bleomycin-induced pulmonary fibrosis hamsters€¦ · in hamsters is associated with collagen...

Bleomycin-induced Pulmonary Fibrosis in Hamsters

AN ALVEOLAR MACROPHAGEPRODUCTINCREASES FIBROBLAST

PROSTAGLANDINE2 AND CYCLIC ADENOSINE MONOPHOSPHATE

AND SUPPRESSESFIBROBLAST PROLIFERATION

AND COLLAGENPRODUCTION

JOAN G. CLARK, KATHERINE M. KOSTAL, and BEVERLY A. MARINO, PulmonaryDivision, Department of Medicine, Washington UniversitySchool of Medicine, St. Louis, Missouri 63110

A B S T R A C T Bleomycin-induced pulmonary fibrosisin hamsters is associated with collagen accumulationthat results from increased lung collagen synthesis rates.However, 1-2 wk after intratracheal instillation ofbleomycin, lung collagen synthesis rates decline towardcontrol values. To evaluate the potential role of thebronchoalveolar macrophage in regulating lung col-lagen production, we studied the effects of macrophagesfrom normal and bleomycin-treated hamsters upon fi-broblasts in vitro. Weobserved: (a) Medium from mac-rophage cultures decreased fibroblast [3H]thymidineincorporation and nondialyzable [3H]hydroxyprolineproduction in a dose-dependent manner. Fibroblast cellcounts were decreased in exposed cultures, and fibro-blast viability was unchanged. Procollagen prolyl hy-droxylation and prolyl-transfer RNA-specific activitywere not altered by macrophage medium; this indicatesthat [3H]hydroxyproline reflects collagen productionrate under the experimental conditions. (b) The sup-pressive effect of macrophage medium was selectivefor collagen since collagen production decreased morethan noncollagen protein production. (c) Medium frombleomycin-treated hamster macrophages suppressed fi-broblast proliferation and collagen production to agreater degree than medium from normal hamstermacrophages. (d) Fibroblast suppression by macro-phage medium was associated with increased fibroblastendogenous prostaglandin E2 production and intracel-

A preliminary report of these studies was presented at theNational Meeting of the American Thoracic Society, LosAngeles, CA, May 1982, and appeared in 1982, Am. Rev.Respir. Dis. 125:217.

Address all reprint requests to Dr. Clark.Received for publication 28 February 1983 and in revised

form 24 August 1983.

lular cyclic AMP(cAMP). (e) Incubation of fibroblastswith indomethacin before exposure completely inhib-ited prostaglandin E2 production and increases incAMP, and eliminated suppression of fibroblast pro-liferation and collagen production. The macrophage-derived suppressive factor has an apparent molecularweight of 20,000-30,000 and is heat stable. It does notappear to be species restricted since both hamster andhuman lung fibroblasts are similarly suppressed. It isat least in part preformed in macrophages obtained bylavage, but its production can also be stimulated invitro. Weconcluded that alveolar macrophages releasea product that stimulates endogenous fibroblast pros-taglandin E2 production and cAMP formation with re-sultant suppression of fibroblast proliferation and col-lagen production. Enhanced release of suppressive fac-tor by macrophages during a time when lung collagenproduction is declining in bleomycin-induced pul-monary fibrosis suggests that macrophages may limitcollagen accumulation in pulmonary fibrosis.

INTRODUCTIONDiffuse interstitial pulmonary fibrosis is a clinicallyheterogeneous condition caused by a variety of injuriesto the lung. Inflammatory cells are invariably present,and a characteristic feature of the disorder is histolog-ically increased and disorganized interstitial collagen(1-3). Because biochemical correlates have been dif-ficult to obtain in patients, studies of experimentallyinduced pulmonary fibrosis in animals have been usefulin delineating the pathogenesis of abnormal collagendeposition in the disorder.

