Genome-wide analysis of AP2/ERF gene family in Camellia oleifera and its potential roles in ethylene-induced fruit abscission

Abstract

Background

Camellia oleifera is a commercially important woody edible oil tree species in China. Yet, heavy premature fruit drop causes substantial yield losses and significantly limits the economic potential of the oil tea industry. Ethylene is well recognized as a promoter of organ abscission in plant, and the APETALA2/Ethylene Responsive Factor (AP2/ERF) family plays critical roles in response to ethylene. However, the AP2/ERF gene family in Camellia oleifera remains largely unexplored.

Result

Enhanced fruit abscission in Camellia oleifera was associated with increased accumulation of 1-aminocyclopropane-1-carboxylic acid (ACC), the immediate precursor of ethylene, in fruit pedicels. Compared with the high-yield Camellia oleifera cultivar ‘1712’, the low-yield cultivar ‘143’ exhibited significantly higher fruit abscission rates (e.g., 9.4% vs. 17.2% in August) and was more sensitive to ethylene. Ethephon treatment increased the cumulative fruit drop rates of cultivar ‘143’ to 14.1% at 3 days, while that of cultivar ‘1712’ was only 3.4%. Microscopic observations of the pedicel abscission zone revealed a reduction in lignified cells following abscission-promoting treatments in Camellia oleifera. Genome-wide analysis identified 143 CoAP2/ERF genes distributed across 15 chromosomes, which were classified into 16 subfamilies with 57 duplicated gene pairs. Transcriptomic data revealed tissue-specific expression patterns of CoAP2/ERF family. Among these genes, 79 CoAP2/ERFs were significantly differentially expressed in responses to ethephon. Notably, nine CoAP2/ERFs were highly expressed in pedicel, among which five down-regulated in response to abscission signal. Collectively, these five CoAP2/ERFs serve as the candidate regulators of ethylene-induced fruit abscission in Camellia oleifera, and they likely govern this process by modulating the development of abscission zone tissues.

Conclusions

Camellia oleifera shows high sensitivity to ethylene signaling, which is closely correlated with the premature abscission of young fruits. Our study systematically characterized the AP2/ERF gene family and identified five CoAP2/ERF genes as potential key regulators of fruit abscission. These findings deepen the understanding of abscission mechanisms in Camellia oleifera and lay a solid scientific foundation for further functional characterization of CoAP2/ERF gene family.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08731-3.

Introduction

Plant organ abscission is a precisely programmed process under the regulation of developmental signals and environmental cues [1], and ethylene is widely recognized as a key regulator in abscission [2]. For example, the application of ethephon accelerates the abscission of Litchi chinensis [3], Areca catechu [4] and Rosa hybrida [5]. In contrast, inhibited ethylene action significantly reduced fruit drop of Malus pumila [6]. Following elucidation of ethylene-mediated abscission mechanisms, accumulating evidence has established that diverse transcription factors are functionally implicated in the regulation of plant organ abscission [3, 7–11]. Among this key regulators, NAC (NAM, ATAF, and CUC transcription factor) and MYB (Myeloblastosis transcription factor) play key roles in ethylene-induced leaf and petal senescence [7, 9], while SlJA2L and PpNAC1/5 govern fruit ripening and maturation in tomato and peach, respectively [8, 12]. Notably, the functions of NAC and MYB are subject to regulation by AP2/ERF proteins. For instance, AP2/ERF-NAC regulatory modules have also been functionally delineated during Ficus caria fruit softening [13], and eggplant SmMYB113 directly modulates ethylene-dependent flower abscission, with SmERF38 potentiating its trans-activation activity at the SmACS1 promoter [14].

C. Oleifera (Camellia oleifera Abel), one of the most valuable woody oleiferous species, has been cultivated in China for over 2300 years [15]. The seeds oil extracted from C. oleifera is abundant in a variety of bioactive substances beneficial to human health, such as increasing increasing vascular permeability, protecting the central nervous system, and strengthening the immune system [16–19]. Fruit abscission is a common event in C. oleifera development that has been found associated with fertilization fail, nutrition shortage, unbalance hormones, and various abiotic/biotic stresses [20–24]. Heavy premature fruit drop results in severe yield losses, thus hampering the healthy development of the oil tea industry. Previous studies reported that ethylene plays an important role in inducing premature fruit drop of C. oleifera [25, 26]. While the molecular mechanisms underlying ethylene-mediated fruit abscission in C. oleifera remain largely uncharacterized. Existing transcriptomic analysis in C. oleifera have associated NAC with leaf senescence [27], and MYB with flavonoid biosynthesis [28] and lipid accumulation [29], however, the function of them in abscission remains poorly understood. Notably exogenous ethylene application is confirmed to accelerate C. oleifera fruit drop [25, 26], and AP2/ERF genes have been identified as pivotal hub genes linked to premature fruit abscission [30]. Collectively, these findings underscore the functional importance of AP2/ERF genes in this process. Systematic and in-depth functional characterization of AP2/ERF superfamily in C. oleifera remains largely unexplored.

As the most studied transcription factor in response to ethylene, APETALA2/Ethylene Responsive Factor (AP2/ERF) family is characterized by the presence of the AP2 structure (consisting of 60–70 amino acids) that is the crucial DNA binding domains [31]. Depending on the number of AP2 and other DNA-binding domains, AP2/ERF transcription factors can be divided into five subfamilies: APETALA2 (AP2), ethylene responsive factor (ERF), dehydration responsive element binding protein (DREB), related to ABI3 (Abscisic acid Insensitive 3)/VP1 (Viviparous 1) (Related to Abscisic acid Insensitive 3/Viviparous 1, RAV), and soloist [32]. Different AP2/ERF subfamilies exhibit distinct cellular functions and play a pivotal role in regulating diverse biological processes in plants [33]. Many ERF family proteins are involved in biological and metabolic regulation as well as in response to abiotic stress [34, 35]. Ethylene participates in stress response and developmental regulation in plants by activating specific transcription factors in the ERF subfamily. DREBs primarily regulate plant responses to abiotic stressors such as drought and cold stress [36, 37], while AP2 subfamily, featuring two tandemly arranged AP2 domains, have important roles in flower organ development [38], embryonic development [39], leaf development [40], and other organ architecture and development in plants [33, 41]. Members of RAV subfamily, containing both an AP2 and a B3 domain, however, is typically involved in hormone signaling and is a major regulator of abiotic stress responses [42, 43]. Critically, the ERF subfamily acts as the primary executor of ethylene-dependent transcriptional regulation, directly binding to GCC-box and DRE cis-elements in the promoters of downstream target genes to regulate cell wall development [10, 11].

