RESUMO
A strategy to construct multivariate biomarkers for exposure to algal neurotoxins via correlating changes to the profiles of a series of neurotransmitters and their metabolites in the central nervous system (CNS) of exposed test organism is reported. 3-Month-old marine medaka (Oryzais melastigma) were exposed to waterborne brevetoxin PbTx-1 at two sub-lethal dose levels (0.5 and 2.5⯵g-PbTx-1â¯L-1) for a duration of 12â¯h before quantification of 43 selected neurotransmitters and metabolites in their CNS were measured via dansyl chloride derivatization and LC-MS/MS determination. The profiling data were analyzed by multivariate statistical analyses, including principle component analysis (PCA), projection on latent structure-discriminate analysis (PLS-DA) and orthogonal projection on latent structure-discriminate analysis (OPLS-DA). Neurotransmitters and metabolites related to activation of voltage-gated sodium channels (VGSCs), N-methyl-D-aspartic acid receptors (NMDARs) and cholinergic neurotransmission were found to contribute significantly to class separation in the corresponding OPLS-DA models. Those models obtained from different exposure dosages were correlated by the Shared and Unique Structures Plot (SUS-plot) to identify appropriate variables for the construction of exposure biomarkers in the form of multivariate predictive scores. The established biomarkers for male and female medaka fish were able to predict acute sub-lethal exposure to PbTx-1 with good sensitivity and specificity (male fish: sensitivity 94.7%, specificity 80.0%; female fish: sensitivity 91.4%, specificity 83.3%). Neurotransmitter profiles in the CNS of medaka fish that should have recovered from exposure to PbTx-1 have also been determined to reveal long-term impacts to the CNS of the affected organism even after the exposure has been interrupted.
