Organized by Seogchan Kang, Pennsylvania State University, University Park
At present, the genomes of more than 150 microorganisms have been sequenced or sequencing projects are underway. Although plant pathogens have been underrepresented in this group, this situation is rapidly improving due to the community-wide efforts to promote the genomics of plant pathogenic organisms. Information and technology resources derived from these efforts will significantly enhance our ability to develop effective measures to control important plant diseases. This symposium will not only highlight the progress of genome sequencing efforts for selected plant pathogens, but also present how genome sequence data have been utilized to investigate the biology and evolution of pathogens and the mechanisms of their interactions with host plants.
Comparative Genomic Analysis of Fungal Plant Pathogens: Secondary Metabolites and Mechanisms of Pathogenesis
B. G. Turgeon1,2, S. Kroken2,3, B.-N. Lee2, S. E. Baker2, P. Amedeo2, N. Catlett2, U. Gunawardena2, E. Wagner2, B. Robbertse2, J. Wu2, O. C. Yoder2, N. L. Glass3, J. W. Taylor3
(1) Plant Pathology, Cornell University, Ithaca, NY 14853;
(2) Torrey Mesa Research Institute, 3115 Merryfield Row, San Diego, CA 92121;
(3) Plant & Microbial Biology, University of California, Berkeley CA 94720.
Communication between microbe and plant is mediated by components of communication circuits (signal transduction pathways) that start with or lead to production of signaling effectors, some of which are likely to be small molecules elaborated via the enzymes of secondary metabolism. Our genomics program focuses on several filamentous ascomycete pathogens (approximately 35 Mb in size) representing wide phylogenetic distribution, and on two pathogens from within the same genus (Yoder and Turgeon 2002). Our primary focus is the maize pathogen, Cochliobolus heterostrophus, chosen as a model pathogen for functional analyses because it is among the most genetically tractable (sequenced to >5X coverage). For comparative purposes, we have included Gibberella fujikuroi (Fusarium verticillioides), causal agent of stalk rot of corn (sequenced to >5X coverage), and G. zeae (Fusarium graminearum), a pathogen of wheat, corn, rice and barley (sequenced to 2X coverage), and Botrytis cinerea, a cosmopolitan pathogen of dicots, including Arabidopsis (sequenced to >5X coverage). In addition to their phylogenetic distribution, these fungi use different modes of infection and are of varying degrees of economic importance.
Of major significance is the fact that Cochliobolus spp. and a sister genus, the Alternaria spp., plus the Gibberella spp., are all notorious for their ability to produce a wealth of secondary metabolites, some of which have clear roles as pathogenicity/virulence determinants such as the host-specific toxins produced by Cochliobolus spp. (Yoder et al., 1997) and the Alternaria spp. (Hatta et al., 2002) and the trichothecenes produced by G. zeae (Cardwell et al, 2001). Little is known about the importance of secondary metabolites in pathogenic behavior of B. cinerea. We have employed phylogenomic analyses to help evaluate roles in pathogenesis of two important classes of enzyme involved in secondary metabolism- polyketide synthases and non-ribosomal peptide synthetases, responsible for production of polyketides and small peptides, respectively. All genes for these enzymes in the sequenced genomes have been identified and annotated, and compared to their counterparts in the genomes of other fungi that have been sequenced: e.g., the early diverging saprophytic ascomycetes Saccharomyces cerevisiae and Schizosaccharomyces pombe, and the human pathogen, Candida albicans, and other filamentous ascomycetes, i.e., the saprophyte Neurospora crassa, the human pathogen Aspergillus fumigatus, the plant pathogens Magnaporthe grisea (rice) and Ashbya gossypii (cotton bolls) and the microsporidian, Enchephalitozoon cuniculi.
We have discovered that:
• The early diverging ascomycetes have few or none of these secondary metabolism genes, while all filamentous pathogens, both plant and animal, have many. The filamentous saprophyte Neurospora has an intermediate number.
• Saprophytes and plant pathogens have the same classes of secondary metabolite genes.
• The plant pathogens have different sets of orthologous genes.
• Few PKS or NRPS orthologs are shared among three polyphyletic plant pathogens.
• Even closely related fungi vary greatly in presence/absence of these genes.
