For this research, we used tomato seeds (Solanum lycopersicum cv. Moneymaker) which were sown in a 0.5 L plastic pot with a mixture of 1:1:1 of perlite, vermiculite, and peat as substrate. After 15 days post germination, seedlings were transferred to 2 L plastic pots containing the same substrate. The plants were watered once a week with a nutrient solution containing 1.1 g/L of Murashige and Skoog salts (MS; Murashige and Skoog, 1962). Tomato plants were cultivated and propagated in growth chambers at 25°C with a long-day photoperiod (16/8 hours light/dark). After 10-12 weeks, tomato plants were subjected to saline stress treatments by adding 400 mL of 300 mM NaCl for 72 hours. Three plants for each condition were considered. Leaf and root were collected separately for each plant in the periods 0, 3, 6, 12, 24, 48 and 72 hours of treatment with salt stress, to measure transcript levels. The collected tissue was quickly frozen with liquid nitrogen and stored at -80°C for further analysis.

The Solgenomic database (http://www.solgenomics.net) was used to search for S. lycopersicum protein displaying sequence identity with the SNARE-like protein from S. brachiata (accession number KF111691). The sequence of the gene Solyc06g068080.2 (SlSLSP6) was selected by the highest score of similarity of the gene product. Sequences were aligned using the MUSCLE software (https://www.ebi.ac.uk/Tools/msa/muscle/). Comparative analyses were performed for each protein with the BLASTp program (http://www.ncbi.nlm.nih.gov), InterPro (https://www.ebi.ac.uk/interpro/) and Pfam databases. A phylogenetic tree was built using the MEGA7 (http://www.megasoftware.net) (Kumar et al., 2016) with the neighbor-joining method (Saitou and Nei, 1987) and bootstrap analysis of 1000 replications.

Total RNA was extracted from leaves and roots of S. lycopersicum plants using the SV Total RNA Isolation System commercial kit (Promega, Madison, WI, USA). Each sample was treated with DNase I (Ambion^® TURBO DNA-free™) to remove remaining genomic DNA. RNA concentration and purity were estimated in a NanoQuant spectrophotometer (UV-160A, Kyoto, Japan).

First-strand cDNA synthesis was performed with 1–2 μg of isolated RNA using the Maxima H Minus First Strand cDNA Synthesis kit (Thermofisher, Massachusetts, USA). cDNA samples were stored at -20°C.

The PCR reactions were carried out in a final volume of 25 μL which contained DNA, 1X PCR buffer, MgCl₂, dNTPs, the corresponding oligonucleotides ( Supplementary Table 1 ) and the Taq polymerase enzyme (Promega Corporation). The PCR product was analyzed in agarose gel electrophoresis (1X TAE Buffer). RT-qPCR analysis was carried out using the Brilliant SYBR Green qPCR MasterMix System (Stratagene, La Jolla, CA) according to the manufacturer’s instructions on a DNA engine Opticon 2 Cycler System (MJ Research, Watertown, MA). For each sample, RT-qPCR was carried out in triplicate (technical replicates). The 2^−ΔΔCt method was applied to calculate the fold change of gene transcript levels (Livak and Schmittgen, 2001). Ubiquitin3 gene (accession X58253) was used as a housekeeping gene (Yáñez et al., 2009).

