J. Microbiol. Biotechnol. (2017), 27(7), 1223–1232 https://doi.org/10.4014/jmb.1701.01043 Research Article Review jmb

Characterization of Newly Bred militaris Strains for Higher Production of Cordycepin through HPLC and URP-PCR Analysis Hyun-Hee Lee1†, Naru Kang1,2†, Inmyoung Park1,3†, Jungwook Park1, Inyoung Kim1, Jieun Kim1, Namgyu Kim1, Jae-Yun Lee2, and Young-Su Seo1*

1Department of Microbiology, Pusan National University, Busan 46241, Republic of Korea 2CMG Biofarm, Busan 46958, Republic of Korea 3Department of Asian Food and Culinary Arts, Youngsan University, Busan 48015, Republic of Korea

Received: January 16, 2017 Revised: April 19, 2017 , a member of , a mushroom referred to as caterpillar Dong- Accepted: April 23, 2017 chung-ha-cho, is commercially valuable because of its high content of bioactive substances, including cordycepin, and its potential for artificial cultivation. Cordycepin (3’- deoxyadenosine) is highly associated with the pharmacological effects of C. militaris.

First published online C. militaris is heterothallic in that two mating-type loci, idiomorph MAT1-1 and MAT1-2, exist April 25, 2017 discretely in two different spores. In this study, nine C. militaris strains were mated with each

*Corresponding author other to prepare newly bred strains that produced a larger amount of cordycepin than the Phone: +82-51-510-2267; parent strains. Nine strains of C. militaris were identified by comparing the internal Fax: +82-51-514-1778; transcribed spacer sequence, and a total of 12 single spores were isolated from the nine strains E-mail: [email protected] of C. militaris. After the MAT idiomorph was confirmed by PCR, 36 mating combinations were † These authors contributed performed with six single spores with MAT1-1 and the others with MAT1-2. Eight mating equally to this work. combinations were successfully mated, producing stroma with perithecia. Cordycepin content analysis of all strains by high-performance liquid chromatography revealed that the KASP4- bred strain produced the maximum cordycepin among all strains, regardless of the medium and stroma parts. Finally, universal rice primer–PCR was performed to demonstrate that the bred strains were genetically different from the parental strains and new C. militaris strains. These results may be related to the recombination of genes during mating. The newly produced strains can be used to meet the industrial demand for cordycepin. In addition, breeding through mating suggests the possibility of producing numerous cordycepin- pISSN 1017-7825, eISSN 1738-8872 producing C. militaris strains. Copyright© 2017 by The Korean Society for Microbiology Keywords: Cordyceps militaris, cordycepin, MAT gene, mating, Dong-chung-ha-cho and Biotechnology

Introduction and germination occurs within a few hours. After several weeks, the species grows continuously and creates stroma The Cordyceps is an entomopathogenic and supported by the host exocuticula [3]. endoparasitic that belongs to the division Ascomycota, Many Cordyceps spp. have been used as health foods or class , order , and family medicines in and South-East Asia for a long time. Clavicipitaceae and contains more than 750 species [1]. Cordyceps sinensis contains numerous bioactive substances Cordyceps spp. are mainly distributed in subtropical regions [4], but is very costly because it is extremely difficulty to with high temperatures and humidity and parasitize artificially culture. The demand for C. militaris has gradually or from lepidopteran larva and pupae to increased because of artificial culture and its numerous imago to obtain nourishment from the host [2]. Infection bioactive substances. The bioactive substances in C. militaris begins by the adherence of conidia on the ’s exoskeleton include polysaccharides, ergosterol, cordycepic acid, and

July 2017 ⎪ Vol. 27⎪ No. 7 1224 Lee et al.