PKSs can be grouped according to the small molecule produced: linear or cyclic toxins, pigments, or 6-methylsalycylic acid-based toxins. NRPSs can be classified as those producing siderophores, penicillin-like compounds, phytotoxins and mycotoxins, cyclosporin-like toxins, and those that fall into three novel clades with no characterized genes. Among these, we are particularly interested in genes of unknown function, conserved across the fungi, which fall into virulence clades, as these are candidates for roles as general virulence factors; interrupting function of these genes or their products may offer a solution for general disease control.
OneOneAt least one example has been identified in the NRPS gene set. The gene is also present in Neurospora, leading us to speculate that the product may have been recruited from its original function for a specific aspect of the pathogenic lifestyle.
Phylogenomic analyses have revealed orthologs of secondary metabolite genes that have discontinuous distribution among species. For example, in G. fujikuroi the FUM1 gene encodes a PKS involved in the biosynthesis of fumonisin, a potent mycotoxin. FUM1 resides in a cluster of fifteen genes all involved in the biosynthesis of fumonisin (Proctor et al, 2002). None of the genes in the FUM1 cluster is found in a sister species, G. zeae, which does not produce fumonisin. Yet, the FUM1 cluster is found in distantly related C. heterostrophus, although there is no microsynteny between the two clusters and the C. heterostrophus cluster lacks five of the genes found in the G. fujikuroi cluster. Comparisons of the two clusters and knowledge of the structure of fumonisin allows us to predict the structure of the hypothetical fumonisin analog produced by C. heterostrophus.
To summarize, there is a correlation between the pathogenic life style in fungi and numbers of genes in the genome involved in biosynthesis of secondary metabolites. Pathogenic fungi have many such genes, saprophytes have few or none. Moreover, among the gene sets in pathogenic fungi there are few orthologs, an observation that predicts an astonishingly large number of small molecules produced by fungi on this planet.
Cardwell, K.F., Desjardins, A.E., Henry, S.H., Munkvold, G., Robens, J. 2001. Mycotoxins: The Cost of Achieving Food Security and Food Quality
Hatta R, Ito K, Hosaki Y, Tanaka T, Tanaka A, Yamamoto M, Akimitsu K, Tsuge T. 2002. A Conditionally Dispensable Chromosome Controls Host-Specific Pathogenicity in the Fungal Plant Pathogen Alternaria alternata. Genetics. 161:59-70
Proctor R.H., Brown, D.W, Plattner, R.D., and Desjardins, A.E. 2002. Co-expression of fifteen contiguous genes delineates a fumonisin biosynthetic gene cluster in Gibberella moniliformis. Fungal Genetics and Biology. In press.
Yoder O.C., Turgeon B.G. 2001. Fungal genomics and pathogenicity.
Current Opinion in Plant Biology 4:315-21
Yoder O.C., Macko, V., Wolpert, T.J., and Turgeon B.G. 1997. Cochliobolus spp. and their host-specific toxins. pp145-166. In The Mycota: Plant relationships. Vol. 5. Part A. edited by G. Carroll and P. Tudzynski. Springer Verlag, Berlin.
Functional Genomics of Phytophthora-Plant Interactions
Sophien Kamoun, The Ohio State University, Ohio Agricultural Research and Development Center, Wooster
Oomycetes, such as Phytophthora, form a unique branch of eukaryotic pathogens and are arguably the most devastating pathogens of dicot plants. Molecular phylogenies based on ribosomal RNA sequences, compiled amino acid data for mitochondrial proteins, and several protein encoding chromosomal genes strongly suggest that oomycetes form a unique lineage of stramenopile eukaryotes, unrelated to "true fungi", but closely related to heterokont (brown) algae. It is evident from these analyses that oomycetes evolved the ability to infect plants independently from other eukaryotic plant pathogens, and are likely to have unique mechanisms to do so.
The most destructive and best-studied oomycete is Phytophthora infestans, the cause of late blight, a disease that results in multibillion-dollar losses in potato production and is considered a significant threat to global food security. Similar to other plant pathogens, P. infestans has the remarkable ability to manipulate biochemical, physiological and morphological processes in its host plants through a diverse array of virulence or avirulence molecules, defined here as effectors. In susceptible plants, these effectors promote infection by suppressing defense responses, enhancing susceptibility, or inducing disease symptoms. Alternatively, in resistant plants, effectors are recognized by the products of plant resistance genes resulting in the hypersensitive response and effective defense responses. One central objective in studying the molecular basis of pathogenicity of P. infestans is to identify and functionally characterize effector genes.
Structural genomic studies of P. infestans are well under way within the framework of the Phytophthora Genome Consortium and the Syngenta Phytophthora Consortium, as well as efforts in individual laboratories. Ongoing and pending projects consist of:
• cDNA sequencing.
