Following previous research, the Arabidopsis Information Resource (TAIR, https://www.arabidopsis.org/) was used to obtain the Arabidopsis thaliana histone methylation modification genes protein sequences. The whole genome data of apples (Malus domestica) was obtained from the Apple Genome Database (http://bioinformatics.cau.edu.cn/AppleMDO/) (Da et al., 2019). And the gene identifier (ID) can be found in the Supplementary Table S1 .

Candidate histone methylation modification genes (PRMTs and JMJs) in apple were identified using; two basic local alignment search tool (BLAST) methods implemented in TBtools (Chen et al., 2020). The Arabidopsis PRMTs and JMJs protein sequences were used as queries for BLAST analysis. Additionally, relevant hidden Markov models were obtained from the Pfam database (http://pfam.xfam.org/), and employed to search the apple protein sequence data using the HMMER 3.0 software with a stringent threshold of E ≤ 10–²⁰. After removing redundancy and duplicate sequences, a preliminary set of candidate sequences was determined. The intersection of the gene family candidate members was obtained by combining the results obtained from the above-mentioned methods. Subsequently, the conserved protein domains of the target gene family in apples were analyzed using the Web CD-Search tool on the NCBI website (https://www.ncbi.nlm.nih.gov/Structure/bwrb/bwrpsb.cgi/). This analysis aimed to determine whether the conserved domain related to the target gene family protein was present in each candidate sequence. Only the candidate sequences containing complete domains were retained for further analysis. In the case of those sequences that contained different or incomplete domains, their sequence integrity was assessed by submitting them to the SoftBerry website (http://linux1.softberry.com/) before subjecting them to the Batch Web CD-Search tool for comparing the similarity between their domains and those of Arabidopsis PRMTs and JMJs sequences.

The protein physicochemical properties of gene family members were calculated using the ExPASy online software ProtParam (https://web.expasy.org/protparam/). These properties encompass various parameters such as the number of amino acids, molecular weight, theoretical pI, instability index, aliphatic index, and grand average of hydropathicity. The chromosome location of each confirmed histone methylation modification gene was retrieved from the GFF3 file of apple genome, and then visualized on apple chromosomes using TBtools (Chen et al., 2020).

The amino acid sequences of PRMTs, and JMJs in Arabidopsis and apple were obtained individually from the TAIR and AppleMDO databases. These sequences were aligned using MUSCLE in MEGA 11 software. Then, the optimal model was determined using MEGA 11 software, and the phylogenetic tree was constructed using the maximum likelihood method(ML) (Tamura et al., 2021). The bootstrap consensus tree inferred from 1000 replicates is taken to represent the evolutionary history of the taxa analyzed (Felsenstein, 1985). The phylogenetic tree was visualized and optimized by TBtools.

The gene location analysis was performed using the Amazing Gene Location tool in TBtools, which mapped the PRMTs and JMJs to their respective chromosomes based on the information extracted from the GTF/GFF files. To visualize the synteny blocks within the apple genome and between the genomes of apple, Arabidopsis, rice and maize, the Dual Synteny Plotter tool in TBtools was employed. These synteny blocks were generated using the One Step MCScanX-Super Fast method. Rice and maize genome data were both downloaded from the EnsemblPlants database (https://plants.ensembl.org/index.html). The Simple Ka/Ks calculator (NG) in TBtools was utilized to calculate the synonymous (Ks) and non-synonymous (Ka) values for duplicated genes.

We employed the DupGen_finder tool (https://github.com/qiao-xin/DupGen_finder) to classify the types of duplicated genes. DupGen_finder_unique categorizes plant genome duplicated genes into five categories based on a specific algorithm: whole-genome duplication (WGD), proximal duplication (PD), transposed duplication (TRD), dispersed duplication (DSD), and tandem duplication (TD) (Qiao et al., 2019). Using Arabidopsis as an outgroup for apple, we followed the recommended workflow of DupGen_finder for our analysis and parameter configuration as previously described (Kenchanmane Raju et al., 2023).

The conserved features of PRMTs and JMJs sequences were analyzed and visualized based on motifs using the Multiple Em for Motif Elicitation (MEME) suite 5.4.1 (https://meme-suite.org/meme/tools/meme) (Bailey et al., 2015) and TBtools. The gene structure of PRMT and JMJ genes were assessed using Gene Structure Shower in TBtools, utilizing information from the apple genome GFF3 file. To identify the cis-elements in the promoters of PRMT and JMJ genes, PlantCare (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/) was employed (Lescot et al., 2002).

