The TaSOC1‐TaVRN1 module integrates photoperiod and vernalization signals to regulate wheat flowering

Summary Wheat needs different durations of vernalization, which accelerates flowering by exposure to cold temperature, to ensure reproductive development at the optimum time, as that is critical for adaptability and high yield. TaVRN1 is the central flowering regulator in the vernalization pathway and encodes a MADS‐box transcription factor (TF) that usually works by forming hetero‐ or homo‐dimers. We previously identified that TaVRN1 bound to an MADS‐box TF TaSOC1 whose orthologues are flowering activators in other plants. The specific function of TaSOC1 and the biological implication of its interaction with TaVRN1 remained unknown. Here, we demonstrated that TaSOC1 was a flowering repressor in the vernalization and photoperiod pathways by overexpression and knockout assays. We confirmed the physical interaction between TaSOC1 and TaVRN1 in wheat protoplasts and in planta, and further validated their genetic interplay. A Flowering Promoting Factor 1‐like gene TaFPF1‐2B was identified as a common downstream target of TaSOC1 and TaVRN1 through transcriptome and chromatin immunoprecipitation analyses. TaSOC1 competed with TaVRT2, another MADS‐box flowering regulator, to bind to TaVRN1; their coding genes synergistically control TaFPF1‐2B expression and flowering initiation in response to photoperiod and low temperature. We identified major haplotypes of TaSOC1 and found that TaSOC1‐Hap1 conferred earlier flowering than TaSOC1‐Hap2 and had been subjected to positive selection in wheat breeding. We also revealed that wheat SOC1 family members were important domestication loci and expanded by tandem and segmental duplication events. These findings offer new insights into the regulatory mechanism underlying flowering control along with useful genetic resources for wheat improvement.


Introduction
Flowering at the optimum time is critical for crop reproduction and yield formation.Control of flowering time (usually represented by heading date in crops) is an important target for crop breeding to harness adaptation to local environments and increase yield potential.The flexible regulatory machinery underlying the decision to flower enhances adaptability of crops to a broad range of agronomic and climatic conditions.Wheat (Triticum aestivum) is the most widely grown crop and provides staple foodstuff for approximately one-third of the world population (FAO, https://www.idrc.ca/en/article/facts-Figuresfood-and-biodiversity).Identification and functional dissection of flowering genes in wheat are therefore essential to improve environmental adaptability and yield potential.
The requirement for prolonged exposure to low temperature, known as vernalization, prevents plants from flowering under freezing conditions, especially during long winter in temperate regions (Kim et al., 2009).Wheat cultivars grown in different environments need diverse vernalization characteristics to ensure flowering initiation and reproductive development at the optimum time, as that is critical for high yield and crop rotations (Milec et al., 2014).Wheat cultivars are classified as winter and spring types based on vernalization requirements.
TaVRN1 encoding a MADS-box transcription factor (TF) is highly homologous to AP1, a floral meristem identity gene in Arabidopsis (Danyluk et al., 2003;Mandel et al., 1992;Trevaskis et al., 2003;Yan et al., 2003).MADS-box TFs usually work by forming hetero-or homo-dimers (de Folter et al., 2005;Kaufmann et al., 2005).Hence, identification of MADS-box proteins interacting with TaVRN1 is extremely helpful to expand its regulatory network in flowering pathways.We previously identified a SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1 (SOC1)-like MADS-box TF, herein designated as TaSOC1, as a partner of TaVRN1 through a yeast two-hybrid library screening (Cao and Yan, 2013).In addition, Li et al. (2013) validated the interaction using pull-down assays.In model plants, SOC1 integrates multiple flowering signals, such as photoperiod, temperature, hormone and plant age (Cao et al., 2021;Hyun et al., 2016).HvSOC1-like1 in barley also responds to vernalization (Papaefthimiou et al., 2012).These findings suggested that SOC1-like genes of temperate cereals might be involved in vernalization-induced flowering.However, the function of SOC1 homologues remains largely unknown in wheat.
Considering that TaSOC1 can interact with TaVRN1, the central factor in vernalization flowering pathway, it is necessary to elucidate the specific function of TaSOC1 and its genetic relationship with TaVRN1 in flowering initiation.In the present study, TaSOC1 was validated as a flowering repressor in the vernalization and photoperiod pathways.We confirmed the interaction of TaSOC1 and TaVRN1 both physically and genetically.We also identified Flowering Promoting Factor 1-like gene TaFPF1-2B as a common downstream target of TaSOC1 and TaVRN1.Moreover, TaSOC1 could impair the interaction between TaVRN1 and TaVRT2, another MADS-box flowering promoter, and hereby a model underpinning the synergistic flowering regulation in response to vernalization and photoperiod was proposed.We further identified natural variations at TaSOC1 and its orthologues, and examined their genetic effects on flowering time in wheat cultivars.These results not only improve understanding of the regulatory network underlying flowering initiation but also provide important genetic resources for wheat breeding.

