A rice tonoplastic calcium exchanger, OsCCX2 mediates Ca2+/cation transport in yeast

In plant cell, cations gradient in cellular compartments is maintained by synergistic action of various exchangers, pumps and channels. The Arabidopsis exchanger family members (AtCCX3 and AtCCX5) were previously studied and belong to CaCA (calcium cation exchangers) superfamily while none of the rice CCXs has been functionally characterized for their cation transport activities till date. Rice genome encode four CCXs and only OsCCX2 transcript showed differential expression under abiotic stresses and Ca2+ starvation conditions. The OsCCX2 localized to tonoplast and suppresses the Ca2+ sensitivity of K667 (low affinity Ca2+ uptake deficient) yeast mutant under excess CaCl2 conditions. In contrast to AtCCXs, OsCCX2 expressing K667 yeast cells show tolerance towards excess Na+, Li+, Fe2+, Zn2+ and Co2+ and suggest its ability to transport both mono as well as divalent cations in yeast. Additionally, in contrast to previously characterized AtCCXs, OsCCX2 is unable to complement yeast trk1trk2 double mutant suggesting inability to transport K+ in yeast system. These finding suggest that OsCCX2 having distinct metal transport properties than previously characterized plant CCXs. OsCCX2 can be used as potential candidate for enhancing the abiotic stress tolerance in plants as well as for phytoremediation of heavy metal polluted soil.


Results
OsCCX2 transcript is differentially expressed under abiotic stresses. The expression profile of a gene provides a clue about its functional relevance and paves way for its further characterization. Four CCX genes namely OsCCX1, OsCCX2, OsCCX3 and OsCCX4 were identified in rice genome. The probesets of OsCCX1 to OsCCX3 were available while probesets for OsCCX4 were not labelled on microarray chip (Agilent and Affimetrix). To determine the transcript expression profile of OsCCX1-3 under abiotic stresses (drought, salt and cold), publically available rice microarray database, Rice DB:Oryza information portal (http://ricedb.plantenergy.uwa.edu.au/) was used. The OsCCX2 transcript was differentially upregulated under drought and salt stresses while no significant change was observed under cold stress (Fig. 1A). The expression profile of two other CCXs, OsCCX1 and OsCCX3 was not significantly changed under drought, salt and cold stresses (Fig. 1A). Expression data of 3 hrs exogenous application of ABA to rice seedlings 35 showed that transcript of OsCCX2 was highly upregulated while OsCCX3 transcript was downregulated (Fig. S1a). The microarray expression profiles of OsCCX2 in drought, salt and ABA treatment were also validated by qPCR and transcript was found to be upregulated upon longer exposure to these stresses ( Fig. 1B-D). Seedlings treated with longer exposure to drought depict increasing transcript abundance and after 6 hrs of drought stress, OsCCX2 transcript shows more than twenty fold increase in transcript abundance (Fig. 1B). Likewise, OsCCX2 transcript showed gradual transcripts abundance under longer exposure to salt or exogenous ABA application (Fig. 1C,D). The results were statistically supported with significant p-value. ABA and drought responsive transcript abundance is further strengthened by presence of various ABA responsive and dehydration responsive elements in 1 kb upstream promoter region of OsCCX2 (Fig. S1b).
