Complexity of the Ca2+ signals and sensors

In plants, Ca2+ is involved in a variety of cellular and developmental processes such as guard-cell turgor control, pollen-tube growth, and root-hair elongation (Blatt 2000; Holdaway-Clarke et al. 1997; Wymer et al. 1997). Moreover, it also plays a critical role in mediating a wide range of extracellular stimuli such as light, plant hormones, pathogen attack, and abiotic stresses (Ehrhardt et al. 1996; Rudd and Franklin-Tong 2001; Sanders et al. 1999). How can the simple cation Ca2+ serve as a second messenger in so many diverse signal transduction pathways and yet manage to elicit appropriate plant responses? The answer to this question lies in the complexity of Ca2+ signal itself and numerous Ca2+ sensors with different properties.

Ca2+ signals can be complex because they are constituted not only of concentration but also the temporal and spatial parameters, which include frequency, magnitude, duration, and subcellular localization of the transient Ca2+ increases in the cytosol (Dolmetsch et al. 1997, 1998; Evans et al. 2001; Rudd and Franklin-Tong 2001).

Three major families of EF-hand Ca2+-binding proteins

In order for the diverse Ca2+ signals to elicit stimulus-specific responses, plant cells should be equipped with various Ca2+ sensors which detect and transduce changes in the Ca2+ parameters. As a matter of fact, plants possess numerous Ca2+ sensors that are mostly proteins with the EF-hand Ca2+-binding motif (Kawasaki et al. 1998). These Ca2+-binding proteins can be classified into three major families; calmodulin (CaM) and CaM-like proteins, Ca2+-dependent protein kinases (CDPKs), and calcineurin B-like (CBL) proteins.

First, the family members of CaM and CaM-like proteins, which do not possess any enzymatic activity themselves, act as “sensor relays” (Luan et al. 2002; Zielinski 1998). Upon Ca2+ binding, they undergo conformational changes to associate with and regulate various target proteins such as NAD kinase, glutamate decarboxylase, Ca2+-ATPase, and protein kinases (Luan et al. 2002; Reddy et al. 2002; Snedden and Fromm 2001; Yang and Poovaiah 2003; Zielinski 1998). The Arabidopsis genome contains 50 CaM-like genes in addition to the six genomic loci encoding typical CaMs (Luan et al. 2002; McCormack and Braam 2003).

Second, the CDPK family members are classified as “sensor responders”, because they harbor a kinase domain at the N-terminal end along with the C-terminal CaM-like Ca2+ sensor (Sanders et al. 2002). It is well known that Ca2+ can activate the kinase activity of CDPKs by binding to the CaM-like region (Chandran et al. 2006; Cheng et al. 2002; Harper et al. 2004; Hrabak et al. 2003). The Arabidopsis genome was predicted to carry 34 CDPK genes, which appear to possess different properties in terms of expression pattern, subcellular localization, Ca2+-binding affinity, and substrate specificity (Asano et al. 2005; Harper et al. 2004; Hernandez Sebastia et al. 2004; Hrabak et al. 2003; Lee et al. 1998; Rodriguez Milla et al. 2006; White and Broadley 2003).

The third major family of EF-hand Ca2+ sensors is CBL proteins, which are most similar to the regulatory B subunit of the protein phosphatase calcineurin (CNB) in animals (Kudla et al. 1999; Liu and Zhu 1998). Following the initial discovery from Arabidopsis, many CBLs were identified from various plant species including rice, maize, and pea (Kolukisaoglu et al. 2004; Mahajan et al. 2006; Wang et al. 2007). Therefore, it is believed that the CBL proteins are ubiquitously present in the plant kingdom (Batistic and Kudla 2009). Interestingly, the CBL family is comprised of 10 genes in both Arabidopsis and rice plants (Kolukisaoglu et al. 2004; Luan et al. 2002).

