To retrieve the AP2/ERF gene sequences, the A. trifoliata genome and protein sequences were retrieved under accession numbers PRJNA750300, SAMN20447974. The assembled chromosome-level genome of A. trifoliata was 726.85 Mb, consisting of a scaffold N50 of 42.49 Mb and a contig N50 of 2.29 Mb, with a repeat rate of 42% and 39,443 protein-coding genes annotated (Niu et al., 2024). The candidate AP2/ERF proteins were searched from the A. trifoliata genomic database by using the Hidden Markov Model (HMM) mapping (PF00847). The AP2/ERF superfamily genes and their corresponding protein sequences of Arabidopsis and rice were retrieved from PlantTFDB (https://planttfdb.gao-lab.org/) and HMMER3.0. (https://www.ebi.ac.uk/Tools/hmmer/home), respectively. The ERF protein sequences of Arabidopsis and rice were 128 and 139, and a BLAST search of the genome of A. trifoliata was performed. The obtained sequences were used to confirm the presence of the AP2 structural domain using Pfam, the National Center for Biotechnology Information, and the Conserved Structure Database. Using the SwissProt database, sequences with high E-values were selected for comparison. Finally, as a quality check, the sequences were analyzed through the Simple Modular Architecture Research Tool (SMART) (https://smart.embl.de/smart/change_mode.cgi) to confirm the presence of the AP2 domain. The physical and chemical characteristics, sequence length, molecular weight (MWs), theoretical isoelectric point (pI), and subcellular localization of AtrAP2/ERF were predicted using the online analysis tools ProtParam and pLoc-mPlant, respectively.

MEGA-X software aligned the AP2 domains of AtrAP2/ERF proteins based on conserved domain sequences. The Jalview program was used to manually correct the predicted amino acid sequences of the AP2 motifs. A maximum likelihood phylogenetic tree was created using MEGA7.0 and the online tool iTOL. AtERF and OsERF classification systems were used to classify identified AtrAP2/ERF genes into several groups. The NCBI CDD and MEME program was used to obtain the AP2/ERF conserved domains and motifs, respectively. The biological data toolkit TBtools was used to visualize the exon-intron arrangements, protein structures, motifs, and chromosomal locations of the AtrAP2/ERF genes (Chen C. et al., 2020).

The chromosomal locations of the AP2/ERF gene family members were extracted from the gff3 file of the A. trifoliata genome annotation, and a map of the chromosomal gene distribution was constructed with Circos software (Feng et al., 2022). Gene duplication and collinearity between A. trifoliata and other species (A. thaliana, Oryza sativa, and Aquilegia vulgaris) were examined by using MCscanX (Wang et al., 2012). The synonymous (Ks) and non-synonymous (Ka) substitutions in each duplicated AtrAP2/ERFs were determined by the KaKs calculator (version 2.0) (Wang et al., 2010).

The 2.0-kb sequence upstream of the promoter region was analyzed in the PlantCare database to obtain the cis-element of the AtrAP2/ERF gene. TBtools was used to examine the location and numbers of the cis-element classified into three main categories: stress-responsive, developmental-responsive, and phytohormone-responsive. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases was used to identify the potential AtrAP2/ERF target genes.

The A. trifoliata transcriptome data were used to examine the expression patterns of AtrAP2/ERF genes in different tissues and stages ( Supplementary Table S1 ). For tissue-specific RNA-Seq analysis, nine tissues including different fruit cracking period (non-cracked fruits, PS; initially cracked fruits, PM; totally cracked fruits, PL) (Niu et al., 2020), flesh stages (hard flesh: UR; pericarp begins to soften; however, the flesh is hard: HR; soft flesh: FR) (Niu et al., 2021), and seed development (August 20: T; September 4: U; September 19: I) (Zhong et al., 2022) were considered and the raw data were retrieved under the accession number PRJNA524995, PRJNA750300 and PRJNA79843. All the transcriptome sequencing data were obtained from the HiSeq 2500 sequencing system. The raw data were filtered to trim low-quality reads and discard transcripts less than 200 bp in length. The remaining clean reads were mapped to the reference A. trifoliata genome using HISAT2 or String Tie, to obtain unigenes. P-Unigene expression levels were calculated as fragments per kilobase of exon model per million mapped fragments (FPKM) using the Cufflinks software package, and Hochberg’s approach to adjust the p value. The differential expression of AtrAP2/ERF genes was visualized by creating heat maps using TBtools. Using the STRING 11.5 database (http://string-db.org), cluster analysis and functional network analysis were performed on the orthologous gene pairs between AtrERF and AtERF, and the interaction network was displayed (Damian et al., 2017). Spearman’s correlation was used to predict the key genes for AtrAP2/ERF based on the degree of association between the genes.

