A¯ atoxin M1 in raw and ultra high temperature-treatedmilk commercialized in Portugal

Maria Lõ  gia Martins* and Hermõ  nia MarinaMartinsL aborato rio Nacional de InvestigacË aÄ o V eterina ria, Departamentode Higiene Publica/Micologia, Estrada de Ben® ca, 701, 1549- 011L isbon, Portugal

(Received 29 February 2000; revised 28 April 2000; accepted19 May 2000)

During June to September 1999, 31 samples of rawmilk were obtained from individual farms and 70samples of ultrahigh-temperature-treated (UHT ) milk(18 samples of whole milk, 22 of semi-skimmed and 30samples of skimmed milk) were collected in super-markets in L isbon, Portugal. The total number ofsamples analysed was 101. The incidence of a¯ atoxinM1 (AFM1) contamination in the milk samples ana-lysed was very high (83.2% ). AFM1 was detected in80.6% of raw milk and 84.2% of UHT milk.Concerning raw milk, 17 samples (54.8% ) containedlow levels (0.005± 0.010 ·g/l), two samples (6.5% ) hadlevels ranging from 0.011 to 0.020 ·g/l and six samples(19.3% ) had levels between 0.021 and 0.050 ·g/l. Of70 samples of UHT milk analysed, 10 samples hadlevels below 0.005 ·g/l (one whole milk, two semi-skimmed milk and seven skimmed milk). NinesamplesÐ two whole milk (11.1% ) and seven skimmedmilk (23.3% )Ð were contaminated with levels between0.005 and 0.010 ·g/l. Twenty-® ve samples (eight wholemilk, 44.4% ; one semi-skimmed milk, 4.5% ; and 16skimmed milk, 53.4% ) had AFM1 contamination levelsof 0.011± 0.020 ·g/l. Six samples of whole milk(33.3% ) and 18 samples of semi-skimmed milk(81.9% ) had levels of 0.021± 0.050 ·g/l. Only twosamples were contaminated at levels above the legallimit for AFM1 (one whole milk and one semi-skimmedmilk, 0.059 and 0.061 ·g/l, respectively). At the mo-ment, the contamination of milk with AFM1 does notappear to be a risk to public health, although the

presence of this toxin was detected in 83.2% of thesamples analysed.

Keywords : a¯ atoxin M1, HPLC, milk, occurrence

Introduction

A¯ atoxin M1 (AFM1 ) occurs as a hydroxylated me-tabolite of a¯ atoxin B1 (AFB1 ) in milk from man andanimals consuming feed contaminated withAspergillus ¯ avus, A. parasiticus or A. nomius(Betina 1989). It is also present in the organs, e.g.kidneys and liver, and excreta of animals exposed toAFB1 (Allcrof t et al. 1966 ). AFM1 is a relativelystable metabolite, though it is mutagenic and carci-nogenic (Green et al. 1982).

AFM1 is less carcinogenic than AFB1 according togenotoxicity studies with rainbow trout and rats(Sinnhuber et al. 1974, Canton et al. 1975).

The amounts of AFM1 excreted in milk as a percent-age of AFB1 averages 1± 2%, varying from animal toanimal, from day to day and from one milking to theother. The AFM1 could be detected in milk 12± 24 hafter the ® rst AFB1 ingestion, reaching a high levelafter a few days. When the intake of AFB1 is ® nished,the AFM1 concentration in the milk decreases to anundetectable level after 72 h (van Egmond 1989).

As milk is a major nutrient for infants, children,convalescents and old people, the presence of AFM1in commercial milk and dairy products is a concern.

The residues of AFM1 remain stable when milk isprocessed by heat or is fermented. There is no evi-dence that cold storage, freezing, heat-treating, fer-menting, concentrating or drying of the contaminatedmilk changes the level of AFM1 (Yousef and Marth1985). Nevertheless, the partition of AFM1 variesaccording to the manufactured product ; therefore,cheese, cream and butter made from contaminatedmilk contain ca 40%± 60%, ca 10% and less than 2%

Food Additives and Contaminants , 2000, Vol. 17, No. 10, 871 ± 874

*To whom correspondence should be addressed.

Food Additives and Contaminant s ISSN 0265± 203X print/ISSN 1464± 5122 online # 2000 Taylor & Francis Ltdhttp://www.tandf.co.uk/journals

of the AFM1 present in milk, respectively (Brackettand Marth 1982, van Egmond 1983).

