Deoxynivalenol in wheat and wheat products from a harvest affected by fusarium head blight

As a result of its nutritional properties, wheat (Triticum aestivum L.) is one of the most frequently eaten cereals all over the world by both humans and animals (Vieira, 2006). Brazil is not self-sufficient in the production of this cereal; producing less than half of its consumption needs, and such production is subject to market fluctuations. Because of milder temperatures, the southern region of Brazil accounts for 94.6% of the national production of wheat. However, given the characteristics of the cropping system, the average grain yield in this region is not the highest in Brazil (Empresa Brasileira de Pesquisa Agropecuária, 2013). In addition, the quality of products after wheat processing is directly related to the quality of grains to be processed; therefore, careful cultivation, harvesting and storage of this cereal are crucial (Vieira, 2006).


Introduction
As a result of its nutritional properties, wheat (Triticum aestivum L.) is one of the most frequently eaten cereals all over the world by both humans and animals (Vieira, 2006).Brazil is not self-sufficient in the production of this cereal; producing less than half of its consumption needs, and such production is subject to market fluctuations.Because of milder temperatures, the southern region of Brazil accounts for 94.6% of the national production of wheat.However, given the characteristics of the cropping system, the average grain yield in this region is not the highest in Brazil (Empresa Brasileira de Pesquisa Agropecuária, 2013).In addition, the quality of products after wheat processing is directly related to the quality of grains to be processed; therefore, careful cultivation, harvesting and storage of this cereal are crucial (Vieira, 2006).
In plantations, wheat can be contaminated by various diseases because of weather conditions, soil type and crop susceptibility.Dne of the best-known diseases that commonly affect this cereal is Fusarium head blight, triggered by infection of Fusarium fungi, which not only cause diseases in plants but also produce toxic substances known as mycotoxins through their secondary metabolism (Calori-Domingues et al., 2007).
Mycotoxins are naturally-occurring toxic compounds, produced by a variety of fungal species that grow on agricultural products, during their growth in the field and in storage, as well as in processed food and animal feed (Scussel, 2002).They cause significant economic impact because they reduce plant and animal productivity, and also toxicological impact, with clinical manifestations in both humans and animals (Santos, 2009).Deoxynivalenol (DDN) is a trichothecene mycotoxin, produced mainly by Fusarium graminearum, and it is one of the most frequently found in small cereal grains (Bando et al., 2007;Freire et al., 2007).Production of DDN in wheat is considered a virulence factor of F. graminearum (Maier et al., 2006).DDN is also known as vomitoxin, because it causes vomiting, especially when consumed by pigs.It is toxicologically relevant, and its synthesis is stimulated by high humidity (Miller, 1995;Beyer et al., 2005), with high incidence in winter cereals, such as wheat and wheat products consumed by humans (Miller, 1995;Mallmann et al., 2003;Calori-Domingues et al., 2007).The World Health Drganization (WHD) considers DDN as a neurotoxin with a teratogenic nature and immunosuppressive characteristics, and, as trichothecenes in general, it has been associated with chronic and fatal intoxication of humans and animals through consumption of contaminated food (Rotter et al., 1996).
Because the occurrence of fungal infection serves as a warning against mycotoxin contamination, it is worrying that wheat crops harvested in 2014/2015 in southern Brazil were affected by high levels of Fusarium head blight.In this context, the aim of this study was to evaluate the mycotoxicological quality of Brazilian wheat grains and wheat products (wheat flour and wheat bran) for DDN contamination during the year of 2014.

Materials and methods
DDN analyses were conducted on the premises of the Laboratory of Mycotoxicological Analyses (LAMIC), Federal University of Santa Maria -RS, Brazil.

Samples
A total of 1504 samples of wheat and wheat products (668 of wheat grain, 697 of wheat flour and 139 of wheat bran) were collected from southern Brazil between January and December 2014, and analyzed according they were acquired.The samples were being sold and/or used by cooperatives and agriculture companies.

