Gain-of-Function Mutants of the Cytokinin Receptors AHK2 and AHK3 Regulate Plant Organ Size, Flowering Time and Plant Longevity

Abstract

The phytohormone cytokinin is a regulator of numerous processes in plants. In Arabidopsis (Arabidopsis thaliana), the cytokinin signal is perceived by three membrane-located receptors named ARABIDOPSIS HISTIDINE KINASE2 (AHK2), AHK3, and AHK4/CRE1. How the signal is transmitted across the membrane is an entirely unknown process. The three receptors have been shown to operate mostly in a redundant fashion, and very few specific roles have been attributed to single receptors. Using a forward genetic approach, we isolated constitutively active gain-of-function variants of the AHK2 and AHK3 genes, named repressor of cytokinin deficiency2 (rock2) and rock3, respectively. It is hypothesized that the structural changes caused by these mutations in the sensory and adjacent transmembrane domains emulate the structural changes caused by cytokinin binding, resulting in domain motion propagating the signal across the membrane. Detailed analysis of lines carrying rock2 and rock3 alleles revealed how plants respond to locally enhanced cytokinin signaling. Early flowering time, a prolonged reproductive growth phase, and, thereby, increased seed yield suggest that cytokinin regulates various aspects of reproductive growth. In particular, it counteracts the global proliferative arrest, a correlative inhibition of maternal growth by seeds, an as yet unknown activity of the hormone.

Figure 1.

The rock2 and rock3 mutations suppress the CKX1ox phenotype. A, Morphology of wild-type (WT), CKX1ox, rock2 CKX1ox, and rock3 CKX1ox plants at the rosette stage. Plants were grown for 25 d under long-day conditions. B, rock2 CKX1ox and rock3 CKX1ox seedlings have larger cotyledons than wild-type and CKX1ox seedlings. The photographs were taken 10 d after germination. C, Adult phenotypes of 46-d-old wild-type, CKX1ox, rock2 CKX1ox, and rock3 CKX1ox plants. D, Flowers of wild-type, CKX1ox, rock2 CKX1ox, and rock3 CKX1ox plants (from left to right). E and F, Flowering time of wild-type and mutant plants grown under long-day (E) or short-day (F) conditions. Crosses indicate that no transition to flowering occurred (n = 20). DAG, Days after germination. G, Relative leaf chlorophyll content of the fourth and fifth leaves of 3-week-old soil-grown plants after 7 d in the dark. Chlorophyll content before the start of dark incubation was set to 100% (n = 3). H, Root elongation of seedlings between day 3 and day 9 after germination (n ≥ 25). Error bars represent sd. The statistical significance of differences was calculated by two-way ANOVA. *, P < 0.01 compared with the wild type; °, P < 0.05 compared with CKX1ox plants.

Figure 2.

rock2 and rock3 mutations alter cytokinin sensitivity. A, Real-time quantitative (q)PCR analysis of CKX1 transcript levels in seedlings grown in vitro for 10 d. Expression values were normalized to PP2AA2, and expression in the wild type (WT) was set to 1. Values are averages of three biological replicates ± se. B, Total cytokinin (CK) contents of 2-week-old seedlings. Contents of individual cytokinin metabolites are shown in Supplemental Table S1. Error bars represent sd. *, P < 0.05 compared with the wild type as calculated by two-way ANOVA. FW, Fresh weight. C, Phenotypes of 14-d-old seedlings grown on medium without (−CK) or with 25 nm benzyladenine (BA; +CK). rock2 CKX1ox and rock3 CKX1ox develop pale yellow leaves and show reduced shoot growth.

Figure 3.

rock2 and rock3 are novel gain-of-function alleles of the AHK2 and AHK3 genes. A, Two segments of the sequence alignment between the cytokinin receptor proteins AHK2, AHK3, and AHK4. The amino acid residues that are mutated (shown in red) in rock2 and rock3 are conserved in all three receptors. Full-length AHK protein sequences were aligned using ClustalW. B, Schematic representation of the AHK2 and AHK3 protein domains and the positions of amino acid substitutions (red arrows) corresponding to the rock2 and rock3 mutations. TM, Transmembrane. C to J, Genetic complementation of the CKX1ox phenotype by AHK2:rock2 and AHK3:rock3. Two independent transgenic AHK2:rock2 (G and H) and AHK3:rock3 (I and J) lines in the CKX1ox background are shown in comparison with the wild type (C), CKX1ox (D), rock2 CKX1ox (E), and rock3 CKX1ox (F) at 18 d after germination.

Figure 4.

