Stabilization of dimeric PYR/PYL/RCAR family members relieves abscisic acid-induced inhibition of seed germination

Abstract

Abscisic acid (ABA) is the primary preventing factor of seed germination, which is crucial to plant survival and propagation. ABA-induced seed germination inhibition is mainly mediated by the dimeric PYR/PYL/RCAR (PYLs) family members. However, little is known about the relevance between dimeric stability of PYLs and seed germination. Here, we reveal that stabilization of PYL dimer can relieve ABA-induced inhibition of seed germination using chemical genetic approaches. Di-nitrobensulfamide (DBSA), a computationally designed chemical probe, yields around ten-fold improvement in receptor affinity relative to ABA. DBSA reverses ABA-induced inhibition of seed germination mainly through dimeric receptors and recovers the expression of ABA-responsive genes. DBSA maintains PYR1 in dimeric state during protein oligomeric state experiment. X-ray crystallography shows that DBSA targets a pocket in PYL dimer interface and may stabilize PYL dimer by forming hydrogen networks. Our results illustrate the potential of PYL dimer stabilization in preventing ABA-induced seed germination inhibition.

Fig. 1. Computational design of DBSA targeting a binding pocket at the PYL dimer interface.

a An extended binding pocket was discovered in the gate-open apo-PYL1 dimer interface (PDB code: 3KAY). This binding pocket partly overlaps with the ABA-binding pocket in the gate-closed PYR1 dimer (PDB code: 3K3K). b The binding pockets differ due to conformational changes in the gate loop and latch loop. The gate loop and latch loop transitions to closed state from open state after ABA binding. c Ser112 at gate loop as well as Pro115 and His142 at latch loop of PYL1 along with their homologous residues in PYR1 (Ser85, Pro88, and His115, respectively) exhibit conformational changes during conformation transition. d Fragment screening was performed to discover chemical probes targeting PYL1 dimer interface. Structural optimization and molecular docking were performed on pyrabactin to suit gate open and latch open PYL1 (PDB code: 3KAY) and PYL2 (PDB code: 3NR4), and find a conformation towards dimer interface. NBSA exhibited the lowest binding free energy. Based on structural optimization, DBSA showed the largest improvement of binding free energy in structural optimization. e NBSA binds to the pocket at the dimer interface and forms several hydrogen bonds with the PYL1 dimer. f DBSA forms an intramolecular hydrogen bond that stabilizes its conformation and forms a series of hydrogen bonds with residues at the PYL1 dimer interface. The hydrogen bonds are shown by red dotted lines.

Fig. 2. DBSA shows low ABA receptor binding affinity and is a potent antagonist of multiple ABA receptors.

a The K_d values of ABA and DBSA to PYL1 using ITC. b The K_d values of ABA and DBSA to PYR1 using ITC. c Antagonistic effect of DBSA on HAB1 activity through a phosphatase assay. PYLs and HAB1 were present at a molar ratio of 1:1 (0.4 μM: 0.4 μM) for PYR1/PYL1/PYL2/PYL3 and 2:1 (0.8 μM: 0.4 μM) for PYL5/PYL6/PYL10. Various PYL-HAB1 combinations were incubated with the indicated chemicals (5 μM ABA, 50 μM DBSA or 5 μM ABA and 50 μM DBSA). n = 3 biologically replicates. d Antagonistic effect of various concentrations of DBSA on HAB1 activity through different PYL members. EC₅₀ values were obtained by nonlinear fits of dose-response data. DBSA were tested at 0 μM to 200 μM, and the concentration of ABA was 5 μM. The concentration of PYR1, PYL5 and PYL10 were 0.4 μM, 0.8 μM and 0.8 μM, respectively, while HAB1 proteins were used at the molar ratio of 0.4 μM. n = 3 biologically replicates. For c, d, the data are presented as the mean ± SD. For (c), the line within the box marks the median and the whiskers represent the minimum and maximum values. Source data are provided as a Source Data file.