Fig. 1. PCR and maximum-likelihood phylogenetic analysis of the partial ITS1-5.8S-1TS2-partial 28S rDNA sequence among Cordyceps spp. (A) ITS sequences including 5.8S rRNA were amplified from 10 strains. The positive control was C. militaris (EFCC-C738). (B) The maximum- likelihood phylogenetic tree was created by the MEGA 6.06 program. Bootstrap values on branches are indicated for values greater than 50. The scale bar represents the degree of difference of the nucleotide sequences. cordycepin [5]. Among these, studies of C. militaris have type idiomorphs MAT1-1 and MAT1-2 exist discretely in focused on cordycepin [6, 7]. Cordyceps militaris that can be two different spores [17-19]. The MAT1-1 locus includes purchased in the market contains 300 mg of cordycepin per MAT1-1-1 and MAT1-1-2, and the MAT1-2 locus includes 100 g of total weight on average. The chemical structure of MAT1-2-1 [19]. The MAT idiomorph encodes the transcription cordycepin is very similar to that of adenosine, except that factor that regulates genes related to sexual reproduction the hydroxyl group of the 3’-carbon is not present in [20]. MAT1-1 and MAT1-2 each encode separate transcription cordycepin (Fig. 1). Cordycepin has many pharmacological factors containing DNA-binding motifs. MAT1-1 is a effects, including antitumor, anti-aging, antiviral, antibacterial, transcription factor containing an α-domain as the DNA- and anti-inflammatory activities [8-10]. Cordycepin, as a binding motif. MAT1-2 is a transcription factor harboring polyadenylation inhibitor, has antitumor activity by an HMG-domain as a DNA-binding motif [21]. Mating can inducing apoptosis and inhibiting the proliferation of be initiated when two single spores with different MAT cancer cells [11]. However, the biosynthetic pathway of idiomorphs meet through self- and non-self-recognition [22]. cordycepin of Cordyceps spp. remains unclear. It was When two single spores with different MAT idiomorphs hypothesized that genes involved in the adenine metabolic meet, pheromone receptors recognize the pheromones of pathway participate in cordycepin biosynthesis [2, 12]. C. militaris and the genes related with sexual development Studies to increase the natural cordycepin amount in are expressed by the MAT transcription factor. Until now, Cordyceps spp. have been conducted in order to produce mutations by UV and high-energy ion beam irradiation C. militaris containing a high cordycepin content, which is have been employed to create new C. militaris containing industrially valuable [13, 14]. high cordycepin contents [13]. However, the mating of Cordyceps militaris proliferates asexually by producing parental strains is the best way to improve the production conidia and sexually by producing [15, 16]. In of cordycepin, irrespective of controversial issues related to sexual reproduction, C. militaris exhibits heterothallism, in genetically modified organisms or food, because cordycepin that two mating-type loci consisting of the opposite mating- is used either as C. militaris or its extracts.

J. Microbiol. Biotechnol. Characterization of Newly Bred Cordyceps militaris Strains 1225

However, C. militaris containing these mutations are not dextrose broth (PDB) in a shaking-type photobioreactor (150 rpm) preferred because of ethical problems related to genetically at 25°C for 7 days. modified organisms or food. In this study, new C. militaris strains with high cordycepin contents were developed by PCR of rDNA Region Spanning the ITS1, ITS2, and 5.8S rRNA gene recombination via natural mating without mutation. Gene Genomic DNA of 15 C. militaris individuals was extracted from C. militaris strains were mated with other strains to increase the fresh mycelium by using the cetyltrimethylammonium cordycepin content. Nine strains of C. militaris were used to bromide method [23]. The ITS4R and ITS5F primers were used to perform sexual reproduction by mating. After single spores amplify the rDNA region spanning the ITS1, ITS2 (internal were separated from the mycelium or stroma of nine transcribed spacer), and 5.8S rRNA gene [24]. PCR was performed C. militaris strains, mating was performed for 50 