Particulate air pollutants are complex mixtures derived from a variety of anthropogenic and natural sources. Ambient particulate matter (PM) can cause effects on the respiratory system that are similar to particulates found in occupational settings; however, extrapulmonary changes (e.g., cardiovascular effects) have also been documented from peak exposures to ambient PM. While research of the effects of ambient PM over the last few decades have focused on the PM in the coarse- and fine-sized fractions, considerable efforts have recently been spent understanding the role that ultrafine particles may play in either contributing or exacerbating cardiopulmonary disease in normal and susceptible populations. It should be noted, however, particularly in the era of nanotechnology, that researchers are now referring to ambient ultrafine particles as “nanosized particles produced by incidental means” to make the distinction from engineered nanoparticles.

Dissimilar size fractions of ambient PM have very different physicochemical characteristics, partly because of their emission sources. Coarse particles are generally comprised of natural materials (i.e., minerals, silicates, pollen) derived from weathering and disturbance of earth soils, whereas fine particles usually originate from anthropogenic sources (i.e., combustion processes, industrial emissions) and are comprised of a mixture of elemental and organic carbon, sulfate, nitrate, minerals, and metals. Ultrafine PM is also generated through combustion processes, but can quickly aggregate to form larger sized particles (9–11).

Episodes of intense particulate air pollution have been known to cause increased morbidity and mortality for at least a half-century. The Donora death fog in 1948 and the London fog in 1952 were notable air pollution events that led to upwards of a few thousand excess deaths and, more recently, the work from the Harvard six cities studies in 1993 showed that peak particulate air pollution events led to increased deaths from lung cancer and cardiopulmonary disease (49). Over the last decade, over 100 studies of more than 35 different cities have investigated the acute effects of ambient PM showing increased hospital admissions and deaths from cardiopulmonary disease (e.g., asthma, chronic obstructive pulmonary disease, arrhythmia, heart attack) (50–52). The effects appear to best correlate with PM_2.5 with an increased mortality of 0.5 to 1.5% for every incremental concentration increase of 5 µg/m³ (51). These studies have served as a basis for the National Ambient Air Quality Standard for PM (10).

There is considerably less information about the chronic effects of ambient particulate pollution. Although lung tissue remodeling is known to occur from exposure to ambient PM, these changes tend to be less prevalent and severe compared to that observed in occupational settings. Autopsy studies have provided some means for assessing the potential long-term effects of exposure to ambient PM. Findings from these studies have suggested that long-term exposures to ambient PM can lead to remodeling of the lungs with the most significant changes occurring in the respiratory bronchioles and bronchoalveolar duct junctions or centriacinar regions of the lungs. Pinkerton et al. (2000), for example, evaluated lungs of deceased young males in California’s Central Valley who died of non-respiratory causes. The individuals had been exposed to ambient conditions consisting primarily of mineral and carbonaceous dusts. Dust deposition was principally observed in tissue sections at the terminal bronchioles and first-generation respiratory bronchioles, with less deposition observed in the upper airways (Figure 3). There was significant wall thickening via inflammatory cell accumulation (dust laden macrophages), increased collagen deposition, and smooth muscle cell hypertrophy, resulting in terminal and first-generation respiratory bronchiole structural remodeling. Changes in lung structure and retention of particles appeared to be inversely proportional to the distance from the center of the lung acinus; particle retention decreased from the first-generation, to second-generation, to third-generation of respiratory bronchioles (Figure 4). The effects seen in the respiratory bronchioles were also observed in the lungs of both smokers and non-smokers, with more severe changes occurring in smokers. It was suggested that there might be synergistic effects between ambient PM and cigarette smoke, but despite any interactions, respiratory bronchiole remodeling as a result of exposure to ambient PM could be detected irrespective of the smoking status (53).