OsCCX2 transcript is down-regulated under Ca 2+ deficiency. To understand the calcium dependent function of rice OsCCXs, microarray data 36 was analysed to determine the expression profile of these genes under Ca 2+ deficient conditions. Under Ca 2+ deficiency conditions of 5 and 14 days, the transcript levels of OsCCX1 and OsCCX3 remained unaffected, whereas OsCCX2 showed significant down regulation. However, OsCCX2 showed upregulation after 6 hrs of Ca 2+ resupply to prolonged Ca 2+ deficiency (14 days) treated seedling, indicating role of OsCCX2 in calcium signaling and transport Scientific RepoRts | 5:17117 | DOI: 10.1038/srep17117   (Fig. 1E). The microarray expression profile of OsCCX2 was validated by qPCR showing similar expression profile under Ca 2+ deficiency and 6 hrs of Ca 2+ resupply conditions (Fig. 1F). Among all rice CCXs, only OsCCX2, showed a distinct and differential expression pattern under Ca 2+ deficiency. The results were statistically supported with significant p-value. This down regulation in transcript abundance of OsCCX2 is indirectly supported by presence of various calcium responsive and CaM binding elements such as ABRERATCAL and CGCGBOXAT 37,38 in 1 kb upstream promoter region of OsCCX2 (Fig. S1b). The presence of three ABRERATCAL elements and sixteen CGCGBOXAT boxes in the promoter region of OsCCX2 signify its role in calcium responsive pathways. Along with Ca 2+ binding and ABA responsive elements, we also analysed the presence of nutrition responsive cis-acting elements and four sulphur responsive element, SURECOREATSULTR11, were found at 1 kb upstream of OsCCX2 gene but other nutrition responsive elements were absent. We have also analysed the expression data of rice CAXs and CCXs under different nutrient deficient conditions. The OsCCX2 was up-regulated under low phosphate (P) conditions (72 h, 24 h and 6 h) while down regulated in shoot under P deficiency. OsCCX2 was also down regulated under potassium (K + ) deficiency. The expression of other rice CCXs were not much affected in major nutrient deficient conditions. The expression of rice CAX1a, CAX1b and CAX1c were altered in low P and Fe (iron) conditions while transcripts of other rice CAXs were not much affected under different nutrient deficient conditions (Fig. S2). OsCCX2 has characteristic Ca 2+ /cation transport signature elements. Calcium exchanger proteins of rice comprised of six Ca 2+ /H + exchangers (CAXs), four cation/Ca 2+ exchangers, two EF-CAX, two MHX and single OsNCKX1 23,27 . The OsCCX2 (LOC_Os03g45370) open reading frame contains 1,728 nucleotides, translating into a putative 60 kDa protein (576 amino acids) and its domain analysis reveals that it is a putative Na + /Ca 2+ exchanger protein. Domain structure analysis of OsCCX2 reveals that it has typical 12 transmembrane domains, two Na + /Ca 2+ exchanger domains and characteristic α -1 and α -2 motifs, which are considered necessary and sufficient for cation exchange activity. The presence of unique signature motifs GNG(A/S)PD in α -1 and (G/S)(N/D)SxGD motif in α -2 repeat regions 23,28 also provide indirect evidence regarding Ca 2+ ion selectivity, binding, transport of OsCCX2 protein (Fig. S3). It shows more than 60% protein sequence homology with cation/Ca 2+ exchanger of monocot species (Brachypodium, Sorghum and Zea mays) and 41% protein sequence homology with Arabidopsis CCX1. In phylogenetic analysis, OsCCX2 groups with Arabidopsis and rice CCXs and its closest Arabidopsis orthologs are AtCCX1 and AtCCX2 27 . A high degree of sequence homology among Arabidopsis and rice CCXs suggests possible involvement of OsCCX2 in similar calcium signaling pathways and prompted us to perform in-depth functional characterization to understand the cation/Ca 2+ exchange properties of OsCCX2.

OsCCX2 localized to tonoplast in
OsCCX2 suppress Ca 2+ sensitivity of K667 yeast mutant. Previously, yeast heterologous expression system approach has been used successfully to decipher the functional role of plant exchanger proteins [31][32][33]40 . We adopted the same approach to determine the role of OsCCX2 in transport of Ca 2+ and other cations. K667, a low affinity Ca 2+ uptake deficient triple S. cerevisiae mutant strain, lacking vacuolar ATPase, PMC1; vacuolar exchanger, VCX1; and cytosolic regulatory CNB1 subunit 41 . The growth of K667 mutant is compromised in media having excess Ca 2+ concentration and is due to defective vacuolar Ca 2+ transport 41 .
Moreover, it was also reported that Arabidopsis CCX5 and CCX3 were unable to suppress the high Ca 2+ sensitivity of yeast mutant K667, where the later was characterized as endomembrane localized H + driven K + transporter 32

GFP
Bright Field Merge B

GFP
Bright Field Merge C

Figure 2. Subcellular localization of OsCCX2 proteins in Nicotiana benthamiana epidermal cells.
