Occurrence of aflatoxin in agricultural produce from local markets in Burundi and Eastern Democratic Republic of Congo

Abstract Aflatoxins are noxious secondary metabolites, of certain fungal species, found in food and feed. Contamination of a commodity with aflatoxins is associated with production and storage losses, and subsequently less food availability. Aflatoxins can also pose human health risks and represent a barrier to the development of trade, in both domestic and international markets. In this study, samples of cassava, maize, groundnut, beans, soybean, sorghum and milk, and their processed products were collected from local markets in Burundi and Eastern DRC. In order to investigate the levels of aflatoxin, crop samples were analyzed using a single step lateral flow immunochromatographic assay (Reveal Q+), while enzyme‐linked immune‐sorbent assay (ELISA) was used to analyze aflatoxin‐M1 in milk, yogurt, and cheese samples. The results revealed the presence of aflatoxins in all samples from both countries, with levels ranging from 1.3 to 2,410 μg/kg. Samples collected from Burundi contained relatively higher (p > 0.0.5) levels of aflatoxins. In 51% of all the crops samples, total aflatoxin contamination was above the EU maximum tolerable level of 4 μg/kg. Processed products, particularly from groundnut, maize, and sorghum, had the highest levels of aflatoxin contamination when compared to unprocessed grain. With regard to milk and dairy products, the level of aflatoxin‐M1 ranged from 4.8 to 261.1 ng/kg. Approximately 29% of milk and yogurt samples had aflatoxin‐M1 higher than the EU regulatory limit of 50 ng/kg, whereas 20% of cheese samples were found to be contaminated at levels higher than the maximum limit of 250 ng/kg. These results can serve as the basis for pre‐ and postharvest approaches to reduce aflatoxin contamination in agricultural commodities in Burundi and Eastern DRC in order to reduce health risk, avoid reduced production in livestock, and open up export markets.


| INTRODUC TI ON
Nutritional security is effectively achieved when all people at all times consume food of sufficient quantity and quality, in terms of variety, diversity, nutrient content, and safety, to meet their dietary needs and food preferences for an active and a healthy life (FAO/ AGN 2012). Contaminated food is one of the major causes of undernutrition, morbidity, and mortality in sub-Saharan Africa, particularly among children, who are more vulnerable to diseases (Paudyal et al., 2017). Ensuring food safety through the reduction of aflatoxin contamination can contribute significantly to alleviating poverty, increasing food security, and improving nutrition. Also, this has likely positive impacts on enhancing farm productivity, conserving natural resources, as well as improving economic growth by meeting standards in domestic, regional, and international trade.
Among the various mycotoxins, aflatoxins have garnered significant attention due to their negative, and carcinogenic, effects on human and animal health (Klingelhöfer et al., 2018). Although aflatoxins are produced by several Aspergillus species, the major causal agent of contamination globally is A. flavus (Klich, 2007). There are four major aflatoxins, including B 1 , B 2 , G 1 , and G 2 ; however, aflatoxin-B 1 is the most toxic and prevalent and is classified as a Group 1A carcinogen by the International Agency for Research on Cancer (IARC 2002). High-dose exposure to aflatoxins concentrations can cause acute health effects such as vomiting, abdominal pain, and even possible death (Probst, Njapau, & Cotty, 2007;Sherif, Salama, & Abdel-Wahhab, 2009), while sublethal chronic exposure may lead to liver cancer, stunting in children, and immune system suppression (Chan-Hon-Tong, Charles, Forhan, Heude, & Sirot, 2013;Wu & Khlangwiset, 2010). In 1981, for instance, the outbreak of aflatoxicosis as a result of ingestion of maize contaminated with 3.2-12 mg/kg of aflatoxin-B 1 caused fatalities in Kenya (Obura, 2013). In another severe aflatoxicosis outbreak, Azziz-Baumgartner et al. (2005) also reported that aflatoxin contamination was found to be the cause of over 125 deaths during [2004][2005] in Eastern province of Kenya. Williams et al. (2004) estimated that over 5 billion people living in low-income countries are at risk of chronic exposure to aflatoxins.
The incidence of aflatoxin contamination in major food crops such as maize, groundnut, sorghum, tree nuts, and dried fruits and spices as well as milk and meat products is widespread in warm climates (CAST 2003;Chala et al. 2014;Mutegi, Ngugi, Hendriks, & Jones, 2009;Perrone et al., 2014;Williams et al., 2004). In animals, aflatoxins may lower resistance to diseases, interrupt vaccineinduced immunity, and adversely affect growth and reproduction, causing serious economic losses (CAST 2003;Fink-Gremmels, 1999).
Infection and production of aflatoxins by ubiquitous, air-borne, and soil-inhabiting species of fungi begin at preharvest stages and may continue to increase until the grain is consumed (Waliyar, Ntare, Diallo, Kodio, & Diarra, 2007;Waliyar et al., 2015).
The interplay between the safety of food and the adequacy of food is therefore crucial when addressing the aflatoxins problem in low-income countries. An earlier study by Brudzynski, Van Pee, and Kornazewski (1977), for example, showed the presence of aflatoxins up to 1,000 μg/kg in maize and groundnut from the DRC. Recently, Kamika and Takoy (2011) reported that 95% of groundnut samples collected during the dry and the rainy seasons in Kinshasa contained aflatoxin-B 1 over the maximum limit of 2 μg/kg prescribed by the EU as the standard for direct human consumption. These studies were, however, limited to maize and groundnut conducted only in the DRC.
To expand insight, we conducted a comprehensive investigation on the incidence of aflatoxin contamination in raw and processed materials from cassava, maize, sorghum, beans, soybean, groundnut, and milk in the local markets of Burundi and Eastern DRC. The main food crops in Gitega province and Kabare district include cereal crops such as maize and legumes, while in Chibitoke province and Uvira district the main crops are cassava and legumes.

