The first stable aflatoxin precursor was confirmed to be norsolorinic acid (NOR) [59,60,88] (Figure 2). A hexanoyl starter unit is the initial substrate for aflatoxin formation [65]. Two fatty acid synthases (FAS) and a polyketide synthase (NR-PKS, PksA) are involved in the synthesis of the polyketide from a hexanoyl starter unit. Seven iterative, malonyl-derived ketide extensions are required to produce norsolorinic acid anthrone (noranthrone) [76,83,84,85,89,90,91,92,93,94,95,85,89]. Mahanti et al. [ 96] cloned, by genetic complementation, a 7.5-kb large transcript which is required for NOR formation in a blocked A. parasiticus mutant. Its protein has high degree of similarity (67%) and identity (48%) to the beta-subunit of FASs (FAS1) of Saccharomyces cerevisiae and Yarrowia lipolytica. Metabolite feeding and gene disruption experiments further confirmed that uvm8 encodes a subunit of a novel fatty acid synthase (FAS) directly involved in the backbone formation of the polyketide precursor of NOR during aflatoxin biosynthesis, therefore, on the basis of its function, the uvm8 gene was renamed fas-1A. In the revised naming scheme, the fas-1A gene was renamed as fas-1, it encodes fatty acid synthase-1 in the aflatoxin biosynthetic pathway gene cluster (Figure 1). Another large transcript (fas-2A) which encodes an alpha-subunit of fatty acid synthase in the aflatoxin gene cluster was reported [96]. The gene fas-1A and fas-2A were renamed as fas-1 and fas-2. They encode two fatty acid synthases (FASα and FASβ) [97]. In A. nidulans the involvement of FASs in sterigmatocystin (ST) biosynthesis was also confirmed and were named stcJ and stcK in the ST cluster [90,95]. The biochemical evidence for the role of a fatty acid synthase and a polyketide synthase (PKS) in the biosynthesis of aflatoxin was demonstrated [98]. Further details on the early stage of aflatoxin biosynthesis involving fatty acid synthases and polyketide synthases were reported [89,91,92,99]. The N-acetylcysteamine thioester of hexanoic acid was incorporated into NOR in a fas-1 disrupted transformant. A polyketide synthase gene (pksA) in A. parasiticus was demonstrated by gene disruption to be required for aflatoxin biosynthesis [73]. The predicted amino acid sequences of these PKSs contain the typical four conserved domains commonly found in other known PKS proteins: β-ketoacyl synthase (KS), acyltransferase (AT), acyl carrier protein (ACP), and thioesterase (TE) [73]. Townsend’s group has dissected the functional domains of the PKS for aflatoxin biosynthesis [76,93,100]. These include domains for the starter unit acyl transferase (SAT) which recognizes hexanoyl CoA and the N-acetylcysteamine thioester of hexanoic acid, the acyl carrier protein (ACP), ketosynthase (KS), malonyl-CoA:ACP transacylase (MAT), product template (PT) allowing the iterative steps in forming the polyketide, and a thioesterase/Claisen-like cyclase (TE/CLC) [76]. The predicted product converted by PksA is noranthrone. The conversion of noranthrone to NOR, the first stable intermediate in the pathway [53,60,62,69,101,102,103], is poorly defined, but it has been proposed to be catalyzed by a noranthrone oxidase, a monooxygenase, or to occur spontaneously [64]. Sequence analysis and enzymatic studies support the contention that the hypC (a gene in the intergenic region of pksA and nor-1) gene product is the required noranthrone oxidase involved in the catalysis of the orxidation of norsolorinic acid anthrone to NOR [80] The fas-1, fas-2, and pksA genes were renamed as aflA, aflB, and aflC respectively [81,83,87] (Figure 1). The aflA, aflB and aflC gene homologues in A. nidulans are stcJ, stcK, and stcA, respectively [90].

