Engine-Operating Load Influences Diesel Exhaust Composition and Cardiopulmonary and Immune Responses
Background: The composition of diesel engine exhaust (DEE) varies by engine type and condition, fuel, engine operation, and exhaust after treatment such as particle traps. DEE has been shown to increase inflammation, susceptibility to infection, and cardiovascular responses in experimentally exposed rodents and humans. Engines used in these studies have been operated at idle, at different steady-state loads, or on variable-load cycles, but exposures are often reported only as the mass concentration of particulate matter (PM), and the effects of different engine loads and the resulting differences in DEE composition are unknown.
Objectives: We assessed the impacts of load-related differences in DEE composition on models of inflammation, susceptibility to infection, and cardiovascular toxicity.
Methods: We assessed inflammation and susceptibility to viral infection in C57BL/6 mice and cardiovascular toxicity in APOE–/– mice after being exposed to DEE generated from a single-cylinder diesel generator operated at partial or full load.
Results: At the same PM mass concentration, partial load resulted in higher proportions of particle organic carbon content and a smaller particle size than did high load. Vapor-phase hydrocarbon content was greater at partial load. Compared with high-load DEE, partial-load DEE caused greater responses in heart rate and T-wave morphology, in terms of both magnitude and rapidity of onset of effects, consistent with previous findings that systemic effects may be driven largely by the gas phase of the exposure atmospheres. However, high-load DEE caused more lung inflammation and greater susceptibility to viral infection than did partial load.
Conclusions: Differences in engine load, as well as other operating variables, are important determinants of the type and magnitude of responses to inhaled DEE. PM mass concentration alone is not a sufficient basis for comparing or combining results from studies using DEE generated under different conditions.
Advances in engine, fuel, and after-treatment technologies such as particle traps have markedly reduced emissions from recent model diesel engines, but some older engines remain in use, and diesel engine exhaust (DEE) has likely contributed to associations between health and proximity to heavy traffic (reviewed by Boothe and Shendell 2008). Potential cancer risks remain of interest, but studies involving experimental exposures of humans and animals now focus predominantly on a broad range of noncancer effects (Hesterberg et al. 2009; Mauderly and Garshick 2009).
Variations in the composition of DEE can influence biological responses (McDonald et al. 2004c; Stevens et al. 2009). The engines, fuels, and the operating modes of engines used for experimental exposures vary widely among laboratories; in fact, no “standard” exists for experimental DEE. Neither is there a standard for describing the exposure atmospheres; the detail with which exposures are characterized varies considerably, and most comparisons among studies are made on the basis of particulate matter (PM) mass concentrations. Although long recognized among the engineering community, the substantial variation in DEE among operating conditions and its potential impact on comparability among studies has received only modest acknowledgment among the health research community.
Hypotheses concerning the “toxic” components of DEE have historically focused on the PM component and include associations with carbon, particle size, particle number, particle surface area, and mass. DEE also consists of variable amounts of nitrogen oxides, sulfur oxides, and organic gases and vapors. Indeed, PM is typically a minor part of the total mass emissions of even older engines (McDonald et al. 2004b). The composition of DEE varies with differences in engine performance, engine characteristics (e.g., size, fuel injection strategy), fuel and lubricant type, engine load, environmental conditions, and after-treatment devices (Hesterberg et al. 2009; Lloyd and Cackette 2001). Engine load alone can affect the ratio of organic to elemental carbon in diesel PM, with a higher organic:elemental ratio at low loads (Shi et al. 2000).
Although some experimental exposures of humans to DEE have used engines operated on realistic variable-duty cycles (e.g., Barath et al. 2010), most have used engines operated in steady-state modes. For example, Umeå University in Sweden conducted studies with healthy volunteers who inhaled exhaust from an idling engine in a laboratory setting. The results of these investigations showed increased airway inflammation and reactivity (Rudell et al. 1996), reduced forearm blood flow in response to vasodilators and fibrinolytic dysfunction (Mills et al. 2005; Törnqvist et al. 2007), increased ex vivo thrombus formation (Lucking et al. 2008), increased arterial stiffness (Lundbäck et al. 2009), and changes in the electrical activity in the brain (Crüts et al. 2008), but no significant change was observed in ex vivo fibrin clot formation (Metassan et al. 2010). That laboratory also showed ischemic alterations of the electrocardiogram (ECG) of men with previous myocardial infarctions (Mills et al. 2007). In a study conducted at the University of Washington in a human exposure laboratory, researchers found that exhaust from an engine operated in steady state at 75% load altered gene expression in circulating monocytes (Peretz et al. 2007) and caused peripheral vasoconstriction and increased circulating vasoactive factors (Peretz et al. 2008b) but did not significantly affect blood coagulation factors (Carlsten et al. 2008), prothrombotic markers (Carlsten et al. 2008), or heart rate (HR) abnormalities (Peretz et al. 2008a). Because none of the above publications provided detailed descriptions of the DEE, it is difficult to determine how differences in engine load may have influenced composition and the comparability of results among laboratories or whether the same outcomes might result from exposure to DEE emitted from vehicles during typical on-road use.
