Ochratoxin A (OA) is a nephrotoxic and nephrocarcinogenic mycotoxin produced by Penicillium verrucosum in temperate or cold climates and by a number of species of Aspergillus such as A. ochraceus in warmer and tropical areas of the world. OA may be involved in the etiology of Balkan Endemic Nephropathy in the Balkan countries.^11,12 It is commonly found in both human food and animal feed, especially in temperate areas including Denmark, Germany^13,14, Canada¹¹, United Kingdom¹⁵ and Sweden.¹⁶ Currently eight countries have specific regulations for OA in one or more food commodities. In addition, a few countries have proposals for OA regulations. The current proposed limits range from 1 to 5 m g/kg for children and infant foods, from 2 to 50 m g/kg for food and from 5 to 300 m g/kg for animal feed.¹⁷In Europe, tolerance levels for OA have been suggested at 1 m g/kg for infant food and 5 m g/kg for cereals.¹⁸ Regulations for OA in some food commodities including cereals has been recently set at 3 m g/kg in Italy.¹⁹

The high humidity and moderate temperatures in the subtropical Caspian littoral of northern Iran could promote the growth of these toxigenic fungi and the production of mycotoxins. In this regard, several toxigenic fungi such as the Aspergillus, Penicillium and Fusarium species have been isolated from cereals produced in these regions.^20,24 Furthermore, it has been shown that these fungi could produce mycotoxins in culture.^20,24,25 There are some reports on natural occurrence of mycotoxins in certain Iranian food commodities.^26,27 Surveys have shown natural occurrence of AF in bread^28,29 and corn.³⁰ With regard to OA, its natural occurrence in red wheat has been reported.²⁰

At present, establishment of regulations for some mycotoxins is underway by the Iranian Standards Organization in collaboration with the Ministry of Health and Medical Education.

No information is so far available in the literature on the natural occurrence of AF in barley and OA in barley and corn samples of northern Iran. The present study was therefore undertaken to determine the natural contamination levels and possible co-occurrence of AF and OA in corn and barley in northern Iran.

During the year 1998, 9 and 14 samples of farmers’ corn and barley lots, intended for animal feed, were collected at random from consignments sold to the Iranian Agriculture Office in Mazandaran and Golestan provinces, respectively. Samples (0.5 to 5 kg) were collected from the lots, personally carried to Italy, and individually stored in zip-lock plastic bags at 4° C until analysis in which each sample was ground to a meal and after thorough mixing, a 25-50 g subsample was withdrawn for analysis.

Extraction and immunoaffinity clean-up:

1) Barley

Samples were analyzed using the method of Stroka.³¹ Ground sample (50 g) was weighed into an Erlenmeyer flask and after adding 5 g of sodium chloride and 300 mL of methanol: water (8:2), the flask was shaken on a wrist-action shaker set at medium speed for 45 minutes. After filtration through a filter paper, 10 mL of the filtrate was diluted with 60 mL of phosphate buffer solution (PBS) (pH 7.4). Afla Test immunoaffinity column (Vicam, Watertown, MA) was conditioned with 10 mL PBS at a flow rate of 2-3 mL/min.

Whole-diluted extract was then applied to the column, at a steady flow rate of 3 mL/min. After washing the column with 15 mL of bidistilled water, air was drawn through the column until dry. AF was eluted from the column into a 3-mL calibrated volumetric flask in a 2-step procedure. At first, 0.5-mL methanol was applied on the column and passed by gravity. After 1 min, a second portion of 0.75 mL was applied and most of the applied evasive solvent collected by forcing air through. Finally, it was filled to the mark with water and mixed.

2) Corn

Samples were analyzed using the Afla Test procedure.³² In brief, 50 g of ground sample (plus 5 g of sodium chloride) was extracted with 100 mL of methanol: water (8:2). Then, 10 mL of filtered extract was diluted with 40 mL of purified water, mixed well and filtered through glass microfiber filter into a clean vessel. Ten milliliters of filtered diluted extract was passed through Afla Test immunoaffinity column. The column was then washed with 10 mL purified water. AF were eluted from the column using 1 mL HPLC grade methanol. One mL purified water was then added to the eluate.

