Ultrafine Carbon Black Particles Inhibit Human Lung ... - ATS Journals

Ultrafine Carbon Black Particles Inhibit Human Lung Fibroblast-Mediated Collagen Gel Contraction Huijung Kim, Xiangde Liu, Tetsu Kobayashi, Tadashi Kohyama, Fu-Qiang Wen, Debra J. Romberger, Heather Conner, Peter S. Gilmour, Kenneth Donaldson, William MacNee, and Stephen I. Rennard Pulmonary and Critical Care Medicine Section, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, Nebraska; and ELEGI/Colt Research Laboratory, University of Edinburgh, Edinburgh, Scotland, United Kingdom

Both acute and chronic exposure to particulates have been associated with increased mortality and morbidity from a number of causes, including chronic obstructive pulmonary disease and other chronic lung diseases. The current study evaluated the hypothesis that ultrafine carbon particles, a component of ambient particulates, could affect tissue repair. To assess this, the three-dimensional collagen gel contraction model was used. Ultrafine carbon black particles, but not fine carbon black, inhibited fibroblast-mediated collagen gel contraction. Although previous research has indicated that inflammatory effects of ultrafine carbon black particles are mediated by oxidant mechanisms, the current study suggests that ultrafine carbon black’s inhibition of fibroblast gel contraction is mediated by the binding of both fibronectin and transforming growth factor (TGF)-␤ to the ultrafine particles. Binding of TGF-␤ was associated with a reduction in nuclear localization of Smads, indicative of inhibition of TGF-␤ signal transduction. There was also a decrease in fibronectin mRNA, consistent with a decrease in TGF-␤– mediated response. Taken together, these results demonstrate the ability of ultrafine particles to contribute to altered tissue repair and extend the known mechanisms by which these biologically active particles exert their effects.

High levels of ambient respirable pollutants are associated with increased respiratory and cardiovascular morbidity and mortality (1–4). Even brief, high-level exposure to particulates ⬎ 10 ␮m in diameter (PM10) have been linked to increases in mortality (5, 6). In addition, chronic exposure to these types of particulates has accompanied an increased risk for chronic disease, including chronic obstructive pulmonary disease (COPD) (7, 8). Ultrafine particles that emanate from combustion processes make up a considerable proportion by number of urban PM10 (6). A number of studies have suggested that ultrafine particles ⬎ 0.1 ␮M in diameter, are more pathogenic than larger particles of the same mineral type (9–12) and mass.

(Received in original form December 10, 2001 and in revised form August 22, 2002) Address correspondence to: Stephen I. Rennard, M.D., University of Nebraska Medical Center, 985125 Nebraska Medical Center, Omaha, NE 68198-5125. E-mail: [email protected] Abbreviations: carbon black, CB; chronic obstructive pulmonary disease, COPD; deferoxamine mesylate, DFM; Dulbecco’s modified Eagle’s medium, DMEM; fetal calf serum, FCS; fluorescein isothiocyanate, FITC; human fetal lung fibroblasts, HFL-1; interleukin, IL; lactic dehydrogenase, LDH; N-acetylL-cysteine, NAC; phosphate-buffered saline, PBS; serum-free DMEM, SF-DMEM; transforming growth factor, TGF; ultrafine carbon black, ufCB. Am. J. Respir. Cell Mol. Biol. Vol. 28, pp. 111–121, 2003 DOI: 10.1165/rcmb.4796 Internet address: www.atsjournals.org

Ultrafine particles are capable of inducing an intense inflammatory reaction (13, 14). In in vitro studies, these particles are capable of inducing the release of proinflammatory cytokines, including interleukin (IL)-8, IL-6 and tumor necrosis factor (TNF)-␣ from human bronchial epithelial cells (15–19). This effect appears to be mediated through the induction of oxidative stress (20). Both the metal chelator deferoxamine and the free radical scavenger N-acetyl-L-cysteine (NAC) can inhibit this in vitro response (18, 21). Although the chemical composition of inhaled particles undoubtedly contributes to their toxicity, current evidence suggests that particle size per se also plays a role. In this regard, ultrafine carbon black instilled into the rat lung induces a marked influx of neutrophils (14–22). Chronic lung disease, to which inhalation of particulate matter appears to contribute, is characterized by limitation of airflow due to alterations in tissue structure. In addition to tissue damage caused by inflammation, tissue remodeling contributes to the structural and functional alterations in the lungs. The current study was designed to determine if ultrafine particles could modify tissue remodeling processes. Fibroblasts cultured in three-dimensional collagen gels are capable of remodeling their surrounding matrix. Fibroblasts attach to the collagen gels by integrin-mediated mechanisms, and by exerting mechanical tension can cause these gels to contract (23–25). This contractile process can be modulated by a variety of exogenous mediators and has been used to model the tissue contraction that characterizes both the formation of fibrous scar and fibrosis (23, 24, 26, 27). Using this system, the current study demonstrates that ultrafine carbon black particles can inhibit fibroblast contraction of collagen gels and that this process is due to an oxidant-independent mechanism, namely binding of procontractile factors to the particles.

Materials and Methods Cell Culture Human fetal lung fibroblasts (HFL-1; lung, diploid, human) were obtained from the American Type Culture Collection (Rockville, MD). The cells were cultured in 100-mm tissue culture dishes (Falcon; Becton-Dickinson Labware, Lincoln Park, NJ) in Dulbecco’s Modified Eagle’s Medium (DMEM), supplemented with 10% fetal calf serum (FCS), 50 U/ml penicillin G sodium, 50 ␮g/ml streptomycin sulfate (penicillin-streptomycin; Invitrogen, Life Technologies, Grand Island, NY), and 1 ␮g/ml amphotericin B (Pharma-Tek, Elvira, NY). The fibroblasts were refed three times weekly and cells between passages 16 and 20 were used. Confluent fibroblasts were detached by 0.25% trypsin in 0.5% mM EDTA and resuspended

112

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 28 2003

in DMEM without serum (SF-DMEM) for mixing with collagen to make gels.

Materials Fine carbon black particles (CB) have a mean diameter of 260.2 nm ⫾ 13.7, surface area of 7.9 m2/g and no detectable iron. Ultrafine carbon black particles (ufCB) have a mean diameter of 14.3 nm ⫾ 0.6, surface area of 253.99 m2/g, and contain low amounts of iron (19 ng/mg). These CB and ufCB particles were kindly provided by Dr. Peter Gilmour (Respiratory Medicine Unit, ELEGI/Colt Laboratory, Edinburgh University). Type I Collagen (rat tail tendon collagen) was extracted from rat-tail tendons using a previously published method (28). Briefly, tendons were excised from rat tails. After washing these tendons several times with Tris-buffered saline (0.9% NaCl, 10 mM Tris, Ph 7.5), and 50%, 75%, 95%, and 100% ethanol, the collagen was extracted in 6 mM hydrochloric acid. Protein concentration was determined by weighing a lyophilized aliquot from each batch of collagen. The RTTC was stored at 4⬚C until use. NAC, mannitol, catalase, and deferoxamine mesylate (DFM) were purchased from Sigma (St. Louis, MO).

