SEARCH

SEARCH BY CITATION

Keywords:

  • cold acclimation;
  • compatible solutes;
  • osmo-adaptation;
  • plant metabolomics

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Metabolomic investigation of the freezing-tolerant Arabidopsis mutant esk1 revealed large alterations in polar metabolite content in roots and shoots. Stress metabolic markers were found to be among the most significant metabolic markers associated with the mutation, but also compounds related to growth regulation or nutrition. The metabolic phenotype of esk1 was also compared to that of wild type (WT) under various environmental constraints, namely cold, salinity and dehydration. The mutant was shown to express constitutively a subset of metabolic responses which fits with the core of stress metabolic responses in the WT. But remarkably, the most specific metabolic responses to cold acclimation were not phenocopied by esk1 mutation and remained fully inducible in the mutant at low temperature. Under salt stress, esk1 accumulated lower amounts of Na+ in leaves than the WT, and under dehydration stress its metabolic profile and osmotic potential were only slightly impacted. These phenotypes are consistent with the hypothesis of an altered water status in esk1, which actually exhibited basic lower water content (WC) and transpiration rate (TR) than the WT. Taken together, the results suggest that ESK1 does not function as a specific cold acclimation gene, but could rather be involved in water homeostasis.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Abiotic stress refers to environmental constraints, such as extreme light, UV, heavy metals and other chemical pollution, extreme temperatures, drought and hypersalinity. Crops are generally susceptible to these stresses, and improvement of their tolerance remains an agronomic challenge. As a complement to the more current genetic and breeding techniques, understanding of tolerance mechanisms has become a key point in developing the potential background for genetic engineering in the post-genomic era (Bohnert & Jensen 1996). Ecotypes, mutant and transgenic genotypes of the model plant Arabidopsis thaliana (TAGI 2000) were thus screened, to find a genetic basis associated with improved or reduced tolerance. This functional genomic approach to abiotic stress responses requires phenotypic dissection at different levels (Granier et al. 2006), including morphology, development and molecular changes in the transcriptome (Kreps et al. 2002), proteome (Bae et al. 2003), enzyme activities (Gibon et al. 2004) and metabolome (Roessner et al. 2001).

The Arabidopsis mutant esk1 was isolated by Xin & Browse (1998) for its constitutive freezing tolerance without previous acclimation. Cold acclimation designates a range of processes occurring in taxons from temperate regions, like the Arabidopsis ecotype Columbia, improving tolerance to freezing temperatures after a few days at low positive temperatures. The mutant tissues were found to constitutively accumulate proline, a compatible solute, and exhibited changes in lipids composition. This phenotype suggested that ESK1 locus might govern genes involved in cold stress response, and could therefore be an attractive candidate for regulon engineering. More recently, ESK1 was identified as At3g55990, member of a 45 gene family encoding proteins that share the domain of unknown function DUF231, specific to plants (Xin et al. 2007). In the same study, the transcriptome of the mutant was shown, surprisingly, to better overlap with those of drought, ABA and salt-treated plants than with those of cold acclimation. Cold, drought or high salinity all exert cellular osmotic strains, so cross talks between their respective signalling pathways can result in partially similar responses (Xiong, Schumaker & Zhu 2002), like overlapping transcriptome profiles, cell wall and membrane adjustments, oxidative stress management and accumulation of compatible solutes (Hare, Cress & Van Staden 1998; Guan, Zhao & Scadalios 2000; Kreps et al. 2002; Cameron, Teece & Smart 2006). Interestingly, esk1 mutant does not exhibit constitutive expression of cold-regulated (COR) genes, which are cold and dehydration responsive genes induced by transcription factors known as C-repeat binding factors/dehydration-responsive binding factors 1 (CBF/DREB1) (Xin & Browse 1998). ESK1 thus apparently functions in a CBF-independent pathway whose role and positioning among stress response cascades remain to be established.

In the present work, the esk1 metabolome was explored by gas chromatography hyphenated to quadrupole-based mass spectrometry (GC–Q-MS) to provide a broad, comparative, non-targeted screening of its characteristic metabolic phenotypes, and uncover pathways possibly interacting with ESK1 product. In a second approach, primary metabolic phenotypes associated with esk1 mutation, and cold, salt and dehydration responses were compared. Both qualitative and quantitative aspects of metabolic and osmotic reprogramming were considered, with the aim of dissecting ESK1 involvement in acclimation to abiotic stress.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Plant material

Eskimo mutant (esk1-1) and wild-type (WT) Col-0 seeds have been kindly provided by Dr J. Browse (Xin & Browse 1998). Plants were sown on 0.7% agar medium and placed 2 d at 4 °C before transfer at 23 °C for germination. Five-day-old seedlings were placed in hydroponic cultivation system on liquid medium (KNO3 2.5 mm, Ca(NO3)2 2.5 mm, KH2PO4 0.5 mm, MgSO4 1 mm, Na2O3Si 0.1 mm, MnSO4 10 µm, ZnSO4µm, CuSO4 0.5 µm, H3BO3 30 µm, Na2MoO4µm, Co(NO3)2, 6H2O 0.5 µm, Fe-EDTA 27 µm), in a growth chamber (21 °C, 70% relative humidity, 105 µmol photons m−2 s−1, 12 h per day). Esk1 plants were sown 7 d before WT to reach the same growth stage at treatment time.

Treatments

All treatments were applied at the same time in the same growth chamber (see parameters) on 32-day-old (39 for esk1) plants at vegetative stage 1.12 according to Boyes et al. (2001). Salt treatment consisted in 150 mm NaCl added to growth medium. Dehydration treatment was applied by complete removing of liquid growth medium. For cold treatment, plants were placed in polystyrene boxes filled with ice and two ice-packs previously frozen at −80 °C and replaced every 3 h to maintain a temperature of 6 ± 2 °C.

Harvesting and sampling

All treatments lasted 48 h and were followed by a 48 h recovery period under standard conditions. Control and treated plants were harvested at different time-points: 0 (before treatment); 2, 6, 10, 24 and 48 h (end of treatment); and 72 and 96 h (end of recovery). Each sample consisted in seven rosettes pooled and immediately flash frozen, lyophilized at 0.12 mbar during 48 h and stored at −80 °C for subsequent analysis.

