Pulmonary Responses to Oil Fly Ash Particles in the Rat Differ by ...

TOHCOLOGICAL SCIENCES 43, 204-212 (1998)ARTICLE NO. TX982460

Pulmonary Responses to Oil Fly Ash Particles in the Rat Differby Virtue of Their Specific Soluble Metals1

Urmila P. Kodavanti, Russ Hauser,* David C. Christiani,* Zhi H. Meng.fJohn McGee, Allen Ledbetter, Judy Richards, and Daniel L. Costa

Pulmonary Toxicology Branch, Experimental Toxicology Division, National Health and Environmental Effects Research Laboratory, U.S. EnvironmentalProtection Agency, Research Triangle Park, North Carolina 27711; \Duke University Medical Center, Durham, North Carolina; and *Department

of Environmental Health, Occupational Health Program, Harvard School of Public Health, Boston, Massachusetts 02115

Received November 25, 1997; accepted March 26, 1998

Pulmonary Responses to Oil Fly Ash Particles in the Rat Differby Virtue of Their Specific Soluble Metals. Kodavanti, U. P.,Hauser, R., Christiani, D. C, Meng, Z. H., McGee, J., Ledbetter,A., Richards, J., and Costa, D. L. (1998). ToxicoL Sci. 43,204-212.

Occupational exposure to residual oil fly ash (ROFA) particu-late has been associated with adverse respiratory health effects inhumans. We hypothesized that ROFA collected at different siteswithin an oil burning power plant, by virtue of its differing metaland sulfate composition, will induce differential lung injury. TenROFA samples collected at various sites within a power plant wereanalyzed for water- and 1.0 M HCl-leachable arsenic (As), beryl-lium (Be), cadmium (Cd), cobalt (Co), chromium (Cr), copper(Cu), iron (Fe), manganese (Mn), nickel (Ni), lead (Pb), vanadium(V), zinc (Zn), and sulfur by inductively coupled plasma-atomicemission spectroscopy. All ROFA samples contained variableamounts of teachable (water-extractable) and 1.0 M HCl-extract-able Fe, V, and/or Ni. All other metals, except Zn (ROFA No. 1contained 3.43 and No. 3, 6.35 fig/mg Zn), were present in negli-gible quantities (<1.0 fig/mg) in the water extract In vivo pulmo-nary injury from exposure to whole saline suspensions of theseROFA was evaluated. Male, SD rats (60 days old) were intratra-cheally instilled with either saline or saline suspension of wholeROFA (<3.0 mass median aerodynamic diameter) at three con-centrations (0.833, 3.33, or 8.33 mg/kg). After 24 h, lungs werelavaged and bronchoalveolar lavage fluid (BALF) was analyzedfor cellular influx and protein content as well as lactate dehydro-genase (LDH) and N-acetyl glucosaminidase (NAG) activity andtotal hemoglobin as indicators of lung injury. ROFA-induced in-creases in BALF protein and LDH, but not neutrophilic inflam-mation, were associated with its water-leachable total metal, Ni,Fe, and sulfate content. However, the neutrophilic response fol-lowing ROFA exposure was positively correlated with its water-leachable V content Modest lung injury was observed with theROFA samples which contained the smallest amounts of water-

1 The research described in this article has been reviewed by the NationalHealth and Environmental Effects Research Laboratory, U.S. EnvironmentalProtection Agency, and approved for publication. Approval does not signifythat the contents necessarily reflect the views and the policies of the Agencynor does mention of trade names or commercial products constitute endorse-ment or recommendation for use.

teachable metals. The ability of ROFA to induce oxidative burst inalveolar macrophage (AM) was determined in vitro using a chemi-luminescence (CL) assay. AM CL signals in vitro were greatestwith ROFA containing primarily soluble V and were less withROFA containing Ni plus V. In summary, ROFA-induced in vivoacute pulmonary inflammation appears to be associated with itswater-leachable V content; however, protein leakage appears to beassociated with its water-leachable Ni content. ROFA-induced invitro activation of AM was highest with ROFA containing leach-able V but not with Ni plus V, suggesting that the potency and themechanism of pulmonary injury will differ between emissionscontaining V and Ni . c 1998 Sod*? of Toifcoiogy.

Adverse respiratory health effects of occupational exposuresto metal-containing particulates (PM) have been well docu-mented (Levy et al, 1984; Rendall et al, 1994; Hauser et al,1995a,b). PM emissions from oil-burning power plants andother industries, which contribute to air pollution, contain largeamounts of metals such as vanadium (V), iron (Fe), nickel (Ni),and zinc (Zn) (Levy et al, 1984; Gulyas and Gercken, 1988;Rendall et al, 1994; Hauser et al, 1995a,b). Vanadium hasbeen frequently associated with boilermaker's bronchitis (Levyet al, 1984) and Ni has been classified as a human carcinogen(IARC, 1990) in addition to its sensitizing effects (Menne etal, 1989). Environmental PM collected in industrial areas hasbeen noted to contain these toxic metals (Spengler and Thur-ston, 1983; Lamm et al., 1994). Respiratory diseases in humanshave been associated with PM air pollutants in industrial areasemitting metallic PM (Lamm et al, 1994). However, the de-finitive role of metal versus other causative constituents has notbeen confirmed in epidemiologic studies associating PM airpollution with mortality of susceptible populations (Dockery etal, 1993; Schwartz, 1994). Overall, although PM-associatedmetals are considered a plausible causative constituent (Hatchet al, 1985; Ghio et al, 1996; Costa and Dreher, 1997),definitive experimental evidence is lacking to confirm, or re-fute, its role in pulmonary injury with air PM at environmen-tally relevant exposures.

1096-6080/98 $25.00Copyright O 1998 by the Society of Toxicology.All rights of reproduction in any form reserved.

204

Downloaded from https://academic.oup.com/toxsci/article-abstract/43/2/204/1657384by gueston 06 February 2018

FLY ASH METALS AND LUNG INJURY 205

Pulmonary toxicity of anthropogenic combustion PM suchas residual oil fly ash (ROFA) in experimental animals hasbeen extensively studied (Hatch et al, 1985; Geertz et al,1994; Pritchard et al, 1996; Ghio et al, 1996; Dreher et al,1997; Kodavanti et al., 1997a,b). A prototypic oil combustionPM containing highly water soluble Fe, V, and Ni has beenshown to cause acute pulmonary injury (Dreher et al, 1997)and fibrosis when instilled intratracheally into rats at highdoses (Kodavanti et al, 1997a). Subsequently, with this par-ticular ROFA, the role of V, Ni, and Fe in causing pulmonaryinjury has also been investigated in the rat in vivo (Costa andDreher, 1997; Dreher et al, 1997; Kodavanti et al, 1997b).Our studies demonstrated that in vivo Ni was the most toxic ofthe three primary water-leachable metals present in this ROFAin terms of severity of the lung pathology and the persistencyof cytokine gene induction (Kodavanti et al, 1997b). ROFA,when applied in vitro to airway epithelial cells, has been shownto induce significant injury and inflammatory cytokine produc-tion. Vanadium but not Ni has been shown to account for muchof these in vitro effects (Becker et al, 1996; Dye et al,1997a,b; Samet et al, 1997). It has also been reported that Nicauses inhibition of alveolar macrophage (AM) phagocytosis atnoncytotoxic concentrations; however, V causes extensive celllysis and death without reducing phagocytosis response (Gra-ham et al, 1975). Based on this earlier work, we hypothesizedthat ROFA containing water-leachable Ni and V may causepulmonary injury by different mechanisms.

