Development of oligonucleotide microarrays for simultaneous multi‐species identification of P hellinus tree‐pathogenic fungi

Summary Polyporoid P hellinus fungi are ubiquitously present in the environment and play an important role in shaping forest ecology. Several species of P hellinus are notorious pathogens that can affect a broad variety of tree species in forest, plantation, orchard and urban habitats; however, current detection methods are overly complex and lack the sensitivity required to identify these pathogens at the species level in a timely fashion for effective infestation control. Here, we describe eight oligonucleotide microarray platforms for the simultaneous and specific detection of 17 important P hellinus species, using probes generated from the internal transcribed spacer regions unique to each species. The sensitivity, robustness and efficiency of this P hellinus microarray system was subsequently confirmed against template DNA from two key P hellinus species, as well as field samples collected from tree roots, trunks and surrounding soil. This system can provide early, specific and convenient detection of P hellinus species for forestry, arboriculture and quarantine inspection, and could potentially help to mitigate the environmental and economic impact of P hellinus‐related diseases.


Introduction
The Phellinus genus sensu lato currently comprises 154 species (Larsen & Cobb-Poulle, 1990;Kirk, 2014) of resupinate, sessile, polyporoid fungi, several of which are known to cause diseases such as stem rot, butt rot, root rot or tree wilt in a wide range of tree species (Van der Kamp, 1991;Castello et al., 1995). These tree-pathogenic fungi include some of the most aggressive wooddecay species identified thus far, infestations of which can devastate forest ecosystems, impact economic viability and render urban environments more vulnerable to tree hazards (Hansen and Goheen, 2000;Ann et al., 2002;Burdon et al., 2006). To date, the identification of diseased trees has primarily relied on visual inspection for signs and symptoms, pathogen isolation and characterization on selective media and other biochemical and immunological techniques (Nobles, 1965;Anselmi and Bragaloni, 1992;Jellison and Jasalavich, 2000;Clausen, 2003). However, the diagnostic process is typically laborious, time consuming and heavily reliant on experienced etiologists; moreover, the low sensitivity of such methods makes it unlikely that infestations can be detected and controlled during the relatively manageable early stages of disease (Thorn et al., 1996;Adair et al., 2002;McCartney et al., 2003;Luisi and Campanile, 2004). In addition, current methods are limited in their ability to identify the causative agent of Phellinus-related diseases at the species level, the knowledge of which is necessary for deploying appropriate control measures (Nam et al., 2002) and for ascertaining whether the infestation is native or exotic in nature (Hansen and Goheen, 2000). Phellinus species differentiation has traditionally relied on the morphological examination of fruiting bodies, spores and basidiocarps, but these may not appear until long after infection, by which time it is often too late to save the diseased tree (Nam et al., 2002). Therefore, the development of accurate, fast and specific diagnostic tools that can be easily used by personnel with a minimum of training is essential for the prevention and practical management of Phellinus infestations.
Recent advances in molecular biology offer the possibility of alternative approaches that can efficiently identify Phellinus pathogens, including nucleic acid-based techniques such as dot-blot hybridization, restriction fragment length polymorphism analysis, single-strand conformation polymorphism analysis and polymerase chain reaction (PCR) assays (Olive and Bean, 1999;Borneman and Hartin, 2000;Tsui et al., 2011). Polymerase chain reaction-based methods have been used to identify Phellinus s.l. at a generic rank (Guglielmo et al., 2007;2008), but a single assay that can efficiently pinpoint the exact disease causative agent and provide differentiation between rapid-decaying and slow-progressive Phellinus species remains elusive . However, methods combining nucleic acid amplification and DNA arrays have demonstrated strong potential, and their good sensitivity and specificity may allow for early and accurate detection of Phellinus infestation at the species level (Martin et al., 2000;L evesque, 2001;Tsui et al., 2011). In these strategies, DNA sequence samples are first amplified, then labelled with universal primers that target conserved regions flanking variable domains, and subsequently hybridized with speciesspecific oligonucleotide probes on DNA arrays to screen for fungal pathogens (Saiki et al., 1989;McCartney et al., 2003;Tsui et al., 2011). But while it is true that array technology currently offers the best chance of realizing a simple, efficient, high-throughput pathogen detection platform that can be readily deployed in the field, one key factor has limited development thus fara lack of discriminatory genetic regions available for species identification (Everett et al., 2010;Frey et al., 2010).
Here, we report the development of a robust highthroughput oligonucleotide microarray system capable of simultaneously screening for multiple Phellinus species. The system utilizes species-specific probes generated from internal transcribed spacer (ITS) regions, areas of non-coding DNA located between the small subunit and large subunit ribosomal RNA (rRNA)-coding genes, which are removed after transcription of the rRNA cistron (Lafontaine and Tollervey, 2001). It has been proposed that ITS regions can serve as a universal barcode marker for fungi, as they provide the broadest range of inter-and intra-species differentiation currently known (Schoch et al., 2012). Internal transcribed spacer sequences have previously been used to develop primers for PCR-based identification of Phellinus species (Nam et al., 2002;Gonthier et al., 2015), but to the best of our knowledge, this is the first study to utilize ITS regions in the development of DNA microarrays for detection of the Phellinus genus s.l. We were able to specifically resolve 17 key Phellinus species on our microarray system, including Phellinus apiahynus, P. cesatii, P. gilvus, P. linteus, P. inermis, P. laevigatus, P. melleoporus, P. membranaceus, P. noxius, P. pini, P. quercinus, P. ribis, P. igniarius, P. formosanus, P. pachyphloeus, P. torulosus (now reclassified as Fuscoporia torulosa) and P. weirii. Tests with template DNA sequences and field samples confirmed the sensitivity and specificity of our microarray system, which was also shown to reliably detect infections in trees before any visually identifiable symptoms of disease were observed. There is strong demand now in forestry, arboriculture and phytosanitation for pathogen detection systems that can be easily mastered and readily deployed to yield early, accurate and specific results in a very short space of time. Our Phellinus microarray system fulfils most of these practical requirements, and could potentially play a part in reducing the environmental and economic impact of Phellinus infestation worldwide.

