Respiratory health effects of diesel particulate matter


  • The Authors: Zoran Ristovski, BE Hons, PhD, is an Associate Professor of Aerosol Physics at the Queensland University of Technology and is leading a team that investigates the relationship between the physico-chemical characteristics of particulate matter and human health.

  • Branka Miljevic, BSc Hons, PhD, is a postdoctoral Research Fellow at the Queensland University of Technology and has interests in the physico-chemical characterization and health effects of ambient particulate matter and vehicle emissions.

  • Nicholas Surawski, BEnvSc Hons, MSc, is a PhD student at the Queensland University of Technology, with research experience in the physico-chemistry of diesel particulate matter emissions.

  • Lidia Morawska, BSc Hons, PhD, is a Professor of Aerosol Physics at the Queensland University of Technology and is director of the International Laboratory for Air Quality and Health, and leads a multidisciplinary research team studying various issues of air pollution and human health.

  • Felicia Goh, BSc Hons, PhD, is a postdoctoral Research Fellow at the UQ Thoracic Research Centre at The Prince Charles Hospital, and has experience in innate immunity and immunology, and has research interests in lung cancer, inflammation in COPD and genomics of air pollution responses.

  • Associate Professor Ian Yang, MBBS Hons, PhD, FRACP, Grad Dip Clin Epid, is a Thoracic Physician at The Prince Charles Hospital and is Head of the UQ Northside Clinical School, and has research interests in COPD, asthma and air pollution.

  • Professor Kwun Fong, MBBS, FRACP, PhD, is a Thoracic Physician at The Prince Charles Hospital and is Director of the UQ Thoracic Research Centre, and leads a multidisciplinary research team studying lung cancer and airways disease.


Zoran Ristovski, 2 George Street, Brisbane, Qld 4001, Australia. Email:


Particulate matter (PM) emissions involve a complex mixture of solid and liquid particles suspended in a gas, where it is noted that PM emissions from diesel engines are a major contributor to the ambient air pollution problem. While epidemiological studies have shown a link between increased ambient PM emissions and respiratory morbidity and mortality, studies of this design are not able to identify the PM constituents responsible for driving adverse respiratory health effects. This review explores in detail the physico-chemical properties of diesel PM (DPM) and identifies the constituents of this pollution source that are responsible for the development of respiratory disease. In particular, this review shows that the DPM surface area and adsorbed organic compounds play a significant role in manifesting chemical and cellular processes that if sustained can lead to the development of adverse respiratory health effects. The mechanisms of injury involved included inflammation, innate and acquired immunity, and oxidative stress. Understanding the mechanisms of lung injury from DPM will enhance efforts to protect at-risk individuals from the harmful respiratory effects of air pollutants.


Achieving good air quality is paramount in maintaining healthy human lung function, especially when one considers that an adult male typically inhales 10.8 m3 of air per day (at rest) and that inhalation is the primary exposure mechanism to toxic airborne pollutants.1 While the air we breathe is comprised mainly of gaseous compounds (i.e. 78% N2, and 21% O2),2 trace amounts of suspended particles (i.e. aerosols) that are present in air have a profound impact on human respiratory health.3 There are a variety of anthropogenic and natural particle sources present in ambient air. Examples of natural particle sources found in air include sea spray and particles produced by marine biota, terrestrial dust, volcanoes, forests and bio-aerosols (such as fungi and pollens), whereas anthropogenic particle sources include nanomaterials production, industrial processes, environmental tobacco smoke and most notably vehicle emissions.2,4,5 Given that vehicle emissions are the predominant source of fine particulate matter(PM) (PM2.5—PM with an aerodynamic diameter <2.5 µm) in the urban environment6 where most people live globally,7 we focus this review on the human respiratory health effects of PM emitted from vehicles.