The first stable aflatoxin (AF) intermediate was identified as NOR produced in A. parasiticus uv-generated disruption mutants [60,69,103,104] and in A. flavus [53,102]. The NOR-accumulating mutants are leaky mutants whose aflatoxin biosynthesis is not completely blocked. By genetic complementation, the gene, aflD (nor-1), encoding a reductase was cloned [105]. A recombinant Nor-1 protein expressed in E. coli catalyzed the reduction of NOR. Therefore, aflD (nor-1) encodes the ketoreductase needed for the conversion of the 1'-keto group in NOR to the 1'-hydroxyl group of AVN [106]. Disruption of the aflD (nor-1) gene also confirmed its involvement in conversion of NOR to AVN in aflatoxin biosynthesis [107]. The aflD (nor-1) homologous gene in A. nidulans is stcE [90]. Genes homologous to aflD (nor-1), in the AF cluster, such as aflE (norA) and aflF (norB) are predicted to encode short chain aryl alcohol dehydrogenases. These proteins may also be able to catalyze the reduction of NOR to AVN depending on the reductive environment of the cell and may explain the leakiness of the nor-1 mutation if they are able to complement Nor-1’s function [108].

Radioisotope incorporation experiments provide the earliest evidence in establishing the conversion of AVN to HAVN [109,110]. There are three enzymatic steps that account for the conversion of NOR to averufin (AVF) [111]: (i) NOR to AVN catalyzed by a reductase, (ii) NOR to HAVN catalyzed by a monooxygenase, and (iii) HAVN to AVF catalyzed by a second dehydrogenase. It was also proposed that the oxidation reactions are reversible and that NADPH was the preferred cofactor [112]. The gene previously named ord-1 encoding a P-450 monooxygenase was cloned and disrupted [113]. Substrate feeding studies of the ord-1 mutant confirmed that HAVN is the intermediate in the conversion of AVN to AVF. The ord-1 gene, which has a high degree of sequence similarity to A. nidulans stcF [90], was renamed aflG (avnA).

Averufin is one of the key intermediates in aflatoxin formation [103,114,115,116,117,118,119]. Several intermediates were reported to be involved in the conversion from AVN to AVF [70,118]. One of these is averufanin (AVNN). Studies later demonstrated that it is a shunt metabolite and not a genuine aflatoxin intermediate [94,120]. The cluster gene aflH (adhA) in A. parasiticus was characterized to encode an alcohol dehydrogenase [103,118,121]. It was shown that adhA deletion mutants accumulated predominantly HAVN and after prolonged growth, the mutants were able to produce small amounts of AVNN consistant with AVNN being a shunt metabolite. Thus, HAVN might be converted directly to AVF or indirectly to AVF by an additional cytosolic enzyme. Sakuno et al. [120] characterized two cytosolic enzymes and a new aflatoxin intermediate named 5'-oxoaverantin (OAVN) as an intermediate between HAVN and AVF. The enzyme for the conversion from HAVN to OAVN is encoded by the aflH (adhA) gene. The adhA gene deletion mutant is leaky indicating that additional enzyme(s) or gene(s) may be involved in the conversion from OAVN to AVF. The enzymatic steps for aflatoxin biosynthesis and the possible involvement of additional enzymes have also been proposed [80,86,122]. The aflH (adhA) gene in A. flavus and the adhA gene in A. parasiticus share no significant homology at either the DNA or the amino acid level.

The conversion of AVF to VHA involves the cytochrome P450 monooxidase, CypX, and another gene, aflI (avfA). Although aflI is required for the conversion, its oxidative role is unclear [123]. A. nidulans also has an aflI gene homolog (stcO) [90,123]. Complementation of an averufin-accumulating mutant, A. parasiticus SRRC 165, with the aflI gene of A. flavus restored the strain’s ability to convert AVF to VHA and to produce aflatoxins [123]. It is likely that the aflI (avfA) encoded protein along with CypX gene product is involved in the ring-closure step in the formation of hydroxyversicolorone. It is possible that the avfA gene product is assocated with the P450 monooxygenase to carry out the conversion as no additional intermediates other that AVF result from the disruption of either gene.

It has been demonstrated that an esterase is involvement in the conversion of VHA to VHOH (VAL) [57,111,112,124,125,126,127,128]. The esterase was purified in A. parasiticus [125,126]. An esterase gene, aflJ (estA), in the aflatoxin gene cluster was identified [129]. The homologous gene in the A. nidulans ST biosynthetic gene cluster is stcI. In the A. parasiticus aflJ (estA) deletion mutants, the accumulated metabolites were mainly VHA and versicolorin A (VERA) [130]. A small amount of versiconol acetate (VOAc) and other downstream aflatoxin intermediates, including VHOH and versicolorin B also accumulated. A metabolic grid containing VHA, VOAc, VHOH, and versiconol (VOH) was previously described and it was suggested that the reactions from VHA to VHOH and from VOAc to VOH are catalyzed by the same esterase [111]. Later, another metabolic grid containing versicolorone (VONE), VOAc, and VHA was identified [131]. Indeed, it has now been proven that the estA-encoded esterase catalyzes the conversion of both VHA to VHOH and VOAc to VOH during aflatoxin biosynthesis [130].