CBLs associate with a group of serine/threonine protein kinases, CIPKs

The CBL members appeared to be sensor relays like CaM, and therefore they should function by interacting with and modulating their target proteins in a Ca2+-dependent manner. In fact, several research groups demonstrated that CBLs have interaction partners which are mainly a group of serine/threonine protein kinases designated as CBL-interacting protein kinases (CIPKs) (Batistic and Kudla 2009; Cheong et al. 2007; Halfter et al. 2000; Huang et al. 2011; Kim et al. 2000; Kolukisaoglu et al. 2004; Quan et al. 2007; Shi et al. 1999; Waadt et al. 2008). The CIPK proteins are unique to plants in that they contain a distinct regulatory domain in the C-terminus, although their N-terminal kinase domain is most related to those of the yeast SNF1 protein kinase (Sucrose NonFermenting 1) and the mammalian AMP-dependent protein kinase (Shi et al. 1999).

Sequence analysis predicted that the CIPK family consists of 25 and 30 genes in the Arabidopsis and rice genomes, respectively (Kolukisaoglu et al. 2004). It appears that each CIPK can interact physically and specifically with more than one CBL members at different affinities through the conserved NAF (or FISL) motif present in the C-terminal non-kinase domain (Albrecht et al. 2001; Batistic and Kudla 2004; Halfter et al. 2000; Jeong et al. 2005; Kim et al. 2000; Kolukisaoglu et al. 2004; Luan et al. 2002).

CBLs activates target CIPKs in a Ca2+-dependent manner

CBL1 formed a complex with CIPK1 only in the presence of Ca2+ (Shi et al. 1999). In contrast, other CBL members, such as CBL2 and CBL4 (also known as SOS3), did not display such Ca2+-dependency: CBL2 and CBL4 interacted regardless of Ca2+ with CIPK14 and CIPK24 (also known as SOS2), respectively (Akaboshi et al. 2008; Halfter et al. 2000). These findings together suggest that the Ca2+ requirement for the complex formation can vary among the participating CBL and CIPK members. Nevertheless, it is very important to note that in both cases Ca2+ is still necessary for the CBL–CIPK complex to be active and phosphorylate substrates (Halfter et al. 2000) (Fig. 1).

Fig. 1
figure 1

Two hypothetical models for the activation of CIPKs by Ca2+-bound CBLs. a Ca2+-dependent interaction. Upon Ca2+ binding, CBL1 undergoes conformational change and can associate with the NAF motif (cross) of CIPK1. The association disrupts the intramolecular interaction of CIPK1, thereby activating the kinase activity. b Ca2+-independent interaction. CBL4 can interact with the NAF motif of CIPK24 in the absence of Ca2+. However, the interaction itself is not enough to block the autoinhibitory interaction of CIPK24. CBL4 requires Ca2+ to activate the enzyme activity

Full size image

Deletion analysis with CIPK24 provided a molecular mechanism by which the CBL–CIPK interaction results in activation of the CIPK enzyme activity (Guo et al. 2001). The results of this study indicated that the NAF-carrying C-terminal domain serves as an auto-inhibitory module by interacting intramoleculary with the kinase domain, and thereby blocks the active site from binding with a substrate. It seems that upon interaction with a Ca2+-bound CBL partner, CIPKs undergo a conformational change which disrupts the auto-inhibitory effect and allows the kinase domain to gain the phosphorylation activity.

CBL–CIPK complexes mediate a variety of extracellular stimuli

Extensive studies with Arabidopsis mutants have demonstrated that the CBL–CIPK complexes are involved in mediating Ca2+ signals elicited by diverse stimuli such as cold, ABA, salinity, osmotic stress, low K+ concentration, and high pH (Cheong et al. 2003, 2007; D’Angelo et al. 2006; Fuglsang et al. 2007; Kim et al. 2007; Li et al. 2006; Pandey et al. 2004; Quan et al. 2007; Xu et al. 2006). It appears that each pair of the CBL–CIPK complexes plays a specific role in relaying different signals (Fig. 2).

