UV-B irradiation-activated E3 ligase GmILPAI modulates gibberellin catabolism to increase plant height in soybean

The study addresses how ultraviolet-B (UV-B) light influences plant height via gibberellin (GA) metabolism in soybean, focusing on whether post-translational regulation contributes to GA homeostasis under UV-B. Plant height is a central ideotype trait impacting yield and lodging resistance, with semi-dwarf varieties underpinning the Green Revolution. While many QTLs for soybean height exist, few causal genes are characterized. GA biosynthesis and signaling components (e.g., GA20ox, GA3ox, GA2ox, DELLA) are known to impact height; GA2ox enzymes inactivate GAs and often suppress stem elongation. Environmental cues, including light and UV-B, modulate GA gene transcription, typically reducing GA biosynthesis gene transcripts and increasing GA2ox expression to inhibit elongation. However, post-translational regulation of GA2ox proteins remains largely unexplored. The ubiquitin–proteasome system (UPS), particularly E3 ligases like APC/C, orchestrates selective protein degradation and affects plant development and hormone signaling, yet APC/C’s role in environmental response of growth has not been reported. This work aims to identify genetic components mediating UV-B effects on height and delineate mechanisms linking UV-B to GA catabolism via protein ubiquitination and degradation.

• GA pathway and height: DELLA repressors, GA20ox/GA3ox biosynthesis genes (loss-of-function causes dwarfism), GA2ox catabolic enzymes (overexpression suppresses elongation). Several transcription factors upregulate GA2ox to reduce growth.
• Environmental regulation: UV-B via UVR8/COP1 elevates HY5/HYH, promotes PIF4/5 degradation, accumulates DELLAs, downregulates GA3ox/GA20ox, and upregulates GA2ox transcripts to inhibit elongation. Prior studies emphasize transcriptional changes, not protein-level control of GA2ox.
• UPS and APC/C: E3 ligases (HECT, RING/U-box, SCF, APC/C) confer substrate specificity for proteasomal degradation. Plant APC/C subunits influence growth traits and hormone signaling; rice apc6 mutants are shorter and GA-insensitive. Known plant APC/C substrates include cyclins, DRB4, MOC1, and RAA1, but environmental regulation via APC/C was unreported.
• UV-B morphology and hormones: UV-B acclimation involves GA dynamics; evidence in Arabidopsis links UV-B to GA pathway modulation requiring UVR8, yet protein-level regulation of GA catabolism enzymes was unclear.

• Forward genetic screen and phenotyping: EMS-mutagenized Hedou 12 (H12) soybean screened for dwarfism; mutant Gmuid1 identified and phenotyped (plant height, internode length, cell length, agronomic traits) under field, glasshouse, and controlled light (white with/without UV-B) conditions.
• Genetic mapping: F2 population from Williams 82 × Gmuid1; segregation analysis (3:1, χ2 P=0.89), INDEL marker mapping to 4.5 Mb on chr11, fine-mapping to 68 kb between SSR markers Gm1015–Gm1016 using 2250 F2/F3. Bulk segregant analysis (40 WT-like and 40 mutant F2s; 40× coverage) identified a splicing-site SNP in Glyma.11G026400 (GmILPA1). Complementation via 35S:GmILPA1-GFP in gmilpa1-2 background.
• Growth and treatments: Controlled white light (600 µmol m−2 s−1) with/without UV-B (1.5 µmol m−2 s−1, 12 h photoperiod). GA3 application (100 µM), PAC treatments, and UV-B exposure regimes.
• Protein interaction assays: Yeast two-hybrid (Y2H) screens against soybean cDNA; validation by Y2H, bimolecular fluorescence complementation (BiFC), and co-immunoprecipitation (Co-IP) in Nicotiana benthamiana for GmILPA1, GmGA2ox-like, and GmUBL1. Subcellular localization via transient expression.
• UV-B effect on interactions: BiFC quantification of cytoplasmic YFP signal ± UV-B (1 h, 21 µmol m−2 s−1); Co-IP after 12 h UV-B to assess interaction strength.
• Ubiquitination assays: In vivo ubiquitination using FLAG-Ub and GmGA2ox-like-GFP co-expression, immunoprecipitation with anti-GFP and detection with anti-FLAG/anti-Ub. Site-directed mutagenesis of GmGA2ox-like lysines (K394R, K407R, double) to identify ubiquitination sites. In vitro APC/C ubiquitination using APC/C immunoprecipitated from 35S:GmILPA1-GFP plants, recombinant MBP-GmGA2ox-like substrate, and commercial E1/E2/Ub. P62-agarose enrichment of ubiquitinated proteins from transgenic soybean expressing 35S:GmGA2ox-like-GFP.
• Protein stability/degradation: Western blot of endogenous GmGA2ox-like in WT and gmilpa1-2 under white vs white+UV-B; time-course after transfer to UV-B; cell-free degradation assays using extracts ± UV-B and MG132; degradation assays in N. benthamiana extracts expressing GmILPA1 and/or GmUBL1.
• GA measurements: LC-ESI-MS/MS quantification of GA1, GA4, precursors (GA20, GA9/GA7), and catabolites (GA8, GA34) from apical buds of V2 seedlings ± UV-B.
• Gene expression and protein accumulation: RT-qPCR for GmILPA1 after GA3 or UV-B; immunoblot time-course for GmILPA1 accumulation.
• Functional genetics: RNAi knockdown of GmGA2ox-like in gmilpa1-2 and phenotypic rescue under UV-B.
• Haplotype and selection analyses: 444-accession panel (87 wild, 158 landrace, 199 cultivars); 10 SNPs + 1 InDel across 5.7 kb around GmILPA1 to define six haplotypes; selection scans (FST, π, Tajima’s D) across 100 kb; phylogeny and haplotype network. Geographic distribution analysis across seven Chinese regions.
• Promoter activity assays: Transient ProGmILPA1-Hap1–Hap6:GUS and Hap5/indel constructs in N. benthamiana; histochemical and fluorometric GUS ± UV-B (1.5 µmol m−2 s−1, 12 h). Statistics by Student’s t-test and one-way ANOVA with Tukey comparisons.

