Background

G. elata, a famous Chinese herbal medicine used for more than 2000 years, mainly treats dizziness, headache, migraine, rheumatism, neuralgia and paralysis [1,2,3]. G. elata is a rootless and leafless orchid that is completely heterotrophic and dependent on fungi for nutrition. During the germination of G. elata seeds, they are infected by Mycena strains, become symbiotic with them, and then digest the infected fungi to nourish themselves, so that G. elata seeds sprout and grow into protocorms. As the protocorm of G. elata grows, it needs to be symbiotic with Armillaria strains, and digests them as nutrition to promote the further expansion and growth of G. elata tuber and complete the life cycle [4,5,6]. G. elata depends on Armillaria for nutrition in most periods. The key to successful artificial cultivation of G. elata is whether there are abundant sources of Armillaria. Different Armillaria strains affect the yield and quality of G. elata [7, 8]. Therefore, in the field of G. elata research, the symbiotic molecular mechanism between G. elata and A. mellea has attracted extensive attention.

The symbiotic mechanism between G. elata and Armillaria has rarely been revealed from the perspective of Armillaria, and most studies have focused on G. elata. Previous G. elata genomic analyses indicated an expansion of genes for novel metabolic processes and mycorrhizal association [9, 10]. Yuan et al. [9] found that strigolactone could stimulate the hyphal branching and development of A. mellea. G. elata may preferentially guide the colonization of Armillaria in its cortex through the ABC transporter mediating the secretion of strigolactone to the extracellular space. In addition, some antibacterial components (e.g., phytoalexin gastrodin and Gastrodia antifungal protein) are secreted in the cortex to prevent Armillaria overgrowth in the tubers and cause a disease. Transcriptome analysis of the symbiosis between G. elata and A. mellea showed that the symbiotic G. elata synthesizes organic nutrients and energy through the digestion of the invaded A. mella for the growth of the tuber [11]. Sucrose has been suggested to be transported from the fungus to G. elata as a carbon source, because high concentrations of sucrose and two sucrose transporter-like genes are highly expressed in young Armillaria-colonized tubers [12].

Armillaria species are pathogens causing root white rot disease [13, 14], but a few Armillaria species can establish a symbiotic relationship with G. elata. Previous studies have shown that the growth of Armillaria is closely related to the activities of extracellular enzymes, it secretes, such as laccase, cellulase, xylanase, pectinase and amylase. Extracellular enzymes play a decisive role in the ability of wood rot fungi to degrade nutrients, and their species and activity are related to the species of fungi [15]. The secretion of these extracellular enzymes provides a material basis for Armillaria to infect the epidermis of G. elata. It was also found that different extracellular enzymes play different roles in Armillaria growth. Laccase can degrade lignin and phenols, and its activity affects the ability of Armillaria to degrade lignin [16, 17]. Cellulase hydrolyzes cellulose to produce glucose, which can provide carbon source for the growth of the strain [15, 18]. Xylanase can decompose the polysaccharide structure xylan located in the secondary wall of plants [19]. Amylase mainly hydrolyses plant polysaccharide starch to provide nutrition. Pectinase can degrade pectin in plant cell stroma and primary cell walls [16, 19]. The study of the extracellular enzyme activity of Armillaria is of great significance for the growth of Armillaria and G. elata.

The pathogenicity and preferential saprophytic characteristics of Armillaria may affect its symbiotic relationship with G. elata. The virulence tests on the five Armillaria species show that A. tabescens has the weakest virulence, and A. mellea has the strongest virulence, followed by A. ostoyae, A. gallica and A. cepistipes [16, 20, 21]. Weakly pathogenic and preferentially saprotrophic Armillaria (e.g., A. gallica, and A. cepistipes) may readily establish a symbiotic relationship with G. elata [4]. However, A. mellea can also establish a symbiotic relationship with G. elata. This may be due to variation in virulence within a species; intraspecific variation has been found among strains of some Armillaria species (e.g., A. ostoyae) [16, 22]. Therefore, it is necessary to sequence the whole genome of Armillaria and analyse its genome to identify the genetic variation that is conducive to the establishment of a symbiotic relationship between Armillaria and G. elata. These results provide genomic information for studying the symbiotic mechanism between Armillaria and G. elata at the gene level and developing useful Armillaria strains to improve the yield of G. elata.