TaSOC1 is a flowering repressor in the vernalization pathway
To determine the function of TaSOC1, we generated overexpression (OE) lines in winter wheat cultivar Kenong 199 (KN199).Thirtythree independent positive transgenic lines were created and three representative lines, TaSOC1-OE2, TaSOC1-OE3 and TaSOC1-OE25 were used in subsequent analyses.Phenotypic analyses showed that TaSOC1 overexpression significantly delayed flowering under non-vernalization, incomplete vernalization (4 °C for 14 days) and complete vernalization (4 °C for 30 days) conditions (Figure 1a-f).Remarkably, TaSOC1 inhibited flowering more significantly under non-vernalization conditions than after vernalization treatments (Figure 1a-f).TaSOC1 overexpression lines (TaSOC1-OE) flowered approximately 4.3 days later than transgenic-null lines (TNL) under complete vernalization conditions, whereas TaSOC1 overexpression delayed flowering 6.9 and 20.8 days under incomplete vernalization and non-vernalization conditions, respectively (Dataset S1).
TaSOC1, a SOC1-like gene on chromosome 4B, has orthologues, temporarily designated TaSOC1-5A and TaSOC1-4D, on chromosomes 5A and 4D.Considering that orthologous genes usually have functional redundancy in wheat, we created triple knockout lines of the SOC1-like genes (TaSOC1-KO) in KN199.Phenotypic investigation showed that TaSOC1-KO lines had similar heading date with KN199 under complete vernalization conditions, whereas TaSOC1-KO significantly accelerated flowering approximately 1.0 and 8.0 days under incomplete and nonvernalization conditions, respectively (Figure 2a-g; Dataset S2).This result confirmed that TaSOC1 acted as a flowering repressor in the vernalization pathway.

TaSOC1 physically interacts with TaVRN1
The MADS-box TF TaSOC1 was identified as a partner of TaVRN1 by Y2H assays in a previous study (Cao and Yan, 2013).To confirm the interaction, we firstly performed a bimolecular fluorescence complementation (BiFC) assay and found that TaVRN1 and TaSOC1 could bind with each other in wheat protoplasts (Figure 3a).Their interaction signals appeared mainly in the nuclei, suggesting that TaVRN1 and TaSOC1 formed a transcriptional complex to regulate downstream genes.We also employed luciferase complementation imaging (LCI) assays to test the interaction between TaSOC1 and TaVRN1 in leaves of Nicotiana benthamiana.Significant luciferase (LUC) signals were detected in the area where TaSOC1-cLUC and TaVRN1-nLUC were co-infiltrated, whereas there were no LUC signals in the negative controls (Figure 3b), confirming the interaction of TaVRN1 and TaSOC1 in planta.
We investigated the expression patterns of TaVRN1 and TaSOC1 in Wheat Expression Browser (http://wheat-expression.com/), which showed that both genes were highly expressed in leaves, the most important tissue to response to environmental cues for flowering initiation (Figure S1a,b).Considering that TaVRN1 is the central regulator in the vernalization flowering pathway and interacts with TaSOC1, we compared their response to vernalization.Reverse transcription quantitative PCR (RT-qPCR) assays showed that TaVRN1 and TaSOC1 had high expression activity in leaves of KN199 and also significantly responded to vernalization (Figure 3c,d).Strikingly, vernalization induced the expression of TaVRN1 but repressed expression of TaSOC1 (Figure 3c,d).These results indicated that TaVRN1 and TaSOC1 had an overlapping spatiotemporal expression window to initiate flowering in the vernalization pathway.
TaSOC1 probably has a genetic interaction with TaVRN1 in the vernalization flowering pathway To further validate the interaction of TaSOC1 and TaVRN1 in the vernalization flowering pathway, we created a segregating population from a cross of TaSOC1-OE and TaVRN1 overexpression lines (TaVRN1-OE).We genotyped individuals in the population to identify TaSOC1-OE, TaVRN1-OE, their double OE lines and NC (negative control).RT-qPCR assays showed that overexpressed TaSOC1 and TaVRN1 in TaSOC1 + TaVRN1 lines had similar expression levels to their counterparts in TaSOC1-OE and TaVRN1-OE lines (Figure 4a).Compared to non-vernalization, incomplete vernalization promoted flowering of TaSOC1-OE, TaVRN1-OE, TaSOC1 + TaVRN1 and NC (Dataset S1).Moreover, the plants in each line had more consistent flowering time under incomplete vernalization conditions than under non-vernalization conditions, so we investigated flowering time for a genetic analysis under incomplete vernalization conditions.Phenotypic investigation revealed that the TaSOC1 + TaVRN1 flowered earlier than TaSOC1-OE and later than TaVRN1-OE (Figure 4b; Dataset S3).Statistical analyses showed that both TaSOC1 and TaVRN1 had significant effects on flowering time (P < 0.05) in ª 2023 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 635-649 the segregating population (Figure 4c,d).A significant genetic epistatic effect (P < 0.05) between TaSOC1 and TaVRN1 was also detected (Figure 4d).
Flowering Promoting Factor 1-like gene TaFPF1-2B is a common downstream target of TaSOC1 and TaVRN1 Since TaSOC1 interacts with TaVRN1 physically and genetically, they can form a complex to regulate common downstream flowering genes.We conducted transcriptome analyses to expand the regulatory network underlying the cooperative modulating flowering controlled by TaVRN1 and TaSOC1.Both TaVRN1 and TaSOC1 caused differential expression of many genes in leaves at the beginning of stem elongation (floret differentiation), a key spatiotemporal window of flowering initiation in response to environmental cues (Xie et al., 2021).In total, 365 and 314 (|Log 2 (Fold change)| >1 and P < 0.01) differentially expressed genes (DEGs) were detected in the TaVRN1-OE and TaSOC1-OE, compared with their respective TNL (Figure 5a; Datasets S4 and S5).Fifteen of those genes were common downstream genes of TaVRN1 and TaSOC1, indicating some proportion of overlap in their downstream targets at the initial stage of flowering (Dataset S6).Interestingly, a Flowering Promoting Factor 1-like gene TaFPF1-2B (TraesCS2B02G059500) was identified as a common downstream target of TaVRN1 and TaSOC1.Flowering Promoting Factor 1-like genes were validated as flowering accelerators in the photoperiod and GA pathways in Arabidopsis (Kania et al., 1997;Melzer et al., 1999), and overexpression of NtFPF1 promotes flowering in Nicotiana plants (Smykal et al., 2004).TaFPF1-2B was significantly up-regulated in TaVRN1-OE and downregulated in TaSOC1-OE, which was confirmed by RT-qPCR assays (Figure 5b,c).We also investigated the expression of TaFPF1-2B in the segregating population derived from the cross between TaSOC1-OE and TaVRN1-OE.TaSOC1 and TaVRN1 showed significant interaction effects on expression of TaFPF1-2B (Figure 5d).Thus, TaSOC1 and TaVRN1 probably orchestrate expression of TaFPF1-2B to modulate flowering.
TaSOC1 and TaVRN1 have overlapping binding sites in the TaFPF1-2B promoter We conducted chromatin immunoprecipitation quantitative PCR (ChIP-qPCR) assays to determine whether TaFPF1-2B is a direct downstream target of TaSOC1 and TaVRN1.MADS-box transcription factors such as TaVRN1 and TaSOC1, were shown to bind with CArG motifs within the promoters of their target genes (Dubcovsky et al., 2008;Kane et al., 2007;Theißen and Gramzow, 2016;Xie et al., 2021).We thus designed primer pairs to query the enrichment of TaSOC1 and TaVRN1 over 11 regions of the TaFPF1-2B promoter covering the CArG-like motifs in the respective overexpression lines carrying TaSOC1-or TaVRN1-YFP fusion constructs (Figures 5e and S2).TaSOC1 and TaVRN1 were both significantly enriched at three sites containing CArG-like motifs near the transcription start site (TSS) in the TaFPF1-2B promoter (Figure 5f).Moreover, TaSOC1 and TaVRN1 shared two binding sites, suggesting that they could coordinately modulate TaFPF1-2B expression through direct binding with its promoter.We further used yeast one-hybrid (Y1H) to validate the interaction of TaSOC1 and TaVRN1 with the representative sites 5, 7, 8 and 9 that are close to transcription start site of TaFPF1 (Figures 5e and S2), for determining target sites and possible competition.The site 9 can generate auto-activation in yeast, so we cannot identify its binding to TaSOC1 and TaVRN1 in Y1H.Notably, TaSOC1 and TaVRN1 specifically bind to the sites 8 and 5, respectively, and both can interact with the site 7 (Figure 5g).The mutated sites cannot bind to TaSOC1 and TaVRN1 (Figure 5g).The results of Y1H are consistent with those of ChIP-qPCR assays and show that TaSOC1 and TaVRN1 may competitively bind to the site 7.