Increasing experimental evidence indicates that AP2/ERFs are involved in the abscission and aging of plant organs in response to ethylene signal. In litchi, LcERF2 contributes to differential abscission rates and responses to ethylene by orchestrating cell wall metabolism, and thus pedicel growth [3]. An AP2/ERF TF (Transcription Factor) suppresses lignin deposition in the AZ (abscission zone) and thus enhances rice seed shattering [10]. However, RhERF1 and RhERF4 delay the abscission of Rosa hybrida petals by reducing the expression of β-GALACTOSIDASE 1 [11]. In general, AP2/ERF TFs regulate organ abscission mainly through modeling the cell wall metabolism, but different AP2/ERF members targeting different key cell wall metabolism genes. The AP2/ERF superfamily has been genomically characterized in many species. There are 163 gene members identified in Oryza sativa and 147 members published in Arabidopsis thaliana [32]. In fruit trees, 126 members of AP2/ERF were reported in Citrus reticulata [44], 149 in Vitis vinifera [45], 158 in Actinidia eriantha [46], 119 in Chinese jujube [47], and a total number of 208 AP2/ERF family members were identified from Citrus maxima [48].

Despite the well-documented role of AP2/ERF TFs in organ abscission, systematic characterization of the CoAP2/ERF family and its association with C. oleifera fruit abscission remains lacking. In this study, we identified AP2/ERF transcription factors in C. oleifera genome and comprehensively analyzed their bioinformation including basic physical and chemical properties, gene structure, conserved motif, chromosomal distribution, cis-element and predicted protein interaction networks of CoAP2/ERF proteins. Putative key AP2/ERF members associated with fruit drop regulation were targeted according to the spatiotemporal expression patterns and the expression in response to fruit drop inducing treatments. Our findings are expected to provide a foundation for further studies on fruit drop regulation mechanism in C. oleifera.

Materials and Methods

Plant materials and field trials

The trees of two C. oleifera cultivars exhibiting contrasting yield phenotypes were selected: the high-yield cultivar ‘1712’ (developed through a targeted breeding program by our research group) and the low-yield cultivar ‘143’. All trees were planted in an orchard located at Guangdong Academy of Forestry, Guangzhou, Guangdong Province, China. Both cultivars were grown in the same field under uniform soil conditions, irrigation, and fertilization management. Six healthy C. Oleifera trees with uniform growth vigor and no disease or pest damage were randomly selected, with each cultivar comprising three independent biological replicates (three trees for ‘1712’ and three trees for ‘143’). According to fruit developmental stages [25, 30], 10 fruiting shoots per tree were randomly tagged to calculate the fruit abscission rate in the peak fruit abscission periods (June, July and August). Concurrently, five additional fruiting shoots were harvested per tree, and intact fruits with attached pedicels were isolated for 1-aminocyclopropane-1-carboxylic acid (ACC) determination. All samples were prepared in three independent biological replicates, with each replicate derived from a distinct individual tree to ensure statistical reliability. For transcriptome profiling, two key developmental stages were specifically selected to capture the critical transcriptional transitions underlying fruit development and abscission. Briefly, young fruits were harvested in April, prior to the onset of significant fruit drop. Pedicels, leaves, peels, and seeds were collected in July, which represents the peak period of fruit abscission [25]. Each tissue sample was represented by three independent biological replicates. Statistical analysis of ACC content and fruit abscission rate was performed independently at each sampling month. Significant differences between the two cultivars were determined by Student’s t-test (P < 0.05).

Thirty bearing shoots from healthy C. oleifera cultivars ‘1712’ and ‘143’ trees were selected for ethephon spray (1 g/L) or girdling plus defoliation (C. oleifera cultivar ‘1712’) treatment on April. It was used water spraying or without treated shoot as control. Single tree replicate and three replicates were set up. Twenty bearing shoots were used to investigate the fruit drop rate, whereas the other ten shoots were used for sampling. The samples were divided into pedicels and fruits and then stored at −80 ℃ until ACC content measurement, RNA extraction or transcriptome sequencing.

Histological and SEM (Scanning Electron Microscopy) observations

The tissues of C. oleifera pedicels were cut into longitudinal sections prior to phloroglucinol lignin staining. Phloroglucinol-HCl reagent was prepared by mixing two volumes of 10% (w/v) phloroglucinol in 95% (v/v) ethanol with one volume of concentrated HCl. The cut tissues were placed in the prepared solution before being observed under a stereoscope (SV11; ZEISS) and photographed with a digital camera (D3200; Nikon).

Samples were fixed and dehydrated, and a critical-point drying was performed in an electron microscopy assay. Specimens were mounted on carbon tape, and sputter coated with gold palladium before being viewed with a 5 K accelerating voltage on an S-4800 field emission SEM (Hitachi, https://www.hitachi.com).

