The AP2 transcription factor BARE RECEPTACLE regulates floral organogenesis via auxin pathways in woodland strawberry

Abstract

During flower development, various floral organs are formed to ensure fertilization and fruit set. Although the genetic networks underlying flower development are increasingly well understood, less is known about the mechanistic basis in different species. Here, we identified a mutant of woodland strawberry (Fragaria vesca), bare receptacle (bre), which produces flowers with greatly reduced carpels and other floral organs. Genetic analysis revealed that BRE encodes an APETALA2 (AP2) transcription factor. BRE was highly expressed in floral meristems and floral organ primordia. BRE could directly bind the GCC-box motif in the YUCCA (YUC) auxin biosynthesis genes FveYUC4 and FveYUC2 and promote their expression. The yuc4 mutant had fewer floral organs, and the bre yuc4 double mutant had similar numbers of petals and carpels to bre. Auxin homeostasis and distribution were severely disrupted in bre. Although auxin application or FveYUC4 overexpression did not rescue the bre phenotypes, bre was hypersensitive to treatment with the polar auxin transport inhibitor N-1-naphthylphthalamic acid (NPA). In addition, BRE was able to directly bind and regulate the expression of five other auxin pathway genes. Overall, these results demonstrate that BRE is required for floral organogenesis, particularly carpel initiation, and acts through the auxin pathway in strawberry.

An APETALA2 transcription factor, BARE RECEPTACLE modulates the expression of several auxin-related genes and thus is essential for floral organogenesis in woodland strawberry.

Introduction

The flower is important for sexual reproduction in plants. Four specialized floral organs, namely sepals, petals, stamens, and carpels, are formed from a flower meristem (FM) to produce seeds or fruit. In the diploid woodland strawberry (Fragaria vesca), each flower typically produces 5 sepals (plus 5 bracts), 5 petals, approximately 20 stamens, and more than 100 carpels, following an intrinsic developmental process that was divided into 12 stages (Hollender et al. 2012). In contrast to other well-known plant models, the numerous carpels are born individually on a raised receptacle dome and remain unfused to each other. A few genes have been reported to regulate floral organ initiation in strawberry. For instance, the double mutant of MicroRNA164 and CUP-SHAPED COTYLEDON2, fvemir164a fvecuc2a, lacks petal and stamen whorls (Zheng et al. 2019). The loss of axillary meristems (lam) flowers fail to produce any stamens in most cases (Feng et al. 2021). However, the carpels are formed normally in these mutants. It is not known whether there are vital genes in strawberry that are responsible for floral organogenesis, in particular carpel initiation.

Floral organ primordia are generated from FMs. In this process, auxin response maxima co-localize with sites of floral organ primordia initiation (van Mourik 2012; Goldental-Cohen et al. 2017); thus, dynamic auxin synthesis, polar transport, and signal responses play essential roles in organ initiation. The YUCCA (YUC) genes encode flavin-containing monooxygenases that convert indole-3-pyruvic acid (IPyA) to indole-3-acetic acid (IAA), a key step in auxin biosynthesis (Mashiguchi et al. 2011; Stepanova et al. 2011; Won et al. 2011). The GRETCHEN HAGEN 3 (GH3) family of proteins conjugates various amino acids to IAA for storage or degradation (LeClere et al. 2002; Staswick et al. 2005). To attenuate bioactive IAA levels, the enzymes DIOXYGENASE FOR AUXIN OXIDATION (DAO) and UGT74D1 sequentially convert IAA to 2-oxindole-3-acetic acid (OxIAA) and OxIAA-Glc (Zhao et al. 2013; Tanaka et al. 2014; Porco et al. 2016; Zhang et al. 2016). For precise auxin distribution, the PIN-FORMED (PIN) proteins are polar auxin efflux carriers whose activity is regulated by PINOID (PID) kinase (Benková et al. 2003; Friml et al. 2004; Adamowski and Friml 2015). In the signaling pathway, free auxin is perceived by the TRANSPORT INHIBITOR RESPONSE1/AUXIN-RELATED F-BOX (TIR1/AFB) receptors, which allow the binding and degradation of AUX/IAA (INDOLE-3-ACETIC ACID) to release the repression of AUXIN RESPONSE FACTORs (ARFs) (Tan et al. 2007; Dos Santos Maraschin et al. 2009). Mutations in auxin pathway genes often cause defects in flower development or even the absence of flowers (Okada et al. 1991; Cheng et al. 2006; Wu et al. 2020). The number of each type of floral organ is substantially altered in the fveyuc4 mutants (Lu et al. 2023), suggesting that the auxin pathway also controls floral organ development in strawberry.

In Arabidopsis (Arabidopsis thaliana), the AP2 transcription factor ENHANCER OF SHOOT REGENERATION 1 (ESR1)/DORNRÖSCHEN (DRN) was found to be a positive regulator of shoot regeneration (Banno et al. 2001), meristem activity, and lateral organ development (Kirch et al. 2003). DRN and/or its close homolog DRN-LIKE (DRNL) are also involved in embryo patterning (Chandler et al. 2007), stamen emergence (Nag et al. 2007), FM identity, and floral organ initiation (Chandler and Werr 2017). The rice (Oryza sativa) homolog DEFECTIVE STIGMA AND PANICLE (DSP) controls stigma, panicle, and tiller formation (Yu et al. 2023). EARLY BUD BREAK1 (EBB1) accelerates bud burst after winter dormancy in poplar and other woody perennials (Yordanov et al. 2014; Anh Tuan et al. 2016; Zhao et al. 2020). The tomato (Solanum lycopersicum) homolog LEAFLESS (LFS) is essential for the production of cotyledons and leaves (Capua and Eshed 2017). Knockout of CsDRNL results in a lack of oil glands in leaves and a smooth leaf margin phenotype in Citrus (Wang et al. 2024). These reports indicate that disruption of DRN/DRNL homologs causes diverse defects in different plant species.

DRN and DRNL show a complex interaction with the auxin pathway. The ARF protein MONOPTEROS (MP) directly controls DRN and DRNL expression by binding to auxin response elements (AuxREs) (Cole et al. 2009; Luo et al. 2018b; Dai et al. 2023), suggesting that DRN/L act downstream of auxin. However, DRNL is specifically expressed in floral organ founder cells prior to the establishment of auxin maxima, which is not affected by mutation of AuxREs (Chandler et al. 2011). DNRL can physically interact with MP to activate the cytokinin pathway (Dai et al. 2023). In rice, DSP directly promotes the expression of PID during stigma development (Yu et al. 2023). PpEBB1 could bind to a GCC-box-like element and activate the expression of YUC1 (Zhao et al. 2021). These results suggest that DRN/L also interacts closely with auxin genes to influence signaling outputs. However, the known target genes of DRN or DRNL in the auxin pathway are still very limited.

In this work, we have identified an ethyl methanesulfonate (EMS)-induced woodland strawberry mutant bare receptacle (bre), which develops flowers with much fewer floral organs of abnormal morphology. The causal gene BRE encodes a single homolog of DRN and DRNL in woodland strawberry. Furthermore, BRE can directly bind the GCC-box motif in several genes involved in auxin homeostasis, transport, and signaling during floral organogenesis. These results provide further insights into the biological role and molecular mechanism of an AP2 transcription factor during flower development in strawberry.

