The mechanisms by which transport-related air pollution induces respiratory morbidity have also been investigated. Inhaled particles encounter the epithelial lining fluid, which contains antioxidants that may alter the effects of these particles. The antioxidants are depleted in the presence of at least some particles, though this has not yet been shown for urban-air particles (Zielinski et al., 1999). As with many other particles, CAPs induce oxygen radical-mediated lesions in cell-free DNA and intact cells, as measured by different methods (Donaldson et al., 1997; Smith & Aust, 1997; Prahalad et al., 2001; Knaapen et al., 2002; Shi et al., 2003). For PM10 or finer particles collected in different locations, non-enzymatic and enzymatic antioxidants reduced the formation of oxidants in most samples. Some studies showed that trivalent cations are associated with the oxidative effect (Upadhyay et al., 2003). In addition to metal ions, organic compounds (such as semiquinone radicals derived from the PM2.5 fraction) seem to be able also to contribute to redox (oxidation—reduction) cycling (Dellinger et al., 2001). Such compounds would be expected on combustion particles, including diesel particles. Two studies showed that the coarse fraction was more potent in damaging DNA than was the fine fraction, but antioxidants ameliorated the particle-induced effect (Greenwell et al., 2002; Shi et al., 2003). Both soluble and insoluble metals seemed to contribute to the formation of radicals. Samples from Hettstedt and Zerbst, Germany, varied in potency over different weeks of sampling, indicating variations in emissions and PM composition over time (Shi et al., 2003). Also, CAPs induced oxidative DNA damage in epithelial cells. The DNA damage was much greater in the presence of particles of residual oil fly ash than in the presence of the CAPs tested (Prahalad et al., 2001). In line with the temporary variations of oxidative potential in the Hettstedt and Zerbst samples, different PM10 particles from Mexico City exhibited spatial variation ofconcen-tration-dependent DNA damage. The particles from the southern part ofthe city were the least potent, while CAPs from the northern part were the most potent (Alfaro-Moreno et al., 2002). Particles from all regions of the city were found to induce apoptosis in different types of cells. In macrophages, the apoptotic effect ofstandardized reference material (SRM) 1648 (St Louis particles) seemed to be mediated by the activation of scavenger receptors, not by the soluble fractions of the particles (Obot et al., 2002).
Pro-inflammatory mediators are involved in the development of inflammation, which is an important factor in many diseases. Different epithelial cells and macrophages exposed to ambient particles up-regulate ribonucleic acid (RNA) levels, protein release ofpro-inflammatory mediators or both; the pro-inflammatory mediators include IL-8, tumour necrosis factor alpha (TNF-a), IL-1ß, IL-6, monocyte chemoattractant protein 1 (MCP-1), granulocyte-macrophage colony-stimulating factor (GM-CSF) and ICAM-1 (Stringer et al., 1996; Kennedy et al., 1998; Fujii et al., 2001; Soukup & Becker, 2001; Huang et al., 2003). The effects of the particles seem to be related to soluble factors in some cases (Huang et al., 2003; Kennedy et al., 1998), but not all (Fujii et al., 2001). In some studies, the coarse fraction elicited a stronger response than the fine fraction. This response was most prominent with the insoluble coarse fraction and was to some extent attributable to endotoxin. Another study found a response to both endotoxin and soluble metal (Bonner et al., 1998). In other cells, however, the PM1.0 fraction elicited a stronger response than did the larger fractions (Huang et al., 2003). Both the coarse and fine fraction of CAPs collected near a busy highway in Downey, California decreased the ratio of reduced-to-oxidized glutathione in macrophages. Also, haem oxygenase 1 was induced by the particles, and this response seemed to depend on PAHs, rather than on metals (Li et al., 2002). Long et al. (2001) compared CAPs from the Boston area, sampled indoors and outdoors; they found that endotoxin appears to play a significant role in eliciting cytokine release, but other components may also be involved. Also, particles in indoor air tended to be more potent than those in outdoor air.
The in vitro data on combined exposure to microbial factors and particles do not render a clear picture. Particles still might increase an inflammatory response to microbes to a degree that damages the lung cells or inhibit the inflammatory response, and thus facilitate microbial attack.
