Regulatory Roles of Cytokinins and Cytokinin Signaling in Response to Potassium Deficiency in Arabidopsis

Potassium (K) is an important plant macronutrient that has various functions throughout the whole plant over its entire life span. Cytokinins (CKs) are known to regulate macronutrient homeostasis by controlling the expression of nitrate, phosphate and sulfate transporters. Although several studies have described how CKs signal deficiencies for some macronutrients, the roles of CKs in K signaling are poorly understood. CK content has been shown to decrease under K-starved conditions. Specifically, a CK-deficient mutant was more tolerant to low K than wild-type; however, a plant with an overaccumulation of CKs was more sensitive to low K. These results suggest that K deprivation alters CK metabolism, leading to a decrease in CK content. To investigate this phenomenon further, several Arabidopsis lines, including a CK-deficient mutant and CK receptor mutants, were analyzed in low K conditions using molecular, genetic and biochemical approaches. ROS accumulation and root hair growth in low K were also influenced by CKs. CK receptor mutants lost the responsiveness to K-deficient signaling, including ROS accumulation and root hair growth, but the CK-deficient mutant accumulated more ROS and exhibited up-regulated expression of HAK5, which is a high-affinity K uptake transporter gene that is rapidly induced by low K stress in ROS- and ethylene-dependent manner in response to low K. From these results, we conclude that a reduction in CK levels subsequently allows fast and effective stimulation of low K-induced ROS accumulation, root hair growth and HAK5 expression, leading to plant adaptation to low K conditions.


