Mycologia, 103(4), 2011, pp. 731–740. DOI: 10.3852/10-331 # 2011 by The Mycological Society of America, Lawrence, KS 66044-8897 Intragenomic variation in the ITS rDNA region obscures phylogenetic relationships and inflates estimates of operational taxonomic units in genus Laetiporus Daniel L. Lindner1 spacer region, intragenomic variation, molecular Mark T. Banik drive, sulfur shelf US Forest Service, Northern Research Station, Center for Forest Mycology Research, One Gifford Pinchot Drive, Madison, Wisconsin 53726 INTRODUCTION Genus Laetiporus Murrill (Basidiomycota, Polypo- rales) contains important polypore species with Abstract: Regions of rDNA are commonly used to worldwide distribution and the ability to produce infer phylogenetic relationships among fungal species cubical brown rot in living and dead wood of conifers and as DNA barcodes for identification. These and angiosperms. Ribosomal DNA sequences, includ- regions occur in large tandem arrays, and concerted ing sequences from the internal transcribed spacer evolution is believed to reduce intragenomic variation (ITS) and large subunit (LSU) regions, have been among copies within these arrays, although some used to define species and infer phylogenetic variation still might exist. Phylogenetic studies typi- relationships in Laetiporus and to confirm the cally use consensus sequencing, which effectively existence of cryptic species (Lindner and Banik conceals most intragenomic variation, but cloned 2008, Ota and Hattori 2008, Tomsovsky and Jankovsky sequences containing intragenomic variation are 2008, Ota et al. 2009, Vasaitis et al. 2009) described becoming prevalent in DNA databases. To under- with mating compatibility, ITS-RFLP, morphology stand effects of using cloned rDNA sequences in and host preference data (Banik et al. 1998, Banik phylogenetic analyses we amplified and cloned the and Burdsall 1999, Banik and Burdsall 2000, Burdsall ITS region from pure cultures of six Laetiporus and Banik 2001). Currently at least eight described species and one Wolfiporia species (Basidiomycota, Laetiporus species worldwide clearly fall in the core Polyporales). An average of 66 clones were selected Laetiporus clade as defined by Lindner and Banik randomly and sequenced from 21 cultures, producing (2008) and ITS data indicate the existence of four a total of 1399 interpretable sequences. Significant undescribed Laetiporus species (Banik et al. 2010). variation ($ 5% variation in sequence similarity) was The regions of rDNA used to infer evolutionary observed among ITS copies within six cultures from relationships in genus Laetiporus also have been used three species clades (L. cincinnatus, L. sp. clade J, to infer phylogenetic relationships among a wide and Wolfiporia dilatohypha) and phylogenetic analyses range of fungal species (Bridge et al. 2005), as well as with the cloned sequences produced different trees among diverse lineages of plants (Baldwin et al. 1995, relative to analyses with consensus sequences. Cloned Chaw et al. 1997), animals (Mallatt and Winchell sequences from L. cincinnatus fell into more than 2002) and prokaryotes (Ludwig et al. 1998). Due to one species clade and numerous cloned L. cincinna- the high copy number of rDNA and extensive datasets tus sequences fell into entirely new clades, which if available, these regions also are used commonly to analyzed on their own most likely would be recog- detect and identify species. For fungi the ITS region is nized as ‘‘undescribed’’ or ‘‘novel’’ taxa. The use of a widely regarded as the preferred region for species 95% cut off for defining operational taxonomic units identification and detection (Horton and Bruns 2001, (OTUs) produced seven Laetiporus OTUs with Ko˜ljalg et al. 2005, Peay et al. 2008, Avis et al. 2010) consensus ITS sequences and 20 OTUs with cloned and this region most likely will be chosen as one of ITS sequences. The use of cloned rDNA sequences the universal barcodes for fungal species (Seifert might be problematic in fungal phylogenetic analyses, 2009, Begerow et al. 2010). as well as in fungal bar-coding initiatives and efforts to Despite the ITS region’s reputation as useful for detect fungal pathogens in environmental samples. both species identification and phylogenetics, it has Key words: evolution, Fungi, internal transcribed recognized problems. One significant problem is intraspecies ITS variations, which have been identi- fied for some fungal species (Ka˚re´n et al. 1997, Aanen Submitted 26 Oct 2010; accepted for publication 21 Dec 2010. et al. 2001, Smith et al. 2007), thus making it 1 Corresponding author. E-mail: [email protected] necessary to consider a range of ITS sequence variants 731 732 MYCOLOGIA across a population of individuals (Horton 2002, ined, Ganley