A fumonisin biosynthetic gene cluster in Fusarium oxysporum strain O-1890 and the genetic basis for B versus C fumonisin production.

Most species of Fusarium that produce fumonisin mycotoxins produce predominantly B fumonisins (FBs). However, Fusarium oxysporum strain O-1890 produces predominantly C fumonisins (FCs). In this study, the nucleotide sequence of the fumonisin biosynthetic gene (FUM) cluster in strain O-1890 was determined. The order and orientation of FUM genes were the same as in the previously described clusters in Fusarium verticillioides and Fusarium proliferatum. Coding regions of F. oxysporum and F. verticillioides FUM genes were 88-92% identical, but regions flanking the clusters did not share significant identity. The FUM cluster gene FUM8 encodes an alpha-oxoamine synthase, and fum8 mutants of F. verticillioides do not produce fumonisins. Complementation of a fum8 mutant with the F. verticillioidesFUM8 restored FB production. Complementation with F. oxysporumFUM8 also restored production, but the fumonisins produced were predominantly FCs. These data indicate that different orthologues of FUM8 determine whether Fusarium produces predominantly FBs or FCs.

Deletion analysis of FUM genes involved in tricarballylic ester formation during fumonisin biosynthesis.

Fumonisins are carcinogenic mycotoxins produced by the maize ear rot pathogen Gibberella moniliformis (anamorph Fusarium verticillioides). These toxins consist of a linear polyketide-derived backbone substituted at various positions with an amine, one to four hydroxyl, two methyl, and two tricarballylic ester functions. In this study, we generated and characterized deletion mutants of G. moniliformis for five genes, FUM7, FUM10, FUM11, FUM14, and FUM16 in the fumonisin biosynthetic gene cluster. Functional analysis of mutants in four genes, predicted to encode unrelated proteins, affected formation of the tricarballylic esters. FUM7 deletion mutants produced a previously undescribed homologue of fumonisin B1 with an alkene function in both tricarballylic esters, FUM10 and FUM14 deletion mutants produced homologues of fumonisin B3 and fumonisin B4 that lack tricarballylic ester functions, and FUM11 deletion mutants produced fumonisins that lack one of the tricarballylic ester functions. These phenotypes indicated specific roles for FUM7, FUM10, FUM11, and FUM14 in fumonisin biosynthesis that are consistent with the predicted proteins encoded by each gene. Deletion of FUM16 had no apparent effect on fumonisin production. The phenotypes of the deletion mutants provide further insight into the order of steps in fumonisin biosynthesis.

Fumonisin production in the maize pathogen Fusarium verticillioides: genetic basis of naturally occurring chemical variation.

Fumonisins are polyketide-derived mycotoxins produced by the maize pathogen Fusarium verticillioides. Previous analyses identified naturally occurring variants of the fungus that are deficient in fumonisin C-10 hydroxylation or that do not produce any fumonisins. In the current study, gene deletion and genetic complementation analyses localized the C-10 hydroxylation deficiency to a cytochrome P450 monooxygenase gene in the fumonisin biosynthetic gene (FUM) cluster. Sequence analysis indicated that the hydroxylation deficiency resulted from a single nucleotide insertion that caused a frame shift in the coding region of the gene. Genetic complementation localized the fumonisin-nonproduction phenotype to the polyketide synthase gene in the FUM cluster, and sequence analysis indicated that the nonproduction phenotype resulted from a nucleotide substitution, which introduced a premature stop codon in the coding region. These results provide the first direct evidence that altered fumonisin production phenotypes of naturally occurring F. verticillioides variants can result from single point mutations in the FUM cluster.

Fusarium graminearum TRI14 is required for high virulence and DON production on wheat but not for DON synthesis in vitro.

