Genome-wide identification and expression analysis of the MORF gene family in celery reveals their potential role in chloroplast development Authors: Pei-Zhuo Liu, Ya-Hui Wang, Yue-Hua Sun, Yong-Ju Wei, Xu Sun, Meng-Yao Li, Guo-Fei Tan, Ai-Sheng Xiong Published: 2024 ABSTRACT AbstractChlorophyll is an important nutrient in celery and one of the main indexes of quality evaluation. RNA editing in chloroplasts is an important factor affecting chloroplast development and chlorophyll biosynthesis. Multisite organelle RNA editing factor (MORF) protein is a necessary regulator of chloroplast RNA editing. In this study, a total of 8 MORF genes in celery were identified, which were named AgMORF1a, AgMORF1b, AgMORF2a, AgMORF2b, AgMORF3, AgMORF7, AgMORF8 and AgMORF9 according to their subfamily classification. The physicochemical property, conserved motifs, cis-acting elements and protein interaction were predicted according to the sequences. The phylogenetic relationships and evolutionary selective pressure between MORF genes in celery and other Apiaceae plants were further analyzed. The results showed that AgMORF1b, AgMORF2a, AgMORF2b and AgMORF9 were predicted to be localized in chloroplasts. The evolution of MORF genes in 4 Apiaceae plants including celery, carrot, coriander and water dropwort was influenced by purify selection. Transcriptome data showed that the transcriptional levels of AgMORF2a, AgMORF2b, AgMORF8 and AgMORF9 were relatively higher among all MORF genes in petioles of celery, indicating their major role. RT-qPCR data showed that the expression levels of the above 4 genes were significantly higher in petioles of green celery than those of white celery. This study provided a basis for analyzing the effects of MORF proteins on chloroplast development of celery with different chlorophyll accumulation.Keywords: MORF, Celery, Chloroplast development, Sequence analysis, Expression profile Introduction Celery (Apium graveolens L.) belongs to Apiaceae family, which has rich nutritional, medicinal and economic value.1 Celery is recognized as a kind of healthy vegetable in the world as its rich nutrients such as chlorophyll, proteins, minerals and essential macro- and microelements for the human body. It can be eaten raw in salads or after stir-fry, serving as an essential aromatic vegetable in people’s recipes.2, 3 As one of the main sources of natural antioxidants, celery is rich in phenolic acids, flavonoids, ascorbic acid and other antioxidants, and has been used to treat a range of metabolic syndromes, cancer and diabetes. The unique aroma and pharmacological active substances give the high economic value of celery processed products, such as celery seed oil, which has the effect of lowering blood pressure, regulating gastrointestinal function, increasing appetite, favored by consumers around the world.4, 5 The petioles and leaves of celery are usually used as edible organs.6 The petioles of celery exhibit different colors such as white, yellow and green due to the discrepancy of chlorophyll accumulation.7 The synthesis of chlorophyll is affected by the regulation of enzymes related to chlorophyll biosynthesis and chloroplast development, and it was found that the abnormal development of chloroplasts may be the reason why white celery does not accumulate chlorophyll in petioles.8. Chloroplasts are the organelles of photosynthesis in higher plants, and the photosynthetic process depends on the precise regulation of gene expression in the nucleus and chloroplasts, which is crucial for plant growth and development.9 There are generally 30–40 RNA editing processes from cytosine (C) to uracil (U) in plant chloroplast transcripts.10 Abnormal RNA editing in chloroplasts will lead to dysplastic chloroplast development and decreased chlorophyll content.11 Current studies have shown that the chlorophyll content of rape (Brassica napus) leaf color mutants raised with the increase of RNA editing efficiency.12 RNA C-U editing efficiency and expression levels of chloroplast atpF, rps14, petL and ndhD