الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 145 |  |  |
| | | from | 145 | | |
Abstract
Aflatoxins are secondary metabolites produced by a variety of fungal strains, including Aspergillus Flavus and Aspergillus Parasiticus, and have been found as contaminants in a variety of foods and feeds. Aflatoxin B1, B2, G1, and G2 are the most common aflatoxins, with aflatoxin B1 being the most potent naturally occurring carcinogen. Aflatoxins, particularly M1, are most likely present in dairy products as a B1 metabolite as a result of animals being fed. Aflatoxin M1 (AFM1) is a heat-stable compound with a high detection rate in pasteurized milk as well as other dairy products derived from contaminated raw milk. This includes many products like whey protein, infant formulas, and colostrum-based products.
According to the International Agency for Research on Cancer (IARC), AFM1 is a possible human carcinogen (group 2B). According to the EU Commission Regulation No. 165/2010/EC and the recently promulgated binding technical rules by the national regulatory agency, limits of 0.05 μg/kg and 0.025 μg/kg were applied for milk-based products and dietary foods intended specifically for infants.
Hence, this study aimed to develop a simple solid-liquid extraction protocol for efficient AFM1 extraction from colostrum-based supplements and whey protein-based sports food, followed by rapid and accurate determination with an LC-ESI-MS/MS system. Furthermore, method validation must be carried out by EU guidelines to comply with EU regulatory limits. selectivity, recovery percentages, repeatability, intermediate precision, LOD, LOQ, linearity range, and linear regression coefficient (R2) are all validation parameters. To further validate the novel SLE method, we not only compared it to established protocols like IAC and QuEChERS, but also statistically assessed their performance using a one-way analysis of variance (ANOVA). Thirty samples representing the target commodities (colostrum and whey protein products) were analyzed using all three techniques, allowing for a robust comparison of accuracy, precision, and efficiency. The validated assay’s practicality was tested by applying it to real commercial samples and proficiency testing samples (PT).
The obtained results can be summarized as following:
5.1. LC-MS/MS Optimization
A reference standard solution of AFM1 was directly infused into ESI-MS/MS under automatic full scanning mode to achieve the best sensitivity and maximum ion stability possible. A parent ion [M+1]+ of 329.1 m/z and two ions (273.2 and 259.1 m/z) were selected for quantitation and further confirmation, respectively.
Optimizing chromatographic separation for rapid and sensitive AFM1 detection involved a two-pronged approach. First, we evaluated three candidate columns from different suppliers with varying particle sizes under multi-step gradient elution programs. This allowed us to identify the column that offered the most efficient separation and peak resolution. The Agilent Poroshell 120 EC-C18 column (50 mm × 4.6 mm, 2.7 μm) emerged as the dominated, paired with an elution system composed of 10-mM ammonium formate in water-MeOH (9:1 v/v, pH 5.0 ± 0.05) and MeOH. Second, we fine-tuned the injection volume, testing 1, 5, and 10 µL to achieve optimal sensitivity and peak characteristics. A 5 µL injection yielded the best balance between sensitivity and acceptable matrix effect percentage (ME%).
5.2. Sample Preparation Optimization
At a dilution factor of 1X to 5X, sample sizes of 2.5 and 5.0 g from the studied commodities were studied. A 5.0 g sample size was found to be appropriate for conducting AFM1 testing in the commodities studied at a 3X dilution level while maintaining tolerable MEs and achieving lower LOD and LOQ values.
5.3. Extraction optimization and matrix effect study
MeOH and MeCN, two commonly used extraction solvents, were tested at different mixing ratios with water (1:1, 3:2, 7:3, and 4:1, v/v) for direct extraction of AFM1 from whey protein-based sports food and colostrum-based supplements. Other than the failed water bath and ultrasonic, mechanical shaking at 700 RPM for 20 minutes yielded acceptable recovery percentages and tolerable ME% for MeCN-water (3:2, v/v). The sample purification took only 5 minutes at 16020 rcf.
5.4. Method validation
The recovery percentage, repeatability, intermediate precision, LOD, LOQ, linearity range, and linear regression coefficient (R2) results were 94.3–104.1% with RSDs of 4.70–8.40%, 4.49–9.7%, 7.89–10.6%, 0.0015 μg/kg, 0.005 μg/kg, 0.005–1.000 ng/mL, and 0.9999, respectively. In both tested commodities, the expanded uncertainty calculation has revealed a value of approximately ± 30% dispersion of the values that could be reasonably attributed to the measurand.
The validated SLE method was not only compared with established protocols like IAC and QuEChERS, but also statistically assessed their performance using a one-way analysis of variance (ANOVA). Thirty samples representing the target commodities (colostrum and whey protein products) were analyzed using all three techniques, allowing for a robust comparison of accuracy, precision, and efficiency.
Applicability was demonstrated on two proficiency testing (PT) samples and 223 domestic samples. Results have confirmed the practicality of the proposed assay protocol, with concentrations ranging from 0.010 to 4.423 μg/kg for posi0tive samples. Out of the tested samples, 1.5% of the whey protein-based sports foods violated the EU’s established limits.