Collagen Gel Preparation and Contraction Assay

Figure 1. Effect of ultrafine carbon black (ufCB) particles on the fibroblast-mediated collagen gel contraction. Fibroblasts were cast into collagen gels with or without ufCB (100 ␮g/ml; triangles) or CB (100 ␮g/ml; squares). Diamonds, control. Gels were released and cultured in 5 ml of SF-DMEM for 5 d. Vertical axis: gel area expressed as % of initial area; horizontal axis: time in days. Data are mean ⫾ SEM for three separate experiments, each performed in triplicate. *P ⬍ 0.001, corrected by Bonferroni test.

Collagen gels were prepared by mixing collagen solution, distilled water, and 4⫻ DMEM with or without ufCB or CB particles so that the final concentration was 1⫻ DMEM, collagen concentration was 0.75 mg/ml, and cell concentration was 3 ⫻ 105 cells/ml. Finally, 500 ␮l of this gel solution was cast into each well of a 24-well tissue culture plate (Falcon; Becton-Dickson). The solution was polymerized for 20 min at room temperature. The gels were gently released from the culture plates and transferred to 60-mm tissue culture dishes, which contained 5 ml of SF-DMEM. The gels were then incubated at 37⬚C in a 5% CO2 atmosphere. The area of each gel was measured daily with an image analyzer (Optomax, Burlington, MA). Data are expressed as the percentage of area compared with the initial gel area.

with MTT solution (0.25 mg/ml in serum-free DMEM, 0.5 ml/gel) for 12 h. Gels were washed once with distilled water. Formazan crystals were then dissolved with dimethyl sulfoxide (500 ␮l/gel) by shaking 4–6 h at room temperature, then absorbance at wavelength of 540 nm was determined with a microplate reader (BioRad). Absolute OD value was obtained and expressed as percent of control. Finally, LIVE/DEAD staining was performed following manufacturer’s instruction (Molecular Probe, Eugene, OR). After staining, cell viability and morphology were observed and pictured under fluorescence microscope.

Hydroxyproline Assay Hydroxyproline content was determined by a modification of previously published methods (29, 30). Briefly, gels were transferred to Eppendorf tubes, where they were solubilized by heating at 65⬚C for 10 min. The samples then were centrifuged at 2,000 ⫻ g for 5 min, and the supernatant harvested for the hydroxyproline assay. Samples (20 ␮l) were mixed with 30 ␮l of 3.3 N NaOH and then autoclaved at 120⬚C for 20 min. Afterward, 450 ␮l of 0.056 M Chloramine-T reagent was added and incubated at room temperature for 25 min. Ehrlich’s aldehyde reagent (0.25 mg/ml, 0.5 ml/ sample) was then added and the chromophore was developed by incubating the samples at 65⬚C for 20 min. Absorbance was measured and analyzed at 540 nm with Bench Mark and Microplate Manager III software (Bio-Rad, Hercules, CA).

Cell Viability and Morphology Assays To determine cell viability, three assays were performed with replicate samples. First, conditioned media were harvested on Days 3 and 5, and lactic dehydrogenase (LDH) amount was measured following manufacturer’s instruction (LDH kit, Cat # 500; Sigma). Second, the MTT assay was performed on Days 3 and 5, respectively, with modifications of previously published methods (31). Briefly, on Days 3 or 5, fibroblast-populated gels were incubated

Measurement of Fibronectin by Enzyme-Linked Immunosorbent Assay The conditioned medium was harvested and gels were solubilized with collagenase (0.25 mg/ml, 0.5 ml/gel) at 37⬚C for 1–2 h. The supernatant was quantified by enzyme-linked immunosorbent assay (ELISA). The amount of fibronectin in the conditioned media and solubilized gels was measured after culturing the gels for 5 d, using an ELISA that is specific for human fibronectin, and which does not detect bovine fibronectin. (32).

Measurement of Transforming Growth Factor-␤1 by ELISA Transforming growth factor (TGF)-␤1 concentration in supernatant from solubilized gels and conditioned media was determined by ELISA, using commercially available materials (R&D Systems, Minneapolis, MN). Briefly, 96-well, flat-bottomed microtiter plates (Dynatech, Chantilly, VA) were coated with monoclonal anti– TGF-␤1 antibodies (clone 9,016.2), at 4⬚C overnight. After washing three times (5 min each), standard and samples (with or without acidification: 500 ␮l sample was mixed with 100 ␮l of 1 N HCL and sat at room temperature for 10 min, then neutralized with 100 ␮l of 1.2 N NaOH/0.5M HEPES) were added and incubated at

Kim, Liu, Kobayashi, et al.: Particulates Inhibit Collagen Contraction

113

tion of genomic DNA, 1 ␮g of total RNA was treated with DNase I (Invitrogen) for 15 min at room temperature. The reaction was stopped with 25 mM EDTA and heated at 65⬚C for 10 min followed by 95⬚C for 5 min. For complementary DNA (cDNA) synthesis, ⬃ 400 ng of total RNA was transcribed with cDNA transcription reagents (Applied Biosystems, Foster City, CA) with use of random hexamers, and the cDNA was used for quantitative real-time polymerase chain reaction (PCR).

Quantitative Reverse-Transcriptase

Figure 2. Concentration-dependence effect of ufCB on the fibroblast-mediated collagen gel contraction. Fibroblasts were cast into collagen gels with or without different concentrations of ufCB (open diamonds, 100 ␮g/ml; open squares, 75 ␮g/ml; open triangles, 50 ␮g/ml; filled squares, 25 ␮g/ml; filled triangles, 10 ␮g/ml; open circles, 1 ␮g/ml; filled circles, control). Gels were cultured in 5 ml of SF-DMEM and measured daily for 5 d. Vertical axis: gel area expressed as % of initial area; horizontal axis: time in days. Data are mean ⫾ SEM for three separate experiments each performed in triplicate. *P ⬍ 0.004 at each time point by Bonferroni procedure.

room temperature for 2 h. After washing again, biotinylated anti–TGF-␤1 antibodies (100 ng/ml) were applied, then the plate was incubated at room temperature for 1 h. After another washing, HRP-streptoavidin (1:20,000 dilution; Zymed, San Francisco, CA) was added and incubated for 1 h at room temperature. After a final wash, substrate TMB (3,3⬘, 5⬘5-tetramethylbenzidine; Sigma) was added and allowed to develop for 30 min to 1 h at room temperature. The reaction was stopped by 1 M H2SO4 and read at 450 nm with a microplate reader.

