INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Trichothecenes are sesquiterpene epoxide mycotoxins produced by species of Fusarium, Trichothecium, and Myrothecium that act as potent inhibitors of eukaryotic protein synthesis (21). The broad spectrum of trichothecene toxicity probably is due to their ability to interact with highly conserved elements of the protein-synthetic apparatus. While trichothecene vertebrate toxicity has been the focus of numerous studies due to the frequent occurrence of trichothecenes in agricultural products, the observed phytotoxicity of trichothecenes has stimulated research into the role these toxins may play in plant diseases caused by some Fusarium species (4, 6, 11, 24).

The biosynthesis of the trichothecene T-2 toxin in Fusarium sporotrichioides has been studied with the aid of mutant strains blocked at specific steps in the trichothecene pathway (5, 12, 18-20). Many of the trichothecene pathway genes in F. sporotrichioides (12) are localized in a gene cluster of at least nine genes including genes for P450 oxygenases (1, 9), an acetyltransferase (19), a transcription factor required for pathway gene expression (25), and a toxin efflux pump (2).

Trichothecenes are antibiotics, and their biosynthesis likely requires special adaptations by the producing organisms for self-protection. Antibiotic-producing microorganisms use various mechanisms for self-protection including alteration of target proteins, pumps to reduce the intracellular concentration of the antibiotic, and metabolism to reduce toxicity (3). Trichothecene 3-O-acetyltransferase (TRI101) catalyzes the conversion of toxic Fusarium trichothecenes to less-toxic products and has, therefore, been proposed as a metabolic self-protection mechanism in Fusarium graminearum (13). TRI101 is not tightly linked to other trichothecene biosynthetic genes in either F. graminearum or F. sporotrichioides (13-15). The enzyme encoded by TRI101 can modify a number of trichothecenes and, when expressed in Schizosaccharomyces pombe, confers resistance to trichothecenes (13). It was suggested that TRI101 acetylation, rather than modification or replacement of target ribosomes, is the primary defense mechanism against 3-hydroxylated trichothecenes in Fusarium and that a mutation resulting in a loss of TRI101 would be lethal (13).

Our objectives in this study were to determine (i) if inactivation of TRI101 was lethal and (ii) if TRI101 functions in both Fusarium self-protection against trichothecenes and trichothecene biosynthesis. We found that disruption of TRI101 was not lethal and that TRI101 deletion mutants accumulated isotrichodermol, a 3-hydroxytrichothecene. These mutants could both germinate and grow in the presence of isotrichodermol and other 3-hydroxytrichothecenes. These results show that although expression of TRI101 in Saccharomyces cerevisiae and S. pombe increased their resistance to trichothecenes, TRI101 is not an essential self-defense mechanism for F. sporotrichioides. Our results also show that TRI101 is an essential trichothecene-biosynthetic gene and are consistent with the hypothesis that much of the T-2 toxin pathway is via 3-acetylated intermediates.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains. F. sporotrichioides NRRL 3299 was obtained from the USDA Agricultural Research Service Culture Collection at the National Center for Agricultural Utilization Research, Peoria, Ill., and maintained on V-8 juice agar slants (29). F. sporotrichioides FsTri101-3D contains disrupted sequences for TRI101 as described below. S. cerevisiae RW2802 (PDR5 leu2 ura3-52 met5) and JG436 (pdr5::Tn5 leu2 ura3-35 met5) were obtained from J. Golin, The Catholic University of America (22).

Media and culture conditions. All Fusarium cultures were grown initially on V-8 juice agar plates under an alternating cycle of 12 h of light at 25°C and 12 h of dark at 22°C. Conidia were washed from the plates and grown in YPG medium (0.3% yeast extract, 1% peptone, 2% glucose) for DNA isolation or in GYEP medium (5% glucose, 0.1% yeast extract, 0.1% peptone) for toxin production (33). For yeast transformations, cells were grown on YPD (1% yeast extract, 2% peptone, 2% glucose) plates for 1 to 3 days; otherwise, cultures were maintained on glucose minimal media with appropriate supplements (1 g of leucine, 0.2 g of uracil, and 0.2 g of methionine/liter). For feeding studies, yeasts were grown in Ygal (1% yeast extract, 2% peptone, 2% galactose) to induce plasmid expression.

