Direct Analysis of Wood-Inhabiting Fungi Using Denaturing Gradient Gel Electrophoresis of Ampliﬁed Ribosomal DNA

Mycol. Res. 104 (8): 927–936 (August 2000). Printed in the United Kingdom. 927

Direct analysis of wood-inhabiting fungi using denaturing gradient gel electrophoresis of ampliﬁed ribosomal DNA

Eeva J. VAINIO* and Jarkko HANTULA Finnish Forest Research Institute, P.O. Box 18, FIN-01301, Vantaa, Finland. E-mail: eeva.vainio!metla.ﬁ.

Received 13 August 1999; accepted 20 November 1999.

Four different primer pairs were designed for the preferential PCR amplification of fungal SSU rDNA directly from environmental samples. Most of the amplification products obtained from a reference collection of 46 wood-decomposing fungi could be separated by denaturing gradient gel electrophoresis (DGGE) using 1650 bp rDNA fragments produced by primer pair FR1jNS1. A relatively high level of resolution was also achieved using 390 bp products amplified with primer pair FR1jFF390. In contrast, the separation of amplification products obtained by the remaining two primer pairs (FR1jFF700 and FR1jFF1100; 700 bp and 1100 bp, respectively) was inadequate when applied to our fungal collection. Differentiation between all the species tested was achieved by combined analysis of the rDNA fragments produced by primer pairs FR1jNS1 and FR1jFF390. The DGGE analysis of environmental samples collected from Norway spruce stumps showed that the analysis of fungal DNA extracted directly from wood was usually in accordance with the investigation of mycelial cultures isolated from the same decay column. In some cases, however, disagreement was observed, which suggests that these two fundamentally different techniques present different views about fungal diversity. The investigation of fungal species profiles directly from environmental samples using DGGE analysis of PCR-amplified SSU rDNA molecules can be used to improve the detection of fungal groups that are difficult to cultivate.

INTRODUCTION et al. 1997, Ka/ re! n et al. 1997, Johanneson & Stenlid 1999, Jonsson et al. 1999a, Jonsson et al. 1999b). Many fungi, Ecological inventories of wood-inhabiting fungi have tra- however, contain within-species variation in their ITS ditionally been based on inspecting the occurrence of sequences (Anderson & Stasovski 1992, O’Donnel 1992, fruit bodies (Hintikka 1993, Renvall et al. 1995). Since, however, Kasuga et al. 1993, Feibelman, Bayman & Cibula 1994, fruit body development is affected by environmental factors Neuve! glise et al. 1994, Hallenberg, Larsson & Mahlapuu 1996, and many microfungi do not produce visible sporocarps, Norman & Egger 1996, Ka/ re! n et al. 1997, Fatehi & Bridge fruit body distribution may present a limited view of the 1998), which may cause overestimation of species diversity fungal diversity existing as vegetative mycelia (Gardes & when analysis is based on different ITS-types. A more Bruns 1996, Dahlberg, Jonsson & Nylund 1997, Johannesson conserved region of the ribosomal gene cluster might, & Stenlid 1999, Jonsson et al. 1999a, Jonsson et al. 1999b). therefore, be more suitable for the analysis of fungal The use of artificial cultivation media enables the isolation communities containing many unknown species. A promising of vegetative forms of fungi independently of sporocarp alternative would be the SSU (small subunit or 18S) rDNA formation (Rayner & Boddy 1988). The adequacy of culture- gene, which shows very limited within-species variation, but based methods is, however, limited by the selectivity of the different species can usually be separated from each other by isolation procedure used, mainly due to the lack of suitable their sequence (Berbee & Taylor 1993, Olsen & Woese 1993, culturing media for some species. This can be overcome by the Mitchell, Roberts & Moss 1995). analysis of total DNA extracted directly from an environ- RFLP analysis has been the most common method used to mental sample. In theory, the species composition of a mixed investigate sequence variation within the rRNA gene cluster DNA sample can be revealed by PCR-analysis, provided that from uncultured fungal communities (Gardes & Bruns 1996, the amplification efficiencies from different templates are Dahlberg et al. 1997, Ka/ re! n et al. 1997, Johannesson et al. similar. 1999, Jonsson et al. 1999a, Jonsson et al. 1999b). This The ITS region of the rRNA gene cluster has been technique as well as sequence determination, requires a successfully used for studying fungal species profiles directly laborious cloning step in situations where each sample is from environmental samples (Gardes & Bruns 1996, Dahlberg inhabited by more than one species (Helgason et al. 1998). This can be circumvented by the use of denaturing gradient * Corresponding author. gel electrophoresis (DGGE), that separates DNA-fragments Direct analysis of fungal diversity 928 according to their sequences (in addition to length poly- plates (Difco, Detroit MI, USA) checking each sample daily in morphisms) and allows the simultaneous analysis of several order to detect all the emerging fungi. The resulting fungal different sequences PCR-amplified from a single environmental cultures were transferred to MOS (modified orange serum) sample (Muyzer, De Waal & Uitterlinden 1993). Among agar plates (Mu$ ller, Kantola & Kitunen 1994) supplemented several bacterial studies (Ferris, Muyzer & Ward 1996, Heuer with cellophane membranes to facilitate the removal of et al. 1997, Kowalchuk et al. 1997b, Vallaeys et al. 1997, Smalla mycelia for DNA extraction. et al. 1998), DGGE analysis has recently been applied also for eukaryotic communities represented by protozoan (Ciliophora) DNA extraction populations from activated sludge (Marsh et al. 1998), and fungal inhabitants of Marram grass roots (Kowalchuk, Gerards Extraction of DNA directly from wood samples was carried & Woldendorp 1997a) and the wheat rhizosphere (Smit et al. out using a multistep procedure beginning with the homo- 1999). genization of the wood chips using a glass rod and quartz The aim of this investigation was to compare fungal species sand (granulation size 0n1–0n5 mm; Riedel-deHae$ n, Seelze, profiles detected from Norway spruce stumps using two Germany) and disrupting the cells in extraction buffer fundamentally unrelated methods: (i) isolation of mycelial (50 m Tris\HCl, pH 7n2; 50 m EDTA; 3% SDS; 1% beta- pure cultures or (ii) analysis of fungal DNA extracted directly mercaptoethanol) at 65 mC for 1 h. The cell lysate was from wood. In order to test whether these techniques present extracted for five times with phenol\chloroform\isoamyl different views about fungal diversity fungus-specific primers alcohol (25:24:1, by vol.) and twice with chloroform\isoamyl were designed for the amplification of partial SSU rDNA alcohol (24:1, by vol). Further purification was carried out fragments and the DGGE technique was optimised for their using the High Pure PCR Template Preparation Kit (Boehringer analysis. In addition, a DNA purification procedure was Mannheim, GmbH, Indianapolis, USA) according to the developed to enable the PCR-amplification of fungal rDNA manufacturer’s instructions. Finally, the DNA was selectively directly from wood extracts. precipitated by adding 0n6 vol. of a solution containing 20% (w\v) polyethylene glycol (PEG 6000) and 2n5  NaCl and incubating on ice for 20 min. The samples were pelleted by MATERIALS AND METHODS centrifugation in a microcentrifuge (14000 rpm for 20 min), Reference organisms washed with 70% ethanol, dried under vacuum and resus- pended in TE-buffer (6 m Tris\HCl, pH 8n0; 1 m EDTA). The fungal isolates used for primer testing and DGGE The same extraction procedure was used for the fungal pure optimisation are listed in Table 1. This reference collection cultures, except omitting the use of the High Pure PCR includes common inhabitants of coniferous logging residues in Template Preparation Kit. boreal forests. Axenic tissues of Scots pine (Pinus sylvestris), birch (Betula sp.) and Norway spruce (Picea abies), and three different bacteria (Escherichia coli DH5α, Agrobacterium Primer design tumefaciens and Anabaena sp.) were used for the primer Potential primer target regions were located by comparing the specificity testing. complete SSU rDNA sequences of a phylogenetically diverse collection of organisms including 13 ascomycetes, six basidiomycetes, nine animals and four plants. Sequences with the Field samples following GenBank accession numbers were used: D14165, Sample discs were collected from seven Norway spruce L37537, L37539, L37735, M83257, M83258, M83263, stumps two growth seasons after felling (autumn 1996) in a U00975, X58056, X69845, X69848, X69850, Z27393 single forest stand located in Nummi-Pusula (southern Finland). (ascomycetes); D13460, L22259, L36658, M94337, M94339, The outermost layer (ca 5 cm) of the stumps was discarded U00973 (basidiomycetes); D14365, D15067, L10826, L10827, and discs cut below this surface were visually examined to L49053, U29494, U36270, X53047, Z19562 (multicellular select distinct uniformly coloured decay columns of different animals); and D38245, U18632, X16077, X56105 (plants). The sizes to be used for cultivation of mycelia and direct extraction sequences were aligned using GCG (Genetics Computer of DNA. Group, University of Wisconsin, Madison, USA) software. Three adjacent samples were taken from each decay column The sequence alignment revealed a region near the 3h end as follows: the surface wood was removed aseptically and of the SSU rDNA, which was invariant in the fungi, but discarded, and the first cultivation sample (c1, wood fragment contained differences compared to all the other organisms $ of ca 10 mm ) was taken below the fresh surface with a sterile included in the analysis. This region was used as a target site knife. The next sample (w) was used for direct extraction of for a PCR-primer designated as ‘fungus-specific reverse DNA and it was considerably larger in volume (approx. primer 1’ (FR1), as it directs DNA synthesis towards the 5h $ 200–300 mm ) to obtain a sufficient amount of DNA. The end of the SSU rDNA gene. The possible annealing of primer $ second cultivation sample (c2, ca 10 mm ) was taken from the FR1 to non-fungal rDNA templates was evaluated by FastA opposite side of this DNA sample to ensure that each decay searches in the GenBank database. Matching sequences were column was uniform in culturable fungal composition. observed in a high number of fungi as expected, but also in The cultivation samples were incubated at room tem- certain representatives of other groups (including Echino- perature (approx. 20 mC) on 1n5% (w\v) malt extract agar dermata, Chaenoflagellida, Mollusca, Arthropoda and Magno- E. J. Vainio and J. Hantula 929

