Reduction of aflatoxin B₁ during tortilla production and identification of degradation by-products by direct-injection electrospray mass spectrometry (DIESI-MS)

Reducción de la aflatoxina B₁ durante la producción de tortilla y la identificación de los productos de degradación por espectrometría de masas con ionización por electrospray de inyección-directa (DIESI-MS)

Abigail Moreno-Pedraza, QFB,⁽¹⁾ Laura Valdés-Santiago, D en C,⁽¹⁾ Laura Josefina Hernández-Valadez, QFB,⁽¹⁾ Alicia Rodríguez-Sixtos Higuera, MSc,⁽¹⁾ Robert Winkler, D en C,⁽¹⁾ Dora Linda Guzmán-de Peña, D en C.⁽¹⁾

(1) Departamento de Biotecnología y Bioquímica, Centro de Investigación y de Estudios Avanzados-IPN, unidad Irapuato. Guanajuato, México.

Corresponding author

Abstract

Objective. To determine the effect of pH, and exposure time over the inactivation of aflatoxin B₁ (AFB₁) during the tortilla making process as well as the degradative molecules generated.

Materials and methods. Inactivation of AFB₁ in maize-dough with alkaline pH and in alkaline methanolic solutions was determined by HPLC. Kinetics of time exposure of AFB₁ in methanolic solution and the degradative products were analyzed by direct injection electrospray mass spectometry (DIESI-MS).

Results. The alkaline pH of the maize-dough after nixtamalización between 10.2, and 30-40 minutes of resting at room temperature allows the 100% reduction of AFB₁. DIESI-MS analysis of the extracts indicated the presence of two degradation molecules from AFB₁.

Conclusion. The alkaline pH of maize-dough and resting time are the principal factors involved in diminishing AFB1 levels in tortillas. A procedure to the tortilla making process is proposed, which allows the reduction of remnant AFB₁, avoiding the accumulative effect over consumers.

Key words: aflatoxin B₁; degradation products; alkalinity; Aspergillus; Mexico.

Resumen

Material y métodos. La inactivación de la AFB₁ a pH alcalino y diferentes tiempos de exposición en masa nixtamalizada y en soluciones metanólicas fueron determinadas por HPLC. La cinética de degradación de AFB₁, y los productos de degradación en soluciones metanólicas se determinaron por DIESI-MS.

Resultados. El pH alcalino de la masa y 30 a 40 minutos de reposo redujeron en 100% la AFB1 adicionada. Se identificaron dos moléculas de degradación.

Conclusión. Los principales factores involucrados en la disminución de la AFB1 durante la producción de tortillas son la hidrólisis alcalina y el tiempo de reposo. Se propone un procedimiento para la producción de tortilla que reducirá la AFB₁ residual evitando el efecto acumulativo en los consumidores.

Palabras clave: aflatoxina B₁; productos de degradación; alkalinidad; Aspergillus; México.

Aflatoxin B₁ (AFB₁) is a fungal secondary metabolite produced by Aspergillus flavus and A. parasiticus.¹ This mycotoxin is considered a potent genotoxic and carcinogenic substance, classified in the category 1 by the International Agency for Research on cancer in 1993.² Exposure to AFB₁ is related with acute non-viral hepatitis, aflatoxin-related immune suppression, liver cancer, nutrition-related problems and even death.^3,4

Due to this important global issue, maximum levels of AFB₁ have been established at Codex alimentarius in animal and human foods in at least 99 countries.⁵ The level of AFB₁ contamination varies widely: in Africa, 355µg of AFB₁/kg were reported in 2005 whereas much higher levels (955µg AFB₁/kg) were detected in Mexico in 1989.^3,6 Some Latin America countries have the regulatory limit of 5-20µg/kg, depending of the kind of food.⁷ In Mexico maize is mainly used for human consumption. Around 12 millions tons are consumed per year, 6.3 millions tons of which in form of tortilla (in rural areas 217.9g/day and in urban areas 155g/day per capita).8 According to the Mexican Norm,⁹ the permitted maximum level of AFB₁ in Mexico is 20µg/kg in maize, and 12µg/kg in harina nixtamalizada (maize flour to make tortillas).

