Impairment of NO-Dependent Relaxation in Intralobar Pulmonary Arteries: Comparison of Urban Particulate Matter and Manufactured Nanoparticles

Background and Objectives Because pulmonary circulation is the primary vascular target of inhaled particulate matter (PM), and nitric oxide is a major vasculoprotective agent, in this study we investigated the effect of various particles on the NO–cyclic guanosine monophosphate (cGMP) pathway in pulmonary arteries. Methods We used intrapulmonary arteries and/or endothelial cells, either exposed in vitro to particles or removed from PM-instilled animals for assessment of vasomotricity, cGMP and reactive oxygen species (ROS) levels, and cytokine/chemokine release. Results Endothelial NO-dependent relaxation and cGMP accumulation induced by acetylcholine (ACh) were both decreased after 24 hr exposure of rat intrapulmonary arteries to standard reference material 1648 (SRM1648; urban PM). Relaxation due to NO donors was also decreased by SRM1648, whereas responsiveness to cGMP analogue remained unaffected. Unlike SRM1648, ultrafine carbon black and ultrafine and fine titanium dioxide (TiO2) manufactured particles did not impair NO-mediated relaxation. SRM1648-induced decrease in relaxation response to ACh was prevented by dexamethasone (an anti-inflammatory agent) but not by antioxidants. Accordingly, SRM1648 increased the release of proinflammatory mediators (tumor necrosis factor-α, interleukin-8) from intrapulmonary arteries or pulmonary artery endothelial cells, but did not elevate ROS levels within intrapulmonary arteries. Decreased relaxation in response to ACh was also evidenced in intrapulmonary arteries removed from rats intratracheally instilled with SRM1648, but not with fine TiO2. Conclusion In contrast to manufactured particles (including nanoparticles), urban PM impairs NO but not cGMP responsiveness in intrapulmonary arteries. We attribute this effect to oxidative-stress–independent inflammatory response, resulting in decreased guanylyl cyclase activation by NO. Such impairment of the NO pathway may contribute to urban-PM–induced cardiovascular dysfunction.


Research
Epidemiologic studies demonstrate a correlation between exposure to particulate matter (PM) pollution and cardiovascular morbidity (Hoek et al. 2001) and mortality (Pope et al. 2002). PM is heterogeneous in size (aerodynamic diameter < 0.1 µm for ultrafine, PM 0.1 ; < 2.5 µm for fine, PM 2.5 ; < 10 µm for coarse, PM 10 ) and in composition (with various adsorbed constituents, e.g., transition metals and inorganic and organic compounds).
Among other adverse effects, exposure to PM may induce pulmonary and systemic inflammation and dysfunction (Salvi et al. 1999). Two major hypotheses, which are not mutually exclusive, have been proposed to account for the effects of inhaled particles that can deeply penetrate into the lungs. One hypothesis is that, once deposited into the lung, particles initiate a local inflammation, which triggers a secondary systemic inflammation that could exacerbate cardiovascular dysfunctions (Seaton et al. 1995). Another hypothesis, although controversial, involves the documented passage of the finest particles, especially nanoparticles, into the blood after inhalation (Nemmar et al. 2001(Nemmar et al. , 2002Wiebert et al. 2006), suggesting direct effects of translocated particles, or of some of their constituents, on remote target tissues.
Constriction of systemic or pulmonary arteries in response to PM is generally observed in in vitro and in vivo animal and human studies (Batalha et al. 2002;Brook et al. 2002;Huang et al. 2002;Li et al. 2005). Exposure to PM also induces a decrease in endotheliumdependent relaxation in systemic arteries (Ikeda et al. 1995;Nurkiewicz et al. 2004Nurkiewicz et al. , 2006. Nitric oxide is a major endotheliumderived vasculoprotective factor that, among other effects (Gewaltig and Kojda 2002), decreases vascular tone through heme-dependent stimulation of soluble guanylyl cyclase and subsequent activation of cyclic guanosine monophosphate (cGMP)-dependent protein kinases (Schlossmann et al. 2003). A decrease in endothelial NO production and/or bioactivity is a key event in the pathogenesis of many cardiovascular disorders (Li and Forstermann 2000). Inflammation and oxidative stress, two major effects accounting for some adverse effects of PM (Bai et al. 2007;Oberdörster et al. 2005), play a central role in endothelial dysfunction in many pathologic blood vessels (Feletou and Vanhoutte 2006), including pulmonary arteries (Fresquet et al. 2006). Although impairment of NO-dependent pathway may contribute to deleterious effects of PM on the cardiovascular system, this issue has never been specifically addressed in pulmonary circulation, which is a primary target of inhaled particles.
