SEARCH

SEARCH BY CITATION

Keywords:

  • Arabidopsis thaliana;
  • cytokinin;
  • cytokinin receptor;
  • plant development;
  • signal specificity

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References

Arabidopsis thaliana has three membrane-located cytokinin receptors (AHK2, AHK3 and CRE1/AHK4), which are sensor histidine kinases containing a ligand-binding CHASE domain. Despite their structural similarity the role of these receptors differs in planta. Here we have explored which parameters contribute to signal specification. In a bacterial assay, the CHASE domain of AHK2 has a similar ligand binding spectrum as CRE1/AHK4. It shows the highest affinity for isopentenyladenine (iP) and trans-zeatin (tZ) with an apparent KD of 1.4 and 4.0 nm, respectively. Real-time PCR analysis of cytokinin primary response genes in double mutants retaining only single receptors revealed that all receptors are activated in planta by cytokinin concentrations in the low nanomolar range. However, there are differences in sensitivity towards the principal cytokinins iP and tZ. The activation of the cytokinin-sensitive PARR5:GUS reporter gene in three different double mutants shows specific, but also overlapping, spatial domains of activity, which were for all receptors predominantly in the shoot apical meristems and root cap columella. AHK2 and AHK3 signal specifically in leaf parenchyma cells, AHK3 in stomata cells, and CRE1/AHK4 in the root vasculature. Promoter-swap experiments demonstrate that CRE1/AHK4 can functionally replace AHK2 but not AHK3. However, the cytoplasmic AHK3 histidine kinase (Hk) domain can be replaced by the CRE1/AHK4 Hk domain, which suggests that functionality is mediated in this case by the extracytosolic domain. Together, the data show that both differential gene expression and ligand preference contribute to specify the receptor activity.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References

The plant hormone cytokinin has numerous functions in the regulation of developmental and physiological aspects throughout the plant’s life cycle (Argueso et al., 2009; Werner and Schmülling, 2009). It is perceived in Arabidopsis thaliana by three membrane-located sensor histidine kinases, AHK2, AHK3 and CRE1/AHK4 (Inoue et al., 2001; Suzuki et al., 2001; Ueguchi et al., 2001; Yamada et al., 2001). The output reaction of a given tissue to cytokinin can be regulated at many different levels, for example by the receptors’ expression levels, receptor sensitivity, coupling to downstream signal transduction proteins or by other factors. Genetic studies have shown that there is a considerable degree of functional redundancy of these three receptors (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006). Loss-of-function mutations of single receptors displayed no or only subtle effects on most of the phenotypes studied. However, in some cases individual receptors were shown to mediate specific cytokinin activities. These activities include a role for CRE1/AHK4 alone in regulating the sensitivity to cytokinin in root elongation (Inoue et al., 2001; Riefler et al., 2006) and a major role for AHK3 (together with AHK2) in regulating leaf senescence (Kim et al., 2006; Riefler et al., 2006). Cell differentiation in the transition zone of the root meristem appears to be mainly regulated by AHK3 (Dello Ioio et al., 2007), while CRE1/AHK4 regulates embryonic root patterning (Mähönen et al., 2000; Müller and Sheen, 2007), the phosphate starvation response (Franco-Zorrilla et al., 2002, 2005) and sulfate assimilation (Maruyama-Nakashita et al., 2004). No specific function has been assigned so far to AHK2, although this receptor alone was sufficient to maintain normal plant growth under standard growth conditions (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006).

The receptor genes were reported to be expressed in almost all tissues. CRE1/AHK4 showed a higher expression in the root, while AHK2 and, in particular, AHK3 transcripts are more abundant in the shoot (Ueguchi et al., 2001; Higuchi et al., 2004). Analysis of promoter:GUS fusion genes showed that the receptor genes have overlapping expression domains and are transcribed in almost all cells in the different organs although with different strength (Nishimura et al., 2004).

Studies of the biochemical properties of the Arabidopsis AHK3 and CRE1/AHK4 cytokinin receptors were performed in E. coli- or yeast-based test systems expressing individual receptors of Arabidopsis (Suzuki et al., 2001; Yamada et al., 2001; Spíchal et al., 2004; Romanov et al., 2005, 2006). These studies revealed a KD for cytokinins in the low nanomolar range for both receptors (Yamada et al., 2001; Romanov et al., 2005, 2006). Interestingly, CRE1/AHK4 showed a high affinity to both principal cytokinins, iP and tZ, and did not recognize other cytokinins well. In contrast, AHK3 showed a broader spectrum of ligand recognition (Spíchal et al., 2004). A difference in ligand binding of the two receptors is also suggested by their different response to PI-55, an inhibitor of cytokinin receptors (Spíchal et al., 2009). Competition experiments directly performed on receptor-expressing E. coli indicated an about 10-fold lower affinity for iP and iPR compared with tZ and tZR for AHK3 (Romanov et al., 2006). Differences in affinities for iP- and tZ-type cytokinin were also reported for maize cytokinin receptors (Yonekura-Sakakibara et al., 2004). It has been hypothesized that AHK3 might be sensitive specifically to root-derived cytokinin (tZ-type) and less to shoot cytokinin (iP-type) (Romanov et al., 2006; Hirose et al., 2008; Romanov, 2009). Heterologous expression of AHK2 has been proven to be difficult and, therefore, its ligand recognition has not yet been studied.

