APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 1996, p. 3432–3438 Vol. 62, No. 90099-2240/96/$04.0010Copyright q 1996, American Society for Microbiology

Identification of Some Ectomycorrhizal Basidiomycetes by PCRAmplification of Their gpd (Glyceraldehyde-3-Phosphate

Dehydrogenase) GenesNORBERT KREUZINGER,1† RENATE PODEU,1‡ FRANZ GRUBER,1§ FRIEDERIKE GOBL,2

AND CHRISTIAN P. KUBICEK1*

Institut fur Biochemische Technologie und Mikrobiologie, Abteilung fur Mikrobielle Biochemie, Technische UniversitatWien, A-1060 Vienna,1 and Forstliche Bundesversuchsanstalt, A-6020 Innsbruck,2 Austria

Received 15 April 1996/Accepted 9 July 1996

Degenerated oligonucleotide primers designed to flank an approximately 1.2-kb fragment of the geneencoding glyceraldehyde-3-phosphate dehydrogenase (gpd) from ascomycetes and basidiomycetes were used toamplify the corresponding gpd fragments from several species of the ectomycorrhizal fungal taxa Boletus,Amanita, and Lactarius. Those from B. edulis, A. muscaria, and L. deterrimus were cloned and sequenced. Therespective nucleotide sequences of these gene fragments showed a moderate degree of similarity (72 to 76%) inthe protein-encoding regions and only a low degree of similarity in the introns (56 to 66%). Introns, wherepresent, occurred at conserved positions, but the respective positions and numbers of introns in a given taxonvaried. The amplified fragment from a given taxon could be distinguished from that of others by bothrestriction nuclease cleavage analysis and Southern hybridization. A procedure for labeling DNA probes withfluorescein-12-dUTP by PCR was developed. These probes were used in a nonradioactive hybridization assay,with which the gene could be detected in 2 ng of chromosomal DNA of L. deterrimus on slot blots. Taxon-specificamplification was achieved by the design of specific oligonucleotide primers. The application of the gpd genefor the identification of mycorrhizal fungi under field conditions was demonstrated, with Picea abies (spruce)mycorrhizal roots harvested from a northern alpine forest area as well as from a plant-breeding nursery. Theinterference by inhibitory substances, which sometimes occurred in the DNA extracted from the root-fungusmixture, could be overcome by using very diluted concentrations of template DNA for a first round of PCRamplification followed by a second round with nested oligonucleotide primers. We conclude that gpd can beused to detect ectomycorrhizal fungi during symbiotic interaction.

Ectomycorrhiza, a mutualistic plant-fungus symbiosis, playsa pivotal role in the growth and nutrition of forest trees (1, 16,28). The specificity of mycorrhizal interaction varies widely,and it may be responsible for various patterns of efficacy underdifferent ecological conditions. Unfortunately, these fungi arevery difficult to identify in their vegetative state. In addition,they are often difficult to grow in pure culture.With the advent of PCR techniques, several groups have

employed molecular tools to identify ectomycorrhizal fungi bymeans of selective amplification of appropriate genes (such asgenes coding for rRNA) followed by restriction analysis orhybridization (2, 6, 9–11, 13, 14, 20). Mehmann et al. (22)recently discussed an alternative approach by amplifying frag-ments of a gene present only in the fungal partner (i.e., chitinsynthase) followed by restriction analysis.Theoretically, any gene may be used for the identification of

the fungal symbiont of ectomycorrhizae provided that it con-tains conserved regions, suitable for amplification from a widevariety of fungi with the same set of primers, and variable

regions, which would make differentiation by hybridization orrestriction analysis possible. We are interested in the geneticmanipulation of ectomycorrhizal fungi and for this purpose arestudying genes encoding glycolytic enzymes as tools for theconstruction of expression cassettes and transformation vec-tors, since they are often expressed constitutively and at acomparably high level (15). Besides this, however, they fulfillthe criteria given above for optimal probes very well.During our study on the cloning and characterization of

glycolytic genes from ectomycorrhizal fungi, we have thereforeinvestigated whether these genes can be used to detect ecto-mycorrhizal fungi. We will show here that amplification oflarge (1.2-kb) fragments of the glyceraldehyde-3-phosphate de-hydrogenase (gpd) genes from species of the ectomycorrhizalfungal taxa Boletus, Amanita, and Lactarius can be used todetect and distinguish mycorrhizal symbionts on the plant root.

