XXI Fungal Genetics Conference
Asilomar, California
March 2001

Genomics and Proteomics

125 Physical mapping and functional analysis of chromosome IV pericentric region in Aspergillus nidulans. Alexei Aleksenko and Michael Lynge Nielsen. BioCentrum, Technical University of Denmark, Bygn. 223, Lyngby 2800 Denmark

Chromosome walk between genes methG, uvsB, hisA, gdhB and frA of A. nidulans produced two continuous segments of a physical map of chromosome IV covering 250kb and 160 kb. Integration of selectable markers into the chromosome targeted by unique cosmid fragments made it possible to establish that the two fragments flank the centromere, with the order of markers methG-uvsB-CEN-hisA-gdhB-frA. The gap in the physical map over the centromere itself was estimated to be between 70 kb and 250 kb. It was shown that the frequency of meiotic recombination in the vicinity of the centromere is strongly reduced. Pericentric fragnets contain highly repetitive AT-rich DNA and degenerated retrotransposons. Centromere-proximal DNA fragments were subcloned in circular and linear vectors, and their behavior in transformation was studied.

126 Heterologous transposition in the phytopathogenic fungus Ustilago maydis using the Caenorhabditis elegans element Tc1. Oliver Ladendorf and Joerg Kaemper. Max Planck Institute for Terrestrial Microbiology, Karl-von-Frisch-Str., D-35043 Marburg, Germany

Development of molecular techniques for phytopathogenic fungi aimes at the identification of genes whose products are essential for a successful infection of the host plant. Initial approaches have relied on isolating genes and generating knockouts by transformation-mediated insertional mutagenesis. This method presents several drawbacks like the formation of tandem integrations or the creation of DNA rearrangements. As an alternative strategy we have now exploited the versatility of transposons for the generation of random mutations in the phytopathogenic basidiomycete Ustilago maydis. For this purpose we have taken advantage of the Tc1 element from Caenorhabditis elegans, known to transpose independently of host-specific factors in evolutionary distant organisms. We have constructed an U. maydis strain constitutively expressing the Tc1 transposase and harbouring the Tc1 transposon with a hygromycin resistance gene on a free replicating plasmid. The use of the nar1 gene (coding for nitrate reductase) as a marker allowed both the selection (on nitrate medium) and counterselection (on chlorate medium) of the transposon-donor-plasmid. Upon curing of the plasmid, hygromycin resistant colonies, indicative for integration events of Tc1 into the genome, were identified. Via inverse PCR it was possible to determine the positions of integrated Tc1 elements. The repetitive ends of the transposon were not altered and the integration was accompanied with a duplication of the dinucleotide TA at the target site, consistent with data for transposition events in C. elegans. The experiments demonstrate for the first time the transposition of a heterologous transposon in U. maydis and underline the potential of Tc1 as a molecular tool.

127 Identification of major membrane glycosylphosphatidylinositol-anchored proteins in Aspergillus fumigatus. Jean-Michel Bruneau1 , Thierry Magnin2 , Eric Tagat1 , Raymond Legrand3 , Muriel Bernard2 , Michel Diaquin2 , Claude Fudali3 , and Jean-Paul Latge2. 1 Infectious Disease Group, Biochemistry Department, and 3 Core Research Functions, Biophysics Department, Aventis Pharma, 93235 Romainville, France and 2 Unit des Aspergillus, Institut Pasteur, 75724 Paris, France

Previous studies (Mouyna I., Fontaine T., Vai M., Monod M., Fonzi W.A., Diaquin M., Popolo L., Hartland R.P., and Latg J.-P. (2000) J. Biol. Chem. 275: 14882- 14889) have shown that glycosylphosphatidylinositol anchored proteins play an important role in fungal cell wall biosynthesis. GPI-anchored proteins of Aspergillus fumigatus have been purified and separated by 2-dimensional electrophoresis. They were characterized by their peptide mass fingerprint through MALDI-TOF mass spectrometry and by internal amino-acid sequences obtained by Edman sequencing, nano-ES-MS/MS or Q-TOF-MS/MS. Most identified proteins were homologous to putative GPI-anchored proteins present in other fungi. One protein only displayed amino-acid sequences not found in sequence databases. Several of the GPI-anchored proteins identified in A. fumigatus were orthologs of genes (CRH1, CRH2, ECM33, GAS1) known to play a role in yeast cell wall biogenesis. Furthermore, a comparative study performed with chitin synthase and glucanosyl transferase mutants of A. fumigatus showed an altered GPI-anchored proteins pattern present in the mutant cell membrane compared to the wild type strain. This result suggests that a number of GPI-anchored proteins identified in this study may be involved in A. fumigatus cell wall biosynthesis and / or remodeling.

128 Genomes to phenomes: genome scale mutagenesis and function analysis in the filamentous fungi. Jeffery R. Shuster, Kiichi Adachi, Matthew Tanzer, Sanjoy Mahanty, Lakshman Ramamurthy, Maria V. Montenegro-Chamorro, Clive Lo, Rex Tarpey, Amy Skalchunes, Ryan Heiniger, Sheryl Frank, Blaise Darveaux, Todd Dezwaan, Grant Nelson, and Lisbeth Hamer. Paradigm Genetics Inc., 108 Alexander Drive, RTP, NC 27709.

The filamentous fungi include a number of organisms that have significant effects in the areas of human health and nutrition. Some fungi are major pathogens of the world's food crops while other are causal agents of human diseases. Fungi are also employed as biological factories producing a number of important metabolites such as such as antibiotics and other pharmaceutical compounds. They are powerful producers of many industrial enzymes and organic acids. Although there is a paucity of full genome information for the filamentous fungi, recent advances in DNA-based technologies allow for full functional genomic analyses. A fungal phenome may be defined as the total discernable phenotypes arising from a genome wide mutagenesis program. A method to define the fungal phenome will be described. The method is comprised of a high throughput mutagenesis method, Transposon Arrayed Gene Knock Out (TAG-KOTM), a bioinformatics pipeline, and a broad spectrum high throughput phenotypic analysis program. 1

129 Analysis of ESTs from two time of day-specific libraries of Neurospora crassa reveals novel clock-controlled genes. Minou Nowrousian1, Hildur V. Colot1, Hua Zhu2, Doris Kupfer2, Gloria Berrocal-Tito1, Hongshing Lai2, Deborah Bell-Pedersen3, Bruce A. Roe2, Jennifer J. Loros1, Jay C. Dunlap1. 1Dartmouth Medical School, Genetics and Biochemistry, Hanover, NH, 03755. 2University of Oklahoma, Chemistry & Biochemistry, Norman OK USA. 3Texas A&M University, Biological Sciences, College Station TX USA

In an effort to determine genes which are expressed in mycelial cultures of Neurospora crassa over the course of the circadian day, we have participated in a collaborative effort to sequence 13,000 cDNA clones from two time of day- specific libraries (morning and evening library) generating approximately 20,000 sequences. Contig analysis allowed the identification of 445 unique sequences and 986 genes present in multiple cDNA clones. For about 50% of the sequences, significant matches to sequences in the NCBI database (of known or unknown function) were detected. The remaining ESTs showed no similarity to previously identified genes. EST sequences and results of comparisons can be obtained from the following web site: http://www.genome.ou.edu/fungal.html. We hybridized Northern blots with probes derived from 26 clones chosen from contigs identified by multiple cDNA clones. Our results indicate that the representation of genes among the morning and evening sequences, respectively, in most cases does not reflect their expression patterns over the course of the day. Nevertheless, we were able to identify four new clock controlled genes. Based upon these data we predict that a significant proportion of Neurospora genes may be regulated by the circadian clock. The 1431 different genes which are represented among the ESTs from both libraries represent an estimated 10% of all Neurospora genes. They may be used for further large scale analyses such as microarray approaches to elucidate the extent of clock control in Neurospora crassa .

130 Identification of stage specific proteins in Phytophthora spp. Pieter van West, Shuang Li, Samantha Shepherd, and Neil A.R. Gow. University of Aberdeen, Department of Molecular & Cell Biology, IMS, Foresterhill AB25 2ZD, Aberdeen, Scotland.

Pathogens of the oomycete genera Phytophthora cause destructive diseases of hundreds of commercially important plant species. Despite their economic importance, little is known about the molecular mechanisms accounting for the success of oomycetes as plant pathogens, or the fundamental molecular processes underlying their development. Oomycetes have many fungus-like characteristics, but are not true-fungi. A number of studies have indicated that they should be classified as Stramenopiles that include the golden-brown algae (Kamoun et al., 1999, TIPS 4: 195-200). This implies that oomycetes have distinct genetic and biochemical mechanisms involved in the interaction with plants that may be significantly different from those in true fungi. Oomycetes have several clearly defined developmental stages in their life cycle. They produce biflagellated zoospores that are able to swim in water surfaces on plant tissues and in the soil, enabling the pathogen to select its infection site. The zoospore encysts, forms a germ tube, from which an appressorium is produced that enables the pathogen to penetrate the plant surface. All these stages of the life cycle are experimentally tractable since they can be produced in vitro (van West et al., 1998, FGB 23, 126-138). A proteomics approach is being employed to identify stage-specific and extra-cellular proteins from two economically important Phytophthora species, P. infestans and P. palmivora. Our aim is to characterise stage specific and secreted proteins at the molecular level. Such proteins might play essential roles in the development of Phytophthora species.

131 Analysis of transcriptional control circuits and their gene targets using transcript profiling in Candida albicans. Christophe d'Enfert1, A. Munir A. Murad2, Fredj Tekaia3, Driss Talibi4, Hélène Tournu2, Daniel Maréchal4, Claude Gaillardin5, and Alistair J.P. Brown2 . 1Unit de Physiologie Cellulaire and 3Unité de Génétique Moléculaire des Levures, Institut Pasteur, 75724 Paris Cedex 15, France; 2Molecular and Cell Biology, Institute of Medical Sciences, University of Aberdeen, Aberdeen AB25 2ZD, United Kingdom; 4Eurogentec, Parc Scientifique du Sart Tilman, 4102 Seraing, Belgium; 5Laboratoire de Génétique Moléculaire et Cellulaire, CNRS URA1925 INRA UMR216 INA-PG, 78850 Thiverval-Grignon, France

Saccharomyces cerevisiae Tup1 is a general transcriptional repressor which is targeted to specific promoters through an interactions with the specific DNA-binding proteins, including ScNrg1 and ScMig1. Inactivation of the homologue of ScTup1 in the human pathogen Candida albicans results in a constitutive filamentous growth which is an important determinant of C. albicans virulence (Braun and Johnson, Science 277:105, 1997). Inactivation of CaNrg1 also causes constitutive filamentous and invasive growth in C. albicans while a filamentous phenotype has not been associated to the inactivation of CaMig1 (Zaragoza et al. J. Bacteriol. 182:320, 2000). In this study, we have used transcript profiling to dissect the roles of CaTup1, CaNrg1 and CaMig1. Using the public data of the C. albicans genome sequencing program (www-sequence.stanford.edu/group/candida/index.html), 2002 open reading frames (ORFs) were identified and PCR products corresponding to their 3' end were spotted on nylon membranes. Hybridization profiles were obtained using RNAs prepared from C. albicans wild-type, tup1/tup1, nrg1nrg1 and mig1/mig1 strains and compared. Our results show that CaNrg1 and CaMig1 co-regulate separate but overlapping subsets of CaTup1-regulated genes, consistent with CaNrg1 and CaMig1 being CaTup1 targetting proteins. In particular, CaNrg1 regulates known hypha-specific genes and other virulence factors while CaMig1 regulates many metabolic functions. Genes with unknown functions identified as targets of CaNrg1 or CaMig1 are likely to participate in these same processes. In this regard, transcript profiling represents a powerful approach to identify genes that are potentially relevant to the virulence of C. albicans.

