Comparative analysis of the complete multi-chromosomal mitogenome of Anoectochilus emeiensis (Orchidaceae) Authors: Song Lu Published: 2026 ABSTRACT AbstractBackgroundAnoectochilus emeiensis K. Y. Lang, an orchid with high ornamental and medicinal value, serves as a core raw material in numerous herbal medicines owing to its pharmacological effects. In 2015, its wild populations comprised six local groups with approximately 5,000 individuals. By 2023, only two populations with fewer than 500 individuals remained, largely resulting from habitat destruction and overexploitation. Deciphering the mitochondrial genome of A. emeiensis will help elucidate the structural diversity and functional conservation of Anoectochilus mitogenomes, and offer insights into genome evolution and adaptive strategies. To date, the A. emeiensis mitogenome has not been documented.ResultsThe mitogenome of A. emeiensis exhibits a multi-chromosomal structure consisting of 20 circular isoforms. Individual chromosome lengths vary from 13,437 bp to 43,704 bp, with a total length of 490,926 bp and an average GC content of 45.07%. The mitogenome harbors 29 protein-coding genes, 21 tRNA genes, and 3 rRNA genes. Codon usage analysis revealed neutral characteristics with A/U bias at the third codon position. Repetitive sequences, intracellular gene migration, genome size and structure, as well as RNA editing sites were all analyzed. These findings offer key clues for exploring the evolutionary process of multi-chromosomal mitochondria in Orchidaceae.ConclusionHybrid sequencing integrating second- and third-generation technologies enabled successful assembly of the A emeiensis mitogenome. This study deepens the understanding of multi-chromosomal mitogenomes in Orchidaceae, illuminates their evolutionary history, and provides valuable resources for germplasm conservation and breeding of this endangered herb.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12870-026-09180-8.Keywords: Anoectochilus emeiensis, Multi-chromosomal mitogenome, Comparative analysis Introduction Orchidaceae is one of the two largest families of angiosperms, comprising 814 genera and approximately 27,000 species [1, 2]. Orchids are classified into epiphytic, mycoheterotrophic, and terrestrial types according to their ecological habitats, and most have important horticultural and pharmacological value [3]. Orchids are well known for their extraordinary morphological diversity and highly complex genomic structures. Previous studies have mainly focused on species taxonomy, horticultural cultivation, mycorrhizal symbiosis, functional gene exploration, and plastid genome analysis, with a few nuclear genomic investigations. In contrast, orchid mitogenomes remain poorly explored [4–6]. Anoectochilus spp., belonging to the subfamily of Orchidoideae [1], possess high ornamental and medicinal value. Up to now, no complete mitogenome has been reported for any species in this genus. As maternally inherited organelles, plant mitochondria possess a semi-autonomous genetic system independent of the nucleus. Plant mitochondrial genomes are structurally more complex and larger than those of animals [7–9]. To date, the smallest known mitochondrial genome among terrestrial plants is from the aerial hemiparasite Viscum scurruloideum Barlow (66 kb), which has lost nine genes encoding subunits of electron-transfer complexes Ⅰ [10]. In contrast, the largest is from the gymnosperm Cathaya argyrophylla Chun & Kuang (18.99 Mb), which expanded via recombination mediated by repetitive sequences and exogenous gene transfer [11]. Generally, mitochondrial gene content and function are relatively conserved. Angiosperm mitogenomes typically contain 50 to 60 genes, including 24 core genes and 17 variable genes [12, 13]. In terms of gene order and structure, plant mitochondrial genomes show much higher variability than plastid genomes [14]. The plant mitochondrial genome remains one of the most poorly understood genomic systems in many plant lineages. Exploring the dynamic structural evolution of orchid mitogenomes has attracted much attention recently. However, high structural complexity, frequent gene rearrangements, inversions, and intracellular gene transfer make mitogenome assembly in Orchidaceae challenging. To date, only a limited number of orchid mitogenomes have been fully assembled. Most available complete mitogenomes belong to the subfamilies of Epidendroideae and Cypripedioideae, with a few reported in Orchidoideae [4–6]. Notably, most sequenced mitogenomes in Orchidaceae display a multi-chromosomal architecture. Anoectochilus Blume (subtribe Goodyerinae, Orchidaceae) includes approximately 