The TaSnRK1‐TabHLH489 module integrates brassinosteroid and sugar signalling to regulate the grain length in bread wheat

Summary Regulation of grain size is a crucial strategy for improving the crop yield and is also a fundamental aspect of developmental biology. However, the underlying molecular mechanisms governing grain development in wheat remain largely unknown. In this study, we identified a wheat atypical basic helix–loop–helix (bHLH) transcription factor, TabHLH489, which is tightly associated with grain length through genome‐wide association study and map‐based cloning. Knockout of TabHLH489 and its homologous genes resulted in increased grain length and weight, whereas the overexpression led to decreased grain length and weight. TaSnRK1α1, the α‐catalytic subunit of plant energy sensor SnRK1, interacted with and phosphorylated TabHLH489 to induce its degradation, thereby promoting wheat grain development. Sugar treatment induced TaSnRK1α1 protein accumulation while reducing TabHLH489 protein levels. Moreover, brassinosteroid (BR) promotes grain development by decreasing TabHLH489 expression through the transcription factor BRASSINAZOLE RESISTANT1 (BZR1). Importantly, natural variations in the promoter region of TabHLH489 affect the TaBZR1 binding ability, thereby influencing TabHLH489 expression. Taken together, our findings reveal that the TaSnRK1α1‐TabHLH489 regulatory module integrates BR and sugar signalling to regulate grain length, presenting potential targets for enhancing grain size in wheat.


Introduction
Bread wheat (Triticum aestivum L.) is a vital staple food crops worldwide, providing sustenance for 40% of the global population (Gupta et al., 2008).With the demand for food increasing and farmland decreasing, breeding for increased wheat yield has become a crucial objective.Among the various components, grain weight is one of the intensively studied traits due to its higher stability and heritability even under changing environmental conditions.Grain weight is determined by grain size and grain-filling rate, with genetic pathways such as proteasomal degradation, G-protein signalling and phytohormone signalling mainly associated with controlling grain size (Li and Yang, 2017).These pathways are genetically conserved in monocotyledonous and dicotyledonous plant species despite some differences in functional mode.In regulating grain size, the components of the ubiquitination-proteasome pathway play critical roles, such as GW2 (Hao et al., 2021;Song et al., 2007), WTG1/OsOTUB1 (Huang et al., 2017;Wang et al., 2017), OsUBP15 (Shi et al., 2019) and HDR3 (Gao et al., 2021).All subunits of heterotrimeric G proteins complex are involved in regulating grain size in rice (Cui et al., 2020;Liu et al., 2022;Sun et al., 2018), including Ga subunit (D1/RGA1), Gb subunit (RGB1) and Gc subunits (RGG1, RGG2, GS3, DEP1 and OsGGC2).
The plant steroid phytohormone brassinosteroid (BR) has been reported to play critical roles in grain development (Tong and Chu, 2018;Wu et al., 2008).BR binding to the extracellular domain of the plasma membrane localized receptor kinase BRASSINOSTEROID INSENSITIVE1 (BRI1) triggers a series of phosphorylation and dephosphorylation to inhibit the activity of GSK3-like kinase BRASSINOSTEROID-INSENSITIVE2 (BIN2) and then activating the BRASSINAZOLE RESISTANT1 (BZR1) family transcription factors, which directly regulate BR-responsive gene expression and plant development (Kim and Wang, 2010;Sun et al., 2010;Yu et al., 2011).The rice defective BR biosynthesis or signalling mutants displayed the reduced grain length and decreased grain weight, such as brd1, brd2, d2, d61, gw5, gsk2, gl2 and Osbzr1 (Bai et al., 2007;Che et al., 2015;Hong et al., 2002Hong et al., , 2003;;Huang et al., 2022;Liu et al., 2017;Yamamuro et al., 2000).Similar, genes affecting BR biosynthesis and signalling in wheat have also been shown to control grain size.For example, the wheat BR-deficient Tad11 mutant produced shorter grain compared to wild type and knock-down the expression of TaBRI1 or increasing TaGSK3 protein stability in Tasg-D1 mutant displayed the less sensitivity to exogenous BR and decreased grain size (Cheng et al., 2020;Xu et al., 2022).
Sucrose nonfermenting-1-related protein kinase 1 (SnRK1) is an evolutionarily conserved energy sensor in plants that shares homology with SNF1 in yeasts and AMP-activated protein kinase (AMPK) in mammals.SnRK1 kinase functions with a heterotrimeric complex, containing catalytic a subunit, regulatory b subunit and c subunit.SnRK1 promotes catabolism and reduces anabolism to maintain cellular energy homeostasis under stress conditions by phosphorylating key enzymes of diverse metabolic processes and master regulators of various signalling pathways.Additionally, SnRK1 plays important roles in plant growth, development and stress adaptations in response to different energy status.Recent studies have demonstrated that SnRK1 is involved in grain development and filling.In rice, the loss-offunction Ossnrk1a mutant displayed the reduced grain filling rate and low seed setting rate compared to wild-type plants (Hu et al., 2022;Li et al., 2022).OsSnRK1a regulates the nonstructural carbohydrates distribution between leaf sheath and panicle to control the grain filling, thereby affecting the grain weight and crop yield (Hu et al., 2022).
Some members of the basic helix-loop-helix (bHLH) family have been reported to participate in response to endogenous hormonal and environmental signals and regulation of cell elongation (Jung et al., 2016;Oh et al., 2014;Pedmale et al., 2016).Here, in this study, an atypical bHLH gene, TabHLH489, in the 2D chromosome of wheat was identified by genome-wide association study (GWAS) and map-based cloning in wheat that controls grain length.The expression of TabHLH489 in developing grains was found to have a negative correlation with wheat grain length.Knockout of TabHLH489 and its homologues led to an increased grain length and weight, whereas the overexpression resulted in decreased grain length and weight.Moreover, TaSnRK1a1 interacts with and phosphorylates TabHLH489 to promote its degradation, thus promoting grain development.Additionally, BR decreases the transcript levels of TabHLH489 through TaBZR1 to promote grain development.Natural variations in the promoter could influence the expression of TabHLH489 by affecting the binding ability of TaBZR1.These variations may have been selected during the breeding process, implying that natural variations in TabHLH489 contribute to grain length diversity in wheat and may be a promising candidate for improving wheat yield.