days. The in a 20-µl reaction containing 2 µl 10× Taq PCR buffer, 1.6 µl cordycepin contents were compared between bred strains and dNTPs (2.5 mM stock), 1 µl Primer 1 (10 µM), 1 µl Primer 2 parent strains by high-performance liquid chromatography (10 µM), 0.1 µl Taq DNA polymerase (Takara, Japan), 1 µl DMSO, (HPLC) analyses. Finally, universal rice primer (URP)–PCR and 50 ng/µl template, using the Sure Cycler 8800 Thermal Cycler was performed to demonstrate that the bred strains with (Agilent Technologies, USA). The PCR was performed as follows: high cordycepin content are genetically different from the 96°C for 5 min, followed by 30 cycles of 96°C for 40 sec, 48°C for parent strains. This result suggests that breeding through 40 sec, and 72°C for 40 sec, with a final extension at 72°C for mating may lead to the development of new C. militaris 10 min. The amplified PCR products were purified using a PCR purification kit (GeneAll Biotechnology, Korea) and sequenced strains with higher production of cordycepin. using a 3730xl DNA analyzer (Macrogen, Inc., Korea). A phylogenetic tree was constructed for the nine C. militaris strains Materials and Methods and other Cordyceps species. All amplified sequences were registered with the National Center for Biotechnology (https:// Fungal Strains and Culture Conditions www.ncbi.nlm.nih.gov/). All sequences were aligned using MUSCLE Nine strains of C. militaris were used in this study. Each strain and curated with Gblocks (http://molevol.cmima.csic.es/castresana/ was obtained from the Korean Agricultural Culture Collection Gblocks/Gb, http://www.drive5.com/muscle). The phylogenetic (KACC), Systems Plant Microbiology Laboratory of Pusan National tree was constructed with MEGA ver. 6.06 using maximum University (SPNU), Entomopathogenic Fungal Culture Collection likelihood. Bootstrap analysis was conducted with 2,000 replicates. (EFCC), and Mycological Laboratory of Pusan National University (MPNU). Six strains were in the mycelium state (KACC44459, Single Spore Isolation and Identification of MAT Idiomorph by KACC50001, KACC44462, SPNU1000, SPNU1001, SPNU1005) and PCR the others were in the stroma state (SPNU1002, SPNU1003, For single spore isolation from stroma, the spore drop/shooting SPNU1004) (Table 1). Fungal strains were maintained on potato method was performed [25]. Stroma with perithecia was attached dextrose agar (PDA) at 25°C. For genomic DNA isolation, the to the inner side of a petri dish containing 2% water agar (WA) at mycelium in PDA was cultured for 7 days in 50 ml of potato 25°C under light for 5 days. To separate a single spore from the mycelium, the dilution and smearing method was performed. The mycelium was cultured in 50 ml of PDB containing glass beads in Table 1. Cordyceps militaris strains used in this study. a shaking incubator at 150 rpm for 7 days at 25°C under light. The Strains Source State supernatant of the culture medium was collected and filtered KACC44459 KACC Mycelium (haploid) through filter paper to remove the mycelium mass. After serially KACC50001 KACC Mycelium (haploid) diluting the supernatant containing conidia in water, 200 µl of the KACC44462 KACC Mycelium (haploid) conidia suspension was smeared in WA medium. The WA plates SPNU1000 SPNU Mycelium (haploid) were cultured at 25°C under light for 5 days to germinate single spores. To confirm the mating idiomorphs of single spores, MAT- SPNU1001 SPNU Mycelium (haploid) PCR was performed using MAT1-1-1 and MAT1-2-1 primers to SPNU1002 SPNU Stroma (diploid) target MAT1-1-1 and MAT1-2-1 (MAT1-1-1-F, 5’-ATGGAACAC SPNU1003 SPNU Stroma (diploid) AGATCGAGCGACAC-3’ and MAT1-1-1-R, 5’-ATATACCTTCGC SPNU1004 SPNU Stroma (diploid) GATCATTGCCCAG-3’; MAT1-2-1-F, 5’-TGTTTTGTCGCGATG SPNU1005 SPNU Mycelium (haploid) GTTCTGG-3’ and MAT1-2-1-R, 5’-CCTCTGGAGGTTCTGCAT EFCC-C738 EFCC Mycelium (haploid) TCCA-3’). MPNU8001-1 and MPNU8001-2 were used as a control strain by confirming the MAT idiomorph through MAT PCR and MPNU80001-1 MPNU Mycelium (haploid) sequencing (Table 1). The PCR was performed in a 20-µl reaction MPNU80001-2 MPNU Mycelium (haploid) consisting of 2 µl 10× Taq PCR buffer, 1.6 µl dNTPs (2.5 mM