• Genome sequencing.
• Bacterial artificial chromosome (BAC) clone-end sequencing.
• Sequencing of selected BAC clones.
• Genetic and physical maps.
• Mitochondrial genomics.
Genomic science promises to significantly impact our understanding of P. infestans biology and pathology. The main applications of genomics are:
• Understanding the molecular basis of P. infestans pathogenicity through the identification of genes that contribute to the infection process.
• Identifying P. infestans targets for genetic control, specifically avirulence genes that function at host and nonhost level.
• Identifying P. infestans targets for chemical control, such as essential genes.
• Understanding P. infestans population structure and evolution, through improved molecular markers and markers that determine phenotypes.
P. infestans genomics has already started to impact our understanding of the molecular basis of pathogenicity through the identification of effector genes. The challenge in the post-genome era is to link a sequence to a phenotype. An overview of ongoing functional genomic studies on P. infestans is described in Kamoun et al. (From sequence to phenotype: Functional genomics of Phytophthora. Canadian Journal of Plant Pathology, 24:6-9, 2002.). These studies have allowed us to unravel a battery of novel P. infestans genes that trigger a variety of cellular and molecular responses in plant cells and to start to establish functional connections between P. infestans genes and plant processes.
To link sequences to phenotypes, we applied the functional genomics paradigm for the discovery of novel effector genes and fungicide targets in P. infestans. First, we developed computational tools for mining the sequence data, then we applied robust high throughput functional assays to validate the predicted function of the candidate genes.
To select candidate genes from sequence databases, we used the following selection criteria:
• Genes that encode degradative enzymes. These genes are predicted to encode putative virulence factor involved in host tissue penetration and degradation.
• Genes that encode extracellular proteins. These are more likely to be involved in cross-talk with host plant.
• Genes that are up-regulated during infection. These encode putative virulence or pathogenicity factor and could serve as fungicide targets.
• Genes that are conserved across oomycetes. These could serve as fungicide target.
Examples of candidate genes identified using the various data mining strategies includes genes predicted to function in adhesion to host cells during appressorial formation, genes encoding degradative enzymes that may facilitate penetration of plant tissue and formation of haustoria, as well as genes that function in infection, such as putative virulence and avirulence genes.
To validate the various candidate genes and to identify novel ones, high throughput assays need to be implemented. At this stage, there is still a need for improvement and adaptation to large scale analyses of the gene knockout and complementation assays currently available for P. infestans and other Phytophthora species. However, ectopic expression of pathogen genes in plant cells can be performed at a remarkable high throughput rate using potato virus X (PVX) and Agrobacterium tumefaciens-based vectors. Therefore, we have been using virus-mediated gene expression to carry out high throughput functional screens of Phytophthora genes in plants. Preliminary PVX based functional screens unraveled a battery of novel Phytophthora effector genes that trigger hypersensitive-like necrosis in Nicotiana, tomato, and potato as well as genes that alter the interaction between Phytophthora and plants.
In summary, P. infestans genomics has already generated numerous candidate genes, several of which are being validated using various functional assays. This is allowing us to understand the molecular basis of pathogenicity, to identify targets for genetic and chemical control, and to understand population structure and evolution in P. infestans.
In collaboration with others in the P. infestans community, we are now seeking to complete the sequence of the 237 Mb genome of P. infestans before the year 2005. Grant proposals to this effect are pending and additional proposals will be submitted in the coming years. The sequence will be of tremendous value for the oomycete and plant breeding community by providing significant insight into key molecular processes regulating this economically important pathosystem and offering novel tools for improvement of late blight resistance in potato. We therefore urge the late blight and plant communities at large to support our efforts and to submit letters of support to help this initiative.
Phytophthora genome Consortium
Phytophthora infestans genome sequencing
Global Initiative on Late Blight
NSF Potato Functional Genomics
Molecular Plant Pathology - Pathogen Profile "Phytophthora infestans enters the genomics era"
Canadian Journal of Plant Pathology article "From sequence to phenotype: functional genomics of Phytophthora"
Kamoun Lab web page
Pseudomonas syringae pv. tomato DC3000: Genomics and phytopathogenicity
A. Collmer1, J. R. Alfano2, A. M. Baldo3, C. R. Buell4, S. Cartinhour3, A. K. Chatterjee5, T. P. Delaney1, S. G. Lazarowitz1, G. B. Martin1, D. J. Schneider3, X. Tang6
(1) Cornell University, Ithaca, NY;
(2) University of Nebraska, Lincoln;
(3) USDA-ARS, Ithaca, NY;
(4) Institute for Genomic Research, Rockville, MD;
(5) University of Missouri, Columbia;
(6) Kansas State University, Manhattan
Bacterial speck of tomato caused by Pseudomonas syringae pv. tomato DC3000 (Click image for larger view).