We conducted the present study to determine the influence of engine-operating conditions on selected responses of mice exposed to DEE by inhalation. In particular, we investigated the influence of engine work (load) under two different operating conditions. We measured indices of lung oxidative stress and inflammatory potential, susceptibility to lung infection, and alterations in HR and HR abnormalities because these parameters had been found at the Lovelace Respiratory Research Institute to be affected by DEE at the PM concentration used in the present study (Campen et al. 2003; Harrod et al. 2003; Seagrave et al. 2005). As expected, the composition of DEE differed at high and low engine loads. High engine load caused greater lung responses than did low engine load, but varying engine load did not affect HR responses.
Materials and Methods
We compared responses to DEE under the two engine load conditions at the same PM mass concentrations, which ranged from 250 to approximately 3,500 μg/m3. These concentrations span PM concentrations of DEE that may exist in high ambient to occupational exposure in humans (McDonald et al. 2002). We measured biological responses and DEE composition identically under the two conditions.
Exposure. The exposure system has been described (McDonald et al. 2004a) and is briefly summarized here. DEE was produced by a 5,500-W single-cylinder diesel generator (model YDG 5500E; Yanmar, Osaka, Japan) that contained a 406-cc–displacement air-cooled engine. The engine was fueled with no. 2 diesel pre-2007 certification fuel (Phillips Chemical Co., Borger, TX). Manufacturer-specified fuel properties included a 370-ppm sulfur content, 29% (by volume) aromatics, and a 47.6 cetane index. Engine lubricating oil (Rotella T 15w-40; Shell Oil Company, The Hague, Netherlands) was changed immediately before each 1-week exposure period. A bank of 500-W halogen lights pulled electrical current from the generator and provided a constant load that differed by the number of lights in operation. The engine operated at partial (55%) load with six lights and at high (100%) load with 11 lights. Engine load was defined as the number of watts drawn from the lights as a fraction of the total engine capacity.
The animal exposure chambers were 1 m3 whole-body inhalation chambers (H-1000; Hazelton Systems, Maywood, NJ) operated at a flow rate (250 L/min) that produced approximately 15 air exchanges per hour. We monitored and electronically recorded the temperature, relative humidity (RH), and flow using LabView software (version 6; National Instruments Corp., Austin, TX). Chamber temperature was maintained at 22–26°C.
We conducted DEE exposures 6 hr/day for either 3 (cardiovascular toxicity) or 7 (all other end points) consecutive days. We maintained exposure levels by adjusting the clean air dilution to a predetermined PM mass concentration, which was measured both in real time and as integrated over 30-min periods as described below. Concurrent exposures to HEPA- and charcoal-filtered air accompanied each DEE exposure. Other than exposure, we treated filtered-air control and DEE-exposed mice identically.
PM mass concentration was measured gravimetrically by sampling for 30-min intervals on 47-mm Pallflex filters (T60A20; Pall-Gelman, East Hills, NY) connected to the exposure chamber with a 20.3-cm-long × 0.6-cm internal-diameter probe connected to an aluminum in-line filter holder (In-Tox Products, Inc., Albuquerque, NM). We electronically measured pre- and postsample filter weights (MT5 balance; Mettler-Toledo, Columbus, OH). A static discharger was used before weighing filters to avoid interference from electrical charge. We characterized the composition of the exposure atmospheres in detail as reported (McDonald et al. 2004a).
