Correspondence: Hiroko Kawasaki, The International Center for Biotechnology, Osaka University, 2-1 Yamada-oka, Suita-city, Osaka 565-0871, Japan. Tel.: +81 6 68794085; fax: +81 6 68797454; e-mail: email@example.com
Bamboos are common and useful plants in Japan but little information is available about their endophytes. In this study, 257 fungal strains were isolated from bamboo tissues, and 71 representative strains were characterized by 18S rRNA gene and internal transcriber spacer region sequence analysis. Phylogenetic analysis showed that the fungal isolates were located in Sordariomycetes and Dothideomycetes. Xylariales was the dominant group within bamboos. Several rRNA gene sequences were not similar to any current sequence in the database and might be a novel species or genera that could represent sources of novel biological compounds. These findings reveal that bamboos are a huge reservoir of microorganisms that should be extensively investigated.
Bamboos are highly diverse and are distributed throughout Asia, especially in China and Japan. Bamboo products are useful for many purposes, e.g. construction, traditional craft, food, and medicines. Some antifungal proteins have been isolated from bamboo shoots (Wang & Ng, 2003). At present, the greatest diversity of bamboo endophytes occurs in Asia, as of the roughly 500 species recorded, 38% were recorded in Japan (Hyde et al., 2002). Data on bamboo-associated fungi have been recorded by Hino & Katumoto (1961). Recently, Tanaka & Harada (2004) described some noteworthy bamboo-associated fungi from Hino and Katumoto records. Nevertheless, most fungal records are based on symptoms and their morphology. Hence, a more extensive study of internal bamboo-associated fungi is needed. Nowadays, molecular technologies offer rapid methods of analysis that have many applications including in taxonomy. In this research, we focused on the screening, isolation and classification of endophytes from bamboos located in Japan by using 18S rRNA gene and the internal transcribed spacer (ITS) region as an indicator of the relationships of fungi with their relatives.
Materials and methods
Isolation of endophytic fungi
Endophytes were isolated from 23 individual bamboos at three sampling sites in Osaka. Two types of plants, Phyllostachy spp. (Take) and Sasa spp. (Sasa), were collected from the Suita campus of Osaka University in the winter, and from Mt. Minoh and the Hakunoshima, Minoh city, in the spring of 2005. Whether or not the external plant tissues exhibited symptoms was ignored. Plant tissues were cut with knives from leaves, nodes and internodes, into small pieces of less than 1 cm. Each tissue was surface sterilized by washing in 70% ethanol for 1 min, then immersed in 3% sodium hypochlorite for 3–5 min, and rinsed with 70% ethanol for 30 s. After that, the tissues were washed in sterile distilled water. The segments were dried and placed on PDA medium containing 20% potato and 1% dextrose for isolation. Samples were incubated at temperatures below 25°C. Each fungus was purified and fungal colonies were classified by colony-forming characters, e.g. colour, shape, growth-rate and pigments.
Genomic DNA extraction and PCR amplification
The total genomic DNA of representative strains from each morphotype were extracted. The DNA was recovered directly from actively growing mycelium scraped from PDA plates. The DNA was extracted using fungal miniprep kits (E.Z.N.A.™, Omega Bio-tek, Inc.). The 18S rRNA genes were amplified in a 50-μL PCR reaction using the primer pairs NS1-1 (5′-TAGTCATWTGCTTGTCTYAAA-3′)/SR6-1 (5′-TTTTASTTCCTCTAAAYGACC-3′) (NS1-1 modified from NS1 (White et al., 1990) and SR6-1 modified from SR6 (Bruns et al., 1992) and the ITS region was amplified using the primer pairs ITS1-1 (5′-TCCGTWKGTGAACCWGCG-3′)/ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) (modified from White et al., 1990). The PCR amplifications were performed using Ex Taq polymerase (Takara Shuzo) in a Gene Amp® PCR system 9700 (PE Applied Biosystems) under both the 18S rRNA gene and the ITS region primer conditions. The details were as follows: 5 min at 94°C for initial denaturation; 30 s at 94°C for denaturation, 30 s at 55°C for annealing of 18S rRNA gene but 50°C for annealing of the ITS region, 30 s at 72°C for extension (for a total of 35 cycles of amplification) and, finally, extension for 7 min at 72°C. PCR products were purified using DNA Gel Extraction kit (Bio-Rad).