Two main internal combustion engine types contribute to degraded air quality in the urban environment, namely spark-ignition engines (or petrol engines) and compression ignition (CI) engines (or diesel engines). These two internal combustion engines types differ considerably in PM emissions, where it is noted that it is much more difficult to control PM with diesel engines by virtue of the heterogeneity of the combustion process.8 As a result, hereafter, we focus this review on the human respiratory health effects arising from diesel engine particle emissions given that they emit a larger mass of PM and number of particles (typically by factors of 10–100) than their petrol engine counterpart.9

Naturally, there exists a vast body of epidemiological literature relating increases in ambient PM exposure to a range of respiratory health outcomes such as asthma, lung function decrements (e.g. FEV1), lung cancer and COPD.10–12 In their comprehensive synthesis of a range of human-based cohort studies, Pope and Dockery3 suggest a 0.6–2.2% increase in respiratory mortality risk for a 10-µg/m3 increase in ambient PM. Human-based cohort studies are therefore useful in establishing a correlation between increased ambient PM exposure and excess respiratory mortality risk, such as that associated with COPD. However, given that ambient PM is a complex mixture of a range of anthropogenic and natural sources,13 a major shortcoming of studies relating ‘whole’ PM exposure to respiratory health outcomes is that studies of this design cannot attribute health outcomes to specific sources (such as diesel engines) nor can they attribute health outcomes to specific constituents of PM (such as elemental and organic carbon). Identifying the PM sources and constituents responsible for the development of adverse respiratory health outcomes is a more beneficial approach, as it potentially enables source specific engineering measures to be implemented, consequently improving the PM constituent mixture emitted from a given source.14

In response to the above-mentioned limitations of epidemiological research, information relating PM sources and constituents to adverse respiratory health outcomes is examined in this review via two-step methodology to provide the desired information. This two-step research methodology involves firstly characterizing the physico-chemical properties of diesel PM (DPM) and then relating specific DPM constituents to cellular level mechanisms (inflammation, innate and acquired immunity, and oxidative stress) that are precursors to adverse respiratory health outcomes.


Complete combustion of a fuel containing hydrocarbons in an internal combustion engine yields only CO2 and H2O as combustion products. While CI engines offer a relatively high combustion efficiency (≥98%) in terms of the percentage of fuel that is burnt,8 the small fraction of unburnt fuel and lubricating oil yield a great number of incomplete combustion products that affect urban air quality and consequently, human respiratory health.15 DPM is a complex, multipollutant mixture of solid and liquid particles suspended in a gas.16 It is also a very dynamic physical and chemical system that exhibits very strong spatial and temporal dependency in terms of its composition.17 The composition of DPM depends on many factors, such as the level of dilution and its subsequent atmospheric processing after being emitted from the tailpipe,18,19 the engine operating condition (e.g. speed/load, injection timing and strategy), the presence of after-treatment devices (such as a diesel particle filter),20 the maintenance status of the engine,21 and the type of fuel and lubricants used.

The primary cause of DPM emissions is due to the presence of a fuel-rich mixture, characterized by a high equivalence ratio (i.e. the fuel-air ratio).9 CI engines operate on the principle of internal mixture preparation, whereby fuel is injected into the combustion chamber and subsequently has to mix with an oxidant (i.e. intake air) before combustion can commence.8 As a result, the CI engine combustion process is characterized by a great degree of heterogeneity, a process that is described as diffusion flame combustion due to the requirement of this air–fuel mixing process. Diffusion flame combustion is the primary cause of particulate emissions in a CI engine.

DPM physico-chemical properties

A recent paper by Giechaskiel et al.22 highlighted that the health impacts (or biological activity) of DPM are influenced by both physical and chemical factors; hence, it is appropriate to review the physico-chemical properties of DPM that are responsible for causing respiratory health effects in humans upon inhalation, as is done herein.

Physical properties of DPM include factors that describe its size and structure such as the mass, surface area, and number/size distribution of particles, and also their physical mixing state.23 The physical properties of DPM influence respiratory health in different ways. For example, the particle surface area influences how toxic compounds adsorb or condense upon particles'; particle size is a critical parameter which governs where DPM deposits in the human respiratory tract (see Fig. 4); particle number emissions govern the ability of particles to coagulate and therefore grow to bigger sizes; whilst the physical mixing state of particles is another factor to take into account when determining respiratory health impacts.

Figure 4.

A graph showing the relationship between various diesel particulate matter particle metrics (number, surface and mass) and the deposition of particles in the tracheo-bronchial and alveolar regions of the human lung (adapted from Kittelson et al.36 with permission).

