Skip to main content
NCBI home page
Frontiers in Plant Science logoLink to Frontiers in Plant Science
. 2026 Sep 14;17:1955008. doi: 10.3389/fpls.2026.1955008

Characterization of the complete mitochondrial genome of Rhododendron simsii var. putuoense and its comparative evolutionary analysis

Hong Zhu 1,2, Chunlei Yue 1,2,*
PMCID: PMC13617138  PMID: 42807159

Abstract

Rhododendron simsii var. putuoense is a coastal-endemic plant with significant ecological and genetic resource value. Although its chloroplast genome (cpDNA) was recently characterized, its mitochondrial genome (mtDNA) remains unexplored. Here, we report the first complete assembly and annotation of the R. simsii var. putuoense mtDNA using PacBio HiFi long-read sequencing. The genome sequence is 658,161 bp in length, organized as two subgenomic linear molecules (a major one of 633,105 bp and a minor one of 25,056 bp), and encodes 33 protein-coding genes (PCGs), 26 tRNAs, three rRNAs and 323 ORFs. Repeat analysis revealed a predominance of dispersed repeats (72%), alongside 128 SSRs and 20 tandem repeats. The mtDNA exhibits a strong A/U-ending codon usage bias, with 463 RNA editing sites predicted. Six homologous segments (0.91% of the mtDNA) provide evidence of interorganellar gene transfer (IGT). Notably, pairwise Ka/Ks analysis revealed an elevated ratio in the matR gene, suggesting at lineage-specific evolutionary dynamics that warrant further site-specific codon model testing. Phylogenetic and synteny analyses further confirmed a close relationship with R. simsii, while revealing structural divergence consistent with ecological specialization. Overall, these findings deepen our understanding of mitochondrial genomic diversity within Rhododendron and establish a genetic framework for conserving lineage-specific resources in this ecologically significant variety.

Keywords: comparative genomics, interorganellar gene transfer, mitochondrial genome annotation, phylogenetic evolution, Rhododendron simsii var. putuoense

1. Introduction

The study of organelle genomes is of great significance for understanding plant genetics, evolution, and adaptation (Gualberto and Newton, 2017; Møller et al., 2021). As the central organelle for energy metabolism, the mitochondrion harbors key genes related to the respiratory chain (Nolfi-donegan et al., 2020). Compared with the high-density coding of nuclear genomes and the relative conservation of chloroplast genomes (cpDNA), the most prominent features of plant mitochondrial genomes (mtDNA) are their highly dynamic structures and diverse origins of content (Palmer et al., 2000). With the advancement of high-throughput sequencing technologies and the deepening of comparative genomics, analyzing the mtDNA of non-model plants has become a critical entry point for exploring plant adaptive evolution (Zeng et al., 2024; Zhu and Lv, 2025; Zhang et al., 2025).

As one of the most evolutionarily active groups within the Ericaceae family (Shrestha et al., 2018), the genus Rhododendron has long been a challenge in plant systematics due to its complex taxonomy system (Mo et al., 2022; Zhu et al., 2025a). Currently, organelle genome research in this genus focuses predominantly on cp genomes (Ma et al., 2024; Zhou et al., 2024; Zhu et al., 2025b), which provide essential evidence for phylogenetic reconstruction. However, the structural characteristics, evolutionary pressures, and their relationship with ecological adaptation of the mtDNA within this genus have not yet been systematically resolved. This gap, to some extent, restricts a comprehensive understanding of the evolutionary mechanisms of Rhododendron.

Rhododendron simsii var. putuoense, a variety of R. simsii, is endemic to the coastal islands of East China, with a narrow geographical range (Li et al., 2010). Its unique coastal habitat, characterized by environmental stressors such as high salinity and strong winds, has shaped distinct evolutionary traits at the genomic level (Zhu et al., 2026). Although its cp genome was recently characterized (Zhu et al., 2025b), its mtDNA remains unexplored, and a comprehensive analysis of its organellar genomes remains lacking.

In this study, we performed the first assembly and annotation of the R. simsii var. putuoense mtDNA, leveraging the high-quality, chromosome-level nuclear genome (Zhu et al., 2026) and cp genomes previously sequenced for this taxon. By employing comparative genomics, collinearity analysis, phylogenetic reconstruction, and selective pressure analysis, we systematically revealed its genomic structure, evolutionary pressures, and phylogenetic relationships with closely related species. These findings aim to provide important baseline evidence for understanding both the structural plasticity and evolutionary dynamics of Rhododendron mtDNA, while offering preliminary genomic insights into potential adaptive divergence. Overall, this work establishes a robust genetic foundation for the conservation and utilization of this taxon.

2. Materials and methods

2.1. Raw data collection and filtering

The raw data used for the assembly of the R. simsii var. putuoense mt genome were extracted from the whole-genome sequencing (WGS) dataset of this taxon (Zhu et al., 2026), which is publicly available in the National Center for Biotechnology Information (NCBI) database under the accession number JBQRKF000000000 (BioProject ID: PRJNA1305678). According to the previously published study by our research group, the samples used in this study were collected from wild mature individuals at Wulei Mountain, Dinghai District, Zhoushan City, Zhejiang province (coordinates: 29.43°N, 122.20°E; altitude: 359 m) in March 2025. The sample was identified and collected by the authors, and a specimen was deposited in the Herbarium of Zhejiang Academy of Forestry (HZJAF, contact: Hong Zhu, Email: 1107401987@qq.com) under the voucher number HZ2025-DH-05.