The findings from Pinkerton and colleagues have also been observed in other PM exposed populations. Subjects living in high-PM areas in Canada showed increased particle deposition in the respiratory bronchioles and at airway bifurcations that correlated with airway remodeling (54, 55). These studies have shown correlations of high ambient PM exposure to particle deposition and airway remodeling in the lungs with indications that fine (≤PM_2.5) and aggregations of ultrafine particles (≤PM_0.1) are likely the particle size fraction contributing to these effects because of their prevalence in lung tissue digests (53, 56). Further studies comparing individuals exposed to non-occupational, high ambient PM have indicated that ultrafine particles retained in the bronchiole airway walls are associated with fibrogenic small airway remodeling and may produce a chronic airflow obstruction (57). Although these results may suggest possible effects from long-term exposures to ambient particulate air pollution, it is important to keep in mind that other confounders (e.g., genetic, lifestyle, smoking, occupational exposures, other ambient pollutants) may have some contribution to these effects and may not be fully accounted for in the individual case history (53, 55–59).

As a means to understand the potential chronic effects of ambient PM, researchers have utilized tracheal explants as a model system for controlling the administered PM dose to a particular region of the lungs. More specifically, it has been shown that particles (collected ambient PM, mineral dusts, and diesel exhaust PM) administered into the airway leads to expression of mediators promoting fibrosis and smooth muscle hyperplasia. The expression occurs without exogenous inflammatory cells, and suggests PM may directly cause epithelial cell injury, airway remodeling, and possible obstruction even in the absence of inflammation (60, 61). Further, studies of concentrated ambient particles (CAPs) have shown both pulmonary and extrapulmonary effects, including up-regulation of pro-inflammatory genes and markers of oxidative stress in the lungs, as well as systemic effects suggesting an increased risk of arthrosclerosis (59, 62, 63).

Oxidative stress has been identified as one of the primary mechanisms by which ambient particulate air pollution exerts adverse health effects. Of the varying sizes of ambient PM, nanosized particles are thought to be potentially the most hazardous due to the their small size, large surface area, and high relative content of redox-cycling organic chemicals with the ability to penetrate, deposit, and be retained in the deep lung (28, 31). Many of the components of nanosized ambient PM, such as metals and organic carbon compounds, are capable of ROS generation through Fenton and Haber Weiss chemistry, as well as redox cycling of organic chemicals (i.e., quinones), which form superoxide radicals (30, 31). The ability of ultrafine particles (PM_0.1) to generate more free radicals than coarse (PM₁₀) and fine particles (PM_2.5), as measured by induction of heme oxygenase and depletion of intracellular glutathione, may be due to ultrafine particles having a large surface area for adsorption of ROS-generating components (64). Inherent ROS generation is commonly associated with transition metals, such as iron (Fe), copper (Cu), nickel (Ni), cobalt (Co), and vanadium (V). These metals are associated with anthropomorphic sources, with Fe having an order of magnitude greater concentration in most samples from polluted regions (65). Transition metals such as Fe catalyze the production of hydroxyl radical from hydrogen peroxide via Fenton-like reactions that are enhanced in the presence of physiological concentrations of ascorbate (Figure 1) (66). Using metal chelators or antioxidants to ameliorate the oxidative stress induced by PM have shown that metals play an important role in the pro-oxidant and proinflammatory effects of ambient PM (31, 67).

In addition to transition metals, environmental PM contains PAHs and quinones and that undergo redox cycling to generate ROS. In the burning of hydrocarbons, radicals formed early in combustion interact, forming PAHs, including carcinogens, from less complex structures. PAHs will aggregate into nanoparticles, which can extend into branched-chain structures observed as black smoke or soot (68). Quinones are derived from PAH components and likely include such compounds as 1,4-naphthoquinone; 5,12-naphthacenequinone; bez[a]anthracene-7,12-dione; and anthracene-9,10-dione. These quinones undergo cyclic reduction reactions with oxygen followed by oxidative coupling with either NADPH or iron to form semiquinones leading to the formation of superoxide radicals (69). Quinones are not only by-products of fuel combustion, but also are generated by the enzymatic conversion of PAH in the lungs (31, 70, 71).

