Fig 1.
The KAI2 gene underwent duplication prior to diversification of the legumes.
Phylogenetic tree of KAI2 and D14 rooted with bacterial RbsQ from indicated species (Lotus japonicus; Glycine max; Pisum sativum; Medicago truncatula; Arabidopsis thaliana; Populus trichocarpa; Oryza sativa; Zea mays; Sorghum bicolor; Marchantia polymorpha). MEGAX was used to align the protein sequences with MUSCLE and generate a tree inferred by Maximum Likelihood method [72]. The tree with the highest log likelihood (-7359.19) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Values below 50 were ignored. KAI2 duplication in the legumes is highlighted by red and blue branches.
Fig 2.
Lotus japonicus D14, KAI2a and KAI2b can replace D14 and KAI2 in Arabidopsis, respectively.
(A-B) Hypocotyl length of A. thaliana wild-type (Ler), kai2-2 and kai2-2 lines complemented by AtD14, AtKAI2, LjD14, LjKAI2a and LjKAI2b, driven by the AtKAI2 promoter at 6 days post germination (dpg). Seedlings were grown in 16h light/8h dark periods (n = 37–122). Scale bar = 5mm. (C) Shoots of A. thaliana Col-0 and d14-1, d14-1, with an empty vector (EV) or complemented with AtD14, AtKAI2, LjD14, LjKAI2a and LjKAI2b, driven by the AtD14 promoter at 26 dpg. Scale bar = 10 cm. (C) Rosette branch number at 26 dpg of A. thaliana wild-type (Col-0), d14-1 and d14-1 lines carrying an empty vector (EV) or plasmids containing AtD14, AtKAI2, LjD14, LjKAI2a and LjKAI2b, driven by the AtD14 promoter (n = 24). Letters indicate different statistical groups (ANOVA, post-hoc Tukey test).
Fig 3.
Lotus japonicus KAI2a, KAI2b and rice D14L confer divergent hypocotyl growth responses to KAR1 and KAR2 in Arabidopsis.
(A) Structures of KAR1, KAR2, GR245DS and GR24ent-5DS. (B-C) Hypocotyl length of A. thaliana kai2 mutants complemented with KAI2 from A. thaliana, L. japonicus and rice, after treatment with solvent (Mock), 1 μM KAR1 or KAR2 at 6 dpg. (B) Ler wild-type, kai2-2 and kai2-2 lines complemented with AtKAI2, LjKAI2a and LjKAI2b, driven by the AtKAI2 promoter (n = 33–128). (C) Ler and Col-0 wild-type, htl-2 (Ler), K02821-line transgenic for p35S:OsD14L (Col-0), and two homozygous F3 lines from the htl-2 x K02821 cross [16] (n = 80–138). (D) Hypocotyl length of A. thaliana Col-0 wild-type, d14-1 kai2-2 double mutants, and d14-1 kai2-2 lines complemented with LjKAI2a and LjKAI2b, driven by the AtKAI2 promoter after treatment with solvent (Mock), 1 μM GR245DS or GR24ent-5DS (n = 59–134). (B-D) Seedlings were grown in 8h light/16h dark periods. Letters indicate different statistical groups (ANOVA, post-hoc Tukey test).
Fig 4.
Binding of GR24ent-5DS to LjKAI2a is determined by three amino acids.
(A-B) The ligand-binding cavity regions of LjKAI2a and LjKAI2b proteins after structural homology modelling on the KAI2 crystal structure of A. thaliana [5]. Conserved residues in the cavity that differ between the KAI2a and KAI2b clades, and that are also different between LjKAI2b and AtKAI2, are shown in green. The phenylalanine residue in LjKAI2a, which is changed to tryptophan in LjKAI2b, is shown in violet. The catalytic triad is coloured in red. (C-J) DSF curves of purified SUMO fusion proteins of (C-D) wild-type LjKAI2a and LjKAI2b, and (E-J) versions with swapped amino acids (E-F) LjKAI2aM160,L190, LjKAI2bL161,S191, (G-H) LjKAI2aW157,M160,L190, LjKAI2bF158,L161,S191, (I-J) LjKAI2aW157, LjKAI2bF158 at the indicated concentrations of GR24ent-5DS. The first derivative of the change of fluorescence was plotted against the temperature. Each curve is the arithmetic mean of three sets of reactions, each comprising four technical replicates. Peaks indicate the protein melting temperature. The shift of the peak in LjKAI2a indicates ligand-induced thermal destabilisation consistent with a protein-ligand interaction. Insets plot the minimum value of (-dF/dT) at the melting point of the protein as determined in the absence of ligand (means ± SE, n = 3). Asterisks indicate significant differences to the solvent control (ANOVA, post-hoc Dunnett test, N.S.>0.05, *≤0.05, **≤0.01, ***≤0.001, ****≤0.0001).
Fig 5.
Amino acid swaps reverse sensitivity of LjKAI2a and LjKAI2b to GR24ent-5DS in Arabidopsis hypocotyls.
