Gearing up for comparative genomics: analyses of the fungal class Dothideomycetes


  • Stephen B. Goodwin,

    Corresponding author
    1. USDA-ARS, Crop Production and Pest Control Research Unit, Department of Botany and Plant Pathology, 915 West State Street, Purdue University, West Lafayette, IN 47907-2054, USA;
    Search for more papers by this author
  • Gerrit H. J. Kema

    1. Plant Research International B.V., Wageningen University and Research Centre, P.O. Box 16, 6700 AA Wageningen, The Netherlands
    Search for more papers by this author

(Author for correspondence:
tel +1 (765) 494-4635; email or

Dothideomycetes Comparative Genomics session: 25th Fungal Genetics Conference, Pacific Grove, CA, USA, March 2009

The Fungal Genetics Conferences, held biannually on the grounds of the Asilomar Conference Center in Pacific Grove, CA, USA, are the premier gatherings for fungal geneticists world-wide. This year's conference, the 25th (, provided four very full days of talks, workshops and posters for almost 1000 participants. Fungi exceed every other kingdom of Eukaryotes in the number of genomic sequences completed or in progress. With this rich resource of available sequences, the focus is now switching from analyses of single genomes to comparative analyses across multiple genomes as a way to understand fungal biology and the interactions of fungi with plants.

Fungi in the class Dothideomycetes are leading the way, and a full concurrent session was devoted to Dothideomycetes Comparative Genomics. This subject was chosen because the Dothideomycetes is one of the largest and most important groups of fungi that collectively infect almost every major monocot and dicot crop, whether for food, feed, fiber or fuel. In addition to plant pathogens, the class includes fungi with an unparalleled ecological, life history and metabolic diversity. Dothideomycetes are present on every continent, including Antarctica (Selbmann et al., 2005), and are important to ecosystem health and global carbon cycling as saprophytes and degraders of plant biomass. Many are lichenized (Del Prado et al., 2006) or are otherwise tolerant of environmental extremes including heat, cold and humidity (Ewaze et al., 2007). Some produce enzymes that help degrade rocks while others can capture and metabolize ethanol vapors (Tribe et al., 2006). A few are pathogens of humans or livestock. Cladosporium herbarum and Alternaria alternata are ubiquitous colonizers of dead plant biomass that play an important role in global carbon cycling, but in addition they are the two most commonly detected human allergens and a leading cause of asthma (Gioulekas et al., 2004). Thus, Dothideomycetes are extremely important to human as well as plant health.

With seven Dothideomycetes genomes sequenced from the orders Capnodiales and Pleosporales (Table 1) and more on the way, a critical mass for comparative analysis has been achieved. The purpose of the Dothideomycetes Comparative Genomics session was to highlight recent progress on comparative analyses of this group. Talks were selected from submitted abstracts because they presented new tools for high-throughput functional analysis that could be applied to other sequenced genomes, or used a comparative genomics approach to reveal new insights into the biology of these fungi. Most Dothideomycetes are plant pathogens, and understanding how they interact with their plant hosts was a major focus of the session.

Table 1.  Sequence resources available currently for comparative analyses of Dothideomycetes genomes
SpeciesOrderCoverageGenome size (Mb)Resources
  1. JGI, The Joint Genome Institute.

Alternaria brassicicolaPleosporales6.4×30.3Washington University (
Cochliobolus heterostrophusPleosporales10.0×34.9JGI (
Mycosphaerella fijiensis Capnodiales7.1×73.4JGI (
Mycosphaerella graminicolaCapnodialesFinished39.7JGI (
Pyrenophora tritici-repentisPleosporales6.9×37.8Broad (
Stagonospora nodorumPleosporales> 10×37.2Broad (

Genetic analysis of Cochliobolus sativus

Shaobin Zhong (North Dakota State University, Fargo, ND, USA) began the session with a talk on new tools for genomic analyses of Cochliobolus sativus (asexual stage: Bipolaris sorokiniana), a pathogen that causes kernel blight, root rot and seedling blight of barley (Hordeum vulgare), wheat (Triticum aestivum) and other grasses. It also causes spot blotch, a very important foliar disease in the northern USA and in tropical environments of Southeast Asia. Genetic and genomic tools are being developed for unraveling the pathogenicity of this fungus, which is highly specialized into several races on barley. Genetic mapping and pulsed-field gel electrophoresis identified the genomic locations for single factors controlling the virulence of the fungus in barley and wheat, which are now being pursued through map-based cloning strategies. Techniques for transformation and RNA-mediated gene silencing were developed for functional analyses that will help to pin down the virulence factors once a genomic sequence becomes available. This is an up-and-coming system with great potential for elucidating host–pathogen interactions in the Dothideomycetes.

