Characterization and Phylogenetic Analysis of the Complete Mitogenomes of Valsa mali and Valsa pyri

TL;DR

This study reports the first complete mitochondrial genomes of two apple canker-causing fungi, revealing insights into their evolution and genetic differences.

Contribution

The first characterization of mitogenomes in the genus Valsa, providing new data on their structure, composition, and evolutionary relationships.

Findings

The mitogenomes of Valsa mali and Valsa pyri are circular DNA molecules with distinct sizes and base composition.

Significant differences in intronic regions and repetitive sequences were found between the two species.

Phylogenetic analysis confirmed V. mali and V. pyri as independent taxa with strong statistical support.

Abstract

Apple Valsa canker, caused by Valsa mali and Valsa pyri, is a devastating disease of apple trees and poses a severe threat to the sustainable development of apple production. Although the two species’ whole genomes have been sequenced, their mitochondrial genomes are still uncharacterized. In this study, the complete mitochondrial genomes of V. mali and V. pyri were assembled, annotated, and compared by bioinformatic methods. The results indicate that the mitogenomes are both circular DNA molecules with sizes of 213,406 bp and 128,022 bp, respectively. The AT skew values of the two Valsa species’ mitogenomes were positive, while the GC skew values were negative. Comparative mitogenome analysis revealed that the length and base composition of protein-coding genes (PCGs), rRNA genes, and tRNA genes differed between the two Valsa species. It was found that the expansion of V. mali was primarily attributable to the intronic regions. There are large numbers of interspersed repetitive sequences (IRS) in both Valsa mitogenomes; however, the proportion of IRS in V. mali (43.56%) was much higher than that in V. pyri (2.41%). The alignment of large fragments between the mitochondrial and nuclear genomes of both V. mali (1.73 kb) and V. pyri (5.17 kb) indicates that gene transfer between mitochondrial and nuclear genomes occurred during evolution. The ka/ks ratios for 15 core PCGs were below one, suggesting that these genes were subjected to purifying selection pressure. Comparative mitogenomics revealed that the two fungi had significant mitogenomic collinearity and large-scale gene rearrangements. The results of phylogenetic analysis based on Bayesian inference (BI) and maximum likelihood (ML) using a combined mitochondrial gene set confirmed that V. mali and V. pyri were fully independent taxa with a high bootstrap value of 100 (ML) and a high posterior probability of 1.0 (BI). This is the first report on the mitogenomes within the genus Valsa. These results will pave the way to understanding the evolution and differentiation of mitogenomes in the genus Valsa.

Figures

Full text

1. Introduction

Apple Valsa canker is one of the most economically important and destructive diseases in apple-producing areas of China [1,2,3], Japan [4,5], Korea [6], and the far eastern region of Russia [6]. The pathogen was initially identified as Valsa mali Miyabe et Yamada [7] and was later classified as a synonym of V. certosperma [8] Maire [9]. However, it was recently proven, based on the evidence from morphological characterization and molecular phylogenetic analysis, that V. mali is an evolutionarily independent species that is distinct from V. certosperma [3,10]. Furthermore, numerous cryptic divergences have been detected among V. mali isolates from apple (Malus domestica Borkh) and pear (Pyrus communis) [2]. Lu (1992) found that V. mali strains from pear were significantly different from those derived from apple on isozyme electrophoresis, although they have similar characteristics in culture [11]. Consequently, the V. mali strains from pear were classified as an independent variety and named V. mali var. pyri (Vmp), while the strains from apple were defined as V. mali var. mali (Vmm) for easier distinction [2,10].

In addition, an increasing amount of evidence supports the classification of the two varieties as independent species. They have been respectively reclassified as V. mali and V. pyri based on the combined sequences of rDNA-ITS, β-tubulin, and EF1α [2]. The genetic divergence in the rDNA-ITS sequence was lower, with a value of 1.4%, while in contrast, the divergences in the β-tubulin and EF1α sequences were significantly higher, with mean pairwise distances of 8.0% and 5.1%, respectively [2]. Furthermore, they exhibit distinct characteristics in culture on PDA plates and possess varying abilities to resist higher temperatures. The V. pyri colonies are almost always milky white, while V. mali colonies vary from white to light brown. The former grows normally on PDA plates at 37 °C, whereas the latter cannot grow at this temperature. More importantly, it is very interesting that V. mali strains are more aggressive on apple trees than on pear trees, while V. pyri exhibits higher virulence on pear trees than on apple trees. Finally, the whole-genome sequences revealed that the two species differ in size: the genome size of V. mali is 44.7 Mb, while the genome size of V. pyri is considerably smaller, at only 35.7 Mb [12]. The V. mali genome also contains a significantly greater number of repetitive elements, although the two species have nearly equal numbers of the genes associated with pathogenicity, secreted proteins, and proteases. The major difference between the two species lies in secondary metabolite gene clusters. Although many differences have been detected between the two species and phylogenetic analysis has been carried out based on multiple loci, including rDNA-ITS, β-tubulin, and EF1α, a more precise phylogenetic analysis at a broader scale, such as at the level of the mitogenome, has not yet been performed.