genes were significantly reduced in yellow-green leaf color mutant of cabbage (Brassica oleracea var. capitata).13 In rice (Oryza sativa), the editing efficiency of many plastid-encoding genes in a white panicle2 (WP2) mutant was lower than that of the wild type.14 Most RNA editing events require the involvement of the multisite organelle RNA editing factor (MORF) to improve the efficiency.15 MORF gene family was first identified in Arabidopsis (Arabidopsis thaliana) and its members were usually localized in chloroplasts or mitochondria.16 Several studies have shown that MORF proteins can interact with other proteins to regulate RNA editing.17 Tomato (Solanum lycopersicum) MORF gene (SlRIP1b) was involved in the C to U editing of cytochrome C related genes to maintain mitochondrial function and regulate fruit development.15 ZmMORF1/8 interacted with PPR proteins (PCW1, EMP7, and bCCP1) to collaboratively complete RNA editing in maize (Zea mays) mitochondria.18 MORF proteins have a conserved residue (MORF-box), which can form homoplastids or heteroplasmons to interact with the selected PPR proteins, enhance its binding ability to the target RNA, and improve the editing efficiency of RNA.19 The loss of single MORF protein will abolish or reduce RNA editing at multiple editing sites. In particular, the editing efficiency of almost all chloroplast RNA editing sites in Arabidopsis morf2 or morf9 mutants was reduced.20 MORF proteins play an important regulatory role in chloroplast development and chlorophyll biosynthesis. The knockout of MORF2 and MORF9 genes in Arabidopsis resulted in a series of defective phenotypes such as albinism, yellow or mottling in Arabidopsis seedlings and impeded embryonic development, and the accumulation of tetrapyrrole metabolites was seriously affected.16 Rice (Oryza sativa) MORF9 played a key role in chloroplast ribosome biogenesis, chloroplast development and seedling survival.21 Chlorophyll is one of the important indexes to measure the quality of celery. Li et al. analyzed the differences of pigment accumulation in celery petioles of different colors through metabolomics and found that chlorophyll is the key pigment that determines the appearance color of celery petioles.7 Song et al. used Arabidopsis gene sequences as templates to identify 27 celery chlorophyll biosynthesis genes, and found that the expression levels of 16 genes encoding 11 enzymes in green celery was significantly higher than that of white and red cultivars with the help of transcriptome data.22 However, less research has been done on MORF proteins in celery. The MORF gene family members in Arabidopsis, rice and maize have been identified, which provides a good reference for the identification and phylogenetic analysis of MORF genes in celery.16, 18, 23 In this study, we identified all MORF gene family members in celery, and performed analysis on physicochemical property, conserved motifs, cis-acting elements, protein interaction, and specific expression profiles in different color petioles. Meanwhile, phylogenetic relationships and evolutionary selective pressure between MORF genes in celery and other proximate plants including carrot (Daucus carota), coriander (Coriandrum sativum) and water dropwort (Oenanthe sinense) were also analyzed. The results of this study may provide a basis for analyzing the molecular mechanism of celery chlorophyll biosynthesis regulation by MORF proteins. Materials and methods Genome data of different plants Genome data of celery, carrot and coriander were obtained through “TVIR database”(https://tvir.bio2db.com/)24, genome data of rice and maize were obtained through “Gramene database” (https://www.gramene.org/),25 tomato genome data were obtained through “Phytozome13” (https://phytozome-next.jgi.doe.gov/),26 Arabidopsis genome data were obtained through “Tair database” (https://www.arabidopsis.org/),27 and water dropwort genome data were obtained through “Water dropwort database” (https://apiaceae.njau.edu.cn/waterdropwortdb).28. Identification of MORF genes A total