Plants with consistent growth size, tree age, and fruit development were used for the ethylene and 1-MCP treatments, and a minimum of 45 fruits per plant were chosen as the experimental unit. These plants were used 20 days before fruit harvesting. Ethylene (300 mg/L), 1-MCP (10 mg/L), abscisic acid (ABA) (0.5 mMol/L) and its inhibitor fluridone (0.5 mMol/L) were sprayed on September 12, 2021, with water as the control. On October 2, three replicates of the samples were collected for RT-qPCR analysis.

Total RNA was extracted and reverse transcribed into cDNA using Plant RNA Extraction Kit (Accurate Biotechnology Co., Ltd., Hunan, China) and Evo M-MLV RT, respectively, followed by RT-qPCR analysis using the SYBR^®qRT-PCR Kit (Accurate Biotechnology Co., Ltd., Hunan, China) on a multicolor Real-Time PCR Detection System (Bio-Rad Laboratories, Berkeley, CA, USA). In this study, the EF-1α gene was used as the internal control gene, which was detected by the de novo transcriptome sequencing of A. trifoliata (Yang et al., 2016). The gene primer sequences are listed in Supplementary Table S2 . The 2^-ΔΔCt formula was used to calculate the gene expression and the histograms were made using GraphPad Prism 8 software. The gene network correlation analysis of key AtrERF genes was displayed by Gephi 0.9.5 (Kauffman et al., 2014).

The full-length open reading frame sequence (ORF) of AtrERF001 (missing stop codon) containing the BamHI locus was inserted into the PBI21-CaMV35S-GFP vector to generate the PBI21-AtrERF001-GFP recombinant plasmid. The expression vector PBI21-AtrERF001-GFP after transformation of Agrobacterium tumefaciens receptor cells GV3101 was suspended to OD₆₀₀ = 0.8-1.0 after expanded culture with MES buffer (10 mM MgCl₂, 10 mM MES, pH 5.6; 150 μM acetosyringone), and after infesting the leaves of Nicotiana benthamiana for 36–48 h, then cut the tobacco leaf near the injection hole with scissors and place it on a slide. The fluorescent protein in tobacco cells was observed by laser confocal electron microscopy to determine the subcellular localization of the AtrERF001 gene.

The ORF sequence of AtrERF001 (missing stop codon) and the prey vector was constructed by ligating it to the pGADT7 vector using the EcoRI and BamHI restriction sites. The bait vector was constructed by cloning the promoter regions of AtrACO1 (500 bp) and AtrPE (708 bp) containing the AtrERF001 binding pattern (GCCGAC) into the pAbAi vector using SmaI and SalI restriction endonucleases. The recombinant bait-pAbAi plasmid was then digested using the BstBI endonuclease and transformed into the Y1H yeast strain. Y1HGold [bait/AbAi] positive strains were verified using colony PCR. The inhibitory concentrations of aureobasidin A (AbA) were evaluated by distributing gradient concentrations on SD/-Ura/AbA plates. The recombinant pGADT7-AtrERF001 plasmid was then transformed into the Y1HGold strain made from the bait recombinant vector and cultured on SD/-Leu/AbA plates. Protein-DNA interactions were determined based on the ability of transformed yeast cells to grow on SD/-Leu/AbA medium. The interaction of pGADT7–53 and p53-AbAi, pGADT7 and AbAi, were used as positive and negative controls, respectively. All transformations and screenings were performed at least three times. Primers used for this assay are listed in Supplementary Table S3 .