Table 1. List of fungal isolates used in the reference collection and the DGGE migration types of their partial SSU rDNA fragments. The isolates used in initial DGGE optimization are indicated with an asterisk (*).

SSU rDNA with DGGE type‡ extended length§ (bp) FR1jNS1\ Isolate† FR1jNS1 FR1jFF390 FR1jFF390

Amylocystis lapponica 11715\1; RP 7a (1) 23 (2) — Amylostereum areolatum Ref2\36; RV 10c (6) 14 (2) — A. chailletii Ref2\34; RV 10c (6) 23 (2) — Armillaria borealis* 92146\4; KK 7c (1) 8 (2) — A. cepistipes 92159\VE; KK 12d (1) 16 (1) — A. ostoyeae 94276\2; KK 10d (1) 6 (2) — Ascocoryne sp. M13\4; EJV 12c (1) 11 (1) — Bjerkandera adusta* Ref1\8; AMH 9 (1) 21 (4) — Ceratocystis sp.* MU5\K4; EJV 17 (1) 28 (3) — Chondrostereum purpureum Ref1\16; AMH 2a (1) 17 (1) 2075\— Coniophora arida 6C7; AMH 3 (1) 28 (3) — Cylindrobasidion evolvens Var1\45; AMH n.d. n.d. — Exophiala sp.* Var3\10; AMH 14b (1) 27 (2) — Fomitopsis pinicola* 7C8; AMH 12b (3) 14 (2) — Gliocladium sp.* M84\1; EJV 5d (1) 4 (1) 2775\730 Heterobasidion annosum K61\7; EJV 10c (6) 28 (3) — H. parviporum* K42\4; EJV 10c (6) 26 (1) — Hypholoma capnoides* 1iA11; AMH 5b (1) 19 (2) — Ischnoderma benzoinum Ref1\33; AMH 10a (1) 25 (1) — Merulius lacrymans 10a10; AMH 5c (1) 21 (4) — Nectria sp. anamorph* K61\11; EJV 8b (1) 6 (2) — Neobulgaria premnophila* 11d9; AMH 18 (1) 8 (2) 2000\— Panellus mitis* Ref1\75; AMH 10c (6) 29 (1) — Penicillium sp.* M71\A1; EJV 11 (1) 30 (1) — Peniophora pithya* Ref1\31; AMH 16 (1) 21 (4) — Phaeolus schweinitzii Ref2\25; AMH n.d. n.d. — Phialophora sp.* Var7\27; AMH 15 (1) 24 (1) — Phlebiopsis gigantea* K31\7; EJV 13 (2) 21 (4) — Polyporus borealis Ref1\36; AMH 13 (2) 7 (1) — P. brumalis Ref1\37; AMH 12b (3) 22 (2) — Resinicium bicolor 12a7; AMH 6a (2) 9 (1) — Rhinocladiella sp.* MU7\5; EJV 14a (1) 12 (1) 2075\— Sistotrema brinkmannii* 202\1; AMH 10c (6) 19 (2) — Stereum sanguinolentum* K41\10; EJV 6a (2) 22 (2) — Trametes zonata Ref1\20; AMH 12b (3) 20 (1) — Trichaptum sp. FB9b; EJV 7b (1) 15 (1) — Trichoderma sp.* K62\11; EJV 4 (1) 10 (1) — Verticicladiella procera Ref2\24; AMH 14c (1) 27 (2) — Unidentiﬁed basidiomycetes Ba\sp. 1 2c3; EJV 6b (1) 18 (2) — Ba\sp. 2 K51\1; EJV 8a (1) 18 (2) — Ba\sp. 3 M72\1; EJV 12a (1) 13 (1) — Ba\sp. 4 M33\1; EJV n.d. n.d. — Unidentiﬁed mitosporic fungi De\sp. 1 K32\13; EJV 10b (1) 5 (1) 2150\850 De\sp. 2 M31\1; EJV 1 (1) 1 (1) — De\sp. 3* M52\3; EJV 2b (1) 2 (1) — De\sp. 4 KU1\12; EJV 5a (1) 3 (1) —

† Cultures were provided by Reijo Penttila$ (RP), Rimvydas Vasiliauskas (RV), Anna-Maija Hallaksela (AMH) or Kari Korhonen (KK). The strains isolated by Eeva Vainio (EJV) were cultivated from stump samples of Norway spruce or Scots pine and identified or preliminarily characterized (unidentified basidiomycetes or mitosporic isolates) by A.-M. Hallaksela. Five different isolates of H. parviporum, S. sanguinolentum, P. gigantea, Trichoderma sp., and De\sp.1 were used. ‡ Relative position of the SSU rDNA fragment in denaturant gradient gels. The bands are numbered according to their migration rate, the slowest-moving l 1. For primer pair FR1jNS1, letters are used to describe groups that were differentiated using denaturant gradients of 18–38%, plain numbers indicating groups separable by gradients of 18–43%. The number of species showing the same migration rate is indicated in parenthesis. Ba\sp.4, with insufficient amplification, and C. evolvens and P. schweinitzii, that produced multiple PCR-fragments, were excluded from the analysis (n.d.). § The PCR products obtained with primer pairs FR1jNS1 or FR1jFF390 were normally approx. 1650 bp and 390 bp long, respectively. The lengths of unusually long PCR-products are indicated for both primer pairs as rough estimates obtained from agarose gel comparisons. Direct analysis of fungal diversity 930

NS1 FF1100 FF700 FF390

IGS SSU rDNA ITS

FR1

300 bp Fig. 1. Schematic representation showing the annealing sites of the PCR primers along the nuclear gene coding for the SSU rDNA. The relative positions of the primers and their direction of extension are indicated by arrows.

Table 2. PCR primers and thermocycling conditions used for the ampliﬁcation of partial SSU rRNA genes from fungi.

