Comprehensive REVIEWS iiFioi Seien» ni Fiod Safety Filamentous Fungi and Mycotoxins in Cheese: A Review Nolwenn Hymery, Valerie Vasseur, Monika Coton, Jeröme Mounier, Jean-Luc Jany, Georges Barbier, and Emmanuel Coton Abstract: Important fungi growing on cheese include Penicillium, Aspergillus, Cladosporium, Geotrichum, Mucor, and Trichoderma. For some cheeses, such as Camembert, Roquefort, molds are intentionally added. However, some contaminating or technological fungal species have the potential to produce undesirable metabolites such as mycotoxins. The most hazardous mycotoxins found in cheese, ochratoxin A and anatoxin Ml, are produced by unwanted fungal species either via direct cheese contamination or indirect milk contamination (animal feed contamination), respectively. To date, no human food poisoning cases have been associated with contaminated cheese consumption. However, although some studies state that cheese is an unfavorable matrix for mycotoxin production; these metabolites are actually detected in cheeses at various concentrations. In this context, questions can be raised concerning mycotoxin production in cheese, the biotic and abiotic factors influencing their production, mycotoxin relative toxicity as well as the methods used for detection and quantification. This review emphasizes future challenges that need to be addressed by the scientific community, fungal culture manufacturers, and artisanal and industrial cheese producers. Keywords: Mold, filamentous fungi, mycotoxins, cheeses, diversity, mycotoxin toxicity, mycotoxin biosynthesis, mycotoxin ecological role, mycotoxin control Introduction Many food spoilage fungi belonging to the genera Aspergillus, Fusarium, and Penicillium produce mycotoxins that are toxic for vertebrate and other animal groups when introduced via a natural route (food ingestion). Naturally present in ambient air, soil, and crops (Yiannikouris and Jouany 2002), mycotoxigenic fungi are considered to be among the most significant contaminants in foods in terms of impact on public health, food safety, and the national economy of several countries (Steyn 1995; Pitt 2000). The term mycotoxin, derived from the Greek words "mykes" meaning fungi or slime molds and "toxicum" referring to toxin, was utilized for the first time in 1962 following a veterinary crisis that resulted in the death of about 100000 young turkeys. These birds had ingested contaminated peanut meal containing secondary metabolites (anatoxins [AFs]) produced by Aspergillus jiavus (Blount 1961; Hundley 2001). Mycotoxins are defined as secondary metabolites, hence not directly essential for normal fungal growth, and are often low-molecular-weight compounds. They are naturally occurring molecules and are thought to confer a selective advantage to the producer strain within complex ecosystems. Their role is to communicate and compete with other organisms or to inhibit competi- MS 20131758 Submitted 11/26/2013, Accepted 12/2/2014. Authors are with Laboratoire Universitaire de Biodiversité et d'Ecologie Microbienne, ES1AB, Technopóle de Brest Iroise, Universitě de Brest, EA3882, 29280 Plouzané, France. Direct inquiries to author Hymery (E-mail: nolwenn.hymery@univ-brest.frJ. This article was originally published on 23 June 2014. Subsequently Figure 2 was corrected to include a previously missing part. The corrected version was published on 10 July 2014. © 2014 Institute of Food Technologists8 doi: 10.111 1/1 541-4337.12069 1 tor growth and reproduction in the same trophic niche (Magan and Aldred 2007; Fox and Howlett 2008). Early mycotoxin production could also allow molds to rapidly colonize the environment. These metabolites are commonly found in various food and feed commodities. Indeed, molds are able to grow on a wide range of substrates (including cereals, meats, nuts, cheeses, grapes, coffee beans, apples, and derived products), at any stage of production in the field, during postharvest storage, and under a wide range of climatic conditions. Moreover, mycotoxins are known to be resistant to industrial processing. It has been established that mycotoxin production profiles may change under different fungal growth conditions including substrate composition (Kokkonen and others 2005a), ecophysiological factors including temperature, water activity, pH or oxygen concentration, and biotic factors including microbial interactions (Magan and Aldred 2007). Analytical methods for identifying and quantifying mycotoxins in foods and feeds have been widely developed. Maximum levels for AFs, deoxynivalenol, fumonisins, ochratoxin A (OTA), patulin, T2 and HT-2 toxins, and zearalenone are regulated in the EU (Commission Regulation [EC] Nr. 1881/2006) as well as worldwide due to their well-characterized adverse health effects after acute, short-term, and/or long-term exposure. In cheese, anatoxin Ml (AFM1) is the only mycotoxin for which maximum levels (0.05 and 0.5 ppb in the milk used for cheese-making in the EU and United States and China, respectively) have been set. In cheese, the most hazardous mycotoxins are OTA and AFM1. They are, respectively, produced by unwanted fungal species either via direct cheese contamination (spoilage agents) or indirect contamination of the milk used for cheese-making due to feed contamination. Among fungal cheese spoilers, the most problematic dI.13,2014 • Comprehensive Reviews in Food Science and Food Safety 437 Filamentous fungi and mycotoxins in cheese. mycotoxigenic ones belong to Penicillium spp. as they are particularly well adapted for growth on and/ or in the cheese matrix. Filamentous fungi also play a central role during cheese-making. Two well-known fungal ripening cultures, Penicillium camemberti and Penicillium roqueforti, which play a significant role in appearance, texture, and flavor development of mold-ripened and blue-veined cheeses, are also known for their ability to produce several mycotoxins. For example, P. camemberti produces cyclopi-azonic acid (CPA), while roquefortine C and mycophenolic acid (MPA) are produced by P. roqueforti. Yet, these metabolites exhibit lower toxicity than those reported above and are present at low levels in cheese. Due to their long history of safe use in cheese in the U.S.A. and EU (no cases of food poisoning associated with the consumption of contaminated cheese were described in literature), use of these fungal cultures is regarded as safe in their intended conditions of use (cheese-making). In addition, these microorganisms are granted a generally recognized as safe (GRAS) status by the U.S Food and Drug Administration (FDA). However, in the EU, while the European Food Safety Agency (EFSA) acknowledged that there are no reports of adverse health effects for cheeses produced with these microorganisms, the qualified presumption of safety (QPS) status was not granted for these species. This was mainly due to a lack of knowledge on mycotoxin production by the strains intentionally used, the occurrence of these mycotoxins in cheese and their toxicity (EFSA 2011). In this context, scientific questions remain concerning the presence of mycotoxins in cheese, their variability in terms of quality and quantity, relative toxicity levels, impact of biotic and abiotic factors on their production as well as their control. The aim of the review is to summarize the current knowledge on filamentous fungi and mycotoxins in cheese. The paper deals with fungal diversity, the main mycotoxins reported in cheeses, mycotoxin toxicity, and their detection and quantification. The biosynthetic pathways described to date in cheese-related species as well as abiotic and biotic factors impacting mycotoxin production in cheese are also presented. Finally, physical, chemical, and biological methods potentially usable to prevent fungal growth and indirectly eliminate or reduce mycotoxin levels in cheese are discussed (from farm to fork). Filamentous Fungi in Cheeses: The Good and the Bad Few studies have extensively investigated fungal communities in cheeses. Barrios and others (1998) provided a global view of mold diversity in 52 commercial cheese susing a culture-dependent approach. Penicillium species were the most frequent (found in 63% of cheeses), followed by Mucor spp. (27%), Geotrichum candidum (17%), and 12 other fungal genera identified in 2 to 10% of the samples. Such culture-dependent methods have also identified Penicillium spp. as the prominent mold in sheep and goat cheeses from different origins (Montagna and others 2004; Hayaloglu and Kirbag 2007; Bath and others 2012). More recently, culture-independent methods have allowed to easily evaluate fungal (both undesirable and technological) community dynamics in cheese (Jany and Barbier 2008). In this context, Florez and Mayo (2006) used PCR-DGGE to study the dynamics of Cabrales blue-veined cheese microbial communities. In artisanal Cabrales cheese, they found a fungal community composed of P. roqueforti, P. chrysogenum, or P. grise-ofulvum that did not change during ripening. Concerning mold surface-ripened cheeses, Arteau and Labrie (2010) recently studied the fungal communities of Camembert and Brie cheeses using a culture-independent T-RFLP method. In this study, Cladosporium cladosporioides, G. candidum, M. racemosus, P. camemberti, P. caseicola, P. chrysogenum, P. commune, and P. roqueforti filamentous fungal species were observed. Finally, use of both culture-dependent and culture-independent methods allowed Panelli and others (2012) to reveal the prominence of P. commune over other fungal species such as Cladosporium spp. in Taleggio cheese, whereas in Fossa cheese, the fungal community was dominated by Penicillium and Aspergillus species (De Santi and others 2010). Microorganisms play a central role during cheese-making, as for other fermented products. When the primary microbiota involved in acid production is mainly composed of starter lactic acid bacteria (LAB), the secondary microbiota includes salt-tolerant bacteria, yeasts, and filamentous fungi that mainly contribute to organoleptic properties of cheeses. The role of filamentous fungi is particularly important for both surface mold-ripened and core mold-ripened cheeses. Indeed, functional features and degradation of milk constituents, such as proteins and lipids, can lead to texture, flavor, and nutritional cheese quality improvement (Fox and McSweeney 2004). Among the species encountered in cheeses, P. camemberti, P. roqueforti, Mucorfuscus, M. lanceolatus, G. candidum, Fusarium domesticum, Sporendonema casei, Scopulariopsis flava, and S. fusca are the most common filamentous fungi used as ripening cultures (Hermet and others 2012; Ropars and others 2012). Filamentous fungi encountered in cheeses may originate from raw materials such as milk or may be introduced during cheese-making either from the environment or are deliberately inoculated using commercial ripening cultures. Delavenne and others (2011) showed that cow milk samples contained high fungal diversity with up to 15 species in a single sample, whereas a maximum of 4 or 6 different species were recovered in goat and sheep milks, respectively. These authors identified fungi belonging to the following genera: Aspergillus, Chrysosporium, Cladosporium, Fusarium, Penicillium, and Torrubiella. Vacheyrou and others (2011) concluded that most fungal species found in cow milk originated from a transfer from the stable to the milking pari and then to milk. Fungi occurring in raw milk or in the cheese-making environment are not very likely to persist and/or grow in cheese due to the biochemical composition of cheese and the hurdle effect exerted by the cheese microbiota. Industrial processes may also select certain fungal strains/species resulting in a fungal community shift. Moreover, raw milk heat treatment is often performed prior to manufacturing and eliminates most fungi in cheeses where the presence of molds is undesirable. Filamentous fungi contamination can also be detrimental to cheese quality, causing appearance defects, off-flav, and potentially toxic secondary metabolite production. P. commune, P. palitans, P. nalgiovense, and P. verrucosum species are the most frequent contaminants as suggested by Lund and others (1995) in a study based on spoilage microbiota of hard, semihard, and soft cheeses. Reports of P. verrucosum occurrence should, however, be treated with care since a high number of P. verrucosum isolates were reclassified as P. nordicum a species closely related to P. verrucosum that is much more adapted to salt and protein-rich substrates than P. nordicum (Larsen and others 2001). Fungal species cannot always be classified as either technological or contaminant species, since the same species can be used as ripening culture in some cases and be undesirable in other contexts (for example, P. roqueforti is the major ripening culture in blue-veined cheeses, whereas it is a contaminant in other cheeses). In some cases, the positive (technological species) or negative (spoilage organism or putative mycotoxin producer) role is not always clear and has to be investigated such as in the case of P. commune in Taleggio cheeses (Panelli and others 2012). In the near future, new high-throughput methodologies (such as pyrosequencing) will certainly provide better insight into cheese fungal communities. Among various filamentous fungi 438 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. Table 1 -Examples of mycotoxins found in cheeses including blue mold cheese. ND: Not determined. Cheeses Mold species Secondary metabolites Technical analysis Range References Comments Blue P. roqueforti Roquefortine, PR Thin-layer 3.4 /xg/g, ND, ND Scott and others toxin, chromatography (1977); Lund and Isofumigaclavine A others (1995) Blue P. roqueforti 100% penicillicacid, Thin-layer ND, ND van Egmond (1983) 37% PR toxin, chromatography patulin Blue-molded P. roqueforti Mycophenolic acid Thin-layer 90% 1 to 5 mg/kg Lafont and others chromatography (1979) Blue-molded P. roqueforti Patulin, penicillic acid, Thin-layer ND, ND, ND, 2.1 to 3.8 Erdogan and Sert tulum cheese PR toxin, chromatography mg/kg (2004) roquefortine Blue veined P. roqueforti Penicillicacid Thin-layer ND Moubasherand others Production of chromatography (1978) penicillic acid is to be due to P. car mum (Boysen and others 1996) Blue OTA LC-ESI-MS/MS 0.2 to 0.3 /xg/kg Dall'Asta and others (2008) Manchego P. roqueforti Roquefortine, Thin-layer >30/xg/kg Lopez-Diaz and others mycophenolic acid chromatography > 20/xg/kg (1996) Cheddar P. roqueforti Penicillicacid, patulin Thin-layer ND, ND Olivigni and Production of chromatography Bullerman (1977) penicillic acid is to be due to P. car mum (Boysen and others 1996) Patulin wasthe sole metabolite produced by P. paneum and P. carneum but was not produced by P. roqueforti (Nielsen and others 2006) Tulum cheese P. roqueforti Penicillicacid, PR Thin-layer ND, ND, ND Erdogan and others toxin, patulin, chromatography (2003) roquefortine Tulum cheese AFM1 LC-MS/MS 11 to 202 ng/kg Curses and others (2004) Blue veined P. roqueforti Mycophenolic acid Thin-layer 250to 5000/xg/kg Engel and others chromatography (1982) Blue mold P. roqueforti Roquefortine C LC-MS/MS 0.0008 to Kokkonen and others Mycophenolic acid 0.012 /xg/kg (2005a,2005b) 0.0003 /xg/kg Cheese P. camemberti Cyclopiazonicacid Thin-layer 0.5 to 1.5 /xg/g Le Bars (1979) chromatography Couda and Edam A. versicolor Sterigmatocystin Thin-layer 5 to 600/xg/kg Northolt and others chromatography (1980) Semhard cheese Penicillium sp. Patulin, ochratoxin High-performance 1 5 to 460 /xg/kg and Pattono and others liquid 1 to 262 /xg/kg (2013) chromatography Turkish cheeses Aspergillus sp. Aflatoxin Ml ELISA 58 to 850 ng/kg Aydemirand others (2010) Italian white P. camemberti Cyclopiazonicacid High-performance 0.02 to 0.08 mg/kg Zambonin and others cheeses liquid (2001) chromatography species encountered in cheese, the mycotoxin-producing ability for some has already been well established, while for others questions still remain. Mycotoxins in Cheese In the 1960s, the first studies related to mycotoxins in cheeses were published and led to the discovery of AFs. Other studies were performed in the 1970s with special interest on the occurrence of CPA in cheese. Since 2000 and the development of powerful analytical techniques such as LC-MS/MS, numerous studies have emerged on the subject. As stated above, various mold species can be isolated from cheeses and correspond to ripening cultures or contaminating molds (Taniwaki and others 2001) that can potentially produce © 2014 Institute of Food Technologists® Vol. 1 mycotoxins. Analytical methods for identifying and quantifying mycotoxins in foods and feeds have widely been developed and governmental agencies have establishe regulatory limits and guidelines for safe doses. Various mycotoxins have been detected in cheese (Table 1) and the most commonly identified are described below. Roquefortine C has been quantified in blue cheese (3.4 /xg/g) (Scott and others 1977), Manchego cheese (Lopez-Diaz and others 1996), and in commercial Finnish cheeses at levels reaching 12000 /xg/kg (Kokkonen and others 2005a, 2005b). MPA has been reported from different types of blue mold cheeses (1.5 mg/kg by Lafont and othersl979 