Introduction

As a group of edible mushrooms with excellent flavour and high medicinal value (Fu et al. 2013; Su et al. 2013; Hu et al. 2013), morels (Pezizomycetes, Ascomycota) are found in most parts of the world and are extensively traded (Pilz et al. 2007; Pildain et al. 2014). In recent years, morels have been commercially cultivated to meet the ever-growing need in China.

Phylogenetic analyses indicate that Morchella consists of three clades: the Elata Clade (black morels), the Esculenta Clade (yellow morels) and the rufobrunnea Clade (O’Donnell et al. 2011). Morchella importuna, Mel-20 and M. sextelata are included in the Elata Clade (Du et al. 2014; Richard et al. 2015). In a previous study, some black morel species were confirmed to form secondary ectomycorrhizal symbioses with spruce (Buscot 1994), while M. importuna and M. sextelata are considered to be saprophytic fungi that can be cultivated artificially (Peng et al. 2016).

Although several species of Morchella have been successfully cultivated (Ower et al. 1986; Miller 2005; Masaphy 2010; Peng et al. 2016), their genetic information, life cycles and reproductive systems remain poorly understood. Based on SNP, SCAR, and AFLP markers, Pagliaccia et al. (2011) consider members of the Morchella sp. Mel-12 phylogenetic lineage to be heterothallic and to outcross in nature. Multiple nuclei are present in a single ascospore or one hyphal segment of Morchella (Hervey et al. 1978; Volk and Leonard 1990). However, whether the ascospores are heterokaryotic or homokaryotic remains unclear.

The sexual reproduction of Ascomycota fungi is controlled by a single mating-type locus (MAT-1) (Coppin et al. 1997) with two highly divergent nonhomologous idiomorphs (Metzenberg and Glass, 1990), designated MAT1–1 and MAT1–2 (Turgeon and Yoder 2000). The MAT1–1 idiomorph is characterized by the MAT1–1-1 gene, which encodes protein with an alpha-box domain, while the gene MAT1–2-1, which encodes a high-mobility group (HMG) domain, is generally located on the MAT1–2 idiomorph (Debuchy and Turgeon 2006). Individual isolates usually contain either the MAT1–1 or MAT1–2 idiomorph in heterothallic species. In contrast, in homothallic species, both mating types are contained in the haploid genome, usually tightly linked, although exceptions exist (Nelson 1996; Debuchy and Turgeon 2006). In Pezizomycetes, edible mushrooms that are as highly praised as morels, Tuber melanosporum and T. indicum, have been identified as heterothallic (Rubini et al. 2011; Belfoori et al. 2013).

In our previous studies, a draft genome database of the Mes-15 strain YAAS2689 included in the Esculenta Clade was obtained (data unpublished), and the sequences of the genes MAT1–1-1 (KP776983) and MAT1–2-1 (KP776984) were verified. In this study, the genomes of two single ascospore isolates from M. importuna (YPL6) with opposite mating-type were sequenced. The aim was (1) to identify the MAT locus structure of M. importuna, (2) to clone and characterise the MAT1–1-1 and MAT1–2-1 mating-type genes of M. importuna, Mel-20 and M. sextelata, and (3) to analyse whether these three black morel species are heterothallic or homothallic.

Materials and methods

Strains and isolations of ascospores

The fruiting body WXLBD7 of Mel-20 was collected from the forests of Weixi (Yunnan province). YPL2 and YPL6 of M. importuna and HL1 of M. sextelata were obtained from morel cultivation fields in Chuxiong (Yunnan province) and Huili (Sichuan province), respectively (Table 1). The corresponding strains were obtained by tissue isolation. Single-ascospore strains were isolated from the fruiting bodies after spore dilution. Every ascospore isolate was numbered with an Arabic numeral suffix corresponding to the ascocarp. For example, YPL6–3 refers to the 3rd ascospore isolate from ascocarp YPL6. All the cultures were identified by phlogenetic analyses (Table S1; Figs. S1, S2) and deposited in Mushroom Center of Yunnan Crops Genebank (in Yunnan Academy of Agricultural Sciences, YAAS), Kunming, China.