Design of oligonucleotide probes and DNA microarrays
To develop a rapid and reliable detection system capable of inter-species differentiation, we first amplified the ITS1-5.8S-ITS2 genetic region of 17 Phellinus species, using universal primers previously described (White et al., 1990). The resulting PCR amplicons varied between 600 bp and 750 bp, and were further utilized in the design of oligonucleotide probes. The two ITS regions, ITS1 and ITS2, contain a high degree of sequence variations, which could potentially be used to generate specific probes capable of resolving different Phellinus species. A total of 48 probes, ranging between 28 bp and 60 bp in length and targeted to the ITS1 and ITS2 regions, were subsequently synthesized. Species-specific sequences were designed to be located at the centre of each probe, and the probes were subjected to extensive screening with hybridization assays to confirm specificity. Through this screening process, 17 oligonucleotide probes (Table 1), one for each key Phellinus species targeted, were eventually selected for the development of a reverse dot-blot hybridization DNA microarray (Fig. 1). In order to assess the applicability of the selected probes to different array systems, and to validate the reliability and efficiency of our microarrays, PCR amplicons from 27 target reference strains and 20 non-target strains were labelled with digoxigenindeoxynucleoside triphosphate (DIG-dNTP), DIG-tagged primers, biotin-dNTP or biotin-tagged primers, and then reverse-hybridized to probes spotted on either nylon membranes or polyvinyl chloride (PVC) chips, to derive a total of eight different array platforms ( Fig. 1; Fig. S1).

