Sulfuric Acid Aerosol Exposure in Humans Assessed by Bronchoalveolar Lavage 1-3

MARK W. FRAMPTON, KAREN Z. VOTER, PAUL E. MORROW, NORBERT J. ROBERTS, JR., DAVID J. CULP, CHRISTOPHER COX, and MARK J. UTELL With technical assistance from Harriet Beiter, F. Raymond Gibb, Joan Nichols, Donna Speers, and Ying Tsai

Introduction

Pollutants released above inversion layers from the tall stacks of utility plants and metal smelting operations lead to the generation of acidic aerosols. These aerosols, predominantly sulfuric acid (H 2S04 ) , remain suspended and are carried for hundreds of miles, where they cause environmental effects as "acid rain" in the northeastern United States and Canada (1). Recent measurements have demonstrated 24-h average H 2S04 concentrations exceeding20 ug/m" and peak concentrations reaching 100 ug/m" (2). Epidemiologic and experimental evidence suggests acidic aerosols may have important human health effects. For example, Bates and Sizto (3) observed a relationship betweendaily ozone, sulfate, and temperature and admissions to hospitals for respiratory illness in Ontario, Canada. Concentrations of total sulfate are considered a surrogate for acidic sulfates, suggesting that sulfuric acid aerosols contribute to respiratory illness. This conclusion is supported by data from the Harvard six cities studies, in which respiratory symptoms in children were found to be associated more closely with atmospheric hydrogen ion concentration than with concentrations of total particulate matter in the communities studied (4). Also, acidic aerosols may potentiate the effects of exposure to atmospheric oxidants such as ozone and nitrogen dioxide (N02 ) (5, 6). Controlled experimental exposures have been used to examine the respiratory effects resulting from the inhalation of acidic aerosols, with measured outcomes limited almost exclusively to symptoms and airway function. Although exposures to concentrations of H 2S04 below 1,000 ug/m" generally do not alter airway function of normal subjects (7), Utell and colleagues (8) observed an increase in throat irritation and 626

SUMMARY Epidemiologic and experimental evidence suggests that exposure to acidic aerosols may affect human health. Brief exposures to acidic aerosols alter mucoclllary clearance and Increase airway responsiveness, but effects on host defense mechanisms at the alveolar level have not been studied In humans. 1Welvehealthy, nonsmoking volunteers between 20 and 39 yr of age were exposed for 2 h to aerosols of approximately 1,000 J,Ag/m 3 sulfuric acid (H2SO.) or sodium chloride (NaCI [control]), with Intermittent exercise, In a randomized, double-blind fashion. Each SUbject received both exposures, separated by at least 2 wk. Bronchoalveolar lavage (BAL) was performed 18 h after exposure In order to detect evidence of an Inflammatory response, changes In alveolar cell subpopulatlons, or changes In alveolar macrophage (AM) function, which Is Important In host defense. When compared with NaCl, exposure to H2SO. did not Increase polymorphonuclear leukocytes In BAL fluid. The percentage of T lymphocytes decreased In association with H2SO. exposure, but the difference was not statistically significant (14.9% after NaCl, 11.5% after H2SO.i p = 0.14). Antibody-mediated cytotoxicity of AM Increased In association with H2SO. exposure (percent lysis 19.1 after NaCl, 23.6 after H2SO.i p = 0.16). No sign Iflcant change was seen In release of superoxlde anion or Inactivation of Influenza virus In vitro. Brief exposures to H2SO. aerosol at 1,000 J,Ag/m 3 do not cause an Influx of Inflammatory cells Into the alveolar space, and no evidence was found for alteration In antimicrobial defense 18 h after exposure. AM REV RESPIR DIS 1992; 148:828-832

airway reactivity to carbachol inhalation in normal volunteers 24 h after exposure to H 2S04 at 450 ug/m", Asthmatic individuals demonstrate a much greater sensitivity to the effects of acidic aerosols, with adolescent asthmatics showing decrements in pulmonary function after exposure to concentrations as low as 68 ug/m" (9). The delayed increase in airway reactivity seen in normal subjects could be caused in part by airway inflammation in response to the exposure. Similarly, the increased sensitivity of persons with asthma may be related to the underlying airway inflammation known to characterize this disorder. It is important to consider whether repeated episodes of distal airway inflammation could contribute to the health effects of exposure to acidic aerosols. Single, brief exposures to ozone have been shown to result in distal airway inflammatory responses (10), but similar studies have not been reported previously with regard to acidic aerosols. Exposure to H 2S04 aerosols alters mucociliary clearance in both animals and humans (11), with speeding of clear-

ance occurring at concentrations of approximately 100 to 200 ug/m" and slowing of clearance at approximately 1,000 ug/rn". Because atmospheric H 2S04 aerosols consist of submicronic particles with significant deposition in the alveolar region of the lungs, exposure may also be expected to affect alveolar clearance. Indeed, animal exposure studies have suggested that exposure to acidic aerosols alters alveolar clearance (12). The cell

(Receivedin originalform November 22, 1991 and in revisedform March 16, 1992) 1 From the Departments of Medicine, Pediatrics, Biophysics, Dental Research, and Biostatistics, University of Rochester School of Medicine and Dentistry, Rochester, New York. 2 Supported by the Electric Power Research Institute, by Grant No. ROI ES02679 from the National Institutes of Health, and by Grant No, RROOO44 from the Division of Research Resources. 3 Correspondence and requests for reprints should be addressed to Mark W. Frampton, M.D., Pulmonary Unit - Box 692, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, NY 14642-8692.