Gene expression was measured with the use of an ABI Prism 7700 Sequence Detection System (Applied Biosystems) as described previously (33). Primers and TaqMan probes were designed using the Primer Express TM 1.0 (Applied Biosystems) software to amplify fewer than 150 base pairs. Probes were labeled at the 5⬘ end with the reporter dye molecule FAM (6-carboxy-fluorescein) and at the 3⬘ end with quencher dye molecule TAMARA (6-carboxytetramethyl-rhodamine). Real-time PCRs of DNA specimens were conducted in a total volume of 50 ␮l with 1⫻ TaqMan Master Mix (Applied Biosystems) and primers at 300 nM and probes at 200 nM. Primer sequences were as follows: TGF-␤1 (forward), 5⬘-CGA GCC TGA GGC CGA CTA C-3⬘; TGF-␤1 (backward), 5⬘-AGA ATT CGT TGT GGG TTT CCA-3⬘; TGF-␤1 (probe), 6FAMCCA AGG AGG TCA CCC GCG TGC-TAMRA; Fibronectin EIIIA (forward), 5⬘-ATG TCG ATT CCA TCA AAA TTG CT3⬘; Fibronectin EIIIA (backward), 5⬘-CTG CAG TGT CTT CTT CAC CAT CA-3⬘; Fibronectin EIIIA (probe), 6FAM-CCT ACT CGA GCC CTG AGG ATG GAA TCC-TAMRA; Fibronectin EIIIB (forward), 5⬘-GAG GTG GAC CCC GCT AAA CT-3⬘; Fibronectin EIIIB (backward), 5⬘-TAC CTT CTC CTG CCG CAA CTA-3⬘; Fibronectin EIIIB (probe), 6FAM-TCC ACC ATT ATT GGG TAC CGC ATC ACA -TAMRA; GAPDH (forward), 5⬘-CCA GGA AAT GAG CTT GAG AAA GT-3⬘; GAPDH (reverse), 5⬘-CCC ACT CCT CCA CCT TTG AC-3⬘; GAPDH (probe), FAM-CGT TGA GGG CAA TGC CAG CCC-TAMRA. Thermal cycler parameters included 2 min at 50⬚C, 10 min at 95⬚C, and 40 cycles involving denaturation at 95⬚C for 15 s and annealing/extension at 60⬚C for 1 min. Target genes were expressed relative to 105 ⫻ GAPDH units.

RNA Isolation and Complementary DNA Synthesis Total RNA was extracted from cell pellets with acid guanidine monothiocynate, precipitated with isopropyl alcohol, and dissolved in TE buffer. Total RNA was quantified with a spectrophotometer (Pharmacia Biotech, Piscataway, NJ). To rid possible contamina-

Immunohistochemical Staining for Fibronectin The localization of fibronectin around ufCB particles was studied by immunohistochemical staining. For this purpose, fibroblasts (15,000 cells/ml) were cast into collagen gels with or without ufCB

Figure 3. Comparison of morphology of HFL-1 cells cultured in the ufCB and CB in collagen gels. Fibroblast-populated collagen gels with or without ufCB or CB (100 ␮g/ml) were prepared and cultured for 3 d. Cells were stained with a LIVE/DEAD kit for 20 min at 37⬚C at 5% CO2 atmosphere. Cytoplasm was stained with green fluorescence and observed under fluorescence microscopy (⫻400). (A ) Control gels. (B ) Gels containing ufCB. (C ) Gels containing CB.

114


Figure 4. Concentration-dependent effect of ufCB on the release of TGF-␤ and fibronectin in the conditioned media. Fibroblasts were cast into collagen gels with or without different concentrations of ufCB. Gels were cultured in SF-DMEM for 5 d. Media in which gels were floated (open bars), as well as collagen gels (solid bars) following collagenous digestion, were collected for quantification of TGF-␤1 (A ) and fibronectin (B ) by ELISA. Vertical axis: TGF-␤1/ fibronectin concentration; horizontal axis: conditions. Data are mean ⫾ SEM for one representative experiment done in triplicate. Repeat experiments (n ⫽ 3) yielded similar results, although the absolute concentrations produced varied. *P ⬍ 0.0001, compared with control, corrected by Bonferroni test.

100 ␮g/ml. Cultures at 0 and after 5 d were harvested for fibronectin staining. Gels were fixed with 10% formalin solution and embedded in paraffin. The paraffin-embedded section was then baked at 70⬚C for 20 min, de-paraffinized in Xylene three times for 10 min, rehydrated in different concentrations of ethanol, and washed in PBS. Then, gels were blocked with 1% normal horse serum/PBS for 30 min and incubated with primary antibodies (rabbit polyclonal anti-fibronectin antibody; DAKO Corp., Carpinteria, CA) for 2 h at room temperature. Then, after washing three times, gels were incubated with biotinylated goat anti-rabbit secondary antibody (Vectorstain ABC kit; Vector Laboratories, Burlingame, CA) for 45 min at room temperature. After the gels were washed again, they were incubated with FITC-streptoavidin for fluorescence staining for 30 min at room temperature. Localization of stained fibronectin was examined by fluorescence with fluorescence microscopy.

well) were seeded into 8-well Lab-Tek II chamber glass slides (Nalge Nunc International, Naperville, IL) with 10% FCS-DMEM. After reaching ⬃ 70% confluence, media was changed to SFDMEM with or without ufCB 100 ␮g/ml for 3 d. Following washing with DPBS, the cells were fixed with 4% formaldehyde/PBS for 10 min. After washing again with PBS, the cells were permeabilized with 0.5% Triton X-100/PBS for 10 min., blocked with 10% rabbit serum/PBS at 37⬚C for 1 h and incubated overnight with 4 ␮g/ml primary antibodies (goat polyclonal anti-Smad 2, 3, and 4; Santa Cruz Biotechnology, Santa Cruz, CA) in 4⬚C. Finally, after washing three times, cells were incubated with FITC-conjugated anti-goat secondary antibody (Sigma) as well as Hoechst 33,342 for 2 h at room temperature. After the cells were stained, cellular localization of fluorescence was examined by fluorescence or confocal microscopy.

Statistical Analysis Immunohistochemical Staining for Smad The expression and nuclear localization of endogenous Smads in the presence or absence of ufCB particle were studied by immunohistochemical staining. For this purpose, fibroblasts (15,000 cells/

All data are expressed as mean ⫾ SEM. Statistical analyses of results were performed using the Student t test and by analyzing variances using ANOVA, with correction by Bonferroni test. P value ⬍ 0.05 was taken as significant for the t test.


