An in vitro testing strategy towards mimicking the inhalation of high aspect ratio nanoparticles

Background: The challenge remains to reliably mimic human exposure to high aspect ratio nanoparticles (HARN) via inhalation. Sophisticated, multi-cellular in vitro models are a particular advantageous solution to this issue, especially when considering the need to provide realistic and efficient alternatives to invasive animal experimentation for HARN hazard assessment. By incorporating a systematic test-bed of material characterisation techniques, a specific air-liquid cell exposure system with real-time monitoring of the cell-delivered HARN dose in addition to key biochemical endpoints, here we demonstrate a successful approach towards investigation of the hazard of HARN aerosols in vitro. Methods: Cellulose nanocrystals (CNCs) derived from cotton and tunicates, with differing aspect ratios (similar to 9 and similar to 80), were employed as model HARN samples. Specifically, well-dispersed and characterised CNC suspensions were aerosolised using an Air Liquid Interface Cell Exposure System (ALICE) at realistic, cell-delivered concentrations ranging from 0.14 to 1.57 mu g/cm(2). The biological impact (cytotoxicity, oxidative stress levels and pro-inflammatory effects) of each HARN sample was then assessed using a 3D multi-cellular in vitro model of the human epithelial airway barrier at the air liquid interface (ALI) 24 hours post-exposure. Additionally, the testing strategy was validated using both crystalline quartz (DQ12) as a positive particulate control in the ALICE system and long fibre amosite asbestos (LFA) to confirm the susceptibility of the in vitro model to a fibrous insult. Results: A rapid (<= 4 min), controlled nebulisation of CNC suspensions enabled a dose-controlled and spatially homogeneous CNC deposition onto cells cultured under ALI conditions. Real-time monitoring of the cell-delivered CNC dose with a quartz crystal microbalance was accomplished. Independent of CNC aspect ratio, no significant cytotoxicity (p > 0.05), induction of oxidative stress, or (pro)-inflammatory responses were observed up to the highest concentration of 1.57 mu g/cm(2). Both DQ12 and LFA elicited a significant (p < 0.05) pro-inflammatory response at sub-lethal concentrations in vitro. Conclusion: In summary, whilst the present study highlights the benign nature of CNCs, it is the advanced technological and mechanistic approach presented that allows for a state of the art testing strategy to realistically and efficiently determine the in vitro hazard concerning inhalation exposure of HARN.