It is well known that the mitochondrion is an important double-membrane organelle in eukaryotic cells that is primarily responsible for producing adenosine triphosphate (ATP), the energy currency in living organisms [13]. Moreover, the mitochondrion is a semi-autonomous organelle in eukaryotic cells because mitochondria carry their own genetic material and possess the components for protein synthesis [14]. Beyond energy metabolism, the mitochondria play a very important role in other cell functions, such as Fe/S-cluster biosynthesis and amino acid metabolism, and they are also closely associated with apoptosis, cell senescence, virulence, and drug resistance [15,16,17]. Fungal mitogenomes usually contain 14 core protein-coding genes (PCGs), including three encoding ATP synthase subunits (apt6, apt6, and atp9), seven encoding subunits of reduced nicotinamide adenine dinucleotide ubiquinone oxidoreductase (nad1, nad2, nad3, nad4, nad4L, nad5, and nad6), three encoding subunits of cytochrome c oxidase (cox1, cox2, and cox3), and one encoding apocytochrome b oxidoreductase (cob) [14,17]. Moreover, the fungal mitogenome also encodes a ribosomal protein (rps3) and the RNA component for RNAse P (rnpB gene), along with two ribosomal RNA (rRNA) genes and 20~31 transfer RNA (tRNA) genes [18].

It is widely accepted that the mitogenome is a powerful tool for phylogenetic analysis in fungi due to its faster rate of evolution rate compared to the nuclear genome, as well as its lower recombination rate, conserved gene content, and uniparental inheritance; this finding has also been successfully leveraged in phylogenetic studies [19,20,21,22,23]. In addition, the relatively small size, circular-mapping topology, and multicopy nature of mitogenomes make their sequencing, assembly, and comparative analysis much easier than these are for the nuclear genome [17]. Recently, with the development of next-generation high-throughput sequencing technology, as well as the declining cost of sequencing, upgrades to related bioinformatics software, and algorithm optimization, more and more fungal mitogenomes have been sequenced, assembled, and annotated. According to the statistics, to date, a total of 3331 (accessed date: 28 March 2025) complete fungal mitogenome sequences have been submitted to the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov/genome/browse#!/organelles/). However, there is no available mitogenome information on Valsa species, and this has become a hindrance to fully understanding the important branch disease caused by the pathogenic Valsa fungi.

It has been reported that the size of the fungal mitogenome varies widely even within a genus. Kanzi et al. (2016) found that the mitogenome size of Chrysoporthe deuterocubensis was 124,412 bp and that of C. austroafricana was 190,834 bp, while that of C. cubensis was only 89,084 bp [24]. This variability can be attributed to differences in the length and number of introns, intergenic regions, repetitive DNA, open reading frames (ORFs) without defined functions, and homing endonuclease genes [19,25,26,27,28]. Among these factors, differences in the number and length of introns are considered to be the main ones contributing to the expansion or contraction of fungal mitochondrial genomes [17,29]. The introns in the fungal mitogenome can be divided into two groups, named group I and group II [13]. The introns in both groups are self-splicing DNA sequences [17]. The group I introns are the most numerous and can encode homing endonucleases [30,31]. In contrast, group II introns encode mostly reverse transcriptase-like ORFs [30]. However, to date, there have been no reports on mitochondrial introns in the genus Valsa.

Therefore, in this study, the mitogenomes of causal agents of apple tree Valsa canker, specifically V. mali and V. pyri, were assembled, annotated, and compared with the mitogenomes of other related pathogens. The principal aims of this study are as follows: (i) to reveal the mitogenome features of the two Valsa species; (ii) to investigate interspecific mitogenome variation between the two Valsa species; (iii) to clarify the phylogenetic position of the two Valsa species based on the datasets containing the mitogenome-encoded genes. This study represents the first report on mitogenomes of Valsa species. It will provide foundational reference sequences for further investigations of Valsa mitochondrial genomes and pave the way for a deeper understanding of the genetic evolution and population genetics of Valsa, as well as of species differentiation within the genus in the future.