July 2017 ⎪ Vol. 27⎪ No. 7 1226 Lee et al.

stock), 1 µl Primer 1 (10 µM), 1 µl Primer 2 (10 µM), 0.1 µl Taq AGGACTCGATAACAGGCTCC-3’) URP primers were used in DNA polymerase (Takara), 1 µl DMSO, and 50 ng/µl template this study [12, 26]. Template DNA from single spores of the mated (Agilent Technologies). The PCR was performed as follows: 95°C and parent strains were used. PCR was performed in a 20-µl for 1 min, followed by 30 cycles of 95°C for 30 sec, 58°C for 30 sec, reaction consisting of 2 µl 10× Taq PCR buffer, 1.6 µl dNTPs and 72°C for 40 sec, and with a final extension at 72°C for 5 min. (2.5 mM stock), 2 µl URP-Primer (10 mM), 0.1 µl Taq DNA polymerase (Takara), 1 µl DMSO, and 50 ng/µl template. The Mating Experiment and Stroma Induction URP-PCR conditions were as follows: one cycle for 5 min at 94°C, Nine parent strains were induced to produce stroma in brown 35 cycles for 1 min at 94°C, 1 min at 55°C, 2 min at 72°C, and a rice, silkworm pupae, and PDB medium. The inoculum was final extension for 5 min at 72°C. prepared in 50 ml of PDB in a 250 ml Erlenmeyer flask by inoculating multiple spores. Inocula were cultured in a shaking- Results type photobioreactor (120 rpm) at 25°C for 7 days. Brown rice medium was prepared by mixing 40 g brown rice and 4 g silkworm Identification and Phylogenetic Tree of Nine Strains pupae in 64 ml of liquid medium (20 g sucrose, 20 g peptone, 1 g The ITS region including 5.8S rRNA has been used as a MgSO ·7H O, 0.5 g KH PO , and distilled water to 1 L) in a 4 2 2 4 genetic marker to investigate the phylogenetic relationships polypropylene mushroom stroma bottle. The silkworm pupae of closely related fungal species. As mating is possible medium was prepared by mixing 20 g brown rice and 40 g silkworm within the same Cordyceps species, identification of nine pupae in 64 ml of liquid medium. Next, 15–20 ml of multiple spore inoculum set to 12 g/l in dry cell weight (DCW) was strains was performed by using ITS sequences including inoculated in all three media. The bred strain was also induced to 5.8S rRNA. ITS sequences including 5.8S rRNA of nine produce stroma in brown rice, silkworm pupae, and PDB C. militaris strains were successfully amplified using the medium. A single spore inoculum was prepared in PDB. Next, 15– ITS4F and ITS5R primers (Fig. 1). Amplified target sequences 20 ml of inoculum set to 12 g/l in DCW and composed of two were approximately 530–560 bp. As the sequences of all spores originating from different parent strains was inoculated strains showed 98% identity or more with the ITS sequence into all media. After incubation for 7 days in the dark at 25°C to including the 5.8S rRNA of known C. militaris (AF153264), promote vegetative growth, all media were incubated at 20°C for all tested strains were identified as C. militaris. A phylogenetic 50 days in a growth chamber (JEIO TECH, Korea) under a 12-h tree was created with nine strains of C. militaris and other light/dark cycle of 500–1,000 lux and high humidity (90%). All Cordyceps species by MEGA 6.06, using maximum-likelihood bred strains with stroma were photographed using a camera phylogenetic analysis to demonstrate that all tested strains (Nikon, Japan). belonged to C. militaris. The phylogenetic tree revealed that Cordycepin Analysis by High-Performance Liquid Chromatography 21 strains of C. militaris were genetically distant from other The average cordycepin content of the stroma and sclerotium Cordyceps spp. (Fig. 1). was set to the cordycepin content of each strain cultivated in brown rice and silkworm pupae medium. In the case of mycelium Isolation and Identification of MAT Idiomorph in Each in the PDB, the amount of cordycepin was measured by Single Spore separating the extracellular cordycepin from the culture medium Single spores were isolated from six C. militaris strains in and intracellular cordycepin in the mycelium. The samples were the mycelium state and three C. militaris strains in the freeze-dried and ground into a powder (diameter of approximate stroma state. Ascospores were dropped from the stroma by 50 mesh) in liquid nitrogen, followed by heating at 110°C for 1 h the drop/shooting method and germinated in WA after in H O except for the PDB culture medium. Samples were filtered 2 5 days. The conidia suspension prepared using the dilution through a 0.22-µm syringe filter (Chromdisc; E.Chrom Science, method was germinated in WA after 5 days. Single spore Inc., Korea). HPLC analysis was performed on a Shiseido isolation for all strains was repeated five times, to isolate a Nanospace SI-2 HPLC (Japan) with a Shiseido CAPCELL PAK C18 AQ Column (250 × 4.6 mm, 5 µm). The standard for cordycepin total of 5 × 9 = 45 single spores of C. militaris. The MAT was purchased from Sigma (USA) and injected at five sample idiomorph of a C. militaris single spore was identified by volumes of 7.8125–500 µg/ml to draw the calibration curve. PCR amplification using specific MAT primers. MAT1-1-1, Detection was performed at a mobile phase ratio of water: a partial gene of MAT1-1, was amplified by specific MAT1- acetonitrile = 92: 8 (v/v) and flow rate of 500 µl/ml at 30°C. The 1-1 primers. MAT1-2-1, a partial gene of MAT1-2, was injection volume was 10 µl and detection wavelength was 260 nm. targeted by specific MAT1-2-1 primers. The PCR product of partial MAT1-1-1 was 457 bp, and that of partial MAT1-2-1 URP–PCR was 368 bp. The PCR results of 45 single spores indicated The 2R (5’-CCCAGCAACTGATCGCACAC-3’) and 6R (5’- that 12 single spores were isolated from the nine parent

J. Microbiol. Biotechnol. Characterization of Newly Bred Cordyceps militaris Strains 1227

spores, six single spores were isolated from six parent C. militaris strains in the mycelium state. The other six single spores were isolated from three parent C. militaris strains in the stroma state.

Mating and Stroma Induction Mating was performed based on 12 single spores composed of six single spores with MAT1-1 and six single spores with MAT1-2 by induction of stroma in three replicate media. Mating of 36 combinations was performed through stroma induction in all media. Eight of the 36 combinations were successfully mated, producing stroma with perithecia in brown rice medium 50 days after inoculation (Fig. 3). Some combinations did not produce stroma, while the other Fig. 2. Identification of MAT idiomorphs of single spores combinations produced abnormal stroma without perithecia. derived from either stroma or mycelium by MAT-PCR. Multi spores isolated from the parent strain were also The MAT1-1-1 partial gene of MAT1-1, and MAT1-2-1 partial gene of induced to stroma in all media. Only a few multi spores MAT1-2 were amplified. Positive control and negative control used isolated from SPNU1002, SPNU1003, and SPNU1004 single spores (MPNU80001-1, MPNU80001-2), which are known as produced the stroma in brown rice medium. No multi MAT idiomorphs. (A) MAT1-1-1 was successfully amplified in the spores from the parent strain could produce stroma in MAT1-1 idiomorph with six single spores. (B) MAT1-2-1 was silkworm pupae and PDB media. Finally, eight newly bred amplified in the MAT1-2 idiomorph with six single spores. strains were developed (Table 2). strains. Six single spores with MAT1-1 and six single spores Cordycepin Content in Brown Rice Medium with MAT1-2 were isolated (Fig. 2). Among the 12 single The cordycepin content of the parent and bred strains

Fig. 3. Eight mated C. militaris strains in brown rice medium.

