Plant Molecular Biology

, Volume 74, Issue 4, pp 367–380

The woody plant poplar has a functionally conserved salt overly sensitive pathway in response to salinity stress


  • Ren-Jie Tang
    • National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological SciencesChinese Academy of Sciences
  • Hua Liu
    • National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological SciencesChinese Academy of Sciences
  • Yan Bao
    • National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological SciencesChinese Academy of Sciences
  • Qun-Dan Lv
    • National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological SciencesChinese Academy of Sciences
  • Lei Yang
    • National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological SciencesChinese Academy of Sciences
    • National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological SciencesChinese Academy of Sciences

DOI: 10.1007/s11103-010-9680-x

Cite this article as:
Tang, R., Liu, H., Bao, Y. et al. Plant Mol Biol (2010) 74: 367. doi:10.1007/s11103-010-9680-x


In Arabidopsis thaliana, the salt overly sensitive (SOS) pathway plays an essential role in maintaining ion homeostasis and conferring salt tolerance. Here we identified three SOS components in the woody plant Populus trichocarpa, designated as PtSOS1, PtSOS2 and PtSOS3. These putative SOS genes exhibited an overlapping but distinct expression pattern in poplar plants and the transcript levels of SOS1 and SOS2 were responsive to salinity stress. In poplar mesophyll protoplasts, PtSOS1 was specifically localized in the plasma membrane, whereas PtSOS2 was distributed throughout the cell, and PtSOS3 was predominantly targeted to the plasma membrane. Heterologous expression of PtSOS1, PtSOS2 and PtSOS3 could rescue salt-sensitive phenotypes of the corresponding Arabidopsis sos mutants, demonstrating that the Populus SOS proteins are functional homologues of their Arabidopsis counterpart. In addition, PtSOS3 interacted with, and recruited PtSOS2 to the plasma membrane in yeast and in planta. Reconstitution of poplar SOS pathway in yeast cells revealed that PtSOS2 and PtSOS3 acted coordinately to activate PtSOS1. Moreover, expression of the constitutively activated form of PtSOS2 partially complemented the sos3 mutant but not sos1, suggesting that PtSOS2 functions genetically downstream of SOS3 and upstream of SOS1. These results indicate a strong functional conservation of SOS pathway responsible for salt stress signaling from herbaceous to woody plants.


SOS pathwaySalt stressPopulusArabidopsisFunctional conservation


Soil salinity has become one of the major worldwide agricultural problems. It is a severe limiting factor that adversely affects plant growth in general and crop productivity in particular. For most plants, high salt concentrations impose both an ionic and an osmotic stress (Zhu 2001), coupled with secondary stresses such as oxidative stress (Borsani et al. 2001), genotoxicity (Albinsky et al. 1999) and nutritional disorders (Parida et al. 2004). Sodium (Na+) toxicity represents the major ionic stress associated with high salinity. The excessive accumulation of Na+ in the cytoplasm has injurious effects on plant cells in that it prevents uptake of the essential mineral nutrient potassium (K+), leading to insufficient cellular K+ amount for enzymatic reactions and osmotic adjustment. Therefore, maintaining an optimal cytosolic K+/Na+ homeostasis is crucial for plant salt tolerance (Niu et al. 1995; Hasegawa et al. 2000; Zhu 2003; Volkov et al. 2004). In order to preserve a low Na+ concentration in the cytoplasm, plants have evolved several machineries at the cellular level (Blumwald 2000; Horie and Schroeder 2004; Chen et al. 2007). One possible way is to restrict unidirectional Na+ uptake by roots (Rubio et al. 1995; Laurie et al. 2002). Alternatively, Na+ extrusion away from the cytoplasm (Shi et al. 2002) and Na+ compartmentation into the vacuole (Apse et al. 1999) may serve as active means to minimize sodium toxicity in plant cells. These vital Na+ translocation processes are mediated by a battery of ion transporters (Horie and Schroeder 2004) under the control of delicate signaling networks that modulate their activities and affinities (Xiong et al. 2002).

In Arabidopsis, the salt overly sensitive (SOS) pathway is responsible for Na+ homeostasis and salt tolerance under salt stress. This pathway was established through genetic identification of several salt overly sensitive (sos) mutants followed by molecular and biochemical characterization of the coding SOS proteins (Zhu 2002). A myristoylated EF-hand calcium-binding protein encoded by SOS3 presumably senses and interprets the cellular calcium signal elicited by salt stress (Liu and Zhu 1998; Ishitani et al. 2000). One of the primary downstream targets SOS2, a serine/threonine protein kinase (Liu et al. 2000), is activated and recruited to plasma membrane via direct interaction with SOS3 (Halfter et al. 2000; Quintero et al. 2002). Subsequently, the SOS2-SOS3 complex phosphorylates the Na+/H+ antiporter SOS1 (Shi et al. 2000) to stimulate its Na+/H+ exchange activity at the plasma membrane (Quintero et al. 2002; Qiu et al. 2002). In yeast, co-expression of SOS2, SOS3, together with SOS1 increases salt tolerance of a yeast mutant deficient in sodium transport more dramatically than expression of one or two SOS proteins, suggesting that full activation of SOS1 depends on the SOS2-SOS3 kinase complex (Quintero et al. 2002). In Arabidopsis, vesicles of sos1, sos2 and sos3 plants all display reduced plasma membrane Na+/H+ exchange activity compared to those of wild type plants, and a constitutively activated SOS2 protein enhance the Na+/H+ exchange activity in a SOS1-dependent and SOS3-independent manner (Qiu et al. 2002). Accordingly, overexpression of SOS1 (Shi et al. 2003) or a constitutively activated form of SOS2 (Guo et al. 2004) or the putative upstream sensor SOS3 (Yang et al. 2009) all confers enhanced salt tolerance on transgenic plants. Likewise, the rice SOS1 homolog OsSOS1 functions as a plasma membrane Na+/H+ antiporter and is phosphorylated by the SOS2–SOS3 kinase complex. Ectopic expression of OsSOS1 suppressed the growth defects of Arabidopsis sos1 mutant upon salt treatment. The SOS2/SOS3 homolog in rice, OsCIPK24/OsCBL4, acted coordinately to activate OsSOS1 in yeast cells, and suppressed salt sensitivity of Arabidopsis sos2/sos3 mutant, indicating that the SOS pathway for salt tolerance is also operational in cereals (Martinez-Atienza et al. 2007).