Our data presented here show that the AHK2 CHASE domain binds cytokinins with high affinity and resembles the CRE1/AHK4 CHASE domain in its ligand specificity. Furthermore, we compared the sensitivity of each of the three receptors to iP and tZ in planta, explored the tissue-specific signalling domains of the three individual receptors, studied the relevance of specificity of receptor expression in a complementation assay, and investigated if the cytoplasmic histidine kinase domain also has a function in signal specification.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References

The CHASE domain of AHK2 binds cytokinin with high affinity

In order to study the cytokinin-binding activity of AHK2, we aimed to express AHK2 in E. coli. However, several attempts to express the full length receptor failed, indicating that the protein may be toxic for E. coli. Because the CHASE domain of CRE1/AHK4 is sufficient for ligand binding (Heyl et al., 2007), we cloned and expressed only AHK2-CHASE including the two adjacent transmembrane domains (CHASE-TM). As a control, the CHASE-TM domain of CRE1/AHK4 was expressed in the same vector system. The mean affinity constant KD for CRE1/AHK4 CHASE-TM was 9.1 ± 1.5 nm with tZ as a ligand (Table 1 and Figure 1b), which is close to the KD of 3.1 nm (Romanov et al., 2006) and 4.4 nm (this study, Table 1) reported for the full length receptor. The KD of AHK2 CHASE-TM with tZ was also in that range, namely 4.0 ± 0.7 nm (Table 1 and Figure 1a). Scatchard plots indicated single non-cooperative binding sites for both the AHK2-CHASE-TM and CRE1/AHK4-CHASE-TM domains (Figure 1a,b).

Table 1.   Comparison of the cytokinin affinity of AHK2 and CRE1/AHK4
CytokininAbbreviationApparent KD (nm)a
CHASE domain AHK2CHASE domain CRE1/AHK4Full length CRE1/AHK4
  1. aAll tests without standard deviation (SD) have been performed once, all other binding tests have been performed twice.

trans-ZeatintZ4.0 ± 0.79.1 ± 1.54.4 ± 1.1
cis-ZeatincZ189 ± 23286454 ± 134
IsopentenyladenineiP1.43 ± 0.0223.5 ± 1.6
trans-Zeatin ribosidetZR16.0 ± 1.47546 ± 0.5
Isopentenyladenine ribosideiPR6.5 ± 0.822
DihydrozeatinDHZ214 ± 47162224 ± 6
BenzyladenineBA45 ± 62664 ± 1.5
ThidiazuronTD2.520
trans-Zeatin O-glucosidetZOG>10 000>10 000>10 000
AdenineAde>10 000
image

Figure 1.  Ligand binding by the CHASE-TM domains of AHK2 and CRE1/AHK4. (a, c) AHK2 CHASE-TM domain. (b, d) CRE1/AHK4 CHASE-TM domain. (a, b) [3H]trans-zeatin binding in the presence of different concentrations of unlabelled trans-zeatin. The inserts show Scatchard plots created from data in the respective figure. Bs, specifically bound trans-zeatin; U, unbound trans-zeatin. (c, d) Competition of different cytokinins with [3H]trans-zeatin for binding to CHASE domains. The curves were calculated using the simple ligand binding option of the SigmaPlot 9 program (Systat Software, San Jose, CA, USA). BA, benzyladenine; cZ, cis-zeatin; DHZ, dihydrozeatin; iP, isopentenyladenine; iPR, isopentenyladenine riboside; TD, thidiazuron; tZ, trans-zeatin; tZOG, trans-zeatin O-glucoside; tZR, trans-zeatin riboside.

Download figure to PowerPoint

The ligand preferences of the AHK2 and CRE1/AHK4 CHASE-TM domains were measured in a competition assay and found to be rather similar and in reasonable accord with the physiological activities of different cytokinins. Both CHASE-TM domains showed a high affinity to iP and tZ and much lower affinity to the ribosides of iP and tZ (Figure 1c,d and Table 1); notably, their affinities to cis-zeatin (cZ) and dihydrozeatin were more than one order of magnitude lower (Figure 1c,d and Table 1). The relative affinities for different cytokinins of CRE1/AHK4 CHASE-TM and the full length CRE1/AHK4 receptor (Romanov et al., 2005, 2006; Table 1) were in good accord.

