Posters III: Plant and Animal Fungal Pathogenesis

A non pathogenic strain of Colletotrichum magna (path-1) was recently isolated. Unlike the C. gloeosporioides and the C. magna, wt isolate, path-1 did not cause disease symptoms in either the pericarp nor the mesocarp of avocado fruit. Isoelectric focusing of culture filtrate of Colletotrichum strains revealed the presence of PL and PG with no significant activity in Path-l. DNA and RNA analysis of the three Colletotrichum isolates hybridized to a pel probe from C. gloeosporiodes indicating that at the DNA level no deletion in the pe1 gene was evident and at the mRNA level similar expression levels were found. However, when PL antibodies were cross-reacted with cell lysate of the Colletotrichum strains, PL accumulated in the non pathogenic mutant suggesting that differential pathogenicity accounts for malfunction in extracellular transport of PL protein.

2. Cloning and characterization of a "tomatinase" gene from Botrytis cinerea.

Saponin-detoxifying enzymes occur in several plant pathogenic fungi and have been shown to determine host-specificity e.g. in Gaeumannomyces graminis (Osbourn et al. 1994, Bowyer et al. 1995). Using the tomatinase gene of Septoria lycopersici as a probe (A. E. Osbourn, pers. commun.) we isolated a corresponding gene from a genomic lambda library of B. cinerea. The putative tomatinase gene (toml) shows significant homology with the probe (up to 60% at aminoacid level) and with the tomatinase gene of Gaeumannomyces graminis. Sequences homologous to tom1 were present in all field isolates of B. cinerea tested so far, though not all of the strains show tomatinase activity. The role of the "tomatinase" in pathogenicity of B. cinerea is under investigation. Osbourn A., Bowyer P., Bryan G., Lunnes P, Clarke B., Daniels M.J.: in Advances in Molecular Genetics of Plant- Microbe Interaktions, Vol. 3:215-221 (1994); Bowyer P., Clarke B., Lunnes P., Daniels M.J., Osbourn, A., Science (in press)

3. Purification and characterization of two enzymes, a bifunctional beta-xylosidase/alpha- arabanosidase and an alpha-arabanosidase, from the maize fungal pathogen Cochliobolus carbonum.

Two enzymes, an alpha-L-arabinosidase and a bifunctional beta-D-xylosidase/alpha- L arabinosidase secreted by C. carbonum grown on medium supplemented with corn cell walls, were purified to homogeneity from culture filtrates . The Mr Of the alpha-L-arabanosidase (CCAR) was 30-32 kD and the Mr of the bifunctional beta-D-xylosidase/alpha-L-arabanosidase (CCXA) was 42-44 kD. As is found with most other secreted cell-wall-degrading enzymes, CCAR and CCXA are catabolite repressed. The CCAR enzyme accounts for approximately 75% of the total alpha-L-arabanosidase activity, with the remainder contributed by the CCXA. Amino acid sequence of two peptides from an AspN-protease digest of CCXA showed more than 50% identity with a bifunctional beta-D-xylosidase/alpha-L- arabanosidase from Bacteriodes ovatus. Oligonucleotides derived from these sequences are being used to screen a C. carbonum genomic library, and peptide fragments of CCAR are being sequenced. The final goal of this research is to disrupt the genes encoding CCXA and CCAR to determine their role in the pathogenicity of C. carbonum on maize.

4. Cell wall degrading enzymes of fungal plant pathogens.

Cell wall degrading enzymes (CWDEs) are secreted by all phytopathogenic fungi and have been thought to play a role in the penetration and ramification by these pathogens into healthy plant tissues. In Cochliobolus carbonum genes have been cloned that encode endo-polygalacturonase, exo- polygalacturonase, three endo xylanases, cellulase, endo-beta-1,4 glucanase, exo-beta-1,3-glucanase, beta- 1,3-1,4 glucanase, and an alkaline proteinase. Using a transformation-mediated gene disruption strategy, many of these genes have been mutated singly in pathogenic wild-type C. carbonum strains. In every case, the disruption of the CWDE gene and the concomitant loss of CWDE activity did not cause any detectable change in the pathogenicity of the mutant strain. Strains are currently being constructed containing multiple CWDE gene disruptions to assess effects on pathogenicity. Our goal is to identify all the CWDEs and their genes from C. carbonum. However, many CWDEs are difficult to assay and purify, and therefore their genes cannot be approached biochemically. To circumvent this problem, a subtracted cDNA library approach is being taken. Genes that are expressed when C. carbonum is grown on corn cell walls as carbon source but not when grown on sucrose will be identified by chemical cross-linking subtraction using the cross linking agent 2,5-diaziridinyl-1,4- benzoquinone (DZQ). The cDNAs isolated will be sequenced and putative CWDE encoding genes tested for a role in pathogenicity.

5. Characterization of a species-specific DNA probe for the in situ identification of Entomophaga maimaiga infected gypsy moth larvae.

The zygomycete, Entomophaga maimaiga, is a member of the Entomophaga aulicae species complex and is responsible for the recent collapse of gypsy moth populations in Northeastern U.S.A. We report here the isolation of a DNA fragment cloned from E. maimaiga, which hybridized to DNA of geographically diverse E. maimaiga isolates, but did not hybridize to DNA from other entomophthoralean fungi, including other isolates of the E. aulicae species complex. Sequence analyses revealed that this species-specific DNA fragment is composed of several simple sequence motifs. The DNA sequence contained several stop codons in all six reading frames. Although weak similarity was observed with a Drosophilia LTR, no significant homology at the nucleotide or amino acid levels, was found in searches of existing databases. Quantitative dot blot hybridizations indicated that the sequence contained a repetitive element present at approximately 161 copies per E. maimaiga nucleus. The repetitive nature of this sequence proved highly useful and allowed for the detection of an E. maimaiga infection within a single gypsy moth larva. (Supported by an NSERC Canada)

6. Assessing the role of endopolygalacturonase in the pathogenicity of white-rot basidiomycete Chrondrostereum purpureum.

The white-rot basidiomycete Chondrostereum purpureum is being developed as a biological control agent for the suppression of unwanted hardwood species in forest renewal sites. We are investigating the role of endopolygalacturonase I (endoPG) in the pathogenicity of C. purpureum by the overexpression and/or transformation mediated disruption of the endoPG gene. EndoPG gene sequences from fungi and plants were aligned and regions of sequence homology identified in order to isolate an endoPG gene fragment from C. purpureum by PCR amplification. A putative endoPG gene fragment (214 bp) was amplified which shared significant similarities with other known endoPG gene sequences. An EMBL-3 lamda DNA Iibrary is being screened to select the full-length gene. Overexpression and disruption of endoPG requires an efficient transformation system for C. purpureum. Transformation experiments with dominant selectable marker driven by promoters derived from ascomycetes (pAN-7 and pPS-57) have been unsuccessful. Transformation with the phleomycin resistence marker gene driven by gpd promoter of the basidiomycete Schizophyllum commune (F.H.J. Schuren, pers. com.) has apparently been successful. Analysis of the putative transformants is in process. The gpd promoter will be switched with the C. purpureum endoPG promoter in order to further increase transformation efficiency.