To date, several whole genome sequencing datasets of Armillaria have already been published, which revealing plant cell wall degradation enzymes and some secreted proteins [23,24,25]. Even among closely related Armillaria species, fungal mitogenomes are highly variable in size, gene content, and genome organization [26]. Mobile genetic elements invading introns and intergenic sequences of Armillaria mitogenomes play an important role in shaping their genome structure [26]. Mobile and highly repetitive elements (REs) play a significant role in the replication and formation of nucleoprotein complexes and affect the expression of genes [27]. Most REs are derived from transposable elements (TEs) [27]. Fungal genomes are densely packed with TEs [28, 29]. TEs may play important roles in the adaptation of fungi to new ecological niches [30]. In Magnaporthe oryzae, TE was found to be associated with the genes involved in host specialization [31]. However, there are few studies on the genome of Armillaria symbiotic with G. elata. Zhan found that there were approximately 23.6% repetitive sequences in the genome of a diploid A. gallica strain isolated from the tuber of G. elata, but most of the repetitive sequences were unknown [32]. The molecular mechanism of the symbiotic relationship between Armillaria and G. elata is still unclear. Therefore, genomic data are critical to resolve the complex properties of Armillaria species and to better study the symbiotic relationship between Armillaria and G. elata. It is necessary to study the genome of Armillaria species to clarify the characteristics of Armillaria and the symbiotic mechanism between Armillaria and G. elata.

In this study, we used PacBio Sequel and Illumina NovaSeq PE150 to analyse the whole genome sequence of the haploid A. gallica Jzi34 strain, which symbiotic with G. elata, and to assess specific and extended gene families. The results will provide enhanced insights for solving the complex characteristics of Armillaria species and help to study the symbiotic mechanism between Armillaria and G. elata.

Results

Genome assembly and characteristics analysis

The A. gallica Jzi34 genome assembly contained ~ 79.9 Mbp. The final assembly results showed that the genome assembly consisted of 60 contigs with an N50 of 2,535,910 bp. The maximum length of the assembled contig was 8,033,447 bp. The total contig length was 79,897,101 bp. The average GC content of the resulting A. gallica Jzi34 genome was 47.44% (Table 1).

BUSCO v3.0.2 software was used to assess the integrity of the genome assembly and annotation completeness. The results showed that 95.9% (278 BUSCOs) were complete genes based on the BUSCO assessment, while 1.4% (4 BUSCOs) were fragmented and 2.7% (8 BUSCOs) were missing (Table 1).

Table 1 Genome assembly parameters of A. gallica Jzi34, A. gallica 012m, A. gallica Ar21-2, A. cepistipes B5, A. ostoyae C18/9, A. solidipes 28 − 4, A. mellea DSM 3731, and A. fuscipes CMW2740 [23,24,25, 32]
Full size table

Gene Prediction

Functional annotation revealed a total of 16,280 protein coding genes with an average gene length of 1,258 bp in the A. gallica Jzi34 genome (Table 1). These genes, with a total length of 20,474,301 bp, accounted for 25.63% of the total genome length. Furthermore, 14,044 RE sequences were identified in the A. gallica Jzi34 genome assembly. Dispersed repeat sequences (6,505) were classified into six groups, namely, LTR, DNA, LINE, SINE, RC, and unknown and their average lengths were 550, 92, 77, 70, 436 and 59 bp, respectively (Table 2). The LTR was widely represented in A. gallica Jzi34, accounting for 3.1105% of the total genome. A total of 7,539 tandem repeat sequences (TRs) were classified into TR (4,127), Minisatellite DNA (3,118), and Microsatellite DNA (294); they were recorded for 0.3926%, 0.1814% and 0.0137% of the total genome with repeat sizes of 1 ~ 1,891 bp, 10 ~ 60 bp, and 2 ~ 6 bp, respectively (Table 3). The RNAs in A. gallica Jzi34 are summarized in Table 4. The number of tRNAs (308) was highest, followed by snRNAs (24).

