Ectopic Expression of GsPPCK3 and SCMRP in Medicago sativa Enhances Plant Alkaline Stress Tolerance and Methionine Content

So far, it has been suggested that phosphoenolpyruvate carboxylases (PEPCs) and PEPC kinases (PPCKs) fulfill several important non-photosynthetic functions. However, the biological functions of soybean PPCKs, especially in alkali stress response, are not yet well known. In previous studies, we constructed a Glycine soja transcriptional profile, and identified three PPCK genes (GsPPCK1, GsPPCK2 and GsPPCK3) as potential alkali stress responsive genes. In this study, we confirmed the induced expression of GsPPCK3 under alkali stress and investigated its tissue expression specificity by using quantitative real-time PCR analysis. Then we ectopically expressed GsPPCK3 in Medicago sativa and found that GsPPCK3 overexpression improved plant alkali tolerance, as evidenced by lower levels of relative ion leakage and MDA content and higher levels of chlorophyll content and root activity. In this respect, we further co-transformed the GsPPCK3 and SCMRP genes into alfalfa, and demonstrated the increased alkali tolerance of GsPPCK3-SCMRP transgenic lines. Further investigation revealed that GsPPCK3-SCMRP co-overexpression promoted the PEPC activity, net photosynthetic rate and citric acid content of transgenic alfalfa under alkali stress. Moreover, we also observed the up-regulated expression of PEPC, CS (citrate synthase), H+-ATPase and NADP-ME genes in GsPPCK3-SCMRP transgenic alfalfa under alkali stress. As expected, we demonstrated that GsPPCK3-SCMRP transgenic lines displayed higher methionine content than wild type alfalfa. Taken together, results presented in this study supported the positive role of GsPPCK3 in plant response to alkali stress, and provided an effective way to simultaneously improve plant alkaline tolerance and methionine content, at least in legume crops.


Introduction
As a versatile crop, alfalfa (Medicago sativa L.) is used for pasture, hay, silage and green-chop, and acts in crop rotation through its positive effects on soil fertility and soil structure [1]. Due to its versatility, high productivity, high feed value and potential roles in soil improvement and soil conservation, alfalfa is grown over a wide range of climatic conditions [1,2]. However, environmental challenges, especially soil salinity and alkalinity, not only seriously restrict alfalfa yield, but also affect nodules formation and symbiotic nitrogen-fixation capacity [3]. With the global climate change and the global shrinkage of arable lands, a grimmer reality of soil salinity and alkalinity is painted, hence more and more attentions have been paid to exploring alfalfa cultivation on marginal lands.
As reported earlier, of the 831 million hectares salt-alkali soils on the world, saline and alkaline soils underline 397 (47%) and 434 (53%) million hectares, respectively [4,5]. What's worse, in northeast China, over 70% of land area is alkaline grassland [6]. Alkaline soil is characterized by high pH, high exchangeable sodium, poor fertility, dispersed physical properties and low water content [7]. Recently, a handful of researches suggested that, compared with salt stress, alkaline stress always caused much stronger inhibition of plant growth and development [8]. Unfortunately, until now, little attention has been paid on the molecular mechanisms of plant adaptation to alkaline stress.
Phosphoenolpyruvate carboxylase (PEPC; EC4.1.1.31) is a kind of tightly controlled cytosolic enzyme which functions in carbon fixation during photosynthesis [9][10][11]. PEPC kinase (PPCK) controls the phosphorylation state and bioactivity of PEPCs. In recent years, PPCK genes have been identified in different higher plants, including two for Arabidopsis [12], three for rice [13], and four for soybean [14]. Recently, several lines of direct evidence supported that PEPCs and PPCKs fulfilled important nonphotosynthetic functions, particularly in plant response to environmental challenges. One of the best studied examples was that salt stress remarkably increased PPCK activity [15][16][17]. Further investigation revealed that salt stress not only increased PPCK gene expression levels but also decreased PPCK protein degradation rates [17]. Moreover, PEPC/PPCK activity in Arabidopsis and poplar was also regulated by alkali stress [12,18].
Besides high stress tolerance, another important desired trait for alfalfa is high nutritional value in terms of essential amino acids. As we know, legume plants are deficient in the sulfur-containing amino acids, namely, methionine and cysteine [19]. Methionine is nutritionally essential for mammals, but at low levels in legume. Unlike plants, mammals could not synthesize methionine; hence they have to obtain it from their diets. In this respect, methionine deficiency obviously limited the nutritional value of legumes; therefore, increasing the methionine content has become another important goal for legumes breeding. To solve this problem, in a previous study, we designed and synthesized the SCMRP gene according to the maize methionine-rich 10 ku zein protein [20,21]. A similar study reported that overexpression of the maize 10 ku zein gene increased the sulphur-containing amino acids of transgenic potato [22].
For these reasons, in this study, we aimed to generate the transgenic alfalfa not only with higher alkali tolerance but also with higher methionine content. Based on the Glycine soja (G07256) microarray data, we isolated and characterized an alkali stress responsive gene GsPPCK3. Our results demonstrated that GsPPCK3 overexpression in alfalfa improved plant tolerance to alkali stress. In this case, we co-transformed the GsPPCK3 and SCMRP genes into alfalfa, and demonstrated that the transgenic alfalfa displayed not only higher alkali tolerance but also higher methionine content. Taken together, results presented in this study demonstrated the biological function of GsPPCK3 under alkali stress, and provided an effective way to simultaneously improve plant alkaline tolerance and methionine content, at least in legume crops.

