Aflatoxins (AF) are fungal secondary toxic metabolites that have mutagenic and carcinogenic effects in humans. Aflatoxin B1 (AFB1), the most toxic Aflatoxin, contaminates cattle feed and can be metabolized and excreted as the hydroxylate Aflatoxin M1 (AFM1). Aflatoxin B2 (AFB2) is excreted as Aflatoxin M2 (AFM2) in milk, and dairy products such as cheese can concentrate these carcinogens. Artisanal Oaxaca-type cheeses were sampled in Veracruz City, Mexico in 2005 and 2006, and three different extraction methods which were representative of many other reported methods- were selected for testing and validation. The R-Biopharm method was chosen and used to analyze the 30 samples that were derivatized and quantified by HPLC. The validation methods gave limits of detection (LODs) of 0.01 ng g-1 for AFM1 and 0.05 ng g-1 for AFM2; the limits of quantification (LOQs) for each Aflatoxin were four times the respective LOD. The recovery percentages were 95% for AFM1 and 93% for AFM2. The retention times were in the range of 8.514 to11.849 min for AFM1 and 20.208 to 22.447 min for AFM2. The extraction method, derivatization, and quantification (which were achieved using an HPLC-fluorescence detector) showed that 16 of the 30 samples (53%) were contaminated with AFM1, at concentrations ranging from 0.01 to 44 μg kg-1, whereas AFM2 contamination was less frequent. AFM2 contamination was found in only 3 of the 30 samples and was below the LOD, and concentrations ranged from 0.67 to 3.43 μg kg-1. These two ranges of AF contamination surpassed the tolerance limits stated by NAFTA (0.5 μg kg-1) and by Codex Alimentarius and the European Union (0.05 μg kg-1).
Keywords: Aflatoxins M1 and M2; Artisanal Oaxaca cheese; Carcinogens; Cheese contamination
AFs: Aflatoxins; AFL: Aflatoxicol; AFB1: Aflatoxin B1; AFB2: Aflatoxin B2; AFG1: Aflatoxin G1; AFG2: Aflatoxin G2; AFM1: Aflatoxin M1; AFM2: Aflatoxin M2; AFP1: Aflatoxin P1; AOAC: Association of Official Analytical Chemists; ºC: Degrees Celsius; IC(β1): Confidence Interval for the Slope to Origin; R2: Determination Coefficient; g: Gram(s); HPLC: High Performance Liquid Chromatography; h: Hours; pH: Hydrogen-Ion Concentration; IAC: Immunoaffinity Column; LD50: Lethal Dose; LOD: Limit of Detection; LOQ: Limit of Quantification; L: Liter; MeOH: Methanol; μg kg-1: Micrograms per Kilogram; μm: Micrometers; Ml: Mililiter; mg kg-1: Milligrams per Kilogram; mm: Millimeters; min: Minutes; MW: Moecular Weight; ng: Nanograms; ng g-1: Nanograms per Gram; nm: Nanometers; NAFTA: North American Free Trade Agreement; bo: Ordinate to Origin; %: Percentage; PBS: Phosphate-Buffered Saline; rpm: Revolutions per Minute; b1: Slope Value; NaCl: Sodium Chloride; SD: Standard Deviation; Sy/x: Standard Deviation of the Regression; % CV : Variation Coefficient Percentage; v/v: Volume per Volume
In 2014, Veracruz State was the sixth leading state in Mexico for milk production, processing 693,950 L of raw cow milk. The dairy industry is the third most important food industry in Veracruz State and 53% of the milk produced here is sold as cheese. It is estimated that 11,130 tons of milk are produced in Veracruz State each year . Handmade production of dairy products processes 51% of the milk for cheese (827,120 L), whereas industrial production processes 21% of the milk .
No quality control is in place in the artisanal non-industrialized milk industry, rendering handmade cheeses susceptible to Aflatoxin (AF) contamination a health risk for consumers; however, no published studies have measured AF levels in cheese.
