Physiologically based modelling of nanoparticle biodistribution and biokinetics

To predict the toxicity of nanoparticles (1-100 nm), it is crucial to understand their biokinetics i.e. how they are taken up, distributed, dissolved and removed from the body. Such information can be gained from biodistribution studies in animals. However, to make predictions for other types of nanoparticles, exposure conditions and species, including humans, extrapolations from such studies are required. Use of models, such as physiologically based pharmacokinetic (PBPK) models, makes extrapolations feasible, given that the models are sufficiently validated. In this thesis, a conceptual nanospecific PBPK model for intravenous administration to rats was developed and applied to different types of inert nanoparticles using experimental data from recent scientific publications (Papers I and II). The model represents systemic distribution and serves as a foundation for expansion to other species and other exposure routes (inhalation, dermal, oral). The PBPK simulations suggest that the model is able to describe the biokinetics of different types of inert nanoparticles given intravenously despite large differences in properties and exposure conditions. Our model is the first to include separate compartments for phagocytic cells and saturable phagocytosis. The simulations show that (1) phagocytosis needs to be incorporated in nano PBPK models, (2) the dose has a clear impact on biokinetics, but (3) further refinements are needed to better reflect processes such as agglomeration, corona formation and dissolution. The model was slightly modified to describe the biodistribution and biokinetics of nanoceria of different sizes and administered via other routes (Paper III). While the model could well predict the biokinetics after intravenous dosing, the predictions of inhalation, instillation and ingestion data were poor. The poor agreement may be partly due to low absorption via these routes, resulting in low nanoceria levels in tissues and organs, often close to or below the detection limit, in tissues. However, low absorption is hardly the only explanation, as the experimentally observed concentration time courses of nanoceria in tissues suggest that the biokinetics depend not only on the nanoparticle properties (size, coating) but also on the exposure conditions (dose, exposure route). The PBPK model was further developed to account for the complexity of inhalation exposure to nanoparticles (Paper IV). The modified model includes regional particle deposition in the respiratory tract, mucociliary clearance and phagocytosis in the lungs, olfactory uptake, and transport into the systemic circulation by alveolar wall translocation. The PBPK model described the biodistribution well and again suggested phagocytosis to be very important. The PBPK simulations were performed assuming that the nanoparticles are inert, i.e. do not dissolve or degrade in the body. However, when modelling the experimental data it seemed that the biokinetics might be better explained by introducing dissolution in the PBPK model. A related problem is that most experimental studies of metal nanoparticles use elemental analysis such as inductively coupled plasma mass spectrometry (ICP-MS). Such analyses do not discriminate between different forms of metal and therefore obscures the biokinetics. To test if gold nanoparticles dissolve in biological media, we developed an in vitro method to characterize dissolution of gold nanoparticles in contact with cell medium, macrophages and lipopolysaccharide (LPS)-triggered macrophages, simulating a disease state (Paper V). We demonstrated that gold nanoparticles are dissolved by cell medium and macrophages and even more so by LPS-triggered macrophages. The dissolution rate was higher for 5 nm than for 50 nm gold particles