CIPK6, a CBL-interacting protein kinase is required for development and salt tolerance in plants


*(fax +91 11 26741658; e-mail


Calcineurin B-like proteins (CBL) and CBL-interacting protein kinases (CIPK) mediate plant responses to a variety of external stresses. Here we report that Arabidopsis CIPK6 is also required for the growth and development of plants. Phenotype of tobacco plants ectopically expressing a homologous gene (CaCIPK6) from the leguminous plant chickpea (Cicer arietinum) indicated its functional conservation. A lesion inAtCIPK6 significantly reduced shoot-to-root and root basipetal auxin transport, and the plants exhibited developmental defects such as fused cotyledons, swollen hypocotyls and compromised lateral root formation, in conjunction with reduced expression of a number of genes involved in auxin transport and abiotic stress response. The Arabidopsis mutant was more sensitive to salt stress compared to wild-type, while overexpression of a constitutively active mutant of CaCIPK6 promoted salt tolerance in transgenic tobacco. Furthermore, tobacco seedlings expressing the constitutively active mutant of CaCIPK6 showed a developed root system, increased basipetal auxin transport and hypersensitivity to auxin. Our results provide evidence for involvement of a CIPK in auxin transport and consequently in root development, as well as in the salt-stress response, by regulating the expression of genes.


Stress and other extracellular stimuli alter the cytosolic Ca2+ concentration in the cell (Knight et al., 1991, 1998). Calcium sensors and their interacting proteins mediate the signal to stimulate cellular responses. In addition to several calcium-binding enzymatic and non-enzymatic proteins, a group of calcium-sensing proteins named calcineurin B-like proteins (CBL) have been identified that specifically target a group of sucrose non-fermenting-related serine/threonine kinases (SnRK3), named CBL-interacting protein kinases (CIPK), to mediate the sensed calcium signal (Kudla et al., 1999; Shi et al., 1999; Luan et al., 2002; Batistic and Kudla, 2004; Kolukisaoglu et al., 2004). Ca2+-bound CBLs interact with the C-terminal NAF (Asn-Ala-Phe)/FISL (Phe-Ile-Ser-Leu) motif of CIPK to activate its catalytic domain (Sanchez-Barrena et al., 2007; Akaboshi et al., 2008). The CIPK catalytic domain contains an activation loop between the N-terminus and the C-terminal domain that may be the target of an unknown kinase. Replacement of a specific threonine with aspartic acid (T/D) in this domain converts the kinase to a hyperactive form that does not require CBL for kinase activity (Guo et al., 2001; Gong et al., 2002a,b,c). Individual CBL proteins show specificity in targeting different CIPKs. The inter-play of a complex network of CBL–CIPK interactions is believed to determine the specificity of the response. In Arabidopsis, the CBL–CIPK network of signal transduction depends upon differential expression of ten CBLs and 25 partner CIPKs in response to various stimuli in a tissue-specific manner. The interaction specificity of most of the CBLs and CIPKs has been determined out by yeast two-hybrid assays (Kim et al., 2000; Guo et al., 2001). However, functional information is available for only a few of them at present. The first genetically defined CIPK was identified in a genetic screen for a salt overly sensitive (SOS) phenotype (Liu et al., 2000). SOS2 (CIPK24) interacts with CBL4 (SOS3) under salt-stress conditions, and in turn activates a plasma membrane-localized Na+/H+ antiporter (SOS1) and vacuolar H+-ATPase to promote salt tolerance (Qiu et al., 2002; Batelli et al., 2007). Expression of CIPK3/PKS12 is enhanced during the early stages of seedling development, and the cipk3 mutant shows abscisic acid (ABA) hypersensitivity during seed germination (Kim et al., 2003). In contrast, expression of CIPK20/PKS18(T169D) rendered the transgenic plants hypersensitive to ABA, while RNAi lines showed insensitivity to ABA (Gong et al., 2002b). Overexpression of constitutively active CIPK8/PKS11(T161D) conferred resistance to high concentration of glucose (Gong et al., 2002a). CIPK23/PKS17 was isolated in a screen for sensitivity to low potassium concentrations. Its kinase activity is regulated by the upstream regulators CBL1 and CBL9. CIPK23 directly interacts with and positively regulates the potassium transporter AKT1 by phosphorylation. Homozygous cipk23 plants are hypersensitive to low K+ concentrations (Li et al., 2006; Xu et al., 2006). Arabidopsis CIPK6 and CIPK16 also interacted with AKT1 in a yeast two-hybrid assay and enhanced the activity of AKT1 in a CBL-dependent manner in Xenopus oocytes (Lee et al., 2007). PKS5 phosphorylates and inactivates plasma membrane H+-ATPase (Fuglsang et al., 2007). A similar complex network seems to exist in rice, as indicated by the presence of ten CBLs and 30 CIPKs and a conserved SOS pathway (Martnez-Atienza et al., 2007;Xiang et al., 2007). Although the functional significance of most of these molecular interactions is not clear at present, they indicate a complex, multi-layered and multi-genic signal transduction network regulated by this calcium sensor–kinase system.

However, neither ectopic expression of constitutively active forms or loss-of-function mutants of any of the kinases studied so far have been linked to any developmental phenotypic changes compared to wild-type plants under normal growth conditions. To investigate the functional role of a drought- and salt-induced CIPK homologue from the leguminous crop chickpea (Cicer arietinum), we expressed its constitutively active form in a heterologous system (tobacco). To our surprise, we observed that the transgenic seedlings displayed increased root length, overall root branching and hypersensitivity to auxin. A T-DNA insertion allele of the corresponding gene of Arabidopsis also showed defects in seedling development and auxin transport. Gene expression analysis under salt-stress and normal growth conditions provided evidence of a dual role for CIPK6 in stress responses and seedling development.