AFs are relatively small molecules (MW<700) and are fungal secondary metabolites [3,4]. Structurally, AFs are bis-dihydro furanocoumarins and they are produced by Aspergillus fungi, such as Aspergillus flavus, A. parasiticus and A. nomius [5-7]. These fungi produce spores that contain AFB1, AFB2, AFG1 and AFG2, which can contaminate field crops or storage warehouses; these toxins are a potential health risk for animals and humans [5,6]. AFB1 can contaminate cereals such as maize, sorghum, rice, barley and wheat, i.e., the grains that are used to make nutritionally balanced feed for dairy cows. AFB1 is a Group I (proven) carcinogen for humans [7-11]. When a cow eats AFB1-contaminated feed, it excretes the toxin as AFM1 hydroxylate and AFM2 hydroxylate in its milk. The liver metabolizes AFB1 by adding a hydroxyl group, which increases the water solubility of AFB1 and facilitates its excretion in urine or milk. This reaction also lowers the toxicity of AFB1 by converting it into AFM1. AFB2 in the feed is metabolized to AFM2 excreted in the milk [3,12,13]. AFM1 is metabolized by the cytochrome P450 enzymes that are located in the endoplasmic reticulum [14-16].
The warm tropical weather of Veracruz favors fungal growth and AF production [17,18]. Cheese made from AFM1 contaminated milk has higher AFM1 concentrations than are found in the original milk [19,20]. Milk and cheese are the vehicles for the introduction of AFM hydroxylates into the human diet . There are many cases of AFM1 contamination of cheese in Brazil, Turkey, Mexico, Kuwait, Lebanon, Italy, and Serbia [7,12,16,20,22-25]. The stability of AFM1 is not affected during the elaboration of cheese. AFM1 seems to bind to casein, and so cheese curd has higher levels of AFM1 than milk and is a risk factor when added during cheese elaboration [5,26]. AFM1 is 2 to 10% less carcinogenic than AFB1, but it is classified as a Group 2B (possible) human carcinogen because it is hepatotoxic, mutagenic and carcinogenic . Owing to its toxicity, many countries have established maximum tolerance levels. The European Regulatory Commission 1881/2006 established a maximum limit of 0.05 μg kg-1 for AFM1 in milk, whereas the USA has set a limit of 0.5 μg kg-1 [27,28]. The maximum tolerance limit of AFM1 in cheese in some European countries, including Switzerland, France, Austria and Turkey, is 0.25 μg kg-1 . Oaxaca-type cheese is produced by acidifying milk, adding curd and making a dough that is submerged in hot water (65º to 70ºC) until strands are formed [30,31]. Oaxaca is one of the more popular cheeses consumed in the City of Veracruz [1,32]. In Mexico, the presence of AFM1 in milk has been reported, but the presence of AFM1 in artisanal cheeses consumed in Veracruz has not been studied .
The purpose of this research was to quantify the levels of AFM1 and AFM2 in artisanal Oaxaca-type cheeses sampled in Veracruz.
The study consisted of 30 samples of Oaxaca-type cheese, (each weighing 750 g), that were purchased in groceries and markets of the Port of Veracruz, Mexico. A Matlab algorithm was performed to randomly select the places from which samples were purchased. Samples of Oaxaca-type cheeses were purchased in Veracruz City in March 2016, which is in the dry season, when the cows are fed with nutritionally balanced feed; the rest of the year, cows typically eat grass (Figures 1 and 2).
Methods tested for the extraction of AFM1 from cheese
A bibliographic data base was consulted, and three different extraction methods for cheese were selected; these methods were representative of many other reported methods as they used the same solvents in different quantities. The three methods are summarized below.
1) The AOAC (2006) method is representative of other methods all of which were performed using chloroform, diatomaceous earth and salt. In this study, the AOAC method was performed according to the following protocol [14,34-36].
