Acute toxicity of dihydroanatoxin-a from Microcoleus autumnalis in comparison to anatoxin-a

Abstract The cyanobacterium Microcoleus autumnalis grows as thick benthic mats in rivers and is becoming increasingly prevalent around the world. M. autumnalis can produce high concentrations of anatoxins and ingestion of benthic mats has led to multiple dog deaths over the past two decades. M. autumnalis produces a suite of different anatoxin congeners including anatoxin-a (ATX), dihydroanatoxin-a, (dhATX), homoanatoxin-a and dihydrohomoanatoxin-a. Benthic mat samples often contain high levels of dhATX, but there is little toxicology information on this congener. In the present study, natural versions of dhATX and ATX were purified from cyanobacteria to determine the acute toxicity by different routes of administration using mice. Nuclear magnetic resonance spectroscopy was used to confirm the putative structure of dhATX. By intraperitoneal (ip) injection, the median lethal dose (LD50) for dhATX was 0.73 mg/kg, indicating a reduced toxicity compared to ATX (LD50 of 0.23 mg/kg). However, by oral administration (both gavage and feeding), dhATX was more toxic than ATX (gavage LD50 of 2.5 mg/kg for dhATX and 10.6 mg/kg for ATX; feeding LD50 of 8 mg/kg for dhATX and 25 mg/kg for ATX). The relative nicotinic acetylcholine receptor (nAChR)-binding affinities of ATX and dhATX were determined using the Torpedo electroplaque assay which showed consistency with the relative toxicity determined by ip injection. This work highlights that toxicity studies based solely on ip injection may not yield LD50 values that are relevant to those derived via oral administration, and hence, do not provide a good estimate of the risk posed to human and animal health in situations where oral ingestion is the likely route of exposure. The high acute oral toxicity of dhATX, and its abundance in M. autumnalis proliferations, demonstrates that it is an important environmental contaminant that warrants further investigation.