INFLUENCE OF STORAGE PERIOD ON OCCURRENCE AND DISTRIBUTION OF AFLATOXINS AND FUNGI IN MAIZE KERNELS

This study had two major objectives: 1) to investigate the influence of a seven-month storage period on maize contamination with aflatoxins (AFs) and 2) to examine the distribution of total count of fungi (TCF), A. flavus and aflatoxin B1 (AFB1) in the stored maize kernels. In the first part of this survey, 700 maize samples were collected in the period from October 2012 to April 2013. Presence of AFs was detected in 72% of maize, before entering the silos. The survey results indicated that the percentage of contaminated maize samples as well as the distribution of determined AFs concentrations changed during the storage period of seven months. By the prolongation of storage period, the percentage of non-contaminated maize samples decreased from 28% to 16%, while the percentage of contaminated samples in the concentration range 20-50 μg/kg increased from 18% to 25%. In the second part of this study, 32 maize samples from four different silos were analysed. The results showed that TCF, A. flavus and AFB1 were unevenly distributed. According to Duncan’s multiple range test, in each silo, statistically significant differences (p<0.05) were noted for all tested parameters. Furthermore, within each silo, statistically significant correlation (r=0.76 at p<0.05) between the number of A. flavus colonies and AFB1 concentrations were observed. According to the obtained results and taking into account the trends in climatic changes in Serbia, further improvement in maize management system concerning AFs is warranted. The findings of this study could be of great importance to increase the knowledge related to the AFs management strategies in Serbia as well as in neighboring countries.


INTRODUCTION
Maize (Zea mays L.) is an important cereal worldwide and one of the major crops grown in Serbia. Most of the total maize production in Serbia is mainly used for animal feed (80%), while the rest is used for human consumption and starch production. Furthermore, in the recent years Serbia has been one of the largest maize producer and exporter in Europe as well as in the whole world (Maslac, 2013;Maslac, 2015).
One of the most frequent contaminants of maize is mycotoxigenic fungus Aspergillus (A.) flavus which can result in aflatoxins (AFs) production (Cotty and Mellon, 2006). AFs represent one of the most important mycotoxins group due to their prevalence and intensive toxic, mutagenic, teratogenic and carcinogenic effects, with proven acute and chronic toxicity in humans and animals (IARC, 2012). Among eighteen identified AFs, aflatoxin B1 (AFB1) is the most frequent one and potent liver carcinogen known in mammals. Besides AFB1, International Agency for Research on Cancer (IARC) defined AFB2, AFG1, and AFG2 as primary carcinogenic compounds (IARC, 2012). A. flavus is a very common and widespread fungus in nature which grows and produces AFs during different phases of maize growth, harvest, processsing, transportation and storage (Ellis et al., 1991). Since maize as starch-based material represents a good substrate for aflatoxigenic fungi, the most influential factors affecting the occurrence of A. flavus and the amount of AFs in maize are environmental and storage conditions. The influence of environmental conditions on the production of AFs during maize growing season was reported in many studies (Abbas et al., 2002;Juan et al., 2008;Kos et al., 2013). In recent study, Karami-Osboo et al. (2012) highlighted drought as the main factor affecting the production of AFs in maize. Furthermore, agronomic factors (type of hybrid, soil, tillage, previous crop), postharvest management, inappropriate storage conditions (temperature, humidity, handling, presence of insects, rodents and birds) as well as extended storage period may influence on increasing risk of A. flavus and AFs presence in stored maize (Hell et al., 2000;Garrido et al., 2012). Some authors reported that contamination with A. flavus and subsequent production of AFs during storage is considered as one of the most serious safety problems throughout the world. In addition, it is known that AFs are stable compounds that cannot be destroyed during most of the maize processing operations, therefore improvements in agricultural practices, in particular grain production and storage environments is required (Williams et al., 2004).
Until 2012, occurrence of AFs in maize was not considered a typical agricultural issue in Serbia Kokić et al., 2009). However, weather condition changes during maize growing season in 2012 resulted in AFs outbreak in maize kernels. Our previous investigation showed that prolonged drought influenced AFs presence in 137 (68.5%) maize samples among total 200 samples analysed (Kos et al., 2013). Furthermore, presence of AFs in maize resulted in toxin presence in feedstuffs (Lević et al., 2013a) as well as in food of animal origin, especially in milk (Kos et al., 2014). Occurrence of AFs in Serbia during 2012/2013 period caused "aflatoxins crisis" followed with numerous protests of agricultural workers and milk producers in the street, several changes of Regulations regarding maximum level (ML) and significant economic losses for processors, producers and marketers (Torović, 2015). It was assumed that the "aflatoxin crisis" during 2012/2013 could have cost Serbia between 100 and 125 million euros (Radović and Keković, 2014).
Even though the presence of AFs in 2012-2013 maize had led to huge economic losses, available official data as well as the national database on AFs occurrence have still been insufficient. Furthermore, significant changes in climatic conditions during previous years in Serbia require further investigations especially in relation to improvement in maize management system concerning the presence of A. flavus and AFs. Therefore, the first objective of this study was to examine the effect of storage period (seven months) on AFs contamination of maize kernels and the second objective was to investigate the distribution of total count of fungi (TCF), A. flavus and AFB1 in the stored maize.