Fig. 3. DBSA relieves ABA-induced inhibition of seed germination and seedling growth, and does not induce ABA-responsive gene expression in Arabidopsis thaliana.

a Germination rate of seeds exposed to ABA (1 µM), DBSA (1 µM) or ABA and DBSA (1 µM:1 µM). DMSO (0.05%) was used as a control. PYL dimer quadruple deletion mutant (pyr1/pyl1/pyl2/pyl4, 1124), PYL monomer multiple deletion mutant (pyl3/pyl7/pyl9/pyl11/pyl12, 3791112), PYL4 overexpression (4OE) seeds are considered germinated when green cotyledons expand (n = 3 biologically replicates). b Seeding growth of WT and 1124 seeds under the treatment of ABA (1 µM), DBSA (1 µM) or ABA and DBSA (1 µM:1 µM). n = 3 biologically replicates. c DBSA and ABA treatment caused different gene expression patterns. d The transcript levels of the common DBSA- and ABA-responsive genes were poorly correlated. The scatter plot shows the log2-transformed expression levels of DBSA-responsive DEGs (y-axis) and ABA-responsive DEGs (x-axis) relative to the DMSO control. e The representative GO terms and pathways enriched in ABA-specific response and DBSA-specific response DEGs based on functional enrichment analysis (p < 0.01). f Induction of abiotic stress marker genes in Arabidopsis thaliana (Col-0) after ABA or DBSA treatment as determined by qRT-PCR. Ten-day-old seedlings were treated with 10 μM ABA or 100 μM DBSA or 10 μM ABA and 100 μM DBSA for 6 h before RNA extraction, and 0.05% DMSO was used as control (n = 3 biologically replicates). For a and f, the data are presented as the mean ± SD, the line within the box marks the median and the whiskers represent the minimum and maximum values. For f, P values are indicated by two-tailed Student’s t test. Source data are provided as a Source Data file.

Fig. 4. DBSA targets PYL1 dimer interface and stabilizes the PYL1 dimer via hydrogen bond networks.

a Crystal structure of PYL1 with DBSA (PDB code 9J6I, green) and ABA (PDB code 3JRS, yellow). ABA induced a gate-close and latch-close conformation, which DBSA bind to a DBSA induced a gate-open and latch- open conformation. b PYL1 was maintained in the latch-open and gate-open state in the crystal structure (shown in green stick model). DBSA forms hydrogen bonds with the side chain of Arg143. DBSA was predicted to bind to latch-open and gate-open PYR1 and interact with R116 (shown in pink stick model). c Alanine mutation result showed that DBSA may form strong interaction with R116 of PYR1. d The aggregation states of PYR1 dimer induced by ABA, DBSA or ABA/DBSA co-treatment were detected by SEC-MALLS. e The motion trends of ABA-PYR1 and DBSA-PYL1 complexes. f The dominant conformations were determined using FEL analysis. g The hydrogen networks mediated by ABA of PYR1 or DBSA of PYL1 in the dimer interfaces.

Fig. 5. Mode of action of DBSA compared with those of reported ABA receptor agonists and antagonists.

a Binding mode of different ligands with PYLs. PYL1-DBSA (PDB code 9J6I), PYR1-AS6 (PDB code 3WG8), PYL10-antabactin (PDB code 7MLD), PYR1-pyrabactin (PDB code 5UR4), PYL2-quinabactin (PDB code 4LA7) and PYL10-3CB (analog of opabactin, PDB code 6NWC). PYR1-pyrabactin, PYL2-quribaction, PYL10−3CB, PYL10-antabactin and PYL5-AS6 are in latch-closed and gate-closed conformations, whereas PYL1-DBSA is in latch-open and gate-open conformation. For the latch-closed and gate-closed agonist binding conformations, a conserved Trp lock of PP2Cs were insert into the PYL pockets. b Mode of action of ABA receptor agonists, ABA-mimic receptor antagonists AS6, antabactin and DBSA. ABA receptor agonists cause gate-closed conformations and PYL dimer dissociation, which inhibits PP2Cs. ABA-mimic receptor antagonists AS6 and antabactin also cause gate-closed conformations and PYL dimer dissociation, but AS6 obstructs the interaction between PP2C and PYLs by occupying the 3’ tunnel, and antabactin blocks the conserved Trp lock of HAB1 to PYL. DBSA stabilizes the PYL dimer, which results in PP2C activation.