Research Sebacinales are common mycorrhizal associates of Ericaceae Blackwell Publishing Ltd Marc-André Selosse1*, Sabrina Setaro2*, Florent Glatard1*, Franck Richard1, Carlos Urcelay3 and Michael Weiß2 1 Centre d’Ecologie Fonctionnelle et Evolutive (CNRS, UMR 5175), Equipe coévolution, 1919 Route de Mende, 34293 Montpellier cedex 5, France; 2 Botanisches Institut, Universität Tübingen, Auf der Morgenstelle 1, D-72076 Tübingen, Germany; 3Instituto Multidisciplinario de Biologia Vegetal and FCEFyN, Universidad Nacional de Cordoba, CONICET, CC 495, 5000 Cordoba, Argentina; *These authors contributed equally to this work. Summary Author for correspondence: Marc-André Selosse Tel: +33 467 61 32 31 Fax: +33 467 41 21 38 Email: ma.selosse@wanadoo.fr Received: 8 December 2006 Accepted: 11 February 2007 • Previous reports of sequences of Sebacinales (basal Hymenomycetes) from ericoid mycorrhizas raised the question as to whether Sebacinales are common mycorrhizal associates of Ericaceae, which are usually considered to associate with ascomycetes. • Here, we sampled 239 mycorrhizas from 36 ericoid mycorrhizal species across the world (Vaccinioideae and Ericoideae) and 361 mycorrhizas from four species of basal Ericaceae lineages (Arbutoideae and Monotropoideae) that do not form ericoid mycorrhizas, but ectendomycorrhizas. Sebacinales were detected using sebacinoidspecific primers for nuclear 28S ribosomal DNA, and some samples were investigated by transmission electron microscopy (TEM). • Diverging Sebacinales sequences were recovered from 76 ericoid mycorrhizas, all belonging to Sebacinales clade B. Indeed, some intracellular hyphal coils had ultrastructural TEM features expected for Sebacinales, and occurred in living cells. Sebacinales belonging to clade A were found on 13 investigated roots of the basal Ericaceae, and TEM revealed typical ectendomycorrhizal structures. • Basal Ericaceae lineages thus form ectendomycorrhizas with clade A Sebacinales, a clade that also harbours ectomycorrhizal fungi. This further supports the proposition that Ericaceae ectendomycorrhizas involve ectomycorrhizal fungal taxa. When ericoid mycorrhizas evolved secondarily in Ericaceae, a shift of mycobionts occurred to ascomycetes and clade B Sebacinales, hitherto not described as ericoid mycorrhizal fungi. Key words: ectendomycorrhizas, Ericaceae evolution, ericoid mycorrhizas, molecular ecology, Sebacinales. New Phytologist (2007) 174: 864–878 © The Authors (2007). Journal compilation © New Phytologist (2007) doi: 10.1111/j.1469-8137.2007.02064.x Introduction Recent works on mycorrhizal communities have stimulated considerable interest in a neglected group of fungi related to the genus Sebacina, recently raised to the order Sebacinales (Weiß et al., 2004). This basal order of Hymenomycetes (Basidiomycetes) encompasses fungi with longitudinally septate basidia and imperforate parenthesomes (i.e. the derivates of the endoplasmic reticulum covering septal pores and allowing communication between cells). They also lack cystidia and structures formed during cytokinesis on some basidiomycetous 864 hyphae, the so-called clamp connections. Cultivable species exhibit monilioid hyphae (i.e. hyphal cells that look like pearls in a chain). This is why sebacinoids with no known sexual stage were placed in the polyphyletic form genus Rhizoctonia. Although some sebacinoids are described as valid species, and some can grow in pure culture, most of our knowledge on Sebacinales and their diverse host species comes from molecular ecology studies during the last 4 yr, that is, from direct amplification of fungal ribosomal DNA (rDNA) of environmental samples. Phylogenetic analysis, using sequences from cultures, fruitbodies or environmental samples, revealed that www.newphytologist.org Research Sebacinales are divided into two clades, A and B, that differ in their ecology (Weiß et al., 2004). The first clade A sequences were derived from fruitbodies (Weiß & Oberwinkler, 2001), and subsequently from mycorrhizas of the achlorophyllous orchids Neottia nidus-avis (McKendrick et al., 2002; Selosse et al., 2002a) and Hexalectris spicata (Taylor et al., 2003), as well as of some photosynthetic orchids related to N. nidus-avis that are partly heterotrophic (Selosse et al., 2004; Julou et al., 2005). At the same time, many clade A sebacinoids, including those from the orchids just mentioned, were demonstrated to form ectomycorrhizas on tree roots (Selosse et al., 2002a,b; Urban et al., 2003; Walker & Parrent, 2004; Moyersoen, 2006). They are among the most common ectomycorrhizal species in temperate and Mediterranean forests (Glen et al., 2002; Avis et al., 2003; Kennedy et al., 2003; Walker et al., 2004; Richard et al., 2005; Tedersoo et al., 2006). Up to now, fruitbodies are only known from clade A sebacinoids. Clade B displays a larger array of associations: some species are mycorrhizal on green, autotrophic orchids (Warcup, 1988; Bougoure et al., 2005), and others associate with liverwort thalli (Kottke et al., 2003). Clade B also contains the nonspecific root endophyte Piriformospora indica (Verma et al., 1998, 2001; PeSkan-Berghöfer et al., 2004). Recently, clade B rDNA sequences were amplified from roots of several Ericaceae (Berch et al., 2002; Allen et al., 2003; Bougoure & Cairney, 2005; Setaro et al., 2006a), raising the intriguing possibility that sebacinoids could be overlooked mycorrhizal fungi of these plants. Ericaceae form a large, ecologically relevant family of trees and shrubs growing all over the world (Kron et al., 2002a), especially at high latitude and high elevation stands. Most Ericaceae have a particular mycorrhizal association adapting them to the nutrient-poor, acidic soils of such areas (Smith & Read, 1997). In the so-called ericoid mycorrhiza (ERM), fungi colonize fine roots called ‘hair roots’, lacking cortical parenchyma, and form hyphal coils in large epidermal cells. Most cells are independently colonized from the soil (Bergero et al., 2000), with some cell-to-cell hyphal connections (Massicotte et al., 2005), so that each hair root harbours several fungi (Perotto et al., 1996). All ERM fungi reported so far are ascomycetes (Smith & Read, 1997; McLean et al., 1999; Cairney & Ashford, 2002; Midgley et al., 2004). However, the cloning of fungal rDNA sequences amplified from ERM roots suggested that sebacinoids colonized roots of Gaultheria shallon (Berch et al., 2002; Allen et al., 2003) and Epacris pulchella (Bougoure & Cairney, 2005). These sebacinoids are likely to have been ignored by classical approaches, as: (i) clampless hyphae of sebacinoids cannot be distinguished from ascomycetous hyphae by light microscopy; and (ii) in vitro isolation failed to reveal them because they have been difficult to get into culture up until now (Berch et al., 2002). However, electron microscope investigations (Bonfante-Fasolo, 1980; Peterson et al., 1980; Allen et al., 1989; Setaro et al., 2006b) have sometimes revealed that basidiomycetes with imperforate parenthesomes, the type present in Sebacinales, may colonize living roots of ERM Ericaceae. Recently, an unusual mycorrhizal association with clade B sebacinoids was reported from the neotropical Cavendishia nobilis (Ericaceae), a member of the Vaccinioideae (Kron et al., 2002a,b), with both intracellular colonization, as in ERM, and growth between and around the cortical parenchyma cells (cavendishioid mycorrhizas; Setaro et al., 2006a). These data raise three questions: (i) are sebacinoid partners common in other ERM species, especially from the large ERM clade Ericoideae; (ii) if so, do all these ERM sebacinoids belong to clade B, as suggested by available sequences (Weiß et al., 2004; Setaro et al., 2006a); and (iii) what are the ultrastructural features of sebacinoid infection in hair roots, as compared with those of ERM ascomycetes? ERM association is thought to have arisen once during Ericaceae evolution (Cullings, 1996; Kron et al., 2002a), and basal subfamilies (Pyroleae and Arbutoideae) display totally different mycorrhizas. They form ectendomycorrhizas (EEM), where fungi colonize both the cortical cells and the root surface, forming a sheath, an intercellular network between cortical cells, and intracellular structures that depend on the host taxon (Smith & Read, 1997). Here, a single fungal individual colonizes each root tip in most cases. Ascomycetes and/or basidiomycetes are present on EEM roots in Pyroleae (Robertson & Robertson, 1985; Bidartondo, 2005; Tedersoo et al., 2007) and Arbutoideae (Giovannetti & Lioi, 1990; Münzenberger et al., 1992; Richard et al., 2005). These fungi belong to species that usually form ectomycorrhizas on tree roots. Sebacinoids were reported to occur on EEM roots of Arbutus unedo (Arbutoideae; Richard et al., 2005) and Orthilia secunda (Pyroleae; Tedersoo et al., 2007), again raising three questions: (i) do sebacinoids often colonize EEM Ericaceae; (ii) which Sebacinales clade(s) are EEM plants associated with; and (iii) what are the ultrastructural features of their infection? To answer these questions on identity, diversity and structural interactions of sebacinoids associated with Ericaceae, we collected ERM and EEM roots, using worldwide sampling for ERM roots. First, detection of sebacinoids was achieved using specific PCR primers. Second, the sequences obtained were used to find their phylogenetic position. Third, microscope investigations allowed characterization of the ultrastructural features, and thus the mycorrhizal status, of these sebacinoids. Materials and Methods Root sampling Ericoid mycorrhiza roots were sampled from various sites in Europe, La Réunion Island (Indian Ocean), and North and South America between 2002 and 2004 (Table 1; Supplementary Material, Table S1). Freshly harvested ERM roots were checked for connection to aerial plant parts to ensure host identity. They were washed carefully under a dissection microscope in order to remove soil and organic matter particles © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 865 866 Research Table 1 A summary of ericoid mycorrhizal roots of Ericaceae (Vaccinioideae and Ericoideae) investigated and successful PCR amplification and sequencing of a sebacinoid ribosomal DNA sequence, using primers ITS3Seb and TW13 (for more details and GenBank accession numbers, see Supplementary Material) Taxa Region of origina No. of samples investigated Vaccinioideae Agauria buxi Agauria salicifolia Andromeda glauca Andromeda polifolia Chamaedaphne calyculata Chiogenes hispidula Gaultheria poeppigii Gaultheria procumbens Gaultheria sp. Vaccinium angustifolia Vaccinium macrocarpum Vaccinium myrtilloides Vaccinium myrtillus Vaccinium oxycoccos Vaccinium uliginosum Vaccinium vitis-idaea Re (1) Re (1) Ca (1) Eu (2) Ca (2) Ca (2) Ar (3) Ca (3) Ar (1) Ca (3) Ca (1) Ca (2) Eu (12) Ca (2), Eu (1) Eu (7) Eu (2) 115 3 5 2 8 4 7 22 6 2 5 2 4 20 6 11 8 64 (55.6%) 0 4 1 7 0 4 13 1 1 1 2 0 10 4 10 6 45 (39.1%) 0 3 (all identical) 1 5c 0 3 8 (5 identical) 1 0 1 2 0 7 2 7 5c Ericoideae Calluna vulgaris Empetrum nigrum Erica arborea Erica ciliaris Erica cinerea Erica gallioides Erica multiflora Erica reunionensis Erica vagans Kalmia angustifolia Kalmia polifolia Ledum palustre Loiseleuria procumbens Rhododendron canadensis Rhododendron conicum Rhododendron decorum Rhododendron ferrugineum Rhododendron fortunei Rhododendron groenlandicum Rhododendron japonicum Eu (19) Eu (3) Eu (3) Eu (1) Eu (9) Re (3) Eu (2) Re (4) Eu (2) Ca (2) Ca (2) Ca (1) Eu (1) Ca (2) Eu (1) Eu (1) Eu (1) Eu (1) Ca (3) Eu (1) 124 34 6 8 1 15 7 8 7 3 5 5 3 2 3 2 2 4 2 5 2 55 (44.3%) 20 3 2 1 7 1 6 3 1 1 0 3 0 1 0 1 3 1 1 0 31 (25.0%) 14 3 2 1c 4 0 2 1 1 0 0 1 0 0 0 0 1c 0 1 0 Subtotal for each global regiond Europe Canada Argentina Réunion Island Total 140 53 24 22 239 No. of successful amplificationsb 81 (57.8%)a 16 (30.2%)b 14 (58.3%)a 8 (40.9%)a b 119 (49.8%) No. of sequences recoveredb 54 10 8 4 76c,e (31.8%) a Ar, Argentina; Ca, Canada; Eu, Europe; Re, La Réunion. Brackets contain the number of independent sampled sites (i.e. areas of < 10 m2 separated by > 1 km from the other sites). b Percentages in brackets are related to the total number of samples investigated. c Five sequences were obtained as a consensus of at least seven clones obtained by cloning the PCR product (namely, EF030919 and EF030867 on Andromeda polifolia, EF030944 on Erica ciliaris, EF030936 on Rhododendron ferrugineum and EF030875 on Vaccinium vitis-idaea). d Values followed by different letters differ according to a chi-squared test (P < 0.001). e Given that some sequences were found on different plants (for Agauria salicifolia and Gaultheria poeppigii), this means a total of 70 different sequences. New Phytologist (2007) 174: 864–878 www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007) Research Table 2 A summary of sebacinoid sequences amplified from ectendomycorrhizal (EEM) Ericaceae using primers ITS3Seb and TW13 Species Site Arbutus unedob Fango forest, France (42°20′N; 8°49′E) Kukka bog, Estonia (58°14′N; 22°00′E) Crêt de l’Oeillon, France (45°24′N; 4°37′E) Chauriat forest, France (45°46′N; 3°17′E) Chauriat forest, France (45°46′N; 3°17′E) Abrahams Lake, Canada (45°09′N; 62°36′W) Arctoctaphylos uva-ursi Pyrola chlorantha Orthilia secunda No. of EEM samples No. of successful PCRs Sequences (GB accession numbers)a 7 7b EF030913 (n = 7) 70 0 – 35 0 – 100 1 EF030896c (n = 1) 100 4 49 1 EF030895c (n = 3) EF030946 (n = 1) EF030894 (n = 1) a With number (n) of samples producing the sequence. Only sebacinoid morphotypes after Richard et al. (2005) were harvested and sequenced for A. unedo. c One of the sequences obtained from O. secunda at the Chauriat forest (EF030895) was 100% identical to that retrieved from P. chlorantha (EF030896) at the same site. b as well as old and nonmycorrhizal roots. Then a 0.1–0.5 g (fresh weight) subsample, containing several hair roots, was frozen at −80°C, either directly or after storage in 60% ethanol in water (v/v) for < 5 d before transport. The samples from Argentina were oven-dried at 60°C for 72 h before transport, and stored in silicagel until DNA extraction. Samples from Calluna vulgaris and Erica cinerea from Lanno in Belle-Isle en Mer (France, one sampling of a cluster of co-occurring roots for each species) and Gaultheria poeppigii from Pampa de Achala (Argentina, two samplings of a root cluster; see Table S1 for locations) were prepared for transmission electron microscopy (TEM) in a fixating buffer (2% glutaraldehyde in 0.1 M phosphate buffer (0.2 M KH2PO4/Na2HPO4), pH = 7.2) at 4°C. For EEM roots, each sample was stored separately at −80°C. For Arbutus unedo, mycorrhizal root tips were harvested in the Fango forest (Corsica, France, Table 2). From two soil cores situated 10 m apart, four and three tips, respectively, were selected that fitted the sebacinoid morphotype described from the same site by Richard et al. (2005). This is a trichotomic mycorrhiza found in large aggregates, exclusively in the organic soil layer. These mycorrhizas are highly clavate, have no emanating hyphae or rhizomorphs, and are orange in colour and almost yellow at the tips. The seven selected tips were cut in two parts: one was placed in TEM fixating buffer and the other was used for DNA extraction. Four additional sequences from the Fango forest were obtained (GenBank accession numbers