2. Materials and Methods

2.1. Sample Collection and Fungal Pure Culture Obtain

Apple branches or twigs with typical canker symptoms were collected from a commercial orchard in Lin County, Henan Province, China. Pure cultures were obtained from the tissues located between the diseased tissues and healthy areas using the tissue-isolation methods described by Wang et al. (2011) and Zang et al. (2012) [10,32]. A total of 24 isolates were obtained. The isolates were identified at the species level based on morphological characteristics, sequence analysis of rDNA-ITS, and a PCR protocol utilizing species-specific primers developed by Zang et al. (2012) [32]. The obtained pure-culture isolates were stored at −80 °C in the ultra-low-temperature freezer (MECCAN DW-HL530, Hefei, China) at the Fungal Institute of Henan Agricultural University. The V. pyri strain was purchased from the Agricultural Culture Collection of China (ACCC36131).

2.2. Fungal DNA Extraction and Sequencing

Mycelium cakes were first obtained from the margins of fungal colonies that had been cultured at 25 °C for three days and then transferred onto the surface of potato dextrose medium (PDA) media, which was overlaid with a thin layer of sterile cellophane. The fungal total genomic DNA was extracted by the sodium dodecyl sulfate (SDS) method. The mycelia were harvested and ground into a powder with a sterile mortar under liquid nitrogen. The powder was suspended in 800 μL SDS extraction buffer, which contains 3% SDS, 50 mM EDTA, and 100 mM Tris-HCl (pH 8.0), transferred to a 2 mL-PCR tube, and then incubated at 65 °C for 1 h. Subsequently, an equal volume of phenol/chloroform/isoamyl alcohol (25:24:1) was added to the tubes, which were then gently inverted several times. The tubes were centrifuged at 4 °C and 12,000 rpm for 15 min. The supernatant (600 μL) was transferred to a new sterile 1.5 mL tube and eluted with an equal volume of chloroform. After centrifugation at 4 °C, 12,000 rpm for 10 min, the aqueous phase (350 μL) was collected and DNA was precipitated in isopropanol (250 μL). Finally, the mixed liquor was centrifuged at 4 °C and 12,000 rpm for 15 min, and the sedimented DNA could be seen at the bottom of the tubes. The supernatant was removed, and the DNA was washed twice with 75% ice-cold ethyl alcohol (500 μL) and then once with 100% ice-cold ethyl alcohol. A volume of 30 μL sterile deionized water was added dissolve the DNA after the ethyl alcohol had evaporated. The quality of the harvested DNA was qualitatively evaluated by agarose gel electrophoresis and quantified on the Qubit 2.0 fluorometer system (Thermo Scientific, Waltham, MA, USA). High-quality DNA was sent to the Novogene Company (Cambridge, UK), where the 350 bp sequencing library was constructed for short-read Illumina sequencing, and another 10K sequencing library was also constructed for long-read Nanopore sequencing.

2.3. Mitogenome Assembly and Annotation

The whole-genome sequencing was performed on the Illumina NovaSeq PE150 and Nanopore PromethION platforms (Illumina, San Diego, CA, USA). The adapters and low-quality short reads were filtered using Fastp v0.12.4 software [33]. The obtained cleaned paired-end data were used to assemble the complete mitogenome with NOVOPlasty v4.3.1 or Get_organelle v1.7.5 software. The cox1sequence of Diaporthe longicolla was used as the seed sequence in NOVOPlasty analysis. The parameters of Get_organelle were set as follows: k-mers 21,45,65,85,105, --max-rounds 15, -F fungus_mt.

If the assembly results could not be circularized using the two software programs, the Nanopore long-reads were assembled using Minimap2.1 and Racon 1.5.0. The mitochondrial-related contigs were identified by performing a Blastn search against the mitogenome sequence of Diaporthe longicolla, which served as the reference sequence. The software program bwa 0.7.17-r1188 was used to align the sequenced reads to the assembled mitogenome [34]. Samtools (v1.15.1) was used to convert the SAM file to a sorted BAM file.

The complete mitogenomes of both V. mali and V. pyri were annotated by combining the results of the MFannot website (https://megasun.bch.umontreal.ca/apps/mfannot/, accessed on 12 March 2024) annotation and MITOS WebServer (http://mitos2.bioinf.uni-leipzig.de/index.py, accessed on 12 March 2024) annotation using genetic code 4 (The Mold, Protozoan, and Coelenterate Mitochondrial Code and the Mycoplasma/Spiroplasma code). The annotation results of MFannot were checked by GeSeq [35] and then manually checked to confirm the gene boundaries [22]. The tRNA secondary structures were predicted by MITOS with default parameters and redrawn with Adobe Illustrator CS6. The graphical maps of the two Valsa mitogenomes were drawn using the Organellar Genome Draw Maps (OGDRAW) v.1.3.1 tool [36].