Water content (WC) and osmotic parameters measurement and calculation

Water status was investigated at time-points 0, 48 and 96 h, by measurement of three rosettes fresh weight (FW) immediately after harvest and dry weight (DW) after drying 48 h at 60 °C. Shoot relative WC was calculated as WC = (FW − DW)/DW.

To determine transpiration rate (TR), four plants of each genotype were grown until stage 1.12 on individual pots filled with liquid medium; then pots were closed hermetically at rosette base and weighted at 100 h interval to calculate water loss rate (gH2O h−1); rosettes were then cut and dried 48 h at 60 °C. The TR was expressed relatively to shoot DW (gH2O h−1 g−1 DW).

Total osmotically active particles (TOAPs) in shoots were estimated by measuring osmolality (OSL) of 100 µL polar extracts (extraction method is explained in the chapter ‘Quantitative Metabolic Profiling’) with a cryo-osmometer (Digital osmometer, Roebling, Berlin, Germany): TOAP (osmol g−1 DW) = (extract OSL) × (extraction volume)/(extracted DW).

Shoot cell osmotic potential was then calculated according to Van't Hoff equation: π (MPa) = −R.T.TOAP/WC, where T is the absolute temperature and R is the gas constant.

To study treatments and mutation impact on osmotic potential variations, those values were expressed relatively to WT under control condition (WT-ctrl). The osmotic potential measured for a given sample was decomposed into the sum of osmotic potential due to TOAP calculated at constant WC and osmotic potential due to water loss: πsample = πTOAP + πwater loss; πTOAP = −R.T.TOAPsample/WCWT-ctrl and πwater loss = πsample − πTOAP.

Metabolomic analysis

GC–MS profiling analysis was performed essentially according to Colebatch et al. (2004), refined according to Desbrosses, Kopka & Udvardi (2005).

Data acquisition was performed by Masslab 1.4v software (ThermoQuest, Manchester, UK), and chromatograms were deconvoluted manually with automated mass spectral deconvolution and identification system (AMDIS) associated to NIST libraries, by searching two mass spectral fragments for each component. Every peak area of the selected mass traces was expressed as a ratio to the maximum peak surface encountered among all chromatograms for this mass fragment, then the two peak surfaces of a given compound were averaged to obtain the final peak surface; this simple mathematics treatment reduces the impact of extreme values on the final peak surface. Analyte identification was performed on the basis of retention time and mass spectrum match with libraries of reference compounds (Kopka et al. 2005). Data were subsequently expressed relatively to ribitol signal and initial DW of each sample. In this way, it was possible to detect and identify 67 polar metabolites. In addition, 31 compound mass spectra matched well with known metabolites, but differed in retention time; they were thus labelled with the name of library best match compound between brackets, followed by a number when several analytes matched the same compound. Another 30 analytes matched with no known structure and were labelled Xn. Information about identification (database identifier, retention time, match score) and quantification (selected ions for manual deconvolution) is available in Supporting Table S1.

Quantitative metabolic profiling

Dry-frozen plant material was ground in liquid nitrogen, and 25 mg was sampled before addition of 700 µL ethanol containing internal standards at known concentration (l-norleucine and phenyl-βd-glucopyranoside). Samples were shaken and placed in a dry bath at 85 °C until complete evaporation of ethanol. Pellets were then dissolved in 700 µL of ultra-pure water, shaken 1 h at 8 °C and centrifuged 15 min at 12 000 g and 4 °C. The polar extract consisted in the supernatant.

Non-structural carbohydrates (NSCs) and organic acids (OAs) were analysed by GC-FID according to the modified method of Adams et al. (1999). Then, 70 µL of polar extract was frozen at −80 °C and dry-frozen during 24 h. The dry residue was dissolved in 70 µL of 20 mg mL−1 methoxyamine hydrochloride in pyridine at 30 °C for 90 min, afterwards 30 µL of N,O-bis(trimethylsisyl)trifluoroacetamide (BSTFA) was added and samples were submitted to 5 min ultrasonication cycles at 37 °C for 30 min. One microlitre of the mixture was injected in a Trace 2000 GC-FID (Thamo-Fisher Scientific, Waltham, CA, USA) fit out with an AS2000 Autosampler (Thamo-Fisher Scientific), a split/splitless injector (split mode set to 1:25) at 230 °C, a J&W DB5 30 m × 0.32 mm × 0.25 µm column and a FID detector at 250 °C. The gradient temperature was: 5 min at 70 °C, 5 °C min−1 until 220 °C, 2 °C min−1 until 260 °C, 20 °C min−1 until 300 °C and finally 5 min at 300 °C.

Amino acids (AAs) were analysed by HPLC-UV with Waters kit (Waters Corporation, Milford, MA, USA), according to the method described by Cohen & Michaud (1993) and adapted by Bouchereau et al. (1999a).

Peak identification and integration were processed manually with Apex Chromatography Workstation v2.15 (Apex Chromatography Pvt. Ltd, Andhra Pradesh, India). Among 46 peaks, 35 metabolites were reliably identified by comparison of sample chromatograms to standard mixtures of known concentration and quantified in absolute amount after normalization against internal standards and plant material DW (µmol g−1 DW); all calibration curves of standard compounds were linear (R2 > 0.96). The reproducibility of quantification was found between 2 and 22%, depending on metabolites, inherently to their abundance and stability after derivatization. Besides, four compounds were identified by comparison of GC-FID profiles to GC–MS profiles, but could not be quantified (their names are expressed between brackets). Five peaks of unidentified analytes were also considered because their intensity was at least 10-fold above noise and they appeared to discriminate samples (named ‘Un’ for unidentified). Amounts of these 14 non-quantified compounds were expressed in arbitrary unit (i.e. peak surface after normalization). Asparagine and serine, as well as arginine and threonine, could not be separated by HPLC-UV, so the sum of their abundance was taken into account, noted Asn-Ser and Arg-Thr. Full data set is available in Supporting Table S2.

Data mining

Statistical analysis has been performed on Microsoft Excel, Statistica v7.1 (StatSoft, Tulsa, OK, USA) and Metagenalyse, the Golm MPIMP online program (Daub, Kloska & Selbig 2003). Data normalization was improved by log-transformation before performing non-parametric Mann–Whitney U-test. Zero values from signal below-detection limit were replaced by an arbitrary very small value (0.001) for subsequent multivariate analyses: hierarchical cluster analysis (HCA), principal component analysis (PCA) and independent component analysis (ICA).

