Development of new strains and related SCAR markers for an edible mushroom, Hypsizygus marmoreus


Correspondence: Hyeon-Su Ro, Department of Microbiology and Research Institute of Life Sciences, Gyeongsang National University, 900 Gajwa-Dong, Chinju, Korea. Tel.: +82 55 7515946; fax: +82 55 7590187; e-mail:


New fast-growing and less bitter varieties of Hypsizygus marmoreus were developed by crossing monokaryotic mycelia from a commercial strain (Hm1-1) and a wild strain (Hm3-10). Six of the better tasting new strains with a shorter cultivation period were selected from 400 crosses in a large-scale cultivation experiment. We attempted to develop sequence characterized amplified region (SCAR) markers to identify the new strain from other commercial strains. For the SCAR markers, we conducted molecular genetic analysis on a wild strain and the eight most cultivated H. marmoreus strains collected from various areas in East Asia by randomly amplified polymorphic DNA. Ten unique DNA bands for a commercial Hm1-1 strain and the Hm3-10 strain were extracted and their sequences were determined. Primer sets were designed based on the determined sequences. PCR reactions with the primer sets revealed that four primer sets successfully discriminated the new strains from other commercial strains and are thus suitable for commercial purposes.


Hypsizygus marmoreus is one of the major mushroom products in East Asia. In most commercial farms, semi-automatic cultivation of this mushroom occurs in a solid substrate in wide-mouth polypropylene bottles (Lee et al., 2009). Many commercial farms have produced various versions of H. marmoreus with their own strains and varieties as spawn. Strains of high commercial value spread to farms in different areas and are sold under different names. This causes confusion among mushroom growers and consumers. Therefore it is important to develop new commercial strains and methods to verify them.

Breeding of edible mushrooms is carried out by hyphal fusion of monokaryotic mycelia, which are derived from basidiospores. Mating of tetrapolar mushrooms is regulated by two mating-type loci. The A locus contains homeodomain family transcription factor genes HD1 and HD2, and the B locus contains genes for pheromone receptors and pheromones (Kronstad & Staben, 1997; Brown & Casselton, 2001). Compatible pairings of genes at both loci are essential for successful mating. Because mating involves two genes in two loci, the theoretical frequency of compatible mating is 25%. However, because the genes in both loci are multi-allelic, the mating frequency can far exceed 25% (Brown & Casselton, 2001). The compatibility study on the mating of Pleurotus tuberregium from different geographic origins showed that the mating frequency could reach as high as 84% (Isikhuemhen et al., 2000). Recent comparative genomic analysis of mating-type loci of Flammulina velutipes also showed that the multi-allelic nature of mating genes depends on geographical distribution (Van Peer et al., 2011). This emphasizes the importance of geographic and genetic diversity of parental strains in the breeding of mushroom strains.

Verification of fungi has been done either by the comparison of a few selected marker DNA sequences, such as small subunit ribosomal DNA (SSU rDNA) (Berbee & Taylor, 1992), internal transcribed spacer (Chen et al., 2001) and multiple nuclear genes (Hansen et al., 2005), or by PCR-based DNA fingerprinting with various methodologies, including randomly amplified polymorphic DNA (RAPD; Lopandic et al., 2005), amplified fragment length polymorphism (Vos et al., 1995), and inter-simple sequence repeat PCR (Nazrul & Bian, 2010). In general, sequence-based approaches have been employed for the verification of phylogenetic relationships of fungal groups of different species. However, they have been unsuccessful in resolving the strains and varieties of the same species (Zhang et al., 2005). In contrast, PCR-based DNA fingerprinting has successfully differentiated strains of Pleurotus eryngii (Ro et al., 2007) and Fellomyces (Lopandic et al., 2005). This method, however, has inherent limitations in its reproducibility because it uses short random primers, which makes the PCR sensitive to the reaction components, including buffer, thermostable polymerase, and annealing temperature. Sequence characterized amplified region (SCAR) marker is a RAPD-derived DNA marker that overcomes the limitations of RAPD by using longer primers designed from the sequence of an extracted unique DNA band in the RAPD gel. SCAR markers have been employed for the identification of F. velutipes (Su et al., 2008), Lentinula edodes (Tanaka et al., 2004), and Laccaria bicolor (Weber et al., 2002).