Introduction
Plants are able to grow under various nutritional environments by adapting to the conditions in which they live. If nutrients are scarce, plants regulate their metabolism through various signaling pathways in order to survive. Nutrient sensing and signaling are active throughout a plant's life span and are important for optimal plant growth. When nutrients are limiting, plants grow at a slower rate, change their nutrient utilization and acquisition, and adjust their metabolism and morphology in order to more effectively acquire the nutrients [1,2]. In an agricultural system, a balanced supply of soil macronutrients, especially nitrogen, phosphorus, and potassium (K), is necessary to produce the optimum quantity and quality of crops [3]. Within the plant, K is the most abundant inorganic cation, consisting of up to 1/10 of a plant's dry weight [4]. Potassium plays various roles in the plant, such as the activation of enzymes, stabilization of protein synthesis, neutralization of negative charges on proteins, maintenance of cytoplasmic pH homeostasis [5] and osmotic balance, and the movement of other ions [3]. Potassium deprivation rapidly induces the expression of two K transporters, HAK5, a high-affinity K uptake transporter and KEA5 in 6-week-old roots [6,7], whose expression is regulated by reactive oxygen species (ROS) [7,8]. However, while HAK5 expression is induced at any developmental stages of roots, KEA5 expression is not, making HAK5 a preferable marker gene in studies of low K responses.
The relationship between the acquisition of different nutrients by mineral nutrient transporters and the imbalances triggered by a mineral deficiency are well documented [5]. For instance, nitrate transporters are down-regulated when a plant is deprived of K [9]; several nutrient transporters are up-regulated by K and phosphorus deprivation in tomato roots [10]; and when plants experience K, nitrogen, phosphorus, and sulfur deprivation, they produce ROS in roots [2,7,8]. Furthermore, the correlation between phytohormone signaling and nutrient signaling is well known. The K transporter TRH1 is required for root hair development and root gravitropism and functions in the auxin transporter system in Arabidopsis roots [11]. The genes involved in auxin biosynthesis were down-regulated by K re-supply in K-starved roots [9]. In addition, an Arabidopsis transcription factor, MYB77, has been shown to modulate the low K-dependent reduction of the lateral root density through auxin signal transduction [12]. Ethylene is involved in the low K signaling pathway by inducing the production of ROS in roots and then changing root hair and primary root growth and up-regulating HAK5 expression in Arabidopsis [13]. Moreover, many genes respond to K starvation, which leads to increased pathogen susceptibility; a process that is linked to jasmonic acid [9].
The cytokinins (CKs) regulate various processes within plants, including cell division and root and shoot morphogenesis. In Arabidopsis, the key CK biosynthetic enzymes are adenosine phosphate-isopentenyltransferases (IPTs) [14]. There are two classes of IPTs in Arabidopsis. ATP/ADP IPTs are involved in the synthesis of N 6 -(D 2 -isopentenyl)adenine (iP)-and trans-zeatin (tZ)type CKs, whereas tRNA IPTs are responsible for the biosynthesis of cis-zeatin (cZ)-type CKs [14]. Additionally, it was suggested that the iP-and tZ-type CKs are the major forms and are more physiologically active than cZ-type CKs in Arabidopsis [15]. To exert their biological functions, CK signaling is mediated by a multi-step phosphorelay that consists of CK receptor histidine kinases (AHKs), phosphotransfer proteins (AHPs) and response regulators (ARRs). The AHKs respond to CKs by autophosphorylation and transfer of a phosphoryl group to the ARRs through the AHPs, resulting in the activation of downstream proteins [16]. Among the 8 AHKs, AHK2, AHK3 and AHK4 are implicated in CK signaling [16,17].
It is fairly well known that interactions between nutrients and CKs influence nutrient signaling and adaptive responses in plants. Nitrate treatment induces the biosynthesis of CKs by up-regulating IPT3 [18] and also triggers the expression of type-A ARRs in Arabidopsis [19]. CKs are also linked systemically to phosphate deprivation signaling by repressing the expression of genes that are induced by phosphate starvation conditions [20]. Through characterization of plants carrying mutations in the receptor kinases AHK3 and AHK4, it was revealed that these kinase encoding genes contribute to the repression of phosphatestarvation-responsive genes [21]. In addition, CKs were found to exert a negative effect on expression of SULTR1;1 and SULTR1;2, resulting in a reduction of sulfate uptake in roots [22]. AHK3 and AHK4 are also involved in the root iron uptake machinery in Arabidopsis by negatively regulating the expression of genes which are induced by iron deficiency [23]. Taken together, these studies demonstrate that CKs play a role in the response to the limitations of various nutrients in plants. However, the roles of CKs in low K signaling are still unclear at the present time.
Here, we show that CK receptor mutants lose their responsiveness to low K signaling through the measurement of ROS accumulation and root growth under low K conditions. Additionally, we found that CKs affected the induction of HAK5 expression and function under low K conditions. Finally, we provide evidence that CKs negatively regulate low K response.

CK Measurements
For measuring CK content, four-day-old seedlings were transferred to +K or 2K LSM, and then the roots and shoots from Arabidopsis grown in either +K or 2K conditions for 1, 3 or 7 days were harvested. More than 6 replicates per condition were analyzed. Extraction and determination of CKs were performed as previously described [25]. Statistical differences were evaluated with a t-test using the Graphpad Prism 5.01 software program.

Root Assay
All seeds were planted on normal LSM and vernalized at 4uC for 3 days. Four-day-old seedlings were transferred to +K or 2K medium for root assays. The plates were scanned at day 7 th and day 12 th and the images were analyzed using the Image J program. The primary root length and the number of lateral roots of more than 30 plants per treatment were measured. For the root hair analysis, two-day-old seedlings were transferred to and grown in K-sufficient and K-deficient liquid media without sucrose for 24 h. Root images were taken on a LEICA M165 FC microscope equipped with a LEICA DFC310FX camera using the Leica application suite version 3.4.1 software program. Root hair numbers were counted in a 3 mm region from the differentiation zone. To measure the length of root hairs, the longest root hairs (n = 15) in a 3 mm region from the end of the root were measured on more than 10 plants using Image J. All experiments were performed more than three times and the data presented are a representative set. Statistical differences were evaluated with One-way ANOVA and Tukey's multiple comparison test by the Graphpad Prism 5.01 software program.

