stress during germination, seedling establishment, and subsequent vegetative growth. Since the three mutants showed enhanced salt tolerance response during all studied developmental stages and reduced content of Na + in photosynthetic organs, we also analysed the molecular mechanisms involved in the enhanced salt tolerance of the Cucurbita etr mutants, including accumulation of osmoprotectants and activation of gene networks involved in the biosynthesis of ABA, Ca 2+ signalling elements, and Na + and K + membrane transporters reducing the accumulation of toxic Na + in shoots and leaves.

Results
Tolerance of etr1b, etr1a, and etr2b to salt stress during germination and early stages of seedling development To determine the ability of etr mutants to germinate in the presence of NaCl, WT and etr1b, etr1a, and etr2b mutant seeds were germinated in both water and 100 mM of NaCl up to 55 h, recording the initiation of seed germination every 2 h. The results are shown in Fig. 1. In water, both WT and the three etr mutants showed a similar germination rate, although the mutant seed was slightly delayed with respect to WT (Fig. 1A). The NaCl treatment delayed germination of both WT and etrs, but the delayed time was much higher in the WT, meaning that the three etr mutants germinated faster than WT under salt stress.
The salt treatment affected WT and mutant seed differently for different germination parameters, including germination initiation, average time for 50% germination, and average germination time (Fig. 1B, D). In water, the assessment of the three germination parameters in etr seeds was highly similar to that of WT. Under salt stress, however, there was a significant increase in germination initiation, 50% of germination and average germination time in WT and mutant seeds, but the percentage of increase of these three parameters in NaCl with respect to water was significantly lower in the three mutants compared with their corresponding WT genotypes (Fig. 1B). Taken together, the data revealed that the three etr mutants are all more tolerant to salt stress than their corresponding WT during germination.
Seedling growth was also differentially reduced by salinity in WT and ethylene-insensitive mutants (Fig. 2). Radicle and hypocotyl growth rates were both reduced in response to NaCl treatments in WT and etr mutants, but the mutant seedlings were always less affected than WT ones (Fig. 2). When germinated and grown in water, the length of the radicle 48 h after germination was similar in WT and mutants, but the reduction of the radicle length under salt stress conditions was much more noticeable in the WT seedlings ( Fig. 2A). The same was true for the length of the hypocotyl 3 days after germination, a parameter that was much more reduced in WT than in etr mutants (Fig. 2B). Under salt stress, in fact, WT seedlings reduced the length of their hypocotyls by approximately 50%, while etr mutants exhibited a reduction of only 20-35% ( Fig. 2A, B). Figure 3A, B shows the effect of salt stress on the root and shoot growth, and root balls of WT and etr mutants, 20 days post-germination. Under control conditions, the root and leaf biomass of mutant seedlings was much higher than that of WT, indicating a higher vigour in the three mutant plants (Fig. 3C, D). Although root biomass was decreased considerably under salt stress, that of mutant plants was similar to that of the WT control plants grown in water (Fig. 3C). The biomass of the aerial part of the plant was also significantly higher in the etr mutants, and although reduced by the NaCl treatment, the leaf biomass of the mutant plants under salt stress was also higher than that of the WT (Fig. 3D). The reduction in leaf and root biomass in response to salt stress was not significantly different between WT and mutant plants (Fig. 3C, D). These data demonstrate that etr1b, etr1a and etr2b seedlings were more vigorous than those of WT under control and salt conditions, but the responsiveness of WT and mutant plants to salt stress did not significantly differ, at least during the first 20 days of vegetative development. The growth of roots, shoots and leaves was always higher in the mutant (Fig. 4) under both control and salt stress, which confirmed the higher vigour of the ethylene receptor mutants observed in previous experiments, and the higher salt tolerance of the mutant. However, the relative response of WT and mutant plants to salt stress differed throughout plant development. At early stages (5 and 10 DAS) etr2b and WT responded similarly to salt stress, reducing both plant height and root length (Fig. 4A, B). The reduction in leaf and root biomass Fig. 1 Effect of salt stress on germination parameters of WT and etr1b, etr1a, and etr2b. A Germination rates of WT and etr1b, etr1a, and etr2b ethylene receptor mutants under control and NaCl conditions. The percentage of germination was analysed every 2â€‰h at the indicated time points. The data represent means of three independent replicates with at least 50 seeds counted per replicate. B, C, and D Effect of NaCl stress treatment on germination initiation, time at which 50% of seed is germinated, and average germination time. The bottom graphs show the percentage of increase of each parameter in response to salt stress in WT and mutant plants with respect to plants of the same genotype grown under control conditions. Means were obtained from four independent replicates with at least 50 seeds per replicate. Different letters indicate statistically significant differences (Pâ€‰<â€‰0.05) between samples between 5 and 30 DAS was also similar in WT and mutant plants (Fig. 4C, D). At 45 DAS, however, the salt sensitivity of the mutant was significantly lower than that of the WT, with etr2b exhibiting a significantly lower percentage of reduction in leaf and root biomass than WT (Fig. 4C, D). These data indicate that etr2b, and probably the other two ethylene receptor mutants, have an enhanced tolerance to salt stress not only during germination, but also during plant vegetative development. Table 1 shows the effect of salinity on the nutrient content of WT and etr2b leaves and roots. For most of the nutrients, no significant differences were found between WT and etr2b plants in either roots or in leaves. For K and Ca, no differences were identified between WT and etr2b, except that the Ca content was slightly lower in the mutant under non-saline conditions. The total N content was reduced in response to salt stress in both WT and mutant leaves, but no difference was detected between the response of the two genotypes. As expected, salt-stressed plants increased their content of the phytotoxic elements Cl − and Na. In leaves of salt-stressed plants, Na accumulated at least 4.5 mg/kg more in WT than in etr2b, but Cl − content was found to be similar in the two genotypes. In roots, however, no significant difference was found between WT and etr2b for either Na or Cl − ( Table 1).