2.2. Mt genome assembly and annotation

The PacBio HiFi sequencing raw data were assembled using PMAT software (v2.0.2, with the parameter -g 5G) (Bi et al., 2024). The resulting assembly was visualized and manually adjusted using Bandage (v0.8.1) (Wick et al., 2015) to generate a preliminary draft sequence. Subsequently, the HiFi data were aligned to the assembled sequence using Minimap2 software (Li, 2018), and the assembly was polished using NextPolish (Hu et al., 2020) to obtain the final sequence, followed by corresponding adjustments of the graphical fragment assembly (GFA) file. The annotation of the mt genome structure was conducted in several sequential steps. Protein-coding genes (PCGs), tRNAs, and rRNAs were annotated using the PMGA online platform (http://47.96.249.172:16084/annotate.html) and further refined through manual correction to obtain the final annotation results. Open reading frames (ORFs) with a minimum length threshold of 300 bp were identified using the NCBI ORF Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html). After excluding redundant sequences and those overlapping with known genes, the ORFs were annotated by alignment against the NCBI non-redundant (nr) protein database. Finally, the physical map of the mt genome was generated using OGDRAW (v1.3.1) (https://chlorobox.mpimp-golm.mpg.de/OGDraw.html).

2.3. RNA editing site analysis and codon usage of PCGs

RNA editing sites were predicted based on the sequences of all protein-coding genes (PCGs) encoded by the mt genome. Specifically, cytosine (C)-to-uracil (U) RNA editing sites within the mt PCGs were identified using the Deepred-Mt tool (Edera et al., 2021). To ensure accuracy, only sites with a probability value greater than 0.9 were retained for subsequent analysis. The protein-coding sequences of the mtDNA were extracted using PhyloSuite (v1.1.16) (Zhang et al., 2020). Codon usage bias and relative synonymous codon usage (RSCU) were calculated using MEGA (v7.0) (Kumar et al., 2016). An RSCU value > 1 indicates that a codon is preferred, while an RSCU value < 1 indicates lower usage frequency, and RSCU = 1 represents no usage bias.

2.4. Analysis of repetitive sequences and interorganellar gene transfer

Repetitive sequences, including simple sequence repeats (SSRs), tandem repeats, and dispersed repeats, were identified using various bioinformatics tools. Specifically, SSRs were identified using MISA (v1.0), tandem repeats were detected via Tandem Repeats Finder (v4.09) (http://tandem.bu.edu/trf/trf.submit.options.html), and dispersed repeats were characterized using BLASTN (v2.10.1) (Altschul et al., 1990). All identified repeats were visualized using Circos (v0.69-5).

The cp genome of R. simsii var. putuoense (GenBank accession number: PV614722) had been previously published by our research group and was retrieved from the NCBI database. Homologous sequences between the cp and mt genomes were identified using BLASTN (v2.10.1) with a similarity threshold of 70% and an E-value of 1e-5. The results of the inter-organellar gene transfer (IGT) analysis were visualized using Circos (v0.69-5).

2.5. Ka/Ks and nucleotide diversity (pi) analysis

Homologous gene pairs were extracted by grouping species in a pairwise manner. These gene pairs were subsequently aligned using MAFFT (v7.50) (Katoh and Standley, 2013). Following alignment, the Ka and Ks values for each gene pair were calculated using KaKs_Calculator (v2.0) (Wang et al., 2010), employing the MLWL method. The Ka/Ks ratios for each gene pair were then statistically analyzed and visualized via box plots. For nucleotide diversity analysis, global alignments of homologous gene sequences from different species were performed, and the nucleotide diversity (pi) for each gene was calculated using DnaSP (v5) (Librado and Rozas, 2009).

2.6. Phylogenetic and synteny analysis

To investigate the evolutionary position of R. simsii var. putuoense, a comprehensive analysis of phylogeny and synteny was performed on this taxon and its representative related species. Complete mt genome sequences of 22 species from seven angiosperm families (Ericaceae, Actinidiaceae, Theaceae, Ebenaceae, Primulaceae, Symplocaceae and Aquifoliaceae) were retrieved from the NCBI database. Two Ilex species from the Aquifoliaceae family, I. rotunda (OR666441.1) and I. macrocarpa (OR066215.1), were used as outgroups. A Maximum Likelihood (ML) phylogenetic tree was reconstructed based on the coding sequences (CDS). Multiple sequence alignment was performed using MAFFT (v7.50) (Katoh and Standley, 2013), and the aligned sequences were concatenated. Subsequently, the sequences were trimmed using trimAl (v1.4.rev15) (Capella-gutiérrez et al., 2009) with the parameter “-gt 0.7”. The substitution model was predicted using jModelTest 2.1.10 (https://github.com/ddarriba/jmodeltest2), which identified the GTR model as the best-fit model. The ML tree was then constructed using RAxML v8.2.10 (https://cme.h-its.org/exelixis/software.html) under the GTRGAMMA model with 1,000 bootstrap replicates. To visualize the genomic relationships among these species, genome-to-genome alignments were performed using MUMmer (v4.0.0beta2) (Kurtz et al., 2004) with the “--maxmatch” parameter, and dot plots were generated.

3. Results

3.1. Basic characteristics of the R. simsii var. putuoense mtDNA

R. simsii var. putuoense was sequenced using the HiFi sequencing mode of the PacBio Sequel II platform. After processing the raw data, a total of 1,622,590 high-quality HiFi reads were obtained, yielding 34.27 Gb of data. The average read length was 21,119 bp, and the N50 was 21,174 bp. The assembled mtDNA of R. simsii var. putuoense has a total length of 658,161 bp and exhibits a multi-branched conformational structure with a GC content of 45.84% (Figure 1a). It consisted of two linear subgenomic molecules (a major and a minor molecule), with lengths of 633,105 bp (GC content: 45.83%) and 25,056 bp (GC content: 46.07%), respectively (Figures 1b, c).

Figure 1.

