Food Contaminants - American Chemical Society

---

Chapter 9

Mycotoxins in Alcoholic Beverages and Fruit Juices: Occurrence and Analysis Peter M. Scott Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

Bureau of Chemical Safety, Health Canada, 2203D, Ottawa, Ontario, K1A 0K9 Canada

A metabolite of black aspergilli growing on grapes, ochratoxin A has often been found in wine on six continents. The European Union (EU) has a regulatory limit of 2.0 ppb in wine and grape juice. Up to 15.6 ppb of ochratoxin A has been found in red wine and 15.3 ppb in a specialty wine (mistelle). Ochratoxin A also occurs in grape juice (up to 6.7 ppb) and frequently in beer in many countries (up to 8.1 ppb). It can be removed experimentally from wine with active dry yeast, yeast lees and fining agents. Aflatoxins have rarely been found in wine but occasionally in beer (ppt levels). Other mycotoxins determined in wine include mycophenolic acid, trichothecin and the Alternaria toxins alternariol and alternariol monomethyl ether. Apple juice frequently contains patulin (originating from Penicillium expansum), which has also been found in grape and other fruit juices. The Fusarium toxins fumonisins and deoxynivalenol (a trichothecene) are additional important contaminants of beer.

170

Published 2008 American Chemical Society.

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

171 The occurrence of mycotoxins in fruits can lead to contamination of processed food products such as fruit juices and wine, while their occurrence in grains such as barley and corn adjunct can give rise to mycotoxins in beer. The subject of occurrence of mycotoxins in these beverages is very broad. This brief review focuses in more detail on the occurrence and analysis of ochratoxin A (OTA) in wine andfruitjuices, a topic of great current interest in those countries where wine is produced, including the EU where there is a maximum allowable limit of OTA in grape juice and wine. Also covered in some detail are occurrence and detection of patulin in apple and in other fruit juices as well as OTA in beer. Background information on the fungal origin and toxicology of those mycotoxins relevant to human health - including carcinogenicity, immunotoxicity, hepatotoxicity, nephrotoxicity, teratogenicity and neurotoxicity - can be found in the report on mycotoxins of the Council for Agricultural Science and Technology (CAST) in the USA (7).

Ochratoxin A in Wine and Grape Juice Occurrence The subject of mycotoxins infruitproducts such as wine andfruitjuices has been reviewed {2-4). OTA in wine has been an important topic of concern to food regulatory authorities (5- 8). The main fungus associated with presence of OTA in grapes and responsible for its occurrence in wine musts is the black aspergillus Aspergillus carbonarius (9-12). A related ochratoxigenic species A. tubigensis can also contaminate grapes (75). OTA was first found in wine and grape juice in Switzerland by Zimmerli and Dick (14) just over 10 years ago. Since then there have been numerous reports on its occurrence in wine (Table I) with fewer reports of occurrence in grape juice (Table II). An analysis of published Europeanfindingsby Belli et al. (6) demonstrated that the mean incidence of OTA positive samples in red wines was 71%, 66% in rosé wines and in white wines 45%. A similar trend was observed for mean OTA concentrations. These differences were first noted by Majerus and Otteneder (38). The difference in normal distributions of OTA concentrations in red and white wines from the Mediterranean region was represented graphically in a Canadian study (57). According to the European Commission Reports on Tasks for Scientific Cooperation (SCOOP), the mean level of OTA found in European wine was 0.36 ppb (24), while for grape juice it was 0.56 ppb (24). The highest levels reported worldwide were 15.6 ppb in a red Italian wine (24) and in a Moroccan wine (55) and 15.3 ppb in a Spanish dessert wine (36). The maximum level of OTA determined in grape juice was 6.7 ppb in a German sample (57). In Turkey, a grape juice produced from moldy grapes is concentrated to a syrup called pekmez, which can contain as much as 50 ppb of OTA (34).

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

172

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

Table L Ochratoxin A in Wine Originating from Various Countries

NOTE: References for most occurrences are cited in (37); where this is not the case other references are indicated and some more recent references are also added.

Table II. Ochratoxin A in Grape Juice Originating from Various Countries

NOTE: References for most occurrences are cited in (31); where this is not the case other references are indicated and some more recent references are also added.

Levels of OTA in red wines from the south of Europe are generally greater than in more northerly European wines (37, 38-40). This appears to be due to climate rather than wine making differences. Concentration increases of OTA in wines from central and southwest France towards the Mediterranean (41), from northern to southern Greece or the Greek islands (42, 43) and from northern to southern Italy (44) have also been observed. Studies on Carignan grape musts in Europe indicate that the OTA concentrations increase the closer the vineyard to the Mediterranean Sea (45). The level of OTA was found to be stable in wine for at least a year (46).

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

173 OTA has rarely been detected in other fruit juices. It was found in grapefruit juice analysed in Morocco (47) but in Switzerland the chromatographic peak seen on analysis of grapefruit juice was regarded as an interference (14). Very low levels of OTA (up to 0.06 ppb) were reported in some samples of black currant and tomato juices in Germany (41) and infruitdrinks in Poland (32).