Cloning and expression analysis of a chickpea cDNA encoding a CIPK structural homologue

We previously described differential screening to isolate drought-inducible chickpea genes (Boominathan et al., 2004). Among several candidates, a partial cDNA clone (GenBank accession number CF340748) with deduced amino acid sequence homology to CBL-interacting protein kinases was chosen for further analysis. 5′ and 3′ RACE (Appendix S1) resulted in a 1917 bp cDNA clone (EU492906) consisting of a 157 bp 5′ UTR and a 425 bp 3′ UTR, and a 1335 bp ORF. The ORF encodes a polypeptide of 444 amino acids with a calculated molecular mass of 50.35 kDa. Database analysis using BLASTP revealed that the deduced amino acid sequence has high homology (E-value = 0.0) to a common bean (Phaseolus vulgaris) protein kinase (BAG06675), poplar (Populus trichocarpa) CIPK9 (ABJ91216) and a soybean (Glycine max) SOS2-like protein kinase (AAM83095). An Arabidopsis protein database search identified AtCIPK6/PKS4 (AF339145) as the closest homologue (E-value = 8e-180) of the chickpea protein (hereafter referred to as CaCIPK6). Protein sequence alignment (Figure 1a) shows that CaCIPK6, like its Arabidopsis homologue, exhibits a two-domain structure, with an N-terminal SNF1-related serine/threonine protein kinase domain and a C-terminal regulatory domain with a CBL-interacting NAF/FISL module. Apart from a few mismatches, the activation loop (DFG…APE) is conserved in the chickpea protein. Also conserved is the threonine residue (Thr181), substitution of which by aspartic acid in SOS2 and other Arabidopsis CIPKs resulted in a constitutively active kinase that does not require CBL for activity (Guo et al., 2001).

Figure 1.

 Predicted amino acid sequence of CaCIPK6 and its expression in chickpea in response to various treatments.
(a) Alignment of the deduced amino acid sequences of CaCIPK6 and AtCIPK6. The activation loop and FISL motif are indicated, and the Thr181 residue within the activation loop is indicated by a hash symbol (#).
(b, c) Expression pattern of CaCIPK6 in response to various stresses and hormone treatments, as indicated, for 6-day-old soil-grown whole chickpea seedlings (b) and in various tissues (c). The lower panels show hybridization with ribosomal RNAs as loading controls.

Genomic Southern blot analysis suggested thatCaCIPK6 is present as a single copy in the chickpea genome (data not shown). Sequence comparison of the gene and the cDNA revealed that the gene is intronless. RNA gel-blot analysis (Appendix S1) of 6-day-old whole seedlings with a CaCIPK6-specific probe showed a low basal level of expression under normal growth conditions, indicating its requirement in the normal developmental process. The transcript slowly reached the maximum level after 3 h of dehydration treatment, and this was maintained up to 24 h. Induction of gene expression was detected within 30 min of treatment with 250 mm NaCl or 100 μm ABA, reached at maximum level at 1 h, and then declined slowly to basal level at 24 h (Figure 1b). In all cases, the maximum level of induction was only approximately three times the basal expression. CaCIPK6 was strongly expressed in the stem and leaves, with almost no detectable expression in the root. However, root-specific expression increased several fold under stress conditions, indicating a root-specific function during stress (Figure 1c). A transient increase in expression was observed after 3 h of cold (4°C) treatment (data not shown). CaCIPK6 expression was analyzed in response to plant hormones, signaling molecules and physical factors. Indole-3-acetic acid (IAA) at 50 μm concentration induced accumulation of CaCIPK6 transcript within 1 h, and this persisted for up to 5 h of treatment (Figure 1b). Jasmonic acid (100 μm), salicylic acid (5 mm) and wound treatments for up to 5 h did not alter CaCIPK6 expression (Figure 1d and data not shown).

An immunoblot with a CaCIPK6 C-terminus-specific antibody detected more than fivefold increased protein expression in response to dehydration (3 h) and 250 mm NaCl (1 h) (Figure S6a), suggesting post-transcriptional regulation of CaCIPK6 expression. To determine the kinase activity of CaCIPK6 under dehydration and salt-stress conditions, equivalent amounts of cell extracts were immunoprecipitated with anti-CaCIPK6 antibody, and an autophosphorylation assay (Figure S6b) shows an approximately 2.5-fold increase in the kinase activity of CaCIPK6 under these stress conditions.

Expression of CaCIPK6 in tobacco

In order to biochemically characterize the enzyme, the cDNA was expressed in Escherichia coli as a GST-fused protein. Purified expressed protein was tested for auto- and substrate phosphorylation. We tried the commonly used protein or peptide substrates, e.g. myelin basic protein, histone H1, casein and a synthetic peptide p3 (ALARAASAAALARR) used for SOS2 (Guo et al., 2001), but were unable to demonstrate any substrate phosphorylation for CaCIPK6. GST–CaCIPK6 showed a very low level of autokinase activity in magnesium chloride buffer as previously described for other CIPKs (Liu et al., 2000; Gong et al., 2002a). Substitution of Thr181 by aspartic acid (Appendix S1) resulted in a mutated kinase (CaCIPK6T181D) that showed approximately 2.7-fold higher autophosphorylation activity (Figure 2a).