627

ACIDIC AEROSOL EXPOSURE

primarily responsible for alveolar clearance is the alveolar macrophage (AM), assisted by the polymorphonuclear leukocyte (PMN) during an inflammatory response. Exposure to H 2S04 aerosols impaired AM phagocytic function (12) and release of interferon following a viral challenge in vitro (13). Furthermore, rabbits exposed to H 2S04 at 1,000JA,g/m3 for 1 h demonstrated greater recovery of PMN by bronchoalveolar lavage (BAL) 12to 24 h after exposure than from shamexposed animals (14), although repeated f-h exposures to 250 JA,g/m3 H 2S04 did not yield evidence of inflammatory cell infiltration (15). These studies suggest that exposure to acidic aerosols alters respiratory defense at the alveolar level. Effects of exposure to acidic aerosols on alveolar inflammation and AM function have not been previously studied in humans. The current study was undertaken to determine whether single exposures to H 2S04 aerosols cause cellular responses in the alveolar space important in host defense. The exposure concentration was 1,000 JA,g/m3 , the maximum level adopted by the Occupational Safety and Health Administration for exposures in the workplace. The same aerosol concentration of sodium chloride (NaCl) was used as a negative control. BAL was performed 18 h after exposure because epidemiologic (16), animal (14), and human (8) studies suggest a delayed onset of health effects after exposure to acidic aerosols. In addition, this timing provided a basis for comparison with observations following exposure to ozone, in which an influx of inflammatory cells is evident 18 h after exposure (10). Three aspects of the response to exposure were studied: influx of inflammatory cells, alteration in the distribution of AM and lymphocyte subpopulations, and changes in AM function, which is important in antimicrobial host defense.Studies of antimicrobial function included release of superoxide anion (0 2 -), antibodydependent cellular cytotoxicity, and inactivation of influenza virus. Methods

Subjects These studies were approved by the Research Subjects Review Board at the University of Rochester, New York. 1\velvehealthy volunteers between the ages of 20 and 39 yr (mean ± SD, 25 ± 5 yr) participated in these studies after giving informed consent. The population included two women and ten men. All were lifetime nonsmokers and gave no history of chronic respiratory disease nor recent

upper respiratory infection. All had normal physical examinations and baseline pulmonary function as well as absence of airway hyperreactivityto inhaled carbachol. Methods for airwaychallenge with carbachol have been described previously (17). Subjects who demonstrated a > 400/0 decrease in specific airway conductance at any concentration of carbachol were excluded from the study.

Facilities Exposures were performed in a 45-m3 environmental chamber in the Clinical Research Center at the University of Rochester. Characteristics of this facility have been described previously (18). Pulmonary function testing equipment and an exercise bicycle ergometer are housed in the main exposure room. For these studies, temperature and relative humidity in the chamber were maintained at 22 ± 10 C and 40 ± 5%, respectively. Background levels of ozone, nitrogen oxides, sulfur oxides, and particles were continuously monitored during exposures and found to be ~ 0.005 ppm, 0.01 ppm, 0.005 ppm, and 4 JJg/m3 , respectively. Amounts of ammonia in the occupied and unoccupied chamber were < 100 ppb. Aerosol generation was accomplished using a modified Dautrebande D-30nebulizer (R. E. Reynolds Co., Rochester, NY) assembled as an eight-jet unit fed from a lO-L reservoir. Output from the generator could be regulated by adjusting the operating pressure without significantly altering the aerosol particle size. The aerosol was passed through a 85Kr deionizer (TSI Inc, St. Paul, MN), mixed with filtered air from the air-conditioned ventilation system of the hospital, and distributed through five ceiling diffusers into the chamber. Sulfate and NaCl aerosol mass concentration was monitored by nephelometry (MIE Inc., Bedford, MA), and by collection on 1 JJmFluoropore filters (Millipore, Bedford, MA) for subsequent analysis by ion chromatography (Model 16;Dionex, Sunnyvale, CAl. Aerodynamic mass distribution was measured on an eight-stage inertial classifier (Model 02100;IN-lOX, Albuquerque, NM). The aerosol generated for these studies had an aver-

age mass median aerodynamic diameter (MMAD) of 0.9 JJm and a geometric standard deviation (GSD) of 1.9.