Incomplete combustion of a hydrocarbon fuel containing trace amounts of sulphur, nitrogen and significant amounts of oxygen (in the case of biofuels) will emit a very large number of incomplete combustion products. The presence of sulphur, metals and ash from incomplete combustion of lubricating oil will also contribute to this ‘cocktail’ of chemicals emitted by the internal combustion engine. It has been estimated that diesel exhaust contains about 20 000 different chemical compounds24 with around 700 having been positively identified in the diesel emissions literature.25 Characterizing the chemical composition of DPM involves investigating the presence of broad classes of particle constituents (such as organics, sulphates and elemental carbon), the presence of metallic ash and metal oxides,26 inorganic ions,27 and also the presence of toxic compounds such as polycyclic aromatic hydrocarbons (PAH),23 reactive oxygen species (ROS),28 carbonyls29 and quinones.30

A visual representation of the physico-chemical structure and composition of DPM is provided in Figure 1.31 A diesel particle consists of many primary carbonaceous particles that agglomerate together to produce a complex, fractal-like morphology16 (see Fig. 2 for a diesel particle analysed by electron microscopy). The carbonaceous component of DPM provides a surface for other compounds (like organics, sulphates and metal oxides) to adsorb or condense upon. The organic compounds present in DPM are derived from heavy hydrocarbons (with a high boiling point) that originate from unburnt fuel and lubricating oil. Lighter unburnt hydrocarbons are present in the gas phase, and it should be noted that the organic component of DPM has the ability to partition between the gas and particle phases dependent upon the level of dilution and cooling employed during DPM sampling.32 The sulphate (such as sulphuric acid (H2SO4) and ammonium sulphate ((NH3)2SO4)) component of DPM also originates from sulphur present in the fuel and lubricating oil. Metallic ash (such as zinc oxide (ZnO) and iron oxide (Fe2O3)) can also adsorb to the DPM particle surface, with lubricating oil providing a metallic source during combustion.26

Figure 1.

The physico-chemical composition and structure of diesel particulate matter (adapted from Maricq31 with permission). (inline image) soot; (inline image) condensed HC/SO4; (inline image) nucleation mode; (inline image) imbedded metallic ash.

Figure 2.

A transmission electron microscope image of diesel particulate matter taken from a modern diesel engine.

Besides the physico-chemical composition of DPM, another important feature in Figure 1 relates to the physical mixing status of the various DPM constituents. A carbonaceous agglomerate with other adsorbed or condensed species (such as organics, sulphates or metal oxides) refers to a situation termed internal mixing. In internal mixing, the various DPM constituents mix together to form a single, incorporated particle. The presence of organic droplets (in the nucleation mode) can also be detected from Figure 1. When the DPM constituents are physically separated into distinct particle types, this situation is referred to as external mixing. The structure, or mixing state, of DPM is another important aspect to consider for assessing the respiratory health effects of this pollutant.

Figure 1 provided a fairly comprehensive sketch of DPM physico-chemical composition and structure. However, for the purposes of ascertaining respiratory health effects, this rather complicated DPM composition can be simplified somewhat, especially when one considers that diesel fuel sulphur content has decreased significantly over the last few years in Australia. Indeed, on January 1st 2009, the Australian Department of the Environment and Heritage33 promulgated a 10-ppm diesel fuel sulphur content standard, which implies that the sulphate contribution to DPM is much less than that reported in vehicle emissions studies even from just 5 to10 years ago. An article by Cowley et al.34 shows for a range of fuel sulphur conversion efficiencies (in the 1–2% range) that the contribution of sulphates to overall DPM mass is negligible (less than a few per cent) for fuel sulphur contents below 0.1% or 1000 ppm. Furthermore, the sensitivity of nanoparticle emissions (<50 nm) to fuel sulphur content is removed with ultra-low sulphur diesel35—a problem that has plagued the automotive industry for quite some time. As a result, the conceptual view of DPM (in terms of its respiratory health effects) can be simplified by ignoring the sulphate contribution—a situation that has arisen due to aggressive reduction in diesel fuel sulphur content.