Physical analyses. Gas chromatography (GC) measurements were made by flame ionization detection with a Hewlett-Packard 5890 gas chromatograph fitted with a 30-m fused-silica capillary column (DB1; 0.25 µm; J&W Scientific Co., Palo Alto, Calif.). For routine screening of the trichothecene toxin phenotype, the column was held at 120°C at injection, then heated to 210°C at 15°C/min and held for 1 min, and then heated to 260°C at 5°C/min and held for 8 min. Low-resolution mass spectra were obtained by GC-mass spectrometry (MS) with a Hewlett-Packard 5891 mass-selective detector fitted with a DB-5 MS column (15-m by 0.25-mm film thickness).

DNA and RNA manipulations. The cDNA expression library was constructed with mRNA from an F. sporotrichioides NRRL 3299 culture grown for 23 h in GYEP. cDNA was cloned into the yeast expression vector pYES2 (Invitrogen, Carlsbad, Calif.). Plasmid preparation methods were as previously described (2). A cosmid library of F. sporotrichioides NRRL 3299 genomic DNA was made in the SuperCos1 vector (Stratagene, La Jolla, Calif.) in accordance with the manufacturer's instructions and was screened with a ³²P-labeled probe corresponding to the TRI101 coding region. DNA labeling was performed with the Prime-a-Gene kit (Stratagene). Sequencing was performed with the DYEdeoxy sequencing kit (Applied Biosystems, Foster City, Calif.). Southern hybridizations were performed as described by Sambrook et al. (26).

Gene disruption and transformations. To disrupt TRI101 we subcloned a PCR (Pfu polymerase; Stratagene) fragment of about 5 kb from cosmid 5-1 that contains TRI101 into plasmid pBluescript II SK(+) (Stratagene) digested with BamHI/Ecl 136 to form pTriR-2. A chimeric hygromycin B phosphotransferase gene (hyg) containing promoter 1 from Cochliobolus heterostrophus (32) was cloned into the unique NcoI site located 90 bp downstream of the ATG start site in TRI101. The resulting plasmid, pTriR-3 (Fig. 1), was used to transform F. sporotrichioides as previously described (25). The PCR analysis employed primers 955 (5' GCGCTGCAGATCAAAATGGCCGAACAAGC 3'), 957 (5' GTTTCCTTCGCTGATGCC 3'), 997 (5' GGCGGTACCACAGAAAAGAGTAAAAGG 3'), 1134 (5' GTCGATCGATACGCACGC 3'), and 1135 (5' CTGGTCGTTGTATGTAGCC 3'). The underlined sequence in 955 indicates an added PstI site, and that in 997 indicates an added KpnI site. These primers have complementary sequences that are located both inside and outside the TRI101 sequences present in plasmid pTriR-3 (Fig. 1). In the Southern analysis, genomic DNA of transformants was digested with BamHI/XbaI and hybridized to a ³²P-labeled probe consisting of the TRI101 coding regions.

View larger version (24K): [in this window] [in a new window]   FIG. 1.   Disruption vector for gene disruption of TRI101. B, BamHI site; K, KpnI site; N, NdeI site; ATG, start codon of TRI101. Numbers reference primers used in PCR analysis. pTriR-3 is the chimeric plasmid. The middle diagram indicates the wild-type DNA with the crossover events; the bottom diagram indicates a gene replacement to form the gene disruption.

Germination and growth on toxin-containing media. To determine what effect the loss of TRI101 had on the growth of F. sporotrichioides in the presence of toxin, YPG agar plates containing 4,15-diacetoxyscirpenol (4,15-DAS) or T-2 toxin at 0, 100, 500, or 1,000 µg/ml were inoculated with a dilution containing approximately 10 spores of F. sporotrichioides NRRL 3299 or mutant strain FsTri101-3D. Plates were incubated at 28°C.

Trichothecene toxin assays. Liquid cultures of F. sporotrichioides were harvested after 7 days, extracted with ethyl acetate, and analyzed by gas-liquid chromatography (GLC) as described previously (19). Trichothecene concentrations were determined from the appropriate standard curves. Compound identifications were confirmed by GC-MS. For deoxynivalenol assays, trimethylsilyl ether derivatives were prepared with Tri-Sil TBT (Pierce, Rockford, Ill.).