132 Annotation and gene expression analysis of BAC clone 6J18 on chromosome 7 of Magnaporthe grisea. Woobong Choi1, Youngjin Koh2, and Ralph A. Dean1. 1Fungal Genomics Lab. North Carolina State University, Raleigh, NC, USA. 2Sunchon National University, Sunchon, Korea.

A 112 kb BAC clone located on chromosome 7 of the rice blast fungus, Magnaporthe grisea, was sequenced and annotated. Sequencing was accomplished in two phases. In the first or production phase, the BAC clone was fragmented into 1-2 kb fragments and cloned into pUC18. These clones were sequenced at random using M13 forward primer to produce ~8-fold coverage. In the second or finishing phase, the other ends of selected clones were sequenced using M13 reverse primer. Gene prediction annotation software, GenScan, identified 36 putative ORFs. 21 of these showed significant matches to known proteins as revealed by a BlastP search. Analysis of ESTs for M. grisea chromosome 7 identified 13 genes that mapped onto the BAC clone. Genomic Southern blot analysis revealed that all 13 genes existed in single/low copy number in the rice blast genome. Northern blot analysis showed that several genes are differentially expressed during appressorium formation. Sequence alignment of the BAC against the rice blast BAC-ends sequence (STC) database identified 70 neighboring BACs, the majority of which were found in the same fingerprint contig as assembled by the software program FPC. These results demonstrate that the sequencing BAC clones when combined with expressed sequence tags (ESTs) and BAC end sequence (STCs) analysis is a powerful and practical approach for elucidating the organization and function of the rice blast genome.

133 Gene expression during growth and microsclerotia development in the vascular wilt pathogen Verticillium dahliae. M.J. Neumann and K.F. Dobinson. Agriculture and Agri-Food Canada, 1391 Sandford St., London, ON Canada N5V 4T3, and Department of Plant Sciences, University of Western Ontario, London, ON Canada N6A 5B7

Vascular colonization by Verticillium dahliae is characterized by both filamentous and budding (yeast-like) growth. This dimorphic growth pattern, characteristic of vascular wilt fungi, is thought to be essential for the systemic colonization of xylem vessels. In the late stages of disease the pathogen enters a limited saprophytic growth phase, obtaining nutrients from the dying plant, and producing microsclerotia. cDNA libraries were constructed from cultures of V. dahliae grown either in a simulated xylem fluid medium (SXM), or under conditions that favour near-synchronous development of microsclerotia (MS). Up to 45% of the expressed sequence tags (ESTs) from approximately 1000 randomly chosen clones from each library had weak or no similarity to protein sequences in the NCBI databases. ESTs with significant scoring pairs (BlastX scores >50; E values <10-7) have been assigned to functional groups. Homologues of melanin biosynthetic genes are present in the MS library, and not in the SXM library. A number of ESTs from the SXM library had high similarity to genes from other phytopathogenic fungi that are expressed preferentially in planta, and/or under nutrient conditions that favour pathogenic growth. These data suggest that the two libraries are representative of different phases within the life cycle of V. dahliae, and will be useful for further study of microsclerotial development and pathogenic growth. A detailed comparison of gene expression during the two growth phases will be presented, and the significance of the data discussed.

134 Transposon impala from Fusarium oxysporum: a novel tool for tagging genes and genome analysis in ascomycetes. M. J. Daboussi1, S. Demais1, A. Hua-Van1, C. d'Enfert1, A. Firon1 , M. C. Grosjean-Cournoyer2, T. Langin1, M. H. Lebrun2, M. G. Li Destri Nicosia1, C. Masson1, C. Scazzocchio1, and F. Villalba2. 1IGM and IBP, Universite Paris-Sud, Orsay, France. 2+Biotechnologies and UMR CNRS-Aventis, Aventis CropScience, Lyon, France; Institut Pasteur, Paris, France.

impala is an active transposable element of Fusarium oxysporum. The imp160 copy, integrated into the niaD gene encoding nitrate reductase, behaves as an autonomous element since it excises from niaD and reinserts at a new genomic location in fungal isolates lacking active elements. A defective impala element with the transposase gene replaced by the hygromycin resistance gene was used to demonstrate the absence of endogenous transposase in impala-free isolates and the ability of different genomic copies of impala to promote transposition in trans. Autonomous and engineered impala copies were introduced in different fungal species such as F. moniliforme, an endophytic fungus, Aspergillus nidulans, a model ascomycetal fungus, A. fumigatus, a human pathogenic fungus, and Magnaporthe grisea, a plant pathogenic fungus. In all these species, impala transposes with a good efficiency in a manner similar to that observed in F. oxysporum. Analysis of a collection of strains with transposed impala elements shows that it can inactivate and tag genes including pathogenicity genes. The ability of impala to function in a two-component system and to transpose in different fungal species demonstrates its potential as a transposon tagging tool in fungi. Furthemore, analysis of the behavior of impala in different species or genetic backgrounds should give significant insights on the factors that modulate transposition efficiency.

135 Genome wide analysis of Cryptococcus neoformans. Steen, B.R., Tangen, K., MacDonald, K., Lian, T.S., *Marra, M., *Jones, S. Kronstad, J.. Biotechnology Lab, University of British Columbia. *Genome Sequence Centre, B.C. Cancer Agency.

Cryptococcus neoformans is an important opportunistic human pathogen. The development of new strategies to control cryptococcal infections will be greatly facilitated by a more detailed understanding of the genome of C. neoformans. We are taking a genome wide approach to discover and characterize C. neoformans genes involved in signaling and virulence in the context of adaptation to growth conditions in the host environment. For example, we are using serial analysis of gene expression (SAGE) to study the ability of C. neoformans strains H99 (serotype A) and JEC21 (serotype D) to adapt to growth at 37oC. Four SAGE libraries were prepared representing each strain and each temperature. Because of the availability of the genome sequence (~3X shotgun coverage), we have initially focused our analysis on the two SAGE temperature libraries for strain JEC21. We have obtained 2,935 and 1,159 sequence reads for the JEC21 25oC and 37oC libraries, respectively and 1,561 and 1,948 sequence reads for the H99 25oC and 37oC libraries, respectively. From these, we have derived a total of 68,284 tags, 13,836 of which are unique, for the JEC21 25oC library, 19,064 total tags, 7,310 of which are unique, for the JEC 21 37oC library, 42,358 total tags, 10,179 of which are unique for the H99 25oC library and 61,976 total tags, 11,298 of which are unique for the H99 37oC library. These libraries are being compared to identify which genes are more highly expressed at infective temperatures (37oC) compared to noninfective temperatures (25oC). In addition, we are constructing physical maps of Cryptococcus in order to compare the serotype A strain with the serotype D strain. Digestion of 3000 BAC clones (HindIII digested) produced fingerprints which resulted in 27 contigs for H99 and 42 contigs for JEC21. Overgo hybridization for JEC21 and H99 specific sequences is being done in order to sew together the existing contigs and for comparison of the physical maps for the two strains. As well, these high resolution maps will aid in assembling the Stanford genome sequence for strain JEC21. These studies should contribute to a global understanding of this fungal pathogen and aid in identifying genes important in virulence.

136 Adaptation of a PCR-based cloning method for use in filamentous fungi. Michael Nielsen, Uffe H Mortensen, and Alexei Aleksenko. Technical University of Denmark, BioCentrum, Kgs. Lyngby, Denmark

The rapid accumulation of genomic sequences from a broad range of organisms fuels the need for specific genome manipulation. A cloning-free PCR based method for allele replacement exists where PCR products are fused together to produce two tailored DNA fragments suitable for co-transformation and subsequent integration into the genome of a given host. The fusion is accomplished by the use of adaptamers, which are PCR primers with overhangs that differentially tag the 5' and 3' of the amplified substrate. Complimentary adaptamers of two denatured PCR fragments anneal, thus fusing the products prior to PCR amplification of the whole fragment. Using this technique one fragment is generated that contains sequences matching the desired site of integration fused to the 5' 2/3 of a selectable gene and another where the integrative sequences are fused to the 3' 2/3 of the same selectable gene. After transformation, homologous recombination in the cell fuses the two fragments to reconstruct the entire selectable marker and inserts the fragments into the genome at the desired site. If a counter selectable marker is used, it can be excised from the genome by a direct repeat recombination event, leaving only the desired genomic alteration. So far this technique only been used in Saccharomyces cerevisiae because its genomic sequence is known, but as more genomic sequences become available the approach may become applicable to other organisms. To demonstrate this, we have adapted the method for use in the filamentous fungus, Aspergillus nidulans, by using it to replace the wild type yA allele with a mutant allele.

137 Agrobacterium tumefaciens-mediated transformation of Magnaporthe grisea. Hee-Sool Rho, Seogchan Kang1, and Yong-Hwan Lee. School of Agricultural Biotechnology, Seoul National University, Suwon 441-744, Korea. 1Department of Plant Pathology, The Pensylvania State University, University Park, PA 16802, USA

Agrobacterium tumefaciens-mediated transformation (ATMT) has long been used to transfer genes to a wide variety of plants and has also served as an efficient tool for insertional mutagenesis. ATMT was established in the rice blast fungus, Magnaporthe grisea, to identify genes important for pathogenicity. Employing two binary vectors, carrying the bacterial hygromycin B phosphotransferase under the control of the Aspergillus trpC promoter as a selectable marker, led to the production of about 300 hygromycin B- resistant transformants per 1 x 107 conidia of M. grisea. Transformation efficiency correlated with the number of conidia used and the duration of co-cultivation with Agrobacterium cells. All transformants tested remained mitotically stable, maintaining their hygromycin B resistance after several generations of growth in the absence of hygromycin B. Genomic Southern blot analysis showed that over 60% of the transformants contained a single T-DNA insert per genome. Considering the efficiency and flexibility of this ATMT protocol, ATMT appears to be a highly efficient alternative to other insertional mutagenesis techniques in characterizing those genes important for the pathogenicity of M. grisea and potentially that of other fungal pathogens.