30 species worldwide, 11 of which are native to China [1, 2]. All are perennial terrestrial orchids distributed from tropical Asia to Oceania [2]. Many Anoectochilus species, including A. emeiensis K. Y. Lang, A. roxburghii (Wall.) Lindl., and A. formosanus Hayata, are collectively known as “King Medicine” or “Jinxianlian” in traditional Chinese folk medicine [15, 16]. The whole plant is widely used as a medicinal herb or booster-treatment for heat-clearing, detoxification, immune enhancement, and other effects [15, 17]. As a primary original plant of “Jinxianlian”, A. emeiensis is narrowly endemic to Mount Emei in Sichuan, China. It was once treated as part of A. roxburghii and later recognized as an independent taxon, attracting increasing research attention over the past decade [15, 18]. A survey in 2015 recorded six wild populations with nearly 5,000 individuals. By 2023, only two populations with fewer than 500 individuals remained owing to habitat destruction and over-exploitation [16]. Comparative mitochondrial genomic analysis provides insights into the balance between structural variability and functional conservation and sheds light on genome evolution, adaptive strategies, energy metabolism, and population genetic structure. Herein, the inaugural mitogenome of the genus Anoectochilus was assembled and annotated via a combination of Illumina and Nanopore sequencing. Analyses of genome features, as well as repetitive repeats, RNA editing, sequence migration, and codon usage preferences were conducted. Comparison with other Orchidaceae species was also performed. The mitogenome divergence among the subfamilies of Orchidaceae was also explored. These findings improve our understanding on the Anoectochilus mitogenome and provide molecular markers for evolutionary analysis, population genetics, conservation biology, and molecular breeding. Methods and materials Plant materials and genome sequencing Capsules of A. emeiensis were collected from Mount Emei, Sichuan, China. Seeds were cultured and propagated in vitro on MS medium, and seedlings were used for DNA extraction. Sample collection complied with the Convention on the Trade in Endangered Species of Wild Fauna and Flora (CITES) (https://www.cites.org/). The wild samples were identified by Dr. Song Lu. Voucher specimens were deposited at the Emeishan Biological Station under accession number: 17-EJ-07. Qualified sequencing libraries were constructed using second- and third-generation technologies by Genepioneer Biotechnologies (Nanjing, China). Sequencing was performed on the Illumina Novaseq6000 and Oxford Nanopore PromethION sequencer platforms. Mitogenome assembly and annotation Raw Illumina reads were filtered using fastp [19]. Adapters, primer sequences, and low-quality reads were removed to obtain high-quality clean reads. Single stretches of the chloroplast and nuclear contigs were manually removed and visualized using Bandage. Third-generation reads were filtered using Filtlong (v0.2.1, https://github.com/rrwick/Filtlong). Then, Minimap2 (v2.1) was used to map long reads against plant mitochondrial core gene sequences [20]. High-quality sequences containing multiple mitochondrial genes were selected as seed sequences. Filtered long reads were corrected using Canu [21]. Illumina short reads were aligned to corrected long reads using Bowtie2 (v2.3.5.1) [22]. Hybrid assembly was performed using Unicycler (v0.4.8) with the default parameters, followed by manual adjustment [23]. The complete mitogenome of A. emeiensis was annotated using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi), tRNAscanSE (http://lowelab.ucsc.edu/tRNAscan-SE/) and Open Reading Frame Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html) for protein-coding genes, tRNAs, rRNAs, and ORFs, respectively. Mitochondrial genome maps were visualized using OGDRAW (1.3.1) [24]. Codon usage analysis Relative Synonymous Codon Usage (RSCU) reflects unequal usage of codons shaped by natural selection, mutation, and genetic drift. A self-coding Perl script was written to screen coding-gene sequences and analyze codon composition of the A. emeiensis mitogenome. The number of codons and GC content at the first, second, and third positions (GC1, GC2, and GC3) were calculated. RNA editing sites prediction RNA editing sites were identified using two complementary methods. First, PREP-Mt was used to predict the addition, deletion, and substitution sites in the protein-coding genes of the A. emeiensis mitogenome [25]. To improve accuracy, a second method was employed using transcriptome data (accession number: SRR8138703) as the reference [16]. The transcriptome data were mapped to