RNA isolation and gene expression analysis

Total RNA in each sample was extracted, using a Trizol reagent (Vazyme, Nanjing, China). First-strand cDNA synthesis was performed using 2 μg of total RNA with the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix Kit, according to the manufacturer’s instructions (TransGen, Beijing, China). CoAP2/ERFs transcript levels were determined by qRT-PCR (Quantitative Real-Time Polymerase Chain Reaction) analysis according to Yi et al. [3]. Each sample was quantified in three biological replicates and normalized to the crossing point (Cp) values of the CoCESA housekeeping gene [49]. The gene-specific primer pairs are listed in Table S1. TBtools (version 1.120) was used to construct a heat map for gene expression pattern [50].

ACC content determination

Fruit or pedicel samples from C. oleifera were thoroughly ground into powder in liquid nitrogen, and 100 mg of the powder was added to 1 mL of PBS (0.01 M, pH 7.4) solution for homogenate extraction. The mixture was then centrifuged at 12,000 g for 10 min at 4 ℃. The supernatant was used to quantify ACC content through an enzyme-linked immunosorbent assay (ELISA) with a double-antibody one-step reagent kit. In the micro-wells pre-coated with ACC antibody, the extracted supernatant or standard samples, and horseradish peroxidase-labeled detection antibody were sequentially added. The ACC concentration was determined by measuring the absorbance at a wavelength of 450 nm.

Identification of AP2/ERF gene family in C. oleifera

Genome-data of C. oleifera were obtained from National Genomics Data Center (https://ngdc.cncb.ac.cn/gwh/). The genome assembly accession numbers are GCA_025200525.1. The amino acid sequences of 141 AtAP2/ERF proteins in the A. thaliana were downloaded from the TAIR (The Arabidopsis Information Resource) database (https://www.arabidopsis.org/) to serve as query sequences for homology identification. Preliminary candidate genes were screened via BLAST (Basic Local Alignment Search Tool) (E-value ≤ 1e-5, identity ≥ 30%), and redundant sequences (≥ 99% identity) were removed. The protein sequence and gene sequence of the C. oleifera AP2/ERF gene family were obtained by analyzing AP2 domain (PF00847) using the Pfam (version 38.1, Protein Family Database) database (http://pfam.xfam.org) [51], SMART 6 (Simple Modular Architecture Research Tool) (http://smart.embl-heidelberg.de/) [52], and NCBI (National Center for Biotechnology Information) Conserved Domain Database (CDD) (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) [53]. All identified CoAP2/ERF candidates carried complete ORFs, and their physicochemical properties (molecular weight, MW and theoretical isoelectric point, pI), were analyzed via the ExPASy (http://www.expasy.org/) [54].

Phylogenetic analysis

The AP2/ERF genes families of C. oleifera and A. thaliana were compared using the ClustalW tool in MEGA 11.0.13 [55]. Consistent with the genomic research on C. Oleifera [56], A phylogenetic tree was constructed using the adjacency method with default parameters, with poisson correction model for amino acid substitution, and 1000 bootstrap replicates. Subsequently, the tree was visualized using the iToL software v7 (https://itol.embl.de/) [57]. CoAP2/ERF genes were classified based on a previously reported classification of the Arabidopsis thaliana AtAP2/ERF and Passiflora edulis PeAP2/ERF transcription factor families [58].

Conserved motif and gene structure analysis of CoAP2/ERFs

TBtools was used to analyze and visualize the phylogenetic and the exon–intron structures of CoAP2/ERFs. MEME (version 5.5.8, https://meme-suite.org/meme/) was used to analysis the conserved motif of CoAP2/ERF proteins [59]. The motif number of parameters was set to 10, while all other parameters were default settings as described in the reference [56]. The motif results in XML format obtained from MEME were visualized using TBtools (version 1.120) [50].

Colinearity and chromosomal distribution of CoAP2/ERF genes within and across species

Chromosomal distribution of CoAP2/ERF genes was derived from the C. oleifera genome and annotation files and was visualized using TBtools. Genome and annotation data for A. thaliana (Arabidopsis thaliana) were obtained from TAIR, whereas gene annotation files for O. sativa (Oryza sativa) were retrieved from National Genomics Data Center. To explore gene duplication events and perform Ka/ks value analysis, the ‘One Step MCScanX Wrapper’ module within TBtools was employed, and the results were visualizations with the Circos under default parameters [56].

Prediction of Cis-element regulatory and GO (Gene Ontology) annotation analysis of CoAP2/ERF genes

The region 2000 bp upstream of the CDS of CoAP2/ERFs was extracted using TBtools software based on the GFF3 (General Feature Format 3) file. PlantCARE online search tool was used to predict the cis-elements in the promoter regions (https://bioinformatics.psb. ugent.be/webtools/plantcare/html/). Finally, TBtools was used to visualize the cis-elements. GO annotations for CoERFs were obtained using TBtools.

Protein–protein interaction network of CoAP2/ERFs

All CoAP2/ERF protein sequences were submitted to STRING (version 11, http://string-db.org), and A. thaliana were selected as the reference species [60]. The results were imported into Cytoscape (version 3.6.1), and then used for network analysis [61]. BLAST was used to construct protein interaction networks based on the highest scoring homologues in A. thaliana.

Results

The critical role of ethylene in the fruit drop of C. oleifera

Our previous research found that ‘1712’ is a high-yield C. oleifera cultivar. In this study, we compared the physiological fruit drop rate of two camellia cultivars, high-yield ‘1712’ and low-yield ‘143’, during peak fruit drop period. The abscission rate both increased from June to August of ‘143’ and ‘1712’ cultivars, which of ‘1712’ was significantly lower than ‘143’ (Fig. 1A). Ethylene has previously been reported to enhance fruit drop in C. oleifera [25]. Here, we compared and characterized the ethylene sensitivity between the two cultivars. As shown in Fig. 1A-B and figure S3, there is a significant positive correlation between cumulative fruit drop rate and ACC (ethylene precursor) content in different C. oleifera cultivars from June to August. Specifically, the correlation coefficient (R) for cultivar ‘1712’ was 0.9897 (P < 0.01) and 0.9971 (P < 0.01) for cultivar ‘143’. But there was no difference in ACC content was detected between ‘1712’ and ‘143 cultivar (Fig. 1B). Ethephon (2-chloroethylphosphate acid) commonly as an ethylene-releasing molecule in plant tissues [62]. Treatment with ethephon significantly increased the cumulative fruit drop rate of ‘143’ at 3 days, whereas that of ‘1712’ cultivar remained unchanged until 5 days (Fig. 1C-D).