7. Isolation of an alfalfa gene induced by Colletotrichum trifolii during a compatible interaction using differential display.

We are interested in understanding how plants and their pathogens communicate at the molecular level. During a compatible interaction between a phytopathogenic fungus and its host, specific genes are induced or repressed in both organisms. Such changes in gene expression may facilitate disease development. To identify genes in either alfalfa or Colletotrichum trifolii whose regulation is altered at the time of infection, differential display was performed. Populations of mRNA expressed in alfalfa, C. trifolii and C. trifolii infected alfalfa leaves were reverse transcribed using four sets of anchored oligo(dT) primers that contain two additional bases, M and N, at their 3' ends. The resulting cDNA populations were amplified by PCR using a set of arbitrary decamers and radioactive nucleotide, and the PCR products were separated by PAGE. Numerous differentially-amplified products were isolated. Northern and Southern analysis identified one product as an alfalfa gene whose expression is induced during infection. Characterization of this gene will be discussed.

8. Genetic analysis of cultivar specificity and race evolution in the soybean pathogen, Phytophthora sojae.

There are 12 major resistance (Rps) genes in soybean against the oomycete pathogen P. sojae, and 37 races of the pathogen. To test whether avirulence against these Rps genes is controlled by single dominant genes, we crossed three isolates of P. sojae. Since P. sojae is homothallic we used RAPDs to identify F1 hybrids from mixed cultures. The F1 hybrids were then selfed to produce F2 progeny. Avirulence was dominant or semi-dominant in the F1 progeny for all 10 Rps genes tested. RFLP and RAPD markers segregated in regular Mendelian fashion among the F2 progeny of one cross, but not of a second cross. In the first cross, avirulence against the six Rps genes tested (Rps1a, Rps1b, Rps3a, Rps3c, Rps4 and Rps6) segregated as a single Mendelian trait. Several sets of linked RFLP markers were identified including a RFLPs linked to Avr(Rps1b). Analysis of the distribution of RFLP markers and avirulence phenotypes among 48 field isolates encompassing 25 race types indicated that new races have arisen in this pathogen both by mutations (presumably in avirulence genes) and by reassortment of avirulence genes following rare outcrosses.

9. Fusarium oxysporum from cyclamens.

Seventy-nine Fusarium isolates were recovered from cyclamen plant material and solicited from culture collections. These isolates were categorized based upon pathogenicity towards cyclamen, vegetative compatibility tests, hybridization with a repetitive DNA (fingerprint) probe, and the organization of the ribosomal intergenic spacer (IGS) sequences. Fifty-three pathogenic and 26 nonpathogenic isolates were identified. The pathogenic isolates could be subdivided into three groups. Each of these groups was limited to a single vegetative compatibility group (VCG) and all members had similar IGS organization and DNA fingerprints. The nonpathogenic isolates could be distinguished from the pathogenic isolates, and usually one another, using either the VCG or the fingerprint criteria, but not necessarily the IGS criterion.

10. A polyketide synthase (PKS) is required for fungal virulence toward Texas male sterile corn.

In 1970, a new race (T) of the fungal pathogen, Cochlobolus heterostrophus appeared. Race T is highly virulent toward Texas male sterile (T) corn and differs from its progenitor, race O, at a locus (TOX1) which is responsible for production of T-toxin, a polyketide. TOX1 is associated with a complex genetic rearrangement involving a reciprocal translocation, AT-rich repeats, and an insertion. We used REMI to mutagenize and tag TOX1. Sequencing of the DNA at the insertion site and translation of a 7.8 kb ORF revealed a multifunctional PKS encoding gene with six enzymatic domains: ketoacyl-ACP synthase (KS), acyl transferase (AT), dehydrogenase (DH), enoyl reductase (ER), keto-reductase (KR), and acyl carrier protein (ACP), all with conserved motifs. The KS motif has the highest homology with other polyketide synthases. The PKS appears to be organized as a single module surrounded by highly repetitive, AT-rich DNA. Non-toxin-producing race O strains lack this gene. When the PKS-encoding gene was disrupted in race T, T toxin production and high virulence were eliminated. These results reveal, for the first time, that a PKS is required for virulence and provide a firm foundation for investigation of how a new pathogenic race arises.

11. The role of PM-toxin in pathogenesis by Mycosphaerella zeae-maydis.

M. zeae-maydis is a homothallic Ascomycete taxonomically unconnected to Cochliobolus heterostrophus race T. Yet both fungi are highly virulent on corn containing Texas male sterile (T) cytoplasm (but cause little damage to N-cytoplasm corn) and each produces a linear polyketide which specifically affects mitochondria of T-cytoplasm corn. Genetic analyses have shown that the C. heterostrophus race T polyketide (T-toxin) is required for high virulence to T-cytoplasm corn. To determine the role of the M. zeae-maydis polyketide (PM-toxin) in pathogenesis, the restriction enzyme mediated integration (REMI) procedure was used to produce Tox mutants via transformation of fungal protoplasts [prepared by digesting mycelium with Driselase and cellulase (each 10 mg/ml) in 0.7 M KCL and 0.2 M CaCl2, pH 6]. From 504 mitotically stable transformants (selected for resistance to hygromycin B, blasticidin S, or benomyl), five Tox mutants were obtained, which when selfed, produced F1 and F2 progenies that were 100% Tox . These mutants were identical to wild type in all respects except that they failed to produce PM-toxin. Two of them were inoculated on T- and N-cytoplasm corn; both had drastically reduced virulence, indicating that the M. zeae-maydis polyketide, like the one from C. heterostrophus, is required for pathogenesis.

12. The host range of a plant pathogenic fungus is determined by saponin detoxification.

Saponins are glycosylated steroidal compounds which occur in many plant species. The toxic effects of saponins are attributed to their ability to form complexes with membrane sterols resulting in loss of membrane integrity. Some pathogenic fungi have intrinsic resistance to these compounds as their membranes contain no sterols, whereas others produce enzymes which can enzymatically detoxify particular saponins. This implies that at least for some interactions, "saponin-saponinase" combinations may determine host range. A gene encoding a saponin-detoxifying enzyme was cloned from the cereal infecting fungus, Gaeumannomyces graminis. Fungal mutants generated by targeted gene disruption were no longer able to infect the saponin-containing host oats but retained full pathogenicity to wheat (which does not contain saponins). Thus the ability of a plant pathogenic fungus to detoxify a plant saponin can determine its host range.

