During the last few years, research on toxicologically relevant properties of engineered nanoparticles has increased tremendously. A number of international research projects and additional activities are ongoing in the EU and the US, nourishing the expectation that more relevant technical and toxicological data will be published. Their widespread use allows for potential exposure to engineered nanoparticles during the whole lifecycle of a variety of products. When looking at possible exposure routes for manufactured Nanoparticles, inhalation, dermal and oral exposure are the most obvious, depending on the type of product in which Nanoparticles are used. This review shows that (1) Nanoparticles can deposit in the respiratory tract after inhalation. For a number of nanoparticles, oxidative stress-related inflammatory reactions have been observed. Tumour-related effects have only been observed in rats, and might be related to overload conditions. There are also a few reports that indicate uptake of nanoparticles in the brain via the olfactory epithelium. Nanoparticle translocation into the systemic circulation may occur after inhalation but conflicting evidence is present on the extent of translocation. These findings urge the need for additional studies to further elucidate these findings and to characterize the physiological impact. (2) There is currently little evidence from skin penetration studies that dermal applications of metal oxide nanoparticles used in sunscreens lead to systemic exposure. However, the question has been raised whether the usual testing with healthy, intact skin will be sufficient. (3) Uptake of nanoparticles in the gastrointestinal tract after oral uptake is a known phenomenon, of which use is intentionally made in the design of food and pharmacological components. Finally, this review indicates that only few specific nanoparticles have been investigated in a limited number of test systems and extrapolation of this data to other materials is not possible. Air pollution studies have generated indirect evidence for the role of combustion derived nanoparticles (CDNP) in driving adverse health effects in susceptible groups. Experimental studies with some bulk nanoparticles (carbon black, titanium dioxide, iron oxides) that have been used for decades suggest various adverse effects. However, engineered nanomaterials with new chemical and physical properties are being produced constantly and the toxicity of these is unknown. Therefore, despite the existing database on nanoparticles, no blanket statements about human toxicity can be given at this time. In addition, limited ecotoxicological data for nanomaterials precludes a systematic assessment of the impact of Nanoparticles on ecosystems.

Titanium dioxide (TiO₂), also known as titanium (IV) oxide or anatase, is the naturally occurring oxide of titanium. It is also one of the most commercially used form. To date, no parameter has been set for the average ambient air concentration of TiO₂ nanoparticles (NP) by any regulatory agency. Previously conducted studies had established these nanoparticles to be mainly non-cyto- and -genotoxic, although they had been found to generate free radicals both acellularly (specially through photocatalytic activity) and intracellularly. The present study determines the role of TiO₂-NP (anatase, ∅ < 100 nm) using several parameters such as cyto- and genotoxicity, DNA-adduct formation and generation of free radicals following its uptake by human lung cells in vitro. For comparison, iron containing nanoparticles (hematite, Fe₂O₃, ∅ < 100 nm) were used. The results of this study showed that both types of NP were located in the cytosol near the nucleus. No particles were found inside the nucleus, in mitochondria or ribosomes. Human lung fibroblasts (IMR-90) were more sensitive regarding cyto- and genotoxic effects caused by the NP than human bronchial epithelial cells (BEAS-2B). In contrast to hematite NP, TiO₂-NP did not induce DNA-breakage measured by the Comet-assay in both cell types. Generation of reactive oxygen species (ROS) was measured acellularly (without any photocatalytic activity) as well as intracellularly for both types of particles, however, the iron-containing NP needed special reducing conditions before pronounced radical generation. A high level of DNA adduct formation (8-OHdG) was observed in IMR-90 cells exposed to TiO₂-NP, but not in cells exposed to hematite NP. Our study demonstrates different modes of action for TiO₂- and Fe₂O₃-NP. Whereas TiO₂-NP were able to generate elevated amounts of free radicals, which induced indirect genotoxicity mainly by DNA-adduct formation, Fe₂O₃-NP were clastogenic (induction of DNA-breakage) and required reducing conditions for radical formation.

Poorly soluble particles such as TiO₂, carbon black, and diesel exhaust particles have been evaluated for their genotoxity using both in vitro and in vivo assays, since inhalation of these compounds by rats at high concentrations has been found to lead to tumor formation. Two principle modes of genotoxic action can be considered for particles, referred to as primary and secondary genotoxicity. Primary genotoxicity is defined as genetic damage elicited by particles in the absence of pulmonary inflammation, whereas secondary genotoxicity implies a pathway of genetic damage resulting from the oxidative DNA attack by reactive oxygen/nitrogen species (ROS/RNS), generated during particle-elicited inflammation. Conceptually, primary genotoxicity might operate via various mechanisms, such as the actions of ROS (e.g., as generated from reactive particle surfaces), or DNA–adduct formation by reactive metabolites of particle-associated organic compounds (e.g., polycyclic aromatic hydrocarbons). Currently available literature data, however, merely indicate that the tumorigenesis of poorly soluble particles involves a mechanism of secondary genotoxicity. However, further research is urgently required, since (1) causality between pulmonary inflammation and genotoxicity has not yet been established, and (2) effects of inflammation on fundamental DNA damage responses that orchestrate mutagenesis and carcinogenic outcome,that is, cell cycle arrest, DNA repair, proliferation, and apoptosis, are currently poorly understood.

With regard to genotoxicity testing and cancer risk assessment, particles and fibers form a rather specific group among all toxicants. First, the physicochemical behavior of fibrous and nonfibrous particles is usually very different from that of nonparticulate, chemical carcinogens. Reactive oxygen species (ROS) are believed to play a major role in primary genotoxicity of particles, which may derive from their surface properties, the presence of transition metals, intracellular iron mobilization, and lipid peroxidation. Other aspects relevant to primary genotoxicity are particle size, shape, crystallinity (e.g., silica), and solubility, and may also include particle uptake, interaction with cell division machinery (e.g., asbestos), and the presence of mutagens carried with the particle (e.g., diesel exhaust particles, DEP). Excessive and persistent formation of ROS from inflammatory cells is considered as the hallmark of the secondary genotoxicity of nonfibrous and fibrous particles. Since lung inflammation is known to occur and persist only at sufficient particle dose, this secondary pathway is considered to contain a threshold (Greim et al., 2001). Identification of (mechanisms of) particle genotoxicity has been/can be achieved via (1) acellular assays, (2) in vitro tests, (3) in vivo studies, usually in mice or rats, and finally (4) biomarker studies in humans with (occupational) exposure. The significance of acellular assays and biomarker studies for risk assessment is limited, but has provided some mechanistic insights (e.g., in oxidant generating properties of quartz and asbestos) and may also contribute to hazard identification. In vitro studies have lead to identification of primary genotoxic properties of particles, whereas recent in vivo studies provide further support for the correlation between particle-induced lung inflammation and secondary genotoxicity. Proper risk assessment of particles necessitates identification of the relative impact of primary versus secondary genotoxicity in realistic exposure conditions. However, since it is impossible to discern between primary and secondary genotoxicity with current in vivo tests, concomitant in vitro assays are required to determine primary genotoxicity. In vivo tests should ideally be designed using different doses to allow dose-effect analysis for both inflammation and genotoxicity.