Nixtamalización is as old as Aztec civilization. It is estimated that nixtamalización began 1 200 years B.C.¹⁰ The traditional nixtamalización has been reported highly effective in removing AFB₁ from spiked maize-dough. Using radioactive aflatoxin solution, it was observed that most of the radioactivity was discarded into the cooking liquid and the first wash, confirming the elimination of AFB₁ by-products.¹¹ The effectiveness of nixtamalización process was also reported by Mendez-Albores and colleagues.^12,13 They reported a reduction of AFB₁ of 92 and 93.2% respectively. Nixtamalización process can be divided into four steps: boiling maize with water and limestone, soaking the mixture; washing the cooking maize and milling nixtamal to obtain the maize-dough (masa).¹⁴ Many reports have demonstrated that nixtamalización process inactivated AFB₁ by 85-95%.^6,15 However, for high aflatoxin contamination (520µg/kg), the percentage of AFB₁ inactivation was reduced to 93%,6 therefore 7% of AFB₁ remains in the tortilla, representing a health hazard.¹⁶ Mexican people would daily ingest 0.95µg of AFB₁ for each tortilla and 4.75µg AFB₁/5 tortillas.⁶ There are some evidences that the pH has an essential role in the inactivation of AFB₁ in different substrates.¹⁷ Hence, in the present work we analyzed the role of alkaline pH of maize-dough during the tortilla making process in the AFB₁ contamination. Additionally, we analyzed the AFB₁ degradation by-products generated during alkaline treatment.

The AFB₁ standard was obtained from Sigma-Aldrich (St. Louis, MO, USA). All solvents, including acetone, acetonitrile, benzene, chloroform and methanol, were HPLC grade and obtained from J.T. Baker (Mallinckrodt Baker, Inc. Mexico). Limestone was obtained from a local market in Irapuato, Guanajuato, Mexico.

Preparation of maize-dough

White maize (1 kg) was boiled for 45 min at 90°C in 3L of water with 10g of lime (minimum content of Ca(OH)2) = 90%) and left to soak overnight (18h at 24°C). Afterwards, the cooked maize was rinsed once with tap water (to remove pericarps) yielding 2kg of alkaline (nixtamal) maize ready for grinding. When this maize was ground into maize-dough, it had a pH of 10.2. Ten samples of maize-dough (50g each) were analyzed to determine the natural aflatoxin contamination.

Treatments of the maize-dough

Fifty grams samples of maize-dough (pH 10.2) were spiked with known amounts of AFB₁ standards solution 6.25µg/ 50g (125µg/kg of AFB₁). The AFB₁ was spiked in the middle of samples, after 5 min the corn dough was mixed with a spatula. Ten samples were maintained at room temperature (24°C), the other ten samples were flattened and cooked at 150°C during 5 min each side (tortillas). To evaluate the effect of the treatment over the spiked mycotoxin, AFB₁ determination and quantitation was done after 18h by HPLC-UV, HPLC(High-Pressure Liquid Chromatography with UV detector)-fluorescence and DIESI-MS. The negative control was maize-dough without addition of AFB₁.

Kinetic of time exposure of AFB₁ in alkaline maize dough pH 10.2

Effect of pH on AFB₁ standard solution

The standard solution of AFB₁ in methanol was used to analyze the effect of varying pH on aflatoxin stability. Three ml of AFB₁ standards were placed in glass tubes and 3ml of methanol adjusted to different pH values with 0.1 M NaOH (pH 8.0, 11.9 and 12.5) was added. A standard AFB₁ solution in non-buffered methanol was used as the reference. Three replicates for each pH value were prepared and they were placed in darkness at 14°C for 24 h. Afterwards, the liquid was removed by evaporation in water bath at 80°C and the contents re-dissolved in 1ml methanol. After filtration through an Extract CleanTM SPE C18 column (50mg per 1.5ml) (Grace Davison Discovery Science, Deerfield, Il 60015) the extracts were quantified by HPLC-UV and HPLC-fluorescence as described below.

Kinetic of time exposure of AFB₁ in alkaline methanolic solution

Methanol solution of standard AFB₁ (66.5μg/ml) at pH 10.5 was prepared and analyzed by DIESI-MS after different incubation times 10, 20, 30 and 40 min.