the mitogenome using BWA [26]. Then, a self-coding Perl script was used to identify consistent nucleotide differences between DNA and RNA sequences supported by high read coverage. Repeat sequence analysis Simple sequence repeats (SSRs) were detected using MISA v1.0 with thresholds: 10, 5, 4, 3, 3, 3 for mononucleotide, dinucleotide, trinucleotide, tetranucleotide, pentanucleotide, and hexanucleotide, respectively [27]. Dispersed repeats were identified using BLASTN (parameters: word_size 7, E-value 1e− 5, remove redundancy, remove tandem repeats). Tandem repeats were detected using Tandem Repeats Finder v4.09 (parameters: 2 7 7 80 10 50 2000 -f -d -m). All repeats were visualized using Circos v0.69-5 [28]. Migration analysis of chloroplast-to-mitochondrial DNA Homologous sequences between chloroplast and mitochondrial genomes of A. emeiensis were identified using BLAST with criteria: E-value ≤ 1e− 5 and sequence similarity ≧ 70%. Results were visualized using Circos v0.69-5 [28]. Comparative analysis of the mitogenome structure Mitogenome structures of A. emeiensis and eight other Orchidaceae species were compared using CGView with default set (http://stothard.afns.ualberta.ca/cgview_server/) (Table S1). Six species belong to Epidendroideae, and one each to Cypripedioideae and Apostasioideae [1, 2, 29]. The mitogenome of A. emeiensis was set as the reference of Orchidoideae. Collinearity analysis was carried out using Nucmer v4.0.0 beta2 (– maxmatch parameter) and dot-plots were generated [30]. Ka/Ks and nucleotide variability (Pi) analysis Ka/Ks ratios of protein-coding genes (PCGs) were calculated between A. emeiensis and eight Orchidaceae species (Table S1). Homologous genes were aligned using Mafft v7.427 (auto mode) (https://mafft.cbrc.jp/alignment/software/) [31]. KaKs values were computed using KaKs Calculator v2.0 with MLWL method (https://sourceforge.net/projects/kakscalculator2/) [32]. Nucleotide variability (Pi) was calculated using DnaSP v5.10.1 [33]. Phylogenetic analysis A total of 22 mitochondrial genomes were downloaded from the NCBI Organelle Genome Database (https://www.ncbi.nlm.nih.gov/) (Table S1). A batch of 23 shared PCGs were aligned using MAFFT v7.427 (automatic mode) and trimmed with trimAl v1.4.rev15 (parameter: -gt 0.7) [20, 34]. The best-fit evolutionary model was determined using jModeltest v2.1.10 [35]. The maximum likelihood (ML) phylogenetic analysis was performed using RAxML v8.2.10 (https://cme.h-its.org/exelixis/software.html) under the GTRGAMMA model with 1,000 bootstrap replicates [36]. The phylogenetic tree was visualized using Figtree v1.4.4 [37]. Results General features of the Anoectochilus emeiensis mitogenome The complete mitogenome of A. emeiensis was successfully assembled via hybrid Illumina and Nanopore sequencing, with an average read length of 11,531 bp (Table 1, Table S2). It exhibits a unique multi-chromosomal structure comprising 20 circular chromosomes, with sizes ranging from 13,437 bp to 43,704 bp (Fig. 1). The GC content of individual chromosomes varies from 42.69% to 48.38% (Table 1). In total, 53 genes were annotated, including 29 PCGs, 3 rRNAs, and 21 tRNAs (Table 1, Table S3). Moreover, 100 open reading frames (ORFs) were predicted (Table 2, Table S4). These 29 PCGs were categorized into nine groups (numbers represent n): respiratory complex I (9), ATP synthase (5), cytochrome c oxidase (3), and cytochrome c biogenesis (4), maturase (1), transport membrane protein (1), ubiquinol-cytochrome c reductase (1), and large (1) and small (4) ribosomal proteins (Table 3). Table 1General features of the Anoectochilus emeiensis mitogenomeGenome FeatureAnoectochilus emeiensisGenome size (bp)490,926Numbers of contigs20Contig length (bp)13,437 bp to 43,704 bpGenBank NosPX764234.1-PX764253.1GC content (%)42.69% to 48.38%Length of the coding region (%)29,574 (6.02%)Length of the rRNA genes (%)5,421 (1.10%)Length of the tRNA genes (%)1,534 (0.31%)Number of protein-coding genes29Number of rRNA genes3Number of tRNA genes21Total genes53 General features of the Anoectochilus emeiensis mitogenome Mitochondrial genome map of A. emeiensis. Genomic features transcribed clockwise and counterclockwise are drawn inside and outside of the outer circle, respectively. Genes are color-coded based on their functions. Dark gray plots on the inner circle represent GC content Table 2Gene and ORF content of the mitogenome of A. emeiensisChromosomeLength (bp)GC content (%)GenesORFChr 143,70445.57nad2, rrn18, rrn5, trnY-GTA, trnC-GCA, trnH-GTGORF80, ORF93, ORF97, ORF118, ORF134, ORF159, ORF196, ORF205, ORF211, ORF236, ORF243, ORF268, ORF281Chr 