OUP user menu

Molecular biology of mycotoxin biosynthesis

Michael J. Sweeney, Alan D.W. Dobson
DOI: http://dx.doi.org/10.1111/j.1574-6968.1999.tb13614.x 149-163 First published online: 1 June 1999

Abstract

Mycotoxins are secondary metabolites produced by many important phytopathogenic and food spoilage fungi including Aspergillus, Fusarium and Penicillium species. The toxicity of four of the most agriculturally important mycotoxins (the trichothecenes, and the polyketide-derived mycotoxins; aflatoxins, fumonisins and sterigmatocystin) are discussed and their chemical structure described. The steps involved in the biosynthesis of aflatoxin and sterigmatocystin and the experimental techniques used in the cloning and molecular characterisation of the genes involved in the pathway are described in detail. The biosynthetic genes involved in the fumonisin and trichothecene biosynthetic pathways are also outlined. The potential benefits gained from an increased knowledge of the molecular organisation of these pathways together with the mechanisms involved in their regulation are also discussed.

Keywords
  • Mycotoxin
  • Biosynthetic pathway
  • Cloning

1 Introduction

Mycotoxins are a group of secondary metabolites which are produced by various filamentous fungi, and which can cause a toxic response, termed a mycotoxicosis, if ingested by higher vertebrates and other animals. The mycotoxigenic fungi involved in the human food chain belong mainly to three main genera: Aspergillus, Penicillium and Fusarium. The former two genera are commonly found as contaminants of foods during drying and storage while the latter are plant pathogens which produce mycotoxins before, or immediately after harvesting.

Aflatoxins are the group of mycotoxins which are of greatest significance in foods and feeds, and are produced mainly by Aspergillus flavus and Aspergillus parasiticus. The four main aflatoxins produced are B1, B2, G1 and G2. They are difuranocoumarin derivatives (Fig. 1), with the B and G nomenclature deriving from the blue and green fluorescent colours produced under UV light on thin layer chromatography plates; with the subscript numbers indicating major and minor compounds respectively. Aflatoxin B1 (AFB1) is widely regarded as the most potent liver carcinogen known for a wide variety of animal species, including humans [1]. Aflatoxin M1 and M2 are hydroxylated derivatives of AFB1 and AFB2, which are formed and excreted in the milk of lactating animals including humans, that have consumed aflatoxin-contaminated foods. A precursor of the aflatoxins, sterigmatocystin (ST), is a mycotoxin which is characterised by a xanthone moiety fused to a dihydrodifuran or tetrahydrofuran moiety; it is also acutely toxic and carcinogenic [2].

Figure 1

Common toxins produced by Aspergillus and Penicillium species.

The trichothecenes are a group of mycotoxins which are produced by several fungal genera including Fusarium, Trichoderma, Trichothecium, Stachybotrys, Cephalosporium and Myrothecium. The trichothecenes are chemically very diverse (Fig. 2), but are all tricyclic sesquiterpenes with a 12,13-epoxy-trichothec-9-ene ring. They can be designated into four subclasses: type A having a functional group other than a ketone at position C-8; type B having a ketone at position C-8; type C having a second epoxy group at C-7, C-8, C-9 or C-10; and type D containing a macrocyclic ring between C-4 and C-5 with two ester linkages [2]. The non-macrocyclic trichothecenes are frequently less cytotoxic, e.g. T-2 toxin, diacetoxyscirpenol and deoxynivalenol, and are primarily produced by Fusarium equiseti, F. graminearum and F. sporotrichioides, whilst the more complex macrocyclic trichothecenes are commonly associated with members of the genus Myrothecium [3, 4]. Interestingly, two members of the Brazilian plant genus, Baccharis, produce macrocyclic trichothecenes, although evidence suggests that the mycotoxin may also be synthesised by an endophytic fungus [4]. Trichothecenes are known to cause alimentary toxic aleukia, fusariotoxicoses and to be cytotoxic to mammalian cells. In addition they are immunotoxic and potent inhibitors of protein synthesis, which can result in a predisposition to other diseases and mask the underlying toxicoses [5]. In addition to their role in animal and human health, many trichothecene-producing fungi are phytopathogenic and its has been suggested that trichothecenes may function as virulence factors in plant disease [4, 6, 49].

Figure 2

Important toxins produced by Fusarium species.

