Bioavailability and impact of sediment-bound endocrine disrupting chemicals on fish in context of flood events

  • Bioverfügbarkeit und Auswirkungen von sedimentgebundenen endokrin wirksamen Chemikalien auf Fische im Kontext von Hochwasserbedingungen

Müller, Anne-Katrin; Hollert, Henner (Thesis advisor); Schäffer, Andreas (Thesis advisor); Segner, Helmut (Thesis advisor)

Aachen (2020)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2020


To date, numerous studies worldwide have demonstrated that sediments function as a sink for a great variety of environmental pollutants, among them are substances interfering with the endocrine system, so called endocrine disrupting chemicals (EDCs). Estrogenic activity evaluated in sediment samples across Europe via in vitro bioassays ranged from 0.02 up to 55 ng 17β-estradiol (E2) equivalents/g sediment. This is of particular concern since it is well documented that waterborne exposure to even low ng/L concentrations of EDCs can impair the reproduction of freshwater fish species. Feminization of male fish is one of the most notable adverse impacts of exposure to EDCs and the production of the female egg yolk protein vitellogenin (vtg) has been observed to coincide with impairment of gonadal development evident as intersex and, ultimately, reproductive failure. In contrast, little is known about the bioavailability and effects of sediment-associated EDCs on fish. Particularly when sediments are perturbed, e.g., during flood events, sediment-bound substances may become bioavailable. During the past decades, several extreme flood events have occurred in central Europe, including Germany. The likelihood and intensity of flood events have been predicted to increase as a result of global climate change. As consequence, the European Parliament established the Directive 2007/60/EC on the assessment and management of flood risk. In order to minimize adverse consequences of flood events to humans and the environment part of such risk assessment is the evaluation of potential sources of environmental pollution as result of flooding. The main objectives of the present thesis, as part of the interdisciplinary Project House Water - a project supported by the German Excellence initiative - , were to (i) investigate the bioavailability of sediment-bound EDCs under flood-like conditions when the sediment is subject to suspension; (ii) evaluate the uptake of sediment-bound EDCs during such a simulated flood event into fish and whether this leads to endocrine responses in the fish and (iii) assess the impact of sediment-bound EDCs to freshwater fish species inhabiting a "hot-spot" of EDC contamination in sediment under field conditions. In doing so, this thesis aimed to provide implications for risk evaluation of sediment associated contaminants with special emphasis on flood events. In order to assess the bioavailability of sediment-bound EDCs, the Luppe River previously described as a "hotspot" for EDC accumulation in sediment was chosen as a study site. The concentration of target EDCs and estrogenic activity of sediments from the Luppe River were investigated using chemical analysis (LC-MS/MS) in addition to effect-based methods, such as a novel screening tool (planar Yeast Estrogen Screen; p-YES) that utilizes high performance thin-layer chromatography plates in combination with an in vitro bioassay (YES). Estrone (50%, E1) and nonylphenol (35%, NP) accounted for the majority of estrogenic activity reported for sediment with up to 20 ± 2.4 ng 17β-estradiol equivalents (EEQ)/g dry weight in the Luppe sediments. E2 accounted for approximately 14% of the estrogenic effect, whereas the estrogenic effect attributed to 17α-ethynylestradiol (EE2), when present, was negligible (approx. 1%) from sediment across all Luppe sampling sites. Two types of passive samplers (polar organic chemical integrative sampler (POCIS) and Chemcatcher) were used to investigate the bioavailability of EDCs from suspended sediment under laboratory conditions. NP, E1, E2 and EE2 were remobilized from Luppe sediment when subjected to turbulent conditions, such as in a flood event, and were readily bioavailable at ecotoxicologically relevant concentrations (NP 18 µg/L, E1 14 ng/L, E2 0.2 ng/L, EE2 0.5 ng/L). Both types of passive samplers were applicable in a sediment-water suspension system, with the Chemcatcher displaying higher sampling rates compared to the POCIS. A laboratory exposure study with juvenile rainbow trout (Oncorhynchus mykiss) was conducted to evaluate uptake and ecotoxicological impact of remobilized sediment-bound EDCs from the Luppe River. Therefore, rainbow trout were exposed over 21 days to constantly suspended sediment in the following treatments: i) a contaminated sediment from the Luppe River ii) a control sediment (exhibiting only background contamination), iii) a serial dilution of Luppe sediment with the sediment control (1:8; 1:4; 1:2), and iv) a water-only control. Measured estrogenic activity using in vitro bioassays as well as target analysis of NP and E1 via LC-MS/MS in sediment, water, fish plasma, as well as bile samples, demonstrated that sediment-bound EDCs became bioavailable during the simulated flood event. EDCs were dissolved in the water phase, as indicated by passive samplers, and were readily taken up by the exposed trout. Interestingly, similar patterns of EDCs were observed in the water and fish blood and bile, suggesting that EDCs partitioned from sediment into the water and subsequently absorbed by the fish, indicating that freely dissolved aqueous concentrations of EDCs might be a major route for uptake of EDCs in fish. An estrogenic response of fish to Luppe sediment was indicated by increased abundance of transcripts of typical estrogen responsive genes, i.e. vitelline envelope protein α, in the liver and vitellogenin induction in the skin mucus. Hepatic gene expression profiles by RNA-sequencing were altered in Luppe exposed fish compared to controls, whereas the repression of a great number of genes involved in cell cycle in combination with induction of apoptotic markers suggest a broader response. However, similar downregulation of cell cycle genes was observed with the sediment control. Together with histological alterations, i.e. local areas of cell lysis, infiltration of immune cells and degenerative nuclear alterations, observed in the liver of fish throughout all treatments, indicates that exposure to suspended particles might elicit stress at the cellular level. Moreover, tench (Tinca tinca) and roach (Rutilus rutilus) as a benthic and pelagic living fish species, respectively, were sampled at the Luppe River. A field reference site, the Laucha River, in addition to fish from a commercial fish farm as a reference were studied. Blackworms (Lumbriculus variegatus), which are a source of prey for fish, were exposed to sediment of the Luppe River and estrogenic activity of worm tissue was investigated using in vitro bioassays. A 153-fold greater estrogenic activity was measured using in vitro bioassays in sediment of the Luppe River compared the Laucha River. Estrogenic activity of Luppe exposed worm tissue (14 ng EEQ/mg) indicated that food might act as secondary source to EDCs. While there were no differences in concentrations of NP in plasma of tench from the Luppe and Laucha, vtg as biomarker for exposure to EDCs was induced in male tench and roach from the Luppe River compared to both the Laucha and commercially cultured fish by a factor of 264 and 90, respectively. However, no histological alterations in testis of these Luppe exposed fish were observed. Our findings suggest that sediments substantially contribute to the overall EDC exposure of both benthic and pelagic fish in the field but that the exposure did not translated to adverse effects on the gonad level and, thus, might not be of relevance for the reproductive success of these populations in the wild. The present thesis demonstrated that sediments not only function as a sink for EDCs but can turn into a significant source of pollution when sediments are resuspended. The results demonstrated that sediment-bound EDCs were readily bioavailable for fish under conditions similar to those of a flood event. Partitioning of EDCs into the water phase might be a major route for uptake of remobilized sediment-bound EDCs into the fish. Passive sampling was a useful tool to assess the bioavailability of sediment-bound EDCs and could be a good indicator of sediment toxicity in a regulatory context. Overall, the work described in this thesis greatly contribute to the assessment of sediment-bound EDCs in the context of flood risk.