Present address: Institute of Plant Sciences, Swiss Federal Institute of Technology (ETH), 8092 Zürich, Switzerland.
These authors contributed equally to this work.
Calcium ions represent both an integrative signal and an important convergence point of many disparate signaling pathways. Calcium-binding proteins, like calcineurin B-like (CBL) proteins, have been implicated as important relays in calcium signaling. Here, we report the in vivo study of CBL1 function in Arabidopsis. Analyses of loss-of-function as well as CBL1-overexpressing lines indicate a crucial function of this calcium sensor protein in abiotic stress responses. Mutation of CBL1 impairs plant responses to drought and salt stresses and affects gene expression of cold-regulated genes, but does not affect abscisic acid (ABA) responsiveness. Conversely, overexpression of CBL1 reduces transpirational water loss and induces the expression of early stress-responsive transcription factors and stress adaptation genes in non-stressed plants. Together, our data indicate that the calcium sensor protein CBL1 may constitute an integrative node in plant responses to abiotic stimuli and contributes to the regulation of early stress-related transcription factors of the C-Repeat-Binding Factor/dehydration-responsive element (CBF/DREB) type.
Plants, as sessile organisms, have evolved an enormous capacity to realize their genetically predetermined developmental program despite ever-changing environmental conditions. A multitude of exogenous stimuli like light, temperature, nutrient, and water availability needs to be perceived and processed simultaneously to achieve an integrated response, ensuring optimal adaptation to the environment (McCarty and Chory, 2000). Therefore, the signaling and adaptation responses of the plant concurrently need to warrant both the required specificity to a certain response and the necessary coordination and interconnection of stimulus-induced reactions (Gilroy and Trewavas, 2001). During the past years, it has become increasingly clear that plant signaling systems not only consist of linear pathways, but also form a complex signaling network with extensive overlaps and nodes interconnecting its branches (Gilroy and Trewavas, 2001; Knight and Knight, 2001). So far, the identification and characterization of such interconnecting signaling nodes, which are likely to hold key positions in coordinating plant stress responses and adaptation, have remained in its infancy.
Elevation of the cytosolic calcium concentration is a primary event in the response to many different stresses like cold, drought, high salinity, and osmolarity (Knight and Knight, 2001; Sanders et al., 1999; Trewavas and Malho, 1998). Although the specific signature of these calcium transients can encode information and specificity on its own, an additional level of regulation in calcium signaling is achieved via the action of calcium-binding proteins (Allen and Schroeder, 2001; McAinsh and Hetherington, 1998; Sanders et al., 2002). Such proteins sense changes in the local calcium concentration and relay this information to their target proteins. If these calcium sensors do not have an enzymatic activity on their own, like, for example, calmodulin and frequenin, binding of calcium ions usually leads to an increased affinity for and subsequent activation or de-activation of their target proteins (Snedden and Fromm, 1998). Typical targets representing primary downstream transducers of calcium signals are phosphorylation cascades consisting of tightly regulated protein kinases and phosphatases (Hunter, 1995; Soderling, 1999).
Alternatively, the calcium signal can be sensed and transmitted by a single protein. This is, for example, the case with plant calcium-dependent protein kinases (CDPKs), which harbor a calcium-binding domain as well as a catalytic kinase domain (Sanders et al., 2002). Functional analyses of plant CDPKs have recently provided strong evidence for a crucial function of these kinases in processes like hormone and stress signaling as well as plant pathogen response (Harmon et al., 2000; Romeis et al., 2001).
Recent studies have identified a novel family of calcineurin B-like (CBL) calcium sensor proteins from Arabidopsis (Kudla et al., 1999). These proteins harbor EF-hand motifs as structural basis for calcium binding and interact specifically with a group of serine–threonine protein kinases designated as CBL-interacting protein kinases (CIPKs; Shi et al., 1999). Subsequent studies identified the NAF-domain, a conserved 24 amino acid motif within the C-terminal region of these kinases, as necessary and sufficient to mediate CBL–CIPK interaction, thus defining these kinases as targets of calcium signals transduced by CBL proteins (Albrecht et al., 2001). Yeast two-hybrid screenings as well as bioinformatics analyses of the complete Arabidopsis genome sequence unraveled a complex signaling network comprising 10 CBL-type calcium sensors and 25 CIPK-type target kinases (Luan et al., 2002). Preferential complex formation of individual CBLs with defined subsets of CIPKs has been found to contribute to generating specificity in this signaling network (Albrecht et al., 2001). Positional cloning of the Arabidopsis Salt Overly Sensitive SOS3 and SOS2 loci revealed that these genes encode proteins belonging to the CBL (SOS3/CBL4) and CIPK (SOS2/CIPK24) signaling system (Liu and Zhu, 1998; Liu et al., 2000). Interestingly, these mutations render plants hypersensitive, specifically to sodium chloride. Moreover, biochemical and genetic studies established a direct interaction of the two proteins and their function in one and the same signaling pathway, suggesting that components of a plant salt stress-specific signaling cascade were identified (Halfter et al., 2000).