2.4. The Sequences Analysis

The base compositions of the two Valsa mitogenomes were analyzed using BioEdit 7.2.5. Strand asymmetry was assessed according to the following formulas: AT skew = [A − T]/[A+T], and GC skew = [G − C]/[G+C] [37]. MegaX and KaKs_Calculator2.0 were used to calculate the nonsynonymous substitution rates (Ka) and the synonymous substitution rates (Ks) for the 15 core PCGs in the two Valsa mitogenomes [38,39]. The Codon Usage module was used to analyze codon-usage bias in the Sequence Manipulation Suite (https://www.bioinformatics.org/sms2/codon_usage.html, accessed on 28 March 2024), based on genetic code 4 [40]. Intronic pairs were examined by EMBOSS Stretcher global alignment, and the intron–exon borders of PCGs were checked using exonerate 2.4.0 [41]. Genome syntenies of the two Valsa mitogenomes and representative species in other genera were analyzed using Mauve v2.4.0 [42].

2.5. The Repetitive Elements Analysis

The BLASTn searches of the whole mitogenomes against themselves were performed to detect whether there were interspersed repeats or intragenomic duplications of large fragments based on an E-value of 1 × 10^−10^. Simple repeat sequences (SSRs) were identified using the MISA-web microsatellite identification tools website (https://webblast.ipk-gatersleben.de/misa/, accessed on 9 April 2024). The tandem repeats in the two Valsa mitogenomes were searched using the Tandem Repeats Finder with the default parameters. Duplication events between the mitogenomes and nuclear genomes were detected using BLAST package 2.9.0+ [43].

2.6. Phylogenetic Tree Construction and Phylogenetic Analysis

To investigate the phylogenetic positions of V. mali and V. pyri in the order Diaporthales, a phylogenetic tree was constructed with 92 Ascomycetes species using the concatenated mitochondrial gene set, which included 14 core PCGs and the rps3 gene. Taphrina deforman and T. wiesneri from Taphrinomycetes were selected as outgroups (Table S7). The single mitochondrial genes were selected using PhyloSuite v1.2.2 and aligned using the MAFFT plug-in in this software [44]. The aligned sequences were concatenated using the Concatenate Sequence module in this software to obtain a combined mitochondrial gene set. Partition-Finder 2.1.1 was applied to determine the best-fit models of evolution and the partitioning scheme for the gene set [45]. The phylogenetic tree was constructed using the maximum likelihood (ML) method and the Bayesian inference (BI) method. The BI analysis was carried out using MrBayes 3.2.6 software [46]. Two independent runs with four Markov chains each were conducted for 3,000,000 generations. Each run was sampled every 100 generations. The first 25% of samples were discarded as burn-in, and the remaining trees were used to calculate Bayesian posterior probabilities (BPP) in a 50% majority-rule consensus tree [47]. The software iqtree (v2.1.3) was used to construct the ML phylogenetic tree based on the combined gene set, and the bootstrap values (BS) were calculated using 1000 replicates to assess the node support [48].

3. Results

3.1. The Mitogenome Size and Organization Characteristics

A total of 5.49 G raw data from short reads and 7.92 G raw data from long reads were generated for V. mali. After processing, there were 5.55 G and 8.30 G of corresponding sequencing data for V. pyri and V. mali, respectively. The mitogenomes of the two Valsa species, V. mali, and V. pyri, were both assembled to circular DNA, with total sizes of 213,406 bp and 128,022 bp, respectively (Figure 1). It is clear that the V. mali mitogenome is significantly larger than that of V. pyri. The GC content of the V. mali mitogenome was 32.77%, while the value for V. pyri was 31.31%. These mitogenome GC contents were very similar. The AT skew values of two Valsa species mitogenomes was positive, while the GC skew value was negative. (Table S1). Whole sets of core PCGs were detected in the V. mali and V. pyri mitogenomes, including 15 typical core genes (apt6, apt8, apt9, cob, cox1, cox2, cox3, nad1-6, nad4L, and rps3). The V. mali mitogenome contains 35 introns distributed in the nad1, nad2, nad5, cox1, cox2, cox3, and cob genes, of which 33 belonged to group I and two belonged to group II. Similarly, The V. pyri mitogenome contained 22 introns distributed in the atp6, nad1, nad5, cox1, cox2, cox3, and cob genes, and all of them belong to group I. In addition, 54 and 22 free-standing ORFs (non-intron encoding ORFs) were respectively found in the V. mali and V. pyri mitogenomes. Meanwhile, 60 and 33 intronic ORFs encoding LAGLIDADG homing endonucleases, GIY-YIG homing endonucleases, and proteins with unknown functions were detected in the mitogenomes of V. mali and V. pyri, respectively. There was a slight difference in the numbers of ORFs encoding LAGLIDADG endonucleases and GIY-YIG homing endonucleases: 23 and 15, respectively, in the V. pyri mitogenome and 63 and 22, respectively, in the V. mali mitogenome. Notably, the V. mali mitogenome ORFs included almost three times as many ORFs encoding LAGLIDADG endonucleases as ORFs encoding GIY-YIG homing endonucleases (Table S2).

