Evaluation of a multiplex PCR assay for concurrent detection of four major mycotoxigenic fungi from foods

Authors


Correspondence

H.V. Batra, Sc H, Defence Food Research Laboratory, Siddarthanagar, Mysore, Karnataka, India – 570011. E-mail: h_v_batra@rediffmail.com

Abstract

Aim

To develop and evaluate a multiplex polymerase chain reaction assay (mPCR) for the concurrent detection of four major mycotoxin metabolic pathway genes, viz. nor1 (aflatoxin), Tri6 (trichothecene), FUM13 (fumonisin) and otanps (ochratoxin A).

Methods and Results

A mPCR assay with competitive internal amplification control, employing specific primers for each of the aforementioned four genes, was optimized and validated using 10 reference strains and 60 pure culture isolates. The standardized mPCR assay detected all four mycotoxin metabolic genes in artificially contaminated maize samples with a sensitivity of 2 × 103 CFU g−1 for nor1-positive Aspergillus strains, Tri6 and FUM13-positive Fusarium strains and 2 × 104 CFU g−1 for otanps-positive Penicillium strains. When the developed mPCR assay was applied to 40 natural foods, 35% (14 of 40) of the samples were contaminated with either one or more mycotoxins. The mPCR results were further evaluated with high-performance liquid chromatography (HPLC), and in general, both the methods provided unequivocal results.

Conclusion

The current mPCR assay is a rapid and reliable tool for simultaneous specific and sensitive detection of aflatoxigenic Aspergillus strains, trichothecene- and fumonisin-producing Fusarium strains, and ochratoxigenic Penicillium species from naturally contaminated foods.

Significance and Impact of the Study

This mPCR assay could be a supplementary strategy to current conventional mycotoxin analytical techniques such as thin-layer chromatography (TLC), high performance thin layer chromatography, HPLC, etc., and a reliable tool for high-throughput monitoring of major mycotoxin-producing fungi during the processing steps of food and feed commodities.

Introduction

Mycotoxins are a diverse group of secondary fungal metabolites, which are ubiquitous in nature, occurring regularly in food and fodder (Kuhn and Ghannoum 2003). Agricultural crops are generally colonized by saprophytic fungi either during crop production or after harvest while in storage. Poor pre- and postharvest practices, improper drying and handling, packaging, storage and transport conditions contribute to fungal growth and mycotoxin presence in food commodities. The Food and Agriculture Organization (FAO) estimated that worldwide approximately 25% of crops are contaminated with mycotoxins every year (Lawlor and Lynch 2005). At a considerably high level of contamination in foods, these mycotoxins can have toxic effects ranging from acute (liver or kidney deterioration) to chronic (mutagenic, teratogenic, carcinogenic) manifestations in humans and animals (ICMSF, 1996).

The most commonly encountered mycotoxins in foods are aflatoxins, fumonisins, trichothecenes and ochratoxins. Aflatoxins are the largest contaminants of food and feeds, produced by Aspergillus flavus and A. parasiticus. They are teratogenic and carcinogenic, mainly causing liver cancer (Creppy 2002). Tricothecenes and fumonisins, produced by the species of Fusarium genus (Fusarium graminearum, F. culmorum and F. sporotrichioides – trichothecene producers; F. verticillioides and F. proliferatum – fumonisin producers), infect food grains such as maize, wheat, barley, rice, finger millet, oats and rye. Diacetoxyscirpenol, deoxynivalenol and T-2 toxins are the best studied of the trichothecenes and are associated with skin inflammation, digestive disorders, tachycardia, oedema, haemorrhages, haemolytic disorders, impairment of immune responses and nervous disorders (Cole and Cox 1981). Fumonisins, on the other hand, are another group of mycotoxins chiefly associated with pink ear rot of maize. They affect animals by interfering with sphingolipid metabolism (Merrill et al. 2001). Ochratoxin A (OTA) is predominantly produced by Penicillium verrucosum and A. ochraceus. It is a potent teratogen, immunosuppressant, carcinogen and causes nephrotoxicity and hepatotoxicity and endemic Balkan nephropathy (Pfohl-Leszkowicz et al. 2002).

