versão On-line ISSN 1678-457X
Ciênc. Tecnol. Aliment. v.20 n.3 Campinas set./dez. 2000
AFLATOXIN M1 IN SAMPLES OF "MINAS" CHEESE COMMERCIALIZED IN THE CITY OF BELO HORIZONTE MINAS GERAIS/BRAZIL1
Milk products such as cheeses may be contaminated by aflatoxin M1 when dairy cattle have consumed feeds contaminated with aflatoxin B1. Samples of "Minas" cheeses (fresh, canastra and standard) were collected by the Inspection Service in the Mercado Central in Belo Horizonte city, Minas Gerais Brazil. A purified extract was obtained by extraction with dichloromethane followed by a washing with nhexane and immunoaffinity column purification. The quantification of aflatoxin M1 was done by high performance liquid chromatography (HPLC) using a fluorescence detector. Recoveries were about 75%. In 56 of the 75 samples (74.7%), the presence of aflatoxin M1 was detected in concentrations ranging between 0.02 and 6.92ng/g of cheese. In the positive cases (³ 0.02ng/g) the mean contamination level of aflatoxin M1 was 0.08ng/g in fresh cheese, 0.36ng/g in canastra cheese and 0.62ng/g in standard cheese. No aflatoxin M1 maximum tolerance level in cheese has been established in Brazil.
Keywords: Aflatoxin M1; cheese; HPLC.
AFLOTOXINA M1 EM AMOSTRAS DE QUEIJO "MINAS" COMERCIALIZADA NA CIDADE DE BELO HORIZONTE MINAS GERAIS/ BRASIL. Produtos derivados de leite, como queijo, podem estar contaminados com aflatoxina M1 quando o gado leiteiro consome ração contaminada com aflatoxina B1. Amostras de queijo "Minas" ( frescal, canastra e padrão ) foram coletados pela Vigilância Sanitária de Minas Gerais Brasil. Foi obtido um extrato purificado através de extração com diclorometano, seguido de lavagem com n-hexano e purificação em coluna de imunoafinidade. A quantificação da aflatoxina M1 foi feita por cromatografia líquida de alta eficiência (CLAE) usando detetor de fluorescência. Os valores de recuperação foram em torno de 75%. A presença de aflatoxina M1 foi detectada em 56 das 75 amostras (74,7%) e a faixa de concentração observada foi 0,02 a 6,92ng/g de queijo. Dentre os valores positivos (³ 0,02ng/g) a média de contaminação de aflatoxina M1 foi 0,08ng/g para queijo frescal, 0,36ng/g para queijo canastra e 0,62ng/g para queijo padrão. O Brasil não apresenta nível de tolerância máxima para aflatoxina M1 em queijo.
Palavras-chave: Aflatoxina M1; queijo; CLAE.
1 - INTRODUCTION
Aflatoxin M1 (AFM1) is the main aflatoxin B1 metabolite found in the milk of lactating animals that have ingested contaminated feed. The population can therefore be indirectly exposed to aflatoxins by the consumption of milk or milk products such as cheese, yoghurt, yoghurt-cheese [4, 11] .
The presence of AFM1 in milk and milk products is considered to be undesirable because is a toxic and carcinogenic metabolite [12, 15] . This presence is a worldwide concern since these products are consumed largely by children, including infants, who are considered more susceptible to the adverse effects of mycotoxins  .
Stability tests of AFM1 during industrial treatments of pasteurization, sterilization and production of milk and milk products showed a decrease in its level after treatments, but these processes do not destroy AFM1. However, contrasting data have been reported [7, 8, 13] .
AFM1 is associated with the protein fraction of milk, and hence it is present in cheeses aproximately a 3-5 fold enrichment over that in milk [8, 11] .
Many countries have established regulation for AFM1 in cheese: 0.20-0.25ng/g  . However, in Brazil, no AFM1 levels have been officially set for cheese.
Surveys have been carried out in various countries to verify the occurrence of AFM1 in cheese [1, 2, 4, 9,11, 12] , but in Brazil only few data have been published  .
The purpose of this work was to investigate the occurrence of AFM1 in 75 samples of "Minas" cheese, manufactured in Minas Gerais/Brazil.
2 - MATERIALS AND METHODS
2.1 - Samples
A total of seventy-five samples of "Minas" cheese (fresh, canastra, and standard) were analysed. The samples were collected by Inspection Service in the Mercado Central in Belo Horizonte city, Minas Gerais / Brazil (1996-1998). The samples were frozen on collection and kept at 20o C until the time of analysis  .
2.2 Extraction and Quantification of AFM1
The method used to extract AFM1 in cheese was described by DRAGACCI et al  . Purification was done as described by TUINSTRA et al  , using immunoaffinity colum (G 1007- VICAM Inc, USA). The final extract was dissolved in 500-1000mL mobile phase of injection water: acetonitrile (30:70).