In contrast, the in vivo function of most of the other CBLs and CIPKs has remained elusive (Luan et al., 2002). Nevertheless, expression studies of CBL1, the founding member of the CBL gene family, have provided an indication for potential functions in planta: expression of CBL1 was found to be strongly and transiently induced by drought and cold stresses, suggesting a role of this calcium sensor in the respective signaling cascades (Kudla et al., 1999).
In this study, we have investigated the function of CBL1 by analyzing a T-DNA-induced knock-out mutant as well as plants overexpressing this calcium sensor protein. Loss of CBL1 impairs plant responses to drought and salt stresses, and affects gene expression of cold-regulated genes, but does not affect abscisic acid (ABA) sensitivity. Conversely, constitutive overexpression of CBL1 induces the expression of known stress-regulated genes and enhances stress tolerance. Taken together, our analyses suggest that the calcium sensor CBL1 may constitute an integrative node in plant abiotic stress signaling and contributes to the regulation of early stress-related transcription factors of the CBF/DREB type.
Isolation and characterization of a T-DNA-induced knock-out mutant of CBL1
Previous expression studies have suggested a role of CBL1 in plant responses to abiotic stresses (Kudla et al., 1999). Therefore, we sought to establish the in vivo function of CBL1 in these processes taking a reverse genetic approach.
To this end, we performed a PCR-based screen of a T-DNA insertion collection representing 36 500 T-DNA-transformed Arabidopsis lines, which led to the isolation of a potential cbl1 T-DNA insertion mutant. Sequence analysis of the PCR product amplified by the combination of T-DNA- and CBL1-specific primers located the T-DNA insertion within the first intron of the CBL1 gene, 265 nt upstream of the ATG initiator codon (Figure 1a). Southern blot and PCR data as well as co-segregation analyses of the kanamycin marker gene with the T-DNA presence indicated a single insertion event in this mutant line (Experimental procedures; data not shown). To analyze expression of CBL1, we subsequently applied RT-PCR with CBL1-specific primers. Because of the existence of gene families with highly similar members in the Arabidopsis genome (e.g. CBL9 shares 85% nucleotide sequence identity with CBL1), gene-specific quantitative RT-PCR assays according to established procedures (Halliday et al., 1999; Ni et al., 1998) were favored over Northern blot analyses and used in all subsequent gene expression studies. RT-PCR using homozygous progeny plants revealed that the T-DNA insertion reduces the CBL1 mRNA to undetectable levels, suggesting that this line harbors a cbl1 loss-of-function allele (Figure 1b). Despite the virtually complete abolishment of CBL1 expression, cbl1 plants did not exhibit any obvious phenotypes under normal long- or short-day growth conditions.
Loss of CBL1 function renders plants drought sensitive
To address the potential involvement of CBL1 in plant drought stress responses, mutant and wild-type plants were grown on soil in continuous light and were consecutively withheld from watering for 10 days (Figure 2a). While wild-type plants survived this treatment, all cbl1 mutant plants showed severe dehydration symptoms under these drought stress conditions. Additionally, the cbl1 mutants could not recover when water was supplied after a 13-day drought period (Figure 2a). We subsequently performed comparative water loss analyses to evaluate drought resistance in a more quantitative manner (Figure 2b). For this purpose, detached leaves of wild-type and mutant plants were exposed to normal humidity (40% humidity in a growth chamber) and the FW was determined at the indicated time points (Figure 2b). These analyses clearly established that the cbl1 mutant exhibits an accelerated water loss, which is likely the cause of the observed drought sensitivity.