DEPS sampled from different engines and the standard diesel particles, SRM1650, have been used in in vitro studies, such as those of Steerenberg et al. (1998), Takizawa et al. (1999) and Boland et al. (2001). Boland et al. (2001) observed that the SRM1650 particles and their own DEPs elicited similar effects on airway epithelial cells, while DEPs from an engine with an oxidation catalyst seemed less toxic. DEPS induced cell death in normal human bronchial epithelial cells, which were more sensitive to them than the other cell types tested. The cytotoxicity of DEPs increased with decreasing glutathione content in the cells. Antioxidants, metal chelators and inhibitors of nitrogen oxide synthase reduced DEP cytotoxicity (Matsuo et al., 2003). A comparison of the effects of DEPs indicated that epithelial cells were less protected against oxidant damage than macrophage cells.
The organic fraction of DEPs has induced more cell death in epithelial cells than in macrophages. In macrophages, this response was partly reversed in the presence of the antioxidant N-acetylcysteine (Li et al., 2002). Both the aromatic and the polar fraction appeared to contribute to the response. PAHs, which are abundant on DEPs, increase oxidative stress in several different cell types (Burchiel & Luster, 2001; Garcon et al., 2001). Benzo[a]pyrene increased the oxidative DNA damage induced by ultraviolet light in two different mammalian cell types (Shyong et al., 2003). On the other hand, neutrophils amplified the formation of benzo[a]pyrene-DNA adducts in human blood neutrophils (Borm et al., 1997). PAHs were found to induce apoptosis in lymphocytes, thus exerting an immunosuppressive effect (Yamaguchi et al., 1997; Page et al., 2002). Apoptosis induced by benzo[a]pyrene metabolites seemed to depend on the activation of a receptor and induction of a PAH-metabolizing enzyme. In addition, protein kinases involved in both survival and death pathways inside the cell were activated (Chen et al., 2003; Solhaug et al., 2004). In another study, the apoptosis induced by PAHs was found to be distinguishable from clonal deletion, since some signal proteins involved in clonal deletion were not activated (Ryu et al., 2003). Suppression of mitogenesis of lymphocytes and inhibition of differentiation of monocytes to macrophages have also been reported. These effects seemed to be receptor dependent (Davila et al., 1996; van Grevenynghe et al., 2003).
DEPS and organic compounds from them have elicited the release of proinflammatory cytokines (IL-6, IL-8, GM-CSF IL-1P and eotaxin) from different types ofepithelial cells and macrophages (Ohtoshi et al., 1998; Steerenberg et al., 1998; Boland et al., 1999; Bonvallot et al., 2001; Li et al., 2002; Takizawa et al., 2003). In contrast, macrophages (in BALB/c mice) and monocytes (RAW264.7) exposed to DEPs (300 ^.g/m3) exhibited reductions in protein or RNA levels of TNF-a and IL-12, with no changes in IL-18 (Saito et al., 2002). Li et al. (2002) found that the effect of the extracts depended on the induction of metabolizing enzymes and that the concentration-dependent changes in IL-8 production were modulated by the induction of apoptosis in the epithelial cells. Bonvallot et al. (2001) demonstrated that GM-CSF production was elicited most strongly by the organic fraction of DEPs, while stripped DEPs exhibited only a small effect. The effect appeared to depend on a ROS-sensitive signal pathway. Upon exposure to benzene extracts of DEPs, an immortalized human bronchial epithelial cell line, BEAS-2B cells, produced increased amounts of IL-8 RNA and protein. Also, this study indicated the involvement ofthe transcription factor NF-kB and ROS (Kawasaki et al., 2001). A DNA microarray analysis revealed the increase offour oxidant defence-related genes in macrophages exposed to extracts of DEPs. An increase in enzymes possibly related to DNA repair was also noted (Koike et al., 2002).