Map-based cloning of TabHLH489
To identify genes associated with grain length, a previously characterized common wheat population composed of 343 cultivars and 21 landraces (Wang et al., 2020), capturing the genetic diversity and geographical distribution of the wheat population, were utilized in GWAS with a marker significance threshold P value of 1.0EÀ04 (Figure S1a,b).The locus qGL2D on chromosome 2D, evidenced by the leading SNP AX-111195248 (2D:599896645, P = 2.63EÀ06), was of particular interest as it accounted for 15.17% of grain length variation in the population (Figure 1a).The associated physical interval of the leading SNP was estimated to be 6.22 Mb (linkage disequilibrium block 2, 594.19 Mb-600.4Mb) (Figure 1b).The leading SNP was able to differentiate the association panel into two haplotypes and 60 accessions with the favourable A-type haplotype exhibit longer grain and higher grain weight compared to those with the G-type haplotype (Figure 1c).To identify the candidate gene in this region, an SSR marker Xwmc181.2 and an InDel marker IN507 were used to construct the BC 3 F 3 population using two wheat varieties, Chinese spring (CS) of the G-type and Shixin 828 (SX) of the A-type, which showed significant differences in grain length and grain weight.Total 9368 individuals of the BC 3 F 3 population were generated and used to narrow down the locus qGL2D associated with grain length.The result of fine mapping analysis indicated that the candidate gene of qGL2D is situated within the 346.2Kb region flanked by markers CAPS91 and IN95, which included a large intergenic region and four genes annotated with high confidence according to IWGSC RefSeq V1.1 (Figure 1d-f).Among these four genes, TraesCS2D02G499300 and TraesCS2D02G499400 had notably low expression levels in wheat grains, while TraesCS2D02G499200 and TraesCS2D02G499500 displayed high expression levels with significant differences between CS and SX varieties (Figure S1c).Additionally, we compared the expression levels of TraesCS2D02G499200 and TraesCS2D02G499500 in 10-DPA (days post-anthesis) seeds from 102 wheat varieties, revealing a significant negative correlation (R = À0.27,P = 0.0082) between the expression of TraesCS2D02G499200 and grain length (Figure 1g) and a weaker and less significant correlation (R = À0.24,P = 0.014) between the expression of TraesCS2D02G499500 and grain length (Figure S1d).It is worth noting that TraesCS2D02G499500 encodes a homologue of Arabidopsis PELOTA (At4G27650), which is involved in RNA quality control systems during translation (Eberhart and Wasserman, 1995;Sz adeczky-Kardoss et al., 2018a,b).In contrast, TraesCS2D02G499200 encodes an atypical bHLH transcription factor known as TabHLH489 in wheat based on previous research (Wei and Chen, 2018).TabHLH489 is a member of the bHLH subfamily 26, exhibiting high homology with UPB1, AIF1-AIF4 and IBH1, all of which play roles in inhibiting plant cell elongation (Bai et al., 2012;Wei and Chen, 2018;Zhang et al., 2009).Therefore, we hypothesized that TraesCS2D02G499200, also known as TabHLH489, may be the potential candidate gene responsible for controlling grain length.

TabHLH489 inhibits the grain development
To validate the role of qGL2D in the regulation of grain length, we constructed the near-isogenic line of qGL2D (BC 5 F 2 ) which was the substitution line in CS genetic background.Compared to CS with the grain length of 6.43 mm and 1000-grain weight of 36.42 g, NIL qGL2D and SX have large grains with the grain length of 6.78 mm and 6.95 mm and 1000-grain weight of 44.22 g and 52.54 g, respectively (Figure 2a-c).Quantitative RT-PCR analysis confirmed that the relative expression level of TabHLH489 in NIL qGL2D was reduced to 46.4% compared to CS in 5-DPA seeds (Figure 2d).Additionally, TabHLH489 exhibited varied expression across different stages of wheat development, with the highest expression observed in the third leaf of wheat seedlings (Figure S2a).TabHLH489 was located in the nuclear in tobacco leaves (Figure S2b).To identify whether TabHLH489 is the key gene of qGL2D, which contributes to longer grain length in SX, we generated the knockout mutants of TabHLH489 gene and its homologous genes on chromosome 2A and/or 2B using CRISPR/Cas9 method in wheat cultivars Fielder background (Figure S3a S2.(e) Progeny testing of homozygous recombinants (BC 3 F 4 ) was used to narrow down the candidate region.The grains (n > 100) came from six individual plants on average.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05, n ≥ 6).(f) Predicted candidate genes in the narrowed region according to the International Wheat Genome Sequencing Consortium wheat genome (IWGSC, RefSeq V1.1).(g) Association analysis of TraesCS2D02G499200 expression level and grain length in 10-DPA seeds of 102 representative wheat varieties.Different lowercase letters indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).TaSnRK1 and TabHLH489 regulate wheat grain development 1991 plant height compared to Fielder, but had longer and heavier grains than Fielder, suggesting TabHLH489 plays an important role in wheat grain development .However, the triple mutant tabhlh489-aabbdd showed a slight decrease of grain width compared to Fielder, indicating there may be a negative effect of grain width resulted by knocking out of three homologous TabHLH489 genes (Figure 2h,j).Although the expression level of TabHLH489 was negatively correlated with grain length in 102 wheat varieties, the protein sequence alignment of TabHLH489 (from CS) and TabHLH489-S (from SX) revealed 8 amino acid substitutions and 3 amino acid insertions in TabHLH489-S compared to TabHLH489 (Figure S5).To examine whether these CDS variations are involved in the grain length regulation in the qGL2D, the transgenic plants performing similar expression levels of TabHLH489 or TabHLH489-S were selected for further phenotypic analysis (Figure S6a).The pUbi:TabHLH489-YFP (OE-TabHLH489) and pUbi:TabHLH489-S-YFP (OE-TabHLH489-S) transgenic plants both displayed significantly decreased plant height, grain length and grain weight compared to Fielder (Figures 2e-i, S6b-f).These results suggested that the difference in grain length between CS and SX was probably not due to the coding sequence differences between TabHLH489 and TabHLH489-S (Figure S5), but rather may be attributed to the different promoter sequences of TabHLH489 in CS and SX, which results in varying expression levels of TabHLH489 in these two varieties (Figures S1c,S7 and S8).Together, these results indicated that TabHLH489 is a negative regulator of wheat grain development.

TabHLH489 reduces the grain length by inhibiting cell elongation
To further investigate the functional mechanism of TabHLH489 in wheat grain development, we performed the transcriptomic analysis using 5-DPA seeds of Fielder and OE-TabHLH489 plants.Overall, we identified 1226 differentially expressed genes, with 815 upregulated and 411 downregulated by more than 1.5-fold in OE-TabHLH489 compared to Fielder plants (Figure 3a, Table S1).Gene ontology (GO) analysis showed that genes involved in carbohydrate metabolism, cell wall organization, water deprivation, lipid catabolism, oxidation-reduction process, xyloglucan catabolism and BR response were highly enriched in TabHLH489-regulated genes (Figure 3b).Cell wall organization is critical for the cell elongation, suggesting that TabHLH489 may regulate wheat grain development by modulating cell elongation.Remarkably, two family genes, Xyloglucan endotransglucosylase/hydrolases (XTHs) and Expansins (EXPs), which were related to cell elongation, performed significantly different expression profiles in OE-TabHLH489 and Fielder (Figure 3c).Quantitative RT-PCR analysis further confirmed that the expression levels of cell elongation-related genes TaEXPA2 (TraesCS3A02G344800) and TaEXPA4 (TraesCS1D02G299700) in 5-DPA seeds were significantly decreased in OE-TabHLH489 plants, while increased in tabhlh489-aabbdd triple mutant (Figures 3d, S6g).Consistent with this, histological sectioning analysis showed that the tabhlh489-aaBBdd double mutant and tabhlh489-aabbdd mutant had significantly enlarged cell length of grain pericarp, while the overexpression of TabHLH489 or TabHLH489-S both displayed the reduced cell length (Figures 3e,f,S6h,i).These results indicated that TabHLH489 regulates the wheat grain length by repressing the expression of cell elongation-related genes in early grain developing stage.

