An in vitro – in vivo integrated approach for hazard and risk assessment of silver nanoparticles for soil organisms

Summary

Owing to their small sizes, nanoparticles (NPs) exhibit completely different and novel characteristics compared to their bulk counterparts of the same chemical composition. These novel properties include increased reactivity due to large specific surface area, fluorescence and colour changes, increased biological barrier crossings and increased material strengthening combined with light-weight. Virtually all fields of human endeavours are exploiting nanotechnology to combat different challenges. This has led to an increase in the production and potential release of NPs into the environment. The novel properties of these NPs however, mean an enhanced potential for interactions with biological systems that are different from the interactions of known conventional chemicals, thus raising environmental and public health/safety concerns. Available literature has reported NP uptake in different organisms along with associated hazards. Therefore, to safeguard human and environmental health and safety, regulatory measures are necessary. Such measures must be based on sound scientific evidence and be risk-based rather than hazard-based. As such, the need to understand the fate of NPs after environmental release and their potential to pose hazards and risks to the environment is critical for a proper risk assessment and further development of policy strategies on the future regulation of the use of NPs.

Some studies have demonstrated different and sometimes conflicting effects of NP properties on their uptake in different organisms. Given that exposure determines whether hazards will turn into risks, there is a critical need for further systematic evaluation of the physico-chemical properties of engineered or manufactured NPs that influence uptake in terrestrial organisms, and also of how soil properties may affect these processes. The objective of this project was to determine the influence of size and surface coating (charge), two important physico-chemical properties of NPs, on their bioavailability, uptake and toxicity. The red earthworm Lumbricus rubellus, common in most parts of Europe, was used as a model soil organism. Silver nanoparticles (AgNPs) have been identified as one of the most commonly used NPs in many products, and their production is expected to continue to increase. Therefore, we selected AgNPs as our model NPs. For our investigations, we applied an integrated in vitro - in vivo approach, utilising high throughput in vitro methods as well as well-established in vivo toxicity end-points in the earthworm. A systematic experimental approach was developed for which AgNPs were synthesized in three sizes: 20, 35 and 50 nm. Surface-coating with bovine serum albumin (AgNP_BSA), chitosan (AgNP_Chit), or polyvinylpyrrolidone (AgNP_PVP) resulted in negative, positive and neutral particles respectively.

Firstly, macrophage cells (RAW 264.7) were exposed to AgNPs at 0 – 200 µg/mL (nominal concentrations) and uptake dynamics, cell viability, as well as induction of tumour necrosis factor (TNF)-α and reactive oxygen species (ROS) were assessed (Chapter 2). Generally, the adverse effects of exposure to the tested AgNPs resulted in reduced overall viability of the cells, which was similar for all AgNPs tested. On adenosine triphosphate (ATP) production and specific mechanisms of toxicity (TNF-α and ROS production) however, we observed that the AgNPs differed significantly, with the negatively charged AgNP_BSA being the most toxic. Significant ROS induction was only observed after exposure to the 20 nm positively charged AgNP_Chit. Effect of size was less prominent than that of surface coating, showing mostly limited differences that were not statistically significant under our experimental conditions. Live confocal imaging of exposed cells allowed the monitoring of the uptake dynamics and subcellular cytoplasmic accumulation of AgNPs. We observed fast uptake of AgNPs within 2.5 hours which is essential in case of exposure durations of 6 and 24 hours, as applied in our experiments. However, similar uptake did not always result in similar effects.

With the insights obtained from the in vitro assessments, we investigated the effects of size and surface coating (charge) of AgNPs on the bioaccumulation in, and toxicity (survival, growth, cocoon production) to the earthworm L. rubellus. Currently, metal engineered NPs in tissues are generally quantified based on total metal concentrations after acid destruction of samples. Such destructive methods are limited in providing information on the speciation and the forms of NPs which is essential for characterising the fate of NPs. In the present thesis, we developed a method using a combination of enzymatic tissue processing and single particle inductively coupled plasma–mass spectrometry (sp-ICP-MS) to characterise and quantify AgNPs in tissues of earthworms (Chapter 3). Subcellular fractionation of tissues was also applied to investigate potential association of AgNPs with the cellular metallothionein (MT) containing fraction of the earthworm tissues. This study provided, to the best of our knowledge, the first estimates of tissue Ag concentrations in both particulate and ionic forms in earthworms exposed in vivo to AgNPs via soil. The results obtained showed fairly low uptake of AgNPs, with earthworms exposed to a commercially obtained PVP-coated AgNP showing approximately 34% of their total Ag tissue burden being in particulate form. This indicates that although AgNPs accumulated in tissues of earthworms in their primary form, the dissolution of Ag in the soil, organism, or both played an important role in determining the ultimate fate of the AgNPs. Although the biological uptake of AgNPs was generally low, the method described in Chapter 3 was still capable of extracting NPs in quantities sufficient for identification, quantification and characterisation. It should be noted however, that the lower size detection threshold for the ICP-MS instrument used for these analyses is approximately 30 nm. Consequently, information on NPs smaller than 30 nm was not available. With the increasing optimisation of analytical systems that combine sp-ICP-MS, or other detection methods with, for example, asymmetric flow field-flow fractionation (AF4) which pre-sort different particle sizes, the potential for application of methods described in this thesis will be even greater.