TaSOC1 is a flowering repressor in the photoperiod pathway
In addition to response to vernalization, SOC1 can integrate photoperiod and gibberellin (GA) signals to regulate flowering in model plants (Lee et al., 2004;Moon et al., 2003;Ryu et al., 2009).To further uncover the function of TaSOC1 in flowering initiation, we first investigated its dynamic expression pattern in leaves of KN199 under GA treatment.The foliage spray of GA significantly promoted seedling growth of KN199, demonstrating the effectiveness of GA treatments.However, there was no significant response in TaSOC1 expression to GA (Figure 6a).We then compared the expression profiles of TaSOC1 under different photoperiod treatments and found that TaSOC1 in completely vernalized KN199 had higher expression levels under short-day (SD) conditions than long-day (LD) conditions (Figure 6b).The time-course transcriptional variation of TaSOC1 over a full day under different photoperiod conditions also showed that SD upregulated TaSOC1 expression (Figure 6c).Phenotypic investigation also showed that GA treatments have similar effects on flowering time of TaSOC1-OE and TNL (Figure 6d,e; Dataset S7).By contrast, photoperiod treatment had little effect on TaVRN1 expression (Figure S3a,b).Upon being completely vernalized, TaVRN1-OE had similar flowering time to TNL under LD or SD  flowering time in a dose-dependent way (Dataset S7).We further investigated the flowering time of TaSOC1-KO and KN199 under SD conditions and found that two of the three TaSOC1-KO lines had significantly earlier flowering time than KN199 (Figure 6h,i; Dataset S8).No significant difference in flowering time was observed between TaSOC1-KO and KN199 under LD conditions (Figure 2b,c).Collectively, TaSOC1 plays a key role in repressing wheat flowering under SD conditions.