Isolation and Sequence Analysis of the GsPPCK3 Gene
In a previous study, we constructed a transcriptional profile of Glycine soja (G07256) roots and leaves in response to salt-alkali stress (50 mM NaHCO 3 , pH 8.5), by using the AffymetrixH Soybean Genome Array [23]. Three PEPC kinase genes GsPPCK1, GsPPCK2 and GsPPCK3 were identified as potential stress responsive genes (Fig. 1). Contrarily, the last PPCK gene GsPPCK4 essentially did not respond to salt-alkali stress, only with an increase at 6 h in leaves.
In this study, we obtained the full length coding region of GsPPCK3 by using homologous cloning strategy. GsPPCK3 contained a complete open reading frame (ORF) of 915 bp encoding 304 amino acids with an estimated molecular weight (Mr) of 34 000 and a theoretical pI of 5.35. A BLASTP search at NCBI showed that GsPPCK3 shared 90%, 84%, 69% and 61% sequence identity with the Glycine max GmPPCK3 (GeneBank Accession: NP_001238645), GmPPCK2 (NP_001236660), GmPPCK1 (NP_001241581) and GmPPCK4 (NP_001237016), respectively. It is worth noting that GsPPCK3 contained a 30 amino acid length sequence (KLLLASVFLFDIIFGGFCVD-GIFGCFVFVG) at its C-terminus, which did not exist in any soybean and Arabidopsis PPCKs. By comparing the DNA and mRNA sequences of GsPPCK3 and GmPPCK3, we found that this 30 aa sequence was encoded by the 90 bp length intron of GmPPCK3.
In order to get better understanding of GsPPCK3 structure, we compared the amino acid sequences of GsPPCK3 with PPCK homologs from Glycine max and Arabidopsis thaliana (AtPPCK1 and AtPPCK2) (Fig. 2). Protein sequence analysis revealed that GsPPCK3 comprised a minimal Ser/Thr kinase domain, closely related to the catalytic domain of plant calcium-dependent protein kinases (CDPKs) without the N-and C-terminal extensions.