EF030880, EF030929, EF030911 and EF030912) by further sequencing A. unedo EEM DNA samples that produced sebacinoid internal transcribed spacer (ITS) sequences in a previous study (sebacinoid #1 to #4 in Richard et al., 2005). For the Pyroleae, EEM tips or root sections that showed external hyphae were recovered from a 6 m2 patch of Orthilia secunda and a 4 m2 patch of Pyrola chlorantha (100 EEM per species) at the Chauriat forest (Puy de Dôme, France, Table 2). A set of 49 mycorrhizal roots of Orthilia secunda were collected from a 4 m2 patch at Abrahams Lake (Nova Scotia, Canada, Table 2). For Arctostaphylos uva-ursi, 70 mycorrhizal roots were harvested from a 6 m2 patch at Kukka bog reserve (Hiiumaa island, Estonia, Table 2) and 35 other roots from several plants at Crêt de l’Oeillon (Monts du Pilat, France, Table 2). The last four species were not sampled for TEM analysis, as no description of sebacinoid morphotypes was available. DNA extraction, PCR and sequencing All roots were submitted to DNA extraction using the DNeasy Plant Mini Kit (Qiagen, Courtaboeuf, France), according to the manufacturer’s instructions, and DNA was recovered in 40 µl of distilled water. Quality of DNA extract and fungal colonization were tested by PCR amplification of the fungal ITS, using a set of primers universal for fungi (ITS1F + ITS4) as in Selosse et al. (2002a). They were then submitted to amplification of a fragment of partial ITS plus the 5′ part of the 28S rDNA (D1/D2 region), using the primers ITS3Seb (5′-TGAGTGTCATTGTAATCTCAC-3′, internal to ITS and specific for Sebacinales, kindly provided by M. Berbee) and TW13 (5′-GGTCCGTGTTTCAAGACG3′, universal for fungi and internal to 28S rDNA; White et al., 1990), using DNA from a sebacinoid fruitbody as a positive control. PCR was carried out in 50 µl, with final concentrations of 66 µM for each dNTP, 0.6 µM for each of the primers (Laboratoires Eurobio, Les Ulis, France), 10 mM Tris-HCl, 50 mM KCl, 1.5 mM MgCl2, 0.2 mg ml−1 gelatin, 0.1% (v/v) Triton X100, 5% (v/v) dimethyl sulphoxide and 1.5 units of Taq DNA polymerase (Quantum Appligène, Illkirch, France). We used 1 µl of the extracted DNA solution, but higher © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 867 868 Research volumes (up to 10 µl) or dilutes (up to 10×) were tested when amplification failed. Reactions were performed in a TRIOThermoblock (Biometra, Göttingen, Germany) under the following thermoprofile: initial denaturation at 94°C for 4 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 53°C for 30 s, and extension at 72°C for 30 s. After the 35th cycle, a terminal extension of 10 min at 72°C was carried out. PCR products were sequenced as in Selosse et al. (2002a). Five randomly chosen PCR products for which no direct sequencing was possible were submitted to cloning (Table S1), as in Julou et al. (2005), and at least seven clones were sequenced for each of them. Sequences were edited, aligned using Sequencher 4.5 from Genes Codes (Ann Arbor, MI, USA), and confirmed as sebacinoid sequences by BLAST analysis (Altschul et al., 1997) against GenBank (NCBI; www.ncbi.nlm.nih.gov). All sequences (including consensus of cloned sequences) were deposited in GenBank (EF030867 to EF030946). Transmission electron microscopy For ERM, only the narrowest roots (hair roots) were used for investigation. About 20 hair roots were selected per glutaraldehyde-fixed sample and processed as in Setaro et al. (2006b) to obtain semithin sections (1 µm) and, when appropriate, ultrathin sections (0.05 µm). For EEM tips, ultrathin sections were obtained directly, by the same procedure as in Selosse et al. (2002b). Phylogenetic reconstruction The nucLSU portions of the retrieved DNA sequences were aligned with representative reference sequences taken from GenBank using MAFFT, version 5.850 (Katoh et al., 2002). To estimate phylogenetic relationships, the alignment was analysed using heuristic maximum likelihood (ML) as implemented in the PHYML software, version 2.4.4 (Guindon & Gascuel, 2003), starting from a BIONJ tree (Gascuel, 1997), with a general time-reversible model of nucleotide substitution and additionally assuming a percentage of invariant sites and gamma-distributed substitution rates at the remaining sites (GTR + I + G). The gamma distribution was approximated with four discrete rate categories. All model parameters were estimated using ML. Branch support was inferred from 1000 replicates of nonparametric bootstrapping (Felsenstein, 1985), with model parameters estimated via ML individually for each bootstrapped alignment. Additionally, we performed a Bayesian Markov chain Monte Carlo (MCMC) analysis using MrBayes 3.1 (Ronquist & Huelsenbeck, 2003). We ran two independent MCMC analyses, each involving four incrementally heated chains over five million generations, using the GTR + I + G model of nucleotide substitution and starting from random trees. Model parameters were not fixed but sampled during MCMC. Trees were sampled every 100 generations, resulting in an overall sampling of 50 000 trees New Phytologist (2007) 174: 864–878 per run, from which the first 20 000 trees of each run were discarded (burn in). The remaining 30 000 trees sampled from each run were pooled and used to compute a majority rule consensus tree to get estimates for the posterior probabilities. Stationarity of the process was assessed using the Tracer software (Rambaut & Drummond, 2003). Results Sebacinoids from ERM samples In all, 239 ERM root samples were recovered from 36 Ericaceae species (Table 1), that is, from 50 sites worldwide (Table S1). ITS amplification using the universal fungal primers ITS1F and ITS4 produced multiple PCR fragments for all samples, demonstrating that several fungal species were present on these roots (data not shown). The sebacinoid-specific primer set ITS3Seb and TW13 produced either a single 900 bp fragment (in 49.8% of the samples, i.e. in 29 out of the 36 investigated species, Table 1), or no PCR product. Although variable percentages of successful PCR amplification were obtained, none of these differences was significant (using pairwise chi-squared tests; P > 0.05, not shown) among Vaccinioideae and Ericoideae, nor among the most extensively sampled species (V. myrtillus, C. vulgaris and E. cinerea), nor even among the most extensively sampled genera (Gaultheria, Vaccinium, Erica and Rhododendron). Only samples from Canada yielded significantly fewer PCR products than those from Europe and Argentina according to chi-squared tests (P < 0.001, Table 1). Sequences were obtained by direct sequencing of the 900 bp fragment from 71 samples (i.e. 59.6% of samples producing a PCR product – Table S1). BLAST analysis confirmed that they were closely related to sebacinoid sequences from GenBank. In some cases (e.g. EF030869, EF030888 or EF030899), two bases were present at some positions, suggesting that two close sequences were sequenced as a result of heterozygosities or the presence of two closely related sebacinoids. All other PCR products yielded a mix of different sebacinoid sequences, so that the conserved areas were the same but many polymorphic sites were detected in the variable regions (not shown). Five of these PCR products were cloned and sequenced (Table 1). In all cases, a unique sequence (with the exception of a few PCR and cloning errors, not shown) was found among clones, suggesting that at least one sebacinoid was present in these samples. (Perhaps other minor sequences were present that were not seen because only seven clones were sequenced). In all, sebacinoid sequences were obtained in 31.8% of the samples (Table 1) and no nonsebacinoid sequences were recovered. All sequences