Microb Ecol (2011) 62:106–120 DOI 10.1007/s00248-011-9838-3 INVERTEBRATE MICROBIOLOGY Fungus Symbionts Colonizing the Galleries of the Ambrosia Beetle Platypus quercivorus Rikiya Endoh & Motofumi Suzuki & Gen Okada & Yuko Takeuchi & Kazuyoshi Futai Received: 27 September 2010 / Accepted: 21 February 2011 / Published online: 8 March 2011 # Springer Science+Business Media, LLC 2011 Abstract Isolations were made to determine the fungal symbionts colonizing Platypus quercivorus beetle galleries of dead or dying Quercus laurifolia, Castanopsis cuspidata, Quercus serrata, Quercus crispula, and Quercus robur. For these studies, logs from oak wilt-killed trees were collected from Kyoto Prefecture, Japan. Fungi were isolated from the: (1) entrances of beetle galleries, (2) vertical galleries, (3) lateral galleries, and (4) the larval cradle of P. quercivorus in each host tree. Among the fungus colonies which appeared on YM agar plates, 1,219 were isolated as the representative isolates for fungus species inhabiting in the galleries based on their cultural characteristics. The validity of the visual classification of the fungus colonies was checked and if necessary properly corrected using microsatellite-primed PCR fingerprints. The nucleotide sequence of the D1/D2 region of the large subunit nuclear rRNA gene detected 38 fungus species (104 strains) of which three species, i.e., Candida sp. 3, Candida kashinagacola (both yeasts), and the filamentous fungus Raffaelea quercivora were isolated from all the tree species. The two yeasts were most prevalent in the interior of galleries, regardless of host tree species, suggesting their close Electronic supplementary material The online version of this article (doi:10.1007/s00248-011-9838-3) contains supplementary material, which is available to authorized users. R. Endoh (*) : Y. Takeuchi : K. Futai Laboratory of Environmental Mycoscience, Division of Environmental Science and Technology, Graduate School of Agriculture, Kyoto University, Kitashirakawa Oiwake-cho, Sakyo-ku, Kyoto 606–8502, Japan e-mail: rikiyasu@kais.kyoto-u.ac.jp R. Endoh : M. Suzuki : G. Okada Microbe Division/Japan Collection of Microorganisms (JCM), RIKEN BioResource Center, Wako, Saitama 351–0198, Japan association with the beetle. A culture-independent method, terminal restriction fragment length polymorphism (TRFLP) analysis was also used to characterize the fungus flora of beetle galleries. T-RFLP patterns showed that yeast species belonging to the genus Ambrosiozyma frequently occurred on the gallery walls along with the two Candida species. Ours is the first report showing the specific fungi inhabiting the galleries of a platypodid ambrosia beetle. Introduction Coleopteran species, including bark and ambrosia beetles, and their fungus associates represent a rich source of information in the field of symbiosis. Reciprocal effects are commonly found between them and a diverse array of interactions including commensalism [19]. Some insects can obtain food directly from fungi or from plants weakened by such fungi. Other beetles feed upon fungi in dead plant material. In return, the fungus benefits by being efficiently transmitted directly to a suitable habitat by the vector beetle [41]. The partnership between bark beetles and blue-stain fungi is one of the most familiar examples of such a relationship. Blue-stain fungi including Ophiostoma novo-ulmi are the principal causal agent of Dutch elm disease in Europe, North America, and parts of Asia [4, 5]. They are transmitted by various elm bark beetles, often Ips spp. [18] where the fungus causes wilting and mass mortality of host trees, mainly Ulmus spp. Ambrosia beetles, as well as bark beetles, have an intriguing partnership with fungi involving questions of interest in microbial ecology, evolutionary microbiology, and fungal systematics. Generally, studies concerning such relationships have arisen as they relate to certain plant diseases and/ or accompanying economic damage to logs and lumber. Symbiotic Fungi of Platypus quercivorus The last decade has witnessed the outbreak of Japanese Oak Wilt disease in Japan, especially in the regions along the Japan Sea on the Honshu Island [14]. Japanese Oak Wilt, caused by the fungus Raffaelea quercivora Kubono & Shin. Ito, is vectored by the beetle Platypus quercivorus (Murayama) (Coleoptera, Platypodidae) which bores into the host’s sapwood, occasionally even the heartwood of host oaks, especially Quercus crispula and Quercus serrata [13, 20–22, 38]. Affected trees wilt and die shortly thereafter, similar to the situation in Dutch elm disease. As shown by phylogenetic analyses based on the nucleotide sequences of small subunit rRNA gene (SSU rDNA) [15], Raffaelea belongs to the fungus group (mitosporic) Ophiostomataceae. The primary vector of the pathogen R. quercivora is the ambrosia beetle P. quercivorus, which is an aggressive tree killer which carries not only the pathogen but also dietary fungus symbionts in special structures called “mycangia.” For reproduction, the beetles construct galleries in the sapwood of a host tree, where the symbiotic fungi, mainly yeast-like microbes, are disseminated and grow on the wall of the galleries in the sapwood. After constructing sufficient branches of galleries, female beetle lays eggs in the gallery. Beetle larvae that hatch feed on fungi growing on the gallery wall. The ambrosia beetles have a characteristic feeding habit called xylomycetophagy, in which the successful development of their offspring largely depends on growth of the ambrosia fungi which are cultivated on the beetle galleries [16]. Although the fungi associated with scolytid bark beetles are relatively well known [18], much less is known about the fungus associates of platypodid ambrosia beetles. Indeed, the fungal symbionts which have been identified for only a small percentage of ambrosia beetles, and for many of these it is not clear if those which have been identified are the primary symbiont or a contaminating fungus [10]. As is the case for other ambrosia beetles, the primary symbiont(s) of P. quercivorus has not been identified. The main purpose of this work was to identify the primary symbiotic fungi of the ambrosia beetle P. quercivorus. Initially, we analyzed the fungal species colonizing the galleries of P. quercivorus. This was done by isolating and identifying the fungi inhabiting beetle galleries in sapwood of Quercus laurifolia, Castanopsis cuspidata, Q. serrata, Q. crispula, and Quercus robur. Isolations were made from the gallery entrance, vertical gallery, lateral gallery, and larval cradle in the sapwood of these five tree species. The isolated fungi were classified according to their cultural characteristics and their polymerase chain reaction (PCR) fingerprints generated by microsatellite-primed PCR. Sequencing of the D1/D2 region of the large subunit rRNA gene (LSU rDNA) was 107 done to determine their phylogenetic positions. The species composition in the fungal flora on the galleries was also determined based on the colony counts. Also, we characterized the fungal flora using a culture-independent method, i.e., terminal restriction fragment length polymorphism (T-RFLP) analysis. Materials and Methods Beetle Galleries The sampling method used to obtain gallery material for fungus isolation has been given [6] with some modifications being used for Q. serrata and Q. crispula galleries. A 5-cm thick disk was cut and broken into wood chips (20– 30×20–30×2–12 mm). The material included a single beetle gallery bored by P. quercivorus. Wood chips