Introduction

Freezing tolerance is essential for winter cereals in a temperate climate if they are to survive the winter and produce sufficient yield. Maximum freezing tolerance can be achieved through a mechanism known as cold acclimation [1, 2]. During this process, low, non-freezing temperatures generate rapid signaling events leading to the induced expression of transcription factors and cold-regulated genes, resulting in enhanced freezing tolerance [3]. This complex response can be divided into two major systems depending on the involvement of abscisic acid (ABA) [4, 5]. One of the major pathways involved in this process is the ICE–CBF–COR system (ICE: inducer of CBF expression, CBF: C-repeat binding factor, COR: cold-regulated gene), which appears to be independent of ABA and exhibits a rapid and immediate response to cold [6]. Many members of this system were described in Arabidopsis thaliana L. [6, 7] and homologous genes were also isolated from barley (Hordeum vulgare L.) [8, 9]. In barley, more than 15 different genes of CBF transcription factors were discovered, localized in closely linked clusters on the long arm of chromosome 5H [9, 10]. Many of them show cold-induced expression within hours [11–13]. However, the individual functions of the CBFs are still not clear, since they show different levels of induction in different species [11, 14] and even in different genotypes [13, 15]. In cereals, CBF14 was found to have the greatest effect on freezing tolerance [16–18].

One positive regulator of CBF expression is the ICE transcription factor [6, 19]. The ICE1 gene is constitutively expressed and its activation by low temperature occurs at the protein level [6]. One of the first detectable gene expression induced by low temperature is the rapid, controlled induction of CBF genes which increases parallel with cold tolerance [15, 16], and the CBF response is followed by the expression of the COR14b gene [13, 20]. COR14b is encoded in the nucleus, but the protein itself is localized in the stroma of chloroplasts [21, 22]. Besides cold and possibly CBFs, the induced expression of COR14b is also regulated by other factors, the most important of which are red (but not far-red) or blue light and some still unknown chloroplast factor [22, 23]. In the Triticeae, COR expression levels are correlated with freezing tolerance: genotypes with greater freezing tolerance accumulate COR gene transcripts in higher amounts compared to genotypes with lower freezing tolerance [24–28]. In barley, a higher degree of freezing tolerance is associated with a higher threshold induction temperature for the accumulation of COR proteins [29]. Altogether, COR gene activation is thought to be a marker of cold acclimation in cereals.

The other pathway participating in the cold acclimation process is ABA-dependent and is induced by dehydration instead of low temperature. This response is slow and it includes bZIP transcription factors known as ABA-Responsive Element Binding Protein/Factor (AREB/ABF) [30]. The two pathways are not independent of each other, but are linked through complex crosstalk [31, 32]. Besides ABA, other phytohormones have also been described as important factors or indicators during the cold acclimation process, such as jasmonic acid (JA) [33] and salicylic acid (SA) [34].

Since cold acclimation is a very complex system involving cold response, light effect, and complex systemic mechanisms at the level of the whole organism, it is difficult to examine the cold-induced gene expression free of any additional interacting and cross-talking reactions. Therefore, the present research focus on a simplified system, in which cellular mechanisms related to cold tolerance can be investigated independently of the whole plant level. Moreover, in the case of cereals freezing tolerance does not mean the survival of each cell in organisms exposed to subzero temperature, but is correlated in practice to the survival of the pluripotent meristemic cells in the crown, since an intact meristem allows the plant to recover from freezing damage even despite the complete loss of the leaves and other superterrestrial parts of the shoot. Thus, widely preferred freezing test protocols for cereals at the whole plant level are based on the determination of meristemic functionality by measuring regeneration potential during the recovery period, after generating frost damage and trimming the upper parts of the plants [35–38].

Since adequate methods are not yet available in the case of the Triticeae tribe for the accurate isolation and long-term maintenance of meristemic cells in tissue culture, callus cultures can be used as an alternative to gain information about the stress responses of dedifferentiated plant cells. Apparently, in case of long-term cultivation of callus, care must be taken during the experimental design, because the increased degree of somaclonal variation might cause false results and misinterpretation of the research data. Several groups have already been reported the cold-responsiveness of cell cultures in a few plant species [39–42], but no data is available about the cold induction of CBF–COR system in this special plant material in cereals.