| Sample collection
For each sample, 1 kg of the commodity was bought and collected from different parts of the seller's container and thoroughly mixed, while 1 L of milk and yoghurt was purchased as sellers prepared in plastic bottle. The sellers were randomly selected from each market, and the type of samples was collected depending on available samples from each seller. After collection, samples were labeled with the name of village and collection date and then subdivided into three portions. The first portion was kept as a backup, while the second was directly used for moisture analysis. The third was examined to determine the level of aflatoxin contamination. All samples were sealed in polyethylene plastic bags under normal atmospheric conditions, whereas the milk and dairy products were kept in plastic bottles. The package seal was carefully inspected to avoid any possibility of leakage. Subsequently, the sealed packages were stored at a temperature of 4°C for dried samples and −4°C for milk and dairy products about 2 weeks, without direct sunlight until further analysis.

| Moisture content
Moisture content (MC) of each sample was determined by drying the samples in the hot air oven at 105°C for 12 hr, following technique 950.46 (AOAC 2006). The tests were conducted in triplicates, and the moisture content was calculated using the following formula: [(original weight of sample -weight of sample after drying)/original weight of sample] *100. To analyze the aflatoxins concentration, a Reveal Q+ test kit (Neogen Corporation, USA) was used as a single step lateral flow immunochromatographic assay based on a competitive immunoassay format. A total of 500 μl diluent was mixed with 100 μl of the sample filtrate and then carefully mixed by pipetting up and down five times in a dilution cup. A 100 μl portion of the mixture was transferred to a new clear sample cup. Subsequently, a Reveal Q+ for aflatoxin test strip was placed into the sample cup for 6 min; the strip was removed and inserted to the AccuScan ® reader (AccuScan Pro, model AX-2; Neogen Corporation, Australia). Aflatoxin concentration was displayed in parts per billion (ppb). All samples were analyzed in duplicate from a separate 10 g measure.