Functional conservation and variation of SOC1 homologues in modulating flowering
The majority of studied flowering plants have SOC1-like genes (Cao et al., 2021).SOC1 is a key flowering promoter in Arabidopsis (Hyun et al., 2016;Onouchi et al., 2000;Samach et al., 2000) and many SOC1 homologues have also been identified as flowering activators in other plants, such as rice (Oryza sativa) (Lee et al., 2004), soybean (Glycine max) (Na et al., 2013;Zhong et al., 2012), maize (Zea mays) (Zhao et al., 2014), bamboo (Phyllostachys heterocycla) (Hou et al., 2021), mustard (Sinapis alba) (Borner et al., 2000), Chrysanthemum (Fu et al., 2013), Eriobotrya japonica (Jiang et al., 2019), Medicago truncatula (Fudge et al., 2018), Petunia (Ferrario et al., 2004) and Phyllostachys violascens (Liu et al., 2016).Rice OsMADS50/OsSOC1 and OsMADS56 have the highest similarity with SOC1 in Arabidopsis.OsMADS50 is upregulated during the floral transition and its overexpression leads to extremely early flowering (Lee et al., 2004).OsMADS50 is expressed in vegetative tissues with elevated expression at the time of floral initiation and its overexpression in Arabidopsis also accelerates flowering (Lee et al., 2004;Tadege et al., 2003).However, OsMADS56 functions as a flowering repressor (Ryu et al., 2009).Here, we showed that TaSOC1 behaved as a flowering repressor.Phylogenetic analyses across major cereal crops and Arabidopsis showed that TaSOC1 was homologous to SOC1, OsMADS50 and OsMADS56 (Figure S4a).Synteny analysis revealed that TaSOC1 was orthologous to OsMADS50 (Figure S4b).Since OsMADS50 is a flowering activator in rice (Lee et al., 2004), it is evident that a functional differentiation event occurred between the orthologous genes TaSOC1 and OsMADS50.Surprisingly, no TaSOC1 orthologue was detected in barley, a relative of wheat (Figure S4c).We also identified two other groups of SOC1-like genes, designated as TaSOC1-like1 and TaSOC1-like2, on wheat homoeologous group-1 chromosomes.Genome assembly showed that TaSOC1-like2 (TraesC-S1A02G199900 and TraesCS1D02G203400) were derived from TaSOC1-like1 (TraesCS1A02G199600, TraesCS1B02G214500 and TraesCS1D02G203300) through tandem duplication events.Synteny analysis revealed that the TaSOC1-like genes had genomic collinearity with OsMADS56 (Figure S4d).The responses of TaSOC1-like genes to photoperiod and vernalization treatments were also investigated.Like TaSOC1, the TaSOC1-like1 and TaSOC1-like2 genes were repressed by vernalization although the latter exhibited weaker responses than the former.However, SD could not induce the expression of the TaSOC1-like genes.Additionally, TaSOC1-like genes were repressed by SD, which is reverse of the response of TaSOC1 to photoperiod treatment (Figure S5).HvSOC1-like1, a SOC1 homologue in barley and orthologue to OsMADS56 in rice, was induced by vernalization (Figure S4d) (Papaefthimiou et al., 2012).HvSOC1-like1 was also orthologous to TaSOC1-like genes in wheat (Figure S4d).Overall, SOC1 homologues in wheat have expanded greatly and undergone considerable functional differentiation although all of them are probably involved in flowering regulation.It is worth exploring the specific functions of each TaSOC1-like gene to dissect the flowering regulatory mechanisms in wheat.
The molecular network of TaSOC1-modulated flowering through integration of the vernalization and photoperiod pathways Understanding the plasticity of genetic architecture underlying flowering time is prerequisite to design resilient crops.SOC1 is an integrator of several flowering pathways, including photoperiod, temperature, hormone and plant age (Lee and Lee, 2010).Here, we confirmed that the wheat SOC1 homologue TaSOC1 also integrated photoperiod and vernalization signals to modulate flowering, suggesting that this gene plays an essential role in response to environmental cues to initiate flowering.TaSOC1 directly interacted with TaVRN1 to repress flowering in the vernalization pathway.A previous study showed that TaVRN1 could achieve positive feedback self-regulation through binding with TaVRT2 (Xie et al., 2021).Yeast three-hybrid (Y3H) assays revealed that TaSOC1 could disrupt interaction between TaVRT2 and TaVRN1, and TaVRT2 could also disrupt the TaVRN1-TaSOC1 interaction (Figure S6).This result indicates that TaSOC1 and TaVRT2 competitively bind to TaVRN1.Flowering promoting factor 1 (FPF1) genes were reported to accelerate flowering in Arabidopsis.Transcriptome assays showed that the FPF1-like gene TaFPF1-2B was repressed by TaSOC1 and induced by TaVRN1 (Figure 5b-d S7b,c).According to the model, TaVRN1 in winter wheat is expressed at very low levels after germination in the fall and cannot promote expression of TaFPF1.TaVRT2 is gradually upregulated during cold winter temperatures (vernalization) and subsequently enhances expression of TaVRN1 (Xie et al., 2021).The increased transcript abundance of TaVRN1 induces expression of TaFPF1.However, TaSOC1 is up-regulated by short photoperiod in winter and thus represses expression of TaFPF1.Accordingly, TaFPF1 still exhibits low expression level in winter due to the combined effects of TaSOC1, TaVRN1 and TaVRT2.With increasing daylength and temperature in spring, TaVRT2 expression sharply declines, leading to a reduction in TaVRN1 expression.Soon afterwards, TaVRN1 expression continues to increase due to its self-regulatory loop mediated by the TaVRN3/TaFDL2 complex (Chen and Dubcovsky, 2012;Deng et al., 2015;Li and Dubcovsky, 2008).Concurrently, TaSOC1 expression is dramatically down-regulated and then has a modest increase.However, TaSOC1 remains a much lower expression level than TaVRN1, causing TaFPF1 to be rapidly up-regulated and ultimately allowing concomitant flower initiation.Overall, TaSOC1, TaVRN1, TaVRT2 and TaFPF1 form a dynamic interactive network to control wheat flowering in response to photoperiod and vernalization.Among them, TaSOC1 acts as an integrator to coordinate cross-talk between the photoperiod and vernalization flowering pathways.
This study confirmed that TaSOC1 expression was regulated by both photoperiod and vernalization, although the underlying regulatory mechanism remained largely unknown.SOC1 can be activated by CONSTANS (CO) through FT in long photoperiods and repressed by FLC, a core component of the vernalization pathway in Arabidopsis (Lee and Lee, 2010).The wheat homologues of FT, CO and FLC have conserved function in flowering time, so their roles in regulating TaSOC1 expression should be investigated (Li et al., 2011;Nemoto et al., 2003;Sharma et al., 2017;Shaw et al., 2020;Yan et al., 2006).Ppd-1 is a key gene in the wheat photoperiod flowering pathway (Shaw et al., 2012(Shaw et al., , 2013)).Phytochrome C is another major regulator to promote flowering under LD conditions in wheat (Chen et al., 2014).TaVRN2 is involved in the vernalization through TaVRN1 and its expression is repressed by SD (Dubcovsky et al., 2006;Li et al., 2011;Trevaskis et al., 2006;Yan et al., 2004b).Thus, it is necessary to identify the relationship of TaSOC1 with Ppd-1, Phytochrome C and TaVRN2 in the photoperiod pathway.Overall, TaSOC1 is an important checkpoint in dissecting regulatory network underlying wheat flowering in response to environmental cues.