A great advantage of using callus cultures is that they can be maintained in both the light and the dark. Light-grown cells form green pigmented calli containing structured chloroplasts and their growth centers often exhibit direct organogenesis or indirect somatic embryogenesis [43], while dark-grown cultures are not pigmented, and lack active chloroplasts. Only photosynthetically inactive, amyloplast-like proplastids without a prolamellar body are found in dark-grown callus cultures [44]. The growth centers of dark-grown callus clumps remain dedifferentiated for a prolonged period during cultivation [45], and they are less efficient in forming plant organs than light-induced cultures [46]. Consequently, their utilization for the long-term maintenance of dedifferentiated cell populations is preferable to that of light-grown cultures. In addition, dark-grown cells lack most of the light-sensing and light-induced signaling systems, therefore they can be utilized to study basic cold-induced cellular mechanisms without any disturbing light-related effect.

In plants, meristemic cells are localized in specialized niches where signals from surrounding cells keep them in an undifferentiated state [47, 48]. In the case of monocot calli, the induction and maintenance of dedifferentiated status require a relatively high dosage of plant growth regulators (PGRs), such as phytohormones (usually auxins) or their artificial analogs, such as 3,6-dichloro-2-methoxybenzoic acid (Dicamba) [49, 50]. Since external treatment with hormones may cause alterations in the endogenous hormone metabolism [51], as well as in gene expression and other cellular processes, investigations on the endogenous hormone system are essential if the cellular and physiological responses of callus cultures are to be understood. Notwithstanding that several results are available where the effect of exogenously added hormones on the stress tolerance of the cultures were investigated, up to our knowledge there is only limited amount of data available about the endogenous hormone levels of dedifferentiated cells upon stress [42].

To sum up, dark-grown callus system is an ideal model for studying parts of the basic cellular cold response, independently of the systemic reaction and completely separated from the effect of light or any chloroplast-related mechanism, even after an extended period of cultivation. Our particular aim was to determine whether one of the most important cold-inducible regulatory system, the CBF–COR14b pathway is able to respond to cold in cell cultures maintained without light. Moreover, we collected new data about the cold-responsiveness of the endogenous hormone system of cultured cells.

Results

Callus cultures were induced from mature barley embryos, and grown in a controlled environment. Since cultivation occurred without light, the development of functional chloroplasts was blocked and all other light-sensing and signaling systems were impeded. This experimental setting provided a possible tool for monitoring the pure cold-induced response of dedifferentiated cells.

In the first experiment, the aim was to compare the pattern of cold-induced activation of the HvCBF and HvCOR genes in the crowns (containing the shoot apical meristem) of young seedlings and in callus cultures grown under normal tissue cultivation conditions (in the dark, treated with 1 mg/l Dicamba) using the quantitative real-time PCR (qRT-PCR) method. The cold response of light-grown cultures was not investigated in this study, since they have presumably different developmental program than dark-grown cultures causing significant difference between the two culture types, therefore their detailed comparison might be the subject of a further study. In a previous pre-screen, HvCBF1, HvCBF2, HvCBF3, HvCBF4, HvCBF5, HvCBF6, HvCBF7, HvCBF9, HvCBF11, HvCBF12, HvCBF13, HvCBF14, and HvCOR14b were tested using semi-quantitative PCR, and HvCBF1, HvCBF5, HvCBF6, HvCBF9, HvCBF11, HvCBF14, and HvCOR14b were cold-inducible in the crown tissue of seedlings, while HvCBF9, HvCBF14, and HvCOR14b showed a significant level of cold induction in both seedlings and calli (data not shown). These latter results were confirmed using the qRT-PCR method (Fig. 1). Although the seedlings were grown in normal light conditions, while the calli were maintained in the dark, a comparison of the two systems indicated that the kinetics of cold induction was similar. In the crowns the two HvCBF genes showed a rapid, transient increase after the beginning of the cold treatment, reaching a maximum level at 6–12 h, followed by a falloff for around 48 h. In the calli, the expression level and kinetics of HvCBF9 were similar to those measured in the crowns. Cold response of the HvCBF14 gene was delayed in the calli in comparison with crowns of seedlings, however, the maximum expression level was reached in 12 h, at the same time as determined in the crowns of seedlings. The time course of the HvCOR14b cold response was almost identical in the two systems, but the level was several orders of magnitude higher in the crowns of the seedlings, indicating the presence of other (e.g., systemic or light-induced) factors in the cold acclimation process at the HvCOR14b gene activation level.