were surface-sterilized carefully using a flame. The wood chip samples of Q. serrata and Q. crispula were immersed in 0.9% (wt./vol.) sodium chloride solution, the volume of the solution was 4 and 2 ml, respectively. Samples used for fungus isolation were collected from different beetle galleries in which beetles had reproduced, i.e., healthy larvae and pupae, or both, were present. The sample locations within the beetle galleries were the: entrance (gallery which runs laterally across a growth ring slightly below the bark), vertical gallery (gallery which runs vertically along the vascular bundle), lateral gallery (gallery which runs laterally along with a growth ring), and larval cradle (gallery which runs vertically up or down only about 5–7 mm in length diverging from a lateral gallery). Table 1 gives the abbreviations for the wood chips used for fungus isolation, fungus isolate, gallery type, and tree species. Diameter of breast height (DBH), sampling site, and collection date are given in Table 1 in the Electronic Supplementary Materials. Cultivation, Colony Count, Typing of Fungi, and Culturing A microbial suspension in 0.9% (wt./vol.) sodium chloride solution was collected from the gallery wall and prepared according to the method of Masuya et al. (Proceeding of the 117th Conference of the Japanese Forest Society, Tokyo. http://www.jstage.jst.go.jp/article/jfsc/116/0/359/_pdf/-char/ ja/) as modified by Endoh et al. [7]. Fungus isolation was done using the standard plating method. One hundred microliters of the microbial suspension of each decimal dilution was distributed onto three replicates (plates) of YM agar (Difco). Colony counting and typing was done for three plates (master plates) that produced more than 50 fungus colonies. Fungus colonies that appeared on the master plates were counted and visually classified into 108 Table 1 Sources of fungus isolation a All wood chips were obtained from the beetle galleries of Platypus quercivorus x strain numbers R. Endoh et al. Abbreviation for wood chip and fungus isolatea Gallery type QlE-i, QmPlEG-2-x QlE-ii, QmPlEG-3-x QlL-i, QmPlLG-1-x QlL-ii, QmPlLG-2-x QlL-iii, QmPlLG-3-x QlV-i, QmPlVG-1-x QlV-ii, QmPlVG-2-x QlV-iii, QmPlVG-3-x Entrance Entrance Lateral gallery Lateral gallery Lateral gallery Vertical gallery Vertical gallery Vertical gallery QlC-i, QmPlPB-1-x QlC-ii, QmPlPB-2-x QlC-iii, QmPlPB-3-x CcE-i, CcPqEG-1-x CcE-ii, CcPqEG-2-x CcE-iii, CcPqEG-3-x Larval cradle Larval cradle Larval cradle Entrance Entrance Entrance 6 7 8 5 3 2 CcL-i, CcPqLG-1-x CcL-ii, CcPqLG-2-x CcL-iii, CcPqLG-3-x CcC-i, CcPqPC-1-x CcC-ii, CcPqPC-2-x CcC-iii, CcPqPC-3-x CcC-iv, CcPqPC-4-x CcC-v, CcPqPC-5-x Qs3C-i, QsPq3PC-1-x Qs3C-ii, QsPq3PC-2-x Qs3C-iii, QsPq3PC-3-x Qc2C-i, QcPq2PC-1-x Qc2C-ii, QcPq2PC-2-x Qc2C-iii, QcPq2PC-3-x Qc2C-iv, QcPq2PC-4-x QrL-i, QrPqLG-1-x QrL-ii, QrPqLG-2-x QrL-iii, QrPqLG-3-x Lateral gallery Lateral gallery Lateral gallery Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Larval cradle Lateral gallery Lateral gallery Lateral gallery 5 3 4 3 3.5 3.5 6.5 9.5 3.5 4 4 3.5 3 5.5 5 4 5 3 QrL-iv, QrPqLG-4-x QrC-i, QrPqPC-1-x Lateral gallery Larval cradle 5 6 several fungus groups according to their cultural characteristics. For each of the 34 wood chips, usually eight or more colonies were selected per fungus group (first representatives). Single colony isolation was done at least twice. The isolates were frozen in ca. 15% (vol./vol.) glycerol solution at −80°C for subsequent laboratory study. Representative isolates (Table 2; Table 2 in the Electronic Supplementary Materials) have been deposited in the Japan Collection of Microorganisms (JCM; http://www.jcm.riken.jp/), except for isolate QmPlVG-1-41 (Peniophoraceae sp.) which subsequently failed to grow in vitro, and for a few in the Centraalbureau voor Schimmelcultures (CBS; http://www. cbs.knaw.nl/) collection (Table 2; Table 2 in the Electronic Supplementary Materials). Gallery length (mm) 4 6 4 5 6 7 12 12 Tree species Quercus laurifolia Castanopsis cuspidata Quercus serrata Quercus crispula Quercus robur Genomic DNA Extraction Two methods were used to extract the genomic DNA from the fungi, depending on their cultural morphology as described below. The genomic DNA extracted was used for the subsequent microsatellite-primed PCR (MSP-PCR) fingerprinting and sequencing of the D1/D2 region of the LSU rDNA. a. DNA extraction from yeast-like fungi. The genomic DNA of fungi that grew in yeast form in liquid medium was extracted using the glass beads method as described in Endoh et al. [6] with some modifications. R. quercivora and related species Ophiostomataceae sp. Symbiotic Fungi of Platypus quercivorus 109 Table 2 Fungus species isolated during this study Species D1/D2 of the LSU rDNA GenBank accession no. Designated closest relativea (GenBank accession no.) and the sequence similarity to it (%) Isolate no. Voucher strain QmPlEG-2-13 JCM 14990T =CBS 10899T AB291675 A. kamigamensis (AB291675), 100 QmPlVG-3-9 JCM 14991 AB296361 A. kamigamensis (AB291675), 99.8 Ambrosiozyma neoplatypodis QmPlVG-2-21 JCM 14992T =CBS 10900T AB291676 A. neoplatypodis (AB291676), 100 Candida guilliermondii CcPqEG-2-1 JCM 16743 AB552926 Meyerozyma guilliermondii (AB260129), 100 Candida kashinagacola QmPlEG-2-14 JCM 15019T =CBS 10903T AB291672 C. kashinagacola (AB291672), 100 Candida pseudovanderkliftii QmPlLG-2-9 JCM 15025T =CBS 10904T AB291673 C. pseudovanderkliftii (AB291673), 100 Candida vanderkliftii QmPlLG-1-42 JCM 15029T =CBS 10905T AB291674 C. vanderkliftii (AB291674), 100 Candida sp. 3 QmPlEG-2-8 JCM 15000=CBS 10902 AB291677 Candida insectalens (U62304), <90 Candida sp. 4 QmPlEG-2-5 JCM 14994 AB291678 Ogataea philodendri (U75522), 98.6 Candida sp. 10-2 QmPlPB-1-59 JCM 15018 AB291684 Candida silvatica (U76201), <90 Candida sp. 10-3 QmPlLG-1-45 JCM 16747 AB552927 C. silvatica (U76201), 95.8 Candida sp. 12 QmPlVG-2-49 JCM 15013 AB291685 Saccharomycopsis crataegensis (U40079), 91.4 Ascomycota Yeasts Ambrosiozyma kamigamensis Candida sp. 13 QmPlVG-3-41 JCM 15017 AB291686 Candida nemodendra (U70246), 97.3 Candida sp. 17-1 CcPqEG-1-1 JCM 16754 AB552929 Metschnikowia agaves (U84243), 95.3 Candida sp. 17-2 CcPqEG-2-9 JCM 16758 AB552930 M. agaves (U84243), 95.6 Candida sp. 22 QmPlLG-1-48 JCM 16738 AB552924 S'copsis selenospora (U40099), 98.2 Millerozyma phetchabunensis QmPlEG-2-12 JCM 14999 AB291679 Millerozyma phetchabunensis (AB371638), 100 Pichia galeiformis (U75738), 99.1 Pichia sp. 10-1 QmPlLG-3-1 JCM 16750 AB291683 Saccharomycopsis sp. QsPq3PC-2-1 JCM 16741 AB552925 S'copsis malanga (U40135), <90 Wickerhamomyces sp. QsPq3PC-2-18 JCM 16752 AB552928 Candida ponderosae (AF271085), 98.1 Filamentous fungi Clavicipitaceae sp. QsPq3PC-2-42 JCM 16797 AB552945 Chaunopychnis pustulata (AF373282), 99.5 Hypocreales sp. 23-3 CcPqLG-3-23 JCM 16796 AB552944 Melanospora fallax (U17404), 99.5 Hypocreaceae sp. 23-1 CcPqEG-2-25 JCM 16794 AB552942 Trichoderma koningiopsis (FJ430784), 100 Ophiostomataceae sp. 7 QmPlEG-2-38 JCM 16978 AB576770 Raffaelea ambrosiae (EU984297), 95.4 Ophiostomataceae sp. 8 QmPlEG-3-41 JCM 16780 AB291681 Raffaelea sulcati (EU177462), 93.1 Ophiostomataceae sp. 9 QmPlEG-3-47 JCM 16782 AB291682 R. ambrosiae (EU177453), 96.2 Ophiostoma sp. 19-1 CcPqEG-1-11 JCM 16786 AB552939 Ophiostoma nigrocarpum (EF506941), 99.5 Penicillium sp. 18-1 CcPqEG-1-9 JCM 16790 AB552940 Penicillium paneum (AB479273), 100 Penicillium sp. 18-2 CcPqEG-3-26 JCM 16792 AB552941 Penicillium glabrum (AF033407), 100 Raffaelea quercivora QmPlLG-2-21 JCM 15683 AB552937 Raffaelea quercivora (AB496474), 100 