TabHLH489 interacts with TaSnRK1a1 to regulate grain development
To investigate the functional mechanism of TabHLH489 in wheat grain development, we performed a yeast two-hybrid (Y2H) screen to identify interaction proteins using the full length TabHLH489 protein as bait and the Y2H complementary DNA library that was constructed with different organs of wheat, including root, leaf, young spike and seeds of different days post anthesis.Among the putative TabHLH489-interacting proteins, TraesCS1D02G353300 exhibited the strong interaction with TabHLH489.TraesCS1D02G353300 is an orthologous gene of the catalytic a-subunit of SnRK1 in wheat, thus was named as TaSnRK1a1.SnRK1 plays critical roles in gain filling and development in rice (Hu et al., 2022;Li et al., 2022).We selected TaSnRK1a1 for further research to reveal the mechanism by which the TaSnRK1a1-TabHLH489 module regulates grain development.Additional Y2H assays confirmed that TabHLH489 interacted with TaSnRK1a1 in yeast (Figure 4a).The in vitro protein pull-down assays revealed that glutathione S-transferase (GST) fusion protein GST-TaSnRK1a1 interacted with maltose binding protein (MBP) fusion protein MBP-TabHLH489, but not with MBP alone (Figure 4b).To determine whether TaSnRK1a1 interacts with TabHLH489 in planta, we performed transient ratiometric bimolecular fluorescence complementation (rBiFC) assays in the mesophyll protoplast of leaves.The strong yellow fluorescent signals were observed in the protoplast co-transformed with TaSnRK1a1-cYFP and TabHLH489-nYFP, but not in those co-transformed with TaSnRK1a1-cYFP and Histone3-nYFP (Figure 4c).Co-immunoprecipitation (Co-IP) assays using A. tumefaciens-mediated transient infiltration in N. benthamiana expressing p35S:TanRK1a1-Myc only or coexpressing p35S:TaSnRK1a1-Myc and p35S:TabHLH489-YFP further confirmed the interaction between TaSnRK1a1 and TabHLH489 in planta (Figure 4d).Together, these results indicated that TaSnRK1a1 interacts with TabHLH489 in vitro and in vivo.
To investigate the potential role of TaSnRK1a1 in wheat grain development, we generated the TaSnRK1a1 overexpression plants (pUbi:TaSnRK1a1-YFP, OE-TaSnRK1a1) and the triple knockout mutants of TaSnRK1a1 (situated on chromosome 1D) and its two homologues TaSnRK1a1-1A, TaSnRK1a1-1B (tasnrk1a1-ko) in Fielder background (Figure S9a-f).The overexpression of TaSnRK1a1 led to increased grain length and grain weight (Figures 4e-g, S9b,c), as well as longer cell length in grain pericarps compared to Fielder plants (Figure 4h,i).The tasnrk1a1ko mutants displayed no significant difference in grain length and grain pericarp cell length, likely due to the redundancy of 15 TaSnRK1 homologue genes in wheat genome (Figure S9g-j).Furthermore, quantitative RT-PCR analysis showed that the overexpression of TaSnRK1a1 strikingly increased the transcript levels of TaEXPA2 and TaEXPA4 in 5-DPA seeds (Figure 4j).These results suggested that TaSnRK1a1 acts as a positive regulator of wheat grain development.

TaSnRK1a1 phosphorylates TabHLH489 to promote its degradation
To examine the impact of the interaction between TaSnRK1a1 and TabHLH489, we generated the progeny from a cross between OE-TaSnRK1a1 and OE-TabHLH489 plants.Phenotypic analysis revealed that grains from the crossed progeny, as well as pericarp cell length, were longer than those from OE-TabHLH489 plants.Additionally, grain weight in the crossed progeny was recovered compared to Fielder, suggesting that TabHLH489 activity was partially inhibited by TaSnRK1a1 (Figures 5a-c, S10a-c).Quantitative RT-PCR analysis showed that the expression levels of TabHLH489 in 5-DPA seeds of the OE-TaSnRK1a1/OE-TabHLH489 plants were similar to those of OE-TabHLH489 plants (Figure S10a).Nevertheless, the results of the western blot analysis revealed that the protein levels of TabHLH489 were significantly reduced in the OE-TaSnRK1a1/OE-TabHLH489 plants compared to OE-TabHLH489 plants (Figure 5d), suggesting that the overexpression of TaSnRK1a1 results in the decreased protein stability of TabHLH489.
Considering that TaSnRK1a1 is an evolutionarily conserved energy sensor kinase and TabHLH489 interacts with TaSnRK1a1, the kinase assays were performed to determine whether TabHLH489 is the substrate of TaSnRK1a1.The results showed that MBP-TaSnRK1a1 alone failed to phosphorylate MBP-TabHLH489.However, in the presence of the upstream kinase MBP-SnAK2, MBP-TaSnRK1a1 is activated by SnAK2-mediated phosphorylation.Activated MBP-TaSnRK1a1 phosphorylated MBP-TabHLH489, but not MBP only (Figure 5e).To analyse the effects of TaSnRK1a1-mediated phosphorylation on the function of TabHLH489, we investigated the degradation rate of MBP-TabHLH489 in cell-free extracts of OE-TaSnRK1a1 and wild type plants.The result showed that the degradation rate of MBP-TabHLH489 was 1.85 times faster in the extracts derived from OE-TaSnRK1a1 compared to those obtained from Fielder, as determined by the half-life time (T 1/2 ) of MBP-TabHLH489, suggesting that TaSnRK1a1-mediated phosphorylation reduces the protein stability of TabHLH489 (Figure 5f,g).Moreover, the sucrose treatment promoted the accumulation of TaSnRK1a1, but decreased the protein stability of TabHLH489 (Figure 5h).Taken together, TaSnRK1a1 promotes wheat grain development through decreasing TabHLH489 protein stability by phosphorylation.

TabHLH489 reduces the BR sensitivity of wheat
The angle of wheat leaf inclination between leaf blade and the culm is an important agronomic trait and contributes to the wheat architecture and grain yields.BR plays a unique and crucial role in determine leaf inclination (Gao et al., 2019;Min et al., 2019;Wu et al., 2016).The overexpression of TabHLH489 or TabHLH489-S both resulted in the decreased flag leaf inclination, whereas tabhlh489-aaBBdd and tabhlh489-aabbdd mutants both showed the significant increased flag leaf inclination (Figures 6a,b, S6j,k).The dwarfism phenotype and erect leaves of TabHLH489 and TabHLH489-S overexpression plants led us to speculate that TabHLH489 may be a negative regulator of BR signalling pathway.To test this hypothesis, we analysed the BR sensitivity of tabhlh489-aabbdd mutant and TabHLH489 overexpression lines.BR treatment increased the leaf inclination of Fielder plants in a dose-dependent manner, while such promoting effects were enhanced in tabhlh489-aabbdd mutant, but reduced in OE-TabHLH489 plants (Figure 6c,d).Quantitative RT-PCR analysis showed that BR treatment significantly reduced the expression of TaCPD and TaD2 in Fielder plants, but had weak effects in OE-TabHLH489 plants (Figure 6e).These results indicated that TabHLH489 functions as a negative regulator of BR signalling in wheat.