Fig. 1
figure 1

Comparison of cold-induced CBF and COR14b expression in seedlings and calli. Relative expression (fold change ± SE) of the HvCBF9 (A, B), HvCBF14 (C, D), and HvCOR14b (E, F) genes in crowns of seedlings (A, C, E) and calli (B, D, F) at different time points during cold-hardening (+4 °C) assayed by qRT-PCR. The calli were grown under control tissue cultivation conditions (dark, 1 mg/l Dicamba), whereas the seedlings were grown in the light, without exogenous hormone treatment. Samples were taken after 2, 4, 6, 12, 24, and 48 h of cold-hardening. Three biological repeats are given for each data point. The results show the expression level in cold-treated samples relative to the unhardened control (grown at +24 °C). The data were normalized to the level of the housekeeping gene HvACTIN (endogene control) for each sample. Bars with the same letter are not significantly different at p = 0.05 according to one-way ANOVA followed by Tukey’s HSD post hoc test analysis

Full size image

The comparison of young seedlings and calli showed similar kinetics for the cold induction of the CBF–COR system, so the further characterization of the cold response in dedifferentiated cell cultures was performed. In these experiments it had to be taken into consideration that the induction and maintenance of the calli were performed using a synthetic auxin-type growth regulator (Dicamba) in the medium, which means that the relative high dosage of exogenous hormone analog might exert an effect on the cold response. Therefore, after the induction and selection of embryogen callus cultures some of the cultures were cultivated on Dicamba-free medium for 9 weeks (Dic(−) calli), while the remaining cultures were kept on medium with Dicamba to investigate the effect of this exogenous auxin analog on the cold response of dedifferentiated cell cultures. The Dicamba content in the Dic(−) cultures (mean ± SE) was 16.52 ± 2.59 pmol/gFW compared to 10757.84 ± 1449.31 pmol/gFW in the Dic(+) cultures proving that the Dicamba content had been successfully reduced. During this process neither macroscopic nor microscopic alterations were detected. Moreover, the phenotype (embryogenic form, friable structure) and growth rate (175 ± 14 % of initial weight at +24 °C, and 148 ± 18 % of initial weight at +4 °C after seven days of cultivation) of the Dic(−) cultures were similar to those of cultures maintained on medium containing Dicamba.

To investigate the effect of Dicamba on cold-regulated gene expression, the cold induction of the HvCBF9, HvCBF14, and HvCOR14b genes was compared in Dic(+) and Dic(−) callus cultures. First, the effect of Dicamba on the background levels of HvCBF9, HvCBF14, and HvCOR14b transcripts (under control conditions, without cold treatment) was determined. There was no difference between the Dic(−) and Dic(+) cultures in the basal expression level of HvCOR14b, while HvCBF14 and HvCBF9 showed slightly higher (1.7–2.4 fold) expression in Dic(+) cultures relative to Dicamba-free control samples. To investigate the effect of Dicamba on the cold response of these genes, cold treatment was performed and the induction of HvCBF9, HvCBF14, and HvCOR14b was analysed in Dic(+) and Dic(−) calli. Samples were taken at maximum induction (12 and 24 h after the beginning of cold treatment) and gene expression levels were determined. In addition, gene expression levels were measured at later stages of cold treatment (after 4 and 7 days) to prove that the HvCBF expression system was down-regulated after 24 h in dedifferentiated cell cultures, as described by Stockinger et al. [13] in barley plants. In the case of HvCBF9 there was no significant difference between the Dic(+) and Dic(−) cultures (Fig. 2A). In contrast, HvCBF14 (Fig. 2B) and HvCOR14b (Fig. 2C) showed much lower expression without Dicamba, albeit the cold induction was clear and the characteristics of the induction were similar in both treatments. The HvCBFs showed a transient response with a clear and complete falloff after 24 h.

