BSPP Presidential Meeting 2003

Plant Pathogen Genomics - From Sequence To Application


Session 4: Emerging Insights Into Pathogens and their Interactions With Plant Hosts

Bioinformatic resources for studying phytopathogenic fungi: The COGEME phytopathogen EST database

Darren M. Soanes, and Nicholas J. Talbot

School of Biological Sciences, University of Exeter, Washington Singer Laboratories, Perry Road, Exeter, EX4 4JQ, United Kingdom.

Currently, there are only a limited number of publicly accessible genomic resources available to researchers studying phytopathogenic fungi. The COGEME EST database (http://cogeme.ex.ac.uk/) contains expressed sequence tag (EST) sequences collected from 13 species of phytopathogenic fungi and oomycetes. The EST sequences have been clustered into unisequences to reduce redundancy and to allow comparative genomic analysis of phytopathogenic species. The relational database currently contains 30,819 unisequences. Putative products have been assigned to each unisequence by homology to known sequences and they have been classified according to function using the MIPS ontology scheme. The database has been implemented using MySQL and has an easy to use web interface that allows the user to identify sequences using simple text queries, by functional classification, and by homology to Saccharomyces cerevisiae genes. The web interface also allows the use of BLAST algorithms to search for unisequences that have homology to entered sequences. Complete unisequence datasets are also available to download. The database also contains a tool to analyse transcript abundance based on representation of EST sequences in raw data-sets from the rice blast fungus Magnaporthe grisea, assigning a probability value to the frequency of occurrence in cDNA libraries derived from the fungus at different developmental stages. A demonstration tool that uses probability analysis to predict the occurrence of amino acid biosynthetic pathways in the different species of phytopathogen from incomplete datasets is also provided in the suite of COGEME databases.

References:

Soanes, D.M., Skinner, W., Keon, J., Hargreaves, J., Talbot, N.J. (2002) Genomics of phytopathogenic fungi and the development of bioinformatic resources. Molec. Plant-Microbe Interact. 15: 421-427

Tunlid, A., Talbot, N.J. (2002) Genomics of parasitic and symbiotic fungi. Current Opinion in Microbiology 5: 513-519

Giles, P.M., Soanes, D.M., and Talbot, N.J. (2003) A relational database for the discovery of genes encoding amino acid biosynthetic enzymes in pathogenic fungi. Comparative and Functional Genomics 4:4-15


Microbial history revealed by genome analysis

Siv G. E. Andersson, Bastien Bosseau, Wagied Davids, Carolin Frank, Hans-Henrik Fuxelius, Olof Karlberg, Lisa Klasson, Boris Legault, Hillevi Lindross, Gabor Nyori

Dept. of Molecular Evolution, Evolutionary Biology Center, Uppsala University, Norbyvagen 18C, 752 36 Uppsala, Sweden 

Some members of the alpha-proteobacteria have evolved elaborate interactions with plants, while others are pathogens of wild and domestic animals. Yet other species in this subdivision are pathogens of humans causing diseases such as typhus, trench fever and cat scratch disease. The recent sequencing of a dozen alpha-proteobacterial genomes, including our own genome data from Rickettsia prowazekii, Bartonella quintana and Bartonella henselae, enables a global genomic comparison of human, animal and plant-associated bacteria. Here, we present a phylogenetic reconstruction of the alpha-proteobacteria for which complete genome sequence data is available and discuss genomic features that are shared between human, animal and plant pathogens.  We identify differences in gene numbers, gene contents and genomic architectures that correlate with major lifestyle changes. We show that genome size expansions of a few thousand genes have accompanied the evolution of the plant-associated bacteria. In contrast, shifts to intracellular animal environments and vector-mediated transmission pathways are characterized by massive gene loss. We present a reconstruction analysis of the alpha-proteobacterial ancestral genome and discuss its relationship with mitochondria. We conclude that the genomic evolution of the alpha-proteobacteria is driven by the many different interactions with animals, humans and plants.


