Variation in tissue Na+ content and the activity of SOS1 genes among two species and two related genera of Chrysanthemum

Background Chrysanthemum, a leading ornamental species, does not tolerate salinity stress, although some of its related species do. The current level of understanding regarding the mechanisms underlying salinity tolerance in this botanical group is still limited. Results A comparison of the physiological responses to salinity stress was made between Chrysanthemum morifolium ‘Jinba’ and its more tolerant relatives Crossostephium chinense, Artemisia japonica and Chrysanthemum crassum. The stress induced a higher accumulation of Na+ and more reduction of K+ in C. morifolium than in C. chinense, C. crassum and A. japonica, which also showed higher K+/Na+ ratio. Homologs of an Na+/H+ antiporter (SOS1) were isolated from each species. The gene carried by the tolerant plants were more strongly induced by salt stress than those carried by the non-tolerant ones. When expressed heterologously, they also conferred a greater degree of tolerance to a yeast mutant lacking Na+-pumping ATPase and plasma membrane Na+/H+ antiporter activity. The data suggested that the products of AjSOS1, CrcSOS1 and CcSOS1 functioned more effectively as Na+ excluders than those of CmSOS1. Over expression of four SOS1s improves the salinity tolerance of transgenic plants and the overexpressing plants of SOS1s from salt tolerant plants were more tolerant than that from salt sensitive plants. In addition, the importance of certain AjSOS1 residues for effective ion transport activity and salinity tolerance was established by site-directed mutagenesis and heterologous expression in yeast. Conclusions AjSOS1, CrcSOS1 and CcSOS1 have potential as transgenes for enhancing salinity tolerance. Some of the mutations identified here may offer opportunities to better understand the mechanistic basis of salinity tolerance in the chrysanthemum complex. Electronic supplementary material The online version of this article (doi:10.1186/s12870-016-0781-9) contains supplementary material, which is available to authorized users.


Background
Soil salinity is becoming a severe environmental stress all over the world. Currently, over 800 million hectares of the world's arable land are adversely affected by salinity [1]. The major toxic cation present in saline soils is Na + , so under saline conditions, plants must minimize their cytosolic Na + concentration to withstand the stress [2]. Three strategies have evolved to avoid the build-up of Na + in the plant shoot: the first restricts the movement of the ion from the soil into the root, the second sequesters Na + in the vacuole, and the third actively pumps Na + out of the cytoplasm into the soil [1,[3][4][5]. Various ion transporters are involved in these processes, but a particularly prominent class is represented by the Na + /H + antiporters. So far, two types of Na + /H + antiporter NHE/NHX1 and NHA/ SOS1 have been well characterized [2,6].
AtSOS1 is the first plasma membrane Na + /H + antiporter gene cloned from higher plant, primarily expression of AtSOS1 in epidermal cells at the root tip and in parenchyma at the xylem-symplast boundary of roots, stems and leaves, implying a role of this transporter in extruding Na + to the growth medium and controlling longdistance Na + transport in plants. Furthmore, under moderate salinity, sos1 mutant accumulated less Na + in its shoots than WT (wild-type) plants, also indicating that SOS1 participates in loading of Na + into the xylem [7,8]. Recently, several similar studies indicated this critical function in tomato [9] and in Thellungiella salsuginea [10]. SOS1 might also be involved in K + nutrition in plants and under salt stress it is more vital for the plant to keep a high K + /Na + ratio [6]. The sos1 mutant showed significantly reduced high affinity K + uptake and K + content [11,12], while higher K + efflux from sos1 root than that in WT plants [13]. Qi and Spalding (2004) demonstrated that SOS1 was required for protecting K + uptake through AKT1 and compromised K + nutrition during salt stress [14]. In addition, ZxSOS1 controls long distance transport and spatial distribution of Na + and K + and maintains Na + , K + homeostasis in the xerophyte Zygophyllum xanthoxylum [15]. Together, SOS1 is essential for plant to cope with salt stress by maintaining ions homeostasis and controlling longdistance Na + transport via the xylem [16,17].