Although previous studies have highlighted important roles of SOS pathway in the dicot herbaceous plant Arabidopsis and the monocot plant rice, little is known about the conservation of this pathway in trees. Populus has become an important model in woody plants (Jansson and Douglas 2007), due to its worldwide distribution, sequenced genome and relative ease of genetic manipulation (Tuskan et al. 2006). Poplars have also received increasing attention as a renewable source of biomass for energy and short-fiber pulp for paper making (Luo and Polle 2009; Pilate et al. 2002). Recently, a putative plasma membrane Na+/H+ antiporter PeSOS1 from Populus euphratica, has been isolated and shown to partially suppress the salt-sensitivity of the Escherichia coli mutant strain EP432 (Wu et al. 2007), but analysis of its in planta function is still lacking. In order to better understand the molecular mechanism underlying salt tolerance in woody plants, we isolated three genes coding for key components in the SOS pathway from the sequenced poplar genotype Populus trichocarpa. By molecular analysis and functional characterization of these three PtSOS genes, we show that the perennial woody plant poplar has a conserved SOS pathway in response to salinity stress.

Materials and methods

Plant materials, growth conditions and stress treatment

Two poplar genotypes Populus trichocarpa and Populus alba × P. Berolinensis cv. Yin-Zhong were used in this study. Generally, in vitro grown plants were subcultured every month by aseptically transferring shoot apices to the fresh MS medium (Murashige and Skoog 1962) supplemented with 0.03 mg/L NAA for rooting. Three-week-old micro-propagated plantlets were subjected to salt stress by directly pouring 30 mL of 200 mM NaCl solution into the tissue-culture containers. Plantlets were also transferred into individual pots (volume 6 L) after 3-day acclimation to ex vitro conditions, and grown in the greenhouse under an 14-h photoperiod comprising natural daylight supplemented with lamps (Philips, 600 W), giving a minimum quantum flux density of 150 μEm−2s−1. The temperature was kept at about 21–24°C in the daytime and 15–18°C at night. All the plants were well irrigated according to evaporation demands during different growth stages and watered with half strength of Hoagland nutrient solutions every other week. After 6 months’ growth, the trees would reach a height of approximately 240 cm and a diameter of 1.2 cm. Various tissues were isolated from vigorously growing trees including apical buds, young leaves, mature leaves, elongating internodes, stem segments and roots. Xylem and phloem tissues were simply separated by stripping off the bark with a sharp blade. The samples were frozen in liquid nitrogen immediately and stored at −80°C.

Arabidopsis thaliana plants were grown in the greenhouse under long-day conditions (16 h light/8 h dark) at the temperature 22–23°C. Sterilized seeds were plated on MS medium solidified with 0.8% agar. All of the sos1, sos2 and sos3 mutants in this study were in the Col-0 gl1 background.

Reverse transcription PCR and quantitative real-time PCR analyses

Total RNA was extracted with the RNAiso Reagent (Takara, Japan) from different organs or tissues of two types of poplar plants at the half-year growth stage in the green house. Following the manufacturer’s instruction, RNA samples were obtained with the same reagent from the salt-treated poplar plantlets and 10-day old Arabidopsis seedlings. After being treated with DNase I (Promega), 2 μg of total RNA was subjected to reverse transcription reaction using the reverse transcriptase ReverTra Ace (TOYOBO, Japan) at 42°C for 1 h. The resulting cDNA was then used for PCR amplification with the gene-specific primers listed below. For PtSOS1, the primers S1RT-F (5′-AAAGCCCATGGTTTCCAAGATG-3′), and S1RT-R (5′-CCGC-TTCAAATGCTGCAATATC-3′) were used; For PtSOS2, the primers S2RT-F (5′-AAAGCCCATGGTTTCCAAGATG-3′), and S2RT-R (5′-AGTCATTGTTCGAAGCAGGCAG-3′) were used; For PtSOS3, the primers S3RT-F (5′-TGGAAAAATCGATCCTGACGAGTG-3′), and S3RT-R (5′-GCTGGTTTAAAATGCACGGATCAC-3′) were used; For EF1β, the primers EF1β-F (5′-GACAAGAAGGCAGCGGAGGAGAG-3′), and EF1β-R (5′-CAATGAGGGAATCCACTGACACAAG-3′) were used; For ACTIN2, the primers ACT2-F (5′-GGAAGGATCTGTACGGTAAC-3′) and ACT2-R (5′-GGACCTGCCTCATCATACT-3′) were used.

Quantitative real-time PCR analysis was performed with the RotorGene 3000 system (Corbett Research) using the SYBR Green Realtime PCR Master Mix (TOYOBO, Japan) to monitor double-stranded DNA products. Data analysis was performed with Rotor-Gene software version 6.0 and relative amounts of mRNA were calculated based on the comparative threshold cycle method. The relative expression of each target gene was double-normalized using the housekeeping gene EF1β and using the control expression values measured at 0 h.