The conditions or mycotoxicosis caused by mycotoxins are not pathognomonic; therefore, determining the cause of the specific condition or disease requires confirmation of the toxin(s) in a representative sample of the feed, food, tissue or fluid (Richard et al. 1993). Mycotoxicosis would be a greater threat considering the co-occurrence of many mycotoxigenic fungi in food and the environment, and as a result, mycotoxins in the food supply are constantly evaluated and closely monitored. These mycotoxins are strictly regulated, and most countries have set thresholds for accepted levels of mycotoxins in food and feeds. Traditionally, analytical methods such as thin-layer chromatography (TLC), high-performance liquid chromatography (HPLC) and gas chromatography (GC), etc. are highly recommended techniques for detecting mycotoxins because of their accuracy, precision and specificity (ICMSF, 1996). Various in vitro short-term biological assays have also been employed to screen the presence of several mycotoxins, viz. Girardi heart cell assay, Brine shrimp larvae assay and Zebra fish larvae assay (Watson and Lindsay 1982). Many researchers have developed rapid and sensitive molecular (PCR based) and immunological assays, which have replaced conventional methods for the efficient detection of mycotoxin-producing fungi (Niessen et al. 2005 and Niessen 2007).

The objective of the present work was to standardize a multiplex polymerase chain reaction (mPCR) assay for concurrently detecting major mycotoxin-producing Aspergillus, Fusarium and Penicillium species by targeting the biosynthetic pathway genes involved in toxin production. A competitive internal amplification control (IAC) was also incorporated to account for the false-negative results during the PCR. The standardized mPCR assay was evaluated with 40 natural food samples and the assay results were compared with HPLC.

Materials and methods

Fungal strains and media

All the fungal strains used in this study are listed in Table 1. Standard strains were obtained from various culture collection centres across India: National Collection of Industrial Microorganisms (NCIM), Pune, Microbial Type Culture Collection (MTCC), Chandigarh, and Indian Type Culture Collection (ITCC), New Delhi. DFR strains were isolated and maintained at the Defence Food Research Laboratory Mysore, India. All media were procured from Himedia, Mumbai, India. Standard strains and isolates were inoculated in potato dextrose agar/broth (PDA/PDB) and incubated for 5 days at 25 ± 1°C. The fungal colonies were periodically subcultured on to PDA plates, and 7-day-old cultures were used for DNA extraction.

Table 1. Standard fungal strains used this study
OrganismNameSource
  1. a

    Microbial Type Culture Collection (MTCC).

  2. b

    Indian Type Culture Collection (ITCC).

  3. c

    National Collection of Industrial Microorganisms.

  4. d

    Defence Food Research Laboratory.

Aspergillus niger MTCC 282MTCCa, India
Fusarium graminearum MTCC 2089MTCC, India
F. proliferatum MTCC 286MTCC, India
Penicillium verrucosum MTCC 4935MTCC, India
Ralstonia stolonifer MTCC 162MTCC, India
Trichoderma viride MTCC 793MTCC, India
A. parasiticus ITCC 456ITCCb, India
F. culmorum ITCC 149ITCC, India
F. solani ITCC 3359ITCC, India
P. citrinum NCIM 765NCIMc, India
Aflatoxigenic isolates
 Aspergillus sp.DFR A4, A5, A8, A16, A20, A26Maize isolate, D.F.R.Ld, Mysore
 Aspergillus sp.DFR A32, A35, A36, A37Paddy isolate, D.F.R.L, Mysore
 Aspergillus sp.DFR A42, A43, A44, A47Finger millet isolate, D.F.R.L, Mysore
 Aspergillus sp.DFR A50, A54, A60Peanut, D.F.R.L, Mysore
Trichothecene isolates
 Fusarium sp.DFR F2, F6, F8, F15, M17, M19Maize isolate, D.F.R.L, Mysore
 Fusarium sp.DFR F30, F34, F35, F39Paddy isolate, D.F.R.L, Mysore
 Fusarium sp.DFR F45, F48, F50Sorghum isolate, D.F.R.L, Mysore
 Fusarium sp.DFR F56, F58, F64Finger millet isolate, D.F.R.L, Mysore
Fumonisin isolates
 Fusarium sp.DFR F7, F14, F16Maize isolate, D.F.R.L, Mysore
 Fusarium sp.DFR F25, F32, F40Paddy isolate, D.F.R.L, Mysore
 Fusarium sp.DFR F51, F57Sorghum isolate, D.F.R.L, Mysore
 Fusarium sp.DFR F68, F70, F75Finger millet isolate, D.F.R.L, Mysore
Ochratoxigenic isolates
 Penicillium sp.DFRP5Maize isolate, D.F.R.L, Mysore
 Penicillium sp.DFRP15Paddy isolate, D.F.R.L, Mysore
Nontoxigenic isolates
 Aspergillus sp.DFRA13Maize isolate, D.F.R.L, Mysore
 Aspergillus sp.DFRA540Paddy isolate, D.F.R.L, Mysore
 Aspergillus sp.DFRA56Peanut, D.F.R.L, Mysore
 Fusarium sp.DFRF20Maize isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF22Paddy isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF54Sorghum isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF60Finger millet isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF18Maize isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF42Paddy isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF49Sorghum isolate, D.F.R.L, Mysore
 Fusarium sp.DFRF72Finger millet isolate, D.F.R.L, Mysore
 Penicillium sp.DFRP8Maize isolate, D.F.R.L, Mysore
 Penicillium sp.DFRP13Maize isolate, D.F.R.L, Mysore
 Penicillium sp.DFRP20Paddy isolate, D.F.R.L, Mysore