Plant-viroid interactions

Ricardo Flores

Universidad Politcnica-CSIC, Valencia, Spain 

Viroids are small (246-401 nt), non-coding, circular, single-stranded RNAs that replicate autonomously into certain host plants and incite specific diseases in some of them. Disorders affect plants of interest for the food industry (potato, tomato, cucumber, hop, grapevine, and tropical and temperate fruit trees that include coconut palm, citrus, avocado, peach, plum, apple and pear) and ornamentals (chrysanthemum and coleus). Some viroids have destructive effects whereas others are asymptomatic.

The lack of coding capacity of viroids establishes a fundamental distinction with viruses, with implications on their replication (viroids strongly depend on pre-existing cell enzymes) and pathogenesis (viroids must induce diseases by direct interactions with host factors). The underlying mechanisms of replication and pathogenesis are most likely diverse because some viroids like Potato spindle tuber viroid (PSTVd), the type species of family Pospiviroidae, replicate and accumulate in the nucleus whereas others like Avocado sunblotch viroid (ASBVd), the type species of family Avsunviroidae, replicate and accumulate in the chloroplast.

Viroids replicate through an RNA-based rolling circle mechanism with three steps: i) synthesis of longer-than-unit strands, ii) cleavage to unit-length, and iii) circularization. Depending on whether cleavage and circularization occurs only in the plus (the most abundant in vivo) or in both polarity strands, the mechanism is termed asymmetric or symmetric, respectively. The first step is catalyzed by the nuclear RNA polymerase II in family Pospiviroidae, which follows the asymmetric pathway, and by a chloroplastic RNA polymerase in family Avsunviroidae, which follows the symmetric pathway. Remarkably, the second step is mediated in family Avsunviroidae by hammerhead ribozymes embedded in both polarity strands, whereas a host RNase is presumed to act in family Pospiviroidae. The third step has been proposed to occur via an RNA ligase or autocatalytically. However, viroid replication *as well as the intracellular, cell-to-cell and long-distance movement* most likely need additional host proteins whose nature remains poorly understood.

The host range of different viroids is very variable: some can infect a broad range of species, while others are restricted to a few related members of the same genus or family. One of the most intriguing aspects of viroids is how they elicit pathogenic effects into their hosts. In members of both families, changes affecting to approximately 1% of the sequence, the so-called virulence-modulating region or pathogenicity determinant, are sufficient to transform a severe into a latent strain. These changes most likely influence the interaction of the viroid with a host factor(s), presumably a protein, which as a result of being diverted from its normal physiological role, may induce a primary alteration and through a signal transduction pathway lead to the onset of symptoms. However, the recent finding in tissues infected by representatives of both families of the viroid-specific small interfering RNAs (siRNAs) associated with post-transcriptional gene silencing (PTGS), opens the possibility that these siRNAs could mediate viroid pathogenicity as well as cross-protection phenomena which, like in viruses, have been also reported in viroids.


Pathogenomics and finding effectors, their function and the affected

John Mansfield, Marta de Torres, George Tsiamis, Hassan Ammouneh; Department of Agricultural Sciences, Imperial College, Wye Campus, Ashford, Kent TN25 5AH, Andy Pitman, Dawn Arnold, CRIPS, Biological Sciences, UWE Bristol, BS16 1QY, Rob Jackson; Department of Plant Sciences, University of Oxford, OX1 3RB, Jens Boch; Martin-Luther-Universitt, Institut fr Genetik, Halle(Saale), Germany 

This talk will focus on avr gene mobility as illustrated by avrPphB and the function of the effector VirPphA. Genes for avirulence and virulence (encoding effectors) have been cloned from Pseudomonas syringae pvs. phaseolicola (Pph) and pisi (Ppi)using four approaches. The first three are based on functional screens involving 1: exchange of genomic libraries between avirulent and virulent races of each pathovar (e.g. avrPphB, avrPphE, avrPphF, avrPpiA and avrPpiB), 2: library exchange between the pathovars (avrPphD, avrPpiC),and 3: restoration of virulence to plasmid-cured avirulent strains of Pph (virPphA, avrPphF and avrPphC).  The fourth approach has been through use of PCR primers based on regions of DNA found to flank avrPpiA (giving avrPpiC and avrPpiG).  Certain genes such as avrPphF and avrPpiC have been recovered using more than one approach.  Dual avirulence and virulence functions have been assigned to several genes.  Homologues of some of the effectors cloned by function have been detected through bioinformatics-based mining of the sequenced P.s. pv. tomato and P.s. pv. maculicola genomes.  Intriguingly, the effector proteins all seem to be injected by bacteria into plant cells through an 8nm diameter needle, the Hrp pilus.