Subcellular localization of PtSOS1, PtSOS2 and PtSOS3

To determine the subcellular localization of PtSOS proteins, stop-codon-less coding region of PtSOS1, PtSOS2 or PtSOS3 was in-frame fused upstream to the YFP sequence in the pA7-YFP vector, respectively. The resulting constructs containing PtSOS-YFP translational fusions were transfected into poplar mesophyll protoplasts, essentially as described by Sheen and colleges (Sheen 2001; Yoo et al. 2007). Fluorescence of YFP in the transformed protoplasts was imagined using a confocal laser scanning microscope (LSM510, Carl Zeiss) after the protoplasts were incubated at 23°C for 16 h.

Yeast experiments

Yeast two hybrid assays were based on Matchmaker GAL4 Two-Hybrid System 3 (Clontech). The PtSOS2 coding sequence was excised from the pBluescript II KS construct using EcoRI-XhoI double digestion and subcloned into to pGADT7 vector to generate an in-frame fusion with AD. Similarly, PtSOS3 was inserted into pGBDT7 vector through EcoRI-SalI sites to generate an in-frame fusion with BD. To generate PtSOS2 devoid of the FISL motif, inverse PCR-based mutagenesis was carried out on double-stranded pKS-PtSOS2 plasmid, resulting in pKS-PtSOS2ΔF. Then PtSOS2ΔF was subcloned into pGADT7 vector. These constructs and empty vector controls were transformed into yeast strain AH109 by the PEG/LiAc method. Yeast cells were plated onto SD/-Ade/-His/-Trp/-Leu medium for stringent screening of the possible interactions. β-galactosidase assays were performed as described in the Clontech Yeast Protocols Handbook.

For experiments with the yeast Ras recruitment system (RRS), the p426GPD-RAS vector was used to express RAS fusion baits and the p425GPD vector (Mumberg et al. 1995) was used to express prey proteins. For p426GPD-RAS construction, a Ras fragment without the stop codon was produced by PCR amplification using the pYES-RRS plasmid (Broder et al. 1998) as a template with the primers RAS-F (5′-CCCACTAGTATGACGGAATATAAGCTGG-3′) and RAS-R (5′-GGGGGATCC CTTGCAGCTCATGCAGC-3′), and then cloned into the p426GPD vector via SpeI-BamHI double digestion. Afterwards, the coding sequence of PtSOS2, PtSOSΔF or AtSOS2 was in-frame fused downstream to RAS in the p426GPD-RAS construct between BamHI and SalI sites. PtSOS3 or AtSOS3 was directly inserted into the p425GPD construct. Several combinations of these plasmids and empty vector controls were transferred into the yeast strain cdc25-2 (MATα, ade2, his3, leu2, lys2, trp1, ura3, cdc25-2) and transformed yeast cells were selected on SD/-Ura/-Leu medium. For interaction analysis, serial decimal dilutions of yeast cultures (stating from OD = 1.0) were spotted onto YPD plates at the permissive (24°C) or the restrictive (36°C) temperature. Five randomly selected yeast colonies after transformation were subjected to this assay.

To express SOS genes in yeast, PtSOS1 or AtSOS1 was constructed into the p426PMA vector, a derivant from p426GPD in which the GPD promoter was replaced by the PMA1 promoter (Capieaux et al. 1989). The yeast vector p414GPD (Mumberg et al. 1995) was employed to express PtSOS2 or PtSOS3 alone. To coordinate expression of PtSOS2 and PtSOS3 within a single plasmid, we constructed PtSOS2:CYC1-GPD:PtSOS3 on the backbone of pBluescript II KS plasmid by successively introducing four PCR fragments (the PtSOS2 CDS; the CYC1 terminator; the GPD promoter; the PtSOS3 CDS) into the multiple cloning site (MCS). The quadruple cassette was then subcloned into the p414GPD plasmid via BamHI-XhoI double digestion, resulting in the final construct harboring two expression cassettes in tandem (GPD:PtSOS2:CYC1-GPD:PtSOS3:CYC1). The Saccharomyces cerevisiae strain B31 (MATα, ade2, can1, his3, leu2, trp1, ura3, mall0, Δena1::HIS3::ena4, Δnha1::LEU2, Bañuelos et al. 1998) was transformed with one of or combinations of these expression constructs. Na+ tolerance tests were performed in arginine-phosphate (AP) medium (8 mM phosphoric acid, 10 mM l-Arg, 2 mM MgSO4, 0.2 mM CaCl2, 2% glucose, plus vitamins and trace elements, pH = 6.0). Aliquots (3 μL) were spotted onto AP plates supplemented with 0, 120 and 240 mM NaCl in addition to 1 mM KCl, and grown for 60 h at 30°C.

Bimolecular fluorescence complementation (BiFC) assays

For generation of the BiFC vectors, the coding region of PtSOS2 was subcloned via BamHI-SmaI into pSPYNE-35S, resulting in PtSOS2YFPN; and the coding region of PtSOS3 was cloned via BamHI-XhoI into pSPYCE-35S, resulting in PtSOS3YFPC. Infiltration of Nicotiana benthamiana leaves was performed as previously described by Walter et al. (2004). Protoplasts were prepared 3 days after infiltration by cutting leaf discs into small pieces and incubating for 3 h in the enzyme solution (0.4 M mannitol, 20 mM MES-K, 10 mM CaCl2, 5 mM β-mercaptoethanol, 0.1% BSA, 1% cellulase R10, 0.3% macerozyme R10, pH 5.7). Fluorescence of YFP in the leaf epidermal cells or isolated protoplasts was imagined by a confocal laser scanning microscope (LSM510, Carl Zeiss).

