mycological research 110 (2006) 511– 520 available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/mycres Molecular data place Trypetheliaceae in Dothideomycetes Ruth DEL PRADOa, Imke SCHMITTa, Stefanie KAUTZb, Zdenek PALICEc, Robert LU¨ CKINGa, H. Thorsten LUMBSCHa,* aDepartment of Botany, The Field Museum, 1400 S. Lake Shore Drive, Chicago, IL 60605, U.S.A. bFachbereich Biologie und Geografie, Universita¨t Duisburg-Essen, Campus Essen, Universita¨tsstraße 5, D-45517 Essen, Germany cInstitute of Botany, Academy of Sciences of the Czech Republic, CZ-25243 Pruhonice, Czech Republic article info abstract Article history: The phylogenetic position of Trypetheliaceae was studied using partial sequences of the Received 15 July 2005 mtSSU and nuLSU rDNA of 100 and 110 ascomycetes, respectively, including 48 newly ob- Accepted 31 August 2005 tained sequences. Our analysis confirms Trypetheliaceae as monophyletic and places the Published online 18 April 2006 family in Dothideomycetes. Pyrenulaceae, which were previously classified with Trypethelia- Corresponding Editor: Martin Grube ceae in Pyrenulales or Melanommatales, are supported as belonging to Chaetothyriomycetes. Monophyly of Pyrenulales, including Trypetheliaceae is rejected using three independent Keywords: test methods. Monophyly of Arthopyreniaceae plus Trypetheliaceae, the two families includ- Ascomycota ing lichen-forming fungi in Dothideomycetes, is also rejected, as well as a placement of Try- Lichens petheliaceae in Pleosporales (incl. Melanommatales). Molecular phylogeny ª 2006 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. Pleosporales Pyrenulales Trypethelium Introduction The latter genus has previously been excluded from Trypethe- liaceae by Veˇzda (1968) based on the thin-walled, muriform as- Trypetheliaceae is a medium-sized family of tropical and sub- cospores, and classified in a separate family Laureraceae (Poelt tropical crustose pyrenocarpous mainly lichenized fungi with 1974), but Eriksson (1981) showed that Laureraceae fits well about 200 species (Trevisan 1861; Malme 1924; Letrouit-Galinou into Trypetheliaceae. Subsequently, three additional genera 1957, 1958; Harris 1984, 1990, 1991, 1995, 1998; Makhija & Pat- were described in the family by Aptroot (1991). wardhan 1988, 1993; Aptroot 1991; Aptroot et al. 1997). Most Based on morphological characters, such as bitunicate asci species grow endophloedically on bark and occur in lowland and graphidean ascospores, Trypetheliaceae have always been to submontane tropical rainforests, gallery forests, and man- regarded as closely related to Pyrenulaceae (Barr 1981; Eriksson groves. The family is characterized by bitunicate asci, asco- 1981; Henssen & Jahns 1973; Poelt 1974; Aptroot 1991; Harris spores with angular-wall thickenings and diamond-shaped 1995), and the family is currently placed in Pyrenulales (Eriksson lumina (syngraphidean sensu Sherwood 1981), rather thin et al. 2004). Thus far, molecular data to test this view are scarce. and richly branched and anastomosing pseudoparaphyses, The only molecular study including Trypetheliaceae was published and if present, a Trentepohlia photobiont. Most taxa have asco- by Lutzoni et al. (2004),whoincludedoneTrypethelium sp. se- mata that are concentrated in pseudostroma. Nine genera quence which fell into Dothideomycetes. However, the relation- were placed in this family by Harris (1984), including Laurera. ships lacked support and hence no conclusions were drawn. * Corresponding author. E-mail address: tlumbsch@fieldmuseum.org 0953-7562/$ – see front matter ª 2006 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.mycres.2005.08.013 512 R. del Prado et al. A placement of Trypetheliaceae in Dothideomycetes would be nu-LSU-1125-30 (¼LR6) (Vilgalys & Hester 1990), nu-LSU-0654- surprising, as the supposedly related Pyrenulaceae belong to 50 (¼LR3R), nu-LSU-0635-30 (¼LR3) and nu-LSU-0948-30 (¼LR5) Chaetothyriomycetes (Lumbsch et al. 2004; Lutzoni et al. 2001, (Rytas J. Vilgays, website; http//www.biology.duke.edu/fungi/ 2004; Schmitt et al. 2005). The families containing lichen- mycolab/primers.htm); and (2) for the mtSSU rDNA: mrSSU1, forming fungi in the Chaetothyriomycetes, Pyrenulaceae and mrSSU2, mrSSU2R, mrSSU3R (Zoller et al. 1999) and MSU 7 Verrucariaceae are characterized by ascohymenial ascoma de- (Zhou & Stanosz 2001). The 25 ml PCR reactions contained velopment (Doppelbaur 1960; Janex-Favre 1970a, b), while 2.5 ml buffer, 2.5 ml dNTP mix, 2 ml of each primer (20 mm), 5 ml Dothideomycetes include species with ascolocular ascoma BSA, 2 ml Taq, 2.5 ml genomic DNA extract and 6.5 ml distilled ontogeny (Janex-Favre 1970a, b; Nannfeldt 1932; Parguey- water. Thermal cycling parameters were: initial