Analysis and quantification of mycotoxins in cereals using time-scheduled MRMHR workflow

Using the X500R QTOF system with SCIEX OS software

Yang Zong 1, Cheng Haiyan 1, Zhong Hongjian 2, Li Lijun 1, Jin Wenhai 1 Phil Taylor3
1
SCIEX, China; 2Henan Academy of Agricultural Sciences; China; 3SCIEX, UK

Abstract

Mycotoxins are toxic secondary metabolites produced by fungi which can cause illness in humans and animals if they contaminate the food supply and many countries have set strict limits.1 Contamination of the food types such as grain can also ruin product and cause large financial loss. Rapid, high-throughput, and highly accurate mycotoxin detection methods are needed to test the safety of grain supplies. Here a targeted method for quantification of 15 different mycotoxins in wheat and corn was developed using the SCIEX X500R system. 

x500r_550x737

Introduction

Mycotoxins are toxic secondary metabolites produced by fungi under favorable conditions which can cause acute and chronic illness in humans and animals.1 The most common mycotoxins that threaten human health include aflatoxin (AF), deoxynivalenol (DON), ochratoxin A (OTA), fumonisin (FB), zearalenone (ZEN), and T-2 toxin(T2),2 which are widely found in many grains and oils, and their products. To date, many countries and regions have set strict mycotoxin limits,3 and China has revised its Food Safety Law in 2015 to explicitly include biotoxins in contaminants of major concern for the first time. China has also set limits for mycotoxins such as deoxynivalenol (DON), zearalenone (ZEN), aflatoxin B1(AFB1), and ochratoxin (OTA); for instance, the state food safety standard for cereals set the limit for ZEN at 60µg/kg, aflatoxin B1(AFB1) at 5-20µg/kg, and OTA at 5µg/kg.4

Abnormal climate conditions and insect damage have caused extensive mycotoxin contamination of grain in recent years, leading to severe damage and huge losses. As an example, in China, 31 million tons of grain are contaminated with mycotoxins annually during production, storage, and transport; this accounts for approximately 6.2% of total grain production. Rapid, high-throughput, and highly accurate mycotoxin detection methods with simple sample pretreatment are critical to ensuring the quality and safety of grain supplies, strengthening the monitoring for mycotoxins in grain, as well as protecting human and animal health.

Figure 1. Scheduled MRMHR workflow selectivity compared to TOF MS for FB1. Increased selectivity is demonstrated by monitoring the ion transition using MRMHR workflow compared to monitoring solely the target precursor ion and reduces the potential for false negatives or incorrect integration of matrix peaks.

Key advantages of method for mycotoxins

  • Sample pretreatment is simple and fast, only 10 minutes from sample preparation to testing
  • Fifteen common mycotoxins are included in the panel, using high-resolution MS/MS spectral library for identification without reference products
  • The use of MRMHR workflow, with retention time scheduling, optimized MS parameters, and high-resolution MS/MS reduce matrix interferences and increase method reproducibility and accuracy.
  • The method is easily applied, saving on development time and costs. 

 

Methods

Sample preparation: This sample preparation method is simple, fast, and is suitable for wheat, corn, rice, and sesame; a total of 10 grain samples were collected from different production regions. The sample preparation workflow is described in Figure 2.

LC-MS/MS analysis:  After processing, samples were injected onto an ExionLC AC system coupled to a SCIEX X500R QTOF system. Chromatographic separation was optimized to provide the best separation and ionization of the mycotoxins. Mass spectrometry analysis was performed using both positive and negative electrospray ionization. The compound information was entered using the MRMHR Method Editor by importing the method tables5 containing the precursor m/z and retention times, m/z for 5 fragment ions, as well as specific MS settings such as declustering potential (DP) and collision energy (CE). The Import and Autofill window is shown in Figure 3.

Data processing: Data was processed using SCIEX OS software. Post-acquisition extracted ion chromatograms were generated using ±10 ppm extraction window.

Figure 2. Sample preparation workflow.

Figure 3. Import and autofill MS/MS scan information. Simple interface for adding detailed compound information into targeted method.

Targeted MRMHR workflow

The X500R QTOF system can perform a targeted quantitative workflow similar to a triple quadrupole system, where the precursor ion of the target analyte is isolated and fragmented for analysis across the expected retention time of the analyte. With the MRMHR workflow, a full scan MS/MS is collected at high resolution on the fragment ions which enables high resolution extraction ion chromatograms to be generated post-acquisition, providing specificity and sensitivity.

 

Mobile phase optimization

Comparison of the influence of mobile phases with varying ratios of modifier on ionization of various mycotoxins shows that ammonium acetate causes a more obvious signal increase than ammonium formate, while a lower salt concentration (2mmol/L ammonium acetate) has a stronger effect than higher concentrations (5mmol/L and 10mmol/L). FB1 and FB2 show a strong effect on signal production when formic acid is added. Ultimately, a weak aqueous eluent containing formic acid and low salt was selected as the mobile phase A.

Target ion selection

This study optimizes the effects of adduct ions of various toxins in a detailed manner. For example, T-2 toxin has peaks for hydrogen, ammonium, and sodium adducts; given that the sodium peak is not fragile, this study compared hydrogen and ammonium adduct peaks and found that the ammonium peak signal was 30 times that of the protonated peak (Figure 5). In addition, the effects of formic acid adduct in NIV and DON were greater than those of dehydrogenation peaks (Figure 4 and 5).