Table 2 The dispersed repeat sequences (DRs) of A. gallica Jzi34 genome on class
Full size table
Table 3 Tandem repeat sequences (TRs) of A. gallica Jzi34 genome on class
Full size table
Table 4 A. gallica Jzi34 genome data related RNAs
Full size table

Gene Annotation

In the GO analysis, 8,358 coding genes were annotated in A. gallica; and classified into molecular function, cellular component, and biological process (Fig. 1). Among these categories, six functional entries had more than 2,400 annotated genes. These genes were significantly enriched in binding (4,552), metabolic process (4,369), cellular process (3,960), catalytic activity (3,871), cell (2,493) and cell part (2,493).

Fig. 1
figure 1

GO functional annotation of the A. gallica Jzi34 genome. GO annotation is divided into three major categories and 47 subclasses. A different colour represents each subclass. The x-axis represents the class of genes and the y-axis represents the percent of genes (%). The z-axis represents the number of genes

Full size image

To further reveal the biological function of these coding genes in A. gallica, KEGG pathway annotation was performed. The KEGG database annotated 9,757 genes. The biological pathways are classified into six categories. Among the categories, metabolic pathways had the greatest number of annotated genes (2,311), with a high number of genes for carbohydrate metabolism and amino acid metabolism (Fig. 2).

Fig. 2
figure 2

KEGG pathway annotation of the A. gallica Jzi34 genome. KEGG pathway annotation is divided into six major classes and 45 subclasses. The x-axis indicates the gene number of the concerned subclass. Each subclass is represented by a different colour

Full size image

Based on KOG annotation, 1,982 genes of functional categories were annotated. The protein functions were mainly enriched in general function prediction only (272); post-translation modification, protein turnover, chaperones (224); translation, ribosomal structure and biology (215); energy production and conversion (203); amino acid transport and metabolism (147); and other aspects (Fig. 3).

Fig. 3
figure 3

KOG functional annotation of proteins in the A. gallica Jzi34 genome. KOG function classification is summarized in 26 classes. The x-axis indicates each class and the y-axis shows the number of matched genes. The names of groups and number of genes are mentioned

Full size image

The Carbohydrate Enzyme Classification and Annotation

A total of 664 CAZymes were annotated in the genome of A. gallica Jzi34, which included 267 glycoside hydrolases (GHs), 158 auxiliary activity enzymes (AAs), 121 glycosyl transferases (GTs), 83 carbohydrate-binding modules (CBMs), 50 carbohydrate esters (CEs), and 27 polysaccharide lyases (PLs) (Fig. 4). Compared with the genome assembly of other Armillaria, A. gallica Jzi34 had fewer CAZymes, while it had the largest set of GTs involved in glycosylation and the synthesis of polysaccharides (Table 5).

Fig. 4
figure 4

Carbohydrate enzyme functional classification. The CAZy classification is divided into 6 classes, including AA, CBM, CE, GH, GT and PL. The x-axis shows the CAZy class and the y-axis indicates the number of genes

Full size image

Although A. gallica Jzi34 had fewer AAs, its AA3-2 subfamilies (77) are more abundant than those of A. gallica 012m (45) and A. gallica Ar21-2 (58). The copy number of the AA3-2 subfamily was also higher than those of A. cepistipes B5, A. ostoyae C18/9, and A. solidipes 28 − 4 (Table 5).

