Attenuation of cytokinin signaling via proteolysis of a Type-B response regulator

Abstract

Type-B response regulators (ARRs) are a group of transcription factors that are activated by cytokinin-initiated phospho-relays and regulate the expression of cytokinin-responsive genes. Recently, we reported that proteolysis of ARR2 in Arabidopsis is facilitated by cytokinins, resulting in attenuation of the signaling output of two-component circuitry. Interestingly, despite similarities in the primary structures and conserved receiver domains, the proteolytic properties of ARR2 are distinct from those of the other type-B ARRs. Using a gain-of-function mutant (ARR2^K90G) resistant to protein degradation, we clearly demonstrated that increased levels of ARR2 affected cytokinin-mediated processes such as primary root growth, callus induction, hypocotyl elongation and leaf senescence. At the molecular level, expression of type-A ARRs was increased in transgenics expressing ARR2^K90G, resulting in enhanced cytokinin sensitivity. Here, we describe these findings and how they may be incorporated into the currently accepted model for the regulation of cytokinin signaling. In addition, we describe the proteomic approaches used to identify proteins that interact with ARR2. The putative roles of ARR2 proteolysis are also addressed with regard to other developmental processes. In conclusion, cytokinin-facilitated degradation of ARR2 must be appreciated as a post-translational event important for regulating cytokinin signaling intensity.

Cytokinins are a group of phytohormones that mediate various developmental processes (e.g., seed germination, vasculature development, meristem maintenance, apical dominance and leaf senescence) and (a)biotic stress adaptation processes in plants.^1,2 The physiological effects of cytokinin responses are largely modulated by cytokinin levels and the activation of target signaling pathways.³ At the cellular level, cytokinins are recognized by CHASE-domain-containing histidine kinases located in the plasma membrane⁴ and/or the endoplasmic reticulum (ER).^5-7 Upon cytokinin binding, type-B response regulators in the nucleus are activated via His-Asp multi-step phospho-relays similar to those observed in bacterial two-component systems (TCS). In Arabidopsis, 11 type-B response regulators (ARRs) have been identified and act as transcription factors with redundant function that participate in various developmental processes mediated by cytokinins.⁸

The output intensities of TCS signaling are primarily regulated by post-transcriptional and post-translational modifications of TCS components.³ For example, in Arabidopsis, the expression levels of AHKs (histidine kinase proteins), AHPs (histidine phosphotransfer proteins) and ARRs fluctuate in response to changing environmental conditions.³ At the protein level, sequential phosphorylations initiated by cytokinin receptors are essential for propagating cytokinin signals. However, the duration and intensity of these phosphorylation-propagated signals must be properly mediated by negative regulatory mechanisms. AHK4, a histidine kinase receptor, acts as a phosphatase in the absence of cytokinins, inhibiting signal transduction to type-B ARRs.⁹ AHP6, a phosphotransfer protein, negatively regulates cytokinin-mediated vasculature formation in the roots.¹⁰ Type-A ARRs, which are direct regulons of type-B ARRs, are also believed to mitigate cytokinin signaling via negative feedback, although the mechanisms that mediate this effect remain unknown.^11,12 In addition, Kim et al. (2012) reported recently that the cytokinin-facilitated proteolysis of ARR2 could be another post-translational regulatory event that moderates cytokinin signaling.¹³

Although reduced levels of ARR2 accumulation were observed in response to ethylene,¹⁴ the implications of this observation in cytokinin signaling were not explored. However, Kim et al. (2012) provided physiological evidence that protein degradation of ARR2 was a regulatory event in TCS important for various cytokinin-mediated developmental processes such as root/hypocotyl elongation and leaf senescence.¹³ At the molecular level, a degradation-resistant mutant of ARR2, ARR2^K90G, conferred sustained upregulation of ARR6, a type-A ARR that is a potential target of ARR2. Expression of ARR2^K90G also increased the expression of other type-A ARRs. Although the precise physiological roles of type-A ARRs are still largely unknown, it is evident that proteolysis of ARR2 affects a range of downstream molecular events.