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

Methods of Analysis for Ochratoxin A in Wine OTA in wine is usually determined by reversed phase liquid chroma­ tography (LC) with fluorescence (48-53) or mass spectrometry detection (5458). Limits of OTA quantitation using fluorescence are in the range 0.02-0.08 ppb. Automation has been applied (50, 58). Cleanup of the wine by an immunoaffinity column is used in most cases and diluted wine may be applied directly to the immunoaffinity column, as in the Official Method of AOAC International (49). Gas chromatography with mass spectrometry detection was evaluated by Soleas et al. (59) as a confirmatory system but it was not recommended for quantitation of OTA in wfne. Castellari et al. (60) compared different immunoaffinity column cleanup procedures while Leitner et al. (54) compared d and immunoaffinity column methods. C i solid phase extraction seems to be a suitable alternative to immunoaffinity column cleanup (52, 57, 61). Other types of solid phase extraction columns, including anion exchange/reversed phase, phenylsilane and polymeric based columns have also been investigated (52, 56, 62). Liquid phase micro-extraction using 1-octanol immobilized in the pores of a porous hollow fiber (63) and solid phase microextraction where the fiber is simply immersed in diluted wine (64) have been investigated for isolation of OTA from wine; the latter technique gave with reversed phase LC and fluorescence a limit of detection of 0.07 ppb. Two attempts have been made to use a molecularly imprinted polymer to selectively bind OTA in wine analysis. One study found that an imprinted polymer performed the same as a non-imprinted polymer, indicating that the imprinted binding sites played a minor role in selective OTA retention (65); the other utilized molecularly imprinted polypyrrole modified carbon nanotubes to give an LC detection limit of 0.012 ppb (66). Dall'Asta et al. (67) avoided cleanup altogether and directly injected wine into the LC system, still obtaining a detection limit by fluorescence of 0.05 ppb. Immunochemical methods did not achieve the same level of sensitivity as LC methods. An electrochemical immunosensor (68) and a membrane dot immunoassay (69) gave detection limits of 0.9 ppb and 1 ppb, respectively, but an array biosensor was not sensitive (detection limit 38 ppb) (70) and would not be suitable for enforcement of the EU regulations, which have a maximum limit for OTA of 2.0 ppb in wine, grape juice and grape must (77). 8

8

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

174

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

Ochratoxin A in Wine: Prevention and Treatment In view of the widespread occurrence of OTA in wine, considerable effort has been made in recent years that aimed to prevent its formation in grapes (vineyard management), to study its fate in the winemaking process and carry out research on treatment of wine and grape juice to remove OTA. Certain fungicides can stop growth of Aspergillus carbonarius on grapes, which together with complementary protection against the grape berry moth results in lower levels of OTA in the musts (72-74). In the processes of crushing and maceration, OTA levels increase in the must because of extraction from the grapes but in the subsequent alcoholic fermentation, racking and malo-lactic fermentation of the wine, decreases in OTA concentrations occur (75). However, poor winery sanitation may result in contamination of the equipment with molds which can contribute to OTA in the must (76). The main losses of OTA occur in the solid-liquid separation after fermentation (77, 78). Overall losses of more than 90% of OTA initially present in the grapes have been reported in red and white wine ready for consumption (78) . Removal of OTA during fermentation was dependent on the yeast strain (79) . In a study where the must was spiked before vinification to port wine, the levels of OTA dropped by up to 92% (80). Radiolabeling of OTA with tritium and analysis by radio-LC of thefiltrateand solid after fermentation indicated that OTA did not appear to be transformed to other products by the yeast (81). Fining (clarification) is a common winery practice. Potassium caseinate and activated charcoal were found experimentally to be the best fining agents to remove OTA from wine (82). Other effective materials were silica gel, gelatin, and bentonite (76, 82, 83). Microbiological adsorbents of OTA from wine or grape juice are active dry yeast (84), yeast lees (84), heat treated yeast (Saccharomyces cerevisiae) (85), Lactobacillus plantarum (86), and living and dead conidia of black Aspergilli (87,88).

Other Mycotoxins in Wine Ochratoxin C The toxic ethyl ester of OTA, ochratoxin C, has been found in wine at about 10% (range 5-25%) of the level of OTA (39), probably formed from OTA and ethanol in the acidic environment.

Alternaría Toxins Alternarla spp. frequently contaminate many kinds of fruits, including grapes (89, 90). Recently, Canadian and imported red and white wines were

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

175 analyzed for two of the toxins of A. alternate*, alternariol (AOH) and alternariol monomethyl ether (AME) (90, 91). As determined by LC-MS/MS, AOH was found in 13 out of 17 Canadian red wines at levels of 0.03 to 5.02 ppb and in 7 out of 7 imported red wines at 0.27-19.4 ppb, usually accompanied by lower concentrations of AME (up to 0.2 ppb in Canadian red wine) (90).

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

Mycophenolic Acid The Pénicillium mycotoxin mycophenolic acid was found in 91% of red wine samples analyzed in Germany (92). The maximum concentration found was 130 ppb and samples from southern Europe (Greece and Spain) were the most contaminated. Another Pénicillium mycotoxin, citrinin, was not detectable in any sample.