Extraction and quantification of aflatoxin by HPLC

Extraction of AFB₁ from samples of maize-dough was done according to the modification of the method 1 Association of Official Analytical Chemists (AOAC).¹⁸ Quantitative determination of AFB₁ in the extracts were made by HPLC using a Supelco C18 column 4.6 x 250nm in a HPLC Agilent Technologies 1200 (ABC Instrumentación Análitica SA de CV), Software Agilent chemstation. The isocratic mobile phase was a mixture of water:acetonitrile:methanol (60:20:20 by volume). Elution of aflatoxin was recorded at 364nm. The detection level of this system was: 21ng AFB₁ in 20µl of extract.

Prederivatization with trifluoroacetic acid

In order to confirm the negative values of the extract treatments, a derivatization with trifluoroacetic acid was performed according to AOAC,¹⁹ and a fluorescence detector was used at 360nm of excitation and 440nm of emission at 40°C in the Agilent 1200 HPLC equipment. The detection limit of this technique is 0.26ng of AFB₁ in 10µl of extract.

DIESI-MS identification of possible derivative AFB₁ products

Samples were filtered through Extract Clean TM SPE C18 column. Next, the samples were subjected to mass spectrometric analysis. The maize-dough and methanol extracts were analyzed by DIESI-MS,^20,21 with the following conditions: flow rate: 10µL/ min, spray voltage: 5kV, capillary temperature: 350°C, capillary voltage: 15V tube lens: 60V mode. The second technique was full positive scan. Open MS, ToppView version 1.10.0 and mMass 3.0 were used to data processing.

Data analysis

Each experiment was repeated three times, and aflatoxin levels were logarithmically transformed before statistical analysis, which was performed using the R version 2.15.2 (The R foundation for Statistical Computing).²² Differences between values were evaluated by Tukey's test (p=0.05) using SAS (version 6.12; SAS Institute, Cary, NC, USA).

Effect of the tortilla making process over AFB₁

Ten non-spiked samples (containing low levels of AFB₁, identified as natural contamination level 0.034μg/50g) and ten more samples spiked with 6.25µg AFB₁/50g of corn, a value equivalent to 125µg/kg were used to make tortillas. Samples of maize-dough containing low levels of AFB₁ did not present any measurable AFB₁, suggesting 100% destruction. Meanwhile, 2.9µg/kg (0.145µg/50g) remained in those samples spiked with high levels of AFB₁ (125µg/kg). These data suggested that under the conditions used in this study, the tortilla making process destroys about 97.7% of aflatoxins (data not shown). When a more sensitive method (fluorescence determination) was used, a detection limit of 0.26µg/10µl was reached, the same samples presented 12.52µg AFB₁/kg corresponding to 90% of aflatoxin reduction. By DIESI-MS it was not found any ionized AFB₁ indicating 100% reduction.

On the other hand, the effect of temperature over AFB₁ content during flattening and cooking of the thin maize-cake was addressed. Samples of 50g of maize-dough pH 10.2 with 30 min of resting at room temperature spiked with AFB₁ 125µg/kg (6.25µg AFB₁/50 g) was used (data not shown).

Effect of alkaline conditions over AFB₁ reduction

The time-course of aflatoxin degradation in maize-dough at pH 10.2 suggested that most (~100%) of the AFB₁ was eliminated with 30 min exposure. Prederivatization of samples and fluorometric detection of AFB₁ yielded similar data (table I). Nevertheless, the kinetics of AFB₁ reduction analyzed with DIESI-MS in methanol solution indicated that the optimal time was 40 min (figure 1).

Tests with two concentration of AFB₁ in buffered solutions showed that pH values around neutral were much less effective in destroying aflatoxins than more alkaline conditions. At pH 8.0 only 19.10% of AFB₁ was reduced and 80% remained in the solution. However, at pH 11.8 and 12.5, the aflatoxin was reduced 90 and 100% respectively (table II).

Degradative products of AFB₁ under alkaline conditions

The alkaline treatment of standard AFB₁ produced at least two degradation molecules that were observed in the DIESI-MS spectra, these two increased peaks in the treatment were not presented in the control (where the AFB₁ was not subjected to alkaline conditions) (figure 2A). The molecular formula of modified-AFB₁ was established; the smallest mass peak was identified being m/z 301.25 corresponding to C₁₇H₁₆O₅ with a mass calculated of 300.30 (figure 2B-2). The largest mass peak was observed at m/z 325.33, the mass calculated for C₁₇H₁₈O₅ was 302.32, indicating a derivative of aflatoxin B₁ (figure 2B-3).