MATERIALS AND METHODS

Selection of ectomycorrhizal basidiomycetes and DNA extraction. Basidiom-ycetous taxa were chosen according to the distribution of ectomycorrhizal specieswithin fungal families in the higher Alpine area according to Moser (24) and aregiven in Table 1. Confirmed specimens of the fungi used are deposited at theInstitute of Microbiology, University of Innsbruck, Innsbruck Austria. To obtainfungal biomass for DNA preparation, pieces of the basidiocarp of individualfungi were grown on Moser 6 agar (23) and covered by a cellophane sheet. Theplates were placed into closed Styropor containers into which a beaker with wetcotton had also been placed to prevent dehydration. DNA was isolated by themethod of Lee and Taylor (19).Extraction of DNA from mycorrhized roots.Mycorrhizal roots were harvested

from soil, soil clumps were removed by shaking, and the roots were then thor-oughly washed with tap water. Once cleaned, they were usually processed within24 h but could be stored at 2708C for at least 12 months if required. To extractDNA, only the root tips (length, 2 to 5 mm) were used. Root tip material (5 to

* Corresponding author. Phone: 43-1-58801-4707. Fax: 43-1-581-62-66. Electronic mail address: ckubicek@fbch.tuwien.ac.at.† Present address: Institut fur Wassergute und Abfallwirtschaft,

Abteilung fur Chemie und Biologie des Wassers, TU Wien, A-1040Vienna, Austria.‡ Present address: Innere Medizin I, Abteilung fur Arbeitsmedizin,

Allgemeines Krankenhaus, Wahringer Gurtel 18-20, A-1090 Vienna,Austria.§ Present address: Immuno AG, Abteilung Gentechnik, A-2304

Orth a/d Donau, Austria.

3432

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from

50 mg [fresh weight]) was placed on filter paper to remove excess water (whichwould otherwise alter the concentration of additives in the following extractionsteps), subsequently placed in Eppendorf vials, and then homogenized in 0.5 mlof CTAB buffer (9) with a motor-driven conical grinder (Treff AG, Degersheim,Switzerland), which fit exactly into the Eppendorf tube (8), until a fine suspen-sion was obtained. All subsequent steps were identical to those described inreference 9.PCR conditions. A DNA region within the open reading frame of the genes

encoding glyceraldehyde-3-phosphate dehydrogenase was amplified. PrimersCTK-052 (sense) and CTK-031rev (antisense), corresponding to sequences in the59- and 39-terminal areas within the gpd region, respectively, were extended by anadditional BamHI, NdeI, or NsiI polylinker sequence to facilitate subsequentcloning (Table 2); they were designed by comparison of the sequences of severalother fungal gpd genes (Table 1) for conserved areas. The reaction mix foramplification contained 10 ml of 103 reaction buffer (Promega, Madison, Wis.),2.5 mM MgCl2, 50 nM (each) deoxynucleoside triphosphate, 0.1 mM primerCTK-052 or CTK-031rev, 0.1 to 1 mg of chromosomal DNA, and 5 or 2.5 U of Taqpolymerase (Promega) or BiTaq polymerase (Biomedica, London, United King-dom) in a final reaction volume of 100 ml (adjusted with bidistilled water). Theamplification reaction was carried out in a thermal cycler (Hybaid, London,United Kingdom) as follows: 8 cycles consisting of denaturation for 90 s at 958C,primer annealing for 30 s at 548C and 90 s at 488C, and extension for 90 s at 728Cfollowed by 30 cycles consisting of 60 s at 958C, 30 s at 548C, 50 s at 488C, and 60 sat 728C and finally a single cycle involving 60 s at 958C, 20 s at 548C, 20 s at 508C,and 10 min at 728C. The holding temperature was 288C. The amplificationproducts were directly subjected to agarose gel electrophoresis.For specific amplification of the gpd fragment from Lactarius deterrimus only,

the Lactarius-specific primers CTK-107 and CTK-108rev were designed (Table2). Amplification was carried out by using the same reaction conditions describedabove but with the following PCR program: 35 cycles consisting of 35 s at 958C,60 s at 628C, and 60 s at 728C followed by a single cycle of 90 s at 958C, 60 s at558C, and 10 min at 728C.To amplify fungal gpd from mycorrhizal roots, a PCR amplification with

primers CTK-052 and CTK-031rev carried out as described above was followedby a second, nested PCR step with 10 ml of the reaction products from the firstamplification (diluted 200-fold with the reaction buffer) and primers CTK-132and CTK-133rev under the following conditions: 10 cycles of 90 s at 958C, 30 s at

