Unravelling hazards of nanoparticles to earthworms, from gene to population

Nanotechnology is an expeditiously growing field, where engineered nanoparticles are being incorporated in many different applications, from food to waste water treatment (Dekkers et al. 2011; Gottschalk and Nowack 2011; Savage and Diallo 2005). Due to this large scale production and use of nanoparticles, their release into the environment seems inevitable (Crane et al. 2008; Handy et al. 2008a; Oberdörster et al. 2005). Actual exposure levels of nanoparticles under field conditions and the hazards of nanoparticle exposure to the environment are poorly understood, especially for the soil environment (Kahru and Dubourguier 2010; Navarro et al. 2008; Shoults-Wilson et al. 2011a). Given the need for better characterization of hazards of engineered nanoparticles to the environment and soil organisms in particular, the aim of the present thesis was to investigate effects of nanoparticle exposure on the earthworm Lumbricus rubellus, as a model organism for soil ecotoxicology, and to contribute to the development of effect markers for engineered nanoparticle exposure in this model. The present thesis was divided in different chapters. Chapter 1 provides an introduction to the topic and discusses the importance of research on the hazards of exposure to engineered nanoparticles. Furthermore, the aim and outline of the thesis are presented, with background information on the model organism, effect markers and nanoparticles. In chapter 2 effects of exposure to the fullerene C60 (nominal concentrations 0, 15.4 and 154 mg C60/kg soil) on survival and growth during the different life stages of L. rubellus (cocoon, juvenile, subadult and adult), as well as reproduction were quantified. These important individual endpoints for population dynamics were incorporated in a continuous-time life-history model (Baveco and De Roos 1996; De Roos 2008). In this way, effects of C60 exposure on the individual endpoints could be extrapolated to implications for population growth rate and life stage distribution, i.e. the development of the population in terms of number of individuals in the different life stages. These implications at the population level may be more relevant for the ecological impact of C60 than effects on endpoints at the individual level (Klok et al. 2006; Widarto et al. 2004). At the individual level C60 exposure caused significant adverse effects on cocoon production, juvenile growth rate and survival. When these endpoints were used to model effects on the population level, reduced population growth rates with increasing C60 concentrations were observed. Furthermore, a shift in life stage structure was shown for C60 exposed populations, towards a larger proportion of juveniles. This result implies that the lower juvenile growth rate induced by C60 exposure resulted in a larger proportion of juveniles, despite increased mortality among juveniles. Overall, this study implied serious consequences of C60 exposure for L. rubellus earthworm populations, even at the lowest level of exposure tested. Furthermore, it showed that juveniles were more sensitive to C60 exposure than adults. To complement the observations made on survival, growth and reproduction described in chapter 2, subsequent investigations on cellular and molecular responses of the earthworms to C60 exposure were performed (chapter 3). A set of established effect markers was used, which reflect different levels of biological organisation in the earthworm and may inform on the toxic mechanisms of adverse effects induced by C60 exposure (Handy et al. 2002; Heckmann et al. 2008). At the molecular level, four specific effect markers were selected, including markers for generic stress (heat shock protein 70 (HSP70) (van Straalen and Roelofs 2006), for oxidative stress (catalase and glutathione-S-transferase (GST) (Kohen and Nyska 2002) and for an immune response (coelomic cytolytic factor-1 (CCF-1) (Olivares Fontt et al. 2002). At the tissue level, histological analyses were used to identify damage to cells and tissues, and indications of inflammation in the tissues. In these investigations, exposure to C60 (0, 15 or 154 mg C60/kg soil) affected gene expression of HSP70 significantly. Gene expression of CCF-1 did not alter in adult earthworms exposed for four weeks, but was significantly down-regulated after lifelong exposure (from cocoon stage to adulthood) of earthworms, already to the lowest C60 exposure level. No significant trends were noted for catalase and glutathione-S-transferase (GST) gene expression or enzyme