The expression pattern of PRMT and JMJ genes following abiotic, hormone and biotic stress treatment were obtained from SRA database in NCBI website or National Genomics Data Center. The study IDs were as follows: for abiotic stress treatment: salt treatment (SRP229388), cold stress (CRA002596) and drought treatment (SRP347250); for hormone treatment: 1-Methylcyclopropene (1-MCP) treatment (SRP334206), Gibberellin Acid 3 (GA₃) and 1-Naphthaleneacetic acid (NAA) treatment (SRP185711) and Indole-3-butyric acid (IBA) treatment (SRP330812); for biotic stress treatment: Alternaria alternata (SRP091754), Pythium ultimum (SRP048684), Fusarium solani (SRP239526) and ring rot (SRP153065). Kallisto, an RNA-seq quantification program were used to calculate the expression of gene (Bray et al., 2016). In this study, the gene expression was estimated using the fragment number per kilobase per million mapped exons (FPKM) and transcripts per kilobase of exon model per million mapped reads (TPM) method.

Apple (‘GL3’) tissue culture seedlings were transferred to MS medium plus 200 mM sodium chloride (NaCl), 10% polyethylene glycol (PEG) 6000. Leaves were sampled at 0, 6, 12 and 24h after treatment for RNA isolation. Total RNA was extracted from apple leaves using the hot borate method as previously described (Ma et al., 2006). For reverse transcription-PCR, first-strand cDNA was synthesized from 1 μg of total RNA using Reverse Transcriptase M-MLV (RNase H-) (Takara Biomedical Technology Co., Ltd., Shiga, Japan). qRT-PCR was conducted with Takara SYBR Premix Ex Taq II (Takara Biomedical Technology Co., Ltd.) using a Light Cycler 480 instrument (Roche, Basel, Switzerland). MdEF-1α were used as internal controls. All primers used are listed in Supplementary Table S8 .

To identify the PRMTs and JMJs gene families in the apple genome, a comprehensive investigation was carried out using the protein sequences of the corresponding genes from Arabidopsis as search queries. After eliminating redundant and duplicate sequences, we identified 14 and 32 genes in PRMT and JMJ families, respectively, in the apple genome ( Table 1 ). The predicted protein sequence lengths of the PRMT and JMJ families ranged from 232 (MdPRMT11) to 722 (MdPRMT3) and 187 (MdJMJ2) to 1843 (MdJMJ14) amino acids, respectively. The molecular weight (MW) of the PRMT and KDM families ranged from 25.45 (MdPRMT11) to 81.12 (MdPRMT3) and 21.43 (MdJMJ2) to 210 (MdJMJ14) kDa, respectively. The isoelectric point (pI) values for the PRMT and JMJ families varied from 25.45 (MdPRMT11) to 81.12 (MdPRMT3) and 5.05 (MdJMJ8) to 9.03 (MdJMJ28), respectively. The complete information about the genes, including gene locus ID, chromosomal position, coding sequence length, and protein sequence length, was shown in Table 1 . The nomenclature of the genes followed a sequential order based on their arrangement on the chromosome.

In Figure 1A , the chromosomal distribution of PRMTs, and JMJs in apple was shown. The analysis revealed that the distribution of these genes is irregular. Particularly, the MdJMJ genes were distributed on all chromosomes apart from chromosomes 2, 9, 11, and 17. The chromosomal distribution of MdPRMT genes was comparatively more restricted, limited to chromosomes 2, 4, 7, 8, 13, 15, and 16. Notably, chromosome 15 showed the most concentrated distribution with a total of 6 genes (MdPRMT8, MdPRMT9, MdPRMT10, MdPRMT11, MdPRMT12 and MdPRMT13).

Gene duplication events are a common occurrence across all species, playing a crucial role in generating novel functional genes and driving species evolution (Xu et al., 2012). In order to explore gene duplication events within the JMJ and PRMT gene families in apple, we conducted genome synteny analysis. The analysis revealed the presence of 10 potential duplicated genes in the MdJMJ gene family and 4 in the MdPRMT gene family ( Figure 1B ). This finding suggests that the expansion of the MdJMJ gene family is primarily driven by segmental duplication, as indicated by the high number of potential duplicated genes. Notably, there is a significantly gene cluster located on chromosome 15 for the MdPRMT gene family ( Figure 1A ).