Product size† Primer Sequence (5h 4 3h) (bp) Thermocycling programme

FR1* AIC CAT TCA ATC GGT AIT — — FF390 CGA TAA CGA ACG AGA CCT 390 8 min at 95 mC, followed by 30 cycles of: 30 s at 95 mC, 45 s at 50 mC, 2 min at 72 mC, and final extension for 10 min at 72 mC FF700 GAT ACC GTI GTA GTC T 700 8 min at 95 mC, followed by 35 cycles of: 30 s at 95 mC, 45 s at 47 mC, 2 min at 72 mC, and final extension for 10 min at 72 mC FF1100 CCA GCT CCA ATA GCG TAT ATT A 1100 8 min at 95 mC, followed by 35 cycles of: 30 s at 95 mC, 45 s at 47 mC, 2 min at 72 mC, and final extension for 10 min at 72 mC NS1 GTA GTC ATA TGC TTG TCT C 1650 8 min at 95 mC, followed by 35 cycles of: 30 s at 95 mC, 45 s at 47 mC, 3 min at 72 mC, and final extension for 10 min at 72 mC * Fungal specific reverse primer used in combination with each of the forward primers FF390, FF700, FF1100 and NS1 (Fig. 1). † The approximate product length observed in the majority of reference species, longer fragments occurred in isolates described in Table 1. liophyta). Thus, although primer FR1 can be considered to Table 3. Denaturing gradient gel electrophoresis conditions used for the have highly enhanced selectivity for fungi, it may also anneal analysis of partial SSU rDNA fragments of different lengths. to a limited set of other organisms. Denaturing Three different forward primers (FF390, FF700, and FF1100) gradient were designed to be used in combination with primer FR1 to (%) Electrophoresis allow the amplification of partial SSU rDNA fragments of FR1jFF390 45–60 18 h at 50 V and 58 mC different lengths. The relative locations of the primers along FR1jFF700 40–55 18 h at 78 V and 58 mC the SSU rDNA gene are shown in Fig. 1 and the primer FR1jFF1100 30–55 17 h at 130 V and 58 mC sequences are described in Table 2. Inosine nucleotides were FR1jNS1 18–43 and 18–38 17 h at 180 V and 58 mC used to accommodate variation in the fungal rDNA sequences covered by oligonucleotides FR1 and FF700 (using two and Denaturing gradient gel electrophoresis (DGGE) one inosine residues, respectively). The partial SSU rDNA fragments were analysed by the D GENE system (Bio-Rad) using 7 5% (w\v) acrylamide\ PCR amplification of partial SSU rDNA fragments 4 n bisacrylamide (37n5:1) gels. The denaturant gradients were PCR amplifications of the partial SSU rDNA fragments were produced with 100% denaturing solution containing 40% carried out in 50 µl reaction mixtures containing 5 µlof deionised formamide and 7  urea. The gels were run in 10ireaction buffer (100 m Tris\HCl, pH 8n8; 1iTAE-buffer (40 m Tris\Acetate, pH 8; 1 m EDTA) at a 15 m MgCl#; 500 m KCl; 1% Triton X-100), 200 µ of constant temperature of 58 mC and the rDNA fragments were each dNTP, 1 U of Dynazyme II DNA polymerase (Finnzymes visualised by ethidium bromide staining under uv-light. The Ltd, Finland), 0n5 µ of primer FR1 and 0n5 µ of one of the DGGE mobilities of different samples were compared in forward primers (FF390, FF700, FF1100 or NS1). The PCR parallel on the same gels in order to reveal differences or to conditions were determined experimentally for each primer confirm similarity. combination as described in Table 2. In order to obtain Melting behaviour of the PCR-fragments was initially efficient separation in DGGE, a GC-clamp sequence (under- analysed by perpendicular denaturant gradients (primer pairs lined) was added to primer FR1 resulting in a 58-meric primer: FR1jFF390 and FR1jFF700) or parallel gradients (primer 5h CCC CCG CCG CGC GCG GCG GGC GGG GCG pairs FR1jFF1100 and FR1jNS1) using a wide denaturant GGG GCA CGG GCC GAI CCA TTC AAT CGG TAI T 3h concentration range (20–80%). The gradients were subse- (where I l inosine). The Expand4-High Fidelity PCR System quently narrowed down on the basis of the observed band (Boehringer Mannheim) was used according to the manu- mobility to allow optimal separation for a taxonomically facturer’s instructions to allow the amplification of Norway diverse collection of 20 reference fungi (Table 1). The spruce rDNA for control purposes. optimised DGGE conditions are described in Table 3. E. J. Vainio and J. Hantula 931