To gain further insights into the molecular basis of the observed phenotypes, we next investigated the expression of well-characterized stress-responsive marker genes by quantitative RT-PCR in wild-type and mutant plants after drought treatment (Figure 2c). Besides CBL1, the expression patterns of CBF1 (DREB1b), CBF2 (DREB1c), and RD29A (LTI78, COR78) were analyzed as known stress-regulated genes (Gilmour et al., 1998). In addition, the transcription factor CBF4 (Haake et al., 2002) was chosen as drought-specific marker gene, and the pathogen and wound-induced PR1 gene was included in these analyses as a specificity control, whereas the constitutively expressed actin2 gene served as an internal standard. The transient induction of CBF4 and the rapid (1–3 h) and transient induction of CBL1 mRNA in wild-type plants indicated the effectiveness of this treatment. As reported previously (Gilmour et al., 1998), the steady state mRNA levels of the transcription factors CBF1/DREB1b and CBF2/DREB1c displayed no significant changes in response to drought stress. Likewise, expression of the wound-induced PR1 gene was not affected. Although the expression of CBF4 was induced by the drought treatment, no alteration of the expression profile was detectable between wild-type and CBL1 mutant plants. In contrast, the kinetics of RD29A mRNA induction was drastically altered in the cbl1 mutant. While in wild-type plants, RD29A transcripts increased progressively for 12 h, this was not the case in cbl1 mutant plants. Here, RD29A transcript levels peaked within the first hour after stress exposure and continuously declined thereafter. Taken together, the observed drought sensitivity of the mutant plants coincided with an altered induction kinetics of RD29A, a gene implicated in plant drought stress response (Xiong et al., 2002a).
CBL1 also functions in plant salt stress response
We next addressed whether CBL1 also mediates salt stress responses in Arabidopsis. To this end, the survival rate of seedlings grown on Murashige & Skoog (MS) media was analyzed after transfer onto plates containing 100 mm NaCl. Plants developing true leaves under these conditions were designated as survivors (see Figure 3a for a representative data set). Quantitative analysis of seedling survivors revealed a survival rate of 85% for wild-type seedlings, whereas only 25% of the cbl1 mutants survived this treatment (Figure 3b). We also investigated whether CBL1 function is required for salt tolerance at the adult plant level. For this purpose, 8-week-old soil-grown plants were treated once with a 100-mm NaCl solution and, as a control, with 200 mm mannitol. As depicted in Figure 3(c), 1 week after the stress treatment, the formation of extensive lesions was seen in mutant plants, while wild-type plants developed lesions to a lesser extent. No lesions were observed in the mutant and wild-type plants after mannitol treatment, indicating the salt specificity of the detected phenotype (data not shown). In addition, we performed root growth measurements at different salt concentrations. These analyses revealed no significant differences between the cbl1 mutant and wild-type plants (data not shown). This finding most likely reflects the predominant role of SOS3/CBL4 in mediating root-specific salt stress responses, and is in accordance with our finding that CBL1 expression in roots is confined to the root tips (D. Blazevic and J. Kudla, unpublished).
The investigation of the CBL1 expression level in 2-week-old Arabidopsis seedlings treated with 100 mm NaCl showed a rapid induction of CBL1 mRNA amounts within 1 h and a subsequent decline in pre-induction levels within 6 h (Figure 3d). We next analyzed the effects of the CBL1 knock-out on the expression of stress-induced marker genes (Figure 3d) under these conditions. After application of 100 mm NaCl, CBF1 as well as CBF2 gene expression were induced in wild-type plants, reaching a maximum 12 h after the onset of stimulus application. RD29A mRNA expression was also induced in wild-type plants peaking 6 h after salt application. In contrast, cbl1 plants displayed a strongly altered pattern in the expression of all three genes. Here, mRNA expression was immediately induced within 1 h after stimulus application, and rapidly declined thereafter (Figure 3d). These data suggest that cbl1 mutant plants cannot sustain the signaling and adaptation responses required to establish adjustment to saline environments for a sufficiently long period.
The cbl1 mutation affects cold-induced signaling but not ABA responses
Accumulating experimental evidence points to a common set of signal transduction pathways triggered during stress responses, especially during adaptation to cold, drought, and salt stresses (Kreps et al., 2002; Xiong et al., 2002a). Also, it is well established that the plant hormone ABA is involved in numerous interactions between the plant and its environment by modulating signaling and adaptation reactions in response to stress stimuli (Bonetta and McCourt, 1998; Finkelstein et al., 2002). We therefore comparatively analyzed the expression of stress-responsive marker genes in wild-type and mutant plants exposed to either cold treatment or ABA. For cold treatments, 2-week-old soil-grown plants were transferred from 21 to 2°C, and gene expression was monitored in samples harvested at the indicated time points (Figure 4a). The expression of the PR1 gene was not affected in wild-type or cbl1 mutant plants, again confirming the specificity of the applied stimulus. The amount of CBL1 as well as CBF1, CBF2, and RD29A mRNAs increased steadily over the investigated 24-h time period. In contrast, in mutant plants, the CBF1, CBF2, and RD29A genes exhibited an accelerated but unsustained induction of expression as that observed before with salt stress response. CBF1 and CBF2 expression in mutant plants peaked within 1 h and declined thereafter. The level of RD29A mRNA reached a maximum after 3 h and continuously decreased during further exposure to cold. Taken together, these data indicate that cold-induced signaling and adaptation are altered by the loss of CBL1 function.
