INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: air pollution; lung diseases; inflammation; neutrophils; bronchoscopy; oxidants; iron

Ambient air pollution particles currently present a serious risk to human health with significant increases in both morbidity and mortality associated with their exposure. The World Health Organization estimates that inhalation of particulate matter (PM) in ambient air is responsible for 500,000 excess deaths each year worldwide (1). However, there has been little description of acute physiologic, biochemical, or molecular alterations after experimental exposure of humans to PM to corroborate epidemiologic evidence of increased morbidity and mortality. Furthermore, there is neither a consensus on a plausible mechanism nor is there any agreement on which component of PM is responsible for biologic activity.

The Utah Valley provided a unique opportunity to evaluate the health effects of PM in humans. The area has had intermittently high particle levels with the principal point source being a steel mill. While operational, this plant contributed greater than 80% of industrially related PM in the valley (2). Because of a labor dispute, the mill was shut down for 13 mo, from August 1, 1986 to September 1, 1987 resulting in a substantial reduction of PM levels in the Utah Valley. Alterations in composition were also assumed. Associated with the closure of the mill, and both the reduction in total PM mass and changes in its composition, were decreases in elementary school absences (3), bronchitis and asthma admissions for pre-school-age children (4), total respiratory hospital admissions for pneumonia, pleurisy, bronchitis, and asthma (2), pulmonary function abnormalities (5), mortality (6), and age-adjusted death rates for malignant and nonmalignant respiratory disease (7). Changes in total mass did not account for all variation in the biologic effects of PM in Utah Valley between those years before the closure of the steel mill, during its shutdown, and following its reopening (6).

The closure and reopening of the steel mill allowed for an examination of potential correlates between epidemiologic observations and measures of the biologic effect of PM with experimental human exposure. We tested the hypothesis that the biological effects of PM collected from January to March of 1986, 1987, and 1988 and instilled into the human lung on an equal mass basis would reflect the findings of the epidemiologic investigation. We also explored the possibility that a metal-catalyzed oxidative stress could contribute to dissimilarities between the effects of exposures to PM from Utah Valley obtained from January to March of 1986, 1987, and 1988.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Corresponding to the times of previous epidemiologic investigation (2), filters containing PM₁₀ were collected from January to March of 1986, 1987, and 1988 (n = 34 each year; Lindon site, Utah Department of Environmental Quality). Each filter was agitated in a 50-ml tube containing 40 ml deionized water for 24 h. After centrifugation, the supernatant fluid from all 34 filters of each year was pooled, lyophilized, weighed, and stored at 80° C. Each of the three extracts was treated with ultraviolet radiation and autoclaved. Extracts were prepared in 1996 and stored at 80° C until used in the investigation. Endotoxin levels of all three extracts were below detectable limits (E-toxin, Sigma, St. Louis, MO).

The study population included 24 healthy, nonsmoking volunteers (21 male, three female) with a mean age of 26.4 yr and a standard deviation of 2.2 yr. The protocol and consent form were approved by the University of North Carolina School of Medicine Committee on the Protection of the Rights of Human Subjects. Prior to inclusion in the study, subjects were informed of the procedures and potential risks, and each signed a statement of informed consent.

The subjects underwent bronchoscopy. A sterile flexible catheter was inserted through the biopsy channel of the bronchoscope and extended 2.5 cm into the orifice of a segmental bronchus of the lingula. Ten milliliters of sterile saline containing 500 µg extract from either 1986 (n = 8), 1987 (n = 8), or 1988 (n = 8) were instilled through the catheter followed by 10.0 ml saline. Twenty milliliters of saline with no particles were similarly instilled into a subsegment of the right middle lobe.

Twenty-four hours later, subjects underwent lavage of the same subsegment in which the extract and the saline (control) had been placed. Cells were counted by hemocytometer and differentials were determined. Lavage protein and albumin concentrations in the lavage fluid were determined using the Pierce Coomassie Plus Protein Assay Reagent (Pierce Chemical Co., Rockford, IL) and an immunoprecipitin assay (Diasorin, Stillwater, MN), respectively. Fibronectin and -1-antitrypsin were measured employing ELISAs developed using antibodies to human fibronectin (Sigma) and antitrypsin (Calbiochem, La Jolla, CA). Concentrations of tissue factor and fibrinogen in the lavage fluid were determined using an ELISA kit (R&D Systems, Minneapolis, MN) and an immunoprecipitin assay (Diasorin), respectively. IL-8, TNF, and IL-1 were measured using ELISA methodology (R&D Systems).

Fibrinogen was stained immunohistologically employing lavaged cells after cytocentrifugation. The primary antibody (rabbit antihuman -fibrinogen antibody; Sigma) was applied at a dilution of 1:100 in PBS. The counterstain employed was hematoxylin.

Oxidant generation by the three extracts was measured using thiobarbituric acid (TBA) reactive products of deoxyribose as an end point.