Random amplified microsatellites (RAMS) analysis in the absence of fungal DNA, however, weak non-specific PCR-products were obtained from plant templates. The PCR amplification of the RAMS markers was carried out as selectivity of the primer pair was further tested using described in Hantula, Dusabenyagasani & Hamelin (1996), experimental PCR mixtures containing 0n1 ng of fungal DNA using the CGA-primer: 5h DHB(CGA)&, where D l G\A\T, (Ascocoryne sp., Ceratocystis sp., Trichoderma sp., Heterobasidion H l A\T\C, B l G\T\C. annosum, Stereum sanguinolentum or Phlebiopsis gigantea) mixed with 100 ng of Norway spruce DNA, thus creating a 1000- RESULTS fold excess of plant DNA. Examples of these PCR-mixtures Optimisation of PCR and DGGE conditions are shown in Fig. 2. The possible amplification of spruce templates was checked by DGGE analysis comparing the The PCR conditions were optimised for the four different mixed PCR products (see Fig. 2, lanes 1, 5 and 8) to primer combinations used (FR1jFF390, FR1jFF700, spruce-originating SSU rDNA fragments (produced by using FR1jFF1100 and FR1jNS1) to allow successful ampli- the Expand4 PCR System, which provided less specificity for fication of fungal DNA extracted both from cultivated isolates fungi than the Dynazyme II polymerase used otherwise in this and directly from wood samples while selecting against study: Fig. 2, lane 9), and no plant-originating bands were bacterial and plant DNA (Table 2). Optimal DGGE conditions detected among the amplification products. were determined experimentally for each of the four different fragment types to examine the resolution capacities of different regions of the SSU rDNA. The optimised electro- Primer design: amplification efficiency from cultured phoresis conditions and denaturant concentrations are de- fungal isolates scribed in Table 3. The amplification efficiencies of the different primer pairs were initially tested using a collection of 20 fungal species (indicated Primer design: selectivity for fungi with asterisks in Table 1), which resulted in successful Three different bacterial isolates and axenic tissues of Scots amplification from all isolates and primer combinations. Primer pine, birch and Norway spruce were used for the primer pairs FR1jNS1 and FR1jFF390 were further tested using specificity testing. No detectable amplification products were all of the 46 reference species (Table 1), only one of which (the obtained from either bacterial or plant templates using three unidentified basidiomycete Ba\sp.4) showed insufficient ampli- of the primer pairs (FR1jFF390, FR1jFF700 or fication with primer pair FRjNS1 and also produced a low FR1jFF1100) under the PCR conditions described in Table 2. yield using primer pair FR1jFF390. Efficient amplification Also the fourth primer pair (primer FR1 combined with was obtained from the remaining 45 species, indicating that primer NS1 obtained from White et al. 1990) allowed full both primer pairs are applicable for analysing a wide range of selectivity against bacterial DNA. When using this primer pair different ascomycetous and basidiomycetous taxa.

M12345678910

Fig. 2. DGGE analysis of experimental PCR mixtures produced using pairwise combinations of two different DNA templates. The DNA mixtures were amplified using primer pair FR1jNS1 and analysed in 18–43% denaturant gradient gels. Lanes 2–4 and 6–7 show serial dilutions of two different fungal templates, while lanes 1, 5 and 8 contain fungal DNA amplified in combination with 1000-fold excess of Norway spruce DNA. Lanes 9 and 10 contain SSU rDNA amplified separately from axenic Norway spruce templates using the Expand4-High Fidelity PCR System (lane 9) or Dynazyme II polymerase (lane 10). The following species and template DNA ratios were used in the mixtures: P. gigantea and Norway spruce 1:1000 (lane 1); P. gigantea and S. sanguinolentum 10:1 (lane 2), 1:1 (lane 3), 1:10 (lane 4); S. sanguinolentum and Norway spruce 1:1000 (lane 5); S. sanguinolentum and Ascocoryne sp. 10:1 (lane 6), 1:10 (lane 7); Ascocoryne sp. and Norway spruce 1:1000 (lane 8). Lane M contains marker DNA (100 bp DNA ladder). Direct analysis of fungal diversity 932

(a) using primer pair FR1jFF390, which can be expected since M12345678910M this primer set covers only a small part of the SSU rDNA region ampliﬁed by primer pair FR1jNS1.