The full-length coding sequence of AtrERF001 was cloned and inserted into the pET-32A vector and transformed into Escherichia coli strain BL21 (DE3) (Novagen, Denmark). Protein expression was induced using 0.2 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) at 16°C for 20 h. The recombinant protein was purified with a Capturem His-Tagged Purification Miniprep Kit (Clontech, CA). The triple tandem copies of the coupled element 1 (AtrACO1) motif (FP1:AATTTGAGTCGGCATTTTTT, RP1:AAAAAATGCCGACTCAAATT) and AtrPE motif (FP2:TTCCAAGCCGACATGGCCGA, RP2: TCGGCCATGTCGGCTTGGAA) were labeled with biotin at the 5′ terminus (BioRun, Wuhan, China) and used as a probe, and the unlabeled same sequences was used for the competition assay. The specificity of binding was examined by competition with the unlabeled probes (10×, 50×). Electrophoretic mobility shift assay (EMSA) was performed using a Light Shift Chemiluminescent EMSA kit (Thermo Scientific, MA) according to the manufacturer’s instructions.

After amplifying the AtrERF001 coding region with cloning primers ( Supplementary Table S4 ), the PCR product was ligated into a cloning kit (TransGen Biotech, China). The ORF sequence of AtrERF001 was ligated into pBI121 and pTRV2 vectors that had been digested with BamHI and transformed into Agrobacterium tumefaciens receptor cell EHA105. The injection of AtrERF001 on A. trifoliata fruits was performed as previously described for cherry fruits (Qi et al., 2017). Strains carrying overexpressed (PBI121-AtrERF001) and repressed expression vectors (pTRV2-AtrERF001) were cultured and expanded on LB medium to an OD₆₀₀ of 1.0-1.5 at 28°C. Then bacterial solution was resuspended in 2-Morpholinoethanesulphonic acid (MES) infiltration buffer with a final OD₆₀₀ of 0.8-1.0. Suspensions containing pTRV1+pTRV2, pTRV1+pTRV2-AtrERF001, PBI121 and PBI121-AtrERF001 were infiltrated into A. trifoliata fruits at 130 days (September 20, 2022) after flowering using a syringe, respectively. The injected fruits were bagged to increase the humidity of the surrounding air. Three replications, each replicate injected with 9 fruits were used for the experiment from 8-year-old fruit trees in the same area.

Determination of pectin content was described as previous studies (He et al., 2020). One gram of powder sample was washed twice with 95% ethanol (v/v) in a boiling water bath for 30 min. After centrifugation, the precipitate was resuspended in 30 ml of distilled water and heated at 50°C for 30 min. The supernatant after centrifugation was used to determine the water-soluble pectin (WSP) content. The remaining precipitate was dissolved in 25 ml H₂SO₄ (0.5 M) and boiled for one hour in a boiling water bath. Acid-soluble pectin (ASP) content was measured in the supernatant after centrifugation. WSP and ASP were determined using the colorimetry of the carbazole–sulfuric acid technique. Hemicellulose and cellulose content were measured as described in a previous study (Sun et al., 2020). After centrifugation, cellulose and hemicellulose were separated with dilute hydrochloric acid (HCl). Firstly, the samples were dissolved in 20 mL of 2N HCl, and hemicelluloses were measured in the supernatant. Seventy-two percent H₂SO₄ was used to dissolve the remaining residue and incubated for 1 h. After filtration, the cellulose content of the supernatant was measured. Finally, the cellulose and hemicellulose contents were determined using the anthranilic sulfuric acid method.

Mid-sections of cotyledons (7–8 days after germination) of wild-type tomato plants (Lycopersicon esculentum Mill. cv. Micro-Tom) were used as material following previously described methods (Liu et al., 2020). After a total of three days of incubation, transformed leaf healing tissues were transferred to regeneration medium by culturing on Murashig and Skoog (MS) medium (MS+100 mg/l kanamycin+1.0 mg/l ZT+1.0 mg/l IAA+300 mg/l Timentin). After rooting, the regenerated plantlets were transferred to soil ( Supplementary Figure S1 ). RNA was isolated from the leaves of wild-type and transgenic plants for PCR validation, and six major transgenic lines were identified, and T-26 was confirmed to carry the inserted AtrERF001 transgene. The first-strand cDNA of different developmental stages of fruits was synthesized and used in RT-qPCR analysis.