Figure 2.

 Root phenotype of the tobacco seedlings expressing CaCIPK6.
(a) Autophosphorylation assay of bacterially expressed GST-fused CaCIPK6 and CaCIPK6T181D proteins. A Coomassie blue-stained gel after SDS–PAGE shows the amount of protein loaded.
(b) Tissue-specific expression of NtCBL3 mRNA in root (R), stem (S) and leaf (L) of tobacco (top). Northern analysis was performed as in Figure 1. Interaction CaCIPK6 and NtCBL3 in a yeast two-hybrid assay (bottom). The bait plasmid pGBKT7 with (2 and 4) or without (1 and 3) the C-terminal domain of CaCIPK6 was transformed into a yeast strain AH109 containing the prey plasmid pGAD with (3 and 4) or without (1 and 2) NtCBL3. Yeast was grown on YPD medium (left panel) or synthetic drop-out medium without histidine, trypsin, adenine or leucine (right panel).
(c) Expression of CaCIPK6 in T1 seedlings of untransformed (Wt) and tobacco plants transformed with pBI121 without (vector) or with the CaCIPK6 or CaCIPK6T181D ORFs. RT-PCR was performed using 1 μg total RNA. A reaction to which no reverse transcriptase was added was also performed (−RT). Actin was used as an internal loading control.
(d) Morphology of vertically grown 6-day-old tobacco T1 seedlings untransformed (Wt) or transformed with a pBI121 vector control or with pBI121 containing CaCIPK6 and CaCIPK6T181D ORFs. Two representative lines each of CaCIPK6- and CaCIPK6T181D-expressing plants are shown.
(e) Comparison of root lengths of the vector control and two independent lines each of CaCIPK6 and CaCIPK6T181D-expressing plants. The means of three measurements of 30 seedlings each are shown.
(f) Confocal imaging of root apices from vector control (A) and CaCIPK6T181D-expressing (B) tobacco seedlings at equal magnification stained using the fluorescent dye FM4-64. The arrow marks the start of elongation zone. Scale bar = 100 μm.

We expressed CaCIPK6 and its active mutant CaCIPK6T181D under the control of the CaMV 35S promoter in tobacco to analyze the functional properties of the enzyme. A functional interaction between CBL and CIPK is necessary for enzyme activity. Arabidopsis CIPK6 has been shown to interact with CBL3 and 4 (Kolukisaoglu et al., 2004). A search in a SOL genomics network ( database identified a tobacco unigene SGNU372502 that shows 93% homology to Arabidopsis CBL3 (At4G26570) at the amino acid level (E-value = 1e-107) and was therefore designated NtCBL3. NtCBL3 is expressed in stem, root and leaf. To determine whether CaCIPK6 interacts with tobacco CBL, we used tobacco CBL3 and CaCIPK6 in a yeast two-hybrid assay (Appendix S1). The NtCBL3 ORF was cloned by reverse transcription followed by PCR using tobacco leaf mRNA, and introduced into a yeast strain harboring the C-terminal regulatory domain of CaCIPK6 (amino acids 311–444). Auxotropic selection shows that CaCIPK6 is able to interact with NtCBL3 (Figure 2b). More than ten transgenic lines for each construct, together with control lines harboring the plasmid pBI121 only, were generated using an Agrobacterium-mediated transformation method. Six-day-old T1 seedlings from two independent lines with single insertions and expressing equivalent levels of each transgene (Figure 2c) were compared with the vector control and wild-type seedlings with regard to morphological phenotype. The vector control and wild-type plants did not exhibit any phenotypic difference in these experiments. No difference was observed between the transgenic lines and the control seeds with regard to the percentage and time taken for germination. There were no apparent differences in the stem and leaf morphology of the seedlings. The roots of CaCIPK6-overexpressing (CaCIPK6OX) lines were approximately 25% (< 0.03) longer than those of control seedlings, and CaCIPK6T181D-overexpressing (CaCIPK6T181DOX) lines had almost twofold (192%) longer root lengths (Figure 2d,e). The difference in root length between control and CaCIPK6T181DOX lines was more evident in the soil-grown 12-day-old seedlings (Figure S1). CaCIPK6T181DOX lines also had longer root hairs (Figure S2). The root morphology of the overexpressing lines is surprising given the very low level of expression of the gene in the root under control conditions. However, root-specific expression of CaCIPK6 drastically increases in response to salt and dehydration stress, which suggests that this gene may function to modulate root morphology during these stresses. The higher root growth in CaCIPK6T181DOX lines compared with those expressing CaCIPK6 shows that the enzyme activity is required for the root phenotype. To investigate whether the longer root phenotype caused by CaCIPK6T181D overexpression is the result of altered cellular organization at the root tip, cell positions were visualized by confocal imaging after staining the root tips with a fluorescent dye. CaCIPK6T181DOX root tips showed no gross deformation with regard to cellular arrangement in comparison to those of the wild-type (data not shown) and the vector control line. However, CaCIPK6T181DOX roots had a significantly higher number of cells in the region above the apical meristem. Increased cell numbers are evident in the epidermal, cortical and steller layers, producing extra cell layers in CaCIPK6T181DOX roots, indicating frequent cell division. In conjunction with enhanced cell division, the zone of cell division is much longer in these seedlings (Figure 2f).

