The effect of buckwheat groats processing on the content of mycotoxins and phenolic compounds

ABSTRACT Two buckwheat groats processing methods were used for production of final commercial product. The first one involved thermal processing (steamed) and then dehulling, and the second one dehulling without thermal treatment (raw). The research evidenced that the raw groats and hulls were several times more contaminated with aflatoxin B1 compared with steamed ones. High concentrations of aflatoxin B1 (75.8 µg kg−1) and T-2 toxin (351.0 µg kg−1) were detected in the raw hull samples. The total phenolics responded more sensitively to thermal treatment than phenolic acids. More than 20 times higher concentrations of quercetin (65.47 ± 6.3 µg g–1) were determined in steamed hulls compared to other raw and steamed samples. Buckwheat groats and hulls, containing the highest concentrations of quercetin and hydroxybenzoic acids, were found to be 10-fold less contaminated with aflatoxin B1 and T-2 toxins; however, the correlations between the phenolics and mycotoxins were statistically insignificant.


Introduction
Buckwheat (Fagopyrum Esculentum Moench) is a popular plant on the global health market. Owing to its chemical composition, buckwheat grain is suitable for use in the production of functional foods and food additives (Matić, Mastiloviã, Čabarkapa, & Mandiã, 2009). The colour of buckwheat groats depends on the dehulling method. Two methods of buckwheat dehulling are usedthermal and non-thermal. In the thermal treatment, buckwheat grains are saturated with moisture up to 22% of the dry weight and are steamed at 150-164°C temperature, sometimes from 130°C to 160°C under 5.5-6 bar pressure, then cooled, conditioned, separated by sieving into fractions and hulled. The result is a brown colour of grain (Antonov, Makevnina, Ivko, Iskusnov, & Korobov, 2013). According to the manufacturing company, during the nonthermal treatment raw grains are dehulled directlybuckwheat grains are moistened and dried until hulls become dry and kernels are moist and soft. Dry hulls break during the rolling process and moist kernels remain intact and of natural colour. Afterwards the kernels are dried. The final yield of kernels ranges from 50% to 60%.
Dehulling processes of buckwheat grain affect the content of biologically active compounds and nutritional value of groats. Research evidenced that thermal treatment of buckwheat flour is the cause of changes in the protein, crude lipid, fibre and ash contents (Pandey, Senthi, & Fatema, 2015), and that thermal processing methods caused a decrease in the total phenolics, total flavonoids and antioxidative activities of Tartary buckwheat flour extract (Dziedzic, Górecka, Marques, Rudzińska, & Podolska, 2015;Zhang, Chen, Li, Pei, & Liang, 2010). Another criterion associated with a decrease in the quality and quantity of food raw materials in food chain is the presence of mycotoxins. Unfortunately, buckwheat has been recently included in the list of cereals and cereal-derived products which may be susceptible to invasion by Aspergillus flavus, and consequently, aflatoxin contamination (EFSA, Supporting Publications 2013: EN-406), especially, if buckwheat grain is stored for a long time. Aflatoxins remain essentially unaltered during food processing and, for this reason, could be present in processed foodsflour, pasta, baked goods and can reduce the quality and safety of products obtained from buckwheat (Chitarrini et al., 2014). Buckwheat hulls have found application in the manufacture of household goods, such as pillows, mattresses, toys. Therefore, it is vital that not only groats, which are a dietary product, but also hulls are high quality, free from toxin and mould fungi contamination, which can cause allergenic reactions (Fritz & Gold, 2003). However, buckwheat is rich in phenolic compounds, which occur naturally in a range of food plants and display antimicrobial, antifungal activity, therefore it could be used for improving food safety as markers for tolerance against mycotoxigenic pathogens (Ansari, Anurag, Fatima, & Hameed, 2013;Lattanzio, Lattanzio, & Cardinali, 2006;Samapundo et al., 2007). The toxicity of phenolic compounds to pathogens is dependent on phenolic compounds' chemical structural features and is related to the presence of hydroxyl functions in the aromatic structure, which evidences that increased hydroxylation results in increased toxicity (Lattanzio et al., 2006;Teixeira, Gaspar, Garrido, Garrido, & Borges, 2013). Flavonoids account for a considerable part of chemical compounds found in buckwheat (Fabjan et al., 2003). Rutin, quercetin are included in the plant cellular structures and are characterized by a significant antioxidant activity (Dietrych-Szostak, 2004;Lattanzio et al., 2006). According to Chitarrini et al. (2014), quercetin, derived from rutin has strong antifungal inhibition against mould fungi. Lattanzio et al. (2006) suggest that free phenolic compounds, such as rutin and quercetin have stronger antifungal properties than bound forms of phenolics. Generally, found esterified or bound to cell wall in buckwheat are phenolic acids, and only a minor fraction exists as free compounds; however, phenolic acids may account for about one-third of the total phenolic compounds in our diet and can help to preserve stability of processed products in food industry (Lattanzio et al., 2006;Teixeira et al., 2013).
The present study was aimed to determine the concentration of mycotoxins, establish the change of phenolic compound concentration in response to different treatments of buckwheat groats and hulls and evaluate the relationship between phenolic compounds and mycotoxins.