Table 1 Ascocarps used in this study
Full size table

DNA extraction and PCR amplification

Each isolate was incubated on PDA at 23 °C for 7 days. Mycelia were scraped and ground in liquid nitrogen. DNA extraction was performed using an EZgene™ Fungal gDNA Kit (BIOMIGA Inc.). All primer pairs were designed by Primer Premier v.5 and synthesized by Beijing Tsingke Biotechnology. The PCR amplifications were performed in a 25-μL mixture containing 12.5 μL of PCR Mix (TSINGKE), 1 μL of primer (10 μM/L) and 10 ng of DNA. Long-range PCR amplification was performed using LA Taq DNA polymerase (TaKaRa Biotechnology (Dalian)). The primer sequences and main PCR cycling parameters are listed in Table 2.

Table 2 primer sequences and PCR cycling parameters used in this study
Full size table

Genome sequencing

According to the conserved amino acid sequences of the MAT1–2-1 gene HMG domain of ascomycetes, primer pair p1 was designed to amplify the corresponding gene fragments from the single-ascospore isolate group which contained 105 isolates of M. importuna YPL6. The isolates that contained the predicted amplification bands were considered to contain the MAT1–2 idiomorph, while the isolates that lacked detectable bands were hypothesised to contain the MAT1–1 idiomorph.

Genomic DNA from the isolates YPL6–1 with the MAT1–2 idiomorph and YPL6–3 with the MAT1–1 idiomorph were submitted to the Beijing Genomics Institute (BGI). Genomic libraries were constructed and paired-End sequences were produced using Illumina HiSeq™ 2000. The clean reads were assembled into contigs by the SOAPdenovo assembly programme (http://soap.genomics.org.cn/soapdenovo.html), and the scaffolds were built by analysing the information of paired reads that covered different contigs. Genes were predicted using GeneMark-ES (v.2.3e)software and corresponding gene annotations were conducted by BLASTx analysis with the KEGG, KOG, SwissProt, NR and GO databases.

Identification and verification of the MAT1–1 and MAT1–2 idiomorphs of M. importuna

The MAT1–2 and MAT1–1 idiomorphs were identified by alignment analyses of two scaffolds containing MAT1–2-1 or MAT1–1-1. Ten PCR primer pairs (p2–1, p2–2, p2–3, p2–4, p2–5, p3–1, p3–2, p3–3, p3–4, and p3–5) (Table 2) were employed to verify the two idiomorphs based on their genome sequences. PCR was performed under standard conditions (Table 2). The products were recovered from agarose gels and purified using a Gel/PCR Extraction Kit (BIOMIGA), then cloned into the pGM-T vector and sequenced. Sequence comparison and dotplot analysis were conducted to verify the two idiomorphs of M. importuna.

Isolation of MAT1–2-1 and MAT1–1-1 from Mel-20 and M. sextelata

Conserved sequences could be found by analysis of the flanking regions of MAT1–2-1 gene in M. importuna (KY782629) and Mes-15 (KY782632). A pair of specific primers (p2–3f and p2–3r) were designed based on the conserved regions to amplify the complete sequences of the MAT1–2-1 gene in Mel-20 and M. sextelata. The PCR amplicons (approximately 5 kb) were cloned and sequenced.

Because no conserved portions could be located in the 10-kb flanking regions from the upstream to the downstream region of the MAT1–1-1 gene in M. importuna (KY782630) and Mes-15 (KY782631), a primer pair (p7f and p7r) was designed to sequences within the MAT1–1-1 gene to amplify the corresponding DNA fragments. To obtain the complete sequence of the MAT1–1-1 gene, the DNA walking method was performed using the Genome Walking Kit (TaKaRa Biotechnology (Dalian)) with SP Primers (p9-1f, p9-2f, p9-3f, p9-1r, p9-2r, and p9-3r) based on MAT1–1-1 fragments of Mel-20 and M. sextelata.