We next compared the survival rates of mutant and wild-type plants, testing several freezing stress regimes (see Experimental procedures). These analyses revealed no significant differences independent of inclusion of an adaptation period at 4°C prior to application of the freezing stress. It remains to be established whether functional overlaps with related CBL proteins or rather the inherent high freezing tolerance of the investigated Wassilewskija ecotype disguised potential physiological differences in cold adaptation between wild-type and mutant plants.
To explore the responses to the hormone ABA, plants grown in liquid culture were treated with 250 µm ABA, and gene expression was monitored over a 24-h time period. In contrast to all previous experiments, CBL1 expression did not respond to ABA application. On the other hand, in both wild-type and cbl1 mutant plants, ABA treatment caused the induction of CBF1, CBF2, and RD29A mRNAs, with very similar kinetics in mutant and wild-type plants. These results clearly indicate that loss of CBL1 function does not interfere with ABA-regulated gene expression. To further corroborate these findings, we comparatively analyzed seed germination inhibition by ABA (Figure 4c) as well as the survival rates of ABA-treated seedlings (data not shown). Again, no differences were observed between cbl1 mutant plants and the wild type. In addition, we investigated the responses to the plant hormones auxin, ethylene, cytokinin, and jasmonic acid by either hypocotyl elongation or root growth assays. Like for ABA, we did not detect an influence of the cbl1 mutation on any of these plant hormone responses (data not shown). Taken together, these results suggest that loss of CBL1 function specifically and simultaneously affects the plant's capability to cope with cold, drought, and salt stresses.
Constitutive overexpression of CBL1 induces stress-regulated genes and enhances drought stress tolerance
Our analyses of the CBL1 loss-of-function allele revealed the requirement of this calcium sensor protein for appropriately mediating responses to abiotic stresses in plants. To further verify the direct involvement of CBL1 in these processes, we complemented the mutant by transformation with a 6.8-kb genomic fragment encompassing the entire coding region of CBL1 as well as the promoter and 3′-untranslated sequences. The kinetics of water loss in the transformed plants was indistinguishable from that of the wild type (data not shown), indicating cbl1 mutation to be the direct cause of the observed drought phenotype.
To unravel the consequences of the induction of CBL1 expression during stress responses, we analyzed the effects of constitutive CBL1 overexpression. Transgenic plants expressing the CBL1 cDNA under the control of the CaMV 35S promoter were generated, and expression of CBL1 was monitored in 20 independently transformed lines under non-stress conditions. Most of these lines exhibited moderate-to-strong (4- to >20-fold enhanced) expression of the transgene. Accordingly, representative lines were designated as moderate (+, 4-fold), medium (++, 8-fold), and strong (+++, 20-fold) overexpressors (Figure 5a). Comparative expression analyses of the described stress marker genes were performed by quantitative RT-PCR on RNA isolated from plants grown under non-stress and under salt-stress conditions (Figure 5). As expected, in non-stressed wild-type and CBL1 mutant plants, no significant induction of the mRNA amounts of all investigated genes (CBF1, CBF2, RD29A, PR1, actin2) was detected. In contrast, already moderate constitutive overexpression of CBL1 caused an enhancement of RD29A as well as CBF1 and CBF2 gene expression (Figure 5a). Further increase of CBL1 mRNA amounts led to a corresponding rise of RD29A mRNA levels. Surprisingly, we reproducibly observed a decrease of CBF1 and CBF2 mRNA amounts (as compared to the moderate overexpressing lines), when CBL1 expression exceeded a certain threshold in medium and strongly overexpressing plants, possibly suggesting negative feedback regulation (see Discussion). Nevertheless, in these lines, CBF1 and CBF2 mRNA levels remained higher than in the wild-type. We next investigated the expression of the stress-induced marker genes in wild-type and CBL1-overexpressing plants treated with 100 mm NaCl. Interestingly, under these conditions, the induction of the marker genes showed the same time course as observed in wild-type plants. For PR1 and actin, no induction was detectable. In contrast to the more rapid induction of marker genes in cbl1 mutants, a similar effect of gene expression was not detectable in CBL1 overexpressors. In contrast, in the CBL1-overexpressing plants, all marker genes showed a higher expression level compared to wild-type plants before stress application, as well as at every analyzed time point of the stress response (Figure 5b). Remarkably, CBL1 expression in the overexpressing transgenic lines was not further inducible.