(A) OsCCX2 fused with N-terminal of GFP appears as circular pre-vesicles inside the lumen of the cell and seems to be localized to vacuolar membrane. (B) The OsCCX2 fused at the C-terminal of GFP also showed tonoplast localization. (C) Cells transformed with CaMV 35S -GFP were used as a control. Fluorescence was detected under a confocal laser-scanning microscope (wavelength: 488 nm). (D) The GFP-OsCCX2 colocalized completely with globular vesicles as shown by tonoplast markers (vac-rk). GFP fusions to OsCCX2 proteins are shown in green, mCherry vacuole markers are shown in red and overlay of two-mentioned proteins in dark field view. Scale bar = 40 μ m. glucose or galactose (Fig. 3A,B). The calcium sensitivity of K667 was suppressed by OsCCX2 in presence of galactose but not in glucose; confirming the inducibility and expression of OsCCX2 protein under higher CaCl 2 concentrations (Fig. 3A,B). It was found that K667 expressing OsCCX2 yeast cells could grow better while growth of empty vector expressing K667 cells was highly compromised at higher CaCl 2 (up to 150 mM) (Fig. 3B).
Moreover, multiple sequence alignment of Arabidopsis CCXs with CAXs revealed that CCXs do not have N-terminal autoinhibitory domain in contrast to CAXs 32 . The full length ORF of OsCCX2 was able to complement K667 mutant, which further indicate that it do not have a N-terminal autoinhibitory domain while showing Ca 2+ exchange properties like other CAXs.
OsCCX2 suppress Na + and Li + sensitivity of K667 yeast mutant. Arabidopsis CAXs and CCXs have broad substrate specificity and can transport a variety of monovalent (Na + , K + , Li + ) as well as divalent (Ca 2+ , Mn 2+ , Zn 2+ , Cd 2+ , Co 2+ ) cations. Although some CAXs such as AtCAX1 and Saccharomyces cerevisiae VCX1 (ScVCX1) are specific for Ca 2+ 27,29,32,41-43 transport, the mammalian CCX family member NCKX6, has been demonstrated to mediate Na + /Ca 2+ and Li + /Ca 2+ exchange 44 . The growth of K667 mutant is also compromised at higher concentration of NaCl 42 . In order to determine the Na + transport properties of OsCCX2, we have performed the complementation assay in K667 yeast mutant at higher concentration of exogenous NaCl. In contrast to empty vector control, OsCCX2 expressing mutant cells grew better on media supplemented with high concentration of NaCl (200-500 mM) (Fig. 4A). This reveal that OsCCX2 efficiently suppress the Na + hypersensitivity of K667 in a manner similar to the suppression of Ca 2+ sensitive phenotype of this mutant (Fig. 4A). K667 yeast mutant shows higher growth sensitivity to excess LiCl than NaCl under increasing concentration of LiCl from 25 mM to 100 mM (Fig. 4B). Empty vector expressing K667 mutant cells were not able to grow at 100 mM LiCl while growth of OsCCX2 expressing yeast mutant is comparable to wild type strain.
OsCCX2 provide metal tolerance to K667 yeast mutant. On the basis of phylogenetic analysis of CAXs, type IB CAXs (CAX2, CAX5 and CAX6) are involved in transporting other metal ions with broader substrate specificity for divalent cations, such as Ca 2+ , Mn 2+ , Zn 2+ , Co 2+ and Cd 2+20,29,31,45-47 . Whereas, in the case of CCXs, only AtCCX3 and AtCCX5 show broad range of substrate specificity from monovalent K + and Na + to divalent cations such as Mn 2+ and Zn 2+ 32,33 .