For the transformation of tomato plants, the coding sequence of SlSLSP6 was amplified from cDNA of leaves using the primers described in Supplementary Table 1 . After the amplification reaction, the DNA fragments were cloned into pGEM-T vector (Promega) and sequenced. Then, the complete cDNA of the SlSLSP6 was inserted between the XbaI-BamHI restriction sites, replacing β-glucuronidase gene (GUS gene) of the binary vector pBI121 (Clontech, Palo Alto, CA, USA), and staying under the control of the constitutive promoter CaMV 35S, generating the 35S::SlSLSP6 cassette. Tomato plants were transformed with the 35S::SlSLSP6 cassette and with the unmodified pBI121 for control plants (Vector). Chemo-competent Agrobacterium tumefaciens GV3101 was cultured in YM solid medium (0.04% w/v yeast extract, 1% w/v mannitol, 1.7 mM NaCl, 0.8 mM MgSO₄.7H₂O and 1.5 w/v of agar) and grown at 28°C for 48 hours with the followed antibiotics: kanamycin (50 mg/mL), gentamicin (100 mg/mL) and rifampin (100 mg/mL). The colonies were analyzed by PCR. A culture of the PCR-confirmed strains was used to transform tomato cotyledons from 10-day-old seedlings, according to the method of Fillati et al. (1987). After six months of organogenesis, the shoots obtained were individualized. The plants were selected in kanamycin antibiotic (50 mg/mL) and characterized by PCR and RT-qPCR. Sixteen SlSLSP6 overexpressor lines were selected (L1-L16).

SlSLSP6 overexpressing plants (L9 and L10) and control plants (WT and Vector) with similar size and number of leaves were subjected to salt stress conditions grown in pots with perlite and vermiculite substrate (1:1). These plants were subjected to irrigation with 200 mL of a 300 mM NaCl saline solution every 72 hours. Plant performance and tolerance parameters were evaluated after 5, 15 and 25 days of stress treatment. Each time of the experimental trial considers 3 plants as biological replicates. Relative water content (RWC) was estimated according to the modified protocol of Maggio et al. (2007) and Vasquez-Robinet et al. (2008), where RWC = [(FW – DW)/(TW – DW)] x 100. Three leaves of about the same size and location on the plant were weighed (fresh weight, FW) and then incubated in distilled water for 24 hours at room temperature in darkness. Then, the tissue was weighed to calculate the turgid weight (TW) and finally, the samples were dried at 60°C for 48 hours for the determination of the dry weight (DW). Additionally, the efficiency of photosystem II represented as Fv/Fm and the performance index (PI) were evaluated. These measurements were obtained using a pocket fluorometer (Hansatech, King’s Lynn, Norfolk, England) following the procedure described by Giorio (2011). The chlorophyll content was evaluated in leaf discs following the previously described protocols (Orellana et al., 2010; Salinas-Cornejo et al., 2021). According to this, the leaf discs were ground with 80% acetone (v/v) and the extract was centrifuged. The absorbance of the supernatant was measured using the NanoQuant spectrophotometer (UV-160A, Kyoto, Japan), using the method described by Lichtenthaler and Wellburn (1983).

To determine lipid damage, the lipid peroxidation marker (MDA) was determined according to the modified protocol of Dionisio-Sese and Tobita (1998) and Ghanem et al. (2008). The concentration was obtained by subtracting the absorbance between 600 and 532 nm. Then, the value was multiplied using the molar extinction coefficient of 155 mmol^-1cm^-1. In addition, the ROS accumulation was observed in cells of the apical zone of the root (root length of 0.5 cm) of transgenic tomato plants (L9 and L10 lines) and control plants (WT and Vector). Tomato seedlings exposed for 40 minutes to liquid MS medium supplemented with 200 mM NaCl were subsequently incubated for 20 minutes with the fluorescent tracer H₂DCFDA 10 μM (2’,7’-dichlorodihydrofluorescein diacetate, ThermoFisher Scientific, Massachusetts, USA) as described by Leshem et al. (2006) and Zhao et al. (2009). To delimit the plasma membrane of each cell, the seedlings were stained with FM4-64 5 μM for 15 minutes in darkness at 4°C. Then, the roots were visualized immediately.

The subcellular accumulation of SlSLSP6 was performed in 10-day-old Arabidopsis thaliana seedling lines expressing different organelle markers described by Geldner et al., 2009. Each line was transiently transformed using the 35S::GFP-SlSLSP6 construct, contained in the pAM1 vector. The SlSLSP6 gene, amplified as described in section 2.5, was inserted into the pAM1 vector in the C-terminal orientation of the GFP gene. We use an adaptation of the technique carried out by Wu et al. (2014), called AGROBEST. The Arabidopsis marker lines RabG3f-mCherry (Germplasm ID CS781670; late endosome/tonoplast), RabF2a-mCherry (Germplasm ID CS781672; late endosome/prevacuolar compartments), VAMP711-mCherry (Germplasm ID CS781673; tonoplast), VTI12-mCherry (Germplasm ID CS781675;early endosome/TGN), SYP32-mCherry (Germplasm ID CS781677; Golgi complex) and PIP1;4-mCherry (Germplasm ID CS781687; plasma membrane) were grown in similar MS agar plates in growth chambers.