Accordingly, this study reports the generation of new hybrid strains by basidiospore mating and the development of SCAR markers for the identification of the generated H. marmoreus strains.

Materials and methods

Strains and culture conditions

Hypsizygus marmoreus strains Hm0-4 and Hm2-10 were from the Green Peace Mushroom Research Institute (GPMI). Strains Hm0-7 and Hm1-1 were collected from Japan. Hm1-6 and Hm2-7 were from China and Taiwan, respectively. Hm3-6 and Hm3-8 were from the National Institute of Agricultural Science and Technology (NIAST), Korea. Hm3-10 was collected from Deog-Yu mountain, Korea. All strains were maintained on mushroom complete media by periodic transfer. To evaluate the fruiting body cultivation characteristics of the strains, the strains were cultivated with the substrate consisting of pine sawdust (23%), corncob (32%), rice bran (32%), and soybean hull (22%). The cultivation of H. marmoreus was carried out at 15 °C in an incubating room with 3000–4000 mg L−1 CO2 and 95% relative humidity. To extract mushroom total cellular DNA, the mushroom mycelia were grown in a potato dextrose broth (Ventech Bio Co., Korea) containing potato starch (4 g L−1) and glucose (20 g L−1) for 3 weeks at 25 °C.

Total cellular DNA extraction and RAPD analysis

Total cellular DNA extraction and RAPD analysis were performed as previously described (Ro et al., 2007). For the RAPD analysis, primers OPS-1 (5′-CTA CTG CGC T-3′), OPS-10 (5′-ACC GTT CCA G-3′), and OPL-13 (5′-ACC GCC TGC T-3′) were employed for the random amplification of mushroom genomic DNA fragments. PCR was conducted with following conditions: 94 °C for 5 min; 35 cycles at 94 °C for 45 s, 45 °C for 45 s, and 72 °C for 2 min; 72 °C for 10 min. Cluster analysis of the pattern of DNA bands was performed by the unweighted pair-group method with arithmetic average (UPGMA) methodology reassembled with 1000 repeats of Jackknifing, as described previously (Ro et al., 2007).

Development of new hybrid strains

Breeding was conducted by mating the basidiospores from two parental strains. Spores of parental strains were spread on a potato-dextrose agar (PDA) plate. Single colonies were picked from each parental strain and cultured on separate PDA plates until mycelia were fully developed. The 20 fastest growing independent mycelia with distinct colony shapes were selected from each parental strain. Mating was conducted by placing mycelial blocks (3 × 3 mm) from opposite strains on the same PDA plate 1 cm apart. Mating was confirmed by the formation of clamp connections under the microscope after incubation at 30 °C for 7 days.

Development of strain-specific SCAR markers

RAPD analysis has shown some strain-specific DNA bands in the gel. To develop the unique DNA bands as the strain-specific DNA markers, the DNA bands were excised and extracted with a DNA gel extraction kit (Solgent Co., Korea). The extracted DNA was cloned into pGEM-T-easy cloning vector (Promega Co., USA). The insert DNA sequence was determined by a commercial DNA sequencing service. The determined DNA sequences were deposited into GenBank (NCBI) with the accession numbers given in Table 1. Primer sets of 15 nucleotides in length were designed using the 5′- and 3′-ends of the determined sequences (Table 1). Target-specific primer sets were employed for the detection of specific strains using the following conditions: 94 °C for 5 min; 30 cycles at 94 °C for 45 s, 60 °C for 45 s, and 72 °C for 2 min; 72 °C for 10 min.

Table 1. Design of primer sets for the strain-specific SCAR markers
Primer setGenBank accession no.Size (bp)SequenceTm* (°C)
  1. NA, not available. *Tm: melting temperature.



Characterization of H. marmoreus strains

RAPD analyses with three random primers were conducted for the verification of nine H. marmoreus strains. The RAPD with primers OPS-1, OPS-10, and OPL-13 yielded 22, 16, and 21 distinct DNA bands, respectively, with the sizes ranging from 0.5 to 3.5 kbp (Fig. 1a). The DNA band pattern was clustered using the UPGMA method. The resulting dendrogram, which was a reflection of genetic background, showed that the H. marmoreus strains could be clustered into three groups (Fig. 1b). The largest group consisted of Hm0-7, Hm1-1, Hm1-6, Hm2-7, Hm3-6, and Hm3-8. Hm1-1 and Hm1-6 were essentially the same strain. Strains Hm3-6 and Hm3-8 are Korean varieties based on the Japanese strain Hm0-7. Taiwanese Hm2-7 could be derived from Hm0-7. Strains Hm0-4 and Hm2-10 were included in the second cluster. These strains were from a mushroom stock belonging to a commercial farm. Hm3-10 showed the most distinct DNA band pattern and thus formed an independent single-member group in the dendrogram. Hm3-10 was a wild strain collected from a mountain in the middle of Korea.

