Food Additives and Contaminants, February 2008; 25(2): 133

Foreword
On the 44th year of its establishment, The Scientific and Technological Council of Turkey (TUBITAK), with support from the Ankara Test and Analysis Laboratory (ATAL), hosted the XIIth International IUPAC [International Union of Pure and Applied Chemistry] Symposium on Mycotoxins and Phycotoxins in the Askeri Museum, Istanbul, from 21 to 25 May 2007. The IUPAC symposia have become the principal interdisciplinary meeting on mycotoxins and phycotoxins facilitating an opportunity to gain an overview of research and development in analytical chemistry, risk assessment, effects on human health, control and remediation strategies. More than 580 participants from 65 countries attended the symposium. They came from universities, research institutes, governmental establishments, and the private sector. They were all involved with mycotoxins and phycotoxins from the perspective of human and animal health, toxicology, food chemistry, food engineering, and risk analysis. The keynote lectures by Professor Timothy Phillips, Reducing human exposure to aflatoxins through use of clays, and Dr Benjamn A. SuarezIsla, The need for new functional and analytical methods for marine biotoxins, set an excellent tone and inspiration for the entire symposium. Professors Phillips and Suarez-Isla provided the audience with an insight into cutting-edge developments in both the mycotoxin and phycotoxin areas. The scientific programme of the symposium comprised twelve main sessions with presentations from 27 invited speakers who are well known worldwide and have renowned expertise in their fields. There were 105 oral presentations and 300 poster presentations selected by the symposiums Scientific Committee, an exhibition of analytical instruments, and displays from food manufacturers. Moreover, workshops, satellite meetings, and parallel meetings hosted by international and European companies were held concurrently. Food and feed are essential elements for human and animal life. Thus, for a healthy life, ideally foods should be free from pathogenic microorganisms and harmful biocontaminants as mycotoxins and phycotoxins. A wide array of aspects

of these toxins were topics of study in the XIIth IUPAC Symposium on Mycotoxins and Phycotoxins. The symposium sessions covered many mycotoxins issues including mycotoxins and human health, analytical techniques for mycotoxins, animal feed and dairy products, dried fruits, spices, botanicals and derived products, nuts, cereals and cereal-based products, coffee, cocoa and derived products, risk assessment, regulations and international trade, advances in toxicology, and some phycotoxin issues as well as microcystins and other freshwater toxins, lipophylic shellfish toxins, paralytic shellfish poisoning toxins, amnesic shellfish poisoning toxins and other marine toxins. In addition, the 3rd Turkish National Mycotoxin Symposium was held as a separate session. Every meeting provides a renewal of the feeling of dynamics of the field. An area so critical to many industries, and practised by so many dedicated and passionate researchers and analysts, will continue to produce novel developments that warrant the attention given to this field at the IUPAC symposia. Our gratitude goes to all the members of the local Organizing Committee and the Scientific Committee who invaluably contributed to making this symposium such a success, in terms of overall experience of the meeting, programme content, and social programme. Additionally, we would like to thank the directorate of TUBITAK-ATAL for all their support. Their experience and dedication during the preparation process and the symposium was an integral contribution to the achievements of this meeting. We have received many letters of appreciation for organizing the Symposium and we are thankful that the high expectations in scientific terms and networking opportunities were achieved. This special issue of Food Additives and Contaminants is devoted to a selection of papers from invited speakers of the XIIth International IUPAC Symposium on Mycotoxins and Phycotoxins. We are grateful to them for sharing their knowledge to a broad audience. Hamide Z. Senyuva & Hans P. van Egmond

ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701788865

Food Additives and Contaminants, February 2008; 25(2): 134145

Reducing human exposure to aflatoxin through the use of clay: A review

T. D. PHILLIPS1, E. AFRIYIE-GYAWU1, J. WILLIAMS2, H. HUEBNER1, N.-A. ANKRAH3, D. OFORI-ADJEI3, P. JOLLY4, N. JOHNSON1, J. TAYLOR1, A. MARROQUIN-CARDONA1, L. XU5, L. TANG5, & J.-S. WANG5
Department of VIBS-MS 4458, Texas A&M University, Veterinary Integrative Biosciences, College of Veterinary Medicine, College Station, TX 77843, USA, 2Peanut Collaborative Research Support Program, University of Georgia, Griffin, GA, USA, 3School of Public Health, Noguchi Memorial Institute for Medical Research, University of Ghana, Legon, Accra, Ghana, 4Epidemiology, University of Alabama at Birmingham, Birmingham, AL, USA, and 5Institute of Environmental/Human Health, Texas Tech University, Box 41163, TTU/TIEHH, Lubbock, TX 79409-1163, USA
(Received 18 May 2007; accepted 10 July 2007)
1

Abstract Innovative sorption strategies for the detoxification of aflatoxins have been developed. NovaSil clay (NS) has been shown to prevent aflatoxicosis in a variety of animals when included in their diet. Results have shown that NS clay binds aflatoxins with high affinity and high capacity in the gastrointestinal tract, resulting in a notable reduction in the bioavailability of these toxins without interfering with the utilization of vitamins and other micronutrients. This strategy is being evaluated as a potential remedy for acute aflatoxicosis, and as a sustainable human intervention for aflatoxins via the diet. Phase I and II clinical trials confirmed the apparent safety of NS for further study in humans. A recent study in Ghanaians at high risk for aflatoxicosis has indicated that NS (at a dose level of 0.25%) is effective in decreasing biomarkers of aflatoxin exposure and does not interfere with the levels of serum vitamins A and E, and iron and zinc. In summary, enterosorption strategies/ therapies based on NS clay are promising for the management of aflatoxins and as a sustainable public health intervention. The NS clay remedy is novel, inexpensive and easily disseminated. Based on the present research, aflatoxin sequestering clays should be rigorously evaluated in vitro and in vivo, and should meet the following criteria: (1) favourable thermodynamic characteristics of mycotoxin sorption, (2) tolerable levels of priority metals, dioxins/furans and other hazardous contaminants, (3) safety and efficacy in multiple animal species, (4) safety and efficacy in long-term studies, and (5) negligible interactions with vitamins, iron and zinc and other micronutrients.

Keywords: NovaSil clay, mycotoxins, aflatoxins, aflatoxin-binding agent, aflatoxin sorbent, aflatoxin-sequestering agent, clinical trial, Ghana

Introduction Historical perspective on moulds Moulds (and their metabolic by-products) can be beneficial, as well as harmful, to humans and animals. Moulds have been used since ancient times in the production of various foods including cheese and salami and in the fermentation of beer and wine (Peraica et al. 1999). The secondary metabolites from these same types of moulds have been used as very effective antibiotics for the treatment of disease and as drugs for other important
Correspondence: T. D. Phillips. E-mail: tphillips@cvm.tamu.edu ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701567467

medicinal purposes. For example, the Chinese used moulds for obstetrical purposes nearly 5000 years ago (Hesseltine 1979). Unlike bacterial toxins, the mycotoxins are not protein in nature, but consist of highly diverse organic structures characterized by a variety of heteroatom-containing functional groups. Many of these potent organic chemicals, though invisible to the naked eye, can be found in mouldcontaminated food sources and may be harmful if ingested in high enough quantities or over a long enough period of time. Fortunately, the toxicity of these compounds is dose related, and their levels in

Clay-based remedy for aflatoxins foods and feeds are typically low to non-detectable. However, during extended periods of drought the production of certain hazardous mycotoxins can be unavoidable and may result in contaminated food and feed products that present significant health risks to man and animals (Phillips et al. 2002, 2006; Huebner et al. 2004; Williams et al. 2004).

135

Mouldy food poisoning and human disease Although the term mycotoxin was not commonly used until the mid-20th century, earlier records suggest that mycotoxin contamination of food and major outbreaks of disease associated with the consumption of mouldy food have occurred frequently throughout history. A variety of toxic effects of moulds can be traced to very early civilizations, including the Chinese almost 5000 years ago (Ramsbottom 1953; Van Rensburg and Altenkirk 1974). Of the more than 300 mycotoxins that have been identified and chemically characterized, many have been found as contaminants of food and have been linked to the aetiology of disease in humans (and animals). Of these, the aflatoxins have been extensively studied due to their frequent occurrence in foods (especially in developing countries) and their mutagenicity and carcinogenicity (Wogan 1992; Wild and Hall 2000; Wild and Turner 2002).
Figure 1. Molecular model of aflatoxin B1 showing the spatial arrangement of atoms. The molecule is planar, except for the terminal furan which is kinked in the cis configuration (coming out of the page). White, hydrogen; dark grey, oxygen; grey, carbon.

Aflatoxins Discovery There was a lack of understanding of the consequences of aflatoxin exposure on human and animal health, until the early 1960s, when mouldy feed was associated with the loss of thousands of young turkeys in the UK. In this incident, the affected animals showed signs of severe liver necrosis as well as fatty degeneration, fibrosis and extensive bile-duct hyperplasia (Siller and Ostler 1961). Upon investigation it was discovered that the Turkeys had been fed Brazilian peanut meal containing the mould Aspergillus flavus along with four metabolic by-products, namely aflatoxins B1 (AfB1) (Figure 1), B2 (AfB2), G1 (AfG1) and G2 (AfG2) (Asao et al. 1963). For confirmation the same symptoms were produced in a variety of other species, including ducklings (Sargeant et al. 1961) and rats (Lancaster et al. 1961) following ingestion of the contaminated peanut meal. The dramatic effects of the aflatoxins resulted in significant scientific interest in delineating the chemical structures and toxicological properties of aflatoxins (and other mycotoxins).

Toxic effects The first observations regarding the toxicity of aflatoxins were recorded during the Turkey X incident in the UK in 1960. In these birds, acute necrosis and bile duct proliferation of the liver was observed (Lancaster et al. 1961). Soon after the same types of symptoms were reported in ducklings and pheasants (Sargeant et al. 1961). A number of acute dosing studies have been performed with AfB1 to determine LD50 values for a wide range of animals (Council for Agricultural Science and Technology (CAST) 1989). While AfB1 was found to be toxic for many species, some animals were identified as highly sensitive, including duckling, rabbit, and rainbow trout (Muller et al. 1970). Other common symptoms of acute poisoning by AfB1 include depression and anorexia, as seen in the recent contamination of pet food in South Texas (Garland and Reagor 2001) and South Carolina (Lang 2006). Chronic toxicity studies have also been conducted with lower levels

136

T. D. Phillips et al.

of aflatoxin exposure. One of the major effects is a general reduction in weight gain for a variety of production animals, including pigs, cattle and poultry. Also, milk production in dairy animals can be decreased in the presence of aflatoxincontaminated food (CAST 1989) and the milk carries a metabolite of known as AfM1. This chemical is highly regulated because infants and young children may consume large quantities of milk, and the young of all species are more susceptible (than adults) to the effects of aflatoxins (Leeson et al. 1995). The main effect of chronic exposure to aflatoxin in humans is hepatocellular carcinoma (HCC). While the incidence of this disease is low in the USA, in parts of Africa and Asia these numbers can be very high (Groopman et al. 1988). Epidemiological studies have shown a positive correlation between the intake of aflatoxin and liver cancer among African and Asian populations. For example, a study by Bulatao-Jayme et al. (1982) showed that the relative risk for liver cancer, when consuming a large level of aflatoxin with a minor amount of alcohol, was 17.5 as compared with a relative risk of 3.9 when alcohol consumption was heavy and aflatoxin intake was light. A diet dependent on foods such as cassava, peanuts, sweet potato and corn increases the likelihood of consuming mouldinfested food. A serious confounder in these studies was the high rate of hepatitis B virus (HBV) in these populations. The potential interaction of HBV with aflatoxin is not fully understood (Groopman et al. 1988). The carcinogenic potency of AFB1 in individuals positive for hepatitis B virus (HBV) surface antigen (HBsAg) has been reported to be considerably higher compared with individuals who are negative for HBsAg. Most of the epidemiological data are from geographical areas where both the prevalence of HbsAg individuals and aflatoxins are high; the relationship between these risk factors in areas of low aflatoxin contamination and low HBV prevalence is unclear (World Health Organization (WHO) 1998). Biochemical mode of action AfB1 is a direct-acting mutagen and identification of a DNA adduct was made by Essigmann et al. (1977). It was shown that the 8,9 vinyl ether group is transformed to an epoxide through cytochrome P450-mediated oxidation; carbon 8 on aflatoxin reacts with guanine at the N7 position to form an 8,9-dihydro-8-(N7-guanyl)-9-hydroxy-AfB1 adduct (Figure 2). Studies have utilized nuclear magnetic resonance (NMR) to characterize the intercalation and adduct formation of AfB1 with two oligodeoxynucleotide sequences

Figure 2. Molecular model showing AfB1 (in medium grey) intercalated between the strands of DNA.

and d(ATAFB [d(ATCAFBGAT).d(ATCGAT) G-CAT)2] (Gopalakrishnan et al. 1990). Importantly, AfB1 has been shown to disrupt several genes involved in the growth of cancer. When administered to rats, it was demonstrated that the resulting liver tumours contained an activated form of a proto-oncogene of the c-Ki-ras family. In addition, the hepatocellular tumours of AfB1-treated rats showed high expression of the proto-oncogenes c-H-ras and c-myc (McMahon et al. 1986, 1987). Proto-oncogenes stimulate growth in the normal cell; when mutated, loss of function leads to rapid growth characteristic of cancer. Tumour suppressor genes provide a complement to this system, keeping growth in check in the normal cell. AfB1 has been shown to produce a G ! T transversion in the p53 tumour suppressor gene in human hepatocytes (Aguilar et al. 1993). In addition to DNA and RNA damage, AfB1 has also been shown to interact with RNA and intercellular proteins. Interactions with protein (Sabbioni et al. 1987; Guengerich et al. 2002a, 2002b) may explain some of the noncarcinogenic effects following exposure to aflatoxin (Eaton et al. 1994). Consequences of exposure AfB1 has been described as a human carcinogen (Group 1A) and implicated in HCC (International Agency for Research on Cancer (IARC) 1976, 1987, 1993). Also, studies suggest that AFs impair the cellular and humoral immune system in animals, and that low-level exposure to these toxins can cause immunosuppression and increased susceptibility to

Clay-based remedy for aflatoxins disease (Rodricks and Stoloff 1977; Miller et al. 1978; Richard et al. 1978; Peska and Bondy 1994; Hinton et al. 2003). A study by Turner et al. (2003) in Gambian children showed evidence that secretory IgA in saliva may be reduced from dietary aflatoxin exposure. This was the first report of immunosuppression in humans associated with AF biomarker measures. Jiang et al. (2005) recently confirmed this finding in adult humans in Ghana and showed a significant correlation between aflatoxin exposure and suppression of the immune system. Other consequences of dietary exposure to aflatoxins include adverse effects on growth and antinutritional effects in animals. For example, AFB1 has been shown to reduce hepatic vitamin A significantly in a variety of animals, including chickens (CAST 1989; Pimpukdee et al. 2004; Williams et al. 2004). Importantly, Gong et al. (2002, 2004) and Turner et al. (2007) reported an association between biomarkers of aflatoxin exposure and growth impairment in children in West Africa. Although many countries have regulatory limits for aflatoxins in foods/feeds, outbreaks of poisoning frequently occur. A recent outbreak of aflatoxin poisoning in Kenya resulted in a 39% case fatality rate and was linked to consumption of foods containing toxin levels as high as 8000 ng g1 (Centers for Disease Control and Prevention (CDC) 2004). Drought stress exacerbates fungal infection, thus enhancing production of the aflatoxins. This is especially true between a latitude of 40 N and 40 S of the equator, a hot zone which encompasses many developing countries where aflatoxins in the diet of humans and animals are largely uncontrolled (Williams et al. 2004). The poorest people who are most likely to consume foods contaminated with aflatoxins suffer the most severe effects, including disease and even death following acute exposure (Lewis et al. 2005). Additionally, it is estimated that 80% of all HCC cases occur in developing countries (Wild and Hall 2000). Thus, feasible interventions and therapies to diminish human and animal exposure to aflatoxins are imperative; dietary calcium montmorillonite clay, used as an aflatoxin enterosorbent, may provide a practical, cost-effective, and sustainable solution to the problem.

137

Intervention approaches Chemopreventive agents Because avoiding consumption of aflatoxincontaminated foods for many is simply not feasible, effective means for reducing dietary exposure to aflatoxins are highly desirable (Phillips et al. 2006). Chemoprevention is one strategy used to solve the

problem in high-risk populations. This involves the use of natural or synthetic agents to block, retard, reverse or modulate the carcinogenic process (Gupta and DuBois 2001; Sporn and Suh 2002). A variety of chemopreventative agents exist as natural constituents in the human diet; many of these include phytochemicals derived from various sources. Although efficacious against a wide range of carcinogens, most of these compounds occur at very low levels in a nutritionally balanced diet and are poorly absorbed in the gastrointestinal tract (Hayatsu et al. 1988; Dragsted et al. 1993). The chemopreventative agent oltipraz, an antischistosomal drug, has been evaluated for use in humans exposed to dietary aflatoxins in China (Kensler et al. 1999; Wang et al. 1999). In clinical trials oltipraz, when administered to individuals exposed to dietary aflatoxins, increased the level of glutathione S-transferasemediated conjugation of aflatoxin 8,9-epoxide, but also inhibited cytochrome P450 1A2 activity, a key enzyme that activates aflatoxin to the reactive epoxide (Wang et al. 1996, 1999; Kensler et al. 1999). Oltipraz may also inhibit hepatitis B virus (HBV) transcription through elevation of p53 providing an additional contribution to HCC chemoprevention (Chi et al. 1998). Chlorophyllins are natural occurring constituents of the human diet that have been shown to be effective anticarcinogens in several animal models (Dashwood et al. 1998). They are hypothesized to act as interceptor molecules by binding with carcinogens, such as AfB1, thereby diminishing bioavailability by impeding their absorption. (Breinholt et al. 1995). In a 4-month clinical trial in China, consumption of 100 mg of chlorophyllin at each meal led to an overall 55% reduction in median urinary levels of aflatoxinN7-guanine adducts versus the placebo (Egner et al. 2001). Application of these compounds in humans would require careful evaluation including long-term effects of enzyme modulation and potential interferences with the uptake of essential nutrients from the diet. Green tea-derived polyphenols, which are highly effective agents against cancer in various animal models, are also being considered as possible interventions for populations at high risk for HCC. Research has indicated that green tea inhibits the initiation of AfB1-induced hepatocarcinogenesis in the rat by modulating metabolism of AfB1 (Qin et al. 1997). Administration of green tea (3% in water) prevented hepatic focal lesion growth induced by dieldrin in B6C3F1 mice (Klaunig and Kamendulis 1999). In humans, inverse associations between the level of green-tea consumption and the risk of development and/or time of cancer onset have also been observed (Nakachi et al. 2000; Fujiki et al. 2002).

138

T. D. Phillips et al. a calcium montmorillonite, can prevent the adverse effects of exposure to dietary aflatoxins. Initially, NS (which was referred to as HSCAS in the early literature), was sold as an anticaking additive for animal feeds. It was reported to sorb aflatoxin B1 with high affinity and high capacity in aqueous solutions and was shown to rescue broiler and Leghorn chicks from the toxic effects of 7500 ppb aflatoxin in the diet (Phillips et al. 1987, 1988, 2006). In subsequent studies, NS and other similar montmorillonite and smectite clays have been reported to protect against aflatoxin toxicity in a variety of young animals including rodents, chicks, turkey poults, ducklings, lambs, pigs, mink and trout (Phillips 1999; Phillips et al. 1990, 1991, 1994, 1995; Colvin et al. 1989; Bonna et al. 1991; Harvey et al. 1991a, 1991b; 1993, 1994; Voss et al. 1993; Kubena et al. 1990a, 1990b, 1991, 1993; Ledoux et al. 1999; Smith et al. 1994; Marquez and Hernandez 1995; Cerdchai et al. 1990; Lindemann et al. 1993; Abdel-Wahhab et al. 1998; Nahm 1995; Jayaprakash et al. 1992; Ellis et al. 2000). In studies using radiolabelled aflatoxins, NS clay has also been shown to decrease the bioavailability of aflatoxins and reduce aflatoxin residues in poultry (Davidson et al. 1987; Jayaprakash et al. 1992), rats (Sarr et al. 1995; Mayura et al. 1998) and pigs (Beaver et al. 1990). Aflatoxin M1 levels in milk from lactating dairy cattle and goats were also decreased when NS was included in the diet (Ellis et al. 1990; Harvey et al. 1991b; Smith et al. 1994). Mechanisms of aflatoxin sorption to NS. The suggested mechanism of AfB1 sorption by NS is an electron donor acceptor (EDA) mechanism. The platelets of NS clay are negatively charged due to isomorphic substitution, and thus they attract positively charged ions to balance this charge. Compounds with areas of electron deficiencies (partial positive areas) can also be attracted to the platelets (Haderlein et al. 1996). The carbons comprising the dicarbonyl system in aflatoxins are partially positive and have been shown to be essential to the adsorption process. AfB1 is planar with the exception of the terminal furan (Figure 1). The importance of the spatial orientation of AfB1 was demonstrated when stereochemical differences of some aflatoxin analogues resulted in significant effects on the tightness of binding. These results also suggested that the sorption of aflatoxin onto NS may favour an orientation where the furan is aligned away from the surface. Adsorption isotherms on heat-collapsed NS have demonstrated that the interlamellar region of NS is the primary site of binding with external surfaces accounting for only minor sorptions of aflatoxins. Based on the

Interventions that reduce the dose of aflatoxins from contaminated foods Food surveillance. Surveillance and subsequent regulation of susceptible commodities, such as groundnuts and maize for aflatoxins and other mycotoxins, are routinely used as a primary intervention to safeguard the health of consumers as well as the economic interests of producers and traders in various countries. These surveillance data are frequently used to establish regulatory guidelines that define the limits of aflatoxins and other mycotoxins in foods. However, in many developing countries, these guidelines are not adequately enforced and result in populations at high risk for aflatoxicosis, i.e. recent outbreak of acute aflatoxin poisoning in Kenya (CDC 2004; Lewis et al. 2005). Community education. One of the most practical and fundamental interventions at the subsistencefarm level in developing countries, is the use of low-technology approaches, such as community education on food handling and storage, as described by Turner et al. (2005). These primary approaches have been shown to reduce significantly the level of aflatoxin contamination in post-harvest foods and associated exposure in human populations at high risk for aflatoxicosis. Aflatoxin enterosorption (NovaSil clay). Another strategy for reducing food-borne exposure to mycotoxins is the inclusion of various binding agents or sorbents in the diet. Many of these binding agents are purported to prevent the deleterious effects of diverse mycotoxins in a variety of animals (primarily poultry and swine). As early as 1979, adsorbent clay minerals were reported to bind AfB1 in liquids (Masimango et al. 1979). Additionally, bleaching clays used to process canola oil were found to lessen the effects of T-2 toxin (Carson and Smith 1983; Smith 1984). The dietary consumption of earth (i.e. geophagy) has been observed for centuries and across all continents in both humans and animals (Carretero 2002). Clay eating has been recorded from traditional human societies and is considered culturally acceptable in many African countries and China (Johns and Duquette 1991; Diamond 1999). A practical approach of current interest for the prevention of aflatoxicoses is the incorporation of non-nutritive clay minerals in contaminated food/ feed to sorb aflatoxins in the stomach and intestinal tract, thus reducing toxin bioavailability and distribution to the blood, liver and other target organs (Phillips et al. 1995, 2002, 2006; Phillips 1999). Using multiple animal models, the present authors laboratory has shown that NovaSilTM (NS) clay,

Clay-based remedy for aflatoxins thermodynamics, AfB1 binds strongly to NS, exhibited by an estimated heat of sorption (enthalpy) of 50 KJ mol1. Interference from compounds with stereochemical restrictive groups could also play an important role in the adsorption process. For the analogues that contain functional groups that make them larger than AfB1, their insertion, docking and adsorption at surfaces in the interlamellar channel might be restricted. Our results also indicate a good correlation between the magnitude of partial positive charges on carbons C11 and C1 of the -dicarbonyl system and the strength of adsorption of planar ligands, suggesting an EDA mechanism with the surface of the clay. Other mechanisms of AfB1 sorption to NS surfaces involve the potential chelation of interlayer cations (especially Ca2) and various edge-site metals (Grant 1998; Phillips 1999; Phillips et al. 2002, 2006). Selectivity of NS clay. Research has demonstrated that NS clay has a notable preference (and capacity) for aflatoxins. NS at a level of 0.5% w/w in the diet of poultry did not impair phytate or inorganic phosphorous utilization (Chung and Baker 1990). In other studies in poultry, the addition of NS at concentrations of 0.5% (which is recommended for anticaking in feeds), did not impair the utilization of riboflavin, vitamin A, manganese, or zinc (Chung et al. 1990). NS (0.5% w/w) was also shown to protect young chickens from aflatoxin levels as high as 7500 ppb; these levels are not likely to be found in human food; although, the recent exposure of humans in Kenya was linked to toxin levels as high as 8000 ppb. While clay-based interventions are clearly effective for aflatoxins, the same effectiveness has not been demonstrated for other mycotoxins. Importantly, unmodified NS clays have not been shown to strongly bind other structurally diverse mycotoxins, e.g. zearalenone, deoxynivalenol, T-2 toxin, ochratoxin A, cyclopiazonic acid, ergotamine, and fumonisins, nor do they significantly prevent the adverse effects of these mycotoxins when included in the diet of animals. For example, in enterosorbent studies in poultry with mycotoxins other than aflatoxin, the inclusion of NS clay in the diet did not prevent the adverse effects of cyclopiazonic acid (Dwyer et al. 1997), T-2 toxin (Kubena et al. 1990a), diacetoxyscirpenol (Kubena et al. 1993), ochratoxin A (Huff et al. 1992), and fumonisins (Lemke 2000). The inclusion of clay in the zearalenone-contaminated diets of mink alleviated some fetotoxicity, but did not reduce the hyperestrogenic effects (Bursian et al. 1992). Also the average daily weight gain was unchanged in pigs exposed to deoxynivalenol when clay was added to the diet at 0.5 and 1.0% w/w. The only effective

139

method for decreasing deoxynivalenol toxicity was dilution of the contaminated maize (Patterson and Young 1993). Although in vitro tests showed potential for protection of ergotamine toxicity with NS (Chestnut et al. 1992; Huebner et al. 1999), NS at levels of 2.0% w/w did not protect rats or sheep from fescue toxicosis. NSs selectivity was further demonstrated in our laboratory in studies involving nanostructured thin films of NS on quartz that were used as affinity probes for aflatoxins in contaminated media. Our findings show that this composite exhibited comparable selectivity to the Aflatest affinity column from Vicam (Huebner and Phillips 2003; Huebner et al. 2004). Long-term safety study in rats. In earlier studies in animals, no observable adverse effects from NS were reported following ingestion of doses up to 2.0% w/w in the diet. For example, SpragueDawley rats that ingested NS clay at dietary concentrations as high as 2% throughout pregnancy, did not show significant trace metal bioavailability in a variety of tissues and showed neither maternal nor foetal toxicity (Wiles et al. 2004). Since most of our preliminary work was based on short-term exposures not greater than 6 weeks in duration, a long-term exposure was warranted to establish the safety of NS further. Before an adverse events/dosimetry trial in humans, a rodent model was used to evaluate the relative safety of chronic exposure to NS clay in the diet. Male and female SpragueDawley rats were fed rations containing 0, 0.25, 0.5, 1.0, and 2.0% levels of NS clay ad libitum over a period of 6.5 months. No morbidity or mortality was observed in the animals throughout the study duration. The results of this study indicated that rats treated with 0.252% NS clay in the diet did not exhibit dose-dependent or NS-related adverse effects on body weight gains, feed conversion ratios, relative organ weights, gross anatomy and histological appearance of major organs, haematology, and serum biochemistry parameters. Additionally, levels of selected nutrients including vitamins A and E, Fe, and Zn were unaffected (Afriyie-Gyawu et al. 2005). Given the safety and efficacy of NS, as demonstrated in a variety of animal models, it was hypothesized that NS-based interventions might be beneficial for the treatment of humans who are at high risk for aflatoxicosis (Phillips et al. 2006). Adverse events/dosimetry trial with NS in humans. As a precursor to a phase IIa clinical intervention trial with NS in Ghana, a short-term (2-week) safety evaluation of NS was carried out at Texas Tech University in 50 healthy adults (Wang et al. 2005). The overall design followed the guidelines

140

T. D. Phillips et al.

for a randomized and double-blind phase I clinical trial. This study was conducted: (1) to evaluate the short-term safety and tolerance of NS capsules in normal human subjects; and (2) to establish optimal protocols for human intervention studies. NS capsules were produced under sterile conditions using US Good Manufacturing Practices. Also, the NS was sterilized at 121 C before encapsulation. Preceding the chronic animal and short-term human studies, NS was analysed for concentrations of various environmental contaminants, including priority toxic metals and dioxins/furans to ensure compliance with federal and international standards. A total of 50 adults, ages 2045, who met the recruiting criteria were randomly divided into two study groups. The high-dose group (HD) took three capsules of NS three times a day (a total of 3.0 g), and the low dose group (LD) took three capsules of NS three times a day (a total of 1.5 g). All capsules were of the same colour and size. The two dose levels were extrapolated from previously published dosimetry data from animal studies (Phillips 1999; Phillips et al. 2002, 2004; Afriyie-Gyawu et al. 2005). After 14 days of capsule ingestion, NS (up to 3.0 g day1) was considered safe for further human studies based on physical examination, biochemistry and haematology results. Concentrations of the standard parameters analysed after the trial were statistically similar to those levels determined before trial. Also, no significant difference was observed for any reported adverse symptoms between LD and HD groups. This study confirms the selectivity of NS clay for aflatoxins, in that no statistical differences were observed in the levels of serum vitamins A and E, and iron and zinc in the participants after 2 weeks of NS ingestion. This evidence further confirms that NS demonstrates binding specificity for aflatoxins and lack of interaction with vitamins A and E. The adverse events trial provided the basis for the phase IIa human intervention study at the Ejura-Sekyedumase district (ESD) of the Ashanti region of Ghana, West Africa. Screening of clays in Ghana for aflatoxin binders. Before the initiation of a 3-month clinical intervention trial in Ghana, 73 edible clays from the marketplace at ESD and 11 clays used in the ceramic industry from other locations in Ghana were tested for aflatoxin sorption using isothermal analyses in our laboratory. Our rationale for screening clays near the study site in Ghana was to identify those that were similar to NS clay for subsequent human studies. It was hypothesized that study participants could be categorized as geophagic or non-geophagic in the context of aflatoxin exposure. However, upon analysis of 84 samples from different geographic

Figure 3. Representative isothermal plots of the most effective and least effective clay samples from Ghana for the sorption of AfB1. These plots are compared with a standard isotherm for NovaSil. The most effective sample from Ghana was obtained from the Brong/Ahafo Region (Qmax 0.07), and the least effective sample was obtained from the Ashanti Region None was comparable with NovaSil (Qmax 0.00). (Qmax 0.40), and would not be expected to decrease the bioavailability of AfB1.

locations in the country, we found that none of these sorbed aflatoxins with high affinity and capacity (Figure 3). Phase IIa clinical intervention trial with NS in Ghana. Aflatoxin contamination in food products remains a serious burden in the developing world where a lack of untainted food supplies and poverty present a major and persistent challenge to many people in affected areas (McAlpin et al. 2002; Shephard 2003). Avoiding consumption of aflatoxin-contaminated foods is one of the most fundamental approaches for reducing risk of aflatoxicosis in humans. However, this is simply not feasible for many communities in developing countries and therefore emphasizes the need for viable intervention strategies to manage aflatoxin contaminated diets and treat aflatoxicosis. A recent study involved a 3-month double-blind and placebo controlled, phase IIa clinical trial conducted in the EjuraSekyedumase district, Ashanti Region, Ghana (Afriyie-Gyawu et al. 2007). The objective was to evaluate the safety, efficacy, and tolerance of dietary NS when administered to humans for the prevention of aflatoxin exposure and toxicity. The study protocol was approved by the Institutional Review Boards of Texas A&M University and its counterpart in Ghana for Ethical Clearance. Five hundred and seven volunteers were clinically screened to evaluate their general health, pregnancy status, and blood AfB1-albumin adduct levels, and 177 of them were enrolled as study participants. Subjects were randomly assigned to three groups: high-dose (HD),

Clay-based remedy for aflatoxins

141

in the blood and urine from study participants (Wang et al. 2007). Further studies are planned to optimize the dosimetry and delivery methods for NS clay. Also, phase IIb, phase III intervention, and epidemiological studies are needed to confirm the safety and efficacy of NS for long-term therapy and the potential inclusion in foods for humans in areas at high risk for aflatoxicosis.

Summary NS clay is commonly used as an anti-caking agent in animal feeds. Importantly, its inclusion in feed (at relatively low levels) may also serve to protect animals by tightly sorbing aflatoxins in the stomach and intestines resulting in decreased bioavailability. Based on numerous studies, it is anticipated that NS clay-based enterosorption of aflatoxins in animals will result in improved growth rates, feed conversions and general health along with diminished aflatoxin residues in foods of animal origin (such as milk). Our recent findings from clinical intervention trials with NS are of particular relevance to populations in developing countries where the incidence of HCC and adverse health impacts from frequent exposures to aflatoxin are often elevated. Moreover, the use of NS for the protection of humans that are at high risk for HCC and aflatoxicosis appears to be culturally acceptable and sustainable. Eventually, the preferred delivery of NS may be through its inclusion in salt (like iodine), taking advantage of its anticaking properties, or as an additive in common groundnut and maize-based foods. Extensive research with NS clay and other sorbent materials suggests that potential mycotoxin enterosorbents (e.g. chemical and biological binders and/or sequestering agents) should be rigorously evaluated in vitro and in vivo. These should meet the following criteria: . Favourable thermodynamic characteristics of sorption. . Tolerable levels of priority metals, dioxins/ furans and other hazardous substances. . Safety and efficacy in multiple animal species. . Safety and efficacy in long-term studies. . Negligible interactions with vitamins, iron and zinc.

Figure 4. NS capsules for the phase IIa clinical intervention trial in Ghana. Treatment doses were 0, 1.5 or 3.0 g NS day1 before meals with water (microcrystalline cellulose was used as a placebo).

low-dose (LD) and placebo-control (PL) groups that received 3.0, 1.5 and 0 g NS day1, respectively, in capsules (Figure 4). To ensure compliance to treatment regimens and participant well-being, trained study monitors supervised administration of the encapsulated NS to participants and recorded side effects daily. On-site physicians performed physical examinations monthly. Blood and urine samples were collected for laboratory analysis. Over 90% of the participants completed the study, and compliance rate was more than 97%. Also, 99% of the time participants reported no side effects throughout the study. Mild to moderate adverse health events (approximately 0.5% of the time) were recorded in some participants but none of them appeared to be associated with NS treatment. No NS-related, significant differences were shown in haematology, liver and kidney functions, and electrolytes among the three groups. In the serum biochemical analysis, isolated statistical differences in a few parameters were detected but no trends of association or dose-dependency were observed, and were all within the normal physiological ranges (Afriyie-Gyawu et al. 2007). This study represents the first phase IIa clinical intervention trial to evaluate the safety and efficacy of NS clay in human subjects. Results suggest that short-term inclusion of NS in the diet at a minimal effective dose (MED) of 0.25% (w/w) would not likely produce overt toxicity in humans. Moreover, the results of this study support the application of NS for the management of aflatoxicosis in humans who are acutely exposed to high levels of dietary aflatoxins. Additional results from this study indicate that ingestion of capsules containing an MED of NS, significantly reduce biomarkers of aflatoxin exposure

Acknowledgements This work was supported by research grants from the US Agency for International Development (USAID LAG-G-00-96-90013-00) through Peanut CRSP of the University of Georgia and NIEHS P42-ES04917.

142

T. D. Phillips et al.
Chung T, Erdman Jr J, Baker D. 1990. Hydrated sodium calcium aluminosilicate: effects on zinc, manganese, vitamin A and riboflavin utilization. Poultry Science 69:1364. Colvin B, Sangster L, Hayden K, Bequer R, Wilson D. 1989. Effect of high affinity aluminosilicate sorbent on prevention of aflatoxicosis in growing pigs. Veterinary and Human Toxicology 31:46. Council for Agricultural Science and Technology (CAST). 1989. Mycotoxins: economic and health risks. Task Force Report No. 116. Ames, IA: CAST. p. 191. Dashwood R, Negishi T, Hayatsu H, Breinholt V, Hendricks J, Bailey G. 1998. Chemopreventive properties of chlorophylls towards aflatoxin B1: A review of the antimutagenicity and anticarcinogenicity data in rainbow trout. Mutation Research 399:245. Davidson J, Babish J, Delaney K, Taylor D, Phillips TD. 1987. Hydrated sodium calcium aluminosilicate decreases the bioavailability of aflatoxin in the chicken. Poultry Science 66:89. Diamond J. 1999. Evolutionary biology: Dirty eating for healthy living. Nature 400:120. Dragsted LO, Strube M, Larsen JC. 1993. Cancer-protective factors in fruits and vegetables: Biochemical and biological background. Pharmacology and Toxicology 72(Suppl. 1): 116135. Dwyer M, Kubena L, Harvey R, Mayura K, Sarr A, Buckley S, Bailey R, Phillips TD. 1997. Effects of inorganic adsorbents and cyclopiazonic acid in broiler chickens. Poultry Science 76:1141. Eaton D, Ramsdell H, Neal G. 1994. Biotransformation of aflatoxins. In: Eaton L, Groopman J, editors. The toxicology of aflatoxins, human health, veterinary agricultural significance. New York, NY: Academic Press. Egner P, Wang J, Zhu Y, Zhang B, Wu Y, Zhang Q, Qian G, Kuang S, Gange S, Jacobson L, et al. 2001. Chlorophyllin intervention reduces aflatoxinDNA adducts in individuals at high risk for liver cancer. Proceedings of the National Academy of Sciences, USA 98: 14601. Ellis J, Harvey R, Kubena L, Bailey R, Clement B, Phillips TD. 1990. Reduction of aflatoxin M1 residues in milk utilizing hydrated sodium calcium aluminosilicate. Toxicologist 10:163 (abst). Ellis R, Clements M, Tibbetts A, Winfree R. 2000. Reduction of the bioavailability of 20 mg/kg aflatoxin in trout feed containing clay. Aquaculture 183:179. Essigmann JM, Croy RG, Nadzan AM, Busby WF, Reinhold VN, Buchi GB, Wogan GN. 1977. Structural identification of the major DNA adduct formed by aflatoxin B1 in vitro. Proceedings of the National Academy of Sciences, USA 74: 18701874. Fujiki H, Suganuma M, Imai K, Nakachi K. 2002. Green tea: Cancer preventive beverage and/or drug. Cancer Letters 188:9. Garland T, Reagor J. 2001. Chronic canine aflatoxicosis and management of an epidemic. In: deKoe W, Samson R, Van Egmond H, Gilbert J, Sabina M, editors. Mycotoxins and phycotoxins in perspective at the turn of the millennium. Wageningen: Ponsen & Looven. pp 231236. Gong YY, Cardwell K, Hounsa A, Egal S, Turner PC, Hall AJ, Wild CP. 2002. Dietary aflatoxin exposure and impaired growth in young children from Benin and Togo: Cross sectional study. British Medical Journal 325: 2021. Gong YY, Cardwell K, Hounsa A, Egal S, Turner PC, Sutcliffe AE, Hall AJ, Cardwell K, Wild CP. 2004. Postweaning exposure to aflatoxin results in impaired child growth: A longitudinal study in Benin, West Africa. Environmental Health Perspectives 112:13341338.

References
Abdel-Wahhab M, Nada S, Farag I, Abbas N, Amra H. 1998. Potential protective effect of HSCAS and bentonite against dietary aflatoxicosis in rat: With special reference to chromosomal aberration. Natural Toxins 6:211. Afriyie-Gyawu E, Ankrah N-A, Huebner H, Ofosuhene M, Kumi J, Johnson N, Tang L, Xu L, Jolly P, Ellis P, et al. 2007. NovaSil clay intervention in Ghanaians at high risk for aflatoxicosis, Part I: Study design and clinical outcomes. Food Additives and Contaminants (accepted). Afriyie-Gyawu E, Mackie J, Dash B, Wiles M, Taylor J, Huebner H, Tang L, Guan H, Wang JS, Phillips T. 2005. Chronic toxicological evaluation of dietary NovaSil clay in SpragueDawley rats. Food Additives and Contaminants 22:259. Aguilar F, Hussain S, Cerutti P. 1993. Aflatoxin B1 induces the transversion of G ! T in codon 249 of the p53 tumor suppressor gene in human hepatocytes. Proceedings of the National. Academy of Sciences, USA 90: 8586. Asao T, Buechi G, Abdel-Kader MM, Chang SB, Wick EL, Wogan GN. 1963. Aflatoxins B and G. Journal of the American Chemical Society 85:17061707. Beaver R, Wilson D, James M, Haydon K. 1990. Distribution of aflatoxins in tissues of growing pigs fed an aflatoxincontaminated diet amended with a high affinity aluminosilicate sorbent. Veterinary and Human Toxicology 32:16. Bonna R, Aulerich R, Bursian S, Poppenga R, Braselton W, Watson G. 1991. Efficacy of hydrated sodium calcium aluminosilicate and activated charcoal in reducing the toxicity of dietary aflatoxin to mink. Archives of Environmental. Contamination and Toxicology 20:441. Breinholt V, Schimerlik M, Dashwood R, Bailey G. 1995. Mechanisms of chlorophyllin anticarcinogenesis against aflatoxin B1: Complex formation with the carcinogen. Chemical Research in Toxicology 8:506. Bulatao-Jayme J, Almero E, Castro M, Jardeleza M, Salamat L. 1982. A case-control dietary study of primary liver cancer risk from aflatoxin exposure. International Journal of Epidemiology 11:112119. Bursian S, Aulerich R, Cameron J, Ames N, Steficek B. 1992. Efficacy of hydrated sodium calcium aluminosilicate in reducing the toxicity of dietary zearalenone to mink. Journal of Applied Toxicology 12:8590. Carretero MI. 2002. Clay minerals and their beneficial effects upon human health. Applied Clay Science 21:155163. Carson M, Smith T. 1983. Role of bentonite in prevention of T-2 toxicosis in rats. Journal of Animal Science 57:1498. Centers for Disease Control and Prevention (CDC). 2004. Outbreak of aflatoxin poisoning Eastern and Central Provinces, Kenya. Morbidity and Mortality Weekly Report 53: 790793. Cerdchai R, Paisansarakit A, Khajarern J. 1990. Effect of hydrated sodium calcium aluminosilcate (NovaSil) on reducing aflatoxicosis in ducks. In Proceedings of the 7th FAVA Congress, Pattaya. p. 391. Chestnut A, Anderson P, Cochran M, Fribourg H, Gwinn K. 1992. Effects of hydrated sodium calcium aluminosilicate on fescue toxicosis and mineral absorption. Journal of Animal Sciences 70:2838. Chi WJ, Doong SL, Lin-Shiau SY, Boone CW, Kelloff GJ, Lin JK. 1998. Oltipraz, a novel inhibitor of hepatitis B virus transcription through elevation of p53 protein. Carcinogenesis 19:21332138. Chung T, Baker D. 1990. Phosphorus utilization in chicks fed hydrated sodium calcium aluminosilicate. Journal of Animal Science 68:1364.

Clay-based remedy for aflatoxins

Gopalakrishnan S, Harris TM, Stone MP. 1990. Intercalation of aflatoxin B1 in two oligodeoxynucleotide adducts: Comparative 1 H NMR analysis of d(ATCAFBGAT).d(ATCGAT) and d(ATAFBGCAT)2. Biochemistry 29:1043810448. Grant P. 1998. Investigation of the mechanism of aflatoxin B1 adsorption to clays and sorbents through the use of isothermal analysis. PhD dissertation, Texas A&M University, College Station, TX. Groopman J, Cain L, Kensler T. 1988. Aflatoxin exposure in human populations: Measurements and relationship to cancer. CRC Critical Reviews in Toxicology 19:113145. Guengerich FP, Arneson KO, Williams KM, Deng Z, Harris TM. 2002a. Reaction of aflatoxin B1 oxidation products with lysine. Chemical Research in Toxicology 15:780793. Guengerich FP, Voehler M, Williams KM, Deng Z, Harris TM. 2002b. Structure of the aflatoxin B(1) dialdehyde adduct formed from reaction with metylamine. Chemical Research in Toxicology 15:793798. Gupta R, Dubois R. 2001. Colorectal cancer prevention and treatment by inhibition of cycloxygenase-2. Nature Reviews Cancer 1:1121. Haderlein S, Weissmahr K, Schwarzenbach R. 1996. Specific adsorption of nitroaromatic explosives and pesticides to clay minerals. Environmental Science and Technology 30:612. Harvey R, Kubena L, Elissalde M, Corrier D, Phillips TD. 1994. Comparison of two hydrated sodium calcium aluminosilicate compounds to experimentally protect growing barrows from aflatoxicosis. Journal of Veterinary Diagnostic Investigation 6:88. Harvey R, Kubena L, Elissalde M, Phillips TD. 1993. Efficacy of zeolitic ore compounds on the toxicity of aflatoxin to growing broiler chickens. Avian Diseases 37:67. Harvey R, Kubena L, Phillips TD, Corrier D, Elissalde M, Huff W. 1991a. Diminution of aflatoxin toxicity to growing lambs by dietary supplementation with hydrated sodium calcium aluminosilicate. American Journal of Veterinary Research 52:152. Harvey R, Phillips TD, Ellis J, Kubena L, Huff W, Petersen D. 1991b. Effects of aflatoxin M1 residues in milk by addition of hydrated sodium calcium aluminosilicate to aflatoxincontaminated diets of dairy cows. American Journal of Veterinary Research 52:1556. Hayatsu H, Arimoto S, Negishi T. 1988. Dietary inhibitors of mutagenesis and carcinogenesis. Mutation Research 202:429446. Hesseltine CW. 1979. Some important fermented foods of MidAsia, the Middle East, and Africa. Journal of the American Oil Chemists Society 56:367374. Hinton DM, Myers MJ, Raybourne RA, Francke-Carroll S, Sotomayor RE, Shaddock J, Warbritton A, Chou MW. 2003. Immunotoxicity of aflatoxins in rats: Effects on lymphocytes and the inflammatory response in a chronic intermittent dosing study. Toxicological Sciences 73:362377. Huebner H, Herrera P, Phillips TD. 2004. Clay-based interventions for the control of chemical and microbial hazards in food and water. In: Beier R, Pallai S, Phillips TD, editors. Preharvest and postharvest food safety: Contemporary issues and future directions. Ames, IA: IFT Press. pp 389402. Huebner H, Lemke S, Ottinger S, Mayura K, Phillips TD. 1999. Molecular characterization of high affinity, high capacity clays for the equilibrium sorption of ergotamine. Food Additives and Contaminants 16:159. Huebner H, Phillips TD. 2003. Clay-based affinity probes for elective cleanup and determination of Aflatoxin B1 Using Nanostructured Montmorillonite on Quartz. Journal of AOAC International 86:534.

143

Huff W, Kubena L, Harvey R, Phillips TD. 1992. Efficacy of a hydrated sodium calcium aluminosilicate to reduce the individual and combined toxicity of aflatoxin and ochratoxin A. Poultry Science 71:64. International Agency for Research on Cancer (IARC). 1976. Some naturally occurring substances. Monographs on the evaluation of the carcinogenic risk of chemicals to man, Vol. 10. Lyon: IARC. International Agency for Research on Cancer (IARC). 1987. Aflatoxins. Monograph on the evaluation of carcinogenic risk to humans, Suppl. 7. Lyon: IARC. International Agency for Research on Cancer (IARC). 1993. Aflatoxins. Monograph on the evaluation of carcinogenic risk to humans, Suppl. 7. Lyon: IARC. Jayaprakash M, Gowda R, Vijayasarathi S, Seshadri S. 1992. Adsorbent efficacy of hydrated sodium calcium aluminosilicate in induced aflatoxicosis in broilers. Indian Journal of Veterinary Pathology 16:102. Jiang Y, Jolly P, Ellis W, Wang J, Phillips T, Williams J. 2005. Aflatoxin B1 albumin adduct levels and cellular immune status in Ghanaians. International Immunology 17:807814. Johns T, Duquette M. 1991. Detoxification and mineral supplementation as functions of geophagy. American Journal of Clinical Nutrition 53:448. Kensler TW, Groopman JD, Sutter TR, Curphey TJ, Roebuck BD. 1999. Development of cancer chemopreventive agents: oltipraz as a paradigm. Chemical Research in Toxicology 12:113126. Klaunig J, Kamendulis L. 1999. Mechanisms of cancer chemoprevention in hepatic carcinogenesis: Modulation of focal lesion growth in mice. Toxicological Sciences 52:101. Kubena L, Harvey R, Huff W, Corrier D, Phillips TD. 1990b. Ameliorating properties of a hydrated sodium calcium aluminosilicate on the toxicity of aflatoxin and T-2 toxin. Poultry Science 69:1078. Kubena L, Harvey R, Huff W, Yersin A, Elissalde M, Witzel D. 1993. Efficacy of a hydrated sodium calcium aluminosilicate to reduce the toxicity of aflatoxin and diacetoxyscirpenol. Poultry Science 72:51. Kubena L, Harvey R, Phillips TD, Corrier D, Huff W. 1990a. Diminution of aflatoxicosis in growing chickens by dietary addition of a hydrated sodium calcium aluminosilicate. Poultry Science 69:727. Kubena L, Huff W, Harvey R, Yersin A, Elissalde M, Witzel D, Giroir L, Phillips TD. 1991. Effects of hydrated sodium calcium aluminosilicate on growing turkey poults during aflatoxicosis. Poultry Science 70:1823. Lancaster M, Jenkins F, Philp J. 1961. Toxicity associated with certain samples of groundnuts. Nature 192:1095. Lang SS. 2006. Dogs keep dying: Too many owners unaware of toxic dog food. Cornell University ChronicleOnline, 6 January (available at: http://www.news.cornell.edu/stories/Jan06/ dogs.dying.ssl.html). Ledoux D, Rottinhaus G, Bermudez A, Alonso-Debolt M. 1999. Efficacy of a hydrated sodium calcium aluminosilicate to ameliorate the toxic effects of aflatoxin in broiler chicks. Poultry Science 78:204. Leeson S, Diaz G, Summers J. 1995. In: Poultry metabolic disorders and mycotoxins. Guelph: University Books. Lemke S. 2000. Investigation of clay-based strategies for the protection of animals from the toxic effects of selected mycotoxins. PhD dissertation, Texas A&M University, College Station, TX. Lewis L, Onsongo M, Njapau H, Schurz-Rogers H, Luber G, Kieszak S, Nyamongo J, Backer L, Dahiye AM, Misore A, et al. 2005. Aflatoxin contamination of commercial maize products during an outbreak of acute aflatoxicosis in eastern and central Kenya. Environmental Health Perspectives 113:17631767.

144

T. D. Phillips et al.
Phillips TD, Clement B, Park D. 1994. Approaches to reduction of aflatoxin. In: Eaton L, Groopman J, editors. The toxicology of aflatoxins, human health, veterinary agricultural significance. New York, NY: Academic Press. Phillips TD, Kubena L, Harvey R, Taylor D, Heidelbaugh N. 1987. Mycotoxin hazards in agriculture: new approach to control. Journal of American Veterinary Medical Association 190:1617 (abst). Phillips TD, Kubena L, Harvey R, Taylor D, Heidelbaugh N. 1988. Hydrated sodium calcium aluminosilicate: A high affinity sorbent for aflatoxin. Poultry Science 67:243. Phillips TD, Lemke SL, Grant PG. 2002. Characterization of clay-based enterosorbents for the prevention of aflatoxicosis. Advances in Experimental Medicine and Biology 504:157171. Phillips TD, Sarr A, Grant P. 1995. Selective chemisorption and detoxification of aflatoxins by phyllosilcate clay. Natural Toxins 3:204. Phillips TD, Sarr AB, Clement B, Kubena L, Harvey R. 1991. Prevention of aflatoxicosis in farm animals via selective chemisorption of aflatoxin. In: Bray G, Ryan D, editors. Mycotoxins, cancer and health. Vol. 1. Baton Rouge, LA: Louisiana State University Press. Pimpukdee K, Kubena LF, Bailey CA, Huebner HJ, Afriyie-Gyawu E, Phillips TD. 2004. Aflatoxin-induced toxicity and depletion of hepatic vitamin A in young broiler chicks: Protection of chicks in the presence of low levels of NovaSil PLUS in the diet. Poultry Science 83:737. Qin G, Gopalan-Kriczky P, Su J, Ning Y, Lotlikar P. 1997. Inhibition of aflatoxin B1-induced initiation of hepatocarcinogenesis in the rat by green tea. Cancer Letters 112:149. RamsbottomJ.1953.Mushroomsandtoadstools. London: Collins. Richard JL, Thurston JR, Pier AC. 1978. Effects of mycotoxins on immunity. Toxin. In: Rosenberg P, editor. Animal, plant and microbial. New York, NY: Pergamon. pp 801817. Rodricks JV, Stoloff L. 1977. Aflatoxin residues from contaminated feed in edible tissues of food-producing animals. In: Mycotoxins in Human and Animal Health, Proceedings of a Conference. pp 6779. Sabbioni G, Skipper PL, Buchi G, Tannenbaum SR. 1987. Isolation and characterization of the major serum albumin adduct formed by aflatoxin B1 in vivo in rats. Carcinogenesis 8:819824. Sargeant K, OKelly J, Carnaghan R, Allcroft R. 1961. The assay of a toxic principle in certain groundnut meals. Veterinary Record 73:1219. Sarr A, Mayura K, Kubena L, Harvey R, Phillips TD. 1995. Effects of phyllosilicate clay on the metabolic profile of aflatoxin B1 in Fischer-344 rats. Toxicology Letters 75:145. Shephard G. 2003. Aflatoxin and food safety: recent African perspectives. Journal of Toxicology, Toxin Reviews 22: 267286. Siller W, Ostler D. 1961. The histopathology of an enterohepatic syndrome of turkey poults. Veterinary Record 73:134. Smith E, Phillips TD, Ellis J, Harvey R, Kubena L, Thompson J, Newton G. 1994. Dietary hydrated sodium calcium aluminosilicate reduction of aflatoxin M1 residue in dairy goat milk and effects on milk production and components. Journal of Animal Science, 72:677. Smith T. 1984. Spent canola oil bleaching clays: Potential for treatment of T-2 toxicosis in rats and short-term inclusion in diets of immature swine. Canadian Journal of Animal Science 64:725. Sporn M, Suh N. 2002. Chemoprevention: An essential approach to controlling cancer. Nature Reviews Cancer 2:537543.

Lindemann M, Blodgett D, Kornegay E, Schurig G. 1993. Potential ameliorators of aflatoxicosis in weanling/growing swine. Journal of Animal Science 71:171. Marquez R, Hernandez I. 1995. Aflatoxin adsorbent capacity of two Mexican aluminosilicates in experimentally contaminated chick diets. Food Additives and Contaminants 12:431. Masimango N, Remacle J, Ramaut J. 1979. Elimination, par des argiles gonflantes, de Laflatoxine B1 des milieus contamines. Annales de la Nutrition et de lAlimentation 33:137. Mayura K, Abdel-Wahhab M, McKenzie K, Sarr A, Edwards J, Naguib K, Phillips TD. 1998. Prevention of maternal and developmental toxicity in rats via dietary inclusion of common aflatoxin sorbents: Potential for hidden risks. Toxicological Sciences 41:175. McAlpin CE, Wicklow DT, Horn BW. 2002. DNA fingerprinting analysis of vegetative compatibility groups in Aspergillus flavus from a peanut field in Georgia. Plant Diseases 86:254. McMahon G, Davis E, Wogan GN. 1987. Characterization of c-Ki-ras oncogene alleles by direct sequencing of enzymically amplified DNA from carcinogen-induced tumors. Proceedings of the National Academy of Sciences, USA 84: 49744978. McMahon G, Hanson L, Lee JJ, Wogan GN. 1986. Identification of an activated c-Ki-ras oncogene in rat liver tumors induced by aflatoxin B1. Proceedings of the National Academy of Sciences, USA 83: 94189422. Miller DM, Stuart SP, Crowell WA, Cole RJ, Goven AJ, Brown J. 1978. Aflatoxicosis in swine: Its effect on immunity and relationship to salmonellosis. Proceedings of the Annual Meeting of the American Association of Veterinary Lab Diagnostics 21:135146. Muller R, Carlson C, Semeniuk G, Harshfield G. 1970. Response of chicks, ducklings, goslings, pheasants, and poults to graded levels of aflatoxins. Poultry Science 49:1346. Nahm K. 1995. Prevention of aflatoxicosis by addition of antioxidants and hydrated sodium calcium aluminosilicate to the diet of young chicks. Japanese Journal of Poultry Science 32:117. Nakachi K, Matsuyama S, Miyake S, Suganuma M, Imai K. 2000. Preventive effects of drinking green tea on cancer and cardiovascular disease: Epidemiological evidence for multiple targeting prevention. BioFactors 13:49. Patterson R, Young L. 1993. Efficacy of hydrated sodium calcium aluminosilicate, screening and dilution in reducing the effects of mold contaminated corn in pigs. Canadian Journal of Animal Science 73:616. Peraica M, Radic B, Lucic A, Pavlovic M. 1999. Toxic effects of mycotoxins in humans. Bulletin of the World Health Organization 77:75466. Peska J, Bondy G. 1994. Immunotoxic effects of mycotoxins. In: Miller J, Trenholm H, editors. Mycotoxins in grain: Compounds other than aflatoxin. St Paul, MN: Eagan. Phillips TD. 1999. Dietary clay in the chemoprevention of aflatoxin-induced disease. Toxicology Sciences 52:118. Phillips TD, Afriyie-Gyawu E, Wang JS, Williams J, Huebner H. 2006. The potential of aflatoxin sequestering clay. In: Barug D, Bhatnagar D, Van Egmond H, Van der Kamp J, Van Osenbruggen W, Visconti A, editors. The mycotoxin fact book. Wageningen: Wageningen Academic Publ. pp 329346. Phillips TD, Clement B, Kubena L, Harvey R. 1990. Detection and detoxification of aflatoxins: Prevention of aflatoxicosis and aflatoxin residues with hydrated sodium calcium aluminosilicates. Veterinary and Human Toxicology 32:15.

Clay-based remedy for aflatoxins

Turner PC, Collinson AC, Cheung YB, Gong YY, Hall AJ, Prentice AM, Wild CP. 2007. Aflatoxin exposure in utero causes growth faltering in Gambian infants. International Journal of Epidemiology Advance Access (18 June). Turner PC, Moore SE, Hall AJ, Prentice AM, Wild CP. 2003. Modification of immune function through exposure to dietary aflatoxin in Gambian children. Environmental Health Perspectives 111:217220. Turner PC, Sylla A, Gong YY, Diallo MS, Sutcliffe AE, Hall AJ, Wild CP. 2005. Reduction in exposure to carcinogenic aflatoxins by postharvest intervention measures in West Africa: A community-based intervention study. Lancet 365:19501956. Van Rensburg SJ, Altenkirk B. 1974. Claviceps purpurea ergotism. In: IFH, editor. Mycotoxins purchase. Elsevier. pp 6996. Voss K, Dorner J, Cole R. 1993. Amelioration of aflatoxicosis in rats by volclay NF-BC, microfine bentonite. Journal of Food Protection 56:595. Wang JS, Luo H, Billam M, Wang Z, Guan H, Tang L, Goldston T, Afriyie-Gyawu E, Lovett C, Griswold J, et al. 2005. Short-term safety evaluation of processed calcium montmorillonite clay (NovaSil) in humans. Food Additives and Contaminants 22:270279. Wang JS, Qian GS, Zarba A, He X, Zhu YR, Zhang BC, Jacobson L, Gange SJ, Munoz A. 1996. Temporal patterns of aflatoxinalbumin adducts in hepatitis B surface antigen-positive and antigen-negative residents of Daxin, Qidong County, Peoples Republic of China. Cancer Epidemiology, Biomarkers and Prevention 5:253261. Wang P, Afriyie-Ggawu E, Tang Y, Johnson NM, Xu L, Tang L, Huebner HJ, Ankrah NA, Ofori-Adjei D, Ellis W, et al. 2007. NovaSil clay intervention in Ghanaians at high risk

145

for aflatoxicosis: II. Reduction in biomarkers of aflatoxin in blood and urine. Food Additives and Contaminants (in press). Wang JS, Shen X, He X, Zhu YR, Zhang BC, Wang JB, Qian GS, Kuang SY, Zarba A, Egner PA, et al. 1999. Protective alterations in phase 1 and 2 metabolism of aflatoxin B1 by oltipraz in residents of Qidong, Peoples Republic of China. Journal of the National Cancer Institute 91:347354. Wild CP, Hall AJ. 2000. Primary prevention of hepatocellular carcinoma in developing countries. Mutation Research 462:381393. Wild CP, Turner PC. 2002. The toxicology of aflatoxins as a basis for public health decisions. Mutagenesis 17:471481. Wiles M, Huebner H, Afriyie-Gyawu E, Taylor R, Bratton G, Phillips TD. 2004. Toxicological evaluation and metal bioavailability in pregnant rats following exposure to clay minerals in the diet. Journal of Toxicology and Environmental Health: Part A 67:863874. Williams JH, Phillips TD, Jolly PE, Stiles JK, Jolly CM, Aggarwal D. 2004. Human aflatoxicosis in developing countries: A reviews of toxicology, exposure, potential health consequences, and interventions. American Journal of Clinical Nutrition 80:11061122. Wogan GN. 1992. Molecular epidemiology in cancer risk assessment and prevention: Recent progress and avenues for future research. Environmental Health Perspectives 11:4754. World Health Organization (WHO). 1998. Safety evaluation of certain food activities and contaminants. WHO Food Additives Series No. 40 Prepared by the Forty-ninth Meeting of the Joint FAO/WHO Expert Committee on Food Additives. Geneva: WHO, International Programme on Chemical Safety.