240,86544.14nad6, trnN-GTT, trnD-GTCORF52, ORF65, ORF87, ORF116, ORF134, ORF135, ORF137, ORF140, ORF141, ORF156, ORF171, ORF179, ORF201, ORF227, ORF228, ORF235Chr 333,68444.59rrn26, trnM-CATORF33, ORF82, ORF130, ORF132, ORF163, ORF177, ORF211, ORF215Chr 430,96745.38rps13, mttB, trnH-GTG, trnQ-TTG, trnR-ACG, trnN-GTTORF32, ORF75, ORF81, ORF127, ORF131, ORF137, ORF149, ORF163, ORF164, ORF165, ORF170, ORF209, ORF211, ORF216Chr 530,17945.49nad2, trnM-CAT, trnK-TTTORF33, ORF155, ORF178, ORF184Chr 627,92346.9nad7, trnK-TTT, trnM-CAT, trnE-TTGORF141Chr 727,85244.42nad9, trnF-GAA, trnW-CCA, trnP-TGGORF46, ORF56, ORF117, ORF175, ORF177Chr 827,20344.57atp6, trnM-CATORF93, ORF98, ORF109Chr 925,45546.7 nad5 ORF64, ORF82, ORF99, ORF115, ORF126, ORF181Chr 1022,78945.32 cox3 ORF74, ORF138, ORF140Chr 1122,19842.69cox2, trnR-TCTORF37, ORF45, ORF65, ORF71, ORF85, ORF86, ORF87, ORF105Chr 1221,38748.38nad3, nad4, rps12, trnQ-TTGORF167Chr 1320,47445.52atp1, ccmFnORF42, ORF67, ORF98Chr 1418,71745.09 nad5 ORF29, ORF69, ORF117, ORF124Chr 1518,16944.41atp4, atp8, nad4LORF42, ORF54, ORF71, ORF75, ORF76, ORF91Chr 1617,98043.43cox1, cob, atp9, rps7, rps14, rpl5ORF56, ORF105Chr 1717,29244.21 ccmFc Chr 1816,39146.38nad1, matRORF83Chr 1914,26042.97 ccmC ORF59Chr 2013,43743.65 ccmB ORF11 Gene and ORF content of the mitogenome of A. emeiensis Table 3Encoding gene content and organization of the mitogenome of A. emeiensisGroup of genesGene nameLengthStart codonStop codonAmino acidATP synthase atp1 1530ATGTGA510 atp4 582ATGTGA194 atp6 807ATGTAG269 atp8 480ATGTAA160 atp9 225ATGTGA75Cytochrome c biogenesis ccmB 615ATGTGA205 ccmC 762ATGTGA254 ccmFc 1320ATGTGA440 ccmFn 1713ATGTGA571Ubiquinone cytochrome c reductase cob 1182ATGTGA394Cytochrome c oxidase cox1 1584ATGTAA528 cox2 801ATGTAG267 cox3 798ATGTGA266Maturases matR 1899ATGTAG633Transport membrane protein mttB 351ATGTAG117NADH dehydrogenase nad1 978ACG (ATG)TAA326 nad2 1509ATGTGA503 nad2 1509ATGTGA503 nad3 357ATGTAA119 nad4 1488ATGTGA496 nad4L 303ATGTAA101 nad5 2016ATGTAA672 nad5 2016ATGTAA672 nad6 915ATGTAG305 nad7 1185ATGTAG395 nad9 612ATGTAA204Ribosomal proteins (LSU) rpl5 558ATGTAA186Ribosomal proteins (SSU) rps12 378ATGTGA126 rps13 351ATGTGA117 rps14 303ATGTAG101 rps7 447ATGTAA149Ribosomal RNAs rrn18 1920 rrn26 3385 rrn5 116Transfer RNAs trnC-GCA 73 trnD-GTC 74 trnE-TTC 72 trnF-GAA 73 trnH-GTG 74 trnH-GTG 74 trnK-TTT 72 trnK-TTT 73 trnM-CAT 64 trnM-CAT 73 trnM-CAT 74 trnM-CAT 73 trnN-GTT 72 trnN-GTT 72 trnP-TGG 74 trnQ-TTG 72 trnQ-TTG 72 trnR-ACG 74 trnR-TCT 72 trnW-CCA 74 trnY-GTA 83other Encoding gene content and organization of the mitogenome of A. emeiensis Consistent with findings in most angiosperms, the mitogenome of A. emeiensis contains three rRNA genes [38], rrn18 (1,920 bp), rrn26 (3,385 bp), and rrn5 (116 bp) (Table 3). Moreover, 15 different tRNA genes were annotated, with multiple copies for trnH-GTG (2), trnM-CAT (4), trnN-GTT (2), and trnQ-TTG (2). The length of tRNA genes varied from 64 to 83 bp (Table 3). Intron numbers in terrestrial plants generally range from 1 to 4 [14]. A total of 23 introns were identified in the mitogenomic genes (Table S5). Most introns (19) occur in NADH dehydrogenase genes, in which nad1, nad2, nad5, and nad7 contain four introns, and nad4 contains three. ccmFc and cox1 each have one intron, and cox2 has two (Table S5). Exons of nad1, nad2, and nad5 are located on different chromosomes (Table 3, Table S6). The lengths of these intronic regions span from 189 (nad2) bp to 7,222 bp (nad4) (Table 3, Table S6). Repeat sequences in the mitogenome of A. emeiensis A total of 862 dispersed repeats were identified, including 370 forward repeats and 492 palindrome repeats, with lengths from 27 bp to 541 bp, covering 61,611 bp (12.55%) of the mitogenome (Fig. 2-A and -B, Table S7). Approximately 50.2% of the repeats (i.e., 433 fragments) are medium-sized (50 to 300 bp) (Table S7). Moreover, 142 SSRs were detected, dominated by monomeric and tetrameric types, accounting for 30.28% and 28.17%, respectively (Fig. 2-C, Table S8). A total of 23 tandem repeats were identified, with lengths ranging from 15 bp to 70 bp and sequence identity ≧ 80% (Table S9). The repeat sequences in A. emeiensis mitogenome. A The repeat sequences identified in the mitogenome. The inner circle represents dispersed repeats on different chromosomes, marked by colored arcs extending outward from the center. The tandem repeats are displayed as short bars in the middle ring, and microsatellite sequences are annotated on the outer circle. The scale is shown with intervals of 10 kb. B The length distribution of dispersed repeats. C The numbers of identified SSR motif types Analysis of codon usage In total, 8,683 codons were