of 10 Arabidopsis MORF gene sequences were used as templates to blast the potential MORF genes in the genome data of celery, carrot, coriander, water dropwort and tomato. Blast results with an E-value < 10−5 were retained, and the conserved domains of retained sequences were compared by “Interpro database” (https://www.ebi.ac.uk/interpro/).29 The sequence that did not contain conserved domain MORF-box (IPR039206) was removed to finally identify the target genes in the 5 plants. Information of protein length was obtained by “GXF State” function of “TBtools” software. Information of molecular weight, isoelectric point, instability index, aliphatic index and grand average of hydropathicity was obtained by “Protein Parameter Calc” function.30 Subcellular localization was predicted through online tool “Cell-PLoc” (https://www.csbio.sjtu.edu.cn/bioinf/Cell-PLoc/).31 Construction of phylogenetic tree Muscle program was used to compare the protein sequences encoded by MORF genes of celery, carrot, coriander, water dropwort, tomato, rice, maize and Arabidopsis. “MEGA-X” software was applied to construct the phylogenetic tree of MORF genes through maximum likelihood method.32 The validation parameter (bootstrap) was set to 1000. Online tool “Circle Tree” was applied to modify and annotate the phylogenetic tree.33 Structural analysis of MORF genes Online tool “MEME” (https://meme-suite.org/) was used for the conserved motif analysis of the MORF protein sequences, and the motif number was set as 6.34 Coding DNA sequences (CDS) and promoter sequences of celery MORF genes and corresponding genome sequences were obtained by “Gtf/Gff3 Sequences Extract” function.30 Intron/exon characteristics of MORF genes were analyzed by online tool “GSDS 2.0” (https://gsds.gao-lab.org/).35 Cis-acting elements in promoter sequences were analyzed through online tool “PlantCARE” (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/).36 The figures for the analysis of promoter sequences were drawn by “TBtools” software.30 Protein interaction analysis of MORF genes Celery MORF protein sequences were used as templates, the prediction of protein interactions and the production of a map of the interaction network were carried out through online tool “STRING” (https://cn.string-db.org/).37 The top 20 proteins with the strongest interaction relationship were performed for gene ontology enrichment analysis and the analysis was completed through online tool “David” (https://david.ncifcrf.gov/home.jsp).38 The column diagram of gene ontology enrichment was drawn through online platform “SRplot”.39 Evolutionary analysis of MORF genes Chromosome locations and gene distributions were obtained from annotation information of celery, carrot, coriander, and water dropwort genome databases. Collinearity analysis and Ka (Nonsynonymous)/Ks (Synonymous) analysis were performed by “One Step MCScanx” and “Simple Ka/Ks Calculator” function. The figures analyzed above were drawn by “Gene Location Visualize”, “Advanced Circos”, “Multiple Synteny Plot” function of “TBtools” software.30 Plant materials As shown in Fig. 1, the plant materials were green celery ‘Jinnanshiqin’ (JNSQ), ‘Qingqin’ (QQ) and white celery ‘Saixue’ (SX), ‘Baiganyihao’ (BGYH) cultivated in the celery germplasm resource nursery of Suqian Research Institute of Nanjing Agricultural University. The celery seedlings were transplanted in the open air in October 2023, and the plant materials were collected at 90 days after sowing. The samples used for RNA isolation and real-time quantitative polymerase chain reaction (RT-qPCR) experiments were the petioles of different celery cultivars. Three petioles were collected from each celery plant, and a total of 9 petioles were collected from 3 celery seedlings of each cultivar as samples. Plant materials of different color celery cultivars JNSQ: “Jinnanshiqin”; QQ: “Qingqin”; SX: “Saixue”; BGYH: “Baiganyihao”. Expression analysis of MORF genes To analyze the expression of MORF genes in different celery cultivars, the published transcriptome data