Fig. 2
figure 2

The effect of Dicamba on CBF and COR14b induction in calli. Relative expression (fold change ± SE) of the HvCBF9 (A), HvCBF14 (B), and HvCOR14b (C) genes in Dicamba-treated, Dic(+), and Dicamba-free, Dic(−), calli at different time points during short- and long-term cold-hardening (+4 °C) assayed by qRT-PCR. Samples were taken after 12 and 24 h (short-term hardening) and 96 and 168 h (4 and 7 days) (long-term hardening) of cold treatment. Three biological repeats are given for each data point. The results show the expression level in cold-treated samples relative to the unhardened control (grown at +24 °C). The data were normalized to the level of the housekeeping gene HvACTIN (endogene control) for each sample. Bars with the same letter are not significantly different at p = 0.05 according to one-way ANOVA followed by Tukey’s HSD post hoc test analysis

Full size image

The cold inducibility of the HvCBF9, HvCBF14, and HvCOR14b genes suggested the existence of a “plant-type” cellular cold response in dedifferentiated cell cultures as well. For the further analysis of this process, the cold acclimation ability of the cultures was determined. In this experiment calli were cold hardened at +4°C without light for 4 or 7 days and a subsequent freezing test was performed to monitor the potential of cell cultures to acclimate to lower temperatures. The survival rate was determined using both TTC assay (cell viability test) and a growth test (determining the growth rate) to verify the results by measuring two different aspects of survival. It was clear that four-day hardening did not result in a significant increase in survival and there was no significant difference between the Dic(+) and Dic(−) calli, as all the samples perished (Fig. 3). By contrast, 7-day hardening led to an increased survival rate measured by both methods, demonstrating the existence of cold acclimation ability in dedifferentiated cells.

Fig. 3
figure 3

Freezing tolerance of callus cultures. Relative cell viability (% of unfrozen control) (mean ± SE) according to TTC assay (A) and the relative growth rate (% of unfrozen control) (mean ± SE) in the growth test (B) after induced freezing at −4 °C. Control calli were grown at +24 °C; long-term hardening lasting 4 or 7 days was carried out at +4 °C. Three biological repeats are given for each data point. Bars with the same letter are not significantly different at p = 0.05 according to one-way ANOVA followed by Tukey’s HSD post hoc test analysis

Full size image

Interestingly, if the survival of Dic(−) cultures was compared to that of Dic(+) calli, after 7 days of hardening the Dic(−) calli showed a lower survival rate than cultures grown on hormone-containing medium, suggesting that Dicamba-treated cultures had an enhanced stress tolerance.

Due to the high concentration of exogenous auxin analog used for callus induction and maintenance, it was interesting to investigate how endogenous levels of stress-related hormones were affected by high Dicamba content. In this experiment, stress hormones JA, SA, ABA and its catabolites were measured in Dic(+) and Dic(−) cultures. The levels of indole-3-acetic acid (IAA) and its two conjugates were also determined to see how the endogenous auxin level is affected by an exogenous analog. Table 1 presents the hormone concentrations measured in Dic(+) and Dic(−) calli grown under control conditions. The level of IAA did not differ in the Dic(+) and Dic(−) cultures, while a lower concentration of the two inactive conjugates, IAA-aspartate (IAA-Asp), and IAA-glucose ester (IAA-GE), was measured in Dic(+) calli, however, this decrease was not significant. The content of the stress hormone SA was slightly elevated, while JA and ABA showed almost 3- and 4.5-fold higher amounts, respectively, in Dic(+) calli. The amount of the ABA catabolite dihydrophaseic acid (DPA) was not affected by Dicamba, whereas the concentrations of neophaseic acid (Neo-PA) and phaseic acid (PA) were about threefold higher in Dic(+) calli. These data suggest that Dicamba itself exerts a stress-like impact on callus. In order to clarify the effect of abiotic stress on endogenous hormone levels, concentrations were measured after 4 and 7 days of cold treatment in both types of calli. The hormonal response to cold was calculated as a percentage of the basal levels recorded in the previous experiment in the Dic(+) and Dic(−) cultures. Since 4-day-long cold-hardening did not induce phenotypic changes in frost tolerance, and 4-day data were confirmed by 7-day data in almost all cases, hormone analysis results at this time point are presented in Supplemental figures. The IAA conjugates IAA-Asp and IAA-GE showed no response to cold, irrespective of the length of treatment, whereas IAA slightly increased during cold treatment in Dic(+) cultures (Fig. 4, Supplementary Fig. 1). Dicamba depletion caused a negative response to cold in point of the amount of IAA derivatives. The ABA content did not change significantly after cold treatment in either Dic(+) or Dic(−) cultures, while one of its catabolite, DPA, showed a moderate cold-induced increase in both cultures. This response was independent of the Dicamba content and the length of the cold treatment (Fig. 5a, Supplementary Fig. 2A). The most interesting observation was the dramatic increase in the biologically inactive ABA catabolites PA and Neo-PA (Fig. 5b, Supplementary Fig. 2B) in response to cold treatment. In Dic(+) cultures, the cold-induced increase in these compounds exceeded 1,500 % (or even 2,000 % in the case of PA) after 7-day cold treatment (Fig. 5b). The amplitude of this response was much lower in Dicamba-depleted calli, especially in the case of PA. Among the other stress-related hormones, SA gave no significant response (Fig. 6, Supplementary Fig. 3), while the amount of JA increased significantly after 7 days of cold treatment in Dic(+) cultures (Fig. 6).