Discussion

Aflatoxin has been named, quite rightly, an invisible food hazard. Aflatoxin remnant contamination has been reported in tortillas,¹¹ and even when very small amounts of aflatoxin are consumed there is a risk, due to its accumulative effect. Chronic exposure to low levels of AFB₁ in patients with hepatitis B virus could contribute to the development of hepatocellular carcinoma.²³ In humans 1.4-2.3% of the AFB₁ ingested binds covalently to serum albumin, and ingestion of 2-6mg/day for a month can cause acute hepatitis and death.²⁴ Therefore, a procedure to obtain tortilla free of aflatoxin is very important, since its consumption has increased in the last years as a consequence of the migration of Latinamerican population, and the adoption of this food in those countries.

During the tortilla making process, two main degradation mechanisms of AFB₁ could be identified: one resulting from the alkaline pH of the maize-dough, the other due to the applied high temperature for cooking the flattened dough. Resting the dough at least 40 min (figure 1), at room temperature (24°C) at elevated pH (10.2) (tables I and II) eliminated nearby 100%, which indicates that alkalinity is the most effective component of the tortilla making process. The results strongly suggested that the effect of alkaline pH is the primary factor responsible for the AFB₁ destruction. The loss of fluorescence indicates a structure change of AFB₁ due to the alkaline conditions. It is well documented that alkalinity enhances the opening of the lactone ring of AFB₁ to form a substituted o-coumaric acid B.¹⁷ However, also more drastic changes have been observed, for example, decarboxylation of the carboxylic acid and further decomposition.^17,25

We observed similar effects: Highly alkaline conditions (pH 12.5) rapidly destroyed 100% of AFB₁ (table II). Those results confirmed the effectiveness of nixtamalización process in the reduction of AFB₁ as it has been reported by different authors.^11-13 However, several authors suggested that the nonfluorescent molecule formed during nixtamalización is able to reforming to AFB₁ when the tortilla reaches the stomach pH.^13,26 In opposition to this hypothesis, Anguiano-Ruvalcaba and colleagues⁶ found that under gastrointestinal tract similar conditions (pH 1.5, 37 °C, and 30 min the time that food remains in the stomach) there is not a recovering of AFB₁. This conclusion is supported by Yates and colleagues.²⁷ They demonstrated by C-NMR that at low pH, the lactone rings closed but not via a simple reversal or the ring opening, rather, prolonged incubation times are necessary. Additionally, Anguiano-Ruvalcaba and colleagues⁶ showed that young chicks fed with maize dough from nixtamalization did not present any symptoms compared with young chicks fed with contaminated dough. It is important to mention that only less of 10% remains during the nixtamalización since according with radiolabel AFB₁ assays, 90% is eliminated in the cooking liquid and the first wash.¹¹ In this way, only the remnant AFB₁ could have an accumulative effect in the consumers.

Three rinses produce a less alkaline maize-dough, Perez-Flores and colleagues²⁹ had suggested that higher amounts of lime should be used to obtain more alkaline maize-dough. However, addition of 1% lime and subsequent rinsing once with tap water produces a maize-dough with a pH about 10.2.

In order to find out the chemical modification that AFB₁ suffers during the process, samples containing AFB₁ standard under alkaline conditions were analized by mass spectrometry. The ionized AFB₁ has a monoisotopic weight of 335.27 m/z, corresponding to the sodium adduct (figure 2A). Two new signals at 301.25 m/z and 325.33 m/z (figures 2B-2 and 2B-3) indicated the generation of degradation products. Fragmentation of the ions was not possible due to their low abundance. However, both signals exhibit isotopic patterns which are consistent with compounds related to AFB₁. Revising possible chemical structures, we suggest the formation of compounds 2 and 3 (figure 2B), where compound 2 is known as aflatoxin D1 a non-toxic one.³⁰

In summary, the alkalinity of the maize-dough is the primary factor to aflatoxin removal. For this reason, toxin-free tortillas can be produced by alkaline conditions and at least 40 min resting time of the maize-dough. Changes in the procedure used to make tortillas (figure 3) and rigorous implementation of this regime is a simple but efficient way of reducing or eliminating AFB₁ from maize-based foods. Our findings also might be transferred to reduce of AFB₁ of other maize food products.

