Comparison of freezing tolerance, compatible solutes and polyamines in geographically diverse collections of Thellungiella sp. and Arabidopsis thaliana accessions

Background Thellungiella has been proposed as an extremophile alternative to Arabidopsis to investigate environmental stress tolerance. However, Arabidopsis accessions show large natural variation in their freezing tolerance and here the tolerance ranges of collections of accessions in the two species were compared. Results Leaf freezing tolerance of 16 Thellungiella accessions was assessed with an electrolyte leakage assay before and after 14 days of cold acclimation at 4°C. Soluble sugars (glucose, fructose, sucrose, raffinose) and free polyamines (putrescine, spermidine, spermine) were quantified by HPLC, proline photometrically. The ranges in nonacclimated freezing tolerance completely overlapped between Arabidopsis and Thellungiella. After cold acclimation, some Thellungiella accessions were more freezing tolerant than any Arabidopsis accessions. Acclimated freezing tolerance was correlated with sucrose levels in both species, but raffinose accumulation was lower in Thellungiella and only correlated with freezing tolerance in Arabidopsis. The reverse was true for leaf proline contents. Polyamine levels were generally similar between the species. Only spermine content was higher in nonacclimated Thellungiella plants, but decreased during acclimation and was negatively correlated with freezing tolerance. Conclusion Thellungiella is not an extremophile with regard to freezing tolerance, but some accessions significantly expand the range present in Arabidopsis. The metabolite data indicate different metabolic adaptation strategies between the species.


Background
Low temperatures and freezing impose major limitations on plant growth and development and limit the productivity of crop plants in large parts of the world. Plants from temperate regions increase in freezing tolerance during exposure to low but nonfreezing temperatures for a period of days to weeks, a process termed cold acclimation. This is accompanied by massive changes in gene expression and metabolite composition [1][2][3], including increased levels of compatible solutes such as sugars, proline and polyamines that potentially contribute to cellular freezing tolerance.
The majority of molecular studies of plant freezing tolerance and cold acclimation have been performed in Arabidopsis thaliana. In addition to forward and reverse genetics, the analysis of natural variation has become an increasingly useful approach in the analysis of complex adaptive traits in this species (see [4][5][6] for reviews). Arabidopsis accessions are widely distributed throughout the Northern hemisphere, spanning diverse growth environments. It can therefore be expected that they harbour phenotypic and genetic variation that is advantageous for adaptation to various climatic conditions. Several studies have shown significant natural variation in the responses of Arabidopsis accessions to low temperature [7][8][9][10][11][12][13]. However, Arabidopsis is not an extremophile and it could be expected that more freezing tolerant species have evolved different or additional protective mechanisms that cannot be found in this species.
Thellungiella salsuginea is an emerging plant model species that has been suggested to possess the characteristics of an extremophile, i.e. high tolerance of salinity, freezing, nitrogen-deficiency and drought stress [14][15][16][17][18][19]. The genus Thellungiella is part of the Brassicaceae family and therefore related to Arabidopsis thaliana [20,21]. T. salsuginea resembles Arabidopsis in many features such as short life cycle, self-fertility, transformation by the floral-dip method and a genome size approximately twice that of Arabidopsis [17]. The genome of the closely related species T. parvula has recently been sequenced [22]. Similar to Arabidopsis, also in T. salsuginea different accessions have been identified and the Shandong and Yukon accessions, which originate from China and Canada, respectively, have frequently been used to investigate responses to abiotic stresses [21]. However, no systematic investigation of natural variation in the stress tolerance of Thellungiella has been published to date.
Here we present such a study, investigating the freezing tolerance and cold acclimation responses of 14 T. salsuginea accessions and of the two closely related species T. halophila and T. botschantzevii. We compare these data to the results of a recent study on 54 Arabidopsis accessions [13]. Our results suggest that the freezing tolerance after cold acclimation of the Thellungiella accessions extends to lower temperatures than the freezing tolerance of the most tolerant Arabidopsis accessions. In addition, the data provide the first evidence for a different metabolic acclimation strategy in Thellungiella compared to Arabidopsis.  [7,13].

