Isolation and Functional Characterization of a LEAFY Gene in Mango (Mangifera indica L.)

LEAFY (LFY) plays an important role in the flowering process of plants, controlling flowering time and mediating floral meristem differentiation. Owing to its considerable importance, the mango LFY gene (MiLFY; GenBank accession no. HQ585988) was isolated, and its expression pattern and function were characterized in the present study. The cDNA sequence of MiLFY was 1152 bp, and it encoded a 383 amino acid protein. MiLFY was expressed in all tested tissues and was highly expressed in flowers and buds. Temporal expression analysis showed that MiLFY expression was correlated with floral development stage, and two relative expression peaks were detected in the early stages of floral transition and floral organ differentiation. Moreover, 35S::GFP-MiLFY fusion protein was shown to be localized to the nucleus of cells. Overexpression of MiLFY in Arabidopsis promoted early flowering and the conversion of lateral meristems into terminal flowers. In addition, transgenic plants exhibited obvious morphological changes, such as differences in cauline leaf shape, and the number of lateral branches. When driven by the MiLFY promoter, GFP was highly expressed in leaves, floral organs, stems, and roots, during the flowering period. Exogenous gibberellin (GA3) treatment downregulated MiLFY promoter expression, but paclobutrazol (PPP333) upregulated it. Bimolecular fluorescence complementation (BiFC) assays showed that the MiLFY protein can interact with zinc-finger protein 4 (ZFP4) and SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (MiSOC1D). Taken together, these results indicate that MiLFY plays a pivotal role in controlling mango flowering, and that it is regulated by gibberellin and paclobutrazol.


Introduction
Flowering is an important process in the plant life cycle, and the time of flowering (early or late) essentially determines the time of harvest. In recent decades, much progress has been made in understanding the physiological and molecular mechanisms underlying flowering time in plants. It is clear that plant flowering is mainly affected by various environmental and endogenous signals, such as daylength, temperature, drought, plant age, endogenous phytohormones, and exogenous plant growth regulators (PGRs) [1,2]. A complex gene regulatory network has been revealed in Arabidopsis. At least six flowering pathways and several genes, such as FLOWERING LOCUS T (FT), SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1), and LEAFY (LFY), which act as floral pathway integrators to activate downstream floral meristem identity genes, such as LFY and APETALA1 (AP1), cooperate to promote flowering [3].
LFY is a plant-specific transcription factor that is a master regulator of flower initiation and, as such, it determines floral fate in Arabidopsis [4]. Many plant genomes contain

Sequence and Phylogenetic Analysis of LFY
The sequence of a MiLFY homolog in mango was obtained. The cDNA length was 1152 bp, encoding a 383 amino acid protein (GenBank accession no. HQ585988). The full-length DNA sequence of MiLFY was 2170 bp, comprising three exons and two introns ( Figure 1A). The predicted protein molecular weight was 43.31 kDa, and the isoelectric point was 6.37. The N-and C-terminal regions of the MiLFY proteins were highly conserved across species ( Figure 1B). The C-and N-terminal regions contained a DNA-binding domain and a sterile alpha motif (SAM) domain. The MiLFY protein had 61.79% sequence homology with the Arabidopsis thaliana ortholog. Phylogenetic tree analysis showed that MiLFY was closely related to DlLFY (Dimocarpus longan), CsLFY (Citrus sinensis), and ClLFY (Clausena lansium) ( Figure 1C).

Analysis of MiLFY Gene Expression in Mango
The expression patterns of the MiLFY gene were analyzed via qRT-PCR. As shown in Figure 2A, MiLFY was expressed in all tested tissues, albeit at different levels. The expression of MiLFY was higher in the tissues of flowering branches than in those of nonflowering branches. The expression level was highest in flowers and lowest in leaves. The expression patterns of MiLFY in mature stems, mature leaves, and buds, at different flower development stages ( Figure 2B-D), were determined by qRT-PCR. MiLFY was more highly expressed in buds than in mature leaves. The highest expression level was found during the floral organ differentiation stage (January 2019) in buds, and during the flowering transition stage in leaves. However, MiLFY was highly expressed in mature stems at the flowering stage and flowering transition stage in the TaiNong No. 1 cultivar, but was highly expressed at the floral organ differentiation stage in the 'SiJiMi' cultivar.