558C, 30 s at 528C, and 90 s at 728C followed by 30 cycles of 45 s at 958C, 60 s at548C, and 60 s at 728C and terminated by a single final cycle of 60 s at 958C, 60 sat 548C, 10 min at 728C, and 10 min at 288C. Appropriate controls, including theomission of DNA from the amplification mix, were used for each set of reactionsin order to check for carryover contaminations.Cloning and sequencing. PCR products were separated on low-melting-point

agarose gels, isolated, filled in with Klenow enzyme, and blunt-end cloned intopGEM-5Zf(1) previously linearized with EcoRV (Promega Biotec, Madison,Wis.). DNA manipulations and transformation of Escherichia coli were per-formed by standard methods (29). At least two independently amplified glycer-aldehyde-3-phosphate dehydrogenase gene fragments from each species weresequenced (30) with universal sequencing primers and primers designed accord-ing to obtained sequences. PCR primers and sequence-specific primers for DNAsequencing were synthesized by M. Muller, Vienna Biocenter, Vienna, Austria.Analysis of gpd fragments by nonradioactive hybridization. Two different

methods were used to detect strain-specific gpd fragments on the gels: (i) South-ern hybridization and (ii) slot blot hybridization. For Southern hybridization,PCR products were separated on agarose gels and transferred to nylon mem-branes (Hybaid) according to standardized protocols. Hybridization for 1 h at688C was followed by two 20-min washings at 688C in 43 SSC (13 SSC is 0.15 MNaCl plus 0.015 M sodium citrate) plus 0.1% (wt/vol) sodium dodecyl sulfate(SDS) and in 0.13 SSC plus 0.1% (wt/vol) SDS. For slot blot hybridization, DNAsamples were heated (908C) in a water bath for 3 min, then applied in variousdilutions onto nylon membranes previously soaked in 203 SSC (10 min), andinserted into a slot blot apparatus (Hoefer Scientific Instruments, San Francisco,Calif.). After application, the membranes were dried in air, and the DNA wascovalently linked to the membrane by a 4-min UV treatment. Hybridization wascarried out as described above. Fluorescein-labeled DNA probes were used forhybridization in both cases. They were synthesized by PCR, with 3 to 5 mg ofchromosomal DNA from Boletus edulis, L. deterrimus, or Amanita muscaria asthe template; 50 nM dATP, dGTP, and dCTP; 0.33 nM dTTP; and 0.17 nMfluorescein-12-dUTP. All other PCR conditions were essentially the same asthose described above. After completion of the PCR program, 2 ml of 0.2 MEDTA and 2.5 ml of 1 M LiCl were added, and the DNA was precipitated by theaddition of 75 ml of 96% (wt/vol) ethanol. The DNA was washed with ethanol,dried in vacuo, and finally taken up in 50 ml of Tris-EDTA buffer. This solutionwas used directly for hybridization without further treatment.Immunostaining was used to detect the hybridizing fluorescein-labeled probe

with a monoclonal anti-fluorescein immunoglobulin G coupled to alkaline phos-phatase and 5-bromo-4-chloro-3-indolylphosphate and nitroblue tetrazoliumsalts as chromogenic substrates, according to the manufacturers instructions (5).Nucleotide sequence accession number. The gpd sequences reported in this

paper have been deposited in GenBank under accession no. U30665 (A. mus-caria), U30625 (B. edulis), and U30876 (L. deterrimus).

RESULTS

PCR amplification and restriction fragment length poly-morphism (RFLP) pattern of gpd genes from several ectomy-corrhizal basidiomycetes. PCR was performed with chromo-somal DNA extracted from mycelia or fruiting bodies ofseveral species of the taxa Boletus, Amanita, and Lactarius.Fragments were directly assessed by viewing the amplifiedfragments after gel electrophoresis (Fig. 1a). Among the 11ectomycorrhizal fungi, the amplification products derived fromthe two primers varied from one to four DNA fragments,among which a major band with a size of 1.2 kb was alwayspresent, although sometimes faintly so, along with fragments

TABLE 1. Mycorrhizal isolates investigated in this study

Isolatea Place of origin Yr ofisolation

Amanita muscaria Haggen, Tyrol, Austria 1980Amanita porphyria Matzenkopfl, Tyrol, Austria 1985Amanita rubescens Klausboden, Tyrol, Austria 1974Boletus edulis Schulterberg, Tyrol, Austriab 1990Boletus aereus Borgotora, Italy 1985Boletus erythropus Paida, Tyrol, Austria 1978Boletus calopus Ehrwald, Tyrol, Austria 1987Lactarius deterrimus Schulterberg, Tyrol, Austriab 1990Lactarius deliciosus Mutters, Tyrol, Austria 1983Lactarius porninsis Haggen, Tyrol, Austria 1980Lactarius scrobiculatus Wattener Lizum, Tyrol, Austria 1992

a All isolates are typical symbionts of Picea abies, except B. aereus is a typicalsymbiont of Fagus spp. and Quercus spp. and L. deliciosus and L. porninsis aretypical symbionts of Pinus spp. and Larix spp., respectively.b Habitats within the Achenkirch area.