Seeds of the
Seeds of the Thellungiella accessions were sown in soil and exposed to 4°C in a growth cabinet at 16 h day length with 90 μE m -2 s -1 for one week to promote germination. Seedlings were transferred to a greenhouse at 16 h day length with light supplementation to reach at least 200 μE m -2 s -1 at a temperature of 20°C during the day and 18°C during the night for 8 weeks (nonacclimated plants). For cold acclimation, plants were transferred to a 4°C growth cabinet under the conditions described above for an additional 14 days. Arabidopsis plants were grown and acclimated under identical conditions [7,11], but were only grown under nonacclimating conditions for 6 weeks to reach the same developmental state.

Freezing tolerance assays
Freezing damage was determined as electrolyte leakage after freezing of detached leaves to different temperatures as described in detail in previous publications [7,11]. Briefly, series consisting of three rosette leaves taken from three individual plants were placed in glass tubes containing 300 μl of distilled water. The tubes were transferred to a programmable cooling bath set to −1°C, control samples were left on ice during the entire experiment. After 30 min of temperature equilibration at −1°C, ice crystals were added to the tubes to initiate freezing. After another 30 min, the samples were cooled at a rate of 4°C/h. Over a temperature range of −1°C to −30°C samples were taken from the bath and thawed slowly on ice over night. Leaves were then immersed in distilled water and placed on a shaker for 16 h at 4°C. Electrolyte leakage was determined as the ratio of conductivity measured in the water before and after boiling the samples. The temperature of 50% electrolyte leakage (LT 50 ) was calculated as the LOG EC 50 value of sigmoidal curves fitted to the leakage values using the software GraphPad Prism3.

Sugar analysis
Two leaves from plants that were also used in the freezing tolerance assays were frozen in liquid nitrogen immediately after sampling and homogenized using a ball mill "Retsch MM 200" (Retsch, Haan, Germany). Soluble sugars were extracted and quantified by high performance anion exchange chromatography (HPAEC) using a CarboPac PA-100 column on an ICS3000 chromatography system (Dionex, Sunnyvale, CA) as described previously [24].

Proline measurements
Proline content was measured from the ethanolic extracts that were also used for sugar determination following a method modified from a previously described procedure [25,26]. The extracts were diluted 10-fold with distilled water and 100 μl were combined with 100 μl of glacial acetic acid and 100 μl of 2.5% (w/v) acid ninhydrine reagent [26]. The mixture was incubated at 95°C for 1 h and then for 10 min on ice. The reaction mixture was extracted with 500 μl of toluene and the ninhydrine absorbance was measured in the toluene phase at 520 nm in a spectrophotometer.

Polyamine measurements
Leaf samples (100-200 mg) were homogenized with a ball mill, extracted in 1 ml of 0.2 N perchloric acid for 1 h at 4°C to extract free polyamines and centrifuged at 16000 x g at 4°C for 30 min. Since we detected only very low levels of bound polyamines in our samples (data not shown), these were not further investigated. To 100 μl aliquots of the supernatants, 110 μl of 1.5 M sodium carbonate and 200 μl of dansyl chloride (7.5 mg/ml in acetone; Sigma, Munich, Germany) were added. In addition, 10 μl of 0.5 mM diaminohexane were added as an internal standard. After 1 h incubation at 60°C in the dark, 50 μl of a 100 mg/ml proline solution was added to bind free dansyl chloride [27]. After 30 min incubation at 60°C Latitude and longitude of the collection sites is extracted as described [23].
in the dark, dansylated polyamines were extracted with 250 μl toluene, dried in a vacuum centrifuge and dissolved in 100 μl methanol. Analyses were performed with a reverse phase LC-18 column (Supelco, Munich, Germany) on a HPLC system (Dionex) consisting of a gradient pump (model P 580), an automated sample injector (ASI-100) and a fluorescence detector (RF 2000). Twenty μl samples were injected, polyamines were eluted with a linear gradient of from 70% to 100% (v/v) methanol in water at a flow rate of 1 ml/min and detected at an excitation wavelength of 365 nm and an emission wavelength of 510 nm. Data were analyzed using the Dionex Chromeleon software and quantification was performed with calibration curves obtained from the pure substances.

Statistics
Correlation tests were performed using Pearson's product-moment correlation analysis in the R statistics package [28].