Table 5 CAZymes and Cytochrome P450 of A. gallica Jzi34, A. gallica 012m, A. gallica Ar21-2, A. ostoyae C18/9, A. solidipes 28 − 4, A. cepistipes B5 [25, 32]
Full size table

P450

P450 enzymes are involved in the construction of important metabolites in fungi, but also play a key role in adaptation to specific environments by modifying different compounds. In total, 349 P450 genes were identified in A. gallica Jzi34. The number was greater than those of A. gallica Ar21-2 (330), A. cepistipes B5 (324), A. solidipes 28 − 4 (279), A. ostoyae C18/9 (271), and A. gallica 012m (271) (Table 5). According to the annotation of these genes in the P450 database, they were divided into 8 groups. Most (223) of the genes were predicted to encode “E-class P450, group I”, followed by “Cytochrome P450” (52) and “E − classP450, group IV” (46) (Fig. 5).

Fig. 5
figure 5

Functional classification of cytochrome P450 in the A. gallica Jzi34 genome. cytochrome P450 annotation is divided into 12 classes. The x-axis represents the P450 class. The y-axis represents the number of matched genes

Full size image

Pathogen Host Interaction (PHI)

A total of 885 genes were annotated by searching the PHI database. The genes for pathogenicity of A. gallica Jzi34 in the pathogen host interaction (PHI) database were annotated to 8 categories. It revealed that the largest number (407) of the genes was associated with the functional class of “reduced virulence”, followed by the functional class of “unaffected pathogenicity” (217) and “loss of pathogenicity” (87). It also found that 67, 37 and 10 genes were identified as “lethal function”, “increased virulence” and “effector”, respectively. In addition, 57 genes were found to be involved in the NA function class (Fig. 6).

Fig. 6
figure 6

PHI classification in the A. gallica Jzi34 genome. PHI annotation is classified into 8 classes. The x -axis shows each class, and the y-axis represents the number of matched genes

Full size image

Synteny analysis of P450 genes

To further explore the evolutionary relationship of P450 proteins in A. gallica Jzi34 (AgP450s) and other Armillaria, the syntenic relationship was traced between AgP450s and homologs in other Armillaria species. The number of homolog pairs between AgP450s and P450s in the other Armillaria, including A. gallica Ar21-2, A. cepistipes B5, A. solidipes 28 − 4, and A. ostoyae C18/9, was 248, 263, 215, and 226. A. gallica Jzi34 had many specific P450 genes (Table 6). The results showed that the genetic relationship between AgP450s and P450s in A. gallica Ar21-2, A. cepistipes B5, A. solidipes 28 − 4, and A. ostoyae C18/9 was complex (Fig. 7A-D).

Fig. 7
figure 7

Syntenic analysis of P450 genes between A.gallica Jzi34 and four other Armillaria. The gray lines at the bottom indicate the collinear blocks within A.gallicaJzi34 and other Armillaria genomes. The red lines indicate the pairs of P450 genes. The results of the syntenic analysis between A.gallica Jzi34 and other Armillaria, including A. cepistipes B5, A. gallicaAr21-2, A. ostoyae C18/9, and A. solidipes 28-4 (A–D)

Full size image
Table 6 Homologous gene pairs between AgP450s and P450s in the other Armillaria
Full size table

To better understand the evolutionary constraints affecting the P450 gene family, the Ka / Ks ratios of P450 gene pairs were calculated. The Ka / Ks ratio of the gene pairs between AgP450 genes and the P450 orthologous genes was less than 1, which suggests that the P450 gene family have undergone a strong selective purge pressure during evolution (Additional file 1 Table S 1).

Discussion

The genome size of A. gallica Jzi34 (~ 79.9 Mbp) was within the genome size range of other Armillaria species. The five draft genomes of Armillaria were assembled into 63–29,300 scaffolds comprising 53–87 M [25, 32]. However, the genomes of A. gallica Ar21-2 and A. solidipes 28 − 4, as pathogens of wood, contain many ambiguous bases [25, 32]. In this study, we obtained a 79.9 Mbp genome of A. gallica Jzi34, which consisted of 60 contigs with an N50 of 2,535,910 bp and 47.44% GC content. A total of 16,280 genes were identified in the A. gallica Jzi34 genome, with an average length of 1258 bp (Table 1). This was less than the number of genes in closely related species, which contained 14,473–26,261 genes [23, 25, 32]. However, even among closely related Armillara species, fungal mitogenomes are significantly different in size, gene content, and genome organization [26].

