Identification of RmNHXs in Rosa multiflora
A total of 11 NHXs in Rosa multiflora was obtained according to the genome-wide analysis (Fig. 1a). The basic parameters about these RmNHXs were exhibited in Supplementary Table 1, including gene ID, CDS length, molecular weight (Mw) and isoelectric point (pI). As presented in Supplementary Table 1, the RmNHXs proteins varied from 268 to 695 aa in size with pI being between 5.03-9.03 and MW being between 29.47-62.51 kD. In addition, the phylogenetic analysis was performed based on the RmNHXs and NHXs from other plant species, which could be classified as vacuolar (Class I), endosomal (Class II), and plasma membrane types (Class Ⅲ) (Fig. 1b). Nine RmNHXs were categorized in Class I, one in Class II and one in Class Ⅲ, respectively (Fig. 1b).
The intron-exon structure analysis was conducted in the conserved Na+/H+ exchanger domains of the RmNHXs genes (Fig. 2a). As shown in Fig. 2a, the gene structure of 11 RmNHXs showed diverse intron-exon patterns with the number of introns ranging from 7 to 24. On the other hand, the longest intron of RmNHX2 is 2.9 kb. As a result, the genome sequences of RmNHX2 reached 7.8 kb (Fig. 2b). These data revealed that the changes of the gene structure in RmNHXs might lead to their function divergences during the evolution.
To gain insights into the function of RmNHXs, the 2.0 kb promoter region of the RmNHXs was adopted to identify the cis-acting elements using PlantCARE database. According to Supplementary Fig. 1, the promoter region of RmNHXs had various cis-acting elements (e.g., CAAT-box, CCAAT-box, and TATA-box), thereby demonstrating that the RmNHXs are closely related to the plant growth and development. Moreover, plant hormone-responsive and stress-responsive elements were also identified in the promoter region of RmNHXs, including salicylic acid responsive elements (TCA-elements), auxin responsive elements (TGA-elements), MeJA-responsive elements (CGTCA-motif, TGACG-motif), ABA responsive elements (ABREs), gibberellinresponsive elements (TATC-box), low-temperature responsive elements (LTR), defense and stress responsiveness elements (TC-rich repeats) (Supplementary Fig. 1). It should be noticed that there were different number of hormone and stress response motif in the promoter region of RmNHXs (Fig. 2c). These results indicated that RmNHXs might have different role in hormone and stress response.
Expression pattern of RmNHXs in Rosa multiflora
To further explore the potential functions of RmNHXs, we conducted qRT-PCR to assess the expression pattern of RmNHXs in response to salt stress. When subjected to salt stress, most of RmNHXs showed alteration of expression levels, while the expression pattern of each RmNHXs were different, demonstrating differences in the potential roles of RmNHXs under salt response (Fig. 3a). Among them, the expression level of RmNHX2 was drastically up-regulated by the salt treatment (Fig. 3a). RmNHX2 expression efficiently increased in 0.5 h in response to salt stress, and it reached the maximum expression in 5 h (about 6.8-fold), followed by gradually declining until last time point (48 h) (Fig. 3a).
As shown in Fig. 3b, RmNHX2 expression efficiently increased in 0.5 h in response to dehydration, after which it began to decline. When subjected to cold stress, the expressing state of RmNHX2 did not noticeably change, suggesting that RmNHX2 was not cold-inducible (Fig. 3b). In the case of exogenous ABA treatment, RmNHX2 mRNAs initially remained steady in 1h, and then elevated by 2.2-fold at 12 h (Fig. 3b). Compared with dehydration and ABA, the salt treatment caused a greater induction, indicating that RmNHX2 might impact plant abiotic stresses resistance, especially salt. The RmNHX2 was chosen for further function analysis.
Analysis of RmNHX2 protein sequence and its subcellular localization
The coding sequence of RmNHX2 was 1632 bp, encoding a 543 amino acids protein. RmNHX2 was deposited in GenBank with accession number MW358917. As revealed from multiple sequence alignment, RmNHX2 shared 12 conserved transmembrane domains highly consistent with FvNHX2 (94.11 %) (Fig. 3c). Motif scanning suggested that the amiloride binding site (87-LFFIYLLPPI-96) existed in the N-terminal of RmNHX2 (Fig. 3c). The structure and function of plant NHXs were closely related to their subcellular locations.