In bleomycin-induced pulmonary fibrosis in hamsters,collagen content increases (4, 5) as the result of greatlyincreased collagen synthesis rates that occur promptly

2082 J. Clin. Invest. © The American Society for Clinical Investigation, Inc. * 0021-9738/83/12/2082/10 $1.00Volume 72 December 1983 2082-2091

after administration of the drug (6). However, collagensynthesis rates decline 1-2 wk later and are subse-quently restored to normal even though histologicalfibrosis persists and collagen content remains elevated.We recently described a factor released from normallung explant cultures that suppresses lung fibroblastproliferation and collagen synthesis by stimulating en-dogenous fibroblast prostaglandin (PG)'E2 productionand intracellular cyclic AMP(cAMP) formation (7). Inthe first week after bleomycin administration, fibroblast-suppressive activity is decreased, but at 2 wk increasesto greater than normal levels. At this time, lung collagensynthesis rate is declining and a mononuclear cell in-filtrate is present along with increased numbers ofbronchoalveolar lavage macrophages (4-6, 8). Theseobservations suggest that mononuclear phagocytesmight play a role in the regulation of lung fibroblastproliferation and/or collagen synthesis. In this report,we present studies of the effects of products from cul-tured bronchoalveolar macrophages (BAM) on lung fi-broblasts. The results suggest that bronchoalveolarmacrophages may participate in limiting lung collagenaccumulation in pulmonary fibrosis.

METHODSBleomycin treatment. 10-wk-old male Syrian hamsters

received a single intratracheal dose of 1 U of bleomycin(Blenoxane; Bristol Laboratories, Syracuse, NY) in 0.1 ml ofsterile saline (0.9 g/dl). Age-matched, uninjected hamstersserved as controls. Bleomycin-treated animals were studiedat 1-3 wk after instillation of the drug.

Conditioned medium from BAMcultures. BAMwereobtained by bronchoalveolar lavage of anesthetized hamsterswith an 18-gauge intratracheal catheter by using 10 ml of0.9 g/dl NaCl containing 3 mMNa2EDTA. The lavage fluidwas centrifuged at 250 g for 15 min at 25°C, and the cellpellet was suspended in Dulbecco's modified Eagle's medium(DMEM) and recentrifuged. Cells were suspended in DMEMat 5 X 106 cells/ml. The percentage of macrophages wasestimated by examination of cytocentrifuge preparationsstained with hematoxylin-eosin. 5 X 105 BAMwere platedin 1.0 ml of DMEMin 1.7-cm diameter multiple well cultureplates (Multiwell Tissue Culture Plate, Falcon Labware, Ox-nard, CA). After incubation for 1 h at 370C in 5% C02:95%air atmosphere, cell layers were rinsed once with DMEMtoremove nonadherent cells, and adherent cells were culturedin basal medium (DMEMsupplemented with 10% fetal calfserum, 200 U/ml penicillin, 200 Ag/ml stretomycin, and 2mMglutamine). In one experiment, the serum concentrationof basal medium was varied. Lavage cells were also culturedin suspension in polypropylene tubes. Conditioned medium(CM) was removed after 24 h and stored at -200C. Adherentcells were >95% viable as indicated by trypan blue exclusion,and >95% demonstrated specific esterase positivity. Before

' Abbreviations used in this paper: BAM, bronchoalveolarmacrophages; CM, conditioned medium; DMEM,Dulbecco'smodified Eagle's medium; [3H]dThd; tritiated thymidine; PG,prostaglandin.

addition to fibroblast cultures, CMwas dialyzed (12,000 molwt cutoff, Spectrum Medical Industries, Los Angeles, CA)against three changes of DMEM(1OOX sample volume) for36 h at 4VC and filter-sterilized (0.2 gm pore size).

To determine if protein synthesis was required for ap-pearance of activity in CM, adherent BAMwere cultured inbasal medium containing cycloheximide (100 Mtg/ml). As acontrol, cycloheximide (100 Mg/ml) was added to BAM-CMbefore dialysis against fresh medium. Parallel BAMcultureswere incubated with [3H]proline and the CMwas dialyzedexhaustively against tap water. Nondialyzable radioactivitywas measured as an indication of newly synthesized protein.In other BAMcultures, Sepharose 4B (10% wt/vol) was addedto the incubation medium to increase release of suppressiveactivity (7). BAM lysate was obtained from a BAMsampleby freezing (-70'C) and thawing (250C) the cells three times.

Fibroblast cultures. Human fetal lung fibroblasts (IMR-90, Institute for Medical Research, Camden, NJ) were platedat 3 X 104 cells/1.7-cm diameter well in 1 ml of basal mediumand incubated in a 5% C02:95% air atmosphere at 370C.Cells were in the tenth to fifteenth passage (1:4 split ratio).

