Latest NEWS!

Report of the Planning Workshop for All Fungi DNA Barcoding
Amy Rossman
Published in: Inoculum, Nov. 2007.
The first meeting to the All Fungi Barcode Initiative was held on 13-15 May 2007 at Front Royal, Virginia. Thirty-seven invited participants (see photograph) from twelve countries met for two days to discuss DNA barcoding of all true Fungi as well as the Oomycota. The goals of the meeting were to arrive at a consensus on a standard gene region for barcoding fungi, brainstorm and prioritize projects, and establish a steering committee. The first day was devoted to reporting experiences with use of the CO1 gene as a barcode followed by those who had used other genes. Although the CO1 gene works for the non-fungal group Oomycota and a few groups of true Fungi, it does not work well for most true Fungi. The consensus of the participants was that the most appropriate gene known at present for DNA barcoding of true Fungi is the ITS region of the nuclear rDNA. The next day researchers reported on fungal DNA barcoding projects already in progress and brainstormed about possible big science projects that involved DNA barcoding. General topics such as vouchering specimens from DNA barcoding, obtaining DNA from herbarium and culture collections, and standards for DNA barcodes were woven into these discussions. The Workshop was organized by Mary Palm, Amy Rossman, JimWhite and a Steering Committee and was funded by the Sloan Foundation.
With 90-99% of the fungi still unknown, it is expected that a great deal of diversity will be revealed through DNA barcoding, as demonstrated by studies using environmental sampling in which major new lineages have been revealed. Those most excited about the application of this technology are ecologists working with environmental samples who would like to determine the role of fungi in, for example, belowground systems associated with roots of forest trees. In addition, DNA barcoding of fungi would be extremely useful to biosecurity and agricultural quarantine agencies around the world to prevent the spread of invasive fungi. DNA barcoding can be regarded as ‘horizontal genomics,’ i.e. determining one gene for all species instead of all genes for one species, analogous to functional genomics in identifying “up and down regulated” species rather than genes in environmental samples. Following is a summary of discussion at this workshop. A more detailed report can be found at: http://www.allfungi.org/
Front L to R: Francois Lutzoni, Jean-Marc Moncalvo, Todd DeSantos, Conrad Schoch, Keith Seifert, Andre Levesque, Pedro Crous, Karen Hughes, Mary Palm, Barbara Robbertse with Alex. Back: Ursula Eberhardt, Rytas Vilgalys, Irina Druzhinina, Dirk Redecker, Barbara Paulus, Joe Bischoff, Sung-Oui Suh, Tom Bruns, Teun Boekhout, Gerald Bills, Dominik Bergerow, Martin Bitardondo, Mehrdad Hajibabaei, Andy Miller, Martha Powell, John Taylor, Emma Steenkemp, Richard Hamelin, Lisa Castlebury, Matteo Garboletto, Amy Rossman, David Hibbett, Cletus Kurtzman, Jim White. Missing: Urmas Koljalg and Juan McEwan

Standard DNA Barcode for True Fungi
Requirements for DNA barcode for true fungi
The problem of selecting a barcoding locus was based on the realization that the locus needed to be multicopy, have robust primer sites, and vary between species. The multicopy requirement was based on the need to obtain sequences from old and degraded specimens limiting our choices to rDNA repeat segments or mtDNA loci. The robust primer requirement was based on the need to amplify from environmental samples with unknown contents, and only rDNA segments meet this requirement. The need for interspecies variation is obvious and it was acknowledged that no single locus would always meet this need.
Experience with CO1 for fungi - CO1 is the designated standard for animals and there would be an obvious advantage to having one locus for all groups of organisms. The main problem with CO1 is that in true Fungi the regions used for PCR priming evolve more rapidly than those used for rDNA.With large organisms, whether in collections or in nature, this variation is not a problem because biologists seeking to employ DNA barcoding can select an appropriate pair of primers based on the phenotype of the organism. However, with microbes in the environment, this variation is a fatal flaw. For example, different CO1 primer pairs are needed for different genera of fungi, so a mixture of primer pairs necessary to recover even a few known genera and unknown groups of fungi would be missed in environmental sampling. Although CO1 is valuable when examining known fungi in collections, e.g., Penicillium, or compact clades of Oomycota, it will not be useful in environmental surveys of fungi.
Problems with CO1 for fungi - There are additional problems with CO1 in true Fungi involving amplification because of extreme length variation in this gene observed in the mitochondrial genomes sequenced to date. This results from the unpredictable presence of multiple and sometimes mobile introns of differing lengths, and the presence of multiple copies of the gene including possible nuclear copies. While the presence of multiple copies could be handled using reverse transcriptase PCR to avoid introns and work only with copies that are transcribed, the attendant problems of handling RNA are less amenable to high throughput methods and not suitable for herbarium specimens. The occurrence of multiple copies of CO1 in strains of Fusarium and the Aspergillus niger complex made it difficult to evaluate the resolution of the gene in these fungi. CO1 did not provide sufficient resolution between some taxa, such as the oyster mushroom fungi related to Pleurotus ostreatus. Cloning may be necessary in order to successfully sequence some of the rusts but this is also true for sequencing ITS in many rust fungi. It was noted that the variability in CO1 across true Fungi was much greater than that represented by animals and that the mitochondrial genome evolves in a different way in true Fungi than in animals. Thus, it is not surprising that, although CO1 works for the animals, it does not work well for the highly variable true Fungi. The Oomycota, which do not appear to have introns in CO1, could be different case, but so far the sampling of Oomycota is mostly limited to Pythium and Phytophthora spp.