Three biological replicates were collected for each sample and the data in the figures were plotted with mean ± standard deviation. Statistical analysis was carried out by ANOVA SPSS 21.0 software (Chicago, IL, USA), and significant differences at p < 0.05 were determined using Duncan’s multiple range tests. The asterisks represented extremely remarkable differences (**: p< 0.01; ***: p< 0.001) between the different samples.

Based on HMM, CDD and Pfam data, 131 AP2/ERF genes were identified in A. trifoliata according to the number of AP2 domains. These genes were renamed AtrAP201 to AtrAP221, AtrERF001 to AtrERF108, and RAV1 to RAV2 according to their corresponding positions on the chromosome. The further characterization of the protein length, molecular weight and isoelectric point (PI) reported that the AtrAP2/ERF proteins included 121–709 amino acids, with an MW of 13.58–79.65 kDa and a projected PI of 4.5–10.7. The AtrAP2/ERF TFs are predicted to be localized in the nucleus ( Supplementary Figure S2 ; Supplementary Table S5 ). Multiple sequence alignments showed that the conserved sites-Gly (G)-4, Arg (R)-6, Glu (E)-16, Trp (W)-28, Leu (L)-29, G-30, and Ala-38 are present in most of the 131 AP2/ERF family proteins ( Supplementary Figure S3 ). The AP2/ERF family can be divided into two main groups, group A (I, II, III, IV) and B (V, VI, VII, VIII, IX, X, XI-L) based on the phylogenetic tree created according to A. thaliana and O. sativa and ERF groups (Toshitsugu et al., 2006). Groups A and B contained 40 and 91 AtERFs, respectively, and each subfamily tended to have one or more homologs or orthologs in A. thaliana and O. sativa, respectively. The ERF groups for A. trifoliata were highly consistent with the phylogenetic tree, indicating its authenticity. However, no AtrERF genes belonging to the Vb-L subgroup were observed ( Supplementary Figure S4 ).

The gene structure of AtrERF family members was investigated by comparison with genomic DNA sequence to learn more about their exon-intron and evolutionary structures. A total of 102 genes contained coding region sequences (77.9%), 29 of which were interrupted by introns. While none of the genes in Group I or III had introns, all genes in Group VII have two introns. A total of 20 conserved motifs were identified in the AtrAP2/ERF family using the MEME online software. All 131 proteins, except ERF030 and ERF089, possessed motifs 1 and 2. Most AtrAP2/ERF proteins in the same subgroup share a common motif, such as motifs 1, 2, 3, 4, and 5 (groups VI–L), motif 20 (Group III), and motif 19 (Group IX) on the phylogenetic tree, indicating that these ERF proteins have a similar evolutionary relationship ( Supplementary Figure S5 ).

Gene expansion and the creation of novel functionalities greatly benefit gene replication. A total of 58 pairs of gene repeats were identified in A. trifoliata, which were divided into tandem repeats and segmental duplication. There were only 3 pairs of tandem repeats (TD, 5.2%), while the remaining 55 pairs were classified as segmental duplicated pairs (SD, 94.8%) ( Figure 1A ; Supplementary Table S6 ), suggesting that SD made a valuable contribution in the gene family evolution. Synonymous substitution rates (Ka/Ks) of gene pairs are useful tools for understanding evolutionary dynamics after the occurrence of gene duplication. Except for AtrERF047/049 (Ka/Ks = 1.66, > 1 correlates with positive selection), the Ka/Ks ratio for tandem and segmental duplicated gene pairs in A. trifoliata was less than 1 (Ka/Ks < 1 indicates purifying/negative selection) ( Supplementary Table S7 ), suggesting negative selection pressure. Comparative syntenic maps of the representative dicot and monocot plant species A. trifoliata, A. thaliana, Solanum lycopersicum, O. sativa and Zea mays indicated that all four species shared 41 collinear gene pairs ( Supplementary Table S8 ; Figure 1B ). A. trifoliata shared 91 orthologous gene pairs with monocotyledons (54/37) and 154 with dicotyledons (88/66). Furthermore, 95% (39/41) of the identified gene pairs between A. trifoliata and dicotyledons differed from those shared by A. trifoliata and monocotyledons, indicating that these ERF orthologous genes evolved after the divergence of dicots from monocots plants.

