S.K. Hanna, A. Montoro Bustos, L.D. Scanlan, S. Hosbas Coskun, M.E. Johnson, M.R. Winchester, B.C. Nelson, J.T. Elliott, E.J. Petersen
National Institute of Standards and Technology,
Keywords: artifact, nanoparticle, nanomaterial, toxicity, assay, C. elegans, standard
Summary:While engineered nanoparticles (ENPs) are revolutionizing some processes and products, their sustainable commercialization depends on their safe implementation. To understand their potential environmental impacts, we must understand their toxicity. However, most standard toxicity assays are designed for soluble chemicals and do not take into consideration the unique properties associated with ENPs. Because of the lack of standard nanotoxicity assays, researchers typically use assays not intended for ENPs, adapt existing assays, or design their own assay. Without careful consideration, ENPs may lead to unforeseen interactions and artifacts in toxicity assays, resulting in inaccurate or irreproducible results. This has hindered efforts to understand the potential environmental consequences of ENP release. Here, we detail our efforts to adapt an International Organization for Standardization (ISO) C. elegans toxicity assay for use with ENPs, including a cause-and-effect analysis to identify potential sources of error followed by a sensitivity analysis to quantify the error. We found that shaking plates during the assay decreased growth by as much as 36 % and changing the concentration of feed in the assay greatly altered the toxicity of the positive control, benzylcetyldimethylammonium chloride (BAC-C16). We also found that three commonly used C. elegans media formulations produced similar results in the toxicity assay, allowing for flexibility to help reduce agglomeration of ENPs. Additionally, in our testing we discovered that positively charged polystyrene nanoparticles (PSNPs) elicit toxic effects on C. elegans. However, those impacts were due to an interaction with E. coli that is used as feed in the assay and impacts on growth were more variable for PSNPs (52 %) compared to BAC-C16 (9 %). E. coli and positively charged PSNPs formed heteroagglomerates, leading to large particles that the nematodes cannot ingest. This, in turn, reduces growth and reproduction. We repeated this test using Au ENPs with surface coatings ranging from positively to negatively charged. Positively charged Au ENPs caused toxic effects (nematode growth inhibition) and showed heteroagglomeration with bacteria, however the neutral and negatively charged Au ENPs did not cause toxicity or show heteroagglomeration with bacteria. To confirm our hypothesis that the impacts seen were due to an indirect effect on the E. coli, we performed toxicity assays with all ENPs in axenic (E. coli free) medium and found little to no toxic impact from the PSNPs or from any of the Au ENPs. This suggests that caution should be used with this assay as a result of a potential bias from ENP heteroagglomeration with the bacterial food source.