Determination of ibotenic acid and muscimol, the Amanita mushroom toxins, in human serum by liquid chromatography–tandem mass spectrometry

In our previous article, we reported the analysis of ibotenic acid and muscimol in Amanita mushrooms by liquid chromatography–tandem mass spectrometry (LC–MS–MS). The levels of ibotenic acid and muscimol in the mushroom were as high as 210 and 107 μg/g, respectively. We have since tried to measure the same toxins in human serum obtained from a poisoned subject, who ingested the Amanita mushrooms, by the same method. However, the levels of the toxins in the human serum were about three orders of magnitude lower than those in Amanita mushrooms. In addition, the recovery rates for ibotenic acid and muscimol in human serum were found to be much lower than those in the previous study for the mushrooms. Therefore, we optimized the solid-phase extraction procedure again, and reevaluated the data for validation at much lower levels of ibotenic acid and muscimol in human serum. A 100-μl aliquot of human serum containing the target toxins was mixed with 100 ng of acivicin as internal standard (IS), 200 μl of distilled water, and 100 μl of 0.5 % ammonium hydroxide in distilled water, and vortexed well for 10 s. The mixture was loaded on an Oasis MAX 3cc (60 mg) extraction cartridge. The cartridge was washed with 0.5 ml of distilled water and 1.0 ml of methanol. The target compounds and IS were eluted with 4 ml of 0.05 % trifluoroacetic acid in methanol. The eluate was evaporated to dryness and reconstituted in methanol, and its small volume was subjected to LC–MS–MS analysis with the same TSK-GEL Amide-80 separation column. The LC elution was made in the gradient and isocratic modes. The selected reaction monitoring chromatograms showed clear peaks at 5.3, 3.5, and 3.6 min for ibotenic acid, muscimol, and IS, respectively; the blank serum sample without the target compounds or IS gave no peaks at the respective retention times except for an impurity peak at 6.5 min. There was good linearity from 10 to 1,000 ng/ml for both ibotenic acid and muscimol with correlation coefficients not <0.999. The detection limits (signal-to-noise ratio = 3) were 1.0 and 2.5 ng/ml for ibotenic acid and muscimol, respectively. The recovery rates of the target compounds in sera at five different concentrations were 87.9–103 %. The intraday and interday accuracy and precision data were also generally satisfactory. Using the modified method, the actual concentrations of ibotenic acid and muscimol were measured for a serum sample obtained from an ill patient thought to have ingested Amanita ibotengutake; they were 95.9 and 105 ng/ml, respectively. To our knowledge, this is the first report of analysis of ibotenic acid and muscimol in human serum by an MS technique, which we believe will be very useful in forensic and clinical toxicology.