Phylogeny of nematode-trapping fungi based on18S rDNA sequences

Dag Ahreèn a;*, Bjoërn M. Ursing b, Anders Tunlid a

a Department of Microbial Ecology, Lund University, Ecology Building, S-223 62 Lund, Swedenb Department of Evolutionary Molecular Systematics, Soëlvegatan 29, S-223 62 Lund, Sweden

Received 17 October 1997; revised 17 November 1997; accepted 17 November 1997

Abstract

The small subunit (SSU) ribosomal DNA (18S rDNA) from 15 species of nematode-trapping fungi and closely related non-parasitic species were sequenced. Phylogenetic analysis indicated that species within the genera of Arthrobotrys, Dactylaria,Dactylella, Monacrosporium and Duddingtonia formed a monophyletic and isolated clade among an unresolved cluster ofapothecial ascomycetes. The phylogenetic patterns within this clade were not concordant with the morphology of the conidianor the conidiophores, but rather with that of the infection structures. The results from the different methods of treereconstruction supported three lineages; the species having constricting rings, the non-parasitic species and the species havingvarious adhesive structures (nets, hyphae, knobs and non-constricting rings) to infect nematodes. z 1998 Federation ofEuropean Microbiological Societies. Published by Elsevier Science B.V.

Keywords: Nematophagous fungus; Phylogeny; Evolution of infection structure

1. Introduction

Soils contain a diverse range of fungi that caninfect nematodes either by forming special mycelialstructures (nematode-trapping fungi) or by infectingthe nematodes as spores (endoparasites). A thirdgroup of nematophagous fungi are egg parasitesthat can penetrate the eggshell and infect the eggsof root-knot and cyst nematodes [1]. The interestsof studying the infection biology of these fungi aredue to their potential use as biological control agents

for plant and animal parasitic nematodes [2]. Fur-thermore, the ease by which most nematophagousfungi and nematodes can be kept and studied inthe laboratory have also rendered these organismsas important models for examining the mechanismsof infection of parasitic fungi [3,4].

The nematode-trapping fungi can grow both assaprophytes using a vegetative mycelium and as par-asites by forming special trapping structures includ-ing constricting rings, adhesive hyphae, nets, andknobs. Based on the morphology, nematode-trap-ping fungi have been identi¢ed as zygomycetes, ba-sidiomycetes and hyphomycetes [1]. Among them,the hyphomycetes are the largest group containing

0378-1097 / 98 / $19.00 ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V.PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 5 1 9 - 3

FEMSLE 7947 9-1-98

* Corresponding author. Tel. : +46 (46) 2223758;Fax: +46 (46) 2224158; E-mail: dag.ahren@mbioekol.lu.se

FEMS Microbiology Letters 158 (1998) 179^184

genera such as Arthrobotrys, Monacrosporium, Dac-tylaria and Dactylella. The teleomorphs of severalArthrobotrys species have recently been identi¢ed asOrbilia belonging to the discomycetes [5].

In this paper, we have used sequences of the smallsubunit (SSU) ribosomal DNA (18S rDNA) to fur-ther examine the phylogenetic relationships of thenematode-trapping hyphomycetes. We selected taxathat could ask the following questions: (1) Are thenematode-trapping hyphomycetes a monophyleticgroup? (2) Is the phylogenies based on moleculardata concordant with traditional taxonomy (basedon the morphology of conidia and conidiophores)or with the morphology of various trapping struc-tures?

2. Materials and methods

2.1. Fungal isolates and DNA extraction

Fifteen species of nematode-trapping fungi andclosely related non-parasitic species were analyzed(Table 1). Mycelia were grown in liquid soya pep-tone medium (1% w/v) at room temperature on arotary shaker. Genomic DNA was extracted essen-tially as described by Persson et al. [6].