denaturation Leduc 1966, 1967). Loculoascomycetes sensu Luttrell (1955) in- for 3 min at 94 C, followed by 34 cycles of 45 s at 94 C, 1 min clude species that are currently placed in either Chaetothyrio- at 50 C (mtSSU primers) or 54 C (nu-LSU-0155-50/LR6), mycetes or Dothideomycetes (Barr 1981; Barr & Huhndorf 2000; 1.5 min at 72 C, and a final elongation for 10 min at 72 C. Eriksson et al. 2004), although ascoma development of non- Amplification products were viewed on 1 % agarose gels lichenized Chaetothyriomycetes is poorly understood. Most mo- stained with ethidium bromide and subsequently purified us- lecular studies confirmed that Loculascomycetes sensu Luttrell ing the QIAquick PCR Purification Kit (Qiagen) or Nucleo Spin (1955) are not monophyletic (Berbee 1996; Winka et al. 1998; DNA purification kit (Macherey-Nagel, Dueren, Germany). Silva-Hanlin & Hanlin 1999; Lumbsch et al. 2000; Lindemuth Fragments were sequenced using the Big Dye Terminator et al. 2001): Chaetothyriomycetes are closely related to Eurotiomy- reaction kit (ABI PRISM, Applied Biosystems, Forster City, cetes (Berbee 1996; Liu et al. 1999; Silva-Hanlin & Hanlin 1999; USA). Sequencing and PCR amplifications were performed us- Lumbsch et al. 2000), sometimes classified as a subclass Chae- ing the same sets of primers. Cycle sequencing was executed tothyriomycetidae of Eurotiomycetes (Lutzoni et al. 2004), while with the following program: initial denaturation for 1 min at Dothideomycetes are related to Arthoniomycetes (Lumbsch et al. 96 C followed by 32 cycles of 96 C for 15 s, 50 C for 10 s, 2004; Lutzoni et al. 2004). The only recent study that recovered 60 C for 4 min. Sequenced products were precipitated with a monophyletic Loculoascomycetes, based on RPB2 sequences 10 ml sterile dH2O, 2 ml of 3 m NaOAc, and 50 ml of 95 % ethanol (Liu & Hall 2004), lacked significant support in the critical before they were loaded on an ABI 3100 or 3730 (Applied Bio- nodes. systems) automatic sequencer. Sequence fragments obtained The aim of the present study is to test the placement of Try- were assembled with SeqMan 4.03 (DNASTAR, Madison, USA) petheliaceae within Dothideomycetes by means of molecular and manually adjusted. data. For this purpose, we generated new sequences of several lichen-forming pyrenomycetes aiming at resolving the phylo- Sequence alignments and phylogenetic analysis genetic position of Trypetheliaceae. The mitochondrial data set contains sequence portions that are highly variable. Standard multiple alignment programs, Material and methods such as Clustal (Thompson et al. 1994) become less reliable when sequences show a high degree of divergence. Therefore Taxon sampling we used an alignment procedure that uses a linear Hidden Markov Model (HMM) as implemented in the software SAM New sequence data of the nuLSU rDNA and mtSSU rDNA were (Sequence Alignment and Modelling system) (Karplus et al. obtained from 28 pyrenocarpous lichen species. Two separate 1998) for the mitochondrial alignment. Regions that were analyses were performed: (1) A data matrix of 115 samples not aligned with statistical confidence were excluded from from 110 species was assembled using mitochondrial small the phylogenetic analysis. The nuLSU rDNA is much less vari- subunit rDNA sequences. Twenty-nine sequences were newly able and alignment was straightforward. This data set was obtained and 86 downloaded from GenBank; four taxa of Sor- aligned using Clustal X and all ambiguous regions were ex- dariomycetes were included as outgroup. (2) A combined data cluded from the alignments following an alignment done us- matrix of nuLSU sequences and mtSSU rDNA sequences of ing SAM (SAM deleted the last part of the alignment, as it 103 samples from 100 species was produced, including 19 was missing in some sequences, therefore the SAM alignment newly obtained nuLSU sequences. Specimens and sequences has not been used, but the modified longer alignment instead). used for the molecular analyses are compiled in Table 1. In the combined data sets only those species were included for which sequences of both gene portions were available. DNA extraction, PCR, and sequencing The alignments were analysed using minimum evolution (ME) and a Bayesian approach (B/MCMC) (Huelsenbeck et al. Total DNA was extracted from freshly collected material and 2001; Larget & Simon 1999). herbarium specimens, using the DNeasy Plant Mini Kit ME analyses were performed using the program PAUP* (Qiagen, Hilden, Germany) following the instructions of the (Swofford 2003). A heuristic search using the general time re- manufacturer. Dilutions (10ÿ1 up to 10ÿ3) or undiluted DNA versible nucleotide substitution model (Rodriguez et al. 1990) was used for PCR amplifications