Panel a shows a colored graph with connected segments labeled by length and coverage, representing mitochondrial genome assembly. Panel b displays a circular genome map for contig 1, 833,105 base pairs, annotated with gene positions and functions. Panel c presents a similar circular map for contig 2, 25,056 base pairs, with gene annotations around the ring.

Open in a new tab

The assembly result of the R. simsii var. putuoense mtDNA. (a) The assembly graph displaying a multi-branched conformational structure; (b, c) Circular maps of the two independent linear subgenomic molecules, designated as contig 1 (major molecule, 633,105 bp) and contig 2 (minor molecule, 25,056 bp), respectively.

A total of 64 genes were annotated in the linear mtDNA of R. simsii var. putuoense, including 26 tRNA genes, three rRNA genes, two pseudogenes, and 33 PCGs. Additionally, 323 ORFs were identified. The 33 PCGs can be classified into 11 categories based on their functions: seven categories belong to core genes related to the core respiratory chain, including ATP synthase (5 genes), cytochrome c biogenesis (4 genes), ubiquinol-cytochrome c reductase (1 gene), cytochrome c oxidase (3 genes), maturases (1 gene), transport membrane protein (1 gene), and NADH dehydrogenase (9 genes). The remaining four categories are non-core functional genes, including large ribosomal proteins (LSU) (3 genes), small ribosomal proteins (SSU) (6 genes), succinate dehydrogenase (2 genes), and ribosomal RNAs (3 genes) (Table 1).

Table 1.

Mt gene annotation of R. simsii var. putuoense..

Group of genes Names of genes
ATP synthase atp1, atp4, atp6, atp8, atp9
Cytochrome c biogenesis ccmB, ccmC, ccmFc*, ccmFn
Ubiquinol cytochrome c reductase cob
Cytochrome c oxidase cox1, cox2, cox3
Maturases matR
Transport membrane protein mttB
NADH dehydrogenase nad1****, nad2****, nad3, nad4**, nad4L, nad5****, nad6, nad7****, nad9
Large ribosomal proteins (LSU) #rpl16, rpl10, rpl5
Small ribosomal proteins (SSU) #rps19, rps1, rps12, rps14*, rps3*, rps4
Succinate dehydrogenase sdh3, sdh4
Ribosomal RNAs rrn18, rrn26, rrn5
Transfer RNAs trnA-TGC*, trnC-GCA, trnD-GTC, trnE-TTC, trnF-AAA*, trnF-GAA, trnG-GCC(2), trnH-GTG, trnI-TAT, trnK-TTT, trnK-TTT*, trnM-CAT(6), trnN-GTT, trnP-TGG, trnQ-TTG, trnR-TCT*, trnS-GCT, trnS-TGA, trnW-CCA, trnY-GTA
Open in a new tab

*indicates intron number; #indicates pseudogene; (2) indicates the number of copies of multi-copy genes.

3.2. Analysis of RNA editing sites

In total, 463 RNA editing sites were predicted across the 33 mt PCGs. Among them, the mttB gene possessed the highest number of RNA editing sites (39), followed by ccmB and nad2, each containing 38 RNA editing sites (Figure 2). Among the predicted editing events, 11.76% resulted in silent hydrophilic substitution (hydrophilic-to-hydrophilic), while 30.72% retained hydrophobicity (hydrophobic-to-hydrophobic). However, 49.67% of the amino acids transitioned from hydrophilic to hydrophobic, while 7.63% shifted from hydrophobic to hydrophilic. Additionally, 0.22% of the sequences resulted in premature termination of the coding process (stop codons).

Figure 2.

Bar chart illustrating the number of RNA-editing sites for various genes, with mttB, ccmB, and nad4L having the highest counts at thirty-nine, thirty-eight, and thirty-eight sites respectively, while several genes have fewer than ten sites.

Open in a new tab

Distribution of RNA editing sites among R. simsii var. putuoense mt protein-coding genes.

3.3. Analysis of codon usage bias

A total of 9,987 codons were detected across all genes. Leucine (Leu), encoded by six synonymous codons, was the most abundant amino acid in this mtDNA, with 1,142 codons (11.43%). It was followed by serine (Ser) with 900 codons (9.01%) and arginine (Arg) with 617 codons (6.18%). In contrast, tryptophan (Trp) accounted for only 1.54% of the total amino acids. The RSCU values ranged from a maximum of 1.73 (UUA) to a minimum of 0.364 (UAG). Thirty-one codons had RSCU values greater than 1, indicating that these codons are used more frequently than other synonymous codons. Among these, 28 codons (90.32%) ended with A or U, while only three ended with C or G. This demonstrates a clear preference for A/U-terminated codons in the frequently used codons of the R. simsii var. putuoense mtDNA (Figure 3).

Figure 3.

Stacked bar chart illustrating Relative Synonymous Codon Usage (RSCU) for different amino acids, with color-coded segments representing individual codons arranged above their corresponding nucleotide sequences in a table below each bar.

Open in a new tab

Relative synonymous codon usage (RSCU) histogram of R. simsii var. putuoense mt genome.

3.4. Analysis of repeat sequences

A total of 527 repetitive sequences were identified in the R. simsii var. putuoense mtDNA. Dispersed repeats were the most predominant, accounting for 379 (72%) of the total, including 170 pairs of forward repeats (F) and 209 pairs of palindromic repeats (P). Simple sequence repeats (SSRs) were the next most common, with 128 (24%) identified. Tandem repeats were the least frequent, representing only 20 (4%) of the total repetitions (Figure 4).

Figure 4.

Circular genomic plot illustrating two chromosomes, labeled one and two, with coordinates in kilobases marked around the circumference. Internal curved lines in light blue represent sequence similarities or genomic links between regions of the two chromosomes.