Aliquots of the aqueous extracts were each agitated in 1.0 N HCl (1.0 mg/1.0 ml) for 1 h at room temperature and centrifuged, and the supernatant was removed for analysis. Metals were individually analyzed employing inductively coupled plasma emission spectroscopy (ICPES, Model P40; Perkin Elmer, Norwalk, CT).

Data are expressed as mean values ± SE. Differences were determined using one-way analysis of variance (ANOVA) and two-way ANOVA (8).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The total mass of PM sequestered on the filters selected for this study was 2,770, 1,324, and 2,132 mg from 1986, 1987, and 1988, whereas the soluble component recovered accounted for 429, 343, and 422 mg, respectively. The levels of ambient PM in the Utah Valley were not notably different from those years immediately preceding the study.

The total number of viable cells in bronchoalveolar lavage fluid (BALF) after instillation of all extracts was increased relative to saline (Figure 1A). The percentage of neutrophils was increased after instillation of PM extracts from 1986 and 1988 relative to that from 1987 and saline (Figure 1C).

View larger version (27K): [in this window] [in a new window]   Figure 1.   A, B, and C. Total number of viable cells and neutrophil counts in the lavage fluid after exposure of healthy volunteers to PM extracts. There was an increase in total viable cells in the lavage of subjects instilled with any extract. Neutrophils were also elevated after exposure of subjects to those extracts collected while the mill was in operation (1986 and 1988). This was significant whether the numbers of neutrophils were expressed as an absolute number or a percentage. Asterisks denote a significant difference in the post-hoc test (Scheffe's) relative to the saline control with p < 0.05.

A neutrophilic influx into the lung can frequently be associated with a lung injury presumed to result, in part, from a release of proteases and endogenous oxidants from this cell. Compared with saline controls and extract from 1987, BALF protein and albumin concentrations were greater after instillation of PM extracts from 1986 and 1988 obtained when the steel mill was in operation (Figures 2A and B). Elevations in concentrations of the glycoproteins fibronectin and -1-antitrypsin in BALF are also reflective of a lung injury. Elevations of both were significantly greater after instillations of extracts from filters obtained during 1986 and 1988 relative to that from 1987 and the saline control (Figure 2C and D).

View larger version (38K): [in this window] [in a new window]   Figure 2.   A, B, C, and D. Lung injury indices after exposure of healthy volunteers to PM extracts. There was a significant increase in all indices of lung injury in those volunteers exposed to extracts collected while the steel mill was functioning (1986 and 1988). Asterisks denote a significant difference in the post-hoc test (Scheffe's) relative to the saline control with p < 0.05.

The equilibrium between mechanisms of thrombosis and fibrinolysis that normally exists in the alveolar epithelial lining fluid can be disrupted in an injury (e.g., adult respiratory distress syndrome and asbestos exposures) (9, 10). A subsequent activation of coagulation pathways with fibrin formation in the lung can amplify an inflammatory response by providing a chemotactic signal for neutrophils. Staining of alveolar macrophages recovered in BALF for fibrinogen demonstrated an increased binding of the antibody after exposure to an extract relative to saline (Figure 3A and B). There were also decreases in the concentrations of tissue factor and fibrinogen in the BALF after instillation of extracts from 1986 and 1988 relative to both 1987 and saline (Figure 3C and D). This decrease likely reflects a deposition of fibrinogen within the alveolar compartment after exposure to the extracts with clearance through binding by the macrophages (11).

View larger version (132K): [in this window] [in a new window]   View larger version (139K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   Figure 3.   A, B, C, and D. Activation of coagulation in the lungs of healthy volunteers exposed to filter extracts. Immunohistochemistry demonstrates an increased concentration of fibrinogen in those cells collected from subjects instilled with a PM extract (1986) (B) relative to those exposed to saline (A). Tissue factor and fibrinogen were both decreased in those subjects instilled with extracts collected while the steel mill was in operation (1986 and 1988) relative to saline. Asterisks denote a significant difference in the post-hoc test (Scheffe's) relative to the saline control with p < 0.05.

An influx of neutrophils into a tissue can be directed by numerous chemotactic mediators. Interleukin (IL)-8, tumor necrosis factor (TNF), and IL-1 are among those cytokines with such a capacity and their in vitro expression by respiratory cells can be effected by particle exposure (12). Concentrations of IL-8, TNF, and IL-1 in BALF were significantly elevated in those subjects instilled with extracts from particles collected while the mill was operating relative to that acquired during shutdown (Figures 4A-C).

View larger version (37K): [in this window] [in a new window]   Figure 4.   A, B, and C. Concentrations of proinflammatory mediators in the BALF of healthy volunteers exposed to PM extracts. Concentrations of the cytokines were elevated after exposure to extracts obtained from Utah Valley while the mill was in operation (1986 and 1988) relative to saline instillation. Asterisks denote significant difference in the post-hoc test (Scheffe's) relative to the saline control with p < 0.05.