and 5 mg/kg by Engel and others 1982) using thin-layer chromatography (TLC) and at much lower levels (0.0003 /xg/kg) in cheese using LC-MS-MS (Kokkonen 3,2014 • Comprehensive Reviews in Food Science and Food Safety 439 Filamentous fungi and mycotoxins in cheese. and others 2005a, b). Due to AFM1 occurrence in milk, this mycotoxin can also be present in cheese. Various studies have reported AFM1 in cheese (Giirses and others 2004; Aydemir and others 2010). On the other hand, very few studies have described AFG1 and AFB1 in cheese. In these cases, the mycotoxins were directly added to cheese and their stability was followed over time. Both AFB1 and AFG1 were stable in Swiss cheese (Lieu and Bullerman 1977). Similar studies were also carried out by inoculating A. jiavus and Aspergillus parasiticus mycotoxinogenic strains in cheese and both mycotoxins were again (Lie and Marth 1967). Sterigmatocystin (STC) was detected and quantified in the surface layer of A. versicolor contaminated hard cheeses at levels ranging from 5 and 600 /xg/kg (Northolt and others 1980). In white cheese, CPA was detected from 0.5 to 1.5 Aig/g (Le Bars 1979) and from 0.02 to 0.08 mg/kg (Zambonin and others 2001). Stott and Bullerman (1976) showed that when patulin was added to cheddar cheese, it was not detected after storage by TLC. However, in semihard cheeses, patulin and ochratoxin were detected at concentrations of 500 and 2620 /xg/kg, respectively (Pattono and others 2013). Since PR toxin was discovered (Wei and others 1973), no studies have described this mycotoxin in cheese as it is unstable and leads to PR imine (Scott and Kanhere 1979; Siemens and Zawitowski 1993). Mycotoxin cheese contamination can occur indirectly via milk contamination or directly by mycotoxin producing spoilage or technological filamentous fungi (O'Brien and others 2004). Concerning indirect contamination, the main concern is AFM1 in cheese. AFM1 in the final product is due to the addition of this mycotoxin into the milk by dairy animals (Eaton and others 1994; Gallo and others 2008) that have fed on moldy AFB1-contaminated feedstuffs. Indeed, AFB1 is converted into AFM1 by enzymes mainly present in the animal liver and then passed into their urine and milk. Many studies have reported AFM1 in various cheeses (Finoli and others 1983; Taniwaki and van Dender 1992; Giirses and others 2004; Kokkonen and others 2005b). Low storage temperature of cheese can induce a decrease in AFM1 concentration. The indirect contamination of milk with other mycotoxins, such as STC (van Egmond and Paulsch 1986), fumonisins (Maragos and Richard 1994), or CPA (Dorner and others 1994), has also been reported. Concerning this type of contamination in cheese, it is much more complex because it depends on the manufacturing processes used. Direct mycotoxin contamination can be exogenous via the presence of contaminating molds during cheese-making or endogenous by the presence of mycotoxin-producing commercial fungal cultures. Cheeses on which unintentional mold growth has occurred have been reported to contain mycotoxins. However, mold growth on the cheese surface does not automatically imply that mycotoxins are present in cheese; moreover, mycotoxin production can be quite different among mycotoxicogenic fungi (Moss 1991; Sweeney and Dobson 1998). In the case of direct contamination from an endogenous origin, several cheeses that intentionally have added mold species during cheese-making for appearance, texture and flavor development can be at risk. Although many cheeses use molds for ripening, the 2 main categories are: soft mold-ripened cheeses (such as "Camembert") associated with P. camemberti and blue-veined cheeses (such as "Gorgonzola," "St Nectaire," and "Cantal") associated with P. roqueforti. P. camemberti, and P. roqueforti have long been used in the manufacture of mold-ripened cheeses consumed throughout the world. P. roqueforti is an essential component of the microflora of many cheeses including Roquefort (France), Stilton (U.K.), Tulum (Turkey), Gorgonzola (Italy), Cabrales (Spain), Blauschimmelkase (Switzerland), and Danish Blue (Denmark) (Bullerman 1981; O'Brien and others 2004), while P. camemberti is associated with Camembert and Brie production. Both species are known for their potential mycotoxin production; however, due to cheese production practices and storage conditions, only low to very low concentrations have been observed (Bullerman 1981; Teuber and Engel 1983). Overall, numerous mycotoxins can be found in cheeses at highly variable concentrations depending on cheese origin. Ecological role of mycotoxins in cheese So far, the ecological roles of mycotoxins have not been completely elucidated but hypothesized functions in natural environments have been proposed. In the cheesemaking context, this role has not been studied to our knowledge. Therefore, ecological role hypotheses in this food matrix can only be proposed based on studies performed in other environments. Role in competition with other organisms. Mycotoxins are naturally occurring molecules that are thought to confer a selective advantage to the producer within complex ecosystems to communicate and compete with other organisms (Magan and Al-dred 2007). Their widespread nature and the preservation of these multigenic biosynthetic pathways suggest that secondary metabolites are used for survival functions in organisms (Demain and Fang 2000). The genes required for the biosynthesis of secondary metabolites are clustered. As stated by Calvo and others (2002), "logic suggests that their maintenance could only be selected for if the final product conferred some advantage to the producing organism." In this context, mycotoxins have been proposed to act as chemical signals for communication, a competitive weapon to defend the habitat, or to inhibit the growth and reproduction of competitors of the same trophic niche (Magan and Aldred 2007; Fox and Howlett 2008). The presence of competing microorganisms is essential to mycotoxin production, these hypotheses are demonstrated by the fact that serial transfers on culture media in laboratory conditions (lack of exposure to competitive and natural stressed conditions) result in the loss of AF production in strains of A.fiavus (Horn and Dorner 2001). Adverse conditions such as micro- and macronutri-ent depletion and competition with other fungi have a stabilizing effect in preventing AF production loss. In addition, early production of mycotoxins could allow molds to rapidly colonize the environment. In the cheese ecological niche, both spoilage fungi and fungal cheese cultures are in competition with other microorganisms including bacteria and yeasts. One study showed bacteriostatic effect of roquefortine on gram-positive bacteria, whereas gram-negative bacteria were not influenced at all. This mycotoxin affected bacterial respiration (Kopp-Holtwiesche and Rehm 1990). However, to our best knowledge, the effects of microbial competition on the production of mycotoxins have not yet been studied in cheeses or in cheese-mimicking models. Role in fitness. Another possible role of mycotoxin production is that it can be related to strain fitness. Mycotoxin production could correspond to a mechanism for adaptation to the environment. For instance, OTA biosynthesis by P. nordicum and P. verrucosum increases their fitness under high NaCl concentration conditions as exhibited in food products like dry-cured meat or brined olives (Schmidt-Heydt and others 2012). OTA 440 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. Table 2-Examples of toxic effects of some mycotoxins produced by Penicillium species. Mycotoxins Effects Models References Mycophenolicacid LD50 (Letal Dose) rats 2500 mg/kg IPLD50rats500mg/kg IV LD50 mice 700 mg/kg IP LD50 mice 450 mg/kg IV Mutagenic: chromosomes aberrations, Ames test Immunosuppressive ICso (inhibitory concentration) > 100 /xg/mL Roquefortine LD50 rats 1 520 mg/kg LD50 mice 169 mg/kg M LD50 mice 184 mg/kg F ntoxications Amestest-neurotoxic ICso 48 /xg/mL PR toxin LD50 mice 5.8 mg/kg IP LD50 rat 11.6 mg/kg IP LD50 rat 8.2 mg/kg IV Target organs: /"capillary permeability (heart, iver, lungs, kidneys damages) inhibits RNA and protein synthesis inhibits DNA polymerase mitochondrial respiration mutagenic (Ames+), carcinogenic Pen i ci 11 i c acid Synergistic with OTA In vivo rats and mice In vivo human Detroit 98 and Cirardi Heart In vitro Salmonella sp. In vitro human macrophage In vitro Caco cells In vivo rats and mice In vivo dogs In vitro Salmonella sp. In vivo day-old cockere In vitro Caco cells In vivo mice and rats In vitro E. coli In vitro rat liver mitochondria In vitro Salmonella sp. In vitro S. cerevisiae In vivo pig Wilson (1971) Engel and others (1982) Umeda and others (1977); Webnerand others (1978) Bentley (2000) Rasmussen and others (2011) Ueno and others (1978) Scott and others (1977) Puis and Ladyman (1988) Schochand others (1984) Wagenerand others (1980) Rasmussen and others (2011) Chen and others (1982) Moule and others (1976) Wei and others (1984) Ueno and others (1978) Wei and others (1979) Stoevand others (2001) contains a chloride atom within its molecule and the constant OTA biosynthesis ensures permanent excretion of chlorine out of the cell for Penicillia (Schmidt-Heydt and others 2012). The more constantly a strain produces OTA, the higher its growth rate will be under high NaCl conditions. Consequently, a strain competes better on a salt-rich habitat in comparison to a strain that produces low amounts of OTA (Schmidt-Heydt and others 2013). Cheese contains a fair amount of NaCl and it is also likely that the NaCl level may influence OTA production in cheeses. Role in detoxification. Finally, mycotoxin production could correspond to a waste elimination process. Indeed, mycotoxins could contribute to cell detoxification from the accumulation of reactive oxygen species (Reverberi and others 2010; Sanzani and others 2012). Fungi are aerobic organisms producing reactive oxygen species in their environment that can damage cell membranes and metabolic pathways. Generally, a cell can tolerate oxidative stress by increasing antioxidant synthesis and also secondary metabolite production, such as mycotoxins, which are under the control of reactive oxygen species (Reverberi and others 2010). In cheese, most LAB are able to produce hydrogen peroxide (H2O2) in the presence of oxygen. Hydrogen peroxide can then be converted to hypothiocyanate (HOSCN) via the reaction of hydrogen peroxide and thiocyanate through catalysis by lactoperoxidase that are both naturally present in milk. These reactive oxygen species, besides the fact that they have known antifungal activities (Schnurer and Magnusson 2005), may also, in turn, affect mycotoxin production in cheese-related fungi. We could also speculate that the production of MPA may be a route for detoxifying acetate, another well-known fungal inhibitor produced, for example, by nonstarter heterofermentative LAB such as Lactobacillus spp., however, no scientific evidence yet exists to support that these hypotheses are true. Mycotoxin production and their biological roles must be considered in an ecological context, which is not easy to elucidate in laboratory conditions. In fact, mycotoxigenic molds do not occur alone, they form part of an ecological niche in which diverse communities of microorganisms compete. A better understanding of the regulation and role of mycotoxin production during cheese-making could help to develop approaches to control mycotoxin contamination in cheeses. Nature and toxicity The usual routes for mycotoxin exposure are ingestion of mycotoxin-contaminated foods and feeds (Hayes 1980; Sweeney and Dobson 1998; Hussein and Brasel 2001; Bennett and Klich 2003), physical contact (CAST 2003), and inhalation (Bunger and others 2004) (Table 2). Mycotoxins can cause either acute or chronic diseases in animals and humans (Stoloff 1977; Pitt 2000) depending on the dose, ingestion duration, and animal species. Mycotoxins present in cheese due to fungal contamination. Aflatoxins. AFs derive from polyketide metabolism and consist of at least 16 structurally related furanocoumarins of which AFB1, AFB2, AFG1, and AFG2 are the most abundant (Huff-mann and others 2010). AFs are considered to be among the most significant classes of known mycotoxins in food and feed. They are produced by certain Aspergillus species. Highest AF levels are produced as a result of inadequate postharvest storage when high-moisture content and warm temperatures facilitate rapid mold growth. Therefore, the presence of AFs in animal feedstuffs (cereals, forages, and so on) (Garon and others 2006) can be an important contamination source for the dairy industry. As stated above, various studies have reported AFM1 in cheese. Concerning other AFs, AFG1, and AFB1 have been reported in cheese, but only after spiking samples with the mycotoxins or mycotox-inogenic strains (Giirses and others 2004; Aydemir and others 2010). Lethal dose (LD50) values range from 0.5 to 10 mg/kg according to the considered AF. Animal species respond differently in regard to their susceptibility to chronic and acute AF toxicity. During aflatoxicosis, the liver is the primarily affected organ, leading to bile duct proliferation, centrilobular necrosis, fatty acid infiltration, and hepatic lesions (Heinonen and others 1996; Bhatnagar and others 2002), which ultimately lead to liver cancer (Chao and others 1991; Sweeney and Dobson 1998). Other organs such as the kidney, lung, heart, gizzard, or spleen may also be affected (Yang and others 2012). AFs induce DNA damage, negatively affect damage repair, and alter DNA base compositions of genes. AF is associated with both toxicity and carcinogenicity in human and animal populations. Within the AF family, IARC (International Agency for Research ©2014 Institute of Food Technologists' Vol. 13,2014 • Comprehensive Reviews in Food Science and Food Safety 441 Filamentous fungi and mycotoxins in cheese. on Cancer 1993) has classified AFB1, as a group class 1 proven to cause cancer in human. Although many different cheese types can contain AFM1, it is almost invariably at such low levels that no danger to human health exists (Blanco and others 1988). But recent studies showed that high AF levels up to 850 ng/kg in regional Turkish cheese (Aydemir and others 2010). Therefore, the health risk should not be underestimated. Finally, due to AFM1 toxicity (classified by the IARC 2002, as class 2B, possible human carcinogen), many countries tightly monitor AFM1 levels in milk and milk products and, as stated above, the European Union has established a 0.05 /xg/kg maximal limit (EFSA 2004). Sterigmatocystin. STC is a xanthone and is structurally similar to AF (Hamasaki and Hatsuda 1977); however, it is less toxic (Sweeney and Dobson 1998). This mycotoxin is produced by a few Aspergillus species (Cole and Cox 1981; Versilovskis and De Saeger 2010). STC has indeed been detected in A. versicolor-contaminated cheeses (Scott 1989). Northolt and others (1980) reported STC occurrence during cheese (Gouda and Edam) ripening in warehouses in the Netherlands. STC was stable in contaminated cheeses for a 3-mo period at various temperatures (—18 to +16 °C) (van Egmond 1983). However, low temperatures (5 to 7 °C) used during ripening and storage should prevent A. versicolor growth and STC production. At these temperatures, molds that develop on cheese rather correspond to Penicillium species, which could not produce AFs or STC (Bullerman 1981). Nevertheless, STC was analyzed in cheese samples, produced in Latvia (8) and Belgium (13). Two cheeses (9.5%) were positive for STC with concentrations of 1.23 and 0.52 g/kg, respectively. Five (24%) samples contained STC above the limit of detection (0.03 g/kg) (Versilovskis and others 2009). STC is activated to sterigmatocystin-l,2-epoxide by liver enzymes, which binds covalently to DNA and induces various tumors (Baertschi and others 1989; Raney and others 1992). STC produces higher cytotoxicity than AFB1 (IC^o = 10 /xM) in A-549 cultures (lung cell line) (Palanee and others 2001). STC is carcinogenic in animal models and possibly carcinogenic to humans (IARC 1976). Citrinin. Citrinin can be produced by a dozen Penicillium species including a few P. camemberti strains and several Aspergillus species (Manabe 2001). P. citrinum, P. viridicatum, P. expansum, P. ver-rucosum, A. carneus, A. niveus, and A. terreus have all been reported to be the most important citrinin-producing fungi in nature, but other species such as Aspergillus candidus, have also been found to produce citrinin under laboratory conditions (Raper and Thorn 1949; Frisvad and others 2006). In cheese, P. citrinum and P. verruco-sum only are known to produce this mycotoxin. Citrinin appeared to be very stable in cheeses, including goat cheese, Saint Marcellin, Soignon (Bailly and others 2002). This mycotoxin is frequently associated with the presence of OTA both produced by P. verrucosum (Cole and Cox 1981; Ostry and others 2013). Citrinin was shown to be nephrotoxic on all performed animal experiments (Kitchen and others 1977). LD50 is 57 mg/kg for ducks, 95 mg/kg for chickens, and 134 mg/kg for rabbits. EFSA (2012) estimated