Animals. We used two strains of mice previously established as disease or injury models in either compromised or healthy subjects. For assays of lung inflammation and resistance to infection, we used 8- to 10-week-old C57BL/6 male mice (Charles River Laboratories, Inc., Wilmington, MA) that were fed a normal ration (Teklad Rodent Diet; Harland, Madison, WI). For assays of HR and HR abnormalities, we used 8- to 12-week-old ApoE–/– male mice (Jackson Laboratories, Bar Harbor, ME) that were fed a high-fat ration (Teklad JK050814; Harland) for 16 weeks before exposure. The ApoE–/– mice are a susceptible model for cardiovascular toxicity because they have a markedly altered plasma lipid profile compared with normal mice and rapidly develop atherosclerotic lesions. All mice were provided water ad libitum and feed outside of exposure hours. All mice were housed in an Association for Assessment and Accreditation of Laboratory Animal Care–approved pathogen-free, rodent housing facility at controlled temperature (20–24°C) and humidity (30–60% RH) and quarantined for 2 weeks before use. Routine serological screens indicated an absence of mouse pathogens. The Lovelace Respiratory Research Institute Animal Care and Use Committee approved all procedures. The animals were treated humanely and with regard for alleviating suffering.
Biological response assays. Inflammatory signaling proteins (cytokines) were measured in homogenates of the right caudal and middle lung lobes (C57BL/6 mice, six per group). Immediately after sacrifice, the lungs were flash frozen. Before analysis, lungs were removed from the freezer and brought to room temperature. Lungs were homogenized for 1 min at full speed in a Tissuemizer (Tekmar, Mason, OH) and centrifuged for 5 min at 14,000 × n microfuge tubes and kept on ice. Heme oxygenase-1 (HO-1) and cytokines [tumor necrosis factor-α (TNF-α) and interferon-γ (IFN-γ)] were determined (two measurements for each cytokine for each sample) by enzyme-linked immunosorbent assays using commercially available analysis kits (Biosource International, Camarillo, CA). To normalize the cytokine measurements to total protein, we diluted the supernatants to 2 mg/mL in phosphate-buffered saline (PBS), and total protein was assayed by the Coomassie dye binding assay (Pierce, Rockford, IL) with bovine serum albumin as the standard.
Inflammation was assessed by the infiltration of inflammatory cells measured in bronchoalveolar lavage fluid. Lungs were lavaged three times with 1 mL PBS. The total number of leukocytes in the recovered lavage fluid was determined using a hemocytometer, and a cytocentrifuge slide was prepared and stained with Diff-Quick for evaluation of the differential cell types in the lavage fluid. These results are not reported because we observed no increase in lung inflammatory cells.
The effect of exposure on resistance to respiratory infection was studied in mice exposed to DEE diluted to approximately 250 μg/m3 PM. The study used a respiratory syncytial virus (RSV) model as previously described (Harrod et al. 2003). Immediately after the seventh day of exposure, groups of exposed and control C57BL/6 mice (n = 8 per group) were instilled intratracheally with 107 plaque-forming units of cultured RSV. Parallel groups of exposed and control mice without infection were sacrificed immediately after exposure as controls for RSV pathology. Infected mice were housed individually in pathogen-free conditions in a Biosafety Level 2 animal facility for 4 days until sacrifice. One lobe of the lung was analyzed for the presence of RSV by densitometric analysis of virus-specific mRNA transcripts that were isolated by gel electrophoresis after gene amplification. Sections for histopathology were obtained from another lung lobe, stained with hematoxylin and eosin (H&E), and examined by a pathologist under a light microscope. Cross sections were obtained approximately 500 μm caudal to the junction of the mainstream bronchus. The condition of the airways and vessels was scored (0–4 scale) without knowledge of the treatment group.
The ApoE–/– mouse model of cardiovascular response to air pollutants, including both its justification and its weaknesses in predicting human response to air pollutants, has been previously described (Campen et al. 2005). After 14 weeks on a high-fat diet, the mice were surgically implanted with radiotelemetry devices (DataSciences, Inc., St. Paul, MN) using a technique similar to that reported for rats (Campen et al. 2003) except that the telemeter remained subcutaneous, rather than being placed in the peritoneal cavity. All procedures were conducted under sterile conditions, and we allowed the mice 7 days to recover before acquisition of control data. The mice remained on the high-fat diet throughout the study period.