Since the recovery of AFB1 in the two highest level AF-contaminated samples was about 70%, we tried to verify the reason for this matter. In this regard, the method of analysis of AF in barley³¹ was modified for analysis of these samples as follows: after extraction and filtration, 10 mL of filtered extract was diluted with 60 mL of PBS, mixed well and filtered through glass microfiber filter. For extract clean-up, two different immuno-affinity columns including Afla Test and AFLAPREP (Rhone-Diagnostics Technologies Ltd., UK) were used for each sample. Sixty mL of filtered diluted extract was passed through immunoaffinity column and AF were eluted as described for barley.

Samples were analyzed by HPLC equipped with a JASCO AS-950-10 autosampler and a JASCO FP-920 fluorescence detector after post-column derivatization with bromine. The analytical column used was C18 (Alltech), 5 mcm, 250 mm ´ 4.6 mm i.d. Mobile phase was water: methanol: acetonitrile (54:29:17) at a flow rate of 1 mL/min. The post-column pump was Waters 510 and the post-column reagent was saturated bromine in water at a flow rate of 0.4 mL/min. The fluorescence detector was operated at an excitation wavelength of 365 nm and an emission wavelength of 435 nm.

Two different five-point calibration curves, with triplicate injections, covering the range of 0.07-0.65 m g/kg and 0.70-6.5 m g/kg for AFB1 and AFG1, 0.04-2.6 m g/kg and 0.2-1.3 m g/kg for AFG2 and 0.02-0.13 m g/kg and 0.2-1.3 m g/kg for AFB2 were built and checked for the linearity. For both calibration curves, correlation coefficients were higher than 0.999 for all four AFs. Quantification was based on an external standard of AF. Limits of detection for toxin peaks were defined as 3 times baseline signal and were 0.07 m g/kg for AFB1 and AFG1, 0.04 m g/kg for AFG2 and 0.02 m g/kg for AFB2.

Extraction and immunoaffinity clean-up:

1) Barley

Samples were analyzed using an in-house validated method. Ground sample (25 g) was extracted with 100 mL of acetonitrile: water (60:40) by a wrist-action shaker set at medium speed for 45 minutes. Four mL of the filtrate was diluted with 44 mL of PBS and filtered diluted extract was then added to the Ochra Test immunoaffinity column followed by 10 mL water wash. OA was eluted from the column using four 1 mL portions of methanol. The eluate was evaporated under a stream of nitrogen at 40° C and then redissolved in 1-mL mobile phase.

2) Corn

Samples were analyzed using the Ochra Test (Vicam, Watertown, MA) procedure.³³ The extraction of ground sample (50 g) and extract dilution were carried out as described for analysis of AF in corn using the Afla Test procedure. Ten mL of filtered diluted extract was passed through Ochra Test immunoaffinity column. The column was then washed with 10 mL mycotoxin wash buffer solution followed by 10 mL purified water. OA was eluted from the column using 1.5 mL HPLC grade methanol. Finally, 1.5 mL purified water was added to eluate.

Samples were analyzed by HPLC equipped with a JASCO AS-950-10 autosampler and a JASCO FP-920 fluorescence detector. The analytical column used was C18 (Alltech), 5 mcm, 250 mm ´ 4.6 mm i.d. Mobile phase was acetic acid 2%: acetonitrile (49:51) at a flow rate of 1 mL/min. The fluorescence detector was operated at an excitation wavelength of 333 nm and an emission wavelength of 470 nm.

A five-point calibration curve, with triplicate injections, covering the range of 0.24-20 m g/kg was built and checked for the linearity. Correlation coefficient was 0.9999 for OA. Quantification was based on an external standard of OA. Limit of detection for OA was 0.24 m g/kg (signal/noise: 3/1).

One spiked sample was analyzed with each series of samples (usually 5-7) to check percentage of recovery of OA and AFG1, AFG2, AFB1 and AFB2. After spiking of samples with known amounts of mycotoxins, the solvent was allowed to evaporate overnight. Then the sample was simultaneously analyzed together with other samples and the recovery calculated. All analytical results were corrected for recovery. The obtained recovery factors are shown in Table 1.

