1. Introduction
Aflatoxin is a potent mycotoxin naturally produced as a secondary metabolite by several fungal Aspergilli species, primarily A. flavus and A. parasiticus [1]. Aflatoxin contaminates a variety of food and crop commodities such as corn, peanuts, spices, tree nuts, figs, honey, fruit juices, wine, and cotton [2-9]. It is estimated that 25% of the world’s crops are affected by mycotoxins, according to the evaluation of the Food and Agriculture Organization [10]. Aflatoxin B1 is labeled as the most potent naturally produced carcinogen with its primary mammalian target being the liver [11,12]. The four main aflatoxins frequently encountered are B1, B2, G1 and G2. They are designated by the corresponding color they fluoresce (blue or green) under UV light, and the numbers assigned to them are given based on the relative distance traveled on a thin layer chromatography (TLC) plate [1,13]. A. parasiticus produces aflatoxin B1, B2, G1 and G2 while A. flavus produces aflatoxins B1 and B2 [14,15]. Aflatoxin B1 is the most toxic of the four toxins [16,17]. Upon ingestion, aflatoxin B1 may also be metabolized into aflatoxin M1, which is expressed in milk of lactating animals, and thereby contaminating dairy products [18,19]. A. flavus and A. parasiticus are saprobic fungi ubiquitous in nature, commonly found living off of decaying debris in the soil and on crops [20,21]. Because of aflatoxin’s carcinogenic nature, it is regulated in feed and food in more than 100 countries [22]. The US Food and Drug Administration has set action levels for aflatoxin at 20 parts per billion (ppb) for all foods intended for human consumption, except for milk which is set at 0.5 ppb [23]. The feed limit for finishing beef cattle is 300 ppb and for dairy cattle the feed limit is 20 ppb.
The United States is the world’s largest grower of corn (Zea mays), which is impacted by aflatoxin contamination annually. In addition to its usage in food and feed products, corn is the main substrate used in ethanol production [24]. The composition of dent corn is 72% starch and 9.5% protein [25]. The high starch composition in corn can easily be broken down into the simple sugars necessary to make ethanol. When corn is fermented, two key co-products are formed: coarse unfermented distillers grains or dried corn solids (DCS), and the liquid or stillage which is composed of yeast, fine grain particles, and soluble nutrients. When these products are combined and dried they are referred to as dried distillers grains and solubles (DDGS) [26]. These co-products are sold to livestock producer for animal feed [27,28].
A simplistic outline of ethanol production [29-31] using corn starts by grinding entire corn kernels into coarse flour, known as meal, which is then mixed with water to form “slurry”. Alpha-amylase is added to break down the starch from corn into short chain dextrin sugars, and the slurry is heated for liquefaction. After liquefaction, the slurry is referred to as “corn mash”. A second enzyme, glucoamylase, is then added to break down the dextrins to form glucose in a process known as saccharification. Finally, yeast is added to the mash and is transferred to fermenters where the yeast converts the simple sugars into ethanol and carbon dioxide. In commercial application, carbon dioxide is usually captured and sold for use in carbonating soft drinks and for use in manufacturing dry ice. After fermentation, ethanol is distilled from the fermented mash with the remaining “stillage” containing solids from the grain, yeast and water. The distilled ethanol can then be prepared and sold as fuel. The stillage can be centrifuged, separating the water from the solubles, which can be sent back to the slurry tanks to help recycle water input. Schingoethe [28] reports that 100 kg of corn produces approximately 40.2 L of ethanol, 32.3 kg of DDGS, and 32.3 kg of carbon dioxide.
The economic success of ethanol plants relies in part by the marketability of its DDGS [26]. It is therefore crucial to know how aflatoxin is affected by the fermentation process for the sustainability of ethanol plants, the health of animals fed DDGS and for the health and safety of humans who may consume food processed from these animals. Several studies have already investigated mycotoxin contaminated meal through ethanol fermentation processes [32-35]. All previous studies agreed that no mycotoxins were detected in the distilled alcohol. However, it is of interest to note that several studies were at variance with one another terms of the fate of aflatoxin during fermentation. Two studies reported that aflatoxin is degraded during fermentation [33,36]; one [36] was based on traditional brewing and the other [33] simulated industrial fermentation procedures. Two additional studies based on industrial fermentation procedures reported that aflatoxin did not degrade during the fermentation [34,35].