Fig. 2
figure 2

The CBL–CIPK networks mediating various stress responses. The signaling network was generated using the results of genetic and biochemical analyses. Different line colors indicate different stress signaling pathways. Broken lines represent signaling pathways which need further solid evidence to make sure (color figure online)

Full size image

Abscisic acid

Abscisic acid (ABA) is a phytohormone involved in plant response to abiotic stresses such as drought and high salt (Yamaguchi-Shinozaki and Shinozaki 2006; Zhu 2002). A specific Ca2+ signature is known to be implicated in an early step of the ABA signaling pathways (Allen et al. 2000, 2001; Leung and Giraudat 1998), suggesting involvement of Ca2+ sensors in this signaling pathway. In fact, several lines of recent evidence suggested that the CBL–CIPK pathways should play a role in the ABA signaling.

The Arabidopsis mutant plants with disrupted CBL9 (cbl9) were hypersensitive to ABA in the early developmental stages such as seed germination and post-germination seedling growth (Pandey et al. 2004). The cbl9 mutant seedlings also accumulated much higher levels of ABA than the wild-type plants under the osmotic stress conditions caused by mannitol and salt. Furthermore, the expression levels of the genes involved in ABA signaling, such as ABA-INSENSITIVE 4 and 5, were increased to a greater extent in the cbl9 mutant plants under the osmotic stress conditions or exogenous ABA (Pandey et al. 2004). These results strongly indicate that CBL9 plays an important role in both the biosynthesis and sensitivity of ABA in Arabidopsis. It was demonstrated that CBL9 can form a specific complex with CIPK3 to act together in regulating the ABA responses (Pandey et al. 2008). The CIPK3 knock-out Arabidopsis mutant (cipk3) plants were more sensitive to exogenous ABA during seed germination, and they also expressed significantly lower levels of ABA-induced genes such as RD22 and RD29B (Kim et al. 2003).

CBL1 is the most similar isoform of CBL9 in Arabidopsis. The Arabidopsis plants lacking the CBL1 activity (cbl1) did not show changes in the ABA responsiveness, although they exhibited less tolerance to drought and salt stress (Albrecht et al. 2003; Cheong et al. 2003). These findings indicated that CBL1 is not implicated in the ABA signaling processes unlike CBL9. Meanwhile, it is interesting to note that CIPK1, which can interact with CBL1 and CBL9, mediates ABA responses as well as osmotic stress responses such as drought and salt. Loss of CIPK1 rendered the plants hypersensitive to the osmotic stresses and it also impaired the ABA responsiveness (D’Angelo et al. 2006). Based on these findings, it is believed that CIPK1 can mediate ABA-independent and ABA-dependent stress responses by forming an alternative complex with either CBL1 or CBL9.

Silencing of the Arabidopsis CIPK15 gene (designated PKS3 in the paper) by an RNA interference technique rendered the plants hypersensitive to ABA in seed germination, seedling growth, stomatal closing, and gene expression (Guo et al. 2002). These findings implicated CIPK15 as a global negative regulator of ABA responses. Because yeast two-hybrid assays demonstrated that CIPK15 associates with a group of CBL family members including CBL1, 2, 3, 5, 8, and 9 (Kim, unpublished data), it is likely that CBL9 may also form a complex with CIPK15 to mediate the ABA responses. However, further genetic analysis is required to demonstrate this speculation.

