Reduction of aflatoxins during brewing of a Malawian maize‐based non‐alcoholic beverage, thobwa

Abstract This study reports onthe effect of various production steps on levels of aflatoxins during preparation of thobwa, a traditional maize‐based fermented non‐alcoholic beverage, brewed across Malawi. The effect of boiling, fermentation and their interaction on the level of aflatoxins, the trends of aflatoxin reduction during brewing, and the distribution of aflatoxins between the solid and liquid phases of the beverage were studied using VICAM AflaTest immunoaffinity fluorometric assay. Fermenting and boiling thobwa pre‐mix‐, with initial aflatoxin content of 45–183 μg/kg, resulted in aflatoxin reduction of 47% (13–61 μg/kg) on average. Fermentation and boiling contributed about 20 and 33% aflatoxin reduction, respectively, but without interactive effect between the two factors. Fermenting the thobwa for 24 h led to further reduction of aflatoxins to about 37% of the initial content, and remainedconstant for up to 8 days. Thobwa is a popular beverage in Malawi which is consumed in large quantities by all gender categories including infants, therefore the presence of aflatoxins may constitute a significant health risk factor. This study highlights the need to use raw materials with low levels of aflatoxins for production of maize‐based non‐alcoholic beverages to ensure consumer safety.


| INTRODUC TI ON
Thobwa is a popular traditional fermented non-alcoholic beverage prepared from various grains and their respective malts in Malawi.
In central region of Malawi, thobwa is principally prepared by fermenting a brewed mixture of ground maize and malted maize. It is a light brown, relatively thick opaque gruel with a sweet-acidic taste and is consumed by anyone including children. However, both maize and malted maize are prone to toxigenic fungal colonization and mycotoxin contamination including aflatoxins. Aflatoxins are acutely toxic, carcinogenic, mutagenic, teratogenic, and immunosuppressive, and are classified as group 1 human carcinogens (International Agency for Research on Cancer (IARC), 2002;Mohsenzadeh et al., 2016;Ostry et al., 2017).
The risk of aflatoxin contamination during malting may increase several folds due to increased moisture content especially under uncontrolled environments (Novellie & De Schaepdrijver, 1986;Schwarz et al., 1995;Wolf-Hall, 2007) and high aflatoxin levels of up to 1020 μg/kg have been reported in malted maize in Malawi (Kenji et al., 2000). In central part of Malawi, some households reserve low grade shriveled maize for brewing thobwa and alcoholic beer for celebrations and income. Studies have indicated that alcoholic beer products prepared from low grade maize contain appreciable aflatoxin levels in the range of 90 ± 95 μg/kg (Matumba et al., 2014).
In general, brewing of alcoholic or non-alcoholic beverages may involve malting, milling, mashing, lautering, wort boiling, wort cooling and fermentation followed by aging (lagering) and filtering (finishing) (Catarino & Mendes, 2011;Eblinger & Narzib, 2012;Embashu & Nantanga, 2019). Some of these brewing steps have been found to significantly reduce mycotoxin levels (Ezekiel et al., 2018;Lulamba et al., 2019a;Pascari et al., 2018). However, there is paucity of information on the effect of the traditional brewing steps, used in the production of thobwa, on the on the reduction aflatoxin content in thobwa. Therefore, the present study reports on (i) trends of aflatoxin reduction during brewing (ii) effects of boiling and fermentation and their interaction on content of aflatoxin in thobwa; and (iii) the distribution of aflatoxin between the solid and liquid phases of the studied beverage.

| Maize and malt samples
Six visually moldy dried maize malts (6-10 kg each) were bought from a rural market in Lilongwe district (Malawi). Visually moldy maize spared for brewing fermented beverage was sourced from households in surrounding villages. Both maize and malt were separately ground into powder that passed through sieve # 20 (opening size = 0.85 mm) after which they were thoroughly mixed and used in thobwa preparation.