Comparison of stress metabolites and gene expression in WT and etr2b in response to salt stress
To gain insight into the mechanisms that regulate the enhanced salt tolerance of etr2b, the content of some metabolites and the expression of genes related to salt stress in different plant systems were measured. Figure 5 shows the contents of proline, total carbohydrate, and anthocyanin in leaves and roots of WT and mutant plants grown under either control or salinity conditions for 45 days. Under control conditions, most of the assessments were similar in WT and etr2b plants, although etr2b roots showed a decreased content of proline, and etr2b leaves reduced their content in total carbohydrates (Fig. 5). In salt-stressed plants, the response of WT and mutant plants was completely dissimilar. Salt induced the accumulation of proline, total carbohydrates and anthocyanins in both roots and leaves of the mutant plants, but hardly changed their contents in WT in either roots or in leaves ( Fig. 5A-C). The expression of genes associated with abiotic stress tolerance was also compared in WT and etr2b plants grown over 45 days under standard and saline stress conditions. Since the C. pepo genome is duplicated, we investigated the expression of paralogs from both A and B subgenomes (indicated by the letter A or B at the end of the gene name, respectively). The phylogenetic relationship between C. pepo selected genes (Supplementary  Table S2) and Arabidopsis homologues with known functions was previously examined for each gene family ( Supplementary Fig. S2), thus providing a likely function of the analysed genes in C. pepo. In fact, the name that we assigned to each C. pepo gene corresponds to the Arabidopsis gene, which had the most conserved protein identity ( Supplementary Fig. S2).
The expression of most of the genes associated with salt tolerance was much more induced in the mutant than in the WT plants, indicating an enhanced response of etr2b plants to salt stress (Fig. 6). K + transporter genes, including K + uptake permeases or KUPs (CpKUP6-1A, CpKUP6-1B and CpKUP6-2A, CpKUP6-2B), and K + /H + efflux antiporters or KEAs (CpKEA4-1A, CpKEA4-1B and CpKEA4-2A, CpKEA4-2B), with the exception of CpKUP6-2A, were upregulated by NaCl in both WT and etr2b, but the upregulation in the mutant was between 2 and 25 times higher than in the WT (Fig. 6A, B). Under control conditions, some of them, including, CpKEA4-1A, CpKEA4-1B, were also more expressed in mutant than in WT plants.
The same is true for genes encoding Na + /H + exchanger CpNHX1-3B and Na + transporter CpHKT1A, which were only upregulated in etr2b plants when grown under saline conditions (Fig. 6C). Of particular interest is the CpHKT1A gene, whose expression was already highest in non-stressed mutant plants, and was upregulated more than 300-fold in response to salinity in only the etr2b mutant (Fig. 6C). Genes involved in abiotic stress signalling pathways, including Ca 2+ signalling gene Calmodulin-binding receptor-like kinases CpCRCK2A and CpCRCK2B, and ABA biosynthesis genes CpNCED3A and CpNCED3B, were also more highly induced in etr2b plants than in WT in response to salt stress (Fig. 6D, E).