To address the physiological consequences of the observed changes in gene expression, we quantitatively determined the water loss of medium CBL1-overexpressing plants (Figure 5c). These analyses revealed a moderate but stably reproducible further reduction of transpirational water loss in CBL1-overexpressing lines as compared to wild-type plants. These findings suggest that the amount of CBL1 quantitatively regulates stress response. Taken together, the molecular and physiological phenotypes of CBL1 loss-of-function as well as overexpressing plants point to a crucial and convergent role of this calcium sensor protein in establishing and coordinating plant signaling and adaptation responses to abiotic stress factors.
Previous studies have identified a complex plant signaling network comprising CBL calcium sensor proteins and CIPK target kinases (Albrecht et al., 2001; Kudla et al., 1999; Shi et al., 1999). Despite accumulating evidence for the involvement of these proteins in deciphering calcium signals, the function of most of these signal transducers in planta has remained elusive. Therefore, the major aim of this study was to determine the in vivo function of the calcium sensor CBL1. The data presented here provide evidence for a crucial role of CBL1 in transducing and coordinating calcium-mediated plant signaling and adaptation responses during cold, drought, and salt stresses. The conclusion that CBL1 represents an integrative node in the multifaceted adaptation to environmental cues has important implications for the recognition of potential mechanisms, warranting the required specificity as well as the necessary cross-talk in signal transmission. Moreover, our finding that CBL1 function does not interfere with ABA responses, but regulates the master CBF/DREB-type stress transcription factors, points to a pivotal position of CBL1 in plant calcium signaling.
In the course of this work, we have isolated a T-DNA induced knock-out allele of CBL1 and have also generated transgenic plants constitutively overexpressing this calcium sensor protein. These plant materials allowed us to explore the physiological and molecular aspects of CBL1 function. We have shown that the CBL1 knock-out allele renders plants hypersensitive to drought and salt stresses, thus establishing the importance of this calcium sensor for appropriate adaptation to abiotic environmental factors. At the molecular level, mutation of CBL1 leads to a severely altered regulation of stress-responsive genes like CBF1, CBF2, and RD29A during drought, cold, and salt stresses. In wild-type plants, CBL1 mRNA induction during drought and salt stresses precedes (and during cold stress parallels) the increased accumulation of these stress-responsive transcripts. In the cbl1 mutant, these genes are more rapidly induced by abiotic stresses, and mRNA levels tend to decline faster than in the wild-type. This inability to faithfully regulate stress-induced genes most likely reflects the observed failure of mutant plants to properly adapt to the respective changes in the environmental conditions. The fact that the observed hypersensitivity of the cbl1 mutant plants to drought stress is not accompanied by changes in CBF4 expression could indicate that CBL1 contributes to plant drought stress responses by a CBF4-independent pathway. Moreover, these findings might point to a yet unexplored complexity in drought stress adaptation in Arabidopsis. The accelerated induction of CBF1, CBF2, and RD29A during stress impact in mutant plants represents a rather surprising finding and currently remains difficult to explain. It seems unlikely that it is because of a negative regulatory role of CBL1 in abiotic stress signaling, as the cbl1 mutant exhibits diminished stress tolerance. Several alternative interpretations of this observation are equally tenable and not mutually exclusive. First, binding and thereby buffering of Ca2+ ions by CBL1 could be required for the appropriate generation of calcium transients, which may then temporally regulate downstream target genes. Second, in the non-induced and non-stressed stage, CBL1 might be involved in keeping stress-responsive genes in a repressed state (ready for fast transcriptional activation by Ca2+), which is released upon continued stress exposure. In the cbl1 mutant, this repressive function would be missing resulting in an accelerated induction during the early phase of stress exposure. It is noteworthy that recent genetic studies have identified fry2/CPL1 as a transcriptional regulator repressing CBF/DREB transcription factors (Koiwa et al., 2002; Xiong et al., 2002b). Similar repressor proteins could represent targets of CBL1 signaling.
