Aflatoxins in mesir paste

ABSTRACT In this research, 42 mesir paste samples produced by a limited number of manufacturers in Manisa, Turkey, were analysed for aflatoxin content by high-performance liquid chromatography with a fluorescence detector after pre-separation using immunoaffinity columns. A good correlation was found with good performance in terms of precision for the method. Limit of detection-limit of quantitation of AFG2-AFB2 and AFG1-AFB1 were 0.05/0.15 and 0.03/0.09 µg/kg, respectively. AFG2, AFG1 and AFB1 were detected in 29%, 71% and 52% samples, whereas AFB2 was not detected in any sample. AFtotal, AFG2, AFG1 and AFB1 content varied between 0.04/10.20 µg/kg, 0.07/0.42 µg/kg, 2.02/10.11 µg/kg and 0.04/0.11 µg/kg, respectively. About 9.5% of samples were found to be above the maxium limit (ML) for AFtotal (10 µg/kg), but none of samples exceeded the ML for AFB1 (5 µg/kg) as set by the European Union for spices. According to the results, it was concluded that traditional products like mesir paste which contains different types of spices should be examined in terms of aflatoxins.

For mesir paste production, up to 42 different spices are weighed and grinded in a mortar. The mixture of spices is boiled with honey, sugar, glucose syrup and citric acid in boilers until a paste consistency is obtained (Giritlioglu, Avcikurt, & Savas, 2010;Hışıl, 1994). Differences were observed in spices used in the production of mesir paste in the recent years. The number of spices used was 20-25 in some products while it was 30-31 and 40-41 in some others. The composition of spices used in mesir paste gives the characteristic aroma and taste to the paste, while production, formulation and process parameters vary depending on the manufacturer. Mesir paste is produced by a limited number of manufacturers, only seven firms in Manisa and by a total of 10 firms in Turkey. Thus, it is difficult to speak of a standard production method for mesir paste. As a result, standardization in production carried out by the same firm at different times, or of mesir paste products produced by different companies, is recommended (Artık et al., 1999;Giritlioglu et al., 2010;Güven, 2010;Nergiz & Yıldız, 1995).
Spices that have a high possibility of containing mycotoxins are used in the preparation of mesir paste, which is a traditional confectionery product. Drying the plants used as the spices in mesir paste production is a critical step in the course of mould growth and mycotoxin contamination. Cultivation of spices in warm and humid areas also increases the mould and bacterial contamination risks (Ali, Hashim, & Shuib, 2015;Romagnoli, Menna, Gruppioni, & Bergamini, 2007). Mycotoxins are the secondary metabolites of mould subspecies growing on agricultural products, including spices at various stages of preparation from field to consumption, and depend for the growth on ecological conditions. Among 400 mycotoxins, aflatoxins (AF) are the most dangerous ones in terms of human health and are commonly seen in spices. Although 20 aflatoxin types have been identified, AFG 2 , AFG 1 , AFB 2 and AFB 1 are the most common ones and also, AFB 1 is the most toxic one. The International Agency for Research on Cancer has included AFB 1 , one of the most dangerous AF, in its Group 1 classification owing to evidence of its carcinogenic effect in humans. (International Agency for Research on Cancer, 1993;Marin, Ramos, Cano-Sancho, & Sanchis, 2013;O' Riordan & Wilkinson, 2008).

Samples
A total of 42 samples of mesir paste were purchased from 7 firms in Manisa province from April 2014 to May 2015. The 6 different samples belonged to different batches were collected in each firms, and each sample weighed at least 1 kg. The samples were saved in airtight glass jars, and stored at 4°C until the analysis.

Chemicals and reagents
The mixed standard of AFG 2 , AFG 1 , AFB 2 and AFB 1 was supplied by Supelco (Bellefonte, PA, USA). The mixture in each bottle consists of 0.3 µg/mL AFG 2 and 1 µg/mL AFG 1 , 0.3 µg/mL AFB 2 and 1 µg/mL AFB 1 in methanol. Working standard solutions were prepared daily from these standard solutions. The clean-up step immunoaffinity columns (AflaTest™) were purchased from VICAM, Watertown, MA, USA. Ultra-pure water was produced by a Milli-Q system from Thermo Scientific, Milford, MA, USA. Solvents like acetonitrile and methanol (HPLC grade) were purchased from Sigma Aldrich, St. Louis, MO, USA. All other chemicals and reagents were used at least of analytical grade. Due to the potential toxicity, all experimental supplies were disposable products.

