View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

ORIGINAL ARTICLE 10.1111/j.1469-0691.2005.01356.x

Rapid identification of clinical longibrachiatum isolates by cellulose-acetate electrophoresis-mediated isoenzyme analysis A. Szekeres1,M.La´day2, L. Kredics3, J. Varga1, Z. Antal3, L. Hatvani1, L. Manczinger1,C.Va´gvo¨lgyi1 and E. Nagy3,4

1Department of Microbiology, University of Szeged, Szeged, 2Plant Protection Institute, Hungarian Academy of Sciences, Budapest, 3Microbiological Research Group, Hungarian Academy of Sciences and University of Szeged and 4Department of Clinical Microbiology, Faculty of Medicine, University of Szeged, Szeged, Hungary

ABSTRACT Cellulose-acetate electrophoresis was used to investigate isoenzyme polymorphism among ten clinical and 11 non-clinical isolates of Trichoderma. Initial testing of 13 enzyme systems for activity and resolution of bands showed that seven were appropriate for identifying the different species. Each of the enzyme systems investigated (glucose-6-phosphate dehydrogenase, glucose-6-phosphate isomerase, 6-phosphogluconate dehydrogenase, peptidases A, B and D, and phosphoglucomutase) was diagnostic for at least one species. On the basis of the results of isoenzyme analysis, several isolates identified originally as Trichoderma pseudokoningii, T. koningii or T. citrinoviride were re-identified as T. longibra- chiatum, in agreement with sequence analysis data for the internal transcribed spacer region of the isolates. The availability of a quick, inexpensive and reliable diagnostic tool for the identification of T. longibrachiatum isolates is important, as most clinical Trichoderma isolates belong to T. longibrachiatum. Furthermore, as many different enzyme systems are available, the method may also be suitable for the identification of other clinically relevant fungal species. Keywords Cellulose-acetate electrophoresis, diagnosis, identification of fungi, isoenzyme analysis, Trichoderma spp. Original Submission: 10 January 2005; Revised Submission: 19 July 2005; Accepted: 13 September 2005 Clin Microbiol Infect 2006; 12: 369–375

has been published previously [3]. The number INTRODUCTION of reported cases has grown from year to year, Species of the imperfect filamentous fungal possibly because of the expanding population of genus Trichoderma with teleomorphs belonging immunocompromised hosts. Four further cases to the order of the are mentioned in a study concerning extracellu- division are of great economic importance as lar protease production by clinical isolates of sources of enzymes and antibiotics, as plant T. longibrachiatum [4], and an additional report growth promoters, as decomposers of xenobiot- describes the infection of a cerebrospinal fluid ics, and as commercial biofungicides [1,2]. shunt device in a non-immunocompromised However, this genus is also on the growing list patient [5]. of emerging fungal pathogens, and has been Data concerning the antifungal susceptibilities reported in an increasing number of cases as an of clinical Trichoderma isolates indicate high aetiological agent in human infections. A review levels of resistance to the antifungal drugs used of 40 cases of infection associated with Tricho- routinely, which may cause difficulties in the derma spp., mostly Trichoderma longibrachiatum, treatment of infected patients [6]. Antifungal susceptibility data often correlate with the taxo- nomic position of isolates; therefore, accurate Corresponding author and reprint requests: A. Szekeres, identification of opportunistic fungi is crucial Department of Microbiology, University of Szeged, PO Box 533, H-6701 Szeged, Hungary for adequate therapeutic interventions. As most E-mail: [email protected] clinical Trichoderma isolates are T. longibrachia-

