ESR9 Carina Schönsee – University of Copenhagen

NaToxAq > PhD Projects > ESR9 Carina Schönsee

ESR9 Carina Schönsee

Project: Physicochemical property determination of natural toxins

Principal supervisor: Dr. Thomas Bucheli

Intro to project:

Natural toxins are not yet commonly regarded as environmental contaminants of concern for water quality. As hardly any experimental data describing their aquatic mobility are available and in silico prediction tools show limited applicability, effective environmental risk assessment is difficult. Thus, systematic research on environmental distribution processes is urgently needed to determine whether or not natural toxins are mobile enough to end up or enrich in potential drinking water abstraction sites.

Therefore, this project aims to establish HPLC-based high-throughput methods for the systematic evaluation of model phase partitioning constants (Kow, Koc) under changing environmental conditions. Generated data will be used to validate and improve prediction models and ultimately provide indications on those natural toxins potentially posing a threat to water quality.

Within this project, two OECD methods were optimized for the determination of pH-dependent octanol-water partitioning coefficients (Dow) of natural toxins in the environmentally relevant pH range from pH 4 to 10. As HPLC based methods, both approaches show the capability to be largely automatized for more efficient, less error-prone analysis and thus allow the reliable determination of Dow in the for potential aquatic contaminants relevant range of log Dow < 4.

The methods have been applied to about 50 natural toxins of different origin and from several compound classes (phyto-, mycotoxins; alkaloids, polyketides, steroids, terpenoids; e.g., colchicine from Colchicum autumnale, autumn crocus). With regards to predicted toxicity, persistence and mobility as well as plant occurrence, specific alkaloid subclasses such as pyrrolizidine alkaloids from Senecio spp. are investigated in more detail. A publication of the generated data set is currently in preparation.

Predicted (EPISuite, ACDLabs), literature and in this project experimentally determined (each run n=3) log Kow values for the reference phytotoxin colchicine from autumn crocus (photo from Beat Bäumler – Bürenberg (BE), https://www.infoflora.ch/de/flora/colchicum-autumnale.html).

As an indicator for the partitioning of natural toxins from aqueous media to organic matrices, Dow can be seen as first proxy estimating natural toxin mobility in the aquatic environment. Thus, experimental data helps in prioritization of toxins for further research activities, including field studies and lab-based characterization of fate processes within NaToxAq.

Column packing performed in
our laboratory at Agroscope.

Additionally, a column chromatography system has been set up to systematically study sorption of natural toxins to different geosorbents. For this purpose, HPLC columns are manually packed in our labs with the sorbents of interests . Using organic matter as a first sorbent, organic carbon-water partitioning coefficients (Koc) are derived as primary toxin mobility indicator. First results show that column chromatography can be reliably applied in sorption studies of large diverse sets of mobile compounds such as natural toxins. The very short analysis time and little material requirements easily allow systematic investigations of varying influences (e.g., pH) on sorption.

Thus, detailed mechanistic insights are gained that are of great value for understanding transport and fate processes in the environment. Applying the method to other sorbents in the future (e.g., minerals, activated carbon) will help to determine their individual contribution to the natural toxins’ overall mobility on their path from source to tap (drinking water).

A case study will exemplify how improved physicochemical properties and experimental data in general will help predicting environmental fate of a specific group of model toxins. As a final step, generated data will be used in evaluation of commonly used prediction models. The basis for both is the close collaboration with other ESRs, particularly during upcoming secondments to the University of Copenhagen and Stockholm University.