Many studies have investigated into the detoxification of aflatoxin using heat, gamma radiation, nixtamalization, and ammoniation [35,37-41] and have shown varying amounts of success in reducing aflatoxin in feeds and foods. Another alternative that is gaining attention is the option of incorporating binders/sorbents (sequestering agents) in the contaminated feed meal that would bind aflatoxin, thereby making the meal non-toxic and safe for animal consumption [21]. Combining binders with the DDGS could be an easy and cost effective strategy, requiring minimal modifications to the ethanol plant facilities. Currently, the US Food and Drug Administration (FDA) allows the use of binders as anticaking agents in animal feed not exceeding 2 percent, but does not approve the use of binders for removing aflatoxin as it is concerned with the uniformity and degree of mycotoxin binding [42-45]. Some recent studies have assessed a variety of binders for detoxification of aflatoxin and how they affect the physiology of animals that consume them [44-49]. These studies have shown a variety of sorbents and binders can lower aflatoxin in a diverse collection of animals when fed aflatoxin contaminated feed that contained these adsorbents. These studies primarily focused on the aflatoxin binding properties of hydrated sodium calcium aluminosilicates (HSCAS), activated carbons, and esterified glucomannans with many showing promising binding of aflatoxin exceeding 95% bound at low concentration of adsorbent. This study was undertaken to determine the fate of aflatoxin during ethanol fermentation of corn and to explore a strategy using binders to help salvage contaminated DDGs for market.
2. Materials and Methods
2.1. Corn Sample Information
Store bought corn meal (yellow 100% whole grain) was used as a negative control (no aflatoxin was detected using the Veratox® method described below). To ensure that aflatoxin levels could be detected among the aliquots taken during ethanol fermentation of corn, four samples of highly aflatoxin-contaminated ground corn were obtained from the United States Department of Agriculture, Agricultural Research Service (USDA ARS), Corn Host Resistance Unit, Mississippi State, MS. The corn was planted in a randomized complete block design with four reps with approximately 20 plants per row. The corn was infected by side needle injections with A. flavus (NRRL 3357) inoculum 14 days after mid-silk (mid-silk is when 50% of the plants on a row have visible silks). Inoculum injected was a 3.4 mL suspension containing 3 × 108 A. flavus conidia. The ears were harvested by hand approximately 60 days after inoculation and dried at 38˚C for 5 to 7 days and then machine shelled. Samples were ground using a Romer mill (Romer Laboratories Inc., Union, MO, USA). The mean aflatoxin level across all reps was 10,800 ppb.
2.2. Preparation and Fermentation of Corn
Four corn samples and their replicates were subjected to a lab-scale ethanol fermentation process. All samples were treated the same. Both subsamples of corn underwent the fermentation process concurrently. One hundred grams of ground corn sample were combined in a 2 L flask with water until a final volume of 1 L was reached. The pH was then measured using litmus paper. All samples were within the pH range of 5.5 - 6.0. The corn meal slurry was brought to a rapid boil on a hot plate while mixing with a stir bar and then allowed to cool to between 45˚C - 74˚C. Next, 0.1 g of α-amylase (Mile Hi Distilling; Lakewood, CO) was added. This converts starch to maltose and oligosaccharides by randomly hydrolyzing α-1,4-glycosidic bonds. Ten mL of 4000 mg/L Ca2+ solution (1 M CaNO3 and distilled water) was added to the slurry, stabilizing and activating the enzyme. The mixture was allowed to incubate for 45 minutes to allow the enzymes to liquefy the mash. Afterwards, the pH was adjusted to 4.0 before adding 0.1 g of glucoamylase (Mile Hi Distilling; Lakewood, CO). This converts the long chain sugars produced by α-amylase into glucose monomers by cleaving α-1,4-glycosidic and α-1,6-glycosidic bonds from the non-reducing ends. The efficiency of the enzymatic conversion of starch to sugar was measured in the mash using a Brix refractometer.
The digested mash was then allowed to cool and settle for 30 minutes. The supernatant from the mash was then removed from the solids (coarse grains). Supernatant was used to wash the solids to retrieve any bound sugars from the supernatant. The coarse grain portion was then set aside and allowed to completely dry forming dried corn solids (DCS). The supernatant was cooled to 37˚C using an ice bath before stirring in approximately 0.3 g of yeast, Saccharomyces cerevisiae (ATTC, Manassas, VA). The mash was incubated at 29˚C for four days in a loosely sealed container.
