Mycotoxins produced from Fusarium
species are a serious problem for grains and maize, and many researchers have reported that Fusarium
toxins may affect livestock and humans in various countries [1
toxins have traditionally been associated with the temperature at which the cultivation, harvest, and storage of cereals occur. These fungi are mesophilic with an optimum temperature for growth and mycotoxin production between 20 and 30 °C. Therefore, many reports have demonstrated the global scale of grain contamination with a number of Fusarium
toxins including fumonisins (FUMs), zearalenone (ZEN), and deoxynivalenol (DON) to name a few [3
Trichothecenes are mycotoxins produced by various Fusarium
genera that are classified into A, B, C, and D type groups according to their chemical properties. DON belongs to type B [6
], which has a double bond with oxygen at the number 8 carbon in the molecular structure [7
]. The chemical structure of DON is shown in Figure 1
DON was discovered by Morooka et al. in 1972 as produced by Fusarium graminearum
on cereals such as wheat, maize, and barley [8
]. DON is a colorless bed-type crystal granule, and has heat stability at 120 °C. DON is soluble in water and some organic solvents [9
]. Protein synthesis is inhibited by this mycotoxin through interference with peptidyl transferase activity combined with intracellular ribosomes [10
]. DON induces impaired growth and weight gain in animals [11
]. In addition, DON causes typical intoxication of livestock and also induces vomiting, leading to a decrease in feed intake and feeding refusal symptoms at high concentrations. In a chronic toxicity test at low concentrations, DON did not induce any changes related to animal behavior; biochemical, hematological, or biochemical characteristics; or immunological parameters [12
]. Swine was highly sensitive to DON. This toxin is classified as a group 3 carcinogen by the International Agency for Research on Cancer (IARC) [9
The occurrence of DON in food and animal feed is a significant problem for the livestock industry and for the supply chain and international trade in food and feed [1
]. Rodrigues and Naehrer [14
] reported a three-year survey related to the presence of mycotoxins in feedstuffs and feed worldwide. Aflatoxins (AFs), ZEN, DON, FUMs, and ochratoxin A (OTA) were detected in 33, 45, 59, 64, and 28% of all analyzed samples including soybean meal, wheat, corn, finished feed samples, and dried distillers grains with solubles (DDGS) from the Americas, Europe, and Asia. DON showed the second highest occurrence.
In China, a total of 56 wheat samples were analyzed for DON contamination levels [15
]. Among them, 89.3% of cereal samples were contaminated with DON in the range of 259 to 4975 μg/kg. In a study in Tunisia, durum wheat samples (65) were contaminated with DON [16
] that ranged from 12.8 ± 5% to 30.5 ± 13.3% μg/g. Pietsch et al. [17
] reported that 81.8% of commercial fish feed samples (11) collected from central Europe were contaminated with DON. The average contamination value was 289 μg/kg feed. In South Korea, DON was detected in 91.3% and 53.3% of compound feeds and feed ingredients, respectively [18
The European Commission (EC) provided recommended guidance values for mycotoxins in animal feed for aflatoxin (AF), ochratoxin (OCT), and other Fusarium
mycotoxins. The USA has also managed aflatoxin, DON, and fumonisins (FUM) in feeds [19
]. In South Korea, the guidance level for AF and OCT in animal feed has been managed according to the guidelines for livestock and fish feed up to 2014. However, recently, levels of Fusarium
mycotoxins in feed have been controlled based on continuous monitoring results and the EC’s recommended guidance values. The values for DON management in the EU [19
] and South Korea [20
] are shown in Table 1
Animal feedstuffs are composed of grains and grain byproducts with vegetable proteins. In 2012, 15,350 tons of feed ingredients were imported by South Korea from many countries such as China, USA, Europe, Canada, South Africa, South East Asia, Australia, and India [21
]. This indicates that not only animal feedstuffs but also feed ingredients including grain, grain by-products, and meal should be controlled for feed safety. Therefore, this study was performed to monitor DON contamination levels in feed ingredients, as well as in animal feedstuffs, over an eight-year period to estimate the tendency of DON contamination.