Pseudomonas syringae pv tomato DC3000 is an important model in molecular plant pathology because it is a genetically tractable pathogen of the well-studied hosts tomato and Arabidopsis, its interactions with plants appear representative of many common bacterial and fungal pathogens, and its parasitic abilities appear to be based largely on a variety of virulence "effector" proteins that are injected into plant cells and which provide a tractable starting point for the molecular dissection of pathogenesis. The NSF Plant Genome Research Program has funded a multi-institutional project to explore the functional genomics P. syringae interactions with plants. The Pseudomonas-Plant Interaction (PPI) project involves researchers at Cornell University, Boyce Thompson Institute for Plant Research, Kansas State University, University of Missouri, University of Nebraska, and The Institute for Genomics Research (which is sequencing the DC3000 genome). Publication of the complete and annotated genome sequence is pending, but all draft editions of the genome have been made immediately available for public use from TIGR (see the TIGR Microbial Database), and these draft sequences have already supported several papers from members of the NSF project and the larger scientific community.
A primary focus of researchers working with the draft sequence has been the identification of novel Hops (Hrp-dependent Outer Proteins), which are secreted by the P. syringae Hrp (type III secretion) system. Some Hops may function as extracellular components of the type III "injector" system, but most are thought to be effectors that are injected into plant cells to promote bacterial parasitism. Finding the genes encoding such effectors has been difficult heretofore because mutations in these genes typically do not have a strong effect on virulence, apparently because of redundancy. Thus, despite their presumed central role in bacterial parasitism, effector genes are typically missed in mutant screens based on reduced bacterial virulence. To overcome this problem the NSF PPI project, in collaboration with the USDA/ARS Center for Agricultural Bioinformatics at Cornell University, has used an iterative process involving computational methods and gene expression and protein secretion assays to characterize the Hrp regulon and potential protein targeting signals and to identify more than 22 Hops in P. s. tomato DC3000.
Sequence logo representing the Hrp promoter Hidden Markov Model (Click image for larger view).
The first stage in this search for novel effectors was founded on the knowledge that the limited number of effector genes that were previously known (many designated "avr" because of the avirulence phenotype they conferred) in various strains of P. syringae are preceded by "Hrp box" promoter sequences. Because there is substantial variation in these promoter sequences, a reporter transposon approach was used with P. s. tomato DC3000 to identify additional functional Hrp promoters. This enlarged training set of functional sequence variants was used to generate a Hidden Markov Model, which was subsequently used to globally identify potential Hrp promoter sequences in the DC3000 genome (Fouts, D. E., Abramovitch, R. B., Alfano, J. R., Baldo, A. M., Buell, C. R., Cartinhour, S., Chatterjee, A. K., D'Ascenzo, M., Gwinn, M. L., Lazarowitz, S. G., Lin, N.-C., Martin, G. B., Rehm, A. H., Schneider, D. J., van Dijk, K., Tang, X., and Collmer, A. 2002. Genomewide identification of Pseudomonas syringae pv. tomato DC3000 promoters controlled by the HrpL alternative sigma factor. Proc. Natl. Acad. Sci. USA 99:2275-2280).
The second stage in the search for novel effectors addressed the signals at the N-terminus of Hops that target these proteins to the type III secretion pathway. Several of the candidate effector genes that are linked to Hrp promoters were experimentally demonstrated to encode proteins secreted in a Hrp-dependent manner. Analysis of the enlarged collection of P. syringae proteins traveling the Hrp pathway revealed a pattern of equivalent solvent-exposed amino acids in the N-terminal five positions, a lack of acidic amino acids in the first 12 positions, and other predictive properties. These characteristics were used to search the DC3000 genome, yielding 32 additional genes that are likely to encode effectors based on potential targeting signals and other properties. Interestingly, one of these encodes a homolog of SrfC, which is a candidate effector in Salmonella enterica (Petnicki-Ocwieja, T., Schneider, D. J., Tam, V. C., Chancey, S. T., Shan, L., Jamir, Y., Schechter, L. M., Buell, C. R., Tang, X., Collmer, A., and Alfano, J. R. 2002. Genomewide identification of proteins secreted by the Hrp type III protein secretion system of Pseudomonas syringae pv. tomato DC3000. Proc. Natl. Acad. Sci. USA 99:7652-7657).