PM-associated metals or metal salts have been shown todiffer in their oxidative reactions in biological media in vitroand these have been implicated in tissue injury in vivo (Hatchetal, 1980, 1985; Aust et al, 1985; Fisher etal, 1986;Gulyasand Gercken, 1988; Berg et al, 1993; Geertz et al, 1994;Schluter et al, 1995; Stohs and Bagchi, 1995; Pritchard et al,1996). These studies show that V and Ni differ significantly intheir oxidation potential in vitro; however, it is not known howit may relate to differences in acute pulmonary injury in vivo.The purpose of this study is to demonstrate how ROFA withdiffering V and Ni may differ in terms of their ability to causein vivo acute pulmonary injury and oxidative reactions withAM in vitro. The approach included three phases: (1) collec-tion, size reduction, and metal characterization of 10 ROFAsamples from various locations within an oil-burning powerplant (with the assumption that dependent on collection sitetheir metal composition will differ); (2) assessment of in vivoacute pulmonary injury/inflammation following R O F A expo-sure in the rat; and (3) determination of the ROFA-induced invitro activation of AM. Our studies indicate that ROFA con-taining water-leachable Ni and V differentially induce pulmo-nary injury. Exposure of rats to ROFA containing water-leachable V was positively associated with pulmonaryinflammatory response and, in vitro, marked activation of AM,whereas exposure to ROFA containing water-leachable Ni wasassociated with significant in vivo pulmonary microvascularleakage but minimal activation of AM in vitro.

TOP

LOWER

ASH PIT BOTTOM

FIG. 1. A simplified sketch of the power plant showing the 10 ROFAcollection sites. The power plant is located in the Boston area.

METHODS AND MATERIALS

Animals

SLxty-day-old male Sprague-Dawley (SD) rats were obtained from CharlesRiver Laboratories (Raleigh, NC) and were housed in an AAALAC-approvedanimal facility (72 ± 2°F, 50 ± 5% relative humidity, 12-h light/dark cycle)during quarantine and after intratracheal instillation. All animals received astandard Punna rat chow (Richmond, IN) and water ad libitum before andduring experiments.

Phase I: Collection, Size Reduction, and Characterization of ROFA

ROFA samples were collected during the shutdown and overhaul of a largeoil-buming boiler at an electricity-generating power plant in Boston, Massa-chusetts. Ten ROFA samples were collected from different locations within thepower plant (Fig. 1). This included 4 samples from different areas within theprecipitator unit, 2 from different baskets of the air heater (air-heat exchangeunit), 2 from different areas in the bottom of the boiler (ash pit), and 2 fromopposite sites of boiler (burner location). Multiple samples were taken fromeach location to account for the variability in ROFA metal content that may bepresent at each location. For instance, the precipitator unit contains ash that iscollected over the course of several weeks to months. Therefore, precipitatorash at the top of the precipitator represents combustion of oil stock that maydiffer substantially from ash at the bottom of the precipitator from an earlier oilstock. The samples were collected and stored in sealed polyethylene tubes ina cool dry area for several months. Each PM sample was designated a number1 through 10 (appropriate number designation with reference to location isdepicted in Fig. 1) and this number designation is followed through entirepaper. Each of these samples was placed into a stainless-steel cup along witha stainless-steel ball and was clamped into the ball mill shaker. The cup wasshaken vigorously for 30-60 min to reduce particle size. The ground PMsamples were then placed in a 15-ml sterile centrifuge tube with a special capcontaining 100-fim mesh nylon screen. The tube was inverted and shaken byhand. The particles which passed through the mesh were analyzed for sizeusing a TSI Model 331OA aerodynamic particle sizer (ISI, Inc., St. Paul, MN).The particles were made airborne with a TSI Model 3433 small-scale powderdispenser. The mass median aerodynamic diameter (MMAD) of all ground PM


206 KODAVANTl ET AL.

samples was within 1.99-2.59 (irn with geometric standard deviation of1.98-2.26.

After sizing, known amounts of each ROFA samples were suspended ineither double-distilled water or 1.0 M HCl and mixed using a mixer for 30 nun.The tubes were then centrifuged at 14,000g for 30 min to remove the pellet.The supematants of water and HCl extracts were filtered using Teflon syringefilters, 0.2-/xm pore size (Poretics Corp., Livermore, CA). Filtrates wereanalyzed for arsenic (As), beryllium (Be), cadmium (Cd), chromium (Cr),copper (Cu), Fe, lead (Pb), manganese (Mn), Ni, V, Zn, and sulfate (SO,,) byinductively coupled plasma—atomic emission spectrometer (ICP-AES; ModelP40; Perkin-Elmer, Norwalk, CT), following closely the protocol for EPAMethod 200.7, "Determination of Metals and Trace Elements in Water andWastes by ICP-AES" (U.S. Environmental Protection Agency, 1992). Com-mercially available, NIST-traceable single-element metal and sulfate solutions(Aldrich Chemical Co., Milwaukee, WI) were used to calibrate the instrumentand generate standard curves; standard curves were checked with a multiele-ment standard from an independent source (QC-21; SPEX Industries, Inc.,Edison, NJ). The metals extracted in water were termed water-leachable orsoluble and the ones extracted in 1.0 M HCl were termed acid- or 1 M HC1-extractable. The quantity of sulfur extracted in water or HCl remained similarfor each PM sample.

Phase II: ROFA-Induced In Vivo Pulmonary Injury

ROFA Instillation

Each weighed ROFA was suspended in bacteriostatic saline at three con-centrations (0.83, 3.33, or 8.3 mg/kg/ml) and mixed for 20 min. Either thesewhole saline suspensions of ROFA or saline alone was intratracheally instilled(in) in the rat under halothane anesthesia (Costa et al., 1986). The selection ofROFA dose (8.33 mg/kg) was based on our previous studies (Kodavanti et al.,1997a,b). Two smaller doses were selected to cover medium effect and low orno effect levels.