The rate of plasma membrane internalization was determined in root apical zone cells (root length 0.5 cm) of both transgenic tomato seedlings (L9 and L10 lines) and control plants (WT and Vector) using the membrane tracer FM4-64 ( Molecular Probes, Eugene, OR). Roots were stained with FM4-64 (5 μM) in a solution of MS liquid medium at 4°C for 20 minutes in darkness. At 5-, 15-, and 30-minutes at room temperature, different seedlings were by imaged. To determine the effect of salt stress on the endocytic rate, a group of seedlings, previously stained with FM4-64 (5μM), was subjected to salt stress by incubation with liquid MS medium supplemented with 200 mM NaCl for 30 minutes at room temperature.

The accumulation of sodium into the vacuole was determined using the Na⁺ indicator Sodium Green™ (ThermoFisher Scientific, Massachusetts, USA). Tomato seedlings with root length of 0.5 cm were incubated in MS medium supplemented with 200 mM NaCl for a period of 16 hours. After treatment, the seedlings were incubated with 5 μM Sodium Green™ for 2 hours and a non-ionic surfactant Pluronic F-127 20% w/v (ThermoFisher Scientific, Massachusetts, USA) to improve the uptake of the dye, following an adaptation of the method described by Tarte et al., 2015 (San Martín-Davison et al., 2017; Salinas-Cornejo et al., 2021). Then, the roots were stained for 20 minutes with FM4-64 5 μM (in the dark at 4°C) to delimit the plasma membrane and visualized.

To quantify colocalization data, colocalization between GFP-SlSLSP6 and mCherry markers was quantified using the square of Pearson correlation coefficient (R²). The colocalization between GFP-SlSLSP6 and PIP1;4-mCherry was scored from a defined region of interest (ROI) for calculation of the Pearson correlation coefficient. R² was estimated considering 10 images using the Zeiss Zen 2010 software (zeiss.com).

The FM4-64 fluorescence was quantified manually at the intracellular and plasma membrane compartments using ImageJ software (http://rsb.info.nih.gov/ij/). FM4-64 internalization rate was determined as the ratio of both signals in each cell, using the equation described by Varma and Mayor (1998) and Baral et al. (2015a).

The quantification of the signal of H₂DCFDA and Sodium Green markers was performed using ImageJ software. The mean of fluorescence was scored from a manually defined ROI in each image. Images were processed using a color-coded intensity, with the lowest fluorescence intensity corresponding to blue and highest fluorescence intensity corresponding to green color.

To quantify the fluorescence intensities, laser, gain and pinhole settings of the confocal microscope were identical and constant among different treatments. All confocal experiments were independently repeated at least three times, considering for each experiment at least 25 cells from 5 roots of 5 different plants for each genotype.

Root cells were visualized with Zeiss-LSM 710 confocal microscope (Zeiss Axio Observer Z1 equipped with LSM700 module and PTM detector). An Fluar 40x/1.30 oil M27 objective or N-Achroplan 10x/0.25 M27 lens were used. Digital zoom 1X or 4X. The parameters were: laser 555 nm/emission>588–596 nm for FM4-64 or fluorescent protein mCherry, laser 488 nm/emission 517–561 nm for the fluorescent protein GFP, Sodium-Green or H₂DCFDA (filter SP 555 Zeiss), pixel dwell of 25.2 μs, 488/555 nm lasers set at 2%, pinhole diameter of 70 μm, and beam splitter MBS 405/488/555/639.