Samples of dry, ground Oaxaca cheese (weighing 15 g each) were blended (Waring ETL laboratory blender 7010S Model WF 2211214 Torrington, Connecticut USA) with 5 g of diatomaceous earth, 100 mL of chloroform and 2 g of NaCl for 1 min. This mixture was centrifuged (ALC 4235 Cool Working System, Milano, Italy) at 4500 rpm for 15 min and 6.7 mL of the supernatant (equivalent 1 g of sample) was diluted in 24 mL of pH 7.4 phosphate-buffered saline (PBS). The diluted supernatant was applied to an immunoaffinity column (IAC), which was previously balanced with 20 mL of PBS. The sample was then washed with 20 mL of distilled water and air was passed through the IAC to dry it. AFM1 and AFM2 were eluted from the agarose gel using HPLC-grade MeOH. The eluates were dried at 40ºC in an oven (Novatech Model BTC 9100, Houston Texas, USA) and were derivatized.
2) The R-Biopharm (2012) method is almost the same as that used in a study by Iha et al. and uses methanol, water and salt [37,38]. This method has been recommended for use with Total aflatoxin Easi-Extract IACs (R-Biopharm Rhône Ltd., Glasgow, Scotland, UK). In this study, the R-Biopharm method was performed according to the following protocol.
Samples of dry, ground Oaxaca cheese (weighing 15 g each) were blended with a mixture of 100 mL of MeOH/water (80:20 v/v) and 2 g NaCl (to clarify the extract) for 2 min at high speed. The mixture was centrifuged at 4500 rpm for 15 min and 6.7 mL of the supernatant (equivalent to 1 g of sample) was dissolved in 24 mL PBS. Before samples were added, each IAC that was used to detect total aflatoxin levels was balanced with 20 mL of PBS at pH 7.4, which was applied at a flux of 5 mL/min. The buffered sample was passed through the IAC and AFM1 and AFM2 were eluted using 1.5 mL of HPLC-grade MeOH, followed by 1.5 mL of distilled water with reflux. The eluate was dried at 40ºC in an oven and then derivatized.
3) The Cavaliere et al. method uses acetone and silica gel with activated charcoal columns, methanol and water. In this study, the Cavaliere et al. method was performed according to the following protocol .
Samples of dry, ground Oaxaca cheese (weighing 15 g each) were blended with acetone (150 mL), as recommended for fresh cheese. From this mixture, 6.7 mL of the supernatant (equivalent to 1 g of sample) was analyzed. The sample extraction included a preliminary matrix solid-phase dispersion-extraction step before the solid-phase extraction clean-up step, which was achieved using a Carbograph-4 cartridge. The cheese sample was applied to the cartridge, which was washed with 5 mL of methanol containing 2% acetic acid. AFM1 was eluted from the cartridge using 10 mL dichloromethane/methanol/ acetic acid (88:10:2 v/v/v). The eluate was dried in an oven, derivatized, and 60 μL was used for HPLC.
We used established methods to derivatize AFs and thereby increase their fluorescence [40,41]. The dry AFM1 and AFM2 standards (Sigma- Aldrich, St. Louis MO, USA) were dissolved in 200 μL acetonitrile (ACN), and 800 μL derivatizing solution was added. The derivatizing solution was composed of 5 mL trifluoroacetic acid (Sigma-Aldrich, St. Louis MO, USA), 2.5 mL glacial acetic acid (Merck, Naucalpan, Estado de Mexico, Mexico) and 17.5 mL deionized distilled water, and was vortexed (Vortex G-560, Bohemia NY, USA) for 30 s. The vials containing the dry eluates were heated in a vapor bath at 65°C for 10 min. The samples were cooled to room temperature, and triplicates of 60 μL were analyzed by HPLC.
Validation of the extraction method
The validation of the analytical method was performed using known methods . To decide which method was the best for further analysis of the 30 Oaxaca cheese samples, the following parameters were taken into account.