Verification of the coding regions of MAT1–1-1 and MAT1–2-1 genes

Total RNA was extracted from the mycelia of YPL6, YPL6–1, YPL6–3, HL1, HL1–1, HL1–47, WXLBD7, WXLBD7–11 and WXLBD7–23 using a TaKaRa MiniBEST Plant RNA Extraction Kit. The first-strand cDNAs were synthesized using Oligo dT primer (TaKaRa PrimeScript™ II 1st Strand cDNA Synthesis Kit). The primer pairs p8 for the MAT1–1-1 gene and p10 for the MAT1–2-1 gene were designed using predicted sequences adjacent to the start site and end site of the two genes. The design strategy also took advantage of conserved regions, so the primers could be used in all three morel species. Introns were verified manually by comparing the RT-PCR sequences with the corresponding DNA sequences.

Mating type ratios of ascospores from the same ascocarp

Single-ascospore populations of WXLBD7, YPL2, YPL6 and HL1 were obtained to determine the occurrence and frequency of the mating type idiomorphs. Standard PCR tests with p8 and p10 primer pairs were conducted to examine the four single-ascospore populations. A chi-square test was applied to compare the MAT1–1/MAT1–2 ratio at the significance level P = 0.05.

Results

MAT loci structures of M. importuna

The results of genome assembly and gene annotation of isolations YPL6–1 and YPL6–3 showed that the genome sizes and components of two isolations are similar (Tables S2, S3). The MAT1–1-1 gene was observed only on Scaffold 152 (KY782630) of the M. importuna isolate YPL6–3 genome, which was 80 kb long and contained 18 predicted genes. The MAT1–2-1 gene was found on Scaffold 91 (KY782629) of the YPL6–1 genome, which was 75 kb long and contained 26 predicted genes. Sequence alignment of the two scaffold regions revealed that the MAT1–1 and MAT1–2 idiomorphs were 10.5 and 6.7 kb long, respectively.

The MAT1–2 idiomorph contained only a single MAT1–2-1 gene. However, in addition to the MAT1–1-1 gene, there were two predicted coding sequences GME3123 and GME3124 on the MAT1–1 idiomorph (Fig. 1). They were not homologues to any of the mating genes that have been published; therefore, we speculated that they are novel mating genes unique to morels. GME3123 was found to be 1688 bp containing four introns and encoding 488 amino acids, and GME 3124 was 1825 bp containing five introns and encoding 512 amino acids. BLASTx analysis revealed that GME3123 had 24% and GME3124 21% amino acids similarity to hypothetical protein (AIU38081.1) of Tuber borchii (Belfiori et al. 2016).

Fig. 1
figure 1

Organization of MAT locus in M. importuna. The scale bar indicates sizes of only the MAT idiomorphs, the distance and sizes of other genes are not to scale. Introns in the MAT genes are indicated by black boxes. Thick and thin arrows indicate the orientation of genes and prime pairs respectively. Gene or superfamily names refer to a gene encoding a protein showing the highest percentage identity to the putative gene product following BLASTx analysis. APN2 DNA-(apurinic or apyrimidinic site) lyase 2; cox13 cytochrome c oxidase subunit 6A, mitochondrial; CPSF6: cleavage and polyadenylation specificity factor subunit 6; tfa1 transcription initiation factor IIE subunit alpha; atp-3 ATP synthase subunit 4, mitochondrial; SDH2 succinate dehydrogenase (ubiquinone) iron-sulfur subunit, mitochondrial; Mba1 mitochondrial inner membrane protein; end4 (also known as SLA2) endocytosis protein end4

Full size image

To compare this analysis with the published morel genome databases, a sequence completely identical to the MAT1–1 idiomorph was found on the Scaffold 48 of the M. conica genome (http://genome.jgi.doe.gov/Morco1/Morco1.home.html) between 239.6 kb and 250.1 kb. Three genes, 10317, 10316 and 10315, were in this area on the Scaffold 48 of the M. conica genome, and they were completely identical with the genes MAT1–1-1, GME3123 and GME3124, respectively. No sequences identical to the MAT1–2 idiomorph were founded in this genome database. Sequences completely identical to the MAT1–2 idiomorph were located on the Scaffold 20 of the M. importuna strain SCYDJ1-A1 genome (http://genome.jgi.doe.gov/Morimp1/Morimp1.home.html) between 230.6 kb and 237.3 kb. Only an ORF known as fgenesh1_pg.20_#_79 was found in this area, and its sequence was the same as the MAT1–2-1 gene. The MAT1–1 idiomorph could not be found in this genome database.