Tools for functional analysis in the Pleosporales

Salim Bourras (INRA-Bioger, Versailes, France) moved the focus to high-throughput tools for functional analysis of Leptosphaeria maculans, the black leg fungus of canola (Brassica napus) and other Brassicaceous crops. This fungus alternates necrotrophic and biotrophic growth within a field season. The genome sequence has approximately 12 000 genes but their density varies with position between gene-poor, AT-rich regions (with 1 gene per 29 kb) and GC-equilibrated isochores containing one gene approximately every 2.4 kb. To assign function to these genes, Agrobacterium tumefaciens-mediated transformation (ATMT) was developed to generate a large mutant library. The number of targeted genes was 279, of which 169 have an assigned function. Forty-three genes were recovered that potentially were involved in pathogenicity. The program will be extended and validated with similar large insertional-mutant libraries that are available for the rice blast pathogen, Magnaporthe oryzae, and should help to elucidate the mechanisms by which L. maculans interacts with its Brassica hosts.

High-throughput mutant generation by ATMT was further elaborated by Carrie Smith (Oklahoma State University, Stillwater, OK, USA) working on Phoma medicaginis, a pathogen that causes a defoliating leaf spot of alfalfa (Medicago sativa) in large areas of the USA and also infects the model legume Medicago truncatula. This fungus is asexual and readily forms conidia in pycnidia on artificial media as well as on host plants. ATMT was used to generate an insertional-mutant library and identify genes involved in pathogenicity. Over 1000 mutants were generated (0.03% efficiency) and 10 were selected for an in-depth analysis. Several mutants with insertions in hypothetical proteins had obvious, visible phenotypes such as cracked pycnidia or lack of melanization. One mutant had fluffy, white hyphae that produced no pycnidia or conidia, while one with an insertion in a poly-A RNA polymerase gene produced small and few pycnidia and had reduced pathogenicity. This gene was similar to caffeine-induced death (Cid) genes in Schizosaccharomyces pombe. Homologs were identified in the Dothideomycetes Stagonospora nodorum and Pyrenophora tritici-repentis. With a sequenced host genome and an efficient T-DNA tagging system, P. medicaginis is poised to become an important organism for the analysis of Dothideomycetes–host interactions.

Tools for comparative genomics analysis

The development of tools for comparative analysis has lagged behind the accumulation of genomic sequence data. To address this problem, The Joint Genome Institute (JGI) of the US Department of Energy (DOE) has pioneered a unique approach to software development by asking communities of users to request new tools that are directly needed to further their research. Andrea Aerts (DOE-JGI, Walnut Creek, CA, USA) presented the JGI annotation pipeline for Dothideomycetes genomes. To support the JGI mission of community annotation, six Dothideomycetes genomes (Table 1) are currently available on their web portal, including three that were sequenced outside JGI and, in addition, they have organized on-site annotation jamborees to facilitate whole-genome annotation. The annotation pipeline comprises repeat masking, data mapping, gene prediction, annotation and validation, and culminates in public release of the data. Functional categories according to GO (Gene Ontology), KOG (euKaryotic Orthologous Groups) and KEGG (Kyoto Encyclopedia of Genes and Genomes) assignments can be browsed for each genome, which provides an excellent tool for comparative genomics analyses. Side-by-side analysis of all six genomes facilitates rapid and efficient comparisons, and it is simple to switch from one species to another. An overview of all browsers, synteny, VISTA ( conservation and protein cluster viewers was provided, demonstrating the excellent support that the JGI provides to the Dothideomycetes community. VISTA conservation tracks for all Dothideomycetes genomes and the synteny viewer allow syntenic relationships to be identified and analyzed efficiently. The Dot Plot viewer readily showed the differences between essential and dispensable chromosomes of the Mycosphaerella graminicola genome. These new tools provide an excellent resource for the Dothideomycetes research community and also should be useful for comparative genomics analyses in other organisms, including plants.

Effectors shared among species

Several talks were related to the finished genome of the septoria tritici blotch fungus of wheat, Mycosphaerella graminicola, or the draft sequence of its close relative, the banana black Sigatoka (or leaf streak) pathogen, Mycosphaerella fijiensis. Ioannis Stergiopoulos (Wageningen University, The Netherlands) reported on a comparative study between M. fijiensis and Cladosporium fulvum (syn.: Passalora fulva), a nonobligate, biotrophic pathogen of tomato (Solanum lycopersicum) and reference organism for host–pathogen interactions. To date, 10 effectors have been identified from C. fulvum. All are small, cysteine-rich proteins that are secreted into the apoplast and whose recognition in tomato is mediated by cognate Cf (for C. fulvum) resistance proteins. Although demonstrated for only a few, all of the effector proteins are assumed to be virulence factors (Stergiopoulos & De Wit, 2009). So far no homologs of the C. fulvum effectors had been identified in other fungi, and thus they were considered to be species specific. However, a careful search in the genome sequence of M. fijiensis, based on a combination of comparative genomics and structural analysis, identified putative homologs for at least three of the C. fulvum effectors, two of which are also present in M. graminicola and other Dothideomycetes. Based on these findings, the current hypothesis is that closely related pathogens share some effectors that are most probably inherited from a common ancestor and are adapted to specific host plants. This is the first report on effectors shared among pathogens of such widely divergent hosts, suggesting that similar strategies might have been employed by related plant pathogens that infect unrelated host plants. These results have led to a new model for effectoromics, in which related Dothideomycetes fungi share a set of common effectors that facilitate colonization of many different host species, but have diverged for other effectors that enable host specialization.