Samples, kits and chemicals
For the first part of investigation, seven hundred (n=700) maize kernel samples were collected. Every month, in the period from October 2012 to April 2013, 100 maize samples were collected, transported to the laboratory and analyzed. Only in October 2012, 100 samples were analyzed immediately after harvest without previous storage. The samples were taken from different silos which were representative in size, composition and storage method applied. It means that during the storage period there was a constant effort to minimize any possibility of maize quality reduction caused by changes in storage conditions (temperature and humidity), as well as mechanical damage influenced by insects, rodents and birds.
In order to examine the distribution of TCF, A. flavus and AFB1 concentration, thirty two (n=32) maize samples from four silos were included in the second part of investigation. Storage capacity of silos was around 800 t. The sampling was performed during the elevation procedure from eight different sections of each silo.
With the aim to provide the representative maize samples for the first and the second part of investigations, sampling procedure was performed according to the European Union (EU) requirements (European Commission Regulation, 2006a). Aggregate samples of approximately 5-10 kg were composed of different number of incremental samples. These samples were homogenized and quartered to obtain a 500 g of laboratory samples which were kept in freezer at -18 °C until the analysis. The obtained representative samples were ground to 1 mm particle size using a laboratory mill (KnifetecTM 1095 mill, Foss, Hoganas, Sweden).

Aflatoxins analysis
Due to high number of analysed samples (n=700), samples from the first part of this study were analyzed using validated Enzyme Linked Immunosorbent Assay (ELISA) method, while 32 samples from the second part were analyzed using High Performance Liquid Chromatography with Fluorescence Detector (HPLC-FLD).

ELISA method
Determination of AFs (AFB1, AFB2, AFG1 and AFG2) in 700 maize samples was performed by ELISA method using two test kits produced by Neogen Corporation (Quantitative AFs HS Test kit and AFs Quantitative Test kit, Neogen Veratox®, Lansing, USA). Chemicals used for ELISA method were distilled water (Millipore, BedFord, MA, USA) and methanol of analytical purity (Merck, Darmstad, Germany). Subsamples of 5 g were extracted with 25 ml of methanol:water mixture (70:30, v/v) and shaken vigorously for three minutes on laboratory Griffin flask shaker (Griffin and George, Wembley, England). Extracts were filtered through the Whatman No. 1 filter paper (Whatman International Ltd., Maidstone, UK). The instructions given by the manufacturer for AFs determination were strictly followed.

HPLC-FLD method
HPLC-FLD method was applied for the analysis of AFB1 in 32 maize samples from the second part of investigation. For HPLC analysis acetonitrile, methanol, nhexane and trifluoroacetic acid (TFA) were purchased from Merck (Darmstadt, Germany). Used water was ultrapure (Milli-Q from Millipore, USA). Aflatoxin B1 standard with certificated concentration of 2 µg/ml was purchased from Sigma Aldrich (Prague, Czech Republic). Twenty-five grams of maize samples were extracted with 100 ml acetonitrile:water (84:16, v/v) and shaken vigorously for thirty minutes using laboratory Griffin flask shaker (Griffin and George, Wembley, England). After extraction, extract was filtered through filter paper (Whatman No. 4, Maidstone, UK) and 5 ml of filtered extract was cleaned up with Mycosep® 224 AflaZon multifunctional columns (Romer Labs. Inc., Union, MO, USA). The purified extract was then evaporated to dryness (Reacti Term, Thermo Fisher, Scientific Bellefonte, P. A. USA). Due to the poor fluorescence of AFB1, post derivatization step for enhancing its fluorescence on HPLC-FLD was required. Derivatization was achieved by adding 100 µl of TFA and 200 µl of nhexane to sample extracts or AFB1 working standards. This mixture was vortexed for 30 s and kept in the dark at 40 °C for 10 min. After evaporation, 400 µl of acetonitrile:water (1:9, v/v) mixture was added to the vials and vortexed again for 30 s. The HPLC instrument was an Agilent 1200 system equipped with a fluorescence detector (FLD), a binary pump, a vacuum degasser, an autosampler and Agilent column (Eclipse XDB-C18, 1.8 µm, 4.6 x 50 mm). The AFB1 analysis was performed with a mobile phase consisted of 65% of water and 35% of a mixture of acetonitrile:methanol (50:50, v/v) with a flow rate of 0.2 ml/min. Ten microliters of standards or samples were injected into the HPLC column. The fluorescence detector was set to an excitation and emission wavelengths of 365 nm and 440 nm, respectively.