Protocol The study protocol is summarized in figure 1. Volunteers were exposed for 2 h to aerosols of 1,000 JJg/m3 H 2S04 or NaCl (control) in a randomized, double-blind fashion. The alternate exposure, NaCl or H 2S04 , was performed at least 2 wk later. Exercise was performed on a bicycle ergometer for 10 min of each half-hour at a work load sufficient to produce a minute ventilation (VE)of approximately 40 L/min. Minute ventilation was monitored by inductive plethysmography (Respitrace Model300C; Ambulatory Monitoring, Ardsley, NY). Pulmonary function was measured in the exposure chamber before, immediately following, and 18 h after exposure. Testing included thoracic gas volume (TOV) and airwayresistancedetermined during panting in an integrated flow, pressure-corrected body plethysmograph, and measurement of FVC and FEV 1 using a pneumotachograph (ModelllooA; Medical Graphics, St. Paul, MN). The best performance of at least two measurements was chosen. Specific airway conductance (SGaw) was determined as the reciprocal of airway resistance (corrected for the resistanceof the pneumotachograph) multiplied by TOV. In order to confirm the nonsmoking status of the subjects, saliva was collected before each exposure, stored at -70 0 C, and analyzed for cotinine by gas chromatography (courtesy of Dr. Neal Benowitz, San Francisco, CAl. Subjects brushed their teeth and rinsed with mouthwash before and midway through each exposure in order to minimize the neutralization of the acidic aerosol by oral ammonia (19).Subjects were polled by questionnaire immediately after and 18 h after each exposure regarding respiratory symptoms, nasal or eye irritation, and ability to detect an odor in the atmosphere. BAL was performed 18h after both H 2S04 and NaCl aerosol exposures. Under topical anesthesia with lidocaine, a fiberoptic bronchoscope (Model FB-19H; Pentax, Orange-

EXPOSURE PROTOCOL

Fig. 1. Experimental protocol. MRinse" indicates useof lemongargleandtooth brushing to minimize oral ammonia. "PFTs" denotes measurement of pulmonary function. Solid bars indicateexerciseperiods (ex1 to eX4). Bronchoalveolar lavage(BAL) wasperformed18h af· ter exposure.

Rinse

Rinse

l

!

PFTs

PFTs

!

! ex1

ex2

ex3

ex4

\ o

1 TIME (hrs)

2

I 18

628 burg, NY) was inserted orally and gently wedged in a subsegment of the lingula. Three 50-ml aliquots of sterile normal saline were instilled and withdrawn under gentle suction. The bronchoscope was then repositioned in a subsegment of the right middle lobe and the lavage repeated. All fluids were collected on iceand immediately transported to the laboratory for processing. After separation of cellsby centrifugation, total protein was measured on the supernatant fluids using the method of Lowry and coworkers (20), with crystalline bovine serum albumin as the standard. Albumin was measured using a modified ELISA as described previously (21). All determinations were performed concurrentlyon samples obtained after exposure to NaCI and H 2S04 ,

Cell Quantitation and Characterization Analysis of cells recovered by BAL was designed to detect influx of inflammatory cells or changes in the distribution of alveolar cell subpopulations in response to the exposure. Total cell counts were performed on fluids obtained from the right middle lobe and the lingula separately, using a hemocytometer, to provide an assessment of regional variability of any inflammatory response. Viability was assessed using trypan blue dye exclusion. Cytospin slides (Shandon Inc., Pittsburgh, PA) were prepared from aliquots of BAL fluid of sufficient volume to contain 5 x 104 cells. One slide from each lung segment was stained with Diff-Quickf (American Scientific Products, McGraw Park, IL) for differential counts; at least 500 cells from each slide were counted. A separate slide of cells from the right middle lobe was stained with Mayer's hematoxyline and toluidine blue for enumeration of mast cells (22). Flow cytometry was used to provide a sensitive method for evaluating changes in lightscattering properties of alveolar cells, and to evaluate lymphocyte subsets. After a single wash in cold phosphate buffered saline, BAL cells were pooled and run immediately on a FACScan flow cytometer (Becton Dickinson, Mountain View, CA) equipped with a 15mw argon ion laser at 488 nm. Narrow-angle and wide-angle light scatter and autofluorescence in the red and green spectra were recorded in list mode and subsequently analyzed by an investigator blinded to the nature of the exposure. Lymphocytes in BAL expressing CD3, CD4, and CD8 antigens were quantitated using direct immunofluorescence staining with Leu-series monoclonal antibodies (Becton Dickinson) and subsequent analysis by flow cytometry. Alveolar Macrophage Function Antibody-dependent cellular cytotoxicity (ADCC) by AM was determined using human 0+ R 1R2 (CDe/cDE) red blood cells (RBC) labeled with 50 J.1Ci 51Cr as target cells. RBC were treated with Rho (D) immune globulin (Cutter Biological, Elkhart, IN). Washed alveolar cellsin RPMI 1640(GIBCD,

FRAMPTON, VOTER, MORROW, ROBERTS, CULp, COX, AND UTELL

Grand Island, NY) were incubated in flatbottomed microtiter wellswith target cells at a ratio of 1:1 for 4 h. 51 Cr release was determined by counting 100 J.1L of supernatant from each well in a gamma counter. Percent lysis was calculated using the following formula: OJoADCC