A further simplification to the conceptual view of DPM is possible, and this relates to the presence of imbedded metallic ash. Electron microscopy work conducted by the authors (see Fig. 2) for investigating the morphological properties of DPM emissions showed no evidence of metallic ash particles attaching to the carbonaceous containing primary particles. Alternatively, European DPM studies conducted with diesel engines with advanced after-treatment quite often observed imbedded metallic ash in nanoparticles (see e.g. Mayer et al.26). Given that advanced after-treatment is not typically used on Australian diesel engines, it appears that metallic ash emissions from diesel engines in Australia do not replicate the problem seen in Europe.

Neglecting the sulphate and metallic ash contribution yields the DPM composition and structure that will be adopted in this review for assessing its respiratory health effects (see Fig. 3). Two features evident from Figure 3 are important from a respiratory health perspective. The first observation is that DPM in the accumulation mode (30 ≤ particle diameter (Dp) ≤ 500 nm) consists of agglomerated primary particles composed of elemental carbon with a thin organic coating. Unpublished data from the authors of this review have observed an organic layer thickness from an off-road engine fuelled with diesel fuel of 5 nm for particles with an overall diameter of 125 nm. When running the same engine on 80% biodiesel and 20% diesel fuel, we observed organic layer thicknesses of 15 nm for particles with an overall diameter of 100 nm. The second observation from Figure 3 is that depending on the organic species present and the level of cooling and dilution they undergo, liquid organic particles can be formed in the nucleation mode (Dp < 30 nm). Later in this review, DPM respiratory health effect mechanisms and health outcomes will be critically examined using the DPM composition and structure outlined in Figure 3.

Figure 3.

Diesel particulate matter composition and structure for assessing its respiratory health effects. (inline image) primary carbon particle; (inline image) adsorbed organics; (inline image) organic particle.

Metrics for characterizing DPM emissions

The previous section discussed the physico-chemical properties of DPM that are relevant to respiratory health, where mention was made of different measures, or metrics, that characterize its physical properties. The mass, surface and number of particles emitted are commonly reported metrics in the air pollution literature. Particle mass is the simplest and most commonly used metric in air pollution studies, and is merely the mass of particles emitted per unit volume of air. Particle mass emissions are referred to as PM emissions in the air pollution literature, where it is common for the analyst to collect PM below a specified aerodynamic diameter that acts as a cut-point to provide a size-resolved PM sample. Common cut-points involve 2.5 µm and 10 µm, known as PM2.5 (fine particles) and PM10 (coarse particles), respectively. Similarly, the particle surface area (S) and particle number emissions (N) represent, respectively, the surface area and number of particles emitted per unit volume of air.

Figure 4 shows particle mass, surface and number size distributions plotted with respect to particle diameter. Note very well how the modal value of each parameter shifts to larger particle diameters, as the order of the dependence on particle diameter increases. Particle number has a zeroth order dependence on particle diameter and has the smallest modal diameter of about 10 nm. Similarly, particle surface area has a second order dependence on particle diameter and has the next largest modal diameter of around 150 nm. PM emissions have a cubic dependence on particle diameter and display the largest mode of over 200 nm. The location of these modes is very important because if a different particle metric is chosen to explore respiratory health effects, then this metric is found in a different part of the size distribution. This is of particular importance for the deposition of DPM in the human lung, for example, because the number of particles deposited exhibits a very strong dependence on particle diameter,37 as can be seen from Figure 4.


Inhalation of DPM

The primary exposure mechanism to DPM (and other particle sources) is via inhalation.38 Upon inhalation, particles deposit in the human respiratory system in a size-dependent manner, as can be observed from Figure 4. The International Commission on Radiological Protection deposition curve, for combined alveolar and tracheo-bronchial deposition, exhibits two distinct peaks.39 The larger deposited fraction peak at a few micrometres entails deposition of larger particles due to their inertia, whereas the smaller peak centred around 20 nm involves deposition of nanoparticles (<50 nm) due to their high diffusivity.37 In between these two peaks, the deposition of particles is governed primarily by sedimentation.38 Of particular interest is the very high deposited fraction for nanoparticles (∼0.4–0.7), which can readily gain access to the alveolar region of the lung. Particle size is therefore a crucial parameter to take into account when assessing the respiratory health impacts of DPM.37

Mechanisms of adverse health effects from DPM

Many cohort studies have demonstrated that airborne PM, of which DPM is a major contributor,6 is responsible for causing respiratory mortality and morbidity.3 Despite a substantial body of research addressing this topic, the underlying toxicological mechanisms by which DPM induces adverse health effects are not yet entirely understood. However, in recent years, there have been a number of studies indicating that inflammation mediated via oxidative stress is the mechanism by which DPM may exert toxicity.10,40 In this section of the review, the inflammatory and oxidative stress processes will be discussed, together with the specific pathway of innate immunity, which is a major pathway through which DPM exhibits adverse respiratory health effects.