138 Identification of virulence and morphogenetic factors in Histoplasma capsulatum using a shotgun DNA microarray. Lena Hwang and Jasper Rine. University of CA, Berkeley, Molecular Cell Biology, Berkeley, CA. USA

Histoplasma capsulatum is a dimorphic, pathogenic fungus that infects humans, causing pulmonary and systemic disease. It exists in mycelial form in the soil, and when aerosolized, the conidia are inhaled by the host. At 37C, within the host, H. capsulatum undergoes a phase transition from mycelia to yeast and is phagocytosed by macrophages. This organism survives within the phagolysosome and is capable of inhibiting the acidification of that compartment. In addition, H. capsulatum is also able to modulate the pH of its media to near neutral when grown in culture. In collaboration with Anita Sil at U.C. San Francisco, we have created a 9600-element H. capsulatum DNA microarray. The microarray contains random genomic fragments, representing approximately 1/3 of the genome. We have used the array to identify genes that are specifically expressed in either the yeast or mycelial form. The microarray is also being used to identify genes that respond to changes in the pH of the media. Genes that display a pH-dependent transcriptional response will be further characterized for their role in the modulation of pH in the phagolysosome of macrophages, as well as their role in virulence.

139 Linkage disequilibrium is associated with an inversion in the het-6 region of Neurospora crassa. Cristina O. Micali, N. Mirrashed and M. L. Smith. Carleton University, Ottawa, ON. Canada

In N. crassa, heterokaryon incompatibility occurs upon fusion of individuals that differ any one of 11 het loci or the mating-type locus. Among these loci, allelic differences at het-6 cause one of the most dramatic incompatibility reactions, resulting in death of heterokaryotic cells. The het-6 region was previously shown to contain two distinct incompatibility genes, un-24 (ribonucleotide reductase large subunit) and het-6. Two alleles, Oakridge (OR) and Panama (PA), have been identified at each locus. The two genes are in severe linkage disequilibrium; only OR/OR or PA/PA combinations were detected in a survey of more than 150 strains from nature. This suggests the un-24-het-6 region acts as a het-gene complex. Here we compare the structure of the PA and OR forms of this region. PA strains carry a ~20 kbp inversion with respect to OR strains. PCR/RFLP markers provide evidence for recombination within and around the inverted region. DNA sequence was determined across ~12 kbp. Sequences at the inversion breakpoints are divergent in OR compared to PA strains. We hypothesize that these physical differences reduce recombination events that would give OR/PA or PA/OR combinations at un-24 and het-6, respectively, and that the inversion polymorphism is maintained by association with the het-gene complex.

140 Genome-scale gene identification, disruption and function assignment in filamentous fungi. Lisbeth Hamer, Kiichi Adachi, Maria V. Montenegro-Chamorro, Matthew Tanzer, Sanjoy Mahanty, Clive Lo, Rex W. Tarpey, Amy Skalchunes, Ryan Heiniger, Sheryl Frank, Blaise Darveaux, Ted Slater, Lakshman Ramamurthy, Todd Dezwaan, Grant Nelson, Jeffery Shuster & Jeff Woessner. Paradigm Genetics Inc., 104 Alexander Dr., Bldg. 2, RTP, NC 27709 USA

The filamentous fungi are a large group of diverse and economically important microorganisms. Because of their complex genomes and a low rate of targeted integration during transformation, high throughput gene disruption and functional analyses are not readily applicable to these organisms. We have developed TAGKOTM, a genome scale functional analysis approach to rapidly identify, disrupt and determine the function of genes without any prior sequence information. We show recent results using this technology in the pathogenic fungi Magnaporthe grisea and Mycosphaerella graminicola, the cause of rice blast and wheat blotch, respectively.

141 Genomic efforts to study the plant pathogen Fusarium graminearum. Linda Harris, Hélène Rocheleau, Sharon Allard, Tricia Glassco, Anju Koul, Thérèse Ouellet. Eastern Cereal & Oilseed Research Centre, Agriculture & Agri-Food Canada, Ottawa, Ontario, Canada K1A 0C6

Fusarium graminearum (teleomorph: Gibberella zeae) is a broad host pathogen, attacking a range of plant species including wheat and barley (head blight/scab), and maize (gibberella ear and stalk rot). It has been the cause of several recent epidemics causing millions of dollars in lost revenue to Canadian and U.S. cereal producers and has been identified as the most important cereal problem in Canada in terms of yield and grade losses and toxicity of infected grains. We are using a genomics approach to identify and characterize genes from F. graminearum which may be involved in the host-pathogen interaction and in fungal pathogenicity. Our goal is to build a large collection of ESTs which will be used to develop micro arrays. Several cDNA libraries have been constructed from F. graminearum at various developmental stages and under different growth conditions. EST sequencing from those libraries has been initiated. Our progress on this project will be presented.

142 MycoPath™: a new resource integrating protein-specific information for the major fungal pathogens of humans. Maria C. Costanzo, Laura S. Robertson, Glenn D. Krumholz, Janice E. Kranz, Pinar Kondu, Kevin J. Roberg-Perez, and James I. Garrels. Proteome, Inc., Beverly, MA, USA.

Information relevant to specific proteins of any particular fungal pathogen is found not only in the biological literature about that pathogen, but also in the literature concerning similar proteins of other fungi. Extracting and integrating information from all of these sources can be a formidable task for an individual researcher. To address this need, Proteome, Inc. maintains a growing database, the BioKnowledgeÖÖ Library, which contains extensive and up-to-date gene- and protein-specific information collected and organized by expert curators and editors. Four volumes of the Library (http://www.proteome.com/databases) are concerned with fungi and model organisms: YPD, PombePD, WormPD, and CalPD, covering the proteomes of S. cerevisiae, S. pombe, C. elegans, and C. albicans, respectively. The next volume to be released will be MycoPathPD, which will incorporate CalPD and add information about genes and proteins of several important human pathogens: Aspergillus species (fumigatus, flavus, and niger;) B. dermatitidis; several Candida species in addition to albicans; C. immitis; C. neoformans; H. capsulatum; and P. carinii. While complete genomic sequences are not yet available for any of these organisms, compiling all available sequence and literature information into one resource with the C. albicans sequences, in a format that facilitates comparisons with the S. pombe and S. cerevisiae proteomes, yields a valuable resource for antifungal research. Furthermore, the availability of this gene-indexed, searchable, and interlinked collection of fungal protein-specific information will greatly facilitate prediction of the functions of proteins encoded by newly sequenced fungal genomes.

143 High throughput transformation and gene disruption of the phytopathogenic fungus Mycosphaerella graminicola. Kiichi Adachi, Grant Nelson, Rex W. Tarpey, Maria V. Montenegro-Chamorro, Sheri Frank, Todd Dezwaan, Amy Skalchunes, Ryan Heiniger, Blaise Dareaux, Sanjoy Mahanty, Matthew Tanzer, Lisbeth Hamer, Lakshuman Ramamurthy and Jeffery Shuster. Paradigm Genetics, Inc., RTP, NC, USA

The Ascomycete fungus Mycosphaerella graminicola is a causal agent of Septoria leaf blotch of wheat. Despite intensive research, gene manipulation techniques such as DNA-mediated transformation and gene disruption are not well established for this fungus. Here, we present a highly efficient electro -transformation system and gene disruption data for M. graminicola. The electroporation conditions were optimized using a hygromycin B-resistant marker and transformation efficiency increased up to 100 transformants/1x107 protoplasts/ g DNA. Using this method we successfully disrupted LEUC (3-isopropylmalate dehydrogenase) and ATR2 (ABC transporter) genes for pilot experiments. Targeted integration frequency was 4.3% for LEUC with 3.1 kb total flanking length, and 5.3% for ATR2 with 6.7 kb total flanking length. To improve targeted integration frequency we invented TAGKOTM (transposon-arrayed gene knockout) technology to generate gene disruption vectors, which always contain long flanking regions of homologous DNA (~40 kb). High throughput gene disruption experiments of M. graminicola were initiated using TAGKOTM constructs. TAGKOTM technology increased targeted integration frequency up to 28%.

144 Searching for coding regions in Neurospora crassa using a simple codon bias algorithm and consensus sequences. Judith Galbraith1,2, Dr. Don Natvig1, Dr. Mary Anne Nelson1, Dr. Laura Salter1, Cara Slutter1,2. 1University of New Mexico, 2.Albuquerque High Performance Computing Center

Current Gene Finding Algorithms depend largely on elevated GC content in coding regions, and hexanucleotide or dicodon counts. Many of these tools are tailored for model organisms such as humans or yeast, and perform poorly with other species such as the filamentous fungus Neurospora crassa.

To locate coding regions in sequences that have no similarity to known genes in the public databases, characteristics distinctive of Neurospora have been examined. One such characteristic is the exaggerated difference between the counts of cytosine(C) and adenosine(A) residues in the third position of codons. This difference is measured using a log ratio, abbreviated log (C/A).

Since Neurospora has many short exons at the 5 prime ends of specific genes, other quantitative measures must also be employed for full gene prediction. The Kozak consensus sequence which surrounds the start codon and splice site consensus sequences flanking introns, with frequencies specific to Neurospora, are evaluated for each potential exon. The current algorithm does an exhaustive search for all possible exons using either a Kozak pattern or a 3 prime splice site as a beginning of coding and a 5 prime splice site or STOP codon as the end. Scoring has been done on all conserved patterns and on the content or codon bias of each putative exon. Logistic regression has been performed to test the significant effect of each descriptive variable on the binary response variable (coding or non-coding). A linear relationship is derived from this analysis and a probability value assigned for each putative exon. Then testing has been done for sensitivity and specificity. A web based tool has been designed and is available at: http://www.ahpcc.unm.edu/Research/CompBio/GeneFinder/NEX/

145 Molecular evolution and comparative function characterization of the septin gene family. Jiong Zhao and Michelle Momany. Department of Botany, University of Georgia, Athens, Georgia 30602

The septins are members of a highly conserved protein family that was first identified in budding yeast as proteins associated with the neck rings. Members of the septin gene family have been found in budding and fission yeast, filamentous fungi, fruit fly, worm, mouse, and man. Despite the differences in protein sequences, most of septins seem to be essential in both fungal and animal cells. Septins were initially thought to play a role in controlling cytokinesis, as evidenced by their concentration at the mother-bud neck in budding yeast and at the cleavage furrow in dividing animal cells. However, recent data suggest that they may function in a much wider array of contexts such as cell surface organization and vesicle fusion processes. In this study we identified and compared 55 septin genes from fungi and metazoa. Our phylogenetic analysis indicates that most septins can be distributed into five orthologous classes, which include two fungal-specific classes, one animal- specific class and two mixed fungal-animal classes. Based on intron-exon comparison, most fungal septins do not show paralogs in the same species. However, most of animal septins have paralogs. Based on our phylogenetic analysis, we suggest function for some uncharacterized septins.