| Analysis of aflatoxins in milk and dairy products
To determine aflatoxin-M 1 in milk and yogurt, the method developed by Gizachew, Szonyi, Tegegne, Hanson, and Grace (2016) was adapted, while the method of Škrbić, Antić, and Živančev (2015) was modified to determine aflatoxin-M 1 in cheese products. One hundred milliliter of milk and yogurt samples was warmed to 37°C in a water bath and then centrifuged at 10°C with 3500 g for 10 min (model TDL-5-A; Lab companion, Korea). After discarding the upper cream layer, the remaining skimmed milk was filtered through Whatman No. 4 filter paper before aflatoxin-M 1 analysis. For the cheese products, 2 g of homogenized samples were weighed and blended with 40 ml dichloromethane for 15 min. The filtrates were evaporated via rotary evaporator (model R-II; Büchi, Postfach, Switzerland) at 60°C.
A solution of 0.5 ml methanol, 0.5 ml phosphate buffer saline (PBS), and 1 ml hexane was added to the residue and then centrifuged at 15°C with 2,700 g for 15 min. The lower methanolic phase was collected. Prior aflatoxin-M 1 determination, 100 μl of this methanolic phase was diluted with PBS to achieve a dilution of 1:5.
Assay procedure was followed according to the protocol pro- Samples that were beyond the range of the highest standard concentration were diluted, and the ELISA experiments were repeated.

| Total aflatoxin and aflatoxin-M 1 validation
To test the sensitivity of the method, the total aflatoxin standard solution at two different concentrations was added to the all samples. The extraction and the recovery of the spiked samples were performed as previously described, in duplicate. The validation of Reveal Q+ and ELISA methods was carried out with the determination of the recoveries and the coefficient of variation (%CV) as presented in Table 1.

| Moisture content of samples
The MC of grain samples collected from local markets in Burundi and Eastern DRC is shown in Table 2. There was no significant difference in mean MC of the samples from the two countries. In the market, grain samples were mostly kept in open containers, while processed samples were stored in plastic closed containers or paper bag. Only traditionally fermented cassava foods (ubuswage) were wrapped in plantain leaves. Overall, the MC ranged between 6.7% and 15.0% for grain samples and between 5.5% and 13.6% for flour, with the lowest being recorded in groundnut flour and the highest in cassava flour.
Much higher MC content was recorded in the cassava prepared for ubuswage, with an average MC of 59.8%. In addition, the MC of fresh milk, yogurt, and cheese ranged between 79.3% and 89.3%.

| Occurrence of aflatoxins in crop samples
The occurrence and concentration of total aflatoxins in crop samples collected from Burundi and Eastern DRC are summarized in Tables 3 and 4. All the 218 samples were contaminated with aflatoxins, which ranged from 1.3 to 2,410.0 μg/kg. Nowadays, the EU has set the strictest standards, such that any products for direct human consumption can only be marketed with concentrations of aflatoxin-B1 and total aflatoxins not >2 and 4 mg/kg, respectively (EC, 2007(EC, , 2010. Likewise, US regulations have specified the maximum acceptable limit for total aflatoxins at 20 mg/kg (Wu, 2006). In India, a tolerance limit of 30 mg/kg for aflatoxins in all foods has been defined. Kenya adopted a maximum allowed level of 10 mg/kg of aflatoxin-B1 in groundnuts and several grain foods. Brazil has fixed the limit of total aflatoxins in nuts at 30 mg/kg (Freitas-Silva & Venâncio, 2011). As Burundi and DRC do not have regulations for aflatoxins, in this study, we applied the EU standard as the strictest standards to compare for all crop samples.
About 60% of these samples contained aflatoxins above the EU maximum permissible limit (4 μg/kg) for total aflatoxins in maize  is relatively hotter and drier, a situation that favors the growth of mycotoxin-producing fungi. Further details of the incidence of aflatoxin contamination in specific crops are also presented below.
More than 88% of the samples met the EU regulatory threshold for aflatoxin of 4.0 μg/kg. All the samples met the proposed East Africa regulatory threshold of 10 μg/kg. Similar observations regarding the low contamination by aflatoxins in cassava are reported in Ghana (Wareing, Westby, Gibbs, Allotey, & Halm, 2001), Republic of Benin (Adjovi et al., 2014;Gnonlonfin et al., 2012) and Tanzania (Sulyok et al., 2015) The occurrence of aflatoxins in cassava chips from Cameroon was only detected after 4 weeks' storage (Essono et al., 2009