Materials and methods
Plant materials, growth conditions and flowering time measurement KN199, an elite semi-winter wheat cultivar, needs approximately 4 weeks at low temperature (2-6 °C) to fully meet its vernalization requirement.KN199 was used as a transgenic recipient and grown with the transgenic lines in a greenhouse under normal conditions (15-18 °C, 16 h light/8 h dark).For vernalization treatments, 3-week-old seedlings were moved into a cold room at 2-6 °C for 30 days (complete vernalization) or 14 days (incomplete vernalization).Photoperiodic treatments of long daylight (16 h light/8 h dark) and short daylight (8 h light/ 16 h dark) were applied to fully vernalized three-leaf seedlings for 3 weeks.For gibberellin treatments, 3-week-old plants were sprayed with a solution containing 0.1 mM GA 3 (Yuan Ye Biotech, Shanghai, China) and 0.05% Tween 20 (BIORIGIN, Beijing) once a week for 5 weeks; water with 0.05% Tween 20 acted as the control.A diverse panel of 166 elite wheat cultivars from the Yellow and Huai River Valleys, the largest wheatproducing region in China, were used to validate the genetic effects of the genes of interest (Datasets S10 and S11).All lines were planted in six environments (Dataset S11).The field trial in each environment was conducted in a completely randomized block design with three replications and a two-row plot with 2 m row length and 20 cm row spacing.
Wheat flowering time was evaluated as heading date which was measured as the number of days from seed germination to heading of the main stem.Approximately, 40 plants of each line were used for statistical analyses.Heading date data for the diversity panel were available from Li et al. (2019).

LCI and BiFC experiments
Nicotiana benthamiana was grown in a growth chamber (20-23 °C, 16 h light/8 h dark) for LCI experiments.LCI assays were carried out as described previously (He et al., 2019).The coding sequences (CDSs) of TaVRN1 (GenBank accession number JQ915056) and TaSOC1 (GenBank accession number AM502888) were cloned into vectors pCAMBIA1300-nLUC and pCAMBIA1300-cLUC, respectively, which were then co-transformed into N. benthamiana leaves by agroinfiltration.LUC activities in the leaves were measured at 50 h post infiltration.The infiltrated leaves were sprayed with 100 lL luciferase assay substrate (Promega, Beijing) and imaged using LB985 NightSHADE (Berthhold Technologies, Bad Wildbad, Germany).Sample analyses were based on at least three biological replicates.
For BiFC analysis, the CDSs of TaVRN1 and TaSOC1 were cloned into pUC-SPYNE(R)173 and pUC-SPYCE(M), respectively.Wheat protoplasts from young leaves of KN199 were prepared and transfected as described in Shan et al. (2014).Fluorescence signals of yellow fluorescent protein (YFP) were analysed with a confocal laser scanning microscope (LSM710, Carl Zeiss).Primers are summarized in Dataset S12.

