Cytokinins and the CRE1 receptor influence endogenous gibberellin levels in Medicago truncatula

ABSTRACT

Gibberellins (GAs) and cytokinins (CKs) are hormones that play antagonistic roles in several developmental processes in plants. However, there has been little exploration of their reciprocal interactions. Recent work in Medicago truncatula has revealed that GA signalling can regulate CK levels and response in roots. Here, we examine the reciprocal interaction, by assessing how CKs and the CRE1 (Cytokinin Response 1) CK receptor may influence endogenous GA levels. Real-Time RT-PCR analyses revealed that the expression of key GA biosynthesis genes is regulated in response to a short-term CK treatment and requires the CRE1 receptor. Similarly, GA quantifications indicated that a short-term CK treatment decreases the GA₁ pool in wild-type plants and that GA levels are increased in the cre1 mutant compared to the wild-type. These data suggest that the M. truncatula CRE1-dependent CK signaling pathway negatively regulates bioactive GA levels.

KEYWORDS:

• Cytokinin
• gibberellin
• legume
• Medicago truncatula

Gibberellin (GA) and cytokinin (CK) hormones are known to play antagonistic roles in several plant developmental processes, and some studies have reported development-dependent interactions between these two hormonal pathways.¹ In Arabidopsis thaliana seedlings, the up-regulation of GA level and/or GA responses was reported to inhibit CK signaling,² and similarly, GA signaling negatively regulates CK levels and/or signaling in A. thaliana and Medicago truncatula roots.^3,4 However, the reciprocal regulation of endogenous GA levels by CKs is less clear. In A. thaliana inflorescences, semi-quantitative RNA gel blot analyses suggested that CKs had no effect on the expression of a GA-responsive gene or a GA biosynthesis gene.² However, in the Arabidopsis shoot apical meristem, a CK-treatment induced the expression of AtGA2oxidase (GA2ox) gene, which encodes an enzyme that deactivates bioactive GA,⁵” and a genome-wide expression analysis reported the downregulation of several GA biosynthesis genes after a short-term CK treatment in whole Arabidopsis seedlings.⁶ Similarly, transcriptomic studies performed in M. truncatula roots revealed a significant change in the expression of GA biosynthesis and response genes after short-term CK treatment,⁷ suggesting that CKs may suppress endogenous GA levels. However, no study has examined the actual changes in endogenous GA levels in response to CKs. Here, we explored whether CKs and the CRE1 (Cytokinin Response 1) receptor that mediates CK responses⁸ influenced endogenous GA levels in M. truncatula.

Based on transcriptomic data gained in M. truncatula roots,⁷ we selected two genes rapidly regulated by a CK treatment; GA20ox1 (GA₂₀-oxidase) and GA2ox1 (GA₂-oxidase). The GA20ox1 gene is predicted to encode an enzyme that catalyzes the formation of the inactive GA precursor GA₂₀, which can then be converted to bioactive GA₁ by a GA3ox enzyme.⁹ The GA2ox1 gene encodes a putative enzyme that catalyzes the deactivation of GA₁ into inactive GA₈⁹. The expression of these genes was analyzed by real time RT-PCR as previously described in Fonouni-Farde et al.,⁴ in wild-type (WT) and cre1 mutant roots in response to a short-term CK treatment (BenzylAminoPurine [BAP; Sigma]; 3h; 10⁻⁷M) (Fig. 1A). In WT, the expression of GA2ox1 was significantly induced by CKs, whereas the expression of GA20ox1 was conversely repressed. These CK-dependent regulations were abolished in the cre1 mutant (Fig. 1A), indicating that a subset of GA metabolic genes requires the CRE1 receptor to be transcriptionally regulated by CKs.