of petiole in cultivars of white celery ‘Saixue’ (SX), green celery ‘Jinnanshiqin’ (JNSQ), yellow celery ‘Best’, purple celery ‘Nanyunziqin’ (NYZQ), ‘Ziganyihao’ (ZGYH) and ‘Ziganqincai’ (ZGQC) were used.40 “Amazing Heatmap” function was used to draw the gene expression heatmap.30 Operation procedure and data processing method of RNA isolation and RT-qPCR experiments were carried out according to process of Huang et al, TUB-B was used as the celery reference gene to normalize the relative gene expression.41 The primers used for RT-qPCR experiments were designed through “PrimerPremier6” software and the primer sequences of MORF genes were shown in Table 1. The histograms of celery MORF gene expression were mapped using “Excel” software, “SPSS20” software was used to analyze the discrepancy of gene expression in different cultivars. Table 1Primer sequences of MORF genes used for RT-qPCR in celery petioles.Gene IDOligonucleotide sequences of forward primersOligonucleotide sequences of reverse primersAg10G00872.1GACACCACTTCGCCATAGCATCCTTACTAGCACCTTCTCAg2G01300.1ACTATCTCGCTCACTCACTCACAATCAACCAATGCTCATAg2G00008.1ATTGGTTGGTTGTTGTTGAGCTTCATCCTTGCTTCTTCCTAg7G01163.1CAGTTACAGAATTGCCACATGTTACGAGGTCTATCATTGCAg8G01732.1AAGAGATACCTGGAGTGTTGCTGCCTATCATCATACCTGTAAg4G01423.1AGGACTATGGAGGTGATAAGTATGTCGTTCAGGAGGAGGACCATCAg6G02830.1GAGGTGTCAGAGGAAGTATCAGGAGGTGGTCTATGTGTTAg1G00697.1ATCGCAGCAGAATTATGGTATGGTGTATTGTTGGTGAGATUB-BTGGTGGCACTGGATCTGGTATGGACTTTCGGAGGAGGGAAGACTGAA Primer sequences of MORF genes used for RT-qPCR in celery petioles. Result Screening and identification of MORF genes in celery Arabidopsis MORF gene sequences were used for blast to screen target genes, and the rice and maize MORF gene sequences were further used to help analyze gene distributions in different subfamilies. Based on the results of homologous alignment and domain analysis, 8 MORF genes were identified in celery. A total of 40 MORF genes in other 4 plants were identified, including 7 in carrot, 10 in coriander, 15 in water dropwort and 8 in tomato. The phylogenetic tree constructed with the identified genes and MORF genes of Arabidopsis, rice and maize was shown in Fig. 2. According to the branches of the phylogenetic tree, 72 MORF genes were divided into 7 subgroups. The 3 genes in the subgroup 1 did not belong to any subfamily, and the members were only found in tomato and coriander. According to the subfamily classification of Arabidopsis, rice and maize MORF genes, celery MORF genes were distributed in 6 subfamilies including MORF1, MORF2, MORF3, MORF7, MORF8 and MORF9. The celery MORF1 and MORF2 subfamilies both contained two members, while the rest had only one member. Phylogenetic tree of MORF proteins in 8 plants. The red stars mark the members of MORF protein family in celery. Physicochemical properties analysis of MORF genes in celery Table 2 showed the physicochemical properties of celery MORF genes. According to sequence analysis, the number of amino acids in celery MORF genes coding proteins ranged from 226 (AgMORF9) to 501 (AgMORF1b), their molecular weights ranged from 25.70121 (AgMORF9) to 52.37987 (AgMORF1b) kDa, their isoelectric points ranged from 7.75 (AgMORF1a) to 9.36 (AgMORF7 & AgMORF8), which were all basic proteins. The instability coefficients ranged from 39.71–63.12, with protein AgMORF1b being the most stable and protein AgMORF9 the most unstable. The aliphatic Index ranged from 41.28–75.43, and the grand average of hydropathicity of all members were negative (GRAVY ranged from −1.175 to −0.565), indicating that all MORF proteins have good hydrophilicity. According to subcellular prediction results, most MORF genes were localized in chloroplasts, including AgMORF1b, AgMORF2a, AgMORF2b and AgMORF9. Gene AgMORF1a, AgMORF3 and AgMORF7 were localized in nucleus. Gene AgMORF8 was the only gene that localized in mitochondrion. Table 2Analysis of physicochemical properties of MORF genes in celery.NameGene IDProtein length (aa)Molecular weight (kDa)Isoelectric pointInstability indexAliphatic indexGrand average