The AFM1 contents were then measured with a high-pressure liquid chromatography (HPLC) system, using a fluorescence detector (Shimadzu LC-10 AD Model), 366 nm excitation, 428nm emission, with Shim-Pack CLC-ODS column, 5mm, 4.6 x 250mm, preceded by a guard column Shim-Pack G-ODS, 5mm, 4 x 10mm. The mobile phase was water : isopropanol: acetonitrile (80:12:8) in a flow rate of a 1mL/minute. For quantification a AFM1 calibration curve was used as described by PRADO et al  . All analyses were done in duplicate.
The efficiency of the method was evaluated by the contamination of the "Minas" cheese samples ( standard and fresh ), which were AFM1 free, with 0.24 and 0.48 ng/g of the AFM1 standard. Samples were spiked by mixing them with a known volume of standard AFM1 solution previously diluted in acetonitrile plus water 1:3  .
3 - RESULTS AND DISCUSSION
The recovery rates after spiking at 0.24 and 0.48ng/g of AFM1 were 72.2% and 79.1%, respectively. These results are showed on Table 1. DRAGACCI et al  found in samples contaminated with 0.2 and 0.8ng/g, recovery in the range of 72.0 to 87.6% and 71.4 to 81.0%, respectively. The Coefficients of Variation (C. V.) were satisfactory (i. e. < 20 %).
A summary of the results of the survey, not corrected for recovery, is shown in Table 2. Of the total of 75 samples of cheese analysed, 56 (74.7%) showed contamination with AFM1 and 20 (26.7%) presented levels higher than 0.25 ng/g, which is the tolerance limit to AFM1 in cheese in some European countries  . These results revealed a higher contamination by AFM1 when compared to the results obtained in Europe. In Italy, a total of 223 samples of Grana Padano cheese, manufactured in 4 years (1991-1994), were checked for AFM1. Most samples (91%) were in the range 5-100 ng/Kg and only 1 (0.5%) was > 250ng/Kg  . BARBIERI et al  verified the presence of AFM1 in 200 fresh parmesan cheeses in Modena, Italy, in 1991, and only 18 samples showed AFM1 in low concentrations (max.0.19ng/g). DRAGACCI et al  . A survey for AFM1 in cheese in France (1990-1995), has detected levels of AFM1 higher than 0.2ng/g, which is the maximum acceptable level, in only 3 samples from a total of 257 samples analysed. When investigating 414 samples of Italian, German and French cheeses, OLIVIERO et al  likewise found this mycotoxin at a percentage of 31.4% and only 9 samples exceeded 0.25ng/g. However, BARRIOS et al  , in the South of Spain, verified the presence of AFM1 in 16 of the 35 samples (45.7%) of cheese analysed (9 fresh, 9 semiripened and 17 ripened), in concentrations ranging between 20 and 200ng/g of cheese. SYLOS et al  detected no AFM1 in cheese commercialized, in 1990, in Campinas/Brazil. However, the method used in this paper was a thin-layer chromatography (TLC), which is less sensitive than high performance liquid chromatography (HPLC).
The presence of AFM1 was detected in concentrations between 0.02 and 6.92 ng/g of cheese and the mean contamination level was 0.08ng/g in fresh cheese,0.36 ng/g in canastra cheese and 0.62 ng/g in standard cheese (Table 3). Standard cheese presented levels of AFM1 higher than the detected in fresh and canastra cheese. BARRIOS et al  also observed that the AFM1 concentration detected in the ripened and in the semi-ripened cheeses was higher than that found in the fresh cheeses. This could be justified by the affinity of AFM1 to casein or else a greater analytical recovery related to a decrease in the moisture content of the substrate and to the concentration of protein and fat components of the milk [2, 3] .
In order to produce 1 Kg of cheese, approximately 10 litres of milk are used in its manufacture. If the milk utilized in the cheese production process presents AFM1 levels of 50ng/L or 500ng/L (maximum tolerated level set by European Community Countries and Brazil, respectively), the theoritical contamination found in cheese will be 0,5ng/g or 5,0ng/g . Therefore, the initial proposed regulatory limits for AFM1 in cheese, could be establish in this concentration range.
The results of this study suggest a established regulation for AFM1 for cheese in Brazil and a constant monitoring in this product.