2.2. 18S rDNA ampli¢cation and sequencing

PCR primers used to amplify 18S rDNA wereNS1 and NS8 [7]. PCR reactions were optimizedand ampli¢ed by denaturing the DNA at 94³C for4 min, 30 times cycling of 1 min at 50³C, 1 min 30 sat 72³C, 1 min 94³C, an extension at 72³C for 10min. The PCR products were puri¢ed by using Mi-crocon 50 columns (Amicon, Inc.). Apart fromabove mentioned NS1 and NS8, eight additional pri-mers were designed for sequencing: 18S1rev : 5P AG-GAT TGGGT AATTT GCG 3P, 18S2for : 5PACGGG TAACG GGGAA TAAGG 3P, 18S2rev :5P ATATT CGAGC AATAC GCCTG C 3P,18S3for : 5P GGCGA ACCAG GACTT TTAC 3P,18s3rev : 5P TGTCA ATCCT TATTG TGTCT GG3P, 18s4for : 5P CTTAA AGGAA TTGAC GGAAGG 3P, 18S4rev : 5P GCTGA TGACT TGCGC TTAC3P and 18S5for : 5P TGTGA TGCCC TTAGA ACG3P. Twenty ng of the puri¢ed PCR product wereampli¢ed with these primers using the ABI PrismDye Terminator Sequencing Ready Reaction Kit ac-cording to protocol P/N402078 (Perkin Elmer). Thesequence reactions were analyzed in a 310 ABI PrismGenetic Analyzer (Perkin Elmer) according to themanufacturer.

FEMSLE 7947 9-1-98

Table 1List of nematode-trapping fungi examined and sequence database accession numbers

Speciesa Strain no.b Capture organs Accession no.(EMBL)

Monacrosporium gephyropagum (Dactylella gephyropaga) CBS178.37 adhesive branches AJ001996Monacrosporium ellipsosporum CBS225.54 adhesive knobs AJ001995Monacrosporium haptotylum (Dactylaria candida) CBS200.50 adhesive knobs

(non-constricting ring)AJ001990

Arthrobotrys oligospora var. oligospora ATCC24927 (CBS115.81) adhesive nets AJ001986Arthrobotrys oligospora var. oligospora CBS289.82 adhesive hyphae AJ001987Arthrobotrys pyriformis CBS340.94 adhesive nets AJ001988Arthrobotrys musiformis CBS266.83 adhesive nets AJ001985Arthrobotrys conoides CBS109.52 adhesive nets AJ001983Duddingtonia £agrans CBS565.50 adhesive nets AJ001991Monacrosporium psychrophilum (Dactylaria psychrophila) L9203 adhesive nets AJ001998Arthrobotrys superba CBS107.81 adhesive nets (hyphae) AJ001989Monacrosporium doedycoides (Dactylaria dodycoides) CBS223.54 constricting rings AJ001994Arthrobotrys dactyloides (Arthrobotrys anchonia) CBS264.83 constricting rings AJ001997Dactylella rhopalota CBS493.67 not nematophagous AJ001992Dactylella oxyspora AR922 not nematophagous AJ001993

aDesignation of species follows those of Rubner [21] and Oorschot [26]. Commonly used synonymes are in paranthesis.bCBS, Centraalbureau voor Schimmelcultures, The Netherlands; ATCC, American Type Culture Collection, USA; L, isolate obtained fromLotta Persmark (Department of Microbial Ecology, Lund University, Lund, Sweden).

D. Ahreèn et al. / FEMS Microbiology Letters 158 (1998) 179^184180

2.3. Phylogenetic analysis

The sequences were aligned as described below.The three most commonly used methods of tree re-construction were used, maximum parsimony, MP[8] ; neighbor joining, NJ [9]; and maximum likeli-hood, ML [10]; implemented by the PHYLIP [10],MOLPHY [11] or PUZZLE v. 3.0 [12] programpackages. Support values for NJ and MP were estab-lished by bootstrapping with 1000 replicates. Thedistance matrix for NJ was calculated with theHKY model for sequence evolution [13]. The quartetpuzzling (QP)/ML analyses were performed with1000 puzzling steps and the TN [14] model for ntsequence evolution.

3. Results

Phylogenetic analyses were carried out on two sep-arate alignments. The interordinal relationships wereanalysed with an alignment containing 32 speciesand the alignment used for the analyses for the rela-tionships among the nematode-trapping fungi in-cluded 19 species. After excluding gaps and ambig-uous sites adjacent to gaps, the lengths of the twodi¡erent alignments were 1359 (32 species) and 1595(19 species) nucleotides. The numbers of parsimoni-ous informative sites were 63 and 107, and the var-iable sites were 157 and 325, respectively. By usingthe second alignment for the phylogenetic analysisamong the nematophagous fungi, the number of var-

FEMSLE 7947 9-1-98

Fig. 1. A cladistic tree showing the relationship among the nematode-trapping fungi and the position of this clade in an unresolved poly-furcation together with species from Pezizales (P), Leotiales (L) and Caliciales (C). The tree is a consensus of the phylogentic analyses us-ing MP, NJ and ML/QP on two separate alignments. Neolecta vitellina was used as an outgroup for these analysis [15]. The support val-ues for the branches A, B, C, D and E are shown in Table 2. Only branches with support values above 50 are shown. The followingseqences were obtained from data banks: Orbilia auricolor [16], Ascobollus lineolatus, Peziza badia, Leotia lubrica, Sphaerophorus globosus[17], Cudonia confusa, Inermisia aggregata, Neolecta vitellina, Spathularia £avida [18], Cazia £exiascus, Choiromyces venosus, Discoitis veno-sa, Discina macrospora, Peziza quelepidotia, Rhizina undulata [19], Saccharomyces cerevisiae [27], Neurospora crassa [28].