The cis-regulatory elements in promoter sequences were analyzed using PlantCARE software, leading to an understanding of the evolutionary and functional diversification of the AtrAP2/ERF genes. A sum of 16 cis-acting elements were identified from the promoter region of AtrAP2/ERF genes, which was divided into three groups: phytohormone-responsive (including TGA-element, abscisic acid-responsive: ABRE, ethylene-responsive: ERE, TGACG-motif, GARE-motif, DRE core: GCC box, DRE1:dehydration responsive element/C-repeat, P-box, and AuxRR-core), stress-responsive (including MBS, WUN-motif, G-Box, and W-box), and related to plant development (including the CAT-box and O2-site). Moreover, the AP2 domain of ERF can bind to multiple cis-elements present in the promoter of ethylene-responsive genes, such as GCC box and DRE/CRT, and activate or inhibit the expression of these genes to regulate fruit development and ripening. Most of the AtrAP2/ERF genes contain light-responsive (G-box), ABRE, and ERE elements, which were the most prevalent cis-acting elements in phytohormone-responsive plants ( Figure 2 , Supplementary Table S9 ), indicating that the AtrAP2/ERF genes play a significant role in light, stress, hormones, and plant development, as evidenced by the composition of their cis-acting elements.

The assembled A. trifoliata genome contained 28,411 genes with at least one ERE (ATTTTAAA) and one DRE (ACCGAGA/GCCGAC/TACCGACAT) in their putative promoters. Of these, 4499 genes had a CCGAC core motif, and 7836 genes had at least one DRE. Consequently, 4499 genes were selected for additional functional enrichment analyses ( Supplementary Table S10 ). A few target genes were enriched in stress response and polysaccharide metabolic pathways in the GO terms ( Supplementary Figure S6A ). Numerous target genes were enriched in environmental adaptability and plant hormone signal transduction according to the KEGG pathway analysis ( Supplementary Figure S6B ). Protein-protein interactions provide intuitive and rapid understanding of the gene function of family genes, and are also important for the regulatory network relationship between family proteins. To predict the molecular interactions between AtrAP2/ERF and other proteins, the PPI network of AtrAP2/ERF and cell wall related genes was constructed based on Arabidopsis homologous proteins ( Supplementary Figure S7 ). It was predicted that cell wall, and ethylene signaling pathway related genes interacted with AtrAP2/ERF genes in the networks. For example, AtrERF043 can interact with AtrNAC, AtrPL4, AtrPE3, AtrACO; AtrERF001 can interact with AtrPE4, AtrPL2, AtrCYP707A2 ( Supplementary Figure S7B ), which played roles in fruit ripening.

Transcriptome data from several developmental stages were used to study the expression patterns of the AtrAP2/ERF genes, with some genes preferentially expressed in the detected tissues. For example, five (23.8%), nine (42.9%), and seven (33.3%) of these were expressed (FPKM value >1) in the pericarp, pulp, and seed, respectively. Among the 131 AtrAP2/ERF genes, 11 were expressed in all nine samples evaluated (FPKM >0). Fourteen genes (AtrERF011/014/019/027/028/031/035/063/074/076/077/093/101/102; AtrAP203/04/13) identified in the seed samples revealed significant transcript abundance during fruit ripening. The highest transcript abundance (FPKM >2 in at least one of the other tissues) was seen in three genes (AtrERF005/020/048; AtrAP203/04/13) in the initial crack stage (PM), ten genes in the complete cracking stage (PL) (AtrERF001/009/015/018/034/043/054/068/082/090; AtrAP214), five genes in the HR pulp stage (AtrERF019/020/027/031/035), and three genes in the mature pulp stage (FR) (AtrERF005/079/080; AtrAP213) ( Supplementary Figure S8 , Supplementary Table S1 ). Our study found that 14 AtrAP2/ERF genes were significantly upregulated at different developmental stages, and the highest expression levels were observed during the mature period.

