Figure 1 depicts 10-minutes data segments of systolic and diastolic arterial BP, HR, body temperature (T), and activity level (Act) of one rat for the time course of baseline (day 0), exposure (day 1), and recovery (day 2–5) periods. The data reflect typical circadian rhythmicity of physiological and behavioural (activity) parameters of SHRs as characterized by elevated BP, HR, T, and Act values during the dark periods.

Comparison of baseline values between 6 and 7 months old SHRs indicate that cardiac performance was not altered by the 4 week time gap between control (6. months) and exposure (7. months) conditions (Figures 2 a–d). Baseline values of mBP (control/exposure: 176 ± 1.2/177 ± 1.2 mmHg), HR (control/exposure: 320 ± 3.5/318 ± 3.4 bpm), T (control/exposure: 38.1 ± 0.03/38 ± 0.03°C) and Act (control/exposure: 63.6 ± 1.4/63 ± 0.06 event count) remain unchanged by the 4 week time gap between control and exposure conditions.

Figures 2a and 2b display mean values (± SE) for mean arterial blood pressure (mBP) and HR during dark periods of baseline, exposure and recovery. Following 24 h exposure to UfCPs mBP increased by about 6 mmHg, (control/exposure: 176 ± 2.0/182 ± 2.1 mmHg; about 4%, p < 0.05) on the first and second day of recovery in exposed SHRs, and returned to baseline levels on the fourth day of recovery. This is due to the concurrent increase in systolic (control/exposure: 194 ± 2.8/204 ± 2.5 mmHg) and diastolic BP (control/exposure: 158 ± 1.5/160 ± 1.8 mmHg) with the systolic being more pronounced. In comparison to the BP response, HR responded with a lag of one day, it increased on the second and third day of recovery by about 17 bpm, (control/exposure: 328 ± 3.5/345 ± 3.5 bpm; about 5%, p < 0.05) and reached baseline values on the fourth day of recovery. UfCP exposure did not affect body temperature or activity levels of the animals. Both remained unaffected in UfCPs exposed SHRs compared to their control (Figures 2c and 2d).

The standard deviation of all normal adjacent sinus intervals (SDNN), a measure of the overall heart-rate variability (HRV), was decreased by about 30% (p < 0.05) during the recovery period on the second and third day (Figure 3). The square root of the mean of squared differences between adjacent normal to normal intervals (RMSSD) and the low-frequency to high-frequency ratio (LF/HF), showed a comparable response as SDNN, but failed to be statistically significant (Figure 3). No individual changes in absolute LF (baseline: 21.5 ± 1.13 nu; second recovery day: 17.6 ± 1.3 nu; third recovery day: 18.1 ± 0.89 nu) and HF (baseline: 70.2 ± 1.9 nu; second recovery day: 67.8 ± 2.8 nu; third recovery day: 72.5 ± 1.6 nu) power have been observed. The observed transient increase in HR associated with overall decrease in HRV (SDNN) suggests an altered sympatho-vagal balance in response to UfCPs inhalation.