Open in a new tab

The distribution of repeats in R. simsii var. putuoense mtDNA. The outermost circle represents the mt genome sequence, and from outside to inside are simple sequence repeats, tandem repeats and dispersed repeats, respectively.

3.5. Analysis of homologous segments between the mt and cp genomes

A total of six homologous segments were identified between the mt and cp genomes, with a total length of 5,967 bp, accounting for 0.91% of the mtDNA. Among these, five segments were entirely located within the homologous sequences and were all tRNA genes (trnA-TGC, trnW-CCA, trnD-GTC, trnN-GTT and trnM-CAT). One segment was partially located (44.51%) within the homologous sequence, which belongs to the ribosomal RNA gene rrn18 (Figure 5).

Figure 5.

Circular genome map diagram comparing two chromosomes, labeled “1” and “2,” illustrating gene locations with color-coded blocks representing different gene categories and connecting lines showing regions of similarity. Legend at the bottom left explains gene category colors including maturases, ATP synthase, and ribosomal RNAs. Distances in kilobases are marked around the circumference, and gene names are annotated at intervals.

Open in a new tab

Interorganellar sequence transfer between cp and mtDNA of R. simsii var. putuoense.

3.6. Ka/Ks analysis

In this study, the Ka/Ks ratios were calculated for the shared PCGs among five Rhododendron species: R. simsii (MW030508.1), R. × pulchrum (OM283814.1), R. decorum (OQ450179.1), R. vialii (PQ537097.1) and R. nivale subsp. boreale (PQ537104.1). When comparing the mtDNA of R. simsii var. putuoense with R. simsii, the overall Ka/Ks value was 0.65, with three PCGs exhibiting Ka/Ks < 1. Compared to R. × pulchrum, the Ka/Ks value was 0.61, with three PCGs showing Ka/Ks < 1. Relative to R. decorum, the Ka/Ks value was 0.47, and 13 PCGs had Ka/Ks < 1. In comparison with R. vialii, the Ka/Ks value was 0.32, with nine PCGs showing Ka/Ks < 1, and one PCG showing Ka/Ks > 1. Compared to R. nivale subsp. boreale, the Ka/Ks value was 0.31, with 13 PCGs showing Ka/Ks < 1. Notably, almost all Ka/Ks ratios were less than 1, indicating that most PCGs have undergone purifying selection during evolution. In contrast, one gene (matR) exhibited a Ka/K ratio > 1, pointing to a potential signature of elevated evolutionary rate or positive selection (Figure 6).

Figure 6.

Box plot showing Ka/Ks ratios for various genes, with each gene labeled on the x-axis. MatR displays a notably high Ka/Ks ratio above 4.0, while other genes show much lower ratios. A red dashed line is drawn at a Ka/Ks value of 1.0 to indicate neutral evolution. The y-axis is labeled Ka/Ks and the x-axis is labeled Genes.

Open in a new tab

Boxplots of the pairwise Ka/Ks values among all shared mt genes of the 6 Rhododendron species.

3.7. Phylogenetic and synteny analysis

The phylogenetic analysis of mt genomes showed that R. simsii var. putuoense clustered with R. simsii and R. x pulchrum in an independent subclade (bootstrap value = 100%), indicating an extremely close phylogenetic relationship among the three taxa (Figure 7).

Figure 7.

Phylogenetic tree diagram showing relationships among several plant species within families such as Ericaceae, Actinidiaceae, Theaceae, Ebenaceae, Primulaceae, Symplocaceae, and Aquifoliaceae. One clade of Rhododendron species is highlighted, and “Rhododendron simsii var. putuoense (In this Research)” is marked in red text. Bootstrap values appear at nodes, and a scale bar represents genetic distance.

Open in a new tab

Phylogenetic tree of R. simsii var. putuoense constructed using mt genomes of 22 species. Bootstrap support values for each node are presented as percentages based on 1,000 replicates.

Synteny analysis revealed significant structural variations in the mt genomes between R. simsii var. putuoense and other species within the same genus (Figure 8). Among them, R. simsii var. putuoense exhibited the highest level of synteny with R. simsii (82.97% and 81.78%, respectively), indicating strong genomic conservation and homology. This was followed by R. x pulchrum (75.85% and 71.82%, respectively). In contrast, the synteny level between R. simsii var. putuoense and R. nivale subsp. boreale was markedly the lowest (47.64% and 62.45%, respectively), reflecting extensive genomic rearrangements and greater evolutionary divergence.

Figure 8.

Five scatter plots compare Rhododendron simsii var. putuoense to five other Rhododendron taxa, each labeled with a percentage. Colored diagonal line segments populate each plot, with axes representing genome coordinates. Plots display varying densities and orientations of red and blue lines.

Open in a new tab

Collinearity analysis of R. simsii var. putuoense compared to five species in Rhododendron. The values in parentheses indicate the proportion of homologous sequences relative to the total genome. The red lines within the box represent forward alignments, and the blue lines represent reverse complementary alignments.

4. Discussion

4.1. Characteristics of post-transcriptional regulation and evolutionary stability of the mt genome

Mt gene expression in plants is primarily regulated post-transcriptionally, with RNA editing, codon usage bias, and repetitive sequence structures serving as key mechanisms (Binder and Brennicke, 2003; Gualberto and Newton, 2017). In this study, 463 RNA editing sites were predicted in the R. simsii var. putuoense mtDNA. This number is comparable to those reported for other Rhododendron species such as R. nivale subsp. boreale and R. vialii (Lyu et al., 2024), indicating a complete RNA editing system. The observed predominance of hydrophilic-to-hydrophobic transitions after RNA editing is highly significant, as it suggests a concerted mechanism to preserve the hydrophobicity of trans-membrane proteins within the mt respiratory chain (Jobson and Qiu, 2008). Additionally, approximately 0.22% of editing events regulate protein expression by introducing premature stop codons. These precise post-transcriptional regulations compensate for mutations at the genomic level, ensuring the normal execution of mt physiological functions for R. simsii var. putuoense under complex environments.