Arabidopsis transformation and complementation test

To prepare constructs for plant transformation, the open reading frame of PtSOS1, PtSOS2 or PtSOS3 was digested from the pBluescript II KS vectors with SmaI and SalI, respectively. To produce the constitutively activated form of PtSOS2, a Thr169-to-Asp mutation was introduced into pKS-PtSOS2 by PCR-based Site-Directed Mutagenesis Kit (Stratagene) using the primers 5′-TTGGACTTCTTCATGACACATGTGGAACCCCG-3′ and 5′-CCCCTTCTGCGGCAACGCACTCAATCC-3′. The yielding pKS-PtSOS2TD plasmid was also digested with BamHI and SalI to release PtSOS2TD coding sequence. All the fragments were then respectively inserted into a modified pCAMBIA-2301 vector downstream of the cauliflower mosaic virus 35S promoter. These plasmids were separately introduced into the Agrobacterium GV3101 strain for Arabidopsis transformation. Putative transgenic plants were screened on MS medium containing 50 μg/L kanamycin and transferred to soil for propagations. Kanamycin-resistant plants of T2 generation were selected and subjected to transgene expression analyses. Seeds of homozygous T2 lines were harvested and the T3 generation was employed for complementation tests on MS agar plates supplemented with NaCl, as indicted for each case.


Isolation of PtSOS1, PtSOS2 and PtSOS3 from Populus

We performed homology-based BLAST searches (Altschul et al. 1997) to collect candidate gene sequences coding for poplar SOS pathway components from the Joint Genome Initiative poplar database (JGI, Populus trichocarpa genome portal v1.1; 1_1/Poptr1_1.home.html). Two SOS1 (with 89% identity), SOS2 (with 91% identity) and SOS3 (with 94% identity) homologues were identified from the whole Populus genome. Due to high similarity within individual Populus SOS candidates, we chose only one gene of each as a representative, with the closest homology to Arabidopsis SOS counterpart, for subsequent studies. These three genes were designated as PtSOS1, PtSOS2 and PtSOS3. Total RNA was extracted from salt-treated plantlets of Populus trichocarpa and first-strand cDNA was used to amplify the coding region of each gene on the basis of public genomic sequence. The amplified fragments were subcloned to pBluescript II KS vector and then sequenced independently three times to confirm proper removal of introns and fidelity of the polymerase. After manual revision of the isolated CDS and its coding protein, sequence alignment was conducted between PtSOS1, 2, 3 and AtSOS1, 2, 3 respectively (Figs. S1, S2, S3). The three putative SOS proteins in poplar bear significant resemblance to their Arabidopsis counterparts in size and structure.

The poplar SOS components show overlapping but distinct expression pattern

As the first step to characterize SOS pathway in woody plants, expression of poplar SOS genes was analyzed to clarify whether the three components co-localize and thereby co-function in planta. We performed RT–PCR to determine relative transcript abundance in various poplar tissues from two genotypes in our lab, i.e. the sequenced genotype Populus trichocarpa and an aspen genotype Populus alba × P. Berolinensis (a common cultivar in the North of China). The final amount of EF1β products verified that all samples contained approximately equal amount of cDNA. In Populus trichocarpa, PtSOS1 and PtSOS2 exhibited generally high, uniform and ubiquitous expression in all tissues (Fig. 1a). Strong expression of PtSOS1 was detected in root, juvenile leaf, stem, xylem and apex while expression of PtSOS2 might be slightly higher in young green tissues such as juvenile leaf and apex than in other organs. The transcript of PtSOS3 was low in most poplar tissues but relatively abundant in root. In Populus alba × P. Berolinensis, a very similar expression pattern was observed with regard to PabSOS1 and PabSOS2, while preferential expression of PabSOS3 in root tissues was even more evident (Fig. 1b). In conclusion, the poplar SOS components show an overlapping but also a distinct expression pattern, which may be analogous among different Populus species.
Fig. 1

Expression analysis of the SOS1, 2, 3 genes in different tissues of Populus trichocarpa (a) and Populus alba × P. Berolinensis (b). cDNA derived from the indicated tissues was used as a template for PCR with the gene specific primers at different numbers of PCR cycles. Amplification of the genomic DNA sample as a control template was also carried out to rule out its possible contamination in each cDNA sample. PCR products were visualized by agarose gel electrophoresis and ethidium bromide staining. The elongation factor gene EF1β was employed as an internal control

Expression of SOS genes in the poplar is regulated by salinity stress

To further examine whether the expression of poplar SOS components is modified by salinity stress, 3-week-old micro-propagated poplar plantlets were subjected to 200 mM NaCl for increasing exposure periods and the transcript level of SOS1, SOS2 and SOS3 was analyzed by quantitative real-time PCR. Since the sequenced genotype Populus trichocarpa is extremely susceptible to salinity and became wilted during our treatment period, we employed Populus alba × P. Berolinensis in this test as our experimental materials. In roots, PabSOS1 was rapidly induced after the onset of salt stress and increased up to the maximal level (around fivefold) at the twelfth hour during the NaCl treatment. Expression of PabSOS2 coordinately increased within 3 h, reached a maximal threefold to fourfold induction between 6 and 12 h, and thereafter declined. In contrast, expression of PabSOS3 in roots was hardly responsive to salinity stress during our experiment (Fig. 2a). In shoot tissues (Fig. 2b), transcripts of both PabSOS1 and PabSOS2 began to accumulate within the first 3 h of NaCl treatment and retained continuously higher expression than basal level from 6 to 24 h. However, the transcript level of PabSOS3 in above-ground tissues was reduced almost by half upon 1-h salt treatment and recovered to the near-basal level after 12 h of salt treatment. These results indicate that SOS1 and SOS2, but not SOS3, are salt responsive genes in the poplar SOS pathway.
Fig. 2