The fumonisins are another important group of mycotoxins produced primarily by the cereal pathogen, Fusarium moniliforme. A number of other fungal species also produce fumonisins including F. proliferatum, F. anthophilum, F. dlamini, F. napiforme and Alternaria alternata f. sp. lycopersici [7]. Their chemical structure, which is a C-20, diester of propane-1,2,3-tricarboxylic acid and a pentahydroxyicosane containing a primary amino group (Fig. 2), resembles sphingosine (sphinganine), which forms the backbone of sphingolipids. The fumonisins are competitive inhibitors of sphingosine N-acetyltransferase which results in the blocking of complex sphingolipid biosynthesis and the accumulation of sphingosine [7]. The most abundant fumonisin produced in nature is fumonisin B1 (FB1) which can cause leukoencephalomalacia in horses and pulmonary oedema syndrome in pigs and is hepatotoxic and hepatocarcinogenic in rats [7, 8]. Fumonisins may also be implicated in the epidemiology of oesophageal cancer in humans. Although no experimental evidence exists the occurrence of F. moniliforme-infected maize and production of fumonisins has been correlated with a higher incidence of oesophageal cancer in specific geographical regions of China and South Africa [7, 9].

2 Biosynthetic pathways

2.1 Aflatoxins and sterigmatocystin

The aflatoxin biosynthetic pathway is well understood and has recently been reviewed [10, 11, 50]. Initially, acetate and malonyl CoA are converted to a hexanoyl starter unit by a fatty acid synthase, which is then extended by a polyketide synthase to norsolorinic acid, the first stable precursor in the pathway. The polyketide then undergoes approximately 12–17 enzymatic conversions, through a series of pathway intermediates, which are summarised in Fig. 3. Following the formation of versicolorin B, the pathway branches to form AFB1 and AFG1 which contain dihydrobisfuran rings and are produced from demethylsterigmatocystin (DMST); and the other branch forms AFB2 and AFG2, which contain tetrabisfuran rings and are produced from dihydrodemethylsterigmatocystin (DHDMST).

Figure 3

Aflatoxin and sterigmatocystin biosynthetic pathway. Enzymes involved: (a) fatty acid synthase, (b) polyketide synthase, (c) norsolorinic acid reductase, (d) versiconal hemiacetal acetate reductase, (e) esterase, (f1) versicolorin B synthase, (f2) versiconyl cyclase, (g) desaturase, (h) O-methyltransferase (MT-II), (i) O-methyltransferase, (j) O-methyltransferase (MT-I). Genes involved in aflatoxin biosynthesis: [A] fas1A and fas2A, [B] pksA, [C] nor1, norA, [D] avnA, [E] avf1 (aflB and aflW), [F] vbs, [G] verB, [H] ver1A, aflS, [I] omtA and [J] ord1. Genes involved in sterigmatocystin biosynthesis: [a] stcJ and stcK, [b] stcA, [c] stcE, [d] stcF, [e] stcI, [f] stcN, [g] stcL, [h] stcS, [i] stcU and [j] stcP. (Compiled from [11, 18, 21, 50].)

While aflatoxins are produced only by certain strains of A. parasiticus, A. flavus and A. nominus, numerous ascomycetes and deuteromycetes including A. nidulans produce the mycotoxin sterigmatocystin (ST), the penultimate intermediate in the AF biosynthetic pathway. The ST pathway is believed to include at least 15 enzymatic activities involving each enzyme activity from the AF pathway bar the penultimate steps involving the conversion of ST to AF.

Several of the enzymes involved in the AF pathway have been purified to homogeneity (for review see [11]). A 78-kDa versicolorin B synthase enzyme has been isolated which is involved in the cyclisation of versiconal to versicolorin B. This enzyme is believed to be the pivotal enzyme in determining the stereochemistry of the bisfuran ring in the aflatoxins. In addition, a tetrahydrobisfuran cyclising enzyme vericonyl cyclase has been purified which is responsible for the conversion of versiconal to versicolorin B. Three different enzymes have been characterised which appear to be involved in the conversion of norsolorinic acid to averantin: a 38-kDa norsolorinic reductase (NAR), a 43-kDa isozyme of the reductase and a 140-kDa NAR. Two versiconal hemiacetal (VHA) reductases (VHA I and II), which convert VHA to versiconal acetate, have been purified from A. parasiticus. Also three esterases which catalyse the conversion of versiconal acetate to versiconal acetate have recently been isolated.

A number of methyltransferases involved in the pathway have also been characterised. A 168-kDa O-methyltransferase and a 40-kDa methyltransferase corresponding to MT-II, which are involved in the conversion of ST to O-methylsterigmatocystin and dihydrosterigmatocystin to dihydro-O-methylsterigmatocystin, respectively, have been purified. More recently another O-methyltransferase (MT-1), with a molecular mass of 150 kDa, which is involved in the conversion of DMST to ST and of DHDMST to dihydrosterigmatocystin (DHST), has been purified [12].

