2022 Volume 63 Issue 3 Pages 88-95
We performed in-vitro germination tests on seeds from five Gastrodia orchids (G. confusa, G. elata var. elata, G. elata var. pallens, G. nipponica, and G. pubilabiata) using one Marasmiaceae and two Mycena isolates. Mycena sp. 1 promoted germination of all five Gastrodia orchids, with root and/or tuber formation observed in G. confusa, G. nipponica, and G. pubilabiata. No additional growth was observed in the other two orchids. Mycena sp. 2 induced G. confusa, G. elata var. elata, and G. nipponica germination, whereas Marasmiaceae sp. 1 induced G. nipponica and G. pubilabiata germination. Phylogenetic analyses indicated that the two Mycena isolates represent distinct lineages within the Mycenaceae. Mycena sp. 1 and Marasmiaceae sp. 1 are closely related to Mycena abramsii and Marasmiellus rhizomorphogenus, respectively. Our results imply that Mycena and marasmioid fungi play important roles in early development in Gastrodia species, and that Mycena fungi in particular may be common mycobionts of Gastrodia species. Root and/or tuber development was observed with four plant-fungus combinations, implying that these associations persist throughout the life cycle, whereas G. elata var. elata may require different associates over time. Our findings will contribute to elucidating the mycorrhizal associations of mycoheterotrophic orchids throughout their life cycle.
Gastrodia is a genus of fully mycoheterotrophic orchids that depend entirely on mycorrhizal fungi for their carbon needs (Merckx, 2013). Gastrodia is among the largest mycoheterotrophic genera and includes about 100 species (Ogura-Tsujita, Yukawa, & Kinoshita, 2021). Fungal associates of Gastrodia species are highly divergent; they comprise mainly leaf-litter- and wood-decaying fungi, but also include ectomycorrhizal fungi and common orchid mycorrhizal fungi, rhizoctonias (Kusano, 1911; Martos et al., 2009; Kinoshita et al., 2016). Such mycorrhizal associations have been identified primarily by examining adult plants, whereas mycobionts associated with early developmental stages remain largely unknown.
All orchids produce fine, dust-like seeds containing few storage nutrients, and require carbon and nutrients from mycobionts during seed germination and seedling development (Rasumssen & Rasmussen, 2009). Mycobiont composition turnover during the transition from seedling to mature plant has been observed in many orchid species (Bidartondo & Read, 2008; Jacquemyn, Brys, Cammue, Honnay, & Lievens, 2011; Vogt-Schilb et al., 2020). Turnover scenarios vary among orchid species and include no turnover, nested gain/loss, and partial or total turnover (Ventre Lespiaucq, Jacquemyn, Rasmussen, & Méndez, 2021). Such turnover is likely due to differences in plant nutrient demands resulting from ontogeny or environmental changes, such as seasonal variation in temperature or water availability (Ventre Lespiaucq et al., 2021). In G. elata Blume, litter-decomposing fungi of the genus Mycena promote seed germination, whereas wood-decaying Armillaria species are associated with mature plants (Xu & Guo, 2000). Such ontogenetic turnover in mycobiont associations was also observed in G. confusoides T.C.Hsu, S.W.Chung & C.M.Kuo (Li, Boeraeve, Cho, Jacquemyn, & Lee, 2022). Information about the mycobionts involved in the early stages of development is needed for a comprehensive understanding of mycorrhizal symbiosis in Gastrodia and other orchid species. Furthermore, some Gastrodia species are rare or threatened (Suetsugu, 2016; Yagara & Nakajima, 2018), and additional information on mycorrhizal associations, particularly with respect to seed germination, is critical for their conservation.
