Generic Methodologies Used In Cyanotoxin Analysis
Sample Extraction for Analysis
Most cyanotoxins are retained in the cell-bound fraction of samples taken from waterbodies (Chapter 2). Only when cyanobacterial populations experience cell lysis, for example, in surface scums exposed to high temperatures and light intensities, substantial amounts of toxins are released to the surrounding water. An exception is cylindrospermopsin that can be released from viable cell and found in large proportions in the cell-free fraction (section 2.2 Box 5.1 and Chapter 10). Further, extraction and sample clean-up is of critical importance for the analysis of cyanotoxins in foods to avoid both under- and overestimating concentrations (see section 5.3.4).
To make cell-bound cyanotoxins accessible to chemical analysis, they need to be extracted from the cells with an appropriate solvent. The solvent needs to be selected to efficiently extract toxins from cells in a few (maximum of three) extraction cycles and needs to be compatible with downstream analytical methods. Extraction procedures for individual classes of toxins will be discussed in the respective sections.
There is no single method available that is suitable for extracting all classes of cyanotoxins; hence, it may be necessary to collect several samples or subdivide samples prior to processing (see Chapter 12).
Solid-Phase Extraction (SPE)
While some immunoassay (ELISA) methods and LC-MS/MS (see below) may be sufficiently sensitive for monitoring in compliance with guideline values, trace analysis of cyanotoxins in water (i.e., concentrations below 1 pg/L) typically requires sample concentration using solid-phase extraction (SPE). This processing involves passing a known volume (typically 100- 500 mL) of a (particle-free) water sample through a solid-phase cartridge to concentrate dissolved cyanotoxins. SPE requires a vacuum manifold system, PTFE (polytetrafluorethylen) connectors and tubing (for a minimum adsorption of analytes) and a vacuum pump. The equipment for SPE can be used for different classes of cyanotoxin; however, specific cartridges need to be used which retain the cyanotoxin of interest. A large selection of ready-to-use SPE cartridges is available with different sorbent materials and varying sorbent volumes. Sorbent materials primarily differ in the degree of hydrophobicity and have to be selected in correspondence with the cyanotoxin of interest, for example, C18 for microcystins or graphitised carbon for highly hydrophilic cylindrospermopsins.
Filtering water prior to SPE is essential to minimise clogging of the cartridges. Nonetheless, during times of blooms, the higher load of very small particles may significantly increase loading times. In this case, it is advised to note the volume of sample that has already passed and discontinue the sample loading. Once the water sample has passed through the cartridge, a washing step follows before the toxin adsorbed on the cartridge is eluted in a small volume (typically 3-5 times the sorbent bed volume) of solvent into a collection tube (Figure 14.2). This can be analysed directly or further concentrated by drying to enhance the detectability. It is tempting to load as large a volume of water sample as possible to allow the detection of low levels of cyanotoxin; however, this is not always appropriate. Larger volumes can take many hours to load onto the cartridge and matrix contaminants are also being concentrated; hence, larger sample volumes increase matrix interference. Loading samples onto SPE cartridges can be time-consuming so it is wise to determine the most appropriate volume of water. This can be done by processing some test samples of spiking water (preferably similar to the samples) and determining how long the process takes. Typically, a volume is chosen that takes no more than 3 h to pass through the SPE cartridge, allowing time for sample preparation, elution and processing before analysis.
It is useful to validate methods by spiking known amounts of the cyanotoxin of interest into water samples. Spiking should not be performed with high-purity water (e.g., Milli-Q) as this may lead to poor recoveries, and as this is in no way representative of the samples being processed. Using the typical raw or tap water to be tested is best (if tap water is chlorinated, using a chlorine-quenching agent, i.e., sodium thiosulphate or ascorbic acid) as this will provide a good indication of expected performance.