Fusarium head blight (FHB) of wheat (Triticum aestivum L.), caused by the fungus Fusarium graminearum, is a major concern worldwide. FHB grain is reduced in yield, may fail to germinate, and is often contaminated with deoxynivalenol, a trichothecene mycotoxin linked to a variety of animal diseases and feed refusals. Annual losses in the tens of millions of dollars due to FHB underscore the need to develop improved methods of disease control and prevention. Previous research has identified deoxynivalenol biosynthesis as a virulence factor on wheat. Recently, we found that the TRI14 gene of F. sporotrichioides, closely related to F. graminearum, was not required for synthesis of a related trichothecene, T-2 toxin. TRI14 does not share similarity with any previously described genes in the databases. In this study, we examined the role that F. graminearum TRI14 may play in both deoxynivalenol synthesis and in virulence on wheat. TRI14 deletion mutants synthesize deoxynivalenol on cracked maize kernel medium and exhibit wild-type colony morphology and growth rate on complex and minimal agar media. However, FHB assays on greenhouse-grown wheat indicate that FgDeltaTri14 mutants cause 50-80% less disease than wild type and do not produce a detectable quantity of deoxynivalenol on plants. We discuss a number of possible roles that TRI14 may play in the disease process.

Identification and heritability of fumonisin insensitivity in Zea mays.

Landraces of maize (Zea mays ssp. mays) and its wild teosinte relatives (Zea mays spp. parviglumis and mexicana) were surveyed for sensitivity to fumonisin B(1), a phytotoxin produced by the maize pathogen Gibberella moniliformis. Only two of 42 Z. mays samples were highly insensitive to FB(1) (ED(50) = ca. 200 microM). The teosintes and 76% of the maize landraces were moderately or highly sensitive to FB(1) (ED(50) < or = 30 microM), which indicates that FB(1) sensitivity is likely to be an ancestral trait in Z. mays. F(1) generations derived from crosses between FB(1)-sensitive maize inbred B73 and insensitive landraces were significantly less sensitive than B73. Thus, our data indicate that FB(1)-insensitivity is a relatively rare but heritable trait in maize. We also report the sensitivity of maize to other Gibberella toxins - beauvericin, diacetoxyscirpenol, and moniliformin.

Comparative analysis of 87,000 expressed sequence tags from the fumonisin-producing fungus Fusarium verticillioides.

Fusarium verticillioides (teleomorph Gibberella moniliformis) is a pathogen of maize worldwide and produces fumonisins, a family of mycotoxins that have been associated with several animal diseases as well as cancer in humans. In this study, we sought to identify fungal genes that affect fumonisin production and/or the plant-fungal interaction. We generated over 87,000 expressed sequence tags from nine different cDNA libraries that correspond to 11,119 unique sequences and are estimated to represent 80% of the genomic complement of genes. A comparative analysis of the libraries showed that all 15 genes in the fumonisin gene cluster were differentially expressed. In addition, nine candidate fumonisin regulatory genes and a number of genes that may play a role in plant-fungal interaction were identified. Analysis of over 700 FUM gene transcripts from five different libraries provided evidence for transcripts with unspliced introns and spliced introns with alternative 3' splice sites. The abundance of the alternative splice forms and the frequency with which they were found for genes involved in the biosynthesis of a single family of metabolites as well as their differential expression suggest they may have a biological function. Finally, analysis of an EST that aligns to genomic sequence between FUM12 and FUM13 provided evidence for a previously unidentified gene (FUM20) in the FUM gene cluster.

Discontinuous distribution of fumonisin biosynthetic genes in the Gibberella fujikuroi species complex.

Production of the carcinogenic mycotoxins fumonisins has been reported in several Fusarium species, most of which are members of the Gibberella fujikuroi (Gf) complex. In this study, we examined 15 Fusarium species in the Gf complex and 12 other species for fumonisin production and the presence of fumonisin biosynthetic genes (FUM). Among the species within the Gf complex, fumonisin production was detected only in F. fujikuroi, F. globosum, F. proliferatum, F. nygamai, F. oxysporum and F. verticillioides. These five species include members of two of the three major clades delineated in the Gf complex. The FUM genes were detected in these same five species and in F. anthophilum, a member of the third clade. Among the species outside the Gf complex, fumonisin production and FUM genes were detected only in F. oxysporum. Phylogenetic analyses of nucleotide sequences from two FUM gene fragments inferred relationships similar but not identical to those inferred from previous analyses of other genes. The results indicate the FUM genes are discontinuously distributed in the Gf complex and that this distribution gives rise to the differences in the abilities of closely related Fusarium species to produce fumonisins.

Patterns of trichothecene production, genetic variability, and virulence to wheat of Fusarium graminearum from smallholder farms in Nepal.