To determine putative role of RmNHX2 in Rosa multiflora, its ORF without stop codon was inserted into pHBT-GFP-NOS vector as driven by CaMV 35S. In addition, the tonoplast marker protein VAC-RK was fused to C-terminal RFP. The mentioned plasmids were co-transformed into Arabidopsis protoplast cells. As demonstrated from transient expression assays, RmNHX2-GFP fluorescence perfectly overlapped with VAC-RK-RFP fluorescent signals (Fig. 3d). The mentioned results suggested that RmNHX2 was localized at the tonoplast.
RmNHX2 overexpression in tobacco led to enhanced salt tolerance
As RmNHX2 expression was strongly up-regulated by salt, this study speculated that RmNHX2 might perform important functions against salt stress. To test the role of RmNHX2 in salt tolerance in depth, tobacco transgenic plants were generated through the leaf culture dish transformation. Totally, twelve positive transgenic lines were identified by genomic PCR analysis (Supplementary Fig. 3). And three RmNHX2 overexpressing (hereafter designated as OE1, OE7 and OE9, Supplementary Fig. 3) with higher expressing levels of RmNHX2 were chosen for further experiment. To determine whether RmNHX2 overexpression led to the increased salt stress resistance of the transgenic plants, the fully expanded leaves detached from 40-day-old RmNHX2-overexpresing lines and wild type tobacco (WTt) were administrated with 200 mM NaCl for 24 h. Before salt treatment, WTt and transgenic tobacco lines did not show any noticeable difference in morphology. However, the WTt leaves showed serious wilting as compared with the transgenic tobacco lines after incubation in 200 mM salt solution (Fig. 4a). Electrolyte leakage (EL) and malondialdehyde (MDA) as a vital indicator of cell damage was measured after salt stress between transgenic lines and WTt (Wang et al., 2019). EL of the transgenic lines (25.7 % for OE1, 34.5 % for OE7 and 37.1 % for OE9) were significantly lower than 52.7 % of WTt (Fig. 4b). According to Fig. 4c, the MDA level was prominently lower in the transgenic lines than in WTt after the salt stress. In addition, we also determined the long-time salt stress tolerance of the potted plants. When 1-month-old plants were subjected to 200 mM NaCl for 14 d, the WTt exhibited a more serious wilting or necrosis compared with the RmNHX2 overexpressing lines (Fig. 4d). Consistent with the enhanced salt tolerance phenotype, the content of EL and MDA in the transgenic lines was significantly lower than in the WTt after salt stress (Fig. 4e, f), indicating that cellular damage was more serious in the WTt. Furthermore, higher Chl level in transgenic lines compared with WTt was observed (Fig. 4g). The mentioned results demonstrated that RmNHX2 overexpression could increase the salt tolerance of transgenic tobacco.
Silencing of RmNHX2 in Rosa multiflora conferred sensitivity to salt stress
To elucidate the function of RmNHX2 in salt tolerance, the VIGS system was conducted to silence RmNHX2 expression in Rosa multiflora. The transcript level of RmNHX2 in the VIGS plants was down-regulated by 30-60% in comparison with that of the empty vector (pTRV2) transformed control plants (WTr) (Supplementary Fig. 4). We further assessed the salt tolerance of the mentioned two pTRV2-RmNHX2 VIGS plants (designated as pTRV2-1 and pTRV2-2) exposed to salt stress. First, both pTRV2-RmNHX2 VIGS and WTr plants grown in hydroponic solution were exposed to 300 mM NaCl for 7 d. After the salt treatment, the VIGS lines showed more severe wilting phenotypes in comparison with the WTr (Fig. 5a). In addition, EL and MDA, related to the cell damages, were determined. At the end of salt stress, the EL of pTRV2-1 (62.0 %) and pTRV2-2 (59.0 %) was significantly higher as compared with 46.4 % of WTr (Fig. 5b). Meanwhile, the pTRV2-1 and pTRV2-2 presented higher MDA relative to WTr after the salt stress (Fig. 5c). As shown in Fig. 5d, the total chlorophyll content of the RmNHX2-VIGS plants were nearly 2.0 folds lower than that of WTr under salt stress.
To gain insights into whether RmNHX2 was associated with salt tolerance, the RmNHX2-VIGS plants and WTr were treated at 300 mM NaCl. No visible phenotypic differences were identified between the RmNHX2-VIGS plants and WTr without salt stress. After the salt treatment for 10 d, the pTRV2-1 and pTRV2-2 plants suggested more serious damage relative to the WTr plants (Fig. 5e). When the salt treatment was completed, the Chlorophyll extracting of WTr displayed paler in color than the RmNHX2-VIGS plants (Fig. 5f). Consistent with the observed phenotype, the EL and MDA in the RmNHX2-VIGS plants were prominently higher than as compared with those of the WTr, implying that the VIGS plants were damaged to greater degree (Fig. 5g, h). Furthermore, the pTRV2-1 and pTRV2-2 plants achieved significantly lower levels of total chlorophyll than those of WTr after salt stress (Fig. 5i). Taken together, the mentioned data suggested that silencing of RmNHX2 in Rosa multiflora led to the elevated the salt sensitivity.