13. Avenacinase-like proteins in fungi from the Gaeumannomyces-Phialophora complex.

Gaeumannomyces is a genus of ascomycete fungi with four known species that infect roots of grasses, cereals, or sedges. Take-all, caused by Gaeumannomyces graminis, is the most damaging root disease of wheat world-wide, and is among the most important cereal diseases in the United Kingdom. G. graminis varieties tritici, avenae, and graminis have Phialophora-like anamorphs and, together with other non-pathogenic Gaeumannomyces and Phialophora species found on cereal roots, constitute the Gaeumannomyces-Phialophora complex. cDNA encoding the saponin detoxifying enzyme avenacinase from G. graminis var. avenae, cross hybridises with DNA from other varieties of G. graminis. In addition, avenacinase antisera cross-reacts with a secreted protein with a similar molecular weight to avenacinase from these same G. graminis isolates. We have shown that these avenacinase-like proteins (ALPs) have very similar physiochemical and immunological properties to avenacinase, but unlike avenacinase, they cannot effectively deglucosylate avenacin. Cloning and DNA sequence comparisons of ALPs with avenacinase and another saponin detoxifying enzyme, tomatinase (from Septoria lycopersici) will be presented. Gene disruption experiments are in progress, and the effects of ALP gene disruption on host range and pathogenicity will be assessed.

14. Saponin detoxification by plant pathogenic fungi.

Saponins occur in many plant species, and because of their antifungal properties they have been implicated as pre-formed determinants of resistance to fungal attack. Some fungi produce enzymes which remove sugars from saponins, to give molecules which are less fungitoxic. Mutants of the cereal-infecting fungus Gaeumannomyces graminis var. avenae which do not produce the saponin glucosyl hydrolase avenacinase can no longer infect the saponin-containing host oats (but are still fully pathogenic to wheat, which does not contain saponins) (see accompanying poster by Bowyer et al). Southern blots using avenacinase cDNA as a probe revealed cross-hybridising DNA sequences in a number of other phytopathogenic fungi, suggesting that enzymes related to avenacinase may be widespread. We have demonstrated for one of these fungi (the tomato pathogen, Septoria lycopersici), that the cross-hybridising DNA in this fungus also encodes a saponin detoxifying enzyme (in this case tomatinase). While avenacinase and tomatinase are clearly related, the relative activities of these enzymes towards avenacin and tomatine reflect the host specificity of the fungi from which they originate. Structure/function analysis of these two highly conserved saponin glucosyl hydrolases should allow us to identify the regions of the enzymes which are important for activity and for substrate specificity, and to design inhibitors of enzyme action which may have significance for crop protection strategies. The occurrence of DNA sequences which hybridise to avenacinase cDNA in genomic DNA of other phytopathogenic fungi suggests that saponin-saponinase combinations may be more important in determining host range than has previously been appreciated.

15. Mapping of avirulence genes in the rust fungus, Puccinia graminis.

Extensive genetic and physiological studies of cereals rust diseases have provided the foundation for many of our current concepts about host-parasite interactions. However, we know very little about the molecular biology of rust, in part due to the obligate nature of these pathogens. As a model system for rust fungi we have chosen to study the wheat stem rust fungus, Puccinia graminis f.sp. tritici. A genetic mapping population has been generated by crossing two North American isolates and selfing a single F1 progeny. In this F2 population we have selected ten avirulence genes (avr6, avr8a, avr9a, avr9d, avr10, avr21, avr28, avr30, avrflk, and avrU) for further study. Seven of these genes segregate as single dominant genes, while one (avr10) appears to be co-dominant and one (avr9d) recessive. Avrflk is segregating with a 13:3 ratio, indicating that two genes may be involved. We are currently using RAPDs and bulked segregant analysis to identify genetic markers closely linked to these genes in order to isolate them by map-based cloning.

16. 1,3-beta-glucanase and xylanase from Phytophthora parasitica: enzyme characterization and gene cloning.

Phytophthora parasitica produces a number of extracellular polysaccharide-hydrolyzing enzymes. Some of these enzymes are putatively involved in pathogenesis: xylanase (XYL) is believed to facilitate the penetration of the cell wall of host plants and 1,3-beta-glucanase (GLU) is assumed to participate in overcoming a callose barrier built by the host in a defence reaction against the pathogen. Zymogram analysis of isoelectric focusing gels revealed two forms of GLU with pI around 5.0 and four forms of XYL with pI around 9.0. GLU isoenzymes were further resolved into four bands on native PAGE gels. Two forms of XYL with Mrs of 50 kD and 70 kD were identified by activity staining of SDS gels after renaturation. Xylanase from culture filtrate was partially purified by ion exchange chromatography. A cDNA library of P. parasitica was constructed in lambda ZAP and screened for activity of fusion proteins (Karlovsky, P. and Wolf, G.A. (1993) Meth. Mol. Cell. Biol. 4:40-45). Clones encoding GLU and XYL were isolated, converted into plasmids and expressed in E. coli. Sequences of protein products deduced from nucleotide sequences of cDNA clones were compared with sequences of known GLU and XYL enzymes.

17. Rapid simultaneous identification of virulence genes.

Transposon mutagenesis has been used to identify bacterial virulence genes by testing individual mutants for altered virulence in living plants and animals, but comprehensive screening of genomes for virulence genes has not been possible because of the inability to identify mutants with attenuated virulence within pools of mutagenized bacteria, and the impracticability of separately assessing the virulence of each one of the few thousand mutants necessary to screen a bacterial genome. To circumvent this problem, we developed a novel transposon mutagenesis system in which each mutant is tagged with a unique DNA sequence. We used Salmonella typhimurium to illustrate the method because it has excellent molecular genetics, and the disease it causes in mice closely resembles typhoid fever. We show that it is possible to identify mutants with reduced virulence simultaneously, by comparing the identity of bacterial cells recovered from spleens of animals with those of a complex pool of mutant cells representing the inoculum. Screening of 1000 mutants resulted in the identification of thirty genes affecting virulence. These were either previously identified S. typhimurium virulence genes, homologues of known genes (including virulence genes) of S. typhimurium and other bacteria, or genes without similarity to entries in DNA databases. The screening system should permit the rapid isolation of virulence genes of other bacterial pathogens, and may be applicable to fungal pathogens, using the restriction enzyme mediated integration (REMI) procedure to generate random insertional mutations.