Using Arabidopsis as an outgroup for apple, we followed the recommended workflow of DupGen_finder and found that after the divergence of the two species, the MdJMJ family contained 10 WGD-pairs, 1 TD-pairs (MdJMJ1-MdJMJ2), and 3 TRD-pairs (MdJMJ3-MdJMJ6, MdJMJ4-MdJMJ7, MdJMJ11-MdJMJ12). No DSD-pairs and PD-pairs were found in the MdJMJ family. The MdPRMT family had 4 WGD-pairs (MdPRMT7- MdPRMT14, MdPRMT5-MdPRMT8, MdPRMT2-MdPRMT13, MdPRMT1-MdPRMT12), 2 PD-pairs (MdPRMT11-MdPRMT9, MdPRMT9-MdPRMT10) and 1 TRD- pairs (MdPRMT6-MdPRMT13). No DSD-pairs and TD-pairs were found in the MdPRMT family. Detailed analytical results can be found in Supplementary Table S9 . The analysis outcomes indicate that the primary mode of evolution for the MdJMJ and MdPRMT gene families is whole-genome duplication.

To elucidate the evolutionary relationships of the PRMT and JMJ gene families, a maximum likelihood (ML) phylogenetic tree was constructed using the PRMT and JMJ proteins from apple and Arabidopsis ( Figure 2 ). The classification of the gene families was based on the established Arabidopsis JMJ gene families. The JMJ subfamily includes KDM5, KDM4, KDM3, KDM1, JMJC, and JMJD6, with 8, 6, 10, 4, 1, and 3 MdJMJ proteins, respectively ( Figure 2A ).

To gain deeper insights into the origin, evolutionary history, and functional characteristics of PRMT and JMJ genes, we performed a comparative synteny analysis between the genomes of apple, Arabidopsis, rice and maize ( Figure 3 ). The apple genome contained 19 JMJ genes with high synteny to the Arabidopsis genome, forming a total of 21 syntenic gene pairs, as shown in Figure 3A . These syntenic gene pairs include MdJMJ10-AtJMJ16, MdJMJ11-AtJMJ17, MdJMJ11-AtJMJ18, MdJMJ14-AtJMJ19, MdJMJ15-AtJMJ20, MdJMJ17-AtJMJ21, MdJMJ19-AtJMJ22, MdJMJ21-AtJMJ23, MdJMJ22-AtJMJ24, MdJMJ23-AtJMJ25, MdJMJ23-AtJMJ26, MdJMJ24-AtJMJ27, MdJMJ25-AtJMJ28, MdJMJ27-AtJMJ29, MdJMJ28-AtJMJ30, MdJMJ29-AtJMJ31, MdJMJ3-AtJMJ32, MdJMJ31-AtJMJ33, MdJMJ4-AtJMJ34, MdJMJ5-AtJMJ35, and MdJMJ9-AtJMJ36 ( Figure 3A ). 7 syntenic gene pairs were found between apples and maize, only one ortholog pair (AtJMJ21-OsFBO14) were found in Arabidopsis and rice. In addition, there were 7 PRMT genes in the apple genome that showed synteny with the Arabidopsis genome, forming a total of 11 syntenic gene pairs. These syntenic gene pairs include MdPRMT13-AtPRMT1a, MdPRMT13-AtPRMT1b, MdPRMT14-AtPRMT4b, MdPRMT14-AtPRMT4a, MdPRMT2-AtPRMT1a, MdPRMT2-AtPRMT1b, MdPRMT3-AtPRMT7, MdPRMT5-AtPRMT5, MdPRMT7-AtPRMT4b, MdPRMT7-AtPRMT4a, and MdPRMT8-AtPRMT5 ( Figure 3B ). Five pairs of co-linear genes (ZmPRMT6-MdPRMT6, ZmPRMT3-MdPRMT7, ZmPRMT3-MdPRMT14, ZmPRMT4-MdPRMT7, and ZmPRMT5-MdPRMT6) were identified between maize and apple, while no co-linear genes were found between Arabidopsis and rice. These results indicate a high conservation of JMJ and PRMT genes between the apple and Arabidopsis genomes.