So far, SOS3/CBL4 has been the only calcium sensor protein implicated in salt stress responses, and consequently, a linear pathway for calcium-mediated salt stress signaling has been postulated (Xiong et al., 2002a). Our finding that CBL1 contributes to salt stress reactions suggests a more complex signaling network mediating salt adaptation and, in addition, suggests a molecular basis for the interconnection with other stress adaptation reactions. This hypothesis is further substantiated by our most recent finding that CBL1 physically interacts with the kinase SOS2/CIPK24 (Albrecht, 2002). Moreover, our observation that in different tissues different CBL proteins might specifically convey the adjustment to saline conditions indicates a tissue-specific action of different CBL family members in the cellular adaptation to saline conditions. In addition, the existence of 10 CBL calcium sensor proteins and 25 interacting CIPK-type kinases would allow a significant amount of functional overlaps of these proteins. This could, at least in part, explain the rather subtile phenotypes of the cbl1 mutant.
Despite the observed effects of the cbl1 knock-out on salt, drought, and cold stress responses, this mutant is indistinguishable from wild-type plants with regard to its responsiveness to the phytohormone ABA. Considering the well-established intense cross-interference of ABA with drought, salt, and cold-induced gene regulation and physiological responses, this observation points to a function of CBL1 upstream (or at least independent) of any modulation of stress signaling by ABA. This facet of our results is difficult to reconcile with a recently published report in which CBL1 (there designated as SCaBP5) was postulated to function as a global regulator of ABA responses (Guo et al., 2002b). The authors of this study pursued an RNA interference (RNAi)-based approach to downregulate CBL1 mRNA by overexpression of a dsRNA corresponding to positions +35 to +495 of the CBL1 cDNA. A problem with this approach is that the Arabidopsis genome harbors another calcium sensor gene (CBL9, Accession no. AF411958), which is very closely related to CBL1. The region of the CBL1 cDNA used for RNAi shares 85% nucleotide sequence identity with the corresponding CBL9 cDNA fragment. Notably, it harbors five stretches of absolutely identical sequences. These stretches range from 20 to 35 nucleotides – lengths that are generally considered to be sufficient for efficiently triggering RNAi (Elbashir et al., 2001). Therefore, it appears likely that the RNAi approach accidentally led to the effective silencing of the weaker expressed CBL9 gene. This interpretation is in agreement with our preliminary characterization of a T-DNA-induced cbl9 knock-out mutant that exhibits a specific impairment of ABA responses (C. D'Angelo and J. Kudla, unpublished). Our finding that CBL1 is involved in abiotic stress signaling rather than in mediating ABA responses is further supported by an independent study of an unrelated cbl1 knock-out mutant (S. Luan, personal communication).
Our analysis of plant lines constitutively overexpressing CBL1 revealed a proportional positive correlation of CBL1 transcript levels with the amount of RD29A mRNA under non-stressed conditions. Together with our finding that CBL1 overexpression confers enhanced stress tolerance (e.g. by reducing transpirational water loss), this might suggest the amount of CBL1 as a rate-limiting step in calcium-mediated stress signaling, at least in respect to the strength of expression of downstream target genes. Our observation that CBL1 overexpression does not change the kinetics of target gene induction compared to wild-type suggests independent mechanisms fine-tuning plant stress adaptation. Again, this might explain the rather subtile phenotypes observed in this and other stress-signaling mutants. The fact that the pathogen and wound-inducible PR1 gene expression is not influenced in CBL1-overexpressing lines furthermore suggests that the function of this calcium sensor is confined to abiotic stress signaling. Interestingly, overexpression of CBL1 stress-independently induced the early response transcription factors CBF1 and CBF2, indicating that these important regulators of stress responses represent targets of CBL1-relayed calcium signals. In contrast, the finding that CBF4 gene expression is not influenced in the cbl1 mutant during drought response might indicate that this transcription factor is not part of a CBL1-regulated signaling circuit. Our observation that CBL1 overexpression, when exceeding a certain threshold, resulted in attenuation of CBF/DREB induction (but not RD29A induction) may be indicative of an effective feedback regulation of these transcription factors. The recent isolation and analysis of the LOS1 locus (Guo et al., 2002a) encoding a translation elongation factor 2-like protein have already provided evidence for such a feedback regulation. The los1-1 mutation appeared to block events downstream of the induction of CBF/DREB genes, but yet resulted in an overaccumulation of CBF mRNAs, suggesting a disturbed feedback regulation of CBF expression. In an attractive alternative model, induction of CBL1 over a critical threshold level would induce the downregulation of the CBF transcription factors and thus abolish their transient accumulation. Interestingly, the kinetics of CBF1, CBF2, and RD29A expression in cbl1 mutant plants appear to resemble the expression of these genes after ABA treatment. A possible explanation for this situation could be that ABA and CBL1 signaling are functionally opposite to each other in regulating the promoter activity of these genes. Genetic approaches have recently identified numerous factors involved in plant abiotic stress signaling and adaptation (Xiong et al., 2002a). Their molecular analysis revealed increasingly complex, if not confusing, effects of the characterized mutations on the expression of stress-responsive genes. To the best of our knowledge, all previous mutant analyses revealed an impairment of only a subset of abiotic stress responses. This clearly distinguishes CBL1 from other known signaling components in that the calcium sensor CBL1 is a candidate for serving an integrative and general function in abiotic stress response by acting very close to the inception of these stresses and upstream of ABA involvement in stress signal transduction. Our recent finding that CBL1 specifically, albeit with gradual differences in affinity, interacts with a subset of six CIPKs (Albrecht, 2002) may provide a plausible molecular basis for the specific downstream channeling of the received abiotic signals.
