‘Molecular Ecology (1993) 2, 113-118 ITS primers with enhanced specificity for basidiomycetes — application to the identification of mycorrhizae and rusts M. GARDES and T. D. BRUNS Department of Plant Pathology, University of California, Berkeley, CA 94720, USA Abstract We have designed two taxon-selective primers for the internal transcribed spacer (ITS) region in the nuclear ribosomal repeat unit. These primers, ITS1-F and ITS4-B, were intended to be specific to fungi and basidiomycetes, respectively. We have tested the specificity of these primers against 13 species of ascomycetes, 14 of basidiomycetes, and 150f plants. Our results showed that ITS4-B, when paired with eithera ‘universal’ primer ITS1 or the fungal-specific primer ITS1-F, efficiently amplified DNA from all basidiomycetes and discriminated against ascomycete DNAs. The results with plants were not as clearcut. The ITS1-F/ITS4-B primer pair produced a small amount of PCR product for certain plant species, but the quantity was in most cases less than that produced by the ‘universal’ ITS primers. However, under conditions where both plant and fungal DNAs were present, the fungal DNA was amplified to the apparentexclusion of plant DNA. ITS1-F/ITS4-B preferential amplification was shown to be particularly useful for detection and analysis of the basidiomycete component in ectomycorrhizae and in rust-infected tissues. These primers can be used to study the structure of ectomycorrhizal communities or the distribution of rusts on alternate hosts. Keywords: fungal ecology, PCR, taxon-specific primer Received 19 November 1992; revision accepted 4 January 1993 Introduction The polymerase chain reaction (Mullis & Faloona 1987) offers the opportunity to characterize fungal symbionts by amplification of specific sequences. Identification of these symbionts can be achieved by analysis of PCR amplified products by RFLPs, sequencing, or oligonucleotide prob- ing (Gardes etal. 1991). However, in order for most of these analyses to be performed, fungal DNA must often be spe- cifically amplified from mixtures of plant and fungal DNAs. The internal transcribed spacer (ITS) region contains two variable non-coding regions that are nested within the 1DNA repeat between the highly conserved small subunit, 5.85, and large subunit rRNA genes (Fig. 1). Several fea- tures make it a convenient target region for molecular identification of fungi: (i) in fungi, the entire ITS region is often between 600 and 800 bp and can be readily amplified with ‘universal primers’ that are complementary to se- quences within the rRNA genes (White et al. 1990), (ii) the multicopy nature of the rDNA repeat makes the ITS region easy to amplify from small, dilute, or highly degraded DNA samples, and (iii) several studies have demonstrated that the ITS region is often highly variableamong morpho- logically distinct fungal species (Gardeset al. 1991; Gardes & Bruns 1991; Baura et al. 1992; Chen et al. 1992; Lee & Taylor 1992) but, with one known exception (O'Donnell 1992), intraspecific variation is low (Gardes et al. 1991; Gardes & Bruns 1991; Anderson & Stasovski 1992; Baura et al, 1992; Chen et al. 1992; Lee & Taylor 1992) One potential disadvantage of using the ITS region for identification of fungal symbionts is that currently avail- able ITS primers were designed to amplify a broad range of organisms including fungi, plants, protists and animals (Whiteet al, 1990). In many natural situations, fungal DNA may be rare in comparison to plant host DNA, and thus specific or preferential amplification of fungal DNA would be desirable.114 M. GARDES & T. D. BRUNS. In this study, we have designed and tested two new primers that in combination preferentially amplify the ITS region of basidiomycetes. Our focus on basidiomycetes has two immediate ecological applications: the identifica- tion of ectomycorrhizal fungi and rust fungi directly from infected plant material. In the former case, ectomycor- thizal fungi are crucial root symbionts of trees and shrubs in many terrestrial ecosystems (Harley & Smith 1983; Kropp & Langlois 1990; Read 1991). Basidiomycetes repre- sent the overwhelming majority of the ectomycorrhizal fungal species, yet little is known of how guilds are struc- tured because most of the ectomycorrhizal