RT-qPCR analyses
Total RNA was extracted using an EasyPure Plant RNA Kit (ER301; Transgene, Beijing), and first-strand cDNA was synthesized using a PrimeScript RT Reagent Kit with gDNA Eraser (RR047A; TaKaRa, Dalian) according to the manufacturer's instructions.The RT-qPCR mixture comprised 3 lL cDNA template, 0.75 lL of each primer (10 lM) and 7.5 lL 29 Universal SYBR Green Fast qPCR Mix (RK21203; ABclonal, Wuhan) in a final volume of 15 lL.Amplification was performed on a BioRad CFX system following the manufacturer's protocol.The expression levels of target genes were normalized to a wheat actin gene (GenBank accession number AB181991) and calculated by the 2 ÀDDCt method (Schmittgen and Livak, 2008).Each sample was analysed as three biological replicates.The primers used for RT-qPCR are listed in Dataset S12.

Transgenic experiments
For the generation of TaSOC1 overexpression transgenic plants, the CDS of TaSOC1 was cloned into the entry vector pDONR207 and then transferred into the destination vector pUbiGW following the handbook for Gateway cloning (Invitrogen, Carlsbad, CA).The resultant construct was transformed into immature embryos of KN199 by the Agrobacterium (strain EHA105)-mediated method (Ishida et al., 2015).TaVRN1 overexpression lines in this study were generated by Xie et al. (2021).

RNA-seq assays
Samples used for RNA-seq were prepared from leaves of transgenic lines and their TNL at the stem elongation stage (from double ridge to floret differentiation), with three biological replicates.RNA extraction, library construction and sequencing were carried out by Novogene (Beijing).A total 10 Gb of transcriptomic data for each sample was obtained from the Illumina HiSeq 4000 platform.Differentially expressed genes (DEGs) were defined according to |fold change| >2 and P < 0.01.

ChIP-qPCR systems
ChIP-qPCR assays were performed according to the protocol from Cao et al. (2014) with a little modification.The young leaves from 3-week-old seedling of KN199 were harvested.The leaf tissue (1.5 g) was ground in liquid nitrogen and fully resuspended in 30 mL nuclei isolation buffer (10 mM HEPES pH 7.6, 400 mM sucrose, 5 mM KCl, 5 mM MgCl 2 , 5 mM EDTA, 1% formaldehyde, 14.4 mM 2-mercaptoethanol, 0.6% Triton X-100 and 0.4 mM PMSF).The crosslinking reaction in the nuclei was carried out at room temperature for 10 min and stopped by 2 mL 2 M glycine.The nuclei were filtered, centrifuged at 2800 g for 15 min and resuspended in 220 lL nuclei lysis buffer (50 mM Tris-HCl pH 8.0, 1.0% SDS, 10 mM EDTA pH 8.0).The chromatin released from nuclei was sonicated by the Bioruptor Pico device (Diagenode) at 15 s on/90 s off for seven cycles and centrifuged at 20 000 g for 5 min.The supernatant containing sheared chromatin was equally dispensed into two fresh tubes and each was adjusted to 1.1 mL with ChIP dilution buffer (0.01% SDS, 1.1% TritonX-100, 1.2 mM EDTA, 16.7 mM Tris-HCl PH 8.0, 167 mM NaCl).Resultant solution (100 lL) from each tube was transferred into a clean tube and stored at À20 °C as an input.Subsequently, 40 lL protein A + G magnetic beads (Cat#16-663; Merck Millipore) together with 2 lg GFP antibody (Cat#ab290; Abcam) or without antibody as negative control (NC) were added to the tubes containing sheared chromatin; the mixtures were shaken overnight at 4 °C.After incubation, the beads were repeatedly washed in 1 mL low-salt immune complex wash buffer (0.1% SDS, 1.0% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.0, 150 mM NaCl), 1 mL high-salt immune complex buffer (0.1% SDS, 1.0% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.0, 500 mM NaCl), 1 mL LiCl immune complex wash buffer (0.25 M LiCl, 1.0% NP-40, 1.0% sodium dexycholate, 1 mM EDTA, 10 mM Tris-HCl pH 8.0) and 1 mL TE, and eluted by 200 lL elution buffer (1.0% SDS, 100 mM NaHCO 3 ).The eluate was mixed with 20 lL of 5 M NaCl and de-crosslinked by incubation overnight at 65 °C.Afterwards, 4 lL of 0.5 M EDTA, 8 lL of 1 M Tris-HCl (pH 6.5) and 2 lL of proteinase K (20 mg/ mL; Fermentas, Shanghai) were incubated with the de-crosslinked eluate for 2 h at 45 °C.The DNA was extracted from the eluate with chloroform/isoamyl alcohol (24:1) and precipitated with an equal volume of anhydrous ethanol in the presence of 0.3 M sodium acetate (pH 5.2) and 5 lL glycogen (20 mg/mL; Macklin, Shanghai, China).The DNA pellet was washed with 70% ethanol and then dissolved in 50 lL TE.The enrichment of coimmunoprecipitated DNA was calculated by the 2 ÀDDCt method (Schmittgen and Livak, 2008).Primers are listed in Dataset S12.