Responses to UfCPs were studied considering particle number and mass concentrations, which approximate peak ambient particulate matter burdens 2627. Assuming a mean minute ventilation of 214 ml minutes^-1 28 the rats in this study inhaled 308 liter of aerosol (172 μg m^-3) in 24 h, resulting in an inhaled cumulative dose of approximately 53 μg UfCPs. It has been shown 29 that inhalation exposure of rat to ultrafine gold particle (mass median diameter: 49 nm) under similar conditions results in an alveolar deposition rate of 20%. Considering this deposition rate, the alveolar burden of UfCPs in our present experiment is ~10.6 μg or 5.5 × 10¹¹ particles. With respect to our previous studies conducted in rats 1724 about 20% of the deposited dose, i.e. 2 μg/rat, is supposed to be rapidly translocated into pulmonary tissues and may exert direct intracellular effects in these tissues. According to the translocation studies 17 much less than 1 μg may be systemically available and less than 0.1 μg may reach the heart. These dose estimates corresponds to the dose that a healthy human will accumulate over one year when being exposed to common ambient levels of insoluble UFPs 30, i.e. 3 × 10¹¹ UFPs will accumulate in the lungs and 6 × 10⁸ UFPs in each of the secondary target organs (the estimate is based on a particle concentration of 1 × 10³ cm^-3, a daily inhaled gas volume of 1 × 10⁴ l/d, and a deposition fraction in the peripheral lung of 0.3). However, ambient UFP concentration near busy roads may exceed the assumed particle number concentration substantially up to a factor of 100 31 and the deposition fraction may be higher in individuals suffering from respiratory diseases, e.g. ~70% in asthmatics 32. This suggest that the biological responses detected in the present study may be related to peak ambient ultrafine particle exposures to which humans may incidentally or accidentally be exposed to.

The cardiophysiological response in SHRs was characterized by a prolonged increase in BP and HR by about 5% during first to third day of the recovery period (p < 0.05). The extent of HR increase in UfCPs-exposed SHRs is consistent with other studies 3334 and that of our previous study 9 showing a significant elevation of HR by ~5% in WKY rats exposed to a comparable UfCPs concentration (180 μg m^-3). However, in our previous study the normotensive (WKY) rats showed only increased HRs during the exposure period and did not exhibit any changes in BP. Similarly Gordent et al. 35 reported small but significant increases of HR (approximately 5%) in healthy rats during 6 h nose-only exposures to concentrated ambient particles (132 μg m^-3-184 μg m^-3). In SHRs about 8% increase of HR has been observed at much higher diesel particle concentrations [1000 μg m^-3; 33. Furthermore, the mild but significant BP increase is comparable with the physiological endpoints from F-344 rats after exposure to urban ambient particles 36. Overall, BP and HR changes are admittedly small; however, with respect to the UfCPs exposure level selected, they are reasonable and likely to occur also in humans. Comparison of response patterns in healthy and compromised animals suggests that hypertensive animals are more susceptible to UfCPs particle exposure than normotensive. We detected a significant increase in BP and HR in SHRs over a period of 3 days after exposure to UfCPs. Whereas in other studies with uncompromised animal model, alteration of cardiovascular perfomance (HR and/or BP) was noted during the exposure and the values reach baseline levels rapidly after exposure 93335.

A series of epidemiological studies have shown that exposure to ambient ultrafine particles is associated with pulmonary inflammation caused by deposition of particles in the alveoli 1037. Particle exposure may results in local and systemic inflammatory responses 611 and ultimately may lead to an activation of the coagulatory system 1819, increased plasma viscosity 12, vascular and endothelial dysfunction 1415. Therefore, we assessed pulmonary and systemic inflammatory responses in order to shed light on pathophysiological pathways, which might be the plausible cause of UfCPs induced cardiovascular effects. No apparent inflammatory response has been detected in pulmonary tissue but a low grade inflammatory reaction cannot be ruled out. Plasma level of acute phase proteins such as HP and CRP were not affected by UfCPs exposure, but the observed small increases in the fraction of neutrophils and lymphocytes following 24 h exposure to UfCPs may indicate a small degree of systemic inflammation. Similar observations have been made by others and such phenomenon are suggested to contribute to the progression of atherosclerosis, hypertension and increased the risk of cardiopulmonary disease 3839. The precise mechanism(s) of how systemic impairment is produced by inhalation of UfCPs and its consequence on cardiophysiology need to be explored.

