Abstract:

An analytical method based on dispersive solid-phase extraction (d-SPE) and ultrafast liquid chromatography-tandem mass spectrometry (UFLC-MS/MS) was employed for the determination of 43 mycotoxins in chestnut flour and wheat flour. A total of 128 samples consisting of 48 chestnut samples and 80 wheat flour samples were collected randomly and subjected to analysis. Finally, five specific toxins were selected as markers to identify these two foodstuffs. Acetonitrile-water (84∶16, v/v) was used to extract mycotoxins from chestnut flour and wheat flour. After extraction, the supernatant was transferred to the d-SPE equipment, using which purification was performed with C₁₈ and EMR-Lipid (lipid adsorbent). Chromatographic separation was carried out by gradient elution with eluent A (ESI⁺: 0.1% formic acid, ESI^-: water) and eluent B (ESI⁺: methanol-acetonitrile (1∶1) containing 0.1% formic acid, ESI^-: acetonitrile) on a BEH C₁₈ column (100 mm×2.1 mm, 1.7 μm). Quantitative analysis was performed with the aid of matrix-matched curves. When establishing the method, the experimental matrix for optimization was designed by central-composite design based on the response surface methodology. Quadratic polynomial equations were deduced to describe the relationships between the responses and variables, and assess the interaction effects among the variables to acquire the true optimal conditions with less workload. Using the optimum experimental conditions, the accuracy of the proposed method was determined through three-level spiking tests, while the precision was evaluated in terms of the repeatability (six replications per level). Satisfactory precisions (RSDs≤7.5% in chestnut flour and RSDs≤9.3% in wheat flour) were achieved in all tested assays. The recoveries were also acceptable, and ranged from 72.4% to 109.4% for chestnut flour and from 70.7% to 112.9% for wheat flour. The matrix effects of mycotoxins were 48%-128% in wheat flour and 41%-112% in chestnut flour. The detectability of mycotoxins in the two matrices was assessed by spiking the blank extracts with various low concentrations, and determined as the lowest values that can produce chromatographic peaks at a signal-to-noise ratio (S/N) of 3∶1. The obtained limits of quantification varied from 0.10 μg/kg to 20 μg/kg (bongkrekic acid) in both investigated matrices. Satisfactory linearities were obtained, with correlation coefficients>0.9991 for all the analytes. After validation, the contamination status of the multiple mycotoxins was evaluated for various concentration ranges. Based on the obtained data, both wheat flour and chestnut flour were severely contaminated, with 17 mycotoxins detected in them. Particularly, chaetoglobosin A, ochratoxin B, and penicillic acid were only detected in chestnut flour, while 3-acetyl-deoxynivalenol, deoxynivalenol, and nivalenol were detected in wheat flour. Further, the positive rates and contamination concentrations of chaetoglobosin A, ochratoxin B, and penicillic acid were not significant; hence, they did not qualify as identification markers. On the other hand, the incidence of deoxynivalenol in wheat flour almost reached 100%, which is very significant. Finally, deoxynivalenol and its four derivatives (3-acetyl-deoxynivalenol, 15-acetyl-deoxynivalenol, deepoxy-deoxynivalenol, and nivalenol) were treated as adulteration markers for the two foodstuffs. To improve the reliability of the conclusion, all samples were re-tested using the first method prescribed by the National Food Safety Standard, i. e., GB 5009.111-2016. Ten chestnut flour samples were also randomly selected to prepare moldy samples under suitable environmental conditions for the growth of Fusarium, to verify the production and release of deoxynivalenol and its derivative mycotoxins under the extreme conditions. The distribution data for these mycotoxins were consistent with those obtained by d-SPE, confirming that the adulteration criterion is trustworthy.
The established method is simple, rapid, sensitive, and accurate, and can effectively meet the requirements for the simultaneous determination of multiple mycotoxins in chestnut flour and wheat flour. Moreover, the adulteration results, which were obtained for natural contaminants (deoxynivalenol and its four derivatives), are less affected by humans and hence, much more accurate and reliable.

Key words: dispersive solid-phase extraction (d-SPE), ultrafast liquid chromatography (UFLC), tandem mass spectrometry (MS/MS), mycotoxin, chestnut flour, wheat flour, adulteration, identification