Therefore, in this study we investigated the influence of PM on the NO-cGMP relaxant pathway in intrapulmonary arteries. We investigated urban PM [standard reference material mean diameter of 0.4 µm and consist of > 63% inorganic carbon and 4-7% organic carbon. Major constituent elements (> 1% mass fraction) are silicon, sulfur, aluminum, iron, potassium, and sodium.
Other particles we used in this study included ultrafine carbon black (ufcb) FW2 and P60 (from Degussa, Frankfurt, Germany), with average primary particle sizes of 13 and 21 nm, respectively. We obtained ultrafine TiO 2 (ufTiO 2 ; average primary particle size, 15 nm) and fine TiO 2 (fTiO 2 ; mean diameter, 0.14 µm, determined by transmission electron microscopy) from Sigma. Before experiments, we freshly suspended particles (10 mg/mL) in distilled deionized water. We chose the concentrations of PM for in vitro or in vivo experiments based on the literature (Li et al. , 2006Mutlu et al. 2007;Nurkiewicz et al. 2004).
Animals, exposure to particles, and tissue preparation. Male Wistar rats (12-14 weeks old; Elevage Janvier, Le Genest Saint Isle, France) were treated humanely and with regard for the alleviation of suffering, according to the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources 1996). We dissected intrapulmonary arteries as previously described (Leblais et al. 2004). For in vitro experiments, we incubated segments in Dulbecco's modified Eagle's medium (DMEM), in absence or presence of particles, for 24 hr at 37°C in a humidified atmosphere (95% air/5% carbon dioxide). In some experiments, we added dexamethasone (10 µM), tempol (1 mM), or ascorbate (200 µM) to DMEM for the 24-hr incubation period. We anesthetized some rats by intraperitoneal injection of ketamine (50 mg/kg) and xylazine (4 mg/kg), and intratracheally instilled them with 5 mg SRM1648 or fTiO 2 in 500 µL saline (NaCl 0.9%), or with saline alone. After a recovery period (6-72 hr), we euthanized rats and removed lungs.
We exposed intrapulmonary arteries to KCl (5-100 mM) or PGF 2α (30 nM to 30 µM). After washout, we submaximally contracted them with PGF 2α to achieve approximately 50% of the tension obtained with 80 mM KCl. Once we obtained stable contraction, we added cumulative concentrations of ACh, sodium nitroprusside (SNP), DEA-NO, 8-bromo-cGMP (8-Br-cGMP), YC-1, isoproterenol, forskolin, or levcromakalim. Figure 1A illustrates the targets of drugs activating the NO-cGMP pathway. We used YC-1 to activate soluble guanylyl cyclase in an NOindependent manner. However, this compound also increases the sensitivity of the enzyme to NO (Friebe and Koesling 2003). To minimize the influence of the latter mechanism, we studied effect of YC-1 in endothelium-denuded arteries treated with ODQ, an irreversible inhibitor of NO-dependent activation of soluble guanylyl cyclase. When we studied isoproterenol, we pretreated arteries with phenoxybenzamine (1 µM) to irreversibly inactivate α-adrenoceptors (Pourageaud et al. 2005). In some experiments, we added tempol (1 mM) to the organ bath before precontraction with PGF 2α .
Detection of reactive oxygen species. We prepared sections of intrapulmonary arteries and exposed them to the fluorescent dye DHE (2.5 µM) as previously described (Fresquet et al. 2006). We examined slides under a laser scanning confocal microscope equipped with a krypton/argon laser (excitation, 488 nm; emission, 610 nm) and obtained final images after stacking.

