553

Journal ofFood Protection, Vol. 45, No.6, Pages 553·556 1,4pril1982) Copyright ©, International Association of Milk, Food, and Environmental Sanitarians

Fate of Aflatoxin M1 in Brick and Limburger-like / ROBERT E. BRACKETT, RHONA S. APPLEBAUM, DANA W. WISEMAN and ELMER H. MARTH* Department ofFood Science and the Food Research Institute, University of -Madison. Madison, Wisconsin 53706

{Received for publication September 25, 1981) Downloaded from http://meridian.allenpress.com/jfp/article-pdf/45/6/553/1655068/0362-028x-45_6_553.pdf by guest on 24 September 2021

ABSTRACT are ripened by a surface growth of certain bacteria and

fungi. Such ripening might affect AFM 1 differently from Three batches of were prepared, using milk that observed with Cheddar cheese. which was naturally contaminated with aflatoxin M1 (AFM 1). were allowed to ripen with a smear for 2, 3 or 4 weeks, MATERIALS Ac'IID METHODS and then were either waxed or wrapped in foil to simulate production of mild brick, aged brick or Limburger-like cheese, Preparation of toxic milk respectively. These cheeses were analyzed for AFM 1 at intervals Two Holstein cows were given daily doses of 13 mg of AFB, (Sigma for about 26 weeks. There was an average 1.7-fold enrichment Chemical Co., St. Louis, MO). Aflatoxin B1 was administered by placing contaminated feed concentrate directly into the rumen through oftoxin in the curd over that in milk. Levels of AFM 1 in cheese started low, appeared to increase at about 4 weeks of age and a fistula (2). Milk was collected twice daily and held at 7°C until there then dropped to initial levels in cheese ripened with a smear for was enough to manufacture a lot of cheese.

3 or 4 weeks. At no time did amounts of AFM 1 drop below Preparation of cheese initial levels. Toxin concentrations appeared to increase most in One hundred liters of contaminated milk was pasteurized for 30 min the rind of cheeses ripened with a smear for 2 or 3 weeks. When at 62.8°C. Established methods (13) were then used to manufacture the cheese in pilot plant-size vats. Blocks of cheese were allowed to ripen such ripening was for 4 weeks, levels of AFM 1 in the rind decreased, whereas levels in the center of the brick remained with a surface smear at 17 .2°C and 90o/o relative humidity for 2, 3 or 4 constant or increased. weeks to simulate production of mild brick, aged brick and Limburger-like cheese, respectively. Cheeses ripened for 2 or 3 weeks were then rinsed and coated with wax. Cheeses ripened for 4 weeks were rinsed and wrapped in foil. All cheeses were then held at 10°C for the remainder of the test period. Moisture content, fat content and pH Aflatoxin M 1 (AFM 1) is a toxic metabolite found in the values of the experimental cheeses are given in Table 1. milk of lactating animals which consume aflatoxin B 1 (AFB ), a toxic metabolite produced by certain fungi 1 TABLE 1. Moisture, fat and salt contents and pH of brick during growth on feedstuffs (1). cheese containing AFM1• Aflatoxin M 1 has been found in commercial milk (6,12) and cheeses (7,12). In addition, investigations have Fata pH Salt Trial Moisture been done to see how the toxin behaves in experimental (%) (%) (initial)b (%) cheeses which were made from milk that was naturally lc 41.8 31 5.25 1.6 contaminated with AFM 1. Several reports (1,5,10) have noted an enrichment of AFM in the curd of cheese made 2d 44.3 29 5.32 1.7 1 3e from naturally contaminated milk. However, washing the 43.8 28.5 5.27 1.7 curd reduced its content of toxin (4). a Percent fat in dry matter: Triall, 53.5; Trial2, 52.1; Trial 3, Aging of naturally contaminated unripened, Camem­ 50.7. bert or for 3 months did not inactivate bMeasured immediately before brining. AFM (5). Also, Brackett and Marth (3) found that the cpH at 12 weeks: ripened 2 weeks, 5.47; 3 weeks, 5.52; 4 weeks, 1 5.73. AFM concentration seemed to increase at about 18 1 dpH at 12 weeks: ripened 2 weeks, 5.50; 3 weeks, 5.52; 4 weeks, weeks of age in Cheddar cheese. This increase was 5.75. attributed to an increase in recovery of the toxin, possibly epH at 12 weeks: ripened 2 weeks, 5.46; 3 weeks, 5.48; 4 weeks, brought about by the processes that occur during 5.73. ripening of the cheese.