146 Variants among Neurospora: an analysis of alternative splicing in Neurospora crassa. Patricia L. Dolan, Kathryn J. Gruchalla, Diego Martinez, Gary Montry, Donald O. Natvig,Mary Anne Nelson. University of New Mexico

Alternative splicing of precursor mRNAs is an important mechanism for regulating gene expression and generating protein diversity. It allows the selection of different splice sites to produce different transcripts and sometimes variant proteins from a single gene. Alternative pre-mRNA splicing is a widespread phenomenon, occurring in organisms as diverse as the nematode Caenorhabditis elegans, the fruitfly Drosophila melanogaster,and humans. Evidence of alternative splicing has been found for about one-third of human genes, with the majority of alternative forms occurring from splicing events in the 5ÆÆ untranslated regions. A study to estimate the extent of alternative splicing in Neurospora crassapre-mRNA has been undertaken. Using TIGR Assembler and Phrap EST assemblies, and where possible subsequent matching of ESTs to genomic sequences, alternative splicing variants have been identified for a number of genes. The issue of alternative splicing versus inefficient splicing (or partially processed nuclear intermediates) is discussed. With continued accumulation of Neurospora EST and genomic data, further insights into the patterns of alternative splicing will be possible. Identifying those proteins with variant forms will allow a greater and more complete understanding of the expression of Neurospora's proteome.

147 Simultaneous mapping of multiple vic loci in Gibberella fujikuroi MP-A. K. A. Zeller1, J. E. Jurgenson2, and J. F. Leslie1. 1Dept. of Plant Pathology, Kansas State University, Manhattan, KS; 2Dept. of Biology, University of Northern Iowa, Cedar Falls, IA.

We have used AFLP markers to map vic (vegetative incompatibility) loci that segregate in the mapping population of Gibberella fujikuroi MP-A [Fusarium verticillioides (=F. moniliforme)]. We crossed two nit1- strains from the mapping population's parents genetic backgrounds, and selected for recombinant progeny with vegetative compatibility (VC) to a strain of known AFLP genotype. This selection for VC should simultaneously select for identity in all segregating vic loci, and will skew segregation ratios for markers linked to each locus. We have used AFLP fingerprints from 104 progeny to define QTL-like associations between markers and VC phenotype. We have examined segregation of > 240 mapped polymorphic AFLP markers generated with 22 selective primer combinations. From these data we have identified 9 unlinked genomic regions with segregation ratios that deviate significantly (alpha < 0.05) from 1:1. In at least two of these regions we have identified AFLP markers that map within 2-3 map units of putative vic loci. These data will allow us to rapidly isolate and characterize vic loci from G. fujikuroi MP-A, and will give us a model for understanding VC interactions among other, less genetically tractable, members of Gibberella and Fusarium.

148 Microarray analysis of frequency - regulated gene expression in Neurospora crassa. Zachary A. Lewis1, Alejandro Correa1, Xie Xin2, Daniel J. Ebbole2 and Deborah Bell-Pedersen1. 1Dept of Biology, Texas A& M University. 2Dept of Plant Pathology, Texas A&M Univeristy

The filamentous fungus Neurospora crassa is a well-established model system for the study of circadian rhythms. The clock gene frequency (frq) was identified through classical genetic studies and has since been shown to be an important component of the pacemaker machinery. frqmRNA and FRQ protein levels oscillate over the course of the day, and FRQ is believed to either directly or indirectly control rhythmic processes in the fungus. Consistent with this possibility, known clock-controlled genes have been shown to be transcriptionally depressed or elevated in frq-null strains. We are using a transcriptional profiling approach to identify candidate genes that are positively or negatively regulated by FRQ protein. A total of 1800 clones from the University of New Mexico EST library have been arrayed on glass slides. These arrays have been probed with cDNA made from RNA isolated from frq+ and frq 10 (frq-null) cultures harvested at the time of peak FRQ levels. Currently, we are optimizing hybridization conditions. In addition,we are performing Northern analysis to confirm the results obtained from array hybridization. This approach should identify components of the FRQ regulatory pathway along with other pacemaker components. Candidate genes will be further characterized by Northern analysis, gene inactivation, and analysis of mutant phenotypes.

149 Systematic identification of essential genes in the human pathogenic fungus Aspergillus fumigatus. A. Firon1, M.-C. Grosjean-Cournoyer2, A. Beauvais3 and C.d'Enfert1 . 1Unit de Physiologie Cellulaire, Institut Pasteur, Paris, France,2 Biotechnology Department, Aventis Crop Science, Lyon, France and 3Unit des Aspergillus, Institut Pasteur, France.

Invasive aspergillosis has become the most frequent air-borne fungal infection in patients with a deficient immune system. Because of a difficult diagnosis and the lack of efficient antifungal treatments, it is associated with a mortality rate as high as 85%. Invasive aspergillosis is mostly due to Aspergillus fumigatus, an opportunist pathogen. Despite intensive efforts, no genuine virulence factors have been identified in this species suggesting that virulence is a multifactorial process. Another way to identify novel antifungal targets is to define A. fumigatus genes that are essential for fungal growth . Here, we have developed a strategy combining insertional mutagenesis and parasexual genetics for the identification of essential genes in A. fumigatus. Although A. fumigatus is haploid and devoid of a sexual cycle, we were able to generate heterokarions using haploid pyrG strains with complementary auxotrophic and spore-colour markers and to identify stable diploid strains. Disruption of one allele of the AfFKS gene encoding 1,3-beta-D-glucane synthase in this strain resulted in a heterozygous diploid which did not produce haploid progenies upon benomyl-induced haploidization when the selective pressure corresponding to the AfFKS disruption was maintained. This result demonstrated that beta 1,3 glucan synthesis is an essential process for growth in A. fumigatus. Furthermore, it suggested that the identification of heterozygous diploids unable to produce haploid progenies could be used to define essential genes in A. fumigatus. A collection of heterozygous diploids was obtained by insertional mutagenesis of the diploid strain with a A. nidulans pyrG-bearing plasmid using electroporation of intact conidia. 8% of these heterozygous diploids were unable to produce haploid progenies when the selection for the pyrG gene was maintained during haploidization. When haploid progenies were obtained in the absence of a selective pressure, none had a pyrG+ phenotype, suggesting that in all cases tested insertion had occured in an gene essential for A. fumigatus growth. Molecular analysis, using a semi-random PCR strategy, of heterozygous diploids containing a disrupted essential gene will be presented. Our data show the potential of using heterozygous diploids to demonstrate the essentiality cloned A. fumigatus genes or to identify on a random basis genes that are essential for A. fumigatus growth.

150 Phospholipid induced extracellular protein analysis in A. fumigatus. Michael J Bromley, Mike Birch, Jayne L Brookman. School of Biological Sciences, University of Manchester, England.

Aspergillus fumigatus is the most common mould infection of humans world-wide with disease incidence increasing particularly amongst immuno-compromised individuals. Current treatments for A. fumigatus infection are limited and resistant strains have been identified.

Functional analysis of A. fumigatus currently relies on genetic techniques that have been adapted from those successfully employed in prokaryotes and yeasts. Unfortunately it has become clear that these technologies do not provide the high throughput analysis necessary for an economically viable search for new drug targets. Proteome analysis using two-dimensional gel electrophoesis may provide the answer.

Comparative proteome analysis has been performed on A. fumigatus samples grown in the presence/absence of a phospholipid which is the primary constituent of lung tissue at the initial site of infection in invasive pulmonary aspergillosis. We have identified several proteins that are unique to the secreted complement of A. fumigatus proteins when grown in the presence of the phospholipid. Detailed analyses of these proteins will be presented.

151 Analyses of ESTs and promoters of useful expression from Aspergillus oryzae. Hiroko Hagiwara, Motoaki Sano, Sumiko Kunihiro, Kumiko Takase, Midori Yamamoto, Masayuki Machida. Natl. Inst. Biosci. & Human-Technol., Tsukuba, Ibaraki, Japan

Aspergillus oryzae is an important filamentous fungus in Japanese beverage and fermentation industries, and for the production of industrially valuable enzymes. To accumulate the basic knowledge of nucleotide sequences and expression of individual genes, we randomly sequenced cDNA libraries, yielding 5' expressed sequence tags (ESTs). The entire project was done by collaboration of Natl. Res. Inst. of Brewing, Natl. Food Res. Inst., Tohoku Univ., Tokyo Univ. Agric. Technol., Nagoya Univ. and Univ. Tokyo, using libraries from several different culture conditions including solid state culture, alkaline condition and at germination. The total number of the analyzed ESTs and contigs after clustering were approximately 17000 and 6000, respectively. We estimated that partial sequences of roughly a half of the total A. oryzae genes, which are estimated to be 8000-9000, had been sequenced. The total lengths of the contigs reached 4.7 Mb, which was equivalent to approximately 16% of the A. oryzae genome. In the present work, we focused on ESTs from liquid culture with and without carbon source consisting of 2478 and 1790 ESTs, respectively. These ESTs are expected to be useful as references to ESTs from other culture conditions. Database search by the BLAST algorithm showed that approximately 40% of the contigs had no similarity with any genes or proteins registered in the public databases to date. We have started sequencing promoters of strong expression and induction, which are important for protein production, and promoters of industrially valuable genes. Functional analyses of the promoters in bioinformatics and biological approaches will facilitate the understanding the genetic network in A. oryzae, which should enable a flexible control of this organism in the industrial application. A part of the above data is available from our web site (http:///www.aist.go.jp/RIODB/ffdb/index.html).