differed by at least some base pairs, even for plants growing on the same site: as the only exceptions, identical sequences were found in five co-occurring Gaultheria poeppigii samples from Argentina (EF030889) and three co-occurring Agauria salicifolia samples from La Réunion www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007) Research Fig. 1 Semithin sections of ericoid mycorrhiza of Gaultheria poeppiggi. The cortical cells are well colonized by fungal hyphae. Bar, 10 µm. (EF030898, Table S1). As a result, 70 different sequences were found. Up to five different sequences were found from the same site, that is from < 10 m2 (Table S1), and different sequences were recovered from the same host at the same site (e.g. Vaccinium myrtillus, V. uliginosum or Calluna vulgaris produced two different sebacinoid sequences at some sites; Table S1). Light and transmission electron microscopy of fungal colonization in ERMs Semithin sections of clusters of hair roots from C. vulgaris (n = 7), Erica cinerea (n = 2) and Gaultheria poeppigii (n = 8) displayed typical ERM colonization, with intracellular hyphal colonization (Fig. 1). TEM revealed both ascomycetes and hymenomycetous basidiomycetes as associated fungi (Fig. 2), with ultrastructure clearly distinguishing them. Ascomycetes had simple septal pores with Woronin bodies (Fig. 2b,c) and electron-light cell walls (Fig. 2b,c,e), whereas basidiomycetes had imperforate parenthesomes (i.e. the parenthesomal type also present in Sebacinales; Fig. 2d,e) and electron-dense hyphal cell walls (Fig. 2a,b,d,e). Sebacinales-like basidiomycetes were found on E. cinerea (in n = 1 ultrathin section) and G. poeppigii (n = 2). Both asco- and basidiomycetes formed intracellular coils in living cortical cells of hair roots. The vitality of the host cells was indicated by the mitochondriarich cytoplasm (Fig. 2a,b,f ). Basidiomycetous colonization of cortical cells was found to be exclusive (Fig. 2a) or dual, that is, with both Sebacinales-like basidiomycetes and ascomycetes forming coils within a single host cell (in G. poeppigii, Fig. 2b,e). sebacinoid sequence (EF030913, Table 2). It differed from the four sequences already obtained by Richard et al. (2005) at this site, suggesting that sebacinoids are diverse in this forest. Two Pyroleae (O. secunda and P. chlorantha) from the Chauriat forest produced the same sequence (EF030895 = EF030896, Table 2), so that the related sebacinoid was probably not host-specific. A second sebacinoid occurred in O. secunda at the Chauriat. A single EEM O. secunda root from Abrahams Lake harboured a sebacinoid sequence that differed from the previous ones (Table 2). No sequences were obtained from A. uva-ursi at the two sites investigated (Table 2). Because of the low frequency of sebacinoid colonization on other hosts, this does not imply that A. uva-ursi is not a host (pairwise chi-squared tests not significant (P > 0.05), not shown). Ultrastructure of fungal colonization in Arbutus unedo EEMs Transmission electron microscopy performed on three randomly selected EEM tips of Arbutus unedo (n = 3) from the Fango forest revealed typical arbutoid ectendomycorrhizas with hyphal sheath, Hartig net and intracellular colonization (Fig. 3a,b). Although vitality of the plant cell is not visible, these structures indicate an ectomycorrhizal association. Peripheral hyphae showed thick cell walls (Fig. 3d), as described for ectomycorrhizal sebacinoids (Selosse et al., 2002b). The associated fungi have dolipores and imperforate parenthesomes, as is typical for Sebacinales (Fig. 3b,c). Phylogeny of sebacinoids from Ericaceae Sebacinoids from EEM samples The seven EEM root tips of A. unedo from the Fango forest showing a sebacinoid morphotype harboured a unique The results of our molecular phylogenetic analysis involving the nucLSU sequences from the present study and a comprehensive set of reference sequences published in GenBank are shown in © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 869 870 Research Fig. 2 Transmission electron micrographs of ericoid mycorrhiza (ERM) colonized by sebacinoids. (a) Cortical cell of an Erica cinerea hair root colonized by sebacinoid hyphae (s). Black arrowhead, a dolipore of an intracellular hypha; white arrowheads, mitochondria of the plant cell. Bar, 2 µm. (b) Cortical cell of an ERM of Gaultheria poeppigii with dual colonization of ascomycetes (a) and Sebacinales (s). The plant cell cytoplasm is alive, indicated by the nucleus (NU) and mitochondria (white arrowheads); the black arrowhead points to an ascomycetous septum with Woronin bodies. Bar, 2 µm. (c) Detail of the ascomycetous hyphae with septum and Woronin body (arrowhead) of panel (b); M, mitochondria; *, the electron-light hyphal wall. Bar, 1 µm. (d) Dolipore of a sebacinoid hypha forming ERM with G. poeppigii. Arrowheads point to imperforate parenthesomes; *, the electron-dense hyphal wall. Bar, 0.1 µm. (e) Sebacinoid hyphae (s) with dolipore (arrowhead) and imperforate parenthesomes forming dual colonization with ascomycetes (a) in a cortical cell of G. poeppigii. White and black asterisks mark the electron-light cell walls of the ascomycetes and the electron-dense cell walls of the sebacinoids, respectively. Bar, 2 µm. (f) Detail of mitochondria (M) in the cytoplasm of the plant cell dually colonized by Sebacinales and ascomycetes. Bar, 0.1 µm. New Phytologist (2007) 174: 864–878 www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007) Research Fig. 3 Transmission electron micrographs (TEMs) of ectendomycorrhiza (EEM) from Arbutus unedo and a sebacinoid (corresponding to sequence EF030880). (a) Mycorrhiza with formations of hyphal sheath (hs) and Hartig net (arrowheads) as well as hyphal growth within cortical cells (CC+). CC–, cortical cells without hyphal colonization. Bar, 4 µm. (b) Cortical cell (CC) with intracellular hyphae (ih) and Hartig net (Hn) by a sebacinoid; arrowhead, a dolipore. Bar, 1 µm. (c) Detail of dolipore with imperforate parenthesomes (arrowheads) of a sebacinoid hypha. (d) External sheath hyphae with thick walls (on the left; inner mantle on the right). Bar, 1 µm. (a–c, pictures from Robert Bauer; d, picture from Houria Horcine). Fig. 4. The outcome of the heuristic ML analysis and the two independent MCMC runs was widely consistent, with few exceptions, the most prominent of which is the different placement of a sebacinoid detected in Vaccinium vitis-idaea (EF030882). In the ML tree, this sequence clustered, although without significant support, together with a similar sequence from Vaccinium vitis-idaea (EF030883) at the base of Sebacinales group A (Fig. 4). In MCMC analysis, however, the sequence EF030882 was placed within Sebacinales clade B with a posterior probability of 100%, while EF030883 was still placed at the base of clade A with a posterior probability of 97% (not shown). Apart from these exceptions, both ML and MCMC analyses separated the Sebacinales into the known clades A and B. Consistent with earlier studies, group A also includes all sequences retrieved from fruitbodies, from ectomycorrhizal samples and from heterotrophic orchids (Fig. 4). Sebacinoid sequences from EEMs (from Arbutus unedo, Orthilia secunda, Pyrola chlorantha) also appear in this group. All the sequences obtained from ERMs, on the other hand, clustered within Sebacinales clade B, the clade that also includes all sequences from cavendishioid mycorrhizas and liverwort thalli (Fig. 4). Clade B also contains the sequences of Sebacina vermifera isolates obtained mostly from Australian green, autotrophic orchids (Warcup, 1988), as well as of Piriformospora indica. Within clades A and B, the sequences were not strictly grouped