The rock2 and rock3 genes code for constitutively active receptors. A, Suppression of the lethal sln1∆ mutation by His kinase activity of AHK4^rock2 and AHK4^rock3. The growth of the sln1∆ yeast strain expressing AHK4^rock2 or AHK4^rock3 was independent of the presence of cytokinin. Error bars represent sd (n = 3). B, Relative expression levels of ARR5 in the rock mutants compared with wild type (WT). qPCR analysis was performed using 5-d-old seedlings grown in vitro. Values are averages of three biological replicates ± se. The statistical significance of differences compared with the wild type was calculated by ANOVA. *, P < 0.01. C to K, Histochemical analysis of ARR5:GUS activity in the wild-type (C–E), rock2 (F–H), and rock3 (I–K) backgrounds. Images show whole seedlings and shoot apices of 5-d-old plants stained overnight and primary root tips of 7-d-old plants after 30 min of staining (from left to right).

Figure 5.

Enhanced shoot growth of rock2 and rock3 mutants and transgenic plants. A and B, Height of the main inflorescence stem after the termination of flowering (50 d after germination). C and D, Primary inflorescence stems 3 cm above the rosette (C) and their diameter (D). Bar in C = 2 mm. The statistical significance of differences in B and D compared with the wild type (WT) was calculated by ANOVA. *, P < 0.001. Error bars represent sd. E, Stem sections of wild-type and AHK3:rock3 plants at the base of primary inflorescence stems. Sections were stained with Toluidine Blue.

Figure 6.

Leaf phenotypes of rock2 and rock3 mutants and transgenic plants. A, Cotyledon size of 5-d-old seedlings. Bar = 1 mm. B, Cotyledons and rosette leaves in the order of appearance (from left to right) at 24 d after germination. Bar = 1 cm. C, Fresh weight of rosette leaves at 32 d after germination. D, Average size of abaxial epidermal cells of the sixth rosette leaf of wild-type (WT) and rock2 plants (n = 14). E, Leaf area of the sixth fully grown rosette leaf. The statistical significance of differences compared with the wild type was calculated by ANOVA. *, P < 0.001.

Figure 7.

Natural and dark-induced leaf senescence. A, Age-dependent senescence phenotypes of the sixth leaf of the wild type (WT) and rock2/rock3 mutants grown under long-day conditions starting at 16 d after leaf emergence (DAE). B, F_v/F_m of the sixth leaf at the time points shown in A. C, Dark-induced senescence in a detached leaf assay. The chlorophyll content of the sixth leaf was examined after 7 d in the dark. The leaf chlorophyll content before the start of dark incubation was set at 100% for each genotype tested (n = 10). Error bars represent sd. The statistical significance of differences compared with the wild type was calculated by ANOVA. *, P < 0.001.

Figure 8.

rock2 and rock3 positively regulate flowering time and flower size. A, Number of rosette leaves at the start of flowering of plants grown under long-day conditions. B, rock2 and rock3 mutants and transgenic plants flower longer than the wild type (WT). Shown are days until the termination of flowering (n = 10). C, Flowers of rock2 and rock3 mutants and transgenic lines compared with the wild type. Bar length is 500 μm. D, Average size of abaxial epidermal cells of petals at stage 13 (Smyth et al., 1990; n = 5). E, Number of siliques on the main stem. Siliques, including unfilled and partially filled siliques, were counted after the end of flowering (n = 15). F, Seed yield of rock2 and rock3 mutants and transgenic lines compared with the wild type. The seed yield of the wild type was set to 100%. Error bars represent sd. The statistical significance of differences compared with the wild type was calculated by Student’s t test. *, P < 0.005.

Figure 9.

The rock2 and rock3 mutants have reduced root systems. A, Elongation of primary roots between day 4 and day 12 after germination (n = 20). B, Root meristems of the wild type (WT), rock2, and rock3. Roots were analyzed 5 d after germination. White and black arrowheads indicate, respectively, the quiescent center and the start of the transition zone. Bars = 50 µm. C, Number of cortex cells between the quiescent center and the start of the transition zone (n = 10). D, Number of lateral roots at 12 d after germination (n = 20). Error bars represent sd. The statistical significance of differences compared with the wild type was calculated by Student’s t test. *, P < 0.005.

Figure 10.

Model of conformational changes associated with transmembrane signaling following cytokinin perception. The schematic topology of the sensory module of cytokinin receptors is based on the crystal structure of AHK4 (Hothorn et al., 2011). The N-terminal helices (α1, α2, and its neighboring 3₁₀-helix) of the CHASE domain are shown in orange, and the two PAS domains are depicted schematically. The model predicts that cytokinin binding causes reversible conformational changes (double-headed arrows), causing a piston-type displacement of the different subdomains and ultimately resulting in the transmission of the signal across the membrane. rock2 and rock3 mutations (arrows) are predicted to mimic those changes locking the receptor in a constitutively active conformation.