18. Lineage structure, avirulence locus polymorphism and the organization of pathotype diversity in the rice blast fungus.

The rice blast fungus, Pyricularia grisea (Magnaporthe grisea), exhibits high levels of pathotype polymorphism that have handicapped efforts to breed durably resistant rice cultivars. Homologous- DNA fingerprints indicate that pathogen diversity is typically organized in historically distinct genetic lineages. Each lineage has a limited cultivar range although most express multiple pathotypes that are varied combinations of lineage-wide compatibilities. Lineages are thus marked as expressing particular, historically conserved avirulences. Complementary evidence of this organization is now provided by lineage-specific RFLPs associated with P. grisea avirulence gene probes and by wholesale lineage incompatibility with single gene resistances in near-isogenic testers. Based on these features we have identified and are evaluating particular resistance gene combinations that should exclude all of the resident lineages (based on their current pathotype repertoire) in various rice-growing regions.

19. Analysis of RAPDs by graphic display reveals genetic relationships among Australian and American genotypes of bean rust.

Over 200 race phenotypes of the bean rust fungus Uromyces appendiculatus (Pers.) Unger var. appendiculatus, have been recorded worldwide. This study assessed genetic diversity of bean rust in Australia in relation to the Americas. Initially, phenetic analysis of 12 representative Australian isolates using RFLPs (10 cDNA probes) and RAPDs (10 decanucleotide PCR primers) revealed three clusters of isolates designated A, B, and AB. Collation of polymorphic bands characteristic of each cluster ("Graphic Display") showed that AB exhibited most bands characteristic of A and B, but gave no characteristic bands of its own, suggesting that AB isolates arose from hybridization between genotypes A and B. Further analysis using RAPDs (10 primers) compared 5 representative American isolates with an extended set of 42 Australian isolates. One American isolate clustered with the Australian isolates of genotype B, while the other four American isolates formed a separate cluster (C). Although genotypes A and B were phenetically closer to each other than genotype C, Graphic Display revealed a group of polymorphic RAPD markers common to genotypes A and C but absent from B, consistent with A being derived by recombination between progenitors of B and C. Results are discussed in relation to the derivation of current genotypes from disparate centers of origin in the Americas.

20. Molecular analysis of pathogenicity genes from the plant pathogenic fungus Glomerella cingulata.

Glomerella cingulata causes disease on a wide variety of fruit crops. The maceration of fruit tissue is caused by the action of an array of pectinolytic and other hydrolytic enzymes secreted by the pathogen. We are using gene disruption as a tool for determining the importance of these enzymes in pathogenicity. Oligonucleotides were designed to conserved amino acid domains from pectin and pectate lyases and used to amplify G. cingulata genomic DNA. Three clones with homology to pectin lyase and one with homology to pectate lyase were identified. One of the pectin lyases designated pnlA was cloned and sequenced. Two gene-disruption vectors were constructed, a replacement and a truncation-disruption, and used to inactivate the pnlA locus. Only transformation with the truncation-disruption vector gave single-copy integrations at the desired locus. Disruption of pnlA was shown using Southern analysis. The pnlA transcript was not detected on Northern gels and PNLA could not be detected by activity staining IEF gels. Although PNLA was the most active isozyme secreted in vitro, the disruption of the pnlA locus had no effect on pathogenicity of G. cingulata on apple or capsicum. Progress on the analysis of other pathogenicity factors such as the secreted aspartic protease was also presented.

21. Host-associated genetic differentiation in Erysiphe cichoracearum: a mixture of diffuse cospeciation and colonization.

The powdery mildew species, Erysiphe cichoracearum, is reported to infect more than 300 hosts from among 8 plant families. Evidence suggests, however, that the species is a complex of host-specialized biotypes. Our analyses of this species with RFLP's of amplified ribosomal DNA (rDNA) segments demonstrate the presence of several, distinct rDNA haplotypes among biotypes of E. cichoracearum. Samples with identical rDNA haplotypes were always collected from the same host family. However, multiple, distinct haplotypes were recovered from sunflower family biotypes. rDNA haplotypes differences also clearly distinguished samples from the morphologically similar species, E. galeopsidis, and from the mildew genus Sphaerotheca. Comparative phylogenetic analyses suggest that host associations of these obligate parasites are derived from a mixture of diffuse cospeciations at the host familial level and colonizations of different host families. Thus, there is only limited support for Fahrenholz's rule, that parasite-host associations evolve by cospeciation in a species-for-species manner. However, the specificity of individual rDNA types suggests that related hosts share basic compatibility with only a subset of these powdery mildews. The evolutionary consequences of interactions between trophic dependence and dispersal in fungal plant pathogens will also be discussed.

22. The role of polygalacturonases in the interaction between Penicillium olsonii and Arabidopsis thaliana

The possible involvement of fungal pectin degrading enzymes in the pathogenicity of Penicillium olsonii to Arabidopsis thaliana is examined. Two fungal polygalacturonases can be induced with pectin as sole carbon source in vitro. Both enzymes are also produced by the fungus in vivo during plant infection. From known sequences of fungal PGs, oligonucleotides have been constructed which lead to the amplification of a 586 bp fragment by PCR. This PCR fragment was cloned into a vector for transformation mediated direct gene disruption. Of 80 transformants, 4 were reduced in PG activity. Southern hybridization and enzyme analysis demonstrated the disruption of one PG gene. The vector also contained the GUS gene under the control of a constitutively transcribed fungal promotor. GUS- positive transformants with unchanged PG activity compared with PG-reduced mutants showed no difference in colonizing leaves during the first six days of infection. Investigations of later infection stages and the cloning of the second PG gene to construct a PG-deficient Penicillium are planned.

23. Cutinase is produced during infection of pea by Nectria haematococca but has no influence on virulence.

A new infection assay was established for the pathogen-plant system Nectria haematococca garden pea. Intact plant seedlings were inoculated with fungal mycelium and the infection process was assayed over a time period of 14 days. A new transformation system was established and a cutinase- deficient transformant was transformed to express the beta-glucuronidase gene. Fungal growth and vitality of cutinase-deficient and cutinase-producing strains was monitored on infected plants by measuring beta-glucuronidase production. Additionally, general esterase and cutinase activity was analysed during the infection process. Cutinase-deficient and cutinase-producing strains generated the same infection symptoms and comparable beta-glucuronidase activities. Cutinase activity was detectable in epicotyls as well as roots infected with Nectria haematococca wild type. No cutinase activity was observed during plant infection with the cutinase-deficient mutant. For the interaction of N. haematococca with pea, we draw the following conclusions: 1. Cutinase is neither essential for pathogenicity nor involved in virulence. 2. It is a rather unspecific enzyme that is found also during root colonization. 3. No additional fungal cutinase is detectable during plant infection.

