Cloning of the afl-2 gene involved in aflatoxin biosynthesis from Aspergillus flavus.

Aflatoxins are extremely potent carcinogens produced by Aspergillus flavus and Aspergillus parasiticus. Cloning of genes in the aflatoxin pathway provides a specific approach to understanding the regulation of aflatoxin biosynthesis and, subsequently, to the control of aflatoxin contamination of food and feed. This paper reports the isolation of a gene involved in aflatoxin biosynthesis by complementation of an aflatoxin-nonproducing mutant with a wild-type genomic cosmid library of A. flavus. Strain 650-33, blocked in aflatoxin biosynthesis at the afl-2 allele, was complemented by a 32-kb cosmid clone (B9), resulting in the production of aflatoxin. The onset and profile of aflatoxin accumulation was similar for the transformed strain and the wild-type strain (NRRL 3357) of the fungus, indicating that the integrated gene is under the same control as in wild-type strains. Complementation analyses with DNA fragments from B9 indicated that the gene resides within a 2.2-kb fragment. Because this gene complements the mutated afl-2 allele, it was designated afl-2. Genetic evidence obtained from a double mutant showed that afl-2 is involved in aflatoxin biosynthesis before the formation of norsolorinic acid, the first stable intermediate identified in the pathway. Further, metabolite feeding studies with the mutant, transformed, and wild-type cultures and enzymatic activity measurements in cell extracts of these cultures suggest that afl-2 regulates gene expression or the activity of other aflatoxin pathway enzymes. This is the first reported isolation of a gene for aflatoxin biosynthesis in A. flavus.     

Development of polyclonal antibodies for detection of aflatoxigenic molds involving culture filtrate and chimeric proteins expressed in Escherichia coli.

Polyclonal antibodies (PAb) were raised against an aflatoxigenic strain of Aspergillus parasiticus by using two different sources for antibody elicitation: (i) filtrate of a culture on which the fungus had been grown (ii) and two chimeric proteins, expressed in Escherichia coli as separate products, of the genes ver-1 and apa-2, which are involved in aflatoxin biosynthesis. The gene products were amplified by PCR, and each was cloned into the E. coli expression vector pGEX2T. Upon induction, the bacteria overexpressed 38- and 33-kDa chimeric proteins corresponding to the N-terminal domains of the genes ver-1 and apa-2, respectively. The chimeric proteins were isolated and affinity purified for use as antigens. The specificity of the raised antibodies was examined by enzyme-linked immunosorbent assay (ELISA). The PAbs raised against the culture filtrate reacted with all the species of Aspergillus and Penicillium tested but not with Fusarium species or corn gain. However, the PAbs elicited against the chimeric proteins were highly specific, showing significantly higher ELISA absorbance values (A405) against A. parasiticus and A. flavus than against the other fungi tested and the corn grain. The approach of utilizing gene products associated with aflatoxin biosynthesis for antibody production therefore appears to be feasible. Such a multiantibody system combined with the PCR technique, could provide a useful tool for the rapid, sensitive, and accurate detection of aflatoxin producers present in grains and foods.     

Characterization of the function of the ver-1A and ver-1B genes, involved in aflatoxin biosynthesis in Aspergillus parasiticus.

The ver-1A gene was cloned and its nucleotide sequence was determined as part of a previous study on aflatoxin B1 (AFB1) biosynthesis in the filamentous fungus Aspergillus parasiticus SU-1. A second copy of this gene, ver-1B, was tentatively identified in this fungal strain. In this study, ver-1B was cloned by screening an A. parasiticus cosmid library with a ver-1A probe. The nucleotide sequence of ver-1B was determined. The predicted amino acid sequence of ver-1B had 95% identity with ver-1A. A translational stop codon, found in the ver-1B gene coding region, indicated that it encodes a truncated polypeptide. To confirm the function of the ver-1 genes in AFB1 synthesis, a plasmid (pDV-VA) was designed to disrupt ver-1A and/or ver-1B by transformation of the AFB1 producer A. parasiticus NR-1. One disruptant, VAD-102, which accumulated the pathway intermediate versicolorin A was obtained. Southern hybridization analysis of VAD-102 revealed that ver-1A but not ver-1B was disrupted. A functional ver-1A gene was transformed back into strain VAD-102. Transformants which received ver-1A produced AFB1, confirming that ver-1A is the only functional ver-1 gene in A. parasiticus SU-1 and that its gene product is involved in the conversion of versicolorin A to sterigmatocystin in AFB1 biosynthesis. A duplicated chromosomal region (approximately 12 kb) was identified upstream from ver-1A and ver-1B by Southern hybridization analysis. This duplicated region contained the aflR gene, which is proposed to be one regulator of AFB1, synthesis. A similar gene duplication was also identified in several other strains of A. parasiticus.     