Food Additives and Contaminants, February 2008; 25(2): 146151

Impact of mycotoxins on human health in developing countries

G. S. SHEPHARD
PROMEC Unit, Medical Research Council, PO Box 19070, Tygerberg 7505, South Africa
(Received 26 April 2007; accepted 10 July 2007)

Abstract Adverse human health effects from the consumption of mycotoxins have occurred for many centuries. Although mycotoxin contamination of agricultural products still occurs in the developed world, the application of modern agricultural practices and the presence of a legislatively regulated food processing and marketing system have greatly reduced mycotoxin exposure in these populations. At the mycotoxin contamination levels generally found in food products traded in these market economies, adverse human health effects have largely been overcome. However, in the developing world, where climatic and crop storage conditions are frequently conducive to fungal growth and mycotoxin production, much of the population relies on subsistence farming or on unregulated local markets. The extent to which mycotoxins affect human health is difficult to investigate in countries whose health systems lack capacity and in which resources are limited. Aflatoxin B1, the toxin on which major resources have been expended, has long been linked to liver cancer, yet its other effects, such as immune suppression and growth faltering previously observed in veterinary studies, are only now being investigated and characterized in human populations. The extent to which factors such as immune suppression contribute to the overall burden of infectious disease is difficult to quantify, but is undoubtedly significant. Thus, food safety remains an important opportunity for addressing current health problems in developing countries.

Keywords: Aflatoxin, health, risk assessment, mycotoxin, fumonisin, cancer, developing countries, Africa, aflatoxicosis

Introduction Although only recognized in recent times as a source of ill-health in humans, the agricultural problems associated with contamination of crops by fungal action have been noted for over two millennia (Campbell 1996). Consequently, human mycotoxicoses have probably also existed since the development of settled agricultural communities reliant on grain stores. Recent attempts to interpret the Biblical plagues of ancient Egypt over three millennia ago have suggested that the tenth plague, mentioned in the Book of Exodus and involving the death of the eldest sons, was due to macrocyclic trichothecene mycotoxins (possibly from Stachybotrys atra) occurring in the stored grains the eldest being the first to access these stores (Marr & Malloy 1996). During the Middle Ages in Europe a common affliction known as St Anthonys Fire was prevalent and caused thousands of deaths over a period of many centuries (Marasas & Nelson 1987a).
Correspondence: G. S. Shephard. E-mail: gordon.shephard@mrc.ac.za ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701567442

More recent understanding has identified this condition as ergotism, produced by the ingestion of the ergots of Claviceps purpurea, a fungus occurring on staple food grains such as rye and wheat. It has also been suggested that the bewitchment displayed by persons in the then British colony of Massachusetts in North America in 1692 and which lead to the Salem witchcraft trials and executions were a result of convulsive ergotism (Woolf 2000). The deaths during the Second World War of thousands in the former Soviet Union from the haemorrhagic syndrome known as alimentary toxic aleukia, caused primarily by T-2 toxin produced by Fusarium sporotrichioides and F. poae contaminating cereal overwintered in fields (Marasas & Nelson 1987b), and the discovery in 1960 of aflatoxins, produced by Aspergillus flavus and A. parasiticus, focused attention on the adverse human health implications of the secondary metabolites of fungi.

Mycotoxins and health in developing countries Based on their known and suspected effects on human and animal health and their production by fungal pathogens of staple food crops, aflatoxin, fumonisin, deoxynivalenol (DON), ochratoxin A (OTA) and zearalenone (ZON)) are recognized as the five most important agricultural mycotoxins. A vast literature has arisen concerning aspects of the toxicology, chemical analysis and natural occurrence of these toxins. This has enabled risk assessments to be undertaken in various countries, which have highlighted the divide between the developed and developing worlds with respect to food safety. Firstly, mycotoxin exposures are generally lower in developed countries. The reasons for this are numerous. In the developed world, human diets are extremely varied, commercial food suppliers in these market economies compete with quality to meet the highest standards, while legislated maximum tolerated levels for mycotoxins are widely enforced and cover a majority of the population accessing food from retail markets. By contrast, diets consumed by the population of developing countries tend to be less varied and, when not grown on a subsistence basis, food items may be obtained from local markets with less attention to quality issues. In addition, there is less emphasis on legislating maximum tolerated levels and even when such legislation exists, the capacity to enforce it is frequently lacking. Food grown in these areas is frequently consumed irrespective of quality due to food scarcity problems. The problem of excessive consumption of a single cereal can readily be seen in many African diets which rely on maize consumed at levels between 400 and 500 g per person day1. Even moderate levels of mycotoxin contamination can result in exposure which exceeds the maximum tolerable daily intake (TDI) set by the Joint FAO/WHO Expert Committee on Food Additives (JECFA). In the case of fumonisin, a 60 kg person consuming 500 g maize day1 would exceed the provisional maximum TDI of 2 mg kg1 body weight day1 set by JECFA if the fumonisin contamination were to exceed 240 mg kg1 (Bolger et al. 2001). Yet contamination levels in home-grown subsistence maize frequently exceeds this level. By contrast, maize consumption within the countries of the European Union is approximately two orders of magnitude lower and fumonisin exposure is unlikely to approach the TDI (Bolger et al. 2001). A further comparison of developing and developed countries reveals that for a human carcinogen such as aflatoxin B1 (AFB1), the carcinogenic potency used in calculating the population cancer risk is greater in developing countries. This is a consequence of AFB1 being synergistic with hepatitis B virus (HBV) infection, which has a greater prevalence in the developing world. The overall potency

147

of AFB1 in risk characterization calculations is adjusted to reflect the proportion of hepatitis B surface antigen-positive individuals in the population (Henry et al. 1998). Thus for both parameters involved in risk characterization, namely population exposure and potency, values reflecting the population in developing countries are higher than typical values for developed countries.

Recognized effects on human health The improvements in food safety in developed countries mentioned above have eliminated acute human mycotoxicoses such as ergotism, which was previously well known in the Middle Ages in Western Europe. However, such outbreaks still occur in rural communities in the developing world, as evidenced by documented cases in Ethiopia, East Africa, where there were outbreaks of gangrenous ergotism in 1978 after consumption of grain contaminated with Claviceps purpurea (Demeke et al. 1979; King 1979). The most tragic recent outbreaks of human mycotoxicosis have happened in Keny, where deaths due to aflatoxin exposure have occurred over a number of years. Aflatoxicosis is a toxic hepatitis leading to jaundice and, in severe cases, death. Repetitive incidents of this nature have occurred in Kenya during 1981, 2001, 2004 and 2005 (Shephard 2004; Lewis et al. 2005). During January to July 2004, in the eastern and central districts of Kenya, 317 cases were admitted to hospital with jaundice. Of these, 125 deaths were recorded. Maize collected from the affected areas had high AFB1 levels with 55% of samples contaminated above the Kenyan legal limit of 20 mg kg1, 35% had levels above 100 mg kg1, and 7% above 1000 mg kg1. The maximum level found was as high as 8000 mg kg1. A similar outbreak in the eastern districts of Kenya during 2005 led to 75 cases being admitted to hospital (as of 21 June 2005) and 32 deaths (Centers for Disease Control (CDC) 2006). Of the maize samples collected in a survey of the affected region, 42% were contaminated with AFB1 levels above 20 mg kg1, 24% were above 100 mg kg1, and 7% above 1000 mg kg1. Acute human aflatoxicosis has also been reported from Asian countries such as India (Bhat 1991) and Malaysia (Lye et al. 1995). AFB1 has been extensively linked to human primary liver cancer in which it acts synergistically with HBV infection and was classified by the International Agency for Research on Cancer (IARC) as a human carcinogen (group 1 carcinogen) (IARC 1993a). This combination represents a heavy cancer burden in developing countries. A recent comparison of the estimated population risk between

148

G. S. Shephard of human disease, mostly in developing countries. A number of occurrences of acute food-borne illness in India and China involving gastrointestinal symptoms have been attributed to the consumption of DON-contaminated cereals (Luo 1988; Bhat et al. 1989). OTA has long been associated with Balkan endemic nephropathy (BEN), a fatal renal disease with histopathological similarities to OTA-induced nephropathy in swine, and has been associated with the incidence of epithelial tumours of the upper urinary tract (Benford et al. 2001; Castegnaro et al. 2006). OTA was classified by the IARC as possibly carcinogenic to humans (group 2B carcinogen) (IARC 1993b). ZON is a naturally occurring endocrine-disrupting chemical and has been associated with clinical manifestations of hyper-oestrogenism in humans, including an outbreak of precocious pubertal changes in young children in Puerto Rico in the Caribbean (Saenz de Rodrigues et al. 1985) and gynecomastia with testicular atrophy in rural males in southern Africa (Campbell 1991).

Kenya and France highlighted the greater burden that can be placed on developing countries (Shephard 2006). Based on respective estimates for aflatoxin exposure of 133 and 0.12 ng kg1 body weight day1 and respective HBV prevalence of 25 and 1%, the liver cancer risk would be 11 vs. 0.0015 cancers per year per 100 000 population, respectively. Given recently published liver cancer incidence rates in the European Union of 10.0 per 100 000 for males and 3.3 per 100 000 for females (Bray et al. 2002), it is clear that aflatoxin plays a significant role in liver cancer in developing countries, but not in the developed world where other risk factors such as cirrhosis are more important. Fumonisins have been implicated in one incident of acute food-borne disease in India in which the occurrence of borborygmy, abdominal pain, and diarrhoea was associated with the consumption of maize and sorghum contaminated with high levels of fumonisins (Bhat et al. 1997). Fumonisin B1, the most abundant of the numerous fumonisin analogues, was classified by the IARC as a group 2B carcinogen (possibly carcinogenic in humans) (IARC 2002). Studies in the former Transkei region of South Africa and in Linxian and Cixian counties, China, have demonstrated an association between fumonisin exposure in rural subsistence farming areas and a high incidence of oesophageal cancer (Rheeder et al. 1992; Zhang et al. 1997). Fumonisins, which inhibit the uptake of folic acid via the folate receptor (Stevens and Tang 1997), have also been implicated in the high incidence of neural tube defects in rural populations known to consume contaminated maize, such as the former Transkei region of South Africa and areas of Northern China (Marasas et al. 2004). The other three agriculturally important mycotoxins have also been associated with various outbreaks

Emerging effects on human health The identified harmful effects of mycotoxin exposure on human health, as exemplified by acute aflatoxicosis and primary liver cancer, are increasingly being recognized as only the tip of the iceberg of morbidity associated with chronic mycotoxin exposure (Miller 1998; Williams et al. 2004). The WHO World Health Report 2002 (2002) identified the top ten health risks for developing countries with high mortality. These are listed in Table I, together with their associated disease burden as measured in disability-adjusted life years (DALYs) and the main type of affected outcome. The WHO report lists ten types of

Table I. Leading health risks for developing countries with high mortality, the corresponding contribution to the burden of disease in attributable disability-adjusted life years (DALYs) and the main type of affected outcome as given in the WHO World Health Report 2002 (2002). Leading health risks Underweight Unsafe sex Unsafe water, sanitation and hygiene Indoor smoke from solid fuels Zinc deficiency Iron deficiency Vitamin A deficiency Blood pressure Tobacco Cholesterol
1 2

Burden in DALYs1 14.9 10.2 5.5 3.6 3.2 3.1 3.0 2.5 2.0 1.9

Main component of outcome infectious and parasitic disease infectious and parasitic disease infectious and parasitic disease infectious and parasitic disease infectious and parasitic disease maternal and perinatal conditions infectious and parasitic disease cardiovascular disease cardiovascular disease2 cardiovascular disease

As a percentage of a total of 833 million DALYs. Infectious disease, cancer and chronic respiratory disease are also significant outcomes.

Mycotoxins and health in developing countries affected outcomes associated with a total loss of 833 million DALYs in developing countries with high mortality. Of these outcomes, infectious and parasitic diseases represent over 90% of the outcomes in the top five risk factors (underweight, unsafe sex, unsafe water, indoor smoke from solid fuels and zinc deficiency) and account for 300 million DALYs overall. Although mycotoxins are not specifically mentioned, they may be seen to play a modulating role in a certain number of these factors. According to the WHO estimate, being underweight caused 3.7 million deaths in 2000, mostly in children under five years of age in developing countries. The mortality and morbidity were due to the effect of poor nutrition on immune function, which led to diarrhoeal diseases, malaria, measles and pneumonia. Recently published results indicate that both being underweight and immune function are directly affected by aflatoxin exposure in developing countries. Studies in Benin and Togo, West Africa, have shown that aflatoxin exposure in infants, which increases at weaning, can lead directly to growth impairment and stunting (Gong et al. 2002). Aflatoxin has long been linked to kwashiorkor, a disease usually considered a form of protein energy malnutrition, although some characteristics of the disease are known to be among the pathological effects caused by aflatoxins in animals. It has been suggested that either aflatoxins could play a causal role in the disease (Hendrickse 1991) or children suffering from the disease are at greater risk to the hazards of dietary aflatoxin (Adhikari et al. 1994). Studies in Gambia and Ghana, also in West Africa, have begun to elucidate the role of aflatoxin in immune suppression in human populations. Aflatoxin exposure was associated with reduced levels of secretory immunoglobulin A (IgA) in Gambian children (Turner et al. 2003). Changes in differential subset distributions and functional alterations of specific lymphocyte subsets have been correlated with aflatoxin exposure in Ghanaian adults and indicate that aflatoxins could cause impairment of human cellular immunity that could decrease resistance to infections (Jiang et al. 2005). Of the other health risk factors, the morbidity and mortality associated with unsafe sex, unsafe water and indoor smoke, arises from infectious diseases, such as HIV/AIDS, infectious diarrhoea and lower respiratory tract infection, respectively. The immunological suppression associated with aflatoxin and possibly DON could adversely affect all these outcomes. The modulating effect of aflatoxins in cases of zinc, iron and vitamin A deficiency in human health is less clear, but evidence from animal nutrition would suggest it could be significant (Williams et al. 2004). Challenges and conclusions

149

Exposure to mycotoxins is a serious risk to human health, especially in developing countries where the effects of poverty and malnutrition lead to an exacerbation of the detrimental effects of these food-borne toxins by circumscribing biochemical detoxification mechanisms. Even where the health risks of mouldy grain to human health are recognized, the contaminated grain may be fed to livestock, decreasing animal productivity and the food supply, and increasing poverty (Wu et al. 2005). Apart from the fact that aflatoxin is a causative factor for primary liver cancer, strong epidemiological evidence for mycotoxins as causative factors of various diseases is still lacking. Nevertheless, exposure to these compounds should be addressed as an urgent food safety issue as they place a significant constraint on attempts to improve human health in developing countries. Measures to achieve the UN Millennium Development Goals are aimed at impacting directly on poverty in the developing world and hence on food choices and food safety. In addressing mycotoxins as a food safety problem, the difficulty arises in prioritizing this issue in communities in which food sufficiency has not been attained. In the search for a means to address the question, it is not clear whether appropriate technologies addressing specific subsistence farmer needs (Turner et al. 2005) or developed world technologies such as improved varieties, genetically modified organisms and agricultural modelling (Gressel et al. 2004) (or a combination of both) will be more effective in improving food safety. Finally, the effects of climate change on fungal distribution and toxin production may present a series of new food safety challenges.

References
Adhikari M, Ramjee G, Berjak P. 1994. Aflatoxin, kwashiorkor, and morbidity. Natural Toxins 2:13. Benford D, Boyle C, Dekant W, Fuchs R, Gaylor DW, Hard G, McGregor DB, Pitt JI, Plestina R, Shephard G, et al. Verger PJP, Walker R. 2001. Ochratoxin A. In: Safety evaluation of certain mycotoxins in food. WHO Food Additives Series No. 47, FAO Food and Nutrition Paper No. 74, Prepared by the 56th Meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA). Geneva: WHO. pp 281415. Bhat RV. 1991. Aflatoxins: Successes and failures of three decades of research. ACIAR Proceedings 36:7782. Bhat RV, Beedu SR, Ramakrishna Y, Munshi KL. 1989. Outbreak of trichothecene mycotoxicosis associated with consumption of mould-damaged wheat production in Kashmir Valley, India. Lancet i: 3537.

150

G. S. Shephard
contamination of commercial maize products during an outbreak of acute aflatoxicosis in eastern and central Kenya. Environmental Health Perspectives 113:17631767. Luo XY. 1988. Outbreaks of moldy cereals poisoning in China. In: Issues in food safety. Washington, DC: Toxicology Forum. pp 5663. Lye MS, Ghazali AA, Mohan J, Alwin N, Mair RC. 1995. An outbreak of acute hepatic encephalopathy due to severe aflatoxicosis in Malaysia. American Journal of Tropical Medicine and Hygiene 53:6872. Marasas WFO, Nelson PE. 1987a. Ergotism. Mycotoxicology. University Park, PA: Pennsylvania State University Press. pp 1922. Marasas WFO, Nelson PE. 1987b. Hemmorhagic syndrome. In: Mycotoxicology. University Park, PA: Pennsylvania State University Press. pp 4144. Marasas WFO, Riley RT, Hendricks KA, Stevens VL, Sadler TW, Gelineau-van Waes J, Missmer SA, Cabrera J, Torres O, Gelderblom WCA, et al. 2004. Fumonisins disrupt sphingolipid metabolism, folate transport, and neural tube development in embryo culture and in vivo: A potential risk factor for human neural tube defects among populations consuming fumonisin-contaminated maize. Journal of Nutrition 134:711716. Marr JS, Malloy CD. 1996. An epidemiologic analysis of the ten plagues of Egypt. Caduceus 12:724. Miller JD. 1998. Global significance of mycotoxins. In: Miraglia M, Van Egmond HP, Brera C, Gilbert J, editors. Mycotoxins and phycotoxins: developments in chemistry, toxicology and food safety. Fort Collins: Alaken. pp 315. Rheeder JP, Marasas WFO, Thiel PG, Sydenham EW, Shephard GS, Van Schalkwyk DJ. 1992. Fusarium moniliforme and fumonisins in corn in relation to human esophageal cancer in Transkei. Phytopathology 82:353357. Saenz de Rodrigues CA, Bongiovanni AM, Conde de Borrego L. 1985. An epidemic of precocious development in Puerto Rican children. Journal of Pediatrics 107:393396. Shephard GS. 2004. Mycotoxins worldwide: Current issues in Africa. In: Barug D, Van Egmond H, Lopez-Garcia R, Van Ossenbruggen T, Visconti A, editors. Meeting the mycotoxin menace. Wageningen: Wageningen Academic. pp 8188. Shephard GS. 2006. Mycotoxins in the context of food risks and nutrition issues. In: Barug D, Bhatnagar D, Van Egmond HP, Van der Kamp JW, Van Ossenbruggen WA, Visconti A, editors. The mycotoxin fact book. Wageningen: Wageningen Academic. pp 2136. Stevens VL, Tang J. 1997. Fumonisin B1-induced sphingolipid depletion inhibits vitamin uptake via the glycosylphosphatidylinositol-anchored folate receptor. Journal of Biological Chemistry 272:1802018025. Turner PC, Moore SE, Hall AJ, Prentice AM, Wild CP. 2003. Modification of immune function through exposure to dietary aflatoxin in Gambian children. Environmental Health Perspectives 111:217220. Turner PC, Sylla A, Gong YY, Diallo MS, Sutcliffe AE, Hall AJ, Wild CP. 2005. Reduction in exposure to carcinogenic aflatoxins by postharvest intervention measures in West Africa: A community-based intervention study. Lancet 365:19501956. Williams JH, Phillips TD, Jolly PE, Stiles JK, Jolly CM, Aggarwal D. 2004. Human aflatoxicosis in developing countries: A review of toxicology, exposure, potential health consequences, and interventions. American Journal of Clinical Nutrition 80:11061122. Woolf A. 2000. Witchcraft or mycotoxin? The Salem witch trials. Clinical Toxicology 38:457460.

Bhat RV, Shetty PH, Amruth RP, Sudershan RV. 1997. A foodborne disease outbreak due to the consumption of moldy sorghum and maize containing fumonisin mycotoxins. Journal of Toxicology Clinical Toxicology 35:249255. Bolger M, Coker RD, DiNovi M, Gaylor D, Gelderblom W, Olsen M, Paster N, Riley RT, Shephard G, Speijers GJA. 2001. Fumonisins. In: Safety Evaluation of Certain Mycotoxins in Food. WHO Food Additives Series 47, FAO Food and Nutrition Paper 74, Prepared by the 56th Meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA). Geneva, Switzerland: WHO. pp 103279. Bray F, Sankila R, Ferlay J, Parkin DM. 2002. Estimates of cancer incidence and mortality in Europe in 1995. European Journal of Cancer 38:99166. Campbell GD. 1991. Trichothecene mycotoxicosis B a new entity? South African Medical Journal 80:361362. Campbell GD. 1996. Mycotoxicosis humankinds greatest affliction? Nutrition and Health 10:323329. Castegnaro M, Canadas D, Vrabcheva T, Petkova-Bocharova T, Chernozemsky IN, Pfohl-Leszkowicz A. 2006. Balkan endemic nephropathy: Role of ochratoxin A through biomarkers. Molecular Nutrition and Food Research 50:519529. Centers for Disease Control (CDC). 2006. Aflatoxicosis Outbreak, Kenya 2005. Report of outbreak investigation presented by the Centers for Disease Control and Prevention and its partners to the Kenya Ministry of Health. Demeke T, Kidane Y, Wuhib E. 1979. Ergotism: A report on an epidemic, 197778. Ethiopian Medical Journal 17:107113. Gong YY, Cardwell K, Hounsa A, Egal S, Turner PC, Hall AJ, Wild CP. 2002. Dietary aflatoxin exposure and impaired growth in young children from Benin and Togo: Cross sectional study. British Medical Journal 325:2021. Gressel J, Hanafi A, Head G, Marasas W, Obilana AB, Ochanda J, Souissi T, Tzotzos G. 2004. Major heretofore intractable biotic constraints to African food security that may be amenable to novel biotechnological solutions. Crop Protection 23:661689. Hendrickse RG. 1991. Clinical implications of food contaminated by aflatoxins. Annals of the Academy of Medicine 20:8490. Henry S, Bosch FX, Bowers JC, Portier CJ, Petersen BJ, Barraj L. 1998. Aflatoxins. In: Safety evaluation of certain food additives and contaminants. Prepared by the 49th Meeting of the Joint FAO/WHO Expert Committee on Food Additives. Geneva: WHO. pp 359468. International Agency for Research on Cancer (IARC). 1993a. Some naturally occurring substances: Food items and constituents, heterocyclic aromatic amines and mycotoxins. Aflatoxins. WHO IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 56. Lyon: IARC. pp 245395. International Agency for Research on Cancer (IARC). 1993b. Some naturally occurring substances: Food items and constituents, heterocyclic aromatic amines and mycotoxins. Ochratoxin A. WHO IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 56. Lyon: IARC. pp 489521. International Agency for Research on Cancer (IARC). 2002. Some traditional herbal medicines, some mycotoxins, naphthalene and styrene. Fumonisin B1. WHO IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 82. Lyon: IARC. pp 301366. Jiang YI, Jolly PE, Ellis WO, Wang J-S, Phillips TD, Williams JH. 2005. Aflatoxin B1 albumin adduct levels and cellular immune status in Ghanaians. International Immunology 17:807814. King B. 1979. Outbreak of ergotism in Wollo, Ethiopia. Lancet 1:1411. Lewis L, Onsongo M, Njapau H, Schurz-Rogers H, Luber G, Kieszak S, Nyamongo J, Backer L, Dahiye, AM, Misore A, et al., Kenya Aflatoxicosis Investigation Group. 2005. Aflatoxin

Mycotoxins and health in developing countries

World Health Organization (WHO). 2002. The world health report 2002. Reducing risks promoting healthy life. Available: http://www.who.int/whr/en/. Accessed 30 March 2007. Wu F, Miller JD, Casman EA. 2005. Bt corn and mycotoxin reduction: An economic perspective. In: Abbas HK, editor.

151

Aflatoxin and food safety. New York, NY: Taylor & Francis. pp 459482. Zhang H, Nagashima H, Goto T. 1997. Natural occurrence of mycotoxins in corn samples from high and low risk areas for human esophageal cancer in China. Mycotoxins 44:2935.

Food Additives and Contaminants, February 2008; 25(2): 152163

Mycotoxin analysis: An update

RUDOLF KRSKA1, PATRICIA SCHUBERT-ULLRICH1, ALEXANDRA MOLINELLI1, MICHAEL SULYOK1, SUSAN MACDONALD2, & COLIN CREWS2
1

Christian Doppler Laboratory for Mycotoxin Research, Center for Analytical Chemistry, Department for Agrobiotechnology (IFA Tulln), University of Natural Resources and Applied Life Sciences Vienna, Konrad Lorenz Strasse 20, A-3430 Tulln, Austria and 2Central Science Laboratory (CSL), Sand Hutton, York YO41 1LZ, UK

(Received 10 July 2007; accepted 17 October 2007)

Abstract Mycotoxin contamination of cereals and related products used for feed can cause intoxication, especially in farm animals. Therefore, efficient analytical tools for the qualitative and quantitative analysis of toxic fungal metabolites in feed are required. Current methods usually include an extraction step, a clean-up step to reduce or eliminate unwanted co-extracted matrix components and a separation step with suitably specific detection ability. Quantitative methods of analysis for most mycotoxins use immunoaffinity clean-up with high-performance liquid chromatography (HPLC) separation in combination with UV and/or fluorescence detection. Screening of samples contaminated with mycotoxins is frequently performed by thin layer chromatography (TLC), which yields qualitative or semi-quantitative results. Nowadays, enzyme-linked immunosorbent assays (ELISA) are often used for rapid screening. A number of promising methods, such as fluorescence polarization immunoassays, dipsticks, and even newer methods such as biosensors and non-invasive techniques based on infrared spectroscopy, have shown great potential for mycotoxin analysis. Currently, there is a strong trend towards the use of multi-mycotoxin methods for the simultaneous analysis of several of the important Fusarium mycotoxins, which is best achieved by LCMS/MS (liquid chromatography with tandem mass spectrometry). This review focuses on recent developments in the determination of mycotoxins with a special emphasis on LCMS/MS and emerging rapid methods.

Keywords: Mycotoxin, analysis, LCMS/MS, sample preparation, chromatographic methods

Introduction Mycotoxins are natural, secondary metabolites produced by fungi on agricultural commodities in the field and during storage under a wide range of climatic conditions. About 200 different filamentous fungi species, e.g. Aspergillus, Penicillium and Fusarium species (sp.), have been identified. Several hundred different mycotoxins have been discovered so far, exhibiting great structural diversity, which results in different chemical and physicochemical properties. Aflatoxins and ochratoxins (produced mainly by Aspergillus sp.), fumonisins, trichothecenes and zearalenone (produced by Fusarium sp.), patulin (produced by Penicillium sp.), and ergot alkaloids (produced in the sclerotia of Claviceps sp.) receive the most attention due to
Correspondence: Rudolf Krska. E-mail: rudolf.krska@boku.ac.at ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701765723

their frequent occurrence and their severe effects on animal and human health (Bennett and Klich 2003; DMello and MacDonald 1997). Mycotoxins are potent toxins and have a wide range of actions on animals and humans, e.g. cyto-, nephro- and neurotoxic, carcinogenic, mutagenic, immunosuppressive and estrogenic effects. Although mycotoxicoses caused by direct consumption of contaminated food and feedstuffs poses the greatest risk to animals and humans, the entry of mycotoxins or their metabolites into the food chain by carry over into milk, animal tissue or eggs, for example, should not be underestimated. National and international institutions and organisations, such as the European Commission (EC), the US Food and Drug Administration (FDA), the

Mycotoxin analysis: An update World Health Organisation (WHO) and the Food and Agriculture Organisation (FAO) of the United Nations, have recognized the potential health risks to animals and humans posed by food- and feed-borne mycotoxin intoxication and addressed this problem by adopting regulatory limits for major mycotoxin classes and selected individual mycotoxins. The FAO has compiled comprehensive worldwide regulations and directives regarding mycotoxins in food and feed as of December 2003 (FAO 2004). The Joint Expert Committee on Food Additives (JECFA), a scientific advisory body of FAO and WHO, provides mechanisms for assessing the toxicity of food additives, veterinary drug residues and contaminants, and has recently evaluated the hazards related to several mycotoxins, including fumonisins B1, B2 and B3, ochratoxin A, deoxynivalenol, T-2 toxin, HT-2 toxin, and aflatoxin M1 (WHO 2002). The report explains the nature of each toxin, including its absorption and excretion, as well as toxicological studies, and it includes general considerations of analytical methods, sampling, associated intake issues and control mechanisms. The EC has set maximum levels for some mycotoxins, including several aflatoxins, ochratoxin A, patulin, deoxynivalenol and zearalenone, in certain foodstuffs. Maximum levels for fumonisins B1 and B2 came into force in October 2007. Consideration of a review of the maximum levels for deoxynivalenol, zearalenone and fumonisins B1 and B2 as well as the appropriateness of setting a maximum level for T-2 and HT-2 toxins in cereals and cereal products should be completed by July 2008 (EC 2006). The requirement to apply these regulatory limits has prompted the development of a vast number of analytical methods for the identification and quantification of mycotoxins in various samples, such as food, feed, and other biological matrices. The chemical diversity of mycotoxins and their varying concentration ranges in a wide range of agricultural commodities, foods and biological samples poses a great challenge to analytical chemists. The different chemical and physicochemical properties of the mycotoxins require specific extraction, cleanup, separation and detection methods. Therefore, most methods target only individual mycotoxins or at best a group of closely related mycotoxins. These methods are usually based on labour-intensive sample preparation protocols followed by traditional chromatographic separation (mostly liquid chromatography, LC). Gas chromatography (GC) either with electron capture detection (ECD) or mass spectrometric (MS) detection is used in mycotoxin analysis, e.g. for trichothecene or patulin determination, but less frequently than alternative methods. In some cases, fast and accurate screening methods

153

based on enzyme-linked immunosorbent assay (ELISA) are applied instead of the more labourintensive LC methods. Thin-layer chromatography (TLC) provides a cheaper alternative to LC-based methods and has an important role, especially in developing countries, for surveillance purposes and control of regulatory limits (Gilbert and Anklam 2002). Modern sample clean-up techniques, such as immunoaffinity columns (IAC) or solid-phase extraction (SPE) methods, help to simplify protocols, improve selectivity and, thus, performance characteristics. To deal with the increasing number of sample matrices and mycotoxins of interest, fast and accurate analytical methods are needed. This demand has led to the development of rapid screening methods for single mycotoxins or whole mycotoxin classes based on immunochemical techniques (e.g. ELISA), biosensors (e.g. protein chips, antibody/protein-coated electrodes) and non-invasive optical techniques. On the other hand, highly sophisticated multi-mycotoxin methods based on LC coupled to multiple-stage MS are being developed to allow accurate and precise determination and unambiguous identification of mycotoxins without the need for tedious sample preparation and clean-up procedures. Sample selection and representative sample collection are often underestimated as sources of error. The design of sampling procedures for various mycotoxins and sample materials has been an international concern for several years. (FAO 2004; EC 2006b; FDA 2007). To obtain comparable data, the EC has laid down certain requirements for sampling and performance criteria for analytical methods (EC 2006b). Therefore, the whole analytical method (including sampling, sample preparation, clean-up and final determination) used by enforcement laboratories for the implementation and control of legislation and regulatory limits must be subject to a validation procedure to show that the method produces reliable results and meets the set performance criteria. Several protocols and guidelines for method validation have been published (Thompson 1993; ISO 1994; Eurachem 1998). There are a multitude of analytical methods available that have been validated and accepted by official authorities, such as the European Committee for Standardization (CEN), the Association of Official Analytical Chemists (AOAC International), and the International Organisation for Standardization (ISO) (Gilbert and Anklam 2002, AOAC 2005). Each laboratory should implement quality assurance measures such as frequently checking the accuracy and precision of their methods by analysing (certified) reference materials (CRM) and by regular participation in proficiency

154

R. Krska et al. determination in dried figs, groundnuts and nuts. However, it has to be considered that handling and sample preparation of large quantities pose great difficulties in implementing such sampling plans, especially with regard to the validation process of sampling methods. The EC regulation, therefore, stipulates the division of aggregate samples intended for direct human consumption into up to three laboratory samples of 10 kg for homogenisation and analysis. This subdivision is not necessary for products intended for further sorting or processing before human consumption or for use as an ingredient (SANCO 2005). The Grain Inspection, Packers and Stockyards Administration (GIPSA) of the United States Department of Agriculture (USDA) developed general sampling guidelines for grain (GIPSA 1995), rice (GIPSA 1994) and hops (GIPSA 1998). The FDA is constantly updating their Investigative Operation Manual (IOM) that describes general procedures for field investigators and inspectors and includes information on sampling and sampling schedules for various occasions (FDA 2007). It contains a detailed sampling schedule for mycotoxin analysis that lists sample sizes dependent on the type of product and distinguishes between samples taken for surveillance (initial sample) and follow-up samples in case of positive findings. The minimum total sample size ranges from 4.5 kg up to 34 kg, depending on the heterogeneity expected of the sample type. Kay (2001) compared a number of approaches for sampling grain, developed by different national and international authorities, and provided an interpretation of the major differences between the methods in terms of lot size, tolerances, sampling techniques, sample size, rates, etc.

testing trials. Although much has been done in the past years regarding the production and certification of reference materials (RM) and calibrants for mycotoxin analysis in various matrices, there is still a need for more RMs appropriate for the different sample matrices and concentration ranges encountered in foods and feeds. The objective of this review is to summarize recent developments in the determination of mycotoxins with a special emphasis on LCMS/MS and emerging rapid methods.

Sampling Sampling plays a crucial part in the precision and determination of mycotoxin levels due to the sometimes very heterogeneous distribution of the toxins in agricultural commodities and products intended for human and animal consumption. The distribution of mycotoxins in the sample material is an important factor to be considered in establishing regulatory sampling criteria. In the past, this has been recognized by many national and international authorities and organisations, such as the EC (2006a), the FDA (2007) and the FAO (2004). In Commission Regulation No. 401/2006 (EC 2006b), the EC has laid down the methods of sampling and analysis for the official control of levels of various mycotoxins in foodstuffs, repealing all former directives and amendments on this subject. It includes methods of sampling and analysis of aflatoxins, ochratoxin A, patulin and Fusarium toxins in cereal, dried fruit, fruit juices, must and wine, groundnuts and nuts, spices, milk, coffee, products derived from the above mentioned basic materials as well as baby foods and food for infants and young children. While it can be assumed that mycotoxins in liquid samples are homogeneously distributed, some mycotoxins, especially in funguscontaminated grain, may be concentrated in socalled hot-spots. Mycotoxins, especially those produced by Aspergillus sp., e.g. aflatoxins, can be distributed very heterogeneously in food products with large particle size such as dried figs or groundnuts. The number of contaminated particles may be very low, but the contamination level within a particle can be very high. To obtain the same representativeness for batches of food products with large particle sizes, the weight of the incremental sample taken has to be larger than in cases of batches with smaller particle size. Commission Regulation No. 401/2006 regulates the number of incremental samples to be taken from different places of a lot depending on the weight of the entire lot. This may result in rather large aggregate samples, up to 30 kg in the case of aflatoxin

Sample preparation and clean-up Only a few analytical techniques, i.e. optical techniques based on IR spectroscopy (Kos et al. 2003), are capable of detecting mycotoxin contamination directly in ground cereal samples without the necessity of further sample preparation, such as solvent extraction or clean-up. However, the application of such techniques is still limited to screening purposes due to a high matrix dependence and lack of appropriate calibration materials. Analytical methods based on chromatography or immunoassays usually require solvent extraction to liberate the mycotoxin from the sample matrix, and subsequent clean-up of the extract to reduce matrix effects. Various combinations of solvents, sometimes with the addition of modifiers (e.g. acids, bases, etc.), are used for extraction, depending on the

Mycotoxin analysis: An update physicochemical properties of the mycotoxins, the sample matrix and the type of clean-up used afterwards (Zollner and Mayer-Helm 2006). Accelerated solvent extraction (ASE), also known as pressurised liquid extraction (PLE) (Royer et al. 2004; Urraca et al. 2004; Juan et al. 2005; Pallaroni and van Holst 2003, 2004) or microwave-assisted extraction (MAE) (Pallaroni et al. 2002) help to speed-up and automate the extraction process, and offer a robust and time-saving alternative to classical solvent extraction techniques. So far, the high cost of an ASE apparatus has, however, limited the application of this technique in the field of mycotoxin analysis to a few laboratories. Supercritical fluid extraction (SFE), especially with supercritical CO2 as an environmentally safe extraction medium, received a lot of attention in the 1990s. The extraction selectivity of the non-polar supercritical CO2 is influenced by temperature and pressure and can be varied in a wide range by adding modifiers (polar solvents, complexing agents, etc.). Although in the past, SFE has received much attention regarding agricultural applications, only a few papers deal with it as an extraction method in mycotoxins analysis (Huopalahti et al. 1997; Krska 1998; Ambrosino et al. 2004; Liau et al. 2007). Liquidliquid partitioning of the mycotoxin containing aqueous acetonitrile/methanol sample extract with hexane is sometimes used for de-fatting or protein precipitation (Srensen and Elbk 2005; Kokkonen et al. 2005). For further purification and analyte enrichment, liquid samples and extracts are predominantly submitted to solid-phase extraction (SPE) protocols for which a wide variety of sorbent materials are available. A comprehensive compilation of different clean-up approaches for various mycotoxins has been published by Zollner and Mayer-Helm (2006). Conventional SPE procedures use reversed-phase (RP) materials (e.g. C8, C18), strong cation or anion exchangers (SCX, SAX) or polymeric materials with combined properties. Modern clean-up procedures employ multifunc tional MycoSep (Krska 1998; Radova et al. 1998; Biselli and Hummert 2005; Ren et al. 2007) or immunoaffinity columns (IAC) (Krska 1998), although these methods are more expensive than conventional clean-up methodologies. MycoSep columns contain a mixture of charcoal, ion-exchange resins and other materials and are suitable for aflatoxins, trichothecenes, ochratoxins, zearalenone, moniliformin and patulin (Romerlabs 2007). Mycotoxin specific molecularly imprinted polymers (MIPs) are also considered as a potential and cheaper alternative for clean-up, which, contrary to IACs, do not suffer from storage limitations and stability problems regarding organic solvents. MIPs have been developed with recognition

155

properties towards several mycotoxins including deoxynivalenol (Weiss et al. 2003), zearalenone (Weiss et al. 2003; Urraca et al. 2006a, b), ochratoxin A (Baggiani et al. 2001; Jodlbauer et al. 2002; Maier et al. 2004; Turner et al. 2004) and moniliformin (Appell et al. 2007). Currently, there is a strong trend towards the use of IACs in mycotoxins analysis as a clean-up and enrichment technique for sample extracts or liquid samples. IACs contain immobilised antibodies that exclusively retain a certain mycotoxin or mycotoxin class. Due to their high specificity, IACs produce cleaner extracts with a minimum level of interfering matrix components and excellent signal-to-noise ratios compared to less selective SPE sorbent materials. IACs have been developed for most major mycotoxins and mycotoxin classes such as aflatoxins, ochratoxin A, trichothecenes, zearalenone and their metabolites (Zollner and Mayer-Helm 2006). The AOAC International and the EC have already validated a few IAC methods; however, these address only a limited number of food commodities. For some mycotoxins, such as ochratoxin A, IACs are already used in routine analysis, e.g. coupled with LC with fluorescence detection (FLD). When comparing both conventional clean-up and IAC approaches in the analysis of selected mycotoxins (aflatoxins, B-fumonisins, and ochratoxin A), discrepancies are found for certain food and feed matrices (Castegnaro et al. 2006; Sugita-Konishi et al. 2006). These problems highlight the necessity to validate methods for each complex matrix separately to provide reliable, comparable and traceable analytical data. Careful selection of the clean-up method is, however, essential for the effectiveness of an analytical method. Immunoaffinity materials are expensive and distinctly less feasible for multitoxin analysis since they are highly specific for only one target mycotoxin (or class). Some scientists even talk about overkill when using highly specific clean-up techniques, such as IACs in combination with liquid chromatography with mass spectrometry (LC/MS), since compound-specific detection stands in contradiction to the multi-analyte detection capabil ities of MS (Leitner et al. 2002; Zollner et al. 1999). However, there are already combined immunoaffinity materials on the market that are specific to a wider range of mycotoxins (MacDonald et al. 2007). It has been shown that, in many cases, the quality of the analytical result does not suffer when conventional SPE approaches are used (Leitner et al. 2002; Reinsch et al. 2005). Of course, this also depends on the selectivity of the MS equipment itself. Single-stage MS in selected ion-monitoring mode might need selective clean-up to remove matrix interferences, while those interferences

156

R. Krska et al. conventional detection techniques, such as UV or fluorescence, mass spectrometry offers increased selectivity and sensitivity (although fluorescence detection might be more sensitive for certain mycotoxins, e.g. aflatoxins), unambiguous confirmation of the molecular identity of the analyte and the option to use isotopically labelled substances as internal standards. Furthermore, it is possible to investigate the molecular structure of metabolites and sugar conjugates (such as masked mycotoxins; Berthiller et al. 2005b) and to omit timeconsuming and error-prone derivatization and clean-up steps. However, it must be kept in mind that a reduction of the sample preparation inevitably emphasizes the Achilles heel of LC/MS, i.e. relatively poor method accuracy and precision due to the irreproducible and unpredictable influence of co-eluting matrix components on the signal intensity of the analytes. Due to the large number of LC/MS-based methods for the quantitative determination of single mycotoxin classes, their exhaustive examination goes beyond the scope of this work and, therefore, the interested reader is referred to the reviews of Zollner and Mayer-Helm (2006) and Sforza et al. (2006).

might not be visible with multi-stage MS in selected reaction-monitoring mode (Zollner and MayerHelm 2006). However, matrix-induced signal suppression or enhancement should always be taken into consideration and can normally be omitted by clean-up of the extract or by using an appropriate calibration method (e.g. matrix-matched calibration standards, standard addition, or the use of adequate internal standards, i.e. isotope-labelled standards, etc.).

Analytical techniques Conventional analytical techniques The term conventional method usually refers to a chromatographic separation coupled to a suitable detection system. The currently used quantitative methods for the determination of regulated mycotoxins, such as the fumonisins, zearalenone, type-A (e.g. T2-toxin) and -B trichothecenes (e.g. deoxynivalenol), ochratoxin A and the aflatoxins, in food and feed mainly use immunoaffinity clean-up with high-performance liquid chromatography (HPLC) or gas chromatography (GC) in combination with a variety of detectors, such as fluorescence detection (FLD) with either a pre- or post-column derivatisation step, UV detection, flame ionisation detection (FID), electron capture detection (ECD) or mass spectrometry (MS). Reviews of these methods have been summarized and published elsewhere (Krska et al. 2001, 2005; Krska and Josephs 2001). From the multitude of available procedures, CEN is trying to standardize methods for mycotoxin analysis. CEN establishes performance criteria for mycotoxin methods usually on the basis of collaborative studies. CEN methods are official reference methods and are used for official control and surveillance and in cases of dispute. CENapproved methods exist for aflatoxins, ochratoxin A, fumonisins, patulin and deoxynivalenol, for example, in various foods. Further methods for various mycotoxins in feed will be issued in the near future (Gilbert and Anklam, 2002). Liquid chromatography/mass spectrometry (LC/MS) Within the last 10 years, liquid chromatography/ mass spectrometry has become the universal approach for mycotoxin analysis, as more or less all potential analytes are compatible with the conditions applied during separation and detection. Nevertheless, the breakthrough of this approach did not occur until the mid-1990s, when suitable interfaces, such as atmospheric pressure ionization, became accessible on a routine basis. Compared to

Multi-mycotoxin methods In the last few years, increased efforts have been made to develop analytical methods for the simultaneous determination of different classes of mycotoxins using LCMS/MS. This trend is a result of the discovery of co-occurrence of different toxins and related synergistic toxic effects that raise concerns about the health hazard from contaminated food and feed (Creppy et al. 2004; Speijers and Speijers 2004). In addition, it would be desirable to cover the toxins addressed by Commission Regulation 1881/ 2006 (aflatoxin B1, B2, G1, G2 and M1, ochratoxin A, patulin, deoxynivalenol, zearalenone, fumonisin B1 and B2, HT-2 and T-2 toxin) with a single method as this increases sample throughput and decreases the costs per analysis. Although mass spectrometry offers sufficient selectivity (especially if tandem-mass spectrometry is applied) and multianalyte capabilities, its realization in the field of multi-mycotoxin analysis has been hampered mainly by the chemical diversity of the different toxin classes, which include acidic (fumonisins), basic (ergot alkaloids) as well as polar (moniliformin, nivalenol) and apolar (zearalenone, beauvericin) compounds. Therefore, compromises have to be made in the choice of extraction solvent and mobile phase, and the conditions may be far from optimal for certain analytes.

Mycotoxin analysis: An update The initial stimulus for LC/MS-based multimycotoxin methods came from the field of mycology, where mass spectrometry is used to identify mould species according to their metabolite profile (Smedsgaard and Frisvad 1996). Beside the development of databases dealing with qualitative LC/MS of mycotoxins (Nielsen and Smedsgaard 2003), this has led to early quantitative methods for the simultaneous determination of Aspergillus and Penicillium mycotoxins in building materials (Tuomi et al. 2001) and in an artificial food matrix (Rundberget and Wilkins 2002). While the former method suffered from low recoveries of some analytes, excellent accuracy and precision were obtained in the latter case through use of a de-fatting step applied to the raw extract, and use of matrixmatched calibration to compensate for matrix effects. Some years later, this method was applied for the simultaneous determination of aflatoxins, ochratoxin A, mycophenolic acid, penicillic acid and roquefortine C (Kokkonen et al. 2005) after a slight modification of the extraction solvent. After this initial phase, the focus of multimycotoxin analysis shifted to Fusarium mycotoxins. Royer et al. (2004) developed a method for the quantitative analysis of deoxynivalenol, fumonisin B1 and zearalenone in maize, including accelerated solvent extraction, a two-step SPE procedure and internal standards for each analyte. The LODs were below the maximum concentration levels permitted in the EU, but the method suffered from a low recovery for zearalenone. The next generation of methods included several A- and B- trichothecenes as well as zearalenone, and used Mycosep columns for clean-up of the raw extracts. Zearalanone was used as internal standard for zearalenone in the method of Berthiller et al. (2005a), and Biselli and Hummert (2005) applied matrix matched calibration for this analyte. Cavaliere et al. (2005) added -zearalenol and three fumonisins to the list of analytes and performed de-fatting and solid-phase extraction of the raw extracts of corn meal. While the efficiency of the extraction step was greater than 84% for all analytes, matrix effects were still present and required matrix-matched calibration. A method for the simultaneous determination of Fusarium, Aspergillus and Penicillium toxins (ochratoxin A, zearalenone, - and -zearalenol, - and -zearalanol, fumonisins B1 and B2, T2- and HT2-toxin, T2-triol, mono- and diacetoxyscirpenol, deoxynivalenol, 3- and 15-acetyldeoxynivalenol, deepoxy-deoxynivalenol and aflatoxin M1) was reported by Sorensen and Elbk (2005). Bovine milk samples were defatted and after adjustment of the pH, an SPE procedure was applied. Signal suppression/enhancement was minimized and recoveries 476% were obtained. However, the major

157

drawback of this method was the necessity of using two chromatographic runs with different columns and eluents. The two most recent reports, which include a clean-up of the raw extract using MultiSep #226 cartridges, introduced instrumental improvements to multi-mycotoxin analysis in foodstuffs. A time-of-flight mass spectrometer was used by Tanaka et al. (2006), while, in the method of Ren et al. (2007), analysis time was significantly decreased through the application of ultra-performance liquid chromatography. In both methods, recoveries 470% were obtained for all analytes and no significant matrix effects were reported. As all these methods rely on some sort of clean-up, certain toxin classes are excluded as they are not compatible with the clean-up and/or extraction conditions (for example, the fumonisins are not determined by the methods of Tanaka et al. (2006) and Ren et al. (2007). In particular, neither ergot alkaloids, moniliformin, enniatins nor masked mycotoxins are included in any of these reports. To overcome these problems, some existing methods omit a clean-up of the sample and inject raw extracts into the LC/MS. This clearly increases the demands on the selectivity of the detector as well as on the investigation of matrix effects, especially if complicated food matrices are analysed. Spanjer et al. (2005) determined 22 mycotoxins (including the ergot alkaloid ergotamine) in different food matrices. Samples were extracted with an acetonitrile/water mixture and were diluted with water prior to injection. Matrix effects were investigated for every analyte/matrix combination and validation data obtained that suggested that the analysis of diluted raw extracts is indeed feasible and at the same time sensitive enough for determining most mycotoxin levels set in the legislation. Our own contribution in this field was the quantitative determination of a set of 39 analytes (including moniliformin, beauvericin, enniatins and masked mycotoxins) in wheat and maize (Sulyok et al. 2006). In both matrices, linear calibration curves were obtained (with the exception of moniliformin) after spiking blank matrices at multiple concentration levels, with coefficients of variance of the overall process of <5.1 and <3.0%, respectively. Significant matrix effects were observed for maize, but these could be overcome by matrix-matched calibration. LODs ranged from 0.03 to 220 mg kg1 and the trueness of the method was confirmed for deoxynivalenol and zearalenone though the analysis of certified reference materials. Very recently, this method has been extended to the simultaneous determination of 87 mycotoxins and has successfully been applied to mouldy food samples (Sulyok et al. 2007).

158

R. Krska et al. (Tudos et al. 2003) and may become an alternative method for rapid screening, which also enables the simultaneous detection of multiple mycotoxins using serial connected flow cells (van der Gaag et al. 2003). In a further label-free immunochemical approach for the detection of aflatoxin B1 and ochratoxin A, optical waveguide lightmode spectroscopy (OWLS) was used with integrated optical waveguide sensor chips measuring the resonance incoupling angle of polarized light, thus determining the surface coverage (Adanyi et al. 2007). A complementary tool for the screening of cereal samples may be DNA microarray-based chips using PCR followed by microarray colorimetric detection, which has been developed for the fast detection and identification of 14 trichothecene- and moniliformin-producing Fusarium species occurring on cereals (Kristensen et al. 2007). In recent years, interest in rapid membrane-based immunoassay methods, such as flow-through immunoassays and lateral flow devices (LFDs), has strongly increased due to the need for rapid on-site (pre)-screening. A flow-through enzyme immunoassay was developed for the detection of ochratoxin A in roasted coffee (Sibanda et al. 2002). Requiring no sample preparation other than an extraction step, LFDs allow qualitative or semi-quantitative determination of mycotoxins on one-step strip tests within a few minutes. Such LFDs have been developed for selected mycotoxins, such as aflatoxin B1 (Delmulle et al. 2005) and fumonisin B1 (Wang et al. 2006). The strong interest is furthermore reflected in the increasing number of commercially available test kits for field use, based mostly on direct competitive assays. Non-invasive techniques Optical methods, such as Fourier Transform midinfrared spectroscopy with attenuated total reflection (Kos et al. 2003) or near-infrared transmittance spectroscopy (Pettersson and Aberg 2003), are promising techniques for the fast and non-destructive detection of mycotoxins in grains. The approaches allow sample preparation to be reduced to an absolute minimum and to be integrated into on-line monitoring systems. Nevertheless, since rapid data interpretation is based on the output of chemometric analysis, the high matrix dependence and the lack of appropriate calibration materials are still major restrictions. Similarly, electronic noses, featuring an array of electronic chemical sensors with pattern recognition systems, have also been developed (Logrieco et al. 2005). In this approach, volatile organic compounds of low molecular weight, which are released by many fungi as products of secondary metabolism, are

In the near future, a strong trend towards multi-mycotoxin methods, which do not involve a clean-up of the sample, can be expected, as these methods can be relatively easily adapted to new analytes and matrices, and the obvious time- and cost-savings compensate for the expense of initial validation. Advances in the technology and in the instrumental design in mass spectrometry will further decrease the influence of matrix effects, which certainly constitute the main drawback of this approach at the moment.

Fast screening methods Immunochemical techniques Rapid methods based on immunochemical techniques often have the advantage of not requiring any clean-up or analyte enrichment steps. ELISAs have become routinely used tools for rapid monitoring of most mycotoxins, especially for the screening of raw materials (Gilbert and Anklam 2002; Fremy and Usleber 2003). Although ELISA tests may show a high matrix dependence and possible overestimation of levels, the advantages of the microtitre-plate format are speed, ease of operation, sensitivity and high sample throughput. ELISA test kits are commercially available for most of the major mycotoxins (EMAN 2007). Alternatives to ELISAs include a number of immunosensors as well as upcoming methods using immunochemical platforms, such as fluorescence polarization immunoassays (FPI) (Ngundi et al. 2005) or surface plasmon resonance (SPR) with mycotoxinprotein conjugates immobilized onto a sensor chip surface (Tudos et al. 2003). Immunosensors are emerging as a cost-effective alternative for screening and quantitative determination of mycotoxins (Maragos 2004). Array biosensors have been developed using competitive-based immunoassays for the simultaneous detection of multiple mycotoxins, including ochratoxin A, fumonisin B, aflatoxin B1, and deoxynivalenol, on a single waveguide surface by imaging the fluorescent pattern onto a CCD (charge-coupled device) camera (Sapsford et al. 2006). Other formats with fluorescence detection include automated flow-through immunosensors with enzyme-labelled mycotoxin derivatives (Urraca et al. 2005). Electrochemical immunosensors with surface-adsorbed antibodies on screen-printed carbon electrodes have been fabricated for the detection of aflatoxin M1 in milk (Micheli et al. 2005) and, in an array configuration, for the detection of aflatoxin B1 (Pemberton et al. 2006). Affinity-based surface plasmon resonance sensors (SPR) have the advantage of not requiring any labelling of the target mycotoxin

Mycotoxin analysis: An update adsorbed onto the sensor surface and measured with a variety of transduction systems based on electrical-, optical-, or mass-transduction, such as metal oxide sensors (MOS) and surface acoustic wave sensors (SAW), for example (Olsson et al. 2002). The high demand for rapid screening methods for mycotoxin analysis reflects the need for fast and cost-effective on-site determination of the level of mycotoxin contamination in food and feed. Immunochemical-based screening methods have shown great potential and are increasingly applied in routine analysis and monitoring of mycotoxins. Nevertheless, although rapid and selective, a loss of sensitivity may have to be taken into account in easyto-use-assays due to the necessary simplification of the system which usually employ no washing step. Future trends in screening methods include the further development of fast and simple tests requiring no instrumentation and improved detection capability for the simultaneous measurement of multiple mycotoxins.

159

methods for the determination of several mycotoxins, including aflatoxins, ochratoxin A, patulin, fumonisins, deoxynivalenol and zearalenone, in various matrices, such as cereals, nuts, milk, fruits and juices, as well as their products, intended for human consumption and animal feed.

Reference materials and intercomparison studies Reference materials (RM) or certified reference materials (CRM) are materials with a defined sample constitution and known or certified content of analyte(s) along with its uncertainty. (C)RMs play an important role not only during the validation process of a method but also as a measure to assure the quality of analytical data during routine analysis (in terms of trueness, comparability and traceability). (C)RMs can be classified as pure substances (standards), standard solutions (calibrators) or matrix materials (spiked or naturally contaminated). Until recently, the Institute of Reference Materials and Measurements (IRMM) of the EC has been the only provider of mycotoxin CRMs (IRMM 2007). These include RMs for aflatoxins in peanut (BCR 263, BCR 264, BCR 401), compound feed (BCR 375), milk powder (ERM-BC 282, ERM-BC 283, ERM-BC 284), ochratoxin A in wheat (BCR 471), deoxynivalenol in maize (BCR 377, BCR 378) and wheat flour (BCR 396), and zearalenone in maize (ERM-BC 716, ERM-BC 717). The IRMM also provides standard solutions for calibration purposes (calibrators) of deoxynivalenol (IRMM 315) and nivalenol (IRMM 316) in acetonitrile. The US National Institute of Standards and Technology (NIST) has recently issued a standard RM (SRM 2387) for aflatoxin determination in peanut butter (NIST 2007a). A full compilation of (C)RMs currently available and in production can be found on the homepages of various institutions, e.g. COMAR (2007), CORDIS (2007), IAEA (2007), IRMM (2007) and NIST (2007b). Despite past and current efforts to produce (C)RMs and calibrators for mycotoxin analysis, there is still an eminent need for more RMs that are appropriate for the different sample matrices and concentration ranges, especially with regard to the implementation and monitoring of regulatory limits (maximum levels) of mycotoxins in food and feed set by national and international authorities. Intercomparison studies (collaborative studies or proficiency testing schemes) play an important role in the validation of analytical methods and the production of RM (Gilbert and Anklam 2002; Josephs et al. 2004) as well as acting as quality assurance tools for laboratories. To improve the

Quality assurance Method validation As mentioned above, a multitude of methods have been published for the determination of mycotoxins in food and feed over the years. However, only a limited number of these publications include performance characteristics data obtained by method validation, which is a prerequisite for the production of reliable results in terms of comparability and traceability. Typical performance characteristics to be evaluated for the validation of a quantitative method are the limit of detection and quantification (LOD/LOQ), linearity, precision (repeatability and reproducibility), selectivity (interference of other compounds and/or matrix components), robustness/ruggedness, working range and trueness/bias (Josephs et al. 2004). Several protocols and guidelines for method validation have been published, e.g. the ISO standard 5725 (ISO 1994) or the guide The fitness for purpose of analytical methods (Eurachem 1998). There are various methods for mycotoxin analysis available that are validated and have been accepted by official authorities such as CEN, the AOAC, and ISO etc. (AOAC 2005; Gilbert and Anklam 2002). CEN usually evaluates the performance criteria of a method on the basis of collaborative studies. Most CEN methods are also AOAC- and ISO-approved. The latest edition of the Official Methods of Analysis (OMA) from the AOAC is available online and contains about 60 validated methods for mycotoxin analysis (AOAC 2005). Gilbert and Anklam (2002) have compiled validated and official analytical

160

R. Krska et al.
Berthiller F, Schuhmacher R, Buttinger G, Krska R. 2005. Rapid simultaneous determination of major type A- and B-trichothecenes as well as zearalenone in maize by high performance liquid chromatographytandem mass spectrometry. Journal of Chromatography A 1062:209216. Berthiller F, Dall0 Asta C, Schuhmacher R, Lemmens M, Adam G, Krska R. 2005. Masked mycotoxins: Determination of a deoxynivalenol glucoside in artificially and naturally contaminated wheat by liquid chromatography-tandem mass spectrometry. Journal of Agricultural and Food Chemistry 53:34213425. Biselli S, Hummert C. 2005. Development of a multicomponent method for Fusarium toxins using LCMS/MS and its application during a survey for the content of T-2 toxin and deoxynivalenol in various feed and food samples. Food Additives and Contaminants 22:752760. Castegnaro M, Tozlovanu M, Wild C, Molini A, Sylla A, PfohlLeszkowicz A. 2006. Advantages and drawbacks of immunoaffinity columns in analysis of mycotoxins in food. Molecular Nutrition and Food Research 50:480487. Cavaliere C, Foglia P, Pastorini E, Samperi R, Lagana A. 2005. Development of a multiresidue method for analysis of major Fusarium mycotoxins in corn meal using liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrometry 19:20852093. COMAR. 2007. COMAR Homepage. International database for certified reference materials. Berlin, Germany: BAM Federal Institute for Materials Research and Testing. Available: http:// www.comar.bam.de. Accessed 10 July 2007. CORDIS. 2007. CORDIS Homepage. Community Research & Development Information Service. Luxembourg: European Union. Available: http://cordis.europa.eu. Accessed 10 July 2007. Creppy EE, Chiarappa P, Baudrimont I, Boracci P, Moukha S, Carratu MR. 2004. Synergistic effects of fumonisin B1 and ochratoxin A: Are in vitro cytotoxicity data predictive of in vivo acute toxicity? Toxicology 201:115123. Delmulle BS, De Saeger SMDG, Sibanda L, Barna-Vetro I, Van Peteghem CH. 2005. Development of an immunoassay-based lateral flow dipstick for the rapid detection of aflatoxin B1 in pig feed. Journal of Agricultural and Food Chemistry 53:33643368. DMello JPF, MacDonald AMC. 1997. Mycotoxins. Animal Feed Science and Technology 69:155166. EC. 2006a. Commission Regulation (EC) No 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Union L364:524. EC. 2006b. Commission Regulation (EC) No 401/2006 of 23 February 2006 laying down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs. Official Journal of the European Union L70:1234. EMAN. 2007. EMAN Homepage. European Mycotoxin Awareness Network. Available: http://www.mycotoxins.org. Accessed 10 July 2007. Eurachem. 1998. The fitness for purpose of analytical methods: A laboratory guide to method validation and related topics. Teddington, UK: Eurachem. FAO. 2004. Worldwide regulation for mycotoxins in food and feed in 2003. FAO Food and Nutrition paper 81. Rome: Food and Agriculture Organization of the United Nations. Available: http://www.fao.org/docrep/007/y5499e/y5499e00.htm. Accessed 10 July 2007. FAPAS. 2007. FAPAS Homepage. Available: http://www.fapas.com. Accessed 10 July 2007.

comparability and traceability of analytical data in Europe, several intercomparison studies in the field of mycotoxin analysis, especially for Fusarium mycotoxins, have been performed in the past 10 years within projects funded by the EC (Josephs et al. 2004; Krska et al. 2005). Additionally, the Food Analysis Performance Assessment Scheme (FAPAS) continually organises proficiency testing schemes in mycotoxin analysis (FAPAS 2005).