The enzymatic evidence that VHOH is converted toVER B by a cyclase was first provided by Lin and Anderson [132]. This enzyme was identified as versicolorin B synthase and was studied intensively by Townsend’s laboratory [114,133,134,135,136]. The gene was cloned and named vbs [114,134,136]. The expected cyclase activity was demonstrated by the expressed recombinant protein of the vbs gene [134,135]. The VHOH cyclase [132] and VER B synthase [133] were independently isolated from A. parasiticus. The enzyme catalyzes the side chain cyclodehydration of racemic VHA to VER B. This is another key step in aflatoxin formation since it closes the bisfuran ring of aflatoxin, the moiety ultimately responsible for aflatoxin’s toxicity and carcinogenicity. The vbs gene was re-named aflK (vbs) [81]. The homologous gene in the A. nidulans ST biosynthetic gene cluster is stcN.

The critical branch point leading to the formation of either AFB₁/AFG₁ or AFB₂/AFG₂ is VER B. Similar to AFB₂/AFG₂, VER B contains a tetrahydrobisfuran ring and, like AFB₁/AFG_1, VER A contains a dihydrobisfuran ring. The conversion of VER B to VER A requires desaturation of the bisfuran ring of VER B by an unstable microsomal enzyme that requires NADPH [137]. Disruption of stcL in A. nidulans [138] abolished ST synthesis and resulted in the accumulation of VER B. The stcL gene encodes a cytochrome P-450 monooxygenase. The homologue, aflL (verB), is present in the aflatoxin gene cluster of A. parasiticus and A. flavus strains. Cultural conditions appear to markedly affect the activity of VER B desaturase and thereby, the final ratio of AFB₁ to AFB₂ and AFG₁ to AFG₂[94].

The biochemical conversion steps from VER A to DMST (and VerB to DHDMST) have been described in great detail [139]. The aflM (ver-1) gene [140], cloned by genetic complementation of VER A-accumulating A. parasiticus CS10 (a mutant strain created by knocking out the verA gene in aflatoxin-producing strain A. parasiticus SU-1), was shown to be responsible for the conversion of VER A to an intermediate that has not been isolated. The aflM (ver-1) gene was predicted to encode a ketoreductase, similar Nor-1. The ver-1 homologue, stcU, (previously named verA) was identified in A. nidulans [141]. Double mutation of stcU and stcL resulted in accumulation of only VER A [141]. The stcS gene(previously named verB), another cytochrome P-450 monooxygenase gene, was also identified and studies showed that it is also involved in the conversion of VER A to an intermediate in the formation of DMST (possibly the first intermediate, which is then acted upon by Ver-1). Disruption of stcS resulted in the accumulation of VER A as did disruption of Ver-1 [142]. Thus, both stcU and stcS are required for the conversion of VER A to DMST. The stcS homologue in A. parasiticus, named aflN (verA), has also been identified [81,87]. A third enzyme is required for the conversion: hypA (aflY). This gene is predicted to encode a Baeyer-Villiger monooxygenase. Disruption of this gene also led to accumulation of VERA suggesting that, like VER-1, it acts as part of an enzyme complex without allowing the formation of an intermediate. A fourth enzyme, OrdB has also been implicated in the conversion, and like AvfA, its homolog, may be a helper protein for the monooxygenase, CypX.

Two O-methyltransferases, (I and II), are confirmed to be involved in aflatoxin biosynthesis [143]. O-methyltransferase I catalyzes the transfer of the methyl from S-adenosylmethionine (SAM) to the hydroxyls of DMST and DHDMST to produce ST and DHST, respectively. This 43-kDa enzyme was purified from A. parasiticus and characterized [144,145]. The corresponding gene, dmtA, was isolated from A. parasiticus based on a partial amino acid sequence of the purified enzyme [146]. Yu et al. [123] concurrently isolated the same gene but named it aflO (omtB) (for O-methyltransferase B) from A. parasiticus, A. flavus and A. sojae. The predicted dmtA-encoded protein contains a consensus SAM-binding motif [146]. The aflO (omtB) homolog in A. nidulans was identified as stcP this gene is required for the conversion of DMST to ST in A. nidulans as shown by gene disruption [147].

