ECFG 9 Parallel Session 1.Functional and Comparative Genomics
Yeasts and filamentous fungi, the great divide?
University O Cambridge, Cambridge, United Kingdom
The Kingdom Fungi is a large and very diverse one. The extent of sequence divergence within the Saccharomycotina (budding yeasts) is as great as that within the phylum Chordata of the animal kingdom. However, these unicellular eukaryotes all have a similar growth habit, which appears very different from that of the filamentous fungi. Indeed, the Pezizomycotina (filamentous ascomycetes) diverged from Saccharomycotina some 900-1000 million years ago. In terms of their growth habit, the filamentous fungi would appear more similar to the Oomycetes (like Phytophthora) than they are to the budding yeasts. Yet the Fungi are more closely related to the animals than they are to the Oomycetes. What, then, are the important characteristics of filamentous organisms, and just how different are the filamentous fungi from the yeasts? To what extent has gene exchange between the members of these groups, and between them and other kinds of organisms, played a part in their evolution? What is the position of dimorphic organisms, and are there common characteristics to all forms of polar growth in the Fungi? This talk will use a comparative genomics approach to address these questions by exploiting the resources of the e-Fungi data warehouse (http://www.e-fungi.org.uk).
Comparative metabolic analysis of Aspergilli at the genome level
Chalmers, Gothenburg, Sweden
Aspergilli are used extensively in the fermentation industry for the production of a range of different industrial enzymes, organic acids and high-value secondary metabolites. In connection with further development of bioprocesses for the production of fuels and chemicals these fungi are interesting versatile cell factories as they tolerates low pH, can utilize a wide range of carbon sources and has relatively high conversion rates. In connection with developing A. niger or A. oryzae as general cell factory platforms we have established a number of systems biology technologies that can be used for enhancing the metabolic engineering of these organism in the future. These techniques include genome scale metabolic models, both with about 1,400 biochemical reactions. In these models there is a link between reactions, enzymes and genes and hereby it is possible to rapidly map transcriptional changes on a global metabolic map. The models contain a large number of extracellular reactions for the degradation of complex carbohydrates, and these reactions form a complex network which has been analyzed using genome-wide transcription analysis during growth on different carbon sources, e.g. glucose, xylose, xylan, starch and arabinan. For transcription analysis we are using a home designed Affymetrix DNA array, that covers all the annotated ORFs of A. niger, A. oryzae and A. nidulans. This DNA array has been used for transcriptional profiling at different growth conditions. Specifically we have looked into the effect of growth on different carbon sources and the influence of pH.
Aspergillus niger genomics: a tale of two strains
Scott Baker1, Mikael Andersen2, Margarita Salazar2, Diego Martinez3, Igor Grigoriev4, Jane Grimwood5, Jens Nielsen2
1Pacific Northwest National Laboratory, Richland, WA, United States, 2Technical University of Denmark, Lyngby, Denmark, 3Joint Genome Institute/LANL, Los Alamos, NM, United States, 4Joint Genome Institute/PGF, Walnut Creek, CA, United States, 5Stanford University, Palo Alto, CA, United States
Used as both an organic acid and protein production strain, Aspergillus niger is a scientifically and economically important filamentous ascomycete. High quality draft genomic sequences for two strains of Aspergillus niger have been released. Genomic sequence was generated by the US DOE Joint Genome Institute (JGI) for A. niger strain ATCC 1015, an organic acid producer. The approximate size of the A. niger strain ATCC 1015 genome is approximately 34.85Mb, approximately 0.9Mb more sequence compared to the currently released genome of A. niger strain CBS 513.88, a protein production strain. The A. niger strain ATCC 1015 genome is spread across 24 scaffolds, none of which contain gaps. One half of the telomeres were "captured" in the eight chromosome A. niger strain ATCC 1015 sequence. Aspergillus niger strains ATCC 1015 and CBS 513.88 were compared at both the genome and transcriptome levels. At the genomic level, there are major chromosomal rearrangements between the two strains, including a chromosomal translocation and two large inversions. Using the JGI gene models as the basis for comparison there are differences in codon usage. Furthermore, the level of single nucleotide polymorphisms is greater between A. niger strains CBS 513.88 and ATCC 1015 than between A. niger strains ATCC 9029 and ATCC 1015. At the level of the transcriptome comparing batch fermentations of A. niger strains ATCC 1015 and CBS 513.888, we observed significant differences amino acid biosynthetic pathways. In addition, while the level of glucoamylase gene expression was ~3-fold higher in A. niger strain CBS 513.88 compared to strain ATCC 1015, measured glucoamylase activity was ~6-fold higher. In summary, there are significant and intriguing differences at the genome and transcriptome level between the two A. niger strains.