The leading ornamental species chrysanthemum does not readily tolerate salinity stress, although some of its many related species do. The current level of understanding the mechanisms of salinity tolerance in this botanical group is still limited [32]. Here, the morphological effects of salinity stress, along with the extent of Na + and K + accumulation in chrysanthemum and its three more tolerant related species (C. chinense, A. japonica and C. crassum) have been explored. The SOS1 homologs present in each of the four species has been isolated and their contribution to salinity tolerance assessed by heterologously expressing them in a yeast mutant ANT3, and in transgenic chrysanthemum and A. thaliana. Furthermore, some important amino acid polymorphism for effective ion transport activity and salinity tolerance was also identified by mutagenesis.

Variation for salinity tolerance in the chrysanthemum complex
Most of the leaves of C. morifolium plants became wilted and chlorotic following a ten day exposure to the salinity stress, and their lower leaves were largely necrotic C. crassum plants were less severely affected by the treatment, while there was no evidence of any damage to either C. chinense or A. japonica plants, the leaves of which stayed green, with the plants maintaining a near-normal level of growth for up to 14 days (Fig. 1). Under the non-stressed growing conditions, there was no variation in tissue Na + concent between the four test species. However, when the plants were exposed to salinity, the tissue Na + content throughout the plant was increased in all four species. The mean increase was notably lower for C. chinense and A. japonica: in these two species, the Na + content in the roots (compared to the levels in non-stressed plants) rose by only 142.0 % and 156.0 % respectively, while for C. crassum and C. morifolium plants, the increase was 300.0 % and 324.0 % (Fig. 2a). The leaves behaved similarly, with the Na + content rising more markedly in C. morifolium than in others. The Na + content in the leaves of C. morifolium was 125.0 %, 169.0 % and 189.0 % that present in C. crassum, C. chinense and A. japonica plants exposed to the NaCl stress respectively (Fig. 2c). Moreover, the Na + content in the stems behaved in a consistent way, it was highest in C. morifolium, moderate in C. crassum and low in both A. japonica and C. chinense (Fig. 2b).
K + concent in the roots of C. chinense, A. japonica and C. crassum plants show nearly unchanged between control and salt stress except that of C. morifolium plants, whose K + contents were significantly decreased (Fig. 2d). K + level in both A. japonica and C. chinense stems was also unchanged, on the contrary, K + content in the stems of C. crassum and C. morifolium was distinctly reduced by 23.2 % and 43.6 %, respectively (Fig. 2e). While K + content in the leaves of all the plants tended to decrease, the reduction was far more significant in C. morifolium, i.e., 54.71 % (Fig. 2f ). Relative to normal condition, salinity appreciably decreased the K + /Na + ratio throughout the plants. C. chinense and A. japonica exhibited the highest K + /Na + ratio of 5.7 and 5.6, respectively. In comparison, C. morifolium showed a minimum value in this ratio (1.8), while C. crassum showed an intermediate ratio of 3.1 (Fig. 2g-i). Overall, K + /Na + ratio of the salt-sensitive plants was much lower than that of the salt-tolerant plants, indicating that salttolerant plants excluded Na + and imported K + more effectively than salt-sensitive plants did.