24. Regulation of the nor-1 gene involved in aflatoxin biosynthesis in Aspergillus parasiticus.

The phytopathogenic fungus, Aspergillus parasiticus, produces the toxic secondary metabolite aflatoxin. Colonization of peanuts, cotton, corn and other crops by the fungus may result in aflatoxin contaminated food and feed supplies. The first stable intermediate in the aflatoxin biosynthetic pathway is the decaketide-derived norsolorinic acid, which is converted to averufin via one or more proposed alternative pathways. One of the enzymes involved in this conversion is the nor-1 gene. Coordinate expression of the nor-1 gene and several other genes associated with aflatoxin biosynthesis indicates that these genes are similarly regulated. Analysis of the promoter regions of these genes should provide some insight into the mechanism of this regulation. Proteins were isolated from nuclei of A. parasiticus grown for 48 hrs in either aflatoxin inducing (GMS) or noninducing (PMS) medium and Mobility Shift DNA Binding assays were performed on a 224bp region upstream of the transcript initiation site. Two sites were identified that bound proteins unique to the GMS cultures. An understanding of the biosynthesis and regulation of aflatoxin production will be useful in developing novel control methods for this pathogen and in clarifying the role of aflatoxin in the biology and pathogenicity of the fungus.

25. A central role of salicylic acid in plant disease resistance.

Transgenic tobacco and Arabidopsis thaliana expressing the bacterial enzyme salicylate hydroxylase cannot accumulate salicylic acid (SA). This defect not only makes the plants unable to induce systemic acquired resistance, but also leads to increased susceptibility to viral, fungal and bacterial pathogens. The enhanced susceptibility extends even to host-pathogen combinations that would normally result in genetic resistance. Therefore, SA accumulation is essential for the expression of multiple modes of plant disease resistance.

26. Genetics of virulence in Phytophthora sojae.

Phytophthora sojae belongs to the Oomycetes which are characterized by gametangial meiosis, thus having a diploid somatic phase, which contrasts with the majority of the true fungi which are haploid for most of their life cycle. Until recently the genetics of virulence/avirulence in P. sojae was considered intractable due to its homothallic nature. A race 1 and a race 7 isolate were co-cultured in vitro and, using RAPD markers, ten hybrids were identified among 354 oospores analyzed. One F1 hybrid was allowed to self fertilize and produce an F2 population of 247 individuals. Fifty-three F2 individuals were selected at random for genetic analysis. A genetic linkage map has been constructed from this cross consisting of 15 major linkage groups and ten small linkage groups using 233 RAPD markers, 30 dominant RFLP markers, 10 co-dominant RFLP markers and four avirulence genes. Segregation of virulence against soybean resistance genes Rps1a, 3a and 5 revealed that the avirulence genes Avr1a, 3a and 5 were dominant to virulence. Avirulence against these three resistance genes appeared to be conditioned by one locus for Avr1a and two independent, complementary dominant loci for both Avr3a and Avr5. Segregation of virulence against Rps6 was in the ratio of 1:2:1 (avirulent : mixed reaction : virulent), suggesting a semi-dominant allele at a single locus.

27. Extraordinary heterogeneity in the ribosomal non-transcribed spacer region of Acremonium.

Acremonium species are a group of asexual filamentous Ascomycetes that form mutualistic associations with grasses of the Pooideae sub-family (Scott and Schardl, 1993). The ubiquitous pasture grass Lolium perenne (perennial ryegrass) forms associations with Acremonium spp. from two taxonomic groupings: LpTG-1 (=A. lolii) and LpTG-2. Recent work has shown that isolates of LpTG- l are haploid whereas isolates from LpTG-2 are heteroploid, and probably derived from inter specific hybridization event(s) between the ryegrass choke pathogen, Epichloe typhina, and LpTG-l (Schardl et al . 1994). Characterization of the ribosomal region (rDNA) of an LpTG-2 Acremonium isolate revealed an extraordinary amount of length heterogeneity in the non-transcribed spacer (NTS) region. Single spore purification showed that this heterogeneity is intragenomic. The NTS pattern is largely conserved through single sporing, but some distinct length differences were observed. The extreme level of heterogeneity found in the NTS region is not found in the two putative parental species - LpTG- 1 and E. typhina, and it appears to be specific to this region of the genome. Sequencing has shown that the NTS contains two closely-related sub-repeats, one 107bp in length and the other 120bp in length. These repeats are arranged tandemly with the organization of the two repeat classes showing no obvious pattern. We propose that the heterogeneity found in the LpTG-2 NTS region is the result of an unequal recombination mechanism, possibly in the register of these sub-NTS repeats. The appearance of this heterogeneity apparently as a consequence of hybridization has implications for the study of control of homogeneity in tandemly repeated genes.

28. Disruption of Mak1 and structural analysis of the Mak1 containing chromosome in Nectria haematococca.

Nectria haematococca is a pathogen on a variety of host plants. We are studying its interaction with chickpea (Cicer arietinum). Our interests are focused on determining: 1) If Mak1, a gene that detoxifies the chickpea phytoalexin maackiain, is important for pathogenicity on chickpea; and 2) If there are hotspots for chromosome breakage during meiosis on the B-chromosome that contains Mak1. Mak1 was recently cloned from a chromosome-specific cosmid library. We replaced an internal 400bp fragment of Mak1 with a hygromycin B expression cassette and are using this construct to create a transformation mediated gene disruption in N. haematococca. The pathogenicity of transformants with a successful gene replacement will be tested in a virulence assay on chickpea. In previous genetic crosses it was found that in isolates that lost part of the Mak1 chromosome, the remaining fragments of the chromosome were most often 0.9-1.0 megabases in size (Miao et al. 1991 Science 254:1773- 1776). This result suggests that specific regions of the chromosome are particularly susceptible to breakage during meiosis. We are currently creating a contiguous cosmid map of the Mak1 chromosome using a procedure called sampling without replacement. The chromosome-specific library will be probed with a variety of deleted forms of the Mak1 chromosome to identify the breakage points generated during meiosis. Progress towards these goals will be presented.

29. A preliminary intraspecific phylogeny of Fusarium oxysporum based on the partial sequence of the intergenic spacer (IGS) region of the rDNA.