QmPlPB-2-17 JCM 15682 AB552938 R. quercivora (AB496474), 100 Trichocomaceae sp. 6 QmPlEG-2-45 JCM 16788 AB291680 Penicillium resedanum (AF033398), 95.4 Trichoderma sp. 23-2 CcPqEG-2-26 JCM 16795 AB552943 Hypocrea lixii (FJ890420), 100 Basidiomycota Yeasts Microbotryomycetidae sp. 21-1 CcPqEG-2-17 JCM 16760 AB552931 Rhodotorula diffluens (AF075485), 97.6 Microbotryomycetidae sp. 21-3 CcPqLG-3-25 JCM 16762 AB552932 Sporobolomyces inositophilus (AF189987), 96.6 Microbotryomycetidae sp. 21-5 CcPqEG-3-20 JCM 16764 AB552933 Rhodotorula fushanensis (AB176591), 96.9 Tremellaceae sp. 21-2 CcPqEG-3-16 JCM 16766 AB552934 Tremella fuciformis (AF042228), 98.3 Tremellaceae sp. 21-4 CcPqEG-1-19 JCM 16767 AB552935 Tremella globispora (EF551317), 96.8 Tremellaceae sp. 24 CcPqEG-2-36 JCM 16768 AB552936 Sterigmatosporidium polymorphum (AF075480), 99.0 AB576771 Peniophora cinerea subsp. fagicola (AF506424), 96.6 Filamentous fungus Peniophoraceae sp. QmPlVG-1-41 T type strain a Highest match inferred by BLASTn search 110 7, 8, 9, and Ophiostoma sp. 19–1 (Table 2; Table 2 in the Electronic Supplementary Materials), as well as yeasts, were subcultured in 4% (wt./vol.) malt extract broth (MEB; Difco) at 25°C for 2–7 days in a shaker incubator at 200 rpm, where R. quercivora grows mainly in yeast form [11] which permitted use of the glass beads method for DNA extraction. A 2-ml aliquot of cell suspension was centrifuged in a microtube at 2,300×g for 5 min at 4°C to obtain the yeast-like cell mass and the supernatant was decanted. This procedure was repeated, if necessary, until enough about 100 μl of cell mass was obtained. The cell mass was rinsed with sterile 0.9% (wt./vol.) sodium chloride at least twice, and re-suspended in 500 μl of lysis buffer (50 mM Tris–HCl, 250 mM sodium chloride, 50-mM ethylenediaminetetraacetic acid (EDTA), 0.3% (wt./vol.) sodium dodecyl sulfate, pH 8). Approximately 200 μl (0.3 g) of glass beads (0.8 mm in diameter) were added to each tube, and the tubes were vortexed for 1.5 min to disrupt the cells. The tubes were incubated for 1 h at 65°C and chilled on ice. Next the tubes were mixed again for 30 s and then centrifuged at 20,400×g for 20 min at 4°C. Each supernatant (300 μl) was transferred into a new tube, to which an equivalent volume of phenolchloroform-isoamyl alcohol (25:24:1) was added and mixed to emulsify. After centrifugation at 20,400×g for 15 min at 16°C, the supernatant (200 μl) was put into a new tube, to which an equivalent volume of chloroformisoamyl alcohol (24:1) was added and mixed to emulsify. After centrifugation at 20,400×g for 15 min at 16°C, the supernatant (100 μl) was put into a new tube and 1/10 volume of 3 M sodium acetate (pH 5.2) and equal volume of isopropyl alcohol were added. The mixture was incubated at −20°C for at least 30 min and centrifuged at 20,400×g for 20 min at 4°C to sediment DNA. The resulting DNA pellet was rinsed with ice-cold 70% (vol./vol.) ethanol and dried. The pellet was resuspended in 200 μl of TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8) and kept at −20°C until used. b. DNA extraction from filamentous fungi. Filamentous fungi, except for R. quercivora, Ophiostomataceae sp. 7, 8, 9, and Ophiostoma sp. 19–1, for which the yeast phase was not easily obtained in liquid culture, were subcultured in MEB and the cell mass was prepared as above. The cell mass was rinsed with sterile 0.9% (wt./vol.) sodium chloride solution and then re-suspended in 500 μl of the lysis buffer. The solution in the tube was frozen in liquid nitrogen and thawed in water or a metal block bath at 65°C. This procedure was repeated three times, following incubation for 1 h at 65°C. Subsequently, the cell mass was disrupted with a micro-pestle and centrifuged at 20,400×g for 20 min at 4°C. Each supernatant (300 μl) was transferred into a new tube. R. Endoh et al. Next, the following manipulations of the DNA extraction were carried out as described in method (a) above. MSP-PCR Fingerprinting To determine the validity of visual grouping of the fungi and to select strains for rDNA sequencing, MSP-PCR fingerprinting was done for 1,202 out of 1,219 isolates which were selected as the representatives. The residual 17 isolates (13 from Q. laurifolia, three from C. cuspidata, and one from Q. robur) were excluded from MSP-PCR fingerprinting due to the failure in successive culturing. For MSP-PCR fingerprinting of fungal isolates, the microsatellite primer (GTG)5, synthesized by Invitrogen (Tokyo), was used [27]. The method for MSP-PCR fingerprinting was as reported previously [7]. When definite MSP-PCR fingerprints were not obtained, the extracted DNA was diluted ten times with sterile, distilled water, which was used as a template DNA for MSP-PCR. To determine the reproducibility of the banding pattern, MSP-PCR fingerprinting for the representative isolates was repeated at least twice. Group designations of MSP-PCR fingerprints were made by visual inspection of the banding pattern. Representative isolates for each of the 106 fungal strains were selected at random (Table 2; Table 2 in the Electronic Supplementary Materials), of which sequencing of the D1/D2 region of the LSU rDNA was done. Sequencing of the D1/D2 Region of the LSU rDNA, Similarity Search, and Phylogenetic Analysis Sequencing of the D1/D2 region of the LSU rDNA was done as described previously [7] for all the 106 strains given in Table 2 and Table 2 in the Electronic Supplementary Materials. Sequence data were aligned by using a GENETYX-MAC program (Software Development, Tokyo). The nucleotide sequences determined in this study were deposited in the DDBJ/EMBL/GenBank under the accession numbers listed in Table 2. The sequences obtained were compared with those available in the GenBank database at the DDBJ using BLASTn program [1]. Similarity of the nucleotide sequence to that of the closest relative was calculated by pair wise alignment using ClustalW [37] and corrected by manual inspection. For yeast species related to Ambrosiozyma, nucleotide substitution rates were determined from Kimura’s two-parameter method [17] and a phylogenetic tree was constructed with the distant matrix-neighbor-joining (NJ) method [30] available in the ClustalW package. Phylogenetic trees were also constructed with the maximum parsimony (MP) method using PAUP* ver. 4.0b10 [36]. Maximum parsimony analyses Symbiotic Fungi of Platypus quercivorus were done with the heuristic search option using the treebisection-reconstruction algorithm as the branch-swapping options with 100 random sequence additions. All sites were treated as unordered and unweighted, with gaps treated as missing data. Statistics including tree length, consistency index (CI), retention index (RI), rescaled consistency index (RC), and homoplasy index (HI) were calculated with PAUP*. The topology of the phylogenetic tree was tested by bootstrap analysis (1,000 replications) [8] for both the NJ and MP analyses. Abundance, Dominancy, and Fungus Diversity Fungus abundance per length (millimeters) of beetle gallery was calculated using total fungal colony counts, dilution rate for plating, and gallery length in a wood chip. Dominancy of each fungus species was calculated as the ratio of the colony count of the species compared with the total colony counts. Validity of visual grouping was checked by comparing MSP-PCR fingerprints of each fungal group. If a different fingerprint type(s) was detected in a visually designated group, colony count of the group was emended based on the error rate. Error rate was calculated (number of isolates giving a