Detection of aflatoxigenic molds in grains by PCR.

Aflatoxins are carcinogenic metabolites produced by several members of the Aspergillus flavus group in grains and floods. Three genes, ver-1, omt-1, and apa-2, coding for key enzymes and a regulatory factor in aflatoxin biosynthesis, respectively, have been identified, and their DNA sequences have been published. In the present study, three primer pairs, each complementing the coding portion of one of the genes, were generated. DNA extracted from mycelia of five Aspergillus species, four Penicillium species, and two Fusarium species was used as PCR template for each of the primer pairs. DNA extracted from peanut, corn, and three insect species commonly found in stored grains was also tested. Positive results (DNA amplification) were achieved only with DNA of the aflatoxigenic molds Aspergillus parasiticus and A. flavus in all three primer pairs. The detection limit of the PCR was determined by using the primer pairs complementing the omt-1 and ver-1 genes. Sterile corn flour was inoculated separately with six different molds, each at several spore concentrations. Positive results were obtained only after a 24-h incubation in enriched media, with extracts of corn inoculated with A. parasiticus or A. flavus, even at the lowest spore concentration applied (10(2) spores per g). No DNA spores per g). It is concluded that genes involved in the aflatoxin biosynthetic pathway may form the basis for an accurate, sensitive, and specific detection system, using PCR, for aflatoxigenic strains in grains and foods.     

Molecular characterization of the afl-1 locus in Aspergillus flavus.

An unusual mutation at the afl-1 locus, affecting aflatoxin biosynthesis in Aspergillus flavus 649, was investigated. The inability of strain 649 to produce aflatoxin was found to be the result of a large (greater than 60 kb) deletion that included a cluster of aflatoxin biosynthesis genes. Diploids formed by parasexual crosses between strain 649 and the aflatoxigenic strain 86 did not produce aflatoxin, indicating the dominant nature of the afl-1 mutation in strain 649. In metabolite feeding experiments, the diploids did not convert three intermediates in the aflatoxin pathway to aflatoxin. Northern (RNA blot) analysis of the diploids grown in medium conducive for aflatoxin production indicated that the aflatoxin pathway genes nor1, ver1, and omt1 were not expressed; however, there was low-level expression of the regulatory gene aflR. Pulsed-field electrophoresis gels indicated a larger (6 Mb) chromosome in strain 649 than the apparently homologous (4.9 Mb) chromosome in strain 86. The larger chromosome in strain 649 suggests that a rearrangement occurred in addition to the deletion. From these data, we proposed that a trans-sensing mechanism in diploids is responsible for the dominant phenotype associated with the afl-1 locus in strain 649. Such a mechanism is known in Drosophila melanogaster but has not been described for fungi.     

Identification of aflatoxin biosynthesis genes by genetic complementation in an Aspergillus flavus mutant lacking the aflatoxin gene cluster.

Aspergillus flavus mutant strain 649, which has a genomic DNA deletion of at least 120 kb covering the aflatoxin biosynthesis cluster, was transformed with a series of overlapping cosmids that contained DNA harboring the cluster of genes. The mutant phenotype of strain 649 was rescued by transformation with a combination of cosmid clones 5E6, 8B9, and 13B9, indicating that the cluster of genes involved in aflatoxin biosynthesis resides in the 90 kb of A. flavus genomic DNA carried by these clones. Transformants 5E6 and 20B11 and transformants 5E6 and 8B9 accumulated intermediate metabolites of the aflatoxin pathway, which were identified as averufanin and/or averufin, respectively.These data suggest that avf1, which is involved in the conversion of averufin to versiconal hemiacetal acetate, was present in the cosmid 13B9. Deletion analysis of 13B9 located the gene on a 7-kb DNA fragment of the cosmid. Transformants containing cosmid 8B9 converted exogenously supplied O-methylsterigmatocystin to aflatoxin, indicating that the oxidoreductase gene (ord1), which mediates the conversion of O-methylsterigmatocystin to aflatoxin, is carried by this cosmid. The analysis of transformants containing deletions of 8B9 led to the localization of ord1 on a 3.3-kb A. flavus genomic DNA fragment of the cosmid.     