based on a 90-d toxicity study in rats, a no-observed-adverse-effect level (NOAEL) of 20 /xg citrinin/kg b.w per day for nephrotoxicity. Citrinin can act synergistically with OTA to depress RNA synthesis in kidneys of mice (Sansing and others 1976). Citrinin also induces a loss of selective membrane permeability and cell death (possibly apoptotic) (Paterson and Lima 2010). Citrinin has a high /C50 in vitro on Caco-2 cells (83 /xg/mL) but can exhibit nephrotoxic effects in mammals (Rasmussen and others 2011). Citrinin exhibits immunotoxic effects, inducing lymphocyte proliferation inhibition in pigs. An immunosuppressive effect is estimated at /C50 equal to 10 mg/L (Wichmann and others 2002). On embryonic stem cells, citrinin induced apoptosis via ROS and mitochondria-dependent pathway (Chan 2007). Ochratoxin. OTA is produced by several Aspergillus (Benford and others 2007) and Penicillium species. In temperate and cold climates, P. verrucosum is the major producer of OTA in cereals. P. verrucosum is recovered mainly from plant-derived material; P. nordicum is generally recovered from meat and cheese products (Cabanes and others 2001). OTA has been detected in a wide range of foods and feed, including cereals, bread and beer (Visconti and others 2010) as well as meat and cheese products (Gareis and Scheuer 2000). OTA was originally discovered as an Aspergillus ochraceus metabolite during a large metabolite screening procedure (Van der Merwe and others 2000) and shortly these after recognized as a potent nephrotoxin, in particular in swine and poultry. Among the numerous ochratoxin analogs (at least 20), OTA is produced at the highest levels and is the most toxic. Although OTA transfer into milk has been demonstrated in other animal species, in ruminants, the resident rumen microflora decreases bioavailability through hydrolysis of OTA to OTa (Skaug 1965). Nevertheless, OTA has been detected in different commercial samples of blue-veined cheeses (Gorgonzola and Roquefort), although only at very low levels (DallAsta and others 1999). The same authors showed that 32.6% of blue cheeses inoculated with only P. roqueforti (non-OTA producers) were actually contaminated with OTA (DallAsta and others 2008). OTA levels detected in different cheese portions demonstrated a nonhomogeneous contamination highly associated with the blue—green veins of the cheese where fungal growth occurs. Moreover, the observed levels increased during storage, indicating that contamination was not derived from contaminated milk. This implied that OTA-producing fungal strains may have contaminated the cheese during the manufacturing process. Ochratoxins have been known for some time to be a major etiological agent in endemic nephropathy (Radic and others 2008) and, by analogy, they have been considered to be associated with the chronic progressive kidney disease in humans known as Balkan endemic nephropathy (BEN) and urinary tract tumors (Castegnaro and others 1997; Tanchev and Dorossiev 1991; O'Brien and Dietrich 1991). Initiation of carcinogenic process is the covalent binding of OTA to DNA, forming DNA adducts (Lebrun and others 2005). OTA toxicity involves mitochondrial respiratory function, OTA interferes in modulating CYP 450 enzymes, COX, and lipox expression (Pfohl-Leszkowicz and Man-derville 2006). OTA induced Ca2+ increase into the cells, tRNA-synthase inhibition, accompanied by reduced protein synthesis and enhanced lipid peroxidation via the generation of free radicals (Hohler 2007). Intake of OTA from cheese seems to be of limited importance for the general population in comparison to potential OTA intake from cereals and their derived products. Mycotoxins potentially produced by the commercial cheese mold P. roqueforti. Roquefortine C. Roquefortine C is the most common metabolite produced by Penicillium genus members (25/58 species) (Frisvad and others 1998). Moreover, according to Scott (2004) and Vi-nokurova and others (1981), within the P. roqueforti species, virtually all studied strains produce roquefortine C and its precursor roquefortine D. 442 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. Roquefortine C was previously incriminated in intoxication cases in cattle in Germany (Haggblom 2001; Auerbach and others 1990). To date, there are no toxicity studies performed on laboratory animals that meet good laboratory practice standards. Intraperitoneal (IP) lethal dose 50 (LD50) were 169 mg/kg and 184 mg/kg in male and female mice, respectively (Arnold and others 1998). Wagener and others (1987) reported paralytic activity in day-old cockerels exposed to roquefortine C. In vitro, roquefortine C exhibits an inhibitory concentration 50 (/C50) of 48 /xg/mL on Caco-2 cells (2011Rasmussen and others 1980). However, no recent data are available to estimate human exposure (Table 2). To date, roquefortine C has never shown any mutagenic activity by the Ames test (Schoch and others 1984). Roquefortine C is known to be neurotoxic (Frisvad and others 2007). There are a number of dog toxicosis cases linked to roquefortine C (Puis and Ladyman 1988), but penitrems or other compounds might have also been present. Mycophenolic acid. Few studies reporting the presence of MPA produced by P. roqueforti in cheeses have been carried out on blue cheeses (Lafont and others 1979; Engel and others 1982; Lopez-Diaz and others 1996). MPA is a phthalide and although it is a mycotoxin with potential toxicological implications, it is also a very important drug with several potential applications (Bentley 2000) . It is presently used as an immunosuppressant in kidney, heart, and liver transplant patients to avoid organ rejection. Toxicity for mammals appears to be low: LD50 in rats is 2500 mg/kg (oral pathway) and 500 mg/kg by intravenous (IV) pathway in mice the LD50 is 700 mg/kg and 450 mg/kg IV (Wilson 1971). Chronic tests of 80 and 320 mg/kg daily oral doses over 1 y did not cause apparent toxicity signs in rabbits (Adams and others 1975). It is not unlikely that MPA could lower the immune system if ingested often enough thereby paving the way for bacterial infections (Bentley 2000; Frisvad and others 2007). Mutagenic activity was evaluated and showed that this mycotoxin was able to induce chromosome aberration (Umeda and others 1977) but was negative for the Ames test (Webner and others 1978). Concerning MPA, with the exception of immunotoxicological effects, toxicological data are scarce. Other mycotoxins. In regards to other mycotoxins produced by P. roqueforti in cheese, only few toxicological data are available. Isofumigaclavines (A and B), festuclavine, and agroclavine are alkaloids potentially produced by P. roqueforti. They belong to the ergot alkaloid mycotoxin family (Brase and others 2009). Isofumigaclavines A and B have been detected at low levels in commercial blue cheeses (O'Connor and O'Brien 2000). Isofumigaclavine B corresponds to the isofumigaclavine A hydrolysis product. These toxins were reported by Ohmomo and others (1975, 1977) and by Kozlovskii and others (1979). Agroclavine increased natural killer (NK) cell activity under nonstress conditions in vivo and in vitro, thus exhibiting an immunostimulatory effect (Starec and others 2001) . PR toxin (Table 2), another known mycotoxin produced by P. roqueforti, is unstable in the cheese environment and is converted to PR imine, an unstable and less toxic molecule. PR toxin can also be converted to PR amide in the presence of basic and neutral amino acids (Chang and others 1993). Other secondary metabolites related to PR toxin synthesis by P. roqueforti, such as eremofortins A, B, C, and D, have also been observed. Mycotoxins potentially produced by the commercial cheese mold P. camemberti. Cyclopiazonic acid. CPA, an indole tetramic acid, is produced by P. cyclopium, some P. camemberti strains, and several Aspergillus species (Pitt and others 1986). Spahr and others (2000) considered that CPA is one of the few mycotoxins that can be transferred to milk. CPA is a calcium-dependent ATPase inhibitor and induces ion transport alteration across cell membranes (Riley and Goeger 1992; Burdock and Flamm 2000). Clinical intoxication signs include weight loss, anorexia, diarrhea, dehydration, pyrexia, ataxia, immobility, and death in dogs, rats, pigs, sheep, and chickens (Bry-den 1991). The possible role of CPA in aflatoxicosis cases should be explored as it could have a synergistic effect. Target organs are kidneys and gut tract in mammals. In acute toxicity, LD50 is 64 mg/kg in female mice and 12 mg/kg in chickens. In humans, CPA is suspected to be responsible for acute mycotoxicosis (named "kodua") inducing nerve troubles. There is limited information and only few data concerning mycotoxin toxicities associated with P. roqueforti and P. camemberti. Not enough is currently known about the actual toxicity of mycotoxins found in cheeses and the toxicity of metabolites that can co-occur with mycotoxins. Methods for qualitative and quantitative analysis (Table 1) Due to mycotoxin toxicity, their frequent occurrence in foodstuffs, and driven by regulatory authorities worldwide, there is a distinct need for highly selective and accurate methods to identify and quantify mycotoxins in diverse agricultural and biological foods and feeds. For example, the maximum AF levels for animal feed are 50 and 20 ppb in the United States and Europe, respectively. There are no published guidelines for mycotoxin analysis. Numerous methods have been developed to analyze the large structural diversity of mycotoxins occurring in different matrices (Zöllner and Mayer-Helm 2006; Malik and others 2010). Chromatography. Concerning cheese, classical analytical methods for mycotoxin detection and/or quantification include TLC, high-performance liquid chromatography (HPLC), gas chromatography (GC), and mass spectrometry (MS) coupled with chromatography. One of the disadvantages for GC analyses for mycotoxin detection is the necessity to derivatize samples, a time-consuming and prone to error step; as a result, GC methods are less frequently used. In recent years, most of these methods have been coupled with immunoaffmity techniques to simplify extraction and improve mycotoxin recovery and measurement from foodstuffs (Bony 2000; Chu 2000; Miraglia and Brera 2000; Stroka and others 2000a, 2000b; Anon 2001; Josephs and others 2001; Krska and Josephs 2001; Turner and others 2009). As cheese has a high fat content, hexane was simultaneously used to remove lipid compounds for the partitioning and extraction steps. However, some lipids persist during extraction; therefore, pure ACN, or methanol is also used to dissolve the sample in the final step. HPLC is widely used to quantify different mycotoxins in cheeses, including AFM1 (Kamkar 2008). HPLC reference methods have been developed for most of the major mycotoxins and are quite sensitive with reasonably low detection levels. For example, patulin and ochratoxin are detected in traditional semihard cheeses at levels ranging from 15 to 460 and 1 to 262 /xg/kg, respectively (Pattono and others 2013). LC/MS or GC/MS have excellent detection sensitivity and selectivity. The better sensitivity of new MS instruments, such as those equipped with ion trap analyzer, in combination with better GC columns, can lead to use the GC/MS technique to significantly low patulin levels for example. However, the necessity of derivatization makes exact quantification more adapted for ©2014 Institute of Food Technologists' Vol. 13,2014 • Comprehensive Reviews in Food Science and Food Safety 443 Filamentous fungi and mycotoxins in cheese. LC rather than GC/MS, unless a stable isotope-labeled patulin is available as internal standard. LC/MS is an analytical technique independent of molecular weight so that mycotoxins can be detected and quantified in complex matrices like cheeses. However, efforts must be made to develop a wide range of protocols taking into account analyte polarity diversities and different ionization capacities. Liquid chromatography tandem mass spectrometry (LC/MS/MS) is a suitable technique as it can be used to simultaneously analyze all mycotoxins produced by 1 fungus (Xu and others 2006). The HPLC/MS/MS performance for cheese samples was defined by recovery, repeatability, and limit of quantifications (LOQs). No regulatory limits exist for cheese, so LOQs cannot be evaluated. Recently, Kokkonen and others (2005b) proposed an easy LC/MS/MS method for quantifying several mycotoxins in blue and white mold cheeses; the proposed method detected 9 mycotoxins produced by Aspergillus and Penitillium species. Mycotoxins (AFs, OTA, MPA, and roquefortin C) were quantified from cheese samples at low levels ranging from 0.6 to 5.0 /xg/kg. Liquid chromatography tandem mass spectrometry (LC/MS/MS) analytical methods are usually used to confirm data since they require expensive instrumentation and highly trained staff. Recently, the natural occurrence of masked mycotoxins (conjugated toxin) has been reported, and their detection will require even more, selective, and sensitive methods (Berthiller and others 2005). Bioassay techniques. Bioassays have become increasingly useful for mycotoxin detection (Watson and Lindsay 1982; Yates 1986) as a rapid screening procedure before chemical analysis. Screening assays used are in majority immunochemical methods including enzyme-linked immunosorbent assays (ELISAs), lateral flow devices (LFDs), dipstick tests, fluorescence polarization immunoassay (FPIA), immunofiltration assays, and more recently, biosensor assays. Immunochemical methods can also be considered as convenient and sensitive alternatives for detecting various mycotoxins in cheese (Chu 1991; Mak and others 2010; Mohajeri and others 2013). For example, both direct and indirect ELISA strategies have clear advantages (cost and time efficiency) and only few limitations including cross-reactivity and immunoreactivity of the primary antibody resulting in nonspecific signals (Turner and others 2009). There are a lot of commercially available bioassay kits. For example, available ELISA kits include, Agraquant (RomerLabs Diagnotica GmbH, Tulln, Austria) and Veratox (Neogen Corporation, Lesher Place Lansing, Mich., U.S.A), the LFD kit "RIDAQUICK" (R-Biopharm AG, Darmstadt, Germany) are avaible in Europe but no bioassay kit is specific to cheese samples. Additionally, not all mycotoxins (including patulin) can be measured by these methods. As presented, numerous techniques are currently available to analyze and detect a wide spectrum of mycotoxins. However, powerful methods still need to be developed to detect and quantify multiple toxins from a single matrix. Due to the high complexity of the cheese matrix, LC/MS/MS currently seems to be the most sensitive method to detect and quantify mycotoxins in cheese. Mycotoxin Biosynthetic Pathways As stated above, mycotoxins correspond to a structurally diverse group varying from simple C-compounds to more complex substances (Brase and others 2009). These compounds are part of fungal secondary metabolism, and interestingly, unlike primary metabolism-associated genes, a large number of the mycotoxin production-associated genes have been shown to be clustered on a single genetic locus (Hoffmeister and Keller 2007). This feature, combined with the fact that some of the most important mycotoxins are synthesized by polyketide synthases (PKSs) (AFs, fumonisins, ochratoxins, and zearalenone), nonribosomal peptide synthetases (NRPS) (gliotoxin, enniatin, and HC-toxin), or PKS-NRPS hybrid proteins (CPA), allowed to identify biosynthetic pathways and the involved genes in various species (Figure 1). Biosynthetic pathways in undesirable species For cheese-contaminating fungi, the pathways associated with mycotoxin production have been studied in depth from a genetic point of view. Aflatoxins/sterigmatocystin. The biosynthetic genes and regulation involved in AF and STC synthesis have been extensively documented in Aspergillus species (Keller and Hohn 1997; Minto and Townsend 1997; Payne and Brown 1998; Brown and others 1999; Yu and others 2004; Yabe and Nakajima 2004; Georgianna and Payne 2009; Huffman and others 2010) including A. flavus (pathway discovered in this AF producing cheese contaminant) (Barrios and others 1998; Mogensen and others 2010) and A. parasiticus (Yu and others 2004). In regards to cheese contamination and as previously stated, detection of these mycotoxins in cheeses is typically due to indirect cheese contamination via feed-stuffs (this is the case for AFs and AFM1 produced by Aspergillus species on feedstuffs that are further detected in milk and cheese) or postcontamination (mostly relates to STC in cheese due to A. versicolor mold growth on cheese surfaces (Northolt and others 1980). Among the studied chesse-related and noncheese-related species, gene organization can differ and may indicate different evolutionary histories; this is the case of A. nidulans and A. flavus (Cary and others 2009). For cheese-related Aspergillus spp., the AF biosynthetic pathway was shown to contain 25 genes clustered within a 70-kb region and involves at least 23 enzymatic reactions (Yu and others 2004; Huffman and others 2010). The starter units correspond to acetate and malonyl-CoA that are converted sequentially to hexanoate by a fatty acid synthase (FAS) and to nor-solorinic acid (common to both AF and STC biosynthesis) by a PKS (Crawford and others 2006, 2008a, 2008b, 2009). Expression of all pathway genes, except for aflS (formerly known as aflj and also involved in AF regulation), has been shown to be positively regulated by a transcription factor, aflR. It is known that STC is the final metabolic product of the biosynthetic pathway in A. nidulans, while this mycotoxin acts as a precurseur for AF biosynthesis in other Aspergillus species including A. flavus and A. parasiticus (Yu and others 2004; Huffman and others 2010). These biosynthetic pathways have been extensively reviewed by Yu and others (2004) and Huffman and others (2010), and the detailed gene cluster is provided in Figure 1 for A. flavus, a species involved in indirect mycotoxin contamination in milk and cheese. Ochratoxin. A pathway for OTA biosynthesis was recently reviewed by Huffman and others (2010) and includes a polyketide synthase for PK dihydroisocoumarin synthesis, a methyltransferase, a P450-type oxidation enzyme for carboxyl group formation at C7, a NRPS to catalyze ligation between phenylalanine and PK and a halogenase to incorporate the chlorine atom. To date, characterization of part of the OTA biosynthetic gene cluster has been described in the cheese contaminating species Penicillium nordicum (Karolewiez and Geisen 2005; Geisen and others 2006). Geisen and others (2006) sequenced a 10-kb-length fragment containing 3 genes corresponding to a partial PKS (otapksPN), a complete NRPS homolog (npsPN) and a complete alkaline serine protease homolog (aspPN). Moreover, O'Callaghan and others 444 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. Sterigmatocystin / Aflatoxin-A flavus aflY aflX aflW aflV aflK aflQ aflP aflO afll aflL aflG aflN aflM aflE aflj aflH aflS aflR aflB aflA D-C a/ZD aflC aflT aflU aflF Citrinin - M. aurantiacus ctnD cdnE orf6 orfl ctnA Ochratoxin - P. nordicum asp npsPN otapksPN OCxTxx Roquefortine C - P. chrysogenum rpt gmt P450 N-hydrola.se hydrolase rds ......