Codon usage bias analysis showed that synonymous codons ending in A/U predominate (90.32%) in the R. simsii var. putuoense mtDNA. This bias is consistent with the typical features of most angiosperm mtDNA, reflecting a balance between mutational pressure and natural selection at the translational level (Smith, 2016). The codon with the highest RSCU value was UUA (1.73), further reinforcing the A/U preference. Leucine, as the most abundant amino acid (11.43%), may be related to its high-frequency distribution in mt transmembrane proteins, such as NADH dehydrogenase subunits.

A total of 527 repetitive sequences were identified in the R. simsii var. putuoense mtDNA, with dispersed repeats being predominant (72%). The presence of numerous repetitive sequences, particularly forward (F) and palindromic (P) repeats, is generally considered a primary driver of frequent inter-molecular recombination in angiosperm mtDNA (Alverson et al., 2011; Gui et al., 2025). This provides a mechanistic explanation for the multi-branched conformational structure observed in R. simsii var. putuoense and the substantial synteny variation detected among congeneric species. Similar repeat-mediated structural dynamics have also been reported in Clausena lansium ‘Ji Xin’, where dispersed repeats drive mitogenome reorganization into a multipartite architecture (Bi et al., 2026). Notably, such dispersed repeat-mediated recombination primarily operates at the level of large-scale genomic architecture, generating structural polymorphisms among related taxa. In contrast, SSRs and tandem repeats, which together account for the remaining 28% of all identified repeats, are more commonly associated with local sequence variation at finer scales. The relatively low abundance of these small-scale repetitive elements suggests a reduced propensity for localized instability, although this does not preclude the extensive macro-rearrangements driven by the abundant dispersed repeats described above. These two classes of repeats thus contribute to mt genome evolution at distinct structural levels—dispersed repeats shaping global genome architecture through homologous recombination, and SSRs/tandem repeats influencing local sequence dynamics. Furthermore, dispersed repeats may provide recombination sites for IGT, potentially correlating with the six mt-cp homologous segments identified in this study.

In summary, the R. simsii var. putuoense mt genome exhibits typical characteristics of angiosperm mitochondria in terms of RNA editing, codon usage, and repeat structure. Simultaneously, it displays regulatory features potentially related to its coastal habitat in specific details, such as changes in post-editing amino acid properties and the predominance of dispersed repeats. The combination of RNA editing, codon bias, and repeat-mediated recombination constitutes a multi-layered regulatory diversity that may facilitate phenotypic plasticity in response to coastal environmental fluctuations. These post-transcriptional regulatory mechanisms collectively ensure the accuracy and stability of mt gene expression, providing a vital foundation for the proper assembly and function of respiratory chain complexes.

4.2. Evolutionary pressure and habitat adaptation of the mtDNA

In angiosperm mt genomes, matR is the only retained intron-encoded maturase gene. Although its reverse transcriptase and endonuclease domains have degenerated during evolution, it maintains a highly conserved X domain (Brown et al., 2014). Despite this structural reduction, matR remains widespread across land plants, and its transcripts undergo extensive RNA editing that restores essential amino acid residues, suggesting that it still plays a critical biological role in plant mt functions. For instance, it is required for the splicing of various group II introns (Sultan et al., 2016). Notably, the utility of matR as a robust phylogenetic marker is well-established in angiosperms and other seed plants. Previous studies have shown that, compared to plastid genes like rbcL and atpB, the slowly evolving matR gene produces substitutions that are less homoplasious yet highly informative (Zhu et al., 2007). The positive selection on matR observed in this study could be interpreted as an instance of adaptive molecular diversity arising from habitat-specific pressures, contributing to the overall genomic diversity of the genus. Our Ka/Ks analysis revealed that while the majority of PCGs in the R. simsii var. putuoense mtDNA are under purifying selection to maintain core physiological stability, the matR gene exhibited a Ka/Ks ratio greater than 1. Elevated Ka/Ks ratios in organellar genes have also been observed in other Rhododendron species adapted to harsh habitats (Lyu et al., 2024). However, this gene-wide result should be interpreted cautiously, as a pairwise Ka/Ks > 1 averages rates across all codons and cannot definitively prove positive selection or account for genetic drift. In contrast, linking this signature directly to coastal adaptations (e.g., salinity or wind stress) remains speculative when based on a single variety. Rather than conclusive proof, the elevated Ka/Ks in matR is a preliminary hint of lineage-specific divergence. Rigorous confirmation requires site-specific codon models, population-level sampling, and functional assays. Overall, these findings provide a valuable baseline for future ecological and functional studies.

4.3. Interorganellar horizontal gene transfer and its evolutionary significance

In eukaryotic cells, DNA transfer between mt and cp is a major driving force for plant genome evolution, a process known as IGT (Leister, 2005; Kleine et al., 2009). It represents a specific form of intracellular horizontal gene transfer (HGT) occurring at the organellar level (Filip and Skuza, 2021). Compared to some species with massive plastid-to-mitochondrion DNA transfers (Park et al., 2020), the 0.91% proportion in R. simsii var. putuoense suggests a relatively stable genomic boundary, possibly reflecting a lineage-specific pattern of organellar co-evolution. Notably, among the six identified homologous segments, five are complete tRNA genes. This high enrichment of tRNA genes is not random but reflects a functional compensation mechanism of the mt genome. Since native mt tRNA genes are frequently lost during evolution, acquiring cp-derived tRNAs through IGT can maintain the integrity of the mt translation system (Gualberto and Newton, 2017). These transferred tRNA segments retain structural integrity in the R. simsii var. putuoense mtDNA, suggesting they may remain transcriptionally active and participate in mt protein synthesis. In addition to tRNAs, a partially transferred ribosomal RNA gene (rrn18) was also identified. Such large-segment transfers typically involve complex genomic recombination events. These homologous sequences not only increase the size of the mt genome but may also serve as sites for non-homologous recombination, further driving structural variations in the mt genome. Combined with the significant structural differences identified in the synteny analysis (Figure 8), it can be inferred that these cross-organellar sequence transfers are one of the important drivers leading to the high polymorphism and structural complexity of Rhododendron mtDNA.