Aspergillus nidulans verA is required for production of the mycotoxin sterigmatocystin.

Aspergillus nidulans produces the carcinogenic mycotoxin sterigmatocystin (ST), the next-to-last precursor in the aflatoxin (AF) biosynthetic pathway found in the closely related fungi Aspergillus flavus and Aspergillus parasiticus. We identified and characterized an A. nidulans gene, verA, that is required for converting the AF precursor versicolorin A to ST. verA is closely related to several polyketide biosynthetic genes involved in polyketide production in Streptomyces spp. and exhibits extended sequence similarity to A. parasiticus ver-1, a gene proposed to encode an enzyme involved in converting versicolorin A to ST. By performing a sequence analysis of the region 3' to verA, we identified two additional open reading frames, designated ORF1 and ORF2. ORF2 is closely related to a number of cytochrome P-450 monooxygenases, while ORF1 shares identity with the gamma subunit of translation elongation factor 1. Given that several steps in the ST-AF pathway may require monooxygenase activity and that AF biosynthetic genes are clustered in A. flavus and A. parasiticus, we suggest that verA may be part of a cluster of genes required for ST biosynthesis. We disrupted the verA coding region by inserting the A. nidulans argB gene into the center of the coding region and transformed an A. nidulans argB2 mutant to arginine prototrophy. Seven transformants that produced DNA patterns indicative of a verA disruption event were grown under ST-inducing conditions, and all of the transformants produced versicolorin A but negligible amounts of ST (200-fold to almost 1,000-fold less than the wild type), confirming the hypothesis that verA encodes an enzyme necessary for converting versicolorin A to ST.     

Intrafungal distribution of aflatoxins among conidia and sclerotia of Aspergillus flavus and Aspergillus parasiticus

This research examines the distribution of aflatoxins among conidia and sclerotia of toxigenic strains of Aspergillus flavus Link and Aspergillus parasiticus Speare cultured on Czapek agar (21 days, 28 degrees C). Total aflatoxin levels in conidia and sclerotia varied considerably both within (intrafungal) and among strains. Aspergillus flavus NRRL 6554 accumulated the highest levels of aflatoxin (conidia: B1, 84000 ppb; G1, 566000 ppb; sclerotia: B1, 135000 ppb; G1, 968000 ppb). Substantial aflatoxin levels in conidia could place at risk those agricultural workers exposed to dust containing large numbers of A. flavus conidia. Cellular ratios of aflatoxin B1 to aflatoxin G1 were nearly identical in conidia and sclerotia even though levels of total aflatoxins in these propagule types may have differed greatly. Aflatoxin G1 was detected in sclerotia of all A. flavus strains but in the conidia of only one strain. Each of the A. parasiticus strains examined accumulated aflatoxin G1 in both sclerotia and conidia. These results are examined in the context of current evolutionary theory predicting an increase in the chemical defense systems of fungal sclerotia, propagules critical to the survival of these organisms.     

Cloning of a sugar utilization gene cluster in Aspergillus parasiticus

At one end of the 70 kb aflatoxin biosynthetic pathway gene cluster in Aspergillus parasiticus and Aspergillus flavus reported earlier, we have cloned a group of four genes that constitute a well-defined gene cluster related to sugar utilization in A. parasiticus: (1) sugR, (2) hxtA, (3) glcA and (4) nadA. No similar well-defined sugar gene cluster has been reported so far in any other related Aspergillus species such as A. flavus, A. nidulans, A. sojae, A. niger, A. oryzae and A. fumigatus. The expression of the hxtA gene, encoding a hexose transporter protein, was found to be concurrent with the aflatoxin pathway cluster genes, in aflatoxin-conducive medium. This is significant since a close linkage between the two gene clusters could potentially explain the induction of aflatoxin biosynthesis by simple sugars such as glucose or sucrose.     