We exposed ApoE–/– mice (n = 5 per group) for 6 hr/day for 3 consecutive days. The DEE concentration was 3.5 mg PM/m3. Signals from telemetry devices were captured by a digital acquisition system (Gould Ponemah Life Science Suite, Valley View, OH). HR was determined from beat-to-beat intervals, averaged over 5-min periods. For T-wave morphology analysis, a specialized ECG analysis package (ECGAuto; EMKA Technologies, Falls Church, VA) reviewed the collected waveforms. T-wave area was calculated as millivolts times seconds, based on the deviation from the isoelectric line defined as the onset of the Q-wave. We normalized values to each individual subject’s baseline values over the 24-hr period preceding exposure.
Statistical analysis of data. For each response category, analysis of variance (ANOVA) was used to evaluate mean differences between responses to DEE and concurrent filtered-air exposure. Levene’s test (Levene 1960) was used to evaluate the appropriateness of the standard ANOVA assumption of equality of variances among experimental group responses. These tests showed significant evidence of inequality of variances (p < 0.05) for all end points except lung histopathology. To alleviate this, a weighted least squares analysis (Neter et al. 1996) was employed using the reciprocals of the estimated variances in experimental groups as weights. F-test contrasts (Searle 1971) were used to compare differences in responses between emission and concurrent filtered-air control groups.
For HR and ECG, group data were tested by two-way ANOVA (time and DEE concentration as factors for whole-body ECG data; DEE and analyte concentrations for vessel assays). Bonferroni post hoc tests were used to determine specific differences. Probability values < 0.05 were considered significant. We used SAS (version 9.0; SAS Institute, Inc., Cary, NC) to perform all the statistical analyses.
Composition of DEE. A summary of the DEE composition and particle size is shown in Figure 1. The high-load operating mode produced exhaust with less carbon monoxide (CO) and less organic material than did the partial-load mode, which was dominated by organic carbon in both the gas and particle phases. The clean-air exposure contained low quantities of both gases and PM. PM mass size distributions did not differ between high- and partial-load modes, but the particle number size distributions did. The PM mass size distributions of both atmospheres had mass median diameters of approximately 0.15 μm. The particle number size distributions had medians of 0.085 μm at high load and 0.030 μm at partial load.
Lung inflammation.Figure 2 shows the concentrations of HO-1, IFN-γ, and TNF-α (normalized for total protein) in lung homogenate from C57BL/6 mice that were sacrificed immediately after DEE or filtered-air exposure (no infection). HO-1 was significantly up-regulated by both DEE exposures, although high load caused a stronger response than did particle load. Interestingly, we found significant (p < 0.01, ANOVA) reductions of both IFN-γ and TNF-α at partial load (relative to filtered-air concurrent controls) and significant (p < 0.01) increases at high load.
Respiratory infection.Figure 3 summarizes the lung responses of mice to RSV challenge, showing the histopathological score and the viral persistence, or the amount of RSV that was still present in the lungs 4 days after instillation. Exposure at high load caused a significant increase in pathological score compared with the clean-air controls, whereas exposure at partial load caused only a slight, nonsignificant increase. Exposure at high load also decreased clearance of RSV from the lungs (p = 0.002), but exposure at partial load did not reduce viral clearance. Histological inflammation was evident at baseline, due to RSV inoculation alone (Figure 4A). Exposure at high load considerably exacerbated inflammation and epithelial pathology (Figure 4B). Exposure at partial load increased inflammation somewhat but did not cause epithelial disruption (Figure 4C).
HR and ECG. High-load DEE exposure induced a significant reduction in HR in ApoE–/– mice (Figure 5), which was consistent with previous findings. The HR responses were gradual, achieving statistical significance during the fourth hour of exposure at roughly 100 beats/min (BPM) below control (Figure 5A). In contrast, exposure at partial load caused a nearly instantaneous significant reduction in HR, with average values in the first hour of roughly 175 BPM below control. Maximal reductions during the 6-hr exposure were roughly 215 BPM below baseline values at partial load. The HR returned to normal values within 1–2 hr after exposure at either load level (data not shown).
At partial load, T-wave alterations mirrored the HR changes, both in magnitude and in time to onset (Figure 5B). We saw a very rapid change in the T-wave in the first hour, resulting in a 150% increase in T-wave area. The high load caused no effects on T-wave.