Particulate matter and pulmonary circulation
Environmental Health Perspectives • VOLUME 116 | NUMBER 10 | October 2008 Bold arrows indicate targets of the different drugs we used. (B) Wall tension recordings of the effect of ACh in PGF 2α -precontracted rat intrapulmonary arteries incubated for 24 hr in the absence (black trace) or presence (gray trace) of SRM1648 (200 μg/mL). Oscillations during contraction are not a characteristic feature of SRM1648-exposed arteries, because we also observed them in unexposed arteries.
Statistical analysis. We expressed relaxation responses as the percentages of the initial tone induced by PGF 2α ; all data are means ± SE from n rats. We compared concentration-response curves using two-way analysis of variance (ANOVA) and performed other statistical comparisons with one-way ANOVA. We considered differences statistically significant when p < 0.05.

SRM1648 selectively impairs NO responsiveness.
After 24-hr preincubation without particles, maximal relaxation to 30 µM ACh was 50.02 ± 3.03% (n = 25). Such relaxation was slightly lower than that obtained in freshly isolated tissue (in which maximum reaches about 75%), as observed in other arterial models (Jiménez-Altayó et al. 2006). Nevertheless, the effect of ACh was almost totally abolished by the NO-synthase inhibitor L-NAME (5.0 ± 3.6% relaxation with 30 µM ACh plus 300 µM L-NAME, n = 4) or after endothelium removal with CHAPS (1.0 ± 1.7% relaxation with 30 µM ACh, n = 4). Neither contractile nor endothelium-independent relaxation maximal VOLUME 116 | NUMBER 10 | October 2008 • Environmental Health Perspectives Data are mean ± SE from 4-10 rats. (B) Levels of cGMP in rat intrapulmonary arteries incubated in the absence (white bar) or in the presence of SRM1648 (200 μg/ml for 24 hr, black bar) and subsequently exposed to ACh (10 μM). Data are mean ± SE from six rats. NS, not significant. capacity of arteries was altered by endothelium removal, because effects of KCl (80 mM) and of the NO donor SNP (10 µM) were not significantly different between CHAPS-treated and untreated arteries (data not shown). Preincubation of rat intrapulmonary arteries for 24 hr with SRM1648 did not modify contraction to KCl (5-100 mM) or PGF 2α (30 nM to 30 µM) (data not shown). We observed a significant decrease in AChinduced relaxation in intrapulmonary arteries after 24-hr exposure to SRM1648 at concentrations ≥ 100 µg/mL (Figures 1B, 2A). AChinduced cGMP accumulation was also diminished in intrapulmonary arteries after exposure to 200 µg/mL SRM1648 ( Figure  2B). Relaxation induced by either SNP or DEA-NO, two compounds releasing NO by distinct mechanisms (which might explain differences in their respective maximal effect), was diminished after exposure to 200 µg/mL SRM1648, whereas relaxation to 8-Br-cGMP, a membrane-permeable cGMP analog that directly stimulates cGMP-dependent protein kinase, or to YC-1, an activator of soluble guanylyl cyclase, remained unaffected ( Figure 2A).
Nanoparticles ufcb or ufTiO 2 , as well as fTiO 2 , do not impair NO responsiveness. For comparison with SRM1648, we studied the effects of ufcb, ufTiO 2 , and fTiO 2 . As shown in Figure 4, exposure of intrapulmonary arteries for 24 hr to these particles at 200 µg/mL did not significantly impair ACh-induced relaxation.

SRM1648 impairs NO responsiveness through oxidative-stress-independent inflammatory response.
The steroidal anti-inflammatory agent dexamethasone (10 µM, added concomitantly to SRM1648) fully prevented SRM1648-induced impairment of the relaxation response to ACh in intrapulmonary arteries, without modifying its relaxant effect in untreated arteries ( Figure 5A). Exposure of intrapulmonary arteries to SRM1648 (200 µg/mL) resulted in an increased release of the pro-inflammatory mediators TNF-α and MIP2 ( Figure 5B), the functional analogue of human IL-8. We also observed an increased level of IL-8 in incubation medium of pulmonary artery endothelial cells that we exposed to SRM1648 (200 µg/mL for 24 hr) ( Figure 5C).