D. Ahreèn et al. / FEMS Microbiology Letters 158 (1998) 179^184 181

iable sites increased from 87 to 157, and the numberof parsimonious informative sites from 30 to 63.

Fig. 1 shows a consensus tree of all the phyloge-netic analyses (MP, NJ, ML/QP) of both alignments.It is a cladistic tree so the lengths of the branches donot represent evolutionary distances. The supportvalues for important nodes are presented in Table2. Although the apothecial fungi (discomycetes) clus-tered in an unresolved polyfurcation, the nematode-trapping hyphomycetes clustered together in amonophyletic clade. The data set did not permitany resolution between di¡erent lineages among Pe-zizales, Leotiales, Calisiales and the nematode-trap-ping fungi. There was no (6 50) support for mono-phyly within Pezizales or Leotiales. Furthermore, thesupport for the basal position of Neurospora crassato the apothecial fungi was not conclusive.

Within the clade of nematode-trapping fungi, theresults from the di¡erent methods of tree reconstruc-tion supported three lineages: the species havingconstricting rings (Monacrosporium doedycoides andArthrobotrys dactyloides) (MP: 77, NJ: 79, QP: 94),the non-parasitic species Dactylella oxyspora andDactylella minuta (MP: 31, NJ: 76, QP: 84), andthe species having various adhesive structures (nets,hyphae, knobs and non-constricting rings) to infectnematodes (MP: 51, NJ: 52, QP: 58). The last groupwas less conclusive. Although the results gave nocon¢dence in resolving the topology within thesetree lineages, there was an indication that the nem-atophagous fungi using constricting rings should bebasal with non-nematophagous species and thegroup using adhesive nematode-trapping structuresas sister groups. Within the group of species usingadhesive nematode-trapping structures, we did not

get resolution for monophyletic groups of speciesusing similar structures of nematode-trapping or-gans. As expected, Orbilia auricolo, the teleomorphof Arthrobotrys oligospora [16] was positioned closeto an isolate of this fungus.

4. Discussion

Although we obtained strong support for the nem-atode-trapping fungi (including two closely relatednon-parasitic species) forming a monophyletic group,it was not possible to position this clade among theapothecial fungi. The low resolution in the phyloge-netic analysis of the species from the orders of Pezi-zales, Leotiales and Caliciales is in accordance withthe data presented by Gargas and Taylor [17]. Theapparent higher resolutions in the phylogenetic treeof species within the Pezizales obtained by Landviket al. [18] and O'Donnell et al. [19] are probably dueto di¡erences in the alignment procedures (e.g. in thetreatment of gaps and adjacent sequences).

The taxonomic classi¢cation of nematode-trappinghyphomycetes goes mainly back to the work byDrechsler [20]. His concept was to classify the didy-mosporous species to Arthrobotrys, and the phrag-mosporous species to Dactylella when bearing a sin-gle conidium on the tip of the conidiophore and toDactylaria when more than one conidium is pro-duced in a sympodial manner. More recently, a num-ber of other genera including Monacrosporium, Dud-dingtonia, Drechsleromyces and Lactydina have beenused for classifying the nematode-trapping hypho-mycetes (for a review see [21]). Notably, the phylog-enies based on the 18S rDNA sequences were notconcordant with the morphology of the conidia northe conidiophores, but rather with the morphologyof the infection structures.

The indication that the species forming constrict-ing rings are found in a separate lineage than thespecies capturing nematodes using various adhesivehyphal structures is well supported by their di¡er-ences in morphology and physiology. The adhesivecells of hyphal branches, nets, knobs and non-con-stricting rings are all covered by a layer of extracel-lular polymers that are accumulated at the site ofadhesion between the two organisms [3,4]. Such pol-ymers are not found on the cell surface of the con-

FEMSLE 7947 9-1-98

Table 2Support values for the nodes of the phylogenetic tree presentedin Fig. 1a

Node MP NJ ML/QP

A 100 100 69B 45 52 ^b

C 50 58 51D 82 71 89E 51 46 58

aMP, maximum parsimonous; NJ, neighbor joining; ML/QP,maximum likelihood/quartet puzzling.bNot resolved.

