Ultra fine particles (UFP) emitted from diesel engines induced oxidative stress responses in endothelial cells [20, 37]. In this study, we demonstrate that UFP emitted from an idling diesel truck (UFP1) versus the same truck under UDDS driving cycle (UFP2) induced differential levels of oxidative stress and pro-inflammatory responses in human aortic endothelial cells. UFP2 contained a higher level of water soluble organic carbon (WSOC) and consistently higher levels of organic compounds, including PAHs. Both UFP1 and UFP2 induced oxidative stress in HAEC. However, only UFP2 induced a pro-inflammatory response via NF-κB-mediated gene expression, and subsequently, monocyte binding to HAEC.
UFP1 and UFP2 had different levels of chemical composition; i.e., water soluble organic carbon and organic compounds on a per PM mass basis (Fig. 1). Water soluble organic carbon was reported to be closely associated with the consumption rate of dithiothreitol (DTT), a molecular assay that has been used as a surrogate measure for oxidative potential . UFP2 contained a higher level of WSOC content (51.5 mg/g of PM) compared to UFP1 (43.4 mg/g of PM). However, our data suggested that UFP1 was a more potent inducer of oxidative stress than UFP2 in the context of lower organic carbon compounds (Figs. 2 and 3). Organic carbon refers to a complex mixture of organic compounds with various functional groups. Of particular interest are organic tracers of vehicle emissions such as PAHs, hopanes and steranes [1, 2]. Li et al.  demonstrated that particle phase PAHs triggered oxidative stress in the cells via the formation of quinones. Hopanes and steranes are mostly found in lubricating oil and they are considered to be important biomarkers of fuel emissions [40, 41]. Although the mechanisms by which these compounds cause health hazards are largely unknown, these organic species, along with other oil-derived components, have been reported to contribute to the inflammatory effects of inhaled emissions . The per PM mass PAH content was consistently higher in UFP2 than in UFP1, and the average ratio (UFP2 to UFP1) was 5.5 ± 3.2, with the highest ratio of 11.7 observed for benzo (b) fluoranthene. Although our data suggest that higher level of organic compounds, such as WSOC and PAHs species in the total mass of UFP2 may account for its pro-inflammatory effects, additional studies need to be done to confirm the roles of individual compound or group of compounds.
Trace elements and metals are also important constituents of diesel engine emissions. While the sum of these species does not represent a substantial fraction of the PM mass, most of them are reported to promote free radical-based reactions in cellular or non-cellular bioassays . As shown in Fig. 1B, Ca, Zn, P and S were the most dominant species in both UFP1 and UFP2. These elements are primarily ingredients of lubricant oil and sulfur containing fuels . Higher fraction of metals and trace elements were found in UFP2 than in UFP1. These metal and trace element profiles displayed a different trend compared to organic compounds. The mass ratio of redox active transition metals (e.g. Iron, Chromium and Nickel) was 4.3 ± 1.9-fold higher in UFP2 than in UFP1. These metals may contribute to the pro-oxidant properties of UFP and up-regulation of pro-inflammatory gene expression . Certain elements and metals have higher mass fractions in UFP1 compared to UFP2, notably Sodium, Aluminum, Phosphorous, Potassium, Manganese and Zinc. Mass fractions of Lead and Cadmium in UFP1 were 35-fold and 17-fold higher than in UFP2 respectively. Because of their ability to react with the function of thiol groups, the higher levels of Lead and Cadmium may be partially responsible for the higher oxidative stress induced by UFP1 than UFP2.
Oxidative stress is an emerging hypothesis to particulate matter (PM)-mediated cardiovascular diseases [45, 46]. Exposure to UFP promoted atherosclerosis via systemic oxidative stress in ApoE-null mice . In our study, both UFP1 and UFP2 induced a significant increase in superoxide production and anti-oxidant gene expression (Fig. 2), consistent with our previous report . Exposure to different dose of PM is reported to induce hierarchical oxidative stress ; namely, a low dose exposure leads to antioxidant responses, whereas a high dose leads to a pro-inflammatory response. When HAEC were treated with the same dose of UFP1 and UFP2, only UFP2 stimulated the pro-inflammatory responses (Fig. 3 and Fig. 4,). However, UFP1 induced a higher level of oxidative stress response despite the lower average mass ratio of the redox active metals, such as Iron, Chromium and Nickel, in the total mass of UFP1 (Fig. 2). In this context, the hierarchical oxidative stress hypothesis alone may be insufficient to explain the differential effects induced by UFP1 and UFP2. Rather, the difference may also be due to the higher mass fractions of WSOC and PAHs species in the total mass of UFP2. The pro-inflammatory properties of UFP2 observed via NF-κB activation may be attributed to its enriched contents of organic species.
Pro-inflammatory oxidants such as Ox-LDL and Ox-PL, activate endothelial cells to express chemokines and adhesion molecules, and subsequently, monocyte recruitment to the endothelial cells . Our study revealed that UFP, analogous to Ox-LDL and Ox-PL, also induced similar pro-inflammatory responses, suggesting a possible mechanism for the in vivo observation that UFP exposure promoted atherosclerosis in ApoE-null mice .
NF-κB is a major signal pathway that conveys pro-inflammatory responses to inflammatory stimuli such as TNF-α and endotoxin [35, 36]. Some studies suggest that NF-κB is implicated in the pro-inflammatory effects of PM. For example, exposure to diesel exhaust particles (DEP) activated nuclear translocation of NF-κB in human bronchial epithelium. Isolated rat capillaries exposed to DEP released TNF-α, a potent activator of NF-κB . Our data demonstrated that UFP2, but not UFP1, activated the NF-κB signal pathway. UFP2 also induced inflammatory chemokines expression such as IL-8, MCP-1 and adhesion molecule such as VCAM. The NF-κB inhibitor, CAY10512, completely abrogated UFP2-induced pro-inflammatory gene expression, illustrating the important role of NF-κB signaling in response to UFP2. Dagher et al. reported that PM induced apoptosis of lung epithelial cells by TNF-α, a strong activator of NF-κB signaling . Hartz et al. reported that DEP released TNF-α to bind to the TNF-α receptor, leading to P-glycoprotein up-regulation. However, inhibition of NF-κB did not block DEP-induced P-glycoprotein expression . Therefore, the exact mechanisms by which UFP induce pro-inflammatory responses in the presence of specific chemical components warrant further investigations.
While both UFP1 and UFP2 stimulated monocyte binding to endothelial cells, UFP2 was a more potent inducer. UFP1 did not activate the NF-κB signal pathway, and CAY10512 did not completely inhibit UFP2-stimulated monocyte binding. These findings suggest that alternative mechanisms may contribute to UFP-stimulated monocyte binding.