Results
Establishment of a collection of natural Thellungiella spec. accessions We investigated the cold acclimation and freezing tolerance of 16 different Thellungiella accessions (Table 1). Of these, 14 belong to the species T. salsuginea and one each to T. halophila (Bayanaul) and T. botschantzevii (Saratov). Four of the accessions originate from the continental USA or Canada and five from China and substantial work has been performed previously on the accessions Yukon and Shandong (see [21] for a review).
In addition, seven accessions were collected for this study from different sites in Russia and Kazakhastan to enrich our collection for accessions from very cold climates (Table 1). Thus the geographical origins of these accessions span the Northern hemisphere (between 33°N and 61°N) from 130°E to 135°W.  Natural variation in the freezing tolerance of Thellungiella accessions The freezing tolerance of the Thellungiella accessions was determined before (nonacclimated; NA) and after (acclimated; ACC) two weeks of cold acclimation at 4°C (Figure 1). The results show strong natural variation in the freezing tolerance of Thellungiella. Higher variation was found in acclimated (LT 50 from −9.12°C (Jiangsu) to −15.21°C (Tuva)), than in nonacclimated plants (LT 50 from −5.70°C (Xinjiang) to −7.40°C (Bayanaul)). In addition, Tuva showed the highest acclimation capacity (8.22°C difference in LT 50 between NA and ACC plants) and Jiangsu the lowest (3.28°C).
Thellungiella is generally considered to be much more freezing tolerant than Arabidopsis [29]. The fact that we have recently determined the freezing tolerance of 54 Arabidopsis accessions under exactly the same conditions as used here for Thellungiella [13] provided a unique opportunity to test this assumption. Figure 1 clearly shows that the range of LT 50 values was not different between Arabidopsis and Thellungiella in the nonacclimated state, but that some Thellungiella accessions (Tuva, Saratov, Altai 1 and 2, Bayanaul) reached lower LT 50 values after cold acclimation.
No significant correlations at p < 0.05 were found between the latitude of the geographical origin of the accessions and their LT 50 either before or after cold acclimation. However, LT 50 ACC was significantly correlated with the average minimum habitat temperature recorded during the coldest month of the growth season, while no such correlation was found before cold acclimation ( Figure 2).

Accumulation of sugars and proline in response to cold
The accumulation of compatible solutes such as sugars and proline is commonly observed during cold acclimation [2,3]. We therefore measured the amounts of glucose (Glc), fructose (Fru), sucrose (Suc), raffinose (Raf ) and proline (Pro). Figures 3 and 4 show that the contents of sugars and Pro increased strongly in leaf samples of most Thellungiella accessions during cold acclimation. As observed previously for Arabidopsis [13], there were also some Thellungiella accessions that failed to accumulate a particular solute. For instance, Yakutsk showed an extremely low level of Fru after acclimation, while Dillibrough did not accumulate any Pro in the cold. We further explored the functional significance of these compatible solutes in leaf freezing tolerance by correlation analysis. The contents of sugars and Pro were not significantly correlated with LT 50 under nonacclimated condition except for Glc (r = −0.619, p = 0.011). After cold acclimation, only the contents of Suc ( Figure 5) and Pro ( Figure 6) were significantly positively correlated with freezing tolerance (i.e. negative correlation with LT 50 ), while the content of Fru was negatively correlated. In other words, the contents of Suc and Pro were higher in the more freezing tolerant accessions, while the contents of Fru was higher in the more sensitive accessions.
Since we had previously also determined the sugar and Pro contents of the leaves of 54 Arabidopsis accessions [13], we could now directly compare the role of compatible solutes in the acclimated freezing tolerance of these species (Figures 5 and 6). While Glc, Fru and Suc contents were significantly positively correlated with freezing tolerance in Arabidopsis, this was only true for Suc in Thellungiella. However, the overall pool sizes of these sugars were similar, although some Arabidopsis accessions accumulated two-to three-fold higher amounts of Glc. The most striking differences were found for Raf and Pro. The amounts of Raf in the leaves of the most freezing tolerant acclimated Arabidopsis accessions were several-fold higher than those of any Thellungiella accessions. For example, the most freezing tolerant  Fig. 1 for LT 50 and in Fig. 3 for sugar contents. The data for Arabidopsis (open symbols) are taken from [13].
Arabidopsis accession (N14) contained about 10.5 μmol Raf g -1 FW, while all Thellungiella accessions accumulated less than 3 μmol g -1 FW. On the other hand, Pro levels were much higher in Thellungiella than in Arabidopsis leaves and there was no significant correlation between Pro contents and LT 50 ACC in Arabidopsis ( Figure 6). Some Thellungiella accessions already contained more Pro in their leaves in the nonacclimated state (up to 18.5 μmol g FW -1 ) than any Arabidopsis accession after cold acclimation (up to 14.9 μmol g FW -1 ).