Y1H and Y3H assays
Y1H was used to validate the interaction of TaSOC1 and TaVRN1 with representative sites 5, 7, 8 and 9 close to the TaFPF1-2B transcription start site.Each CDS of TaSOC1 and TaVRN1 was inserted into the pB42AD vector (Cat#ZT0295, Clontech), respectively.Each of the sites was cloned into the pLacZi reporter vector (Cat#631707, Clontech).The resultant pB42AD and pLacZi vectors were co-transformed into the yeast strain EGY48 using the lithium acetate method following the user manual in Clontech (https://www.takarabio.com/documents/User%20Manual/PT3024/PT3024-1.pdf).Transformants were selected on SD-Trp/Ura plates at 30 °C for 2-4 days and positive transformants were transferred to X-Gal (5-bromo-4-chloro-3-indolyl-b-d-galactopyranoside) plates for galactosidase activity analysis.Primers for Y1H are listed in Dataset S12.
Y3H assays were used to test whether TaSOC1 can interfere with the interaction between TaVRT2 and TaVRN1.The CDSs of TaVRT2 (GenBank accession number AAY43789) and TaSOC1 were inserted in MCS I or MCS II, respectively, of the pBridge vector (TaKaRa, Dalian), while that of TaVRN1 was fused with the pGADT7 vector (TaKaRa).The resulting constructs and the empty vectors (used as controls) were co-transferred into yeast AH109 cells using the lithium acetate method.Transformants were selected on the synthetic defined (SD) medium without leucine (L) and tryptophan (W) (SD-L/W).Positive transformants were used to identify the interaction among the proteins of interest on the SD mediums without L, W and histidine (H) (SD-L/W/H) or without L, W, H and methionine (M) (SD-L/W/M/H).The detailed procedure of yeast transformation and protein interaction assays were performed as described in the Yeast Protocols Handbook (https://www.takarabio.com/documents/User%20Manual/PT3024/PT3024-1.pdf).Primers for Y3H are listed in Dataset S12.

Identification of allelic variations and development of gene-specific markers
Sequence variants of the genes of interest were identified from the WheatUnion databases (Guo et al., 2020; http://wheat.cau.edu.cn/WheatUnion/).Haplotypes of target genes were defined by the genomic variation information of 100 modern Chinese cultivars (Hao et al., 2020;Dataset S9).Gene-specific markers were developed to genotype allelic variations.Genotypic and phenotypic data of the 166 elite wheat cultivars from the Yellow and Huai River Valleys of China were used for association analysis to mine superior haplotypes.

Statistical analyses
Student's t-tests were performed to assess significant differences (P < 0.05) in flowering time, gene expression levels and chromatin enrichment between contrasting materials using SAS 9.4.Epistatic effects of TaSOC1 and TaVRN1 on flowering time were analysed by two-way ANOVA in SAS 9.4.Mean values and error bars were calculated with the AVERAGE and STDEV functions, respectively, in Microsoft Excel.and barley cultivars are labelled in the left panel.Red arrowhead shows TaSOC1.(d) Synteny comparison of TaSOC1-like genes with rice OsMADS56 and barley HvSOC1-like1.Figure S5 Responses of TaSOC1-like genes to vernalization (a) and photoperiod (b).In (a): V0, the day before vernalization; V7, V21 and V35, 7th, 21st and 35th day during vernalization; V35N7 and V35N21 indicate the 7th and 21st day after vernalization, respectively; orange and blue lines represent vernalized and nonvernalized (negative control) treatments, respectively.In (b), sampling time-points include the day before photoperiod treatments (1), 5th (2), 10th (3), 15th (4) and 20th (5) day during photoperiod treatments, and 5th (6) and 10th (7) day after photoperiod treatment; orange and blue lines represent short-day (SD) and long-day (LD) treatments, respectively; error bars, standard deviations of three biological replicates.Figure S6 Yeast three hybrid arrays for competitive interaction of TaSOC1 and TaVRT2 with TaVRN1.SD-L/W, SD-L/W/H, SD-L/W/ M/H represent the plates with synthetic defined (SD) media lacking Leu/Trp, Leu/Trp/His, Leu/Trp/Met/His, respectively; 10 0 , 10 À1 , 10 À2 and 10 À3 indicate gradient dilution of yeast concentration.AD, activation domain; BD, DNA binding domain.

Figure 2
Figure 2 TaSOC1 knockout lines (TaSOC1-KO) and phenotypic investigation of TaSOC1-KO and KN199 under differing vernalization conditions.(a) Sequencing-based identification of TaSOC1 knockout (KO) mutants.PAM and mutant sites are shown in red and blue, respectively; orange bars and black lines represent exons and introns, respectively; ATG, start codon; sgRNA, small guide RNA; TGA, stop codon.Statistical analyses (b) and phenotype display (c) of the heading date of TaSOC1-KO and KN199 following 30 days (complete) of vernalization.Statistical analyses (d) and phenotype display (e) of heading dates of TaSOC1-KO and KN199 following 14 days (partial) vernalization.Statistical analyses (f) and phenotype display (g) of heading dates of TaSOC1-KO and KN199 under non-vernalization conditions (n = 40 plants).KO1, KO2 and KO3 are representative TaSOC1-KO.KN199 was used as the negative control (NC).*P < 0.05, **P < 0.01, ns, not significant; scale bar, 30 cm.
; Dataset S6).Moreover, TaSOC1 and TaVRN1 were enriched at certain sites close to the TSS motifs in ª 2023 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 635-649 TaFPF1-2B (Figure 5f).Thus, TaFPF1-2B could function as a common downstream gene of TaSOC1 and TaVRN1 to modulate flowering.To further identify the relationship among TaSOC1, TaVRN1, TaVRT2 and TaFPF1 in flowering initiation, we investigated changes in expression of these genes in response to pre-flowering vernalization and photoperiod patterns simulating temperature and photoperiod conditions in the field (Figure S7a).Based on results, we proposed a regulatory model of TaSOC1 in orchestrating TaVRN1, TaVRT2 and TaFPF1 to modulate flowering through integration of photoperiod and vernalization ª 2023 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 635-649 signals (Figure