Linearity of the system (Calibration curves)
Solutions of different AFM1 and AFM2 concentrations were prepared from a stock concentration of 1000 ng AFM. For the 0.25 mg AFM standards, dry dusts produced by Aspergillus flavus were diluted with benzene:acetonitrile (98:2 v/v), following known methodology, so that the pure AFMs did not decompose .
a. The spectrophotometer that we used (Genesys 10 UV Thermo Electron Corporation. Madison Wisconsin, USA) was calibrated before the experiments to measure the absorbance of the AFM standard solutions at 357 nm.
b. The following formula was applied to calculate AFM concentrations:
Concentration of AFM (μg mL-1)=Absorbance × molecular weight × 1000 × Correction factor of the equipment/Extinction coefficient.
c. We calculated the inverse of the AFM concentration (1/x), which represents the amount of AFM that has to be added to 1 mL of MeOH.
d. The amount of AFM was subtracted from 1000 ng to give the volume of MeOH needed for a 1000 ng stock solution.
e. Twelve concentrations (0.01, 0.05, 0.1, 0.5, 1, 2, 4, 8, 16, 32, 64 and 128 ng) of AFM were created from the 1000 ng stock solution. These samples were then used to plot the analytic signal (area below the curve of each chromatographic peak) against the AFM1 and AFM2 concentrations; the curve equation and statistical parameters were obtained. The slope value (b1), ordinate to origin (bo), determination coefficient (R2), confidence interval for the slope to origin (IC(β1)), variation coefficient percentage (% CV), standard deviation (SD) and the LOD and LOQ were calculated using Excel 2003.
The LOD of the equipment was established in relation to the noise in the chromatogram. The LOD equals the AFM1 concentration that gives a signal that is three times greater than the noise. The LOQ equals the AFM1 concentration that is 10 times greater than the noise .
To calculate the LOD, the following equation was used:
The LOQ was calculated using the following equation:
where Sy/x is the standard deviation of the regression, and b1 is the value of the slope.
The recovery percentage is a measure of the accuracy of the method and expresses the proximity between theoretical and experimental values. The recovery percentage of each spiked sample was calculated by dividing the recovered AFM quantities by the spiked amount. The arithmetic average, standard deviation, percentage of variation coefficient and confidence interval were calculated for the population media. To obtain accurate measurements, 15 g samples of dried, ground Oaxaca cheese and aliquots containing 1 g of each sample were spiked with three different concentrations (5, 20 and 40 μg kg-1) of AFM1; one aliquot without spiked AFM1 was used as the control. The samples were individually processed with each one of the three extraction methods as explained above [34,37,39]. The control sample gave the basal level of contamination. The AFM1 of 1 g of sample - measured with and without spiking to account for the basal AFM1 contamination - diluted in PBS (1:4 v/v) was purified and concentrated using an IAC and quantified by HPLC. The percentage of recovery (i.e., the accuracy) of each method was obtained.
Extraction, purification and concentration of AFM1
After the validation of the three extraction methods, AFM1 extraction was performed using the R-Biopharm Rhône methodology and the obtained AFM1 was then derivatized . When the derivatizing mixture cooled to room temperature, triplicates of 60 μL of each sample were analyzed by HPLC.
The chromatographic system was an Agilent Series 1200 HPLC (Agilent Technologies, Inc., USA) and consisted of an isocratic pump (Model G1310A), a fluorescence detector (Model G1310A Series DE62957044, Agilent Technologies, Inc., USA; which was set to an excitation wavelength of 360 nm and an emission maximum of 450 nm, and an autosampler (G1329A Series DE64761666). The chromatography column was a VDS Optilab VDSpher 100 C18-E 5 μm 250 × 4.6 mm maintained at room temperature (22°C) with a mobile phase of water:ACN:methanol (65:15:20 v/v/v) that was degasified for 30 min by vacuum filtration and added at a flux of 1.0 mL/min.
The statistical analysis was performed using Minitab version 16. Variance analyses with the Tukey test at 95% were performed in triplicate, considering each cheese as an experimental unit. The graphs showing the data from the Tukey test and the standard deviations were produced using Kaleidagraph version 3.5.
AFB1 is a genotoxic carcinogen that is the most carcinogenic of all aflatoxins, and animal experiments have shown that AFB1 causes liver cancer in most species. AFM1 is a metabolite of AFB1 that is found in the milk of animals that have ingested AFB1. The liver is the target of the carcinogenic effects of AFM1. The genotoxicity of AFM1 has been demonstrated by in vitro and in vivo experiments. The carcinogenic potency of AFM1 is 2 to 10% weaker than that of AFB1 . Hence, the Food Safety Commission of Japan has concluded that the AFB1 present in animal feed is extremely unlikely to affect the health of humans that have consumed contaminated milk or other livestock products. However, AFM1 and its metabolites are also genotoxic carcinogens, and are more likely to be found in livestock products, so AFB1 contamination in feed and AFM1 contamination in milk need to be reduced as much as possible. In particular, attention should be paid to the fact that the intake of milk per 1 kg of body weight is higher in infants than in other age groups .