Polyamine contents in Thellungiella and Arabidopsis accessions
There is evidence from several studies that polyamines may play important roles in the development of plant freezing tolerance (see [30] for a recent review). We have therefore measured the amounts of free putrescine (Put), spermidine (Spd) and spermine (Spm) in leaf samples from all Thellungiella accessions both before and after cold acclimation (Figure 7). Since no published data on the polyamine contents of different Arabidopsis accessions under these conditions were available, we also determined the respective polyamines in nine Arabidopsis accessions that span a wide range of freezing tolerance [7,13]. In general, the levels of Put and Spd were similar in Thellungiella and Arabidopsis and they either increased during cold acclimation or remained unaltered in some accessions (e.g. Dillibrough and Hebei; Te-0 and Can-0). However, the levels of Spm were much higher in nonacclimated Thellungiella leaves, but were drastically reduced during cold acclimation. In Arabidopsis, Spm levels were generally lower and only decreased in a few accessions during acclimation. In both species free Spd was the predominant polyamine under both conditions. No significant correlations were found among the Put, Spd or Spm contents and LT 50 NA (not shown) or Put or Spd contents and LT 50 ACC in Thellungiella (Figure 8). However, there was a significant correlation between Spm content and LT 50 ACC in Thellungiella, indicating that higher leaf freezing tolerance was correlated with a lower pool size of free Spm. In Arabidopsis, no correlations among LT 50 and polyamine pool sizes were observed under either condition.

