Inter- and intraspecific variation of homologous hydrophobin (H1) gene sequences among Xanthoria spp. (lichen-forming ascomycetes)

Authors

  • Sandra Scherrer,

    1. Institute of Plant Biology, University of Zurich, Zollikerstrasse 107, CH–8008 Zurich, Switzerland
    Search for more papers by this author
  • Rosmarie Honegger

    Corresponding author
    1. Institute of Plant Biology, University of Zurich, Zollikerstrasse 107, CH–8008 Zurich, Switzerland
      Author for correspondence: Rosmarie Honegger Tel: +41 1634 8243 Fax: +41 1634 8204 Email: rohonegg@botinst.unizh.ch
    Search for more papers by this author

Author for correspondence: Rosmarie Honegger Tel: +41 1634 8243 Fax: +41 1634 8204 Email: rohonegg@botinst.unizh.ch

Summary

  • • Inter- and intra-specific variability of the hydrophobin gene H1 was investigated in a range of Xanthoria species collected in their natural habitat.
  • • On Southern blots the XEH1 probe (from Xanthoria ectaneoides) hybridized under standard conditions to the genomic DNA of Xanthoria calcicola, Xanthoria parietina and Xanthoria flammea. The corresponding H1 genes in these and additional species (Xanthoria turbinata, Xanthoria spp.), and ribosomal gene regions (ITS 1 and 2, 5.8S rDNA) were amplified with polymerase chain reaction, sequenced and similarities calculated. Restriction fragment length polymorphism and cleaved amplified polymorphic sequence analyses provided molecular tools for species delimitation.
  • • Ribosomal gene data were largely in parallel with the hydrophobin H1 sequence data, the protein-coding H1 sequence being even more variable as the noncoding rDNA regions.
  • • Very low intraspecific variation was detected in the H1 gene of X. parietina samples collected on four continents, with the highest variability occurring among Mediterranean samples. Xanthoria ectaneoides from Brittany, France, and X. calcicola appear to be coastal and inland forms of the same fungal species but differ from Sicilian X. ectaneoides. The African X. flammea and X. turbinata are related to the X. parietina cluster.

Introduction

Populations of golden yellow to orange lichens of the genus Xanthoria, as found from the coast to mountain areas, are often conspicuous landmarks which indicate nutrient-rich sites. In Europe, Xanthoria parietina, symbiotic with unicellular green algae of the genus Trebouxia, is one of the most common lichen-forming ascomycetes and certainly one of the best investigated (Honegger, 1990, 1996). As a heteromerous macrolichen, the symbiotic phenotype of X. parietina comprises two structurally and functionally distinct zones: hydrophilic, conglutinate pseudoparenchyma as peripheral upper and lower cortical layers, which passively absorb water and dissolved mineral nutrients and provide the thallus with elasticity and mechanical stability; and the gas-filled thalline interior, built up by a system of loosely interwoven aerial hyphae with hydrophobic wall surfaces, some of which contact the algal cells by means of intraparietal haustoria (Honegger, 1990). Sealing of the algal wall surface and thus of the apoplastic continuum with mycobiont-derived hydrophobic cell wall surface components was shown to be an essential feature for thalline water relations and for the symbiotic relationship per se (Honegger, 1993, 1997; Scherrer et al., 2002; for review see Dyer, 2002).

Previous studies focused on the fine structure, biochemistry and molecular genetics of wall surface hydrophobicity in X. parietina and in the closely related Xanthoria ectaneoides, which was shown to be hydrophobin-based (Scherrer et al., 2000, 2002). Hydrophobins are small (100 ± 25 amino acids) secreted fungal proteins with very low amino acid homology except for eight cysteines in a conserved pattern (Wessels, 1997, 1999; Wösten & Wessels, 1997; Kershaw & Talbot, 1998; Wösten, 2001) and with interfacial self-assembly into an amphiphilic protein film with distinct, semicrystalline rodlet pattern (Wösten et al., 1993).

The present study is the first approach to investigate the intra- and inter-specific variation of the hydrophobin gene H1 in a group of closely related lichen-forming ascomycetes, the focus being on the genus Xanthoria. The two class 1 hydrophobins so far known from lichen-forming ascomycetes, XPH1 (X. parietina) and XEH1 (X. ectaneoides), revealed a high homology and were shown to be encoded by the fungal, not by the algal partner of the lichen thallus (Scherrer et al., 2000); high gene expression was detected in medullary hyphae (Scherrer et al., 2002). The mature hydrophobin XEH1 of X. ectaneoides was isolated from washed cell wall fragments from specimens collected on coastal rocks in Roscoff (Brittany, France). In the present study, the XEH1 probe and primers designed for the amplification of XPH1 were used for testing the interspecific variation of the H1 gene in a range of Teloschistaceae (lichen-forming ascomycetes). Because of considerable phenotypic plasticity, the species delimitation in the type species of this common and widespread group of macrolichens is rather vague, conventional taxonomy being based on features of the symbiotic phenotype such as morphotype and chemotype. Therefore, parts of the rDNA gene region (ITS1, 5.8S rDNA, and ITS2) were analysed in a series of parallel experiments.

Materials and Methods

Collection of specimens

Thalli of representatives of the Teloschistaceae, predominantly Xanthoria spp., from different collecting sites, were examined (Table 1; see Fig. 2). Voucher specimens are deposited in the personal lichen collection of R. Honegger (Institute of Plant Biology, University of Zurich, Switzerland). The majority of specimens used in this project are stored in the desiccated state at −20°C, where they retain their full viability over prolonged periods.

Table 1.  Collecting sites of the lichen samples and database accession numbers
Species and areaCollecting site and yearCollected byDet.3Voucher numberDatabase accession No. of H1 gene4Database accession No. of rDNA region4
  • 1

    Specimens stored at room temperature; all others were kept at −20°C, where they retain full viability.