Global transcription regulation and genome organization in Candida albicans
Karen E.S. Tang, Kelly Bouchonville, Anna Selmecki, Anja Forche, Judith Berman
University of Minnesota, Minneapolis, United States
With the goal of understanding global features of transcription wiring networks, we analyzed all available Candida albicans microarray data using the Iterative Signature Algorithm (ISA) to cluster genes and experimental conditions into 'transcription modules'. Genes in the same transcription module co-regulate under the module conditions. Any given gene or experimental condition can be present in multiple modules, and module membership depends on the input stringency of co-regulation. The overall organization of the ISA results are displayed as a module ‘tree’, with the 'branches' representing major biological processes (e.g., peptide catabolism). In 2004, we analyzed the then available ~250 C. albicans microarray data sets using the ISA. We have now extended this analysis to a much larger set of available data with ~500 different experiments generously provided by members of the Candida community. Major conclusions to date from the new analysis are: 1) The major module branches have been preserved between the old and new analyses. These reflect major processes like protein synthesis, carbohydrate metabolism, stress responses, hyphal growth and opaque-specific gene expression. 2) Novel branches appeared in the new analysis that were not present in the previous study. We have also applied several whole genome approaches including comparative genome hybridization and SNP microarrays to monitor aneuploidy and loss of heterozygosity in this diploid parasexual human pathogen. We used the data to predict the proportion of whole chromosome and segemental aneuploidy present in strains that were studied and found that aneuploidy occurs in ~50% of drug resistant strains and ~10% of strains that have undergone laboratory manipulations such as transformation with DNA. Specific aneuploidies such as isochromosome 5L contribute to drug resistance by providing two extra copies of two specific genes, ERG11 and TAC1, on Ch5L. Furthermore, strains that contain one or more aneuploid chromosomes have a higher likelihood of changing their chromosome profile upon transformation. Our studies provide a global view of the patterns of gene expression and the patterns of genome alterations that confer different phenotypes like drug resistance.
Building networks: functional gene analysis during hyphal fusion in Neurospora crassa
Alex Lichius, Nick D. Read
University of Edinburgh, Edinburgh, United Kingdom
Polarised tip growth and hyphal fusion are essential for the establishment of an interconnected colony network, and are key features of filamentous fungi. Hyphal fusion comprises the orchestration of cell-cell recognition, homing, attachment, depolarisation, cell wall remodelling involving fusion pore formation, and plasma membrane fusion. Hyphal fusion in Neurospora crassa and appressorium-mediated penetration in the plant pathogen Magnaporthe oryzae share a number of these processes in common. By applying a comparative genomics approach and analyzing deletion mutants of genes in N. crassa orthologous to genes essential for pathogenicity in M. oryzae we have identified 12 novel fusion mutants. Amongst the encoded proteins are members of all three MAPK signalling pathways in N. crassa and M. oryzae. In addition, other components identified provide the first evidence for the involvement of GTPase-mediated signalling and reactive oxygen species-dependent regulation of hyphal fusion. We are now using live-cell imaging techniques to analyse the roles of these proteins during hyphal fusion to develop a model of the underlying signalling network. The identification of additional novel genes involved in hyphal fusion, but which have not yet been analysed in M. oryzae, should help us predict novel pathogenicity genes in plant pathogens in the future.
Comparative genomics as tool for the identification of species-specific virulence factors in smut fungi
Jan Schirawski1, Thomas Brefort1, Kerstin Schipper1, Julia Schöning1, Olaf Müller1, Gertrud Mannhaupt2, Regine Kahmann1
1Max Planck Institute for Terrestrial Microbiology, 35043 Marburg, Germany, 2Institute for Bioinformatics (MIPS), GSF National Research Center, 85764 Neuherberg, Germany
Sporisorium reilianum and Ustilago maydis are two closely related plant pathogenic fungi that parasitize the same host plant, Zea mays. S. reilianum is a typical smut fungus that infects the plant at seedling stage and leads to a systemic, initially symptomless infection of the whole plant. The infection becomes obvious only after the plant has set flowers when spore-filled sori replace the inflorescences. In contrast, U. maydis locally ramifies in the tissue after penetration and induces the formation of tumors, in which the fungal spores develop.