Results

The bare receptacle mutant in woodland strawberry has fewer floral organs

To isolate essential genes regulating flower development, an EMS mutagenized population of woodland strawberry was screened at the M₂ generation. One mutant showed greatly reduced numbers of floral organs in all four whorls, particularly the carpels, and was therefore named bare receptacle (bre) (Fig. 1, A to C). Specifically, the mean number of sepals and bracts is 10.2 in wild type (WT) and 2.5 in bre, the mean number of petals is 5.2 in WT and 0.2 in bre, the mean number of stamens is 20 in WT and 2.6 in bre, and the mean number of carpels is 159.5 in WT and 5.5 in bre (Fig. 1B). The severity of the developmental defects varied between flowers. Most flowers had closed sepals, while a small fraction of flowers had no sepals, petals, stamens, or carpels (Supplementary Fig. S1). The bre petals had deeper serrations at the tip (Fig. 1D). When the sepals or bracts were removed, a few abnormal carpels and stamens were visible in bre (Fig. 1E). In particular, each wild-type carpel had a long style, a green-walled ovary, and a white ovule inside (Fig. 1F). In contrast, the bre carpels had only the white naked ovule-like organs, resulting in female sterility due to the missing styles (Fig. 1F). In WT, each carpel became an achene with a seed inside after fertilization, while the receptacle became the juicy flesh (Fig. 1G). In bre, most receptacles did not enlarge due to the lack of fertilized achenes. Occasionally, the bre receptacle was enlarged without any achenes (Fig. 1G).

Figure 1.

Phenotypic characterization of the bre mutant in woodland strawberry. A) Plants of wild type (WT) and the bare receptacle (bre) mutant. B) Number of the floral organs per flower in WT and bre. Data are mean ± SD. ***, P < 0.001, Student's t-test. C) Flowers of WT and bre. D) Petals of WT and bre. E) Flowers of WT and bre with the sepals and petals removed. F) Carpels of WT and bre. Inset of the WT image is the dissected ovule. G) Fruits of WT and bre. H) SEM images showing the flower buds of WT and bre at different developmental stages. Br, bract; Pe, petal; Se, sepal; St, stamen; Ca, carpel. Scale bars: (A) 1 cm; (C) 5 mm for WT and 1 mm for bre; (D and E) 1 mm; (F and H) 100 μm; (G) 5 mm.

To compare the flower development in bre and WT, scanning electron microscopy was utilized to examine dissected flower buds at early developmental stages. In WT, the five sepals first emerged from the periphery of the flower meristem dome, the petal primordia were initiated inside the bracts, then the stamen primordia arose rapidly, and finally, the carpel primordia emerged acropetally from the receptacle dome (Fig. 1H). In contrast, floral organ development was severely disrupted in bre. Much fewer sepal and bract primordia were initiated and disorganized, making it difficult to define the developmental stages. At later stages, very few carpel primordia were initiated, and the stamens were also malformed (Fig. 1H). Notably, epidermal hairs arose from the bre receptacle, which were completely absent in WT. We also found that the size of the FMs was smaller in bre than in WT (Supplementary Fig. S2), but this is not the main reason for fewer floral organs. In addition, a large proportion of leaves had one leaflet or two leaflets in bre instead of three leaflets in WT, suggesting a reduced leaf complexity (Supplementary Fig. S3, A and B). We also noticed that the majority of bre seedlings had only one cotyledon, suggesting a disruption in embryogenesis (Supplementary Fig. S3C). These observations indicate that BRE is primarily required for floral organogenesis and also plays a role in vegetative organ growth in strawberry.

BRE encodes an AP2 transcription factor homologous to Arabidopsis DRNL and DRN

The bre mutant was backcrossed to WT. The F₁ progeny are all WT, suggesting that this is a recessive mutation. To isolate the causal gene, the WT and bre mutant plants in the backcrossed F₂ population were bulk segregated for genome resequencing (Supplementary Table S1). Data analysis revealed a missense mutation closely associated with the mutant phenotype. This mutation occurred in the coding region of FvH4_4g25630 (Supplementary Table S2) and caused an amino acid change from Gly to Arg at the residue 81, which is highly conserved in the N-terminal AP2 domain (Fig. 2, A and B). To confirm this result, 50 F₂ mutants were individually sequenced at this locus by Sanger sequencing. All individuals were homozygous for this mutation, and at least one individual was heterozygous for other candidates. In addition, we found a second EMS mutant, called bre-2, which exhibited the same leaf and flower phenotypes as bre (Supplementary Fig. S4). In bre-2, a mutation was found in FvH4_4g25630 that causes a premature stop codon (Fig. 2, A and B). Therefore, FvH4_4g25630 is likely BRE.

Figure 2.

BRE encodes an AP2 transcription factor. A) Schematic diagram depicting the gene model of BRE (FvH4_4g25630) and the causative mutations in bre and bre-2. AP2 and ESR indicate the two conserved domains. B) Sequence logo showing the conserved amino acids in the AP2 domain. BRE and 12 close homologs were used for sequence alignment. C) Phylogenetic tree of BRE and its homologs in different plant species. Bootstrap values are percentages from 1,000 replicates. D) Subcellular localization of BRE-GFP in Nicotiana benthamiana leaves. Green indicates the BRE-GFP fluorescence. Magenta indicates the fluorescence of nuclear marker VirD2NLS-mCherry. E) Open flowers of WT, bre, and pBRE:BRE bre (Comp for short, L1 and L2). Images were digitally extracted for comparison. F) Number of the floral organs per flower in WT, bre, and Comp (L1 and L2). Data are mean ± SD, and significant differences at the P < 0.05 level are indicated by different letters as determined by Tukey's test. G) Relative expression levels of BRE in WT, bre, and Comp (L1 and L2) examined by RT-qPCR. Data are mean ± SD from three biological replicates. **, P < 0.01; ***, P < 0.001; Student's t-test. Scale bars: (D) 20 μm; (E) 1 mm for bre and 5 mm for others.

According to the phylogenetic tree, BRE/FvH4_4g25630 encodes the single strawberry homolog of the Arabidopsis DRN and DRNL, with a higher similarity to DRNL than to DRN (Fig. 2C). BRE was also grouped with tomato LEAFLESS and poplar EBB1 (Yordanov et al. 2014; Capua and Eshed 2017). FvePUCHI (FvH4_4g17630) and FveLFP (LEAFYPETIOLE, FvH4_4g22310) were used as outgroups. However, the similarity between BRE and its homologs is rather low outside the AP2 and ENHANCER OF SHOOT REGENERATION (ESR) domains (Supplementary Fig. S5). A transient expression assay in the Nicotiana benthamiana leaves revealed that the fused BRE-GFP protein was exclusively localized to the nucleus, consistent with its role as a transcription factor (Fig. 2D).

To further validate the gene identity, the complementation construct pBRE:BRE, including the 1,928-bp promoter, the full-length coding sequence, and 766-bp downstream sequence after the stop codon, was stably transformed into the bre mutant. As a result, all 12 transgenic lines (called Comp for short) showed similar phenotypes. Compared to bre, flower morphology was generally restored by pBRE:BRE in the T₀ generation (Fig. 2E). The number of floral organs was significantly increased in Comp transgenic lines compared to bre and was comparable to WT (Fig. 2F). In addition, Comp leaves were mostly composed of three leaflets and resembled WT leaves except for the upward curl (Supplementary Fig. S6). According to the RT-qPCR results (Fig. 2G), the expression level of BRE was significantly higher in bre than in WT, suggesting feedback regulation per se, and even higher in the Comp lines than in bre, which could be caused by the insertion sites of the transgene or the absence of some regulatory sequences in the construct. Nevertheless, these results show that BRE encodes an AP2 transcription factor.

