Mycotoxins in grain

This application note presents a validated LC-MS/MS method to detect 9 Fusarium toxins and Ochratoxin A in diluted crude extracts of grain. The developed sample preparation procedure is quick, easy, robust, and inexpensive. The LC-MS/MS method in Multiple Reaction Monitoring (MRM) detects all compounds in a single run with Limits of Quantitation (LOQ) between 0.3 μg/kg and 10 μg/kg. The complete method was validated for the analysis of wheat, rye, barley, and oat samples.

Mycotoxins are known to harm the health of humans and animals. They are known either as carcinogenic or cytotoxic and impair the immune system. Therefore, different countries have set regulations on mycotoxins. In the EU, mycotoxin limits are harmonized in the regulation for contaminants in foodstuffs (EC 1881/2006 of December 19, 2006) and the amended regulation (EC 1126/2007 of September 28, 2007). The focus of the legislation and the European monitoring program is on Fusarium toxins like Deoxynivalenol, Zearalenon, HT-2 and T-2 toxins because of their frequent and increasing occurrence in grain. In addition, the European Commission has decided to set maximum levels for T-2 and HT-2 toxin by July 1, 2008. The expected limit for the sum of both Fusarium toxins could be 50 μg/kg or less.^1-3 

Cereals and grains are often contaminated with Ochratoxin A and Fusarium toxins. Fusarium toxins can be found in all types of grains such as wheat, rye, maize, barley (malt), and oats. The contamination is dependent on climate conditions during growth, harvest, and storage. Because of bad weather conditions in 2007 high contaminations of Deoxynivalenol (DON) were found in wheat and T-2 and HT-2 toxin in wheat and oats.

Due to this, it is necessary to have a reliable, sensitive, robust, and fast method to analyze a high number of mycotoxins in grain. Methods used so far have not shown the required sensitivity. Thus necessary clean up steps with immunoaffinity columns have resulted in time consuming and expensive methods.

A method for the detection of 9 Fusarium toxins: DON, Zearalenon (ZON), 3-Acetyldeoxynivalenol (3- AcDON), 15- Acetyldeoxynivalenol (15-AcDON), HT-2, T-2, Fusarenon X (FUS X), Nivalenol (NIV), Diacetoxyscirpenol (DAS) and Ochratoxin A (OTA) was developed (Figure 1). Diluted crude extract were analyzed using Liquid Chromatography and tandem Mass Spectrometry (LC-MS/MS) in a single run on an API 4000^TMLC/MS/MS system. No time consuming and possibly expensive sample preparation is needed. The method was validated for wheat, rye, barley and oat and applied for the analysis approximately 220 grain samples. The LOQ vary between 0.3 μg/kg and 10 μg/kg depending on the compound.

 

Method Details

Sample Preparation:

1) Mill and homogenize 10 g of grain sample.
2) Add 40 mL of acetonitrile/water (84/16).
3) Extract by mixing for 90 min (220 rpm).
4) Filter throughWhatman S&S 1573 ½.
5) Dilute filtrate 1:10 with water + 5mM ammonium acetate.
6) Inject 100 μL into LC/MS/MS.

HPLC Conditions:

A Shimadzu Prominence LC system consisting of system controller, two pumps, degasser, autosampler, and column oven was used. Separation was performed on an Agilent ZORBAX Eclipse XDB C18, 100x4.6 mm (1.8 μm) column. The column oven temperature was set to 40°C. A gradient of eluent A: water + 5 mM ammonium acetate and eluent B: methanol + 5 mM ammonium acetate was used at a flow rate of 500 μL/min. Details of the gradient are given in Table 1. The injection volume was set to 100 μL.  

MS/MS Conditions: 

An API 4000^TMLC/MS/MS system equipped with Turbo V^TM source and Electrospray Ionization (ESI) probe was used. The method contained three periods with alternating polarities (0.0-7.6 min negative; 7.6-9.2 min positive; 9.2-16.0 min negative).