• Identification of causal gene: The dwarf mutant Gmuid1 (renamed gmilpa1-2) maps to Glyma.11G026400 (GmILPA1), encoding an APC8-like APC/C subunit. A G→A SNP at the first intron splice site causes aberrant splicing and premature termination. F2 segregation fits a 3:1 ratio (153 WT:48 mutant; χ2 P=0.89). Overexpression of GmILPA1 in gmilpa1-2 rescues height.
• UV-B dependency: gmilpa1-2 shows no dwarfism in glasshouse without UV-B but is significantly shorter under white+UV-B; UV-B is essential for the mutant phenotype (P=2.42×10−6 and 1.80×10−8 vs WT for gmilpa1-2 and gmilpa1, respectively).
• Protein interactions: GmILPA1 interacts with GmGA2ox-like (a functional GA 2-oxidase) and with GmUBL1 (a RING/U-box E3) in Y2H, BiFC, and Co-IP. GmGA2ox-like also interacts with GmUBL1.
• UV-B enhances interaction: UV-B increases GmILPA1–GmGA2ox-like interaction (BiFC cytoplasmic YFP intensity ~3-fold higher with UV-B; P=1.30×10−7) and strengthens Co-IP signal; UV-B does not affect interactions involving GmUBL1.
• Ubiquitination and degradation: GmGA2ox-like is polyubiquitinated in vivo. K394 is critical for ubiquitination: K394R (and K394R/K407R) abolishes Ub smear; K407R retains smear. In vitro APC/C ubiquitination assays show GmILPA1 promotes GmGA2ox-like ubiquitination; K394R substrate is not ubiquitinated. In planta, co-expression with GmILPA1 or GmUBL1 enhances GmGA2ox-like polyubiquitination. P62-agarose enrichment confirms ubiquitinated GmGA2ox-like in soybean.
• UV-B–dependent degradation: Endogenous GmGA2ox-like protein decreases under UV-B in WT but not in gmilpa1-2. Time-course shows degradation in WT begins ~10 h after UV-B; no reduction in mutant. Cell-free assays reveal UV-B–dependent degradation in WT extracts starting ~60 min; blocked by MG132; absent in gmilpa1-2 extracts.
• Role of GmUBL1: GmUBL1 alone ubiquitinates but does not directly degrade GmGA2ox-like in extracts; it enhances GmILPA1-mediated degradation (degradation apparent 90–150 min when both E3s present), indicating cooperative action.
• GA physiology: Exogenous GA3 (100 µM) restores gmilpa1-2 height to WT under sunlight; PAC suppresses growth in all genotypes; GA3 rescues PAC or UV-B-induced dwarfism. Under UV-B, both genotypes show decreased GA levels overall, but WT maintains higher active GA1 and GA4 (and precursors GA20/GA7) than gmilpa1-2, while non-bioactive GA8 and GA34 accumulate more in gmilpa1-2, indicating greater depletion of active GA in the mutant (significant P-values across GA measurements; e.g., GA1 P=6×10−8).
• Regulation of GmILPA1: GA3 and UV-B induce GmILPA1 protein accumulation within 12 h; GmILPA1 transcripts increase with GA3 but not with UV-B.
• Genetic rescue: RNAi knockdown of GmGA2ox-like in gmilpa1-2 significantly increases plant height under UV-B, rescuing the mutant phenotype (ANOVA P-values ≤9.29×10−5).
• Domestication and haplotypes: Selection signals (elevated FST, reduced π, Tajima’s D shifts) across a 100-kb region around GmILPA1 indicate artificial selection. Six major haplotypes identified; Hap1 and Hap5 enriched in cultivars, Hap2 and Hap6 prevalent in landraces, Hap3/Hap4 specific to wild soybeans. Frequencies of Hap1 and particularly Hap5 increased during domestication.
• Promoter INDEL and UV-B response: Hap5 uniquely carries a 13-bp deletion (Indel-665) in a light-responsive cis-element in the GmILPA1 promoter. Transient GUS assays show lower UV-B-induced promoter activity for Hap5 vs Hap1/2/6; restoring the deleted element in Hap5 (Hap5/indel) increases UV-B responsiveness, implicating the INDEL as causal for reduced transcription under UV-B.
• Agronomic associations and geography: Accessions with Hap1/2/6 are taller than Hap5 under natural conditions; Hap5 tends to have fewer pods but higher hundred-seed and grain weight per plant, indicating yield potential. Geographically, Hap5 occurs in northern China (higher UV-B), while Hap1/Hap6 increase toward southern regions. Upon UV-B exposure, GmILPA1 transcripts increase more in Hap1 than Hap5; Hap1 accessions maintain greater height under UV-B than Hap5.
• Model: Under UV-B, GmILPA1 accumulates and, with GmUBL1, ubiquitinates and promotes degradation of GmGA2ox-like, mitigating UV-B-induced reductions in active GA to sustain growth. In Hap5, reduced promoter responsiveness to UV-B limits GmILPA1 transcriptional induction, contributing to shorter height under UV-B.