| Study design
Three experiments were performed to understand the effect of various production steps of thobwa on aflatoxin content. The first experiment was performed to establish the trend of aflatoxin reduction during thobwa brewing. The second study was aimed at determining the effect of boiling, fermentation and their interaction on aflatoxin reduction in thobwa. The third (last) experiment was performed to determine the distribution of aflatoxins between the solid and liquid phases of thobwa.
2.2.1 | Determining the trend of aflatoxin reduction during brewing (experiment # 1) Two sets of experienced thobwa brewers (women) were engaged to prepare thobwa at two different sites using a standardized procedure informed by preliminary trials and local expertise. The brewers were cautioned not to drink the thobwa as it contained high aflatoxin levels. The standardized procedure involved preparing a porridge from 15 liters of water and 4.5 kg of aflatoxin contaminated maize flour and boiling the mixture for 80 min. The porridge was then cooled to about 45°C followed by addition of 1.5 kg of ground malt and vigorous stirring. A sample was then drawn and immediately analyzed for aflatoxin concentration as

| Total aflatoxins determination
Extraction and clean-up of aflatoxin from sample test portions were performed using AflaTest® immuno-affinity fluorometric procedures for popcorn according to the manufacturer's instruction with slight modification (VICAM, 2014). Aflatoxin was extracted from the thobwa and filtrate test portions (50 g test portion +5 g NaCl) using 200 mL of HPLC grade methanol (100%) and blended at high speed for 2 min. The triturated mixture was filtered through Sartorius grade 1289 fluted filter paper to remove particulate matter. A 10 mL volume of the filtered extract was diluted using 40 mL distilled water, thoroughly mixed and filtered through a glass-fiber filter paper of 1.5 μm diameter. A 20 mL (1 g sample equivalent) of the diluted extract was passed through AflaTest® immuno-affinity column at a flow rate of about 1 drop/second. The column was washed twice using (10 mL) distilled water at a rate of 1 drop/second to remove maize intrinsic compounds and finally the aflatoxins were selectively eluted with 1 mL of 100% HPLC grade methanol into a glass cuvette. 1 mL of AflaTest® developer (developer/water of 1 + 9, v/v) was added to eluate in the cuvette and mixed using a vortex and absorbance read using a calibrated VICAM Series-4EX Fluorometer, Milford, MA, USA. The method had limit of detection of 1 μg/kg and was validated using artificially contaminated thobwa at a concentration of 25 μg/kg aflatoxin B 1 (Sigma Aldrich) which yielded recovery of 95% with relative standard deviation (RSD) of 2%.

| Statistical analysis
The differences among means were analyzed by one-way analysis of variance (ANOVA). Results showing significant differences were subjected to post-hoc Tukey's test. The linear regression analysis was performed to study the effects of the boiling, fermentation and their interaction on aflatoxin reduction. The level of confidence required for significance was set at p ≤ .05 for all statistical analyses. These analyses performed using XLSTAT (version 19.01; Addinsoft, New York) and SPSS version 21 (IBM Corp, Armonk, New York).

| Reduction of aflatoxin levels during brewing
The reduction of aflatoxin content in six thobwa samples whose original concentrations ranged from 45 to 183 μg/kg is graphically presented in Figure 2. Aflatoxins were detected at levels of about eighty percent (77 ± 9%) of the initial aflatoxin levels after 17 h of fermentation (Stage B). Boiling the 17-h fermented F I G U R E 1 Thobwa brewing (experiment # 1).

| Effect of boiling and fermentation on aflatoxin
The results of effect of boiling, fermentation and their combination on aflatoxin levels in three thobwa samples are displayed in Figure 3.
Boiling thobwa mixtures prior to fermentation (not a normal procedure, only performed in this experiment to elucidate singular F I G U R E 2 Fate of aflatoxin during brewing and storage of Thobwa. 'A' represents, warm mixture of porridge and malt on the onset of the brewing process; 'B' is 17-h fermented mixture before boiling; 'C' is 17-h fermented mixture (beverage) after boiling for 40 min, 'D' through to 'K' are fermented beverage 1-8 days after boiling, respectively. 'B1' represents beverage samples from brewers set 1 and 'B2' represents beer samples from brewers set 2.