Having developed a method for extracting Ag from tissues, we exposed earthworms to all nine synthesized AgNPs as well as to AgNO3 at two concentrations below known EC50s to control for ionic effects of Ag in a 28-day sub-chronic reproduction toxicity test in soil in Chapter 4. Uptake was observed to be generally highest for the negatively charged AgNP_BSA especially at the lower exposure concentration ranges. Total Ag concentrations in earthworm tissues reached a plateau level of about 80 mg Ag/kg dry weight (DW) for exposure concentrations between 15 – 100 mg Ag/kg soil DW. Reproduction was impaired at high nominal soil concentrations of all AgNPs tested, with AgNP_BSA particles being the most toxic. Size had an influence on uptake of the AgNP_PVP, showing both uptake and effect on reproduction of the 20 nm sized group to be significantly more than those of the 35 and 50 nm AgNP_PVP. This size effect however, did not hold for AgNP_BSA nor AgNP_Chit. Higher uptake from the soil may consequently lead to a higher potential for toxicity in organisms. Interestingly, internal total Ag tissue concentrations measured after 72 hour exposure were better at predicting the effect on reproduction than tissue concentrations after 28 days exposure. It is likely therefore, that reproduction was affected already in the 72 hour exposure window.

In order to further elucidate the likely mechanisms by which these AgNPs were exerting their effects, we conducted a toxicogenomic study in Chapter 5. Although AgNPs have been increasingly investigated, information regarding their effect on the gene expression profile of especially soil organisms is yet inadequate. Using RNAseq, we investigated the transcriptome and gene expression profiles of the earthworm L. rubellus, following exposure to the nine AgNPs. Overall, exposure to medium sized AgNPs at a concentration close to the EC50 for effects on cocoon production caused most pronounced responses at the transcriptional level. There was a correlation however, between the numbers of differentially expressed genes (DEGs) and internal Ag concentrations in the earthworms. Within the medium size AgNPs, AgNP_BSA caused extensive transcriptional responses, with 684 genes affected. In contrast ionic silver (AgNO3) did not affect gene expression at low as well as higher exposure levels. Only one gene was regulated by all AgNP and Ag+ treatments, indicating that there was hardly any functional overlap between the responses of the organisms to AgNPs with different coatings. Remarkably, this gene was metallothionein, a cysteine-rich peptide known to strongly bind free metal ions for chelation and detoxification, which was strongly up-regulated. Gene ontology enrichment analysis for 35 nm AgNP_BSA exposures revealed a total of 33 significantly enriched gene ontology terms related to biological processes. These included responses to pH, proton transport, cell differentiation, microtubule organisation, and and MT induction. Surface coating (BSA) was important in triggering the AgNP-induced differential gene expression profiles in earthworms. The importance of physicochemical properties of NPs in influencing their fate and toxicity is thus elucidated in the current study.

The studies reported in the current thesis showed that within the range of 20 to 50 nm, effects of the size of AgNPs on toxicokinetics and toxicodynamics are limited. However, effects of surface coating were consistent over the different levels of biological integration. Generally, the negatively charged AgNP_BSA accumulated to a higher extent in the earthworms, especially at lower concentrations. The in vitro uptake was fast for all NPs, but also showed the highest uptake of AgNP_BSA. The negatively charged AgNPs were also the most toxic, likely related to their increased uptake. This was evident at all levels: gene expression, cellular, and individual (population dynamic parameters) levels. At the in vitro level, this applied mostly to effects on specific modes of action (TNF-α induction, ROS production). For more general cytotoxic effects, the effects of surface coatings were less evident. Except in cells exposed to AgNP_Chit 20 nm, where there was a slight increase in ROS production, this set of AgNPs under the experimental conditions applied, did not appear to induce the production of ROS. This was supported by the lack of expression of any ROS-related gene in the gene expression profile analyses.

Based on the results of the current research, it can be concluded that the physico-chemical properties of NPs do influence their environmental fate and toxicity. It should be noted however that general predictions on the outcome of exposure to NPs are difficult to make, and NPs should be evaluated on a case by case basis. Our research supports the use of in vitro models to limit and prioritize further in vivo studies. Studies investigating the fate and effects of NPs for soil organisms are vital for a holistic approach towards a comprehensive and adequate environmental risk assessment (ERA). The studies described in this thesis contribute to this knowledge, thereby improving our understanding of the hazards and risks due to exposure to AgNPs, thus enabling their adequate and comprehensive ERA.