2.2 Trichothecenes

The trichothecene pathway begins with the cyclisation of farnesyl pyrophosphate (FPP) to trichodiene by the enzyme trichodiene synthase (Fig. 4). This is the only enzyme in the biosynthetic pathway that has been purified and characterised to date, and the dimer, with a subunit molecular mass of 45 kDa, has been isolated from four fusaria including Fusarium sporotrichioides [13]. The subsequent pathway involves a number of oxygenations, isomerisations, cyclisations and esterifications leading from trichodiene to diacetoxyscirpenol, T-2 toxin and 3-acetyldeoxynivalenol. All of the intermediates except those involved in the earlier steps of the non-macrocyclic biosynthetic pathway have been confirmed by feeding studies [14]. In contrast, the macrocyclic biosynthetic pathway is much less understood; only the end products and late intermediates of the pathway have been isolated and characterised [15].

Figure 4

Trichothecene biosynthetic pathway in Fusarium species. Genes involved: [A] Tri 5, [B] Tri 4 and [C] Tri 3. (Taken from [44].)

2.3 Fumonisins

Fumonisins are thought to be synthesised through the condensation of the amino acid alanine to an acetate-derived precursor. Branched-chain methyl groups are added at C-12 and C-16 by an S-adenosyl methionine transferase. The subsequent biosynthetic steps involving oxygenation and esterification of the acetate-derived backbone are as yet unknown. It is not clear whether oxygenation and methylation occur before or after condensation with alanine. However, it appears likely that less oxygenated trichothecenes such as FB2, FB3 and FB4 are precursors of the more highly oxygenated FB1 (Fig. 5) [16]. To date, no enzymes involved in the fumonisin biosynthetic pathway have been isolated.

Figure 5

Proposed pathway for fumonisin biosynthesis. R1 designates tricarballylic acid esters. Genes involved: [A] fum 1, [B] fum 3 and [C] fum 2. (Taken from [16].)

3 Cloning and molecular characterisation of the aflatoxin and sterigmatocystin biosynthetic genes

Early studies on the genetics of AF/ST biosynthesis employed A. flavus and A. parasiticus mutants, either partially or fully blocked in AF production, indicating the possibility that some of the AF genes were clustered [17]. Molecular genetics confirmed these initial reports, as the development of efficient DNA transformation systems for both A. flavus and A. parasiticus allowed the elucidation of genes corresponding to the enzyme activities of the AF/ST pathway [10, 11, 18]. Two gene transfer techniques, gene complementation which involves the restoration of gene function in AF blocked mutants and gene disruption where ‘knock-out’ strains are formed, have been invaluable in assigning the function of the isolated AF/ST genes. The AF biosynthetic genes are clustered in A. flavus and A. parasiticus and the gene cluster has been located on a 4.9-Mb chromosome in A. flavus [19]. Mapping studies and complementation of an AF gene cluster-deleted mutant of A. flavus indicated that all of the cloned aflatoxin biosynthetic genes are located within a 75–90-kb region [20, 21]. The physical order of the genes in the cluster appears to largely coincide with the sequential enzymatic steps of the pathway and both gene organisation and structure are very conserved within A. flavus and A. parasiticus. The significance of this gene clustering is not known although the involvement of chromosome structure in gene regulation may be possible. In addition the conserved nature of the AF gene cluster suggests that the function or regulation of AF biosynthesis may rely on an intact structural organisation [10]. Many of the genes encoding enzymes involved in the AF biosynthetic pathway in both A. flavus and A. parasiticus have been cloned (Table 1A and Fig. 3). The genetics of the ST biosynthetic pathway of A. nidulans has been recently elucidated, which has furthered our understanding of the AF pathway (Table 1B and Fig. 3). A 60-kb region of chromosome IV of A. nidulans has been identified, which encodes 25 co-regulated transcripts which are thought to encompass most or all of the genes involved in ST biosynthesis [22]. The functions of many of these stc genes have been assigned experimentally, whilst the putative functions of the others have been assigned from the identity of the deduced amino acid sequences to enzymes predicted to the involved in the ST/AF biosynthetic pathway. The ST biosynthetic genes in A. nidulans are functionally and physically conserved with the AF genes of A. flavus and A. parasiticus, although differences in gene order and the direction of transcription are evident. The characterisation of the genes of the AF/ST biosynthetic pathway, together with the techniques employed for their isolation and confirmation of function, will be outlined.