Using PCR, we amplified and sequenced 1,000bp of the 5' end of the intergenic spacer (IGS) in 15 isolates of Fusarium oxysporum and one isolate of F. subglutinans. The IGS region was chosen for sequence analysis because RFLPs revealed variation within F. oxysporum. Isolates were selected based on the genetic markers, virulence, race, vegetative compatibility group (VCG), mitochondrial DNA (mtDNA) haplotypes, IGS haplotypes, and DNA fingerprint, to represent the diversity in our collection. The objective of this research was to clarify the origin of virulence within F. oxysporum, the evolution of the different races of F. o. melonis, and the relationship between pathogenic and nonpathogenic strains. Parsimony analysis of the partial IGS sequence data identified a phylogenetic tree with highly significant branches. The two F. o. melonis VCGs, 0131 and 0134, were separated into distinct lineages closely related to F. o. cubense. Race was not distinguished by IGS sequence differences within the pathogen VCGs, except for one isolate. This race 1 isolate was first thought to be a new race in VCG 0131, but it was also associated with the same mtDNA and IGS haplotypes, and DNA fingerprint as isolates in VCG 0134. Two IGS sequence types were found in this isolate, representing both VCG 0131 and 0134, suggesting a somatic interaction or recombination event occurred between F. o. melonis, VCG 0131 and 0134. Nonpathogens that were vegetatively compatible with the pathogen, were not closely related to the pathogen and do not appear to be potential sources of new pathogenic races.

30. Tri10: A new gene in the trichothecene gene cluster in Fusarium sporotrichioides.

Fusarium sporotrichioides produces a spectrum of trichothecenes with the primary metabolite being T-2 toxin. Genetically, the synthesis of these sequiterpenoid mycotoxins is determined, in part, by pathway genes that are organized in a large gene cluster. Preliminary DNA sequence information indicated the presence of two small open reading frames, ORF-1 and ORF-2, downstream from the Tri5 gene which encodes the pathway enzyme, trichodiene synthase. Northern analyses conducted with DNA probes within and/or overlapping these two ORFs demonstrated that each probe hybridized to a transcript approximately 1400 bases in length. A pair of primers within ORF-2 has also yielded a PCR product of identical size from both genomic DNA and DNA prepared from a cDNA library. In addition, the pattern of expression of this transcript parallels that of the other known pathway genes located in the trichothecene gene cluster. These results indicate that we have identified a new gene, Tri10, within the trichothecene gene cluster. Experiments are underway to determine the complete sequence of the Tri10 cDNA and to investigate the function of this gene via gene disruption.

31. Pneumocystis carinii strain and species dynamics in natural and induced infections monitored by electrophoretic karyotyping.

Pneumocystis are a group of eukaryotic organisms that cause a lethal pneumonia in immunocompromised mammalian hosts. Gene sequence data show they are phylogenetically related to fungi. Two putative species of rat-derived Pneumocystis carinii, P. carinii carinii and P. carinii rattus (previously called prototype and variant) were defined by several molecular genetic criteria. At least 8 forms of P. c. carinii could be distinguished on the basis of electrophoretic karyotype patterns. Although P. c. carinii and P. c. rattus were found as co-infections within rat lungs, co- infections with 2 forms of P. c. carinii have not been observed. Monitoring of the natural infection in commercial rat colonies by electrophoretic karyotyping, hybridization with repetitive and single gene probes, and by gene sequencing showed that a single karyotype form of P. c. carinii could remain stable in a colony 3 years or more if housing conditions were unperturbed. In colonies where rats with co-infections of the 2 species were introduced, the numbers of rats harboring co-infections were variable. Direct co-inoculation of mixtures of equal numbers of 2 defined P. c. carinii forms into immunosuppressed rats produced infections of only a single karyotype form. A delay of 10 days or 20 days between inoculations resulted in infections bearing the karyotype of the form first inoculated and in some cases, a significant reduction in organism burden. These results indicate that not all forms of P. c. carinii are compatible in vivo and suggest that mechanisms of interference may be operational.

32. MPG1 hydrophobin and development in Magnaporthe grisea.

Upon germination on a hydrophobic substrate, conidia of the rice blast fungus (Magnaporthe grisea) form infection structures called appressoria that allow direct penetration of plant cells. We are examining expression of a hydrophobin of M. grisea named MPG1 (Magnaporthe Pathogenicity Gene 1). MPG1 was so named because mutant strains are no longer able to efficiently form appressoria and are therefore less pathogenic than wild-type. We hypothesize that this protein plays a role in recognition of surface hydrophobicity. Expression of this gene is observed during the early stages of infection of rice plants and in aerial hyphae and conidiophores. It has been reported that nutrients do not affect development of appressoria. However, we have found that nutrient conditions that repress MPG1 expression in liquid culture also strongly inhibit appressorium formation on hydrophobic surfaces. Exogenous application of cAMP results in restoration of appressorium formation and induces MPG1 expression in otherwise repressing growth medium. The data suggest a relationship between nutritional status, hydrophobic surface recognition, and cAMP signalling in appressorium formation.

33. Electrophoretic karyotypic variation in the genus Pythium.

Significant variation is observed in electrophoretic karyotypes in the genus Pythium. While isolates of a single species tend to have a similar distribution of chromosome sizes, differences in the number of chromosomal bands and their individual sizes were observed. While some morphologically similar species share a common range in size distribution, this was not observed for all species comparisons made. One mechanism which appears to contribute to intraspecific polymorphism in karyotypes is meiotic instability, however, depending on the species investigated different types of polymorphisms were generated. For the homothallic species P. oligandrum, the only polymorphism detected was the loss of a putative supernumerary chromosome in 3.5% of the progeny. While karyotypes of progeny of the homothallic species P. spinosum were nearly identical to the parental isolate, 60% of the isolates were polymorphic for 1-2 chromosomal bands. In contrast to the results with the homothallic species, significant levels of meiotic instability were detected in the heterothallic species P. sylvaticum; in total, 80% of the progeny chromosomal bands were nonparental in size or location of specific coding regions. The contributions of mitotic instability on the generation of karyotypic polymorphisms also will be discussed.

34. Glume blotch of wheat: Adaptation of the pathogen to varieties with quantitative resistance.

The fungal pathogen Stagonospora nodorum (syn. Septoria nodorum) causes glume blotch of wheat. The aim of our research is to investigate variation in field populations of S. nodorum and, in particular, to determine how this variation is distributed among wheat host varieties differing in their resistance to the pathogen. RFLP markers were used to measure the amount and distribution of genetic variation in nine Swiss pathogen populations. Nine hundred isolates of S. nodorum were collected from three fields each of three varieties. DNA variation was assayed with ten probes that hybridized to single RFLP loci, one probe that showed a DNA fingerprint and one mtDNA probe. A maximum amount of genotypic diversity was found based on the DNA fingerprint. However, single locus analysis of RFLP data indicated that allele frequencies were generally very similar in all populations, including two additional populations from Texas and Oregon. There was no evidence for gametic disequilibrium in any of the populations. The results suggest a high level of sexual recombination and that the RFLP markers were selectively neutral. On an intercontinental level, the calculated gene flow (Nm>11) was strong enough to counteract genetic drift and to prevent differentiation among populations.