Mycological survey
Analyses of TCF in maize kernel samples were performed according to ISO standard (ISO, 21527-2, 2008). Colonies that were assumed to belong to Aspergillus species were transferred on Czapek Yeast Extract Agar (CYA).
Seeded surfaces were incubated during the period of 7 days at 25 °C. The criteria described by Samson et al. (2004) and Pitt and Hocking (2009) were applied for species identification. Taxonomic classifycation was determined on the basis of macromorphological and micromorphological characteristics of growing colonies.

Quality control
Analyses of all maize kernel samples were carried out in accredited laboratory of the Institute of Food Technology, University of Novi Sad. The laboratory was accredited in agreement with standard SCS ISO/IEC 17025 (2006).
The analytical quality of the applied ELISA and HPLC-FLD methods were assured by the use of certified reference material (CRM). Naturally contaminated maize sample (Progetto Trieste, Test Veritas, Padova Italy) with certified AFB1 (4.98±3.54 µg/kg) and AFs (6.46 ± 4.42 µg/kg) content was used as the CRM. The validation parameters were calculated and expressed using European Official Decision procedure (European Commission Regulation, 2002).

Statistical analysis
Statistical analysis of variance was carried out by Duncan's multiple comparison tests using STATISTICA software version 10 (StatSoft Inc. 2011, USA). P values < 0.05 were regarded as significant.