Inflammation in the lungs involves a complex set of molecular and cellular responses resulting from exposure to exogenous stimuli such as pathogens or noxious substances41—including particles42—or endogenous stimuli such as cytokines or danger signals. The pathways associated with acute inflammation in response to particle exposure involve a carefully orchestrated sequence of events, mediated in part by chemotactic molecules (chemokines).42 Upon deposition of a particle, phagocytic cells such as neutrophils and macrophages are recruited to the foreign particle by chemokines; these particles, after being engulfed, are transported by the mucociliary escalator for removal to the gastrointestinal tract.43 DPM induces the release of inflammatory cytokines, such as IL-6, IL-8, granulocyte-macrophage colony-stimulating factor and tumour necrosis factor-alpha (TNF-α)44 from immune cells (e.g. macrophages), as well as structural airway cells (e.g. bronchial epithelial cells).45 While the pathways associated with acute inflammation in response to PM exposure are reasonably well defined and understood, the relationship between chronic inflammation and the progression of respiratory disease is not.41

Immune system responses

The immune system plays a key role in mediating the inflammatory response to DPM and other air pollutants. Immune responses to air pollutants show similarities in the response to pathogens encountered in infections. In the innate immune response, structural components of pathogens necessary for the pathogen's survival and not found in the host organism (termed pathogen-associated molecular patterns or PAMPs) are recognized by pathogen recognition receptors (PRRs). PRRs are expressed on professional antigen-presenting cells (APCs) such as dendritic cells (DCs), macrophages and B cells, and include members of the Toll-like receptor (TLR), C-type lectin receptor and nucleotide-binding oligomerization domain-like receptor families.

A number of downstream mediators are important in the TLR pathway. TLRs initiate overlapping but non-identical signalling pathways, as they utilize different combinations of adaptor proteins.46,47 These adaptor proteins include myeloid differentiation factor 88 (MyD88), MyD88 adaptor-like protein, Toll/IL-1 receptor domain-containing adaptor inducing interferon-β (TRIF) and TRIF-related adaptor molecule (TRAM; Fig. 5). TLRs 2, 4 and 9 can all signal through MyD88, which leads to the translocation of the transcription factor nuclear factor-κB (NF-κB) to the nucleus after dissociation of the inhibitory IκB subunit. The release of IκB allows NF-κB, in conjunction with other factors such as interferon regulatory factor (IRF) 7, to upregulate expression of a distinct set of inflammatory genes (e.g. interferon-α and TNF-α). However, TLR4 can also signal via MyD88-independent mechanisms, while TLR3 solely utilizes a MyD88-independent signalling pathway. That is, TLR3 engages TRIF, and TLR4 engages both the adaptors TRIF and TRAM, activating IRF3 and then NF-κB,48 leading to upregulation of a different set of inflammatory genes (e.g. interferon-β). The genes upregulated by these pathways have multiple functions, including initiating antimicrobial actions, regulating cell survival and death, triggering inflammation, and priming the antigen-specific adaptive immune response.49

Figure 5.

Toll-like receptor (TLR) signalling pathways. TLR ligands initiate signalling through adaptor proteins myeloid differentiation factor 88 (MyD88), MyD88 adaptor-like protein, Toll/IL-1 receptor domain-containing adaptor inducing interferon-β (TRIF) and TRIF-related adaptor molecule according to which TLR they bind. Downstream signalling pathways involve recruitment of various factors, including interferon regulatory factor (IRF) and nuclear factor (NF)-κB, to upregulate certain inflammatory genes.