according to the mycorrhizal types from which they were obtained. Furthermore, no larger biogeographical subgroup was resolved. For example, a well-supported subgroup (97% in ML bootstrap, 100% MCMC support) contains sequences from Australia, Argentina, Ecuador, Canada, France, Germany, and Estonia. © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 871 872 Research Discussion Sebacinoids as common mycorrhizal fungi in Ericaceae roots Out of 600 samples of Ericaceae roots, a total of 89 produced a sebacinoid sequence (Tables 1, 2), representing 74 different sequences (because of some identities). Among them, five were obtained after cloning, suggesting that the 48 PCR products from ERM roots that we were not able to sequence directly contain sebacinoid sequences (Table S1). Crosscontamination of samples seems unlikely, as most sequences are different. Moreover, we provide evidence that hyphae with sebacinoid dolipores and parenthesomes can be detected by TEM analysis and that they occur in (or around) living cells, forming typical ERM (Fig. 2) or EEM (Fig. 3) on the roots of at least three host species. This underlines the utility of combining molecular and TEM approaches, if possible on the same sample, as already used to demonstrate that sebacinoids form ectomycorrhizas (Selosse et al., 2002b) and cavendishioid mycorrhizas (Setaro et al., 2006a). In this study, the use of specific primers probably enhanced molecular detection. Sebacinales form ERM-like interactions in hair roots of Ericaceae, whereas only ascomycetes were reported to be involved so far (Smith & Read, 1997). Although more species should be investigated at the ultrastructural level, this is in accordance with the molecular diversity of Sebacinales previously found on hair roots of Vaccinioideae and Styphelioideae species from Canada (Gaultheria shallon, Berch et al., 2002; Allen et al., 2003), Australia (Epacris pulchella, Bougoure & Cairney, 2005), and the Neotropics (Cavendishia nobilis, Setaro et al., 2006a). It extends this diversity to the Ericoideae tribe. With the exception of C. nobilis (Setaro et al., 2006a), and probably Calluna vulgaris (Bonfante-Fasolo, 1980), we are not aware of direct observation of sebacinoids in ERM. Some species investigated in the present study failed to produce positive PCRs (Table 1). This could be the result of undersampling, as Sebacinales abundance might vary among species. On G. shallon they represented more than half of cloned ITS sequences (Allen et al., 2003; accordingly 75% of roots did not yield fungal isolates) compared with less than a quarter on E. pulchella (Bougoure & Cairney, 2005; 8% of roots did not yield fungal isolation). This suggested that Sebacinales colonized larger root portions on G. shallon than on E. pulchella. However, these percentages are not directly comparable, since isolation procedures were not identical. Can one say that Sebacinales are ERM mycorrhizal? If mycorrhizal interaction is defined as a morphogenetic process uniting roots and soil fungi, a definition we favour and use here, then we observed ERM mycorrhizas involving sebacinoids. If the definition additionally implies a mutualistic relationship, a feature sometimes violated in associations considered as mycorrhizal (Kiers & van der Heijden, 2006) and difficult to establish, then the question requires further inves- New Phytologist (2007) 174: 864–878 tigation. Mutualism has been experimentally shown for relatives of the ERM sebacinoids in clade B. Piriformospora indica is beneficial for the growth of several plants (Rai et al., 2001; Varma et al., 2001; PeSkan-Berghöfer et al., 2004; Waller et al., 2005) and even induces systemic resistance to fungal diseases and tolerance to salt stress in barley (Waller et al., 2005). Together with another clade B sebacinoid, it enhanced growth of Nicotiana tabacum, but at the expense of herbivore resistance (Barazani et al., 2005). Similar positive effects on plant growth have also been shown for several isolates of the Sebacina vermifera species complex (Deshmukh et al., 2006). Such beneficial effects might thus be expected in the interaction between group B sebacinoids and Ericaceae, but await further studies, for example using in vitro synthetic associations. Evolution of mycorrhizal structures in the Sebacinales Mycorrhizal types are not evenly distributed over the phylogeny of Sebacinales (Fig. 4), in agreement with previous investigations (Weiß et al., 2004). Clade B encompasses species that grow intracellularly, in hepatics (Kottke et al., 2003), ERM roots (our data), orchid roots (Warcup, 1988), and many other hosts for P. indica (the model species Arabidopsis thaliana (PeSkan-Berghöfer et al., 2004) and Nicotiana tabacum (Barazani et al., 2005), but also species in Fabaceae and Rhamnaceae (Varma et al., 2001), Asteraceae and Solanaceae (Rai et al., 2001) as well as Poaceae (Waller et al., 2005)). However, the interaction is somewhat different at the cellular level for P. indica, which was shown to enhance apoptosis of plant cells, and colonize dead cells instead of living cells as observed here (Deshmukh et al., 2006). Several types of interaction can therefore be expected at the structural level in clade B. Some well-supported terminal clades within group B do not at present contain sequences from different host types (Fig. 4), but this might well be the result of an undersampling of Sebacinales diversity. Within clade B, species involved in cavendishioid mycorrhizas (Setaro et al., 2006a) form abundant extracellular mycelium, in addition to the usual intracellular colonization. This results in an EEM-like association that probably evolved secondarily among ERM plants (Setaro et al., 2006b). By contrast, clade A mycorrhizal association usually presents a hyphal sheath (Selosse et al., 2002b; Urban et al., 2003), and sometimes intracellular colonization, as in some ectomycorrhizas (Selosse et al., 2002b) and in some heterotrophic orchids (Selosse et al., 2002a; Taylor et al., 2003). All EEM sebacinoids from Ericaceae belong to clade A, as observed in independent studies of Monotropoideae mycorrhizal associates (Pyrola chlorantha and Orthilia secunda, Tedersoo et al., 2007; P. rotundifolia, M.-A. Selosse, unpublished; discussed later). This corroborates the observation that EEM fungi usually belong to clades with ectomycorrhizal abilities, in both Arbutoideae (Richard et al., 2005) and Monotropoideae (Bidartondo, 2005; Tedersoo et al., 2007). www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007) Research Fig. 4 Phylogenetic relationships within the Sebacinales, with part of the tree relating to clade A and clade B. Phylogram derived by heuristic maximum-likelihood (ML) analysis from an alignment of nucLSU sequences, using a time-reversible model of nucleotide substitution, additionally assuming a portion of invariable sites and gamma-distributed substitution rates at the remaining sites (GTR + I + G). Branch support values were calculated from 1000 replicates of nonparametric ML bootstrap analysis (first numbers), and from Bayesian Markov chain Monte Carlo analysis, also using the GTR + I + G substitution model (second numbers). Values below 50% are indicated with asterisks or omitted. Branch lengths are scaled in terms of expected numbers of nucleotide substitutions per site. The tree was rooted with Auricularia auricula-judae. Mycorrhizal types: ECM, ectomycorrhiza; CVM, cavendishioid mycorrhiza; EEM, ectendomycorrhiza; ERM, ericoid mycorrhiza; JMM, jungermannialean mycorrhiza-like thalli; ORM, orchid mycorrhiza. Sample origin: A, Austria; AUS, Australia; ARG, Argentina; CAN; Canada; CHL, Chile; CHN, P.R. China; ECU, Ecuador; EST, Estonia; FRA, France; GER, Germany; GUY, Guyana; IND, India; NOR, Norway; REU, la