Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence
1.4 CHEMICAL SCIENCES
The most widely used reference method in Europe for the detection and monitoring of lipophilic marine toxins is the Mouse Bioassay (MBA) as first described by Yasumoto et al., (1978). The MBA offers a good level of protection to human health and is capable of detecting the overall toxicity of previously known toxins. However, there are drawbacks associated with the use of the MBA as it is both expensive and time consuming and can give false positives due to interferences (Suzuki et al, 1976). The problems associated with the original mouse bioassay have led to several modifications (Yasumoto et al, 1984: Lee et al, 1987). Recent EU Regulations 853/2004, 854/2004, 2074/2005 and 1664/2006 (Anon 2004a; Anon2004b; Anon 2005a and Anon 2006a) set out details of which toxins should be monitored and the corresponding regulatory limits and methods to be used. They also permit the use of alternative methods, provided they are fully validated and can offer at least an equivalent level of protection for human health. LCMS is emerging as one of the most promising analytical methods available for the analysis of marine toxins. However, none of the available LCMS based methods are fully validated for all of the regulated toxin groups (OA, DTXs, YTX, PTX and AZAs) and therefore the replacement of the MBA as a reference method is not yet feasible (Hess et al, 2006) This study focuses on particular aspects in the development of LCMS methodology and was carried out as part of the EU funded project called Biotox which was brought about to develop and validate alternative methods to the MBA. Several aspects of the LC method were examined including, the column type and gradient elution conditions. The columns put forward were examined using: resolution between components of the mixture and the theoretical plate model of chromatography (plate number (N) and plate height (H)). The BDS Hypersil C8 emerged as the column to be advised for the majority of the lipophilic toxins included in the regulations (OA, DTXs, DTX2, PTX2,AZA1, AZA2 and AZA3. An additional LC method was developed using a basic mobile phase to include the detection of YTX. A study examining the MS conditions (ionisation mode, acquisition mode and number of transitions) showed that the choice of MS conditions plays a significant role in the results obtained. It was found that the analysis of the OA toxin group in negative ionisation mode gave more accurate results. The choice of acquisition mode for OA was not found to cause a significant variability in results. For AZAs and PTX2 (analysed in positive ionisation mode only) the choice of acquisition mode was important; parent ion monitoring was shown to give the most variable results compared to single and double transition monitoring. Oyster and Scallop tissue were used to examine trends of matrix effects in shellfish extracts. Similar trends were found between two different MS detectors (Quadrupole Time of Flight and Triple Stage Quadruple) of the same manufacturer and equipped with identical ionisation sources. Ion suppression effects were observed for AZA1 (up to 15%) and ion enhancement effects were observed for OA (ranging form 0 to 40%) and PTX2 (ranging form 45 to 100%) Two sample clean-up schemes, Liquid-liquid extraction (LLE) and Solid Phase Extraction (SPE) were investigated, with a view to removing or at least minimising matrix effects. A prerequisite for each of the sample clean-up schemes was good recovery of all toxins using one procedure. The LLE procedure used a hexane extraction to remove any fats from the extract, and a dichloromethane (DCM) partitioning step to isolate the toxins. Recovery losses were incurred with additional partitioning steps; further losses were attributed to the evaporation/reconstitution step. Using the optimised conditions LLE recoveries of approximately 80% were obtained for OA and AZA1. An array of different SPE sorbent phases were evaluated: Oasis HLB tm(Waters), Strata SDB-L and Strata X (Phenomex), Isolute Env + (Biotage/IST) and Bond Elut LRC Certify (Varian). The load, wash and elution steps were optimised. The Strata-X and the Oasis HLB tm cartridges (co-polymer sorbents) gave the best recoveries for OA, AZA1, PTX2 and YTX respectively. Both clean-up schemes were evaluated for their effectiveness in the removal of matrix effects using oyster, mussel and scallop extracts. In the LLE study matrix effects were observed for OA (ion enhancement, 14%) and AZA1 (ion suppressions, 36%) in the crude extract. LLE demonstrated a clean up effect for AZA1 only. A small-scale study between three laboratories highlighted the difficulties for evaluating the clean-up effect for SPE. At the Marine Institute (MI) the matrix effects arising form the extracts were variable. No matrix effects were observed for OA or AZA1 whereas ion enhancements of 70% was observed for PTX2. Substantially different degrees of matrix effects were reported between laboratories (using the same tissues and spiking standards) and only one laboratory reported a significant clean-up effect using SPE. From the results obtained during the course of this thesis LLE and SPE have potential in removing matrix effects under certain conditions.
McElhinney, M. (2007). Matrix effects, development of clean-up and LC-techniques contributing towards a reference LCMS method for the analysis of lipophilic marine toxins. Technological University Dublin. doi:10.21427/D76K7K