Auxin-induced AUXIN RESPONSE FACTOR4 activates APETALA1 and FRUITFULL to promote flowering in woodland strawberry

Flowering time is known to be regulated by numerous pathways, such as the autonomous, gibberellin, aging, photoperiod-mediated, and vernalization pathways. These regulatory mechanisms involve both environmental triggers and endogenous hormonal cues. Additional flowering control mechanisms mediated by other phytohormones, such as auxin, are less well understood. We found that in cultivated strawberry (Fragaria × ananassa), the expression of auxin response factor4 (FaARF4) was higher in the flowering stage than in the vegetative stage. Overexpression of FaARF4 in Arabidopsis thaliana and woodland strawberry (Fragaria vesca) resulted in transgenic plants flowering earlier than control plants. In addition, FveARF4-silenced strawberry plants showed delayed flowering compared to control plants, indicating that FaARF4 and FveARF4 function similarly in regulating flowering. Further studies showed that ARF4 can bind to the promoters of the floral meristem identity genes APETALA1 (AP1) and FRUITFULL (FUL), inducing their expression and, consequently, flowering in woodland strawberry. Our studies reveal an auxin-mediated flowering pathway in strawberry involving the induction of ARF4 expression.


Introduction
Flowering marks a transition from vegetative to reproductive growth in plants and involves numerous physiological processes, metabolic pathways, and gene regulatory mechanisms 1,2 . These mechanisms involve intracellular and intercellular signal transduction cascades and the specific spatiotemporal expression of flowering genes [3][4][5] . To elucidate the molecular underpinnings of flowering, many studies have been performed on the model species Arabidopsis thaliana, resulting in the identification of genes involved in flowering regulation, as well as multiple signaling pathways: photoperiodic, vernalization, ambient temperature, autonomous, aging, and gibberellin (GA) [6][7][8] . However, it is recognized that given the interactions among phytohormones in regulating plant development, other signaling systems may also be involved 9 . For example, auxin is known to control many processes of plant growth and development and broadly regulates gene expression; however, its role in flowering and the associated molecular mechanisms remain poorly understood 10,11 .
Central to auxin-regulated transcription are three families of primary auxin-responsive genes: Aux/IAA (AUXIN/INDOLE ACETIC ACID), GH3 (GRETCHEN HAGEN3), and SAUR (SMALL AUXIN UP RNA) 12 . All of the gene promoters from these families contain the TGTCNC motif (auxin response element, AuxRE), which is bound to AUXIN RESPONSE FACTOR (ARF) proteins to mediate auxin responses 13 . In A. thaliana, ARF functions have been well studied, and it is known that AtARF1, AtARF2, AtARF6, and AtARF8 are involved in floral organ development 14 , while AtARF7, AtARF16, and AtARF19 are associated with root development 13,15 , AtARF12-15 regulates embryogenesis and seed development, and AtARF20-22 has similar functions to AtARF12-15 13,16 .