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases 370 Clinical Microbiology and Infection, Volume 12 Number 4, April 2006 tum, the availability of a quick, inexpensive and including 99 isoenzyme bands revealed by reliable diagnostic tool for the identification of PAGE with six enzyme systems. It has also been T. longibrachiatum isolates would be very shown that isoenzyme profiles of extracellular important. cell-wall-degrading enzymes, separated by Isoenzyme analysis is a powerful method for SDS-PAGE, show correlation with rDNA taxo- taxonomic, genetic and population studies. Zamir nomic species [14]. From the results of these and Chet [7] demonstrated that enzyme electro- studies, it was concluded that isoenzyme patterns phoresis could be used to distinguish between might also be useful as diagnostic tools for the isolates of Trichoderma harzianum. Stasz et al. [8] routine laboratory identification of clinical evaluated 63 enzyme assays by horizontal starch T. longibrachiatum isolates. gel electrophoresis for their applicability to the Cellulose-acetate electrophoresis (CAE) is a study of Trichoderma spp., and then performed promising alternative technique for isoenzyme cladistic analysis of isoenzyme polymorphism analysis, and has been applied successfully for among 16 enzymes for the evaluation of phenetic the identification of fungal species within the species and phylogenetic relationships within the genera Ganoderma [15], Fusarium [16] and Phyt- genus [9]. Isoenzyme analysis of 23 enzyme ophthora [17,18]. In the present study, CAE- systems on starch gels was used to compare the based multilocus enzyme electrophoresis was Trichoderma anamorphs of two spp., and used for the investigation of isoenzyme poly- with this method it was possible to distinguish morphisms among ten clinical and 11 environ- between strains of T. longibrachiatum, Trichoderma mental Trichoderma isolates in order to identify pseudokoningii and Trichoderma reesei [10]. In a isoenzymes that could be used for species subsequent study, Leuchtmann et al. [11] deter- identification. mined isoenzyme subgroups within Trichoderma section Longibrachiatum, based on the examination MATERIALS AND METHODS of 78 strains with ten enzyme systems. The results of this study supported Bisset’s morphology- Fungal strains based taxonomic scheme for this section [12]. The ten clinical and 11 non-clinical Trichoderma isolates Grondona et al. [13] characterised 15 isolates of examined are listed in Table 1. More information about the T. harzianum on the basis of morphological, phy- clinical isolates and related case reports are available else- where [3]. siological, molecular and biochemical features,

Table 1. List of Trichoderma isolates used in this study

Morphology-based, Culture Identity according to GenBank accession Geographical original description collection no. ITS sequence analysis no. of ITS 1 Source origin ET

Clinical isolates T. citrinoviride UAMH 9573 T. longibrachiatum AY328038 Peritoneal catheter Newfoundland, Canada VIII T. longibrachiatum ATCC 201044 T. longibrachiatum AY585879 Skin lesion Houston, TX, USA X CBS 446.95 T. longibrachiatum AY328039 Lung Austria XII UAMH 9515 T. longibrachiatum AY328035 Peritoneal effluent Newfoundland, Canada XII ATCC 208859 T. longibrachiatum AY328042 HIV+ patient Texas, USA XII T. koningii CM 382 T. longibrachiatum AY328034 Peritoneal effluent Canary Islands, Spain XII T. pseudokoningii UAMH 7955 T. longibrachiatum AY328040 Acute invasive sinusitis Pennsylvania, USA XII UAMH 7956 T. longibrachiatum AY328041 Lung, liver Iowa, USA XI IP 2110.92 T. longibrachiatum Z82902 Lung, brain, heart, stomach France XII T. viridea 043.99a T. koningii AY328036 Nasal mucus Austria I Non-clinical isolates T. citrinoviride UAMH 999 T. citrinoviride X93940 Lung of woodchuck Canada VI T. koningii CECT 2412 T. longibrachiatum AF027216 Mushroom compost Wales XII T. longibrachiatum CECT 2606 T. longibrachiatum X93929 Soil Sierra Leone VII T. pseudokoningii CECT 2937 T. longibrachiatum Z82912 Soil Antarctica IX T. virens SZMC 0931 T. virens DQ118083 Agricultural soil Hungary III T. virens SZMC 0561 T. virens DQ118085 Agricultural soil Hungary III T. harzianum SZMC 0559 T. harzianum DQ118087 Agricultural soil Hungary V T. harzianum SZMC 0560 T. harzianum DQ118084 Agricultural soil Hungary V T. harzianum SZMC 0566 T. harzianum DQ118088 Agricultural soil Hungary IV T. harzianum SZMC 0919 T. harzianum DQ118089 Agricultural soil Hungary IV T. harzianum SZMC 0930 T. brevicompactum DQ118086 Agricultural soil Hungary II

ET, electrophoretic type; ITS 1, internal transcribed spacer 1; HIV, human immunodeficiency virus; ATCC, American Type Culture Collection; CBS, Centraalbureau voor Schimmelcultures; CECT, Coleccio´n Espan˜ola de Cultivos Tipo; CM, Centro Nacional de Microbiologia, Immunologia y Virologia Sanitaria; IP, Institut Pasteur; UAMH, University of Alberta Microfungus Collection and Herbarium; SZMC, Szeged Microbiological Collection. aENT University Hospital, Graz, Austria [41].