The gene for O-methyltransferase required for the conversion of ST to OMST and DHST to DHOMST was first cloned [148] from A. parasiticus by reverse genetics using antibodies raised against the purified A. parasiticus O-methyltransferase A [78]. This gene was initially named omt-1, then omtA and finally renamed aflP (omtA) [148]. The recombinant enzyme was expressed in E. coli and its activity to convert ST to OMST was demonstrated by substrate feeding studies [148]. O-methyltransferase A has strict substrate-specificity and cannot methylate DMST or DHDMST. Thus, the O-methyltransferases A encoded by aflP (omtA) is the enzyme responsible for the conversion of ST to OMST and DHST to DHOMST. The genomic DNA sequence of this gene (omtA) was cloned from A. parasiticus and A. flavus [149]. This aflP (omtA) gene homologue was also detected in other aflatoxigenic and non-aflatoxigenic Aspergillus species [150]. The absence of the aflP orthologue in A. nidulans is the reason that A. nidulans produces ST as the end product instead of aflatoxins.

Based on feeding experiments, the relationship between B-group and G-group aflatoxin formation was proposed [151]. A P-450 monooxygenase gene in A. flavus named ord-1 was shown to be necessory for this reaction [152,153]. This P-450 monooxygenase gene, aflQ (ordA), was cloned in A. parasiticus and demonstrated in a yeast system that it is involved in the conversion of OMST to AFB₁/AFG₁, and DHOMST to AFB₂/AFG₂ [154]. Whether aflQ (ordA) gene product, OrdA, catalyzes two successive monooxygenase reactions in the later steps of aflatoxin biosynthesis is not clear. Studies [154] suggested that additional enzyme(s) is required for the synthesis of G-group aflatoxins. After the cloning and characterization of the cypA gene, it is clear that cypA encoded a cytochrome P450 monooxygenase for the formation of G-group aflatoxins [155]. Most recently, the nadA gene, which was shown, by gene profiling studies using microarray, to be a member of the aflatoxin gene cluster [156,157] rather than belonging to the adjoining sugar utilization cluster as originally proposed [158], was found to play a role in AFG₁/AFG₂ formation. Yabe’s group recently disrupted the nadA gene and reported that NadA is a cytosolic enzyme for the conversion from a new aflatoxin intermediate named NADA, which is between OMST and AFG₁, to AFG₁ [159]. The aflE (norA) gene was initially believed to be involved in the conversion of NOR due to certain degree of sequence similarity to the aflD (nor-1) gene [108]. However, recent studies support the hypothesis that the aflE (norA) is involved in the final two steps in AFB₁ formation [80]. In the same report, the transcript, hypB, a homolog of hypC, may be involved in one of the oxidation steps in the conversion of OMST to aflatoxins. A. flavus produces only AFB₁ and AFB₂, whereas A. parasiticus produces all four major aflatoxins, AFB₁, AFB₂, AFG₁, and AFG₂. Coincidently, only the G-group aflatoxin producer, A. parasiticus, has intact nadA and norB genes. Preliminary data suggests that norB encodes another enzyme predominantly involved in AFG₁/AFG₂ formation [160].

The aflatoxin M₁ and M₂ are mammalian bioconversion products of AFB₁ and AFB₂ respectively and are originally isolated and identified from bovine milk [161,162,163,164,165]. After entering the mammaliam body (human or animals), aflatoxins are metabolized by the liver cytochrome P450 enzymes to a reactive epoxide intermediate which becomes more carcinogenic, or be hydroxylated and become the less harmful aflatoxins M₁ and M₂. However, recent studies by feeding with aspertoxin (12c-hydroxy-OMST) [166] indicated that A. parasiticus produces the minor aflatoxins M₁ (AFM₁), M₂ (AFM₂), GM₁ (AFGM₁), and GM₂ (AFGM₂), as well as the major aflatoxins B₁ (AFB₁), B₂ (AFB₂), G₁ (AFG₁), and G₂ (AFG₂). It demonstrated that aspertoxin is a precursor of AFM₁ and AFGM₁. Feeding of the same fungus with O-methylsterigmatocystin (OMST), AFM₁ and AFGM₁ were formed together with AFB₁ and AFG₁; feeding with dihydro-OMST (DHOMST), AFM₂ and AFGM₂ were formed together with AFB₂ and AFG₂. It showed that the enzyme OrdA catalyzes both 12c-hydroxylation reaction from OMST to aspertoxin and the successive reaction from aspertoxin to AFM₁ and the AFB₁ is not a precursor of AFM₁.