BR induces the wheat grain length by repressing TabHLH489 expression
Considering the critical roles of BR in grain development and lamina joint morphogenesis and the negative regulation of wheat grain size and leaf angle by TabHLH489, we speculated that BR may regulate wheat grain development through modulating TabHLH489 expression levels.Exogenous BR treatment significantly reduced TabHLH489 expression and this inhibitory effect became more pronounced with the increase of treatment time (Figure S11).To further test this hypothesis, we generated a series of BR-related wheat materials.Based on the sequence similarity analysis, the homologous genes of BRI1, BIN2 and BZR1 in wheat were identified, including TaBRI1, TaSK2 and TaBZR1.We generated the knockout mutants of tabri1-aaBBdd, tabri1-AAbbdd, tabri1-aabbDD (the triple mutant of tabri1-aabbdd is unable to survive) (Figure S12a) and task2-aabbdd (the triple mutant of TaSK2 [from chromosome 1B] and its two homologues TaSK2-1A and TaSK2-1D) (Figure S13a) using the CRISPR/Cas9 method and the overexpression of TaSK2 (OE-TaSK2) (Figure S13b) and TaBZR1 (OE-TaBZR1) (Figure S14a) by expressing the pUbi:TaSK2-YFP and pUbi:TaBZR1-YFP in Fielder background.Phenotypic analysis showed that loss-of-function mutants tabri1-aaBBdd, tabri1-AAbbdd, tabri1-aabbDD and the overexpression of TaSK2 caused the dwarf wheat, decreased flag leaf angle, short grains and decreased grain weight compared to Fielder (Figures 7a-g, S12b-h and S13c-i).In contrast, the task2-aabbdd mutant and TaBZR1 overexpression lines displayed the similar plant height, increased flag leaf angle, longer grains and increased grain weight compared to Fielder (Figures 7a-g, S13c-i and S14b-h).Histologic sectioning analysis showed that tabri1-aaBBdd, tabri1-AAbbdd, tabri1-aabbDD and OE-TaSK2 plants displayed the reduced cell length of grain pericarps, but task2aabbdd and OE-TaBZR1 plants showed that increased cell length of grain pericarps (Figures 7h,i, S12i,j, S13j,k and S14i,j).These results indicated that BR promotes grain development in wheat.Next, we analysed the expression levels of TabHLH489 expression levels in 5-DPA seeds of different BR-related materials.Quantitative RT-PCR analysis showed that the transcript levels of TabHLH489 is significantly increased in 5-DPA seeds of tabri1-AAbbdd mutant and OE-TaSK2 lines, but decreased in 5-DPA seeds of task2-aabbdd mutant and TaBZR1 overexpression transgenic plants (Figure 7j).Furthermore, the transcript levels of TaEXPs are significantly increased in 5-DPA seeds of OE-TaBZR1 plants (Figure S14k).Furthermore, we examined the expression levels of TaBRI1, TaSK2, TaEXPA2 and TaEXPA4 in 7-day-old seedlings of Fielder and TabHLH489 single, double and triple knockout mutants.The findings revealed that the knockout of TabHLH489 led to the downregulation of TaBRI1 and TaSK2, as well as the upregulation of TaEXPA2 and TaEXPA4, (Figure S15a, b), indicating that TabHLH489 may regulate wheat grain development in a dose-dependent manner.These results suggested that BR might promote wheat grain length by inhibiting TabHLH489 expression.

DNA variations in the TabHLH489 promoter confer grain length variations
BZRs are the key transcription factors in BR signalling and BRRE motif (CGTGT/CG) is responsible for the BZR1 binding.Genomic analysis revealed that TabHLH489 has an identical genetic sequence in two wheat varieties, CS and Fielder, but has different genetic sequence in SX variety.Motif searching revealed the promoter region of TabHLH489 underwent DNA variation, causing four putative BRRE motifs to change in both CS and SX (Figures 8a, S16).Among these, SNP-281 caused a BRRE motif change in the promoter of TabHLH489 in CS, located in a chromatin-accessible peak at À500 ~À272 bp during 6-DPA of CS endosperm (Figure 8a).This suggests that TaBZR1 may directly regulate TabHLH489 expression by binding to the TabHLH489 promoter, resulting in differential expression of TabHLH489 in CS and SX.The chromatin immunoprecipitation (ChIP)-qPCR assay with OE-TaBZR1 transgenic wheat further showed that TaBZR1 directly bound to TabHLH489 promoter fragments P3 and P6 (Figure 8b).Analysis of the P6 region in CS and SX revealed a typical BZR1-binding motif in SX and an atypical BZR1-binding motif in CS due to one SNP.TaBZR1 exhibited a higher binding ability to the P6 region of SX than CS in DNA protein pull-down assays (Figure 8c,d).The transient expression assays showed a significant decrease in TaBZR1 transfected luciferase activity derived from pTabHLH489:LUC and pTabHLH489-S:LUC, which was significantly reduced by point mutation of TaBZR1-binding motif in the P6 region of TabHLH489-S promoter (Figure 8e).In summary, TaBZR1 contributes to the differential expression patterns of TabHLH489 in CS and SX.
Next, we wonder how the different haplotypes of TabHLH489 could impact its function and regulation of grain length.To investigate the correlation between the SNP at the TaBZR1 binding site with TabHLH489 expression level in 10-DPA grains at the population level, we developed a PCR marker targeting the P6 region and genotyped a core collection consisting of 102 representative varieties from the GWAS panel.Compared to the TabHLH489 haplotype (Hap-C), 15 accessions of TabHLH489-S haplotype (Hap-S) had significantly longer grain and remarkably lower expression level (Figure 8f,g).These findings indicate that natural variations in the proximal (P6) region may contribute to the TabHLH489 expression differences and grain length variations within wheat population, potentially by affecting binding of TaBZR1.

Breeding selection of TabHLH489
To investigate whether the SNP in P6 was selected during the breeding process in China, we further genotyped the Chinese wheat mini-core collection (MCC), which consists of 262 accessions and represents much of the genetic diversity in the Chinese national collection.The percentage of accessions with Hap-S was lower in introduced cultivars from other countries (21.05%) compared to landraces (41.96%) and modern cultivars (33.64%) (Figure 9a).Interestingly, the frequency of Hap-S was observed to decline sharply during the early stages of the breeding process, from 50.00% in stage 1949-1957 to 19.64% in stage 1958-1978 and remained low in frequency, present in only 17.65% of cultivars post-2000 (Figure 9b).Additionally, we observed distinct distribution characteristics of the alleles in the major Chinese agro-ecological zones, with the zone III and VI exhibiting a low frequency of Hap-S compared to other zones (Figure 9c).However, the grain length of varieties in these major Chinese agro-ecological zones was increased accompanied by the frequency of Hap-S (Figure 9d).Collectively, our findings indicate the potential selection of the short grain Hap-C of TabHLH489 during the breeding process in China.

Discussion
Wheat grain weight is a complex quantitative trait and is largely dependent on grain size.In the present study, we demonstrated that the critical roles of TabHLH489 and TaSnRK1a1 in wheat grain development.Overexpressing TaSnRK1a1 or knocking out TabHLH489 both promotes wheat grain development and increases grain weight.Sugar induces the accumulation of TaSnRK1a1, which interacts with and phosphorylates TabHLH489 to facilitate its degradation.Additionally, we found that BR promotes wheat grain development through reducing TabHLH489 expression through transcription factor TaBZR1.By integrating BR and sugar signals, the TaSnRK1a1-TabHLH489 module regulates wheat grain development (Figure 9e).

TabHLH489 acts as a negative regulator of wheat grain development
Grain weight is an important agronomic trait for cereal crops and several genes affecting wheat grain have been isolated through QTL mapping or homologous gene analysis.Here, a key locus qGL2D was identified as a regulator of grain length and grain weight through GWAS.This locus has been previously identified by different research groups studying varieties with different grain sizes (Cui et al., 2014;Ramya et al., 2010;Su et al., 2018;Xie et al., 2015).Our research further revealed that TabHLH489 is a potential gene responsible for qGL2D locus.The correlation analysis and transgenic materials both showed lower expression levels of TabHLH489 resulting increased grain length.Furthermore, the knockout of TabHLH489 on chromosome 2D alone promoted grain length and additional knockouts on chromosome 2A and 2B can further enhance the grain length and regulate the downstream gene expression.These findings suggested that TabHLH489 negatively regulates wheat grain development in a dosagedependent manner.
The qGL2D locus consists of four genes, two of which are not expressed in wheat grains.However, TabHLH489 and TraesCS2D02G499500 exhibit the high expression levels in wheat grains and their expression levels are negatively correlated with wheat grain length.We have chosen to focus on studying TabHLH489 rather than TraesCS2D02G499500 due to the following two considerations.Firstly, the negative correlation between the expression level of TraesCS2D02G499500 in seeds and wheat grain length was significantly lower compared to that of TabHLH489.Secondly, TraesCS2D02G499500 encodes a homologue of Arabidopsis PELOTA, which is involved in RNA quality control systems during translation and plays an important role in immunity and salt stress responses (Ge et al., 2023;Kong et al., 2021).Conversely, TabHLH489 encodes homologous genes of Arabidopsis, UPB1, AIF1-AIF4 and IBH1, all of which are involved in inhibiting plant cell elongation (Bai et al., 2012;Wei and Chen, 2018;Zhang et al., 2009).However, the effect of qGL2D on grain length surpassed that of knocking out TabHLH489.It implies that qGL2D may be a compound locus in regulating grain length.Thus, the function of TraesCS2D02G499500 in grain development needs to be studied in the future.