D. Ahreèn et al. / FEMS Microbiology Letters 158 (1998) 179^184182

stricting rings. These cells possess a unique andhighly ordered ultrastructure that is collapsed duringthe in£ation of the traps [4]. Furthermore, the ringforming species have a physiology and metabolismthat appear to be more adapted for parasitic growththan that of the net forming species. On the otherhand, the net forming species are more e¤cient sap-rophytes than the ring forming species [4]. The di¤-culties in resolving the clades within the lineage offungi using various adhesive structures to capturenematodes can be expected because the morphologyand the development of these trapping devices aresimilar [4]. For example, the adhesive branches ofM. gephyropagum tend to anastomose and form sim-ple, two-dimensional nets [1]. So far it is possible toidentify two clades of fungi capturing nematodes us-ing adhesive nets. One group contained amongothers the common nematode-trapping fungi A. oli-gospora. The second group contained Arthrobotryssuperba, an isolate of A. oligospora and Arthrobotryspyriformis. Notably, it has been shown that this iso-late of A. oligospora as well as A. superba can cap-ture nematodes not only on three-dimensional net-works but also on undi¡erentiated hyphae [22,23]. Asimilar infection mechanism has to our knowledgenot been reported for A. pyriformis.

Already Drechsler [24] isolated several hyphomy-cetes including D. rhopalota having a morphologyvery similar to those of nematode-trapping fungibut not being able to capture nematodes. The closerelationship of this fungus with the nematode-trapping hyphomycetes is also suggested by the iden-ti¢cation of the teleomorph of D. rhopalota as anOrbilia sp. [25]. The fact that D. rhopalota as wellas another non-parasitic species (D. oxyspora) clus-tered together in a group within the clade of nema-tode-trapping hyphomycetes indicates that the abilityto capture nematodes have arisen at least twice,once in a lineage leading to the species forming con-stricting rings and once in a lineage leading to theformation of adhesive infection structures. Alterna-tively, the ability to infect nematodes have been lostin the lineage formed by D. rhopalota and D. oxy-spora.

In conclusion, we think that the phylogenetic anal-ysis of the 18S rDNA sequences presented in thispaper provides an important conceptual focus forfurther and more detailed analysis of the molecular

biology and evolution of parasitism within nema-tode-trapping fungi.

Acknowledgments

This study was supported by grants from theSwedish Natural Science Research Council. Wethank Prof. Birgit Nordbring-Hertz and Prof. An-ders Tehler for stimulating discussions.

References

[1] Barron, G.L. (1977) The Nematode Destroying Fungi. Topicsin Mycobiology 1, 140 pp. Canadian Biological Publications,Guelph, Ontario, Canada.

[2] Jansson, H.-B., Tunlid, A. and Nordbring-Hertz, B. (1997)Biological control: nematodes. In: Fungal Biotechnology(Anke, T., Ed.), pp. 38^50. Chapman and Hall, Weinheim.

[3] Tunlid, A., Jansson, H.-B. and Nordbring-Hertz, B. (1992)Fungal attachment to nematodes. Mycol. Res. 96, 401^412.

[4] Dijksterhuis, J., Veenhuis, M., Harder, W. and Nordbring-Hertz, B. (1994) Nematophagous fungi : physiological aspectsand structure-function relationships. In: Advances of Micro-bial Physiology (Rose, A.H. and Tempest, D.W., Eds.), vol.36, pp. 111^143. Academic Press, London.

[5] P¢ster, D.H. and Liftik, M.E. (1995) Two Arthrobotrys ana-morphs from Orbilia auricolor. Mycologia 87, 684^688.

[6] Persson, Y., Erland, S. and Jansson, H.-B. (1996) Identi¢ca-tion of nematode-trapping fungi using RFLP analysis of thePCR-ampli¢ed ITS region of ribosomal DNA. Mycol. Res.100, 531^534.

[7] White, T.J., Bruns, T., Lee, S. and Taylor, J. (1990) Ampli¢-cation and direct sequencing of fungal ribosomal RNA genesfor phylogenetics. In: PCR Protocols: A Guide to Methodsand Applications (Innis, M.A., Gelfand, D.H., Sninsky, J.J.and Sussman, A.S., Eds.), pp. 315^322. Academic Press, NewYork.

[8] Fitch, W.M. (1971) Towards de¢ning the course of evolution,minimum change from a speci¢c tree topology. Syst. Zool. 20,406^415.