activity. Tissue samples of the C60 exposed earthworms from both experiments and exposure levels, showed a damaged cuticle with underlying pathologies of epidermis and muscles. Additionally, the gut barrier was not fully intact. However, tissue repair was also observed in these earthworms. In conclusion, this study demonstrated effects of sub-lethal C60 exposure on L. rubellus earthworms, at the level of gene expression and tissue integrity. Although tissue injury is generally associated with an inflammatory response, as part of tissue repair (Cikutovic et al. 1999; Goven et al. 1994), the tissue damage observed for the in vivo C60 exposed earthworms in chapter 3 appeareded to occur without accompanying induced immune responses. The CCF-1 gene expression level was reduced in the C60 exposed earthworms, and histological observations did not show infiltration of damaged tissues by immune cells. In order to obtain further insight in mechanisms of effects observed at the molecular and tissue level on immune related parameters, the sensitivity of immune cells (coelomocytes) of L. rubellus earthworms towards exposure to selected nanoparticles was investigated in vitro (chapter 4). To this end, coelomocytes were isolated from unexposed adult L. rubellus earthworms and exposed to C60 in vitro. After exposure, these coelomocytes were tested for cellular viability, phagocytic activity and CCF-1 gene expression levels. The gene expression of CCF-1 was most affected, demonstrating a strong reduction, which indicated immunosuppression. Experiments with NR8383 rat macrophage cells and tri-block copolymer nanoparticles were used to compare sensitivity of the cell types and showed the usefulness of coelomocytes as a test system for nano-immunotoxicity in general. Overall, this study indicated that the absence of an immune response, in case of tissue injuries observed after in vivo C60 exposure, is likely caused by immunosuppression rather than coelomocyte mortality. In subsequent investigations, the experiments performed for C60 were also carried out with silver nanoparticles (AgNP), both in vivo and in vitro (chapter 5). Effects of AgNP were assessed in vivo at nominal concentrations of 0, 1.5 (low), 15.4 (medium) and 154 (high) mg Ag/kg soil and compared to effects of silver ions, added as AgNO3 (nominal concentration 15.4 mg Ag/kg soil). In a four week reproduction assay, the high AgNP and AgNO3 treatments had a significant effect on cocoon production and high AgNP exposure also caused a reduction in weight gain of the adult earthworms. No juveniles survived the high AgNP treatment, therefore only F1 earthworms from the other exposure treatments were monitored for survival and growth, until adulthood. These individual endpoints were used to model effects on the population level. The low and medium AgNP as well as the AgNO3 treatments significantly reduced the population growth rate. The high AgNP treatment caused complete failure of the population growth. Furthermore, histological examination of the earthworms from all AgNP exposure treatments demonstrated tissue damage, with injuries mainly at the external barriers, e.g. the cuticle and the gut epithelium. In addition, effects of AgNP exposure were assessed in vitro and a reduction of coelomocyte viability was observed in a concentration-dependent manner, although the EC50 was fourteen times higher compared with that for Ag ions, added as AgNO3. Furthermore, characterisation of the in vivo exposure media implied that AgNP remained present in the soil in single and aggregated state, releasing Ag to the soil pore water up to at least eleven months. The ionic fraction of Ag in soils has been suggested to be bioavailable to organisms and (largely) responsible for the observed AgNP toxicity (Coutris et al. 2012; Koo, et al. 2011; Shoults-Wilson et al. 2011b). In comparison, the AgNO3 seemed to dissolve rapidly, as is also known for this metal salt, and fixation of Ag ions by the soil presumably led to a quick reduction of Ag bioavailability (Atkins and Jones 2000; Coutris et al. 2012; Ratte 1999). This is in line with the observation that effects were more prolonged in the AgNP treatments in comparison with the AgNO3 exposed animals. In conclusion, this study indicated that AgNP exposure may seriously affect earthworm populations, with the ability to cause immunotoxicity, injury to the external barriers of the earthworm body and a reduction in growth, reproduction and juvenile survival. Finally, chapter 6 presents a discussion on the findings of the present thesis and provides suggestions for future research.