the exobasidiales: an evolutionary hypothesis 1 · 2018. 6. 3. · n the new system of...

n the new system of Ustilaginomycetes (BAUER, OBER-WINKLER & VÁNKY 1997), all species lacking teliospo-res belong either to the Microstromatales (BEGEROW,

BAUER & OBERWINKLER 2001) or Exobasidiales. BAUER et al.(2001a) used the order Exobasidiales for species having localcomplex host-parasite interaction apparatus with the formationof interaction rings. Molecular analyses confirmed this group(BEGEROW, BAUER & OBERWINKLER 1997; BAUER et al.2001a). Morphologically, however, the Exobasidiales possessa high degree of divergence (BAUER, BEGEROW & OBER-WINKLER 1998, BAUER et al. 2001a). In addition, phylogeneticand evolutionary aspects within this group are poorly under-stood. For example, the G+C content of the DNA, RFLP ana-lyses of the nuclear large subunit rRNA gene as well as mole-cular phylogenetic analyses of the nuclear small subunit rRNAgene show a great heterogeneity even among the Exobasi-dium species tested (BLANZ & OBERWINKLER 1983, BLANZ &DÖRING 1995, DÖRING & BLANZ 2000). Therefore, in the pre-sent study we compare morphological, ultrastructural, andmolecular characters of members of the Exobasidiales, in orderto estimate the phylogenetic relationship in this group. Basedon the results, evolutionary strategies are discussed.

Material and methods

Specimens, the respective characters studied, and the originof the sequences are listed in Table 1.

For the study of sori, freehand sections through infectedareas of leaves were mounted in 3% KOH and examined withlight microscope using phase contrast optics.

For the study of the ballistosporic or gastroid nature ofLaurobasidium lauri, the ballistospore-fall method was used(DERX 1930): living sori were fixed inside the caps of Petridishes containing 2% water agar and kept at room tempera-ture. After two days the agar underneath the sori was cut outin pieces of about 15 mm in diameter, transferred to slides,and observed by light microscopy.

For TEM, except for Exobasidium pachysporum all spe-cimens were fixed with 2 % glutaraldehyde in 0.1 M sodiumcacodylate buffer (pH 7.2) at room temperature overnight.Following six transfers in 0.1 M sodium cacodylate buffer,samples were postfixed in 1 % osmium tetroxide in the samebuffer for 1 h in the dark, washed in distilled water, and stai-ned in 1 % aqueous uranyl acetate for 1 h in the dark. Afterfive washes in distilled water, samples were dehydrated in ace-tone, using 10 min changes at 25 %, 50 %, 70 %, 95 %, and 3times in 100 % acetone. Samples were embedded in Spurr’splastic and sectioned with a diamond knife. Ultrathin serialsections were mounted on formvar-coated, single-slot coppergrids, stained with lead citrate at room temperature for 5 min,and washed with distilled water. They were examined usinga transmission electron microscope operating at 80 kV.

The Exobasidiales: an evolutionary hypothesis1

Dominik BEGEROW2,*, Robert BAUER2, and Franz OBERWINKLER2

To gain insight in the phylogenetic relationships within the Exobasidiales, septal pore apparatus, host-parasite interac-tions, sori, hymenia, basidia, and nucleotide sequences from the 5’ terminal domain of the nuclear large subunit rRNA genewere studied and compared.

The results of our molecular phylogenetic analyses correlate well with the morphological data and both reflect the distri-bution of parasites on several host groups. Thus, the Exobasidiales seem to be divided into four groups, which are distin-guishable by basidial morphology and host range as follows: (i) the Exobasidiaceae parasitizing mainly Ericanae arecharacterized by an abaxial orientation of the hilar appendices of the ballistosporic basidiospores on the elongate basidia,(ii) the Cryptobasidiaceae occurring mainly on Lauraceae sporulate inside the host tissue with elongate gastroid basidia,(iii) the Brachybasidiaceae living on monocots are characterized by elongate basidia bearing two ballistosporic basidiosporeswith adaxially oriented hilar appendices, and (iv) the Graphiolaceae occurring on palms produce chains of gastroid basi-dia in distinct basidiocarps. The arrangement of the four groups and the tree topology within Exobasidium derived from themolecular analyses essentially parallel phylogenetic host relationships, suggesting cospeciation. Based on our results, how-ever, the radiation of Exobasidium on Vaccinioideae cannot be explained by cospeciation alone.

Mycological Progress 1(2): 187–199, 2002 187

1 Part 204 of the series “Studies in Heterobasidiomycetes” from theBotanical Institute, University of Tübingen

2 Universität Tübingen, Botanisches Institut, Lehrstuhl Spezielle Bo-tanik und Mykologie, Auf der Morgenstelle 1, D-72076 Tübingen,Germany

* Email: [email protected]

© DGfM 2002

I

188 Evolution of the Exobasidiales

© DGfM 2002

Tab. 1. Specimens and characters studied

Specimens Hosts Characters Sequences2 Source3

studied1

1. Exobasidiales

Arcticomyces warmingii (Rostr.) Savile Saxifraga bryoides L. H, S AF 487380* R.B. 3081

Botryoconis tumefaciens (Winter) laurel H Rbh., Fgi. eur. 3295 (M)H. & P. Sydow

Brachybasidium pinangae (Rac.) Gäum. Pinanga kuhlii Blume H Sydow, Fgi. exot. exs. 455 (M)

Clinoconidium bullatum H. Syd. Phoebe neurophylla H, S AF 487383* Sydow, Fgi. exot. exs. 553 (M)Metz & Pittier

Clinoconidium cf. bullatum Apollonias barbujana H, S AF 487382* R.B. 30024

(Cav.) Bornm.

Clinoconidium sp. Cinnamomum japonicum H, S AF 487381* R.B. 30825

Sieb.