Mycosphaerella graminicola to the fore

Sarrah Ben M’Barek (Plant Research International, Wageningen, The Netherlands) presented the current status of the M. graminicola sequence analysis, the first finished genome of a filamentous fungus. High-density linkage mapping with over 2000 sequenced markers enabled a perfect alignment of the genetic map with the genome sequence over all 21 chromosomes. Genome plasticity in this fungus was evident from polymorphisms in chromosome length and number. Meiosis, in contrast to mitosis, frequently generates chromosome number polymorphisms (CNPs), particularly for chromosomes 14–21. These chromosomes are missing frequently in progeny with no visible effect on viability or virulence and appear to be dispensable (Wittenberg et al., in press). Comparative genomic hybridizations using a NimbleGen (Roche NimbleGen Inc., Madison, WI, USA) whole-genome array confirmed the origin of CNPs. The dispensable chromosomes are smaller, have lower gene densities, have a higher density of transposons and contain many unclassified genes that could code for novel proteins. They also contain a high number of redundant copies of genes that are unique on the essential chromosomes. For example, extra copies of tubulin genes on the dispensable chromosomes appeared to be pseudogenes compared with those on the essential chromosomes. The JGI browsers (described by Andrea Aerts) enabled high-resolution self-synteny analyses showing that the dispensable chromosomes are a mosaic of redundant blocks of virtually all other chromosomes. However, the synteny with other Dothideomycetes such as Stagonospora nodorum and M. fijiensis is extremely low. The presence of eight dispensable chromosomes is unique for fungi. However, their function is unknown and will be the subject of future analyses.

Braham Dhillon (Purdue University, West Lafayette, IN, USA) continued analyses of the M. graminicola genome with research on the amplification and apparent inactivation of a gene for cytosine methylation. Analysis of the repetitive fraction of the genome identified a family of 28 repeats, 23 of which contained a region similar to a DNA methyltransferase (DNMT) gene. All copies except for one were located near the telomeres, with the remaining copy on chromosome 6. The DNMT gene was single copy in nine other fungal genomes and, based on syntenic relationships identified by a comparative analysis with the genome sequence of M. fijiensis, the copy on chromosome 6 probably was the original before amplification. Interestingly, all copies of the DNMT gene, including the putative original on chromosome 6, showed evidence of repeat-induced point mutation (RIP), a mechanism in fungi for inactivating transposons by introducing stop codons into reading frames. Tests for methylation showed that M. graminicola lacks cytosine methylation, even though it was present in close relatives. Accidental amplification followed by RIP appears to be a novel mechanism for inactivation of single-copy genes in fungi.

In the final talk of the session, Eva Stukenbrock (University of Aarhus, Aarhus, Denmark) reported on the putative genetic basis of speciation in Mycosphaerella based on comparative genomics. A previous analysis of sequence data had shown very recent divergence times among M. graminicola, two undescribed new species from wild grasses in Iran and the somewhat more distantly related barley pathogen, Septoria passerinii (Stukenbrock et al., 2007). Therefore, genomic comparisons with M. graminicola could reveal evolutionary changes that occurred during speciation. Resequencing one of these related species revealed two haplotypes in the mitochondrial sequences, indicating the first possible report of heteroplasmy in the Dothideomycetes. The nuclear genome sequence was very similar and showed a high degree of synteny with the essential chromosomes 1–13 of the finished genome of M. graminicola. In contrast, synteny with the dispensable chromosomes 14–21 was considerably lower. Regions of repetitive DNA correlated well with nonaligned DNA except for the eight dispensable chromosomes of M. graminicola, probably as a result of their higher content of transposons. A browser developed for comparative genomics showed a significant difference in the ratio of substitutions in both coding and noncoding DNA of the essential and dispensable chromosomes, suggesting that they have different evolutionary patterns. In particular, the ratio of nonsynonymous to synonymous substitutions (dN/dS) was markedly higher on the dispensable chromosomes, giving a strong indication of possible directional selection.


As more genomes are sequenced, the potential power of comparative analyses increases. Being able to search complete genomes for homologs of interesting genes has revealed new insights into the evolution of effectors and the genes on dispensable chromosomes, and has also revealed a potentially new mode of inactivation of an otherwise single-copy gene. Estimates of selection on all genes in an organism will identify those that are under directional selection and might be involved in host–pathogen interactions and speciation. The power of these analyses to illuminate host–pathogen interactions will increase greatly once genome sequences also are available for the host plants. The future of comparative genomics is increasingly bright. The recent session on Dothideomycetes Comparative Genomics provided a first small hint of what is to come as more fungal and plant genomes are sequenced.


We thank Francine Govers and Jay Dunlap for programming the Dothideomycetes Comparative Genomics session of the 25th Fungal Genetics Conference, and Andrea Aerts, Sarrah Ben M’Barek, Salim Bourras, Braham Dhillon, Carrie Smith, Ioannis Stergiopoulos, Eva Stukenbrock, and Shaobin Zhong for speaking at the session and for verifying a previous draft of this report.