BRE is highly expressed in the young floral organs

Transcriptome data from woodland strawberry (Li et al. 2019a) showed that BRE was highly expressed in the REM (stage 6 to 7 receptacle), SFM (stages 1 to 4 flower meristem), young carpels of stages 6 to 7, and embryo 3 (heart-stage embryo), but low or absent in fruit tissues (cortex, pith, and wall) (Fig. 3A). To confirm its expression pattern, a total of 10 pBRE:GUS transgenic woodland strawberry lines were generated with similar expression patterns as examined by GUS staining. According to GUS signals, BRE was widely expressed throughout young flower buds, including young stamens and carpels, and was preferentially expressed in the styles at stage 9 (Fig. 3B). As the transgenes may be ectopically expressed, RNA in situ hybridization for BRE was also performed on WT flower buds. Similarly, BRE transcripts were detected throughout the flower meristem, young carpels and stamens, stigmas, ovules, leaf primordia, and young leaf blades (Fig. 3C). These results indicate that BRE may play a vital role in the initiation and development of floral organs.

Figure 3.

BRE expression patterns in wild-type woodland strawberry. A)  BRE expression levels in different tissues of woodland strawberry represented by TPM (Transcripts Per Million) obtained from RNA-seq (Li et al. 2019a). B) GUS signals in flower buds at stages 2 to 9 of pBRE:GUS transgenic lines at wild-type background. Longitudinal sections are shown for the flower buds at stages 6 and 9. Scale bars: 100 μm. C) RNA in situ hybridization images showing BRE expression in floral organs and a shoot tip using the antisense probe and the sense probe as control. Br/Se, bract/sepal; St, stamen; Ca, carpel; Sti, stigma; Ov, ovule; Po, pollen; LP, leaf primordium. Scale bars: 100 µm.

Identification of BRE-regulated genes by RNA-seq and DAP-seq

To identify genes regulated by BRE, transcriptome analysis was performed for the young flower buds of bre and WT at stages 1 to 7 with three biological replicates (Supplementary Table S3). Pairwise comparison revealed that a total of 1,449 genes were significantly upregulated, while only 387 genes were downregulated in the bre flowers compared to WT (fold change >2, P-adj < 0.05) (Fig. 4A; Supplementary Data Set 1). The downregulated genes are enriched in the Gene Ontology (GO) in terms of organ morphogenesis and auxin homeostasis (Fig. 4B; Supplementary Data Set 2). In particular, the class A (FveAP1) and class B (FveAP3, FvePIa, and FvePIb) flower development genes were greatly reduced in bre flowers (Fig. 4C), consistent with the loss of most floral organs. Notably, several auxin biosynthesis, transport, and signaling genes were also differentially expressed in bre flowers (Fig. 4C), such as FveYUC2, FveYUC4, FvePIN1, FvePIN3, FvePID2, FveIAAs, and FveGH3.1, suggesting that the auxin pathway was severely disrupted in the bre mutant.

Figure 4.

Identification of BRE-regulated genes by RNA-seq and DAP-seq. A) Pie chart showing the number of differentially expressed genes (DEGs) in flower buds of bre compared to WT (fold change >2, P < 0.05). B) Enriched Gene Ontology (GO) terms of the downregulated genes in bre (P < 0.05). y-Axis indicates -log₁₀ (P-value). C) Heatmap showing expression levels of the selected DEGs involved in flower development and the auxin pathway. The number in each box indicates the average TPM (Transcripts Per Million). D) Genomic distribution of the enriched BRE-IP peaks in DAP-seq (q < 0.05). E) Sequence logos showing the enriched BRE-binding motifs derived from DAP-seq. F) Common and specific genes between the potential BRE-binding genes obtained in DAP-seq and DEGs in the flower buds of bre obtained in RNA-Seq.

To identify the BRE-binding sites at a genome-wide scale, DNA affinity purification sequencing (DAP-seq) was employed (O'Malley et al. 2016; Bartlett et al. 2017). A total of 9,427 genes were identified from the 10,801 enriched BRE-IP peaks across the entire genome (Supplementary Data Set 3, q < 0.05). Genomic distribution of these peaks revealed that 34.58% were located in the first exon, 13.42% were located in other exons, 16.3% were located within the promoter region (−2 kb to the transcription start site), 9.27% were located in the 5'-UTRs, and small proportions were located in introns and 3'-UTRs (Fig. 4D). The enriched BRE-binding motifs mostly contain the GCC-box core sequence GCCGCC (Fig. 4E), a typical motif recognized by the AP2/ERF (ethylene responsive factor) transcription factors (Chandler 2018). Among the differentially expressed genes (DEGs) in the bre flowers, 119 downregulated and 518 upregulated genes contain potential BRE-binding sites (Fig. 4F), providing promising candidate downstream genes.

BRE directly binds the genic regions of FveYUC4 and FveYUC2

According to the DAP-seq data, two BRE-IP peaks were located at the promoter and second exon of FveYUC4, and one BRE-IP peak was present in the first exon of FveYUC2, suggesting that these two genes might be direct targets of BRE (Fig. 5A). To test this hypothesis, the yeast one-hybrid (Y1H) assay was performed. Positive interaction signals were detected in the three target regions of FveYUC4 and FveYUC2 by BRE (Fig. 5B). In the electrophoretic mobility shift assay (EMSA), the DNA-protein complex showed retarded migration when the purified recombinant BRE was mixed with the target sequences in FveYUC4 or FveYUC2 (Fig. 5C). Complex formation was attenuated by adding increasing amounts of the unlabeled competitor. For the FveYUC4 promoter, no binding was observed when mutant BRE protein with the bre mutation or mutant DNA target sequence at the GCC-box was added. For the FveYUC4 and FveYUC2 exons, the mutated target sequence can’t compete with the wild-type target sequence at the ×25 concentration and therefore didn’t reduce the binding affinity between BRE and the wild-type target sequence. The dual luciferase reporter assay was used to investigate the effects of BRE on the transcriptional activity of the target sequences. Co-transformation of 35S:BRE with the FveYUC4 and FveYUC2 promoter- or exon-containing LUC reporter significantly increased the LUC/REN ratio compared to the empty vector control, whereas the mutated BRE protein failed to do so (Fig. 5D), indicating that BRE is a positive regulator of FveYUC4 and FveYUC2 expression. Furthermore, we generated 35S:BRE-GFP transgenic woodland strawberry lines by stable transformation (Supplementary Fig. S7), and performed a chromatin immunoprecipitation (ChIP)-qPCR assay to verify BRE binding to these motifs in vivo. Consistently, the results showed significant enrichment of BRE in the GCC-box motifs of FveYUC4 and FveYUC2 after immunoprecipitation (Fig. 5E). Moreover, the expression levels of FveYUC4 and FveYUC2 were significantly decreased in bre and increased in the Comp lines (pBRE:BRE bre) (Fig. 5F). These results indicate that FveYUC4 and FveYUC2 may serve as direct targets of BRE in strawberry.

Figure 5.

Direct binding of BRE to the genic regions of FveYUC4 and FveYUC2. A) BRE-IP peaks in the promoter and exon of FveYUC4 (FvH4_2g29930) and the exon of FveYUC2 (FvH4_1g04800) obtained by DAP-seq. P1, P2, and P3 indicate the regions examined by ChIP-qPCR. B) Yeast one-hybrid assay showing the interaction between BRE and the promoter or exon of FveYUC4/2. C) EMSA showing the interaction between BRE and the promoter or exon of FveYUC4/2. Blue letters in the oligo indicate the core binding motif. Probes are labeled with FAM. BRE^m contains the bre mutation. D) Dual luciferase assay showing BRE-induced expression of FveYUC4/2. BRE was driven by 35S promoter as an effector. The promoter and exon of FveYUC4/2 were used to drive LUC as a reporter. EV indicates empty vector. BRE^m contains the bre mutation. Data are mean ± SD of three biological replicates, and significant differences at the P < 0.05 level are indicated by different letters as determined by Tukey's test. E) ChIP-qPCR assay in young flower buds of 35S:BRE-GFP in the regions indicated in (A). F) Expression levels of FveYUC4/2 in WT, bre, and Comp (pBRE:BRE bre, L1 and L2) examined by RT-qPCR. In (E) and (F), data are mean ± SD from three biological replicates. *, P < 0.05; **, P < 0.01; Student's t-test.