The LOD for AFM1 was 0.01 ng g-1, the range for the retention times was 8.514 to 8.769 min, the R2 was 0.9834, and the recovery percentage was 95%. For AFM2, the LOD was 0.05 ng g-1, the range for the retention times was 20.208 to 22.447 min, the R2 was 0.9946, and the recovery percentage was 93% (Table 1).
|Method||Recovery percentages of the extraction methods for Oaxaca cheese|
|Basal AFM1||AFM1 (ng g-1) spiked||Total AFM1||Recovered AFM1||Recovered percentage|
|AOAC (5 ng g-1)||5||11.82||4.86||41.1|
|AOAC (20 ng g-1)||20||26.82||26.37||98.3|
|AOAC (40 ng g-1)||40||46.82||42.70||91.2|
|Average of controls AOAC||6.82||77%|
|Biopharm (5 ng g-1)||5||9.71||8.98||92.5|
|Biopharm (20 ng g-1)||20||24.71||24.51||99.2|
|Biopharm (40 ng g-1)||40||44.71||42.26||94.5|
|Average of controls Biopharm||4.71||95%|
|Cavaliere et al. (5 ng g-1)||5||9.04||6.0||66.4|
|Cavaliere et al. (20 ng g-1)||20||24.04||16.86||70.1|
|Cavaliere et al. (40 ng g-1)||40||44.04||41.38||94.0|
|Average of controls Cavaliere||4.04||85%|
AFM1: Aflatoxin M1; AFM2: AFlatoxin M2; AFMt: Total M aflatoxins; SD: Standard Deviation
Table 1: Recovery percentages of the three extraction methods for Oaxaca cheese .
The recovery percentages of each extraction method are presented in Table 1; examination of these criteria led to the decision to use the R-Biopharm extraction method for the rest of the study. This method was safer than the AOAC method, which, dissolved the plastic of the blender cover and used more dangerous solvents.
Quantification of AFM in the Oaxaca-type cheese samples
The levels and SDs of AFM1 and AFM2 in the media are presented in Table 2 and Figure 3. AFM 1 contamination was found in more samples (16 out of 30, 53%) than was AFM2 contamination (3 out of 30, 10%); three samples had AFM2 concentrations that were below the LOD. The AFM1 levels ranged from 0.1 to 44 μg kg-1 and AFM2 concentrations ranged from 0.67 to 3.43 μg kg-1.These two ranges surpass the tolerable limits stated by NAFTA (of 0.5 μg kg-1) and by Codex Alimentarius and the European Union (0.05 μg kg-1) . A previous study has also reported high levels of AFM1 in 27% of samples of fresh panela cheese from Mexico, but we found high AFM1 levels in 53% of samples of Oaxaca-type cheese . To date, no human cases of food poisoning have been associated with the consumption of contaminated cheese, even though high levels of carcinogens can be found in cheese.
|Sampling point||Origin||M Aflatoxins (µg kg-1)|
|5||Soledad de Doblado||0.76||0.72||0.02||0||0||0||0.76|
|13||Soledad de Doblado||2.66||2.53||0.18||0||0||2.66|
|20||Boca del Río,Veracruz||0||0||0||0||0||0||0|
|22||Soledad de Doblado||0||0||0||2.76||0.013||0.010||2.76|
|24||Soledad de Doblado||0.01||0||0||0||0||0||0.01|
|29||(La Joya) Acajete, Ver||0||0||0||0||0||0||0|
AFM1: Aflatoxin M1; AFM2: Aflatoxin M2; AFMt: Total M aflatoxins; SD: Standard Deviation
Table 2: Concentrations of AFM1 and AFM2 (μg kg-1) in artisanal Oaxaca cheese from the City of Veracruz.