In Portugal, very few literature data are available onthe occurrence of AFM1 in raw and commercial milk.Several countries control contamination of milk anddairy products by AFM1, and they have establishedregulations for AFM1 in these products. The currentpermissible levels for AFM1 in milk range from 0.05to 0.5 mg/kg, except for infant milk for which lowerlevels exist (FAO 1997). In Portugal the o� cialregulatory limits for a¯ atoxins B1, B2, G1 and G2 infoods and a¯ atoxin B1 in feeds are established by theRegulation (CE) 1525/98 (O� cial Journal of theEuropean Communities 1998) and Directive 199/29/CE (O� cial Journal of the European Communities1999).

The aim of this study was to investigate the occur-rence of AFM1 in raw milk and in UHT milkmarketed in Portugal.

Materials and methods

Samples

During June to September 1999, 31 samples of rawmilk were obtained from individual farms and 70samples of ultrahigh-temperature-treated (UHT) milk(18 samples of whole milk, 22 of semi-skimmed and30 samples of skimmed milk) were collected in super-markets in Lisbon, Portugal. The total number ofsamples analysed was 101.

The samples of raw milk obtained were kept on iceduring transportation to the laboratory and analysedon the next day.

Method of analysis

AFM1 in milk was extracted using an immunoa� nitycolumn and quanti® ed by high pressure liquid chro-matography (HPLC) according to the method de-scribed by the International Dairy Federation (IDF1995). One hundred ml of warm milk (35± 378C) were® ltered through Whatman No. 4 ® lter paper and50 ml were collected in a syringe barrel and passed

through the immunoa� nity column (Rhoà ne-diagnos-tics Technologies Ltd ref. P04). Then the column waswashed with 10 ml of water and the AFM1 was slowlyeluted with 4 ml of acetonitrile. The ® nal extract wasevaporated and redissolved in 5 ml of water.

Standard AFM1 was purchased from Sigma (A-6428)(Quimica S.A., Spain). The stock solution and work-ing standards were prepared as described by Tuinstraet al. (1993). The concentration of stock solution wasdetermined by UV spectophotometry (U 2000-Hitachi) (IDF 1995).

Determinations of AFM1 levels in milk extracts werecarried out by isocratic HPLC using a LiChrospher100 RP-18, 5 mm column 25 £ 4.6 mm i.d. ) EcoPack(Merck, Portugal), with ¯ uorescence detector andcomputing integrator Merck Hitachi (CompaqDeskpro) ; excitation and emission wavelengths were360 nm and 435 nm. The mobile phase was acetoni-trile± water (25 :75) and the ¯ ow rate was 0.8 ml/min.The calibration curve was determined, using a seriesof dilutions containing di� erent levels of a¯ atoxin M1and 10% acetonitrile was used as the diluent(Tuinstra et al. 1993).

Recoveries of AFM1 added to milk at 0.05, 0.15 and0.20 mg /l averaged 97% (CV = 11.6%).

Samples with AFM1 levels below 0.005 mg/l wereconsidered negative, and for positive samples thecon® rmation of identity of AFM1 was performed asprescribed by AOAC (1990).

Results and discussion

AFM1 was determined in 101 samples of milk col-lected from individual farms (raw milk, n ˆ 31) andfrom di� erent supermarkets (UHT milk, n ˆ 70)during the summer season of 1999. The incidence ofAFM1 contamination in milk samples was very high(83.2%). AFM1 was detected in 80.6% of raw milk(table 1) and in UHT milk the percentage of positiveswas : 94.4% for whole milk, 90.9% for semi-skimmedmilk and 76.7% for skimmed milk (table 1).Concerning raw milk, six samples (19.4%) had levelsbelow 0.005 mg/l ; in 17 samples (54.8%) AFM1 wasdetected at low levels (0.005 to 0.010 mg/l) (table1) ;two samples (6.5%) had levels of 0.011 and 0.020 mg/land six samples (19.3%) had levels between 0.021 and0.050 mg/l (table 1). Of 70 samples of UHT milk