Quantitative real-time PCR analyses of poplar SOS components in response to salt stress. Three-week-old poplar plantlets grown on MS medium were treated with 200 mM NaCl. Samples were collected at 0, 1, 3, 6, 12 and 24 h after the initiation of treatment. Total RNA was extracted from roots a and shoots b respectively, and real-time PCR analyses were performed with SOS1, 2 and 3 gene specific primers. The relative expression of each target gene was double-normalized using the housekeeping gene EF1β and using the control expression values measured at 0 h. Data represents the average of three independent experiments ±SD

Subcellular localization of PtSOS1, PtSOS2 and PtSOS3 in poplar cells

Proteins that participate in the same pathway are supposed to occupy overlapped locations within the cell. In order to investigate subcellular localization of the Populus SOS components in plant cells, we fused PtSOS1, PtSOS2 and PtSOS3 in-frame to the N-terminus of yellow fluorescent protein (YFP), respectively. The chimeric construct was delivered into mesophyll protoplasts prepared from poplar leaf tissues by the polyethylene glycol transformation method (Sheen 2001; Yoo et al. 2007). Each fusion protein was transiently expressed under the control of the constitutive 35S promoter and its subcellular localization was visualized by confocal microscopy. As expected in the case of the positive control YFP alone, a diffuse green fluorescence typical of cytosolic location was observed with a tendency to accumulate in the nucleus and near the plasma membrane (Fig. 3, top panels). When fused to YFP, PtSOS1 was exclusively localized in the plasma membrane of poplar protoplasts (Fig. 3, second panels from top), consistent with its identity as a Na+/H+ exchanger at the cell membrane. PtSOS2-YFP, however, was dispersed throughout the cytoplasm, within the nucleus and also at the cell periphery, quite similar to YFP alone (Fig. 3, third panels from top). PtSOS3-YFP was predominantly targeted to the plasma membrane with leaky intracellular localization (Fig. 3, bottom panels). These observations are in favor of the physical possibility that PtSOS1, PtSOS2 and PtSOS3 could be involved in the same pathway at the plasma membrane of poplar cells.
Fig. 3

Subcellular localization of PtSOS1-, 2- and 3-YFP fusion proteins. Confocal microscopy analysis of YFP signals from poplar mesophyll protoplasts transiently expressing YFP or PtSOS-YFP fusions as indicated. Scale bar = 5 μm

Populus SOS proteins are functional homologues of their Arabidopsis counterparts

The Arabidopsis mutants in the SOS pathway, i.e. sos1, sos2 and sos3, are specifically hypersensitive to high external Na+ concentrations. To dissect the biological function of PtSOS proteins, we constructed transgenic Arabidopsis lines heterologously expressing each PtSOS gene in the corresponding sos mutant background for functional complementation tests. We selected four independent complemented lines from rescued sos1, sos2 and sos3 mutants, respectively. By RT–PCR analysis, we confirmed the expression of PtSOS1 transgene in sos1, PtSOS2 transgene in sos2, and PtSOS3 transgene in sos3 (Fig. 4a). Preliminary assays on salt tolerance of these transgenic lines for each gene showed that they had similar salt tolerance capacity, and therefore two representative transgenic lines (CS12, CS16; CS26, CS29; CS35, CS36) of each gene were chosen for subsequent phenotypic analyses. In the absence of NaCl treatment, wild type plants, sos mutants and PtSOS complemented lines displayed relatively uniform growth phenotype, although roots of the seedlings in sos1 background were slightly shorter (Fig. 4b, top panels). When treated with 60 mM NaCl, growth of sos1, sos2 and sos3 mutants were all inhibited compared with wild type plants, albeit to different extent. Expression of PtSOS1 in sos1, PtSOS2 in sos2 and PtSOS3 in sos3 substantially alleviate the growth retardation imposed by moderate salt stress (Fig. 4b, middle panels). Under 120 mM NaCl treatment, growth inhibition of sos2 and sos3 mutants was even more pronounced. Seedlings of sos1 were killed by this high concentration of salt stress. Nevertheless, the sos1 mutants expressing PtSOS1 could survive at this lethal NaCl level. The salt-sensitive phenotype was also greatly suppressed in PtSOS2- and PtSOS3-complemented lines (Fig. 4b, bottom panels). A detailed analysis of root length and seedling fresh weight showed that all the rescued transgenic plants by PtSOS genes were generally restored to a near-wild-type level of salt tolerance, although with differential extents between different lines (Fig. 4c–h). These data indicate the poplar SOS proteins can substitute for the corresponding Arabidopsis SOS components, thereby being functional orthologues of their Arabidopsis counterparts.
Fig. 4

Functional complementation of Arabidopsis sos1, sos2 and sos3 mutants by poplar PtSOS1, PtSOS2 and PtSOS3, respectively. a Expression of PtSOS1, 2, 3 in each sos mutant as indicated. RT–PCR products were amplified from seedlings of wild type, sos mutants and four corresponding complementary lines. The Arabidopsis ACTIN2 gene was used as an internal control. b Four-day-old seedlings of wild-type, sos mutants and complementary lines CS12, CS16, CS26, CS29, CS35, CS36 were transferred to MS medium containing 0, 60 and 120 mM NaCl, respectively. The pictures were taken after 8 days of treatment. Seedling fresh weight ce and root length fh were measured at day 8 after the transfer. Error bars represent SD (n = 10)

