Are standard ecotoxicological bioassays a blessing or a curse?

David Spurgeon
Centre for Ecology and Hydrology
dasp@ceh.ac.uk
ECG Bulletin January 2016

The need to manage the use of chemicals in a sustainable manner has led to the development of scientific and regulatory approaches for assessing prospective and retrospective risk. As an emerging area of industrial development that can have potential environmental impact, nanotechnology requires similar regulation. With the growing production of a wide array of nanomaterials with different sizes, compositions, surface coatings and other properties, there is concern that some of these nanomaterials may be toxic in ways that are not predictable based on bulk material properties. An initial challenge for the regulation of nanotechnology products is to assess which components of the current chemical risk assessment methods should be used with nanomaterials. Reliable methods are needed for studying how different nanomaterials may behave in natural systems and how they interact with living organisms.

Credit: kwanchai.c/Shutterstock

As it grows, the field of nano-ecotoxicology has been able to draw extensively on existing tools and techniques for assessing chemical risk. Within regulatory ecotoxicology, a key recognised requirement is to have standardised protocols available for toxicity testing. Such agreed protocols are vital because they ensure that studies conducted in different laboratories can be compared and can inform risk assessments for nanomaterial regulation under different jurisdictions. Standardised toxicity tests have also been developed to inform nanomaterial regulation. However, these standard test methods are often also used in more fundamental nanotoxicology research. Often these uses are entirely appropriate for the question under study. However, in other cases, limiting studies to standard species, test methods, and endpoints may lead researchers to miss key aspects and issues in eco-nanotoxicology that would be revealed by the use of nonstandard test systems. For example, a growing number of studies have assessed the responses of soil-dwelling species to nanomaterial exposure, recognising that soils are likely sinks for nanomaterials. These studies illustrate the importance of going beyond a standard set of species, soils, and endpoints to understand the toxicity effects of nanomaterials.

Selection of species for testing
Toxicity can rarely be examined for a wide range of species or at the ecosystem scale, except for the chemicals of  highest environmental concern. To protect ecosystem structure and function, risk assessment generally relies on data for a very limited number of (usually short-lived) species. To assess nanomaterial toxicity for soil ecosystems, the most commonly used test species are the earthworm Eisenia fetida, the springtail Folsomia candida and the nematode Caenorhabditis elegans (1, 2, 3). Standardised tests for these species have been available for over 10 years and are widely used in nano-ecotoxicology research. Of the 79 papers published on nanomaterial toxicity to soil organisms by June 2015, 64% studied just these three species (E. fetida, 28%; F. candida, 9%; C. elegans, 27%).

These studies have led to a number of insights. At least for metal and metal oxide nanomaterials made from silver, zinc oxide and copper oxide, nanomaterial toxicity is most frequently lower than the toxicity of the metallic component alone in ionic form. Indeed, meta-analyses by Notter et al. (4) revealed that nanomaterial toxicity was at or below ionic metal toxicity in 93.8%, 100% and 81% of paired studies for Ag, CuO and ZnO nanomaterials respectively. However, the focus on so few species limits the chance to identify unexpected effects in species that are rarely, if ever, used for study. On a more practical level, focus on just a few organisms makes it difficult to develop robust species sensitivity distributions. For certain metals and pesticides (e.g. zinc, chlorpyrifos), it has been possible to build robust species sensitivity distributions as an excellent basis from which to derived scientifically credible environmental quality standards (5, 6). The current focus on using mainly standard test species may delay such comprehensive assessments for nanomaterials in soil.

Use of a standard test medium
One of the most important innovations to allow the standardisation of terrestrial toxicity tests has been the use of standard test soils. Natural (e.g. LUFA 2.2) and/or artificial soils have been proposed and used for this purpose. The deployment of the same media across tests provides a constant set of conditions that govern bioavailability for the same soil in different laboratories. However, the bioavailability and toxicity of metals are strongly affected by different soil properties, such as soil pH, organic matter content, clay content, cation exchange capacity, and iron oxide content. These soil properties may also affect how nanomaterials behave in soil. For example, the chemical composition of the nanomaterial, pH, organic matter content and soil clay composition can all affect nanomaterial uptake by organisms and the resulting toxic effects (7, 8). For metals, key innovations such as the terrestrial biotic ligand model (9) provide a mechanistic framework for identifying major soil physiochemical drivers that influence nanomaterial toxicity; knowledge of these drivers can be used to support site specific risk assessments. To extrapolate nanomaterial toxicity observed in one soil type to another, there is an urgent need for the development of variants of these models. Such developments require data on toxicity in different soil types, not just standard tests with a single soil type.

Measurement endpoints
The classic endpoint measured in toxicity tests is mortality. However, this endpoint is a crude measure of possible ecological effect. In many standard tests for soil species, the capacity also exists to measure sublethal effects e.g. on reproduction, growth, and even multiple life cycle traits. Going beyond organism life cycles, there remains a dearth of studies of nanomaterial effects on genetic/epigenetic, physiological and behavioural properties of sensitive soil organisms. Recent studies on the effects of metal, pharmaceutical and pesticides have shown that these biological traits may be much more sensitive to chemical influence than are more classically assessed endpoints.
 