Table 1 Effect of Dicamba on endogenous hormone levels in control cultures
Full size table
Fig. 4
figure 4

Effect of Dicamba on hormone levels of endogenous auxins after 7 days of cold treatment. Relative hormone levels of endogenous auxins: indole-3-acetic acid (IAA), IAA-aspartate (IAA-Asp), and IAA-glucose ester (IAA-GE) in Dicamba-treated, Dic(+), and Dicamba-free, Dic(−), calli (mean ± SE). The results show the hormone levels in cold-treated samples after 7 days of hardening relative to the unhardened control (grown at +24 °C) (%). Three biological repeats are given for each data point. Statistical analysis was performed using unpaired (Student’s) and Welch’s t test (*0.01 < p < 0.05; **0.001 < p < 0.01)

Full size image
Fig. 5
figure 5

Effect of Dicamba on hormone levels of ABA and its catabolites after 7 days of cold treatment. Relative hormone levels of endogenous abscisic acid (ABA) and dihydrophaseic acid (DPA) (A), and phaseic acid (PA) and neophaseic acid (Neo-PA) (B) in Dicamba-treated, Dic(+), and Dicamba-free, Dic(−), calli (mean ± SE). The results show the hormone levels in cold-treated samples after 7 days of hardening relative to the unhardened control (grown at +24 °C) (%). Three biological repeats are given for each data point. Statistical analysis was performed using unpaired (Student’s) and Welch’s t test (*0.01 < p < 0.05; **0.001 < p < 0.01)

Full size image
Fig. 6
figure 6

Effect of Dicamba on hormone levels of jasmonic acid and salicylic acid after 7 days of cold treatment. Relative hormone levels of endogenous jasmonic acid (JA) and salicylic acid (SA) in Dicamba-treated, Dic(+), and Dicamba-free, Dic(−), calli (mean ± SE). The results show the hormone levels in cold-treated samples after 7 days of hardening relative to the unhardened control (grown at +24 °C) (%). Three biological repeats are given for each data point. Statistical analysis was performed using unpaired (Student’s) and Welch’s t test (*0.01 < p < 0.05; **0.001 < p < 0.01)

Full size image

Discussion

The study reported here attempted to investigate the cold-induced expression of the CBF–COR system in dark-grown cell culture, and to investigate the effects of the main exogenous hormone component in callus cultivation system on the cold response and the endogenous hormonal pools during low-temperature stress.

Growth centers of embryogenic callus cultures have the potential to model the behavior of meristemic cells during cold stress. Moreover, cell culture is an ideal system for the performance of various chemical treatments, thus allowing the basic mechanisms involved in the cold stress response of cereals to be studied via a pharmacological approach, using inhibitors or inducers of certain cellular processes. However, if dedifferentiated embryogenic cell cultures are to be used as a model of meristemic cells and the results are to be clearly interpreted, a deeper knowledge will be required on the functioning and nature of this special cell type. For this purpose, the aim was to provide new information about the behavior of monocot callus cultures under abiotic stress conditions, also taking into consideration the possible effects of the exogenous hormone treatment required for tissue cultivation.

Cereal cell suspension cultures hardened in the dark are known to have cold acclimation ability [41, 52]. In the present study, these findings were confirmed in barley callus cultures and linked to the cold-induced activation of the CBF–COR regulatory system. The activation of the CBF–COR system in seedlings of the genotype used in this study was previously described [13], but the cold response of meristemoid cells, which lack light-sensing machinery, remained unknown. Interestingly, besides HvCBF9 and HvCBF14, Stockinger et al. [13] found several other HvCBF genes to be cold-inducible, but the 2-week difference in the developmental phase could explain this discrepancy.