The pulmonary response of exposed SHRs is characterized by a significant induction of HO-1 (~2.5-fold), a sensitive marker for UfCPs mediated oxidative stress 4041, on the third day of recovery. The induced HO-1 is an indicator of a host defense mechanism for oxidative stress and is likely to be directly activated by UfCPs which have been shown to provide a substantial oxidative potency 42. The observed induction of PAI-1 (~2.5-fold) can be considered to be a downstream effect of particle induced oxidative stress 43. Increased levels of PAI-1, have been recognized as hallmarks of impaired endothelial function and are a common denominator of increased risk for cardiovascular disease 44. PAI-1 is known to be an inhibitor of fibrinolysis and regulator of vasoactivity 45, thus, increased levels of PAI-1 mRNA expression may also be involved in the observed increase in BP. Further, an increased TF expression (~2.5-fold) was detected in the lungs. Increased TF levels have been related to an increased risk of cardiac events because induction of TF is highly correlated with thrombogenesis and endothelial dysfunction 4647. Activation of TF, the extrinsic coagulation pathway, in association with impaired fibrinolysis via PAI-1 activation suggests that UfCPs exposure induces endothelial dysfunction and activates the coagulatory pathway, both of which are correlated with overall cardiovascular risk 454647.

Additionally, induction of ETA and ETB (~2.5-fold) along with a 6-fold induction of pulomonary ET-1 further supports the notion of an UfCPs induced endothelial dysfunction in the pulmonary circulation. The simultaneous increase in ETA and ET-1 indicates a synergistic effect because ET-1 contributes to endothelial dysfunction predominantly via ETA receptor stimulation 48. Potential health implications are obvious since several reports imply that both ETA and ETB are contributing to ET-1 induced hypertension 4950. Moreover, up-regulation of ETB receptors is most notable in heart failure, hypertension and in artherosclesis formation 51.

Hence, the observed induction of the endothelin system (ET-1, ETA and ETB) and increased expression of coagulatory factors (PAI-1 and TF) in pulmonary tissue suggests that UfCPs exposure triggers important pathophysiological pathways in the lungs which have been associated with impaired cardiovascular performance and an increases risk for cardiovascular events.

Interestingly, in the heart of UfCPs exposed SHRs, the only marker being affected was the oxidative stress-inducible defense enzyme HO-1. Expression of HO-1 mRNA was more than 2-fold repressed in contrast to its up-regulation in the lung tissue. Our data do not allow to provide a final explanation for this observation, but under hypoxic conditions HO-1 mRNA is reported to be induced by a number of studies while others found a repression 52. It was suggested that the differential regulation and the different expression levels of HO-1 represent an adaptation or unrecognized defense strategy to stress and that the response can differ depending on cell and tissue type 5253. Accordingly, we may value the different regulations observed for HO-1 in lung and cardiac tissue in the present study. However, we have to specify that we do not have any hint that UfCP exposure results in hypoxic conditions in cardiac tissue.

The most interesting UfCPs-mediated systemic effect is the close association of increased BP in exposed SHRs (Figure 2a) with a significant increase of circulating PRC at the same time points (Figure 5a). The tendency of increased Ang I and II concentrations on the first day of recovery further supports the notion that the RAS is primarily involved in the UfCPs-mediated BP increase detected in exposed SHRs 545556. On the other hand, the enzymatic activity of renin appears not to be affected by UfCPs exposure. Renin is a rate-limiting enzyme determining the overall RAS activity. It converts plasma angiotensinogen to Ang I, which is subsequently converted to Ang II, one of the most powerful vasoconstrictor. The RAS system is typically associated with the perfusion of the kidney, however, several recent studies reported that local RAS exist which is physiologically active in different tissues like lung, heart, kidney, or brain 555758. Besides regulating regional perfusion, local RAS may also play an important role in regulating systemic blood pressure 555758. Therefore, induced RAS in plasma might be due to direct effect of UfCPs in the pulmonary tissue, resulting in local RAS activation and systemic blood pressure elevation.