872 M. L . Martins and H. M. Martins

analysed, 10 samples had levels below 0.005 mg/l (onewhole milk, two semi-skimmed milk and sevenskimmed milk) (table 1). Nine samples (two wholemilk, 11.1% and seven skimmed milk, 23.3%) werecontaminated with levels between 0.005 and 0.010 mg/l.Twenty-® ve samples (eight whole milk, 44.4%; onesemi-skimmed milk, 4.5%; and 16 skimmed milk,53.4%) had contamination levels of 0.011 to0.020 mg/l. Six samples of whole milk (33.3%) and18 samples of semi-skimmed milk (81.9%) had anincidence of 0.021 to 0.050 mg/l. Only two sampleswere contaminated at levels above the EU legal limitof 0.05 mg/kg (one whole milk and one semi-skimmedmilk , 0.059 and 0.061 mg/l, respectively ) (table 1).

The overall incidence of contamination with AFM1was higher than in another survey carried out in 1982in Portugal (Ouakinin and Martins 1982). The resultsobtained in this ® rst survey showed that of 74 samplesanalysed, only 39% were contaminated by AFM1with levels ranging from 0.060 to 0.065 mg/l and thisshould be attributed to the limit of detection of theTLC analytical method used (0.050 mg/l). In this pres-ent survey, the AFM1 in milk was extracted using theimmunoa� nity column, with increased speci® city,since speci® c antibodies are used that bind with anyAFM1 (antigen) present in the sample of milk. Thenewer analytical methods and equipment also allowmuch lower limits of detection (0.005 mg/l).

Galvano et al. (1998) detected AFM1 in 97 samples ofdry milk for infant formula using HPLC; in 81samples (84%) amounts ranged from <0.001 to0.101 mg/l. The occurrence of AFM1 in commercialpasteurized milk as determined with ELISA andHPLC (Markaki and Melissari 1997) showed that of81 samples analysed, 32 contained AFM1 at levels of0.003± 0.005 mg/l, none had more than 0.005 mg/l, and31 contained only traces of AFM1. In Brazil, De Syloset al. (1996) found only four samples contaminatedwith 0.073± 0.370 mg/l in 52 samples analysed.

According to Trucksess (1997), in samples collectedfrom Argentina, Brazil, Cyprus, Egypt, Germany,India, Italy and Spain the results indicated thatAFM1 levels in samples from Germany were thelowest (the maximum contamination level found intank truck milk was 0.026 mg/l). Trucksess (1998,1999) reported that in recent European studies oncontamination of milk and milk products with AFM1the levels were very low.

It seems that the results of the present survey aresimilar to those presented in other European studies.Therefore we can conclude that the incidence ofAFM1 in milk commercialized in Lisbon, Portugal,does not appear to be a serious public health hazardat the moment, since only two samples exceeded theEU limit of 0.05 mg/kg milk (FAO 1997). However,more samples will have to be analysed and surveysconducted over a more extended period.

Acknowledgements

The authors wish to thank Mr Aure lio Caldeira, ofthe Technical Department of Merck Portugal for hisever-willing collaboration and to Anabela Ramos andM. Candida Camejo for their technical assistance.

References

ALLCROFT, R., ROGERS, H., LEWIS, G ., NABNEY, J., and BEST, P. E.,1966, Metabolism of a¯ atoxin in sheep: excretion of the `milktoxin’ . Nature (L ondon), 209, 154± 155.

AOAC, 1990, O� cial Methods of Analysis. V olume II (Arlington, VA :Association of O� cial Analytical Chemists), section 980.21F.

873A¯ atoxin M1 in milk in Portugal

Table 1. Incidence and levels of a¯ atoxin M1 in raw milk and UHT milk.

Incidence Levels in mg/l

Sample No% < 0.005 a 0.005± 0.010 0.011± 0.020 0.021± 0.050 >0.051

Raw milk 25/31 (80.6) 6 (19.4%) 17 (54.8%) 2 (6.5%) 6 (19.3%) 0UHT-whole milk 17/18 (94.4) 1 (5.6%) 2 (11.1%) 8 (44.4%) 6 (33.3%) 1 (5.6%)UHT semi-skimmed milk 20/22 (90.9) 2 (9.1%) 0 1 (4.5%) 18 (81.9%) 1 (4.5%)UHT skimmed milk 23/30 (76.7) 7 (23.3%) 7 (23.3%) 16 (53.4%) 0 0

a <0.005 m/lÐ negative for AFM1.