Expression Patterns of GsPPCK3 in Response to Alkaline Stress and in Different Tissues
In an attempt to verify the induced expression of GsPPCK3 under alkali stress, we carried out the quantitative real-time PCR analyses. Consistent with the microarray data, GsPPCK3 expression in both leaves and roots of Glycine soja seedlings was greatly and rapidly induced by alkali stress (Fig. 3A). After 50 mM NaHCO 3 (pH8.5) treatment, GsPPCK3 displayed an obvious increase and reached a maximum point at 1 h. It is noteworthy that the induction degree of GsPPCK3 in roots was significantly higher than that in leaves. The possible reasons for the stronger response might be that plant roots were the exact sites of perception and injury for stresses, or there were different response mechanisms for GsPPCK3 expression between roots and leaves. Anyway, it is obvious that GsPPCK3 expression was greatly induced by alkali stress, suggesting an important role of GsPPCK3 in plant response to alkali stress.
In order to get better understanding of GsPPCK3 expression patterns, we further investigated the spatial specific expression of GsPPCK3 in Glycine soja seedlings (Fig. 3B). The real-time PCR results showed that, among the eight tissues detected in this study, GsPPCK3 displayed the highest expression level in leaves, which could be explained by the important role of GsPPCK3 in photosynthesis. Contrarily, the GsPPCK3 transcript level in roots was relatively lower, even though it exhibited a greater degree of alkali stress induction in roots. This difference led us to propose the possibility of different mechanisms and different roles of GsPPCK3 between roots and leaves in plant response to alkali stress.

GsPPCK3 Overexpression Confers Enhanced Alkaline Tolerance in Transgenic Alfalfa
The stress induced expression of GsPPCK3 gave us an insight into its potential role in response to alkali stress. To precisely investigate the biological and physiological function of GsPPCK3 under alkaline stress, we transformed GsPPCK3 into the wild type (WT) M. Sativa through the Agrobacterium tumefaciens-mediated transformation strategy ( Figure S1A). By using PCR and semiquantitative RT-PCR analyses, we identified a total of nine transgenic lines (Fig. 4A), and two of them, with relatively higher expression levels (3-16 and 3-26), were selected to examine the response to alkaline stress.
Under control conditions, transgenic lines showed no obvious differences in growth performance compared with WT. After 50 or 100 mM NaHCO 3 treatment for 14 d, both the WT and transgenic lines showed growth retardation in a dose-dependent manner. However, the growth inhibition of transgenic lines was less severe than that of WT (Fig. 4B). In details, WT plants exhibited severe chlorosis and even death, whereas the transgenic lines maintained continuous growth after 100 mM NaHCO 3 treatment.
We further compared the relative membrane permeability ( 4F) of WT and GsPPCK3 transgenic lines, respectively. No obvious differences were observed between WT and transgenic lines under control condition. As expected, alkali stress significantly increased the relative membrane permeability and MDA content, but decreased the total chlorohpyll content and root activity of both WT and transgenic plants. However, the transgenic lines showed relatively lower levels of electrolyte leakage and MDA content (Fig. 4C, D), but higher levels of total chlorohpyll content and root activity than WT (Fig. 4E, F) (*P, 0.05; **P,0.01 by Student's t-test). Taken together, these results strongly suggested that GsPPCK3 overexpression improved the alkali tolerance of transgenic alfalfa.

Increased Alkaline Tolerance of GsPPCK3-SCMRP Overexpression Transgenic Alfalfa
Considering the deficiency in sulfur-containing amino acids, we then co-transformed the GsPPCK3 and SCMRP genes into M. Sativa (Fig. S1B), in an attempt to obtain the transgenic alfalfa with not only higher stress tolerance but also higher methionine content. We identified a total of four transgenic lines by using PCR and semi-quantitative RT-PCR assays (Fig. 5A). Based on western blot analysis, two transgenic lines (PS-2 and PS-31) were selected for further studies (Fig. 5B).
We firstly compared the alkali tolerance between WT and GsPPCK3-SCMRP transgenic alfalfa. As shown in Fig. 5C, alkali stress obviously inhibited the growth of both WT and transgenic plants; however, transgenic lines displayed significantly taller (Fig. 5D) and more biomass accumulation than WT (Fig. 5E). Specifically, in the presence of 150 mM NaHCO 3 , the shoot length was 29.14 cm for WT, 32.31 cm for PS-2 and 32.72 cm for PS-31, and the shoot/root weight was 1.46/0.81 g for WT, 1.91/ 1.08 g for PS-2 and 1.9/1.08 g for PS-31, respectively (*P,0.05; **P,0.01 by Student's t-test). These results suggested that GsPPCK3 and SCMRP co-transformation promoted seedling growth of the transgenic alfalfa under alkali stress.