The purpose of the work reported here was to Methods r~(analysis determine the fate of AFM 1 in brick cheese and Amounts of AFM 1 were determined by using the method of Limburger-like cheese. These cheeses, unlike Cheddar, Stubblefield (14). Thin-layer chromatography (TLC) was done using

JOURNAL OF FOOD PROTECTION, VOL. 45, APRIL 1982 554 BRACKETT, APPLEBAUM, WISEMAN AND MARTH

Redi/Plate TLC plates precoated with Silica Gel G (Fisher Scientific Figures 1 through 3 give the AFM 1 content of cheese Co., Pittsburgh, PA) and quantification was done using a fluorometer that was initially ripened for 2, 3 or 4 weeks and then (G. K. Turner Associates, Palo Alto, CA). All solvents used for analysis stored until the cheese was 26 weeks old. In all trials, and column chromatography were from Burdick and Jackson Laboratories (Muskegon. Ml), and were of spectrophotometric grade. cheese which was surface-ripened for 2 weeks (Fig. 1) had

Chloroform was of analytical grade. Aflatoxin M1 standards were maximum values of AFM 1 at about 4 weeks of age, and obtained from Eureca Laboratories (Sacramento, CA). All results are then the values decreased to near initial concentrations expressed as means of duplicate analyses. for the remainder of the test period.

RESULTS 20.--.--.--.---.--.--.--.---~-,--.--.---.-, 16 w 16 e TRIAL 1 Table 2 gives the AFM 1 content of cheese, curd, whey, Vl t; TRIAL2 t:J 14 0 TRIAL 3 wash water and milk from which the cheese was made. 5 12 The average increase in AFM1 concentration in unbrined curd as compared to milk was 1.7-fold. Thus only about 12 o/o of the total toxin in milk could be accounted for in Downloaded from http://meridian.allenpress.com/jfp/article-pdf/45/6/553/1655068/0362-028x-45_6_553.pdf by guest on 24 September 2021

the unbrined curd. Brining had little effect on the AFM 1 OOL ___1_2 --+-~6'-.. 8 10 12 14 16 18 20 22 24 2& concentration in the curd. The whey contained about the TIME !WEEKS) same concentration and the wash water contained Figure 1. Al:M content of brick cheese ripened with a smjace slightly less AFM 1 than the milk. 1 Surface-ripening for 2 or 3 weeks appeared to increase smear for 2 weeks.

the concentration of the AFM 1 in cheese but 4 weeks of Toxin in cheeses which were surface-ripened for 3 surface-ripening appeared to decrease the concentration weeks (Fig. 2) showed less uniform behavior. Cheese of

of the toxin. There was a difference in the AFM 1 Trial 1 had a steady increase in the amount of AFM 1 up concentration at different locations in the cheese at the to about the lOth week of storage, at which time the

end of the ripening periods. Table 3 gives the AFM 1 amount of toxin decreased slowly throughout the concentrations of the rind and the inner portion of cheese remainder of the test period. Toxin in cheese of Trial 2 blocks. There appeared to be a higher concentration of remained at near the initial concentration until about the AFM 1 in the rind of the cheeses which were allowed to 12th week of storage. At that time, the concentration ripen with a surface smear for 2 or 3 weeks, than in those increased until about the 15th week and then remained that were ripened with a surface smear for 4 weeks. near that level. In Trial 3, the toxin level in cheese

These differences in amount of AFM 1 in the inside of decreased slightly for the first 12 weeks, after which the cheese and the rind ranged from 16 to 224%. behavior of the toxin was much like in Trial2.

