FACT SHEET - www.micotoxinas.com.br
Leatherhead Food Research Association, Randalls Road, Leatherhead, Surrey KT22 7RY, England
Most trichothecenes have a double bond at position C-9,10, a 12,13-epoxide ring and a number of hydroxyl and acetoxy groups. T-2 and HT-2 belong to the type A trichothecenes with an oxygen function different from a carbonyl function at the C-8 position. The type B trichothecenes possess a carbonyl function at this position.
Fig. 1: Structural formula of type A trichothecenes:
T-2: (R1 = OAc)
HT-2 (R1 = OH)
Fig. 2: Structural formula of type B trichothecenes:
DON (R1 = OH, R2 = H, R3 = OH, R4 = OH)
NIV (R1 = OH, R2 = OH, R3 = OH, R4 = OH)
3-AcDON (R1 = OAc, R2 = H, R3 = OH, R4 = OH)
15-AcDON (R1 = OH, R2 = H, R3 = OAc, R4 = OH)
FUS-X (R1 = OH, R2 = OAc, R3 = OH, R4 = OH)
At the 56th meeting of the Joint FAO/WHO Expert Committee on Food Additives (Geneva, 6-15 February 2001) the state of the art in DON, T2 and HT2 analysis was evaluated. Official methods and other validated methods have been developed for the analysis of deoxynivalenol in cereals and foodstuffs. No official methods for the determination of T-2 and HT-2 toxins have been published, although some methods with intralaboratory validation have been reported. Interlaboratory comparison studies clearly showed that improvements are needed for deoxynivalenol and T-2 and HT-2 toxin analysis with respect to the recovery, accuracy, and precision of the measurements. Wider availability of reference materials and international comparative studies are needed to ensure improved quality assurance.
Analytical procedures differ in extraction, clean-up and determination steps, based on which of the two groups of trichothecenes are to be analysed.
Reviews of HPLC and GC methods were given by Scott et al.  and Langseth & Rundberget . Other reviews of mycotoxins, including trichothecenes, were given by Cole  and Chu , amongst others, and in the World Health Report Environmental Health Criteria 105 of the WHO Immunochemical methods were reviewed by Morgan  and Park et al. .
Type A Trichothecenes
Immunoassays are the only screening methods available for T-2 and HT-2 for routine analysis of cereals. Extraction is usually performed with acetonitrile/water, methanol or chloroform/ethanol. TLC and ELISAs are the only reliable means available for T-2- and HT-2 detection. The detection limits of the assays can be as low as 0.2 to 50 ng/g for T-2.
Different methods of analysis are employed for both types of trichothecenes. HPLC with UV detection is not applicable to type A trichothecenes (as they lack a keto group at the C-8 position).
Combinations of solvents, usually acetonitrile/water and methanol/water, have been used for the extraction of grain, food, and feeds. Extraction is performed by high-speed blending or mechanical shaking. For subsequent clean-up, prepacked silica gel, florisil, cyano and C18 SPE-cartridges are used. Multifunctional MycoSepÔ columns are becoming more and more established.
GC analysis is the preferred method for the determination of type A trichothecenes. Most methods are based on trimethylsilylation or fluoroacylation to increase volatility and sensitivity. Mass spectrometric (MS) detection or the use of an electron capture detector (ECD) with a fluoroacylation reagent is recommended. The formation of fluoroacyl derivatives by trifluoroacetic anhydride (TFA), pentafluoropropionyl (PFP) or heptafluorobutyryl (HFB) derivatives are also applied in order to increase the sensitivity of the ECD.
Typical detection limits of quantitative methods for determination of T-2 and HT-2 in cereals are 3 ng/g (T-2, LC-MS), 1 ng/g (HT-2, LC-MS), 10 ng/g (T-2 and HT-2, GC-MS) and 10 ng/g (T-2 and HT-2, GC-ECD). Typical recoveries range from 70-120%.
Type B Trichothecenes
Methods based on thin-layer chromatography (TLC) are still common and, with the introduction of high-performance TLC (HPTLC) and scanning instruments, separation efficiency and precision have increased. Reagents (e.g. sulfuric acid or para-anisaldehyde) are necessary to visualise the only short wavelength absorbing DON. Other spray reagents include 4-para-nitrobenzyl-pyridine or nicotinamide in combination with 2-acetyl-pyridine) or AlCl3, which is the most useful visualisation reagent for DON. Typical detection limits by TLC are in the range of 20-300 ng/g.
In addition to TLC, enzyme-linked immunosorbent assays (ELISAs) have recently been developed for rapid screening purposes. For less complex matrices (pure cereals), these methods can be used for quantification. Clean-up (if any) is minimal. The most sensitive ELISA methods have been developed for tri-acetylated DON, with an LOD of 0.3-1 ng/g (requiring acetylation and giving the sum of DON and its acetylated derivatives as a result). In addition, ELISAs have been developed, which enable direct determination of DON with less sensitivity (LOD=20-300 ng/g). Accurate quantification by ELISAs is often limited, owing to cross reactivities.