FveYUC4 acts downstream of BRE in floral organogenesis

The fveyuc4 mutants in woodland strawberry have an altered number of floral organs (Lu et al. 2023). To better understand the role of FveYUC4 in floral organ initiation, young flower buds of the strong mutant fveyuc4-2 (yuc4-2 for short) were observed by SEM. At stage 6, the stamens and receptacle were exposed due to fewer sepals and bracts (Fig. 6A). In some yuc4-2 flowers, no carpels were formed at the receptacle apex at later stages (Fig. 6A). These results suggest that the yuc4-2 mutants have similar floral organ initiation defects to bre, although to a lesser extent. To better understand the relationship between BRE and FveYUC4, their expression patterns were compared. Based on woodland strawberry transcriptome data (Kang et al. 2013; Hollender et al. 2014; Li et al. 2019b), BRE, FveYUC4 and FveYUC2 were co-expressed in SFM, REM, young carpels at stages 7 to 10, embryos 3 to 4 and young leaves (Fig. 6B). Previous transcriptome analysis revealed 278 downregulated genes and 559 upregulated genes in yuc4-2 young flower buds at stages 1 to 7 compared to WT (fold change >2, P-value < 0.05) (Lu et al. 2023). We found that 36 genes were commonly downregulated, and 311 genes were commonly upregulated in bre and yuc4-2 (Fig. 6C). The expression levels of the five commonly downregulated genes, including FveAP3, FvePIa, FvePIb, FveLFYd (FvH4_3g03570) and FveGH3.2 (FvH4_1g16980), were validated in the bre and yuc4-1 flowers by RT-qPCR (Fig. 6D). These data indicate that BRE and FveYUC4 regulate a subset of common genes during flower development. To test the genetic interactions between FveYUC4 and BRE, we constructed the yuc4-1 bre double mutant (Fig. 6E). The number of bracts/sepals or stamens in bre yuc4-1 was lower than that in bre (Fig. 6F), suggesting a synergistic interaction between BRE and YUC4. However, the number of petals or carpels in bre yuc4-1 was not significantly different from that in bre (Fig. 6F), suggesting that the two genes work in a common pathway. The defects of the yuc4 mutants were significantly milder than those of the bre mutants (Fig. 6, E and F), suggesting that BRE may also regulate other genes in addition to FveYUC4.

Figure 6.

FveYUC4 acts downstream of BRE in floral organ initiation. A) SEM images of yuc4-2 flower buds. Br, bract; Se, sepal; St, stamen; Ca, carpel. Arrows indicate the receptacle apex without growing carpels. Scale bars: 100 μm. B) Heatmap showing the expression levels of BRE, FveYUC2, and FveYUC4 in different tissues of woodland strawberry, represented by log₂TPM obtained from RNA-seq. Asterisks indicate the highly expressed tissues of these genes. C) Venn diagram showing the common DEGs in flower buds of bre and yuc4-2 compared to WT (fold change >2, P < 0.05). D) Expression levels of the five common genes in WT, bre, and yuc4-1 examined by RT-qPCR. Data are mean ± SD of three biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; Student's t-test. E) Flowers of WT, bre, yuc4-1, and bre yuc4-1. Scale bars: 1 mm. F) Number of the floral organs per flower in WT, bre, yuc4-1, and bre yuc4-1, respectively. Data are mean ± SD, and significant differences at the P < 0.05 level are indicated by different letters as determined by Tukey's test. n = 28 for WT, n = 30 for bre, n = 25 for yuc4-1 and yuc4-1 bre.

The auxin pathway is severely disturbed in the bre mutant

Since the YUC family genes FveYUC4 and FveYUC2 were transcriptionally regulated by BRE, we hypothesized that endogenous IAA levels might be altered in bre. To test this hypothesis, stages 1 to 7 flower buds of WT and bre were used to quantify auxin content. As a result, there was no significant difference in free auxin between bre and WT (Fig. 7A), although two of the three biological replicates had obviously lower free auxin levels in bre than in WT. However, the precursor for IAA synthesis, tryptophan (Trp), and the IAA-amino acid conjugate IAA-Ala were significantly increased, whereas OxIAA and IAA-Asp were significantly decreased in the bre mutant (Fig. 7A). These results suggest that auxin homeostasis is disrupted by BRE. To further observe the auxin signaling output in bre, a previously described DR5ver2:GUS reporter line in woodland strawberry was crossed into bre (Feng et al. 2019). GUS staining showed that the blue signal was distributed throughout the flower buds at stages 2 to 7 and restricted to the stigma at stage 9 after 7 h of staining in WT, whereas the blue signal was much weaker and mostly located in the vascular bundles in bre (Fig. 7B). Longer GUS staining (19 h) gave similar results (Supplementary Fig. S8), suggesting a reduced auxin signaling output in bre. Furthermore, the bre mutant was cultured and treated with different concentrations of indole-3-butyric acid (IBA), but the floral organ defects were not rescued by this treatment (Fig. 7C). Similarly, application of naphthalene acetic acid (NAA) also did not restore the leaf phenotypes (Supplementary Fig. S9). Previously, we obtained pFveYUC4:FveYUC4 transgenic plants that exhibited a high auxin phenotype due to FveYUC4 overexpression (Lu et al. 2023). This construct was introduced into bre by crossing to increase endogenous IAA levels (Fig. 7D). The expression level of FveYUC4 was significantly increased in pYUC4:YUC4 bre compared to that in bre according to RT-qPCR (Fig. 7E). Consistently, the leaf petioles of bre became longer upon the introduction of pYUC4:YUC4 (Fig. 7F). However, the floral organ defects remained the same in pYUC4:YUC4 bre (Fig. 7G), probably due to the complex interactions between BRE and the auxin pathway. To test the interaction between BRE and the auxin pathway, the auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) was applied to WT and the bre mutant. As a result, NPA treatment caused fewer floral organs, abnormal carpels without style or consisting of a naked ovule only (Fig. 7, H and I), and a reduced leaf complexity phenotype in WT (Supplementary Fig. S10), similar to the bre mutant. More importantly, bre was hypersensitive to NPA treatment, producing a pin-like inflorescence/flower without floral organs even at the low concentration (Fig. 7H, and Supplementary Fig. S10). These results indicate that the auxin pathway is severely disrupted in bre.

Figure 7.

BRE interacts with the auxin pathway. A) IAA metabolite levels in young flowers of wild type and bre. B)  DR5ver2:GUS expression in young flowers of WT and bre after 7 h of staining. Br/Se, bract/sepal; St, stamen; Ca, carpel; Pe, petal. C)  bre mutant treated with different concentrations of IBA. Red arrows indicate flowers. D) Plant of WT, bre, pYUC4:YUC4, pYUC4:YUC4 bre. E) Expression levels of FveYUC4 in WT, bre, pYUC4:YUC4, and pYUC4:YUC4 bre examined by RT-qPCR. In (A and E), data are mean ± SD of three biological replicates. **, P < 0.01; ***, P < 0.001; ns, not significant; Student's t-test. F) Length of leaf petioles in WT, bre, pYUC4:YUC4, and pYUC4:YUC4 bre. Data are mean ± SD of 15 samples, and significant differences at the P < 0.05 level are indicated by different letters as determined by Tukey's test. G) Flower of WT, bre, pYUC4:YUC4, pYUC4:YUC4 bre. H) Effects of NPA application on WT and bre flowers at stage 8. I) Effects of NPA application on WT flowers at stage 12. Scale bars: (B and H) 100 µm, (C) 5 mm, (D) 1 cm, (G) 5 mm for WT and 1 mm for bre, (I) 100 µm for carpels and 1 mm for others.