115

Figure 5. Binding of fibronectin and TGF-␤1 by ufCB particles. After culturing fibroblasts in monolayer for 2 d, media were harvested and incubated with ufCB (0–100 ␮g/ml) and CB (0–3 mg/ml) for 24 h at 4⬚C. After centrifuging (2,000 rpm, 5 min), supernatant media were collected for measuring TGF-␤1 (left vertical axes and diamonds) and fibronectin (right vertical axes and squares). (A ) ufCB (0–100 ␮g/ml, 0–2.54 ⫻ 10⫺2 M2 of surface area). (B ) CB (0–3 mg/ml, 0–2.54 ⫻ 10⫺2 M2 of surface area). Vertical axes: TGF-␤1/fibronectin concentration; horizontal axes: concentration and corresponding surface area of ufCB and CB. Data are mean ⫾ SEM for one representative experiment done in triplicate. Repeat experiments (n ⫽ 3) yielded similar results, although the absolute concentrations produced varied (*P ⬍ 0.05).

Results Effect of ufCB and CB Particles on Fibroblast-Mediated Collagen Gel Contraction Control gels and CB (100 ␮g/ml)-exposed gels contracted progressively over the period of observation. After 5 d, the gels were 32.30 ⫾ 0.58% and 37.89 ⫾ 1.01% of their initial area, respectively (Figure 1). In contrast, gels exposed to ufCB (100 ␮g/m) were 97.17 ⫾ 1.24% of their original size after 5 d. The effect of ufCB was both concentration- and time-dependent (Figure 2). To determine if inhibition of ufCB on fibroblast-mediated collagen gel contraction resulted from the cytotoxicity of the particles, cell viability and DNA content corresponding to cell number in the gels were measured. Viable cell number (determined by MTT assay and LDH assay) and DNA content in the gels decreased as a function of time, especially from Day 3 to Day 5. However, there were no significant differences in viable

cell number, nor in DNA content between control and ufCB-treated groups (data not shown). Furthermore, ufCB did not change the morphology of the fibroblasts in the gels (Figure 3) or induce degradation of collagen gels, as assessed by hydroxyproline assay (data not shown). To investigate the possibility of a reactive oxygen species-mediated mechanism for ufCB inhibition of fibroblastmediated gel contraction, antioxidants (NAC, catalase, mannitol, and deferoxamine mesylate) were added to the culture media for ufCB-exposed gels. None of the antioxidants were effective in restoring the gel contraction (data not shown). Effect of ufCB on the Release of TGF-␤1 and Fibronectin To determine the effect of ufCB on TGF-␤1 and fibronectin release, both TGF-␤1 and fibronectin were quantified by ELISA. The level of TGF-␤1 in conditioned media as well

116


Figure 6. Effect of the ufCB or CB depletion on enhancing activity for fibroblast-mediated collagen gel contraction contained in fibroblast-conditioned media. Fibroblasts were cultured in gels with six different kinds of media. Gel area was measured daily for 5 d. SF-DMEM: cultured gels with SF-DMEM. Fibroblast-conditioned media (CM): media from monolayer culture of fibroblasts for 3 d. (Fibroblast ⫹ ufCB) CM: media from monolayer culture of fibroblasts incubated with ufCB (100 ␮g/ml) for 3 d. Fibroblast CM absorbed with ufCB: media from centrifuged supernatant of mixture of monolayer cultured control media with ufCB (100 ␮g/ml) after sitting 24 h at 4⬚C. Fibroblast CM absorbed in CB: media from centrifuged supernatant of mixture of monolayer cultured control media with CB (100 ␮g/ml) after sitting 24 h at 4⬚C. (Fibroblast ⫹ CB) CM: media from monolayer culture of fibroblasts incubated with CB (100 ␮g/ml) for 3 d. Vertical axis: gel area expressed as % of initial area; horizontal axis: conditions. In all cases, particles were removed from media by centrifugation before the contraction experiments. Data are mean ⫾ SEM for three separate experiments each performed in triplicate. *P ⬍ 0.05, **P ⬍ 0.01.

as solubilized gels was significantly and concentrationdependently decreased in gels containing 25–100 ␮g/ml of ufCB, but TGF-␤ was significantly increased in gels containing 1␮g/ml of ufCB (Figure 4A). Fibronectin in conditioned media and solubilized gels was also significantly and concentration-dependently decreased in gels containing 50– 100 ␮g/ml of ufCB. In addition, fibronectin was significantly increased in gels containing 1 ␮g/ml of ufCB (Figure 4B). Concentrations of both TGF-␤ and fibronectin in media of fibroblasts maintained in routine dish culture were also reduced by ufCB, whereas CB (100 ␮g/ml) had no effect (data not shown). ufCB Particles Bind to TGF-␤1 and Fibronectin in Culture Media The possibility that ufCB affected gel contraction by binding to TGF-␤ and/or fibronectin was evaluated. After 2 d, conditioned media of HFL-1 cells cultured in monolayer were harvested and exposed to CB particles and ufCB particles for 24 h. To be able to compare ufCB and CB particles over a range of both concentrations and surface areas, these

Figure 7. Effect of exogenous fibronectin on ufCB-exposed gel contraction. Fibroblasts were cast into collagen gels with or without ufCB (100 ␮g/ml) as indicated and cultured in SF-DMEM with or without plasma type fibronectin (50 ␮g/ml). The gel area was measured on Day 2. (*P ⫽ 0.0017, **P ⬍ 0.0001, by Bonferroni procedure). Vertical axis: gel area expressed as % of initial area; horizontal axis: conditions. Data are mean ⫾ SEM for three separate experiments, each performed in triplicate.

absorption experiments were performed with a range of particle concentrations. ufCB particles were much more effective at absorbing and therefore reducing the concentration of TGF-␤1 in conditioned medium than were CB particles. Similar results were found in fibronectin (Figure 5B). Interestingly, ultrafine particles were more effective in absorbing TGF-␤ and fibronectin than were CB particles whether the particles were compared on a mass basis or when compared based on particles’ surface area. Effect of ufCB or CB Particles on Fibroblast-Mediated Collagen Gel Contraction Stimulated by Conditioned Media from Monolayer Culture To further evaluate the effect of depletion of TGF-␤1 and fibronectin by ufCB on collagen gel contraction, fibroblasts in collagen gels were cultured with conditioned media “depleted” by the addition of 100 ␮g/ml of CB or ufCB particles, which were then removed by centrifugation (Figure 6). Gels cultured in fibroblast-conditioned media from monolayer culture were able to contract rapidly during the first 48 h of incubation, reaching 45.73 ⫾ 3.14% of their original area, and further contracted to 27.55 ⫾ 3.61% by Day 5. Gels cultured in fibroblast ⫹ CB–conditioned media or in fibroblast-conditioned media absorbed in CB also contracted to about the same size as control gels over 5 d. Gels cultured in fibroblasts with ufCB-conditioned media also contracted, but significantly less so. Gels cultured in fibroblast-conditioned media absorbed by ufCB particles contracted to the same degree as gels in SF-DMEM, indicating that the ufCB could deplete the contraction-enhancing activity contained in conditioned media. Effect of Exogenous Fibronectin on ufCB-Exposed, Fibroblast-Mediated Gel Contraction To evaluate the role of fibronectin on gel contraction, 50 ␮g of exogenous plasma-type fibronectin was added to the