The present study demonstrates that the operating load of a diesel engine has a significant effect on the composition of the exhaust and, not surprisingly, the relative toxic potency of inhaled exhaust. In a somewhat surprising contrast, lung responses in C57BL/6 mice were greater at high load, whereas cardiac responses in APOE–/– mice were greater at partial load at the same PM mass concentration. We obtained these novel observations using fresh DEE and inhalation exposures of previously characterized models of toxicity. Exhaust produced under high load had a lower total organic content (both PM and gaseous), a higher content of elemental (black) carbon, and a slightly larger number-based particle size than did that produced at partial load. At both loads, the particles were in the ultrafine particle size range; however, the median aerodynamic diameter produced at high load was approximately 2.5–3 times that produced at partial load. Both atmospheres showed low sulfate concentrations (< 5% of PM), despite the use of fuel that had approximately 300 ppm sulfur content. Fuel after 2007 has < 15 ppm sulfur content, so the contribution of diesel to sulfur would decrease with more modern fuel relative to the pre-2007 fuel used in the present study. Exposure to high-load DEE produced lung responses in both the noninfected and infected mouse models. The responses were evident but clearly less at partial engine load. In some cases cytokines, including TNF-α, were suppressed after exposure to the partial-load atmosphere. TNF-α is classified as a “proinflammatory cytokine,” and a toxic effect is typically signaled by an increase. Because the physiological basis for the decrease in TNF-α at partial load is uncertain, this response is difficult to interpret.
The greater lung responses at high load suggest elemental carbon as a putative causal component. Elemental carbon was the only measured constituent that was present at a higher concentration at high load than at partial load. Of course, because we measured only select organic species in this study (polycyclic aromatic hydrocarbons, acids), some unmeasured organic compounds may be responsible or play a role in the observed effects. Furthermore, the higher CO levels in the low-load DEE may offer a protective effect similar to that described by Ryter and Choi (2010), where CO reduced inflammation in disease models when administered therapeutically. However, several epidemiology studies have shown associations between cardiovascular effects in people and exposure to black carbon as a proxy for traffic exposure (e.g., Zanobetti et al. 2010). Although those studies do not typically single out the toxicity of black carbon explicitly, the strong association lends plausibility that it may itself cause biological effects. A common misconception regarding the composition of elemental carbon is that it is analogous to “carbon, elemental,” which is often considered benign with respect to the biological end points studied here. “Elemental carbon” is actually a term defining the most light-absorbing portion of carbon, which is also most commonly defined functionally by its measurement as part of thermal–optical measurement methods (e.g., Chow et al. 2005). As described by Setyan et al. (2009), elemental carbon can contain functional groups and can have varying surface chemistry. As an example of the potential for elemental carbon with enhanced surface chemistry to interact with biological proteins, Seagrave et al. (2004) showed that elemental carbon with differential functional groups and surface chemistry can result in different interactions with inflammatory cytokines. An important point to consider is that if elemental carbon plays a role in the health effects of DEE, it is noteworthy that the diesel particle trap technology implemented to meet the U.S. 2007 on-road emission standards effectively removes most elemental carbon from exhaust. This was evident in our previous study, where we used the engine system described here coupled with a diesel particle trap and showed that elemental carbon was removed (and health effects were also reduced) (McDonald et al. 2004c).
The effects of DEE on the clearance of pathogens or susceptibility to infection have been reported in previous publications (Harrod et al. 2003; Reed and Berger 2006). We conducted those studies on a different engine system with a variable duty cycle. Indeed, both DEE and carbon have been shown in previous studies to have an effect on clearance of RSV. We have reported studies in which DEE either had an effect on RSV clearance (Harrod et al. 2003; McDonald et al. 2004c) or did not have an effect on RSV clearance (Reed and Berger 2006). Those studies used different DEE systems. In addition, the pathogenicity of the two lots of RSV used in the two studies differed, which may explain the apparent disparity of results. Further investigations confirmed the importance of viral passage in its pathogenicity, because we determined the characteristics and lineage of the virus to be important determinants of responses. In the present study, we treated all groups with RSV from the same viral passage and lineage conditions. As a result, the differences observed between high and partial load are believed to relate to differences in exhaust composition and not to differences in the virus.
