Ranking of nanomaterial potency to induce pathway perturbations associated with lung responses.

Global transcriptomic responses in lungs of mice exposed for 24 h to individual multi-walled carbon nanotubes (MWCNTs, ten different types), nano titanium dioxide (nano TiO 2, nine different types) and one type of carbon black (CB) nanomaterials (NMs) were investigated using toxicogenomics tools to determine if gene or pathway dose-response modelling can be used to rank the potential of NMs to induce in vivo acute lung inflammation. In each study, female adult C57BL/6 mouse lungs were intratracheally exposed once with 18, 54 or 162 μg/mouse doses of individual NMs and control mice were exposed to vehicle only. A high dose of 486 μg/mouse was used in only one study that involved nanoTiO 2 . The pathway perturbations associated with NM features and the underlying toxicity mechanisms were identified using bioinformatics tools. Bench Mark Dose (BMD) response analysis was conducted to derive transcriptional BMD estimates for each differentially expressed gene and the associated pathways in NM-treated lungs compared to vehicle-treated controls. The resulting BMDs were used to rank the potency of different NMs to induce perturbations in pathways that mark the occurrence of acute lung inflammation in mice. The transcriptional BMDs were compared with the BMDs of an apical endpoint derived for the lung neutrophil influx in bronchoalveolar lavage fluid from the same mice, a commonly measured pro-inflammatory endpoint in vivo. The results showed that similar gene and pathway responses were induced by the NM variants investigated. Among the MWCNT types, NM-401 and Mitsui-7, the two long and rigid fibres exhibited the most potency and, the short and tangled NRCWE-026 exhibited the least potency compared to the other nine varieties. Although mixed rutile/anatase nano TiO 2 showed higher potency compared to other nano TiO 2 , owing to the large confidence intervals, a clear distinction in potency could not be made. CB behaved similarly to the less potent MWCNTs. The transcriptomic BMDs were largely comparable to the neutrophil influx BMDs, with mass-based BMDs showing better correlation for all approaches. The mass-based BMDs were more conservative compared to the surface area based BMDs. Although the NMs investigated did not show large differences in their responses that would allow a clear ranking, the concept described demonstrates that quantitative pathway responses can be used to rank the potency of NMs to induce in vivo lung inflammation