117

Figure 8. Fibronectin immunolocalization in gels exposed to ufCB particles. Fibroblasts in collagen gels were incubated with or without ufCB particles. (A ) ufCB particles in the gel (⫻400) on Day 0. (B ) ufCB particles in the gel (⫻400) on Day 5. (C ) Fibronectin staining in ufCB contained gel on Day 0. (D ) Fibronectin staining in gels containing ufCB on Day 5. (E ) Fibronectin staining in control gel on Day 0. (F ) Fibronectin staining in control gel on Day 5.

collagen gels. In control conditions, gels with added fibronectin contracted significantly more than gels without added fibronectin. With ufCB exposure, gels with added fibronectin also contracted significantly more than gels without added fibronectin (Figure 7). Fibronectin Staining of the ufCB-Exposed Gels To confirm the binding of fibronectin to ufCB particles, fibronectin was localized in the control gels and in those exposed to ufCB (Figure 8). The carbon particles were readily apparent under light microscopy (Figure 8A), and remained intact for 5 d (Figure 8B). Fibronectin was absent at Day 0 (Figures 8C and 8E), but accumulated with time. In control cultures, fibronectin was present in the cells and matrix after 5 d (Figure 8F). In contrast, in gels containing ufCB, fibronectin was localized on the CB particles after 5 d (Figure 8D). Effect of ufCB on TGF-␤1 and Fibronectin mRNA Expression To determine if the reduction in TGF-␤1 and fibronectin might also have resulted from a reduction in gene expression, the level of expression of TGF-␤1 and fibronectin mRNA was assessed. TGF-␤1 and fibronectin mRNA levels were significantly decreased in fibroblasts from gels containing 100 ␮g/ml of ufCB, compared with control gels or

those with 10 ␮g/ml of ufCB (Figures 9A and 9B). The decrease in fibronectin mRNA was observed for both the EIIIA and EIIIB splice variants. Smads Immunocytochemical Staining of the ufCB-Exposed Cells To determine if the reduced mRNA expression might be due to binding of TGF-␤ and reduced TGF-␤ signaling, the effect of ufCB particles on the immunohistochemical staining for Smad 2, 3, and 4 was performed in fibroblasts in monolayer cultures, incubated with or without ufCB particles. Relatively high levels of endogenous Smad 2, 3, and 4 were detected in control fibroblasts (Figures 10A, 10B, and 10C). Smad 3 immunofluorescence was localized primarily in the nucleus, with a very low level of cytoplasmic accumulation. Upon exposure with ufCB particle (Figures 10D, 10E, and 10F), however, the localization of Smad3associated fluorescence into the nucleus was markedly altered (Figure 10B versus Figure 10E). Examination of endogenous Smad 2 and 4 localization showed a reduction in nuclear localization after ufCB treatment, although the effect was less marked than with Smad 3 (Figure 10A versus Figure 10C, and Figure 10D versus Figure 10F).

Discussion Using ufCB as a surrogate for the ultrafine component of PM10, the current study demonstrates that ufCB particles

118


Figure 9. Effect of ufCB on the mRNA expression of TGF-␤1/ fibronectin fragments (EIIIA/EIIIB) in fibroblasts from collagen gels. Fibroblasts were cultured in collagen gels with or without ufCB for 3 d. The cell pellets were obtained for mRNA measurement by centrifuging the cells (2,000 rpm, 3 min) after dissolving the gels with collagenase (0.25 mg/ml). The mRNA levels of TGF-␤1 (A ) and fibronectin fragments (EIIIA/EIIIB; B ) were quantified by real-time PCR. In B, striped bars indicate control, open bars ufCB 10 ␮g/ml, and filled bars ufCB 100 ␮g/ml. Vertical axis: mRNA amount, expressed as ratio of 105XGAPDH unit. Horizontal axis: conditions. Data are mean ⫾ SEM for three separate experiments, each performed in triplicate. *P ⬍ 0.05; **P ⬍ 0.01.

inhibit the ability of fibroblasts to contract three-dimensional collagen gels. This effect is not likely due to a cytotoxic effect from the ufCB particles on the fibroblasts, because DNA content in the cells, LDH content in the media, formazan crystal absorbance by the cells (MTT assay), and cell morphology do not change and exogenous fibronectin can reverse the effect. In contrast to ultrafine particles, fine CB particles had no measurable effect. Whereas previous studies on the effects of ufCB have identified oxidant-mediated mechanisms (14, 18, 20, 22, 23), an oxidant-independent mechanism for ufCB attenuation of collagen gel contraction is supported by the lack of inhibitory effect of antioxidants. ufCB significantly inhibited the release of both fibronectin and TGF-␤ into the media in which collagen gels were floated. This effect appears to be partially mediated by the binding of both fibronectin and TGF-␤ to the surfaces of the ufCB particles. A decrease in the production of

fibronectin is also suggested because a decrease in fibronectin mRNA was also observed following exposure to ufCB. A decrease in TGF-␤–mediated signaling is supported by a decrease in intranuclear localization of Smad proteins. Because TGF-␤ stimulates fibronectin mRNA and protein production, sequestration of TGF-␤ and reduced TGF-␤ signaling could account for the decreased fibronectin production. The culturing of fibroblasts in three-dimensional collagen gels is a model system often used to evaluate several aspects of wound repair and tissue remodeling (34). Fibroblasts cultured in native collagen gels attach to the collagen fibers through integrin-mediated mechanisms and are distinctly different from fibroblasts cultured in routine “dish” culture. Three-dimensional cultures are generally believed to behave in a more “tissue-like” manner (34, 35). Fibroblasts are capable of exerting a mechanical tension and, as a consequence, can cause contraction of three-dimensional gels maintained in floating culture. This is believed to mimic the contraction that takes place during scar formation or the development of fibrosis. Integrins and a number of cytokines, including both TGF-␤ and fibronectin, have been suggested to play a role in augmenting this process (23, 24, 26, 27, 36). Both acute and chronic exposures to particulates have been associated with increased mortality and increased morbidity from a number of causes, including chronic lung disease (1–4). Particles in the ambient environment which are derived from many sources vary in size and have complex compositions, including ammonium salts, carbon, metals, and a variety of crystal and adsorbed biologic materials (6, 37). Ultrafine particles, such as diesel soot and other combustion-derived particles, have been suggested to be a major mediator of adverse health effects. To study the effect of particle size, ultrafine particles have been compared with fine particles of the same composition (i.e., ultrafine and fine TiO2). These simpler chemical compounds have been used to demonstrate size-related toxicity with ufCB particles, inducing more inflammation than CB when instilled into the airspaces of rats (14, 22). ufCB also has been shown to increase the resting cytosolic calcium concentration of a human monocytic cell line MonoMac 6 (MM6). This effect was not found with the same dose of larger carbon black particles (38). In the current study, only ufCB, not CB, inhibited collagen gel contraction. The inhibitory effect of ufCB on gel contraction was not due to the cytotoxicity of the particles to the gels, as demonstrated by three different assays of cell viability and morphologic observation. The mechanism by which ultrafine particles inhibit gel contraction appears to be due to absorption of TGF-␤ and fibronectin to the particle surface. Interestingly, this property appears to depend on more than just nominal particle surface area, as addition of CB particles at sufficiently high concentrations to have similar expected surface areas did not result in equivalent absorption of proteins. An effect of surface area is still theoretically possible, as the estimated surface areas depend on the particle shape. A change in the geometry of the particles in the culture conditions may affect the particle surface area calculations. Other differences between the ultrafine and carbon black


