Proteasome-dependent proteolysis has a critical role in fine-tuning the feedback inhibition of cytokinin signaling

Abstract

The ubiquitin/26S proteasome-dependent proteolysis of response regulators is a critical element of many plant hormone signaling pathways. We have recently shown that cytokinin signaling requires the AXR1 component of the related to ubiquitin (RUB) protein modification pathway to promote the proteasome-dependent degradation of the cytokinin response inhibitor ARR5. Here, we show that ARR5 also accumulates in the 26S proteasome mutant rpn12a-1, and leads to a marked resistance to cytokinins. Collectively, these results suggest that proteasome-dependent proteolysis of feedback inhibitors such as ARR5 is essential for the maintenance of optimal responsivity and plasticity in cytokinin signaling.

Cytokinins are plant hormones that regulate cell division, elongation and differentiation, and are therefore essential for every aspect of plant growth and development.^1-5 Due to the importance of cytokinins for the plant life cycle, the cytokinin signal transduction cascade has been a focus of intense studies and was shown to include a two-component system.^6-8 In Arabidopsis, cytokinin binding stimulates a family of three receptor histidine kinases to transfer a phosphoryl group onto a set of histidine phosphotransfer proteins that in turn phosphorylate type-A and type-B Arabidopsis response regulators (Fig. 1). Type-B Arabidopsis response regulators (ARRs) are transcriptional activators, and they promote cytokinin responses.

Figure 1. AXR1 promotes the cytokinin response by suppressing the accumulation of response inhibitors. Cytokinin (CK) is perceived by the AHK receptors leading to their autophosphorylation. The phosphoryl group is then transferred from the AHKs via the AHPs onto the response-activating type-B ARRs and the response-inhibiting type-A ARRs. AXR1, a key enzyme of the RUB protein modification pathway, promotes RUB attachment on the CUL subunit of the CRL class of ubiquitin ligases. RUB modification of an as-yet-unidentified CRL increases its affinity for the type-A ARRs and leads to their polyubiquitination and subsequent degradation by the 26S proteasome.

Similar to other signaling pathways, the cytokinin response pathway includes negative feedback control that limits the intensity and the duration of the response. The negative feedback control depends on the type-A ARRs.^6-8 The activity of these response repressors is upregulated by cytokinins in a phosphorelay-dependent manner by two mechanisms: (1) the type-B ARR-dependent transcriptional activation of type-A ARR genes and (2) the AHP-dependent phosphorylation of type-A ARRs, which leads to their increased stability and accumulation.^6-9

Some—but not all—of the type-A ARRs were shown to be unstable proteins degraded by the proteasome.^5,10 It, however, remained untested if their stabilization indeed causes the cytokinin insensitivity frequently observed in ubiquitin/26S proteasome pathway mutants.^11-14 We have recently shown that type-A ARR5 stabilization is a major cause for the cytokinin resistance of axr1 mutants in which the activity of a key enzyme in the Related to Ubiquitin (RUB) pathway of protein modification is downregulated or abolished.^15-19 Because RUB modification is required for the optimal function of the Cullin Ring Ubiquitin ligases (CRLs), we proposed that loss of AXR1 function led to a decrease in activity of an ARR5-targeting CRL and consequently, to an increase ARR5 abundance and decreased cytokinin sensitivity (Fig. 1).^16,20

An unusual finding described in our AXR1 study is that loss of function of ARR5 was sufficient to suppress significantly the cytokinin resistance of axr1 seedlings, even though the Arabidopsis genome encodes for other unstable type-A ARRs that presumably also accumulate in an axr1 background.^5,9,10,16 This finding accentuates the importance of ARR5 proteolysis for maintaining cytokinin responsiveness, and suggests that aberrant accumulation of ARR5 may also cause the cytokinin insensitivity described for some other ubiquitin/proteasome pathway mutants.^11,12

We have previously shown that the proteasome mutant rpn12a-1 combines decreased cytokinin sensitivity with a modest decrease in ubiquitin-dependent proteolysis, suggesting the partial stabilization of a cytokinin response repressor.^11,21 To test if ARR5 stabilization plays a role in the rpn12a-1 cytokinin phenotype, we introgressed an ARR5 overexpression transgene into the rpn12a-1 background, and analyzed ARR5 accumulation and cytokinin growth responses (Fig. 2). Even though rpn12a-1 seedlings have only a mild decrease in 26S proteasome activity,²¹ the FLAG-ARR5 level was noticeably higher in rpn12a-1 compared with the Col-0 wild type background (1.8 ± 0.2-fold increase; Fig. 2A). This increase in ARR5 abundance was similar to that observed for axr1 seedlings (1.9 ± 0.3).¹⁶ To test if accumulation of ARR5 in the rpn12a-1 background also enhanced the cytokinin insensitivity, we assayed two cytokinin responses, the cytokinin-dependent inhibition of root growth and shoot development (Fig. 2B and C). Indeed, rpn12a-1 seedlings containing FLAG-ARR5 were significantly more resistant to the cytokinin benzyladenine (BA) compared with both rpn12a-1 and the wild type expressing FLAG-ARR5 (Figs. 2B and 2C). Strikingly, the rosettes of rpn12a-1 seedlings expressing FLAG-ARR5 were able to expand on media containing 1 μM BA, a concentration that caused a near complete growth inhibition in rpn12a-1 and FLAG-ARR5 (Col-0). This indicated a strong synergistic effect between the ARR5 overexpression and the loss of proteasome activity on cytokinin sensitivity.

Figure 2. ARR5 accumulates in the rpn12a-1 genetic background. (A) Increased abundance of FLAG-ARR5 in the proteasome mutant rpn12a-1. Membrane probed with anti-ARR5 antibodies was re-probed with anti-tubulin α (TUA) antibodies to illustrate equal loading. The anti-RPN12 antibodies were used to confirm the genetic background. RPN12a-NPTII, a fusion protein generated by the insertion of the activation T-DNA tag in the RPN12a gene.¹¹ (B) Cytokinin inhibition of root elongation growth. Four-day-old seedlings germinated on MS/2 medium were transferred to media containing the indicated BA concentrations, and grown for another 8 d. The root length of untreated seedlings of each genotype was set at 1, and mean values ± SEM (n ≥ 6) are shown as fold induction compared with the respective control. To simplify the graph, we only show the significance of the differences between FLAG-ARR5 ox in the Col-0 and rpn12a-1 backgrounds (*, p < 0.01; ***, p < 0.001, ****, p < 0.0001; two-way ANOVA with Bonferroni's multiple comparison test). (C) Cytokinin inhibition of rosette growth in plants incubated for 14 d on a range of BA concentrations. Three representative plants per line and per treatment are shown.

A conspicuous common thread linking the AXR1 study¹⁶ and the results presented here (Fig. 2) is that a mild increase in ARR5 accumulation (1.9-fold in axr1-3 and 1.8-fold in rpn12a-1) is sufficient to cause a significant change in cytokinin responsiveness. This emphasizes the importance of protein stability control in the negative feedback regulation of cytokinin signaling.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.