For evolutionary studies, Ka (non-synonymous substitution) and Ks (synonymous substitution) values can be used to predict the selective pressure on duplicated genes (Zhao et al., 2016). When the Ka/Ks ratio falls below 1, it signifies that the genes have undergone purifying selection, indicating the absence of significant functional differentiation among these genes during the evolutionary process (Hurst, 2002). We utilized the Simple Ka/Ks CalculatorNG program in TBtools to analyze 14 pairs of collinear genes in MdJMJ and MdPRMT to further investigate the evolutionary relationships among these collinear genes. The results showed that all the collinear gene pairs had a Ka/Ks ratio less than 1, indicating that these genes have undergone purifying selection and have not experienced functional differentiation ( Table 2 ). However, it is important to note that a Ka/Ks ratio less than 1 does not imply that these collinear genes have not undergone functional differentiation or evolution. It simply suggests that, within the observed time scale, the functional differences among these genes have had a relatively minor impact on the evolutionary process.

A phylogenetic tree for the MdJMJ and MdPRMT gene families was constructed using the Maximum Likelihood method in MEGA11.0 software ( Figures 4A, E ). It is evident that the genes clustered together in both gene families exhibit similar conserved domains. MdJMJ4 and MdJMJ31 contained only motif 1, while MdJMJ17 lacked any motifs. Additionally, motif 15, motif 19, and motif 20 were unique to MdJMJ19, MdJMJ29, MdJMJ28, and MdJMJ15, respectively ( Figures 4B ). The motif analysis was further supported by conserved domain analysis further supported the similar domain structures within gene cluster, which showed MdJMJ19, MdJMJ29, MdJMJ28, and MdJMJ15 possessed the ZZ superfamily domain, PLN02529 superfamily domain, and PLN02328 superfamily domain, respectively ( Figures 4C ). These three conserved domains were specific to these four genes, consistent with the motif analysis. Gene structure analysis revealed significant variation in the structure of MdJMJ genes, with clustering genes showing similar structures ( Figure 4D ). The number of introns ranged from 0 to 32, with MdJMJ28 and MdJMJ15 lacking introns, while MdJMJ3 and MdJMJ16 had the highest number of 32 introns.

For the MdPRMT gene family, which had a smaller number of members, clustering of genes based on conserved motifs revealed similar motif structures within the clusters ( Figures 4F ). Conserved domain analysis clearly reveals that MdPRMT10, MdPRMT5, MdPRMT2, and MdPRMT13 genes only contain the AdoMet_MTases conserved domain ( Figures 4G ). MdPRMT4 and MdPRMT8 genes only possess the COG4076 superfamily conserved domain. The PRMT5_C and PRMT5 conserved domains are present only in MdPRMT7, MdPRMT12, and MdPRMT1 genes. All the members of the MdPRMT gene family have gene structures with introns, ranging from 6 (MdPRMT10 and MdPRMT5) to 22 (MdPRMT7 and MdPRMT12), indicating variations in gene structure within this gene family ( Figures 4H ).

The specific amino acid sequences of motifs of MdJMJ, MdPRMT and the detailed information of the MEME sites analysis were shown in the Supplementary Tables S2 and  S3 , respectively.

An analysis was conducted on the promoter fragments (-2000 bp) of all MdPRMT and MdJMJ genes ( Figures 5A, C ). A large number of cis-regulatory elements were identified in their promoter regions, mainly including light response, cell cycle, low temperature, circadian rhythm, stress defense, growth and development (endosperm, seed-specific, meristem, cell differentiation), hormone signaling (MeJA, ABA, GA, Auxin, and SA), anoxic specific inducibility, and some transcription factors (TF), including MYB and ATBP. Additionally, we also found a wound responsive element in each of the MdJMJ. By counting the number of cis-regulatory elements ( Figures 5B, D ), it was found that light-responsive cis-elements were present in the promoter regions of all members of the MdPRMT and MdJMJ. MeJA, ABRE, and ARE elements were also widely distributed. Therefore, it can be inferred that MdPRMTs and MdJMJs play important roles in apple growth and development as well as hormone responses. On the other hand, less abundant cis-regulatory elements should not be overlooked. For example, the low-temperature elements were present in a significant number of MdJMJs and MdPRMTs, with 18 and 6 genes involved, respectively. Stress defense elements were found in 11, and 3 genes, while MYB-drought elements were present in 20 and 8 genes of MdJMJ and MdPRMT, respectively. These genes may play crucial roles in stress tolerance in apple. Additionally, one gene in MdJMJs contained an Endosperm-specific negative expression element and a Wound responsive element, while none of the MdPRMTs genes contained these two elements.