Standards and solvents
The DDN standard was purchased from Sigma (Sigma-Aldrich, Alcobendas, Spain).Acetonitrile and methanol were purchased from J. T. Baker (Deventer, The Netherlands).All solvents were LC grade.Water was purified with a Milli-Q purification system.

Preparation of the sample
The toxin was extracted according to the official method no.986.17 from Association of Dfficial Analytical Chemistry (1995), with some modifications.Three grams of the previously ground sample were extracted in vortex mixers with 24 ml of the solvent mixture containing methanol: water (70: 30, v/v).Sample cleanup was not performed in this method.The extracts of the samples were diluted in a combination of aqueous phase, water: ammonium acetate (995: 5, v/v) and organic phase (water: ammonium acetate: methanol (95: 990: 5, v/v/v), and 5 μL were injected into a liquid chromatography system.

Determination of DON
DDN was determined by separation with high-performance liquid chromatography, and detected by sequential mass spectrometry (LC-MS/MS).The isotopes 13 C 20 -Don and 13 C 20 -Zea were used as internal standard.For quantification of mycotoxins, a seven-point calibration curve was prepared (200,400,600,1000,2000, 4000 and 8000 µg.Kg -1 ).A validated method was used for sample preparation, extraction and dilution (LDD: 50 µg.Kg -1 ; LDQ: 200 µg.Kg -1 ).For the collected data, it was assumed that the calibration curve showed a linear behavior (r 2 >0.9).

Statistical analysis
The average contamination of all samples and average contamination of positive samples, prevalence and percentage of samples above LMT of DDN in wheat grain, wheat flour and wheat bran were calculated in each month of 2014.Tukey test at 5% significance was used to compare the average results of each month.