The coding sequences flanking the two idiomorphs were highly conserved. The 5′ flanking region of both idiomorphs contained APN2, which is connected to the cox13 gene, and there were four additional predicted ORFs between the cox13 gene and the MAT idiomorphs (Fig. 1). The APN2 gene displayed 66% amino acid sequence similarity to CCX30259.1 (Pyronema omphalodes), and the cox13 gene displayed 62% similarity to GAP92444.1 (Rosellinia necatrix). The end4 (also known as SLA2) gene was the second gene that located on the 3′ flanking region of two idiomorphs, and it displayed 66% amino acid sequence similarity to CCX33915.1 (Pyronema omphalodes).

Structure of the MAT1–1-1 gene of three morel species

The lengths of the MAT1–1-1 gene sequences of M. importuna, M. sextelata and Mel-20 were 1694, 1730, and 1651 bp, respectively. RT-PCR confirmed that the MAT1–1-1 ORFs contained two introns and encoded proteins of 525, 537 and 511 amino acids, respectively (Table 3). Sequence alignments and dotplot analyses indicated that the MAT1–1-1 gene in M. importuna displayed 92.3 and 87.2% sequence identity to those in M. sextelata and Mel-20, respectively.

Table 3 The variance of MAT1–1-1 and MAT-2-1 gene in three morel species
Full size table

The deduced amino acid sequences of these three morel species contained a conserved alpha-box motif of the MAT1–1-1 protein of ascomycetes (Fig. 2a). In addition, the three proteins shared more than 20% amino acid identities to those in Tuber indicum (AHE80942), T. melanosporum (ADU56595) and T. borchii (AIU38080).

Fig. 2
figure 2

Amino acid alignment of conserved regions of MAT proteins of Morchella spp. with those of other ascomycetes. a Amino acid alignment of the conserved alpha-box domain of the following species: (1) Stemphylium callistephi(AAR04468); (2) Alternaria brassicicola(AAK85542); (3) Aspergillus fischeri(XP_001263836); (4) Talaromyces marneffei (ABC68484); (5) Histoplasma capsulatum (ABO26868); (6) Ophiocordyceps sinensis (ALH25059); (7) Cordyceps militaris (AKM95188); (8) Fusarium poae (CAD59610.3); (9) Colletotrichum musae (CAD59611.3); (10) Fusarium avenaceum (CAD59608.4); (11) Morchella sextelata; (12) Morchella importuna; (13) Mel-20; (14) Tuber indicum (AHE80942); (15) Tuber melanosporum (ADU56595); (16) Tuber borchii (AIU38080); (17) Xanthoria polycarpa (CAI59771); (18) Pyrenopeziza brassicae (CAA06844); (19) Sclerotinia homoeocarpa (AJW31369). b Amino acid alignment of the conserved HMG domain of following species: (1) Aspergillus fumigatus (XP_754989.2); (2) Penicillium chrysogenum (CAP17333); (3) Sclerotinia sclerotiorum (AGB05594); (4) Sclerotinia trifoliorum (ANN44262); (5) Sclerotinia homoeocarpa (AJW31335); (6) Sordaria fimicola (CAB63226); (7) Cordyceps militaris (AKM95197); (8) Ophiocordyceps sinensis (AEH27625); (9) Colletotrichum gloeosporioides (AKO22190); (10) Rusavskia elegans (CAI59778.2); (11) Dufourea flammea (CAI59780.2); (12) Passalora fulva (ABG49507); (13) Leptosphaeria maculans (AAO37761); (14) Bipolaris sorokiniana (AAF87724); (15) Morchella importuna; (16) Morchella sextelata; (17) Mel-20; (18) Tuber borchii (AIU38078); (19) Tuber indicum (AHE80950)

Full size image

Structure of the MAT1–2-1 gene of three morel species

The lengths of MAT1–2-1 gene sequences of M. importuna, M. sextelata and Mel-20 were 1197, 1199 and 1247 bp, respectively. The results of RT-PCR analyses confirmed that the MAT1–2-1 ORFs contained three introns and encoded proteins of 329, 329 and 344 amino acids, respectively (Table 3). Sequence alignments and dotplot analyses indicated that the MAT1–2-1 gene in M. importuna shared 94.3 and 86.3% sequence identity with the corresponding genes in M. sextelata and Mel-20, respectively. The alignment of the MAT1–2-1 protein with corresponding sequences deposited in GenBank indicated that the HMG-box motifs of three morels were conserved (Fig. 2b).