Similarly, to compare the differences in cis-regulatory elements between apple and Arabidopsis, the same method was used to construct a map of cis-regulatory elements in Arabidopsis promoters and categorize them according to the phylogenetic tree ( Supplementary Figure S2 ). The results indicate that, on average, Arabidopsis has a fewer variety of cis-regulatory elements compared to apple. Like apple, Arabidopsis AtJMJ and AtPRMT genes also contain a large number of light-responsive and stress-response elements, which can provide some assistance for the function and evolution of the apple gene families.

In this study, the expression profiles of MdJMJs and MdPRMTs in different tissues and developmental stages of apple were analyzed using publicly available transcriptome data ( Figure 6 ). According to the results, the majority of histone modification-related genes showed relatively low expression levels during fruit maturation (SRP034165). It is worth noting that 11 histone modification-related genes (MdPRMT1, MdPRMT2, MdPRMT4, MdPRMT5, MdPRMT6, MdPRMT7, MdPRMT12, MdPRMT13, MdPRMT14, MdJMJ6, and MdJMJ19) exhibited higher expression levels in the early stage of fruit maturation, and their expression levels decreased to varying degrees as the fruit matured. Among them, the MdPRMT gene family had the largest number of genes, with approximately 64% of the genes showing decreased expression levels during fruit maturation. To delineate the expression patterns of histone modification-related genes in major organs of apple, transcriptome data from seedling and tree shoot apex, dormant buds and bud break, and different parts of apple flowers were analyzed. In shoot apex tissues at different stages, members of each gene family exhibited different expression patterns (SRP050139). For example, most members of the MdPRMT gene family showed varying degrees of down-regulation during the transition from the seedling stage to the mature stage, which was consistent with the down-regulated genes during fruit maturation. This suggests that the MdPRMTs may play an important role in the growth and development of apple. Furthermore, during apple plant development, 6 down-regulated genes were identified in the MdJMJs. Thus, it can be inferred that they collectively play a role in apple plant development.

Dormant buds play a crucial role in the growth cycle of plants and serve as a survival strategy for apple trees under adverse conditions such as winter or drought (Rohde and Bhalerao, 2007). They are also an important means of reproduction for apple trees. We analyzed the expression patterns of histone modification-related genes in apple dormant buds during dormancy and bud break (SPR099578). The results showed that MdPRMTs and MdJMJs had 5 and 6 genes, respectively, with higher expression levels. Among them, MdPRMT7, MdJMJ3 and MdJMJ25 showed relatively minor down-regulation, while the majority of genes exhibited lower expression levels or irregular trends.

Flowers are important reproductive organs of apple. We analyzed the transcriptome data from different parts of apple flowers. The results showed that histone modification-related genes had significantly higher average expression levels in pollen compared to other parts. MdPRMTs, and MdJMJs had 1 (MdPRMT8) and 5 (MdJMJ3, MdJMJ13, MdJMJ21, MdJMJ23, MdJMJ32) genes with significantly higher expression levels in pollen, respectively. Furthermore, MdJMJ19 was expressed in all parts of the flower, with significantly higher expression levels in petals and sepals compared to other parts. Finally, it is noteworthy that MdJMJ2 showed either no expression or relatively low expression levels in fruit development, shoot apex, bud, different parts of flowers, and young leaves.

In East Asia, Alternaria alternata apple pathotype (AAAP) is one of the main pathogens causing apple Alternaria blotch disease, significantly affecting the growth and development of apple trees and fruit yield (Abe et al., 2010). Some MdPRMTs and MdJMJs showed decreased expression levels after infection, such as MdJMJ6, MdJMJ15, and MdJMJ28. Some genes showed an upregulation, such as MdPRMT1, MdPRMT12 and MdJMJ25. In addition, certain genes demonstrated fluctuating expression levels, such as MdPRMT14 and MdPRMT7, which increased in expression level from 0 to 18 hours after infection and then decreased below the expression level at 0 hours of infection. On the other hand, MdJMJ27 and MdJMJ3 showed a decrease in expression level at the early stage of infection and then an increase in expression level after 18 hours of infection. It is worth noting that the expression level of MdJMJ13 was significantly higher at 18 and 72 hours after infection compared to other time points.