Extraction of DON and other type B trichothecenes is performed by shaking or blending with aqueous acetonitrile. Also other solvents, such as water/PEG, and chloroform/methanol. Trenholm assessed extraction procedures for DON and could demonstrate that longer extraction times are required for naturally contaminated samples than for spiked ones. Extraction efficiency should therefore be evaluated with naturally contaminated samples.
Clean-up is usually not required for immunoassays, while physicochemical methods employ extensive clean-up procedures. The main procedures are liquid-liquid partitioning, solid phase extraction (SPE), column chromatography, immunoaffinity columns (IAC) and multifunctional clean-up columns. Interfering lipids can be removed with n-hexane.
The most frequently employed column packing material in column chromatography is a mixture of charcoal/alumina/celite. Modern SPE columns are an alternative to conventional column chromatographic methods.
The application of IACs for purification has already been a success for several mycotoxins. However, commercial IACs are not available for most trichothecenes. The analyte molecules are bound to antibodies and, subsequent to a washing step, the toxins can be eluted. Although low recoveries were obtained for DON, a new HPLC method employing both charcoal/alumina/celite columns and a commercial IAC gave reasonable recoveries and a LOQ of 50 ng/g.
Another recent development in clean-up methods is the widely used commercial MycoSepÔ column (#225 for trichothecenes). It allows quick sample purification (30 seconds), and no rinsing steps are required. Analytical interfering substances are retained while trichothecenes are not adsorbed on the packing material.
Separation and detection techniques
Gas chromatography (GC) is carried out with electron capture detection (ECD), mass spectrometric (MS), or tandem mass spectrometric detection (MS/MS) after derivatisation. The choice of derivatisation reagent depends on the type of trichothecene and the method of detection. Derivatisation mixtures such as TRI-SIL TBT® and Sylon BTZ are used, which contain TMSI (40 ± 5 %), BSA (35 ± 5 %) and TMCS (25 ± 5 %). This is to avoid two peaks caused by incomplete derivatisation.
Several HPLC methods have been published for the determination in food and cereals. Separation is usually achieved on a C18 reversed-phase column with methanol/water mixtures as mobile phase. The use of acetonitrile instead of methanol in aqueous mixtures seem to be the better choice. However, time-consuming clean-up is required for the determination of DON in complex matrices (food and feed). The German VDLUFA standard method for DON should be mentioned here, which is being validated. The proposed HPLC method includes a time-consuming three-step clean-up including liquid/liquid partitioning with petrol ether, SPE with charcoal/alumina/celite mixture and an immunoaffinity column.
Analytical performance characteristics comparable to GC methods can be achieved with HPLC methods and pre- or post-column derivatisation. Post-column derivatisation, which has been developed by Sano, involves alkaline decomposition of DON and NIV to generate formaldehyde and reaction with acetoacetate and ammonium acetate to form a fluorescent derivative. Recently, LC-MS instruments, particularly using APCI (atmospheric pressure chemical ionisation) interfaces, have been employed for the determination and identification of trichothecenes at trace levels. Electrochemical detection has also been described for the determination of DON.
Results from an intercomparison study reflect the current state-of-the-art techniques. For purification of DON from the raw extracts SPE techniques were employed (MycoSepTM columns or florisil/active charcoal columns). Additionally IACs and the Extrelutâ technique in combination with SPE florisil was applied. Determination was carried out by GC-ECD, GC-MS, HPLC-UV/DAD and HPLC-FLD. Furthermore, ELISAs were also employed. Generally, the LODs of the DON methods employed ranged from 0.3 to 107.9 ng/g. Individually determined recoveries varied from 60 to 100% and the reported CVs ranged from 3 to 15%.
Typical detection limits for DON in cereals are 100-1600 ng/g (HPLC-UV), 6-40 ng/g (HPLC-MS), 20 ng/g (HPLC-FLD), 20-50 ng/g and even lower (GC-ECD) and down to approximately 5 ng/g (GC-MS with HFB and PFP derivatives), with typical recoveries ranging from 70-110%.
 Scott, P.M. (1993) In Betina, V. (ed): Chromatography of mycotoxins - techniques and applications, Elsevier, Amsterdam, pp373-425.
 Langseth, W., & Rundberget, T. (1998) Instrumental methods for determination of nonmacrocyclic trichothecenes in cereals, foodstuff and cultures. J. Chromatography A, 815, 103-121.
 Cole, R.J. (1986) Modern methods in the analysis and structural elucidation of mycotoxins. Academic press, Orlando
 Chu, F.S. (1991) In: Sharma, R.P. & Salunkhe, D.K. (eds): Mycotoxins and phytoalexins, CRC press, Boca Raton), pp33-79.
 Morgan, M.R.A. (1989) Mycotoxin immunoassays: with special reference to ELISAs. Tetrahedron, 45, 2237-2249.
 Park, J.Ch., Zong, M.S., & Chang, I.-M. (1991) Survey of the presence of the fusarium mycotoxins nivalenol, deoxynivalenol and T-2 toxin in Korean cereals of the 1989 harvest. Food Additives and Contaminants, 8(4), 447-451.
 Trenholm, H.L., Warner, R.M., & Prelusky, D.B. (1985) Assessment of extraction procedures in the analysis of naturally contaminated grain products deoxynivalenol (vomitoxin). J.Assoc.Off.Anal.Chem., 68(4), 645-649.