DNA extraction

The DNA extraction protocol of Ramana et al. (2011) was followed, with minor modifications. A pinch of 7-day-old mycelia from the surface of the PDA plates was transferred to 50 ml PDB and was subjected to constant agitation for 3 days at 25 ± 1°C. After incubation, the mycelia were separated from the medium by 20 min of centrifugation at 8000 g. The mycelia pellet was ground in liquid nitrogen using a pestle and mortar. DNA was extracted using a DNeasy plant Minikit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Nuclease-free water (Qiagen) was used as eluent, and the DNA concentration was estimated by NanoDrop ND-1000 spectrophotometer (Thermo Scientific, Bengaluru, India). The DNA was stored at −80°C until further use. DNA from the spore suspensions was extracted by the thermal lysis method (Cenis 1992). All standardization experiments were performed with DNA/spores of A. paraciticus ITCC 456, F. graminearum MTCC 2089, P. verrucosum MTCC 4935 and F. proliferatum MTCC 286.

Primer designing

The primers used in the study were custom synthesized by Eurofins, Bangalore, India. Specific genes to detect toxigenic fungal species were identified after a thorough literature survey. Four metabolic pathway genes were recognized that were found to be specific to major toxigenic fungal species, namely nor1 for aflatoxigenic Aspergilli, Tri6 and FUM13 for trichothecene- and fumonisin-producing Fusarium species, respectively, otanps for ochratoxigenic Penicillium species. A new set of primers for specific detection of the nor1 gene (GenBank accession no. AY371490.1) was kindly provided by S.R. Priyanka, DFRL, India. Specific primers for amplification of the FUM13 gene (accession no. AF155773.5) were designed using Gene Runner software (http://www.generunner.net/). Primers enabling the amplification of the otanps and Tri6 genes were selected from Bogs et al. 2006 and Ramana et al. 2011, respectively. The sequences of all primers (Table 2) were evaluated using Primer-Blast and BlastN tools (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) to identify any nonspecific targets and to anticipate the specificity of the PCR assay.

Table 2. Primers used in this study
Primer namePrimer sequence (5′–3′)Gene targetedAmplicon size (bp)References
  1. a

    Tri6 gene overhangs are given in uppercase and the priming region in pUC 19 plasmid are in lower case.

nor1 FACCGCTACGCCGGCACTCTCGG nor1 396This work
nor1 RGGCCGCCAGCTTCGACACTCCG
tri6 FGATCTAAACGACTATGAATCACC Tri6 541Ramana et al. 2011
tri6 RGCCTATAGTGATCTCGCATGT
otanps FAGTCTTCGCTGGGTGCTTCC otanps 750Bogs et al. 2006
otanps RCAGCACTTTTCCCTCCATCTATCC
fum 13 FGAGCTTGTCCTTCTCACTGG FUM13 982This work
fum 13 RGAGCCGACATCATAATCAGT
IAC FGATCTAAACGACTATGAATCACCac atcgaactggatctcaacagcapUC 19 flanked with Tri6280This work
IAC RGCCTATAGTGATCTCGCATGTactgcataattctcttactgtca