^^x^^ >■■■■ H " N^ Mycophenolic acid - P. brevicompactum mpaH mpaG mpaF mpaE mpaD mpaC mpaB mpA PR toxin - P. roqueforti ari I (_ ......i=>..... A- Cyclopiazonic acid - A. flavus cpctO cpaD cpaS pksCT or/5 ctnF or/7 ctnR or/8 ctnG ctnH OH 0 Figure 1 -Known fungal gene clusters associated with production of mycotoxins encountered in cheese. Based on the following Cenebank accession numbers: sterigmatocystin/aflatoxin (NW_002477243), citrinin (AB243687), ochratoxin (AY557343), roquefortine C (AM920436), mycophenolic acid (HQ731031), PR toxin (L05193), and cyclopiazonic acid (JN712210). (2013) recently described a 3-gene cluster encoding an oxidore-ductase (OtaE), a PKS (OtapksPV), and a transporter protein (OtaT) involved in OTA biosynthesis in P. verrucosum, another cheese contaminating species. Other Penitillium species carry inactive homologs for these genes (P. nalgiovense, also isolated from cheese) or do not carry these genes. However, detailed genetic and biochemical studies are still required to elucidate the molecular mechanisms and confirm the proposed OTA biosynthetic pathway. Citrinin. Citrinin is known to be produced by some cheese contaminating Penitillium species including P. titrinum, P. verrucosum, and P. expansum (Ei-Banna and others 1987; Ku-rata 1990; Cabanes and others 2007). However, to this date, citrinin biosynthetic pathway has not been characterized in these species but can be extrapolated from putative citrinin biosynthetic 16-gene cluster recently described in Monascus purpureus (Shimizu and others 2005, 2007) and M. aurantiacus (Li and others 2012). In these species, the pathway genes were identified to be part of a putative 43-kb biosynthetic gene cluster including a polyketide synthase gene (pksCT), a transcriptional activator gene (ctnA), a membrane transporter gene (orf5), a fatty acyl-CoA synthetase (ctnl), an oxygenase (orf3), 3 genes for post-PKS-modifying en- zymes, oxidoreductase genes (ctnD, orf4), and 3 dehydrogenases (ctnE, orfl, ctnH) among others (Shimizu and others 2005; Li and others 2012). Biosynthesis of citrinin, originating from a polyketide, often co-occurs with OTA. Further studies are still required to fully understand the regulation of citrinin biosynthesis. Biosynthetic pathways in cheese ripening species Concerning the potentially mycotoxigenic fungal species used as ripening cultures for cheese, P. roqueforti and P. camemberti, biosyn-thethic pathway-associated genes have not been described to this date, with the exception of the PKS involved in PR toxin formation. However, pathways have been identified in other mold species. These studies can be used as a base for studying mycotoxin-related genes in fungi used as ripening cultures in cheesemaking. Roquefortine C. Studies have shown that roquefortine C is an intermediate of the oxaline and epi-neoxaline pathway in P. glandicola (Reshetilova and others 1995) and P. tulipae (Overy and others 2005), respectively. In 2011, a biosynthesis pathway was proposed in P. chrysogenum (Garcia-Estrada and others 2011). In this study, by using gene silencing, the authors indicated that a single-gene cluster containing 6 genes involved in the biosynthesis and secretion of 2 mycotoxins: roquefortine C and meleagrin. ©2014 Institute of Food Technologists' Vol. 13,2014 • Comprehensive Reviews in Food Science and Food Safety 445 Filamentous fungi and mycotoxins in cheese. The cluster organization showed the presence of genes encoding a nonribosomal cyclodipeptide synthetase, a prenyltransferase, and an N-hydroxylase leading from histidine and tryptophan to roquefortine D. Under the action of a monooxygenase, the latter substance is metabolized into roquefortine C. The production of meleagrin is obtained via the actions of 3 enzymes leading sequentially to glandicoline A and glandicoline B before meleagrin. More recently, 2 publications refined the biosynthetic pathway and reassigned ascribed synthetases. According to Hazrat and others (2013), the NRPS RoqA is indeed responsible for the production of histidinyltryptophanyldiketopiperazine (HTD). The molecule can be transformed by the sequential action of RoqR followed by RoqD, or conversely by the action of RoqR and then RoqD, to lead to roquefortine C. This is done without the action of RoqM contrary to what was proposed by Garcia-Estrada and others (2011). The same research group showed further branching in the metabolic pathway by the characterization of roquefortine F and neoxaline as well as the identification of Roquefortine L and its degradation products (Ries and others 2013). Noteworthy, in P. roqueforti and in the cheese context, only roquefortine C has been observed, thus suggesting that a part of the pathway leading from roquefortin C to meleagrin might either be absent or inactive in this species. Mycophenolic acid. As described by Birch and others (1957), MPA consists of an acetate-derived phthalide nucleus and a terpene-derived side chain indicating the involvement of polyke-tide and isoprenoid pathways. Proposed biosynthetic pathways (Bentley 2000; Regueira and others 2011) indicate that acetyl-CoA, malonyl-CoA (x3), and S-adenosyl methionine (SAM) would form 5-methylorsellinic acid (5-MOA) through the action of a polyketide synthase. Following phthalide formation, a reaction with farnesyl diphosphate yields 6-farnesyl-5, 7-dihydroxy-4-methylphthalide. Degradation of the side chain via oxidative cleavage leads to demethylmycophenolic acid (DMPA), which is finally methylated by a SAM: DMPA O-methyltransferase to yield MPA. This last step was confirmed by the purification and characterization of SAM: DMPA O-methyltransferase in P. stoloniferum. Regueira and others (2011) identified and characterized the MPA biosynthetic gene cluster (8 putative genes) in P. brevkompactum. Gene cluster determination was performed by detection of a resistance gene encoding an inosinate dehydrogenase (IMPDH) and deletion of the mpaC gene encoding a PKS, which abolished the production of metabolites (including MPA), associated with MPA biosynthesis. In the described cluster, the mpaG gene encodes the SAM: DMPA O-methyltransferase. Recently, Hansen and others (2012) showed through the creation of a chimeric MpaDE protein, consisting of a cytochrome P450 (encoded by the mpaD gene) and a hydrolase (encoded by the mpaE gene), that the step following 5-MOA production in the pathway could be carried out by a natural-fusion MpaDE enzyme. PR toxin. PR toxin metabolic pathways have been proposed in P. roqueforti (Moreau and others 1980; Chalmers and others 1981; Jelen 2002). In the pathway, Chalmers and others (1981) proposed that the pathway precursor, acetate, is sequentially converted to mevalonate, dimethylallyl pyrophosphate (DMAPP), and farnesyl pyrophosphate (FPP). Cyclizations by the aristolochene synthase lead to aristolochene, the precursor for eremophilane-type sesquiterpenes, with germacrene A as an intermediate, as demonstrated by Calvert and others (2002). PR toxin arises from aristolochene via a synthetic pathway including eremefortin B, A, and C. Eremefortin C is finally converted to PR toxin. The ac-etal form of eremofortin C leads to eremofortin D (Moreau and others 1980). Recently, Brock and Dickschat (2013) used a combination of CLSA/GC-MS and 13C NMR to obtain an in-depth view of the biochemical pathway and described several unknown sesquiterpenes leading from FPP and aristolochene as well as side products. PR toxin can be degraded in PR imide, PR amide (eremofortin E) (Chang and others 1993), and PR acid (Chang and others 1996). Concerning the genes and enzymes associated with this pathway, until recently, the genes involved in this pathway had not been described, with the exception of aristolochene synthase, a sesquiterpene cyclase, implicated in PR toxin biosynthesis in P. roqueforti (Proctor and Hohn 1993). The 1129-bp-long Aril gene, exhibiting 2 introns, encodes a 342-amino acid protein with a calculated molecular weight of 39200 (Proctor and Hohn 1993). The protein was purified by Hohn and Plattner (1989) and was shown to be Mg2+-dependent. The crystal structure (2.5-A resolution) was determined by Caruthers and others (2000) and confirmed that the enzyme is responsible for the cyclization of the FPP, to form aristolochene. The enzyme eremefortine C, responsible for the conversion of eremefortin C to PR toxin, was studied by Chang and others (1985). Recently, Hidalgo and others (2014) cloned and sequenced a 4-gene cluster from P. roqueforti. The cluster included the aril gene and silencing of the other genes (encoding for putative oxidoreductases and an alcohol dehydrogenase) demonstrated their involvement in the PR synthetic pathway. Interestingly, silencing of the PR toxin production caused a large increase in MPA synthesis suggesting a cross talk between these 2 mycotoxin pathways. Cyclopiazonic acid. CPA, produced by P. camemberti in cheese, derives from tryptophan, acetyl CoA, malonyl CoA, and DMAPP, as a prenyl donor, via the formation of 2 intermediates, cycloacetoacetyl L-tryptophan (cAATrp) and yS-CPA (Holzapfel and Wilkins 1971). As stated above and as for the other mycotoxins described in this section, to date no complete biosynthetic cluster associated to cheese ripening cultures has been elucidated. However, for CPA, gene clusters were described in A. flavus (cheese-contaminating species) and A. oryzae and can be used as a base to elucidate the corresponding pathway in P. camemberti. In A. flavus, the CPA biosynthetic pathway consisted of 3 genes (cpaS, cpaD, and cpaO) and was situated next to the AF biosynthesis gene cluster (Chang and others 2009b). In A. oryzae, the ability to form CPA is strain-dependent (Chang and others 2009a); while the cluster is complete in the NBRC 4177 strain, the RIB40 strain is unable to form the mycotoxin due to a truncation of the PKS—NRPS (Shinohara and others 2011). Studies performed by Liu and Walsh (2009a, 2009b) demonstrated that cpaS encodes a PKS—NRPS responsible for cAATrp formation, while cpaD encodes a cAATrp-dimethylallyltransferase, leading to yS-CPA. In this cluster, the cpaO gene encoding a monoamine oxidase would lead to CPA, thus correlating well with the previously proposed CPA biosynthetic pathway. The study by Chang and others (2009b) showed that disruption of this gene (named maoA in the study) abolished the production of CPA in A. flavus. Concerning mycotoxin biosynthetic pathways, the already described gene clusters as well as the growing access to complete fungal genomes will allow identifying the corresponding pathways in cheese-related fungi. Moreover, the presence of mycotoxin-related genes in the subphylum Mucoromycotina, so far considered as a nonmycotoxin-producing group, could be evaluated. Factors Modulating Mycotoxin Biosynthesis in Cheese It is generally accepted that fungal growth and mycotoxin production in food is influenced by multiple factors including abiotic 446 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese... 1 Figure 2-Main biotic and abiotic factors modulating mycotoxin production in cheese. and biotic factors (Filtenborg and others 1996). In cheese, biotic factors modulating mycotoxin production include the intrinsic capacity of a given mold species or strain to produce mycotoxins, its physiological state and the interactions that may occur between this organism and the other members of the cheese microbiota (Figure 2). The main abiotic factors that are likely to modulate mycotoxin production in cheese can be divided into manufacturing and environmental factors such as temperature, relative humidity (RH), ripening time, and storage atmosphere used during cheesemaking, as well as physicochemical factors of the cheese itself, including chemical composition (carbon and nitrogen sources, C/N ratio, NaCl content. . .), pH, water activity (aj), and redox potential (E°) (Figure 2). Because over 1000 cheese varieties exist worldwide, it is difficult to draw a general picture of abiotic factor variations in cheeses. For example, in surface-ripened cheeses comprising soft, semihard, and hard types, cheeses are ripened at an RH ranging from 90% to 80% to 85% and at temperatures of 4 to 20 °C for 14 to 540 d depending on cheese type (Robinson 1995). Moreover, pH, NaCl, and aw at the cheese surface layer may vary at the end of ripening from 7% to 5.6%, 2% to 0.6%, and 0.98% to 0.95%, respectively (Robinson 1995). The physical and chemical characteristics of cheeses are also subjected to both spatial and temporal changes during ripening. First, environmental conditions used in cheese manufacture, such as RH and temperature in ripening rooms and biological activities (glycolysis, proteolysis, and lipolysis) of cheese microorganisms, affect physicochemical cheese characteristics during ripening. Moreover, cheese is also characterized by a heterogeneous physicochemical composition with gradients in pH, NaCl, moisture, aw, O2 level, and so on from the surface to the core. For example, in the core and surface of the blue-veined cheese Danablu, pH, NaCl, and aw gradients ranged from 4.6% to 4.4%, 1% to 6.0% and 0.99% to 0.87%, respectively, after 1 week of ripening while they ranged from 6.4% to 5%, 2% to 4.5% and 0.94% to 0.91%, respectively, after 5 wk of ripening (Cantor and others 2004). Also in Danablu cheese, van den Tempel and others (2002) indicated that average oxygen saturation decreased from 63% to 0.1% at the center and from 69% to 0.1% at the cheese edges from 1 to 19 wk. It is also interesting to note that in the same study, after 3 wk of ripening, oxygen was not detected in the cheese, except in the 0.25 mm surface layer and in apparent small air pockets within the cheese containing about 3% oxygen. Effect of nutritional factors on mycotoxin production While it is generally admitted that the chemical composition of food exerts a significant effect on mycotoxin production, there is controversy in the literature on whether cheese is a food matrix that is well suited for mycotoxin production. In earlier and much more recent studies, cheese was considered a poor substrate for mycotoxin formation because of its low C/N ratio (Scott and ©2014 Institute of Food Technologists' Vol. 13,2014 • Comprehensive Reviews in Food Science and Food Safety 447 Filamentous fungi and mycotoxins in cheese. others 1977; Engel 1978; Olivigni and Bullerman 1978; Finoli and others 2001) and low ripening temperature (Marth 1979). Finoli and others (2001) investigated the ability of P. roqueforti fungal ripening cultures to synthesize roquefortine C and PR toxin in vitro in 10% reconstituted skim milk and yeast extract sucrose (YES) medium, which is characterized by a high C/N ratio due to its high concentration of sucrose (150 g/L). They also analyzed, using a HPLC-UV method, roquefortine C and PR toxin contents in 30 blue-veined cheeses from where the strains originated. The cheeses contained very small amounts of roquefortine C (<1.44 mg/kg) and no PR toxin was detected. Roquefortine C and PR toxin were generated in very low quantities (<3.08 mg/L), if any, in milk, whereas all the studied strains produced these 2 mycotoxins in YES medium