2026/03/15 by Huixue Dong, Danning Yang, Yu He +6 · 1 voice
Agricultural and Biological Sciences · #Plant nutrient uptake and metabolism #Seed Germination and Physiology #Wheat and Barley Genetics and Pathology
paper · doi:10.1111/jipb.70234
openalex publication_date 2026/03/15 · openalex created_date 2026/03/17 · openalex updated_date 2026/07/28
GLYCOGEN SYNTHASE KINASE 3 (GSK3), a negative regulator of brassinosteroid signaling, phosphorylates and stabilizes the seed dormancy protein DOG1L4, which in turn activates ABSCISIC ACID INSENSITIVE 5 (ABI5) to enhance seed dormancy. GSK3 also phosphorylates ABI5. This dual-target mechanism suggests a strategy for controlling seed dormancy and germination in crops. Pre-harvest sprouting (PHS) is a major challenge in global wheat production, leading to premature seed germination and reducing both yield and flour quality. Sustained rainfall during harvest increases PHS by weakening seed dormancy, which is controlled by the antagonistic phytohormones abscisic acid (ABA) and gibberellic acid (GA) (Debeaujon and Koornneef, 2000). Besides, Delay of Germination 1 (DOG1) was first identified as the causal gene of a major quantitative trait locus (QTL) that positively regulates seed dormancy (Bentsink et al., 2006). DOG1, TaDOG1-like (TaDOG1L), and HvDOG1-like (HvDOG1L) were shown to be required for seed dormancy (Ashikawa et al., 2010). A recent study has shown that TaPP2C-a6, a clade A PP2C that negatively regulates ABA signaling, interacts with TaDOG1Ls to regulate seed dormancy and germination in wheat (Zhang et al., 2025). In addition, brassinosteroid (BR) signaling indirectly regulates seed dormancy and germination by interacting with ABA/GA metabolism and signaling pathways (Tong et al., 2014). Here, we demonstrate that BR-negative regulator GLYCOGEN SYNTHASE KINASE 3 (GSK3) physically interacts with DOG1L4 and regulates its accumulation to mediate seed dormancy in wheat. We previously cloned the GSK3 kinase-encoding gene TaGSK3 (TraesCS3D02G137200) in wheat. The gain-of-function mutations of the conserved TREE286 motif in TaGSK3 to TREK286 or TRK285E produce a dwarf phenotype associated with the Green Revolution gene Rht-B1b (Dong et al., 2023). In this study, GSK3E285K overexpression (OE) transgenic plants in the modern cultivar KN199 (wild type, WT) background displayed BR-insensitive phenotypes (Figure S1) and significantly enhanced PHS resistance (Figures 1A, S2). Threshed seeds from the OE lines showed markedly delayed germination compared with WT (Figures 1B, S3). Notably, this effect was abolished in dormancy-released seeds (Figure S4), indicating that GSK3 specifically regulates seed dormancy rather than the germination process directly. These results indicate that GSK3 functions as a negative regulator of PHS in common wheat. GSK3 regulates pre-harvest sprouting resistance in wheat (A) GSK3E285K-OE exhibited a higher level of PHS resistance than KN199. (B, C) Germination percentages (GPs) in the threshed seeds of KN199 and the OE plants without (B) or with fluridone (FL) (C). Data are presented as means ± SD of five biological replications. Statistical differences were determined by ANOVA followed by post-hoc Tukey's test (P < 0.05). FL, fluridone. (D) Y2H assays showed the interaction between GSK3 and DOG1L4. (E) Pull-down assays show that GSK3 interacts with DOG1L4 in vitro. (F) GSK3 phosphorylates DOG1L4 in vitro. Red asterisk indicates phosphorylated DOG1L4. CIAP, Calf Intestinal Alkaline Phosphatase. (G, H) The stability of DOG1L4-MBP (G) and DOG1L4S294A-MBP mutated protein (H) in cell-free KN199 and OE plant seeds extracts. (I) Immunoblotting to detect the accumulation of DOG1L4 in the seeds of KN199 and OE at 20 and 25 DPA (days post-anthesis). Anti-GAPDH antibody served as an internal reference in (G–I). (J) Proposed model for the roles of GSK3 and DOG1L4 in regulating seed dormancy in wheat. In wild-type (WT), DOG1L4 is unstable and degraded, resulting in reduced seed dormancy. In GSK3E285K-OE plants, accumulated GSK3E285K phosphorylates