Fusarium graminearum causes wheat head blight and contaminates grain with the trichothecenes 4-deoxynivalenol and nivalenol. Sequence analysis of trichothecene genes indicates that nivalenol production is the ancestral trait; however, deoxynivalenol producers occur worldwide and predominate in North and South America and in Europe. Analysis of a large field population (>500 strains) from Nepal identified three groups that were both genetically distinct and polymorphic for trichothecene production: SCAR1 comprising 95% deoxynivalenol producers, SCAR2 comprising 94% nivalenol producers, and SCAR3/5 comprising 34% deoxynivalenol producers/63% nivalenol producers. The ability to cause wheat head blight differed between SCAR groups and trichothecene chemotypes: deoxynivalenol producers were more virulent than nivalenol producers across all three SCAR groups and within the SCAR3/5 genetic background. These data support the hypothesis that production of deoxynivalenol rather than nivalenol confers a selective advantage to this important wheat pathogen.

Functional demarcation of the Fusarium core trichothecene gene cluster.

Many Fusarium species produce toxic sesquiterpenoids known as trichothecenes, including deoxynivalenol and nivalenol by Fusarium graminearum and T-2 toxin by Fusarium sporotrichioides. These toxins are potent inhibitors of protein synthesis and are a significant agricultural problem due to their adverse affect on human, animal, and plant health. Previously, 10-12 co-regulated orthologous genes within a 26-kb region were identified in F. graminearum and F. sporotrichioides, respectively. A majority of these clustered genes have been shown to be involved in different aspects of trichothecene metabolism including 7 of 15 biosynthetic steps. Three other biosynthetic steps are carried out by genes located elsewhere in the genome. In this study, we sequenced 14-16 kb of DNA on both sides of the core clusters and identified 12 new ORFs in both Fusarium species. Although the predicted functions of some of the new ORFs are consistent with some unassigned biochemical reactions, gene expression and gene deletion studies indicate that none are required for trichothecene biosynthesis. These results provide evidence to demarcate both ends of the core trichothecene gene cluster. Index descriptors: Fungal secondary metabolite, Pathogenic fungi, Gene cluster, Fusarium, Trichothecene, DON

Characterization of a fusarium 2-gene cluster involved in trichothecene C-8 modification.

The Fusarium trichothecenes T-2 toxin and deoxynivalenol (DON) are potent inhibitors of eukaryotic protein synthesis and are a significant agricultural problem. Three coregulated loci are required for T-2 toxin synthesis by Fusarium sporotrichioides. The core-trichothecene gene cluster consists of 12 genes (Tri3-Tri14) while the second locus consists of a single gene (Tri101). The third locus was recently partially described and encodes 1-2 biosynthetic enzymes and a putative regulatory gene. Here, we describe a detailed characterization of this locus. Located adjacent to Tri1 is Tri16, which is required for esterification of the C-8 hydroxyl. A putative regulatory gene, also adjacent to Tri1, is not required for T-2 toxin synthesis. The genomic sequence of Fusarium graminearum (a DON producer) contains a putative functional Tri1 and a nonfunctional Tri16. The presence of the Tri16 pseudogene is consistent with the chemical structure of DON, which has a C-8 keto group rather than the C-8 ester of T-2 toxin.

FUM9 is required for C-5 hydroxylation of fumonisins and complements the meitotically defined Fum3 locus in Gibberella moniliformis.

Deletion of the Gibberella moniliformis FUM9 gene resulted in mutants that produce only fumonisins that lack a C-5 hydroxyl group. This phenotype is identical to that of previously described mutants with defective alleles at the meiotically defined Fum3 locus. Transformation with a wild-type FUM9 gene into a Fum3-defective mutant restored wild-type fumonisin production. These results indicate that the FUM9 protein catalyzes the C-5 hydroxylation of fumonisins and that FUM9 and the Fum3 locus are the same gene.

FUM13 encodes a short chain dehydrogenase/reductase required for C-3 carbonyl reduction during fumonisin biosynthesis in Gibberella moniliformis.