Improvement of PEPC Activity, Photosynthetic Rate and Citric Acid Content in Transgenic Alfalfa
To further elucidate the influence of GsPPCK3 overexpression in alfalfa, PEPC activity, which was regulated by PPCK phosphorylation [28,29], was determined by measuring the seedling crude extracts. As shown in Fig. 6A, alkaline stress improved the PEPC activity in both WT and transgenic lines, however, an obvious upregulation of PEPC activity was observed in the transgenic lines. Specifically speaking, PEPC activity of the transgenic lines was 31.64% (PS-2) and 34.78% (PS-31) higher than that of WT, respectively (*P,0.05; **P,0.01 by Student's t-test).
It has been suggested that photosynthetic rate decreased proportionally, along with the increase of NaHCO 3 concentration [30]. Considering the function of PEPC/PPCK in photosynthesis, we measured the net photosynthesis (Pn) rates of both WT and transgenic lines under alkali stress. The quantitative analysis revealed that Pn of transgenic alfalfa was significantly higher than that of WT, in the presence of 100 or 150 mM NaHCO 3 (Fig. 6B) (*P,0.05; **P,0.01 by Student's t-test). These results suggested that overexpression of GsPPCK3 led to an obvious increase of Pn under alkali stress, maybe by activating PEPCs.
Citric acid content was considered to be an indicator of plant response to pH challenge resulting from alkaline stress [31]. As expected, plants accumulated more citric acids under alkali stress than under control conditions. Compared with WT, the transgenic lines exhibited significantly higher levels of citric acid content, after 100 or 150 mM NaHCO 3 treatment (Fig. 6C) (*P,0.05; **P, 0.01 by Student's t-test). Collectively, these results demonstrated that the alleviation of high pH damage caused by alkali stress in transgenic alfalfa might be related to the elevated levels of PEPC activity, photosynthetic rate and citric acid content, due to GsPPKC3 overexpression.

GsPPCK3 Overexpression Up-regulated the Expression Levels of Several Stress Responsive Genes
Previous studies showed that H + -ATPase and NADP-ME maintained the homeostasis of the cytosolic pH value [32,33]. In view of the increased PEPC activity and citric acid content of transgenic alfalfa, we examined the expression patterns of PEPC, CS (Citrate synthase), H + -ATPase and NADP-ME after 100 mM NaHCO 3 treatment. The real-time PCR analysis indicated that expression of PEPC, CS, H + -ATPase and NADP-ME was induced by alkali stress in both WT and transgenic lines. Expectedly, the expression levels in transgenic plants were significantly higher than that in WT (Fig. 6D-G), which explained the up-regulation of the PEPC activity and citrate acid content. These results implied that GsPPCK3 overexpression promoted the accumulation of transcript expression levels of the stress responsive genes, which might be helpful for the intracellular pH regulation under alkali stress. In addition to the enhanced alkali tolerance, we also determined the content of 16 amino acids in leaves of both WT and transgenic plants, to verify if GsPPCK3-SCMRP co-transformation increased the methionine content ( Table 1). As expected, GsPPCK3-SCMRP transgenic lines displayed significantly higher levels of methionine content than WT (*P,0.05; **P,0.01 by Student's t-test), without obvious changes for other amino acids. Specifically speaking, the methionine contents were 0.9760.04% in WT plants, 2.2760.89% (2.34 folds to WT) in transgenic line PS-2 and 2.2360.67% (2.29 folds) in line PS-31. Taken together, these results suggested that GsPPCK3-SCMRP co-overexpression not only enhanced the alkaline tolerance, but also increased the methionine content of transgenic alfalfa.