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 369–375 Szekeres et al. Identification of clinical Trichoderma isolates 371

PHYLTOOLS and the NEIGHBOR and CONSENSE programs Sample preparation of the PHYLIP v. 3.57c software package [21,22]. Phylogenetic For isoenzyme analysis, isolates were grown in 250-mL trees were prepared by the neighbour-joining method [23], Erlenmeyer flasks containing liquid yeast extract medium using the NEIGHBOR program. Data were analysed using the ⁄ ⁄ ⁄ (yeast extract 5 g L, glucose 10 g L, KH2PO4 5g L in water). outgroup rooting method, with strain Trichoderma koningii Flasks were inoculated with conidial suspensions of the 043.99 designated as the outgroup. Trichoderma isolates to a final concentration of 105 conidia ⁄ mL, The internal transcribed spacer 1 (ITS 1) sequences were and were incubated for 4 days on an orbital shaker at 200 rpm gathered from the GenBank database (see Table 1 for accession and 25C. Mycelia were vacuum-filtered through No. 1 filter numbers). TrichoBLAST, a publicly available database of paper (Whatman, Maidstone, UK), washed three times with Trichoderma and Hypocrea sequences supported by sequence 50 mL of distilled water, and lyophilised. Dried mycelia were diagnosis, and similarity search tools [24] were used for powdered with a pestle and mortar and stored at )20C until sequence analysis. DNA sequences were aligned first with protein extraction. CLUSTAL X v. 1.83 [25] and were adjusted manually with Genedoc v. 2.6 (http://www.psc.edu/biomed/genedoc). Phylogenetic trees were prepared by the neighbour-joining Protein extraction and enzyme electrophoresis method with CLUSTAL X software, with bootstrap values Protein extraction was performed as described by La´day and calculated from 1000 replications of the bootstrap procedure. Sze´csi [16]. CAE was as described by Hebert and Beaton [19], Strain T. koningii 043.99 was used as the outlier during the with a CAE system from Helena Laboratories (Beaumont, TX, analyses. USA). Titan III cellulose-acetate gels (Helena Laboratories) were soaked for ‡ 30 min in electrophoresis buffer (0.25 mM Tris-glycine, pH 8.5) and were then blotted dry between sheets RESULTS of filter paper. The protein extracts were applied from the sample plate to the gel with a Super Z-12 Applicator. When the Thirteen enzyme systems were tested initially for staining activity was low, the extracts were blotted two or activity, resolution and consistent appearance of three times. Electrophoresis was carried out at 180 V for the bands (Table 2). Seven enzyme systems show- 20 min. The gels were stained for 13 enzyme systems (Table 2), ing clear, reproducible and resolvable banding and the enzyme activities were detected using agar overlays. Staining protocols were as described previously [19]. All patterns were selected for the full sample set, samples were extracted and analysed on three occasions in from which 40 distinct bands were scored and separate runs. Bands were ordered alphabetically, based on used in the analysis (Table 3). their relative mobility; the band located next to the anode was Electromorphs of glucose-6-phosphate isomerase designated as ‘A’ for each enzyme assay. (GPI), peptidase A, peptidase B, phosphoglu- comutase and 6-phosphogluconate dehydro- Analysis of data genase showed single bands for all the isolates, Matrices were created from the isoenzyme data, based on the whereas peptidase D (PEP D) showed double and presence or absence of a band with a given mobility. Simple matching coefficients were calculated with the PHYLTOOLS triple bands. The main bands of peptidase A and v. 1.32 software package [20]. Bootstrap values were collected peptidase B were identical, but phosphogluco- from 1000 replications of the bootstrap procedure using mutase, GPI, 6-phosphogluconate dehydroge- nase and PEP D also showed faint (and not always well-resolved) bands that were not scored. Table 2. Enzyme systems tested and their Enzyme Com- Glucose-6-phosphate dehydrogenase showed more mission (EC) numbers

No. of different Table 3. Electrophoretic types (ETs) of the Trichoderma electrophoretic Enzyme Abbreviation EC no. Activity patterns isolates with isoenzyme phenotypes at seven loci