Conclusion and outlook In the laboratory, sample extracts are preferably purified and enriched in a clean-up step procedure using mainly SPE and IAC, with multifunctional IACs currently being put to the test. TLC and LC are still the most frequently employed analytical methods for the (official) determination of mycotoxins; however, LCMS/MS is increasingly used for the simultaneous determination and identification of large numbers of mycotoxins, currently up to 87. The use of CRMs and certified calibrants is one of the key issues of quality assurance in the analytical laboratory. The established state-of-the-art chromatographybased methods for mycotoxin analysis are increasingly being complemented by a number of new screening methods, including LFDs, biosensors and IR-screening techniques, that are fast and costeffective. Nevertheless, these techniques will have to compete with both classical confirmatory methods and MTP-ELISAs, which are now widely used for mycotoxin screening.

References
Adanyi N, Levkovets IA, Rodriguez-Gil S, Ronald A, Varadi M, 00 Szendro I. 2007. Development of immunosensor based on OWLS technique for determining aflatoxin B1 and ochratoxin A. Biosensors and Bioelectronics 22:797802. Ambrosino P, Galvano F, Fogliano V, Logrieco A, Fresa R, Ritieni A. 2004. Supercritical fluid extraction of beauvericin from maize. Talanta 62:523530. AOAC. 2005. Official Methods of Analysis of AOAC International. 18th ed. Gaithersburg (MD): AOAC International. Available: http://www.eoma.aoac.org. Accessed 10 July 2007. Appell M, Kendra DF, Kim EK, Maragos CM. 2007. Synthesis and evaluation of molecularly imprinted polymers as sorbents of moniliformin. Food Additives and Contaminants 24:4352. Baggiani C, Giraudi G, Vanni A. 2001. A molecular imprinted polymer with recognition properties towards the carcinogenic mycotoxin ochratoxin A. Bioseparation 10:389394. Bennett JW, Klich M. 2003. Mycotoxins. Clinical Microbiology Reviews 16:497516.

Mycotoxin analysis: An update

FDA. 2007. Investigative Operation Manual. Washington (DC): Food and Drug Administration. Available: http://www.fda.gov/ ora/inspect_ref/iom/contents/ss_toc.html. Accessed 10 July 2007. Fremy JM, Usleber E. 2003. Policy on characterization of antibodies used in immunochemical methods of analysis for mycotoxins and phycotoxins. Journal of AOAC International 86:868871. Gilbert J, Anklam E. 2002. Validation of analytical methods for determining mycotoxins in foodstuffs. Trends in Analytical Chemistry 21:468486. GIPSA. 1994. Rice Inspection Handbook. Chapter 2: Sampling. Washington (DC): USDA Grain Inspection, Packers and Stockyards Administration. Available: http://www.gipsa.usda.gov/GIPSA/webapp?areahome&subjectlr&topichb-ri. Accessed 10 July 2007. GIPSA. 1995. Grain Inspection Handbook. Book I: Grain Sampling. Washington (DC): USDA Grain Inspection, Packers and Stockyards Administration. Available: http:// archive.gipsa.usda.gov/reference-library/handbooks/grain-insp/ grbook1/gihbk1.htm. Accessed 10 July 2007. GIPSA. 1998. Hop Inspection Handbook. Chapter 2: Sampling. Washington (DC): USDA Grain Inspection, Packers and Stockyards Administration. Available: http://www.gipsa.usda.gov/GIPSA/webapp?areahome&subjectlr&topichb-hi. Accessed 10 July 2007. Huopalahti RP, Ebel J Jr., Henion JD. 1997. Supercritical fluid extraction of mycotoxins from feeds with analysis by LC/UV and LC/MS. Journal of Liquid Chromatography and Related Technologies 20:537551. IAEA. 2007. IAEA Homepage. Vienna: International Atomic Energy Agency. Available: http://www.iaea.org/. Accessed 10 July 2007. IRMM. 2007. IRMM reference materials catalogue. Geel, Belgium: Institute of Reference Materials and Measurements of the European Commissions Directorate General Joint Research. Available: http://www.irmm.jrc.be/html/reference_materials_catalogue/catalogue/index.htm. Accessed 10 July 2007. Jodlbauer J, Maier NM, Lindner W. 2002. Towards ochratoxin A selective molecularly imprinted polymers for solid-phase extraction. Journal of Chromatography A 945:4563. Josephs RD, Derbyshire M, Stroka J, Emons H Anklam E. 2004. Trichothecenes: reference materials and method validation. Toxicological Letters 153:123132. Juan C, Gonzalez L, Soriano JM, Molto JC, Manes J. 2005. Accelerated solvent extraction of ochratoxin A from rice samples. Journal of Agricultural and Food Chemistry 53:934851. Kay S. 2001. Comparison of sampling approaches for grain lots. EUR 20134 EN. Ispra, Italy: European Commission, Directorate General JRC, Institute for Health and Consumer Protection. Available: http://biotech.jrc.it/home/doc/sampling_ comparison_Kay_paper.pdf. Accessed 10 July 2007. Kokkonen M, Jestoi M, Rizzo A. 2005. Determination of selected mycotoxins in moulded cheeses with liquid chromatography coupled to tandem mass spectrometry. Food Additives and Contaminants 22:449456. Kos G, Lohninger H, Krska R. 2003. Development of a method for the determination of Fusarium fungi on corn using midinfrared spectroscopy with attenuated total reflection and chemometrics. Analytical Chemistry 75:12111217. Kristensen R, Gauthier G, Berdal KG, Hamels S, Remacle J, Holst-Jensen A. 2007. DNA microarray to detect and identify trichothecenes- and moniliformin-producing Fusarium species. Journal of Applied Microbiology 102:10601070.

161

Krska R. 1998. Performance of modern sample preparation techniques in the analysis of Fusarium mycotoxins in cereals. Journal of Chromatography A 815:4957. Krska R, Josephs R. 2001. The state-of-the-art in the analysis of estrogenic mycotoxins in cereals. Fresenius Journal of Analytical Chemistry 369:469476. Krska R, Baumgartner S, Josephs R. 2001. The state-of-the-art in the analysis of type-A and -B trichothecene mycotoxins in cereals. Fresenius Journal of Analytical Chemistry 371:285299. Krska R, Welzig E, Berthiller F, Molinelli A, Mizaikoff B. 2005. Advances in the analysis of mycotoxins and its quality assurance. Food Additives and Contaminants 22:345353. Leitner A, Zollner P, Paolillo A, Stroka J, PapadopoulouBouraoui A, Jaborek S, Anklamb E, Lindner W. 2002. Comparison of methods for the determination of ochratoxin A in wine. Analytica Chimica Acta 453:3341. Liau BC, Jong TT, Lee MR, Chang CMJ. 2007. Supercritical fluid extraction and quantification of aflatoxins in Zizyphi fructus by liquid chromatography/ atmospheric pressure chemical ionization tandem mass spectrometry. Rapid Communications in Mass Spectrometry 21:667673. Logrieco A, Arrigan DWM, Brengel-Pesce K, Siciliano P, Tothill I. 2005. DNA arrays, electronic noses and tongues, biosensors and receptors for rapid detection of toxigenic fungi and mycotoxins: A review. Food Additives and Contaminants 22:335344. Maier NM, Buttinger G, Welhartizki S, Gavioli E, Lindner W. 2004. Molecularly imprinted polymer-assisted sample clean-up of ochratoxin A from red wine: Merits and limitations. Journal of Chromatography B 804:103111. MacDonald SJ, Kelleher B, Donelly C, Hird S. 2007. Multimycotoxin analysis using immunoaffinity column clean-up and LC-MS/MS determination. Poster 1438 at the XII IUPAC Symposium on Mycotoxins and Phycotoxins. Istanbul; May 2125, 2007. Maragos CM. 2004. Emerging technologies for mycotoxin detection. Journal of Toxicology Toxin Review 23:317344. Micheli L, Grecco R, Badea M, Moscone D, Palleschi G. 2005. An electrochemical immunosensor for aflatoxin M1 determination in milk using screen-printed electrodes. Biosensors and Bioelectronics 21:588596. Ngundi MM, Shriver-Lake LC, Moore MH, Lassman ME, Ligler FS, Taitt CR. 2005. Array biosensor for detection of ochratoxin A in cereals and beverages. Analytical Chemistry 77:148154. Nielsen KF, Smedsgaard J. 2003. Fungal metabolite screening: Database of 474 mycotoxins and fungal metabolites for dereplication by standardised liquid chromatographyUV mass spectrometry methodology. Journal of Chromatography A 1002:111136. NIST. 2007a. Certificate of Analysis. Standard Reference Material 2387: Peanut Butter. Gaithersburg (MD): National Institute of Standards & Technology. Available: http://srmors.nist.gov/view_cert.cfm?srm2387. Accessed 10 July 2007. NIST. 2007b. NIST Homepage. Gaithersburg (MD): National Institute of Standards & Technology. Available: http:// www.nist.gov. Accessed 10 July 2007. Olsson J, Borjesson T, Lundstedt T, Schnurer J. 2002. Detection and quantification of ochratoxin A and deoxynivalenol in barley grains by GCMS and electronic nose. International Journal of Food Microbiology 72:203214.

162

R. Krska et al.
Spanjer M, Rensen P, Scholten J. 2005. Multimycotoxin analysis by LCMS/MS in a single extract. Presentation. In: Book of abstracts, The World Mycotoxin Forum The Third Conference, 1011 November 2005, Noordwijk aan Zee, The Netherlands. Speijers GJH, Speijers MHM. 2004. Combined toxic effects of mycotoxins. Toxicology Letters 153:98. Sugita-Konishi Y, Tanaka T, Nakajima M, Fujita K, Norizuki H, Mochizuki N, Takatori K. 2006. The comparison of two clean-up procedures, multifunctional column and immunoaffinity column, for HPLC determination of ochratoxin A in cereals, raisins and green coffee beans. Talanta 69:650655. Sulyok M, Berthiller F, Krska R, Schuhmacher R. 2006. Development and validation of a liquid chromatography/ tandem mass spectrometric method for the determination of 39 mycotoxins in wheat and maize. Rapid Communications in Mass Spectrometry 20:26492659. Sulyok M, Berthiller, F, Krska R, Schuhmacher R. 2007. Liquid chromatography/tandem mass spectrometric multimycotoxin method comprising 87 analytes and its application to moldy food samples. Analytical and Bioanalytical Chemistry. 389: 15051523. Tanaka H, Takino M, Sugita-Konishi Y, Tanaka T. 2006. Development of a liquid chromatography/time-of-flight mass spectrometric method for the simultaneous determination of trichothecenes, zearalenone and aflatoxins in foodstuffs. Rapid Communications in Mass Spectrometry 20:14221428. Tudos AJ, Lucas-van den Bos ER, Stigter ECA. 2003. Rapid surface plasmon resonance-based inhibition assay of deoxynivalenol. Journal of Agricultural and Food Chemistry 51:58435848. Tuomi T, Johnsson T, Hintikka EL, Reijula K. 2001. Detection of aflatoxins (G1-2, B1-2), sterigmatocystin, citrinine and ochratoxin A in samples contaminated by microbes. Analyst 126:15451550. Turner NW, Piletska EV, Karim K, Whitcombe M, Malecha M, Magan N, Baggiani C, Piletsky SA. 2004. Effect of the solvent on recognition properties of molecularly imprinted polymer specific for ochratoxin A. Biosensors and Bioelectronics 20:10601067. Urraca JL, Marazuela MD, Moreno-Bondi MC. 2004. Analysis for zearalenone and -zearalenol in cereals and swine feed using accelerated solvent extraction and liquid chromatography with fluorescence detection. Analytica Chimica Acta 524:175183. Urraca JL, Benito-Pena E, Perez-Conde C, Moreno-Bondi M, Pestka JJ. 2005. Analysis of zearalenone in cereal and swine feed samples using an automated flow-through immunosensor. Journal of Agricultural and Food Chemistry 53:33383344. Urraca JL, Marazuela MD, Moreno-Bondi MC. 2006a. Molecularly imprinted polymers applied to the clean-up of zearalenone and -zearalenol from cereal and swine feed sample extracts. Analytical and Bioanalytical Chemistry 385:11551161. Urraca JL, Marazuela MD, Merino ER, Orellana G, MorenoBondi MC. 2006b. Molecularly imprinted polymers with a streamlined mimic for zearalenone analysis. Journal of Chromatography A 1116:127134. Van der Gaag B, Spath S, Dietrich H, Stigter E, Boonzaaijer G, van Osenbruggen T, Koopal K. 2003. Biosensors and multiple mycotoxin analysis. Food Control 14:251254. Wang S, Quan Y, Lee N, Kennedy IR. 2006. Rapid determination of fumonisin B1 in food samples by enzyme-linked immunosorbent assay and colloidal gold immunoassay. Journal of Agricultural and Food Chemistry 54:24912495.

Pallaroni L, von Holst C. 2003. Determination of zearalenone from wheat and corn by pressurized liquid extraction and liquid chromatographyelectrospray mass spectrometry. Journal of Chromatography A 993:3945. Pallaroni L, von Holst C. 2004. Development of an extraction method for the determination of zearalenone in corn using less organic solvents. Journal of Chromatography A 1055:247249. Pallaroni L, von Holst C, Sparr Eskilsson C, Bjorklund E. 2002. Microwave-assisted extraction of zearalenone from wheat and corn. Analytical and Bioanalytical Chemistry 374:161166. Pemberton RM, Pittson R, Biddle N, Drago GA, Hart JP. 2006. Studies towards the development of a screen-printed carbon electrochemical immunosensor array for mycotoxins: A sensor for aflatoxin B1. Analytical Letters 39:15731586. Pettersson H, Aberg L. 2003. Near-infrared spectroscopy for determination of mycotoxins in cereals. Food Control 14:229232. Radova Z, Holadova K, Hajslova J. 1998. Comparison of two clean-up principles for determination of trichothecenes in grain extract. Journal of Chromatography A 829:259267. Reinsch M, Topfer A, Lehmann A, Nehls I. 2005. Determination of ochratoxin A in wine by liquid chromatography tandem mass spectrometry after combined anion-exchange/reversedphase clean-up. Analytical and Bioanalytical Chemistry 381:15921595. Ren Y, Zhang Y, Shao S, Cai Z, Feng L, Pan H, Wang Z. 2007. Simultaneous determination of multi-component mycotoxin contaminants in foods and feeds by ultra-performance liquid chromatography tandem mass spectrometry. Journal of Chromatography A 1143:4864. Romerlabs 2007. Romerlabs Homepage. Tulln, Austria: Romer Labs Diagnostic GmbH. Available: http://www.romerlabs.com. Accessed 10 July 2007. Royer D, Humpf HU, Guy PA. 2004. Quantitative analysis of Fusarium mycotoxins in maize using accelerated solvent extraction before liquid chromatography/atmospheric pressure chemical ionization tandem mass spectrometry. Food Additives and Contaminants 21:678692. Rundberget T, Wilkins AL. 2002. Determination of Penicillium mycotoxins in foods and feeds using liquid chromatographymass spectrometry. Journal of Chromatography A 964:189197. SANCO. 2005. Guidance document for competent authorities for the control of compliance with EU legislation on aflatoxins. SANCO/1208/2005-rev1. Sapsford KE, Ngundi MM, Moore MH, Lassman ME, ShriverLake LC, Taitt CR, Ligler FS. 2006. Rapid detection of foodborne contaminants using an array biosensor. Sensors Actuators B 113:599607. Sforza S, Dall0 Asta C, Marchelli R. 2006. Recent advances in mycotoxin determination in food and feed by hyphenated chromatographic techniques/mass spectrometry. Mass Spectrometry Reviews 25:5476. Sibanda L, De Saeger S, Barna-Vetro I, Van Peteghem C. 2002. Development of a solid-phase cleanup and portable rapid flowthrough enzyme immunoassay for the detection of ochratoxin A in roasted coffee. Journal of Agricultural and Food Chemistry 50:69646967. Smedsgaard J, Frisvad JC. 1996. Using direct electrospray mass spectrometry in taxonomy and secondary metabolite profiling of crude fungal extracts. Journal of Microbiological Methods 25:517. Srensen LK, Elbk TH. 2005. Determination of mycotoxins in bovine milk by liquid chromatography tandem mass spectrometry. Journal of Chromatography B 820:183196.

Mycotoxin analysis: An update

Weiss R, Freudenschuss M, Krska R, Mizaikoff B. 2003, Improving methods of analysis for mycotoxins: Molecularly imprinted polymers for deoxynivalenol and zearalenone. Food Additives and Contaminants 20:386395. WHO. 2002. Safety evaluation of certain mycotoxins in food. Fifty-sixth report of the Joint FAO/WHO Expert Committee on Food Additives. WHO Technical Report Series 906. Geneva: World Health Organisation (WHO). Available: http://whqlibdoc.who.int/trs/WHO_TRS_906.pdf. Accessed 10 July 2007. Zollner P, Jodlbauer J, Lindner W. 1999. Determination of zearalenone in grains by high-performance liquid

163

chromatographytandem mass spectrometry after solid-phase extraction with RP-18 columns or immunoaffinity columns. Journal of Chromatography A 858:167174. Zollner P, Leitner A, Lubda D, Cabrera K, Lindner W. 2000. Application of a Chromolith SpeedROD RP-18e HPLC column: Determination of ochratoxin A in different wines by high-performance liquid chromatography-tandem mass spectrometry. Chromatographia 52:818820. Zollner P, Mayer-Helm B. 2006. Trace mycotoxin analysis in complex biological and food matrices by liquid chromatographyatmospheric pressure ionisation mass spectrometry. Journal of Chromatography A 1136:123169.

Food Additives and Contaminants, February 2008; 25(2): 164171

Use of cyclodextrins as modifiers of fluorescence in the detection of mycotoxins

C. M. MARAGOS1, M. APPELL1, V. LIPPOLIS2, A. VISCONTI2, L. CATUCCI3, & M. PASCALE2

Mycotoxin Research Unit, USDA-ARS-NCAUR, 1815 N. University Street, Peoria, IL 61604, USA, Istituto di Scienze delle Produzioni Alimentari (ISPA), Consiglio Nazionale delle Ricerche, Via G. Amendola, 122/o, Bari 70126, Italy, and 3Dipartimento di Chimica, ` Universita di Bari, Via Orabona 4, Bari 70126, Italy
2 1

(Received 11 May 2007; revised 28 June 2007; accepted 10 July 2007)

Abstract Cyclodextrins, cyclic oligosaccharides composed of amylose subunits, are known to interact with mycotoxins. The interactions may be useful to analytical chemists by altering the properties of the mycotoxin of interest, namely the chromatographic properties, electrophoretic properties, fluorescence, or absorption of these fungal metabolites. Practical applications of these effects have been the incorporation of cyclodextrins into high-performance liquid chromatography and capillary electrophoresis methods for mycotoxin detection. Specific mycotoxins include those with a native fluorescence such as the aflatoxins, ochratoxin A (OTA) and zearalenone (ZEN) as well as those that can be rendered fluorescent through derivatization, such as T-2 toxin. The literature describing the applications of cyclodextrins in mycotoxin analysis is reviewed and an attempt to extend the use of cyclodextrins to the detection of labelled T-2 toxin is presented. Twenty cyclodextrins were evaluated for their ability to enhance the fluorescence emission of T-2 toxin derivatized with pyrene1-carbonyl cyanide (T2-Pyr). This evaluation revealed that heptakis (2,6-di-O-methyl)--cyclodextrin (DIMEB), in particular, enhanced T2-Pyr fluorescence. DIMEB was used as a buffer modifier in a capillary electrophoresis-laser-induced fluorescence (CE-LIF) method for detecting T-2 in maize. Because of the effects that certain cyclodextrins have, especially under aqueous conditions, they may make useful additives for a variety of mycotoxin analytical methods.

Keywords: Mycotoxin, trichothecene, zearalenone, cyclodextrin, fluorescence, analysis

Introduction Among the mycotoxins known to have a native fluorescence emission are certain of the aflatoxins, ochratoxins, zearalenone, and related congeners. The structures, toxic effects, and fluorescence characteristics of these three groups of toxins differ immensely. The fluorophores are derived from a variety of molecular structures. The aflatoxins contain the coumarin moiety, which is highly fluorescent. Similarly, the ochratoxins contain an isocoumarin moiety linked to the amino acid phenylalanine, while zearalenone and related compounds are resorcyclic acid lactones (Figure 1). Even small changes in a toxins structure can dramatically influence fluorescence. The fluorescence emission of
Correspondence: C. M. Maragos. E-mail: chris.maragos@ars.usda.gov ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701564555

certain of the aflatoxins, those containing a double bond in the furan moiety, can be enhanced by a number of techniques, including halogenation, hydrolysis and rearrangement to the more fluorescent phenolate ions (aflatoxins B2a, G2a, M2a), or photochemical reaction. The fluorescence emission of many fluorophores are also known to be sensitive to the local environment, and this is true for the aflatoxins, ochratoxin A (OTA) and zearalenone (ZEN) as well. Fluorescence emission of aflatoxins B1 and G1 (AFB1, AFG1) are substantially greater in solvents such as methanol or chloroform than in water (Vazquez et al. 1991). The emission maximum is also influenced by solvation, with a shift towards shorter wavelengths (blue shift) as the environment

Cyclodextrins and mycotoxin fluorescence

165

fluorophore and water in the presence of the CD. Thus the fluorescence emission of fluorophores small enough to form inclusion complexes may be affected (Easton & Lincoln 1999). Certain of the mycotoxins fit these criteria. This manuscript provides a short review of the applications of this phenomenon to mycotoxin analysis and a description of an attempt to extend the usefulness of the effect to a mycotoxin lacking a native fluorescence (T-2 toxin) after labelling it with the fluorophore pyrene-1-carbonyl cyanide (PCC), yielding T2-pyrene (T2-Pyr).

Applications of CDs to the analysis of mycotoxins having a native fluorescence The aflatoxins have a native fluorescence that can be excited with ultraviolet light (360365 nm). The nomenclature of the B or G aflatoxins was derived from the colour of the fluorescence: AFB1 and AFB2 have a blue fluorescence (emission circa 440 nm), while AFG1 and AFG2 have a blue/green fluorescence (emission circa 460 nm). Aflatoxin emission can be affected by solvent composition, temperature, and interactions with solid phases, such as silica gel (Robertson & Pons 1968, Van Duuren et al. 1968, Dirr & Schabort 1987, Vazquez et al. 1991). The effects of cyclodextrins on substituted coumarins, such as the aflatoxins, have also been known for some time (Francis et al. 1988, Vazquez et al. 1991) and have been applied in several different ways. Among these include the use of CDs in fungal culture media as an aid in the identification of toxinproducing isolates (Fente et al. 2001, Abbas et al. 2004, Jaimez et al. 2004, Rojas et al. 2005), and the use of CDs as reagents in chemical analyses for the aflatoxins (Francis et al. 1988, Cepeda et al. 1996, Franco et al. 1998, Vazquez et al. 1999, DallAsta et al. 2003). The latter include pre-column (Chiavaro et al. 2001) and post-column (Francis et al. 1988, Cepeda et al. 1996, Vazquez et al. 1999) HPLC assays. Chiavaro et al. (2001) used 6 mM of various CDs added to the mobile phase pre-column and reported that the greatest fluorescence enhancements for AFB1 and AFM1 were obtained with succinyl--CD, while the best enhancement for AFG1 was obtained with heptakis (2,6-di O-methyl)--cyclodextrin (DIMEB). Vazquez et al. (1999) reported better resolution and lower back pressure with the post-column approach. The earliest report (Francis et al. 1988) used unmodified -CD, while the later reports used DIMEB, which was found to provide greater enhancement (Cepeda et al. 1996). The CDs have also been examined for use as aids in the separation of aflatoxins by

Figure 1. Representative structures of some of the mycotoxins. Aflatoxin B1(AFB1), ochratoxin A (OTA), and zearalenone (ZEN).

becomes less polar, an effect observed for AFB1 in various solvents (Dirr & Schabort 1987, Vazquez et al. 1991). Cyclodextrins (CDs) have been shown previously to enhance the fluorescence emission from a number of hydrophobic fluorophores. The CDs are cyclic oligosaccharides composed of multiple subunits of glucose in an (14) configuration. They are classified by the number of subunits ( 6,  7, 8) and by the type and degree of substitution. The CDs have a cavity (pore) that may accommodate small molecules as guests, forming inclusion complexes. The size of the pore and the environment within it can be modulated through changes to the subunits, with cavity diameters of 4.7, 6.8, and 7.5 A for the -, -, and -CD, respectively, and annular depths of 7.98.0 A (Easton & Lincoln 1999). A possible mechanism for the fluorescence enhancement is believed to be by providing favourable interactions between the fluorophore and the cyclodextrin. The effect might also be derived from a reduction of the interactions between the

166

C. M. Maragos et al. decreased, increasing AFB1 fluorescence (RamrezGalicia et al. 2007). Based on models of the docking of AFB1 with - and -CDs it has been suggested that inclusion of the fluorophore into the CD cavity may reduce the quenching effect of the solvent, thereby enhancing fluorescence (Amadasi et al. 2007). The exact mechanism of interaction of the CDs with aflatoxins has not been reported, although formation of a 1 : 1 inclusion complex has been suggested (DallAsta et al. 2003). Recently modelling studies have also suggested the dihydrofuran portion of AFB1 is inserted into -CD, with possible hydrogen bonding between the carbonyl groups of the toxin and the secondary hydroxyl groups of the CD (Amadasi et al. 2007). It should be noted that significant enhancement of aflatoxin emission has generally required using high ratios of CD: aflatoxin, ranging from more than 103 : 1 to 105 : 1. Therefore, regardless of whether the stoichiometry of the host guest complex may be 1 : 1, 2 : 1, or 1 : 2, the need for high ratios of CDs suggest the interaction of aflatoxins with the CDs may be more complicated than the formation of stable inclusion complexes, perhaps including interactions with the outer surface of the CDs (Vazquez et al. 1991). Ochratoxin A (OTA), is a substituted isocoumarin related to the fungal metabolite mellein, which is also fluorescent (Sachse 1992). The fluorescence emission of OTA is sensitive to the hydrophobicity and pH of the environment (Golinski & Chelkowski 1978, Hunt et al. 1979). The absorption spectrum of OTA changes dramatically with pH, with the band near 320 nm decreasing and the band near 370 nm increasing with increasing pH over the range from 3.5 to 11 (Bohs et al. 1995, Verrone et al. 2007). The fluorescence excitation maxima for OTA on silica gel was reported as 340 nm, with an emission maximum of 475 nm (Chu 1970), and HPLC of OTA with excitation at 330340 nm and emission at 460470 nm is commonly used (Scott 2002). The effect of -CD on the spectroscopic properties of OTA in aqueous solution has been recently investigated by Verrone et al. (2007). A 1 : 1 stoichiometry of OTA/-CD was observed at all tested pHs (range 3.59.5) with an increase in emission intensity of up to about twofold (excitation at 330 nm, emission at 450 nm). It was reported that -CD interacted with both the protonated and deprotonated forms of OTA, with binding constants of 2140 and 9790 M1 at pHs 3.5 and 9.5, respectively. This suggests that the dianion form of OTA interacts more strongly with -CD than the protonated form. Molecular modelling simulations have also suggested that the phenylalanine portion of OTA is involved in the inclusion complexation with -CD (Amadasi et al. 2007).

capillary electrophoresis. The initial report by Holland & Sepaniak (1993) tested -, -, and CDs in combination with sodium dodecyl sulfate (SDS) and acetonitrile as buffer additives in the separation of ten mycotoxin standards. Unfortunately, the study used absorbance, rather than fluorescence for detection, so potential effects on fluorescence were not reported. Additionally, -, -, and -CDs had minimal effects upon retention of aflatoxins B1, B2, G1, and G2 (Holland & Sepaniak 1993). The first report to incorporate the fluorescence enhancement effects of CDs with capillary electrophoresis of aflatoxins was that of Wei et al. (2000). The approach used a titanium: sapphire laser (730770 nm) to excite aflatoxins B1, B2, G1, and G2, with either carboxymethyl--CD or sulfated-CD present at 210 mM in the electrophoretic buffer. The carboxymethyl--CD was useful in helping to resolve the aflatoxins. The result was a rapid (80 s) separation of AFB2, AFG1, and AFB1, and limits of detection of 0.20.4 nM. Several CDs have been screened for their ability to enhance aflatoxin fluorescence, and many of these caused a blue shift in the emission from 440 to 435 nm, suggestive of the formation of an inclusion complex (DallAsta et al. 2003). The greatest relative enhancement was observed with succinyl-CD, but others such as DIMEB and carboxymethyl--CD were also quite effective. Based upon binding constants of aflatoxins B1, B2, G1, G2, and M1, the authors suggested the furan moiety of the aflatoxin was involved in the CD inclusion complex (DallAsta et al. 2003). Interestingly, three hydroxylated aflatoxins (AFQ1, AFP1, AFM1) did not show a shift in excitation or emission maxima in the presence of CDs (Franco et al. 1998). AFM1 differs from AFB1 by a hydroxyl group on the dihdryofuran moiety at the C14 position, while AFQ1 differs from AFB1 by a hydroxyl group on the other end of the molecule: the cyclopentenone moiety. AFM1 exhibited a fivefold enhancement of the emission when tested with DIMEB, while AFQ1 exhibited a much greater enhancement of 39-fold (Franco et al. 1998). Furthermore, aflatoxins B2 and G2, which are analogues of AFB1 and AFG1 with the furan double bond reduced, show little enhancement with CDs (Vazquez et al. 1991). These spectroscopic results support the hypothesis that the DIMEB is interacting with the dihydrofuran portion of the molecule: an interaction which is hindered in the case of AFM1 by the presence of the hydroxyl in this region. Alternatively, a theoretical study has proposed a mechanism whereby vibrational coupling of the carbonyl groups of AFB1 with water allows deexcitation of the AFB1. When -CD is added the interactions of the carbonyls with the water may be

Cyclodextrins and mycotoxin fluorescence While the effect of -CD on OTA fluorescence intensity is not as large as it is for the aflatoxins, the CDs have still been found to be useful in chromatographic assays. Specifically, -CD was reported to facilitate the separation of OTA from ZEN in reverse phase HPLC (Seidel et al. 1993), and to facilitate the separation of OTA from moniliformin in a capillary electrophoresis assay (Bohs et al. 1995). With HPLC, a slight (15%) increase was observed in the sensitivity of the assay to OTA with the inclusion of -CD, which corresponded with a higher ultraviolet light absorption of OTA over the range 310350 nm (Seidel et al. 1993). The increased emission was slight, perhaps because the mobile phase that was used (methanol/water 45 : 55) may have had sufficient solvent strength to obscure the effect. With capillary electrophoresis, 20 mM hydroxypropyl--CD (HP--CD) added to the electrophoresis buffer enhanced the separation factor for OTA from moniliformin, but did not affect the separation factors for ZEN/OTA or OTA/ochratoxin B pairs (Bohs et al. 1995). The apparent mobility of OTA was increased when -CD, HP--CD, HP- -CD, or -CD were added to the electrophoresis buffer. Conversely, the migration of OTA in a micellar electrokinetic capillary chromatography assay was not affected appreciably by 7 mM of either -, -, or -CD, although effects upon OTA fluorescence were not examined (Holland & Sepaniak 1993). The report of Seidel et al. (1993), as described previously, also suggested that ZEN forms an inclusion complex with -CD. Despite the use of fluorescence detection, the effect upon fluorescence intensity of ZEN caused by formation of the inclusion complex was not reported perhaps, as with the OTA, because of the strength of the mobile phase (45% methanol), or the low concentration of -CD used (0.15 mM). The lower CD concentration was recommended in order to reduce backpressure of the HPLC system. The use of CDs to influence the electrophoretic separation of ZEN has also been investigated (Holland & Sepaniak 1993, Bohs et al. 1995, Maragos & Appell 2007). The early work described a strong interaction between -CD and ZEN as measured by an effect upon electrophoretic migration (Holland & Sepaniak 1993). Conversely, -CD and -CD did not show significant effects upon ZEN migration. In the latter report ultraviolet light (254 nm) rather than fluorescence detection was used, and the potential effect of the inclusion complex formation on ZEN fluorescence was therefore not described. Bohs et al. (1995) examined the effects of several CDs on the separation of mycotoxins, including ZEN. CDs examined included: -CD, 2,3,6-methyl--CD, HP--CD, -CD, and HP- -CD. The apparent mobility of

167

ZEN was reduced for the 2,3,6-methyl--CD and increased for the other CDs, relative to the mobility in the absence of a CD. Of the CDs, the methylated -CD also yielded the greatest separation factor between ZEN and OTA. It was suggested that ZEN may be too large to form inclusion complexes with HP--CD (Bohs et al. 1995). Although the authors examined several CDs, the method of detection was by photodiode array, and therefore potential effects of the CDs upon ZEN fluorescence were not reported. Recently a systematic description of the effects of 22 cyclodextrins upon ZEN fluorescence was reported (Maragos & Appell 2007). In that work, several of the CDs were shown to influence both the magnitude of the fluorescence emission and electrophoretic mobility in a CE-LIF format. The -CDs had little effect upon fluorescence, suggesting the cavity of the -CDs may be too small for ZEN to form an inclusion complex. The CDs giving the greatest enhancement of fluorescence (325 nm excitation) were DIMEB, 6-monodeoxy-6-monoamino--CD, carboxyethylated--CD, and -CD. The presence of DIMEB increased the mobility of ZEN, while the presence of carboxyethylated--CD reduced it. DIMEB, which gave the greatest fluorescence enhancement, was used to develop a CE-LIF method for detecting ZEN in maize, with a limit of quantitation of 5 ng ZEN g1 maize (Maragos & Appell 2007).

Application of CDs to analysis of fluorescently labelled T-2 toxin T-2 toxin has been reported to interact with CD-bonded phases in an HPLC format (Armstrong et al. 1985). Given this, the effect of CDs upon the native fluorescence of mycotoxins, and the ability of CDs to influence chromatographic and electrophoretic separations, we endeavoured to determine the effect of commonly available CDs upon the fluorescence of a pyrene derivative of T-2 toxin (T2-Pry). The T2-Pyr was made by reacting T-2 toxin with pyrene-1-carbonyl cyanide (PCC) using reaction chemistry equivalent to that of the 1-anthroylnitrile (1-anthroylcyanide) derivative (Pascale et al. 2003, Lippolis et al., personal communication). T-2 toxin was solubilized with 0.05 ml of 15 mM 4-dimethylaminopyridine (DMAP) in toluene. The solution was mixed and 0.1 ml of PCC (1.8 mM in toluene) was added, vortexed, and held at 50 C for 20 min. The mixture was dried at 50 C under nitrogen. Dried mixtures were dissolved in 0.6 ml of acetonitrile, diluted with 0.4 ml of water, and applied to a C18 Sep-Pak Plus cartridge (Waters

168

C. M. Maragos et al. (sodium salt), 6-monodeoxy-6-monoamino -CD, mono-6-N-allylammonium-6-deoxy--CD chloride, -CD, succinyl- -CD, carboxymethylated -CD, and carboxyethylated -CD. The effect of CDs upon the fluorescence emission of the T2-Pyr was greatly influenced by the type of CD used. The relative fluorescence response of T2-Pyr with each CD is listed in Table I. Of the CDs tested, the three with the greatest effect were DIMEB, HP--CD, and hexakis(2,3,6-triO-methyl)--CD. With a 64-fold enhancement, DIMEB gave the greatest effect and was used in the development of a CE-LIF method for T2-Pyr, discussed below. In aqueous solution DIMEB also provided substantial fluorescence enhancement (up to 30-fold) of T-2 toxin labelled with the 1-anthroylnitrile (data not shown). It is worth noting that DIMEB is the same modified CD found to be among the most effective at improving the fluorescence emission of AFB1, AFG1 (DallAsta et al. 2003), and ZEN (Maragos & Appell 2007). The fluorescence vibronic structure of pyrene depends upon the hydrophobicity of the environment (Kalyanasundaram & Thomas 1977) and the influence of CDs upon the fluorescence of pyrene is well documented (Street 1987). From the effects of -CD, -CD, and -CD upon pyrene fluorescence, it has been suggested that pyrene interacts with but does not completely fit into the cavity of -CD

Corp., Milford, MA, USA) previously conditioned with 7 3 methanol/water (v/v). The cartridge was washed with 6 ml of methanol/water (7 3), and the T2-Pyr was eluted with 1.5 ml acetonitrile into a silane-treated vial. The contents were dried under nitrogen at 50 C. The derivatization of T-2 toxin replaced the C-3 hydroxyl of the toxin with a pyrene ester rendering the product, T2-Pyr, more hydrophobic. One of the aspects of pyrene that makes it useful as a probe for hydrophobicity is the ability to self associate into complexes that have different emission spectra than the monomers. Specifically, pyrene dimers have a fluorescence emission, known as an excimer emission at a wavelength (489 nm) higher than that of the pyrene monomer (389 and 409 nm) (Dyck et al. 2003). Because modifications to CDs can affect the incorporation of guest molecules, we tested 20 CDs for the ability to enhance the fluorescence of T2-Pyr when held in phosphate buffer, pH 7.3, at ambient temperature. The CDs examined included: -CD, methyl--CD, hexakis(2,3,6-triO-methyl)--CD, (2-hydroxy)propyl--CD, carboxymethylated -CD, carboxyethylated -CD, -CD phosphate (sodium salt), -CD, methyl-CD, DIMEB, heptakis(2,3,6-tri-O-methyl)-CD, HP--CD, carboxymethylated -CD, carboxyethylated -CD, -CD phosphate

Table I. Effect of 20 cyclodextrins upon the fluorescence of T2-Pyr.a Relative enhancement of responsec Cyclodextrins No cyclodextrin -Cyclodextrins Modification of cyclodextrinb at 389 nm (1.0) 1.2 2.5 9.6 2.6 1.2 2.4 1.9 3.4 64.3 6.3 12.1 2.0 6.1 1.7 1.0 2.5 4.8 3.4 4.4 1.5 at 489 nm (1.0) 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.8 0.8 0.9 1.0 1.0 1.0 1.0 1.0 0.9 1.0 1.0 0.8

-Cyclodextrins

-Cyclodextrins

Unmodified Methyl Hexakis(2,3,6-tri-O-methyl) (2-Hydroxy)propyl Carboxymethylated Carboxyethylated Phosphate Unmodified Heptakis (2,6-di-O-methyl) Heptakis (2,3,6-tri-O-methyl) (2-Hydroxy)propyl Carboxymethylated Carboxyethylated Phosphate Sulfate 6-Monodeoxy-6-monoamino Unmodified Succinyl Carboxymethylated Carboxyethylated

a Determined using 1 mM T2-Pyr and 10 mg ml1 of the indicated CD. Data were collected with a FluoroMax fluorometer (SPEX, Edison, NJ, USA), 355 nm excitation, with excitation and emission monochromators set to allow 1.5 nm band-pass. b Chemical modification of the indicated cyclodextrin (CD) backbone. c Calculated as the response with 10 mg ml1 added CD divided by the response without added CD.

Cyclodextrins and mycotoxin fluorescence

169

Figure 2. Molecular representation of interactions between T2Pyr and DIMEB. Shown are a possible configuration of a 1:1 ratio (a) and possible configuration of a 1:2 ratio (b) of T2Pyr:DIMEB. Modelling was done with HyperChem, version 7.52 (Gainesville, FL, USA).

Figure 3. Effect of 2 to 50 mM DIMEB upon the fluorescence emission spectrum of 1 mM T2-Pyr. Solutions were in 10 mM sodium phosphate buffer, pH 7.3 (excitation 355 nm). Emission spectra were collected in 1 nm increments with an integration time of 0.25 s.

(Street 1987). Pyrene is known to complex very rapidly with -CD with stoichiometries of 1 : 1, 1 : 2 and 2 : 2 (pyrene: CD) forming within milliseconds (Dyck et al. 2003). Spectroscopic studies suggest a substantial portion of the pyrene resides within the CD cavity and is aligned parallel to its axis (Easton & Lincoln 1999). The stoichiometry and the binding constant of the T2-Pyr/DIMEB inclusion complex were determined by means of the modified Benesi Hildebrand equation (Indirapriyadharshini et al. 2001). In the range of concentrations studied a 1 : 2 (T2-Pyr: DIMEB) stoichiometry was found, with a binding constant of 7700 M1, suggesting a strong interaction between the T-2 derivative and the cyclodextrin. Molecular representations of T2-Pyr inclusion complexes are shown in Figure 2. The configuration relevant to the interaction between T2-Pyr and the first CD molecule, shows the pyrene moiety partially inserted into DIMEB (Figure 2(a)). The iso-valerate portion of T-2 toxin

interacting with a second DIMEB cavity leads to the 1 : 2 configuration depicted in Figure 2(b). The interaction of the T2-Pyr with the DIMEB was greatly influenced by the DIMEB concentration over the range from 2 to 50 mM in phosphate buffer (Figure 3). The dependence of the emission intensity upon the CD concentration is an effect that has also been reported for AFB1, AFG1 (DallAsta et al. 2003), and OTA (Verrone et al. 2007). We observed a gradual decrease in the emission corresponding to the pyrene excimer (489 nm, Figure 3) and a concomitant increase in the emission corresponding to the pyrene monomer (389 nm, Figure 3) as the concentration of DIMEB increased. We speculate that the T2-Pyr stock solution contained T2-Pyr as dimers that, upon addition to the CD solution, dissociate, with the resulting monomers associating with the DIMEB, increasing the fluorescence at 389 nm 409 nm. The hydrophobicity of T2-Pyr suggested that non-aqueous CE, or approaches using high solvent strength or detergents, such as micellar electrokinetic capillary chromatography (MECC), would be necessary in addition to the CD in order for the T2-Pyr to remain soluble. Control maize containing nondetectable levels of T-2 toxin and HT-2 toxin was spiked with T-2 toxin, extracted with 9 1 methanol/ water (v/v), and isolated by immunoaffinity column (IAC) cleanup as described by Visconti et al. (2005). T-2 toxin from samples, or T-2 standards, were derivatized and the T2-Pyr isolated as described above. Derivatized samples were dissolved in 0.15 ml of acetonitrile and 0.45 ml of SDS-CD buffer, composed of 42.5 mM SDS, 2.5 mM sodium borate, 4.3 mM dibasic sodium phosphate and 10 mM DIMEB in water. The electrophoresis buffer consisted of a mixture of 42.5 mM SDS, 2.5 mM sodium borate, 4.3 mM dibasic sodium phosphate, 10 mM DIMEB and 25% (v/v) acetonitrile, pH 9.2. A variety of solvents were tested in combination with DIMEB to arrive at the electrophoresis buffer described above (data not shown). The buffer was found to enhance the emission at 389 nm by 220-fold relative to T2-Pyr in phosphate buffer alone. This seems to be because in phosphate buffer the T2-Pyr appears to exist almost solely as the excimer, while in the more hydrophobic environment of the electrophoresis buffer it exists predominantly as the monomer. Using this buffer the T2-Pyr was separated from the reactants and non-specific products within 10 min (Figure 4). The CE-LIF method could be used to detect T-2 toxin in maize containing as little as 50 ng g1. This demonstrates that detection of a fluorescently derivatized mycotoxin by CE-LIF using CDs is certainly possible. However, the sensitivity of the CE-LIF method relative to an HPLC-FL method

170

C. M. Maragos et al. available for HPLC detectors. Furthermore, with CE-LIF analysis it was necessary to isolate the T2-Pyr after derivatization. This additional step was not required for HPLC analysis of the same derivative. This combination of factors suggests that, while CE-LIF of T2-Pyr is feasible, HPLC detection of this derivative is preferable.

Conclusions Certain of the cyclodextrins, in particular heptakis (2,6-di-O-methyl)--CD (DIMEB), enhance the fluorescence of mycotoxins with a native fluorescence such as the AFB1, AFG1, and ZEN. This effect has proven useful in HPLC and CE-fluorescence-based systems for detecting these toxins. This paper demonstrated that the enhancement of fluorescence produced by CDs can also be observed with fluorescently derivatized T-2 toxin (T2-Pyr). The type of CD used and the concentration were factors that influenced the magnitude of the effect. The most effective of 20 CDs that were tested, DIMEB, was used in the development of a capillary electrophoresis-laser induced fluorescence method for T-2 toxin in maize. The method was capable of detecting as little as 50 ng g1 of T-2 toxin in maize. However, the additional clean-up steps for the CE method, which are not required in the HPLC method, suggest that HPLC is preferred for measuring the T2-Pyr derivative. Despite this result, the ability of CDs to enhance fluorescence of mycotoxins or their derivatives is still a useful effect and suggests examination of aqueous-based systems, where this effect can be used to its full potential, should be further explored.
Figure 4. Electropherograms of T2-Pyr. Standard T-2 equivalent to 50 ng T-2 g1 (a), unspiked control maize (b), and maize spiked with 50 ng T-2 g1 (c). Data were collected with a Beckman Coulter P/ACE MDQ system equipped with a LIF detector (Beckman Coulter, Fullerton, CA, USA). Excitation at 325 nm was provided by a model 100 He/Cd laser (Omnichrome, Chino, CA, USA). Emitted light was filtered through a 405 nm bandpass filter (Andover Corporation, Salem, NH, USA). Before each injection the capillary was rinsed for 2 min at 20 psi with electrophoresis buffer (described in the text). Samples were injected for 5 s at 0.5 psi. Separation was performed with a 75 mm i.d. 50.2 cm fused silica capillary at 30 C by applying 30 KV (approximately 102 mA).

Acknowledgements The authors express their appreciation to Mr John Bobell for excellent technical assistance. They would also like to thank the National Research Council, Italy, for the short Short-Term Mobility Award, 2005. Mention of trade names or commercial products in this paper is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture.

of the same derivative (Lippolis et al., personal communication) was poorer. In part this was likely due to the excitation wavelength selected. The excitation source, a He/Cd laser, provided coherent light at 325 nm, while the excitation maximum for the T2-Pyr was circa 355 nm. Using a 355 nm light source would likely have further improved sensitivity, and such sources are readily

References
Abbas HK, Zablotowicz RM, Weaver MA, Horn BW, Xie W, Shier WT. 2004. Comparison of cultural and analytical methods for determination of aflatoxin production by Mississippi delta Aspergillus isolates. Canadian Journal of Microbiology 50:193199.

Cyclodextrins and mycotoxin fluorescence

Amadasi A, DallAsta C, Ingletto G, Pela R, Marchelli R, Cozzini P. 2007. Explaining cyclodextrin-mycotoxin interactions using a natural force field. Bioorganic and Medicinal Chemistry 15:45854594. Armstrong DW, Alak A, DeMond W, Hinze WL, Riehl TE. 1985. Separation of mycotoxins, polycyclic aromatic hydrocarbons, quinones, and heterocyclic compounds on cyclodextrin bonded phases: An alternative LC packing. Journal of Liquid Chromatography 8:261269. Bohs B, Seidel V, Lindner W. 1995. Analysis of selected mycotoxins by capillary electrophoresis. Chromatographia 41:631637. Cepeda A, Franco CM, Fente CA, Vazquez BI, Rodrguez JL, Prognon P, Mahuzier G. 1996. Postcolumn excitation of aflatoxins using cyclodextrins in liquid chromatography for food analysis. Journal of Chromatography A 721:6974. Chiavaro E, DallAsta C, Galaverna G, Biancardi A, Gambarelli E, Dossena A, Marchelli R. 2001. New reversed-phase liquid chromatographic method to detect aflatoxins in food and feed with cyclodextrins as fluorescence enhances added to the eluent. Journal of Chromatography A 937:3140. Chu FS. 1970. Note on solid state fluorescence emission of ochratoxins A and B on silica gel. Journal of the Association of Official Analytical Chemists 53:696697. DallAsta C, Ingletto G, Corradini R, Galaverna G, Marchelli R. 2003. Fluorescence enhancement of aflatoxins using native and substituted cyclodextrins. Journal of Inclusion Phenomena and Macrocyclic Chemistry 45:257263. Dirr HW, Schabort JC. 1987. Characterization of the aflatoxin B1 binding site of rat albumin. Biochimica et Biophysica Acta 913:300307. Dyck ASM, Kisiel U, Bohne C. 2003. Dynamics for the assembly of pyrene- -cyclodextrin hostguest complexes. Journal of Physical Chemistry B 107:1165211659. Easton CJ, Lincoln SF. 1999. Modified cyclodextrins. Scaffolds and templates for supramolecular chemistry. London: Imperial College Press. p 5. Fente CA, Ordaz JJ, Vazquez BI, Franco CM, Cepeda A. 2001. New additive for culture media for rapid identification of aflatoxin-producing Aspergillus strains. Applied and Environmental Microbiology 67:48584862. Francis Jr OJ, Kirschenheuter GP, Ware GM, Carman AS, Kuan SS. 1988. -Cyclodextrin post-column fluorescence enhancement of aflatoxins for reverse-phase liquid chromatographic determination in corn. Journal of the Association of Official Analytical Chemists 71:725728. Franco CM, Fente CA, Vazquez BI, Cepeda A, Mahuzier G, Prognon P. 1998. Interaction between cyclodextrins and aflatoxins Q1, M1 and P1 fluorescence and chromatographic studies. Journal of Chromatography A 815:2129. Golinski P, Chelkowski J. 1978. Spectral behavior of ochratoxin A in different solvents. Journal of the Association of Official Analytical Chemists 61:586589. Holland RD, Sepaniak MJ. 1993. Qualitative analysis of mycotoxins using micellar electokinetic capillary chromatography. Analytical Chemistry 65:11401146. Hunt DC, Philip LA, Crosby NT. 1979. Determination of ochratoxin A in pigs kidney using enzymic digestion, dialysis and high-performance liquid chromatography with postcolumn derivatization. The Analyst 104:11711175. Indirapriyadharshini VK, Karunanity P, Ramamurthy P. 2001. Inclusion of resorcinol-based acridinedione dyes in cyclodextrins: fluorescence enhancement. Langmuir 17:40564060. Jaimez J, Fente CA, Franco CM, Cepeda A, Vazquez BI. 2004. A survey of the fungal contamination and presence of ochratoxin A and zearalenone on Spanish feed and raw materials. Journal of the Science of Food and Agriculture 84:832840.

171

Kalyanasundaram K, Thomas JK. 1977. Environmental effects on vibronic band intensities in pyrene monomer fluorescence and their application in studies of micellar systems. Journal of the American Chemical Society 99:20392044. Maragos CM, Appell M. 2007. Capillary electrophoresis of the mycotoxin zearalenone using cyclodextrin-enhanced fluorescence. Journal of Chromatography A 1143:252257. Pascale M, Haidukowski M, Visconti A. 2003. Determination of T-2 toxin in cereal grains by liquid chromatography with fluorescence detection after immunoaffinity column clean-up and derivatization with 1-anthroylnitrile. Journal of Chromatography A 989:257264. Ramrez-Galicia G, Garduno-Juarez R, Vargas MG. 2007. Effect of water molecules on the fluorescence enhancement of aflatoxin B1 mediated by aflatoxin B1: -cyclodextrin complexes. A theoretical study. Photochemical and Photobiological Sciences 6:110118. Robertson JA, Pons WA. 1968. Solid state fluorescence emission of aflatoxins on silica gel. Journal of the Association of Official Analytical Chemists 51:11901192. Rojas TR, Sampayo CAF, Vazquez BI, Franco CM, Cepeda A. 2005. Study of interferences by several metabolites from Aspergillus spp. in the detection of aflatoxigenic strains in media added with cyclodextrin. Food Control 16: 445450. Sachse J. 1992. Identification and determination of mullein in cultures of the fungus Septoria nodorum (berk.) by thin-layer and high-performance liquid chromatography. Journal of Chromatography 609:349353. Scott PM. 2002. Methods of analysis for ochratoxin A. In: DeVries JW, Trucksess MW, Jackson LS, editors, Vol. 504 Advances in Experimental Medicine and Biology. New York, NY: Kluwer. pp 117134. Seidel V, Poglits E, Schiller K, Lindner W. 1993. Simultaneous determination of ochratoxin A and zearalenone in maize by reversed-phase high-performance liquid chromatography with fluorescence detection and -cyclodextrin as mobile phase additive. Journal of Chromatography A 635:227235. Street Jr KW. 1987. Cyclodextrin cavity polarity and chromatographic implications. Journal of Liquid Chromatography 10:655662. Van Duuren BL, Chan TZ, Irani FM. 1968. Luminescence characteristics of aflatoxins B1 and G1. Analytical Chemistry 40:20242027. Vazquez ML, Cepeda A, Prognon P, Mahuzier G, Blaus J. 1991. Cyclodextrins as modifiers of the luminescence characteristics of aflatoxins. Analytica Chimica Acta 255:343350. Vazquez ML, Fente CA, Franco CM, Cepeda A, Mahuzier G, Prognon P. 1999. Preliminary study on fluorimetric detection of aflatoxins Q1, P1 and B1 using heptakis-di-O-methyl-cyclodextrin as post-column HPLC reagent. Analytical Communications 36:57. Verrone R, Catucci L, Cosma P, Fini P, Agostiano A, Lippolis V, Pascale M. 2007. Effect of -cyclodextrin on spectroscopic properties of ochratoxin A in aqueous solution. Journal of Inclusion Phenomena and Macrocyclic Chemistry 57:475479. Visconti A, Lattanzio VMT, Pascale M, Haidukowski M. 2005. Analysis of T-2 and HT-2 toxins in cereal grains by immunoaffinity clean-up and liquid chromatography with fluorescent detection. Journal of Chromatography A 1075:151158. Wei J, Okerberg E, Dunlap J, Ly C, Shear JB. 2000. Determination of biological toxins using capillary electrokinetic chromatography with multiphoton-excited fluorescence. Analytical Chemistry 72:13601363.

Food Additives and Contaminants, February 2008; 25(2): 172180

Mycotoxins in cattle feeds and carry-over to dairy milk: A review

JOHANNA FINK-GREMMELS
Faculty of Veterinary Medicine, Division of Veterinary Pharmacology, Pharmacy and Toxicology, Utrecht University, Utrecht, the Netherlands
(Received 17 August 2007; accepted 22 November 2007)

Abstract The complex diet of ruminants, consisting of forages, concentrates, and preserved feeds, can be a source of very diverse mycotoxins that contaminate individual feed components. A number of mycotoxins are successfully inactivated by the rumen flora, whereas others pass unchanged or are converted into metabolites that retain biological activity. Hence, the barrier function of the rumen largely determines the susceptibility of dairy cows and other ruminant species towards individual mycotoxins. An impairment of this barrier function due to diseases or the direct antimicrobial effect of certain mycotoxins may increase absorption rates. The rate of absorption determines not only the internal dose and risk for adverse health effects, but also the excretion of mycotoxins and the biologically active metabolites into milk.

Keywords: Mycotoxins, ruminants, dairy cows, aflatoxins, ergot alkaloids, carry over, milk, biological barriers

Introduction The contamination of feedstuffs with mycotoxins is of increasing concern as changes in agricultural practice and probably climatic changes seem to have increased the prevalence of mycotoxin contamination. Contamination of feeds with mycotoxins accounts for significant economic losses in animal husbandry, as well as in undesirable trade barriers for raw materials and consumable products (Wu 2006). Experimental data and clinical experience suggest that ruminants are less susceptible than other animal species to the adverse health effects associated with mycotoxin exposure. This assumption is based on the finding that the forestomach (rumen) flora can convert a number of mycotoxins into metabolites that are less potent or even biologically inactive at common exposure levels. This does not apply, however, to all mycotoxins that contaminate feed materials. It is the aim of this brief review to identify and describe the uncertainties in the assessment of mycotoxins in the diet of dairy cows in terms of exposure assessment with reference to the physiological and pathophysiological parameters that
Correspondence: Johanna Fink-Gremmels. E-mail: j.fink@uu.nl ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701823142

modulate mycotoxin exposure. Moreover, the mechanisms involved in the excretion of mycotoxins with milk, and potential risk factors associated with the transfer to milk are reviewed.