  • 2

    2 No material left (photographic documentation)

  • 3

    3 Det., determined by: C, the collector; 1, R. Honegger; 2, P. W. James; 3, P. L. Nimis.

  • 4

    4 NP, no product obtained in polymerase chain reaction; ND: not determined.

Xanthoria parietina (L.) Th. Fr.
Central EuropeZurich, Switzerland, 2000S. ScherrerC98AJ287227AJ320118
Winterthur, Switzerland, 2000S. ScherrerC88AJ320091AJ320146
Western EuropeRoscoff, Brittany, France, 1998R. HoneggerC2AJ287226ND
Le Fret, Brittany, France, 2000R. HoneggerC91/1
91/2
AJ320094
AJ320094
AJ320148
AJ320148
Cudden Point, Penzance, Cornwall, UK, 2000J. Gray1106AJ320092AJ320136
Mt. Louis, Cerdagne, France, 1990R. HoneggerC60AJ320107AJ320120
Irchester, Northamptonshire, UK, 2000J.-J. Pittet118AJ320099AJ320129
Northern EuropeTjörn, Bohuslän, Sweden, 2000S. ScherrerC99AJ320102AJ320137
Ljugarn, Gotland, Sweden, 2000S. ScherrerC97AJ320103AJ320128
Oslo, Norway, 2000T. TonsbergC16AJ320100AJ320126
Oulunsallo, Finland, 2000P. Ervasti189AJ320108AJ320127
Southern EuropeLo Zingaro, Sicily, Italy, 1995R. HoneggerC4AJ320093AJ320147
North AmericaBangor, ME, USA, 2000J. HindsC26

27
AJ320095

AJ320096
AJ320121

AJ320122
Willamette Valley, Corvallis, OR, USA, 2000B. Mc CuneC6AJ320097AJ320125
Bay Bulls, Newfoundland, Canada, 2000M. TrembleyC100AJ320101AJ320119
AustralasiaStonor, Tasmania, Australia, 2000G. KantvilasC10AJ320098AJ320144
Roxburgh, Central Otago, New Zealand, 2000D. J. GallowayC1/1
1/2
AJ320104
AJ320105
AJ320123
AJ320124
AfricaLambert's Bay, South Africa, 19931H.-P. Ruffner114AJ320106AJ320145
Xanthoria calcicola Oxner
Central EuropeBurgdorf, Switzerland, 1999R. HoneggerC44AJ287223AJ320133
Lausanne, Switzerland, 2000S. ScherrerC105/1
105/2
AJ320109
AJ320109
AJ320152
AJ320151
Western EuropeBrockenhurst, Hampshire, UK, 1999P. W. JamesC80AJ287222AJ320150
Xanthoria ectaneoides (Nyl.) Zahlbr.
Western EuropeRoscoff, Brittany, France, 1998R. Honegger22
2
AJ271716/A6
AJ287221/B5
AJ320130/A6
AJ320131/B5
Le Fret, Brittany, France, 2000R. HoneggerC90/1
90/2
AJ320112
AJ320113
AJ320149
AJ320149
Cudden Point, Penzance, Cornwall, UK, 2000J. Gray183AJ320110AJ320135
Cudden Point, Penzance, Cornwall, UK, 2000J. Gray184AJ320111AJ320135
Southern EuropeLo Zingaro, Sicily, Italy, 1995R. Honegger343AJ287224AJ320134
Xanthoria elegans (Link) Th. Fr.
Central EuropeWinterthur, Switzerland, 1999S. Scherrer 2NPAJ320139
North AmericaEllesmere Island, Arctis, Canada, 19961A. Jolles1172NPND
Xanthoria flammea (L. f.) Hillm. (= Xanthodactylon flammeum (L. f.) Dodge)
AfricaYzerfontein, South Africa, 1999H. P. Ruffner1101AJ287225AJ320132
Xanthoria turbinata Vain.
AfricaAlexander Bay, South Africa, 2000R. Dudler13AJ320114AJ320143
Xanthoria candelaria (L.) Th. Fr.
AntarcticLivingston Island, UK, 19901S. OttC169NPAJ320138
Xanthoria fallax (Hepp) Arnold
Central EuropeChur, Switzerland, 1998U. Jauch168NPNP
Southwest EuropeMt Louis, Cerdagne, France, 1990R. HoneggerC61NPND
Xanthoria polycarpa (Hoffm.) Rieber
North AmericaSan Juan Island, WA, USA, 19951R. HoneggerC171NPND
Xanthoria sp.
Southeast EuropePaphos, Cyprus, 2000A. Birchmeier 5AJ320115AJ320142
Mt Eros, Hydra, Greece, 2000O. W. Purvis 87AJ320116AJ320141
Mt Eros, Hydra, Greece, 2000O. W. Purvis 85AJ320117AJ320140
Xanthoria novozelandica Hillm.
AustralasiaRoxburgh, Central Otago, New Zealand, 2000D. J. GallowayC66NPAJ320153
Teloschistes chrysophthalmus (L.) Th. Fr.
South AmericaPasochoa Reservat, Quito, Ecuador, 19971H.-R. Hohl1173NPND
Caloplaca regalis (Vainio) Zahlbr.
South AmericaPaine National Park, Chile, 1993B. SchroeterC170NPND
Figure 2.