4.4. Phylogenetic relationships and taxonomic status

Given the high species diversity and frequent hybridization within Rhododendron, which complicate its taxonomic classification, mtDNA sequences provide a valuable tool for phylogenetic reconstruction. In this study, a ML phylogenetic tree was reconstructed based on complete mt genome sequences. The results showed that R. simsii var. putuoense clustered with R. simsii and R. x pulchrum into a single clade with 100% bootstrap support. This outcome provides molecular-level support for the taxonomic treatment of R. simsii var. putuoense as a variety of R. simsii. This finding is also congruent with previous mtDNA studies (Shen et al., 2022), which reported that R. simsii is the sister species to R. × pulchrum based on mt genome phylogeny. However, given that the mtDNA is typically uniparentally inherited, this evidence primarily reflects maternal lineage and is insufficient to definitively confirm the parental origin of R. x pulchrum. Future studies incorporating biparental nuclear markers are necessary to provide a more comprehensive understanding of its hybrid parentage. Synteny analysis (Figure 8) further supports this phylogenetic relationship. R. simsii var. putuoense exhibited the highest synteny proportion with R. simsii, followed by R. x pulchrum, indicating a high degree of genomic structural conservation among the three. In contrast, the synteny proportion between R. simsii var. putuoense and R. nivale subsp. boreale was significantly lower, reflecting substantial genomic rearrangement and evolutionary divergence, which is consistent with their positions in different sub-clades on the phylogenetic tree. Taken together, the mitogenomic evidence provides a new perspective on the evolutionary divergence between R. simsii var. putuoense and its close relative, R. simsii. While their high sequence similarity confirms a recent common ancestry, the distinct selective signatures, particularly the adaptive evolution of the matR gene discussed in Section 4.2, highlight a clear trajectory of niche differentiation. This molecular divergence aligns with the specialized coastal island habitat of R. simsii var. putuoense, suggesting that intense environmental pressures have bypassed neutral genetic drift to drive functional adaptations in the mt genome. Consequently, these findings suggest that R. simsii var. putuoense has developed a unique genetic identity at the organellar level, supporting its status as a distinct evolutionary unit shaped by ecological specialization. Understanding the mitogenomic diversity of R. simsii var. putuoense not only clarifies its evolutionary distinctiveness but also underscores the importance of preserving genetically unique populations for maintaining regional biodiversity. However, the taxon sampling of Rhododendron mt genomes in this study was limited, particularly regarding R. simsii varieties and closely related taxa. Future studies should broaden sampling to rigorously resolve the taxonomic status of R. simsii var. putuoense and clarify its evolutionary relationships within the genus.

5. Conclusions

In this study, we report the first complete assembly and annotation of the R. simsii var. putuoense mt genome, unveiling a multi-branched conformational architecture composed of two linear subgenomic molecules totaling 658,161 bp in length. The genome exhibits typical characteristics of angiosperm mt genomes in gene composition, RNA editing, codon usage bias, and repetitive sequence distribution. Importantly, it also reveals evolutionary dynamics, including IGT and sequence divergence in the matR gene, which provide preliminary insights into the mitogenomic variability of this coastal endemic taxon. Phylogenetic and synteny analyses further confirm its close relationship with R. simsii, while the distinct selective pressures reveal a trajectory of ecological specialization. In summary, our findings not only illuminate the structural and evolutionary complexity of this mt genome but also highlight the conservation value of lineage-specific genomic diversity, providing a critical genetic foundation for preserving the adaptive potential of this variety and guiding future explorations of cyto-nuclear coevolution within the genus Rhododendron.

Acknowledgments

We thank Genepioneer Biotechnologies Co.,Ltd. (Nanjing, China) for providing genomic sequencing and bioinformatics analysis.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This research was supported by Zhejiang Provincial Natural Science Foundation of China (LQ24C030002).

Footnotes

Edited by: Changwei Bi, Nanjing Forestry University, China

Reviewed by: Wenbo Xu, China Pharmaceutical University, China

Arief Priyadi, National Research and Innovation Agency (BRIN), Indonesia

Data availability statement

The datasets generated during and/or analyzed during the current study are available in the NCBI repository (PZ910021-PZ910022).

Author contributions

HZ: Conceptualization, Formal analysis, Funding acquisition, Validation, Visualization, Writing – original draft, Writing – review & editing. CY: Formal analysis, Investigation, Writing – review & editing.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. During the preparation of this manuscript, the authors used generative AI tools (specifically, large language models) solely for language polishing, grammar correction, and overall readability improvement. No AI was used for data generation, scientific interpretation, hypothesis formulation, experimental design, or drawing conclusions. All authors have thoroughly reviewed the final content and take full responsibility for its accuracy and originality.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

