H. coronarium was cultivated in the open field at the College of Horticulture, South China Agricultural University (Guangzhou, China, 23.16°N, 113.36°E). The plant materials (root, rhizome, leaf, leaf buds, inflorescence buds) were collected from 19:00 on June 15, 2024. Root, rhizome and leaf were collected during reproductive growth. All tissues were immediately frozen in liquid N and stored at − 80 until use.

Arabidopsis PEBP sequences were downloaded from TAIR (https://www.arabidopsis.org/). The H. coronarium genome (unpublished) was provided by Beijing Novogene Bioinformatics Technology Corporation (China). Arabidopsis PEBP sequences were used as queries to identify HcPEBP sequences from the H. coronarium genome using the BLAST module in TBtools (E-value ≤ 1.0 × 10^-5) (Chen et al., 2020). Additionally, The HMM profile of the PEBP domain (PF01161) from the Pfam database (http://pfam.xfam.org/) (Finn et al., 2006) was used in the Simple HMM search module in TBtools. Sequences identified by both BLAST and HMM searches were merged, and duplicates were removed. All candidate sequences were analyzed for the PEBP domain using the NCBI-Conserved Domain Database (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) (Morris et al., 2015). Only sequences containing the complete PEBP domain were retained for further analysis.

Chromosomal details (length, gene density, and gene positions) were extracted from the H. coronarium genome using TBtools (Chen et al., 2020). The chromosomal locations of HcPEBP genes were visualized using TBtools. Genome sequences and annotations for wild banana, pineapple, and rice were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/) and Ensemble Plants (https://plants.ensembl.org). Gene duplication events and collinear relationships were analyzed using the One-step MCScanX-Super Fast module in TBtools with default settings. Collinear relationships were visualized using the Dual Synteny Plotter in TBtools.

Conserved motifs within HcPEBP proteins were analyzed using the MEME suite (https://meme-suite.org/meme/tools/meme) (Bailey et al., 2015). The motif count was set to eight, with other parameters kept at default settings. The identified motifs and gene structures were visualized using the Gene Structure View (advanced) module in TBtools.

Physicochemical properties of HcPEBP proteins, including molecular weight (MW), isoelectric point (pI), instability index, aliphatic index, and GRAVY (grand average of hydropathicity), were analyzed using the ProtParam tool on the ExPASy online platform (http://www.expasy.ch/tools-/pi_tool.html) (Artimo et al., 2012). Subcellular localization predictions for HcPEBP proteins were made using the WoLF PSORT tool (Horton et al., 2007).

PEBP protein sequences from Arabidopsis and rice were downloaded from TAIR and the Rice Genome Annotation Project (RGAP, http://rice.uga.edu/) databases, respectively. The GenBank accession numbers for PEBPs in Arabidopsis and rice are listed in Supplementary Table S1 . Multiple sequence alignments of PEBP proteins from H. coronarium, Arabidopsis, and Oryza sativa were conducted using TBtools. A phylogenetic tree comprising 40 PEBP proteins from H. coronarium, Arabidopsis, and rice was constructed with TBtools using default parameters and visualized with Evolview (https://www.evolgenius.info/evolview-v2) (Zhang et al., 2012). Amino acid sequences of HcPEBPs were aligned using the ClustalW algorithm in MEGA11 and displayed using GeneDoc (Tamura et al., 2021).

Sequences 2000 bp upstream of the transcription start site (ATG) for HcPEBP genes were extracted from the genome sequence using TBtools. Cis-acting elements within the promoter sequences of HcPEBPs were predicted using the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al., 2002), with results analyzed, classified, and visualized using TBtools.

The expression patterns of HcPEBP genes across various tissues of H. coronarium were investigated. Relative expression levels in five tissues (root, rhizome, leaf, inflorescence bud, leaf bud) were measured by qRT-PCR. Morphological characteristics of these tissues are depicted in Supplementary Figure S1 . Primer sequences are provided in Supplementary Table S2 .

Total RNA was isolated from plant materials using the HiPure Plant RNA Mini Kit (Magen, Guangzhou, China), following the manufacturer’s protocol. DNA was extracted using the DNA Quick Plant System (Tian Gen, Beijing, China), as per the provided manual. For gene cloning, cDNA was synthesized using the PrimeScript™ RT Reagent Kit with gDNA Eraser (TaKaRa, Japan). For qRT-PCR, cDNA was reverse-transcribed using the Evo M-MLV RT Mix Kit with gDNA Clean for qPCR Ver.2 (Accurate Biology, Hunan, China). qRT-PCR was performed using Hieff^® qPCR SYBR Green Master Mix (Yeasen Biotechnology, Shanghai, China) on an ABI 7500 Fast Real-Time PCR system (Applied Biosystems, USA). Reaction conditions followed a previously described protocol (Wang et al., 2021). The GAPDH gene was used as an internal reference for normalization. Relative gene expression levels were calculated using the 2−ΔΔCt method. Statistical analyses for Significant differences were determined using IBM SPSS Statistics and Origin 2021.

