ESR2 Xiaomeng Liang

Project: Identification of plant produced natural toxins leaching from soil

Principal supervisor: Prof. Jan H. Christensen

Intro to project: Phytotoxins are naturally occurring small organic molecules (< 1000 Da) synthesized by plants as secondary metabolites (PSM). They have recently been recognized as very important environmental micropollutants (MPs) which when leaching into the aquatic system might become a threat to public health. In this project, we aim to establish a novel analytical workflow, employ it for chemical fingerprinting of plant and environmental matrices, and determine those environmental-relevant phytotoxins. The workflow are expected to encompass a fast and efficient extraction methodology which can be used to selectively extract phytotoxins from complex plant and soil matrices for LC-ESI-HRMS based analysis, a sensitive, reliable and reproducible non-target analysis methodology which can be used to detect phytotoxins in a multiple of matrices, and a state-of-the-art data-mining strategy which can be adapted in non-target analysis for structure elucidation and compound identification; we also expect to establish a comprehensive in-house database of the desired phytotoxins. We consider the non-target approach is crucial in determining environmental-relevant phytotoxins because i) the compounds may have large variances in distribution depending on plant species, geographical conditions and the climate, ii) unlike anthropogenic analytes, phytotoxins are labile and prone to form transformation products which are hard to predict and analyze using traditional target or suspect analysis approach.

Outcome
The project started in Jun 2017 with initial focus on two classes of natural compounds steroids and aromatic polyketides. Later we included alkaloids, flavonoids and a few other class plant secondary metabolites into our interest list. A literature review on related analytical methodologies was prepared and an in-house database was built and served as a reference in LC-MS based compound screening.

Sample preparation is the first step in almost all types of analytical studies; the extraction efficiency and selectivity is crucial for data quality. The challenge of the sample preparation step in this project is mainly associated with the high matrix complexity (eg. chlorophylls and lipids in plant, organic matter in soil, salts in natural water), trace level in presence and possibly light-/heat- unstable. We therefore developed a one-step selective pressurized liquid extraction (sPLE) method which can be used to simultaneously extract secondary metabolites and remove potential interferences in solid matrices (plant and soil). Parameters of extraction temperature, solvent, the pre-heat/heat time were investigated and optimized meanwhile a sorbent layer of C18 was applied to remove interfering co-extractives. (https://doi.org/10.1016/j.acax.2020.100040)

In the next, an LC-ESI-HRMS method protocol was developed for chemical fingerprinting of phytotoxins in all three matrices involved in the project (soil, water and plant); the protocol covers from non-target screening, suspect screening to target analysis for different research scenarios in practice. Key factors of mobile phase additive/pH, organic modifier, column temperature and gradient time were investigated and optimized in both ESI+ and ESI- mode. (Link)

In parallel, we also investigated the possibility of LC×LC-MS setup as the additional dimension can be an advantage in comprehensive analysis, especially with respect to resolving isobaric compounds. We also looked into data-independent MSE acquisition apart from MS/MS acquisition and investigated the possibility of adapting curve resolution techniques in nontarget/suspect analysis. These two researches are however designated as a future expectation of the project therefore not expected to be finalized within the time-period of employment.

Furthermore, we integrated the workflow in a half-year Lupin field monitoring research. The field is located in Reckenholz, Switzerland; in there Lupinus angustifolius L. (blue lupin) was sowed in April 2019 and harvested in August in the same year. Soil, drainage water and surface water samples were collected prior to ploughing, through the entire period of plant growth and development, and continued afterwards for approximately a month; alongside we also collected plant samples and further distributed into leaves, stems, roots, etc. After screening and comparison of the chemical fingerprints with those in plant samples, we (tentatively) identified over thirty lupin-produced toxins in soil and water samples, mainly being flavonoids and alkaloids. We also captured a leaching pattern of some of the phytotoxins into the drainage during and after irrigation, which may indicate a wash-off event (rain, irrigation, etc) can facilitate the occurrence of phytotoxins in aquatic environment. See link to publication list.