TabHLH489 regulates wheat grain development downstream of the BR signalling pathway
Brassinosteroids are a class of growth-promoting hormones and play pivotal roles in crop grain development.In rice, mutations that results in BR deficiency or BR insensitivity led to shorter and lighter grains (Cheng et al., 2020;Gao et al., 2019;Hong et al., 2003;Wu et al., 2016).This study confirms the critical role of BRs in wheat grain development.Enhancing BR signal by knocking out TaSK2 or overexpressing TaBZR1 led to the increased grain length and grain weight, while weakening the BR signal by knocking out TaBRI1 or the overexpression of TaSK2 resulted in the decreased grain length and grain weight.The expression of TabHLH489 was decreased in the transgenic materials with the knockout of TaSK2 or the overexpression of TaBZR1, but increased in the materials with the knockout of TaBRI1 or the overexpression of TaSK2.Additionally, TaBZR1 directly binds to the promoter of TabHLH489 to inhibit its ª 2024 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22,[1989][1990][1991][1992][1993][1994][1995][1996][1997][1998][1999][2000][2001][2002][2003][2004][2005][2006] expression.These findings suggest that TabHLH489 regulates wheat grain development downstream of BR signalling pathway.
Plant height and grain size are two crucial agronomic traits that greatly impact the crop yield.Excessively tall plants are prone to lodging, while excessively short plants tend to produce smaller grains, more ineffective tillers and exhibit weak disease resistance (Liu et al., 2018).Hence, achieving an optimal balance in plant height and grain size is vital for improving the crop yield.BR plays a significant role in regulating these traits in crops.Numerous BRrelated genes have been identified in rice, most of which influence both plant height and grain size.Mutants associated with BR synthesis in rice, such as brd2, d11, d2 and brd1, exhibit varying degrees of reduced plant height and grain size (Hong et al., 2003(Hong et al., , 2005;;Mori et al., 2002;Tanabe et al., 2005).In our study, we uncovered a novel component of the BR signalling pathway called TabHLH489, which negatively regulates plant height and grain development.Overexpressing TabHLH489 in wheat led to decreased plant height and grain size, while knocking out TabHLH489 had no significant effects on plant height and resulted in increased grain weight.Additionally, both BR and TabHLH489 were found to regulate leaf angle, which is pivotal for photosynthesis and plant density.Manipulating TabHLH489 expression to reduce BR signalling resulted in more upright leaves.Therefore, targeting BR and TabHLH489 presents an excellent opportunity for wheat breeding.By modulating the activity of TabHLH489, the intensity of BR signal transduction can be adjusted, ultimately yielding wheat plants with semi-dwarfed height, upright leaves and appropriately sized grains.

TabHLH489 is a potential target for breeding practices
In breeding practices, grain shape and size are important factors in determining the production and market value of bread wheat (Gegas et al., 2010).The significant increase in grain width in wheat breeding has led to a shift from long and thin grains to larger, shorter and wider grains, primarily through modifications in grain length during domestication (Eckardt, 2010;Gegas et al., 2010;Yoshioka et al., 2019).Despite the trade-off between number and size of grains, genetic gains in grain yield potential have been predominantly achieved by increasing grain number per spike, whilst selecting for stable or even reduced grain size (Jia et al., 2013;Shukla et al., 2015;Zhai et al., 2018).Analysis of each agro-ecological zone revealed that grain length was negatively correlated with the grain number per spike (Figure S17a-c).However, the grain number per spike has no remarkable correlation with these two haplotypes (Figure S17b,  c).The widespread use of the short-grain Hap-C of TabHLH489 in past Chinese breeding processes may be an indirect outcome of the balance between selecting for shorter but wider grains and the number versus size of grains.Thus, the short-grain Hap-C of TabHLH489 allele may be crucial in achieving larger grains with minimal trade-off in terms of grain number per spike, although there is still room for improvement, particularly in grain length in zones III and VI.
In conclusion, TabHLH489 plays a crucial role in wheat grain development, acting downstream of the BR signalling pathway.TaSnRK1a1 interacted with and phosphorylated TabHLH489 to induce its degradation, thereby promoting wheat grain development.Natural variation in the proximal (P6) region of TabHLH489 promoter contributes to the expression difference between the two haplotypes of TabHLH489 by affecting TaBZR1 binding capacity.However, the superior haplotype within the proximal (P6) region of the TabHLH489 promoter did not emerge prominently during domestication and breeding practices.These findings suggest that further investigation is needed to uncover the underlying mechanisms of TabHLH489 in regulating wheat development and grain yield, with the aim of potentially reintegrating the superior haplotype into wheat breeding.

Phenotyping and GWAS
A total of 364 wheat accessions were evaluated in six different environments using a plant density of 2.7 million ha À1 .Field trials were conducted at Shijiazhuang (37.85°N, 114.82°E) and Dezhou (37.43°N, 116.35°E) in China over three consecutive cropping seasons (2015-2018), with two replicates carried out in a randomized complete block design.Each block consisted of six rows, each three meters in length and spaced 21 cm apart.Agronomic management followed local cultivation practices and no serious damage from plant diseases, insect pests, or lodging occurred during the growing seasons.A total of 30 representational main spikes from the inner rows were harvested and used to measure grain shape and size.Grain weight was determined with three replicates, using more than 200 random grains for each replicate and then transformed into 1000-grain weight (TGW, g).The respective average values for grain length (GL, mm), grain width (GW, mm), grain length/width (GLW), grain area (GA, mm 2 ) and grain circumference (GC, mm) were calculated using a Crop Grain Appearance Quality Scanning Machine (SC-G, Hangzhou Wanshen Technology Co., Ltd., China).Each accession was genotyped using Affymetrix Wheat660K SNP arrays by Capital Bio Corporation (Beijing, China).SNPs were filtered based on the following criteria to ensure high quality SNP markers: (1) Minor allele frequency (MAF) not less than 5%; (2) Missing rate in the population does not exceed 10%; (3) Genotype hybrid rate less than 5%; and (4) Unique mapping to the reference genome IWGSC RefSeq V1.0.High-quality SNPs from 364 samples were analysed for association with phenotypic data using Tassel v5.2 using the mixed linear model (introducing PCA as a fixed effect and Kship matrix as a random effect in the model).The independent effective markers (Ne) were estimated based on linkage disequilibrium using GEC software (Li et al., 2012).The P value of 1/Ne was used as the suggestive threshold for declaring a significant association, in accordance with the adjusted Bonferroni method.Manhattan plots and quantile-quantile plots were generated using R package 'CMplot' (https://github.com/YinLiLin/R-CMplot). Pairwise R 2 values were calculated and displayed with LD plots using Haploview 4.2 software (Barrett et al., 2005).