Analysis of MiLFY Gene Expression in Mango
The expression patterns of the MiLFY gene were analyzed via qRT-PCR. As shown in Figure 2A, MiLFY was expressed in all tested tissues, albeit at different levels. The expression of MiLFY was higher in the tissues of flowering branches than in those of nonflowering branches. The expression level was highest in flowers and lowest in leaves. The expression patterns of MiLFY in mature stems, mature leaves, and buds, at different flower development stages ( Figure 2B-D), were determined by qRT-PCR. MiLFY was more highly expressed in buds than in mature leaves. The highest expression level was found during the floral organ differentiation stage (January 2019) in buds, and during the flowering transition stage in leaves. However, MiLFY was highly expressed in mature stems at the flowering stage and flowering transition stage in the TaiNong No. 1 cultivar, but was highly expressed at the floral organ differentiation stage in the 'SiJiMi' cultivar.

Subcellular Localization of MiLFY
To determine the subcellular localization of MiLFY, the protein was fused to green fluorescent protein (GFP) under the control of the constitutive cauliflower mosaic virus 35S (CaMV35S) promoter. The 35S::GFP-MiLFY vector and 35S::GFP-P1300 vector (control) were introduced into onion epidermal cells via Agrobacterium. As shown in Figure 3, the fluorescence signal of the 35S::GFP-P1300 control vector was observed throughout the cell. However, 35S::GFP-MiLFY fusion proteins were only visible in the nucleus, and were colocalized with 4′,6-diamidino-2-phenylindole (DAPI).

Subcellular Localization of MiLFY
To determine the subcellular localization of MiLFY, the protein was fused to green fluorescent protein (GFP) under the control of the constitutive cauliflower mosaic virus 35S (CaMV35S) promoter. The 35S::GFP-MiLFY vector and 35S::GFP-P1300 vector (control) were introduced into onion epidermal cells via Agrobacterium. As shown in Figure 3, the fluorescence signal of the 35S::GFP-P1300 control vector was observed throughout the cell. However, 35S::GFP-MiLFY fusion proteins were only visible in the nucleus, and were colocalized with 4 ,6-diamidino-2-phenylindole (DAPI).

Phenotypic Analysis of Transgenic Plants with Overexpression
A vector overexpressing the MiLFY gene was transformed into wild-type (WT) Arabidopsis to study the function of MiLFY. WT plants were used as negative controls, and WT plants transformed with the pBI121 empty vector were used as positive controls. Semiquantitative PCR was used to detect the expression level of the exogenous MiLFY gene in transgenic plants, and qRT-PCR was used to detect the expression level of the Arabidopsis flowering-related genes AtFT, AtAP1, and AtSOC1 in transgenic and control plants.
MiLFY was expressed normally in transgenic Arabidopsis thaliana, but not in the control plants ( Figure 4A Transgenic plants overexpressing the MiLFY gene exhibited different morphological phenotypes ( Figure 4A). All axillary rosette branches (basal branches) and main stem branches (stem branches) developed single flowers ( Figure 4A-b,c,f). In addition, the pedicels were wrapped by curly cauline leaves, and the parts of the single flowers were opposite or whorled on the main stem ( Figure 4A-c,d). Unusual siliques developed on the upper stems and basal branches ( Figure 4A-e,g,i), and the transgenic plants had fewer flower petals than the control plants ( Figure 4A-h,j). All transgenic MiLFY1-overexpressing Arabidopsis lines except OE-13 had the same morphological characteristics. The morphology of the OE-13 line was similar to that of the control lines.
The experimental materials were sampled when the transgenic plants were flowering and the control plants were not. RNA was extracted from the aerial portions of the plants 27 d after germination, and this RNA was subsequently reverse-transcribed into cDNA. qRT-PCR was used to determine whether the expression of endogenous genes in transgenic Arabidopsis was affected by MiLFY overexpression ( Figure 5). Compared with the WT Arabidopsis plants, and the plants transformed with the empty vector, the transgenic