July 2017 ⎪ Vol. 27⎪ No. 7 1228 Lee et al.

cultivated in brown rice medium was measured. The medium was measured in triplicate. The average content of retention time value of standard cordycepin was 13.29 min cordycepin in the nine parent strains was 2.13 mg/g. under optimum chromatographic conditions (Fig. 4). The Among the top three strains of parental strains (SPNU1001, calibration curve was plotted with standard sample data SPNU1002, SPNU1005), SPNU1002 produced the maximum measured in triplicates from 7.8125 to 500 μg/ml. The cordycepin (5.64 mg/g), followed by SPNU1001 (3.17 mg/g) standard curve equation of cordycepin was Y = 3.38E+10X (by ANOVA) (Fig. 5). The average content of cordycepin in - 833885.435 (R2 = 0.9997). The content of cordycepin in the the eight bred strains was 3.75 mg/g. The bred strain nine parent strains and eight bred strains in brown rice producing the highest cordycepin content was KASP4

Table 2. List of C. militaris bred strains generated through mating in this study. MAT 1-1 MAT 1-2 Bred strain SPNU1002 SPNU1000 KASP1 SPNU1003 SPNU1000 KASP2 SPNU1002 SPNU1001 KASP3 SPNU1005 SPNU1001 KASP4 SPNU1003 SPNU1002 KASP5 KACC44459 SPNU1003 KASP6 SPNU1002 SPNU1003 KASP7 SPNU1005 SPNU1003 KASP8

Fig. 5. Cordycepin content in brown rice medium. Cordycepin content of nine parent strains (A) and eight bred strains (B) in brown rice medium. Error bars indicate the standard deviation Fig. 4. Representative HPLC profiles of cordycepin. of three independent experiments. Different letters above the (A) Standard cordycepin sample (62.5 µg/ml). (B) Hot water extracts cordycepin content bar indicate significant differences according to of KASP4. The retention time of cordycepin was 13.29 min. strain (p < 0.001, one-way ANOVA followed by Duncan’s test).

J. Microbiol. Biotechnol. Characterization of Newly Bred Cordyceps militaris Strains 1229

(7.71 mg/g) followed by KASP3 (5.11 mg/g) (by ANOVA) Table 3. Average cordycepin contents in all tested strains (Fig. 5). Comparison of the cordycepin content among the based on growth in different media. top three bred strains and top three parent strains based on Medium Cordycepin content (mg/g) the results of cordycepin content in all strains revealed that Brown rice 2.89 ± 1.99b KASP4 produced the largest amount of cordycepin in Silk worm pupae 4.37 ± 2.32a brown rice medium. KASP4 produced more cordycepin PDB 1.16 ± 1.23c than the parent strain SPNU1002. The cordycepin content *F-value (P) - 36.296 (.000***). in KASP4 was increased by 36% compared with that in SPNU1002, producing the largest amount of cordycepin among the parent strains in brown rice. Production of Cordycepin from Fungal Sclerotium and Stroma Production of Cordycepin from Different Culture Media The cordycepin content of the stroma was compared The cordycepin contents of all strains cultivated in with that of the sclerotium in the tested strains cultivated brown rice, PDA, and silkworm media were analyzed. The in brown rice medium. First, the cordycepin content in the cordycepin contents of the top three strains among the stroma was analyzed among all bred strains and three parent strains and top three strains among the bred strains parent strains (SPNU1002, SPNU1003, and SPNU1004). were determined among all tested strains (Fig. 6). The top Although the content differed according to stroma parts, three strains were determined by calculating the average KASP3 strain showed the highest cordycepin content (by cordycepin content of strains cultivated in all media. ANOVA) (Fig. 7). In the range of cordycepin based on KASP4 produced more cordycepin than SPNU1002 in not stroma development, 5.43 mg/g of cordycepin content was only brown rice, but also PDB and silkworm pupae media measured at 2 cm length of stroma, whereas 3.78 and (by ANOVA), demonstrating that KASP4 produced the 3.61 mg/g of cordycepin contents were measured at 5 and maximum cordycepin among all strains. Additionally, 10 cm length of stroma, respectively. The cordycepin C. militaris strains produced the maximum cordycepin in contents were higher at the early stage (2 cm length) than silkworm pupae medium among the three different media later stages of stroma development. In addition, KASP3 (by ANOVA) (Table 3). showed higher productivity of cordycepin compared with that in the sclerotium of the parent strain (Fig. 7). The

Fig. 6. Influence of medium on cordycepin production. The cordycepin content of the top three mated and top three parental Fig. 7. Cordycepin production in fungal sclerotium and strains is indicated according to media. Bred strain KASP4 produced stroma. the highest cordycepin content in all media. Error bars indicate the The cordycepin content is indicated in sclerotium and stroma parts standard deviation of cordycepin content. Different letters above the among the three parent strains and all bred strains. Different letters cordycepin content bar indicate significant difference according to above the cordycepin content bar indicate differences according to the strain and indicate independent dependence on media (p < 0.001, strain and independent dependence on stroma part (p < 0.001, one- one-way ANOVA followed by Duncan’s test). way ANOVA followed by Duncan’s test).