fingerprint mismatched to the others of the group)/(number of the representatives (usually eight or more) of the group). For example, when the colony count which appeared on a master plate and was visually grouped as C. kashinagacola was 120, with the error rate of one of eight among the first representatives, then the colony count of C. kashinagacola was emended to 105. If the one representative out of eight of which the fingerprint was mismatched to that of C. kashinagacola gave a fingerprint matching that of C. vanderkliftii, then the residual 15 was added to the colony count of C. vanderkliftii. To evaluate the structure and diversity of fungi in beetle galleries, Shannon’s diversity index [33] and Pielou’s evenness index [29] were also calculated for each community observed from a single beetle gallery. Morphology, Yeast Ascospore Formation, and Sporulating Condition of Filamentous Fungi Some of morphological characteristics and ascospore formation were examined for the selected strains of yeasts listed in Table 2 according to Yarrow [42] and recent publications [25, 26] to designate genus affiliation. Morphology of the vegetative cells was observed by inoculating pre-incubated culture grown on YM agar plate at 25°C for 2–3 days. Individual strains were inoculated into YM broth (Difco) and incubated at 25°C. Cultures were examined microscopically for 3 days, or more if necessary, after inoculation. 111 Ascospore formation was tested for yeasts by inoculating the pre-incubated culture grown as described above. Individual yeast strains listed in Table 2 and Table 2 in the Electronic Supplementary Materials, except Candida sp. 17–1 and Candida sp. 17–2, were inoculated onto YM agar- and corn meal agar (Difco) slants and incubated at 25°C. For Candida sp. 17–1 and Candida sp. 17–2, 1/20diluted V8 agar slants were used and incubated at 15°C. Cultures were examined microscopically each week for 6 weeks. Crossing of two strains for ascospore formation was not performed. Sporulating conditions in the selected filamentous fungi were tested by incubating the cultures on PDA (Nissui, Tokyo) plates. Cultures were arbitrarily examined microscopically for up to 4 weeks. DNA Extraction from Microbial Pellet on Gallery Walls Gallery specimens used for T-RFLP analysis were also collected from five trees attacked by P. quercivorus in four locations in Kyoto Prefecture, Japan. Except for Q. crispula, which was collected from Hachodaira, Kyoto, the sample trees used for T-RFLP analysis were different ones from the trees which the wood chips listed in Table 1 were obtained for the fungus isolation by plating. Each gallery sample came from different beetle galleries. Samples used for the T-RFLP analysis were also collected from the galleries in which reproduction was successful and seemingly healthy larvae or pupae were active. Sample identity, gallery type, gallery length, host tree species, sampling site, and sampling date are given in Table 3 in the Electronic Supplementary Materials. Wood chips containing a single beetle gallery were obtained as described in the methods for fungus isolation. Wood chips were further broken into smaller pieces to obtain 30–50% of the entire gallery wall. A microbial pellet was removed from the wall with a small, sterile spatula and then suspended in 1 ml of sterile 0.9% (wt./vol.) sodium chloride solution, and washed at least once. The DNA extraction and purification were done using ISOFECAL for beads beating DNA Isolation kit (Nippon Gene, Toyama). The microbial pellet sample was resuspended in DNA extraction buffer provided with the kit. Subsequent manipulations were performed based on the manufacturer’s instructions. The extracted DNA was dissolved in 50 μl of TE buffer and kept at −20°C until used. T-RFLP Analysis T-RFLP analysis was done using gallery wall samples and selected single species (strains used for T-RFLP analysis are shown in parenthesizes), i.e., Candida sp. 3 (JCM 15000), C. kashinagacola (JCM 15019T), Ambrosiozyma kamigamensis (JCM 14990T), Ambrosiozyma neoplatypodis (JCM 14992T), 112 and R. quercivora (JCM 15682, JCM 15683). The D1/D2 region of the LSU rDNA was amplified using universal primers NL1-FAM (forward) and tailed-NL4 (reverse). For T-RFLP analysis NL1 was labeled at the 5′ end with 6carboxyfluorescein. Tailed-NL4 consisted of NL4 with seven nucleotide sequence (information protected by the manufacturer) at the 5′ end. Both primers were synthesized by Applied Biosystems (Foster City, Calif.). Amplification reactions with primers NL1-FAM and tailed-NL4 were performed in 50 μl of reaction mixture containing 5 μl of dissolved DNA (single species: 100 ng; gallery wall sample: intact dissolved DNA), 1.25 U of ExTaq DNA polymerase (TaKaRa Bio Inc., Ohtsu), 5 μl of the reaction buffer, 5 μl of dNTP mix (2.5 mM each), and 10 pmol of each primer. LSU rDNAs were amplified in GeneAmp® PCR System 9700 (Applied Biosystems) using the following program: an initial denaturation at 94°C for 5 min, followed by 36 cycles of 10 s at 98°C, 30 s at 55°C, 1 min at 72°C, and a final extension of 7 min at 72°C. Amplified DNA was verified by electrophoresis of aliquots of PCR mixtures in 1.5% agarose in TAE buffer. The PCR products were purified by polyethylene glycol (PEG) precipitation method [31] with some modifications. A 50-μl aliquots of the PCR product was mixed with 12 μl of 3 M sodium acetate, and 30 μl of PEG solution (40% (wt./vol.) PEG 6000 and 10-mM magnesium dichloride), and gently shaken for 10 min at room temperature, and centrifuged 20,400×g for 15 min at 16°C. The supernatant was removed carefully by pipetting, and then the precipitated DNA was washed twice with 70% (vol./vol.) ethanol and dissolved in 40 μl of sterile distilled water. Purified PCR products were stored at −20°C in darkness until used. Terminal restriction fragment lengths (T-RFLs) of the D1/D2 region of the LSU rDNA of all the fungus species isolated here, all designated Ambrosiozyma species, and Candida llanquihuensis were predicted using the BioEdit software version 7.0.3 [9]. Based on the predicted T-RFLs, restriction enzymes were selected to detect species-specific terminal restriction fragment (T-RF) peaks for Candida sp. 3, C. kashinagacola, A. kamigamensis, A. neoplatypodis, and R. quercivora. The purified PCR product (100 ng of the amplicon) was digested with either 8 U of RsaI, 8 U of Hpy188III, 8 U of SfaNI (New England Biolabs, Beverly, Mass.), or 3 U of EcoRV (TaKaRa Bio Inc.), in a total volume of 10 μl at 37° C for 6 h with vortex-mixing at every 2 h. EcoRV was additionally used only in case that a species non-specific 121 bp T-RF peak was observed when digested with Hpy188III. For RsaI, Hpy188III, and SfaNI, preliminary experiments conducted with various enzyme concentrations (4, 8, and 12 U) for 6 h using the purified PCR products from gallery samples demonstrated that 8 U was sufficient for complete digestion. In the case of a rare cutter EcoRV, R. Endoh et al. preliminary experiments were similarly conducted as described above with various concentrations (1.5, 3, and 4.5 U) for 6 h using the purified PCR product from A. neoplatypodis demonstrated that 3 U was sufficient for complete digestion. Two μl of the reaction digest product was mixed with 7.5 μl of Hi-Di™ formamide (Applied Biosystems) and 0.5 μl of DNA fragment length standard GeneScan 1200 LIZ Size Standard (Applied Biosystems). Each sample was denatured at 98°C for 3 min and then immediately chilled on ice. T-RFL was determined on an ABI PRISM 3130 Genetic Analyzer (Applied Biosystems) GeneMapper