Inorganic solutes contents

Nitrate, sulphate and chloride were quantified in 100-fold diluted samples by ionic chromatography on a Dionex DX120 (Dionex Corporation, Sunnyvale, CA, USA) with a AS9HC column and ion Pac AG9 HC precolumn. Ions were eluted by Na2CO3 and detected by conductimetry. Sodium and potassium were quantified by flame photometry (Bibby Scientific Ltd T/As Jenway, Essex, England) in 200-fold diluted extracts.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

To explore metabolic changes caused by the ESK1 mutation, WT and esk1 plants were compared in the vegetative growth phase, under control and stress conditions. Metabolite profiling by GC–Q-MS was used for semi-quantitative screening of changes induced by the mutation, in shoot and root tissues, under non-stress conditions. In addition, further analyses were performed on shoots for a fully quantitative assessment of metabolite pool size. This was achieved by combining results from GC-FID, a mostly carbon-sensitive detector with a larger dynamic range compared to quadrupole MS, for the quantification of NSC and OA, and conventional reversed-phase HPLC-UV for a specific detection of AA.

Screening for metabolic responses using GC–MS-based metabolic phenotyping

Unsupervised descriptive multivariate statistics tools were applied to metabolomic results in order to reduce data dimension and point out relevant information related to highest variance (Fiehn et al. 2000). ICA was processed after data reduction by PCA. This method has a higher discriminating power than PCA, by calculating independent components with maximized non-gaussianity from a defined number of principal components of a previous PCA (Scholz et al. 2004).

ICA separated well esk1 and WT samples, and provided better clusters fitting along independent components than PCA, allowing easy determination of discriminating variables (Fig. 1). Metabolomic difference between genotypes thus appeared to be a major framework of the dataset. Another important part of the variance was related to biological replicates, reflecting the high plasticity of the metabolome despite reproducible growth conditions. Plots of variables indicated that differences between mutant and WT are not restricted to a few pathways, but conversely involve most detected metabolites in shoots and roots.

image

Figure 1. Independent component analysis (ICA) plot of WT (○) and esk1 (▵) GC-Q-MS metabolomic data sets of shoots (94 analytes) and roots (112 analytes). ICA was performed from three principal components of a principal component analysis (PCA) and allowed a better ‘genotype–component’ association (groups delimited by dotted line have no statistical significance). Shoots (a and b) and roots (c and d) ICA plots traduce 81.8 and 90.9% of the respective data sets variance. (a) ICA plot of shoots samples (two independent experiments with three biological replicates, each made of seven rosettes pooled). (b) ICA plot of shoots variables. Most discriminative metabolites are tagged. (c) ICA plot of roots samples (one experiment with three biological replicates, each made of seven roots pooled). (d) ICA plot of roots variables. Most discriminative metabolites are tagged.

Download figure to PowerPoint

A few metabolites showed very contrasted content in both genotypes (Supporting Table S1), some of which are labelled on the plots. Proline and the raffinose family oligosaccharide (RFO) relatives: myo-inositol, galactinol, melibiose and raffinose, were all found from 1.4- to 4.7-fold more abundant in esk1 shoots and from 2.1- to 8-fold in roots. Mannitol and GABA were also found respectively 3.7- and 4.9-fold more abundant in esk1 roots. It is worth noting that these particular metabolites are well known to be involved in stress responses, as are polyamines (Cohen 1998; Bouchereau et al. 1999b), which showed constitutively in esk1 tissues some profiles comparable to that expected in WT under stress: namely a fall of spermidine and a putrescine increase (respectively 0.6 and 1.7 times the WT shoot content). Other discriminating compounds were more unexpected, such as nicotianamine, detected only in WT roots.

A logarithmic plot of mean content ratios (Fig. 2) illustrates the relevant differences between genotypes. Metabolites were considered only if their variation was greater than 50% and/or was significant according to a Mann–Whitney U-test (P < 0.05). Most metabolites showed greater levels in the mutant than in the WT, while only few were lowered, such as glutamate and aspartate in shoots. It thus appears that esk1 mutation widely impacts shoot and root metabolomes. Most metabolites, including many abiotic stress markers, are over-accumulated in esk1, which might quantitatively affect cellular osmotic regulation. Together, the results indicate an integrated reprogramming of metabolism in the mutant, rather than isolated metabolic alterations.

image

Figure 2. Comparison of metabolite levels in wild-type (WT) and esk1 shoots and roots. Plot on a logarithmic scale of quotients of mean relative response ratio from esk1 over WT (shoots n = 5; roots n = 3). Negative values represent a lower and positive values a higher metabolite content in esk1. Black bars indicate Mann–Whitney U-test significant differences (P < 0.05). Names between square brackets arise from identifications by best mass spectral match only; they represent analytes that have mass spectral similarity to known metabolites. ‘Xn’ represents unidentified analytes.

Download figure to PowerPoint

Comparison of metabolic profiles induced by esk1 mutation and abiotic stresses

Mutant and WT seedlings were grown for 3 weeks and then submitted for 2 d to non-freezing low temperature, high salinity or water deprivation, before a 2 d recovery period. Rosettes were collected at different time-points for quantitative metabolite content analysis; the results presented are averages of two completely independent experiments (Supporting Table S2).