Mycotoxins in feeds for dairy cattle In professional animal operations, monogastric species such as pigs, poultry, and fish receive a standardized diet designed to meet the nutritional requirements of the species and age group. The components used in the production of these mixed feeds can be monitored and allow the formulation of diets that are well tolerated by the animals. In contrast, the unique physiology of ruminant species, which is characterized by a pre-systemic fermentation and digestion of plant constituents such as cellulose by microbes comprising the ruminal flora, requires feeding regimes that include sufficient amounts of roughage to maintain a functional rumen flora. Genetic selection for high milk yield made it necessary to add increasing quantities of digestible

Mycotoxins in cattle feeds and carry-over to dairy milk: A review energy-rich feed components to the ruminant diet. In extensive farming, grazing makes up a large portion of the diet and the intake of concentrates is limited to a few per cent of the total feed intake. In contrast, in dairy cattle operation, concentrates may feature up to 70% of the daily feed ration. A direct consequence of the complex and variable composition of ruminant diets is the risk of exposure to more than one mycotoxins or mycotoxin cluster; the term cluster refers to a set of mycotoxins produced by an individual fungal species (Table I). The first identified source of mycotoxins in ruminant diets was the contamination of concentrates with aflatoxins. Aflatoxins occur in many typical energy-rich concentrates as, for example, cereal grains, corn gluten, soybean products, as well as in press cakes from oil plants such as peanuts, sunflower seeds, cotton seeds, palm kernels, and copra. Other prominent mycotoxins, such as fumonisins and zearalenone, occur in maize (and maize derived products), whereas cereal grains are contaminated frequently with trichothecenes, particularly with deoxynivalenol, ochratoxins, and ergot alkaloids (Nawaz et al. 1997; Scudamore, Nawaz, Hetmanski 1998; Scudamore, Nawaz, Hetmanski, Rainbird 1998; Placinta et al. 1999). At the same time, ruminants might be exposed to entirely different classes of mycotoxins that occur in forages (pasture grasses), such as the Neotyphodium toxins of the lolitrem-paxilline group and ergovaline, as well as other ergot alkaloids. The level of contamination of (cold season) grasses shows significant geographical differences (Cheeke 1995) and is gaining increasing attention. The third source of mycotoxins in the diet of dairy cows results from the consumption of preserved feeding stuffs such as silage, hay, and straw (OBrien et al. 2005; Mansfield and Kuldau 2007). Particularly after a longer storage period, silage can be spoiled by a variety of fungal species, which are acid-tolerant and micro-aerobe. Mycological investigation identified Penicillium spp. such as P. roqueforti and P. paneum, Aspergilli (A. fumigatus and A. flavus) (Cole et al. 1977), Monascus spp. (Schneweis et al. 2001), and Byssochlamys nivea (Escoula 1975) as the most prevalent fungal species in silage. Mycotoxins produced by these fungi include patulin, mycophenolic acid, penicillic acid, roquefortins, marcfortine A, andrastin A, gliotoxin, and toxins of the verruculogen/fumitremorgen group (Garon et al. 2006; OBrien et al. 2006). It should be mentioned that mycotoxins originating from pre-harvest contamination of forages that are ensiled are often unaffected by the ensiling process, and add to the overall mycotoxin contamination. The ratio in which these different feed sources are used in the diet of dairy cows varies considerably and

173

Table I. Examples of possible co-exposure of dairy cows to mycotoxins in different components of a ruminant diet. Diet component Concentrates Pasture grasses Mycotoxins aflatoxins, fumonisins, zearalenone, trichothecenes (DON), ergot alkaloids lolitrems, paspalitrems, penitrem A, ergovaline and associated ergot alkaloids, trichothecenes patulin, mycophenolic acid, roquefortines, fumitremorgens, verruculogen, monacolines, and others

Preserved feeds (silage)

is determined by regional differences, the production stage of the animal, and farm management. This also implies that a generally applicable exposure assessment is not feasible, and hence data relating exposure to multiple toxins to quantifiable markers of animal health and productivity are scarce.

Clinical mycotoxicoses in dairy cows Typical clinical intoxications that are well described in ruminants are fescue toxicosis and staggers (reviewed by Fink-Gremmels 2005). Both disease complexes are related to pasture grasses that are infected with endophytes. Tall fescue (Festuca arundinacea) is a major forage grass in North America, and infection with the endophyte Neotyphodium coenophialum is associated with the production of various alkaloids, the most prominent of which is ergovaline. Cattle exposed to contaminated forages or hay develops a heat intolerance characterized by malign hyperthermia, and peripheral gangrene (fescue foot), reflecting the vasocontrictive properties of ergot alkaloids. In addition, ergovaline acts as a dopamine-receptor agonist, causing a reduced milk yield and lower conception rates (Browning 2000). Recent data show that following ingestion of ergovaline-contaminated tall fescue straw, not only ergovaline, but also lysergic acid is detectable in the urine and faeces of exposed cattle. A mass-balance study revealed that the toxin concentration in ruminal fluid apparently increases over time (De Lorme et al. 2007). These unexpected findings resulted in the hypothesis that the rumen fermentation processes can liberate non-extractable toxins (escaping the initial feed analysis) and metabolize ergovaline into lysergic acid that might be even be absorbed through the ruminal wall (Hill et al. 2001). Although, certainly, further investigations are necessary to support these assumptions, the data clearly indicate that the rumen metabolism does not necessarily result in toxin inactivation. At the same

174

J. Fink-Gremmels showed that ochratoxin A is mainly degraded by rumen protozoae, and that in healthy cattle up to 12 mg of ochratoxin A per kg feed could be inactivated. This effective deactivation explains the high comparable high tolerance of ruminants to ochratoxin A exposure (Hult et al. 1976; Pettersson et al. 1982). Drastic changes in the feed composition, and a high percentage of protein-rich concentrates in the daily diet modify, however, the cleavage capacity of rumen microorganisms (Xiao et al. 1991; Muller et al. 2001), which explains why incidentally small amounts of ochratoxin A could be detected in milk (Skaug 1999). The susceptibility of ruminants to deoxynivalenol (DON) is low, as DON is converted almost completely into the less toxic DOM (the de-epoxidized metabolite of DON) by the rumen flora. Studies by Ingalls (1996) showed that ruminating cattle may tolerate diets containing up to 8.5 mg g1 DON for several weeks without major health effects. In a recent study, dietary DON concentrations ranging between 3.1 and 3.5 mg g1 feed (88% DM) did not cause any significant adverse health effects, but transiently increased post-prandial ammonia concentrations (Danicke et al. 2005; Seeling, Boghun 2006; Seeling, Danicke, et al. 2006). Aflatoxins are only partly degraded by the ruminal flora, and a typical secondary metabolite of rumen metabolism is aflatoxicol. Exposure to aflatoxins results in an impairment of liver function and reduced feed intake, which might also explain the reduced milk production in dairy cattle exposed to aflatoxins. The impairment of hepatic functions might also account for the photosensitization associated with aflatoxin exposure (Miller and Wilson 1994). Zearalenone is converted by the rumen flora into its hydroxy-metabolite -zearalenol (approximately 90%) and to a lesser extend to -zearalenol (Kiessling et al. 1984; Kennedy et al. 1998). Although -zearalenol has a higher oestrogenic potency compared with the parent zearalenone, its lower rate of absorption and its interconversion in the liver to the less potent -zearalenol might account for the low susceptibility of dairy cattle (Diekman and Green 1992; Danicke et al. 2005; Seeling et al. 2005). Zearalenone and its metabolites can be excreted with milk, but levels are very low often remaining below the limit of quantification (Seeling et al. 2005). Fumonisins pass the rumen, and an intake of 3 mg fumonisin B1 kg1 body weight day1 by Jersey cows for 14 days led to a decreased feed intake and milk production (Richard et al. 1996; Caloni et al. 2000). Signs of intoxication also included an elevated serum enzyme activity of diagnostic liver enzymes (aspartate aminotransferase (AST) and

time, these data also demonstrate the uncertainties in correlating in-feed concentrations of mycotoxins to the internal dose and to predictable biological effects. The staggers syndrome observed in cattle and sheep, and also horses, is associated with the exposure to lolitrems and probably paxilline. Perennial ryegrass (Lolium perenne) harbours the endophyte Neotyphodium lolii, and for many years outbreaks of intoxication were only reported in Australia and New Zealand. However, already in the 1990s, clinical intoxications were reported from North and South America as well as from Europe (Fink-Gremmels 2005, and references therein). Typical clinical symptoms are muscle fasciculation, tremor, and ataxia, which might even progress into tonic convulsions. As probable mechanisms of action, the impairment of the GABAergic pathways (Smith et al. 1997; Wang et al. 2003), was recently challenged by the finding that lolitrem has a potent effect on potassium channel conductance (Dalziel et al. 2005). Neothyphodium lolii produces not only tremorgenic toxins, but also peramine, which exerts a potent repellent effect that protects infected plants from insect plagues and earthworm damage, exemplifying the unique symbiosis between endophytes and their hosts. In addition to these typical mycotoxicoses, the potential adverse effects of other mycotoxins are less well documented in dairy cows. Various reports describe a reduced feed consumption and other adverse health effects associated with the consumption of mouldy feed (hay, silage), and feed materials contaminated with mycotoxins (Osweiler 2000; Hussein and Brasel 2001; Puntenny et al. 2003). A critical analysis of the individual reports indicates, however, obvious gaps in many of the individual case descriptions, as often essential data, such as the amount of feed consumed per day, the animals body weight, the time of exposure, the presence of other contaminants in the diet, and the animal health status, are not reported. This also applies to the reported outbreaks of acute mycotoxicoses such as Aspergillus clavatus toxicosis (Sabater-Vilar et al. 2004), pithomycotoxicosis, which is only incidentally observed in Europe (Pinto et al. 2005), and Diploida maydis toxicosis (Odrizola et al. 2005). Conversion of mycotoxins by the rumen flora Kiessling et al. already stated in 1984 that ruminating animals are developing mycotoxicoses less frequently as the rumen flora acts as a first line of defence against mycotoxins (Kiessling et al. 1984). For example, ochratoxin A is rapidly converted into the less toxic ochratoxin  (lacking the phenylalanine moiety) by the forestomach flora, and only very small amounts of intact ochratoxin A are absorbed. In vitro studies

Mycotoxins in cattle feeds and carry-over to dairy milk: A review gamma-glutamyl-transferase (GGT)), suggesting mild hepatocellular injury. Feeder calves showed signs of immunotoxicity in the form of a significantly reduced lymphoblastogenesis (Osweiler et al. 1993). These effects were observed at feed concentrations that corresponded to exposure rates varying between 2.4 and 3.5 mg g1 body weight. Mould contamination might also change the digestibility of individual feed components. Seeling, Danicke, et al. (2006) describe, for example, an increased crude protein degradation and a lower molar percentage of propionate in the rumen when Fusarium-contaminated wheat was fed to dairy cows. Certain mycotoxins such as, for example, patulin affect the rumen fermentation (Morgavi et al. 2003) and decrease acetic acid production and protein synthesis (Escoula 1992). Taken together, these examples demonstrate the correlation between the capacity of the rumen to inactivate mycotoxins and the likelihood of adverse health effects in cattle. At the same time, it becomes evident that for many toxins that can be expected in the diet of dairy cows the rumen stability and the oral bioavailability have not yet been investigated. In addition, the increasing use of protected concentrated (proteins) that are designed to bypass the rumen might influence the oral bioavailability of mycotoxins.

175

Feed-to-milk transmission of aflatoxins in dairy cows as an example of intra-species variability related to different feeding regimes Aflatoxins are the most intensively studied mycotoxins in dairy cattle as the excretion of aflatoxin M1 in dairy milk is of public health concern. Following ingestion of aflatoxin-contaminated feeds, a part of the ingested aflatoxin B1 is degraded in the rumen, resulting in the formation of aflatoxicol. The remaining fraction is absorbed in the digestive tract by passive diffusion and is hydroxylated in the liver to aflatoxin M1 (Kuilman et al. 2000). Aflatoxin M1 is either conjugated to glucuronic acid, and subsequently excreted via bile, or enters the systemic circulation. Circulating aflatoxin M1 can be excreted in the urine or appear in milk. Initially, the excreted amount of aflatoxin M1 in milk of dairy cows was estimated to represent 12% of the ingested aflatoxin B1 (for a review, see Van Egmond 1989). The extent of transfer from feed to milk (carry-over) is influenced by various nutritional and physiological factors, including feeding regimens, rate of ingestion, rate of digestion, health of the animal, hepatic biotransformation capacity, and actual milk production. This implies that the rate of absorption of aflatoxins, and the

excretion of aflatoxin M1 in milk, varies between individual animals, from day to day, and from one milking to the next. In high-yielding cows, the consumption of significantly higher amounts of concentrated feeds might result in carry-over percentages as high as 6.2% (Veldman et al. 1992). The carcinogenic potency of aflatoxin M1 is almost as high as that of aflatoxin B1, and the toxicological properties are generally comparable (Henry et al. 2001). In consideration of these toxicological findings, many countries have set maximum acceptable levels for aflatoxin M1 in milk and dairy products. Following the risk evaluation by the Joint Expert Committee on Food Additives (JECFA) Codex Alimentarius, regulatory bodies in many countries, including the US Food and Drug Administration (USFDA) set a maximum permissible level for aflatoxin M1 in milk of 0.5 g Kg1. In contrast, in Europe, as well as some countries in Africa, Asia and Latin America, the maximum acceptable level is set at 0.05 g aflatoxin M1 kg1 milk, with reference to the relative high consumption of milk and dairy products by children (reviewed by Van Egmond et al. 2007). To achieve this objective, statutory limits were defined for animal feeds, including feeds for dairy cows. Subsequently, several authors have tried to determine whether the current legislation on aflatoxin B1 in feed (2002/32/EC (OJL 140, 30.05.2002)) for lactating animals is sufficient to keep aflatoxin M1 levels in milk below the threshold of 0.05 g Kg1. Pettersson has already established a model calculation to determine the carry-over of ingested aflatoxin B1 to aflatoxin M1 in milk (Pettersson 1998). This equation was based on ten observations from five controlled experiments, and is expressed as follows (r2 0.915): Aflatoxin M1 ng kg1 milk 10:95 0:787 mg aflatoxin B1 intake day1 : However, a data analysis performed in 2004 on all trials in which daily feeding contained less than 150 g aflatoxin B1 kg1 feed (21 observations from six individual studies) yielded a lower regression coefficient (r2 0.417), pointing towards a larger margin of uncertainty. In addition, a model calculation for a worst-case scenario of aflatoxin carry-over into milk was performed for the major milk-producing animal species, including dairy cattle, sheep, goats, camels, and buffaloes, and included carry-over rates of 2% (assumed average) and 6% (high yielding cows) (European Food Safety Authority (EFSA) 2004). This model calculation indicated that in a worst-case situation, aflatoxin M1 levels in milk might exceed the maximum acceptable level of 0.05 g kg1, set by the European Union, even if

176

J. Fink-Gremmels study in Mexico, conducted between 1996 and 1998, measured aflatoxicol levels in 580 samples of (ultra) pasteurized milk from different regions in Mexico (Carvajal et al. 2003). Aflatoxicol was present in 13% of the samples at concentrations of !0.05 g l1 and in 8% of the samples at !0.5 g l1, and levels were not influenced by pasteurization. These results need to be confirmed as they suggest a need to monitor the occurrence of aflatoxicol in milk and dairy products. The likely reason why aflatoxicol has not been described earlier in milk is the lack of fluorescence of aflatoxicol, while aflatoxin M1, M2 and M4 retain the fluorescence spectrum typical for aflatoxins. The lack of fluorescence requires detection methods that are different from those commonly used of aflatoxin M1.

the given feed materials comply with the current feed legislation. This might occur in all the mentioned animal species. As yet, aflatoxin M1 has been considered to be the major metabolite excreted with milk in dairy cows and other ruminants. In addition, aflatoxin M2 and M4, originating from hepatic-biotransformation reactions of other natural aflatoxins, have been found to be excreted with milk, albeit at very low amounts. However, recent data show that aflatoxicol is also excreted with milk (Carvajal et al. 2003). As mentioned above, aflatoxicol is the major metabolite of aflatoxin B1 produced by microorganisms of the rumen flora. This could be elegantly demonstrated in in vitro studies using radiolabelled aflatoxin B1 (Auerbach et al. 1998). Studies with isolated functional bovine hepatocytes, however, failed to show any formation of aflatoxicol, excluding that hepatic biotransformation contributes to aflatoxicol tissue levels (Kuilman et al. 2000). The carcinogenicity of aflatoxicol has been investigated only in the rainbow trout, an experimental animal model known to be very sensitive to the hepato-carcinogenicity of aflatoxin B1. Results demonstrated that the carcinogenic potency of aflatoxicol is comparable with that of aflatoxin B1, and that it is even more potent than aflatoxin M1 in this model (Hendricks et al. 1980; Schoenhard et al. 1981; Hendricks 1994). A recent

Feed-to-milk transmission of other mycotoxins and factors affecting transmission rates As yet, aflatoxin M1 is the only mycotoxins for which maximum permissible levels in milk have been established. However, considering the wide range of mycotoxins that might occur in ruminant diets, the number of available studies addressing the transfer of mycotoxins into milk is very limited. Table II provides a summary of the available data (for detailed

Table II. Products of ruminal bioconversion and transfer of mycotoxins from feed to milka. Main product of rumen metabolism aflatoxicol aflatoxin M1d unchanged unchanged ochratoxin- various de-epoxy-DON (DOM) -zearalenol unchanged unchanged unchanged Reduction of biological potency minor minor unchanged unchanged significantf significant significant none unchanged unchanged unchanged Estimated carry-over rates n.d.b 012.4 g l1c 2.06.2% n.d. 6.40.7 g l1e 00.05% n.d. 0.052% DON: 0.00010.0002 DOM: 0.00040.0024g 0.060.08%h n.d. n.d. n.d.

Mycotoxin Aflatoxin B1 Cyclopiazonic acid Fumonisin B1 Ochratoxin A T-2 toxin DON (and related trichothecenes) Zearalenone Patulini Ergovalin Lolitrem

Note: aAccording to Galtier (1998, 1999), Yiannikouris and Jouany (2002), and other sources as indicated. b n.d., Not determined. c Aflatoxicol has been detected, however, in commercial milk samples (Carvajal et al. 2003). d Aflatoxin M1 is not a product of rumen metabolism but originates from hepatic metabolism of aflatoxin B1. e According to Oliviera et al. (2006). f Ochratoxin- is considered to be less toxic than ochratoxin A, but it can be esterified to yield ochratoxin C, which is a toxic form. g According to Seeling, Boghun et al. (2006). h Total milk analysis shows also minor concentrations of -zearalenol. i Patulin is metabolised in the liver.

Mycotoxins in cattle feeds and carry-over to dairy milk: A review references, see Jouany and Diaz 2005). These studies, addressing the transfer of mycotoxins into milk, have been conducted in healthy animals with an intact bloodmilk barrier. However, various systemic diseases and local (mammary) infections might alter the functionality of this barrier and, hence, transmission rates may be higher in daily practice. The bloodmilk barrier comprises different anatomical structures and active transport processes. The physical barrier is formed by the epithelium of the blood capillaries that span and supply the secretory epithelium of the mammary gland. Polar substances and large molecules cannot pass this barrier by passive diffusion. Factors that determine the excretion with milk are the molecular weight and lipophilicity of a compound (including mycotoxins and their metabolites), as well as the degree of binding to plasma proteins, as only the unbound fraction can be transported. The transport rate is also influenced by the pH gradient between blood plasma and milk. In a healthy animal, the pH of milk is lower than the plasma pH, whereas in a diseased animal (suffering, for example, from mastitis) the pH of milk is equal to or even exceeds the blood plasma pH. These differences modulate the rate of transfer into milk as demonstrated for various drugs. Recently, a distinct class of transmembrane transporters, which facilitate the active excretion of endogenous and exogenous compounds from the bloodstream into milk, gains increasing attention. In the mammary gland, BCRC (the gene product of ABC transporter ABCG2) has been described as a major excretory transporter (Borst and Oude Elferink 2002). Substrates for these transporters are likely to appear in high concentrations in milk. Recently, it was shown that, for example, ochratoxin A is a substrate for BCRP (Schrickx et al. 2006), which correlates with the high prevalence of ochratoxin A in human breast milk samples. In addition, it has been shown that aflatoxin B1 is a substrate for BCRP, making it likely that aflatoxin M1 and aflatoxicol are also excreted actively into milk (Van Herwaarden et al. 2006). These findings offer the possibility of estimating the likelihood of galactogenic excretion through rapid in vitro assays, which can be applied to all chemical classes of mycotoxins. It is worthwhile mentioning that investigations devoted to the galactogenic excretion of various veterinary medicinal products did show significant differences in rumen metabolism and transfer into milk between individual ruminating species, such as dairy sheep, goats, and buffaloes (Merino et al. 2006). The increasing market for milk products from these animal species underlines the need for data from these animals as well.

177

The excretion of mycotoxins with milk is generally reviewed with respect to potential adverse effects on human health, particularly children, who are high milk consumers. Contamination of dairy milk with mycotoxins might, however, also impair milk quality and the use of milk for typical fermented dairy products such as yoghurt. As mentioned above, various mycotoxins, particularly those from silage moulds, exert strong antimicrobial effects. Even minor amounts of these toxins might affect milk technologies and the control of tank milk for undesirable residues of therapeutic antibiotics (false-positive results).

Current uncertainties in the assessment mycotoxins in the diet of dairy cows The uncertainties in the exposure of dairy cows are attributable to significant differences in the composition of individual diets, depending on the feeding regimen, the availability of natural pastures, and the methods of feed preservation. Subsequently, the animals are potentially exposed to highly variable and complex mixtures of mycotoxins, and the health consequences of these mixtures are difficult to assess. Major points of interests are as follows: . Various mycotoxins have the ability to modify the rumen flora due to their antimicrobial activity. This may decrease the degrading capacity of the rumen resulting in an unexpected passage rate of intact toxins from other sources. A comparable effect can be also be expected in cases in which the rumen flora is affected in the course of metabolic diseases, as, for example, rumen acidosis. . Toxintoxin interactions at the level of absorption and biotransformation are likely, but the clinical significance of these interactions remains to be elucidated. . The excretion of mycotoxins with milk is generally low. Changes in the bloodmilk barrier due to systemic, and particularly local, infections (mastitis) affect the integrity of the bloodmilk barrier and the pH gradient between blood and milk. This may, in turn, alter the rate of excretion and facilitate the excretion of mycotoxins that are not expected in milk. As mentioned above, a number of recent reports refer to the likelihood that the excretion of mycotoxins influences the standard tests for undesirable residues of antibiotics in milk. . Various mycotoxins exert a modulating effect on the immune system, even at low doses. This effect might result in an increased prevalence of infectious diseases

178

J. Fink-Gremmels or an acceleration of minor infections. An increased incidence of mastitis and lower leg problems in dairy cows has been associated with a poor quality of the given silage (Nyman et al. 2007). To what extend this phenomenon is directly attributable to mycotoxins in the silage needs to be investigated.
nutrient turnover, microbial protein synthesis and metabolism of deoxynivalenol and zearalenone in the rumen of dairy cows. Journal of Animal Physiolology Animal Nutrition (Berl) 89:303315. De Lorme MJ, Lodge-Ivey SL, Craig AM. 2007. Physiological and digestive effects of Neotyphodium coenophialum-infected tall fescue fed to lambs. Journal of Animal Science 85:11991206. Diekman MA, Green ML. 1992. Mycotoxins and reproduction in domestic livestock. Journal of Animal Science 70:16151627. Escoula L. 1975. [Toxinogenic moulds of silage. IV. Patulin production in liquid medium using fungus species isolated in silages (authors trans.)]. Annual Research Veterinary 6:303310. Escoula L. 1992. Patulin production by Penicillium granulatum and inhibition of ruminal flora. Journal of Environmental Pathology, Toxicology and Oncology 11:4548. European Food Safety Authority (EFSA). 2004. Opinion of the Scientific Panel on Contaminants in the Food Chain on a request from the Commission related to Aflatoxin B1 as undesirable substance in animal feed. Request No. EFSA-Q2003-035. EFSA. Parma, Italy. Fink-Gremmels J. 2005. Mycotoxins in forages. In: Diaz DE, editor. The mycotoxin blue book. Nottingham (UK): Nottingham University Press. pp 249268. Galtier P. 1998. Biological fate of mycotoxins in animals. Revue de medecine veterinaire 149:549554. Galtier P. 1999. Biotransformation and fate of mycotoxins. Journal of Toxicology Toxin Review 18:295312. Garon D, Richard E, Sage L, Bouchart V, Pottier D, Lebailly P. 2006. Mycoflora and multimycotoxin detection in corn silage: experimental study. Journal of Agricultural and Food Chemistry 54:34793484. Hendricks JD. 1994. Carcinogenicity of aflatoxins in nonmammalian organisms. In: Eaton DL, Groopman JD, editors. The toxicology of aflatoxins: human health, veterinary, and agricultural significance. San Diego (CA): Academic Press. p 103. Hendricks JD, Wales JH, Sinnhuber RO, Nixon JE, Loveland PM, Scanlan RA. 1980. Rainbow trout (Salmo gairdneri) embryos: a sensitive animal model for experimental carcinogenesis. Federation Proceedings 39:32223229. Henry SH, Whitaker T, Rabbini I, Bowers J, Park D, Price WD, Bosch FX, Pennington J, Verger P, Yoshizawa T, et al. 2001. Aflatoxin M1. In: Safety evaluation of certain mycotoxins in food. Prepared by the Fifty-sixth Meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA). FAO Food and Nutrition Paper No. 74. Rome (Italy): Food and Agriculture Organization of the United Nations. Hill NS, Thompson FN, Stuedemann JA, Rottinghaus GW, Ju HJ, Dawe DL, Hiatt EE III. 2001. Ergot alkaloid transport across ruminant gastric tissues. Journal of Animal Science 79:542549. Hult K, Teiling A, Gatenbeck S. 1976. Degradation of ochratoxin A by a ruminant. Applied and Environmental Microbiology 32: 443444. Hussein HS, Brasel JM. 2001. Toxicity, metabolism, and impact of mycotoxins on humans and animals. Toxicology 167:101134. Ingalls JR. 1996. Influence of deoxynivalenol on feed consumption by dairy cows. Animal Feed Science and Technology 60:297300. Jouany J, Diaz D. 2005. Effects of mycotoxins in ruminants. In: Diaz DE, editor. The mycotoxin blue book. Nottingham (UK): Nottingham University Press. pp 295321.

In conclusion, dairy cows are protected against exposure to mycotoxins by their rumen flora. Various mycotoxins, however, pass this barrier or are converted into metabolites that retain biological activity. The assessment of undesirable effects exerted in ruminants should include the antimicrobial activity of various mycotoxins that results in an impairment of the function of the rumen flora, followed by a poor feed utilization and reduced weight gain and productivity.

Acknowledgements The valuable support of Dr N. Visser in finalizing this manuscript was highly appreciated.

References
Auerbach H, Maas RFM, Op den Camp HJM, Pol A, Fink-Gremmels J. 1998. Biodegradation of Aflatoxin B1 by bovine rumen micro-organisms in vitro and its effects on rumen fermentation. Proceedings of Mycotox 98 (Toulouse). Revue de Medecine Veterinaire 573. Borst P, Oude Elferink R. 2002. Mammalian ABC transporters in health and disease. Annual Review of Biochemistry 71:537592. Browning Jr R. 2000. Physiological responses of Brahman and Hereford steers to an acute ergotamine challenge. Journal of Animal Science 78:124130. Caloni F, Spotti M, Auerbach H, Op den Camp H, Fink-Gremmels J, Pompa G. 2000. In vitro metabolism of fumonisin B1 by ruminal microflora. Veterinary Research Communications 24:379387. Carvajal M, Rojo F, Mendez I, Bolanos A. 2003. Aflatoxin B1 and its interconverting metabolite aflatoxicol in milk: the situation in Mexico. Food Additives and Contaminants 20:10771086. Cheeke PR. 1995. Endogenous toxins and mycotoxins in forage grasses and their effects on livestock. Journal of Animal Science 73:909918. Cole RJ, Kirksey JW, Dorner JW, Wilson DM, Johnson Jr J, Bedell D, Springer JP, Chexal KK, Clardy J, Cox RH. 1977. Mycotoxins produced by Aspergillus fumigatus isolated from silage. Annual Nutrition Alimentation 31:685691. Dalziel JE, Finch SC, Dunlop J. 2005. The fungal neurotoxin lolitrem B inhibits the function of human large conductance calcium-activated potassium channels. Toxicological Letters 155:421426. Danicke S, Matthaus K, Lebzien P, Valenta H, Stemme K, Ueberschar KH, Razzazi-Fazeli E, Bohm J, Flachowsky G. 2005. Effects of Fusarium toxin-contaminated wheat grain on

Mycotoxins in cattle feeds and carry-over to dairy milk: A review

Kennedy DG, Hewitt SA, McEvoy JD, Currie JW, Cannavan A, Blanchflower WJ, Elliot CT. 1998. Zeranol is formed from Fusarium spp. toxins in cattle in vivo. Food Additives and Contaminants 15:393400. Kiessling KH, Pettersson H, Sandholm K, Olsen M. 1984. Metabolism of aflatoxin, ochratoxin, zearalenone, and three trichothecenes by intact rumen fluid, rumen protozoa, and rumen bacteria. Applied and Environmental Microbiology 47:10701073. Kuilman ME, Maas RF, Fink-Gremmels J. 2000. Cytochrome P450-mediated metabolism and cytotoxicity of aflatoxin B(1) in bovine hepatocytes. Toxicology In Vitro 14:321327. Mansfield MA, Kuldau GA. 2007. Microbiological and molecular determination of mycobiota in fresh and ensiled maize silage. Mycologia 99:269278. Merino G, Alvarez AI, Pulido MM, Molina AJ, Schinkel AH, Prieto JG. 2006. Breast cancer resistance protein (BCRP/ ABCG2) transports fluoroquinolone antibiotics and affects their oral availability, pharmacokinetics, and milk secretion. Drug Metabolism and Disposition 34:690695. Miller DM, Wilson DM. 1994. Veterinary diseases related to aflatoxins. In: Eaton DL, Groopman JD, editors. The toxicology of aflatoxins. San Diego (CA): Academic Press. pp 347364. Morgavi DP, Boudra H, Jouany J-P, Graviou D. 2003. Prevention of patulin toxicity on rumen microbial fermentation by SH containing reducing agents. Journal of Agricultural and Food Chemistry 51:69066910. Muller HM, Muller K, Steingass H. 2001. Effect of feeding regime on the metabolism of ochratoxin A during the in vitro incubation in buffered rumen fluid from cows. Archiv fur Tierernahrung 54:265279. Nawaz S, Scudamore KA, Rainbird SC. 1997. Mycotoxins in ingredients of animal feeding stuffs: I. Determination of Alternaria mycotoxins in oilseed rape meal and sunflower seed meal. Food Additives and Contaminants 14:249262. Nyman AK, Ekman T, Emanuelson U, Gustafsson AH, Holtenius K, Waller KP, Sandgren CH. 2007. Risk factors associated with the incidence of veterinary-treated clinical mastitis in Swedish dairy herds with a high milk yield and a low prevalence of subclinical mastitis. Preventive Veterinary Medicine 78:142160. OBrien M, Nielsen KF, OKiely P, Forristal PD, Fuller HT, Frisvad JC. 2006. Mycotoxins and other secondary metabolites produced in vitro by Penicillium paneum Frisvad and Penicillium roqueforti Thom isolated from baled grass silage in Ireland. Journal of Agricultural and Food Chemistry 54:92689276. OBrien M, OKiely P, Forristal PD, Fuller HT. 2005. Fungi isolated from contaminated baled grass silage on farms in the Irish Midlands. FEMS Microbiol Lett 247:131135. Odrizola E, Odeon A, Cabton G, Clemente G, Escande A. 2005. Diploida maydis: a cause of death of cattle in Argentina. New Zealand Veterinary Journal 53:160161. Oliviera CA, Rosmaninho J, Rosim R. 2006. Aflatoxin M1 and cyclopiazonic acid in fluid milk traded in Sao Paulo, Brazil. Food Additives and Contaminants 23:196201. Osweiler GD. 2000. Mycotoxins. Contemporary issues of food animal health and productivity. The Veterinary Clinics of North America. Food Animal Practice 16:511530, vii. Osweiler GD, Kehrli ME, Stabel JR, Thurston JR, Ross PF, Wilson TM. 1993. Effects of fumonisin-contaminated corn screenings on growth and health of feeder calves. Journal of Animal Science 71:459466.

179

Pettersson H. 1998. Concerning Swedish derogation on aflatoxin. Complement to the Memo of 97-03-03 on Carry-over of aflatoxin from feedingstuffs to milk. Uppsala (Sweden): Department of Animal Nutrition and management, Swedish University of Agricultural Sciences. Pettersson H, Kiessling KH, Ciszuk P. 1982. Degradation of ochratoxin A in rumen. In: Proceedings of the Vth International IUPAC Symposium Mycotoxins and Phycotoxins. Vienna (Austria): Austrian Chemical Society. Pinto C, Santos VM, Dinis J, Peleteiro MC, Fitzgerald JM, Hawkes AD, Smith BL. 2005. Pithomycotoxicosis (facial eczema) in ruminants in the Azores, Portugal. The Veterinary Record 157:805810. Placinta CM, dMello JPF, MacDonald EK. 1999. A review of worldwide contamination of animal feed with fusavium mycotoxins. Animal Feed Science and Technology 78:2137. Puntenny SB, Wang Y, Forsberg NE. 2003. Mycotic infections in livestock: recent insight and studies on aetiology, diagnosis and presentation of hemorrhagic bowel syndrome. Paper presented at the Southwest Nutrition & Management Conference, Phoenix, Az, USA. Tuscon (AZ): University of Arizona, Department of Animal Science. p. 4963. Richard JL, Meerdink G, Maragos CM, Tumbleson M, Bordson G, Rice LG, Ross PF. 1996. Absence of detectable fumonisins in the milk of cows fed Fusarium proliferatum (Matsushima) Nirenberg culture material. Mycopathologia 133:123126. Sabater-Vilar M, Maas RF, De Bosschere H, Ducatelle R, FinkGremmels J. 2004. Patulin produced by Aspergillus clavatus isolated from feed containing malting residues associated with a lethal neurotoxicosis in cattle. Mycopathologia 158:419426. Schneweis I, Meyer K, Hormansdorfer S, Bauer J. 2001. Metabolites of Monascus ruber in silages. Journal of Animal Physiolology Animal Nutrition (Berl) 85:3844. Schoenhard GL, Hendricks JD, Nixon JE, Lee DJ, Wales JH, Sinnhuber RO, Pawlowski NE. 1981. Aflatoxicol-induced hepatocellular carcinoma in rainbow trout (Salmo gairdneri) and the synergistic effects of cyclopropenoid fatty acids. Cancer Research 41:10111014. Schrickx J, Lektarau Y, Fink-Gremmels J. 2006. Ochratoxin A secretion by ATP-dependent membrane transporters in Caco-2 cells. Archives Toxicology 80:243249. Scudamore KA, Nawaz S, Hetmanski MT, Rainbird SC. 1998b. Mycotoxins in ingredients of animal feeding stuffs: III. Determination of mycotoxins in rice bran. Food Additives and Contaminants 15:185194. Scudamore KA, Nawaz S, Hetmanski MT. 1998a. Mycotoxins in ingredients of animal feeding stuffs: II. Determination of mycotoxins in maize and maize products. Food Additives and Contaminants 15:3055. Seeling K, Boghun J, Strobel E, Danicke S, Valenta H, Ueberschar KH, Rodehutscord M. 2006. On the effects of Fusarium toxin contaminated wheat and what chaff on nutrient utilization and turnover of deoxynivalenol and zearlanone in vitro (Rusitec). Toxicology in Vitro 20:703711. Seeling K, Danicke S, Ueberschar KH, Lebzien P, Flachowsky G. 2005. On the effects of Fusarium toxin-contaminated wheat and the feed intake level on the metabolism and carry over of zearalenone in dairy cows. Food Additives and Contaminants 22:847855. Seeling K, Danicke S, Valenta H, Van Egmond HP, Schothorst RC, Jekel AA, Lebzien P, Schollenberger M, Razzazi-Fazeli E, Flachowsky G. 2006a. Effects of Fusarium toxin-contaminated wheat and feed intake level on the biotransformation and carryover of deoxynivalenol in dairy cows. Food Additives and Contaminants 23:10081020.

180

J. Fink-Gremmels
carcinogens aflatoxin B1, IQ and Trp-P-1 but also mediates their secretion into breast milk. Carcinogenesis 27:123130. Veldman A, Meijst JAC, Borggreve GJ, Heeres-van Tol JJ. 1992. Carry-over of aflatoxin from cows food to milk. Animal Production 55:163168. Wang L, Cross AL, Allen KL, Smith BL, McLeay LM. 2003. Tremorgenic mycotoxins increase gastric smooth muscle activity of sheep reticulum and rumen in vitro. Research in Veterinary Science 74:93100. Wu F. 2006. Mycotoxin reduction in Bt corn: potential economic, health, and regulatory impacts. Transgenic Research 15:277289. Xiao H, Marquardt RR, Frohlich AA, Ohilipps GD, Vitti TG. 1991. Effect of a hay and grain diet in the rate of hydrolysis of ochratoxin A in the rumen of sheep. Journal of Animal Science 69:37063014. Yiannikouris A, Jouany J-P. 2002. Mycotoxins in feeds and their fate in animals: a review. Animal Research 51:8199.

Skaug MA. 1999. Analysis of Norwegian milk and infant formulas for ochratoxin A. Food Additives and Contaminants 16:7578. Smith BL, McLeay LM, Embling PP. 1997. Effect of the mycotoxins penitrem, paxilline and lolitrem B on the electromyographic activity of skeletal and gastrointestinal smooth muscle of sheep. Research in Veterinary Science 62:111116. Van Egmond HP, Schothorst RC, Jonker MA. 2007. Regulations relating to mycotoxins in food: perspectives in a global and European context. Analytical and Bioanalytical Chemistry 389:147157. Van Egmond HP. 1989. Aflatoxin M1: occurrence, toxicity, regulation. In: Van Egmond HP, editor. Mycotoxins in dairy products. London (UK): Elsevier Applied Science. Van Herwaarden AE, Wagenaar E, Karnekamp B, Merino G, Jonker JW, Schinkel AH. 2006. Breast cancer resistance protein (Bcrp1/Abcg2) reduces systemic exposure of the dietary

Food Additives and Contaminants, February 2008; 25(2): 181192

Mycotoxins in botanicals and dried fruits: A review

M. W. TRUCKSESS1 & P. M. SCOTT2

Food and Drug Administration, 5100 Paint Branch Parkway, College Park, MD 20740, USA and 2Health Canada, 2203D, 251 Sir Frederick Banting Driveway, Ottawa, Ontario K1A 0K9, Canada
(Received 18 May 2007; accepted 10 July 2007)
1

Abstract Botanicals are used in many countries for medicinal and general health-promoting purposes. Numerous natural occurrences of mycotoxins in botanicals and dried fruits have been reported. Aflatoxins or ochratoxin A (OTA) have been found in botanicals such as ginseng, ginger, liquorice, turmeric, and kava-kava in the USA, Spain, Argentina, India, and some other countries, while fumonisins have been found in medicinal wild plants in South Africa and in herbal tea and medicinal plants in Turkey. Zearalenone was identified in ginseng root. Dried fruits can be contaminated with aflatoxins, OTA, kojic acid, and, occasionally, with patulin or zearalenone. One main area of concern is aflatoxins in dried figs; bright greenish yellow fluorescence under ultraviolet light is associated with aflatoxin contamination. OTA in dried vine fruits (raisins, sultanas, and currants) is another concern. There are also reports of aflatoxins in raisins and OTA in dried figs, apricots, dried plums (prunes), dates, and quince. Maximum permitted levels in the European Union include 4 mg kg1 for total aflatoxins in dried fruit intended for direct consumption and 10 mg kg1 for OTA in dried vine fruit. This review discusses the occurrence of mycotoxins in botanicals and dried fruits and analytical issues such as sampling, sample preparation, and methods for analysis. Fungal contamination of these products, the influence of sorting, storage, and processing, and prevention are also considered.

Keywords: Botanicals, dried fruits, garlic, ginseng, ginger, capsicum, liquorice, raisins, figs, fumonisins, ochratoxin A, aflatoxins

Introduction Botanical products and various dried fruits are in great demand in the health food markets. As people of differing nationalities live in the same communities, ethnic foods become increasingly popular and available. Varying processing and storage conditions can provide mould growth and mycotoxin development. This review will discuss the incidence and occurrence, methods of analysis, and prevention of occurrence of mycotoxins in botanicals and dried fruits. There have been many investigations of the occurrence of mycotoxins in these products. Dried fruits, particularly dried vine fruit (raisins, sultanas and currants), and figs (Drusch and Ragab 2003; Drusch and Aumann 2005) have been of considerable interest.

Mycotoxins in botanicals The term botanicals simply means plants or plant products. Usually botanicals have been understood as medicinal plants and botanical supplements. In many cases it is difficult to distinguish between these two categories. Numerous botanical products enter markets around the world as foods and as dietary supplements. The food or herbal supplements may be whole plants, plant parts or plant extracts. Medicinal plants have been used as medicines for centuries and some of them are now used in developed countries as alternative or complementary medicines. Traditional remedies have been used in China for over 5000 years, and currently it is estimated that half of all healthcare delivered in China is based on traditional Chinese

Correspondence: M. W. Trucksess. E-mail: mary.trucksess@fda.hhs.gov ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701567459

182

M. W Trucksess & P. M. Scott . but very few studies on occurrence of mycotoxins in botanicals. It may be due to the lack of methods of analysis suitable for determining mycotoxins in botanicals as the use of botanicals as supplements is relatively new. This section covers mycotoxin contamination in medicinal plants and some of the common botanical supplements such as garlic, ginseng, ginger, capsicum, turmeric, liquorice roots, kava-kava (also known as kava) and red mould rice (RMR) (also known as red yeast rice). Medicinal plants Almost all countries have indigenous plant species that have been used to prevent illness or cure diseases. Surveys of fumonisins in medicinal plants from 2001 to 2006 indicated that widespread low levels of fumonisin B1 (FB1) occurred in Portugal, Turkey, and South Africa. In Portugal, a total of 87 samples (69 samples of four different medicinal plants and 18 samples of black tea) were purchased from supermarkets and analysed by liquid chromatography (LC) for FB1 and fumonisin B2 (FB2) (Martins et al. 2001). Fifty-five samples were contaminated with FB1, but no FB2 was found in any of these samples. Sixteen samples of the black tea and 39 samples of the medicinal plants contained FB1 at levels ranging from 80 to 280 mg kg1 and from 20 to 700 mg kg1 FB1, respectively. In Turkey, a total of 115 herbal tea and medicinal plant samples were analysed using an LC method. FB1 was detected in two samples at 160 and 1487 mg kg1 and no FB2 was found (Omurtag and Yazicioglu 2004). In South Africa, 30 medicinal plants were analysed for FB1 and aflatoxin B1 (AFB1) using immunoaffinity column (IAC) clean-up and LC with fluorescence (FL) detection (Sewram et al. 2006). None of the plant extracts contained AFB1. Four samples were positive for FB1 at levels ranging from 8 to 1553 mg kg1 and its presence was confirmed by LC-tandem mass spectrometry (LC-MS/MS). Studies from China, Egypt, India, Sri Lanka, Thailand, Malaysia, and Indonesia reported finding AFB1 in medicinal plants and one study from India found citrinin (CIT) in addition to AFB1. Three of 19 traditional Chinese medicines contained aflatoxins (up to 28 mg kg1; Yang et al. 2005). In Egypt, nine of the 31 herbs and medicinal plants analysed contained average AFB1 levels of 49 mg kg1 by LC with ultraviolet light (UV) detection (Selim et al. 1996). In India, 14 out of 15 different drug plant samples collected from storehouses contained AFB1 at 0.11.2 mg kg1 (Roy et al. 1988). In another study in India, 36 out of 60 samples of seeds of medicinal plants were positive for AFB1 at levels ranging from 20 to

medicines (Shaw 1998). A 1998 report shows that 89% of people in developing countries rely on traditional herbal medicines. These are widely available to consumers worldwide and some of them are used as alternative medicines. Herbal and other alternative treatments are also popular in the developed world, being used by about 50% of Australians and 33% of Americans (Shaw 1998). In industrialized countries it is believed that between 30 and 50% of the population regularly use herbal medicines and/or vitamin and mineral supplements. The most common reasons given for taking supplements are it makes me feel better and to live longer. In the USA since 1994, with the passage of the Dietary Supplement Health and Education Act (DSHEA) (US Food and Drug Administration (FDA) 1994), herbal supplements easily enter commercial markets. Herbal supplement usage has increased tremendously since the passage of the DSHEA. The DSHEA states that the government cannot prohibit the sale of dietary supplements unless the products prove to be unsafe. Consequently, they are sold over the counter in drug stores, supermarkets, health supply shops and through the internet. Some are used daily by consumers for various reasons. In many cases, the functions and toxicities of the inherent bioactive compounds in the botanicals are largely unknown. Contamination with chemicals such as mycotoxins, heavy metals, pesticides and synthetic drugs, microorganisms such as bacteria and fungi, or undeclared constituents can contribute to adverse human health problems (Schilter et al. 2003).

Incidence and occurrence In spite of the long history and wide use of botanicals as foods and traditional medicines, there are few publications on their contamination with moulds and mycotoxins compared with numerous publications on the contamination of grains and oil seeds. Several surveys of toxigenic moulds in botanicals have found high levels of Aspergillus, Penicillium and Fusarium spp. (Abeywickrama and Bean 1991; Halt 1998; Rizzo et al. 2004). While the presence of moulds might not be correlated with the presence of mycotoxins, there are reports of aflatoxins, ochratoxin A (OTA) and fumonisins in medicinal plants, tea and other botanicals. Raw materials for medicinal use and herbal supplements are frequently contaminated with toxigenic fungi generated from the soil, or the plants themselves, during harvesting or in storage. Contamination with mycotoxins produced by these fungi could pose human health problems. There are more than 20 000 botanicals available in the market,

Mycotoxins in botanicals and dried fruits 1180 mg kg1 and 11 were positive for CIT at 10760 mg kg1 (Roy and Kumari 1991). OTA was found at levels up to 2340 mg kg1 in 55 out of 129 herbal samples destined for the preparation of Ayurvedic medicines (Roy and Kumar 1993). Other work from the same group in India found 23 out of 50 samples of liver curative herbal medicines contained AFB1; the maximum level was 2230 mg kg1 in Asparagus racemosus (Kumar and Roy 1993). In Sri Lanka, 500 mg kg1 AFB1 was detected by thin-layer chromatography (TLC) in one of six Asian medicinal plants, Aerra lanata (Abeywickrama and Bean 1991). In Thailand, five out of 28 herbal medicinal products were contaminated with aflatoxins at 1.714.3 mg kg1 using an IAC/LC method (Tassaneeyakul et al. 2004). In Malaysia and Indonesia, 16 of the 23 traditional herbal medicines jamu and makjun analysed by an IAC/LC method contained a low level of AFB1 (0.3 mg kg1) (Ali et al. 2005). In Croatia, seven samples of medicinal plant material were analysed for AFB1, OTA, and zearalenone (ZON). Only a trace amount of OTA was found in one sample (Halt 1998). There was no mention of the method of analysis and the limit of detection (LOD). In Italy, 27 aromatic herbs, 48 herbal infusions and medicinal plants were analysed for aflatoxins by LC with post-column derivatization and FL detection; none of the samples was contaminated, even if it originated from tropical countries (Romagnoli et al. 2007). In Japan, 49 powdered herbal drugs were analysed for aflatoxins, sterigmatocystin and OTA; mycotoxins were not detected in any samples (Hitokoto et al. 1978). Aflatoxins were found in one out of ten tablets of Cascara sagrada dried bark sold in Argentina as well as in two out of nine samples of the raw material (Rizzo et al. 1999). Garlic Garlic is the most widely used botanical supplement in the USA. People take garlic supplements for the following uses: antibacterial function, antifungal function, to lower blood pressure, to lower cholesterol levels, to improve the circulation, as a cardioprotective, as an anti-oxidant, and for coughs and colds, stomach conditions and cancer treatment. Many of these uses have not been evaluated by sound scientific studies. There are only two reports on analysis for aflatoxin in garlic. In the UK, a survey of mycotoxins in ethnic food including four garlic samples was conducted. No aflatoxins were found at 40.1 mg kg1 (Patel et al. 1996). Another study found one garlic powder to contain 3.3 mg kg1 total aflatoxins (MacDonald and Castle 1996). The same study showed that aflatoxin levels in spiced

183

sauces were not reduced by cooking or microwave heating. Ginseng Ginseng has been used in Chinese medicine for more than 5000 years. The term ginseng means man root in Chinese. The botanical name Panax means heal all in Greek. There are only two reports of finding aflatoxins in ginseng: one in ginseng roots and one in ginseng supplements. In the USA, 11 simulated wild and 12 cultivated ginseng root samples were analysed by DOvidio et al. (2006). All cultivated roots were found to be aflatoxin-free. Two of the simulated wild roots contained aflatoxins at 15.1 and 15.2 mg kg1. One mouldy ginseng root purchased from a grocery store was contaminated with aflatoxins at 16 mg kg1. Ten ginseng products were purchased from herbal supply stores, grocery stores and drug stores in the USA (Trucksess et al. 2007). Three products were found to contain about 0.1 mg kg1 AFB1 and four products had OTA at levels ranging from 0.4 to 1.8 mg kg1. The remaining samples contained no aflatoxins or OTA (<0.1 mg kg1). ZON has been found in ginseng. Crude extracts of dried ginseng roots from four sources were screened for ZON content using direct competitive enzyme-linked immunosorbent assay (ELISA) (Gray et al. 2004). ZON in one wild-crafted Panax quinquefolius was 680 mg kg1 and in two P. ginseng and one P. quinquefolius were 183, 386, and 177 mg kg1, respectively. However, ZON levels in the same four extracts measured by LC were much lower: 2.60, 11.7, 6.13, and 0.25 mg kg1. The extracts all showed binding activities to oestrogen receptors  and . Ginger Ginger root is widely used for digestive problems. There are several reports of finding aflatoxins in ginger powder (MacDonald and Castle 1996; Patel et al. 1996; Reddy et al. 2002). In India, OTA was found by ELISA in two of 25 samples of ginger at 23 and 80 mg kg1 (Thirumala-Devi et al. 2001). There had not been any earlier report of finding mycotoxins in ginger supplements. The first authors laboratory analysed 25 bottles of ginger capsules (60 capsules/bottle, 625 mg/capsule) purchased from a botanical supplier. All of the samples contained aflatoxins and OTA at levels of 6.212.3 and 1.82.9 mg kg1, with averages of 8.7 and 2.2 mg kg1, respectively. The aflatoxin G1 (AFG1) level was about 10% higher than that of the AFB1, which was unexpected since AFB1 is usually the major aflatoxin found in most grains, nuts, and other agricultural commodities.

184

M. W Trucksess & P. M. Scott . ginseng (Miller 1998). Several reports have been published on OTA in liquorice: three in Germany and one in Spain. In Germany, results of analysis of 83 liquorice roots and liquorice products used as food or supplements for medicinal purposes indicated that high levels of OTA were found in supplement products (0.364.3 mg kg1, median 4.48 mg kg1) but not in childrens tea (0.13 0.44 mg kg1) (Majerus et al. 2000). The peeled and unpeeled roots contained OTA averaging <0.1 and 4.3 mg kg1, respectively. The OTA levels (2.650.3 mg kg1) in medicinal products depended on the content of liquorice. Ten of 11 liquorice tablet samples were contaminated with OTA at 0.72.6 mg kg1 and the other liquorice products were at 0.4646.2 mg kg1, with a median of 1.1 mg kg1. Nineteen samples of liquorice root and 19 liquorice sweets were analysed by LC/MS-MS (Bresch et al. 2000). Nine of the roots and 18 sweets contained OTA at levels of 0.3217 and 0.53.0 mg kg1, respectively. Thirty samples of liquorice root, liquorice confectionery, liquorice block, and liquorice extract were analysed in Spain by LC and confirmed by methyl ester formation (Arino et al. 2007). All samples contained OTA. The 15 dry roots samples had an average OTA of 63.6 mg kg1 (1.4253 mg kg1); the eight fresh roots averaged 9.2 mg kg1 (3.314.7 mg kg1); the four sweets averaged 3.8 mg kg1 (0.58.2 mg kg1); one liquid liquorice extract and one solid block contained 16 and 40 mg OTA kg1, respectively. Red mould rice (RMR) CIT has been found in red mould (yeast, or fermented) rice. RMR containing the cholesterollowering agent monacolin K is a bright reddish purple fermented rice, which acquires its colour from being cultivated with the mould Monascus purpureus. It is a commonly used red food colouring in East Asia and has also been used in Chinese herbal medicines. It is also sold as an overthe-counter dietary supplement for lowering cholesterol. However, in 1999 a study in the Netherlands detected the mycotoxin CIT in all the commercial Monascus samples at concentrations varying between 0.2 and 17.1 mg kg1 (Sabater-Vilar et al. 1999). In the USA, nine different commercially available RMR dietary supplements were tested and CIT was found in seven of them (Heber et al. 2001). In Taiwan, all Monascus products were found to contain CIT at concentrations of 0.286.29 mg kg1 (Liu et al. 2005). A mutant strain, Monascus sp. M12-69, was acquired by treatment with mutagenic agents of a wild strain M12 of Monascus screened from RMR samples gathered in China (Chen and Hu 2005). This strain can produce RMR

Capsicum Capsicum is derived from the Greek word to bite. Thus, Capsicum is known as the plant that bites back. The genus includes chilli peppers, cayenne, pimento, paprika, red peppers, Tabasco peppers, and bell peppers and has been cultivated for thousands of years in tropical America, Africa, and India. There have been many reports of finding aflatoxins in capsicum. In Ethiopia, out of 60 samples each of ground red pepper and Shiro, eight (13%) and five (8.3%) were positive for aflatoxins, respectively (Fufa and Urga 1996). Only AFB1 was detected in both types of foodstuff, ranging from 100 to 500 mg kg1 and from 250 to 525 mg kg1, respectively. In India, samples of grades 13 chilli pod collected in 1998 and 1999 from the principal market yards and cold storage facilities were analysed for AFB1 by an indirect competitive ELISA (Reddy et al. 2001). Of the 182 chilli samples 59% were contaminated with AFB1 and 18% contained the toxin at non-permissible levels (430 mg kg1). The highest AFB1 concentration of 969 mg kg1 was found in one sample representing grade 3. Overall the greatest percentage of chilli pods showing AFB1 levels higher than 30 mg kg1 was in grade 3. The lowest AFB1 levels were found in chilli stored in cold facilities. Another study examined 100 chilli samples for OTA: 26 were found to contain over 10 mg kg1 OTA (Thirumala-Devi et al. 2000). In 12 samples the OTA concentration varied from 10 to 30 mg kg1, in ten samples from 30 to 50 mg kg1, in three samples from 50 to 100 mg kg1, and in one sample it was 120 mg kg1. In Hungary, 91 spice samples (70 ground red pepper, six black pepper, five white pepper, five spice mix and five chilli samples) were analysed for aflatoxins and OTA by IAC/LC (Fazekas et al. 2005). Eighteen of the 70 ground red pepper samples contained AFB1, and concentrations in seven of these ranged from 6.1 to 15.7 mg kg1. Thirty-two of the 70 ground red pepper samples contained OTA, eight of them in a concentration range of 10.666.2 mg kg1. One chilli sample was contaminated with OTA at 2.1 mg kg1. Liquorice Liquorice is derived from the ancient Greek words for sweet root. Glycyrrhizin, a sweetener found in liquorice, is more than 50 times as sweet as sucrose and has pharmaceutical properties. The related Chinese liquorice (G. uralensis), which is used extensively in traditional Chinese medicine, contains this chemical in much greater concentration. Liquorice is a widely consumed herb and is the second most prescribed herb in China, after

Mycotoxins in botanicals and dried fruits with a high concentration of monacolin K and low concentrations of CIT under optimum conditions. Turmeric and kava-kava There was one report in India of finding OTA in nine out of 25 turmeric sample at levels ranging from 11102 mg kg1 (Thirumala-Devi et al. 2001). In the USA, 150 mg kg1 FB1 was found in one of three powdered turmeric samples and OTA in three powdered kava-kava samples at levels of 0.3, 5.6, and 11.2 mg kg1 (Trucksess et al. 2006).

185

Methods of analysis for mycotoxins in botanicals In general, the above limited data indicated that the levels of mycotoxins in botanicals are much lower than in grains and nuts. The results were based on traditional analytical methods with or without modifications. Commonly used methods are LC with FL detection, ELISA and occasionally TLC. Recently, LC-MS and LC-MS/MS techniques have also been applied. Analytical methods for fumonisins Fumonisins in medicinal plants were determined by methanolwater extraction, strong ion exchange solid-phase extraction column clean-up, o-phthaldialdehyde derivatization, reversed-phase LC separation, and FL detection (Martins et al. 2001; Omurtag and Yazicioglu 2004). The LODs for the derivatization methods were 2025 mg kg1. A recent study used a similar extraction solvent but an IAC was used for purification and isolation of FB1 before LC-MS determination without derivatization (Sewram et al. 2006). The LOD was about 10 mg kg1. Analytical methods for aflatoxins The most common methods of analysis for aflatoxins include IAC clean-up, post-column bromination with electrochemical reaction cell (Kobra), or UV irradiation with a photochemical derivatization cell (PHRED), and LC-FL methods (Ali et al. 2005; Arranz et al. 2006; Romagnoli et al. 2007) or solid-phase extraction clean-up and LC-MS/MS (Ventura et al. 2004). One of these methods was tested in a mini-collaborative study by four laboratories for AFB1 in senna pods (Cassia angustifolia), Devils claw (Harpagophytum procumbens) and ginger roots (Zingiber officinale) (Arranz et al. 2006). The limit of quantitation (LOQ) was 1 mg kg1 and recoveries varied depending on the kind of herb. Another IAC/LC method had an LOQ of 0.050.1 mg kg1 (Gomez-Catalan et al. 2005).

When applied to several herbs, recoveries of added aflatoxins were 5060% for some of the herbs even though the reproducibility was good. The IAC procedure was modified by Ip and Che (2006) who replaced the phosphate-buffered saline (PBS) pH 7.4 with 0.1 M phosphate buffer pH 8.0 in order to improve recoveries of aflatoxins from certain Chinese highly acidic medicinal herbs. Three derivatization techniques to enhance the fluorescence of aflatoxin after IAC clean-up were compared: pre-column trifluoroacetic acid, post-column bromination (Kobra cell) and post-column UV irradiation (PHRED or UV cell) (Trucksess et al. 2006). Results of the three derivatization techniques were all comparable for ginseng, ginger, liquorice, and kava-kava. Recoveries of aflatoxin added to ginseng at levels from 2 to 16 mg kg1 were 80%. The LOD was about 1 mg kg1. Reif and Metzger (1995) developed an IAC/LC method for aflatoxins in medicinal herbs which had AFB1 recoveries of 7999% from valerian, fennel seed, and gourd seed. Aflatoxin in medicinal plants such as Rhammus purshiana was determined by LC-MS after methanolwater extraction and polymeric solid-phase clean-up (Ventura et al. 2004). A single-quadrupole MS using an electrospray ionization source operating in the positive-ion mode was used. Mean recoveries of added aflatoxins were about 77110%. The LOD was 10 mg kg1 and the LOQ was 25 mg kg1. TLC methods were used in several older studies (Hitokoto et al. 1978; Trucksess and Stoloff 1980; Tanaka et al. 1988; Efuntoye 1999). There was no mention of LODs in most of these methods except that one method had an LOD of 15 mg kg1 (Tanaka et al. 1988). Non-specific reactions of antibodies to the sample matrix were often encountered when ELISA was used (Thirumala-Devi et al. 2000; Gray et al. 2004). This could result in over- or under-estimated toxin levels. Ochratoxin A LC with FL detection after extraction and purification is widely used for OTA in botanicals. Most of these methods are for liquorice because OTA is well known as a contaminant in liquorice roots and liquorice products. OTA in test samples was extracted by boiling in water for 10 min (Bresch et al. 2000). After the addition of sodium bicarbonate, adjustment of the pH to 7.4, and dilution with buffer, the diluted extract was purified on an IAC. LC-MS/MS turbo ion-spray ionization and multiple reaction monitoring were performed for the separation and quantitation of OTA in the samples. The LOQ of the method was about 0.3 mg kg1. Liquorice products for food,

186

M. W Trucksess & P. M. Scott . Zearalenone (ZON) There have not been many studies on the analysis of botanicals for ZON. When ZON was found in extracts of ginseng roots (Gray et al. 2004), a commercially available ELISA kit and a published method (Ware et al. 1999) were used. There was no mention of LOD or LOQ.

supplements or medicinal purposes were extracted with sodium bicarbonate and methanol and the extract was passed through a phenyl cartridge then an IAC before LC-MS/MS analysis (Majerus et al. 2000). This method enabled the detection of OTA at levels 40.3 mg kg1 in all products analysed. A method based on CEN (European Committee for Standardization) method EN 14132:2003 was applied to liquorice and derived products (Arino et al. 2007). The sample was extract with sodium bicarbonatemethanol, centrifuged, and passed through an IAC. LC-FD was used for separation and quantitation; the LOQ was 0.5 mg kg1 and recoveries were 91%. A multi-toxin IAC column was recently used for aflatoxins and OTA in ginger and ginseng (Trucksess et al. 2007). This is the first report of using multi-toxin IAC for botanicals. After LC-FD separation and quantitation, recoveries of added aflatoxins and OTA were 7080%. The LOD was about 0.1 mg kg1 and the LOQ was 1 mg kg1for AFB1 and OTA. OTA contamination was found in powdered ginger and turmeric powder using an indirect competitive ELISA method (Thirumala-Devi et al. 2001). Samples were extracted with 0.5% potassium chloride in 70% methanol, then diluted before ELISA. The LOD and LOQ were 10 and 35 ng kg1, respectively. Citrinin (CIT) CIT was found in cultures inoculated with Monascus species (Blanc et al. 1995). An excellent review of analysis of CIT was published by Xu et al. (2006). The Monascus product known as RMR has been found to contain the cholesterol-lowering agent monacolin K (MK), including the lactone form (MKL) and the acid form (MKA), and CIT. High recovery rates for CIT, MKL and MKA are achieved by extracting the RMR with 95% ethanol at 60 C for 30 min, and separating the peaks of the three analytes using reversed-phase LC on a C-18 column, acetonitrilewatertrifluoroacetic acid (55 45 0.05, v/v/v) mobile phase, and UV and FL detectors (Lee et al. 2006). After extraction of CIT from Monascus cultures into acetonitrilewater (3 2, v/v), centrifugation and reverse-phase LC separation with FD detection at 334 nm, the recovery of added CIT was 97% (Xu et al. 2003b). A selective solvent extraction using tolueneethyl acetateformic acid was used to extract CIT from Monascus culture material providing red colorants for foods; recoveries were 87126% at 12 mg kg1 spiking levels (Xu et al. 2003a). The LOD was 15 mg kg1 when an ELISA method was applied to determine CIT in Chinese RMR dietary supplements (Heber et al. 2001).