ROS Detection Assay
To observe ROS in Arabidopsis roots, two-day-old seedlings were floated and grown for one day in K-sufficient and K-deficient liquid medium with or without 10 nM t-zeatin and without sucrose. The seedlings were treated with 20 mM of DFFDA (Invitrogen) as described previously [13]. The images of roots were taken with a LEICA M165 FC microscope equipped with a LEICA DFC310FX camera and the Leica application suite version 3.4.1 software. For ROS quantification, each image was selected within a 0.5 mm region from the starting point of the root hair differentiation zone (RHDZ). Green pixel intensity from the histogram was analyzed by the Adobe Photoshop program. The same microscopy parameters, such as exposure time, gain and contrast, were used for all repeated experiments. All ROS detection analyses were repeated more than five times. More than 15 seedlings were used for each experiment and statistical analyses were performed using the Graphpad Prism 5.01 software program. One representative set of data is shown.

Quantitative Real-time PCR
Four-day-old seedlings were transferred and grown on Ksufficient and K-deficient media for 7 days. Roots of the seedlings were harvested in four biological samples (3 seedlings were used for each sample) for total RNA extraction using Trizol-reagent following the manufacturer's instructions (Invitrogen). After treating with DNase I (Invitrogen), RNA was checked for genomic DNA contamination using PCR analysis. RNA (2 mg) was used as the template for reverse transcription using Superscript III (Invitrogen). Quantitative real-time PCR was performed using the THUNDERBIRD SYBR qPCR mix (Toyobo) and the MX3000P PCR machine (Agilent Technologies). An actin gene (AT3G18780) was used as a control and amplified with sense (59-CTGGATCGGTGGTTCCATTC-39) and antisense (59-CCTGGACCTGCCTCATCATAC-39) primers. For analyzing HAK5 gene (AT4G13420) expression, a sense primer (59-CGAGACGGACAAAGAAGAGGAACC-39) and antisense primer (59-CACGACCCTTCCCGACCTAATCT-39) were used. MX3000P version 4 (Agilent Technologies) was used for calculating and analyzing the threshold cycle (Ct) values. The fold change of target transcript compared with the K-sufficientgrown wild-type (WT) control was calculated by normalization with the Actin Ct values. The expression level of Actin was not affected by K starvation. Statistical differences in target gene transcripts were evaluated by One-way ANOVA using DDCt values [26].