Salt stress

A series of genetic and biochemical analyses with the salt overly sensitive (SOS) mutants provided a molecular mechanism by which a CBL–CIPK complex mediates the salt stress-induced Ca2+ signal and give rise to salt tolerance (Zhu et al. 1998). CBL4 and CIPK24, encoded, respectively, by the genomic loci SOS3 and SOS2, form a complex and activate the CIPK24 kinase activity in a Ca2+-dependent manner (Halfter et al. 2000; Liu et al. 2000; Liu and Zhu 1998). Then, the CBL4/SOS3-CIPK24/SOS2 complex phosphorylates and activates the downstream component SOS1, a Na+/H+ antiporter (Shi et al. 2000). The activated SOS1 protein functions to get rid of excess Na+ in plant cells, thereby conferring salt tolerance (Qiu et al. 2002; Quintero et al. 2002).

Another CBL family member, CBL10, was later included in the salt-tolerance pathway. Like CBL4/SOS3, CBL10 also associates with and activates the kinase activity of CIPK24/SOS2 (Kim et al. 2007; Quan et al. 2007). Although both CBL4/SOS3 and CBL10 proteins are involved in mediating salt tolerance, they do differ in executing their functions due to their distinct expression patterns and subcellular localizations.

CBL4/SOS3 is expressed mainly in the roots and localized at the plasma membrane. Therefore, CIPK24/SOS2 is recruited to the plasma membrane of the root cells where it phosphorylates the Na+/H+ antiporter SOS1, thereby enhancing the Na+ efflux rate (Fig. 3a). In contrast, CBL10 is expressed predominantly in the shoots and leaves, and it is localized at the vacuolar membrane (tonoplast). The knock-out Arabidopsis mutant lacking the CBL10 activity (cbl10) showed the salt-sensitive phenotype specifically in the leaves or shoots under the high salt conditions (Kim et al. 2007). Interestingly, the cbl10 mutant plants accumulated much less Na+ than the wild-type under either normal or high salt conditions. This unique feature of the cbl10 mutant plants, along with the tonoplast localization of the CBL10 protein, suggests that CBL10 is required for the sequestration of Na+ into the vacuole. It is likely that CIPK24/SOS2 recruited by CBL10 to the tonoplast may phosphorylate and activate an as-yet-unknown Na+ channel or transporter, which is tonoplast-bound and plays a role in transporting cytosolic Na+ into the vacuole (Fig. 3b).

Fig. 3
figure 3

Hypothetical model of CBL–CIPK function in salt stress response. a In the roots, CIPK24 activates the the Na+/H+ antiporter SOS1 at the plasma membrane by forming a complex with either CBL1 or CBL4. b In the leaves, CIPK24 regulates Na+ sequestration into the vacuole by interacting with the tonoplast-bound CBL10. Question mark a component of unknown identity

Full size image

Meanwhile, it should be noted that another CBL family member, CBL1, is also involved in plant response to salt stress. The cbl1 mutant plants showed less tolerance to the salt stress (Albrecht et al. 2003; Cheong et al. 2003). Like CBL4/SOS3, CBL1 is present in the plasma membrane and interacts with CIPK24/SOS2. Therefore, it is very likely that CBL1 also confers salt tolerance via CIPK24/SOS2-mediated SOS1 regulation (D’Angelo et al. 2006). Unlike CBL4/SOS3, however, CBL1 expression is not restricted the roots. CBL1 is actually expressed in both root and shoot tissues. Such an expression pattern implies that Na+ extrusion mediated by the CIPK24/SOS2-SOS1 machinery may also occur in the shoots. In summary, CIPK24/SOS2 can regulate two different processes depending on its interacting CBL partners. Through the association with CBL1 or CBL4/SOS3, CIPK24/SOS2 facilitates Na+ export across the plasma membrane by activating the Na+/H+ antiporter SOS1. Furthermore, CIPK24/SOS2 can also play a role in sequestrating Na+ into the vacuole by forming a complex with CBL10.