July 2017 ⎪ Vol. 27⎪ No. 7 1230 Lee et al.

Table 4. Average cordycepin contents in stroma and sclerotium of all tested strains. Stroma part Cordycepin contenta (mg/g) Sclerotium 4.34 ± 2.40 Stroma 2.66 ± 1.7 aCordycepin contents in 2 cm length of stroma. *F-value(P) – 4.175 (.045**). sclerotium generally produced more cordycepin than stroma in all tested strains (t-test, p < 0.001) (Table 4).

Identification of Newly Bred Strain by URP-PCR URP-PCR was performed to confirm whether KASP3 producing the most cordycepin was a newly bred strain with a genotype different from that of the parental strains. Polymorphism in the URP-PCR patterns among the parent Fig. 8. URP-PCR analysis of KASP4. strains (SPNU1001, SPNU1005) and bred strain (KASP4) SPNU1001 and SPNU1005 are parent strains of KASP4. (A) URP-PCR were investigated. Polymorphisms among the mating using a 6R primer and (B) a 2R primer. strains and parental strains were the most distinctive when the 2R primer and 6R primer were used (Fig. 8). The URP- PCR results indicated that the amplified band patterns of were isolated from the parent strains in the stroma state. As the parental strains SPNU1001 and SPNU1005, and KASP4 the proportion of ascospores with MAT1-1 is generally the were clearly different. In the URP-PCR results using the 2R same as that of ascospores with MAT1-2 in the of the primer, although polymorphism between SPNU1001 and perithecia, two MAT idiomorph ascospores were isolated SPNU1005 was not clear, polymorphisms between the from one strain in the stroma state. However, only one KASP4 and KASP4 parent strains (SPNU1001, SPNU1005) MAT1-1 or MAT1-2 idiomorph conidium could be isolated were clearly observed. When PCR was performed using the from the mycelium. This may be related to the properties of 6R primer, differences in the amplified band pattern parent strains in the mycelium state. First, the conidium among the three strains (SPNU1001, SPNU1005, KASP4) originating from the mycelium germinated soon after the were apparent. The URP-PCR results demonstrated that conidium was smeared onto WA medium. Short mycelium KASP4 has a genotype different from that of SPNU1001 generally overlapped with other MAT idiomorph mycelium, and SPNU1005. resulting in the isolation of multiple spores rather than a single spore [2, 16]. As ascospores isolated from the stroma Discussion germinated soon after application to the WA medium, they are less likely to overlap with other MAT idiomorph Cordyceps militaris is well known for its cordycepin mycelia. Second, when MAT-PCR was performed with content, which has medicinal effects, including antitumor, parental strains in the mycelium state, the ratio of MAT1-1 antiaging, antiviral, and antileukemia activities. Studies to idiomorph conidium in the parent strains was not the same increase cordycepin productivity of C. militaris by changing as the ratio of MAT1-2 idiomorph conidium. the media conditions and mutation of genes related to Mating was performed by inoculating isolated single cordycepin synthesis have been performed [13, 14]. In this spores in brown rice media. Eight combinations were mated study, C. militaris with high cordycepin was developed successfully among 35 combinations. Not all combinations by mating. Because it remains unclear which mating were mated, likely because not only the combination of combination increases the cordycepin content, the mating different mating types of single spores but also other combination was randomly performed to form newly bred factors are required to induce the sexual cycle by mating. strains with higher cordycepin content than the parent strain. For example, fungal activity and expression of genes related Twelve single spores were isolated from nine parent to the sexual cycle are likely correlated. The reduction strains. Both MAT1-1 and MAT1-2 idiomorph ascospores of fungal activity was associated with degeneracy that