AtARF3 and AtARF4 are involved in the transition from vegetative to reproductive growth, and a unique feature of ARF3/4 regulation is that their transcripts are posttranscriptionally cleaved by trans-acting short-interfering RNAs, such as tasiRNA3, an endogenous trans-acting small-interfering RNA 13 . The production of tasiRNA3 is initiated by cleavage of the non-protein-coding TAS3 RNA by miR390. A complex of the miR390-cleaved transcript bound to ARGONAUTE7 (AGO7) is then used as a template for polymerization by RNA-DEPENDENT RNA POLYMERASE6 (RDR6) and SUPPRESSOR OF GENE SILENCING3 (SGS3) 10 . The resulting doublestranded RNA is cleaved by DICER-LIKE4 (DCL4) to generate the 21-nucleotide-long tasiRNA3 17 . Thus, AGO7/RDR6/SGS3/DCL4 is required for the production and/or stability of tasiRNA3, which targets both ARF3 and ARF4. Mutations in AGO7/RDR6/SGS3/DCL4 in A. thaliana accelerate the juvenile-to-adult transition, and the expression of ARF3 and ARF4 is upregulated in the ago7/rdr6/sgs3/dcl4 mutant 17 . This indirectly demonstrates that the ago7/rdr6/sgs3/dcl4 phenotype is attributable to the absence of tasiRNA3-mediated repression of ARF3 and ARF4 10 . However, the specific function of AtARF3 and AtARF4 in the phase transition is unclear.
A potentially useful experimental system to study flowering transition is strawberry (Fragaria) 18 . Early flowering is a great advantage for fruit crop breeding and cultivation since it shortens the process of generating new cultivars and reduces the time to produce a marketable crop, resulting in economic benefits 7,19 . Octoploid cultivated strawberry (Fragaria × ananassa) is widely cultivated in all arable regions around the globe and has high economic value 20 . However, although F. × ananassa is of great agricultural and research importance, genetic transformation and associated functional studies are hindered by its complex genetics, which include a large genome and high degree of heterozygosity 21 . In contrast, diploid woodland strawberry (Fragaria vesca) is suitable for regeneration, and a transformation system has been established in vitro 22 . Moreover, F. vesca has a substantial seed set, rapid growth, a small size, a relatively small genome, and existing genetic maps, which makes it a good option as a model plant for the development of genomic tools, enabling us to perform functional studies using forward or reverse genetic approaches 23 . Notably, F. vesca shares considerable sequence similarity with F. ×ananassa and other rosaceous plants, and the high levels of functional conservation at the genetic level mean that knowledge of diploid strawberry gene function has the potential to be applied to octoploid strawberry or other members of the Rosaceae family 7,23 .
In this study, we investigated the role and mechanism of auxin-regulated flowering in strawberry and found that the expression level of auxin-induced FaARF4 was upregulated in the flowering stage rather than during vegetative growth, indicating that FaARF4 is related to flowering. Using transgenic plants, we discovered that FaARF4 promoted flowering and that FveARF4-silenced plants showed delayed flowering. Further studies revealed that FaARF4 and FveARF4 bound to the promoters of FveAP1 and FveFUL to induce their expression. Our findings showed that auxin is involved in the flowering pathway in strawberry through its regulation of ARF4 expression.