Fumonisins are polyketide-derived mycotoxins produced by the filamentous fungus Gibberella moniliformis (anamorph Fusarium verticillioides). Wild-type strains of the fungus produce predominantly four B-series fumonisins, designated FB(1), FB(2), FB(3), and FB(4). Recently, a cluster of 15 putative fumonisin biosynthetic genes (FUM) was described in G. moniliformis. We have now conducted a functional analysis of FUM13, a gene in the cluster that is predicted by amino acid sequence similarity to encode a short chain dehydrogenase/reductase (SDR). Mass spectrometric analysis of metabolites from FUM13 deletion mutants revealed that they produce approximately 10% of wild-type levels of B-series fumonisins as well as two previously uncharacterized compounds. NMR analysis revealed that the new compounds are similar in structure to FB(3) and FB(4) but that they have a carbonyl function rather than a hydroxyl function at carbon atom 3 (C-3). These results indicate that the FUM13 protein catalyzes the reduction of the C-3 carbonyl to a hydroxyl group and are the first biochemical evidence directly linking a FUM gene to a specific reaction during fumonisin biosynthesis. The production of low levels of FB(1), FB(2), FB(3), and FB(4), which have a C-3 hydroxyl, by the FUM13 mutants suggests that G. moniliformis has an additional C-3 carbonyl reductase activity but that this enzyme functions less efficiently than the FUM13 protein.

Co-expression of 15 contiguous genes delineates a fumonisin biosynthetic gene cluster in Gibberella moniliformis.

Fumonisins are mycotoxins produced by the maize pathogen Gibberella moniliformis and are associated with cancer in rodents. In this study, we determined the nucleotide sequence of a 75-kb region of G. moniliformis DNA and identified 18 heretofore undescribed genes flanking a cluster of five previously identified fumonisin biosynthetic (FUM) genes. Ten of the newly identified genes downstream of the cluster were coregulated with FUM genes and exhibited patterns of expression that were correlated with fumonisin production. BLASTX analyses indicated that the predicted functions of proteins encoded by the 10 genes were consistent with activities expected for fumonisin biosynthesis or self-protection. These data indicate that the 10 newly identified genes and the previously identified FUM genes constitute a fumonisin biosynthetic gene cluster. Disruption of two of the new genes, encoding longevity assurance factors, had no apparent effect on fumonisin production, but disruption of a third, encoding an ABC transporter, had a subtle effect on ratios of fumonisins produced.

Determination of deoxynivalenol and nivalenol in corn and wheat by liquid chromatography with electrospray mass spectrometry.

The fungus Fusarium graminearum is a pathogen of both wheat and corn. Strains of the fungus from the United States produce a toxin, deoxynivalenol (DON); strains of the fungus from Asia and Europe produce DON or a related toxin, nivalenol. These toxins can cause disease in livestock, and their potential presence in feed and foods is a concern for animal and human health. A method was developed to detect both toxins in corn and wheat by liquid chromatography/mass spectrometry of an extract of ground grain. The method requires no sample cleanup and can detect the toxins at 0.05 microg/g.

Rapid fluorescence polarization immunoassay for the mycotoxin deoxynivalenol in wheat.

The fungus Fusarium graminearum, a pathogen of both wheat and maize, produces a toxin, deoxynivalenol (DON), that causes disease in livestock. A rapid test for DON in wheat was developed using the principle of fluorescence polarization (FP) immunoassay. The assay was based on the competition between DON and a novel DON-fluorescein tracer (DON-FL2) for a DON-specific monoclonal antibody in solution. The method, which is a substantial improvement over our previous DON FP immunoassay, combined a rapid (3 min) extraction step with a rapid (2 min) detection step. A series of naturally contaminated wheat and maize samples were analyzed by both FP immunoassay and liquid chromatography (HPLC-UV). For wheat the HPLC-UV and FP methods agreed well (linear regression r(2) = 0.936), but for maize the two methods did not (r (2) = 0.849). We conclude that the FP method is useful for screening wheat, but not maize, for DON.

N-(1-deoxy-D-fructos-1-yl) fumonisin B(1), the initial reaction product of fumonisin B(1) and D-glucose.