Qualifications of ITS region of rDNA - Within the true Fungi the ITS region varies from 527-700 bp and thus is easy to sequence using high throughput techniques. A huge database of ITS sequences already exists and this gene has been sequenced as part of the AFTOL project and the UNITE consortium (see Appendix 2). ITS has been sequenced for a sampling of all the major groups of true Fungi. In addition fungal-specific ITS primers exist, have been widely applied, and are known to work well across the major groups of fungi. Participants agreed that they could meet CBOL’s requirements for proposing a non-COI barcode region and they prepared preliminary responses to these requirements. A formal proposal to make ITS the barcode region for fungi will be submitted to CBOL. In addition, the ITS has been almost universally used by mycologists in sequencing fungi as evidenced by the DNA barcoding projects listed below all of which have used the ITS region, although some also sequenced additional genes. The ITS region evolves rapidly because only a few regions are constrained by the need for the ITS to be spliced out of the rDNA transcript. The rest of the region can evolve without much selection, hence change is rapid. While useful at the species or close to species level, alignment of the ITS region between major groups of fungi and even within order and families is usually not possible, thus a global ITS alignment cannot be presented here. Assuming the ITS sequences were alignable, the name of this gene as deposited in GenBank has varied with the over 65,000 accessions as has the length of the sequence submitted making extraction of all sequences of true Fungi difficult.
Problems with ITS region - Although the ITS region of the rDNA appears to be the best of choice for a universal DNA barcode for true Fungi, there are some problems with its use. In some groups of fungi the ITS is too variable to determine major group. Yeasts and AM fungi were mentioned in this regard and this problem was the main argument for LSU, either alone or in addition to ITS. However, most participants viewed the ‘too-muchvariation’ problem as an issue that would be solved by increased sampling. In addition there were a number of good examples, especially from plant pathogenic ascomycetes, where ITS clearly does not separate distinct species. In addition, the ITS has paralogs as determined for some Fusarium species and the Glomeromycota (AM fungi). The solution suggested was to use ITS as the first step or first key and a second gene for a precise identification. This approach has been used successfully in the TrichoKey DNA barcoding system. For each group of fungi this locus will need to be determined, although the translation elongation factor 1-alpha (EF1a) and RPB2 have been shown to be useful in several groups.
Other possible regions for fungi - ITS plus the D1/D2 LSU rDNA region was also considered but rejected because, for most groups of fungi, the D1/D2 region does not give sufficient signal at the species level. In addition, more than one primer in each direction is required, thus it does not lend itself to high throughput. Considering that sequencing is becoming less expensive and the real cost is in the collection of the samples, sequencing this longer region might be desirable in the future. With so many fungi still not sequenced, the ITS often does not place an unknown with even a close match and thus the D1/D2 region works for placing unknown species especially when isolating fungi from the environment.
Decision - After considerable discussion the decision was made that the entire ITS region should serve as the DNA barcoding locus for the true Fungi, but not necessarily the Oomycota. The two other contenders for the DNA barcoding locus were CO1 or the 5’ end of LSU.

Potential Projects
Air spora.
Sampling devices exist for sampling the air spora that are known to include many different kinds of fungi with some species known to travel many thousands of miles. DNA barcodes of these fungi would be extremely useful to scientists such as ecologists tracking the distribution of fungi from habitats such as the Saharan desert, plant quarantine officials tracking the movement of invasive fungi and plant pathogens such as soybean rust, and physicians checking the distribution of fungi that cause allergies. Fungi associated with global trade. Knowledge and rapid identification techniques for the fungi associated with agricultural and horticultural commodities are needed to prevent the spread of invasive species. Initially this would be a USDA project to barcode fungi that are currently being intercepted on imported plant products at the US ports of entry. Many of these are interesting, unusual, and even unknown species in addition to being difficult to identify. Cultures for sequencing would be obtained from intercepted specimens; specimens would be deposited at BPI with cultures at CBS.
Fungi of boreal forests. As the climate changes, the boreal forests are likely to be impacted as already evidenced by changes in fungal disease distribution. DNA barcoding of fungi in these forests would provide a baseline for future changes to the organisms that depend on fungi especially the trees themselves that require mycorrhizal fungi for survival.
Fungi of the National Parks of Canada and the US and/or UNESCO Heritage sites. Knowledge of the fungi of these important preserved natural areas is critical for monitoring changes that may occur over time. The project on the Great Smoky Mt. National Park described in Appendix 2 serves as a model for an expanded project in which many groups of fungi are collected, identified and vouchered over time with DNA barcodes obtained. DNA barcodes would be linked to biodiversity data.