Morphology of the symbiotic phenotype of Xanthoria ssp. (a) X. parietina, Zürich (Switzerland), voucher No. 98; (b) X. parietina, Sicily (Italy), No. 4; (c) X. ectaneoides, Le Fret (France), No. 90; (d) X. ectaneoides, Penzance (UK), No. 83; (e) X. ectaneoides, Penzance, No. 84; (f) X. ectaneoides, Lo Zingaro, Sicily, No. 43; (g) X. calcicola, Brockenhurst (UK), No. 80; (h) X. sp., Hydra (Greece), No. 87; (i) X. sp., Hydra, No. 85; (j) X. sp., Paphos (Cyprus), No. 5; (k) (l) X. turbinata, Alexander Bay (South Africa), No. 3; (m) (n) X. flammea, Yzerfontein (South Africa), No. 101; (o) X. novozelandica, Roxburgh (New Zealand), No. 66. Same magnification in a–k, n–o.

Isolation and analysis of genomic DNA

For PCR reactions genomic DNA was isolated from single apothecia, or if absent, from individual lobes. Air-dried fragments were ground with a micropestle in a 1.5-ml reaction tube which was placed in liquid nitrogen. Subsequently, genomic DNA was isolated using the GFX polymerase chain reaction (PCR), DNA and Gel Band Purification Kit (Amersham Pharmacia Biotech Inc., Piscataway, NJ, USA). A total of 100 µl of capture buffer was added to the ground lichen fragments and the samples were incubated at 60°C for 10 min After an additional centrifugation step, the supernatant was loaded on columns preloaded with 100 µl capture buffer. Subsequent steps were carried out following the manufacturer's instructions.

For genomic DNA blot analysis (Southern), fragments of lichen thalli were frozen and ground in liquid nitrogen after the addition of aluminium oxide. Genomic DNA was extracted according to the protocol of M.I. Borges et al. available at the following URL: http://www.fgsc.net/fgn37/borges.html. For digestion of genomic DNA for Southern blot analysis, 5 µg of DNA was diluted in 1× buffer to a final volume of 400 µl (10× OPA+ buffer, supplied with the enzyme EcoRI; Amersham Pharmacia Biotech) to reduce the concentration of inhibiting compounds, mixed and incubated for 1 h on ice for equal distribution. Half of the enzyme volume was added, the other half after 10 min incubation on ice (total 100 U). Subsequently, spermidine was added to a final concentration of 1.75 mm and digestion was performed overnight at 37°C. DNA electrophoresis and blotting were carried out using standard procedures (Sambrook et al., 1989). Hybridization and detection was performed with the nonradioactive AlkPhos direct Kit (Amersham Pharmacia Biotech) according to the instructions of the manufacturer. The cDNA fragment of XEH1 cloned in pCR 2.1 (Invitrogen) was used to produce a probe (for details see Scherrer et al., 2000).

Gene amplification by polymerase chain reaction

To amplify the H1 gene from genomic DNA, PCR was performed with either 1.25 U HotStarTaq polymerase (Qiagen, Hilden, Germany) or 1.25 U Taq DNA Polymerase (Amersham Pharmacia Biotech) in 50 µl reaction volume. The reaction mixture contained PCR buffer (supplied with the enzyme, with 1.5 mm MgCl2), 200 µm of each dNTP, 0.4 µm forward primer (XpP6, 5′-ACATCATCGCCTTTGTCTCC-3′; XpP5, 5′-CCGAGAACCTGGTCGAGC-3′) and 0.4 µm reverse primer (XpP8, 5′-CCTTGCCATAGACTATGGAC-3′; XpP9, 5′-AATTCGCTCCGAGGCGAATG-3′ or XpP10, 5′-CTACAARGAGACSGGCACRC-3′). Thirty-five cycles with the following profile were programmed: denaturation for 1 min at 94°C; annealing for 30 s at 50°C; extension for 1 min at 72°C. All primers were designed after the sequence of XPH1 from aposymbiotically cultured X. parietina, which was obtained by screening a genomic library (Scherrer, 2000). rDNA fragments were amplified according to the protocol of White et al. (1990) with the primers ITS4 and ITS5.

Cleaved amplified polymorphic sequence analysis

Restriction enzyme reactions were performed for 3 h in a total volume of 10 µl following the manufacturer's instructions (HapII, HindIII, PstI and SmaI from Amersham Pharmacia Biotech; HinfI and RsaI from New England BioLabs, Beverley, MA, USA).

Fragments of the amplified and cleaved rDNA regions (primers ITS4 and ITS5) were analysed on 3% agarose gels containing 1% ultra pure agarose (Gibco BRL, Basel, Switzerland) and 2% NuSieve GTG agarose (FMC Bioproducts, Rockland, ME, USA). Fragments of the amplified and cleaved H1 regions (primers XpP6 and XpP8) were analysed on 1.2% agarose gels (Gibco BRL). The length of the fragments was estimated by comparison with the 100 bp DNA Ladder (New England BioLabs). Fragment sizes of selected samples were confirmed by restriction digests in silicio using the programs DNA strider 1.3 (Marck, 1988) (Ch Marck, Service de Biochimie, Gif-sur-Yvette, Cedex, France) or GCG (Genetics Computer Group, Wisconsin Sequence Analysis Package, Madison, WI, USA). Because the sequences analysed were shorter than the PCR product used for restriction analyses, the missing, strongly conserved flanking regions were added by comparison with published sequences of other organisms to get the actual fragment sizes.

Sequencing and sequence analysis

The PCR products were cut out from the agarose gel, purified with the GFX columns (Gel extraction Kit, Amersham Pharmacia Biotech) and directly sequenced with the described primers. All H1 gene sequences were determined at least twice on both strands. The rDNA fragments were sequenced at least once from both strands. The sequencing reaction was performed with an ABI PRISM BigDye· Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) following the instructions of the manufacturer and analysed in a 373 DNA Sequencer (Applied Biosystems). Electropherograms were checked with the chromas program version 1.43 (C. McCarthy, Griffith University, Brisbane, Australia) or Sequencing Analysis 3.0 (Applied Biosystems, a division of Perkin Elmer) and sequences were analysed using the program DNA strider 1.3 or GCG version 10 software (Genetics Computer Group, Wisconsin Sequence Analysis Package). Sequences were aligned with the pileup program of GCG and checked by eye. Neighbour-joining distance trees were calculated using paup* version 4.0 b10 (Swofford, 2002) based on uncorrected (‘p’) distances. Bootstrap analysis was performed with 1000 replicates.