We used comparative genomics to identify the molecular basis of the difference in life style after plant penetration. To this end, we have sequenced the genome of S. reilianum and have aligned the available sequences with the recently published genome of U. maydis. While both genomes are highly syntenic and most putative gene products show more than 70% amino acid identity, a gene-for-gene comparison revealed the existence of regions of low sequence conservation. Among these are the six gene clusters encoding potentially secreted proteins that were previously identified in U. maydis as containing virulence determinants. Four of these gene clusters were deleted in S. reilianum. While two of the respective cluster deletions had a dramatic effect on virulence, the two other cluster deletions did not influence virulence of S. reilianum. This shows that genome comparison can reveal gene clusters containing species-specific virulence factors in smut fungi.
Novel ways to explore sequenced fungal genomes
Kim Hammond-Kosack, Andrew Beacham, Thomas Baldwin, Martin Urban, Salvador Gazan, John Antoniw
Rothamsted Research, Harpenden, United Kingdom
Complete genomic sequence information is now readily available for many pathogenic and non-pathogenic fungal species and in the near future this will extend to different strains of a single species. For some species this new sequence information has been aligned to an existing genetic map. By exploring and visualising gene content and gene type within the chromosome landscape new information on the evolution of fungal genomes can be obtained.
The sequenced genome of the cereal attacking Ascomycete pathogen Fusarium graminearum (teleomorph Gibberella zeae) has been explored at various scales using a new visualisation tool in combination with statistics, Affymetrix gene expression data, gene function categories and recombination frequencies.
These combined analyses have revealed that genes coding for specific protein types are non randomly distributed over the four F. graminearum chromosomes. Some gene types are located only in centromeric regions, whereas others only reside in sub-telomeric regions, for example in planta expressed genes. In addition, regions of high recombination tend to harbour specific protein classes, for example polyketide synthases, whereas other gene types are solely located in regions of low /no recombination. These analyses have also revealed the existence on chromosome 1 of a ~25 kb micro-region rich in homologues of pathogenicity genes and that most of the experimentally verified F.graminearum pathogenicity genes do not reside in either sub-telomeric regions or regions with a high recombination frequency.
The four large F. graminearum chromosomes are hypothesised to have arisen from a series of earlier chromosome fusion events. The gene content of these genomic sub-regions is unique and reflects this suspected evolutionary path.
Insights in short term evolution of Aspergillus niger strains
Peter J.I. van de Vondervoort1, Mikael R. Andersen2, Scott E. Baker3, Martin Meijer4, Noël N.M.E. van Peij1, Herman J. Pel1, Hans A. Roubos1, Robert A. Samson4, Peter J. Schaap5
1DSM Food Specialties, Delft, Netherlands, 2Technical University of Denmark, Lyngby, Denmark, 3Pacific Northwest National Laboratory, Richland, United States, 4Centraal Bureau voor Schimmelcultures, Utrecht, Netherlands, 5Wageningen University, Wageningen, Netherlands
In most post-genomic studies genomes are treated as static entities describing the genomic potential of a species. To cope with the challenges of a constantly changing environment in reality genomes are dynamic systems subject to constant changes. Due to the costs involved eukaryotic species are rarely sequenced twice. This is not so for Aspergillus niger, an important fungal species used in the commercial production of various extracellular enzymes and organic acids. Currently a near complete genome sequence with a high quality annotation is publicly available of strain CBS513.88, a selected industrial enzyme producer (Pel et al., Nature Biotech. 2007) and a draft sequence of strain ATCC1015 a historic citric acid producer can be retrieved from DOE. (http://genome.jgi-psf.org/Aspni1/Aspni1.home.html).
Both genome sequences have been assembled successfully into a set of large scaffolds but both sets still suffer from a reasonable number of sequence gaps. To improve annotation and to allow a detailed high quality genome comparison between the respective A. niger strains PCR-sequencing is being used for gap closure in the A. niger CBS513.88 using the ATCC1015 genome sequence as reference.
A comparison of the revised CBS513.88 genome sequence with ATCC1015 genome sequence reveals that a number of the approximately 14,700 protein-coding genes in CBS513.88 is lacking in the ATCC1015 strain. Overall these strains are virtually identical at DNA level, however a number of large and small chromosomal rearrangements are observed. Some of these chromosomal recombination events and their consequences will be discussed in more detail.