18S rRNA gene, ITS region sequencing and phylogenetic analysis
The PCR products of 18S rRNA gene and the ITS region were sequenced using the ABI PRISM 310 Genetic Analyzer (PE Applied Biosystems) with Big Dye Terminator Cycle Sequencing Kit Version 3.1 by using primer P10 (5′-GACTAACTACTGCGAAAG C-3′) (Yamada et al., 1999) for partial 18S rRNA gene and ITS1/ITS4 for the ITS region (White et al., 1990). Preliminary multiple alignment and reversed compliment were performed and adjusted using mega software version 3.1 (Kumar et al., 2004). Ambiguous positions that may not be homologous were eliminated, and gap positions were treated as missing data. Phylogenetic analyses were simulated based on truncated sequences of gene alignment; phylogenetic trees were inferred by the neighbour-joining method (Kimura 2-parameter distance calculation) and by maximum parsimony using the heuristic search (CNI level=1) option with the 100 random addition sequences (Nei & Kumar, 2000). The bootstrap method was used with 1000 replications to evaluate the reliability of tree topologies.
Results and discussion
Isolation of endophytic fungi in bamboo tissues
The 257 fungal strains were isolated and categorized into 78 morphotypes on the basis of colony distinction. The representative strain of each group underwent sequencing analysis based on 18S rRNA gene and the ITS region. A list of the data used in drawing the trees and of sampling information is given in Table 1.
The morphological characteristics of the taxa were very diverse and ranged from 4 to 17 morphotypes for each plant individual. Some taxa – g02, g12 and g26 – were widely colonized and can be recognized in different areas. Conversely, some morphotypes were colonized preferably within the same sampling sites and host plants. For instance, g10 and g16 were isolated from the Osaka University campus, whereas g49, g52, g59, and g62 were isolated from Mt. Minoh. Also, g74, g75 and g76 were only found in the Hakunoshima area. In particular, all plant samples from Mt. Minoh were Sasa and differed from the plant samples from the Hakunoshima area, which included only Take. This finding might suggest that the diversity of endophytic fungi has been influenced by habitat, region and the range of their hosts, and it would be interesting to further investigate their ecological community.
Phylogeny based on the partial sequence of 18S rRNA gene
The partial 18S rRNA gene phylogram showed overviews of endophyte clusters in bamboos. Selected truncated sequences (610 bp) of 60 strains from 69 groups were aligned, and their trees drawn. The topology of both the neighbour-joining and maximum parsimony trees were similar. In Fig. 1, the maximum parsimonious tree was mainly illustrated by bootstrap analysis from the maximum parsimony method of greater than 50% and those from the neighbour-joining method (maximum parsimony/neighbour-joining) of greater than 90%.
The phylogenetic tree showed that our samples contained diverse fungal group profiles and were distributed within Ascomycota. They were dominated by two subclasses: Sordariomycetes and Dothideomycetes. Sixty strains were clustered into 12 groups (cluster A–L). Forty-two isolates belonging to Sordariomycetes were localized into seven orders; Xylariales, Phyllachorales, Hypocreales, Diaporthales, Ophiostomatales, Sordariales and Coniochaetales. Fifteen strains were related to Pleosporales within Dothideomycetes.
Xylariales was the dominant group of fungi associated with bamboos in this study. Nine strains were localized into cluster A. Three strains were similar to Microdochium and Pestalotiopsis in cluster B. Five strains were distinctively located in cluster C with Xylaria and Poronia ssp. Only strain g49 was located in cluster C, but it was separated from Xylaria. Strain g34 was rather similar to Sordaria in cluster D. Fourteen strains were mixed within Hypocreales and Phyllachorales in cluster E. Nine strains were similar to Diaporthales and Ophiostomatales in clusters G and H. Strain g48 was located in Coniochaetales in cluster F.
Three strains (g67, g37 and g73) had quite similar sequences to Guignardia belonging to Botryosphaeriaceae, Cladosporium belonging to Mycosphaerellaceae and Peziza to Pezizales, respectively. Those strains were located with high bootstrap support in their own cluster (I, J and K, respectively).
Fifteen strains were localized in Pleosporales belonging to Dothideomycetes in cluster L. Three strains (g05, g42 and g75) were located in Pleosporales, but they have no closely related references. Strain g41 was closely related to Alternaria with a high bootstrap value. Eleven other strains (g58, g78, g70, g43, g51, g72, g27, g76, g74, g59 and g19) were intermingling among Pleosporales with several reference genera, e.g. Leptosphaeria, Shiraia, Paraphaeosphaeria and Phoma.
18S rRNA gene results showed a high divergence of bamboo endophytes among Ascomycota. Several isolates had no similar sequence in the database, which suggests that either they are novel species/genera or that the database of 18S rRNA gene sequences is not complete. Hence, the ITS region sequences would need to be further identified at the genus and/or species level.