Extraction and isolation of trichothecenes. We isolated trichothecenes produced by the mutant strain from liquid shake cultures of F. sporotrichioides FsTri101-3D grown for 7 days at 28°C and 200 rpm on YPD (1 liter in 2-liter Erlenmeyer flasks). Cultures were extracted twice with 600 ml of ethyl acetate, and the combined extracts were concentrated under reduced pressure. The syrup was separated on a silica gel column eluted with dichloromethane-methanol (95:5). Twelve 40-ml fractions were collected, and separation was monitored by thin-layer chromatography and GLC. Fractions 4 to 6 contained isotrichodermol. Fractions 8 to 10 contained 3,15-didecalonectrin.

Trichothecene standards. 4,15-DAS was isolated from F. sporotrichioides 1716cos9-1 (12); 15-monoacetoxyscirpenol was prepared by treating 4,15-DAS with 0.1 N NaOH; 3,4,15-triacetoxyscirpenol (3,4,15-TAS) was prepared from 4,15-DAS treated with acetic anhydride in pyridine. Isotrichodermin, 8-hydroxyisotrichodermin, and 8-hydroxyisotrichodermol were isolated from F. sporotrichioides mutant Allb (18); isotrichodermol was prepared by hydrolysis of isotrichodermin with 0.1 N NaOH. Deoxynivalenol and 15-acetyldeoxynivalenol were isolated from Gibberella zeae 3639 (24) grown on cornmeal. T-2 toxin was isolated from F. sporotrichioides NRRL 3299. 15-Decalonectrin and 3,15-didecalonectrin were isolated from F. sporotrichioides mutant strain O2 (19). All standards were greater than 95% pure as determined by GLC.

Whole-cell feeding experiments. Liquid cultures of F. sporotrichioides FsTri101-3D were inoculated with conidia washed from the V-8 plates, at a starting density of 5 × 10⁴ conidia/ml in 10 ml of GYEP medium in a 50-ml Erlenmeyer flask, and incubated on a gyratory shaker (200 rpm) at 28°C. After 24 h, a 25 mM stock solution of the trichothecene in acetone was added to the culture to a final concentration of 250 µM (1% acetone). Six substrates were tested: 4,15-DAS, 3,4,15-TAS, isotrichodermin, 15-decalonectrin, 3,15-didecalonectrin, and 8-hydroxyisotrichodermin. Control cultures had acetone added to a final concentration of 1% acetone. Cultures were incubated on a rotary shaker (200 rpm) at 28°C for up to six additional days and then were extracted with ethyl acetate and analyzed by GLC.

Cell-free system. Cell extracts of F. sporotrichioides NRRL 3299 and FsTri101-3D were made from liquid GYEP cultures incubated for 42 h on a gyratory shaker (200 rpm) at 28°C in the dark. Cultures were vacuum filtered, washed with nitrogen, and extracted with 3.5 ml of 10 mM potassium phosphate buffer (pH 7.0) containing 1 mM 2-mercaptoethanol. The extract was centrifuged at 3,000 × g for 10 min at 4°C, and the supernatant was decanted and centrifuged at 3,000 × g for an additional 5 min. Assays were run at 30°C and were initiated by the addition of 100 µl of the cell extract to a reaction mixture containing 250 µl of potassium phosphate buffer (pH 7.5), 10 µl of the trichothecene substrate in acetone (1 mg/50 µl of acetone), 100 µl of 20 mM MgSO₄, and 50 µl of acetyl coenzyme A (Sigma, St. Louis, Mo.) in water (25 mg/200 µl). Controls contained an additional 100 µl of potassium phosphate buffer in place of the cell extract. Immediately after addition of the cell extract and at 1, 2, and 4 h following addition, 100-µl aliquots of the reaction mixture were transferred to a glass vial containing 60 µl of ethanol. The samples were dried under a stream of nitrogen, redissolved in 50 µl of ethyl acetate or methanol, and analyzed by GLC. Six substrates were tested: isotrichodermol, 4,15-DAS, 15-monoacetoxyscirpenol, deoxynivalenol, 15-acetyldeoxynivalenol, and T-2 toxin.

Isolation of TRI101. Trichothecene-sensitive S. cerevisiae JG436 was used as a host for the expression of an F. sporotrichioides cDNA library. The increased trichothecene sensitivity of yeast strain JG436 results from the absence of a functional pleiotropic drug resistance gene, PDR5 (22). cDNA library construction employed the yeast expression vector pYES2 and mRNA harvested under growth conditions supporting maximum levels of trichothecene pathway gene expression. Following the initial transformation of JG436, cells were plated under plasmid selection conditions and the transformants were pooled. Dilutions of the transformant pools were then selected for resistance to trichothecenes by plating them onto media containing 4,15-DAS. Of the 45 resistant colonies isolated, 16 were analyzed further. Plasmids rescued from these transformants could retransform JG436 to the trichothecene resistance phenotype. Of the 16 analyzed clones, 14 contained the same cDNA.