Figure 4. Mobile phase optimization studies. Shown are the chromatograms of T-2 under different chromatography tests. The ammonium adduct (top) produced a greater absolute signal versus protonated precursor (bottom).

Figure 5. Chromatogram of DON. In this case, the formate adduct (top) produced greater absolute signal than dehydrogenated precursor ion (bottom).

Strong resistance to matrix interference

Sample preparation involved a simple liquid-liquid extraction and dilution, which does not remove much background sample matrix and can therefore leave the sample subject to interference effects. Four grain matrices were assessed for matrix effects based on the ratio of the areas of the target in the blank solvent and the matrix solvent. As the ratio approached 100%, matrix effects became insignificant; above 100% matrix effects were considered “enhancement”; below 100% there was matrix suppression. Matrix effects and presence of interferences were compared between using the TOF MS data and the MRMHR data. When TOF MS data is used for quantification of the 15 mycotoxins, matrix effects ranged between 43.1% and 125.6%, and when MRMHR workflow data was used, matrix effects ranged between 88.5% and 109.2%. The results showed that the MRMHR workflow is much more selective and resistant to matrix interference than primary TOF MS data. Figures 1 and 6 show that after the addition of corn matrix to fumonisin B1, there was strong interference with the TOF MS data, while the MRMHR data showed less noise and greater selectivity. Aflatoxin AFB1 is highly sensitive, and matrix inhibition effects are minimal when TOF MS data is used. Thus, this study shows that the MRMHR workflow is more selective and resistant to matrix inhibition, increasing method accuracy and reproducibility, which can in turn greatly decrease positive and false negative results.

Figure 6. Scheduled MRMHR selectivity for AFB2 compared to TOF MS; increased selectivity with MRMHR reduces the potential for false negatives or incorrect integration of different peaks.

Simultaneous quantification and qualification

At concentrations between 0.05 and 50ng/mL, there were strong linear responses for all mycotoxins (Figure 7); the correlation coefficients were greater than 0.99, fully meeting quantitative analysis requirements.

SCIEX OS software calculates ion abundance ratios in a similar manner to quadrupole devices, and different compounds have qualitative and quantitative differences in terms of ion abundance ratios. International and EU regulations (Table 1) can be used to determine the range of ratios.

When data of actual samples were analyzed, as shown in Figure 8, they were processed quickly and easily, and the interface was intuitive. The result showed a standard curve, a sample and matrix spiked ion abundance ratio, accuracy, sample concentration, and retention time. One can intuitively determine that wheat produced in a certain area contains aflatoxin B1 (AFB1) based on a sample ion abundance ratio of 0.357, a matrix spiked ion abundance ratio of 0.343, and concentration of 2.5µg/kg. This study examined 10 samples including wheat, corn, etc., from various provinces. See Table 2 for the detection results.

Figure 7. The regression calibrations of mycotoxins.

Table 1. EU regulations from the SANTE/11945/2015 document.

Figure 8. Quantification results for AFB1 in wheat as they appear in the SCIEX OS software user interface.

Table 2.  Mycotoxin content of wheat (n=6) and corn (n=4) from different regions. A blank space in the table indicates not detected.

Conclusions

This report describes the use of the X500R QTOF system to establish techniques for the detection of common mycotoxins in grain samples. The targeted MRMHR workflow allows collection on high resolution MS/MS spectra with high sensitivity, for both quantitative product ion detection (peak area) and qualitative confirmation (ion abundance ratios) at ultra-rapid scanning speeds. Because this capability is selective and resistant to matrix interference, sample preparation techniques were able to be further simplified, decreasing labor-intensive steps and improving work efficiency. Since mycotoxin reference samples are expensive and difficult to obtain, a high-resolution secondary library of mycotoxins was created. This removes the need for reference samples and permits easy identification of mycotoxin species in a sample. In summary, this method is sensitive, meets international and EU requirements for mycotoxin limits, and fully complies with customer analytical requirements. It can serve as a valuable reference in the application of high-resolution instrumentation to the detection of mycotoxins in grain samples.

This study included a total of 10 samples of wheat and corn from different production regions. Testing showed AFB1, DON, NIV, ZEN, 3-AcDON and 15-AcDON were widely distributed throughout the wheat and corn sample, illustrating the prevalence of mycotoxin contamination of grain samples and highlighting the need for greater monitoring and oversight.

 

References

  1. Hussein H S, Brasel J M. Toxicity, metabolism, and impact of mycotoxins on humans and animals. Toxicology, (2001) 167(2): 101-134.
  2. Sulyok M, Berthiller F, Kraska R, et al. Development and validation of a liquid chromatography/tandem mass spectrometric method for the determination of 39 mycotoxins in wheat and maize. Rapid Commun Mass Spectrom, (2006) 20(18): 2649-2659.
  3. Commission Regulation (EC) No 1881/2006, Setting Maximum Level for Certain Contaminants in Food Stuffs
  4. GB2761-2011, People’s Republic of China State Standard for Mycotoxin Limits in Food Products