Discussion
Ethylene is a key modulator of plant response to salt stress, but its specific role in different plant species and plant developmental stages is unclear 24,27,32 . In Arabidopsis and other plants, including maize and tomato, ethylene positively regulates salt stress tolerance 25,[45][46][47] , however, in other plant species, such as rice and tobacco, ethylene plays a negative role in salinity stress response 24,32 . In this paper, we demonstrate that ethylene is also involved in the salt stress response of C. pepo. All of the physiological and molecular data presented in this paper indicate that gain-of-function mutations in three C. pepo ethylene receptor genes increase salt stress tolerance at germination and during seedling and plant vegetative development, suggesting that ethylene is a negative regulator of salt tolerance in C. pepo. Seed germination is severely affected by salinity, being the first process involved in the stress tolerance response 7 .
The phenotypes of loss-of-function mutants in Arabidopsis have demonstrated that ETR1 and EIN4 inhibit, while ETR2 enhances, seed germination under salt stress, and ERS1 and ERS2 have no significant effect on seed  36,37,39 . Accordingly, Arabidopsis loss-offunction mutations for ETR1, including etr1-6 and etr1-7, are more tolerant to salt stress and germinate before WT; whereas, gain-of-function mutants for ETR1, including etr1-1, etr1-2 and etr1-3, are more sensitive to salt stress and germinate later than WT under salt stress [48][49][50] . The accelerated germination of the three analysed gain-of-function etr1b, etr1a, and etr2b of C.
pepo under salt stress indicate that CpETR1B, CpETR1A, and CpETR2B are positive regulators of squash seed germination under salt stress. If this function is dependent on ethylene and the ethylene signal transduction pathway, ethylene would play a negative role in C. pepo salt tolerance, which is similar to what occurs in rice 24 . However, given that in Arabidopsis the function of ETRs in seed germination can take place independently of the canonical  ethylene signal transduction pathway, but appears to be mediated by ABA signalling 36,39 , it is also likely that the mechanisms underlying the function of squash ETRs under salt stress may also occur through ABA rather than the ethylene signalling pathway. The fact that the three etrs exhibit similar salt stress tolerance, but differ in the magnitude of their ethylene response (etr1b showed the most residual responsiveness and etr2b the least 43 ), suggests that the function of C. pepo ETRs on germination under salt stress could take place independently of ethylene. The reduced germination rate of ethylene-insensitive etr mutants in water may also be the consequence of an increased biosynthesis or sensitivity to ABA, as found in Arabidopsis ethylene-insensitive mutants etr1 and ein2 51 . Furthermore, the higher induction of the ABA biosynthesis genes CpNCED3A and CpNCED3B in etr2b plants under salt stress also supports the involvement of ABA in the salt tolerance of this mutant. The assessment of ABA sensitivity and ABA biosynthesis of C. pepo etr mutants in the presence and the absence of NaCl will provide insight into the interactions between ABA and ethylene signalling cascades during seed germination. We also observed that the three C. pepo etr mutations promote seedling and plant growth, resulting in higher plant height, and higher root and leaf biomass when grown under control standard conditions. This higher vegetative vigour of etr mutants was also identified in adult plants 42,43 , indicating that ethylene is a negative regulator throughout the vegetative development cycle of the plant. The stimulating effect of ethylene insensitivity on vegetative growth was also found in Arabidopsis ethyleneinsensitive gain-of-function mutants 49,52,53 , although other studies found no differences in total leaf area between WT and the etr1-1 mutant at earlier stages of vegetative development 54 . The higher constitutive growth and vigour of squash ethylene-insensitive mutants was correlated with their higher salt tolerance during seedling and vegetative plant development. These data contrast with those found in Arabidopsis, in which the higher vegetative growth of the ethylene-insensitive gain-of-function etr1-1 mutant and the transgenic Arabidopsis plants overexpressing the tobacco ethylene receptor NTHK1 were associated with a higher salt sensitivity, while the reduced seedling growth of the etr1-7 loss-of-function mutant was associated with greater salt tolerance 49,55 . The reduced ABA sensitivity of etr1-7 and the enhanced ABA sensitivity of etr1-1 may account for differences in plant growth and salt-tolerance, as explained for germination 49,51 . The enhanced salt tolerance of C. pepo etrs during seedling and plant vegetative growth could result from the inhibitory role of ethylene receptors in the ethylene signalling pathway 42,43 . However, it is also likely that the vegetative growth regulation of ethylene receptors occurs through the abscisic acid signalling pathway, as has been observed for ETRs and EIN2 in Arabidopsis 49,51,56 . Genes involved in both ABA biosynthesis and intracellular Ca 2+ signalling pathway were more induced in the mutant than in WT, which indicates that these two signalling pathways can coordinate the tolerance response of etr2b to salt stress. ABA is known to control the expression of ion transport genes and the influx of Ca 2+ in the guard cells that leads to stomata closure limiting water loss in leaves 57,58 , but also a number of ABA responsive genes that are involved in ion homoeostasis and osmotic adjustment 59 .