Discussion
Thellungiella has been proposed as an alternative model species to Arabidopsis to investigate plant abiotic stress tolerance mechanisms. Thellungiella shares many features with Arabidopsis that make it an attractive candidate for both physiological and molecular studies [14,21,29]. The main argument in favor of Thellungiella, however, is that it is considered an "extremophile" that is much more tolerant to various stresses than Arabidopsis. On the other hand, it has been shown that there is considerable natural variation between different accessions of Arabidopsis that results in different levels of tolerance under various environmental growth and stress conditions (see e.g. [6] for a recent review). This natural variation has been investigated most extensively for cold acclimation and freezing tolerance [7,8,10,12,13]. Since natural accessions are also available for Thellungiella this opens the unique possibility to directly compare the range of stress tolerance and possible differences in adaptive mechanisms between these species.
In the present study, we have for the first time compared the range of natural variation in the freezing tolerance of Arabidopsis and Thellungiella. We conclude from the wide overlap in the freezing tolerance that at least with regard to this trait Thellungiella should not be considered an extremophile. Its range of freezing tolerance, however, extends to lower temperatures than that of Arabidopsis with about one-third of the available Thellungiella accessions more freezing tolerant than any Arabidopsis accession. The acclimated freezing tolerance of Thellungiella was positively correlated with the average minimum habitat temperature recorded during the coldest month of the growth season, consistent with previous results for Arabidopsis [7,12].
Only the freezing tolerance of the Yukon accession of Thellungiella has previously been reported in the literature [16]. LT 50 values of −13°C for nonacclimated and −18.5°C for cold acclimated plants were recorded when whole-plant survival was evaluated. These temperatures are substantially lower than the −6.4°C (NA) and −11.7°C (ACC) obtained from our electrolyte leakage measurements. However, corresponding electrolyte leakage data in [16] suggest a similar temperature range to our results although no LT 50 values were given. In addition, since no direct comparison with Arabidopsis was presented, any comparison between the species remained speculative in this paper. From the comparison presented here we suggest that although Thellungiella may not be an extremophile with regard to freezing tolerance, its range of freezing tolerance after cold acclimation clearly extends beyond Arabidopsis. We therefore consider Thellungiella a useful additional model species to identify superior or alternative freezing tolerance mechanisms.
During cold acclimation in Arabidopsis, the composition of the metabolome is strongly changed (see [1] for a review). The pool sizes of several metabolites are increased and there are significant differences in the cold-responsive metabolomes of different Arabidopsis accessions [7,31,32]. Significantly, the leaf contents of the four sugars Glc, Fru, Suc and Raf were linearly correlated with leaf freezing tolerance [8,11,13] and these sugars were also found among a small group of metabolites that could be used to predict the freezing tolerance of several Arabidopsis genotypes with high accuracy [32]. In addition, although the Pro contents of the leaves also increased during cold acclimation, there was no correlation with freezing tolerance among the 54 accessions investigated previously [13] and Pro was also not among the predictive metabolites [32].
The present data suggest that the role of these five compatible solutes may be significantly different between Arabidopsis and Thellungiella. Among the sugars, a positive correlation with acclimated freezing tolerance was only observed for Suc, while there was actually a negative correlation for Fru. In addition, the Thellungiella accessions did not accumulate Raf to the same extent as Arabidopsis. Instead, Thellungiella accumulated much higher amounts of Pro during cold acclimation and we found a significant correlation with acclimated freezing tolerance. The accumulation of compatible solutes, particularly Suc and Pro, was not only found in Thellungiella plants during cold acclimation. Especially Pro contents also increased much more than in Arabidopsis when plants were challenged with high NaCl concentrations [15,33,34] suggesting a different metabolic adaptation strategy between the species under abiotic stress conditions. Obviously, this hypothesis has to be tested in the future by metabolomic approaches using appropriate collections of accessions from both species.
We would like to stress at this point that it is highly unlikely that the differences in compatible solute content are the only reason for the observed differences in freezing tolerance. Although the constitutively freezing tolerant esk1 mutant in Arabidopsis shows a high accumulation of Pro under nonacclimated conditions [35], it also shows hundreds of changes in gene expression, making it impossible to attribute the higher freezing tolerance to a single factor [36]. Similarly, although freezing tolerance in Arabidopsis is strongly correlated with Raf content, a knock-out mutant of the raffinose synthase gene in Col-0 resulted in the absence of Raf in the cold acclimated leaves without an impairment of freezing tolerance [37]. All these findings emphasize the wellknown fact that plant freezing tolerance is a multigenic, quantitative trait. In addition, the present data indicate that even in closely related species, different metabolites may be important.
One additional class of metabolites that has frequently been implicated in plant freezing tolerance are polyamines [30]. They are thought to be involved in many aspects of plant growth, development and stress tolerance (see [38][39][40] for reviews). Their exact functions in these processes have not been completely elucidated, but it was demonstrated that Put is an essential component of the cold acclimation process in Arabidopsis [41]. This is at least in part mediated through a role in the regulation of ABA biosynthesis. The measurement of free polyamine levels in several accessions of both Arabidopsis and Thellungiella revealed that not all accessions showed an increase in the content of Put or Spd during cold acclimation. Also, the levels of free Put and Spd were not correlated with leaf freezing tolerance. In fact, the most freezing tolerant Arabidopsis accession in this study (Te-0) showed no increase in the pool size of either polyamine. In addition, the overall amounts of Put and Spd were very similar in all studied plants. Only the contents of free Spm showed higher levels in Thellungiella under nonacclimating conditions than in Arabidopsis. This was, however, strongly decreased during cold acclimation, leading to similar pool sizes between the species in the acclimated state. In Thellungiella we found a negative correlation between Spm contents and LT 50 ACC, indicating that low levels of Spm may be a requirement for efficient cold acclimation. A similar reduction of Spm levels was previously already observed in the Arabidopsis accession Col-0 [41] and in wheat [42] in response to cold exposure. However, the functional relevance of this reduction of free Spm levels is currently unknown. The natural variation in Spm content revealed in this study may offer an interesting possibility to elucidate the molecular basis and functional significance of this phenomenon.

Conclusion
While Thellungiella is generally assumed to be an extremophile with regard to its abiotic stress tolerance, the presented data indicate that this is not true with regard to its freezing tolerance. Some accessions, however, significantly expand the range present in Arabidopsis, stressing the utility of Thellungiella as an additional model species. The metabolite data indicate different metabolic adaptation strategies between these rather closely related species that need to be followed up with appropriate profiling technologies.