152 Chromosomal polymorphisms, gene mapping and ploidy of Paracoccidioides brasiliensis. Luciano dos Santos Feitosa1, Màrcia R. Machado dos Santos1, Renato A. Mortara1, Jos Franco da Silveira1, Patrícia S. Cisalpino2, Zoilo Pires de Camargo1.1Departamento de Microbiologia, Imunologia e Parasitologia, Universidade Federal de São Paulo, São Paulo, Brasil. 2Departamento de Microbiologia, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brasil

Paracoccidioides brasiliensis (anamorph; teleomorphic stage unknown), a thermo dimorphic fungus, is the ethiological agent of paracoccidiodomycosis (PCM), a human granulomatous disease prevalent in Latin America. Genetic composition and genomic organization of P. brasiliensis is poorly understood. We worked on the electrophoretic karyotype (PFGE, as in Cano et al., 1998) of twelve fungal isolates from different geographic areas obtained from patients with chronic and acute PCM, armadillo and soil. Stable and reproducible karyotypes were observed. Our results were consistent with a haploid number of 4 - 5 chromosomal bands in the range of 2,5 to 10,0 Mpb. The haploid genome was estimated to be 24-30 Mbp. Six distinct karyotypic profiles were observed. Chromosomal polymorphisms made it difficult to correlate the banding pattern among isolates. In order to determine chromosome identity, nine specific gene probes have been used to hybridize Southern blots containing intact chromosomal bands. One genetic linkage group and eight distinct profiles were observed. There was a significant regularity when matching genes to chromosomes within certain size range, suggesting that the overall structure of the genome could be quite constant. We also estimated the DNA content of DAPI-stained nuclei of the twelve isolates by confocal microscopy. Estimation of the genome sizes by PFGE and microfluorimetry indicated the possible existence of both haploid and diploid fungal isolates. Supported by: PRONEX, FAPEMIG, FAPESP.

153 Transcript profiling during ectomycorrhiza development. Tomas Johansson, Antoine Le Quéré, Dag Ahrén, Joakim Lundeberg, Rikard Erlandsson, Bengt Söderström, Anders Tunlid. Lund University, Microbial Ecology Lund, Sweden. and Royal Institute of Techn., Biochemistry & Biotechn., Stockholm Sweden

Mycorrhiza, a symbiotic interaction between plant and fungi, is found on a vast majority of all plant species including many important forest trees and crop plants. The symbiotic interaction, which involves the development of specific infection structures in the plant rhizosphere, improve the growth of the host plant by facilitating nutrient uptake from the soil, while the fungus gains carbohydrates from the plant. The developmental process is anticipated to depend on a well-tuned regulation at both temporal and spatial levels, and the objective of this research is to isolate and characterize symbiosis-related (SR) genes by analyses of gene expression during the ectomycorrhizal development between the fungus Paxillus involutus and the plant Betula pendula (birch). Several cDNA libraries have been constructed from symbiotic ectomycorrhizal tissue, from free-living hyphae of P. involutus and from non-infected roots of B. pendula, respectively. So far we have obtained 5 -end EST sequence information from approximately 10,000 EST clones. The annotation and assignment of putative cellular function(s) is still in progress, but score frequencies from database comparisons yields significant differences between libraries. It is obvious that the transcript profile of the mycorrhiza is significantly different from that of the fungus and the plant, and clones are here observed to be of both fungal and plant origin. Among the high-score transcripts there are a number of fungal and plant genes (ôôcandidatesöö) that has been suggested to be involved in the development and function of the mycorrhizal symbiosis or in the interaction between fungal pathogens and plants.

154 Smart man genomics: Expression profiling of organisms with unsequenced genomes. Ted van der Lende, Cees van den Hondel and Frank Schuren. Department of Applied Microbiology and Gene Technology, TNO Nutrition and Food Research, P.O. Box 360, 3700 AJ Zeist, The Netherlands.

In our view the basis of genomics is the holistic, non-targeted study of in principle all biomolecules in an organism. This can be at the level of RNA (transcriptomics), proteins (proteomics) or metabolites (metabolomics). In our lab we have fully implemented the infra-structure for expression profiling (transcriptomics) of organisms of which the entire nucleotide sequence is available. In addition, we developed the ôôsmart man genomicsöö concept, a strategy which circumvents full genome sequencing but nevertheless enables transcriptome analysis of in principle every microorganism. We show that, in spite of the dogma that one needs a full genome sequence of an organism to study transcription profiles, expression profiles of organisms with unsequenced genomes is feasible. An explanation of this strategy and some preliminary data will be presented.

155 Microbial Metabolomics. Mariët J. van der Werf* and Cees A.M.J.J. van den Hondel. Department of Applied Microbiology and Gene Technology, TNO Nutrition and Food Research, P.O. Box 360, 3700 AJ Zeist, The Netherlands.

Applied genomics technologies (transcriptomics, proteomics, and metabolomics), aimed at analyzing the keystone biomolecules of a cell in its entirety, have ushered a new era in scientific methodology in the life sciences. Metabolomics is the most recent addition to the applied genomics toolbox. In analogy with transcriptomics and proteomics, metabolomics involves the non-targeted, holistic determination of changes in the complete set of metabolites (low molecular weight molecules) in the cell (the metabolome) in response to environmental changes. Subsequently, knowledge is gained by analyzing different metabolomes using bioinformatic tools (pattern recognition) that either identify metabolites relevant for a specific environmental condition or interpret the quality/status of the complete metabolome (cluster analysis). Moreover, the information obtained from a metabolomics experiment gives information as to which metabolic pathways are used by the organism or even that are functioning in a specific compartment. This is of special importance when working with (metabolically) poorly characterized organisms. Metabolomics is expected to become of crucial importance for metabolic engineering, as it can be applied to determine bottle necks in biosynthesis. Due to the intrinsic characteristics of the metabolomics technology (measured metabolite concentrations, and its holistic and comparative approach), it is expected to replace metabolic flux analysis as the primary tool for deciding were to change a metabolic network in order to improve productivity. Moreover, metabolomics can be applied for growth medium optimization for increasing microbial growth rates or productivity, to determine the key molecules that induce (desired or undesired) secondary metabolite or enzyme production, to predict the effect of individual medium components on the overall flavour profile, to establish regulatory networks, and to predict the quality of the end-product form the quality of the microbial starter culture in food fermentations.

156 Sequencing of the centromeric region of chromosome V in Podospora anserina : a draft project for the complete sequence of the genome. Christian Barreau1, Alain Billault2, Laurence Cattolico3, Robert Debuchy4, Simone Duprat3, Sebastien Kicka4, Annie Sainsard-Chanet5, Carole H. Sellem5, Philippe Silar4 and Béatrice Turcq1. 1Institut de Biochimie et Génétique Cellulaires, UPR CNRS 9026, 1 rue Camille Saint-Saens, F-33077 Bordeaux, France. 2Centre d'Etude du Polymorphisme Humain, 75010 Paris, France. Present address: Molecular Engines Laboratories, 20 rue Bouvier, F-75011 Paris, France.3Génoscope, Centre Nationale de Séquenéage, 2 rue Gaston Cr mieux, CP 5706, 91057 Evry Cedex, France. 4Institut de Génétique et Microbiologie, UMR 8621, bâtiment 400, Université Paris-Sud, F-91405 Orsay cedex, France. 5Centre de Génétique Moléculaire, UPR CNRS 9061, bâtiment 24, Allee de la terrasse, F-91190 Gif sur Yvette, France .

Podospora anserina is used as a model system for investigations about translation, control of sexual development, stability of mitochondrial genome, mitochondrial metabolism, non-conventionnal infectious elements or vegetative incompatibility. Each of these fundamental processes is controled by a large array of genes. The analysis of these processes will beneficiate from the access to the complete genomic sequence of P. anserina. In addition, comparative genomic using P. anserina genome sequence should yield interesting results about the evolution of fungal species. A BAC library has been constructed in the vector pBHYG, which is derived from pBeloBAC11 by the addition of a selectable hygromycine resistance gene in P. anserina. Several BAC which contain sequences from the locus of the centromere of chromosome V have been isolated and their sequencing is in progress. Preliminary results suggest that the region which is genetically defined as the centromere of chromosome V is not present in the library. Several genes present on either sides of the centromere have been identified in the BAC sequences. The progress of the work can be monitored on http://www.cns.fr/ or http://cgdc3.igmors.u-psud.fr/.

157 AFLP linkage map of Gibberella zeae. J. E. Jurgenson1, R. L. Bowden2, K. A. Zeller2, J.F. Leslie2, N. A. Alexander3, and R. D. Plattner3. 1Department of Biology, University of Northern Iowa, 2Department of Plant Pathology, Kansas State University, 3Mycotoxin Research Unit, USDA/ARS National Center for Agricultural Utilization

A genetic linkage map of Gibberella zeae (Fusarium graminearum) was constructed by crossing nitrate nonutilizing (nit) mutants of G. zeae strains R- 5470 (from Japan) and Z-3639 (from Kansas). Ninety- nine nit+ progeny were selected and analyzed for polymorphisms using AFLP markers. Thirty-one pairs of two-base selective primers revealed 1044 polymorphic markers that mapped to 444 unique loci on nine linkage groups. The total map length of the genome from this analysis was 1247 centimorgans with an average interval of 2.8 map units between loci. Three linkage groups had high levels of segregation distortion. Selection of nit+ recombinant progeny accounts for two of the skewed regions. One linkage group had an abundance of symmetrical crossovers and appeared to have an intercalary inversion. Loci governing red pigment, trichothecene toxin amount, toxin type (deoxynivalenol vs. nivalenol), and the Tri5 gene were mapped.

158 High throughput mutation by targeted disruption in the rice blast fungus, Magnaporthe grisea. Grant Nelson, Rex W. Tarpey, Kiichi Adachi, Maria V. Montenegro-Chamorro, Sheri Frank, Todd Dezwaan, Amy Skalchunes, Ryan Heiniger, Blaise Dareaux, Sanjoy Mahanty, Matthew Tanzer, Lisbeth Hamer, Lakshuman Ramamurthy and Jeffery Shuster. Paradigm Genetics, Microbial Department, RTP, NC, USA

Paradigm Genetics is industrializing the process of gene function determination. The Microbial Research Group has developed a functional genomics platform to explore fungal genomes and uncover gene function across an array of economically important organisms. The fungus M. grisea is the causative agent of rice blast disease and is responsible for severe economic losses worldwide. Like most of the filamentous fungi, the frequency of homologous recombination in M. grisea is generally low. Our functional genomics platform utilizes TAGKOTM (transposon-arrayed gene knockout) technology to rapidly generate 'knock out vectors' of the entire genome. Furthermore, all of these cosmids contain flanking regions of ~40 kilobases. Subsequent sequencing and bioinformatics provide large numbers of targets with corresponding knock out vectors. To determine the function of these targets, we have developed a method for high throughput in vivo disruption. This method is based upon maximizing the frequency of homologous recombination in filamentous fungi and rapidly screening transformants to identify knock out mutants. Data will be presented, illustrating that transformation with TAGKOTM vectors dramatically increases the frequency of targeted disruptions in M. grisea.