The toxic potential of PM in ambient air or from combustion processes has been shown to correlate with the chemical composition and the capacity to induce oxidative stress (31, 72, 73). Studies of concentrated ambient particles (CAPs) show that nanosized particles, which contain a significant amount of PAHs and quinones, have a greater potency to induce oxidative stress in macrophages and epithelial cells than coarse PM, which is mostly comprised of crustal elements (31, 72). Similarly, quinones and PAHs in diesel exhaust particles have been shown to contribute to oxidant injury of the lung (31, 74–76). It has also been suggested that oxidative stress generated from organic chemicals on the surface of combustion particles are responsible for not only proinflammatory effects, but also adjuvant effects in the respiratory tract, which can lead to nonspecific and allergic inflammatory processes (31, 73). Interestingly, pretreatment of SOD or nitric oxide synthase inhibitors significantly diminished inflammatory cell infiltration, mucus and nitric oxide production, and increased airway hyperreactivity (effects involved in the pathogenesis of asthma), emphasizing the role of oxidant stress on inflammatory processes and pathological outcomes (31, 77–79). Similarly, thiol-antioxidants (i.e., N-acetyl-cysteine) suppressed adjuvant effects of diesel exhaust particles on ovalbumin induced allergic responses (31, 80). The importance of the role of antioxidant and enzyme detoxification pathways has been further emphasized by the observation of increased nasal allergic and allergen-specific IgE responses in individuals who exhibit GST M1-null genotype following exposure to diesel exhaust particle (31, 81). In conclusion, while recent research has provided a better understanding of the potential drivers of oxidative stress from ambient PM, the complexity of particle-cellular interactions and the associated intrinsic and extrinsic sources of ROS still leaves much to be evaluated in terms of defining the role and molecular pathways by which oxidative stress leads to PM induced disease processes.

Although nanoscale particles (<100 nm diameter) resulting from manufacturing (i.e., ultrafine titanium dioxide [TiO₂] or carbon black) or combustion processes (i.e., vehicle exhaust, air pollution) have been studied extensively for decades, the potential biocompatibility and toxicity of engineered nanomaterials have only recently received attention from the scientific community. Carbon-based engineered nanoparticles, such as single-walled carbon nanotubes (SWCNTs), multi-walled carbon nanotubes (MWCNTs), and fullerenes, have received notable attention due to their superior electronic, optical, mechanical, chemical, and even biological properties. Questions have been raised as to whether the unique properties of these materials may exert biological effects distinct from their parent material (82, 83). Carbon nanotubes are hollow graphite tubes that can be visualized as a single sheet of graphite rolled to form a cylinder. Carbon nanotubes are comprised of either a single layer (e.g., SWCNT) or multiple layers of individual SWCNTs stacked within one another (e.g., MWCNT), and are manufactured either by electrical arc discharge, laser ablation, or chemical vapor deposition processes (84, 85).

The type of carbon nanoparticle (i.e., SWCNT, MWCNT, fullerene), method of processing (i.e., refined or unrefined), presence of residual transition metal catalysts, and functionality of different reactive groups are a few of the parameters researchers have tested in cultured cells to better understand which physicochemical characteristics influence toxicity (86–89). SWCNTs, for example, appear to have greater toxic effects on cultured human fibroblasts than MWCNT, active carbon, carbon black, and graphite carbon. In addition, acid treatment (a method of carbon nanotube refinement and removal of residual metal catalysts) of SWCNT produced more toxicity than its unrefined counterpart (89). These findings are supported by other studies that show that acid treatment and subsequent functionalization of SWCNT or fullerenes influence the extent of toxicity on human lung-tumor cell lines and primary immune cells (86–88). The addition of carbonyl-, carboxyl, or hydroxyl groups on the surface of carbon nanotubes induces cell death in lung tumor cells (87). Functionalization of SWCNTs with water-soluble functional groups appears to influence cellular specific uptake and tolerance by primary immune cells, whereas nonfunctionalized carbon nanotubes induce oxidative stress and apoptosis in a variety of cell systems (90–95).

