Relevance of oxidative stress

The development of rat lung tumours in response to chronic inhalation of PSP is a well-established phenomenon. However, most if not all of these particulate materials are not considered to be inherently genotoxic. This lack of association implies the involvement of a secondary mechanism leading to the genetic changes necessary for neoplastic transformation of cells. Evidence shows that excessive and persistent formation of ROS and RNS is a significant prerequisite in particle induced fibrosis and also in rat lung tumour formation. During particle exposure, ROS and RNS are mainly generated from the oxidative burst of pulmonary inflammatory cells, i.e. neutrophils and AMs. The importance of inflammatory-mediated ROS/RNS for the development of secondary driven genotoxicity and subsequent mutagenesis was initially provided by Driscoll and coworkers (1995, 1997).

A comprehensive overview of potential mode of actions of particle-induced genotoxicity is discussed in Knaapen et al (2004). In this respect, it has been hypothesised that particle induced persistent inflammation in the rat lung and heretofore related release of ROS can provoke indirectly genetic damage in epithelial cells (Weitzman and Stossel, 1982; Driscoll 1995). During concurrently enhanced epithelial cell proliferation such mutations may then be fixed in dividing cells and clonally expressed. Implicit in such an inflammatory/proliferative driven mechanism is the existence of a threshold for such responses, i.e. that particle exposures not eliciting persistent pulmonary inflammation should also not pose a lung tumorigenic hazard. Driscoll et al (1996) demonstrated for the first time a dose-dependent increase of mutation frequency in alveolar epithelial cells following exposure of rats for 13 weeks against 1.1, 7.1 and 52.8 mg/m3 carbon black particulate including 3 and 8 months of recovery. Measured endpoints included mutations in the HRPT gene of alveolar epithelial cells, changes in bronchi alveolar lavage fluid markers, expression of mRNA for chemokines as well as histopathology. Whereas subchronic inhalation of 1.1 mg/m3 carbon black resulted in no detectable lung effects; especially no inflammatory responses, increased mutation frequencies in the presence of significant pulmonary inflammation and epithelial cell hyperplasia were observed immediately after the exposure period in the mid and high dose groups. Subsequent studies (Driscoll et al, 1997) revealed a relationship between the severity of pulmonary inflammation and ex vivo induced mutations by co-incubating lung lavage inflammatory cells (macrophages, neutrophils, lymphocytes) from carbon black exposed rats with rat lung epithelial cells. Increases of the HRPT mutation rate were observed when the percentage of neutrophils in the lavage fluid increase above 50%. Since this mutagenic response was inhibited in the presence of catalase as antioxidant, the authors concluded that oxidative stress in form of the generation and release of ROS and/or depletion of antioxidants may significantly contribute to this response. These findings support the discussed mechanism that inflammatory cell-derived reactive oxidants and increased cell proliferation play a key role in the pathogenesis of rat lung tumours in response to PSP and are consistent with the existence of a threshold (Oberdörster 1996, Greim et al, 2001). Since then, various additional studies investigating the mechanisms for mutagenicity induction following particulate exposures have demonstrated that reactive oxygen/nitrogen species generated during particle-induced pulmonary inflammation can oxidatively attack DNA resulting in structural alterations of the DNA and ultimately in fixed mutations (Weitzman and Gordon 1990, Risom et al, 2005, Singh et al, 2009).

Several studies in rats have supported the observation that persistent inflammation of the lung is driving particle induced genotoxicity in vivo Beside HPRT mutations in alveolar epithelial cells (Driscoll et al, 1997), also alternative genotoxicity markers like 8-hydroxy-2`-deoxyguanosine (8-OH-dG) (Gallagher et al, 2003, Seiler et al, 2004), DNA strand breaks (Knaapen et al, 2002) or micronuclei (Albrecht et al, 2004) 8-OHdG represents the best investigated oxidative and premutagenic DNA lesion, and consequently is also considered a marker of oxidative stress. Oxidative stress as generally defined by Sies (1991) is "a disturbance in the prooxidant-antioxidant balance in favour of the former, leading to potential damage". Inflammatory cells, like neutrophils, eosinophils and macrophages, possess a NADPH oxidase, which is induced during cell activation to produce superoxide anions. The influx and subsequent activation of such inflammatory cells into the lung therefore may lead to an oxidative burst which overwhelms the pulmonary antioxidative defences, resulting in oxidative DNA damage. Hence, genetic damage as well as proliferative effects to target cells (i.e. type II epithelial cells, Clara cells) can be induced through the production and release of oxidants.

This promutagenic environment provides favourable conditions for producing both, fibrotic cores and clones of mutated cells that eventually culminate in malignant lung disease if antioxidative defences, DNA-repair and selective apoptosis fail as protective mechanisms.