Sequence analysis of the SOS1 homologs
A summary description of the AjSOS1, CrcSOS1, CcSOS1 and CmSOS1 sequences is given in Additional file 1: Table  S2. The ORF (open reading frames) sequence of the derived CcSOS1 fully matched that given in [23], but the UTR sequence differed slightly. All four SOS1 sequences were predicted to encode a Na + /H + antiporter. Both AjSOS1 and CrcSOS1 harbored a 1147aa ORF, whereas the CcSOS1 and CmSOS1 products were two residues shorter. The secondary structure of the four SOS1 proteins featured 12 transmembrane domains in their N terminal region according to TMPRED and included a long hydrophilic cytoplasmic tail in their C terminal segment (Fig. 3). The levels of peptide identity between AjSOS1 and the other three proteins were 98.5 % (CrcSOS1), 97.0 % (CcSOS1) and 97.4 % (CmSOS1); those between CrcSOS1 and the other two proteins were 97.1 % (CcSOS1) and 97.3 % (CmSOS1); and that between CcSOS1 and CmSOS1 was 99.5 %. Comparisons with other plant SOS1s revealed a high degree of sequence conservation: for example the level of amino acid sequence identity between four cloned SOS1s and A. thaliana AtSOS1 was 90.3 %, with the tomato SlSOS1 91.8 %, with rice OsSOS1 90.1 % and with Helianthus tuberosus HtSOS1 94.7 %. A phylogenetic analysis between four SOS1s and other palnt SOS1 transports [7, 9, 15, 18, 19, 21, 24-27, 29, 33-46] showed that AjSOS1 and CrcSOS1 were closely relatives, as were CcSOS1 and CmSOS1, while the nearest relatives of the four SOS1s as a group were HtSOS1 and SlSOS1 (Fig. 4). The presence of three conserved domains is required for the activity and regulation of the SOS1 protein: these are Nhap (an Na + /H + exchanger domain spanning the transmembrane region), InhiBD (an auto-inhibitory domain) and S2P (a phosphorylation motif recognized by SOS2) [19,47], and all three were present in the four SOS1s analysed here (Fig. 3).

SOS1 transcription profiling
In the roots, the abundance of AjSOS1 transcript increased gradually of salinity stressed plants, reaching a level of 3.87 fold above the base level after a 24 h exposure to 200 mM NaCl. CmSOS1 expression level increased only slowly over the first four hours of the treatment, peaking by 12 h, then decreased slightly, while the transcripts of CrcSOS1 and CcSOS1 maintained relatively constant (Fig. 5a). In the stems, all four SOS1s were up-regulated by the stress, their transcripts were greatest after 12 h (Fig. 5b). In the leaves, the level of transcription of both Fig. 1 The phenotypic response of chrysanthemum and its three close relatives to a ten day exposure to 200 mM NaCl. a-h Side view; i-p Vertical view from above; a-d and i-l Plants grown in the absence of stress; e-h and m-p plants exposed to NaCl. a, e, i, m C. chinense, b, f, j, n A. japonica, c, g, k, o C. crassum, d, h, l, p C. morifolium. Bar = 1.0 cm CrcSOS1 and CmSOS1 was highest at 24 h; that of AjSOS1 rosed most sharply between 4 h and 12 h, thereafter declined; and that of CcSOS1 remained relatively constant, with a two fold up-regulation occurring at 4 h (Fig. 5c). In essence, all four SOS1s were up-regulated by exposure to salinity, and the abundance of SOS1 transcript was greater in the more salinity tolerant plants.

Complementation of the yeast mutant with four SOS1s
All the yeast cells grew freely on YPDA in the absence of NaCl and one of the four related SOS1s transformed ANT3 cells grew much better on AP NaCl-containing medium than the control strain ( Fig. 6a-d). A comparison of the ability of four SOS1s transformants to grow in the presence of salt especially at 70 mM NaCl showed that the inclusion of AjSOS1 was the most beneficial, followed by that of CrcSOS1; the strain carrying CcSOS1 was better than CmSOS1, but was worse than CrcSOS1, while the inclusion of CmSOS1 was the least salinity tolerant of the transformed cells (Fig. 6d). Furthermore, qPCR (quantitative real-time polymerase chain reaction) analysis of the SOS1 expression levels were almost the same between yeast transformants for four SOS1s (Fig. 6e). The data demonstrated that Na + /H + antiporter activity of four SOS1s was essential and AjSOS1, CrcSOS1 and CcSOS1 were fully able to exclude Na + when expressed in yeast.