119

Figure 10. Smads staining of fibroblasts in the monolayer culture with or without ufCB particles. Fibroblasts with or without ufCB 100 ␮g/ml in SF-DMEM were cultured for 3 d, after which immunohistochemistry for anti-Smad 2, Smad 3, or Smad 4 was performed and examined by confocal microscopy. (A, B, and C ) Smad 2, 3, 4 localization in control culture condition. (D, E, and F ) Smad 2, 3, 4 localization in ufCB-exposed cultures.

particles could also play a role. Subtle differences in chemistry, for example, could affect protein absorption. Similarly, differences in particle size or shape could alter particle cell surface charge, which in turn could affect protein absorption. Thus, although the mechanisms are incompletely defined, ultrafine particles appear to be particularly potent in absorbing proteins, hence accounting for their ability to interfere with fibroblast-mediated contraction of threedimensional collagen gels. Oxidative stress is a prominent mechanism by which a variety of toxicants mediate their effects. In this regard, ultrafine particles are potent inducers of intracellular oxidant generation (14, 20, 22, 38). Particle-generated oxidants are believed to mediate increased production of IL-8, IL-6, and TNF-␣, and hence can account for ultrafine particle– driven inflammation (15–19). In the current study, oxidants do not appear to play a key role in ufCB inhibition of collagen gel contraction, because none of the antioxidants tested exerted any inhibitory effect on ufCB-mediated toxicity. This was in marked contrast to an antioxidant blockade of cigarette smoke inhibition of collagen gel contraction (personal observation). Rather, the inhibitory effect of ultrafine particles appears to be due to absorption of fibronectin and TGF-␤, and possibly other mediators, from the pericellular milieu. TGF-␤ is believed to play a particularly important role in wound repair, stimulating the production of fibronectin together with a variety of extracellular matrix

macromolecules (39–42). TGF-␤ also can modulate cellular expression of integrins (43, 44). In the current study, low concentrations of ufCB, which were inadequate to absorb TGF-␤ and did not block collagen gel contraction, had a slight stimulatory effect on fibronectin and TGF-␤ release. Low concentrations of ultrafines, therefore, may have a profibrotic effect. This observation is consistent with profibrotic responses observed with particulates in other studies (45, 46). The mechanism for stimulation of profibrotic mediator release by low concentrations of ultrafines remains to be determined. At higher concentrations, however, both fibronectin and TGF-␤ were bound to the ufCB particles. Two lines of evidence suggest direct absorption. First, fibronectin was localized on the surface of particles by immunohistochemistry. Second, media containing both fibronectin and TGF-␤ could be depleted of fibronectin and TGF-␤ by incubation with ufCB particles followed by centrifugation. These results indicate that ufCB particles may contribute to the development of fibrosis at low concentrations, whereas at high concentrations, ufCB may inhibit tissue repair processes by absorbing extracellular matrices and growth factors. TGF-␤ is produced as a latent precursor, and the majority of TGF-␤ present in cell cultures is found in the latent form (47, 48). In the current study, all of the detectable TGF-␤ was present in the latent form. This does not exclude, however, the possibility that TGF-␤ is functioning, because

120


an autocrine or paracrine mediator in the current assay system as active TGF-␤ may have been present below the limit of detectability of our immunoassay system. Alternatively, conformational activation of latent TGF-␤ has been described (49, 50). Such conformational changes are also undetectable by conventional immunoassays for active TGF-␤. Several lines of evidence suggest that ufCB is reducing TGF-␤ activity. TGF-␤ signals by binding to a heterodimeric receptor which leads to the phosphorylation and activation of two signaling proteins, Smad 2 and Smad 3. These proteins then form dimers with Smad 4 and translocate to the nucleus. Our control cultures demonstrated spontaneous nuclear localization of Smads 2, 3, and 4, suggesting constitutive activation of the Smad signaling pathway (51–54). Following incubation with ufCB, a detectable reduction in nuclear localization of Smads was observed, consistent with reduced signaling through Smad proteins. Although other cytokines, for example, the activins, can also signal through the Smads, decreased nuclear localization of Smad 2, 3, and 4 is consistent with decreased TGF-␤ signaling (55, 56). Decreased TGF-␤ bioactivity would account for the reduction in fibronectin mRNA expression noted in the current study. Decreased fibronectin production, in addition to absorption to ufCB particle surfaces, could account for the decreased fibronectin release and localization in the extracellular matrix of the three-dimensional collagen gels. It is likely that reduced TGF-␤ signaling could, either directly or indirectly by decreasing extracellular fibronectin, result in reduced contraction of three-dimensional collagen gels. In addition, ufCB may also absorb other growth factors and extracellular matrices, which could subsequently contribute to the inhibition of collagen gel contraction by fibroblasts. In summary, the current study demonstrates that ultrafine particulate matter can bind biologically active molecules and, as a result, alter the behavior of fibroblasts. Whether such effects would be observed in vivo in people inhaling urban PM10 rich in ultrafines would require further studies. However, by removing fibronectin and TGF-␤ from a milieu in which injury had occurred, such an effect could contribute to ineffective tissue repair and possibly the development of pulmonary emphysema. It is also possible to envisage that components of PM10 may interact with each other. Thus, in a PM10 sample, transition metals could initiate injury, via oxidant stress, whereas ultrafine particles might exacerbate the problem by binding active molecules. Alternatively, it is possible that fibronectin, TGF-␤, and potentially other biologically active moieties absorbed to particles could serve as a reservoir for biologic activity. The localization of such molecules could be assessed immunohistochemically on specimens taken from lung tissues from either patients or animals exposed to particulates. These specimens, of course, are likely to contain a complex mix of particulates resulting from complex exposures. It may be difficult, therefore, to definitively attribute binding to ultrafines. Nevertheless, the present study provides an in vitro mechanism by which ultrafine particles could contribute to altered repair mechanisms and, therefore, to the

development of chronic lung disease, which can be tested in vivo. Acknowledgments: CB and ufCB particles were kindly provided by Dr. Peter Gilmour (Respiratory Medicine Unit, ELEGI/Colt Laboratory, Edinburgh University, Edinburgh, UK). K.D. is the BLF Transco Fellow in Air Pollution and Respiratory Health. The authors gratefully acknowledge the secretarial support of Ms. Lillian Richards and the editorial assistance of Ms. Mary Tourek. This study was supported by the Larson Endowment, University of Nebraska Medical Center, The Colt Foundation/Medical Research Council.