Low K Status in planta Results in Reduction of CK Content
To investigate whether CK metabolism is affected by low K status, the content of CKs in K-deficient and K-sufficient-grown Arabidopsis WT Col-0 shoots and roots were analyzed in a timecourse manner ( Figure 1 and Table S1). Most samples contained undetectable amounts of the dihydrozeatin-type (DZ) CKs. The contents of cis-zeatin (cZ), cis-zeatin riboside (cZR), and cis-zeantin riboside phosphates (cZRPs) were not significantly altered by low K (Table S1). Among the bioactive CKs, the levels of the tZ-type ( Figure 1A) and iP-type CKs ( Figure 1B) clearly decreased in the roots of plants grown on media with lower K for three or seven days. In shoot tissues, tZ-type CKs were mainly reduced in plants grown on media with lower K for one or three days ( Figure 1A); and iP-type CKs were reduced in plants grown on lower K media for one or seven days ( Figure 1B). These results demonstrate that CK content is negatively affected by low K conditions.
The CK-deficient ipt1,3,5,7 Mutant Enhanced Root Growth while the CK-overaccumulating IPT3-ox Suppressed Root Growth Under Low K Conditions The ipt1,3,5,7 line carries mutations in ATP/ADP isopentenyltransferases, leading to markedly reduced bioactive endogenous tZ-type and iP-type CKs [14]. On the other hand, the transgenic plant overexpressing IPT3 (IPT3-ox) highly accumulates CKs in relative comparison to WT plants [24]. Primary root growth and lateral root numbers were analyzed in ipt1,3,5,7 and IPT3-ox plants under both K-sufficient (1.75 mM KCl) and K-deficient (10 mM KCl) conditions in order to understand whether low K signaling is affected by the level of endogenous CKs. To optimize the nutrient composition for Arabidopsis and to modify K content, LSM was used for all experiments [7,8,13]. All seedlings were germinated on full nutrient LSM and at four days were transferred to medium containing 1.75 mM KCl (+K) or 10 mM KCl (-K). When ipt1,3,5,7, IPT3-ox and WT plants were grown under Kdeficient conditions for 7 days and compared with K-sufficientgrown plants, the primary root length of IPT3-ox plants (32% reduction) showed 15% more reduction compared to WT (17% reduction) but ipt1,3,5,7 roots showed no significant changes ( Figure 2A). Twelve-day-old K-deficient-grown WT showed a 36% reduction of primary root growth compared to Ksufficient-grown WT. On the other hand, IPT3-ox plants showed a 54% reduction and ipt1,3,5,7 only showed a 20% decrease in root growth ( Figure S1B). Since ipt1,3,5,7 is not a complete CK-null mutant, even though there was no difference in root growth for 7 day K-sufficient-grown and Kdeficient-grown ipt1,3,5,7, longer K-deficient treatment (12 days) showed mild reduction in root growth ( Figure S1B). The discrepancy between the data observed with 7-and 12-day-low K treated plants might be attributed to the age-dependent expression of the ATP/ADP IPT6 encoding gene. This gene weakly expresses in young seedlings, and its expression is increased with the aging of the plants [27]. In addition, Kdeficient-grown ipt1,3,5,7 had a 38% reduction in lateral root numbers compared to K-sufficient-grown ipt1,3,5,7. However, WT had a 61% decrease in lateral root numbers responding to K deficiency and IPT3-ox had a 68% decrease ( Figure 2B). These results suggest that CKs negatively influence root growth rate and an increase in endogenous CK content results in reduced tolerance to K deficiency.

Loss of Root Growth Response to K Deprivation in ahk Mutants
To further understand the role of CK signaling under low K conditions, primary root length and lateral root numbers of the WT, ahk single mutants (ahk2, ahk3 and ahk4) and ahk double mutants (ahk2ahk3, ahk2ahk4 and ahk3ahk4) were analyzed. Statistically significant differences of P,0.05 among WT ahk single and double mutants were determined using ANOVA; and significances were corrected post hoc using Tukey's HSD comparisons. Unlike WT plants, which exhibited decreased primary root length under K-starved conditions, the primary root growth of both ahk single (except ahk4) and ahk double mutants was not affected by the 2K conditions ( Figure 3A). While lateral root numbers in WT and ahk single mutants exhibited similar responses to low K conditions ( Figure 3B), the ahk double mutants, especially ahk2ahk3, showed a reduction in responsiveness of lateral root growth under the same conditions ( Figure 3B). These results suggested that the repression of primary root growth, and to some extent the lateral root growth, by K starvation was mediated by CK signaling, especially through AHK2 and AHK3.

CKs Function in Low K-dependent ROS Accumulation
Previous studies have demonstrated that low K conditions induce ROS accumulation in Arabidopsis roots [7,13]. To examine whether CK signaling is involved in the accumulation of ROS under low K conditions, ROS accumulation in RHDZ of WT, ahk2ahk3 and ipt1,3,5,7 was analyzed under +K and 2K conditions using the membrane-permeable fluorescent dye 5-(and 6-) carboxyl-29,79-difluorodihydrofluorescein diacetate (DFFDA; Figure 4A). After obtaining images, the signal intensity within a region 0.5 mm from the starting point of the RHDZ was calculated ( Figure 4B). Under K-sufficient conditions, there was greater accumulation of ROS in the RHDZ of ahk2ahk3 relative to WT. Similar to what we observed with primary root growth, ROS accumulation was not increased in the ahk2ahk3 mutant under K-deficient conditions. This finding suggests that AHK2-and AHK3-dependent CK signaling is required for low K-dependent ROS accumulation (Figure 4). Similar to the result of ahk2ahk3, higher ROS levels were detected in ipt1,3,5,7 and IPT3-ox RHDZ than that of WT under +K conditions. We did not observe significant change in ROS level in K-deficient grown IPT3-ox plants. However, a significant increase in ROS level was noted for K-deficient grown ipt1,3,5,7 (Figure 4). The enhanced ROS production under K deficiency conditions in plants with low level of CKs supports the hypothesis that low CK levels are associated with enhanced low K stress tolerance, which is also consistent with the observed reduction of CK content under K-deficient conditions (Figure 1).