and Kobayashi (2007) concluded that Nilsson et al. 2008). In addition intragenomic concerted evolution generally acts in a highly efficient variation (sometimes referred to as intrastrain or fashion to eradicate variation. intra-individual variation) among the various copies The small amount of intragenomic variation of the ribosomal array has been recognized in a observed by Ganley and Kobayashi (2007) in Saccha- number of fungal species (O’Donnell and Cigelnik romyces cerevisiae and the conclusion that concerted 1997, Lachance et al. 2003, Pawlowska and Taylor evolution acts efficiently is in contrast to the findings 2004, Avis et al. 2005, Ganley and Kobayashi 2007, of James et al. (2009), who observed significant Smith et al. 2007, Simon and Weiss 2008, James et al. intragenomic variation among 34 strains of S. 2009, Connell et al. 2010, Huang et al. 2010, Santos et cerevisiae for which whole genome data were available. al. 2010), and this same phenomenon has been James et al. (2009) observed significant variation observed in a wide range of organisms, from animals throughout the ribosomal region but noted that the (Harris and Crandall 2000, Leo and Barker 2002, highest variation was confined to the intergenic Wo¨rheide et al. 2004, Sa´nchez and Dorado 2008, spacer (IGS) region. James et al. (2009) also noted Elderkin 2009) to dinoflagellates (Thornhill et al. that many single nucleotide polymorphisms (SNPs) 2007) to prokaryotes (Stewart and Cavanaugh 2007). were unresolved, meaning that specific SNPs oc- Intragenomic variation is believed to be due to a curred on some copies of the ribosomal array and relaxation of concerted evolution, the process that introduced the term pSNP to describe these poly- homogenizes variation among the ribosomal DNA morphisms. pSNPs were more common in strains with repeats in tandem arrays. Evidence suggests that mosaic/hybrid genomes than in strains with typically concerted evolution acts through unequal crossing structured genomes, suggesting that hybridization over between repeating units, gene conversion or plays a role in intragenomic variation. gene amplification, although the exact mechanisms Based on whole genome fungal data, SNPs and that govern concerted evolution are largely unknown pSNPs appear to be prevalent in the ribosomal array (Dover 1993, Elder and Turner 1995, Liao 1999). In of some fungal species, which raises the question of fungi tandem arrays may contain 45–200 copies of the whether these variations are detectible when high ribosomal region (Maleszka and Clark-Walker 1990, throughput cloning and sequencing or next genera- Ganley and Kobayashi 2007) and these copies may be tion sequencing (e.g. pyro-sequencing) techniques distributed across one or more chromosomal loca- are applied to environmental samples. Cloning and tions (Pasero and Marilley 1993), thus allowing for sequencing of ribosomal regions has become popular significant variation within the genome of one as a culture-independent detection method in eco- individual. Both intragenomic variation in the ITS logical and environmental studies, and this approach region, as well as intraspecies variation in ITS at the now has been applied to a wide range of environ- population level, present significant challenges for ments and fungal groups (Jumpponen 2003, Ander- phylogenetic analyses and species identification. son and Cairney 2004, O’Brien et al. 2005, Lindahl et While intraspecies and intragenomic variation in al. 2007, Arnold et al. 2007, Fierer et al. 2007, ribosomal units has been recognized for some time in Fro¨hlich-Nowoisky et al. 2009). Pyro-sequencing is a fungi (e.g. Ka˚re´n et al. 1997, O’Donnell and Cigelnik relatively new technique in fungal ecology but is 1997), the issue of intragenomic variation recently has being used more frequently on environmental sam- received increased attention due to the availability of ples (Bue´e et al. 2009, Gillevet et al. 2009, Jumpponen whole genome data. Whole genome data let Rooney and Jones 2009, O¨ pik et al. 2009, Jumpponen et al. and Ward (2005) analyze many copies of the 5S 2010, Tedersoo et al. 2010) despite a lack of region from four species of Ascomycota, Aspergillus information regarding the biases and limitations of nidulans, Fusarium graminearum, Magnaporthe grisea this new technology (Nilsson et al. 2009, Kunin et al. and Neurospora crassa. Rooney and Ward (2005) 2010, Medinger et al. 2010, Tedersoo et al. 2010). found multiple 5S gene types and pseudogenes within Due to the rate at which sequence data are individual genomes and concluded that the 5S gene is generated with these high throughput technologies, characterized by ‘‘birth-and-death’’ evolution under ‘‘non-consensus’’ rDNA sequences (e.g. cloned or strong selection pressure.