[9] Saitou, N.N.M. (1987) The neighbor-joining method, a newmethod for reconstructing phylogenetic trees. Mol. Biol. Evol.4, 406^425.

[10] Felsenstein, J. (1991) PHYLIP: Phylogeny inference package,version 3.5. University of Washington, Seattle, WA.

[11] Adachi, J.H.M. (1996) MOLPHY version 2.3: Programs formolecular phylogenetics based on maximum likelihood. Com-put. Sci. Monogr. 28, 1^150.

[12] Strimmer, K.V.H.A. (1996) Quartet puzzling: A quartet max-imum-likelihood method for reconstructing tree topologies.Mol. Biol. Evol. 13, 964^969.

[13] Hasegawa, M., Kishino, H. and Yano, K. (1985) Dating of

FEMSLE 7947 9-1-98

D. Ahreèn et al. / FEMS Microbiology Letters 158 (1998) 179^184 183

the human-ape splitting by a molecular clock of mitochondrialDNA. J. Mol. Evol. 22, 160^174.

[14] Tamura, K.N.M. (1993) Estimation of the number of nucleo-tide substitutions in the control region of mitochondrial DNAin humans and Chimpanzees. Mol. Biol. Evol. 10, 512^526.

[15] Landvik, S., Eriksson, O.E., Gargas, A. and Gustafsson, P.(1993) Relationship of the genus Neolecta (Neolectales ordonov., Ascomycotina) inferred from 18S rDNA sequences. Syst.Ascomycetum 11, 107^118.

[16] P¢ster, D.H. (1997) Castor, Pollux and life histories of fungi.Mycologia 89, 1^23.

[17] Gargas, A. and Taylor, J.T. (1995) Phylogeny of discomycetesand early radiation of the apothecial ascomycotina inferredfrom SSU rDNA sequence data. Exp. Mycol. 19, 7^15.

[18] Landvik, S., Egger, K.N. and Schumacher, T. (1996) Towardsa subordinal classi¢cation of the Pezizales (Asacomycota). In:Phylogenetic rDNA Studies of Discomycetes (Ascomycota),Thesis. Department of Ecological Botany, Umeaî University,Umeaî, Sweden.

[19] O'Donnell, K., Gigelnik, E., Weber, N.C. and Trappe, J.M.(1997) Phylogenetic relationships among ascomycetous tru¥esand the true and false morels inferred from 18S and 28Sribosomal DNA sequence analysis. Mycologia 89, 48^65.

[20] Drechsler, C. (1937) Some hyphomycetes that prey on free-living terricolous nematodes. Mycologia 29, 447^552.

[21] Rubner, A. (1996) Revision of predacious hyphomycetes inthe Dactylella-Monacrosporium complex. Stud. Mycol. 39.

[22] Jansson, H.-B. and Nordbring-Hertz, B. (1981) Trap and con-idiophore formation in Arthrobotrys superba. Trans. Br. My-col. Soc. 77, 205^207.

[23] DenBelder, E. and Jansen, E. (1994) Capture of plant para-sitic nematodes by an adhesive hyphae forming isolate of Ar-throbotrys oligospora and some other nematode-trapping fun-gi. Nematologica 40, 423^437.

[24] Drechsler, C. (1943) A new nematode-capturing Dactylellaand several related hyphomycetes. Mycologia 35, 339^362.

[25] Thakur, S. and Zachariah, K. (1989) Responses of the fungusDactylella rhopalota to bacteria. Plant Soil 120, 87^93.

[26] Van Oorschot, C.A.N. (1985) Taxonomy of the Dactylariacomplex V. A review of Arthrobotrys and allied genera. In:Taxonomy of the Dactylaria Complex IV-VI (De Hoog, G.S.,Ed.), Stud. Mycol. 26, 445^451.

[27] Rubtsov, P.M., Musakhanov, M.M., Zakharyev, V.M.,Krayev, A.S., Skryabin, K.G. and Bayev, A.A. (1980) Thestructure of the yeast ribosomal RNA genes. I. The completenucleotide sequence of the 18S ribosomal RNA gene fromSaccharomyces cerevisiae. Nucleic Acids Res. 8, 5779^5794.

[28] Sogin, M.L., Miotto, K. and Miller, L. (1986) Primary struc-ture of the Neurospora crassa small subunit ribosomal RNAcoding region. Nucleic Acids Res. 14 (23), 9540.

FEMSLE 7947 9-1-98

D. Ahreèn et al. / FEMS Microbiology Letters 158 (1998) 179^184184