PtSOS3 interacts with and recruits PtSOS2 to the plasma membrane

In Arabidopsis, SOS2 interacted with SOS3 to form a kinase complex for proper biological functions (Halfter et al. 2000; Gong et al. 2004). Here we determined the physical interaction between PtSOS2 and PtSOS3 by yeast two hybrid assays. A combination of the bait plasmid pGBKT7-PtSOS3 with the prey construct pGADT7-SOS2 resulted in full transcriptional activation of HIS, ADE and lacZ reporters, whereas co-transformation of either empty vector with the construct pGBKT7-PtSOS3 or pGADT7-SOS2 did not activate any of the reporter genes. The FISL motif (also known as NAF domain) in the SOS2 regulatory region is necessary and sufficient for SOS3 interaction (Guo et al. 2001; Albrecht et al. 2001). Deletion of this conserved FISL motif away from PtSOS2 protein (PtSOS2ΔF) abolished its interaction with PtSOS3 (Fig. 5a), suggesting that PtSOS2 also interacts with PtSOS3 via the FISL motif.
Fig. 5

Interaction between PtSOS2 and PtSOS3 in yeast and plant cells. (a) Yeast strain AH109 harboring BD-PtSOS3 fusion in pGBKT7 vector (or pGBKT7 empty vector) was transformed with AD-PtSOS2, PtSOS2ΔF or PtSOS2TD fusion in pGADT7 vector (or pGADT7 empty vector). The interaction was monitored by the ability of yeast transformants to grow on selective medium minus Leu, Trp, His and Ade, or by β-galactosidase assay on a piece of filter paper. (b) Recruitment of PtSOS2 to the plasma membrane by PtSOS3. The yeast cdc25-2 mutant was transformed with the plasmid combinations as indicated on the left. Three microliters of serial decimal dilutions were spotted onto YPD plates at the permissive (24°C) or at the restrictive (36°C) temperature. Cells expressing RAS-PtSOS2 in combination with PtSOS3 or RAS-AtSOS2 with AtSOS3 combination were able to grow at 36°C, indicating interaction between the two proteins at the plasma membrane. (c) PtSOS2 interacted with PtSOS3 in planta. Empty YFP construct (as a positive control) or different combinations of the N-YFP and C-YFP fusion constructs as indicated were co-transfected into the tobacco (N. benthamiana) leaves. The images were collected under the Zeiss confocal microscope. The left panels depict the fluorescence in intact leaf epidermal cells (together with bright-field images on the right), and the right panels exhibit the fluorescence signals in protoplasts prepared from the same leaf (together with bright-field images on the right)

To further address spatial specificity of PtSOS2–PtSOS3 complex formation, we employed the Ras recruitment system (RRS) to monitor membrane targeting of PtSOS2 by PtSOS3 interaction. This system is based on the ability of a mammalian Ras variant to be localized to the plasma membrane through the interaction between two proteins. Ras membrane localization via protein–protein interaction results in complementation of the yeast thermo-sensitive mutant cdc25-2 deficient in the Ras guanyl nucleotide exchange factor (Broder et al. 1998). As shown in Fig. 6b, expression of RAS-PtSOS2 fusion alone without PtSOS3 failed to restore yeast growth at the restrictive temperature (36°C), indicating a cytosolic localization. However, the RAS-SOS2 fusion was efficiently recruited to the plasma membrane upon co-expression of PtSOS3, allowing cdc25-2 strain to grow at 36°C. RAS fusion with the PtSOS2 mutant lacking FISL motif disrupted PtSOS3 interaction and thus did not restore the thermo-tolerance in the presence of PtSOS3. In conclusion, PtSOS3 recruited PtSOS2 to the plasma membrane through their physical interaction.
Fig. 6

Reconstitution of poplar SOS pathway in yeast cells. The yeast B31 (Δena1-4, Δnha1) strain transformed with an empty vector or expressing the indicated combination of Populus SOS genes were grown overnight in liquid selective medium. Three microliters of serial decimal dilutions were spotted onto plates of arginine phosphate (AP) medium with 1 mM KCl and supplemented with 0, 120 and 240 mM NaCl. Plates were incubated at 30°C for 60 h. The Arabidopsis SOS1 gene AtSOS1 was used as a positive control. Plasmids used for expression of SOS genes were described in “Materials and methods

We also investigated the interaction between PtSOS2 and PtSOS3 in planta using bimolecular fluorescence complementation (BiFC) assays (Walter et al. 2004). The BiFC results shown in Fig. 5c not only confirmed the in vivo PtSOS2-PtSOS3 interaction but also clearly reinforced the evidence that specific interaction of the two proteins take place at or near the plasma membrane of plant cells.

Reconstitution of the poplar SOS pathway in yeast cells

In yeast Saccharomyces cerevisiae, the Na+-pumping ATPase ENA1-4 and the Na+/H+ antiporter NHA1 are two major sodium efflux proteins at the plasma membrane responsible for most of the sodium extrusion. The B31 strain devoid of these two Na+ transporters was defective in sodium efflux and extremely sensitive to high external Na+ concentrations (Bañuelos et al. 1998; Kinclova et al. 2001). To test whether the putative Na+/H+ antiporter PtSOS1 could be activated by PtSOS2–PtSOS3 complex, we reconstituted poplar SOS regulatory pathway in yeast cells by expressing different combinations of PtSOS proteins in the B31 background. As depicted in Fig. 6, expression of PtSOS1 alone only marginally elevated halotolerance of yeast cells. Additional expression of PtSOS2, but not PtSOS3, enabled the yeast strain to survive better under moderate salinity stress (120 mM NaCl), above levels imparted by SOS1 alone, but still failed to afford yeast growth in high saline condition (240 mM NaCl). Concurrent expression of three poplar SOS components PtSOS1, PtSOS2 and PtSOS3 dramatically augmented the salt tolerance of yeast cells, capable to sustain growth in AP medium containing up to 240 mM NaCl. This enhancement in salt tolerance was observed only in the presence of PtSOS1, excluding the possibility that the PtSOS2–PtSOS3 complex was unmasking endogenous yeast activity. PtSOS2–PtSOS3 complex was also able to activate AtSOS1 (as a positive control in our experiment), with even greater efficacy, suggesting that the regulation of SOS1 activity might be evolutionally conserved between woody and herbaceous plant species.