Phenotypic Analysis of Transgenic Plants with Overexpression
A vector overexpressing the MiLFY gene was transformed into wild-type (WT) Arabidopsis to study the function of MiLFY. WT plants were used as negative controls, and WT plants transformed with the pBI121 empty vector were used as positive controls. Semiquantitative PCR was used to detect the expression level of the exogenous MiLFY gene in transgenic plants, and qRT-PCR was used to detect the expression level of the Arabidopsis flowering-related genes AtFT, AtAP1, and AtSOC1 in transgenic and control plants.
MiLFY was expressed normally in transgenic Arabidopsis thaliana, but not in the control plants ( Figure 4A Transgenic plants overexpressing the MiLFY gene exhibited different morphological phenotypes ( Figure 4A). All axillary rosette branches (basal branches) and main stem branches (stem branches) developed single flowers ( Figure 4A-b,c,f). In addition, the pedicels were wrapped by curly cauline leaves, and the parts of the single flowers were opposite or whorled on the main stem ( Figure 4A-c,d). Unusual siliques developed on the upper stems and basal branches ( Figure 4A-e,g,i), and the transgenic plants had fewer flower petals than the control plants ( Figure 4A-h,j). All transgenic MiLFY1-overexpressing Arabidopsis lines except OE-13 had the same morphological characteristics. The morphology of the OE-13 line was similar to that of the control lines.

Activity of GUS Driven by the MiLFY Promoter in Transgenic Arabidopsis Lines
In our previous study, the MiLFY promoter was cloned, and its cis-elements were analyzed [16]. In the present study, the 35S promoter in the pBI121 vector was replaced with the MiLFY promoter to drive GUS gene expression. The pLFY-GUS vector was subsequently transformed into WT Arabidopsis, and WT plants and pBI121-GUS vector-transformed plants were used as the controls. Different transgenic and control seedlings, as well as organs of mature plants, were subjected to histochemical staining for GUS to detect MiLFY promoter expression. As shown in Figure 6, the stem and stem apex in the transgenic Arabidopsis plants exhibited sites of GUS expression at 5 d (two-true leaf stage) and 10 d (four-true leaf stage). Moreover, GUS expression was detected in all mature organs, including flowers, stems, leaves, siliques, and roots. The expression level of the MiLFY promoter was relatively high in floral organs and stems. However, no staining was detected in control samples.
Many PGR response elements have been found in the MiLFY promoter [17]. In this study, 15 d old transgenic plants harboring the MiLFY promoter were treated with GA3 and PPP333, and the control lines were treated with ddH2O. Promoter activity was affected by the PGRs, as determined by analyzing the expression level of the GUS gene. Compared with those in the plants treated with ddH2O, the GUS expression levels in the transgenic plants harboring the MiLFY promoter were reduced by GA3 treatment. In contrast, the GUS expression level was significantly increased by PPP333 treatment (Figure 7).

Activity of GUS Driven by the MiLFY Promoter in Transgenic Arabidopsis Lines
In our previous study, the MiLFY promoter was cloned, and its cis-elements were analyzed [16]. In the present study, the 35S promoter in the pBI121 vector was replaced with the MiLFY promoter to drive GUS gene expression. The pLFY-GUS vector was subsequently transformed into WT Arabidopsis, and WT plants and pBI121-GUS vectortransformed plants were used as the controls. Different transgenic and control seedlings, as well as organs of mature plants, were subjected to histochemical staining for GUS to detect MiLFY promoter expression. As shown in Figure 6, the stem and stem apex in the transgenic Arabidopsis plants exhibited sites of GUS expression at 5 d (two-true leaf stage) and 10 d (four-true leaf stage). Moreover, GUS expression was detected in all mature organs, including flowers, stems, leaves, siliques, and roots. The expression level of the MiLFY promoter was relatively high in floral organs and stems. However, no staining was detected in control samples.
Many PGR response elements have been found in the MiLFY promoter [17]. In this study, 15 d old transgenic plants harboring the MiLFY promoter were treated with GA 3 and PPP 333 , and the control lines were treated with ddH 2 O. Promoter activity was affected by the PGRs, as determined by analyzing the expression level of the GUS gene. Compared with those in the plants treated with ddH 2 O, the GUS expression levels in the transgenic plants harboring the MiLFY promoter were reduced by GA 3 treatment. In contrast, the GUS expression level was significantly increased by PPP 333 treatment (Figure 7).

Yeast Two-Hybrid (Y2H) Screening and Confirmation by Bimolecular Fluorescence Complementation (BiFC) Assays
Using Y2H screens, we identified the following two candidate-interacting proteins: zinc-finger protein 4 (ZFP4) and MiSOC1D. Further BiFC assays using a split yellow fluorescent protein (YFP) system in onion epidermal cells were used to verify the interactions. The coexpression of MiLFY and MiZFP4, and of MiLFY and MiSOC1D, resulted in YFP signals in the nucleus of onion epidermal cells (Figure 8).