All three cytokinin receptors show high sensitivity in planta

The analysis of the AHK2 CHASE-TM domain has completed the data about the ligand-binding properties of Arabidopsis cytokinin receptors. To test if bacterial models truly reflect the behaviour of receptors in planta, we studied the activity of each of the three receptors singly in planta, focussing on a comparison of tZ and iP and using two different approaches. Firstly, we measured the transcriptional activation of the two primary cytokinin response genes ARR5 and ARR6. Secondly, we analyzed the activation of the cytokinin-sensitive reporter gene PARR5:GUS (D’Agostino et al., 2000; Romanov et al., 2002). All experiments were carried out in three different double-receptor mutants, each retaining a different single receptor, and in the wild type.

Figure 2(a) shows that the activation of ARR5 and ARR6 in wild type by iP and tZ was almost identical. One nanomolar of each was sufficient to induce a significant increase in transcript abundance and a further increase was obtained with 5 nm of both cytokinins and for both response genes. The sensitivity towards iP and tZ differed in mutants that retained AHK3 or CRE1/AHK4 as the sole cytokinin receptor. ahk2 cre1 mutants showed a lower sensitivity to iP, while ahk2 ahk3 mutants were less sensitive to tZ (Figure 2c,d). In comparison, AHK2 showed a lower overall sensitivity than the two other receptors. Transcript levels of ARR5 and ARR6 were not significantly increased by 1 nm iP or tZ (Figure 2b). At a higher concentration (5 nm), gene activation by tZ was stronger than that by iP. A full response in the wild type, as well as in the receptor mutants retaining single receptors, was achieved with 10–40 nm iP and tZ. Even higher concentrations did not increase the transcript levels further, which indicated saturation of the reaction (data not shown).

image

Figure 2.  Activation of immediate early cytokinin response genes. Induction of ARR5 and ARR6 expression by different cytokinins in wild type (a); and the double-receptor mutants ahk3 cre1 (b); ahk2 cre1 (c); and ahk2 ahk3 (d). Five-day-old seedlings were induced for 20 min with different concentrations (1 nm or 5 nm) of cytokinin (iP or tZ). Expression levels of the two cytokinin response genes ARR5 and ARR6 were analyzed by qRT-PCR. The expression level of untreated samples was set 1. Data represent mean values (±SD) of three technical replicates. anova was used to compare values of plants treated with equal concentrations of iP and tZ. *< 0.05; **< 0.01; ***< 0.001.

Download figure to PowerPoint

To study the differential response to iP and tZ further, the cytokinin-specific reporter gene PARR5:GUS (D’Agostino et al., 2000) was introgressed into the double-receptor mutants. Figure 3(a) shows that the reporter gene was activated in wild-type background by both iP and tZ, with a lower activity of iP in the low concentration range. Plants retaining only AHK2 or CRE1/AHK4 showed a comparable activation by both iP and tZ (Figure 3b,d). Full activation was already obtained by 10 nm of either cytokinin in CRE1/AHK4, while 40 nm were required for AHK2. AHK3 responded to tZ with a similar sensitivity as did CRE1/AHK4, however a several fold higher concentration of iP was required for the same response (Figure 3c). It is interesting to note that the receptor output activities are very similar for all three receptors (i.e. in the range of 150–200 nmol 4-methylumbelliferone (4-MU) mg protein−1 h−1) and that the sum of single receptor activities corresponds to the activity measured in the wild type. The response of the reporter gene in the wild type, as well as in all three mutant lines, did not increase further at concentrations more than 10–40 nmtZ (Figure 3), which indicated saturation of the reaction similar to that described above for the transcriptional activation of immediate early response genes (Figure 2).

image

Figure 3. PARR5:GUS reporter gene activation by different cytokinins in double cytokinin receptor mutants. Induction of the PARR5:GUS reporter gene by different cytokinins in wild type (a); and the double-receptor mutants ahk3 cre1 (b); ahk2 cre1 (c); and ahk2 ahk3 (d). Five-day-old seedlings were incubated with tZ or iP for 5 h at 22°C. GUS activity was determined as nmol 4-methylumbelliferone (4-MU) mg protein−1 h−1. Data represent mean values ± SD (= 2).

Download figure to PowerPoint

Taken together, AHK2 shows a lower sensitivity in planta in these assays than the other two receptors, while AHK3 shows a lower sensitivity to iP compared with tZ.