Fig. 1.

Fruit drop rate and sensitivity to ethylene of different C. oleifera cultivars. A and B The fruit drop rate and ACC content of different C. oleifera cultivars in June, July and August; Statistical analysis was performed independently at each sampling month using Student’s t-test.* P < 0.05 represent significant differences between the two cultivars. Values are mean ± SE (n = 3 biological replicates). C and D The cumulative fruit drop rate in response to ethephon of ‘143’ and ‘1712’. Statistical analysis was conducted independently at each time using Student’s t-test. Data are presented the mean ± SEs of three biological replicates. Asterisks indicate significant difference (*p < 0.05; ** < 0.01)

The effects of fruit drop inducing treatments on ACC production and pedicel structure

The girdling plus defoliation is an important treatment causing fruit abscission in other horticultural tree species according to the carbon and hormone interruption [63–65]. Hence, this study adopted the girdling plus defoliation technique to investigate its effect on the abscission of C. oleifera. Similarly, girdling plus defoliation induced severe fruit abscission. The fruit drop rate up to 18.7% at day 7 after treatment, while only 4.5% in the control (Fig. 2A). Interestingly significant increases of ACC content in the pedicel but not in the fruit were observed (Fig. 2B-C). To determine possible morphological changes in the pedicel caused by ethephon and girdling plus defoliation, the longitudinal sections of pedicels were observed and compared (Fig. 2D-E). Upon staining with phloroglucinol-HCl, lignified cells within the AZ were fewer in ‘143’ than those in ‘1712’. Moreover, lignified cells further decreased after girdling plus defoliation and ethephon treatment, compared to control. The anatomical observations of AZ cells were analyzed in response to girdling plus defoliation and ethephon treatments using SEM. In comparison with the controls, the AZ showed the weak adhesion between the adjacent cells, and contained many broken elements after girdling plus defoliation. The abscission fracture plane was divided into two distinct layers after sprayed ethephon. However, the adjacent cells of AZ in the control remained adhered together. The results indicate that ethylene may accelerate abscission by influencing the structure of the AZ and cell proliferation.

Fig. 2.

The fruit drop rate and the morphological changes of pedicels in response to abscission treatments. A The fruit drop rate in response to girdling plus defoliation treatment; (B) and (C) The ACC content in pedicel and fruit in response to girdling plus defoliation treatment. D Observations of lignified cells and AZ changes in two C. oleifera varieties, or after girdling plus defoliation treatment using longitudinal section. E Observations of lignified cells and AZ changes after ethephon treatment using longitudinal section and SEM. The black arrow indicates the position of AZ. Data are presented the mean ± SD of three biological replicates. Asterisks indicate significant difference (*p < 0.05; ** < 0.01)

Identification and characterization of AP2/ERFs in C. oleifera

A total of 180 CoAP2/ERF candidate genes were initially identified from C. oleifera genome based on sequence similarity to A. thaliana AP2/ERF genes. After eliminating redundant sequences and characterizing the AP2/ERF structural domain (PF00847), a total of 143 AP2/ERF genes were retained (Table S2).

Physiochemical analyses based on sequence information showed that the 143 CoAP2/ERF proteins ranged from 128 to 1168 amino acids in length, with a molecular weight of 14.23 kDa (LOK49_LG01G01852) to 125.7 kDa (LOK49_LG08G01853), and a theoretical isoelectric point of 4.4 (LOK49_LG11G02780) to 11.4 (LOK49_LG14G01129). Notably, 48 proteins exhibited PI (Isoelectric Point) ≥ 7, indicating a positive charge in acidic solutions. Through the prediction of protein hydrophilicity/hydrophobicity, the GEAVY values of 143 AP2/ERF members were all negative, indicating that 143 CoAP2/ERFs were hydrophilic proteins. Except for 5 of LOK49_LG01G03320, LOK49_LG04G00916, LOK49_ LG14G02219, LOK49_LG03G00790 and LOK49_LG04G00982, all other members showed instability coefficients of greater than 40, indicting high instability. Hence, the 143 CoAP2/ERFs showed significant variations in their physical and chemical properties (Table S2).

Phylogenetic relationships of the CoAP2/ERF proteins

To elucidate the evolutionary relationships within the CoAP2/ERF family, we performed phylogenetic analysis by aligning full-length protein sequences using 141 AP2/ERFs in A. thaliana and 143 AP2/ERFs in C. oleifera (Fig. 3). The AP2/ERF proteins were classified into five subfamilies. Notably, no CoAP2/ERF proteins were grouped into RAV subfamily. Most member were grouped into ERF (86 members) and DREB (38 members), but subfamilies contained only a single AP2 domain. Based on the differences in their structural domains, theses subfamilies were further subdivided into 13 subgroups, A1 to A6 and B1 to B7, respectively. The AP2 subfamily containing two AP2 domains had 17 members, and the Soloist domain that contained specialized structural domains had only 2 members.

Fig. 3.