with up to 8.44 and 60 mg/L for roquefortine C and PR toxin, respectively. In contrast, Kokkonen and others (2005a), using LC/MS/MS, showed that roquefortine C was present in 11 blue mold cheese samples purchased from Finnish supermarkets at concentrations ranging from 0.8 to 12 mg/kg that were about 10 times higher than those reported by Finoli and others (2001). However, it is difficult to compare the concentrations of roquefortine C reported in the latter 2 studies because different methods were used for quantification. More importantly, different cheese varieties likely obtained with different strains of P. roqueforti and ripening conditions were analyzed. More studies on a large number of blue cheese samples and varieties are therefore needed to clarify whether or not cheese is well suited for roquefortine C production and the relationships existing between roquefortine C content, manufacturing practices and the toxigenic potential of P. roqueforti strains. This observation is also true for other mycotoxins such as CPA and MPA. Kokkonen and others (2005b) also studied the ability of P. crus-tosum, P. nordkum, and P. verrucosum to produce mycotoxins in 3 different substrates: YES agar as well as cheese and bread analogs. P. nordkum, a frequent contaminant encountered in cheese, only synthesized moderate or undetectable amount of OTA and none of the substrates favored its production. In contrast, the substrates had a significant impact on the secondary metabolism of the other tested fungi. P. crustosum produced roquefortine C on all the substrates, with the highest amount on the cheese analog, while it only synthesized penitrem A on the cheese analog that was a favorable medium for the production of this mycotoxin. The authors hypothesized that the ability of P. crustosum to produce this secondary metabolite was linked with the high protein content of cheese because amino acids are required for the synthesis of tremorgenic mycotoxins. In contrast, only 1 out of 8 P. verrucosum strains produced OTA on the cheese analog, while high levels of ochatoxin A and citrinin were produced on the bread analog. Interestingly, fresh goat cheese was also a more favorable substrate for citrinin biosynthesis by P. citrinum and P. expansum than YES medium with up to 600 mg citrinin per kg of cheese produced after 10 d of incubation at 20 °C (Bailly and others 2002). However, as discussed by Kokkonen and others (2005b), the presence of precursors in the substrate cannot solely explain the ability of a given species/strain to produce higher levels of mycotoxins, and the factors and their interrelations that modulate mycotoxin production in food are mostly unclear. For example, in the study of Finoli and others (2001), roquefortine C production in cheese or milk by P. roqueforti was poor, despite the fact that the cheese contained high protein levels including the mycotoxin precursors (tryptophan and histidine). In conclusion, it cannot be assumed that cheese is a poor substrate in terms of its nutritional composition for mycotoxin pro- duction. Most probably, the ability of a mold to produce mycotoxins in cheese is rather species-or strain-specific. Moreover, it also appears that the nature and the quantity of mycotoxins produced by a given mold strain are influenced by the substrate composition as well as some other intrinsic or environmental factors discussed below. Effect of NaCI on mycotoxin production Sodium chloride in cheese contributes to flavor and acts as a preservative through water activity reduction and microorganism inhibition due to solubilized ions. It also enables the growth of salt-tolerant microorganisms that mostly contribute to the sensory properties of cheese. However, NaCI may also affect mycotoxin production. Indeed, Schmidt-Heydt and others (2012) showed that high amounts of OTA were produced by P. nordkum over a wide concentration range of NaCI (5 to 100 g/L), with a weak optimum at 20 g/L in YES medium. A mutant strain of P. nordkum, unable to produce OTA, displayed important growth reduction and elevated chloride content in mycelium under elevated NaCI concentration. It was also shown that P. verrucosum shifted from production of citrinin to ochratoxin when NaCI concentration was equal to or above 20 g/L with a peak at 40 g/L NaCI. As presented above for the possible roles of mycotoxin production in molds, these observations suggest that the biosynthesis and excretion of OTA, containing a chloride atom, plays a role in maintaining chloride homeostasis in the fungal cell. Finoli and others (2001) observed that there was a positive relation (r2 = 0.85) between the amount of roquefortine C and NaCI concentration in various blue cheeses. However, due to the number of varying intrinsic factors, it could not be extrapolated from the latter study whether NaCI alone had an effect on mycotoxin production by P. roqueforti. For example, Finoli and others (2001) also found a positive relation (r2 = 0.58) between ripening time and roquefortine C concentration. Further work is therefore needed to understand how the range of NaCI concentrations commonly encountered in cheeses affect production of mycotoxins other than citrinin and OTA. Effects of temperature, aw, pH, and atmosphere composition on mycotoxin production Whatever the fungal species (contaminant or fungal ripening cultures), observations have been made, both in vitro and in cheese, that mycotoxin-producing fungi are inversely correlated with temperature (low or no production at refrigerated temperatures, intermediate production at 12 °C, and optimal production at 20 to 25 °C) (Le Bars 1979; Wagener and others 1980; Chang and others 1991; Sweeney and Dobson 1998; Finoli and others 2001; Taniwaki and others 2001; Bailly and others 2002; Erdogan and others 2003). This is not surprising, since low temperature also affects fungal growth that is often, but not systematically, correlated with mycotoxin production. Very little information is available on the effects of aw and pH on mycotoxin production by cheese-related fungi (ripening cultures and common contaminants, see "Filamentous Fungi in Cheeses: The Good and the Bad" section). Most fungi able to grow in cheese are salt-tolerant and can grow at relatively low aw conditions. In P. verrucosum, OTA production in YES medium was enhanced at both near-optimal and under mild pH and aw stress conditions (Schmidt and others 2008). However, this does not mean that mycotoxin production did not occur with other conditions since production also occurred at other pH and aw values (Schmidt and others 2008). In P. roqueforti, the optimum pH for 448 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. eremofortin C and PR toxin production was around 4.0 in modified YES medium (Chang and others 1991). However, no information is currently available on the effect of aw on the production of these secondary metabolites or other secondary metabolites such as roquefortine C. More studies are also needed regarding the influence of these factors on mycotoxin production by the most common ripening cultures (P. camemberti, P. roqueforti, F. domesticum, S. flava, and S. fusca) and contaminants (P. commune, P. palitans, P. nalgiovense, and P. nordicum) mentioned in "Filamentous Fungi in Cheeses: The Good and the Bad" section. Gas composition (CO2 and O2 levels) of the atmosphere may also exert a significant impact on mycotoxin production by cheese-related fungi (Taniwaki and others 2001, 2009, 2010). Modified atmosphere packaging (MAP), which combines the inhibitory effect of low oxygen and elevated carbon dioxide levels, is increasingly used as a hurdle technology to prevent spoilage fungi to grow in cheese (Haasum and Nielsen 1998). Taniwaki and others (2001) compared mycotoxin production by strains of A. jiavus, P. roqueforti, and P. commune on sliced-Cheddar cheese in air and the combined presence of C02 (20% or 40%) and 02 (1% or 5%). Mycotoxin production was greatly decreased under modified atmosphere conditions compared to production in air. For example, roquefortine C and CPA production were divided by factors of 7- to 20-fold and 12- to 850-fold under modified atmospheres as compared to mycotoxin production in air by P. roqueforti and P. commune, respectively. On the other hand, the gas composition in the cheese core may also influence secondary metabolite production by P. roqueforti during blue-cheese ripening. In blue-veined cheeses such as Danablu cheese, O2 levels decrease rapidly during the 1st weeks of ripening, while the highest CO2 level reported in Danablu cheese was 20% (van den Tempel and Nielsen 2000). Consequently, these atmosphere conditions are most likely unfavorable for mycotoxin production in cheese. Effects of biotic factors To our knowledge, the intraspecific mycotoxigenic potential of fungal species found in cheese has not yet been extensively studied. Nevertheless, the toxigenic potential of different strains of P. roqueforti and P. camemberti was highly variable (Le Bars 1979; Chang and others 1991; Finoli and others 2001). As already mentioned, the quantity and nature of secondary metabolites produced may be a strain-dependent trait. For example, high intraspecific differences in MPA/PR toxin ratios have been reported in P. roqueforti (O'Brien and others 2006). Microbial interactions play an important role in cheese quality and safety (Irlinger and Mounier 2009). Cheese microbiota or the addition of protective cultures may provide a hurdle effect (competitive interactions, production of inhibitory compounds) toward potentially toxigenic fungal contaminants that, in turn, are unable to produce mycotoxins (Nielsen and others 1998; Irlinger and Mounier 2009; Dalie and others 2010). Finally, it is worth noting that very little attention has been given to studying the effects of cheese cultures with beneficial use on mycotoxin production by fungal ripening cultures. As mentioned above, further work is still required to fully understand the effect of biotic and abiotic factors on fungal growth and subsequent mycotoxin production in cheese. The effects of single abiotic factors and their interactions on mycotoxin formation and the intraspecific variability within mycotoxigenic spoilage fungi and fungal ripening cultures deserve further attention. With the development of mathematical models able to predict fungal growth and mycotoxin production, the cheese industry will be able to ap- ply them to both existing and newly developed products. This will allow companies to better define manufacturing processes, storage conditions, shelf-life, and will contribute to the foodborne mycotoxin risk assessment (Membre and Lambert 2008; Garcia and others 2009). Control of Mycotoxins in Cheeses As previously stated, some authors consider that cheese is very susceptible to mold growth as well as mycotoxin production (Sengun and others 2008). On the contrary, other authors (Gourama 1997) have indicated that conflicting reports can be found in the literature concerning mycotoxin production in cheese and dairy products, and concluded that cheese is actually a better medium for mold growth than for mycotoxin production. The impact of toxic molds on cheese is relatively low and nontoxic molds overgrow toxic ones; moreover, cheese is a poor substrate for mycotoxin production if correctly stored at a low temperature (5 to 7 °C) (Bullerman 1981). However, even if the risk is considered to be low, it should be controlled. The 1st goal is to avoid mycotoxins in milk. Then, during cheese production and storage until consumption, 2 different aspects should be considered, fungal contaminants that potentially produce mycotoxins are undesirable and should be avoided and fungal ripening cultures intentionally added during cheesemaking (for example, blue-veined cheeses or camembert-type cheeses) should not be mycotoxigenic. Control of mycotoxins in milk Good animal feeding practices are necessary to avoid mycotoxins in feed products and subsequently, in milk (mainly AFM1). Molds able to produce mycotoxins may grow on plant materials in the fields before harvest, during handling and storage, as well as during transformation into feed products (Jouany 2007). During harvest and postharvest steps, mycotoxin control includes: early harvest usually associated with low concentrations of mycotoxins, high cutting heights to avoid contamination from soil particles and elimination of damaged grains that can favor fungal growth, low humidity during storage (the main factor to avoid fungal growth, critical limit of 14% w/w), and cool storage temperatures and ventilation that are important secondary factors to reduce the growth risk (Jouany 2007). For silage, humidity remains relatively high; therefore, is it necessary to respect the optimal moisture content at harvest (33% to 35% dry matter for the entire plant [Demarquilly 1994] and use good agricultural practices when filling silos). By following these guidelines, anaerobic conditions are encountered and lactic fermentation takes place thus avoiding fungal growth. LAB, known for their antifungal activities, can be inoculated during silage production and are now more frequently used than chemical preservatives (Kalac 2011). A Lactobacillus casei subsp. pseu-doplantarum 371 isolate from silage, was studied for its ability to inhibit mold growth and A. flavus subsp. parasiticus NRRL 2999 AF production (Gourama and Bullerman 1997). Numerous biological additives, usually containing LAB, are currently marketed for silage control. However, the main recommendations for usual practice are to quickly create stable anaerobic conditions and to apply propionic acid, well known for its antimold activity (Kalac 2011). In Europe, propionic acid (E code = E280) is authorized for use in feed without a time limit and, consequently, different acidic additives for silage containing propionic acid are currently sold. Concerning milk containing mycotoxins, curative physical, chemical, and/or biological treatments to eliminate mycotoxins ©2014 Institute of Food Technologists' Vol. 13,2014 • Comprehensive Reviews in Food Science and Food Safety 449 Filamentous fungi and mycotoxins in cheese. are poorly efficient and too expensive to bring viable solutions for feed decontamination (Jouany 2007; Kolossova and others 2009). Different physical and chemical methods have been recommended to detoxify food and feed; however, only a few (ammonia treatment to destroy AF for instance) are practically used (Bata and Lasztity 1999). Up to 95% to 98% of AF is decomposed by am-moniation that is consequently used in various countries for animal feedstuffs decontamination (van Egmond 2004). Milk-producing animals are more or less efficient barriers against mycotoxins. Rumen microorganisms and the liver produce toxic metabolites (Kalac 2011). However, of healthy animal rumen microbial communities also contribute to prevent health risks as they biotransform different mycotoxins (for instance, de-oxynivalenol or zearalenone, the main mycotoxins produced in silage) into harmless metabolites. Consequently, milk usually contains low concentrations of mycotoxins. Mycotoxins are generally stable compounds. However, heat treatment and ultraviolet light application have been shown to be efficient methods to reduce AFM1 content in milk (Rustom 1997). Control of mycotoxigenic fungal contaminants in cheese Good hygiene and sanitation standards throughout the entire food chain permit to minimize mold presence and growth on cheese (Bullerman 1981). Cold Storage. Cold storage (5 to 7 °C) is an important step in mycotoxin prevention (Sengun and others 2008). Penkillium species are the only molds growing at such temperatures (Bullerman 1981) and are less problematic as they do not produce AFs or STC. Such species could produce less toxic and less stable metabolites in cheese (such as MPA). At the consumer level, cheeses should not be maintained at room temperatures unless other prevention techniques are employed. Chemical preservatives. Chemical preservatives can be used to control mycotoxins (Sengun and others 2008). For instance, pi-maricin (= natamycin, E code = E235, produced by the acti-nobacterium Streptomyces natalensis) was shown to delay fungal growth and as a consequence to avoid mycotoxins (Kiermeier and Zierer 1975; Nilson and others 1975). Its direct effect on production of mycotoxins in cheese remains poorly documented. This antifungal agent is now permitted and used in cheese in Europe and the United States. Private companies sell natamycin formulations for use in cheese production. In Europe, the admissible daily intake value is 0.3 mg/kg of body weight. Mold growth was inhibited in vacuum-packed Kashar cheese containing natamycin, for a 5-mo ripening period at 4 °C (Var and others 2006). Preservatives such as sorbic, benzoic, and propionic acids can inhibit mold growth, except for P. roqueforti. This species is resistant and, consequently, the most important spoilage species, except for blue-veined cheeses (Filtenborg and others 1996). The U.S. Federal Standard of Identity permits the use of sorbate in cheeses at levels under 0.3%, calculated as sorbic acid (Sengun and others 2008). Propionates are permitted for the surface treatment (such as wrapping