52_220

Original PaperOriginal Paper

Method for Determination of Aatoxin M1 in Cheese and Butterby HPLC Using an Immunoa$nity Column

(Received December 2, 2010)

Hisako S6@JB61, , Yoichi K6B6I61, Yoshiko SJI6-KDC>H=>1 and Hiroshi K6L6@6B>2

1National Institute of Health Science: 1181 Kamiyoga, Setagaya-ku, Tokyo 1588501, Japan;2Kyoritsu Womens University: 221 Hitotsubashi, Chiyoda-ku, Tokyo 1018437, Japan;

Corresponding author

A rapid, sensitive convenient method for determination of aatoxin M1 (AFM1) in cheese andbutter by HPLC was developed and validated. The method employs a safe extraction solution(mixture of acetonitrile, methanol and water) and an immunoa$nity column (IAC) for clean-up.Compared with the widely used method employing chloroform and a Florisil column, the IACmethod has a short analytical time and there are no interference peaks. The limits of quantica-tion (LOQ) of the IAC method were 0.12 and 0.14 g/kg, while those of the Florisil column methodwere 0.47 and 0.23 g/kg in cheese and bu#er, respectively. The recovery and relative standarddeviation (RSD) for cheese (spiked at 0.5 g/kg) in the IAC method were 92 and 7, respectively,while for the Florisil column method the corresponding values were 76 and 10. The recoveryand RSD for butter (spiked at 0.5 g/kg) in the IAC method were 97 and 9, and those in theFlorisil method were 74 and 9, respectively. In the IAC method, the values of in-houseprecision (n2, day5) of cheese and butter (spiked at 0.5 g/kg) were 9 and 13, respectively.The IAC method is superior to the Florisil column method in terms of safety, ease of handling,sensitivity and reliability. A survey of AFM1 contamination in imported cheese and butter inJapan was conducted by the IAC method. AFM1 was not detected in 60 samples of cheese and 30samples of butter.

Key words: aatoxin M1; cheese; butter; immunoa$nity column; HPLC

Introduction

Aatoxins, a group of potent carcinogenic and terato-genic compounds, are secondary metabolic products ofAspergillus avas, A. parasiticus and A. nomius, whichmay grow in various agricultural commodities.Aatoxin B1 (AFB1), the most toxic compound of thegroup, is metabolized by the hepatic microsomal mixed-function oxidase system1) Aatoxin M1 (AFM1) is anAFB1 metabolite that may contaminate mammalianmilk2) Since AFM1 is a stable compound, it may alsocontaminate dairy products such as cheese and buttermanufactured with contaminated milk.

The experimentally determined carcinogenic potencyof AFM1 is one order of magnitude less than that ofAFB13) AFM1 was classied as a possibly carcinogeniccompound to humans (Group 2B) by the InternationalAgency for Research on Cancer (IARC)4)

Many countries have established regulations forAFM1 in milk and dairy products. In 2001, the CodexAlimentarius Commission established the maximumpermitted level (MPL) of 0.5 g/kg for AFM1 in milk3)The MPL for milk is 0.5 g/kg in the United States1

1 U.S. Compliance Policy Guideline 527.400.

and 0.05 g/kg in the European Union5) An MPL of0.25 g/kg for AFM1 in cheese has been established bySwitzerland and Austria6)

Many analytical methods, including TLC, HPLC andLCMS/MS, have been used to determine aatoxins.Immunoa$nity columns (IACs)7) and solid-phaseextraction8) are available for clean-up of AFM1. Formilk, IAC is often used as a clean-up procedure, and hasbeen adopted in the O$cial Methods of Analysis ofAOAC International9) However, no convenient andreliable method for AFM1 determination in cheese andbutter has been validated. Methods to determine AFM1in foods using Florisil columns with chloroform aregenerally accepted and have been used for a survey ofAFM1 in cheese in Japan8) However, chloroform isdened by the IARC as a possibly carcinogenic com-pound to humans (Group 2B)10) Although some studiesin Europe used IACs to clean-up AFM1 in cheese,dichloromethane was used as the extract solvent11), 12)Dichloromethane is also dened as a possibly carci-nogenic compound to humans (Group 2B)13)

Therefore, in this study, we developed a precise, accu-rate, sensitive, convenient and safe analytical methodfor determining AFM1 in cheese and butter. Wevalidated this newly developed IAC method in a singlelaboratory according to the Japanese guideline for pesti-

Food Hyg. Saf. Sci. Vol. 52, No. 4220

cides2 We then employed this validated method tosurvey AFM1 in 60 samples of imported cheese and 30samples of imported butter.