Quantitative analysis of PCR-amplification using mixed DNA templates The sensitivity of PCR amplification to detect different species from mixed DNA templates was investigated using primer pair FR1jNS1. Experimental template mixtures were constructed using pairwise combinations of DNAs from two different fungi containing either equal amounts of both DNAs or one of the templates being diluted in 1:10 ratio. The following five species combinations were tested: Phlebiopsis gigantea with Heterobasidion annosum or with Stereum sanguinolentum; and Ascocoryne sp. with S. sanguinolentum, Trichoderma sp. or Ceratocystis sp. In all cases, both species (b) could be detected by DGGE even when one of them M123 45678910M constituted only 10% of the total DNA mixture (Fig. 2, lanes 2, 4, 6 and 7), although some reduction from the expected relative band intensity was observed in Ceratocystis sp. In addition, community fingerprints were successfully generated from experimental PCR-mixtures containing equal amounts of DNAs from eight different species: P. gigantea, Coniophora arida, Nectria sp., Penicillium sp., Peniophora pithya, Hypholoma capnoides, Ascocoryne sp. and S. sanguinolentum.

DGGE of cultivated isolates: differentiation between species The PCR-fragments produced by primer pair FR1jFF1100 did not allow adequate separation of different species, as most of the isolates occupied a relatively narrow area on the gels under the optimised DGGE conditions. A wider amplitude of band migration in the denaturant gradient was achieved using primer pair FR1jFF700, suggesting that this primer set Fig. 3. DGGE analysis of partial SSU rDNA fragments amplified might be adequate for fungal community studies, but the from different species of fungi. Lane 1, Chondrostereum purpureum; resolution was not optimal for the reference collection used in 2, Trichoderma sp.; 3, mitosporic isolate De\sp.4; 4, Stereum this investigation. sanguinolentum;5,Panellus mitis;6,Ascocoryne sp.; 7, Phialophora sp.; Primer pair FR1jNS1 was used to produce long PCR- 8, Peniophora pithya;9,Ceratocystis sp.; 10, Neobulgaria premnophila. fragments covering most of the entire SSU rDNA molecule (ca Lane M contains marker DNA (100 bp DNA ladder). (a) Fragments 1650 bp). Optimal resolution for most of the products was amplified by primer pair FR1jNS1 and analysed in denaturant achieved using denaturant gradients of 18–43% (Fig. 3a). (It gradient of 18–43%. (b) Fragments amplified by primer pair should be noted, that artifact bands probably representing FR1jFF390 and analysed in denaturant gradient of 45–60%. single-stranded DNA appeared in the gels when higher denaturant concentrations were used.) Amplification products The approximate size of the amplification products was from some of the tested species could, however, only be usually 390 bp for primer pair FR1jFF390 and 1650 bp for separated using a more narrow denaturant gradient of 18–38%, primer pair FR1jNS1. Unusually long PCR-products were, but this could not be applied to all the tested fragments due however, observed in some fungal isolates (Table 1), to their fast movement in low denaturant concentrations. suggesting that introns or other insertions would occur in Initial analysis of unknown samples was, therefore, carried out their SSU rDNA genes. Using primer pair FR1jNS1, this was using denaturant gradients of 18–43%, and isolate groups observed in Rhinocladiella sp., Gliocladium sp., Chondrostereum with similar migration rate were subsequently analysed by purpureum, Neobulgaria premnophila, the mitosporic De\sp.1, gradients of 18–38%. Since this SSU rDNA region showed and two of the fungal strains cultivated from the stump promising resolution capacity, the entire reference fungal samples (isolates 5Ac1b and 5Ac2b producing fragments collection (excluding Ba\sp.4 with insufficient amplification, of ca 2075 bp; and 7Cc1a and 7Cc2a, ca 2600 bp). Only and C. evolvens and P. schweinitzii that produced multiple PCR- some of these isolates (Gliocladium sp., De\sp1 and 7Cc1a fragments) was analysed using both denaturant gradients. As and 7Cc2a) produced PCR-fragments with extended length a result, 43 reference species could be divided into 34 distinct E. J. Vainio and J. Hantula 933 mobility groups (e.g. groups consisting of molecules with the 21 direct DNA extracts analysed did not provide sufficient identical migration rate in DGGE), 30 of which contained only PCR amplification. one species (Table 1). A relatively high level of resolution was also achieved Field samples: selection of uniformly colonised decay using the amplification products obtained with primer pair regions FR1jFF390. For this fragment, denaturant gradients of 45–60% showed optimal separation for most of the reference Comparison between the directly extracted DNA-samples isolates (Fig. 3b). The analysis of 43 reference species revealed and cultivation samples was carried out by DGGE-analysis of 30 mobility groups, 20 of which were composed of only one their partial SSU rDNA fragments. Before this it was verified species (Table 1). As with primer pair FR1jNS1 narrowing of that each decay region was actually uniform in its culturable the denaturant gradient might, however, further increase the fungal composition. Two separate cultivation samples were separation capacity of this SSU-fragment for certain species. taken from the opposite sides of each direct DNA sample and Using primer pair FR1jNS1, four groups were formed the resulting pure cultures were analysed using the highly from isolates with identical migration rate in DGGE (Table 1). sensitive RAMS-fingerprinting technique to confirm that the The largest group was composed of six different species, same species occurred in both of the cultivation samples. The including both close relatives (Heterobasidion annosum and H. RAMS fingerprinting revealed 21 uniformly colonised regions, parviporum,orAmylostereum areolatum and A. chaillettii), and 18 of which contained only one culturable fungal strain, while species distantly related to them (Panellus mitis). Using primer in three cases (decay regions 5A, 6B and 7C) two different pair FR1jFF390, identical migration rate was in some cases fungi grew out from both cultivation samples. These 21 observed between ascomycetous and basidiomycetous taxa regions were considered suitable for direct DNA analysis (Armillaria ostoyae with Nectria sp., and A. borealis with since it could be expected that each direct wood sample would Neobulgaria premnophila, Table 1). When the reference isolates be identical in fungal composition with the cultivated samples were analysed by both primer pairs, however, all the 43 adjacent to it. species analysed could be successfully separated. Field samples: comparison between direct DNA samples and cultivated isolates DGGE of cultivated isolates: variation within species Primer pair FR1jNS1 was selected due to its high resolution Multiple isolates were analysed from some species to find out capacity for the DGGE analysis of altogether 66 field samples whether fragments produced by primer pairs FR1jNS1 and (20 direct wood samples and 46 fungal pure cultures, excluding FR1jFF390 would show within-species variation. No decay region 2B with insufficient PCR amplification). The intraspecific variation was detected among five different analysis of 17 decay regions containing a single culturable isolates representing the following species: P. gigantea, H. fungal strain revealed 13 cases where the DGGE migration parviporum, S. sanguinolentum, Trichoderma sp., and De\sp.1 rate of the directly isolated DNA sample was identical to that that produced an unusually long amplification fragment (Table of the cultivation samples (Fig. 4a). 1). In cases where two different fungi grew out from each In some cases, however, more than one band position cultivation sample (decay regions 5A, 6B and 7C), however, (sequence type) was revealed by DGGE analysis from single only one of them could be detected from the corresponding fungal isolates. Using primer pair FR1jNS1 two minor bands direct DNA sample. Experimental PCR mixtures were were observed in addition to one bright band in the PCR constructed from the cultivated isolates to elucidate whether product of Cylindrobasidion evolvens and with primer pair this was due to selectivity of PCR amplification against the FR1jFF390 the same situation occurred in P. schweinitzii.A undetected species. Unlike the direct DNA samples (see Fig. faint additional band was observed also in De\sp.3 using 4b, lane 9), the mixed PCR-products (Fig. 4b, lane 8) contained primer pair FR1jFF390. SSU-fragments amplified from both species (Fig. 4b, lanes 7 and 10), although some reduction of relative band intensity was observed in the species that could not be detected from Field samples: extraction of DNA directly from wood the direct DNA samples. This suggests that those species that As inspected by agarose gel electrophoresis, most of the were detected from the direct DNA samples were also wood samples provided low amounts of DNA: the yields dominant by their amount of mycelia. It should also be noted −" were usually approximately 0n5ngµl or lower, while those that although species determination of the cultures was not an " from fungal pure cultures were normally ca 10–20 ng µl− . object of this study, all the undetected strains clearly belonged Thus, the amount of wood material required for DNA to slow-growing ascomycete groups. # isolation (200–300 mm ) was substantially higher compared In addition to these double samples, there were four cases # to that sufficient for isolating fungal pure cultures (10 mm ). (decay regions 4B, 4C, 4D and 7B) in which the directly Due to their low DNA concentration, the undiluted extracts extracted DNA sample was completely different in DGGE were used as such for the PCR-amplifications, thus allowing ca mobility from the culturable strains occupying the same decay 50 reactions from each sample. The PCR-yields were usually region (Fig. 4b, lanes 1–3 and 4–6). In three of these cases (4B, fully sufficient for DGGE analysis (although sometimes slightly 4C, 4D) the cultivated isolates were identified as Trichoderma lower compared to fungal pure cultures), as only one out of sp., and in one case (7B) as Phialemonium sp. Direct analysis of fungal diversity 934