Recently, an Arabidopsis mutant lacking CIPK6 activity was reported to be more sensitive to salt stress compared to the wild-type, suggesting involvement of CIPK6 in salt tolerance (Tripathi et al. 2009). Because CIPK6 was known to associate with CBL4/SOS3 in the yeast two-hybrid system (Kim et al. 2000), there is a possibility that CBL4/SOS3 may also target CIPK6 in vivo in addition to CIPK24/SOS2. Further genetic and biochemical analyses will provide solid evidence, which will determine whether or not the CBL4/SOS3-CIPK6 complex is actually working in plant cells to mediate salt tolerance. Anyway, it is obvious that multiple members of the CBL and CIPK families are implicated in salt tolerance.

Drought and osmotic stress

Loss-of-function Arabidopsis mutants lacking CBL1, CBL9, or CIPK1 were found to be more sensitive to drought and osmotic stress than the wild-type plants (Pandey et al. 2004; D’Angelo et al. 2006). Moreover, several lines of evidence demonstrated that CIPK1 interacts with either CBL1 or its closest isoform CBL9 at the plasma membrane (Shi et al. 1999; Kim et al. 2000; Albrecht et al. 2001; D’Angelo et al. 2006). These data together suggested that both CBL1 and CBL9 can target CIPK1 to form two distinct complexes, CBL1–CIPK1 and CBL9–CIPK1, thereby mediating the stress responses. As described above, they are, respectively, involved in ABA-independent and ABA-dependent signaling pathways.

Cold

The cbl1 mutant plants exhibited more tolerance to cold stress and displayed enhanced cold induction of stress genes such as RD29A and Kin1 (Albrecht et al. 2003; Cheong et al. 2003). These results indicate that CBL1 plays a negative role in regulating plant cold response. Recently, a study showed that CBL1 can interact with CIPK7 whose expression is induced by cold stress (Huang et al. 2011). This suggests that CBL1 may target CIPK7 to mediate cold response in plants.

Meanwhile, expression patterns of RD29A and Kin1/Kin2 genes were altered in the cipk3 mutant plants in response to cold stress (Kim et al. 2003), implying the involvement of CIPK3 in cold response. Because expressions of the stress genes are independent of endogenous ABA production (Shinozaki and Yamaguchi-Shinozaki 2000; Thomashow 1999), these findings suggest that CIPK3 can act as a cross-talk node between the ABA-dependent and ABA-independent pathways in stress responses.

Potassium deficiency

Two independent research groups have provided solid evidence demonstrating that CBL1 and CBL9 calcium sensors and their target CIPK23 are involved in plant response to low K+ conditions (Li et al. 2006; Xu et al. 2006). According to their work, the Arabidopsis mutant plants with the disrupted CIPK23 gene, cipk23, were more sensitive to low K+ concentrations than the wild-type plants. The similar phenotype was also observed with the cbl1/cbl9 double mutant plants, but not with the cbl1 or cbl9 single mutants, which implies that the two CBL members have redundant functions in the low K+ signaling pathways.

Biochemical assays have demonstrated that CIPK23 can be activated by the association with either CBL1 or CBL9, thereby phosphorylating the C-terminal region of K+ transporter AKT1 in the plasma membrane. Based on these findings, a model was proposed for the K+ -deficiency response in plants: CBL1 and CBL9 detect the Ca2+ signatures elicited by low K+ conditions. Upon binding with Ca2+, they recruit CIPK23 to the plasma membrane and promote the kinase activity. The activated CIPK23 then phosphorylates the C-terminal region of the AKT1 transporter, thereby enhancing its K+-transporting activity.

High pH

The Arabidopsis mutant plants lacking the functional protein kinase CIPK11 (designated PKS5 in the paper) displayed enhanced tolerance to high external pH compared with the wild-type plants because of the increased rate of H+ secretion to the extracellular space (Fuglsang et al. 2007). Further studies revealed that CIPK11 phosphorylates the Ser-931 residue in the C-terminal regulatory domain of the plasma membrane proton pump (PM H+-ATPase 2; AHA2), which results in the inhibition of AHA2 activity by preventing its association with an activating 14-3-3 (Fuglsang et al. 2007). Although CBL2 was originally proposed to be responsible for activating CIPK11 in the presence of cytosolic Ca2+ signals elicited by external high pH conditions, the involvement of CBL2 was later ruled out due to its exclusive localization at the tonoplast (Batistic et al. 2008). Therefore, it needs to be discovered which member of the CBL family is responsible for Ca2+-dependent modulation of the AHA2 activity in planta.