detected in the coding genes. The mitogenome shows a strong AT-bias, with an average GC content of 44.28% across the three codon positions. GC1 (49.78%), GC2 (44.13%), and GC3 (38.93%) are all below 50%. Most genes use ATG as the start codon, except nad1 (ACG) (Table 3). The usage frequencies of the three stop codons are TGA (45.16%), TAA (32.26%), and TAG (22.58%) (Table 3). Leucine (Leu) (10.78%) is the most frequent amino acid, followed by serine (Ser) (9.76%) and isoleucine (Ile) (7.27%) (Table S10). Excluding the stop codon UGA, 29 codons have RSCU values greater than 1, encoding 19 amino acids, with a strong A/U preference (90%) at the third codon position (Fig. 3). Codon usage bias of mitochondrial protein-coding genes in A. emeiensis. The bottom square represents codons encoding amino acids. The height of the upper column indicates the sum of RSCU values of all codons Substitution rates of protein-coding genes Ka/Ks ratios of the 29 PCGs were compared across nine Orchidaceae species (Fig. 4, Table S11). Values range from 0.056 (cox1) to 6.868 (nad6). Most genes are under purifying selection (Ka/Ks < 1). The nad6 gene shows an average Ka/Ks value (1.25), indicating positive selection. cox1 (0.24) has the lowest average Ka/Ks value, suggesting strong purifying selection (Fig. 4, Table S11). Boxplots of the pairwise KaKs values among shared genes of the 9 species in Orchidaceae. The “X” axis shows the name of protein-coding genes, and the “Y” axis shows the Ka/Ks values Synteny analysis and mitogenome comparison The synteny dot-plots show moderate to high similarity (12.04%–31.99%) between A. emeiensis and other orchids (Fig. 5). The highest similarity was found between A. emeiensis and Cymbidium lancifolium Hook. Extensive syntenic blocks indicate both conserved sequences and frequent rearrangements in orchid mitogenomes (Fig. 5). Chromosome 12 shows higher collinearity than other chromosomes (Fig. 5). Dot-plot graphs show the synteny regions between mitogenomes in related species compared to A. emeiensis. The red line segment indicates forward direction, and the blue line segment represents reverse direction Gene content is largely conserved across nine Orchidaceae species (Fig. 6). All contain 5 ATP synthase genes, 4 cytochrome c biogenesis genes, and 1 ubiquinol-cytochrome c reductase gene. Apostasia shenzhenica Z. J. Liu & L. J. Chen lacks cox2; Cymbidium macrorhizon lacks the matR; Dendrobium amplum Lindl. and Hygrochilus parishii Pfitzer lack mttB. Five species, including A. emeiensis, retain all 9 NADH dehydrogenase genes. The A. emeiensis mitogenome retains only rpl14 among the large ribosomal subunit genes, while most rpl and rps genes are lost. The sdh4 gene is intact in Paphiopedilum micranthum Tang & F. T. Wang, but pseudogenized in A. emeiensis (Fig. 6). Comparison of mitochondrial gene content in nine Orchidaceae species. Grey boxes represent that the gene is lacking in the mt genomes. Blue boxes depict the occurrence of the particular mitochondrial genomes Mitogenome sizes range from 473,821 to 1,250,579 bp in nine Orchidaceae species. Gastrodia elata Bl. has the largest, and D. amplum the smallest (Table S12). Eight of the nine mitogenomes are multi-chromosomal, with 17 to 35 chromosomes. Nucleotide diversity (Pi) Nucleotide diversity (Pi) of 29 PCGs and 3 rRNA genes among nine Orchidaceae species ranges from 0.002 to 0.046, with most values below 0.02 (Fig. 7). The most variable genes are atp8 (Pi = 0.046), mttB (Pi = 0.045), and ccmFc (Pi = 0.037). The most conservative protein-coding gene is nad7 (Pi = 0.009) (Fig. S3). Nucleotide diversity (Pi) analysis of protein-coding genes (PCGs) in the mitochondrial genome of A. emeiensis and eight Orchidaceae species. The “X” axis shows the names of protein-coding genes, and the “Y” axis shows the Pi values RNA editing sites In this study, 541 RNA editing sites were predicted within 33 protein-coding genes. Editing sites per gene range from 1 to 78 (Table S13). After editing, 158 (29.21%) hydrophobic and 72 (13.31%) hydrophilic amino acids remain unchanged, whereas 256 (47.32%) hydrophilic sites convert to hydrophobic amino acids, and 51 (9.43%) hydrophobic amino acids become hydrophilic (Table S13). Four (0.74%) editing events create premature stop codons: atp8, atp9, ccmFc (CGA→TGA), and atp6 (CAA→TAA) (Table S13). All editing events are cytosine-to-thymine conversions in the A. emeiensis mitogenome. Editing occurs mainly at the first (31.31%) and second (62.25%) codon positions. No editing was detected at the third position. A special editing type was observed that simultaneous changes at two bases convert proline (CCC or CCT) to phenylalanine (TTC or TTT) (Table S13). Phylogenetic analysis Phylogenetic analysis was performed based on 29 core PCGs from 23 species, with Magnolia officinalis Rehd. et Wils. and Ginkgo biloba L. as outgroups. All bootstrap values exceeded 80 (Fig. 8), supporting robust relationships. Nine orchids form a monophyletic clade sister to the Liliaceae clade. The placement of different clades generated from the mitogenomic phylogenetic analysis is in agreement with the result of the APG Ⅳ classification [2]. Distribution of homologous fragments between the chloroplast and mitochondrial genome. The comparison illustrates the MTPT (mitochondrial plastid DNA) sequences. Colored curves depict the linkage of homologous sequence fragments between two genomes Horizontal gene transfer in the A. emeiensis mitogenome Homologous sequences between the A. emeiensis mitogenome and the chloroplast genome were detected. 71 chloroplast-derived fragments were identified, totaling 11.51% (56,487 bp) of the mitogenome, ranging from 51 bp to 4,027 bp. The chloroplast-origin sequences are distributed across 14 chromosomes and include 13 nearly intact tRNAs and 2 partial rrn genes (Fig. 9, Tables S14). Moreover, many ATP synthesis-related genes, NADH dehydrogenase genes, and photosynthesis-related genes were found to migrate from the chloroplast to the mitochondrion, with sequence similarities ranging from 74.15% to 99.27% (Fig. 9, Tables S14). The phylogenetic relationships of A. emeiensis with other plant species based on the nucleotide sequences of protein-coding genes from the mitogenome. The analysis was performed with the best evolutionary GTRGAMMA model based on the maximum likelihood (ML) analysis with bootstrap = 1000 Discussion General features of the A. emeiensis mitogenome Plant mitochondria contain semi-autonomous genetic systems encoding a limited set of proteins, rRNAs, and tRNAs, whereas the other elements related to DNA replication, transcription, and translation are encoded in the nucleus [39, 40]. Plant mitogenomes exhibit diverse structures: single circular, linear, multi-circular, or a circular molecule with linear tails [39, 41, 42]. Their intricate structural features make it challenging to sequence and assemble plant mitochondrial genomes. This study presents the first complete mitogenome for the genus Anoectochilus, assembled via hybrid short- and long-read sequencing. The Illumina sequencing depth exceeded 200× with 100% coverage, ensuring high accuracy [Fig. S1]. All core mitochondrial genes are intact, confirming the assembly integrity. Unlike the typical single-circle structure in most angiosperms, the A. emeiensis mitogenome comprises 20 circular chromosomes. Multi-chromosomal architecture is common in Orchidaceae, as observed in Cymbidium ensifolium (19 circular isoforms) [3], Dendrobium wilsonii Rolfe (22 circular isoforms) [43], Dendrobium henanense J. L. Lu & L. X. Gao (24 circular isoforms) [44], and Gastrodia angusta S. Chow & S. C. Chen (36 circular isoforms) [45]. Repeat-mediated recombination is widely recognized to facilitate the formation of multi-isoform mitogenomes in angiosperms [46]. In many cases, long repeats (≥ 1,000 bp) give rise to the dynamic, and oversized multi-isoform mitogenomes [43, 44]. For example, recombination and rearrangement were identified in D. wilsonii and D. henanense, which harbor six and five long repeats (≥ 1,000 bp), respectively [43, 44]. Recombination mediated by short repeats usually occurs at a low frequency and is difficult to detect [47], whereas it does exist in certain cases, as exemplified by the mitogenome of Glycine max [48]. Long-read coverage confirms that the multi-loop structure of the A. emeiensis mitogenome does not arise from long-read mediated recombination [Fig. S1]. A similar pattern was also reported in C. ensifolium, which lacks long repeats too (longest repeat: 759 bp) [3]. The mitogenome of A. emeiensis is rich in short repeats (< 100 bp). Although recombination mediated by short repetitive fragments was not directly detected in the A. emeiensis mitogenome, its potential occurrence cannot be entirely excluded. RNA editing events RNA editing prevails in plant organelles and modifies transcripts at the post-transcriptional level. The 541 RNA editing sites identified in the A. emeiensis mitogenome are within the range reported for most angiosperms, similar to D. wilsonii (571), D. henanense (567) and Curcuma amarissima Roscoe (492) [43, 44, 49], but lower than in some gymnosperm species such as Cycas taitungensis C. F. Shen (1,214) and G. biloba (1,306) [50, 51]. Most editing events are nonsynonymous and enhance hydrophobicity (totally 76.53%). All