We have previously established that high levels of DEE can induce bradycardia and T-wave alterations in mice and that ApoE–/– mice were more susceptible than their genetic background, C57BL/6 mice (Campen et al. 2005). These effects appear to be highly congruous with data from controlled diesel exposures with coronary artery disease patients, wherein significant T-wave abnormalities develop during combined exercise and exposure (Mills et al. 2007). The effects in previous rodent studies were primarily linked to the gas-phase constituents, as evidenced by the robust bradycardic responses elicited by an atmosphere with the particles removed by a ceramic filter. The present study is therefore consistent, showing a greater cardiovascular response caused by the partial-load atmosphere, which was characterized as having greater levels of gaseous pollutants. The nature of the cardiac responses was not fully characterized, but earlier work suggested that the bradycardia and bradyarrhythmias (observed in the present study but not quantified) were a result of vagal stimulation (Watkinson et al. 2001). The T-wave morphology, however, was a finding seemingly unique to models of cardiac disease, namely, the spontaneously hypertensive rat (Kodavanti et al. 2000) and the ApoE–/– mouse (Campen et al. 2005). We postulate that the high-fat diet regimen advances the natural vascular pathology in this model sufficiently to predispose the heart to occlusive ischemic events, similar to those reported by Caligiuri et al. (1999).
The present study highlights the role of DEE composition in driving divergent pathologies. These findings have implications for the design and interpretation of experiments involving experimental exposures of humans and animals to DEE. Using a difference of only 55% and 100% rated load, this study made clear that engine load alone can alter the nature and magnitude of lung and cardiac responses. It can be speculated, if not yet determined, that idle versus partial or full load might induce greater differences in response and that any other factor affecting the composition of DEE can also affect biological responses and cause results to differ among studies. Few would argue this point, but little attention has been given to the composition of DEE in reporting and comparing studies. Because real-world exposures result from a spectrum of engine types operated in a wide range of modes using different fuels, lubricating oils, and exhaust after-treatment, it is probably not useful to contemplate establishing a “standard” DEE for experimental exposures. Together, however, the present and previous results clearly indicate the importance of characterizing the composition of DEE used in laboratory exposures to a greater extent than has been typical of the literature to date. The present results demonstrate clearly that PM mass concentration alone is an inadequate metric for comparing results among studies or for estimating hazards or risks from exposure–response slopes using results from studies conducted under different conditions of engine type and operation.
This work was supported by the Office of Freedom Car and Vehicle Technologies, U.S. Department of Energy and by a grant from the National Institutes of Health (ES 014639).
This article was prepared as an account of work sponsored by an agency of the U.S. government. Neither the U.S. government nor any agency thereof, nor any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed or represents that its use would not infringe on privately owned rights. Reference herein to any specific commercial product, process, service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the U.S. government or any agency thereof. The views and opinions of the authors expressed herein do not necessarily state or reflect those of the U.S. government or any agency thereof. No copyright is asserted in the works of U.S. government employees.
The authors declare they have no actual or potential competing financial interests.