Discussion
Saline-alkaline stress, as a kind of widespread environmental stress with significantly negative impact on plant growth, severely reduces crop productivity and affects agricultural production worldwide. It has been suggested that alkaline soil, characterized by high NaHCO 3 /Na 2 CO 3 content, caused injury to plants not only through salt stress, but also through alkali stress [34][35][36]. Plants could homeostatically maintain the intracellular pH value and ion concentration in a temperate range [32,[37][38][39]. Unfortunately, up to now, most studies emphasized on salt stress [40][41][42][43], and no slightly definite mechanism was proposed about plant response to alkaline stress.
Glycine soja, the wild ancestor of cultivated soybean (Glycine max), could normally germinate and grow in the alkaline soil with a pH value at 9.02 [44]. In previous studies, we constructed the global transcriptional profile of Glycine soja (G07256) under alkali stress (50 mM NaHCO 3 , pH 8.5), and three of the four PPCK genes GsPPCK1, GsPPCK2 and GsPPCK3 were identified as putative stress responsive genes (Fig. 1) [23]. Contrarily, GsPPCK4 essentially did not respond to alkali stress, only with an increase at 6 h in leaves. Consistent with previous researches, these results further supported the evolutionary relationship of soybean PPCKs [14,45]. PPCK1, PPCK2 and PPCK3 shared a high similarity and belonged  to the same legume PPCK subfamily, while PPCK4 represented a high divergence, outlier to the legume PPCK.
In a previous study, we isolated and characterized one of the alkali stress responsive PPCK genes GsPPCK1, and found that overexpression of GsPPCK1 in alfalfa significantly improved plant tolerance to alkali stress [46]. In the present study, we focused on the expression pattern and biological function of GsPPCK3. We cloned the full length GsPPCK3, and found that it shared 90%, 84%, 69% and 61% sequence identity with GmPPCK3, GmPPCK2, GmPPCK1 and GmPPCK4, respectively. Protein sequence analysis revealed a unique motif consisting of 30 amino acids in subdomain XI of GsPPCK3, which was not found in any soybean and Arabidopsis PPCKs. This unique sequence was encoded by the 90 bp length intron of GmPPCK3, indicating the existence of different transcripts in Glycine soja. Furthermore, GsPPCK3 protein contained all conserved subdomains required for kinase activity [24,25] (Fig. 2), including a protein kinase ATP-binding signature, a Ser/Thr kinase active site signature [26] and a conserved G-T/ S-XX-Y/F-X-APE motif, as well as two transmembrane domains.
GsPPCK3 expression was greatly and rapidly induced by alkali stress in both leaves and roots (Fig. 3A). Previous researches also suggested the induced expression of PPCKs under environmental stress. For example, salt stress significantly increased the transcript accumulation of SbPPCK1 and SbPPCK2 in sorghum [17]. Moreover, GsPPCK1 was also induced by alkali stress and positively regulated the alkali tolerance of transgenic alfalfa [46]. Therefore, the alkali induced expression indicated an important role of GsPPCK3 in plant response to alkali stress. It is worth noting that the stress induction degree of GsPPCK3 in roots was significantly higher than that in leaves. It is possible that plant roots were the exact sites of perception and injury for stresses, or there were different response mechanisms for GsPPCK3 expression between roots and leaves.
It has been reported that PPCK genes showed spatial expression specificity. For example, OsPPCK1 and OsPPCK3 displayed obviously high expression levels in roots [13], while GmPPCK2 and GmPPCK3 showed high transcription levels in root nodules [14]. Except for root nodules, GmPPCK3 also showed relatively higher expression levels in stems, flowers and young leaves. In this To explore expression patterns of stress-responsive genes, the 4-week-old WT and GsPPCK3-SCMRP transgenic seedlings after shoot cottage were treated with 50 mM NaHCO 3 (pH 8.5) for 0, 3, 6, 12 and 24 h, respectively. Relative transcript levels were determined by quantitative real-time PCR with GAPDH as internal reference, and were normalized to WT plants at 0 h. Values represented the means of three fully independent biological replicates, and three technological replicates for each. doi:10.1371/journal.pone.0089578.g006 study, we found that GsPPCK3 displayed the highest level in leaves but relatively lower in roots (Fig. 3B). As we know, PEPC undergoes an important function in carbon fixation during photosynthesis, and its activity was largely regulated by PPCK phosphorylation [28,29]. It is reasonable to speculate that the high levels of GsPPCK3 transcripts in leaves could make sure of the high PEPC activity and thereby the effective photosynthesis of plant leaves. Considering the difference in the transcript levels and alkali induction degrees of GsPPCK3 between roots and leaves, one could speculate that in leaves, most GsPPCK3 products were used for carbon fixation in photosynthesis process, but in roots, GsPPCK3 mainly functioned in alkali response.
To further confirm the biological function of GsPPCK3 in alkali response, we transformed GsPPCK3 into alfalfa and carried out the alkali tolerance assays. We gave four lines of direct evidence showing the increased tolerance and possible mechanisms of GsPPCK3 transgenic alfalfa in response to alkali stress (Fig. 4, 5). Firstly, GsPPCK3 overexpression significantly promoted plant growth under alkali stress. GsPPCK3 transgenic lines displayed much better at plant height, shoot weight and root weight than WT under alkali stress (Fig. 5D, E). Secondly, GsPPCK3 overexpression alleviated the damage caused by alkali stress, as evidenced by an obvious decrease in the relative ion leakage (Fig. 4C) and MDA content (Fig. 4D), but an increase in root activity (Fig. 4E) [37,[47][48][49]. Thirdly, GsPPCK3 overexpression in alfalfa resulted in increased PEPC activity (Fig. 6A), increased chlorophyll content (Fig. 4F) and thereby the increased Pn rate (Fig. 6B), which in turn could promote plant growth under alkali stress. We also observed the higher expression levels of PEPC gene in transgenic alfalfa (Fig. 6D). In this context, we proposed the hypothesis that GsPPCK3 regulated the PEPC activity not only through protein phosphorylation process, but also through gene transcription regulation. Finally, overexpression of GsPPCK3 led to more effective response to high pH stress by increasing the citrate acid content (Fig. 6C) [50]. The up-regulation of CS gene expression (Fig. 6G) in transgenic plants explained the increase of citrate acid content [51,52]. Meanwhile, we also suggested relatively higher expression level of H + -ATPase in transgenic lines under alkali stress (Fig. 6F), which was also helpful for intracellular pH regulation. Based on these results, we speculated that GsPPCK3 regulated the gene expression and enzyme activity involved in photosynthesis and pH regulation, alleviated damage caused by alkali stress, and finally promoted plant growth under alkali stress.
On the other hand, we also obtained the transgenic alfalfa with not only higher alkali tolerance but also higher methionine content, by co-transforming the GsPPCK3 and SCMRP genes. As a kind of nutritionally essential amino acid, methionine was found to be at low level in legume. In this study, we significantly improved the methionine content of transgenic alfalfa by ectopically expressing the SCMRP gene, which was designed and synthesized according to the maize methionine-rich 10 ku zein protein [20,21]. Similarly, it has been reported that overexpression of the maize 10 ku zein gene in potato could increase the contents of sulphurcontaining amino acids [22]. During the past decades, a series of efforts have been made to increase the methionine content by using genetic engineering methods, for example altering expression levels of the methionine-rich storage proteins [53][54][55], or increasing the soluble content of methionine [56][57][58][59][60].
Collectively, here we provided an effective way to simultaneously improve plant alkaline tolerance and methionine content, at least in legume crops. For the first time, we gave exact evidence for a PPCK protein from Glycine soja, and provided insights into a plausible mechanism by which GsPPCK3 positively controlled plant tolerance to alkali stress.