(a) partial SSU rDNA-fragments using a single PCR amplification M1234567891011 12 M step while discriminating against plant and bacterial templates. This eliminates some of the possible bias due to preferential amplification as compared to the commonly used nested protocols including two successive PCR amplifications with separate conserved and group-specific primer sets (Suzuki & Giovannoni 1996, Heuer et al. 1997, Kowalchuk et al. 1997a). Even in one-step PCR-amplifications, however, certain DNA templates may be preferentially amplified leading to different quantities of PCR products from different species in a mixed fungal community (Dutton, Paynton & Sommer 1993, Farrelly, Rainey & Stackebrandt 1995, Vallaeys et al. 1997). In this study, some reduction of relative band intensity was observed in three ascomycetes when their DNA was amplified in the presence of competing templates from other fungi. When using experimental PCR-mixtures from five different (b) species combinations, however, the amounts of amplification M12345678910 products correlated well with the respective template concentrations, and all the species tested could be detected in the presence of tenfold amounts of competing DNA. Furthermore, community patterns could be generated from model mixtures containing eight different fungal species, which is considerably more than is likely to occur within most samples of decaying wood. Although some bias may, therefore, result from preferential PCR amplification, this seems not to be a problem for the majority of species. No within-species variation was observed when multiple isolates were compared from certain species. Additional bands were, however, observed in the amplification products of three species, which might reflect the occurrence of more than Fig. 4. DGGE analysis of SSU rDNA fragments amplified from one SSU rDNA sequence type within single isolates. Previous fungal pure cultures and direct DNA samples isolated from stump studies have also reported intraspecific variation within decay columns. The DNAs were amplified using primer pair introns occurring in unusually long SSU rDNA sequences FR1 NS1 and analysed in 18–43% denaturant gradient gels. Lane j (DePriest & Been 1992, Crockard et al. 1998, Jones & M contains marker DNA (100 bp DNA ladder). (a) Analysis of four different decay regions uniform in fungal composition. Decay regions Blackwell 1998, Perotto & Bonfante 1998). Fortunately, this 1C (lanes 1–3), 6A (lanes 4–6), 3C (lanes 7–9) and 3D (lanes 10–12). kind of within-species variation is relatively rare, and probably Lanes 1, 3, 4, 6, 7, 9, 10 and 12 contain rDNA amplified from fungal will not cause a major overestimation of species diversity. pure cultures, while lanes 2, 5, 8, and 11 show fragments produced As expected (Heuer et al. 1997, Vallaeys et al. 1997), from direct DNA samples extracted from the corresponding decay variation was observed in the DGGE separation capacity region. The following samples were used: lane 1, 1Cc1; 2, 1Cw; 3, achieved using the different primer combinations. A relatively 1Cc1; 4, 6Ac1; 5, 6Aw; 6, 6Ac2; 7, 3Cc1; 8, 3Cw; 9, 3Cc2; 10, high level of resolution was achieved using the shortest 3Dc1; 11, 3Dw; 12, 3Dc2. (b) Analysis of decay regions showing fragment (ca 390 bp), whereas the intermediate-sized frag- different fungal composition using direct extraction of DNA or ments (ca 700 and 1100 bp) were not optimal for the reference cultivation of mycelia. Decay regions 4D (lanes 1–3), 7B (lanes 4–6), collection used in this study. The highest resolution capacity and 7C (lanes 7–10). Lanes 1, 3, 4, 6, 7, and 10 contain rDNA was achieved using long fragments covering almost the entire fragments obtained from fungal pure cultures, while lanes 2, 5 and 9 ca show amplification products from DNA extracted directly from the SSU rDNA ( 1650 bp), which might be surprising due to the corresponding wood samples. Lane 8 contains an experimental PCR known decrease in DGGE resolution of DNA molecules with mixture produced by combining the DNAs from the pure cultures large melted regions (Myers et al. 1985). The DGGE technique shown in lanes 7 and 10 in 1:1 template ratio. The following samples was originally designed for the analysis of point mutations were used: lane 1, 4Dc1; 2, 4Dw; 3, 4Dc2; 4, 7Bc1; 5, 7Bw; 6, 7Bc2; from relatively short DNA fragments (Fischer & Lerman 7, 7Cc2a; 8, 7Cc2aj7Cc1b; 9, 7Cw; 10, 7Cc1b. 1983). When using very long molecules, part of the single base-pair differences certainly may remain undetected, but it seems that in this study the additional variation occurring in DISCUSSION the longer molecules compensated for the reduced resolution The aim of this investigation was to develop a DNA-based capacity. For community analyses, the practical separation technique for studying fungal species profiles directly from potential is more significant than whether each individual base decomposing wood and to compare this methodology with substitution can be detected. conventional cultivation techniques. Four primer sets were Two-year old Norway spruce stumps were used as natural designed that allowed the amplification of fungus-specific test material to examine the fungal species profiles from E. J. Vainio and J. Hantula 935 apparently uniform decay columns usually colonised by single Fischer, S. G. & Lerman, L. S. (1983) DNA fragments differing by single base- fungal individuals (Rayner & Boddy 1988). Some decay pair substitutions are separated in denaturing gradient gels: correspondence Proceedings of the National Academy of Science USA regions showed a completely different fungal composition with melting theory. , 80: 1579–1583. when analysed using mycelial pure cultures or direct extraction Gardes, M. & Bruns, T. D. (1996) Community structure of ectomycorrhizal of DNA. This demonstrated the selectivity of the culture fungi in a Pinus muricata forest: above- and below-ground views. Canadian conditions preferring the growth of certain fungal strains, Journal of Botany 74: 1572–1583. while selecting against other species that could only be Hallenberg, N., Larsson, E. & Mahlapuu, M. (1996) Phylogenetic studies in detected using direct extraction of DNA. Thus, the fungal Peniophora. Mycological Research 100: 179–187. Hantula, J., Dusabenyagasani, M. & Hamelin, R. C. (1996) Random amplified species detectable from the direct DNA samples were most microsatellites (RAMS) – a novel method for characterizing genetic likely dominant in the original wood sample, but were variation within fungi. European Journal of Forest