fungi are not identifiable either as mycorthizae (the symbiotic root- fungus structure) or from pure cultures. Virtually all eco- logical studies of ectomycorthizae either use an ecosystem approach to place entire complex guilds, containing scores of unrelated fungi, into a single black-box, or infer species, ‘composition from fruiting records. However, recent stud- ies have shown that fruit-body appearance may not be a good indicator of mycorrhizae composition (Gibson & Deacon 1988; Danielson & Pruden 1989; Dahlberg 1991). In the case of rust fungi, they represent one of the largest group of obligately parasitic plant pathogens. They fre- quently produce five morphologically distinct spore stages and alternate between two unrelated plant hosts. In ‘many cases identification of particular rusts on one of the two hosts is difficult or impossible by morphology alone. Thus, like the ectomycorrhizal fungi, difficulties with identification have hampered our understanding of the basic autoecology of many rusts species. Material and methods DNA extraction DNAs from freeze-dried tissues were extracted as follows: 300 ul of 2x CTAB lysis buffer (100-mw Tris-HCI (pH 8.0), 1.4-m NaCl, 20-mm EDTA, 2% CTAB, 0.2% f-mercapto- ethanol ) was added to 1.5-ml centrifuge tubes containing fresh or dried tissues. Tissue samples consisted of one or two mycorrhizal root tips, or 5-20 mg of mycelium, carpophore, or plant tissue. The buffer-suspended sam- ples were frozen and thawed three times by passing the tubes between a dry-ice-ethanol bath and a 65°C heat block. After the final thaw, the tissue was crushed with a micropestle (Kontes) and incubated at 65°C for 30 min to 1h, One volume of chloroform was added to each sample and mixed by briefly vortexing. The samples were then centrifuged for 15 min at room temperature. The aquous phase was removed from each tube and transferred to a new centrifuge tube. DNA was precipitated from it with an equal volume of cold isopropanol. After 10 min the pre- cipitant was pelleted by a 10-min centifugation at full speed. The supernatant was discarded and the pellet was washed with a 70% ice-cold ethanol solution before resuspension in 40-60 jil of 0.1x TE (1-ma Tris-Hel,0.1-m EDTA (pH8.0)). Prior to amplification, each DNA sample was diluted 200- to 1000-fold in sterile double-distilled water. Dilution was found to be necessary to prevent inhi- bition of the amplification reaction. With this extraction procedure we could process up to 150 samples in a single day and obtain DNAs that amplified reliably. Pinus ponderosa and Suillus pungens DNA concentra- tions were determined using a DNA fluorometer model TKO 100 (Hoefer Scientific Instruments), updated for measurements with capillary tubes. PCR amplification and analysis Aliquots of 12.5 ll of the diluted DNAs were combined with an equal volume of PCR mix containing buffer, nucleotide triphosphates, and Taq polymerase. The final concentrations of the components were: 200 um each of ATP, dCTP, dGTP and dTTP, 50-mv KCl, 10-mM Tris- HCI (pH 83), 0.1-mg/ml gelatin, 05 units of Tag DNA polymerase per 25-l reaction (Boehringer Mannheim) and 1 us of each of the two primers used in a particular reaction. Fach reaction was overlaid with 1 drop of min- eral oil (Sigma). Temperature cycling was carried out us- ing a programmable heat block (DNA Thermal Cycler, ‘Techne model PHC-2). An initial denaturation step of 94°C for 85s was followed by 35 amplification cycles of de- naturation, annealing, and extension. The temperature and times for these steps in the first 13 cycles were 95°C for 35 s, 55°C for 55 s, and 72°C for 45 s, Cycles 14-26 and 27— 35 used the same parameters except that the extension steps were lengthened to 120 and 180 s, within the two set of cycles respectively. After the 35 cycles were completed, the samples were incubated an additional 10 min at 72°C. Negative controls (no DNA template) were used in every experiment to test for the presence of DNA contamination of reagents and reaction mixtures. The two universal ITS primers used, ITS1 and ITS4, have been described previously (White et al. 1990; Gardes et al. 1991). We designed two taxon-selective primers for the ITS region, ITSI-F and ITS4-B. These were intended to be specific for the higher fungi and