INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Epidemiologic studies have identified significant associations between ambient air particles, especially particles with a mass median diameter of less than 10 µm (PM₁₀), and increased cardiopulmonary morbidity and mortality (1). The biologic mechanisms responsible for this particle-induced increase in morbidity and mortality are unclear. Seaton and colleagues postulated that the inhalation of fine particles provokes a low-grade inflammatory response in the lung that aggravates lung disease and a change in blood coagulability that increased pulmonary and cardiovascular deaths (4). Alveolar macrophages (AM) are the most likely link between the inflammatory process in the lung and the systemic response because they are cells responsible for ingesting and clearing inhaled particles (5). The interaction of AM with atmospheric particles increases their phagocytic activity, oxidant production, and the release of inflammatory mediators such as tumor necrosis factor-alpha (TNF-) (6, 7).

An intriguing aspect of the epidemiologic data is that the particle health effects are also associated with disease of the cardiovascular system (1) that does not come in direct contact with the particles. Studies from our laboratory have shown that particle exposure causes a leukocytosis in humans (8) and animals (9), suggesting that a systemic inflammatory response is a feature of breathing polluted air. The instillation of mediators from AM exposed to PM₁₀ ex vivo produced a bone marrow response similar to that produced by instilling the particles themselves into the lung (9, 10). This suggests that AM are capable of initiating both a local and systemic inflammatory response when PM₁₀ are deposited in the lung.

The activation and mobilization of inflammatory cells (9, 10), the production of acute-phase proteins (11), and the production of circulating inflammatory mediators characterize the systemic inflammatory response. An integral component of this response is stimulation of the hematopoietic system, specifically the bone marrow that results in an increase in circulating leukocytes. Several large population-based studies have shown this level of leukocytosis is a predictor of total mortality, independent of smoking (12, 13). Military recruits exposed to high concentrations of particulate matter air pollution during the forest fires in Southeast Asia in 1997, developed leukocytosis that was associated with bone marrow stimulation (8). We postulate that atmospheric particles stimulate AM to produce mediators that are capable of eliciting a systemic inflammatory response.

The present study was designed to survey the cytokines produced by human AM exposed to particles of different kinds (14- 16) and determine which of these cytokines were detectable in the blood of subjects during the Southeast Asian haze of 1997.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Particles

The urban air dust preparation, EHC 93, was obtained from Environmental Health Directorate, Health Canada, Ottawa, Ontario (15, 16). Residual oil fly ash (ROFA) was a kind donation by Dr. Dan Costa, Environmental Protection Agency, NC. Latex beads of different sizes (< 0.1 µm, ~ 1 µm and ~ 10 µm) were obtained from Sigma (St. Louis, MO) and inert carbon particles from Fount India Drawing Ink (Pelikan, Germany) (see details in the online data supplement).

AM

Human AM were harvested from bronchoalveolar lavage (BAL) fluid from a noninvolved segment or lobe of lungs resected for small peripheral tumors (n = 16). Human AM harvested were more than 90% viable (trypan blue exclusion); cells consisted of 98% AM and 2% lymphocytes and neutrophils. All specimens were tested for endotoxin contamination using the Limulus amebocyte lysate test (E-TOXATE; Sigma Chemical Co., St. Louis, MO) and rejected if positive. The particles (0.01 to 0.1 mg/ml) were suspended in RPMI-1640 and 10% fetal calf serum, sonicated, and gently mixed for 20 min. Human AM (0.5 × 10⁶/ml) were incubated with particles in a time (2, 4, 8, 12, and 24 h) and concentration fashion at 37° C in 5% CO₂ in a 24-well plate. AM from the same subject was used as a control and incubated with the vehicle without the particles. Supernatants were harvested, filtered through 0.2-µm filters to remove suspended particles, and centrifuged at 10,000 rpm for 10 min before storage at 70° C.

Circulating Cytokine in PM₁₀-exposed Subjects

Serum samples were collected from 30 healthy males (age 19 to 24) who performed compulsory national service in Singapore during the forest fires in Southeast Asia in 1997. The details of this exposure are described elsewhere (8). Samples were collected during the haze period (September-October) and after the haze had cleared (November-December). Serum samples were frozen and kept at 70° C and cytokine concentrations measured in batch by ELISA (see details in online data supplement).