Map-based cloning of TabHLH489
SX is a commercial variety known for its increased grain length and grain weight, while CS is a landrace with short grain length and low grain weight.To develop F 1 hybrids, SX was used as the male parent and CS as the female parent.The SSR marker Xwmc181.2 and InDel marker IN507, which were adjacent to the locus qGL2D, were used as assistant markers for genetic population construction due to their polymorphism between two parental lines.The BC 3 F 3 population for fine mapping of TabHLH489 and near-isogenic lines (NIL qGL2D ) were generated by three or five generations of backcrossing CS 9 SX F 1 progeny to CS.The resulting BC 3 F 2 plants were self-pollinated to produce BC 3 F 3 progenies, which were used to perform fine mapping analysis of TabHLH489.The BC 3 F 3 populations were grown at Yuanxia village, Yantai, Shandong, China (37.53°N, 121.15°E), with spacing of 0.4 meters between rows and 0.2 meters between individual plants during the 2021-2022 growing season.To isolated the recombinant lines, DNAs were extracted from the immature leaves of the 9368 individual plants in the BC 3 F 3 populations using the CTAB method.PCR genotyping was conducted by using markers in the qGL2D region.Markers' information was shown in Table S2.The progenies of recombinants were used to test the grain length trait.
The wheat wild type and all transformation backgrounds used in this study were the Fielder variety.Transgenic plants and nearisogenic lines for phenotype analysis were planted in a greenhouse with a 16-h light (22 °C)/8-h dark (16 °C) cycle and approximately 50% relative humidity.

Histological assay
Grains from three individual plants at 15-DPA stage were sampled and fixed in Carnoy's fluid (absolute alcohol: acetic acid = 3:1, v/v) at room temperature for 24 h.After fixation, grains were dehydrated in 100%, 90%, 80% and 70% ethanol for 1 h sequentially and stored in 70% ethanol at last.Pericarp pieces of 1-2 mm 2 were taken from the middle of grains and stained using a modified Pseudo-Schiff Propidium Iodide (PS-PI) method (Truernit et al., 2008).Cell lengths of pericarps were pictured by the LSM-900 laser scanning confocal microscope (Zeiss, Germany) and measured by ImageJ 1.51 (http://imagej.net/ij/).

Pull-down assay
The recombinant GST-fused TaSnRK1a1 and MBP-fused TabHLH489 were extracted from bacteria using glutathione beads (GE Healthcare, USA) and amylose resin (New England Biolabs), respectively.To validate the protein interaction, GST-TaSnRK1a1 (1 lg) was incubated with MBP-TabHLH489 (1 lg) or MBP (1 lg) at 4 °C for 1 h and the beads were washed six times with wash buffer.The proteins were eluted from glutathione beads by boiling in 50 lL 2 9 SDS loading buffer and separated by an 8% SDS-PAGE gel.Gel blots were analysed using anti-MBP (New England Biolabs, 1:5000 dilution) and anti-GST antibodies (Merck, 1:3000 dilution, Germany).
The stability assays for both TaSnRK1a1 and TabHLH489 proteins were carried out on 7-day-old seedlings of OE-TaSnRK1-a1 and OE-TabHLH489 plants after a 6-h treatment with 1% sucrose under light conditions.The total protein extraction and GFP-tagged protein abundance were performed as mentioned above.

Cell-free degradation assay
Total proteins were extracted from 7-day-old seedlings of Fielder and OE-TaSnRK1a1 using the modified cell-free degradation buffer (25 mM Tris-HCl, pH 7.5, 10 mM NaCl, 10 mM MgCl 2 , 1 mM PMSF, 5 mM DTT and 2 mM ATP) as previously described (Wang et al., 2009).The extracted protein concentrations from Fielder and OE-TaSnRK1a1 were adjusted to the equal concentration of 5 lg/lL in the degradation buffer.The recombinant proteins MBP-TabHLH489 and MBP were purified from Escherichia coli and incubated in 100 lL total protein extracts from Fielder or OE-TaSnRK1a1 for 5 min, 10 min and 15 min at 22 °C, respectively.Samples were taken from each reaction at the indicated time points.MBP-TabHLH489 and MBP abundance were determined by immunoblot analysis as mentioned above and the degradation ratio of MBP-TabHLH489 and MBP in Filder and OE-TaSnRK1a1 were calculated as the ratios between MBP-TabHLH489 and TaRubisco, which was used as a loading control and stained with CBB R250.The protein abundance of each time point was quantified by ImageJ 1.51.

Chromatin immunoprecipitation (ChIP)-qPCR assay
The young leaves from 7-day-old Fielder and OE-TaBZR1 plants were harvested and crosslinked with 1.0% formaldehyde (v/v) for 30 min by vacuum infiltration.The crosslinking reaction was stopped by replacing the formaldehyde solution with 0.25 M ª 2024 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22,[1989][1990][1991][1992][1993][1994][1995][1996][1997][1998][1999][2000][2001][2002][2003][2004][2005][2006] glycine.The anti-GFP polyclonal antibody (ab290, Abcam, England) was incubated with Protein A/G Magnetic Beads at 4 °C for 2 h on a rotator.For immunoprecipitation, the sonicated chromatin extracts were combined with the antigen-antibodybeads complex and incubated at 4 °C for 2 h.The beads were washed sequentially with low salt wash buffer, high salt wash buffer, LiCl buffer and TE buffer, as previously described (Tian et al., 2018).The protein-DNA complex was eluted with the elution buffer (0.1 M NaHCO 3 and 1% SDS, w/v) and decrosslinked with 0.2 M NaCl at 65 °C overnight.Proteins were removed using proteinase K (Thermo Scientific) and DNA was precipitated using 3 volumes of ethanol and 1/10 volume of 3 M sodium acetate.The fold enrichment of the promoter of TabHLH489 was calculated as the ratio between Fielder and OE-TaBZR1.The promoter of TaADPRF was used as an internal control.The ChIP-qPCR experiments were carried out with three independent biological repeats.Primer pairs used for ChIP-qPCR analysis were listed in Table S2.

RNA-seq
For RNA-Seq analysis, 5-DPA seeds from Fielder plants and OE-TabHLH489 plants grown in the greenhouse were collected to isolate the total mRNA and construct the mRNA sequencing libraries.Sequencing was performed on the BGISEQ-500 platform at Beijing Genomics Institute.The RNA-seq clean reads were mapped to the Chinese Spring reference genome (International Wheat Genome Sequencing Consortium RefSeq v1.1) downloaded from Ensembl Plants using Bowtie2 software.Htseq-count with default parameters was used to calculate the gene read counts, which were then normalized to TPM (Transcripts Per Kilobase per Million mapped reads) and calculated the gene relative expression level by Salmon.Differential gene expression analysis was carried out using DESeq2 with genes showing log 2 (fold change) ≥ 0.5849 and adjusted P value (FDR) < 0.05 were considered to be differentially expressed ones.Enriched GO analysis of TabHLH489-regulated genes was identified using Cluster Profiler.

Figure 1
Figure 1 Map-based cloning of grain length regulating gene TabHLH489.(a, b) Manhattan plot (a) and LD heatmap (b) showing the significant peak for wheat grain length on chromosome 2D.The horizontal dashed black line indicates the significance threshold P value (P = 1.0EÀ04) for marker-trait associations.The 10-Mb region surrounding peak marker of qGL2D was indicated by dashed red vertical lines and a LD heatmap for pairwise R 2 value of markers in this region was shown with white to black represents R 2 = 0-1.The leading SNP and gene TabHLH489 were indicated by red and blue asterisks, respectively.(c) The distribution of grain length between haplotypes defined by the leading SNP on chromosome 2D.Significance was determined by Student's t-test (n = 364).(d) The candidate gene for grain length was fine-mapped to a 346.2 Kb region between markers CAPS91 and IN95.Boxs in white and black indicate SX and CS genotype of regions, respectively.'r' indicates number of recombinants.Markers' information was listed in TableS2.(e) Progeny testing of homozygous recombinants (BC 3 F 4 ) was used to narrow down the candidate region.The grains (n > 100) came from six individual plants on average.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05, n ≥ 6).(f) Predicted candidate genes in the narrowed region according to the International Wheat Genome Sequencing Consortium wheat genome (IWGSC, RefSeq V1.1).(g) Association analysis of TraesCS2D02G499200 expression level and grain length in 10-DPA seeds of 102 representative wheat varieties.Different lowercase letters indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).

