101 Dothideomycetes genomes: A test case for predicting lifestyles and emergence of pathogens.

Dothideomycetes is the largest class of kingdom Fungi and comprises an incredible diversity of lifestyles, many of which have evolved multiple times. Plant pathogens represent a major ecological niche of the class Dothideomycetes and they are known to infect most major food crops and feedstocks for biomass and biofuel production. Studying the ecology and evolution of Dothideomycetes has significant implications for our fundamental understanding of fungal evolution, their adaptation to stress and host specificity, and practical implications with regard to the effects of climate change and on the food, feed, and livestock elements of the agro-economy. In this study, we present the first large-scale, whole-genome comparison of 101 Dothideomycetes introducing 55 newly sequenced species. The availability of whole-genome data produced a high-confidence phylogeny leading to reclassification of 25 organisms, provided a clearer picture of the relationships among the various families, and indicated that pathogenicity evolved multiple times within this class. We also identified gene family expansions and contractions across the Dothideomycetes phylogeny linked to ecological niches providing insights into genome evolution and adaptation across this group. Using machine-learning methods we classified fungi into lifestyle classes with >95 % accuracy and identified a small number of gene families that positively correlated with these distinctions. This can become a valuable tool for genome-based prediction of species lifestyle, especially for rarely seen and poorly studied species.

Development of a fungal transformation system based on selection of sequences with promoter activity.

A novel strategy was used to develop a transformation system for the plant pathogenic fungus Cochliobolus heterostrophus. Sequences capable of driving the expression of a gene conferring resistance to the antibiotic hygromycin B in C. heterostrophus were selected from a library of genomic DNA fragments and used, with the selectable marker, as the basis for transformation. The library of random 0.5- to 2.0-kilobase-pair fragments of C. heterostrophus genomic DNA was inserted at the 5' end of a truncated, promoterless Escherichia coli hygromycin B phosphotransferase gene (hygB) whose product confers resistance to hygromycin B. C. heterostrophus protoplasts were transformed with the library and selected for resistance. Resistant colonies arose at low frequency. Each colony contained a transformation vector stably integrated into chromosomal DNA. When the transforming DNA was recovered from the genome and introduced into C. heterostrophus, resistant colonies appeared at higher frequency. We determined the sequences of two of the C. heterostrophus DNA fragments which had been inserted at the 5' end of hygB in the promoter library and found that both made translational fusions with hygB. One of the two fusions apparently adds 65 and the other at least 86 amino acids to the N-terminus of the hygB product. Plasmids containing hygB-C. heterostrophus promoter fusions can be used unaltered to drive hygB expression in several other filamentous ascomycetes. This approach to achieving transformation may have general utility, especially for organisms with relatively undeveloped genetics.

A fungal kinesin required for organelle motility, hyphal growth, and morphogenesis.

A gene (NhKIN1) encoding a kinesin was cloned from Nectria haematococca genomic DNA by polymerase chain reaction amplification, using primers corresponding to conserved regions of known kinesin-encoding genes. Sequence analysis showed that NhKIN1 belongs to the subfamily of conventional kinesins and is distinct from any of the currently designated kinesin-related protein subfamilies. Deletion of NhKIN1 by transformation-mediated homologous recombination caused several dramatic phenotypes: a 50% reduction in colony growth rate, helical or wavy hyphae with reduced diameter, and subcellular abnormalities including withdrawal of mitochondria from the growing hyphal apex and reduction in the size of the Spitzenkörper, an apical aggregate of secretory vesicles. The effects on mitochondria and Spitzenkörper were not due to altered microtubule distribution, as microtubules were abundant throughout the length of hyphal tip cells of the mutant. The rate of spindle elongation during anaphase B of mitosis was reduced 11%, but the rate was not significantly different from that of wild type. This lack of a substantial mitotic phenotype is consistent with the primary role of the conventional kinesins in organelle motility rather than mitosis. Our results provide further evidence that the microtubule-based motility mechanism has a direct role in apical transport of secretory vesicles and the first evidence for its role in apical transport of mitochondria in a filamentous fungus. They also include a unique demonstration that a microtubule-based motor protein is essential for normal positioning of the Spitzenkörper, thus providing a new insight into the cellular basis for the aberrant hyphal morphology.