FaARF4 expression analysis
Elite strawberry cultivars, free of plant pathogenic fungi and bacteria, are commonly produced by micropropagation, although several problems, such as epigenetic variation, can be encountered using this approach 24 . In our previous study, we reported that the flowering characteristics of micropropagated strawberry (F. × ananassa) changed so that it exhibited early flowering, along with a low level of miR390 expression 25 . Next, we overexpressed miR390 in tobacco and found that the juvenile-to-adult phase transition was significantly delayed 26 . Since miR390 negatively regulates the expression of ARF3/4, we speculated that ARF3/4 may be related to flowering in cultivated strawberry. To study the expression patterns of FaARF3 and FaARF4 in response to flowering, we determined their expression profiles in stem tips at different developmental stages in the cultivated strawberry 'Yanli' by quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis. As shown in Fig. 1a, the expression levels of FaARF3 and FaARF4 were higher in the flowering stage than in the vegetative stage, indicating that these two genes are involved in flowering. Since the expression level of FaARF4 changed more than that of FaARF3, FaARF4 was chosen for further study.

Cloning and phylogenetic analysis of FaARF4
Primers for FaARF4 from the cultivated strawberry 'Yanli' were designed based on the FveARF4 sequence in the NCBI database (https://www.ncbi.nlm.nih.gov/). The full-length coding sequence (CDS) of FaARF4 (GenBank No. MG765454) was found to be 2403 bp long and encode an 800-amino-acid protein with many α-helices, β-sheets, and hydrophilic and flexible regions (Fig. S1). An alignment of the amino acid sequences of FaARF4, FveARF4, and AtARF4 revealed 85.52% sequence identity (Fig. 1b). The conserved regions were primarily confined to three domains, the DNA binding domain (DBD), the middle region (MR), and the protein binding (PB1) domain, consistent with FaARF4 being a member of the ARF family. As shown in Fig. 1c, FaARF4 contains a DBD, an MR, a PB1, and two tasiRNA3 target sites.
To study the phylogenetic relationships of ARF4 with different species, the ARF4 amino acid sequences from members of the Rosaceae and other plant species were retrieved from the NCBI database. A phylogenetic analysis ( Fig. 1d) showed that the FaARF4 and FveARF4 sequences were the most closely related, followed by other Rosaceae proteins and then proteins from other species, consistent with their phylogenetic placement. FaARF4 expression profile in strawberry qRT-PCR analysis of FaARF4 in different strawberry tissues revealed that this gene was universally expressed; the highest expression level was found in sepals, where the expression level was approximately 7-fold higher than that in roots, which had the lowest expression level (Fig. 1e). Next, we measured FaARF4 expression during fruit development and ripening. Strawberry fruit maturity can be divided into five stages: small green (SG), large green (LG), white (W), turning (T), and red (R) 27 . As shown in Fig. 1f, FaARF4 was strongly expressed in the SG stage, followed by a gradual decline until the W stage, with no change in the T and R stages. FaARF4 expression in the SG stage was~3-fold higher than that in the W stage.
To test the hypothesis that auxin controls FaARF4 expression, we analyzed its expression in 'Yanli' fruit after treatment with water (control), 500 μM NAA (α-naphthalene acetic acid, a synthetic auxin), or 500 μM NPA (N-1-naphthylphthalamic acid, an auxin inhibitor) by qRT-PCR. As shown in Fig. 1g, the expression of FaARF4 was upregulated after NAA treatment and downregulated after NPA treatment, suggesting that auxin induces the expression of FaARF4.