The mycorrhizal associates of 11 Gastrodia species have been identified (Ogura-Tsujita et al., 2021; Li et al., 2022). Nine species have been found to associate with Mycena and six with marasmioids (including Marasmiaceae and Omphalotaceae), indicating that these fungi are the main mycobionts of Gastrodia. The contribution of Mycena fungi to the growth of G. elata at germination, rather than maturity (Xu & Guo, 2000), implies that Mycena and marasmioid fungi play important roles in the early development of Gastrodia species. In fact, in-situ seedlings obtained by seed baiting were found to harbor Mycena and/or marasmioid fungi in G. confusa Honda & Tuyama, G. nipponica (Honda) Tuyama, G. pubilabiata Y.Sawa (Kinoshita et al., 2016), and G. confusoides (Li et al., 2022). Symbiotic compatibility between fungus and plant in the early stages of plant development can be evaluated by in-vitro germination tests consisting of aseptic culture with orchid seeds and a single fungal isolate (Otero, Ackerman, & Bayman, 2004; Bonnardeaux et al., 2007). However, such germination tests have not been conducted for most Gastrodia species, aside from the medicinal species G. elata, which has been the focus of several studies (Xu & Guo, 1989; Park, Lee, & Ahn, 2012). In-vitro symbiotic germination was achieved in G. nipponica (Umata & Nishi, 2010) and G. pubilabiata (Umata, Yamauchi, & Hashimoto, 2000) using unidentified fungal isolates from mycorrhizal roots and fruiting bodies. In our previous study, we successfully induced in-vitro symbiotic germination of G. pubilabiata with Mycena and marasmioid isolates (Shimazaki, Higaki, Rammitsu, & Ogura-Tsujita, 2021). These results imply that in-vitro symbiotic culture systems occur in Gastrodia species other than G. elata.
In this study, we used in-vitro germination tests conducted with the seeds of five Gastrodia orchids (G. confusa, G. elata var. elata, G. elata var. pallens Kitag., G. nipponica, and G. pubilabiata) and three fungal isolates obtained from Gastrodia roots (two Mycena and one Marasmiaceae species) to evaluate the symbiotic compatibility of Gastrodia plants and their mycorrhizal fungi during the early stages of plant development. Phylogenetic analyses of the fungal isolates were also performed to confirm their taxonomic positions.
Roots of two individuals of G. pubilabiata and one individual of G. confusa were collected from an evergreen broadleaf forest and a bamboo forest, respectively, in Fukuoka Prefecture, Japan (Supplementary Table S1). The roots were washed with sterilized water and sectioned by hand into 1-mm-thick discs. Each disc was immersed in 200 µL distilled water and crushed with a scalpel under a stereomicroscope to release the fungal coils from the root cells. The fungal coils were collected using a micropipette and placed on 1.5% water agar medium containing 50 mg/L streptomycin and tetracycline. The plates were incubated at 25 °C under dark conditions for 3-7 d. Grown mycelium was subcultured on 2% malt extract agar medium (1.5% agar) at 25 °C. The latter culture condition previously yielded the best mycelial growth rates for fungi isolated from G. pubilabiata (Shimazaki et al., 2021). The fungal isolates used in this study were deposited in the NITE Biological Resource Center (Table 1; NBRC114082, NBRC114332, NBRC114083).
| Fungal taxon | Isolation origin | Isolate ID | NBRC No. * | Accession No. | |
| ITS | LSU | ||||
| Mycena sp. 1 | Gastrodia pubilabiata | F183 | NBRC 114082 | LC651187 | LC651188 |
| Mycena sp. 2 | Gastrodia confusa | F380 | NBRC 114332 | LC651189 | LC651190 |
| Marasmiaceae sp. 1 | Gastrodia pubilabiata | F184 | NBRC 114083 | LC651186 | - |
* NBRC: NITE Biological Resource Center.