=

A-B C _ B x 100

where A = mean counts per min (cpm) of wells containing antibody-coated RBC plus effector cells, B = mean cpm of wells containing non-antibody-coated RBC plus effector cells, and C = mean total cpm of target cells added to each well. Mean spontaneous release of 51Cr from RBC was less than 4% in this assay. O 2- release was quantitated using superoxide dismutase (SOD) inhibitable cytochrome c reduction in unstimulated AM and after stimulation with 1 J.1g phorbol myristate acetate (PMA) and 5 J.1g dihydrocytochalasin B. One million cells were incubated with 1.25mg cytochrome c with or without 20 J.1g (68 U) SOD for 30min in a 37° C shaking-water bath, and absorbance of the supernatant was determined at 550 nm using a spectrophotometer (Model OMS 100;Varian, Springvale, Australia). Absorbance was converted to nmol O 2- using the extinction coefficient (23). Effects of exposure on the ability of AM to inactivate influenza virus was evaluated in vitro, as reported previously (24). Infection of AM with influenza virus is abortive, with a net decrease in infectious virus in surrounding medium following infection (25). Influenza A/AA/Marton/43 HINI virus was grown in allantoic cavities of lO-day-oldembryonated hens' eggs and stored at -70° C. The virus stock titered at 108 . 0 to 108.2 TCIDso/mL (the dilution of virus infecting 50% of inoculated cell samples) when assayed in MadinDarby canine kidney cells (26). Aliquots (1 to 1.2 x 106 ) of lavaged cells were incubated with virus at a virus-to-cell ratio of 10 in serum-free medium for 1 hat 37° C in a 95% air, 5% CO 2 atmosphere (26). Cells were then washed repeatedly, counted in a hemocytometer, resuspended in medium M199 supplemented with 5% fetal bovine serum, divided into five aliquots, incubated from zero to 4 days at 37° C, and then collected and frozen at - 70° C until assayed for virus. Assay of remaining infectious virus was performed concurrently on samples obtained from each subject after exposure to NaCI and H 2S04 , as previously described (24).

Statistics Exposure to NaCI aerosol served as the control in these studies, so that each subject served as his/her own control. Randomization of the order of the exposures was used to avoid order effects. Data were analyzed using the paired crossover Student's t test (27), which allows for possible period and interaction effects, and by calculation of 95% confidence intervals (CI) using the crossover t statistic.

Results

All subjects completed exposures to both NaCI and H 2S0 4 , followed by BAL 18 h later. Measurement of salivary cotinine levelsbefore each exposure confrrmed the nonsmoking status of the volunteers; cotinine was below the limits of detection (10 ng/mL) prior to all but two exposures. Saliva from one subject contained 25.6 ng/mL of cotinine before the NaCI exposure, and another subject's saliva contained 10.2ng/mL; both values are consistent with passive smoke exposure (28), and findings from these two exposures did not differ from the group as a whole. The total intake of aerosol during exposure was estimated for each subject using the product of VE during both rest and exercise and the mass concentration of the aerosol determined by ion chromatography. The total intake of H 2S04 was slightly higher than NaCI for the group as a whole (mean ± SD: 2,726 ± 606 J,1g H 2S04 versus 2,325 ± 303 J,1g NaCI), because the average mass concentration of both aerosols was slightly different from the initially targeted concentration of 1,000 J,1g/m3 (1,175 ± 234 J,1g/m3 H 2S0 4 versus 940 ± 55 J,1g/m3 NaCl). There were no significant differences in MMAD between the H 2S04 and NaCI aerosols.

Symptoms and Pulmonary Function Three subjects experienced cough and four experienced throat irritation during H 2S04 exposure; one subject experienced cough and three experienced throat irritation during NaCI exposure. Four of 12 subjects detected an odor or taste during H 2S04 exposure, whereas no odor or taste was detected during NaCI exposure. Essentially all subjects were asymptomatic 18 h after exposure. No changes in FVC, FEV h or SGaw were observed immediately following or 18 h after exposure to NaCI or H 2S04 aerosols when compared with pre-exposure baseline measurements, and comparison of pulmonary function data from the H 2S04 and NaCI exposures revealed no differences. For example, FVC decreased from a pre-exposure baseline of 5.12 ± 0.17 to 5.11 ± 0.18 L (mean ± SE) immediately after H 2S04 exposure. FEV 1 was 4.24 ± 0.13 L before" and 4.24 ± 0.14 L after exposure to H 2S04 , The subjects who reported symptoms during exposure to H 2S04 were not different from the group as a whole with regard to pulmonary function responses.

629

ACIDIC AEROSOL EXPOSURE

TABLE 1

1.20J0). For sixsubjects the percentage of CD3+ lymphocytes was more than 5OJo lower after H 2S04 than NaCl exposure, with one subject showing a 21.4OJo difference. One subject had more CD3+ cells after H 2S04 exposure, and the remaining five subjects showed minimal difference between exposures. The lower number of total T lymphocytes in association with H 2S04 exposure was accounted for by a reduction in CD4+ lymphocytes (6.7OJo) compared with NaCl exposure (9.5OJo), again not a statisticallysignificant difference (p == 0.22, CI -7.7 to 2.7OJo).