There is also some evidence that nanomaterials may differ from their ionic chemical counterparts in how they are taken up by organisms. Transcriptomic and biological imaging studies for soil organisms have suggested the possibility of active uptake of nanomaterials though cellular endocytosis pathways (10, 11). This uptake mechanism leads to a higher overall accumulation of nanomaterials than for their metal constituents. The long-term physiological consequences of this “over-accumulation” of nanomaterials have not been fully established. It would benefit the field greatly to understand how these effects may underpin impacts on sensitive endpoints and the resulting long-term effects. This will require tests that go beyond durations routinely used in standard tests.

Long-term / transgenerational toxicity
Perhaps the greatest challenges for interpreting the effects of exposure to nanomaterials from current standardised toxicity tests is the relatively short-term nature of these assays. For many aquatic toxicity tests, the period of exposure is very short (48 or 96 hours). For soil toxicity tests, exposure periods are often longer (e.g. 28 days). However, considering that an earthworm may be able to survive for up to 5 years, this is a relatively short exposure given the organism’s lifespan. Furthermore, chemical and nanomaterial interactions with the different components of the epigenome may affect to what extent toxic effects are transferred to subsequent unexposed generations (12). Further, a recent study conducted in our laboratory with the nematode C. elegans has shown that long-term multi-generational exposure results sensitises subsequent generations to silver nanoparticles (and also silver ions), and that this sensitisation is retained even if exposure is removed for five generations. Thus, although short-term tests may be ideal for identifying acute toxicity, new approaches will be needed to understand complex, long-term effects.

Outlook
Given the expected increase in nanotechnology product use, we will become more reliant than ever on the availability of robust risk assessment scheme to help balance risks and benefits of these novel materials. Work conducted to date has shown that key methods for chemical fate modelling, toxicity testing, and biological endpoint assessment are often fit for purpose for applications in regulatory nanomaterial assessments. Existing standardised toxicity tests provide tractable and useable tools for the rapid generation of initial toxicity data that are comparable between laboratories, with only a few modifications needed for their use  with nanomaterials. Although valuable for routine studies, standard tests are, however, not a panacea for all cases. Because standard tests encourage the use of a very limited number of species, many species, and indeed whole phyla, will rarely if ever be tested. This raises the real possibility that vulnerable species are, and will continue, to be missed. Further, it is difficult to know just how relevant short-term data in a single standard test medium are for different environments, over extended exposure times, and for all species in exposed communities.

When considering the use of standard tests for nano-ecotoxicology, the words of Sumpter and Johnson (13) on the lesson learned from studies of endocrine disrupting chemicals in the aquatic environment should be borne in mind. They concluded: “One lesson ... during the 10 years of research on endocrine disruption is that the current testing regime used to determine the toxicity of a chemical to aquatic organisms has “failed”, in the sense that it has not detected the endocrine activity of many chemicals” and that “even more surprisingly, it transpired that EE2 [ethinyl estradiol, a potent endocrine disrupting chemical] is acutely toxic to fish, but this effect is delayed long enough that it is not detected in the acute toxicity test protocols used currently.” These comments suggest that we should be careful as scientists not to overly rely on standard tests for ecologically meaningful risk assessment of nanomaterials released into the environment.

References
1. OECD, Guidelines for the testing of chemicals. No. 207 Earthworm acute toxicity tests (OECD, Paris, France, 1984).
2. International Organization for Standardization, Soil quality - Inhibition of reproduction of collembola (Folsomia candida) by soil pollutants (ISO, Geneva, 1999).
3. ASTM, Standard Guide for Conducting Laboratory Soil Toxicity Tests with the Nematode  Caenorhabditis elegans (ASTM International, Conshohocken, USA, 2014).
4. D. Notter, D. M. Mitrano, B. Nowack, Environmental Toxicology and Chemistry 33, 2733 (2014).
5. G. K. Frampton, S. Jansch, J. J. ScottFordsmand, J. Rombke, P. J. VandenBrink, Environmental Toxicology and Chemistry 25, 2480 (2006).
6. P. A. VanSprang, F. A. M. Verdonck, F. VanAssche, L. Regoli, K. A. C. DeSchamphelaere, Science of the Total Environment 407, 5373 (2009).
7. L. R. Heggelund, M. Diez-Ortiz, S. Lofts, E. Lahive, K. Jurkschat, J. Wojnarowicz, N. Cedergreen, D. Spurgeon, C. Svendsen, Nanotoxicology 8, 559 (2014).
8. L. Settimio, M. J. McLaughlin, J. K. Kirby, K. A. Langdon, L. Janik, S. Smith, Environmental Pollution 199, 174 (2015).
9. A. Peters, Graham Merrington, Biotic ligand models, metal bioavailability and regulatory application, ECG Environmental Brief XI, ECG Bulletin, January 2016, pp. 23-24 (2016).
10. O. V. Tsyusko, J. M. Unrine, D. Spurgeon, E. Blalock, D. Starnes, M. Tseng, G. Joice, P. M. Bertsch, Environmental Science & Technology 46, 4115 (2012).
11. M. Novo, E. Lahive, M. Diez-Ortiz, M. Matzke, A. J. Morgan, D. J. Spurgeon, C. Svendsen, P. Kille, Environmental Pollution, 205, 385 (2015).
12. C. Voelker, C. Boedicker, C., Daubenthaler, J., Oetken, M. & Oehlmann, J. (2013) Comparative Toxicity Assessment of Nanosilver on Three Daphnia Species in Acute, Chronic and Multi-Generation Experiments. Plos One 8, e75026 (2013).
13. J. P. Sumpter, A. C. Johnson, Environmental Science & Technology 39, 4321 (2005).