View this table:
Table 1

Identity of AF pathway genes (A) and ST pathway genes (B) with other protein sequences

GenePutative activityAccession numberGene product (aa) (kDa)Similar polypeptides (aa)% identityNumber of aa comparedOrganismAccession number
A: AF pathway genes
fas1Afatty acid synthase αL481831980β-subunit of FAS1147159Saccharomyces cerevisiaeM31034
β-subunit of FAS1240345S. cerevisiaeM31034
pksA/pksL1polyketide synthaseZ471982109StcA/pksST (2181)64overallAspergillus nidulansL39121
wA PKS364100A. nidulansX65866
wA PKS432130A. nidulansX65866
type 1 PKS5 (3519)29100Streptomyces antibioticusL09654
type 1 PKS6 (10288)25100Saccharopolyspora erythraeaM63677
nor1dehydrogenaseL27801271 (29)StcE (260)56overallA. nidulansU34740
PKS7 (272)32103Streptomyces violaceoruberX16300
VER1A (262)26120Aspergillus parasiticusM91369
NAM dehydrogenase8 (272)23.2211Flavobacterium sp. 141-8D90316
norA/adh2dehydrogenaseU24698388 (43.7)NORA (388)99overallAspergillus flavusU32377
AAD9 (385)49overallPhanaerochaete chrysosporiumL08964
ADH1 (349)23overallA. flavusL27434
NOR1 (271)22overallA. parasiticusL27801
avnACYP-450 monooxygenaseU62774495 (56.3)StcF (506)66overallA. nidulansU34740
StcL (500)37overallA. nidulansU34740
StcB (435)15.6overallA. nidulansU34740
StcS (505)9.3overallA. nidulansU34740
vbsoxidase/dehydrogenaseU51327643 (70.3)GOX10 (605)38overallA. nigerX16061
CDH11 (556)34overallEscherichia coliX52905
ver1AketoreductaseM91369262VER1B12 (86)95overallA. parasiticusU63994
StcU/VERA (397)85overallA. nidulansL27825
T4HN13 (283)56overallMagnaporthe griseaL22309
ketoreductase14 (261)52overallStreptomyces coelicolorM1953141
omtA/omt1O-methyltransferaseL25834428 (46)OMT1 (418)97overallA. flavusL25836
ord1CYP-450 monooxygenaseU81806528 (60.2)StcF (506)69overallA. nidulansU34740
ord2unknownL40840286 (30.6)StcO (297)52overallA. nidulansU34740
StcQ (274)30overallA. nidulansU34740
aflR/apa2transcription factorL26220437 (46.7)AFLR/AFL2 (437)>95overallA. flavusL32577
AFLR (433)7142A. nidulansU34740
AFLR (433)31overallA. nidulansU34740
adh1alcohol dehydrogenaseL27434349ADH115 (349)82overallA. nidulansM16196
ADH316 (352)82overallA. nidulansX02764
ADH117 (348)57overallA. nidulansJ01313
aflJunknownAF002660438AFLJ (435)>95overallA. flavusAF0077975
B: ST pathway genes
stcJfatty acid synthase αU343701559FAS21844overallPenicillium patulumM37461
stcKfatty acid synthase βU343701914FAS119 (2076)37overallYarrowia lipolyticaX53868
stcA/pksSTpolyketide synthaseL391212181wA PKS3,442overallA. nidulansX65866
stcEketoreductaseU34740260NOR156overallA. parasiticusL27801
stcFCYP-450 monooxygenaseU34740506StcL41overallA. nidulansU34740
ORD169overallA. parasiticusU81806
stcIlipase/esteraseU34740276lipase20 (433)sig. ident.overallMoraxella sp. TA144X53868
lipase21 (308)sig. ident.overallPseudomonas sp. B11-1AF034088
stcNGMC oxidoreductase32U34740unp seq.GDH22 (612)sim. ident.overallDrosophila melanogasterM29298
CDH11 (556)sim. ident.overallE. coliX52905
GOX10 (605)sim. ident.overallA. nigerX16061