PROTEIN CONCENTRATION IN BRONCHOALVEOLAR LAVAGE FLUID"

NaCI H2SO4 Cit

Volume Returned (m/)

% Returned

Total Protein (lJglm/)

Albumin (lJglm/)

106 :t: 2 104 :t: 3 -8,4

72 :t: 1 70:t: 2 -6,2

110 :t: 8 116:t: 11 -12,24

30 :t: 4 26:t: 3 -9,1

* Values are mean :t SE. 95% confidence Interval, H2S04 minus NaC!.

t

TABLE 2 CELLS RECOVERED AFTER EXPOSURE TO SODIUM CHLORIDE OR SULFURIC ACID AEROSOLS" Cell Concentration (x 104/m/)

Alveolar Macrophages (%)

Lymphocytes (%)

Polymorphonuclear Leukocytes (%)

Mast Cellst

20.7 :t: 1.9 21.2:t: 2.6 -3.6,4.7

83.8:t: 1.4 84.7:t: 1.5 -2.2,4.1

13.1 :t: 1.3 12.2 :t: 1.4 -3.6, 1.8

2.2 :t: 0.3 2.5 :t: 0.4 -0.8, 1.4

8.0:t: 2.0 7.3 :t: 2.5 -7.6,6.3

• Values are mean :t SE. Metachromatic cells per 5 x 10 4 "bronchoalveolar cells. 95% confidence interval, H2SO4 minus NaC!.

t

*

Findings from Bronchoalveolar Lavage Fiberoptic bronchoscopy after both NaCl and H 2S04 aerosol exposures revealed normal appearance of airways and bronchial mucosa. Lavage fluid return was 72 ± IOJo following inhalation of NaCl and 70 ± 2010 following H 2S04 (p == 0.24). The concentration oftotal protein and albumin in BAL fluid is shown in table 1.There Was a slight increase in total protein and a slight decrease in albumin concentration after exposure to H 2S04 compared with NaCl, but the difference was not statistically significant. The total number and differential counts of cells retrieved by BAL are shown in table 2. No statistically significant changes in cell recovery were observed. Analyses performed separately on cells obtained from the lingula and right middle lobe revealed no significant re-

gional differences in cell differential counts (data not shown). Flow cytometry revealed two predominant populations of alveolar cells showing light-scattering and autofluorescence properties characteristic of lymphocytes and AM. When compared with the NaCl (control) exposures, no significant changes in light-scattering properties were observed following exposure to H 2S04 aerosols, suggesting that influx of smaller, less differentiated macrophages or monocytes did not occur. Figure 2 shows the distribution of lymphocyte subsets in BAL determined by immunofluorescence staining and flow cytometry. There wereno statistically significant differences between H 2S04 and NaCl exposures. T lymphocytes (CD3+) represented 14.9OJn of total cells after NaCI exposure and 11.5OJo after H 2S04 exposure (p == 0.14, 95OJo CI, -7.9 to

Fig. 2. Lymphocytesubsets in BAL. Immunofluorescence staining with Leuseries monoclonal antibodies to CD3 (left panel), CD4 (C8nler panef), and CD8 (rightpane/) was performed on cells obtained by BAL. Flow cytometrywasused to determinethe number of fluorescence positivecells in the lymphocytegate,and the result expressed as percent of total cells.

Alveolar Macrophage Function The ability of AM to lyse antibodycoated RBC was assessed in vitro 18 h after exposure to NaCl and H 2S04 aerosols. The cytotoxicity assay was successfully completed on both exposure days for nine subjects, and these data are shown in table 3. A small increase in lysis of RBC was seen following exposure to H 2S04 when compared with NaCI exposure. AM from seven subjects increased ADCC following acid exposure; cells from the remaining two subjects decreased ADCC following acid exposure. The difference was not statistically significant (p == 0.16). Sufficient cells were available after both NaCI and H 2S04 exposure to assess release of O 2 - by AM in response to PMA in 10 subjects, and in addition, from fresh unstimulated cells in eight subjects. The data are shown in table 3. PMA stimulation resulted in an approximately 12-fold increase in the release of O 2 - , but no significant difference in baseline or stimulated release of O2 was found between H 2S04 and NaCI exposures. .The ability of AM to become associated with and inactivate influenza virus was evaluated in vitro 18 h after exposure. Lavaged cells were exposed to influenza virus for 1 hat 37° C, washed, incubated for zero to 4 days, and then assayed for infectious virus. The experiment was successfully completed for both exposures in nine subjects. Figure 3 shows the rate of decline in titers of cell-associated infectious virus. There was no statistically significant difference between H 2S04 and NaCI exposure at any time point (e.g., p = 0.5 for Day 2 of incubation). Discussion

NACL

H2S04

NACL

H2S04

In the atmosphere, acidic aerosols exist as submicronic particles with the potential for significant deposition in the alveolar compartment of the lung. Ex-

630

FRAMPTON, VOTER, MORROW, ROBERTS, CULp, COX, AND UTELL

TABLE 3 CYTOTOXICITY AND SUPEROXIDE ANION RELEASE BY ALVEOLAR MACROPHAGES·

ADCC

Superoxide Anion (Unstimulated)

Superoxide Anion (PMA}:j:

(% lysis) (n = 9)

(nmol/108 cellsl30 min) (n = 8)

(nmo1/10 8 cells130 min) (n • 10)

19.1 ± 1.6 23.6 ± 1.9 -2.1,10.8

1.0 ± 0.4 1.1 ± 0.5 -0.4,2.0

13.7 ± 2.9 12.9 ± 2.7 -5.5,3.3

• Values are mean ± SE. 95% confidence interval, H,SO. minus NaCI. Stimulated with 1 119 phorbol myristate acetate.