Hamster lung fibroblasts were cultivated as previously de-scribed from 10-wk-old male Syrian hamsters (7). Hamsterlung fibroblasts obtained by this method required 2-3 wk toreach visual confluence, while IMR-90 fibroblasts plated atthe same density were confluent in less than 1 wk. Becauseof their slow growth, hamster lung fibroblasts were platedat high density and used in the first to third passage. Unlessotherwise specified, we used IMR-90 cells, which are referredto as fibroblasts.

When fibroblasts were nearly confluent (usually 5 d afterpassage), varying amounts of CMfrom cultured BAMwereadded with fresh basal medium to a total volume of 500 Mllwell. In one experiment, serum concentration in basal mediumwas reduced 2 d before addition of CM. Fibroblasts wereincubated in the presence of CM, and proliferation and col-lagen synthesis assayed as described below.

Fibroblast proliferation assays. After incubation of thefibroblasts with CM, 1 MCi tritiated thymidine ([3H]dThd)(25 Ci/mol; Amersham Corp., Arlington Heights, IL) in 100Ml of basal medium was added to each well, and the incubationwas continued for an additional 4 h. Incorporation of [3H]dThdwas measured as previously described (7). In some experi-ments, cell layer protein (9) or DNA (10) was measured asan index of cell number. Cell counts were performed witha hemocytometer. Cell viability was determined by trypanblue exclusion.

Fibroblast collagen production assay. After incubationof the fibroblasts with CM, 10 ,uCi [3H]proline (130 Ci/mol;Amersham Corp.) in 100 Ml of basal medium, containingascorbic acid and 0-aminopropionitrile to yield final con-centrations of 50 and 20 ,g/ml, respectively, was added toeach well, and the incubation continued for an additional 6h. The medium was removed and combined with a 500 MlHanks' balanced salt solution wash of the cell layer. The celllayer was scraped into 1 ml of 0.5 M acetic acid. After theaddition of protease inhibitors (10 mMN-ethylmaleimide,20 mMdisodium EDTA, and 0.3 mMphenylmethylsulfon-ylfluoride in final concentration), the medium and cell layerwere dialyzed separately for 24 h against running tap water,hydrolyzed in 6 N HCI at 110°C for 18 h, evaporated, anddissolved in 0.2 M sodium citrate buffer, pH 2.2.[3H]hydroxyproline and [3H]proline were separated chro-matographically as previously described (11). Since[3H]hydroxyproline in medium was proportional to total[3H]hydroxyproline, measurements were made on mediumalone unless otherwise indicated. In some experiments, me-

Collagen Production in Pulmonary Fibrosis 2083

dium 3H-proteins were digested with bacterial collagenase(type VI, Sigma Chemical Co., St. Louis, MO), which waspurified as described by Peterkovsky and Diegelmann (12).

Specific radioactivity of proline in fibroblast intracellularfree amino acid pool and amino acyl-transfer RNA(tRNA)pool were measured by a method described by Airhart et al.(13) and modified by Hildebran et al. (14). In these exper-iments, confluent fibroblasts were cultured in P-60 culturedishes (Falcon Labware) with 2 ml of CM(30% concentration)or basal medium containing 200 jiCi [3H]proline, ascorbicacid (50 ug/ml), f3-aminopropionitrile (20 mg/ml), and pro-line (0.2 mM) for 18 h. The following is a brief descriptionof the analysis. Medium was removed and analyzed for[3H]hydroxyproline and [3H]proline as described above. Therinsed cell layer was solubilized with cacodylate-bufferedsodium dodecyl sulfate. An aliquot of the solubilized celllayer was precipitated with trichloroacetic acid; the super-natant containing the intracellular free amino acid pool wasanalyzed as described below to determine intracellular freeproline-specific radioactivity. The remainder of the cell layerto be analyzed for prolyl-tRNA-specific radioactivity was de-proteinized by addition of freshly distilled phenol. Nucleicacids were precipitated from the aqueous phase by additionof ice-cold ethanol, and amino acids bound to tRNA werereleased by alkaline hydrolysis after which nucleic acids werereprecipitated by addition of HCL.

Specific radioactivity of proline in the intracellular freeamino acid pool and in the aminoacyl-tRNA pool was de-termined after reaction with ['4C]dansyl chloride and sepa-ration of dansyl derivatives by two-dimensional ascendingthin-layer chromatography. Dansyl proline (cold) was addedto the sample before chromatography to clearly identify theposition of proline on the chromatography plate. That spotwas cut from the plate and 3H and 4C counted in a liquidscintillation counter equipped with automatic quench com-pensation (Beckman LS8000, Beckman Instruments, Inc.,Fullerton, CA). Specific radioactivity of proline was calculatedby the formula:

proline specific activity

= 3H(dpm) dansyl chloride specific activity* K (1)14C(dpm)where K equals moles of dansyl group bound per moles ofproline, which equals 1. Because dansyl chloride specificactivity was not known with certainty, the values are pre-sented as arbitrary numbers for comparison of control andCM-exposed cultures.