Male spontaneously hypertensive rats (SHR; 6 month) were used for the present study. Animals were housed under filtered air and specific pathogen free (SPF) conditions at a mean temperature of 22 ± 2°C, a mean relative humidity of 50 ± 5%, and a 12 h light-dark cycle (6 a.m. to 6 p.m. light on) with pelleted feed and filtered water being supplied ad libitum. Experimental protocols were approved by the Animal Care and Use Committee of the HelmholtzZentrum München – German Research Center for Environmental Health and by the Bavarian Animal Research Authority (211-2531-88/2001).

The methodology of UfCPs generation and the setup of the whole body exposure system for rodent have been previously described 959. UfCPs showed a monomodal number distribution in the exposure chamber with a median particle size ± arithmetic SD of 31 ± 0.3 nm. Measured mass and number concentration was 172 μg m^-3 and 9× 10⁶cm^-3, respectively. This translates into a surface area concentration of 0.139 m²(particle) m^-3 (air) because the mass specific surface area (according to the BET method) of the UfCPs was determined to be 807 m² g^-1. Based on the polydispersity of the particle distribution (geometric standard deviation is 1.51) a median mass diameter of 46 nm is calculated.

Table 2 provides an overview of the experimental design. Primarily, the cardiophysiological responses, i.e. effects on HR and BP were measured in 7 SHRs following 24 h UfCPs inhalation exposure using a radio telemetry system. Since BP and HR were increased on first to third day of recovery, the subsequent exposures were conducted in additional, non-telemetry SHRs to obtain blood, BALF and tissue samples from the first and third day of recovery. Each exposure used 16 SHRs, 8 animals were exposed to filtered air (controls) while the other 8 animals were exposed to UfCPs for 24 h (exposed). In the first study, animals were sacrificed in the morning of the first day of recovery. Prior to BALF collection, blood samples (blood A) of 8 SHRs were collected from retro orbital sinus for analysis of haematological parameters and from the abdominal aorta for analysis of biomarkers. Six animals were used to collect BALF and tissue samples (heart and lung) for further assessment of pulmonary and systemic response, the remaining 2 SHRs of each group were used for pulmonary and cardiac histopathology. Animal distribution and sample collection of the control group was similar to that of the exposed group.

In the second inhalation study, SHRs were sacrificed in the morning on third day of recovery. Blood sampling from retro orbital sinus and abdominal aorta (blood A), BALF and tissue samples collection, as well as cardiac histopathology were carried out as described above. In addition, 400 μl of blood from the caudal vein of each animal (blood B) was collected in the morning before exposure (base line) and on the first, second and third day of recovery to assess plasma renin concentration and activity.

The whole heart of each non lavaged animal (n = 2) was submerged in fixative until processing for histology. Paraffin blocks were prepared from dehydrated tissues and 3- to 4-μm sections were stained with hematoxylin and eosin for light microscopic evaluation of the cardiac histopathology 9.

Haematological analysis, measurement of different biomarkers from plasma and serum were used for the assessment of systemic response following UfCPs exposure. Blood samples (blood A, Table 2) of each animal were collected from retro orbital sinus (haematology) and from abdominal aorta (biomarkers) on first and third day of recovery. For analysis of plasma renin concentration and activity, blood samples (blood B, Table 2) were also collected from caudal vein at several time points.

After checking for the normal distribution assumption the differences between exposure and control groups were compared by using the t-test. Cardiovascular response parameters were described by a linear mixed regression model for repeated measurements. Based on this model group differences between the exposure and control group were tested. For expression analysis of various parameters from lung and heart tissues, a two-way analysis of variance (ANOVA) was used to analyze differences between the groups. However, for the plasma rennin and angiotensin (I and II) data the normailtiy assumption does not hold. Therefore, for plasma rennin and angiotensin (I and II) concentration the Wilcoxon rank sum test was performed. P values less than 0.05 were stated as statistically significant. All computations were done by the software packages Statgraphics plus v5.0 (Manugistics, Rockville, MD) and SAS V9.1 (Cary, NC). Data are presented as arithmetic mean values of n observations ± the standard error (SE), unless otherwise indicated.