Standardization of multiplex PCR

Initially, the presence of all genes in the standard strains was confirmed by performing monoplex PCRs. Standardizing the mPCR was performed by empirically varying critical factors that affect multiplexing, primer concentrations, amount of MgCl2 and annealing time and temperature. Multiplex PCR was performed in a Master Cycler-Pro thermal cycler (Eppendorf, Germany) in a 30-μl reaction mix containing 50 ng of template DNA of A. paraciticus ITCC 456, F. graminearum MTCC 2089, P. verrucosum MTCC 4935 and F. proliferatum MTCC 286, 1 × PCR buffer (with 2 mmol l−1 MgCl2), 200 μmol l−1 each dNTP, 0·5 μmol l−1 nor1 primers, 0·8 μmol l−1 Tri6 primers, 2 μmol l−1 otanps primers, 0·8 μmol l−1 FUM13 primers and 1 unit of Taq polymerase (Fermentas, New Delhi, India). The PCR cycling conditions included an initial denaturation at 94°C for 4 min, followed by 30 cycles of denaturation at 94°C for 1 min, annealing at 58°C for 1 min and extension at 72°C for 1 min with a final extension of 72°C for 8 min. The PCR products were electrophoresed on 1·2% agarose gel stained with ethidium bromide and visualized under a UV transilluminator (G-Box; Syngene, Gurgaon, India).

Design of internal amplification control

A synthetic competitive IAC was designed from pUC 19 plasmid with overhangs corresponding to Tri6 gene following composite primer technique (Siebert and Larrick 1992). By using this technique, the IAC and the Tri6 genes were amplified with one common set of primers (tri6 F and tri6 R) under the same conditions and in the same PCR tube (Hoorfar et al. 2004). A schematic representation of the IAC gene is shown in Fig. S1. The competition by IAC can, however, lower the amplification efficiency of the PCR and may thereby result in a lower detection limit for the gene in competition. Considering this critical parameter, the IAC was designed in competition with the gene that was predominantly amplified in the developed mPCR, that is, Tri6 gene (Fig. S2). The lowest reproducible IAC DNA concentration was determined by performing a mPCR assay with tenfold serial dilutions of 10 ng IAC along with 50 ng each fungal template DNA.

Robustness test

Three replicates of the multiplex PCR constituents were prepared with 50 ng of each DNA from the four standard strains that were used to standardize the assay, 0·1 pg IAC template and the optimized primer cocktail. Each replicate varied only with PCR buffer concentration (×0·8, ×1 and ×1·2). The experiments were performed at annealing temperatures of 56, 58 and 60°C. All other conditions were kept constant as described above.

Specificity of mPCR

The cross-reactivity of the primers used in the developed mPCR was analysed by testing the optimized mPCR on the different fungal species listed in Table 1. In the process, each primer set was assessed for its potential to amplify any spurious amplicons when nonspecific fungal DNA along with 0·1 pg IAC template was used in the PCR with the primer cocktail. The PCR reagent concentrations and conditions were essentially same as mentioned above.

Sensitivity of mPCR

The sensitivity of the mPCR was estimated using two strategies: DNA dilutions (DNA extracted from a culture was tenfold serially diluted) and spore dilutions (tenfold serial dilutions of spores were prepared and DNA was extracted from each dilution). 100 ng of DNA extracted from A. paraciticus ITCC 456, F. graminearum MTCC 2089, P. verrucosum MTCC 4935 and F. proliferatum MTCC 286 were tenfold serially diluted and used in the mPCR assay. Alternatively, tenfold serial dilution of spore suspensions was made from the aforementioned fungi, and 5 ml of each dilution was centrifuged at 12 000 g for 5 min. The spore pellet was resuspended in 50 μl water, and DNA was extracted by the thermal lysis method. For analysis, 5 μl of each DNA sample was used in the mPCR assay. Total spore counts were recorded after plating 100 μl of each spore dilution on to PDA plates and incubated at 28°C for 3 days.

Multiplex PCR with artificially contaminated food samples

To estimate the interference of food matrices on the sensitivity, detectable concentrations of fungal spore suspensions (108 spores) were tenfold serially diluted in 10 ml of saline. Each dilution was artificially inoculated (spiked) on to 5 g of grounded and autoclaved maize grains, and incubated in a shaker at 200 rev min−1 at 28°C for 48 h. Later, 2 ml of debris-free supernatant was collected and centrifuged at 12 000 g for 5 min. The pellet obtained was resuspended in 100 μl of distilled water and treated by the thermal lysis method to extract DNA. For analysis, 2 μl of each sample was included in the mPCR assay.