Statistical analyses of the performance of analytical methods used for analysis of AF and OA

Table 1. Statistical analysis of the performance of analysis methods of aflatoxins (AF) and ochratoxin A (OA) in corn and barley samples.

* Coefficient of variation reproducibility.

** Samples analyzed by modified method of analysis of AF in barley and Afla Test (Vicam, Watertown, MA) column.

*** Sample analyzed by modified method of analysis of AF in barley and AFLAPREP (Rhone-Diagnostic Technologies Ltd, UK) column.

in corn and barley samples are shown in Table 1.

The recovery of AF and OA in barley and corn samples were good and in the acceptable range (Table 1). The coefficient of variation for reproducibility of all spiked samples with AF or OA ranged from 0.8 to 21.2%.

The recovery of AF in corn samples spiked with 40 ppb of AFB1 and 40 ppb of AFG1 and analyzed with Afla Test procedure were between 65-68%. In order to verify the reason of this matter, spiked samples were analyzed with a modified method of analysis of AF in barley, using two different immunoaffinity columns including Afla Test and AFLAPREP. In this method, we increased the ratio of extraction solvent to corn from 2 to 6. In addition, we passed about 1.4 g of corn through the column in comparison to 1 g in the first analysis but the recoveries were not significantly changed (69.2-74.4%) (Table 1).

1) Corn

The concentration, mean, median and incidence of positive samples for AF in corn samples in northern Iran are given in Table 2.

On the whole, 88.8% of the samples were contaminated with AFB1 at concentrations ranging from 0.09 ppb to 73.10 ppb (mean and median level, 15.83 and 1.72 ppb, respectively). Furthermore, 66.6% of the samples were positive for AFB2 at concentrations ranging from 0.02 to 12.70 ppb (mean and median level 2.99 and 1ppb, respectively). AFG1 and AFG2 were not detected in any of the corn samples. The concentration of aflatoxin total (AFT) ranged from 0.09 to 85.30 ppb with a mean and median level of 18.82 and 1.83 ppb, respectively. Typical chromatograms relative to AF standards and corn sample are shown in Figure 1.

As shown in Table 2, no significant difference was observed between the AF concentrations of samples, which were analyzed with both Afla Test and AFLAPREP immunoaffinity columns.

On the co-occurrence of both carcinogenic mycotoxins, only one of the AF-contaminated corn samples (No. 6) was co-contaminated with OA.

oats, wheat and corn grown in Denmark and other Scandinavian countries as well as in the Balkan countries and India are particularly susceptible to high levels of OA contamination.^35-37 In addition, in Canada, Europe and Australia, animal feed may also be highly contaminated with OA, sometimes at concentrations exceeding 5000 ppb.³⁵ In the present study, only 11 % of corn samples were contaminated with OA at very low concentrations and it was not detected in any of the barley samples. Clearly, the incidence of OA found in corn kernels from Mazandaran Province compare favorably with the incidence found in wheat reported by Lacey.²⁰ He reported that among 141 red wheat samples only one was contaminated with OA. Concerning the simultaneous occurrence of three carcinogenic mycotoxins, aflatoxins, fumonisins and ochratoxin A, the highest level of AFT-and FM-contaminated corn (sample No. 6) was co-contaminated with OA.

The present study reports the natural occurrence of OA and co-occurrence with AF in the corn of northern Iran for the first time. In addition, along with the results of analysis of FM in Mazandaran corn samples²⁷ this study described, for the first time the simultaneous occurrence of Aspergillus toxins (AFT and OA) together with fumonisins in corn from northern Iran. Although the reported levels of AF and especially OA were low (due to suitable climate of these areas for fungal contamination), a high level of contamination may occur. On the other hand, toxigenic fungi have also been isolated from cereals produced in other parts of Iran.^21,22 Taking into consideration all of the above concerns, and with regard to the presence of some diseases with unknown etiologies (like esophageal cancer in Northeast of Iran), comprehensive studies including a periodical survey on natural co-occurrence of mycotoxins in cereals is needed in order to have a broader perspective on the extent of mycotoxin problem in Iran.