MOX23 (664)sim. ident.overallHansenula polymorphaX02425
stcLCYP-450 monooxygenaseU34740500StcF (506)41overallA. nidulansU34740
ORD1 (528)40overallA. parasiticusU81806
CYP-450 monooxygenase24 (506)29overallNectria haematococcaX73145
stcS/verBCYP-450 monooxygenaseU34740505CYP 4A1125 (519)23overallHomo sapiensS67581
stcU/verAketoreductaseU34740264VER185
stcPO-methyltransferaseU34740unpub. seq.OMT1 (428)sig. ident.overallA. parasiticusL22091
O-methyltransferase26 (306)sig. ident.overallS. erythraeaX60379
N-methyltransferase27 (381)sig. ident.overallS. erythraeaX51891
stcBCYP-450 monooxygenaseU34370435CYP-450 monooxygenase24 (506)24overallN. haematococcaX73145
stcCoxidaseU34370311Chloroperoxidase (321)29overallCaldariomyces fumagoM19025
stcGdehydrogenaseU34740unpub. seq.budABC operon encoded polypeptides28sig. ident.overallKlebsiella terrigenaL04507
α-acetolactase decarboxylase29 (259)sig. ident.overallEnterobacter aerogenesL04506
stcOunknownU34740297ORD2 (286)52overallA. parasiticusL40840
StcO (297)30overallA. nidulansU34740
stcQunknownU34370274ORD2 (286)30overallA. parasiticusL40840
StcO (297)30overallA. nidulansU34740
stcTelongation factor 1γU34370215elongation factor 1γ30 (430)37overallArtemia sp.M28020
stcVdehydrogenaseU34370387AAD31 (385)46overallP. chrysosporiumL08964
stcWFAD monooxygenaseU34370488cyclohexanone monooxygenase (542)27overallAcinetobacter sp. NCIB 9871M19029
  • Protein sequences included in this analysis are as follows: 1, FAS1, β-subunit enoyl reductase domain of β-subunit of fatty acid synthetase; 2, malonyl/palmityl transferase domain of β-subunit of fatty acid synthetase; 3, β-ketoacyl-acyl-carrier protein synthase functional domain of wA polyketide synthase; 4, acyltransferase domain of wA polyketide synthase; 5, polyketide synthase putatively involved in oleandomycin biosynthesis; 6, erythromycin-producing polyketide synthase (eryA gene); 7, granaticin-producing polyketide synthase; 8, N-acyl-d-mannosamine dehydrogenase (nam gene); 9, ligninolytic aryl-alcohol dehydrogenase; 10, glucose oxidase; 11, choline dehydrogenase (betA gene); 12, non-functional truncated polypeptide; 13, tetrahydroxynaphthalene reductase (scytalone reductase); 14, ketoreductase (actIII gene); 15, alcohol dehydrogenase I (alcA gene); 16, alcohol dehydrogenase III; 17, alcohol dehydrogenase I; 18, α-subunit of fatty acid synthase; 19, β-subunit of fatty acid synthase; 20, triacylglycerol lipase (lip2 gene); 21, cold-adapted lipase (lipP gene); 22, glucose dehydrogenase; 23, methanol oxidase; 24, phytoalexin pisatin demethylase (PDA6-1 gene); 25, fatty acid ω-hydrolyase; 26, erythromycin O-methyltransferase (eryG gene); 27, N-6-aminoadenine-N-methyltransferase (ermE gene); 28, α-acetolactase decarboxylase (budA gene), α-acetolactate synthase (budB gene), acetoin (diaceyl) reductase (budC gene); 29; α-acetolactase decarboxylase (budA gene); 30, translation elongation factor; 31, aryl alcohol dehydrogenase; 32, glucose/methanol/choline oxidoreductase.