Prevention and treatment of mycotoxins in botanicals Surveys of fungal contamination in botanicals indicate many toxigenic moulds such A. flavus, A. parasiticus, and Penicillium spp. Although mycotoxins are not always detected in botanicals, under warm, humid conditions some moulds could produce aflatoxins and other toxins. Good manufacturing practice after harvesting botanicals such as cleaning, drying, and packaging will minimize mould growth and proliferation. Changing cultural practices such as wetting capsicum (chilli pods) by sprinkling with water before marketing will avoid conditions that favour mould growth and aflatoxin production (Reddy et al. 2001). Microbiological decontamination of botanicals can be achieved by irradiation (Katusin-Razem et al. 1983, 2001; Owczarczyk et al. 2000). However, some important measures and steps should be adopted (Xingwang and Jilan 1998). In order to have no significant biological or toxicological changes in traditional Chinese medicines after irradiation, proper conditions such as irradiating with 7 kGy for herbal medicine and with 5 kGy for some special herbal medicines are required. Herbs should be stored in a dry state and traditional Chinese medicines should be mixed with honey, forming a bolus to minimize decomposition. Levels of OTA can be reduced by peeling the outer skin of liquorice roots (Majerus et al. 2000). It would be worthwhile to investigate whether this procedure is applicable for other roots to determine if OTA and AF levels in ginger, for example, can be lowered by this physical procedure.

Mycotoxins in dried fruit Mycotoxins in dried vine fruit Parallel to their occurrence in wine and grape juice resulting from contaminated grapes (Varga and Kozakiewicz 2006; Scott 2008), mycotoxins are often present in dried vine fruits. Surveys for OTA have been carried out in many countries, including Argentina, Canada, the Czech Republic, Finland, France, Germany, Greece, Hungary, Sweden, the

Mycotoxins in botanicals and dried fruits Netherlands, and the UK (Varga and Kozakiewicz 2006). Some maximum levels of OTA reported in the literature are 35 mg g1 in raisins analysed in Sweden (Moller and Nyberg 2003), 250 mg kg1 in raisins in Egypt (Youssef et al. 2000), 26 mg kg1 in sultanas analysed in Canada (Lombaert et al. 2004), 54 mg kg1 in unprocessed sultanas in Turkey (Meyvaci et al. 2005), 100 mg kg1 in processed Turkish sultanas (Aksoy et al. 2007), and 54 mg kg1 in currants analysed in the UK (MacDonald et al. 1999). Taking some of the data from four of these surveys as examples: Lombaert et al. (2004) found 79% of 85 samples of raisins and 59% of 66 samples of sultanas contained OTA above 0.1 mg kg1 with overall mean levels in all samples of 1.8 mg kg1 for both raisins and sultanas; Meyvaci et al. (2005) covered three production years for unprocessed sultanas in the Aegean region of Turkey, finding considerable variations in incidence and median levels with an overall mean level 3.4 mg kg1; while Aksoy et al. (2007) reported varying incidences over five years in processed Turkish sultanas (with a overall mean concentration 1.4 mg kg1). MacDonald et al. (1999) reported incidences of 95, 85, and 85% OTA40.2 mg kg1 in 20 samples each of currants, raisins, and sultanas, respectively. Aspergillus carbonarius was the predominant fungal source of OTA in dried vine fruits examined in Spain and Argentina (Abarca et al. 2003; Magnoli et al. 2004; Valero et al. 2005). Other OTA-producing fungi which have been isolated from grapes or dried vine fruit and are probable contributors to its presence in dried vine fruits include A. niger var. niger (Abarca et al. 2003; Magnoli et al. 2004; Romero et al. 2005) and A. tubingensis (Medina et al. 2005; Varga et al. 2006). In Greece, the highest frequency of OTA in currants and sultanas occurred in samples from sea level and the lowest occurred in samples from the highest altitudes (6001000 m) (Pateraki et al. 2005). This parallels other findings that OTA concentrations in European Carignan grape musts increased the closer the vineyard was to the Mediterranean Sea (Roset 2003). Sampling and subsampling of dried vine fruits for OTA is an important problem. Sample sizes that have been chosen for analysis are 500 g (Moller and Nyberg 2003; Lombaert et al. 2004), 1000 g (MacDonald et al. 1999) and 3000 g (Meyvaci et al. 2005). The heterogeneity of OTA in dried vine fruits is well illustrated by the study of Moller and Nyberg (2003), who prepared slurries of 150275 g subsamples of raisins or currants in aqueous sodium bicarbonate. There were large differences between subsamples in some cases, e.g. 4.1/13.7, 0.9/4.5, 0.1/3.1, 0.2/34.6, 1.2/5.2 and 6.0/ 19.0 mg kg1.

187

In contrast to the large volume of work on OTA, there are fewer reports on the presence of aflatoxins in dried vine fruits. Maximum concentrations of AFB1 found in raisins were 180 mg kg1 in India (Saxena and Mehrotra 1990), 550 and 300 mg kg1 in two studies in Egypt (Abdel-Sater and Saber 1999; Youssef et al. 2000), and 20 mg kg1 in imports to Greece (Apergi et al. 1998). Up to only 2 mg kg1 aflatoxins B1 and B2 were found in white sultanas analysed in Brazil (Iamanaka et al. 2007). Aflatoxigenic A. flavus has been isolated from grapes in Lebanon (El Khoury et al. 2006). Mycotoxins in dried figs Figs have also been found to contain aflatoxins and OTA as found in numerous surveys (Drusch and Ragab 2003; Drusch and Aumann 2005; Senyuva et al. 2007). Contamination of figs with aflatoxins begins during sun drying on the tree and continues during drying on the ground (Buchanan et al. 1975; Boyacioglu and Gonul 1990; Ozay and Alperden 1991). Levels can be very high: up to 76 000 mg kg1 AB1 (measured by TLC) (Steiner et al. 1993); up to 72 mg kg1 aflatoxin B2 (AB2) (Boyacioglu and Gonul 1990); up to 180 000 mg kg1 aflatoxin G1 (AG1) (measured by TLC) (Steiner et al. 1993); and up to 12 300 mg kg1 OTA (by LC) (Steiner et al. 1993). A four-year survey of aflatoxins in figs intended for export from Turkey showed 2.6, 3.0, 5.1, and 2.7% incidences of total aflatoxins above 4 mg kg1 (Senyuva et al. 2007). Other mycotoxins kojic acid (up to 8600 mg kg1) (Steiner et al. 1993) and patulin (up to 152 mg kg1) (Karaca and Nas 2006) have been also found in dried figs as determined by GC-MS and LC, respectively. More than one mycotoxin can occur in the same sample of figs: AFB1 and OTA (Senyuva et al. 2005); AFB1, AFG1, kojic acid, and OTA (Steiner et al. 1993); and aflatoxins and patulin (Karaca and Nas 2006). In 99 samples of figs exported from Turkey only four contained both AFB1 and OTA (Senyuva et al. 2005). In the study of Karaca and Nas (2006), total aflatoxin concentrations were significantly correlated with the patulin concentration in bright greenish-yellow (BGY) fluorescent figs from the Aegean region in Turkey. The little known species Aspergillus alliaceus may be responsible for the occasional contamination of figs in California with OTA (Doster et al. 1996, Bayman et al. 2002). The sampling situation for aflatoxins in dried figs is similar to that in peanuts in that high levels of aflatoxins are associated with individual fruits. A large sample size is required, e.g. 20 kg from lots of 1820 tonnes (Sharman et al. 1991, Hussain and Vojir 1993); 30 kg from 1530 tonnes (European Commission (EC) 2006a); and 40 kg from 20 tonnes

188

M. W Trucksess & P. M. Scott . European Communities. The maximum level for OTA in dried vine fruit (currants, raisins, sultanas) is 10 mg kg1 (EC 2006b). The maximum level for AFB1 in dried fruit for human consumption or food ingredient is 2 mg kg1 and for total aflatoxins B1, B2, G1, and G2 is 4 mg kg1 (EC 2006b); for dried fruit to be subjected to sorting or other physical treatment the maximum levels are 5 mg kg1 AFB1 and 10 mg kg1 aflatoxins B1, B2, G1, and G2. There are also sampling directives for dried figs and other dried fruits (EC 2006a). The European regulations for patulin (EC 2006b) have maximum limits of 25 mg kg1 patulin in solid apple products (which presumably would include dried apple rings) and 10 mg kg1 patulin in solid apple products for infants and young children. Methods of analysis Two methods of analysis for mycotoxins in dried fruit or processed dry fruit have been subjected to interlaboratory study. One was for the determination of OTA in currants, raisins, sultanas, dried figs, and mixed dried fruit (comprising dried pineapple, papaya, prunes, dates, and banana chips as well as sultanas) using extraction with acidic methanol, IAC clean-up and LC-FL. The method was tested in 24 laboratories at low mg kg1 levels (1.1 11.4 mg kg1) so that the European Union maximum level can be enforced (MacDonald et al. 2003). It ller and should be noted that other workers (Mo Nyberg 2003; Senyuva et al. 2005) have used a basic extraction with methanolaqueous sodium bicarbonate, which gives better recoveries of OTA from figs and dried vine fruits than acid extraction. The second collaboratively studied method, also based on IAC/LC, was for the determination of aflatoxins in fig paste; it was studied in 16 laboratories with AFB1 spiking levels of 1 and 4 mg kg1 and total aflatoxin spiking levels of 2.4 and 9.6 mg kg1 (Stroka et al. 2000). This method used post-column bromination. In a separate study electrochemical bromination and photochemical derivatization were found to be equally suitable for the analysis of fig paste for aflatoxins (Papadopoulou-Bouraoui and Anklam 2002). LC-MS has been used to determine OTA in raisins (Buttinger et al. 2004; Lindenmeier et al. 2004) and aflatoxins in figs (Vahl and Jrgensen 1998). Prevention of mycotoxins in dried fruit As for the botanicals (see the section on Prevention and treatment of mycotoxins in botanicals), exposure of humans to mycotoxins from consumption of dried fruits can be reduced (Drusch and Ragab 2003). The considerable research on

(Bruland et al. 1992). Senyuva et al. (2007) found that three 10-kg subsamples were usually not uniformly contaminated. The sample may be slurried with water before taking subsamples for analysis (Sharman et al. 1991). For fig paste the sample size was 5 kg (Sharman et al. 1991). More aflatoxin was found in BGY fluorescent figs (Steiner et al. 1988; Akerstrand and Moller 1989). Thus, preliminary sorting by removing figs with BGY fluorescence can lower the aflatoxin contamination level of the lot. The AFB1 level decreased from 23 to 0.3 mg kg1 by removal of all BGY fluorescent figs from a 56-kg sample (Steiner et al. 1988). However, absence of BGY fluorescence does not necessarily mean there are no aflatoxins in the sample (Hussain and Vojir 1993). In California figs, BGY fluorescence was not considered a promising screening tool (Doster and Michailides 1998). BGY fluorescence is more likely to be visible after cutting the fig open and this might be of use for figs used to make fig paste where figs are cut into quarters (Doster and Michailides 1998). BGY fluorescence cannot screen for OTA. Mycotoxins in other dried fruit Other dried fruits may also contain mycotoxins, particularly OTA and aflatoxin (Zohri and Abdel-Gawad 1993; Drusch and Ragab 2003; Drusch and Aumann 2005). The natural occurrence of OTA has been reported in dried apricots (Zohri and Abdel-Gawad 1993); dried plums (prunes) (Zohri and Abdel-Gawad 1993; Engel 2000; Iamanaka et al. 2005) and dates (Majerus et al. 1993, Abdel-Sater and Saber 1999; Miraglia and Brera 2002; Iamanaka et al. 2005). Dried slices of quince from markets in India contained up to 1630 mg kg1 of OTA (Sharma and Sumbali 1999). Co-occurrence of OTA and CIT in dry copra was reported from India (Kumari and Nusrath 1987). AFB1 has also been found in dried apricots (Apergi et al. 1998), prunes (Apergi et al. 1998), and dates (Abdel-Sater and Saber 1999; Herry and Lemetayer 1992). Dried pomegranate peel from India, which was analysed in the USA, contained 105 mg kg1 AFB1 (Selim et al. 1996). The occurrence of aflatoxins in dry stored copra is well known and levels of up to 4000 mg kg1 AFB1 have been reported (Samarajeewa and Arseculeratne 1983; Kumari et al. 1984). Saxena and Mehrotra (1990) found CIT as well as aflatoxins in dried coconut. ZON has been detected in dates in Egypt (Abdel-Sater and Saber 1999). Maximum levels for mycotoxins in dried fruit European regulations for certain mycotoxins in dried fruit have been published by the Commission of the

Mycotoxins in botanicals and dried fruits fungicides and vineyard management (soil cultivation, irrigation, and soil amendments) that has been carried out on grapes for the prevention of OTA in wine (Drouillard et al. 2003; Bell et al. 2006; Leong et al. 2006) is applicable with regard to dried vine fruit. The code of sound vitivinicultural practices to minimize levels of OTA in vine-based products put forward by the Office International de la Vigne et du Vin (OIV) includes hygiene of the containers, measures to avoid fruit fly infestation, avoidance of overstacking, sorting, and drying conditions (Castellucci 2005). For the problem of aflatoxins in figs, it is important to have good techniques for orchard management, harvesting and drying, including decreasing A. flavus spores in the orchard, removal of damaged fruits and additional drying in a solar drier (Le Bars 1990; Ozay et al. 1995). As indicated above, fluorescence sorting of dried figs can help prevent aflatoxincontaminated figs from reaching the consumer (Steiner et al. 1988; Akerstrand and Moller 1989). Irradiation of fresh fruits, including grapes and figs, can significantly decrease fungal counts (Aziz and Moussa 2002). Microflora can be eliminated after treatment of fruits with sulfur dioxide followed by drying; sodium and potassium metabisulfite were shown to reduce aflatoxins and mould counts in figs (Ozay et al. 1995).

189

Conclusions On basis of this review, it is clear that botanicals and dried fruits sold commercially can be contaminated with mycotoxins at levels exceeding regulations in some countries. It is essential to investigate further the presence of mycotoxins in these commodities. It is also essential to develop and apply strategies to prevent the formation of mycotoxins in them in order to ensure that botanicals and dried fruits are wholesome and safe for consumers.

References
Abarca ML, Accensi F, Bragulat MR, Castella G, Cabanes FJ. 2003. Aspergillus carbonarius as the main source of ochratoxin A contamination in dried vine fruits from the Spanish market. Journal of Food Protection 66:504506. Abdel-Sater MA, Saber SM. 1999. Mycoflora and mycotoxins of some Egyptian dried fruits. Bulletin of the Faculty of Science of Assiut University D.28: 92107 (Chemical Abstracts 133: 3874p. 2000). Abeywickrama K, Bean G. 1991. Toxigenic Aspergillus flavus and aflatoxins in Sri Lankan medicinal plant material. Mycopathologia 113:187190.

Akerstrand K, Moller T. 1989. The approach in Sweden to the aflatoxin problem in dried figs and measures to be taken in the processing plants to avoid aflatoxins in figs. In: Contributions to the International Dried Figs and Aflatoxin Synposium, Cesme, Turkey, 48 April 1989. Aksoy U, Eltem R, Meyvaci KB Altindisli A, Karabat S. 2007. Five-year survey of ochratoxin A in processed sultanas from Turkey. Food Additives and Contaminants 24: 292296. Ali N, Nashim NH, Saad B, Safan K, Nakajima M, Yoshizawa T. 2005. Evaluation of a method to determine the natural occurrence of aflatoxins in commercial traditional herbal medicines from Malaysia and Indonesia. Food and Chemical Toxicology 43:17631772. Apergi E, Gardikis JP, Panagiotopoulou V-Y. 1998. Occurrence of aflatoxins B1, B2, G1 and G2 in imported goods in Greece during 1995. In: Miraglia M, Van Egmond H, Brera C, Gilbert J, editors. Mycotoxins and phycotoxins developments in chemistry, toxicology and food safety. Fort Collins, CO: Alaken. pp 105110. Arino A, Herrera M, Estopanan G, Juan T. 2007. High levels of ochratoxin A in licorice and derived products. International Journal of Food Microbiology 114:366369. Arranz I, Sizoo E, Van Egmond H, Kroeger K, Legarda TM, Burdaspal P, Reif K, Stroka J. 2006. Determination of aflatoxin B1 in medicinal herbs: interlaboratory study. Journal of AOAC International 89:595605. Aziz NH, Moussa LAA. 2002. Influence of gamma-radiation on mycotoxin producing moulds and mycotoxins in fruits. Food Control 13:281288. Bayman P, Baker JL, Doster MA, Michailides TJ, Mahoney NE. 2002. Ochratoxin production by the Aspergillus ochraceus group and Aspergillus alliaceus. Applied and Environmental Microbiology 68:23262329. Bell N, Marin S, Sanchis V, Ramos AJ. 2006. Impact of fungicides on Aspergillus carbonarius growth and ochratoxin A production on synthetic grape-like medium and on grapes. Food Additives and Contaminants 23:10211029. Blanc PJ, Loret MO, Goma G. 1995. Production of citrinin by various species of Monascus. Biotechnology Letters 17:291294. Boyacioglu D, Gonul M. 1990. Survey of aflatoxin contamination of dried figs grown in Turkey in 1986. Food Additives and Contaminants 7:235237. Bresch H, Urbanek M, Nusser M. 2000. Ochratoxin A in food containing liquorice. Die Nahrung 44:276278. Bruland HG, Matthiaschk G, Sanitz W, Vierkotter S, Weber R, Wenzel H. 1992. Aflatoxine in Trockenfeigen ein pragmatischer Probenahmevorschlag. Deutsche LebensmittelRundschau 88:183185. Buchanan JR, Sommer NF, Fortlage RJ. 1975. Aspergillus flavus infection and aflatoxin production in fig fruits. Journal of Applied Microbiology 30:238241. Buttinger G, Fuchs E, Knapp H, Berthiller F, Schuhmacher R, Binder EM, Krska R. 2004. Performance of new clean-up column for the determination of ochratoxin A in cereals and foodstuffs by HPLC-FLD. Food Additives and Contaminants 21:11071114. Castellucci F. 2005. Resolution Viti-Oeno 1/2005. Code of sound vitivinicultural practices in order to minimise levels of ochratoxin A in vine-based products. Paris: Office International de la Vigne et du Vin (available at: http://news.reseau-concept.net/images/oiv_uk/Client/VITIOENO_1-2005_EN.pdf). Chen F, Hu X. 2005. Study on red fermented rice with high concentration of monacolin K and low concentration of citrinin. International Journal of Food Microbiology 103:331337.

190

M. W Trucksess & P. M. Scott .

Iamanaka BT, de-Menezes HC, Vicente E, Leite RSF, Taniwaki MH. 2007. Aflatoxigenic fungi and aflatoxins occurrence in sultanas and dried figs commercialized in Brazil. Food Control 18:454457. Iamanaka BT, Taniwaki MH, Menezes HC, Vicente E, Fungaro MH. 2005. Incidence of toxigenic fungi and ochratoxin A in dried fruits sold in Brazil. Food Additives and Contaminants 22:12581263. Ip S-P, Che C-T. 2006. Determination of aflatoxins in Chinese medicinal herbs by high-performance liquid chromatography using immunoaffinity column clean-up. Improvement of recovery. Journal of Chromatography A 1135:241244. Karaca H, Nas S. 2006. Aflatoxins, patulin and ergosterol contents of dried figs in Turkey. Food Additives and Contaminants 23:502508. Katusin-Razem B, Novak B, Razem D. 2001. Microbiological decontamination of botanical raw materials and corresponding pharmaceutical products by irradiation. Radiation Physics and Chemistry 62:261275. Katusin-Razem B, Razem D, Dvornik I, Matic S. 1983. Radiation treatment of herb tea for the reduction of microbial contamination (Flores chamomillae). Radiation Physics and Chemistry 22:707713. Kumar S, Roy AK. 1993. Occurrence of aflatoxin in some liver curative herbal medicines. Letters in Applied Microbiology 17:112114. Kumari CK, Nusrath M. 1987. Natural occurrence of citrinin and ochratoxin A in coconut products. National Academy of Science Letters 10:303305. Kumari CK, Nusrath M, Reddy BN. 1984. Aflatoxin contamination of coconut products from Andhra Pradesh. Indian Phytopathology 37:284287. Le Bars J. 1990. Contribution to a practical strategy for preventing aflatoxin contamination of dried figs. Microbiologie Aliments Nutrition 8:265270. Lee C-L, Wang J-J, Pan T-M. 2006. Synchronous analysis method for detection of citrinin and lactone and acid forms of monacolin K in red mold rice. Journal of AOAC International 89:669677. Leong SL, Hocking AD, Pitt JI, Kazi BA, Emmett RW, Scott ES. 2006. Australian research on ochratoxigenic fungi and ochratoxin A. International Journal of Food Microbiology 111(Suppl. 1): S10S17. Lindenmeier M, Schieberle P, Rychlik M. 2004. Quantification of ochratoxin A in foods by a stable isotope dilution assay using high-performance LC-tandem mass spectrometry. Journal of Chromatography A 1023:5766. Liu B-H, Wu TS, Su M-C, Ping Chung C, Yu F-Y. 2005. Evaluation of citrinin occurrence and Monascus fermentation in Monascus fermentation products. Journal of Agricultural and Food Chemistry 53:170175. Lombaert GA, Pellaers P, Neumann G, Kitchen D, Huzel V, Trelka R, Kotello S, Scott PM. 2004. Ochratoxin A in dried vine fruits on the Canadian retail market. Food Additives and Contaminants 21:578585. MacDonald S, Castle L. 1996. A UK retail survey of aflatoxins in herbs and spices and their fate during cooking. Food Additives and Contaminants 13:121128. MacDonald S, Wilson P, Barnes K, Damant A, Massey R, Mortby E, Shepherd MJ. 1999. Ochratoxin A in dried fruit: Method development and survey. Food Additives and Contaminants 16:253260. MacDonald SJ, Anderson S, Brereton P, Wood R. 2003. Determination of ochratoxin A in currants, raisins, sultanas, mixed dried fruit, and dried figs by immunoaffinity column clean-up with liquid chromatography: Interlaboratory study. Journal of AOAC International 86:11641171.

Doster MA, Michailides TJ. 1998. Production of bright greenish yellow fluorescence in figs infected by Aspergillus species in California orchards. Plant Disease 82:669673. Doster MA, Michailides TJ, Morgan DP. 1996. Aspergillus species and mycotoxins in figs from California orchards. Plant Disease 80:484489. DOvidio K, Trucksess M, Weaver C, Horn E, McIntosh M, Bean G. 2006. Aflatoxins in ginseng roots. Food Additives and Contaminants 23:174180. Drouillard J-B, Sage L, Pladeau V, Dubernet M. 2003. ` Ochratoxine A dans les vins. Un partenariat filiere pour des solutions pratiques au vignoble. Phytoma 565: 3035 (Chemical Abstracts 140: 365250p, 2004). Drusch S, Aumann J. 2005. Mycotoxins in fruits: Microbiology, occurrence, and changes during fruit processing. Advances in Food and Nutrition Research 50:3378. Drusch S, Ragab W. 2003. Mycotoxins in fruits, fruit juices, and dried fruits. Journal of Food Protection 66:15141527. Efuntoye MO. 1999. Mycotoxins of fungal strains from stored herbal plants and mycotoxin contents of Nigerian crude herbal drugs. Mycopathologia 147:4348. El Khoury A, Rizk T, Lteif R, Azouri H, Delia M-L, Lebrihi A. 2006. Occurrence of ochratoxin A- and aflatoxin B1-producing fungi in Lebanese grapes and ochratoxin A content in musts and finished wines during 2004. Journal of Agricultural and Food Chemistry 54:89778982. Engel G. 2000. Ochratoxin A in sweets, oil seeds and dairy products. Archiv fur Lebensmittelhygiene 51:98101. European Commission (EC). 2006a. Commission Regulation (EC) No. 401/2006 of 23 February 2006 laying down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs. Official Journal of the European Union L 70: 1234. European Commission (EC). 2006b. Commission Regulation EC No. 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Union L 364: 524. Fazekas B, Tar A, Kovacs M. 2005. Aflatoxin and ochratoxin A content of spices in Hungary. Food Additives and Contaminants 22:856863. Fufa H, Urga K. 1996. Screening of aflatoxins in Shiro and ground red pepper in Addis Ababa. Ethiopian Medical Journal 34:243249. Gomez-Catalan J, Pique E, Falco G, Borrego N, Rodamilans M, Llobet JM. 2005. Determination of aflatoxins in medicinal herbs by HPLC. An efficient method for routine analysis. Phytochemical Analysis 16:196204. Gray SL, Lackey BR, Tate PL, Riley MB, Camper ND. 2004. Mycotoxins in root extracts of American and Asian ginseng bind estrogen receptors  and . Experimental Biology and Medicine 229:560568. Halt M. 1998. Moulds and mycotoxins in herb tea and medicinal plants. European Journal of Epidemiology 14:269274. Heber D, Lembertas A, Lu Q-Y, Bowerman S, Co VLW. 2001. An analysis of nine proprietary Chinese red yeast rice dietary supplements: Implications of variability in chemical profile and contents. Journal of Alternative and Complementary Medicine 7:133139. Herry M-P, Lemetayer N. 1992. Contamination par laflatoxine B1 de fruits oleagineux, de fruits seches et depices. Microbiologie Aliments Nutrition 10:261266. Hitokoto H, Morozumi S, Wauke T, Sakai S, Kurata H. 1978. Fungal contamination and mycotoxin detection of powdered herbal drugs. Applied and Environmental Microbiology 36:252256. Hussain M, Vojir F. 1993. Stichprobenplan fur die Abnahmeprufung beim Import getrockneter Feigen. Deutsche Lebensmittel-Rundschau 89:379383.

Mycotoxins in botanicals and dried fruits

Magnoli C, Astoreca A, Ponsone L, Combina M, Palacio G, Rosa CA, Dalcero AM. 2004. Survey of mycoflora and ochratoxin A in dried vine fruits from Argentina markets. Letters in Applied Microbiology 39:326331. Majerus P, Cutka I, Dreyer A, El-Dessouki S, Eyrich W, Teusch H, Schurer B, Waiblinger HU. 1993. Zur Belastingssituation von Ochratoxin A in Lebensmitteln pflanzlichen Ursprung. Deutsche Lebensmittel-Rundschau 89:112114. Majerus P, Max M, Klaffke H, Plavinskas R. 2000. Ochratoxin A in Su ssholz, Lakritze und daraus hergestellten Erzeugnissen. Deustche Lebensmittel-Rundschau 96:451454. Martins ML, Martins HM, Bernardo F. 2001. Fumonisins B1 and B2 in black tea and medicinal plants. Journal of Food Protection 64:12681270. Medina A, Mateo R, Lopez-Ocana L, Valle-Algarra FM, Mateo F, Jimenez M. 2005. Study of Spanish grape mycobiota and ochratoxin A production by isolates of Aspergillus tubingensis and other members of Aspergillus section nigri. Applied and Environmental Microbiology 71:46964702. Meyvaci KB, Altindisli A, Aksoy U, Eltem R, Turgut H, Arasiler Z, Kartal N. 2005. Ochratoxin A in sultanas from Turkey I: Survey of unprocessed sultanas from vineyards and packing-houses. Food Additives and Contaminants 22:11381143. Miller LG. 1998. Herbal medicinals: selected clinical considerations focusing on known or potential drugherb interactions. Archives of Internal Medicine 158:22002211. Miraglia M, Brera C, editors. 2002. Assessment of dietary intake of ochratoxin A by the population of EU Member States report of experts participating in Task 3.2.7. In: Reports on Tasks for Scientific Cooperation. Directorate-General Health and Consumer Protection of the European Commission (available at: http://ec.europa.eu/food/fs/scoop/index_en.html). Moller TE, Nyberg M. 2003. Ochratoxin A in raisins, & currants: Basic extraction procedure used in two small marketing surveys of the occurrence and control of the heterogeneity of the toxins in samples.. Food Additives and Contaminants 20:10721076. Omurtag GZ, Yazicioglu D. 2004. Determination of fumonisin B1 and B2 in herbal tea and medicinal plants in Turkey by high-performance liquid chromatography. Journal of Food Protection 67:17821786. Owczarczyk HB, Migdal W, Kedzia B. 2000. The pharmacological activity of medical herbs after microbiological decontamination by irradiation. Radiation Physics and Chemistry 57:331335. Ozay G, Alperden I. 1991. Aflatoxin and ochratoxin a contamination of dried figs (Ficus carina L.) from the 1988 crop. Mycotoxin Research 7:8591. Ozay G, Aran N, Pala M. 1995. Influence of harvesting and drying techniques on microflora and mycotoxin contamination of figs. Die Nahrung 39:156165. Papadopoulou-Bouraoui A, Anklam E. 2002. Comparison of two post-column derivatization systems, ultraviolet irradiation and electrochemical determination, for the liquid chromatographic determination of aflatoxins in food. Journal of AOAC International 85:411416. Patel S, Hazel CM, Winterton AGM, Mortby E. 1996. Survey of ethnic foods for mycotoxins. Food Additives and Contaminants 13:833841. Pateraki M, Dekanea A, Lydakis D, Mitchell D, Magan N. 2005. Ecology and control of ochratoxin in grapes and dried vine fruits. In: Congress Proceedings BCPC International Congress: Crop Science & Technology; 31 October2 November 2005, Glasgow, UK. Alton: British Crop Protection Council. p. 413 410 (Chemical Abstracts 145: 270386d, 2006). Reddy DVR, Thirumala-Devi K, Reddy SV, Waligyar F, Mayo MA, Rama-Devi K, Ortiz R, Lenne JM. (2002). Estimate of aflatoxin levels in selected foods and feeds in India.

191

In: Hannak E, Boutrif E. Fabre P, Pineira M, editors. Proceedings of the International Workshop Food Safety Management in Developing Countries CVRAD-FAO. Montpellier: CVRAD-FAO. p. 1113. Reddy SV, Mayi DK, Reddy MU, Thirumala-Devi K, Reddy DVR. 2001. Aflatoxins B1 in different grades of chillies (Capsicum annum L.) in India as determined by indirect competitive-ELISA. Food Additives and Contaminants 18:553558. Reif K, Metzger W. 1995. Determination of aflatoxins in medicinal herbs and plant extracts. Journal of Chromatography A 692:131136. Rizzo I, Varzavsky E, Vedoya G, Haidukowski M, Frade H, Chiale C. 1999. Mycotoxicological control on raw material and tablets of cascara sagrada (Rhamnus purshiana). Mycotoxin Research 15:9195. Rizzo I, Vedoya G, Maurutto S, Haidukowski M, Varsavsky E. 2004. Assessment of toxigenic fungi on Argentinean medicinal herbs. Microbiological Research 159:113120. Romagnoli B, Menna V, Gruppioni N, Bergamini C. 2007. Aflatoxins in spices, aromatic herbs, herb teas and medicinal plants marketed in Italy. Food Control 18:697701. Romero SM, Comerio RM, Larumbe G, Ritieni A, Vaamonde G, Fernandez Pinto V. 2005. Toxigenic fungi isolated from dried vine fruits in Argentina. International Journal of Food Microbiology 104:4349. Roset M. 2003. Survey on ochratoxin A in grape juice. A review of analytical data and research realised by the French grape industry since 1999. Fruit Process 13:167172. Roy AK, Kumar S. 1993. Occurrence of ochratoxin A in herbal drugs of Indian origin a report. Mycotoxin Research 9:9498. Roy AK, Kumari V. 1991. Aflatoxin and citrinin in seeds of some medicinal plants under storage. International Journal of Pharmacognosy 29:6265. Roy AK, Sinha KK, Chourasia HK. 1988. Aflatoxin contamination of some common drug plants. Applied and Environmental Microbiology 54:842843. Sabater-Vilar M, Maas RFM, Fink-Gremmels J. 1999. Mutagenicity of commercial Monascus fermentation products and the role of citrinin contamination. Mutation ResearchGenetic Toxicology and Environmental Mutagenesis 444:716. Samarajeewa U, Arseculeratne SN. 1983. A survey of aflatoxin contamination of coconut products in Sri Lanka; incidence, origins and recommendations. Journal of the National Science Council of Sri Lanka 11:225235. Saxena J, Mehrotra BS. 1990. The occurrence of mycotoxins in some dry fruits retail marketed in Nainital district of India. Acta Alimentaria 19:221224. Schilter B, Andersson C, Anton R, Constable A, Kleiner J, OBrien J, Renwick AG, Korver O, Smit F, Walker R. 2003. Guidance for the safety assessment of botanicals and botanical preparations for the use in food and food supplements. Food and Chemical Toxicology 41:16251649. Scott PM. 2008. Mycotoxins in alcoholic beverages and fruit juices: Occurrence and analysis. In: Siantar DP, Trucksess MW, Herman EM, editors. Food allergens and mycotoxins. Oxford: Oxford University Press, in press. Selim MI, Popendorf W, Ibrahim MS, El Sharkawy S, El Kashory ES. 1996. Aflatoxin B1 in common Egyptian foods. Journal of AOAC International 79:11241129. Senyuva HZ, Gilbert J, Ozcan S, Ulken U. 2005. Survey for co-occurrence of ochratoxin A and aflatoxin B1 in dried figs in Turkey by using a single laboratory-validated alkaline extraction method for ochratoxin A. Journal of Food Protection 68:15121515.

192

M. W Trucksess & P. M. Scott .

Trucksess MW, Weaver CM, Oles CJ, DOvidio K, Rader JI. 2006. Determination of aflatoxins and ochratoxin A in ginseng and other botanical roots by immunoaffinity column clean-up and liquid chromatography with fluorescence detection. Journal of AOAC International 89:624630. Trucksess MW, Weaver CM, Oles CJ, White KD, Betz JM, Rader JI. 2007. The use of multi-toxin immunoaffinity columns for determination of aflatoxins and ochratoxin a in ginseng and ginger. Journal of AOAC International 90 (in press). US Food and Drug Administration (FDA). 1994. Dietary Supplement Health and Education Act of 1994. College Park, MD: US FDA. Vahl M, Jrgensen K. 1998. Determination of aflatoxins in food using LC/MS/MS. Zeitschrift fur Lebensmittel -Untersuchung und Forschung 206:243245. Valero A, Marin S, Ramos AJ, Sanchis V. 2005. Ochratoxin Aproducing species in grapes and sun-dried grapes and their relation to ecophysiological factors. Letters in Applied Microbiology 41:196201. Varga J, Kocsube S, Koncz Z, Teren J. 2006. Mycobiota and ochratoxin A in raisins purchased in Hungary. Acta Alimentaria 35:289294. Varga J, Kozakiewicz Z. 2006. Ochratoxin A in grapes and grapederived products. Trends in Food Science and Technology 17:7281. Ventura M, Gomez A, Anaya I, Diaz J, Broto F, Agut M, Comellas L. 2004. Determination of aflatoxins B1, G1, B2, and G2 in medicinal herbs by liquid chromatography-tandem mass spectrometry. Journal of Chromatography A 1048:2529. Ware GM, Zhao Y, Kuan SS, Carman AS. 1999. Preparative method for isolating alpha-zearalenol and zearalenone using extracting disk. Journal of AOAC International 82:9094. Xingwang F, Jilan W. 1998. Feasibility of sterilizing traditional Chinese medicines by gamma-irradiation. Radiation Physics and Chemistry 52:5358. Xu G, Chen Y, Yu H, Cameleyre X, Blanc PJ. 2003a. HPLC fluorescence method for determination of citrinin in Monascus cultures. Archiv fur Lebensmittelhygiene 54:8284. Xu B-J, Jia X-Q, Gu L-J, Sung C-K. 2006. Review on the qualitative and quantitative analysis of the mycotoxin citrinin. Food Control 17:271285. Xu BJ, Wang QJ, Lee JH, Jia XQ, Sung CK. 2003b. HPLC analysis of citrinin in red yeast rice. Food Science and Biotechnology 12:376380. Yang M-H, Chen J-M, Zhang X-H. 2005. Immunoaffinity column clean-up and liquid chromatography with postcolumn derivatization for analysis of aflatoxins in traditional Chinese medicine. Chromatographia 62:499504. Youssef MS, Abo-Dahab NF, Abou-Seidah AA. 2000. Mycobiota and mycotoxin contamination of dried raisins in Egypt. African Journal of Mycology and Biotechnology 8:6986. (Chemical Abstracts 135: 317617r, 2001). Zohri AA, Abdel-Gawad KM. 1993. Survey of mycoflora and mycotoxins of some dried fruits in Egypt. Journal of Basic Microbiology 33:279288.

Senyuva HZ, Gilbert J, Ulken U. 2007. Aflatoxins in Turkish dried figs intended for export to the European Union. Journal of Food Protection 70:10291032. Sewram V, Shephard GS, Van der Merwe L, Jacobs TV. 2006. Mycotoxin contamination in the Eastern Cape Province of South Africa. Journal of Agricultural and Food Chemistry 54:56885693. Sharma YP, Sumbali G. 1999. Natural occurrence of ochratoxin A and toxigenic Aspergillus ochraceus strains in dry fruit slices of quinces from Jammu and Kashmir. Indian Phytopathology 52:148150. Sharman M, Patey AL, Gilbert J. 1991. Surveillance and control of aflatoxin contamination of dried figs and fig paste imported into the United Kingdom. Food Additives and Contaminants 8:299304. Shaw D. 1998. Risks or remedies? Safety aspects of herbal remedies in the UK. Journal of the Royal Society of Medicine 91:294296. Steiner W, Brunschweiler K, Lembacher E, Schneider R. 1993. Aflatoxine B1 und G1, Cyclopiazonsaure, Kojisaure und Ochratoxin A in Trockenfeigen mit BGY-Fluoreszenz. Mitteilungen aus dem Gebiete der LebensmittelUntersuchung und Hygiene 84:523536. Steiner W, Rieker RH, Battaglia R. 1988. Aflatoxin contamination in dried figs: distribution and association with fluorescence. Journal of Agricultural and Food Chemistry 36:8891. Stroka J, Anklam E, Jorissen U, Gilbert J. 2000. Immunoaffinity column clean-up with liquid chromatography using post-column bromination for determination of aflatoxins in peanuts butter, pistachio paste, fig paste and paprika powder: Collaborative study. Journal of AOAC International 83:320340. Tanaka T, Hasegawa A, Yamamoto S, Aoki N, Toyazaki N, Matsuda Y, Udagawa S, Sekita S, Narita N, Suzuki M, Harada M. 1988. Natural occurrence of aflatoxins in traditional herbal drugs from Indonesia. Proceedings of the Japanese Association of Mycotoxicology 28:3335. Tassaneeyakul W, Razzazi-Fazeli E, Porasuphatana S, Bohm J. 2004. Contamination of aflatoxins in herbal medicinal products in Thailand. Mycopathologia 158:239244. Thirumala-Devi K, Mayo MA, Reddy G, Reddy SV, Delfosse P, Reddy DVR. 2000. Production of polyclonal antibodies against ochratoxin A and its detection in chillies by ELISA. Journal of Agricultural and Food Chemistry 48:50795082. Thirumala-Devi K, Mayo MA, Reddy G, Tangni EK, Larondelle Y, Reddy DV. 2001. Occurrence of ochratoxin A in black pepper, coriander, ginger and turmeric in India. Food Additives and Contaminants 18:830835. Trucksess MW, Stoloff L. 1980. Thin layer chromatographic determination of aflatoxins in dry ginger root and ginger oleoresin. Journal of AOAC International 63:10521054.

Food Additives and Contaminants, February 2008; 25(2): 193202

Managing ochratoxin A risk in the grape-wine food chain

ANGELO VISCONTI, GIANCARLO PERRONE, GIUSEPPE COZZI, & MICHELE SOLFRIZZO

Institute of Sciences of Food Production (ISPA), National Research Council (CNR), Via Amendola 122/O, 70126 Bari, Italy
(Received 5 September 2007; accepted 9 October 2007)

Abstract The main source of ochratoxin A (OTA) in the wine food chain is the infection of grapes by black aspergilli in the field. OTA-producing black aspergilli include principally Aspergillus carbonarius, followed by A. niger and possibly A. tubingensis. They are opportunistic fungi that develop particularly on damaged berries at ripening, although they may occur and form OTA on grapes from veraison to harvest. Climatic conditions (high humidity and temperature) and geographical location are important factors favouring OTA accumulation in grape berries. The severity of aspergillus rot is influenced by excessive irrigation and rainfall prior to harvest, which causes berry splitting. In addition, berry wounds caused by insect attack provide preferential entries for black aspergilli. High OTA levels occur in grapes severely damaged by the grape moth, Lobesia botrana, particularly in Mediterranean areas. Some grape varieties display greater susceptibility to aspergillus rot due to intrinsic genetic characteristics and bunch conformation (i.e. compact4sparse). Control measures for toxigenic mycoflora in the vineyards must consider these critical control points. Proper fungicidal and insecticidal treatments can reduce OTA contamination. Nevertheless, knowledge about the fate of OTA and its distribution in wine and winery byproducts is important to manage OTA risk in contaminated stock. In our wine-making experiments, only 4% of the OTA present in grapes remained in the wine the majority is retained in pressed grape pomaces. OTA concentration remained unchanged in wine after a 1-year aging as well as in all liquid fractions collected during vinification (i.e. must, free run wine, and wine after first and second decantation). Activated carbon can reduce OTA levels in wine but negatively affects wine quality.

Keywords: Ochratoxin A, black aspergilli, mycotoxins, grape, wine, food safety

Introduction Ochratoxin A (OTA) is a major mycotoxin, produced by several species of Aspergillus and Penicillium, naturally occurring in a variety of food commodities prior to harvest or more commonly during storage. Numerous animal studies have shown that OTA is a potent nephrotoxin with the degree of renal injury depending on toxin, dose and exposure time; decreasing nephrotoxic sensitivity was observed from pig to rat to mice. OTA is immunotoxic, neurotoxic in vitro and in vivo in rats, teratogenic in mice, rats and rabbits (JECFA 2001). Based on renal carcinogenicity shown in rats and mice, the International Agency for Research on Cancer (IARC) has classified OTA as a Group 2B
Correspondence: Angelo Visconti. E-mail: angelo.visconti@ispa.cnr.it ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701744546

carcinogen, i.e. carcinogenic to animals and possible carcinogenic to human (IARC 1993). Studies on the genotoxicity of OTA remain controversial. Recent scientific evidence indicates that the sitespecific renal toxicity as well as the DNA damage and genotoxic effects of OTA, measured in various in vivo and in vitro studies, are most likely attributable to cellular oxidative damage (EFSA 2006). Various studies in humans have associated OTA with an endemic kidney disease observed in the Balkans (Balkan Endemic Nephropathy and related Urinary Tract Tumors), but convincing epidemiological evidence associated with OTA exposure is currently lacking. It has been frequently found in human blood, urine and milk and a widespread individual exposure to low levels

194

A. Visconti et al.
Table I. Occurrence of ochratoxin A in wine. No. of analyzed samples 118 144 30 60 46 192 55 420 111 1470 35 601 80 150 Incidence of contamination (%) 70 42 50 56 41 82 87 48 82 59 63 15 85 80 Range(mg kg1) <0.0050.4 <0.017.0 <0.010.3 <0.010.8 <0.010.2 <0.010.6 <0.017.6 <0.013.3 <0.0013.8 <0.0115.6 <0.023.2 <0.021.0 <0.012.9 <0.015.2

Reference Zimmerli and Dick (1996) Majerus and Otteneder (1996) Ospital et al. (1996) MAFF (1997, 1998) Ueno (1998) Burdaspal and Legarda (1999) Visconti et al. (1999) Otteneder and Majerus (2000) Pietri et al. (2001) European Commission (2002) Soufleros et al. (2003) Hocking et al. (2003) Tateo and Bonomi (2003) Finoli et al. (2004)

Origin Europe Europe France Europe Japan Spain Italy Europe Italy Europe Greece Australia Italy Italy

of OTA in Europe and other continents has been demonstrated (EFSA 2006). The widespread human exposure to OTA is well documented by a number of surveys reporting the occurrence of OTA in a variety of food products. An assessment of the dietary intake of OTA by the population of the European Community has been performed showing that the main contributors to OTA exposure are cereals and cereal products (European Commission 2002). Wine, coffee and beer were identified as significant contributors to human OTA exposure. Dried vine fruit and grape juice contribute to a significant extent to the OTA exposure of vulnerable groups of consumers, such as children. The EFSA Scientific Panel on contaminants in the food chain has recently adopted an updated scientific opinion relating to OTA in food, taking into account new scientific information and derived a tolerable weekly intake (TWI) of 120 ng kg1 body weight (EFSA 2006). Based on the available scientific toxicological and exposure data, the European Union established maximum permitted limits for OTA in a variety of food products that have been updated with the EC Regulation 1881/2006 (European Commission 2006). Maximum levels have been set at 2 mg kg1 for wine, fruit wine, grape juice, grape nectar and grape must intended for direct human consumption, and at 10 mg kg1 for dried vine fruit (currants, raisins and sultanas). OTA was detected in wine for the first time in 1996 (Zimmerli and Dick 1996). Thereafter, several surveys have been conducted, mainly in Europe, on the occurrence of the toxin in wine and related products, showing it as a problem mainly for Southern Europe. The results of several reports from different countries are reported in Table I for a total of 3512 samples. A high incidence

of contamination (from 40 to 87%) was reported in all surveys with the exception of the Australian survey having an incidence of 15% (Hocking et al. 2003); the maximum OTA level was recorded in Italy at 15.6 mg kg1 (European Commission 2002). OTA levels showed a decreasing gradient from red to rose to white wines, and the same trend was observed for grape juice. Wines from southern and warmer regions of Europe showed incidence and levels of contamination (72.3%, mean value 0.64 mg kg1, n 635) higher than those from northern European areas (incidence 50.3%, average 0.18 mg kg1, n 835). Wines produced in Southern Italy, where climatic conditions favour the growth of OTA-producing fungi in grapes, generally show incidence and levels of contamination higher than wines produced in Northern and Central Italy. Recently, a survey of retailed wine samples from Southern Italy showed a positive correlation between high levels of OTA and resveratrol-related compounds (Perrone et al. 2007b). Considering that a significant reduction of hepatic and renal damage caused by OTA was reported in mice fed with grape juice contaminated with OTA (Jeswal 1998), there may be some possibility that toxic levels of OTA in red wine are, to some extent, counterbalanced by the beneficial effects of resveratrol derivatives. In the following sections, different aspects of OTA formation in grapes and strategies for prevention and control in the field are reported together with OTA distribution and possible corrective actions during wine processing. Good agriculture and manufacturing practices are currently under discussion by Codex Alimentarius for the adoption of a code of practice for the prevention and reduction of OTA contamination in wine, based on a code of sound vinicultural practices adopted by the Organisation

Managing ochratoxin A risk in the grape-wine food chain Internationale de la Vigne et du Vin (OIV) in 2005 (FAO/WHO 2007).

195

Fungi responsible for OTA accumulation in grapes: Taxonomy and ecology Fungi responsible for OTA accumulation in cereals, i.e. Aspergillus ochraceus and Penicillium verrucosum, were initially thought to be involved in OTA formation in grapes. However, a number of studies performed over the last decade have provided evidence showing that all fungi responsible for OTA in grapes belong to Aspergillus section Nigri, the so called black aspergilli (Battilani et al. 2003a). Most epidemiological surveys, performed in Mediterranean, Australian and South American countries, have shown that, within the black aspergilli, the biseriate species, Aspergillus carbonarius and A. niger aggregate, and the uniseriate species, A. aculeatus and A. japonicus, are prevalent on grapes (Da Rocha Rosa et al. 2002; Battilani et al. 2003b; Leong et al. 2006a). The taxonomy of this Section is still not completely resolved, especially within the A. niger aggregate (a group of morphologically indistinguishable species), often leading to misidentification of the species distribution in food. A comprehensive molecular characterization of the black aspergilli occurring in grape in Europe was performed within the EU project Wine-Ochra Risk (QLK1CT-2001-01761) using representative strains isolated from 107 vineyards of the Mediterranean basin (Bau et al. 2006; Perrone et al. 2006a,b). These studies led to the identification of four main populations, namely A. carbonarius, A. tubingensis, A. niger and a group of Aspergillus uniseriate, which could be separated using molecular methods, including amplified fragment length polymorphism (AFLP), restriction fragment length polymorphism (RFLP) and sequences analysis. The Aspergillus uniseriate group was clearly separated from A. japonicus and A. aculeatus by molecular techniques but was morphologically indistinguishable (Perrone et al. 2006a,b). Ecological and morphological differences between these species are summarized below. A. carbonarius was easily distinguished from other biseriates species due to its large, spiny conidia; a high percentage of strains of this species (98100%) have been shown to produce OTA. Spore germination of A. carbonarius is very rapid, occurring within 24 h at water activity (aw) 0.900.99 and temperature 2535 C. Optimal growth is at 3235 C and aw 0.950.98 (min 10 C and max 45 C). Optimal conditions for OTA

production by A. carbonarius are at 2025 C and aw 0.95/0.98 (Belli et al. 2005). The A. niger aggregate comprise four different species indistinguishable by morphological characteristics. The most frequent species isolated from grapes are A. niger and A. tubingensis, while A. foetidus and A. brasiliensis have been detected to a minor extent (Perrone et al. 2007c). A. niger aggregate optimal growth conditions are 3537 C and aw 0.930.98 (min 68 C and max 47 C). This is one of the most common species group found in a wide range of fresh and dry fruits, cereals, etc., and is seen in food processing as GRAS (generally regarded as safe). OTA production by A. niger aggregate normally occurs at 2025 C and aw 0.95/0.98 (Esteban et al. 2004). A low percentage of OTA-producing strains (510%) were detected among the A. niger aggregate (Perrone et al., 2006a). Among the uniseriate group, A. aculeatus and A. japonicus are often isolated from grapes, but have not been proven to produce OTA although they grow under similar conditions to A. carbonarius. Recently, the Aspergillus uniseriate population from grapes in Europe was characterized molecularly as a population quite different from A. japonicus and A. aculeatus (Perrone et al. 2006b). This population is being described as a new species to be called A. uvarum (Perrone et al. 2007d). Several reports from South America, claiming production of OTA by strains of A. japonicus or that A. niger is the major culprit for OTA accumulation in grapes, are based on morphological identification of the producing strains (Dalcero et al. 2002; Chulze et al. 2006; Ponsone et al. 2007). This evidence has not been confirmed by molecular identification of the species and should be regarded judiciously to avoid confusion in the literature. In our survey of about 600 strains of black aspergilli, representative of a 3-year sampling, 5% of A. niger aggregate strains (360) were OTA producers, while all A. carbonarius strains (200) and none of the A. uniseriate strains (50) were positive for OTA production (Cozzi et al. 2007).

Black aspergilli and OTA occurrence in vineyards: Role of environmental, ecological and agronomical factors Black aspergilli cause a black-rot disease of grape due to high fungal sporulation on berries which renders them completely shrunken and dry. The incidence of colonised berries is more closely related to seasonal conditions during the year of cultivation than the grape growth stage (Battilani et al. 2003b; Leong et al. 2006a; Cozzi et al. 2007).

196

A. Visconti et al.
Black aspergilli - Log (CFU+1)g1

Fungal conidia are usually present on berry skins from setting and increase in number from early veraison to harvest, with a peak at ripening. Black aspergilli overwinter in soil and frequent soil cultivation can favour fungal infection in vineyard. The severity of aspergillus rot is influenced by excessive irrigation prior to ripening, which causes berry splitting. Rain prior to harvest is a common cause of berry damage, favouring Aspergillus infection (Leong et al. 2006a; Cozzi et al. 2007). Berry damage caused by insects, birds or other fungal infection is the primary factor affecting disease development and OTA accumulation in berries. OTA is produced in vineyards and is normally absent up to early veraison. Bunches without visible symptoms can also contain OTA although berries with visible black mould normally show higher contamination levels. The absence of OTA at early growth stages can be explained by major difficulties encountered by fungi in berry penetration (Cozzi et al. 2007). The distribution of black aspergilli in vineyards can be summarized as follows: (i) A. niger aggregate is the principal group at all growth stages; (ii) A. carbonarius incidence is 23 times less than A. niger aggregate and increases from ripening to harvest; (iii) Aspergillus uniseriate is the least represented group, sporadically occurring in Portugal, Greece and Spain, and more frequently in Italy, France and Israel (Battilani et al. 2006). Based on the results of the Wine-Ochra Risk project carried out in six Mediterranean countries, the incidence of berries infected by black aspergilli at harvesting is significantly correlated with latitude and longitude, with a positive WestEast and NorthSouth gradient (Battilani et al. 2006). Following the geostatistical approach described by Battilani et al. (2006), data on the incidence of A. carbonarius were run with ArcView and a predictive map was drawn. The incidence of A. carbonarius was significantly correlated with geographic coordinates showing a positive gradient going towards the South of Europe. Based on the combination of degree-day and rainfall parameters in late Augustearly September in several countries of the Mediterranean basin, discriminant analysis gave promising perspectives for predicting OTA presence in vineyards by the development of thermo-wetness maps (Battilani et al. 2006). Meteorological conditions as well as closeness to the sea have been shown to play a major role in determining OTA occurrence in grapes (Cozzi et al. 2007). A 3-year survey (20042006), performed in eight vineyards located in the Salento peninsula of Southern Italy, showed a wide variability of OTA levels between different cultivation years. In particular, the 2005 crop was the most conducive to black

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

A. niger A. carbonarius A. japonicus OTA

6 5 4 3 2 1 OTA (g kg1)

2004

2005

2006

Figure 1. Occurrence of black aspergilli and OTA in eight vineyards of Primitivo and Negroamaro varieties in Apulia over three grape-harvest seasons (20042006).

aspergilli contamination due to the higher relative humidity and rainfall levels associated with hot temperatures at ripening and harvest time (late AugustSeptember). Figure 1 shows the occurrence of black aspergilli and OTA in eight vineyards of the two major grape varieties (Negroamaro and Primitivo) cultivated in this area during the three harvest seasons (Cozzi et al. 2007). Aspergillus niger aggregate was predominant from early veraison to ripening, representing 8085% of contamination. A. carbonarius increased from veraison to ripening. OTA contamination of processed berries was assessed and results were correlated with the incidence of black aspergilli population, in particular with the increasing colony-forming unit (CFU) values of A. carbonarius. The incidence of A. carbonarius increased from ripening to harvest, when vineyard relative humidity usually increased, with high risk of OTA accumulation always associated with hot temperatures (Cozzi et al. 2007). Despite the widespread occurrence of OTA in various types of wine, there is limited information on the ability of black aspergilli to infect berries and produce OTA in different grape varieties (Battilani et al. 2004). In in vitro experiments, grape variety was shown to affect the incidence of black aspergilli and the level of OTA contamination. Three of the 12 tested varieties, namely Bianco di Alessano, Pampanuto and Uva di Troia, showed low OTA contamination after artificial infection with a mixture of five OTA-producing strains, whereas the most susceptible variety (Cabernet Sauvignon) contained over 200 mg kg1 OTA and $80% incidence of colonized berries (Battilani et al. 2004). The role of the cropping system was monitored in a 2-year survey carried out on four different systems, namely spur-pruned cordon, bower system, head (or small tree) system and espalier, in eight vineyards in Apulia. In both years, the espalier cropping system produced the most contaminated grapes in terms of A. carbonarius infection and OTA accumulation, as

Managing ochratoxin A risk in the grape-wine food chain

2.0
Black aspergilli - Log (CFU+1)g1
A. niger A. carbonarius OTA

197
In Ar Lb

10 8 OTA (g kg1) 6

1000.00 OTA (g kg1) 100.00 10.00 1.00 0.10 0.01 0.0 2.0 4.0 6.0 Log (CFU+1)g1 8.0

1.5

1.0 4 0.5 A 0.0 Spur-pruned cordon Bower system Head system Espalier AB 0 AB 2

Figure 2. Influence of the training systems on the epiphitic black aspergilli in CFU g1 of grape berries samples and the OTA contamination in the Primitivo variety during 2004/2005. OTA levels with same letters are not significantly different according to the Duncan test (p < 0.01).

