Profiling the fatty acids from a strain of the microalgae Alexandrium tamarense by means of high-speed counter-current chromatographyand gas chromatography coupled with mass spectrometry

tFatty acids of microalgae have been studied as potential chemotaxonomic markers, to reveal plausi-ble lipid phycotoxins or in the context of mass production of algal biofuels. The planctonic microalgaeAlexandrium tamarense (Dinophyceae) is a common harmful algal bloom species that often proliferatesin eutrophic costal waters. Alexandrium blooms are the proximal source of toxins associated with par-alytic shellfish poisoning (PSP), a neurological affliction that has caused human illness for centuries viaconsumption of contaminated shellfish. However, data on the fatty acid composition of A. tamarense iscurrently limited. For this reason, we cultivated a well-defined strain of A. tamarense (Alex2, group I, NorthAmerican clade) in order to study both its major and minor fatty acids. The harvested microalgae weretransesterified and the fatty acid methyl esters were fractionated by means of high-speed counter-currentchromatography (HSCCC). The resulting 31 HSCCC fractions were analyzed by gas chromatography withmass spectrometry (GC/MS). Unknown substances were identified by transferring assorted HSCCC frac-tions into picolinyl or pyrrolidide derivatives. Twenty fatty acids (range 0.2–22.9% contribution to totalfatty acids) were identified in the unfractionated sample with 14:0, 16:0, 18:1n-9, 18:4n-3, 18:5n-3 and22:6n-3 representing > 80% of the total fatty acids. HSCCC fractionation enabled the identification of fur-ther 22 trace fatty acids contributing between ∼0.01 and 0.2% to total fatty acids. The fatty acids includedseveral branched-chain fatty acids as well as scarcely reported fatty acids like 11-methyl-18:1n-6tr or18:2�4,9. In order to enable a better comparability and repeatability of HSCCC fractionations, we calcu-lated for each HSCCC fraction the total volume of mobile phase, which had passed the HSCCC. From thisvolume we subtracted the volume of extruded stationary phase and divided the corrected volume by thetotal coil volume. These elution values were in good agreement with the partition ratios of randomlychosen fatty acid methyl esters obtained in shake flask tests, which allows the prediction of the elutionfrom the HSCCC system when the partition ratio is known