Influence of surface chemical properties on the toxicity of engineered zinc oxide nanoparticles to embryonic zebrafish

Abstract

SummaryZinc oxide nanoparticles (ZnO NPs) are widely used in a variety of products, thus understanding their health and environmental impacts is necessary to appropriately manage their risks. To keep pace with the rapid increase in products utilizing engineered ZnO NPs, rapid in silico toxicity test methods based on knowledge of comprehensive in vivo and in vitro toxic responses are beneficial in determining potential nanoparticle impacts. To achieve or enhance their desired function, chemical modifications are often performed on the NPs surface; however, the roles of these alterations play in determining the toxicity of ZnO NPs are still not well understood. As such, we investigated the toxicity of 17 diverse ZnO NPs varying in both size and surface chemistry to developing zebrafish (exposure concentrations ranging from 0.016 to 250 mg/L). Despite assessing a suite of 19 different developmental, behavioural and morphological endpoints in addition to mortality in this study, mortality was the most common endpoint observed for all of the ZnO NP types tested. ZnO NPs with surface chemical modification, regardless of the type, resulted in mortality at 24 hours post-fertilization (hpf) while uncoated particles did not induce significant mortality until 120 hpf. Using eight intrinsic chemical properties that relate to the outermost surface chemistry of the engineered ZnO nanoparticles, the highly dimensional toxicity data were converted to a 2-dimensional data set through principal component analysis (PCA). Euclidean distance was used to partition different NPs into several groups based on converted data (score) which were directly related to changes in the outermost surface chemistry. Kriging estimations were then used to develop a contour map based on mortality data as a response. This study illustrates how the intrinsic properties of NPs, including surface chemical modifications and capping agents, are useful to separate and identify ZnO NP toxicity to zebrafish (Danio rerio).

Exposure to Alumina Nanoparticles in Female Mice During Pregnancy Induces Neurodevelopmental Toxicity in the Offspring

“…Previous research on metal oxidative NPs found developmental toxicity induced by nano-silver ( Xin et al, 2015 ; Charehsaz et al, 2016 ; Ema et al, 2017 ; Pecoraro et al, 2017b ), nano-silica ( Yi et al, 2016 ), graphene oxide ( Buccheri et al, 2016 ; Pecoraro et al, 2017a ), zinc oxide ( Bonfanti et al, 2015 ; Zhou et al, 2015 ; Lee et al, 2016 ; Spence et al, 2016 ), iron oxide ( Di Bona et al, 2015 ), copper oxide ( Maisano et al, 2015 ; Ganesan et al, 2016 ; Torres-Duarte et al, 2016 ), magnesium oxide ( Ghobadian et al, 2015 ), titanium dioxide ( Scuderi et al, 2014 ; Parivar et al, 2015 ; Rollerova et al, 2015 ; Brundo et al, 2016 ), and cerium oxide ( Andersen et al, 2016 ). Moreover, studies have shown that bulk-Al, once entering the brain, persists for a very long time with a half-life ranging from 20% of the lifespan to greater than the entire lifespan ( Yokel, 2000 , 2002 ; Priest, 2004 ; Exley and House, 2011 ).…”

Computational models for the assessment of manufactured nanomaterials: Development of model reporting standards and mapping of the model landscape

“…In addition, an “ecotoxicological endpoint” is defined as an aggregation of different ecotoxicity-related endpoints such as LC 50 [19] , or percentage of mortality [28] and is found in 11 models ( Fig. 3 ).…”

“…The inventory reports 5 models for ecotoxicological endpoints. Of these, only 1 model reports algorithms predicting REACH-relevant endpoints, as the others suggest new biological metrics that integrate multiple toxicological endpoints [27] , [28] , [72] , [73] , [74] . Chen et al [19] reported a series of global and species-specific models that could be applied to predict REACH endpoints.…”

Identifying diverse metal oxide nanomaterials with lethal effects on embryonic zebrafish using machine learning

“…Here, molecular descriptors were calculated for the gas phase, neutral forms of the molecular constituents of the surface components. (This approach is similar to previous modelling studies of NBI Knowledgebase data, where, in addition to exploring the identities of the surface functional groups as variables related to biological effects, surface chemical composition was encoded using molecular descriptors computed for the gas phase molecules [28,29], with pH value-dependent ionization state reported to have been taken into account in one study [29].) In reality, these molecules are likely to be bound to the nanomaterial surface [62][63][64] and it is possible this binding could even alter the molecular structure [64].…”