Expression of Constitutively Active PtSOS2 (PtSOS2TD) complemented salt-sensitive phenotype of sos2 mutant entirely, and that of sos3 partially, but did not rescue salt hypersensitivity of sos1 at all.

Physical interaction and yeast reconstitution experiments in this study have demonstrated that the putative poplar SOS components should be working in the same pathway as those in Arabidopsis. To further consolidate this notion in planta, we tested whether PtSOS2 functions dependent of SOS1 and downstream of SOS3 in salt tolerance. We generated a constitutively active form of PtSOS2 (designated as PtSOS2TD) by mutating the 169th amino acid threonine (T) to aspirate (D) in the activation loop of this kinase (Fig. S2). This single-point mutation in SOS2 substantially increased the kinase activity (Guo et al. 2001) but did not affect its interaction with SOS3 (Fig. 5a), resulting in a constitutively activated and functional variant of SOS2 both in vitro and in vivo (Guo et al. 2004). The Arabidopsis sos1, sos2 and sos3 mutants were transformed with PtSOS2TD under the control of cauliflower mosaic virus 35S promoter. By RT–PCR assay using PtSOS2 specific primers, we identified 10–15 transgenic lines overexpressing PtSOS2TD in sos1, sos2 or sos3 mutant background and selected two representative lines of each exhibiting similar expression levels for subsequent phenotypic analysis (Fig. 7a). Expression of PtSOS2TD in sos1 background failed to rescue its salt overly sensitive phenotype, as is shown in Fig. 7b. The transgenic sos1 plants and sos1 parental mutants displayed equally severe growth defect on MS plates supplemented with 120 mM NaCl (Fig. 7c, f). However, transgenic sos2 plants harboring PtSOS2TD grew like the wild-type plants upon high salt treatment (Fig. 7b, middle panels). A statistical analysis of root length (Fig. 7d) and fresh weight (Fig. 7g) further indicated that salt hypersensitivity of sos2 was fully rescued by constitutive expression of the poplar activated form of SOS2. When expressed in sos3 background, PtSOS2TD partially suppressed the salt-sensitive phenotype of sos3 mutants (Fig. 7b, right panels). The sos3 transformants harboring PtSOS2TD grew bigger than sos3 mutants on saline medium, as revealed by partially recovered fresh weight (Fig. 7h) and less inhibited root length (Fig. 7e). Taken together, the enhanced activity or functionality of PtSOS2 is not sufficient for salt tolerance in the absence of SOS1 but somewhat bypass the SOS3 deficiency in planta. By virtue of strong conservation between Populus and Arabidopsis SOS proteins, the functional complementation tests here could imply that in poplar trees, PtSOS2 might function downstream of PtSOS3 and upstream of PtSOS1, as well.
Fig. 7

Expression of a constitutively activated poplar SOS2 version (PtSOS2TD) in Arabidopsis sos mutants. a Transgene expression of PtSOS2TD (the 169 amino acid Thr was mutated to Asp) in sos1 (LS11, LS12), sos2 (LS21, LS22) and sos3 (LS31, LS32) background. The Arabidopsis ACTIN2 gene was used as an internal control. b Four-day-old seedlings of wild-type, sos mutants and selected transgenic lines in different sos mutant background were transferred to MS medium containing 0 or 120 mM NaCl, respectively. The pictures were taken after 10 days of treatment. Seedling fresh weight ce and root length fh were measured at day 10 after the transfer. Error bars represent SD (n = 10)


Adaptation to high salinity is one of the important concerns for woody perennials during their lifespan and over evolutionary timescales. The molecular mechanisms underlying salt tolerance and salt stress signaling are poorly known in woody plants. With the completion of genome sequencing in Populus trichocarpa (Tuskan et al. 2006) and rapidly developing genomic tools, a handful of genes related to salt stress tolerance (Ottow et al. 2005; Wu et al. 2007; Wang et al. 2008; Martin et al. 2009; Chen et al. 2009; Ye et al. 2009) were isolated and characterized from the genus Populus that has become the accepted model for perennial trees (Jansson and Douglas 2007). However, most of those studies were focused on the biological function of individual genes, without interpretation of a gene regulatory network leading to fine-tuning of adaptive processes. In this report, for the first time, we characterized the Populus SOS pathway by integrated analysis of three component genes. Several lines of molecular evidence support the conclusion that poplar plants have a functional salt overly sensitive pathway orthologous to the well-documented Arabidopsis SOS pathway. First, the putative PtSOS components share high similarity with AtSOS proteins in structure implying their similar biochemical properties. Second, the overlap in expression and subcellular localization of three poplar SOS elements provides a prerequisite for their coordinated actions in the same pathway governing salt stress adaptation. Third, the poplar SOS orthologues can substitute for the endogenous Arabidopsis counterparts revealed by functional complementation tests in sos mutants. Above all, PtSOS3 recruited PtSOS2 to the plasma membrane via direct physical interaction and the PtSOS2–PtSOS3 complex activates PtSOS1 in yeast cells.