Discussion
LFY not only regulates flowering time but also has a specific function as a floral meristem identity gene in the flower development pathway [18]. Moreover, LFY links floral

Yeast Two-Hybrid (Y2H) Screening and Confirmation by Bimolecular Fluorescence Complementation (BiFC) Assays
Using Y2H screens, we identified the following two candidate-interacting proteins: zinc-finger protein 4 (ZFP4) and MiSOC1D. Further BiFC assays using a split yellow fluorescent protein (YFP) system in onion epidermal cells were used to verify the interactions.

Yeast Two-Hybrid (Y2H) Screening and Confirmation by Bimolecular Fluorescence Complementation (BiFC) Assays
Using Y2H screens, we identified the following two candidate-interacting proteins: zinc-finger protein 4 (ZFP4) and MiSOC1D. Further BiFC assays using a split yellow fluorescent protein (YFP) system in onion epidermal cells were used to verify the interactions. The coexpression of MiLFY and MiZFP4, and of MiLFY and MiSOC1D, resulted in YFP signals in the nucleus of onion epidermal cells (Figure 8).

Discussion
LFY not only regulates flowering time but also has a specific function as a floral meristem identity gene in the flower development pathway [18]. Moreover, LFY links floral

Discussion
LFY not only regulates flowering time but also has a specific function as a floral meristem identity gene in the flower development pathway [18]. Moreover, LFY links floral induction with flower initiation [19]. In the present study, a LFY homolog from the mango cultivar SiJiMi was identified and named MiLFY. The function of the MiLFY gene and the regulation of its promoter activity were systematically analyzed, thereby helping to determine the genetic and molecular mechanisms underlying the involvement of LFY genes in mango flowering.
Many woody fruit tree species, such as navel orange [6], grapevine [20], Ziziphus jujube [21], and longan [22], contain only a single copy of the LFY gene. However, multiple copies are present in various species, namely, pear [23], Chinese quince [23], and loquat [24], all of which contain two copies. Only a single copy of MiLFY was found in mango. Phylogenetic analysis indicated that MiLFY is more closely related to its orthologs in longan, citrus, and wampee than to those in other woody fruit tree species. The N-and C-terminal regions of the LFY protein are highly conserved across species; this conserved structure assures the similarity of LFY gene function [25].
The expression pattern of the LFY gene varies among woody species. For example, the London planetree PlacLFY gene was found to be expressed mainly in male and female inflorescences and was only weakly expressed in stems and young leaves [26]. Precocious trifoliate orange CiLFY is highly expressed in mature apex bud, flower, and stem tissues, but not in juvenile tissues; high CiLFY expression was found to be maintained from December to March [27]. Jatropha curcas JcLFY is expressed in inflorescence buds, flower buds, and carpels, with the highest expression occurring in the early developmental stage of flower buds [28]. Prunus mume PmLFY is highly expressed in floral buds, leaf buds, pistils, and seeds, with the highest expression occurring in floral buds during the floral differentiation stage [7]. However, peach PpLFL was expressed mainly in leaves and in the petal primordia of the shoot apical meristem during the floral induction period [29]. Populus tomentosa PtLFY mRNA was found to be highly abundant in the roots and floral buds of both male and female flowers [30]. In the present study, MiLFY was more highly expressed in the flowers and stems of flowering branches than in those of nonflowering branches, and its expression was correlated with the floral development stage in different organs in different cultivars. Litchi LcLFY was expressed primarily in flower buds, but was barely detectable in stems, mature leaves, petioles, and pedicels [31].
Overexpression of the LFY gene causes early flowering in many plant species, although the overall plant morphology remains normal [7,29,31,32]. However, some differences do occur. JcLFY overexpression induces early flowering but causes the production of single flowers and terminal flowers in Arabidopsis [28]. Overexpression of VpLFY2 without overexpression of VpLFY1 causes precocious flowering in Arabidopsis [31]. AfLFY expression in transgenic tobacco plants promotes precocious flowering, and these transgenic plants exhibit obvious changes in leaf shape [33]. In the present study, overexpression of MiLFY promoted early flowering in Arabidopsis, but considerable morphological variation was observed. For example, indeterminant inflorescences became single flowers, and cauline leaves were curled and intertwined with pedicels. In addition, the number of flower petals was decreased in the transgenic plants, and some pods were short or curved. The phenotype of the transgenic plants was similar to that of plants overexpressing PmLFY1 [7]. These results suggest that the LFY gene can promote flower formation in different species but differentially influences plant development.
Gene expression is regulated by cis-elements in promoters. PlacLFY promoter activity was detected in the shoot apices, young leaves, young fruits, petioles, and young/old stems of pPlacLFY::GUS transgenic tobacco, consistent with the expression pattern of PlacLFY in London planetree [26]. CcLFY promoter expression is influenced by low temperature and dark conditions [34]. The promoter of MiLFY in mango was previously characterized [17]. In the present study, the MiLFY gene promoter was transformed into Arabidopsis, and analysis of GUS staining showed that the MiLFY promoter was active mainly in flowers, stems, leaves, and roots during the flowering period, consistent with the expression patterns of MiLFY in mango (Figure 2). Exogenous GA 3 promotes Arabidopsis flowering but inhibits mango flowering, and PPP 333 , a synthetic inhibitor of GA 3 , promotes mango flower formation [1]. In the present study, we treated MiLFY promoter-containing transgenic seedlings with exogenous GA 3 and PPP 333 and found that the expression of the MiLFY promoter was significantly inhibited by GA 3 but enhanced by PPP 333 . These results suggest that the MiLFY gene is involved in phytohormone-mediated regulation of mango flowering.
As a floral integration factor, the LFY gene plays an important regulatory role in the floral network [35]. Winter (2011) identified direct LFY target genes throughout the genome. These target genes were found to be involved in flowering time, floral organ development, phytohormone responses, and biotic stimulus responses. In the present study, we found that two proteins, MiZFP4 and MiSOC1D, can directly interact with LFY. ZFPs constitute one of the largest transcription factor families, whose members are highly involved in transcriptional regulation of flowering induction, floral organ morphogenesis, and stress responses [36]. SOC1 is a floral integration factor, and overexpression of MiSOC1 promotes early flowering in A. thaliana [37].
In conclusion, we functionally characterized the MiLFY gene in mango and found that overexpression of MiLFY significantly promoted flowering in transgenic plants. Moreover, transgene expression led to significant morphological variation, including changes in floral organs and leaf morphology, in the transgenic plants. For the first time, we found that GA 3 treatment inhibits, but PPP 333 promotes, MiLFY promoter activity. This pattern is consistent with the finding that GA 3 inhibits mango flower development while PPP 333 promotes mango flowering. Protein interaction analysis showed that by directly interacting with MiZFP4 and MiSOC1D, the MiLFY protein regulates mango flowering.