The presence of functional alleles of the avr genes in Pph determines virulence on differential cultivars of bean.  Race change leads to breakdown of resistance and can be achieved through mutation (avrPphE) or gene loss (avrPphB). Analysis of the race 4 * race 2 alteration caused by loss of avrPphB from strain 1302A has identified deletion of a 100 kb region of the race 4 chromosome which is flanked by tRNA genes.  Sequencing the deletion containing avrPphB reveals a region with characteristic features of an ICEland (Integrative conjugal element) with sequence similarity to P.s. pv. syringae B728A and also Xylella fastidiosa.  Unlike some other effector genes, homologues of avrPphB are not widely distributed in strains of P. syringae, however, regions flanking avrPphB were used to amplify DNA from other P.syringae pvs and this approach recovered avrRpt2 from P.s. pv. tomato.  The avrPphB gene is located within a mobile mosaic of apparently interchanged gene cassettes and provides a useful model for the analysis of avr gene mobility.

In contrast to avrPphB, homologues of virPphA are widespread and include avrPtoB which was identified as an avr gene "interacting" with Pto in tomato.  In Pph virPphA, avrPphC  and avrPphF are located on a pathogenicity island (PAI) containing several avr genes and mobile elements.  We have examined the virulence function of virPphA homologues.  Completion of a pathogenicity matrix using alleles and tests on bean, soybean and Arabidopsis reveals differential effects. Both avrPtoB and virPphA suppress the HR in bean.  In soybean, virPphA but not avrPtoB acts as an avr gene.  In Arabidopsis, only avrPtoB acts as a virulence factor but does so by suppressing basal resistance associated with cell wall alterations rather than the HR.  Expression of AvrPtoB in Arabidopsis leads to enhanced symptom production by non-pathogenic bacteria.  Targets for AvrPtoB activity in Arabidopsis are being examined using yeast two hybrid and immunoprecipitation approaches.  Putative targets, including both constitutively expressed (e.g. fibrillin) and induced proteins (e.g. receptor like kinase) have been identified.  The emerging pattern is that mobile effectors and PAIs have evolved to overcome basal defences in plants in addition to modulating activation of the HR.


Pathogenicity and trichothecene mycotoxin production by the ear blight pathogens Fusarium graminearum and F. culmorum  

Martin Urban1, Thomas Baldwin 1, Arsalan Daudi1, Dimitry Kornyukhin1, Frances Trail2 and Kim E. Hammond-Kosack1  

1 Wheat Pathogenesis Programme, Rothamsted Research, Plant-Pathogen Interactions Division, Harpenden, Herts, AL5 2JQ, UK
2 Departments of Plant Biology and Plant Pathology, Michigan State University, East Lansing MI 48824, USA  

This Fusarium ear blight project is part of the new Wheat Pathogenesis programme at Rothamsted Research which aims to identify the common molecular themes underlying disease formation on cereal hosts caused by non-biotrophic fungi.  The two other species investigated are the leaf blotch pathogen Mycosphaerella graminicola (Septoria tritici) and the eyespot pathogen Tapesia yallundae.

In the laboratory, we are undertaking an experimental approach that combines forward and reverse genetics with comparative bioinformatics analyses of other sequenced fungal genomes and EST collections to identify the Fusarium genes required for pathogenicity and mycotoxin production.  