6-Phosphogluconate 6PGDH 1.1.1.44 + 5 Isoenzyme locus dehydrogenase Aconitase ACN 4.2.1.3 – – Electrophoretic types PGM G6PDH PEP A PEP B PEP D 6PGDH GPI Glucose-6-phosphate G6PDH 1.1.1.49 + 6 dehydrogenase ET I ADAABBA Glucose-6-phosphate GPI 5.3.1.9 + 6 ET II CBDBHAC isomerase ET III ECBDHCC Glycerol-3-phosphate GPDH 1.1.1.8 – – ET IV BADCACB dehydrogenase ET V BADCBCC Malate dehydrogenase MDH 1.1.1.37 – – ET VI DEEECEE Peroxidase PRX 1.11.1.7 – – ET VII DECCFCD Peptidase A (Gly-Leu) PEP A 3.4.11 ⁄ 13 + 5 ET VIII DCCBFBD Peptidase B (Leu-Gly-Gly) PEP B 3.4.11 ⁄ 13 + 5 ET IX DFCCEDF Peptidase D (Phe-Pro) PEP D 3.4.11 ⁄ 13 + 8 ET X DEBCGBF Phosphoglucomutase PGM 5.4.2.2 + 5 ET XI DECCDCF Shikimate dehydrogenase SKDH 1.1.1.25 – – ET XII DECCDBF Succinate dehydrogenase SUD 1.3.99.1 – – For enzyme abbreviations, see Table 2.

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 369–375 372 Clinical Microbiology and Infection, Volume 12 Number 4, April 2006 complex banding patterns, but only the three (or IX, XI and XII were originally misidentified, and occasionally two) main reproducible bands were that all of them might be T. longibrachiatum. considered for the determination of electro- A phylogenetic tree was prepared from the morphs (Table 2). isoenzyme data matrix (Fig. 1A). The isoenzyme All of the enzyme systems examined were patterns of T. koningii 043.99 (ET I), an isolate polymorphic for the full sample set (Table 3). representing section Trichoderma, clade Rufa, High intra-specific polymorphism was observed according to the currently suggested species among the T. longibrachiatum isolates, with the concept for the genus [26,27], exhibited isoen- PEP D, 6-phosphogluconate dehydrogenase and zyme patterns that were clearly different from all glucose-6-phosphate dehydrogenase patterns the other isolates examined; therefore, it was showing four, three and three banding patterns, designated as the outgroup taxon. The remaining respectively. The variability was lower for pep- isolates clustered into two major groups on the tidase A, peptidase B and GPI, which each dendrogram based on isoenzyme data. showed two banding patterns. The PEP D and The first cluster contained seven non-clinical GPI isoenzymes also showed intra-specific poly- isolates representing section Pachybasium B of morphism for the T. harzianum isolates. High the genus [26,27], including two T. virens iso- inter-specific polymorphism was detected in the lates (clade Virens, ET III) and four T. harzianum patterns of all isoenzymes (Table 3). isolates (clade Lixii ⁄ Catoptron) belonging to The enzyme patterns were used to group the two different ETs, ET IV and ET V. ET II is isolates into 12 different electrophoretic types also located within this clade, and is represent- (ETs) (Table 3). These were arranged according to ed by an isolate belonging to the newly the original, morphology-based identification of described species Trichoderma brevicompactum the isolates: ET I for T. koningii; ET II, ET IV and (clade Lutea) [28] based on ITS sequence analy- ET V for T. harzianum; ET III for Trichoderma sis (A. Szekeres et al., unpublished data; Gen- virens; ET VI and ET VIII for Trichoderma citrino- Bank accession number DQ118086). To our viride; ET VII and ET X for T. longibrachiatum; knowledge, this is the first report of the occur- ET IX and ET XI for T. pseudokoningii; and ET XII rence of this species in . for certain isolates of three species—T. longibra- The second main cluster corresponds to section chiatum, T. pseudokoningii and T. koningii Longibrachiatum of the genus Trichoderma.It (Table 1). The results of these experiments sug- includes isolates exhibiting ETs VI–XII, with gested that certain isolates belonging to ETs VIII, strain T. citrinoviride UAMH 999 (ET VI) forming

Fig. 1. (A) Neighbour-joining dendrogram resulting from the analysis of isoenzymes with different electrophoretic types. (B) Neighbour-joining tree based on phylogenetic analysis of ITS 1 sequence data of the isolates. The species names indicated correspond to the ITS-based identification. The numbers below branches indicate bootstrap values.