Biotic stress

Plants generate cytosolic Ca2+ signals in response to pathogen attacks (Rudd and Franklin-Tong 2001). Currently, however, little is known about whether the CBL family members are involved in mediating the Ca2+ signals elicited by the biotic stress. Only circumstantial or indirect evidence suggests that CBLs may be implicated in plant response to the biotic stress; salicylic acid, mediating biotic stress response in plants, stimulated the expression of pea CBL, called PsCBL (Mahajan et al. 2006).

Perspectives

Identification of the novel Ca2+ sensors CBLs and their interaction partners CIPKs has certainly enhanced our knowledge on the Ca2+ signal transduction pathways in higher plants. Furthermore, the findings of the in vivo substrates phosphorylated by the CBL4/SOS3–CIPK24/SOS2 and CBL1 (CBL9)–CIPK23 complexes have provided molecular mechanisms by which plants respond to salt and low K+ stresses, respectively (Qiu et al. 2002; Quintero et al. 2002; Li et al. 2006; Xu et al. 2006). Currently, only three CIPK substrates are known, all of which happen to be membrane-bound proteins, SOS1, AKT1, and AHA2 (Fuglsang et al. 2007; Li et al. 2006; Qiu et al. 2002; Quintero et al. 2002; Xu et al. 2006). Therefore, it becomes critical to identify as yet unknown in vivo substrates or target molecules for the CIPK family members in order to gain further insight into the CBL–CIPK Ca2+ signaling network.

Analyses of the CBL and CIPK knock-out mutants have indicated that the CBL–CIPK Ca2+ signaling pathways are involved in regulating expression of stress genes such as RD29A/B, Kin1/2, and DREB1A/B (Albrecht et al. 2003; Cheong et al. 2003; D’Angelo et al. 2006; Kim et al. 2003; Pandey et al. 2004). However, little is known about how the CBL–CIPK complexes exert their influence on the transcriptional controls. In the cases of the CBL1 and CBL9 proteins, which recruit interacting CIPK partners to the plasma membrane (Albrecht et al. 2003; Cheong et al. 2003; D’Angelo et al. 2006; Pandey et al. 2004), we can speculate the existence of messenger molecules delivering information from the plasma membrane-bound CBL–CIPK complexes into the nucleus, where gene regulation occurs. In this regard, ECT1 can be considered as a strong candidate protein executing such function, because it can associate with CIPK1 in the cytoplasm and can be translocated into the nucleus (Ok et al. 2005). Further analyses will show whether ECT1 is actually involved in relaying a signal from the CBL1–CIPK1 complex.

Meanwhile, it is important to note a recent report that CBL3 can interact with and modulate the Arabidopsis 5′-methylthioadenosine nucleosidase (AtMTAN) in a Ca2+-dependent manner (Oh et al. 2008). Since AtMTAN associates only with CBL3 but not with other CBL family members such as CBL1 and CBL4, it is reasonable to speculate that other CBL members may also have as yet unidentified interaction partners which do not belong the CIPK family. In this case, the CBL members can target a greater number of proteins with different biochemical properties. Such target diversity confers an additional level of complexity on the CBL-mediated Ca2+-signaling pathways, thereby allowing the CBL Ca2+ sensors to regulate a wider range of cellular and physiological processes in higher plants. Anyway, the finding above clearly indicates that the CBL family can relay the Ca2+ signals in more diverse ways than currently known. Extensive further studies will be necessary to fully understand the Ca2+-signaling networks mediated by the CBL family members.