Protein-coding regions occupied the largest proportion of the V. pyri mitogenome, accounting for 34.11% of the mitogenome (Figure 2). The second-largest region was the intronic region, which accounting for 30.28%. The ncRNA (tRNA and rRNA) region and the intergenic region occupied almost the same proportions, respectively 17.91% and 17.7%. By contrast, in the mitogenome of V. mali, the largest region was the intronic region, which occupied 33.82% of the mitogenome. The second-largest region was the protein-coding region, which accounted for 31.65%, and the third-largest region was the ncRNA region, which accounted for 21.04% (Figure 2). Finally, the intergenic region took up the smallest proportion, 13.50%. The mitogenome of V. mali was 85,384 bp larger than that of V. pyri. The four regions played an important role in V. mali mitogenome expansion. The intronic region contributed the most to the V. mali mitogenome expansion, accounting for 39.11% of the difference. In addition, the ncRNA region, the protein-coding region, and the intergenic region also contributed to the expansion of V. mali mitogenome, respectively accounting for 25.72%, 27.95%, and 7.20% (Figure 2).

3.2. rRNA, tRNA, and Codon Usage Analysis

The small subunit ribosomal RNA (rns) and large subunit ribosomal RNA (rnl) were both detected in both Valsa mitogenomes. One and six introns were respectively detected in the rns and rnl of the V. pyri mitogenome, whereas there is no intron was found in these genes in the V. mali mitogenome. In total, 26 and 24 tRNA genes were respectively detected in the mitogenomes of V. mali and V. pyri, with gene lengths ranging from 71 to 85 bp, mainly due to size variation in the extra arms. All tRNAs in the two Valsa mitogenomes can be folded into typical cloverleaf structures because no introns were found in these tRNA genes, which encode twenty kinds of standard amino acids (Figure 3). The two mitogenomes contained two tRNAs with different anticodons coding for serine and arginine and one tRNA with the same anticodon coding for methionine. Almost all tRNA genes were present in the two Valsa mitogenomes as single copies, except for the trnL, trnM, trnR, and trnS genes. trnM was present in four copies in V. mali mitogenome, while it was present in three copies in the V. pyri mitogenome. Additionally, three other tRNA genes (trnL, trnR and trnS) were observed to be present in multiple copies: the trnR gene was present in three copies, while trnL and trnS were present in two copies in each of the two mitogenomes. The V. mali mitogenome contained the trnM-4 and trnW genes, whereas these two genes were absent from the V. pyri mitogenome.

The results of codon-usage analysis showed that ATG was the most commonly used start codon in the core PCGs of the two Valsa mitogenomes. However, the Cox1 gene of the two Valsa species used TTG as a start codon. The nad4 gene of V. mali used TTG as the start codon, while that of V. pyri still used ATG as the start codon. The most frequently used stop codon was TAA, followed by TAG (Table S3). Meanwhile, the results also indicated that the most frequently used codons in the two Valsa mitogenomes were TTT (for phenylalanine; Phe), AAA (for lysine; Lys), TTA (for leucine; Leu), TAT (for tyrosine; Tyr), ATT (for isoleucine; Ile), and AAT (for asparagine; Asn). To some extent, the frequency of A and T used in codons leads to a relatively higher value of AT content in the two Valsa species (AT average: 67.96%) (Figure 4).

3.3. The Repeatitive Elements Analysis

A total of 21 and 1299 interspersed repetitive elements were respectively detected in the mitogenomes of V. pyri and V. mali through BLASTn searches of the two Valsa mitogenomes against themselves. The lengths of repeat sequences in the V. pyri mitogenome ranged from 52 to 418 bp, with pair-wise nucleotide similarities ranging from 79.11% to 96.15%. The largest repeat in V. pyri was located in the intergenic region between nad5 and atp8. The repeat sequences accounted for 2.41% of the V. pyri mitogenome. The length of repeat sequences in the V. mali mitogenome ranged from 37 to 802 bp, with pair-wise nucleotide similarities ranging from 75.12% to 100%. The largest repeats in V. mali were observed in the intergenic region between trnR and orf102. Repeat sequences accounted for 43.56% of the V. mali mitogenome length. The repeat sequence of V. mali was much longer than that of V. pyri (Table S4).

A total of 19 and 33 tandem repeats were detected in the mitogenomes of V. pyri and V. mali, respectively. The longest tandem repeat sequence was found in V. pyri and extended over 99 bp while encompassing two repeat loci. Most tandem repeat sequences were duplicated once or thrice in the two Valsa mitogenomes, with the highest replication number (42) observed in the V. mali mitogenome. However, the longest repeat sequence was composed of 42 adenines (A), and this kind of tandem repeat sequence was absent from the V. pyri mitogenome. Tandem repeat sequences account for 0.72% and 0.10% of the mitogenome lengths of V. pyri and V. mali (Table S5).