Auxin transport and sensitivity in CaCIPK6-overexpressing lines

The plant hormone auxin regulates a number of cellular and developmental processes in various tissues, including the root. Root length and lateral root formation are severely compromised in plants that are defective in auxin transport. Therefore, auxin transport at the root tips of the CaCIPK6-overexpressing lines was compared with that of the vector control seedlings by an experimental procedure as described previously (Shin et al., 2005). A significant increase in basipetal auxin transport was observed in CaCIPK6T181DOX lines when measured in tissue 2 mm above an agar block containing radioactive IAA (Figure 3a). Auxin sensitivity was tested by exposing the seedling roots to exogenous auxin. While the control roots grew to some extent during incubation with auxin, growth of CaCIPK6T181DOX roots was completely inhibited in medium supplemented with NAA (∞-naphthalene acetic acid) (Figure 3b). Prolonged incubation in the same medium resulted in robust callus formation in the CaCIPK6T181DOX roots (Figure 3c). It is notable that the Arabidopsis superroot mutant that over-produces auxin spontaneously forms rooty masses (Boerjan et al., 1995).

Figure 3.

 Auxin-transport in the tobacco seedlings expressing CaCIPK6.
(a) Comparison of basipetal root auxin transport in 6-day-old vector control and two independent lines of CaCIPK6T181DOX seedlings in a root segment 2 mm above the site of auxin application at the root tip. Data are presented as the percentage auxin transport relative to control tobacco (= three groups of 15 seedlings each).
(b) Inhibition of root growth in the presence of auxin. Six-day-old vector control (A) and two lines of CaCIPK6T181DOX (B and C) seedlings were transferred to medium supplemented with 5 μm auxin and kept for 3 days. The positions of the root tips at the time of transfer to auxin-containing medium are indicated.
(c) Vector control (A) and CaCIPK6T181DOX (B and C) seedlings as in (b) were allowed to grow in auxin-supplemented medium for a further 6 days for callus formation.

Improved salt resistance in CaCIPK6-overexpressing lines

Expression of CaCIPK6 increases during dehydration and salt stress and also under treatment with ABA, suggesting involvement of the gene in abiotic stress and ABA responses in plants. To evaluate the effect of CaCIPK6 overexpression in the abiotic stress response, seeds were germinated on 0.3 m mannitol-containing medium. Seed germination of CaCIPK6T181DOX lines was delayed in comparison to the control seeds in mannitol medium. However, after germination, the growth rate of CaCIPK6T181DOX roots was greater than that of control roots. To investigate the reason behind the delay in seed germination, the seeds were germinated in medium containing 0.3 m mannitol and 100 μm gibberellic acid (GA3) or 100 μm fluridone. Both the treatments caused germination of vector control and CaCIPK6OX seeds at the same time (Figure S3). GA3 is known to antagonize the ABA effect on seed germination, and fluridone is a carotenoid biosynthesis inhibitor, and hence an inhibitor of ABA synthesis (Hughes and Galau, 1991). Therefore, the hypersensitivity of transgenic tobacco seed germination to mannitol suggests involvement of CaCIPK6 in ABA accumulation or perception or signaling. We did not observe any significant difference in relative fresh weights between the control and transgenic seedlings under these experimental conditions.

Eight-day-old seedlings were exposed to medium supplemented with either 0.4 m mannitol or 250 mm NaCl. After 8 days, the salt-treated seedlings were transferred to control medium for recovery and grown for a further 8 days. Leaves were scored for chlorosis, and the fresh weights of the seedlings were obtained and compared with those for corresponding seedlings continuously grown in normal growth medium for the same period, and presented as relative fresh weight. Evaluation of three experiments with 50 seedlings of each line in each experiment showed that approximately 70% of the control and 60% of CaCIPK6OX leaves have undergone chlorosis. In contrast, only 10% of CaCIPK6T181DOX leaves exhibited partial chlorosis, with no total bleaching apparent (Figure 4a,b). The relative fresh weight of the control seedlings was 40% of those grown in normal medium; the fresh weight of the CaCIPK6T181DOX seedlings was 80% of that of seedlings grown in the normal medium (Figure 4c). Seedlings exposed to 0.4 m mannitol medium were kept for 20 days. We did not observe any difference in leaf colour; however, the relative fresh weight of the CaCIPK6T181DOX seedlings was twice of that of the controls (Figure 4c). To gain insight into the changes in the gene expression level, a subtracted cDNA library of 8-day-old CaCIPK6T181DOX seedlings against the vector control has been constructed, and expression of the ESTs was validated by reverse Northern analysis. A number of ESTs related to the abiotic stress response and reported as stress-tolerant determinants were identified as showing more than twofold increased expression in the transgenic plants, such as MRP4, dehydrin, ER5LEA, RAP2.4 and inositol-3-phosphate synthase (Table S1). The expression of six ESTs putatively related to auxin signaling (IAA) and abiotic stress tolerance (dehydrin, MRP4) was validated by quantitative RT-PCR (Figure 6b).

Figure 4.

 Abiotic stress-response of the tobacco seedlings expressing CaCIPK6.
(a) Effect of salt stress on tobacco seedlings. Eight-day-old vector control and two independent lines each of CaCIPK6OX and CaCIPK6T181DOX T1 seedlings were transferred to medium with (left) or without (right) 250 mm NaCl for 8 days and returned to normal medium for recovery for 8 days.
(b) Leaves were scored for chlorosis and bleaching, and the results are presented relative to the total number of leaves.
(c) Fresh weights of the seedlings from the same experiment (salt) are presented relative to the fresh weights of corresponding seedlings grown in normal medium for the same period. The experiment in (a) was repeated with 0.4 m mannitol replacing NaCl. Eight-day-old seedlings were transferred to mannitol-supplemented medium and grown for 20 days. The results (mannitol) are presented in the same way as for the salt experiment. Data for three experiments with 15 seedlings each are presented.