Identification of the transcription factor FaARF4
FaARF4 sequence analysis predicted that it localizes to the nucleus (http://www.csbio.sjtu.edu.cn/bioinf/Cell-PLoc/), and this hypothesis was tested by expressing a green fluorescent protein (GFP):ARF4 fusion (FaARF4-GFP) driven by the constitutive CaMV 35S promoter ( Fig. 2a) in A. thaliana protoplasts and Nicotiana benthamiana leaves. A constitutively expressed GFP control was similarly tested. The green fluorescent signal of FaARF4-GFP was only detected in the nucleus, while GFP alone was found throughout the entire cell (Figs. 2b and S2), suggesting that FaARF4 specifically accumulates in the nucleus.
To investigate whether FaARF4 has transcriptional activation activity, we fused its CDS with the GAL4 DBD and created the pGBT9-FaARF4 construct (BD-FaARF4). BD-FaARF4 and the empty vector pGBT9 (BD), as a negative control, were then transformed into yeast for transcriptional activity analysis. As shown in Fig. 2c, the yeast strain containing BD-FaARF4 grew well on SD/-Trp-His-Ade media, while the yeast strain containing BD grew only in SD/-Trp, demonstrating that FaARF4 is a transcriptional activator.
Mutation of FaARF4 tasiRNA3 target sites tasiRNAs are a group of endogenous non-coding small RNAs that function on their target genes by complementary base pairing. It is well known that ARF4, the target gene of tasiRNA3, is negatively regulated by tasiRNA3 in A. thaliana 10,13 . To verify whether tasiRNA3 cleaves FaARF4 in strawberry, an RLM-5ʹ rapid RACE assay was performed. As shown in Fig. 3a, FaARF4 contains two tasiRNA3 target sites (site 1 and site 2) located in the coding region that were cleaved by tasiRNA3. To prevent FaARF4 cleavage by endogenous tasiRNA3, we performed rapid PCR site-directed mutagenesis of FaARF4 using the methods of Picard et al. (1994) 28 . We obtained the nucleotide sequence of a tasiRNA3-insensitive mutant (named FaARF4mut) and confirmed that the amino acid sequence was the same as that of FaARF4 and  that only the nucleotide sequence was affected (Fig. 3b). FaARF4mut was used for further functional studies.
FaARF4 promotes flowering in A. thaliana and woodland strawberry A. thaliana is often used to assess the function of genes from other plant species 20 , and here we transformed both A. thaliana (Columbia) and 'Ruegen' (F. vesca) with the pRI 101-FaARF4mut plasmid (Fig. 4a). All transgenic lines were identified by detection of the FaARF4 gene using the 35S and ARF4-R primers. A 2000-3000 bp band was amplified from the genomic DNA of all transgenic lines, and no corresponding bands were amplified from control plants (Fig. S3a). We obtained four transgenic A. thaliana lines and four transgenic strawberry lines. FaARF4 expression was investigated in the transgenic lines and control plants, and it was clearly higher in the transgenic plants (Fig. S3b, c).
The phenotypic differences between the transgenic lines and wild-type A. thaliana plants are shown in Fig. S3d. We found that transgenic plants flowered earlier than the wildtype and that the vegetative growth stage was approximately 52 days for the wild-type and 44 days for the transgenic plants (Fig. S3e). These results indicated that FaARF4 can regulate the flowering process in A. thaliana.
To better observe the phenotype of the transgenic strawberry plants, we transplanted them into a greenhouse. We observed an early flowering phenotype in the transgenic lines (Fig. 4b) and found that the transgenic plants flowered 15 days earlier than the control plants (Fig. 4c). Thus, FaARF4 also regulates flowering in strawberry.
To confirm the function of FveARF4, an RNAi vector was created and transformed into 'Ruegen' using Agrobacterium-mediated transformation (Fig. 4d). Transgenic lines were confirmed by PCR amplification of a 250-500 bp band that was not amplified from control plants (Fig. S3f). We also analyzed FveARF4 expression by qRT-PCR in the RNAi lines. Compared with the control plants, the expression of FveARF4 in the RNAi lines was lower (Fig.  S3g). We observed that the flowering time of the RNAi lines was delayed by 8 days compared to that of the control plants, indicating that FveARF4 regulates the flowering process in strawberry (Fig. 4e, f). To explore the cause of early flowering in strawberry, we measured the expression of these flowering-related genes. qRT-PCR analysis showed that the expression levels of FveAP1 and FveFUL were higher in FaARF4-overexpressing plants than in control plants (which was set to 1) (Fig. 4g), while the expression of FveFT, FveSOC1, FveLFY, and FveTFL1 was not significantly different between transgenic plants and control plants (Fig. S4). FveAP1 and FveFUL expression was lower in the RNAi lines than in control plants (Fig. 4h).
To verify whether FaARF4 binds to the FveAP1 and FveFUL promoters, yeast one-hybrid (Y1H) analysis and luciferase reporter assay were carried out. The Y1H assay indicated that FaARF4 bound to the AuxRE and HUD boxes in the FveAP1 and FveFUL promoters in vitro (Fig.  5a, b). The FveAP1 and FveFUL promoters were then individually inserted into the pRI-mini35S-LUC vector (luciferase reporter vector) as reporters, and FaARF4 driven by the 35S promoter was used as the effector (Fig.  5c, d) in a transient tobacco leaf expression assay. We observed that FaARF4 also bound to the FveAP1 and FveFUL promoters in vivo (Fig. 5c, d). The Y1H and luciferase reporter assays suggested that FaARF4 directly binds to the FveAP1 and FveFUL promoters, and activates the transcription of the corresponding genes. Since the expression of FveAP1 and FveFUL was reduced in the RNAi lines, we speculated that FveARF4 and FaARF4 had the same mechanism. Supporting this hypothesis were the results of the transactivation activity assay demonstrating that FveARF4 is also a transcriptional activator (Fig. S5) and those of the luciferase reporter assay indicating that FveARF4 can bind to the promoters of both FveAP1 and FveFUL (Fig. 5e, f).