TABLE 2. Oligonucleotide primers used in the present study

Primera Purpose Structureb

CTK-052 General (59) 59-ATGGATCCATATGCATCGGCCGTATCGTCCTCCGTAATGC-39

CTK-032rev General (39) 59-ATGGATCCATATGCATGAGTAACCGCATTCGTTATCGTACC-39CTK-107 L. deterrimus specific (59) 59-GCTCCTAGACCCTCGTGTCAAGG-39CTK-108rev L. deterrimus specific (39) 59-ATGCCCTTGAGAGGGCCATCGGC-39CTK-132 Nested primer (59) 59-GTCTACATGTTCAAGTACGACTC-39CTK-133rev Nested primer (39) 59-CCGATGAAGTCAGTTGACACTAC-39

a Primers CTK-052 and CTK-031 were designed according to homologous nucleotide areas in the gpd genes of the following organisms (EMBL accession numbersare given in parentheses): Aspergillus nidulans (M19694), Cryphonectria parasitica (X53996), Ustilago mayidis (X07879), Agaricus bisporus 1 (M 81727), A. bisporus 2 (M81728), Phanerochaete chrysosporium (M81754), and Schizophyllum commune (M81724). The nucleotide areas used to design the remaining four primers are shown inFig. 2.b The positions of the restriction cleavage sites on the termini of primers CTK-052 and CTK-031rev are indicated by underlining (BamHI), double underlining (NsiI),

and overlining (NdeI).

VOL. 62, 1996 gpd GENE FROM ECTOMYCORRHIZAL FUNGI 3433

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from

with different sizes. The size of this 1.2-kb nucleotide bandcorresponded perfectly with the expected size of the fragmentcalculated from the corresponding sequences of A. bisporus, P.chrysosporium, and Schizophyllum communes (12). Proof of theidentity of this nucleotide band with a gpd fragment was ob-tained by sequencing (see below).The amplified, putative gpd fragments from three selected

species corresponding to the three taxa, B. edulis, A. muscaria,and L. deterrimus, were eluted from the gel, subcloned, andsubjected to RFLP analysis with BamHI, NdeI, and NsiI, re-spectively. These species were chosen because they representmycorrhizal symbionts of Picea abies, important in the Austriannorthern alpine forest areas (11a, 25a). The three restrictionnucleases allowed a distinction between these three fungi (Fig.1b), hence suggesting that RFLP analysis of gpd may allowdistinction between the three basidiomycetous species.Analysis of putative gpd fragments by sequencing. To prove

the identity of the 1.2-kb DNA fragments from B. edulis, A.muscaria, and L. deterrimus with the gene encoding glyceral-dehyde-3-phosphate-dehydrogenase, the DNA sequences ofthe amplified fragments were determined (Fig. 2). Multiplealignments of these nucleotide sequences with known genesencoding glyceraldehyde-3-phosphate dehydrogenase (12, 18,25, 26, 31, 32) suggested the presence of various introns (Fig.3). All introns were confirmed by identifying the consensusexon/intron splice junction sequences seen in other gpd genes.The relative positions and length variations of the introns dif-fered in different fungal classes and families and may thus bepotential taxonomic markers. As observed for other fungalgenes as well (4), the basidiomycetous gpd genes contained alarger number of introns (four to nine) than did ascomycetousgpd genes (two to six). Interestingly, L. deterrimus gpd exhibitedthe same intron pattern as Agaricus bisporus, whereas the B.

edulis gpd gene lacked one intron present at nucleotides 647and 587 in the A. muscaria and L. deterrimus gpd genes, re-spectively.The nucleotide sequence analysis of the three cloned genes

encoding glyceraldehyde-3-phosphate dehydrogenase showedabout 72 to 76% similarity at the nucleotide level when onlythe exons were compared. The degree of similarity was partic-ularly high in those areas encoding the hallmark amino acidsequences required by the enzyme for catalysis, whereas otherareas (e.g., the 59 and 39 ends of the amplified fragment)showed very low degrees of similarity. The introns showed leastsimilarity (56 to 66%). In the most conserved areas, mostnucleotide alterations were observed in the third position butdid not alter the encoded amino acid. Consequently, the en-coded proteins showed higher degrees of similarity than didthe genes.Use of isolated gpd fragments for specific detection of my-

corrhizal fungi by hybridization. The roughly 25% differencesin the nucleotide sequences of gpd between L. deterrimus, B.edulis, and A. muscaria suggested that stringent hybridizationconditions could be used to specifically detect one of thesegenes in mycorrhizal roots. In order to develop this method forroutine analysis in an average laboratory, a nonradioactiveassay was also adopted. Under carefully optimized high-strin-gency conditions (see Materials and Methods), an L. deterri-mus fluorescein-12-dUTP-labeled gpd fragment hybridized bySouthern blotting to its own amplification products only (Fig.4). Similarly, B. edulis- and A. muscaria-derived gpd gene frag-ments were successfully used to detect amplification productsof these species from DNA mixtures (data not shown). It wasthereby noted that the gpd probes older than 5 days (irrespec-tively of how they were stored) had to be avoided, since oth-erwise-unspecific hybridization was observed. While the reasonfor this remains to be studied, the data show that with freshlyprepared probes the use of the gpd gene allows the discrimi-nation between mycorrhizal fungi at the DNA level.To examine the sensitivity of hybridization with the fluores-