F. graminearum and F. culmorum cause ear blight disease on small grain cereals1. Infections at anthesis, not only lower grain quality but also lower grain safety because both fungal species produce a range of trichothecene mycotoxins post-infection. Within different genetic lineages of F. graminearum unique trichothecene mycotoxin chemotypes exist. Each of these genetic lineages has a distinct global distribution pattern and the different mycotoxin chemotypes appear to have independently evolved in each lineage2.  In the UK and North West Europe, F. culmorum is the major pathogenic species, whereas in North America and Asia, F. graminearum predominates. It is not known whether F. graminearum and F. culmorum differ in their infection biology of the wheat ear and whether their mycotoxin chemotype influences the invasion process.  If species differences do exist, this may influence the development of disease control options. 

At the molecular level, it is not known which genes and gene networks are essential for Fusarium pathogenicity on wheat ears and whether the pathogenicity genes required differ between the two Fusarium species. A molecular genetic analysis has indicated that deoxynivalenol (DON) mycotoxin production is not essential for F. graminearum to cause disease in wheat ears3.  Also our laboratory and others have demonstrated that two distinct Mitogen Activated Protein Kinases (MAPKs) Map1 and Mgv1, are independently required for infection and subsequent spread within the wheat ear4.

Recently, our group has shown that both F. culmorum and F. graminearum can infect Arabidopsis floral tissue causing disease symptoms. During these floral infections DON mycotoxin synthesis also occurs5.  This experimental system can be used to explore the Fusarium genes required to penetrate and cause floral disease in cereal and non-cereal host species separated by over 100 million years of evolution.

The resources and techniques available to undertake a large scale exploration of Fusarium gene function are exceptionally good. They include 10x genome sequence coverage for F. graminearum (Whitehead Institute, Cambridge, USA) (http://www-genome.wi.mit.edu/annotation/fungi/fusarium/index.html), various libraries of expressed sequence tags (ESTs) (http://cogeme.ex.ac.uk) and efficient transformation systems to create specific gene knockouts within 4-6 weeks. 

References 1 http://www.scabusa.org; 2 O'Donnell K et al.  (2000) Proc. Natl. Acad. Sci. USA, 97: 7905-7910, Ward TJ et al. (2002) Proc. Natl. Acad. Sci. USA, 99: 9278-9283; 3 Proctor RH et al. (1995) Mol. Plant-Microbe Interact. 5: 249-256; 4 Hou Z et al.  (2002) Mol. Plant-Microbe Interact. 15: 1119-1127, Jenczmionka NJ et al.  (2003) Current Genetics 43: 87-95, Urban M et al. (2003) Mol. Plant Pathol. 4: 347-359, 5 Urban M. et al.  (2002) The Plant Journal  32, 961-973. 

Rothamsted Research receives grant-aided support from the Biotechnology and Biological Sciences Research Council (BBSRC) of the UK. This work is supported in part by Monsanto, Syngenta and the Rothamsted International Fellows Programme.


Molecular Dissection of the Stagonospora nodorum wheat interaction 

Richard P. Oliver, Peter S. Solomon, Kar-Chun Tan, T.J. Greer Wilson & Robert Lee  

Australian Centre for Necrotrophic Fungal Pathogens, Murdoch University, Perth 6150, Western Australia, Australia. 

The Australian Centre for Necrotrophic Fungal Pathogens (ACNFP) was created to develop an understanding of necrotrophic fungal pathogen/host interactions at the molecular level, particularly those affecting Australian crops. One such project within the centre is focused on the interaction between the fungus Stagonospora nodorum and wheat. Stagonospora nodorum is the causal agent of leaf and glume blotch on wheat and is responsible for $60M (AUD) of crop loss in Australia each year. Whilst also appearing to be an economically important pathogen throughout the world, very little is known at a molecular level about how the fungus infects wheat. Using a range of classical microscopic techniques supplemented by the use of GFP-expressing strains, we have characterised the infection biology of the fungus. The fungus appears to penetrate the leaf both via stomata and directly through the cuticle. Small penetration structures, which we term hyphopodia, are normally found at penetration sites. Within 2-3 days the fungus ramifies through the depth of the leaf. This precedes obvious cell death. Pycnidia develop within 5-8 days or as soon as the leaf senesces.