Multiplex PCR with naturally contaminated food samples

Forty diverse food samples were collected from various retail markets of Mysore, Karnataka, India (Table 3). Each sample was transferred in a sealed sterile polyethylene bag to the mycological laboratory. These samples were surface disinfected by washing them sequentially with 70% ethanol, 2% sodium hypochlorite and sterile distilled water. The surface sterile samples were ground and inoculated into 50 ml of PDB and incubated for 2 days at 25°C. Two millilitres of supernatant was centrifuged at 12 000 g for 5 min, and the pellet obtained was resuspended in 100 μl of distilled water and further subjected to DNA isolation by the thermal lysis method. For analysis, 5 μl of each sample was included in the multiplex PCR assay. The rest of the 2-day-old culture broth was further incubated for 7 days at 25°C for mycotoxin production, as described by Amadi and Adeniyi (2009).

Table 3. Application of multiplex polymerase chain reaction assay (mPCR) to naturally contaminated foods
SampleID nor-1 Tri6 otanps FUM13 IAC
mPCRHPLCmPCRHPLCmPCRHPLCmPCRHPLC
  1. HPLC, high-performance liquid chromatography; IAC, internal amplification control.

MaizeM1+++++
MaizeM2+
MaizeM3+++
MaizeM4+
MaizeM5+++++++
MaizeM6+
MaizeM7+
MaizeM8++
MaizeM9+
MaizeM10++
MaizeM11+
MaizeM12+
PaddyP13+++
PaddyP14+++
PaddyP15++
PaddyP16+++++
PaddyP17+
PaddyP18+
PaddyP19+
PaddyP20+
PaddyP21+
PaddyP22+
GingerG23+
GingerG24+
GingerG25+++
GingerG26+
GingerG27+
Finger milletF28+++
Finger milletF29+
Finger milletF30+
Finger milletF31+
Finger milletF32+
Finger milletF33+
Finger milletF34+
SorghumS35+++
SorghumS36+
SorghumS37+++
SorghumS38+
SorghumS39+
SorghumS40+++

Extraction and clean-up of mycotoxins from culture broth

The toxin extraction protocol for aflatoxin B1 (AFB1), OTA, trichothecene (DON) and fumonisin B1 (FB1) was followed as given in the study by Scudamore et al. (1997) with minor modifications. The culture broth was centrifuged at 12 000 g for 10 min, and the supernatant was divided into four aliquots of 10 ml each. Every aliquot was added with 20 ml of acetonitrile/water (60 : 40) (for AFB1 and OTA) or methanol/water (4 : 1) (for DON and FB1) and was mixed for 20 min. The extract was filtered through 0·45 μm Whatman syringe filters. Specific immuno-affinity columns (VICAM, Watertown, MA, USA) were used for the clean-up of mycotoxins following the manufacturer's instructions. Elutions were performed in 3 ml of methanol. Standard aflatoxin B1, OTA, deoxynivalenol and FB1 (Sigma-Aldrich, Bengaluru, India) were prepared according to the supplier's instructions.

High-performance liquid chromatography

Ten microlitres of toxin extracts was injected into the RP-C18 column (Jasco, Great Dunmow, Essex, UK) with dimensions of 3 μm and 250 × 46 mm for HPLC analysis. For the analytes AFB1, OTA and FB1, a methanol/water solution in the ratio of 7 : 3 v/v (isocratic elution) and for DON, a methanol/water solution in the ratio of 8 : 2 v/v were used as mobile phases. A Jasco HPLC system (Jasco) with fluorescence detector and wavelength settings of excitation 365 nm and emission 455 nm with a flow rate of 0·8 ml min−1 was used for AFB1, OTA and FB1 determination. The Jasco HPLC system with UV detector and wavelength settings of 218 nm with a flow rate of 0·6 ml min−1 was used for DON determination.