Figure 2
Figure 2 TabHLH489 negatively regulates wheat grain development.(a-c) Wheat grain morphology of CS, SX and NIL qGL2D .The grains (n > 100) came from six individual plants on average.Scale bar = 1 cm.(d) Quantitative RT-PCR analysis of the expression level of TabHLH489 in 5-DPA seeds of CS and NIL qGL2D .Error bars indicate AESD from three biological repeats.TaADPRF was used as an internal control.(e, f) Plant architecture of Fielder, OE-TabHLH489 and TabHLH489 knockout plants at the heading stage.Error bars indicate AESD (n ≥ 6).Scale bar = 10 cm.(g-j) Wheat grain morphology of Fielder, OE-TabHLH489 and TabHLH489 knockout plants.The grains (n > 300) came from six individual plants on average.Scale bar = 1 cm.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).'***' indicates statistically significant differences between samples (Student's t-test, P < 0.001).

Figure 4
Figure 4 TaSnRK1a1 interacts with TabHLH489 and promotes wheat grain development.(a) TaSnRK1a1 interacts with TabHLH489 in yeast.(b) TaSnRK1a1 directly interacts with TabHLH489 in vitro.Glutathione agarose beads loaded with GST-TaSnRK1a1 were incubated with equal amounts of MBP or MBP-TabHLH489.Proteins bound to GST-TaSnRK1a1 were detected by immunoblot analysis with an anti-MBP antibody.The red asterisk indicates the MBP-TabHLH489 protein band.(c) Confocal images of ratiometric bimolecular fluorescence complementation assays showed that TaSnRK1a1 interacts with TabHLH489 in protoplast.nYFP-H3 indicates nYFP-Histone3.Scale bar = 20 lm.(d) TaSnRK1a1 interacts with TabHLH489 in plants.Immunoprecipitation was performed using the tobacco leaves transient expressing p35S:TaSnRK1a1-Myc only or co-expressing p35S:TaSnRK1a1-Myc and p35S:TabHLH489-YFP.The co-immunoprecipitation experiments were performed using GFP-Trap agarose beads and the immunoblots were probed with anti-Myc or anti-GFP antibodies.(e-g) Wheat grain morphology of Fielder and TaSnRK1a1 overexpression lines.The grains (n > 300) came from six individual plants on average.Scale bar = 1 cm.(h, i) The grain pericarp cell length of Fielder and OE-TaSnRK1a1 plants.The cells (n > 100) came from six individual plants on average.Scale bar = 50 lm.(j) Quantitative RT-PCR analysis of the expression of TaEXPA2 and TaEXPA4 in 5-DPA seeds of Fielder and OE-TaSnRK1a1 plants.Error bars indicate AESD (n = 3).TaADPRF was used as an internal control.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).'***' indicates statistically significant differences between samples (Student's t-test, P < 0.001).

Figure 5
Figure 5 TaSnRK1a1 phosphorylates and destabilizes TabHLH489.(a-c) Wheat grain morphology of Fielder, OE-TaSnRK1a1, OE-TabHLH489 and the progenies derived from a cross between OE-TaSnRK1a1 and OE-TabHLH489 plants.The grains (n > 100) came from more than five individual plants on average.Scale bar = 1 cm.(d) Immunoblot assay of the protein levels of TabHLH489 and TaSnRK1a1 in Fielder, OE-TabHLH489, OE-TaSnRK1a1 and OE-TaSnRK1a1/OE-TabHLH489 plants.TaActin was used as the loading control.(e) TaSnRK1a1 phosphorylates TabHLH489 in vitro.Upper image was the gel containing proteins labelled with ATP-c-32 p, while the following image was the gel staining with coomassie brilliant blue.(f, g) Degradation of MBP and MBP-TabHLH489 in cell-free system from Fielder and OE-TaSnRK1a1 extracts.Recombinant purified MBP and MBP-TabHLH489 were added to the protein extracts and incubated at 22 °C for different times.Protein abundance was evaluated using an anti-MBP antibody.The band intensity was quantified by ImageJ.(h) Sucrose promoted TaSnRK1a1 accumulation but induced TabHLH489 degradation.Total proteins were extracted from 7-day-old seedlings of OE-TaSnRK1a1 and OE-TabHLH489 plants after a 6-h treatment with 1% sucrose under light conditions.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).'*' indicate statistically significant differences between samples (Student's t-test, P < 0.05).

Figure 6
Figure 6 TabHLH489 reduces the BR sensitivity of wheat.(a, b) The flag leaf angles of Fielder, OE-TabHLH489 and TabHLH489 knockout mutants.Error bars indicate AESE, scale bar = 5 cm.(c, d) The leaf angles of Fielder, OE-TabHLH489 and tabhlh489-aabbdd mutant in the presence of different concentrations of eBL.Error bars indicate AESE, scale bar = 5 cm.(e) Quantitative RT-PCR analysis of the expression level of TaCPD and TaD2 in 7-day-old Fielder, OE-TabHLH489 plants with or without 100 nM eBL foliar-sprayed treatment for 3 h.Error bars indicate AESD from three biological repeats.TaADPRF was used as an internal control.Different lowercase letters indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).'*' indicate statistically significant differences between samples (Student's t-test, P < 0.05).

Figure 8
Figure 8 Natural variation in the promoter of TabHLH489 confer grain length variation.(a) A schematic diagram showing the natural variation, putative BRRE motifs (Blue triangles) and open chromatin regions (Red peaks) in the promoter of TabHLH489.TabHLH489 and TabHLH489-S indicate genes from wheat CS and SX, respectively.'[0-30]' indicates the range of ATAC peaks (Reads Per Kilobase per Million mapped reads, RPKM).Grey boxes indicate the promoter region.Black boxes with arrows indicate the coding sequence.(b) TaBZR1 directly binds to the promoter of TabHLH489.The enrichment value was calculated as the ratio between Fielder and OE-TaBZR1.Promoter of TaADPRF was used as reference control.Error bars indicate AESD (n = 3).(c, d) DNA-protein pull down assay showed the different binding ability of TaBZR1 to P6 fragment from CS and SX.Error bars indicate AESD (n = 3).(e) The mutation in P6-SX (pTabHLH489-S-mut, containing TCCAaaATCA) declines the TaBZR1's inhibition of pTabHLH489-S.Wheat protoplasts were transformed with dual luciferase reporter constructs containing pTabHLH489:LUC, pTabHLH489-S:LUC, pTabHLH489-S-mut:LUC and/or constructs overexpressing the effector TaBZR1-GFP.LUC (Firefly luciferase) activity was normalized to REN (Renilla luciferase) and the LUC/REN ratios of control samples were normalized to one.Error bars indicate AESD (n = 3).(f) Expression difference of the haplotype TabHLH489-S (Hap-S) or TabHLH489 (Hap-C) in total 102 representative varieties.(g) Grain length variance of 102 representative varieties containing Hap-S or Hap-C.'*', '**' and '***' indicate statistically significant differences between samples (Student's t-test, P < 0.05, P < 0.01 and P < 0.001, respectively).