Cytokinin receptors mediate the cytokinin response in distinct cell types

Next, we took advantage of the receptor mutant lines that harboured the PARR5:GUS reporter, which has been used previously to record the endogenous cytokinin status (e.g. D’Agostino et al., 2000; Werner et al., 2003; Lohar et al., 2004; Aloni et al., 2005), to analyze the tissue-specificity of the cytokinin response in plants that contained only a single receptor. Studies using promoter:GUS fusion genes for each of the three receptors have revealed overlapping, but also partly distinct, expression zones (Higuchi et al., 2004; Nishimura et al., 2004), which was also supported by microarray analyses (Zimmermann et al., 2004; Brady et al., 2007; Winter et al., 2007). However, it is not known if the promoters used to regulate the reporter gene properly reflect the expression domains of the endogenous genes and whether or not the receptors would indeed be active in the different tissues that expressed the gene. Furthermore, we were interested in analyzing whether or not the PARR5:GUS reporter gene would be activated in the common expression zones by different receptors with different efficiency or whether they would show differences in cytokinin responsiveness.

First, we studied the spatial domains of PARR5:GUS gene activation in 6-day-old Arabidopsis seedlings without cytokinin. Figure 4(a) shows that the strongest activity was seen in all genotypes in the shoot and root tips. After a prolonged incubation period of up to 24 h or treatment with cytokinin (100 nmtZ), most tissues displayed GUS activity indicating that many different cell types possess at least a basal ability to sense and respond to cytokinin (Figure 4b,c; see also D’Agostino et al., 2000).

image

Figure 4.  Expression of PARR5:GUS in wild type and cytokinin receptor double mutants. (a) Whole seedlings stained for 5 h. (b) Shoots and (c) roots stained for 24 h. (d) Close-up of the upper part of the shoot (5 h staining). (e) Cotyledons. (f) Epidermal cells of cotyledons with stomata (first row) and spongy parenchyma cells (second row) stained for 24 h. (g, h) Root tips stained for 12 min and 5 h, respectively. Arabidopsis seedlings were cultivated in liquid media and stained 6 dag (days after germination). White arrowheads, quiescent centre; black arrowheads, emerging GUS staining in root vasculature. Cytokinin treatment, as indicated by +CK in (b, d, e, h), was for 5 h with 10 nm or 100 nmtZ prior to the start of the staining procedure. Experiments were repeated twice and representative phenotypes are shown (= 30).

Download figure to PowerPoint

Closer inspection revealed that the response domains were not identical for different receptors. In the shoot apex, CRE1/AHK4 transmitted the cytokinin signal only in a limited central region, while both AHK2 and AHK3 activated the reporter in a larger domain (Figure 4d). Following treatment with cytokinin (10 nmtZ) the CRE1/AHK4 activity domain was enlarged while those of AHK2 and AHK3 remained unchanged (Figure 4d). In cotyledons of wild-type seedlings, expression of the reporter gene was strongest in the vasculature but only AHK2 and AHK3 were able to activate the reporter gene in this tissue (Figure 4e). In addition, AHK2 activated expression in parenchyma cells and AHK3 activated expression in both parenchyma cells and stomata (Figure 4f). In contrast, no expression of the reporter gene was achieved in these cells by CRE1/AHK4. Cytokinin treatment caused a stronger leaf expression of PARR5:GUS in the presence of AHK2 or AHK3 but did not induce expression in the line harbouring only CRE1/AHK4 (Figure 4e,f).

Differences in the activity of single receptors also became apparent in root tissue. In wild-type background, the strongest staining was noted in the root cap columella cells and the vasculature (Figure 4g,h). Staining of the vasculature was mainly attributed to the activity of CRE1/AHK4 as it disappeared in cre1 mutants. However, after longer staining and after cytokinin treatment, AHK2 and AHK3 also activated the reporter gene in the root vasculature, indicating that both receptors are present but contribute to only a small part of the total signal output (Figure 4c). The first staining in the root tip was visible already after few minutes in all double mutants in the columella cells, while the quiescent centre was not stained (Figure 4g). There was also only low apparent receptor activity in the cell division zone of the root meristem that was slightly enhanced upon cytokinin treatment (Figure 4h).

CRE1/AHK4 can functionally replace AHK2 but not AHK3

The data reported above have revealed that AHK2 and CRE1/AHK4 share a similar ligand preference, while that of AHK3 differs. Moreover, all three receptor genes showed partially distinct domains of activity. In order to study the functional relevance of the differential regulation of expression, we performed promoter-swap experiments. As a tester line for functional activity, we used the ahk2 ahk3 double mutant, which shows a distinct shoot phenotype that consisted of a reduced rosette size and shoot height, and insensitivity to cytokinin in a chlorophyll retention assay in dark-induced leaf senescence. In contrast, the ahk2 and ahk3 single mutants grew in a similar way to the wild type (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006).