Bronchoalveolar Lavage (BAL) and BAL Fluid (BALF) Analysis forDetermining Lung Injury. At 24 h following iti of saline or saline suspen-sions of ROFA, rats were anesthetized with sodium pentobarbital (Nembutal;Abbott Lab., Chicago, IL; 50 mg/kg body wt, ip) and exsanguinated via theabdominal aorta. Tracheas were cannulated and lung lavage was performedwith phosphate-buffered saline (pH 7.4) using a volume equal to 28 ml/kgbody wt (approximately 75% total lung capacity) as modified from Hatch et al.(1986). Three in-and-out washes were performed using the same fluid. Onealiquot of whole lavage fluid was used for determining total cells using aCoulter Counter (Coulter, Inc., Miami, FL), and a second aliquot was centri-fuged using Shandon 3 Cytospin (Shandon) for preparing cell differentialslides. The slides were dried at room temperature and stained with LeukoStat(Fisher Scientific Co., Pittsburgh, PA). The numbers of macrophages, neutro-phils, and eosinophils were counted using light microscopy. The remainingBALF was centrifuged at lOOOg to remove cells and die supernatant fluidswere analyzed for protein, albumin, lactate dehydrogenase (LDH), and n-acetylglucosaminidase (NAG). Assays for protein, LDH, NAG, and hemoglo-bin were modified and adapted for the use on a Hoffmann-La Roche CobasFara II clinical analyzer (Roche Diagnostics, Branchburg, NJ). Total proteincontent was determined using a Coomassie Plus protein assay kit (Pierce,Rockford, IL). LDH activity was determined using a Kit 228 and standardsfrom Sigma Chemical Co. (SL Louis, MO). The kits and controls for NAGanalysis were from Boehringer Mannheim Corporation Products (Indianapolis,IN). Concentrations of hemoglobin in the BALF pellet which contained redblood cells (in case of pulmonary leakage) was determined by a commerciallyavailable kit for total hemoglobin (Kit Protocol 525) from Sigma Chemical Co.The sample values were normalized to the original BALF volume that wasused in obtaining pellets.

Phase III: In Vitro AM Activation

Preparation of AM for In Vitro Assays

Healthy male SD rats (60-70 days old) were anesthetized and lavaged asshown earlier. Whole BALF fluid was centrifuged at lOOOg for 15 min and thesupernatant fluids, as well as cell pellets, were isolated. Cells were suspendedin Gey's balanced salt solution (GBSS) (Gibco BRL, Life Technologies, Inc.,Grand Island, NY) at a concentration of 0.6-1.0 X 10* cells/ml, kept on ice,and used within 2 h of isolation. Cells were counted with a hemocytometerunder a light microscope and viability was determined using trypan blue dye.BALF cells isolated from healthy rats consisted of >90% AM and viabilitywas >80% at the beginning and the end of the assay.

Preparation of ROFA Fractions for In Vitro Assays

For in vitro assays, ROFA particles were suspended in GBSS buffer at aconcentration of 1 mg/ml and mixed for 10 min at room temperature. In someexperiments this suspension was centrifuged at 4°C, 14,000g for 15 min Theclear supernatant, termed the leachate, was then separated and stored until use.After collection of supematants, the pellet fractions were washed three timeswith fresh GBSS buffer and then suspended in the original volume of GBSS.This fraction was termed the ROFA pellet.

In Vitro Chemiluminescence Assay

Macrophages in vitro, upon activation, undergo oxidative changes resultingin chemiluminescence (CL), which is measurable photometrically (Hatch etal., 1978, 1980; Brennan and Kirchners, 1985; Ghio et al., 1997). The mea-surement of luminol-enhanced CL by AM in vitro has been used as an indexof oxidation or oxidative burst (Hatch et al., 1978). CL from the oxidative burstof activating agents, such as infectious material or zymosan, has been exten-sively characterized (Hatch et al., 1978). We utilized luminol-enhanced CL todetermine ROFA-induced activation of AM. The analysis protocol has beenpreviously described (Ghio et al., 1997). Briefly, AM (0.3-0.5 X 106 cells/ml)suspended in GBSS were incubated with albumin-conjugated luminol in theabsence or presence of ROFA (whole suspension, leachate, or washed ROFAfraction). The light intensity resulting from oxidation of luminol was moni-tored every 15 or 30 min for up to 150 min using an FTP photometer (Model3000; SAI Technology Co., San Diego, CA) and recorded on a strip chartrecorder. The data are expressed in arbitrary units of V/0.5 X 10* AM.

Statistics

General linear models procedures (SAS 516; SAS Institute, Cary, NC) andDunnett's t tests were used to compare dose-related effects of each ROFA onvarious indices of lung injury. Corrected pairwise comparisons were used assubtests to evaluate differences between exposure groups. The Type I error ratewas set at 0.05 for significance. Control values were pooled because nostatistical differences were noted between the saline-instilled animals at eachtime. Multiple regression and correlation analysis were used to investigate therelationship between BALF parameters and ROFA constituents (i.e., metalsand sulfate concentrations). Data are represented as the means ± SE of 18-20controls and 6 exposed rats. In vitro data were analyzed using one-way analysisof variance followed by Dunnett's t test and samples were run in duplicate atleast two times.

RESULTS

Characteristics and Composition of ROFA Samples

All ROFA samples were ground to similar sizes within therespirable range before analyzing for metals or using them foranimal studies. Water- and 1.0 M HCl-leachable metals wereanalyzed in all 10 ROFA samples (Table 1). Samples were



TABLE 1Water- and 1 M HCl-Leachable Metals of ROFA Collected at Various Locations within a Residual Oil-Burning Power Plant"

ROFAsample*

123456789

10

Collection site

Lower ptecipitatorTop precipilalotLower precipitate*Top precipitatorAsh pit bottomAsh pit bottomAir heater basketAir heater basketFire box/bumersFire box/bumers

pH insaline'

2.8910.493.504 276346435.813645.845.78

Particlesize, /un(MMAD)

2.272.232.592.362.452.261.992.252 672.52

Geometricstandarddeviation

(«,)

2.122.062.092.082.042.082.262.141982.24

Water-leachable

total metal(/ig/mg)

57.03.80

47.23.0

25.935.822.616239 836.5

HCI-leachable

total metal(Mg/mg)

82.4133.778.7

136.0117.8149.0901

133.81343127 0

Water-leachable

Fe(Hg/mg)

21.20.04.72.00.00.00.00.20.00 0

HC1-leachable

Fe(fig/mg)

39.0114.0

17.7133.3

12.414.710913.773

10.9

Water-leachable

V(/ig/mg)