Trichothecin and Related Metabolites Trichothecium roseum occurs on grapes in Germany and its metabolites trichothecin (up to 500 ppb) and the related compounds trichothecolone and rosenonolactone have been detected in grapes and wines in that country (93, 94). Trichothecin is cytotoxic (93). The mycotoxin containing wines were Auslese, Trockenbeerenauslese and Muscatel. Trichothecolone, rosenonolactone and isotrichothecin, in addition to two related metabolites, were formed from trichothecin during the fermentation of grape juice (95).

Aflatoxins Several surveys from 1967-1977 reported the presence of aflatoxins in wine (96-99). Two of 33 German white wines contained less than 1 ppb anatoxin Bj as determined by thin layer chromatography (TLC) (96) but identity was not confirmed by an additional test. Another TLC study found aflatoxins Bi, B , Gi and G in 5 out of 22 European wine samples (from 1.2 to 2.6 ppb total aflatoxins in three red wines and two dessert wines); confirmation of identity was by sulfuric acid spray (97). Eleven other red and dessert wines appeared to contain traces of aflatoxins (< 1 ppb). An unpublished report (98), again using TLC, but with identification in 9 solvent systems, indicated the presence, of aflatoxins Bi, B , Gi and G in 33 samples of mainly European red, white and dessert wines (up to, respectively, 4.1, 14 and 49 ppb total aflatoxins B B , Gj and G ). In a U.S. survey, the hemiacetal aflatoxin B was detected by LC in a French Sauterne wine (0.3 ppb) and in a plum wine (0.05 ppb) (99). The subject of aflatoxin occurrence in wines should be revisited using modern analytical 2

2

2

2

h

2

2a

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

2

176 methods. The hemiacetals aflatoxins B2« and G could result from the acidity and ethanol content of wine but are not regarded as very toxic. 2 a

Other Mycotoxins in Fruit Juices

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

Patulin Patulin is one of the most important mycotoxins occurring infruitjuices (7J). Its chemical properties, toxicity, methods of analysis, occurrence and control have been reviewed (700, 707). Originating from Pénicillium expansum growing on apples, patulin occurs frequently in apple juice as shown in surveys from many countries, including Australia, Austria, Belgium, Brazil, Canada, France, Germany, Italy, Japan, New Zealand, Taiwan, Turkey, South Africa, USA and the United Kingdom. The highest level ever determined was 45 000 ppb in freshly pressed apple cider in the USA (700). Patulin has also been found in other fruit juices, including grape, pear, peach, cherry, raspberry, mulberry, strawberry, pineapple, sour cherry, blackcurrant and orange juices (7, 707-705). The commonly used methods of analysis use extraction of juice with ethyl acetate, cleanup by extraction with aqueous sodium carbonate and reversed phase LC with UV detection at 276 nm; cloudy apple juices are pretreated with pectinase to clarify them (700, 707). Solid phase extraction techniques have also been used for cleanup. No enzyme-linked immunosorbent assay (ELISA) has been developed for detection of patulin. The regulatory limit for patulin in the EU is 50 ppb infruitjuices and fruit nectar, in particular apple juice, and 10 ppb in apple juice and solid apple products for infants and young children (706). The 50 ppb limit for apple juice has been adopted by the Codex Alimentarius Commission, while a 50 ppb limit forfruitjuices in general has been adopted by some countries (707). It is well known that the yeast Saccharomyces cerevisiae greatly reduces patulin levels during alcoholic fermentation of apple juice (108-110), with formation of E- and Z-ascladiol (770). Surprisingly then, patulin has been found in French commercial alcoholic apple ciders (777). In line with the fermentation losses in apple juice, patulin was not detected in wine made from patulincontaining grape juice (772). Storage of apple juice concentrates for one month at 22 °C reduced patulin levels by 45-64%, and after four months it was not detectable (773). There was greater reduction of patulin in both apple juice and a juice-like model system in the presence of L-ascorbic acid (774, 775). On the other hand, patulin was stable in grape juice (and wine) for 1 month at room temperature (116). Various methods have been investigated for reduction of patulin in apple juice: sorting of infected apples before processing, addition of ascorbic acid, sulphur dioxide or charcoal (which is not feasible on a commercial scale), fermentation, irradiation, pressing followed by filtration, enzyme treatment and fining with bentonite (700, 707, 777). Efficiency of these methods varied.

In Food Contaminants; Siantar, D., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2008.

177 Alternaría Toxins

Downloaded by PENNSYLVANIA STATE UNIV on July 2, 2012 | http://pubs.acs.org Publication Date: October 20, 2008 | doi: 10.1021/bk-2008-1001.ch009

In contrast to wine, red grape juices contained only sub-ppb levels of alternariol (AOH) or alternariol monomethyl ether (AME), except for one sample (of U.S. origin) which contained 39 ppb AME (90). White wines, white grape juices and cranberry juices contained little or no AOH/AME. AOH (