Statistical analyzes were performed using the R program (version 1.7). Statistical significance was determined by one-way ANOVA analysis (p < 0.05).

Using the Solgenomics database, the sequence coding for a putative SNARE-like in Solanum lycopersicum was identified. This gene was named SlSLSP6 (Solyc06g068080.2), which has not been previously characterized. The analysis using NCBI Blastp tool indicated that the amino acid sequence of SbSLSP from Salicornia brachiata (Accession AGV05388) displayed an 84.35% identity to the sequence SlSLSP6 from Solanum lycopersicum (Accession XP_010322672). SlSLSP6 is a protein of 147 amino acids with no transmembrane domains in its structure as expected for a SNARE protein. However, SlSLSP6 displayed putative palmitoylation site at the cysteine residue located at position 115 (Cys115) at the LDKYGKICLCLDEIV motif that may allow it to interact with membranes ( Supplementary Figure 1 ). Multiple alignment and comparison of the amino acid sequences of SlSLSP6 with SNARE-like from Arabidopsis thaliana (AtSLSP), Salicornia brachiata (SbSLSP) and two other sequences identified in Solanum lycopersicum (SlSLSP3 and SlSLSP11) revealed a high identity score as well as the presence of conserved domains as a SNARE-like superfamily protein ( Supplementary Figure 1 ). The phylogenetic analysis of plants SNARE-like proteins showed that SlSLSP6 clustered with proteins from the halophytic plants S. brachiata and Zostera marina (Park et al., 2016; Singh et al., 2016) ( Supplementary Figure 2 ). SISLSP6 is grouped also with the proteins from Nicotiana tomentosiformis and Solanum pennelli, considered salt-tolerant plants. Consequently, based on the high sequence identity and the presence of the structural domains of the SNARE-like superfamily proteins, the SlSLSP6 is predicted to function as a SNARE-like protein in S. lycopersicum.

The S. lycopersicum plants were subjected to saline stress (300 mM NaCl) for 72 hours, to evaluate whether the SlSLSP6 responds to such a challenge ( Figure 1 ). The salt treatment indeed was a stressful condition since the salt stress-marker-genes AREB1 (Orellana et al., 2010; Madrid-Espinoza et al., 2019) and TSW12 (Torres-Schumann et al., 1992; San Martín-Davison et al., 2017) responded as expected, increasing their transcript level along the treatment ( Figures 1B, C ). Interestingly, SlSLSP6 transcript levels increased at 3 hours of salt stress in both roots and leaves. This increase was maintained as far as 12 hours of treatment ( Figure 1A ). After 24 hours of salt stress, the SlSLSP6 transcript level becomes similar to the beginning of the trial (time 0) in both tissues. These results indicate that SlSLSP6 is induced under salt stress in leaves and roots suggesting the participation of the SlSLSP6 gene product in the primary response of S. lycopersicum to saline stress.

To evaluate the participation of SlSLSP6 in the response to salt stress, we study the effects of its overexpression in S. lycopersicum. Then, cDNA of the SlSLSP6 gene was cloned into the binary vector pBI121 to express the coding sequence under the 35S promoter in plants. S. lycopersicum cv. Moneymaker was transformed mediated by Agrobacterium tumefaciens and transformant explants were selected by kanamycin resistance and verified by PCR. The same protocol was followed transforming tomato plants with the empty vector to be used as a control (Vector) for the following experiments. The expression levels of SlSLSP6 were evaluated by qRT-PCR on the SlSLSP6 transformant plants. Three lines were identified as overexpressors of SlSLSP6 ( Supplementary Figure 3 ). Two of them, L9 and L10 were selected for the further characterization.