Acknowledgment

We thank William Broughton for his critical review and constructive suggestions to this manuscript.

1. Tritscher A, Miyagishima K, Nishida C, Branca F. Ensuring food safety and nutrition security to protect consumer health: 50 years of the Codex Alimentarius Commission. Bull World Health Organ 2013;91:468-8A.         [ Links ]

2. International Agency for Research on Cancer. IARC Some naturally occurring substances: food items and constituents, heterocyclic aromatic amines and mycotoxins. Monographs on the evaluation of carcinogenic Risks of Chemicals to humans World Health Organization 1993;56:245-540.         [ Links ]

3. Azziz-Baumgartner E, Lindblade K, Gieseker K, Rogers HS, Kieszak S, Njapau H, et al. Case-control study of an acute aflatoxicosis outbreak, Kenya 2004. Environ Health Perspect 2005;113:1779-1783.         [ Links ]

4. Chen CJ, Wang LY, Lu SN, Wu MH, You SL, Zhang YJ, et al. Elevated aflatoxin exposure and increased risk of hepatocellular carcinoma. Hepatol 1996;24:38-42.         [ Links ]

5. Food and Agriculture Organization. Worldwide Regulations for Mycotoxins in Food and Feed in 2003. FAO Food and Nutrition Paper 81. Rome, Italy: FAO/UN, 2004.         [ Links ]

6. Anguiano-Ruvalcaba GL, Verver y Vargas-Cortina A, Guzman-de Peña D. Inactivation of aflatoxin B₁ and aflatoxicol through traditional "nixtamalización" of corn and their regeneration by acidification of corn dough. Salud Publica Mex 2005;47:369-375.         [ Links ]

7. Commission Regulation (EU) No. 165/2010 of February 2010 amending Regulation (EC) No. 1881/2006 setting maximum levels for certain contaminants in foodstuffs as regards aflatoxin. 2010:8-12.         [ Links ]

8. Secretaria de Economía. Análisis de la cadena de valor maíz-tortilla: situación actual y factores de competencia local [documento en internet]. 2013 [consultado en marzo 2014]. Disponible en: www.economia.gob.mx.         [ Links ]

9. Secretaría de Salud. NOM-188-SSA1-2002. Productos y Servicios. Control de aflatoxinas en cereales para consumo humano y animal. Especificaciones sanitarias [documento en internet]. 2002 [consultado en marzo de 2014]. Disponible en: http://www.salud.gob.mx/unidades/cdi/nom/188ssa12.html        [ Links ]

10. Krickeberg W. Las antiguas culturas mexicanas. México: FCE, 1961.         [ Links ]

11. Guzmán-de-Peña D, Trudel L, Wogan GN. Corn "nixtamalización" and the fate of radiolabelled aflatoxin B₁ in the tortilla making process. Bull Environ Contam Toxicol 1995;55:858-864.         [ Links ]

12. Mendez-Albores JA, Arámbula-Villa G, Loarca-Pina, G, González-Hernández J, Castaño-Tostado E, Moreno Martínez E. Aflatoxins fate during the nixtamalization of contaminated maize by two tortilla-making processes. J Stored Prod Res 2004a;40:87-94.         [ Links ]

13. Méndez-Albores A, De Jesús-Flores F, Castañeda-Roldán E, Arámbula-Villa G, Moreno Martínez E. Aflatoxin-detoxification achieved with mexican traditional nixtamalization process (MTNP) is reversible. J Sci Food Agric 2004b;84:1611-1614.         [ Links ]

14. Guzmán-de-Peña D. The destruction of aflatoxins in corn by "nixtamalización". En: Mycotoxins in food, feed and bioweapons. Berlin: Springer, 2010:39-49.         [ Links ]

15. Elias-Orozco R, Castellanos-Nava A, Gaytán-Martínez M, Figueroa-Cárdenas JD, Loarca-Pina G. Comparison of nixtamalization and extrusion processes for a reduction in aflatoxin content. Food Addit Contam 2002;19:878-885.         [ Links ]