Based on the above findings, the tumourigenic effect of PSP in rats seems to be closely linked to the induction of persistent inflammation evoking genotoxic events and mutations as prerequisites of tumour formation. The mechanistic understanding by which such secondary genotoxic mechanisms occur following exposure to high concentrations of PSPs is well described in a NIOSH report on the health hazards of TiO2 (NIOSH, 2011) as well as in appendix R7-1 to Chapter R7a of the ECHA guidance (ECHA 2012). In contrast to rats, no other animal species, including mice and hamsters have developed lung tumours following chronic exposure to PSP. Interestingly, following intra-tracheal instillation of comparable high doses of quartz to rats and hamsters, both species exhibited elevated levels of neutrophils with changes significantly higher in rats compared to hamsters. Rats also showed greater expression of several pro-inflammatory mediators and lower levels of anti-inflammatory mediators (Carter and Driscoll, 2001). Although genotoxic events measured as 8-oxoguanine were observed in both species, increased cell proliferation was only noted in rats but not in hamsters (Seiler et al, 2001). The authors concluded that the differences in pro- and anti-inflammatory mediators may not only contribute to the higher inflammatory reactions seen in rats, but do also indicate less potential for scavenging the mutagenic effects in rats. Moreover, the increased cell proliferation which was only detectable in rats but not in hamsters, may contribute to the observed species differences in sensitivity to tumorigenic effects of PSP.

Ziemann et al (2011) have investigated the in vivo induction of genotoxic effects of fine versus ultrafine particles in the lungs of rats following repeated intra-tracheal instillation of crystalline silica (DQ12, 1300 nm), amorphous silica (Aerosil® 150, 14 nm) and carbon black (Printex® 90, 14 nm) for 3 months. After immunohistochemical quantification, the results of the measured genotoxicity biomarkers poly-ADP-Ribose (PAR), 8-hydroxy-2`-deoxyguanosine (8-OH-dG), 8-oxoguanine DNA (OGG1) and gamma-H2AX were indicative of distinct genotoxic stress, the occurrence of DNA double strand breaks and oxidative DNA damage in lung epithelial cells. Comparisons of these results with existing instillation data from a 3 month BAL study and life time carcinogenicity assay with the same particles, revealed also a good correlation between the biomarker results with the tumour potency data and the histopathologically scored pulmonary inflammation. The authors concluded that the findings after subchronic instillation of these particles are consistent with the picture of an overload phenomenon in the rat lung in that chronic inflammation in the lungs is leading to a persistent exposure of lung epithelial cells to released ROS, oxidative stress and subsequent oxidative DNA-damage including cell death (cell proliferation), double strand breaks, mutations and finally tumour development in the rat lung.

More recently, Gebel (2012) investigated differences as to the pulmonary carcinogenic potency of nanomaterials versus fine dusts. Based on the assumption that inflammatory responses are driving the development of pulmonary tumours in the rat and especially because the particle, whether nanosized or microsized, is considered the toxicological principle as scientifically accepted common mode of action, a meta-analysis of existing rat inhalation carcinogenicity studies was performed. Taking mass concentration as dose metric, relative potency factors based on cumulative particle mass following chronic exposure indicate a 4- to 5-fold higher carcinogenic potency of nanosized particles compared to microsized particles. However, these analyses have not taken into account that tumour induction is age dependent and that particle-induced pulmonary tumours in the rat appear late, generally later than 24 month after start of the exposure (Mauderly et al, 1987). Taking into account that the median study duration for nanosized particles was 4 months longer than for micromaterials, Gebel (2012) therefore made adjustments for exposure duration as well as for total study duration to consider the age-dependency in tumour formation. Based on his results, the relative carcinogenic potency differences between nano and micro materials is small and indicate only a 2- to 2.5 fold higher potency compared to the 4- to 5-fold difference previously assumed (Gebel 2012).

In conclusion, recent findings have shown that chronic exposure to high concentrations of PSP, whether in nanosize or microsize, will increase the lung burden until a steady state between deposition and clearance is reached. Above this threshold, the mucociliary and alveolar macrophage mediated clearance mechanisms of the lung become overloaded, leading to accumulation of the particles and finally to sustained pulmonary inflammatory responses. Below this lung overload threshold, particles will be removed from the lung at normal clearance rates without any appreciable adverse response. During lung overload conditions, inflammation dependent increase of oxidative stress becomes dominant resulting in secondary genotoxic events. Compared to other species, the inflammatory and subsequent pathological responses are much more pronounced in rats which indicate the special sensitivity and specificity of the rat concerning lung overload driven lung responses. Additionally, from all available data it is concluded that inflammatory processes are responsible for the pathogenic lung responses and that mechanistically no difference between nano- or microsized particles exist. Scientifically accepted is that as common mode of action the particle nature as such can be regarded the causative principle. A critical analysis of data obtained in rats after exposure to high PSP concentrations - whether micro- or nanosized with regard to their relevance for human health risk assessments is therefore crucial.