BRE directly regulates the expression of several other auxin pathway genes

We searched for other target genes of BRE in the auxin pathway using RNA-seq and DAP-seq. Five promising candidates with differential expression and enriched binding peaks were selected, including FvePIN1, FvePIN3, FvePID2, FveGH3.1, and FveIAA14 (Fig. 8A). RT-qPCR results showed that the expression of FvePIN1 and FvePID2 was significantly decreased in the bre mutant and increased in the Comp lines compared to the WT. In contrast, the expression of FvePIN3, FveGH3.1 and FveIAA14 was significantly upregulated in the bre mutant and downregulated in the Comp lines (Fig. 8B). EMSA showed that BRE bound directly to the GCC-box-containing motifs of these genes, whereas the mutated GCC-box abolished the interactions (Fig. 8C). ChIP-qPCR was then performed for these genes using the 35S:BRE-GFP transgenic woodland strawberry lines. As a result, the peak regions of all five genes in DAP-seq were significantly enriched after immunoprecipitation against GFP antibody than input, whereas the control regions were not enriched (Fig. 8D). These results indicate that BRE directly regulates the expression of several auxin pathway genes involved in auxin conjugation, transport and signaling during flower development in woodland strawberry.

Figure 8.

Direct binding of BRE to the genic regions of 5 auxin pathway genes. A) BRE-IP peaks in the genic regions of FvePIN1, FvePIN3, FvePID2, FveGH3.1, and FveIAA14 obtained by DAP-seq. P1, P2, and P3 indicate the regions examined by ChIP-qPCR. B) Expression levels of the five genes in WT, bre, and Comp (pBRE:BRE bre, L1) examined by RT-qPCR. C) EMSA showing the interaction between BRE and the genic regions of the five genes. Blue letters in the oligo indicate the core binding motif. Probes are labeled with FAM. BRE^m contains the bre mutation. D) ChIP-qPCR assay in young flower buds of 35S:BRE-GFP in the regions indicated in (A). In (B and D), data are mean ± SD of three biological replicates. *, P < 0.05; **, P < 0.01; **, P < 0.001; ns, not significant; Student's t-test.

Discussion

A distinctive feature of strawberry flowers is the numerous individual carpels that develop independently on the receptacle, making it an excellent model for studying floral organogenesis. In this study, we isolated the woodland strawberry bre mutants with very few or no floral organs, especially the carpels, although the receptacle appears to form normally. Multiple approaches show that BRE can directly regulate the expression of several genes in the auxin pathway, including biosynthesis, conjugation, transport, and signaling. These results suggest that BRE promotes floral organ initiation in strawberry by regulating the auxin pathway at different steps. This provides further insights into understanding the conserved and diverse functions of DRN/L homologs and the mechanistic basis of floral organ initiation in strawberry (Fig. 9).

Figure 9.

A proposed model summarizing the functions of BRE in strawberry flower development. BRE expression is induced in the flower meristem. BRE directly regulates the expression of the auxin biosynthesis genes FveYUC4 and FveYUC2, the auxin conjugation gene FveGH3.1, the auxin polar transport genes FvePIN1, FvePIN3, FvePID2, and the auxin signaling gene FveIAA14, and then the auxin signaling outputs promote floral organogenesis in strawberry flowers. BRE may also regulate other genes independently of the auxin pathway. The color shading represents the expression pattern of BRE in floral organs. The BRE mutation resulted in a reduced number of floral organs with abnormal morphology.

BRE is a master regulator of floral organogenesis in strawberry

The BRE homologs, such as Arabidopsis DRN/L, tomato LFS and poplar EBB1, play diverse roles in tissue regeneration, embryo patterning, leaf development, bud dormancy, root and oil gland formation in different species (Banno et al. 2001; Chandler et al. 2007; Yordanov et al. 2014; Capua and Eshed 2017; Lee et al. 2024; Wang et al. 2024). Their roles in lateral organ initiation, including flowers, leaves, axillary buds, and roots, are relatively conserved across angiosperms (Chandler et al. 2011; Chandler and Werr 2017; Zhang et al. 2018; Dai et al. 2023; Yu et al. 2023; Lee et al. 2024; Wang et al. 2024). However, the mutants of these genes show distinct defects in different species. For example, bre leaves have fewer leaflets and deeper serrations, and most bre seedlings have only one cotyledon (Supplementary Fig. S3); in contrast, the csdrnl mutant has smooth leaf margins, but the leaf complexity is not affected (Wang et al. 2024). The Arabidopsis drn drnl double mutant and the tomato lfs mutant have naked shoot tips without leaves (Chandler et al. 2011; Capua and Eshed 2017), but the lfs flowers have essentially normal petals and the same number of stamens, a weaker floral phenotype than bre. These results indicate that BRE and its homologs make species-specific contributions to plant growth and development.

The floral defects that occur in bre are much more severe than other reported strawberry mutants, such as lam, fvemir64a fvecuc2a, fvesep3, fveyuc4, and fatm6^CR (Martin-Pizarro et al. 2019; Zheng et al. 2019; Feng et al. 2021; Pi et al. 2021; Lu et al. 2023), highlighting the significant impact of BRE on strawberry flower development. In addition to the reduction in number, the emerging floral organs in bre show severe morphological defects, such as carpels without carpel wall or style (Fig. 1, E and F), indicating a widespread role of BRE in floral organogenesis. In pBRE:BRE bre (Comp) plants, BRE is overexpressed, but floral organs are not further increased compared to WT (Fig. 2, F and G). This could be due to the need for other genes in addition to BRE, a disorganized expression pattern or a limited increase in expression levels.

BRE interacts with the auxin pathway at different steps in floral organ initiation

DRN/L interactions with the auxin pathway in floral organ development have only been well studied in Arabidopsis and rice (Chandler et al. 2011; Chandler and Werr 2017; Dai et al. 2023; Yu et al. 2023). Genetic analysis between drnl and auxin biosynthesis or polar transport mutants revealed that they have synergistic interactions in flower development (Chandler et al. 2011). DRNL indirectly regulates YUC4 expression through STY1 in Arabidopsis (Eklund et al. 2010, 2011), whereas YUC expression is not significantly different between WT and dsp in rice (Yu et al. 2023), suggesting different gene wiring downstream of DRN/L.

Several lines of evidence suggest that BRE and FveYUC4 are closely linked in flower development. First, the bre and fveyuc4 mutants have similar floral organ reduction phenotypes, and bre yuc4 has a similar number of petals and carpels to bre (Fig. 6, E and F). Second, FveYUC4 expression levels are significantly reduced in bre and greatly increased in pBRE:BRE bre (Fig. 5F). Third, a subset of common DEGs was identified in the flower buds of bre and fveyuc4 (Fig. 6, C and D). Fourth, the auxin metabolites, such as OxIAA and IAA-Asp, were similarly reduced in bre and yuc4 (Fig. 7A) (Lu et al. 2023). Lastly, four independent assays validated that BRE is able to bind the FveYUC4 promoter and exon via the GCC-box motif (Fig. 5), suggesting direct regulation of FveYUC4 by BRE. Peach PpEBB1 has been reported to bind to the GCC-box and activate the expression of PpYUC1 (Zhao et al. 2021), indicating a conserved molecular function of BRE in promoting auxin synthesis. In addition to YUC4, YUC2 is another target gene of BRE. YUC4 and YUC2 usually play redundant roles in plant development, such as regulating root architecture in maize (Zea mays) (Zheng et al. 2023). We speculate that the yuc4 yuc2 double mutant should have more severe floral defects than yuc4 and an obvious reduction in free IAA.