  • Abbreviations: unpub. seq., unpublished sequences; sig. ident., significant identity; sim. ident., similar identity.

Two of the genes of the ST gene cluster in A. nidulans, stcJ and stcK, encode the α- and β-subunit of a fatty acid synthase (FAS) which is specific for the formation of the hexanoate starter of ST. Disrupted stcJ/stcK mutants do not synthesise ST, but retain the ability to do so when provided with hexanoic acid [23]. Two functional homologues of stcJ and stcK were isolated in A. parasiticus, fas1A and fas2A. fas1A was cloned by complementation of an A. parasiticus double mutant (blocked at nor1, responsible for the conversion of norsolorinic acid, and at a preceding step) resulting in the restoration of norsolorinic acid (NA) production [24]. The subsequent disruption of fas1A prevented NA accumulation in a nor1-blocked A. parasiticus mutant which normally accumulated NA. Homology of the predicted product of nor1 with functional domains of the β-subunit of a yeast FAS suggests that the fas1A gene encodes a FAS which synthesises part of the hexanoate starter of AF. A second FAS gene, fas2A, has been located next to fas1A and is suspected to function as the β-subunit of this specific FAS [25].

The subsequent extension of the hexanoate starter by a specialised polyketide synthase (PKS) was supported by the isolation of stcA (formerly pksST). stcA was revealed by transcriptional mapping of genomic DNA cosmids of A. nidulans which hybridised to a nor1 cDNA fragment and a 55-kb deleted region of a non-sterigmatocystin-producing mutant, and shows significant amino acid identity to two other PKSs of A. nidulans [26, 27]. The functional homologue of stcA in A. parasiticus, pksA (formerly pksL1), was independently cloned by polymerase chain reaction (PCR) amplification with degenerate primers and by gene disruption of an O-methylsterigmatocystin-accumulating strain of A. parasiticus, which resulted in a mutant unable to produce NA [28, 29]. The next step in the AF/ST pathway, where NA is converted to averantin (AVN), may involve the putative ketoreductase encoding gene in A. nidulans, stcE, whilst subsequent conversion of AVN to averufin (AVF) may involve stcF which encodes a putative P-450 monooxygenase, although this has not been confirmed [22]. The apparent homologue of stcE, nor1, was isolated by the complementation of an A. parasiticus mutant which accumulated NA. The role of nor1 was confirmed by the disruption of nor1 in aflatoxigenic isolates of A. parasiticus which resulted in the accumulation of NA [30]. The identity of nor1 as an NADPH-dependent reductase was confirmed by a nor1/maltose-binding protein fusion assay where transformed Escherichia coli converted NA to AVN in the presence of NADPH. A second reductase capable of converting NA, which shows little homology to nor1, norA, was located in the AF gene cluster of A. flavus and A. parasiticus and was cloned using monoclonal antibodies to NAR [31]. stcV in A. nidulans encodes a similar deduced product to that of norA; however, the function of stcV has not been confirmed [22]. The putative homologue of stcF, avnA (formerly ord1), was initially identified from the region between ver1 and omtA and associated with an oxidoreductive step in the AF pathway. Disruption of avnA resulted in a non-aflatoxigenic averantin-accumulating mutant and precursor feeding studies with AF intermediates with this mutant indicated that avnA, which encodes a cytochrome P-450 type enzyme, is involved in the conversion of AVN to AVF [32].

The avf1 locus is believed to be involved in the conversion of AVF to VHA [21]. An AF gene cluster-deleted mutant of A. flavus transformed with a series of overlapping cosmids containing the AF cluster accumulated averufanin (AVNN) and AVF and was relieved by complementation with the avf1 locus. Sequence analysis of the avf1 locus revealed two genes, aflB and aflW, which encode products with similar amino acid identity to stcB and stcW respectively in A. nidulans but their functions are as yet unknown [18].

Conversion of versiconal acetate to versiconal and subsequently to versicolorin B (VER B) may involve the putative esterase gene stcI, which has no known AF gene homologue, and stcN, a putative oxidoreductase [22]. The homologue of stcN, vbs, was cloned using degenerate primers designed from the amino acid sequences of peptide fragments of the VBS protein which catalyses the conversion of versiconal to VER B [20].

The stcL gene encodes a P-450 monooxygenase putatively involved in the conversion of VER B to versicolorin A (VER A) as gene inactivation of stcL resulted in the accumulation of dihydrosterigmatocystin (DHST) [27]. verB, the apparent homologue of stcL in A. parasiticus and A. flavus, has been cloned [18]. The conversion of VER A and VER B to demethylsterigmatocystin (DMST) and dihydrodemethylsterigmatocystin (DHDMST) involves stcS and stcU as gene disruption of the two genes resulted in the accumulation of VER A [3335]. In addition, a disrupted double mutant (stcL and stcU) of A. nidulans accumulated VER B demonstrating the specific requirement of stcU for the conversion of VER B to DHDMST [27]. aflS, a gene similar to stcS, has been located between ver1A and avnA in the AF gene cluster [18]. ver1A (homologue of stcU), which encodes a NADPH-dependent ketoreductase involved in the conversion of VER A to ST, was isolated by the complementation of an A. parasiticus mutant that accumulated VER [36]. The subsequent conversion of DHDMST to DHST and DMST to ST may involve the methyltransferase stcP, which has no known AF gene homologue, as disruption of stcP results in the accumulation of DMST [35].

Homologues of omtA and ord1, genes involved in the final conversion steps of the AF biosynthetic pathway, are notably absent in the non-aflatoxigenic ST-producing A. nidulans. omtA (formerly omt1) was isolated by antibodies raised against the purified enzyme, OMT-A, and subsequently cloned into an E. coli expression system which overexpressed an OMT-A-β-galactosidase fusion protein capable of converting ST to O-methylsterigmatocystin (OMST) [37]. The ord1 gene was identified by transforming an A. flavus AF gene cluster-deleted mutant with a 3.3-kb genomic fragment and the regulatory gene aflR of the AF gene cluster which allowed the transformant to convert OMST to AFB1. Sequence analysis of the inserted AF fragment revealed ord1 which encodes a cytochrome P-450-type monooxygenase. Transformation of Saccharomyces cerevisiae with ord1 resulted in the ability to convert OMST to AFB1, indicating that the ord1 gene product is sufficient to complete the last step in the AF pathway [21].