Figure 5
Figure 5 Identification of common downstream target genes regulated by TaSOC1 and TaVRN1 and detection of TaSOC1 and TaVRN1 enrichment at different sites of TaFPF1.(a) Identification of differentially expressed genes (DEGs) in TaSOC1 overexpression lines (TaSOC1-OE) and TaVRN1 overexpression lines (TaVRN1-OE) compared to respective transgenic null lines (TNL) based on RNA-seq assays.The overlapping part of the blue and orange circles indicates 15 DEGs shared by TaSOC1 and TaVRN1.RNA-seq (b) and reverse transcription quantitative PCR (RT-qPCR) (c) assays for expression levels of TaFPF1-2B in TaSOC1-OE and TaVRN1-OE (n = three biological replicates).FPKM, fragments per kilobase million.OE, overexpression line; TNL, transgenic null line.Note: TaFPF1 expression levels in TaSOC1-OE and TaVRN1-OE were compared to the counterparts of their respective TNLs.(d) TaFPF1-2B expression analyses in a F 2 segregating population derived from cross TaSOC1-OE 9 TaVRN1-OE using RT-qPCR (n = three biological replicates).NC, negative control (TNLs identified from the F 2 population).(e) Schematic of detection sites in the TaFPF1 promoter.The blue arrowheads indicate the positions of CArG-like motifs, and black boxes 1-11 represent different detection sites; orange box, first exon; TSS, transcription start site; ATG, start codon.(f) TaSOC1 and TaVRN1 enrichment at different sites of the TaFPF1 promoter using chromatin immunoprecipitation quantitative PCR (ChIP-qPCR) assays (n = three biological replicates).Anti-GFP, GFP antibody; NIC, non-immune control; *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant.(g) Validation of TaSOC1 and TaVRN1 binding to the representative sites 5, 7 and 8, close to transcription sites of TaFPF1, by yeast one-hybrid (Y1H) assays.5, 7 and 8 indicate the sites 5, 7 and 8 in the TaFPF1 promoter; 5 m, 7 m and 8 m represent their respective mutated versions.The promoter in LacZi empty vector is used as a negative DNA bait.AD represents activation domain in the vector pB42AD.
Figure S7 model of wheat flowering synergistically modulated by TaSOC1, TaVRN1, TaVRT2 and TaFPF1.(a) Expression analyses of TaSOC1, TaVRN1, TaVRT2 and TaFPF1 under simulated winter conditions using quantitative real-time PCR (n = three biological replicates).The orange and blue lines represent short-day (SD) (simulated winter) and long-day (LD) (control group) treatments during winter (low temperature), respectively; LD were applied before and after winter.The numbers on the abscissa represent the sampling time points, including the 10th (BV1) and 5th (BV2) day before winter, the 5th (VSD5), 10th (VSD10), 15th (VSD15) and 20th (VSD20) day during winter, and the 5th (NS5), 10th (NS10), 15th (NS15) and 20th (NS20) day after winter.(b) Schematics of the dynamic expression abundances of TaSOC1, TaVRN1, TaVRT2 and TaFPF1 before, during and after winter.(c) Molecular model of TaSOC1, TaVRN1 and TaVRT2 modulating TaFPF1 expression.Colour intensity indicates the level of protein accumulation and arrow thickness represents the transcriptional activity of TaFPF1.Note: single TaVRN1 or TaSOC1 as well as their homo-dimers are possible to bind to the TaFPF1 promoter.Dataset S1 Statistical analysis on the heading date of TaSOC1-OE and TNL under different vernalization conditions.Dataset S2 Statistical analysis on the heading date of TaSOC1-KO and NC under different vernalization conditions.Dataset S3 Statistical analysis of the heading date of segregating population from the cross of TaSOC1-OE and TaVRN1-OE lines.Dataset S4 Differentially expressed genes between TaSOC1 overexpression lines and their transgenic negative lines.Dataset S5 Differentially expressed genes between TaVRN1 overexpression lines and their transgenic negative lines.Dataset S6 Overlapping downstream differentially expressed genes of TaSOC1 and TaVRN1.Dataset S7 Statistical analysis on the heading date of TaSOC1-OE and TNL under GA or photoperiod treatments.Dataset S8 Statistical analysis on the heading date of TaSOC1-KO and NC under different photoperiod treatments.Dataset S9 Identification for the haplotypes of TaSOC1 and TaSOC1-5A based on resequencing data of 100 Chinese wheat cultivars.Dataset S10 Haplotypes of TaSOC1 and TaSOC1-5A, and heading date of 166 wheat cultivars from the Yellow and Huai River Valleys.Dataset S11 Association analyses of TaSOC1 and TaSOC1-5A haplotypes with heading date.Dataset S12 The list of primers used in this study.ª 2023 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 635-649