We transformed ahk2 ahk3 with the AHK2, AHK3 or CRE1/AHK4 genes retaining approximately 2 kb of their own 5′ promoter regions as well as with chimeric constructs containing the same promoter regions of AHK2 or AHK3 linked to the CRE1/AHK4 coding region (Figure 5a). Figure 5(b) shows that both PAHK2:AHK2 and PAHK3:AHK3 were able to complement the ahk2 ahk3 mutant. In contrast, all of the nine transformants expressing the PCRE1:CRE1/AHK4 gene retained the ahk2 ahk3 mutant phenotype (data not shown). Evidence of the functionality of the PCRE1:CRE1/AHK4 gene was obtained by transforming plants that were homozygous ahk2 cre1 mutants and retained a single active ahk3 allele. Selfed progeny of these plants segregated 1:3 for the triple mutant phenotype. In contrast, none of 42 transformed plants containing the PCRE1:CRE1/AHK4 gene displayed the triple mutant phenotype, 14 transformants showed the ahk2 ahk3 phenotype and 28 showed a wild-type phenotype. Together this finding indicates complementation of the cre1 allele by the PCRE1:CRE1/AHK4 gene. We conclude that CRE1/AHK4 was unable to functionally replace either of the two other receptors when expressed under control of its own promoter. In contrast, ahk2 ahk3 double mutants that expressed the CRE1/AHK4 gene under control of the AHK2 promoter (Figure 5c) showed almost full complementation of the mutant phenotype. Rosette diameter and shoot length were similar to the wild type (Figure 5d,e) and the ability to respond to cytokinin in a dark-induced senescence assay was re-established (Figure 5f). In contrast, expressing CRE1/AHK4 under the transcriptional control of the AHK3 promoter (Figure 5c) did not complement the ahk2 ahk3 phenotype (Figure 5d–f).

image

Figure 5.  Promoter-swap experiments. (a) Schematic drawing of promoter-swap constructs. The receptor genes were cloned as full length cDNAs or genomic DNA as indicated. (b) Complementation of ahk2 ahk3 mutant by PAHK2:CRE1/AHK4. Plants are shown 24 dag. For each construct two independent homozygous lines were tested (lower row, line I; upper row, line II). (c) RT-PCR analysis shows the expression of the CRE1/AHK4 transgene in the promoter-swap lines. The Actin2 gene was used as a control. (d, e) Rosette diameter (d); and shoot height (e); of wild type, double mutant ahk2 ahk3 and two representatives (I, II) of ahk2 ahk3 plants transformed with the constructs shown in (a). Rosette diameter was measured 25 dag, shoot height was determined 46 dag. Data represent mean values ± SD (= 20). The independent transformants I (left bar) and II (right bar) were compared with the ahk2 ahk3 mutant in two independent experiments. The significance of differences was evaluated separately for both experiments using anova. *< 0.025. (f) Chlorophyll retention test in a detached leaf assay. Leaf chlorophyll content before start of dark incubation was set to 100% for each genotype. Data represent mean values ± SD of two probes, each consisting of leaf number six of five independent plants.

Download figure to PowerPoint

The CHASE domain, but not the histidine kinase domain, of AHK3 is required to determine receptor specificity

The failure of PAHK3:CRE1/AHK4 to complement the ahk2 ahk3 mutant might be due to the inability to recognize the cytokinin signal appropriately and/or to couple to downstream signalling elements. In order to study the relevance of different receptor domains to mediate specificity we coupled the cytoplasmic domain of CRE1/AHK4 to the CHASE-TM of AHK3, resulting in a chimeric protein named AHK3-CRE1 (Figure 6a). Figure 6(b–f) show that expression of the AHK3-CRE1 gene under control of the AHK3 promoter partially complemented the ahk2 ahk3 double-mutant phenotype. ahk2 ahk3 mutants that expressed PAHK3:AHK3-CRE1 had rosette sizes, shoot height and cytokinin responses in the chlorophyll retention assay that were intermediate between the wild type and the ahk2 ahk3 mutant.

image

Figure 6.  Domain-swap experiment. (a) Schematic drawing of the domain-swap construct PAHK3:AHK3-CRE1. The protein domains shown below include the N-terminal CHASE domain with two adjacent transmembrane domains (TM1 and TM2) from AHK3 and the C-terminal part including the histidine kinase domain (HisKA), the ATP-binding domain (HATPse) and receiver domain (REC) from CRE1/AHK4. (b) Partial complementation of ahk2 ahk3 by expression of PAHK3:AHK3-CRE1. Two independent homozygous transgenic lines (I, II) are shown 24 dag. (c) RT-PCR analysis shows the expression of the AHK3-AHK4 transgene in both domain-swap lines. The Actin2 gene was used as a control. (d) Rosette diameter of WT, ahk2 ahk3 double mutant and two independent homozygous domain-swap lines (I, II) 27 dag. (e) Shoot height at the end of flowering. Data represent mean values ± SD (= 25). anova was used to compare values with the ahk2 ahk3 mutant. *< 0.001. (f) Chlorophyll retention test in a detached leaf assay. Leaf chlorophyll content before start of dark incubation was set to 100% for each genotype. Data represent mean values ± SD of two probes, each consisting of leaf number six of five independent plants.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References