Phylogeny based on the ITS region
ITS region trees are shown in Fig. 2. Twenty-four strains were distributed into several genera of Xylariales (Fig. 2a), i.e. Arthrinium, Monochaetia, Pestalotiopsis, Microdochium, Arthrobotrys, Hypoxylon, Biscogniauxia, Camillea, Rosellinia and Xylaria. Eight strains (g15, g39, g22, g62, g40, g04, g32 and g06) were similar to Arthrinium ssp. (94–97% similarities), whereas g14 showed lower similarity (94%). In addition, five strains (g26, g02, g36, g12 and g30) were separated into another lineage close to the genus Arthrinium, and none of the available reference strains showed similarity except fungal endophyte AF413049 with low similarities (91–98%). Strain g01 was highly similar to Pestalotiopsis vismiae. Strain g57 was clustered to Monochaetia monochaeta. Two strains, g44 and g53, were closely related to Microdochium phragmitis and Arthrobotrys follicola. Strain g55 was located in the same lineage as Hypoxylon with a high bootstrap value. Interestingly, strain g49, which showed a close relationship with Xylaria in 18S rRNA gene, is closely related to Biscogniauxia and Camillea spp. in the ITS region tree. However, the similarity score was very low (92%). Strain g71 was similar to genus Rosellinia. Three strains (g47, g66 and g08) were clustered to Xylaria, but their sequence similarities with related species were low (92–98%).
Six strains (g20, g03, g25, g31, g17 and g23) were related to the members of Hypocreales (Fig. 2b). Their closely related genera were Gibberella and Fusarium. The information on the ITS region of Fusarium/Gibberella is quite well established. Many reference species were similar to those strains. The genus Fusarium has been isolated from many plants and causes some plant diseases (Rubini et al., 2005).
Figure 2c shows the 10 strains related to Colletotrichum/Glomerella in Phyllachorales. Six strains (g46, g28, g07, g77, g45 and g60) were most similar to Colletotrichum dematium. Strains g16 and g64 were in the same lineage with high bootstrap values but no reference. Strains g13 and g61 were similar to Glomerella cingulata and Colletotrichum gloeosporioides with 99% similarity.
Five strains belonging to Diaporthales were closely related to Phomopsis amygdali (g11 and g29), Phomopsis liquidemabari (g09 and g24) and Diaporthe helianthi (g21) (Fig. 2d). An overview of Diaporthales based on 28S rRNA gene showed that Diaporthe is a lineage clade belonging to Diaporthaceae (Castlebury et al., 2002) with anamorphs known as the genus Phomopsis. Three strains (g10, g18 and g35) were not similar to any references data and formed a monophyletic lineage with a high bootstrap level. Moreover, two strains (g56 and g48) were supposed to be Ophiostomatales and Coniochaetales (Fig. 2e). Strain g56 was similar to Pyricularia or Magnaporthe and strain g48 was closely related to Phialophora or Lecythophora with a low similarity (94%).
Figure 2f shows the strains located in Pleosporales of Dothideomycetes. Three strains (g05, g74 and g43) were similar to Shiraia sp. ML-2004 (91–94% similarities) and strain g58 was similar to Shiraia bambusicola. Strains g19 and g41 were similar to Ophiosphaerella agrostis (93% similarity) and Alternaria alternata/Alternaria tenuissima (97% similarity), respectively. Three strains (g75, g78 and g42) were similar to the genera Ampelomyces and Phoma with 95–98% similarities. Strain g69 was closely related to Lophiostoma vagabundum (94% similarity). Strain g51 was clustered with Leptosphaeria and Helminthosporium (93% similarity). However, five strains (g59, g70, g27, g72 and g76) were intermingled among Pleosporales with no closely related references.
Bamboo isolates were highly diverse within several fungal groups. Bamboo may represent a huge resource in the search for novel strains, including novel metabolites. Consequently, taxonomic studies involving both morphological and molecular approaches should be intensively performed.
Fungal diversity and community within bamboos hosts
All of the bamboo isolates were highly distributed among Ascomycota, mainly in Sordariomycetes and Dothideomycetes. Many species belonging to those subclasses are known as pathogens (Berbee, 2001), as indicated by the fruiting bodies produced on dead plant material. However, some of the species probably colonized living plants asymptomically. Some genera that were found in our samples were recorded as pathogens, i.e. Alternaria, Leptosphaeria, Fusarium, Gibberella and Glomerella. However, no symptom was recognized in the plant materials. Perhaps some fungi may act as latent pathogens and endophytes in parts of their life cycle.