Once PAMPs are detected by PRRs and processed by APCs, antigens are displayed on the cell surface. These are then detected by naïve T cells (cells that have not encountered antigens previously), thus activating the T cells and inducing them to differentiate into effector cells that are able to perform specific immunological functions including cytokine production, direct killing of pathogen-infected cells and activation of other cell types such as B cells, macrophages, mast cells and eosinophils.50 Effector cells also recognize and react more quickly than naïve T cells to their specific antigens without the need for co-stimulation,51 allowing eradication of that particular pathogen to become more efficient as the response gains momentum.

The subset of T cells expressing the TCR co-receptor CD4 (helper T cells or Th cells) are divided into four further subsets on the basis of phenotype and function: Th1, Th2, Th17 and T regulatory (Treg) cells.52 Th1 cells are usually associated with protective responses against intracellular pathogens, whereas Th2 cells are usually associated with humoral immunity and effective clearance of large extracellular pathogens.53 They are characterized by different cytokine profiles that drive immune responses tailored against the particular pathogens encountered by PRRs.54 For example, activation of the TLR2 pathway by the synthetic ligand Pam3Cys promotes a Th2 response,55 whereas activation of the TLR4 pathway by LPS can promote both a Th1 and Th2 response,56 as TLR4 signalling can occur via both MyD88-dependent and independent pathways.

Immune response to DPM

Innate and acquired immunity are both important in the immune response to DPM. As DPM have varying physico-chemical characteristics, such as size, solubility and composition, it is likely that the effects on the immune system will vary greatly. For example, nanoparticle-rich DPM with a smaller size and increased surface area-to-mass ratio may trigger an enhanced immune response due to a greater ability to penetrate deeper into the lung.57 There is also evidence that different PRRs may be involved in the detection of DPM, including TLR2,58,59 TLR459–62 and receptor for advanced glycation end-products,63 indicating that different components may trigger different signalling pathways. Alternatively, as some TLRs have common downstream elements, it is also possible that DPM may be detected by different receptors but share similar pathways. For example, induction of cytokine responses in macrophages by fine and coarse PM is decreased in TLR2- and TLR4-deficient cells, respectively, but MyD88 is required for the response to both.64

DPM is a transport vector for pathogens, which alter the ensuing inflammatory response, as pathogens can access sites in the lung where they would otherwise not normally be found. For example, DPM affect cytokine levels induced by the bacterial product LPS65,66 and treatment of a PM fraction with polymyxin B, a protein that binds and inhibits LPS, alters the LPS-induced immune response in macrophages.64 DPM increases expression of TLR3 on human respiratory epithelial cells, which may influence downstream signalling in response to TLR3 ligands,67 and suppress ROS release from alveolar macrophages.66 Finally, expression of the lipid mediator cyclooxygenase-2 and the TLR2 ligand Pam3Cys in a human monocytic cell line was increased by co-stimulation with LPS and DPM, but not by DPM alone,68 indicating that DPM alone may not induce a significant immune response.

As exposure to pathogens, as well as PM, would be expected in an outdoor environmental setting, this may have implications in diseases such as COPD, as the impact of low-level infections may therefore have a more severe outcome than would be predicted.69 For example, levels of IL-8, a neutrophil chemoattractant,70 are elevated in respiratory diseases such as COPD71 and asthma.72 PM drives production of IL-8 through an NF-κB-dependent, IκB-independent pathway,73 indicating that the particles may contribute to the high neutrophil count in these diseases. PM also adsorbs IL-8,74 which may skew the results of some in vitro studies to underestimate the effect of air pollution on immune systems.

Indeed, DPM may promote Th2 responses specifically. DPM recruit neutrophils, monocytes and DCs to the lung and upregulate expression of the DC maturation marker CD86,75 which is involved in priming Th2 responses.76 DPM also increases DC maturation and expression of lymphocyte activation markers in murine splenocytes in vitro57 and activates DCs in vivo to drive production of the Th2-associated cytokines IL-4 and IL-5.77 Human basophil production of the Th2-associated cytokine IL-4 is increased by DPM.78 The host organism may not be the only body affected, as DPM acts synergistically with an allergen to reduce IgE levels and airway eosinophilia in adult offspring of pregnant mice.79 LPS-induced interferon-gamma production by splenic natural killer (NK) and NKT cells was inhibited by DPM.80 However, another study found that DPM induce DCs to produce a mixed Th1/Th2 cytokine profile.81 As a Th1- or Th2-promoting environment can significantly alter the response to the same pathogen and is even investigated as a possible way of exploiting immunotherapy treatments,82 it is likely that the potential for DPM to skew this Th1/Th2 balance would be of significant interest.

