Mechanisms of ichthyotoxicity by gymnodinioid dinoflagellates (Kareniaceae, Dinophyta) : the role of lipids, toxins and reactive oxygen species

Fragile, athecate, gymnodinioid dinoflagellates from the family Kareniaceae have long been associated with mortalities of finfish, with the principal site of damage consistently observed at the gills. The agents investigated as potentially responsible for ichthyotoxicity included lipids, particularly the polyunsaturated fatty acid (PUFA) octadecapentaenoic acid (OPA), the recently characterized polyketide phycotoxin karlotoxin (KmTx), and reactive oxygen species (ROS), measured as superoxide (0\\(_2\\)). Lipid composition of 8 species (Karenia brevis, K. milkimotoi, K. papilionacea, K. umbella; Karlodinium decipiens, Ka. veneficum; Takayama helix, T. tasmanica) revealed fatty acids dominated by the putative lipid phycotoxin OPA and docosahexaenoic acid (DHA). Sterol composition consisted mainly of novel 4-methyl ‚Äöv†vú\\(^{8(14)}\\) sterols, brevesterol and gymnodinosterol, in all species except Karenia papilionacea, which contained 4-desmethy ‚Äöv†vú\\(^{8(14)}\\) sterols. Survey of 15 species revealed only 2 closely related species, Karlodinium veneficum and Karlodinium conicum, produce karlotoxin, but with variable hemolytic activity in different strains. The hemolytic potency (HD50%) of karlotoxins was: Km\\(^{swan}\\)Tx 2-1 (65.9 ¬¨¬± 4.8 ng) and Km\\(^{Swan}\\)Tx 2-2 (63.4 ¬¨¬± 3.7 ng); Km\\(^{Huon}\\)Tx 2 (343.¬¨¬± 4.9 ng) and Km\\(_{con1cum}\\) Tx (>4000ng). Karlotoxin was shown to be algicidal towards Karenia papilionacea, reinforcing that 4-desmethyl sterol containing cells, such as cholesterol in fish gill membranes, are vulnerable. Larval finfish bioassays confirmed karlotoxin was cell bound as ichthyotoxicity increased upon cell lysis. Lysed low density Karlodinium veneficum cultures grown under phosphorus limited conditions cause,cl quicker fish mortality than those cultured in replete nutrient conditions. Pure karlotoxins were toxic to sheepshead minnow larvae, the lethal dose (LD\\(_{50}\\)) was: Km\\(^{Huon}\\)Tx 2 508.2 ng/mL; Km\\(^{Swan}\\)Tx 2-1 563.2 ng/mL, and Km\\(_{con1cum}\\) Tx 762.4 ng/mL. Purified monogalactosyl diglyceride (MGDG) lipids containing OPA exclusively, or an octatetraenoic (OTA) and docosahexaenoic (DHA) acid mixture, caused no larval finfish mortality at 120 mg/L. Superoxide production from 15 species of Kareniaceae was 10 times less than those produced by the well-known ichthyotoxic raphidophyte Chattonella marina. The singular or synergistic application of lipids, superoxide and sub-lethal karlotoxin revealed no ichthyotoxicity with any paired combination, except when lipids were combined with karlotoxin. Except for the demonstrated role of karlotoxin in Karlodinium veneficum and Ka. conicum and brevetoxin in Karenia brevis, the fish-killing mechanism by other gymnodinioid dinoflagellates remains largely unresolved. The purported toxin in Karenia mikimotoi, gymnocin, was not detected. Reactive oxygen species in Kareniaceae were quantitatively less important than in fish-killing raphidophytes and bioassays using naturally occurring forms of OPA and OTA indicated that PUFA were non-toxic to sheepshead minnow larvae.