The contamination of compound feeds by DON has been investigated in many countries. In Poland, DON was detected in 93.5% of 217 compound feed samples that were collected from 2006 to 2009 [24
]. The average DON contamination level ranged from 136 to 225 µg/kg and with maximum contamination concentrations in the range of 409 to 2739 µg/kg. In South Africa, a total of 92 commercial compound feed samples were collected and 70.3% of the collected samples were contaminated with DON [25
]. The mean DON contamination was 766.6 µg/kg, and the maximum value was 2352 µg/kg. Two swine feed samples exceeded the DON contamination limit level for South Africa (1000 µg/kg). In Kuwait, DON was found in 88.8% of poultry feed samples [26
]. The mean DON concentration in poultry feeds was estimated at 261.1 µg/kg, within the range of 220 to 1200 µg/kg. In that study, the concentration of DON was found to be lower than EC-recommended levels. In Turkey, DON was detected in 48.3% of entire feed samples that ranged from 18.5 to 500 µg/kg [27
]. The mean DON contamination level was analyzed as 59.8 µg/kg, and no sample exceeded the allowed level for DON in Turkey (5000 µg/kg for adult ruminants feed and 2000 µg/kg for lamb-calf feed). Another group reported a similar study that was conducted to determine DON levels in feed and feedstuff samples in Turkey [28
]. A total of 106 samples were collected from several farms and feedstuffs manufacturers in Turkey. DON occurred in 18.4% of feedstuffs and 43.3% of feed samples. The highest determination level was 4769.6 μg/kg in maize gluten, indicating that DON content in feed and feedstuffs did not exceed the permitted levels. In 2016, Wu et al. [29
] monitored the DON contamination in feed (2013–2015) obtained from several provinces in China. In that study, a total of 127 samples were analyzed using HPLC, and the highest detection level was at 3346.0 μg/kg in pig feed (pellet) in 2015, which exceeded the guidance values in pig complete feed in China (1000 μg/kg).
In this study, DON contamination was measured in 494 compound feed samples comprising 174 cattle feeds, 160 pig feeds, and 160 poultry feeds that were collected in 2009 and then every two years from 2010, 2012, 2014 and 2016. DON was detected in 95.3% of all of the compound feed samples with a range of 0.4 to 2420.0 µg/kg. DON was analyzed in 97.7% of cattle feeds, 93.1% of pig feeds, and 95.0% of poultry feeds. In the case of cattle and poultry feeds, the maximum contamination level was 2420.0 µg/kg and 1175.2 µg/kg, respectively. No sample exceeded the EU commission levels (5000 µg/kg for other feeds except calf and pig feeds). However, among pig feeds, two compound feeds including gestating sow feeds in 2009 (1566.0 µg/kg) and 2010 (1128.8 µg/kg) exceeded the EC guidance value (900 µg/kg for pig feeds).
Research related to DON contamination levels in feed ingredients has been performed in many countries. In Thailand, Poapolathep et al. monitored 90 wheat and wheat product samples [30
] and showed that 18.9% of total samples were contaminated with DON in the range of 130 to 1130 µg/kg with a mean contamination level of 280.6 µg/kg, suggesting that the risk of DON exposure from wheat products is very low. In China, 83 feed ingredients samples were analyzed, and 95.2% of total samples were contaminated with DON with an average concentration of 1670.2 µg/kg [31
]. The maximum contamination level of DON was 13,139.4 µg/kg, which exceeded the EC guidance value (8000 µg/kg). In the Netherlands, 140 maize silage samples and 20 wheat silage samples were collected, and DON was detected in 72% and 10% of the samples, respectively [32
]. Average concentration levels of DON were 854 and 621 µg/kg, respectively, and maximum concentration levels were 3142 and 1165 µg/kg, respectively, and no samples exceeded the guidance value for DON (8000 µg/kg). In Tunisia, 83% of entire durum wheat samples were contaminated with DON [16
], with an average concentration of DON that ranged from 12,800 ± 5% to 13,300 ± 13.3% µg/kg, which exceeded the EC guidance value for wheat (1750 µg/kg). In Serbia, a total of 289 feed ingredient samples were collected from 2004 to 2007 [33
]. Some samples (33.2%) were contaminated with DON, and the average concentration of DON was 253 µg/kg. Three samples (two maize samples and a wheat sample) exceeded the guidance value. Wu et al. [29
] estimated the level of contamination for DON in a total of 443 feed ingredient samples collected in China. Almost all samples were contaminated with DON (83.3% to 100%), and soybean meal showed the lowest incidence of DON (66.7%). Interestingly, in Hungary, wheat (305) and maize (108) were analyzed to estimate DON contamination levels collected from 2008 to 2015 [34
]. In wheat samples, the highest mean contamination level of 2159 μg/mL and the lowest mean level of 181 μg/mL were observed in 2011 and 2012, respectively. However, in maize, the highest mean value (1261 μg/mL) and the lowest mean value (73 μg/mL) were observed in 2014 and 2009, respectively.
In this study, over half of the collected feed ingredient samples (64%) were found to be contaminated in South Korea. The average DON contamination concentration was 555.3 µg/kg with a range of 0.01 to 8480.0 µg/kg. DON was detected at a concentration of 96.9, 1796.4, 240.8, 361.4, and 22.4 µg/kg in grains, grain by-products, meals, fibrous feed, and food by-products, respectively. No samples exceeded the guidance values for the EU or South Korea (8000 and 1000 µg/kg, respectively). In the case of corn bran, a maximum of 8480.0 µg/kg of DON was detected, but this did not exceed the EC guidance value for corn bran (12,000 µg/kg).