The coding sequence (CDS) of HcPEBP11 was amplified from H. coronarium cDNA using Phanta Max Super-Fidelity DNA Polymerase (Vazyme, Nanjing, China), following the manufacturer’s protocol. The CDS was cloned into the pOx vector (provided by the State Key Laboratory for Conservation and Utilization of Subtropical Agro-Bioresources, South China Agricultural University, China) using the ClonExpress II One Step Cloning Kit (Vazyme, Nanjing, China). The constructed plasmid was transformed into Agrobacterium tumefaciens GV3101 (WeDi, Shanghai, China) using the freeze-thaw method. Primers used are listed in Supplementary Table S2 . Tobacco plants (Nicotiana tabacum cv. W38) were transformed using the leaf disc method (Horsch et al., 1986). Agrobacterium cultures containing pOx-HcPEBP11 were grown at 28°C in liquid medium with 50 µg/mL kanamycin and 25 µg/mL rifampicin until reaching an OD600 of 0.6–0.8. Bacteria were pelleted by centrifugation at 5000 rpm for 8 minutes and resuspended in MS liquid medium containing 100 µM acetosyringone to an OD600 of 0.7–0.8. Young leaf explants from sterile tobacco seedlings were pre-cultured for three days in darkness before transformation. Transformed plants were regenerated through a series of cultures: co-culture, bacteriostatic, differentiation, induction, rooting, and screening.

Putative transgenic tobacco plants resistant to 20 mg/L hygromycin B in 1/2 MS medium were screened using PCR (Cordero Otero and Gaillardin, 1996; Thion et al., 2002; Walter et al., 1989). Universal primers pOx-F/R (flanking the multiple cloning sites of the pOx vector) and specific primers HPH-F/R (targeting the hygromycin B phosphotransferase gene) were used to confirm the integration of HcPEBP11 into the tobacco genome. Primer sequences are listed in Supplementary Table S2 . Semi-quantitative RT-PCR (Semi-qRT-PCR) and qRT-PCR were performed to verify HcPEBP11 expression in transgenic lines (Tripathi et al., 2022). Total RNA was extracted from leaves at the flower bud stage. Semi-qRT-PCR was conducted using ABI VeritiPro PCR (Thermo Fisher) and Phanta Max Super-Fidelity DNA Polymerase, with specific primers for HcPEBP11. Transgenic and wild-type tobacco lines were grown under natural light conditions. The time to bolting (appearance of the first flower bud) and flowering (blooming of the first flower) was recorded to assess the role of HcPEBP11 in flowering regulation.

The coding sequence (CDS) of the HcPEBP11 gene, lacking termination codons, was successfully cloned into the p35S-cGFP vector. Subsequently, the resultant recombinant vector underwent transformation into the A. tumefaciens strain GV3101 via the heat shock method. For the transformation assay, leaves of N. benthamiana at the five-leaf stage were utilized (Kokkirala et al., 2010). Post 72 hours of infiltration, the green fluorescent protein (GFP) signals were detected using a confocal laser scanning microscope.

In H. coronarium, 14 PEBP genes were identified and named HcPEBP1–14 based on their chromosomal distribution. Key details about the HcPEBP family and the physicochemical properties of their encoded proteins are summarized in Table 1 and Supplementary Table S3 . Protein length ranges from 142 aa (HcPEBP7) to 201 aa (HcPEBP14), with an average of 175 aa. Molecular weight averages 19.71 kDa, ranging from 15.97 kDa (HcPEBP7) to 22.90 kDa (HcPEBP14). Theoretical isoelectric points span from 5.93 (HcPEBP11) to 11.28 (HcPEBP7). Instability Index ranges from 34.77 (HcPEBP4) to 61.06 (HcPEBP13). Aliphatic index aries between 87.73 (HcPEBP8) and 72.92 (HcPEBP13). All 14 HcPEBP proteins are hydrophilic. Except for HcPEBP13 (mitochondrial), all other HcPEBPs are predicted to be cytoplasmic proteins.