In order to understand transport properties of various metal ions (Fe 2+ , Zn 2+ , Co 2+ Mn 2+ and Cd 2+ ) mediated by OsCCX2, the sensitivity and tolerance of OsCCX2 expressing K667 cells were examined under excess concentrations of above mentioned heavy metals. We found that OsCCX2 expressing K667 yeast cells provide tolerance specifically against excess FeSO 4 , ZnCl 2 , and CoCl 2 than empty vector OsCCX2 suppress metal ion sensitivity of mid1 mutant. As mentioned above, we have shown that OsCCX2 has ability to suppress Ca 2+ , Na + , Li + , Fe 2+ , Zn 2+ and Co 2+ sensitivity of K667 mutant. In order to further validate the metal transport properties in another yeast mutant, we choose the S. cerevisiae mid1 mutant. The MID1 gene is nonessential for vegetative growth. Generally, mid1 mutant cells grow normally in a medium having glucose as a carbon source even at higher concentration of NaCl (upto 500 mM) while mid1 mutant shows impaired growth in presence of galactose at higher concentration of NaCl (upto 500 mM). The galactose inducible pYES2 empty vector expressed in mid1 mutant showed impaired growth at higher concentration of NaCl (500 mM), LiCl (25-100 mM) and FeSO 4 (20 mM) in presence of galactose as carbon source and as an inducer of Gal1 promoter (Fig. 6A-C). Therefore, OsCCX2 expressing mid1 yeast cells can suppress Na + , Li + and Fe 2+ ion sensitivity and provide tolerance at increasing concentration of these metal ions (Fig. 6A-C). This complementation further validates the cation transport activity of OsCCX2 in another yeast mutant background.
Unlike AtCCXs, OsCCX2 could not transport K + . The AtCCX3 and AtCCX5 shows K + transport in yeast mutant. To validate K + transport activity of OsCCX2, we transformed it into W∆6 yeast double mutant 43 . The W∆6 mutant lacks Trk1 and Trk2 genes and hence a high affinity K + transport is compromised in low K + -Na + medium. The arginine-phosphate yeast growth medium was specially designed with lesser K + and Na + content for evaluating the growth sensitivity of W∆6 yeast double mutant 43 . Similar to W∆6 mutant, OsCCX2 expressing W∆6 cells were not able to complement the mutant under micromolar concentration of K + in arginine-phosphate medium and suggest its inability to transport K + under low K + -Na + medium (Fig. 6D).

OsCCX2 expressing K667 cells have lesser cation content. The yeast cells maintain cationic bal-
ance by various signaling pathways. Earlier results of this study suggest that OsCCX2 suppress Ca 2+ , Na + , Li + , Fe 2+ , Zn 2+ and Co 2+ ion sensitivity of K667 yeast mutant (Figs 3-5). To understand the differential accumulation of these above-mentioned ions in the yeast, we have analysed the total cation content of wild type yeast (K601), K667 mutant and K667 mutant expressing OsCCX2 under excess concentration of CaCl 2 , NaCl, FeSO 4 and ZnCl 2 as mentioned in material and methods. Atomic absorption spectrophotometric analysis was done to estimate the total accumulation of Ca 2+ , Na + , Fe 2+ and Zn 2+ elements in yeast cells. The total contents of Ca 2+ , Na + , Fe 2+ and Zn 2+ ions under normal growth condition were comparable to standard metal contents of yeast cells 48 (Fig. 7).
It is already known that total Ca 2+ concentration in K601 strain (WT) is below 0.2 mg/g dry weight 49 , which corresponds well with the results of this study (Fig. 7). The total concentrations of Ca 2+ and Na + ions were in ppm range while total heavy metal Fe 2+ and Zn 2+ contents were in ppb range as reported previously 49 (Fig. 7). The K667 mutant accumulates more Ca 2+ , Na + , Fe 2+ and Zn 2+ ions than WT and OsCCX2 expressing K667 cells. The wild type strain has lowest Ca 2+ , Na + , Fe 2+ and Zn 2+ content than  Figure 4. Suppression of Na + and Li + sensitivity of K667 yeast mutant by OsCCX2. Yeast strains having empty vectors and OsCCX2 were grown in SC-Ura overnight at 30 °C and diluted to 0.5 OD 600 followed by 10 times serial dilutions of 0.5 OD 600 were dotted on SC-URA+ Gal containing different concentrations of NaCl and LiCl. The plates were incubated at 30 0 C for 3 days. OsCCX2 was able to provide tolerance to K667 yeast mutant under excess NaCl and LiCl conditions. Scientific RepoRts | 5:17117 | DOI: 10.1038/srep17117 either OsCCX2 expressing K667 cells or empty vector expressing K667 cells under excess metal concentrations (Fig. 7). These observations were statistically supported with significant p-value.