t

*

posure to acidic aerosols could alter the alveolar microenvironment through direct epithelial injury, through alteration of alveolar inflammatory cell populations, or by altering specific functions of immunocompetent cells in the distal airways. These effects could be mediated through local changes in pH. Holma (29) has shown that mucus viscosity is increased when pH decreases below 7.4, providing a possible mechanism for previously observed changes in mucociliary clearance following exposure to acidic aerosols (30). In addition, acid hydrolases generated by inflammatory cells such as AM or PMN may be activated at acidic pH (31). Acidic particles deposited at the alveolar level could cause altered epithelial permeability, activation of resident AM with release of toxic oxygen species, or recruitment of lymphocytes or PMN from the vascular compartment. In the present studies, exposure to H 2S04 at 1,000 ug/m" for 2 h did not result in alveolar inflammation as assessed

by BAL 18 h after exposure. Total and differential cell counts in BAL fluid were comparable following exposure to H 2S04 and NaCI aerosols, with no evidence for influx of PMN. The CI indicate there is a 95070 assurance that exposure to H 2S04 aerosol, as in these studies, does not increase PMN by more than 1.4070 over control (table 2). Therefore, although these data do not exclude the possibility of a very mild alveolar inflammatory response, they indicate that any increase in PMN is an order of magnitude less than that seen following exposure to 0.4 ppm ozone (10) and less than the minimal increase observed following exposure to 2.0 ppm N0 2 (32). Although this study did not specifically assess airway inflammatory cells, BAL is known to sample the distal airways as well as the alveolar space (33), and diseases involving the airways primarily, such as asthma and smoking, cause changes detectable with BAL (34, 35). Thus, our data suggest that the in-

4.0

rn li:;) CJ

3.0

z

sa::

Fig. 3. Influenza virus inactivation. Cells obtained by SAL after exposure to NaCI or H2S04 were incubated with Influenza virus, divided into five aliquots, Incubated from zero to 4 days at '$TO C, and subsequently assayed for infectious virus. Results are expressed as means ± SE for nine subjects.

oII.

III :;)

ac

2.0

~

l1.

~ o

(;

1.0

~

0.0 L - - - y - - - r - - - - - r - - - - - - . - - - - r - 00 01 02 03 04

DAY OF INCUBATION

crease in airways responsiveness 24 h after exposure to H 2S04 aerosol in healthy volunteers described in a previous study (8) was not caused by inflammation. However, it is possible that small increases in inflammatory cells in the distal airways were masked by dilution from the larger alveolar component of BAL. In future studies, separate analysis of the first lavage aliquot (36) or isolated airway lavage (37) may help to determine whether acidic aerosols induce airway inflammation. We observed a small decrease in the percentage of T lymphocytes retrieved by BAL in association with H 2S04 exposure, which was not statistically significant. No change in the light-scattering properties of alveolar cells was observed 18 h after exposure, suggesting that shifts in size or internal complexity of alveolar cells did not occur. In addition, no evidence was found for translocation of plasma proteins to the alveolar space in association with exposure to the acidic aerosol, because total protein and albumin concentration in BAL fluid was similar following exposure to both NaCI and H 2S04 aerosols. This suggests that epithelial permeability was not significantly altered by the acidic aerosol exposure. Most respiratory infections are viral, and AM playa role in limiting infections at the alveolar level during illnesses such as influenza (38). Mechanisms of AM antiviral function may include release of O 2during a respiratory burst or lysis of virus-infected cells in the presence of antibody. Pollutant exposure could alter the response to viral infection by impairing these functions. Frampton and colleagues (24) examined the effects ofN02 exposure on AM antiviral function in normal humans. Exposure to 0.6 ppm N02 for 3 h was associated with a reduced inactivation of influenza virus in vitro by lavaged AM from four of nine subjects. For the present studies, exposure to H 2S04 was associated with a small increase in the rate of influenza virus inactivation when compared with NaCI exposure. This effect was in the opposite direction from that seen following N02 exposure and was not statistically significant. A small increase in antibody-mediated cytotoxicity by AM in association with H 2S04 exposure was also observed.. Release of O 2 - by AM, both unstimulated and following incubation with PMA, was similar after exposure to NaCI and H 2S04 , No evidence was found linking H 2S04 exposure to impairment of AM antiviral function in these studies.