While clear toxicological differences between carbon nanoparticles functionalized with different chemical moieties have been observed with in vitro cell systems, these same responses are not always seen when administering the same material in vivo. The different toxicity responses of in vitro versus in vivo studies has been specifically observed with functionalized fullerenes administered to different human cell lines (dermal fibroblasts, lung epithelial cells, astrocytes) compared to pulmonary responses of rats administered the same material by intratracheal instillation (88, 96). As for in vivo studies conducted on carbon nanotubes, most investigators report inflammation, progressive fibrosis, and granulomas in rodents exposed to carbon nanotubes via intratracheal installation or pharyngeal aspiration. More specifically, as a result of these exposures, acute dose-dependent changes in alveolar wall thickness, immune cell recruitment, and indicators of cellular damage and oxidative stress (measured by levels of inflammatory cells, cytokines, and protein in bronchoalveolar lavage) were observed (97–101). Carbon nanotubes also produce pulmonary function deficits, impairment of bacterial clearance, aortic plaques, and atherosclerotic lesions (99, 102–104).

In an attempt to understand how different physical and chemical parameters contribute to toxicological effects, researchers have evaluated the impact of the method of carbon nanotube production, as well as the influence of milling carbon nanotubes or altering the content and type of metal catalyst on the toxicity in animals (98, 105, 106). Results suggest that all of the various formulations of carbon nanotubes produce pulmonary lesions (97). A relatively recent published study shows that these effects can be exacerbated by feeding animals a vitamin E-deficient diet, and thereby reducing the antioxidant capacity (glutathione, ascorbate, α-tocopheral) of the lungs while enhancing acute inflammation and fibrotic responses (101).

A few inhalation studies of SWCNTs and MWCNTs using a variety of aerosol delivery systems have recently been published in an attempt to address whether the pulmonary effects of CNTs can be attributed to the method of administration (e.g., pharyngeal aspiration or intratracheal instillation) or the toxicity of the particle itself (107–115). Based on these recent studies, there appears to be a difference in the pattern and extent of pathology across the different types of nanomaterials (SWCNT versus MWCNT), as well as approaches for delivery to the respiratory tract (aspiration versus inhalation). MWCNTs delivered by intratracheal instillation or pharyngeal aspiration produces inflammation and fibrosis biochemically and histologically at delivered doses up to 5 mg per rat (98, 116), whereas inhalation of aerosolized MWCNTs produces mixed results. For example, no pulmonary lesions were observed following exposure to 100 mg/m³ MWCNTs for 6 hr (109) or 5 mg/m³ MWCNTs for 14 days (108). In fact, in one study, inflammatory responses in the spleen were more sensitive to MWCNT exposures than that observed in the lungs (108). In mice exposed by inhalation to 0.3, 1 or 5 mg/m³ MWCNTs for 7 or 14 days (followed for 7 and 14 days post-exposure), alveolar macrophages in bronchoalveolar lavage and in lung tissue sections contained black particles, but without elevations of white blood cell counts in bronchoalveolar lavage or oxidant stress markers or pathology in the lungs. Changes in immunosuppression markers (e.g., T-cell antibody and proliferative response) and cytokine gene expression of IL-10 and NAD(P)H oxidoreductase, however, were observed in the spleen (108).