The bre phenotype is partially caused by impaired auxin biosynthesis, but application of auxin did not alter the defects (Fig. 7C). Similar results were obtained by exogenous auxin treatment of the tomato lfs mutant (Capua and Eshed 2017). The fveyuc4 phenotype can be rescued by expression of pFveYUC4:FveYUC4 rather than by exogenous auxin treatment (Lu et al. 2023), indicating that precise local distribution of endogenous auxin may be required. Therefore, the failure of auxin treatment may be due to a disruption in auxin transport or signaling, as suggested by the reduced DR5 activity and hypersensitivity to NPA treatment in bre (Fig. 7). Consistently, several auxin transport and signaling genes were differentially expressed in bre flower buds (Fig. 4). Our data show that BRE can directly regulate the expression of the auxin conjugation gene FveGH3.1, the auxin transport genes FvePIN1, FvePIN3, and FvePID2, and the auxin signaling gene FveIAA14 (Fig. 8). Loss of AtPIN1 function resulted in naked, pin-like inflorescences, which can be mimicked by application of the polar auxin transport inhibitor NPA (Okada et al. 1991; Benková et al. 2003). Consistent with the involvement of auxin transport genes, NPA treatment caused similar floral defects in the WT as in bre (Fig. 7, H and I). Floral organ development was more sensitive to NPA treatment in the bre mutant (Fig. 7H), suggesting a role for BRE through the auxin pathway. Other differentially expressed auxin genes may be indirectly regulated by BRE.

Other candidate genes regulated by BRE

As a master developmental regulator, BRE can affect a significant number of downstream genes. The binding sites of DRN/L and their homologs have not been identified at a genome-wide scale. DAP-seq is a high-throughput assay that does not require gene-specific antibodies or transgenic lines for this purpose (O'Malley et al. 2016; Bartlett et al. 2017). Using this method, thousands of significant BRE-IP peaks were identified in the woodland strawberry genome (Supplementary Data Set 3). The enrichment of several GCC-box-containing motifs indicates that the results are reliable. Approximately 30% of the downregulated and 35% of the upregulated genes in bre flower buds overlapped with the BRE-IP peaks (Fig. 4F), providing potential downstream regulatory targets of BRE. The validation of BRE-binding sites in the seven auxin genes demonstrates the value of the DAP-seq results. In addition, CUC1, CUC2, STM, AHP6, CKX6, JAZ5, ASA1, and LMI1 are among the DRN/L target genes in Arabidopsis (Ikeda et al. 2006; Tian et al. 2014; Zhang et al. 2018; Dai et al. 2023; Lee et al. 2024). Among their homologs, we found BRE-IP peaks in the 5'-UTR of FveSTM (FvH4_3g01400) and the exons of FveCUC2a (FvH4_3g19410) and FveCUC2b (FvH4_6g06070). However, only FveCUC2b expression is significantly downregulated in bre flower buds. Whether the homologs of these reported DRNL target genes are indeed regulated by BRE requires further investigation. Our data revealed a number of genes encoding transcription factors that could be the target genes of BRE. For example, the higher order mutants of the SHORT INTERNODES/STYLISH (SHI/STY) family members produce a malformed gynecium without stigma and style, and also produce leaves with deeper serrations (Kuusk et al. 2002, 2006). In this gene family, FveSRS5 (FvH4_2g38950) and FveLRP1 (FvH4_3g27920) showed reduced expression in bre flower buds, which may be involved in the flower or leaf defects. BRE-mediated regulation of these and auxin genes may explain why the bre mutant cannot be rescued by the introduction of pYUC4:YUC4.

Materials and methods

Plant materials and growth conditions

The woodland strawberry (Fragaria vesca) varieties Yellow Wonder 5AF7 (YW5AF7) and Hawaii 4 (H4) were used in this study. The seeds were sowed in the pots containing peat soil and vermiculite (2:1 ratio by volume). The plants were grown inside a growth room at 25 °C with a relative humidity of 55% under a 16 h light/8 h dark photoperiod and a light intensity of 100 μmol m⁻² s⁻¹.

Gene isolation of the bre mutant

The WT YW5AF7 seeds were treated with 0.4% methanesulfonate (EMS) for eight hours, and then propagated in a greenhouse. The bre mutants were screened in the M₂ generation. The bre mutants were backcrossed into the parent YW5AF7 to generate an F₂ population. In this population, an equal amount of young leaves was pooled from 26 wild-type plants and 18 mutant plants, respectively. DNA extraction for the mutant and wild-type groups was performed using a CTAB (hexadecyltrimethylammonium bromide) method (Porebski et al. 1997). Genome sequencing was performed using the Illumina HiSeq X Ten platform (Novogene, Tianjin) and analyzed as described previously (Luo et al. 2018a). After read mapping and mutation filtering, the candidate single nucleotide polymorphisms (SNPs) were further confirmed by polymerase chain reaction (PCR) amplification and Sanger sequencing in each individual mutant found in the F₂.

Scanning electron microscopy

The developing floral buds of WT and mutants were dissected under a stereomicroscope and immediately fixed in 2.5% glutaraldehyde at 4 ℃ overnight. The samples were washed with PBS (0.1 m, without NaCl) 3 to 5 times (15 min each time) and then incubated with 1% osmium tetroxide in 25 mm sodium phosphate buffer at 4 °C overnight, washed with PBS (0.1 m, without NaCl) for 3 to 5 times (15 min each time), dehydrated sequentially with 30%, 50%, 70%, 80%, 90%, 95%, and 100% ethanol for 15 min each, transferred to isoamyl acetate for 3 times (20 min each time), critical point dried, coated with gold for 70 s, and photographed under a scanning electron microscope (JSM-6390LV).

Phylogenetic analysis

The protein sequences were downloaded from the Genome Database for Rosaceae (http:/ww.rosaceae.org/), TAIR (https://www.arabidopsis.org/), and the GenBank database (https://ncbi.nlm.nih.gov/). Sequence alignment was performed by DNAMAN. The full-length protein sequences were subjected to multiple alignments by ClustalW, and the neighbor-joining tree was constructed using MEGA (v7.0) software. Bootstrapping was performed with 1,000 replicates. The alignment and Newick-format tree are provided in Supplementary Files 1 and 2.

Auxin quantification

Free auxin contents in the flower buds at stages 1 to 7 of WT and the bre mutant were measured with three biological replicates. For each replicate, approximately 50 mg of fresh tissues were harvested, immediately frozen in liquid nitrogen, ground into powder, and extracted with 1 mL methanol/water/formic acid (15:4:1, v/v/v). The combined extracts were evaporated to dryness under nitrogen gas stream, reconstituted in 100 μL of 80% methanol (v/v), and filtered through a 0.22-μm filter for LC-MS analysis (METWARE, Wuhan).

Plasmid construction

For pBRE:BRE, the genomic sequence of BRE including 1,928-bp upstream of the start codon, 1,287 bp of gene body, and 766-bp downstream of the stop codon was amplified, sub-cloned into the binary vector pCambia1300 at the EcoRI and BamHI sites. For pBRE:GUS, the 1,994-bp promoter of BRE was inserted into the binary vector pMDC162 before the GUS reporter by gateway cloning. For 35S:BRE-GFP, the full-length coding sequence of BRE without the stop codon was inserted into the vector pMDC32 and fused with GFP.

Stable transformation in woodland strawberry

The constructs were transformed into the Agrobacterium tumefaciens strain GV3101. The Agrobacterium-mediated strawberry transformation was performed as described previously (Feng et al. 2019). In this process, positive transgenic calli and regenerated plants were selected using both hygromycin (4 mg L⁻¹) and GFP fluorescence examined under a fluorescence dissecting microscope (OLYMPUS SZX7, Japan).