Oxidative stress

The induction of oxidative stress is a characteristic of exposure to DPM. Oxidative stress develops when there is an imbalance between the production of ROS and the availability of anti-oxidant defences83 (Fig. 6). ROS is a collective term that refers to free radicals such as hydroxyl (HO.) and peroxyl (HOO., ROO.), ions such as superoxide (O2-.) and peroxynitrite (ONOO-), and molecules such as hydrogen peroxide (H2O2) and hydroperoxides (ROOH). The term ‘reactive’ is used to indicate the higher reactivity of ROS relative to molecular oxygen85 due to the presence of unpaired electrons.86 Precursors of ROS such as carbon-centred radicals can also be considered as ROS. ROS can be generated directly on the surface of DPM via redox cycling processes or indirectly through interactions between DPM and cells.38 Oxidative stress initiates redox-sensitive transcription factors, such as the mitogen-activated protein kinase (MAPK) and NF-κB cascades, which work synergistically to activate expression of proinflammatory cytokines such as IL-4, IL-6, IL-8 and TNF-α, as well as chemokines and adhesion receptors.40,87

Figure 6.

An illustration of oxidant/antioxidant balance in vivo and associated biological responses (adapted from Kelly84 with permission).

In vivo and in vitro studies have reported that an increased amount of ROS is generated in cells upon exposure to DPM and other air pollutants.88 For example, exposure of 16-HBE bronchial epithelial cells to DPM induced the production of ROS, as detected by fluorescent probes, as well as gene expression of the phase I (cytochrome P-450 1A1) and phase II (nicotinamide adenine dinucleotide phosphate quinone oxidoreductase-1) xenobiotic metabolizing enzymes.89 Considerable research attention has focussed on the formation of in situ ROS, formed after particle deposition in the human respiratory tract. In situ or endogenous, ROS production can be formed by chemical species on the particle that have the potential to generate ROS (such as quinones14—a direct pathway) or by phagocytic processes initiated by the presence of DPM in the lungs (the indirect pathway). However, in addition to the particle-induced generation of ROS, several recent studies have shown that particles may also contain ROS (termed exogenous ROS).90–94

Physico-chemical properties of DPM that influence inflammation, immune response and oxidative stress

Oxidative stress and inflammation are a coupled, synergistic phenomenon. Ayres et al.10 have described that oxidative stress is a precursor to inflammation, with the occurrence of inflammation being able to generate more oxidative stress. Consequently, the physico-chemical properties of DPM influence these biological processes.

The first risk factor associated with the development of oxidative stress and associated inflammation is the DPM surface area. While PM mass is a regulated ambient air quality parameter and is also a regulated pollutant for internal combustion engines, a consistently reported result in the toxicological literature is that particle mass is not a very appropriate metric for describing the ability of particles to induce oxidative stress and inflammation (see Oberdörster et al.95,96 and references therein). For example, a rodent particle exposure study by Oberdörster97 showed that the particle surface area correlated better with the inflammatory response than did particle mass. An increased particle surface area per unit mass dose provides an increase in the availability of adsorbed toxic substances and provides a locus for which catalytic chemistry can occur, potentially leading to ROS formation. The particle surface area is also the physical quantity in contact with the lung lining fluid, so it gives a good measure of the biological activity of particles.22

Several studies have also demonstrated the importance of the organic fraction in DPM toxicity. The organic fraction of DPM is especially complex, containing hundreds (or even thousands) of compounds including PAHs, which are known human carcinogens.98 These studies have shown that DPM, as well as their organic extracts, were able to induce proinflammatory responses and/or induce apoptosis in lung tissue cells, while DPM that had their organic constituents extracted were no longer able to induce such responses in cells.99–102 Thus the organic component of DPM is implicated in the induction of oxidative stress, which is a viewpoint held by several research groups.10,84,103,104

Additionally, it has been shown that different fractions of DPM organic extracts exert different toxic effects, with aromatic and polar fractions being able to induce a proinflammatory response101 and mitochondrial dysfunction.105 The aliphatic fraction had no distinguishable effect on these cell responses. Chemical analysis of fractionated organic extracts in these studies revealed that the aromatic fraction was enriched with PAHs, while the polar fraction was enriched with quinones.