basidiomycetes, re- spectively. We designed these primers by comparing fun- gal and plant sequences in the 3’ end of the 18S and the 5" end of the 285 (Fig, 1). The sequences of basidiomycetes are from unpublished work of T. D. Bruns and T. M.Szaro (Department of Plant Pathology, U.C. Berkeley). All other sequences were obtained from GENBANK. When we de- signed our primers, a Zamia pumila sequence was the only sequence available for gymnosperms in the 185 while a Saccharomyces cerevisiae sequence was the only sequence available for Ascomycetes in the 23S. Because mismatches at the 3' end are often most critical for efficient amplifica-ITS PRIMERS FOR HIGHER FUNGI AND BASIDIOMYCETES 115, iis CTTGGTCATTTAGAGGAAGTAA Cronartium riticla Pasillusatrotomentosus ‘Suillus 4p. 8 others Boletus satanas& others [Saccharomyces cerevisiae Fig. 1 Design of primers. The approximate loca- tion of the ITS primers is shown along with the known sequences that were used to design the taxon-selective primers. The two primer sequences Besidomyeetes Neurospors crassa [sscomyootes are listed in the standard 5' to 3' direction Zama pumila 6.4 Giveine ep 26.4) pants loeyea sativa Gc 2eamayais cal is 1154-8 Primer ‘aconcTTarACACGGTCCAG Buccs saames Mort Table 1 Intensities of the amplified ITS fragment from various Basidiomycetes |Gomphidus glutinosus A organisms using four different sets of primers Sus snuspa ars mo Paxillus atrotomentue Ascomyostes | Sacenarmyces corensiae AGA AAG AINE sie 11Ste ISHFS TSIEN pnts lOyea sativa wee a ist stb 1734 Toa NNS tion Gommer & Tautz 1989, Kwok eal, 1990), we tried to Plants , w place mismatches between target and non-target se- ABS www quences at the3'end of the primerand mismatchesamong Arctostaphylos sp. YO ® 2 ow intended target sequences in the central or 5'ends of the Lait laicina + of oF oH ‘ Monotropa hypopthys + oo @ 2 primer (Fig. 1). Pinus contorta - 2 wm & oH Restriction digests of the amplified DNAs were per-—P.lambertanut - 2 © WH formed on unpurified PCR products by diluting 8 il of the P. monticola + a w ey + completed PCR reaction 1:1 intoa solution oftwofold con- fy Mil 2 wD we centrated restriction enzyme buffer (supplied by manufac. raiata + @ & & 4 turer) and 1-3 units of Alul, Hinfl or Rsal enzyme (BRL, Poeudotsuga Sa w Be G) ey ” Boehringer Mannheim). The reactions were then incu- Ure Asrie ee bated for 1h at the optimal temperature for each enzyme. Vacrinium parifolium +O) 2 2 PCR products and their restriction digests were electrophoresed for1-4hon T2-cmx12-cmx05-cmgelsof —ASOmYENe® mS 5 6 6 2% Nu-Sieve agarose (FMC) with 1% non-modified Cenococcum geopilum 166 s+ (4) ss) agarose in Tris-acetate buffer (100-mw tris, 125m Coote geoph 349 se = ne NaAcetate, I-mm EDTA (pH 8.1)), The gels were stained pryaicis gtr’ pe 7p with 1-g/ml ethidium bromide for 10-15 min, destained Estrain R947 Hoo ORS with water for 15-30 min, and photographed under ultra-_Evemnscus albus eR f violet ight Monascus purpureus elon f [Nematospora cory Holo Sporothrx schenkii elo w@ an Tuber melanogaster Holo - Results and discussion Wileoxina mikolae Bolt Wileoxina rohit es @ Hw We tested the ability of four primer pairs: ITSI-ITS4, TIS1-ITS4-B, ITSI-F-ITS4, and ITSI-F-ITS4-B to amplify _ Basidiomycetes DNA from various plant, basidiomycete and ascomycete gym imnciel ETE Tt species. The ITS region was amplified differentially from Cortinarius violaceus + Oe ROR these samples (Table 1). Some of these differences corre-__Hiellum pect, too + lated well with the intended specificity of the primers. For Moabe sp. Por ron example, all attempted amplifications of ascomycetous Melanogster fuberformis, =} te aha DNA that included ITS4-B, the basidiomycete primer, re- Paxillus aa + + the sulted ineither no product oran extremely fant product Petri mamas 94294 (Table 1, Fig, 2), while all basidiomycetous DNAs ampli-__Platygloeasehacer , oR ROG fied efficiently with this primer (Table 1), Reactions that Sls pungens toe included the fungalsclective primer, ITSI-F, efficiently (elm Padi top ton amplified all of the ascomycetous and basidiomycetous samples tested + oderate:( trong: amplification,116 M. GARDES & T. D, BRUNS Mi23 4 5 67 891011 121314 0.955 0.585 0.341 0.258 Fig.2 Intensities of the amplified DNA in the ITS region from various