materials) of cheese and cheese products (Nielsen and de Boer 2004). Mold growth on cheeses can also be reduced using high sodium chloride concentrations that decrease water activity required for growth and mycotoxin production (Sengun and others 2008); however, both organoleptic factors and health considerations have to be taken into account in this context. Reduction or absence of oxygen. Since mycotoxin-producing molds are obligate aerobes, mold growth and mycotoxin forma- tion are minimized by low oxygen concentrations and/or high concentrations of other gases (CO2) in the gaseous phase (Bullerman 1981). Consequently, vacuum-packaging is used to inhibit fungal growth, and in such conditions, production of roquefortine C and CPA by P. roqueforti and P. commune is virtually eliminated (Taniwaki and others 2001). Plant extracts. Knowing their organoleptic characteristics, antimicrobial herbs, spices, or their essential oils have long been used to prevent fungal growth in cheese (Sengun and others 2008). On this basis, eugenol and thymol were shown to affect growth and citrinin production by 2 P. citrinum strains in Spanish cheeses (Vazquez and others 2001). Eugenol appeared more efficient than thymol and the effects reported were cheese-dependent. Fungal growth was totally inhibited in Arzu'a-UUoa cheese with 200 mg/mL of eugenol, but no effect was observed in Cebreiro cheese. At lower concentrations, permitting fungal growth, citrinin production inhibition appeared to be limited or even absent. Microbial control. Competing and/or inhibiting microorganisms can also limit fungal growth and mycotoxin production (Sengun and others 2008). When fully viable, the secondary ripening cultures P. camemberti and G. candidum can efficiently compete against contaminants (Nielsen and others 1998; Decker and Nielsen 2005). Species belonging to the Lactococcus and Lactobacillus genera are the most capable of preventing or limiting mycotoxigenic mold growth because of organic acid production, competition for nutrients, and the production of other antagonistic compounds (Dalie and others 2010). Raw milk from ewe, cow, and goat was shown to be a productive reservoir for antifungal bacteria, most of them belonging to Lactobacillus (Delavenne and others 2011). Eleven antifungal Lactobacillus species showed an antifungal activity in milk with strain-dependent activity spectra (fungal species tested were Debaryomyces hansenii, Kluyveromyces lac-tis, Kluyveromyces marxianus, Penkillium brevicompactum, Rhodotorula mucilaginosa, and Yarrowia lipolytka). Lb. harbinensis showed very strong antifungal effect in yogurt inhibiting all tested fungi. Lactobacillus amylovorus DSM 19280 was shown to inhibit Penkillium expansum, P. roqueforti, Aspergillus niger, Aspergillus fumigatus, and Fusarium culmorum. In the presence of the bacterial adjunct, ched-dar cheeses exposed to natural airborne fungi benefited from a 6-d delay in the appearance of mycelia on their surface without detectable negative impact on cheese quality (Lynch and others 2014). Dairy propionibacteria, widely used for Swiss-type cheese production, are known to ferment glucose, lactic acid, and glycerol into propionic acid, acetic acid, and carbon dioxide. Antifungal effect of dairy propionibacteria was studied using glycerol as carbon source for bacterial growth. Five type strains of propionibacteria were tested against the yeast Rhodotorula mucilaginosa and the molds Penkillium commune and Penkillium roqueforti. Increase in glycerol concentrations enhanced the inhibition of the molds when the yeast was less affected. This effect was shown to be due to the production of propionic acid and simultaneous pH reduction of the medium (Lind and others 2010). Efficient biocontrol solutions for cheesemakers could likely emerge from those recent scientific results. Selection of nontoxigenic ripening cultures Mold ripening cultures used to produce mold-ripened cheeses should have low toxigenic capacity (Sengun and others 2008). Variability within the P. camemberti species in regards to toxigenic capacity allows for the selection of weakly toxicogenic strains (Le Bars 1979). In pure cultures, all known P. camemberti strains are 450 Comprehensive Reviews in Food Science and Food Safety • Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. able to produce CPA (Nout 2004); however, it is possible to select CPA-negative mutants. Consequently, selected ripening cultures were CPA-negative and only very low CPA levels were detected in Camembert-type cheeses. The same observations concerning mycotoxin production variability by P. roqueforti have been made (Finoli and others 2001). It was concluded that roquefortine C was the only metabolite found and that its low toxicity combined with the low levels observed make blue cheese consumption safe for consumers. Moreover, mycotoxin production can be stopped by gene inactivation of the biosynthetic pathway using mutation/selection processes or genetic engineering (Geisen and Holzapfel 1996). However, one should keep in mind that such genetically modified organism (GMO) ripening cultures are not socially accepted everywhere (Costa-Font 2011) and that special regulations could apply to the obtained products. Common sense advice has also been expressed (Sengun and others 2008): to avoid the mycotoxin risk, cheeses with spontaneously growing molds should not be consumed. To conclude on mycotoxin control in cheeses, we can recall that good practices in animal feeding are available to avoid mycotoxin presence in milk, and good hygiene and sanitation standards during milking, milk storage, and cheese production are required. These guidelines associated with efficient selection and preparation of ripening cultures, cold storage of cheese, as well as adapted packaging can all prevent growth of fungal contaminants. Conclusion Filamentous fungi encountered in cheeses may originate from raw materials such as milk or may be introduced during cheese-making either from the environment (contamination), or deliberately inoculated using commercial ripening cultures. Contamination by filamentous fungi can be detrimental to cheese quality, causing appearance defects, off-flavors, and/or potentially toxic secondary metabolite production including mycotoxins. Mycotoxin production profiles may change under different fungal growth conditions, including substrate composition, ecophysio-logical factors (temperature, water activity, pH, or oxygen concentration), and biotic factors (strain-dependent mycotoxin production). To detect a wide spectrum of mycotoxins in cheese, numerous techniques are currently available. Due to the high complexity of the cheese matrix, LC/MS/MS seems to be the most sensitive technique currently available to quantify mycotoxins in cheeses. Concerning mycotoxin toxicity, for those present in cheese, there is a lack of recent scientific data. The most common mycotoxins stable in cheese are AF, citrinin, CPA, roquefortine C, STC, and MPA. Others, including patulin, penicillic acid, and PR toxin do not persist in cheese due to the microaerophilic conditions encountered. Toxicological investigations must be made with new tools to understand the toxicity mechanisms of these mycotoxins. Further knowledge on mycotoxin biosynthetic pathways should help to develop methodologies to elucidate the ecophysiological roles and control mycotoxin production in cheeses. The molecular basis of these pathways may also lead to complementary approaches for mycotoxin control. These guidelines, associated with efficient ripening cultures, selection of nontoxigenic strains, cold storage as well as adapted packaging conditions can all prevent the growth of fungal contaminants. This review emphasizes on future challenges that need to be addressed by the scientific community, fungal culture manufacturers, and artisanal and industrial cheese producers. References Adams E, Tood G, Gibson W. 1975. Long term toxicity study of mycophenolic acid in rabbits. Toxicol Appl Pharmacol 34:509—12. Anon. 2001. Safety evaluation of certain mycotoxins in food. FAO food and nutrition paper 74. World Health Organization (WHO) Food Additive Series 47. Geneva: WHO. Arnold DL, Scott PM, McGuire PF, HawigJ, Nera EA. 1987. Acute toxicity studies on roquefortine and PR toxin, metabolites of Penicillium roqueforti in the mouse. Food Cosmet Toxicol 16:369—71. Arteau M, Labrie S. 2010. Terminal-restriction fragment length polymorphism and automated ribosomal intergenic spacer analysis profiling of fungal communities in Camembert cheese. Intl Dairy J 20:545—54. Auerbach H, Oldenburg E, Weissbach F. 1998. Incidence of Penicillium roqueforti and roquefortine C in silages. J Sci Food Agric 76:565—72. Aydemir AM, Adigusal G, Atasever M, Ozturan K. 2010. Determination of anatoxin Ml levels in some cheese types consumed in Erzurum-Turkey. Kafkas Univ Vet Fak Derg 16:87-91. Baertschi SW, Raney KD, Shimada T, Harris TM, Guengerich FP. 1989. Comparison of rates of enzymatic oxidation of anatoxin Bl, anatoxin Gl, and sterigmatocystin and activities of the epoxides in forming guanyl-N7 adducts and inducing different genetic responses. Chem Res Toxicol 2:114-22. Bailly JD, Querin A, Le Bars-Bailly S, Benard G, Guerre P. 2002. Citrinin production and stability in cheese. J Food Prot 65:1317—21. Barrios M, Medina L, Lopez M. 1998. Fungal biota isolated from Spanish cheeses. J Food Saf 18:151—7. Bata A, Lasztity R. 1999. Detoxification of mycotoxin-contaminated food and feed by microorganisms. Trends Food Sci Technol 10:223—8. Bath K, Persson KN, Schniirer J, Leong SIL. 2012. Microbiota of an unpasteurized cellar-stored goat cheese from northern Sweden. Agric Food Sci 21:197-203. Benford D, Boyle C, Dekant W, Fuchs R, Gaylor DW, Hard G, McGregor DB, Pitt JI, Plestina R, Shephard G, Solfrizzo M, Verger PJP. 2001. Ochratoxin A. In: Safety evaluations of certain mycotoxins in food. WHO Food Additives Series 47, FAO Food and Nutrition Paper 74, IPCS International Programme on Chemical Safety. Geneva: WHO. p 281—387 + Appendix A, p 388-415. Bennett JW, Klich M. 2003. Mycotoxins. Clin Microbiol Rev 16:497-516. Bentley R. 2000. Mycophenolic acid: a one hundred year odyssey from antibiotic immunosuppressant. Chem Rev 100:3801—25. Berthiller F, Dall'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. J Agric Food Chem 53:3421—5. Bhatnagar D, YuJ, Ehrlich KG 2002. Toxine of filamentous fungi. Chem Immunol 81:167-206. Birch AJ, English RJ, Massy-Westropp RA, Smith H. 1957. Origin of the terpenoid structures in mycelianamide and mycophenolic acid. Mevalonic acid as an irreversible precursor in terpene biosynthesis. Proc Chem Soc 1957:233-4. Blanco JL, Dominguez L, Gomez-Lucia E, Garayzabal JFF, Garcia JA, Suarez G. 1988. Presence of anatoxin Ml in commercial ultra-high-temperature-treated milk. Appl Environ Microbiol 54:1622—3. Blount WP. 1961. Turkey X disease. J Br Turk Fed 49:52-4. Bony M. 2000. Methods for detecting mycotoxins. Ind Alimentaires Agricoles 117:59-62. Boysen M, Skouboe P, Frisvad J, Rossen L. 1996. Reclassification of the Penicillium roqueforti group into three species on the basis of molecular genetic and biochemical profiles. Microbiology 142:541—49. Brase S, Encinas A, Keck J, Nising CF. 2009. Chemistry and biology of mycotoxins and related fungal metabolites. Chem Rev 109:3903—90. Brock NL, Dickschat JS. 2013. PR toxin biosynthesis in Penicillium roqueforti. Chembiochem 14:1189-93. Brown MP, Brown-Jenco CS, Payne GA. 1999. Genetic and molecular analysis of anatoxin biosynthesis. Fungal Genet Biol 26:81—98. Bryden WL. 1991. Occurrence and biological effects of cyclopiazonic acid. In: Mise K, Richard JL, editors. Emerging food safety problems resulting from microbiological contamination. Tokyo: Toxic Microorganisms Panel of the UJNR. Bullerman LE. 1981. Public health significance of moulds and mycotoxins in fermented dairy products. J Dairy Sci 64:2439—52. ©2014 Institute of Food Technologists' Vol. 13,2014 • Comprehensive Reviews in Food Science and Food Safety 451 Filamentous fungi and mycotoxins in cheese. Bunger J, Westphal G, Monnich A, Hinnendahl B, Hallier E, Muller M. 2004. Cytotoxicity of oc cupationally and environmentally relevant mycotoxins. Toxicology 202:199—211. Burdock GA, Flamm WG. 2000. Review article: safety assessment of the mycotoxin cyclopiazonic acid. IntlJ Toxicol 19:195—218. Cabanes FJ, Abarca ML, Bragault MR, Castella G. 2007. Especies productoras de micotoxinas. In: Soriano JM editor.Micotoxinas en alimentos. Madrid: Ediciones Diaz de Santos, p 29—61. Cabanes FJ, Bragulat MR, Castella G. 2010. Ochratoxin A producing species in the genus penicillium. Toxins 5:1111—20. Calvert MJ, Ashton PR, Allemann RK. 2002. Germacrene A is a product of the aristolochene synthase-mediated conversion of farnesylpyrophosphate to aristolochene. J Am Chem Soc 124:11636—41. Calvo AM, Wilson RA, BokJM, Keller NP. 2002. Relationship between secondary metabolism and fungal development. Microbiol Mol Biol R 66:447-59. Cantor MD, van den Tempel T, Hansen TK, Ardö Y. (2004). Blue cheese. In: Fox PF, editor. Cheese: chemistry, physics and microbiology. Amsterdam: Elsevier, p 175—98. Caruthers JM, Kang I, Rynkiewicz MJ, Cane DE, Christianson DW. 2000. Crystal structure determination of aristolochene synthase from the blue cheese mold, Penicillium roqueforti.] Biol Chem 275:25533—9. CaryJW, Szerszen L, Calvo AM. 2009. Regulation of Aspergillus flavus anatoxin biosynthesis and development. In: Appell M, Kendra DF. Trucksess MW, editors. Mycotoxin prevention and control in agriculture. ACS Symposium Series, vol. 1031. Washington, DC: American Chemical Society, p 183-203. CAST (Council for Agricultural Science and Technolgy). 2003. Mycotoxins: risk in plant, animal and human systems. Task Force Report Nr. 139. Ames. IA: Council for CAST. Castegnaro M, BarekJ, Fremy J-M, Lafontaine M, Miraglia M, Sansone EB. Telling GM. 1991. Laboratory Ddcontamination and destruction of carcinogens in laboratory wastes: some mycotoxins. IARC Scientific Publications Nr. 113. Lyon: International Agency for Research on Cancer, p 63. Chalmers AA, de Jesus AE, Gorst-Allman CP, Steyn PS. 1981. Biosynthesis of PR toxin by Penicillium roqueforti.] Chem Soc Perkinn 1:2899—2903. Chan WS. 2007. Citrinin induces apoptosis via a mitochondria-dependent pathway and inhibition of survival signals in embryonic stem cells, and causes developmental injury in blastocysts. Biochem J 404:317—26. Chang PK, Ehrlich KC, Fujii I. 2009a. Cyclopiazonic acid biosynthesis of Aspergillus flavus and Aspergillus oryzae. Toxins 1:74—99. Chang PK, Horn BW, Dorner JW. 2009b. Clustered genes involved in cyclopiazonic acid production are next to the anatoxin biosynthesis gene cluster in Aspergillus flavus. Fungal Genet Biol 46:176—82. Chang SC, Lu KL, Yeh SF. 1993. Secondary metabolites resulting from degradation of PR toxin by Penicillium roqueforti. Appl Environ Microbiol 59:981-6. Chang SC, Wei YH, Liu ML, Wei PJ3. 1985. Isolation and some properties of the enzyme that transforms eremofortin C to PR toxin. Appl Environ Microbiol 49:1455-60. Chang SC, Wei YH, Wei DL, Chen YY, Jong SC. 1991. Factors affecting the production of eremofortin C and PR toxin in Penicillium roqueforti. Appl Environ Microbiol 57:2581-5. Chang SC, Yeh SF, Li S-Y, Lei WY, Chen MY. 1996. A novel secondary metabolite relative to the degradation of PR toxin by Penicillium roqueforti. Curr Microbiol 32:141-7. Chao TC, Maxwell SM, Wong SU. 1991. An outbreak of aflatoxicosis andboric acid poisoning in Malaysia: a clinicopathological study. J Pathol 164:225-33. Chen FC, Chen CF, Wei RD. 1982. Acute toxicity of PR toxin, a mycotoxin from Penicillium roqueforti. Toxicon 20:433—41. Chu FS. 1991. Current immunochemical methods for mycotoxin analysis. In: Vandelaan M, Stanker LH, Watkins BE, Roberts DW, editors. Immunoassays for trace chemical analysis: monitoring toxic chemicals in humans, food and the environment. Washington, DC: American Chemical Society, p 140-57. Chu FS. 2000. Mycotoxin analysis: immunological techniques. In: Hui YH. Smith PJ\. and Spoerke DG, editors. Foodborne disease handbook, volume 3: plant toxicants. 