Pathology 26: 159–166. overgrown when cultivated under laboratory conditions. Helgason, T., Daniell, T. J., Husband, R., Fitter, A. H. & Young, J. P. W. (1998) Since some culturable isolates could not be detected from the Ploughing up the wood-wide web? Nature 394: 431. direct DNA samples (although their SSU rDNA could be Heuer, H., Krsek, M., Baker, B., Smalla, K. & Wellington, E. (1997) Analysis of actinomycete communities by specific amplification of genes encoding PCR-amplified from mycelial cultures), they probably 16S rRNA and gel-electrophoretic separation in denaturing gradients. originated from spores or very sparse hyphae. Applied and Environmental Microbiology 63: 3233–3241. Thus, the diversity detectable by the DGGE analysis Hintikka, V. (1993) Occurrence of edible fungi and other macromycetes on depends on the relative abundance of each species in mixed tree stumps over a sixteen-year period. Acta Botanica Fennica 149: 11–17. fungal communities and different views are revealed using Johannesson, H. & Stenlid, J. (1999) Molecular identification of wood- Picea abies Forest Ecology cultivation methods and direct extraction of DNA. DGGE inhabiting fungi in an unmanaged forest in Sweden. and Management 115: 203–211. can, therefore, be used to complement conventional methods Jones, K. G. & Blackwell, M. (1998) Phylogenetic analysis of ambrosial species to obtain a more accurate understanding about fungal in the genus Raffaelea based on 18S rDNA sequences. Mycological Research communities, and to improve the detection of fungal groups 102: 661–665. that are difficult to cultivate under laboratory conditions. Jonsson, L., Dahlberg, A., Nilsson, M.-C., Zackrisson, O. & Ka/ re! n, O. (1999a) Ectomycorrhizal fungal communities in late-successional Swedish boreal forests, and their composition following wildfire. Molecular Ecology 8: 205–215. ACKNOWLEDGEMENTS Jonsson, T., Kokalj, S., Finlay, R. & Erland, S. (1999b) Ectomycorrhizal We acknowledge Mr Esa Pohjolainen, Ms Pa$ ivi Gustafsson and Ms Marja- community structure in a limed spruce forest. Mycological Research 103: Leena Santanen for their technical assistance and Dr Anna-Maija Hallaksela 501–508. for identification of the fungal isolates used as reference strains for the DGGE Kasuga, T., Woods, C., Woodward, S. & Mitchelson, K. (1993) Heterobasidion optimization. The study was financially supported by the Maj and Tor annosum 5.8S ribosomal DNA and internal transcribed spacer sequence: Nessling Foundation (E.J.V.) rapid identification of European intersterility groups by ribosomal DNA restriction polymorphism. Current Genetics 24: 433–436. Kowalchuk, G. A., Gerards, S. & Woldendorp, J. W. (1997a) Detection and characterization of fungal infections of Ammophila arenaria (Marram grass) REFERENCES roots by denaturing gradient gel electrophoresis of specifically amplified 18S rDNA. Applied and Environmental Microbiology 63: 3858–3865. Anderson, J. B. & Stasovski, E. (1992) Molecular phylogeny of Northern Kowalchuk, G. A., Stephen, J. R., De Boer, W., Prosser, J. I., Embley, T. M. & Hemisphere species of Armillaria. Mycologia 84: 505–516. Woldendorp, J. W. (1997b) Analysis of ammonia-oxidizing bacteria of the Berbee, M. L. & Taylor, J. W, (1993) Dating the evolutionary radiations of the β subdivision of the class Proteobacteria in coastal sand dunes by denaturing true fungi. Canadian Journal of Botany 71: 1114–1127. gradient gel electrophoresis and sequencing of PCR-amplified 16S ribosomal Crockard, M. A., Fulton, C. E., Bjourson, A. J. & Brown, A. E. (1998) Intron DNA fragments. Applied and Environmental Microbiology 63: 1489–1497. polymorphism in small subunit rDNA of Nectria galligena. Microbiology Ka/ re! n, O., Ho$ gberg, N., Dahlberg, A., Jonsson, L. & Nylund, J.-E. (1997) Inter- 144: 2367–2372. and intraspecific variation in the ITS region of rDNA of ectomycorrhizal Dahlberg, A., Jonsson, L. & Nylund, J.-E. (1997) Species diversity and fungi in Fennoscandia as detected by endonuclease analysis. New Phytologist distribution of biomass above and below ground among ectomycorrhizal 136: 313–325. fungi in an old-growth Norway spruce forest in south Sweden. Canadian Marsh, T. L., Liu, W-T., Forney, L. J. & Cheng, H. (1998) Beginning a Journal of Botany 75: 1323–1335. molecular analysis of the eukaryal community in activated sludge. Water DePriest, P. T. & Been, M. D. (1992) Numerous group I introns with variable Science and Technology 37: 455–460. distributions in the ribosomal DNA of a lichen fungus. Journal of Molecular Mitchell, J. I., Roberts, P. J. & Moss, S. T. (1995) Sequence of structure? A Biology 228: 315–321. short review on the application of nucleic acid sequence information to Dutton, C. M., Paynton, C. & Sommer, S. (1993) General method for fungal taxonomy. Mycologist 9: 67–75. amplifying regions of very high GjC content. Nucleic Acids Research 21: Mu$ ller, M. M., Kantola, R. & Kitunen, V. (1994) Combining sterol and fatty 2953–2954. acid profiles for the characterization of fungi. Mycological Research 98: Farrelly, V., Rainey, F. A. & Stackebrandt, E. (1995) Effect of genome size and 593–603. rrn copy number on PCR amplification of 16S rRNA genes from a mixture Muyzer, G., De Waal, E. C. & Uitterlinden, A. G. (1993) Profiling of complex of bacterial species. Applied and Environmental Microbiology 61: 2798–2801. microbial populations by denaturing gradient gel electrophoresis analysis Fatehi, J. & Bridge, P. (1998) Detection of multiple rRNA-ITS regions in of polymerase chain reaction-amplified genes coding for 16S rRNA. Applied isolates of Ascochyta. Mycological Research 102: 762–766. and Environmental Microbiology 59: 695–700. Feibelman, T., Bayman, P. & Cibula, W. (1994) Length variation in the internal Myers, R. M., Fischer, S. G., Lerman, L. S. & Maniatis, T. (1985) Nearly all transcribed spacer of ribosomal DNA in chantarelles. Mycological Research single base substitutions in DNA fragments joined to a GC-clamp can be 98: 614–618. detected by denaturing gradient gel electrophoresis. Nucleic Acids Research Ferris, M. J., Muyzer, G. & Ward, D. M. (1996) Denaturing gradient gel 13: 3131–3145. electrophoresis profiles of 16S rRNA-defined populations inhabiting a hot Neuve! glise, C., Brygoo, Y., Vercambre, B. & Riba, G. (1994) Comparative spring microbial mat community. Applied and Environmental Microbiology analysis of molecular and biological characteristics of strains of Beauveria 62: 340–346. brongniartii isolated from insects. Mycological Research 98: 322–328. Direct analysis of fungal diversity 936