Coniodictyum chevalieri Har. & Pat. Zizyphus mucronata Willd. H, S AF 487384* R.B. 10006

Dicellomyces calami R. Berndt & Calami cf. rotangis L. H R.B. 30847

N. D. Sharma

Dicellomyces scirpi Raitv. Scirpus sylvaticus L. H, S AF 487385* R.B. 1032

Drepanoconis brasiliensis Schröter & P. Henn. Ocotea sp. H Rbh., Fgi. eur. 4495 (M)

Exobasidiellum graminicola (Bres.) Donk Bromus inermis Leysser H Krieger, Fgi. saxon. 664 (M)

Exobasidium arescens Nannf. Vaccinium myrtillus L. H, S AF 352057BE R.B. 2047

Exobasidium bisporum Sawada ex A. Ezuka Eubotryoides grayana Hara S AF 487386* DSM 4454

Exobasidium gracile (Shirai) Sydow Camellia sp. S AF 487387* DSM 4441

Exobasidium japonicum Shirai Rhododendron lateritium S AF 487388* DSM 4463Planch

Exobasidium karstenii Sacc. & Trott. Andromeda polifolia L. H, S AF 487389* R.B. 2052

Exobasidium myrtilli Siegm. Vaccinium myrtillus L. H, S AF 487390* R.B. 2055

Exobasidium oxycocci Rostrup ex Shear Vaccinium oxycoccos L. H F.O. 18925

Exobasidium oxycocci Rostrup ex Shear Vaccinium oxycoccos L. H, S AF 487391* R.B. 2086

Exobasidium pachysporum Nannf. Vaccinium uliginosum L. H, S AF 487392* R.B. 947

Exobasidium pieridis-ovalifoliae Sawada Lyonia neziki Nakai & Hara S AF 487393* DSM 4455

Exobasidium reticulatum Ito & Sawada Thea sinensis L. S AF 487394* DSM 4520

Exobasidium rhododendri (Fuck.) Cram. Rhododendron ferrugineum L. H, S AF 009858B R.B. 2050

Exobasidium rostrupii Nannf. Vaccinium oxycoccos L. H, S AF 009857B R.B. 949

Exobasidium shiraianum Henn. Rhododendron degronianum S AF 487395* DSM 4522Carr.

Exobasidium sp. Rhododendron sp. H F.O. 39680

Exobasidium sundstroemii Nannf. Andromeda polifolia L. H, S AF 487396* R.B. 2051

Exobasidium symploci-japonicae Kusano & Symplocus sp. S AF 487397* DSM 4523Tokubuchi

Exobasidium vaccinii (Fuck.) Woronin 1 Vaccinium vitis-idaea L. H, S AF 009858B R.B. 945

Exobasidium vaccinii (Fuck.) Woronin 2 Vaccinium vitis-idaea L. H, S AF 487398* R.B. 2073

Exobasidium yoshinagai Henn. Rhododendron reticulatum S AF 487399* DSM 4524D. Don ex G. Don

Graphiola cylindrica Kobayasi palm S AF 487400* JCM 8561

Graphiola cylindrica Kobayasi palm H R.B. 30838

Graphiola phoenicis (Moug.) Poiteau Phoenix canariensis Chaub. H, S AF 009862B F.O. 29350

Kordyana celebensis Gäum. Commelina sp. S AF 487401* HB 17

Kordyana tradescantiae (Pat.) Rac. Tradescantia sp. H, S AF 487402* F.O. 47147

Laurobasidium lauri (Geyler) Jülich Laurus azorica (Seub.) Franco H, S AF 487403* M.P. 2371

Muribasidiospora hesperidium (Maire) Rhus glaucescens A. Rich. H IMI 5074Kamat & Rajendren

Muribasidiospora indica Kamat & Rajendren Rhus lancea E. Mey. H, S AF 352058BE F.O. 47397ex Harv. & Sond.

Proliferobasidium heliconiae Cunningham Heliconiae bihai L. H BPI 726024

Exobasidium pachysporum was prepared by high pressurefreezing and freeze substitution. Infected areas of leaves wereremoved with a 2 mm cork borer. To remove air from inter-cellular spaces, samples were infiltrated with distilled watercontaining 6 % (v/v) (2.5 M) methanol for approximately 5min at room temperature. Single samples were placed in analuminium holder (one half with a hollow of 0.3 mm depthfor the sample and the other a flat top) and frozen immedia-tely in the high pressure freezer HPM 010 (Balzers Union,Lichtenstein) as described in detail by MENDGEN et al. (1991).Substitution medium (1.5 ml per specimen) consisted of 2%osmium tetroxide in acetone which had been dried over cal-cium chloride. Freeze substitution was performed at – 90 °C,– 60 °C, and – 30 °C, 8h for each step, using a Balzers freezesubstitution apparatus FSU 010. The temperature was then rai-sed to approximately 0 °C over a 30 min period and sampleswere washed in dry acetone for another 30 min. Infiltrationwith an Epon/Araldite mixture (WELTER, MÜLLER & MEND-GEN 1988) was performed stepwise: 30% resin in acetone at4 °C for 7 h, 70 % and 100 % resin at 8 °C for 20 h each and100% resin at 18 °C for approximately 12h. Samples werethen transferred to fresh medium and polymerized at 60 °Cfor 10 h. Finally, samples were processed as for chemicallyfixed samples described above except that the sections wereadditionally stained with 1 % aqueous uranyl acetate for 1 h.

Molecular data were obtained using the same methods asdescribed earlier (BEGEROW, BAUER & OBERWINKLER 1997).