of hydropathicitySubcellular localizationAgMORF1aAg6G02830.147550.161457.7547.9741.28−0.920NucleusAgMORF1bAg1G00697.150152.379878.1539.7145.79−0.833ChloroplastAgMORF2aAg2G01300.124327.343059.2854.6175.43−0.565ChloroplastAgMORF2bAg10G00872.123426.468079.2053.5770.85−0.656ChloroplastAgMORF3Ag8G01732.126729.906609.0658.1267.19−0.739NucleusAgMORF7Ag7G01163.125628.102849.3644.8541.84−1.175NucleusAgMORF8Ag2G00008.142946.175359.3659.7046.39−0.878MitochondrionAgMORF9Ag4G01423.122625.701219.2863.1265.53−0.704Chloroplast Analysis of physicochemical properties of MORF genes in celery. Structural analysis of MORF genes in celery The distributions of introns and exons in celery MORF gene sequence were shown in Fig. 3a. Gene AgMOF1a had the longest gene length (more than 4000 bp), while the gene length of AgMORF7 and AgMORF9 was the shortest, which is less than 2000 bp. The number of introns and exons in different MORF genes was similar, AgMORF2a, AgMORF2b, AgMORF3 and AgMORF9 contained 4 exons and 3 introns. The number of introns and exons in gene AgMORF1b and AgMORF8 were 5 and 4, AgMORF1a had the most exons and introns, AgMORF7 had only 3 exons and 2 introns. After summarizing the motifs distributions in MORF protein sequences and the sequence of each motif (Fig. 3b and 3c), it was found that the MORF-BOX domain contained 6 motifs, and all MORF protein sequences contained motif1, indicating that motif1 was an important part of the domain. Analysis of gene structures and conserved motifs of MORF genes in celery a: structures of MORF genes, blue boxes indicate 5′ or 3′ UTR regions, pink boxes indicate exons, and black lines indicate introns; b: conserved motifs of MORF protein sequences; c: Motif sequences. Sequence analysis of the promoter region of MORF genes in celery According to the statistics of the category and number of cis-acting elements in the promoter region of MORF genes, there were totally 15 kinds of cis-acting elements (Fig. 4). Among 8 celery MORF genes, the most abundant cis-acting element was the light response element. The promoter region of gene AgMORF2b, AgMORF7 and AgMORF9 contained 4 MeJA response elements. Gene AgMORF3, AgMORF7 and AgMORF9 contained more anaerobic induction elements. Gene AgMORF2b and AgMORF7 contained more low temperature response elements and MYB binding sites. Gene AgMORF9 contained the most abscisic acid response elements. Statistical diagram of cis-acting elements in promoter region of MORF genes in celery The numbers in the box represent the number of cis-acting elements in promoter sequences, darker colors represent a higher number of cis-acting elements. Collinearity analysis of MORF genes The figure of chromosome localization (Fig. 5) based on annotation information showed that a total of 8 genes were located in 7 chromosomes, AgMORF2a and AgMORF8 were both located in chromosome 2. Chromosomes 1, 4, 6, 7, 8 and 9 all contained only 1 MORF gene. The occurrence of gene duplication expanded the members of MORF gene family in celery. As shown in Fig. 6, gene duplication occurred in two pairs of MORF genes, which occurred between AgMORF1a and AgMORF1b, AgMORF2a and AgMORF2b. The collinearity of MORF genes in 4 plants with high homology in Apiaceae family was further analyzed, and the results were shown in Fig. 7. Seven MORF genes in celery had collinearity with those in carrot, water dropwort and coriander. No collinearity was found between gene AgMORF9 and MORFs in other plants. A total of 25 pairs of orthologous genes and 2 pairs of paralogous genes were found in 4 Apiaceae plants, and the results of Ka/Ks analysis were shown in Table 3. The Ka/Ks values of all gene pairs were less than 1 and all the Ka and Ks values were less than 1. Chromosome localization of MORF genes in celery The different colors in each chromosome indicate the gene density, the positions of the genes on the chromosome are marked below the gene names. Collinearity analysis of MORF genes in celery The different colors and bar graphs in the boxes indicate gene density in each chromosome, the