The following ion source parameters were used: Temperature 600°C, Curtain Gas 25 psi, Gas1 50 psi, Gas2 70 psi, CAD Gas 6 (positive) or 10 (negative), and IS voltage +5000 V or -4000 V, respectively. All mycotoxins were detected using two MRM transitions in the following order: first period NIV, DON, FUS X, AcDON; second period DAS, OTA, HT-2, T-2; and third period ZON. Since 3-AcDON and 15- AcDON were not separated chromatographically, only compound specific transitions were used for detection. The used MRM transitions are listed in Table 2 and product ion spectra of 3-AcDON and 15-AcDON are shown in Figure 2.

 

Figure 3 shows a standard chromatogram of 9 Fusarium toxins at 50 μg/kg and Ochratoxin A at 10 μg/kg. During the evaluation it was shown that the sensitivity of the MRM transitions depends on the quality of the used solvents as well as on the analyzed matrices.

AcDON, DON, FUS X and NIV generally show good sensitivity for both the [M+CH₃COO]^- and the [M-H]^-. However, in matrix samples the in Table 2 listed MRM transitions were used for better S/N, reproducibility, and recovery. A special characteristic of ionization was found for HT-2. The sensitivity of the MRMs either [M+NH4]+ or [M+Na]+ differ in accordance to the analyzed matrices (Figure 4).

A special characteristic of ionization was found for HT-2. The sensitivity of the MRMs either [M+NH4]⁺ or [M+Na]⁺ differ in accordance to the analyzed matrices (Figure 4).

The LOQ were validated as 0.3 μg/kg for OTA, 5 μg/kg for HT-2, T-2 and ZON, and as 10μg/kg for AcDON, DON, FUS X, DAS and NIV. 

The injection volume of 100 μL with a ten times diluted sample showed much better LOQ than 25 μL of a direct injected or 50 μL of a 1/5 diluted sample. Responsible for this finding are the specific initial chromatographic conditions needed for NIV. The acetonitrile in the sample, at the end of the extraction procedure, caused peak broadening for NIV. This could only be eliminated by diluting the sample in 100% water +5 mM ammonium acetate by a factor of 10.

Footnotes to Table 3:

# EC 1881/2006 and the amended EC 1126/2007
* Unprocessed durum wheat and oats
** Unprocessed cereals other than durum wheat and oats
*** Unprocessed cereals
(1) Due to co-occurrences and as "generally low" considered levels no MRL was estimated
(2) Appropriateness of setting a maximum level should be considered by 1 July 2008

The calibration curves of all compounds were linear ranges differ highly. LOQ and upper ends of the linear dynamic range of all detected mycotoxins are shown in Table 3.

The recoveries were determined for ach mycotoxin in each matrix compared to the calibration curves without matrix (Table 4-6).

It was shown that the solutions of the extracted grains are stable over 36 hours under cool conditions (4°C). 

A large carryover of OTA in the injection port was observed when injecting high standard concentrations, thus solvent blanks were injected after standard injections. 

With the here presented validated method about 220 grain samples of the new harvest have been analyzed since July 2007.

The presented data are based on the European Grain Monitoring Program (EGM) and selected data are shown in Tables 4 to 6.

The developed method is appropriate for the analysis of 9 Fusarium toxins and OTA in one single LC-MS/MS run without time consuming sample preparation/enrichment. The LOQ were found at 0.3 μg/kg for OTA, 5 μg/kg for HT-2, T-2 and ZON and
at 10 μg/kg for AcDON, DON, FUS X, DAS and NIV and meet the National and European law required detection limits. Recoveries were determined in the range of 21 to 100%.

References

1 EC 1881/2006 of December 19, 2006
2 Amended regulation EC 1126/2007 of September 28, 2007
3 http://www.mykotoxin.de/Gesetzgebung.htm