Overexpression of four SOS1s enhances salinity tolerance in transgenic chrysanthemum and Arabidopsis plants
Transgenic chrysanthemum lines overexpressing four SOS1s were successfully generated. qPCR analysis showed that compare with wide type (SM), SOS1 transcript abundance was not very high in the eight transgenic lines under control conditions but increased greatly upon 200 mM NaCl treatment (Fig. 7a). When exposure to saline hydroponics, most of the apex and edge of the lower leaves of all plants Fig. 2 Variation in tissue Na + 、K + content and K + /Na + ratio in the four test species in response to salinity stress. a Na + content in the root, b Na + content in the stem, c Na + content in the leaf, d K + content in the root, e K + content in the stem, f K + content in the leaf, g K + /Na + ratio in the root, h K + /Na + ratio in the stem, i K + /Na + ratio in the leaf. *,**: means differ significant from levels in the control treatment (P < 0.05 and < 0.01, respectively) showing signs of yellowing and necrotic after 1 day treatment, while after 3 days, the leaves of SM became severely necrotic and most plants died, the survival ratio of which was only 18 %. In the transgenic plants, symptoms of damage in leaves were much less evident in S1 and S2 than SM plants, but were worse than that of other transgenic plants, most of their upper leaves still remained green, and with less affected by salinity stress, which showed the transgenic plants maintained higher chlorophyll contents. The chlorophyll content is often used as index of salt tolerance in plants under salt stress, such as in Arabidopsis [48] and Residues conserved in at least two proteins are highlighted in white and blue. The black asterisks indicate conserved residues which were replaced in the site-directed mutagenesis experiment (see Additional file 2: Figure S4). Nhap, an Na + /H + exchanger domain spanning the transmembrane region; InhiBD, an auto-inhibitory domain; S2P, SOS2 phosphorylation motif tobacco [49]. The percentage survival of S1 and S2 plants was 33 % and 36 %, respectively, whereas that of other transgenic plants was 49 %-68 % (Fig. 7b-c). Furthermore, each of two transgenic A. thaliana lines overexpressing four SOS1s were selected for further study. For example no expression of exogenous SOS1 was detected in A. thaliana wide type gl1 but in these transgenic lines M-1, M-2, F-1, F-2, D-1, D-2,S-1 and S-2, which showed a high expression level of SOS1 (Additional file 3: Figure S2a). On 1/2 MS medium containing 150 or 75 mM NaCl, the seed germination rates, root length and fresh weight of transgenic A. thaliana wild type or sos1-1 lines were higher than those of the corresponding, and in the transgenic lines, the above index value in gS-1, gS-2, sS-1 and sS-2 were also notably lower than that of other lines (Additional file 3: Figure S2 and Additional file 4: Figure S3). These results indicated that the differences between the SOS1 transcription level in the transgenic lines did not very important effect in their tolerance

Site-Directed Mutagenesis functional analysis in yeast
The site-directed mutagenesis applied to AjSOS1 produced a set of 18 residue polymorphisms and additional one site-directed mutagenesis applied to CcSOS1 (Additional file 2: Figure S4). The hypothesis was that mutations a critical residue in AjSOS1 or CcSOS1 would generate a loss of salinity tolerance, as assayed by the yeast complementation test. The mutated forms were introduced into ANT3, and the drop test was conducted on AP medium containing 70 mM NaCl and 1 mM KCl. As depicted in Fig. 8, mutants G13E, T26S, F143I, V238L, . Actin was used as reference gene Fig. 6 Functional characterization of four SOS1s in the salinity sensitive yeast mutant ANT3 (ena1 nha1) and expression analysis of SOS1 gene in four SOS1s yeast transformants. ANT3 were transformed with plasmid containing four SOS1s (+AjSOS1, +CrcSOS1, +CcSOS1, +CmSOS1), G19 (ena1) and ANT3 (ena1 nha1) were transformed with the empty vector. G19 (ena1) cells were used as a positive control. Transformants were brought to a density 2 x 10 6 per mL, of which 5 μL (serially diluted) were spotted onto YPDA medium containing 0 mM NaCl (a) and AP medium containing 30 (b), 50 (c) and 70 (d) mM NaCl. Plates were incubated at 30°C for 2-4 days. e qPCR analysis of SOS1 expression in four SOS1s yeast transformants. The actin gene was employed as an internal control Y463H, E512G, Y549H, S639L, A919T, YG927HS, G982V, A1027V, N1109K and G1127A failed to complement the growth defect of yeast cells, which suggested that these mutations couldn't mediate Na + efflux in yeast and may be important for transport activity and salt tolerance of AjSOS1. The other AjSOS1-mutants supported more cell growth than either empty vector transformed ANT3 cells or those transformed with CmSOS1, indicating that these mutants were null mutations.