A PCA was performed on a data set including both genotypes under control and stress conditions. The data set was processed with PCA rather than ICA in order to obtain information on the portion of variance ascribed to each principal component. A three-dimensional (3-D) plot of the first PCs (Fig. 3a) shows the main data trends, namely metabolic shifts associated with treatments and/or esk1 mutation. The Arabidopsis metabolome appeared to be reconfigurated progressively during treatments (from 2 to 48 h); however, when plants where placed back under standard conditions (72 and 96 h), it began to revert to its initial configuration (Fig. 3b,d). This observation is consistent with an acclimation/deacclimation process, and validates the experimental set-up.

image

Figure 3. Principal component analysis (PCA) of wild type (WT) (○) and esk1 (▴) shoot metabolome data under control (inline image), cold (inline image), dehydration (inline image) and saline (inline image) conditions. Numbers indicate treatment period, from 0 to 48 h, and recovery (72 and 96 h). (a) Samples plot on three first PCs (arrows indicate progression of the time-scaled metabolic profiles distortions); (b) and (c) samples and metabolites plots on (PC1,PC2); (d) and (e) samples and metabolites plots on (PC1,PC3). Asp, aspartate; Asn + Ser, asparagines + serine; Glu, glutamate; Gly, glycine; Arg + Thr, arginine + threonine; Gln, glutamine; Ala, alanine; Pro, proline; Cys, cysteine; Tyr, tyrosine; Val, valine; Met, methionine; Orn, ornithine; Lys, lysine; Ile, isoleucine; Leu, leucine; Phe, phenylalanine; Trp, tryptophan; Glyco, glycolate; Succ, succinate; Glyce, glycerate; Fum, fumarate; Mal, malate; Cit, citrate; [Dhasc], [deshydroascorbate]; Fru, fructose; Gal, galactose; Glc, glucose; [Sorb], [sorbitol]; [Asc], [ascorbate]; Myo, myo-inositol; Suc, sucrose; [Cello], [cellobiose]; Treh, trehalose; Malt, maltose; Gent, gentiobiose; Meli, melibiose; Gal-ol, galactinol; Raf, raffinose; Ux, unidentified compound.

Download figure to PowerPoint

The first PC (43.6% of total variance) is associated with WT time-scaled metabolic response to stress, which mainly consisted in a reduction of glutamate, aspartate and citrate pools, while almost all other metabolites accumulated, especially proline, GABA, tyrosine, malate and fructose. More specific responses to dehydration and salt are depicted by PC2 (12%) which traduced the lowest amount of aspartate and high levels of leucine, isoleucine, phenylalanine and tryptophan, but also sucrose and [cellobiose] (Fig. 3b,c). The particular response to cold acclimation was represented by PC3 (11%) associated to over-accumulation of trehalose, maltose, alanine, glutamine and sucrose (Fig. 3d,e). Interestingly, control esk1 coordinates on PC1 seemed closer to stressed than control WT, which suggested that ESK1 mutation led to metabolic changes conforming to that induced by abiotic stresses. This trend was actually confirmed by an HCA (Fig. 4) performed on data excluding those arising from recovery period. It follows from these observations that the esk1 metabolome constitutively resembles the WT metabolome induced by either cold, salt or dehydration treatment. The striking fact concerning cold acclimation is that WT response was the simultaneous combination of two metabolic shifts. One shift occurred along PC1, towards a non-specific stress phenotype. This response involved accumulation of metabolites that were also found to increase under dehydration and high salinity, and were constitutively more abundant in esk1. The second shift component occurred along PC3, towards a specific cold-acclimated metabolome configuration. This profile was also inducible in the mutant under cold treatment (Fig. 3d,e). A more precise picture of similarities and differences between mutation and stress phenotypes is given in Fig. 5. Metabolite contents measured at 48 h in three experimental replicates were averaged and ratios were calculated over WT control. A noticeable group of metabolites showed similar accumulation profiles in esk1 under control conditions and in WT after cold, saline and water deprivation treatments. It reassembled most of the already well-known stress-associated compounds: melibiose, raffinose, galactinol, proline, galactose, fructose and GABA. Esk1 might thus participate in the control of these metabolic pools. Some of these metabolites can be further accumulated in the mutant after exposure to stress, such as proline, galactose and fructose, which means that ESK1 shares the control of these metabolic phenotypes with other independent pathways. On the contrary, some metabolites such as glutamine, trehalose and sucrose were not affected by esk1 mutation, but were significantly involved in the WT cold response. These compounds accumulated in the same way when the mutant was submitted to cold. The case of glutamine must be underlined, because its accumulation was very specific to cold acclimation, while its level was on the contrary found constitutively lower in esk1. Therefore, the observed phenocopying between cold acclimation and esk1 mutation was clearly partial, because some specifically cold-induced metabolic changes were independent of ESK1 control.

image

Figure 4. Classification of esk1 and wild-type (WT) shoot samples based upon their metabolites contents, by Ward hierarchical clustering using city block distances (Ct, control; Co, cold; Sa, salt; Dr, dehydration). The two main groups distinguish on one hand esk1 control plants and WT plants stressed for 24–48 h, and on the other hand WT plants under control conditions or early stages of stress. esk1 metabolic profiles under control conditions thus appears closer to WT stress profiles than to WT control profile.

Download figure to PowerPoint

image

Figure 5. Relative shoot metabolites content after 48 h stress, expressed as logarithmic mean ratios. (*Mann–Whitney U-test P < 0.05). Shoot metabolite contents after 48 h stress were averaged (three independent experiments) and ratios calculated over the WT control mean. Metabolites were sorted according to esk1/WT ratio (first column) decreasing order.

Download figure to PowerPoint

Organic and inorganic solute variations and osmotic balance in response to abiotic stress and esk1 mutation

Osmotic potential of shoot crude extracts was estimated as described in experimental procedures (Supporting Table S3; Fig. 6), and its variations were subsequently expressed as the combination of WC variations and the amount of solutes, measured as the TOAP fraction of DW.

image

Figure 6. Shoot osmotic potentials (π) calculated after 48 h of osmotic stresses in wild-type (WT) and esk1 plants. Respective contributions of total osmotically active particles (TOAPs) (grey) and water content (white) variations were calculated as explained in the Materials and methods section. Dot line materializes π level in WT under control conditions.

Download figure to PowerPoint

Cold treatment applied to WT plants slightly reduced the osmotic potential of shoots. This trend reflected a loss of water that overcompensated a small reduction of TOAP content. Osmotic potential resulting from WT exposure to high salinity could be explained by both a limited loss of water and a huge accumulation of sodium and chloride ions. Conversely, the fall of osmotic potential consecutive to dehydration treatment was only because of a massive loss of water. Under control conditions, esk1 mutant showed a moderately lower osmotic potential compared to WT. (πWT = −0.96 MPa and πesk1 = −1.23, not significant U-test). But this fact hides significant changes in TOAP and WC. WT displayed on average 4325 µosmol g−1 DW, and esk1 only 3649 (U-test, P < 0.05), while WC were 11 and 7.5 g g−1 DW, respectively (U-test, P < 0.05).