Phylogenetic relationships of AP2/ERF proteins from A. thaliana and C. oleifera. A total of 141 AtAP2/ERFs from A. thaliana (Black) and 143 CoAP2/ERFs from C. oleifera (Red) were utilized to construct a phylogenetic tree in MEGA 11. Different colors represent different subfamilies and the names of the different subfamilies are labeled in the outermost circle

Gene structure and motif composition analysis of CoAP2/ERF

Divergence in genes structure can provide insight into gene family evolution. We constructed an evolutionary tree of the 143 CoAP2/ERFs and analyzed the conserved motifs and intron/exon structure (Fig. 4A). As shown in Fig. 4B, motif 1 was present in all CoAP2/ERFs genes which may serve as AP2 domain. The AP2 subfamily contained ten motifs and shared the conserved motifs of 1, 4, 6. The ERF subfamily contained motifs of 1, 2, 3, 5, 8, 10,while the DERB subfamily had motifs of 1, 2, 3, 5 and 8. A better understanding of gene expression pattern can be obtained by analyzing introns and extons. In C. oleifera, the AP2 subfamilies contained more exons than others (Fig. 4C). The DREB subfamily members typically lacked introns, except for LOK49_LG08G01853 and LOK49_LG01G01851.

Fig. 4.

Phylogenetic relationships, conserved motifs, and gene structure of CoAP2/ERF. A Neighbor-joining phylogenetic tree of CoAP2/ERF, and labeling of each subfamily name. B CoAP2/ERF conserved motif distribution. Different motifs were represented by boxes of different colors. C The structure of CoAP2/ERF genes is illustrated with exons represented by yellow boxes and introns by black lines

Duplication, synteny and K_a/K_s analysis of CoAP2/ERF genes

The 143 CoAP2/ERFs were randomly distributed across fifteen chromosomes, with 1 to 20 CoAP2/ERFs allocated to each one (Fig. S1). Chromosome 11 (chr11) contained only one CoAP2/ERFs, whereas chromosome 1 had the most (20), followed in turn by chr 14, chr 6, chr 9, chr2, chr7 and chr 5, which had 19, 14, 12, 11, 11 and 10, respectively. In contrast, chr13, chr 3 and chr 10 contained 7, 6, 4 CoAP2/ERFs, respectively. Additionally, LOK49_Contig189G00011 and LOK49_Contig 307G00002 were located on unanchored scaffolds.

As a major driving force in genome evolution, the gene duplication was investigated within the CoAP2/ERF gene family members in this study. In camellia, Overall 57 pairs of replicators were recorded on all the chromosomes (Table S3). Of the 15 chromosomes, chr1 had the most copies, with twenty pairs. There were at least one pair of replicators on chr10, chr11 and chr13 (Fig. 5A). CoAP2/ERFs were hypothesized to have undergone a certain scale of fragment replication events during evolutionary development.

Fig. 5.

Synteny (A) and collinearity (B) analysis of CoAP2/ERF genes. A Grey lines in panel A represented all syntenic blocks in the C. oleifera genome. While red lines and the associated genes indicated duplicated pairs. Each pair of genes connected by the red line are homozygous, and the chromosome number of each homologous gene is indicated in parentheses. B Grey lines in panel B represented all colinear blocks between the C. oleifera and the genomes of A. thaliana and O. sativa, while blue lines highlight collinear gene pairs involving CoAP2/ERF genes

Since the Ka/Ks ratio was a good indicator of the selection pressure occurring at the protein level, the TBtools software was used to estimate the values of Ks (synonymous) and Ka (non-synonymous). The Ka/Ks ratio usually reflects the selective pressure on genes. Specifically, Ka/Ks < 1, Ka/Ks = 1 and Ka/Ks > 1 indicates negative, neutral and positive selection, respectively. The Ka/Ks ratio of 57 duplicate was significantly lower than 1, ranging from 0.07 to 0.61 (Table S3). It indicated that all duplicate genes were under strong purification selection, thereby limiting functional differentiation.

In order to compare the evolutionary of history of AP2/ERFs in C. oleifera with in other species. the colinearity among C. oleifera, A. thaliana and O. sativa were investigated (Fig. 5B). It was showed that the highest colinearity was observed between C. oleifera and A. thaliana, followed by C. oleifera and O. sativa. There were the greatest number of orthologs on chromosome 1 with all other species. Except of the instances of duplication or loss, the AP2/ERF genes were highly conserved because of the maximum number of colinear orhologs in both C. oleifera and A. thaliana.

Analysis of cis-regulatory elements in the promoter of CoAP2/ERFs

To identify potential cis-regulatory elements, the 2000 bp sequence upstream of each CoAP2/ERF transcriptional start site was analyzed using PlantCARE. In addition to the common and core cis-elements, such as TATA and CAAT box, 28 cis-elements fall into various functional categories, including environment responses, hormones, and growth and development (Fig. 6, Table S4). Most cis-elements in the promoter regions of CoAP2/ERFs were the Box 4 motif (490) followed by the ABRE motif (379) and the G-box/G-Box motif (338). Additionally, 49 members had cis-elements involved in defense and stress responsiveness (TC-rich repeats); 102 had abscissic acid binding site (ABRE), 51 had auxin-responsive elements (TGA-element); 159 had cis-elements involved in MeJA responsiveness (TGACG-motif and CGTCA-motif); 57 and 52 had cis-acting elements involved in drought inducibility and low-temperature response, respectively. The gene of LOK49_LG02G00110 contained the most cis-elements, followed by LOK49_LG13G01053. These suggested that the AP2/ERF may play a crucial role in C. oleifera growth, development, and stress response that was consisted with the prior researches [66, 67].

Fig. 6.

Analysis of cis-acting elements in the promoter region of CoAP2/ERF genes. A The distribution of cis-acting elements. B The number of cis-acting elements

Gene ontology functional enrichment analysis of CoAP2/ERF proteins in C. oleifera

In order for a better understand of the functions of CoAP2/ERFs, the Gene Ontology (GO) annotation was created in this study, CoAP2/ERFs were classed into 20 functional groups with three main categories of biological processes, molecular functions and cellular components. Within the biological processes category, most CoAP2/ERFs were involved in cellular response to stimulus (GO:0009987), ethylene-activated signaling pathway (GO:0009873), and response to stress (GO:0006950). In the molecular functions category, CoAP2/ERFs were mainly involved in DNA-binding (GO:0003677,GO:0043565, GO:0000976), and nucleic acid binding (GO:0003676). Furthermore, the cellular components category were chiefly intracellular anatomical structure (GO:0005622) and nucleus (GO:0005634) (Fig. S2, Table S5).