One of the most important findings in this study was that cold-inducible HvCBF genes showed similar responsiveness to low temperature in dark-grown callus cultures as in the crown of light-grown seedlings (Fig. 1). The HvCOR14b gene also responded to cold in callus cultures, but to a much lower extent, which is in agreement with previous reports on the light-dependent activity of this gene in plants and on the localization of its product in the chloroplasts [22, 23]. Obviously, without light and functional chloroplasts the cold induction of HvCOR14b was much weaker, but it was clearly detectable, suggesting the presence of a light-independent, basic activation mechanism in dedifferentiated cells.

Another important aspect of callus system characterization that was unclear in cereal tissue cultures was elucidation of the changes observed in the endogenous hormonal pools under stress conditions, especially as an exogenous hormone analog is used to induce and maintain the dedifferentiated state in callus cells. The effect of exogenous auxin on endogenous hormone levels has already been determined in embryogenic cultures of carrot [53–55], but the response to abiotic stress was not investigated in these studies. In Arabidopsis cell culture, stress response of endogenous ABA has already been determined by Sasaki et al. [42], but other hormones were not investigated. In our experiments hormone concentrations were observed to be about one order of magnitude lower than in common wheat (Triticum aestivum L.) leaves [56]. However, dedifferentiated barley cells are able to produce comparable amounts of phytohormones whether maintained with or without Dicamba, and the hormonal system is capable of responding to cold stress. In general, the response of the endogenous hormones was stronger in the presence of exogenous auxin analog, especially in the case of ABA catabolites. Interestingly, the endogenous ABA concentration did not change significantly following cold, which is in agreement with the data of Sasaki et al. [42] in Arabidopsis suspension culture. On the contrary, the level of ABA catabolites showed dramatic cold-induced increase in Dic(+) cells. The elevated amounts of PA and Neo-PA (Fig. 5b) suggested a high level of previous ABA induction [57], but the Dicamba dependence appeared to be a general enhancement of the stress response rather than a specific effect on ABA metabolism. This was clearly demonstrated by comparing the hormone levels in non-stressed calli, where the stress hormones and their derivatives showed significantly higher concentrations in Dic(+) cultures. This hypothesis was also supported by the freezing test results, in which cold-treated calli grown on Dicamba-containing medium showed a higher rate of hardening than Dicamba-depleted cultures. This may be a consequence of the stronger cold induction of the HvCBF14 (Fig. 2B) and HvCOR14b (Fig. 2C) genes when exogenous hormone was applied. To sum up, the effect of Dicamba appears to be more in the nature of co-induction, since the basal level and the characteristics of cold-induced gene expression were not affected by this exogenous hormone analog, though the strength of the response was generally higher in Dic(+) cultures under stress conditions.

Summarizing the findings it can be concluded that dedifferentiated embryogenic cells maintained on solid medium have the potential to respond to cold at the endogenous hormonal level and they have cold-responsive CBF–COR14b expression system even without light. Although, exogenously applied auxin analog might cause altered cellular metabolism, the endogenous hormonal system and the CBF–COR14b pathway was not specifically affected by Dicamba. Our data along with previous findings show that this system might be a suitable model for studying certain basic cellular cold-induced mechanisms. Taking into account the similarities and differences between embryogenic callus cells and meristemic cells, our concluding hypothesis is that the greater amplitude of the cold response determined in crowns of seedlings is the consequence of the cellular, physiological, and biochemical environment surrounding the shoot apical meristem and providing sufficient amounts of hormones, metabolites, and other factors necessary to reach the maximum level of cold-hardening.

Materials and Methods

Callus Induction, Tissue Cultivation and Growth Conditions

Embryos were dissected under sterile conditions from surface sterilized, mature seeds of winter barley (H. vulgare L. cv. Nure), and placed on callus induction medium I. The basic medium was prepared using modified “Murashige and Skoog medium including vitamins” (Duchefa) supplemented with 1 g/l “N-Z-Amine A” casein hydrolysate (Sigma), 40 g/l maltose and 8 g/l agar, and adjusted to pH 5.8 before autoclaving. Callus induction medium I also contained 2.5 mg/l of the synthetic auxin analog 3,6-dichloro-2-methoxybenzoic acid (Dicamba) and 0.5 mg/l of the synthetic cytokinin analog 6-benzylaminopurine (BAP). The callus cultures were maintained by routine subculture every 3 weeks. Because of higher sugar concentration (standard media contain usually 30 g/l sucrose) and the fact, that maltose shows a slower rate of extracellular hydrolysis, and it is taken up more slowly than sucrose [58, 59], sugar starving due to the depletion of carbon source, which was reported in batch cultures [60], cannot occur. After subculturing five times, actively growing, friable calli were selected and placed on medium II, which contained 1 mg/l Dicamba, but no BAP. After subculturing five more times on medium II, callus cultures were divided into two groups: one group was maintained on medium II (Dic(+) cultures), while the other group (Dic(−) calli) was maintained without Dicamba treatment. Healthy, fast growing cultures were used for carrying out experiments after the 13th subcultivation. All the cultures were grown at +24 °C, 50 % RH, in continuous dark (control conditions for tissue cultivation).