These results indicate that esk1 shoots displayed lower TOAP content despite higher levels of almost all metabolites. This apparent contradiction was investigated with further analysis implementing main mineral ions quantification in crude extracts of the third independent experiment (Supporting Table S3). At least 74% of shoot TOAP content was identified. The most abundant family of metabolites found in WT were OA, followed by AA and finally NSC. In esk1, the proportions were changed towards higher metabolite content (+637 µmol g−1 DW) with the highest quantitative increases concerning fumarate (+289 µmol g−1), malate (+188), proline (+76) and to a lower extent glucose (+43). Meanwhile, the three most important reductions involved NO3- (−1096 µmol g−1), K+ (−262) and glutamine (−14). Esk1 organic compound accumulation was actually compensated by loss of the most abundant minerals NO3- and K+. Interestingly, this trend was roughly the same in WT under cold and water deprivation, where the increase in organic solutes (+629 and +1001 µmol g−1, respectively) only compensated a loss of minerals like NO3- (−448 and −428 µmol g−1), thus maintaining TOAP almost constant. Organic solute accumulation induced by esk1 mutation and abiotic stresses thus appears to be balanced by lower amounts of inorganic solutes.

The osmotic and metabolic effects of the different treatments are depicted in Fig. 7. Cold treatment on WT seedlings induced the greatest changes in shoot NSC (reaching 4.4% of TOAP) and glutamine, while water stress induced the most massive AA accumulation in all treatments (up to 15% of TOAP). Saline treatment distinctiveness relied on sodium and chloride incorporation in the tissues (24% of total TOAP each), confirming that these ions were mainly responsible for the increase in TOAP. The mutant response differed from that of WT: during cold treatment, a much larger proportion of organic solutes was accumulated in esk1 shoots. This illustrates again the intact plasticity of esk1 metabolite content under stress (Fig. 5) linked to ESK1-independent regulations. Compared to WT, a lower Na+ and Cl- influx in esk1 shoots was also observed under salt stress, as well as weaker changes occurring during water deprivation. This was better understood considering the impairment of esk1 water balance (Fig. 8). In fact, relative WC in shoots and whole plant TR were found to be about one-third less than that of WT (U-test, P < 0.05). Such reduction of water flow into the plant could explain the lower ion uptake under salt stress and the relative ineffectiveness of dehydration treatment. In the latter case, the mutant roots were actually observed to be still wet after 48 h stress periods, whereas WT had already completely absorbed the residual root water layer and had started to experience effective water deprivation.

image

Figure 7. Contribution of organic and mineral solutes to the total osmotically active particles (TOAP) determined in wild-type (WT) and esk1 shoots. Circle sizes are proportional to relative TOAP content, expressed as WT control percentage (100 µOsmol g−1 DW). Individual contributions of 35 primary metabolites and five main mineral solutes (K+, NO3-, SO42−, Na+, Cl-) were calculated. Metabolites are grouped by biochemical families (AA, NSC, AO) and those representing more than 0.5% of total soluble fraction are tagged.

Download figure to PowerPoint

image

Figure 8. Transpiration rates and relative water content of wild-type (WT) and esk1 shoots. Differences between the two genotypes are significant in both graphs, according to Mann–Whitney U-test (P < 0.05).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

The esk1 phenotype constitutively mimics phenotypic traits of abiotic stress responses, in terms of development, metabolic profile and osmotic status

Mutation at esk1 locus was found to affect several phenotypic traits in a way similar to abiotic stresses. Growth rate for instance is impaired in the mutant, as its rosette FW is only about 20% that of WT plants (3-week-old plants; data not shown). This refers to growth limitations found under various abiotic stresses (Munns 2002), a clear illustration of which was already given for the WT Col-0 grown at 4 °C, being five times smaller in size than at 22 °C (Xin & Browse 1998). From a metabolic point of view, among the main mutation markers, namely metabolites that showed the largest amplitude of variation between esk1 and WT, many compounds are known to be involved in abiotic stress responses (Fig. 2). The range of changes associated with the mutation at the metabolomic level was also consistent with wide transcript modifications already reported, including those of stress-responsive genes (Xin et al. 2007). Data showed redundancies between the two ‘omic’ approaches concerning key metabolites. For instance, proline accumulation in esk1 shoots correlated both with increased expression of P5CS, encoding a major proline formation enzyme, and reduced expression of the catabolic proline dehydrogenase gene (Table 1). This parallel suggests that transcriptional regulation participates to some extent in the control of related metabolic pools, despite some discrepancies between transcript and metabolic quantitative levels, because of the complex nature of metabolite regulation networks (e.g. AT1G56600 transcript was found 20.8-fold higher in esk1, while galactinol showed only a 3.74-fold increase). The effective metabolic phenotype correlation with their transcriptional regulation is also informative as it substantiates the putative function assigned to At1g56600 and At3g50480.

Table 1.  Consistency between esk1 mutation effects on transcriptome and metabolome
 Transcriptomics (Xin et al. 2007) Metabolomics (this paper) 
AGI IDAnnotationsesk1/WTMetaboliteIdentifier in MSRI library (qt_msri_id_eigtms121) of the Golm Metabolome Database (GMD@CSB.DB)esk1/WT
AT2G39800Delta 1-pyrroline-5-carboxylate synthetase B/P5CS B (P5CS2)4.7ProlineEIQTMS_132003-101_METB_1303.4_DL-Proline (2TMS)4.7
AT1G23090Proline oxidase, mitochondrial/osmotic stress-responsive proline dehydrogenase (POX) (PRO1) (ERD5)0.4   
AT2G22240Inositol-3-phosphate synthase isozyme 2/myo-inositol-1-phosphate synthase 2/MI-1-P synthase 2/IPS 22.5Myo-inositolEIQTMS_209002-101-5_METB_2090.3_Inositol, myo- (6TMS)1.3
AT4G39800Inositol-3-phosphate synthase isozyme 1/myo-inositol-1-phosphate synthase 1/MI-1-P synthase 1/IPS 11.5   
AT1G56600Galactinol synthase, putative20.8GalactinolEIQTMS_299002-101_METB_2972.8_Galactinol (9TMS)3.95
AT5G40390Raffinose synthase family protein1.9RaffinoseEIQTMS_337002-101_METB_3353.3_Raffinose (11TMS)3.74
AT3G50480Nicotianamine synthase, putative0.3NicotianamineEIQTMS_259003-101_METB_2606.3_Nicotianamine (4TMS)0.06