Exposures were repeated in a smaller number of volunteers (n = 2/filter extract) but at a lower mass (100 µg instilled). Lavage collected 24 h later again demonstrated an inflammatory injury after instillation of that extract collected while the steel mill was functioning (Table 1).

An inflammatory response can be coordinated through an oxidant-sensitive activation of specific transcription factors. These factors influence cytokine and other mediator expression by binding promoter sites and affecting mRNA transcription (13). The in vitro oxidant generation by each PM extract (500 µg) was measured as the thiobarbituric acid (TBA) reactive products of deoxyribose. Extracts from 1986 and 1988 produced a significantly greater oxidative stress compared with extract from 1987 and saline (Figure 5). DMTU diminished the absorbance of TBA-reactive products while the metal chelator deferoxamine inhibited all oxidant generation by the three extracts, suggesting that metals included in the PM could participate in the catalysis of reactive oxygen intermediates through their support of electron transport. Subsequently, metals in the PM extracts were quantified. Concentrations were increased in the PM extracts from 1986 and 1988 obtained while the steel mill was in operation relative to values from 1987 procured during its closure (Table 2).

View larger version (27K): [in this window] [in a new window]   Figure 5.   Thiobarbituric (TBA) reactive products of deoxyribose after exposure to PM extracts. The absorbance of oxidized products of deoxyribose was significantly greater after incubation with extracts collected while the steel mill was functioning (1986 and 1988). The absorbance of oxidized products after incubations with all three extracts was significantly greater than that demonstrated by saline. Both dimethylthiourea (DMTU) and deferoxamine diminished the absorbance of oxidized products of TBA.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Volunteers instilled with aqueous extracts of PM filters while the steel mill was open had significant increases in both lung inflammation and injury in contrast to those volunteers instilled with an equal mass of PM extract from filters collected while the mill was closed. The increased biologic effect of the PM extract collected while the mill was open is analogous to measures of human morbidity and mortality that were similarly elevated during this same time period. These findings suggest that mass may not be the most appropriate metric to use in assessing health effects after PM exposure, but rather specific components must be identified and assessed.

In this investigation, 500 µg of extract were instilled into the lingula, which comprises about 10% of total lung volume. An active person might have an average ventilation rate of 15 L/min and thus inspire 21.6 m³ of air per day. During temperature inversions that occur during the winter in the Utah Valley, levels of PM₁₀ can exceed 100 µg/m³. Subsequently, a person could inhale 2,160 µg PM₁₀ in a 24-h period. Approximately 10%, or 216 µg, could be distributed to the lingula. Assuming an average deposition efficiency of 42%, 91 µg PM₁₀ would be deposited in the lobe. This is roughly 5.0 times less than what we instilled. Furthermore, 100 µg of filter extract (collected from January to March 1986) instilled into the lingula of volunteers also elevated levels of neutrophils, protein, and inflammatory cytokines. These calculations suggest that the biologic effects observed in this study could be experienced by persons during a typical winter inversion in the Utah Valley.

After the instillation of PM extract from Utah Valley collected during the operation of steel mill, the number of neutrophils found in the BAL fluid was more than five times higher than that of heavily exercising humans exposed to ozone at concentrations four times higher than the current National Ambient Air Quality Standard (14). Indeed, no other ambient air pollutant has been reported to cause this degree of inflammation (15).

Episodes of temperature inversion are common in the Utah Valley in the winter months. During these episodes, local emissions become trapped in a stagnant mass near the valley floor and PM concentrations become elevated. Biologic monitors (e.g., lichens) confirm that metal concentrations in Utah Valley were extremely high (16). In addition, elevations in metal concentrations are documented by direct monitoring of PM (17). These data support a release of metals by the steel mill. In the investigation, the closure of the mill is associated with decreased concentrations of metals in aqueous extracts of the filters. Metal content, and consequent oxidative stress that paralleled metal concentrations, are potential contributors to the dissimilar biologic effect of the three extracts used in this investigation. Cellular and tissue exposure to metal-catalyzed oxidants can promote activation of both cell signaling and transcription factors, which influence release of inflammatory mediators. The product of this cascade of reactions is an inflammatory injury.

The possibility that the disparate effects of PM extracts observed in this study resulted from physical or chemical differences other than metal content and catalyzed oxidants cannot be excluded. Because a soluble extract was studied and not the original particles themselves, the influence of size, shape, and insoluble components in modulating health effects in the Utah Valley remain unknown.

This investigation took advantage of an unusual situation to demonstrate for the first time a correlation between findings of previous epidemiologic studies and the biologic effects of PM in humans. In addition, the results of this study indicate that equal masses of PM can induce disparate lung injuries, suggesting that particle components may be relevant in assessing health effects after their exposure. Specifically, metals can participate in the biologic effects of PM collected from the Utah Valley.