The response to salt stress of SlSLSP6 overexpressor plants was evaluated and compared to control plants as well as wild type plants. Two-month-old tomato plants were irrigated every 72 hours with a solution of 300 mM NaCl. Plant performance parameters were evaluated just before starting the trial and after 5, 15 and 25 days of salt treatment. After 15 days of treatment, the macroscopic symptoms as yellow leaves induced by salt stress began to appear in wild type and Vector plants; however, they are less abundant on L9 and L10 lines. After 25 days of salt stress, there were evident differences between the control plants and L9 and L10 plants. The symptoms of chlorosis and wilting were dramatic on WT and Vector plants showing a significant loss of the older leaves. However, L9 and L10 plants exhibited less chlorotic leaves and displayed several green leaves after the same period of salt stress treatment, suggesting an increase of salt tolerance due to the higher level of SlSLSP6 gene product ( Figure 2A ). Consistently, the RWC of L9 and L10 lines reached 74% and 64%, respectively while in wild type plants decreased to 56% after 25 days of salt stress ( Figure 2B ). This suggests that there is greater turgidity in the leaves of the overexpressing lines, which is consistent with the better appearance of the plants ( Figure 2A ). The applied salt stress treatment resulted in 70% reduction of chlorophyll levels in wild type and Vector plants. However, the overexpression of SlSLSP6 rescued the drop on chlorophyll (chlorophyll A and chlorophyll B) content induced by 25 days of salt stress ( Figures 2C–E ). These results suggest that the high levels of SlSLSP6 somehow increase the plant tolerance. To confirm that SlSLSP6 overexpression plants are more tolerant to salt stress, we evaluate physiological state of the plants by measuring the performance index (PI) and the efficiency of photosystem II (PSII), represented by the maximum quantum efficiency of the PSII (Fv/Fm). The PI values in the SlSLSP6 overexpressor lines were significantly higher compared to control plants, which decreased by approximately 75% after 25 days of stressful conditions ( Figure 2F ). As expected, the ratio Fv/Fm dropped drastically to 0.4 in control plants after 25 days of salt stress ( Figure 2G ). In contrast, SlSLSP6 overexpressors lines kept the levels of Fv/Fm as plants without any salt stress condition. Consistently, there was no membrane damage on overexpressor plants as result of the salt stress while in control plants lipid peroxidation increases along the salt treatment (1.8 times higher after 25 days) ( Figure 2G ). In addition, we used the fluorescent marker H₂DCFDA to determine the level of ROS in tomato roots under normal conditions ( Supplementary Figure 4 ) and subjected to 200 mM NaCl. The L9 line showed to be less tolerant since it had a similar result to the control plants. Instead, the L10 line showed 15 - 30% decrease of ROS accumulation compared to control plants ( Figures 3A, B ). Overall, the overexpression of SlSLSP6 prevents the detrimental effects of salt stress improving plant performance at the physiological and cellular level. These results demonstrate that the overexpression of SlSLSP6 improves the tolerance of the tomato plants to high salinity conditions.

SbSLSP from S. brachiata, a SNARE-like protein, is localized within the endomembrane system (Singh et al., 2016). Similar to SbSLSP, SlSLSP6 from S. lycopersicum would have a conserved function in the endomembrane system. Although SlSLSP6 does not display a transmembrane domain most likely it is able to be recruited to membranes in order to accomplish its function. To determine the subcellular localization of SlSLSP6, we overexpress transiently a GFP-SlSLSP6 construct in a collection of A. thaliana that express fluorescent endomembrane compartment markers (Geldner et al., 2009). The following mCherry markers were selected: plasma membrane (AtPIP1;4), early endosome/TGN (AtVTI12), Golgi complex (AtSYP32), late endosome/prevacuolar compartments (AtRabF2a), late endosome/tonoplast (AtRabG3f) and tonoplast (AtVAMP711). A. thaliana roots were transiently transformed using a protocol adapted from the AGROBEST technique (Wu et al., 2014; Wang et al., 2018), with two different constructions: the plasmid pAM1 carrying the sequence to express GFP : SlSLSP6 (35S::GFP-SlSLSP6) or with the empty vector (35S::GFP), as a negative control. The results indicated that GFP-SlSLSP6 signal overlapped with the mCherry AtPIP1;4 protein, located in the plasma membrane (yellow arrow Figure 4 and Supplementary Figure 5 ), strongly suggesting that SlSLSP6 is accumulated in the plasma membrane. In contrast, the GFP-SlSLSP6 distribution did not colocalize with the markers of the vacuole and Golgi complex. In addition to the plasma membrane distribution, GFP-SlSLSP6 displayed a punctate pattern (white arrowheads Figure 4 ) with similar characteristics to endosomes or PVCs. However, these bodies did not colocalize with the AtVTI12 nor AtRabF2a proteins, indicating there are different compartments than the early endosome/TGN nor LE/PVC. Therefore, these results suggest that the SlSLSP6 protein would be participating in processes linked to the plasma membrane, such as endocytosis, and may participate in pathways to the early or late endosome.