Analysis of the ligand binding properties of the AHK2-CHASE-TM domain completes the data about the affinities of Arabidopsis cytokinin receptors for different cytokinins and showed that the AHK2 receptor shares a high affinity to tZ with AHK3 and CRE1/AHK4. The KD for tZ was in the low nanomolar range for all three receptors (Romanov et al., 2006; Table 1), which is in good agreement with measurements of endogenous cytokinin concentrations and also in accordance with the cytokinin concentration needed for experimental activation of the receptor output in planta (Figures 2 and 3). Thus, although the ligand-binding CHASE domains share only about 65% amino acid sequence similarity (Spíchal et al., 2004; Heyl et al., 2007), they have retained a fairly high similarity in recognizing different cytokinins indicating their functional relevance and the existence of evolutionary constraints. Notably, the in planta response mediated by AHK2 needed higher concentrations of cytokinin compared to the two other receptors indicating that additional receptor-specific regulatory cellular factors may modulate the responsiveness. Another possibility is a different subcellular localization of the receptors causing a different accessibility for the exogenously added cytokinins.

The ligand preference of AHK2 resembled more closely the one of CRE1/AHK4 than that of AHK3. A higher functional similarity between AHK2 and CRE1/AHK4 is also suggested by the promoter-swap experiment, in which CRE1/AHK4 expressed under control of the AHK2 promoter complemented the ahk2 ahk3 loss-of-function phenotype (Figure 5). In contrast, complementation was not achieved by expression of CRE1/AHK4 under control of the AHK3 promoter. Partial complementation of the ahk2 ahk3 loss-of-function phenotype by a chimeric AHK3 CRE1/AHK4 protein indicated that the inability of CRE1/AHK4 to replace AHK3 is not due to another property of the cytoplasmic domain, but more likely to an, as yet unknown, feature of the AHK3-CHASE-TM domain. However, other reasons such as differences in protein localization and/or stability may also influence the experimental result. The interchangeability of the cytoplasmic domains of AHK3 and CRE1/AHK4 is consistent with their redundant interaction with all five AHP proteins of the two-component signalling system (Dortay et al., 2006) and the dominant-negative activity of the wol mutation on the whole cytokinin pathway (Mähönen et al., 2006). The fact that only a partial complementation was achieved by the chimeric AHK3-CRE1/AHK4 protein may be due to other differences in interacting with downstream partners (Dortay et al., 2008). Receptor-specific response pathways are also suggested by molecular phenotypes of receptor mutants. The cre1-1 mutant, for example, specifically lacks the activation of ARR15 and ARR16 (Kiba et al., 2002).

No specific function in regulation of plant growth has been assigned to AHK2 alone, in contrast to AHK3 and CRE1/AHK4. In many instances, AHK2 seems to operate together with AHK3; for example, neither ahk2 nor ahk3 mutation alone has a major effect on plant growth, while ahk2 ahk3 double mutants show marked developmental impairments, indicating that the two receptors act through a common pathway. In this context, it is interesting that AHK2 and AHK3 are evolutionary more closely related to each other than either of the two with CRE1/AHK4 (Pils and Heyl, 2009). It seems that following gene duplication and independent evolution, these two receptors still cover redundantly common shoot-associated functions. Whereas the ligand recognition properties of AHK2 have remained more similar to those of CRE1/AHK4, AHK3 has developed a relatively lower affinity to iP. An additional partial divergence was achieved by changes in gene expression, for example by establishing or maintaining AHK3 expression in the stomata. However, a largely overlapping expression and activity domain pattern of AHK2 and AHK3 has been stably maintained during evolution, indicating the functional relevance of the simultaneous presence of both receptors. Receptors showed differences in biochemical traits, such as pH- or salt-dependence, which might contribute to their signalling variation (Romanov et al., 2006). It would be interesting to study to what extent AHK2 can overtake AHK3 functions and vice versa.

The relative lower sensitivity of AHK3 to iP measured in bacteria (Romanov et al., 2006) has been confirmed in planta in this study (Figures 2 and 3), although the difference was less strong. The available data suggest that the AHK3 receptor is particularly mediating sensitivity to root-derived tZ-type cytokinin (Hirose et al., 2008), but less so to iP-type cytokinin synthesized in the shoot. This difference may be of specific functional relevance in those cells that express only AHK3, e.g. the stomata. In other cells, AHK2 may overtake (most of) the iP-dependent functions as its activity domain overlaps with AHK3 (Figure 4).