F I G U R E 3
The effect of boiling, fermentation and combination thereof on aflatoxin content in three Thobwa samples.
effect of boiling) caused a mean reduction of 33% in aflatoxin levels.
The linear regression results displayed in Table 2 clearly demonstrated that there was no interaction between boiling and fermenting the thobwa mixtures as regards reduction of aflatoxins as there was no noticeable significant surplus effect of combining the two processes.  can be postulated based on the findings from other studies that fermentation led to irreversible microbial binding/conjugation (Hamad et al., 2017) and/or enzymatic transformations of the aflatoxins thereby making them escape analysis of its parent forms (Adebo et al., 2017;Wang et al., 2018). Likewise, it can be postulated from existing literature that boiling may have also led to hydrolytic opening of the lactone ring in the aflatoxin followed by heat-induced decarboxylation, leading to loss of the methoxy group in the aromatic ring (Farah et al., 1983;Rustom et al., 1993).

| The proportion of aflatoxins present in liquid
Thermal treatments of food containing mycotoxins have been reported to yield metabolites that are as toxic as their parent mycotoxins or even worse (BenáTrivedi, 1993;Dombrink-Kurtzman et al., 2000;Voss & Snook, 2010). However, for aflatoxins, several toxicological studies using cytotoxicity assays on mammalian cell lines or live animal subjects have demonstrated that aqueous boiling and fermentation produce aflatoxin degradation products with reduced toxicity and carcinogenicity (Aiko et al., 2015;Liu et al., 2017).
For instance, aflatoxin B2a, degradation product of bacterial conversion of aflatoxin B1, has shown to exhibit lower DNA-binding capacity than its parent mycotoxin (aflatoxin B1). Therefore, the reduction of the aflatoxin levels observed in the present study suggests that brewing of thobwa had comparatively lowered toxicity than their original mixtures. However, it is noteworthy that in the present experiment, thobwa samples were not completely safe for human consumption since significant proportions (36-46%) of the parent aflatoxin compound were detected. Thus, while brewing may significantly improve the safety of beers, it is necessary to ensure that the raw materials (maize and malts) contain as low aflatoxin as possible to guarantee consumer safety particularly among infants and children.
The stability of aflatoxins during brewing witnessed in this study is of great importance in Malawi considering the common practice of reserving low-grade grains (with high aflatoxin levels) for preparation of thobwa for celebratory and rituals functions by majority of the households. This practice, likely, exposes thobwa consumers to high dietary aflatoxin levels as already demonstrated by earlier studies (Matumba et al., 2014). Under the present investigation, it was observed that fermentation alone was not sufficient to reduce the aflatoxin in the thobwa. Seventeen hours of natural fermentation of the pre-mix-thobwa only degraded 20% of the toxins. Contrarily, 70-100% degradation of aflatoxins was observed when bacteria starters are added (Shu et al., 2018;Wacoo et al., 2019). However, in Malawi, traditional preparation of thobwa does not involve addition of culture starters although it is becoming a popular practice among alcoholic brewers within the area under study. Given the many nutritional benefits associated with probiotics (Nagpal et al., 2012) and the aflatoxin degradation capability of some strains, promoting a culture of probiotic enrichment of thobwa seems plausible.
This study is unique in that it was carefully designed to allow an investigation of the singular effect of boiling pre-mix-thobwa on af- Organization), 1998), consequently not much work has been performed to investigate conditions that may enhance its degradation.
While aflatoxins might indeed be quite stable to dry heating, they seem to be unstable under moist heat and the present findings further reaffirms such phenomenon (Asghar, 2011;Samarajeewa et al., 1990).
The comparatively lower aflatoxin levels detected in filtrate of the beers partly explains the disparity between aflatoxin levels recorded in opaque fermented beverages of Africa and clear beers brewed in industrialized world (Lulamba et al., 2019b). While the absence of aflatoxins in industrial beers could be primarily due the low aflatoxin levels in the raw materials used, the processing effect can- Nonetheless, being the first study of its kind, the current results provide useful insights that would inform the design of future comprehensive studies on fate of aflatoxin during brewing of thobwa.

| CON CLUS ION
While there is substantial reduction of aflatoxins during preparation of thobwa, a significant proportion of aflatoxins may still remain in the final product. These findings highlight the need to use aflatoxin safe raw maize grains and malt during preparation of the beverage if safety of consumers is to be guaranteed (Matumba et al., 2021).
Further, there is need to develop practical strategies to hinder fungal proliferation during malting such as the use of herbs (Abd El-Aziz et al., 2012). The study has provided some insights, however, it necessary to identify and quantify aflatoxin thermal degradation or matrix bound products that arise during the beverage preparation and assess their toxicological safety.

CO N FLI C T O F I NTE R E S T S TATE M E NT
We declare that we have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of the study are available from the corresponding author upon reasonable request.