plants and fungi using universal and taxon-selective ITS primers. Reactions in odd lanes used the ITS1 and ITS4 primer pair and those in even lanes used ITSI-F and ITS4-B primer pair. DNAs were derived from: plants—Pinus monticola (1,2), P. ponderosa (3,4, Pseudotsuga menziesii (5); basidiomycete fungi — Suillus pungens (7,8), Amanita francheti (9,10); and ascomycetes ~ Cenococcum geophilum 166 (11,12), E-strain BDG38 (13, 14). M = molecular weight marker (pUCI9 cleaved with Sau3AD. Results with the plant samples were less clear-cut. The ITS region of many plants proved to be both difficult to amplify efficiently with the universal primers and difficult to eliminate completely with the fungal-selective primers (Table 1, Fig. 2). For example, with the universal primer pair ITSI-ITS4, some conifers (eg. Pinus contorta, Pinus lambertiana) did not amplify while others (e.g. Abies concolor, Pseudotsuga menziesii) amplified poorly (Table 1). With the taxon-specific primers (ITS1-F plant-specific and ITS4-B basidiomycete-specific) the ITS region of many plants were still faintly amplified. However, in most cases these faint amplifications were barely visibly and unclear inappearance compared to amplifications from the fungal samples (Fig. 2). To determine whether an amplifiable DNA concentra- tion and quality was present in the DNA extracted from plant samples, we tested them by using the universal primers NS3 and NS4. These primers target a region of ap- proximately 600 bp in the 185 rRNA gene (White et al. 1990). This region was amplified well in all plant samples (Table 1). The reason for the low efficiency in ITS amplifi- cation of plant samples remains unknown but the size of the target could also bea factor. It has been shown recently that the ITS region in spruce is c. 3 kb and that as many as five different rDNA repeat units existed within an indi- vidual spruce genome (Bobola et al. 1992). If DNA repeats are frequently located in several unliked genomic loca- tions in conifers, then microheterogeneity may be more common in these plants. Such heterogeneity could explain the presence of doublets (e.g. Pseudotsuga menziesii, see Fig. 2) or the presence of ITS fragment unclear in appear- ance. On the other hand, non-specific amplification may also led to the different sized fragments. To examine the specificity of the primers under condi- tions in which both plant and fungal DNAs were present, wwe tested two sets of natural samples and one artificial mixture of plant and fungal DNAs. In all cases the fungal DNA was amplified to the apparent exclusion of plant DNA (Figs3-5). Three of the rust samples (the telial states of both rusts and the aecial state of Pucciniastrum -goeppertianum) and all of the ectomycorrhizal samples ‘were extracted directly from infected plant material with- out any attempt to sort the fungi from the host. In the case ofthe three rust samples itis likely that plant DNA was the predominant template in the polymerase chain reactions, because plant tissue was the main constituent of the ex- tracted material. Yet only rust RFLP patterns were visible from DNA extracted from three different hosts (Fig. 5). In the absence of basidiomycete DNA, the ITS region from at least one of these hosts, Abies concolor, can be amplified faintly with the taxon-specitic primers (Table 1). For the artificial mixture of DNAs, we mixed Pinus ponderosa DNA, the plant sample that amplified the best with the taxon-selective primers, with DNA of the fungus Suillus pungens. Yet, even when tree DNA was 11 times ‘more concentrated than the fungal DNA, only the fungus was visibly amplified (Fig. 3). +—_Hnt1__, PS12345MPS1 a Auli 2345 Fig. 3 Preferential amplification of fungal DNA in the ITS region from plant-fungal DNA mixtures using ITSI-F and ITSL-B as, primers: Hinfl and Alu 1 digest of amplified DNA from: P — Pinus ponderosa ; S~ Suillus pungens ; lanes 1 to 5~mixture of P. ponderosa (P) and S. pungens (5) DNA: 11.51 P:14l$;9 wl P:3 ql S:6pIP:6ylS;3 pl P:9 lS; 1 ul P: 11.515, respectively. 