18.03.3

20.60.0

25.535.021.311.838.635.6

HCI-leachable

V(Mg/mg)

25411.736.9

0,391.0

117.863.3

105.4113.4100.9

Water-leachable

Ni(Mg/mg)

13.9OS

14.20.10.30.51 1380.50.7

HCI-leachable

NiOg/mg)

13.46.0

16.00.2

10.210.614.01227.9

109

Sulfur(Mg/mg)

320.4120.4267.349.213.413.726 975 733 399.0

° Metal leached off in water is termed water-leachable and metal leached off in 1 M HCI is termed HCI-leachable. Total metal is the sum of either water- orHCI-leachable Fe, V, Ni, Zn, Pb, or Cu. Samples 1 and 3 contained small amounts of water-leachable Zn (No. 1, 3.43; and No. 3, 6.35 ^ig/mg ROFA) or 1 MHCl-extractable Zn (No. 1, 3.62; and No. 3, 6.72 /xg/mg ROFA) but no other metals in significant quantities. Arsenic, Cd, Co, Mn, Cr, and Be were also analyzedin all samples; however, these metals were present in minimal quantities (water-leachable < 1.0 /xg/mg and 1 M HCl-extractablc <6.0 /xg/mg).

* Refer to Fig. 1 for matching collection site with the numbers assigned for each ROFA.c pH of leachate (supernatant of 8.33 mg/ml particle suspension).

analyzed for the presence of As, Be, Cd, Co, Cr, Cu, Fe, Mn,Ni, Pb, V, Zn, and sulfate. Of all water-leachable (soluble) or1.0 M HCl-extractable metals, ROFA primarily contained Fe,V, and/or Ni and therefore, only the concentrations of thesethree metals are given in Table 1. The leachable fractions ofROFA No. 1 and No. 3 also contained 3.43 and 6.35 /xg/mgZn, respectively, in addition to Fe, V, and Ni. All other metalswere present in minimal quantities in all ROFA samples (eitherwater-leachable or 1.0 M HCl-extractable). ROFA No. 1 andNo. 3, collected from the lower precipitator area (Fig. 1),contained the most water-leachable metal; they were the mostacidic and had the highest sulfate concentration. ROFA sample1 contained Fe, V, and Ni, whereas No. 3 contained the mostwater-leachable V and Ni, but the least Fe. In contrast, ROFAcollected from the upper precipitator (Fig. 1) contained theleast amounts of water-leachable total metal. Samples fromwithin the boiler, ash pit, or air heater basket (Fig. 1) containedprimarily water-leachable V which accounted for most of theirwater-leachable metal content. HCl-extractable total metalcontent or individual metals varied between samples with nospecific pattern or relationship between their amounts and aparticular collection site. Except for samples 1 and 3, sulfatedoes not seem to relate closely with the amounts of total orindividual leachable metal.

Acute Lung Injury Following ROFA Exposure

The time point selected for determination of acute pulmo-nary injury was 24 h. Our previous studies with other ROFAPM have indicated that at 24 h postinstillation, pathology andindices of pulmonary injury, in general, were greatest forBALF protein, LDH, and neutrophilic inflammation (Dreher etal, 1997, Kodavanti etal, 1997b). In the present study, BALF

analysis (24 h following single iti of ROFA at three differentdoses) demonstrated a dose-dependent increase in protein (Fig.2), LDH, hemoglobin (Table 2; only high dose data areshown), and NAG activity (data not shown). The increase inBALF protein, LDH, NAG, and hemoglobin varied with thetype of ROFA. ROFA containing the highest concentrations of

g/m

i t

c"3oa.• |

<

m

7000

6000

5000

4000

3000

2000

1000

- • - PM*1- • - PM#3- © - PM»«- e - PM#9- A - PM#10- ^ - PM#5- $ - PM#7- © - PM#»

A PM #2T PM »4

ROFA samples in the order ofdecreasing total water- jleachable metal content /

•

i

|

0 2 4 6 8 10Intratracheal Instillation (mg/kg Body Weight)

FIG. 2. Dose-dependent increases in BALF protein in the rat followingexposure to different ROFA samples. Samples are listed in order of decreasingwater-leachable total metal content. Solid lines ( ) indicate ROFA withwater-leachable metals >45 fig/mg and also highest Ni content, broken lines(—) indicate ROFA containing 15-40 /xg/mg metals, primarily V, and dottedlines (.. .) indicate ROFA containing <5 /ig/mg soluble metals. Valuesindicate means ± standard deviation of 18-20 controls and 6 exposed animals.The significant differences are not marked for clarity. Similar kinetics of LDHincreases in the BALF was observed following exposure to the 10 ROFAsamples. (Refer to Fig. 1 or Table 1 for ROFA numbers, collection sites,physical characteristics, and metal composition.)


208 KODAVANTI ET AL.

TABLE 2Changes in BALF Parameters Following Intratracheal Exposure of rats to 10 ROFA Samples at 8.33 mg/kg Body Wt Dose"

Samples in orderof decreasing

water-leachablemetal content

Control139

10657824

BALF LDHactivity (units/L)

50.5 ±447.0 ±261.3 ±149.8 +127.2 ±137.2 ±116.5 ±114.5 +170.3 ±83.2 ±76.2 ±

9.872.1*69.9*15.3*10.9*12.7*6.9*

10.7*16.1*10.96.0

BALFhemoglobin as

an index ofhemorrhage

(mg/ml)

1.51 ± 0.3715.87 ± 3.17*76.47 ± 30.4*

6.67 ± 2.425.41 ± 0.63*

18.5 ± 4.93*3.35 ± 0.358.46 ± 2.02*

15.74 ± 7.12*10.28 ± 2.45*8.30 ± 2.84*

Total lavageablecells/ml X 10~6

0.27 ± 0.050.84 + 0.09*1.24 + 0.38*1.91 +0.33*1.33 + 0.21*1.52 + 0.19*1.29 + 0.11*0.94 + 0.05*1.35 ±0.28*0.85 ±0.130.37 ± 0.05

Neutrophils/mlBALF X 10"6

0.05 ± 0.020.35 ± 0.07*0.78 ± 0.25*1.12 ±0.21*0.75 + 0.11*1.10 + 0.13*0.85 + 0.10*0.57 ± 0.07*1.02 ±0.20*0.64 ±0.12*0.19 + 0.05

Alveolarmacrophages/mlBALF X lO"6

0.19 ±0.020.40 ± 0.04*0.32 ± 0.08*0.63 ±0.11*0.46 ± 0.07*0.31 ±0.05*0.33 ± 0.02*0.28 + 0.01*0.25 + 0.080.16 ±0.030.15 ± 0.01

Eosinophils/mlBALF

0.2431.989.9

104.061 154 162.439.954.623.58.6

x 10"3

± 0.13± 16.2*±41.7*± 15.7*±26.1*± 10.5*± 12.1*± 12.1*±28.1*+ 7.4*+ 2.8*

" Values are given for only 8.33 mg/kg dose of ROFA and indicate means ± standard error of 18-20 saline-instilled and 6 ROFA-instilled rats.* Significant difference from saline-instilled controls (p < O.05).

water-leachable Fe, V, and Ni (ROFA No. 1) or V and Ni(ROFA No. 3) caused the largest increase in these biochemicalindices of lung injury (Fig. 2, Table 2). Incidentally, the sulfateconcentrations of these two PM samples were also the highest(Table 1). ROFA containing primarily soluble V (Nos. 6-10)caused less dramatic increases in protein, LDH, NAG, andhemoglobin. ROFA with the least soluble metals (Nos. 2 and 4)had the least effect on these parameters.