The occurrence of fungal endophytes is very high in cool-season grass (Marshall et al., 1999), and palms in both tropical and temperate regions (Taylor et al., 2000). Our results showed that the colonization of fungal endophytes in bamboo plants is extremely frequent, and the common genera have already been reported by Hino & Katumoto (1961). Some fungi were already known, i.e. Shiraia, Leptosphaeria, Hypoxylon, Xylaria, Phoma, Colletotrichum, and Fusarium. However, reports on only a few species have been published. Several fungi not yet been recorded by Hino were found in this study, e.g. Peziza, Diaporthe, Sordaria, Alternaria and Microdochium. Notably, several species and genera were determined as a result of the rapid identification by molecular tools.
Xylariales are a dominant group in bamboos, especially those of the genus Arthrinium, which are the most common. Arthrinium is cosmopolitan in soil, marine water and the roots of reeds (Wirsel et al., 2001). Samuels et al. (1981) have reported Arthrinium isolates and their telemorph (Apiospora) on bamboo and graminaceous hosts from New Zealand. Two species of Apiospora were found in Japan but the anamorphs of these have not been recorded. Several isolates shown in Fig. 2a were closely related to Arthrinium and might be novel genera. Nevertheless, documentation for this is poor.
Pestalotiopsis are most commonly isolated from rain forests and are a source of secondary metabolites (Strobel, 2002). Strain g01, which is similar to Pestalotiopsis, was isolated in this study. Strain g53 was isolated close to Microdochium, which has been described as a plant pathogen causing diseases of the stem of cereal and turfgrasses (Mahuku et al., 1998). Xylariaceae has been documented in many plants and genus Xylaria has been isolated from almost all tropical and crop plants (Bayman et al., 1998; Davis et al., 2003). Hino recorded one species, Xylaria take, in Japan. Our results showed that three species of Xylaria were found in bamboos. Biscogniauxia and Hypoxylon were normally found in the bark of dead plant material (Ju et al., 1998). Interestingly, two strains that related to those genera were isolated from fresh plant tissues. At present, the role of Xylariaceae in the ecological system is ambiguous. Whalley (1996) hypothesized that the role of Xylariaceae endophytes is as a quiescent colonizer and later a decomposer when the plant begins to senesce. This hypothesis agrees with our results showing that some xylariaceous fungi may act as endophytes by taking up energy from fresh tissues and later from dead plants.
Gibberella and Fusarium (its anamorph) were described as an endophyte and/or phytopathogen. Two strains were similar to Gibberella avenacea, which is known to cause disease in wheat. Some species of Fusarium have the ability to degrade both lignin and laccase (Anderson et al., 2005). Several strains from our study might be a source of useful biological products. Recently, much of the geographical distribution of Fusarium/Gibberella was investigated. O'Donnell et al. (1998) revealed that Fusarium species often originate from the same geographical region. This finding might be of interest in geographic evolutionary studies on Fusarium associated with bamboos in different regions.
Glomerella and Colletotrichum (its anamorph) were widely distributed and had an extremely broad host range. They are known as pathogens of plants worldwide and have frequently been isolated from forests. Many species of Colletotrichum cause anthracnose on important crops, and it is useful genus for studies of pathology and fungal-plant interaction (Perfect et al., 1999) such as Colletotrichum gloeosporioides and Colletotrichum sublineolum, which were found in our samples. Also, several strains were related to Colletotrichum dematium, which produces a chemical compound (Abou-Zaid et al., 1997).
The genera Diaporthe and Phomopsis (its anamorph) were mainly found among Diaporthales in our study. Diaporthe/Phomopsis are frequent colonizers in various hosts from woody dicotyledonous to herbaceous monocots in different geographic areas, both tropical (Lodge et al., 1996) and temperate (Eriksson & Yue, 1998).
In Pleosporales, S. bambusicola was recorded as a parasitic fungus on the twigs of bamboo plants; its uses in Chinese medicine have been documented (Kishi et al., 1991; Cheng et al., 2004). This species is widely distributed in Southern China and has been reported in Japan, along with Phaeosphaeria and Leptosphaeria. Interestingly, we isolated one strain similar to S. bambusicola and three strains referred to as Shiraia species. These strains might represent a new species found in Japanese bamboos that should be more extensively investigated.
This work was supported by a Grant-in-Aid for Cooperative Research (A) (no. 16255001) from the Japan Society for the Promotion of Science, Ministry of Education, Culture, Sports, Science and Technology of Japan. This paper represents a portion of the dissertation submitted by Doungporn Morakotkarn to Osaka University in partial fulfilment of the requirements for a Ph.D. degree.