Reverse dot-blot hybridization of reference strains
Reverse dot-blot hybridization results revealed that all test isolates from target Phellinus species successfully hybridized with their respective oligonucleotide probes, with no cross-hybridization observed ( Fig. 1B and C; Fig. S1). No hybridization signals apart from those generated by the positive controls were observed when non-target strains were screened. Together, these results indicate that our microarray systems were capable of achieving 100% specificity under controlled laboratory conditions. Further-more, the use of PVC chips could potentially allow the screening process to be completed within 2 h (excluding the time required for target DNA amplication), which would represent a significant improvement over traditional Phellinus species identification methods that rely on morphological examination and pathogen isolation.

Sensitivity analysis of microarrays
To determine the sensitivity of our array system, we serially diluted template genomic DNA from P. weirii, a serious threat to Douglas fir viability in the Pacific Northwest of North America (Hansen and Goheen, 2000), and P. noxius, one of the most destructive Phellinus species in Taiwan (Ann et al., 2002). Samples respectively containing 1 ng, 100 pg, 10 pg, 1 pg, 100 fg and 10 fg of starting DNA were prepared, amplified and subjected to agarose gel electrophoresis and hybridization with different microarray systems. While DNA agarose gel bands were only visible with samples containing up to 10 pg of starting DNA (Fig. 2), our microarrays were able to accurately identify the respective Phellinus species with as little as 1 pg of starting DNA (Fig. 2). Biotin labelling was found to be more sensitive than DIG labelling, but no visible differences in detection sensitivity were observed between nylon membranes and PVC chips (Fig. 2).

Analysis of complex samples and field samples
In the natural environment, infestations often involve multiple fungal species, and a robust diagnostic system should have the ability to identify several different pathogenic species simultaneously. We therefore sought to assess whether our microarray system would be capable of detecting multiple Phellinus species within a single assay. We prepared complex samples that combined template genomic DNA from P. noxius, P. melleoporus, P. pini and P. weirii, and subjected the samples to reverse hybridization against our microarrays. The results showed that the arrays were capable of accurately detecting multiple Phellinus species in a single sample (Fig. 3).
We further prepared field samples from trees in Taiwan with suspected or confirmed Phellinus infestations. Dimocarpus longan, Cinnamomum camphora, Grevillea robusta, Ficus microcarpa and Prunus campanulata were among the tree species sampled. Our microarrays were able to detect the disease causative agent, P. noxius ( Fig. S2), before disease symptoms were visually identifiable, and the detection results were subsequently confirmed through conventional morphological analyses and pathogen isolation methods. We further used our microarray system to assess tree seedlings from a local plant nursery for the presence of Phellinus infestations. Our results showed that multiple seedlings of five species (Fraxinus formosana, Cinnamomum camphora, Koelreuteria elegans, Michelia champaca and Acacia confusa) sampled were all free of Phellinus infection (Fig. S3).