Statistical Analysis

All values are expressed as mean ± SEM. Differences in TNF- with different doses of particles were evaluated with a one-way analysis of variance (ANOVA) followed by Bonferroni correction for multiple comparisons. Differences between control and particle-exposed AM cytokine production were analyzed using a two-way paired t test and corrections were done for multiple comparisons. A value of p < 0.05 was accepted as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Particles

The elemental composition of the EHC 93 particles was previously reported by Vincent and colleagues (15). EHC 93 particles contain small amounts of endotoxin (7, 15, 17), but the dose used in this study (0.1 mg/ml) contains just trace amounts of endotoxin (< 2 ng), and this dose has been shown not to cause either a local or systemic effect when instilled into the lung of rabbits (10).

Subjects

Table 1 summarizes the demographic data of subjects from whom the AM were harvested for exposure to EHC 93. Subjects were divided into current smokers and nonsmokers (four never-smokers and one subject who stopped 22 yr before surgery). All the subjects were advised to stop smoking at least 4 wk before the surgery. Diffusing capacity of the lungs for carbon monoxide (DL_CO) was reduced to 67 ± 5.1% of predicted in the smokers.

Phagocytosis of Particles by AM

AM incubated with the smaller latex beads (0.1 and 1 µm) internalize the smaller beads and more than 80% were fluorescent-labeled. The larger beads (10 µm) just adhere to the cells. Particles were visible in 25 ± 6% of AM incubated with a low dose (0.01 mg/ml) of EHC 93 and 56 ± 9% with a high dose (0.1 mg/ml) of EHC 93 showing a dose-dependent internalization of particles. Both EHC 93 and ROFA in doses higher than 0.1 mg/ml are toxic to AM (reduced trypan blue exclusion and TNF- production, data not shown).

TNF- Produced by AM

AM stimulated with either lipopolysaccharide (LPS) or incubated with particles have maximum TNF- production after a 24-h culture period, and all values are from supernatants collected after 24 h. Incubation of AM with LPS causes a dose-dependent increase in TNF- production (see Figure E1 in the online data supplement). AM incubated with latex beads (n = 4) produced a small but dose-dependent increase of TNF- that was independent of the size of the beads (0.1 µm = 121 ± 8%, 1 µm = 158 ± 31%, and 10 µm = 141 ± 20%, percent change from control). EHC 93 caused a dose-dependent increase in TNF- by AM (Figure 1a, n = 5) that was also seen with ROFA (n = 7) and inert carbon particles (n = 3, data not shown). Figure 1b shows the maximum TNF- response produced by the different particles (0.1 mg/ml) in comparison to AM incubated with 1 µg/ml LPS.

View larger version (21K): [in this window] [in a new window]   Figure 1.   (a) TNF- production by AM when incubated for 24 h with EHC 93 (0.01 to 0.1 mg/ml). (b) TNF- production by human AM exposed to different particles, latex beads of 1 µm size (n = 4), EHC 93 (n = 16), inert carbon (n = 3), and ROFA (n = 7) all at a dose of 0.1 mg/ml. LPS (1 µg/ml) was used as positive control. All particles caused an increase in TNF- production compared with control medium but EHC 93 caused significantly higher TNF- production than the other particles (#p < 0.05 versus control medium, *p < 0.01 versus EHC 93). (c, d ) Cytokines production by human AM exposed to ambient urban air particles EHC 93 (0.1 mg/ml). EHC 93 induced several cytokines with a significant increase in IL-6, IL-1, MIP-1, and GM-CSF in smokers (Panel d, n = 11) and IL-6, IL-1, and GM-CSF in nonsmokers (Panel c, n = 5) (*p < 0.05 versus control medium). Values for IL-8, IL-6, MIP-1, and TNF- are × 100 pg/ml and all the other cytokines are × 10 pg/ml. Values are mean ± SEM.

Cytokines produced by AM Incubated with EHC 93

Figures 1c and 1d show cytokines produced by human AM incubated with 0.1 mg/ml EHC 93 for 24 h. These results show a generalized increase in all cytokine produced by AM in both the smoker (Figure 1d) and the nonsmoker (Figure 1c) groups. The magnitude of increase was similar in smokers and nonsmokers. There were significantly higher values for granulocyte macrophage colony-stimulating factor (GM-CSF), interleukin-6 (IL-6), interleukin-1 (IL-1), and macrophage inflammatory protein-1 (MIP-1) in the AM exposed to particles in the smokers and GM-CSF, IL-6, and IL-1 in the nonsmokers.