Discussion
Cation/Ca 2+ exchangers from lower to higher plants indicate their vital role in maintaining calcium-cation homeostasis in the plant cells. To characterize an individual gene or group of genes, expression analysis under various conditions may provide clues about their possible functions in the cell. Rice OsCCX genes are responsive to abiotic stresses, especially transcripts of OsCCX2 were differentially upregulated under drought and salt stresses (Fig. 1A). Earlier reports also indicate that expression of Arabidopsis CCX1 increases under drought and osmotic stresses while CCX2 transcript is upregulated under cold stress 32 . Arabidopsis CCX3 transcript is also upregulated under NaCl and KCl treatment while expression of CCX4 does not change under abiotic stresses 32 . The expressional analysis of OsCCXs, especially OsCCX2, at various time points under abiotic stresses indicates its possible role in abiotic stress signaling in rice.
In plant cell, calcium ATPases and calcium exchangers maintain nanomolar [Ca 2+ ] cyt by efflux or influx of Ca 2+ present either at plasma membrane or at endomembranes 50,51 . Under Ca 2+ deficiency, the transcript level of ATPases and CAXs remains to be either unchanged or downregulated, to maintain optimal level of Ca 2+ in cytosol. Microarray expression data under 5 days or 14 days of Ca 2+ deficiency shows moderate decrease in expression of OsCCX1 and OsCCX3 while OsCCX2 transcripts were highly down regulated (Fig 1E). Moreover, OsCCX2 transcripts were slightly upregulated by six hours of Ca 2+ resupply upon Ca 2+ deficiency (Fig. 1F). Transcriptome analysis revealed that most of the Arabidopsis Ca 2+ ATPases are down-regulated under Ca 2+ deficiency 52  OsCAX1a, CAX2 transcript was significantly down regulated 40 . This role was further supported by transcriptome analysis of Arabidopsis Ca 2+ exchanger, CAX1, which showed upregulation in response to exogenous Ca 2+ application 53 . Therefore, altered expression of OsCCX2 under Ca 2+ deficiency/resupply indicate its role in Ca 2+ homeostasis in plant cell. Subcellular localization of OsCCX2 in Nicotiana benthamina epidermal peel cells revealed that it is localized to vacuolar membrane. In earlier study also, AtCCX3 was shown to be localized to vacuolar membrane of onion peel cells 32 while AtCCX5 was localized to plasma membrane and around the nuclear periphery in yeast cells 33 . Based on these report and the results of the present study, it is speculated that different members of CCXs protein might be involved in transporting the ion across various membrane of cell. However, this assumption requires in-depth functional assays in future studies.
Yeast mutant complementation analysis has been extensively used to assess the functional nature of most of the Ca 2+ exchangers and pumps of plants 40,41,49,50 . Previously, it is well established that plant CAXs have ability to suppress Ca 2+ sensitivity of K667 yeast mutants 30,41,42 but AtCCX3 and AtCCX5 were unable to complement and suppress Ca 2+ sensitivity of this yeast mutant. In contrast to previously characterized CCXs, rice OsCCX2 complements the K667 yeast mutant in presence of excess Ca 2+ like CAXs members. This result possibly infers that OsCCX2 shows similarity in Ca 2+ transport properties like Arabidopsis CAXs and others. Unlike majority of plant CAXs, full length OsCCX2 was able to rescue CaCl 2 sensitivity of K667 mutant, indicating absence of N-terminal regulatory domain.
The OsCCX2 protein complements the NaCl sensitivity of K667 mutant in presence of excess NaCl. It was reported that AtCCX5 showed Na + transport in 9.3 yeast mutant cells (defective in K + and Na + transport) under low Na + -K + medium 33 . It also provide tolerance to Li + indicating its unique Li + ion transport unlike reported for Arabidopsis CAXs and CCXs. As reported previously, Li + could compete with Na + transport in yeast cells 54,55 . Soybean cation/proton antiporter, GmCAX1 expression is induced by Ca 2+ , Na + and Li + treatments 56 . Arabidopsis plants overexpressing GmCAX1 were more tolerant to elevated Li + and Na + levels during germination and suggested that GmCAX1 may function as antiporter The plates were incubated for 3 days. OsCCX2 was able to suppress the Na + , Li + and Fe 2+ sensitivity of mid1 mutant and hence provide tolerance toward these metal ions. (D) The high affinity K + uptake in yeast W∆6 mutant is compromised. OsCCX2 was unable to complement W∆6 mutant and showed growth like vector control under low K + concentrations on solid arginine-phosphate medium as described under "Material and methods" section. The plates were incubated for 5 days.