Réunion; SPA, Spain. © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 873 874 Research Fig. 4 continued New Phytologist (2007) 174: 864–878 www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007) Research Fig. 5 Phylogeny of the Ericaceae subfamilies following Freudenstein (1999) and Kron et al. (2002a), with most parsimonious scenario for evolution of the mycorrhizal types. AM, the ancestral arbuscular mycorrhizal condition; CVM, some species form cavendishioid mycorrhizas. Note that emergence of ericoid mycorrhizas (ERMs) corresponds to the rise of the ‘early anther inversion clade’ (Kron et al., 2002a). EEM, ectomycorrhiza. Clades A and B therefore differ in mycorrhizal associations. The occurrence of ERM sequences basal to group A in Fig. 4 is not strongly supported and might have two alternative explanations: either future works will support this (in which case features specific to clade A are derived in Sebacinales evolution, such as mycorrhizal associations with sheath), or this is an artifactual grouping. Indeed, these sequences appeared in various positions within group B in preliminary analyses, and there is a discrepancy concerning this pair of sequences between ML and MCMC analysis (as detailed in the Results section). For the moment, we regard the placement of these sequences in the phylogenetic tree as unstable. Addition of more sequences might resolve this problem in future analyses. Evolution of mycorrhizal structures in the Ericaceae Available Ericaceae phylogenies (Freudenstein, 1999; Kron et al., 2002a) suggest a single most parsimonious scenario for evolution of mycorrhizal association (Fig. 5). Starting from a plesiomorphic association with arbuscular mycorrhizal fungi, still observed in Enkianthoideae (Abe, 2005), a shift occurred to EEM association involving ectomycorrhizal fungi, as currently retained in Monotropoideae + Arbutoideae. Then, ERM association arose once in the Ericoideae + Styphelioideae + Harrimanelloideae + Cassiopoideae + Vaccinioideae clade (Fig. 5; Cullings, 1996). Among these, more derived conditions evolved, such as the cavendishioid mycorrhizas in the Andean clade of Vaccinioideae (Setaro et al., 2006a,b), or perhaps the additional presence of arbuscular mycorrhizal fungi in Gaultheria poeppigii (Urcelay, 2002), in Vaccinium spp. and Styphelia tameiameiae from Hawaii (Koske et al., 1990). Significantly, although data on the smaller subfamilies Harrimanelloideae and Cassiopoideae are lacking so far, the EEM to ERM shift was associated with a shift from clade A to clade B sebacinoids (Fig. 5). Interestingly, two sequences of clade B sebacinoids were found in cloning of fungal ITS in the Ericaceae Pyrola chlorantha, in addition to EEM mycorrhizal fungi (Tedersoo et al., in press): cloning allows recovery of various endophytic fungi, and this may mean that clade B sebacinoids even occur in EEM Ericaceae as endophytes, together with the main mycorrhizal fungus. Strikingly, Helotiales, the ERM ascomycetous associates of Ericaceae, also encompasses root endophytic species (Vrålstad et al., 2002), raising an intriguing hypothesis. At the emergence of ERM association in Ericaceae, some endophytic fungal clades (namely the Helotiales and the Sebacinales) would have been recruited as mycorrhizal partners and/or excluded the former EEM symbionts in the ancestor of ERM Ericaceae. We did not detect any significant difference in frequency of colonization, or specialization for any sebacinoid subclade in Ericoideae, Vaccinioideae, or in any frequently sampled plant species (Table 1). Mycobionts from the same ericaceous species are distributed over group B in several cases (Fig. 4). However, Sebacinales obtained from some host species from different localities are very similar, if not identical (e.g. Vaccinium myrtillus mycobionts from France and Germany), and there is a cluster (near the top of the clade B tree) uniting mycobionts from several Vaccinium species. Our sampling design and effort are nevertheless not suitable for reliable conclusions to be drawn regarding specificity. Perspectives on Sebacinales Our data add to the amazing diversity and ecological abundance of sebacinoids. So far, no biogeographical subgroup is detectable in Sebacinales. One well-supported subgroup (97%, with a © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 875 876 Research long branch; Fig. 4) contains Sebacinales from Europe, South America and Australia. Some well-supported terminal clades at present do not include sequences from different continents: however, ensuring that they represent regional clades would require much more additional sampling. Finding sebacinoids among ascomycetes in ERM roots (sometimes in the same cell, Fig. 2b) again underlines their frequent coexistence with ascomycetes (Warcup, 1988; Selosse et al., 2002a; Urban et al., 2003; Setaro et al., in press), an intriguing and hitherto unexplained feature. Interestingly, in > 60% of the ERM samples successfully amplified, a single or, at most, two nearly identical sequences (when two bases are present at certain positions) were obtained. In Andean Ericaceae investigated by Setaro et al. (2006), different sebacinoids co-occurred in a given sampled root. The sequence homogeneity observed on our samples encompassing many roots from a given root system, on is of unclear significance, since the meaning of the differences between ITS + 28S rDNA sequences is unknown. If these differences represent interspecific variations, we observe here patchiness in species distribution that often leads to the colonization of a root system by a single species (including possibly several individuals), and a huge specific diversity exists among Sebacinales. If differences between sequences represent intraspecific polymorphism, then a single individual (or several closely related ones) often covers each root system. Although there is evidence from other basidiomycetes that the variability of 28S rDNA sequences is low at the intraspecific level (Fell et al., 2000), no rigorous conclusion can be drawn. This frustrating problem arises for sebacinoids because the absence of meaningful morphological characters makes distinction of morphological species impossible. Similarly, the absence of haploid cultures for most ERM and EEM species does not allow distinction of biological species using interfertility tests. If other loci were available, the phylogenetic species concept (Taylor et al., 2000) would be interesting to apply: the fingerprint of recombination in a taxon is that phylogenies of individuals of the same biological species are not congruent for different loci. Conversely, in a comparison between phylogenetic trees of different loci, the level at which congruence between loci arises delineates biological species (Taylor et al., 2000). This potentially applies to taxa known only from environmental DNA, if at least two loci are available. For sebacinoids, specific primers should be designed, since environmental DNA extracts harbour DNA from many fungal species. This is a common problem with mycorrhizal taxa mostly reported from molecular studies, such as tullasnelloids (Bidartondo et al., 2004) or thelephoroids (Koljalg et al., 2001). More diversity and biological interactions may still be found in Sebacinales. They may colonize many other hosts, as suggested by their isolation from roots that were first supposed to be arbuscular mycorrhizal (Williams, 1985; Milligan & Williams, 1987). In particular, clade B sebacinoids may New Phytologist (2007) 174: 864–878 grow in roots of many plant taxa: some were recovered from Phragmites australis (Neubert et al., 2006) and hepatics (Kottke et al., 2003), and P. indica showed compatibility with tremendously diverse hosts in glasshouse experiments, as mentioned earlier. Additionally, sebacinoid mycelia may be shared by different root systems. Thus, we might expect a major role in interplant interactions, or even in carbon transfers between plants, as already demonstrated for clade A sebacinoids associated with achlorophyllous orchids (McKendrick et al., 2002; Selosse et al., 2002a,b; Taylor et al., 2003). A role for sebacinoids in shaping plant communities is an intriguing possibility raised by their frequent detection in molecular microbial ecology. Acknowledgements The authors wish to thank Sigisfredo Garnica, Thomas Julou, Anne-Marie Mollet, Carine Saison and Claude Selosse for help in root sampling, as well as Robert Bauer and Houria Horcine for TEM analysis of Fig. 3. We also thank two anonymous referees for helpful comments. Marc-André Selosse is funded by the Centre National de la Recherche Scientifique and the Société Française d’Orchidophilie. References Abe J. 2005. An arbuscular mycorrhizal genus in the Ericaceae. Abstracts of the MSA Symposium 6. Allen WK, Allaway WG, Cox GC, Valder PG. 1989. Ultrastructure of mycorrhizas of Dracophyllum secundum R. Br. (Ericales, Epacridaceae). Australian Journal of Plant Physiology 16: 147–535. Allen TR, Millar T, Berch SM, Berbee ML. 2003. Culturing and direct DNA extraction find different fungi from the same ericoid mycorrhizal roots. New Phytologist 160: 255–272. Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W, Lipman DJ. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acid Research 25: 3389–3402. Avis PG, McLaughlin DJ, Dentinger BC, Reich PB. 2003. Long-term increase in nitrogen supply alters above- and below-ground ectomycorrhizal communities and increases the dominance of Russula spp. in a temperate oak savanna. New Phytologist 160: 239–253. Barazani O, Benderoth M, Groten K, Kuhlemeier C, Baldwin IT. 2005. Piriformospora indica and Sebacina vermifera increase growth performance at the expense of herbivore resistance in Nicotiana attenuata. Oecologia 146: 234–243. Berch SM, Allen TR, Berbee ML. 2002. Molecular detection, community structure and phylogeny of ericoid mycorrhizal fungi. Plant and Soil 244: 55 –66. Bergero R, Perotto S, Girlanda M, Vidano G, Luppi AM. 2000. Ericoid mycorrhizal fungi are common root associates of Mediterranean ectomycorrhizal plants (Quercus ilex). Molecular Ecology 9: 1639–1649. Bidartondo MI. 2005. The evolutionary ecology of myco-heterotrophy. New Phytologist 167: 335–352. Bidartondo MI, Burghardt B, Gebauer G, Bruns TD, Read DJ. 2004. Changing partners in the dark: isotopic and molecular evidence of ectomycorrhizal liaisons between forest orchids and trees. Proceedings of the Royal Society London Series B 271: 1799–1806. Bonfante-Fasolo P. 1980. Occurrence of a basidiomycete in living cells of mycorrhizal hair roots of Calluna vulgaris. Transactions of the British Mycological Society 75: 320–325. www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007) Research Bougoure DS, Cairney JWG. 2005. Assemblages of ericoid mycorrhizal and other root-associated fungi from Epacris pulchella. Environmental Microbiology 7: 819– 827. Bougoure JJ, Bougoure DS, Cairney JW, Dearnaley JD. 2005. ITS-RFLP and sequence analysis of endophytes from Acianthus, Caladenia and Pterostylis (Orchidaceae) in southeastern Queensland. Mycological Research 109: 452–460. Cairney JWG, Ashford AE. 2002. Biology of mycorrhizal associations of epacrids (Ericaceae). New Phytologist 154: 305–326. Cullings KW. 1996. Single phylogenetic origin of the world’s heath plants indicated by 28S ribosomal RNA gene sequences and the ericoid mycorrhizal symbiosis. Canadian Journal of Botany 74: 1896–1909. Deshmukh S, Hückelhoven R, Schäfer P, Imani J, Sharma M, Weiß M, Waller F, Kogel KH. 2006. The root endophytic fungus Piriformospora indica requires host cell death for proliferation during mutualistic symbiosis with barley. Proceedings of the National Academy of Sciences, USA 103: 18450 –18457. Fell JW, Boekhout T, Fonseca A, Scorzetti G, Statzell-Tallman A. 2000. Biodiversity and systematics of basidiomycetous yeasts as determined by large-subunit rDNA D1/D2 domain sequence analysis. International Journal of Systematic and Evolutionary Microbiology 50: 1351–1371. Felsenstein J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791. Freudenstein JV. 1999. Relationships and character transformation in Pyroloideae (Ericaceae) based on ITS sequences, morphology, and development. Systematic Botany 24: 398– 408. Gascuel O. 1997. BIONJ: An improved version of the NJ algorithm based on a simple model of sequence data. Molecular Biology and Evolution 14: 685–695. Giovannetti M, Lioi L. 1990. The mycorrhizal status of Arbutus unedo in relation to compatible and incompatible fungi. Canadian Journal of Botany 68: 1239–1244. Glen M, Tommerup IC, Bougher NL, O’Brien PA. 2002. Are Sebacinaceae common and widespread ectomycorrhizal associates of Eucalyptus species in Australian forests? Mycorrhiza 12: 243–247. Guindon S, Gascuel O. 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology 52: 696–704. Julou T, Burghardt B, Gebauer G, Berveiller D, Damesin C, Selosse M-A. 2005. Evolution of mixotrophy in orchids: insight from a comparative study of green and achlorophyllous Cephalanthera damasonium. New Phytologist 166: 639– 653. Katoh K, Misawa K, Kuma K, Miyata T. 2002. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transformation. Nucleic Acids Research 30: 3059–3066. Kennedy PG, Izzo AD, Bruns TD. 2003. There is high potential for the formation of common mycorrhizal networks between understorey and canopy trees in a mixed evergreen forest. Journal of Ecology 91: 1071–1080. Kiers ET, van der Heijden GA. 2006. Mutualistic stability in the arbuscular mycorrhizal symbiosis: exploring hypotheses of evolutionary cooperation. Ecology 87: 1627–1636. Koljalg U, Dahlberg A, Taylor A, Larsson E, Hallenberg N, Stenlid J, Larsson KH, Fransson P, Karen O, Jonsson L. 2001. Diversity and abundance of tomentelloid fungi in Swedish boreal forests. Molecular Ecology 9: 1985–1996. Koske RE, Gemma JN, Englander L. 1990. Vesicular-arbuscular mycorrhizae in Hawaiian Ericales. American Journal of Botany 77: 64 –68. Kottke I, Beiter A, Weiß M, Haug I, Oberwinkler F, Nebel M. 2003. Heterobasidiomycetes form symbiotic associations with hepatics: Jungermanniales have sebacinoid mycobionts while Aneura pinguis (Metzgeriales) is associated with a Tulasnella species. Mycological Research 107: 957–968. Kron KA, Judd WS, Stevens PF, Crayn DM, Anderberg AA, Gadek PA, Quinn CJ, Luteyn JL. 2002a. Phylogenic classification of Ericaceae: molecular and morphological evidence. Botanical Review 68: 335–423. Kron KA, Powell EA, Luteyn JL. 2002b. Phylogenetic relationships within the blueberry tribe (Vaccinieae, Ericaceae) based on sequence data from matK and nuclear ribosomal ITS regions, with comments on the placement of Satyria. American Journal of Botany 89: 327–336. Massicotte HB, Melville LH, Peterson RL. 2005. Structural characteristics of root–fungal interactions for five ericaceous species in eastern Canada. Canadian Journal of Botany 83: 1057–1064. McKendrick SL, Leake JR, Taylor DL, Read DJ. 2002. Symbiotic germination and development of the myco-heterotrophic orchid Neottia nidus-avis in nature and its requirement for locally distributed Sebacina spp. New Phytologist 154: 233–247. McLean CB, Cunnington JH, Lawrie AC. 1999. Molecular diversity within and between ericoid endophytes from the Ericaceae and Epacridaceae. New Phytologist 144: 351–358. Midgley DJ, Chambers