Protein interactions of AP2/ERFs

After the protein interactions prdeiction by Athaliana orthologues of the 143 AP2/ERFs, their functions in C.oleifera was inferred from A. thaliana (Tale S6). The network consisted of 33 nods and 47 edges, with four core targets of LOK49_LG03G02652, LOK49_LG11G02302, LOK49_LG06G00760 and LOK49_LG13G01559 (Fig. 7). The functions of these three core target nodes were similar to those of ERF4, ERF5, ERF6 and ERF13 respectively. LOK49_LG14G01759 and LOK49_LG14G01755 were functionally similar to the wax induced 1–2 (AtWIN1-2).

Fig. 7.

The investigation of AP2/ERF protein–protein interactions involved constructing a network using STRING in C. oleifera. Gene constructed gray lines were functionally related. More gray lines indicate more genes interacting with them

Transcriptional profiling of CoAP2/ERF genes in various tissues

To gain insights into the potential function of CoAP2/ERFs, the expression pattern in five different tissues was analyzed to be visualized as a heatmap (Fig. 8). In camellia seed, the CoAP2/ERF genes of LOK49_LG05G03245, LOK49_LG04G00284 and LOK49_LG09G00727 all exhibited relatively high expression levels. LOK49_LG01G02185 and LOK49_LG03G02616 were predominant expressed in leaf. Several genes with LOK49_LG15G00643, LOK49_LG01G01202, LOK49_LG07G00646, LOK49_LG07G00647, LOK49_LG06G03320 and LOK49_LG01G03735 were showed elevated expression in the pericarp. In the young fruit and pedicel, LOK49_LG04G03161, LOK49_LG05G03245, LOK49_LG01G01204, LOK49_LG07G00647, LOK49_LG03G00790, LOK49_LG13G02731, LOK49_LG08G01090 and LOK49_LG05G03416 were showed highly transcript level, while LOK49_LG01G03122, LOK49_LG08G00829, LOK49_LG07G02640, LOK49_LG06G03320, LOK49_LG11G01430, and LOK49_LG11G02323 were specifically expressed in the pedicel.

Fig. 8.

Expression patterns of CoAP2/ERF genes across various tissues. Data represent the mean of three biological replicates. The color scale indicates the log2 fold change expression, with blue to red indicating low to high expression abundance

The expression of CoAP2/ERFs in response to ethephon and girdling plus defoliation treatments

AP2/ERF family genes, as ethylene-responsive genes, affect plant growth and development [39, 40]. To explore the expression level of AP2/ERFs response to ethephon treatment, the transcriptomic data of pedicel after ethephon treatment was analyzed in C. oleifera in this study. There were 79 CoAP2/ERFs shown the differential expression in response to ethylene (Fig. 9A), among which, 42 differential expressed genes were related to pectin metabolism (Table S7). To elucidate the molecular basis underlying the phenotypic divergence in fruit abscission, comparative transcriptome analysis was conducted on the fruit AZs of cultivars ‘1712’ and ‘143’ at the peak abscission stage. A total of 72 CoAP2/ERF family members were identified as significantly differentially expressed between the two cultivars, displaying expression patterns consistent with the observed differences in fruit abscission rate and ethylene sensitivity (Table S8). Subsequently, nine putative CoAP2/ERF genes with the highest expression level in pedicel were validated using qRT-PCR (Fig. 9B). Similarly with transcriptomic data these nine genes were detected highly expressed in the pedicel.

Fig. 9.

Transcriptome analysis of CoAP2/ERFs. A The heatmap illustrates the relative expression levels of CoAP2/ERF genes response to ethephon treatment. Colors denote the log2 fold change expression, with red indicating high expression, blue representing low expression. B Expression analysis of 9 pedicel specific expressed CoAP2/ERFs using qRT-PCR. C Expression of 9 CoAP2/ERFs in response to ethephon treatment. D Expression of 9 CoAP2/ERFs in response to girding plus defoliation. Data were presented the mean ± SD of three biological replicates. Asterisks indicate significant difference (*p < 0.05; ** < 0.01)

We further examined the expression pattern of these candidate genes in response to ethephon treatment and girdling plus defoliate treatment. Compared with the control group, six out of the nine genes displayed significantly down-regulated expression levels following ethephon spraying (Fig. 9C). Girdling plus defoliation has been shown to elevate ethylene production and accelerate fruit abscission [63], concomitantly, down-regulated expression of five out of the six CoAP2/ERF genes were observed under this treatment (Fig. 9D). Notably, four of these six candidate genes are members of the ANT family (Table 1; Table S6), for which loss-of-function mutants in A. thaliana exhibit reduced seed yield [68].

Table 1.

The basic analysis of ANT genes in C. oleifera

Genes	GeneBank number	Length (AA)	Top BLAST match (Accession)	Homology (%)
LOK49_LG08G00829	KAI8004156.1	618	Durio zibethinus (XP_022766436.1)	64.94
LOK49_LG05G03245	KAI8013017.1	662	Durio zibethinus (XP_022766436.1)	66.99
LOK49_LG04G03161	KAI8020653.1	668	Cornus florida (XP_059624038.1)	70.46
LOK49_LG01G03122	KAI8029282.1	623	Cornus florida (XP_059624038.1)	70.47
LOK49_LG04G00284	KAI8019962.1	521	Rhododendron vialii (XP_058227630.1)	100.00
LOK49_LG11G01430	KAI7994202.1	197	Camellia sinensis (XP_028091783.1)	97.97

Discussion

The role of the AP2/ERF gene family induced fruit abscission of C. oleifera by ethephon treatment was comprehensively analyzed in this study for the first time. Gene family identification, gene structure analysis, motif analysis, gene ontology annotation, expression pattern analysis on abscission treatment (spraying 1 g/L ethephon and girdling plus defoliation) and different tissues, were all discussed.