Plant Material and Growth Conditions
Glycine soja (G07256) seeds, obtained from Jilin Academy of Agricultural Sciences (Changchun, China), were treated with 98% sulfuric acid for 10 min, washed five times with sterile water, and then kept in complete darkness with humidity for 2-3 days to promote germination. The seedlings were transferred and grown in 1/4 Hoagland solution for 3 weeks at 24-26uC and 16 h light/ 8 h dark cycles. To explore the expression profiles of GsPPCK3 under alkali stress, the 3-week-old seedlings were treated with 1/4 Hoagland solution containing 50 mM NaHCO 3 (pH 8.5) for 0, 1, 3, 6, 12 and 24 h, respectively. Equal amounts of leaves and roots were harvested and stored at 280uC after snap-frozen in liquid nitrogen.
Medicago Sativa L. was kindly obtained from Heilongjiang Academy of Agricultural Sciences (Haerbin, China), and grown in green house under controlled environmental conditions (24-26uC, 16 h light/8 h dark cycles, 600 mmol m 22 s 21 , 8065% relative humidity). To investigate the expression patterns of stressresponsive genes, the 4-week-old seedlings after shoot cottage were treated with 1/4 Hoagland solution containing 50 mM NaHCO 3 (pH 8.5) for 0, 3, 6, 12 and 24 h, respectively. Samples were harvested and stored as described above.

