Research Article: Comparative Genomics of Plant Fungal Pathogens: The Ustilago-Sporisorium Paradigm

Date Published: July 3, 2014

Publisher: Public Library of Science

Author(s): Theresa Wollenberg, Jan Schirawski, William E. Goldman.


Partial Text

The closely related smut fungi Ustilago maydis, U. hordei, and Sporisorium reilianum f. sp. zeae are facultatively biotrophic basidiomycetes that occur ubiquitously. Teliospores germinate to produce sporidia of different mating type that grow saprophytically and multiply mitotically by budding [1]. For mass proliferation and sexual genetic exchange, successful colonization of economically important crop plants like maize, barley, and oats is a prerequisite. Mating of compatible haploid yeast cells leads to the formation of dikaryotic filaments that are infection competent. These filaments enter their hosts by penetration of the leaf surface [2]. Once inside the plant, filaments multiply in the affected tissue and induce spore formation in tumors near the penetration site (U. maydis) [3] or spread through the entire plant and form spores in inflorescences (S. reilianum and U. hordei) [4], [5]. Although presence of the fungus is clearly detected [6], defense reactions of native host plants are very limited, allowing fungal spread initially without major plant tissue damage. In fact, a living host plant is required to provide nutrients for massive fungal proliferation and successful spore formation.

Smut fungi have intrigued scientists for more than a century for many different reasons, among which are their host specificity, mating behavior, and ability to cause plant disease [7]. Before the molecular era, smut fungi were typically classified by identification of the plant on which symptoms were found, since smuts have a limited host range and many form spores only on a single plant species [8].

Because of its virulence in maize and its molecular accessibility, U. maydis was the first smut fungus to be sequenced [3]. After sequencing by two private companies, it was also sequenced by the public sector using classical Sanger sequencing. Genome sequences of S. reilianum f. sp. zeae and U. hordei were assembled from Roche/454 sequencing reads, and S. reilianum was one of the first eukaryotes to be de novo sequenced using this technology [4]. In contrast to the S. reilianum genome, which could be well assembled from the sequencing reads, assembly of the U. hordei genome was only possible after data integration of a whole genome shotgun and a 10 kb paired-end, as well as an end-sequenced bacterial artificial chromosome (BAC) clone library [5].

U. maydis and S. reilianum f. sp. zeae are both able to form spores on the same host plant, maize, but cause different symptoms (Figure 1A, left panels). Likely, fungal proteins involved in determining symptom specificity during fungal growth in planta are proteins in need of constant change to escape recognition by the plant. Therefore, these proteins are expected to show weak conservation of encoded amino-acid sequences and might thus be recognized by genome comparison. Comparison of the U. maydis and S. reilianum genomes revealed a very high similarity: about 95% of all genes occur in both organisms, and most of them are located at syntenic positions [4]. This allowed a gene-by-gene comparative analysis, which revealed 43 genomic regions containing at least three consecutive genes with predicted protein sequence identities well below average. Some of these divergence regions corresponded to the gene clusters encoding secreted proteins [4] previously identified in the genome of U. maydis[3]. Notably, all U. maydis clusters whose virulence effect was proven were reidentified as divergence regions, while three of the U. maydis clusters whose deletion did not affect virulence were not. Six newly found divergence regions were deleted in U. maydis, and four of them affected virulence [4]. These results confirmed that virulence factors can be efficiently identified using a comparison approach of related pathogens.

The major chromosomal differences between S. reilianum, U. maydis, and U. hordei can be explained by two independent chromosome rearrangements that happened during speciation and separated the most recent common ancestor with a genome organization as in S. reilianum from the U. maydis and the U. hordei lineages. In the U. hordei lineage, the chromosome rearrangement had a profound effect on fungal biology because it placed the before independently segregating a and b mating type loci on the same chromosome, introducing a physical linkage and forcing a bipolar mating behavior on U. hordei[5]. Accumulation of repetitive elements in the intervening regions between the a and b part of the mating locus of U. hordei may have led to suppression of recombination [5], [19], which represents a step towards evolution of a sex chromosome.

Absolutely. As outlined above, a lot can be learned from sequencing and comparing smut fungal genomes. In addition, sequencing of smut fungal transcriptomes at different stages of plant colonization will tell us when or in which tissues virulence effectors are expressed, which will help in the identification of effector targets in the host plant. With each new genome and transcriptome sequence at hand, the prediction of which factors are responsible for the colonization of particular niches (e.g., host plant or host tissue) will become more precise. At the moment it is possible to compare the effector proteins of U. maydis, S. reilianum f. sp. zeae, and U. hordei[5] and generate lists with effectors conserved among all three species (which would be expected to have a general role in virulence) and those that are conserved only in U. maydis and S. reilianum (and would thus be involved in colonization of maize rather than barley). The problem with the current lists is that they are too long to experimentally verify involvement of the target effectors in adaptation to a particular host and that they likely still include many “false positives”, which decreases the chance of experimental validation. Therefore, we need more genome and transcriptome sequences of closely related fungi colonizing different ecological niches. For example, much progress in virulence effector prediction can be expected from sequencing and comparison of the smut fungi S. scitamineum, a pathogen on sugarcane, U. bromivora, a pathogen on Brachypodium, or Thecaphora thlaspeos, a smut fungus able to infect Brassicaceae. However, the more closely related the compared fungi are, the more likely it is that genomic differences reflect host adaptation. Therefore, sequencing of the sorghum pathogen S. reilianum f. sp. reilianum and comparison to its maize-pathogenic relative S. reilianum f. sp. zeae is most promising to identify genes involved in host adaptation.




0 0 vote
Article Rating
Notify of
Inline Feedbacks
View all comments