We used the 5’ terminal region of the nuclear large subunitrRNA gene to build an alignment of 534 bp. The alignmentwas produced using Clustal X (JEANNMOUGIN et al. 1998) andoptimized visually. Because of problems in the alignment thepositions 35-42 and 373-385 were excluded in the phyloge-netic analyses. Phylogenetic analyses were done with PAUP®

4.0b8a (SWOFFORD 1998). Maximum parsimony was carriedout in three steps. First, a heuristic search was done with100.000 random sequence additions without branch swappingto search for the best islands. Thereafter we performed TBRbranch swapping over the 17 best trees from the first step,which resulted in 104 most parsimonious trees. The consen-sus tree from the 104 trees with 757 steps was computed. Fi-nally, bootstrap values were calculated for 1000 replicates with10 random sequence additions without branch swapping ineach replicate. Modeltest 3.0 (POSADA & CRANDALL 1998)was carried out to determine a model of DNA substitution thatfits the data set and TIMIG was selected from the Akaike In-formation Criterion (base frequencies: πA = 0.2673, πC =0.1741, πG = 0.3026, πT = 0.2326; substitution rates: A/C =G/T = 1.0000, A/G = 2.4778, A/T = C/G = 0.7648, C/T =6.4718; gamma shape parameter = 0.7648; percentage of in-variant sites = 0.3651). Neighbor-joining analysis was doneusing genetic distances according to the specified substitu-tion model (see SWOFFORD et al. 1996). 10.000 replicates wereused for bootstrap analysis. The alignment is available uponrequest.

Mycological Progress 1(2) / 2002 189

© DGfM 2002

2. Other Ustilaginomycetes

Entyloma microsporum (Unger) Schröter Ranunculus repens L. S AF 007530B F.O. 37329

Georgefischeria riveae Thirum. & Naras. Rivea hypocrateriformis S AF 009861B HUV 15614Chois

Microstroma juglandis (Bereng.) Sacc. Juglans regia L. S AF 009867B F.O. 39211

Rhamphospora nymphaeae D. D. Cunn. Nymphaea alba L. S AF 007526B R.B. 862

Tilletia caries (DC.) Tul. Triticum aestivum L. S AJ 235308BO CBS 160.85

Ustilago hordei (Pers.) Lagerh. not cited S L 20286BER D.M. 11.2C

1) H = Hyphal septation, cellular host-parasite interaction and sporulation; S = Sequence 2) Origin of sequences: B = BEGEROW, BAUER & OBERWINKLER (1997), BE = BEGEROW, BAUER & OBERWINKLER (2001), BER = BERRES, SZABO &MCLAUGHLIN (1995), BO = BOEKHOUT, FELL & O´DONNELL (1995), * = new sequences3) BPI = US National Fungus Collections, Beltsville, USA; CBS = Culture collection of the Centraalbureau voor Schimmelcultures, Baarn, TheNetherlands; DSM = Deutsche Sammlung für Mikroorganismen und Zellkulturen GmbH, Braunschweig, Germany; D.M. = D. Hills, Corval-lis, USA; HB = Institut für angewandte Mikrobiologie, Wien, Austria; HUV = Herbarium Ustilaginales Vánky, Tübingen, Germany; F.O. = Her-barium F. Oberwinkler, Tübingen, Germany; IMI = Herbarium of the CABI Bioscience UK Centre, Egham, England; M = Botanische Staats-sammlung, München, Germany; M.P. = Herbarium M. Piepenbring, Frankfurt, Germany; R.B. = Herbarium R. Bauer, Tübingen, Germany4) Material obtained from H. D. Frey, Tübingen, Germany5 ) Material obtained from M. Kakishima, Tsukuba, Japan 6 ) Material obtained from J. P. van der Walt, Pretoria, South Africa7) Material obtained from R. Berndt, Tübingen, Germany8) Material obtained from R. J. Bandoni, Vancouver, Canada

Tab. 1. continued

Results

Septal pore architectureSeptal pore apparatus in the species studied (see Table 1) werecomposed of simple pores with more or less rounded lips en-closed by a membrane cap at either side of the pores. In addi-tion, the pore channels themselves were narrowed by electron-opaque tubes (Figs. 1-2).

Cellular interactionAll species studied interacted with their respective hosts bythe formation of complex interaction apparatus containing in-teraction rings (Fig. 3-5, for a detailed description of this typeof interaction apparatus see BAUER, OBERWINKLER & VÁNKY

1997). In Arcticomyces, Brachybasidium, Dicellomyces, Exo-basidiellum, Exobasidium arescens, E. pachysporum, E.rostrupii, Kordyana, Muribasidiospora, and Proliferobasi-dium these interaction sites were located in intercellular hyphaeattached with host cells (Figs. 3-4), whereas in Botryoconis,Clinoconidium, Coniodictyum, Drepanoconis, Exobasidiumkarstenii, E. myrtilli, E. oxycocci, E. sp. F.O. 39680 (Fig. 5),E. sundstroemii, E. vaccinii, Graphiola, and Laurobasidiumthey were located in haustoria. Haustoria were not constrictedat the penetration point. Usually, they extended a short dis-tance into the host cells where several aseptate short lobeswere formed (Fig. 5). Haustoria in Graphiola, but not in theother taxa studied, consisted of a clamped basal body (Fig. 6)with several lobes extending into the host cell.

SporulationThe species studied of Exobasidium (Fig. 7), Muribasidio-spora (Figs. 8-9), Laurobasidium (Fig. 10), Arcticomyces,Brachybasidium (Fig. 13), Dicellomyces, Exobasidiellum (Fig.14), Kordyana (Fig. 15), and Proliferobasidium sporulated onthe surface of the host organs. The basidia protruded throughstomata or emerged from the disintegrated epidermis with

elongate basidia. In Arcticomyces, Exobasidium (Fig. 7), Muri-basidiospora (Fig. 8), and Laurobasidium (Fig. 10), the hilarappendices of the basidiospores were oriented abaxially at theapex of the basidia, whereas in Brachybasidium (Fig. 13), Di-cellomyces, Exobasidiellum (Fig. 14), Kordyana (Fig. 15), andProliferobasidium they were oriented in the opposite direc-tion. In Arcticomyces, Exobasidium, and Laurobasidium, thenumber of sterigmata per basidium were not fixed, varyingfrom two to eight or more with four as the most frequentnumber. The species of Dicellomyces, Muribasidiospora(Figs. 8-9), Brachybasidium (Fig. 13), Exobasidiellum (Fig.14), and Kordyana (Fig. 15), however, formed generally two-sterigmate basidia. The basidiospores in Exobasidium, Muri-basidiospora, Arcticomyces, Brachybasidium, Dicellomyces,Exobasidiellum, Kordyana, and Proliferobasidium, but not inLaurobasidium, were forcibly discharged. Surprisingly, al-though the basidia in Laurobasidium are clearly of the exo-basidial type (compare Fig. 7 with Fig. 10), they apparentlylack the ballistosporic mechanism. At least in our basidio-spore-fall experiments no spores could be observed on theagar. Probasidial swellings were seen in Muribasidiospora,Laurobasidium, Brachybasidium and Exobasidiellum. In Lau-robasidium, basidia with and without probasidial swellingswere present (Fig. 10).