35. The global population genetic structure of the wheat pathogen Mycosphaerella graminicola.

We have analyzed the genetic structure of populations of the pathogenic fungus Mycosphaerella graminicola (anamorph Septoria tritici) from wheat fields around the world. Anonymous probes that hybridized to individual RFLP loci were used to measure gene diversity, genetic identity, and evaluate the potential for gene flow among populations. Populations of the fungus are remarkably similar worldwide. Genetic diversity is distributed on a small scale in each population. All populations have a large number of different genotypes and individual RFLP loci are at gametic equilibrium, properties expected in random-mating populations. All populations share alleles at RFLP loci, often at similar frequencies. These results suggest the potential for a high degree of gene flow among populations. Yet, DNA fingerprints showed that no clones were shared among populations, so there is not a predominant clone or set of clonal lineages worldwide as has been found for other plant pathogens such as Magnaporthe grisea and Phytophthora infestans. It appears that sexual reproduction and gene flow are the dominant factors affecting the genetic structure of populations of M. graminicola worldwide.

36. Indication of population subdivision within the sorghum anthracnose fungus, Colletotrichum graminicola.

Two populations of Colletotrichum graminicola, one from a commercial sorghum field in South Texas, the other from the sorghum disease nursery in Griffin, Georgia were analyzed for their genetic structure using seven anonymous DNA probes which hybridized to low copy fragments in each isolate. Both populations were characterized by a low number of alleles at each locus and also by a low number of genotypes retrieved from each location, indicating that this fungus exhibits much less genetic variation than previously assumed based on virulence studies. It appears that populations of this fungus are largely clonal in sorghum populations. Upon comparison of the two populations, private alleles were found for some RFLP loci in the two populations. Shared alleles were found in both populations for other RFLP loci, but in these cases allele frequencies were often significantly different. Calculated values for Nm for most probes were less than one, indicating that gene flow is not one of the forces influencing the evolution of this fungus, at least for these two populations.

37. RAPD isolation of fingerprint probes for analysis of Colletotrichum species.

We have been analyzing genetic diversity of isolates of Colletotrichum gloeosporioides and C. acutatum from diverse plant hosts. The availability of reliable and informative fingerprint probes would greatly facilitate analysis of phylogenetic relationships between populations of these fungi. We have used hybridization of labelled genomic DNA to screen a battery of RAPD products to identify repetitive elements from other taxa. RAPD products identified are then used to probe Southern blots of DNA from various isolats and fingerprint patterns compared. Our analyses indicate a dearth of heterodispersed, repetitive elements in these two species of Colletotrichum relative to some other plant pathogenic genera. However, different RAPD products may be readily assembled to generate useful fingerprint probes.

38. Nrs1, a repetitive element linked to pisatin demethylase genes on dispensable chromosomes of Nectria haematococca.

We have identified a repetitive DNA sequence in T-2, an isolate of Nectria haematococca mating population VI. This repetitive family has been called Nrsl. The 2027 bp clone of the Nrsl-2 allele contains a long polyA sequence, imperfect RNA polymerase III promoter sequences, multiple inverted and direct repeats, and the potential for extensive secondary structure similar to known polymerase III transcripts and related retroelements. Ten of the 11 HindIII restriction fragments from T-2 that hybridize to Nrsl-2 segregate in a manner consistent with a 1:1 ratio for random ascospore progeny. The ten RFLP loci define three linkage groups and correspond to three chromosomes from T-2 as separated by pulse field gel electrophoresis. Three RFLP loci defined by hybridization to the gene for pisatin demethylase and located on a 1.6 Mb chromosome are genetically linked to each other and to several of the Nrsl loci. These sequences recombined despite the fact that no obvious homolog exits for the 1.6 Mb chromosome in one of the parent strains. DNA hybridization analysis, using Nrsl-2, shows homology to other Nectria isolates that are pathogenic to Pisum sativum.

39. Progress in cloning cultivar specificity gene AvrCO39 from Magnaporthe grisea

The cultivar specificity gene AvrCO39 was mapped between marker 5-10-F and cosegregating markers CH5-120H and 1.2H on chromosome 1 of Magnaporthe grisea. RecA-Mediated Achilles' cleavage was used to determine that the physical distance between markers 5-10-F and CH5-120H is 600kb. A chromosome walk was initiated towards AvrC039 from markers 1.2H and 5-10-F. Unfortunately, both walks were curtailed before reaching AvrCO39. While walking away from 1.2H, a region of repetitive DNA was encountered which contained, among other repeats, a copy of the grasshopper retrotransposon. Within this repetitive region, it was not possible to find a single copy probe to identify an overlapping cosmid. From the other direction, it was not possible to progress more than six steps beyond marker 5-10-F, despite screening cosmid clones from a second library of more than 20 genome equivalents. Presumably the overlapping cosmid contains a sequence that is "poisonous" to E. coli. A novel method was used to overcome both of these obstacles. Achilles' cleavage was directed to markers 43-2-H and 18-2-F which were identifed while walking. This resulted in the liberation of a 280kb restriction fragment containing AvrCO39. This fragment was resolved using CHEF gel electrophoresis, digested using BamHI and then subcloned into a plasmid vector. In this manner, markers were identified that cosegregate fully with AvrCO39. Conventional chromosome walking was resumed to assemble contigs that span the entire locus. Progress in cloning AvrCO39 was reported.

40. Detecting genetic variation in a fungal population using amplified frangment length polymorphism (AFLP).

AFLP (Amplified Fragment Length Polymorphism) analysis was originally developed by Keygene, Wageningen, as a tool for plant genome mapping. However, these new molecular markers have great potential for studying genetic variation in fungal populations. The main advantage of AFLPs is that variation can be found quickly and reliably, even in fungal species where other methods have failed to detect sufficient variation. The ascomycete Pyrenopeziza brassicae is the causal agent of light leaf spot of Bassica species. The population structure of this economically important pathogen is unknown. We present AFLP data on isolates from six geographical areas in the UK and discuss advantages of using AFLP analysis (as opposed to RAPDs, RELPs or fingerprinting methods) for population studies to answer questions on population substructure and evolution of host - parasite interactions.