Results

Hybridization

Southern blot analysis was performed with genomic DNA of eight species of the genus Xanthoria and of two representatives of other teloschistacean genera using the XEH1 probe (Fig. 1). Homologues were found in X. parietina, Xanthoria calcicola, X. ectaneoides and Xanthoria flammea. Under low stringency conditions weak signals could be detected in Xanthoria candelaria, Xanthoria fallax, Xanthoria polycarpa and Caloplaca regalis, but not in Xanthoria elegans, or Teloschistes chrystophthalmus (not shown). Xanthoria turbinata (Fig. 2k–l) and Xanthoria novozelandica (Fig. 2o) were not tested.

Figure 1.

Southern blot analysis with the cDNA of XEH1 hybridizing to DNA of Xanthoria spp. Genomic DNA was digested with EcoRI. Lanes: 1–4, four thalli of X. calcicola (Brockenhurst, UK), collected at three different sites; 5, X. ectaneoides (B4, Roscoff, France); 6, X. parietina (A1, Roscoff, France); 7 and 8, two thalli of X. parietina (Zürich, Switzerland); 9, X. parietina (Winterthur, Switzerland); 10, X. ectaneoides (Sicily, Italy); 11, X. flammea (South Africa).

Amplification and sequencing of the hydrophobin genes

The H1 hydrophobin gene was amplified by PCR from genomic DNA of the species with positive signals and some additional samples, partly from different collecting sites: X. parietina (Fig. 2a,b), X. calcicola (Fig. 2g), X. ectaneoides (Fig. 2c–f), X. flammea (Fig. 2m–n), X. turbinata (Fig. 2k–l) and a series of samples which could not be unquestionably identified by morphological characters (Xanthoria sp.; Fig. 2h–j). The primers were designed after the sequences of X. parietina and X. ectaneoides; the forward primer annealed in the region coding for the secretion signal, and the reverse primer shortly after the stop codon. Possibly owing to too high variability no PCR product was obtained in either of the other Xanthoria species. All DNA sequences are available in the databases (Table 1).

Intraspecific variation of the genomic H1 sequences

The rather vague species delimitation in some representatives of the genus Xanthoria causes difficulties in the determination of the intraspecific variation. Therefore, the species name is completed by the geographical origin of the samples for a closer description. The intraspecific variation of the genomic sequences of H1 in X. parietina and X. calcicola from Western and central Europe was very low (> 99%; in most cases 100% identity). This was confirmed by additional samples which are not listed in this study. The same applies for H1 of X. ectaneoides from coastal Brittany, France, and Cornwall, UK. However, X. parietina and X. ectaneoides both include a Southern type (X. p., South Africa and Sicily, Fig. 2b; X. e., Sicily, Fig. 2f) whose hydrophobin gene differed considerably from the Northern type (all other origins). The hydrophobin sequences of X. parietina from Sicily and South Africa were only 92.5% identical with X. parietina from the other collecting sites. Interestingly, samples of X. parietina from the American continent and New Zealand were 100% identical with samples from northern and central Europe (Northern type). In X. ectaneoides, the similarity between the Southern and the Northern type (X. e. from Sicily and X. e. from Brittany) was only 89%. These variations resulted in amino acid differences in both taxa. No differences were found in the nucleotide sequences of the H1 gene of X. calcicola collected in Brockenhurst (UK) and Burgdorf (Switzerland) or Lausanne (Switzerland). No data are available on intraspecific variation in X. turbinata and X. flammea, both South African endemics, because only one or two freshly collected samples were available.

Generally, intraspecific differences were found in samples from geographically separated populations. The nucleotide sequences of X. parietina from Zurich and Winterthur (20 km apart) were identical; they showed 96% similarity to the samples collected in Roscoff (France). The variation was mainly caused by either different intron sequences or by a wobbling third base of a codon (see Table 2, Fig. 3). Different genotypes of samples from the same collecting site were also detected (e.g. in X. ectaneoides from Roscoff, where the variation was caused by a deletion of three bases in the first intron; Figs 3 and 5a).

Table 2.  Length of the two introns of the H1 gene in the different Xanthoria species
Species, Collecting sitesIntron IIntron II
  1. In X. parietina and X. calcicola the introns were confirmed by sequencing of the cDNA fragments, in the other species the introns were determined on the basis of sequence homologies and the GT/AG rule.

Xanthoria parietina Roscoff (France), Zurich and Winterthur (Switzerland)50 bp47 bp
Xanthoria calcicola Brockenhurst (UK), Lausanne (Switzerland)50 bp47 bp
Xanthoria ectaneoides Roscoff (France), Penzance (UK)47/50 bp47 bp
Xanthoria ectaneoides Sicily (Italy)50 bp45 bp
Xanthoria flammea Yzerfontein (South Africa)49 bp50 bp
Xanthoria turbinata Alexander Bay (South Africa)52 bp50 bp
Xanthoria sp. Paphos (Cyprus)50 bp45 bp
Xanthoria sp. Hydra (Greece)50 bp45 bp
Figure 3.

Comparison of the H1 gene region of different Xanthoria species. Exons are indicated by upper case letters, introns by lower case letters, and variable positions are marked with asterisks. The underlined three bases in the intron I of Xanthoria ectaneoides from Roscoff, France, were deleted in sample B5 and present in sample A6.