Endocytic trafficking has been proved to be important for salt stress tolerance (San Martín-Davison et al., 2017). In fact, an increased rate of endocytosis has been shown to increase plant tolerance under salt stress conditions (Salinas-Cornejo et al., 2021). As SlSLSP6 could be participating in processes linked to the plasma membrane, we hypothesized that this SNARE-like would have an active role in endocytosis. To test the rate of endocytosis, the internalization of the tracer FM4-64 was analyzed in SlSLSP6 overexpression lines (L9 and L10) and compared to wild type and Vector tomato seedling roots. The L10 overexpressing line grown in control conditions displayed a higher internalization rate than control plant: FM4-64 internalization after 15 and 30 minutes increased 1.6 and 2.0 times higher, respectively, compared to what was observed in the roots of the control plant ( Figures 5A, C ). Surprisingly, the L9 overexpressing line displayed similar levels of FM4-64 internalization than control plants although displayed higher levels of SlSLSP6 transcripts. This result confirms the functional role of SlSLSP6 on endomembrane trafficking as a SNARE-like protein at the plasma membrane. It has been shown that salt stress induces endocytosis in root cells (Salinas-Cornejo et al., 2021). Therefore, we tested whether such a challenge would stimulate further membrane internalization in SlSLSP6 overexpression lines. Then, tomato seedlings were subjected to 30 minutes of salt stress (200 mM NaCl) before the endocytosis rate was evaluated ( Figure 5B ). Indeed, salt treatment induced FM4-64 internalization on control plants ( Figure 5B ). Under this condition, the overexpressing lines L9 and L10 showed a higher internalization rate than control plants: 1.2 and 1.4, respectively, compared to 1.0 ( Figures 5B, D ). These results strongly suggest that the high level of SlSLSP6 gene product resulted in modulation of vesicular trafficking increasing the rate of endocytosis playing an important role on the mechanisms by which plants manage the salt stress.

An increase in the endocytosis rate promotes a greater compartmentalization of sodium in the vacuoles contributing to plant tolerance against salt stress conditions (Salinas-Cornejo et al., 2021). To verify this relationship, sodium accumulation was monitored using the Sodium Green™ fluorescent indicator in tomato roots of plants under normal conditions ( Supplementary Figure 6 ) and subjected to 200 mM NaCl for 16 hours. Both SlSLSP6 overexpressing lines L9 and L10 displayed a higher level of signal of Sodium Green than control plants ( Figures 6A, B ). Indeed, the L9 line displayed an increase of about 60% of fluorescent signal than control plants while L10 displayed the double amount of control plants. It was also observed that the accumulation of sodium occurs inside the vacuoles, both in the epidermal and cortical zone of the tomato roots, unlike what was identified in the control plants, where the accumulation is mainly in epidermal cells. Along with this, we observed that there was a strong accumulation of the sodium marker in structures with a pattern different from that of the vacuoles, whose composition and nature are unknown. These structures were also labeled with FM4-64 (white arrows Figure 6A ). It could be speculated that they correspond to endosomal aggregates from the plasma membrane via the endocytic pathway, but this requires further investigation. Consequently, these results suggest that the greater compartmentalization of sodium in the vacuoles and in endosomal aggregates could be due to a greater endocytic dynamic of the transgenic plants that overexpress SlSLSP6, allowing to explain their enhanced tolerance under high salinity conditions.