When added concomitantly to SRM1648, the antioxidants tempol (1 mM) and ascorbate (200 µM) failed to modify effect of ACh ( Figure 6A). Similarly, tempol did not modify the effect of ACh when added to SRM1648pretreated arteries 15 min before precontraction with PGF 2α ( Figure 6A). Compared with untreated intrapulmonary arteries, those exposed to SRM1648 did not exhibit elevated levels of reactive oxygen species (ROS), as determined using the fluorescent probe DHE ( Figure 6B).
In vivo exposure to SRM1648, but not fTiO 2 , impairs NO responsiveness. The presence of particles (arrows, Figure 7A) was clearly evidenced in lung parenchyma removed 12 or 72 hr after intratracheal instillation of SRM1648. In arteries isolated from SRM1648-instilled animals (12 hr before), relaxation response to ACh was significantly decreased compared with responses obtained in control rats ( Figure 7B). No impairment of ACh-induced relaxation was evidenced after shorter (6 hr) or longer (24 or 72 hr) recovery delay after intratracheal instillation of SRM1648 (data not shown). In contrast, intratracheal instillation of fTiO 2 (5 mg, 12 hr before) did not modify ACh-induced relaxation ( Figure 7B).

Discussion
This study shows, for the first time to the best of our knowledge, that urban PM impairs NOdependent relaxation in small intrapulmonary arteries, not only after in vitro exposure, but Particulate matter and pulmonary circulation Environmental Health Perspectives • VOLUME 116 | NUMBER 10 | October 2008 also after in vivo intratracheal instillation. Manufactured PM, however, including nanoparticles, did not exhibit this effect. The relaxation responses to ACh and NO donors, but not to 8-Br-cGMP, were decreased after exposure of intrapulmonary arteries to SRM1648 (200 µg/mL for 24 hr). In addition to relaxation, ACh-induced cGMP accumulation was also decreased in such conditions. This demonstrates that SRM1648 induced a decrease in responsiveness of smooth muscle to NO, rather than a decrease in endothelial NO production and/or bioactivity, and that SRM1648 impairs the NO signaling pathway upstream of activation of cGMP-dependent protein kinases. YC-1-induced relaxation was not affected by SRM1648, supporting the view that impairment of the NO pathway is likely attributable to a decrease in guanylyl cyclase activation by NO. This argues against a role of decreased expression of guanylyl cyclase in SRM1648-induced impairment of NOdependent relaxation. In addition, this study demonstrates that impairment is selective for the NO pathway, because other relaxant mechanisms (K ATP activation by levcromakalim, adenylyl cyclase activation by forskolin, and β 2 -adrenergic receptor activation by isoproterenol) remained unaffected by SRM1648. Particular core and/or adsorbed constituents may be responsible for PM-induced impairment of NO-dependent relaxation. To address this question, we compared effects of SRM1648 with those of other particles, with different core composition and mean diameter, that are relatively free of adsorbed constituents. Neither carbon black nor TiO 2 particles modified ACh-induced relaxation in rat intrapulmonary arteries. Even though comparison between particle types is difficult, results show that, among particles of similar size range (PM 2.5 ), SRM1648, but not fTiO 2 , decreased ACh-induced relaxation. Thus, adsorbed components of SRM1648, rather than particulate core, are likely responsible for impaired NOdependent relaxation. The water-soluble fraction of inhaled PM is more biologically relevant because its components could more easily reach pulmonary vessels than can whole particles or the insoluble fraction .