Figure 5.

Neighbour-joining distance trees of Xanthoria species from different locations (including country codes and voucher numbers). Bootstrap values over 50% (percentages based on 1000 replicates) are indicated on the main branches. Outgroups are shown in italics. (a) Neighbour-joining distance tree based on the H1 gene region (length of the alignment: 354 bp). (b) Neighbour-joining tree based on ribosomal gene region containing the internal transcribed spacers (ITS 1 and 2) and the 5.8 S rDNA (length of the alignment: 579 bp).

A neighbour-joining distance tree based on all genomic H1 sequences is shown in Fig. 5a. Xanthoria flammea and X. turbinata were used as outgroups. In the analysis of the rDNA data, they appeared to be a monophyletic sister group to the ingroup (Fig. 5b). Two main clusters were observed: one containing X. parietina, the other containing the X. ectaneoides/X. calcicola complex, which could not be further resolved as some sequences from both species were 100% identical. Surprisingly, X. ectaneoides from Sicily does not fall in this cluster. The similarity matrix of 10 selected genomic H1 sequences is presented in. Table 3a. The sequences of these samples code for the 10 divergent H1 amino acid sequences discovered in this study.

Table 3.  Similarity matrices
Species, collecting site, voucher No.Xp 98Xp 4Xc 105Xe 84Xsp 87Xsp 85Xe 43Xsp 5Xf 101Xt 3
(a) Genomic H1 sequences of 10 selected samples with divergent H1 amino acid sequences
Xanthoria parietina Zurich, Switzerland, 98         
Xanthoria parietina Sicily, Italy, 492.8%        
Xanthoria calcicola Lausanne, Switzerland, 10594.0%97.1%       
Xanthoria ectaneoides Penzance, UK, 8493.0%96.0%98.6%      
Xanthoria sp. Hydra, Greece, 8790.0%92.0%93.1%91.9%     
Xanthoria sp. Hydra, Greece, 8586.5%88.2%89.4%88.7%88.8%    
Xanthoria ectaneoides Sicily, Italy, 4386.2%88.5%89.0%88.4%88.5%98.0%   
Xanthoria sp. Paphos, Cyprus, 585.6%86.0%86.8%85.8%86.5%87.6%88.5%  
Xanthoria flammea Yzerfontein, South Africa, 10172.3%74.3%72.8%72.7%71.1%73.9%74.1%73.9% 
Xanthoria turbinata Alexander Bay, South Africa, 369.9%72.2%71.9%71.8%68.8%72.7%72.9%69.5%86.0%
(b) rDNA sequences of the same samples
Xanthoria parietina Zurich, Switzerland, 98         
Xanthoria parietina Sicily, Italy, 498.0%        
Xanthoria calcicola Lausanne, Switzerland, 10595.3%96.5%       
Xanthoria ectaneoides Penzance, UK, 8495.5%96.7%98.9%      
Xanthoria sp. Hydra, Greece, 8798.9%97.6%95.0%95.2%     
Xanthoria sp. Hydra, Greece, 8597.0%97.0%96.1%96.3%96.7%    
Xanthoria ectaneoides Sicily, Italy, 4396.3%96.3%95.6%95.7%96.0%99.3%   
Xanthoria sp. Paphos, Cyprus, 596.8%98.3%95.2%95.8%96.5%96.0%95.2%  
Xanthoria flammea Yzerfontein, South Africa, 10185.7%84.7%84.7%85.1%85.3%85.4%85.2%83.7% 
Xanthoria turbinata Alexander Bay, South Africa, 385.8%84.8%84.8%85.2%85.7%85.0%84.6%83.9%93.2%

Introns

All H1 genes were interrupted by two introns at conserved positions. The sequences of the introns were more variable than the exons, but all of them followed the GT/AG rule and contained a putative internal splicing signal. The size of the introns ranged from 45 to 52 bp (Table 2, Fig. 3).

Amino acid sequences

Translation of the genomic H1 sequences of all samples examined in this study resulted in 10 different amino acid sequences which are aligned in Fig. 4. (The characters in Fig. 4 correspond to those in Table 5 where all samples are listed.) Amino acid exchanges and similarity matrices of pairwise combinations are shown in Table 4. Of 83 positions of the mature H1 protein the amino acids were found to vary interspecifically at 23 positions. However, at 15 positions only conservative amino acid exchanges were evident (Fig. 4). In none of the cases were the cysteine pattern and the hydropathy plot affected, characteristic features of class I hydrophobins. Only one conservative change was observed within the conserved amino acids as determined by Wösten (2001) in an alignment of all known hydrophobins of ascomycetes and basidiomycetes. Most of the intraspecific variation of the H1 gene sequence did not result in amino acid changes; thus only minor intraspecific variation of the amino acid sequence was detected. However, in X. ectaneoides three different sequences were found.

Figure 4.

Alignment of the deduced amino acid sequences of the 10 divergent mature H1 sequences discovered in this study. Variable positions are marked with asterisks, and the typical pattern of eight cysteines is pointed out by bold letters. The characters preceding every line correspond to those in Table 5.

Table 5.  Restriction fragment patterns and amino acid sequences
Species, areaCollecting sitevoucher No.rDNAH1H1 aa sequence
PstISmaIHapIIHindIIIRsaIHinfI
  1. Restriction fragment patterns were determined by enzymatic digests of rDNA and H1 PCR products (625 and 950 bp). In several samples only the 460 bp H1 PCR product was obtained, which was used for sequencing (no H1 restriction fragment pattern). For details see the Materials and Methods section, for fragment sizes see Table 6. The amino acid sequences are shown in Fig. 4. NP, no product obtained. ND, not determined.