General methods, plasmid construction, plant material, and transformation
Molecular biology techniques were performed using standard protocols (Sambrook and Russel, 2001). A list of primers used in this work can be obtained upon request. Plant DNA isolation, Southern blot analyses, hybridizations, DNA sequencing, and computational analyses were performed as described previously by Kudla et al. (1999). For cbl1 mutant complementation, a cosmid harboring the complete CBL1 locus was isolated from an ordered library (Klein et al., 1994). After restriction mapping, a 6.8-kb SalI/SstI subfragment encompassing the CBL1 locus from position −3916 bp upstream of the ATG to +1035 bp downstream of the stop codon was cloned into the binary vector pGPTV-Bar (Überlacker and Werr, 1996). For overexpression of CBL1, the complete cDNA (Accession no. AF076251) was cloned into the binary vector pGPTV-Bar. Plant transformations were performed by vacuum infiltration using the Agrobacterium tumefaciens strain GV3101::pMP90. Seeds were sown in the soil and selected for basta resistance by spraying young seedlings with 0.1% basta solution containing 0.1% Tween-20. Arabidopsis thaliana cv. Wassilewskija plants were used in this work and, except for the indicated stress treatments, cultivated at 21°C in a growth chamber under an 8-h light/16-h dark cycle.
Isolation of a cbl1 T-DNA insertion mutant
In this study, a collection of 36 500 T-DNA insertion lines available from the Nottingham Arabidopsis Stock Center (NASC) was screened for a CBL1 knock-out allele. An individual T-DNA insertion was identified in the Arabidopsis lines generated in the Feldmann laboratory (Forsthoefel et al., 1992) and confirmed by sequencing of the isolated PCR product as described previously by Gaedeke et al. (2001). Individual homozygous cbl1 mutant plants were subsequently identified from a plant population (400 plants) corresponding to the respective DNA pool (derived from 20 plants) by PCR with either two gene-specific primers (yielding no product) or the combination of a T-DNA primer with a CBL1-specific primer (amplifying a PCR product harboring parts of the T-DNA and the CBL1 locus). Southern blot analyses were performed with T-DNA-specific as well as CBL1-specific DNA fragments. For co-segregation analysis, a homozygous cbl1 mutant was crossed with the corresponding wild-type (ecotype Wassilewskija). Segregation analyses of the F2 progeny revealed a 3 : 1 (133 : 43 kanamycin-resistant:kanamycin-sensitive) segregation ratio for the dominant selection marker gene. RT-PCR analyses to analyze the expression of CBL1 and the actin gene serving as a control in mutant and wild-type plants were performed with 200 ng cDNA (38 cycles: 1 min at 94°C, 1 min 30 sec at 58°C, 1 min at 72°C) in a Stratagene robocycler.