CKs Influence Root Hair Development Under K-deficient and K-sufficient Conditions
ROS is known to be an essential signal for root hair elongation [28]. Induction of root hair elongation by low K requires ethylenedependent ROS accumulation [13]. In order to determine whether CKs exert influence on the low K-dependent induction of root hair development, root hair growth in the WT, ahk2ahk3, ipt1,3,5,7 and IPT3-ox plants was analyzed ( Figure 5). As previously reported, the root hairs of K-deficient WT plants were much longer than those of K-sufficient WT plants ( Figure 5) [13]. In ahk2ahk3 and IPT3-ox, root hair length was longer than that in WT under K-sufficient conditions, but the induction degree of root hair length in the ahk2ahk3 (17% increase) and the IPT3-ox (no significant change) by low K treatment was much lower than in WT (60%) ( Figure 5A). K-sufficient-grown ipt1,3,5,7 seedlings have much shorter root hairs than WT and ahk2ahk3, but Kdeficiency led to a dramatic induction in root hair growth of ipt1,3,5,7 in comparison with WT ( Figure 5A). Similar results were obtained from analyses of root hair numbers in the RHDZ responding to low K conditions ( Figure 5B). K-sufficient-grown ipt1,3,5,7 had fewer root hairs than WT, ahk2ahk3 and IPT3-ox but K-deficient-grown ipt1,3,5,7 dramatically increased root hair numbers. However, the numbers of root hairs in ahk2ahk3 and IPT3-ox were not affected by K status ( Figure 5B). These data indicated that the degree of CK reduction was important for the induction of root hair elongation to cope with low K conditions. Similar to ROS accumulation, the induction of root hair growth by K deprivation is also negatively correlated with endogenous CK content and modulated via AHK2/AHK3-meditated CK signaling.

CKs Alter K-deficient Induced HAK5 Expression
A high-affinity Arabidopsis K transporter, HAK5, is one of the key proteins functioning in low K signaling and is transcriptionally regulated by K limitation [29]. To investigate whether CKs regulate low K-induced gene expression, HAK5 expression was analyzed by real-time PCR in the CK receptor mutant, ahk2ahk3, the CK-overaccumulating IPT3-ox line, and the CK-deficient ipt1,3,5,7 mutant under K-sufficient and K-deficient conditions (Table 1). Under K-sufficient conditions, the expression level of HAK5 was lower in the ahk2ahk3 mutant and remarkably higher in IPT3-ox than in WT plants. However, HAK5 expression under insufficient K remained unchanged in the ahk2ahk3 mutant. Interestingly, the induction of HAK5 expression by K deficiency was greatly suppressed in IPT3-ox but highly activated in ipt1,3,5,7 compared to WT (Table 1). These results indicate that the expression of HAK5 under low K conditions is regulated by both CK-dependent and CK-independent mechanisms and CKs negatively regulate HAK5 gene expression in response to K starvation.