Mold flora and aflatoxin contamination of stored and cooked samples of pearl millet in the Paharia tribal belt of Santhal paragana, Bihar, India

Stored and cooked samples of pearl millet (Pennesetum typhoides), which is regularly consumed as food by the Paharia tribe in the hilly regions of Santhal Pargana, Bihar State, India, that were harvested in January 1989 were analyzed for mold flora, natural occurrence of Aspergillus flavus and A. parasiticus, and incidence and levels of aflatoxin B1. Of the 22 fungal species isolated, A. flavus and A. parasiticus were the predominant species (63.8%) during the rainy season, followed by other species of Aspergillus, Penicillium, Fusarium, Rhizopus, Helminthosporium, and Curvularia. Screening of 169 A. flavus and A. parasiticus strains showed that 59 of them were toxigenic, producing various combinations of aflatoxins B1, B2, G1, and G2. The amounts of aflatoxin B1 ranged between 4 and 30 mg/100 ml of liquid medium. Analysis of stored and cooked samples also revealed a high incidence and alarming levels of naturally produced aflatoxin B1. Forty-nine of 75 stored and 16 of 38 cooked samples contained various combinations of aflatoxins. The levels of aflatoxin B1 ranged between 17 and 2,110 ppb in stored samples and 18 and 549 ppb in cooked samples. The correlation of insect damage with A. flavus and A. parasiticus incidence and quantity of aflatoxin B1 was found to be insignificant.     

Precursor recognition by kinetic pulse-labeling in a toxigenic aflatoxin B1-producing strain of Aspergillus.

Kinetic pulse-labeling of aflatoxin pathway compounds was carried out in Aspergillus parasiticus, beginning with radioactive acetate. Norsolorinic acid, averufin, versicolorin A, and sterigmatocystin (all known as compounds which can be incorporated into the aflatoxin molecule) were radiotraced to follow their order of appearance. Aflatoxin species B1, B2, G1, and G2 were included. Norsolorinic acid and averufin appeared as early transient intermediates followed in order by versicolorin A, aflatoxins, and sterigmatocystin. To date, a mutually confirming array of results has been obtained with established precursors in wild-type strains of A. parasiticus and A. versicolor (as well as with an aflatoxin pathway mutant of A. parasiticus), which together establish a practical methodology for recognition of new pathway intermediates. The kinetic of pulse-labeling for sterigmatocystin in relation to aflatoxins suggests that duel branchlets may exist to flatoxins; i.e., sterigmatocystin may not be an obligatory aflatoxin precursor.     

Biosynthetic origin of aflatoxin G1: confirmation of sterigmatocystin and lack of confirmation of aflatoxin B1 as precursors

The origin of aflatoxin G1 was studied using mutant strains of Aspergillus parasiticus blocked early in the pathway and by tracing 14C-labelled aflatoxin B1 (AFB1) in wild-type A. flavus and A. parasiticus strains. Sterigmatocystin (ST) was a precursor of AFB1, AFG1 and AFG2 in the four mutants examined. The identity of AFG1 was confirmed by mass spectrometry. No evidence for conversion of AFB1 to AFG1 was found. A rigorously controlled study of conversions of radioactivity based on preparative thin-layer chromatography of aflatoxins demonstrated that low levels of aflatoxin interconversions previously reported in the literature might actually be artifacts.     

Mold flora and aflatoxin contamination of stored and cooked samples of pearl millet in the Paharia tribal belt of Santhal paragana, Bihar, India.

Stored and cooked samples of pearl millet (Pennesetum typhoides), which is regularly consumed as food by the Paharia tribe in the hilly regions of Santhal Pargana, Bihar State, India, that were harvested in January 1989 were analyzed for mold flora, natural occurrence of Aspergillus flavus and A. parasiticus, and incidence and levels of aflatoxin B1. Of the 22 fungal species isolated, A. flavus and A. parasiticus were the predominant species (63.8%) during the rainy season, followed by other species of Aspergillus, Penicillium, Fusarium, Rhizopus, Helminthosporium, and Curvularia. Screening of 169 A. flavus and A. parasiticus strains showed that 59 of them were toxigenic, producing various combinations of aflatoxins B1, B2, G1, and G2. The amounts of aflatoxin B1 ranged between 4 and 30 mg/100 ml of liquid medium. Analysis of stored and cooked samples also revealed a high incidence and alarming levels of naturally produced aflatoxin B1. Forty-nine of 75 stored and 16 of 38 cooked samples contained various combinations of aflatoxins. The levels of aflatoxin B1 ranged between 17 and 2,110 ppb in stored samples and 18 and 549 ppb in cooked samples. The correlation of insect damage with A. flavus and A. parasiticus incidence and quantity of aflatoxin B1 was found to be insignificant.