| Maize
Among the grain samples, the high concentrations of total aflatoxins were obviously detected in maize, followed by groundnut, sorghum, beans, and soybean, respectively (Table 3)

| Sorghum
In this study, all sorghum samples in grain, flour, and germé forms contained detectable concentrations of aflatoxins, ranging Notes. Value is the mean ± SD. a Ubuswage is the traditional cassava product in Central African region. b Germe is the germinated sorghum for beer processing.  Ubuswage is the traditional cassava product in Central African region. c Germe is the germinated sorghum for beer processing. between 2.5 and 490.0 μg/kg. Additionally, total aflatoxins exceeded the regulatory levels for direct human consumption as set by the EU in 84.6% of the sorghum samples. The levels of aflatoxin contaminations may also be associated with the poor pre-and postharvest practices as well as processing methods.

TA B L E 3 Distribution of total aflatoxins in dried food samples found on local markets in Burundi and Eastern DRC
Sorghum, in particular, is used as a malted grain (germé) in beer production in Burundi. The traditional processing technique, which involves the use of Enterobacteruaceae and molds, may cause aflatoxin contamination in germé (Bationo et al., 2015).

| Beans
Aflatoxin was present in 100% of bean samples from Burundi and Eastern DRC and ranged from 1.9 to 6.6 μg/kg. This low level of aflatoxin contamination in the bean samples is perhaps due to the ability of phenolic compounds, particularly gallic and chlorogenic acids, to inhibit fungal amylase activities (Telles, Kupski, & Furlong, 2017). Pagnussatt, Bretanha, Sílvia, Garda-Buffon, and Badiale-Furlong (2013) also mentioned that the synergistic effect of different compounds in beans can contribute to a defense barrier against development of toxigenic species. Literature reports a few instances of aflatoxins in red kidney beans, split peas, chickpea, and cowpea such as in Pakistan (Lutfullah & Hussain, 2012).

| Soybean
All soybean samples analyzed were positive for total aflatoxins with 40.0% of these samples exceeding 4 μg/kg. The highest concentration of aflatoxins was found in flour than in dried grains. It has been reported that aflatoxin contaminations in soybean are relatively low, but there are conflicting explanations to the possible cause of low aflatoxin contamination in soybean. One of the initial studies associated this phenomenon to the zinc binding ability of phytate in soybean, as it is an important intermediate substrate of aflatoxin biosynthesis (Gupta & Venkitasubramanian, 1975). However, Ehrlich and Ciegler (1985) showed that phytate level does not influence aflatoxin biosynthesis. Burow, Nesbitt, Dunlap, and Keller (1997) hypothesized that lipoxygenase in soybean can produce hydroxyl fatty acids which are capable of inhibiting aflatoxin production in A. parasiticus. With regard to aflatoxin inhibition, Mellon and Cotty (2002) reported that soybean grains with lipoxygenase might not deter increased seed pathogen susceptibility, but seed coat integrity and seed viability may play more determinant role in seed resistance to aflatoxin contamination. There is hence the need for further understanding of the possible cause of low aflatoxin contamination in soybean.