A significant positive correlation was found between ROFA-induced increase in BALF protein levels and ROFA content ofwater-leachable Ni (Fig. 3). Since these ROFA samples (Nos.1 and 3) also contained the most water-leachable total metaland V, regression analysis was used to investigate whether therelationship between ROFA-associated Ni and BALF proteinlevels was confounded by its total soluble metal or V content.The positive association between BALF protein and Ni contentof ROFA remained after inclusion of total metal and V valuesin the regression models. Increased BALF LDH activity fol-lowing ROFA exposure was also associated with its content ofwater-leachable Ni (r = 0.81, p = 0.0001) and this association,likewise, remained when total soluble metals and V valueswere included in the regression models. In contrast, ROFA-induced increases in BALF protein or LDH leakage were notassociated with its content of water-leachable V (r's < 0.1, pvalues >0.6). ROFA-induced increases in BALF hemoglobin(Table 2), taken as an index of capillary damage, and NAGactivity, an index of cell damage (data not shown), followed apattern similar to protein and LDH in terms of their associationwith its water-leachable Ni content.

BALF cell differentials showed increases in various celltypes dependent on the type of ROFA that rats were exposed to(neutrophils, Fig. 4). Of 10 ROFA samples, those with thehighest water-leachable V content, in contrast to those with

highest Ni, induced the greatest increase in BALF neutrophils.Correlation analysis indicated a significant positive associationbetween BALF neutrophil increases (Fig. 5) and water-leach-able V content of ROFA. This positive association remainedafter inclusion of ROFA-associated total metal and Ni in theregression models. On the contrary, ROFA-induced neutro-philic inflammation was not associated with its water-leachableNi or Fe or total metal contents (r's <0.2, p values >0.1).

Total and most subtype lavageable cells were also increasedfollowing ROFA exposure (Table 2). This was largely associ-ated with the water-leachable V content of ROFA. ROFA No.2, which contained a modest amount of water-leachable V butnot other metals, also caused cellular, primarily neutrophilic,influx. However, ROFA No. 4, which did not contain water-leachable V or other metals, produced the smallest cellularresponse. Of the lavageable cells, AM showed the greatestvariability in response to ROFA containing differing amountsof water-leachable V and Ni; no specific associations wereapparent (Table 2). A minimal number of eosinophils weredetected in control rats instilled with saline, but increased in allrats exposed to the ROFA samples (Table 2). Although thedegree of eosinophil influx varied, it appeared to be associatedwith water-leachable V.

Oxidation of AM Determined by Luminol-Enhanced CL

The time course of luminol-enhanced CL by AM fromhealthy rats was measured in vitro in the presence or in theabsence of three ROFA fractions prepared from its water(HBSS in case of in vitro studies) suspension (whole suspen-sion, supernatant, or pellet). Increase in AM CL over time inthe presence of whole ROFA suspension or its leachates isdepicted in Figs. 6A and 6B (data for the pellet fraction are not



10000^

800O;

600OJ

4000^

r = 0.70p= 0.0001

0 10 20 30 40 50 60

Total Soluble Metal (ng/mg PM)

a2Q.

IAAAJ"

8000-

6000-

4000-

2000-

r = 0.79p;= 0.0001

• ^*

•

•

0 5 10 15 20 25

Soluble Fe (|xg/mg PM)10000n

8000-r = 0.87

0 2.5 5 7.5 10 12.5 15

Soluble Ni (ng/mg PM)

FIG. 3. Correlations of ROFA-induced increase in BALF protein withexposure doses of ROFA-associated water-leachable total metal, Fe, and Ni.The instilled doses for water-leachable total metal, Fe, or Ni were calculated bymultiplying concentrations of metal with the instilled dose of ROFA (8.33mg/kg). Individual responses rather than means ± standard error are plotted.

shown). The peak CL occurred during 90-120 min following

the addition of whole ROFA suspensions or its leachate frac-

tions. Whole ROFA suspension (in buffer) containing water-

leachable V caused the most intense CL, while ROFA contain-

ing a combination of Fe, V, and Ni or only V and Ni caused a

less dramatic enhancement of CL. A small but apparent V-

related increase in AM CL could also be detected with whole

suspension or supernatant fractions of ROFA No. 2, which

contained small amounts of soluble V but no Ni. Incidentally,

as shown earlier, ROFA No. 2 also caused neutrophilic inflam-

mation in vivo. ROFA No. 4, which contained no soluble V,

did not cause an increase in AM CL. The metal-specific pattern

1600000

E 1400000

= 1200000

S. 10000000)•= 800000

O* 600000

a>

u.

t400000

200000

0

ROFA samples In the order ofdecreasing total water-teachable Tmetal content

0 2 4 6 8 10

Intratracheal Instillation (mg/kg Body Weight)

FIG. 4. BALF neutrophilic influx in rats following exposure to differentROFA samples. Samples are listed in order of decreasing water-leachable totalmetal content Solid lines ( ) indicate ROFA with soluble metals >45jig/mg and also highest Ni content, broken lines ( ) indicate ROFA con-taining 15-40 ng/mg metals, primarily V, and dotted lines (. ..) indicateROFA containing <5 fig/mg soluble metals. Values indicate means ± stan-dard deviation of 18-20 controls and 6 exposed animals. The significantdifferences are not marked for clarity. (Refer to Fig. 1 or Table 1 for ROFAnumbers, collection sites, physical characteristics, and metal composition.)

of changes in CL during incubation of AM with whole ROFA

suspension became even stronger when the incubations of AM

were done with its supernatant (leachate) fraction, suggesting

that the activation of AM in vitro was primarily dependent

upon the presence of water-leachable V. When the incubations

were done with the washed ROFA fraction (pellet), the asso-

ciation of V with CL became less significant (data not shown),

although the signals were still apparent probably due to surface

Ioo

1500000

2 1000000a

I500000

ffi

r = 0.45p = 0.0003

0 50 100 150 200 250 300 350 400

Water-Leachable V Dose (u,g/kg Body Weight)

FIG. 5. Correlation of ROFA-induced BALF neutrophilic influx and ex-posure dose of ROFA-associated water-leachable V. The instilled dose of Vwas calculated by multiplying water-leachable V concentration of ROFA andthe instilled dose of ROFA (8.33 mg/kg). Individual responses rather thanmeans ± standard error are plotted.