Discussion
At the phenotypic level, C. chinense and A. japonica both appeared to tolerate salinity stress rather better than either C. crassum or C. morifolium (Fig. 1), this finding consistent with the division of 32 chrysanthemum-related taxa into four clusters based on their morphological response to the stress [32]. The primary effect of salinity stress is a disturbance of cellular ion homeostasis, followed by the ingress of toxic levels of Na + into the cytoplasm. Patterns of ion accumulation have been exploited with some success as a means of discriminating between tolerant and sensitive species/cultivars [50]. The present data showed that exposure to 200 mM NaCl induced a smaller increase in tissue Na + content and a less reduction in tissue K + content and K + /Na + ratio in C. chinense and A. japonica than in C. crassum and C. morifolium (Fig. 2), consistent with the ranking based on the species' morphological response. The main conclusion was that the variation in salt tolerance displayed by the four species most likely reflected genetic variation for their ability to exclude the ingress of Na + , most probably thanks to have a more selective ion transport system. Similar conclusions have been drawn from the study of a range of other plant species [51][52][53][54]. Na + transporters are an important class of protein employed by A. thaliana to maintain ion homeostasis during an episode of salinity stress. The activity of AtSOS1 is central to the exclusion of Na + , as well as to its loading and retrieval into and out of the xylem [8]. The existence of an efficient SOS pathway would therefore make a major contribution to the superior salinity stress tolerance of C. chinense and A. japonica.
The SOS1 genes isolated from the four chrysanthemum and its related species all belong to the A. thaliana CPA1 (cation proton antiporter 1) family [6]. They all harbored three conserved functional domains Nhap, InhiBD and S2P (Fig. 3), a characteristic of SOS1 encoded proteins, and thought to be critical for their functionality [47]. In the absence of salinity stress, the abundance of SOS1 transcript in both the root and stem was higher in the more salinity tolerant A. japonica and C. chinense than in either C. crassum or C. morifolium, whereas in the leaf, the SOS1 transcript abundance in the four species differed little. The SOS1 genes were all up-regulated throughout the plant when salinity stress was imposed, inducing much higher transcript levels in the root than in either the stem or the leaf (Fig. 5). Moreover, the transcripts of reference gene Actin in four tested plants after salt treatment were relatively constant (Additional file 5: Figure S1). The behavior of SOS1 genes in a range of glycophytes is quite similar [7,10,15,18,22,29,43,55], although in other species, salinity stress has been found to significantly up-regulate SOS1 in the leaf but not in root [26,40,42,56]. In the former case, the assumption is that the SOS1 protein acts to remove Na + from the root cell, while in the latter, they have been suggested to function as maintainers of a low cytosolic Na + concentration in the leaf to protect photosynthesis. Notably, the abundance of AjSOS1 and CrcSOS1 transcript in salinity-stressed plants was greater than that of CcSOS1 and CmSOS1, which concords with the differences in ion accumulation and salinity tolerance displayed by the four species. Similarly, in a contrast between the salinity tolerant Populus euphratica and the more sensitive Populus popularis, the former was seen to accumulate a higher transcript abundance of genes related to Na + /H + antiporter activity [57]. Likewise, in a comparison of four Brassica spp. accessions, the more salinity tolerant entries displayed the highest level of SOS1 transcription [53], while in bread wheat 'Kharchia 65' , a cultivar known to be an efficient Na + exporter also showed high levels of SOS1 transcription [58]. Finally, in A. thaliana, the level of SOS1 transcription in the root has been shown to be inversely proportional to the accumulation of Na + in the plant [59]. Thus the evidence is very strong to support the notion that SOS1 proteins make an important contribution to salinity tolerance in the chrysanthemum species complex.