Figure 3. Logarithmic graph of the distribution of OTA concentration in grape berries versus black aspergilli contamination levels in groups of berries: In: intact berries; Ar: Aspergillus rotten berries; Lb: Aspergillus rotten berries with L. botrana larvae damages.

shown in Figure 2. This can be explained by the closeness of bunches to the soil, which is the most important source of A. carbonarius inocula, compared to spur pruned cordon and bower system. The higher humidity occurring in the espalier cropping system compared to the head system can explain the different contamination level, despite the similar distance of bunches from the soil (Cozzi et al. 2007). All the above ecoagronomical factors play individual roles in developing A. carbonarius on grapes and consequent OTA accumulation, although the final result in relation to OTA risk is better represented by a combination of all these factors. Developing risk maps based on critical control points can help to prevent and control OTA accumulation in grapes. Lobesia botrana: An OTA risk factor in grapevine management Black aspergilli are opportunistic fungi (saprophytes), mainly responsible for secondary rot of grape berries, that develop through entry sites favoured by berry wounds or splitting caused by either biotic or abiotic factors (insects, fungi, birds, rainfall, hail). Lobesia botrana (Lepidoptera: Tortricidae) is the principal grape-berry moth in the vineyards of Southern Europe and can complete three to four generations a year, depending on weather conditions during late summer. Generally, the first generation larvae of L. botrana damage flowers, while the following generations damage berries at different ripening stages. A good correlation between pest damage and OTA content has been found in grape berries, due to the contribution of L. botrana to berry wounds and fungal spore dissemination (Cozzi et al. 2006). Larvae can either contribute to spore dispersal or act as spore vectors, by trapping conidia in the cuticle ornamentation, then facilitating a rapid fungal penetration by

tunnelling into berries, as demonstrated for Botrytis cinerea (Fermaud et al. 1989; Cozzi et al. 2006). Grape berry damage by L. botrana has been shown to considerably increase the contamination level of black aspergilli and consequent OTA accumulation in grapes. In Figure 3, a comparison between groups of intact berries, black rot berries with and without L. botrana damages is reported. All samples of berries damaged by L. botrana showed considerable levels of OTA contamination (up to 1000 mg kg1 OTA) and black aspergilli infection higher than 106 CFU. OTA was not detected in the group of intact berries, while it was found at levels below 1 mg kg1 in $42% of black-rot berries without grape moth damages (Cozzi et al. 2006). Field trials, performed in 2004 and 2005 using both biological and conventional insecticidal treatments, confirmed that a successful control of the third generation of L. botrana reduced the Aspergilli inocula and the formation of OTA in grapes (Kappes et al. 2005; Perrone et al. 2007a). It is, therefore, important to ensure adequate insect control in combination with fungicide treatment to realize effective pest management. Chemical and biological control of OTA-producing fungi The following chemicals have been shown to be active in reducing, to varying degrees, both fungal growth and OTA levels in grape bunches: mepanipyrim, pyrimethanil, fluazinam, iprodione and the mixture cyprodinil/fludioxonil. The latter mixture was confirmed as effective in a number of field trials carried out in several Mediterranean countries, including France, Spain, Greece and Italy (Tjamos ` et al. 2004; Kappes et al. 2005; Bell et al. 2007). The most effective treatment was observed at 21 days before harvesting and a previous treatment at veraison was suggested in high-risk conditions. This mixture of active ingredients applied against

198

A. Visconti et al. some of these studies were performed either by artificially inoculating grape with toxigenic A. carbonarius (Leong et al. 2006b) or by spiking uncontaminated grapes with OTA (Fernandes et al. 2003). These materials differ from naturally contaminated grapes, which comprise both contaminated and uncontaminated berries. Spiking uncontaminated grapes with OTA produces an apparent reduction in OTA concentration in the resulting wine since most of the spiked OTA is adsorbed by the grape pomaces and biomass (grape skins, pulp, yeasts released in must) (Fernandes et al. 2003). On the other hand, when using naturally contaminated berries for vinification, the grape pomaces and solid biomass have a high OTA concentration and represent the source of OTA in wine (Solfrizzo et al. 2007). Moreover, in most of these studies, OTA was only monitored in the liquid fractions and no measurements were recorded for pressed pomaces and lees; thus, the distribution of OTA between solid and liquid fractions during vinification was not established (Fernandes et al. 2003; Rousseau 2004; Grazioli et al. 2006). Another critical point is sample preparation of the liquid fractions (must and wines before racking), which contain suspended biomass, before OTA analysis. The separation or inclusion of the biomass in the sample to be analysed has a significant effect on the measured OTA concentration due to the high amount of OTA reversibly bound to the biomass. Indeed, Leong et al. (2006b) included the biomass when unsedimented liquid fractions (must and wines before racking) were analysed for OTA. Consequently, the OTA concentrations found in these fractions were much higher than those found in the same liquid fraction analysed after spontaneous sedimentation of biomass (first racking). The fate of OTA and its distribution in wine and winery by-products during vinification of naturally contaminated Negroamaro and Primitivo grapes has been recently reinvestigated at laboratory (microvinification) and industrial level by Solfrizzo et al. (2007). Samples of must (before and after maceration), grape pomaces, wine and lees (after the first and second racking) were analysed for OTA to evaluate the levels at each step of vinification. Before analysis, the liquid fractions were centrifuged to separate the biomass and measure soluble OTA. Results of microvinification experiments showed that only 4% of the OTA present in grapes remains in the wine, whereas 95% of the originally OTA is retained on pressed grape pomaces (98% in the skin and 2% in the seeds) and 1% is retained on the lees. Leong et al (2006b) found that 9% of OTA originally present in grapes passed into wine.

black aspergilli at the same combination and schedule, both in dosage and timing, is effective against grey mould, caused by Botrytis cinerea. Moreover, the insecticide treatment against L. botrana in combination with the fungicide contributed significantly to a reduction of OTA levels in the field, particularly in crop-years at high contamination risk (Kappes et al. 2005). Promising results were also obtained using yeast as a biological control agent, isolated from grapes in Greece and in Italy. In particular, good results were obtained with two strains, Cryptococcus laurentii and Aureobasidium pullulans, in Greece (Dimakopoulou et al. 2005) and with a strain of Hanseniaspora uvarum in Italy using weekly or two-weekly treatments. Distribution of OTA in wine and winery by-products and its fate during vinification of red grapes OTA in grapes is transferred to wine and relevant by-products during vinification. Therefore, the availability of reliable analytical methods for OTA determination in must, wine and relevant by-products is important for the risk management of OTA contamination in the wine food chain. To take prompt corrective action, the availability of rapid methods is necessary in wineries for screening the whole production. Several rapid methods, available for OTA analysis in food products, need to be adapted to wine and by-products (Visconti and De Girolamo 2005). The AOAC official method 2001.01 for OTA determination by HPLC in wine (Visconti et al. 2001) can also be used for must, if the solid fraction is previously separated by centrifugation (Solfrizzo et al. 2006). The fate of OTA during vinification has been studied, with contrasting results. Fernandes et al. (2003) observed an increase of OTA concentration in must during maceration of crushed grapes and a consistent reduction in OTA during pomaces and lees separations. Grazioli et al. (2006) found little or no reduction in OTA concentration in wine after the first racking (separation of lees), while a significant reduction in OTA was observed after spontaneous malo-lactic fermentation occurring between the first and the second racking. In contrast, Rousseau (2004) reported that OTA content in must increased after grape crushing and reached maximum levels during malo-lactic fermentation. Leong et al. (2006b) reported that 24% of OTA originally present in crushed red grape passed into free run wine (must) and a 72% OTA reduction was recorded in wine after the first racking. The differences in approach by these authors could explain the conflicting results. Due to unavailability of naturally contaminated grape

Managing ochratoxin A risk in the grape-wine food chain Therefore, the use of these wine by-products as food ingredients should be avoided or checked for OTA contamination. OTA concentration in must remained nearly constant after maceration, pressing, juice clarification, alcoholic fermentation and lees separations (after first and second racking). The same OTA concentrations were found in wine samples analysed after 1 year. The results obtained with the microvinification were also confirmed at an industrial level. An increase of OTA concentration in must was observed during maceration of Primitivo crushed grapes highly contaminated with OTA. This increase could be explained by the high concentration of OTA in the grapes, which required a longer time for the toxin to equilibrate between must and grape pomaces. Removal of OTA Several fining agents have been tested for their ability to remove OTA from contaminated must/wines (Castellari et al. 2001; Leong et al. 2006c). Oenological decolourising carbon has been reported to remove the highest amount of OTA, although carbon also removes anthocyanins and other coloured polyphenols from wine. The effectiveness of the treatment with oak wood fragments depended upon the quantity of wood chips and powder used (Savino et al. 2007). Removal of OTA from grape juice, must and wine using oenological yeast strains has been reported (Bejaoui et al. 2004; Garcia Moruno et al. 2005; Cecchini et al. 2006). The removal of OTA during fermentation is based on adsorption mechanism other than degradation; however, the efficacy of yeasts for OTA reduction at industrial level, as well as their impact on wine quality parameters (phenol compounds), is unclear. Our laboratory has confirmed OTA reduction by yeasts or inactivated yeast walls, with a consistent

199

reduction in colour index (expressed in terms of the Folin Ciocalteu index). The results obtained in our laboratory on the efficacy of selected adsorbent materials to remove OTA from contaminated red wine are reported in Table II. The best results, in terms of OTA removal, were obtained with carbon or commercial preparations containing carbon (i.e. Mikofree, Myco AD A-Z, Standard Q/FIS). On the other hand, the efficacy in OTA removal was proportional to the reduction in polyphenol content of treated wines.

Conclusions The main source of OTA in the wine production chain is infection by black aspergilli in the field. A. carbonarius is the principal species responsible for OTA accumulation in grape berries from early veraison to ripening. OTA production is influenced by climatic conditions/geographic areas, grape varieties/crop systems and berries damage caused by insects, fungal infection or excessive irrigation/rainfall. Fungicidal and insecticidal treatments can reduce OTA contamination and susceptibility to infection can vary between year and region. Developing of risk maps, based on critical control points, can help to prevent and control OTA accumulation in grapes. Availability of rapid methods for OTA analysis is also important for preventive and corrective intervention at critical control points. After maceration of (red) grapes, OTA remains stable during vinification and after a 1-year aging. During vinification of (red) grapes only 4% OTA remains dissolved in the wine, while 96% is retained by solid winery by-products (grape pomace and lees). Carbon reduces OTA concentrations in wines, but negatively affects quality. Good Agriculture Practices (balanced soil tillage, irrigation, nitrogen

Table II. Percentage removal of ochratoxin A (OTA) from red wine containing 10 mg l1 OTA and treated with different amounts of adsorbent. Adsorbent dosage (g l1) Adsorbent Oenological decolourising carbon (Esseco) Activated carbon (Sigma) Mikofree (Perdomini) Standard Q/FIS (Feed Industry Service) Myco AD AZ (Special Nutrients Inc.) Cholestyramine (SIGMA) Bentonite (Vason) Polygel (AEB Group) Oak wood chips (different types and sizes)c
a b

0.1 80 (9)a 36 9 11 6 1 2

0.2 92 (11) 73 28 25 10 12 2

0.4 b 85 31 28 16

0.5 88 63 (4)

1.0 90 83 (6) 68 53 35

8.0 2642

Percent reduction in polyphenol content (Folin Ciocalteu index) reported in parenthesis. Not tested. c Used on red wine containing 8.5 OTA mg l1.

200

A. Visconti et al.

Table III. Main critical control points plus suggested preventive and corrective measures to reduce ochratoxin A contamination in grapes and wine. Risk factors Field and pre-harvest . Mediterranean basin, closeness to the sea . High temperature and relative humidity from veraison to harvest . . . . . Rainfall during ripening period (berry splitting) Berry damage (high risk with grape berry moth infestation) Grape training system susceptibility (high risk with espalier) High nitrogen fertilization, frequent tillage Grape variety susceptibility . Anticipate harvest time in high OTA risk areas when favouring conditions occur Segregate rot bunches at harvesting Minimize storage time before processing Control OTA contamination in must Use carbon preparations to reduce OTA contamination during fermentation Preventive and corrective measures

. . . . . .

Monitoring climatic conditions from veraison to harvest Monitoring with trap system the Lobesia botrana pressure in the vineyards Monitoring black aspergilli rot berries from veraison to ripening Avoid excess of vigour and vegetation favouring aeration of bunches Avoid tillage from veraison to harvest Combined fungicide/insecticide treatments (1 or 2) when favourable climatic conditions occur

Harvestwine making . Mechanical harvest without selection of bunches

. .

High incidence of rot bunches Long grape storage after harversting (48 h)

. . . .

fertilization, pruning) and Good Manufacturing Practices (reduced harvest to vinification time, segregation of rot bunches) help considerably to reduce OTA contamination risk. The main critical control points, as well as preventive and corrective actions, are summarized in Table III.

Acknowledgements Work partially supported by the Italian Ministry of Education, University and Research, MIUR Project n. 12818 SIVINA (D.M. 593/200).

References
Battilani P, Giorni P, Pietri A. 2003a. Epidemiology of toxin producing fungi and ochratoxin A occurrence in grape. European Journal of Plant Pathology 109:715722. Battilani P, Pietri A, Bertuzzi T, Languasco L, Giorni P, Kozakiewicz Z. 2003b. Occurrence of ochratoxin A-producing fungi in grapes grown in Italy. Journal of Food Protection 66:633636. Battilani P, Logrieco A, Giorni P, Cozzi G, Bertuzzi T, Pietri A. 2004. Ochratoxin A production by Aspergillus carbonarius on some grape varieties grown in Italy. Journal of the Science of Food and Agriculture 84:17361740. Battilani P, Barbano C, Marin S, Sanchis V, Kozakiewicz Z, Magan N. 2006. Mapping of Aspergillus Section Nigri in

Southern Europe and Israel based on geostatistical analysis. International Journal of Food Microbiology 111S1:S72S82. Bau M, Castella G, Bragulat MR, Cabanes FJ. 2006. RFLP characterization of Aspergillus niger aggregate species from grapes from Europe and Israel. International Journal of Food Microbiology 111S1:S18S21. Bejaoui H, Mathieu F, Taillandier P, Lebrihi A. 2004. Ochratoxin A removal in synthetic and natural grape juices by selected oenological Saccharomyces strains. Journal of Applied Microbiology 97:10831044. Bell N, Ramos AJ, Coronas I, Sanchis V, Marn S. 2005. Aspergillus carbonarius growth and ochratoxin A production on a synthetic grape medium in relation to environmental factors. Journal of Applied Microbiology 98:839844. Bell N, Marn S, Argiles E, Ramos AJ, Sanchis V. 2007. Effect of chemical treatments on ochratoxigenic fungi and common mycobiota of grapes (Vitis vinifera). Journal of Food Protection 70:157163. Burdaspal PA, Legarda TM. 1999. Ocratoxina a en vinos, mostos y zumos de uva elaborados en Espana y en otros paises eropeos. Alimentaria 299:107113. Castellari M, Versari A, Fabiani A, Parpinello GP, Galassi S. 2001. Removal of ochratoxin A in red wines by means of adsorption treatments with commercial fining agents. Journal of Agricultural and Food Chemistry 49:39173921. Cecchini F, Morassut M, Garcia Moruno E, Di Stefano R. 2006. Influence of yeast strain on ochratoxin A content during fermentation of white and red must. Food Microbiology 23:411417. Chulze SN, Magnoli CE, Dalcero AM. 2006. Occurrence of ochratoxin a in wine and ochratoxigenic mycoflora in grape and dried vine fruits in South America. International Journal of Food Microbiology 111S1:S5S9.

Managing ochratoxin A risk in the grape-wine food chain

Cozzi G, Pascale M, Perrone G, Visconti A, Logrieco A. 2006. Effect of Lobesia botrana damages on black aspergilli rot and ochratoxin A content in grapes. International Journal of Food Microbiology 111S1:S88S92. Cozzi G, Perrone G, Epifani F, Pascale M, Visconti A. 2007. Epidemiology of ochratoxin A producing fungi in Apulian vineyards. Poster 1422r presented at XII International IUPAC Symposium on Mycotoxins and Phycotoxins, 2007 May 2125; Istanbul. Da Rocha Rosa CA, Palacios V, Combina M, Fraga ME, De Oliveira Rekson A, Magnoli CE, Dalcero AM. 2002. Potential ochratoxin A producers from wine grapes in Argentina and Brazil. Food Additives and Contaminants 19:408414. Dalcero A, Magnoli C, Hallak C, Chiacchiera SM, Palacio G, Rosa CAR. 2002. Detection of ochratoxin A in animal feeds and capacity to produce this mycotoxin by Aspergillus section Nigri in Argentina. Food Additives and Contaminants 19:10651072. Dimakopoulou M, Tjamos SE, Tjamos EC, Antoniou PP. 2005. Chemical and biological control of sour rot caused by black aspergilli in the grapevine variety agiorgitico of Korinth region. Poster presented at the International Workshop on Ochratoxin A in Grapes and Wine: Prevention and Control; Marsala (TP) Italy, 2005 October 2021. p 77. Esteban A, Abarca ML, Bragulat MR, Cabanes FJ. 2004. Effects of temperature and incubation time on production of ochratoxin A by black aspergilli. Research in Microbiology 155:861866. EFSA 2006. European Food Safety Authority. Opinion of the Scientific Panel on contaminants in the Food Chain of the EFSA on a request from the Commission related to ochratoxin A in food. EFSA J. 365: 1-56. Available: http:// www.efsa.europa.eu/etc/medialib/efsa/science/contam/contam_ opinions/1521.Par.0001.File.dat/contam_op_ej365_ochratoxin_ a_food_en1.pdf. Accessed 5 September 2007. European Commission 2002. SCOOP EC Directorate-General Health and Consumer Protection. Assessment of dietary intake of ochratoxin A by the population of EU Member States. Reports on tasks for scientific cooperation 153 pages. Available from: http://ec.europa.eu/food/fs/scoop/3.2.7_en.pdf. Accessed 5 September 2007. European Commission. 2006. Commission Regulation (EC) No 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Union 364:524. FAO/WHO Codex Alimentarius Commission 2007. Proposed draft code of practice for the prevention and reduction of ochratoxin A contamination in wine. Codex Committee on Contaminants in Foods. ALINORM 07/30/41 Session 1: 5760. Available from: http://www.codexalimentarius.net/ download/report/691/al30_41e[1].pdf. Accessed 5 September 2007. Fermaud M, Le Menn R. 1989. Association of Botrytis cinerea with Grape Berry Moth Larvae. Phytopatology 79:651656. Fernandes A, Venancio A, Moura F, Garrido J, Cerdeira A. 2003. Fate of ochratoxin A during a vinification trial. Aspects of Applied Biology 68:7380. Finoli C, Vecchio A, Scarpellini M, Burruano S. 2004. Ochratoxin A occurrence in Italian wines of different origins. Rivista di Viticoltura e di Enologia 57:6377. Garcia Moruno E, Sanlorenzo C, Boccacino B, Di Stefano R. 2005. Treatment with yeast to reduce the concentration of ochratoxin A in red wine. American Journal of Enology and Viticulture 56:7376.

201

Grazioli B, Fumi MD, Silva A. 2006. The role of processing on ochratoxin A content in Italian must and wine: A study on naturally contaminated grapes. International Journal of Food Microbiology 111S1:S93S96. Hocking AD, Varelis P, Pitt JI, Cameron S, Leong S. 2003. Occurrence of ochratoxin A in Australian wine. Australian Journal of Grape and Wine Research 9:7278. IARC. 1993. Monographs on the evaluation of carcinogenic risks to humans. Some naturally occurring substances: food items and constituents, heterocyclic aromatic amines and mycotoxins. Lyon. France: International agency for research on cancer 56:489521. JECFA. 2001. Joint FAO/WHO Expert Committee on Food Additives; 56th Meeting, Geneva, 615 February 2001. Jeswal P. 1998. Antidotal effect of grape juice (Vitis vinifera) on ochratoxin A caused hepatorenal carcinogenesis in mice (Mus musculus). Cytobios 93:123128. Kappes ME, Serrati L, Drouillard JB, Cantus JM, Kazantzidou M. 2005. A crop protection approach to Aspergillus and OTA management in Southern European vineyards. Paper presented at the International Workshop on Ochratoxin A in Grapes and Wine: Prevention and Control; Marsala (TP), Italy, 2005 October 2021. p 24. Leong SL, Hocking AD, Pitt JI, Kazi BA, Emmett RW, Scott ES. 2006a. Australian research on ochratoxigenic fungi and ochratoxin A. International Journal of Food Microbiology 111S1:S10S17. Leong SL, Hocking AD, Varelis P, Giannikopoulos G, Scott ES. 2006b. Fate of ochratoxin A during vinification of Semillon and Shiraz grapes. Journal of Agricultural and Food Chemistry 54:64606464. Leong SL, Hocking AD, Scott ES. 2006c. The effect of juice clarification, static or rotary fermentation and fining on ochratoxin A in wine. Australian Journal of Grape and Wine Research 12:245251. MAFF. 1997. Ministry for agriculture, fisheries and food. Survey of aflatoxins and ochratoxin A in cereals and retail products. London: HMSO. MAFF. 1998. Ministry for agriculture, fisheries and food. Survey of retail products for ochratoxin A. London: HMSO. Majerus P, Otteneder H. 1996. Nachweis und Vorkommen von Ochratoxin A in Wein und Traubsaft. Dtsch LebensmRundsch 92:388390. Ospital M, Cazabeil JM, Betberder AM, Tricard C, Creppy E, Medina B. 1998. LOchratoxine A dans les vins. Revue Francaise dOenologie 169:1618. Otteneder H, Majerus P. 2000. Occurrence of ochratoxin A (OTA) in wines: influence of the type of wine and its geographical origin. Food Additives and Contaminants 17:793798. ` Perrone G, Mule G, Susca A, Battilani P, Pietri A, Logrieco A. 2006a. Ochratoxin A production and AFLP analysis of Aspergillus carbonarius, Aspergillus tubingensis, and Aspergillus niger strains isolated from grapes in Italy. Applied and Environmental Microbiology 72:680685. ` Perrone G, Susca A, Epifani F, Mule G. 2006b. AFLP characterization of Southern Europe population of Aspergillus Section Nigri from grapes. International Journal of Food Microbiology 111S1:S22S27. Perrone G, Cozzi G, Pascale M, Logrieco A, Visconti A. 2007a. Lobesia botrana an ochratoxin A risk factor in grapevine management. Poster 1423 presented at the XII International IUPAC Symposium on Mycotoxins and Phycotoxins; 2007 May 2125; Istanbul. Perrone G, Nicoletti I, Pascale M, De Rossi A, De Girolamo A, Visconti A. 2007b. Positive correlation between high levels of ochratoxin A and resveratrol related compounds in red wines. Journal of Agricultural and Food Chemistry 55:68076812.

202

A. Visconti et al.
Society National Meeting & Exposition. September 1014, 2006, San Francisco, CA; Section D, no. 140. Solfrizzo M, Panzarini G, Visconti A. 2007. Fate of ochratoxin A during vinification of naturally contaminated primitivo and negroamaro grapes. Paper presented at the XIIth International IUPAC Symposium on Mycotoxins and Phycotoxins. May 2125; Istanbul, Turkey. p 1425. Tateo F, Bononi M. Survey on ochratoxin A in wine. More data concerning bottled red wines. Le Bulletin de lOIV 74: 766778. Tjamos SE, Antoniou PP, Kazantzidou A, Antonopoulos DF, Papageorgiou I, Tjamos EC. 2004. Aspergillus niger and Aspergillus carbonarius in Corinth raisin and wineproducing vineyards in Greece: population composition, ochratoxin A production and chemical control. Journal of Phytopatology 152:250255. Ueno Y. 1998. Residue and risk of ochratoxin A in human plasma and beverages in Japan. Mycotoxins 47:2532. Visconti A, De Girolamo A. 2005. Fitness for purpose Ochratoxin A analytical developments. Food Additives and Contaminants Suppl 1:3744. Visconti A, Pascale M, Centonze G. 1999. Determination of ochratoxin A in wine by means of immunoaffinity column clean-up and high-performance liquid chromatography. Journal of Chromatography A 864:89101. Visconti A, Pascale M, Centonze G. 2001. Determination of ochratoxin A in wine and beer by immunoaffinity column cleanup and liquid chromatographic analysis with fluorometric detection: collaborative study. Journal of the Association of Official Analytical Chemists International 84:18181827. Zimmerli B, Dick R. 1996. Ochratoxin A in table wine and grape juice: occurrence and risk assessment. Food Additives and Contaminants 13:655668.

` Perrone G, Susca A, Epifani F, Mule G, Logrieco A. 2007c. Biodiversity of Aspergillus Section Nigri from grapes in Europe. Book of Abstracts of the Aspergillus International Workshop: Aspergillus systematics in the genomic era; 2007 April 1214; CBS, Utrecht, The Netherlands. pp 3941. Perrone G, Varga J, Susca A, Frisvad JC, Stea G, Kocsube S, Toth B, Kozakiewicz Z, Samson RA. 2007d. Aspergillus uvarum sp. nov., an uniseriate black Aspergillus species isolated from grapes in Europe. International Journal of Systematic and Evolutionary Microbiology. In press. Pietri A, Bertuzzi T, Pallaroni L, Piva G. 2001. Occurrence of ochratoxin A in Italian wines. Food Additives and Contaminants 18:647654. Ponsone ML, Combina M, Dalcero A, Chulze S. 2007. Ochratoxin a and ochratoxigenic Aspergillus species in Argentinian wine trapes cultivated under organic and nonorganic systems. International Journal of Food Microbiology 114:131135. Rousseau J. 2004. Ochratoxin A in wines: current knowledge. Vinidea.net. Wine Internet Technical Journal (5). Available: http://www.infowine.com/default.asp?scheda1038. Accessed 5 September 2007. Savino M, Limosani P, Garcia-Moruno E. 2007. Reduction of ochratoxin A contamination in red wines by oak wood fragments. American Journal of Enology and Viticulture 58(1):97101. Soufleros EH, Tricard C, Bouloumpasi EC. 2003. Occurrence of ochratoxin A in Greek wines. Journal of the Science of Food and Agriculture 83:173179. Solfrizzo M., Panzarini G, Pascale M, Visconti A. 2006. Determination of ochratoxin A in grape and winery by-products by immunoaffinity column cleanup and HPLC/fluorescence detection. Paper presented at the 232nd American Chemical

Food Additives and Contaminants, February 2008; 25(2): 203208

Management and prevention of mycotoxins in peanuts

J. W. DORNER
USDA, ARS, National Peanut Research Laboratory, Dawson, GA, USA
(Received 31 May 2007; accepted 19 August 2007)

Abstract Contamination of peanuts with mycotoxins, particularly aflatoxins, is a worldwide problem that affects both food safety and agricultural economies. Most countries have adopted regulations that limit the quantity of aflatoxins in food and feed to 20 mg kg1 or less; however, environmental conditions in most of the world where peanuts are produced and stored often make it difficult or impossible to attain such low concentrations. In addition to aflatoxins, peanuts are often contaminated with cyclopiazonic acid (CPA). Both mycotoxins are produced by Aspergillus flavus, a ubiquitous fungus that can infect and grow in peanuts under both pre- and post-harvest conditions. Management of mycotoxin contamination in peanuts generally involves removal of high-risk components from shelled lots or the removal of individual, highly contaminated nuts. This is accomplished by various processes such as screening, kernel sizing, electronic colour sorting, hand sorting, and blanching followed by electronic colour sorting. Recently, biological control technology has been developed that prevents much of the contamination that might otherwise occur. Biocontrol is based on competitive exclusion whereby a dominant population of a non-toxigenic strain of A. flavus is established in the soil before peanuts are subjected to conditions favouring contamination. The applied strain competes with toxigenic strains for infection sites, resulting in significantly reduced concentrations of aflatoxins in peanuts. Monitoring of the first commercial use of the technology showed that aflatoxins were reduced by an average of 85% in farmers stock peanuts and by as much as 98% in shelled, edible grade peanuts.

Keywords: Peanut, aflatoxin, cyclopiazonic acid, Aspergillus flavus, Aspergillus parasiticus, segregation, electronic colour sorting, biological control, prevention, management

Introduction The birth of mycotoxicology came about with the discovery of the aflatoxins in the early 1960s (Sargeant et al. 1961), and because they were found as contaminants of Brazilian peanut meal, there has been long-standing concern about the safety of peanuts as food and feed. Aflatoxins are produced in peanuts as a result of invasion and growth by Aspergillus flavus and A. parasiticus, and contamination can occur during various stages of production, harvest, handling, and storage (Diener et al. 1987). Pre-harvest aflatoxin contamination of peanuts is associated with late-season drought conditions as peanuts begin to dehydrate in the soil under hot, dry environmental conditions (Cole et al. 1989). Contamination can also occur after peanuts are dug if they are not quickly harvested, dried, and
Correspondence: J. W. Dorner. E-mail: Joe.Dorner@ars.usda.gov ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701658357

maintained at a safe moisture level. In addition to aflatoxins, peanuts can be contaminated with cyclopiazonic acid (CPA), another mycotoxin produced by A. flavus as well as other species of Aspergillus and Penicillium (Lansden & Davidson 1983). Although not as acutely toxic as aflatoxins (the oral LD50 in rodents is approximately 3070 mg kg1), CPA is a potent inhibitor of the reticular form of the Ca2ATPase pump. A thorough review of the toxicology and a safety assessment of CPA has been published recently (Burdock & Flamm 2000). CPA has been found as a natural contaminant of a variety of commodities and foods, and it was implicated in a human poisoning associated with kodo millet (Rao & Husain 1985). The vast majority of CPA contamination of peanuts likely results from A. flavus and quite often co-occurs with aflatoxins (Urano et al. 1992). Because peanuts are primarily produced in

204

J. W Dorner . peanuts from the field to the point of sale when there could be delays in drying; and (4) during storage of farmers stock (FS) or shipment of shelled peanuts when a safe storage moisture content cannot be maintained. Most of the techniques used to manage this contamination involve the physical separation of contaminated from uncontaminated seed. Lot segregation In the USA, the first opportunity to manage aflatoxin contamination occurs when FS lots are delivered by the farmer to the point of sale (buying point). As part of the official grading procedures performed on all FS lots, a sample of peanuts taken from the lot is visually inspected for the presence of aflatoxin-producing fungi. If a peanut is found with visible A. flavus or A. parasiticus, the entire lot is diverted from the edible supply and can be used only for oil (Whitaker et al. 1998). Although no official aflatoxin analysis is conducted on incoming FS peanuts, many buyers perform independent analyses on lots in which visible A. flavus is not found to determine better the aflatoxin risk associated with incoming lots. Buyers can then segregate lots for storage based on the concentration of aflatoxin found. Management systems in place in other peanut-producing countries vary, and some include an aflatoxin analysis on incoming FS peanuts with price deducts based on the amount of aflatoxin. After a period of storage, during which time aflatoxin may be produced or increase, peanuts are subjected to some or all of the management techniques described below to produce shelled lots that meet specific regulatory guidelines. Screening After lot segregation the first step in managing aflatoxin contamination usually involves running peanuts over a screening device to separate certain high-aflatoxin-risk components. High levels of aflatoxin have been shown to be associated with loose shelled kernels (LSK), which are peanuts that have been dislodged from their pods during the harvesting and handling processes (Dorner & Cole 1997). In addition, higher levels of aflatoxin are associated with small, immature pods. Therefore, removing these high-risk components before storing or shelling has the effect of reducing the aflatoxin concentrations subsequently found in shelled lots. For many years these separations were accomplished with vibratory, perforated screens that were not widely utilized because of low peanut flow rates and a tendency for perforations to become clogged. However, development of the belt screen has greatly increased the screening of FS peanuts before shelling. The belt screen is a series of parallel

tropical and subtropical-to-temperate regions, A. flavus and A. parasiticus are the predominant mycotoxigenic fungi associated with peanuts, and significant contamination of peanuts with other mycotoxins is rare. Therefore, the focus of this paper will be on measures to manage and prevent aflatoxin contamination with the understanding that these same measures are also effective for the management and prevention of CPA contamination. Although aflatoxins are potent hepatotoxins, concern about their potent carcinogenicity has forced government regulatory agencies to establish very low tolerances for aflatoxins in food, including peanuts and peanut products (Van Egmond 2002). The European Union upper limit for aflatoxins in peanuts is 2 mg kg1 for aflatoxin B1 and 4 mg kg1 for total aflatoxins (B1 B2 G1 G2) (European Commission 1998). The upper limit set by the US Food and Drug Administration is 20 mg kg1 for total aflatoxins (http://www.cfsan.fda.gov/$lrd/fdaact. html), but the US peanut industry maintains a selfimposed limit of 15 mg kg1 that is administered by the US Department of Agriculture (USDA) (Whitaker et al. 2005). Most other countries have adopted similar regulations (Food and Agriculture Organization 2004). Because individual peanut seeds can be contaminated with aflatoxin concentrations as high several hundred thousand to a million mg kg1 coupled with the fact that usually very few seeds are contaminated, sampling error makes it very difficult to ensure that shelled lots meet the low regulatory limits (Whitaker et al. 1974). The scientific research community, in conjunction with peanut industries, have worked very hard to ensure that edible-grade peanuts contain the lowest aflatoxin concentrations possible. However, it is not always possible to achieve the necessary level, and this produces severe economic pressure on commercial peanut companies during years when aflatoxin contamination is severe. The purpose of this paper is to review the techniques that have been developed for managing aflatoxin contamination when it occurs and also to describe a newly developed methodology for preventing much of that contamination.

Management techniques Aflatoxin management techniques are those that have been developed to manage contamination that has already occurred. That contamination could have taken place during any of several phases in the production of edible-grade peanuts, including: (1) in the field under late-season conditions of drought and heat stress; (2) after peanuts were dug, but before they could be harvested, usually a result of rainy conditions after digging; (3) during transport of

Management of mycotoxins in peanuts belts spaced apart at specific distances which rotate continuously around appropriately positioned sheaves. As peanuts flow across the rotating belts, materials smaller than the spaces between the belts (such as LSK and small pods) fall through while larger pods ride across to the end for collection (Smith et al. 1995). The belts can be spaced to allow for very efficient separation of LSK, small pods, and some foreign material resulting in reduced aflatoxin in final shelled product. In a study of the effect of belt screening 17 loads of FS peanuts, Dowell et al. (1990) found an average 35% reduction in aflatoxin as a result of screening. These devices are now widely used in the USA peanut industry. Density segregation After peanuts are shelled, they are fed to a gravity table that separates material based on specific gravity. This is done primarily to remove foreign material as well as to separate unshelled pods from the shelled kernels. Because highly contaminated kernels are less dense than most kernels, a reduction in aflatoxin contamination of the final product can be achieved by separately collecting the least dense kernels from the gravity table (Davidson et al. 1981). From an initial average FS lot concentration of 60 mg kg1, Davidson et al. found average aflatoxin concentrations of 10.2, 44.5, and 69.6 mg kg1 in the heavy, medium, and light fractions, respectively, after density segregation. Use of this methodology for the specific purpose of managing aflatoxin is not a matter of routine practice because of a lack of efficiency. Too many non-contaminated kernels are lost in the light fraction to make it economical. However, during crop years characterized by unusually high levels of aflatoxin separation of the very lightest kernels from the main flow can remove enough aflatoxin to make the process worthwhile. Kernel sizing After peanuts are shelled, kernels are separated into different size categories by passing peanuts over a series of slotted and round-hole screens (Whitaker et al. 2005). Edible grade runner-type peanuts in the USA are classified as jumbo (kernels that ride a screen with 0.833 cm wide by 1.9 cm long slotted holes [21S]), medium (kernels that fall through the 21S screen but ride a screen with 0.714 cm wide by 1.9 cm long slotted holes [18S]), number one (kernels that fall through the 18S screen but ride a screen with 0.675 cm diameter round holes [17R]), and sound splits (kernels that split during shelling and ride a 17R screen). Kernels that fall through the 17R screen are classified as oil stock and are not used for edible purposes. Although shelled peanuts are not sized for the purpose of managing aflatoxin

205

contamination, that is a by-product of the process because higher concentrations of aflatoxin are associated with smaller size kernels. This association is actually based on the maturity of peanut pods, with immature pods being more susceptible to contamination than mature pods (Dorner et al. 1989). Generally, immature pods contain smaller kernels than mature pods. Therefore, much of the aflatoxin in farmers stock peanuts is found in immature kernels that end up in the oil stock category. As the size of kernels increases, generally lower concentrations of aflatoxin are found. The jumbo and medium size categories can account for about 70% of the total weight of kernels in farmers stock peanuts. In a study of the partitioning of aflatoxin into various size categories using a 45 kg sample from 46 farmers stock lots, Whitaker et al. (2005) found that the initial mean aflatoxin concentration of 73.7 mg kg1 was reduced to means of 42.5 and 66.2 mg kg1 in the jumbo and medium size categories, respectively, but was increased to 93.6, 105.1, and 133.6 mg kg1 in the number one, sound split, and oil stock categories, respectively. Electronic colour sorting The most effective technique for managing aflatoxin contamination in commercial shelling plants is electronic colour sorting (ECS). In shelling plants in the USA all peanuts pass through these highspeed sorters to remove discoloured kernels. This is done to improve overall quality, including the reduction of aflatoxin in the final product. Peanuts that have been colonized by aflatoxigenic fungi are often discoloured, and ECS very efficiently removes a high percentage of the contaminated, discoloured kernels. In a study that evaluated the effect of various post-harvest aflatoxin management techniques, ECS produced a 70% reduction in the amount of aflatoxin in the medium kernel size category (Cole et al. 1995). In recent years continued advances in ECS technology have improved sorter efficiency with the result that fewer good kernels are rejected. However, not all aflatoxin-contaminated kernels are discoloured; therefore, ECS is never 100% effective in aflatoxin removal. Blanching and electronic colour sorting The final technique that can be employed to reduce the aflatoxin concentration in shelled peanut lots is blanching followed by ECS. Blanching is a procedure that removes the testa (seed coat) from kernels. ECS after blanching very efficiently removes aflatoxin-contaminated kernels from the blanched lot because slight discolorations in the kernel tissue that were not visible before testa removal become evident after removal. Because aflatoxin contamination is

206

J. W Dorner . Aflatoxin risk forecasting for early harvesting Pre-harvest aflatoxin contamination of peanuts can be prevented if peanuts are harvested before aflatoxin is actually produced. The time under lateseason drought conditions that is necessary for aflatoxin contamination to occur varies and is dependent on numerous factors, the most important being soil temperature. Data from several years of studies conducted at the USA National Peanut Research Laboratory were used to develop aflatoxin prediction models that could be used to forecast when aflatoxin contamination was likely to occur in farmers fields (Thai et al. 1990; Parmar et al. 1997). Farmers could then use that information to include aflatoxin risk in making harvest decisions. However, that technology never has been seriously utilized in the USA because the marketing system for FS peanuts does not allow for economic penalties based on a measure of aflatoxin. Rather, farmers are penalized if incoming loads are found to contain visible A. flavus, but such a finding is relatively rare except under the harshest of conditions. Therefore, harvest decisions are still primarily made to achieve the highest yield possible. In Australia, however, where penalties are imposed based on the quantity of aflatoxin found in FS loads, a web-based aflatoxin risk-prediction system called AFLOMAN (http://www.apsim.info/ apsim/afloman/) has recently been employed (Wright et al. 2005). Farmers input information on daily rainfall and soil and ambient temperatures via the internet. The Agricultural Production Systems Simulator (APSIM) peanut aflatoxin model is then run for the specific field with results uploaded back to the website. Farmers can view graphs showing changes in the fraction of available soil water, soil temperature, and aflatoxin risk. The risk of aflatoxin contamination can then be taken into account so that earlier-than-normal harvesting can be undertaken to minimize aflatoxin contamination. Biological control New biological control technology has been developed that can prevent much of the contamination of peanuts with aflatoxins and CPA that would otherwise occur. That control is based on competitive exclusion and is achieved by applying a competitive, non-toxigenic strain of A. flavus to the soil of developing peanuts. Most of the research that resulted in development of this technique has been recently reviewed (Dorner 2005). It was shown that biological control is effective for both pre- and post-harvest aflatoxin contamination. The technology has been commercialized and the biocontrol product afla-guard has been registered by the US Environmental Protection Agency (2004) as

often associated with these discolorations, blanching followed by ECS is widely recognized as the best method for reducing aflatoxin in shelled peanut lots, and it is usually performed at a facility specifically designed for this process. When shelled lots are found to contain aflatoxin concentrations above prescribed limits, thus precluding their sale, they are often sent to a blanching facility in order to reduce the concentration to an acceptable level. In the study reported by Cole et al. (1995), blanching/ ECS produced a 91% reduction in the mean aflatoxin concentration of a lot of shelled medium peanuts. The major disadvantage to this form of aflatoxin reduction is the cost, which includes US$0.075/lb in direct charges, the weight loss incurred during blanching, and the loss of kernels by ECS (Dorner & Lamb 2006).

Prevention techniques The best way to control mycotoxin contamination of peanuts is to prevent it in the first place. This is not always possible, but technologies exist which, if available and affordable, can prevent much of the contamination that would otherwise occur. Kernel moisture control Pre-harvest aflatoxin contamination of peanuts essentially can be eliminated with proper and adequate irrigation. Developing and maturing peanuts are not susceptible to colonization by A. flavus and A. parasiticus until kernel moisture (water activity) begins to decrease in response to lateseason drought conditions with increased soil temperature (Dorner et al. 1989). Maintaining high kernel water activity until the time of harvest maintains the natural defence mechanism (phytoalexin production) of peanuts against growth by aflatoxigenic fungi, even if fungal invasion occurs. The only exception to this is under severe insect pressure whereby extensive pod damage may give the fungi the opportunity to overwhelm the ability of kernels to ward off the fungal attack. Unfortunately, many peanut farmers do not have access to supplemental irrigation or the cost is not affordable. After peanuts are dug and harvested, contamination can be prevented by rapidly drying peanuts to or below a water activity (0.83) that cannot support aflatoxin production (Diener & Davis 1970). It is then necessary to maintain that safe storage moisture until peanuts are processed. This can also be difficult or impossible to accomplish because of environmental conditions during harvest as well as during the storage period. Nevertheless, control of kernel moisture is the best way to prevent mycotoxin contamination of peanuts if the means are available.

Management of mycotoxins in peanuts a biopesticide for control of aflatoxin contamination in peanuts. The biopesticide is hulled barley that is coated with conidia of a non-toxigenic strain of A. flavus (NRRL 21882). The strain is not only a nonproducer of aflatoxins, but also it does not produce CPA or other aflatoxin biosynthetic precursors (Dorner 2004). Genetic analysis of the strain revealed a deletion of the entire aflatoxin gene cluster (Chang et al. 2005). Ideally, afla-guard is applied to the peanut crop at 6080 days after planting, or soon after canopy closure. After application and uptake of moisture the coated conidia germinate and grow, producing abundant sporulation that is disseminated into the soil for competition with toxigenic strains that are naturally present. Research studies have shown that this biological control strategy can produce reductions in aflatoxin contamination of approximately 8090% (Dorner et al. 1998; Dorner 2004). In 2004, studies were carried out to monitor the efficacy of commercial applications of afla-guard (Dorner & Lamb 2006). FS peanuts treated with afla-guard from seven locations in Georgia and Alabama in the USA were found to have an overall mean reduction in aflatoxin of 85.2%. At two locations where treated and untreated peanuts were stored for several months and then shelled, mean aflatoxin reductions in edible grade peanuts were 69 and 98%, respectively. At both locations no shelled lots of treated peanuts tested above the USDA limit of 15 mg kg1 compared with 15.8 and 48.4% of untreated peanuts at the respective locations. Economic analysis based on the costs associated with the blanching of failed lots showed that the use of the biopesticide produced a net increase in shelled stock value at the two locations of 6.1 and 15.3%, respectively.

207

before contamination occurs. However, this early harvesting can result in reduced yield and reduced income for the farmer. Biological control technology has recently been commercialized which prevents much of the contamination that would otherwise occur. Best practices to achieve the lowest possible levels of contamination are to combine all possible management and prevention strategies to ensure and maintain a safe supply of peanuts.

References
Burdock GA, Flamm WG. 2000. Review article: Safety assessment of the mycotoxin cyclopiazonic acid. International Journal of Toxicology 19:195218. Chang P-K, Horn BW, Dorner JW. 2005. Sequence breakpoints in the aflatoxin biosynthesis gene cluster and flanking regions in nonaflatoxigenic Aspergillus flavus isolates. Fungal Genetics and Biology 42:914923. Cole RJ, Dorner JW, Holbrook CC. 1995. Advances in mycotoxin elimination and resistance. In: Stalker HT, Pattee HE, editors. Advances in peanut science. Stillwater, OK: American Peanut Research and Education Society. pp 456474. Cole RJ, Sanders TH, Dorner JW, Blankenship PD. 1989. Environmental conditions required to induce preharvest aflatoxin contamination of groundnuts: Summary of six years research. In: Hall SD, editor. Aflatoxin contamination of groundnut: Proceedings of the international workshop, 69 October 1987. Patancheru, AP, India: ICRISAT Center. pp 279287. Davidson Jr JI, Holaday CE, Bennett CT. 1981. Separation and removal of aflatoxin contaminated kernels in peanut shelling plants: Part 1. A case study. Proceedings of the American Peanut Research and Education Society 13:2945. Diener UL, Cole RJ, Sanders TH, Payne GA, Lee LS, Klich MA. 1987. Epidemiology of aflatoxin formation by Aspergillus flavus. Annual Review of Phytopathology 25:249270. Diener UL, Davis ND. 1970. Limiting temperature and relative humidity for aflatoxin production by Aspergillus flavus in stored peanuts. Journal of the American Oil Chemists Society 47:347351. Dorner JW. 2004. Combined effects of biological control formulations, cultivars, and fungicides on preharvest colonization and aflatoxin contamination of peanuts by Aspergillus species. Peanut Science 31:7986. Dorner JW. 2005. Biological control of aflatoxin crop contamination. In: Abbas HK, editor. Aflatoxin and food safety. Boca Raton, FL: CRC Press. pp 333352. Dorner JW, Cole RJ, Blankenship PD. 1998. Effect of inoculum rate of biological control agents on preharvest aflatoxin contamination of peanuts. Biological Control 12:171176. Dorner JW, Cole RJ, Sanders TH, Blankenship PD. 1989. Interrelationship of kernel water activity, soil temperature, maturity, and phytoalexin production in preharvest aflatoxin contamination of drought-stressed peanuts. Mycopathologia 105:117128. Dorner JW, Cole RJ. 1997. Distribution of aflatoxin in grade sample components of farmers stock peanuts. Peanut Science 24:4751. Dorner JW, Lamb MC. 2006. Development and commercial use of afla-guard, an aflatoxin biocontrol agent. Mycotoxin Research 21:3338.

Conclusions Mycotoxin contamination of peanuts, which can occur both before and after harvest, can be effectively managed to produce shelled peanuts that meet strict regulatory guidelines, ensuring a safe food supply. This management primarily involves techniques that remove highly contaminated kernels from the majority that are not contaminated. However, these removal steps are costly, in terms of both processing and unavoidable loss of non-contaminated kernels. It is highly preferable to take steps to prevent contamination if at all possible. Such steps include control of kernel moisture both before and after harvest. If that is not possible, prevention of pre-harvest aflatoxin contamination can be achieved by harvesting peanuts

208

J. W Dorner .
Thai CN, Blankenship PD, Cole RJ, Sanders TH, Dorner JW. 1990. Relationship between aflatoxin production and soil temperature for peanuts under drought stress. Transactions of the ASAB-E 33:324329. Urano T, Trucksess MW, Beaver RW, Wilson DM, Dorner JW, Dowell FE. 1992. Co-occurrence of cyclopiazonic acid and aflatoxins in corn and peanuts. Journal of AOAC International 75:838841. US Environmental Protection Agency. 2004. Available: http:// www.epa.gov/oppbppd1/biopesticides/ingredients/tech_docs/ brad_006500.pdf/ Van Egmond HP. 2002. Worldwide regulations for mycotoxins. In: DeVries JW, Trucksess MW, Jackson LS, editors. Mycotoxins and food safety. New York, NY: Kluwer/Plenum. pp 257269. Whitaker TB, Dickens JW, Monroe RJ. 1974. Variability of aflatoxin test results. Journal of the American Oil Chemists Society 51:214218. Whitaker TB, Dorner JW, Lamb M, Slate AB. 2005. The effect of sorting farmers stock peanuts by size and color on partitioning aflatoxin into various shelled peanut grade sizes. Peanut Science 32:103118. Whitaker TB, Hagler Jr WM, Giesbrecht FG, Dorner JW, Dowell FE, Cole RJ. 1998. Estimating aflatoxin in farmers stock peanut lots by measuring aflatoxin in various peanut-grade components. Journal of AOAC International 81:6167. Wright G, Rachaputi N, Chauhan Y, Robson A. 2005. Increasing productivity and quality of peanuts using novel crop modeling and remote sensing technologies. Summary of the International Peanut Conference, Kasetsart University, Bangkok, Thailand, 912 January 2005.

Dowell FE, Dorner JW, Cole RJ, Davidson JI. 1990. Aflatoxin reduction by screening farmers stock peanuts. Peanut Science 17:68. European Commission. 1998. Commission Regulation (EC) No. 1525/98 of 16 July 1998 amending Regulation (EC) No 194/97 of 31 January 1997 setting maximum levels for certain contaminants in foodstuffs. Official Journal of European Communities L 201:4346. Food and Agriculture Organization. 2004. Worldwide regulations for mycotoxins in food and feed in 2003. FAO Food and Nutrition Paper No. 81. Rome: FAO. Lansden JA, Davidson JI. 1983. Occurrence of cyclopiazonic acid in peanuts. Applied and Environmental Microbiology 45:766769. Parmar RS, McClendon RW, Hoogenboom G, Blankenship PD, Cole RJ, Dorner JW. 1997. Estimation of aflatoxin contamination in preharvest peanuts using neural networks. Transactions of the ASAB-E 40:809813. Rao LB, Husain A. 1985. Presence of cyclopiazonic acid in kodo millet (Paspalum scrobiculatum) causing kodua poisoning in man and its production by associated fungi. Mycopathologia 89:177180. Sargeant K, Sheridan A, OKelly J, Carnaghan RBA. 1961. Toxicity associated with certain samples of groundnuts. Nature 192:10961097. Smith Jr JS, Blankenship PD, McIntosh FP. 1995. Advances in peanut handling, shelling and storage from farmer stock to processing. In: Patee HE, Stalker HT, editors. Advances in peanut science. Stillwater, OK: American Peanut Research and Education Society. pp 500527.

Food Additives and Contaminants, February 2008; 25(2): 209218

Factors influencing fungal and aflatoxin levels in Turkish hazelnuts (Corylus avellana L.) during growth, harvest, drying and storage: A 3-year study

GUNER OZAY, FERDA SEYHAN, CEYDA PEMBECI, SENA SAKLAR, & AYSUN YILMAZ
_ TUBITAK Marmara Research Centre, Food Institute, P Box 21, 41470 Gebze, Kocaeli, Turkey .O.
(Received 28 May 2007; accepted 27 September 2007)

Abstract The levels aflatoxins in Turkish hazelnuts have been monitored over a 3-years period (20022004). Periodical sampling was made in 72 different orchards at different locations representative of the hazelnut-growing areas and post-harvest applications. Various parameters (aflatoxins, water activity, moulds) were analysed and environmental conditions (temperature and relative humidity) recorded during growing and at different stages of harvest and post-harvest processing, involving three different harvesting methods (collection in nets, from the ground, etc.) and four drying techniques (traditional sun-drying, mechanical drying, etc.). Fungal and aflatoxin analyses (HPLC) showed no significant difference except between samples which had been in contact with the ground and those which had not (at 95% confidence level). Aflatoxins levels from the orchard recorded a maximum of 0.77 0.08 ng g1 from a total of 1624 samples. Regarding harvesting and post-harvest processes, the only application where aflatoxins were detected was in samples which had been in direct contact with the ground (max. 3.18 0.03 ng g1). Aflatoxin formation was low during storage (max. 0.34 0.003 ng g1). As a result of mycological studies, a total of 5546 Aspergillus flavus (89%) and A. parasiticus (11%) species were isolated and identified from samples. The results indicated that harvesting hazelnuts into a canvas by shaking the trees, manual harvesting of mature hazelnuts where possible, use of jute instead of nylon sacks and mechanical drying technique would minimize aflatoxin levels in hazelnuts. These recommendations have been implemented and about 4000 people in the hazelnut industry have been trained in these practices.

Keywords: Hazelnut, aflatoxin, moulds, harvesting, drying, storage, prevention

Introduction Aflatoxin contamination is an important issue in the areas of food safety and international trade since they are potent carcinogens and teratogens in humans and farm animals (Moreu 1979; Bullerman 1986; Pitt 2000; Campbell et al. 2003; Magan and Olsen 2004; European Commission 2006). The total aflatoxin action threshold level for nuts is set at 20.0 ng g1 by the US Food and Drug Administration for domestic foods. The European Union (and Japan) has set their threshold levels for imported nuts, intended for direct human consumption or use as an ingredient in foodstuffs, at least five times lower at 4.0 ng g1 for total aflatoxins
Correspondence: Guner Ozay. E-mail: guner.ozay@mam.gov.tr ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701711016

(European Commission 2006). However, the low threshold levels for aflatoxin have significantly increased the probability of the rejection of treenut shipments by the major importing nations of the EU and Japan. Therefore, the determination of aflatoxin risk and preventive measures need to be implemented to minimize food safety concerns and for purely economic reasons. In Turkey, hazelnuts are traditionally sun-dried and may be subject to mould growth and, like other nuts, subsequent aflatoxin formation due to prolonged drying under humid or rainy conditions (Simsek et al. 2002). The global production of hazelnut is increasing by more than 5% annually; Turkey being the most important producer (75%

210

G. Ozay et al. Rhone Ltd. (West of Scotland Science Park, Glasgow, UK). HPLC analyses were done using a Shimadzu Class-Vp 5.0, Shimadzu RF-10AXL (florescence detector); column: ACE C18, 250 4.6 mm I.D., 5 mm particle size (Advanced Chromatography Technologies, Aberdeen, Scotland).

market share), followed by Italy (15%), Spain (3%) and the US (3%) (Annual Statistic 2006). Aspergillus flavus and A. parasiticus are known to be the primary aflatoxin-producing species, and are closely associated with agricultural environments and crops, including tree nuts (Magan and Olsen 2004; Bayman et al. 2002; Denizel et al. 1976a). A. flavus populations are influenced by agricultural and processing practices but, in many cases, the mechanism and reason are unclear. Several environmental factors are known to influence aflatoxin production, but temperature and relative humidity (RH) are critical (Northolt et al. 1976; Chiou et al. 1984; Denizel et al. 1976b). The hard shell of nuts is a good barrier against bacterial and fungal contamination (Bayman et al. 2002; Campbell et al. 2003); nevertheless, the rate and degree of contamination are dependent on temperature, humidity, soil and storage conditions. Prevention, particularly by excluding or reducing toxigenic mould growth and toxin production in susceptible food crops, is the most effective way to restrict aflatoxin contamination (Magan and Olsen 2004; Barung et al. 2006). Additional factors, such as substrate composition (Sakai et al. 1984), storage time, insect damage and presence/absence of a shell, also influence fungal growth and aflatoxin production (Schatzki and Ong 2001; Campbell et al. 2003). It is also important to recognize, however, that the interaction of all these factors may provide varying results as regards fungal growth and mycotoxin production, even on identical substrates. It is estimated that 510% of food and other agricultural crops are unfit for human/animal consumption due to fungal damage, at a cost of nearly 16 billion USD per year (Pitt and Hocking 1997; Moreu 1979). The total cost of tree nut sales lost to aflatoxin contamination averages around $50 million per year (Cardwell et al. 2001). The impact of potential aflatoxin contamination on hazelnuts as regards food safety and international trade has created the impetus to develop methods and strategies for reducing aflatoxins in pre- and postharvest hazelnut products. Therefore, the aim of this study was to determine the level of aflatoxin contamination in Turkish hazelnuts from orchard to storage and implement preventive measures in pre- and post-harvest applications.

Methods of analysis Mould enumeration. An assessment of the level of mould contamination was carried out by suspending 25 g from each hazelnut sample in 225 ml of sterile dilution solution. Homogenised samples were diluted and inoculated on Dichloran Rose Bengal Agar (DRBC-Oxoid) by the pour plate method for enumeration. The plates were incubated at 25 1 C for 57 days. Fungal counts from plates, having between 15 and 150 colony-forming units (cfu), were used in calculating the total mould count per g in each sample (Pitt and Hocking 1997; Samson et al. 1995). Isolation and identification of Aspergillus flavus group. Hazelnut samples (50 kernels) were surface disinfected by a 2 min immersion in 70% ethanol followed by 2 min in 0.4% chlorine; then, 40 kernels were plated directly (2 kernels per plate) onto Aspergillus Flavus and Parasiticus Agar (AFPA; Oxoid). Plates were incubated at 25 1 C for 35 days and examined visually under a stereomicroscope. Fungal growth was recorded after incubation and colonies having a yellow/orange colour on the underside were isolated as possible A. flavus/parasiticus growth. Any colony suspected of belonging to the A. flavus group was sub-cultured on Czapek (CZ; Merck) and Malt Extract Agar (MA; Merck) (Pitt and Hocking 1997; Samson et al. 1995). Morphological characteristics were observed during growth and identification of A. flavus and A. parasiticus was made according to the classification given by Pitt and Hocking (1997) and Samson et al. (1995). Detection of aflatoxin production ability of fungi. Pure isolates of A. flavus and A. parasiticus were transferred to YES (Yeast Extract Sucrose) Agar and incubation for 14 days at 25 C for the production of secondary metabolites (Samson et al. 1995). After the incubation period, the YES medium was mixed with chloroform in a stomacher shaker for 12 min. The chloroform extracts of culture filtrates of A. flavus and A. parasiticus were qualitatively analysed on thin-layer chromatography (TLC) plates under UV light. The TLC plates, spotted with $20 ml of chloroform extract, one spot of

Materials and methods Chemicals & instruments All chemical reagents and standards were obtained from Sigma-Aldrich-Fluka Co. Ltd. (Taufkirchen, Germany), unless otherwise stated. The aflatoxin standards were obtained from R-Biopharm

Factors influencing fungal and aflatoxin levels in Turkish hazelnuts standard aflatoxins (mixture of B1, B2, G1 and G2; R-Biopharm Rhone Ltd.) and one spot of chloroform extract plus standard aflatoxins were developed in a chloroform/acetone solvent system (90:10, v/v). The plates were observed under UV light (at 366 nm) for detection of various aflatoxins by comparison with a standard after 56 min air drying of the plates. Chemical confirmation of aflatoxins was carried out by spraying 25% H2SO4 (Reddy et al. 1970; Stack and Pohland 1975). Determination of aflatoxin (B1, B2, G1 and G2). Qualitative aflatoxin tests for detection of aflatoxin-producing fungi were performed using TLC plates and quantitative test were carried out by HPLC. After each pre-harvest, harvest and postharvest steps, samples were taken and sent to the laboratory under cold-chain conditions (4 C) to conduct aflatoxin B1 and total aflatoxin analyses. The samples were homogenized with a Waring blender using water (1:1, v:v). Aflatoxins were quantified using HPLCIAC with post-column derivatization from AOAC (999.07; AOAC 2000) (limit of detection (LOD): B140.04 ng g1; total40.1 ng g1). Determination of water activity and moisture. Water activity values of all samples were measured using a  Novasina Novatron water activity analyser at 25 C constant temperature. During harvesting and postharvest investigations, a mobile laboratory was setup in the field and moisture content and other physical experiments were conducted from this mobile unit. Moisture content of each sample was measured as previously described (AOAC 2000).

211

Statistical analysis. All results were analysed using analysis of variance followed by least-significant difference (LSD) with SPSS software ver.7.5.1 (SPSS Inc., 1999). Significance was determined at p < 0.05 for all analyses.

Sampling of hazelnuts for analysis Pre-harvest studies. Hazelnut samples were collected from three different parts of the Black Sea region, a major hazelnut-growing area of Turkey (east, middle, and west) over consecutive 3 years (2002, 2003 and 2004). The names of the regions and number of orchards are given in Table I and displayed in Figure 1. The number of the orchards was determined according the hazelnut production capacity of the region. The orchards were selected homogeneously and located at altitudes of 0250, 250500 and 500750 m in each region. However, no significant difference was found in total mould counts or aflatoxin incidence for altitude, which was then ignored during evaluation. Periodical sampling was done for 4 months (May August) from flowering stage until harvest. In each year, the
Table I. Sampling data from three hazelnut-growing regions. Number of orchards Region West Central East Total Orchards no. 124 2544 4572 City Adapazari Duzce Samsun Ordu Giresun Trabzon 2002 12 12 9 12 15 12 72 2003 12 12 9 12 15 12 72 2004 12 12 9 12 15 12 72

Figure 1. Locations of the 72 orchards (each point on the map indicates the location of the orchard from western (124), central (2544) and eastern (4572) areas of the Black Sea region).

212

G. Ozay et al. the effects of early harvesting on aflatoxin formation. Harvesting to a canvas was done within the normal harvesting period by shaking the branches to knock the hazelnuts. Harvesting from the ground was done by collecting the fallen hazelnuts (naturally matured) towards the end of the harvesting period. Hazelnuts were collected into jute or nylon sacks to determine the effect of different materials on aflatoxin formation. The harvesting scheme for each application is given in Table II. Application nos. 1, 3 and 4 were the applications carried out by some farmers under the poorest conditions. Application no. 2 is the postharvest stage performed by most farmers carry out. Application nos. 5A and 5.B were those recommended, using meshed shelves and mechanical driers. Application nos. 2 and 5 were repeated for 2 years. Drying of hazelnuts. Different sun-drying techniques used by farmers (drying on the ground/soil, concrete floor, etc.) were evaluated for levels of aflatoxin formation. Drying was also carried out using two layers of mesh shelves and a mechanical dryer operating at 40 C, which was designed specifically for hazelnut farmers. Details of the drying processes will be discussed in a future article. Hazelnuts were dried until the moisture content decreased to 5%, which is the accepted, safe moisture level. Average ambient air temperature and relative humidity in the province was 23 C/78% in 2002 and 22 C/75% in 2003, but ambient temperature reached 2729 C in the afternoon. During post-harvest treatment, the ambient relative humidity reached 96% on rainy days in both the 2002 and 2003 seasons. Samples were also collected from harvest sites under adverse conditions, i.e. during times of high humidity (rainy days or just after rain) from sites

periods between each sampling and the starting date were set according to climate and regional differences. The number of samplings in each year varied between eight and nine, depending on maturation of the hazelnuts. Sampling was carried out in each orchard according to Z pattern (AOAC 2000) and 11 trees were marked at five points in an orchard. In total, the same 55 marked trees in an orchard were used for subsequent sampling throughout the 3-year period. The sampling was done at different maturation stages until the harvesting period. Hazelnut samples (minimum 3 kg for each case) were placed in a polystyrene box with a cooling gel (pre-frozen to _ 20 C) and transferred to the TUBITAK MRC Food Institute within the same day. Upon arrival, the temperature and weight of the samples were recorded and samples were subjected to total mould count, mould isolation, water activity and aflatoxin analyses. For aflatoxin tests, hazelnut samples from flowering stage to kernel formation were homogenized after removing the green hulls. After kernel formation, the hazelnuts were cracked and shells were removed manually, taking precautions against possible contamination.