A 47 kDa sequence-specific zinc-finger DNA-binding protein (AflR), encoded by the aflR gene, is required for transcriptional activation of most, if not all, the structural genes of the aflatoxin gene cluster [72,74,173,174,175,176,177,178,179,180]. Like other Gal4-type regulatory proteins that bind to palindromic sequences, functional AflR probably binds as a dimer. It binds to the palindromic sequence 5'-TCGN5CGR-3' in the promoter regions of the structural genes [77,181]. The AflR-binding motifs are found to be located from −80 to −600 positions, with the majority at the −100 to −200 positions, relative to the translation start site. AflR binds, in some cases, to a deviated sequence rather than the typical motif such as in the case of aflG (avnA). When there is more than one binding motif, only one of them is the preferred binding site such as in the case of aflC (pksA) [77,181]. The more upstream motif is found to belong to another gene for turning on the expression of hypC. Deletion of aflR in A. parasiticus abolishes the expression of other aflatoxin pathway genes [182]. Overexpression of aflR in A. flavus up-regulates aflatoxin pathway gene transcription and aflatoxin accumulation [176] in a fashion similar to that reported for A. parasiticus [173]. These results demonstrate that AflR is specifically involved in the regulation of aflatoxin biosynthesis. Indeed, all 23 upregulated genes, identified by transcription profiling using DNA microarray assays comparing wild-type and aflR-deleted A. parasiticus strains, have the consensus AflR binding motif in their promoter regions [156,168,183,184].

The aflS (aflJ) gene [168,185], bidirectionally transcribed from aflR, although not demonstrating significant homology with any other encoded proteins found in databases, is necessary for aflatoxin formation. The aflS and aflR share a 737 bp intergenic region. In the A. parasiticus aflR transformants, the production of aflatoxin pathway intermediates was significantly enhanced in transformants that contained an additional aflR plus aflS [173]. Quantitative PCR showed that in the aflS knockout mutants, the lack of aflS transcript is associated with 5- to 20-fold reduction of expression of some aflatoxin pathway genes such as aflC (pksA), aflD (nor-1), aflM (ver-1), and aflP (omtA). The mutants lost the ability to synthesize aflatoxin intermediates and no aflatoxins were produced [168]. However, deletion of aflS (aflJ) did not have a discernible effect on aflR transcription, and vice versa. Du et al. [186] showed that overexpression of A. flavus aflS (aflJ) did not result in elevated transcription of aflM (ver-1), aflP (omtA), or aflR, but it appears to have some effect on aflC (pksA), aflD (nor-1), aflA (fas-1), and aflB (fas-2) [186], which are required for the biosynthesis of the early aflatoxin pathway intermediates. The mechanism(s) by which aflS modulates transcription of these pathway genes in concert with aflR is under investigation.

The novel global regulatory gene, laeA (for lack of aflR expression), was first identified from A. nidulans [169]. This gene is well conserved in fungi as shown by its presence in the genomes of all fungi so far sequenced. LaeA is a nuclear protein which contains an S-adenosylmethionine (SAM) binding motif and activates transcription of several other secondary metabolism gene clusters in addition to the AF cluster. Examples include the sterigmatocystin and penicillin clusters in A. nidulans, the gliotoxin cluster in A. fumigatus, and aflatoxin cluster in A. flavus [169,187]. It also regulates genes required for virulence of A. fumigatus [188]. Perrin et al. [170] carried out a whole-genome comparison of the transcriptional profiles of wild-type and laeA-deleted A. fumigatus strains and found that LaeA positively controls the expression of 20% to 40% of major classes of secondary metabolite biosynthesis genes. It also regulates some genes not associated with secondary metabolite clusters. Similar results were confirmed in gene expression profiling in A. flavus using microarrays to study the genetic mechanism of sclerotium formation. The exact mechanism of how LaeA regulates secondary metabolism gene clusters is not yet known. Interestingly, when an unrelated gene such as argB was placed within the boundary of the ST gene cluster, it was co-regulated with other genes in the cluster. But, when a gene in the cluster, such as aflR was placed elsewhere in the genome, its regulation was not affected by LaeA [189]. One proposed regulatory mechanism is that LaeA differentially methylates histone protein and it alters the chromatin structure for gene expression. Unlike the mentioned signaling factors, the primary role of LaeA is to regulate metabolic gene clusters, not sporulation, based on the obesevation that laeA-deleted strains produced conidia equivalent to wild-type level [169]. Most recent analyses of nonaflatoxigenic A. parasiticus sec- (for secondary metabolism negative) variants generated through serial transfer of mycelia of the sec+ parents show that laeA was expressed in both sec+ and sec- strains [190]. This result suggests that LaeA only exerts its effect on aflatoxin biosynthesis at a certain level and is independent of other regulatory pathways that are involved in fungal development.