Subcellular localization analysis

The coding sequence of BRE was inserted into the PRI101 vector digested with SalI and EcoRI and fused in frame with GFP under the control of the CaMV35S promoter. The Agrobacterium (GV3101) colonies containing 35S:BRE-GFP or the nuclear marker VirD2NLS-mCherry were grown in 10 ml of liquid LB medium at 28 °C overnight. The culture was then spun down and resuspended in the buffer (10 mm MgCl₂, 100 μM acetosyringone, and 10 mm MES) to reach an OD₆₀₀ of 0.1, and then injected into Nicotiana benthamiana leaves. The fluorescence was observed under a confocal laser scanning microscope (Leica SP8SP8, Germany). The excitation/emission wavelength is 488/507 nm for GFP and 587/610 nm for mCherry.

Transcriptome sequencing and data analysis

Total RNAs were isolated from the young flower buds of WT and bre at stages 1 to 7 using a HiPure Plant RNA Mini Kit (Magen, China; cat no. R4151). Each sample had 3 biological replicates. Sequencing libraries were generated using NEBNext UltraTM RNA Library Prep Kit for Illumina (NEB, USA) following the manufacturer's recommendations. Sequencing was performed on an Illumina HiSeq X Ten platform (Novogene, Tianjin, China). Raw reads were firstly trimmed 9 bp at the 5’ end using Trimmomatic (Bolger et al. 2014) and then mapped to the F. vesca reference genome with the ver4.0.a2 annotation using the program hisat2 (Kim et al. 2015). FeatureCounts was used to count the reads mapped to the gene models (Liao et al. 2014). Gene expression levels were calculated as the TPM (Transcripts Per Million reads) with Tbtools (Chen et al. 2020). Differential gene expression analysis was implemented by the R package DESeq2 (adjusted P-value of <0.05, fold change >2) (Love et al. 2014). Venn diagram was made at SRplot (http://www.bioinformatics.com.cn/srplot). The Gene Ontology (GO) terms were annotated for F. vesca proteins by blasting the database eggNOG v5.0 using eggNOG-mapper v2 (Huerta-Cepas et al. 2019; Cantalapiedra et al. 2021). GO enrichment analysis was conducted using the R package clusterProfiler (v4.2.0) (Wu et al. 2021).

RT-qPCR

Total RNAs were extracted from the young flower buds of different plant materials at stages 1 to 7 using a HiPure Plant RNA Mini Kit (Magen, China; cat no. R4151) and treated with DNase I (Promega) according to the manufacturer's instructions. Approximately 1 μg of total RNA was reverse-transcribed into cDNA using a PrimeScript RT reagent kit (TaKaRa, Shiga, Japan; cat no. RR047A). Quantitative real-time PCR was performed as previously described (Pi et al. 2021).

DNA affinity purification sequencing (DAP-seq)

Approximately 10 μg of Genomic DNA was extracted from the young flower buds of YW5AF7 at stages 1 to 7. BRE was fused to an N-terminal HaloTag. DAP-seq assay was performed for BRE by Gene Denovo Biotechnology Co., (Guangzhou, China). The BRE-IP and BRE-input gDNA libraries were sequenced using Illumina NovaSeq 6000. The clean reads were aligned to the F. vesca reference genome with the ver4.0.a2 annotation using Bowtie 2 (version 2.2.5) (Langmead and Salzberg 2012), and counted by deepTools (version 3.2.0) (Ramirez et al. 2016). Enriched peaks were identified by using MACS2 (version: 2.1.2) (Zhang et al. 2008). Peak-related genes were annotated by the R package ChIPseeker (Yu et al. 2015). MEME suite was used to detect the motifs.

GUS staining

The flower buds were dissected from the reporter lines under a stereomicroscope and directly submerged in the staining buffer (0.2% Triton X-100, 50 mm NaPO₄, 10 mm Potassium ferri/ferrocyanide, 2 mm X-Gluc, pH 7.0). After 15 min of vacuum infiltration, tissues were incubated at 37 °C until staining developed. For better observation, the staining buffer was exchanged with the clear solution (chloral hydrate: glycerol: H₂O = 80%: 10%: 10%) for several hours. Finally, the samples were observed under differential interference contrast optics using a Zeiss Axioscope A1 microscope with a 0.5× optical adapter.

RNA in situ hybridization

The young flower buds and leaves of WT woodland strawberry were fixed at 4 ℃ overnight. The digoxigenin-labeled RNA probes were generated by the DIG RNA labeling kit (Roche, Switzerland) according to the manual. Fixation and embedding, pre-hybridization, hybridization, and detection were carried out as previously described (Zhang et al. 2020). Images were obtained with a Zeiss Axioscope A1 microscope.

Dual luciferase reporter assay

The full-length coding sequence of BRE and BRE^m were cloned into the vector PRI101 under the control of the 35S promoter to generate effector constructs. The FveYUC4p (−4929∼−4 bp), FveYUC4e (285∼916 bp), and FveYUC2e (−1598∼592 bp) fragments were inserted into pGreenII0800 to drive expression of the luciferase (LUC) reporter gene. All recombinant constructs were transformed into the Agrobacterium tumefaciens strain GV3101 (pSoup) (WEIDI, Shanghai, China). The cultures were spun down and resuspended in the infiltration buffer (10 mm MgCl₂, 100 μM acetosyringone, 10 mm MES). The effector, reporter, and P19 cell suspensions were mixed to reach a final OD₆₀₀ of 0.1, 0.4, and 0.1, respectively, and infiltrated into the N. benthamiana leaves. After 2 to 3 days, 1 mm D-luciferin (Solarbio, Beijing, China) was applied to the back of leaves, and the fluorescence was detected with an exposure time of 10 s (NightSHADE L985). Moreover, the luciferase activity was determined using a Luciferase Assay System kit (Promega, USA). The primers are listed in Supplementary Data Set 4.

Yeast one-hybrid assay

The coding sequence of BRE was cloned into the prey vector pB42AD. The pFveYUC4 (−4106∼−1554 bp), FveYUC4e (13∼1567 bp), and FveYUC2e (−1598∼592 bp) fragments were ligated into the pLacZi vector. The primers are listed in Supplementary Data Set 4. These constructs or the corresponding empty vectors were co-transformed into the yeast strain EGY48. The transformants were grown on SD/-Trp/-Ura for 3 days and then transferred onto the plates containing 0.08 mg l⁻¹ X-gal for blue-white screening.

Electrophoretic mobility shift assay

The coding sequences of BRE and BRE^m (including the bre mutation) were cloned into the vector pMAL-c2X in frame with MBP, respectively. The constructs were then introduced into Rosetta (DE3) competent cells (WEDI, Shanghai, China). The MBP or MBP-fused proteins were induced using 0.1 mm isopropylβ-D-1-thiogalactopyranoside (IPTG) at 16 °C for 20 h, and affinity-purified using Dextrin beads 6FF (Smart-Lifesciences, China). Wild-type or mutated DNA probes harboring cis-elements were commercially synthesized (Sangon Biotech, China) as single-stranded (ss) DNA and labeled with biotin at their 5′ ends. To make double-stranded (ds) oligos, equal amounts of complementary ss oligos were mixed, treated at 95 ℃ for 2 min, and slowly cooled down to 25 ℃ by 0.1 ℃/s. Competitive probes were made similarly without labeling. EMSA was performed using a chemiluminescent EMSA kit (Thermo Fisher Scientific, USA).