Barath S, Mills NL, Lundbäck M, Törnqvist H, Lucking AJ, Langrish JP. 2010. Impaired vascular function after exposure to diesel exhaust generated at urban transient running conditions.Part Fibre Toxicol 7:19; doi: 10.1186/1743-8977-7-19[Online 23 July 2010] 20653945. Crossref, Medline, Google Scholar Boothe VL, Shendell DG. 2008. Potential health effects associated with residential proximity to freeways and primary roads: review of scientific literature, 1999–2006.J Environ Health 70:33−41 18468221. Medline, Google Scholar Caligiuri G, Levy B, Pernow J, Thorén P, Hansson GK. 1999. Myocardial infarction mediated by endothelin receptor signaling in hypercholesterolemic mice.Proc Natl Acad Sci USA 96:6920-6924 10359814. Crossref, Medline, Google Scholar Campen MJ, Babu NS, Helms GA, Pett S, Wernly J, Mehran R. 2005. Nonparticulate components of diesel exhaust promote constriction in coronary arteries from ApoE-/- mice.Toxicol Sci 88:95-102 16093524. Crossref, Medline, Google Scholar Campen MJ, McDonald JD, Gigliotti AP, Seilkop SK, Reed MD, Benson JM. 2003. Cardiovascular effects of inhaled diesel exhaust in spontaneously hypertensive rats.Cardiovasc Toxicol 3:353-361 14734832. Crossref, Medline, Google Scholar Carlsten C, Kaufman JD, Trenga CA, Allen J, Peretz A, Sullivan JH. 2008. Thrombotic markers in metabolic syndrome subjects exposed to diesel exhaust.Inhal Toxicol 20:917−921 18668408. Crossref, Medline, Google Scholar Chow JC, Watson JG, Louie PK, Chen LW, Sin D. 2005. Comparison of PM2.5 carbon measurement methods in Hong Kong, China.Environ Pollut 137:334−344 15963372. Crossref, Medline, Google Scholar Crüts B, van Etten L, Törnqvist H, Blomberg A, Sandström T, Mills NL. 2008. Exposure to diesel exhaust induces changes in EEG in human volunteers.Part Fibre Toxicol 5:4; doi: 10.1186/1743-8977-5-4[Online 11 March 2008] 18334019. Crossref, Medline, Google Scholar Harrod KS, Jaramillo RJ, Rosenberger CL, Wang SZ, Berger JA, McDonald JD. 2003. Increased susceptibility to RSV infection by exposure to inhaled diesel engine emissions.Am J Respir Cell Mol Biol 28:451-463 12654634. Crossref, Medline, Google Scholar Hesterberg TW, Long CM, Bunn WB, Sax SN, Lapin CA, Valberg PA. 2009. Noncancer health effects of diesel exhaust: a critical assessment of recent human and animal toxicological literature.Crit Rev Toxicol 39:195−227 19280432. Crossref, Medline, Google Scholar Kodavanti UP, Schladweiler MC, Ledbetter AD, Watkinson WP, Campen MJ, Winsett DW. 2000. The spontaneously hypertensive rat as a model of human cardiovascular disease: evidence of exacerbated cardiopulmonary injury and oxidative stress from inhaled emission particulate matter.Toxicol Appl Pharmacol 164:250-263 10799335. Crossref, Medline, Google Scholar Levene H. 1960. Robust tests of equality of variances. In Contributions to Probability and Statistics (Olkin I, ed). Palo Alto, CA:Stanford University Press, 278–292. Google Scholar Lloyd AC, Cackette TA. 2001. Diesel engines: environmental impact and control.J Air Waste Manag Assoc 51:809-847 11417675. Crossref, Medline, Google Scholar Lucking AJ, Lundbäck M, Mills NL, Faratian D, Barath SL, Pourazar J. 2008. Diesel exhaust inhalation increases thrombus formation in man.Eur Heart J 29:3043−3051 18952612. Crossref, Medline, Google Scholar Lundbäck M, Mills NL, Lucking A, Barath S, Donaldson K, Newby DE. 2009. Experimental exposure to diesel exhaust increases arterial stiffness in man.Part Fibre Toxicol 6:7; doi: 10.1186/1743-8977-6-7[Online 13 March 2009] 19284640. Crossref, Medline, Google Scholar Mauderly JL, Garshick E. 2009. Diesel exhaust. In: Environmental Toxicants: Human Exposures and Their Health Effects (Lippmann M., ed). 3rd ed. Hoboken, NJWiley, 551−632. Google Scholar McDonald JD, Barr EB, White RK. 2004a. Design, characterization, and evaluation of a small scale diesel exhaust exposure system.Aerosol Sci Technol 38:62−78. Crossref, Google Scholar McDonald JD, Barr EB, White RK, Chow JC, Schauer JJ, Zielinska B. 2004b. Generation and characterization of four dilutions of diesel engine exhaust for a subchronic inhalation study.Environ Sci Technol 38:2513−2522 15180045. Crossref, Medline, Google Scholar McDonald JD, Harrod KS, Seagrave JC, Seilkop SK, Mauderly JL. 2004c. Effects of low sulfur fuel and a catalyzed particle trap on the composition and toxicity of diesel emissions.Environ Health Perspect 112:1307−1312 15345344. Link, Google Scholar McDonald JD, Zielinska B, Sagebiel JC, McDaniel MR. 2002. Characterization of fine particle material in ambient air and personal samples from an underground