Discussion
ARF4 is involved in the IAA-mediated flowering pathway in strawberry In species such as sweet cherry (Prunus avium) 30 and longan (Dimocarpus longan) 7 , shortening the life cycle can allow multiple generations to be grown commercially in a single season, and the breeding process can be accelerated. It is also possible to increase the yield of plants by prolonging the number of days of vegetative growth in species such as sugar beet (Beta vulgaris) 4 and perennial sugarcane (Saccharum officinarum) 3 . Moreover, flowering genes can also be used to adjust the flowering period, which greatly improves the economic value of ornamental plants 31 . In cultivated strawberry, the regulation of flowering is used to provide fresh berries throughout the year 32 . Most strawberry cultivars are June-bearing, and their flower formation is affected by environmental conditions, especially light and temperature 18 . Growth studies have revealed that short days and/ or cool temperatures can promote flowering, and short days greatly promote and enhance flower initiation 19,33 . At 15°C or during crown-cooling treatments, flowering time was hastened, and the number of flowers increased 34 .
It was also reported that light quality affected flowering time and that strawberry could be induced to flower when exposed to far red light 33 . The role of GA in strawberry flowering has also been studied, and it has been shown to inhibit flowering and promote the formation of runners 32 . Additionally, it was reported that both auxin IAA and cytokinin affect the type of first bud initiation, and FaARF4 binds to FveAP1 promoter fragments (pFveAP1, pFveAP1-1, pFveAP1-2, and pFveAP1-3) (a) and FveFUL promoter fragments (pFveFUL, pFveFUL-1, pFveFUL-2, and pFveFUL-3) (b) in vitro. The empty pGAD424 and pAbAi vectors were used as negative controls. In vivo luciferase reporter assay showing that FaARF4 and FveARF4 positively regulate FveAP1 expression (c, e) and FveFUL expression (d, f), respectively. The FaARF4/FveARF4 effector and either the pFveAP1 or the pFveFUL reporter were coinfiltrated into tobacco leaves, and the luciferase signal was measured exogenously applied IBA (indole-3-butytric acid, an IAA analog) +6-BA (6-benzylaminopurine, a synthetic cytokinin) leads to the production of inflorescences instead of runners 35 . However, it has not been reported how IAA regulates flowering at the molecular level.
IAA regulates plant growth and development by employing signal transduction, and as the center of the auxin signaling pathway, the expression of ARF proteins is known to be regulated by IAA 13 . The expression of ARF4 is also affected by IAA, and it has been shown to be upregulated by IAA treatment in model plants, such as A. thaliana 36 , Medicago truncatula 37 , and Brachypodium distachyon 38 . In a previous study, we also demonstrated that the expression of FveARF4 was induced by IAA in woodland strawberry 39 , and here, we found that the expression of FaARF4 was increased by IAA and reduced by NPA (Fig. 1g). We analyzed the function of both FaARF4 and FveARF4 and found them to be involved in the regulation of flowering time, which helps to elucidate the IAA-mediated flowering pathway in strawberry ( Fig. 4a-f).  13 . It has been suggested that AtARF3 and AtARF4 indirectly activate AtLFY, AtAP1, and AtFUL by enhancing the expression of AtSPL3, the target gene of miR156, which promotes flowering in A. thaliana, but this has not been experimentally confirmed 6 . However, we observed no FveSPL3 transcriptional activation activity in yeast (Li, 2020;unpublished data). To explore how ARF4 regulates flowering in woodland strawberry, we analyzed the expression of flowering-related genes in transgenic and control strawberry and observed significant differences in the expression levels of FveFUL and FveAP1 (Fig.  4g, h). We also showed that FaARF4 and FveARF4 have transcriptional activation activity (Figs. 2b and S5), and Y1H and luciferase reporter assays further confirmed that ARF4 can bind to the AuxRE and HUD-box motifs in the FveFUL and FveAP1 promoters and activate their expression (Fig. 5). FUL and AP1 belong to the AP1/ SQUA gene subfamily and are key genes in flowering regulation, and it has been reported that they promote flowering in various species, such as A. thaliana 7 , soybean (Glycine max) 40 , tomato (Solanum lycopersicum) 41 , maize (Zea mays) 42 , apple (Malus domestica) 43 , and others. Notably, the expression of FUL and AP1 is important in controlling flowering in woodland strawberry: high expression of FUL and AP1 promotes flowering, and if they are not highly expressed, vegetative growth continues 18 . In addition, we found that flowering time correlated with the expression levels of FveFUL and FveAP1 in transgenic strawberry lines (Fig. 4g, h). Taken together, we propose an IAA-mediated flowering pathway in strawberry (Fig. 6), which suggests a target for altering flowering time for horticultural purposes.
Flowering is an important factor in the production of fruit trees, and the existence of a long juvenile phase in fruit trees has been suggested to limit breeding and genetic improvement 32 . Furthermore, most economically important phenotypes related to fruits of hybrid cultivars cannot be identified in this phase, and one of the major goals of fruit breeding programs is to reduce the juvenile phase and accelerate floral production 18 . Genetic engineering has been used to successfully shorten the juvenile phase and promote flowering in several fruit species 7 . In our study, we found that the auxin-induced FaARF4 and FveARF4 genes were involved in the regulation of flowering. ARF4 has a sequence similar to those of homologs from fruit tree species, especially members of the Rosaceae, indicating analogous functions. Therefore, it may be possible to break the long juvenile phase of fruit trees using overexpression of the ARF4 gene.