To detect whether there were gene segments that had transferred between the nuclear and mitogenomes, the two Valsa mitogenomes were BLASTed against their nuclear genomes. Totals of 13 and 18 aligned fragments were respectively detected in the mitogenomes of V. pyri and V. mali. The lengths of these aligned segments ranged from 48 to 5172 bp, with sequence-identity values ranging from 76% to 100%. The largest fragment was located in the protein-coding regions of cob gene in the V. pyri mitogenome and had a length of 5172 bp. The largest aligned fragment in V. mali was detected between the fifth intron of the nad5 gene and orf103 and had a length of 1719 bp. The presence of large fragments that aligned between the mitochondrial and nuclear genomes indicates that genetic transfer between mitochondrial and nuclear genomes may have occurred during the evolution of Valsa species (Table S6).

3.4. Variation, Genetic Distance and Evolutionary Rates of 15 Core PCGs

Among the 15 core PCGs detected in this study, eight genes including nad1, nad2, nad4, nad5, cob, cox3, atp6, and rps3 were found to differ in length between the two Valsa species. Sequence alignment revealed that these variations predominantly occur at the gene initiation and termination regions. All of these genes except cob were obviously longer in V. pyri than V. mali (Figure 5). The GC contents of eight among the 15 core PCGs differed between the two Valsa species, while the other genes had almost the same GC contents. The atp9 gene had the highest GC content, while atp8 had the lowest GC content. The AT skew values of 14 core PCGs (all except rps3) were negative. Similarly, the GC skew values of 14 core PCGs (all except atp8) were positive (Figure 5).

Among the 15 detected core PCGs, the rps3 gene had the largest average value of Kimura-2-parameter (K2P) genetic distance, followed by nad4, nad6, and nad3, which demonstrates that these genes had differentiated greatly during the evolutionary process (Figure 6). The atp8 gene showed the smallest K2P genetic distance among the six analyzed species (Valsa mali, V. pyri, Chrysoporthe austroafricana, C. cubensis, C. deuterocubensis, and Diaporthe longicolla) in the order Diaporthales, indicating that the gene is highly conserved. The rps3 gene exhibited the largest non-synonymous substitution rate (Ka) among the analyzed PCGs, while atp8 had the smallest Ka value. The synonymous substitution rate (Ks) of the nad4 gene was the largest, while that of the atp8 gene was the smallest among the analyzed species (Figure 6). Finally, the overall Ka/Ks values for all detected core PCGs were <1, indicating that these genes were subjected to purifying selection pressure.

3.5. Mitochondrial Gene Arrangement and Collinearity Analysis in Diaporthales

The mitochondrial gene arrangement, including 15 core PCGs and two rRNA genes of eleven species (Valsa mali, V. pyri, Chrysoporthe austroafricana, C. cubensis, C. deuterocubensis, Diaporthe longicolla, D. phaseolorum, D. sojae, D. caulivora, D. nobilis and D. eres) in the order Diaporthales was analyzed. The results showed that the mitochondrial gene arrangement varied greatly in the order Diaporthales. Large-scale gene rearrangements, including gene relocations and position exchanges, were detected between species in different genera. However, the three examined species of Chrysoporthe and six species of Diaporthe each exhibited completely identical gene-arrangement patterns, suggesting these congeners may share closer phylogenetic relationships within their respective genera. By contrast, large-scale gene rearrangements were also detected between species in the same genus; the two Valsa species had different gene arrangements (Figure 7). This result indicated that the two Valsa species underwent different evolutionary processes.

The mitogenome sizes of the six analyzed species (Valsa mali, V. pyri, Chrysoporthe austroafricana, C. cubensis, C. deuterocubensis, and Diaporthe longicolla) in Diaporthales varied greatly, ranging from 53,439 bp to 213,406 bp, with an average size of 133,199 bp. The mitogenome of V. mali was the largest among the analyzed Diaporthales mitogenomes, at almost four times that the size of the smallest mitogenome. In addition, the results of mitogenome collinearity analysis showed that the two Valsa mitogenomes can be divided into seven homologous regions. The relative positions of these homologous regions were highly variable between the two Valsa species (Figure 8). Homologous region G was absent from the genus Chrysoporthe. However, the relative positions of homologous regions were the same.