Heterologous expression in yeast has been exploited by a number of researchers aiming to functionally characterize plant SOS1 genes [10, 18, 20-22, 25, 60]. The ANT3-based system effectively discriminated between the efficacy of the chrysanthemum and its related species SOS1s in terms of their ability to counteract salinity stress. In particular, the assay showed that the AjSOS1 and CrcSOS1 products were able to compensate for the yeast host's lack of Na + -pumping ATPase ENA1-4 and plasma membrane Na + /H + antiporter NHA1 activity and the SOS1 expression levels of yeast transformants for four SOS1s were almost the same (Fig. 6). The Fig. 8 The salinity tolerance of ANT3 cells expressing altered forms of AjSOS1. The yeast cells were cultured overnight and a 5 μL aliquot (serially diluted) was spotted onto either a YPDA medium containing no NaCl (a-c) or an AP medium containing 70 mM NaCl (d-e). Plates were incubated at 30°C for 2-4 days implication is that these proteins mediate Na + efflux at the plasma membrane of yeast. Since AjSOS1, CrcSOS1 and CcSOS1 were much more effective than CmSOS1, it seems probable that these proteins are key determinants of the contrasting ionic homeostasis and levels of salinity tolerance of the four species. Similar conclusions have been drawn by contrasting the effectiveness of an SOS1 gene isolated from the salinity tolerant species Thellungiella salsuginea with that of AtSOS1 [10], and that of the SOS1 genes from the two halophytes Eutrema salsugineum and Schrenkiella parvula [35]. Takahashi et al. (2009) have shown that yeast cells heterologously expressing a PhaNHA1 allele (PhaNHA1-n) isolated from a salinity tolerant reed plants grew better than those harboring an allele (PhaNHA1-u) isolated from a salinity sensitive accession [21].
Several researchs have been shown that transgenic plants over-expression SOS1 improved salt tolerance [18,20,[27][28][29][30][31]61]. In this study, we demonstrated that over expression of four SOS1s also enhanced the salinity tolerance of transgenic chrysanthemum and A. thaliana wild type or sos1-1, and the overexpressing plants of SOS1s from salt tolerant plants were more tolerant than that from salt sensitive plants (Fig. 7, Additional file 3: Figure S2 and Additional file 4: Figure S3). These results were consist with the above functional analysis in the yeast mutant. To understand the reason for the different activities at SOS1s, a multiple alignment of four SOS1s proteins was analyzed and found that the AjSOS1 and CrcSOS1 sequences differed from CcSOS1 and CmSOS1 with respect to eighteen residues, and additional one residues in which CmSOS1 encode amino acid relative to the same ones of the other three SOS1s, and of which six were located in the membrane-spanning region and the other thirteen in the hydrophilic tail (Fig. 3).
When site-directed mutagenesis was carried out, it was found that a number of the altered polypeptides had no deleterious effect on the ability to complement the lesion in the ANT3 cell line, showing that these residues were not determinants of the protein's functionality. However, some of the altered polypeptides (G13E, T26S, F143I, V238L, Y463H, E512G, Y549H, S639L, A919T, YG927HS, G982V, A1027V, N1109K and G1127A) did reduce the level of the yeast's salinity tolerance, implying that these were essential for endowing AjSOS1 with the capacity to compensate for the host's defective Na + -pumping ATPase and plasma membrane Na + /H + antiporter activity (Fig. 8). G13E and T26S lie at the 5′ end of TMD1, F143I and V238L in the TMD4 and TMD6 respectively, while the remaining sites map to the C terminal hydropholic tail. Transmembrane regions in plant NHAs are thought to be important for Na + and H + exchange. The presence of a cytoplasmic tail indicates that the transporter is probably regulated by an external signal: under either salinity or oxidative stress, the AtSOS1 cytoplasmic tail interacts with RCD1, a regulator of the oxidative stress response [62]. Therefore, it is possible that the differential activity of the four SOS1s reflects a dissimilar interaction between their cytoplasmic tail and a signaling protein such as RCD1. Some of the mutations to AjSOS1 are likely to have induced alterations to the protein's secondary structure (Additional file 6: Figure S5), thereby potentially affecting its regulation and functionality. In A. thaliana, the salinity sensitive mutations sos1-3, sos1-8, sos1-9 and sos1-12 each comprise a single residue substitution in AtSOS1 [7], while the substitution E1044V in the putative auto inhibitory domain of E. salsugineum EsSOS1 is necessary, but not sufficient to facilitate the growth of AXT3K (Δena1::HIS3::ena4, Δnha1::LEU2, Δnhx1::KanMX4) yeast cells cultured on a saline medium [35]. In Triticum durum, the mutation of TdSOS1 alleles S1126A and S1128A (DSPS mutated to DAPA) have been associated with a reduced phosphorylation ability by the A. thaliana SOS2 kinase T/DSOS2Δ308, thereby preventing its activation of TdSOS1 [19]. Furthermore the alleles AtSOS1 S1136A and S1138A both interfere with phosphorylation by SOS2, while the G777D variant (sos1-8) is not activated by SOS2 [47]. Further investigations will be needed to provide much more evidences for contribution of the four SOS1s homologs in salt tolerance and to understand the basis of the observed variation in the activity of the AjSOS1 alleles.