DNA was extracted from the fungal isolates using the methods described by Izumitsu et al. (2012). The mycelium was collected using a sterilized toothpick and suspended in 50 µL TE buffer in a 1.5-mL tube. The tubes were microwaved for 1 min, kept at room temperature for 30 s, and then microwaved for an additional 1 min. After storage at -20 °C for 10 min, the tubes were centrifuged at 10,000 rpm for 5 min. The supernatants were used for polymerase chain reaction (PCR) templates. PCR and sequencing were performed as described by Ogura-Tsujita & Yukawa (2008). The internal transcribed spacer (ITS) region of the nuclear rDNA (nuc rDNA) was amplified using the primer combination ITS5/ITS4 (White, Bruns, Lee, & Taylor, 1990). The large subunit (LSU) D1/D2 region of nuc rDNA was also amplified using CTB6/LR5-F (Garbelotto et al., 1997; Tedersoo et al., 2008) for use in the phylogenetic analysis of the Mycena species. The PCR products were purified using a FastGene Gel/PCR extraction kit (Nippon Genetics Co., Ltd., Tokyo, Japan) and sequenced using a BigDye Terminator v3.1 cycle sequencing kit (Thermo Fisher Scientific, Waltham, MA, USA) and 3130 Genetic Analyzer (Applied Biosystems, Austin, Texas, USA). Obtained sequences were aligned with ATGC sequence assembly software (version 7; Genetyx Co., Tokyo, Japan). The aligned sequences were deposited in the DNA Data Bank of Japan under accession numbers LC651186-LC651190 (Table 1).
2.3. Phylogenetic analysesWe constructed four datasets, comprised of the ITS and LSU regions of Mycena and the ITS region of Marasmiaceae, for the performance of phylogenetic analysis of each isolate with high-homology sequences identified by a BLAST search. DNA sequences obtained from our isolates that closely matched sequences (> 90% similarity) from the BLAST search were aligned for analysis. For a Mycenaceae isolate, the ITS and LSU sequences from Chew, Desjardin, Tan, Musa, & Sabaratnam (2015); Oliveira, Sanchez-Ramirez, & Capelari (2014); and Karunarathna et al. (2020) were also included in the analysis. Multiple alignments were conducted using MUSCLE implemented in MEGA 7 (Kumar, Stecher, & Tamura, 2016). The phylogenetic analyses were conducted using the maximum likelihood (ML) method with MEGA 7 software, and Bayesian inference (BI) with MrBayes (version 3.2.6) (Ronquist et al., 2012). The best-fitting models of sequence evolution were determined using Kakusan 4 (Tanabe, 2011). The T92+G model was used for the ITS datasets for Mycena sp. 1 and a Marasmiaceae isolate, whereas GTR+G+I was used for the LSU datasets and the ITS dataset for Mycena sp. 2. Bootstrap (BS) analysis of the ML tree was performed with 1,000 bootstrap replicates. For BI analysis, Markov chains were run twice for 1,500,000 generations, and topologies were sampled every 100 generations. The first 25% were discarded as burn-in. A majority-rule consensus tree was constructed from the remaining trees, and the posterior probabilities (PPs) of clades were calculated. Sequence alignments were deposited in TreeBase under accession no. S28579 (http://www.treebase.org/).
2.4. Symbiotic seed germination experimentMature Gastrodia seeds were collected from natural habitats (Supplementary Table S1). Mature fruits are softer than young fruits and exhibit a slit. The collected seeds were dried for 2 wk in glass bottles with silica gel. The dried seeds were then stored at 5 °C until use. The seeds were sown within 6 mo after collection. The fungal isolates were subcultured on a malt extract agar medium for 2-3 wk. Leaf discs were prepared following Park et al. (2012) and Shimazaki et al. (2021). Dead leaves of Quercus glauca Thunb. were cut into 2 × 2-cm discs and autoclaved for 2 h. The sterilized discs were placed on the subcultured plates of each fungal isolate and incubated for an additional 2-3 wk. Fungus-colonized discs were then transferred to 1.5% water agar medium plates (Fig. 1A). The seeds were sterilized in 5% Ca (ClO)2 with 200 µL Tween 20 for 15 min, then rinsed three times with sterilized distilled water. Approximately 20-50 seeds were sown on each leaf disc, and each treatment was replicated five times. The plates were incubated for 4 mo at 25 °C in the dark. At the end of this period, germinants were counted under a stereomicroscope and assigned to one of five developmental stages following Higaki, Rammitsu, Yamashita, Yukawa, & Ogura-Tsujita (2017): 0, ungerminated (Fig. 1B); 1, 0.2-3-mm-long protocorms (Fig. 1C); 2, 3-10-mm-long protocorms with developed roots (Fig. 1D); 3, >10-mm-long seedlings (Fig. 1E); and 4, tuber development (Fig. 1F). Tubers are distinguishable from roots by the presence of fine hair on their surface. Because G. elata var. elata did not develop roots or develop beyond stage 2, its germination stages were classified as 1 (0.2-3-mm-long protocorms) and 2 (3-10-mm-long protocorms). To confirm fungal colonization of the protocorms, roots, and tubers, each tissue was sectioned with a razor and observed under a light microscope.