  1. Altschul S. F., Gish W., Miller W., Myers E. W., Lipman D. J. (1990). Basic local alignment search tool. J. Mol. Biol. 215, 403–410. doi:  10.1016/S0022-2836(05)80360-2 [DOI] [PubMed] [Google Scholar]
  2. Alverson A. J., Zhuo S., Rice D. W., Sloan D. B., Palmer J. D. (2011). The mitochondrial genome of the legume Vigna radiata and the analysis of recombination across short mitochondrial repeats. PloS One 6, e16404. doi:  10.1371/journal.pone.0016404 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bi C., Bin Z., Jiang A., Qu J., Hao X., Nong L., et al. (2026). Compact multipartite mitogenome of Clausena lansium ‘Ji Xin’ highlights structural diversity within Rutaceae. Int. J. Mol. Sci. 27, 6984. doi:  10.3390/ijms27156984 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bi C., Shen F., Han F., Qu Y., Hou J., Xu K., et al. (2024). PMAT: an efficient plant mitogenome assembly toolkit using low-coverage HiFi sequencing data. Hortic. Res. 11, uhae023. doi:  10.1093/hr/uhae023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Binder S., Brennicke A. (2003). Gene expression in plant mitochondria: transcriptional and post-transcriptional control. Philos. Tran R. Soc B. Biol. Sci. 358, 181–189. doi:  10.1098/rstb.2002.1179 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brown G. G., Colas F.-S. C., Ostersetzer-Biran O. (2014). Group II intron splicing factors in plant mitochondria. Front. Plant Sci. 5, 35. doi:  10.3389/fpls.2014.00035 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Capella-gutiérrez S., Silla-martínez J. M., Gabaldón T. (2009). trimAl: a tool for automated alignment trimming in large-scale phylogenetic analyses. Bioinformatics 25, 1972–1973. doi:  10.1093/bioinformatics/btp348 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Edera A. A., Small I., Milone D. H., Sanchez-puerta M. V. (2021). Deepred-Mt: Deep representation learning for predicting C-to-U RNA editing in plant mitochondria. Comput. Biol. Med. 136, 104682. doi:  10.1016/j.compbiomed.2021.104682 [DOI] [PubMed] [Google Scholar]
  9. Filip E., Skuza L. (2021). Horizontal gene transfer involving chloroplasts. Int. J. Mol. Sci. 22, 4484. doi:  10.3390/ijms22094484 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gualberto J. M., Newton K. J. (2017). Plant mitochondrial genomes: dynamics and mechanisms of mutation. Annu. Rev. Plant Biol. 68, 225–252. doi:  10.1146/annurev-arplant-043015-112232 [DOI] [PubMed] [Google Scholar]
  11. Gui L., Zhang Z., Song L., Feng C., Yu H., Pan L., et al. (2025). Mitogenome of Uncaria rhynchophylla: genome structure, characterization, and phylogenetic relationships. BMC Genom. 26, 199. doi:  10.1186/s12864-025-11372-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hu J., Fan J., Sun Z., Liu S. (2020). NextPolish: a fast and efficient genome polishing tool for long-read assembly. Bioinformatics 36, 2253–2255. doi:  10.1093/bioinformatics/btz891 [DOI] [PubMed] [Google Scholar]
  13. Jobson R. W., Qiu Y. L. (2008). Did RNA editing in plant organellar genomes originate under natural selection or through genetic drift? Biol. Direct 3, 43. doi:  10.1186/1745-6150-3-43 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Katoh K., Standley D. M. (2013). MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30, 772–780. doi:  10.1093/molbev/mst010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kleine T., Maier U. G., Leister D. (2009). DNA transfer from organelles to the nucleus: the idiosyncratic genetics of endosymbiosis. Annu. Rev. Plant Biol. 60, 115–138. doi:  10.1146/annurev.arplant.043008.092119 [DOI] [PubMed] [Google Scholar]
  16. Kumar S., Stecher G., Tamura K. (2016). MEGA7: Molecular evolutionary genetics analysis version 7.0 for bigger datasets. Mol. Biol. Evol. 33, 1870–1874. doi:  10.1093/molbev/msw054 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kurtz S., Phillippy A., Delcher A. L., Smoot M., Shumway M., Antonescu C., et al. (2004). Versatile and open software for comparing large genomes. Genome Biol. 5, R12. doi:  10.1186/gb-2004-5-2-r12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Leister D. (2005). Genomics-based dissection of the cross-talk of chloroplasts with the nucleus and mitochondria in Arabidopsis. Gene 354, 110–116. doi:  10.1016/j.gene.2005.03.039 [DOI] [PubMed] [Google Scholar]
  19. Li G. Y., Chen Z. H., Hu J. F., Jin S. H., Ma D. D., Ou Y. D. (2010). Two new plant varieties from Putuo Island, Zhejiang Province. J. Zhejiang A&F Univ. 27, 908–909. doi:  10.11833/j.issn.2095-0756.2010.06.017 [DOI] [Google Scholar]
  20. Li H. (2018). Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics 34, 3094–3100. doi:  10.1093/bioinformatics/bty191 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Librado P., Rozas J. (2009). DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25, 1451–1452. doi:  10.1093/bioinformatics/btp187 [DOI] [PubMed] [Google Scholar]
  22. Lyu Z., Yang G., Zhou X., Wang S., Zhang R., Shen S. (2024). Deciphering the complex organelle genomes of two Rhododendron species and insights into adaptive evolution patterns in high-altitude. BMC Plant Biol. 24, 1054. doi:  10.1186/s12870-024-05761-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ma W. B., Ou Y., Dayananda B., Ji H. J., Yu T. (2024). The complete chloroplast genome of Rhododendron ambiguum and comparative genomics of related species. Comp. Cytogenet. 18, 143–159. doi:  10.3897/compcytogen.18.119929 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mo Z. Q., Fu C. N., Zhu M. S., Milne R. I., Yang J. B., Cai J., et al. (2022). Resolution, conflict and rate shifts: insights from a densely sampled plastome phylogeny for Rhododendron (Ericaceae). Ann. Bot. 130, 687–701. doi:  10.1093/aob/mcac114 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Møller I. M., Rasmusson A. G., Van aken O. (2021). Plant mitochondria – past, present and future. Plant J. 108, 912–959. doi:  10.1111/tpj.15495 [DOI] [PubMed] [Google Scholar]