Filamentous fungi are important for surface mold-ripened and core mold-ripened cheeses. The degradation of milk constituents, such as proteins and lipids, can lead to improvements in the texture, flavor, and nutritional quality of cheeses . In Aspergillus spp., aflatoxins act as chemical signals for communication, competitive weapons to defend the fungal habitat, and inhibitors of the growth and reproduction of competitors within the same trophic niche [47,48]. In the cheese ecological niche, spoilage fungi and fungal cultures are in competition. The presence of competing microorganisms, including bacteria and yeast, is essential for mycotoxin production, as demonstrated in a study showing that serial transfers on culture media in laboratory conditions (i.e., without competitors and natural stress conditions) result in the loss of AF production in strains of A. flavus [49-51]. Nutrient depletion and competition with other fungi stimulate AF production, and early production of AF allows molds to rapidly colonize the environment. Lethal dose (LD50) values range from 0.5 to 10 mg kg-1 depending on the type of AF. Different animal species have different susceptibilities to chronic and acute AF toxicity, which can ultimately lead to liver cancer .
The incidence of AFM1 in cheese appears vary . Many countries monitor AFM1 levels in milk and dairy products. High AF levels of up to 850 ng kg-1 have been found in cheeses from certain regions of Turkey . In the Bursa Province, Turkey, the AFM1 levels in local cheese samples exceeded 250 ng kg-1 . In the South of Spain, AFM1 was detected at concentrations between 20 and 200 μg g-1 in 16 out of 35 samples (45%) of local cheese . In a study of cheeses in Campinas, Brazil, AFM1 concentrations were low, and only 4 out of 204 samples were contaminated . In Argentina, 2 out of 50 samples were positive for AFM1, with levels of 0.33 and 0.20 μg kg-1 . The consumption of AFB1-contaminated or AFB2-contaminated feed by dairy cows is known to result in the excretion of the monohydroxylated derivatives AFM1 and AFM2 in the cows’ milk within a few hours . If cows ingest 300 ng g-1 AFB1, they will produce milk containing 1-3 ng g-1 AFM1 24 h later . Other studies have reported that the normal carry-over is approximately 0.4-0.6%, so an intake of ≥ 70 μg AFB1 by cows results in the production of milk containing AFM1 levels greater than the regulatory limit (0.05 μg kg-1) [61,62]. The amount of AFM1 formed depends on the individual cow, and its excretion in milk decreases one day after the consumption of AFB1-contaminated feed ceases, although traces can be found over two or three more days. The conversion ratio of AFB1 to AFM1 varies from 1:100 to 1:300 .
The health risk of AFM1 should not be underestimated. Ammonia treatment is the only method that can reduce AF levels; this method can destroy 95% to 98% of AFs and is used to decontaminate animal feed in various countries [64,65].
Cold temperatures (5 to 7°C) prevent fungal growth and mycotoxin production . Only Penicillium spp. can grow at such temperatures, but they do not produce AFs or sterigmatocystin contamination . Other chemicals such as pimaricin (also known as natamycin) - which is produced by the actinobacterium Streptomyces natalensis - delay fungal growth and are used as preservatives to control mycotoxin production [26,67,68].
In conclusion, although the legislation regarding maximum tolerance levels has attempted to decrease the level of AFM1 contamination in cheeses, and although there is no direct evidence of human toxicity resulting from the consumption of cheese contaminated with AFM1, the problem is still present in fresh cheeses such as the artisanal Oaxaca cheese, as we described in this manuscript.
The authors thank the Instituto Tecnológico de Veracruz for the cheese sampling and the Instituto de Biología, Universidad Nacional Autónoma de México (IBUNAM) for the data analysis. The authors also thank IBUNAM´s personnel: Noemí Chávez from the Secretaría Técnica, and Joel Villavicencio, Jorge López, Alfredo Wong, Celina Bernal, Diana Martínez and Julio César Montero provided valuable assistance with imaging, computer analysis and design. Additionally, we thank Georgina Ortega Leite and Gerardo Arévalo for library information.