SM, Cairney JWG. 2004. Utilisation of carbon substrates by multiple genotypes of ericoid mycorrhizal fungal endphytes from eastern Australian Ericaceae. Mycorrhiza 14: 245–251. Milligan MJ, Williams PG. 1987. Orchidaceous rhizoctonias from roots of nonorchids: mycelial and cultural characteristics of field and pot culture isolates. Canadian Journal of Botany 65: 598–606. Moyersoen B. 2006. Pakaraimaea dipterocarpacea is ectomycorrhizal, indicating an ancient Gondwanaland origin for the ectomycorrhizal habit in Dipterocarpaceae New Phytologist 172: 753–762. Münzenberger B, Kottke I, Oberwinkler F. 1992. Ultrastructural investigations of Arbutus unedo – Laccaria amethystea mycorrhiza synthesized in vitro. Trees 7: 40 –47. Neubert K, Mendgen K, Brinkmann H, Wirsel SGR. 2006. Only a few fungal species dominate highly diverse mycofloras associated with the common reed. Applied and Environmental Microbiology 72: 1118 –1128. Perotto S, Actis-Perino E, Perugini J, Bonfante P. 1996. Molecular diversity of fungi from ericoid mycorrhizal roots. Molecular Ecology 5: 123–131. Peskan-Berghöfer T, Shahollari B, Giong PH, Hehl S, Markert C, Blanke V, Kost G, Varma A, Oelmüller R. 2004. Association of Piriformospora indica with Arabidopsis thaliana roots represents a novel system to study beneficial plant–microbe interactions and involves early plant protein modifications in the endoplasmic reticulum and at the plasma membrane. Physiologia Plantarum 122: 465–477. Peterson TA, Mueller WC, Englander L. 1980. Anatomy and ultrastructure of a Rhododendron root–fungus association. Canadian Journal of Botany 58: 2421–3325. Rai M, Acharya D, Singh A, Varma A. 2001. Positive growth responses of the medicinal plants Spilanthes calva and Withania somnifera to inoculation by Piriformospora indica in a field trial. Mycorrhiza 11: 123–128. Rambaut A, Drummond A. 2003. Tracer MCMC trace analysis tool, version 1.2.1. Oxford, UK: University of Oxford. [http://evolve.zoo.ox.ac.uk/software.html.] Richard F, Millot S, Gardes M, Selosse MA. 2005. Diversity and structuration by hosts of the below-ground mycorrhizal community in an old-growth Mediterranean forest dominated by Quercus ilex L. New Phytologist 166: 1011–1023. Robertson DC, Robertson JA. 1985. Ultrastructural aspects of Pyrola mycorrhizae. Canadian Journal of Botany 63: 1089–1098. Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. Selosse MA, Bauer R, Moyersoen B. 2002b. Basal hymenomycetes belonging to the Sebacinaceae are ectomycorrhizal on temperate deciduous trees in silva: microscopic and molecular evidence. New Phytologist 155: 183–195. Selosse MA, Faccio A, Scappaticci G, Bonfante P. 2004. Chlorophyllous and achlorophyllous specimens of Epipactis microphylla (Neottieae, Orchidaceae) are associated with ectomycorrhizal septomycetes, including truffles. Microbial Ecology 47: 416–426. © The Authors (2007). Journal compilation © New Phytologist (2007) www.newphytologist.org New Phytologist (2007) 174: 864–878 877 878 Research Selosse MA, Weiß M, Jany JL, Tillier A. 2002a. Communities and populations of sebacinoid basidiomycetes associated with the achlorophyllous orchid Neottia nidus-avis (L.) L.C.M. Rich. & neighbouring tree ectomycorrhizae. Molecular Ecology 11: 1831–1844. Setaro S, Kottke I, Oberwinkler F. 2006b. Anatomy and ultrastructure of mycorrhizal associations of neotropical Ericaceae. Mycological Progress 5: 243–254. Setaro S, Weiß M, Oberwinkler F, Kottke I. 2006a. Sebacinales form ectendomycorrhizas with Cavendishia nobilis, a member of the Andean clade of Ericaceae, in the mountain rain forest of southern Ecuador. New Phytologist 169: 355–365. Smith SE, Read DJ. 1997. Mycorrhizal symbiosis, 2nd edn. London, UK: Academic Press. Taylor DL, Bruns TD, Szaro TS, Hodges SA. 2003. Divergence in mycorrhizal specialization within Hexalectris spicata (Orchidaceae), a non-photosynthetic desert orchid. American Journal of Botany 90: 1168 –1179. Taylor JW, Jacobson DJ, Kroken S. 2000. Phylogenetic species recognition and species concepts in fungi. Fungal Genetics and Biology 31: 21–32. Tedersoo L, Pellet P, Kõljalg U, Selosse MA. 2007. Parallel evolutionary paths to mycoheterotrophy in understorey Ericaceae and Orchidaceae: ecological evidence for mixotrophy in Pyroleae. Oecologia 151: 206–217. Tedersoo L, Suvi T, Larsson E, Kõljalg U. 2006. Diversity and community structure of ectomycorrhizal fungi an a wooded meadow. Mycological Research 110: 734–748. Urban A, Weiß M, Bauer R. 2003. Ectomycorrhizae involving sebacinoid mycobionts. Mycological Research 107: 3–14. Urcelay C. 2002. Co-occurrence of three fungal root symbionts in Gaultheria poeppigii DC in Central Argentina. Mycorrhiza 12: 89 –92. Varma A, Singh A, Sudha S, Sharma J, Roy A, Kumari M, Rana D, Thakran S, Deka D, Bharti K, Hurek T, Blechert O, Rexer KH, Kost G, Hahn A, Maier W, Walter M, Strack D, Kranner I. 2001. Piriformospora indica – an axenically culturable mycorrhiza-like endosymbiotic fungus. In: Hock B, ed. Mycota IX. Berlin, Germany: Springer, 123–150. Verma S, Varma A, Rexer KH, Hassel A, Kost G, Sarbhoy A, Bisen P, Bütehorn B, Franken P. 1998. Piriformospora indica, gen. et sp. nov., a new root-colonizing fungus. Mycologia 90: 896–903. Vrålstad T, Myhre E, Schumacher T. 2002. Molecular diversity and phylogenetic affinities of symbiotic root-associated ascomycetes of the Helotiales in burnt and metal polluted habitats. New Phytologist 155: 131–148. Walker JF, Miller OK, Horton JL. 2004. Hyperdiversity of ectomycorrhizal fungus assemblages on oak seedlings in mixed forests in the southern Appalachian Mountains. Molecular Ecology 14: 829–838. Walker JF, Parrent JL. 2004. Molecular phylogenetic evidence for the mycorrhizal status of Tremellodendron (Sebacinaceae). Memoirs of the New York Botanical Garden 89: 291–296. New Phytologist (2007) 174: 864–878 Waller F, Achatz B, Baltruschat H, Fodor J, Becker K, Fischer M, Heier T, Hückelhoven R, Neumann C, von Wettstein D, Franken P, Kogel KH. 2005. The endophytic fungus Piriformospora indica reprograms barley to salt-stress tolerance, disease resistance, and higher yield. Proceedings of the National Academy of Sciences, USA 102: 13386–13391. Warcup JH. 1988. Mycorrhizal associations of isolates of Sebacina vermifera. New Phytologist 110: 227–231. Weiß M, Oberwinkler F. 2001. Phylogenetic relationships in Auriculariales and related groups – hypotheses derived from nuclear ribosomal DNA sequences. Mycological Research 105: 403–415. Weiß M, Selosse M-A, Rexer K-H, Urban A, Oberwinkler F. 2004. Sebacinales: a hitherto overlooked cosm of heterobasidiomycetes with a broad mycorrhizal potential. Mycological Research 108: 1003– 1010. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand H, Sninsky JS, White TJ, eds. PCR protocols: a guide to methods and applications. New York, NY, USA: Academic Press, 315–322. Williams PG. 1985. Orchidaceous rhizoctonias in pot cultures of vesicular-arbuscular mycorrhizal fungi. Canadian Journal of Botany 63: 1329 –1333. Supplementary Material The following supplementary material is available for this article online: Table S1 List of the 239 samples of ericoid mycorrhizal (ERM) roots investigated in this work, in alphabetic order of Ericaceae host name, with geographic origin and GenBank accession numbers of the recovered sequences of the ITS +28S rDNA This material is available as part of the online article from: http://www.blackwell-synergy.com/doi/abs/10.1111/ j.1469-8137.2007.02064.x (This link will take you to the article abstract). Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to New Phytologist Central Office. www.newphytologist.org © The Authors (2007). Journal compilation © New Phytologist (2007)