Harvesting and post-harvest studies Harvesting. Tombul variety of hazelnuts (Corylus avellana L.) were harvested during the August 2002 and 2003 seasons in two provinces, located to the east and west of the Black Sea. Harvesting was done using three different techniques; early manual harvesting, harvesting to a canvas (shaking the branches to collect the hazelnuts from canvas laid under the branches) and harvesting from the ground (windfall). Early manual harvesting was done 1 week before the officially announced harvesting period to determine

Table II. Applied harvesting techniques to determine their effects on aflatoxin formation in hazelnuts. Applications Post-harvest stages

No. 1 2003 Manual/early harvesting Nylon 10 Days in nylon sacks

No. 2 2002/2003 Ground* Jute 3 Days on concrete floor

No. 3 2002 Ground Nylon 3 Days on ground

No. 4 2003 Ground Nylon 7 Days on ground plus 3 days in nylon sack Ground

No. 5 2002/2003 Canvas Jute 3 Days on mesh shelves

(1) Harvesting Sacks used (2) Withering

(3) De-hulling (4) Drying

Ground

Concrete floor

Ground

(A) Mesh shelves (B) Mechanical drying

*Ground application was preformed in direct contact with soil.

Factors influencing fungal and aflatoxin levels in Turkish hazelnuts which were in contact with soil. The hazelnuts were protected from rain by nylon canvas without any air circulation. Storage of hazelnuts. Dried hazelnuts were stored under controlled (5 2 C, 65 5% RH) and uncontrolled conditions (storage on-site without any temperature and humidity control). Temperature and environmental conditions in the storage facility were recorded using AZ Data loggers. Training and dissemination of knowledge. To disseminate information about aflatoxin and preventive measures, approximately 4000 people (growers, traders and exporters) were trained over the three years. Posters (20,000) and two series of brochures (37,000) were also published and distributed.

213

Results and discussion Sampling from orchards The sampling was done from the upper and lower branches of the same trees from each orchard on three consecutive years (2002, 2003 and 2004). The number, total mould count and water activities of samples from 72 orchards are given in Table III. Statistical analysis was carried out if there was a significant difference among altitudes, regions or sampling year (SPSS 10.1 version; t-test). However, there were no significant difference among altitude or regions over three consecutive years in terms of total mould counts (p40.05). The samples of different maturation stages were also analysed for their compositional characteristics and enzyme activities to investigate if any correlation existed with aflatoxin formation; these results are reported elsewhere (Seyhan et al. 2007). However, the number of aflatoxin-contaminated samples were insufficient for a correlation study. In Turkish hazelnuts, the mould count ranged between 1.8 101 and 3.8 106 cfu g1 from flowering stages to harvest. Arrus et al. (2005) investigated the mould count of Brazil nuts from various regions and found differences between the regions in the range 106107 cfu g1. In 109 freshly harvested

maize samples, the total mould counts ranged 1.9 1043.5 106 cfu g1, and ranged 1.8 102 1.6 104 (average: 3.4 103) in 32 cashew nut samples from Nigerian markets (Adebajo and Diyaolu 2003; Ono et al. 2006). Total mould count of 103104 cfu g1 have been reported in 143 freshly harvested pistachios from Turkey and the mould count increased to 105106 cfu g1 during storage (Heperkan et al. 1994). Aflatoxin was not detected (LOD for B1 < 0.04 ng g1; total aflatoxin <0.1 ng g1) in 640 and 336 samples from orchards, from the flowering stage to harvest, for 2003 and 2004, respectively. Aflatoxin was detected in seven orchards in 2002 prior to harvest at low levels and the maximum level was 0.77 0.08 ng g1. The percentage of aflatoxin positive samples were only 1.54% of 648 samples in 2002. There was a decreasing trend in water activity of samples prior to harvesting but the level varied between 0.95 and 0.99. The optimum water activity for growth of A. flavus is 0.980.99 (ICMSF 1996; Pitt and Miscamble 1995) and the range for growth of A. flavus and A. parasiticus is between 0.99 and 0.80 (Bresler 1998; Holmquist 1983; Nesci 2003). Studies on hazelnuts and pistachios suggest that optimum temperature and RH for aflatoxin production is 25 30 C and 9799%, respectively (Diener and Davis 1967; Schindler et al. 1967; Northolt et al. 1976; Simsek et al. 2002). Although the recorded water activities of the samples fits the optimum conditions for aflatoxin formation, in this study aflatoxin was not detected during 2003 and 2004. Incidence of A. flavus and A. parasiticus. Hazelnut kernels infected with A. flavus and A. parasiticus (AFP) were positive according to the yellow/orange colour of the underside of colonies. The% incidence for AFP in 2002, 2003 and 2004 is given in Table IV for the three regions listed in Table I. Sampling was carried out in hulls until the start of kernel formation. After kernel formation, sampling involved removing the green hulls surrounding the kernel. The% incidence of A. flavus/parasiticus (AFP%) in developing hazelnut samples differed over the three consecutive years, possibly caused by mycoflora alterations and stress/activating factors affecting fungal growth via environmental changes (Choudhary and Sinha 1993; Bayman et al. 2002; Campbell et al. 2003; Magan and Olsen 2004). In general, although the AFP% fluctuates through the sampling period there is an increasing trend from the start of kernel formation over the three consecutive years. The incidence for AFP fluctuates due to the changes in surface flora affected by alterations in ecological and climate conditions (Prado 1991).

Table III. Total mould count and water activities of samples from 72 orchards. Total mould count [cfu g1] (minmax) 1.00 1023.87 106 1.22 1032.09 105 1.80 1013.40 105

Years 2002 2003 2004

No. of samples 648 640 336

Aw (minmax) 0.950.99 0.960.99 0.950.98

214

G. Ozay et al. producer of aflatoxin with a close relationship to the agricultural environment and crop (Hill et al. 1983; Abdel-Hafez and Saber 1993; Abdel-Gawad and Zohri 1993; Bayman et al. 2002; Sims ek et al. 2002; Adebajo and Diyaolu 2003; Arrus et al. 2005b). Aflatoxin production capacity of isolated fungi Isolates of A. flavus and A. parasiticus were monitored for their toxigenic potential. The percentage of toxigenic fungi differed according to year; however, no significant difference was found among regions (p40.05). In 2002 and 2003, toxigenic fungi represented 4% of 1920 identified fungi but was 48% of 1028 identified fungi in 2004. Toxigenic fungi totalled 566 (19%) of the 2948 identified fungi for the three consecutive years. Toxin-formation potential of some fungi isolated from various foodstuffs is given in Table VII, with the range varying between 5.7 and 55% (Jayaraman and Kalyanasundaram 1990; Mishra and Daradhiyar 1991; Kamphuis et al. 1992; Munimbazi and Bullerman 1996; Vazquez Belda et al. 1996; Freitas Costa and Scussel 2002; Gatti et al. 2003). Kamphuis et al. (1992) determined the percentage of the toxigenic fungi at 5.7% in 35 examined dried maize samples. Munimbazi and Bullerman (1996), in a study of 95 isolates from foodstuffs in Burundi, found that 39% of A. flavus and 100% of A. paraciticus were toxigenic. Similarly, Doster et al. (1994) isolated 11 Aspergillus spp. from Californian pistachios and reported that 65% were toxigenic. The number of isolates (2948) in this study is larger than those listed in Table VII and established that the percentage of toxigenic fungi varied with year ranging 4-48% over the three consecutive years. Harvesting techniques Harvesting methods used in this study were based on the worst conditions operated by farmers and the recommended applications to prevent aflatoxin formation. Hazelnut orchards are located on very steep sites in the Black Sea region of Turkey, between 0 and 1000 m above sea level. Mature

Temperature and relative humidity in the region was obtained from the Turkish State Meteorological Services General Directorate and average, minimum and maximum values are given in Table V. Choudhary and Sinha (1993) studied the relationship between A. flavus and competing moulds in maize, reporting that toxigenic A. flavus was mostly isolated in monsoon (63%) and winter periods (52%). Relative humidity of the environment and moisture of the substrate were significantly correlated with A. flavus in maize (Choudhary and Sinha 1993). A negative correlation was observed between A. flavus growth and rainfall. Significant competition was demonstrated between A. flavus, Penicillium spp. and A. niger, and a significant positive correlation was recorded in seasons having higher temperatures (Choudhary and Sinha 1993). A. flavus and A. parasiticus were identified at the genera level in accordance with morphological characteristics and taxonomic criteria (Samson et al. 1995; Pitt and Hocking 1997). The percentage of A. flavus and A. parasiticus differed in each year (see Table VI) but the percentage of A. flavus was greater than A. parasiticus 89 and 11%, respectively, from 5546 identified samples averaged over 3 years. Systematic studies on the microflora of fresh nuts are limited, but it appears that A. flavus and A. parasiticus have a particular affinity for nuts and oil seeds (Pitt 2000). A. flavus was the dominant fungi isolated from foodstuffs, defined as the primer
Table IV. Incidence for A. flavus/parasiticus (AFP%) in hazelnut samples during 2002, 2003 and 2004. AFP% for sampling period 2002 Region Orchard no. 1 2 East Central West 24 21 27 3 2003 4 1 2 3 4 1 2004 2 3 4

0 6 24 68 1 8 3 56 10 53 26 47 3 5 18 73 1 6 0 63 15 36 27 61 0 7 27 79 3 3 1 63 4 40 39 54

*Sampling periods were selected to coincide with the same period in the three consecutive years, i.e. 1: 530 July (flowering stage); 2: 319 June (kernel development); 3: 2030 June (development stage); 4: 420 August (harvesting stage).

Table V. Air temperature and relative humidity in the region (Turkish State Meteorological Services General Directorate). Average Region West Central East Province Duzce Akcakoca Samsun Ordu Giresun Trabzon RH (%) 68 83 77 74 70 70 Temp ( C) 13.7 13.2 14.8 14.9 15.1 15.2


Min RH (%) 28 37 31 28 19 22 Temp ( C) 7 2 2 2 1 0



Max RH (%) 95 96 96 95 92 95 Temp ( C) 28.5 27.4 27.2 28.4 27.6 27.8

Factors influencing fungal and aflatoxin levels in Turkish hazelnuts hazelnuts naturally drop to the ground and, thus, this harvesting technique is easy and occasionally used in the region; labour costs are low and it is possible to collect all hazelnuts at the same time. However, there is a risk of mould contamination on the surface of hazelnut hulls. In the early manual harvesting technique, labour costs are high and all hazelnuts cannot be collect due to differences in maturity; thus, a number of picking times are requires. Producers sometimes use this technique to sell hazelnuts earlier; however, both mature and immature nuts (low weight/small size hazelnuts) are collected. Schatzki and Pan (1997) studied the distribution of aflatoxin in small pistachios where low weight (small size) might be an indicator of pre-harvest weakness or damage. Smaller nuts showed greater aflatoxin content, but the size dependence was not significant. Harvesting to canvas by shaking is recommended to prevent contact of hazelnuts with the ground; however, it was not practical in very steep orchards due to the confined space. Farmers were recommended to use manual harvesting of mature nuts into small plastic or wooden baskets. Drying of hazelnuts In Turkey, hazelnuts are generally dried under the sun by small producers. The problem is that the weather in the Black Sea region is very humid and the harvesting season can be wet. Drying may extent to 23 weeks during rain, which increases the risk of mould contamination and consequent formation of aflatoxins. Some farmers collect hazelnuts 1 week earlier in the harvesting period for economic
Table VI. Number of isolates and % of A. flavus and A. parasiticus in three consecutive years. No. of A. No. of A. flavus % parasiticus % 2105 1907 943 4955 84 95 90 89 392 92 107 591 16 5 10 11

215

Years 2002 2003 2004 Total/Average

Total no. of isolates 2497 1999 1050 5546

considerations and prefer to use nylon sacks due to ease of availability and cheaper price. However, jute sacks are recommended owing to their high air permeability, but are more expensive than nylon sacks. The hulls of the hazelnuts rot and heat is produced when the sacks are over-packed and kept for a few days before laying under sun to dry. This application is simulated in application no. 1 (see Table II). Due to respiration and reduced air circulation, hazelnuts are become wet and the risk of mould contamination increases. In 2002, jute sacks were used during the applications but aflatoxin levels were not determined in the samples. In 2003, nylon sacks were used instead of jute sacks and although samples were left in them for 10 days (Table II), moisture decreased only 2.27% (from 29.85 to 27.58%). A common application in the region is the harvesting of hazelnuts from the ground and drying on a concrete floor (Table II); the latter being generally preferred in the region due to the shorter drying time. The concrete floor is heated by the sun which helps dry the hazelnuts. Picking hazelnuts takes a long time in large orchards and hazelnuts may lie on the ground, which are favourable conditions for mould contamination and aflatoxin formation. In 2002, hazelnuts lay on the ground for 3 days; in 2003, they were 7 days on the ground and 3 days in nylon sacks (see Table II). These conditions were designed to observe aflatoxin formation. The hazelnuts, which lay for 7 days on the ground plus 3 days in nylon sacks, partially dried and it appears these hazelnuts had a short drying time, but 10 days is the general holding period before normal drying. The hazelnuts lost some moisture during this period depending on the weather conditions. Mechanical drying time was significantly less (33 h) than sun-drying (110 h) and was independent of weather conditions. Although it is timesaving and has benefits for improved safety of the final product, it is not generally practiced by small farmers in the region due to energy costs compared to traditional sun-drying.

Table VII. Percentage incidence of some toxigenic fungi isolated from various foods. Food Foodstuff from Burundi Cheese production from Arzua California pistachios Brazilian coffee Rice bran Maize No. of samples 95 isolates 120 Aspergillus sp. 11 Aspergillus sp. % Incidence 39% A. flavus 100% A. paraciticus 19% 65% 25% mycotoxin 13% aflatoxin 55% toxigenic 5.7% Reference Munimbazi et al. (1996) Vazquez et al. (1996) Doster et al. (1994) Freitas Costa et al. (2002) Jayaraman et al. (1990) Kamphuis et al. (1992)

34 samples 35 samples

216

G. Ozay et al. (just after or during rainfall) in 2003 (29 samples) and 2004 (50 samples). In 2003, the percentage of aflatoxin-positive samples was 43% with a maximum aflatoxin level of 1.02 0.01 ng g1. There were no samples over the EC limit (4.0 ng g1) in 2003. In 2004, the percentage of aflatoxin-positive samples was 16% with a maximum aflatoxin level of 24.36 0.24 ng g1; 6% of the contaminated samples had levels 44.0 ng g1 aflatoxins and 10% had <4.0 ng g1. Water activities of these samples were higher than 0.80 due to the extended drying periods during rain, which increased the risk for aflatoxin formation. The optimum relative humidity conditions for aflatoxin formation are 9799% (at equilibrium conditions, which correspond to 0.970.99 water activity) (Denizel 1976b). It has also been reported that aflatoxin can be formed at water activities of $0.82 (Denizel 1976a). FAO (1995) recommends that nuts should be dried to 0.70 water activities in a short period of time and storage conditions should not exceed 70% relative humidity. Therefore, the risk for aflatoxin formation increases during rain and high humidity conditions. Storage applications Dried hazelnut samples were stored in controlled (5 C, 65 5% RH) and uncontrolled storage conditions. Aflatoxin was not detected in the controlled storage conditions after 2 years of storage, although aflatoxin-producing fungi were recorded in the stored samples. Uncontrolled storage involved conditions in the region similar to those operated by farmers and traders. Temperature and humidity was recorded throughout the storage period; RH ranged 4884% in uncontrolled conditions throughout the 2-year storage period. Aflatoxin was detected in samples that were in direct contact with the ground (soil) during harvesting and drying. The level was 3.18 0.03 ng g1 for hazelnuts harvested from the ground after 3 days (application no. 3 in Table II). Although increased aflatoxin contamination was expected during uncontrolled storage of samples from the no. 3 application, the level was low (max: 0.34 ng g1). The reason for the low level of aflatoxin formation in contaminated samples during uncontrolled conditions may be competitive environmental conditions. Conclusions Fungal contamination and subsequent production of aflatoxin can occur in hazelnuts in the orchard, at harvest, during post-harvest operations and in storage. Our results indicate that, although the risk of aflatoxin formation is present in the orchard, the

Aflatoxins in hazelnuts samples from post-harvest stages Duplicate harvesting and post-harvest applications were performed in eastern and western sites of the Black Sea region and the results of post-harvest aflatoxin analyses are given in Table VIII. No aflatoxin was detected in the samples from application nos. 2 and 5 during 2002 and 2003, but contamination was recorded from applications in which hazelnuts were in direct contact with soil. Results showed that hazelnuts harvested from the ground (1.02 ng g1 total aflatoxin) and lying on the ground for 3 days after the de-hulling process contained aflatoxin (2.18 0.02 ng g1 B1 and 3.18 0.03 ng g1 total aflatoxin). The maximum detected level of total aflatoxin was 3.18 0.03 ng g1 and for B1 was 2.18 0.02 ng g1, which is very close to the EC limits (2.0 ng g1 for B1 aflatoxin and 4.0 ng g1 for total aflatoxin). Direct contact with ground is crucial and any harvesting and drying techniques should prevent direct contact of hazelnuts with the ground. Early manual-harvested hazelnuts, held in nylon sacks for 10 days and dried on the ground, had a total aflatoxin level of 0.6 0.01 ng g1. Similarly, hazelnuts harvested after 7 days on the ground, subsequently held in nylon sacks for 3 days and dried on the ground had a total aflatoxin level of 0.44 0.004 ng g1. Nylon sacks are not recommended as a storage bag due to low air circulation and respiration; the hazelnuts become wet, which enhance mould growth and, consequently, aflatoxin formation. For aflatoxin formation, the effect of direct contact with the ground is clear. On the other hand, the effect of early harvest on aflatoxin formation is uncertain because counting the number of immature nuts in the 200-kg of harvested hazelnuts was not possible and the distribution of immature hazelnuts was not known. Orchard samples under poorest conditions In addition to post-harvest applications, sampling under adverse conditions was also carried out. Samples were taken randomly during the drying stage under conditions of high relative humidity

Table VIII. Aflatoxin analyses during post-harvest. Aflatoxin (ng g1) Sample Application Application Application Application no. no. no. no. 1 3 3 4 (harvest (harvest (harvest (harvest stage stage stage stage 4) 1) 2) 4) B1 0.42 0.004 1.02 0.01 2.18 0.02 0.26 0.002 Total 0.60 0.01 1.02 0.01 3.18 0.03 0.44 0.004

Factors influencing fungal and aflatoxin levels in Turkish hazelnuts most important stages to prevent aflatoxin formation are harvesting and post-harvest, including storage. Aflatoxins detected during sampling from orchards showed a maximum of 0.77 0.08 ng g1 from 1624 samples (20022004). There was no significant difference between regions, altitudes or samples collected from upper and lower branches (p40.05). There was an increasing trend in AFP% as the hazelnuts reached maturity; thus, precautions during harvesting and post-harvest are essential in minimizing the aflatoxin contamination risk. Three harvesting and four drying techniques were studied. Harvesting to a canvas was the recommended collection method as it prevented the contact with the ground. Due to the steepness of the sites, this is not always practical; then manual harvesting of mature hazelnuts into plastic or wooden baskets is recommended. It is important that contact of hazelnuts with the ground is prevented as there is increased risk of contamination for hazelnuts lying on the ground for long periods of time and exposed to humid weather or rain. Mechanical drying of hazelnuts is the recommended technique to effectively prevent aflatoxin formation. It took $33 h to dry 300 kg of hazelnuts from $26 to 5% moisture content at 40 C. Sun-drying techniques are dependent on weather conditions and take longer than mechanical drying. Drying on mesh shelves is hygienic and safe but two layers are insufficient for an entire crop; it needs more layers and drying time, which took $4 days, weather depending. Drying on the ground is not recommended, as direct contact with the ground results in aflatoxin formation and prolonged drying times 33.5 days depending on weather conditions. The maximum detected level of total aflatoxin was 3.18 ng g1 in post-harvest stages. The total number of the samples analysed in this study was 2113 over 3 years, including the sampling under orchards, harvesting and storage conditions. A total of 3.6% of the 2113 samples were aflatoxincontaminated; 3.9% of these samples (max: 24.36 0.24 ng g1) were over 4.0 ng g1 total aflatoxin and 96.1% were lower than the EC limit (4.0 ng g1). Hazelnuts harvested from the ground and allowed to remain on the ground posed the greatest danger; aflatoxin was also detected in hazelnuts kept in nylon sacks and those dried on the ground, but the levels again were lower than the EC limit. In conclusion, pre-harvest contamination of hazelnuts by aflatoxins is a risk factor; thus, future sampling studies should focus on identifying the factors which increase aflatoxin contamination on the tree. Research into competing natural compounds/organisms for the prevention of aflatoxin

217

contamination on the tree and during storage would also be valuable.

Acknowledgements The authors gratefully acknowledge the Hazelnut Promotion Group (Turkey) for financial support, MRC researchers and research technicians for their assistance, Local Directorates of the Ministry of Agriculture and Rural Affairs and farmers for providing their orchards for sampling over the three years, cooperation of local authorities during field studies and Giresun Hazelnut Research Institute for the use of their facilities. References
Abdel Gawad KM, Zohri AA. 1993. Fungal flora and mycotoxins of six kinds of nut seeds for human consumption. Mycopathologia 124:5564. Abdel Hafez AI, Saber SM. 1993. Mycoflora and mycotoxin of hazelnut (Corylus avellana L.) and walnut (Juglans regia L.) seeds in Egypt. Zentralblatt fur Mikrobiologie 148:13747. Adebajo LO, Diyaolu SA. 2003. Mycology and spoilage of retail cashew nuts. African Journal of Biotechnology 2:369373. Annual Statistic 2006. Hazelnut Promotion Group. Available: http://www.ftg.org.tr/devam_eng/export.htm. Accessed 28 May 2007. AOAC. 2000. Official Methods of Analysis of AOAC International. 16th ed. Gaithersburg, MD: AOAC International. Arrus K, Blank G, Abramson D, Clear R, Holley RA. 2005a. Aflatoxin production by Aspergillus flavus in Brazil nuts. Journal of Stored Products Research 41:513527. Arrus K, Blank G, Clear R, Holley RA, Abramson D. 2005b. Microbiological and aflatoxin evaluation of Brazil nut pods and the effect of unit processing operations. Journal of Food Protection 68:10601065. Bansal NK, Garg HP. 1987. Solar crop drying. In: Mujumdar AS, editor. Advances in drying. Vol. 4. New York: Hemisphere Publishing. Barung D, Bhatnagar D, van Egmond HP, van der Kamp JW, van Osenburggen WA, Visconti A. 2006. The Mycotoxin Fact Book: Food and Feed Chain. Wageningen, The Netherlands: Wageningen Academic. p 182. Bayman P, Baker JL, Mahoney NE. 2002. Aspergillus on tree nuts: Incidence and associations. Mycopathologia 155:161169. Beuchat LR. 1975. Incidence of moulds on pecan nuts at different points during harvesting. Applied Microbiology 29:852854. Bullerman LB. 1986. Mycotoxins and food safety. Food Technology 40:5965. Campbell BC, Molyneux RJ, Schatzki TF. 2003. Current research on reducing pre- and post-harvest aflatoxin contamination of US almond, pistachio, and walnut. Journal of Toxicology Toxin Reviews 22:225266. Cardwell, KF, Desjardins, A, Henry, SH, Munkvold, G, Robens, JM, 2001. Mycotoxins: The Cost of Achieving Food Security and Food Quality. APSnet, Mycotoxins: August. Available: http://www.apsnet.org/online/feature/mycotoxin/top.html. Accessed 28 May 2007. Chiou RYY, Koehler PE, Beuchat LR. 1984. Hygroscopic characteristics of peanut components and their influence on growth and aflatoxin production by Aspergillus parasiticus. Journal of Food Protection 47:791794.

218

G. Ozay et al.
Magan N, Olsen M. 2004. Mycotoxins in Food: Detection and Control. Abington, UK: Woodhead Publishing. p 17. Mishra NK, Daradhiyar SK. 1991. Mold flora and aflatoxin contamination of stored and cooked samples of pearl millet in Paharia Tribal Belt of santhal Pargana, Bihar, India. Applied Environmental Microbiology 57:12231226. Moreu C. 1979. Moulds, Toxins and Food. 2nd ed. London: Wiley. p 477. Munimbazi C, Bullerman LB. 1996. Molds and mycotoxins in foods from Burundi. Journal of Food Protection 59:869875. Nejad MK, Tabil GH, Mortazavi A, Kordi AS. 2003. Effect of drying methods on quality of pistachio nuts. Drying Technology 21:821838. Northolt MD, Verhulsdonk CAH, Soentoro PSS, Paulsch WE. 1976. Effect of water activity and temperature on aflatoxin production by Aspergillus parasiticus. Journal of Milk and Food Technology 39:170174. Ono EYS, Biazon L, Silva M, Vizoni E, Sugiura Y, Ueno Y, Hirooka EY. 2006. Fumonisins in corn: correlation with Fusarium sp. count, damaged kernels, protein and lipid content. Brazilian Archives of Biology and Technology 49:6371. Pitt JI. 2000. Toxigenic fungi: which are important? Medical Mycology 38:1722. Pitt JI, Miscamble BF. 1995. Water relations of Aspergillus flavus and closely related species. Journal of Food Protection 58:8690. Pitt JI, Hocking AD. 1997. Fungi and Food Spoilage. London: Blackie Academic. Reddy TV, Vishwanathan L, Subramanian TAV. 1970. Thin layer chromatography of aflatoxins. Analytical Biochemistry 38:568571. Sakai T, Sugihara K, Kozuka H. 1984. Growth and aflatoxin production of Aspergillus parasiticus in plant materials. Journal of Hygienic Chemistry 30:6268. Samson RA, Hoekstra ES, Frisvad JC, Filtenborg O. 1995. Introduction to Foodborne Fungi, 4th ed. Barn, The Netherlands: Centraalbureau voor Schimmelcultures. p 322. Schatzki TF, Pan J. 1997. Distribution of aflatoxin in pistachios. 4. Distribution in small pistachios. Journal of Agricultural Food Chemistry 45:205207. Schatzki TF, Ong MS. 2001. Dependence of aflatoxin in almonds on the type and amount of insect damage. Journal of Agricultural and Food Chemistry 49:45134519. Schindler AF, Palmer J, Eisenberg W. 1967. Aflatoxin production by Aspergillus flavus as related to various temperatures. Applied Microbiology 15:10061009. Seyhan F, Ozay G, Saklar S, Ertas E, Satr G. 2007. Chemical changes of three native Turkish hazelnut (Corylus avellana L.) during fruit development. Food Chemistry 105:590596. Simsek O, Arici M, Demir C. 2002. Mycoflora of hazelnut (Corylus avellana L.) and aflatoxin content in hazelnut kernels artificially infected with Aspergillus parasiticus. Nahrung 46:194196. Stack ME, Pohland AE. 1975. Collaborative study of a method for chemical confirmation of the identity of aflatoxins. Journal of the Association of Official Analytical Chemists 58:110113. Vazquez Belda B, Fente Sampayo CA, Quinto Fernandez E, Franco Abuin C, Rodriguez Otero JL, Cepeda Saez A. 1996. Incidence of toxigenic molds in farm-level cheese making units from Arzua (La Coruna, Spain). Food Science and Technology International 1:9195.

Choudhary AK, Sinha KK. 1993. Competition between a toxigenic Aspergillus flavus strain and other fungi on stored maize kernels. Journal of Stored Products Research 29:7580. Denizel T, Jarvis B, Rolfe EJ. 1976a. A field survey of pistachio (Pistacia vera) nut production and storage in Turkey with particular reference to aflatoxin contamination. Journal of the Science of Food and Agriculture 27:10211026. Denizel T, Rolfe EJ, Jarvis B. 1976b. Moisture-equilibrium relative humidity relationships in pistachio nuts with particular regard to control of aflatoxin formation. Journal of the Science of Food and Agriculture 27:10271034. Diener UL, Davis ND. 1967. Limiting temperature and relative humidity for growth and production of aflatoxin and free fatty acids by Aspergillus flavus in sterile peanuts. Journal of the American Oil Chemists Society 44:259263. Doster MA, Michailides TJ. 1994. Aspergillus molds and aflatoxins in pistachio nuts in California. Phytopathology 84:583590. European Commission. 2006. Commission Regulation (EC) no. 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Communities L 364:524. FAO. 1995. Edible nuts. Non-wood forest products; No. 5. Rome: Food and Agriculture Organization of the United Nations. Freitas Costa LL, Scussel VM. 2002. Toxigenic fungi in beans (Phaseolus vulgaris L.) classes black and color cultivated in the state of Santa Catarina, Brazil. Brazilian Journal of Microbiology 33:138144. Gatti MJ, Fraga ME, Magnoli C, Dalcero AM, Rocha Rosa CA. 2003. Mycological survey for potential aflatoxin and ochratoxin producers and their toxicological properties in harvested Brazilian black paper. Food Additives and Contaminants 20:11201126. Heperkan D, Aran N, Ayfer M. 1994. Mycoflora and aflatoxin contamination in shelled pistachio nuts. Journal of the Science of Food and Agriculture 66:273278. Hill R, Blankenship PD, Cole RJ, Sanders TH. 1983. Effect of soil moisture and temperature on preharvest invasion of peanuts by the Aspergillus flavus group and subsequent aflatoxin development. Applied and Environmental Microbiology 45:628633. Hocking AD. 1982. Aflatoxigenic fungi and their detection. Food Technology in Australia 34:236238. ICMSF. 1996. International Commission on Microbiological Specifications for Foods, Vevey, Switzerland. Jayaraman P, Kalyanasundaram I. 1990. Natural occurrence of toxigenic fungi and mycotoxins in rice bran. Mycopathologia 110:8185. Kamphuis HJ, Horst MI, Samson MA, Ramboust FM, Notermans S. 1992. Mycological condition of maize products. International Journal of Food Microbiology 16:237245. Lopez A, Pique MT, Boatella J, Ferran A, Garcia J, Romero A. 1988. Drying characteristics of the hazelnuts. Drying Technology 16:627649. Lopez A, Pique MT, Boatella J, Parcerisa Romero A, Ferran A, Garcia J. 1997a. Influence of drying on the hazelnut quality. I. Lipid oxidation. Drying Technology 15:965977. Lopez A, Pique MT, Ferran A, Romero A, Boatella J, Garcia J. 1997b. Influence of drying on the hazelnut quality. II. Enzymatic activity. Drying Technology 15:979988.

Food Additives and Contaminants, February 2008; 25(2): 219230

Review

Mycotoxins in small grains and maize: Old problems, new challenges

J. DAVID MILLER
Ottawa-Carleton Institute of Chemistry, Department of Chemistry, Carleton University, Ottawa, Ontario, Canada K1S 5B6
(Received 25 July 2007; accepted 8 October 2007)

Abstract This paper reviews the challenges relating to chronic contamination of small grains and maize with deoxynivalenol and related compounds, fumonisin and the use of ensiled cereals in cool dairy areas. Uncertainties in the tolerable daily intakes for deoxynivalenol and fumonisin are discussed as they have the potential to affect current regulatory limits. In addition, climate change is resulting in more extreme rainfall and drought events which favour formation of deoxynivalenol and fumonisin, respectively. The development and refinement of models for predicting mycotoxin accumulation from weather data will become an essential tool for managing these events. Such models are also important for providing timely food aid to developing countries, which experience increased occurrence of acute toxicities, especially in children. Chronic contamination of silage in some areas with some Penicillium toxins deserves more attention in terms of their economic effects and possible implications for the purity of milk.

Keywords: Deoxynivalenol, neurotoxicity, fumonisin, neural-tube birth defects, silage, roquefortine, festuclavine, PR toxin, climate change

Introduction It seems timely to review the progress made on mycotoxins research in cereals over the past 15 years and consider the challenges remaining and new problems on the horizon. As noted in an earlier review (Miller 1995), toxigenic fungi in crops have been historically divided into two distinct groups. The first includes those that invade and produce toxins before harvest and the second group, which form toxins after harvest, are known as storage fungi. However, the source of the fungi in both instances is the field (Miller 1995). Four types of toxigenic fungi can be identified: (1) plant pathogens, such as Fusarium graminearum; (2) fungi that produce mycotoxins on senescent or stressed plants, such as F verticillioides and Aspergillus flavus on maize and . A. carbonarious on grapes; (3) fungi that colonise the plant and predispose the commodity to mycotoxin contamination after harvest e.g. A. flavus in subtropical maize and (4) fungi that are found in the soil or decaying plant material that occur on the developing kernels in the field and later proliferate in storage if conditions permit, e.g. Penicillium verrucosum on cereals, P. roqueforti
Correspondence: J. David Miller. E-mail: david_miller@carleton.ca ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030701744520

complex on ensiled materials and A. flavus on many commodities. The major toxins that contaminate maize and small grains (wheat, triticale, barley) pre-harvest are deoxynivalenol (replaced in some areas by nivalenol) and zearalenone, fumonisin and aflatoxin on maize. For two of these toxins, namely deoxynivalenol and fumonisin, there are unresolved issues that might affect their hazard assessment. Because they are common in grain, this represents a level of uncertainty that perhaps deserve more attention in this review. There are other mycotoxins that can cause problems occasionally in small grains and maize, the most important being the Fusarium toxin, T-2, which is normally associated with a derivative, HT-2 toxin. Alimentary toxic aleukia (ATA) disease was described prior to 1900 and was associated with the ingestion of overwintered grain. During World War II, Russians were forced to eat grain left in the field. Thousands of people were affected resulting in the elimination of entire villages (Mirocha 1984; Beardall and Miller 1994). Strains of fungi isolated from the grains at the time were later shown to produce T-2 and related toxins ( Joffre and

220

J. D. Miller dominate depends on temperature. These species also vary somewhat in pathogenicity; F. graminearum is regarded as the most virulent, although all three species can cause epidemics. Wheat, maize and barley are most affected by these pathogens (Miller 1994; Mesterhazy 2003; Snijders 2004) and by their toxins. These three crops comprise two thirds of the worlds cereal production. Contamination of oats, rye and triticale has also been reported to contain Fusarium mycotoxins (Scott 1989; Gareis et al. 2003). In parts of Europe, F. poae is also an important producer of nivalenol on small grains (Thrane et al. 2004) and nivalenol is commonly reported in European oat samples (Gareis et al. 2003). F. graminearum is associated with wheat and maize grown in warmer areas (e.g. southern Ontario) and F. culmorum, in cooler areas (e.g. northwestern Europe, but see below). The influence of temperature relates to conditions that allow a sustained period of warm weather (daytime temperatures 430 C) regardless of daily means. The most pathogenic species, F. graminearum and F culmorum, . are generally the most common species found. Since the 1890s, Fusarium head blight has been common in wheat from North America and China (Miller 1994; Wang and Miller 1988; Chen et al. 2000; Goswami and Kistler 2004). In the 1980s and 1990s, F. culmorum was the dominant species in cooler wheat-growing areas, such as Finland, France, Poland and The Netherlands (Snijders and Perkowski 1990; Miller 1994; Toth et al. 2004), but this trend has apparently changed in recent years as European summers have reached record warm temperatures such that F. graminearum largely dominates (Xu et al. 2006). Morphologically identical isolates of F. graminarum (Gibberalla zeae) can produce either DON and zearalenone or nivalenol and zearalenone as the principal toxic metabolites that accumulate in grain. Within the former group, some strains produce DON by the 3-acetylated precursor and others make the 15 acetylated precursor. DON-producing strains with the 15-acetylated precursor dominate in North and South America. DON-producing strains with the 3-acetylated precursor are common in Europe and Asia (Miller et al. 1991). The Asian and New World strains are genetically distinct (ODonnell et al. 2000). Nivalenol-producing strains of F. graminearum are common in parts of Europe, Japan and Australasia but very uncommon in the Americas. F. culmorum produces DON and zearalenone ( Miller et al. 1991; Jennings et al. 2004; and references cited therein; Toth et al. 2004). The crown rot form of F. graminearum Group 1 is now called F. pseudograminearum (G. coronicola; Aoki and ODonnell 1999) but also produces deoxynivalenol and nivalenol (Clear et al. 2006).

Hagen 1977). These are mainly F. sporotrichioides toxins, a species that grows on wet grain left in the field and to some extent on the glumes of small grains (Miller 1994; Miller et al. 1998). In parts of Europe, F. langsethiae is also an important producer of T-2 toxin on small grains (Thrane et al. 2004; Torp and Nirenberg 2004). However, despite the vast literature on T-2, incidence data show that material concentrations of this toxin are uncommon in most growing areas. This is because most grain is harvested under warm, dry conditions. Modest levels of contamination in grain observed at harvest in parts of western Europe, primarily in cooler, wetter areas, are apparent exceptions to this generalization (Gareis et al. 2001). The Provisional Maximum Tolerable Daily Intake (PMTDI) of the Joint Expert Committee on Food Additives and Contaminants of the World Health Organization/Food and Agricultural Organization (JECFA) for T-2/HT-2 toxin of 0.6 mg kg1 bw has a larger safety factor than would normally be indicated. This is primarily due to a lack of experimental data (Larsen et al. 2004). This review and comment will focus on three broad topics. First, a perspective will be offered on research on the Fusarium toxins, deoxynivalenol and fumonisin, in small grains (wheat, barley, oats) and maize. These crops comprise two thirds of cereal supply, which is currently in the order of 350 kg person1 year1 (Dyson 2001). The reason for this emphasis is that, for cereals contaminated by aflatoxin, including rice, the guidelines applying to international trade are clear. This is regardless of whether, for example. the difference between WHO and EU guidelines can be defended on a health basis (Wu 2004). In contrast, there are some uncertainties in the PMTDIs for deoxynivalenol and fumonisin relating to aspects of the mechanism and human health effects that might affect current trade limits. Second, factors that resulted in increased exposure to these toxins will be explored with suggestions about actions required to manage this change. Finally, the increased use of ensiled maize in north temperate dairy-producing areas (e.g. Quebec) will be examined in relation to uncertainties about toxins associated with this feed source. Toxins associated with Fusarium head blight and Gibberella ear rot Fusarium graminearum, F. culmorum and F. crookwellense are closely related species that produce deoxynivalenol (DON) or nivalenol and zearalenone, depending on the geographic origin of the isolate (Miller et al. 1991). These fungi cause Fusarium head blight in small grains and Gibberella ear rot in maize. These diseases are associated with temperate graingrowing regions. Which of the three species will

Mycotoxins in small grains and maize The use of susceptible wheat cultivars and maize hybrids is largely responsible for incidence of F. graminearum. Under epidemic conditions, agronomic practices have modest impact on disease (Miller 1994; Schaafsma et al. 2001; Hooker et al. 2005; Koch et al. 2006; Miller et al. 1998). As far as can be seen, only countries that enforce clear requirements, such as reductions in Fusarium head blight (including DON measurements) (Wilde et al. 2007), have been able to reduce toxin amounts in the harvested crop (Snijders 2004; see also Larsen et al. 2004). Red mold poisoning was reported in rural Japan coincident with an increase in wheat production from 1800. Major epidemics were recorded in Japan for the 1890, 1901, 1914, 1932, 1946, 1958, 1963 and 1970 crops, with human and animal toxicoses reported throughout (Yozhizawa 1983; Udagawa 1988). Japanese researchers and officials were sensitive to the possibility of toxic chemicals from mold-damaged food. The study of mycotoxins began in 1881 when a Japanese researcher showed that ethanol extracts of rice damaged by Penicillium citreonigrum were fatal to dogs, rabbits and guinea pigs. This led to a commercial ban on the sale of rice damaged by that fungus (Pitt 1991). Well-documented reports of human toxicosis from the consumption of Fusarium head blight-damaged wheat and barley are available. These describe the typical symptoms that consistently include nausea, vomiting and diarrhoea (Yozhizawa 1983; Udagawa 1988). Russian officials reported the same symptoms from humans consuming bread baked from scabby grain in 1923 (Prentice and Dickensen 1968). DON was isolated by Japanese researchers from grain that had made humans ill (Morooka et al. 1972). This toxin was responsible for a large-scale incident of human toxicosis in the Kashmir Valley of India in 1988 (Bhat et al. 1989; Medical Research Council of India, unpublished report). The same symptoms were seen in Indians consuming bread made from highly contaminated wheat. Acute human toxicoses have been reported in China, Japan and Korea, among other countries ( Yoshizawa 1983; Beardall and Miller 1994; KuiperGoodman 1994; Li et al. 1999). Fusarium head blight-damaged grain began to be a problem in the Midwest US and Canada coincident with the dominance of Marquis wheat during WW I. By the 1920s, large cultivar-screening programs were underway in Minnesota (Schroeder and Christensen 1963). In 1928, there was a massive epidemic in the mid west, where US scientists showed that damaged barley resulted in emesis in swine (Mundkur 1934). By 1941, a water extract of barley contaminated by a fungus described as G. saubentii [an invalid name that included F. graminearum], induced emesis in swine by gavage

221

(Hoyman 1941). Water extracts and then methanol extracts of maize and barley cultures of F. graminearum given intraperitoneally (i.p.) produced toxic signs in nursing mice and swine, and in swine by gavage, i.p. and intravenously (i.v.) by the mid 1960s (Vesonder and Hesseltine 1981). Using strains isolated from Fusarium head blight-affected cereals provided by W.L. Gordon (Agriculture Canada), Prentice et al. (1959) reported an emetic principle in organic solvent extracts from Fusarium cultures but were unable to determine the chemical structure (Prentice and Dickenson 1968). About the same time, while investigating Fusarium-damaged maize (described as F. culmorum and F graminearum . by Booth) resulting in cattle toxicosis. Australian researchers reported a toxic principle resulting in skin necrosis (Fisher et al. 1967). Finally, US researchers re-reported DON as vomitoxin from F. graminearum-contaminated maize in 1973 that had produced emesis in swine (Vesonder et al. 1973). Humans appear to be quite sensitive to DON (Bhat et al. 1989; Kuiper-Goodman 1994), but the available information does not permit a dose response to be reliably determined. The domestic animal most affected by DON is swine and, as noted, the use of the second trivial name for DON, vomitoxin, arose from the emetic effect in swine. The minimum oral dose required for emesis is in the order of 100 mg kg1 bw (Pestka et al. 1987). The emetic response in dogs appears to occur at a similar dose (Ueno 1983). However, DON seldom causes overt toxicity, including emesis, in swine because its presence in feed limits consumption. This anorexic effect typically results in decreased feed consumption and growth in swine at concentrations of more than 1 mg g1 in diets containing naturally contaminated grains. Trichothecenes in general, including DON, have a variety of immunological effects in laboratory animals at very low exposures. In experimental situations, this leads to increased susceptibility to bacterial, viral and fungal diseases with strong implications for human disease (Bondy and Pestka 2000; Pestka and Smolinski 2005). There is, therefore, a long and clear historic association between DON and animal disease. After consumption of grains affected by Fusarium head blight, similar symptoms in human have been consistently reported in many populations since the turn of the 19th century. There is no uncertainty that consumption of contaminated wheat results in DON exposure in humans (Turner et al. 2007). For the last 25 years, health authorities have acted to reduce human consumption of this toxin. Considering the available toxicology data, Health Canada established a tentative tolerable daily intake in 1982 of 3 mg kg1 bw per day and half that for infants

222

J. D. Miller changes in the membranes of more resistant types (Snijders and Kreching 1992; Cossette and Miller 1995; Miller and Ewen 1997). It is reasonable to speculate that modest effects on membranes associated with the emesis centre might be responsible for the neurotoxicity as the receptor structure would be altered and, hence, binding affinity. The ILSIEU meeting suggested that the establishment of an acute reference dose (ARfD) for DON would be valuable. It was also agreed that the ethical problems in doing a human study would be profound, indicating that perhaps a non-human primate study would be desirable (Larsen et al. 2004), in my opinion, would be less important than determining the mechanism of neurotoxicity. There are human clinical data available from the use of another trichothecene, DAS (also known as anguidine), as a chemotherapy agent in many studies. These studies have demonstrated that nausea and vomiting occurred in $50% of the patients at doses of between 200 and 400 mg kg1 bw i.v. (Bukowski et al. 1982; DeSimone et al. 1997). Considering the relative acute toxicities of DAS to DON, this would translate into an emetic dose for DON in adults of 4800 mg kg1 bw. This suggests that the minimum emetic dose for DON in swine is a reasonable approximation of the human equivalent. It was also suggested that studies of interactions between trichothecenes be performed (Larsen et al. 2004), which, in my opinion, would have little value. It is known that there are interactions between trichothecenes in model systems (Koshinshy and Khachatourians 1992) and in animals (Schiefer et al. 1986; Bhavanishankar et al. 1988), but their dimension is modest (<5) compared to the safety factors in the PMTI for DON. Fumonisins from Fusarium verticillioides and related species F. verticillioides and F. proliferatum are the most common fungi associated with maize. For many years, these fungi have been known to occur systemically in leaves, stems, roots and kernels and can be recovered from virtually all maize kernels worldwide, including healthy kernels. It is important to note that the plating of surface-disinfected kernels is an insensitive method compared to, for example, grinding the sample followed by dilution to extinction on semi-selective media, which largely detects actively growing mycelium in proportion to biomass. Detection is roughly proportional to the number of living cells in kernels and the fungus in the tip cap of the kernel is often not seen. The diseases resulting from Fusarium ear rot/ Fusarium kernel rot is associated with warm, dry years and insect damage, and is mainly caused by

(Kuiper-Goodman 1985). Based on a much expanded database, the JECFA established a PMTI that was slightly lower in 2001 (Canady et al. 2001). These factors (and others discussed in the following section) led to a series of recommendations on future research on DON at an ILSIEU meeting held in Dublin (Larsen et al. 2004). In relation to either increasing or decreasing the PMTDI, two issues were raised that deserve repeating. There is wide agreement that the mechanism causing neurotoxicity (emesis and feed refusal) needed to be determined. In the mid-1990s, a great deal of work was done to try and resolve this question for DON in swine. Dosing by a continuous-exposure osmotic pump, implanted intraperitoneally, resolved that the effects could not be due to taste or learned responses (Prelusky 1997). A single dose of 0.25 mg kg1 bw (i.v.) changed neurotransmitter concentrations in the hypothalamus, frontal cortex and cerebellum up to 8 days post-dosing. Norepinephrine increased in all three tissues, whereas dopamine was decreased. In contrast, serotonin increased and then decreased in the hypothalamus, it was decreased in the frontal cortex and no change was observed in the cerebellum (Prelusky et al. 1992). A lower dose (10 mg kg1 bw i.v.) resulted in changes in cerebral spinal fluid neurotransmitters (Prelusky 1993). Serotonin-receptor antagonists prevented DON-induced vomiting, while 5HT2-receptor antagonists were moderately effective in high doses. Other anticholinergic actives were also effective but by acting directly at the emetic centre preventing emesis regardless of the cause (Prelusky et al. 1992). This suggested that, although there is no doubt that the emetic centre is responsible for vomiting, the effect seems indirect. In plants, it has long been known that trichothecenes disrupt membranes, causing physical damage to membranes resulting in cell lysis. Red blood cells are a compartment for trichothecene metabolism and these cells will lyse in the presence of excess circulating toxin. The amount of toxin required to lyse red blood cells varies according to animal species. The reason is that the proportions of sphingomyelin, phosphatidylcholine, PPTethanolamine and PPTserine in the inner and outer membranes of erthythrocytes vary between animal species. The concentrations of the trichothecene T-2 toxin that cause membrane deformation in human blood cells are comparable to those that affect protein synthesis in HeLa cells (Cundliffe et al. 1974; Gyonhyossy-Issa et al. 1986; Khachatourians 1990). It is known that, in wheat and maize, some genotypes have greater resistance to the membrane-damaging effects of DON. In maize, this is due to differences in the binding affinity of DON, implying that there are unknown functional

Mycotoxins in small grains and maize F verticillioides (G. fujikuroi) and F. proliferatum. . In warmer corn-growing areas, F verticillioides is one . of the most important ear diseases (Miller 2001). F proliferatum (which also produces moniliformin) . becomes dominant under different environmental conditions than F. verticillioides (De La Campa et al. 2005). . Below 2528 C, F graminearum grows well, with growth virtually ceasing above that temperature, and in that range, assuming that there is sufficient rain, this fungus out-competes F. verticillioides. Many studies on fumonisin from natural occurrence and experimental infections have demonstrated the importance of drought rather than temperature stress. F. verticillioides grows well at temperatures above 28 C (Reid et al. 1999) and there is evidence that fumonisin can only accumulate in stressed or senescing kernel tissue (Reid et al. 1999; Miller 2001). This is consistent with considerable field data; for example, in a US study, fumonisin concentrations were inversely proportional to June rainfall (Shelby et al. 1994). In the cool corngrowing area of southern Ontario, accumulation was limited to drought-stressed fields. Comparing three counties with similar temperatures, the three with the highest average FB1 concentrations (1.4 mg g1) had half the rainfall of the counties with the lowest average FB1 (0.4 mg g1; Miller et al. 1995). Since drought stress results in greater insect herbivory on maize, it is not possible to totally separate these variables from other complications (Miller 2001). However, there is a strong consistent relationship between insect damage and Fusarium ear rot. Within a year or two of the availability of fumonisin analytical standards, a field survey demonstrated that the incidence of the European corn borer increased Fusarium kernel rot and fumonisin concentrations (Lew et al. 1991). Maize genotypes containing the anti-insectan Bt protein have reduced amounts of fumonisin compared to non-Bt genotypes (Bakan et al. 2002; Hammond et al. 2004; De La Campa et al. 2005). De La Campa et al. 2005) were able to integrate this information in a study of factors that affected fumonisin accumulation in maize. Insect damage and weather variables in four periods around silking explained most of the variation in fumonisin concentrations at harvest. The first critical period for fumonisin accumulation was 410 days before silking when temperatures of <15 and 434 C (permissive temperatures) reduced fumonisin. Within permissive temperatures, some rainfall increased fumonisin. In the week following the silking period, again within permissive temperatures, some rainfall increased fumonisin. Thereafter, fumonisin was increased by slight drought stress.

223

Since the discovery of fumonisin in 1988, a great deal has been learned about its effects. Consumption of maize contaminated with fumonisin has a number of toxic effects on domestic animals, including equine leucoencephalomalacia in horses (ELEM), and pulmonary edema and immunosuppression in swine. The toxin is carcinogenic in rodents. The mechanism for all these phenomena is directly or indirectly due to the effects of fumonisin on sphingolipid biosynthesis; this work has been reviewed extensively (IPCS 2000; Bolger et al. 2001; SCF 2003). JECFA established a PMTI using the renal toxicity of fumonisin as the endpoint. (Voss et al. 1995; NTP 2001). Fusarium kernel rot was associated with animal disease in the US midwest in 1904 and there were large epidemics of ELEM in the US during the drought years of the 1930s. In 1971, corn contaminated by the fungus, now called F. verticilliodes, was shown to cause ELEM (IPCS 2000). A South African group studying elevated esophageal cancer in the Transkei and a French group working on ELEM independently described fumonisin as the cause of disease in 1988 (Marasas 2001) and then in 1989 (as macrofusin; Laurent et al. 1989). There has also been the association of regular consumption of large amounts of maize-based foods, regularly infected with F. verticilliodes, with esophageal cancer in South Africa and northern Italy (IARC 1993, 2002; IPCS 2000). South Africa has been growing maize at least since the 17th century and it is now grown across Africa (Desjardins and McCarthy 2004; McCann 2005). In Latin America, food is prepared primarily from tortilla flours prepared by heating with base which reduces fumonsin concentrations. However, in Africa, fumonisin is not affected by traditional methods of cooking (De La Campa et al. 2004; Shephard et al. 2002; Fandohan et al. 2005), but sorting does effectively reduce fumonisin concentrations (Desjardins et al. 2000; Riley and Miller 2003). The earliest reports of esophageal cancer in rural black populations (studies from 19551969; Rose 1973 and references cited therein) noted the extraordinarily high rates of this cancer in the Transkei. This was striking compared to other parts of the world and other parts of Africa (Day, 1975). Since no biomarkers are available, it has proven impossible, so far, to establish fumonisin as a causative factor in this pattern of esophageal cancer. In the last IARC review (IARC 2002), fumonisin remained in group 2B (possible human carcinogen) and no change in classification was made despite the fact that, since the 1992 review, fumonisin has been shown to be a rodent carcinogen. This would change should positive evidence be found (currently this is sphingolipid ratio change only, concurrent with

224

J. D. Miller summer or is water-limited. Since 1961, China has increased cereal production 5-fold from 100 to 400 kg person1, with maize and wheat increasing roughly in proportion. Maize has been grown in China since the 16th century (Desjardins and McCarthy 2004) and currently has a much larger production than wheat. For food, wheat is nearly equal to rice, a well-established food crop since the 6th century (Myer 1978). The ratio of rice to wheat maize production has changed from 1.2:1 to 0.8:1 (Tong et al. 2003). Fusarium head blight epidemics have been greatly increasing in frequency in recent years (Chen et al. 2001) and, as noted above, there is exposure to DON in China from both wheat and maize, although it is not well documented (Canady et al. 2001; Meky et al. 2003). The situation in Africa is much different. During the period 19602003, cereal production increased 2.5-fold (half that of China) but, approximately over the same period, declined on a per capita basis from 150 to 125 kg person1 (Dyson 2001). As with China, there has been a modest change in the ratio of maize production to that of the other staple crops (sorghum, millet, rice). However, of these, maize is uniquely susceptible to fumonisin, DON and zearalenone and co-exposures with aflatoxin are certainly common (Doko et al. 1996; Ngoko et al. 2001) Africa has become extremely vulnerable to exposure from mycotoxins found in maize (Riley and Miller 2003; Azziz-Baumgartner et al. 2005). China is producing sufficient food for residents of rural areas to purchase food (Gale et al. 2005), which means that diets are much more diverse in China compared to Africa. In regions where weather conditions result in (more) severe mycotoxin problems, very high exposures are inevitable with the potential for acute toxicoses (Riley and Miller 2003; Azziz-Baumgartner et al. 2006). While this has long been known for aflatoxin in Africa, the PMTDI for fumonisin, as noted, is exceeded in Africa, with the upper 10th percentile of the population being approximately three times that of the PMTDI ( JECFA 2001). In areas where the occurrence of fumonisin is chronic, this materially understates the situation. Shephard et al. (2007) estimated fumonisin exposure in some areas of rural South Africa at 219 times the PMTDI and exposure in rural Bukino Faso was found to be 1260 times the PMTDI (Nikiema et al. 2004). Against this broad background, in both the fully developed market economies and due to the limited diversity of the food supply in developing countries, increased climate variability will produce more frequent epidemics of Fusarium head blight and Gibberella ear rot. As described above, the former requires rain at anthesis or silk emergence and warm

demonstrated dietary exposure) and, if reliably demonstrated, the IARC classification would change from 2B to 2A (probable human carcinogen), which might require a re-evaluation of the PMTDI. After the setting of the JECFA TDI, it was found that fumonisin causes neural tube birth defects (NTDs) in mouse somites (Sadler et al. 2002) and a rodent model in vivo (Gelineau-van Waes et al. 2005). These studies arose from a transient increase in NTDs from 10 to 27 per 10,000 live births in Mexican-Americans in Cameron County Texas (Hendricks 1999; Marasas et al. 2004). A followup study found that increased NTD risk was associated with fumonisins exposure (Missmer et al. 2006) and, for a number of reasons, animal models had failed to predict this possibility (IPCS 2000). The mechanism relates to material exposure to fumonisin prior to the formation of the placenta. Fumonisin affects folate transport, which results in lowered folate in the embryo (Sadler et al. 2002; Marasas et al. 2004). A study of tortilla production in Cameron County revealed that some preparation methods in local facilities left intact fumonisin in the final product (De La Campa et al. 2004) and NTDs are very high in fumonisin endemic areas (Marasas et al. 2004). At the time of writing, there is no published study of NTDs in a regulatory strain of rodent [The strain used in the Gelineau-van Waes (2005) study is specialized for NTD research]. Another factor that might result in the re-evaluation of the JECFA TDI would be the results of a well-designed study in a regulatory strain of fumonisin. Exposure to maize and wheat borne toxins is increasing The existence of a widely accepted JECFA PMTI for DON and fumonisin are major achievements; however, in the recent past, exposure of young children has been close to the PMTDI in the Netherlands (Pieters et al. 2004), Denmark (Rasmussen et al. 2007) and Canada (KuiperGoodman et al. 2008). The PMTDI is exceeded in other countries dramatically so in Africa and in parts of Latin America (JECFA 2001) and there is no doubt that this would be more dramatic if exposures were calculated for wheat-consuming population in endemic areas lacking a diverse source of cereals, as opposed to the standard GEMS diet. The situation for fumonisin exposure is similar, except much worse in parts of Africa (Bolger et al. 2001; Shephard et al. 2005, 2007). As noted above, large areas of arable land have come under wheat and maize production in China since 1961 (Dyson 2001; Tong et al. 2003). Most arable land is in areas prone to humidity in the

Mycotoxins in small grains and maize conditions while Fusarium kernel rot requires dry conditions but permissive temperatures. At the end of 2006, the number of weather-related disasters in Canada has increased 10-fold since 1900 and 4-fold since 1960, most of which are related to heavy rainfall (http://www.ec.gc.ca/TKEI/graphs/ w_disasters_095_e.xls). These disasters have been associated with the flooding of rivers, caused by prolonged rain during rapid snowmelt, of drainage pathways, primarily caused by short-duration, intensive rainfall from thunderstorms or the residue of hurricanes coming up from the US southeast. These phenomena have also been felt in Europe (Ekstrom et al. 2005; Lehner et al. 2006; Wilson 2007). In the principal Canadian maize-production area (Ontario), each of the last five summers has been hotter than the previous 30, on average, which creates one of the conditions for increased risk of fumonisin accumulation. The other condition is drought (Miller et al. 1995; Miller 2001), which is also predicted to occur more often in most of the corn regions over the coming decades (Lehner et al. 2006). An increased prevalence of extreme weather events is now anticipated worldwide over the next century (Zhang et al. 2007). Riley and Miller (2003) argued for increased use of forecasting methods to predict mycotoxins on a countywide-scale. There is a long history of the use of models to predict crop diseases, including Fusarium head blight (De Wolf et al. 2003; Del Ponte et al. 2005; Carranza et al. 2007) however, there are comparatively few reports on models predicting the potential for mycotoxins in field crops the most useful being developed Schaafsma and colleagues (Hooker et al. 2002; Schaafsma et al. 2006; Schaafsma and Hooker 2008). These models need to be developed against a large background dataset of DON and weather within a particular area as the relationship between disease symptoms and toxin accumulation is cultivar-specific (Miller et al. 1984; Paul et al. 2006). Such work has also been attempted for Gibberella ear rot (Mansfield et al. 2006). Although there are some recent studies on models for predicting DON in wheat (Eiblmeier 2006; Forrer et al. 2006), only one predictive model for DON, DONcast, has been published and commercialized (Hooker et al. 2002). This model, which adapted to Uruguay (Schaafsma et al. 2006) and French conditions (Schaafsma and Hooker 2008), allows decision-makers to implement changes in agronomic (fungicide application) or harvesting practices, to be aware where emerging DON problems exist and to take the necessary management steps. These might involve diverting the harvest from the affected fields away from human food-use to alternative uses, including use in

225

DON-tolerant domestic animals (cattle) after appropriate dilution. For DON and fumonisin in maize, Hooker and Schaafsma (2005) and, in greater detail for fumonisin in maize, De La Campa et al. (2006) have demonstrated that such models are feasible. While these models are meant for areas with on-line meteorological data and information on emergent insect populations, some modeling might be feasible from remote-sensing information. Rainfall timing, water stress and permissive temperatures are the key factors for DON and fumonisin accumulation. Modeling of drought and vegetation indices are a component of the Famine Early Warning System, which assesses remotely sensed data, ground-based sources and other factors affecting local food availability (http://www.fews. net/). These data could form the basis of models that might be developed for DON, fumonisin and possibly aflatoxin, adding an important early warning capacity to managing contaminated crops. Potentially, toxin-predictive modelling is an important research direction for both vulnerable populations and due to increased climate variability in countries with commercial agriculture. Uncertainties associated with the increased use of short season maize hybrids and silage In eastern Canada, the use of maize silage in dairy production has increased approximately 5-fold over the past 25 years. This is mainly attributed to the availability of short season maize hybrids suitable for both eastern North America and parts of western Europe. In addition, as long as there is adequate protein available, maize silage has a high starch content and is a useful high-energy feed for cows (e.g. Dawo et al. 2007). Silage production has remained fairly stable in most of Europe in recent decades but there has been a shift towards maize silage in some countries (e.g. Denmark, The Netherlands; Wilkinson and Toivonen 2003). In recent years, there has been increased recognition that silage is quite frequently contaminated by toxins, mainly from Penicillium roqueforti (Seglar et al. 1997; Auerbach et al. 1998; Seglar 1999). This appears to be due to a combination of changes in feed production technology (e.g. increased reliance on maize fodder instead of transporting grain maize from the Midwest US or southern Ontario) and/or increased farm sizes (making it more likely that small changes in herd health will be noticed). At first, this was, and often still is, attributed to the presence of low levels of Fusarium toxins in maize. Since these do not affect cows or cattle, this is not the explanation. In contrast, a number of toxic phenomena in cows, associated silage contaminated by P. roqueforti group, have been observed. Severe toxicoses in cows,

226

J. D. Miller Acknowledgements I thank colleagues at Agriculture Canada and Carleton University, Clive James, Maya Pineiro, Art Schaafsma, David Hooker, Kristian Nielsen, Jens Frisvad, as well as the Natural Sciences & Engineering Research Council of Canada and TUBITAK, for financial support. I thank John Gilbert and Hamide Z. Senyuva for inviting me to discuss this topic.

associated with the latter fungi growing on silage, were first reported from Japan and the US in the 1960s ( Wei et al. 1973; Omomo et al. 1994). This issue remained unresolved and largely ceased to be a practical problem. More recently, P. roquefori sensu lato has been associated with reports of two syndromes in cows: serious toxicoses associated with P. roqueforti and a general ill-thrift associated with P. paneum (Sumarah et al. 2005). Both of these fungi produce roquefortine but the former produces PR toxins and the later festuclavine, a compound long associated with ill-thrift in cows (Nielsen et al. 2006; OBrien et al. 2006). Good silage practice eliminates these fungi (OBrien et al. 2007); hence, the problem can be managed. However, a careful US study reported in 1997 of dairy herds in Florida, Vermont and Wisconsin suffering from ill-thrift, silage was more likely to be contaminated by PR toxin (Seglar et al. 1997). Similarly, a German study reported widespread contamination by roquefortine (Auerbach et al. 1998), which is proxy for contamination either by P. roqueforti or P. paneum. From a public-health perspective, it remains unresolved whether these compounds occur in milk. This is regarded as needing special care owing to its importance in the diets of infants and children. Tuller et al. (1998) demonstrated that sheep fed $1.25 mg kg1 bw day1 roquefortine (recalculated from the authors data) resulted in the compound being detected in liver (1.15 mg kg1), bile (0.12 mg kg1), kidney (0.15 mg kg1) and trace amounts in muscle and fat. This is a high exposure but emphasizes the point that nothing is apparently known about the fate and distribution of roquefortine, PR toxin or festulcavine in milk a question that needs resolution.