Xanthoria parietina
Central EuropeZurich, Switzerland98IIIIIIIIIIIIA
Winterthur, Switzerland88IIIIIIIIIIIIA
Western EuropeRoscoff, Brittany, FranceA1ND  IIIIIIIIA
Le Fret, Brittany, France91IIIIIIIIIIIIA
Penzance, Cornwall, UK106IIIIIIIIIIIIA
Cerdagne, France60IIIIIIIIIIIIA
Irchester, Northamptonshire, UK18IIIIIIIIIIIIA
Southern EuropeLo Zingaro, Sicily, Italy4IIIIIVIIB
Northern EuropeOslo, Norway16IIIIND  A
Tjörn, Bohuslän, Sweden99IIIIND  A
Ljugarn, Gotland, Sweden97IIIIIIIIIIIIA
Oulunsallo, Finland89IIIIIIIIIIIIA
North AmericaBay Bulls, Canada100IIIIIIIIIIIIA
Maine, USA26, 27IIIIIIIIIIIIA
Oregon, USA6IIIIIIIIIIIIA
AustralasiaTasmania, Australia10IIIIIIIIIIIIA
Central Otago, New Zealand1/1, 1/2IIIIIIIIIIIIA
AfricaLambert's Bay, South Africa14NP  NP  B
Xanthoria calcicolaBurgdorf, Switzerland44IIIIIIIIIIC
Brockenhurst, Cornwall, UK80IIIIIIIIIIIC
Lausanne, Switzerland105/1
105/2
II
II
II
II
I
I
ND
ND
  C
C
Xanthoria ectaneoidesRoscoff, Brittany, France(A6, B5)III, IIIIIIIIIC
Penzance, Cornwall, UK83IIIIIINDNDC
Penzance, Cornwall, UK84IIIIIIIVIID
Le Fret, Brittany, France90/1, 90/2IIIIIIIIIIC
Lo Zingaro, Sicily, Italy43IIIIINP  G
Xanthoria flammeaYzerfontein, South Africa101IIIIINP  I
Xanthoria turbinataAlexander Bay, South Africa3IIIIINP  J
Xanthoria sp.Paphos, Cyprus5IIIINP  H
Hydra, Greece87IIIIIIIIIIVE
Hydra, Greece85IIIIINP  F
Table 4.  Substitution and similarity matrices of the 10 divergent H1 amino acid sequences on the basis of the alignment in Fig. 4
Species, collecting site, voucher No.Xp 98Xp 4Xc 105Xe 84Xsp 87Xsp 85Xe 43Xsp 5Xf 101Xt 3
  1. In the top right triangle the number of amino acid substitutions are listed; the bottom left triangle shows the sequence similarities.

Xanthoria parietina Zurich, Switzerland, 98 2 3 310 5 6 71012
Xanthoria parietina Sicily, Italy, 497.6% 1 3 8 5 4 71112
Xanthoria calcicola Lausanne, Switzerland, 10596.4%98.8% 2 9 6 5 71112
Xanthoria ectaneoides Penzance, UK, 8496.4%96.4%97.6%11 9 7 91311
Xanthoria sp. Hydra, Greece, 8788.0%90.4%89.2%86.8% 8 7111718
Xanthoria sp. Hydra, Greece, 8594.0%94.0%92.8%89.2%90.4% 1 61010
Xanthoria ectaneoides Sicily, Italy, 4392.8%95.2%94.0%91.6%91.6%98.8% 51011
Xanthoria sp. Paphos, Cyprus, 591.6%91.6%91.6%89.2%86.8%92.8%94.0%1115
Xanthoria flammea Yzerfontein, South Africa, 10188.0%86.8%86.8%84.3%79.5%88.0%88.0%86.8% 9
Xanthoria turbinata Alexander Bay, South Africa, 385.5%85.5%85.5%86.8%78.3%88.0%86.8%81.9%89.2%

A considerable genetic distance was evident between X. ectaneoides specimens collected either in Cornwall and Brittany or Sicily. By contrast, the amino acid sequences of X. ectaneoides from Roscoff and X. calcicola were fully identical. In addition, two distinct H1 amino acid sequences were found in X. parietina, reflecting a Northern and a Southern type. Highest variability was found in samples from the eastern Mediterranean islands Hydra (two samples) and Paphos (one sample); these did not share the sequence of any other species.

Ribosomal gene region

The internal transcribed spacers (ITS) of the ribosomal clusters were investigated with the aim of comparing the variation of these rapidly evolving regions with the hydrophobin gene data. Polymerase chain reaction amplification of the sequence coding for the 5.8S rRNA, flanked by ITS1 and ITS2, using the primers ITS4 and ITS5 (White et al., 1990), yielded in all samples a product of approximately 625 bp. The amplified fragments were sequenced directly with the same primers; the readable part contained the complete sequences of the internal transcribed spacers and 5.8S rDNA. For alignment, the sequences were shortened to corresponding fragments of approximately 540 bp (18S rDNA, partial, 9 bp; ITS1, approx. 200 bp; 5.8S rDNA, 157 bp; ITS2, approx. 160 bp; 28S rDNA, partial, 19 bp). In Table 3b the similarity matrix of the same samples is presented as chosen for the similarity matrix of the H1 gene in Table 3a. A neighbour-joining distance tree, as calculated for the entire aligned sequences, is shown in Fig. 5b. Xanthoria novozelandica served as outgroup. As observed in the H1 gene, the X. parietina cluster was well supported by bootstrap analysis. No clusters were formed within the X. ectaneoides/X. calcicola complex, and X. ectaneoides from Sicily did not group within the other X. ectaneoides samples.