Plant materials and treatments
The cultivated strawberry (F. × ananassa) cultivar 'Yanli' was grown under greenhouse conditions at Shenyang Agricultural University, China. Stem tips from the 4-leaf and bud stages were collected as vegetative stage and flowering stage samples, respectively. Different organs Fig. 6 Model of how auxin induces ARF4 to promote the flowering pathway. In woodland strawberry, auxin induces the expression of ARF4, and the ARF4 protein binds to and activates the AP1 and FUL promoters. The expression of AP1 and FUL initiates flowering. Symbols/abbreviations: '+', promotion; AuxRE and HUD box, auxin-responsive element (ARF4 binding sites) and fruits at different developmental stages were harvested individually for gene expression analysis, frozen in liquid nitrogen, and stored at -80°C. In the white ripening stage, the fruits were picked and transported to the laboratory, where they were divided into three groups: fruits from one group were submerged in water (as the control); fruits from the second group were submerged in 500 μM NAA (Solarbio, Beijing, China); and fruits from the third group were submerged in 500 μM NPA (a polar auxin transport inhibitor; Shanghai Yuanye Bio-Technology Company Limited, Shanghai, China). All treatments lasted for 30 s. After 1 h, total RNA samples were extracted using the cetyltrimethyl ammonium bromide (CTAB)-based method for the analysis of FaARF4 expression 44 . Three biological replicates were prepared.
The strawberry leaf transformation experiment was performed using 'Ruegen' (F. vesca), which was grown in tissue culture. Agrobacterium-mediated leaf transformation was carried out after plants had grown 4-5 leaves 45 .
Seeds of wild-type Arabidopsis thaliana Columbia were sown on peat:vermiculite:perlite::3:3:1 substrate after 3 days of vernalization at 4°C and were placed in climatecontrolled boxes. Transformation was carried out using the floral dip method 20 .