We symbiotically cultured five Gastrodia orchid seeds (G. confusa, G. elata var. elata, G. elata var. pallens, G. nipponica, and G. pubilabiata) with three fungal isolates [Mycena sp. 1 (F183) and Marasmiaceae sp. 1 (F184) isolated from G. pubilabiata, and Mycena sp. 2 (F380) isolated from G. confusa] (Table 1). Mycena sp. 1 promoted germination and development to at least stage 1 in all five orchids (Fig. 2). Seedling growth was also accelerated in G. confusa, G. nipponica, and G. pubilabiata, with 16-32% of individuals reaching stage 3 or higher and exhibiting root or tuber development. By contrast, the protocorms of G. elata var. elata and G. elata var. pallens exhibited no growth beyond stages 2 and 1, respectively. Mycena sp. 2 stimulated germination and development to stage 1 in G. elata, G. confusa, and G. nipponica. The strongest response was observed in G. confusa, with the germination of 71% of seeds. Marasmiaceae sp. 1 induced germination in G. nipponica and G. pubilabiata. A high rate of germination was observed in G. pubilabiata, and 20% of seedlings reached stage 3 or 4.
Microscopic observations showed that the basal protocorm region was heavily colonized by intracellular hyphal coils, but apical meristem was never colonized by the hyphae (Supplementary Fig. S1A). Young roots and tubers were not colonized by mycorrhizal fungi, while many starch granules were found in tuber cells (Supplementary Fig. S1B). Developed roots were partially colonized by hyphae and dense colonization was observed in cortical cells (Supplementary Fig. S1C). In both protocorms and roots, the third to fourth cell layers contained intact hyphal coils, and the adjacent lower layer contained digested hyphae (Supplementary Fig. S1D, E). Papillae, which are a typical hyphal structure for ptyophagous digestion, were often found on epidermal cell walls (Supplementary Fig. S1F).
3.2. Phylogenetic analysesThe phylogenetic analyses indicated that the two Mycena isolates (F183 and F380) belong to the Mycenaceae lineage (Fig. 3). The Mycena sp. 1 (F183) was grouped with species of Mycena sect. Fragilipedes in the LSU and ITS trees, with a high degree of support (72% BS/0.95 PP in LSU; 100% BS/1.00 PP in ITS; Fig. 3, Supplementary Fig. S2). This isolate was close to M. abramsii (Murrill) Murrill in the ITS tree and formed a monophyletic clade with fungi previously isolated from G. pubilabiata (LC314114 and LC314115), which shared 98-99% sequence similarities with Mycena sp. 1, with 80% BS and 0.99 PP support. The species closest to Mycena sp. 2 (F380) in the BLAST analysis was M. pura (Pers.) P. Kumm., with 89-90% ITS sequence similarity. However, Mycena sp. 2 was located outside of the sect. Calodontes which includes M. pura, in the LSU and ITS trees (Fig. 3, Supplementary Fig. S3). The LSU sequence of Mycena sp. 2 exhibited 100% similarity with a sequence of mycorrhizal fungus isolated from G. confusa (AB454414; Fig. 3). These two sequences formed a monophyletic clade with 100% BS and 0.99 PP support.
The BLAST search indicated that Marasmiaceae sp. 1 (F184) had > 90% similarity with 17 sequences. The phylogenetic analysis indicated that Marasmiaceae sp. 1 formed a monophyletic clade with 10 sequences, comprising mycobionts of G. pubilabiata (LC013342), G. nipponica (LC013366), and the green leafy orchid Cremastra variabilis (Blume) Nakai (LC440286), and the fruiting bodies of Marasmiellus rhizomorphogenus Antonín, R. Ryoo & H.D. Shin (GU319115, GU319116), M. candidus (Fr.) Singer (AB512315), and Crinipellis sp., which shared 99-100% ITS similarities (Supplementary Fig. S4).