Chromatographic apparatus
Chromatographic analysis was carried out by using an Agilent 1260 Infinity system (Palo Alto, CA, USA) with an auto sampler using a fluorescence detector. Chromatographic separations of AFs were achieved with 5 µm Thermo Scientific (Waltham, MA, USA) Hypersil™ ODS C 18 , 250 × 4.6 mm column. The wavelengths for excitation and emission were 360 nm and 440 nm, respectively. The mobile phase was a mixture of water, methanol and acetonitrile (50:30:20%; v/v) with 350 µL of 4 M nitric acid and 120 mg KBr per 1 l of mobile phase. The flow rate was 1 mL min −1 , and column temperature was maintained at 25°C. The injection volume was 100 µL. Post-column derivatisation was carried out with electrochemically generated bromine in Cobra cell (Coring System Diagnostix, GmnH, Gernsheim, Germany) (AOAC, 2000).

Extraction of aflatoxins
The extraction of AFs from mesir paste was carried out as described in AOAC Official Method 999.07 (AOAC 2000) with some modifications. A 25 g of sample was mixed with 2.5 g of NaCl and blended with 150 mL methanol:water (80:20%; v/v) using a Waring blender at high speed for 10 min. The extract was filtered with filter paper (Whatman no: 4) and collected into clean flasks.

Separation with immunoaffinity columns
AFs' clean-up was performed using immunoaffinity columns according to the method AOAC Official Method 999.07 (AOAC 2000) with slight modifications. A 5 mL of the filtrate was diluted with 20 mL phosphate buffered saline solution (pH: 7.3). The diluted extract (25 mL) was vortexed for 1 min and centrifuged at 4000 rpm/min at 4°C and then filtered through glass microfiber filter (1.5 µm, 110 mm, Vicam, Watertown, MA, USA). All filtrate was passed through an AflaTest TM immunoaffinity column attached onto a vacuum manifold at a rate of 2-3 mL min −1 and washed with 15 mL ultrapure water at the same flow rate for cleaning. The column was quickly dried by passing air for 1.5 min. AFs were eluted from the column by applying 1 mL methanol and 1 mL ultra-pure water and collected in a HPLC vial and stored in a refrigerator at 4°C until analysis.

Method validation
The method validation parameters like linearity, limit of detections (LOD), and limit of quantifications (LOQ), recovery of AFG 2 , AFG 1 , AFB 2 and AFB 1 were determined (Iqbal, Rabbani, Asi, & Jinap, 2014). Instrumental precision was demonstrated as instrumental repeatability (within-day precision) and intermediate precision (between-day precision). The linearity of the method was determined by five-point calibration curves. The calibration curves of AFG 2 and AFB 2 were linear between 0.058 and 5.184 µg kg −1 ; AFG 1 and AFB 1 were linear between 0.192 and 17.28 µg kg −1 . The LODs based on a signal-to-noise S/N = 3/1 and LOQs were calculated with a signal-to-noise S/N = 10/1 as described in ISO11843 part 2 (British Standard, 2000). Relative standard deviation (RSD) percentages were calculated to determine for instrumental precision, in terms of instrumental repeatability and intermediate precision at one standard concentration on the same day with six-replicated and on the different three days with three-replicated. The recoveries were ascertained by spiking 1-3 µg/kg of AFG 2 and AFB 2 , 2-5 µg/kg of AFG 1 and AFB 1 in non-contaminated samples in two parallels at least an hour before the analysis. RSD percentages were calculated to determine for the method repeatability.

Statistical analysis
Experiments were performed triplicate using samples. The data of AFs were statistically analysed and presented as mean ± standard deviation (SD), and coefficient of determination (R 2 ) was determined by regression/correlation analysis using SPSS software (IBM, PASW Statistics 19, Armonk, NY, USA).

Method validation
The method performance for AFs determination in mesir paste is summarised in Table 1. The results have shown good linear response with coefficient of determination R 2 ≥ 0.99 for both AFs. The calibration curves of AFG 2 and AFB 2 were linear between 0.058 and 5.184 µg kg −1 ; AFG 1 and AFB 1 were linear between 0.192 and 17.28 µg kg −1 . Calculated LOD and LOQ were 0.05 and 0.15 µg kg −1 for AFG 2 and AFB 2 and 0.03 and 0.09 µg kg −1 for AFG 1 and AFB 1 , respectively. They were well below the legal limits to be controlled. Instrumental precision was determined in terms of instrumental repeatability and intermediate precision at one standard concentration on the same day with six-replicated and on the different three days with three-replicated. The % RSD value was calculated. % RSD of instrumental repeatability for AFG 2 , AFG 1 , AFB 2 , and AFB 1 were 0.78, 0.95, 1.05 and 1.24, respectively. % RSD of intermediate precision for AFG 2 , AFG 1 , AFB 2 and AFB 1 were 2.22, 2.52, 2.67 and 3.06, respectively. Instrumental repeatability was under 2%, intermediate precision was under 5%. The retention times of AFG 2 , AFG 1 , AFB 2 and AFB 1 were 5.94, 6.64, 7.47 and 8.51 min, respectively. The recoveries were ascertained by spiking 1-3 µg/kg of AFG 2 and AFB 2 , 2-5 µg/kg of AFG 1 and AFB 1 in non-contaminated samples in two parallels. For the first spike level: AFG 2 88.40%, AFG 1 96.96%, AFB 2 118.66% and AFB 1 68.00%. For the second spike level, recoveries were AFG 2 92.77%, AFG 1 70.09%, AFB 2 90.39% and AFB 1 69.43%. The method has shown good recoveries for AFs. The RSD percentages were ranged 2.42-10.91 for first spike level and ranged 5.57-8.41 for the second spike level for the four AF.