Mating type ratios of three morel species based on single-ascospore populations

The primer pairs p8 for the MAT1–1-1 gene and p10 for the MAT1–2-1 gene could be used in all three morel species to examine the mating types of single-ascospore populations. Strains that contained only MAT1–1-1 and not MAT1–2-1 were amplified and scored as MAT1–1, and vice versa (Fig. 3). In the four ascospore populations (Table 4), the ratio of MAT1–1 to MAT1–2 ranged from 0.44 to 0.96. Chi-square analysis indicated that the ratio was not significantly different from 1:1 except for the HL1 populations. No sequence difference was detected between the single-ascospore strains and the parent strains in either of the MAT1–1-1 or MAT1–2-1 genes.

Fig. 3
figure 3

MAT gene PCR products of single ascospore populations YPL2, HL1 and WXLBD7. a MAT1–1-1 gene PCR products of YPL2–1 to YPL2–24. b MAT1–2-1 gene PCR products of YPL2–1 to YPL2–24. c MAT1–1-1 gene PCR products of HL1–1 to HL1–24. d MAT1–2-1 gene PCR products of HL1–1 to HL1–24. e MAT1–1-1 gene PCR products of WXLBD7–1 to WXLBD7–24. f MAT1–2-1 gene PCR products of WXLBD7–1 to WXLBD7–24. M DNA marker 2000 (2000,1000, 750, 500, 250 and 100 bp)

Full size image
Table 4 Summary of MAT type ratios of four ascospore populations
Full size table

Discussion

Heterothallism in M. importuna, M. sextelata and Mel-20

In ascomycetes, the mating type corresponds to two allelic forms of a single locus and, accordingly, heterothallism is bipolar (Whitehouse 1949). This property was first demonstrated in Ascobolus magnificus and A. carbonarius (Dodge 1920) and then in several Neurospora species (Shear and Dodge 1927). Molecular analyses of mating-type in filamentous ascomycetes began with the cloning of A and a mating-types from Neurospora crassa (Glass et al. 1988). It has since been confirmed that the mating-type locus contains one of two highly divergent sequences occupying the same chromosomal locus. Metzenberg and Glass (1990) used the word idiomorph to denote that these large sequences are not obviously related by structural similarity. In heterothallic ascomycetes, ascospore populations can be divided into two groups: those harbouring the MAT1–1 and MAT1–2 idiomorphs. The sexual cycle occurs only between two individuals with opposite mating-type, rather than with self-fertilizing as in homothallic reproduction (Pöggeler 2001; Chilvers et al. 2014).

The genomic data of M. importuna YPL6 confirmed that the MAT locus is divided into two idiomorphs. In the YPL6–3 genome, only the MAT1–1 idiomorph harbouring the MAT1–1-1 gene was found. Similarly, the MAT1–2 idiomorph with the MAT1–2-1 gene was only located in the YPL6–1 genome. Furthermore, the conserved flanking sequences of the two idiomorphs confirmed that they occupy the same chromosomal locus. The characterization of the mating locus is consistent with heterothallic ascomycetes. Mating-type analysis of four ascospore populations from three species, M. importuna, M. sextelata and Mel-20, indicated that two sexual groups were present, and the ratios of MAT1–1- and MAT1–2-harbouring idiomorphs were not significantly different from a 1:1 ratio. These results suggested that these morels are all heterothallic.

Because the lack of visible structures, such as the clamp connection in basidiomycetes precludes visible morphological analysis, these molecular data will facilitate crossbreeding to produce high-quality morel cultivars.