In general, the rDNA region was less variable than the H1 gene. As expected, the 5.8S rDNA was highly conserved in all species, but the internal transcribed spacers were more variable. Surprisingly, but in parallel to our data on the H1 gene, the amplified rDNA region of X. ectaneoides from Roscoff and the UK X. calcicola were highly similar, but the sequences of X. ectaneoides from Roscoff and X. ectaneoides from Sicily were only 95.6% identical. The rDNA of Xanthoria ectaneoides from Sicily was more similar to X. parietina than to X. ectaneoides from Roscoff. The sequences of X. ectaneoides from Penzance, Cornwall (voucher numbers 83 and 84) were fully identical with X. ectaneoides from Brittany.

Cleaved amplified polymorphic sequence analyses of the H1 gene region and the rDNA region

The PCR products corresponding to the H1 gene region (950 bp) and the ITS rDNA region (625 bp) were digested with a set of different endonucleases. Restriction phenotypes of each specimen and the approximate fragment sizes are summarized in Tables 5 and 6. The cleavage sites were checked in sequences of selected samples. In general, no discrepancies were observed. For cleaved amplified polymorphic sequence (CAPS) analysis each of the two regions was digested with three enzymes (rDNA with PstI, SmaI and HapII; H1 with HindIII, RsaI and HinfI); up to five phenotypes were distinguished, which were in full agreement with sequencing data and allow species delimitation (Tables 5 and 6). In more distantly related species (e.g. X. flammea) the 950 bp fragment of the H1 gene could not be amplified.

Table 6.  Restriction fragment sizes in base pairs (bp) and restriction fragment patterns
(a) rDNA PstI patternFragment size (bp)SmaI patternFragment size (bp)HapII patternFragment size (bp)
  1. Given sizes are calculated with selected samples; slight variations were possible which did not affect the pattern. The H1 950 bp PCR product was fully sequenced only in X. parietina; fragment sizes of the other patterns are estimated values, by comparison with the DNA ladder on the gel. U, no digestion. 1Bands not detected on agarose gels are written in brackets. 2Fragments could not be separated.

IU 625IU 625I242, 200, 872, 712 (14, 6, 4)1
II409, 221II539, 85II202, 157, 882, 842, 832 (10, 6)1
(b) H1
HindIII patternFragment size (bp)RsaI patternFragment size (bp)HinfI patternFragment size (bp)
IU 951IU 951IU 951
II648, 303II600, 350II500
III698, 216 (37)1III377, 352, 222  
IV600, 300IV650, 400  
V650, 250    

Discussion

The present study is the first report on the variability of the hydrophobin gene in a group of closely related, lichen-forming fungi. Moreover, it is a first approach to determine the variation of a hydrophobin gene in a large number of samples collected in nature in geographically distant areas (four continents) and to compare the variability in this protein-coding gene with noncoding areas of the rDNA. The sequence of the 5.8S rDNA and internal transcribed spacers of Xanthoria resendei (Martin & Winka, 2000) shows high homology to the corresponding region of the species examined in the present study.

Hydrophobin gene sequences as taxonomic markers

In their study on evolutionary relationships in the Aspergillus section Fumigati Geiser et al. (1998) pointed out that hydrophobin gene sequences might be useful tools in molecular systematics because they reveal, in addition to the characteristic cysteine pattern, very low amino acid homology (Wessels, 1997). Our present data on a range of Xanthoria spp. indicate that this also applies to lichen-forming ascomycetes. However, only a very limited number of closely related taxa can be tested with one particular hydrophobin probe. However formerly unrecognized relationships may be detected with this approach. The South African endemics X. flammea and X. turbinata (see later) proved to be more closely related to the X. parietina complex than X. fallax, X. polycarpa, and X. novozelandica. Conventional lichen taxonomy at the species level is based on morphotypic and chemotypic features of the symbiotic phenotype. In a large number of representatives of the genus Xanthoria the species delimitation is rather vague (Kärnefelt, 1989; Purvis et al., 1992; Lindblom, 1997). This applies to a high degree to the European taxa of the X. parietina complex as investigated in the present study.

The rosette-shaped thalli of X. parietina (Fig. 2a,b) form apothecia on young lobe portions. Probably because of its low substrate specificity, its very successful dispersal via thallus fragmentation and fecal pellets of lichenivorous invertebrates, and by its remarkable regenerative capacity, as observed in field and experimental studies (Honegger, 1996; Honegger et al., 1996; Meier et al., 2002), X. parietina is now an almost cosmopolitan species. However there is a high probability that it has been anthropogenically introduced in Australia, New Zealand, New Guinea and in western North America (Galloway, 1985, 1991; Rogers, 1992; Lindblom, 1997). The high sequence identity (up to 100%) of north and central European, North American and Australasian samples, as found in this study, supports this hypothesis. The eastern Mediterranean area seems to harbour a high genetic diversity in this group, as concluded from few randomly sampled, partly unnamed specimens. In Europe, X. parietina occurs on a very wide range of nutrient-rich, natural or anthropogenic substrata from coastal to montane areas. The morphologically quite similar X. ectaneoides (Fig. 2c,f) is often sterile and may form irregular to circular hummocks of overlapping, crenulate to strap-shaped lobes. It occurs in Europe and Africa as a predominantly coastal species on calcareous and granitic rocks and concrete walls (Doidge, 1950; Purvis et al., 1992). In the Xanthorion of coastal rocks in Brittany, relatively broad-lobed X. ectaneoides is the quantitatively predominant species, but owing to considerable morphotypic overlap cannot always be properly distinguished from X. parietina (Figure 2c,d). Usually, no barrage is visible between vicinal thalli of either species.