Materials and Methods

Reagents and samplesAFM1 was purchased from Wako Pure Chemical

Industries, Ltd. (Osaka, Japan). A stock solution ofAFM1 (10 g/mL) in acetonitrile was prepared in ambersilized vials (GL Sciences Inc., Tokyo, Japan) and storedat 20 in the dark. Acetonitrile and methanol wereof high performance liquid chromatography (HPLC)grade (Wako Pure Chemical Industries, Ltd.).Phosphate-bu#ered saline (PBS) was prepared byreconstitution of PBS tablets (MP Biomedicals, LLC,Santa Ana, CA, USA) in water. Cheese (Cheddar) andbutter were purchased from supermarkets located inTokyo, and stored at 4 until use.

Florisil column clean-up procedureCheese and bu#er were cleaned up by the method of

Kamimura et al.8) with a slight modication. Briey, 20 gof ground cheese was placed in a 500-mL blender jar.Ten mL of saturated saline and 100 mL of chloroformwere added, and the sample was homogenized in ahomogenizer (Nihonseiki Kaisha Ltd., Tokyo, Japan) at4,000 rpm for 3 min, and ltered through a No. 4-lterpaper (Whatman plc, Maidstone, Kent, UK). Fifty mL ofthe chloroform layer of the clear ltrate was placed in a50-mL graduated cylinder with a glass stopper. Afteraddition of 10 g of sodium sulfate anhydrous (Na2SO4),the cylinder was vigorously shaken, and this chloro-form solution was used for analysis.

Ten g of chopped butter was placed in a 100-mLblender jar, and 40 mL of chloroform was added. Thesample was homogenized at 4,000 rpm for 2 min. Thesample was maintained at 20 for 15 min, and thenltered through a No. 4-lter paper.

A Florisil column (Sep-Pak Plus, Waters, Milford, MA,USA) was equilibrated in 30 mL of chloroform. In thecase of cheese, the analysis sample (50 mL) was loadedonto the Florisil column. In the case of butter, the entireltrate was loaded onto the Florisil column. In bothcases, the column was washed with 20 mL chloroformand 30 mL of chloroformmethanol (9 : 1, v/v), andAFM1 was eluted with 30 mL of acetonewater (99 : 1,v/v). The eluate was collected in a ask and evaporatedusing a rotary evaporator. The residue was subjectedto HPLC.

Immunoa$nity column clean-up procedureTen g of ground cheese was placed in a 100-mL

blender jar and a mixture of 40 mL of acetonitrilemethanolwater (6 : 1 : 3, v/v/v) was added. Thesample was homogenized at 4,000 rpm for 5 min, then

2 Guideline of method validation of residue pesticides infood (November 15, 2007, Notication No. 1115001). TheMinistry of Health, Labor and Welfare of Japan.

centrifuged with a centrifuge (Kubota Corp., Tokyo,Japan) at 3,000 rpm for 5 min. Ten mL of the superna-tant was diluted in 30 mL of PBS and ltered through aglass lter (934AH, Whatman plc, Maidstone, Kent, UK).

Ten g of chopped butter was placed in a 100-mLblender jar and 40 mL of the mixture of acetonitrilemethanolwater (6 : 1 : 3, v/v/v) was added. Thesample was homogenized at 4,000 rpm for 2 min andltered through a No. 4-lter paper. Ten mL of theresulting ltrate was diluted in 30 mL of PBS, andltered through a 934AH glass lter.

An AFLAKING IAC (Horiba, Ltd., Kyoto, Japan),which was reported by Uchigashima et al.14) to be toler-ant to organic solvents, was used in this study. ThisIAC is compatible with 40 methanol or 20 ac-etonitrile. For both cheese and butter, the IAC wasconditioned with 10 mL of PBS and then 20 mL of theltrate was loaded onto the column. The column wasthen washed with 10 mL of PBS and 10 mL of water.AFM1 was then eluted with 1 mL of acetonitrile threetimes, with gravity ow. The eluate was evaporatedunder nitrogen gas. The residue was subjected toHPLC.