Analysis of H2O2 and O2- in transgenic tobacco and Rosa multiflora silenced plants under salt stress
Abiotic stresses were extensively evidenced to often induce ROS accumulation, thereby causing serious oxidative damage. ROS accumulation refers to a major indicator of stress tolerance. To verify whether RmNHX2 is closely associated with enhanced salt tolerance of plants via ROS scavenging, DAB and NBT were performed to detect H2O2 and O2- productions, respectively, in transgenic and wild-type plants after the salt stress. Under 30-day-old tobaccos grown in soil pots administrated with 200 mM NaCl for 14 d, the WTt leaf discs were stained deeper and more intensely by NBT and DAB in compared with the RmNHX2-overexpressing transgenic lines (Fig. 6a, b). In contrast to the mentioned, deeper DAB and NBT staining were visualized in pTRV2-RmNHX2 plants in contrast to the control plants (Fig. 6c, d), implying that ROS accumulation was enhanced when RmNHX2 was silenced. For the verification of DAB and NBT staining results, the authors determined the H2O2 and O2- levels in the salt-treated leaves as well. Consistent with the staining, the levels of H2O2 and O2- were significantly lower in the transgenic tobacco lines compared with WTt, as indicated by the measured results (Fig. 6e, f). In contrast, the pTRV2-RmNHX2 VIGS plants accumulated more ROS than did WTr (Fig. 6g, h). Both results suggested that the RmNHX2 overexpression led to the lower ROS accumulation in comparison with the WTt after salt treatment, whereas they also showed in an opposite way when RmNHX2 was down-regulated.
Analysis of antioxidant enzyme activities and expression levels of the encoding genes in transgenic tobacco and Rosa multiflora silenced plants under salt stress
Antioxidant enzymes positively impact ROS scavenging under abiotic stresses, which impelled us to assess the CAT, POD and SOD enzyme activities (Gill and Tuteja, 2010). When the 30-day-old potted tobacco plants were exposed to the salt treatment for 14 d, the activities of the CAT, POD and SOD in the transgenic plants were significantly higher than those in the WTt (Fig. 7a-c). Furthermore, the expression levels of CAT, POD and SOD in RmNHX2 overexpression tobacco plants were significantly up-regulated compared with that in the WTt after salt treatment (Fig. 7d-f). In contrast to the mentioned, noticeably lower three enzyme activities were observed in the two VIGS plants (pTRV2-1 and pTRV2-2) than in the WTr plants at the end of salt stress (Fig. 7g-i). However, it was displayed opposite way when the RmNHX2 expression was down-regulated (Fig. 7j-l). The mentioned finding suggested that RmNHX2 functioned in salt tolerance, at least partially, resulted from the increased actions of CAT, POD and SOD.
Opposite accumulation of Na+ and K+ in transgenic tobacco and Rosa multiflora silenced plants under salt stress
Plant's Na+ balance is considered disrupted under salt stress. Given that overexpression or down-regulation of RmNHX2 influenced the salt tolerance of plants, we were urged to assess whether ionic homeostasis was affected under salt stress. To tackle down the mentioned issues, we also measured the content of Na+ and K+ in the leaves and roots. Upon exposure to salt stress, the Na+ levels were 11.91 mg/g dry weight (DW) in WTt tobacco roots, 7.38 mg/g DW in OE1, 8.76 mg/g DW in OE7 and 9.63 mg/g DW in OE9 (Fig. 8a). The leaves Na+ levels were lower in the OE1, OE7 and OE9 as compared with those in the WTt (Fig. 8b). Moreover, the leave and the root in the RmNHX2 overexpression tobacco plants accumulated more K+ than those in the WTt leaves and roots (Fig. 8b). As a result, the Na/K ratios in WTt were noticeably higher than in the transgenic tobacco lines after salt treatment (Fig. 8c). Conversely, the levels of Na+, K+ and Na/K in the two pTRV2-RmNHX2 VIGS lines were significantly higher or lower than those of the WTr plants after salt stress, respectively (Fig. 8d-f). The mentioned data demonstrated that the enhanced tolerance of transgenic tobacco lines and susceptibility of pTRV2-RmNHX2 VIGS plants to salt stress were closely related to ion homeostasis.