Xanthoria calcicola (Fig. 2g), a European species, predominates in relatively warm and dry inland habitats, especially in the east of Britain, but is rather rare in central Europe. It has isidia-like outgrowths and only seldom and irregularly reproduces sexually. In chemotaxonomic studies X. parietina and X. calcicola were shown to belong to two distinct chemosyndromes within the Teloschistaceae, which differ by the relative proportions of the anthraquinone pigments (Steiner & Hauschild, 1970; Søchting, 1997). No corresponding data are available for X. ectaneoides. The photographic documentation of a preliminary report on transplant studies show that coastal X. ectaneoides (referred to as X. parietina) achieved the same morphotype and coloration as neighbouring X. calcicola within 1 yr after transplantation to an inland site and vice versa (Richardson, 1967), which suggests considerable morphotypic plasticity. In our neighbour-joining analysis no resolution was obtained in the X. ectaneoides/X. calcicola cluster. Parsimony analysis of the same datasets resulted in the same tree topology (not shown). In some cases, both DNA regions were 100% identical in samples from both species. The variation in this cluster is comparable to the intraspecific variation in X. parietina. Our present data on the hydrophobin H1 gene and on the ITS region suggest that X. calcicola and X. ectaneoides from Western Europe are inland or coastal forms, respectively, of the same species, which differs from X. parietina. However, this has to be confirmed with additional molecular markers. The Mediterranean X. ectaneoides, with only 94% H1 protein homology to X. ectaneoides from Brittany, requires further investigation.

The CAPS analysis tests of both amplified DNA regions permitted detection of sequence variation between Xanthoria ssp. This method is less time-consuming than sequencing and thus can be used for screening large numbers of specimens. It provides a useful tool in comparative studies on the morphological and genetic variation within a species or species complex.

It is surprising that the XEH1 probe hybridized to X. flammea (Fig. 2m,n), an endemic of western South Africa, whose dorsiventrally organized lobes give rise to prominent, tubular–erect structures with lateral or terminal apothecia (no mature apothecia were available in our samples). Because of this unique morphological feature, the monotypic genus Xanthodactylon Duvignard was created, with Xanthodactylon flammeum (L. fil.) Dodge as type species (Kärnefelt, 1989). Based on morphological features such as lobe structure X. turbinata, another African species, was interpreted as an intermediate form between the genera Xanthoria and Xanthodactylon (Kärnefelt, 1989). The molecular data on the H1 gene and the rDNA region confirm the relatively close relation to X. flammea but, by contrast to the morphological data, the H1 gene sequences imply that X. turbinata is more distantly related to X. parietina than X. flammea, rejecting the hypothesis of X. turbinata as an intermediate form. However, the genetic distance of both African species based on rDNA data is approximately equal.

Intraspecific variation of the H1 gene

Only two amino acid sequences were found to differ at two positions of the mature H1 protein of either X. parietina or X. ectaneoides/X. calcicola. Intraspecific variability was shown to occur in the hydrophobin protein sequences of the basidiomycete Agaricus bisporus. Lugones et al. (1996, 1998) found 98.2% similarity in the ABH1 protein of two different cultivars with either white or brown fruiting bodies, and 95.7% similarity in ABH3 of the vegetative mycelium of the two homokaryons which constitute the white-fruiting dikaryotic cultivar. Intraspecific variation was also shown to occur in a range of other coding genes of nonlichenized fungi (e.g. allelic laccase genes of Lentinus edodes; Zhao & Kwan, 1999).

The very low intraspecific variation of the H1 gene and of the ribosomal gene region within thalli of X. parietina from geographically widely separated populations is surprising, because this species forms abundant sexual reproductive stages. As generally observed, sequence conservation of the coding regions is generally higher than that of noncoding regions. Within different Xanthoria species the sequence and length of the introns of the H1 gene were more variable, but their position was never changed. Intron positions within corresponding genes of different species are generally conserved, but this does not usually apply to their length and sequence (Unkles 1992). Based on rDNA and its group-I introns the intraspecific variation and phylogenetic relationships have been investigated in a range of other lichen-forming ascomycetes (DePriest, 1993a; Gargas et al., 1995), but the molecular biology proper of lichen-forming fungi is largely unexplored. Genetic differences in the nuclear SSU rDNA were detected within populations of Cladonia chlorophaea (DePriest, 1993b). Beard and DePriest (1996) found no variation in rDNA sequences within one mat of Cladina subtenuis, but different mats of one geographic location proved to be genetically diverse. In all of these studies a higher genetic variability was determined than in the Xanthoria spp. from distant geographic locations as examined in the present study.

Speciation in lichen-forming ascomycetes

Very little is known about speciation in lichen-forming fungi. Our present data indicate that the closely related European species of the X. parietina complex might be an interesting system for exploring speciation in lichen-forming ascomycetes. It could well be that the Mediterranean area harbours as yet unrecognized cryptic species, but additional coding and noncoding regions will have to be studied in a wide range of samples. A whole range of cryptic species has been discovered in the genus Letharia in western North America on the basis of comparative sequence analyses of coding and noncoding regions (Kroken & Taylor, 2001). For obtain a better understanding of our data set on the interspecific and intraspecific genetic variation in the X. parietina complex we are currently investigating the mating systems of all taxa. Moreover, it will be interesting to investigate the taxonomic affiliation and genetic diversity of the green algal photobiont in samples collected round the globe, and to explore evolutionary traits in both partners of this fascinating symbiosis.

Acknowledgements

Our sincere thanks to all friends and colleagues who provided us with fresh specimens of Teloschistaceae, partly from remote areas of the globe, as summarized in Table 1, to P. W. James for invaluable help with taxonomic problems, to Annette Haisch and Undine Zippler for competent technical assistance, to J.-J. Pittet for help with the artwork, and to the Swiss National Science Foundation for generous financial support (grants Nos. 31-42392.94 and 31-52981.97 to R.H).

Ancillary