Fortied testAFM1 was spiked into cheese and butter at 0.5 g/kg

for the Florisil column procedure. For the IAC proce-dure, AFM1 was spiked at 0.1 and 0.5 g/kg. Thesamples were held at room temperature in the dark for1 hr according to Japanese guideline for pesticides2and the method of Itoh et al.15), and then extractionsolvents were added as described above.

HPLCThe HPLC procedure adapted for cheese and bu#er

followed the method reported by Nakajima et al.16)The residue was dissolved in 1.0 mL of acetonitrilewater (2 : 8, v/v), and the solvent was ltered through a0.45-m lter (Sartorius Stedim Biotech GmbH,Goettingen, Germany). Standard solutions preparedfrom the stock solution of AFM1 (10 g/mL) were alsoevaporated under nitrogen gas and dissolved in 1.0 mLof acetonitrilewater (2 : 8, v/v). HPLC was performedwith a Shimadzu LC-10A apparatus (Shimadzu, Kyoto,Japan) equipped with a RF-10AXL uorescence detec-tor (excitation at 365 nm and emission at 435 nm). AC18-column (4.6250 mm, 5 m of particles) with aguard column (Inertsil ODS-3, GL Sciences Inc., Tokyo,Japan) was used at 40 with a CTO-10A column oven.The composition of the mobile phase was acetonitrilewater (25 : 75, v/v) at a ow rate of 1.0 mL/min. Injec-tion volumes of samples and standard solutions were 50L and a SIL-10AF auto-injector was used. The concen-tration range of standard solutions for the Florisilcolumn procedure was 0.2510 g/mL with the middlesensitivity setting of the detector, and the concentra-tion range of standard solutions for the IAC procedurewas 0.11.0 g/mL with the high sensitivity setting.

Determination of Aatoxin M1 in Cheese and ButterAugust 2011 221

Evaluation of the methods for quantifying AFM1The Florisil column procedure was compared with

the IAC procedure (n7). Limits of detection (LOD)were calculated at a signal-to-noise (S/N) ratio of 3.3 : 1,and limits of quantication (LOQ) were calculated at aS/N ratio of 10 : 1. The IAC procedure was alsoevaluated by means of a 5-day single laboratory valida-tion (0.5 g/kg spiked, n2 per day) according to Japa-nese guidelines for pesticides2

Surveillance of AFM1 in cheese and butterThe occurrence of AFM1 in 60 cheese and 30 butter

samples was examined by using the validated IACmethod in this study.

Imported cheese was either Cheddar or Gouda im-ported from The Netherlands, Denmark, Germany, theU.S.A., New Zealand or Australia. Butter was importedfrom The Netherlands, Denmark, the U.S.A., New Zea-land or Australia.

Results

Comparison of Florisil and immunoa$nity column clean-up procedures

For the Florisil column procedure, chloroform is usedas the extraction solvent to obtain a high recovery andas the washing solvent, but it is classied as a carcino-gen. However, for the IAC procedure, methanol andacetonitrile (which are relatively safe) provided high

recovery. The IAC procedure required less time, espe-cially in the column clean-up step, than the Florisilprocedure.

Table 1 compares the results obtained for cheeseusing the IAC and Florisil column procedures. Therecovery was 76 and the relative standard deviation(RSD) value was 10 in the Florisil procedure forsamples spiked with 0.5 g/kg AFM1, but the corre-sponding values were 92 and 7 for the IAC proce-dure. For samples spiked with 0.1 g/kg, recovery was90 and RSD was 10 in the IAC procedure. The LODin the Florisil column procedure was 0.15 g/kg andthe LOQ was 0.47 g/kg, while the correspondingvalues were 0.04 and 0.12 g/kg for the IAC procedure.HPLC chromatograms of AFM1 in cheese are shown inFig. 1. A huge peak was detected before the AFM1 peak,in the chromatogram in the Florisil column procedure(Fig. 1A), but in the IAC procedure there was no inter-fering peak before the AFM1 peak (Fig. 1B).