In Populus, although the expression of poplar SOS genes is generally ubiquitous and largely overlapping, each gene may exhibit a distinct expression pattern to some extent. For example, the expression of PtSOS1 in root and in shoot was comparable, whereas PtSOS2 was preferentially expressed in green tissues rather than roots (Fig. 1). In Populus alba × P. Berolinensis, the expression of PabSOS1 and PabSOS2 was rapidly induced upon salt treatment, with a differential manner in root and in shoot (Fig. 2). It seemed that the salt-induced up-regulation of PabSOS1 and PabSOS2 had a long-term effect in shoots with moderate levels, but was more responsive and transient in roots. The simultaneous induction of PabSOS1 and PabSOS2 transcripts by salinity also suggested a fine-tuned regulation of SOS elements at the gene expression level. However, the expression of PabSOS3, the upstream component of SOS pathway, was not responsive to salt treatment in our test, either in root (Fig. 2a) or in shoot (Fig. 2b), presumably because SOS3 senses calcium signals and triggers downstream signaling primarily at protein modification but not the transcription level. We also noticed that the transcript abundance of PtSOS3 was pretty high in root but remarkably diminished in the aerial parts (Fig. 1). This is consistent with the root-specific expression of AtSOS3 in Arabidopsis (Quan et al. 2007). On the other hand, the SOS3-like calcium binding protein 8 (SCaBP8)/CBL10 was recently identified as a shoot-specific player in the SOS pathway (Quan et al. 2007). Like SOS3, it recruited SOS2 to the plasma membrane, but uniquely underwent phosphorylation by SOS2, leading to stabilized SCaBP8–SOS2 complex for SOS1 activation (Lin et al. 2009). Interestingly, we found a SCaBP8 homolog existing in Populus genome to be predominantly expressed in the above-ground tissues (unpublished data). This fact may implicate a tissue-differential regulation of salt response in the poplar SOS pathway.

The sodium efflux protein SOS1 functions as a pivotal effector molecule in the SOS pathway, and it has been suggested to fulfill several facets of important roles pertaining to salt tolerance. To begin with, at the cellular level, SOS1 would act extruding the excessive Na+ to prevent Na+ accumulation in the cytoplasm, as it has been shown in Arabidopsis (Qiu et al. 2003), rice (Martinez-Atienza et al. 2007), wheat (Xu et al. 2008), Thellungiella (Oh et al. 2009) and tomato (Olías et al. 2009). In Populus, the salt-tolerant genotype Populus euphratica exhibited a higher transcript abundance of genes related to Na+/H+ antiport compared to the salt-sensitive P. popularis (Ding et al. 2010). In a recent study based on the scanning ion-selective electrode technique, it was concluded that Na+ extrusion in stressed Populus euphratica roots was the result of an active Na+/H+ antiport across the plasma membrane (Sun et al. 2009), which might be mediated by PeSOS1 reported previously (Wu et al. 2007). In addition, at the whole-plant level, it was suggested that SOS1 controls long-distance Na+ transport (Shi et al. 2002) and affects Na+ partitioning between plant organs (Olías et al. 2009). These Na+ distribution processes would be particularly important for woody poplars because of their big stature and long lifespan. A precise assessment of possible contributions of PtSOS1 in this aspect awaits the availability of RNAi lines and/or overexpressors of transgenic poplars. Furthermore, SOS1 takes part in the cross-talk between the ion-homesotasis and oxidative-stress signaling pathways. It was reported that stress-induced SOS1 mRNA stability in Arabidopsis was mediated by reactive oxygen species (ROS) and SOS1 played negative roles in tolerance to oxidative stress (Chung et al. 2008). SOS1 interacted with RCD1, a regulator of radical-based signaling, through the long C-terminal cytoplasmic tail, further suggesting its involvement in regulating oxidative-stress responses (Katiyar-Agarwal et al. 2006). In NaCl-stressed Populus euphratica cells, H2O2 production and cytosolic Ca2+ signals were integrated to mediate K+/Na+ homeostasis (Sun et al. 2010; Zhang et al. 2007). It needs further experimentation how the poplar SOS pathway functions in such signaling cross-talk. Last but not least, by affecting endocytosis under salinity conditions, SOS1 may also play dynamic roles in vacuole integrity, membrane trafficking, as well as pH homeostasis (Oh et al. 2009; Oh et al. 2010). Further work on poplar SOS genes will be instructed by and focused on all the above findings so as to elucidate functional significance of SOS pathway in woody species. On the other hand, engineering PtSOS genes identified in this study offer good opportunities towards generation of transgenic tree species with improved salt tolerance, which may be grown on saline land to prevent erosion and potentially to ameliorate soils (Singh 1998). At present, genetic transformation of poplar cultivars by overexpressing the three PtSOS components is under way in our laboratory.


We thank Dr. Jörg Kudla (Institut für Botanik, Universität Münster, Germany) for providing the pSPYNE-35S and pSPYCE-35S plasmids. We are also grateful to Dr. Ami Aronheim for RRS materials and Dr. Lydie Maresova for the B31 yeast strain. Special thanks would go to Dr. Rongmin Zhao (University of Toronto, Canada) for his technical guidance on yeast experiments. This work was supported by the following grants: the National Natural Science Foundation of China (30872044), the National Basic Research Program of China (2006CB100106), the National mega project of GMO crops (2008ZX08001-003; 2008ZX08004-002; 2008ZX08010-004); Shanghai Science & Technology Development Fund (0853Z111C1; 08DZ2270800).

Supplementary material

11103_2010_9680_MOESM1_ESM.tif (3.5 mb)
Fig. S1 Protein sequence alignment of AtSOS1 and PtSOS1. The underlined putative transmembrane domains were predicted according to TMHMM2 program ( (TIFF 3597 kb)
11103_2010_9680_MOESM2_ESM.tif (81 kb)
Fig. S2 Protein sequence alignment of AtSOS2 and PtSOS2. FISL motif was underlined. The star positions the point mutation made in this study. The amino acid threonine (T) 169 was mutated to aspirate (D) in the activation loop of PtSOS2. (TIFF 81 kb)
11103_2010_9680_MOESM3_ESM.tif (54 kb)
Fig. S3 Protein sequence alignment of AtSOS3 and PtSOS3. Four EF-hand motifs for calcium binding are underlined. (TIFF 53 kb)

Copyright information

© Springer Science+Business Media B.V. 2010