mode using 36 cm long to detector by 50 μm inner diameter capillaries, POP-7 polymer (Perkin-Elmer, Applied Biosystems), a 16 s injection time, 1.2 kV injection voltage, 15.0 kV run voltage, and 60°C capillary temperature. Fragment sizes were estimated using the Local Southern Method in GeneMapper Software v3.7 (Applied Biosystems) or Peak Scanner Software v1.0 (Applied Biosystems). Only 20–700 bp peaks of GeneScan 1200 LIZ Size Standard were used for sizing because the size of the majority of the amplicons using the primer pair NL1 and NL4 for fungi was known to be less than 700 bp. Fragments were resolved to one base pair by manual alignment of the size standard peaks from different electropherograms. T-RFs with a peak height below 25 fluorescence units were excluded from the analysis. Reproducibility of the patterns was confirmed for repeated T-RFLP analysis of the D1/D2 region of the LSU rDNA using the same DNA extracts from the gallery samples and the single fungus species. Results Isolation, Count and Typing of Fungi, MSP-PCR Fingerprinting, Similarity Search, and Phylogenetic Analysis Fungal colonies grown on the master YM agar plates were first visually classified based on the cultural characteristics of the macro colony and counted. Twelve hundred and nineteen colonies; 557 from 11 galleries in Q. laurifolia, 288 from 11 galleries in C. cuspidata, 99 from three galleries in Q. serrata, 112 from four galleries in Q. crispula, and 163 from five galleries in Q. robur were initially selected and picked up from the master plates as the representatives of each visually designated fungus groups. These isolates were then further grouped by MSPPCR fingerprints generated with the microsatellite primer (GTG)5. The fingerprints were well-conserved within species, but differentiated between species that are closely related, e.g., A. kamigamensis and A. neoplatypodis, although trivial intraspecific differentiations were detected Symbiotic Fungi of Platypus quercivorus 113 designated closest relative inferred by BLASTn search and its percent identity, and sporulating condition are given in Table 2 and Table 2 in the Electronic Supplementary Materials. For two species, R. quercivora and Ophiostomataceae sp. 7, the sequences of the 5′ end of the D1/D2 region of the LSU rDNA could not be sequenced. Thirty-eight fungus species were detected. The sequences of the D1/D2 region of the LSU rDNA were completely conserved within the species except for those of A. kamigamensis. Based on a similarity search for nucleotide sequences, 29 species were determined to be ascomycetous, and seven in some species (data not shown). For the Candida sp. 3, C. kashinagacola, and R. quercivora, grouping which was based on visual observation of fungal colonies gave few errors because of their characteristic colony texture (Table 4 in the Electronic Supplementary Materials). The final representatives (Table 2; Table 2 in the Electronic Supplementary Materials) were selected at random from the first representatives and used for sequencing the D1/D2 region of the LSU rDNA. Species, isolate number of the final representatives, voucher strain, GenBank accession number of nucleotide sequences determined, 100 65 50 0.01 12 Candida sp. JW01-7-11-1-4-y2 (AY242326) Candida llanquihuensis (U70190) 20 Candida sp. ST-246 (DQ404487) 11 4 Candida kashinagacola (AB291672) 5 Candida vanderkliftii (AB291674) 6 Candida pseudovanderkliftii (AB291673) 73 100 77 15 Ambrosiozyma monospora (U40106) 16 Ambrosiozyma cicatricosa (U40128) 53 17 Ambrosiozyma platypodis (U40083) 80 Ambrosiozyma ambrosiae (U73605) 19 64 Ambrosiozyma philentoma (U40113) 18 96 Ambrosiozyma sp. NRRL Y-6106 (EU011596) 14 63 77 7 Ambrosiozyma kamigamensis (AB291675) 8 Ambrosiozyma neoplatypodis (AB291676) 21 Ambrosiozyma angophorae (U75521) Candida sp. BG02-7-20-001A-2-1 (AY520418) 13 98 22 Candida nanaspora (U70187) 59 23 d d nitratophila C Candida h l (U70180) 100 Candida cariosilignicola (U70188) 25 Ogataea methylivora (U75525) 26 49 Ogataea naganishii (EU011601) 48 Candida methanosorbosa (U70186) Candida pini (U70252) 34 94 37 Ogataea dorogensis (EU011620) Candida xyloterini (FJ381703) 38 28 Ogataea kodamae (U75525) Ogataea minuta var. minuta (U75515) 85 40 Ogataea minuta var. nonfermenta (U75518) 41 Ogataea polymorpha (U75524) 29 85 9Candida sp. 4 (AB291678) 99 39 Candida sp. BG02-7-20-019A-1-1 (AY520354) 27 Ogataea philodendri (U75522) Candida succiphila 47 p ((U70186)) 100 Ogataea pini (U75527) 44 95 45 Ogataea henricii (U75519) 56 Ogataea glucozyma (U75520) 46 91 10 Candida sp. 13 (AB291686) Candida nemodendra (U70246) 30 94 42 Ogataea finlandica (U75517) 43 Candida maris (U70181) 31 Candida sonorensis (U70185) Ogataea 32 salicorniae (U75966) 33 Candida pignaliae (U70183) Candida arabinofermentans (AF017248) 24 Kuraishia capsulata (U70178) Saccharomyces cerevisiae (U44806) Figure 1 Phylogenetic tree of the Ambrosiozyma clade and the related taxa based on the nucleotide sequences of the D1/D2 region of the LSU rDNA. The tree was generated by the distant matrixneighbor-joining method. Bootstrap values (each expressed as a percentage of 1,000 replications) of ≥50% are given at the nodes. The species isolated are shown in bold letters. Saccharomyces cerevisiae served as the outgroup 114 R. Endoh et al. had a basidiomycetous affiliation. Among them 19 and six were yeast species, respectively. Based on the morphological observations, results of ascospore formation, and phylogenetic analyses, yeast species belonging to the genera Millerozyma, Pichia, Saccharomycopsis, and Wickerhamomyces were detected along with Ambrosiozyma and Candida (Table 2). Filamentous fungi belonging to the genera Ophiostoma, Penicillium, and Trichoderma were also detected. A phylogenetic tree determined from the neighborjoining method indicated that two Ambrosiozyma species (A. kamigamensis and A. neoplatypodis) and three Candida species (C. kashinagacola, C. pseudovanderkliftii, and C. vanderkliftii) isolated in this study were located in the Ambrosiozyma clade (Fig. 1). MP analysis resulted in trees with similar topology as the NJ tree. The alignment data matrix consists of 573 characters, in which 377 were constant, 59 were parsimony-uninformative, and 137 were parsimony-informative. One hundred twenty-eight equally most parsimonious trees were constructed by the MP analysis (tree length=641, CI=0.432, RI=0.637, RC= 0.275, HI=0.568). Although MP trees differed in the branching of Ambrosiozyma species, A. kamigamensis and A. neoplatypodis were always nested within other Ambrosiozyma. Also, C. kashinagacola, C. pseudovanderkliftii, and C. vanderkliftii were always nested with C. llanquihuensis in the MP trees, as was the case of the NJ tree (Fig. 1). Candida sp. 4 and Candida sp. 13 were located in the neighboring taxon Ogataea clade (Fig. 1). Candida sp. 10–2, Candida sp. 10–3, Candida sp. 12, and Candida sp. 22 related to Saccharomycopsidaceae together with one Saccharomycopsis species. Candida sp. 17–1 and Candida sp. 17–2 were located in the Metschnikowiaceae. Although the closest relative of Candida sp. 3 was C. insectalens as inferred by BLASTn search, the phylogenetic position of the species could not be validated. Figure 2 Distribution and abundance of the fungus species isolated from the beetle galleries of Platypus quercivorus. The abbreviation of the gallery samples is presented in Table 1. The pie charts associated to each area refer to the CFU ratio of the fungus species; the legend for the segments of the pie charts is presented in the lower, right part. Fungus