References
Aoki T, ODonnell K. 1999. Morphological characterization of Gibberella coronicola sp. nov., obtained through mating experiments of Fusarium pseudograminearum. Mycoscience 40:443453. Auerbach H, Oldenburg E, Weissbach F. 1998. Incidence of Penicillium roqueforti and roquefortine C in silages. Journal of the Science Food Agriculture 76:565572. Azziz-Baumgartner E, Lindblade K, Gieseker K, Rogers HS, Kieszak S, Njapau H, Schleicher R, McCoy LF, Misore A, DeCock K, et al.. 2006. Case-control study of an acute aflatoxicosis outbreak, Kenya, 2004. Environmental Health Perspectives 113:17791783. Bakan B, Melcion D, Richard-Molard D, Cahagnier B. 2002. Fungal growth and Fusarium mycotoxin content in isogenic traditional maize and genetically modified maize grown in France and Spain. Journal of Agriculture Food Chemistry 50:728731. Beardall JA, Miller JD. 1994. Human disease in which mycotoxins have been suggested as among the causal factors. In: Miller JD, Trenholm HL, editors. Mycotoxins in grain: compounds other than aflatoxin. Mycotoxins in grain: compounds other than aflatoxin. St. Paul, MN: Egan Press. pp 487540. Bhat RV, Beedu SR, Ramakrishna Y, Munshi KL. 1989. Outbreak of trichothecene mycotoxicosis associated with consumption of mould-damaged wheat in Kashmir Valley, India. Lancet 7:3537. Bhavanishanka TN, Ramesh HP, Shantha T. 1988. Dermal toxicity of Fusarium toxins in combinations. Archives Toxicology 61:241244. Bolger M, Coker RD, DiNovi M, Gaylor D, Gelderblom W, Olsen M, Paster N, Riley RT, Shephard G, Speijers GJA. 2001. Fumonisins. WHO/IPCS safety evaluation of certain mycotoxins in food. WHO Food Additives Series 47:557680. Bondy GS, Pestka JJ. 2000. Immunomodulation by fungal toxins. Journal of Toxicology Environmental Health B Critical Reviews 3:109143. Bukowski R, Vaughn C, Bottomley R, Chen T. 1982. Southwest Oncology Group phase II study of anguidine in 134 patients with gastrointestinal malignancies. Cancer Treatment Reports 66:381383. Canady RA, Coker RD, Egan SK, Krska R, Kuiper-Goodman T, Olsen M, Pestka J, Resnik S, Schlatter J. 2001. Deoxynivalenol. WHO/IPCS safety evaluation of certain. mycotoxins in food. WHO Food Additives Series 47:419555. Carranza MR, Moschini RC, Kraan G, Bariffi JH. 2007. Examination of meteorology-based predictions of Fusarium head blight of wheat grown at two locations in the southern Pampas region of Argentina. Australasian Plant Pathology 36:305306.

Summary All the issues raised in this review regarding toxins have plagued agriculture over the past century and perhaps longer (Matossian 1989). The challenges for the current generation of researchers relates to setting appropriate regulatory limits that improve public health, especially for children. In addition, greater preparedness is needed to manage changes in climate and agricultural technologies which may increase the occurrence of mycotoxins. As some governments adopt increasingly restrictive regulatory limits, officials need to be prepared for supply restrictions leading to price increases and excursions over the TDIs. In developing countries, food sufficiency remains a problem, but risk of acute toxicoses due to certain mycotoxins appears to be increasing.

Mycotoxins in small grains and maize

Chen LF, Bai G-H, Desjardins AE. 2001. Recent advances in wheat head scab research in China. Available: http://www.nal. usda.gov/pgdic/WHS/whsindex.html. Accessed 25 July 2007. Clear RM, Patrick SK, Gaba D, Roscoe M, Turkington TK, Demeke T, Pouleur S, Couture L, Ward TJ, ODonnell K. 2006. Trichothecene and zearalenone production in culture by isolates of Fusarium pseudograminearum from western Canada. Canadian Journal of Plant Pathology 28:131136. Cossette F, Miller JD. 1995. Phytotoxic effect of deoxynivalenol and Gibberella ear rot resistance of corn. Natural Toxins 3:383388. Cundliffe E, Cannon M, Davies J. 1974. Mechanism of inhibition of eukaryotic protein synthesis by trichothecene fungal toxins. Proceedings of the National Academy of Sciences of the United States of America 71:3034. Dawo MI, Wilkinson JM, Sanders FET, Pilbeam DJ. 2007. The yield and quality of fresh and ensiled plant material from intercropped maize (Zea mays) and beans (Phaseolus vulgaris). Journal Science of the Food and Agriculture 87:13911399. Day NE. 1975. Some aspects of the epidemiology of esophageal cancer. Cancer Research 35:33043307. De La Campa R, Miller JD, Hendricks K. 2004. Fumonisin in tortillas produced under small-scale facilities and the effect of traditional variations on masa production on this toxin. Journal of Agriculture Food Chemistry 52:44324437. De La Campa R, Hooker DC, Miller JD, Schaafsma WA, Hammond BG. 2005. Modelling effects of environment, insect damage and BT genotypes on fumonisin accumulation in maize in Argentina and the Philippines. Mycopathologia 159:539552. Del Ponte EM, Fernandes JMC, Willingthon P. 2005. A risk infection simulation model for Fusarium head blight of wheat. Fitopatologia Brasileira 30:634642. DeSimone PA, Greco FA, Lessner HF. 1979. Phase I evaluation of a weekly schedule of anguidine. Phase I evaluation of a weekly schedule of anguidine. Cancer Treatment Reports 63:20152017. Desjardins AE, Manandhar G, Plattner RD, Maragos CM, Shrestha M, McCormick SP. 2000. Occurrence of Fusarium species and mycotoxins in Nepalese maize and wheat and the effect of traditional processing methods on mycotoxin levels. Journal of Agriculture Food Chemistry 48:13771383. Desjardins AE, McCarthy SA. 2004. Milho, makka, and yu mai: early journeys of Zea mays to Asia. Available: http://www. nal.usda.gov/research/maize. Accessed 25 July 2007. De Wolf ED. 2003. Risk assessment models for Fusarium head blight epidemics based on within-season weather data. Phytopathology 93:428435. Doko MB, Canet C, Brown N, Sydenham EW, Mpuchane S, Siame BA. 1996. Natural Co-occurrence of fumonisins and zearalenone in cereals and cereal-based foods from Eastern and Southern Africa. Journal of Agriculture Food Chemistry 44:32403243. Dyson T. 2001. World food trends: a neo-malthusian prospect? Proceedings of the American Philosophical Society 145:438455. Ekstrom M, Fowler HJ, Kilsby CG, Jones PD. 2005. New estimates of future changes in extreme rainfall across the UK using regional climate model integrations. 2. Future estimates and use in impact studies. Journal of Hydrology 300:234251. Fandohan P, Zoumenou D, Hounhouigan DJ, Marasas WFO, Wingfield MJ, Hell K. 2005. Fate of aflatoxins and fumonisins during the processing of maize into food products in Benin. International Journal of Food Microbiology 98:249259. Fisher EE, Kellock AM, Wellington NAM. 1967. Toxic strain of Fusarium culmorum (WG Smith) Sacc. isolated from Zea mays L. associated with sickness in dairy cattle. Nature 315:322.

227

Gareis M, Zimmermann C, Schothorst R, Paulsch W, Vidnes A, Bergsten C. Paulsen B, Brera C, Miraglia M, Grossi S, Debegnach F. 2001. Collection of occurrence data of Fusarium toxins in food and assessment of dietary intake by the population of EU member states. Available: http:// europa.eu.int/comm/food/fs/scoop/task3210.pdf. Accessed 25 July 2007. Gelineau-van Waes J, Starr L, Maddox J, Aleman F, Voss KA, Wilberding J, Riley RT. 2005. Maternal fumonisin exposure and risk for neural tube defects: mechanisms in an in vivo mouse model. Birth defects. Res Clin Mol Teratol 73:487497. Goswami RS, Kistler HC. 2004. Pathogen profile heading for disaster: Fusarium graminearum. Molecular Plant Pathology 5:515525. Gyonggyossy-Issa MIC, Card RT, Fergusson DJ, Khachatourians GC. 1986. Prehaemolytic erythrocyte deformability changes caused by trichothecene T-2 toxin. Blood Cells 11:393403. Hammond BG, Campbell K, Pilcher C, DeGooyer T, Robinson A, McMillen B, Spangler S, Riordan S, Rice L, Richard J. 2004. Lower fumonisin mycotoxin levels in the grain of Bt corn grown in the United States in 20002002. Journal of Agriculture Food Chemistry 52:13901397. Hooker DC, Schaafsma AW, Tamburic-Ilincic L. 2002. Using weather variables pre- and postheading to predict deoxynivalenol in winter wheat. Plant Disease 86:611619. Hooker DC, Schaafsma AW. 2005. Agronomic and environmental impacts concentrations of deoxynivalenol and fumonisin B1 in corn across Ontario. Canadian Journal of Plant Pathology 27:347356. Hoyman WG. 1941. Concentration and characterization of the emetic principle present in barley infected with Gibberella saubenetii. Phytopathology 31:871885. IARC. 1993. Monographs on the evaluation of carcinogenic risks to humans. Some naturally occurring substances: food items and constituents, heterocyclic aromatic amines and mycotoxins Vol. 56. Lyon: World Health Organisation/IARC Press. IARC. 2003. Monographs on the evaluation of carcinogenic risks to humans. Some traditional herbal medicines, some Mycotoxins, naphthalene and Styrene Vol. 82. Lyon: World Health Organisation/IARC Press. IPCS. 2000. Fumonisin B1. Environmental health criteria Vol. 219. Geneva: International Programme on Chemical Safety, WHO. pp 4350. Jennings P, Coate ME, Walsh K, Turner JA, Nicholson P. 2004. Determination of deoxynivalenol- and nivalenol-producing chemotypes of Fusarium graminearum isolated from wheat crops in England and Wales. Plant Pathology 53:643652. Joffe AZ, Yagen B. 1977. Comparative study of the yield of T-2 toxic produced by Fusarium poae, F sporotrichioides and . F. sporotrichioides var. tricinctum strains from different sources. Mycopathologia 60:9397. Khachatourians GC. 1990. Metabolic effects of the trichothecene T-2 toxin. Canadian Journal of Physiology Pharmacology 68:10041008. Koch H-J, Pringas C, Maerlaender B. 2006. Evaluation of environmental and management effects on Fusarium head blight infection and deoxynivalenol concentration in the grain of winter wheat. European Journal of Agronomy 24:357366. Koshinsky HA, Khachatourians GC. 1992. Trichothecene synergism, additivity and antagonism: the significance of the maximally quiescent ratio. Natural Toxins 1:4847. Kuiper-Goodman T. 1985. Potential human health hazards and regulatory aspects. In: Scott PM, Trenholm HL, Sutton MD, editors. Mycotoxins: a Canadian perspective. NRCC Publication 22848. Ottawa: NRCC. ISSN 0316-0114. pp. 103110.

228

J. D. Miller
Miller JD, Culley J, Fraser K, Hubbard S, Meloche F, Ouellet T, Seaman L, Seifert KA, Turkington K, Voldeng H. 1998. Effect of tillage practice on Fusarium head blight of wheat. Canadian Journal of Plant Pathology 20:95103. Miller JD, Greenhalgh R, Wang YZ, Lu M. 1991. Mycotoxin chemotypes of three Fusarium species. Mycologia 83:121130. Miller JD, Savard ME, Schaafsma AW, Seifert KA, Reid LA. 1995. Mycotoxin production by Fusarium moniliforme and Fusarium proliferatum from Ontario and occurrence of fumonisin in the 1993 corn crop. Canadian Journal of Plant Pathologyogy 17:233239. Miller JD, Young JC, Sampson DR. 1984. Deoxynivalenol and Fusarium head blight resistance in spring cereals. Phytopathol Zeitschrift 113:359367. Mirocha CJ. 1984. Mycotoxicoses associated with Fusarium. In: Moss MO, Smith JE, editors. The applied mycology of Fusarium. The applied mycology of fusarium. Cambridge, UK: Cambridge University Press. pp 141155. Missmer SA, Suarez L, Felkner M, Wang E, Merrill Jr AH, Rothman KJ, Hendricks KA. 2006. Exposure to fumonisins and the occurrence of neural tube defects along the Texas-Mexico border. Environmental Health Perspectives 114:237241. Morooka N, Uratsuji N, Yoshizawa T, Yamamoto H. 1972. Studies on the toxic substances in barley infected with Fusarium species. Journal of Food Hygiene Society Japan 13:368375. Mundkur BB. 1934. Some preliminary feeding experiments with scabby barley. Phytopathology 24:12371243. Myer R. 1978. Wheat in China: past, present and future. The China Quarterly 74:297333. Ngoko Z, Marasas WFO, Rheeder JP, Shephard GS, Wingfield MJ, Cardwell KF. 2001. Fungal infection and mycotoxin contamination of maize in the humid forest and the Western Highlands of Cameroon. Phytoparasitica 29:352360. Nielsen KF, Sumarah MW, Frisvad JC, Miller JD. 2006. Production of metabolites by species in the Penicillium roqueforti complex. Journal of Agriculture Food Chemistry 54:37563763. Nikiema PN, Worrillow L, Traore AS, Wild CP, Turner PC. 2004. Fumonisin contamination of maize in Burkina Faso, West Africa. Food Additives and Contaminants 21:865870. NTP. 2001. TR-496 Toxicology and Carcinogenesis Studies of Fumonisin B1 (CAS No. 116355-83-0) in F344/N Rats and B6C3F1 Mice (Feed Studies). US National Toxicology Program: Research Triangle Park, NC. OBrien M, Nielsen KF, OKiely P, Forristal PD, Fuller HT, Frisvad JC. 2006. Mycotoxins and other secondary metabolites produced in vitro by Penicillium paneum Frisvad and Penicillium roqueforti Thom isolated from baled grass silage in Ireland. Journal of Agriculture Food Chemistry 54:92689276. OBrien M, OKiely P, Forristal PD, Fuller HT. 2007. Quantification and identification of fungal propagules in wellmanaged baled grass silage and in normal on-farm produced bales. Animal Feed Science and Technology 132:283297. ODonnell K, Kistler HK, Tacke HK, Casper HK. 2000. Gene genealogies reveal global phylogeographic structure and reproductive isolation among lineages of Fusarium graminearum, the fungus causing wheat scab. Proceedings of the National Academy of Sciences of the United States of America 97:79057910. Ohmomo S, Kitamoto HK, Nakajima T. 1994. Detection of roquefortines in Penicillium roqueforti isolated from molded maize silage. Journal Science of the Food and Agriculture 64:211215.

Kuiper-Goodman T. 1994. Prevention of human mycotoxicosis through risk assessment and risk management. In: Miller JD, Trenholm HL, editors. Mycotoxins in grain. Mycotoxins in grain. St. Paul, MN: Eagan Press. pp 439470. Kuiper-Goodman T, Hilts C, Hayward S, Richard I, Billiard S. 2008. Probabilistic exposure assessment and risk analysis of ochratoxin and deoxynivalenol for all age groups and impact of MLs. In preparation. Larsen JC, Hunt J, Perrin I, Ruckenbauer P. 2004. Workshop on trichothecenes with a focus on DON: summary report. Toxicology Letters 153:122. Laurent DN, Platzer N, Kohler F, Sauvant MP, Pellegrin F. 1989. Macrofusin and micromonilin: two new mycotoxins isolated from corn infected by Fusarium moniliforme. Microbiologie Aliment Nutrition 7:916. Lehner B, Doll P, Alcamo J, Henrichs T, Kaspar F. 2006. Estimating the impact of global change on flood and drought risks in Europe: a continental integrated analysis. Climatic Change 75:273299. Lew H, Adler A, Edinger W. 1991. Moniliformin and the European corn borer (Ostrinia nubilalis). Mycotoxin Research 7:7176. Li F, Luo X, Yoshizawa T. 1999. Mycotoxins (trichothecenes, zearalenone and fumonisins) in cereals associated with human redmold intoxications stored since 1989 and 1991 in China. Natural Toxins 7:9397. Mansfield MA, De Wolf ED, Kuldau GA. 2005. Relationships between weather conditions, agronomic practices, and fermentation characteristics with deoxynivalenol content in fresh and ensiled maize. Plant Disease 89:1151115. Marasas WFO. 2001. Discovery and occurrence of the fumonisins: a historical perspective. Environ Health Perspectives 109 S2:23943. Marasas WFO, Riley RT, Hendricks KA, Stevens VL, Sadler TW, Gelineau-van Waes J, Missmer SA, Cabrera J, Torres O, Gelderblom WCA, et al.. 2004. Fumonisins disrupt sphingolipid metabolism, folate transport, and neural tube development in embryo culture and in vivo: a potential risk factor for human neural tube defects among populations consuming fumonisin-contaminated maize. Journal of Nutrition 134:711716. Matossian MK. 1989. Poisons of the past: molds, epidemics, and history. Yale University Press. McCann JC. 2005. Maize and grace: Africas encounter with a new world crop, 15002000. Harvard University Press. Meky FA, Turner PC, Ashcroft AE, Miller JD, Qiao Y-L, Roth MJ, Wild CP. 2003. Development of a urinary biomarker of human exposure to deoxynivalenol. Food Chemical Toxicology 41:265273. Mesterhazy A. 2003. Breeding wheat for Fusarium head blight resistance in Europe. In: Leonard KJ, Busch RH, editors. Fusarium head blight of wheat and barley. Fusarium head blight of wheat and barley. St. Paul, MN: American Phytopathological Society. pp 211240. Miller JD. 1994. Epidemiology of Fusarium graminearum diseases of wheat and corn. In: Miller JD, Trenholm HL, editors. Mycotoxins in grain: compounds other than aflatoxin. Mycotoxins in grain: compounds other than aflatoxin. St. Paul, MN: Egan Press. pp 1936. Miller JD. 1995. Fungi and mycotoxins in grain: implications for stored product research. Journal of Stored Product Research 31:16. Miller JD. 2001. Factors that affect the occurrence of fumonisin. Environmental Health Perspectives 109 S2:321324. Miller JD, Ewen MA. 1997. Toxic effects of deoxynivalenol on ribosomes and tissues of the spring wheat cultivars Frontana and Casavant. Natural Toxins 5:234237.

Mycotoxins in small grains and maize

Paul PA, Lipps PE, Madden LV. 2006. Meta-analysis of regression coefficients for the relationship between fusarium head blight and deoxynivalenol content of wheat. Phytopathology 96:951961. Pestka JJ, Lin WS, Miller ER. 1987. Emetic activity of the trichothecene 15-acetyldeoxynivalenol in swine. Food and Chemical Toxicology 25:855858. Pestka JJ, Smolinski AT. 2005. Deoxynivalenol: toxicology and potential effects on humans. Journal of Toxicology Environmental Health B Critical Reviews 8:3969. Pieters MN, Bakker M, Slob W. 2004. Reduced intake of deoxynivalenol in The Netherlands: a risk assessment update. Toxicology Letters 153:145153. Pitt JI. 1991. Penicillium toxins. ACIAR Proc 36:99103. Prelusky DB. 1993. The effect of low-level deoyxnivalenol on neurotransmitter levels measured in pig cerebral spinal fluid. Journal of Environmental Science Health B 26:731761. Prelusky DB. 1997. Effect of intraperitoneal infusion of deoxynivalenol on feed consumption and weight gain in the pig. Natural Toxins 5:121125. Prelusky DB, Yeung JM, Thompson BK, Trenholm HL. 1992. Effect of deoxynivalenol on neurotransmitters in discrete regions of swine brain. Archives Environmental Contamination Toxicology 22:3640. Prentice N, Dickson AD. 1968. Emetic material associated with Fusarium species in cereal grains and artificial media. Biotechnology and Bioengineering 10:413427. Rasmussen PH, Petersen A, Ghorbani F. 2007. Annual variation of deoxynivalenol in Danish wheat flour 19982003 and estimated daily intake by the Danish population. Food Additives and Contaminants 24:315325. Riley RT, Miller JD. 2003. Case study: risk analysis of fumonisin. Food safety risk analysis: a manual. Rome: Food and Agriculture Organization, International Life Sciences Institute, Industry Council For Development. Available from FAO on CD ROM or www.fsc.go.jp/ sonota/foodsafety_riskanalysis.pdf. Accessed 25 July 2007. Reid LM, Nicol RW, Ouellet T, Savard M, Miller JD, Young JC, Stewart DW, Schaafsma DW. 1999. Interaction of Fusarium graminearum and F moniliforme in maize ears: disease progress, . fungal biomass and mycotoxin accumulation. Phytopathology 89:10281037. Rose EF. 1973. Esophageal cancer in the Transkei: 19551969. Journal National Cancer Institute 51:716. Sadler TW, Merrill AH, Stevens VL, Sullards MC, Wang E, Wang P. 2002. Prevention of fumonisin B1-induced neural tube defects by folic acid. Teratology 66:169176. SCF. 2003. Scientific Committee on Food. Updated opinion on fumonisin B1, B2 and B3. Available: http://www.europa. eu.int/comm/food/fs/sc/scf/out185en.pdf. Accessed 25 July 2007. Seglar W. 1999. Case studies that implicate silage mycotoxins as the cause of dairy herd problems. Silage: field to feedbunk Vol. 99. Ithaca, NY: Natural Resource Agriculture and Engineering Service. pp 242254. Seglar W, Blake C, Cinch D, Chu FS, Gotlieb AR, Hinds M, Thomas C, Thomas E, Tomes N, Trenholm L, Undersander D. 1997. Comparison of mycotoxins levels among problem and healthy dairy herds. Available from Pioneer Hi-Bred International, Inc. DeMoines, Iowa. Schaafsma AW, Hooker DC. 2008. Climatic models to predict occurrence of Fusarium toxins in wheat and maize. International Journal of Food Microbiology. (in press). Schaafsma AW, Hooker DC, Pineiro M, Daz de Ackermann M, Pereyra S, Castano JP. 2006. Pre-Harvest Forecasting of deoxynivalenol for regulatory action in wheat grain in Uruguay using readily available weather inputs. In: Njapau H, et al. editors. Mycotoxins and phycotoxins: advances in

229

determination, toxicology and exposure management. Mycotoxins and phycotoxins: advances in determination, toxicology and exposure management. Wageningen, The Netherlands: Wageningen Academic Publishers. pp 227238. Schiefer HB, Hancock DS, Bhatti AR. 1986. Systemic effects of topically applied trichothecenes. I. Comparative study of various trichothecenes in mice. Journal of Veterinary Medicine A 33:373383. Schroeder HW, Christensen JJ. 1963. Factors affecting resistance of wheat to scab caused by Gibberella zeae. Phytopathology 53:831838. Scott PM. 1989. The natural occurrence of trichothecenes. In: Beasley VR, editor. Trichothecene toxicosis: pathophysiological effects. Vol. 1. Trichothecene toxicosis: pathophysiological effects. Boca Raton, FL: CRC Press. pp 226. Shelby RA, White DG, Burke EM. 1994. Differential fumonisin production in maize hybrids. Plant Disease 78:582584. Shephard GS, Leggott NL, Stockenstrom S, Somdyala NIM, Marasas WFO. 2002. Preparation of South African maize porridge: effect on fumonisin mycotoxin levels. South African Journal Science 98:393396. Shephard GS, Marasas WF, Burger HM, Somdyala NI, Rheeder JP, Van der Westhuizen L, Gatyeni P, Van Schalkwyk DJ. 2007. Exposure assessment for fumonisins in the former Transkei region of South Africa. Food Additives and Contaminants 24:621629. Shephard GS, van der Westhuizen L, Gatyeni PM, Somdyala NI, Burger HM, Marasas WFO. 2005. Fumonisin mycotoxins in traditional Xhosa maize beer in South Africa. Journal of Agriculture Food Chemistry 30:96349637. Snijders CHA. 1994. Breeding for Fusarium resistance in wheat and maize. In: Miller JD, Trenholm HL, editors. Mycotoxins in grain. Mycotoxins in grain. St. Paul, MN: Eagan Press. pp 3758. Snijders CHA. 2004. Resistance in wheat to Fusarium infection and trichothecene formation. Toxicology Letters 153:3746. Snijders CHA, Krechting CF. 1992. Inhibition of deoxynivalenol translocation and fungal colonization in Fusarium head blight resistant wheat. Canadian Journal of Botany 70:15701576. Snijders CHA, Perkowski J. 1990. Effects of head blight caused by Fusarium culmorum on toxin production and weight of wheat kernels. Phytopathology 80:566570. Sumarah MW, Miller JD, Blackwell BA. 2005. Isolation and metabolite production by Penicillium roqueforti, P paneum . and P. crustosum isolated in Canada. Mycopathologia 159:571577. Thrane U, Adler A, Clasen P-E, Galvano F, Langseth W, Lew H, Logrieco A, Nielsen KF, Ritieni A. 2004. Diversity in metabolite production by Fusarium langsethiae, Fusarium poae, and Fusarium sporotrichioides. International Journal of Food Microbiology 95:257266. Tong C, Hall CHS, Wang H. 2003. Land use change in rice, wheat and maize production in China (19611998). Agriculture, Ecosystems and Environment 95:523536. Torp M, Nirenberg HI. 2004. Fusarium langsethiae sp. nov. on cereals in Europe. International Journal of Food Microbiology 95:247256. Toth B, Mesterhazy A, Nicholson P, Teren P, Varga J. 2004. Mycotoxin production and molecular variability of European and American isolates of Fusarium culmorum. European Journal of Plant Pathology 110:587599. Tuller G, Armbruster G, Wiedenmann S, Hanichen T, Schams D, Bauer J. 1998. Occurrence of roquefortine in silage toxicological relevance to sheep. Journal of Animal Physiolology Animal Nutrition 80:246249.

230

J. D. Miller
Wilde F, Korzun V, Ebmeyer E, Geiger WW, Miedaner T. 2004. Comparison of phenotypic and marker-based selection for Fusarium head blight resistance and DON content in spring wheat. Molecular Breeding 19:357370. Wilkinson JM, Toivonen MI. 2003. World Silage. Lincoln, UK: Chalcombe Publications Wilson M. 2007. Increased frequency and intensity of floods. Catastrophic rapid climate change, 360 Risk Project. London: Lloyds. pp 1621. Wu F. 2004. Mycotoxin risk assessment for the purpose of setting international regulatory standards. Environmental Science and Technology 38:40494055. Xu XM, Parry DW, Nicholson P, Thomsett MA, Simpson D, Edwards SG, Cooke BM, Doohan FM, Brennan JM, Moretti A, et al.. 2005. Predominance and association of pathogenic fungi causing Fusarium ear blight in wheat in four European countries. European Journal of Plant Pathology 112:143154. Yoshizawa T. 1983. Red-mold diseases and natural occurrence in Japan. In: Ueno Y, editor. Trichothecenes chemical, biological and toxicological effects. Developments in Food Science. 44:1952009. Zhang X, Zwiers FW, Hegerl GC, Lambert FH, Gillett NP, Solomon S, Stott PA, Nozawa T. 2007. Detection of human influence on twentieth-century precipitation trends. Nature doi:10.1038/nature06025.

Turner PC, Burley VJ, Rothwell JA, White KLM, Cade JE, Wild CP. 2007. Dietary wheat reduction decreases the level of urinary deoxynivalenol in UK adults. Journal of Exposure Analytical Environmental Epidemiolology. (in press). Ueno Y. 1983. Trichothecenes: chemical, biological and toxicological aspects. Developments in Food Science Vol. 4. Amsterdam: Elsevier. Udegawa S. 1988. Mycotoxicoses the present problem and prevention of mycotoxins. Asian Medical Journal 31:599604. Vesonder RF, Hesseltine CW. 1981. Vomitoxin: natural occurrence on cereal grains and significance as a refusal and emetic factor to swine. Process Biochemistry 16:12,14/15, 44. Vesonder RF, Ciegler A, Jensen AH. 1973. Isolation of the emetic principle from Fusarium-infected corn. Applied Microbiology 26:10081010. Voss KA, Chamberlain WJ, Bacon CW, Herbert RA, Walters DB, Norred WP. 1995. Subchronic feeding study of the mycotoxin fumonisin B1 in B6C3F1 mice and Fischer 344 rats. Fundamental Applied Toxicology 24:102110. Wang YZ, Miller JD. 1988. Screening techniques and sources of resistance to Fusarium head blight. In: Klatt AY, editor. Wheat production constraints in tropical environments. Wheat production constraints in tropical environments. Mexico City: CIMMYT. pp 239250. Wei RD, Still PE, Smalley EB, Schnoes HK, Strong FM. 1973. Isolation and partial characterization of a mycotoxin from Penicillium roqueforti. Applied Microbiology 25:111114.

Food Additives and Contaminants, February 2008; 25(2): 231240

Design and implementation of an integrated management system for ochratoxin A in the coffee production chain

R. LOPEZ-GARCIA1, C. AUGUSTO MALLMANN2, & M. PINEIRO3

Logre International Food Science Consulting, Mexico, D.F., Mexico, 2Laboratorio de Analisis Micotoxicologicas, Universidade Federal de Santa Maria, Santa Maria, Brazil, and 3Food and Agricultural Organization (FAO), AGN, Rome, Italy
(Received 24 May 2007; accepted 28 January 2008)
1

Abstract Coffee is an important export product of Ecuador. Producers are challenged by the implementation of regulatory limits for ochratoxin A. Ecuador has four coffee production areas and the potential for mycotoxin contamination varies due to different environmental conditions and cultural differences in harvesting, storage, processing and commercialization. The major contributors to contamination are the lack of selection during harvesting, delays in drying or rewetting, the lack of proper drying and storage conditions, the mixing of products with different levels of moisture, and the potential for crosscontamination. The long commercialization chain involves different intermediaries that use foreign materials to increase the weight of the product without consideration of quality. An integrated mycotoxin management system using the Hazard Analysis Critical Control Point Systems (HACCP) principles was developed to prevent mycotoxin contamination at each stage of production. Critical control points were developed based on the resources available at the different stages of the production chain. Training programmes helped increase awareness about the impact of contamination, but failed to transform knowledge into improved practices. Thus, different demonstrative models specific for each productive region at all production levels were developed to show the application of prevention mechanisms using limited resources and to demonstrate the increased commercial value of coffee produced using good practices throughout the chain so producers have a better disposition to adopt improved practices. Preliminary results show that coffee managed using the models had a better quality, a lower contamination, a higher yield and better commercial value. The use of local resources and low-cost technology was important in demonstrating the practical approach.

Keywords: Ochratoxin A, coffee, hazard analysis critical control point systems (HACCP), integrated mycotoxin management

Introduction Mould infestation of susceptible agricultural products such as cereal grains, nuts and fruits is common in food supplies worldwide. Mould growth can result in reduced crop yields, losses during storage and mycotoxin contamination. Some mycotoxins such as ochratoxin A (OTA) have been recognized as contaminants of concern from a food safety perspective. OTA is produced by mould from the Penicillium and Aspergillus genera particularly by P. verrucosum, A. ochraceus and A. carbonarius. Of these species, A. ochraceus grows at moderate temperatures and at water activity above 0.8 and

may infect coffee beans producing OTA during drying (Food and Agricultural Organization (FAO) 2006). Contamination may also occur during storage and transportation if coffee is exposed to humid conditions. OTA is a known nephrotoxin, carcinogen and genotoxic agent (Joint Expert Committee on Food Additives (JECFA) 2001). Although coffee is processed at high temperatures during roasting, mycotoxins such as OTA, and to a lesser extent aflatoxin, are known to survive roasting and remain as a potential hazard. Several studies have reported OTA data for raw and roasted coffee since 1980, and occurrence continues to be reported from different regions of the world (Levi 1980,

Correspondence: R. Lopez-Garcia. E-mail: rebecalg@prodigy.net.mx ISSN 0265203X print/ISSN 14645122 online 2008 Taylor & Francis DOI: 10.1080/02652030801946546

232

R. Lopez-Garcia et al. Association (ANECAFE) tried to address mycotoxin contamination by participating in the global project Enhancement of Coffee Quality through the Prevention of Mould Formation (GCP/INT/743/ CFC) financed by the Common Fund for Basic Products of the United Nations Food and Agriculture Organization (FAO) (ANECAFE, Statistics for coffee exports from Ecuador, personal communication, 2006). In 1999, a processed coffee shipment from Ecuador was denied entrance to France due to contamination with OTA at levels of 21.1 and 13.8 mg kg1. Subsequent analysis done in 2002 showed there was OTA contamination ranging from 0.6 to 104.1 mg kg1, with an average contamination of 14.95 mg kg1. Curiously, coffee beans reported as being very damaged by the coffee berry borer (Hypothenemus hampei) and visibly mouldy were not highly contaminated (FAO 2004). During 2003, fourteen field technicians participated in an international training workshop on good agricultural practices (GAP) and Hazard Analysis Critical Control Point Systems (HACCP) for mycotoxin prevention and control sponsored by the FAO. However, due to the detection of mycotoxin contamination, in particular OTA, and the importance of coffee production in the country, Ecuador requested additional technical support from the FAO. This technical cooperation project (FAO 2004) was developed in coordination with the Ecuador Ministry of Agriculture (MAG), ANECAFE and the FAO. The main objective was to develop national awareness in relation to contamination issues due to mould contamination and mycotoxin formation and the impact on food safety. Specific objectives included training on prevention and control strategies at different levels of the coffee production chain, strengthening of the national analytical capabilities, development of communication tools, and development of a concerted National Action Plan to ensure long-term improvement of the safety and quality of coffee from Ecuador.

Tsubouchi et al. 1987; Studer-Rhor et al. 1995; Nakajima et al. 1997). Worldwide contamination trends may be hard to establish since many analyses come from European customs offices, therefore restricting some contamination information to coffee exported to the European Union. Coffee is one of Ecuadors major agricultural export products and together with bananas, cocoa beans and flowers represents the majority of the income from agricultural exports. It is estimated that in 2005 Ecuador had at least 200,000 hectares destined for coffee production. This activity represents the major source of income for approximately 130, 000 families that have regrettably suffered with the different commercial crises that have affected international coffee trade. These estimates do not consider the number of jobs generated by the industrialization and commercialization of coffee beans. According to the Ecuadorian Institute of Statistics, approximately 8% of the population depends directly or indirectly on coffee production (Instituto Nacional de Estadstica y Censos (INEC) 2005). Ecuador is one of the 14 coffee producer countries that, due to its geographical location and conditions, produces both arabica (Coffea arabica) and robusta (Coffea canephora) coffee. Coffee is produced throughout the country in 20 of the 22 provinces and can be generally located in four different distinct regions (INEC 2005): Coast, Andes, Amazon, and Islands, with the majority of production concentrated in the Coast region where Arabica is produced predominantly. Each production region has different challenges due to the environmental conditions, type of production, processing and commercial routes. Coffee production in Ecuador is a family activity done in small farms where primary production is not technified and demands a lot of labour. Lack of productivity and low yields are common problems for most producers. Since the international coffee price crisis that started in 1998 and worsened during 1999, coffee producers have faced serious problems and have struggled to maintain production. In many cases, the commercial price did not cover production costs leading to the loss of income and eventual crop replacement. Although efforts have been made internationally to stabilize prices, producers are still plagued with challenges that are now increased by the new OTA limits established by several countries. Table I presents some of the major strengths, weaknesses and risk factors for OTA contamination in Ecuador. Ecuador exports mainly soluble coffee from domestic or imported coffee beans. Since the major market for Ecuadorian coffee are countries in the European Union with implemented regulations (Table II), the National Coffee Exporters

Integrated mycotoxin management systems In order to establish clear strategies for this project it was important to know the conditions by which coffee is produced, processed, stored and transported in the country. In general, it was determined that there were four distinct coffee production areas with very different systems due to the type of coffee produced and the environmental and market conditions: . The Coast is the area with the highest production of arabica coffee. Harvesting is conducted during the dry season so

Table I. Major weaknesses, strengths and risk factors for contamination with ochratoxin A (OTA). Strengths in the coffee production chain of Ecuador Major risk factors for contamination with OTA Lack of awareness of the problem

Weaknesses in the coffee production chain of Ecuador

Lack of education

Lack of awareness of contamination issues Good industry capability Existence of updated technical norms

Existence of a national organization Consejo Cafetalero Nacional (COFENAC) that provides technical support throughout the coffee production chain Production of both species, robusta and arabica

Low productivity and yield

Decrease of cultivated areas

Lack of a producer organization

Ideal environmental conditions for mould proliferation General lack of sanitary conditions for coffee production Presence of pests such as the coffee borer (Hypothenemus hampei) Lack of selective harvesting

Lack of a commercial organization

Flourishing of a new industry of specialty coffees with access to higher value markets Availability of a laboratory equipped to perform ochratoxin analyses

Lack of access to financial resources

Lack of insurance for agriculture Low income for producers Increase of production costs and decrease of competitiveness

Lack of technification Lack of communication throughout the coffee production chain Low production of washed coffee Lack of incentives to improve quality Disorganization throughout the commercial chain

Decrease of exports Low domestic consumption of coffee

Integrated mycotoxin management system for coffee

Lack of enforcement of existent regulations Purchasing of coffee without consideration of quality issues

Harvested beans mixed with dirt, fallen beans, rocks and other contamination vectors Storage of freshly harvested beans under inadequate conditions for over 8 h Lack of infrastructure for wet processing Lack of adequate infrastructure for drying Product blending with products of different moisture content Lack of adequate storage infrastructure Storage for long periods of time Storage with other grains Lack of adequate transportation conditions Extremely long commercial chain with several intermediaries Lack of bean sorting and selection Lack of maintenance and clearing of processing equipment Cross contamination with peeling operations

233

234

R. Lopez-Garcia et al.
Table II. Regulatory requirements and coffee exports from Ecuador by volume and value (US$), 2005. Regulatory requirements Roasted and ground coffee (ppb)* 5 5 5 5 5 5 5 5 5 5 5 Soluble coffee (ppb)* 10 10 10 10 10 10 10 10 10 10 5 Exports from Ecuador Volume of soluble coffee exported from Ecuador (number of 60-kg bags, 2005) 458, 638.08 125, 744.67 11, 700.00 5427.50 915.64 587.17 390.00 Value of soluble coffee exported from Ecuador (US$ thousands, 2005) 45, 039.13 10, 331.01 1123.50 424.30 68.92 60.98 53.60

Green coffee beans (ppb)* European Union (general) n.d.

Countries with individual regulations Germany** n.d. Czech Republic** 10 Netherlands** Spain** 8 Finland** 5 Italy** 8 Greece** 20 Hungary** 15 Portugal** 8 Switzerland 5

Sources: FAO, ANECAFE. Notes: *Limits were updated to November 2004. The status for the limits differs by country. Some are included in the regulations and others are norms, customs instructions or inspection guidelines. **Limits of roasted and ground coffee and soluble coffee for European Union members have been expressed as the European Union general limit.

environmental conditions do not usually play a major role in the drying process and contamination issues may be prevented with improved practices. . The Amazon produces mostly robusta coffee and faces high relative humidity and rainy conditions during harvesting, thus drying operations are extremely difficult. In addition, transportation is challenging since coffee for exportation to the European Union is transported by land and will take several hours on roads with different conditions going through diverse environmental extremes. . Coffee in the Andes is produced by small producers, many of which are now organized in cooperatives. Although the use of washing is still not widespread, producers in this region are slowly transforming to this method of processing. In general, there are dry conditions during harvesting which facilitate drying operations. However, many producers that wash coffee do not dry it immediately and usually transport it as wet parchment. Many producers in the Andes region have turned to the specialty coffee market and are participating in organic and fair trade certification programmes. Since they are more organized and have been able to penetrate some specialty markets successfully, producer awareness is better. Contamination may be prevented with improved practices during wet processing and storage.

. The Galapagos Islands produce mainly specialty organic coffee. There is one major producer that has access to top-of-the-line organic technology and only uses wet processing. However, dry parchment coffee is then transported by sea to the mainland for further processing such as sorting, grading, polishing, blending, roasting and grinding. This transportation operation may again increase the risk of contamination. In addition, processing facilities also polish and blend coffee from the mainland, so cross-contamination is also a possibility. Increased awareness and improved storage, transportation and processing practices may help prevent and control contamination of coffee from the islands. Throughout the country, with the exception of producers with special certifications, coffee may be sold to a number of different intermediaries who will in turn blend it with coffee from other producers. Sometimes water, dirt or rocks are added to increase weight and get more money during commercialization. Coffee with different levels of moisture is usually blended and stored without any further precaution and in some cases storage conditions are less than ideal. In many cases, coffee is stored piled directly on the ground and mixed with other commodities with different levels of moisture such as cocoa beans, corn, peanuts and peppercorns. Although each region presents a unique challenge, a general HACCP model was developed to represent the major steps in coffee production. Obviously, the

Integrated mycotoxin management system for coffee

PRE-HARVEST

235

HARVEST WASHING DRYING ON THE FARM SORTING

CCP1

DRYING

STORAGE 1

SORTING

TRANSPORTATION SAMPLING AND ANALYSIS (MORPHOLOGY AND MOISTURE)

RECEPTION

SORTING DRYING
CCP2n

SAMPLING AND ANALYSIS (MORPHOLOGY, MOISTURE AND OTA)

STORAGE 2n TRANSPORTATION

RECEPTION

CCPn +1

GENERAL STORAGE

GENERAL STORAGE

INDUSTRIAL PROCESSING

CLEANING

DOMESTIC SALES

EXPORT MA RK ET

BLENDING

Figure 1. General flow chart for coffee production in Ecuador showing potential critical control points (CCPs).

flow diagram may be quite fragmented because different players will be involved. In addition, the storage, reception, and transportation phases may occur more than once due to the very long commercial chain involving many intermediaries. Figure 1 shows a general flow diagram with proposed critical control points, and Table III shows a summary of the critical control points (CCP). It is important to point out that critical limits were established according to the international guidelines emanating from the FAO Global Project (FAO 2006). This is obviously, a modified version of a HACCP plan since some steps had to be adapted to obtain an integrated management approach. In other words, CCPs may not follow the formal

definition because none minimize the risk associated with OTA in a single step; however, correct application of these controls in conjunction with the use of good practices throughout the production chain should prevent and control mould contamination at each stage of production. In addition, some of the monitoring operations are qualitative since they are adapted to the resources available at each stage of production.

Training The Train the Trainer approach was used to teach GAP and HACCP for mycotoxin management.

236

Table III. Summary of Hazard Analysis Critical Control Point Systems (HACCP) Critical Control Points (CCPs). Monitoring

Critical control points Critical limits Where How How often Who

Hazard(s)

Corrective actions

Verification

Records

R. Lopez-Garcia et al.

CCP1 Drying on the farm or drying after wet processing Patios, tents or dryers Every 48 h depending on the drying system Farmer, supervisor or operator Use of mechanical dryers or redistribution of coffee beds on patios to obtain the desired depth Qualitative (supervisor uses traditional sensory methods such as biting onto the bean or cracking it to hear to a characteristic sound) Qualitative (supervisor uses traditional sensory methods such as biting onto the bean or cracking it to hear to a characteristic sound) Or moisture analyses Every 48 h depending on the drying system Laboratory analysis Each lot Farmer, supervisor or operator Product segregation

Biological: presence of toxigenic moulds Chemical: contamination with ochratoxin

Moisture <15% Depth of drying bed <4 cm

Use of a moisture analyser

Lot records and moisture reports

CCP2!n Drying before storage Patios, tents or dryers

Biological: presence of toxigenic moulds Chemical: contamination with ochratoxin

Moisture 12% Depth of drying bed <4 cm Mechanical dryer settings

OTA analysis

Lot records, moisture reports and OTA analysis

CCPn1 Reception at the processing facility Reception area Laboratory

Biological, chemical: damaged coffee or contamination with OTA

Percentage defects according to the Ecuadorian technical norm number nte inen 285 OTA <6 ppb

Buyer/receiving supervisor

Rejection or segregation if the per cent defects is more than allowed or OTA is over the limit

Laboratory analysis of the finished product (in-house or official laboratory)

Documentation for reception Laboratory records for each lot

Integrated mycotoxin management system for coffee A total of 57 field technicians were trained as trainers in two different workshops. These trainers in turn taught 442 regional workshops that reached 4213 people in all coffee production areas. In addition, three good manufacturing practices (GMP) and HACCP workshops were taught to industry personnel with a total participation of 94 people. These workshops left a good base of personnel trained on the basic methods for prevention and control. However, the question remained on how to transform this knowledge into a practical approach since most producers did not understand the contamination issues and did not feel this was a problem to which they were directly related. In addition, producers felt that additional controls were a burden for their already aggravated commercial systems. A common question during the workshops was as follows. If I do this, how much will the international price increase and how much more money will I make? Obviously, these were not answers that could be obtained through the framework of the project. So, although training had a good outcome, its impact was questionable except for the increased awareness that was accomplished. A different approach was needed to encourage producers to adopt controls.

237

Development of demonstrative models Since producers perceived controls as an economic burden to their system without increased price, it was decided to create demonstrative models. Different models were designed according to the environmental conditions for each region. The objective was to demonstrate that regardless of the international price for coffee, the use of good practices throughout the chain would help them increase the yield, quality and marketability of their products increasing value rather than price. These model systems were created using a volunteer producer, merchant or organization that was perceived as a community leader, had awareness of the impact of contamination on the country in general and their commercial infrastructure in particular and was willing to lend his warehouse, drying patio or processing facility. Selective harvesting was encouraged by selecting volunteer producers that would harvest only mature berries and discard over-mature or dropped berries. Green berries were saved until maturation. This expanded the commercial cycle and some producers were able to save berries and sell them at the peak of pricing instead of the low prices experienced at the peak of the harvesting season. The intent was to have these producers share their experiences through personal communication.

Unfortunately, it will take several harvesting cycles to obtain reliable information, but at this point this demonstrative mechanism has been seeded for future reference. A model warehouse was created in the Coast region. A warehouse building was already available; however, common storage practices included mixing of products with different moisture levels or different commodities. Products of extremely different qualities were also mixed in the warehouse. Therefore, if the producer had originally harvested high-quality beans that could have achieved a higher price in the market, by the time it arrived at the warehouse, it had lost its additional value due to blending with products of questionable quality. Thus, the resulting product had low quality and low value and had already lost a potential higher market price. Also, walls on the warehouse had evident mould growth and products were blended with leftovers from the previous harvest. Blended beans stored directly on the floor had such mould contamination that they were white and had high temperatures. Obviously, in addition to losing quality and increasing potential for contamination, the evident contamination is directly related to product yield and, therefore, decreased material to commercialize further. The walls of the model warehouse were white washed as a means for inexpensive disinfection. Small cubic (1 1 1 m) containers were made of cane, a construction material that is widely available in the region and which is successfully used for building traditional houses and storage units. This material, if handled correctly, is extremely resistant to mould and other pests. Also, it is widely available so the cost of the material is not an issue and, in many cases, producers or merchants can harvest it from their own land. Since cane is cut into thin strips, finished containers had open strips that allowed for product ventilation. Additionally, the wooden frame was oversized to allow for stacking and to prevent storage immediately next to the wall. Therefore, wall spacing and spacing between containers to allow for air flow between the containers was guaranteed. These small containers were useful for product sorting, maintaining coffee of different qualities and other commodities segregated. Since, originally, the warehouse capacity depended only on the ability to pile grains on the floor, the use of containers immediately increased the warehouse capability and, thus, the merchants ability to commercialize more products. This improved his perception of the value of good practices and generated a model storage unit that can be easily multiplied by other producers in the area. In the Amazon region, another model warehouse was created. In addition, several artisanal drying mechanisms for small volumes were explored.

238

R. Lopez-Garcia et al. In order to obtain long-term success, producers will have to imitate these models and multiply the efforts. However, one of the major challenges will be their ability to obtain the financial resources needed to improve their production, storage and processing infrastructure. During the project, negotiations were initiated with the Ministry of Agriculture to obtain long-term soft credits of US$3001000/producer or merchant for individuals who had participated in the training programmes. These funds will be used to establish adequate drying patios or, in the case of the Amazon area, drying tents or artisanal dryers as well as improving storage conditions. In the Andes region funds will be used to buy small depulpers and establish adequate drying patios to produce washed arabica coffee through wet processing. It will take several harvest cycles to determine if producers requested the funds and the result of their investments with the subsequent impact on mycotoxin contamination.

Since there is high relative humidity and constant rainfall in this area, it was essential to provide producers with accessible drying technology. Although some merchants own mechanical dryers, one of the major issues is that these dryers have a medium to large capacity (1020 tonnes). Producers are small and a single producer is not able to harvest such large volumes at any given day. Thus, recently harvested coffee berries are stored wet for several days until enough volume is obtained to run the dryer. This process may take several days, therefore there is increased risk of contamination. The use of drying patios is not feasible due to rainfall. Traditionally, producers had tried to use tents to protect their products during drying. However, there are some design issues associated with these tents because moisture accumulates inside and during night condensate falls again onto the product. A demonstrative model for this method of drying included a tent built with a double roof that was designed to let moisture escape during the day, protect the product from rainfall and avoid nighttime condensation. Also, structures were built to raise the product off the ground and allow for air circulation. In addition, an artisanal dryer made of wood and operated with domestic gas was constructed for demonstration. This dryer can be operated with a low volume of coffee as opposed to the commercial dryers that were originally available. This region also faces a very disorganized commercial chain. There are several formal established merchants. However, at the peak of the harvesting season, opportunistic merchants emerge and buy coffee regardless of the quality. This has a direct impact on the efforts to control contamination because producers are able to sell the worst coffee through this route. This coffee is then blended with good coffee, thus getting a final product with lower quality. To address this issue, the organization of a formal merchant association was facilitated during the project. The idea was to empower formal merchants to ensure that they would all buy coffee under the same criteria and segregate coffee of questionable quality. Through this organization, merchants can negotiate directly with exporters and avoid the chain of intermediaries that tend to blend coffee and add water and other materials to increase weight. Thus, organized merchants have a better chance of obtaining better value and avoiding contamination. In the Andes region the use of small drying patios to dry wet parchment coffee before transportation was promoted. In addition, the warehouse of one of the major cooperatives was improved as a demonstrative model for good storage practices.

Strengthening of analytical capabilities At the beginning of the project only a few of the major processing industries had analytical equipment to determine mycotoxin contamination. The official mycotoxin analytical capabilities were extremely limited. Thus, it was important to establish at least one laboratory that would be able to perform analytical services. The laboratory at the Instituto nomo Nacional Auto de Investigaciones Agropecuarias, Estacion Experimental Santa Catalina (INIAP) was selected due to the analytical capability already in place and its status as an official institution. The high-performance liquid chromatography (HPLC) system already available in the laboratory was equipped with a fluorescence detector. Other analytical resources needed in the laboratory were also financed by the project. Twelve laboratory technicians were trained in sampling and analysis methodology for OTA and the method for OTA analysis for green coffee beans was validated. Laboratory personnel were trained in validation methodology for other methods. In addition, the training included the use of thin-layer chromatography and rapid methods for OTA analysis. This laboratory is also preparing for ISO 17025.2005 certification. One of the major issues for this laboratory will be long-term sustainability of analytical services. Therefore, the use of the systems and methodology for other commodity analysis, i.e. cocoa beans, was promoted to increase the number of potential samples to be analysed. This laboratory will need to analyse at least 50 samples per month for sustainability. In addition, the major exporters made a commitment to send samples on a monthly basis

Integrated mycotoxin management system for coffee and promote the use of the services among the producers that work with them. Since in recent years the coffee industry has had to import green coffee beans to produce soluble coffee, this laboratory can also be used for screening contamination in imported coffee. Conclusions

239

Communication materials Long-term outreach and training were also considered essential for the success of the project. Therefore, brochures with basic information on OTA, GAPs and HACCP, sampling and analysis, and drying technologies were developed, printed and distributed throughout the regions where work was conducted. Posters with general information on mycotoxins and OTA were also printed and distributed. Also, two different radio spots were produced in Spanish and Quechua and were constantly aired in the coffee producing areas. A video explaining basic GAPs and the use of integrated management systems was also produced to ensure continuous training. These materials were an essential tool to raise awareness and multiply the original training efforts. In addition, in order to increase awareness of issues related to mycotoxin contamination in the food and feed production chains, several technical presentations were made throughout the two years. These presentations involved stakeholders from industry, government and academia. Also, newspaper articles with general information on mycotoxins and the importance of prevention and control were published on several occasions.

A continuous effort is needed in order to observe long-term results and improved safety and quality of coffee from Ecuador. However, initial project outcomes show that continuous application and followup of the recommendations will help prevent and control not only OTA contamination, but also other contamination issues that might have a direct impact on the safety and quality of coffee and other agricultural products in Ecuador. Table 4 shows results from random sampling and analysis of coffee during the 2006 harvest season in the areas where the project had its major activities. Although these results are not representative of the whole country and are not statistically conclusive, they can be considered indicators of improved management. However, it will take several harvest cycles to determine not only sustainability of the improved practices, but also replication by other producers and long-term impact throughout the country. The use of an integrated approach combining good practices and the HACCP methodology throughout the coffee production chain are essential to achieve improved safety and quality. However, more research is needed on a global basis to refine the limits established as critical control points and validate the controls. The newly equipped laboratory will help analyse products for both domestic and export markets and provide a continuous diagnostic tool to determine mycotoxin contamination issues within the country. Finding mechanisms for the financial support of the laboratory will be essential to continue analytical services and allow for the establishment and validation of other methods. Ecuador does have current technical standards for coffee and enforcement will

Development of a National Action Plan In order to ensure long-term sustainability of the project and continuation of the effort once it concluded, a concerted National Action Plan was developed. This Plan was developed with the input from the different stakeholders of the coffee production chain and established short-, medium-, and long-term strategies and recommendations to achieve the desired outcomes. Short-term strategies were addressed during the development of the FAO project. In addition, an Official National Commission integrated by all stakeholders in the coffee production chain and coordinated by the Consejo Cafetalero Nacional (National Coffee Council; COFENAC), an official government institution with national representation, was established to provide follow up and support to the Plan.
Table IV. Results of random sampling and analysis of samples from different regions of Ecuador at the end of the project (not representative of the whole country). Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 Moisture (%) 12.10 12.7 10.7 11.5 13 12.20 12.00 13.00 13.20 11.00 11.50 12.50 12.40 Average OTA (mg kg1) (HPLC) n.d. 0.16 0.12 0.10 0.14 0.15 0.05 0.08 0.12 n.d. n.d. 0.10 0.17 0.18

240

R. Lopez-Garcia et al.
Food and Agricultural Organization (FAO) [internet]. 2006. Guidelines for the prevention of mould formation in coffee [cited 2006 Apr 13]. Available: http://www.ico.org/documents/ ed1988e.pdf Instituto Nacional de Estadstica y Censos (INEC). 2005. Encuesta de superficie y produccion agropecuaria continua [cited 2006 Sept 11]. Available: http://www.inec.gov.ec/ interna.asp?incenc_tablas_graf&idEncuesta13/ Joint Expert Committee on Food Additives (JECFA). 2001. Ochratoxin A. Joint Food and Agriculture Organization/World Health Organization Expert Committee on Food Additives [cited 2006 Sept 11]. Available: http://www.inchem.org/documents/jecfa/jecmono/v4je04.htm Levi C. 1980. Mycotoxins in coffee. Journal of the Association Official Analytical Chemists 63:12821285. Nakajima M, Tsubouchi H, Miyabe M, Ueno Y. 1997. Survey of aflatoxin B1 and ochratoxin A in commercial green coffee beans by high-performance liquid chromatography linked with immunoaffinity chromatography. Food and Agricultural Immunology 9:7783. SCOOP. 2002. European Commission, SCOOP-task 3.2.7.2002. Assessment of dietary intake of ochratoxin A by the population of EU Member States. European Commission, DirectorateGeneral Health and Consumer Protection, Reports on Tasks for Scientific Cooperation [cited 2007 May 5]. Available: http://europa.eu.int/comm/food/fs/scoop/3.2.7_en.pdf/ Studer-Rhor I, Dietrich DR, Schlatter J, Schlatter C. 1995. The occurrence of ochratoxin A in coffee. Food and Chemical Toxicology 33:341355. Tsubouchi H, Yamamoto K, Hisada K, Sakabe Y, Udagawa S. 1987. Effect of roasting on ochratoxin A level in green coffee beans inoculated with Aspergillus ochraceus. Mycopathologia 97:111115.

also be important in the effort to improve quality, safety and competitiveness of the whole sector. It will take a joint effort from all players in the coffee production chain to obtain long-term success that will not only increase exports, but also improve the quality of life of the different stakeholders.

Acknowledgements The results presented are part of Food and Agricultural Organization (FAO) Project No. TCP/ ECU/3001 (T) and the national counterparts MAG and ANECAFE. The authors wish to thank the FAO Representation in Ecuador: Juan Carlos Villacs, Alberto Larco, Marcelo Macas, Victor Hugo Chala, Ruben Alcvar, Pedro Tulio Gomez, Susana Espn de Rivera, the technical team at COFENAC, and Laura Cardenas for their participation and contributions to the project.

References
Food and Agricultural Organization (FAO) [internet]. n.d. Reducing ochratoxin A in coffee [cited 2006 Sept 11]. Available: http://www.coffee-ota-org/faq/ Food and Agricultural Organization (FAO). 2004. Prevencion de hongos productores de Ocratoxina A (OTA) en el cafe ecuatoriano. FAO. Project Proposal No. TCP/ECU/3001 (T).