Discussion
In this report, we describe the functional analyses of CKs and CK-related signaling in response to K deficiency by investigating the consequences of altered CK contents and the suppression of CK signaling. Results from both gain-and loss-of-function studies suggest that CKs may function as negative regulators in response to low K conditions (Figure 2 and 3). CK content was decreased in low-K-grown roots and shoots ( Figure 1). In addition, the induction level of the HAK5 gene by low K was decreased in IPT3-ox plants (Table 1). Consistent with this result, the expression of HAK5 was more highly induced by low K conditions in the CK-deficient ipt1,3,5,7 mutant as compared to WT (Table 1). Moreover, the growth of both primary and lateral roots in IPT3-ox plants under low K conditions was more highly suppressed than in WT ( Figure 2). Collectively, these data strongly indicate the negative regulatory roles of CKs metabolism in primary and lateral root growth responding to K deficiency.
Previous studies have shown that multiple phytohormones regulate low-K signaling, that can lead to effects on gene expression, reduced primary root growth, reduced lateral root growth, and increased root hair growth [7,9,12,13,30]. One typical phenotype of low-K-grown Arabidopsis plants is the reduction of lateral root numbers. Auxin is known to be a positive regulator of lateral root development [31]. Low K-dependent reductions of lateral root development are known to be regulated by auxin via reductions of auxin levels and transport. Independent from primary and lateral root growth, ethylene has been reported to act as a positive regulator of low K-dependent ROS accumulation, HAK5 expression and the induction of root hair growth [13]. Low K also leads to a decrease in primary root growth [29] that may be regulated by ABA [30]. Jasmonate was also shown to regulate low K-dependent gene expression [9]. In this study, we found that CKs function differently from other hormones by acting as negative regulators in low K-dependent HAK5 expression (Table 1), primary root growth (Figure 2A and 3A) and root hair growth ( Figure 5). These data suggest that CKs might function in parallel with ethylene but in an antagonistic behavior in low K signaling.
CKs are known to exert influence on the acquisition of several macronutrients, such as nitrogen, phosphorus and sulfur. Specifically, the expression of genes encoding multiple macronutrient transporters, including nitrate transporters, sulfate transporters and phosphate transporters, were decreased by CKs [32]. In the case of nitrogen, CKs and nitrogen are reciprocally influenced. CK content was tightly regulated by nitrogen supply. Higher nitrate-grown Arabidopsis had higher CK levels than low nitrategrown Arabidopsis. In addition, CKs act as long distance root-to- shoot signals and local signals for nitrate sensing [33]. CKs could negatively regulate nitrogen uptake via the control of nitrate and ammonium transporter gene expression [32,34]. Other phosphate and sulfate transporters were regulated similar to nitrate transporters [20,22]. In our study, we also showed that CKs negatively regulate the gene expression of the high-affinity K transporter HAK5 (Table 1). Moreover, the levels of bioactive CKs were reduced in both roots and shoots with the most drastic reduction observed in roots after 3 days of K deprivation ( Figure 1A). Collectively, our results support that CKs function as negative regulators of HAK5 gene expression; a regulation that is similar to that of other macronutrient transporters.
In this study, we have also demonstrated that CKs control the response to low K conditions through CK signaling by functional analyses of the ahk mutants in response to K deficiency. The results of root growth assays indicated that among the three CK receptor kinases, AHK2 and AHK3 play major roles in the regulation of the response to K deficiency (Figure 3). The weak correlation between AHK4 and low K signaling may be explained by its dual activity. In the presence of CKs, AHK4 possesses kinase activity and phosphorylates AHPs; however, in the absence of CKs, AHK4 acts as a phosphatase that dephosphorylates AHPs [35]. This finding differs from the regulation of other macronutrients by CKs. AHK3-and/or AHK4-dependent CK signaling was proposed to have dominant roles in the function of nitrate, phosphate and sulfate transporters [16,21,22,23,36]. These data suggest that there might be some specificity of CK signaling to each macronutrient signaling pathway and that AHK2 and AHK3 might have major roles in low K signaling.