2nd ed. New York: Marcel Dekker. p 683—713. Cole R, Cox R. 1981. Handbook of toxic fungal metabolites. New York. NY: Academic Press, p 500. Costa-Font M. 2011. Mapping social and environmental concerns and the acceptability of genetically modified organisms in the European Union. J Socio-Econ 40:903-8. Crawford JM, Dancy BC, Hill EA, Udwary DW, Townsend CA. 2006. Identification of a starter unit acyl-carrier protein transacylase domain in an iterative type I polyketide synthase. Proc Natl Acad Sei USA 103:16728—33. Crawford JM, Korman TP, Labonte JW, Vagstad AL, Hill EA, Kamari-Bidkorpeh O, Tsai SC, Townsend CA. 2009. Structural basis for biosynthetic programming of fungal aromatic polyketide cyclization. Nature 461:1139-43. Crawford JM, Thomas PM, Scheerer JR, Vagstad AL, Kelleher NL, Townsend CA. 2008a. De construction of iterative multidomain polyketide synthase function. Science 320:243—6. Crawford JM, Vagstad AL, Whitworth KP, Ehrlich KC, Townsend CA. 2008b. Synthetic strategy of nonreducing iterative polyketide synthases and the origin of the classical starter-unit effect. Chem Biochem 9:1019—23. Dalie DKD, Deschamps AM, Richard-Forget F. 2010. Lactic acid bacteria - potential for control of mould growth and mycotoxins: a review. Food Control 21:370-80. Dall'Asta C, De Dea Lindner J, Galaverna G, Dossena A, Neviani E. Marchelli R. 2008. The occurrence of ochratoxin A in blue cheese. Food Chem 106:729-34. Decker M, Nielsen PV. 2005. The inhibitory effect of Penicillium camemberti and Geotrichum candidum on the associated funga of white mould cheese. Ind J Food Microbiol 104:51-60. Delavenne E, Mounier J, Asmani K, jany J-L, Barbier G, Le Blay G. 2011. Fungal diversity in cow, goat and ewe milk. IntlJ Food Microbiol 151:247-51. Demain AL, Fang A. 2000. The natural functions of secondary metabolites. Adv Biochem Eng Biotechnnol 69:1—39. Demarquilly C. 1994. Facteurs de variation de la valeur nutritive du mai's ensilage. INRA Prod Animales 7:177-89. De Santi M, Sisti M, Barbieri E, Piccoli G, Brandi G, Stocchi V. 2010. A combined morphologic and molecular approach for characterizing fungal microflora from a traditional Italian cheese (Fossa cheese). Ind Dairy J 20:465-71. Dorner JW, Cole RJ, Erlington DJ, Sukksapath S, Mc Dowell GH, Bryden WL.1994. Cyclopiazonic acid residues in milk and eggs. J Agric Food Chem 42:1616-8. Eaton DL, Ramsdell HS, Neal GE. 1994. Biotransformation of anatoxins. In: Eaton DL, GroopmanJD, editors. The toxicology of anatoxins: human health, veterinary and agricultural significance. San Diego Calif: Academic Press, p 45—7. EFSA (European Food Safety Authority). 2004. Opinion of the scientific panel on contaminants in the Food chain on a request from the request from the commission related to anatoxin Bl substance in animal feed (Request Nr. EFSAQ- 2003-035). Adopted on 3 February 2004. EFSA J 39:1-27. EFSA. 2011. Scientific opinion on the maintenance of the list of QPS biological agents intentionally added to food and feed (2011 update). EFSA J 9:2497-2551. EFSA. 2012. Scientific opinion on the risks for public and animal health related to the presence of citrinin in food and feed. EFSA J 10:2605—87. Ei-Banna A, Pitt JI, Leistner L. 1987. Production of mycotoxins by Penicillium species. Syst Appl Microbiol 10:42—6. Engel G. 1978. Formation of mycotoxins on Tilsit cheese. Milchwiss 33:201-3. Engel G, von Milczewski KE, Prokopek D, Teuber M. 1982. Strain-specific synthesis of mycophenolic acid by Penicillium roqueforti in blue-veined cheese. Appl Environ Microbiol 43:1034-40. Erdogan A, Gurses M, Sert S. 2003. Isolation of moulds capable of producing mycotoxins from blue mouldy Tulum cheeses produced in Turkey. Food Microbiol 85:83-5. Erdogan A, Sert S. 2004. Mycotoxin-forming ability of two Penicillium roqueforti strains in blue moldy tulum cheese ripened at various temperatures. J Food Prot 67:533-5. Filtenborg O, FrisvadJC, Thrane U. 1996. Moulds in food spoilage. IntlJ Food Microbiol 33:85-102. Finoli C, Vecchio A, Bellavita M, Cerruti G. 1983. Sulla presenza di aflatossina Ml in latte e derivati. Latte 8:611—25. Finoli C, Vecchio A, Galli A, Dragoni I. 2001. Roquefortine C occurrence in blue cheese. J Food Prot 64:246—51. 452 Comprehensive Reviews in Food Science and Food Safety ♦ Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. Florez AB, Mayo B. 2006. Microbial diversity and succession during the manufacture and ripening of traditional, Spanish, blue-veined Cabrales cheese, as determined by PCR-DGGE. IntlJ Food Microbiol 110: 165-71. Fox PF, McSweeney PLH. 2004. Cheese: an overview. Cheese: chemistry. Phys Microbiol 1:1-18. Fox EM, Howlett BJ. 2008. Secondary metabolism: regulation and role in fungal biology. Curr Opin Microbiol 11:481—7. FrisvadJC, SmedsgaardJ, Larsen TO, Samson RA, Robert A. 2004. Mycotoxins, drugs and other extrolites produced by species in Penicillium subgenus Penicillium. Stud Mycol 49:201—41. FrisvadJC, Thrane U, Samson RA, PittJL. 2006. Important mycotoxins and the fungi which produce them. In: Hocking AD, Pitt JI, Samson RA. Thrane U, editors. Advances in food mycology. Adv Exp Med Biol 571:3-31. FrisvadJC, Thrane U, Samson RA. 2007. Mycotoxin producers. In: Dijksterhuis J, Samson RA, editors. Food mycology: a multifaceted look at fungi and food. Baco Raton, Florida: CRC Press/Taylor and Francis, p 135-59. Gallo A, Moschini M, Masoero F. 2008. Anatoxins absorption in the gastro-intestinal tract and in the vaginal mucosa in lactating dairy cows. Ital J Anim Sci 8:53—63. Garcia D, Ramos AJ, Sanchis V, Marin S. 2009. Predicting mycotoxins in foods: a review. Food Microbiol 26:757—69. Garcia-Estrada C, Ullan RV, Albillos SM, Fernandez-Bodega MA, Durek P, Von Dohren H, MartinJF. 2011. A single cluster of coregulated genes encodes the biosynthesis of the mycotoxins roquefortine C and meleagrin in Penicillium chrysogenum. Chem Biol 18:1499—512. Gareis M, Scheuer R. 2000. Ochratoxin A in meat and meat products. Arch Fur Lelebensmittelhygiene 51:102—4. Garon D, Richard E, Sage L, Bouchart V, Pottier D, Lebaill P. 2006. Mycoflora and multimycotoxin detection in corn silage: experimental study. J Agric Food Chem 54:3479-84. Geisen R, Holzapfel WH. 1996. Genetically modified starter and protective cultures. IntlJ Food Microbiol 30:315-24. Geisen R, Schmidt-Heydt M, Karolewiez A. 2006. A gene cluster of ochratoxin A biosynthetic genes in Penicillium. Mycotoxin Res 22:134—41. Georgianna DR, Payne GA. 2009. Genetic regulation of anatoxin biosynthesis: from gene to genome. Fungal Genet Biol 46:113—25. Gourama H. 1997. Inhibition of growth and mycotoxin production of Penicillium by Lactobacillus species. Lebensm Wiss Technol 30:279—83. Gourama H, Bullerman LB. 1997Anti-aflatoxigenic activity of Lactobacillus casei pseudoplantarum. IntlJ Food Microbiol 34:131—43. Giirses M, Erdogan A, Qetin B. 2004. Occurence of anatoxin Ml in some cheese types sold in Erzurum, Turkey. Turk J Vet Anim Sci 28:527—30. Haasum I, Nielsen PV. 1998. Ecophysiological characterization of common food-borne fungi in relation to pH and water activity under various atmospheric compositions. J Appl Microbiol 84:451—60. Haggblom P. 1990. Isolation of roquefortine C from Feed Grain. Appl Environ Microbiol 56:2924-6. Hamasaki T, Hatsuda Y. 1977. Sterigmatocystin and related compounds In: Rodricks JV, Hesseltine GW, Mehlman MA, editors. Mycotoxins in human and animal health. Park Forest South, 111.: Pathtox Publishers, Inc. p 597-607. Hansen BG, Mnich E, Nielsen KF, Nielsen JB, Nielsen MT, Mortensen UH, Larsen TO, Patil KR. 2012. Involvement of a natural fusion of a cytochrome P450 and a hydrolase in mycophenolic acid biosynthesis. Appl Environ Microbiol 78:4908-13. Hayaloglu AA, Kirbag S. 2007. Microbial quality and presence of moulds in Kuflu cheese. IntlJ Food Microbiol 115:376-80. Hayes AW. 1980. Mycotoxins: a review of biological effects and their role in human diseases. Clin Toxicol 17:47—83. Hazrat A, Ries MI, NijlandJG, Lankhorst PP, Hankemeier T, Bovenberg RA, Vreeken RJ, Driessen AJM. 2013. A branched biosynthetic pathway is involved in production of roquefortine and related compounds in Penicillium chrysogenum. PLoS ONE 8:e65328. Heinonen JT, Fisher R, Brendel K, Eaton DL. 1996. Determination of anatoxin Bl biotransformation and binding to hepatic macromolecules in human precision liver slices. Toxicol Appl Pharmacol 136:1—7. Hermet A, Meheust D, Mounier J, Barbier G, JanyJL. 2012. Molecular systematics in the genus Mucor with special regards to species encountered in cheese. Fungal Biol 116:692-705. Hidalgo PI, Ullan RV, Albillos SM, Montero O, Fernandez-Bodega MA, Garcia-Estrada C, Fernandez-Aguado M, MartinJF. 2014. Molecular characterization of the PR-toxin gene cluster in Penicillium roqueforti and Penicillium chrysogenum: cross talk of secondary metabolite pathways. Fungal Genet Biol 62:11-24. Hoffmeister D, Keller NP. 2007. Natural products of filamentous fungi: enzymes, genes, and their regulation. Nat Prod Rep 24:393—416. Hohler D. 1998. Ochratoxin A in food and feed: occurrence, legislation and mode of action. Z Ernahrungswiss 37:2—12. Hohn TM, Plattner RD. 1989. Purification and characterization of the sesquiterpene cyclase aristolochene synthase from Penicillium roqueforti. Arch Biochem Biophys 272:137-43. Holzapfel CW, Wilkins DC. 1971. On the biosynthesis of cyclopiazonic acid. Phytochemistry 10:351—8. Horn BW, Dorner JW. 2001. Effect of competition and adverse culture conditions on anatoxin production by Aspergillus jlavus through successive generations. Mycologia 94:741—51. Huffman J, Gerber R, Liangcheng D. 2010. Recent advancements in the biosynthetic mechanisms for polyketide-derived mycotoxins. Biopolymers 93:764-76. Hundley BR. 2001. Mycotoxins and the feed industry. Proceedings of the AFMA Student Symposium. Pietermaritzburg, South Africa: University of Natal; 1-9. Hussein HS, BraselJM. 2001. Toxicity, metabolism and impact of mycotoxins on humans and animals. Toxicology 167:101—34. IARC (International Agency for Research on Cancer). 1976. Some naturally occuring substances. Monographs. Vol. 10. Lyon, France: IARC. p 245—51. IARC (International Agency for Research on Cancer). 1993. Monographs Vol. 56. Lyon, France: IARC. p 489-521. IARC (International Agency for Research on Cancer). 2002. Some traditional herbal medicines, some mycotoxins, naphthalene and styrene. Monographs. Vol. 82. Lyon, France: IARC. p 301-66. Irlinger F, Mounier J. 2009. Microbial interactions in cheese: implications for cheese quality and safety. Curr Opin Biotechnol 20:142—8. Jany J-L, Barbier G. 2008. Culture-independent methods for identifying microbial communities in cheese. Food microbial 25:839—48. Jelen HH. 2002. Volatile sesquiterpene hydrocarbons characteristic for Penicillium roqueforti strains producing PR toxin. J Agric Food Chem 50:6569-74. Josephs RD, Schuhmacher R, Krska R. 2001. International interlaboratory study for the determination of the Fusarium mycotoxins zearalenone and deoxynivalenol in agricultural commodities. Food Add Contam 18: 417-30. Jouany J-P. 2007. Methods for preventing, decontaminating and minimizing the toxicity of mycotoxins in feeds. Anim Feed Sci Technol 137:342—62. Kalac P. 2011. The effects of silage feeding on some sensory and health attributes of cow's milk: a review. Food Chem 125:307—17. Kamkar A. 2008. The study of anatoxin Ml in UHT milk samples by ELISA. J Vestibular Res 63:7-12. Karolewiez A, Geisen R. 2005. Cloning a part of the ochratoxin A biosynthetic gene cluster of Penicillium nordicum and characterization of the ochratoxin polyketide synthase gene. Syst Appl Microbiol 28:588—95. Keller NP, Hohn TM. 1997. Metabolic pathway gene clusters in filamentous fungi. Fungal Genet Biol 21:17—29. Kiermeier F, Zierer E. 1975. Zur wirkung von pimaricin auf schimmelpilze und deren aflatoxinbildung bei kasen. Z Lebensm Unters Forsch 157:253-62. Kitchen DN, Carlton WW, TuiteJ. 1977. Ochratoxin A and citrinin induced nephrotoxicosis in beagle dogs. I. Clinical and clinicopathological features. Vet Pathol 14:154-72. Kokkonen M, Jestoi M, Rizzo A. 2005a. The effect of substrate on mycotoxin production of selected Penicillium strains. IntlJ Food Microbiol 99:207-14. Kokkonen M, Jestoi M, Rizzo A. 2005b. Determination of selected mycotoxins in mould cheeses with liquid chromatography coupled to tandem with mass spectrometry. Food Addit Contam 22:449—56. Kolossova A, StrokaJ, Breidbrach A, Kroeger K, Ambrosio M, Bouten K, Ulberth F. 2009. JRC scientific and technical report: evaluation of the effect of mycotoxin binders in animal feed on the analytical performance of standardized methods for the determination of mycotoxin in feed (EUR 23997 EN). Luxembourg: Office for Official Publications of the European Communities. © 2014 Institute of Food Technologists® Vol. 13, 2014 ♦ Comprehensive Reviews in Food Science and Food Safety 453 Filamentous fungi and mycotoxins in cheese... Kopp-Holtwiesche B, Rehm HJ. 1990. Antimicrobial action of roquefortine. J Environ Pathol Toxicol Oncol 10:41-4. Kozlovskii AG, Reshelilova TA, Medvedeva TN, Arinbasarov MU. Sakharovskij VG, Adanin VM. 1979. Intracellular and extracellular alkaloids of the fungus Penicillium roqueforti. Blokhlmya 44:1691—700. Krška R, Josephs R. 2001. The state-of-the-art in the analysis of estrogenic mycotoxins in cereals, Fresenius J. Anal Chem 369:469—76. Kuráta H. 1990. Mycotoxins and mycotoxicoses. In: Pohland E, Dowell VR. Richards JL, editors. Amicrobial toxins in foods and feeds. New York: Plenum Press, p 249-59. Lafont P, DebeaupuisJ, Gaillardin M, PaydenJ. 1979. Production of mycophenolic acid by Penicillium roqueforti strains. Appl Environ Microbiol 37:365-8. Larsen TO, Svendsen A, SmedsgaardJ. 2001. Biochemical characterization of ochratoxin A-producing strains of the genus Penicillium. Appl Environ Microbiol 67:3630-5. Le Bars J. 1979. Cyclopiazonic acid production by Penicillium camemberti Thorn and natural occurrence of this mycotoxin in cheese. Appl Environ Microbiol 38:1052-5. Lebrun S, Golka K, Schulze H, Follmann W. 2006. Glutathione- S-transferase polymorphisms and ochratoxin A toxicity in primary human urothelial cells. Toxicology 224:81—90. Li YP, Xu Y, Zhi-Bang H. 2012. Isolation and characterization of the citrinin biosynthetic gene cluster from Monascus purpureus. Biotechnol Lett 34:131-6. Lie JL, Marth EH. 1967. Formation of anatoxin in cheddar cheese by Aspergillus flavus and Aspergillus parasiticus. J Dairy Sei 50:1708—10. Lieu FY, Bullerman LB. 1977. Production and stability of anatoxins. penicillic acid, and patulin, in several substrates. J Food Sei 42:1222—4. Lind H, Broberg A, Jacobsson K, Jonsson H, Schnürer J. 2010. Glycerol enhances the antifungal activity of dairy Propionibacteria. IntlJ Microbiol 2010:430873-82. Liu X, Walsh C. 2009a. Cyclopiazonic acid biosynthesis in Aspergillus sp.: characterization of a reductase-like R* domain in cyclopiazonate synthetase that forms and releases cyclo-acetoacetyl-ltryp top han. Biochemistry 48:8746-57. Liu X, Walsh C. 2009b. Characterization of cyclo-acetoacetyl-L-tryptophan dime thy laUyltransferase in cyclopiazonic acid biosynthesis: substrate promiscuity and site directed mutagenesis studies. Biochemistry 48:11032-44. Lopez -Diaz TM, Roman-Bianco C, Garcia-Arias MT, Garcia- Fernandez MC, Garcia-Lopez ML. (1996). Mycotoxins in two Spanish cheese varieties. IntlJ Food Microbiol 30:391—5. Lund F, Filtenborg O, FrisvadJC. 1995. Associated mycoflora of cheese. Food Microbial 12:173-80. Lynch KM, Pawlowska AM, Brosnan B, Coffey A, Zannini E, Furey A. McSweeney PLH, Waters DM, Arendt EK. 2014. Application of Lactobacillus amylovorus as an antifungal adjunct to extend the shelf-life of Cheddar cheese. Intl Dairy J 34:167—73. Magan N, AldredD. 2007. Why do fungi produce mycotoxins? In: Dijkserhuis J, Samson RA, editors. Food mycology: a multifaceted approach to fungi and food. New York: CRC Press/Taylor and Francis Group, p 121-33. Mak AC, Osterfeld SJ, Yu H, Wang SX, Davis RW, Jejelowo OA, Pourmand N. 2010. Sensitive giant magneto resistive-based immunoassay for multiplex mycotoxin detection. Biosens Bioelectron 25:1635—9. Malik AK, Blasco C, Pico Y. 2010. Liquid chromátography-mass spectrometry in food safety. J Chromatograph A 1217:4018—40. Manabe M. 2001. Fermented foods and mycotoxins. Mycotoxins 51:25—8. Maragos CM, Richard JL. 1994. Quantitation and stability of fumonisins Bl and B2 in milk. J AOAC Intl 77:1162-7. Marth H. 1979. Anatoxin in milk, cheese and other dairy products. Proceedings of Marschall International Cheese Conference 6:21. Membré JM, Lambert RJ. 2008. Application of predictive modelling techniques in industry: from food design up to risk assessment. Ind J Food Microbiol 128:10-5. Minto RE, Townsend CA. 1997. Enzymology and molecular biology of anatoxin biosynthesis. Chem Rev 97:2537—56. Miraglia M, Brera C. 2000. Determination of mycotoxins in grains and related products. In: Nollet LML, editor. Food analysis by HPLC. 2nd ed. New York: Marcel Dekker. p 493-522. MogensenJM, FrisvadJC, Thrane U, Nielsen KF. 2010. Production of Fumonisin B2 and B4 by Aspergillus niger on grapes and raisins. Agric Food Chem 58:954-8. Mohajeri FA, Ghalebi SR, Rezaeian M, Gheisari HR, Azad HK, Zolfaghari A, Fallah AA. 2013. Anatoxin Ml contamination in white and Lighvan cheese marketed in Rafsanjan, Iran. Food Control 33:525—7. Montagna MT, Santacroce MP, Spilotros G, Napoli C, Minervini F, Papa A. Dragoni I. 2004. Investigation of fungal contamination in sheep and goat cheeses in southern Italy. Mycopathologia 158:245—9. Moreau S, Lablache-Combier A, BiguetJ. 1980. Production of Eremofortins A, B, and C relative to formation of PR toxin by Penicillium roqueforti. Appl Environ Microbiol 39:770-6. Moss MO. 1991. Economic importance of mycotoxins-recent incidence. Ind Biodeter Biodegr 27:195-204. Moubasher AH, Abdel-Kader MIA, El-Kady IA. 1978. Toxigenic fungi isolated from Roquefort cheese. Mycopathologia 66:187—90. Moule Y, Jemmali M, Rousseau N. 1976. Mechanism of the inhibition of transcription by PR toxin, a mycotoxin from Penicillium roqueforti. Chem Biol Interact 14:207-16. Nielsen MS, FrisvadJC, Nielsen PV. 1998. Protection by fungal straters against growth and secondary metabolite production of fungal spoilers of cheese. IntlJ Food Microbiol 42:91-9. Nielsen PV, de Boer E. 2004. Food preservatives against fungi, 357—363. In: Samson RA, Hoekstra ES, Fisvad JC, editors. Introduction to food- and airborne fungi. Utrecht, The Netherlands: CBS. Nielsen KF, Sumarah MW, FrisvadJC, MillerJD. 