We have begun dissecting this interaction using a variety of molecular techniques. We have developed a small collection of ESTs and developed moderate throughput methods to generate knock-out strains. Transformation and homologous recombination are efficient.

We have investigated genes previously associated with pathogenicity and some novel genes. Amongst the known genes are a heterotrimeric g-alpha protein encoding gene. The mutants have multiple phenotypes including lack of sporulation, lack of pigmentation, reduced osmo-tolerance and reduced secretion. Analysis of this gene has differentiated stomatal and direct penetration and has identified DOPA as the likely precursor of the melanin. Amongst the genes not previously associated with pathogenicity is a mannitol-1-phosphate dehydrogenase. Knock-outs of this gene lack all detectable M1PDH activity and have radically altered neutral sugar contents. Gross symptom production was not changed but the time taken to sporulate was increased markedly. This mutant has focused our attention on the conditions required for sporulation on the host.


Sensing, signalling and stress in the barley powdery mildew fungus

S.J. Gurr, Z. Zhang, C. Henderson, E. Perfect

Department of Plant Sciences, University of Oxford, OX1 3RB, UK.

Blumeria graminis is the causal agent of barley powdery mildew disease. Infection is spread by asexual conidia, which, on contact with the leaf surface, undergo a complex and highly regulated programme of development. They germinate and produce a short primary germ tube (PGT) followed by a second formed germ tube, the appressorium germ tube (AGT) which elongates, swells and produces a specialised, hooked infection structure, the appressorium.

B. graminis is an obligate biotroph; it cannot be grown axenically and consequently, tissue for experiments is limiting. We have described a range of techniques to assess how B. graminis perceives, integrates and relays signals for morphogenesis up to the point of penetration. Previously, our work demonstrated that both physical properties of the leaf surface, such as hydrophobicity, and cuticle-derived chemicals promote B. graminis differentiation. But how does B. graminis transduce signals to drive differentiation and development? Applications of exogenous agonists and antagonists have allowed us, with Richard Olivers group, to demonstrate a role for cAMP signalling and PKA in germling differentiation, but this work also highlights that cAMP alone is not sufficient to trigger the complete programme of differentiation. Hitherto, we have identified other genes involved in signal transduction and cell integrity pathways in B. graminis, notably PKC and MAP kinase cascades.

Plants defend themselves against pathogen attack by invoking a myriad of mechanisms, including the hypersensitive response, the production of antimicrobial metabolites, cell wall fortification and the papilla response. Much is known about the host defence, but, by contrast, virtually nothing is known about how true obligate plant pathogenic fungi cope with the hostile environment of the plant. As the pathogen must tolerate .O2- and H2O2 during colonisation of host tissue, the evolution of antioxidant detoxifying systems may arm it with a major selective advantage. We have turned our attention to look at the pathogens management of host-derived oxidative stress during development and penetration. In particular, we have focussed on the characterisation of the B. graminis catalase B gene, immunolocalisation of the mildew CATB protein, the detection of H2O2 during the infection process and the use of Magnaporthe grisea as a surrogate host in these studies.


Molecular Analysis of Pathogenicity of Leptosphaeria maculans

Barbara Howlett 1, Candace Elliott 1, Alex Idnurm 1,2, Donald Gardiner 1, Soledade Pedras 3, Marie-Helene Balesdent 4, Franoise Blaise 4, Michel Meyer 4, Estelle Remy 4, Thierry Rouxel 4

1 School of Botany, The University of Melbourne, Vic. 3010, Australia
2 Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27110, USA, 3 Department of Chemistry, University of Saskatchewan, Saskatoon, SK, S7N 5C9, Canada, 4 INRA-PMDV, Route de St Cyr, 78026 Versailles Cedex, France

Blackleg caused by the dothideomycete, Leptosphaeria maculans is the major disease of canola-oilseed rape (Brassica napus) worldwide.  This fungus is outcrossing and amenable to molecular genetic analysis.  It has a genome size of about 34 Mb, and its chromosomes are of a size range (0.7 to 3.5 Mb) and number (16) that can be readily resolved by electrophoretic karyotyping. 