abundance is presented as values of (log10 CFU/beetle gallery (millimeters)) 6.15±0.46 82.3±17.7 0.87±0.08 0.64±0.10 5.59–6.57 72.9–90.4 0.81–1.01 0.54–0.75 4.81±0.16 84.1±9.8 0.98±0.13 0.73±0.17 4.61–4.97 75.4–98.3 0.78–1.06 0.56–0.95 3.76±0.61 88.3±4.2 0.88±0.13 0.52±0.19 3.14–4.36 84.3–92.6 0.76–1.02 0.39–0.74 5.93±0.84 86.6±16.0 0.64±0.31 0.45±0.13 4.60–6.97 51.3–97.5 0.38–1.32 0.35–0.74 CFU colony forming unit The data for the entrance of the galleries were excluded from calculation 5.45±0.74 86.4±8.7 0.98±0.23 0.53±0.12 Range Mean±SD Mean±SD Range Range Mean±SD Range Mean±SD Range 4.59–6.56 73.2–96.9 0.58–1.37 0.28–0.66 Fungal abundance (log10 CFU/gallery (mm)) Dominancy of Candida sp. 3+Candida kashinagacola (%) Shannon’s diversity index Pielou’s evenness index Indicator The abundance (including abundance per gallery length [mm]) of the various fungus species isolated from each wood chip is summarized in Table 5 in the Electronic Supplementary Materials. From the beetle galleries in Q. laurifolia, C. cuspidata, Q. serrata, Q. crispula, and Q. robur, 20, 21, 10, five, and seven species were isolated, respectively. Out of the 38 fungal species, only Candida sp. 3, C. kashinagacola, and R. quercivora were isolated from the beetle galleries regardless of host tree species with frequencies of 91.2% (31/34 gallery specimens), 94.1% (32/34), and 85.3% (29/34), respectively. Candida sp. 3 and C. kashinagacola were isolated from all the wood chips except those taken from gallery entrances of C. cuspidata. Sometimes, R. quercivora was not detected even from the interior of galleries in Q. laurifolia and Q. robur. Regarding the percent ratio of fungal colony counts, Candida sp. 3 and C. kashinagacola, or both, were most prevalent fungi (Fig. 2). Indeed, Candida sp. 3 plus C. kashinagacola accounted for about an average 85% of the fungi in all five tree species (see ‘dominancy of Candida sp. 3 + C. kashinagacola’ in Table 3), which were significantly higher than the sum of all the others for all tree species (Student’s t test, P<0.01). Although Candida sp. 3 tended to be the most prevalent fungus, this was not always the case for Q. laurifolia where C. kashinagacola replaced Candida sp. 3 as being the most prevalent (Fig. 2). Conversely, occurrence of the oak wilt pathogen, R. quercivora was rather low (Fig. 2). Although several other species of filamentous fungi such as Ophiostomataceae sp. 8 and Trichocomaceae sp. 6 were isolated, their presence was limited mainly to the entrance of the galleries in Q. laurifolia and substantially rare in the interior parts of galleries (Table 5 in the Electronic Supplementary Materials). Some yeast species were isolated from different tree species. For example, A. kamigamensis was isolated from four tree species except Q. crispula with frequency of 32.4% (15/34), while Candida sp. 13 from Q. laurifolia, Q. serrata, and Q. crispula with frequency of 26.5% (9/34) (Table 5 in the Electronic Supplementary Materials). Average fungus abundance, percent dominancy of Candida sp. 3 plus C. kashinagacola, Shannon’s diversity index, and Pielou’s evenness index in each tree species are summarized in Table 3 where the data for fungus isolations for the entrance of galleries in Q. laurifolia and C. cuspidata were excluded from the calculations. Fungus abundance in the beetle galleries ranged from 103.14 CFU/ gallery (mm; CFU, colony forming unit) in Q. serrata to 106.97 CFU/gallery (mm) in C. cuspidata. As mentioned above, Candida sp. 3 plus C. kashinagacola accounted for Table 3 Characterization of the fungus flora on the gallery walls of Platypus quercivorus in different host trees Frequency, Dominancy, Abundance, and Diversity Mean±SD 115 Quercus laurifolia (n=9) Castanopsis cuspidata (n=8) Quercus serrata (n=3) Quercus crispula (n=4) Quercus robur (n=5) Symbiotic Fungi of Platypus quercivorus 116 R. Endoh et al. Table 4 T-RFLs (bp) for the selected fungus species after digestion with RsaI, Hpy188III, SfaNI, and EcoRV Species Candida sp. 3 Candida kashinagacola Ambrosiozyma kamigamensis Ambrosiozyma neoplatypodis Raffaelea quercivora Frequencya Restriction enzyme RsaI Hpy188III SfaNI EcoRV 283–284 92–93 310 310 179, 181 475–476 217 591 (n.d.) 121 107, 108 521 (n.d.) 462–463 79 462–463 105, 106 521 (n.d.) 591 (n.d.) 591 (n.d.) 92 603, 605 a Bold-faced entries indicate the species-specific T-RFs b n.d. non-digested T-RF c Detected frequency from the 14 gallery samples about 85% (range=from 82.3±17.7 in Q. robur to 88.3± 4.2% in Q. serrata) of fungus colony counts on average in all five tree species. As for Shannon’s diversity index, a low value of 0.38 was obtained for C. cuspidata, but the value occasionally reached 1.32 or 1.37 for C. cuspidata and for Q. laurifolia, respectively. As for Pielou’s evenness index, the values ranged from 0.28 for Q. laurifolia to as high as 0.95 for Q. crispula. Figure 3 T-RFLP patterns of LSU rDNA D1/D2 region from beetle gallery (Ysd-LG1) of Platypus quercivorus digested with RsaI, Hpy188III, SfaNI, and EcoRV. Fungus speciesspecific T-RF peaks are indicated by arrows 14/14 14/14 12/14 7/14 0/14 T-RFLs of the Selected Fungus Species T-RFLs after digestion of the amplified LSU rDNA D1/D2 sequences with each RsaI, Hpy188III, SfaNI, and EcoRV were determined for Candida sp. 3, C. kashinagacola, A. kamigamensis, A. neoplatypodis, and R. quercivora (boldfaced entries in Table 4). Fungus species-specific T-RF peaks of the species listed in the table were obtained when Symbiotic Fungi of Platypus quercivorus digested with some of the four restriction enzymes: i.e., Candida species-specific peaks were generated after digestion with RsaI and Hpy188III, while SfaNI and EcoRV generated A. kamigamensis- and A. neoplatypodis-specific T-RF peaks, respectively (Table 4). T-RFLP Analysis of the Gallery Samples Frequency of the targeted fungi (Table 4) was calculated based on the existence or absence of species-specific T-RF peaks after digestion with four different restriction enzymes. Candida sp. 3 and C. kashinagacola were detected from all the gallery specimens (14/14 gallery specimens) by T-RFLP analyses with relative high T-RF peaks. However, a R. quercivora-specific peak could not be detected from any gallery sample. Two Ambrosiozyma species, which both were seldom isolated from beetle galleries, were detected with frequencies of 85.7% (12/14) for A. kamigamensis and 50% (7/14) for A. neoplatypodis by T-RFLP analyses. However, the relative peak heights (relative fluorescence units) of the Ambrosiozyma species were generally low (Fig. 3, where only the profiles yielded from Ysd-LG1 are shown). The detection frequencies of the two Ambrosiozyma species by T-RFLP analysis were higher than those by the culture-dependent method (44.1% (15/34) and 14.7% (5/34) for A. kamigamensis and A. neoplatypodis, respectively). Discussion In this study, we analyzed the fungal flora colonizing the galleries of the ambrosia beetle P. quercivorus. Since no 117 detailed studies focusing on the ecology of the fungal flora in beetle galleries of P. quercivorus are available we used both a traditional, culture-dependent plating method in combination with MSP-PCR fingerprinting, and a cultureindependent technique, T-RFLP analyses. We used isolation methods suitable for yeasts since the galleries of P. quercivorus were apparently covered with yeast-like fungi when observed by a light microscope. Here we conclude that Candida sp. 3, C. kashinagacola, and R. quercivora are the primary symbionts of P. quercivorus because they were isolated regardless of the host