Compared with the complete BUSCO genes in A. gallica Ar21-2 (98.6%) as wood pathogens, the lower genome completeness of A. gallica 012m (95.8%) and A. gallica Jzi34 (95.9%) established a symbiotic relationship with G. elata (Table 1). The genome completeness of A. gallica Jzi34 was lower than that of A. gallica Ar21-2. This may be related to the extensive loss of lignin and cellulose degrading enzymes involved in the transition from saprophytic to symbiotic [33].

In this study, only 4.1% of RE sequences were identified in the A. gallica Jzi34 genome assembly, but few sequences were unknown (Table 2). The genome of A. gallica strain 012m had approximately 23.6% repetitive sequences, but most of the repetitive sequences were unknown [32]. There was a significant difference in the number of TEs among different Armillaria species [25, 33]. Moreover, LTR elements of the identified TE classes were the most frequent in basidiomycete fungi [34]. The LTR (6,505) was also widely represented in the A. gallica Jzi34 genome, accounting for 3.1% of the A. gallica Jzi34 genome (Table 2).

A. gallica Jzi34 had fewer CAZymes than other Armillaria species. This finding coincided with previous research showing that certain CAZyme families contracted during the evolution of mycorrhizal fungi compared with those of their saprotrophic ancestors [35, 36]. Although the AA family was significantly contracted in A. gallica Jzi34, its AA3-2 subfamily had more genes than the other Armillaria (Table 5). Moreover, compared with the other Armillaria genomes, A. gallica Jzi34 had the largest set of GTs. The AA3-2 subfamily has been considered important enzymes in the biodegradation of lignocellulose [37, 38]. Glycosylation catalysed by glycosyltransferases (GTs) contributes to fungal infection and secondary metabolic synthesis. GTs are critical for fungal growth, host penetration and immune evasion [39, 40]. These may be beneficial for A. gallica Jzi34 to establish a symbiotic relationship with G. elata, and continue to provide nutrition for G. elata.

In fungi, P450 enzymes are widely involved in a variety of physiological reactions, contributing to the adaptation and fecundity of fungi in specific ecological niches [41, 42]. P450 enzymes may play an important role in fungal colonization of plant material [43]. Filamentous fungi symbionts are important in detoxifying host chemical defence compounds [44]. There were more P450 genes identified in A. gallica Jzi34 (349) than in A. gallica Ar21-2 (330), A. cepistipes B5 (324), A. solidipes 28 − 4 (279), A. ostoyae C18/9 (271), and A. gallica 012m (271) (Table 5). Therefore, an increase in the P450 genes may contribute to the growth of A. gallica Jzi34 in a symbiotic relationship with G. elata.

The syntenic analysis on AgP450s and the homologs showed that more P450 orthologous pairs were found between A. gallica Jzi34 and two other Armillaria, including A. gallica Ar21-2 and A. cepistipes B5. However, the homologous gene pairs between A. gallica Jzi34 and A. solidipes 28 − 4 and the homologous P450 genes pairs between A. gallica Jzi34 and A. ostoyae C18/9 were relatively less (Table 6). These results are consistent with the phylogenetic relationship between A. gallica and the other Armillaria [25, 32]. There were still many AgP450s in A. gallica Jzi34 did not exhibit the homologous gene pairs with these Armillaria, including A. gallica Ar21-2, A. cepistipes B5, A. solidipes 28 − 4, and A. ostoyae C18/9 (Table 6). This may be related to the expansion of the cytochrome p450 family in Armillaria species [25], suggesting the unique role in biology of these fungi. It may also be related to the symbiosis of A. gallica Jzi34 and G. elata.