Materials and methods
Samples of buckwheat groats and hulls (n = 24) were collected from a Lithuanian manufacturing company and examined for total phenolics, rutin and quercetin contents. Seven phenolic acids (p-hydroxybenzoic, 3,4-dihydroxybenzoic, p-coumaric, ferulic, vanillic, syringic and sinapic) were identified and quantified. The fungal infection was estimated and identified and mycotoxins T-2 toxin (T-2), ochratoxin A (OCHA), aflatoxin B1 (AFLB1) concentrations were determined. The buckwheat groats samples were prepared from the grain grown organically in Lithuania. Analyses of mycotoxins and phenolic compounds (total and individual) were done in Akademija, Institute of Agriculture, Lithuanian Research Centre for Agriculture and Forestry.

Preparation of extracts
All samples were milled in an IKA A11 Basic mill (Staufen, Germany) and stored at +4°C until analysis. Mycotoxin extraction was performed according to manufacturer's instructions. AFLB1, T-2 and OCHA were extracted according to the R-Biopharm Ridascreen ® (Darmstadt, Germany) instruction.
Phenolic acids were extracted according to Kvasnička et al. (2008). One gram (±0.001 g) of ground sample was weighed and 25.0 ± 0.1 ml of 0.1 M NaOH added, shaken in a water bath Memmert WNB 14 (Memmert, Schwabach, Germany) at 40°C for 60 min, cooled to room temperature, acidified with 2 M HCl to pH 5-6 and supplemented with 20.0 ± 0.1 ml of methanol. The flask was placed in an ultrasonic bath (Bandelin Electronic, Berlin, Germany) for 30 min, cooled to room temperature and made up to volume with methanol. The filtrate after filtration by 0.22 µm membrane filter (Frisenette ApS, Knebel, Denmark) was analysed by high performance liquid chromatography (HPLC).

Mycotoxin and phenols assay methods
Mould fungi infection level on groats and hulls were determined by the agar media method (Mathur & Kongsdal, 2003) on the potato dextrose agar medium (PDA). The fungal grain infection was estimated and identified according to the manual of Sutton, Fothergill and Rinaldi (2001). Analysis of T-2, OCHA, AFLB1 was carried out using an ELISA commercial kit Ridascreen ® (R-Biopharm AG, Darmstadt, Germany). The basis of the test is the antigen-antibody reaction. The optical densities of samples and controls from standard curve were estimated by a multichannel photometer Multiskan Ascent (Thermo Electron Corp., Vantaa, Finland), supplied with internal software, using a filter of 450 nm for T-2 (limit of detection (LOD)~3.5 µg kg −1 ), OCHA (LOD 2.5 µg kg −1 ), AFLB1 (LOD 1 µg kg −1 ). Mycotoxin analyses of each treatment were repeated twice. A Shimadzu HPLC system was employed for determination of phenolic acids, rutin and quercetin content and consisted of the following modules: system controller SBM-20A, auto injector SIL-20A, UV-Vis detector SPD-20A, low-pressure gradient flow control valve LC-20AT, column oven CTO-20A, on-line degasser DGU-20As. Data collection and evaluation were performed by using operating system LCsolution Workstation (Shimadzu, Horiyamashita, Hadano, Kanagawa, Japan). Phenolic acids were separated by the method described by Amarowicz and Weidner (2001). Rutin and quercetin were separated by the method described by Fabjan et al. (2003) - Table 1. TPC was determined using Folin-Ciocalteu reagent after the method described by Sedej, Sakač, Mišan and Mandić (2010) with some modifications. TPC was conducted by mixing 7.9 ml of deionized water, 100 μl extract, 0.5 ml Folin-Ciocalteu and 1.50 ml 20% sodium carbonate (after 6 min at room temperature) was added. The absorbance (after 120 min) was measured at 760 nm using a UV/VIS spectrophotometer PerkinElmer LAMBDA 25 (PerkinElmer, Waltham, Massachusetts, U.S.A.). A standard curve (0.05-1.5 mg ml −1 ) was prepared with rutin. Final results were expressed as mg of rutin eq. g −1 dry weight (d.w.).
Conventional statistical methods were used to calculate means and standard deviations. The data on mycotoxins and phenolic compounds were statistically processed by mean ± standard deviation (SD) using Microsoft Office Excel 2007 (Microsoft, U.S.A.). The correlation and regression type of analysis were performed in order to examine the quantitative relationship between investigated compounds. Statistical analysis was calculated by using the packages a STAT ENG from software SELEKCIJA (Tarakanovas & Raudonius, 2003).