and stabilizes DOG1L4, upregulates ABI5 expression, and phosphorylates and potentially stabilizes ABI5, collectively leading to significantly enhanced seed dormancy. ABA and GA are two key phytohormones involved in the regulation of PHS. Our previous work showed that GSK3 stabilizes the GA-negative regulator DELLA (Rht-B1b) proteins (Dong et al., 2023). We therefore hypothesized that GSK3-mediated PHS resistance might result from prolonged Rht-B1b stability, which would dampen GA signaling and suppress seed germination. We first detected the seed germination of Rht-B1b-OE. Consistent with a previous study (Van De Velde et al., 2021), Rht-B1b did not alter seed dormancy (Figure S5), indicating that GSK3E285K enhances seed dormancy through a DELLA-independent pathway. We next blocked endogenous ABA synthesis with fluridone (FL). As expected, seed germination in both KN199 and the overexpression lines increased from 3 DAI (days after imbibition) owing to reduced endogenous ABA levels. Nevertheless, GSK3E285K still delayed germination and lowered the GP (Figures 1C, S6), indicating that GSK3E285K enhances seed dormancy not fully dependent on endogenous ABA content. Together, these results show that GSK3E285K-OE restrains PHS through GA- and ABA-independent mechanisms. Seed dormancy is another major determinant of PHS, and DOG1 is a key dormancy regulator. We therefore sought to validate whether GSK3 interacts with DOG1 family members to regulate wheat seed dormancy. We found that only DOG1L4 could interact with GSK3 (Figure 1D), and the binding was mediated by its N-terminal DOG1 domain (Figure S7). Furthermore, pull-down, luciferase complementation imaging (LCI), bimolecular fluorescence complementation (BiFC), and co-immunoprecipitation (Co-IP) assays (Figure S8A–C) revealed that GSK3 interacts with DOG1L4 in vitro and in vivo. Generally, GSK3 functions by phosphorylating and stabilizing its downstream substrates. We then tested whether DOG1L4 is a substrate of GSK3. Indeed, GSK3 was able to phosphorylate DOG1L4 (Figure 1F). Liquid chromatography–tandem mass spectrometry (LC–MS) analysis showed that the residue Ser-294 was identified as a putative phosphorylation site of DOG1L4 by GSK3 (Figure S9). We generated the non-phosphorylatable mutant DOG1L4S294A to confirm Ser-294 as the authentic phosphorylation site. This mutation indeed abolished phosphorylation (Figure 1F). We subsequently performed the cell-free protein degradation assays. The rate of DOG1L4-MBP degradation was much slower in GSK3E285K-OE than in WT seed protein extracts (Figure 1G), whereas the rate of DOG1L4S294A-MBP degradation was comparable between the two extracts (Figure 1H). Furthermore, we compared DOG1L4 protein abundance in the seeds of WT and GSK3E285K-OE at 20 and 25 DPA (days post-anthesis) using DOG1L4 antibody (Figure S10). The DOG1L4 protein was significantly more abundant in GSK3E285K-OE compared with WT (Figure 1I), which was also observed in Co-IP assays (Figure S8C). DOG1L4-OE lines also displayed strong PHS resistance (Figure S11). These results demonstrate that overexpression of GSK3 increases DOG1L4 protein stability without altering its transcript level, suggesting post-translational regulation of DOG1L4 and the resulting PHS resistance. Moreover, transcriptome sequencing (RNA-seq) revealed distinct expression patterns of ABA pathway genes in GSK3E285K-OE versus WT plants (Figure S12). RT-qPCR confirmed that several ABA signaling genes, such as ABI5 and VP1, were significantly upregulated in GSK3E285K-OE transgenic plants (Figure S12C). Phylogenetic analysis of upregulated ABI5 homologous revealed four duplicated paralogs on chromosome 3A and a single copy on 3D (Figure S13). We selected one representative from the 3A array and the unique 3D copy for subsequent assays. GSK3 could interact with ABI5-4900 but not with ABI5-1900 (Figure S14A, B), suggesting that GSK3 has a specific and selective effect on ABI5. It has been reported that BIN2 can interact with and phosphorylate ABI5 (Hu and Yu, 2014). We confirmed this