159 Functional genomics of Phytophthora sojae: identification of a necrosis inducing peptide by a heterologous expression assay. Dinah Qutob1, 2, Mark Gijzen1 and Sophien Kamoun3. 1Agriculture and Agri-Food Canada, Southern Crop Protection and Food Research Centre, London, ON, Canada; 2Department of Microbiology and Immunology, University of Western Ontario London, ON, Canada; 3Department of Plant Pathology, The Ohio State University-OARDC, Wooster, OH, USA, 44691.

Phytophthora sojae is an oomycete that causes stem and root rot on soybean plants. To discover pathogen factors that activate plant defense responses, we identified putative secretory proteins from a database of expressed sequence tags (ESTs) (Plant Physiol. 2000 May;123(1):243-54) and assayed selected candidates using a potato virus X (PVX)-based heterologous expression system. First, we screened translational reading frames within ESTs for signal peptide motifs. From an analysis of 3035 ESTs originating from mycelium, zoospore, and infected soybean tissues, we identified over 300 putative proteins with N-terminal transit peptides. We further analyzed the candidates for a series of features of the encoded proteins, including size, hydrophobicity, putative glycosylation, similarity to known proteins, and completeness of the open reading frame. We selected a total of 17 different cDNAs encoding putative secreted proteins ranging in size from 5-24 kD for expression analysis in Nicotiana benthamiana using PVX/Agrobacterium-based assays. The results indicated that recombinant PVX expressing two of the 17 candidates caused necrosis and browning symptoms atypical of wild-type PVX infection. One cDNA was especially active, causing both localized and spreading tissue damage of infected plants. This necrosis-inducing peptide shares sequence similarity to other proteins found in fungal and bacterial species and may be involved in virulence of P. sojae.

160 Identification of a hypervirulent mutant and genes of mitochondrial origin incorporated into the nuclear genome of Cryptococcus neoformans. Nelson, R. T., Handly, L. A., Hua, J., and Lodge, J. K.. Saint Louis University, Doisy Department of Biochemistry and Molecular Biology, St. Louis, MO. 63104

Cryptococcus neoformans is a yeast-like fungus with a world wide distribution. C. neoformans causes cryptococcosis which is a threat to a large patient population with impaired immune function. We have identified a signature tagged insertional mutant which is able to produce approximately 10 fold more CFUs in the brain of mice in a disseminated model of cryptococcosis. The virulence of this mutant does not appear to be a result of increased resistance to the reactive oxygen intermediates nitrous oxide or hydrogen peroxide relative to the parental strain. Analysis of the flanking genomic sequence indicates that the insertion interrupts the 5' UTR for a mitochondrial ND4L gene on one side and sequence that has no match in the existing databases on the other side. CHEF analysis of the insertion site, indicates that this site is in the nuclear genome on a chromosome of approximately 1.2 Mb. Southern blot of CHEF gels of the parental strain, H99 used in the signature tagged mutagenesis, indicates that a number of mitochondrial genes are also present in this chromosome. This is the first report of hypervirulence and the first report of mitochondrial genes present in the nuclear genome of Cryptococcus neoformans var. neoformans strain H99.

161 Targeted gene inactivation by homologous recombination in Cryptococcus neoformans var. neoformans serotype A using linear DNA and biolistic delivery. Nelson, R. T. and Lodge, J. K.. Saint Louis University, Doisy Department of Biochemistry and Molecular Biology, St. Louis, MO 63104

Cryptococcus neoformans var. neoformans is a opportunistic pathogen of immunocompromised individuals. Homologous recombination in Cryptococcus has been studied mainly in serotype D strains. In order to systematically investigate the parameters for homologous recombination and targeted gene inactivation in C. neoformans serotype A, we constructed a number of DNA fragments with varying lengths of sequence (400, 300, 200, 100 and 50 bp) from a serotype A CAP59 gene flanking a selectable marker. Disruption of this gene inhibits capsule polysaccharide synthesis producing an acapsular phenotype. These molecules were transferred into the genome of Cryptococcus using biolistic transformation followed by analysis of phenotype and genotype by PCR and Southern blot. The results indicate that as little as 200 bp of homologous sequence on each flank could specifically target the inactivation of the CAP59 gene with an efficiency of 19%. Increasing the length of the flanking sequence to 400 bp increased the efficiency to 54%. Asymmetrical constructs were created with 400 bp of flanking sequence on one side of a selectable marker and 50 bp on the other. These constructs were as effective (6-22)% in inactivating the CAP59 gene as the symmetrical molecules containing 200 bp of homologous sequence. These results indicate that linear molecules with as few as 200 bp of homologous sequence on both sides of the selectable marker can be used to effectively knock out genes in C. neoformans serotype A and that a short flanking sequence on one side of a selectable marker can be complemented by a longer flanking sequence on the other flank.

162 Chromosome landing across avirulence loci in the potato late blight pathogen, Phytophthora infestans. Stephen Whisson1, Theo van der Lee2, Glenn J Bryan3, Robbie Waugh3, Francine Govers2, and Paul RJ Birch1 . 1Scottish Crop Research Institute, Unit of Mycology, Dundee, Scotland. 2University of Wageningen, Phytopathology, Wageningen The Netherlands. 3Scottish Crop Res. Inst., Unit of Genomics, Dundee Scotland

Phytophthora infestans, which causes late blight of potato and tomato, is a globally important plant pathogen. On potato, resistance to late blight is conditioned by a gene-for-gene interaction between potato resistance (R) genes and P. infestans avirulence (Avr) genes. Six P. infestans Avr genes have been genetically mapped and AFLP markers have been identified tightly linked to these Avr loci. We have adopted a map-based cloning approach for cloning the Avr genes from P. infestans. A 10-fold genome coverage bacterial artificial chromosome (BAC) library was constructed and clones were pooled for screening by AFLP. Overlapping BAC clones which span two avirulence loci (Avr11, Avr4) have been identified, and BAC clones near two other Avr genes (Avr1, Avr2) have also been identified. Current and future efforts will be focussed on confirming the presence of Avr genes in the BAC clones, and identifying the functional Avr genes. Cloning of an Avr gene from P. infestans will yield insights into the molecular interaction between P. infestans and potato, and may form a basis for novel sources of resistance to late blight.

163 A comparative study of Aspergillus fumigatus genes. Anderson MJ, Prebble EJ, Denning DW. University of Manchester, Dept of Medicine, Manchester, Gr Manchester, UK

A comparative analysis of all the published Aspergillus fumigatus genes has been carried out as an aid for the A. fumigatus genome sequencing project. This project was initiated in July 1999 and as of December 2000, funding has been obtained to sequence 50 % of the genome. Funding for the remainder is currently being sought. Two sets of analyses have been performed on the full-length A. fumigatus genes published in the literature or in the DNA databases (~60). Consensus sequences have been determined for the initiation codon and for the 5' and 3' splice sites of introns. These consensus sequences were generated using only those sequences which had been confirmed experimentally. Intron splice site sequences were compiled only in those examples where both the sequence of the gene and mRNA had been determined and for the initiation codon, sequences were complied where the position had been determined by N-terminal protein sequencing, transcriptional start site mapping or by comparison of gene and mRNA sequences. These consensus sequences should be useful for programming gene-finding software. The second set of analyses compared A. fumigatus gene sequences with similar sequences in other Aspergilli to determine the level of conservation in the structure of homologous genes. Similar sequences were identified by searching the fungal directory of the EMBL database using Fasta. These sequences were aligned and the size, position and number of exons were compared. Some interesting examples will be presented including the comparison of catalases, polyketide synthases, chitin synthases and proteases. This type of analysis might assist in determining which homologous genes are most closely related to each other and therefore help define orthologues.

164 Agrobacterium tumefaciens-mediated targeted gene disruption in Coccidioides immitis. Kris Osborn1, Marc J. Orbach2, Anath Das3, Maria Lourdes Lewis4 and John N Galgiani1. 1University of Arizona, Infectious Diseases, Tucson, AZ. 2University of Arizona, Plant Pathology, Tucson AZ USA. 3University of Minnesota, Biochem,Mol Biol,Biophys, St. Paul MN USA. 4Southern AZ VA Healthcare, Research, Tucson AZ USA

Coccidioides immitis, the causative agent of Valley Fever, is a dimorphic fungal pathogen of humans and other mammals. To develop targeted gene disruption in C. immitis we used A. tumefaciens-mediated transformation to delete the gene for Ag2/PRA, a spherule wall protein of unknown function which stimulates protective immunity against coccidioidal infection. To create a knockout cassette, a 1.4 kb modified hygromycin B phosphotransferase gene (HygR) under the control of the Aspergillus nidulans trpC promoter was used to replace the Ag2/PRA gene. This construct was placed between the left and right borders of an A. tumefaciens binary vector to create pKO322. A. tumefaciens (5X108 cells) harboring pKO322 were co-cultivated with 5x106 germinated (24 hour) arthroconidia for 3 days, after which the germlings were plated on 2XGYE containing 20 g/ml hygromycin and 50 g/ml kanamycin. At least 50 discrete HygR C. immitis colonies grew in each of 3 experiments. Selected colonies were screened by PCR. The results of this screen, together with the observation that more than 90% of arthroconidia contain two nuclei, raised the concern that transformed strains were heterokaryons. We passaged arthrocondia from 26 putative transformed strains under more stringent selective conditions to isolate homokaryons. Homokaryons from eleven separate transformation events were recovered. In nine of these, the Ag2/PRA gene appeared to be knocked out as evidenced by a PCR product predicted for replacement of the locus by the HygR cassette and no detectable Ag2/PRA signal. Thus, the rate of homologous integration for this gene was high enough to be readily observed in a single transformation experiment.

165 High-throughput gene disruption in Ashbya gossypii reveals genes required for normal filamentous fungal growth. Thomas Gaffney, Krista Gates, Keri Cavanaugh, John Marhoul, Natasha Springer, Melisa Harrison, and Michelle Kirksey. Syngenta-RTP, Research Triangle Park, NC, USA

The filamentous ascomycete Ashbya gossypii is a useful model organism in a functional genomics approach due to its small genome size (ca. 9 megabases encompassing ca. 5000 genes) and the relative ease with which gene replacement experiments can be conducted. Also, yeast replicons function in Ashbya, allowing complementation of mutant constructs. Since the Ashbya genome has not undergone the extensive duplication observed in the Saccharomyces genome, it is often more straightforward to link phenotype with genotype in the Ashbya background. Our studies with Ashbya have revealed numerous genes whose effect on filamentous growth would have been difficult to predict based upon analysis of related Saccharomyces genes. Also, we have found that several Ashbya homologues of Saccharomyces genes required for yeast pseudohyphal growth do not influence Ashbya filamentous growth.