In contrast with the taxa discussed above, the studied spe-cies of Botryoconis, Coniodictyum (Fig. 11), Clinoconidium(Fig. 12), and Drepanoconis sporulated internally by producingholobasidia in peripheral lacunae of the host galls. During ma-turation, the galls ruptured and liberated the basidiosporemass. The basidia were gastroid without sterigmata. The basi-diospores were usually thick-walled, resembling the uredo-spores of rust fungi or the teliospores of smut fungi. In addi-tion, old fructifications often resembled smut sori.

The two Graphiola species studied formed distinct cylind-rical basidiocarps, in which globose basidia were produced inchains. The passively released basidiospores arose laterallyon the basidia (Fig. 16).


© DGfM 2002

Figs. 1-6. Septal pore apparatus and cellular interaction of some Exobasidiales. Material illustrated in Figs. 1-4 was preparedusing freeze substitution. Fig. 1. Simple septal pore with two membrane caps (arrows) of Exobasidium pachysporum. Notethe tube within the pore (arrowheads). Bar = 0.1 µm. Fig. 2. Median section through a septal pore apparatus of Exobasidiumpachysporum illustrated to show the membrane caps (arrows) and the tube within the pore channel (arrowheads) in detail.The separation of the lower membrane cap from the plasma membrane may represent an artifact of preparation. Bar = 0.1 µm.Fig. 3. Intercellular hypha (ih) of Exobasidium pachysporum in contact with host cell wall (HW) showing a local interactionsite with the exocytosis profile of the cisternal interaction apparatus (arrow) and the interaction ring (arrowheads, three-dimen-sional configuration reconstructed from serial sections). Note the electron-opaque deposit at the host side. Bar = 1 µm. Fig. 4.Intercellular hypha (ih) of Exobasidium pachysporum in contact with host cell wall (HW) illustrated to show the local interac-tion site with the exocytosis profile of the interaction apparatus (arrows), the interaction ring (arrowheads, three-dimensionalconfiguration reconstructed from serial sections) and the electron-opaque deposit (d) between host cell wall and host plasmamembrane in detail. Bar = 0.1 µm. Fig. 5. Short-lobed haustorium of Exobasidium sp. F.O. 39680 with three interaction sites(1, 2, 3) and two interaction apparatus (arrows). Note the sectioned interaction ring in the interaction site 1 (arrowheads, three-dimensional configuration reconstructed from serial sections) and the electron-opaque deposits (d) in the interaction sites 2 and3. The point of penetration the host cell is visible at double arrowheads. Bar = 1 µm. Fig. 6. Section through a haustorium ofGraphiola phoenicis with clamp (arrow) at the base. Bar = 1 µm.

Figs. 7-12. Sori and spores of selected members of the Exobasidiales. Fig. 7. Sorus of Exobasidium oxycocci F.O. 18925 onVaccinium oxycoccos. Note the orientation of the basidiospores on the basidium. Fig. 8. Basidial apex of Muribasidiospora in-dica demonstrating the abaxial orientation of the hilar appendices of the basidiospores. Fig. 9. Basidial layer with hyphidia-likesterile elements of Muribasidiospora hesperidium. Fig. 10. Basidial layer of Laurobasidium lauri illustrating the exobasidia-ceous orientation of the basidiospores on the basidia. Note that some basidia, but not all, have probasidial swellings. Fig. 11.Basidial layer of Coniodictyum chevalieri on Zizyphus mucronata. Note the lack of sterigmata on the basidia. Fig. 12. Basidiallayer with sterile paraphyses of Clinoconidium cf. bullatum. Note the lack of sterigmata on the basidia. Note also that the ger-minating basidiospores are aseptate.


© DGfM 2002

Figs. 13-16. Sori and spores of selected members of the Exobasidiales. Fig. 13. Sorus of Brachybasidium pinangae on Pinangakuhlii. Note the probasidial swellings. Fig. 14. Basidial layer of Exobasidiellum graminicola. Fig. 15. Basidia of Kordyanatradescantiae demonstrating the adaxial orientation of the hilar appendices of the basidiospores. Fig. 16. Basidial chains ofGraphiola phoenicis. Note that mature basidiospores are septate and that the two segments of mature basidiospores become se-parated from each other during germination.


© DGfM 2002

In Botryoconis and Clinoconidium (Fig. 12), but not in theother species studied (Figs. 7, 9-11, 13-15), the basidiosporesremained aseptate during germination.

Sequence analyses

Heuristic maximum parsimony analysis resulted in 104 mostparsimonious trees after exhaustive swapping. The majorityrule consensus topology is illustrated in Fig. 17. Most bifur-cations were not strongly supported by bootstrap values with-in the maximum parsimony analysis, although all illustratedfurcations were supported with at least 75% in the consensustree of the 104 most parsimonious trees. The neighbor-joininganalysis resulted in a similar arrangement of the species be-longing to the Exobasidiales (Fig. 18). The monophyly of themembers of the Exobasidiales was supported in this hypo-thesis with a bootstrap value of 79%. The same analysisgrouped the members of the Exobasidiales in the expectedfamilies (BAUER et al. 2001a). In contrast with the neighbor-joining analysis (Fig. 18), however, in the maximum parsi-mony analysis (Fig. 17) the Brachybasidiaceae did not forma monophyletic group. The Cryptobasidiaceae, Exobasidia-ceae, and Graphiolaceae were supported in both analyses andthe monophyly of the Exobasidiaceae was well supported bybootstrap resampling in the neighbor-joining analysis (Fig.18).