Plant Materials
The mango cultivars SiJiMi and TaiNong No. 1 were grown in an orchard of the College of Agriculture, Guangxi University, Nanning, Guangxi, China. For tissue expression analysis, leaves, stems, and buds or flower tissues were collected from the flowering and nonflowering branches of the same plants on 4 January 2019. For seasonal expression analysis, leaves, stems, and buds or flowers were collected each month from 5 November 2018 to 6 March 2019. All samples were used for experiments before storage at −80 • C. The Arabidopsis ecotype Columbia (Col-0) plants used for transformation were maintained in our laboratory.

Isolation of the MiLFY Gene
Total RNA was extracted from mango leaves by using an RNAprep pure kit (DP441) (Tiangen, Beijing, China). First-strand cDNA was synthesized with M-MLV reverse transcriptase (TaKaRa, Dalian, China) using 1 µg of RNA according to the manufacturer's instructions. Genomic DNA was isolated using the cetyl-trimethylammonium bromide (CTAB) method with minor modifications. In our previous study, we obtained the sequence of a LFY gene from transcriptome data from the mango cultivar SiJiMi (unpublished data) and, in this study, we further verified the correctness of its sequence by RT-PCR with the primers LFY-F and LFY-R (Table S1). PCR amplification was performed as described in a previous study [2]. The primers LFY-F and LFY-R were also used to amplify genomic DNA.

Analysis of MiLFY Gene Expression
The expression pattern of MiLFY was analyzed with an ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) and SYBR Premix Ex Taq II (Takara, Dalian, China) according to the manufacturers' instructions. The reaction mixture contained the following components: reaction solution 10 µL, cDNA 2 µL (50 ng/µL), up-and downstream primers 0.5 µL (10 µM) each, ROX reference dye II 0.8 µL, and sterile water to 20 µL. PCR amplification was performed with the following thermal cycling program: 95 • C for 30 s; 40 cycles at 95 • C for 5 s, 60 • C for 34 s, and 95 • C for 15 s; melting curve analysis was performed at 95 • C for 15 s and 60 • C for 1 min. The gene-specific primers used are listed in Table S1. MiActin1 was used as the internal reference gene in mango [16]. The delta-delta Ct method was used to calculate the relative gene expression [38]. The data are presented as the average of at least three technical replicates.