Figure 6.

 Gene expression analysis in atcipk6kd and CaCIPK6T181DOX plants.
(a) Relative expression analyses by quantitative RT-PCR of 16 selected genes that were identified as downregulated in atcipk6kd by more than 1.5-fold. The expression in 14-day-old wild-type plants (as described above) is shown relative to that in atcipk6kd plants. Ubiquitin was used as a control.
(b) Relative expression analyses by quantitative RT-PCR of selected six genes that were identified as upregulated in CaCIPK6T181DOX. The expression in CaCIPK6T181DOX plants relative to that in vector control plants is shown. The results for two independent transgenic lines are shown. Tobacco actin was used as internal control.

Phenotypic characterization of the atcipk6 knockdown mutant

As CaCIPK6T181D overexpression altered root morphology and the salt-response phenotype of transgenic tobacco seedlings, we wished to study its structural homologue in Arabidopsis. The expression profile of AtCIPK6 mRNA was monitored by quantitative RT-PCR. As in chickpea, the basal level of expression of the Arabidopsis gene is very low in roots (50 times lower than the total expression in seedlings), and high in stem and leaf. Dehydration (3 h) and salt treatment (150 mm, 1 h) induced expression to approximately three times the basal level expression (Figure 5a). GST–AtCIPK6 showed a very low level of autokinase activity, and the mutated kinase (AtCIPK6T182D) showed approximately 2.54-fold higher autophosphorylation activity (Figure 5b).

Figure 5.

 Phenotype characterization of the Arabidopsis atcipk6 knockdown mutant.
(a) Comparison of expression of AtCIPK6 in root (R), stem (S) and leaf (L) and whole seedlings (Col) of 10-day-old Arabidopsis (Columbia) seedlings (left panel), and fold expression under dehydration (3 h) and salt treatment (150 mm, 1 h) by quantitative RT-PCR. Ubiquitin was used as a control.
(b) Autophosphorylation assay of bacterially expressed GST-fused AtCIPK6 and AtCIPK6T182D proteins. A Coomassie blue-stained gel after SDS–PAGE shows the amount of protein loaded.
(c) Northern blot and quantitative RT-PCR verification of AtCIPK6 expression in Arabidopsis atcipk6kd (SALK_080951) and wild-type Columbia seedlings. Ribosomal RNA and ubiquitin were used as controls in Northern blot and quantitative RT-PCR respectively.
(d) Phenotype of germinated seedlings of wild-type (A) and atcipk6kd (B).
(e) Root phenotype of 14-day-old seedlings of atcipk6kd. A close-up view on the right shows initiation of lateral root primordial in atcipk6 roots (arrows).
(f) Comparison of acropetal auxin transport from the shoot apex to the root and from the root–shoot junction to the root and basipetal root transport from the root tip for 6-day-old vertically grown wild-type Columbia and atcipk6kd seedlings. Data are presented as the percentage auxin transport relative to Columbia wild-type (= three groups of 15 seedlings each).
(g) Phenotypes of 10-day-old wild-type (bottom) and atcipk6kd (top) seedlings transferred to medium with (right) or without (left) 150 mm NaCl and grown for 6 days (representative of two experiments of approximately 40 seedlings each).

In order to genetically dissect the in vivo function of AtCIPK6, we used the only available SALK T-DNA insertion line (SALK_080951) of Arabidopsis, in which the insertion is in the AtCIPK6 3′ UTR, and designated the corresponding allele atcipk6kd (for Arabidopsis CIPK6 knockdown). We isolated a homozygous line by PCR using gene-specific and T-DNA-specific primers. The junction of AtCIPK6 and the T-DNA was amplified and sequenced to identify the insertion site as 1901 bp from transcription start site (total length 1958). No PCR product was obtained from atcipk6kd seedlings using gene-specific primers designed for full-length cDNA. Furthermore, Northern analysis with full-length cDNA as a probe and quantitative RT-PCR with primers designed for the 3′ UTR of AtCIPK6 before the T-DNA insertion showed a 50% reduction in gene expression in the mutant line compared to the wild-type (Figure 5c).

No apparent change in the seed germination time or percentage was observed in the atcipk6kd line. Germinated seedlings of atcipk6kd exhibited cotyledons fused at the base and thick hypocotyls as opposed to the open cotyledons of wild-type seedlings (Figure 5d). Mature seedlings did not show any shoot phenotype from the primary leaf onwards. A slight reduction in root length was observed in atcipk6kd seedlings 14 days after sowing; however, the length and frequency of the lateral roots were significantly reduced. Closer observation revealed initiation but delayed emergence of lateral roots in the mutant seedlings (Figure 5e). At 20 days after sowing, we observed the emergence of lateral roots almost at the same frequency as wild-type, but they are thinner and shorter than the wild-type lateral roots (Figure S4). No significant difference was observed in the cellular organization at the root tips between the wild-type and mutant plants (Figure S5), as expected from a very low expression of AtCIPK6 in the root.

A considerable decrease in shoot to root auxin transport at the apex and at the hypocotyl–root junction was observed in atcipk6kd seedlings when incubated with radioactive IAA. In contrast, atcipk6kd seedlings showed a moderate decrease in root basipetal auxin transport when compared with wild-type (Figure 5f).

atcipk6kd Seedlings showed relatively greater sensitivity to salt stress in comparison to the wild-type when 10-day-old seedlings were transferred for 6 days to medium supplemented with 150 mm NaCl (Figure 5g).