References 1. Abbey, D. E., R. J. Burchette, S. F. Knutsen, W. F. McDonnell, M. D. Lebowitz, and P. L. Enright. 1998. Long-term particulate and other air pollutants and lung function in nonsmokers. Am. J. Respir. Crit. Care Med. 158:289–298. 2. Dockery, D. W., C. A. Pope III, X. Xu, J. D. Spengler, J. H. Ware, M. E. Fay, B. G. Ferris, Jr., and F. E. Speizer. 1993. An association between air pollution and mortality in six US cities. N. Engl. J. Med. 329:1753–1759. 3. Pope, C. A. III, D. W. Dockery, J. D. Spengler, and M. E. Raizenne. 1991. Respiratory health and PM10 pollution. A daily time series analysis. Am. Rev. Respir. Dis. 144:668–674. 4. Pope, C. A. III, M. J. Thun, M. M. Namboodiri, D. W. Dockery, J. S. Evans, F. E. Speizer, and C. W. Heath, Jr. 1995. Particulate air pollution as a predictor of mortality in a prospective study of US adults. Am. J. Respir. Crit. Care Med. 151:669–674. 5. Committee of the Environmental and Occupational Health Assembly of the American Thoracic Society. 1996. Health effects of outdoor air pollution. Am. J. Respir. Cell Mol. Biol. 153:3–50. 6. Seaton, A., W. MacNee, K. Donaldson, and D. Godden. 1995. Particulate air pollution and acute health effects. Lancet 345:176–178. 7. Abbey, D. E., P. K. Mills, F. F. Petersen, and W. L. Beeson. 1991. Longterm ambient concentrations of total suspended particulates and oxidants as related to incidence of chronic disease in California Seventh-Day Adventists. Environ. Health Perspect. 94:43–50. 8. Jammes, Y., S. Delpierre, M. J. Delvolgo, C. Humbert-Tena, and H. Burnet. 1998. Long-term exposure of adults to outdoor air pollution is associated with increased airway obstruction and higher prevalence of bronchial hyperresponsiveness. Arch. Environ. Health 53:372–377. 9. Donaldson, K., X. Y. Li, and W. MacNee. 1998. Untrafine (nanometre) particle mediated lung injury. J. Aerosol Sci. 29:553–560. 10. Ferin, J., G. Oberdorster, and D. P. Penney. 1992. Pulmonary retention of ultrafine and fine particles in rats. Am. J. Respir. Cell Mol. Biol. 6:535–542. 11. Oberdorster, G. 1996. Significance of particle parameters in the evaluation of exposure-dose–response relationships of inhaled particles. Inhal. Toxicol. 8:73–89. 12. Oberdorster, G., J. Ferin, and B. E. Lehnert. 1994. Correlation between particle size, in vivo particle persistence, and lung injury. Environ. Health Perspect. 102:173–179. 13. Ferin, J., G. Oberdorster, S. C. Soderholm, and R. Gelein. 1991. Pulmonary tissue access of ultrafine particles. J. Aerosol Med. 4:57–68. 14. Li, X. Y., P. S. Gilmour, K. Donaldson, and W. MacNee. 1996. Free radical activity and pro-inflammatory effects of particulate air pollution (PM10) in vivo and in vitro. Thorax 51:1216–1222. 15. Bayram, H., J. L. Devalia, R. J. Sapsford, T. Ohtoshi, Y. Miyabara, M. Sagai, and R. J. Davies. 1998. The effect of diesel exhaust particles on cell function and release of inflammatory mediators from human bronchial epithelial cells in vitro. Am. J. Respir. Cell Mol. Biol. 18:441–448. 16. Driscoll, K. E., D. G. Hassenbein, J. M. Carter, S. L. Kunkel, T. R. Quinlan, and B. T. Mossman. 1995. TNF alpha and increased chemokine expression in rat lung after particle exposure. Toxicol. Lett. 82–83:483–489. 17. Ohtoshi, T., H. Takizawa, H. Okazaki, S. Kawasaki, N. Takeuchi, K. Ohta, and K. Ito. 1998. Diesel exhaust particles stimulate human airway epithelial cells to produce cytokines relevant to airway inflammation in vitro. J. Allergy Clin. Immunol. 101:778–785. 18. Quay, J. L., W. Reed, J. Samet, and R. B. Devlin. 1998. Air pollution particles induce IL-6 gene expression in human airway epithelial cells via NF-KB activation. Am. J. Respir. Cell Mol. Biol. 19:98–106. 19. Drumm, K., H. Schindler, R. Buhl, E. Ku¨stner, R. Smolarski, and K. Kienast. 1999. Indoor air pollutants stimulate interleukin-8-specific mRNA expression and protein secretion of alveolar macrophages. Lung 177:9–19. 20. Stone, V., J. Shaw, D. M. Brown, W. MacNee, S. P. Faux, and K. Donaldson. 1998. The role of oxidative stress in the prolonged inhibitory effect of ultrafine carbon black on epithelial cell function. Toxicol. In Vitro 12:649– 659. 21. Carter, J. D., A. J. Ghio, J. M. Samet, and R. B. Devlin. 1997. Cytokine production by human airway epithelial cells after exposure to an air pollution particle is metal-dependent. Toxicol. Appl. Pharmacol. 146:180– 188. 22. Brown, D. M., V. Stone, and P. Findlay. 1999. Increased inflammation and


23. 24. 25. 26.

27.

28. 29. 30. 31. 32.