Circulating Cytokines in PM₁₀-exposed Subjects

All subjects were male, seven (23%) were smokers, and the average age was 21 ± 1.7 yr. Air pollution and blood cell count data during and after the haze were previously published (8). The predominant pollutant during the haze period was PM₁₀ that showed an increase of approximately 300% (125.4 ± 44.9 versus 40 ± 14.3 µg/m³, haze versus posthaze). Other air pollutants such as SO₂ (~ 190%), NO₂ (~ 129%), O₃ (163%), and CO (~ 131%) were also increased but to a lesser extent (8).

Figure 2a shows the concentrations of TNF, IL-1, IL-6, and GM-CSF in the serum samples. Only values that were in the detection range of the assay were used (TNF- 3.9 to 250 pg/ml, IL-1 0.781 to 50 pg/ml, IL-6 1.56 to 100 pg/ml, and GM-CSF 2.344 to 150 pg/ml). IL-1 concentrations in the blood significantly decreased after clearing of the haze (p < 0.05) with a trend toward lower concentrations of IL-6 (p = 0.09) and GM-CSF (p = 0.07). Figures 2b and 2c show the relationship between IL-1 and PM₁₀ concentrations (Figure 2b) as well as IL-6 and PM₁₀ concentrations (Figure 2c) during and after the haze. With clearing of the haze and a decrease in the PM₁₀ concentrations (p < 0.001) both the circulating IL-1 and the IL-6 decreased. TNF- was similar during and after clearing of the haze.

View larger version (13K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Figure 2.   (a) Cytokines in the blood of subject during the Southeast Asia forest fires of 1997. The solid bars represent the concentrations of cytokines in the serum during the haze period and the hatched bars after the haze cleared. Values are mean ± SEM of all samples (pre- and posthaze) with values within the detection limit of the assay (n = 27) (*p < 0.05 versus posthaze, ** p < 0.1 versus posthaze). (b, c) Relationship between ambient PM10 levels and circulating cytokine levels during and after the 1997 Southeast Asia haze. PM10 concentrations were significantly lower with clearing of the haze (posthaze, p < 0.001) and correlated with a decrease in IL-1 (p < 0.05) and IL-6 (p < 0.1). Values are mean ± SEM of 27 subjects.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results show that when human AM phagocytose atmospheric particles they produce TNF- in a dose-dependent manner. When incubated with an optimal dose of atmospheric particles (EHC 93) they also produced several other proinflammatory cytokines, particularly GM-CSF, IL-6, and IL-1. These cytokines are known to stimulate the bone marrow to produce and release leukocytes and platelets into the circulation (18) and stimulate the production of acute-phase proteins (11, 18). Several of these cytokines were higher in the blood of subjects during an acute episode of air pollution and decreased in the period after the pollution had cleared. We postulate that these cytokines found in the blood were produced by AM during phagocytosis of particles deposited in the lung. We suspect that the systemic effects of particle exposure are important in the pathogenesis of the cardiopulmonary disease associated with particulate air pollution exposure.

AM are important in processing airborne particles and secrete mediators that are capable of eliciting both a local and a systemic inflammatory response (5). We used TNF- as a marker of the magnitude of AM stimulation by PM₁₀ and showed that incubation of human AM with different types of particles induced a dose-dependent increase of TNF- production that was the most intense after stimulation with ambient particles collected over a major Canadian city (EHC 93). This finding is consistent with previous reports (7, 10, 19) using different urban air pollutants. A dose of particles (EHC 93) that elicited an optimal TNF response (0.1 mg/ml) was used to investigate the nature of other cytokines produced by AM after particle exposure. In the group data (Figure 1c), the TNF response induced by an optimal dose of PM₁₀ was not statistically different from controls. This was due to a large variability in the TNF response between subject both in vitro (Figures 1c and 1d) and in vivo (Figure 2a). The endotoxin that contaminated the particles was not responsible for this cytokine production because the particles dose used (0.1 mg/ml) contained a very small amount of endotoxin (< 2 ng) but produced a TNF response equal to 1 µg of LPS in vitro.