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 369–375 Szekeres et al. Identification of clinical Trichoderma isolates 373 a distinct branch. This cluster includes all T. long- be used for species-specific detection in the genus ibrachiatum isolates, as well as isolate UAMH 9573 Trichoderma. Thus, random amplification of poly- (identified previously as T. citrinoviride), isolates morphic DNA, using both fingerprinting and CM 382 and CECT 2412 (identified previously random primers, has been used successfully as T. koningii), and isolates UAMH 7955, [34], but results are not always reproducible, UAMH 7956, IP 2110.92 and CECT 2937 (identi- and the profiles may be unstable or too sensitive fied previously as T. pseudokoningii). Electropho- to reaction conditions, as has been demonstrated retic types VII–XII formed a well-defined cluster, with Aspergillus fumigatus [35]. demonstrating high biochemical similarity, sug- In general, sequence analysis of the ITS region gesting that the above-mentioned five clinical and of rDNA is considered to be the most reliable two environmental isolates are also isolates of method for phylogenetic analysis and species T. longibrachiatum. identification of fungi. This method has been used Previous studies on the variability of the to identify and investigate the of ITS 1–5.8S rRNA–ITS 2 region in Trichoderma various Trichoderma spp. [26,36–39], including section Longibrachiatum revealed no variation members of section Longibrachiatum [29,31,40]. within the 5.8S rRNA gene, very low variation Species identification based on ITS sequence in ITS 2, and low, but informative, variation in analysis may be rapid enough if all the necessary ITS 1 [29]. Previously sequenced ITS 1 regions of equipment is available, but the costs are high. The the examined isolates [30,31] (A. Szekeres et al., costs become substantially lower if a high- unpublished data) were used for the construction throughput apparatus is shared by a group of of a dendrogram (Fig. 1B), which was consistent laboratories, or if sequence data are obtained on with that based on the isoenzyme data. However, an outsourcing basis, although this results in a CAE appeared to be superior to ITS 1 with respect slower identification process. However, sequen- to the discriminatory power within T. longibrachi- cing techniques are not available and affordable atum. Isolate T. citrinoviride UAMH 999 also for many hospitals, especially those in less-devel- formed a separate branch on the ITS tree. As oped countries. Thus, although ITS sequence expected, all T. longibrachiatum isolates formed a analysis may be better suited for the rapid well-defined cluster, which included the five diagnosis and identification of Trichoderma spp., clinical and two saprophytic isolates originally CAE-mediated isoenzyme analysis seems to be an described as non-T. longibrachiatum, supporting inexpensive and useful alternative for the identi- the hypothesis that they should be re-identified as fication of clinical T. longibrachiatum isolates. T. longibrachiatum. Kuhls et al. [31] have also As there are many available enzyme systems suggested that isolate IP 2110.92 [32] should be for CAE, it is highly probable that markers can be re-identified as T. longibrachiatum on the basis of found for the differentiation and practical iden- PCR fingerprinting and ITS sequence data. tification of other clinically relevant fungal spe- cies, as has been demonstrated for the genus Fusarium [16]. About 300 samples can be exam- DISCUSSION ined per day by CAE with a single set of There is a need for simple, rapid, inexpensive and equipment. The good resolution of isoenzymes reliable diagnostic methods for the identification on pre-cast gels, a short run time (20 min) and of fungi in clinical microbiology laboratories. minimal equipment requirements make CAE a Identification of Trichoderma spp. on the basis promising and inexpensive method for culture- of morphological and cultural characteristics is based diagnosis that can be applied for the difficult, as their traits exhibit variations on a identification of clinical isolates of T. longibrachi- continuous scale that may overlap between the atum and other clinically relevant fungi. The species. Although a detailed taxonomic key based results of the present study support the hypothe- on these characteristics is available for the iden- sis that most Trichoderma isolates involved in tification of T. longibrachiatum [33], accurate iden- opportunistic infections belong to the Longibra- tification requires specialist expertise and is not chiatum section of the genus, and that previous an option for routine diagnostic laboratories reports concerning the isolation of other Tricho- encountering several other clinically relevant derma spp. from immunocompromised patients filamentous fungal species. Molecular assays can may have been incorrect.