2006. Production of metabolites from Penicillium roqueforti complex. J Agric Food Chem 54:3756-63. Nilson KM, Shahani KM, Vakil JR, Kilara A. 1975. Pimaricin and mycostatin for retarding cottage cheese spoilage. J Dairy Sci 58:668—71. Northolt MD, van Egmond HP, Soentoro R, Deijll E. 1980. Fungal growth and the presence of sterigmatocystin in hard cheese. J Assoc Off Anal Chem 63:115-9. Nout MJR. 2004. Useful role of fungi in food processing, 364—374. In: Samson RA, Hoekstra ES, Fisvad JC, editors. Introduction to food- and airborne fungi. Utrecht, The Netherlands: CBS. O'Brien E, Dietrich DR. 2005. Ochratoxin A: the continuing enigma. Crit Rev Toxicol 35:33—60. O'Brien M, Nielsen KF, O'Kiely P, Forristal PD, Fuller HT, FrisvadJC. 2006. Mycotoxins and other secondary metabolites produced in vitro by Penicillium paneum Frisvad and Penicillium roqueforti Thorn isolated from baled grass silage in Ireland. J Agric Food Chem 54:9268—76. O'Brien NM, O'Connor TP., O'Callaghan J. Dobson ADW. 2004. Toxins in cheese biogenic amines and mycotoxins in cheese: chemistry, physics and microbiology. In: Fox PF, Guinee T, Cogan T, McSweeney P, editors. 3rd ed., General Aspects, Vol. 1, chapter 19. Amsterdam: Elsevier Applied Science. O'Callaghan J, Coghlan A, Abbas A, Garcia-Estrada C, Matin JF, Dobson ADW. 2013. Functional characterization of the polyketide synthase gene required for ochratoxin A biosynthesis in Penicillium verrucosum. IntlJ Food Microbiol 161:172-81. O'Connor TP, O'Brien NM. 2000. Nutritional aspects of cheese. In: Fox PF, Guinee T, Cogan T, McSweeney P, editors. Fundamentals of cheese science. Amsterdam: Elsevier Applied Science, p 504—13. Ohmomo S, Sato T, Utagawa T, Abe M. 1975. Isolation of festuclavine and three new indole alkaloids, roquefortine A, B and C from the cultures of Penicillium roqueforti (production of alkaloids and related substances by fungi part XII). J Agric Chem Soc Japan 49:615-23. Ohmomo S, Utagawa T, Abe M. 1977. Identification of roquefortine C produced by Penicillium roqueforti. Agric Biol Chem 41:2097—8. Olivigni FJ, Bullerman LB. 1977. Silmutaneous production of penicillic acid and patulin by a Penicillium species isolated from Cheddar cheese. J Food Sci 42:1654-7. Olivigni FJ, Bullerman LB. 1978. Production of penicillic acid and patulin by an atypical Penicillium roqueforti isolate. Appl Environ Microbiol 35:435—8. Ostry V, Malir F, RuprichJ. 2013. Producers and important dietary sources of ochartoxin A and citrinin. Toxins 5:1574—86. Overy DP, Nielsen KF, SmedsgaardJ. 2005-Roquefortine/oxaline biosynthesis pathway metabolites in Penicillium ser. corymbifera: in planta production and implications for competitive fitness. J Chem Ecol 10:2373-90. 454 Comprehensive Reviews in Food Science and Food Safety ♦ Vol. 13,2014 © 2014 Institute of Food Technologists' Filamentous fungi and mycotoxins in cheese. Palanee T, Dutton M, Chuturgoon A. 2001. Cytotoxicity of aflatoxin Bl and its chemically synthesised epoxide derivative on the A549 human epithelioid lung cell line. Mycopathology 151:155—9. Panelli S, Buffoni JN, Bonacina C, Felligini M. 2012. Identification or moulds from the taleggio cheese environment by the use of barcodes. Food Control 28:285-391. Paterson RRM, Lima N. 2010. Toxicology of mycotoxins. In: Luch A. editor. Molecular, clinical and environmental toxicology clinical toxicology. Vol. 2. Basel: Springer, p 31—63. Pattono D, Grosso A, Stocco PP, Pazzi M, Zeppa G. 2013. Survey of the presence of patulin and ochratoxin A in traditional semi-hard cheeses. Food Control 33:54-7. Payne GA, Brown MP. 1998. Genetics and physiology of aflatoxin biosynthesis. Annu Rev Phytopathol 36:329—62. Pfohl-Leszkowicz A, Manderville RA. 2007. Ochratoxin A: an overview on toxicity and carcinogenicity in animals and humans. Mol Nutr Food Res 51:61-99. Pitt JI. 2000. Toxigenic fungi and mycotoxins. Br Med Bull 56:184-92. Pitt JI, Cruickshank RH, Leistner L. 1986. Penicillium commune, P. camembertii. the origin of white cheese moulds, and the production of cyclopiazonic acid. Food Microbiol 3:363—71. Proctor RH, Hohn TM. 1993. Aristolochene synthase: isolation, characterization, and bacterial expression of a sesquiterpenoid biosynthetic gene (Aril) from Penicillium roqueforti.] Biol Chem 268:4543—6. Puls R, Ladyman E. 1988. Roquefortine toxicity in a dog. Can Vet J 29:569. Raney KD, Shimada T, Kim D-H, Groopman JD, Harris TM, Guengerich FP. 1992. Oxidation of anatoxins and sterigmatocystin by human liver microsomes: significance of aflatoxin-Ql as a detoxication product of aflatoxin-Bl. Chem Res Toxicol 5:202—10. Radic B, Fuchs R, Peraica M, Lucie A. 1997. Ochratoxin A in human sera in the aera with endemic nephropathy in Croatia. Toxicol Lett 91:105—9. Raper KB, Thorn C. 1949. A manual of Penicillia. Baltimore: Williams Wilkins Company. 875 pp. Rasmussen RR, Rasmussen PH, Larsen TO, Bladt TT, Binderup ML. 2011. In vitro cytotoxicity of fungi spoiling maize silage. Food Chem Toxicol 49:31-44. Regueira TB, Kildegaard KR, Hansen BG, Mortensen UH, Hertweck C. Nielsen J. 2011. Molecular basis for mycophenolic acid biosynthesis in Penicillium brevicompactum. Appl Environ Microbiol 77:3035—43. Reshetilova TA, Vinokurova NG, Khmelenina VN, Kozlovsky AG. 1995. The role of roquefortine in the synthesis of alkaloids meleagrin. glandicolines A and B, and oxaline in fungi Penicillium glandicola and P. atramentosum. Microbiology 64:27—9. Reverberi M, Ricelli A, Zjalic S, Fabbri AA, Fanelli C. 2010. Natural functions of mycotoxins and control of their biosynthesis in fungi. Appl Microbiol Biotechnol 87:899-911. Ries MI, Ali H, Lankhorst PP, Hankemeier T, Bovenberg RA, Driessen AJ. Vreeken RJ. 2013. Novel key metabolites reveal further branching of the roquefortine/meleagrin biosynthetic pathway. J Biol Chem. 288:37289—95. Riley RT, Goeger DE. 1992. Cyclopiazonic acid: speculations on its function in fungi. In: Bhatnagar D, Lillehoj EB, Arora DK, editors. Handbook of applied mycology. Mycotoxins in ecological systems. New York, N.Y.: Marcel Dekker. p 385-402. Robinson RK. 1995. A colour guide to cheese and fermented milks. London, UK: Chapmann and Hall. Ropars J, Cruaud C, Lacoste S, Dupont J. 2012. A taxonomic and ecological overview of cheese fungi. IntlJ Food Microbiol 155:199—210. Rustom IYS. 1997. Aflatoxin in food and feed: occurrence, legislation and inactivation by physical methods. Food Chem 59:57—67. Sansing GA, Lillehoj EB, Detroy RW. 1976. Synergistic toxic effect of citrinin, ochratoxin A and penicillic acid in mice. Toxicon 14:213—20. Sanzani SM, Reverberi M, Punelli M, Ippolito A, Fanelli C. 2012. Study on the role of patulin on pathogenicity and virulence ofPenicillium expansum. IntlJ Food Microbiol 153:323-31. Schmidt-Heydt M, Graf E, Stoll D, Geisen R. 2012. The biosynthesis of ochratoxin A by Penicillium as one mechanism of adaptation to NaCl rich foods. Food Microbiol 29:233-41. Schmidt-Heydt M, Magan N, Geisen R. 2008. Stress induction of mycotoxin biosynthesis genes by abiotic factors. FEMS Microbiol Lett 284:142—9. Schmidt-Heydt M, Stoll D, Mrohs J, Geisen R. 2013. Intraspecific variability of HOG1 phosphorylation in Penicillium verrucosum reflects different adaptation levels to salt rich habitats. IntlJ Food Microbiol 165:246—50. SchnurerJ, MagnussonJ. 2005. Antifungal lactic acid bacteria as biopreservatives. Trends Food Sci Technol 16:70—8. Schoch U, Luthy J, Schlatter C. 1984. Mycotoxins in mold-ripened cheese. Mitt Geb Lebensmittelunters Hyg 74:50-9. Scott PM. 1981. Toxins of Penicillium species used in cheese manufacture. J Food Prot 44:702-10. Scott PM. 1989. Mycotoxigenic fungal contaminants of cheese and other dairy products. In: van Egmond, HP editor. Mycotoxins in dairy products — mycotoxigenic fungal contaminants of cheese and other dairy products. London: Elsevier Applied Science, p 193—259. Scott PM, Kanhere SR. 1979. Instability of PR toxin in blue cheese. J Assoc Off Anal Chem 62:141-7. Scott PM, Kennedy BPC, HarwigJ, Blanchfield BJ. 1977. Study of conditions for production of roquefortine and other metabolites of Penicillium roqueforti. Appl Environ Microbiol 33:249—53. Sengun IY, Yaman DB, Gonul SA. 2008. Mycotoxins and mould contamination in cheese: a review. World Mycotoxin J 1:291—8. Shimizu T, Kinoshita H, Ishihara S, Sakai K, Nagai S, Nihira T. 2005. Polyketide synthase gene responsible for citrinin biosynthesis in Monascus purpureus. Appl Environ Microbiol 71:3453—7. Shimizu T, Kinoshita H, Nihira T. 2007. Identification and in vivo functional analysis by gene disruption of ctnA, an activator gene involved in citrinin biosynthesis in Monascus purpureus. Appl Environ Microbiol 73:5097—103. Shinohara Y, Tokuoka M, Koyama Y. 2011. Functional analysis of the cyclopiazonic acid biosynthesis gene cluster in Aspergillus oryzae RIB 40. Biosci Biotechnol Biochem 75:2249—52. Siemens AG, ZawitowskiJ. 1993. Occurrence of PR imine, a metabolite of Penicillium roqueforti, in blue cheese. J Food Prot 56:317—9. Skaug MA. 1999. Analysis of Norwegian milk and infant formulas for ochratoxin A. Food Addit Contam 16:75—8. Spahr U, Walther B, Sieber R. 2000. Transfert des mycotoxines dans le lait : vue d'ensemble. Rev Suisse Agric 32:75—8. Starec M, Fiserova A, RosinaJ, MalekJ, Krsiak M. 2001. Effect of agroclavine on NK activity in vivo under normal and stress conditions in rats. Physiol Res 50:513-9. Steyn PS. 1995. Mycotoxins, general view, chemistry and structure. Toxicol Lett 82(8):843-51. Stoev SD, Vitanov S, Anguelov G, Bocharova TP, Creppy EE. 2001. Experimental porcine nephropathy in pigs provoked by a diet containing ochratoxin A and penicillic acid. Vet Res Commun 25:205—23. StolofFL. 1977. Aflatoxin—an overview. In: RodricksJV, Hasseltine CW. Mehlman MA, editors. Mycotoxins in human and animal health. Park Forest: Pathotox Publishers, p 16—28. Stott WT, Bullerman LB. 1976. Instability of patulin in cheddar cheese. J Food Sci 41:201-3. Stroka JAE, Jorissen U, Gilbert J. 2000a. Immunoaffmity column cleanup with liquid chromatography using post-column bromination for determination of anatoxins in peanut butter, pistachio paste, fig paste and paprika powder: collaborative study. JAOAC Ind 83:320-40 (AOAC Official Method 999.07). Stroka J, Van Otterdijk R, Anklam E. 2000b. Immunoaffinity column clean-up prior to thin layer chromatography for the determination of anatoxins in various food matrices. J Chromatogr A 904:251—6. Sweeney MJ, Dobson AD. 1998. Mycotoxin production by Aspergillus. Fusarium and Penicillium species. IntlJ Food Microbiol 43: 141-58. Tanchev Y, Dorossiev D. 1991. The first clinical description of Balkan endemic nephropathy and its validity 35 years later. In: Castegnaro M. Plestina R, Dirheimer G, Chernozemsky IN, Bartsh H, editors. Mycotoxins, endemic nephropathy and urinary tract tumours. Vol. 115. Lyon: IARC Scientific Publication, p 21—8. Taniwaki MH, van Dender AGF. 1992. Occurrence of toxigenic molds in Brazilian cheese. J Food Prot 55:187—91. Taniwaki MH, Hocking AD, Pitt JI, Fleet GH. 2001. Growth of fungi and mycotoxin production on cheese under modified atmospheres. Ind J Food Microbiol 68:125-33. Taniwaki MH, Hocking AD, Pitt JI, Fleet GH. 2009. Growth and mycotoxin production by food spoilage fungi under high carbon dioxide and low oxygen atmospheres. Ind J Food Microbiol 132:100—8. Taniwaki MH, Hocking AD, Pitt JI, Fleet GH. 2010. Growth and mycotoxin production by fungi in atmospheres containing 80% carbon dioxide and 20% oxygen. IntlJ Food Microbiol 143:218-25. © 2014 Institute of Food Technologists' Vol. 13, 2014 ♦ Comprehensive Reviews in Food Science and Food Safety 455 Filamentous fungi and mycotoxins in cheese. Teuber M, Engel G. 1983. Low risk of mycotoxin production in cheese. MAN 1:193-7. Turner NW, Subrahmanyam S, Piletsky SA. 2009. Analytical methods for determination of mycotoxins: a review. Anal Chim Acta 632:168—80. Ueno Y, Kubota K, Ito T, Nakamura Y. 1978. Mutagenicity of carcinogenic mycotoxins in Salmonella typhimurium. Cancer Res 38:3536—42. Umeda M, Tsutsui T, Saito M. 1977. Mutagenicity and inducibility of DNA single stranded breaks and chromosome aberrations by various mycotoxins. Gann 68:619-25. Vacheyrou M, Normand A-C, Guyot P, Cassagne C, Piarroux R, Bouton Y. 2011. Cultivable microbial communities in raw cow milk and potential transfers from stables of sixteen French farms. Intl J Food Microbiol 146:253-62. van den Tempel T, GundersenJK, Nielsen MS. 2002. The microdistribution of oxygen in Danablu cheese measured by a microsensor during ripening. IndJ Food Microbiol 75:157-61. van den Tempel T, Nielsen MS. 2000. Effects of atmospheric conditions. NaCl and pH on growth and interactions between moulds and yeast related to blue cheese production. Intl J Food Microbiol 57:193—9. van der Merwe KJ, Steyn PS, Fourie L, de Scott B, Theron JJ. 1965. Ochratoxin A, a toxic metabolite produced by Aspergillus ochraceus Willi. Nature 205:1112-3. van Egmond HP. 1983. Mycotoxins in dairy products. Food Chem 11:289-307. van Egmond HP, Paulsch WE. 1986. Mycotoxins in milk and milk products. Neth Milk Dairy J 40:175-88. van Egmond HP. 2004. Mycotoxins: detection, reference materials and regulation. In: Samson RA, Hoekstra ES, FisvadJC, editors. Introduction to food- and airborne fungi. Utrecht, The Netherlands: CBS. p 332—8. Var I, Erginkaya Z, Giiven M, Kabak B. 2006. Effects of antifungal agent and packaging material on microflora of Kashar cheese during storage period. Food Control 17:132-6. Vazquez BI, Fente C, Franco CM, Vazquez MJ, Cepeda A. 2001. Inhibitory effects of eugenol and thymol on Penicillium citrinum strains in culture media and cheese. Intl J Food Microbiol 67:157—63. Versilovskis A, De Saeger S. 2010. Sterigmatocystin: occurrence in foodstuffs and analytical methods-an overview. Mol Nutr Food Res 54:136—47. Versilovskis A, Van Peteghem C, De Saeger S. 2009. Determination of sterigmatocystin in cheese by high-performance liquid chromatography-tandem mass spectrometry. Food Addit Contam 26:127—33. Vinokurova NG, Boichenko DM, Baskunov BP, Zelenkova NF, Vepritskaya IG, Arinbasarov MU, Reshetilova TA. 2001. Minor alkaloids of the fungus Penicillium roqueforti Thorn 1906. Appl Biochem Micro 37:184—7. Visconti A, Pascale M, Centonze G. 2000. Determination of ochratoxin A in domestic and imported beers in Italy by immunoaffinity clean-up and liquid chromatography. J Chromatogr A 888:321—6. Watson DH, Lindsay DG. 1982. A critical review of biological methods for the detection of fungal toxins in foods and feedstuffs. J Sei Food Agric 33:59-67. Wagener P^E, Davis ND, Diener, UL. 1980. Penitrem A and roquefortine production by Penicillium commune. Appl Environ Microbiol 39:882—7. Webner FC, Thiel PG, van Rensburg SJ, Demasius IPC. 1978. Mutagenicity to Salmonella typhimurium of some Aspergillus and Penicillium Mycotoxins. Mutat Res 58:193-203. Wei R, Ong T, Whong W, Frezza D, Bronzetti G, Zeiger E. 1979. Genetic effects of PR toxin in eukaryotic microorganisms. Environ Mutagen 1:45-53. Wei YH, Ding WH, Wei RD. 1984. Biochemical effects of PR toxin on rat liver mitochondrial respiration and oxidative phosphorylation. Arch Biochem Biophys 230:400-11. Wei RD, Still PE, Smalley EB, Schnoes HK, and Strong FM. 1973. Isolation and partial characterization of a mycotoxin from Penicillium roqueforti. Appl Microbiol 25:111-4. Wichmann G, Herbarth O, Lehmann I. 2002. The mycotoxins citrinin. gliotoxin, and patulin affect interferon-gamma rather than interleukin-4 production in human blood cells. Environ. Toxicol 17:211—8. Wilson BJ. 1971. Miscellaneous Penicillium toxins. In: Ciegler A, Kadis S, Ail SJ, editors. Microbial toxins. Vol. VI. New York: Academic Press, Inc. p459-521. Xu BJ, Jia XQ, Gu LJ, Sung CK. 2006. Review on the qualitative and quantitative analysis of the mycotoxin citrinin. Food Control 17:271—85. Yabe K, Nakajima H. 2004. Enzyme reactions and genes in anatoxin biosynthesis. Appl Microbiol Biotechnol 64:745—55. Yang X, Lu H, Li Z, Bian Q, Qm L, Liu Q, Li J, Wang X, Wang S. 2012. Cytochrome P450 2A13 mediates anatoxin Bl-induced cytotoxicity and apoptosis in human bronchial epithelial cells. Toxicology 300: 138-48. Yates IE. 1986. Bioassay systems and their use in diagnosis of myco toxicoses. In RichardsJL, ThurstonJR, editors. Diagnosis of mycotoxicoses. Dordrecht, The Netherlands: Martinus Nijhoff. p 333-81. Yiannikouris A, JouanyJP. 2002. Mycotoxins in feeds and their fate in animals: a review. Anim Res 51:81—99. Yu JJ, Chang PK, Ehrlich KC, Cary JW, Bhatnagar D, Cleveland TE, Payne GA, Linz JE, Woloshuk CK, Bennett JW. 2004. Clustered pathway genes in anatoxin biosynthesis. Appl Environ Microbiol 70:1253—62. Zambonin CG, Monaci L, Aresta A. 2001. Determination of cyclopiazonic acid in cheese samples using solid-phase microextraction and high performance liquid chromatography. Food Chem 75:249—54. Zöllner P, Mayer-Helm B. 2006. Trace mycotoxin analysis in complex biological and food matrices by liquid chromatography-atmospheric pressure ionization mass spectrometry. J Chromatogr A 1136:123—69. 456 Comprehensive Reviews in Food Science and Food Safety ♦ Vol. 13,2014 © 2014 Institute of Food Technologists'