Mechanisms of salt tolerance in C. pepo etr mutants
A number of physiological and molecular responses, including Na + detoxification, ion homoeostasis, osmotic adjustment and ROS scavenging, have been developed in plants to combat salt stress 7 . The exclusion of Na + in photosynthetic organs is a mechanism that is widely used by salt-tolerant genotypes to maintain vegetative growth while dealing with the high toxicity of these elements in leaves 3 . In our experiments, the leaves of the salt-stressed WT plant accumulated seven times more Na than nonstressed plants, but the leaves of the salt-tolerant etr2b only accumulated 5.3 more Na. Given that WT and etr2b roots have similar Na content, these data demonstrate a high ability of the salt-tolerant mutant to restrict the transport of Na + from roots to leaves.
The exclusion of Na + and the higher growth rates of etr2b plants are likely to be regulated by the induced Na + and K + transporter genes in the leaf. As occurs with Arabidopsis AtNHX1 and AtNHX2 Na + /H + antiporters in the tonoplast, the induced CpNHX1-3B may be involved in sequestering Na + into the vacuole, thus reducing the content of Na + in the cytoplasm and alleviating osmotic stress 60,61 . They also function as K + /H + antiporters to maintain K + homoeostasis 13 . The CpHTK1A transporter is particularly interesting because it was upregulated 300 times more in etr2b than in WT. HTKs are high-affinity transporters for both Na + and K + , mediating root Na + uptake, Na + unloading from xylem sap, and leaf Na + refluxing to the phloem, which are mechanisms that increase leaf Na + exclusion [62][63][64][65] .
Gene expression data also suggest a positive role of the K + transporters KEAs and KUPs in combating salt stress. The Arabidopsis KEAs are K + /H + antiporters that mediate pH and K + homoeostasis in the inner and thylakoid membranes of chloroplast (KEA1, KEA2, and KEA3) or in endomembrane compartments (KEA4, KEA5, and KEA6) 66,67 . The C. pepo KEAs genes that were highly induced in etr2b under salt stress are highly homologous to the second clade. KUP/HAK/KT, on the other hand, is a large family of high-affinity K + transporters that function in potassium acquisition and translocation from roots to shoots 68,69 , facilitate K + efflux from the vacuole to regulate osmotic adjustment, and some of them are involved in plant growth and development 58,68,70 . The C. pepo KUPs that were upregulated in salt-stressed etr2b leaf have a higher homology with Arabidopsis KUP6, an ABA responsive K + subfamily transporter that has a key role in osmotic adjustment and K + homoeostasis of guard cells 58 .
C. pepo etr2b plants induced the accumulation of metabolites, such as proline, total sugars (glucose, fructose, sucrose, and trehalose) and anthocyanins at both the roots and shoots under salt stress, which demonstrates that ethylene receptors and the subsequent ethylene or ABA signal transduction pathways are mediating the production of these osmolytes and therefore the osmotic adjustment of cells under salt stress 71 . These osmolytes can lower osmotic potential 72 , but can also act as stabilisers of proteins and cell components against ion toxicity and NaCl-induced oxidative damage 71,73 . Proline is perhaps the main salinity-related osmolyte, and is considered a biochemical marker of salt stress 74 . Exogenous proline treatments and transgenics plants with enhanced production of proline are more tolerant to salt, while mutants that are deficient in proline exhibited a limited growth and development under salt stress 75,76 . The biosynthesis of proline and other osmolytes is induced by ABA in different systems 77 , suggesting again that the enhanced response of mutant ethylene receptors of C. pepo to salt stress is likely mediated by ABA.