Conclusions

In this study, we sequenced the genome and annotated the functional genes of the A. gallica Jzi34 strain, which was symbiotic with G. elata. A total of 16,280 genes were predicted in the genome. Repetitive sequences represent approximately 4.1% of the genome. Genome comparison analyses with other Armillaria strains showed that carbohydrate enzyme genes were significantly contracted in A. gallica Jzi34. However, it had an expansion of cytochrome P450 and AA3-2 subfamily genes. The synteny analysis result of P450 genes showed that the evolutionary relationship of P450 proteins in A. gallica Jzi34, A. gallica Ar21-2, A. cepistipes B5, A. solidipes 28 − 4, and A. ostoyae C18/9 is complex. The study elucidates the characteristics of A. gallica Jzi34. In addition, it provides an important genomic resource for further studies on the symbiotic relationship between Armillaria and G. elata.

Methods

Materials and sequencing

The haploid A. gallica Jzi34 strain used for genome sequencing was isolated from a fruiting body as a single spore isolate. The fruiting body was collected from a plantation field in Luquan, Kunming, Yunnan Province, China, in 2021. For genomic DNA, A. gallica Jzi34 was grown in a liquid culture medium [45]. Then, the culture was shaken for 10 days at 115 rpm/min and 25°C. Mycelium was collected and frozen in liquid nitrogen. Genomic DNA was extracted with the SDS method [46]. The whole genome of A. gallica Jzi34 was sequenced using the PacBio Sequel platform and Illumina NovaSeq PE150.

Genome assembly

SMRT Link v5.0.1 was used to accomplish preliminary assembly [47, 48]. To ensure the accuracy of the subsequent analysis results, the low-quality reads were filtered (less than 500 bp) to obtain clean data. Using the automatic error correction function of the SMRT portal, the long reads were selected (more than 6000 bp) as the seed sequence, and the other shorter reads were aligned to the seed sequence by Blasr, so as to improve the accuracy of the seed sequences. After assembly, we obtained an initial result. The Variant Caller module of SMRT Link software was used to correct and count the variant sites in the preliminary assembly results.

We used BUSCO Version 3.0.2 to assess the completeness of the assemblies. The lineage dataset is: fungi_odb9 (Creation date: 2016-02-13, number of species: 85, number of BUSCOs: 290).

Genome component prediction included the prediction of the coding gene, repetitive sequences and noncoding RNA. The steps were as follows:

  1. 1)

    The Augustus 2.7 program was used to retrieve the related coding genes.

  2. 2)

    The interspersed repetitive sequences were predicted using RepeatMasker (http://www.repeatmasker.org/) [49]. The tandem repeats were analysed by the Tandem Repeats Finder [50].

  3. 3)

    Transfer RNA (tRNA) genes were predicted by tRNAscan-SE [51]. Ribosome RNA (rRNA) genes were analysed by rRNAmmer [52]. sRNA, snRNA and miRNA were predicted by BLAST against the Rfam database (default parameters).

Gene function

We used seven databases to predict gene functions. They were respective GO (Gene Ontology) [53], KEGG (Kyoto Encyclopedia of Genes and Genomes) [54, 55], KOG (Clusters of Orthologous Groups), NR (Non-Redundant Protein Database) [56], TCDB (Transporter Classification Database), P450 [57, 58], and, Swiss-Prot. A whole genome BLAST search (E-value less than 1e-5, minimal alignment length percentage larger than 40%) was performed against the above seven databases. The secretory proteins were predicted by the Signal P database. For pathogenic fungi, we added the pathogenicity analyses. PHI (pathogen host interactions) was used to perform the above analyses [59, 60]. Carbohydrate-active enzymes were predicted by the Carbohydrate-Active Enzymes Database.

Syntenic analysis and Ka/Ks analysis

The four Armillaria species amino acid, genome, and CDS sequence assembly and corresponding annotation were downloaded from the US DoE JGI fungal genomics resource database (https://mycocosm.jgi.doe.gov/) [25]. TBtools [61] was used to analyse the homology and collinearity of the P450 family gene between A. gallica Jzi34 and the other Armillaria (default parameters).