ChIP-qPCR

The 35S:BRE-GFP transgenic lines were used for chromatin extraction. Briefly, 2 g of the samples were fixed in 1% (v/v) formaldehyde with vacuum infiltration for 20 min at room temperature, and neutralized with 0.125 m glycine for 5 min. The samples were ground to a fine powder in liquid nitrogen. The chromatin was extracted with extraction buffer1 (0.4 m sucrose, 10 mm Tris–HCl, 10 mm MgCl₂, 5 mm β-ME, 0.1 mm phenylmethanesulfonyl fluoride (PMSF), PI), extraction buffer2 (0.25 m sucrose, 10 mm Tris–HCl, 10 mm MgCl₂, 1% Triton X-100, 5 mm β-ME, 0.1 mm PMSF, and PI), extraction buffer3 (1.7 m sucrose, 10 mm Tris–HCl, 2 mm MgCl₂, 0.15% Triton X-100, 5 mm β-ME, 0.1 mm PMSF, and PI), and nuclei lysis buffer (50 mm Tris–HCl, 10 mm EDTA, 1% SDS, and PI) sequentially. The chromatin was sonicated to approximately 200 bp to 1-kb fragments. Remove 10 μl from each sample into a clean tube and set aside at −20 °C to serve as the “input DNA control”. Chromatin was cleared with GFP-trap (ChromoTek; gtma) magnetic beads and incubated overnight. Wash the immune complexes with low-salt wash buffer (150 mm NaCl, 0.1% SDS, 1% Triton X-100, 2 mm EDTA, 20 mm Tris–HCl), high-salt wash buffer (500 mm NaCl, 0.1% SDS,1% Triton X-100, 2 mm EDTA, 20 mm Tris–HCl), LiCl wash buffer (0.25 m LiCl, 1% NP-40, 1% DOC, 1 mm EDTA, 10 mm Tris–HCl), and then TE buffer (10 mm Tris–HCl and 1 mm EDTA). Add elution buffer to each sample and incubate at 65 °C for 15 min to elute the immune complexes from the washed beads. The cross-linking of each IP sample and “input DNA control” were reversed by incubation at 65 °C for at least 6 h. Digest with 1 µL RNase A (10 mg/mL) and 2 µL proteinase K (10 mg/mL). DNA was purified with the Qiagen MinElute Kit. After DNA purification, the BRE-binding sites on target genes were evaluated using qPCR and normalized by total chromatin (input). The fold enrichment is calculated as the ratio between the samples of input and IP samples. The FvH4_2g03140 promoter was used as a reference. The primers are listed in Supplementary Data Set 4.

Hormone treatment

For IBA and NAA treatment, the bre mutant plants were grown in the 1/2 mS medium with or without different concentrations of NAA (0, 0.125, 0.25 or 1 μM; CAS#86-87-3, PhytoTech) or IBA (0, 15, or 30 μM; CAS#133-32-4, PhytoTech) in jam bottles for 1.5 months. For NPA treatment, 1 ml solution containing 0, 250, or 1000 μM NPA (CAS#132-66-1, PhytoTech) with 0.01% Silwet L-77 was applied to the shoot center of each plant every 2 days for 1.5 months.

Statistical analyses

Statistical analyses were performed using GraphPad Prism 10. Pairwise comparisons were determined using Student's t-test (*, P < 0.05; **, P < 0.01; ***, P < 0.001). The comparison between multiple samples was determined using Tukey's test, and significant differences at the P < 0.05 level are indicated by different letters. Data for all statistical analyses are shown in Supplementary Data Set 5.

Accession numbers

The sequencing datasets have been submitted to the Sequence Read Archive at NCBI (http://www.ncbi.nlm.nih.gov/sra) under the submission codes of PRJNA858434, PRJNA858241, and PRJNA858408. The strawberry gene sequences from the F. vesca genome ver4.0 can be downloaded from the Genome Database for Rosaceae (http:/ww.rosaceae.org) with accession numbers: BRE, FvH4_4g25630; FveYUC4, FvH4_2g29930; FveYUC2, FvH4_1g04800; FveGH3.1 FvH4_2g04750; FvePIN1, FvH4_5g17310; FvePIN3, FvH4_4g32450; FvePID2, FvH4_1g03770; FveIAA14, FvH4_4g02280.

Author contributions

C.K. and R.L. conceived and designed the experiments; R.L. performed most of the experiments; S.H. analyzed the sequencing data; J.F. helped with the experiments; C.K. and R.L. wrote the manuscript; Z.L. helped with data interpretation and revised the manuscript. All the authors have read and approved the paper.

Supplementary data

The following materials are available in the online version of this article.

Supplementary Figure S1 . Different bre flowers in woodland strawberry.

Supplementary Figure S2 . Flower meristem of WT and bre.

Supplementary Figure S3 . Leaf and cotyledon phenotypes of bre in woodland strawberry.

Supplementary Figure S4 . Leaf and flower phenotypes of bre-2 in woodland strawberry.

Supplementary Figure S5 . Protein sequence alignment of BRE and its homologs.

Supplementary Figure S6 . Leaf phenotypes of the pBRE:BRE bre transgenic lines.

Supplementary Figure S7 . The p35S:BRE-GFP transgenic woodland strawberry lines used for ChIP-qPCR.

Supplementary Figure S8 . DR5ver2:GUS expression in the developing flower primordia of WT and bre.

Supplementary Figure S9 . Treatment of the bre mutant with different concentrations of NAA.

Supplementary Figure S10 . Effects of NPA application on WT and bre plants.

Supplementary Table S1 . Summary of genomic DNA sequencing read statistics.

Supplementary Table S2 . Candidate SNPs in the EMS mutant bre.

Supplementary Table S3 . Summary of RNA-seq read statistics.

Supplementary Data Set 1 . Differentially expressed genes in the pairwise comparison between bre and WT in flowers at stages 1 to 7.

Supplementary Data Set 2 . Enriched GO terms among the DEGs in the flower buds of bre.

Supplementary Data Set 3 . Enriched BRE-IP peaks obtained by DAP-seq.

Supplementary Data Set 4 . Primers used in this study.

Supplementary Data Set 5 . Statistical analyses.

Supplementary File 1 . Alignment for phylogenetic analysis

Supplementary File 2 . Newick-format tree file for phylogenetic analysis.

Funding

This work was supported by the National Key Research and Development Program of China (2023YFF1001100), the National Natural Science Foundation of China (31822044, 32172539 and 32372669), and the Fundamental Research Funds for the Central Universities (2662023PY011).

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• DRNL Gramene: AT1G24590

• DRNL Araport: AT1G24590

• SDS Gramene: AT1G14750

• SDS Araport: AT1G14750

• YUC2 Gramene: AT4G13260

• YUC2 Araport: AT4G13260

• ahp6 Gramene: AT1G80100

• ahp6 Araport: AT1G80100

• ASA1 Gramene: AT5G05730

• ASA1 Araport: AT5G05730

• CKX6 Gramene: AT3G63440

• CKX6 Araport: AT3G63440

• CUC1 Gramene: AT3G15170

• CUC1 Araport: AT3G15170

• CUC2 Gramene: At5g53950

• CUC2 Araport: At5g53950

• STM Gramene: AT1G62360

• STM Araport: AT1G62360

• STY1 Gramene: AT3G51060

• STY1 Araport: AT3G51060

• Esr1 Gramene: Zm00001d027820

• Esr1 Araport: Zm00001d027820

• YUC4 Gramene: AT5G11320

• YUC4 Araport: AT5G11320

• JAZ5 Gramene: AT1G17380

• JAZ5 Araport: AT1G17380

• EMSA Gramene: electrophoretic mobility shift assay

• EMSA Araport: electrophoretic mobility shift assay

• auxin CHEBI: CHEBI:22676