Discussion
Wood rot diseases caused by the Phellinus genus s.l. can induce decay in the roots, trunk and branches of almost all woody plants, and represent a serious threat to forest ecosystems and commercial arboriculture. Among the Phellinus species, P. weirii is considered to be one of the most destructive pests in the economically important Douglas fir forests of the Pacific Northwest in North America (Holah et al., 1993;Thies and Sturrock, 1995;Leckie et al., 2004); P. noxius has ravaged forests, plantations, orchards and urban landscapes across Japan, Taiwan and Southeast Asia (Ann et al., 1999;Mohd Farid et al., 2005;Sahashi et al., 2014); P. ignarius has been implicated as the cause of the grapevine black measles disease currently affecting vineyards in Europe and North America (Chiarappa, 1997;Gatica et al., 2004); and the trunk rot of pine caused by P. pini is viewed as one of the most dangerous invasive threats to plantation forests in Australia (Mireku and Simpson, 2002). Phellinus-related diseases are transmitted when the healthy roots of susceptible trees come into contact    with infected roots, stumps or soil; alternatively, Phellinus spores can also colonize wounds on tree trunks or branches (Leckie et al., 2004). In the initial stages of disease progress, infected trees exhibit few symptoms, and by the time crown symptoms or Phellinus mycelial strands and/or fruiting bodies appear, a significant portion of the root system will have been colonized, decayed, and destroyed, to the point that it is often too late to save the tree (Nam et al., 2002;Leckie et al., 2004). Precise identification of the exact disease-inducing Phellinus species is crucial for the accurate assessment of disease progression and effective deployment of control measures, but conventional methods that rely on morphological examination and pathogen isolation are slow and inefficient (Thorn et al., 1996;Adair et al., 2002;McCartney et al., 2003;Luisi and Campanile, 2004). Another cause for concern involves the increasing convenience of international transport and the effects of global warming, which can facilitate the spread of Phellinus species to new regions or temperate habitats that were previously inaccessible due to reasons of geography or climate (Mireku and Simpson, 2002;Sahashi et al., 2014). Considering that Phellinus species have been known to survive in wood or soil for as long as 50 years (Leckie et al., 2004), sensitive, rapid and convenient assays that can simultaneously screen for multiple Phellinus species will be needed to conduct quarantine inspections of soil, seeds and saplings, timber and other wood products. In this study, we describe an oligonucleotide microarray system that combines all of these key attributes, allowing for fast, sensitive and convenient simultaneous identification of 17 key Phellinus species. This would potentially facilitate the early detection of disease or contamination. Previously, microarrays have seen limited application due to the difficulty of identifying unique sequences that can be used for inter-species differentiation (Everett et al., 2010;Frey et al., 2010), but it has been recognized in recent years that the rRNA gene regions are highly variable, and may be used to generate speciesspecific markers or probes (Schoch et al., 2012). Nuclear large subunit (LSU) ribosomal DNA (rDNA) has been used in taxonomy to establish subdivisions within the Inonotus genus s.l. and Phellinus genus s.l. (Wagner and Fischer, 2001;2002a,b). The ITS regions of the rRNA genes have also been used for fungal identification (Turenne et al., 1999;Nam et al., 2002;Gonthier et al., 2015), and it was recently proposed that ITS sequences can serve as a universal bar code marker for fungi due to their excellent inter-and intra-species differentiation ability in a wide range of fungal species (Schoch et al., 2012). In this study, we initially designed 48 oligonucleotide probes targeting the ITS1 and ITS2 regions in 17 selected Phellinus species. The eventually selected probes ranged from 28-53 bp in length, excluding an additional seven thymine bases added to the 3'-ends to improve hybridization signals (Brown and Anthony, 2000;Peplies et al., 2003;Leaw et al., 2007). It is speculated that the improved effect is due to preferential attachment of the charged support to the added thymine bases at the 3'-end, thus leaving a greater number of oligonucleotides available for hybridization (Brown and Anthony, 2000). We further found that stronger hybridization signals were observed with longer (> 50 bp) probes, which accorded with earlier reports (Letowski et al., 2004;Rhee et al., 2004;Tiquia et al., 2004). Of the 48 oligonucleotide probes we screened, 17 probes with 100% sensitivity to the target DNA and 100% selectivity for the target species were selected for use in microarrays. All of the selected probes were 50 bp or longer in length, with the exception of the 28 bp Phine probe, which was shortened to increase specificity. To cater to the lower melting temperature (Tm) of this shorter probe, the hybridization temperature of our microarrays was adjusted to 48°C; this did not compromise the specificity of longer probes.