The finest particles in ambient air, the ultrafine particles, have received attention only recently. Ultrafine CAPs from Los Angeles were found to generate ROS in epithelial cells and macrophages and to induce haem oxygenase, an enzyme involved in defence against ROS. The ultrafine particles and, to a lesser extent, the fine ones localized to mitochondria, where they might cause further damage (Li et al., 2003). In the Netherlands, ultrafine particles (50 |ig/ml) induced considerably less IL-6 release from macrophages than the coarse fraction, and less than the fine fraction. Ultrafine particles did not affect CD11b expression, yeast-induced oxidative burst and phagocytosis, while these functions were reduced in the presence of the coarse fraction and, to a lesser extent, the fine fraction. The effects ofthe coarse and fine fractions seemed to be partly mediated by endotoxin (Becker et al., 2003). Similar samples from Bilthoven, the Netherlands induced a concentration-dependent increase (up to 400 |g/ml) in IL-8 and IL-6 release from A549 cells. Although there was no significant difference between the size fractions' ability to elicit IL-8 release, the ultrafine particles were most potent in inducing IL-6 release. At higher concentrations, the coarse and ultrafine particles were apparently more toxic than the fine fraction. Surprisingly, ultrafine particles were not able to induce cytokine release from primary rat type-2 cells, in contrast to the coarse and fine fraction or St Louis dust (Hetland et al., 2004). In New York, human bronchial epithelial cells responded to ultrafine ambient particles (up to 100 |g/ml) with an increased release ofGM-CSF (Reibman et al., 2002). The ultrafine particles appeared to exert a stronger effect than the larger size fractions, and the effect varied with collection period. Activation of protein kinases involved in survival and death signalling seemed necessary for the increased release of GM-CSF. Another study, using ultrafine carbon and epithelial cells, described increases in certain transcription factors and the involvement of factors associated with apoptosis (Timblin et al., 2002). Stone et al. (2000) observed that the ultrafine carbon effect was related to the influx of extracellular ionic calcium.
In some areas, the use of studded tyres leads to greatly increased abrasion of the road pavement, which results in substantial increases in PM10 and a much smaller increase in PM2.5. Most of the abrasion-generated PM consists of mineral particles (Hetland et al., 2000). Such particles include a variety of different minerals, such as quartz and amphiboles. Hetland et al. (2000) and Becher et al. (2001) showed that a mineral type, such as plagioclase, had little potential to induce the release of pro-inflammatory cytokines in different human and rat epithelial cells and macrophages. In contrast, stone types, such as mylonite and gabbro, and PM from a tunnel in which the pavement consisted of these stone types, were very efficient in eliciting pro-inflammatory responses. Though some ofthe minerals in these stones were rich in metals and produced some ROS, these factors could not explain the differences in inflammatory potential (Hetland et al., 2001). Thus, some pavement abrasion particles may elicit inflammation in the lungs. Other known components ofdust generated by road transport are tyre debris, including latex, and vehicle wear particles, but no in vitro information on them is available.
The antioxidant defence system of such compounds as certain vitamins and the radical-removing enzymes, such as superoxide dismutase, may be important markers of susceptibility. The growing evidence of the involvement of ROS in particle effects corroborates this notion. ROS, such as those generated by particles, might exert their effects at the cell surface by lipid peroxidation, through activation of nicotinamide adenine dinucleotide phosphate hydrogen oxidase or other enzymes, or stimulation of mitochondrial ROS production; in these two cases, the effects of ROS might be secondary to some other reactions. These reactions may involve the activation ofcertain cell surface receptors, downstream signalling through different types of protein kinases (such as tyrosine kinases and mitogen-activated protein kinases) and transcription factors (Samet et al., 1999; Sauer et al., 2001; Baulig et al., 2003; Brown et al., 2004). Secondary effects might be elicited, including autocrine effects of released mediators. The components of the surfactant are other factors that modulate the inflammatory response to particles influencing susceptibility (Hohlfeld et al., 2002-2003). Hohr et al. (2001) observed a reduced release of inflammatory cytokines, when epithelial cells and macrophages were exposed to particles in the presence of a phospholipid component of surfactant. Surfactant proteins are also deemed important for lung defence, and reduced release of these proteins would conceivably exacerbate pathological conditions in the lungs (Bridges et al., 2000; Hohlfeld et al., 2002-2003; Augusto et al., 2003). Oxidant gases (such as ozone and nitrogen dioxide) and ROS (hydrogen peroxide and ferrous chloride) have been shown to reduce the normal activity of surfactant proteins - either the antimicrobial activity of surfactant protein (SP) A/D or the reduction in surface tension by the concerted activity of SP-B, SP-C and SP-A (Putman et al., 1997; Wu et al., 2003).
Was this article helpful?