Incubation Time (minutes)

40 60 80 100 120 140 160

Incubation Time (minutes)

FIG. 6. Luminol-enhanced CL of AM following in vitro exposure to either whole ROFA suspension (A) or its water leachate (B). AM from normal controlrats (0.3-0.5 X 106 cellsyml) were incubated with albumin-conjugated luminol and ROFA (100-166 jig/lO6 AM) and the resultant CL was measuredphotometrically over a period of 120 or 150 min.

complexed metal. At the end of incubation period neither theviability of AM nor the LDH leakage was changed by anyROFA or any fractions (not shown).

DISCUSSION

Studies of occupational pulmonary injury in workers ex-posed to ROFA from the same power plant where the samplesfor the present experiments were collected have demonstratedupper and lower airway inflammation (Hauser et al, 1995a,b).Our previous animal studies with a fugitive ROFA (similar incharacteristics to the ROFA used in the present study) haveshown in vivo pulmonary inflammation, cytokine gene induc-tion, and fibrosis (Dreher et al, 1997; Kodavanti et at,1997a,b). These studies have demonstrated that Ni, but not V,when instilled in rats as pure metal salt suspension, at concen-trations that existed in the ROFA, account for the majority ofthe pulmonary injury caused by ROFA (Kodavanti et al.,1997b). Additionally, the appearance of lung lesions as well asthe kinetics of pulmonary inflammatory cytokine gene induc-tion differed between Ni and V or between Ni- and ROFA-exposed animals (Kodavanti et al., 1997b). These recent stud-ies (Dreher et al, 1997; Kodavanti et al., 1997b) alsosuggested that there were metal-metal interactions (notably Niand V or Fe) which were seemingly antagonistic in terms ofpathology and cytokine gene expression. All these studies weredone using only one prototypic ROFA sample or pure metalsalt preparation and suggested lesion severity differences be-tween Ni and V, but did not provide information on whatresponses were likely involved in metal-specific manner.

In vitro, V or ROFA from different sources has been shownto activate AM (Ghio et al., 1997) as well as airway epithelialcells via redox-sensitive pathways to cause inflammatory cy-

tokine gene induction and cytotoxicity, while Ni, on the otherhand, has been shown not to cause any of these effects (Beckeret al., 1996; Dye et al., 1997a,b; Samet et al., 1997). Ni hasbeen shown to inhibit AM phagocytosis of latex particles atconcentrations that are noncytotoxic, while V has been shownto be cytotoxic without altering AM phagocytosis in vitro(Graham et al., 1975). These and the above studies led us tohypothesize that ROFAs that differ in water-leachable V andNi content will cause different inflammatory and vascularleakage responses in the lung in vivo. Furthermore, the abilityof these ROFA to activate AM in vitro will also be dependenton their water-leachable Ni and V content. Although puremetal suspensions have been tested previously in this regard(Graham et al., 1975), the role of PM-associated Ni and V inactivating AM has not been compared.

The relationship between in vivo indices of ROFA-inducedpulmonary injury and its water-leachable levels of metals (totalmetals, V, Ni, and Fe) and sulfate (in all 10 ROFA samples) wasinvestigated using regression analysis. Our study demonstrated asignificant positive association between ROFA-induced increasesin BALF protein, LDH, NAG, and hemoglobin (but not neutro-phils) and its content of water-leachable Ni. In contrast, ROFA-induced increases in BALF neutrophils (but not protein, LDH,NAG, or hemoglobin) were positively associated with its water-leachable V content V as metal salt aerosol has been shown toinduce marked pulmonary inflammatory response in the rat(Pierce et al, 1996). The associations between water-leachable Niand V content of ROFA with biochemical indices of injury(BALF protein and LDH) and inflammatory cell response, respec-tively, were independent of each other and the presence of water-leachable total metal and sulfate.

Differences in pulmonary responses to ROFA based upon its



water-leachable Ni and V content could be due to: (1) differ-ences in pulmonary epithelial cell types that directly contactROFA, (2) ROFA-associated V and Ni may elicit pulmonaryinjury by different mechanisms (Goebeler et al., 1993), and (3)acute toxicity can be influenced by the half-life of leachablemetal in the lung and their clearance (Conklin et al, 1982;Hirano et al., 1994; Benson et al., 1995). ROFA-associated Nimay inhibit clearance of particles as it has been shown toinhibit AM phagocytosis (Graham et al., 1975). There is alsosome evidence that Ni, while minimally toxic to the airwayepithelial cells in vitro, causes selective effects on alveolarepithelial cells as reflected by alveolar type II cell hyperplasiaand increased surfactant secretion in vivo (Benson et al., 1986,1995; Camner and Johansson, 1992). It is likely that ROFAwhich contains Ni may be selectively toxic to alveolar paren-chymal cells leading to microvascular and/or cellular leakageassociated with increases in BALF levels of protein, LDH,NAG, and hemoglobin. The influx of inflammatory cells intothe BALF of these rats may be due to the leakage of extraneousfluid and/or the effects of V as one of the components of ROFAsamples. V alone is known to cause inflammatory changes invivo (Pierce et al, 1996).

Metal-metal interactions may also have had some impact onoverall pulmonary injury from ROFA; however, such possibil-ities could not be taken into account with the design of presentexperiments. We have previously shown that when Ni, V, andFe are mixed together, the pathology and cytokine gene induc-tion caused by these three metals together were less severe thanthat caused by Ni alone (Kodavanti et al., 1997). Dreher et al.(1997) have shown that when V and Ni were mixed together,the injury caused by the mixture was also less than the injurycaused by Ni alone. Thus, in the present study, ROFA whichcontained both water-leachable V and Ni (Nos. 1 and 3), it islikely that the effects of water-leachable Ni may have beenantagonized, at least in part, by V. Despite this possibility, ourstudy showed that Ni is associated with greater pulmonaryvascular leakage and damage than V. It is noteworthy thatROFA No. 4 with no detectable water-leachable metal contentcaused the least protein leakage and inflammatory response.