Multiplex PCR-based methods have previously been used to identify Phellinus s.l. at a generic rank (Guglielmo et al., 2007;2008), and ITS sequences have been used in the development of primers for PCR-based Phellinus species identification techniques (Nam et al., 2002;Gonthier et al., 2015). It is known that microarray techniques can facilitate the simultaneous identification of multiple pathogenic species in a high-throughput manner (Martin et al., 2000;L evesque, 2001;Tsui et al., 2011). To the best of our understanding, this is the first study to employ probes generated from ITS regions in DNA microarrays for the identification and differentiation of Phellinus species, and we elected to use reverse dot-blot hybridization to develop our arrays, as this method has been shown to result in fewer non-specific cross-hybridizations (L evesque et al., 1998). We spotted the 17 probes selected for sensitivity and specificity against target Phellinus species on to nylon membranes or PVC chips, and further added an amplification control and hybridization control (Fig. 1A). The amplification control was developed from PC1 control DNA, which can be amplified with universal ITS 1/4 primers in the same manner as Phellinus target DNA (White et al., 1990). To eliminate the possibility of competitive hybridization for amplicons between Phellinus probes and the PC1 amplification control probe, the former were designed to target only sense amplicons, while the latter targets only antisense amplicons. An additional hybridization control (HC1) DNA amplicon was also added to the hybridization buffer. Together, these two controls can provide verification of the amplification and hybridization process within our developed assay.
Amplicons derived from template DNA and field samples were labelled with DIG-dNTP, DIG-tagged primers, biotin-dNTP or biotin-tagged primers. Combined with the use of nylon membrane or PVC chip arrays, this allows for eight different array combinations. Based on the data obtained in this study, the optimal array system utilizes biotin-tagged primers for labelling, and PVC chips for the array platform. Excluding the time required for sequence amplification in samples, our array system can complete screening and detection in less than 2 h, and is capable of detecting P. weirii or P. noxius at 1 pg of starting DNA (Fig. 2). The other seven developed systems (Fig. 1B,  Fig. S1) display comparable levels of robustness, and together, these results demonstrate the broad applicability of our designed Phellinus probes to different array systems.
In this study, we showed that our microarray systems were capable of detecting target Phellinus species at about 1 pg of starting DNA (Fig. 2). Considering that the genome of Phellinus species is around 40 MB in length, at an average molecular weight of 660 Da for each nucleotide base pair, only 20-30 spores or cells would be needed for detection. However, it is important to note that if nontarget DNA amplified with the same set of primers exceeds target DNA, the detection limit may be affected . We also found that the choice of nylon membrane or PVC chip for the array platform did not noticeably affect detection sensitivity; however, the detection limit for biotin-labelled amplicons was slightly lower than DIG-labelled sequences, indicating greater sensitivity with biotin labelling. Since DIG labelling is generally considered to exhibit greater sensitivity than biotin labelling (Rihn et al., 1995a,b;Gauthier and Blais, 2003), we believe our findings may be the result of differences in streptavidin-alkaline phosphatase concentrations, and/or the number of alkaline phosphatases conjugated to each streptavidin molecule.
Our array systems could potentially be used to provide early detection of infestations involving multiple Phellinus species, and while general inferences to the diseasecausing organism can be made from the location and species of infected trees (Hansen and Goheen, 2000;Ann et al., 2002), such an array could provide much more definitive identification of disease causative agents, thus facilitating the effective implementation of control or quarantine measures. Interestingly, traditional Asian medicine has long considered P. linteus to have important medicinal properties, and recent studies have also isolated potentially useful compounds with immuno-stimulatory or anti-cancer properties from this species (Nam et al., 2002;Dai et al., 2010;Wu et al., 2012). However, there are major difficulties in differentiating between P. linteus and other species, such as P. igniarius, P. laevigatus and P. baumii through phenotypic methods. Considering that Phellinus species do not have the same medicinal effects, and that traditional Asian medicine relies on the use of the entire fruiting body or basidiocarp, our microarray system could plausibly be used to provide identification of important medicinal Phellinus species, in addition to more conventional applications in disease detection and phytosanitary inspection.
In conclusion, here we describe a set of eight DNA microarray systems that utilize probes generated from ITS regions to simultaneously detect 17 key Phellinus species. These arrays were shown to be capable of sensitive, specific, rapid and reliable detection, and could thus provide a significant advantage over traditional morphological or biochemical methods.