There was a significant difference in the mean contamination level of DON in compound feed and feed ingredients over time in Korea that indicated a decreasing trend. This is mainly due to the continuous monitoring of Fusarium mycotoxins in feeds for many years and to the designation of guidance values for Fusarium mycotoxins since 2015.
4. Materials and Methods
4.1. Chemicals and Reagents
The standard reagent for DON analysis was purchased from Sigma Chemical Co. (St. Louis, MO, USA). Phosphate-buffered saline (PBS) was also obtained from Sigma-Aldrich for elution of DON in immunoaffinity column chromatography. Acetonitrile and methanol used in DON extraction were products of Honeywell Burdick & Jackson (Morris Plains, NJ, USA). The DONPREP kit (R-Biopharm®, Darmstadt, Germany) and DON Test kit (Vicam®, Palo Alto, CA, USA) were used for DON purification. The DON standard reagent was dissolved in acetonitrile to prepare standard solutions of high concentration, which were then diluted with 20% acetonitrile for use in analysis (acetonitrile: distilled water = 20:80, v:v).
4.2. Sampling of Feeds and Feed Ingredients
Contamination levels of DON were analyzed in 653 different feed samples (494 compound feed samples and 159 feed ingredients) produced in 2009 and every other year from 2010 to 2016. These samples were gathered at livestock feed factories from South Korea according to the annual procedure of the Ministry of Agriculture, Food, and Rural Affairs. The descriptions of compound feed and feed ingredient samples are shown in Table 8
and Table 9
. All the samples were preprocessed according to the general guidelines on sampling from the FAO and WHO [35
]. Random sample collection included choosing one kilogram per every ton of feed samples. Samples were collected four times from the same group, and the mixed sample was divided into another four groups. All of the divided samples were subdivided into 200 g and stored in a refrigerator at 4 °C. Detailed classification data for compound feed and feed ingredient samples are shown in the supplementary data (Tables S1–S5
4.3. Extraction and Purification
The animal feed samples were ground to a particle size of 600 μm using a homogenizer, and 20 g of each feed was used as an analytical sample. The feed samples were mixed with distilled water (120 mL), and the mixture was extracted with a homogenizer at 7000 rpm for 2 min. After filtration of the extract through Whatman No. 4 filter paper (GE Healthcare Life Science, Maidstone, Kent, UK), 3 mL of the filtrate was added to an IAC prepared previously in a Vac Elut 20 Manifold (Agilent Technologies, Santa Clara, CA, USA). For an adequate reaction between the IAC packing material and DON, the flow rate was adjusted to 2 to 3 mL per minute. The extracts were passed through the IAC, and, after washing with 5 mL of distilled water, the distilled water was removed using a vacuum pump. The mycotoxin attached to the IAC was eluted with 3 mL methanol, which was slowly dropped under gravity. To increase the elution efficiency, back flushing was performed three times using a syringe before methanol was completely eluted from the IAC. The eluted solution was completely dried at 50 °C using a nitrogen micro-concentrator and re-dissolved in 20% acetonitrile. The re-dissolved solution was filtered through a syringe filter (0.22 μm pore size) and used as a solution for analyzing.
4.4. HPLC Analysis of DON
The concentration of DON in compound feed and feed ingredient samples was measured using HPLC. In analysis, Agilent 1100 series (Santa Clara, CA, USA) including a degasser, auto sampler, a ZORBAX Eclipse XDB-C18 column (4.6 × 250 mm, 5 μm), and a guard column C18 (4.6 × 10 mm, 5 μm) were used at 30 °C. DON was separated using HPLC for 20 min at a flow rate of 1 mL/min and detected with a diode array detector at 220 nm. The mobile phase was composed of HPLC grade water and acetonitrile, which was used in the gradient mode. The retention time was 4.4 min after injection of 20-μL samples.
4.5. Method Validation
The method of HPLC analysis for DON detection was verified by evaluating linearity, LOD, LOQ, accuracy, and precision. All parameters were calculated according to the ICH guidelines [36
]. To determine linearity, the standard curve range was between 50 and 1000 μg/kg (50, 100, 200, 250, 500, and 1000 µg/kg), and the regression equation was calculated using the peak area and concentration of standard solution as parameters. The regression coefficient (R2
) was used to confirm the linearity. The LOD and LOQ were calculated, which were the signal to noise ratio of 3 and 10. To determine accuracy, the recovery test included spiking a blank sample with various concentrations of DON standards, and the results are expressed as the recovery ratio. In this study, the precision indicated the degree of repeatability, and the percent relative standard deviation (%RSD) was used to calculate precision.