Table 2 shows a comparison of parameters for butterobtained by the two procedures. For the sample spikedwith 0.5 g/kg AFM1, the recovery was 74 and theRSD value was 9 in the Florisil procedure while in theIAC procedure these values were 97 and 9, respec-tively. For the sample spiked with 0.1 g/kg AFM1, therecovery was 105 and RSD was 11 in the IACprocedure. LOD in the Florisil column procedure was0.08 g/kg and LOQ was 0.23 g/kg. LOD and LOQ in

Table 1. Comparison of detection methods using Florisil column andimmunoa$nity column method to detect aatoxin M1 in cheddarcheese

Method withSpiked aatoxin M1

(g/kg)Parameter

Recovery () RSD ()

Florisil column 0.5 76 10Immunoa$nity column 0.1 90 10

0.5 92 7

n7

Fig. 1. Typical HPLC chromatograms of aatoxin M1 in cheddar cheese

The tested cheese was spiked with aatoxin M1 at 0.5 g/kg. A: chromatogram obtained using Florisil columnprocedure, B: chromatogram obtained using IAC procedure


the IAC procedure were 0.05 and 0.14 g/kg, respective-ly. HPLC chromatograms of AFM1 in butter are shownin Fig. 2. Before the AFM1 peak, an interfering peakwas detected in the chromatogram of the Florisilcolumn procedure (Fig. 2A), but the chromatogram ofthe IAC procedure showed a clear baseline (Fig. 2B).

Single laboratory validation of the IAC procedureWe conducted a single laboratory validation accord-

ing to the Japanese guideline for pesticides2 using IACfor AFM1 in cheese and butter. The recovery,repeatability and in-house precision of AFM1 were 85,5 and 9 in cheese, and 91, 9 and 13 in butter,respectively.

Surveillance of AFM1 in cheese and butterAFM1 was not detected in any of the 60 samples of

cheese and 30 samples of butter. Chromatograms of 6samples of cheese showed an AFM1 peak, but the con-centrations were under the LOD. None of thechromatograms of butter showed any peak around theAFM1 retention time.

Discussion

Due to the slightly polar character of AFM1, solventssuch as methanol, acetone, chloroform, and combina-tions of these solvents can be used to extract AFM1from various kinds of food materials. It was reportedthat a mixture of acetonitrile, methanol and water wasuseful for extraction of mycotoxins in foods containinghigh concentrations of salts or sugars, such as choco-

late, a#ording a high recovery15) In this study weapplied this extraction solvent to cheese and butter, asa convenient procedure to extract AFM1.

It is necessary, when using IAC, to dilute the extractwith water or PBS to reduce the amount of organicsolvent17) The IAC used in this study was compatiblewith an organic solvent concentration of up to 17.5.We could apply the extract to IAC without much dilu-tion, so we could obtain good recoveries of AFM1 witheasy handling in a short time.

In cheese (Table 1), the recovery rate of the Florisilcolumn procedure was acceptable at 0.5 g/kg AFM1but was lower than that of the IAC procedure. RSDvalues of both clean-up procedures were acceptable at0.5 g/kg. When AFM1 was spiked at 0.1 g/kg, theIAC procedure showed an acceptable recovery rate andRSD value. The LOD and LOQ of the IAC procedurewere lower than those of the Florisil column procedure.The eluate from the Florisil column showed an interfer-ence peak ahead of the AFM1 peak (Fig. 1). The HPLCinjection sample size was 10 g sample equivalent in theFlorisil column procedure or 1.25 g sample equivalentin the IAC procedure. Interference in chromatogramsinuences the S/N ratio, so LOD and LOQ in the IACprocedure were lower than those in the Florisil columnprocedure.

In the case of butter (Table 2), the recovery rate in theFlorisil column procedure was acceptable from samplesspiked with 0.5 g/kg, but was lower than that of theIAC procedure. RSD values of both clean-up procedureswere acceptable at 0.5 g/kg. When samples were

Table 2. Comparison of detection methods using Florisil column andimmunoa$nity column method to detect aatoxin M1 in butter

Method withSpiked aatoxin M1

(g/kg)Parameter

Recovery () RSD ()

Florisil column 0.5 74 9Immunoa$nity column 0.1 105 11

0.5 97 9

n7

Fig. 2. Typical of HPLC chromatograms of aatoxin M1 in butter

The tested butter was spiked with aatoxin M1 at 0.5 g/kg. A: chromatogram obtained using Florisil columnprocedure, B: chromatogram obtained using IAC procedure


spiked with 0.1 g/kg, the IAC procedure showed ahigh recovery rate and an acceptable RSD value. TheLOD and LOQ in the IAC procedure were also lowerthan in the Florisil column procedure. In the Florisilprocedure, a large peak appeared before the AFM1 peakon the chromatogram (Fig. 2), as was observed forcheese.