On the other hand, it would be interesting to examine whether sustained upregulation of type-A ARRs in ARR2^K90G overexpressing lines results in the enhanced feedback regulations in cytokinin responses.^12,15 However, the primary root lengths of ARR2^K90G overexpressing lines were markedly shorter than those of the ARR2 overexpressing lines, despite high levels of induction of type-A ARRs.¹³ This suggests the presence of a compensatory mechanism insensitive to type-A ARRs-mediated negative feedback of cytokinin responses in ARR2^K90G transgenics. Alternatively, ARR2 degradation itself might be directly or indirectly associated with the negative feedback regulation of type-A ARR proteins. This hypothesis is partially supported by the experimental observation that although transient co-expression of type-A ARRs (i.e., ARR5 and ARR6) generally suppresses the induction of pARR6:luciferase enhanced by ARR2 in protoplasts (Fig. 1), such inhibition was not observed in the presence of ARR2^K90G, indicating that the stable form of ARR2 was insensitive to negative feedback by type-A ARRs. It will be intriguing to determine how type-A ARRs affect ARR2 proteolysis. However, the increased protein stability of ARR5 in the arr1 arr2 arr10 arr12 type-B quadruple mutant¹² suggests that negative feedback regulation of type-A ARRs might be closely related to the activity or stability of type-B ARRs, or vice versa, via protein-protein interaction. As phosphorylation-associated dimerization and oligomerization of response regulators have been suggested to occur in multiple organisms,^16,17 it would be worthwhile to examine the potential physical interaction between type-A ARRs and ARR2. Alternatively, complex formation might be mediated by the interaction of scaffold proteins with type-A and type-B ARRs, a possibility that also needs to be explored in the future.

Figure 1. The effect of ARR2^K90G on the negative regulation of type-A ARRs in cytokinin-induced expression of ARR6. Mesophyll protoplasts were transfected with pARR6:luciferase as a reporter and ARR2-HA, ARR2^K90G-HA, ARR5-HA or ARR6-HA as effectors. Luciferase activity was measured at 3 h after treatment with 100 nM t-zeatin. NE, no effector plasmids; NA, no type-A ARR effector plasmids.

The proteolytic properties of ARR2 are distinct from those of other type-B ARRs (i.e., ARR1, ARR10, ARR12 and ARR18).¹³ Experiments using transient expression systems established from Arabidopsis mesophyll protoplasts revealed that ARR2 undergoes post-translational modifications distinct from those of other type-B ARRs and therefore might play a role in the cytokinin response that is different from that of other type-B ARRs. In support of this, primary root elongation in the arr1 arr2 arr10 triple mutant displayed higher cytokinin sensitivity than the arr1 arr10 double mutant,⁸ which suggests that ARR2 might negatively regulate other type-B ARRs under certain conditions. Since proteolysis of ARR2 was abolished by MG132 treatment, ARR2 degradation must be mediated by an E3 ligase via direct ubiquitination. Currently, the regulatory events or elements that control ARR2 degradation are unknown. However, one clue can be found in the ethylene signaling pathway, in which signaling output is regulated by the transcriptional factor EIN3. Ethylene activates EIN3, which upregulates the expression of an F-box gene (i.e., EBF2) whose encoded product degrades EIN3 proteins in a negative feedback mechanism.^18,19 If ARR2 upregulates the expression of an E3 ligase that specifically targets ARR2 proteins, then cytokinin-dependent degradation of ARR2 may be a regulatory mechanism distinct from that of other type-B ARRs. Therefore, the identification of an E3 ligase that directly targets ARR2 proteins will be crucial for understanding the roles of ARR2 proteolysis in cytokinin signaling. On the other hand, ARR2 degradation kinetics revealed a hyperbolic decrease, in that proteolysis was slowed down after 1 h in Arabidopsis mesophyll protoplasts.¹³ This result implicates the presence of a negative feedback regulation on E3 ligase machinery targeting ARR2, but needs to be validated further.