Pythium ultimum is one of the main pathogens causing Apple Replant Disease (ARD), which leads to growth inhibition and even death of apple seedlings when planted in orchard soil previously used for apple (or closely related species) cultivation (Zhu and Saltzgiver, 2020). The expression levels of MdJMJ and MdPRMT genes generally do not show significant changes after infection. However, it should be noted that MdJMJ13 expression is higher than the control after 4 hours of infection, while MdJMJ18 expression is lower than the control.

ARD causes severe growth and developmental obstacles in apple trees, and previous studies have determined Fusarium solani as the main pathogen causing ARD (Liu et al., 2022). There was a significant downregulation of MdJMJ3, MdJMJ18, MdJMJ25, MdJMJ27, MdJMJ29 and MdPRMT8, while MdPRMT2 showed an upregulation after F.solani infection. These results suggest that the above genes may play a role in apple resistance against F. solani infection.

Apple fruit ring rot (FRR) is a highly destructive disease caused by Botryosphaeria dothidea, which severely affects the apple industry’s development in the Asian region (Shen et al., 2019). The results revealed that, regardless of susceptible or resistant varieties, the expression levels of MdPRMTs and MdJMJs showed no significant changes compared to the corresponding mock after Botryosphaeria dothidea infection ( Figure 7A ).

First, in terms of abiotic stress, we focused on salt stress (SRP229388) ( Figure 6B ). Transcriptional data from apple seedlings subjected to two types of salt stress, KCl and NaCl, showed that the expression levels of most MdPRMTs and MdJMJs were relatively low in the leaves. It is worth noting that MdJMJ6 showed a downregulation trend in the leaves under both types of salt stress. Conversely, MdJMJ3, MdJMJ13, MdJMJ19, and MdPRMT7 exhibited significant upregulation under both KCl and NaCl salt stress, with MdJMJ3 showing higher expression levels during the initial 1 hour of KCl salt stress, followed by a significant decrease after 6 hours, while the opposite trend was observed under NaCl salt stress. In the root system, MdJMJ13 displayed an obvious upregulation trend under both KCl and NaCl salt stress, with a significant increase in expression levels after 6 hours of NaCl stress, much higher than the expression levels under the same period of KCl stress. MdPRMT12 showed a downregulation trend, and MdJMJ19 exhibited an initial upregulation followed by a decrease in expression levels. A comparison between the leaf and root systems also revealed that the expression levels of MdJMJ13, MdPRMT2, MdPRMT13, and MdPRMT14 under both KCl and NaCl salt stress were significantly higher in the root system than in the leaf system, whereas the opposite was observed in MdJMJ6 and MdPRMT7. Neither MdJMJ16 nor MdJMJ2 showed expression.

In the northern region of China, low-temperature freeze injury is one of the common diseases in apple during winter and spring (Yang et al., 2011). We analyzed the RNA-seq transcriptomes of barks (epidermis, phloem, and cambium) from one-year-old branches of two apple cultivars, ‘Golden Delicious’ (non-cold-tolerant varieties) and ‘Jinhong’ (cold-tolerant varieties) under chilling and freezing treatments (CRA002596) ( Figure 7B ). The results showed that the expression level of MdPRMT8 was significantly higher in ‘Golden Delicious’ than in ‘Jinhong,’ and with the continuous decrease in temperature, its expression level showed a significant downregulation. MdPRMT9, MdPRMT10, MdPRMT11, and MdJMJ23 had higher average expression levels in ‘Golden Delicious’ than in ‘Jinhong’, but their expression levels did not show an obvious pattern of change. It is worth noting that MdJMJ19 had very high expression levels at 4°C in both apple cultivars, and as the temperature decreased, its expression levels were significantly downregulated. In particular, MdJMJ19 in ‘Golden Delicious’ showed a stable expression level at -9°C and even lower temperatures, while in ‘Jinhong’, a slight upregulation was observed starting from -4°C, and although the expression levels decreased at -14°C and -29°C, the overall trend was still upregulation. In ‘Jinhong’, the expression level of MdJMJ15 was significantly higher at -4°C and lower temperatures than at 4°C, while in ‘Golden Delicious’, the change in expression levels was not significant. MdPRMT2 and MdPRMT13 had significantly higher expression levels at -4°C in ‘Jinhong’ than at 4°C, and their expression levels were downregulated at lower temperatures, which was not observed in ‘Golden Delicious’. MdJMJ18 showed an upregulation of expression levels between -4°C and -14°C in both cultivars, and its expression levels decreased at even lower temperatures, indicating that this gene may play a role in apple’s resistance to low temperature.