166 Identifying pathogenicity and essential genes in fungi by deletion scanning. Jianguo Wu, Barbara Robbertse, Xun Wang, Olen Yoder and Gillian Turgeon. Novartis Agricultural Discovery Institute, San Diego, California, U.S.A.

Knowledge of the fungal genes essential for life and those controlling molecular mechanisms of pathogenicity would suggest both fungicide targets and strategies by which plants resistant to disease might be developed. We are taking a genome-wide approach to the identification of such genes in Cochliobolus heterostrophus, a pathogen of maize. The project involves three elements: (1) Sequencing the fungal genome; (2) Directed mutagenesis aimed at evaluating candidate genes whose products are suspected of being involved in fungal pathogenesis, and (3) Saturation mutagenesis of the genome. For the third goal, a library designed to delete small random fragments from the genome was constructed and transformed into a wild type strain. Each transformant is tested for viability and pathogenicity. Mutants with either altered virulence or lethality are noted and the plasmid used for transformation sequenced, permitting the deleted DNA to be identified in each case. To validate the result and to identify the gene(s) responsible for the phenotype of interest, each ORF affected by the deletion is targeted individually. A pilot study indicated that mutations affecting viability and virulence could be detected by this strategy.

167 Genome characterization of three plant pathogenic ascomycetes. Paolo Amedeo1, Scott Baker1, Todd Moughamer2, Don Hutchison3, John Thompson2, Trini Miguel3, Hemant Varma2, Darrell Ricke2, Chris Martin2, Stephen Goff3, Olen Yoder1, and Gillian Turgeon1 . Novartis Agricultural Discovery Institute Inc., 3115 Merryfield Row, San Diego, CA, 92121-1125 1 Plant Health 2 Bioinformatics 3 Structural Genomics

The genomes (~35 Mb) of three phylogenetically diverse plant pathogens (Cochliobolus heterostrophus, Gibberella zeae/Fusarium graminearum and Botrytis cinerea) have been sequenced to different extents using a shotgun approach. Each has been characterized individually and by comparison. The data allow us to make comparisons to genomes of other organisms such as the single celled, fungal non-pathogens Saccharomyces cerevisiae and Schizosaccharomyces pombe, the multicellular, filamentous, non-pathogens Neurospora crassa and Aspergillus nidulans, and to other filamentous pathogens. Inter-organism comparisons should allow us to sort genes essential for life, for filamentous growth, and for pathogenicity. The latter is expected to be subdivided into two categories: those common to all or most pathogens, and those unique to particular pathogens with specific host ranges.

168 Genetic analyses of Iprodione resistant mutants of Cochliobolus heterostrophus and cloning of histidine kinase genes. Akira Yoshimi, Chihiro Tanaka and Mitsuya Tsuda. Pesticide Research Institute, Faculty of Agriculture, Kyoto University, Kyoto 606-8502, Japan.

We have isolated Iprodione resistant mutants of Cochliobolus heterostrophus, and identified mutant genes at three loci; Ipr1, Ipr2 and Ipr3 by crossing experiments. Strains carrying mutant alleles at Ipr1 or Ipr2 showed Iprodione resistance (MIC>400 g/ml) and sensitive to high osmolarity (ED50 for KCl, 300 400mM), whereas wild type's MIC for Iprodione was 2.5 g/ml and ED50 for KCl was ca. 600mM. At Ipr3, we have distinguish two types mutant allele by their ED50 for KCl. One type showed ED50 similar range of ED50 to the Ipr1 or Ipr2, the other allele showed highly sensitivity to high osmolarity (ED50, 25 200mM). These were indicating that the osmotic stress responding genes in the pathways such as histidine kinase and MAP kinase would be involved in the Iprodione resistance in this fungus. We made an attempt to clone histidine kinase by a PCR approach. Based on the consensus amino acid sequences in the histidine autokinase domain, the nucleotide-binding domain and ATP-binding domain, degenerated primers were designed and used to amplify genomic fragments of C. heterostrophus. Three kinds of fragments from putative histidine kinase genes (bmhk-1, bmhk-2 and bmhk-3) were obtained. Nucleotide sequence of bmhk-1 and bmhk-2 showed high similarities to Nik-1 (os-1) from Neurospora crassa, Chk-1 from Glomerella cingulata , respectively.The bmhk-3 shared homology with LuxQ from Vibrio cholerae. Gene- disruption studies are currently underway.

169 Agrobacterium tumefaciens T-DNA as a potential insertional mutagen for Blastomyces dermatitidis and Histoplasma capsulatum. Thomas D. Sullivan, Peggy J. Rooney, Colin P. Kealey and Bruce S. Klein. University of Wisconsin, Pediatrics, Madison, WI, USA

B. dermatitidis and H. capsulatum are pathogenic dimorphic fungi responsible for pulmonary infections and disseminated disease. Random mutagenesis by insertion of known DNA sequences may aid in identification and cloning of virulence factors and other genes of these fungi. Integration of T- DNA from A. tumefaciens has been used in this way in plants, and recent studies indicate that A. tumefaciens can deliver DNA into fungal cells. Vectors containing hygromycin phosphotransferase or URA5 selectable markers within T-DNA sequences were introduced into A. tumefaciens. ura5 auxotrophs of H. capsulatum and newly isolated ura5 auxotrophs of B. dermatitidis were used as targets for URA5 marker selection. Cocultivation of A. tumefaciens and yeast-phase cells was followed by selection for hygromycin-resistant or uracil-prototrophic yeast. For auxotrophic strains of both fungi, transformation frequencies were greater than ten-fold higher using URA5 selection than for hygromycin selection. Southern blot analyses confirmed that the T-DNA is often integrated at a single site in the genome, but at different sites in independent transformants. As B. dermatitidis yeast are multinucleate, we sought to obtain cells harboring the same mutation in all nuclei, facilitating isolation of recessive mutations during phenotypic screens. To this end, we also transformed germinating single- nucleate B. dermatitidis conidia by this method.

170 Sequencing the Cryptococcus neoformans genome. Eula Fung, Richard W Hyman, Don Rowley, Dan Bruno, and Ronald W Davis. Stanford Genome Technology Center, Stanford, CA USA

Cryptococcus neoformans is the etiologic agent of cryptococcosis, one of the most serious fungal diseases worldwide. C. neoformans' haploid genome is contained on 13 chromosomes and totals 21 Mb. Strain JEC21, serotype D was chosen for sequencing because it represents the only known isogenic mating pair. Our approach is a whole genome shotgun plus assembly. We will identify and finish as many open reading frames as possible, subject to the usual time and money constraints.

As a service to the research community, all of the raw shotgun sequence data and contigs-in-progress will be posted on an overnight basis on the Stanford Genome Technology Center website. An initial assembly of the data was done in late October 2000 when we had reached 3-fold coverage of the genome. The assembly includes the circular mitochondrial DNA and 3900 contigs covering almost 13 Mbases of the nuclear DNA. As of Dec 15, 2000, we have over 138,000 shotgun reads. Taken together, these reads provide almost 81 Mbases. C. neoformans (strain JEC21, serotype D) genomic DNA was generously provided by Dr. June Kwon-Chung, NIAID, NIH, USA. Our C. neoformans Genome Project is supported by a cooperative agreement (U01 AI47087) from the NIAID, NIH, USA. Funding commenced in March, 2000, and will run for three years.


171 Aspergillus nidulans genome sequencing effort: Strategic scaffold sequencing of Chromosome IV. Patricia Ayoubi1, Nigel Dunn-Coleman2, Nancy Keller3 and Rolf Prade1. 1Department of Microbiology and Molecular Genetics, Oklahoma State University, Stillwater, OK, USA. 2Genencor International, Palo Alto, 925 Page Mill Road, CA, USA. 3Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX, USA.

As a simple multicellular eukaryote, the ascomycete Aspergillus nidulans is an important model for addressing fundamental questions in cellular and molecular biology. A. nidulans has also become a model system for basic study of biotechnology and pathogenesis of related multicellular ascomycetous fungi. In an effort to generate publicly available genome sequence data of A. nidulans, we have started pilot scale sequencing of chromosome IV ( 2.9 Mb). Since the A. nidulans genome is estimated to contain ~8,000 genes over ~30 Mb, this sequencing effort is expected to reveal ~800 genes. Chromosome IV is covered by 83 non-overlapping ordered cosmids called tiles (average 40 kb inserts) with an additional 62 cosmids filling the gap between tiles. This minimum set of overlapping cosmids is a scaffold that provides the necessary framework for maximizing gene discovery from sequencing efforts. Our sequencing strategy involves subcloning (2 kb) and shotgun end-sequencing of about 180 subclones per cosmid. To anchor all cosmids known to hybridize to chromosome IV, 240 cosmids have also been end-sequenced. To date, approximately 25,000 successful reactions have been completed. Up to date progress on the assembly and sequence analysis of these data will be presented.

172 Aspergillus nidulans EST microarrays. Hugh Russell1, Sven Krappmann1, Patricia Ayoubi1 , Nigel Dunn-Coleman2, and Rolf Prade1 1Department of Microbiology and Molecular genetics, Oklahoma State University, Stillwater, OK 74078 USA and 2Genencor International, Palo Alto, 925 Page Mill Road CA 94304-1013 USA

The northern blot has been the standard for transcript monitoring for many years; however, with finished and current genome sequencing projects mounting, new large-scale gene expression analysis tools are needed. Microarrays are DNA/DNA hybridization devices in which ALL or a large number of gene-specific targets (DNA fragments) derived from a genome, have been cross-linked to a support (e.g., glass) and hybridized against labeled probes of high degree of complexity - e.g., reverse transcribed and labeled mRNA populations. DNA microarray gene expression monitoring has been useful in various applications including drug discovery, transcriptional regulated gene expression studies, and gene discovery. Here, we describe the construction of a glass-based microarray containing approximately 4,600 cross-linked targets derived from the A. nidulans expressed sequence tag (EST) collection. PCR amplified targets were copied from a cDNA library, primed by an EST-specific and a vector-common oligonucleotide set. Template targeted regions for amplification were chosen, whenever possible, to map to the 3"-end of the EST (ORF) and produce a 500 bp fragment. Multiple bands amplified with a single EST primer could in theory be utilized, if in the majority of cases they correspond to alternate 3' end terminations. However, we have determined that under our laboratory conditions, alternate 3'-end termination could not be established in PCR reactions showing two or more bands. Thus, PCR reaction which did not result in a single band, were redone using a secondary and/or a tertiary PCR cycling protocol. Finally, new primers for non- validated PCR products were designed. All PCR-products were ethanol- precipitated, targets validated by agarose gel electrophoresis and printed onto glass-slides. Microarrays will be made available upon request by Genencor International and Oklahoma State University.