Plant material
The ethylene receptor mutants analysed in this study, etr1b, etr1a and etr2b, were selected from a highthroughput screening of a Cucurbita pepo mutant collection by using the triple response of etiolated seedlings to ethylene 41 . In addition to their reduced triple response to ethylene, the three mutations convert female into hermaphrodite or male flowers, reducing or preventing self-fertilisation 42,43 . The mutants were therefore maintained in BC 2 S 1 segregating generations, obtained by crossing each mutant twice or more times with the background genotype MUC16, and then selfed. The mutations affect CpETR1B, CpETR1A, and CpETR2B genes; thus, the WT and mutant plants in segregating populations were selected by detecting the WT and etr1b, etr1a, and etr2b alleles using real-time PCR with TaqMan probes 42,43 . DNA was isolated from the cotyledon of seedlings after the development of the first true leaf (5 or 7 days after sowing, DAS) by using the CTAB protocol. The multiplex PCRs were done using the Bioline Sensi-FAST™ Probe No-ROX Kit, a set of forward and reverse primers amplifying the polymorphic sequence, and two allele-specific probes descriptive of the SNP of interest. The WT probe was labelled with FAM dye, while the mutant probe was labelled with HEX reporter dye. BHQ1 quencher molecule was used in both probes (Supplementary Table S1).

Seed germination under salinity stress
Seed germination of WT and etr1b, etr1a, and etr2b was tested under salinity stress. Seeds were sterilised with a 5% sodium hypochlorite solution for 10 min and rinsed in distilled water three times, before being incubated in 50 ml Falcon tubes containing 25 ml of distilled water (control) or 100 mM NaCl for 12 h at 25°C in darkness under continuous shaking. After the imbibition, the seeds were transferred to a dispositive designed to study seed germination ( Supplementary Fig. S1). Seeds were placed in a foam strip between two pieces of filter paper and two panes of glass of 12 × 20 cm. This "sandwich glass" was secured with two clips and situated vertically in a recipient with water (control) or 100 mM NaCl solution for seeds to germinate and grow vertically. The sandwich glass with seeds was then incubated in a growth chamber in darkness at 24°C and 80% RH for 55 h. Three-hundred BC 2 S 1 seeds, segregating for each etr mutant, were germinated and grown using both water and salt in four independent experiments.
The germinated seeds were recorded every 2 h for 55 h through digital images that were processed using ImageJ ® . Seeds were considered germinated when the seed coat was broken and primary root protrusion was visible (>1 mm). Germination initiation, time of germination at 50% of seeds, and average germination time were determined according to procedures described by Ranal and De Santana 78 . Root elongation from both WT and ert mutants was assessed from seedling images at 48 h of initiating germination.

Seedling and plantlets growth under salinity stress
After germination, seeds were transplanted into 54 seedling trays filled with a mixture of perlite and coconut fibre (20-80%), a substrate with low-cation exchange capacity. 150 seeds of each genotype (WT/WT and etr/etr), 75 germinated under salt stress and 75 germinated in water, were distributed in three independent experiments. Trays were incubated in a growth chamber in darkness at 24°C and 80% RH for 72 h, and hypocotyl elongation was assessed in all plants. Control seedlings were irrigated with a nutritive Hoogland solution with a conductivity of 2 dS/m; whereas, for those subjected to salt stress, the nutritive solution was supplemented with 35 mM of NaCl, which increased its conductivity to 5 dS/m. Seedlings of each segregating population were then genotyped with Taqman probes, and 72 WT/WT and 72 etr/etr plants from each mutant family were transplanted into 1 l pots containing the same substrate as previously, and grown for 20 additional days at 24°C under long-day photoperiod (16 h light/8 h dark) and 70% RH in three independent experiments. Half of the plants (36) continued to be irrigated with the standard nutritive solution as previously, while the other half (36) were supplemented with 35 mM of NaCl. Leaf and root biomass were compared between each WT and ert mutant grown under both control and salinity conditions.

Vegetative growth of WT and etr2b under salinity stress
Although etr mutations affect female fertility and prevent selfing 42,43 , we were able to pollinate the mutant flowers several days prior to anthesis, thus forcing selffertilisation of the mutant plants and obtaining 100% mutant offspring. This was only achieved in the etr2b mutant, which allowed the evaluation of a higher number of plants implementing the analysis of growth parameters in additional plant developmental stages, as well as biochemical and gene expression studies in this mutant. In this mutant family, separated WT and ert2b plants were cultivated for up to 45 days under either control or salt conditions following the protocol described in the previous section. The development of different growth parameters, including root length, plant height, and root and leaf fresh and dry weight, were compared between WT and ert2b at 5, 10, 20, 30, and 45 DAS. Three independent replicates of ten plants each were analysed for each genotype and irrigation conditions at each developmental stage. At 45 DAS, plants were also used to analyse the effect of etr2b mutation on the content in micro-and macro-elements, the accumulation of stress metabolites, and the relative expression of stress-related genes.