To create plant material, mature seeds of winter barley (H. vulgare L. cv. Nure) were germinated on wet filter paper in Petri dishes at +24 °C in the light under sterile conditions. The germination rate was ~94 %. After 3 days the seedlings were planted in wooden boxes (42 × 31 × 13 cm) in a ~9 cm deep 4:1 mixture of garden soil and sand. The plantlets were cultivated in climate chambers (PGV-36, Conviron, Canada) under controlled control conditions (irradiation: 250 μmol/m2/s, day/night photoperiod: 16/8 h, RH: 75 %, day/night temperature: +20/+15 °C). One day after planting, cold-hardening was begun at +4 °C. Control and cold-treated samples were collected 0, 2, 4, 6, 12, 24, and 48 h after the beginning of cold treatment. Three replicates, each containing three crown samples of individual plants, were collected at each time point. The plant material was frozen in liquid nitrogen immediately after sampling and kept at −80 °C until further processing.

Experimental Design, Cold Treatment and Freezing Test

Clear evidences of somaclonal variability were demonstrated in barley callus cultures after long-term cultivation [61]. To prevent the effects of this phenomenon, Dic(+) and Dic(−) samples were prepared by mixing, gently homogenizing, and reapportioning of all of the colonies from Dic(+) or Dic(−) treatment group. Thus, one single sample contained callus clumps from hundreds of colonies from the given treatment group. Then, both the Dic(+) and Dic(−) groups of calli were divided into two subgroups: cold-treated and control cultures. Cold-treated samples of each group were placed at +4 °C, 50 % RH in continuous dark, and three biological replicates of each subgroup were collected at the beginning of the experiment and after 2, 4, 6, 12, 24, and 48 h or after 4 and 7 days of cold treatment (depending on the experiment). Samples for gene expression analysis and the quantification of phytohormones and the hormone analog Dicamba were immediately frozen in liquid nitrogen and kept at −80 °C until further processing. The other set of samples were treated according to the protocol for testing the freezing tolerance of callus cultures, described below.

Freezing Test

For testing the freezing tolerance of tissue cultures, gently homogenized clumps of calli were wrapped in two layers of wet, sterile filter paper to initiate ice nucleation, and placed in a plastic bag. The samples were then placed in a liquid freezer (Grant GP200-R4 thermostat) using continuously circulated ethylene glycol as the thermal transfer agent, and the temperature was gradually decreased from +4 to −4 °C (0.266 °C/min). The freezing protocol was selected based on the results of preliminary experiments (LT50 measurements) on control callus cultures (data not shown). After 1 h of freezing at −4 °C the samples were removed from the thermostat and one portion of the cultures was placed on medium II for a long-term growth test, while the other portion was used for a viability test after complete self-thawing at +24 °C. Two independent freezing tests were performed; the results are summarized in a single data set.

Long-Term Growth Test

Three biological replicates of frozen and self-thawed cultures from each treatment group were placed on medium II under sterile conditions after weight measurement. After a 3-week regeneration period at +24 °C, 50 % RH, in continuous dark the weight measurements were repeated. Re-establishment of the growth rate was evaluated as the weight increase (expressed as a percentage of the initial weight).