Intracellular free proline concentration was measured insome experiments by a photochemical method described byTroll and Lindsley (15).

Other calculations included:

Procollagen prolyl hydroxylation (%)

= ([3H]hydroxyproline released by collagenase. total radioactivity released by collagenase) X 100. (2)

Noncollagen protein production

(noncollagen proline incorporation per cell)

= (total 3H-protein/,ug DNA)

-[([3H]hydroxyproline production/ug DNA)

(procollagen prolyl hydroxylation)]+ (prolyl-tRNA specific radioactivity). (3)

Collagen production (hydroxyproline production per cell)

= ([3H]hydroxyproline production/,g DNA)

+ (prolyl-tRNA specific radioactivity). (4)

Collagen production (percentage of total protein production)

= ([3H]hydroxyproline production

procollagen prolyl hydroxylation + 3.71)

[([3H]hydroxyproline production

- procollagen prolyl hydroxylation + 3.71)

+ (noncollagen protein production)], (5)

where 3.71 is a factor to account for the relative proportionof proline in collagen compared with other proteins (16).

Fibroblast PGE2 production and cAMP levels. PGE2 infibroblast culture medium and intracellular cAMPwere mea-sured by radioimmunoassays previously described (7).

Physical properties of CM. Heat stability of suppressiveactivity in CMwas assessed by incubating CMat 560 or80°C for 1 h before incubation with cultured fibroblasts.

CMwas fractionated on a 60 X 2.5-cm column of Sephacryl200 (Pharmacia Fine Chemicals, Piscataway, NJ) equilibratedwith 150 mMNaCl, 0.01 mMTris, pH 7.5, and calibratedwith molecular weight standards. CMor basal medium (3ml) was chromatographed, and pooled fractions were lyoph-ilized and resuspended in 3 ml of basal medium. Aliquots(500 ul) of the pooled fractions were then added to fibroblastcultures, and suppression of [3H]dThd incorporation wasmeasured.

Statistical evaluation. Significance of differences betweenmeans was determined by t test.

RESULTS

CM-mediated suppression of fibroblast prolifera-tion and collagen production. Incubation of lung fi-broblasts with CMfrom normal hamster macrophagesresulted in a dose-related decrease in fibroblast[3H]dThd incorporation (Fig. 1). Lower CMconcen-trations (0.1 and 1.0%) did not affect fibroblast [3H]dThdincorporation. Fibroblast viability (>90%) was not al-tered by CM. Although [3H]dThd incorporation is amore sensitive test of fibroblast proliferation, cell countswere lower in fibroblast cultures exposed to CM. Forexample, nonconfluent cultures exposed to 20% CMfor24 h contained 1.51±0.10 X 105 cells, while culturesin basal medium contained 2.61±0.32 X 105 cells(mean±SEM, n = 3).

Collagen production by fibroblasts was also sup-pressed by macrophage CM. Nondialyzable [3H]-hydroxyproline in fibroblast culture medium decreasedin a dose-related manner as indicated in Fig. 1. Nochange in [3H]hydroxyproline was noted at CMcon-centrations of 0.1 or 1.0%. Table I shows that[3H]hydroxyproline production per microgram of celllayer DNA decreased after CMexposure while pro-

2084 J. G. Clark, K. M. Kostal, and B. A. Marino

31COLLAGENPRODUCTIONPROLIFERATIONCM~~~~~~~ x

z 10 20 0 1020zI 0~~~-.~

0

0.

0 1 0200 02

CM CONCENTRATION(%C)

FIGURE 1 CM-mediated suppression of fibroblast proliferation and collagen production. Fi-broblasts were cultured in the presence of BAM-CMin varying concentration for 18 h. Tomeasure fibroblast proliferation, [3H]dThd (1 MCi) was added for 4 h, and [3H]dThd incorporationwas measured (left). To measure collagen production, [3H]proline (10 MCi), ,3-aminopropionitrile(20 Mg/in1), and ascorbic acid (50 Mg/in1) were added for 6 h, and nondialyzable[3H]hydroxyproline was measured in the medium (right). Values are mean±SEM (n = 4). CM-exposed cultures are significantly different from cultures in basal medium (P < 0.05).