Norman, J. E. & Egger, K. N. (1996) Phylogeny of the genus Plicaria and its Smit, E., Leeflang, P., Glandorf, B., Van Elsas, J. D. & Wernars, K. (1999) relationship to Peziza inferred from ribosomal DNA sequence analysis. Analysis of fungal diversity in the wheat rhizosphere by sequencing of Mycologia 88: 986–995. cloned PCR-amplified genes encoding 18S rRNA and temperature gradient O’Donnel, K. (1992) Ribosomal DNA internal transcribed spacers are highly gel electrophoresis. Applied and Environmental Microbiology 65: 2614–2621. divergent in the phytopathogenic ascomycete Fusarium sambucinum Suzuki, M. T. & Giovannoni, S. J. (1996) Bias caused by template annealing in (Gibberella pulicaris). Current Genetics 22: 213–220. the amplification of mixture of 16S rRNA genes by PCR. Applied and Olsen, G. J. & Woese, C. R. (1993) Ribosomal RNA: a key to phylogeny. Environmental Microbiology 62: 625–630. FASEB Journal 7: 113–123. Vallaeys, T., Topp, E., Muyzer, G., Macheret, V., Laguerre, G., Rigaud, A. & Perotto, S. & Bonfante, P. (1998) Genetic and functional diversity of ericoid Soulas, G. (1997) Evaluation of denaturing gradient gel electrophoresis in mycorrhizal fungi. Symbiosis 25: 19–27. the detection of 16S rDNA sequence variation in rhizobia and Rayner, A. D. M. & Boddy, L. (1988) Fungal Decomposition of Wood: Its biology methanotrophs. FEMS Microbiology Ecology 24: 279–285. and Ecology. John Wiley & Sons, Bath. White, T. J., Bruns, T., Lee, S. & Taylor, J. W. (1990) Amplification and direct Renvall, P. (1995) Community structure and dynamics of wood-rotting sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR basidiomycetes on decomposing conifer trunks in northern Finland. Protocols – A guide to methods and applications (M. Innis, D. H. Gelfand, J. J. Karstenia 35: 1–51. Sninsky & T. J. White, eds): 315–322. Academic Press, San Diego. Smalla, K., Wachtendorf, U., Heuer, H., Liu, W.-T. & Forney, L. (1998) Analysis of BIOLOG GN substrate utilization patterns by microbial communities. Applied and Environmental Microbiology 64: 1220–1225. Corresponding Editor: D. A. Wood