33. 34. 35. 36. 37. 38.

intracelullar calcium caused by ultrafine carbon black is independent of transition metals or other soluble components. Occup. Environ. Med. 57:85–91. Asaga, H., S. Kikuchi, and K. Yoshizato. 1991. Collagen gel contraction by fibroblasts requires cellular fibronectin but not plasma fibronectin. Exp. Cell Res. 193:167–174. Gillery, P., F. X. Maquart, and J. P. Borel. 1986. Fibronectin dependence of the contraction of collagen lattices by human skin fibroblasts. Exp. Cell Res. 167:29–37. Houlgate, P. R., K. S. Dhingra, S. J. Nash, and W. H. Evans. 1989. Determination of formaldehyde and acetaldehyde in mainstream cigarette smoke by high-performance liquid chromatography. Analyst 114:355–360. Reed, M. J., R. B. Vernon, I. B. Abrass, and E. H. Sage. 1994. TGF-␤1 induces the expression of type I collagen and SPARC, and enhances contraction of collagen gels, by fibroblasts from young and aged donors. J. Cell. Physiol. 158:169–179. Mio, T., Y. Adachi, D. J. Romberger, J. R. Spurzem, R. F. Ertl, S. Carnevali, and S. I. Rennard. 1995. Human bronchial epithelial cells modulate collagen gel contraction by fibroblasts. Am. J. Respir. Crit. Care Med. 151:A561. (Abstr.) Mio, T., Y. Adachi, D. J. Romberger, R. F. Ertl, and S. I. Rennard. 1996. Regulation of fibroblast proliferation in three dimensional collagen gel matrix. In Vitro Cell. Dev. Biol. 32:427–433. Bergman, I., and R. Loxley. 1963. Two improved and simplified methods for the spectrophotometric determination of hydroxyproline. Anal. Chem. 35:1961–1965. Edwards, C. A., and W. D. O’Brien. 1980. Modified assay for determination of hydroxyproline in a tissue hydrolyzate. Clin. Chim. Acta 104:161–167. Supino, R. 1995. MTT assays. Methods Mol. Biol. 43:137–149. Rennard, S. I., R. L. Church, D. H. Rohrbach, D. E. Shupp, S. Abe, A. T. Hewitt, J. C. Murray, and G. R. Martin. 1981. Localization of the human fibronectin (FN) gene on chromosome 8 by a specific enzyme immunoassay. Biochem. Genet. 19:551–566. Bustin, S. A. 2000. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J. Mol. Endocrinol. 25: 169–193. Bell, E., B. Ivarsson, and C. Merrill. 1979. Production of a tissue-like structure by contraction of collagen lattices by human fibroblasts of different proliferative potential in vitro. Proc. Natl. Acad. Sci. USA 76:1274–1278. Mauch, C., A. Hatamochi, K. Scharffetter, and T. Krieg. 1988. Regulation of collagen synthesis in fibroblasts within a three-dimensional collagen gel. Exp. Cell Res. 178:493–503. Gullberg, D., A. Tingstrom, A.-C. Thuresson, L. Olsson, L. Terracio, T. K. Borg, and K. Rubin. 1990. ␤1 integrin-mediated collagen gel contraction is stimulated by PDGF. Exp. Cell Res. 186:264–272. Donaldson, K., V. Stone, A. Clouter, L. Renwick, and W. MacNee. 2001. Ultrafine particles. Occup. Environ. Med. 58:211–216. Stone, V., M. Tuinman, J. E. Vamvakopoulos, J. Shaw, D. Brown, S. Petterson, S. P. Faux, P. Borm, W. MacNee, F. Michaelangeli, and K. Donaldson. 2000. Increased calcium influx in a monocytic cell line on exposure to ultrafine carbon black. Eur. Respir. J. 15:297–303.

121

39. Montesano, R., and L. Orci. 1988. Transforming growth factor-␤ stimulates collagen-matrix contraction by fibroblasts: implication for wound healing. Proc. Natl. Acad. Sci. USA 85:4894–4897. 40. Tipton, D. A., and M. K. Dabbous. 1998. Autocrine transforming growth factor beta stimulation of extracellular matrix production by fibroblasts from fibrotic human gingiva. J. Periodontol. 69:609–619. 41. Hocevar, B. A., T. L. Brown, and P. H. Howe. 1999. TGF-beta induces fibronectin synthesis through a c-Jun N-terminal kinase-dependent, Smad4-independent pathway. EMBO J. 18:1345–1356. 42. Shah, M., D. Revis, S. Herrick, R. Baillie, S. Thorgeirson, M. Ferguson, and A. Roberts. 1999. Role of elevated plasma transforming growth factorbeta1 levels in wound healing. Am. J. Pathol. 154:1115–1124. 43. Enenstein, J., N. S. Waleh, and R. H. Kramer. 1992. Basic FGF and TGFbeta differentially modulate integrin expression of human microvascular endothelial cells. Exp. Cell Res. 203:499–503. 44. Arora, P. D., N. Narani, and C. A. McCulloch. 1999. The compliance of collagen gels regulates transforming growth factor-beta induction of alpha-smooth muscle actin in fibroblasts. Am. J. Pathol. 154:871–882. 45. Salvi, S., A. Blomberg, B. Rudell, F. Kelly, T. Sandstrom, S. T. Holgate, and A. Frew. 1999. Acute inflammatory responses in the airways and peripheral blood after short-term exposure to diesel exhaust in healthy human volunteers. Am. J. Respir. Crit. Care Med. 159:702–709. 46. Workshop, I. L. S. I. 2000. The relevance of the rat lung response to particle overload for human risk assessment. Inhal. Toxicol. 12:1–2. 47. Miyazono, K., K. Yuki, F. Takaku, C. Wernstedt, T. Kanzaki, A. Olofsson, U. Hellman, and C.-H. Heldin. 1990. Latent forms of TGF-beta: structure and biology. Ann. N. Y. Acad. Sci. 593:51–58. 48. Khalil, N. 1999. TGF-beta: from latent to active. Microbes Infect. 1:1255– 1263. 49. Lawrence, D. A. 1996. Transforming growth factor-beta: a general review. Eur. Cytokine Netw. 7:363–374. 50. Brown, P. D., L. M. Wakefield, A. D. Levinson, and M. B. Sporn. 1990. Physicochemical activation of recombinant latent transforming growth factor-beta’s 1, 2, and 3. Growth Factors 3:35–43. 51. Massague, J., and D. Wotton. 2000. Transcriptional control by the TGFbeta/Smad signaling system. EMBO J. 19:1745–1754. 52. Piek, E., C. H. Heldin, and P. Ten Dijke. 1999. Specificity, diversity, and regulation in TGF-beta superfamily signaling. FASEB J. 13:2105–2124. 53. Massague, J. 1998. TGF-beta signal transduction. Annu. Rev. Biochem. 67:753–791. 54. Derynck, R., Y. Zhang, and X. H. Feng. 1998. Smads: transcriptional activators of TGF-beta responses. Cell 95:737–740. 55. Nakao, A., T. Imamura, S. Souchelnytskyi, M. Kawabata, A. Ishisaki, E. Oeda, K. Tamaki, J. Hanai, C. H. Heldin, K. Miyazono, and P. ten Dijke. 1997. TGF-beta receptor-mediated signalling through Smad2, Smad3 and Smad4. EMBO J. 16:5353–5362. 56. Eppert, K., S. W. Scherer, H. Ozcelik, R. Pirone, P. Hoodless, H. Kim, L. C. Tsui, B. Bapat, S. Gallinger, I. L. Andrulis, G. H. Thomsen, J. L. Wrana, and L. Attisano. 1996. MADR2 maps to 18q21 and encodes a TGFbetaregulated MAD-related protein that is functionally mutated in colorectal carcinoma. Cell 86:543–552.