Scientific RepoRts | 5:17117 | DOI: 10.1038/srep17117 for Na + and Li +56 . It is quite possible that OsCCX2 might also be acting as Na + and Li + transporter similar to GmCAX1. Interestingly, based on the ∆trk1trk2 yeast mutant complementation analysis, OsCCX2 does not have the ability to transport K + ions like reported for other AtCCXs. Therefore, OsCCX2 have Na + transport activity like AtCCX5 while in contrast to AtCCX3 and AtCCX5, unable to transport K + in yeast mutant cells.
OsCCX2 shows similar properties like IB type of CAXs, which also transport heavy metals such as Fe 2+ , Zn 2+ and Co 2+ and provide tolerance to K667 mutant. Arabidopsis AtCCX3 and AtCCX5 also transport divalent heavy metal ions such as Mn 2+ , Zn 2+ and Co 2+ , Cd 2+ , Li + ions respectively by unknown mechanism 32,33 .
In contrary to our expectation, OsCCX2 was unable to complement the single metal transport defective yeast mutants ccc1, zrc1, cot1 (defective in Fe 2+ , Zn 2+ and Co 2+ transport respectively) under excess of respective metal ions (Fig. S4). In a previous report as well, Arabidopsis ZAT1, a putative vacuolar Zn 2+ transporter, failed to suppress the Zn 2+ sensitivity of zrc1cot1 double mutant strain but it was able to suppress a Schizosaccharomyces pombe Zn 2+ -sensitive mutant devoid of SpZRC1 gene 57 . It is quite possible that OsCCX2 is not able to complement these specific mutants, which are generated in Saccharomyces cerevisiae background. Hence, it is quite apparent that rice OsCCX2 might show functional heterogeneity (variability in transport) towards different metal ions transport and indicate its role in Fe 2+ , Zn 2+ and Co 2+ ions transport.
To further validate the cation transport activity in other mutants than K667, we choose mid1 mutant for complementation analysis in this study. MID1 is N-glycosylated integral membrane Ca 2+ -permeable channel required for Ca 2+ influx and stimulated by mating pheromone α -factor 58 . The mid1 mutant also show impaired growth in presence of excess NaCl and metal ions when galactose is used as a carbon source. We showed that mid1 mutant expressing OsCCX2 shows more tolerant phenotype in presence of galactose along with excess NaCl, LiCl and FeSO 4 . At present, we do not know the mechanism of Figure 7. Total Ca 2+ , Na + , Fe 2+ and Zn 2+ accumulation in K667 yeast mutant expressing OsCCX2 under normal and excess metal concentrations. Yeast strains transformed either with empty vector or OsCCX2 were grown without excess metal (control) or with indicated excess metal ions. The cells were harvested and samples were prepared for atomic absorption spectrophotometric analysis as described under "Material and methods" section. The total metal ion content in yeast was measured by atomic absorption spectrophotometry. Differences between mean values of treatments and controls were compared using Student's t -tests ( * P≤ 0.05, ** P≤ 0.01).
Scientific RepoRts | 5:17117 | DOI: 10.1038/srep17117 tolerance provided to mid1 mutant by OsCCX2 in above-mentioned conditions but it may be speculated that OsCCX2 might be involved in excess metal ions transport and hence impart tolerance to mid1 mutant.
Here, we have shown that total Ca 2+ , Na + , Fe 2+ and Zn 2+ ions accumulation is higher in K667 mutant than OsCCX2 expressing K667 mutant when grown in excess of these metal ions. We hypothesize that the higher accumulation of these cations leads to sensitive phenotype of K667 mutant but this notion require further experimental evidences. It was already reported that another yeast triple mutant, K616, where high affinity Ca 2+ uptake is compromised, accumulates 8 times higher Ca 2+ than WT when subjected to 20 nM external Ca 2+49 . Moreover, deletion of CNB1 blocks ENA1 expression and hence regulates ENA1 mediated efflux of cations. Therefore, calcineurin signaling plays critical role in cation homeostasis and ENA1 mediated cation efflux [59][60][61] . OsCCX2 complements K667 mutant, which lacks PMC1, VCX1 and CNB1 genes. It is possible that in absence of functional calcineurin, expression of ENA1 system genes is compromised and therefore lesser efflux of excess cations like Na + , Li + and other cation through ENA efflux system and hence higher accumulation of cations take place, especially Na + in K667 mutant.