Figure 1.

Molecular genetic analysis of Hypsizygus marmoreus strains and morphology of selected strains. (a) OPS-1, OPS-10, and OPL-13 were employed for RAPD. Numbers on top of each lane are the strain number. The distinct DNA bands are marked with the white arrow head with the band number. (b) Dendrogram derived from the DNA band pattern. (c and d) Morphological characteristics of the wild strain (c, Hm3-10) and the cultivated strain (d, Hm1-1).

Cultivation characteristics of the strains were investigated in terms of mushroom yield, culture period, and taste of fruiting bodies. The results are summarized in the Table 2. Strains Hm1-1, Hm1-6, and Hm3-6 showed the best results among the strains. The former two strains were identical in RAPD and therefore had the same characteristics. The wild strain Hm3-10, which exhibited a distinct genetic background in the RAPD analysis, showed reasonable cultivation characteristics except for poor fruiting body yield, suggesting a potential to be developed as a commercial strain. The morphology of fully grown Hm3-10 and Hm1-1 are shown in Fig. 1c and d, respectively. In Hm3-10, the size of cap was slightly larger and the number of fruiting bodies was less than to Hm1-1, resulting in a lower yield.

Table 2. Cultivation characteristics of hybrid strains of Hypsizygus marmoreus
StrainYield (g)aCultivation period (days)TastebSource
  1. a

    Average weight of mushroom fruiting bodies with the standard error of ±3 g.

  2. b

    Taste of mushroom was rated by 10 selected people.

Hm0-417324++GPMI, Korea
Hm0-713025+++Kyoto, Japan
Hm1-1 (parental)19320+Fukuoka, Japan
Hm1-619420+Beijing, China
Hm2-1016023++GPMI, Korea
Hm3-619224++NIAST, Korea
Hm3-814322++NIAST, Korea
Hm3-10 (parental)12524+++GPMI, Korea

Development of new strains by basidiospore mating

Although Hm1-1 had good production yield and relatively short cultivation period, its commercial value is limited by its bitter taste. Hm3-10 has shown good potential as a commercial strain in terms of taste in spite of its lower yield and longer cultivation period. Therefore, we tried to develop new varieties of H. marmoreus with a better taste by mating these two strains. Basidiospores of Hm1-1 and Hm3-10 were collected and spread on a PDA plate. Twenty monokaryotic mycelia from each strain were selected on the basis of growth rate and mycelial growth pattern. Mating was conducted by placing the monokaryotic mycelial blocks of opposite strains on the same plate. The total number of mated mycelia was 400 (20 spores from Hm1-1 × 20 spores from Hm3-10). Of 400 mating pairs, 343 were observed to make clamp connections, an indication of successful mating. The mating frequency was 85.8%, which was unusually high for a tetrapolar mating system. The expected mating frequency in tetrapolar basidiomycetes is 25% (Kronstad & Staben, 1997). However, the mating of a species in a geographically distinct population could be compatible. For example, the compatibility of P. tuberregium, a tetrapolar mushroom, from a New Caledonia collection and a Nigeria or a Papua New Guinea collection was 83% or 84% (Isikhuemhen et al., 2000). Therefore, the unusual mating frequency of H. marmoreus strains is potentially due to geographic isolation.

The mated dikaryotic mycelia were cultivated on solid substrate, as described previously (Lee et al., 2009). Subsequently, 58 hybrid strains were initially screened in terms of production yield, shape of cap, and cultivation period. We chose six new hybrids with better taste and cultivation characteristics (Table 2). The selected strains Hm15-3, Hm15-4, Hm15-5, Hm16-1, Hm16-2, and Hm17-5 tasted better than parental Hm1-1 strain and had better production yield than Hm3-10 strain. Optimization of cultivation conditions may further increase yield and shorten the cultivation period.