Gene expression profile of the atcipk6 knockdown mutant

To investigate any alteration in gene expression profile caused by the AtCIPK6 mutation and resulting in a phenotypic change, we used an Affymetrix GeneChip Arabidopsis ATH1 genome array (Affymetrix, containing more than 22 500 probe sets representing approximately 24 000 genes to assay changes in the transcriptome under normal growth conditions (Appendix S1). Transcript levels of wild-type Columbia and atcipk6kd seedlings at 14 days after sowing were analyzed in duplicate biological samples. Fold changes in expression were calculated using ArrayAssist software (Stratagene, As the gene expression under normal growth conditions is being compared, a minimum 1.5-fold change was set as the cut-off value for each replicate to indicate increased or decreased expression.

Using this criterion, 517 probe sets, accounting for approximately 2.3% of the total probe sets, were found to be AtCIPK6-responsive genes (some of the probe sets represent more than one gene). A total of 458 genes (other than AtCIPK6 itself) representing 436 of the AtCIPK6-responsive probe sets were found to be downregulated at a range of 1.5- to 26-fold. A total of 84 genes representing for 81 probe sets were upregulated at a range of 1.5- to 3-fold (Table S2). We used quantitative RT-PCR to analyze the expression of a number of genes that have been reported to be involved in auxin transport and abiotic stress tolerance (Figure 6a).

Among the downregulated genes coding for proteins, those that can be linked to the change in root phenotype are the genes encoding auxin efflux carrier family protein (AT2G17500), which is predicted to be membrane-bound and have auxin/hydrogen symporter activity, PIN2 (AT5G57090), which is involved in polar auxin transport, and ARF16 (AT4G30080), IAA17 (AT1G04250) and IAA20 (AT2G46990), which are involved in auxin signalling.

Downregulated genes coding for proteins related to abiotic stress tolerance include those coding for late-embryogenesis abundant protein (AT1G52690), ABC transporter family protein AtMRP4 (AT2G47800), ABA-responsive protein (AT3G02480), β-amylase (BMY1) (AT4G15210). Interestingly, CIPK20 expression is also reduced in atcipk6kd seedlings. A number of disease resistance-related genes were downregulated in the mutant seedlings, indicating a role for AtCIPK6 in the biotic stress response also.

CaCIPK6 functionally complements AtCIPK6 in Arabidopsis

To investigate whether AtCIPK6 and CaCIPK6 encode the same function, CaCIPK6 cDNA was expressed under the control of the CaMV 35S promoter in the atcipk6 knockdown line (atcipk6kd/35S:CaCIPK6). We checked the ability of CaCIPK6 to interact with AtCBL3 (NP_849449) using a yeast two-hybrid assay as for NtCBL3 (data not shown). We compared two representative independent lines expressing CaCIPK6 with a line transformed with empty vector and wild-type plants (Figure 7a). No difference was observed between germinated seedlings of wild-type and those of the mutant lines expressing CaCIPK6, in contrast to the vector-transformed line (atcipk6kd/vector), which showed the same fused cotyledon and thick hypocotyls seen in the mutant line itself (Figure 7b). There were no significant differences in the number and length of the root branches in the CaCIPK6-expressing lines compared with the wild-type at 14 days after sowing, except for the tap root length, which was approximately 15% (< 0.05) greater in the transgenic line than in the wild-type. Small but consistent (< 0.03) 20% increases over the wild-type for root basipetal auxin transport and acropetal auxin transport from the hypocotyl/root junction to the root were observed in both atcipk6 knockdown lines expressing CaCIPK6 from the 35S CaMV promoter (Figure 7d). When the 10-day-old seedlings were transferred to medium containing 150 mm NaCl, CaCIPK6-transformed mutant lines displayed more tolerance (like the wild-type plants) than vector-transformed lines (Figure 7e).

Figure 7.

 Characterization of Arabidopsis atcipk6 knockdown mutant expressing CaCIPK6.
(a) Expression of CaCIPK6 in Arabidopsis (Col) atcipk6kd lines transformed with the empty vector (pCAMBIA1305) and vector harboring CaCIPK6. RT-PCR was performed using 1 μg total RNA. A reaction to which no reverse transcriptase was added was also performed. Actin was used as an internal loading control.
(b) Phenotype of germinated Arabidopsis seedlings of wild-type (A), atcipk6kd lines expressing CaCIPK6 (B and C) and those transformed with the empty vector (D).
(c) Root phenotype of 14-day-old seedlings of wild-type (Col), atcipk6kd lines harboring CaCIPK6 and those transformed with the empty vector.
(d) Comparison of acropetal auxin transport from the root–shoot junction to the root and basipetal root transport from the root tip for 6-day-old vertically grown wild-type Columbia, atcipk6kd/35SCaCIPK6 and atcipk6kd/vector seedlings. Data are presented as the percentage auxin transport relative to Columbia wild-type (= three groups of 15 seedlings each). The results for two independent atcipk6kd:35SCaCIPK6 lines are shown.
(e) Phenotypes of 10-day-old wild-type (A) and two independent atcipk6kd lines expressing CaCIPK6 (B and C) or transformed with the empty vector (D) and transferred to medium with (right) or without (left) 150 mm NaCl and grown for 6 days.