Fungal genomics and pathogenicity.

The filamentous fungal genetics community has enthusiastically embraced the utilization of genomics technologies to resolve long-standing issues in fungal biology. For example, such technologies have been proposed to study the mechanics of tip growth, photoreception, gene silencing, the molecular basis of conidiation, the pathway leading to sexual reproduction, and mechanisms of pathogenesis. These studies have provided a refreshing change of pace in research on filamentous fungi, which has lagged behind that on other eukaryotes in the exploitation of genome-wide methodologies. Despite the late start, several fungal genome sequencing projects are underway. The resulting databases will allow the comprehensive analysis of developmental processes that are characteristic of fungi, including the molecular nature of pathogenicity. DNA databases underpin analyses of the fungal transcriptome, proteome, and metabolome. This combined information will contribute to our basic understanding of not only the mechanics of infection but also the evolution of pathogenicity.

Application of mating type gene technology to problems in fungal biology.

In ascomycetes, the single mating type locus (MAT) controls sexual development. This locus is structurally unusual because the two alternate forms ("alleles") are completely dissimilar sequences, encoding different transcription factors, yet they occupy the same chromosomal position. Recently developed procedures allow efficient cloning of MAT genes from a wide array of filamentous ascomycetes, thereby providing MAT-based technology for application to several ongoing issues in fungal biology. This article first outlines the basic nature of MAT genes, then addresses the following topics: efficient cloning of MAT genes; the unusual molecular characteristics of these genes; phylogenetics using MAT; the issues of why some fungi are self-sterile, others self-fertile, and yet others asexual; the long-standing mystery of possible mating type switching in filamentous fungi; and finally the evolutionary origins of pathogenic capability.

Toxin-deficient mutants from a toxin-sensitive transformant of Cochliobolus heterostrophus.

Tox1 is the only genetic element identified which controls production of T-toxin, a linear polyketide involved in the virulence of Cochliobolus heterostrophus to its host plant, corn. Previous attempts to induce toxin-deficient (Tox-) mutants, using conventional mutagenesis and screening procedures, have been unsuccessful. As a strategy to enrich for Tox- mutants, we constructed a Tox1+ strain that carried the corn T-urf13 gene (which confers T-toxin sensitivity) fused to a fungal mitochondrial signal sequence; the fusion was under control of the inducible Aspergillus nidulans pelA promoter which, in both A. nidulans and C. heterostrophus, is repressed by glucose and induced by polygalacturonic acid (PGA). We expected that a transformant carrying this construction would be sensitive to its own toxin when the T-urf13 gene was expressed. Indeed, the strain grew normally on medium containing glucose but was inhibited on medium containing PGA. Conidia of this strain were treated with ethylmethanesulfonate and plated on PGA medium. Among 362 survivors, 9 were defective in T-toxin production. Authenticity of each mutant was established by the presence of the transformation vector, proper mating type, and a restriction fragment length polymorphism tightly linked to the Tox1+ locus. Progeny of each mutant crossed to a Tox1+ tester segregated 1:1 (for wild type toxin production vs. no or reduced toxin production), indicating a single gene mutation in each case. Progeny of each mutant crossed to a Tox1- tester segregated 1:1 (for no toxin production vs. no or reduced toxin production) indicating that each mutation mapped at the Tox1 locus. Availability of Tox- mutants will permit mapping in the Tox1 region without interference from a known Tox1 linked translocation breakpoint.

The translocation-associated tox1 locus of Cochliobolus heterostrophus is two genetic elements on two different chromosomes.