red lines indicate two pairs of genes that have a collinear relationship. Collinearity analysis of MORF genes in 4 Apiaceae plants The blue lines indicate MORF genes that have a collinear relationship between 4 Apiaceae plants. Table 3Ka/Ks analysis of MORF genes in 4 Apiaceae plants.Gene pairKaKsKa/KsP/OAg2G00008.1- Oenanthe_sinensis0202550.10.0983770.4972290.197850OAg2G00008.1- Oenanthe_sinensis0085490.10.0972670.4625080.210303OAg2G00008.1- Cs02G00268.10.0571860.3013030.189795OAg10G00872.1- DCAR_0185070.0934860.8873230.105358OAg10G00872.1- Oenanthe_sinensis0212720.10.0910570.8239590.110511OAg10G00872.1- Oenanthe_sinensis0035190.10.0289380.2670430.108364OAg10G00872.1- Oenanthe_sinensis0046210.10.0289610.2912780.099428OAg10G00872.1- Oenanthe_sinensis0076200.10.0926910.8820280.105088OAg10G00872.1- Cs02G01410.10.0979610.8080930.121225OAg10G00872.1- Cs08G00576.10.0308280.2057890.149803OAg2G01300.1- DCAR_0125680.0915490.8304190.110245OAg2G01300.1- DCAR_0185070.0252120.2727390.092442OAg2G01300.1- Oenanthe_sinensis0212720.10.0292080.2563500.113937OAg2G01300.1- Oenanthe_sinensis0035190.10.0959840.7097820.135230OAg2G01300.1- Oenanthe_sinensis0046210.10.0938880.6641500.141366OAg2G01300.1- Oenanthe_sinensis0076200.10.0267460.2668050.100245OAg2G01300.1- Cs02G01410.10.0204320.2277950.089693OAg2G01300.1- Cs08G00576.10.1083930.7832570.138388OAg8G01732.1- Cs05G02124.10.0432500.2152810.200898OAg6G02830.1- DCAR_0018280.1379560.4257730.324012OAg6G02830.1- DCAR_0292620.2232930.7923800.281800OAg6G02830.1- Cs06G01731.10.2208080.7779540.283832OAg1G00697.1- DCAR_0018280.2356840.7747010.304226OAg1G00697.1- DCAR_0292620.1460810.3722820.392392OAg1G00697.1- Cs06G01731.10.0794760.2080570.381990OAg1G00697.1- Ag6G02830.10.2317710.7022870.330024PAg2G01300.1- Ag10G00872.10.0935060.8348920.111998PP: paralogs; O: orthologs. Ka/Ks analysis of MORF genes in 4 Apiaceae plants. P: paralogs; O: orthologs. Interaction analysis of MORF proteins in celery By comparing the homology of celery MORF proteins with Arabidopsis MORF proteins, it was found that 8 celery MORF proteins belonged to 6 subfamilies. The protein sequences in celery were used as templates to predict their potential interacting proteins and construct a protein–protein interaction network (Fig. 8). AgMORF1a/b, AgMORF8 and AgMORF9 all had possible direct interactions with each other, AgMORF2a/b may had direct interaction with MORF8 and MOR9, but indirect interaction with MORF1a/b. AgMORF3 and AgMORF7 only had indirect interaction with other celery MORFs, and both of them may interact with PPOX1 protein. The most common interacting proteins were PCMP proteins, which belong to pentatricopeptide repeat-containing proteins. In order to further understand the function of MORF proteins, gene ontology enrichment analysis was performed on MORF interacting proteins (Fig. 9). The most prominent biological processes of enrichment were RNA modification, mRNA processing and cytidine to uridine editing. Among the 20 interacting protein-coding genes, 13 were involved in RNA modification and 12 were involved in mRNA processing. The cellular component enrichment showed that all interacting proteins play roles in plastids, including chloroplasts and mitochondria. A total of 13 interacting proteins had the function of RNA binding, which was the most prominent molecular function, followed by zinc ion binding and mRNA binding. Protein-protein interaction network of MORF proteins in celery. Gene ontology enrichment of MORF interacting proteins BP: Biological process; CC: Cellular component; MF: Molecular function. Expression analysis of MORF genes in celery According to the transcriptome data of petioles in 6 celery cultivars tested at the rapid growth stage (60 days after sowing) and maturity stage (90 days after sowing), the expression analysis of 8 MORF genes were performed based on log2-transformed FPKM values and the results were shown in Fig. 10. The transcriptional levels of genes AgMORF2a, AgMORF2b, AgMORF8 and AgMORF9 were apparently higher than those of other MORF genes. The