The veA gene in A. nidulans [191] is a gene initially found to be crucial for light-dependent conidiation. A comparison of the light effect on sterigmatocystin production by A. nidulans veA+ and veA1 strains showed that both strains produced sterigmatocystin but the highest amount was produced by the veA+ strain grown in darkness. However, veA-deleted A. flavus and A. parasiticus strains completely lost the ability to produce aflatoxin regardless of the illumination conditions [192,193]. Under normal growth conditions, some A. flavus and all A. parasiticus strains produce conidia in both dark and light conditions. Stinnett et al. [193] showed that VeA contains a bipartite nuclear localization signal (NLS) motif and its migration to the nucleus is light-dependent and requires the importin α carrier protein. In the dark VeA is located mainly in the nucleus; under light it is located both in cytoplasm and nucleus. VeA has no recognizable DNA-binding seuqences and likely exerts its effect on sterigmatosyctin and aflatoxin production through protein-protein interactions with other regulatory factors. Post-translational modifications such as phosphylation and dephosphorylation may modulate its activity. Lack of VeA production in the veA-deleted A. flavus and A. parasiticus strains consequently abolishes aflatoxin production because a threshold concentration of nuclear VeA might be necessary to initiate aflatoxin biosynthesis.

Carbon, nitrogen, amino acid, lipid, trace elements and other nutritional factors have long been observed to affect aflatoxin production [198,200,201]. The best-known nutritional factors affecting aflatoxin biosynthesis are carbon and nitrogen sources [202,203,204]. The relationship of carbon source and aflatoxin formation has been well established. Simple sugars such as glucose, sucrose, fructose, and maltose support aflatoxin formation, while peptone, sorbose, or lactose does not [198,205]. Woloshuk et al. reported the connection between alpha amylase activity and aflatoxin production in A. flavus [206]. Yu et al. identified a gene cluster related to sugar utilization in A. parasiticus next to the aflatoxin gene cluster [158]. A close physical linkage between the two gene clusters could point to a relationship between the clusters in the processing of carbohydrates leading to the induction of aflatoxin biosynthesis. Lipid substrate is a good carbon source to support aflatoxin production [207,208,209]. A lipase gene, lipA, was cloned in A. parasiticus and A. flavus [210]. The expression of lipA and subsequent aflatoxin production are induced by lipid substrate. The addition of 0.5% soybean oil to non-aflatoxin-conducive peptone medium induces lipase gene expression and leads to aflatoxin formation [210]. However, the molecular mechanism by which a carbon source is involved in the regulation of aflatoxin pathway gene expression is to be further investigated.

Nitrogen is closely linked to aflatoxin production [198]. Asparagine, aspartate, alanine, ammonium nitrate, ammonium nitrite, ammonium sulfate, glutamate, glutamine, and proline containing media support aflatoxin production; while sodium nitrate and sodium nitrite containing media do not [211,212,213]. It was also suggested that nitrate represses averufin and aflatoxin formation [214,215]. Nitrate was reported to have a suppressive effect on aflatoxin production, and overexpression of aflR gene by additional copies of aflR overcomes the negative regulatory effect on aflatoxin pathway gene transcription [173]. Nitrogen utilization genes and a nitrogen regulator gene, areA from A. parasiticus were cloned [216,217]. In the intergenic region between aflR and aflS (aflJ), several AreA binding motifs have been identified [216,217]. The AreA binding could prevent AflR binding. Certain amino acids can have different effects on aflatoxin production. Recent studies show that tryptophan inhibits aflatoxin formation while tyrosine enhances aflatoxin production in A. flavus [183].