2.3. Distillation of Ethanol
The fermented mash was distilled for ethanol on a microscale still consisting of a Bunsen burner, fractionating column, thermometer, Graham condenser and graduated cylinder for collection of ethanol. Distilled ethanol fractions were collected between 78˚C - 85˚C. To ensure that no contamination of aflatoxin occurred between samples, a rinse solution of 50:50 ethanol: water was distilled through the fractionating column and Graham condenser three times, collected and analyzed via LC-MS/MS.
2.4. Preparation, Extraction and Analysis of Samples
2.4.1. Corn
The total aflatoxin in the four corn samples obtained for this experiment was determined by two immunoassays: AflaTest® and Veratox®, as well as by LC-MS/MS. The samples were extracted and analyzed as indicated and outlined below.
2.4.2. DCSs
The total aflatoxin in the DCSs fractions collected from the experiment were analyzed by AflaTest®, Veratox®, and LC-MS/MS. For the two immunoassays, the samples were extracted and analyzed as indicated by the test kit instructions. For LC-MS/MS analysis, the extraction of aflatoxin from DCSs followed the Veratox® extraction procedure using 5 g of grounded sample in 25 mL of 70:30 methanol: water. The filtrate was then diluted to fall within the standard curve. An additional clean-up step was needed before LC-MS/MS analysis (outlined below).
2.4.3. Fermented Mash
The total aflatoxin in the fermented fractions collected from the experiment were analyzed by LC-MS/MS. Aliquots taken from the fractions collected were cleaned-up using a 0.45 µm nylon filter before injection with no dilution step necessary. Fermented mash fractions were not a suitable matrix for the two-immunoassay analysis as indicated by the test kit instructions.
2.4.4. Distilled Ethanol
The total aflatoxin in the distilled ethanol fractions collected from the experiment were analyzed by LCMS/MS. Aliquots taken from the fractions collected were cleaned-up using a 0.45 µm nylon filter before injection with no dilution step necessary. Distilled ethanol fractions were not a suitable matrix for the two-immunoassay analysis as indicated by the test kit instructions.
2.4.5. Stillage
The total aflatoxin in the distilled ethanol fractions collected from the experiment were analyzed by LCMS/MS. Aliquots taken from the fractions collected were cleaned-up using a 0.45 µm nylon filter before injection with no dilution step necessary. As per the instructions, stillage fractions were not a suitable matrix for the immunoassays.
2.5. Immunoassays Methods
2.5.1. AflaTest® Assay
AflaTest®, manufactured by VICAM, is an immunoaffinity column (IAC) method using monoclonal antibodybased affinity chromatography and is approved by AOAC International and the Federal Grain Inspection Service (FGIS). This method was followed as directed by the instruction manual. Aflatoxin is extracted by mixing 50 grams of ground sample with 100 mL of a 70:30 methanol/water mixture and then filtered. The filtered extract is then diluted and filtered again before passing the filtrate over AflaTest® affinity column. The column is then washed with water and the aflatoxins are eluted with methanol into a cuvette. Developer is then added to the eluate before reading the sample with a calibrated Series-4EX Fluorometer (VICAM, Watertown, MA) to determine aflatoxin concentration. AflaTest® detection limits range from 1 - 300 ppb for grains.
2.5.2. Veratox® Assay
Veratox®, manufactured by Neogen, is a direct competetive ELISA using polyclonal antibodies and is approved by Grain Inspection, Packers and Stockyards Administration (GIPSA) and FGIS. Aflatoxin is extracted by vortexing 5 g of sample that has been ground to a fine particle size with 50 mL of a 70:30 methanol/water solution before filtering through a Whatman Filter #1. The filtrate is then diluted if necessary before being sampled and mixed with enzyme-labeled toxin (conjugate). The mixed solution is transferred to antibody-coated wells, where free toxin and conjugate compete for antibody binding sites. The unbound conjugate and other soluble phase substances are then rinsed away and a substrate is added. Color develops as a result of the presence of bound conjugate. Red stopping reagent is added and the color of the resulting solution is observed. Samples were analyzed on Neogen Stat Fax Microwell Strip Reader (Neogen, Lansing, MI). Absorbance (OD650) readings of the samples are compared with OD650 readings of the controls, and the concentration of total aflatoxin in the samples are calculated in ppb. Veratox® has a lower detection limit of 2 ppb with a range of quantitation from 5 - 50 ppb.
