The impact of Ag nanoparticles on the mortality of aquatic species: ISA-TAB compatible dataset mined from literature (2007-2016)
Several factors have been suggested as origins of NP toxicity (e.g. dissolution, agglomeration), which are correlated with NP stability and consequently their physicochemical characteristics. As a result, the detailed characterisation of NPs is highly significant as it can provide detailed information on their physicochemical properties and their potential effects (e.g. cell damage, toxicity) and has led to extensive study of NPs in an attempt to monitor potential changes of their physicochemical properties over time and their behaviour following immersion in different media. The large amount of potential characterisation parameters (e.g. 36 endpoints listed in OCED Workshop on physical-chemical characterisation of nanomaterials) in combination with a lack of total understanding of which parameters cause toxicity, has led scientists to follow different characterisation routes, which results in a discontinuity in research and difficulties in comparability of the, so far, published results. As a result, the need for a more standardised characterisation protocol is needed, which will offer the needed continuity and comparability and will allow for safer conclusions of the causes of NP toxicity to be reached. Out of the wide range of available NPs, one of the most studied is nanosilver (AgNPs), the versatility of which, its small size and high surface area make them more prone to dissolution and agglomeration, two parameters which have been accused as toxicity factors and can also be affected by other physicochemical properties.
For these reasons, among others, a database (DB) was created as part of the EU FP7 ModNanoTox project, which focused mainly on AgNPs, contained extracted data from published studies dealing with the toxicity of engineered NPs on aquatic species and with inclusion criteria based on the study meeting minimum characterisation requirements. The produced DB contains 108 studies with 170 separate studied NPs and 2,754 NP characterisation entries. Assay details add up to 240 entries, while the study outcomes concerning the biological endpoints add up to 7,612 entries. The % mortality entries for AgNPs were 1,118, which were then categorised according to biotarget and detailed analysis performed for the biotargets (D. magna, D. rerio, C. elegans, E. coli) which contained the most entries (n > 70), so as to identify the characterisation parameters which potentially contribute the most on the mortality (%) of the aquatic species. Due to lack of data linearity, analysis was performed using Categorical Principal Component Analysis (CATPCA) and was verified using Exploratory Factor Analysis (EFA). Results demonstrated that the parameters most correlated with biotarget mortality (%) were dosage, NP exposure time, surface modification, size (and hydrodynamic diameter) and dissolution, although some variance existed depending on the biotarget and can be explained by the different NP characterisation protocols used for each study, which resulted in missing data values.