TABLE 2. Concentration wash water. cheese and which cheese was made.

Curd AFM 1 in cheesea Milke (before brine) AFM in (weeks of smear Trial Wheya.b Washa,b 1 Total Total curd after development) AFM,a kg AFM,a brine a AFM 1(J;g) AFM 1{Jlg) 2 3 4

1 3.7 370 6.55 4.1 26.9 4.6 2.8 6.9 9.6 11.2 3.7 2 4.5 450 7.54 6.0 45.2 4.1 2.7 5.5 5.7 9.2 3.7 3 4.5 450 7.03 12 84.4 3.9 2.3 7.8 11.2 10.3 8.2 a An concentrations are given in j.lglkg or J.LgiL. bsafety restrictions prohibited collection of all whey and wash water; as a result, total amounts present are unknown. CI()O L of milk was used to prepare cheeses.

TABLE 3. Amount in the rind and cheese.

2 weeksb 4 weeks Trial 3 weeks Centerc Rind Center Rind Center Rind

1 S •.::> ·a 13.6 11.0 11.5 4.0 3.4 2 4.5 6.8 7.2 11.2 7.8 3.7 3 14.1 16.3 9.4 11.1 11.4 5.1 a All concentrations expressed in j.lglkg. bweeks of smear development. cRind samples were the outer 1.5 em of block. whereas center samples excluded the outer 1.5 em.

JOURNAL OF FOOD PROTECTION. VOL. 45, APRIL 1982 AFLATOXIN IN BRICK CHEESE 555

~.-.--.--.--.-.--.--.--.-.--.--.-~-. 24 normal yield for this type of cheese is lOo/o, and our yield 22 was less than 8%, some curd obviously was lost in the whey. This could, in part, account for the greater than expected proportion of toxin in the whey and the correspondingly smaller proportion in the cheese. In

addition, one cannot assume the concentrations of AFM 1 to remain the same in whey as drainage of curd e TRIAL 1 A TRIAL2 proceeded, since curd particles tend to form a sediment. 0 TRIAL3 This could lead to higher concentrations of AFM 1 at the

6 8101214161820222426 beginning of whey drainage, when samples were taken, TIME (WEEKS l than later in the process. However, this possibility was not investigated. Figure 2. AFM content of brick cheese ripened with a surface 1 Apparently, the increase in AFM 1 in cheese for up to 3 smear for 3 weeks. weeks resulted from an increase in recovery, since the Downloaded from http://meridian.allenpress.com/jfp/article-pdf/45/6/553/1655068/0362-028x-45_6_553.pdf by guest on 24 September 2021 Cheese which was surface-ripened for 4 weeks (Fig. 3) toxin was not produced in the cheese. The primary activity occurring during the first several weeks of exhibited a slight initial decrease in AFM 1 concentration followed by an increase until about the 16th week. At ripening of brick cheese is reported to be the proteolytic that time the toxin level dropped for the remainder of the action resulting from growth of Brevibacterium linens If test period. However, in no instance did this final (9). AFM 1 is bound to or occluded with casein, as has decrease result in values less than those found initially. been suggested (1,8,10,11), perhaps this proteolytic action could serve to make the toxin more easily recoverable. This possibility was also noted and discussed in an earlier report by Brackett and Marth (3). The