The Florisil column procedure requires chloroform,which is hazardous, for extracting AFM1. It is also moretime-consuming and complicated than the IAC proce-dure.

The IAC procedure was superior to the Florisilcolumn procedure in terms of performance, speed andsafety. The mixture of acetonitrile, methanol, andwater was very e#ective as an extraction solvent in theanalysis of AFM1 in cheese and butter.

The results of this single laboratory validation of theIAC procedure for the analysis of AFM1 in cheese andbutter revealed that the IAC procedure is suitable forthe determination of AFM1 in cheese and butter. TheJapanese guideline of method validation for pesticides2requires that recovery should be 70 to 120 and RSDunder 30 with an in-house precision 35. Theresults obtained from this study satised these condi-tions for AFM1 at 0.5 g/kg in cheese and butter. Toconrm that butter is a suitable sample, we performed apreliminary test using the IAC procedure at 0.5 g/kgas follows. The butter samples were melted at 40, andfortied with AFM1. After 1 hr, the butter wasresolidied at 4. The recovery of the preliminary testwas 88 (n2), and was the same as that of thefortied test according to Japanese guideline for pesti-cides. Therefore, the Japanese guideline method wasused for butter in this study. The commission directive98/53/EC18) on validation procedures of mycotoxinsdescribes the enforcement criteria for o$cial methods.According to these criteria, recovery is 70 to 120 andRSD is consistent with the Horwits equation. The IACprocedure also satised these criteria.

Using the IAC procedure, which was validated in thisstudy, AFM1 was not detected in imported cheese andbutter in Japan. In butter, AFM1 was undetectable.However, chromatograms obtained from a few cheesesamples showed traces of AFM1. But, no sample ex-ceeded the LOD. Many studies have shown that theconcentration of AFM1 in cheese is higher than that ofthe original milk because of the association of AFM1with casein19) AFM1 is more likely to be found incheese than in butter2) Our study conrmed this.

In Japan, studies on the occurrence of AFM1 in im-

ported and domestic cheese were conducted in 1979199220)24) After 1988, AFM1 was not detected incheese in Japan. However some studies in other coun-tries have found AFM1 at concentrations of more than0.25 g/kg in cheese12), 25) The number of importedmilk products is increasing and there is no regulationfor AFM1 in Japan. Therefore, we need to monitorAFM1 in dairy products.

Acknowledgements

This work was commissioned by the Ministry ofHealth, Labor and Welfare of Japan.

References

1) Masri, M. S., Booth, A. N., Hsieh, D. P. H. Comparativemetabolic conversion of aatoxin B1 to M1 and Q1 bymonkey, rat and chicken liver. Life Sci., 15, 203212(1974).

2) H. P. van Egmond, ed. Mycotoxins in dairy products.London, New York, Elisevier Applied Science, 1989, 272p. (ISBN 185166369X)

3) Safety evaluation of certain mycotoxins in food. Inter-national programme on chemical safety, World HealthOrganization-Geneva, WHO food additives series 47,FAO food and nutrition paper 74, 2001, p. 1102. (WHOISBN 92-4-166047-3)

4) Some naturally occurring substances: Food items andconstituents, heterocyclic aromatic amines and myco-toxins. IARC Monographs on the Evaluation of Carci-nogenic Risks to Humans, 56, 245395 (1993).

5) O$cial Journal of European Communities, L77,16.3.2001, 113.

6) H. P. van Egmond. Current situation on regulations formycotoxins. Overview of tolerances and status of stan-dard methods of sampling and analysis. Food Addit.Contam., 6, 139188 (1989).

7) Mortimer, D. N., Gilbert, J., Shepherd, M. J. Rapid andhighly sensitive analysis of aatoxin M1 in liquid andpowdered milks using an a$nity column cleanup. J.Chromatogr., 407, 393398 (1987).

8) Kamimura, H., Nishijima, M., Yasuda, K., Ushiyama, H.,Tabata, S., Matsumoto, S., Nishima, T. Simple, rapidcleanup method for analysis of aatoxins and compari-son with various methods. J. Assoc. O#. Anal. Chem., 68,458461 (1985).

