Marine Drugs (Feb 2022)
Quantitation Overcoming Matrix Effects of Lipophilic Toxins in <i>Mytilus galloprovincialis</i> by Liquid Chromatography-Full Scan High Resolution Mass Spectrometry Analysis (LC-HR-MS)
Abstract
The analysis of marine lipophilic toxins in shellfish products still represents a challenging task due to the complexity and diversity of the sample matrix. Liquid chromatography coupled with mass spectrometry (LC-MS) is the technique of choice for accurate quantitative measurements in complex samples. By combining unambiguous identification with the high selectivity of tandem MS, it provides the required high sensitivity and specificity. However, LC-MS is prone to matrix effects (ME) that need to be evaluated during the development and validation of methods. Furthermore, the large sample-to-sample variability, even between samples of the same species and geographic origin, needs a procedure to evaluate and control ME continuously. Here, we analyzed the toxins okadaic acid (OA), dinophysistoxins (DTX-1 and DTX-2), pectenotoxin (PTX-2), yessotoxin (YTX) and azaspiracid-1 (AZA-1). Samples were mussels (Mytilus galloprovincialis), both fresh and processed, and a toxin-free mussel reference material. We developed an accurate mass-extracted ion chromatogram (AM-XIC) based quantitation method using an Orbitrap instrument, evaluated the ME for different types and extracts of mussel samples, characterized the main compounds co-eluting with the targeted molecules and quantified toxins in samples by following a standard addition method (SAM). An AM-XIC based quantitation of lipophilic toxins in mussel samples using high resolution and accuracy full scan profiles (LC-HR-MS) is a good alternative to multi reaction monitoring (MRM) for instruments with HR capabilities. ME depend on the starting sample matrix and the sample preparation. ME are particularly strong for OA and related toxins, showing values below 50% for fresh mussel samples. Results for other toxins (AZA-1, YTX and PTX-2) are between 75% and 110%. ME in unknown matrices can be evaluated by comparing their full scan LC-HR-MS profiles with those of known samples with known ME. ME can be corrected by following SAM with AM-XIC quantitation if necessary.
Keywords