Previously, Tox1 was defined as a single genetic element controlling the difference between races of Cochliobolus heterostrophus: race T is highly virulent on T-cytoplasm corn and produces the polyketide T-toxin; race O is weakly virulent and does not produce T-toxin. Here we report that Tox1 is two loci, Tox1A and Tox1B, on two different chromosomes. Evidence for two loci derives from: (1) the appearance of 25% Tox+ progeny in crosses between induced Tox1(-) mutants, one defective at Tox1A, the other at Tox1B; (2) the ability of Tox1A- + Tox1B- heterokaryons to complement for T-toxin production; and (3) electrophoretic karyotypes proving that Tox1(-) mutations are physically located on two different chromosomes. Data showing Tox1 as a single genetic element are reconciled with those proving it is two loci by the fact that Tox1 is inseparably linked to the breakpoints of a reciprocal translocation; the translocation results in a four-armed linkage group. In crosses where the translocation is heterozygous (i.e., race T by race O), all markers linked to the four-armed intersection appear linked to each other; in crosses between induced Tox1(-) mutants, complications due to the translocation are eliminated and the two loci segregate independently.

Isolation of a phytoalexin-detoxification gene from the plant pathogenic fungus Nectria haematococca by detecting its expression in Aspergillus nidulans.

Detoxification of the pea phytoalexin pisatin via demethylation, mediated by a cytochrome P-450 monooxygenase, is thought to be important for pathogenicity of the fungus Nectria haematococca on pea. To isolate a fungal gene encoding pisatin demethylating activity (pda), we transformed Aspergillus nidulans with a genomic library of N. haematococca DNA constructed in a cosmid which carried the A. nidulans trpC gene. Transformants were selected for Trp+ and then screened for pda. One transformant among 1250 tested was Pda+ and was less sensitive to pisatin in culture than Pda- A. nidulans. The cosmid containing the gene (PDA) conferring this activity was recovered by phage lambda packaging of transformant genomic DNA. When A. nidulans was transformed with the cloned cosmid, 98% of the Trp+ transformants were Pda+. RNA blots probed with a 3.35 kb subclone carrying PDA indicated that the gene is expressed constitutively in A. nidulans but is inducible by pisatin in N. haematococca.

A cloned tryptophan-synthesis gene from the ascomycete Cochliobolus heterostrophus functions in Escherichia coli, yeast and Aspergillus nidulans.

A gene (TRP1) in the tryptophan biosynthetic pathway of the fungal plant pathogen Cochliobolus heterostrophus was isolated by complementation of an Escherichia coli trpF mutant which lacked phosphoribosylanthranilate isomerase (PRAI) activity. The cloned gene also complemented an E. coli trpC mutant lacking indoleglycerolphosphate synthase (IGPS) activity, a yeast trp1 mutant missing PRAI activity and an Aspergillus nidulans trpC mutant. It functioned in E. coli and A. nidulans without apparent rearrangement but in yeast only after the 5' end of the gene was deleted. The gene was subcloned on a 4.65-kb DNA fragment and the PRAI domain was localized to a 2.9-kb region. It showed homology to the A. nidulans trpC and Neurospora crassa trp-1 genes. There was one predominant transcript of C. heterostrophus TRP1, the same size (2.6-kb) as one of the two functional transcripts produced by A. nidulans trpC. The constitutive activity of the C. heterostrophus TRP1 gene was high whereas that of the A. nidulans trpC gene was low.

Ultrastructure of infection-thread development during the infection of soybean by Rhizobium japonicum.