Figure 1. CKs and CRE1 regulate GA-metabolic genes and endogenous GA levels. (A) Quantification by RT-qPCR of the expression of GA2ox1 and GA20ox1 GA-metabolic genes, previously shown to be regulated by CKs (Ariel et al., 2012), in wild-type (WT) or in cre1 mutant roots after a BenzylAminoPurine (BAP [Sigma]; 3 h; 10^–7 M) treatment. (B) Quantification (in ng/g of Root Fresh Weigh [RFW]) of the bioactive GA1 in WT or cre1 mutant roots. (C-D) Quantification (in ng/g of Shoot Fresh Weigh [SFW]) of the bioactive GA₁ (C) or the precursor GA₂₀ (D) in WT shoots treated or not with BAP (3 h; 10^–7 M). (E-F) Quantification (in ng/g of SFW) of GA₁ (E) or GA₂₀ (F) in WT or cre1 mutant shoots. In (A), transcript levels are normalized relatively to untreated control roots to show fold changes and the dotted line indicates a ratio of 1. Error bars represent standard deviations. Asterisks indicate significant differences compared to the untreated control, based on a Mann-Whitney test (α<0.05). Results are the mean of three biological replicates, each replicate being a pool of 25 plants. In (B-F), error bars represent standard errors of the mean. Letters indicate significant differences based on a Mann–Whitney test (α<0.05). In (B), the GA₁ quantification was based on four biological replicates, each replicate being a pool of at least two plants. In (C-F), GA quantifications were based on at least five biological replicates, each replicate being a pool of at least two plants.

Based on the expression patterns observed in response to the BAP treatment (Fig. 1A), we hypothesized that there may be an increase of bioactive GAs in the cre1 mutant compared to the WT. We therefore quantified GAs in WT and cre1 mutant roots following a protocol derived from Boden et al.¹⁰ (Supp. Methods). Like in the closely related legume pea,¹¹ GA₁ appears to be the predominant bioactive form in M. truncatula, as we could detect GA₁ but not GA₄. A small increase in GA₁ was detected in the roots of the cre1 mutant compared to the WT, although this change was not significant (Fig. 1B).

As GA₁ levels were near the detection limit in M. truncatula roots, we also quantified GAs in shoots of WT plants after a short-term CK treatment (BAP; 3 h; 10⁻⁷M) (Fig. 1C, D) and in the cre1 mutant (Fig. 1E, F). In the WT, the level of the bioactive GA₁ was significantly reduced in response to the CK treatment, whereas a small but not significant increase was observed for the immediate precursor to GA₁, GA₂₀ (Fig. 1C, D). To determine if the CRE1 receptor was involved in the regulation of endogenous GAs level, we quantified GA₁ and GA₂₀ in WT and cre1 shoots (Fig. 1E, F). While the content of GA₁ showed no statistical difference between the WT and the cre1 mutant, the level of GA₂₀ was significantly higher in the cre1 mutant (Fig. 1F). Taken together, these results indicate that CKs and the CRE1 receptor are able to regulate the level of a GA endogenous precursor and/or of a biologically active GA in M. truncatula.

In this study, we show that CKs through the action of the CRE1 receptor influence the levels of GAs in M. truncatula. Together with previous results reporting that GAs can regulate CK levels and response in M. truncatula roots,^4,7 this suggests that there is a bidirectional antagonistic regulation of CK and GA levels that likely contribute to a dynamic equilibrium between these two hormones. As bioactive GAs suppress the accumulation of DELLA proteins,¹² we can speculate that in various developmental contexts the CRE1-dependent CK suppression of GA content may increase the DELLA protein accumulation and therefore modulate the expression of target genes.^13,14

Acknowledgments

The FF laboratory acknowledges the CNRS, the “Agence Nationale de la Recherche” (ANR) Labex “Saclay Plant Science” (SPS), and the Lidex “Plant Phenotyping Pipeline” (3P) for their support. CFF was the recipient of a Paris-Sud University PhD fellowship. The EF laboratory acknowledges the Australian Research Council (ARC) for financial support, Assoc. Prof. Noel W. Davies and Assoc. Prof. John J. Ross for advice with hormone extraction and analysis and Shelley Urquhart for technical assistance.

Additional information

Funding

French National Research Agency (ANR) and Australian Research Council (ARC).