Conclusions
In summary, in the four chrysanthemum and its related species, C. chinense, A. japonica and C. crassum were better tolerate than C. morifolium. They also had a superior capacity to prevent the accumulation of Na + and the reduction of K + in planta and their level of SOS1 transcription was higher. Moreover SOS1 sequence polymorphisms may be responsible for the higher efficacy of the AjSOS1 encoded protein. Taken together, AjSOS1, CrcSOS1 and CcSOS1 might be potential genes for enhancing salinity tolerance through transgenic strategies.

Methods
Plant materials, growing conditions and the assessment of salinity tolerance Samples of C. chinense, A. japonica, C. crassum and C. morifolium were obtained from the Chrysanthemum Germplasm Resource Preserving Centre (Nanjing Agricultural University, China). Uniform cuttings were vegetatively propagated in sand. For experiments designed to estimate Na + and K + content under salinity stress, a set of rooted seedlings at the 6-10 leaf stage \was transplanted into a 1:1 mixture of garden soil and vermiculite, and the plants cultured under a 16 h photoperiod, a day/night temperature of 22°C/18°C and a relative humidity of 68-75 %. The plants were irrigated with 200 mM NaCl every four days and photographed on day 10. Leaves, stems and roots were harvested separately on day 14, baked at 80°C for three days and weighed. A 0.1 g aliquot of each dry sample was digested in 2 mL 10 M HNO 3 , after which the solution were added to 10 mL with distilled water. Na + and K + contents were measured in this extract using an Optima 2100DV inductively coupled plasma optical emission spectrometer (Perkin Elmer, USA) [63]. Each sample was replicated three times. Means were compared using the Student's t test implemented in SPSS v17.0 J software (SPSS Inc., Chicago, IL, USA).
Isolation of SOS1 sequences RNA (ribonucleic acid) was extracted from roots of plants exposed to a hydroponic solution (half strength Hoagland's solution) containing 200 mM NaCl using the RNAiso reagent (TaKaRa Bio, Tokyo, Japan), following the manufacturer's protocol. A 1 μg aliquot of this RNA was reverse transcribed via M-MLV reverse transcriptase (TaKaRa Bio), primed by Oligo d(T) 18 . The SOS1 coding regions were amplified from the first strand cDNA (complementary deoxyribonucleic acid) using fusion PCR/overlap PCR [64] and the primer pairs F1/R1 and F2/R2 (sequences given in Table 1). The full length cDNA sequences were deduced from 5′ and 3′ RACE (rapid amplification of cDNA ends) amplicons [65], and then amplified from cDNA template using the primer pair Full-F/R (sequences given in Table 1). These amplicons were inserted into the pEASY-Blunt Zero Cloning vector (TransGen Biotech, Beijing, China) for sequence-based validation. Open reading frames (ORFs) was identified using the ORF finder program (www.ncbi.nlm.nih.gov/gorf/gorf.html). Hydrophobicity and  [66,67], and normalized against the transcript abundance at 0 h in the root of CmSOS1 at the respective time-point.