1 ul corresponds to 80 pg of DNA. M=molecular-weight marker (pUCI9 cleaved with Sau3AD,ITS PRIMERS FOR HIGHER FUNGI AND BASIDIOMYCETES 117 The utility of these taxon-selective primers for rapid identification of basidiomycetes is demonstrated with the natural samples. In the case of the ectomycorrhizae, we were able to confirm the identity of Amanita francheti mycorrhizae and reject the possibility that mycorrhizae collected below a fruitbody of Amanita pantherina be- longed to this latter species (Fig. 4). With some plants like Pinus banksiana, the universal primers also preferentially amplify the fungal ITS region (Gardes et al. 1990). In other mycorrhizae, such as those of Pterespora andromedea, the universal primers will amplify both the plant and fungal components about equally, but the ITS1-F-ITS4-B combi- nation efficiently eliminates the plant component (T. Bruns & K. Cullings, personal communications). In the case of the rusts, the ability to match patterns from different hosts is illustrated with Pucciriastrum -goeppertianum (Fig, 5). This technique should have im- mediate application to identification of rust species that share a host and are difficult to distinguish morpho- logically on that host. Examples include Cronartium ribicola/C. occidentalis on Ribes spp. and members of the ‘Melampsora epitea complex on Salix species. A. francheti A. pantherina y Mycorrhizae tM Cai core 8) (Goll core AY Fig. 4 Preferential amplification of the fungal ITS region from mycorrhiza DNAs using ITSI-F and ITS4-B as primers. Mycor- thizae were collected from two soil cores that were taken immedi: ately beneath fruitbodies of A. francheti and A. pantherina. DNA ‘was extracted from the mycorthizae, the ITS regions were ampli- fied and digested with Alul. All the mycorthizae investigated from soil core A have an identical RFLP pattern to that of the A. Jrancheti basidiocarp which was collected above them. This result indicates that A. francheti is probably the fungus involved in the formation of these mycorrhizae. All the mycorrhizae investigated from soil core B have a different RFLP pattern from both the A. pantherina basidiocarp collected above them and from that of A. francheti; thus the identity of the fungus involved in the forma- tion of these mycorrhizae remains unknown, M = molecular- weight marker (pUCI9 cleaved with Sau3AI and Taq). qAluil _.» 4 Rsal_, M123 451 2345 Fig. 5 Preferential amplification of the fungal DNA in the ITS re- gion from rust infected plants. using ITSI-F and ITS4-B as prim- ers, Alul and Rsal digests of amplified ITS fragments from: 1 Ribes leaves infected with the telial state of Cronartium ribicola; 2 C. ribicola spores; 3 Vaccinium parvifolium stems infected with the telial state of Pucciniastrum goeppertianuim ;4 Abies concolor needles infected with aecial state of P. gacppertianum; 5 P. goeppertianum from isolated aeciospores. Note that the RFLP patterns of the rust infected plants match with the RFLP pattern of the spores of the corresponding rust species. M=molecular-weight marker (pUCI9 cleaved with Sau3AD, The ITS primers that we have reported here will prob- ably be useful for examining the basidiomycete compo- nent in many other natural substrates. If these substrates contain a single dominant basidiomycete (e.g, small pieces ‘of wood), the RFLP method is probably the fastest way to analyse the amplified products. In more complex commu- ities (e.g. soil), specific taxa could be assayed by oligonucleotide probing when such probes are available (Gardes et al. 1990). In the broader context, taxon-selective amplification of the ITS region is likely to become a com- mon approach in molecular identification strategies. Taxon-selective ITS amplification has already been used. for the detection of the fungal pathogen Verticillium (Nazar et al.1991) and the identification of mosquito larvae (Porter & Collins 1991). PCR primers for the equivalent spacer region in bacteria have been used to identify the symbiotic nitrogen-fixing actinomycete Frankia in plant nodules (Simonet et al. 1991) and to distinguish a human118 M. GARDES & T. D. BRUNS pathogen Clostidrium perfringes from other organisms in the genus (Barry et al. 1991). The useful properties of the ITS region combined with the growing ITS database are likely to make this region increasingly popular in micro- bial ecology. Acknowledgements We thank Drs Randy Molina, and Keith Egger for cultures of ecto- and ectendo-mycorrhizal ascomycetes; Dr Sylvie Laliberte for in-vitro culture of Larix laricina ; Dr Mary Berbee, Ken Cullings and Yunan Li for DNA samples of several ascomycetes and plants; Tim Szaro for access to ‘unpublished sequence data and for computer help; and Dr Det Vogler for his comments on the manuscript. Financial support was provide by NSF grants BSR-9006834 and BSR- 8918454 to TD Bruns. References Anderson JB, Stasovski E (1992) Molecular phylogeny of Northern Hemisphere species of Armillaria, Mycologia, 84,505-516. Barry T, Colleran G, Glennon M, Dunican LK, Gannon F (1991) The 16/23s ribosomal spacer region as a target for DNA probes to identify Eubacteria. PCR Methods ‘and Applications, 1, 51-56 Baura G, Szaro TM, Bruns TD (1992) Gastrosuillus laricinus is a recent derivative of Suillus grevellei: mo- lecular evidence. Mycologia, 84, 592-597. Bobola MS, Smith DE, Klein A (1992) Five major nuclear ribosomal repeats represent a large and variable region of the genomic DNA of Picea rubens and P. mariana, ‘Molecular Biology and Evolution, 9, 125-137. Chen W, Hoy JW, Schneider RW (1992) Species-specific polymorphisms in transcribed ribosomal DNA of five Pythium species. Experimental Mycology, 16, 22-34, Dahlberg A (1991) Ectomycorrhiza in coniferous forest: struc- ttre and dynamics of populations and communities. PRD thesis, Uppsala. Danielson RM, Pruden M (1989) The ectomycorrhizal sta- tus of urban spruce. Mycologia, 81, 335-341 Gardes M, Bruns TD (1991) Rapid characterization of ectomycorrhizae using RFLP pattern of their PCR am- plified-ITS. Mycological Society Newsletter, 41, p.14. Gardes M, White TJ, Fortin JA, Bruns TD, Taylor JW (1991) Identification of indigenous and introduced symbiotic fungi in ectomycorrhizae by amplification of nuclear and mitochondrial ribosomal DNA. Canadian Journal of Botany, 69, 180-190. Gibson F, Deacon JW (1988) Experimental study of estab- lishment of ectomycorrhizas in different regions of birch root systems. Transactions of the British Mycological Society, 91, 239-251. Harley JL, Smith SE (1983) Mycorrhizal Symbiosis. Aca- demic Press, London. Kropp BR, Langlois CG (1990) Ectomycorrhizae in refor- estation. Canadian Journal of Forest Research, 20, 438-451. Kwok S, Kellogg DE, McKinney N, Spasic D, Goda L, Levenson C, Sninsky JJ (1990) Effects of primer-tem- plate mismatches on the polymerase chain reaction: Human immunodeficiency virus type 1 model studies. Nucleic Acids Research, 18, 999-1005. Lee SB, Taylor JW (1992) Phylogeny of five fungus-like protoctistan Phytophtora species, inferred from the in- ternal transcribed spacers of ribosomal DNA. Molecular Biology and Evolution, 9, 636-653. Mullis KB, Faloona FA (1987) Specific synthesis of DNA in vitro via a polymerase-catalysed chain reaction. ‘Methods in Enzymology, 158, 335-350. Nazar RN, Hu X, Schmidt J, Culham D, Robb J (1991) Po- tential use of PCR-amplified ribosomal intergenic se- quences in the detection and differentiation of verti- cillium wilt pathogens. Physiological and Molecular Plant Pathology, 39, 1-11 O’Donnell KL (1992) Ribosomal DNA internal tran- scribed spacers are highly divergent in the phytopathogenic ascomycete Fusarium sambucinum (Gibberella pulicaris). Current Genetics, 22,213-220 Porter CH, Collins FH (1991) Species-diagnostic differ- ences in a ribosomal DNA internal transcribed spacer from the sibling species Anopheles freeborni and Anoph- eles hermsi (Diptera: Culicidae). American Journal of Tropical Medecine and Hygiene, 45, 271-279. Read Dj (1991) Mycorthizas in ecosystems. Experientia, 47, 376-391 Simonet P, Grosjean MC, Misra AK, Nazaret S, ‘Cournoyer B, Normand P (1991) Frankia genus-specific characterization by polymerase chain reaction. Applied and Environmental Microbiology, 57,3278-3286 Sommer R, Tautz D (1989) Minimal homology require- ‘ments for PCR primers. Nucleic Acids Research, 17, 6749. White TJ, Bruns T, Lee S, Taylor J (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ (eds) PCR Protocols: a Guide to Methods and Applications, pp. 315-322. Academic Press, New York. ‘The authors are interested in the ecology and evolution of ecto mycorrhizal fungi and symbiotic interactions. Molecular identifi cation methods have been a necessary focus of our initial work, ‘This paper is one of several that have dealt with various aspects of PCR-based identification of mycorrhizal fungi