Medically important fungi. Fungi cause a number of serious human diseases. The causal fungi belong to a number of groups with the diversity of fungal pathogens increasing especially for humans who are immunocompromised. While many of these fungi are relatively well- known and commercial tests for some species exist to determine their identify, these tests are not standardized using different molecular approaches. A standard DNA barcode for human pathogenic fungi would be extremely useful.
Determine sequence variability among strains of species for proposed DNA barcoding genes. Barcoding is of little value unless extent of resolution is understood.
Diverse species known from multiple strains would be selected and characterized using multigene analyses or genetic crosses to verify that they are really conspecific. In this manner the resolution of these strains from the barcode gene would be tested. Such tests should be conducted across the groups of fungi as part of the basic systematic knowledge upon which DNA barcoding is based.
Development of gene chips for identification of fungi. As each of these projects is completed, they would result in the potential to develop microarrays, i.e. gene chips, for detection and identification of fungi from environmental samples without the need for cultivation or PCR. The project being developed at present involves the 1,200 species of ectomycorrhizal fungi that have been barcoded as part of the UNITE project.
Bioinformatics tools for the application of barcode sequences i.e. integrated barcode methods. Although the planning workshop members selected ITS as the DNA barcode for true Fungi, it is acknowledged that for many groups of fungi this region does not result in an accurate species identification and that a second gene region is required. In addition, sequence similarity search like BLAST has a number of pitfalls and should be applied and interpreted carefully, mainly due to the presence of conserved sequence region 5.8S rRNA inside the barcode locus and database population. Therefore it would be useful to discuss methods of barcode identifications currently used in different groups of true Fungi and formulate some general recommendations valid for the majority of fungi. The International Subcommission on Trichoderma and Hypocrea (ISTH) DNA barcode systems exemplifies this problem and one solution to it.

DNA Barcoding Projects of Fungi Already in Progress, all use ITS
UNITE –
The project provides DNA barcodes for about 1200 species in 120 genera of ectomycorrhizal fungi from northern Europe. The on-line database of sequences has three options for searching including an option to align and make a tree with the unknown sequence. As yet, this database is not populated with ascomycetes.
International Subcommission on Trichoderma and Hypocrea (ISTH) – This Web site provides complete DNA barcodes of all formally described and several potentially new Hypocrea/Trichoderma species (up to 200). This system uses an automated barcode program based on ITS1 and 2 (TrichOKEY) and TrichoBLAST – multiloci sequence similarity search based on ITS1 and 2, two separate introns of tef1, partial rpb2, and chi18-5 (former ech42). The TrichOKEY system includes measures of reliability based on the known intraspecific variability and discriminates between incomplete, poor, non-Trichoderma sequences, or sequences of new alleles/species. The resulting species is linked to a biodiversity profile and the most recent phylogeny of the genus. The whole database is linked to GenBank.
Yeast Fungi – Between Clete Kurtzman, USDA, Jack Fell, RSMAS, University of Miami, and Vincent Robert, CBS, all known species of yeast fungi have been DNA barcoded using the D1/D2 plus ITS gene regions. Most of these sequences are in GenBank and include both the ascomycetous and basidiomycetous yeasts.
Herbarium Specimens of Macrofungi at K and Venice – Using robotics for DNA extraction and purification, Martin Bitardondo sequenced two specimens for each species of two major groups of macrofungi at K. He determined that specimens collected within the last 30 years yielded useful DNA without applying special measures and that many matched “unknown” fungal DNA sequences in GenBank. Matteo Garboletto sequenced 6000 recently deposited herbarium specimens of macrofungi from Venice with considerable success.
Macrofungi from the Great Smoky Mountains National Park All Taxa Biodiversity Inventory – Three NSF projects have been funded for collection and ITS sequencing of fungi in the Great Smoky Mountains National Park: Agaricales (R. H. Petersen and K. W. Hughes), Pyrenomycetes (A. Miller), and Harpellales (B. Lichtwardt and M. White). At least 1,500 ITS sequences are available representing inter- and infraspecies variation from these projects. ITS sequences do not always distinguish morphospecies in some cases but, in other cases, clearly identify cryptic species.
Biodiversity and Prospecting – Gerald Bills (Merck) reported that his company decided to focus its natural products group on the discovery of novel antibiotics against bacterial and fungal pathogens of humans. Their group is cataloging and mapping the phylogeny of fungi with antibiotic properties. They use the map to determine metabolite-rich taxa, if an isolate is new, and to avoid repeatedly screening the same kinds of fungi.
CBS Fungal Culture Collection – CBS is in the process of DNA barcoding all of their cultures. These in- clude many economically important fungi such as human pathogens and food-associated fungi.
Assembling the Fungal Tree of Life (AFToL) – This project provides a comprehensive multigene phylogeny for the true Fungi with representatives of all fungal families including over 1,000 species with documented specimens and well resolved placement. As part of this project the ITS has been sequenced as a DNA barcode gene. The AFTol project successfully developing a bioinformatics system called WASABI that will be considered for use in a DNA barcoding initiative for fungi.
Ecbol.org – This is a list of scientists conducting barcoding projects in Europe.