RESULTS AND DISCUSSION
The results of the validation methodology are shown in Table 1. The validation parameters, for both used ELISA test kits as well as HPLC-FLD method, were in compliance with recommendations given in the Regulation 2006/401/EC (European Commission Regulation, 2006a).
Seven hundred maize kernel samples were analyzed with the aim to investigate influence of storage period of seven months on AFs contamination. The obtained results for each month in the period from October 2012 to April 2013 are summarized and shown in Fig. 1. Determined AFs concentrations were distributed in the following five concentrations ranges: <1 µg/kg, 1 µg/kg-10 µg/kg, 10 µg/kg-20 µg/kg, 20 µg/kg-50 µg/kg and >50 µg/kg. As it can be seen from Fig. 1. significant variation in the distribution of AFs concentrations within investigated months were noticed. The greatest percentage (28%) of maize kernel samples that were not contaminated with AFs (c< 1 µg/kg) was noted in maize samples collected in October 2012. The possible reason for that is the fact that only maize samples from October were analyzed immediately after harvest without previous storage. On the other hand, in the maize samples which were stored and collected from the silos, the percentage of non-contaminated maize samples decreased (from 26% in November to 16% in March) with the prolongation of storage period.  (2006) reported that maize samples stored in Uganda for more than six months were contaminated with higher concentration of AFs in comparison to the maize samples stored in a period from two to six months. It should be emphasized, that storage facilities in Benin as well as in Uganda are significantly different in comparison to storage facilities in Serbia, which were included in this study. Therefore, complete comparison of obtained results is difficult. Similarly, Keller et al. (2013) concluded that prolongation of storage period had a great influence on the increase of AFs contamination in stored maize silage. In the second part of this study, a total of 32 maize samples were collected from four different silos with the aim to investigate distribution of TCF, A. flavus and AFB1 in stored maize kernels. As it can be seen from Table 2, the obtained results indicated significant differences in the distribution of investigated parameters within each silo. In regard to TCF, it could be noticed that in the examined maize samples, TCF varied from 85.0 to 230 cfu/g x 10 3 cfu/g, 105 to 320 cfu/g x 10 3 cfu/g, 90.0 to 320 cfu/g x 10 3 cfu/g and 50.0 to 550 x 10 3 cfu/g in the first, second, third and fourth silo, respectively. Table 2.
Fungal mycobiota and aflatoxin B1 contamination of maize samples collected from four silos Significant differences in terms of TCF were noted within each silo, which means that mycobiota were very unevenly distributed. The currently valid Serbian Regulation for microbiological criteria for food (Pravilnik, 2010), does not define the maximum TCF in maize. However, earlier edition of this Regulation (Pravilnik, 1993) defined 10 x 10 3 cfu/g as maximum TCF allowed in maize kernels. In all 32 analyzed maize samples, TCF was greater than 10 x 10 3 cfu/g. A. flavus was present in all of 32 analyzed maize samples with significant individual variations within each silo. Minimum and maximum counts of A. flavus colonies were distributed in the following way: 6.20-57.5 x 10 3 cfu/g in the first silo, 6.10-105 x 10 3 cfu/g in the second, 7.00-40.0 x 10 3 cfu/g in the third, and 10.0-50.0 x 10 3 cfu/g in the fourth silo. Among 32 analyzed maize samples, AFB1 was not detected in only two samples from the second and one sample from the fourth silo. Moreover, the AFB1 concentrations within each silo were very differ-rent. The AFB1 concentration ranges were: 10.4 -88.1 µg/kg, 16.5 -183.0 µg/kg, 9.50 -71.9 µg/kg and 9.70 -150.0 µg/kg for the first, second, third and fourth silo, respectively. The total count of fungi, A. flavus and AFB1 were unevenly distributed in the examined silos and according to Duncan's multiple range test (P < 0.05) statistically significant differences were noted for each of the examined parameter (TCF, A. flavus and AFB1) in every silo. The largest contamination range was noted in the second silo: the sample contamination with AFB1 and A. flavus spanned from 16.5 µg/kg to 183 µg/kg and from 6.10 to 105 cfu/g x 10 3 cfu/g, respectively. Furthermore, in the same silo, among eight samples, only two were not contaminated with AFB1 at concentration over 1 µg/kg. The largest number of determined A. flavus colonies (105 cfu/g x 10 3 cfu/g) caused the highest determined AFB1 concentration (183 µg/kg) (Silo 2, sample 8). Furthermore, it could be noticed that the smallest number of A. flavus colonies (6.10 cfu/g x 10 3 cfu/g) was determined in the sample with AFB1 concentration below 1 µg/kg (Silos 2, sample 1). However, correlation may not be so straightforward as the presence of A. flavus colonies in large number does not necessary mean the presence of the high concentrations of AFB1 as well as that the lower number of A. flavus colonies does not necessary mean the absence of AFB1. This is in accordance with the study reported by Harley et al. (1997) which indicated that mycotoxins content is not always related to the amount of fungi present. However, the results from our study showed statistically significant correlation (r=0.76 at p<0.05) between the number of A. flavus colonies and AFB1 concentrations observed in each of the four silos. Some authors also confirmed heterogeneous distribution of A. flavus and AFB1 as well as occurrence of AFB1 in pockets with different concentrations in stored contaminated maize (Trung et al., 2008; Luftullah and Hussain, 2012). Castellari et al. (2010) reported that Aspergillus species represent one of the most common fungal genera identified in stored maize. The proliferation of these fungi as well as AFs biosynthesis may be stimulated by the following factors: moisture content, high temperature during storage, long storage period, and intensive infection by fungi before storage and by higher activity of insects, rodents and birds. Therefore, it is important to identify the factors with the strongest impact on the fungal growth and toxicogenesis. Drought conditions recorded during the 2012 maize growing season, especially during the period from June to September, were associated with a decrease in the moisture content (average value was around 13%) which resulted in unfavourable conditions for the growth of most fungal species, except certain Aspergillus species (Maslac, 2013). Lević et al. (2013b) reported that over the period 1967-2008 frequency of A. flavus in maize from Serbia varied from 2.9 to 16.0%. However, in 2012 registered incidence of A. flavus was 95.3%. Authors claimed that extremely stressful agrometeorological condition, over the period from flowering to waxy maturity of maize in 2012, was the main reason for such high incidence of A. flavus on maize. Furthermore, as a second factor causing intensive occurrence of A. flavus and AFs authors highlighted prevalence of European corn borer (ECB) (Ostrinia nubilalis). ECB influenced high damage and numerous injuries which were covered by visible olive-green powdery colonies, typical for A. flavus.
It could be noticed that weather conditions recorded during the 2012 maize growing season greatly influenced the high initial AFs contamination (72%), which was further increased by the prolongation of storage period. During the observed 7-month storage period all factors that could contribute to maize quality deterioration were minimized (temperature and humidity control, pest control). Therefore, based on everything stated above it could be concluded that factors with the strongest influence on the high AFB1 incidence in the stored maize were severe initial infection and contamination of maize kernels before storage as well as the duration of storage period.

CONCLUSIONS
Based on the findings obtained in this study it can be suggested that maize kernels infected by fungi and contaminated with AFs should not be stored without previous application of some procedure for reduction of initial AFs contamination.
Considering the recorded changes in weather conditions in Serbia in recent years, given consequential economic losses and numerous negative impacts of AFs on human and animal health, a practicable and realizable strategy for reducing the risk of AFs contamination of maize should be developed. This strategy should be related to the implementation of monitoring program as well as some physical, chemical and/or biological procedures for AFs decontamination, with the main aim to avoid storage of contaminated maize. In order to prevent AFs presence, it is necessary to make every possible effort to improve agricultural practices (in particular, irrigation system), AFs control, and storage environments.