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 369–375 374 Clinical Microbiology and Infection, Volume 12 Number 4, April 2006

ACKNOWLEDGEMENTS 17. Oudemans P, Coffey MD. Isozyme comparison within and among worldwide sources of three morphologically dis- This work was supported by grant F037663 of the Hungarian tinct species of Phytophthora. Mycol Res 1991; 95: 19–30. Scientific Research Fund and grant OMFB-00219 ⁄ 2002 of the 18. Goodwin SB, Schneider RE, Fry WE. Use of cellulose- Hungarian Ministry of Education. We thank W. Buzina (ENT acetate electrophoresis for rapid identification of allozyme University Hospital, Karl-Franzens-University, Graz, Austria) genotypes of Phytophthora infestans. Plant Dis 1995; 79: for the gift of isolate 043.99. 1181–1185. 19. Hebert PDN, Beaton MJ. Methodologies for allozyme analysis using cellulose acetate electrophoresis. A practical handbook. REFERENCES Beaumont, TX: Helena Laboratories, 1993. 20. Buntjer JB. Phylogenetic computer tools (PhylTools). Version 1. Chet I, Inbar J. Biological control of fungal pathogens. Appl 1.32 for Windows. Wageningen: Laboratory of Plant Biochem Biotechnol 1994; 48: 37–43. Breeding, Wageningen University, 1997. 2. Manczinger L, Antal Z, Kredics L. Ecophysiology and 21. Felsenstein J. Confidence limits on phylogenies: an breeding of mycoparasitic Trichoderma strains. Acta approach using the bootstrap. Evolution 1985; 39: 783–791. Microbiol Immunol Hung 2000; 49: 1–14. 22. Felsenstein J. PHYLIP (Phylogeny Inference Package). Version 3. Kredics L, Antal Z, Do´czi I, Manczinger L, Kevei F, Nagy 3.57c. Seattle: Department of Genetics, University of E. Clinical importance of the genus Trichoderma. Acta Washington, 1995. Microbiol Immunol Hung 2003; 50: 105–117. 23. Saitou N, Nei M. The neighbor-joining method: a new 4. Kredics L, Antal Z, Szekeres A et al. Production of extra- method for reconstructing phylogenetic trees. Mol Biol Evol cellular proteases by human pathogenic Trichoderma long- 1987; 4: 406–425. ibrachiatum strains. Acta Microbiol Immunol Hung 2004; 51: 24. Kopchinskiy A, Komon M, Kubicek CP, Druzhinina I. 283–295. TrichoBLAST: a multilocus database for Trichoderma and 5. Piens MA, Celard M, De Monbrison F et al. Trichoderma Hypocrea identifications. Mycol Res 2005; 109: 657–660. infection of cerebrospinal fluid shunt device in a non im- 25. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, munocompromised patient. J Mycol Med 2004; 14: 49–51. Higgins DG. The Clustal–windows interface: flexible 6. Do´czi I, Do´sa E, Varga J, Antal Z, Kredics L, Nagy E. Etest strategies for multiple sequence alignment aided by for assessing the susceptibility of filamentous fungi. Acta quality analysis tools. Nucleic Acids Res 1997; 25: 4876– Microbiol Immunol Hung 2004; 51: 271–281. 4882. 7. Zamir D, Chet I. Application of enzyme electrophoresis for 26. Druzhinina I, Kubicek CP. Species concepts and biodi- the identification of isolates in Trichoderma harzianum. Can J versity in Trichoderma and Hypocrea: from aggregate Microbiol 1985; 31: 578–580. species to species clusters? J Zhejiang Univ SCI 2005; 6B: 8. Stasz TE, Weeden NF, Harman GE. Methods of isoenzyme 100–112. electrophoresis for Trichoderma and Gliocladium species. 27. Chaverri P, Candoussau F, Samuels GJ. Hypocrea phyllos- Mycologia 1988; 80: 870–874. tachydis and its Trichoderma anamorph, a new bambusico- 9. Stasz TE, Nixon K, Harman GE, Weeden NF, Kuter A. lous species from France. Mycol Progr 2004; 3: 29–36. Evaluation of phenetic species and phylogenetic 28. Kraus GF, Druzhinina I, Gams W et al. Trichoderma brevi- relationships in the genus Trichoderma by cladistic ana- compactum sp. nov. Mycologia 2004; 96: 1059–1073. lysis of isozyme polymorphism. Mycologia 1989; 81: 391– 29. Kuhls K, Lieckfeldt E, Samuels GJ, Meyer W, Kubicek CP, 403. Bo¨rner T. Revision of the Trichoderma sect. Longibrachiatum 10. Samuels GJ, Petrini O, Manguin S. Morphological and including related teleomorphs based on analysis of ribo- macromolecular characterization of Hypocrea schweinitzii somal DNA internal transcribed spacer sequences. Myco- and its Trichoderma anamorph. Mycologia 1994; 86: 421–435. logia 1997; 89: 442–460. 11. Leuchtmann A, Petrini O, Samuels GJ. Isozyme subgroups 30. Kredics L, Antal Z, Varga J et al. Genetic variability and in Trichoderma section Longibrachiatum. Mycologia 1996; 88: antifungal susceptibilities of clinical Trichoderma isolates. 384–394. In: Proceedings of Trends in Medical Mycology: 9th Congress of 12. Bisset J. A revision of the genus Trichoderma. I. Section the European Confederation of Medical Mycology and 7th Longibrachiatum sec. nov. Can J Bot 1984; 62: 924–931. Trends in Invasive Fungal Infections. Amsterdam: Monduzzi 13. Grondona I, Hermosa R, Tejada M et al. Physiological Editore, 2003; 133–136. and biochemical characterization of Trichoderma harzia- 31. Kuhls K, Lieckfeldt E, Bo¨rner T, Gue´ho E. Molecular num, a biological control agent against soilborne fungal reidentification of human pathogenic Trichoderma isolates plant pathogens. Appl Environ Microbiol 1997; 63: 3189– as Trichoderma longibrachiatum and Trichoderma citrinovi- 3198. ride. J Med Mycol 1999; 37: 25–33. 14. Sanz L, Montero M, Grondona I et al. Cell wall-degrading 32. Gautheret A, Dromer F, Bourhis J-H, Andremont A. isoenzyme profiles of Trichoderma biocontrol strains show Trichoderma pseudokoningii as a cause of fatal infection in a correlation with rDNA taxonomic species. Curr Genet 2004; bone marrow transplant recipient. Clin Infect Dis 1995; 20: 46: 277–286. 1063–1064. 15. Smith BJ, Sivasithamparam K. Isozymes of Ganoderma 33. Samuels GJ, Petrini O, Kuhls K, Lieckfeldt E, Kubicek CP. species from Australia. Mycol Res 2000; 104: 952–961. The Hypocrea schweinitzii complex and Trichoderma sect. ´ 16. La´day M, Sze´csi A. Distinct electrophoretic isoenzyme Longibrachiatum. Stud Mycol 1998; 41: 1–54. profiles of Fusarium graminearum and closely related spe- 34. Turner D, Kovacs W, Kuhls K et al. Biogeography and cies. Syst Appl Microbiol 2001; 24: 67–75. phenotypic variation in Trichoderma sect. Longibrachiatum