Mycotoxin concentration in buckwheat groats and hulls
The concentrations of mycotoxins in buckwheat groats and hulls dehulled by thermal and non-thermal treatment are shown in Table 2. All the tested samples were found positive for the investigated toxins; however, the concentrations of OCHA and T-2 mycotoxins in groats in both treatments were low ( Figure 1). The total contamination of groats and hulls with mould fungi was higher in the non-thermal treatment (raw) samples than in those in thermal treatment (steamed).
Aspergillus spp. infection level in steamed groats was four times as high as that in raw groats. Although AFLB1-producing species were detected in the groats samples of both treatments, the AFLB1 concentrations in raw groats were approximately eight-fold higher (2.3 ± 0.6 µg kg −1 ) compared with steamed groats ( Table 2). The concentrations of AFLB1 toxins in baby foods exceeded the allowable limits (Regulation (EC) No.1881/2006, 0.1 µg kg −1 ) several times and reached the highest allowable level for adults (2.0 µg kg −1 ). Since buckwheat is widely used as a dietary product, such groats should not be used for consumption. The contamination of steamed hulls with mould fungi and mycotoxins was low; however, very high concentrations of mycotoxins were determined in the raw hulls samples that had a high mould fungi infection level too. Nearly 100% of raw hulls were contaminated with Penicillium spp. mould fungi (for groats about 10-15%); however, mycotoxin producers Penicillium viridicatum, Aspergillus ochraceus were not detected. Fusarium spp. contamination was minimal, F. sporotrichioides species, producing T-2 toxin (Yli-Mattila, 2010), was prevalent, which might have determined high T-2 concentrations. A high AFLB1 concentration of 75.8 ± 1.6 µg kg −1 was identified in raw hulls. This suggests that hulls play an important role in protecting the grain kernel from mycotoxin contamination. It is likely that inadequate storage conditions of buckwheat raw material increase the likelihood of mould growth and synthesis of mycotoxins. Krysińska-Traczyk, Perkowski and Dutkiewicz (2007) found that the concentration of OCHA in buckwheat grain dust was significantly greater than in grain (p < 0.01) and deoxynivalenol was detected in 79.2% of grain samples and in 100% of grain dust samples. Buckwheat hulls are not used in food industry; however, attention should be drawn to their use as fillers, for example, pillows. Contaminated hulls are recommended to be used as feedstock for fuel production or mulch. Goldberg and Angle (1984) have reported that part of aflatoxins is adsorbed in the soil and biodegraded to water soluble products and CO 2 , therefore the contamination of ground water is highly unlikely. Hulls, including underlying aleurone layer, hilum (navel) and a sizable portion of the germ, are usually more susceptible to mycotoxin contamination compared to other grain parts (Mutungi, Lamuka, Arimi, Gathumbi, & Onyango, 2008). Little is known about the effectiveness of the dehulling process, that industrial manufacturers use, for the decontamination of the buckwheat grains and groats. Analysis of the spelt grain samples revealed that deoxynivalenol concentrations in glumes were five times as high as those in grain. Similar trends were identified for the concentrations of zearalenone and T-2/HT-2 mycotoxins (Jablonskytė-Raščė et al., 2015). Based on the study by Mutungi et al. (2008), aflatoxin levels in whole-grain maize samples with a mean of 97.3 ng/g after dehulling process of the grains significantly decreased (p < 0.001) with a mean value of 57.3 ng/g. The maize byproducts, comprising hulls and fines, had 2-7 times higher levels of aflatoxin than the whole-grain maize. The presence of an unidentified barrier(s) to endosperm contamination with mycotoxins was reported by Abbas and Shier (2009). The observations of mycotoxins in rice hulls and bran show that fumonisin levels in dehulled and milled unpolished rice were very high in hulls (≥17 ppm), low in brown rice (0.9 ppm), moderate in bran (≤4 ppm), but were below the level of detection in polished rice. Llewellyn et al. (1988) indicated that buckwheat and rice hull media inoculated with Fusarium tricinctum yielded trichothecenes (T-2 toxins) in the ppm range, with the buckwheat hull media producing approximately three times more T-2 toxins than the rice hull media. Buckwheat hull media yielded approximately twice the quantity of AFLB1 and AFLG1 than did rice hull media.
Mycotoxins are stable under heat treatment. As a result, treatment with boiling water, roasting or even autoclaving cannot adequately destroy them. Raters and Matissek (2008) have observed that OCHA seems to be stable up to 180°C; however, aflatoxin B1 was almost completely degraded at heating temperatures of 160°C and above. OCH and AFL thermal stability has been widely analysed by Bullerman and Bianchini (2007). According to the authors, OCH is stable during bread baking. Ordinary cooking of rice contaminated with aflatoxin B1 showed an average reduction of 34%. Even further reduction was obtained with pressure cooking (78-88%); boiling corn grits gave an average reduction of aflatoxins of 28%, while frying the boiled grits gave 34-53% total reduction.