Overall, this study suggests that OsCCX2 is involved in providing tolerance to K667 mutant against over-accumulation of Ca 2+ , Na + , Fe 2+ and Zn 2+ in yeast cells. Hence, it leads to survival of OsCCX2 expressing K667 mutant under excess metal ions by some unknown mechanism, which requires further experimental work in near future.
Conclusion and future prospective. In conclusion, rice OsCCX2, belongs to Ca 2+ /cation exchanger family. In contrast to other rice CCXs members, OsCCX2 transcript is highly inducible under abiotic stresses especially drought and salt and also under exogenous application of ABA. The expression of OsCCX2 is downregulated in seedlings grown under Ca 2+ deficient conditions and the protein is localized to vacuolar membrane of N. benthamiana epidermal peel cells. The bioinformatic analysis reveals that OsCCX2 is a Ca 2+ /Na + exchanger, with 12 transmembrane domains and peculiar cation exchange domains. OsCCX2 suppresses the Ca 2+ , Na + , Li + , Fe 2+ , Zn 2+ and Co 2+ sensitivity of low affinity Ca 2+ uptake deficient yeast K667 mutant and Na + , Li + and Fe 2+ sensitivity of mid1 mutant (Fig. 8). Whereas, OsCCX2 could not suppress W∆6 (∆trk1trk2) yeast mutant suggesting inability to transport K + ions under low K + condition (Fig. 8). In conclusion, we have identified rice OsCCX2, a tonoplast localized Ca 2+ /cation exchanger showing a distinct monovalent and divalent cations exchange properties in yeast cells.
In future, in planta characterization of OsCCX2 under abiotic stress and different metal ions will reveal its role in cation homeostasis in rice. Genetically engineered variants of CCX2 proteins can be used as potential candidate for enhancing the abiotic stress tolerance, increased level of calcium content in edible parts of plant to eradicate Ca 2+ malnutrition, and can be used for phytoremediation of heavy metals polluted soil.

Materials and Methods
Microarray based gene expression analysis of OsCCXs. The  Rice DB:Oryza Information Portal (http://ricedb.plantenergy.uwa.edu.au/). Transcript profiling of rice OsCCXs in Ca 2+ deficient conditions was analysed from the microarray expression data of rice seedling 36 . We could not monitor the expression of OsCCX4 because probesets for OsCCX4 were not available in rice microarray chips (Agilent and Affimetrix).

Expression analysis of OsCCX2 by quantitative PCR. Microarray expression data of OsCCX2
under abiotic stress conditions and exogenous ABA application showed significant differential regulation, validated by quantitative PCR (qPCR) using two biological replicates. The microarray expression pattern of OsCCX2 under 5 days and 14 days of Ca 2+ deficiency and 6 hrs of resupply were also validated by qPCR. The primers of OsCCX2 were designed from 3′ UTR region using PRIMER EXPRESS (PE Applied Biosystems, USA) with default parameters and analyzed using BLAST tool of NCBI and dissociation curve analysis after the PCR reaction for their specificity (Table S1). First strand cDNA was synthesized using 2 μ g of DNaseI treated total RNA, in a 50 μ l reaction volume, using high-capacity cDNA Archive kit (Applied Biosystems, USA

Sub-cellular localization of OsCCX2 into Nicotiana benthamiana epidermal peels.