There is growing evidence suggesting that gene expression is involved in the regulation of multiple parts of the AF/ST biosynthetic pathway. The observation of co-ordinate transcription of nor1, ver1 and omtA suggests that AF genes may be regulated, at least in part, at the transcriptional level by a common regulatory factor [37, 38]. A. flavus mutants blocked at aflR (previously afl2) could not convert various intermediates to AF and complementation of these mutants with afl2 restored the expression of several AF pathway enzyme activities, which is characteristic of a gene encoding a trans-acting regulatory factor [39]. aflR (previously apa2) was also isolated from A. parasiticus by complementation of a non-aflatoxigenic aflR mutant [39]. Preliminary studies demonstrated that the transcription of nor1, ver1 and omtA is activated by the aflR gene product AFLR [40]. Inactivation of aflR in A. nidulans results in the absence of expression of stcW, stcV, stcU and stcT transcripts and transformation of A. nidulans with the aflR homologue from A. parasiticus regulates ST production, demonstrating that the AFLR of A. nidulans is a functional homologue of the AFLR of A. parasiticus even though overall amino acid identity is low [41]. The predicted amino acid sequence of aflR contains a cysteine-rich zinc finger DNA-binding domain which is characteristic of some fungal transcriptional activators. The expression of aflR may be autoregulated as AFLR has been shown to specifically bind upstream of the AFLR translation start site [40]. aflJ, which is located adjacent to aflR, is required for the conversion of AF pathway intermediates to AF as disrupted strains of A. flavus at the aflJ locus do not accumulate any AF pathway intermediates and do not convert NA, ST or OMST to AF. Although speculative, the deduced amino acid sequence suggests that the aflJ product may be involved in transmembrane transport of AF intermediates or the localisation of AF pathway enzymes to an organelle [42].

4 Cloning and molecular characterisation of other mycotoxin biosynthetic genes

Several genes of the trichothecene biosynthetic pathway appear to be clustered in F. sporotrichioides [43]. Analysis of the gene cluster revealed nine genes within a 25-kb region and the function of eight of these genes has been assigned [6]. Two of these genes, Tri 3 and Tri 4, were identified following complementation of UV-induced mutants, blocked in trichothecene T-2 toxin production. Tri 3 encodes a 15-O-acetyltransferase, which converts 15-decalonectrin to calonectrin [43](Fig. 4). Tri 4 encodes a cytochrome P-450 monooxygenase involved in the first step in the pathway converting trichodiene to an as yet unidentified oxygenated product. Tri 11, a second cytochrome P-450 monooxygenase, has also been identified which appears to oxygenate the trichothecene ring at the C-15 position [44]. Two additional specific acetyltransferases may also be present in the gene cluster for the hydroxylation at the C-3 and C-4 positions [43]. The other biosynthetic gene, Tri 5, encodes trichodiene synthase involved in the cyclisation of farnesyl pyrophosphate to trichodiene (Fig. 4). The biosynthetic pathway appears to be regulated by the product of the Tri 6 gene, a Cys2, His2 zinc finger protein [45]. More recently a Tri 101 gene has been isolated from F. graminearum, encoding a protein which catalyses the acetyl CoA-dependent O-acetylation of the trichothecene ring at the C-3 position. This O-acetyl group introduction acts as a resistance mechanism for the type B trichothecene producer, F. graminearum [46]. Interestingly, the Tri 101 gene is located between a putative UTP-ammonia ligase gene and the phosphate permease gene and mapping analysis with two of the least overlapping cosmid clones containing Tri 101 revealed that this gene is not clustered with Tri 4, Tri 5 and Tri 6 [46]. In addition to the structural genes and transcription factor, a Tri 12 gene which encodes a putative transport protein has also been identified [6]. The genes of the biosynthetic pathway of macrocyclic trichothecenes have been investigated in Myrothecium roridum [4]. Homologues of the non-macrocyclic trichothecenes pathway genes Tri 4, Tri 5 and Tri 6 have been reported within a 40-kb region of M. roridum. The deduced amino acid sequences of the products of MrTri 6 and MrTri 4 are 75% and 63% identical and similar in molecular mass to the apparent counterpart proteins in F. sporotrichioides. However, MrTri 6 encodes a protein which is almost twice the size of the product of Tri 6 and only the C-terminal region containing the Cys2, His2 zinc finger motif shows significant homology (65% identity) to Tri 6 in F. sporotrichioides The putative cytochrome P-450 monooxygenase product of MrTri 4 appears to be a functional homologue of Tri 4 as complementation of a F. sporotrichioides mutant lacking Tri 4 resulted in the accumulation of T-2 toxin, sambucinol, deoxysambucinol and the intermediates, trichothecene and isotrichodiol. Although mapping data indicate that the macrocyclic genes of M. roridum are clustered, the organisation and orientation of these genes differ from those of the trichothecene gene cluster in F. sporotrichioides. In F. sporotrichioides the Tri 4, Tri 6 and Tri 5 genes are located in that order within an 8-kb region whilst their putative homologues are located within a 40-kb region in M. roridum. In addition, the relative orientation of Tri 6 and Tri 4 differs from that of MrTri 6 and MrTri 4 in M. roridum. While differences in gene organisation have been observed between the AF and ST pathways of A. parasiticus and A. nidulans, the differences in the trichothecene pathways of F. sporotrichioides and M. roridum are more pronounced [4, 10, 22, 38]. The less conserved gene organisation of the trichothecene pathways, in contrast to the AF/ST pathways, may be indicative of the presence of genes required for the production of unique structural features of the trichothecenes of F. sporotrichioides and M. roridum or may reflect the comparison of two more distantly related fungal species, i.e. F. sporotrichioides and M. roridum are taxonomically more distant than A. parasiticus and A. nidulans. Overall, it is evident that the clustering of genes for trichothecene biosynthesis is maintained in distantly related fungi and that the evolution of these gene clusters can involve substantial genetic rearrangements [4].