Nucleotide sequence accession number. The nucleotide sequence of TRI101 has been submitted to GenBank under accession no. AF127176.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

TRI101 expression in yeast. Analysis of trichothecenes from liquid cultures of yeast transformed with TRI101 amended with 4,15-DAS indicated that about 50% of the 4,15-DAS was converted to 3,4,15-TAS in 3 days. Based on the proposed pathway for biosynthesis of T-2 toxin (5), we think that isotrichodermol is the likely substrate for TRI101 in the T-2 toxin pathway. Isotrichodermol is the first intermediate in Fusarium pathways to possess the trichothecene core structure and also has a C-3 hydroxyl group.

Characterization of TRI101. Analysis of the cDNA and corresponding genomic sequences for TRI101 indicated that the coding region consists of 1,380 bp with no introns and that TRI101 encodes a protein of 459 amino acids. A putative TRI6 binding site with the sequence TNAGGCCT (10, 25) is located 315 bp upstream of the start codon. Cosmid clones containing TRI101 do not appear to overlap cosmids carrying the other described trichothecene pathway genes, in agreement with published reports indicating TRI101 is not closely linked to the pathway gene cluster (13, 14).

Disruption of TRI101. Disruption of TRI101 was accomplished via transformation of wild-type strain F. sporotrichioides NRRL 3299 with plasmid pTriR-3, which included the entire TRI101 coding region into which the chimeric hyg gene had been inserted. Four of sixteen transformants, e.g., FsTri101-3D, were identified as TRI101 disruptants by PCR analysis with primers 955 or 1134 paired with 957, 997, or 1135 (Fig. 1). The single fragment amplified from FsTri101-3D genomic DNA with each primer combination was 2.5 kb larger than the single product amplified from genomic DNA of the wild-type progenitor strain. The increased size of FsTri101-3D PCR products is consistent with disruption of TRI101 via two homologous recombination events between pTriR-3 and chromosomal TRI101 sequences, since the hyg gene is 2.5 kb (Fig. 1). If pTriR-3 had integrated ectopically or via a single homologous recombination event, PCR with primer pairs 955 and 957 would have yielded two amplification products: a 2.0-kb fragment corresponding to the native TRI101 and a 2.7-kb fragment corresponding to the hyg-interrupted TRI101 in pTriR-3. The PCR results were confirmed via Southern hybridization with BamHI/XbaI-digested genomic DNA hybridized to a probe consisting of the TRI101 coding region (data not shown). In the TRI101-disrupted strains the hybridizing band was 2.5 kb larger than the hybridizing band from the wild-type progenitor strain.

View larger version (25K): [in this window] [in a new window]   FIG. 2.   Chromatograms of F. sporotrichioides culture extracts analyzed by GLC. Cultures were grown on GYEP medium for 7 days. (A) NRRL 3299 (TRI101+) chromatogram. The locations of T-2 toxin, DAS, and apotrichodiol are indicated. (B) Strain FsTri101-3D (mutant TRI101). The locations of isotrichodermol, didecalonectrin, 3-decalonectrin, and apotrichodiol are indicated.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The sequence of TRI101 from F. sporotrichioides NRRL 3299 has 94% similarity to the sequence from the previously reported TRI101 (14), resulting in a 93% protein similarity. No sequence similarities between TRI101 and the several fungal O-acetyltransferases that have significant similarity to each other were detected (8, 16, 17). Although TRI101 also appears unrelated to TRI3 (13), a trichothecene 15-O-acetyltransferase encoded by a gene that resides within the pathway gene cluster (19), both TRI3 and TRI101 share distinct motifs with a group of plant acyltransferases and their genes may be part of an extended acyltransferase gene family. TRI101 is similar in size to these plant acyltransferases and is conserved in both sequence and position for two motifs thought to be involved in catalysis (30).

The regulation of TRI101 in the presence of exogenously added trichothecenes is rapid and unlike that of other trichothecene biosynthetic genes (13). However, there is a putative TRI6 binding site (TNAGGCCT) (10) located in the TRI101 promoter region, which suggests that TRI101 may be regulated by the pathway transcription factor TRI6.