Agrobacterium tumefaciens (GV3101: pMP90) was transformed with OsCCX2-GFP and GFP-OsCCX2 constructs. Overnight grown culture of transformed Agrobacterium (OD 600 = 0.5) and helper strain p19 cells (OD 600 = 0.3) were pelleted down at 5000g for 15 minutes. Pelleted cells were re-suspended in activation buffer (10 mM MES, 10 mM MgCl 2 , 150 μ M acetosyringone) and kept for 2-4 hours at room tempreture. The re-suspended cells were then used to infiltrate the leaves of 4-6 weeks old N. benthamiana plants. After infiltration, plants were kept for incubation (3-4 days) under 12 h light/12 h dark photoperiodic cycle at 28 °C. Agrobacterium-infiltrated N. benthamiana epidermal peel cells were analyzed in TCS SP5 laser scanning confocal microscope (Leica, Germany) for sub-cellular localization and co-localization experiments. GFP fluorescence signals were detected at 500 to 535 nm after excitation at 488 nm while mCherry was excited at 543 nm and scanned at 600-630 nm. For co-localization experiments, sequential scanning was done for both the channels and then merged together to shows overlapping signals. All the images were further processed using Leica LAS AF Lite software.
DNA manipulation-Cloning of OsCCX2 and plasmid DNA constructs. Rice OsCCX2 was amplified from cDNA of 7 day old IR-64 rice seedlings and was cloned into pJET1.2 vector (Fermentas Inc. USA). The cloned gene was confirmed by sequencing. OsCCX2 was subcloned into galactose inducible yeast expression vector, pYES2 (Clontech Inc. USA). For in planta subcellular localization, OsCCX2 ORF was cloned into pGPTVII-GFP.Kan vector to form fusion protein at the N-terminal of GFP (OsCCX2-GFP) 62 . To generate the fusion protein at C-terminal of GFP (GFP-OsCCX2P), OsCCX2 entry clone was mobilized to pSITE2CA vector 63 . Both the resulting constructs contained cauliflower mosaic virus (CaMV) 35S promoter driving expression of the transporter. All primers used in this study are listed in Table S1.
Yeast strains, media and growth conditions. To test the different cations transport function of OsCCX2 in yeast, OsCCX2/pYES2 along with vector control, were individually transformed into different yeast mutant strains by LiAc/ss carrier DNA/PEG method 64 . The details of different Saccharomyces cerevisiae yeast strains used in this study are listed in Table S2. Transformants were selected for uracil prototrophy by plating on synthetic medium minus uracil (SC-Ura; 6.7 g/L yeast nitrogen base without ammonium sulphate, without amino acids, 5 g/L ammonium sulphate, 1.92 g/L of drop-out mix without uracil, 2% glucose and 2% agar). Single colony of each transformant was grown in SC-Uracil having midlog phase and culture was pelleted, washed thrice with water. The yeast cells were serially diluted with water to obtain 10 −1 , 10 −2 and 10 −3 fold dilutions of 0.5 OD 600 cultures. 5 μ l of each serial dilution was dotted on the SC-Uracil/galactose media with mentioned different concentrations of cations. To observe the growth, plates were incubated at 30 °C for 3 days otherwise mentioned in figures. Arginine-phosphate (AP) medium, first described by Rodriguez-Navarro et al. (1984) is a Na + -K + -free minimal synthetic medium frequently used to examine the effect of Na + and K + on ∆trk1∆trk2 mutant growth 43 . The composition of AP medium: 8 mM phosphoric acid, 10 mM L-arginine, 2.0 mM MgSO 4 , 0.2 mM CaCl 2 , 10 mM MES, 2% glucose, recommended vitamins, trace elements, amino acids and pH 6.5 was adjusted using L-arginine. 1% agarose was used to make solid AP medium plates.
Total Ca 2+ , Na + , Fe 2+ and Zn 2+ analysis in yeast cells. Total Ca 2+ , Na + , Fe 2+ and Zn 2+ were measured using Shimadzu Atomic Absorption Spectrophotometer 6300. Single colony of K601 having pYES2 vector and K667 mutant harbouring pYES2 and OsCCX2/pYES2 cultured in SC-uracil having different concentrations of metal ions (CaCl 2 , 50 mM; NaCl, 300 mM; FeSO 4 , 7.5 mM and ZnCl 2 , 4 mM) at a cell density of ~2 OD 600 /ml. The cultures of each set were harvested in pre-weighed microfuge tubes by centrifugation. The cells grown in CaCl 2 were washed three times with 10 mM EGTA pH 5.5, while cells grown in other metal ions were washed with 10 mM EDTA and then finally with Milli Q water. After aspirating the supernatant, pelleted cells were dried in oven overnight at 60 °C. The dried samples were weighed again and re-suspended in HNO 3 for atomic absorption spectrophotometry.