Here we report that CIPK6, a member of the CIPK family, is a positive regulator of root development and salt tolerance. Its orthologue in chickpea also showed a similar function when expressed in tobacco. This suggests that the roles of CIPK proteins are not restricted to abiotic stress responses, and that the modes of functions of this family of kinases are conserved at least in dicot plants. Although CBL proteins show a high degree of structural homology across the species (Guo et al., 2001), we have demonstrated that CBL–CIPK binding specificity can cross the species barrier. Together, these results support the use of CBL and CIPK molecules from one plant in another.

Under normal growth conditions, the basal function of CIPK6 appears to be restricted to within shoots; root-specific expression was very low compared to that in stem and leaf. Although its expression in the root drastically increases during stress, a question remains as to why there was no altered shoot phenotype in overexpressing tobacco lines, especially in view of the shoot phenotype in the Arabidopsis mutant. One explanation might be that the extent of CIPK6 activity required for its shoot-specific function is not limited, and hence overexpression of the active mutant gene does not cause any extra effect. In vitro kinase activity assays showed that the substrate phosphorylation activity of SOS2 (CIPK24) was negligible in the absence of SOS3 (CBL4), but the SOS2–SOS3 complex posses a basal level of activity for substrate phosphorylation even in the absence of Ca2+ (Halfter et al., 2000). We hypothesize that CIPK6 has a basal level of activity in the presence of physiological concentrations of cytosolic Ca2+, and that this activity is sufficient for its shoot-specific function. However, because of low root-specific expression, its activity in root is limiting, and therefore overexpression of CaCIPK6 produced a developed root phenotype. The slight increase in root length in CaCIPK6-expressing tobacco and atcipk6kd seedlings may be explained by the basal level of CIPK activity. It is notable that constitutively active CIPK mutants are active even in absence of Ca2+ (Guo et al., 2001).

Enhancement of salt and dehydration tolerance by CaCIPK6T181D overexpression may be explained by increased expression of stress-response genes such as AtMRP4 homologue, dehydrin, inositol-3-phosphate synthase etc. in the overexpressing plants. The protein kinase activity of tobacco CIPK6 may be too limiting to provide visual evidence for significant salt and dehydration tolerance under these experimental conditions. Even overexpression of CaCIPK6 did not produce considerable improvement in tolerance against those stresses. Similar results were obtained for SOS2 (AtCIPK24) in Arabidopsis (Guo et al., 2004).

The available SALK line for AtCIPK6 has a T-DNA insertion in the 3′ UTR of the gene, causing a reduction in the steady-state mRNA level. Changes in phenotype even at 50% expression support the concept that the amount/activity of CIPK6 is limiting in Arabidopsis. The shoot phenotype of the Arabidopsis mutant appears to be stage-specific, as we did not notice any abnormality in the shoot after the emergence of primary leaves. The mature mutant plants flowered and produced seeds as the wild-type plants. Mutant seedlings at early stages exhibited initiation of lateral root primordia, but those were fewer in number and their emergence was delayed. Initiation of lateral root primordia at the early stage of seedling development depends on basipetal transport of the high concentration of auxin localized at the root tip, while shoot-to-root auxin transport controls the emergence of lateral roots. In later stages, the root synthesizes a considerable amount of auxin, thereby reducing its dependency on shoot-derived auxin (Bhalerao et al., 2002). The delay in lateral root emergence in the mutant seedlings is most likely due to the reduction in shoot-to-root auxin transport that appears to be the primary effect of the AtCIPK6 mutation given the very low root-specific and high shoot-specific expression of the gene. A reduction in shoot-to-root auxin transport should cause an increase in the auxin pool in the shoot, and is probably the reason for the shoot phenotype.

Kinases function primarily by phosphorylating their substrates. Therefore, CIPK6 function is not restricted to alteration of gene expression as shown in the overexpressing and knockdown mutant plants. Our results suggest transcriptional downregulation of genes involved in auxin transport and signaling as the possible reason for reduction in auxin transport and the corresponding phenotype of the Arabidopsis mutant. The involvement of CIPK proteins in the abiotic stress response is indicated by decreased or increased expression of the downstream genes in the mutant and overexpressing plants, respectively. Interestingly, the lists of downregulated genes in mutant Arabidopsis and upregulated ESTs in overexpressing tobacco include a number of genes encoding, e.g. MRP4, protein phosphatase 2C, late embryogenesis abudant protein (LEA), GH3 family protein, glycine-rich protein, RAP2.4, S-adenosyl methionine synthase, senescence-associated protein, etc.

The altered auxin transport and developmental phenotype, in addition to the enhanced abiotic stress tolerance associated with CIPK6, suggest that it plays a role in auxin–ABA cross-talk. There are several lines of evidence demonstrating that a number of molecules function in both the pathways. Expression of AtNAC2 and the proline biosynthetic gene AtP5CS1, catalyzing the rate-limiting step of proline biosynthesis, is regulated by ABA as well as by auxin (Strizhov et al., 1997; He et al., 2005). Expression of the transcription factor ABI3 that regulates ABA responsiveness is regulated by auxin, and plants with a mutation in the ABI3 gene show defects in lateral root development (Brady et al., 2003). ABI3 shows dual affinity in DNA-binding assays towards ABA- and auxin-responsive elements (ABRE, AuxRE) (Nag et al., 2005). The CIPK6 overexpression and knockdown phenotypes match in a number of ways the overexpression and knockout phenotypes of vacuolar pyrophosphatase AVP1 (Li et al., 2005). Overexpression of both the molecules promoted lateral root development through facilitation of auxin fluxes and enhanced osmotic tolerance (Park et al., 2005). Reduction in expression of both the molecules caused alteration in auxin transport and disrupted growth. To explain the function of AVP1 in regulating auxin transport-mediated root growth, it was proposed that a mutation in AVP1 causes reduction in plasma membrane-localized proton-ATPase (P-H+-ATPase) expression and activity, and thus a change in apoplastic pH (Li et al., 2005). Enough evidence exists to connect the low apoplastic pH created by P-H+-ATPase and initiation of cell wall loosening and extension growth (Hager, 2003). The altered apoplastic pH can affect the intracellular proton gradient and thus the driving force for auxin transport. Microarray data show that there is a reduction in expression of a P-H+-ATPase (AT3G60330) in atcipk6kd (Table S2). In addition, the reduction in expression levels of PIN2 and auxin efflux carrier protein (AT2G17500) may contribute to the altered auxin transport and thus to the phenotype in the atcipk6kd mutant. Reductions in the expression of a calcium-transporting ATPase (AT3G22910) and a cation exchanger CAX3(AT3G51860) most likely contributed to the salt and osmotic sensitivity of the mutant line.