BETINA, V., 1989, A¯ atoxins, sterigmatocystins and versicolorins.Mycotoxins: Chemical, Biological and Environmental Aspects(Elsevier: New York), pp. 114± 150.

BRACKETT, R . E., and MARTH, E. H., 1982, Fate of a¯ atoxin M1 inCheddar cheese and in process cheese spread. Journal of FoodProtection, 45, 549± 552.

CANTON, J. H., K ROES, R., VAN LOGTEN, M . J., VAN SCHOTHORST,M., STAVENUITER, J. F. C., and VERHULSDONK, C. A. H.,1975, The carcinogenicity of a¯ atoxin M1 in rainbow trout.Food and Cosmetics T oxicology, 13, 441± 443.

DE SYLOS, C. M ., R ODRIGUEZ-AMAYA, D.B., and CARVALHO, P. R .,1996, Occurrence of a¯ atoxin M1 in milk and dairy productscommercialized in Campinas, Brazil. Food Additives andContaminants, 13, 169± 172.

FAO (FOOD AND AGRICULTURE ORGANIZATION), 1997, W orldwideRegulations for Mycotoxins 1995. A Compendium. FAOFood and Nutrition Paper 64 (Rome: Food and AgricultureOrganization of the United Nations).

GALVANO, F ., GALOFARO, V., DE ANGELIS, A., GALVANO, M .,BOGNANNO, M ., and G ALVANO, G., 1998, Survey of the occur-rence of a¯ atoxin M1 in dairy products marketed in Italy.Journal of Food Protection, 61, 738± 741.

GREEN, C. E., R ICE, D . W., HSIEH, D . P. H., and BYARD, J. L., 1982,The comparative metabolism and toxic potency of a¯ atoxin B1

and a¯ atoxin M1 in primary cultures of adult rat hepatocytes.Food and Chemical Toxicology, 20, 53± 60.

IDF (INTERNATIONAL DAIRY FEDERATION), 1995, STANDARD171:1995. Determination of A¯ atoxin M1 Content. Clean± up byImmunoa� nity Chromatography and Determination by High-performance L iquid Chromatography (Brussels : IDF).

MARKAKI,P., and MELISSARI,E.,1997, Occurrence of a¯ atoxin M1 incommercial pasteurized milk determined with ELISA andHPLC. Food Additives and Contaminants, 14, 451± 456.

O� cial Journal of the European Communities, 1998, L201, 17/7/98.Reg. (CE) No 1525/98.

O� cial Journal of the European Communities, 1999, L115, 4/5/99(Directive 199/29/CE).

OUAKININ, J., and MARTINS M . L., 1982, Ocorreà ncia de A¯ atoxinaM1 nos Leites. Repositorio de Trabalhos Instituto Nacional deV eterinaria, XIV, 75± 78.

SINNHUBER, R., LEE, D . J., WALES, J. H ., LANDERS, M . K ., andK EYL, A. C., 1974, Hepatic carcinogenesis of a¯ atoxin M1 inrainbow trout (Salmo gairdneri). Journal of the National CancerInstitute, 53, 1285± 1288.

TUINSTRA, L. G. M. TH., R OOS, A. H., and VAN TRIJP M . P., 1993,Liquid chromatographic determination of a¯ atoxin M1 in milkpowder using immunoa� nity columns for cleanup : interlabora-tory study. Journal of AOAC International, 76, 1248± 1254.

VAN EGMOND, H. P., 1983, Mycotoxins in dairy products. FoodChemistry, 11, 289± 307.

VAN EGMOND, H. P.,1989, A¯ atoxin M1 : occurrence, toxicity, regula-tion. Mycotoxins in Dairy Products, edited by Hans P. vanEgmond (New York : Elsevier Applied Science), pp.11± 55.

TRUCKSESS,M .W.,1997, Mycotoxins. Journal of AOAC International,80, 119± 126.

TRUCKSESS,M .W.,1998, Mycotoxins. Journal of AOAC International,81,128± 137.

TRUCKSESS,M .W.,1999, Mycotoxins. Journal of AOAC International,82, 488± 495.

YOUSEF,A.E., and MARTH, E. H.,1985, Degradation of a¯ atoxin M1in milk by ultraviolet energy. Journal of Food Protection, 48,697± 698.

874 A¯ atoxin M1 in milk in Portugal

Copyright of Food Additives & Contaminants is the property of Taylor & Francis Ltd and its content may not

be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.