The location and topography of infection sites in soybean (Glycine max (L.) Merr.) root hairs spot-inoculated with Rhizobium japonicum have been studied at the ultrastructural level. Infections commonly developed at sites created when the induced deformation of an emerging root hair caused a portion of the root-hair cell wall to press against an adjacent epidermal cell, entrapping rhizobia within the pocket between the two host cells. Infections were initiated by bacteria which became embedded in the mucigel in the enclosed groove. Infection-thread formation in soybean appears to involve degradation of mucigel material and localized disruption of the outer layer of the folded hair cell wall by one or more entrapped rhizobia. Rhizobia at the site of penetration are separated from the host cytoplasm by the host plasmalemma and by a layer of wall material that appears similar or identical to the normal inner layer of the hair cell wall. Proliferation of the bacteria results in an irregular, wall-bound sac near the site of penetration. Tubular infection threads, bounded by wall material of the same appearance as that surrounding the sac, emerge from the sac to carry rhizobia roughly single-file into the hair cell. Growing regions of the infection sac or thread are surrounded by host cytoplasm with high concentrations of organelles associated with synthesis and deposition of membrane and cell-wall material. The threads follow a highly irregular path toward the base of the hair cell. Threads commonly run along the base of the hair cell for some distance, and may branch and penetrate into subjacent cortical cells at several points in a manner analagous to the initial penetration of the root hair.

Early Events in the Infection of Soybean (Glycine max L. Merr) by Rhizobium japonicum: I. LOCALIZATION OF INFECTIBLE ROOT CELLS.

The infectible cells of soybean roots appear to be located at any given time just above the zone of root elongation and just below the position of the smallest emergent root hairs. The location of infectible cells on the primary root at the time of inoculation was inferred from the position of subsequent nodule development, correcting for displacement of epidermal cells due to root elongation. Marks were made on the seedling growth pouches at the time of inoculation to indicate the position of the root tip and the zones of root hair development. Virtually all of the seedlings developed nodules on the primary root above the marks made at the root tips at the time of inoculation. None of the plants formed nodules on the root where mature root hairs were present at the time of inoculation. These results and profiles of nodulation frequency indicate that the location of infectible cells is developmentally restricted. When inoculations were delayed for intervals of 1 to 4 hours after marking the positions of the root tips, progressively fewer nodules were formed above the root tip marks, and the uppermost of these nodules were formed at progressively shorter distances above the marks. These results indicate that the infectibility of given host cells is a transient property that appears and then is lost within a few hours. The results also indicate that host responses leading to infection and nodulation are triggered or initiated in less than 2 hours after inoculation. The extent of nodulation above the root tip mark increased in proportion to the logarithm of the number of bacteria in the inoculum.

Deletion of the Cochliobolus heterostrophus mating-type (MAT) locus promotes the function of MAT transgenes.

Cochliobolus heterostrophus has alternate genes (MAT-1 and MAT-2) at its mating-type locus. Transformants of a MAT-1 or a MAT-2 strain carrying a transgene of opposite mating type can self and are dual maters; the transgene, however, promotes development of pseudothecia only, not ascospores. To determine if the resident gene interferes with the function of the transgene, transformation vectors were designed to delete different amounts (2.5 kb, 5.7 kb, and 6.3 kb) of DNA at the MAT locus. Deletions occurred at a higher frequency (about 90% of transformants) with linearized plasmid than with circular plasmid (about 15% of transformants), and all three vectors were equally efficient at gene replacement. Both MAT-1 and MAT-2 could be deleted with the same set of vectors. Re-transformation of deletion strains (regardless of deletion size) with a wild-type copy of MAT restored full mating ability, indicating that the resident MAT gene interferes with function of the MAT transgene. Moreover, sexual development was normal whether the MAT transgene integrated at the homologous or at an ectopic site.

Organization of ribosomal RNA genes in the fungus Cochliobolus heterostrophus.

The genes encoding the 17S, 5.8S and 25S ribosomal RNAs in the Ascomycete Cochliobolus heterostrophus were cloned and analyzed. They are arranged in tandemly repeated units (rDNA) either 9.0 or 9.15 kilobases in length, depending upon the strain. The 5S rRNA genes are not part of the tandemly repeated rDNA. Instead, many and perhaps all of the 5S genes are dispersed in the genome, as they are in the fungi Neurospora, Aspergillus and Schizosaccharomyces. Comparative restriction maps of the rDNA from C. heterostrophus and other filamentous fungi and yeasts are presented. A survey of rDNAs from twenty-three field isolates of C. heterostrophus collected worldwide demonstrated that each isolate has one or the other of two rDNA forms, which differ in length and in the presence or absence of at least three restriction enzyme sites. The differences are all located in spacer DNA outside the coding regions for the rRNA genes. Heterogeneity in the rDNA repeat was not observed within any single isolate. The copy number of the rRNA gene cluster in C. heterostrophus is approximately 130 per haploid genome.