The purpose of the in vitro experiments was to determine whetherthe ability of 10 different ROFA to activate AM differ based upontheir water-leachable Ni and V content, and if so, how is it related tothe type of pulmonary response produced by ROFA in vivo. In thepresent study, luminol-enhanced CL measurements (as an index ofoxidation) by AM exposed to ROFA clearly showed that the samplescontaining water-leachable V were much more potent than thosecontaining Ni in addition to V and Fe. Further, this was confirmed bythe fact that the dependency of CL signal on leachable V becameeven stronger when the leachate fraction of ROFA containing solublemetal was used. Our in vitro results using ROFA suspension are inagreement with previous reports which used pure metal salt prepara-tions and showed that V, but not Ni, can cause activation of AM(Graham et al, 1975) and airway epithelial cells (Samet et al, 1997;Dye et al, 1997b). V also has been shown to cause more oxidant

production than Ni when incubated with AM from various animalspecies (Fisher et al, 1986; Waseem et al, 1993; Geertz et al, 1994;Ghio et al, 1997). V-containing ROFA has been shown to activateredox-sensitive nuclear factors with subsequent activation of inflam-matory cytokine genes in airway epithelial cells (Becker et al, 1996;Samet et al, 1997). Whether these oxidative effects of V or V-containing ROFA are specific to airway AM and epithelial cells, orwhether V can cause similar reactions with alveolar cells, has notbeen investigated. Also, whether inflammation resulting from V-containing ROFA is at least in part due to its ability to oxidativelyactivate AM, or this is an independent effect which does not influenceinflammatory response, remains to be investigated. Our results showthat in vivo inflammation and in vitro macrophage activation areeffects common to V-containing ROFA. ROFA containing water-leachable Ni (in addition to V) were most toxic in vivo in terms ofprotein leakage but less able to activate AM than those containingonly V. This either could be due to lack of Ni effects on AM or couldbe due to inhibition of V-induced AM oxidative burst as has beenreported by Graham et al (1975). Analysis of cell viability afterincubation of cells with ROFA, however, showed that Ni-containingROFA did not cause an increase in LDH release or cell death (datanot shown). In summary, our studies demonstrate that ROFA con-taining water-leachable V and Ni can cause pulmonary injury bydifferent pathways.

ACKNOWLEDGMENTS

The authors acknowledge the technical help of Mr. Richard Jaskot and Mr.James Lehmann. Z.H.M. was supported by EPA cooperative agreement withDuke University Medical Center (Durham, NC) (EPA/CR819093). Thesestudies were also supported in part from Grants ES05947, ES07069, andES00002 from NIEHS and U60/CCU109979 from NIOSH.

R E F E R E N C E S

Aust, S. D., Morehouse, L. A., and Thomas, C. E. (1985). Role of metals inoxygen radical reactions. J. Free Radical Biol. Med. 1, 3-25.

Becker, S., Soukup, J. M., Gilmour, M. I., and Devlin, R. B. (1996). Stimu-lation of human and rat alveolar macrophages by urban air particulates:Effects on oxidant radical generation and cytokine production. Toxicol.Appl. Pharmacol. 141, 637-648.

Benson, J. M., Henderson, R. F., McClellan, R. O., Hanson, R. L., and Rebar,A. H. (1986). Comparative acute toxicity of four nickel compounds to F344rat lung. Fundam. Appl. Toxicol. 7, 340-347.

Benson, J. M., Chang, I.-Y., Sung, S., Hahn, F. F., Kennedy, C. H., Barr, E. B.,Maples, K. R., and Snipes, M. B. (1995). Particle clearance and histopa-thology in lungs of F344/N rats and B6C3F, mice inhaling nickel oxide ornickel sulfate. Fundam. Appl. Toxicol. 28, 232-244.

Berg, I., SchJuter, T., and Gercken, G. (1993). Increase of bovine alveolarmacrophage superoxide anion and hydrogen peroxide release by dusts ofdifferent origin. J. Toxicol. Environ. Health 39, 341—354.

Brennan, P. C , and Kirchner, F. R. (1985). Quantitative assessment of rabbitalveolar macrophage function by chemiluminescence. Environ. Res. 37,452-460.

Camner, P., and Johansson, A. (1992). Reaction of alveolar macrophages toinhaled metal aerosols. Environ. Health Perspecl. 97, 185-188.

Conklin, A. W., Skinner, C. S., Felten, T. L., and Sanders, C. L. (1982).



Clearance and distribution of intratracheally instilled ^vanadium com-pounds in the rat Toxicol. Lett. 11, 199-203.

Costa, D. L., Lehmann, J. R., Harold, W. M., and Drew, R. T. (1986).Transoral trachea) intubation of rodent using a fiberoptic laryngoscope. Lab.Anim. Sci. 36, 256-261.

Costa, D. L., and Dreher, K. L. (1997). Bioavailable transition metals inpaniculate matter mediate cardiopulmonary injury in healthy and compro-mised animal models. Environ. Health Perspect. 105(Suppl. 5), 1053-1060.

Dockery, D. W., Pope, A., Ill, Xu, X., Spengler, J. D., Ware, J. H., Fay, M. E.,Ferris, B. G., Jr., and Speizer, F. E. (1993). An association between airpollution and mortality in six US cities. N. Engl. J. Med 329, 1753-1759.

Dreher, K. L., Jaskot, R. H., Lehmann, J. R., Richards, J. H., Mcgee, J. K.,Ghio, A. J., and Costa, D. L. (1997). Soluble transition metals mediateresidual oil fly ash induced acute lung injury. J. Toxicol. Environ. Health 50,285-305.

Dye, J. A., Adler, K. B., Richards, J. H., and Dreher, K. L. (1997a). Epithelialinjury induced by exposure to residual oil fly-ash particles: Role of reactiveoxygen species? Am. J. Respir. Cell. Mol. Biol. 17, 625-633.

Dye, J. A., Adler, K. B., Richards, J. H., and Dreher, K. L. (1997b). Airwayepithelial cell responses to residual oil fly ash (ROFA) particles: Contribu-tion of soluble transition metals. Am. J. Respir. Crit. Care Med. 155, A197.

Fisher, G. L., McNeill, K. L., and Democko, C. J. (1986). Trace elementinteractions affecting pulmonary macrophage cytotoxicity. Environ. Res. 39,164-171.

Geertz, R., Gulyas, H., and Gercken, G. (1994). Cytotoxicity of dust constit-uents towards alveolar macrophages: Interactions of heavy metal com-pounds. Toxicology 86, 13-27.

Ghio, A. J., Stonehuemer, J., Pritchard, R. J., Piantadosi, C. A., Quigley, D. R.,Dreher, K. L., and Costa, D. L. (19%). Humic-like substances in airpollution particulates correlate with concentrations of transition metals andoxidant generation, lnhal. Toxicol 8, 479-494.