In H. coronarium, 14 HcPEBP genes were identified. 9 genes (HcPEBP1-9) mapped to six chromosomes (Hc-2, Hc-3, Hc-5, Hc-10, Hc-11, Hc-14), while the remaining five genes (HcPEBP10-14) are located on four genome scaffolds. Intraspecific collinearity analysis revealed three pairs of duplicated genes: HcPEBP2 and HcPEBP7, HcPEBP4 and HcPEBP14, HcPEBP6 and HcPEBP12 ( Figure 1A ). To assess the impact of selection pressure on the collinear HcPEBP gene pairs, the Ka/Ks ratios (non-synonymous to synonymous substitutions) were calculated for each duplicated pair. A Ka/Ks = 1 indicates neutral selection, Ka/Ks < 1 indicates purifying selection, and Ka/Ks > 1 suggests positive selection. The Ka/Ks ratios for the three collinear HcPEBP gene pairs ranged from 0.087 to 0.295 ( Supplementary Table S4 ), indicating that these duplicated gene pairs have undergone purifying selection, which maintains functional stability by removing deleterious mutations.

To further explore the evolutionary relationships of HcPEBP genes with those of other species, a collinearity analysis was conducted using O. sativa, Musa balbisiana, and Ananas comosus alongside HcPEBP family members. The results revealed that the HcPEBP gene family exhibits 17 collinearities with wild banana, 2 with rice, and 4 with pineapple ( Figure 1B , Supplementary Table S5 ). Among these species, HcPEBP genes show the closest evolutionary relationship with PEBP members from wild banana. Notably, HcPEBP4 and HcPEBP10 display collinearities with all three species, highlighting their conserved evolutionary roles.

To investigate the evolutionary relationships of HcPEBP genes with PEBP genes from other species, a phylogenetic tree was constructed using multiple sequence alignments of amino acids from 6 PEBP proteins of A. thaliana, 20 PEBP proteins of O. sativa, and 14 PEBP proteins of H. coronarium ( Figure 2 ). The phylogenetic tree is divided into three subgroups: MFT, TFL1, and FT. HcPEBP8 clusters with AtMFT, OsMFT1, and OsMFT2, indicating it belongs to the MFT subgroup. Four genes (HcPEBP2/5/7/13) cluster with AtBFT, AtTFL1, AtATC, and OsRCEs, suggesting these four genes belong to the TFL1 subgroup. Nine genes (HcPEBP1/3/4/6/9/10/11/12/14) cluster with AtFT, AtTSF, Hd3a, and OsFTLs, indicating they belong to the FT subgroup.

Phylogenetic tree analysis revealed that HcPEBP1/3/4/6/9/10/11/12/14 belong to the FT subfamily, while HcPEBP2/5/7/13 belong to the TFL1 subfamily. To further investigate the structural and functional conservation of these proteins, amino acid sequence alignment was performed for HcPEBPs, along with AtFT and AtTFL1 ( Figure 3 ). All HcPEBP proteins contain the GxHR motif, a hallmark of the PEBP family. Except for HcPEBP7, all other HcPEBP proteins possess the DPDxP motif, another key functional domain. The C-terminal amino acid sequence can be divided into four segments (I–IV), each of which is essential for the FT protein’s function. The conserved motif LYN/IYN/in FT-like proteins is located in segment III, playing a critical role in enzymatic activity. In segment IV region, FT-like proteins contain the motif xGxGGR, whereas TFL1-like proteins contain a different motif (TAARRR) in the same region. FT-like proteins HcPEBP1/3/6/10/11/12 contain the key amino acid residues Tyr85, Trp138, and Gln140, which are critical for their function. In contrast, TFL1-like proteins HcPEBP2/5/13 possess the key residues His88 and Asp144, which are characteristic of their functional role. Interestingly, in FT-like proteins HcPEBP4/9/14, the residue Tyr85 is mutated to His, Phe, His, respectively.

A phylogenetic tree containing only HcPEBP genes was constructed, dividing them into three subfamilies ( Figure 4A ). To further investigate the characteristics of HcPEBP genes, we conducted conserved motif and gene structure analyses. In the conserved motif analysis ( Figure 4B ), 8 conserved motifs (named motifs 1–8) were identified, with detailed sequence information provided in Supplementary Table S6 . The analysis revealed the following patterns: Motifs 3, 5, and 7 are shared by all HcPEBPs, indicating their fundamental role in the protein family. Motif 1 is present in all HcPEBPs except HcPEBP7. Motif 2 is found in all HcPEBPs except HcPEBP3. Motif 4 is absent in HcPEBP3 and HcPEBP14. Motif 6 is exclusively present in the FT subfamily, highlighting its specificity to this group. Motif 8 is unique to the TFL1 and MFT subfamilies, suggesting a functional role specific to these groups. Additionally, conserved domain analysis confirmed that all 14 HcPEBPs contain the PEBP domain ( Figure 4C ).