The output profiles of PARR5:GUS triggered by single receptors has identified activity of all three receptors in numerous tissues and, in particular, the SAM and root cap columella cells (Figure 4). The activity domain of CRE1/AHK4 in the shoot apex was less broad than those of AHK2 and AHK3 consistent with previous reports (Mähönen et al., 2006; Gordon et al., 2009). It is evident from the expression pattern, as well as the mutant phenotype, that all three cytokinin receptors have a function in regulation of SAM activity. A very strong receptor output was recorded for all three receptors in root columella cells. Intriguingly, genes encoding proteins of cytokinin synthesis and breakdown are expressed in columella cells, indicating their active cytokinin metabolism (Werner et al., 2003; Miyawaki et al., 2004). A function for cytokinin in the columella cells is not known but roles as the origin of an apical-basal cytokinin gradient in the root and in regulation of root geotropism have been proposed (Aloni et al., 2004, 2006).

Apparently, only few tissues sense cytokinins by mainly one cytokinin receptor. One such tissue is the root vasculature, where CRE1/AHK4 is known to regulate cellular differentiation (Mähönen et al., 2000, 2006). This finding also suggests that cytokinin-dependent responses to nutritional changes that are mediated by CRE1/AHK4 (reviewed by Argueso et al., 2009; Rubio et al., 2009; Werner and Schmülling, 2009) are realized in the vasculature. The stomata are a cell-type that display activity of only AHK3. Cytokinin has been described as regulating the behaviour of stomata (Acharya and Assmann, 2009) and it could be that this activity is linked to the recently described function of cytokinin in regulating the drought response (Rivero et al., 2007).

Together, differences in receptor affinity to different cytokinin metabolites as well as differences in receptor gene regulation both contribute to tissue- and cell-specific responses of the cytokinin system. Certainly, the presence of several receptors in most tissues and their specific coupling to downstream signal transduction chains offers numerous additional ways to fine-tune the biological outcome of a cytokinin signal.

Experimental Procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References

Cytokinin binding assay

[3H]trans-zeatin of high purity and specific activity (851 GBq mmol−1) was provided by the Isotope Laboratory of the Institute of Experimental Botany (Prague, Czech Republic). Non-labelled cytokinins were from Olchemim (Olomouc, http://www.olchemim.cz). DNA sequences encoding the CHASE domains of Arabidopsis AHK genes and its adjacent membrane-spanning segments were cloned in pDEST15 vectors using the GATEWAY® Cloning Technology (Invitrogen, http://www.invitrogen.com) (Heyl et al., 2007) and expressed in E. coli strain BL21DE3pLys. To test the binding activity of the full length CRE1/AHK4 receptor, pDEST15 expressing CRE1/AHK4 was used (Heyl et al., 2007). Bacteria were grown in liquid LB medium supplemented with 50 and 200 μm IPTG for CRE1/AHK4-CHASE and AHK2-CHASE, respectively. Growth was continued with gentle shaking at 24°C for 16 h (AHK2-CHASE) or 30 h (CRE1/AHK4-CHASE). Then bacteria were precipitated at low rpm and LB medium was replaced with salt solution buffered with 50 mm MES-KOH, pH 7 as described (Romanov et al., 2006). Binding was performed according to Romanov et al. (2005). KD were determined as the average of two independent Scatchard analyses (Scatchard, 1949). Mean values and standard errors were calculated according to Student’s t-test.

Plant material and growth conditions

Arabidopsis thaliana accession Columbia (Col-0) was used as wild type. Receptor mutant lines containing the alleles ahk2-5, ahk3-7 and cre1-2 were those described by Riefler et al. (2006). The PARR5:GUS reporter line was described by D’Agostino et al. (2000) and introduced in receptor mutant background by crossing. Experiments were performed with plants homozygous for the reporter gene. Unless otherwise stated, plants were grown on soil or in vitro on MS medium as modified by Kemper et al. (1992) at 22°C and a 16 h light/8 h dark cycle. For in vitro assays seeds were surface sterilized and vernalized before transfer to the growth chamber.