2.6. LC-MS/MS Conditions
For sample separation, a Varian ProStar dual solvent delivery module (Palo Alto, CA) coupled to with a Bruker Esquire Mass Spectrometer (Billerica, MA) capable of tandem MS was used for positive identification. Sample volumes of 50 µL were injected at 25˚C with a Varian ProStar autosampler onto a 100 × 2.1 mm i.e. 3 μm Ultra Aqueous C18 (Restek, Bellefonte, PA, USA) column. A gradient elution method (Table 1) was used for the mobile phase consisting of water with 0.1% formic acid plus 1.25 mM of ammonium acetate and methanol. The retention times of aflatoxins B1, B2, G1, and G2 are as follows based on elution: 16 min for G2, 16.6 min. for G1, 17.1 min. for B2, and 17.6 min for B2. A secondary UV scan at a wavelength of 356nm was conducted using a Waters 2587 module (Waters Corporation, Milford, MA). For the MS method, electrospray ionization was used in the positive mode yielding [M+H]+ parent ions 313, 315, 329, 331 m/z of the aflatoxins B1, B2, G1, and G2, respectively. The most intense daughter ions, resulting from collision-induced dissociation were used for identification for B1 (285.1 + 298.1), B2 (297.1 + 287.1 + 259.1), G1 (311.1 + 301.1 + 283.1 + 243.1), and G2 (313.0 + 285.1 + 303.1).
2.7. Standards Used for LC-MS/MS Analysis
The standard curve for LC-MS/MS analysis was made by combining separate stock solutions of aflatoxins: 2.0 µg/mL B1, 0.5 µg/mL B2, 2.0 µg/mL G1, 0.5 µg/mL G2 (SUPELCO; Bellefonte, PA) into a working solution of 800 ng/mL total aflatoxin, each aflatoxin equaling 200 ng/mL. A seven point standard curve plotting 8, 40, 80, 200, 400, 600, 800 ng/mL total aflatoxin (2, 10, 20, 50, 100, 150, 200 ng/mL for each of the four individual aflatoxins) was injected on the LC-MS/MS using the same method as described above with an overall correlation coefficient, R2, ranging from 0.9979 to 0.9990 of all four aflatoxins.
2.8. Preparation and Analysis of DCSs Mixed with Binders
Evaluation for decontamination of aflatoxin in DCSs using binder, MTB-100®, was carried out on DCSs made from contaminated corn sample 3. Triplet positive control samples were analyzed, each sample containing 5g of contaminated DCSs (approximately 14,000 ppb). Duplicate samples containing binder and DCSs made from contaminated corn sample 3 were analyzed, each sample prepared by mixing 1 g of MTB-100® (Alltech; Lexington, KY) with 5 g of contaminated DCSs. All five samples were stored at 4˚C for seven days to ensure binding before having the aflatoxin extracted as outlined in the Veratox® extraction procedure described above. The filtrate of all the samples were then passed through a 0.45 µm nylon filter as a clean-up step before analysis by the LC-MS/MS method described previously. Quantification was accomplished using peak areas. All solvents were HPLC grade (purchased from Sigma Aldrich, St. Louis, MO).
Table 1. Gradient method used for the mobile phase in separation of samples for the LC-MS/MS method.
2.9. Analysis of Dried Distillers Grains with Solubles
Dried Distillers Grains with Solubles (DDGs) were provided by Bunge-Ergon Vicksburg, LLC ethanol facility. Aflatoxin B1 standard was purchased from Sigma Aldrich and used to create a 3.00 μg/mL spiking solution in methanol. DDG samples were spiked with this aflatoxin solution to reach a concentration of 300 ppb. Each sample was then manually mixed for 6 min. Aflatoxin was then extracted using the procedure outlined previously. The pH portion of the study used a similar methodology, with all testing samples consisting of 9.8 g of DDG, 0.2 g of binder. Following the addition of 50mL of MeOH and vortexing, the pH was modified using dilute HCl and/or NaOH. The pH was then adjusted to the range of 10, 9, 8, 7, 6, 5, and 4. The solutions were then left at this pH for an hour, then readjusted to the pH of 7 for analysis. After settling occurred (one hour), the extraction solvent was carefully decanted and tested using the Veratox® kit. Finally, the temperature range portion of the experiment was conducted in a similar manner to the pH experiment. Samples consisted of 9.8 g of DDG and 0.2 g of binder. 50 mL of MeOH was added and promptly vortexed, again the pH was adjusted to 8, 7, or 6 depending on the experiment. The solutions were then placed in an ice water bath (0˚C) or a hot water bath (40˚C).