Aflatoxin formation is directly affected by temperature [218,219,220,221]. Optimal aflatoxin production is observed at temperatures near 30 °C (28 °C to 35 °C) [221]. When temperature increases to above 36 °C, aflatoxin production is nearly completely inhibited. Genome wide gene profiling using microarray and RT-PCR verification [221] indicated that high temperature is associated with a decrease in the expression of the aflatoxin pathway genes. RT-PCR detected ample amount of transcripts of both regulatory genes aflR and aflS (aflJ) [221]. So it was hypothesized that temperature may affect the activity of AflR or some other unknown regulatory element. High resolution studies using next generation sequencing technologies on aflatoxin pathway gene expression in response to temperature clearly showed that temperature affects the level of aflR and aflS transcripts [222]. High temperature affects aflS more than aflR. Change in the ratio of aflS to aflR renders aflR unfunctional for transcription activation.

Severe aflatoxin outbreaks in corn have been documented to occur under hot weather and drought conditions [223,224]. The mechanism of A. flavus infestation in corn under these conditions is not well understood. The possible scenarios may include a combination of these factors: (a) the plant defense mechanism is weakened under water stress conditions; (b) higher insect feeding and associated injuries to plant tissues, thus providing entry opportunities for fungal invasion; and (c) more fungal spores dispersed into the air under dry climate.

Aflatoxin biosynthesis in A. flavus occurs in acidic media, but is inhibited in alkaline media [7]. The pacC gene is a major transcriptional regulatory factor for pH homeostasis [225]. In the aflR promoter region, at least one PacC binding site has been identified [77,181]. The presence of a putative PacC-binding site close to the aflR transcription start site may play some role in pH regulation on aflatoxin production [225,226]. In the non-aflatoxin-conducive peptone medium, this site was shown to be inhibitory to aflatoxin formation [7]. The regulatory mechanism might be due to the binding of pacC to this site at alkaline conditions to repress the transcription of acid-expressed gene aflR and thus aflatoxin formation [227,228]. The PacC and AreA [217] binding sites in the aflR-aflS (aflJ) intergenic region suggest that gene expression is regulated by environmental signals (pH and nitrate).

Sporulation and sclerotial formation are associated with secondary metabolism [6,8,229,230]. Spore formation and secondary metabolite formation occur at about the same time [85,229]. Some mutants that are deficient in sporulation are unable to produce aflatoxins [68] and some compounds that inhibit sporulation in A. parasiticus also inhibit aflatoxin formation [231]. The aflatoxin-producing ability was gradually decreased in response to a series subculturing. The changes in aflatoxin production ability were accompanied by marked morphological changes [232].

Moreover, chemicals that inhibit polyamine biosynthesis in A. parasiticus and A. nidulans inhibit both sporulation and aflatoxin/ST biosynthesis [233]. A more recent finding reveals that the regulation of sporulation and ST production is through a shared G-protein mediated signaling pathway in A. nidulans [229,234]. Mutations in A. nidulans flbA and fadA genes, early acting members of a G-protein signal transduction pathway, result in loss of ST gene expression, ST production, and sporulation [179,229,235]. The regulation is partially mediated through protein kinase A [179,235]. This G-protein signaling pathway involving FadA in the regulation of aflatoxin production also exists in A. parasiticus and A. flavus [229].

Oxidative stress and aflatoxin biosynthesis are related in A. parasiticus [236,237,238,239]. Oxidative stress induces aflatoxin formation in A. parasiticus [240]. Treatment of A. flavus with tert-butyl hydroperoxide, or gallic acid induced significant increases in aflatoxin production [238]. Similar treatment of A. parasiticus also induced aflatoxin production [239,241]. Hydrolysable tannins significantly inhibit aflatoxin biosynthesis, with the main anti-aflatoxigenic constituents in these tannins being gallic acid [237]. Gallic acid reduces expression of structural genes within the aflatoxin biosynthetic cluster, but surprisingly not the aflatoxin pathway gene regulator, aflR. It appears that gallic acid disrupts signal transduction pathway(s) for aflatoxigenesis. When certain phenolics or other antioxidants, such as ascorbic acid, are added to oxidatively stressed A. flavus, aflatoxin production significantly declines, with no effect on fungal growth [238]. Caffeic acid is another antioxidant that inhibits aflatoxigenesis. Microarray analysis of A. flavus treated with caffeic acid identified a gene, named ahpC2,an alkyl hydroperoxide reductase that is potentially involved in quelling the signal for aflatoxin production. However, no notable effect on expression of laeA, a gene encoding a global regulator for secondary metabolism in Aspergillus [169] was observed when under caffeic acid treatment.