J. Microbiol. Biotechnol. Characterization of Newly Bred Cordyceps militaris Strains 1231

decreased gene expression related to the sexual cycle [27]. stroma parts indicated that the content in the sclerotium of Mating of single spores isolated from the mycelium did C. militaris was more than in the stroma of C. militaris [34]. grow well because of the low activity of single spores In summary, KASP4 is a newly bred strain, as validated isolated from the mycelium. However, mating with by URP-PCR. The phenotype of stroma with perithecia was ascospores showed better results than with single spores the first evidence demonstrating the creation of a newly isolated from the mycelium. Differences in phenotypes bred strain by mating. Compared with the parent strains, a between abnormal stroma and normal stroma were combination of different genotypes by URP-PCR with attributed to differences in signaling pathways. If the higher cordycepin production in KASP4 confirmed the newly activity of a single spore was good and not degenerated, bred strain. The industrial value of the new C. militaris the spore may have induced the stroma without perithecia strains is increased because of higher cordycepin content. via activation of somatic cell division by the MAPK This result indicates that mating can be used to produce a pathway, because the normal stroma of bred strains with novel and high-quality strain with high cordycepin content. perithecia is induced by the PKA pathway [19, 28, 29]. Additionally, this will help to increase the industrial utility A comparison of cordycepin content revealed that KASP4 value of cordycepin as a medicinal and functional material. produced the maximum cordycepin among the tested strains. The parent strain of KASP4 was not SPNU1002, Acknowledgments which showed the highest cordycepin production among the parent strains. Some bred strains from SPNU1002 This research was supported by the Basic Science Research produced lower cordycepin content compared with Program through the National Research Foundation of SPNU1002. These results indicate that the cordycepin Korea (NRF) funded by the Ministry of Science, ICT & content in bred strains was not directly related to that in Future Planning (NRF-2014R1A1A2058139). the parent strains. Gene recombination caused by generation from the stroma resulted in the upregulation or References downregulation of genes related to cordycepin biosynthesis. Thus, the cordycepin content of ascospores isolated from 1. Sung GH, Hywel-Jones NL, Sung JM, Luangsa-Ard JJ, the parent strains was not always proportional to that of Shrestha B, Spatafora JW. 2007. Phylogenetic classification of the parental strains. In addition, mating activates the Cordyceps and the clavicipitaceous fungi. Stud. Mycol. 57: 5-59. signaling pathway involved in cordycepin synthesis, and 2. Shrestha B, Zhang W, Zhang Y, Liu X. 2012. The medicinal the cordycepin content in the diploid state is higher than fungus Cordyceps militaris: research and development. Mycol. Prog. 11: 599-614. that in the haploid state [30]. 3. Spatafora JW, Sung GH, Sung JM, Hywel-Jones NL, White Analysis of the influence of the medium on cordycepin JF Jr. 2007. Phylogenetic evidence for an animal pathogen production indicated that the cordycepin content in origin of ergot and the grass endophytes. Mol. Ecol. 16: silkworm pupae medium was the highest among the three 1701-1711. media tested, although the incidence of stroma in silkworm 4. Liu Y, Wang J, Wang W, Zhang H, Zhang X, Han C. 2015. pupae medium was lower than that in brown rice medium. The chemical constituents and pharmacological actions of This is correlated with the high protein and low carbohydrate Cordyceps sinensis. Evid. Based Complement. Alternat. Med. contents in the silkworm pupae medium [31]. Because the 2015: 575063. component of silkworm pupae and brown rice media used 5. Hur H. 2008. Chemical ingredients of Cordyceps militaris. in this study were similar except for the protein and Mycobiology 36: 233-235. carbohydrate components, these factors may prevent the 6. Lin Q, Long L, Wu L, Zhang F, Wu S, Zhang W, Sun X. stroma from developing and producing high levels of 2016. Evaluation of different agricultural wastes for the production of fruiting bodies and bioactive compounds by cordycepin [32, 33]. It was previously reported that the medicinal mushroom Cordyceps militaris. J. Sci. Food Agric. carbohydrate content of the optimum medium used for DOI: 10.1002/jsfa.8097. C. militaris stroma had a higher protein content [14]. To 7. Zhou X, Cai G, He Y, Tong G. 2016. Separation of obtain only cordycepin from C. militaris, the use of a cordycepin from Cordyceps militaris fermentation supernatant medium with relatively high pupae is necessary. The PDB using preparative HPLC and evaluation of its antibacterial medium showed low stroma and cordycepin productivity activity as an NAD+-dependent DNA ligase inhibitor. Exp. because of its lower nutrient content compared with other Ther. Med. 12: 1812-1816. media. Analysis of the cordycepin content according to 8. Ahn YJ, Park SJ, Lee SG, Shin SC, Choi DH. 2000.

July 2017 ⎪ Vol. 27⎪ No. 7 1232 Lee et al.