Evaluation of stress-associated metabolites in WT and etr2b plants
The concentration of different stress metabolites, including proline, total carbohydrates and anthocyanins, was assessed in dry leaves and dry roots of WT and etr2b plants at 45 DAS under control and salinity stress conditions. All determinations were carried out in triplicate, each containing plant material from four plants.
Proline was determined through the ninhydrin method 79 with minor modifications. 100 mg of dry sample was incubated in a 2 ml of ethanol 60% at 4°C for 12 h. 0.5 ml of this solution was then mixed with 1 ml of ninhydrin 1%, dissolved in 60% acetic acid, and incubated at 95°C for 20 min at room temperature. Proline concentration was finally determined by spectrophotometry at 520 nm, and expressed as µmol/g DW. Total carbohydrates concentration was assessed by the phenol-sulphuric method 80 with minor modifications. 100 mg of dry sample was incubated in 5 ml of ethanol 80% at 80°C for 1 h, and 1 ml of this solution was then mixed with 1 ml of a solution of phenol 5% and 5 ml of sulphuric acid 95-97%. Total carbohydrates were determined at 490 nm, and expressed as mg/g DW. Anthocyanin content was measured according to Mancinelli 81 . 100 mg of dry sample was incubated at 4°C for 12 h in 3 ml of a solution of ethanol acidified with 1% of HCl 37%. The spectrophotometry measurements were done at 530 and 657 nm, and the concentrations expressed as µg/g DW. All spectrophotometric readings were performed on 96-well microplates using the BioTek® UV-Visible Epoch™ spectrophotometer.

Comparison of micro-and macro-elements in WT and etr2b plants
Micro-and macro-elements were measured in 5 g of dry leaves and roots coming from the same three samples for each genotype and salinity condition used in the determination of stress metabolites. The elemental measurements were carried out according to the standard protocols dictated by the International Organisation for Standardisation (ISO) (https://www.iso.org/home.html). Total nitrogen was measured through elemental analysis (ISO-13878), chloride was determined by fragmented flow analysis (ISO-15682), and the rest of macro-and micronutrients studied (phosphorus, potassium, calcium, magnesium, sulphur, iron, manganese, copper, zinc, boron, molybdenum, and sodium) were assessed by ICP-OES Spectrophotometry (ISO-11885).

Assessment of gene expression by qRT-PCR in WT and etr2b plant
The relative expression of different salt-stressassociated genes was assessed by quantitative reverse transcription (qRT)-PCR in WT and etr2b plants grown under control and salt conditions for 45 DAS. The analysis was performed in three biological replicates for each genotype and growing condition, each one derived from a pool of leaves from four plants. Total RNA was isolated from 1 g of leaves according to the protocol of the Gen-eJET Plant RNA Purification Kit (Thermo Fisher). RNA was reverted to cDNA with the ADNc RevertAid™ Kit (Thermo Fisher). The qRT-PCR was performed in 10 μl total volume with 1×Top Green qPCR Super Mix (BioRad) in the CFX96 Touch Real-Time PCR Detection System Thermocycler (BioRad). The gene expression values were calculated using the 2 −ΔΔCT method 82 . EF1α was used as the internal reference gene. Supplementary  Table S2 shows the primers used for qRT-PCR reactions in each analysed gene.

Phylogenetic analysis
MEGA 10 software 83 was used to establish the phylogenetic relationships between C. pepo and Arabidopsis thaliana genes encoding for Na + and K + membrane transporters (KUPs, KEAs, NHXs, and HKTs), abscisic acid biosynthesis enzymes (NCEDs), and Calmodulinbinding receptor-like cytoplasmic kinases (CRCKs). Phylogenetic trees were performed using the Maximum Likelihood method based on the Poisson correction model, with 2000 bootstrap replicates. The protein sequences and information were obtained from the Arabidopsis Information Resource (https://www.arabidopsis. org/) and the Cucurbit Genomic Database (http:// cucurbitgenomics.org/).

Statistical analysis
Data were analysed for multiple comparisons by analysis of variance (ANOVA) using the statistical software Statgraphic Centurion XVIII. Differences between genotypes and treatments were separated by the least significant difference (LSD) at a significance level of P ≤ 0.05.