Incubation of fumonisin B(1) and D-glucose in aqueous solutions resulted in the formation of N-(1-deoxy-D-fructos-1-yl) fumonisin B(1) in addition to the previously reported N-(carboxymethyl) fumonisin B(1). N-(1-Deoxy-D-fructos-1-yl) fumonisin B(1) is the first stable product formed after the Amadori rearrangement of the Schiff base formed by the reaction of the primary amine of fumonisin B(1) and the aldehyde group of D-glucose. N-(1-Deoxy-D-fructos-1-yl) fumonisin B(1) was synthesized by reacting fumonisin B(1) with an excess of D-glucose in methanol and heating for 6 h at 64 degrees C. It was purified using C(18) and strong cation exchange solid-phase extraction cartridges and characterized by nuclear magnetic resonance and liquid chromatography-mass spectrometry. Subsequently, N,N-dimethylformamide was found to be a better reaction solvent, requiring reaction for only 2-3 h at 64 degrees C and eliminating the formation of methyl esters. Alkaline hydrolysis of N-(1-deoxy-D-fructos-1-yl) fumonisin B(1) gave a mixture of hydrolyzed fumonisin B(1) and hydrolyzed N-(carboxymethyl) fumonisin B(1).

Analysis of aberrant virulence of Gibberella zeae following transformation-mediated complementation of a trichothecene-deficient (Tri5) mutant.

Gibberella zeae causes wheat ear blight and produces trichothecene toxins in infected grain. In previous studies, trichothecene production in this fungus was disabled by specific disruption of the trichodiene synthase gene (Tri5) and was restored by two methods: gene reversion and transformation-mediated mutant complementation. In previous field tests of wheat ear blight, trichothecene-nonproducing mutants were less virulent than the wild-type progenitor strain from which they were derived. Trichothecene-producing revertants also were restored to wild-type levels of virulence. In contrast, in the field test of wheat ear blight reported here, trichothecene-producing strains obtained by Tri5 mutant complementation were not restored to wild-type levels of virulence. The complemented mutants showed a slightly reduced radial growth compared to the wild-type strain, but otherwise appeared normal in morphology, pigmentation and sexual fertility. Genetic analysis indicated that the aberrant virulence of a complemented mutant was likely due to non-target effects that occurred during the process of transforming the trichothecene-nonproducing mutant with Tri5. These results confirm previous findings that trichothecenes contribute to the virulence of G. zeae, but also demonstrate that manipulating this fungus in the laboratory may cause it to undergo subtle changes that reduce its virulence.

Distribution of Fumonisins in Maize Ears Infected with Strains of Fusarium moniliforme that Differ in Fumonisin Production.

Strains of Fusarium moniliforme (Gibberella fujikuroi mating population A) that differ in fu-monisin production in vitro were previously identified in a Kansas field population. One strain that produced high levels of fumonisins and two strains that produced very low levels of fu-monisins were applied to maize kernels at planting at the Rocky Ford Farm near Manhattan, Kansas. The distribution of fumonisins in symptomatic and symptomless kernels from individual harvested ears was determined by high performance liquid chromatography, and the distribution of the three applied strains in the kernels was determined by vegetative compatibility group analysis. Both symptomatic and symptomless kernels were extensively colonized with F. moniliforme, but the highest levels of fumonisins were in the symptomatic kernels. All three applied strains were recovered from kernels in 1993, and two of them were recovered from kernels in 1994. However, a high frequency of ear and kernel infection with a strain that produced little fumonisin in vitro did not consistently decrease the level of fumonisins. The frequency of infection with fumonisin low-producing strains may have been too low for competitive exclusion of naturally occurring fumonisin high-producing strains. Also, strains that are low-fumonisin producers under laboratory conditions may be high producers in the field.

Comparative Assessment of Fumonisin in Grain-Based Foods by ELISA, GC-MS, and HPLC.

Seventy-one (71) food samples were analyzed for the mycotoxin fumonisin by a monoclonal antibody based competitive enzyme-linked immunosorbent assay (ELISA). Fumonisins were detected primarily in corn-based products with 7/12, 2/2 and 1/3 and 1/7 yellow cornmeal, blue cornmeal, corn muffin mix, and mixed grain cereal samples yielding positive results, respectively. When the positive samples and randomly selected negative samples were assessed by other methods, correlations (r values) between ELISA and gas chromatography-mass spectrometry (GC-MS), ELISA and high-pressure liquid chromatography (HPLC) and GC-MS and HPLC were 0.478 (p < 0.05), 0.512 (p < 0.05), and 0.946 (p < 0.01), respectively. The results suggested that although the immunoassay could be used for screening of fumonisin in food samples, higher estimates were attained by ELISA than by the other two methods particularly in the more contaminated samples. These observations may result from differences in sample preparation among the methods or because of the presence of structurally related compounds in extracts that are detectable by ELISA but not the other two methods.