Ghio, A. J., Meng, Z. H., Hatch, G. E., and Costa, D. L. (1997). Luminol-enhanced chemiluminescence after in vitro exposures of rat alveolar mac-rophages to oil fly ash is metal dependent lnhal Toxicol. 9, 255-271.

Goebeler, M., Meinardus-Hager, G., Roth, J., Goerdt, S., and Sorg, C. (1993).Nickel chloride and cobalt chloride, two common contact sensitizers, di-rectly induce expression of intracellular adhesion molecule-1 (ICAM-1),vascular cell adhesion molecule-1 (VCAM-1), and endothelial leukocyteadhesion molecule (ELAM-1) by endothelial cells. /. Invest. Dermatol. 100,759-765.

Graham, J. A., Gardner, D. E., Waters, M. D., and Coffin, D. L. (1975). Effectof trace metals on phagocytosis by alveolar macrophages. Infect. Immun. 11,1278-1283.

Gulyas, H., and Gercken, G. (1988). Cytotoxicity to alveolar macrophages ofairborne particles and waste incinerator fly-ash fractions. Environ. Pollut.51, 1-18.

Hatch, G. E., Gardner, G. E., and Menzel, D. B. (1980). Stimulation of oxidantproduction in alveolar macTophages by pollutant and latex particles. Envi-ron. Res. 23, 121-136.

Hatch, G. E., Boykin, E., Graham, J. A., Lewtas, J., Pott, L. F., Loud, K., andMumford, J. L. (1985). Inhalable particles and pulmonary host defense: Invivo and in vitro effects of ambient air and combustion particles. Environ.Res. 36, 67-80.

Hatch, G. E., Gardner, D. E., and Menzel, D. B. (1978). Chemiluminescenceof phagocytic cells caused by N-formylmethionyl peptides. J. Environ. Med147, 182-195.

Hatch, G. E., Slade, R. S., Stead, A. G., and Graham, J. A. (1986). Speciescomparison of acute inhalation toxicity of ozone and phosgene. / . ToxicolEnviron. HeaUh 19, 43-53.

Hauser, R., Elreedy, S., Hoppin, J. A., and Christiani, D. C. (1995a). Airway

obstruction in boilermakers exposed to fuel oil ash: A perspective investi-gation. Am. J. Respir. Crit. Care Med 152, 1478-1484.

Hauser, R., Elreedy, S., Hoppin, J. A., and Christiani, D. C. (1995b). Upperairway response in workers exposed to fuel oil ash: Nasal lavage analysis.Occup. Environ. Med 52, 353-358.

Hirano, S., Shimada, T., Osugi, J., Kodama, N., and Suzuki, K. T. (1994).Pulmonary clearance and inflammatory potency of intratracheally instilledor acutely inhaled nickel sulfate in rats. Arch. Toxicol. 68, 548-554.

International Agency for Research on Cancer (IARC) (1990). Monographs onthe Evaluation of Carcinogenic Risks to Humans, Vol 49, Chromium,Nickel, and Welding, p. 257. International Agency for Research on Cancer,Lyon.

Kodavanti, U. P., Jaskot, R. H., Su, W. Y., Costa, D. L., Ghio, A. J., andDreher, K. L. (1997a). Genetic variability in combustion particle-inducedchronic lung injury. Am. J. Physiol Til, L521-L532.

Kodavanti, U. P., Jaskot, R. H., Costa, D. L., and Dreher, K. L. (1997b).Pulmonary proinflammatory gene induction following acute exposure toresidual oil fly ash: Roles of particle-associated metals, lnhal. Toxicol. 9,679-701.

Lamm, S. H., Hall, T. A., Engel, A., Rueter, F. H., and White, L. D. (1994).PM10 particulates: Are they the major determinant of pediatric respiratoryadmissions in Utah County, Utah (1985-1989). Ann. Occup. Hyg. 38(Suppl.1), 969-972.

Levy, B. S., Hoffman, L., and Gottsegen, S. (1984). Boilermakers' bronchitis:Respiratory tract irritation associated with vanadium pentoxide exposureduring oil-to-coal conversion of a power plant J. Occup. Med 26, 567—570.

Menne, T., Chnstophersen, J., and Green, A. (1989). Epidemiology of nickeldermatitis. In Nickel and the Skin: Immunology and Toxicology (H. I.Maibach and T. Menne, Eds.), pp. 109-116. CRC Press, Boca Raton, FL.

Pierce, L. M., Alessandrini, F., Godleski, J. J., and Paulauskis, J. D. (1996).Vanadium-induced chemokine mRNA expression and pulmonary inflam-mation. Toxicol. Appl. Pharmacol 138, 1-11.

Pritchard, R. J., Ghio, A. J., Lehmann, J. R., Winsett, D. W., Tepper, J. S.,Park, P., Gilmour, M. I., Dreher, K. L., and Costa, D. L. (1996). Oxidantgeneration and lung injury after paniculate exposure increase with theconcentrations of associated metals, lnhal. Toxicol. 8, 457-477.

Rendall, R. E., Phillips, J. A., and Renton, K. A. (1994). Death followingexposure to fine paniculate nickel from a metal arc process. Ann. Occup.Hyg. 38, 921-930.

Samet J. M., Stonehuerner, J., Reed, W., Devlin, R. B., Dailey, L. A.,Kennedy, T. P., Bromberg, P. A., and Ghio, A. J. (1997). In vitro exposureto oil fly ash disrupts intracellular signaling in human bronchial epithelialcells. Am. J. Respir. Crit. Care Med. 155, A432.

Schluter, T., Berg, I., Dorger, M., and Gercken, G. (1995). Effect of heavymetal ions on the release of reactive oxygen intermediates by bovinealveolar macrophages. Toxicology 98, 47-55.

Schwartz, J. (1994). What are people dying of on high air pollution days?Environ. Res. 64, 26-35.

Spengler, J. D., and Thurston, G. D. (1983). Mass and elemental compositionof fine and course particles in six U.S. Cities. Air Pollui. Cont. Assoc. J. 33,1162-1171.

Stohs, S. J., and Bagchi, D. (1995). Oxidative mechanisms in the toxicity ofmetal ions. Free Radical Biol. Med. 18, 321-336.

U.S. Environmental Protection Agency, Environmental Monitoring SystemsLaboratory (1992). In Methods for the determination of metals in environ-mental samples, pp. 33-93. CRC Press, Boca Raton, FL.

Waseem, M., Bajpai, R., and Kaw, J. L. (1993). Reaction of pulmonarymacrophages exposed to nickel and cadmium in vitro. J. Environ. PatholToxicol Oncol 12, 47-54.


Pulmonary Responses to Oil Fly Ash Particles in the Rat Differ by ...

Documents

Transcript of Pulmonary Responses to Oil Fly Ash Particles in the Rat Differ by ...