  26. Nolfi-donegan D., Braganza A., Shiva S. (2020). Mitochondrial electron transport chain: Oxidative phosphorylation, oxidant production, and methods of measurement. Redox Biol. 37, 101674. doi:  10.1016/j.redox.2020.101674 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Palmer J. D., Adams K. L., Cho Y., Parkinson C. L., Qiu Y., Song K. (2000). Dynamic evolution of plant mitochondrial genomes: Mobile genes and introns and highly variable mutation rates. Proc. Natl. Acad. Sci. U.S.A. 97, 6960–6966. doi:  10.1073/pnas.97.13.6960 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Park H. S., Jayakodi M., Lee S. H., Jeon J. H., Lee H. O., Park J. Y., et al. (2020). Mitochondrial plastid DnA can cause DNA barcoding paradox in plants. Sci. Rep. 10, 6112. doi:  10.1038/s41598-020-63233-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Shen J., Li X., Li M., Cheng H., Huang X., Jin S. (2022). Characterization, comparative phylogenetic, and gene transfer analyses of organelle genomes of Rhododendron × pulchrum. Front. Plant Sci. 13, 969765. doi:  10.3389/fpls.2022.969765 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Shrestha N., Wang Z., Su X., Xu X., Lyu L., Liu Y., et al. (2018). Global patterns of Rhododendron diversity: The role of evolutionary time and diversification rates. Global Ecol. Biogeogr. 27, 913–924. doi:  10.1111/geb.1275040046247 [DOI] [Google Scholar]
  31. Smith D. R. (2016). The mutational hazard hypothesis of organelle genome evolution: 10 years on. Mol. Ecol. 25, 3769–3775. doi:  10.1111/mec.13742 [DOI] [PubMed] [Google Scholar]
  32. Sultan L. D., Mileshina D., Grewe F., Rolle K., Abudraham S., Głodowicz P., et al. (2016). The reverse transcriptase/RNA maturase protein matR is required for the splicing of various group II introns in Brassicaceae mitochondria. Plant Cell 28, 2805–2829. doi:  10.1105/tpc.16.00398 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wang D., Zhang Y., Zhang Z., Zhu J., Yu J. (2010). KaKs_Calculator 2.0: a toolkit incorporating gamma-series methods and sliding window strategies. Genom. Proteom Bioinf. 8, 77–80. doi:  10.1016/S1672-0229(10)60008-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Wick R. R., Schultz M. B., Zobel J., Holt K. E. (2015). Bandage: interactive visualization of de novo genome assemblies. Bioinformatics 31, 3350–3352. doi:  10.1093/bioinformatics/btv383 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Zeng Z., Zhang Z., Tso N., Zhang S., Chen Y., Shu Q., et al. (2024). Complete mitochondrial genome of Hippophae tibetana: insights into adaptation to high-altitude environments. Front. Plant Sci. 15, 1449606. doi:  10.3389/fpls.2024.1449606 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Zhang D., Gao F., Jakovlić I., Zou H., Zhang J., Li W. X., et al. (2020). PhyloSuite: An integrated and scalable desktop platform for streamlined molecular sequence data management and evolutionary phylogenetics studies. Mol. Ecol. Resour. 20, 348–355. doi:  10.1111/1755-0998.13096 [DOI] [PubMed] [Google Scholar]
  37. Zhang C. L., Tu X. D., Wang L. Y., Chen D. Q., Zeng M. Y., Yin W. L., et al. (2025). A comprehensive analysis of the complete Dendrobium chrysotoxum mitogenome reveals a multiple-circular structure and horizontal gene transfer from fungi. Ornam Plant Res. 5, e007. doi:  10.48130/opr-0025-0009 [DOI] [Google Scholar]
  38. Zhou D., Wang J., Zhou F., Li Z., Qi Y., Hu T. (2024). The complete chloroplast genome of Rhododendron williamsianum (ericaceae). Mitochondrial DNA B. 9, 1058–1062. doi:  10.1080/23802359.2024.2391085 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zhu X. Y., Chase M. W., Qiu Y. L., Kong H. Z., Dilcher D. L., Li J. H., et al. (2007). Mitochondrial matR sequences help to resolve deep phylogenetic relationships in rosids. BMC Evol. Biol. 7, 217. doi:  10.1186/1471-2148-7-217 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhu H., Gao H., Li H. (2026). Chromosome-level genome and population genomics reveal demographic history, incomplete divergence and coastal adaptation of Rhododendron simsii var. putuoense in East China. Plant Divers. 48, 635–639. doi:  10.1016/j.pld.2025.12.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zhu H., Li D., Yue C., Li H. (2025. a). Development of single nucleotide polymorphism and phylogenetic analysis of Rhododendron species in Zhejiang Province, China, using ddRAD-seq technology. Plants 14, 1548. doi:  10.3390/plants14101548 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhu H., Lv X. (2025). The first complete chloroplast genome and phylogenetic analysis of Rhododendron simsii var. putuoense G. Y. Li & Z. H. Chen (Ericaceae) from East China. Mitochondrial DNA B. 10, 1216–1220. doi:  10.1080/23802359.2025.2590214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhu H., Yue C., Li H. (2025. b). Mitochondrial genome characteristics and comparative genomic analysis of Spartina alterniflora. Curr. Issues Mol. Biol. 47, 107. doi:  10.3390/cimb47020107 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available in the NCBI repository (PZ910021-PZ910022).

Articles from Frontiers in Plant Science are provided here courtesy of Frontiers Media SA

RESOURCES