Structure of MAT idiomorphs

In heterothallic ascomycetes, the structure of the MAT idiomorph varies in different species. The mating idiomorphs of Neurospora crassa were the first to be investigated, and the MAT1–1 idiomorph contains three ORFs: MAT1–1-1, MAT1–1-2 and MAT1–1-3 (mat A-1, A-2 and A-3 in N. crassa). The MAT1–2 idiomorph contains the MAT1–2-1 and MAT1–2-2 ORFs (mat a-1 and a-2 in N. crassa) (Pöggeler and Kück 2000). Similar MAT structures have been found in many other species surveyed later (Debuchy and Turgeon 2006; Wilken et al. 2012; Hutchinson et al. 2016). In this study, M. importuna has been identified as possessing a similar MAT idiomorph structure similar to those previously described. The MAT1–1 idiomorph contains three OFRs, and the MAT1–2 idiomorph contains a single MAT1–2-1 gene. However, two ORFs,GME3123 and GME3124,in the MAT1–1 idiomorph were found to be highly divergent from the homologues of mating genes that have been published. At this point, we can only speculate that GME3123 and GME 3124 are new mating-type genes at the MAT1–1 locus of M. importuna.

APN2, SLA2 and APC5 are ancestral companions of the MAT locus in euascomycetes (Debuchy and Turgeon 2006). Similarly, homologues of APN2 and SLA2 were found on each side of the MAT locus in Sordariomycetes, and the ATG3, cox13 and CWF24 genes were also conserved in the up- and downstream regions of MAT (Debuchy and Turgeon 2006; Xu et al. 2016; Lu et al. 2016). For M. importuna, the APN2 gene was connected to a cox13 gene on the 5′ flanking region of both MAT idiomorphs. However, four predicted genes were found between the cox13 gene and the MAT idiomorphs. The SLA2 (end4) gene was located on the 3′ flanking region of two idiomorphs. These results indicated that the MAT locus structure of M. importuna had some conserved sequences but was highly divergent from other species.

Based on the conserved flanking sequences, MAT1–1 and MAT1–2 idiomorphs have been cloned from many species (Martin et al. 2011; Belfiori et al. 2013). In this study, the coding sequences flanking the two idiomorphs of M. importuna were highly conserved, but the cloning of the whole MAT idiomorph of M. sextelata and Mel-20 using long-distance PCR amplification was not successful (data not shown), possibly due to the variability of the sequence in the different species. Unfortunately, we were not able to identify the entire mating idiomorphs of these two morel species.

Expression of mating genes

In Neurospora crassa and Tuber melanosporum, mating genes are constitutively expressed during the vegetative as well as the sexual phases (Ferreira et al. 1996; Rubini et al. 2011). In this study, we did not analyse the expression of mating genes specially in different cultivation periods and conditions. However, we verified the coding region by successfully amplifying the MAT1–1-1 or MAT1–2-1 gene with cDNA from all the tested strains incubated on PDA for 10 days. For the single-ascospore isolates, MAT1–1-1 of YPL6–3, HL1–1 and WXLBD7–11, and MAT1–2-1 of YPL6–1, HL1–47 and WXLBD7–13 were amplified with their cDNA, respectively. In addition, the expression of the two genes was detected in the mycelia from the YPL6, WXLBD7 and HL1 tissues. These results implied that the mating genes in morels are constitutively expressed.

Mating locus and phylogenetic relationship

In ascomycetes, as throughout the eukaryotic kingdom, genes controlling sex determination are evolving more rapidly than other protein-coding genes (Whitfield et al. 1993; Ferris et al. 1997; Pöggeler 1999), which enables the study of the phylogenetic relationship of closely related fungal species. The sequences from the gpd (glyceraldehyde-3-phosphate dehy-drogenase gene), mat A-1 and mat a-1 genes produced phylogenetic trees with a similar topology and strict separation of homothallic and heterothallic species within the genera Neurospora and Sordaria (Pöggeler 1999). Belfiori et al. (2013) revealed the existence of cryptic species of Tuber indicum by comparing the idiomorph structure and sequences of MAT genes.

Morchella importuna and M. sextelata had a closer phylogenetic relationship than did either to Mel-20 (Fig. S1). Based on comparing the DNA sequences of the MAT1–1-1 and MAT1–2-1 genes, M. importuna showed greater identity to M. sextelata than to Mel-20, which was consistent with the molecular evolution relationship based on the analysis of ITS, RPB1, RPB2 and TEF1 (Du et al. 2012; Richard et al. 2015). Acquiring more sequence information on the mating loci will help to investigate the phylogenetic diversity of morel species, thereby producing new insights.