In contrast to the negative pathological changes in the above noted studies, other researchers have reported thickening of alveolar walls following exposure to 32 mg/m³ MWCNTs for 15 days (107), lung injury and fibrosis under conditions of pre-existing allergic inflammation following exposure to 100 mg/m³ MWCNTs for 6 hr (109), inflammation and granuloma formation following exposure to 0.5 or 2.5 mg/m³ MWCNTs for 90 days or 0.4 mg/m³ MWCNTs for 13 weeks (113, 114), and subpleural fibrosis and mononuclear cell aggregates following a single 6 hr exposure to 30 mg/m³ MWCNTs (115). To evaluate the effects of inhaled carbon nanotubes, researchers have used a variety of systems, such as a nebulizer, jet mill and powder generator, to aerosolize these carbon-based nanomaterials with median mass aerodynamic diameters (less than 2 µm) within the respirable size range.

In comparison, studies of SWCNTs have reported pulmonary inflammation, interstitial fibrosis, and granulomas following exposure by instillation, pharyngeal aspiration, and inhalation (97–101, 110, 117). In two studies published by researchers at NIOSH, it has been suggested that administration of dispersed SWCNTs, either by aspiration or inhalation, increase collagen and alveolar wall thickness compared to less dispersed forms of SWCNTs delivered by aspiration (110, 117). It has been proposed that SWCNTs elicit these responses by generation of ROS mediated through the presence of residual iron catalysts decorated on the surface of the SWCNTs as well as through release of inflammatory mediators. Although it has been shown that MWCNTs are clearly recognized by alveolar macrophages (108, 109), it has been suggested that evasion of SWCNTs from macrophage phagocytosis might contribute to their facilitated translocation to the interstitium, where the production of collagen is stimulated. Whether macrophage clearance and transepithelial migration, or differences in physicochemistry, can explain the dissimilar pathological patterns in the respiratory tract between MWCNTs and SWCNTs will require further investigation.

Because of the elongated and fibrous nature of carbon nanotubes, some researchers have suggested analogies between these engineered nanoparticles and asbestos. In fact, two studies have compared peritoneal responses of MWCNTs to asbestos following intraperitoneal injection. In one study, inflammatory and granuloma responses in the peritoneal cavity were assessed following an intraperitoneal injection (50 µg/mouse) of MWCNTs or amosite asbestos (long fibers of 10–20+ µm length or short fibers of <5 µm length) or nanoparticle carbon black (118). It was found that the granulomatous inflammation was greatest for long MWCNTs and amosite formulations as compared those containing mostly short MWCNTs or amosite fibers. While the response observed with long MWCNTs and amosite fibers were similar, the study did not address whether the inflammatory or granulomatous changes would go on to develop mesotheliomas (118). Although highly criticized for its dosing regimen and animal model selection, Takagi et al. (2008) reported mesotheliomas in p53 heterozygous mice following single 3 mg intraperitoneal injections of MWCNTs or crocidolite fibers but not in mice injected with fullerenes (119). In contrast, a two-year bioassay showed no mesotheliomas as a result of single intraperitoneal injections of MWCNTs (with and without structural defects) in mice, whereas significant incidence rates of mesothelioma tumors were observed with intraperitoneal injections of crocidolite (120). Despite the findings in these three studies, further research is clearly needed to evaluate whether any pathological changes in the appropriate animal model can lead to tumors following acute exposure to CNTs via relevant routes of administration (e.g., inhalation) (118, 121, 122).

Similar to incidental ambient nanoparticles derived from anthropogenic sources, engineered nanoparticles can produce ROS and oxidative stress through intrinsic physicochemical properties of the nanoparticle, as well as through nanoparticle-cellular interactions with immune and epithelial cells in the lungs. There are a number of different proposed mechanisms by which engineered nanoparticles produce ROS (28, 31). First, electron-hole pairs though UV activation could participate in electron donor or capture interactions that generate superoxide or hydroxyl radicals. Second, semiconductor properties could lead to electron jumping from the conduction band to oxygen to generate superoxide. Third, dissolution of the nanoparticle to release metal ions can catalyze ROS generation and, lastly, transition metals on the nanoparticle surface can generate superoxide radicals via Fenton chemistry (28, 31). Also, similar to ambient nanosized particles, engineered nanoparticles can produce oxidant injury via nanoparticle-cellular interactions and ROS derived from inflammatory processes and mitochondrial dysfunction.