tree species in different seasons (Fig. 2). Since the gallery specimens in the five host tree species used for fungus isolation were obtained in different seasons, different locations, and different host tree sizes (DBH), we did not make any statistical comparisons among the host tree species for the indices in Table 3. Although the biomass of yeast species on the galleries might be overestimated compared with those of filamentous fungi as the isolation method we used is suitable for yeasts/ yeast-like fungi, it is very likely that Candida sp. 3 and C. kashinagacola predominated in the beetle galleries considering their highest frequencies of isolation. For Candida sp. 3 and C. kashinagacola, the constant relevance was further confirmed by the cultureindependent method, T-RFLP analyses (Table 4). The primer pair NL1 and NL4 used in this study may not amplify all the D1/D2 region of the LSU rDNA of fungi on the beetle galleries, and not accurately preserve the evenness of the original fungal community DNA template, as we were unable to detect R. quercivora by T-RFLP analyses (Table 4). The negative results may be attributed to an unevenness of the PCR efficacy during amplification Table 5 Fungi isolated from the platypodid beetle-associated sources Fungal lineage Candida sp. 3 relative Ambrosiozyma relativea Raffaelea relative a See Fig. 1 Platypodid ambrosia beetle Platypus quercivorus (Japan; this study) Platypus cylindrus (Portugal; [2,12,28]) Crossotarsus externedentatus (South Africa; [32,40,41]) Candida sp. 3 Candida kashinagacola Candida pseudovanderkliftii Candida vanderkliftii Ambrosiozyma kamigamensis Ambrosiozyma neoplatypodis Raffaelea quercivora Ophiostomataceae sp. 7 Ophiostomataceae sp. 8 Ophiostomataceae sp. 9 Ophiostoma sp. 19-1 Not recorded Ambrosiozyma platypodis Candida insectalens Ambrosiozyma ambrosiae Raffaelea montetyi Raffaelea albimanens 118 of the D1/D2 region of the LSU rDNA of R. quercivora resulting from the many G/C repeats in the region. By combining the subsequent sequencing of the D1/D2 region of the LSU rDNA, 38 fungal species were detected from the beetle galleries. All the fungi isolated could be clearly distinguished by combining the colony morphology, comparison of MSP-PCR fingerprints, and sequencing of the D1/D2 region of the LSU rDNA. Kurtzman and Robnett [23] predicted that strains showing greater than 1% nucleotide substitutions in the D1/D2 region of the LSU rDNA are likely to be different species. According to this guideline, 18 ascomycetous and six basidiomycetous yeasts isolated here might be undescribed species. Out of the undescribed ascomycetous (including mitosporic) yeasts, we have described five new species; A. kamigamensis, A. neoplatypodis, C. kashinagacola, C. pseudovanderkliftii, and C. vanderkliftii, all of which are members of the Ambrosiozyma clade placed near the Ogataea clade (Fig. 1) [6,7]. Kurtzman and Robnett [24] reported that the Ambrosiozyma clade separated from and was assumed a position basal to the Ogataea species determined from phylogenetic analyses of gene sequences for nuclear LSU and SSU rRNA, translation elongation factor-1α, and mitochondrial SSU rRNA. All the species in the Ambrosiozyma clade, including A. kamigamensis, A. neoplatypodis, C. kashinagacola, C. pseudovanderkliftii, and C. vanderkliftii isolated in this study, do not assimilate methanol [6,7,34], while species circumscribed in the Ogataea clade, including Candida sp. 4 and Candida sp. 13, with a few exception assimilate it. We also found several yeast species occurred in other taxa such as Saccharomycopsidaceae (e.g., Saccharomycopsis sp.) and Metschnikowiaceae (e.g., Candida sp. 17– 1). Taking into consideration that all of such species were detected from the galleries with low CFU ratio, however, they were likely to be arbitrary gallery-contaminating yeasts. In the galleries CcE-i, CcE-ii, and CcE-iii, the fungal flora was apparently different from others (Fig. 1). The reason is not known. Texture of the bark of C. cuspidata is smooth and relatively thin, while those of the Quercus species sampled in this study are rather rugged and thick. The fungal flora of the entrance of beetle galleries in C. cuspidata might be particularly subject to invasion or perturbation by non-symbionts of the host beetle due to the thickness and texture of bark. Batra [3] classified the symbiotic fungi of ambrosia beetles in terms of contribution as food sources into two categories, i.e., primary ambrosia fungi (PAF; main food source) and auxiliary ambrosia fungi (AAF; supplementary food source). It remains to be determined if Candida sp. 3, C. kashinagacola, and R. quercivora are PAF or AAF of P. quercivorus in terms of the acquisition of nutrition. In contrast to Candida sp. 3 and C. kashinagacola, Ambrosiozyma species were only occasionally isolated R. Endoh et al. (44.1% (15/34 gallery specimens) and 8.8% [3/34] for A. kamigamensis and A. neoplatypodis, respectively) with relative low CFU ratios (Fig. 2). Since the two Ambrosiozyma species were detected by T-RFLP analyses with higher frequencies of 85.7% (12/14) and 50% (7/14) (Table 4), respectively, and they grow well on YM agar, their lower detection via the culture-dependent method may have resulted from substantially low ratios of cell existence in the beetle galleries compared with those for Candida sp. 3 and C. kashinagacola. Of interest here was the fact that, the closest relative of Candida sp. 3, C. insectalens was isolated from larval cradles of a platypodid ambrosia beetle Crossotarsus externedentatus in South Africa [40]. Also, A. ambrosiae and Raffaelea albimanens, of which relative species were C. kashinagacola (Fig. 1) and R. quercivora, respectively (Table 5), have been isolated from the beetle [32,39]. Likewise, A. platypodis and Raffaelea montetyi were isolated from Platypus cylindrus, which has been suggested as being related to cork oak decline in Portugal and Mediterranean countries [2,12,28]. Closely related species of Candida sp. 3 have not been recorded from P. cylindrus-related sources. These observations demonstrate that, platypodid ambrosia beetles likely closely related with three different lineages of fungi; namely, C. insectalens relatives, Ambrosiozyma relatives, and Raffaelea relatives (Table 5). Although not studied, work needs to be done on how the symbionts of P. quercivorus are transmitted from tree to tree. Suh et al. [35] reported that the guts of a variety of beetles from the southeastern USA and Panama were a hidden habitat of yeasts. Perhaps Candida sp. 3 or C. kashinagacola, or both may be transmitted via the digestive tract as well as the mycangia. The symbiont(s) transmitted specifically via mycangia, if any, would be most important one(s) for a host beetle, because such symbiont(s) is required to provide certain benefits to the host beetle and to increase fitness even at the cost of maintaining the mycangia. The next research challenge is to determine which species are transmitted via the mycangia of P. quercivorus. Acknowledgments This research was supported in part by a Grant-inAid for Scientific Research (A) from the Ministry of Education, Sports, Culture, Science and Technology of Japan (no. 18208015 to KF, 2006– 2008) and a grant from the Japan Society for the Promotion of Science (no. 21·1976 to RE, 2009–2010). We are grateful to Dr. J. Sutherland for advice on preparing the manuscript and correcting the English. We thank Dr. M. Kobayashi (Kyoto Pref. Univ.), the technical staffs at Kamigamo Experimental Station of Kyoto University, Mr. H. Qi (Kyoto Univ.), and Mr. T. Hagus (Kyoto Univ.), and other members of our labortory for their precious assistance during wood sample collection. We also thank Dr. M. Sakamoto (RIKEN), and Dr. Y. Benno (RIKEN) for their technical advice on T-RFLP analysis. Too, we thank Dr. K.