Our results indicate that germination of five Gastrodia orchid was promoted by three fungal isolates, implying that Mycena and Marasmiaceae have symbiotic compatibility with these orchids during the early stages of development in nature. Mycena sp. 1 induced germination in all five Gastrodia orchids. Symbiotic in-vitro germination has been observed in G. elata with various Mycena isolates, such as M. osmundicola J.E. Lange (Xu & Guo, 1989), M. anoectochii L. Fan & S.X. Guo (Guo, Fan, Cao, Xu, & Xiao, 1997), and M. dendrobii L. Fan & S.X. Guo (Guo, Fan, Cao, & Chen, 1999). In addition, Mycena has been found in Gastrodia seedlings that germinated in natural habitats (Kinoshita et al., 2016; Li et al., 2022) or were cultured in organic materials collected from natural habitats of Gastrodia (Higaki et al., 2017; Shimaoka, Fukunaga, Inagaki, & Sawa, 2017). These results suggest that Mycenaceae fungi play an important role in the germination of Gastrodia.
Germinated protocorms of G. confusa, G. nipponica, and G. pubilabiata exhibited continued growth and tuber development when cultured with Mycena sp. 1, as did G. pubilabiata when cultured with Marasmiaceae sp. 1 (Fig. 2). Mycena sp. 1 sequences were detected in adult roots of all three Gastrodia species, whereas Marasmiaceae sp. 1 was detected predominantly in G. pubilabiata roots (Kinoshita et al., 2016). These findings imply that the association between plant and fungus persists from seed germination to seedling maturity. Associations with such core mycobionts may be maintained throughout the plant life cycle, and total replacement of mycobionts, as exhibited by G. elata (Xu & Guo, 2000) and G. confusoides (Li et al., 2022), may not always occur. Whereas tuber formation was observed with these particular plant-fungus combinations, protocorm growth ceased at stage 1 or 2 with other combinations (Fig. 2). As G. elata switches mycobionts from Mycena to Armillaria following seed germination (Xu & Guo, 2000), its protocorm requires Armillaria for continued growth, as demonstrated by Park et al. (2012). Similar mycobiont replacement may be needed with other plant-fungus combinations in which protocorm growth ceased at stage 1 or 2. Seed germination without subsequent growth could also be explained by inadequate culture conditions, especially for Mycena sp. 2. This isolate was obtained from G. confusa collected in a bamboo forest, whereas the other isolates were obtained from G. pubilabiata collected in evergreen broadleaf forests. We used Quercus leaves for all isolates, but bamboo leaves or stems may be a more suitable substrate for the nutrition of Mycena sp. 2. As the type of organic material affects seedling growth in G. pubilabiata (Shimazaki et al., 2021), improvement of the substrate match could induce further growth of G. confusa cultured with Mycena sp. 2.
The effects of Mycena and Marasmiaceae on seed germination varied among Gastrodia species (Fig. 2), implying that fungal specificity during germination differs among plant species. Germination of G. elata var. elata and G. confusa was promoted only by Mycena isolates, whereas that of G. nipponica and G. pubilabiata was stimulated by Mycena and Marasmiaceae. These differences likely reflect the mycorrhizal fungal specificity of adult plants, with the exception of G. elata, whose mycobionts switches from Mycena to Armillaria after germination (Xu & Guo, 2000). Mature G. confusa plants were associated almost exclusively with Mycena species, whereas mature G. nipponica and G. pubilabiata plants were associated with Mycena and Marasmiaceae fungi, with nearly identical relative fungal frequency abundances (Kinoshita et al., 2016). Some Gastrodia species appear to maintain fungal specificities throughout life cycles.