Using EHC 93 to stimulate AM, we surveyed those cytokines that are known to have systemic effects and all these cytokines were increased in the supernatants of AM exposed to particles. In smokers GM-CSF, IL-6, IL-1, and MIP-1 and in the nonsmokers GM-CSF, IL-6, and IL-1 were significantly higher in particle-exposed compared with vehicle-exposed AM (Figures 1c and 1d). Cigarette smoking is known to stimulate AM (20) and might confound the results we obtained in this study. However, the similar results obtained from macrophages of smokers and nonsmokers suggest that PM₁₀ have an independent effect on AM cytokine production. Furthermore, seven of the 30 subjects studied during Southeast Asia haze were smokers and their bone marrow response was similar to the nonsmokers (8), indicating that the effect of PM₁₀ is independent of the effect of cigarette smoking.

The cytokines produced by AM exposed to PM₁₀ could be of particular importance inducing the inflammatory response associated with PM₁₀ exposure. GM-CSF is a hematopoietic growth factor that stimulates granulocyte and monocyte differentiation and release from the bone marrow but also activates circulating leukocytes and prolongs their survival in the circulation (18). GM-CSF has also recently been identified to be an important granulocyte degranulation factor that may enhance tissue damage induced by granulocytes (21). IL-1 is one of the "acute response" cytokines that induces cytokine production by many cells, stimulates hematopoiesis, activates endothelial cells, is pyrogenic, and induces the acute-phase response (18). IL-6 stimulates the liver to produce acute-phase proteins such as C-reactive protein, fibrinogen, and antiproteases by hepatic cells (11, 18); stimulates hematopoiesis, specifically the production of platelets; and has a broad stimulating effect on B and T cells (18). We have recently reported that IL-6 accelerates the transit time of granulocytes through the bone marrow, releases them into the circulation, and promotes their sequestration in microvascular beds (22). We also measured IL-10, a cytokine known to inhibit the production of proinflammatory cytokines, TNF-, IL-1, IL-6, and IL-8 from endotoxin-stimulated human monocytes (23). The concentrations of IL-10 produced by AM exposed to PM₁₀ did not differ from controls (Figures 1c and 1d), suggesting that PM₁₀ do not induce a significant anti-inflammatory cytokine response. Collectively these three cytokines have the ability to elicit a systemic inflammatory response characterized by an increase in circulating leukocytes, platelets, proinflammatory and prothrombotic proteins. They also have the ability to activate circulating leukocytes and the endothelium of the vascular bed to promote leukocyte-endothelial adhesion and migration.

Our study demonstrates that the same cytokines produced by AM exposed to urban particles (Figure 2a) were present in serum samples collected from subjects during the haze period and that their concentrations decreased after the haze cleared. This decrease in circulating cytokines correlates best with a decrease in PM₁₀ levels (Figures 2b and 2c), suggesting that the particulate component of air pollution is important in the cytokine response induced by air pollution. Our data show that cytokines produced in the lung are present in the circulation, and we postulate that these cytokines contribute to the systemic response elicited by air pollution. Further studies are required to establish a link between circulating cytokines and the adverse cardiopulmonary effects observed in susceptible individuals exposed to air pollution.

Bioavailable metals, biogenic, organic components, and the physical characteristics of particles have each been proposed to be important in the PM₁₀-induced inflammatory response in the lung (14, 16, 24). We have used four different particles with different physical characteristics and different organic and inorganic compositions. These different particles have no significant difference in their ability to elicit a TNF response in AM (Figures 1b and 1c). Although the AM were able to adhere but not engulf the larger latex particles (10 µm size), they still produced a dose-dependent TNF response similar to AM that phagocytose smaller latex particles. This suggests that adhesion of particles to AM may initiate the signal transduction pathways responsible for cytokine production. Both ROFA (14) and EHC 93 (15) contain transitional metals that caused pulmonary injury in healthy animals (14, 16). An increase in reactive oxygen intermediates generated by these metals has been implicated in this tissue damage response. These oxygen intermediates also activate nuclear factor (NF)-B and activator protein (AP)-1, which in turn are responsible for the subsequent induction of various cytokine genes (25). Our results show that ambient particles activate a broad range of cytokine genes particularly cytokines that have the ability to induce a systemic response.

In summary, our results show that AM that phagocytoses ambient particulate matter result in a dose-dependent increase in TNF- production and that particles with different composition and size produce a similar response. The AM exposed to particles also produced a broad range of other cytokines with high concentrations of IL-6, IL-1, and GM-CSF, and these cytokines were present in the circulation of subjects during an episode of acute air pollution. The systemic inflammatory response that they elicit includes the production of proinflammatory acute-phase proteins and stimulation of the bone marrow and could be important in the pathogenesis of the pulmonary and cardiovascular disease associated with particulate air pollution.