Within the Exobasidiaceae a few clusters are well sup-ported by neighbor-joining bootstrap values. The groups onTheales and Vaccinioideae are supported by 67 % and 78 %,respectively (Fig. 18). The main groups are also visible in themaximum parsimony analysis, but not as well supported as inthe distance method. This inner topology of the Exobasidia-ceae was consistent with neighbor-joining analysis based onother substitution models (e. g. the Kimura-2-parameter or thegeneral time reversible model; data not shown).

Discussion

Phylogenetic aspects

The species studied here have an essentially identical septalpore apparatus: the simple pores are enclosed at both sides bymembrane caps and the pore channels are narrowed by tubes.Among the basidiomycetes, the Exobasidiales share the pre-sence of membrane caps enclosing the simple pores with theEntylomatales, Microstromatales and Doassansiales of theExobasidiomycetidae, and also with the Urocystales of theUstilaginomycetidae (BAUER, OBERWINKLER & VÁNKY 1997).Among these taxa, the Exobasidiales share the presence oftubes within the pore channels only with the Doassansiales.Accordingly, the Doassansiales may represent the sister groupof the Exobasidiales, as already discussed by BAUER, OBER-WINKLER & VÁNKY (1997). In addition, the interaction appa-ratus of the Doassansiales resembles that of the Exobasidiales,but differs predominantly in the absence of interaction rings.

Thus, complex interaction apparatus with the formation of in-teraction rings characterize the Exobasidiales. The membersof the Exobasidiales are holobasidiate and dimorphic. Theydo not form teliospores in the parasitic phase and ballistoco-nidia in the saprobic phase. In most of the species, the basi-diospores become septate during germination.

The molecular results are consistent with this grouping. Inboth maximum parsimony and neighbor-joining analyses theexobasidealean species form a group that is statistically sup-ported with 54 % and 79 %, respectively. Interestingly, theExobasidiales were better supported by bootstrap values inearlier studies (BEGEROW, BAUER & BOEKHOUT 2000; BEGE-ROW, BAUER & OBERWINKLER 1997, 2001; BAUER et al. 1999;2001a) than in the present analyses, possibly due to the largernumber of non-exobasidealean taxa in the former studies.

Our morphological analyses indicate the existence of fourgroups within the Exobasidiales. Except for Laurobasidium(see below), this phylogenetic interpretation is not in conflictwith the molecular results.

(i) The Exobasidiaceae comprise Arcticomyces, Exobasi-dium and Muribasidiospora (HENNINGS 1900; MIMS, RICHARD-SON & ROBERSON 1987; RAJENDREN 1968; NANNFELDT 1981;OBERWINKLER 1977, 1978, 1982, 1987, 1993; SAVILE 1959a,b). Among the Exobasidiales, the abaxial orientation of thehilar appendices of the ballistosporic basidiospores on the ba-sidia characterizes this group (see Figs. 7-8). However, thisspecific exobasidiaceous basidiospore orientation occurs alsoin members of the Tilletiales, Georgefischeriales and Doassan-siales (BAUER et al. 1999, 2001b; INGOLD 1995; VÁNKY &BAUER 1996). Therefore, we consider this basidial type as apo-morphic for the Exobasidiomycetidae and, therefore, plesio-morphic for the Exobasidiales. Bootstrap support for the Exo-basidiaceae is smaller than 50 % in maximum parsimony, and78 % in neighbor-joining analysis. The Exobasidiaceae sporu-late through stomata or from the disintegrated epidermis, thebasidia are elongate and ballistosporic, and the basidiosporesare thin-walled. In most Exobasidiaceae species, the numberof sterigmata per basidium is not fixed, varying from two toeight or more with four as the most frequent number. Only afew species form generally two-sterigmate basidia. Dependingupon the species, haustoria are present or absent. Interestingly,however, the Exobasidium species studied that lack haustoria(E. arescens, E. rostrupii, and E. pachysporum) are located inboth molecular analyses on a statistically well supportedbranch (see Figs. 17-18).

(ii) The Cryptobasidiaceae comprise Botryoconis, Clino-conidium, Coniodictyum, Drepanoconis and Laurobasidium(DONK 1956; HENDRICHS, BAUER & OBERWINKLER 2002;LENDNER 1920; MALENÇON 1953; MAUBLANC 1914; OBER-WINKLER 1978; 1982, 1993; SYDOW 1926). All species of thisfamily studied form haustoria. Except for Laurobasidium, theCryptobasidiaceae sporulate internally by producing holo-basidia in peripheral lacunae of the host galls. During matu-ration, the galls rupture and liberate the basidiospore mass.The basidia are gastroid and lack sterigmata. The basidiospo-


© DGfM 2002

res are usually thick-walled, resembling the uredospores ofrust fungi or the teliospores of smut fungi. In addition, oldfructifications often resemble smut sori. Surprisingly, in bothmolecular analyses Laurobasidium lauri appears within thecluster representing the Cryptobasidiaceae (Figs. 17-18). Inaddition, in both analyses the union of Laurobasidium and Cli-noconidium ssp. that represent the cryptobasidealean coregroup in the analyses is statistically supported by 100%. Incontrast with the other members of the Cryptobasidiaceae,Laurobasidium lauri sporulates on the surface of the host or-gans and the basidia resemble strongly those of Exobasidium,but they are apparently gastroid like in the other members ofthe Cryptobasidiaceae. In addition, Laurobasidium parasitizeslaurels as the other members of the cryptobasidealean coregroup. Thus, Laurobasidium mediates morphologically bet-ween the Exobasidiaceae and Cryptobasidiaceae.