Afs in mesir paste
AFG 2 , AFG 1 , AFB 2 and AFB 1 analyses were conducted on a total of 42 mesir paste samples in 3 parallel samples. The occurrence and mean contamination of AF in mesir paste samples is shown in Table 2.
The results showed mediocre incidence of these mycotoxins in analysed samples. In mesir paste samples, AFG 2 values varied between 0.07 and 0.42 µg kg -1 , AFG 1 values varied between 2.02 and 10.11 µg kg -1 and AFB 1 values varied between 0.04 and 0.11 µg kg -1 . AFB 2 was not detected in any of the samples. Total aflatoxin values in the samples ranged from 0.04 to 10.20 µg kg -1 , and the average was 5.52 µg kg -1 . Thirty of samples (71%) samples were below the AFG 2 detection limit (0.05 µg kg -1 ), 12 (29%) samples were below the AFG 1 detection limit (0.03 µg kg -1 ) and 20 (48%) samples were below the AFB 1 (0.03 µg kg -1 ) detection limit, and also all of samples were below AFB 2 detection limit (0.05 µg kg -1 ). In the light of the data obtained, it is seen that the results for the analysed samples, obtained  Table 2. Occurrence and mean contamination of aflatoxins in mesir paste (n = 42).
Parameter AFG 2 AFG 1 AFB 2 AFB 1 AFs Total * Positive samples** n (%) 12 (29) 30 (71) -22 (52) 32 (76) Samples above 10 µg kg −1 n (%) ----4 (9.5) Range (µg kg −1 ) LOD-0.42 LOD-10.11 ˂LOD LOD-0.11 LOD-10.20 Mean ± SD (µg kg −1 ) 0.24 ± 0.17 5.74 ± 2.67 -0.07 ± 0.02 5.52 ± 3.01 *Sum of AFG 2 , AFG 1 , AFB 2 and AFB 1 . **AFs' level ˃ LOD. *Suma de AFG 2 , AFG 1 , AFB 2 y AFB 1 ** Nivel de AFs˃ LOD from different producers, and is related to the varying formulations in the preparation of mesir paste. There are legal regulations in Turkey and in the European Union (EU) for levels of AF (total AF and AFB 1 ) permitted for nuts, cereals, dried fruits and some spices, but there are no official regulations in Turkey or in the European Union regarding the maximum allowable total AF and AFB 1 levels in mesir paste. The results of the samples tested here were below the maximum limit (ML) values determined for spices (5 µg kg -1 for AFB 1 ) for comparison with the results obtained in this study. However, the mean of value of total AF was found to be 5.52 µg kg -1 in samples; 4 samples (9.5%) were found to be above the EU MLs (10 µg kg -1 for total AFs). In an extensive literature search, no studies or reports were found that examined AF contamination in mesir paste. Therefore, the other mycotoxin contamination must consider also traditional products like mesir paste.
It is believed that careful selection of raw materials especially spices during the preparation of mesir paste and fulfilling recommended production and storage requirements may prevent the emergence of AF, which have harmful effects on health. Also, the examination of traditional products in terms of mycotoxins will help raise awareness among both producers and consumers of these products about their possible contamination.

Conclusion
In this study, aflatoxin contamination risk for mesir paste was evaluated, considering the negative effects of AF on health. Mesir paste is a traditional confectionery product in Manisa and fondly consumed, having claimed functional properties, including antioxidative and energizing ones. This is the first study that focused on determining AFG 2 , AFG 1 , AFB 2 , and AFB 1 contamination levels in mesir paste. Although mesir paste carries aflatoxin contamination risks, it was not previously investigated in any study in terms of AF and other mycotoxins. This study revealed positive results in terms of aflatoxin contamination, and these results suggest the necessity of investigating this kind of traditional product in terms of its mycotoxin content. Also, the examination of traditional products in terms of mycotoxin content will help raise awareness in both the producers and the consumers of these products. When the aflatoxin contamination were determined of the samples, it was obviously seen that some producers have high awareness because aflatoxin contamination was not found in all batches. Measuring AFG 2 , AFG 1 , AFB 2 , and AFB 1 contamination in traditional mesir paste revealed the necessity of paying greater attention and care in the selection of spices used in mesir paste production and in their storage conditions, from procurement to production.

Disclosure statement
No potential conflict of interest was reported by the author.