41. Detection of terminal restriction fragment length polymorphisms among race 8 of Ustilago hordei and its inbreeding progeny.

Ustilago hordei (Pers.) Lagerth is the causal agent of covered smut on barley (Hordeum vulgare) and has a physiologic race structure that obeys the gene-for-gene model. Race 8 of U. hordei is pathogenic only on the universal susceptible cultivar, Odessa. In a previous study, inbreeding of race 8 increased its pathogenicity in a stepwise progression and was speculated to have given rise to a new race, race 14, pathogenic on all differential cultivars (Phytopath. 69:1207-1212). Inbreeding of race 8 in our laboratory failed to reproduce this increase of pathogenicity and the progeny of successive selfed generations were pathogenic on Odessa only. The electrophoretic karyotypes (EK) of a race 8 strain, F1, and F2 inbreeding progeny from our laboratory and the previous study were resolved. All populations have a similar EK with the exception of the F2 progeny from the previous study, which had an additional chromosome ca. 1100 kb. Estimates of the number of chromosomes in the race 8 strain and the inbreeding progeny were obtained by hybridization of a telomere-specific probe to Southern blotted BamHI, BglII, and EcoRI digests from CHEF fractionated chromosomal bands. The F2 progeny from the previous study had 24 chromosomes while all other populations had 23 chromosomes. Hybridization of the telomere probe to BamHI, BglII, and EcoRI digests of total genomic DNA from all populations detected RFLPs only between race 8 and the F2 progeny from the previous study. The significance of these results with regard to the relatedness of race 8 and race 14 strains of U. hordei was discussed.

42. Molecular analysis of chromosome polymorphism in Ustilago hordei by using chromosome-specific probes and a telomere repeat as markers.

Variation in chromosome number and length have been detected by CHEF pulsed field gel electrophoresis (PFGE) in strains representing the 14 races of Ustilago hordei, the causal agent of covered smut of barley. Our objective was to determine the exact chromosome number for each race and to ascertain whether some strains are aneuploid using 2 approaches. Chromosome-specific libraries were constructed and 3 single copy probes from each library were hybridized onto Southern blots of CHEF gels containing chromosomes from strains of the 14 races. The homologous chromosomes identified were typically monosomic with length polymorphisms ranging to 100 kb, but cases of disomy were also observed in some strains. The second approach involved a telomere-specific repeat from Fusarium oxysporum, which was used as a hybridization probe onto Southern blots of restriction digests of individual chromosomes. Using this technique, 20 to 23 chromosomes were detected in 15 to 19 chromosome bands visualized by CHEF PFGE analysis in the strains representing the 14 races. The telomere-specific probe also generated specific RFLP patterns for each chromosome and allowed comparisons of homologous chromosomes. Each race of U. hordei showed a unique RFLP pattern for each of its chromosomes except for strains representing race 10 and 13. These strains had identical karyotypes and no RFLP differences.

43. Molecular approaches to distinguish physiologic "races" of Ustilago violacea.

Ustilago violacea is a heterobasidiomycete plant pathogen that infects over 70 species of Caryophyllaceae (Pinks). Circumstantial evidence has suggested that physiological races of U. violacea can be defined according to which host species are productively infected, as individual strains of the fungus tend to be only able to infect host species from which they are isolated and such host preference varies among different U. violacea isolates. Since development of the dikaryotic and spore forms of the fungus is obligately parasitic, the molecular bases for such "races" (if bona fide) and for their host preference are of interest. To date, there has been a paucity of genetic linkage data concerning U. violacea races. In the absence of such data, we have investigated several approaches to produce molecular "fingerprints" of U. violacea strains isolated from different host species and to further characterize the relatedness of such strains. Eighteen different sporidial strains representing 7 different races were examined for electrophoretic karyotype, RAPD-PCR profile, and by phylogenetic analysis of intron sequences in the gamma-tubulin gene. Comparison of electrophoretic karyotypes in conjunction with Southern hybridization with gamma-tubulin gene as a probe provided a measure for gauging strain relatedness. The combination of these methods with RAPD profiles identified isolates in a manner consistent with their anecdotal race designations.

44. Genetic variability in Sclerotinia trifoliorum.

S. trifoliorum is a pathogen of forage legumes with an interesting genetic system. Within each ascus there are four large and four small ascospores. The large spores give rise to homothallic (selfing) isolates. The small spores give rise to obligate heterothallic (outcrossing) isolates. A field study was conducted to deterrnine if outcrossing occurred in two infested alfalfa fields near Ithaca, New York. Outcrossing was assessed by using a mycelial incompatibility assay to examine genetic variability between sibling ascospore isolates from individual apothecia. In field #1, which was heavily infested during the previous spring and had relatively low crop density and low numbers of apothecia, 14 of 60 apothecia were outcrossed. In field #2, which was less heavily infested in the spring and had a higher crop density and abundant apothecia, one of 80 apothecia was outcrossed. We conclude outcrossing occurs and gives rise to genetic variability. We also assessed the levels of genetic variability within the field populations by testing single ascospore progeny from the homothallic (selfing) apothecia for mycelial incompatibility. We found that 42 of 46 isolates from field #1 were mycelially incompatible with all other isolates. From field #2, 75 of 79 isolates were mycelially incompatible with all other isolates. We conclude that S. trifoliorum has multiple incompatibility genes and that field populations are genetically diverse.

45. The PWL host-species specificity gene family in the rice blast fungus.

Poor understanding of the mechanisms underlying the specificity of interactions between fungal pathogens and plants has handicapped resistance breeding and other control strategies for these economically devastating diseases. Molecular cloning of fungal genes that determine the specificity of fungal-plant interactions now opens up an opportunity to study the biochemical and genetic basis of these important but elusive interactions. We have cloned and characterized members of the PWL gene family implicated in controlling host- species specificity in an emerging model system for fungal pathogenesis, Magnaporthe grisea. The rice blast fungus M. grisea is the cause of the major disease of this widely cultivated cereal grain. M. grisea also attacks diverse grass species, including millets, wheat, barley, and maize. Members of the PWL gene family are highly divergent and ubiquitous in nearly all host-specific forms of M. grisea. The PWL genes have a mode of action that is analogous to that of classical avirulence genes that determine cultivar specificity. The proteins encoded by PWL genes are predicted to be low molecular weight, glycine-rich, hydrophilic, and secreted. The working model is that the proteins encoded by PWL genes are secreted by M. grisea and directly trigger host defense response.

Return to the Asilomar 1995 page
Return to the Asilomar page
Return to the FGSC main page