Fungal strains and growth conditions
Cultures of Phellinus strains were obtained from the Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan), the Taiwan Forestry Research Institute (TFRI, Taipei, Taiwan) and the USDA Forest Product Laboratory (USDA FPL, Madison, WI, USA) (Table ). All strains were grown on potato dextrose agar medium (PDA; BD Difco, Sparks, MD, USA) in the dark at 24°C for 7 days prior to DNA extraction.

Genomic DNA extraction
Genomic DNA was isolated from fresh fungal cultures, field samples or herbarium specimens, using a modified version of the cetyltrimethylammonium bromide (CTAB) method that was previously described (Doyle and Doyle, 1987). In brief, 0.1 g of mycelium or 0.3 g of field or herbarium specimens were placed in 1.5 ml centrifuge tubes containing 500 ll of preheated (65°C) CTAB isolation buffer (2% hexadecyltrimethylammonium bromide, 1.4 M NaCl, 20 mM EDTA, 100 mM Tris-HCl, pH 8.0) and crushed with a grinder tube, after which 3 ll of 2mercaptoethanol was added. The mixture was vortexed for 30 s and then incubated at 65°C for 10 min. The supernatant was extracted with phenol-chloroform-isoamyl alcohol (25:24:1, v/v) and centrifuged at 10 000 9 g for 2 min, after which 0.6 volumes of isopropanol was used to precipitate nucleic acids. The precipitated DNA was washed with 500 ll of wash buffer (76% ethanol, 10 mM ammonium acetate) and re-suspended in 20 ll of distilled deionized water containing 0.1 ll of RNase A (1 mg/ml concentration; Sigma, St. Louis, MO, USA). Concentration of DNA was determined through spectrophotometry (Nanodrop ND-1000, NanoDrop Technologies, Rockland, DE, USA).

DNA amplification and labelling
Universal fungal primers ITS1/4 (White et al., 1990) were used to amplify the ITS1-5.8S-ITS2 region in target Phellinus species. Amplification was carried out in a 50 ll reaction volume, containing 50 lM of each primer, 10 mM dNTP (GeneTeks BioScience, Taipei, Taiwan), 1 U Prime Taq DNA polymerase (GeNet Bio, Chungnam, South Korea) and 5 ll ( 1 ng) of template DNA, using the GeneAmp PCR 2400 System (PerkinElmer, Waltham, MA, USA). Polymerase chain reaction conditions were as follows: 94°C for 4 min and 35 cycles at 95°C for 30 s, 50°C for 30 s and 72°C for 1 min, with a final extension at 72°C for 7 min. Products of PCR were subjected to electrophoresis in 1.0 % (wt/vol) agarose gel and visualized by UV illumination after ethidium bromide staining.
For array analysis, ITS1/4 primers were also used to amplify ITS regions, and the amplicons were simultaneously labelled with either biotin or DIG. Biotin labelling was conducted using either 5'-biotin-tagged ITS 1/4 primer sets or biotin-16-dUTP mix (Roche Applied Science, Mannheim, Germany), which were added to the amplification reaction. Digoxigenin labelling was performed with either 5'-DIG-tagged ITS 1/4 primers or DIG-11-dUTP (PCR DIG Labeling Mixplus, Roche Applied Science). Polymerase chain reaction conditions for labelling were the same as described above, and the labelled amplicons were used as targets in subsequent microarray hybridization reactions.

Oligonucleotide probe design
Species-specific probes were designed by aligning the ITS sequences of Phellinus species, using the AlignX function in Vector NTI (Invitrogen, Carlsbad, CA, USA), to identify polymorphic regions specific to each species. Probes were designed according to the following criteria: (i) probe length of 25-55 bp; (ii) GC% of 40%; (iii) and Tm of 55-65°C. The Gibbs free energy (DG) of probes was calculated, and the presence of dimers and hairpin loops was assessed in order to minimize the formation of secondary structures. Seven additional thymine bases were added to the 3'-end of each probe to increase sensitivity (Brown and Anthony, 2000). Conserved regions of the 5.8S rRNA gene and ketosynthase domain were respectively used as positive PCR/hybridization or hybridization-only controls.