transcriptional levels of gene AgMORF7 in petioles of 6 celery cultivars were relatively low at both stages. The transcriptional levels of gene AgMORF2a, AgMORF2b and AgMORF9 at 60 DAS (days after sowing) were higher than those at 90 DAS, which might be related to the changes in the growth and development of celery. The transcriptional level of AgMORF9 at 60 DAS in petioles of ‘JNSQ’ was about 1.3 times than that of ‘Best’ and ‘NYZQ’, 1.6 times than that of ‘ZGQC’, 1.9 times than that of ‘SX’, and 4.1 times than that of ‘ZGYH’. The transcriptional levels of AgMORF2b in petioles of all purple celery, yellow celery and white celery cultivars were about half that of green celery ‘JNSQ’. Heatmap of MORF gene expression in petioles of 6 celery cultivars JNSQ: Jinnanshiqin; SX: Saixue; NYZQ: Nanyunziqin; ZGYH: Ziganyihao: ZGQC: Ziganqincai; 60 DAS: 60 days after sowing; 90 DAS: 90 days after sowing. RT-qPCR was used to further determine the differences in MORF gene expression between white and green celery petioles (Fig. 11). Significant differences were found in MORF gene expression between white and green celery petioles. The expression of gene AgMORF7 was not detected in 4 celery petioles and the expression levels of all MORF genes were very low in petioles of ‘SX’. The expression of MORF genes in petioles of ‘BGYH’ was significantly higher than that of ‘SX’, but was generally lower than that of green celery petioles. The expression levels of gene AgMORF1a, AgMORF1b, AgMORF2b, AgMORF8 and AgMORF9 in petioles of ‘JNSQ’ was significantly higher than those in white celery cultivars. Among them, the expression levels of gene AgMORF9 and AgMORF2b were the most different, the fold changes in expression levels of AgMORF9 and AgMORF2b between ‘JNSQ’ and ‘BGYH’ reached 4.1 and 3.0, while the fold changes in expression levels of AgMORF1a, AgMORF1b and AgMORF8 were about 1.9, 1.5 and 2.6. Except AgMORF1a and AgMORF8, there were significant differences in the expression levels of other MORF genes between the petioles of green celery ‘JNSQ’ and ‘QQ’. The expression levels of AgMORF2a and AgMORF3 in petioles of ‘QQ’ were significantly higher than those in ‘JNSQ’ and ‘BGYH’, and the fold changes reached 3.8 and 3.5. Relative expression of MORF genes in petioles of green and white celery cultivars SX: ‘Saixue’; BGYH: ‘Baiganyihao’; JNSQ: ‘Jinnanshiqin’; QQ: ‘Qingqin’. Discussion MORF protein is a universal RNA editing factor involved in about 80 % of editing sites in mitochondria and almost all editing sites in chloroplasts, so it is widely present in a variety of plants.18 The number of MORF gene family members in Arabidopsis, rice and maize were 10,7 and 6 respectively, while 8 MORF genes in celery were identified in this study. There was only one member in each Arabidopsis MORF gene subfamily, and two members in both rice MORF2/8 and maize MORF2 subfamilies. Due to gene replication events, there were both 2 members in MORF1 and MORF2 subfamilies in celery, while only one member was found in MORF3, MORF7, MORF8 and MORF9 subfamilies. The MORF2 subfamily had the largest number of genes in the phylogenetic tree, and most species had more than one member, indicating that gene replication events mainly occurred in the MORF2 subfamily. The most abundant cis-acting element in promoter region of celery MORF genes was the light response element, indicating that MORF genes may be closely related to light response. Currently, most MORF genes with known functions have been found to be located in mitochondria of chloroplasts, while 3 MORF genes in celery were predicted to be located in the nucleus, and the other 5 genes may be involved in the regulation of RNA editing in the organelles. Gene MORF2 and MORF9 in MORF gene family were widely believed to participate in RNA editing in chloroplasts.16, 19 The subcellular localization of gene AgMORF2 and AgMORF9 in celery was also predicted in chloroplasts, and their potential interacting proteins were also associated with RNA editing. Transcriptome data showed that 4 of the 8 MORF