Plant metabolites play some role on aflatoxin formation [242,243,244,245]. Wright et al. reported that, at certain conditions, n-decyl aldehyde reduces not only fungal growth of A. parasiticus but also aflatoxin production by over 95% compared with control [246]. Octanal reduces fungal growth by 60%, however, it increases aflatoxin production by 500%, while hexanal reduces fungal growth by 50%, but it has no effect on aflatoxin production. The 13(S)-hydroperoxide derivative of linoleic acid, the reaction product of lipoxygenase (encoded by L2 LOX gene from maize), is reported to reduce aflatoxin production [247]. Therefore, besides routine detection and screening, management of aflatoxin contamination in commodities must include pre-harvest as well as post-harvest control measures.

Surveillance programs have been established to reduce the risk of aflatoxin consumption in humans and animals. Analytical testing methods for rapid detection of a large number of samples of food stuffs have been developed [83]. Current analytical techniques for accurate characterization and quantitation of aflatoxins include thin layer chromatography (TLC), high pressure liquid chromatography (HPLC) and gas chromatography (GC). Rapid immuno-assay (RIA) and serum assay (ELISA) formats based on sera developed for major aflatoxins [83,248,249] are also available for rapid detection of aflatoxins at levels as low as one tenth of a nanogram per milliliter [250]. A dozen commercial test kits have been developed for field-testing of various agricultural commodities.

Aflatoxin contamination can be reduced somewhat by appropriate cultural practices such as irrigation, control of insect pests etc prior to harvest. In some cases, changes in cultural practices to minimize aflatoxin contamination are not feasible. For example, inclusion of extra irrigation regimes in desert cotton fields is not feasible because it would add significant cost to the grower even if they were available. However, effective control of aflatoxin contamination is expected to be dependent on a detailed understanding of the physiological and environmental factors that affect aflatoxin biosynthesis, the biology and ecology of the fungus, and the parameters of the host plant-fungal interactions. Efforts are underway to study these parameters, primarily by functional genomics approaches [251,252].

Aflatoxin contamination during storage after harvest is prevalent in most tropical countries due to hot, wet climates coupled with sub-adequate methods of harvesting, handling and storage practices, which often lead to severe fungal growth and thus contamination of food and feed [253,254]. Significant emphasis has been placed on improving storage conditions that are less favorable for fungal growth. Detoxification of aflatoxin contaminated grains to reduce aflatoxin to an acceptable level is another strategy in corn and peanuts [255].

Application of nonaflatoxigenic, biocompetitive, native A. flavus strains to outcompete toxigenic isolates in the fields has been effective in significantly reducing aflatoxins contamination [256]. Biocompetition may be most feasible in crops such as cotton where resistance against fungal infections is not available due to limited genetic diversity in the cotton germplasm. In the long term, however, significant control of the aflatoxin problem will likely be linked to introduction of resistant germplasms, which are resistant either to fungal invasion or toxin production or both. Naturally resistant germplasms are available and the identification of specific biochemical factors linked to resistance against A. flavus will assist in enhancing the observed resistance levels in the existing germplasms. Identification of a novel pool of germplasm that demonstrates the desired characteristics is also critical to the success of the current marker-assistance breeding programs. However, because the aflatoxin contamination is a complex problem and a combination of approaches will be required to control the preharvest contamination.

The most economic and effective strategy for reducing and eliminating aflatocin contamination is to breed commercial crop varieties that show resistance to fungal infection and/or aflatoxin formation. Significant efforts are made in plant breeding programs in identifying resistance or tolerance to aflatoxin contamination in preharvest crops. Unfortunately, no highly resistant varieties or germplasm lines have been identified for the major crops such as corn, cotton, and peanut. Some low to medium resistant lines in corn are under testing and development [257,258]. Progress has been made in identifying genes in corn that shows resistance to aflatoxin producing fungus [259].