Based on published studies we hypothesise that CDNP have their effects through common pathways that produce inflammation and that oxidative stress is the lead effect driving the adverse health effects. Table 1 shows the CDNP studied in this paper along with a description of their origin and reported health effects in animals and humans.
There is considerable mechanistic data describing the molecular events flowing from the deposition in the lungs of the different CDNP under discussion here and this is outlined in Figure 2. Figure 2 shows that different components of different CDNP can cause oxidative stress that acts through well-documented redox-sensitive pathways, such as MAPK and NF-κB, to cause inflammation. Although the components that mediate these effects differ greatly between the different CDNP, there is commonality through their ability to cause oxidative stress and inflammation.
Diesel exhaust particles
As described above DEP causes inflammation in rat lungs [28
] and in human lungs [60
] following short-term, high level exposure. Evidence of the oxidative properties of DEP in vivo
is shown by increased level of 8 OH dG, the oxidative adduct of hydroxyl radical, in the lungs of rats following exposure and in cells in culture treated with DEP [61
]. DEP causes oxidative stress in a number of models in vitro
such as oxidation of low density lipoprotein (LDL) [63
] and in exposed epithelial cells [9
]. The component responsible for the oxidative stress and subsequent pro-inflammatory signaling is principally the organic fraction [9
], although transition metals may also be involved [37
]. The organic fraction either contains, or can be metabolized to, species such as quinones that can redox cycle in cells to generate reactive oxygen species [17
Activation of signaling pathways for pro-inflammatory gene expression is seen in a number of studies using DEP; these include MAPK activation [30,67-69] and NF-κB activation [30,70]. As would be anticipated, activation of these pathways culminates in transcription of a number of pro-inflammatory genes such as IL-8 in epithelial cells treated in vitro  and in human lungs exposed by inhalation . TNFα has been reported to be increased in macrophages exposed to DEP in vitro  and IL-6 is released by primed human bronchial epithelial cells exposed to DEP . Increased expression of the GM-CSF gene is reported in human epithelial cells exposed to DEP; in humans exposed to short-term high levels of DEP similar to those encountered in a busy garage, bronchial biopsies showed increased GROα and RANTES expression in the bronchial wall .
Nanoparticulate Carbon black (NPCB)
As described previously, at high exposures carbon black causes overload tumours in rats [27
]. NPCB causes inflammation and the onset of rat lung overload tumours at a lower lung mass burden than larger, respirable CB [76
]. This reflects the reliance of rat lung overload on the particle surface area burden [25
], which is much greater for a given mass of NPCB than the same mass of non-NP respirable CB. Even at low lung burden NPCB showed evidence of mild pro-inflammatory effects whilst respirable CB did not [47
]. Similar greater inflammogenicity of NPCB than respirable CB has been described with instillation models, [11
Reactive oxygen species production has been measured with NPCB using in vitro cell-free systems [78,79] and oxidative stress has been demonstrated in exposed cells [48,80]. The chemical basis of the ability of NPCB to cause oxidative stress  is unknown, but unlike highly soluble welding fume, ROS production is not related to metal or any other soluble component . In a cell free system the NPCB particles and similarly polystyrene NP, induce ROS production , suggesting that the surface reactivity is sufficient. However, in cells this ability may also be related to increased influx of extracellular Ca++ ions seen with NPCB . Oxidative stress raises intracellular calcium by increasing release of Ca2+ from the endoplasmic reticulum, by enhancing ingress of Ca2+ through the plasma membrane calcium channels and through inhibition of Ca2+ transport out through the ATPase pumps in the plasma membrane . Oxidative stress caused by NPCB is translated into activation of NF-κB and IL-8 gene expression in epithelial cells in vitro , while both oxidative stress and calcium are implicated in activation of AP-1 and TNFα production in macrophages . A recent study reports that NPCB causes oxidative stress-mediated proliferation of airway epithelium, involving the Epidermal Growth Factor Receptor and the ERK cascade .
Exposure to welding fume nanoparticulate in humans is associated with inflammatory cytokine increases in the bronchoalveolar lavage (BAL) [86
] and systemic oxidative stress [43
]. The ability of welding fume to generate free radical is abundantly clear, even in a cell-free environment with only H2
acting as a reductant [42
]. Rats exposed to welding fume show marked pulmonary inflammatory responses [42
] and lipid peroxidation indicative of oxidative stress [42
]. In a comprehensive study of the molecular signaling pathways leading to inflammation with welding fume, McNeilly et al demonstrated that the pro-inflammatory effects of welding fume in vitro
] and in vivo
] were entirely driven by oxidative stress arising from the soluble transition metal component. Epithelial cells treated with welding fume or the soluble transition metals from them showed oxidative stress leading to MAPK-dependent (manuscript in preparation) NF-κB and AP-1 activation leading to IL-8 gene transcription [10
]. For welding fume nanoparticles, therefore, the soluble transition metals appear to be the primary mechanism of oxidative stress and inflammation.
Coal fly ash
In the past coal fly ash has been shown to have relatively low toxicity, for example lower than coal or quartz [54
]. Recently there has been increasing interest in the ability of CFA to release bioavailable iron which can redox cycle to produce oxidants [55
]. One study showed that the ability of a CFA standard to induce IL-8 release from epithelial cells was dependent on size, with the smallest size fraction (92
]; this was due to the fact that the bioavailable iron was concentrated in this fraction. There was no attempt in this study to collect a nanoparticle fraction. In a study especially relevant to our review of CDNP, Gilmour et al [12
] demonstrated that the nanoparticulate fraction of sub-bituminous coal was much more potent than any other fraction in causing lung inflammation and cytotoxicity in vitro
, when compared on a mass basis. This was not obviously linked to enrichment of Fe or any other toxic metals in the nanoparticulate fraction. Electron microscopic examination of the nanoparticulate fraction of coal fly-ashes from bituminous and low-rank coals showed abundant discrete crystalline particles rich in Fe, Ti, and Al crystalline phases down to 10 nm in size whilst low-rank samples contained considerable amounts of alkaline-earth element aggregates in the form of phosphates, silicates, and sulfates and mixed species. Importantly, all coal fly-ash samples exhibited carbonaceous particles in the form of soot aggregates with primary particle size typically between 20 and 50 nm sometimes mixed or coated with multi-element inorganic species [93
]. It seems possible that the soot particles were an important component in driving the adverse effects in ways analogous to the effects of diesel soot and NPCB.
There are no further studies on the ability of the NP fraction of CFA to cause oxidative stress or signal for inflammatory gene expression but such studies are warranted and we would predict that, along with the other CDNP discussed here, the pathway shown in Figure 2 would be activated, leading to inflammation.