Cytotoxicity of Prymnesium parvum extracts and prymnesin analogs on epithelial fish gill cells RTgill-W1 and the human colon cell line HCEC-1CT.

Harmful algal blooms kill fish populations worldwide, as exemplified by the haptophyte microalga Prymnesium parvum. The suspected causative agents are prymnesins, categorized as A-, B-, and C-types based on backbone carbon atoms. Impacts of P. parvum extracts and purified prymnesins were tested on the epithelial rainbow trout fish gill cell line RTgill-W1 and on the human colon epithelial cells HCEC-1CT. Cytotoxic potencies ranked A > C > B-type with concentrations spanning from low (A- and C-type) to middle (B-type) nM ranges. Although RTgill-W1 cells were about twofold more sensitive than HCEC-1CT, the cytotoxicity of prymnesins is not limited to fish gills. Both cell lines responded rapidly to prymnesins; with EC50 values for B-types in RTgill-W1 cells of 110 ± 11 nM and 41.5 ± 0.6 nM after incubations times of 3 and 24 h. Results of fluorescence imaging and measured lytic effects suggest plasma membrane interactions. Postulating an osmotic imbalance as mechanisms of toxicity, incubations with prymnesins in media lacking either Cl-, Na+, or Ca2+ were performed. Cl- removal reduced morphometric rearrangements observed in RTgill-W1 and cytotoxicity in HCEC-1CT cells. Ca2+-free medium in RTgill-W1 cells exacerbated effects on the cell nuclei. Prymnesin composition of different P. parvum strains showed that analog composition within one type scarcely influenced the cytotoxic potential, while analog type potentially dictate potency. Overall, A-type prymnesins were the most potent ones in both cell lines followed by the C-types, and lastly B-types. Disturbance of Ca2+ and Cl- ionoregulation may be integral to prymnesin toxicity.

How relevant are sterols in the mode of action of prymnesins?

Prymnesins, produced by the haptophyte Prymnesium parvum, are considered responsible for fish kills when this species blooms. Although their toxic mechanism is not fully understood, membrane disruptive properties have been ascribed to A-type prymnesins. Currently it is suggested that pore-formation is the underlying cause of cell disruption. Here the hypothesis that A-, B-, and C-type prymnesins interact with sterols in order to create pores was tested. Prymnesin mixtures containing various analogs of the same type were applied in hemolysis and cytotoxicity assays using Atlantic salmon Salmo salar erythrocytes or rainbow trout RTgill-W1 cells. The hemolytic potency of the prymnesin types reflected their cytotoxic potential, with approximate concentrations reaching 50 % hemolysis (HC50) of 4 nM (A-type), 54 nM (C-type), and 600 nM (B-type). Variabilities in prymnesin profiles were shown to influence potency. Prymnesin-A (3 Cl) + 2 pentose + hexose was likely responsible for the strong toxicity of A-type samples. Co-incubation with cholesterol and epi-cholesterol pre-hemolysis reduced the potential by about 50 % irrespective of sterol concentration, suggesting interactions with sterols. However, this effect was not observed in RTgill-W1 toxicity. Treatment of RTgill-W1 cells with 10 µM lovastatin or 10 µM methyl-ß-cyclodextrin-cholesterol modified cholesterol levels by 20-30 %. Regardless, prymnesin cytotoxicity remained unaltered in the modified cells. SPR data showed that B-type prymnesins likely bound with a single exponential decay while A-types seemed to have a more complex binding. Overall, interaction with cholesterol appeared to play only a partial role in the cytotoxic mechanism of pore-formation. It is suggested that prymnesins initially interact with cholesterol and stabilize pores through a subsequent, still unknown mechanism possibly including other membrane lipids or proteins.

Human-Relevant Sensitivity of iPSC-Derived Human Motor Neurons to BoNT/A1 and B1.

The application of botulinum neurotoxins (BoNTs) for medical treatments necessitates a potency quantification of these lethal bacterial toxins, resulting in the use of a large number of test animals. Available alternative methods are limited in their relevance, as they are based on rodent cells or neuroblastoma cell lines or applicable for single toxin serotypes only. Here, human motor neurons (MNs), which are the physiological target of BoNTs, were generated from induced pluripotent stem cells (iPSCs) and compared to the neuroblastoma cell line SiMa, which is often used in cell-based assays for BoNT potency determination. In comparison with the mouse bioassay, human MNs exhibit a superior sensitivity to the BoNT serotypes A1 and B1 at levels that are reflective of human sensitivity. SiMa cells were able to detect BoNT/A1, but with much lower sensitivity than human MNs and appear unsuitable to detect any BoNT/B1 activity. The MNs used for these experiments were generated according to three differentiation protocols, which resulted in distinct sensitivity levels. Molecular parameters such as receptor protein concentration and electrical activity of the MNs were analyzed, but are not predictive for BoNT sensitivity. These results show that human MNs from several sources should be considered in BoNT testing and that human MNs are a physiologically relevant model, which could be used to optimize current BoNT potency testing.