cein-labeled probe, serial dilutions of L. deterrimus DNA werecarried out. In order to eliminate electrophoresis and Southerntransfer, slot blot analysis was used for this purpose. A stan-dard calibration curve obtained with a thin-layer chromatog-raphy scanner (Fig. 5) showed that the quantification of theblotted DNA was possible with 1 to 10 ng of applied material.Because of the high factor of amplification by PCR and thepresence of 1% DNA per fungal biomass (dry weight), eventraces of fungal mycelium can be detected by this method.When a comparison of the sensitivities of radioactive (32P)

and nonradioactive hybridization was made, the detection lim-its of both probes were at the same relative DNA dilution (10to 15 ng per blot) (data not shown). We therefore concludethat the use of the nonradioactive hybridization procedure wasnot accompanied by a loss in the sensitivity of the analysis.Detection of L. deterrimus by species-specific amplification.

The presence of several nucleotide areas in the basidiomycet-ous gpd gene which shared little similarity between differentspecies suggested that the gpd gene may also be used fordiagnosis by species-specific amplification. To test this possi-bility, primers CKT107 and CKT108rev which corresponded toa 59 and a 39 nucleotide sequence of L. deterrimus, were de-signed. These primers allowed the amplification of the corre-sponding gpd fragment from a Lactarius sp. but not from var-ious other basidiomycetes and ascomycetes (Fig. 6). Theaccumulation of smaller amounts of the respective gpd frag-ment from Lactarius scrobiculatus than from L. deterrimus sug-gests that the amplification conditions may be further opti-mized to distinguish below the generic level.

FIG. 1. (a) Amplification of gpd fragment by PCR, with primers CTK-031revand CTK-052, from chromosomal DNA of several mycorrhizal basidiomycetousspecies. Lanes: L, a 100-bp ladder; 1, A. muscaria; 2, A. porphyrea; 3, A. rubescens;4, B. aereus; 5, B. calopus; 6, B. edulis; 7, B. edulis; 8, L. deliciosus; 9, L. deterrimus;10, L. porninsis; 11, L. scrobiculatus. (b to d) Restriction cleavage analysis of gpdfragments of A. muscaria, B. edulis, and L. deterrimus, respectively, after cloninginto the EcoRV site of vector pGEM5z1. Lanes: M, 100-bp ladder; 1, intactpGEM5z1; 2, SalI-treated pGEM5Z1; 3, intact pGEM5Z1 containing therespective gpd fragments; 4 to 7, products of cleavage with BamHI, NdeI, NsiI,and SalI, respectively.

3434 KREUZINGER ET AL. APPL. ENVIRON. MICROBIOL.

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from

Similar taxon-specific amplification was observed with prim-ers designed specifically for A. muscaria and B. edulis (data notshown).Amplification of fungal gpd gene fragments from mycorrhi-

zal root tips. Following the protocol given in Materials andMethods, we have attempted to identify the fungal gpd genesselectively and species specifically from mycorrhizal roots.Negative results (e.g., lack of amplification), which occurred ininitial trials, were subsequently shown to be due to (i) the verylow percentage of fungal DNA in the total DNA mixture and(ii) the presence of unknown substances that affected the PCR,which rendered the use of higher concentrations of total DNAimpracticable. To bypass this problem, PCR amplification wascarried out in two subsequent steps: first with very dilutedDNA concentrations and secondly by a nested PCR step toincrease the concentration of the amplified fungal gene frag-ments. To this end, appropriate primers (Table 2) which cor-responded to sequences located approximately 39 and 59 of thesense and antisense primers, respectively, used in the first PCRamplification were designed. Figure 7A shows that this proce-dure yielded the expected 1.2-kb gpd fragment in all but onesample. Controls with only one instead of both primers yieldedno amplicons, demonstrating the specificity of the nested PCR.In order to identify the fungal partners in these symbioses, the

FIG. 2. Nucleotide sequence alignment of PCR-amplified gpd fragments of B. edulis, A. muscaria, and L. deterrimus. Coding sequences are given in capital letters.Dashes indicate identical nucleotides; asterisks indicates gaps. The sequences of the general PCR primers, CKT-052 and CKT-031rev, are not included in the givensequence and occur immediately adjacent to the 59 and 39 ends of the sequence. The nucleotide areas used to design the L. deterrimus-specific primers CKT-107 andCKT-108rev are indicated by dotted underlining. The areas used to design the nested primers CKT-132 and CKT-133rev, respectively, are boxed.