Subcellular Localization Analysis
To analyze the subcellular localization of MiLFY, the full coding sequence (CDS) of MiLFY without the termination codon was inserted into the Xba I and BamH I sites in the P1300 vector to express the 35S::GFP-MiLFY fusion construct under the control of the CaMV35S promoter [39]. The 35S::GFP-MiLFY vector was transformed into Agrobacterium tumefaciens EHA105a, and the transformants were used to infect onion (Allium cepa) epidermal cells. The onion epidermal cells were then observed at a wavelength of 488 nm with a confocal laser-scanning microscope (TCS-SP8MP, Leica, Germany). Nuclei were confirmed by DAPI staining.

Vector Construction and Transformation of Arabidopsis
The full CDS of the MiLFY gene was inserted into the XbaI and XmaI sites in pBI121 under the control of the CaMV35S promoter to construct the overexpression vector. The pBI121-MiLFY and pBI121 vectors were separately transformed into WT Arabidopsis via A. tumefaciens EHA105 using the floral dip method [40]. The transgenic plants were first grown on half-strength Murashige and Skoog (MS) medium supplemented with kanamycin (100 mg/L). The positive transgenic plants were further confirmed via PCR-based amplification of DNA. Homozygous T3 transgenic plants were used for subsequent experiments.
The bolting time, flowering time, number of rosette leaves, and plant height of the WT Arabidopsis and empty vector-transformed Arabidopsis plants (which were used as controls) were recorded or measured under long-day (LD) conditions. For semiquantitative PCR and qRT-PCR analysis, 24-d-old transgenic and WT Arabidopsis seedling leaves were collected for total RNA extraction. Total RNA was extracted, and first-strand cDNA was synthesized as described above. Semi-quantification was mainly used to detect whether MiLFY gene was normally expressed in transgenic plants. qRT-PCR was used to measure the expression levels of the endogenous flowering-related genes in Arabidopsis. The Arabidopsis Actin2 gene was used as the internal control. The semiquantitative PCR and qRT-PCR methods were described in a previous study [15].

MiLFY Promoter Activity Assay
In our previous study, a 1314-bp fragment of the MiLFY promoter was amplified through thermal asymmetric interlaced PCR (TAIL-PCR) [17]. This fragment was inserted into the pBI121 vector in place of the 35S promoter to activate GUS, and plants containing the 35S promoter vector and WT plants were used as the controls. pMiLFY-pBI121 transgenic plants were obtained via the above approach. Homozygous T3 transgenic plants were used for subsequent experiments. Growing plants at different stages were harvested and immersed in GUS staining buffer (Real-Time, Beijing, China) at the same time. The plant material was incubated in a constant-temperature incubator at 37 • C for 24 h in the dark and was then decolorized with 75% ethanol until the control samples became white. Images showing sites of GUS staining were acquired using an ultra-depth-of-field 3D microscopy system (VHX-6000, Leica, Germany).
To determine the effects of plant growth regulators (PGRs) on MiLFY promoter activity, 20-d-old Arabidopsis plants were sprayed with 150 mL of ddH 2 O, 10 µM GA 3 , or 10 µM PPP 333 . Materials were collected at 0 d, 3 d, and 6 d after treatment. Total RNA was extracted from these samples using an RNAprep pure kit (DP441) (Tiangen, Beijing, China), and the GUS gene expression level was measured via qRT-PCR. The gene-specific primers used are listed in Table S1. The method was the same as that described above.

Y2H Screening and Confirmation by BiFC Assay
The full-length CDS of MiLFY was amplified and inserted into the pGBKT7 vector. A cDNA library comprising bud, leaf, and flower samples, obtained from SiJiMi mango trees during flower development, was constructed and stored in our laboratory. cDNA library screening was performed according to the Matchmaker Gold Yeast Two-Hybrid System User Manual (Clontech). All positive clones were sequenced and functionally annotated using the BLAST online search engine tool. To further verify protein-protein interactions, the full-length CDSs of the candidate-interacting proteins were inserted into the pSPYCE vector, and the full-length CDS of MiLFY was inserted into the pSPYNE vector. All fusion constructs were subsequently transformed into A. tumefaciens (strain GV3101). Subsequently, the different fusion vectors were transformed into onion epidermal cells. Fluorescence signals were observed 48 h after infiltration using laser-scanning confocal microscopy (TCS SP8, Leica, Germany)