Isolation and Sequence Analysis of the GsPPCK3 Gene
Total RNA was extracted from the 3-week-old Glycine soja seedlings, by using RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA), and subjected to cDNA synthesis by using SuperScript TM III Reverse Transcriptase kit (Invitrogen, Carlsbad, CA, USA). The full-length coding region of GsPPCK3 was PCR amplified with gene specific primers (59-AAGATAGAATGTGCACAGCCC-TAAAG-39 and 59-TTCTCAAGTGAGTTCAGCCGTGTC-

Quantitative Real-time PCR Analyses
Total RNA extraction and cDNA synthesis were operated as described above. Quantitative real-time PCR analyses were performed by using SYBR Premix ExTaq TM II Mix (TaKaRa, Shiga, Japan) on an ABI 7500 sequence detection system (Applied Biosystems, Carlsbad, CA, USA). The glyceraldehyde-3-phosphate dehydrogenase genes in G. soja (Accession: DQ355800) and M. sativa (Accession: Medtr3g085850) were used as internal references, respectively. Expression levels for all candidate genes were calculated by using the 2 -DDCT method, and the relative intensities were normalized as described previously [61]. To enable statistical analysis, three fully independent biological replicates and three technical repeats were conducted. Primers used for quantitative real-time PCR were designed using Primer 5 software and listed in Table 2.

Generation of Transgenic Alfalfa
In order to investigate the influence of GsPPCK3 and SCMRP on plant stress tolerance and methionine content, we constructed the expression vectors carrying the GsPPCK3 gene alone (Fig. S1A), and the GsPPCK3 and SCMRP genes together (Fig. S1B), respectively. The GsPPCK3 gene was inserted to the bone vector pBEOM (made in our lab) under the control of the cauliflower mosaic virus (CaMV) 35S promoter, with the binding enhancers E12 and omega. The SCMRP gene was under the control of double CaMV35S promoter and omega sequence. The Bar gene was used as the selectable marker. The recombinant vectors were introduced into A. tumefaciens strain EHA105, and then transformed into M. sativa by using the cotyledonary node method. The transformants were selected by using 0.5 mg L 21 glufosinate ammonium, and regenerated shoots were rooted on 1/2 Murashige and Skoog (MS) medium. At last, the glufosinatepositive seedlings were transplanted into soil and grown in green house under controlled conditions. Presence of the GsPPCK3 and SCMRP genes in the glufosinatepositive plants was confirmed by PCR analysis using CaMV35S promoter specific forward primer and Bar gene specific reverse primer (59-CCTGTGCCTCCAGGGAC-39 and 59-GCGGTCTGCACCATCGTC-39). GsPPCK3 transcript levels in the PCR-positive plants were analyzed by semi-quantitative RT-PCR  analysis  using  gene  specific  primers  (59-CCCTCCTTTCACCTCACC-39 and 59-GAACC-GAAGTCCGCCAGT-39). Expression levels of the SCMRP gene were examined by using a pair of specific primers (59-CAG-CAGGGTCTCGCTTCACT-39 and 59-GCAGATTC-CAATGCCACAAT-39). The alfalfa GAPDH gene was used as an internal control. And the PCR-and RT-PCR positive seedlings were subjected to western blot analysis with specific polyclonal antibody to the C-terminus of GsPPCK3 protein (CHPWIL-SAGDTAELT).