3.6. Phylogenetic Relationships Analysis

The phylogenetic relationships analysis was carried out using the maximum likelihood (ML) method and a Bayesian inference (BI) method based on the concatenated mitogenome gene set including 14 core protein-coding genes and the rps3 gene. It generated almost identical and well-supported tree topologies for 92 Ascomycota species (Figure 9). Almost all major branches in the ML tree and BI tree were well supported, with high bootstrap values. The results of the phylogenetic analysis showed that the 92 ascomycetous species can be divided into 17 major clades, corresponding to the orders of Eurotiales, Onygenales, Pleosporales, Botryosphaeriales, Cladosporiales, Mycosphaerellales, Microascales, Diaporthales, Ophiostomatales, Xylariales, Sordariales, Lecanorales, Ostropales, Caliciales, Peltigerales, Erysiphales, and Taphrinales, respectively. The six species within the order Diaporthales could be divided into three groups, wherein the first group was composed of one species in the genus Diaporthe, and the second group consisted of three species in the genus Chrysoporthe with a bootstrap value of 100 and a posterior probability value of 1.00. The two Valsa species are grouped together within the same cluster, with a bootstrap value of 100 and a poster probability value of 1.00. This topological structure shows that the two species have a very close evolutionary relationship; however, they are distinct sister species. This result also indicated that the phylogenetic analysis based on the mitogenome genes was a robust molecular marker by which to analyze the phylogenetic relationships of Ascomycota.

4. Discussion

In the study, the mitogenomes of two Valsa species, V. pyri and V. mali, were assembled and annotated respectively for the first time. The assembled mitogenomes of V. pyri and V. mali were both circularized DNA molecular with a size of 128,022 bp and 213,406 bp, respectively. The mitochondrial genome of V. mali is much bigger than that of V. pyri. It is well known that the size of the fungal mitochondrial genome varies greatly due to introns present within protein-coding genes, intergenic regions, repeat sequences, and plasmid-derived dynamic regions [13,23,26,27]. The V. mali mitogenome contains 35 introns, whereas the V. pyri mitogenome has only 22. In a comparison of the two mitogenomes, the intronic region contributed the most to the V. mali mitogenome expansion. Similar results were also observed in the mitogenomes of Botryosphaeria dothidea, B. kuwatsukai [21], Exserohilum turcicum, and E. rostratum [49]. In addition, the length of the repeat sequence of V. mali is much longer than that of V. pyri, which accounted for 43.56% of the V. mali mitogenome length, while the value in V. pyri is only 2.41%. The repeat sequences also played a very important role in the expansion of V. mali mitogenome. Although the proportion of intergenic regions differed little between the two Valsa mitogenomes, it also contributed to the expansion of the V. mali mitogenome.

The second parity rule states that each base in the complementary DNA strand has almost equal frequencies as long as there is no mutation or selection bias [50]. AT skews and GC skews were respectively detected in V. mali and V. pyri. The presence of AT skews and GC skews in different species demonstrates that mitogenomes of different species underwent different mutations or were subject to different environmental selection pressure. In addition, the Ka/Ks values for all core PCGs in the two Valsa species were <1, which revealed that they were subjected to stronger pressure of purifying selection [51].

The exchange of genetic material between the mitochondrial and nuclear genomes has been proven to be a common occurrence in various fungal species and promotes the differentiation of mitogenomes [37]. Meanwhile, the mitochondrial and nuclear genes work together synergistically to support the growth and development of fungi. It was commonly acknowledged that a majority of mitochondrial genes have been transferred to the nucleus, while some nuclear genes have also been discovered to have transferred to the mitogenome during the long process of evolution [52]. In the present study, it was discovered that more than ten larger fragments aligned, indicating likely transfer between mitochondrial genomes and the nuclear genome in the two Valsa species. This finding suggests that these species have experienced frequent natural gene transfers.

It has been reported that the mitochondrial gene arrangement can be used as an important clue to the phylogenetic relationships and evolutionary histories of eukaryotic species [37,53,54,55]. In the past two decades, with the development of high-throughput sequencing technology, more and more mitogenomes of eukaryotes have been sequenced., Mitochondrial gene rearrangement in animals has been widely studied, and many models associated with it have been established to reveal the mechanisms of mitogenome rearrangement [56]. Mitochondrial gene rearrangements are even more widespread and frequent in fungi than in animals [28,57]. Such gene rearrangements can be observed across the kingdom, with relevant species including but not limited to many mushroom-forming fungi, such as Lyophyllum spp. [47], Pleurotus spp. [58], Cantharellus spp. [59], Russula spp. [60]; plant pathogenic fungi, such as Pseudocercospora fijiensis [8], and Bipolaris oryzae [61]; and entomopathogenic fungi, such as Beauveria caledonica [62]. The gene rearrangements were mainly attritubed to nonhomologous, intrachromosomal recombination and the distribution of tRNA [57]. In the present study, significant mitochondrial gene rearrangement in the two Valsa species involved rRNA genes and core PCGs. The result showed that there are obvious differences between the two Valsa species. According to previous studies, fungal mitochondrial gene rearrangement may be attributed to the accumulation of repetitive sequences, especially in the intergenic regions [57]. In this study, a large number of repeat sequences was detected in the V. mali mitogenome, accounting for 43.56% of the length of the V. mali mitogenome. However, the proportion of repetitive sequences in V. pyri was much lower, only 2.41%. Compared with the other species in Diaporthales, gene rearrangements have been detected in the V. pyri mitogenome, which implies that other driving mechanisms may be responsible for fungal mitogenome rearrangement.