FIG. 3. Comparison of intron positions in fungal gpd fragments. Horizontallines represent the coding regions of each fragment. The vertical bars mark thepositions of introns. Bold lines indicate those genes sequenced in the presentstudy. Data on fungi not studied in this paper were taken from references 12, 18,25, 26, 31, and 32.

VOL. 62, 1996 gpd GENE FROM ECTOMYCORRHIZAL FUNGI 3435

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from

amplification products were treated with DdeI (Fig. 7B). ByRFLP analysis it became evident that most mycorrhizae inves-tigated apparently contained L. deterrimus as the fungal part-ner. In order to provide further evidence for this, the fragmentfrom sample 7 was labeled by nick translation and used as aprobe for high-stringency Southern hybridization of a blot con-taining genomic DNA of various Lactarius, Boletus, and Aman-ita spp. (Fig. 7C). Exclusive hybridization was achieved withonly L. deterrimus DNA. Consistent results were further ob-tained by microscopical examination of the roots from whichsample 7 had been extracted (data not given). Hence, our datashow that by amplification of the gpd fragment and subsequentanalysis the fungal partner in ectomycorrhizal symbiosis can beidentified.

DISCUSSION

In this paper, we have shown that PCR amplification of afragment of the gpd gene can be used to specifically detectmycorrhizal fungi by various methods such as RFLP analysis,Southern or slot blot hybridization, and taxon-specific ampli-fication. Although unlike the gene cluster coding for rRNA,fungal gpd genes are present in single, or double (12), copies inthe genome only. The use of a second, nested PCR amplifica-tion step eliminated the problem of sensitivity and showed thatthe gpd gene can be used for analyzing DNA samples from themycorrhizal roots. The use of gpd genes for taxonomic pur-poses has been criticized because of evidence that this gene isvariable at the species level and because of evidence for thepresence of pseudogenes and horizontal gene transfer (7).These drawbacks are unlikely to influence the methods pro-posed in this paper as long as no attempts are made to use thepresent results for taxonomic conclusion. However, intraspe-cific variation, as has been demonstrated in Agarius bisporus(12), may produce a problem, and the present method is there-fore recommended only for the identification of species in

which the variation of gpd is known, rather than for the typingof unknown isolates.The success in using the gpd gene for identification of the

fungal partners in mycorrhizal symbiosis complements the

FIG. 4. (A) Agarose gel electrophoresis of PCR products from chromosomalDNA of fungi used as the template (1 mg [total concentration] per organism). (B)Southern hybridization analysis of PCR-generated gpd fragment with a PCR-generated fluorescein-dUTP-labeled L. deterrimus gpd gene as the probe. Lanes:1, L. porninsis; 2, B. edulis; 3, A. muscaria; 4, L. deterrimus; 5, L. scrobiculatus; 6,B. erythropus; 7, all six fungi; 8, all fungi cited except L. deterrimus.

FIG. 5. (A) Slot blot analysis with PCR-generated fluorescein-dUTP-labeledprobe of 1.2-kb gpd fragment amplified from L. deterrimus. Rows a and b areduplicates of the sample. The amount (in micrograms) of L. deterrimus DNAused is indicated above each lane. Lanes A and B contain DNA extracted fromdifferent samples of L. deterrimus. (B) Calibration curve drawn from scanning theslots in panel A and plotting the peak area (in arbitrary units [a.u.]) versus theapplied concentration. The dotted line shows the continuation of the pseudo-linear part of the graph, which can be used as a calibration curve.

FIG. 6. Taxon-specific amplification of a gpd fragment with L. deterrimus-specific primers. DNA from the indicated organisms and tissues was subjected toPCR amplification, and the putative products were separated by electrophoresison 1.2% agarose. Lanes: 1, A. muscaria mycelium; 2, B. edulis mycelium; 3, L.deterrimusmycelium; 4, Saccharomyces cerevisiae; 5, Trichoderma reesei; 6 and 10,A. muscaria basidiocarp; 7 and 11, B. edulis basidiocarp; 8 and 12, L. deterrimusbasidiocarp; 9, L. scrobiculatus basidiocarp; 10 to 12, fungal DNA after 3 monthsof storage at 2208C.

3436 KREUZINGER ET AL. APPL. ENVIRON. MICROBIOL.

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from

tools already available (i.e., fragments of the gene cluster cod-ing for rRNA [6, 9, 14] and genes encoding chitin synthase[22]), thereby offering the possibility to cross-check resultsfrom an analysis. The presentation of a nonradioactive methodfor hybridization will further aid in the use of molecular toolsfor mycorrhizal identification, as the whole procedure can nowbe carried out on an average laboratory bench and does notrequire sophisticated equipment.Because of the availability of the gpd gene fragments from

the three taxa of mycorrhizal basidiomycetes, the straightfor-ward isolation of full-length clones of the corresponding geneswill be possible. These clones, in turn, will enable us to use therespective expression signals to construct gene expression cas-settes for these organisms, e.g., to construct appropriate plas-mids for transformant selection. Genetic transformation ofsome mycorrhizal fungi has already been reported (3, 21),although not for the three taxa investigated here. Once condi-tions for these taxa are available, the gpd gene expressionsignals may also be used to introduce reporter genes (e.g., uidA[17]), which will facilitate the investigation of the symbioticinteraction.