Transcripts at the mating type locus of Cochliobolus heterostrophus.

The single mating type locus (MAT) of the heterothallic ascomycete Cochliobolus heterostrophus is composed of a pair of unlike sequences called idiomorphs, each of which encodes one MAT-specific gene (MAT-1 and MAT-2). MAT transcripts were observed in blots of poly(A)+ RNA isolated from cultures grown in minimal medium, but were not detectable after growth of the fungus in complete medium, suggesting that transcription of MAT is tightly regulated. The idiomorphs (MAT-1 = 1297-bp, MAT-2 = 1171-bp) encode transcripts of 2.2 kb (MAT-1) and 2.1 kb (MAT-2), which start 5' and end 3' of the idiomorph within sequences common to both mating types. Analyses of MAT-1 and MAT-2 cDNAs revealed obligatory splicing of one intron (55-bp in MAT-1, 52-bp in MAT-2) within each MAT-specific ORF and optional splicing of two introns (63 and 79-bp) in the long (approximately 0.55 kb) 5' untranslated leader sequences; the 3' untranslated region is 0.46 kb long. Transcription start sites were found 5' of, and within, the 79-bp intron. Optional splicing of the upstream introns and at least two transcription start sites result in three types of transcript: Type I with both 5' introns spliced, Type II with only the 63-bp intron spliced, and Type III with neither 5' intron spliced. The three transcript types are distinguished by various combinations of four short ORFs encoded by the corresponding genomic DNA, in the leader sequences of the MAT mRNAs. The transcript structure suggests several mechanisms by which expression of the MAT genes might be regulated at the level of translation during sexual development.

Efficient cloning of ascomycete mating type genes by PCR amplification of the conserved MAT HMG Box.

Cloning of mating type (MAT) genes from ascomycetes has been hampered by low conservation among them. One of the pair of MAT genes, represented by MAT-2 of Cochliobolus heterostrophus (a loculoascomycete) and mt a of Neurospora crassa (a pyrenomycete), encodes a protein with a conserved DNA binding motif called the high mobility group (HMG) box. PCR with primer pairs corresponding to the borders of the C. heterostrophus and the N. crassa HMG boxes generated an approximately 0.3-kb product from genomic DNAs of MAT-2 and mt a strains, respectively, but not from MAT-1 and mt A strains. The C. heterostrophus primers amplified approximately 0.3-kb products from DNA of most loculoascomycete genera tested but not from DNA of pyrenomycete genera; this specificity was reversed with the N. crassa primers. The validity of the PCR procedure was documented by near sequence identity between the C. heterostrophus MAT-2 HMG box and PCR products from several Cochliobolus spp. and by cosegregation of the PCR product with mating type in progeny of Setosphaeria turcica and of Cryphonectria parasitica. Regions of the MAT locus flanking the HMG box were readily cloned using the TAIL-PCR procedure with a combination of random and specific primers.

Analysis of mating-type genes in the chestnut blight fungus, Cryphonectria parasitica.

In nature, the chestnut blight fungus, Cryphonectria parasitica, has a mixed mating system; i.e., individuals in the same population have the ability to self and outcross. In the laboratory, C. parasitica appears to have a bipolar self-incompatibility system, typical of heterothallic ascomycetes; selfing is rare, although demonstrable. In this report we describe the cloning and sequencing of both mating-type idiomorphs and their flanking regions at the MAT locus in C. parasitica. The two idiomorphs, MAT1-1 and MAT1-2, are structurally similar to those of other pyrenomycetes described to date. MAT1-1 encodes three genes (MAT1-1-1, MAT1-1-2, and MAT1-1-3) and MAT1-2 encodes a single gene (MAT1-2-1). Unlike MAT idiomorphs in some ascomycetes, the sequences at both ends of the idiomorphs in C. parasitica show a relatively gradual, rather than abrupt, transition from identity in the flanking regions to almost complete dissimilarity in the coding regions. The flanking regions have repetitive polypyrimidine (T/C) and polypurine (A/G) tracts; the significance of these repetitive tracts is unknown. Although we found repetitive tracts in the flanks and gradual transition zones at the ends of the idiomorphs, we found no special features that would explain how selfing occurs in an otherwise self-incompatible fungus.