mine.Aerosol Sci Technol 36:1033−1044. Crossref, Google Scholar Metassan S, Routledge MN, Lucking AJ, de Willige SU, Philippou H, Mills NL. 2010. Fibrin clot structure remains unaffected in young healthy individuals after transient exposure to diesel exhaust.Part Fibre Toxicol 7:17; doi: 10.1186/1743-8977-7-17[Online 16 June 2010] 20565709. Crossref, Medline, Google Scholar Mills NL, Törnqvist H, Gonzalez MC, Vink E, Robinson SD, Söderberg S. 2007. Ischemic and thrombotic effects of dilute diesel-exhaust inhalation in men with coronary heart disease.N Engl J Med 357:1075-1082 17855668. Crossref, Medline, Google Scholar Mills NL, Törnqvist H, Robinson SD, Gonzalez M, Darnley K, MacNee W. 2005. Diesel exhaust inhalation causes vascular dysfunction and impaired endogenous fibrinolysis.Circulation 112:3930−3936 16365212. Crossref, Medline, Google Scholar Neter J, Kutner MH, Nactsheim CJ, Wasserman W. 1996. Applied Linear Statistical Models. 4th ed. Chicago, ILIrwin. Google Scholar Peretz A, Kaufman JD, Trenga CA, Allen J, Carlsten C, Aulet MR. 2008a. Effects of diesel exhaust inhalation on heart rate variability in human volunteers.Environ Res 107:178−184 18329013. Crossref, Medline, Google Scholar Peretz A, Peck EC, Bammler TK, Beyer RP, Sullivan JH, Trenga CA. 2007. Diesel exhaust inhalation and assessment of peripheral blood mononuclear cell gene transcription effects: an exploratory study of healthy human volunteers.Inhal Toxicol 19:1107−1119 17987463. Crossref, Medline, Google Scholar Peretz A, Sullivan JH, Leotta DF, Trenga CA, Sands FN, Allen J. 2008b. Diesel exhaust inhalation elicits acute vasoconstriction in vivo.Environ Health Perspect 116:937−942 18629317. Link, Google Scholar Reed MD, Berger JA. 2006. Real time RT-PCR assessment of clearance of respiratory syncytial virus altered by exposure to diesel exhaust and hardwood smoke. [Abstract]Toxicologist 90:232. Google Scholar Rudell B, Ledin MC, Hammarström U, Stjernberg N, Lundbäck B, Sandström T. 1996. Effects on symptoms and lung function in humans experimentally exposed to diesel exhaust.Occup Environ Med 53:658−662 8943829. Crossref, Medline, Google Scholar Ryter SW, Choi AM. 2010. Heme oxygenase-1/carbon monoxide: novel therapeutic strategies in critical care medicine.Curr Drug Targets 11:1485−1494 20704552. Crossref, Medline, Google Scholar Seagrave J, Knall C, McDonald JD, Mauderly JL. 2004. Diesel particulate material binds and concentrates a proinflammatory cytokine that causes neutrophil migration.Inhal Toxicol 16(suppl 1):93-98 15204797. Crossref, Medline, Google Scholar Seagrave JC, McDonald JD, Reed MD, Seilkop SK, Mauderly JL. 2005. Responses to subchronic inhalation of low concentrations of diesel exhaust and hardwood smoke measured in rat bronchoalveolar lavage fluid.Inhal Toxicol 17:657−670 16087572. Crossref, Medline, Google Scholar Searle SR. 1971. Linear Models. New York:John Wiley and Sons. Google Scholar Setyan A, Sauvain JJ, Rossi MJ. 2009. The use of heterogeneous chemistry for the characterization of functional groups at the gas/particle interface of soot and TiO2 nanoparticles.Phys Chem Chem Phys 11:6205-6217 19606331. Crossref, Medline, Google Scholar Shi JP, Mark D, Harrison RM. 2000. Characterization of particles from a current technology heavy-duty diesel engine.Environ Sci Technol 34:748-755. Crossref, Google Scholar Stevens T, Cho SH, Linak WP, Gilmour MI. 2009. Differential potentiation of allergic lung disease in mice exposed to chemically distinct diesel samples.Toxicol Sci 107:522-534 19074765. Crossref, Medline, Google Scholar Törnqvist H, Mills NL, Gonzalez M, Miller MR, Robinson SD, Megson IL. 2007. Persistent endothelial dysfunction in humans after diesel exhaust inhalation.Am J Respir Crit Care Med 176:395−400 17446340. Crossref, Medline, Google Scholar Watkinson WP, Campen MJ, Nolan JP, Costa DL. 2001. Cardiovascular and systemic responses to inhaled pollutants in rodents: effects of ozone and particulate matter.Environ Health Perspect 109(suppl 4):539-546 11544160. Link, Google Scholar Zanobetti A, Gold DR, Stone PH, Suh HH, Schwartz J, Coull BA, Speizer FE. 2010. Reduction in heart rate variability with traffic and air pollution in patients with coronary heart disease.Environ Health Perspect 118(3):324-330 20064780. Link, Google Scholar