Phenolic compounds concentration
Various antifungal compounds, such as phenolics, are used in the prevention of fungi and mycotoxin accumulation in different plant species (Ansari, Anurag, Fatima, & Hameed, 2013;Chitarrini et al., 2014;Samapundo et al., 2007). Buckwheat grain contains a wide range of phenolic compounds. Our previous studies were focused on the quantification of phenolic compounds in buckwheat grain . In the present study, we quantified three groups of phenolic compounds in buckwheat groats and hulls dehulled by thermal and non-thermal treatment: flavonoids (rutin, quercetin), hydroxybenzoic acids (p-hydroxybenzoic, 3,4-dihydroxybenzoic and vanillic) and hydroxycinnamic acids derivate (p-coumaric acid, sinapic and syringic) ( Table 3). Analysis of the phenolic compounds showed that irrespective of the treatment method, phenolic compounds, except for p-hydroxybenzoic acid, were concentrated in hull samples. TPC in raw groats and hulls was 15% and 10% higher than in steamed ones. The effects of various thermal processing methods on the phenolics of Tartary buckwheat were studied by Zhang et al. (2010). The total phenolic and flavonoid content of Tartary buckwheat flour extracts significantly (p < 0.05) decreased by 9% after roasting at 120°C for 40 min and processing by pressured steam-heating at 0.2 MPa for 40 min. However, other authors did not find any statistically significant changes in the total phenolic compound contents after roasting (200°C, 10 min) and extrusion (170°C) of dark buckwheat flour (Sensoy, Rosen, Ho, & Karwe, 2006).
The distribution of individual phenolic compound concentrations in groats and hulls differed between the dehulling treatments. Hulls had twice as high rutin content as groats; however, treatment method did not have any effect on rutin concentration. Very different concentrations were established for quercetin (Table 3). More than 20 times higher concentrations of quercetin (65.47 ± 6.3 µg g −1 ) were determined in steamed hulls compared to other raw and steamed samples. Other researchers reported that thermal treatment of buckwheat groats can affect flavonoid content. Hęś, Dziedzic, Górecka, Drożdżyńska and Gujska (2014) observed that boiled buckwheat groats contained significantly more catechins in comparison to raw buckwheat groats but no change in the rutin content was found. Kreft, Fabjan and Yasumoto (2006) determined that raw groats had~3 times higher rutin content than pre-cooked groats. They indicated that the reduction of rutin in precooked grouts occurred because it might have combined with some other molecules or because of the presence of the rutin degrading enzyme. Yoo, Kim, Yoo, Inglett and Lee (2012) reported that when native buckwheat flour had been mixed with distilled water, the rutin content significantly reduced over time (p < 0.05) and the amount of quercetin increased. However, the rutin content remained constant and quercetin was hardly detected in hydrothermally treated buckwheat flour. It is likely that the activity of rutin-degrading enzymes convert rutin to quercetin and rutinose.
Analysis of phenolic acids showed that 3,4-dihydroxybenzoic acid predominated in groats and hull samples. No differences in phenolic acid content in groats were found between the treatments. Steamed hulls had~20% higher content of p-hydroxybenzoic and 3,4-dihydroxybenzoic acids than raw ones. This resulted in a higher total phenolic acid content in steamed hulls compared with raw ones. p-Coumaric acid had the lowest concentration in all samples. Ferulic, sinapic and syringic acids were detected neither in groats nor hull samples. Vanillic acid was detected at the trace-level concentrations but only in buckwheat hulls.
It was interesting to compare how the concentrations of phenolic compounds changed in buckwheat groats after dehulling compared with intact grain. The samples analysed in this study were prepared from the Lithuania-grown buckwheat grain, phenolic compounds were quantified using the same methods, therefore we were able to compare the results with those of the previous studies . We found that the TPC in raw groats was similar to that of intact buckwheat grain, and in steamed groats phenolic compound content was about 30% lower. The total phenolic acids content in the groats of both treatments was found to be significantly lower than that in intact grain. It is noteworthy that in intact grain there were identified ferulic and sinapic acids which accounted for 14% of the total phenolic acids. The same trend was observed suggesting that vanillic acid is specific to solely hull fraction. It can be assumed that free phenolic compounds are more sensitive to thermal treatment compared with the phenols bound to cell wall. However, it is likely that grain treatment methods affect the diversity of phenolic acids. Contrary to our findings, Zieliński, Michalska, Piskuła and Kozłowska (2006) identified vanillic, ferulic, syringic and p-coumaric acids in buckwheat groats. The hydrothermal processing changed the contents of phenolic acids (except for vanillic acid) in the extruded buckwheat groats. Other authors have indicated that phenolic acids are stable and no changes were found for ferulic or p-coumaric acids after extrusion of brown rice, wheat, maize and barley whole grains (Yang et al., 2014). It is likely that the functionality of active compounds could be maintained by optimization of time and temperature during processing.