Several in vivo studies have demonstrated a decrease in endothelium-dependent relaxation of systemic or pulmonary arteries after exposure to PM (Brook et al. 2002;Nurkiewicz et al. 2004Nurkiewicz et al. , 2006Törnqvist et al. 2007). This study demonstrates that, like in vitro exposure, in vivo exposure to SRM1648 also resulted in a decrease of NO-dependent relaxation in response to ACh in intrapulmonary arteries. Moreover, as in in vitro studies, fTiO 2 failed to alter relaxation to ACh when instilled in animals. This not only argues against a nonspecific response resulting from intratracheal instillation of particles, but also further supports the idea that impairment of NO pathway is caused by adsorbed components of SRM1648, rather than by the particular core. SRM1648 and fTiO 2 have a similar size range, so it seems unlikely that their differential in vivo effects can be attributed to size-related differential penetration in the bronchiolar space. Interestingly, we observed SRM1648-induced decreases in relaxation to ACh 12 hr after instillation, but not after longer delays. Elucidation of the mechanisms underlying this transient aspect of the impairment of ACh-induced relaxation VOLUME 116 | NUMBER 10 | October 2008 • Environmental Health Perspectives   deserves further investigation. Release of antiinflammatory mediators (e.g., transforming growth factor-β or IL-10) may recover or counteract SRM1648-induced alteration of NOdependent relaxation. Because the presence of PM was clearly evidenced in lung parenchyma removed 72 hr after SRM1648 instillation, it is unlikely that recovery is related to elimination of particle deposit from lung parenchyma. Oxidative stress is a major contributor of the adverse effects of PM (Baeza and Marano 2007;Oberdörster et al. 2005). In this study, SRM1648-induced alteration of NO-dependent relaxation was not modified in the presence of the antioxidant tempol. Consistently, intrapulmonary artery exposed to SRM1648 did not display an increase in ROS level. This argues against a role of oxidative stress in the SRM1648-induced impairment of NO-mediated relaxation. This differs from data showing that superoxide dismutase can prevent particleinduced decrease in relaxation to ACh in rat aorta (Ikeda et al. 1995) and that SRM1648 increases production of ROS in pulmonary endothelial cells (Li et al. 2006). Oxidative stress appears as an acute response (within 5-10 min) in endothelial cells (including those from pulmonary artery; Li et al. 2006) or arteries exposed to PM. Even though oxidative stress might be an early event in intrapulmonary arteries exposed to SRM1648, it does not seem to play a role in impairment of NO pathway, because addition of antioxidants concomitantly with SRM1648 failed to prevent impairment of NO-dependent relaxation. Oxidative stress is recognized as a key process underlying endothelial dysfunction in pulmonary arteries (Fresquet et al. 2006). As discussed above, SRM1648 decreased activation of guanylyl cyclase by NO within smooth muscle, a mechanism that may be independent of oxidative stress.
Release of inflammatory mediators is associated with PM-induced impairment of endothelium-dependent vasodilatation in systemic arteries (Nurkiewicz et al. 2006;Törnqvist et al. 2007). In the present study, the anti-inflammatory drug dexamethasone prevented SRM1648-induced impairment of NO-dependent relaxation in intrapulmonary arteries. Moreover, SRM1648 increased the release of proinflammatory mediators (TNF-α, IL-8) from intrapulmonary arteries or endothelial cells. Proinflammatory mediators such as TNF-α are key players in alterations of the NO signaling pathway in the vasculature (Huang and Vita 2006), including in pulmonary arteries (Greenberg et al. 1993). They may induce a decrease in NO, but not cGMP, responsiveness in systemic resistance arteries (Jiménez-Altayó et al. 2006). We show here that SRM1648 also increased IL-8 release. This chemokine is a key mediator in inflammatory pulmonary diseases, not only by attracting neutrophils, but also by acting on vascular cells (Mukaida 2003).
In conclusion, this study shows that urban but not manufactured PM (including nanoparticles) impairs NO-mediated relaxation, without affecting cGMP responsiveness in rat intrapulmonary arteries. We attribute this result to an oxidative-stress-independent inflammatory response, resulting in decreased guanylyl cyclase activation by NO. Such impairment of the NO pathway in pulmonary circulation may favor vasoconstriction, remodeling, and thrombosis, all contributing to enhance arterial resistance, which in turn may have a negative impact on cardiac function.