Stress treatments and water loss measurements
Tissue samples in all stress experiments described were taken after 0, 1, 3, 6, 12, and 24 h, immediately frozen in liquid nitrogen, and stored at −80°C until use. Cold treatments were performed with 14-day-old soil-grown seedlings, which were transferred from 21 to 2°C and kept under these conditions for 24 h in the dark. Freezing tolerance was assayed as follows: wild-type and mutant plants were raised in a growth chamber under an 8-h light (at 22°C)/16-h dark (at 18°C) regime. Following exposure to −2°C for 12 h in the dark, plants were transferred back to 22°C. Leaves from wild-type and cbl1 plants showed equally severe symptoms of freezing-induced damage, but the plants eventually recovered and survived. In an alternative experimental set-up, plants were subjected to cold acclimation at 4°C for 48 h, then treated at −5°C in the dark for 8 h, followed by 16 h at 4°C and subsequently transferred to 22°C. Neither the wild-type control nor the knock-out plants survived this treatment. To obtain drought-induced RNAs for the transcriptional analyses, plants were grown hydroponically in magenta boxes for 14 days on 0.5× MS medium with 0.5% sucrose. The rafts with the plants were taken out of the boxes and transferred into a flow chamber to expose the plants to a stream of dry air in a flow bench for 20 min (resulting in approximately 10% FW loss) and were then returned into the magenta boxes. Afterwards, the magenta boxes were closed and samples were taken at the described time points. The drought resistance experiments were performed with plants grown in soil for 6 weeks under continuous irradiation. Water was withheld for 10 days, and the phenotypes were photographically documented. Control plants were watered every other day during this period. After three more days, plants reached complete dehydration and were subsequently re-watered to analyze the ability to recover. To investigate the salt and ABA responses by RT-PCR, plants were grown in liquid culture (0.5× MS medium with 0.5% sucrose) for 14 days. The cultivation medium was then supplemented with either 100 mm NaCl or 250 µm ABA, and samples were taken at the respective time points. For analysis of salt tolerance, wild-type and cbl1 seeds were germinated on 1× MS plates for 4 days, then transferred to plates supplemented with 100 mm NaCl and grown for another 8 days. The number of surviving plants was estimated by counting plants developing true leaves. To determine the salt tolerance of adult plants, cbl1 and wild-type plants were grown in the soil for 8 weeks under short-day conditions and treated once with 100 ml 100 mm NaCl or alternatively with 200 mm mannitol. Plants were subsequently transferred to the greenhouse (long-day growth conditions), and pictures were taken after 1 week. For the analysis of inhibition of seed germination by exogenously applied ABA, seeds were plated on 0.5× MS medium supplemented with 0.5% sucrose, 0.7% BactoAgar (Difco, Detroit, USA) and appropriate concentrations of ABA and incubated at 4°C for 3 days for stratification of seeds and break of dormancy. After incubation for 4 days at 22°C with 16 h day−1 illumination, the total number of plated seeds was counted and germinated seedlings were scored. Germination rate was calculated as percent germination on ABA-containing medium in comparison to medium without the phytohormone. Germination on medium without ABA was used as 100% value for each batch analyzed. Two independent seed batches of wild type and cbl1 were tested in triplicate. To investigate the sensitivity of seedlings to ABA, 4-day-old seedlings grown without ABA were transferred on plates containing 50 µm ABA. The number of developed true leaves was scored after 12 days of further cultivation. To determine the drought resistance in a quantitative manner, the leaf water loss was assayed as follows: after cultivation in the soil for 8 weeks, six leaves were detached from each plant and placed in a weighing dish. Dishes were kept at 40% humidity in a growth chamber, and the loss of FW was determined at the indicated time points. All stress assays were performed at least in duplicate (drought, cold, ABA, and salt treatment for RT-PCR analyses) or in triplicate (drought and salt resistance, ABA response analyses, freezing tolerance, water loss assays).
Quantitative RT-PCR analyses
RNA was isolated from 150 mg tissue samples using TriFast solution (Peqlab, Erlangen, Germany). cDNA syntheses were performed using 3 µg random hexamer primer and 200 U MMLV reverse transcriptase (Promega, Mannheim, Germany) for 1 h at 42°C. Two hundred nanograms cDNA was used as template in all PCR reactions with gene-specific primers. In preliminary experiments, we first determined the cycle number that were within the linear range of PCR product amplification for each gene to be analyzed. Subsequently, PCR amplifications (1 min at 94°C, 1 min 30 sec at 58°C, 1 min at 72°C; volume 50 µl) were performed for 23 cycles, a cycle number found to be within the linear range for all RNAs assayed. The six transcripts were assayed simultaneously for wild-type and mutant plants in separate tubes using identical cDNA aliquots and the respective gene-specific primers. Ten-microliter aliquots of each amplification reaction were separated by agarose gel electrophoresis and transferred onto nylon membranes. All RT-PCR reactions were performed in triplicate. Hybridizations were carried out with an equimolar mixture of DNA fragments specific for the seven genes and radioactively labeled with a Ready Prime kit (Amersham Pharmacia, Freiburg, Germany). For signal detection and quantification, membranes were either exposed to X-ray films or evaluated by phospho-imaging.
We thank Anke Berger for assistance during plant cultivation and mutant characterization and the Arabidopsis stock center for providing various DNA and seed stocks. This work was supported by a joint grant from the Deutsche Forschungsgemeinschaft to K.H. (HA 2146/3-2) and J.K. (KU 931/3-2).