631

ACIDIC AEROSOL EXPOSURE

Could the absence of an alveolar inflammatory response in these studies be explained by lack of acid deposition in the alveolar space? This is unlikely because our study was designed to assure alveolar deposition of H 2S04 • The respiratory deposition of aerosols with GSD ~ 2.0 can be adequately described by their MMAD as though they were uniform particles (39). The growth of a hygroscopic 0.9 urn droplet in alveolar air is expected to increase in diameter by a maximum of a factor of 1'\.14 (40). To reach this equilibrium size would require approximately one second residence time. Conservatively, therefore, we can estimate the alveolar deposition probability of the 0.91lm H 2S04 aerosol as being determined by an ultimate aerodynamic droplet size range of between 1\.11.8 and 1'\.13.6 um diameter. This provides a size range optimal for alveolar deposition, with a probability of deposition between 1'\.125 and 1\.140070 (41). Thus, the acidic aerosol generated for these studies was of optimal size to penetrate to and deposit in the alveolar region of the lung. Furthermore, the study protocol was designed to minimize neutralization of H 2S04 by oral ammonia (19). Although these findings suggest that exposure to H 2S04 aerosols at 1,000 ug/m" does not cause a significant inflammatory response in the alveolar space, an absence of effects on alveolar host defense has not been established. Early, transient alterations in AM function could have occurred that resolved by 18h after exposure. Even transient impairments in host defense, when applied across large populations of exposed individuals who are repeatedly exposed to respiratory viruses, could havesignificant public health implications. In addition, the limited sample size necessitated by the expense and complexity of these human studies restricts their statistical power; the possibility that exposure to acidic aerosols causes small changes in lymphocyte subpopulations or AM function has not been excluded with certainty. For example, the 95070 CI tell us that the decrease in CD3+ T lymphocytes could be as large as 7.9070, but an increase of 1.2070 is also not excluded. We have calculated that in order to provide 80070 assurance that the observed mean decrease of 3.4070 CD3+ T cells is not due to chance, assuming an a of 0.05, a total of 20 subjects would be required. Of course, excluding or confirming changes of smaller magnitude would require larger numbers of subjects. Similar conclusions can be drawn with

regard to AM ADCC (table 3). The data do not exclude the possibility that ADCC increases by as much as 10.8070 lysis (a 56% increase) in association with the acidic aerosol. Finally, an absence of demonstrable effects on host defense following single, brief exposures to acidic aerosols does not exclude the possibility that more prolonged or repeated exposures - a situation more reflective of environmental conditions - may impair host defense. In conclusion, these studies suggest that single 2-h exposures to H 2S04 aerosols at 1,000 ug/m" do not result in alveolar inflammation, influx of plasma proteins into the alveolar space, or alteration in selected antiviral functions of AM, when assessed 18 h after exposure. Additional studies of larger numbers of subjects are needed to exclude significant effects of H 2S0 4 exposure on recovery of CD3+ T lymphocytes or ADCC by AM. Future studies should incorporate methods to specifically evaluate airway inflammation. If exposure to acidic aerosols alters mucociliary transport in the absence of an airway inflammatory response, as suggested by the data currently available, there is a need to determine whether acidic aerosol exposure alters the composition of airway mucins. In addition, future studies should examine alveolar cellular responses at other time points following single exposures, and after prolonged or repeated exposures. References 1. Dockery OW, Speizer FE. Epidemiological evidence for aggravation and promotion of COPO by acid air pollution. In: Hensley MJ, Saunders NA, eds. Lung biology in health and disease. Vol. 43: Clinical epidemiology of chronic obstructive pulmonary disease.New York:Marcel Dekker, 1989; 201-25. 2. Spengler JO, Keeler OJ, Koutrakis P, Ryan PB, RaizenneM, Franklin CA. Exposures to acidic aerosols. Environ Health Perspect 1989; 79:43-51. 3. Bates OV, Sizto R. Air pollution and hospital admissions in southern Ontario: the acid summer haze effect. Environ Res 1987; 43:317-31. 4. Speizer FE. Studies of acid aerosols in six cities and in a new multi-city investigation: design issues. Environ Health Perspect 1989; 79:61-7. 5. Last JA, Hyde OM, Guth OJ, Warren OL. Synergistic interaction of ozone and respirable aerosols on rat lungs. I. Importance of aerosol acidity. Toxicology 1986; 39:247-57. 6. Frampton MW, Morrow PE, Cox C, et al. Does pre-exposureto acidic aerosols alter airway responses to ozone in humans (abstract)? Am Rev Respir Dis 1992; 145:A428. 7. Utell MJ, Morrow PE, Hyde RW. Airway reactivity to sulfate and sulfuric acid aerosols in normal and asthmatic subjects. J Air Pollut Control Assoc 1984; 34:931-5. 8. Utell MJ, Morrow PE, Hyde RW. Latent development of airway hyperreactivity in human subjects after sulfuric acid aerosol exposure. J Aero-