Gene cloning and plant transformation

All gene constructs were generated by using the MultiSite GATEWAY® Cloning Technology (Invitrogen). Promoters were cloned into the entry vector pDONRP4-P1R using appropriate primers as follows PAHK2-fw, ggggacaactttgtatagaaaagttgTTTTTTTCGCATATGCCAC; PAHK2-rv, ggggactgcttttttgtacaaacttgTTCGACTCCTAATCTCAGATTC; PAHK3-fw, ggggacaac tttgtatagaaaagttgAACCCTGGACGATTTGCAT; PAHK3-rv, ggggactgcttttttgtacaaacttgCCACCACTTGAATACACG; PAHK4-fw, ggggacaactttgtatagaaaagttgCTTTCTCGGAAGAGCACAATG; PAHK4-rv, ggggactgcttttttgtacaaacttgATCTGAGCTACAACAATAGAG. Primer sequences written in lower case letters indicates GATEWAY® vector recombination sites. The coding sequence of the cytokinin receptor CRE1/AHK4 (At2g01830.1) was amplified by PCR using a cDNA library from Arabidopsis thaliana C24 as a template (Minet et al., 1992). The coding sequence of AHK3 (At1g27320.1) was amplified by RT-PCR from total RNA extracted from primary leaves of Arabidopsis thaliana Col-0. The AHK2 gene (At5g35750.1) was amplified from the genomic DNA of BAC clone MIXH1 (Liu et al., 1995). Primer sequences were as follows: AHK2-fw, aaaaagcaggctAAATGTCTATAACTTGTGAGC; AHK2-rv, agaaagctgggtTTAACAAGGTTCAAAGAATCTTGC; AHK3-fw, aaaaagcaggctTGATGAGTCTGTTCCATG; AHK3-rv, agaaagctgggtATTATGATTCTGTATCTG; AHK4-fw, aaaaagcaggctTGATGAACTGGGCACTCAAC; AHK4-rv, agaaagctgggtATTACGACGAAGGTGAG. Primer sequences written with lower case letters indicate GATEWAY® vector recombination sites. The amplification products were cloned via in vitro recombination in the entry vectors pDONR201 (CRE1/AHK4), pDONR222 (AHK3) or pDONR221 (AHK2) and subsequently verified by sequencing. The constructs were finally introduced in the vector pk7m24GW,3 (Karimi et al., 2007) and transformed into Arabidopsis using the floral dip method (Clough and Bent, 1998).

Analysis of gene expression

Expression of transgenes was verified with total RNA extracted from 7-day-old homozygous seedlings using the TRIzol method as described by Brenner et al. (2005). Total RNA was treated with RNase-free DNase I (Fermentas, http://www.fermentas.com) and RT-PCR was performed using the OneStep RT-PCR kit (Qiagen, http://www.qiagen.com) and Actin2 as a control. The primers used for RT-PCR were AHK4 RT-fw (TGCCACAGATGGACGGATTTGAAGC) and attB2-GW-3′-rv (CCACTTTGTACAAGAAAGCTG) to detect the transgenic CRE1/AHK4 transcript and Actin2-fw (TACAACGAGCTTCGTGTTGC) and Actin2-rv (GATTGATCCTCCGATCCAGA) for the Actin2 transcript. Quantitative real-time PCR was performed with RNA of 5-day-old seedlings which were treated 20 min or 30 min with different concentrations of tZ or iP (1–40 nm). Seedlings were grown in Petri dishes containing half-strength MS medium. RNA was purified using the RNeasy kit (Qiagen) and cDNA was synthesized from 5 μg RNA using SuperScript™ III Reverse Transcriptase (Invitrogen). For each probe biological duplicates were prepared and PCR reactions were performed in technical triplicates in one PCR setup. UBC10 (At5g53300) and PDF1 (At3g25800) were used as references genes (Czechowski et al., 2005). Primer sequences were as follows: for ARR5 ARR5-qRT-fw (CTACTCGCAGCTAAAACGC) and ARR5-qRT-rv (GCCGAAAGA ATCAGGACA); for ARR6 ARR6-qRT-fw (GAGCTCTCCGATGCAAAT) and ARR6-qRT-rv (GAAAAAGGCCATAGGGGT); for UBC10 UBC10-qRT-fw (CCATGGGCTAAATGG AAA) and UBC10-qRT-rv (TTCATTTGGTCCTGTCTTCAG); and for PDF1 PDF1-qRT-fw (CCATTAGATCTTGTCTCTCTGCT) and PDF1-qRT-rv (GACAAAACCCGTACCGAG). Data analysis was done using 7500 Software v2.0.1 (Applied Biosystems, http://www.appliedbiosystems.com).

Measurement of GUS activity

Histochemical analysis of the GUS reporter enzyme was performed essentially according to Jefferson et al. (1987). Measurement of GUS enzyme activity was also according to Jefferson et al. (1987) with modifications (Romanov et al., 2002) using methylumbelliferyl-β-d-glucuronide (MUG) as a substrate.

Chlorophyll retention assay

The chlorophyll retention assay and determination of chlorophyll concentration was carried out as described by Riefler et al. (2006) using 80% acetone instead of methanol for chlorophyll extraction.

Determination of morphometric parameters

Rosette diameter of plants grown on soil in the greenhouse was measured with a ruler 25–27 days after germination (dag) from at least 20 different plants taking two independent measurements per plant. Shoot height of the main shoot was measured at 46–48 dag.

Statistical analysis

The statistical significance of differences was evaluated by one-way analysis of variance (anova) using sas v.9.2 software (SAS Institute Inc., http://www.sas.com).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References

We thank Anne Cortleven for an introduction into statistical analysis by anova. This work was supported by grants of the Deutsche Forschungsgemeinschaft in the frame of Sfb 449 and the Arabidopsis Functional Genomics Network (AFGN), of the Russian Foundation of Basic Research (NN 10–04–00638 and 11–04–00614), Program of Presidium RAS ‘Molecular and cell biology’ and DAAD fellowships to S.N.L. and G.A.R.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References