Oligonucleotide array preparation
Species-specific probes were synthesized by MDBio (Taipei, Taiwan), and diluted to a final concentration of 20 lM and 10 lM for spotting on positively charged polyvinyl chloride (PVC) chips (Dr Chip Biotechnology, Miaoli, Taiwan) or positively charged nylon membranes (Bio-Rad Laboratories, Hercules, CA, USA) respectively. Oligonucleotide probes were spotted onto PVC chips using automatic spotter, and fixed with UV Crosslinker (Spectrolinker XL-1000, Spectronics, Westbury, NY, USA). Alternatively, probes were spotted onto nylon membranes using EZspot arrayer (EZlife Technology, Taipei, Taiwan), air-dried and exposed to UV for probe immobilization.

Sensitivity analysis
To assess array sensitivity, P. weirii and P. noxius template genomic DNA was quantified using a NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies), diluted to starting DNA concentrations of 1 ng, 100 pg, 10 pg, 1 pg, 100 fg and 10 fg, and subjected to PCR amplification with biotin-labelled or DIG-labelled ITS1/4 primers. The PCR products were then subjected to electrophoresis on a 1.0% (w/v) agarose gel, as well as hybridization with PVC chip or nylon membrane microarrays, according to the procedures described above.

Detection and identification of phellinus species in complex samples and field samples
To ascertain if our array systems were capable of simultaneously identifying multiple Phellinus species in a single assay, complex samples containing template genomic DNA from 2 (P. melleoporus and P. noxius), 3 (P. melleoporus, P. noxius and P. pini) or 4 (P. melleoporus, P. noxius, P. pini and P. weirii) Phellinus species were prepared, labelled with biotin-tagged primers during amplification and hybridized to a PVC chip array. In addition, field samples collected from roots, stems, soil and herbariums throughout Taiwan were similarly prepared and hybridized to PVC chip arrays. The field samples were also subjected to conventional detection involving cultivation on PDA or semi-selective media (Chang, 1995) followed by microscopic examination; and the results were compared with those derived from microarray analysis. Wagner, T., and Fischer, M. (2002a) Proceedings towards a natural classification of the worldwide taxa Phellinus s.l. and Inonotus s.l., and phylogenetic relationships of allied genera. Mycologia 94: 998-1016. Wagner, T., and Fischer, M. (2002b) Classification and phylogenetic relationships of Hymenochaete and allied genera of the Hymenochaetales, inferred from rDNA sequence data and nuclear behaviour of vegetative mycelium. Mycol Prog 1: 93-104. White,T.J., Bruns, T.D., Lee, S.B., and Taylor, J.W. (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In

Supporting information
Additional Supporting Information may be found in the online version of this article at the publisher's web-site: Fig. S1. Reverse hybridization of differentially labelled amplicons to Phellinus probes spotted on nylon membrane or PVC chip arrays. Probes were arranged on arrays as indicated in Fig. 1A.
A. Digoxigenin-deoxynucleoside triphosphate-labelled amplicons hybridized to probes on nylon membrane. B. Digoxigenin primer-labelled amplicons hybridized to probes on nylon membrane. C. Biotin-dNTP-labelled amplicons hybridized to probes on nylon membrane. D. Digoxigenin-deoxynucleoside triphosphate-labelled amplicons hybridized to probes on PVC chip. E. Biotin-dNTP-labelled amplicons hybridized to probes on PVC chip. F. Biotin-primer-labelled amplicons hybridized to probes on PVC chip. For results of biotin-primer-labelled amplicons hybridized to probes on nylon membrane, and DIG-primerlabelled amplicons hybridized to probes on PVC chip, please see Fig. 1B and C respectively. confusa seedlings were found to be free of Phellinus infestation through microarray analysis.