As a common response to K deficiency, ROS is induced in roots, leading to root hair elongation [2,7,8,13]. The investigation of ROS induction in ahk2ahk3 roots further supports the observation that CK signaling is involved in the response to low K. Results shown in Figure 4 indicated that the ROS accumulation was not altered in the ahk2ahk3 mutant by K availability, whereas a significant difference was observed in ROS accumula- tion in WT either with or without K. Similar to low K-dependent primary root growth, we only observed a slight responsiveness of root hair elongation to low K in the ahk2ahk3 mutant ( Figure 5). Consistent with these data, several ethylene insensitive mutants, including ein2-1, etr1-1, and etr1-3, showed a smaller increase of ROS in low K-grown roots. However, these ethylene insensitive mutants still show the low K-dependent ROS accumulation in roots [13], suggesting that other ethylene receptors or unidentified  Figure 4A. Pixel intensity of the roots was measured from the root hair differentiation zone to 0.5 mm. Different letters indicate significant differences from each other as determined using ANOVA (P,0.05) and significances were corrected post hoc using Tukey's HSD comparisons. (n.20). doi:10.1371/journal.pone.0047797.g004 Figure 5. Root hair growth analysis in WT, ahk2ahk3 and ipt1,3,5,7 plants. Root hair length (A) and number of root hairs (B) in the seedlings were analyzed. The number of root hairs was counted in a 3 mm region from the starting point of the RHDZ. In order to measure root hair length, the longest root hairs (n = 8) per seedling (n.10) were chosen and measured. Different letters indicate significant differences from each other as determined using ANOVA (P,0.05) and significances were corrected post hoc using Tukey's HSD comparisons. doi:10.1371/journal.pone.0047797.g005 ethylene receptors regulate low K signaling. Our data indicates that AHK2 and AHK3 are the major CK receptors for low Kdependent ROS accumulation.
In ipt1,3,5,7 mutant, shorter and fewer root hairs were found under K-sufficient conditions compared to WT. This phenomenon may be due to the lower endogenous CKs content and is not related to K responsiveness. However, the increase in root hair ratio of ipt1,3,5,7 in response to low K is more than 10 times higher than that of WT ( Figure 5). Reduction ratio of CKs responding to low K compared to +K seems to be more important than the absolute value of CKs for root hair growth. In addition, the ipt1,3,5,7 has lower active CK levels in +K conditions. Even though actual decreased CK content responding to low K may be lower in ipt1,3,5,7, plants showed more sensitive response to low K. Because ipt1,3,5,7 has lower CK level, the threshold value for root hair growth responding to low K may be much lower than WT. These data suggest that the CK deficient ipt1,3,5,7 has lower response threshold level to K deficiency-induced root hair growth.
It is well known that CKs and ABA have antagonistic effects on responses to a number of stresses, including drought and high salinity [37,38,39,40,41]. Our data, together with previously published results, demonstrate that CKs and ABA also have opposite effects on the response to low K conditions as well [30]. In accordance with their function, CK levels were decreased in WT plants under K-deficient conditions (Figure 1), whereas ABA levels increased [30].
Taken together, we propose a model in which plants adapt to K deprivation by triggering an efficient regulation of various hormonal biosynthesis and signal cascades, including CKs and CK signaling ( Figure 6) [7,12,13,30]. Specifically, for CKs and CK signaling, we suggest that during K deprivation, CK metabolism is altered, leading to a decrease in CKs. A reduction in CK levels subsequently allows fast and effective stimulation of low K-induced ROS accumulation, root hair growth and HAK5 expression, leading to plant adaptation to low K conditions.

Supporting Information
Figure S1 Root growth assay of WT, IPT3-ox, ipt1,3,5,7 and ahk2ahk3 under +K and -K conditions. (A) WT, IPT3ox, ipt1,3,5,7 and ahk2ahk3 grown under +K and 2K conditions for 7 days. (B) Root growth assay of WT, IPT3-ox and ipt1,3,5,7 plants under +K and 2K conditions for 12 days. Plants were grown under +K conditions for 4 days and then transferred and grown on +K or 2K medium for 12 days. Length of primary roots were analyzed (n.30). Significant differences were represented by different letters on the bars (P,0.05; t-test).