Air pollution is a complex mixture of gases and particulates that varies spatially and temporally, making attribution of adverse effects to specific individual air pollutants a challenge. To disentangle effects of mixtures in a controlled setting, reproducible and realistic co-exposures of human-relevant models to gaseous and particulate pollutants are needed. Although air-liquid interface (ALI) exposures offer considerable promise as non-animal models for inhalation toxicity testing, a lack of studies comparing individual and co-exposures to gaseous and particulate pollutants has thus far prevented assessment of their strengths and limitations for disentangling effects of pollutant mixtures.

Using an integrated ALI exposure system, we characterized the interaction between ozone and particles (25 nm fluorescent polystyrene beads) to assess effects on and reproducibility of critical physical endpoints including temperature, humidity, ozone concentration and particle deposition. Particle deposition and concentration were assessed via three independent methods: fluorescence, quartz crystal microbalance (QCM), and airborne particle count. To evaluate the acute biological effects of an air pollutant mixture in vitro, human lung type 2 epithelial-like cells (A549) were exposed at the ALI to air, ozone (O3), particles, and O3 + particles (co-exposure) for 1 h (n = 4 independent repeats/exposure type). Cell injury and inflammation were quantified by extracellular lactate dehydrogenase (LDH) activity and release of proinflammatory cytokines (interleukin (IL)-8 and IL-6) respectively 0 and 24 h post-exposure.

Exposures were effective at delivering targeted O3 exposures under controlled temperature and relative humidity. In-well particle deposition and airborne concentration exiting the exposure system, quantified through parallel methods, were consistent, and increased in relation to aerosolized particle concentration. Levels of each pollutant were effectively maintained in the presence of the other. O3 alone, and co-exposure to O3 and particles, increased LDH release from A549 cells, indicating pollutant-specific cytotoxicity. In contrast, IL-8 and IL-6 release (24 h > 0 h) were not changed by exposure to the individual pollutants, but tended to increase following co-exposure.

The present work establishes the utility of ALI exposure systems to disentangle individual effects of pollutants from a mixture, and highlights the importance of direct experimental characterization of dosimetry and exposure conditions.