| Groundnut
In this study, total aflatoxins concentration in groundnut products from the local markets in Burundi and Eastern DRC ranged from 2.2 to 2,410.0 μg/kg. The highest contamination level was found in groundnut flour (2,410 μg/kg), followed with roasted groundnut (1,080 μg/kg) and dried kernels (29.3 μg/kg), respectively. About 69.4% of the groundnut samples exceeded the EU aflatoxin regulatory limits. None of the groundnut flour samples were fit for human consumption according to any existing regulation globally, with some samples surpassing the EU maximum permissible limit of 4 μg/kg by 600-fold. Aflatoxins were found more in processed groundnut than in unprocessed dried grains (Tables 2 and 3). Processed groundnut, often prepared from low quality groundnut, can be exposed to a wide range of environmental conditions, such as high temperature and humidity as well as to oxygen and mold, which can trigger further increase in aflatoxin contamination. Nonetheless, other factors including biological, nutritional, and climatic factors can be responsible for aflatoxins contamination, especially in groundnut and maize, some of which are either difficult or impracticable to control. Groundnut is a preferred substrate for aflatoxinproducing fungi (Bankole, Schollenberger, & Drochner, 2006;Ezekiel et al., 2013;Monyo et al., 2012). The range of aflatoxin contamination in groundnut samples in this study was comparable to those reported from local vendors, markets, and retail shops in Nigeria where aflatoxin-B 1 detected in 64.2% of dry roasted groundnut (Bankole, Ogunsanwo, & Eseigbe, 2005).

| Occurrence of aflatoxin-M 1 in milk and dairy products
Milk and dairy products are important for growth and development as well as maintenance of good health in humans, especially babies and children. The occurrence of aflatoxin-M 1 in milk and its products collected in Burundi and Eastern DRC is presented in Tables 5 and 6. According to the EU regulations, the maximum residue level of aflatoxin-M 1 in raw milk and dairy products is 50 ng/L, while this level based on USA regulations was adjusted to 500 ng/kg (Campagnollo et al., 2016;Iqbal et al., 2015;Mulunda & Mike, 2014). Aflatoxin-M 1 was detected in all samples collected for this study, with concentrations ranging between 4.8 and 261.1 ng/kg. Among the 13 fresh milk samples analyzed, 4 (30.8%) contained aflatoxin-M 1 above the maximum permissible limit of 50 ng/kg, as set by the EU for raw milk, heat-treated milk, and milk for the manufacture of milk-based products (EC 2006).
Of the eight yogurt samples, only two samples (25%) were contaminated with aflatoxin-M 1 above the limit of 50 ng/kg, with the concentration ranging between 4.8 and 63.2 ng/kg. Brackett and Marth (1982) explained that the changes in casein structure due to fermentation process may cause adsorption or occlusion of toxins, including aflatoxin-M 1 , in the precipitate. Montaseri et al. (2014) also referred to this behavior as the possible reason why LAB is capable of removing aflatoxin-M 1 from yogurt. Furthermore, the low concentration of aflatoxin-M 1 in yogurt might be associated with processing variables such as pH, formation of organic acids, or other fermented by-products (Govaris, Roussi, Koidis, & Botsoglou, 2002).  The EU permissible level and the WHO advisory level for total AFs are 4 and 10 μg/kg, respectively, for foods intended for direct human consumption. b The first integer is the number, and the integer in parenthesis is the percent of samples containing a specified level of aflatoxins. c Ubuswage is the traditional cassava product in Central African region. d Germe is the germinated sorghum for beer processing.

| CON CLUS IONS
This first report on the incidence of aflatoxin contamination in agricultural products from local markets in Burundi and Eastern DRC showed that of the 244 crops, milk, and their processed products sampled, the percentage of aflatoxin positive samples was 100%. In addition, 50.9% of crop, 28.6% of milk and yogurt, and 20.0% of cheese samples had aflatoxin concentrations higher than the regulatory limits set by the EU. The processed samples presented higher aflatoxin contamination when compared to unprocessed samples. Therefore, the presence of aflatoxin in local food products from Burundi and Eastern DRC is a problem in the context of food sufficiency, public health, and economic benefits.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interests.

E TH I C A L S TATEM ENT
This study does not involve any human or animal testing.