Complementation by the four SOS1 genes in the yeast mutants ANT3
The two bakers' yeast (Saccharomyces cerevisiae) mutant strains G19 (Δena1::HIS3::ena4) and ANT3 (Δena1::HI-S3::ena4, Δnha1::LEU2) were employed for performing a complementation assay. The former mutant lacks the Na + -pumping ATPase ENA1 to ENA4 while the latter,both the Na + -pumping ATPase and the plasma membrane Na + /H + antiporter NHA1 are defective [8,68]. First, the ORF of four SOS1s were amplified using Phusion High Fidelity DNA Polymerase (Thermo Scientific, USA) with the primer pair SOS1s-Sal-F/Not-R (sequences given in Table 1). Both the resulting amplicons and pENTR TM 1A plasmid were digested with Sal I and Not I restriction sites and the corresponding bands were ligated to yield the entry vector pENTR TM 1A-SOS1s. Then the pENTR TM 1A vector containing SOS1s were inserted into the yeast expression vector pAD426GPD [69] using LR Clonase II enzyme mix (Invitrogen, USA). The pAD426GPD-SOS1s constructs were validated by sequencing, then transformed into ANT3 cells, utilizing the Yeastmaker transformation system 2 (Clontech, Mountain View, CA, USA). The empty vector pAD426GPD was transformed into strain ANT3 and G19, which used as positive control. Transformants were selected by culturing on SD standard medium lacking uracil. The yeast cells' salinity tolerance phenotype was explored using a drop test in which a 5 μL aliquot of a saturated yeast culture, along with a similar volume of a serial dilution, was spotted onto alkali cation-free AP plates containing 1 mM KCl and with NaCl as designated [68,70]. The plates were held at 30°C for 2-4 days before being photographed. Moreover, four SOS1s transformants were cultured overnight and extracted RNA by using a fungi RNA extraction kit (Huayueyang biotech, Beijing, China). The reverse transcription and qPCR were performed as described above. The gene specific primer pair were DL-F/R and the reference gene was actin (GenBank accession number AAA34391), primer sequences were given in Table 1.

Generation of four SOS1s over-expressors and their response to salinity treatment
To further analyze the function of the four SOS1s and confirm their importance in plant salt tolerance, the plasmid pENTR TM 1A-SOS1s were subjected to the LR reaction to obtain expression vector pMDC32-SOS1s and introduced the constructs p2 × 35S:: SOS1s into salt sensitive C. morifolium 'Jinba' , A. thaliana (Columbia ecotype) wide type gl1 and mutant sos1-1 via Agrobacterium tumefaciens strain EHA105 mediated leaf disc and floral pollen dip method as described above [71]. RNA was isolated from control and 200 mM NaCl treated putative transgenic chrysanthemum and wide type (SM) plants, and processed for qPCR directed to SOS1 (using the primer pair DL-F/R) as described above. The primer pair CmEF1α-F/R was used to amplify the reference gene CmEF1α. Relative gene expression levels were also estimated using the 2 −ΔΔCt method [66], and normalized against the expression level of SOS1 in wide type plants under control conditions. Twenty plants of each transgenic lines (M1, M2, F1, F2, D1, D2, S1 and S2) and SM plants with three replicates were exposed to either control and a liquid nutrient solution (half strength Hoagland's solution) supplemented with 200 mM NaCl conditions, photographed on day 3 and calculated the survival rate on day 4. In addition, putative Arabidopsis transformants were firstly screened on hygromycin medium and then identified by RT-PCR analysis, based on the primer pair DL-F/R. Transgenic lines were used to assess salinity tolerance on 1/2 MS (Murashige and Skoog) agar medium supplemented with NaCl, as indicted for each case. Each experiment was performed three times and significant differences among treatments were identified by one-way analysis of variance and Tukey's multiple range test (p = 0.05). All statistical analyses were performed using SPSS v17.0 J software (SPSS Inc).