(iii) In interpreting the morphological, ultrastructural, andmolecular data, the Brachybasidiaceae comprise Brachybasi-dium, Dicellomyces, Exobasidiellum, Kordyana, and Proli-ferobasidium (BANDONI & JOHRI 1975; BERNDT & SHARMA

1998; CUNNINGHAM, BAKSHI & LENTZ 1976; GÄUMANN 1922;OBERWINKLER 1978, 1982, 1993; OLIVE 1945). The studiedspecies of this family lack haustoria. Like the Exobasidiaceae,the Brachybasidiaceae sporulate on the surface of the host or-gans. The basidia protrude through stomata or emerge fromthe disintegrated epidermis. The basidia are elongate, ballisto-sporic and two-sterigmate. In contrast with the Exobasidia-ceae, however, available data indicate that the hilar appendicesof the basidiospores are oriented adaxially at the apex of thebasidia (see Fig. 15 in the present study; Figs. 2, 6, 13, 17 inCUNNINGHAM, BAKSHI & LENTZ 1976; Figs. 1.10-2, 1.10-3in OBERWINKLER 1982; Fig. 4 in OBERWINKLER 1993, Fig. 1-Gin INGOLD 1985). Only in neighbor-joining analysis thebrachybasidiaceous representatives appear as a monophyleticgroup, whereas the maximum parsimony topology shows aparaphyletic situation for this family.

(iv) The Graphiolaceae comprise Graphiola and Stylina(FISCHER 1921, 1922; OBERWINKLER et al. 1982). This groupdiffers significantly from the other exobasidealean speciesin the formation of distinct basidiocarps, the production of ba-sidia in chains, the formation of clamped haustoria, and inthe basidial morphology. Fructification starts between thechlorenchyma and hypodermal tissue (COLE 1983). Duringdifferentiation of the cylindrical basidiocarp, the epidermisruptures and globose basidia are produced in chains within thebasidiocarps. The passively released basidiospores arise late-rally on the basidia.

The role of the host in the exobasidealean evolutionThe present results indicate that the host phylogeny plays adominant role in the exobasidealean evolution. There is an ob-vious phylogenetic correlation between the Exobasidiales andtheir respective hosts. Thus, with a few exceptions (see be-low) the Exobasidiaceae live on Ericanae, the Cryptobasidia-ceae on laurels, the Brachybasidiaceae on monocots, and the

Graphiolaceae on palms. In addition, the arrangement of thesefour groups in our molecular trees, though not always wellsupported by bootstrap values, essentially reflects the respec-tive host phylogenies (see APG 1998). In fact, in correlationwith the position of the Lauraceae in molecular phylogeneticanalyses of angiosperms, the Lauraceae-parasitizing Crypto-basidiaceae stand well apart from both the monocots-parasi-tizing Brachybasidiaceae and Graphiolaceae and the eudicots-parasitizing Exobasidiaceae (see Figs. 17-18). In addition, likethe mono- and eudicots in molecular phylogenetic studies ofangiosperms (e. g., APG 1998), the monocots-parasitizingBrachybasidiaceae and Graphiolaceae are separated from theeudicots-parasitizing Exobasidiaceae. In this respect the ma-ximum parsimony analysis differs slightly from the neighbor-joining analysis where the monocots-parasitizing Brachyba-sidiaceae and Graphiolaceae form sister groups (compare Fig.17 with Fig. 18).

In summary, the topology of the four exobasidiaceousgroups discussed above in our molecular trees and their res-pective host distribution can best be interpreted as a result ofcospeciation or association by descent (BROOKS 1988). How-ever, it is likely that the Exobasidiales ancestors have not onlyundergone a period of parallel cladogenesis. The occurrenceof Arcticomyces on Saxifraga (Saxifragaceae), Coniodictyumon Zizyphus (Rhamnaceae) as well as Muribasidiospora onRhus (Anacardiaceae) may reflect jumps to new hosts or re-lictual coevolution (BROOKS & BANDONI 1988). Because theLauraceae and their relatives respresent one of the oldest an-giosperm groups we assume that the Exobasidiales arose aspathogens on early angiosperms.

An evolutionary hypothesis for Exobasidium

The inherent topology of the Exobasidiaceae in our moleculartrees, while not always well supported by bootstrap analysis,essentially correlates with that of the respective hosts (for thehost phylogeny see, e. g., KRON & CHASE 1993, KRON, POWELL

& LUTEYN 2002). In interpreting this correlation, the followingevolutionary scenario for Exobasidium may be plausible: Exo-basidium arose as a pathogen on the Ericanae ancestor. In thefollowing the fungus and its descendants cospeciated for along period. This would explain the basal dichotomy occur-ring between the Exobasidium species on Theales on the onehand and those on the Ericaceae on the other in the neighbor-joining analysis (Fig. 18), and the subsequent dichotomy bet-ween the Exobasidium species on Rhododendron on the onehand and those on Vaccinioideae on the other. Although inmaximum parsimony analysis the Exobasidium species oc-curring on Vaccinioideae do not form a monophylum this ana-lysis is not in conflict with the interpretation (see Fig. 17).

It is clear from our molecular results, however, that thejoint evolution of Exobasidium and members of the Vacci-nioideae was more complex and not only a process of cospe-ciation. Thus, on the one hand only distantly related Exobasi-dium species occur on the same host species, e. g. E. arescens


© DGfM 2002

Fig. 17. Majority-rule consensus tree of 104 most parsimonious trees (757 steps found in heuristic search) calculated from par-tial nuclear LSU sequences, rooted with Ustilago hordei. Percentage bootstrap values of 1.000 replicates are given at eachfurcation. Values smaller than 50% are not shown. The letters refer to the following host groups: E = Ericaceae, L = Lauraceae,M = monocots, R = Rhododendron, Ru = Rhus, S = Saxifraga, T = Theales, V = Vaccinioideae, Z = Zizyphus.

© DGfM 2002


Graphiolaceae

Brachybasidiaceae

Cryptobasidiaceae

Exobasidiaceae

Fig. 18. Topology obtained by neighbor-joining analysis of partial nuclear LSU sequences using TIMIG distances, rooted withUstilago hordei. Percentage bootstrap values of 10.000 replicates are given at each furcation. Values smaller than 50% are notshown. Branch lengths are scaled in terms of expected numbers of nucleotide substitutions per site. The letters refer to the fol-lowing host groups: E = Ericaceae; L = Lauraceae, M = monocots, R = Rhododendron, Ru = Rhus, S = Saxifraga, T = Theales,V = Vaccinioideae, Z = Zizyphus.