3. Results
The total aflatoxin in the four aflatoxin-contaminated corn samples obtained for this experiment were measured by the AflaTest® method and are as follows: Sample 1: 7600 ppb, Sample 2: 10,800 ppb, Sample 3: 14,800 ppb, and Sample 4: 11200 ppb. Veratox® gave comparable values of the contaminated corn samples (shown in Table 2). Compilation of the results for the total aflatoxin of four duplicate samples of aflatoxin-contaminated corn plus a duplicate negative control (no aflatoxin detected) is found in Table 2. The corn and DCS samples were analyzed for total aflatoxin by LC-MS/MS, Veratox®, and AflaTest®. The fermented mash, distilled ethanol, and stillage fractions were analyzed for total aflatoxin only via LC-MS/MS because the matrices were unsuitable for the two-immunoassay methods. No aflatoxin was detected in any of the fractions collected from the negative control corn sample. For the aflatoxin-contaminated corn samples, no aflatoxin was detected in the distilled ethanol fractions. Also, no aflatoxin was detected in the rinse fractions, which were distilled through the fractionating column and Graham condenser after each contaminated sample was distilled for ethanol. Ethanol yields averaged 2.1% (v/v). Aflatoxin was detected in all fractions collected from the fermented mash with total aflatoxin levels ranging from approximately 127 - 196 ppb. Also, aflatoxin was detected in all stillage fractions collected with total aflatoxin levels ranging from approximately 133 - 195 ppb. The highest aflatoxin levels were measured in the DCS fractions. The means with standard deviation total aflatoxin levels for the four DCS samples and their replicates are as follows: 9867 ± 3000 ppb, 9756 ± 3000 ppb, 14,208 ± 4000 ppb, 14,149 ± 3000 ppb. The percentage of aflatoxin recovered was 31% - 58% in the DCSs, 13% in the fermented mash and 13% in the stillage.
Testing the effect temperature and pH have on the binding of aflatoxin in DDGs with MTB-100 indicates a dependence upon the two variables. DDGs were analyzed using the Veratox® method with the initial control concentration of aflatoxin being 300 ppb. Binding efficiency as a function of binder concentration plateaued to 80% with the addition of 6% (w/w) binder, at 10% concentration binder resulted in a binding efficiency of 88%. At a 2% (w/w) binder exhibited 72.5% in aflatoxin binding (Figure 1). Binding efficiency as a function of pH found that at 2% w/w binder in room temperature conditions (25˚C) the optimum binding occurs at pH 6 amongst the pH range of 4 - 10 resulting in 81.3% binding reducing the aflatoxin concentration to 56.0 ppb. Aflatoxin binding decreased at the pH range of 7 and 8, 42.2% and 52.9% respectively, compared to the pH of 6 resulting in an aflatoxin concentration of 173.4 ppb and 141.2 ppb (Figure 2). A separate experiment was conducted to determine what effect pH had on the amount of recovered aflatoxin in the absence of binder (results not shown). The results indicate that there is no variation in the amount of aflatoxin recovered as a function of pH from pH range 4 - 8 with a dramatic reduction of aflatoxin recovery at pH 9 and 10. Binding efficiency as a function of temperature showed reduction in aflatoxin binding capability at both hot and cold temperatures.
Cold temperature had a dramatic effect upon the binding efficiency with temperatures near 0˚C resulting in near negligible binding at pH of 6 and a maximum binding percentage of 19.7% at a pH of 8 resulting in a final aflatoxin concentration of 240.9 ppb (Figure 3). Hot temperature had a somewhat muted effect on the binding efficiency of MTB-100® with the pH of 6 having 36% binding resulting in a final concentration of 192 ppb and the pH of 7 and 8 having a percent binding of 47 and 45% binding resulting in a final aflatoxin concentration of 159 ppb and 165 ppb respectively (Figure 4).
Figure 5 shows the chromatogram obtained from LC-MS/MS analysis of aflatoxin-contaminated DCSs (14,000 ppb) collected from DCSs fractions from corn sample 3 in the ethanol fermentation experiment previously described. The chromatogram shows decrease levels of detected aflatoxin in DCSs mixed with binders (2 replicates) when compared to the positive controls (3 replicates) containing only DCSs. The binders reduced aflatoxin levels by 80%.