Results

Standardization of mPCR

The monoplex PCRs for the four genes revealed distinct bands in 1·2% agarose gel (Fig. 1). During optimization of the present mPCR, empirical variations of primer concentrations, amount of MgCl2 and annealing time and temperature were followed. The standardized mPCR had a primer mix at final concentrations of 0·5 μmol l−1 (nor1), 0·8 μmol l−1 (Tri6), 2 μmol l−1 (otanps) and 0·8 μmol l−1 (FUM13). The tenfold serial dilutions of competitive IAC template DNA utilized in the mPCR assay revealed that 0·1 pg IAC was sufficient for unhindered amplification of its competitive counterpart Tri6 gene, besides amplifying other genes in the assay. The standardization of mPCR revealed the presence of five equally intense bands corresponding to IAC (280 bp), nor1 (396 bp), Tri6 (541 bp), otanps (750 bp) and FUM13 (982 bp) in 1·2% agarose gel (Fig. 2). The IAC was amplified independently, irrespective of the nucleic acid load (ranging from 100 ng to 0 ng) in the assay.

Figure 1.

Agarose gel electrophoresis patterns of monoplex PCR amplicons. Lane M – DNA marker (Gene ruler 1-kb plus DNA Ladder; Fermentas); lane 1 – nor1 gene (Aspergillus parasiticus ITCC 456); lane 2 – Tri6 gene (Fusarium graminearum MTCC 2089); lane 3 – otanps gene (Penicillium verrucosum MTCC 4935); lane 4 – FUM13 gene (F. proliferatum MTCC 286); lane 5 – internal amplification control.

Figure 2.

Gel electrophoresis of PCR amplicons obtained from multiplex polymerase chain reaction assay with standard fungal cultures. Lane M – DNA marker (Gene ruler 1-kb plus DNA Ladder; Fermentas); lane 1 – nor1 gene (Aspergillus parasiticus ITCC 456); lane 2 – Tri6 gene (Fusarium graminearum MTCC 2089); lane 3 – otanps gene (Penicillium verrucosum MTCC 4935); lane 4 – FUM13 gene (F. proliferatum MTCC 286); lane 5 – Pooled DNA; lane 6 – Trichoderma viride MTCC 793 and lane 7 – no template control (NTC). Internal amplification control amplification is seen in all the lanes (280-bp band).

Robustness of the multiplex PCR assay

No significant loss in the visibility of the bands was observed when mPCR assay was performed with lesser (×0·8) or greater (×1·2) amounts of PCR buffer when compared with optimized concentration (×1). Annealing temperature variations of ±2°C from the optimal temperature of 58°C affected the mPCR efficiency. At lower temperatures, only the amplification of nor1 and IAC was observed. At higher temperatures, the amplification of Tri6 and FUM 13 was hindered (Fig. S3).

Specificity and sensitivity of multiplex PCR

The specificity of all the primers was assessed by performing mPCR on the genomic DNA of other fungal strains listed in Table 1. No spurious products were observed when the DNA from nonspecific organisms was used. The IAC gene was amplified under all conditions, demonstrating that the PCR mix and conditions were satisfactory. To determine the minimum amount of fungal template necessary to obtain visible amplification products, the mPCR assay was carried out using serial dilutions of fungal genomic DNA and spore suspensions obtained from A. paraciticus ITCC 456, F. graminearum MTCC 2089, P. verrucosum MTCC 4935 and F. proliferatum MTCC 286. The minimum detection level of purified genomic DNA by the optimized mPCR assay was observed to be 100 pg for all four standard strains. When spore suspension dilutions were used, the sensitivity of the mPCR was 103 spores per ml for otanps and Tri6 genes and 104 spores per ml for the nor1 and FUM13 genes. To evaluate the interference of food matrices on the sensitivity of the developed mPCR assay, artificially inoculated (spiked) maize samples were subjected to mPCR. The detection limit for nor1-positive Aspergillus strains, Tri6- and FUM13-positive Fusarium strains was 2 × 103 CFU g−1, whereas for otanps-positive Penicillium strains was 2 × 104 CFU g−1.

Application of multiplex PCR assay to natural food samples

Forty natural samples including maize (12), paddy (10), ginger (05), finger millet (07) and sorghum (06) collected from different retail markets in Mysore were subjected to mPCR assay after a 2-day enrichment step, as well as HPLC after a mycotoxin production step. Fourteen of these enriched samples (five maize, four paddy, one ginger, one finger millet and three sorghum) contained at least one or many of the four mycotoxin metabolic genes as determined by mPCR assay (Fig. 3) and only 12 samples possessed the corresponding mycotoxins, as determined by HPLC (Table 3). The M1, M5 and P16 samples possessed more than one mycotoxin-producing fungi, after enrichment. No food that was determined to be negative by mPCR assay had any mycotoxin in the sample. All the enriched cultures were placed on to PDA, and the corresponding colonies obtained belonged to a variety of genera including Cladosporium, Trichothecium, Aspergillus, Fusarium, Penicillium, Alternaria and Mycelia sterilia (Data not shown).