editing sites are C-to-T conversions, occurring mainly at the second codon position, which is the most common pattern in plant mitogenomes [52]. RNA editing can cause premature termination and thereby affect gene function. Abundant editing sites in atp, nad and cox genes imply their important roles in mitochondrial respiration, aquatic adaptation and stress response. These findings are useful for further understanding plant mitochondrial evolution and gene expression. Codon usage pattern Codons usage bias reflects evolutionary selection pressure [53]. The codon usage bias in A. emeiensis mitogenome shows typical land plant patterns that ATG serves as the dominant start codon and codons exhibit strong A or T(U) preference at the third codon position. Nearly all codons with RSCU > 1 end in A or T (U). The remaining four codons with RSCU < 1 are located in Ala (GCA, 0.94), Ter (UAA, 0.931), Leu (CUA, 0.888), and Ile (AUA, 0.734). This trend is commonly observed in terrestrial plant species such as C. amarissima, C. ensifolium, and Isochrysis galbana [3, 50, 54]. Chloroplast-origin sequences in A. emeiensis mitogenome Gene transfer events from plastid to mitochondrion frequently occur in flowering plants [55, 56]. In angiosperms, chloroplast-derived sequences generally occupy 3% to 11.5% of the mitogenome [57, 58]. Homologous recombination facilitates organellar DNA exchange [59, 60]. Chloroplast sequences are considered a major source of mitochondrial genome expansion [61]. In A. emeiensis, many chloroplast genes, including the whole set of ndh genes, are transferred to the mitogenome but become non-functional, with tRNA genes being the exception. One hypothesis is that native mitochondrial genes remain functional, making transferred copies redundant. Uneven distribution of chloroplast fragments was detected in the A. emeiensis mitogenome, with chromosomes 4 and 11 harboring 14,363 bp and 12,261 bp, respectively. Moreover, the presence of multiple tRNA copies indicates that these transferred genes have undergone successive recombination events. Mitogenome comparison in Orchidaceae Mitogenome structure and gene content vary substantially among plants. Comparative analysis of nine Orchidaceae mitogenomes revealed frequent gene loss and replication. The mitogenome of A. emeiensis is relatively conserved, encoding 29 out of the 43 core plant mitochondrial genes, but has lost most rps and several rpl genes. Loss of 12 ribosomal protein genes is rare in orchids but common in the genus Silene [62]. Furthermore, the sdh genes which encode succinate dehydrogenase are lost or pseudogenized in all examined orchids, a pattern widespread in Orchidaceae and numerous monocots [43, 63]. To date, only Apostasia (Apostasioideae) possesses a single-ring mitogenome [4, 64], while other Orchidaceae subfamilies have multi-chromosomal architectures [6, 65, 66]. Variations in genomic structure, size, chromosomal number, and gene content reflect high mitogenomic plasticity in Orchidaceae. No clear evolutionary pattern has been identified across these subfamilies based on available data. Besides, most orchids are epiphytic, and about 43 genera include mycoheterotrophic species [67]. No correlation was found between trophic mode and mitogenome complexity too, which is consistent with the result of Kim et al. [67]. Phylogenetic inference Ka/Ks analysis revealed that most Orchidaceae PCGs were under purifying selection, consistent with prior studies [68–71]. Several genes, including ccmFc, nad1, nad6, atp4 and atp9 show elevated substitution rates in Orchidaceae compared with eudicots [45, 65, 72, 73]. Besides, complex I plays an essential role in electron transport and ATP production [74], and genetic mutations may reshape metabolic performance. In this study, nad6 shows positive selection across nine orchids representing terrestrial, epiphytic, and mycoheterotrophic lifestyles. Additionally, mitochondrial genomes serve as reliable markers for resolving phylogenetic relationships within Orchidaceae, even outperforming plastid genomic data in some cases [29, 75, 76]. In this study, the mitochondrial phylogenetic tree placed A. emeiensis in the Goodyerinae clade, with subfamily relationships matching the APG Ⅳ classification: (Apostasioideae (Cypripedioideae (Orchidoideae, Epidendroideae))) [1, 2, 29]. Phylogenetic discordance still remains among nuclear, plastid, and mitochondrial trees in orchid systematics [77–79]. Further exploration into orchid mitogenome evolution will be facilitated by the accumulation of more mitogenomic resources. Mitogenomic evolution can be influenced by the scale of species assemblages. Small and shrinking