© DGfM 2002


Graphiolaceae

Brachybasidiaceae

Cryptobasidiaceae

Exobasidiaceae

© DGfM 2002


and E. myrtilli on Vaccinium myrtillus, E. rostrupii and E. oxy-cocci on Vaccinium oxycoccos, or E. karstenii and E. sund-stroemii on Andromeda polifolia (see Table 1). On the otherhand, in both molecular analyses there is a statistically wellsupported cluster (E. arescens group) of Exobasidium specieson different hosts that all cause local infections on leaves, whe-reas the species outside this cluster cause hypertrophies onshoots or are at least able to do so as in the case of E. vaccinii.As already noticed by NANNFELDT (1981), E. vaccinii cancause both local infections on leaves and hypertrophies onshoot tips. Thus, the two specimens of E. vaccinii used in ourmolecular analyses also differed in the infection mode: spe-cimen 1 caused a local infection on a leaf, whereas specimen2 caused a hypertrophy on a shoot tip. In general, the topologyof the Exobasidium species on Vaccinioideae in our molecu-lar trees suggests that besides cospeciation the change of thesporulation site was an additional mechanism of speciation.In other words, it is possible that on the Vaccinioideae at leasttwo (or more) Exobasidium lineages with different infectionmodes cospeciated independently from each other as a resultof duplication (ROY 2001). However, additional molecularphylogenetic analyses involving more species and also otherDNA regions are necessary to clarify definitely this fascina-ting phenomenon of joint evolution of Exobasidium and Vac-cinioideae.

Acknowledgements

We thank M. Lutz, U. Simon, and M. Weiß for critically rea-ding the manuscript and helpful discussions, the official her-baria and persons cited in Table 1 for specimens, K. Mendgenfor the use of the high pressure freezer, J. Götze, M. Wagner-Eha and F. Albrecht for technical assistance, and the DeutscheForschungsgemeinschaft for financial support.

ReferencesAPG (1998) An ordinal classification for the families of flowering

plants. – Annals of the Missouri Botanical Garden 85: 531-553.

BANDONI RJ, JOHRI BN (1975) Observation on the genus Exobasi-diellum. – Canadian Journal of Botany 53: 2561-2564.

BAUER R, BEGEROW D, OBERWINKLER F (1998) Fortschritte in derSystematik der Brandpilze. – Zeitschrift für Pflanzenkrank-heiten und Pflanzenschutz 105: 224-238.

BAUER R, OBERWINKLER F, VÁNKY K (1997) Ultrastructural markersand systematics in smut fungi and allied taxa. – Canadian Jour-nal of Botany 75: 1273-1314.

BAUER R, VÁNKY K, BEGEROW D, OBERWINKLER F (1999) Ustila-ginomycetes on Selaginella. – Mycologia 91: 475-484.

BAUER R, BEGEROW D, OBERWINKLER F, PIEPENBRING M, BERBEE

ML (2001a) Ustilaginomycetes. In MCLAUGHLIN DJ,MCLAUGHLIN EG, LEMKE PA (eds) Mycota VII Part B. Sys-tematics and evolution, pp. 57-83. Springer Verlag, Heidel-berg, New York.

BAUER R, BEGEROW D, NAGLER A, OBERWINKLER F (2001b) The Ge-orgefischeriales: a phylogenetic hypothesis. – MycologicalResearch 105: 416-424.

BEGEROW D, BAUER R, BOEKHOUT T (2000) Phylogenetic placementsof ustilaginomycetous anamorphs as deduced from nuclearLSU rDNA sequences. – Mycological Research 104: 53-60.

BEGEROW D, BAUER R, OBERWINKLER F (1997) Phylogenetic studieson large subunit ribosomal DNA sequences of smut fungi andrelated taxa. – Canadian Journal of Botany 75: 2045-2056.

BEGEROW D, BAUER R, OBERWINKLER F (2001) Muribasidiospora:Microstromatales or Exobasidiales? – Mycological Research105: 798-810.

BERNDT R, SHARMA ND (1998) Dicellomyces calami sp. nov. fromIndia. – Mycological Research 102: 1484-1486.

BLANZ P, DÖRING H (1995) Taxonomic relationships in the genusExobasidium (Basidiomycetes) based on ribosomal DNA ana-lysis. – Studies in Myology 38: 119-128.

BLANZ P, OBERWINKLER F (1983) A contribution to the species defi-nition in the genus Exobasidium (Basidiomycetes). – Syste-matic and Applied Microbiology 4: 199-206.

BROOKS DR (1988) Macroevolutionary comparisons of host andparasite phylogenies. – Annual Review of Ecology and Sys-tematics 19: 235-259.

BROOKS DR, BANDONI SM (1988) Coevolution and relicts. – Syste-matic Zoology 37: 19-33.

Cole GT (1983) Graphiola phoenicis: a taxonomic enigma. – Myco-logia 75: 93-116.

CUNNINGHAM JL, BAKSHI BK, LENTZ PL (1976) Two new genera ofleaf-parasitic fungi (Basidiomycetidae: Brachybasidiaceae).– Mycologia 68: 640-654.

DERX HG (1930) Étude sur les Sporobolomycètes. – Annales My-cologici 28: 1-23.

DONK MA (1956) The generic names proposed for hymenomycetes-VI. Brachybasidiaceae, Cryptobasidiaceae, Exobasidiaceae.– Reinwardtia 4: 113-118.

DÖRING H, BLANZ P (2000) 18S rDNA-Analysen bei der GattungExobasidium. – Hoppea 61: 85-100.

FISCHER E (1921) Zur Kenntnis von Graphiola and Farysia. – An-nales Mycologici 18: 118-197.

FISCHER E (1922) Weitere Beiträge zur Kenntnis der Gattung Gra-phiola. – Annales Mycologici 20: 228-237.

GÄUMANN E (1922) Über die Gattung Kordyana Rac. – AnnalesMycologici 20: 257-271.