Gene isolation
cDNA samples were synthesized from the total RNA described above using the RNA PCR Kit (Takara, Dalian, China) following the manufacturer's instructions. The full-length FaARF4 CDS was cloned by RT-PCR using the FaARF4-F and FaARF4-R primers (Table S1) with Nde I and Sal I restriction sites added at the 5ʹ and 3ʹ ends, respectively.
PCR was performed under the following conditions: 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 57°C for 30 s, and 72°C for 3 min, with a final extension at 72°C for 5 min. The PCR products were ligated into the pMD18-T vector (Takara, Dalian, China) and subsequently transformed into Escherichia coli strain DH5α. Subsequently, positive colonies were selected, the gene sequence was amplified by PCR, and the product was sequenced at BGI-Shenzhen, China. The ExPasy ProtParam tool (http:// web.expasy.org/compute_pi/) and DNASTAR program (DNASTAR Inc., USA) were used to predict the structure of the encoded proteins. The amino acid sequences of FaARF4, FveARF4, and AtARF4 were aligned using DNAMAN 6.0 software (Lynnon Biosoft, USA).

Phylogenetic analysis
The ARF4 amino acid sequences from A. thaliana radiata (XP_014498033), Vitis vinifera (XP_002285019), Z. mays (XP_008656904), and Ziziphus jujuba (XP_024928576) were obtained from the NCBI nucleotide database (http://www.ncbi.nlm.nih.gov/nucleotide/). A phylogenetic tree was constructed using the neighborjoining method using MEGA 6.0 software with 1000 bootstrap replicates to evaluate the reliability of the phylogenetic grouping. The tree files were viewed and edited using MEGA 6.0 46 .

Quantitative RT-PCR analysis
Quantitative RT-PCR (qRT-PCR) was performed using the 7500 system (Applied Biosystems, Foster City, USA) according to the manufacturer's instructions with SYBR Green II (Takara, Dalian, China) and the FaARF3 and FaARF4 qRT-PCR primers listed in Table S1 (FaARF3-QF/R and FaARF4-QF/R). The experiments were conducted on three biological replicates, and the results were normalized using strawberry 26S rRNA as the housekeeping gene 45 . The qRT-PCR analysis of Fa26S was conducted with the following primers: Fa26S-F and Fa26S-R (Table S1). Tender leaves from transgenic strawberry plants and transgenic A. thaliana plants were harvested for gene expression analysis. Fve26S was used as the housekeeping gene for woodland strawberry, and AtACTIN8 was selected as the housekeeping gene for A. thaliana 20 (Table S1).
PCR was performed using the following thermal cycling conditions: 95°C for 10 min, followed by 40 cycles of 95°C for 10 s and 60°C for 1 min. The transcription levels were calculated using the 2 -ΔΔCT method 47 .

Subcellular localization
The full-length FaARF4 CDS without the termination codon was amplified using LA Taq polymerase (Takara, Dalian, China) with gene-specific primers FaARF4-GFP-F and FaARF4-GFP-R (Table S1) harboring Xba I and Xho I sites. PCR was performed under the following conditions: 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 3 min, with a final extension at 72°C for 5 min. The PCR products were inserted into the pGPTVII-GFP vector between the Xba I and Xho I sites to create the 35S::FaARF4-GFP (FaARF4-GFP) construct. Using pGPTVII-GFP containing 35S::GFP (GFP) as a control, vectors were injected into A. thaliana protoplasts and N. benthamiana leaves 45 . Two days after agroinfiltration, GFP fluorescence was imaged using a laser confocal fluorescence microscope (TCS SP8-SE; Leica, Wetzlar, Germany) with an excitation wavelength of 488 nm and a 505-530 nm bandpass emission filter.