A cytoplasmic dynein required for mitotic aster formation in vivo.

An astral pulling force helps to elongate the mitotic spindle in the filamentous ascomycete, Nectria haematococca. Evidence is mounting that dynein is required for the formation of mitotic spindles and asters. Obviously, this would be an important mitotic function of dynein, since it would be a prerequisite for astral force to be applied to a spindle pole. Missing from the evidence for such a role of dynein in aster formation, however, has been a dynein mutant lacking mitotic asters. To determine whether or not cytoplasmic dynein is involved in mitotic aster formation in N. haematococca, a dynein-deficient mutant was made. Immunocytochemistry visualized few or no mitotic astral microtubules in the mutant cells, and studies of living cells confirmed the veracity of this result by revealing the absence of mitotic aster functions in vivo: intra-astral motility of membranous organelles was not apparent; the rate and extent of spindle elongation during anaphase B were reduced; and spindle pole body separation almost stopped when the anaphase B spindle in the mutant was cut by a laser microbeam, demonstrating unequivocally that no astral pulling force was present. These unique results not only provide a demonstration that cytoplasmic dynein is required for the formation of mitotic asters in N. haematococca; they also represent the first report of mitotic phenotypes in a dynein mutant of any filamentous fungus and the first cytoplasmic dynein mutant of any organism whose mitotic phenotypes demonstrate the requirement of cytoplasmic dynein for aster formation in vivo.

Single mating type-specific genes and their 3' UTRs control mating and fertility in Cochliobolus heterostrophus.

To determine the number of proteins required for mating type (MAT) locus-regulated control of mating in Cochliobolus heterostrophus, MAT fragments of various sizes were expressed in MAT deletion strains. As little as 1.5 kb of MAT sequence, encoding a single unique protein in each mating type (MAT-1 and MAT-2), conferred mating ability, although an additional 160 bp of 3' UTR was needed for production of ascospores. No other mating type-specific genes involved in mating identity or fertility were found. Thus, although homologs of the C. heterostrophus MAT-1 and MAT-2 genes exist in the filamentous ascomycetes Neurospora crassa and Podospora anserina, C. heterostrophus does not appear to have mating type-specific homologs of two additional genes required by both N. crassa and P. anserina for successful sexual reproduction. Three genes were identified in the common DNA flanking the MAT locus: a gene encoding a GTPase-activating protein and an ORF of unknown function lie 5' while a beta-glucosidase encoding gene lies found 3'. None of these genes appears to be involving in the mating process.

Evolution of the fungal self-fertile reproductive life style from self-sterile ancestors.

In most fungal ascomycetes, mating is controlled by a single locus (MAT). Fungi requiring a partner to mate are heterothallic (self-sterile); those not requiring a partner are homothallic (self-fertile). Structural analyses of MAT sequences from homothallic and heterothallic Cochliobolus species support the hypothesis that heterothallism is ancestral. Homothallic species carry both MAT genes in a single nucleus, usually closely linked or fused, in contrast to heterothallic species, which have alternate MAT genes in different nuclei. The structural organization of MAT from all heterothallic species examined is highly conserved; in contrast, the organization of MAT in each homothallic species is unique. The mechanism of conversion from heterothallism to homothallism is a recombination event between islands of identity in otherwise dissimilar MAT sequences. Expression of a fused MAT gene from a homothallic species confers self-fertility on a MAT-null strain of a heterothallic species, suggesting that MAT alone is sufficient to change reproductive life style.