-D. An (RIKEN) for his assistance on optimizing the PCR condition especially for R. quercivora. We deeply appreciate anonymous reviewers for their helpful suggestions on our manuscript. Symbiotic Fungi of Platypus quercivorus References 1. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search program. Nucleic Acids Res 25:3389–3402 2. Baker JM, Kreger-van Rij NJW (1964) Endomycopsis platypodis sp. n. (Ascomycetes): auxiliary ambrosia fungus of Platypus cylindrus Fab. (Col. Platypodidae). Antonie Leeuwenhoek 30:433–441 3. Batra LR (1966) Ambrosia fungi: extent of specificity to ambrosia beetles. Science 153:193–195 4. Brasier CM (1991) Ophiostoma novo-ulmi sp. nov., causative agent of current Dutch elm disease pandemics. Mycopathologia 115:151–161 5. Brasier CM (2000) Intercontinental spread and continuing evolution of the Dutch elm disease pathogens. In: Dunn CP (ed) The elms: breeding, conservation, and disease management. Kluwer Academic Publishers, Dordrecht, pp 61–72 6. Endoh R, Suzuki M, Benno Y (2008) Ambrosiozyma kamigamensis sp. nov. and A. neoplatypodis sp. nov., two new ascomycetous yeasts from ambrosia beetle galleries. Antonie Leeuwenhoek 94:365–376 7. Endoh R, Suzuki M, Benno Y, Futai K (2008) Candida kashinagacola sp. nov., C. pseudovanderkliftii sp. nov. and C. vanderkliftii sp. nov., three new yeasts from ambrosia beetleassociated sources. Antonie Leeuwenhoek 94:389–402 8. Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783–791 9. Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acids Symp Ser 41:95–98 10. Harrington TC (2005) Ecology and evolution of mycophagous bark beetles and their fungal partners. In: Vega FE, Blackwell M (eds) Insect-fungal associations. Oxford University Press, Oxford, pp 257–291 11. Hata K, Ohkusu M, Takeo K, Sone K (2004) Dimorphism of Raffaelea quercivora and Hormoascus sp. obtained from gallery walls of ambrosia beetles on liquid and agar media. Kyusyu J For Res 57:117–119 12. Inácio ML, Henriques J, Lima A, de Sousa EMR (2008) Fungi of Raffaelea genus (Ascomycota: Ophiostomatales) associated to Platypus cylindrus (Coleoptera: Platypodidae) in Portugal. Revista de Ciências Agrárias de Portugal 31:96–104 13. Ito S, Kubono T, Sahashi N, Yamada T (1998) Associated fungi with the mass mortality of oak trees. J Jpn For Soc 80:170–175 14. Ito S, Yamada T (1998) Distribution and spread of the mass mortality of oak trees. J Jpn For Soc 80:229–232 15. Jones KG, Blackwell M (1998) Phylogenetic analyses of ambrosial species in the genus Raffaelea based on 18S rDNA sequences. Mycol Res 102:661–665 16. Kajimura H, Hijii N (1992) Dynamics of the fungal symbionts in the gallery system and the mycangia of the ambrosia beetle, Xylosandrus mutilatus (Blandford) (Coleoptera: Scolytidae) in relation to its life history. Ecol Res 7:107–117 17. Kimura M (1980) A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J Mol Evol 16:111–120 18. Kirisits T (2004) Fungal associates of European bark beetles with special emphasis on the ophiostomatoid fungi. In: Lieutier F, Day KR, Battisti A, Grégoire J-C, Evans HF (eds) Bark and wood boring insects in living trees in Europe, a synthesis. Kluwer, Dordrecht, pp 202–214 19. Klepzig KD, Six DL (2004) Bark beetle-fungal symbiosis: context dependency in complex associations. Symbiosis 37:189–205 119 20. Kobayashi M, Ueda A (2005) Wilt disease of Fagaceae trees caused by Platypus quercivorus (Murayama) (Coleoptera: Platypodidae) and the associated fungus: aim is to clarify the damage factor. J Jpn For Soc 87:435–450 21. Kubono T, Ito S (2002) Raffaelea quercivora sp. nov. associated with mass mortality of Japanese oak, and the ambrosia beetle (Platypus quercivorus). Mycoscience 43:255–260 22. Kuroda K, Yamada T (1996) Discoloration of sapwood and blockage of xylem sap ascent in the trunks of wilting Quercus spp. following attack by Platypus quercivorus. J Jpn For Soc 78:84–88 23. Kurtzman CP, Robnett CJ (1998) Identification and phylogeny of ascomycetous yeasts from analysis of nuclear large subunit (26S) ribosomal DNA partial sequences. Antonie Leeuwenhoek 73:331– 371 24. Kurtzman CP, Robnett CJ (2010) Systematics of methanol assimilating yeasts and neighboring taxa from multigene sequence analysis and the proposal of Peterozyma gen. nov., a new member of the Saccharomycetales. FEMS Yeast Res 10:353–361 25. Kurtzman CP, Robnett CJ, Basehoar-Powers E (2008) Phylogenetic relationships among species of Pichia, Issatchenkia and Williopsis determined from multigene sequence analysis, and the proposal of Barnettozyma gen. nov., Lindnera gen. nov. and Wickerhamomyces gen. nov. FEMS Yeast Res 8:939–954 26. Kurtzman CP, Suzuki M (2010) Phylogenetic analysis of ascomycete yeasts that form coenzyme Q-9 and the proposal of the new genera Babjeviella, Meyerozyma, Millerozyma, Priceomyces, and Scheffersomyces. Mycoscience 51:2–14 27. Meyer W, Latouche GN, Daniel H-M, Thanos M, Mitchell TG, Yarrow D, Schönian G, Sorrell TC (1997) Identification of pathogenic yeasts of the imperfect genus Candida by polymerase chain reaction fingerprinting. Electrophoresis 18:1548–1559 28. Morelet M (1998) Une espèce nouvelle de Raffaelea isolée de Platypus cylindrus, coléoptère xylomycétophage des chênes. Ann Soc Sci Nat Arch Toulon Var 50:185–193 29. Pielou EC (1966) The measurement of diversity in different types of biological collections. J Theoret Biol 13:131–144 30. Saito N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic tree. Mol Biol Evol 4:406–425 31. Sakamoto M, Hayashi H, Benno Y (2003) Terminal restriction fragment length polymorphism analysis for human fecal microbiota and its application for analysis of complex bifidobacterial communities. Microbiol Immunol 47:133–142 32. Scott DB, Du Toit JW (1970) Three new Raffaelea species. Trans Br Mycol Soc 55:181–186 33. Shannon CE (1948) A mathematical theory of communication. Bell Syst Technol J 27:379–423 34. Smith MT (1998) Ambrosiozyma van der Walt. In: Kurtzman CP, Fell JW (eds) The yeasts, a taxonomic study, 4th edn. Elsevier, Amsterdam, pp 129–133 35. Suh S-O, McHugh JV, Pollock DD, Blackwell M (2005) The beetle gut: a hyperdiverse source of novel yeasts. Mycol Res 109:261–265 36. Swofford DL (1998) PAUP* 4.0: phylogenetic analysis using parsimony. Sinauer Associates, Sunderland 37. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673– 4680 38. Urano T (2000) Relationships between mass mortality of two oak species (Quercus mongolica Turcz. var. grosseserrata Rehd. et Wils. and Quercus serrata Thunb.) and infestation by and 120 reproduction of Platypus quercivorus (Murayama) (Coleoptera: Platypodidae). J For Res 5:187–193 39. van der Walt JP (1972) The yeast genus Ambrosiozyma gen. nov. (Ascomycetes). Mycopathol Mycol Appl 46:305–316 40. van der Walt JP, Scott DB, van der Klift WC (1971) Five new Torulopsis species from South Africa insect sources. Antonie Leeuwenhoek 37:461–471 R. Endoh et al. 41. van Zandt PA, Townsend VR Jr, Carlton CE, Blackwell M, Mopper S (2003) Loberus impressus (LeConte) (Coleoptera: Erotylidae) fungal associations and presence in the seed capsules of Iris hexagona. Coleopt Bull 57:281–288 42. Yarrow D (1998) Methods for the isolation, maintenance and identification of yeasts. In: Kurtzman CP, Fell JW (eds) The yeast, a taxonomic study, 4th edn. Elsevier, Amsterdam, pp 77–100