decrease in amount of AFM 1 in cheese, which received 4 weeks of surface-ripening was not surprising, since a distinct odor of ammonia was noted. Presumably the ammonia resulted from proteolysis. Ammonia can inactivate aflatoxins (10). e TRIAL 1 Results given in Table 3 also support the idea that A TRIAL2 o TRIAL3 cheese ripening influences aflatoxin recovery. The rind, 18 20 22 24 26 which is the region of the cheese most affected during TIME !WEEKS I surface-ripening (9,15) also exhibited an increase in its concentration of AFM before this happened in the Figure 3. AFM content of Limburger-like cheese ripened 1 1 center of the cheese. In addition, some of the increase with a surface smear for 4 weeks. could result from a decreased moisture content of the DISCUSSION rind, thus raising the amount of toxin per unit of weight. There also was an increase in toxin concentration in the center of the block, but the increase appeared to occur The average 1.7-fold enrichment of AFM 1 in the brick cheese is lower than that found by others. Kiermeier and more slowly than on the surface. These differences could result from inward diffusion of enzymes involved in Buchner (5) found increases of 3.2-, 3.3- and 3.7-fold in uncured, Camembert, and Tilsit cheeses, respectively. surface-ripening or from faster ripening of the cheese at Brackett and Marth (3) found a 4.3-fold enrichment in the surface without diffusion of enzymes. There is Cheddar cheese. This decrease in enrichment of toxin in evidence for both mechanisms of ripening (9,15). the curd of brick cheese was not completely unexpected As stated earlier, the odor of ammonia was detected in since the curds are washed, and this has been samples of cheese that were ripened with a surface smear demonstrated (4) to decrease the toxin in the curds. This for 4 weeks. However, we did not notice this odor in samples taken from the center of the cheeses. This might observation was supported by the fact that AFM 1 was found in the wash water from all trials in our explain why there was no consistent decrease in amount experiments. of AFM1 in the center of the cheeses and variability in amount of ammonia produced may explain why the Concentrations of AFM 1 found in the whey also differ from results of Kiermeier and Buchner (5) and Brackett amount in the rind decreased less in some cheeses than in and Marth (3). In the work of these authors, the amount others. Ripening of the cheese occurs more rapidly in the rind, which could also explain why the concentration of of AFM 1 in whey was no more than about one-half of that found in the milk from which the whey came. It is toxin in the center ofthe cheese did not change markedly even when the cheese had been allowed to ripen with a possible that the greater concentrations of AFM 1 we found in whey could be related to the low yield of curd we surface smear for 3 or 4 rather than 2 weeks. obtained. Thus the small curd particles in the whey could In all instances, AFM 1 concentrations were greatest

contain an appreciable amount of AFM 1• Since the after ripening had taken place. However, it is unusual