Viability Test (TTC Assay)

The protocol of Towill and Mazur [62] was adapted for barley callus cultures as follows. Approximately 200 mg gently homogenized tissue was washed twice with 50 mM phosphate buffer (pH = 7.5), after which 2 ml of 0.08 % sterile 2,3,5-triphenyltetrazolium chloride (TTC) solution (prepared in 50 mM phosphate buffer, pH = 7.5) was added to the cultures in microcentrifuge tubes. The samples were incubated on an orbital shaker at room temperature for 12 h in the dark. Then, the TTC solution was completely removed, the cultures were washed again with 50 mM phosphate buffer, 1.5 ml ethanol (96 %) was added to the cells and the samples were incubated again on an orbital shaker at room temperature for 12 h in the dark to solubilize formazan. After centrifugation at 9,500×g for 5 min, the supernatant was recovered and the absorbance was measured at 485 nm (Varian Cary100 Scan Spectrophotometer). Cell viability was calculated as the absorbance at 485 nm for 1,000 mg callus culture.

Gene Expression Analysis

Total RNA was extracted using TRIzol reagent (Invitrogen) followed by further purification using an RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. Oligo (dT)15 primer, dNTP Mix, and M-MLV Reverse Transcriptase (Promega) were used for cDNA synthesis from 1 μg of total RNA. A single reverse transcription reaction was performed for each RNA sample.

qRT-PCRs were run on a 7500 Fast Real-Time PCR System (Applied Biosystems) using Fast SYBR Green Master Mix (Applied Biosystems) according to the manufacturer’s instructions. Cycling conditions were 20 s at +95 °C, 40 cycles of 3 s at +95 °C, and 30 s at +60 °C, followed by a melting curve program (60–95 °C). The following primer sets were used for the analysis: HvACTIN fw 5′-GGATCTCGAAGGGTGAGTACGA-3′, rev 5′-GACAACTCGCAACTTAGAAGCACTT-3′, HvCBF9 fw 5′-CATGTAGATAGTTGCGTTCTTCCAG-3′, rev 5′-CACAATCGAGTTCAGTACCAATTG-3′, HvCBF14 fw 5′-GGGAGGCTGCATTTGTACTTG-3′, rev 5′-CCATGGAGTGCTTGAGGTCC-3′, and HvCOR14b [63] fw 5′-AGACCCAGATCGATGGCTTCT-3′, rev 5′-GCACGGCCTGGGAAGAG-3′. The efficiency of each primer set was tested by serial dilution curve analysis before the experiment (data not shown). Relative quantification was performed for each primer set and cDNA template combination using four technical replicates, including a non-template control. In order to validate the results of qPCR analysis, two independent replications were made, both on randomly chosen plates from each experiment and on the whole experiment, using new samples. The expression levels of genes of interest were calculated relative to the ACTIN expression level using the ΔΔCt method reported by Livak and Schmittgen [64]. Constant expression level of HvACTIN gene was checked in different samples at different time points. The fold induction values in cold-treated samples were calculated relative to the control samples (grown at +24 °C) at each time point.

Quantification of Phytohormones and Dicamba

The concentrations of endogenous IAA and IAA derivatives (IAA-Asp, IAA-GE), ABA and ABA catabolites (PA, Neo-PA, DPA), SA, JA and the exogenous auxin analog (Dicamba) were determined using an LC-MS–MS system. Extraction and purification of the samples were carried out using reverse phase and ion exchange chromatography as described previously [65]. Isotope dilution technique was used to account for losses during purification. For this purpose appropriate stable isotope labeled internal standards were added. Phytohormones, including Dicamba, were concentrated in methanolic eluate. After evaporation, the fraction was dissolved into 15 % acetonitrile in water. Quantification was performed after injection of sample aliquots into LC-MS/MS system consisting of HPLC (Ultimate 3000, Dionex, Sunnyvale, CA, USA) coupled to hybrid triple quadrupole/linear ion trap mass spectrometer (3200 Q TRAP, Applied Biosystems, Foster City, CA, USA) set in selected reaction monitoring mode.

Statistical Analysis

In the case of data set presented in Table 1, Figs. 4, 5, 6 and Supplemental Figs. 1, 2, 3 performing unpaired (Student’s) t test analysis was appropriate for comparing hormone content of Dic(–) and Dic(+) control cultures. If the assumption test suggested that the difference between two standard deviations was significant, alternative t test (Welch’s test) was performed.

In order to compare more than two data sets from different time points and treatment groups, instead of simple t test the statistical analysis involved one-way ANOVA and Tukey’s HSD post hoc test analysis (p < 0.05) using SPSS 16.0.1 and InStat (GraphPad) software. The conditions were tested for each data set before the analysis: normal distributions using the Kolmogorov–Smirnov (KS) test and deviation with the Levene test. Bars on the graphs with the same letter are not significantly different at p = 0.05.