To this end, we utilized proteomic approaches to identify proteins that interact with ARR2 in planta. Previously, we investigated the proteolytic properties of individual domains in ARR2 and found that the Gln rich Q-domain (379 amino acid residues at the C-terminal)²⁰ was resistant to degradation (data not shown). In addition, this domain exhibits a relatively variable amino acid sequence compared with the other type-B ARRs. To isolate ARR2 specifically, we generated a recombinant protein containing only the Q-domain of ARR2 fused to glutathione S-transferase (GST). After purification, the ARR2-Q-GST proteins were mixed with Arabidopsis seedling extracts and precipitated using glutathione sepharose resin. The co-eluted proteins were subjected to mass spectroscopy and identified using the MASCOT database (www.matrixscience.com). As shown in Table 1, the results revealed that ARR2 interacts with a wide variety of proteins, including chromatin remodeling factors (i.e., histones) and defense-related regulators (i.e., PR thaumatin family proteins, pectin acetylesterase, disease resistant receptors, JR1, etc). Noticeably, however, no AHP proteins known to interact with type-B ARRs were detected, although this was probably due to the exclusion of conserved receiver domains in the bait protein. Interestingly, an F-box family protein (At1g70390) was co-eluted with ARR2-GST. Although it is premature to address the role of At1g70390 in the proteolysis of ARR2, we hope to elucidate its effect on the proteolysis of ARR2 in the near future.

Table 1. Potential binding partners of ARR2 in Arabidopsis.

AGI ID	TAIR Annotation
AT5G40020	pathogenesis-related thaumatin family protein
AT4G19410	pectinacetylesterase, putative
AT2G30620	histone H1.2
AT5G44900	transmembrane receptor
AT4G13000	protein kinase family protein
AT5G56240	similar to unknown protein
AT1G44900	ATP binding/DNA binding/DNA-dependent ATPase
AT2G07110	similar to 3′ exoribonuclease family domain 1-containing protein
AT2G36870	xyloglucan:xyloglucosyl transferase
AT5G20360	tetratricopeptide repeat (TPR)-containing protein
AT5G38850	disease resistance protein (TIR-NBS-LRR class), putative
AT5G52470	FIB1 (FIBRILLARIN 1)
AT5G20630	GLP3 (GERMIN-LIKE PROTEIN 3); manganese ion binding
AT3G14310	ATPME3 (Arabidopsis thaliana pectin methylesterase 3)
AT2G33080	leucine-rich repeat family protein
AT5G18620	CHR17 (CHROMATIN REMODELING FACTOR17); DNA-dependent ATPase
AT1G16330	CYCB3;1 (CYCLIN B3;1); cyclin-dependent protein kinase
AT1G54690	histone H2A, putative
AT3G16470	JR1 (Jacalin lectin family protein)
AT5G62190	PRH75 (plant RNA helicase 75); ATP-dependent helicase
AT1G74260	catalytic
AT4G05320	UBQ10 (POLYUBIQUITIN 10); protein binding
AT1G70390	F-box family protein
AT3G16030	CES101 (CALLUS EXPRESSION OF RBCS 101); carbohydrate binding/kinase;

ARR2 has been demonstrated to be involved in leaf senescence mediated by AHK3⁴ and in systemic acquired resistance against bacterial pathogens via synergistic interaction with TGA3, a key transcription factor necessary for salicylic acid-dependent systemic acquired resistance.²¹ In addition, recent reports suggest emerging roles for cytokinins in lateral organogenesis,²² bud outgrowth control,²³ osmotic stress resistance²⁴ and dedifferentiation of somatic cells.²⁵ Based on the wide range of functions affected by cytokinins, ARR2 proteolysis might also be engaged in cross-talk with signaling pathways involved in other developmental and environmental adaptation processes. Therefore, the precise identification and analyses of ARR2-dependent regulons will shed light on the physiological roles of cytokinin-facilitated proteolysis of ARR2 and the mechanisms by which cytokinin signals are desensitized or attenuated in various developmental processes.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.