9) Horwitz, W., ed. O$cial Methods of Analysis of AOACInternational. 18th ed. Gaithersburg, USA, AOAC Inter-national, 2005, 49.3.07. p. 4547. (ISBN 0-0935584757)

10) Some chemicals that cause tumours of the kidney orurinary bladder in rodents and some other substances.IARC Monographs on the Evaluation of CarcinogenicRisks to Humans, 73, 131182 (1999).

11) Dragacci, S., Gleizes, E., Fremy, J. M., Candlish, A. A. G.Use of immunoa$nity chromatography as a purication

Table 3. Single laboratory validation of the method using immunoa$nitycolumn for detection of aatoxin M1 in cheddar cheese and butter

Food Parameter

Recovery () Repeatability () In-house precision ()

Cheddar cheese 85 5 9Butter 91 9 13

Five-day single laboratory validation (0.5 g/kg spiked, n2 per day)


step for the determination of aatoxin M1 in cheeses.Food Addit. Contam., 12, 5965 (1995).

12) Elgerbi, A. M., Aidoo, K. E., Candlish, A. A. G., Tester, R.F. Occurrence of aatoxin M1 in randomly selectedNorth African milk and cheese samples. Food Addit.Contam., 21, 592597 (2004).

13) Re-evaluation of some organic chemicals, hydrazine andhydrogen peroxide. IARC Monographs on the Evalua-tion of Carcinogenic Risks to Humans, 71, 251316(1999).

14) Uchigashima, M., Saigusa, M., Yamashita, H., Miyake, S.,Fujita, K., Nakajima, M., Nishijima, M. Development ofa novel immunoa$nity column for aatoxin analysisusing an organic solvent-tolerant monoclonal antibody.J. Agric. Food Chem., 57, 87288734 (2009).

15) Itoh, A., Watanabe, Y., Nakajima, M. A method ofanalysis for determination of aatoxins in chocolate.Mycotoxins, 58, 713 (2008).

16) Nakajima, M., Tabata, S., Akiyama, H., Itoh, Y., Tanaka,T., Sunagawa, H., Tyonan, T., Yoshizawa, T., Kumagai, S.Occurrence of aatoxin M1 in domestic milk in Japanduring the winter season. Food Addit. Contam., 21, 472478 (2004).

17) Scott, P. M., Trucksess, M. W. Application of immu-noa$nity columns to mycotoxin analysis. J. Assoc. O#.Anal. Chem., Int. 80, 941949 (1997).

18) O$cial Journal of European Communities, L201, 17.7.98,93101.

19) Brackett, R. E., Marth, E. H. Association of aatoxin M1

with casein. Z. Lebensm. Unters. Forsch., 174, 439441(1982).

20) Saito, K., Nishijima, M., Kamimura, H., Ibe, A., Ochiai, S.,Naoi, Y. Analytical method and investigation of naturaloccurrence of aatoxins in dairy products. ShokuhinEiseigaku Zasshi (J. Food Hyg. Soc. Jpn.), 20, 2732(1979).

21) Saito, K., Nishijima, M., Yasuda, K., Kamimura, H., Ibe, A.,Nagayama, T., Ushiyama, H., Naoi, Y. Natural occur-rence of aatoxins in commercial cheese. ShokuhinEiseigaku Zasshi (J. Food Hyg. Soc. Jpn.), 21, 472475(1980).

22) Hisada, K., Yamamoto, K., Tsubouchi, H., Sakabe, Y.Natural occurrence of aatoxin M1 in imported anddomestic cheese. Shokuhin Eiseigaku Zasshi (J. FoodHyg. Soc. Jpn.), 25, 543548 (1984).

23) Tabata, S., Kamimura, H., Tamura, Y., Yasuda, K.,Ushiyama, H., Hashimoto, H., Nishijima, M., Nishima, T.Investigation of aatoxins contamination in foods andfoodstu#s. Shokuhin Eiseigaku Zasshi (J. Food Hyg. Soc.Jpn.), 28, 395401 (1987).

24) Taguchi, S., Fukushima, S., Sumimoto, T., Yoshida, S.,Nishimune, T. Aatoxins in foods collected in Osaka,Japan, from 1988 to 1992. J. Assoc. O#. Anal. Chem., Int.78, 325327 (1995).

25) Pietri, A., Bertuzzi, T., Bertuzzi, P., Piva, G. Aatoxin M1occurrence in samples of Grana Padana cheese. FoodAddit. Contam., 14, 341344 (1997).


52_220

Documents