Hendrichs M, Bauer R, Oberwinkler F (2002) The Cryptobasidia-ceae of the new world. Sydowia (subm.).

HENNINGS P (1900) Exobasidiineae. In: ENGLER A, PRANTL K (eds)Die natürlichen Pflanzenfamilien 1,1, pp. 103-105. Verlag vonWilhelm Engelmann, Leipzig.

INGOLD CT (1985) Dicellomyces scirpi: its conidial stage and taxo-nomic position. – Transactions of the British MycologicalSociety 84: 542-545.

INGOLD CT (1995) Products of teliospore germination in Tilletia hya-lospora. – Mycological Research 99: 1247-1248.

JEANNMOUGIN F, THOMPSON JD, GOUY M, HIGGINS DG, GIBSON TJ(1998) Multiple sequence alignment with Clustal X. – Trendsin Biochemical Sciences 23: 403-405.

KRON KA, CHASE MW (1993) Systematics of the Ericaceae, Em-petraceae, Epacridaceae and related taxa based upon rbcL se-quence data. – Annals of the Missouri Botanical Garden 80:735-741.

© DGfM 2002


KRON KA, POWELL EA, LUTEYN JL (2002) Phylogenetic relati-onships within the blueberry tribe (Vaccinieae, Ericaceae) ba-sed on sequence data from MATK and nuclear ribosomal ITSregions, with comments on the placements of Satyria. – Ame-rican Journal of Botany 89: 327-336.

LENDNER A (1920) Un champignon parasite sur une Lauracée dugenre Ocotea. – Bulletin de l`Herbier Boissier Genève 12:122-128.

MALENÇON G (1953) Le Coniodyctium chevalieri Har. et Pat., sa na-ture et ses affinités. – Bulletin de la Société Mycologique deFrance 69: 77-100.

MAUBLANC MA (1914) Les genres Drepanoconis Schr. et Henn. etClinoconidium Pat.: leur structure et leur place dans la clas-sification. – Bulletin de la Société Mycologique de France 30:444-446.

MENDGEN K, WELTER K, SCHEFFOLD F, KNAUF-BEITER G (1991)High pressure freezing of rust infected plant leaves. In MEND-GEN K, LESEMANN, DE (eds) Electron microscopy of plant pa-thogens, pp. 31-42. Springer Verlag, Heidelberg, New York.

MIMS CW, RICHARDSON EA, ROBERSON RW (1987) Ultrastructureof basidium and basidiospore development in three species ofthe fungus Exobasidium. – Canadian Jounal of Botany 65:1236-1244.

NANNFELDT JA (1981) Exobasidium, a taxonomic reassessment ap-plied to the European species. – Symbolae Botanicae Upsa-lienses 23/2: 1-71.

OBERWINKLER F (1977) Das neue System der Basidiomyceten. InFREY W, HURKA H, OBERWINKLER F (eds) Beiträge zur Bio-logie der niederen Pflanzen, pp. 59-105. Gustav Fischer Ver-lag, Stuttgart, Jena, New York.

OBERWINKLER F (1978) Was ist ein Basidiomycet? – Zeitschrift fürMykologie 44: 13-29.

OBERWINKLER F (1982) The significance of the morphology of thebasidium in the phylogeny of basidiomycetes. In WELLS K,WELLS EK (eds) Basidium and basidiocarp, pp. 9-35. Sprin-ger Verlag, New York, Heidelberg, Berlin.

OBERWINKLER F (1987) Heterobasidiomycetes with ontogenetic yeaststages - systematic and phylogenetic aspects. – Studies in My-cology 30: 61-74.

OBERWINKLER F (1993) Diversity and phylogenetic importance oftropical heterobasidiomycetes. In ISAAC S, FRANKLAND JC,WATLING R, WHALLEY AJS (eds) Aspects of tropical myco-logy, pp. 121-147. Cambridge University Press.

OBERWINKLER F, BANDONI RJ, BLANZ P, DEML G, KISIMOVA-HORO-VITZ L (1982). Graphiolales: basidiomycetes parasitic onpalms. – Plant Systematics and Evolution 140: 251-277.

OLIVE LS (1945) A new Dacrymyces-like parasite of Arundinaria.– Mycologia 37: 543-552.

POSADA D, CRANDALL KA (1998) Modeltest: testing the model ofDNA substitution. – Bioinformatics 14: 817-818.

RAJENDREN RB (1968) Muribasidiospora - a new genus of the Exo-basidiaceae. – Mycopathology 36: 218-222.

ROY BA (2001) Patterns of association between crucifers and theirflowering-mimic pathogenes: host jumps are more commonthan coevolution or cospeciation. – Evolution 55: 41-53.

SAVILE DBO (1959a) Notes on Exobasidium. – Canadian Journal ofBotany 37: 641-656.

SAVILE DBO (1959b) The Botany of Sommerset Island, district ofFranklin. – Canadian Journal of Botany 37: 959-1002.

SWOFFORD DL, OLSEN GJ, WADDELL PJ, HILLIS DM (1996) Phylo-genetic inference. In HILLIS DM, MORITZ C, MABLE BK (eds)Molecular systematics, pp 407-514. Sinauer Associates, Sun-derland, MA

SWOFFORD DL (1998) PAUP*. Phylogenetic analysis using parsi-mony (*and other methods). Version 4. Sinauer Associates,Sunderland, MA.

SYDOW H (1926) Fungi in itinere costaricensi collecti. Pars secunda.– Annales Mycologici 24: 283-288

VÁNKY, K, BAUER R (1996) Ingoldiomyces, a new genus of Ustila-ginales. – Mycotaxon 59: 277-287.

WELTER K, MÜLLER M, MENDGEN K (1988) The hyphae of Uromy-ces appendiculatus within the leaf tissue after high pressurefreezing and freeze substitution. – Protoplasma 147: 91-99.

Accepted: 13.3.2002

the exobasidiales: an evolutionary hypothesis 1 · 2018. 6. 3. · n the new system of...

Documents