JOURNAL OF FOOD PROTECTION, VOL. 45, APRIL 1982 556 BRACKETT, APPLEBAUM, WISEMAN AND MARTH that the concentration of the toxin quickly returned to of Dairy Science for their assistance in obtaining milk that was initial values in samples surface-ripened for 2 weeks, naturally contaminated with AFM 1. whereas the amount of AFM 1 in other samples remained at high levels for longer times. This implies that any REFERE~CES effects which caused the increase of AFM were only 1 1. Allcroft, R., and R. B. A. Carnaghan. 1963. Groundnut toxicity: temporary for cheese ripened for 2 weeks but were longer An examination for toxin in human food products from animals lasting for cheeses that were ripened longer. If fed toxic ground nutmeal. Vet. Rec. 75:259-263. surface-ripening was the only cause for increased 2. Applebaum, R. S., R. E. Brackett, D. W. Wiseman, and E. H. Marth. 1982. Responses of dairy cows to dietary aflatoxin: Feed recovery of AFM 1, one would expect the cheeses ripened for 4 weeks to have largest differences in AFM intake and yield, toxin content 14nd quality of milk of cows treated 1 with pure and impure aflatoxin. J. Dairy Sci. (In press). concentration as compared to initial values. This is true 3. Brackett, R. E., and E. H. Marth. 1982. Fate of aflatoxin M1 in because this cheese was wrapped in foil and moderate Cheddar cheese and process cheese ·spread. J. Food Prot. 45: ripening continued (15) during storage. The cheeses 4. Kiermeier, F .. and M. Buchner. 1977. Zur Verteilung von ripened for less time were coated with wax, which Aflatoxin M1 auf Molke und Bruch bei der Kaseherstellung. Z. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/45/6/553/1655068/0362-028x-45_6_553.pdf by guest on 24 September 2021 essentially stops surface-ripening (15). However, cheeses Lebensm. Unters. Forsch. 164:82-86. 5. Kiermeier, F., and M. Buchner. 1977. Verhalten von Aflatoxin M1 ripened 3 or 4 weeks would also allow microorganisms, wahrend der Reifung und Lagerung von Kase. Z. Lebensm. other than those involved in surface-ripening, to grow Unters. Frosch. 164:87-91. more. It is possible that the action of these microorgan­ 6. Kiermeier, F., G. Weiss, G. Behringer, M. Miller, and K. Ranfft. isms could be important. 1977. Vorkommen und Gehalt an Aflatoxin M1 in molkerei­ It appears that binding of AFM occurs (1,3,8,10,11) anlieterungs Milch. Z. Lebensm. Unters. Forsch. 163:171-174. 1 7. Kiermeier, F., G. Weiss, G. Behringer, and M. Miller. 1977. and might affect the amount of toxin recovered during Ueber das Vorkommen und den Gehalt von Aflatoxin M1 in analysis (11). As stated previously (J), it is difficult to Kasen des Handels. Z. Lcbensm. Unters. Forsch. 163:268-271. explain or predict the behavior of AFM1 in cheeses 8. Kraybill, H. F., and R. E. Shapero. 1969. Implications of fungal during ripening without first knowing the relationship of toxicity to human health. pp. 401-441. In L. A. Goldblatt (ed.) AFM to the cheese proteins. Aflatoxins: Scientific background, control and implications. 1 Academic Press, New York. It seems certain from this work that while less AFM 1 9. Langhus, W. L., W. V. Price, H. H. Sommer, and W. C. Frazier. remained in our brick cheese than in Cheddar, 1945. The "smear" of brick cheese and its relation to flavor Camembert, Tilsit or uncured cheese, the toxin was development. J. Dairy Sci. 28:827-838. stable for at least 26 weeks. Since brick or Limburger 10. McKinney, J.D., G. C. Cavanagh. J. T. Bell, A. S. Hoversland, D. cheeses are usually consumed sooner, if the cheese is M. Nelson, J. Pearson, and R. J. Selkirk. 1973. Effects of ammoniation on aflatoxin in rations fed lactating cows. J. Amer. contaminated with AFM 1, it can be assumed to be no Oil Chem. Soc. 50:70-84. safer than the milk from which the cheese was made. 11. McKinney, J. D., and G. C. Cavanagh. 1977. Extraction of "bound" aflatoxins. Zeszyty Problemowe Postepow Nauk Rolniczych. 189:247-253. ACKNOWLEDGMENTS 12. Polzhofer, K. 1977. Aflatoxinbestimmung in Milch and Milchprodukten. Z. Lcbensm. Unters. Forsch. 163:175-177. 13. Price, W. V., and H. J. Buyens. 1967. Brick cheese: pH. moisture, This research was supported by the College of Agricultural and Life and quality control. J. Dairy Sci. 50:12-19.

Sciences. University of Wisconsin-Madison; Dairy Research, Inc., 14. Stubblefield. R. D. 1979. The rapid determination of aflatoxin M1 Rosemont, IL; the National Cheese Institute, Chicago, IL; Kraft. Inc .. in dairy products. J. Amer. Oil Chem. Soc. 56:800-802. Glenview, IL: and the Food Research Institute of the University of 15. Tuckey, S. L., and M. R. Sahasrabudhe. 1957. Studies in the Wisconsin. We thank Dr. A. Hardie and Dr. L. D. Satter, Department ripening of Limburger cheese. I. J. Dairy Sci. 40:1329-1337.

JOURNAL OF FOOD PROTECTION. VOL. 45, APRIL 1982