In addition to the role of organics, several studies have postulated that transition metals, such as iron and copper, are possible mediators of DPM-induced airway inflammation.106–111 Iron and copper are believed to contribute to particle-induced formation of ROS through the Fenton reaction.111,112 In the Fenton reaction, hydroxyl radicals are generated through a transition metal-mediated reduction of hydrogen peroxide. These results, however, are in conflict with the results published by other groups who have found that some cellular responses to PM do not depend on the transition metal content.113,114

Proposed mechanisms by which DPM causes respiratory health effects

To synthesize the results presented in this review related to the respiratory health effects of DPM, an attempt is made (see Fig. 7) to provide a mechanism describing pathways to respiratory illness and disease. As was discussed in the previous subsection, Figure 7 reiterates that there are three primary physico-chemical properties of DPM to consider when investigating their respiratory health effects, namely the DPM surface area, as well as the presence of adsorbed transition metals and organics. The literature consulted in this review indicates that all of these three factors are implicated in the development of oxidative stress. An important consequence related to the development of oxidative stress is that redox-sensitive signalling pathways, such as the MAPK, NF-κB and activator protein 1 cascades are activated, which can result in an inflammatory response mediated through the immune system. In this mechanism, inflammation related to PM is viewed as a potential precursor to the development and acute exacerbation of airway and lung diseases, such as asthma and COPD. Redox-sensitive signalling pathways, along with the organic fraction of DPM, are also implicated in the formation of DNA adducts, such as 8-oxo-2′-deoxy-guanosine,85 which are viewed in this mechanism as a precursor to the development of cancer.

Figure 7.

A potential mechanism, derived from Donaldson et al.40 and Duffin et al.,115 outlining the pathways to respiratory health effects upon diesel particulate matter (DPM) inhalation. AP-1, activator protein 1; NF-κB, nuclear factor-κB.

Other routes of DPM exposure: translocation

While the primary route by which DPM causes health effects is via inhalation through the human respiratory system, other particle exposure pathways are possible. Translocation is a route of exposure, whereby particles can migrate to a secondary organ (such as the brain, liver or spleen) after inhalation, thereby causing health effects in that secondary organ. In rodent particle exposure studies, a well-established link appears to be the translocation of ultra-fine particles to the brain via the olfactory bulb. The work of Oberdörster et al.116 and others shows convincing evidence of translocation of ultra-fine particles to the brain in rats and also in monkeys;117 however, much further work is required to demonstrate similar effects in humans.118 If this route of exposure is viable in humans, inhalation toxicological studies would need to consider the health effects of DPM that has migrated to secondary organs.


This review has used DPM as a surrogate for discussing human respiratory health effects upon inhalation of particles. The physico-chemical properties of DPM that are relevant from a respiratory health perspective were outlined, where it is noted that the organic carbon content of DPM plays an important role. A mechanism was proposed in this review describing potential pathways leading to respiratory illness and disease. This mechanism indicates where mitigation efforts should be directed to protect respiratory health. While this review has used DPM as a surrogate for discussing human respiratory health effects, much of the discussion could be applicable for other combustion particle types, such as those found from petrol engines, coal-fired power stations, industrial smelting processes and waste incineration. Understanding mechanisms of lung injury resulting from DPM exposure will enhance efforts to protect at-risk individuals from the harmful effects of air pollutants.


This work was supported by National Health and Medical Research Council (NHMRC) Career Development Award (IY), NHMRC Practitioner Fellowship (KF), Cancer Council Queensland Senior Research Fellowship (KF), Australian Lung Foundation/Boehringer Ingelheim COPD Research Fellowship (IY), and project grants from Australian Research Council, NHMRC, Queensland Health Smart State, Australian Coal Association Research Program, Institute of Health and Biomedical Innovation, The Prince Charles Hospital Foundation and Asthma Foundation of Queensland.