Results and discussion
Table 1 shows the results of the occurrence of DDN in wheat and wheat products (flour and bran) in 2014.A total of 1000 samples (66.5%) were contaminated with DDN, with mean levels of 1574.75 µg.kg -1 .Contamination was most common in November with 399 (90.9%) of positive samples and less frequent in April, with 38 (36.9%) positive samples.
A study by Calori-Domingues et al. (2007) evaluated the occurrence of DDN in wheat grown in Brazil and imported from abroad.It was found that 94% of the Brazilian wheat and 88% of the imported wheat were contaminated with mean levels of 332 and 90 µg.Kg -1 , respectively.In addition, 4% of the Brazilian wheat samples analyzed had contamination levels above 1250 µg.Kg -1 , which is the maximum permissible level (MPL) in accordance with European Community law; however, this level is still allowed in Brazil.Although the extent of contamination was greater in that study, the mean levels presented by the authors were below those observed in the period evaluated in the present work, which included samples from crops containing high levels of Fusarium head blight (in the second semester).Wheat flour was the fraction showing the lower prevalence, while wheat grain and bran presented the highest prevalence of DDN.A study by Nowicki et al. (1988) showed that the distribution of DDN in ground wheat grain fractions is influenced by the degree of fungal penetration into the grain endosperm, and this susceptibility is dependent on plant variety.Those authors found that when penetration was low, higher levels of infection and DDN were found in the grain surface and, consequently, low concentrations of DDN were recovered in the flour.
A more recent study indicated that in the post-harvest stage, the cleaning, aeration, debranning and milling procedures have an influence in the distribution of mycotoxins in wheat fractions.In the milling process there is no mycotoxin reduction, although mycotoxin concentrations may be redistributed and concentrate according to the milling fractions.Mycotoxins tend to be concentrated in outer fractions intended for animal feed (bran, flour shorts screenings and middlings) and lower in inner fractions intended for human consumption (flour or semolina) (Cheli et al., 2013).
Dn the other hand, the not significant difference between the fractions contamination is not in accordance with findings reported in the literature (Nowicki et al., 1988;Samar et al., 2003;Lancova et al., 2008).However, it should be noted that the samples analyzed in this study were independent, and there is no interrelationship between the whole grain and the different fractions.In addition, knowing this survey deals with samples from two different seasons, involving different wheat cultivars yielded in 2014 and naturally showing different sensitivity to FHB, the variation in levels of contamination may have influenced the average results of DDN and consequently the statistical results.
Twenty one percent (21% n = 145) of the analyzed wheat grain samples had contamination levels above 2000 µg.kg -1 (the maximum permissible level for wheat grain for further processing by current Brazilian law (Agência Nacional de Vigilância Sanitária, 2011Sanitária, , 2013)), and the highest level of contamination found for the samples was 11800 µg.kg -1 , confirming the low mycotoxicological quality of this cereal from the year 2014.Among the wheat flour samples, 10% (n = 71) were out of the limits currently allowed by Brazilian regulation.
According to the Agriculture Federation of Rio Grande do Sul (FARSUL), as a result of the legislation for mycotoxins (enforced in Brazil since 2011), if the allowed limit (1750-2000 µg.Kg -1 ) is exceeded, wheat is not suitable for either human or animal consumption.It should be noted that new values will take effect as of 2017.The same organization also estimates that one million out of the 1.7 million tons produced in the last harvest in Rio Grande do Sul would be affected, and 700,000 tonnes of low quality wheat had been traded until December 9th, 2014 at low prices to international markets willing to absorb this production with low mycotoxicological quality (Batchold, 2015).
Figure 1 shows the mean of monthly concentration and the maximum level found in the evaluated samples of wheat and wheat products.The highest mean of DDN contamination was in November, when the samples of the new wheat crop started to be introduced (2014/2015).
According to Companhia Nacional de Abastecimento ( 2015), in the 2014/2015 harvest in Rio Grande do Sul, the area sown with wheat was as big as 1.14 million hectares, which represents an increase of 9.8% over the previous harvest.However, the same government agency informed that during the growth of the crop, several adverse factors -frost, torrential rains, floods, lack of light, excessive heat, disease attacks in general, hail and lodging -influenced the end result.Buerstmayr et al. (2009) reported that in Southern Brazil (latitude 28 S and 23 S), where wheat crops are located, the main limiting factor for wheat production is the high relative humidity, which favors the occurrence of outbreaks of fungal diseases, especially Fusarium Head Blight (FHB).This disease is influenced by warm temperatures and rainy days during the flowering time, causing losses in grain yield and reductions in baking and seed quality.
Within this framework, Rio Grande do Sul produced 1.5162 thousand tons in 2014, 52.3% less than the harvest in 2013, reflecting a decrease by 56.5% in productivity (Companhia Nacional de Abastecimento, 2015).
Several studies worldwide (Love & Seitz, 1987;Miller, 1995;Trigo-Stockli et al., 1995;Moschini & Fortugno, 1996;Dsborne & Stein, 2007) have correlated periods of high humidity with high incidence of Fusarium head blight in wheat crops, which consequently tends to cause higher levels of DDN in wheat and wheat products (Lori et al., 2003;Vanheule et al., 2014), as noted in our study in the final quarter of 2014.

Conclusions
Considering wheat as a raw material widely used in the food industry and the high levels of DDN contamination observed in certain months of the year, especially in samples from the harvest presenting high incidence of Fusarium head blight, it is crucial to carry out appropriate mycotoxicological monitoring of the quality of wheat and wheat products traded to prevent the population exposition to levels above those allowed by regulation.

Figure 1 .
Figure 1.Mean results and monthly maximum concentrations of DDN in wheat grain, wheat bran and wheat flour analyzed in 2014.
The variables were calculated using Microsoft Excel software (Microsoft Excel 2016 MSD, Microsoft Corporation, Redmond, WA) and analyzed in the statistical program Statgraphics Centurion XV (Statgraphics Centurion 15.2.11,Manugistics Inc., Rockville, MD).

Table 1 .
Dccurrence of deoxynivalenol contamination in Brazilian samples of wheat grain, wheat flour and wheat bran harvested in 2014.