173 Constructing and sequencing a BAC contig of the Phytophthora sojae genome. Felipe Arredondo and Brett Tyler. Department of Plant Pathology University of California, Davis Davis, CA95616.

Phytophthora sojae causes $1-2 billion in damage to soybean crops worldwide each year. To facilitate isolation of pathogenicity genes from P. sojae by map-based cloning, we are constructing a BAC contig of the entire genome of this organism, using a hybridization fingerprinting strategy. We are hybridizing a 13X BAC library with unique mixtures of random probes, most of them repetitive, and with ESTs. Computer software has been developed to collect, simulate and analyze the data. At present we have probed the library with 49 of the mixtures. Each mixture hybridized to around 300 - 500 BACs. 19% of the BACs so far have received the minimum number of hits needed to establish statistically significant overlaps (5 each). Of these BACs, 34% have been placed into contigs. We have confirmed the authenticity of three of the largest contigs by HindIII digestion.

With the long term goal of sequencing the entire genome of P. sojae, we have begun preliminary sequencing of a 200 kb BAC contig spanning two avirulence genes from P. sojae. Sequencing of the first 60 kb BAC is nearly complete. In the region sequenced so far the gene density is extremely high. Both the sequencing data and the BAC hybridization data suggest that the P. sojae genome is composed of gene-rich regions separated by regions rich in repetitive sequences.

174 EST analysis and sequencing of chromosome 7 of the rice blast fungus, Magnaporthe grisea. T. Mitchell1, T. Houfek1, T. Mitchell1, S. Martin1, D. Ebbole2, R. A. Dean1. 1Fungal Genomics Laboratory, North Carolina State University, Raleigh, NC 2Texas A&M University, College Station, TX

Magnaporthe grisea, the causal agent of rice blast disease, is not only the largest threat to rice production worldwide, but is recognized as a leading model system for fungal biology, genetics, and pathology. As part of an international effort, a major initiative has been funded to use a genomics approach for gene discovery and whole genome analysis of this pathogen. We are currently sequencing 35,000 5' ESTs from 8 cDNA libraries prepared from different stages of growth and development. A minimum set of 5,000 unique genes will then be sequenced from the 3' end to validate sequence data and contig assemblies. Concurrently, we are using a BAC-by-BAC approach to sequence chromosome 7 of this fungus. We will obtain ~5X coverage of the minimum tile of 41 BACs covering this chromosome. All EST and BAC sequence data will be publicly available on a BLAST server. Results from initial EST and BAC sequencing efforts will be presented as well as an analysis of sequence data for each part of the project.

175 Durable resources for discovery and development of new gene products from Trichoderma reesei. Phase I cDNA and BAC end sequencing. M. Chellappan1, N. Dunn Coleman2, A. Hillian1, T. Houfek1, T. Mitchell1, P. van Solingen3, P. Teunnissen3, D. Wang2, M. Ward2, J. Yao2, R. A. Dean1. 1Fungal Genomics Laboratory, North Carolina State University, Raleigh, NC 2Genencor International, Palo Alto, CA 3Genencor International, Leiden, The Netherlands

The filamentous fungus Trichoderma reesei is a major producer of cellulytic and other enzymes used in commercial applications. The discovery of new genes is critical to the continued development of commercially competitive biologically based products. In this study, two cDNA libraries were constructed, one with RNA extracted from cells grown in conditions promoting the production of cellulytic enzymes, and the other from cells grown in 18 different conditions. A total of 9,792 ESTs were sequenced using ABI 3700 automated sequencing system. Of these, 6,603 were identified to be of high quality (excluding those clones with no inserts) using criteria to be presented. Using Phrap 4,267 clones were placed into 1,207 contigs leaving 2,336 singlets. Each of the two libraries had the same level of redundancy while the redundancy for the entire project stands at 46%. The total number of unique sequences identified is 3,543. Sequence data was subjected to BLASTX and N similarity searches and processed through an automated gene-indexing package. Approximately half of sequences had significant matches. A HindIII BAC library has been constructed and clones are being sequenced from the ends. Results from cDNA and BAC end sequencing, BLAST searched, and indexing will be presented. BAC filters, BAC clones, and cDNA clones will be made available to academic users via online requests.

176 EST analysis for the discovery of new gene products from Aspergillus niger. N. Dunn-Coleman2, M. Chellappan1, M. Conboy2, A. Hillian1, T. Houfek1, T. Mitchell1, P. van Solingen3, P. Teunnissen3, D. Wang2, M. Ward2, J. Yao2, R. A. Dean1. 1Fungal Genomics Laboratory, North Carolina State University, Raleigh, NC 2Genencor International, Palo Alto, CA 3Genencor International, Leiden, The Netherlands

Many fungi, particularly Aspergillus species, produce various secondary metabolites desirable for commercial development. An EST sequencing project was established to facilitate the discovery of new genes with potential for commercial applications. Two cDNA libraries were constructed using total RNA derived from different growth conditions such as nutrient limitation, stress response and alternative N, C sources. One library was normalized using a novel technique to be presented. Sequence results show an overall success rate of 80% with an average high quality read length of 462 bp. Sequence data has been subjected to BLASTX and N similarity searches and processed through an automated gene-indexing package. BLAST results will be presented and will be available in a searchable and browsable format on a publicly accessible web site. cDNA sequence and clones will be made available to academic users upon request.

177 PipeOnline 2.00: data mining of processed DNA sequence databases. Patricia Ayoubi1, Eduardo Misawa2 and Rolf Prade1 . 1Department of Microbiology & Molecular Genetics and 2School of Mechanical & Aerospace Engineering, Oklahoma State University, Stillwater, OK 74078 USA.

The exponential increase in uncharacterized public genomic DNA and cDNA derived sequence data has driven the development of computational methods for processing data for the purpose of gene identification and inference of biological function. PipeOnline 2.00 is an experimental Web-based resource designed by the OSU Bioinformatics Laboratory to aid investigators in determining metabolic and other protein functions from large-scale DNA sequence collections. Demonstration databases were generated from public cDNA sequence data derived from fungal and plant model organisms. Typically, raw DNA sequence data (collection of trace or FASTA files) are processed automatically by PipeOnline 2.00, a series of script-linked programs which process raw DNA sequence files and produces a database of records that can be retrieved through queries, or comprehensive gene-function browsing. PipeOnline 2.00 produces contig-assembled files, assembled using publicly available software. These contigs are than automatically compared against a local NCBI non-redundant protein database using BLASTX and resulting output files automatically collected, parsed, formatted, assembled, indexed and uploaded to a local MySQL server. PipeOnline also automatically produces a functionally sorted output, facilitating analysis of large DNA sequence collections, specifically data sets with high levels of functional redundancy. Functional sorting was achieved through a proprietary sorting method that utilizes functional information gathered from public databases. The current version of PipeOnline 2.00 employs definitions and the biologically threaded functional tree derived from the Metabolic Pathways Database, MPW dictionary.

178 Comparative studies of gene organization between Magnaporthe grisea and other fungal species. Jun Seop Jeong, and Ralph A. Dean, Fungal Genomics Laboratory, North Carolina State University, Raleigh, NC 27695 USA.

Magnaporthe grisea causes rice blast disease, which results in devastating losses in rice production worldwide. Due to its significance, the genetics and physiology of the fungus have been studied extensively--It has become a leading model system for the investigation of host-pathogen interactions. Numerous resources including genetic maps, large insert BAC libraries, and ESTs provide a foundation for genomic studies. Chromosome 7 has been studied in most detail including the mapping of several hundred ESTs. Emerging sequence data from several fungal species, now, allow for comparative analysis of gene organization to be undertaken. Comparative studies have revealed syntenic relationships among higher eukaryotes, such as grasses and vertebrates. However, detailed comparison of genome organization among fungal species is lacking. cDNA clones from chromosome 7 of M. grisea were identified, and used to investigate syntenic relationships in Aspergillus nidulans and Neurospora crassa by sequence alignment and DNA hybridization analysis. This comparative approach will facilitate genome reconstruction efforts of filamentous fungi and provide insight into genome evolution and function.

179 Genomic sequence comparisons and the role of small molecules in signaling in filamentous ascomycetes. Scott Kroken, John. W. Taylor, and N. Louise Glass. Dept. of Plant and Microbial Biology, UC-Berkeley

Filamentous fungi make and secrete small molecules that may be involved in signaling, defense or scavenging nutrients. In collaboration with Olen Yoder and Gillian Turgeon at Syngenta, we are performing comparative genomics of saprobes (e.g. Neurospora crassa, Saccharomyces cerevisiae and Schizosaccharomyces pombe) and plant pathogens (Cochliobolus heterostrophus, Botrytis cinerea, Fusarium graminearum and Ashbya gossypii). A comparative analysis is aimed at identifying genes that encode secreted ribosomal peptides, polyketides (PKs), and nonribosomal peptides (NRPs). We have cloned a 2530 amino acid-encoding gene from N. crassa that is similar to PKS1 (38% identity), a polyketide synthase (PKS)-encoding gene involved in making a plant toxin (T-toxin) in C. heterostrophus. A N. crassa 5217 amino acid-encoding gene was identified that is similar to HTS1 (33% identity), a nonribosomal peptide synthetase (NRPS)-encoding gene also involved in the production of a plant toxin (HC toxin, a cyclic tetrapeptide) in C. carbonum. Based on analysis of the as yet incomplete N. crassa genome sequence, it is predicted that N. crassa has 3-7 PKS and PS genes, in contrast to C. heterostrophus, which is predicted to have >35 PKS-encoding genes and >35 NRPS-encoding genes. Based on these comparisons, a mutational analysis of genes involved in the production of small molecules will be performed in N. crassa. Metabolite profiling between wild-type N. crassa and mutants will be performed to match PKS and NRPS genes with secreted metabolites. Our goal is to delineate the role of small molecules in signaling pathways that are required for formation of the hyphal network, sporulation, vegetative growth and response to environmental cues in filamentous fungi.

Return to the XXI FGC Abstract Page

Return to the XXI FGC Main Page

Go to the FGSC main page