Development of strain-specific DNA markers

RAPD analysis yielded multiple amplified DNA bands, some of which were unique for a certain strain (Fig. 1). To develop the strain-specific SCAR markers, we selected 10 distinct DNA bands from the three RAPD gels which were amplified with OPS-1, OPS-10, or OPL-13 primers (Fig. 1). Bands 1, 6, and 7 were unique for Hm1-1 and Hm1-6. Bands 2–5 and 8–10 were unique for Hm3-10. The selected DNA bands were cloned into a TA cloning vector and their sequences were determined. The sequences were deposited in GenBank and were used to design the 15-base primer sets using their 5′- and 3′-ends (Table 1).

The specificity of the primer sets was investigated by PCR with an elevated annealing temperature (60 °C). As shown in Fig. 2a, the primer set P6, derived from a 755-bp DNA band of Hm1-1, was able to distinguish Hm3-10 from other strains. P9 and P10 were highly specific for Hm3-10, yielding 255- and 1666-bp PCR products, respectively. P8, which was from Hm3-8, produced a 454-bp DNA fragment only for Hm1-1, Hm1-6, Hm2-10, and Hm3-8. The primer sets P1-P5 and P7 produced DNA bands with corresponding sizes from all H. marmoreus strains and presented no strain specificity (only P3 data are shown in Fig. 2a). P1, P4, and P7 were polymorphic. The primer sets P6 and P8 could be employed for the specific detection of Hm1-1-related strains, while P9 and P10 were specific for Hm3-10.

Figure 2.

PCR with primer sets designed from the determined sequences. (a) Amplification of marker DNAs in cultivated strains of Hypsizygus marmoreus with sequence specific primer sets. (b) Amplification of marker DNAs in hybrid strains generated by monokaryotic mycelia mating of Hm1-1 and Hm3-10.

Marker specificity on hybrid strains

The specificities of the selected primer sets were challenged with the hybrid strains Hm15-3, Hm15-4, Hm15-5, Hm16-1, Hm16-2, and Hm17-5 (Fig. 2b). The P6 marker appeared only for Hm1-1, whereas the P8 marker appeared for most hybrid strains except Hm16-1 and the wild Hm3-10. This is interesting because the P6 marker showed broader specificity than the P8 marker in the identification of strains. The P9 marker appeared on Hm16-1, Hm16-2, Hm17-5, and Hm3-10 and the P10 marker appeared on Hm15-3, Hm15-4, Hm16-1, Hm16-2, and Hm3-10. The hybrid strain Hm15-3 was the only strain that did not contain either the P9 or the P10 marker.


Development of new strains and verification techniques are some of the major issues in mushroom technology. In this study, we crossed a commercial strain of H. marmoreus and a wild strain of H. marmoreus by monokaryotic mycelial mating. The wild strain (Hm3-10) showed distinct morphological and cultivation characteristics. Cultivated H. marmoreus strains originated largely from Japan, where this mushroom is the second most cultivated mushroom. Most of them were raised from a few Japanese parental strains and thus are closely related to each other. Dendrogram analysis based on RAPD demonstrates that cultivated strains can be categorized in two groups (Fig. 1b) and the genetic distance between the groups is closer than that to Hm3-10. Uniqueness of Hm3-10 was further evidenced by the mating experiment. Mating frequency between the commercial Hm1-1 and the wild Hm3-10 strain was 85.8%, which is unusually high for tetrapolar mating, indicating allelic diversification of the mating-type genes in the Korean strains. Similar results were reported in the mating of P. tuberregium from different geographic origins (Isikhuemhen et al., 2000).

RAPD is not, in general, a good method for identification and classification of fungi because of limitations in reproducibility. However, it can be a simple and powerful tool when it is used to make comparisons within a set of samples. It is also a useful tool to generate SCAR markers (Weber et al., 2002; Tanaka et al., 2004). In this work, the primer sets derived from distinct RAPD bands were successfully employed to discriminate specific strains. Our results showed that PCR reactions with the primer sets yielded strain-specific DNA bands, indicating that our strategy to develop SCAR marker is a reasonable approach.


This work was supported by Mushroom Export Research Program and Technology Development Program for Agriculture and Forestry, Ministry of Agriculture and Forestry, Korea.