The resemblance between esk1 constitutive metabolic phenotypes and that of WT under stress was also noticed in the quantitative profiling experiment (Fig. 4). Total shoot metabolite content increased significantly after 48 h of cold, salt or dehydration stress, and was constitutively greater under control conditions in the mutant shoots (Supporting Table S3; Fig. 7). Besides, the same remarkable imbalance between organic and mineral solutes was observed constitutively in esk1 shoots and in WT shoots after cold acclimation or dehydration; in all cases, it maintained the global TOAP content at a roughly constant level (Fig. 7). This observation might be linked to the growth limitation, because decreased inorganic solutes such as nitrate may mean reduced availability of nutrients for nitrogen assimilation, while increased organic solutes may mean less available building-block molecules for growth. Whatever the reasons, this observation constitutes another common trait between esk1 mutation and stress induced phenotypes.

Such results strongly support the hypothesis, already formulated by Xin & Browse (1998), of a relationship between ESK1 function and stress response pathways. They proposed ESK1 could act as an upstream negative regulator of cold acclimation, submitted to post-transcriptional regulation as its transcription level remained unchanged during cold acclimation (Xin et al. 2007). Interestingly, ESK1 was first considered as part of a CBF-independent suite of cold-acclimation responses, because many COR genes did not show any transcriptional increase in the mutant (Xin & Browse 1998). An alternative explanation stated it could negatively control a sub-regulon of CBF pathway (Gilmour et al. 2000); in this case, the constitutive mutant molecular phenotype would be expected to be circumscribed into the WT cold-acclimated phenotype or into CBF over-expressing plant phenotype. But the set of up- or down-regulated genes in esk1 was found to be much larger than that of cold-induced or CBF2 over-expressing plant (Xin et al. 2007). There was an even higher transcriptomic match between esk1 mutation and ABA or drought treatment than between esk1 mutation and cold treatment. This important result directly questioned the role of ESK1 as a specific cold acclimation regulator.

Comparison of cold-acclimated WT and esk1 phenotypes questions the specific involvement of ESK1 in cold acclimation process

In the present work, the esk1 metabolic phenotype clearly does not completely overlap with the WT cold acclimation phenotype (Fig. 3). More precisely, esk1 mutation phenocopies the non-specific stress profile involving melibiose, raffinose, galactinol, proline, galactose, fructose, GABA and glucose, all highly accumulated under all three abiotic stresses (Fig. 5). But some important metabolic traits of cold acclimation, like trehalose, sucrose and glutamine accumulation (Figs 3 & 5), are not constitutive traits of the esk1 mutant. These particular observations might reflect experimental artefacts rather than coordinated cold responses, but literature abounds on the subject of trehalose and sucrose accumulation during cold acclimation (Cook et al. 2004; Klotke et al. 2004) and on their contribution to freezing tolerance by reducing the rate of cellular ice migration (Gusta et al. 2004) and improving membrane stabilization (Crowe et al. 1990; Wolfe & Bryant 1999). The sucrose cryoprotective effect was also demonstrated in transgenic plants over-expressing sucrose phosphate synthase (Strand et al. 2003). Moreover, despite a lack of knowledge about its physiological meaning, glutamine accumulation has been reported so far in cold-acclimated Arabidopsis plants (Cook et al. 2004; Kaplan et al. 2004; Klotke et al. 2004). This phenotype of accumulation appears here to be strictly associated to cold stress (Fig. 5); it thus probably reflects integrated physiological adjustments rather than treatment side effects. Considering transcription of At1g09240 (putative nicotianamine synthase) revealed similar information, because it was found inhibited by half in the mutant (Xin et al. 2007), a trend consistent with the undetectable level of nicotianamine in esk1 roots (Supporting Table S1), while cold treatment of the WT conversely up-regulated At1g09240 expression by 3.7-fold (Kaplan et al. 2007). From these various observations, it follows that the mutant esk1, selected for its constitutive resistance to freezing, is perhaps not mutated on a gene of cold acclimation. The present results rather suggest it could drive a subclass of core metabolic responses which are common to various abiotic stresses.

The mutant shows a dehydrated phenotype that would explain its constitutive freezing tolerance and suggests ESK1 may function in the control of water homeostasis

An interesting trait of the esk1 phenotype is its impairment in water conduction and/or absorption (Fig. 8). Present investigation of water status evolution in the shoots of WT plants also revealed reduced WC after 48 h cold treatment (76% of control, U-test P < 0.05), and a consistent increase of the osmotic potential absolute value (117% of control). Gusta et al. (2004) observed that cold-acclimated Canola (Brassica napus) leaves had a lower WC than non-acclimated leaves, and initiated freezing much more slowly. They concluded that a moderated dehydration could be a component of plant acclimation strategy. It follows that the mutant water status is not contradictory to a possible involvement of ESK1 in cold acclimation process. Another possibility is that ESK1 rather participates in plant dehydration response or water homeostasis. Indeed, it has already been observed that drought treatments can confer subsequent freezing tolerance (Mantyla, Lang & Palva 1995). Dehydration stress tolerance mechanisms are likely to help cold acclimation, based among other mechanisms on supercooling, a process by which elevated cellular concentration of solutes lowers the ice nucleation temperature. This phenomenon was demonstrated to occur in Arabidopsis after cold acclimation and to be constitutive in esk1 (Reyes-Diaz et al. 2006). Enhanced supercooling relies not only upon solute concentration, but also upon their physicochemical properties and subcellular compatmentalization (Gagneul et al. 2007). Such a mechanism is consistent with a lower osmotic potential in esk1 leaves and metabolic reconfiguration towards storage of compatible solutes (Brown & Simpson 1972; Yancey et al. 1982) with osmo- or cryo-protective properties (Gilles 1997; Yancey 2005). In WT plants, comparable metabolic shifts occurred under stress that traduced the reorientation of metabolic fluxes towards such preferential pathways, and such observations open the way for identification of new biochemical stress markers among unidentified substances, like X25 or U10.

CONCLUSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Discussion on ESK1 function has been primarily oriented towards a possible involvement in cold acclimation processes, based on known models of regulatory pathways, including cellular stress probes, transcription factors and protein activation cascades (Xin & Browse 1998; Xin et al. 2007). In the present work, multivariate statistics on metabolic markers and water phenotyping of both esk1 mutation and environmental stress responses led to reconsider the role of ESK1 as an upstream repressor of cold acclimation. Two reasonable hypotheses would match with information available to date about esk1 mutant. (1) The first extends the current theory of a post-transcription-activated negative regulator, positioned upstream in a CBF-independent signalling cascade. The new statement brought by our results is that ESK1 would participate in non-specific environmental stress acclimation responses which all require reconfiguring of the internal water status and/or osmo-adaptation. In this case, the mutant phenotype would reflect the constitutive activation of ESK1-dependant stress responses, in the absence of any external or internal strain. This statement leads to the important conclusion that osmotic changes, altered water and mineral nutrition, as well as growth modulation, are integrated plant responses to stress (some of which controlled by ESK1), rather than primary disruptions caused by environmental pressure. (2) The alternative hypothesis completely questions the involvement of ESK1 in a regulatory process. Indeed, ESK1 belongs to a gene family sharing a domain of unknown function (Xin et al. 2007); there is thus no evidence that ESK1 belongs to a signalling pathway, and one could rather propose that it positively participates in water homeostasis, i.e. by structurally facilitating water conduction in plants. In this case, its inactivation in the homozygous esk1 mutant would logically result in an effective internal dehydration strain, responsible for subsequent stress response activation, nutritional disorders and growth limitation. The reported improved freezing tolerance would then constitute an unexpected side effect of a constitutive acclimation to dehydration.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