 BARBIERI, G.; BERGAMINI, C.; ORI, E.; RESCA, P. Aflatoxin M1 in Parmesan cheese: HPLC determination. J. Food Sci., v. 59, n. 6, p. 1313-1331, 1994. [ Links ]
 BARRIOS, M. J.; GUALDA, M. J.; CABANAS, J. M.; MEDINA, L. M.; JORDANO, R. Occurrence of aflatoxin M1 in cheeses from the south of Spain. J. Food Prot., v. 59, n. 8, p. 898-900, 1996. [ Links ]
 BRACKETT, R. E. & MARTH, E. H. Fate of aflatoxin M1 in Parmesan and Mozzarella cheese. J. Food Prot., v. 45, p. 597-600, 1982. [ Links ]
 DRAGACCI, S. & FREMY J. M. Application of immunoaffinity column cleanup to aflatoxin M1 determination and survey in cheese. J. Food Prot., v. 59, n. 9, p. 1011-1013, 1996. [ Links ]
 DRAGACCI, S.; GLEIZES, E.; FREMY, J. M.; CANDLISH, A. A. G. Use of immunoaffinity chromatography as a purification step for the determination of aflatoxin M1 in cheeses. Food Addit. Contam., v. 12, n. 1, p. 59-65, 1995. [ Links ]
 FOOD AND AGRICULTURAL ORGANIZATION OF THE UNITED NATIONS. Worldwide regulations for mycotoxins. 1995. A Compedium. n. 64, 45 p. Rome, 1996. [ Links ]
 GALVANO, F.; GALOFARO, V.; GALVANO, G. Occurrence and stability of aflatoxin M1 in milk and milk products: a wordwide review. J. Food Prot., v. 59, n. 10, p. 1079-1090, 1996. [ Links ]
 HASSANIN, N. I. Stability of aflatoxin M1 during manufacture and storage of yoghurt, yoghurt-cheese and acidified milk. J. Sci. Food Agric. v. 65, p. 31-34, 1994. [ Links ]
 KARAIOANNOGLOU, P.; MANTIS, A.; KOUFIDIS, D.; KOIDIS, P.; TRIANTAFILLOU, J. Occurrence of aflatoxin M1 in raw and pasteurized milk and in Feta and Teleme cheese samples. Milchwissenschaft, v. 44, n. 12, p. 746-748, 1989. [ Links ]
 OLIVIERO, G. L.; SERPE, G.; MONTELA, G.; MARTINO, F. Aflatoxin M1 in milk samples at the Naples Central Dairy. Ind. Aliment., v. 26, p. 1161-1164, 1987. [ Links ]
 PIETRI, A.; BERTUZZI, T.; BERTUZZI, P.; PIVA, G. Aflatoxin M1 occurrence in samples of Grana Padano cheese. Food Addit. Contam., v. 14, n. 4, p. 341-344, 1997. [ Links ]
 PIVA, G.; PIETRI, A.; GALAZZI, L.; CURTO, O. Aflatoxin M1 occurrence in dairy products in Italy. Food Addit. Contam., v. 5, n. 2, p. 133-139, 1987. [ Links ]
 PRADO, G.; NICÁCIO, M. A. S.; LARA, M. A. Incidência de aflatoxina M1 em leite cru e em pó no Estado de Minas Gerais. Hig. Alim., v. 8, p. 34-36, 1994. [ Links ]
 PRADO, G.; OLIVEIRA, M. S.; ABRANTES, F. M.; SANTOS, L. G.; SOARES, C. R.; VELOSO, T. Ocorrência de aflatoxina M1 em leite consumido na cidade de Belo Horizonte Minas Gerais / Brasil Agosto/98 à Abril/99. Ciênc. Tecnol. Aliment., v. 19, n. 3, p. 420-423, 1999. [ Links ]
 SAITANU, K. Incidence of aflatoxin M1 in Thai milk products. J. Food Prot., v. 60, n. 8, p. 1010-1012, 1997. [ Links ]
 SYLOS, C. M.; RODRIGUEZ-AMAYA, D. B.; CARVALHO, P. R. N. Occurrence of aflatoxin M1 in milk and dairy products commercialized in Campinas, Brazil. Food Addit. Contam., v.13, n. 2, p. 169-172, 1996. [ Links ]
 TUINSTRA, L. G. M.; ROOS, A. H.; TRUP, J. M. P. Liquid chromatographic determination of aflatoxin M1 in milk powder using immunoaffinity columns for cleanup: Interlaboratory study. Food Biol. Contam., v. 76, n. 6, p. 1248-1254, 1993. [ Links ]
The authors wish to thank the CNPq and Health Ministry for the financial support.
1 Recebido para publicação em 12/09/2000. Aceito para publicação em 23/01/01.
2 Fundação Ezequiel Dias Instituto Octávio Magalhães. Núcleo de Micologia e Micotoxinas. Rua Conde Pereira Carneiro, 80. Gameleira. CEP 30510-010. Belo Horizonte, Minas Gerais / Brasil. Tel: (31) 33719462 Fax: (31) 33719553. e-mail: firstname.lastname@example.org
* A quem a correspondência deve ser enviada.