Altogether, our results provide evidence that overexpression of CaCIPK6 enhances auxin sensitivity and basipetal auxin transport in the root, most likely causing increased cell division and overall root growth. A mutated Arabidopsis line with reduced CIPK6 expression exhibited reduced lateral root production at an early stage of development, with a decrease in basipetal root and acropetal shoot-to-root auxin transport. Growth of overexpressing tobacco and mutated Arabidopsis plants in salt-supplemented medium highlighted a role for CIPK6 in abiotic stress tolerance, which is substantiated by gene expression analysis. The data presented here suggest that, in addition to its role in abiotic stress tolerance like other CIPKs studied so far, CIPK6 also functions in auxin transport and root development.

Experimental procedures

Plant materials and treatments

Chickpea (Cicer arietinum) cv. BGD72 was used in this study. Seedlings were grown and subjected to dehydration and abscisic acid treatments as described by Boominathan et al. (2004). For transgenic lines, N.icotiana tabacum var. Xanthi was used as the wild-type. Agrobacterium-mediated transformation of tobacco explants was performed as described previously (Gelvin and Schilperoort, 1994). Integration of the transgene was tested by PCR amplification and Southern analyses of the genomic DNA of antibiotic-selected seedlings. Expression of the transgene was evaluated by reverse transcription followed by PCR amplification using gene-specific primers of the total RNA extracted from leaves. Arabidopsis thaliana plants (ecotype Columbia) were grown on soil in growth chambers with 80 μmol μm−2 sec−1 light under a 16 h light/8 h dark cycle at 21°C. T-DNA insertion lines were screened for homozygosity using methods and primers recommended by the TAIR website (

Expression, purification and in vitro phosphorylation assay of recombinant GST fusion proteins

The coding regions of CaCIPK6, CaCIPK6T181D, AtCIPK6 and AtCIPK6T182D cDNAs were cloned into pGEX-4T2 (Amersham Pharmacia Biotech; Wild-type and mutated GST fusion proteins were transformed into Escherichia coli BL21(DE3)-CodonPlus cells (Stratagene; Recombinant proteins were purified from the bacterial lysates by glutathione–Ssepharose affinity chromatography as described by Gong et al. (2002c). In vitro protein phosphorylation assays were performed in a 40 μl reaction mixture comprising purified recombinant GST proteins incubated in kinase buffer (10 μCi [γ32P]ATP, 20 mm Tris/HCl pH 8.0, 5 mm MnCl2, 1 mm CaCl2, 0.1 mm EDTA and 1 mm DTT) for 30 min at 30°C.

Analyses of transgenic and mutated plants for abiotic stress

To study the morphology of transgenic plants harboring different constructs, T1 seeds were grown vertically on MSH (half-strength MS salts + 1.5% sucrose + 0.8% agar) for 6 days, and root lengths were measured. For salt stress, seedlings grown for 8 days after germination on MSH were transferred to MSH medium supplemented with 250 mm NaCl for 8 days. The seedlings were then washed briefly with water and allowed to grow on MSH for a further 8 days. For dehydration stress treatments, seeds that had germinated on MSH medium containing 0.3 m mannitol, as well as seedlings grown for 8 days after germination on MSH, were transferred on MSH medium containing 0.4 m mannitol and grown for 20 days. For salt treatment of Arabidopsis plants, 10-day-old seedlings were transferred to half-strength MS medium containing 150 mm NaCl for 6 days.

Auxin transport assay

Auxin transport was measured essentially as previously described (Shin et al., 2005). Six-day-old vertically grown seedlings were used for the experiment. Briefly, agar blocks of 1 mm diameter containing 7.7 × 10−8 m3H-IAA (Amersham) were applied at various positions (root tip, shoot apex and hypocotyl/root junction) in three independent experiments. After incubation for 3 h, a 0.5 mm section of the root close to the agar block was dissected and discarded. Two consecutive 2 mm segments below the incision line were then collected separately, and the segments from 15 roots were pooled and placed into glass scintillation vials containing 5 ml scintillation fluid. The radioactivity in these two pools of root segments was measured using a Beckman Coulter LS6500 scintillation counter (Beckman Coulter, The amount of radioactivity was the mean of three separate experiments ±SD. Student’s t test with a paired two-tailed distribution was used for statistical analysis.


The project was funded by the Department of Biotechnology, Government of India, and by a seed grant from the National Institute for Plant Genome Research. V.T. acknowledges support from the Council for Scientific and Industrial Research, and B.P. acknowledges research fellowships from the University Grants Commission.