Figure 3.

Multiplex PCR analysis of natural food samples. Two-day-old enriched food sample cultures were analysed by the developed multiplex polymerase chain reaction assay. Lane M – DNA marker (Gene ruler 1-kb plus DNA ladder; Fermentas); lanes 1 – 12: maize samples M1 – M12; lanes 13 – 22: paddy samples P13 – P22; lanes 23 – 27: ginger samples G23 – G27; lanes 28 – 34: finger millet samples F28 – F34; lanes 36 – 40: sorghum samples S36 – S40.

Discussion

Aspergillus, Fusarium and Penicillium spp. are the leading causes of major foodborne illnesses. Conventional methods for the detection of mycotoxins such as TLC, HPTLC, HPLC, etc., are either time-consuming, costly or laborious. However, for the identification and characterization of toxin metabolic genes in these fungal species, PCR assays have proven to be very useful for rapid and sensitive detection (Niessen et al. 2005 and Niessen 2007). Among the four targeted genes in the present mPCR assay, nor1 is a stable intermediate that encodes for norsolorinic acid reductase involved in aflatoxin biosynthesis (Trail et al. 1994). In Fusarium species, Tri6 encodes an unusual zinc finger protein involved in regulation of trichothecene biosynthesis (Proctor et al. 1995) and FUM13 encoded dehydrogenase/reductase for C-3 carbonyl reduction during fumonisin biosynthesis (Butchko et al. 2003). The otanps gene is responsible for the linkage of the phenylalanine moiety to the polyketide during OTA biosynthetic pathway (Bogs et al. 2006).

The reproducibility of detection by PCR is greatly influenced by the efficiency of the thermal cyclers, DNA polymerases, PCR contaminants and faulty PCR mixes (Hoorfar et al. 2003). These factors can be efficiently dealt with by the integration of an IAC in the diagnostic PCR. The IAC independently amplifies, irrespective of the sample type and nucleic acid load in the PCR master mix. In the present mPCR assay, a competitive IAC designed from pUC 19 plasmid with overhangs corresponding to the Tri6 gene following composite primer technique (Siebert and Larrick 1992) was successfully incorporated. Generally, competitive IAC has favourable advantage over noncompetitive IAC, as the latter requires the standardization of two reactions with different kinetics that proceed simultaneously. Moreover, amplification of noncompetitive IAC may not accurately reflect amplification of the primary target due to differences in the primer sequences (Hoorfar et al. 2004). In our study, the competitive IAC was amplified with the same set of primers (tri6 F & tri6 R) already used in the PCR mix under same conditions.

When naturally contaminated food samples, after enrichment, were subjected to the optimized mPCR assay, two samples (M8 and P15) were found to possess the FUM13 gene but did not contain fumonisin (determined by HPLC) and one sample (M10) possessed the nor1 gene but did not contain aflatoxin (determined by HPLC). The minor discrepancy observed might be the result of many factors that can influence toxin production. Sugar, salt, amino acid levels and moisture content in the production media might also affect toxin production (Jarvis 1971). Or, the intricacies involved in toxin extraction protocol or clean-up columns also would have resulted in a poor yield in mycotoxins. In some cases though, the toxin metabolic pathway genes are present, the ability to produce a particular mycotoxin is found lacking and these isolates may result in false positives in a PCR assay (Russell and Paterson 2006). However, considering the economic importance of these mycotoxins and the pathogenicity of the fungi producing the mycotoxins, a fast and rapid mPCR assay, despite occasional false positives, would still be useful for the high-throughput analysis of fungal isolates because it is the false-negative results that turn a risk into a threat, whereas a false-positive result merely leads to a clarification of the presumptive results by retesting the sample (Hoorfar et al. 2004). The current mPCR assay did not yield any false-negative results.

Conclusion

The mPCR assay we developed for simultaneous detection of major mycotoxin-producing fungi from food samples is a rapid and reliable diagnostic tool. Although not a complete substitute, this method may supplement conventional mycotoxin detection techniques with respect to ease of performance and cost.

Ancillary