genes had higher transcriptional levels, suggesting that AgMORF2a, AgMORF2b, AgMORF8 and AgMORF9 may play a major role, and the transcriptional levels of AgMORF2a, AgMORF2b and AgMORF9 may relate to the growth and development of celery. In addition, transcriptome data and RT-qPCR results showed that the expression of most MORF genes were significantly higher in petioles of green celery than those in white celery. MORF genes can actively regulate the stress response. Chloroplast-preferred MORF proteins (NbMORF2b, NbMORF2c and NbMORF9) in tobacco (Nicotiana benthamiana) positively regulated plant immunity to pathogens.42 In kiwifruit (Actinidia chinensis), the MORF genes overcontrolled the degree of RNA editing to regulate plant pathogen interactions.20 In this study, the expression level of gene AgMORF7 was very low, and its promoter region contained more anaerobic induction elements, low temperature response elements and MYB binding sites, which suggested that it may respond to abiotic stress. MORF proteins have chaperone activity and usually interact with each other. Multiple MORF proteins synergistically affect the efficiency of RNA editing in chloroplasts and regulate structural genes in chlorophyll biosynthesis pathway.19 Previous studies have revealed that Arabidopsis AtMORF2 and AtMORF9 genes work together through chaperone activity to regulate RNA editing and chloroplast development.19 The expression of MORF genes in petioles of green celery was not completely consistent. The expression levels of gene AgMORF2b and AgMORF9 were higher in ‘JNSQ’, while the expression level of AgMORF2a was higher in ‘QQ’. Considering the potential interaction between AgMORF2 and AgMORF9, AgMORF2 and AgMORF9 may play regulatory functions together. ‘JNSQ’ was derived from cross breeding of western celery and eastern celery, ‘QQ’ was an eastern celery cultivar. It is speculated that the differential gene expression may because of the different genetic background. Chloroplasts are important organelles of plants, and the development of chloroplasts affects the growth and development, stress resistance, fruit quality and many other aspects.9 Understanding the self-regulation mechanism of chloroplast development is of great significance to improve quality and stress resistance. The range of cultivation of celery is wide, and celery with multiple petiole colors and different environmental adaptability has been cultivated under different genetic background. At present, the regulation mechanism of chloroplast development in different celery cultivars has been studied from the differences of metabolites and gene expressions, but there are few studies on the post-transcriptional modification stage in chloroplasts.7, 22 The results of this study may provide a basis for understanding the molecular mechanism of chlorophyll development regulated by MORF proteins and understanding the evolution and coloring mechanism of celery. CRediT authorship contribution statement Pei-Zhuo Liu: Writing – original draft, Formal analysis. Ya-Hui Wang: Investigation, Data curation. Yue-Hua Sun: Data curation. Yong-Ju Wei: Formal analysis. Xu Sun: Formal analysis. Meng-Yao Li: Writing – review & editing, Supervision, Conceptualization. Guo-Fei Tan: Writing – review & editing, Supervision, Conceptualization. Ai-Sheng Xiong: Writing – review & editing, Supervision, Conceptualization. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This article was supported by the Key Research and Development Program of Jiangsu (BE2022386), Jiangsu Seed Industry Revitalization Project [JBGS(2021)068], Coordinated Extension of Major Agricultural Technologies Program of Jiangsu (2022-ZYXT-01-3), Key Research and Development Program of Suqian (L202302), Priority Academic Program Development of Jiangsu Higher Education Institutions Project (PAPD), National Natural Science Foundation of China (32002027). Contributor Information Meng-Yao Li, Email: limy@sicau.edu.cn. Guo-Fei Tan, Email: tgfei@foxmail.com. Ai-Sheng Xiong, Email: xiongaisheng@njau.edu.cn. References