The approaches described above are now being amplified to complete the inventory of virulence effector genes in P. s. tomato DC3000. They demonstrate the power of combining computational and experimental methods in exploring pathogenic mechanisms. Many important questions now can be addressed. For example, what are these effector proteins doing in plants to promote parasitism, what additional virulence factors will be found when the complete genomic sequence is analyzed, and how can we use this new information to develop better controls for bacterial plant diseases?
Discovery of plant genes required for disease resistance through a combination of expression profiling and reverse genetics
J. Glazebrook, J. D. Clarke, B. Estes, W. Chen, H.-S. Chang, and T. Zhu, Torrey Mesa Research Institute, San Diego, CA
Plants respond to pathogen attack by activation of a large number of defense responses, including transcriptional activation of hundreds of genes. The signal transduction network controlling activation of defense responses is complex, and most of the genes responsible for its control and for executing defense mechanisms remain to be discovered. Complete elucidation of plant responses to pathogens will require application of high-throughput approaches for identification of participating plant genes and understanding of their roles in disease resistance. In an attempt to contribute to this undertaking, we are using expression profiling and reverse genetics to study the response of Arabidopsis thaliana to infection by the virulent bacterial pathogen Pseudomonas syringae pv. maculicola strain ES4326 (Psm ES4326).
Infection by Psm ES4326 triggers activation of a signal transduction pathway that uses salicylic acid (SA) as a signaling intermediate. Mutations that interfere with SA signaling lead to enhanced susceptibility to this pathogen, indicating that SA-dependent signaling controls activation of defense responses that limit Psm ES4326 growth. Some other pathogens, for example the fungal pathogen Alternaria brassicicola, are not inhibited by SA-dependent responses, but are inhibited by responses under control of a different signal transduction pathway that requires jasmonic acid (JA) and ethylene (ET) as signaling intermediates. These two pathways are interconnected in a complex manner, as SA signaling can inhibit JA/ET signaling, and vice versa, while some genes are positively regulated by both pathways (See Glazebrook, 2001; and Kunkel and Brooks, 2002; for reviews of disease-resistance signaling pathways in Arabidopsis).
To identify plant genes induced in response to infection, we profiled wild-type plants infected with Psm ES4326 on an Affymetrix GeneChip representing approximately 8,000 Arabidopsis genes (Zhu and Wang, 2000). We also profiled a number of similarly-infected mutant plants with defects in SA, JA, or ET signaling, as well as mutant plants with enhanced susceptibility to Psm ES4326 but whose effects on signaling were unknown. Analysis of the data revealed pathogen-induced genes controlled by different signal transduction pathways. One group of genes required SA signaling, and was negatively affected by JA/ET signaling. Another group required JA/ET signaling, and was negatively affected by SA signaling. Other groups with more complex patterns of regulation were also observed. The mutants were sorted into a hierarchy according to observed alterations in gene expression patterns. As expected, mutants with defects in SA signaling formed a group distinct from mutants with defects in SA or ET signaling.
The pathogen-induced genes may include some that are required for effective disease resistance. We have tested this idea using a reverse genetics approach. Syngenta has created a large collection of Arabidopsis plants carrying T-DNA insertions (See Description of SAIL Project at TMRI) . The positions of the insertions were determined by DNA sequencing of the flanking regions of the genome. Using this collection, we have identified plants with mutations in several hundred pathogen-induced genes. We are now testing the mutant plants for enhanced susceptibility to Psm ES4326. Among the first 111 genes tested, 27 reproducibly display an enhanced-susceptibility phenotype, indicating that they contribute to resistance to Psm ES4326. Clearly, this combination of expression profiling and reverse genetics is an efficient method for identifying plant genes required for disease resistance.
Glazebrook, J. (2001). Genes controlling expression of defense responses in Arabidopsis--2001 status. Curr Opin Plant Biol 4, 301-308.
Kunkel, B.N., and Brooks, D.M. (2002). Cross talk between signaling pathways in pathogen defense. Current Opinion in Plant Biology 5, 325-331.
Zhu, T., and Wang, X. (2000). Large-scale profiling of the Arabidopsis transcriptome. Plant Physiol 124, 1472-1476
RETURN TO APSnet FEATURE STORY