This study examining abscission inducing treatments have documented a positive correlation between pedicel ACC content (the immediate precursor of ethylene biosynthesis) and fruit abscission rate (Fig. S3). ACC content was used as a critical indicator of ethylene biosynthetic potential in plants. As shown in Fig. 1, no significant differences in ACC content were detected among the tested C. oleifera varieties, suggesting that intrinsic ethylene production capacity is largely conserved across these genotypes. Conversely, exogenous ethylene application elicited markedly distinct fruit abscission phenotypes [25]. These findings highlight that ethylene response, rather than ethylene synthesis, may serve as a key determinant of fruit abscission.

The ‘143’ of C.oleifera variety had a high rate of fruit drop with more sensitive to ethylene, which the stained lignin cells decreased. It was similar to the abscission treatment (Figs. 1–2). The effects of ethylene on fruit drop in the two cultivars showed consistent results with previous reports in litchi [3]. Observation using a scanning electron microscope, the changes in the morphology of the separated cells were also observed in our study. Compared with control, the intercellular adhesiveness of the cells in AZ was poorer after spray ethephon, and the cells were clearly separated, resulting in two distinct layers [3, 25, 69]. These results indicated that the ethylene may regulate fruit abscission by affecting the development of pedicel. Notably, the observed variation in fruit abscission between the two cultivars is closely correlated with their divergent ethylene sensitivity. Evidence derived from abscission- promoting treatments and structural observations of the pedicel corroborates the hypothesis that ethylene functions as a pivotal signaling molecule governing genotypic differences in C.oleifera by modulating pedicel AZ development, thereby ultimately specifying the fruit abscission phenotype.

As among the largest TF families in plants, AP2/ERF transcription factors interact with cis-elements of specific target genes to influence various biological processes, such as plant development, stress response, and defense [70–72]. With the availability of genome sequencing data of C. oleifera, 143 AP2/ERFs were identified that distributed across 15 chromosomes (Fig. 3, Table S2, Fig. S1). The number of AP2/ERF TFs in C. oleifera is less than that in A. thaliana (147) [32], Zanthoxylum bungeanum (146) [67] and Triticum aestivum (322) [73], yet exceeds that in Citrus reticulata (126) [44] and Chinese jujube (119) [47]. This variation highlights the diversity in the distribution of AP2/ERF among different plant species. such as in Passiflore edulis (91 members: with 53 ERF subfamily) [58] and in Zanthoxylum bungeanum (146 members: with 73 ERF subfamily) [67]. The ERF subfamilies accounted for a large proportion of the total AP2/ERF, indicating that the ERF subfamilies play a broad role in the development of C. oleifera (Figs. 3–4).

Gene duplication is crucial for expanding and diversifying gene pools, facilitating adaptation and evolution in organisms over time [74]. The main models of gene duplication in plants are tandem, segmental, or whole gene duplication, with selection by purging to produce genes with conserved structures [75, 76]. Within the AP2/ERF family, 57 colinear gene pairs were identified in C. oleifera (Fig. 5A). It suggested that these duplication events were derived from chromosome segmentation or large-scale duplication/triplication events. The Ka/Ks ratios of all identified colinear gene pairs were below 1 (Table S3), indicating that AP2/ERFs have undergone strong purifying selection during evolution. In addition, the AP2/ERF families in A. thaliana and O. sativa for comparative analysis were also constructed (Fig. 5B). The colinearity between C. oleifera and A. thaliana suggests that these two species have a particularly close evolutionary relationship.

TFs play the important role in inducing downstream functional gene expression and signal transduction [77, 78]. The cis-acting elements of promoter region regulate accurate initiation and transcription efficiency of gene transcription [79]. In C. oleifera, AP2/ERF gene family contained a large number of cis-acting elements related to hormonal, photosensitivity, biological and abiotic stress response (Fig. 6), which are speculated to play a role in growth and development, stress tolerance and hormone signaling. These observations are consistent with previous characterizations of the AP2/ERF family reported by Zhang et al. [35]. In the analysis of GO annotation analysis indicating that a number of AP2/ERFs were associated with DNA-binding factor activity. In terms of biological processes, AP2/ERFs were also mainly involved in the cellular metabolic and hormone signaling pathways, such as ethylene, cytokinin and jasmonic acid. In addition, some AP2/ERFs enriched in plant organ development, phloem and xylem histogenesis (Fig. S2). Based on the protein–protein interactions (PPIs) visualized in Fig. 7, we further hypothesize that AP2/ERF transcription factors participate in modulating plant cell wall architecture [80, 81]. As the PPI network was predicted using Arabidopsis orthologs, it may not fully recapitulate the authentic in vivo interactions in C.oleifera. Nevertheless, the predicted network provides a valuable preliminary framework for deciphering the functional associations of AP2/ERF genes, and rigorous experimental validation (yeast two-hybrid, co-immunoprecipitation) is ultimately required to verify these interactions.

Previous studies have demonstrated that the AP2/ERF proteins in specific tissues are related to the growth and development of plants [82, 83]. For example, LcERF2 plays a crucial role in the development of the fruit pedicel in litchi [3], and the expression level of RhERF1 was reduced in response to ethylene that regulate the abscission of rose petals [11]. Similar results were also observed in C. oleifera. Ethylene enhanced fruit abscission and regulated the development of pedicel in C. oleifera. A number of members showed relatively low expression level response to ethephon in this study (Fig. 8). There were all 143 AP2/ERF family members identified and 48 CoAP2/ERF genes expression level down-regulated in response to ethylene (Fig. 9A). Furthermore, to avoid the confounding effects of AP2/ERF genes involved in active oil biosynthesis during later stages [84], we specifically compared the transcriptomes of fruit pedicel tissues from two C. oleifera cultivars ‘1712’ and ‘143’ collected in July, the peak period of fruit abscission [25]. This analysis revealed that, compared with cultivar ‘143’, 38 CoAP2/ERF genes were significantly up-regulated and 34 were down-regulated in cultivar ‘1712’ (Table S8). These differentially expressed genes likely represent the core regulatory network driving the phenotypic divergence in abscission rates between the two cultivars Notably, nine CoAP2/ERF genes displayed high transcript level in the pedicel (Fig. 9B), among which five were transcriptionally repressed in response to both ethephon application and the girdling‑plus‑defoliation treatment (Fig. 9C). The consistent down-regulation under ethylene-promoting abscission conditions strongly implies that these five CoAP2/ERF genes may function as key regulators of fruit abscission in C. oleifera.