We are using molecular genetic approaches, in particular random mutagenesis, to analyse fungal genes involved in the disease process.  Using Restriction Enzyme Mediated Integration we identified two pathogenicity genes - one that encodes an open reading frame of 529 amino acids with no matches to databases, and the other encoding isocitrate lyase, an enzyme enabling the fungus to use lipids as a carbon source in planta.  In parallel at Versailles and Melbourne, we have developed an efficient Agrobacterium-based transformation protocol for random insertion of selectable markers such as hygromycin resistance and Green Fluorescent Protein genes into the fungal genome. A range of vectors with properties that optimise insertions into coding regions of genes has been developed. In Melbourne, of 360 transformants analysed, six show reduced or loss-of-pathogenicity. In Versailles nine of 150 transformants analysed showed altered phenotypes in terms of pathogenicity. When five of these latter mutants were crossed with untransformed isolates to check whether the insertion co-segregated with the phenotype, three were shown to be untagged and two were tagged. Of the two tagged mutants, one showed loss-of-pathogenicity with an insertion in the putative promoter regions of two head-to-tail oriented ORFs. The second mutant showed increased pathogenicity in terms of speed of development of symptoms and sporulation on plant tissues. The analysis of all these mutated genes is underway.

We are also examining a cluster of about 15 genes, which is predicted as the pathway for biosynthesis of an epipolythiodioxopiperazine (EPT) phytotoxin named sirodesmin PL that is secreted at high levels by L.maculans. The presence of this cluster in other fungi correlates with the ability of such fungi to secrete EPTs. We are examining the regulation of expression of genes in this cluster and also disrupting them and then determining phenotype of the resultant L. maculans mutants.

Knowledge of the sequence of the L. maculans genome would greatly enhance the ability of the international community of blackleg researchers to advance knowledge of this important disease and would also be valuable for comparative analysis with genomes of closely related phytopathogenic fungi such as Stagonospora nodorum. Our common objective is to acquire the sequence of the L. maculans genome.


Multiple Layers of Host Resistance to Pathogens

J. Parker, B. Feys, L. Moisan, N. Medina-Escobar, M. Wiermer, S. Betsuyaku, P. Muskett, & L. Nol.

Max-Planck Institute for Plant Breeding Research, Dept. Plant-Microbe Interactions, Carl-von-Linn Weg 10, 50829 Cologne, Germany.

Plants are resistance to the majority of pathogen species and this resistance is exerted by a combination of constitutive barriers and inducible defences. Once a particular pathogen strain has evolved mechanisms to invade plant tissues the host is under strong selection pressure to recognize pathogen-derived effectors and engage these to trigger local and systemic immune responses. Genetic analysis of Arabidopsis pathogen interactions has helped us to identify plant recognition components (Resistance of R proteins) and unravel some of the complex defence regulatory circuits involved. Early redox changes and the activities of small signalling molecules, such as salicylic acid and lipid derivatives, are important for induction of appropriate defences to attack by necrotrophs or biotrophs. So far, little is understood about the nature of R protein complexes or how recognition events are transduced to downstream defences. We are particularly interested in the activities of two defence regulators, EDS1 and PAD4, that are required by a structural subset of R proteins in local race-specific resistance. In the absence of specific R protein recognition, EDS1 and PAD4 are more broadly utilized in basal resistance against virulent pathogens as well as in systemic immunity. A combination of genetics, protein biochemistry and cell biology is providing clues about how these proteins operate at different levels of cellular defence. Two other signalling components, RAR1 and SGT1, are required for full expression of race-specific resistance. RAR1 is a small zinc-binding protein necessary for the accumulation of a number of R proteins examined. SGT1 has been shown to have multiple activities in yeast and is essential for SCF (Skp1-Cullin-F Box protein) E3 ligase activity during cell cycle progression. SCF complexes mediate polyubiquitination of target proteins, often directing them for degradation. Recent evidence from yeast and plants points to a role of SGT1 as a molecular co-chaperone that may assist the assembly and/or activation of protein complexes. We are using genetics and protein biochemistry to define the functions of RAR1 and SGT1 more precisely in plant defence and development.