Analysis of cis-acting elements showed that abscisic and ethylene response elements were the most prevalent cis-acting elements in AtrAP2/ERFs. To verify the effect of different hormone treatments on the ripening in A. trifoliata fruit, ethylene and its inhibitor 1-MCP, abscisic acid and its inhibitor fludioxonil were treated on A. trifoliata fruits 20 days before fruit ripening. Ethylene treatment (a) led to early fruit ripening, whereas 1-MCP treatment (c) delayed it. Abscisic acid (d) and its inhibitors (f) had little effect on fruit ripening ( Figure 3A ). The changes of fruit hardness and soluble solids content under different treatments were further analyzed. It was found that the hardness of ethylene-treated fruits was lower than that of the control group, while the soluble solids content increased significantly ( Figures 3B, C ). It indicated that ethylene-based treatment could accelerate fruit ripening and did not affect the edible taste of the fruit, while the results of 1-MCP treatment were just the opposite.

RT-qPCR analysis of the 14 AtrAP2/ERF TFs and 14 predicted target genes was conducted to investigate their expression patterns under ethylene and 1-MCP treatments. The results showed that the expression of AP2 genes (AP2/03/04/13/14) was upregulated under both ethylene and 1-MCP treatments compared to the control, indicating that the different treatments did not have much effect on the expression of AP2 genes. Most of the ERF genes (AtrERF001/005/009/020/043/054/068/082/090) showed higher expression under ethylene treatment while, they were significantly down-regulated under 1-MCP treatment compared with the control fruit ( Figure 4A ). Moreover, predicted target genes (cell wall modification-related genes), including AtrPGs, AtrPEs, AtrGALs, and AtrXTHs, were significantly increased under ethylene treatment, but decreased under 1-MCP treatment ( Figure 4B ).

The co-expression regulatory network was constructed by gene expression profiles under ethylene and 1-MCP stress to further elucidate the interaction between AP2/ERF transcription factor members and cell wall-related genes. AtrERF068, AtrERF001, AtrPE2, AtrACO, and AtrNCED were the most associated genes, which were concentrated in fruits under ethylene treatment ( Figure 5 ). Among these AtrERF genes, one ERF gene (AtrERF001) showed higher expression levels in full maturity and demonstrated a strong association with the cell wall-related genes. Our future work will involve cloning and isolating the homologous gene for AtrERF001 from A. trifoliata to confirm its role in fruit ripening.

In order to study the protein properties of ERF transcription factors in A. trifoliata, subcellular localization and yeast one-hybrid (Y1H) were conducted to elucidate the function of AtrERF001 in vivo and in vitro. The AtrERF001 ORF sequence was cloned into the PBI121 vector without termination codons. The fluorescence signals from the AtrERF001 fusion protein were found to be exclusively localized in the nucleus following the transient expression of these constructs in tobacco leaves, suggesting that AtrERF001 is a nuclear protein ( Figure 6A ).

To investigate the transcriptional activity of AtrERF001, the Y1H assay was used to verify the binding of AtrERF001 to the AtrACO and AtrPE promoters. Sequence analysis of the AtrACO and AtrPE promoters (2000 bp) revealed that there are AtrERF binding sites (GCCGAC). The CDS of AtrERF001 was cloned into the pGADT7 vector for effector construction, and the promoter sequences of AtrACO and AtrPE were cloned to create the reporter construct. The minimum inhibitory concentration of aureobasidin A (AbA) was determined by self-activation assay ( Figure 6B ). The pGADT7-AtrERF001 and pAbAi-AtrACO/AtrPE co-transformants grew well on SD/-Leu/AbA200 medium ( Figure 6C ), while the pGADT7+p53-AbAi negative control could not grow normally on the same medium, indicating that AtrERF001 specifically binds to the binding sites of the AtrACO and AtrPE promoters. To confirm the interaction of AtrERF001 with the AtrACO and AtrPE promoters, the AtrERF001 protein were purified and used for EMSA. However, the site mutation assay result showed that AtrERF001 could not bind to the GCC box (GCCGAC elements) in the promoters of AtrACO and AtrPE ( Figure 6D ). Therefore, AtrERF001 may indirectly regulate AtrACO and AtrPE by interacting with other TFs.