Disruption of the TRI101 gene in F. sporotrichioides confirms that it encodes an enzyme essential for T-2 toxin biosynthesis. A role for TRI101 as a self-protection mechanism in Fusarium was proposed (13) based on the fact that acetylation, like other modifications of the C-3 hydroxyl group of trichothecenes, significantly reduces their toxicity (13, 28). Although we found that yeasts transformed with TRI101 have increased tolerance of 4,15-DAS, the loss of TRI101 in F. sporotrichioides does not result in a significant inhibition of growth on trichothecene-containing media. Therefore, F. sporotrichioides most likely has additional self-protection mechanisms such as the toxin efflux pump encoded by TRI12 (2).

The possibility that multiple 3-O-acetyltransferases are present in Fusarium species has been proposed (14). No residual trichothecene 3-O-acetyltransferase activity was detected in mycelial extracts of the TRI101 disruptant strain, arguing against the possibility that TRI101 activity was still expressed in this mutant or that other trichothecene 3-O-acetyltransferases were expressed in F. sporotrichioides.

Mutants lacking a functional TRI101 produced an altered profile of trichothecenes. Specifically, no T-2 toxin was synthesized and isotrichodermol accumulated as the major trichothecene component. Isotrichodermol accumulation suggests that 3-O-acetylation is required for the subsequent C-15 oxygenation step to occur. Formation of small amounts of 3,15-didecalonectrin and 3-decalonectrin may be due to the presence of relatively high concentrations of isotrichodermol and the ability of the TRI11-encoded C-15 monooxygenase (1) and the TRI3-encoded C-15 acetyltransferase (19) to accept isotrichodermol and didecalonectrin as substrates under these conditions. We think that TRI101 acts as an integral biosynthetic enzyme and that while C-3 acetylation of other trichothecenes may occur, conversion of isotrichodermol to isotrichodermin is the primary biosynthetic reaction catalyzed by TRI101.

The disruptant strain FsTri101-3D can convert isotrichodermin and 15-decalonectrin, both of which have a C-3 acetyl, to T-2 toxin, suggesting that biosynthetic steps after that performed by the TRI101 product are not affected by the disruption reported. Our results are consistent with the hypothesis that most of the F. sporotrichioides trichothecene biosynthetic pathway is via intermediates with a C-3 acetyl protecting group. Mutants with disrupted TRI3 and TRI11 accumulate 15-decalonectrin and isotrichodermin, respectively, indicating that trichothecenes with an acetyl group at C-3 accumulate as predicted. 3,4,15-TAS was not converted to T-2 toxin by the disrupted strain FsTri101-3D, which suggests that there may be deacetylation and reacetylation required at C-3 in the later steps of biosynthesis.

Accepting TRI101 as part of an intracellular protection mechanism implies the existence of an extracellular esterase to activate trichothecenes. In cultures of mutants that accumulate trichothecenes with a C-3 acetyl group, e.g., 15-decalonectrin or isotrichodermin, these acetylated compounds are slowly deacetylated. Four esterases that may remove acetyl groups from various positions of the trichothecene skeleton are known (23). Kimura et al. (14) reported that there was no conversion to an acetylated product when T-2 toxin was added to F. sporotrichioides cultures, although crude extracts of recombinant TRI101 from Escherichia coli and from F. graminearum cultures both could convert T-2 toxin to 3-acetyl T-2 toxin. They interpreted this observation as evidence that TRI101 was not strongly induced by T-2 toxin. An alternate hypothesis is that competing esterases mask TRI101 expression by removing the C-3 acetyl group in the final steps in T-2 toxin biosynthesis. This hypothesis is supported by our results with several trichothecenes containing a C-3 acetyl group. When these compounds were added to cultures of the disruptant strain FsTri101-3D, deacetylation did not appear to be blocked and deacetylated products were recovered.

The presence of a 3-hydroxyl group differentiates Fusarium trichothecenes from most other fungal trichothecenes. Even though the early enzymes and regulatory mechanisms for the biosynthesis of macrocyclic trichothecenes in Myrothecium roridum are related to their counterparts in Fusarium trichothecene pathways (31), macrocyclic trichothecenes do not contain a 3-hydroxyl group. Therefore, identification of isotrichodermol as the primary TRI101 substrate makes this enzyme essential for biosynthesis of the most toxic Fusarium trichothecenes. TRI101 therefore appears to be a critical and unique enzyme in the Fusarium trichothecene biosynthetic pathway.