effective population sizes strengthen genetic drift and reduce the efficiency of natural selection. The mitogenomes are more susceptible to evolutionary bottlenecks and genetic drift compared with nuclear genomes [80]. Listed as a “critically endangered” (CR) species on the IUCN Red List, A. emeiensis has experienced drastic population reduction [16], which may intensify genetic drift in its mitogenome. Future studies with broader sampling are required to clarify mitogenomic evolution in this endangered species. Conclusion The present study reports the inaugural complete mitogenome of the genus Anoectochilus, assembled via second- and third-generation sequencing. The novel mitogenome of A. emeiensis exhibits a multipartite configuration consisting of 20 circular chromosomes. Comprehensive analyses of genomic features, repeats, RNA editing, intracellular gene transfer, codon bias, and phylogeny provide valuable resources for germplasm conservation, evolutionary research, and molecular breeding. Limited data on multi-chromosomal mitogenomes hinder the understanding of repeat-mediated recombination and structural evolution in Orchidaceae. Additional mitogenomes will help clarify these mechanisms. Supplementary Information Supplementary Material 1. Figure S1: Coverage map using third-generation data to check the accuracy of each sequence Supplementary Material 2. Table S1: GenBank accession numbers of mitogenomes for species used in this study. Supplementary Material 3. Table S2: Summary of sequencing statistics. Supplementary Material 4. Table S3: Composition of A. emeiensis mitogenome. Supplementary Material 5. Table S4: Open reading frames annotated in the mitogenome of A. emeiensis. Supplementary Material 6. Table S5: Introns in A. emeiensis mitogenome. Supplementary Material 7. Table S6: The position of 55 genes in chromosomes. Supplementary Material 8. Table S7: Dispersed repeats in the mitogenome of A. emeiensis. Supplementary Material 9. Table S8: The SSR types detected in A. emeiensis. Supplementary Material 10. Table S9: Distribution of tandem repeats in A. emeiensis mitogenome. Supplementary Material 11. Table S10: Relative synonymous codon usage values of A. emeiensis mitogenome. Supplementary Material 12. Table S11: Ka/Ks of genes in nine mitogenomes. Supplementary Material 13. Table S12: Sizes of 9 Orchidaceae plants mitogenomes. Supplementary Material 14. Table S13: Prediction of RNA editing sites. Supplementary Material 15. Table S14: Plastid-derived regions in the mitochondrial genome of A. emeiensis. Acknowledgements The author thanks Dajun Qi for the help in field survey. The author appreciates the efforts of the anonymous reviewers. Abbreviations Murashige and Skoog medium Protein-coding genes Authors’ contributions Concept and method, data analysis, writing, revision, project and funding acquisition, S.L. Funding This work was supported by the Sichuan Natural Science Foundation (No: 2023NSFSC0015). Data availability The annotated sequences of the novel mitogenome of *Anoectochilus emeiensis* was deposited in the NCBI GenBank database with the accession numbers: PX764234-PX764253. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Figure S1: Coverage map using third-generation data to check the accuracy of each sequence Supplementary Material 2. Table S1: GenBank accession numbers of mitogenomes for species used in this study. Supplementary Material 3. Table S2: Summary of sequencing statistics. Supplementary Material 4. Table S3: Composition of A. emeiensis mitogenome. Supplementary Material 5. Table S4: Open reading frames annotated in the mitogenome of A. emeiensis. Supplementary Material 6. Table S5: Introns in A. emeiensis mitogenome. Supplementary Material 7. Table S6: The position of 55 genes in chromosomes. Supplementary Material 8. Table S7: Dispersed repeats in the mitogenome of A. emeiensis. Supplementary Material 9. Table S8: The SSR types detected in A. emeiensis. Supplementary Material 10. Table S9: Distribution of tandem repeats in A. emeiensis mitogenome. Supplementary Material 11. Table S10: Relative synonymous codon usage values of A. emeiensis mitogenome. Supplementary Material 12. Table S11: Ka/Ks of genes in nine mitogenomes. Supplementary Material 13. Table S12: Sizes of 9 Orchidaceae plants mitogenomes. Supplementary Material 14. Table S13: Prediction of RNA editing sites. Supplementary Material 15. Table S14: Plastid-derived regions in the mitochondrial genome of A. emeiensis. Data Availability Statement The annotated sequences of the novel mitogenome of *Anoectochilus emeiensis* was deposited in the NCBI GenBank database with the accession numbers: PX764234-PX764253.