Figure 9
Figure 9 Breeding selection of TabHLH489 and a working model of BR-TabHLH489 regulating grain length.(a, b) The frequency of two haplotypes in different category (a) and breeding periods (b) in wheat mini core collections (n = 287).(c) Distribution of Hap-S and Hap-C in the major Chinese agroecological zones.The size of pie charts in the geographical map showing the number of cultivars, with percentage of two haplotypes in different colours.(d) Correlation of grain length with the frequency of Hap-S in 10 ecological zones.(e) A working model of TabHLH489 and TaSnRK1a1 integrating BR and sugar signals to regulate wheat grain length.TabHLH489 acts as a negative regulator of wheat grain length by suppressing the expression of cell-elongation genes TaEXP2 and TaEXP4.TaSnRK1a1 interacts with and phosphorylates TabHLH489 to promote its degradation, thereby facilitating grain development.Sugar induces the accumulation of TaSnRK1a1 but promotes TabHLH489 degradation.BR promotes wheat grain development by reducing TabHLH489 expression through transcription factor TaBZR1.Natural variations in the promoter of TabHLH489 in CS and SX results in the different expression of TabHLH489 in CS and SX by affecting the binding ability of TaBZR1.Therefore, the TaSnRK1a1-TabHLH489 regulatory module integrates BR and sugar signalling to regulate wheat grain size.Black bars indicate transcriptional repression.Red arrows and bars indicate posttranslational activating and inhibitory effects on protein.Green dashed lines indicate TabHLH489 produced by gene expression.Blue dashed lines show indirect mechanisms.
(n ≥ 6).Scale bar = 10 cm.(d) The mutation sites of tasnrk1a1ko mutants are indicated in red.(e, f) Plant architecture of Fielder and tasnrk1a1-ko mutants at the heading stage.Error bars indicate AESD (n ≥ 6).Scale bar = 10 cm.(g, h) Wheat grain morphology of Fielder and tasnrk1a1-ko mutants.The grains (n > 300) came from six individual plants on average.Scale bar = 1 cm.(i, j) The grain pericarp cell length of Fielder and tasnrk1a1-ko mutants.The cells (n > 100) came from six individual plants on average.Scale bar = 50 lm.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).Figure S10 TaSnRK1a1 inhibits activity of TabHLH489 in the crossing progenies.(a) Quantitative RT-PCR analysis of the expression level of TabHLH489 in 5-DPA seeds of Fielder, OE-TaSnRK1a1, OE-TabHLH489 and OE-TaSnRK1a1/OE-TabHLH489 plants, respectively.Error bars indicate AESD from three biological repeats.TaADPRF was used as an internal control.(b, c) The grain pericarp cell length of Fielder, OE-TabHLH489, OE-TaSnRK1a1 and OE-TaSnRK1a1/OE-TabHLH489 plants.The cells (n > 100) came from six individual plants on average.Scale bar = 50 lm.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).
Figure S11Quantitative RT-PCR analysis of BR effects on TabHLH489 expression.Total RNA was extracted from the leaves of 7-day-old seedlings which were foliar-sprayed with 100 nM eBL (containing 0.01% TritonX-100) for different time.Error bars indicate AESD (n = 3).TaADPRF was used as an internal control.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).

Figure S12
TaBRI1 knocked out affects wheat grain development.(a) The mutation sites of TaBRI1 knockout mutants are indicated in red.(b, c) Plant architecture of Fielder and tabri1 mutants at the heading stage.Error bars indicate AESD (n ≥ 6).Scale bar = 10 cm.(d, e) The flag leaf angles of Fielder and tabri1 mutants.Scale bar = 5 cm.Error bars indicate AESE of different lines.(f-h) Wheat grain morphology of Fielder and tabri1 mutants.The grains (n > 300) came from six individual plants on average.Scale bar = 1 cm.(i, j) The grain pericarp cell length of Fielder and tabri1 mutants.The cells (n > 100) came from six individual plants on average.Scale bar = 50 lm.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).

Figure S13
TaSK2 inhibited wheat grain development.(a) The mutation sites of TaSK2 mutant are indicated in red.(b) Quantitative RT-PCR analysis of the expression level of TaSK2 in 5-DPA seeds of OE-TaSK2 plants.Error bars indicate AESD from three biological repeats.TaADPRF was used as an internal control.(c, d) Plant architecture of Fielder, OE-TaSK2 and task2-aabbdd mutant at the heading stage.Error bars indicate AESD (n ≥ 6).Scale bar = 10 cm.(e, f) The flag leaf angles of Fielder, OE-TaSK2 and task2-aabbdd mutant.Scale bar = 5 cm.Error bars indicate AESE.(g-i) Wheat grain morphology of Fielder, OE-TaSK2 and task2-aabbdd mutant.The grains (n > 300) came from six individual plants on average.Scale bar = 1 cm.(j, k) The grain pericarp cell length of Fielder, OE-TaSK2 and task2-aabbdd mutant.The cells (n > 100) came from six individual plants on average.Scale bar = 50 lm.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).

Figure S14
TaBZR1 promotes wheat grain development.(a) Quantitative RT-PCR analysis of the expression level of TaBZR1 in 5-DPA seeds of Fielder and OE-TaBZR1 plants.Error bars indicate AESD from three biological repeats.TaADPRF was used as an internal control.(b, c) Plant architecture of Fielder and OE-TaBZR1 plants at the heading stage.Error bars indicate AESD (n ≥ 6).Scale bar = 10 cm.(d, e) The flag leaf angles of Fielder and OE-TaBZR1 plants.Error bars indicate AESE, Scale bar = 5 cm.(f-h) Wheat grain morphology of Fielder and OE-TaBZR1 plants.The grains (n > 300) came from six individual plants on average.Scale bar = 1 cm.(i, j) The grain pericarp cell length of Fielder and OE-TaBZR1 plants.The cells (n > 100) came from six individual plants on average.Scale bar = 100 lm.(k) Quantitative RT-PCR analysis of TaEXPA2 and TaEXPA4 in 5-DPA seeds of Fielder and OE-TaBZR1 plants.Error bars indicate AESD (n = 3).TaADPRF was used as an internal control.Different letters above bars indicate statistically significant differences between samples (one-way ANOVA, P < 0.05).'**' and '***' indicates statistically significant differences between samples (Student's t-test, P < 0.01 and P < 0.01, respectively).

Figure S15
TabHLH489 regulates the expression level of BRrelated genes in a dose-dependent manner.(a) Quantitative RT-PCR analysis of TaBRI1 and TaSK2 in 7-day seedlings of Fielder and TabHLH489 single, double and triple mutants.(b) Quantitative RT-PCR analysis of TaEXPA2 and TaEXPA4 in 7-day seedlings of Fielder and TabHLH489 single, double and triple mutants.Error bars indicate AESD (n = 3).TaADPRF was used as an internal control.Different letters above bars indicate statistically significant differences between samples (two-way ANOVA, P < 0.05).Figure S16 Natural variations of the TabHLH489 promoter region alter the binding site of TaBZR1.Green background colour indicates sequence variance.Red Letters indicate the core element of the binding site of TaBZR1.
Figure S17 TabHLH489 contributes a limitation of the trade-off between grain number per spike and 1000-grain weight.(a) The correlation analysis between the grain length and grain number per spike in wheat varieties from the major Chinese agroecological zones.(b) The correlation analysis between the expression level of TabHLH489 and grain number per spike in 102 representative varieties.(c) Grain number per spike of wheat varieties containing Hap-S or Hap-C.

Table S1
DEGs in seeds of Fielder and OE-TabHLH489 at the 5-DPA stage.Table S2 Information of primers used in this study.ª 2024 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 1989-2006