To date, no fumonisin biosynthetic genes have been cloned although several genes have been identified in F. moniliforme by classical genetics. Variants of Gibberella fujikuroi (F. moniliforme), blocked in the production of FB1 and whose phenotypes segregate as single genetic loci were identified by crossing with high-producing FB1 strains. Four classes of putative fumonisin biosynthetic genes, fum 1, fum 2, fum 3 and fum 4, were identified by this meiotic genetic analysis. Fum 1 represents strains that do not produce FB1, FB2, FB3 or FB4, whilst fum 4 represents a single strain that shows reduced fumonisin production and appears to be closely linked to fum 1. Fum 3 and fum 4 affect the hydroxylation of FB1 at C-10 and C-5 respectively and do not affect the overall level of fumonisin production. Both fum 3 and fum 4 are closely linked to fum 1. Tentative estimations suggest that fum 4 and fum 2 are situated 250 kb and 360 kb from fum 1; however, gene order has not been elucidated. Close linkage of the four genes indicates that the biosynthetic genes are arranged in a gene cluster on chromosome 1 of G. fujikuroi. These genetic data are consistent with the scheme in Fig. 5 where fum 2 may encode a C-10 hydroxylase that converts FB4 to FB3 and FB2 to FB1 while fum 3 may encode a C-5 hydroxylase that can convert FB4 to FB2 and FB3 to FB1[16]. However, the fum loci could alternatively encode regulatory genes and not the structural genes of the biosynthetic pathway.

5 Conclusions

The cloning and molecular characterisation of mycotoxin biosynthetic genes is vital in order to gain a fuller understanding of the organisation, regulation and expression of these genes. Firstly this will be valuable in our overall understanding of the number, type and order of the enzymatic steps involved in the various biosynthetic pathways and of the physiological factors controlling these processes. Secondly it will aid in the development of improved molecular-based detection systems for mycotoxins and mycotoxigenic fungi in food systems. For example, sequence variability in the aflR gene generates distinct DNA fingerprints which allow the non-aflatoxigenic species of A. sojae and A. oryzae in the Aspergillus flavi group to be distinguished from the aflatoxigenic species A. parasiticus and A. flavus [47]. In addition PCR has been successfully used to detect aflatoxigenic fungi in grains, using primers based on the coding regions of ver1, omt1 and aflR [48]. Finally this knowledge may allow (through the use of techniques such as gene disruption and AF gene/reporter constructs), the development of strategies for the biological control of mycotoxigenic fungi and the development of genetically engineered resistant crop plants.

The recent field application of atoxigenic strains of the trichothecene-producing wheat pathogen, Gibberella zeae (Fusarium graminearum), obtained by disruption of the Tri 5 gene, resulted in disease reduction and indicates the potential of genetically engineered atoxigenic fungi for the suppression of mycotoxigenic fungi in the field [49].

Acknowledgements

The authors acknowledge the Irish Department of Agriculture Food and Forestry for support under the Food Industry Sub Programme of the EU Structural Funds 1994–1999.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
View Abstract