sol Sci 1983; 14:202-5. 9. Koenig JQ, Covert OS, Pierson WE. Effects of inhalation of acidic compounds on pulmonary function in allergic adolescent subjects. Environ Health Perspect 1989; 79:173-8. 10. Koren HS, Devlin RB, Graham DE, et al. Ozone-induced inflammation in the lower airways of human subjects. Am Rev Respir Dis 1989; 139:407-15. 11. Schlesinger RB. The interaction of inhaled toxicants with respiratory tract clearance mechanisms. Crit Rev Toxicol 1990; 20:257-86. 12. Naumann BD, Schlesinger RB. Assessment of early alveolar particle clearance and macrophage function following an acute inhalation of sulfuric acid mist. Exp Lung Res 1986; 11:13-33. 13. Schwartz LW, Zee YC, Tarkington BK, Moore PF, Osebold JW. Pulmonary responses to sulfuric acid aerosols. In: Lee SO, ed, Assessing toxic effects of environmental pollutants. Ann Arbor: Ann Arbor Science, 1979; 173-87. 14. Schlesinger RB. Factors affecting the response of lung clearance systems to acid aerosols: role of exposure concentration, exposure time and relative acidity.Environ Health Perspect 1989; 79:121-6. 15. Gearhart JM, Schlesinger RB. Sulfuric acidinduced changes in the physiology and structure of the tracheobronchial airways. Environ Health Perspect 1989; 79:127-37. 16. Thurston GO, Ito K, Lippmann M, Hayes C. Reexamination of London, England, mortality in relation to exposure to acidic aerosols during 1963-1972winters. Environ Health Perspect 1989; 79:73-82. 17. Frampton MW, Morrow PE, Oibb FR, Speers OM, Utell MJ. Effects of nitrogen dioxide exposure on pulmonary function and airway reactivityin normal humans. Am Rev Respir Dis 1991; 143:522-7. 18. Utell MJ, Morrow PE, Hyde RW,SchreckRM. Exposure chamber for studies of pollutant gases and aerosols in human subjects: design considerations. J Aerosol Sci 1984; 15:219-21. 19. Utell MJ, Mariglio JA, Morrow PE, Oibb FR, Speers OM. Effects of inhaled acid aerosols on respiratory function: the role of endogenous ammonia. J Aerosol Med 1989; 2:141-7. 20. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Bioi Chern 1951; 193:265-75. 21. Frampton MW, Finkelstein IN, Roberts NJ Jr, Smeglin AM, Morrow PE, Utell MJ. Effects of nitrogen dioxide exposure on bronchoalveolar lavage ptoteins in humans. Am J Respir Cell Mol Bioi 1989; 1:499-505. 22. Sandstrom T, Andersson MC, KolmodinHedman B, Stjernberg N, Angstrom T. Bronchoalveolar mastocytosis and lymphocytosis after nitrogen dioxide exposure in man: a time-kinetic study. Eur Respir J 1990; 3:138-43. 23. Massey V. The microestimation of succinate and the extinction coefficient of cytochrome c. Biochim Biophys Acta 1959; 34:255-6. 24. Frampton MW, Smeglin AM, Roberts NJ Jr, Finkelstein IN, Morrow PE, Utell MJ. Nitrogen dioxide exposure in vivo and human alveolar macrophage inactivation of influenza virus in vitro. Environ Res 1989; 48:179-92. 25. Rodgers BC, Mims CA. Influenza virus replication in human alveolar macrophages. J Med Virol 1982; 9:177-84. 26. Roberts NJ Jr, Douglas RO Jr, Simons RM, Diamond ME. Virus-induced interferon production by human macrophages. J Immunol 1979; 123:365-9. 27. Brown BW Jr. The crossover experiment for clinical trials. Biometrics 1980; 36:69-79. 28. Coultas DB, Howard CA, Peake GT, Skipper W, Samet JM. Salivary cotinine levelsand involun-

632 tary tobacco smoke exposure in children and adults in New Mexico. Am RevRespir Dis 1987; 136:305-9. 29. Holma B. Effects of inhaled acids on airway mucus and its consequences for health. Environ Health Perspect 1989; 79:109-13. 30. Lioy PJ, Lippmann M. Measurement of exposure to acid sulfur aerosols. In: Lee SO, Schneider T, Grant LD, Verkerk PJ, eds. Aerosols. Chelsea, MI: Lewis Publishers, Inc., 1986; 743-52. 31. Schlesinger RB. Effects of inhaled acids on respiratory tract defense mechanisms. Environ Health Perspect 1985; 63:25-38. 32. Frampton MW, Morrow PE, Roberts NJ Jr, Tsai Y-C, Utell MJ. Characterization of the inflammatory response to N01 exposure using bronchoalveolar lavage in humans (abstract). Am Rev Respir Dis 1991; 143:A89.

FRAMPTON, VOTER, MORROW, ROBERTS, CULp, COX, AND UTELL

33. Kelly CA, Kotre CJ, Ward C, Hendrick OJ, Walters HE. Anatomical distribution ofbronchoalveolar lavage fluid as assessed by digital subtraction radiography. Thorax 1987; 42:624-8. 34. Hunninghake OW, Crystal RG. Cigarette smoking and lung destruction. Accumulation of neutrophils in the lungs of cigarette smokers. Am Rev Respir Dis 1983; 128:833-8. 35. Kirby JG, Hargreave FE, Gleich GJ, O'Byrne PM. Bronchoalveolar cell profiles of asthmatic and nonasthmatic subjects. Am Rev Respir Dis 1987; 136:379-83. 36. Rennard SI, Ghafouri MO, Thompson AB, et al. Fractional processing of sequential bronchoalveolar lavage to separate bronchial and alveolar samples. Am Rev Respir Dis 1990; 141:208-17. 37. Eschenbacher WL, Gravelyn TR. A technique

for isolated airway segment lavage. Chest 1987; 92:105-9. 38. Ada GL, Jones PD. The immune response to influenza infection. Curr Top Microbiol Immunol 1986; 128:1-54. 39. Bates D, Fish BR, Hatch TF, Mercer IT, Morrow PE. Deposition and retention models for internal dosimetry of the human respiratory tract. Health Phys 1966; 12:173-208. 40. Ferron GA. The size of soluble aerosol particles as a function of the humidity of the air. J Aerosol Sci 1977; 8:251-67. 41. Heyder J, Gebhart J, Heigwer G, Roth C, Stahlhofen W. Experimental studies of total deposition of aerosol particles in the human respiratory tract. J Aerosol Sci 1973; 4:191-208.