Distribution of mycotoxin and phenolic compounds in buckwheat groats and hulls
Since phenolic compounds in buckwheat grain tend to inhibit pathogen growth and mycotoxin production, we compared the concentrations of mycotoxins and phenolic compounds in differently dehulled buckwheat groats and hulls. A trend was revealed that samples containing the highest concentrations of quercetin and hydroxybenzoic acids in groats and hulls were found to be 10-fold less contaminated with AFLB1 and T-2 toxins. A salient difference ratio between phenolic compounds and mycotoxin concentrations in the hull samples was determined. The difference ratio of quercetin and AFLB1 content in raw hulls was 1:27 and that in steamed hulls 34:1 (Figure 2(a)). The difference of the ratio between phenolic compound and mycotoxins concentrations in groats samples was lower; however, the same trend was revealed suggesting that the samples with higher quercetin, p-hydroxybenzoic acid and 3,4-dihydroxybenzoic acid concentrations had lower levels of AFLB1 and T-2 toxins (Figure 2(b)). However, no significant correlation between the investigated mycotoxins and phenolic compounds in raw or steamed groats and hulls was recorded. Thermal buckwheat grain treatment might have had a greater effect on the reduction of mycotoxin and mould contamination. However, protective effect of rutin and vanillic acid against pathogens and toxic metabolites produced by them has been proven (Beekrum, Govinden, Padayachee, & Odhav, 2003;Lattanzio et al., 2006). According to Chitarrini et al. (2014), hulls of Fagopyrum esculentum not proportionally but are susceptible to A. flavus infection and rutin-derived quercetin appears to be more efficient in inhibiting aflatoxin biosynthesis than their parent compound rutin. Samapundo et al. (2007) determined that application of phenolic compounds did not have any effect on growth of the Aspergillus species in corn, but significantly reduced fumonisin B1 and AFLB1 production. As a result, high concentrations of mycotoxins in raw groats and hulls and utilization of phenolic compounds to reduce mycotoxin contamination in buckwheat groats and hulls warrant further research. 38.9 ± 1.9 40.0 ± 3.1 76.9 ± 6.5 67.9 ± 6.6 Mean ± standard deviation (n = 6); n.d.: not detected. Promedio ± desviación estándar (n = 6); n.d.: no detectado.

Conclusions
The research findings suggest that irrespective of the treatment method, phenolic compound contents were significantly higher in buckwheat hulls compared with groats. The total phenolics responded more sensitively to thermal treatment than phenolic acids. The content of quercetin in steamed buckwheat groats and hulls was from 8 to 20 times as high as that in raw groats and hulls. Steamed groats were characterized by lower mycological contamination and their hulls were either free from mycotoxins contamination or the contamination was at a trace level; however, the correlation between the studied mycotoxins and phenolics was statistically insignificant. In order to use raw buckwheat groats for baby food, the grains need to be thoroughly checked for AFLB1 presence (Regulation (EB) No.1881/2006). Because of the high concentrations of AFLB1 and T-2 mycotoxins and mycological contamination, we recommend that raw hulls should be used for human needs with great caution.

Disclosure statement
No potential conflict of interest was reported by the authors.