Cordycepin: selective growth inhibitor derived from liquid 21. Wang H, Wei J, Lin N, Feng A, Chen M, Bao D. 2010. culture of Cordyceps militaris against Clostridium spp. J. Distribution of mating-type genes in fruiting and non- Agric. Food Chem. 48: 2744-2748. fruiting forms of Cordyceps militaris. Acta Edulis Fungi 17: 1-4. 9. Muller WE, Weiler BE, Charubala R, Pfleiderer W, Leserman 22. Jones SK Jr. Bennett RJ. 2011. Fungal mating pheromones: L, Sobol RW, et al. 1991. Cordycepin analogues of 2’,5’- choreographing the dating game. Fungal Genet. Biol. 48: 668-676. oligoadenylate inhibit human immunodeficiency virus infection 23. Doyle J. 1991. DNA protocols for plants, pp. 283-293. In via inhibition of reverse transcriptase. Biochemistry 30: 2027- Godfrey MH, Andrew WB, Johnson J, Peter WY (eds.), 2033. Techniques in . Springer-Verlag, Berlin-Heidelberg. 10. Ramesh T, Yoo SK, Kim SW, Hwang SY, Sohn SH, Kim IW, Germany. Kim SK. 2012. Cordycepin (3’-deoxyadenosine) attenuates 24. Park J-E, Kim G-Y, Park H-S, Nam B-H, An W-G, Cha J-H, age-related oxidative stress and ameliorates antioxidant et al. 2001. Phylogenetic analysis of caterpillar fungi by capacity in rats. Exp. Gerontol. 47: 979-987. comparing ITS1-5.8S-ITS2 ribosomal DNA sequences. 11. Tian X, Li Y, Shen Y, Li Q, Wang Q, Feng L. 2015. Mycobiology 29: 121-131. Apoptosis and inhibition of proliferation of cancer cells 25. Choi Y-W, Hyde KD, Ho W. 1999. Single spore isolation of induced by cordycepin. Oncol. Lett. 10: 595-599. fungi. Fungal Divers. 3: 29-38 12.Xiang L, Li Y, Zhu Y, Luo H, Li C, Xu X, et al. 2014. 26. Kang H-W, Park D-S, Go S-J, Eun M-Y. 2002. Fingerprinting Transcriptome analysis of the sinensis fruiting of diverse genomes using PCR with universal rice primers body reveals putative genes involved in fruiting body generated from repetitive sequence of Korean weedy rice. development and cordycepin biosynthesis. Genomics 103: Mol. Cells 13: 281-287. 154-159. 27. Shrestha B, Park Y-J, Han S-K, Choi S-K, Sung J-M. 2004. 13. Das SK, Masuda M, Hatashita M, Sakurai A, Sakakibara M. Instability in in vitro fruiting of Cordyceps militaris. J. 2008. A new approach for improving cordycepin productivity Mushroom Sci. Prod. 2: 140-144. in surface liquid culture of Cordyceps militaris using high-energy 28.Lu Y, Xia Y, Luo F, Dong C, Wang C. 2016. Functional ion beam irradiation. Lett. Appl. Microbiol. 47: 534-538. convergence and divergence of mating-type genes fulfilling 14.Wen TC, Li GR, Kang JC, Kang C, Hyde KD. 2014. in Cordyceps militaris. Fungal Genet. Biol. 88: 35-43. Optimization of solid-state fermentation for fruiting body 29. Poggeler S, Nowrousian M, Ringelberg C, Loros JJ, Dunlap JC, growth and cordycepin production by Cordyceps militaris. Kuck U. 2006. Microarray and real-time PCR analyses reveal Chiang Mai J. Sci. 41: 858-872. mating type-dependent gene expression in a homothallic 15.Shrestha B, Han SK, Sung JM, Sung GH. 2012. Fruiting fungus. Mol. Genet. Genomics 275: 492-503. body formation of Cordyceps militaris from multi-ascospore 30. Zheng ZL, Qiu XH, Han RC. 2015. Identification of the genes isolates and their single ascospore progeny strains. Mycobiology involved in the fruiting body production and cordycepin 40: 100-106. formation of Cordyceps militaris fungus. Mycobiology 43: 37-42. 16. Shrestha B, Han SK, Yoon KS, Sung JM. 2005. Morphological 31.Guo M, Guo S, Huaijun Y, Bu N, Dong CH. 2016. characteristics of conidiogenesis in Cordyceps militaris. Comparison of major bioactive compounds of the caterpillar Mycobiology 33: 69-76. medicinal mushroom, Cordyceps militaris (ascomycetes), fruiting 17. Shrestha B, Kim H-K, Sung G-H, Spatafora JW, Sung J-M. bodies cultured on wheat substrate and pupae. Int. J. Med. 2004. Bipolar heterothallism, a principal mating system of Mushrooms 18: 327-336. Cordyceps militaris in vitro. Biotechnol. Bioprocess Eng. 9: 440-446. 32. Cho S-M, Park H-J, Seo G-S, Hong J-D. 2009. Effect of media 18. Yokoyama E, Arakawa M, Yamagishi K, Hara A. 2006. composition on the cordycepin and content nutritional Phylogenetic and structural analyses of the mating-type loci components of Cordyceps militaris. Kor. J. Mycol. 37: 161-166. in Clavicipitaceae. FEMS Microbiol. Lett. 264: 182-191. 33. Wen T-C, Kang J-C, Hyde KD, Li G-R, Kang C, Chen X. 19. Zheng P, Xia Y, Xiao G, Xiong C, Hu X, Zhang S, et al. 2011. 2014. Phenotypic marking of Cordyceps militaris fruiting- Genome sequence of the insect pathogenic fungus Cordyceps bodies and their cordycepin production. Chiang Mai J. Sci. militaris, a valued traditional Chinese medicine. Genome Biol. 41: 846-857. 12: R116. 34. Masuda M, Urabe E, Honda H, Sakurai A, Sakakibara M. 20. Ait Benkhali J, Coppin E, Brun S, Peraza-Reyes L, Martin T, 2007. Enhanced production of cordycepin by surface culture Dixelius C, et al. 2013. A network of HMG-box transcription using the medicinal mushroom Cordyceps militaris. Enzyme factors regulates sexual cycle in the fungus Podospora anserina. Microb. Technol. 40: 1199-1205. PLoS Genet. 9: e1003642.

J. Microbiol. Biotechnol.