ACKNOWLEDGMENTS

We thank F. Herman and J. Camba for their interest and supportand J. G. H. Wessels for disclosing the sequences of other basidiomy-cetes prior to their publication (12).This work was supported by grant 56.810/39-VA2/91 from the Aus-

trian Federal Ministry of Agriculture and Forestry.

REFERENCES

1. Agerer, R. 1993. Mycorrhizae: ectomycorrhiza and ectendomycorrhiza. Prog.Bot. 54:505–529.

2. Armstrong, J. L., N. L. Fowles, and P. T. Rygiewicz. 1989. Restriction frag-

ment length polymorphisms distinguish ectomycorrhizal fungi. Plant Soil116:1–7.

3. Barrett, V., R. K. Dixon, and P. A. Lemke. 1990. Protoplast formation fromselected species of ectomycorrhizal fungi. Appl. Microbiol. Biotechnol. 33:313–316.

4. Birch, P. R. J., P. F. G. Sims, and P. Broda. 1992. Nucleotide sequence of agene from Phanerochaete chrysosporium that shows similarity to the facAgene of Aspergillus nidulans. DNA Sequence 2:319–323.

5. Boehringer Mannheim. 1989. Biochemica—applications manual: DNA la-beling and nonradioactive detection. Boehringer Mannheim GmbH Bio-chemica, Mannheim, Germany.

6. Bruns, T. D., and M. Gardes. 1993. Molecular tools for identification ofectomycorrhizal fungi: taxon-specific oligonucleotide probes for sulloidfungi. Mol. Ecol. 2:233–242.

7. Bruns, T. D., T. J. White, and J. W. Taylor. 1991. Fungal molecular system-atics. Annu. Rev. Ecol. Syst. 22:525–564.

8. Cenis, J. L. 1992. Rapid extraction of fungal DNA for PCR amplification.Nucleic Acids Res. 20:2380.

9. Gardes, M., and T. D. Bruns. 1993. ITS primers with enhanced specificity forbasidiomycetes—application to the identification of mycorrhizae and rusts.Mol. Ecol. 2:113–118.

10. Gardes, M., J. A. Fortin, G. M. Mueller, and B. R. Kropp. 1990. Restrictionfragment length polymorphism in the nuclear ribosomal DNA of four Lac-caria spp.: L. bicolor, L. laccata, L. proxima, and L. amethystina. Phytopa-thology 80:1312–1317.

11. Gardes, M., T. J. White, J. A. Fortin, T. D. Bruns, and J. W. Taylor. 1991.Identification of indigenous and introduced symbiotic fungi in ectomycor-rhizae by amplification of nuclear and mitochondrial ribosomal DNA. Can.J. Bot. 69:180–190.

11a.Gobl, F. Unpublished findings.12. Harmsen, M. C., F. H. J. Schuren, S. M. Mouhka, C. M. van Zuilen, P. J.

Punt, and J. G. H. Wessels. 1992. Sequence analysis of the glyceraldehyde-3-phosphate dehydrogenase genes from the basidiomycetes Schizophyllumcommune, Phanerochaete chrysosporium, and Agaricus bisporus. Curr. Genet.22:447–454.

13. Henrion, B., C. Di Battista, D. Bouchard, D. Vairelles, B. D. Thompson, F.Le Tacon, and F. Martin. 1994. Monitoring the persistence of Laccariabicolor as an ectomycorrhizal symbiont of nursery grown Douglas fir by PCRof the rDNA intergenic spacer. Mol. Ecol. 3:571–580.

14. Henrion, B., F. Le Tacon, and F. Martin. 1992. Rapid identification of

FIG. 7. Detection of basidiomycetous species in mycorrhizae by nested PCR (A), restriction analysis (B), and Southern hybridization (C). DNA was extracted fromthe root tips and subjected to PCR amplification with primers CKT-052 and CKT-031rev as described in Materials and Methods. The amplification products were thendiluted, used for a nested PCR with primers CKT-132 and CKT-133rev, and subjected to agarose gel electrophoresis. Lanes: 4 to 9, various samples collected throughoutOctober and November 1993, in Schulterberg, Tyrol, Austria; 1 to 3, chromosomal DNA from B. edulis, L. deterrimus, and A. muscaria, respectively. (B) Restrictionanalysis of nested PCR products described for panel A with DdeI. Numbering of lanes is identical to that for panel A. (C) Results of Southern hybridization ofmycorrhizal samples with the amplified, random-primed fragment from sample 7. Lanes 2 to 4 indicate controls with chromosomal DNA from B. edulis, L. deterrimus,and A. muscaria (identical to lanes 1 to 3 in panel A), respectively, whereas the other lanes contain amplification products from mycorrhizal samples isolated inSchulterberg, Tyrol, Austria.