We thank Z. Xin and J. Browse for kindly providing esk1 seeds, and A. Erban and I. Fehrle for technical assistance in metabolomics. This work was supported by a grant from La Région Bretagne, France.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information
  • Adams M.A., Chen Z.L., Landman P. & Colmer T.D. (1999) Simultaneous determination by capillary gas chromatography of organic acids, sugars, and sugar alcohols in plant tissue extracts as their trimethylsilyl derivatives. Analytical Biochemistry 266, 7784.
  • Bae M.S., Chop E.J., Choi E.-Y. & Park O.K. (2003) Analysis of the Arabidopsis nuclear proteome and its response to cold stress. The Plant Journal 36, 652663.
  • Bohnert H.J. & Jensen R.G. (1996) Strategies for engineering water-stress tolerance in plants. Trends in Biotechnology 14, 8997.
  • Bouchereau A., Duhaze C., Martin-Tanguy J., Guegan J.P. & Larher F.R. (1999a) Improved analytical methods for determination of nitrogenous stress metabolites occurring in Limonium species. Journal of Chromatography A 836, 209221.
  • Bouchereau A., Aziz A., Larher F. & Martin-Tanguy J. (1999b) Polyamines and environmental challenges, recent development. Plant Science 140, 103125.
  • Boyes D.C., Zayed A.M., Ascenzi R., McCaskill A.J., Hoffman N.E., Davis K.R. & Gorlach J. (2001) Growth stage-based phenotypic analysis of Arabidopsis: a model for high throughput functional genomics in plants. The Plant Cell 13, 14991510.
  • Brown A.D. & Simpson J.R. (1972) Water relations of sugar-tolerant yeasts, the role of intracellular polyols. Journal of General Microbiology 72, 589591.
  • Cameron K.D., Teece M.A. & Smart L.B. (2006) Increased accumulation of cuticular wax and expression of lipid transfer protein in response to periodic drying events in leaves of tree tobacco. Plant Physiology 140, 176183.
  • Cohen S.A. & Michaud D.P. (1993) Synthesis of a fluorescent derivatizing reagent, 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate, and its application for the analysis of hydrolysate amino acids via high-performance liquid chromatography. Analytical Biochemistry 211, 19.
  • Cohen S.S. (1998) Putrescine and osmotic stress. In A Guide to the Polyamines (ed. S.S.Cohen), pp. 404405. Oxford University Press, New York, USA.
  • Colebatch G., Desbrosses G.G., Ott T., Krusell L., Montanari O., Kloska S., Kopka J. & Udvardi M.K. (2004) Global changes in transcription orchestrate metabolic differentiation during symbiotic nitrogen fixation in Lotus japonicus. The Plant Journal 39, 487512.
  • Cook D., Fowler S., Fiehn O. & Thomashow M.F. (2004) A prominent role for the CBF cold response pathway in configuring the low-temperature metabolome of Arabidopsis. Proceedings of the National Academy of Sciences of the United States of America 101, 1524315248.
  • Crowe J.H., Carpenter J.F., Crowe L.M. & Anchordoguy T.J. (1990) Are freezing and dehydration similar stress vectors? A comparison of modes of interaction of stabilizing solutes with biomolecules. Cryobiology 27, 219231.
  • Daub C.O., Kloska S. & Selbig J. (2003) MetaGeneAlyse: analysis of integrated transcriptional and metabolite data. Bioinformatics 19, 23322333.
  • Desbrosses G.G., Kopka J. & Udvardi M.K. (2005) Lotus japonicus metabolic profiling. Development of gas chromatography–mass spectrometry resources for the study of plant–microbe interactions. Plant Physiology 137, 13021318.
  • Fiehn O., Kopka J., Dörmann P., Altmann T., Trethewey R.N. & Willmitzer L. (2000) Metabolite profiling for plant functional genomics. Nature Biotechnology 18, 11571161.
  • Gagneul D., Aïnouche A., Duhazé C., Lugan R., Larher F.R. & Bouchereau A. (2007) A reassesment of the function of the so-called compatible solutes in the halophytic Plumbaginacae Limonium latifolium. Plant Physiology 144, 15981611.
  • Gibon Y., Blaesing O.E., Hanneman J., Carillon P., Höne M., Hendriks J.H.M., Palacios N., Cross J., Selbig J. & Stitt M. (2004) A robot-based platform to measure multiple enzyme activities and transcript levels in Arabidopsis using a set of cycling assays, comparison of changes of enzyme activities and transcript levels during diurnal cycles and in prolonged darkness. The Plant Cell 16, 33043325.
  • Gilles R. (1997) ‘Compensatory’ organic osmolytes in high osmolarity and dehydration stresses, history and perspectives. Comparative Biochemistry and Physiology A 117, 279290.
  • Granier C., Aguirrezabal L., Chenu K., et al. (2006) PHENOPSIS, an automated platform for reproductible phenotyping of plant responses to soil water deficit in Arabidopsis thaliana permitted the identification of an accession with low sensitivity to soil water deficit. New Phytologist 169, 623635.
  • Guan L.M., Zhao J. & Scadalios J.G. (2000) Cis-elements and Trans-factors that regulate expression of the maize Cat1 antioxidant gene in response to ABA and osmotic stress, H2O2 is the likely intermediary signalling molecule for the response. The Plant Journal 22, 8795.
  • Gusta L.V., Wisniewski M., Nesbitt N.T. & Gusta M.L. (2004) The effect of water, sugars, and proteins on the pattern of ice nucleation and propagation in acclimated and nonacclimated canola leaves. Plant Physiology 135, 16421653.
  • Hare P.D., Cress W.A. & Van Staden J. (1998) Dissecting the roles of osmolyte accumulation during stress. Plant, Cell & Environment 21, 535553.
  • Gilmour S.J., Sebolt A.M., Salazar M.P., Everard J.D. & Thomashow M.F. (2000) Overexpression of the Arabidopsis CBF3 transcriptional activator mimics multiple biochemical changes associated with cold acclimation. Plant Physiology 124, 18541865.
  • Kaplan F., Kopka J., Haskell D.W., Zha W., Schiller K.C., Gatzke N., Sung D.Y. & Guy C.L. (2004) Exploring the temperature-stress metabolome of Arabidopsis. Plant Physiology 136, 41594168.
  • Kaplan F., Kopka J., Sung D.Y., Zhao W., Popp M., Porat R. & Guy C.L. (2007) Transcript and metabolite profiling during cold acclimation of Arabidopsis reveals an intricate relationship of cold-regulated gene expression with modifications in metabolite content. The Plant Journal 50, 967981.
  • Klotke J., Kopka J., Gatzke N. & Heyer A.G. (2004) Impact of soluble sugar concentrations on the acquisition of freezing tolerance in accessions of Arabidopsis thaliana with contrasting cold adaptation – evidence for a role of raffinose in cold acclimation. Plant, Cell & Environment 27, 13951404.
  • Kopka J., Schauer N., Krueger S., et al. (2005) GMD@CSB.DB: the Golm Metabolome Database. Bioinformatics 21, 16351638.
  • Kreps J.A., Wu Y., Chang H.-S., Zhu T., Wang X. & Harper J.F. (2002) Transcriptome changes for Arabidopsis in response to salt, osmotic, and cold stress. Plant Physiology 130, 21292141.
  • Mantyla E., Lang V. & Palva E.T. (1995) Role of abscisic acid in drought-induced freezing tolerance, cold acclimation, and accumulation of LT178 and RABI8 proteins in Arabidopsis thaliana. Plant Physiology 107, 141148.
  • Munns R. (2002) Comparative physiology of salt and water stress. Plant, Cell & Environment 25, 239250.
  • Reyes-Diaz M., Ulloa N., Zuniga-Feest A., Gutierrez A., Gidekel M., Alberdi M., Corcuera L.J. & Bravo L.A. (2006) Arabidopsis thaliana avoids freezing by supercooling. Journal of Experimental Botany 57, 36873696.
  • Roessner U., Luedemann A., Brust D., Fiehn O., Linke T., Willmitzer L. & Fernie A.R. (2001) Metabolic profiling allows comprehensive phenotyping of genetically or environmentally modified plant systems. The Plant Cell 13, 1129.
  • Scholz M., Gatzek S., Sterling A., Fiehn O. & Selbig J. (2004) Metabolite fingerprinting: detecting biological features by independent component analysis. Bioinformatics 20, 24472454.
  • Strand A., Foyer C.H., Gustafsson P., Gardestromn P. & Hurry V. (2003) Altering flux through the sucrose biosynthesis pathway in transgenic Arabidopsis thaliana modifies photosynthetic acclimation at low temperatures and the development of freezing tolerance. Plant, Cell & Environment 26, 523535.
  • TAGI (The Arabidopsis Genome Initiative) (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796815.
  • Wolfe J. & Bryant G. (1999) Freezing, drying and/or vitrification of membrane-solute-water systems. Cryobiology 39, 103129.
  • Xin Z. & Browse J. (1998) Eskimo1 mutants of Arabidopsis are constitutively freezing-tolerant. Proceedings of the National Academy of Sciences of the United States of America 95, 77997804.
  • Xin Z., Mandaokar A., Chen J., Last R.L. & Browse J. (2007) Arabidopsis ESK1 encodes a novel regulator of freezing tolerance. The Plant Journal 49, 786799.
  • Xiong L., Schumaker K.S. & Zhu J-K. (2002) Cell signaling during cold, drought, and salt stress. The Plant Cell 14, 51655183.
  • Yancey P.H. (2005) Organic osmolytes as compatible, metabolic and counteracting cryoprotectants in high osmolarity and other stresses. Journal of Experimental Biology 208, 28192830.
  • Yancey P.H., Clark M.E., Hand S.C., Bowlus R.D. & Somero G.N. (1982) Living with water stress; evolution of osmolyte system. Science 217, 12141222.

Supporting Information

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Table S1. Relative contents of metabolites determined by GC-Q-MS, in WT and esk1 shoots and roots. (Rep, replicates; ND, not determined; SE, standard error; U-Test, P value of a Mann–Whitney non-parametric comparison test between esk1 and WT contents).

Table S2. Absolute contents (μmol g−1 DW) of metabolites in esk1 and WT shoots, after periods of abiotic stress treatments and recovery.

Table S3. Absolute contents (μmol g−1 DW) of metabolites and minerals, osmotic and water status of esk1 and WT shoots, after 48 h of various abiotic stress treatments.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
PCE_1898_sm_Table_S1.xls110KSupporting info item
PCE_1898_sm_Table_S2.xls146KSupporting info item
PCE_1898_sm_Table_S3.xls38KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.