Gene structure analysis revealed that the 14 HcPEBP genes share a generally similar structural organization ( Figure 4D ). Among them: nine HcPEBP genes (64.3% of the total) contain four exons and three introns, representing the most common structural pattern in the HcPEBP gene family. Three HcPEBP genes (HcPEBP9/11/14) consist of five exons and four introns. HcPEBP3 with three exons and two introns. HcPEBP7 contain only two exons and one intron.

Cis-acting elements are critical for regulating gene transcription. In this study, we analyzed the 2000 bp upstream sequences of the HcPEBP genes’ start codon to predict and characterize cis-acting elements ( Figure 5A ). A total of 45 distinct cis-acting elements were identified and classified into four functional categories ( Figure 5B ): hormone-responsive (98 elements), development-related (28 elements), light-responsive (161 elements), and defense and stress responsiveness-related (40 elements), summing up to 327 elements. Among these, light-responsive elements were the most abundant, while development-related element were the least frequent. All HcPEBP genes, except HcPEBP4 and HcPEBP14, contain the G-box element (light-Responsive Element). All genes, except HcPEBP10, contain the Box4 element. HcPEBP1/3/6/7 lack any development-related elements. Detailed information on the cis-acting elements in the promoter regions of HcPEBP genes is provided in Supplementary Table S7 .

To investigate tissue-specific gene expression patterns, we analyzed the expression levels of HcPEBP genes across various tissues—root, rhizome, leaf, leaf bud, and inflorescence bud using qRT-PCR ( Supplementary Figure S1 ). HcPEBP5 and HcPEBP8 showed the highest expression in roots, with HcPEBP8 significantly exceeding HcPEBP5 ( Figures 6E, H, O ). HcPEBP1/2/3/4/9/10/12 and HcPEBP14 exhibited significantly elevated expression in leaves compared to other tissues ( Figures 6A–D, I, J, L, N ). Among these, HcPEBP1/2/14 had notably higher expression levels ( Figure 6Q ). HcPEBP6/7/13 were most highly expressed in Rhizomes, with HcPEBP7 levels significantly surpassing those of HcPEBP6 and HcPEBP13 ( Figures 6F–G, M, P ). HcPEBP11 expression was significantly higher in inflorescence buds compared to other tissues ( Figure 6K ). Additionally, transcriptome data and qRT-PCR results indicate that the expression of HcPEBP11 progressively increases during the development of inflorescence buds. ( Figures 7A, C , Supplementary Figure S2 , Supplementary Table S9 ). Seven FT subfamily genes (HcPEBP1/3/4/9/10/12/14) exhibited high expression levels in leaves ( Figures 6A, C, D, I, J, L, N ). To further investigate their roles, we analyzed their expression patterns in leaves at four developmental stages of the apical meristem ( Figures 7B, C ). The results revealed that HcPEBP1 showed a continuous increase in expression from stage 1 to stage 4. HcPEBP3/4/9/10/12 displayed an initial increase followed by a decrease, with HcPEBP9/10/12 peaking at stage 2 and HcPEBP3/4 reaching their highest expression at stage 3. These findings suggest that HcPEBP1/9/10/12 may play roles in floral transition.

Expression analysis of the HcPEBP gene family suggests that HcPEBP11 may influence the timing of flowering. To elucidate HcPEBP11’s regulatory role in flowering, we overexpressed it in tobacco. PCR amplification confirmed the presence of HcPEBP11 in transgenic tobacco strains and the positive control, with consistent band positions, while the negative control displayed no bands ( Supplementary Figures S3A, B ). This indicates the successful integration of HcPEBP11 into the tobacco genome. Semi-quantitative PCR results revealed HcPEBP11 expression in all transgenic lines, absent in wild type (WT) ( Supplementary Figure S3D ), confirming its expression in the transgenic strains. Expression analysis showed that HcPEBP11 levels in the leaves and flower buds of transgenic strains L-1 and L-6 were significantly higher than in WT, with transgenic line L-8 exhibiting even higher levels in the flower bud ( Figures 8B, C ). Transgenic plants flowered earlier than WT, with bolting and flowering times occurring 9–11 and 8–12 days earlier, respectively ( Figure 8A ; Supplementary Table S8 ). Bioinformatic prediction and GFP assay validation indicated that HcPEBP11 proteins localize to the cytoplasm ( Supplementary Table S4 ; Supplementary Figure S4 ). These findings suggest that HcPEBP11 promotes early flowering in tobacco and may function as a cytoplasmic regulator involved in flowering time control.