Drought is also one of the main types of stress faced during apple cultivation, especially in drought-prone areas of China (Berger et al., 2016). We analyzed the RNA-seq data of ‘GL3’ under drought treatment (SRP347250) ( Figure 7B ). The analysis results showed that compared to the control group, most MdPRMTs and MdJMJs showed varying degrees of upregulation under drought stress, with MdPRMT2, MdPRMT6, MdPRMT14, MdJMJ18, MdJMJ13, MdJMJ20, MdJMJ21, MdJMJ25 and MdJMJ27 exhibiting significant upregulation. Therefore, it can be inferred that these genes may be closely related to apple’s resistance to drought.

We selected five MdJMJ and MdPRMT genes that were significantly upregulated under drought stress for qRT-PCR analysis. The results showed a significant and varying upregulation in all five genes upon PEG (as drought) treatment ( Figure 8A ). When two MdJMJ genes (MdJMJ13 and MdJMJ32) were subjected to salt stress, both exhibited a decrease in expression after 6 h, followed by an increase at 12 h ( Figure 8B ). Similarly, the relative expression levels of the four genes under 4°C cold stress displayed a similar pattern, with an initial increase at 6 h and subsequent decline ( Figure 8C ).

Plant hormones play a crucial role in the growth and development of plants. 1-MCP (1-Methylcyclopropene) is commonly used for preserving ethylene-producing or ethylene-sensitive fruits, effectively delaying fruit senescence (Berger et al., 2016). We analyzed the RNA-seq data (SRP334206) of three apple fruit varieties (‘Golden Delicious’, ‘Granny Smith’, and ‘Fuji’) treated with 1-MCP ( Figure 9 ). The analysis revealed that MdPRMT2, MdPRMT7, MdPRMT12, MdPRMT13, and MdPRMT14 showed significantly upregulated expression in the 1-MCP treatment group compared to the control group in all three apple varieties. Additionally, the expression of MdPRMT8 in ‘Fuji’ was significantly lower than the other two varieties. Similarly, the expression of MdJMJ3, MdJMJ18, MdJMJ25, and MdJMJ27 was significantly higher in the 1-MCP treatment group, with MdJMJ3 showing the most noticeable upregulation at 60 d after 1-MCP treatment. This suggests that the aforementioned genes may be important for regulating the response of apple fruits to 1-MCP.

GA₃ and NAA are important plant hormones involved in various plant growth and development processes (Hennerty and Forshey, 1972; Pharis and King, 1985). We analyzed the RNA-seq data (SRP185711) of the main floral organs of ‘Honeycrisp’ apples after 18 and 132 days of NAA and GA₃ treatment. The analysis showed that the expression of MdJMJ6 in the hypanthium, ovary, and ovule was significantly lower after 18 days of NAA and GA₃ treatment compared to the hand-pollinated control (HP). Additionally, after 132 days of GA₃ treatment, all three organs showed lower expression levels of MdJMJ6. MdPRMT8 and MdJMJ19 exhibited significant upregulation in expression after 18 days of NAA and GA₃ treatment compared to HP. Furthermore, the expression levels of MdPRMTs and MdJMJs in all organs were significantly lower after 132 days of GA₃ treatment compared to 18 days ( Figure 10 ).

Finally, we also investigated the response of apple plants at different growth and development stages to Indole-3-butyric acid (IBA), which promotes adventitious root formation ( Supplementary Figure S1 ). The ability of apple plants to generate adventitious roots varies at different stages of growth and development. We analyzed the RNA-seq data (SRP330812) of adult and juvenile apple plants treated with IBA. Most of the MdPRMTs and MdJMJs showed no significant sensitivity to IBA treatment. However, it should be noted that in adult plants, the expression of MdJMJ19 was upregulated after 6 and 12 hours of IBA treatment compared to mock, followed by a decrease with longer treatment time. In juvenile plants, except for the 12-hour treatment, the expression of MdJMJ19 was higher than mock.