Given some of the physicochemical similarities (i.e., presence of transition metals, high aspect ratio, potential biopersistence) to other high aspect ratio particles (e.g., asbestos), researchers have sought to understand how the role oxidative stress integrates with some of the health effects seen in the lung following exposure to CNTs. Studies have generally focused on oxidative responses from the presence of residual transition metal catalysts associated with the production of SWCNTs and MWCNTs and/or from phagocytic cells as a result of phagolysosomal activation. When comparing as-produced SWCNTs (containing iron) to purified SWCNTs, where the iron is removed through acid treatment, it has been shown that iron-rich SWCNTs are more effective in stimulating hydroxyl radicals in zymosan-stimulated RAW 264.7 macrophages and increasing intracellular ROS and decreasing mitochondrial membrane potential in rat macrophages and human lung cells than purified SWCNTs lacking iron (16, 123, 124). While several studies have shown that the presence of transition metals (e.g., iron) is important in determining the redox-dependent responses of macrophages, CNTs have other characteristics that make them unique from other toxicants, such as asbestos. Specifically, it has been shown that CNTs may have the ability to quench and scavenge free radicals, which has been proposed to be related to the presence of structure defects and associated with genotoxic and inflammatory potential (16, 105, 125).

With the massive oxidizing potential of phagolysosomes, it is likely that oxidative modification of phagocytized CNTs may occur. Myeloperoxidase (MPO), a potent oxidant enzyme source in neurtrophils, has been shown to biodegrade SWCNTs in an acellular system (16, 126). Treatment of biodegradation products of SWCNTs with MPO failed to induce pulmonary inflammatory responses in mice after pharyngeal aspiration (16, 126). It was also observed that the clearance of SWCNTs from MPO-deficient mice was markedly reduced compared to wild-type animals (15, 16). Other carbon-based nanoparticles have been shown to undergo similar MPO-catalyzed modifications and degradation (16, 127). It has been proposed that hypochlorite and MPO reactive intermediates are possible pathways in which the oxidative biodegradation process occurs (15, 16, 127, 128). Certainly the biodegradation of SWCNTs has tremendous implications for biopersistence and potential long-term health effects from prolonged exposures.

Recent work by Donaldson and colleagues has shown that CNTs are generally durable materials but may undergo biological modification (16, 129). Testing the durability of four types of CNTs in simulated biological fluid demonstrated that durable CNTs with short or tightly bundled aggregates with no long isolated fibers were less inflammogenic than the longer CNT counterparts (16, 129). Other studies have shown that biodegradation leads to physical and/or chemical modifications. SWCNTs with carboxylated surfaces in artificial phagolysosomal fluid can reduce the length of SWCNTs, as well as result in accumulation of carbonaceous debris. These biodegradation processes certainly have implications for the hazardous potential of these materials.

It has been suggested that CNTs may induce ROS through NADPH oxidase-dependent pathways. In NADPH oxidase-deficient mice, SWCNT exposure resulted in an accumulation of neutrophils, production of pro-inflammatory cytokines, decreases in anti-inflammatory and pro-fibrotic cytokines, and less collagen deposition as compared to wild-type control mice (16, 110). Similarly, deficiency in antioxidant molecules resulted in increased sensitivity to the inflammatory effects of SWCNTs. Lower levels of antioxidants in vitamin E-deficient mice were associated with greater acute inflammation and enhanced pro-fibrotic responses (increase in TGF-β and collagen deposition) following exposure to SWCNTs (16, 101). In summary, the research thus far suggests there are multiple factors (e.g., CNT chemistry, biodegradation and antioxidant capacity/susceptibility) that can influence oxidative stress following exposure to CNTs and further research is needed to understand which variables impact or are associated with any progression of pulmonary disease.