phosphorylation in wheat (Figure S14C), indicating that the GSK3-mediated regulation of ABI5 is evolutionarily conserved. Hormone profiling of GSK3E285K-OE lines revealed elevated ABA and reduced GA and BR levels (Figure S15), consistent with their strong dormancy phenotype. These findings establish a mechanistic link between GSK3 and the canonical ABA-ABI5 module, positioning GSK3 as a central integrator that coordinates multiple hormonal pathways to regulate seed dormancy. DOG1 can regulate the expression of hundreds of genes during seed maturation, including that of ABA-responsive factor ABI5 (Dekkers et al., 2016). As expected, the expression of ABI5 was upregulated in GSK3E285K-OE and DOG1L4-OE (Figure S16). But there is no direct evidence that DOG1L4 enhances its transcription. We cannot rule out the possibility that DOG1L4 affects the ability of other transcription factors to activate ABI5, thereby indirectly influencing the expression of ABI5. Identifying the direct downstream genes of DOG1L4 will be an important investigation in future work. Based on the current and previously published results (Zhang et al., 2025), we propose that accumulation of the gain-of-function mutant protein GSK3E285K stabililizes DOG1L4 protein, conferring a strong PHS resistance phenotype observed in the GSK3E285K-OE (Figure 1J). Our study establishes GSK3, a core BR signaling kinase, as a regulator of wheat PHS through modulation of DOG1L4 stability. Additionally, GSK3 interacts with and phosphorylates ABI5, uncovering a potential complementary mechanism underlying GSK3-mediated enhancement of seed dormancy. We thank Professor Jiaqiang Sun (CAAS, China) for providing the GSK3E285K-OE transgenic wheat in the KN199 background, and the Rht-B1b-OE transgenic wheat in the Fielder background. This work was supported by the National Key Research and Development Program of China (2024YFF1001200) and the National Natural Science Foundation of China (32301810, 32301837, U22A20472). The authors declare no conflicts of interest. J.W. and H.D. conceived this project; H.D., D.Y., and Y.H. performed the experiments; M.L. and M.C. performed some experiments on phenotypic and biochemical assays; H.D. and J.W. wrote the manuscript; X.G., Q.C., and Z.Y. revised the manuscript. All authors have read and approved the contents of this paper. Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.70234/suppinfo Figure S1. BR-insensitive phenotypes of the GSK3E285K-OE. Figure S2. GSK3E285K-OE transgenic plants exhibited enhanced PHS resistance compared with KN199. Figure S3. Seed germination morphology of the wild-type KN199 and GSK3E285K-OE transgenic plants without treatment at the 7 d. Figure S4. Seed germination morphology of the wild-type KN199 and GSK3E285K-OE transgenic plants in three months after harvest. Figure S5. Seed germination morphology of the wild-type Fielder and Rht-B1b-OE transgenic plants. Figure S6. Seed germination morphology of KN199 and GSK3E285K-OE transgenic plants at 7 d with 10 μM FL treatment. Figure S7. GSK3 interacts with the N terminus of DOG1L4. (A) Schematic structure of the DOG1L4 protein. Figure S8. GSK3 interacts with DOG1L4. (A) LCI assays show that GSK3 interacted with DOG1L4 in plants. Figure S9. The DOG1L4 phosphorylation site Ser294 by GSK3 was identified by LC-MS/MS. Figure S10. Detection of tissue-specific expression of DOG1L4 genes and specificity detection of DOG1L4 antibody. Figure S11. Overexpression of DOG1L4 enhanced PHS resistance compared with Fielder. Figure S12. Overexpression of GSK3 induced gene expression in the ABA pathway. Figure S13. Phylogenetic tree analysis of ABI5 amino acid sequences in bread wheat (T. aestivum) and its homologs in maize (Z. mays), rice (O. sativa), and Arabidopsis (A. thaliana). Figure S14. GSK3 interacts with and phosphorylates ABI5. Figure S15. Endogenous ABA/GA/BR levels in developing seeds (25 DAP) of KN199 and GSK3E285K-OE lines. Figure S16. Overexpression of DOG1L4 induces the expression of ABI5 Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.