VOL. 62, 1996 gpd GENE FROM ECTOMYCORRHIZAL FUNGI 3437

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from

genetic variation of ectomycorrhizal fungi by amplification of ribosomalRNA genes. New Phytol. 122:289–298.

15. Holland, M. J., and P. J. Holland. 1980. Structural comparison of twonontandemly repeated yeast glyceraldehyde-3-phosphate dehydrogenasegenes. J. Biol. Chem. 255:2596–2605.

16. Jasper, D. A. 1994. Bioremediation of agricultural and forestry soils withsymbiotic microorganisms. Aust. J. Soil Sci. 32:1301–1319.

17. Jefferson, R. A., T. A. Kavanagh, and M. W. Bevan. 1987. GUS fusions:b-glucuronidase as a sensitive and versatile fusion marker in higher plants.EMBO J. 6:3901–3902.

18. Jungehulsing, U., C. Arntz, R. Smit, and P. Tudzynski. 1994. The Clavicepspurpurea glyceraldehyde-3-phosphate dehydrogenase gene: cloning, charac-terization and use for the improvement of a dominant selection system. Curr.Genet. 25:101–106.

19. Lee, S. B., and J. W. Taylor. 1992. Phylogeny of five fungus-like protocistanPhytophthora species, inferred from the internal transcribed spacers of ribo-somal DNA. Mol. Biol. Evol. 9:636–653.

20. Marmeisse, R., J. C. Debaud, and L. A. Casselton. 1992. DNA probes forspecies and strain identification in the ectomycorrhizal fungus Hebeloma.Mycol. Res. 96:161–165.

21. Marmeisse, R., G. Gay, J.-C. Debaud, and L. A. Casselton. 1992. Genetictransformation of the symbiotic basidiomycetes Hebeloma cylindrosporum.Curr. Genet. 22:41–45.

22. Mehmann, B., I. Brunner, and G. H. Braus. 1994. Nucleotide sequencevariation of chitin synthase genes among mycorrhizal fungi and its potentialuse in taxonomy. Appl. Environ. Microbiol. 60:3105–3111.

23. Moser, M. 1963. Forderung der Mykorrhizenbildung in der forstlichenPraxis. Mitt. Forstl. Bundes-versuchsanst. Wien 60:691–720.

24. Moser, M. 1984. Kleine Kryptogamenflora IIb/2. Gustav Fischer Verlag,Stuttgart, Germany.

25. Osiewacz, H. D., and R. Ridder. 1991. Genome analysis of imperfect fungi:electrophoretic karyotyping, and characterization of the nuclear gene codingfor glyceraldehyde-3-phosphate dehydrogenase (gpd) of Curvularia lunata.Curr. Genet. 20:151–155.

25a.Peintner, U., and M. Moser. 1995. Artenvielfalt und Abundanz von Basid-iomyzeten im Projektgebiet Achenkirch. FBVA-Berichte 87:69–94.

26. Punt, P. J., M. A. Dingemanse, B. J. M. Jabos-Meijising, P. H. Pouwels, andC. A. M. J. J. Van den Hondel. 1988. Isolation and characterization of theglyceraldehyde-3-phosphate dehydrogenase gene of Aspergillus nidulans.Gene 69:49–57.

27. Rossman, M. G., A. Liljas, C.-I. Branden, and L. J. Banaszak. 1975. Evolu-tionary and structural relationships among dehydrogenases, p. 61–102. InP. D. Boyer (ed.), The enzymes. Academic Press, New York.

28. Rygiewicz, P. T., and C. P. Andersen. 1994. Mycorrhizae alter quality andquantity of carbon allocated below ground. Nature (London) 369:58–60.

29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

30. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing withchain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467.

31. Templeton, M. D., E. H. A. Rikkerink, S. L. Solon, and R. N. Crowhurst.1992. Cloning and molecular characterization of the glyceraldehyde-3-phos-phate dehydrogenase-encoding gene and cDNA from the plant pathogenicfungus Glomerella cingulata. Gene 122:225–230.

32. Van Wert, S. L., and O. C. Yoder. 1992. Structure of the Cochliobolusheterostrophus glyceraldehyde-3-phosphate dehydrogenase gene. Curr.Genet. 22:29–35.

3438 KREUZINGER ET AL. APPL. ENVIRON. MICROBIOL.

on May 24, 2021 by guest

http://aem.asm

.org/D

ownloaded from