Accumulating evidence demonstrates that AP2/ERF genes modulate abscission by regulating cell wall metabolism in Litchi chinensis [3], O. sativa [10], and Solanum lycopersicum [85]. In Rosa hybrida, ethylene-induced down-regulation of RhERF1 accelerates pectin degradation, thereby promoting petal abscission [11]. In this study, 42 differentially expressed genes (DEGs) associated with pectin metabolism were identified through transcriptomic profiling of C. oleifera following ethylene treatment (Table S7). However, the role of AP2/ERF in cell wall metabolism and fruit abscission in C. oleifera remains to be further elucidated. However, woody perennial species exhibit long generation cycles and inherent recalcitrance to genetic transformation, which currently represent substantial technical barriers to the comprehensive functional validation of the candidate genes identified in this study. Therefore, future research will be devoted to establishing an efficient and stable genetic transformation system for C. oleifera.

AINTEGUMENTA (ANT) is a member of the AP2/ERF family related to cell proliferation and organ growth [86]. In Arabidopsis, ant mutants exhibit reduction in the number and size of floral and ovule, thereby, organ enlargement in plants overexpressing ANT [68]. Two Zmant1 ortholog mutants displayed small leaves and low yields than WT plants [87]. There were 4 ANT genes in the genome of C. oleifera (Table 1), and these genes showed homology with ANT of A. thaliana (Table S6). Notably, based on the GO functional and transcriptomic analysis, three ANT genes of LOK49_LG05G03245, LOK49_LG08G00829 and LOK49_LG04G03161, responsed to abscission signaling pathway (Figs. 8–9, Fig. S2, Table 1), and showed high expression in pedicel (Fig. 9B). Therefore, it was hypothesized that three ANT genes in C. oleifera, play a regulatory role in the development of pedicel and the shedding of fruit under ethylene stress. Besides, the expression level of LOK49_LG05G03245 was not only higher in pedicel, but also in young fruit and seed (Fig. 8). Whether LOK49_LG05G03245 functions in regulating organ cell meristematic capacity in C. oleifera remains unclear and requires further study.

Conclusion

In summary, ethylene-mediated fruit abscission is a pivotal physiological process directly governing the yield of C. oleifera; yet, the regulatory networks involving AP2/ERF transcription factors in this woody oil crop have remained largely elusive. In this study, we substantiated that ethylene modulates pedicel abscission zone (AZ) development to drive cultivar-specific abscission phenotypes and systematically identified 143 AP2/ERF family members within the C. oleifera genome. This work presents the first comprehensive characterization of the AP2/ERF gene family, integrating phylogenetic bioinformatics, multi-tissue expression profiling, and transcriptional dynamics under two distinct abscission-inducing regimes: exogenous ethephon application and girdling combined with defoliation.

Our transcriptomic analyses delineated two critical expression signatures associated with abscission divergence: 79 CoAP2/ERF genes exhibited significant responsiveness to ethylene induction, whereas 72 members displayed distinct expression trajectories between cultivars with contrasting abscission rates. Through a stringent multi-criteria screening strategy-encompassing AZ-preferential expression, ethylene sensitivity, strong phenotypic correlation, and homology to functionally validated abscission regulators in model species-we prioritized five core candidate genes: LOK49_LG05G03245, LOK49_LG08G00829, LOK49_LG04G03161, LOK49_LG04G00284, and LOK49_LG11G01430. Aligning with our phenotypic and physiological observations, we propose that these five genes serve as potential critical negative regulators in the ethylene signaling cascade controlling young fruit abscission in C. oleifera.

Collectively, this study elucidates the molecular framework of the ethylene-AP2/ERF regulatory module underlying young fruit abscission in Camellia oleifera, and provides valuable genetic targets and theoretical support for molecular breeding strategies aimed at enhancing fruit retention and yield stability in this economically important woody oil crop.

Abbreviations

ABI3

Abscisic Acid Insensitive 3

VP1

Viviparous 1

RAV

Related to Abscisic Acid Insensitive 3/Viviparous 1

TF

Transcription Factor

qRT-PCR

Quantitative Real-Time Polymerase Chain Reaction

TAIR

The Arabidopsis Information Resource

Pfam

Protein Family Database

NCBI

National Center for Biotechnology Information

SMART

Simple Modular Architecture Research Tool

Pe

Passiflora edulis

GFF3

General Feature Format version 3

GO

Gene Ontology

BLAST

Basic Local Alignment Search Tool

PI

Isoelectric Point

SEM

Scanning Electron Microscopy

Ka

Non-synonymous substitution rate

Ks

Synonymous substitution rate

Ka/Ks

Ratio of the non-synonymous to synonymous substitution rate

Authors’ contributions

Y.J.W. conceived and designed the research. Y.J.W. W.H.C. and W.J. wrote the manuscript. Z.Y.Z. provide material of *C. oleifera.* S.Y.Q. performed the bioinformatics analysis. H.J. performed the experiment. All authors have read and approved the manuscript.

Funding

This work was supported by Guangdong Basic and Applied Basic Research Foundation [2024A1515011013], Science and Technology Projects in Guangzhou [2024A04J4669].

Data availability

All data generated or analyzed in this work are including in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.