Phenotypic Analysis of Transgenic Alfalfa Under Alkali Stress
The lignified WT and transgenic alfalfa plants were propagated by stem cuttings, and about 2 weeks later, adventitious roots appeared. The seedlings were transplanted into plastic culture pots filled with a mixture of peat moss: soil (1:1; v/v), irrigated with 1/8 Hoagland nutrient solution and grown in green house under controlled conditions. For phenotypic analysis of GsPPCK3 transgenic alfalfa, the 3-week-old WT and transgenic plants with similar sizes were exposed to alkali stress by irrigating with 1/8 Hoagland solution containing either 0, or 50, or 100 mM NaHCO 3 every 3 days for a total of 12 days. For phenotypic analysis of GsPPCK3-SCMRP transgenic alfalfa, the 4-week-old plants were exposed to alkali stress by irrigating with 1/8 Hoagland solution containing either 0, or 100, or 150 mM NaHCO 3 every 3 days for 15 days.
The total chlorophyll content was determined in 80% (v/v) acetone extract according to the method of Arnon [62]. The relative electrolyte leakage was measured using a conductivity meter (DDSJ-308A, Shanghai, China) as described previously [63]. The malon dialdehyde (MDA) content was determined according to the protocol described by Peever et al. [64]. The biological activity of roots from alkali-treated WT and transgenic plants was assessed by measuring root dehydrogenase activity using triphenyltetrazolium chloride (TTC) reduction technique as described [65]. The citric acid content was determined by using a spectrophotometer (UV-2550, Shimadzu, Japan) at the absorbance of 490 nm, according to the method of Zhu [66]. The net photosynthesis rate was determined by using the LI-6400 chamber (LI-COR Biosciences, Lincoln, NE, USA). During the experiments, the light intensity was 500 mmol photons m 22 s 21 , and the CO 2 concentration was in the range of 350 to 400 mmol mol 21 . The flow rate was adjusted to 500 mmol s 21 , and the leaf temperature was maintained at 25uC. For statistical analysis, three independent experiments were conducted with at least 5 independent plants per genotype and experiment.
PEPC activity was measured by the coupled spectrophotometric method at 340 nm and 30uC as previously described by Gonzalez [67]. Total proteins of the WT and OX alfalfa leaves were extracted by using the 0.1 M Tris-HCl (pH 7.5) solution, containing 20% (v/v) glycerol, 1 mM EDTA, 10 mM MgCl 2 , 10 mg mL 21 chymostatin, 10 mg mL 21 bestatin, 10 mg mL 21 leupeptin, 1 mM PMSF, 1 mg mL 21 microcystin-L/R (L and R are two variable amino acids in the structure of microcystin), and 14 mM b-mercaptoethanol. Protein concentration was determined according to the method of Bradford (1976) [68]. All of the above numerical data were subjected to statistical analyses using EXCEL 2007 and/or IBM SPSS statistics 19, and analyzed by Student's T-test and/or Duncan's Multiple Range Test.

Determination of Amino Acid Content
Leaves of WT and transgenic alfalfa were fixated at 105uC for 15 min, and then dried to constant weight at 80uC. About 100 mg dried leaves were crushed, suspended in 6 M HCl, and then hydrolyzed at 110uC for 22 h with nitrogen replacement. The hydrolysate was dried under reduced pressure, dissolved in 0.02 M HCl, and subjected to the amino acid analyzer (Hitachi, Japan). Figure S1 Schematic representation of the constructs for Agrobacterium tumefaciens-mediated transformation into Medicago sativa. a, Schematic representation of the GsPPCK3 overexpression construct. b, Schematic representation of the GsPPCK3-SCMRP overexpression construct. (TIF)