16. Williams JH, Phillips TD, Jolly PE, Stiles JK, Jolly CM, Aggarwal D. Human aflatoxicosis in developing countries: a review of toxicology, exposure, potential health consequences, and interventions. Am J Clin Nutr 2004;80:1106-1122.         [ Links ]

17. Dollear FG. Detoxification of aflatoxins in foods and feeds. en Aflatoxin: scientific background, control and implications. New York: Academic Press, 1968:359-391.         [ Links ]

18. Guzmán-de-Peña D, Anguiano-Ruvalcaba GL, Medina-Arredondo JJ. Modification of the method1 OACC (CB Method) for the detection of aflatoxins. Bull Environ Contam Toxicol 1992;49:485-489.         [ Links ]

19. AOAC. Official Methods of Analysis of the AOAC International (25 ed). Natural toxins chapter 49 (ed.). Gaithersburg, Maryland: AOAC, 2005:45-47.         [ Links ]

20. García-Flores M, Juárez-Colunga S, Montero-Vargas JM, López-Arciniega JAI, Chagolla A, Tiessen A, et al. Evaluating the physiological state of maize (Zea mays L.) plants by direct-injection electrospray mass spectrometry (DIESI-MS). Mol bioSyst 2012;8:1658-1660.         [ Links ]

21. Montero-Vargas JM, Lindbergh H, González-González E, Gálvez-Ponce E, Ramírez-Chávez J, Molina-Torres J, et al. Metabolic phenotyping for the classification of coffee trees and the exploration of selection markers. Mol bioSyst 2013;9:693-699.         [ Links ]

22. R Development Core Team. R: a language and environment for statistical computing [document en internet]. Austria: R Foundation for Statistical Computing, 2012 [consultado en mayo 2014]. Disponible en: www.R-project.org.         [ Links ]

23. Wild C, Jiang YZ, Allen SJ, Jansen LAM, Hall AJ, Montesano R. Aflatoxin-Albumin adducts in human sera from different regions of the world. Carcinogenesis 1990;11:2271-2274.         [ Links ]

24. Gan LS, Skipper PL, Peng XC, Groopman JD, Chen JS, Wogan GN, et al. Serum albumin adducts in the molecular epidemiology of aflatoxin carcinogenesis: correlation with aflatoxin B₁ intake and urinary excretion of aflatoxin M1. Carcinogenesis 1988; 9:1323-1325        [ Links ]

25. Méndez-Albores A, Nicolás-Vázquez I, Miranda-Ruvalcaba R, Moreno-Martínez E. Mass spectrometry/mass spectrometry/spectrometry study on the degradation of B-aflatoxins in maize with aqueous citric acid. AM J Agril & Biol Sci 2008;3:482-489.         [ Links ]

26. Price RL, Jorgensen KV. Effect of processing on aflatoxin levels and mutagenic potential of tortillas made from natural contaminated corn. J Food Sci 1985;43:635-638.         [ Links ]

27. Yates EA, Phylipp B, Buckley C, Atkinson S, Chhabra SR, Stockett E, et al. Lactones undergo lactonololysis in a pH-temperature, and acyl chain length-dependent manner during growth of Yersinia pseudotuberculosis and Pseudomonas aeruginosa. Infect Inmun 2002;70:5635-5646.         [ Links ]

28. Kabak B. The fate of mycotoxins during thermal food processing. J Sci Food Agric 2009;89:549-554.         [ Links ]

29. Pérez-Flores GC, Moreno-Martínez E, Méndez-Albores A. Effect of microwave heating during alkaline-cooking of aflatoxin contaminated maize. J Food Sci 2011;76:T48-T52.         [ Links ]

30. Grove MD, Plattner RD, Peterson RE. Detection of aflatoxin D1 in ammoniated corn by mass spectrometry-mass spectrometry. Appl Environ Microbiol 1984;48:887-889.         [ Links ]

Corresponding author:
Dra. Dora Linda Guzmán de Peña.
Departamento de Biotecnología y Bioquímica,
Centro de Investigación y de Estudios Avanzados-IPN,
unidad Irapuato. Km 9.6 libramiento Norte Irapuato-León.
36821 Irapuato, Guanajuato, México.
E-mail: dguzman@ira.cinvestav.mx

Declaration of conflict of interests. The authors declare that they have no conflict of interests