Most fungal species in the genus Valsa are important plant pathogens. However, no effective disease control strategy has been established despite many years of interest due to confusion with regard to species definition. It has been reported that identifying the Valsa species and their anamorphs Cytospora to the species level based solely on morphological characteristics is very difficult because of the overlapping and variable morphological characteristics of pathogenic anamorphs and the lack of teleomorphs [63]. Therefore, DNA sequences were introduced to study the phylogenetic relationships and species definitions. Up to now, rDNA internal transcribed spacer (rDNA-ITS), β-tubulin partial sequence and translation elongation factor (EF1α) have been widely used to explore the genetic relationships within the Valsa genus. It is still a difficult task to identify some species with more complex relationships using either single-gene or multiple-gene approaches to construct the phylogenetic trees. Here, a highly supported phylogenetic tree of 92 species in different orders was established based on the combined 15 mitochondrial core genes by using BI and ML phylogenetic analysis. The topological structures of the BI and ML trees indicated that all analyzed species were well clustered together or divided into different independent clades or subclades. The two Valsa species were clustered together with high values of 1.0 (BI tree) or 100 (ML tree) and were shown to be more closely related to the three Chrysoporthe species. This result showed that V. mali and V. pyri are two entirely different taxa. V. mali var. mali, and V. mali var. pyri should be renamed as V. mali and V. pyri, respectively. This result also demonstrates that the mitochondrial gene is an effective tool for the analysis of the phylogenetic relationship between Valsa and related genera. However, more mitogenomes of Valsa species must be obtained if we are to precisely understand the origins and evolutionary patterns of Valsa.

It was believed that the mitogenome was acquired from alphaproteobacteria by eukaryotic ancestors through endosymbiosis [64,65]. The main function of mitochondria is to serve as the suppliers of chemical energy in the form of ATP for aerobic respiration in eukaryotes [37]. However, recently, more and more studies have shown that mitochondria also play a crucial role in virulence because fungal growth, biofilm formation, and hyphal growth are regulated by the expression of mitochondrial genes [17]. It has been reported that the imperfect fission of mitochondria of Pyricularia oryzae affects conidiation, growth, and virulence, and that other changes in mitogenome genes affect the development of infection structures, invasion, and pathogenicity [66,67]. However, the corresponding relationships between the virulences of Valsa pathogens and their mitochondrial gene expression are still unknown. This problem should be studied in depth in the future to better understand their pathogenic mechanisms and to establish effective strategies to control plant disease.

5. Conclusions

In this study, the complete mitogenomes of V. mali and V. pyri were reported and compared with other mitogenomes from the order Diaportliaccac. The size of the V. mali mitogenome is 213,406 bp, while that of the V. pyri mitogenome is only 128,022 bp. V. mali’s mitogenome was significantly larger than that of V. pyri. The V. mali mitogenome included a set of 35 introns and 115 uORFs. By comparison, the V. pyri mitogenome contained 22 introns and 50 uORFs. In addition, there were significant differences in genome characteristics, such as gene length and base composition of PCGs, tRNAs, rRNAs, AT skew, and GC skew between V. mali and V. pyri. The intronic region was considered to be the main factor responsible for the difference in mitogenome size between the two Valsa genera. The two Valsa mitogenomes contain a large number of interspersed repetitive sequences (IRS); however, the proportion of IRS in V. mali (43.56%) was significantly greater than that observed in V. pyri (2.41%). It has also been found that large fragments may have transferred between the mitochondrial and nuclear genomes. The values of Ka/Ks for 15 core PCGs were <1, which indicated that these genes were subjected to purifying selection pressure. Significant mitogenomic collinearity and large-scale gene rearrangements were also found to have occurred in two Valsa mitogenomes. The use of a combined dataset of mitochondrial gene sequences for phylogenetic analysis resulted in well-supported phylogenetic trees using the Bayesian inference and the maximum likelihood methods. The topological structure indicated that V. mali and V. pyri were fully independent species. As this is the first report on mitogenomes in the genus Valsa, these results contribute to our understanding of the genetic evolution and species differentiation in the genus of Valsa.