Transcriptional activation analysis
The FaARF4 and FveARF4 CDSs were individually inserted into the pGBT9 vector to create the pGBT9-FaARF4 (BD-FaARF4) and pGBT9-FveARF4 (BD-FveARF4) constructs, respectively. The primers used are listed in Table S1. BD-FaARF4, BD-FveARF4, and the empty vector pGBT9 (BD) were then transformed into yeast (Saccharomyces cerevisiae) for transcriptional activity analysis using the Yeast Transformation Kit (Shaanxi Protein Interaction Biotechnology Company Limited, Shaanxi, China). The transformed yeast strains were grown on SD/−Trp medium and selected on SD/ −Trp-His-Ade medium. The transcriptional activity of FaARF4 was determined by observing yeast growth according to previously described methods 48 . X-α-gal (5bromo-4-chloro-3-indolyl-α-D-galactoside) was added to the SD/−Trp-His-Ade medium, and the resulting blue coloration was used as an indication of FveARF4 transcriptional activation activity 49 .

Identification of the tasiRNA3 target by RLM-5ʹ RACE
To identify FaARF4 as a tasiRNA3 target gene, 5ʹ-RNA ligase-mediated rapid amplification of cDNA ends (RLM-5ʹ RACE) was performed with the 5ʹ-Full RACE Kit (Takara, Dalian, China). According to the manufacturer's instructions, PCR products were ligated into the pMD18-T vector, and the recombinant vectors were verified by sequencing. All primers used for RLM-5ʹ RACE are listed in Table S1.

Transformation of A. thaliana and strawberry
The full-length FaARF4-coding region with added Nde I and Sal I restriction sites was amplified by PCR and inserted into the polylinker site of the plant overexpression vector pRI 101-AN. An RNAi vector, pART27, was constructed as previously described 45 . Briefly, a 333 bp fragment of FveARF4 was inserted into the left (EcoR I and Xho I) and right (Hind III and Xba I) multiple cloning sites of the pART27 vector. All primers used are listed in Table S1.
A. thaliana was transformed using the Agrobacterium tumefaciens strain GV3101 and the floral dip method 20 . Transgenic plants were selected on 1/2 Murashige and Skoog (MS) medium with 30 mg/L kanamycin. Transgenic A. thaliana plants were grown in climate-controlled boxes at 24°C under a 12/12 h light/dark cycle. A. tumefaciens strain EHA105 was used for 'Ruegen' transformation following the leaf-disk procedure 45 . Transgenic strawberry plants were planted in November in a greenhouse under sunlight, and the day and night temperatures were 15-25°C and 5-10°C, respectively.

Identification of transgenic A. thaliana and strawberry
To identify transgenic A. thaliana and strawberry plants, genomic DNA was extracted using the CTAB method 44 . Transgenic plants were confirmed using the 35S-F/FaARF4-R primer set, targeting the pRI 101-AN vector and the transgene sequence. The following PCR conditions were used: 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 3 min, with a final extension at 72°C for an additional 5 min. The expression of ARF4 in the leaves of transgenic and control plants was analyzed using qRT-PCR with the FaARF4 and Fve26S/AtACTIN8 primers (Table S1).

Yeast one-hybrid assay
The FaARF4 sequence was ligated into the pGAD424 vector. The promoter sequences (1600 base pairs) of FveAP1 and FveFUL were cloned and analyzed to determine whether AuxREs were present. Four FveAP1 promoter fragments and four FveFUL promoter fragments were then amplified and individually inserted into the pAbAi vector. Primers are listed in Table S1. The vectors containing FaARF4 and a fragment of the FveAP1 or FveFUL promoter were cotransformed into the Y1H yeast strain, and the Y1H assay was performed as previously described 50 .

Luciferase reporter assay
The 35S::FaARF4 and 35S::FveARF4 vectors were used as effectors, and the FveAP1 and FveFUL promoters were inserted into the pRI-mini35S-LUC vector and used as reporters. Primers are listed in Table S1. The luciferase reporter assay was carried out following previously described methods 49 . Finally, the luciferase fluorescence and luciferase signal intensity were imaged and measured using a living fluorescence imager (Lb985, Berthold, Germany).

Statistical analysis
The significance of the differences between the means was analyzed using Duncan's test with DPS 7.05 software 48 .
The mean values marked with * and ** indicate significant differences at the 5% and 1% levels, respectively.