Hypocotyl length of A. thaliana Col-0 wild-type, d14-1 kai2-2 double mutants, and d14-1 kai2-2 lines complemented with LjKAI2a and LjKAI2b variants driven by the AtKAI2 promoter and after treatment with solvent (Mock), 1 μM GR245DS or GR24ent-5DS. (A) LjKAI2aM160,L190 and LjKAI2aM160,L190,W157 (n = 46–84). (B) LjKAI2bL161,S191 and LjKAI2bL161,S191,F158 (n = 49–102). (A-B) Seedlings were grown in 8h light/16h dark periods Asterisks indicate significant differences versus mock treatment (Welch t.test, *≤0.05, **≤0.01, ***≤0.001, ****≤0.0001).
Fig 6.
Role of D14, KAI2a, KAI2b and MAX2 in shoot and hypocotyl development of Lotus japonicus.
(A) Schematic representation of the L. japonicus D14, KAI2a, KAI2b and MAX2 genes. Black boxes and lines show exons and introns, respectively. LORE1 insertions are indicated by red triangles and EMS mutations by red stars. (B) Shoot phenotype of L. japonicus wild-type and karrikin and strigolactone perception mutants at 8 weeks post germination (wpg). Scale bars: 7 cm. (C) Number of branches and of L. japonicus wild-type, karrikin and strigolactone perception mutants at 7 wpg (n = 12–21). (D) Leaf size of the indicated genotypes at 9 wpg (n = 12–15 plants with an average of 3 leaves). (E) Hypocotyl length of the indicated genotypes of L. japonicus under short day conditions (8h light/16h dark) at 1 wpg (n = 79–97). (C-E) Letters indicate different statistical groups (ANOVA, post-hoc Tukey test).
Fig 7.
Lotus japonicus root system architecture is affected specifically by treatment with KAR1 but not KAR2.
(A) Primary root length (PRL), post-embryonic root (PER) number and PER density of wild-type plants 2 wpg after treatment with solvent (M) or three different concentrations of KAR1, KAR2 or rac-GR24 (GR24) (n = 32–57). (B) PER density of wild-type plants at 2 wpg and treated with solvent (Mock) 1 μM KAR1, 1 μM KAR2, or 1 μM rac-GR24 (n = 43–51). Plants were transferred onto fresh hormone-containing medium after 5 days. (C-D) RT-qPCR-based expression of DLK2 normalized to Ubiquitin expression in roots at 2 wpg after 2 hours treatment with solvent (Mock), (C) 1 μM KAR1 and 1 μM KAR2, (D) 1 μM rac-GR24 (n = 4). (A and C) Letters indicate different statistical groups (ANOVA, post-hoc Tukey test). (B) Asterisks indicate significant differences (ANOVA, Dunnett test, N.S.>0.05, *≤0.05). (D) Asterisk indicate significant differences versus mock treatment (Welch t.test, *≤0.05, **≤0.01, ***≤0.001).
Fig 8.
LjKAI2a and LjKAI2b operate redundantly in the response of roots to KAR1 (A) Post-embryonic-root (PER) density of L. japonicus plants, 2 wpg after treatment with solvent (M) or 3 μM KAR1 (n = 34–72). Letters indicate different statistical groups only for non-treated mutant roots (ANOVA, post-hoc Tukey test). (B-C) RT-qPCR-based expression of DLK2 in roots of L. japonicus plants at 2 wpg after 2 hours treatment with solvent (Mock) or (B) 3 μM KAR1 or (C) 1μM GR24ent-5DS. Expression values were normalized to those of the housekeeping gene Ubiquitin (n = 3–4). (A-C) Asterisks indicate significant differences versus mock treatment (Welch t.test, *≤0.05, **≤0.01, ***≤0.001).
Fig 9.
L. japonicus KAI2a and KAI2b display organ-specific redundancy and differ in their ligand-binding specificity.
(A) LjKAI2a is required to mediate inhibition of hypocotyl growth in response to KAR1 and KAR2. In roots LjKAI2a and LjKAI2b redundantly promote postembryonic root density, but only in response to KAR1 treatment. (B) In the Arabidopsis kai2-2 background LjKAI2a mediates hypocotyl growth inhibition in response to KAR1, KAR2 and GR24ent-5DS. In the same background, LjKAI2b mediates a stronger response to KAR1 than to KAR2 and no response to GR24ent-5DS (indicated by a red cross). Three divergent amino acids at the binding pocket are indicated in white. (C) Swapping the three divergent amino acids in the binding pocket reconstitutes GR24ent-5DS activity through LjKAI2b and abolishes GR24ent-5DS activity through LjKAI2a. Among the three amino acids F157/W158 are decisive for GR24ent-5DS binding (strong colors), while L160/M161 and S190/L191 play a weaker role (pale colors). Amino acids from LjKAI2a have a red/pale red and amino acids from LjKAI2b a violet/pale background.