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 369–375 Szekeres et al. Identification of clinical Trichoderma isolates 375

and associated Hypocrea species. Mycol Res 1997; 101: 449– eomorphs for T. minutisporum and T. polysporum, and 459. their newly discovered relatives. Mycologia 2004; 96: 310– 35. Lasker BA. Evaluation of performance of four genotyping 342. methods for studying the genetic epidemiology of Asperg- 39. Hermosa MR, Keck E, Chamorro I et al. Genetic diversity illus fumigatus isolates. J Clin Microbiol 2002; 40: 2886–2992. shown in Trichoderma biocontrol isolates. Mycol Res 2004; 36. Kindermann J, El-Ayouti Y, Samuels GJ, Kubicek CP. 108: 897–906. Phylogeny of the genus Trichoderma based on sequence 40. Gherbawy Y, Druzhinina I, Shaban GM et al. Trichoderma analysis of the internal transcribed spacer region 1 of the populations from alkaline agricultural soil in the Nile rDNA cluster. Fungal Genet Biol 1998; 24: 298–309. valley, Egypt, consist of only two species. Mycol Progr 37. Kubicek CP, Bissett J, Druzhinina I, Kullnig-Gradinger C, 2004; 3: 211–218. Szaka´cs G. Genetic and metabolic diversity of Trichoderma: 41. Willinger B, Obradovic A, Selitsch B et al. Detection and a case study on South-East Asian isolates. Fungal Genet Biol identification of fungi from balls of the maxillary 2003; 38: 310–319. sinus by molecular techniques. J Clin Microbiol 2003; 41: 38. Lu BS, Druzhinina I, Chaverri P et al. Hypocrea ⁄ Tricho- 581–585. derma species with pachybasium-like conidiophores: tel-

2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 369–375