To verify the regulation of AtrERF001 on AtrACO and AtrPE expression in A. trifoliata, overexpression vector (PBI21-AtrERF001-GFP: OE-AtrERF001) and transient silencing vector (pTRV2-AtrERF001) were constructed. Then they were transiently transformed into immature A. trifoliata fruits, respectively ( Figure 7A ). Compared with the control fruits (PBI21-GFP and TRV vectors), the ventral suture line was more clearly visible in over-expression fruits (OE-AtrERF001), and fruit cracking began to occur after 14 days of infection. RNA was extracted from injected fruits (AtrERF001 gene overexpression and transient silencing fruits), and the AtrERF001 protein coding sequences carried by PBI121 and TRV vectors were amplified respectively to detect the results of transient transformation (primer sequence, Supplementary Table S4 ). It can be seen from Figure 7B that AtrERF001 gene overexpression fruits (red No. 2-3) and silencing fruits (red No. 7-8) had obvious AtrERF001 gene band size. The control fruits (empty vector, red number 5: PBI21-GFP; red No.6: TRV) did not detect the AtrERF001 gene band, indicating that the target gene AtrERF001 was accumulated in transiently transformed fruit. Moreover, compared with the control, the overexpressing AtrERF001 (OE-AtrERF001) fruit showed higher WSP content, whereas the ASP, cellulose, and hemicellulose contents decreased. Silencing of AtrERF001 in A. trifoliata significantly increased the ASP, cellulose, and hemicellulose contents ( Figure 8 ). This suggests that the degradation of cell wall materials, specifically the conversion of carbonate-soluble pectin to WSP, marks the initiation of fruit ripening (Jiang Y.L. et al., 2022).

RT-qPCR was further used to analyze the changes in gene expression in A. trifoliata fruits after overexpression and gene silencing of AtrERF001. The abundances of AtrERF001 transcripts increased significantly after AtrERF001 infiltration, and the levels of AtrACO and AtrPE transcription were also increased. Remarkably, the expression of the three genes increased approximately 30-fold, 5-fold, and 1-fold, respectively, in OE fruits than the control. Meanwhile, the expression of these three genes was significantly reduced in silenced fruits ( Figure 9 ). These results indicate that AtrERF001 may promote the maturation of A. trifoliata by indirectly affecting the expression of key genes AtrACO and AtrPE involved in ethylene metabolism and cell wall degradation.

Since the regeneration system and genetic transformation of A. trifoliata are more difficult, we chose tomato as the material for the next step of validation, which provides a good basis for the validation of AtrERF001 genes. To confirm the success of the transgene, PCR analysis of T0 tomato leaves was performed using specific primers of AtrERF001. By sequencing, the PCR product of the transgenic tomato leaves was consistent with the AtrERF001 gene at 753 bp, with 99% similarity and genetic distance ( Supplementary Figure S9 ). Transgenic tomato fruits entered the breaking stage earlier than control fruits. RT-qPCR was then used to detect changes in the expression of AtrERF001 at different developmental stages. During fruit development, the expression levels of AtrERF001 increased dramatically, peaking at the red-ripe stage in both the control and transgenic fruits. The expression of AtrERF001 in the OE red-ripe stage fruit increased approximately 1.5-fold compared to that in the control fruits. Moreover, the level of AtrACO and AtrPE transcription was increased significantly after AtrERF001 infiltration at the red-ripe stage ( Figure 10 ). These findings provide strong evidence for the crucial role played by AtrERF001 in modulating fruit ripening.