This work demonstrates that the APC/C subunit GmILPA1 is an environmentally responsive E3 ligase component that safeguards plant height under UV-B by post-translationally downregulating the GA catabolic enzyme GmGA2ox-like. By promoting GmGA2ox-like ubiquitination and UV-B–dependent proteasomal degradation, GmILPA1 counteracts the UV-B-triggered decline in active GAs, maintaining elongation growth. The discovery that K394 in GmGA2ox-like is a critical ubiquitination site and that another E3 ligase, GmUBL1, augments GmILPA1-mediated degradation, reveals a cooperative E3 network fine-tuning GA catabolism. These findings provide, to the authors’ knowledge, the first evidence that APC/C-mediated proteostasis integrates environmental UV-B signals with hormone metabolism to control plant architecture. The genetic and haplotype analyses connect molecular mechanism to breeding outcomes: strong artificial selection at GmILPA1 yielded haplotypes with distinct UV-B transcriptional responsiveness. The Hap5 promoter INDEL weakens UV-B-induced GmILPA1 expression, aligning with shorter height yet potential yield advantages (higher hundred-seed and grain weight) and prevalence in high-UV regions, suggesting local adaptation and ideotype utility. Exogenous GA3 phenotypically rescues the dwarfism of gmilpa1-2 under UV-B, corroborating the mechanistic link to GA homeostasis. Together, the results answer the central question by establishing a UV-B–activated APC/C pathway that modulates GA catabolism through targeted degradation of GA2ox to regulate height.

• Main contributions: Identified GmILPA1 (APC8-like) as a positive regulator of plant height under UV-B; established GmGA2ox-like as a novel APC/C substrate whose degradation is UV-B dependent; showed cooperative action of GmUBL1 with GmILPA1 to enhance GmGA2ox-like degradation; linked mechanism to GA homeostasis and growth outcomes; revealed domestication-driven selection of GmILPA1 haplotypes, with a promoter INDEL (Hap5) diminishing UV-B-responsive transcription and associating with shorter height and regional adaptation.
• Implications: Demonstrates environmental control of hormone catabolism via APC/C, offering targets for molecular breeding to optimize plant architecture and UV-B tolerance. Hap5 and promoter editing (e.g., base or INDEL editing at the light-responsive cis-element) present strategies to tailor height and yield traits under varying UV-B environments.
• Future directions: Define the upstream UV-B signaling components that induce GmILPA1 protein accumulation; dissect how UVR8/COP1 pathways intersect with APC/C activity; resolve the biochemical mechanism by which GmUBL1 enhances APC/C-mediated degradation; explore broader APC/C substrates in GA metabolism and other hormones; validate haplotype performance across environments and integrate GmILPA1 edits into breeding programs.

• Mechanistic upstream gap: While UV-B increases GmILPA1 protein levels and promotes its interaction with GmGA2ox-like, the upstream signaling cascade (e.g., direct links to UVR8/COP1 or post-translational modifications of GmILPA1/APC/C) was not delineated.
• Scope of substrates: The study focuses on GmGA2ox-like; additional APC/C targets within GA metabolism or other pathways may contribute to the phenotype but were not comprehensively surveyed.
• Context and generalizability: Many protein interaction and ubiquitination assays used N. benthamiana transient systems; although corroborated in soybean, broader validation across tissues, stages, and field UV-B spectra would strengthen generalizability.
• Dosage and pleiotropy: GA3 rescued height but not all morphological traits, suggesting hormone dose-sensitivity and additional roles of GmILPA1 (e.g., petiole angle) not fully resolved mechanistically.
• Haplotype-functional link: The promoter INDEL’s effect was inferred from transient assays; in planta genome-edited isogenic lines validating the causal impact on UV-B responsiveness and field performance were not presented.