Microscopic observations suggested that the three Gastrodia species, G. confusa, G. nipponica and G. pubilabiata, exhibit a rare fungal penetration type, ptyophagy. Orchid mycorrhiza has two penetration types: tolypophagy, found in the majority of orchids, and ptyophagy, only known in a few mycoheterotrophic orchids (Burgeff, 1932; Rasmussen, 2002). The structure of G. elata mycorrhizae allows ptyophagous penetration, since its epidermal and/or outer cortex layers host the intact hyphae and are adjacent to the digestion cells in the inner cortex (Wang, Wang, Zhang, Liu & He, 1997; Li, Guo & Lee, 2020). After fungal hyphae penetrated the plant cell, the inner plant cell wall becomes thickened and papillae-like structures are formed surrounding penetrated hyphae (Burgeff, 1932; Martos et al., 2009), which were often found in our Gastrodia samples (Supplementary Fig. S1F). Since ptyophagous penetration is mainly known in Gastrodia species, such as G. javanica, G. callosa and G. similis (Burgeff, 1932; Martos et al., 2009), this penetration type seems to be common for Gastrodia species.
Phylogenetic analyses demonstrated that the two Mycena isolates used in this study positioned in the diffderent lineages (Fig. 3). The LSU and ITS analyses demonstrated that Mycena sp. 1 belongs to section Fragilipedes (Fig. 3, Supplementary Fig. S2). The ITS analysis also indicated that Mycena sp. 1 is closely related to M. abramsii (Supplementary Fig. S2). By contrast, we could not fully determine the taxonomic affinity of Mycena sp. 2 at the species level. This isolate exhibited 100% LSU sequence similarity with the G. confusa mycobiont, but did not cluster with any known Mycena species. In-vitro induction of fruit body formation will be required to further identify Mycena sp. 2. To date, three Mycena species, i.e., M. orchidicola L. Fan & S.X. Guo (Fan, Guo, Cao, Xiao, & Xu, 1996), M. anoectochii (Guo et al., 1997), and M. dendrobii (Guo et al., 1999), have been described based on artificially induced fruit bodies using cultures isolated from orchid roots. Identification of the mycorrhizal associates of Gastrodia species, as well as other orchid species, will contribute to understand diversity and ecology of Mycena.
The phylogenetic placement of the Marasmiaceae isolate (F184) that we suggested previously for Marasmiaceae and Omphalotaceae (Ogura-Tsujita et al., 2021) indicated that this isolate belongs to the campanelloid clade of Marasmiaceae. This clade includes mycobionts of mature individuals of two Gastrodia species and mature individuals of Cremastra variabilis (Yagame, Lallemand, Selosse, Funabiki, & Yukawa, 2021; Supplementary Fig. S4). This finding implies that the members of this clade are mycobionts of leafy orchids in addition to mycoheterotrophic species. The ITS sequences (GU319115, GU319116) used to describe Marasmiellus rhizomorphogenus (Antonín, Ryoo, & Shin, 2010) belonged to the same clade with 99% similarity, implying that isolate F184 may belong to M. rhizomorphogenus.
Our results support the hypothesis that Mycena and marasmioid fungi play important roles in the germination of Gastrodia species. Fungal isolates induced tuber formation may support plant growth throughout the life cycle, and be useful in the propagation of endangered Gastrodia species. Further in-vitro culture is required to confirm whether these fungal isolates promote growth during later developmental stages. Fungal isolates, Mycena sp. 1 and Marasmiaceae sp. 1, are closely related to Mycena abramsii and Marasmiellus rhizomorphogenus, respectively, and the taxonomic position of Mycena sp. 2 remains unclear. Morphological identification using fruit bodies induced by in-vitro culture or collected from habitats where Gastrodia occurs will determine the taxonomic affinity of these fungi at the species level. Our work contributes to a better understanding of the mycorrhizal associations of mycoheterotrophic orchids throughout their life cycles, and the diversity and ecology of Mycena and marasmioid fungi.
The authors declare no conflicts of interest. All the experiments undertaken in this study comply with the current laws of the country where they were performed.
We thank A. Kawarabata and K. Tanaka for collecting samples, K. Rammitsu for technical supports. This study was supported by JSPS KAKENHI Grant Number JP21K06306 for YO and JP18H02500 for YT.