TABLE IEffects of CMon Parameters of Fibroblast Collagen Production

Fibroblast culture conditions'Percent

Parameter Basal medium CM change

[3H]hydroxyproline production 3.21±0.26 1.86±0.141' 42.2(dpm x 10-4/jg DNA) (8) (9)

[3H]proline incorporation 10.72±0.51 8.00±0.13" 25.4(dpm x 10-4/Mg DNA) (8) (9)

Procollagen prolyl 40.5±0.3 41.0±2.1 (1.2)hydroxylation (%) (4) (3)

Collagen production 18.6±1.1 14.2±1.1§ 23.6(% of total soluble (6) (9)protein production)

Prolyl-tRNA specific activity 1.07±0.13 0.97±0.07 (9.3)(dpm/unit weight proline) (5) (9)

Intracellular free proline 1.56±0.05 1.31±0.01" 16.0specific activity (dpm/unit (6) (6)weight proline)

Intracellular free proline 237±11 270±15X 13.9(pg proline/gtg DNA) (8) (7)

Fibroblasts were cultured in basal medium or CM(30% concen-tration) for 18 h. Ascorbic acid and [3H]proline were added; analysesand calculations were performed as described in Methods. Values aremean±SEM(n).I Significantly different from basal medium, P < 0.05.§ Significantly different from basal medium, P < 0.01."Significantly different from basal medium, P < 0.001.

collagen prolyl hydroxylation was not reduced by CM.Although intracellular free proline was increased andintracellular free proline-specific activity decreased af-ter CMexposure, prolyl-tRNA-specific activity was notsignificantly different. Total soluble protein productionand hydroxyproline production were significantly de-creased in CM-exposed cultures when calculated onthe basis of either proline pool. Moreover, collagenproduction was selectively decreased by CMsince theproportion of total soluble protein synthesis directedtoward collagen was less in CM-exposed cultures com-pared with cultures in basal medium (Fig. 2).

Similar suppression of [3H]dThd incorporation and[3H]hydroxyproline production was observed whenhamster lung fibroblasts were exposed to CM (Ta-ble II).

The effects of culture conditions on the appearanceand/or detection of macrophage-derived fibroblastsuppressive activity were tested by varying serum con-centrations in both macrophage and fibroblast cultures.The results (Table III) indicate that fibroblasts grownin low serum concentration (0.04%) or serum-free me-dium are quiescent. Under these conditions, macro-phage-derived suppressive activity was not detectable.In fact, stimulation of [3H]dThd incorporation was ob-served. Cultures of adherent macrophages containedessentially the same level of fibroblast suppressive ac-tivity as suspension cultures.

The appearance of fibroblast suppressive activity afterintratracheal bleomycin was studied in normal hamsters


61

6

Zc30W

-w

.zzo

zOwo

OzZ_

0

-3-i

LO

L)

cr 2

ao.(cWz a.

zs z

i00°

Xa

Iq 0Basal BAM- BOsal BAM-

Medium CM Medium CMBasal BAM-

Medium CM

FIGURE 2 Effect of CMon fibroblast soluble protein production. Fibroblasts were cultured for18 h in basal medium (0) or CM(U; 30% concentration). Ascorbic acid, 0-aminopropionitrile,and [3H]proline were added, and incubation continued for an additional 6 h. [3Hjproline in-corporation and [3H]hydroxyproline production were measured in culture medium; prolyl-tRNA-specific radioactivity and DNA were measured in the cell layer. Noncollagen proteinproduction per cell (A), collagen production per cell (B), and collagen production as a percentof total soluble protein production (C) were calculated as described in Methods. Units ofnoncollagen protein production and collagen production are comparable but arbitrary becausethe specific activity of ['4C]dansyl chloride which was used to determine prolyl-tRNA-specificradioactivity was not known with certainty. Values are mean±SEM (n = 5-9). CMexposureresulted in significant suppression of noncollagen protein production (P < 0.001), collagenproduction (P < 0.001), and percent collagen production (P < 0.01).

and hamsters treated 1, 2, or 3 wk before bronchoal-veolar lavage. The results (Table IV) indicate that sup-pressive activity derived from cultures containing equalnumbers of macrophages increased over the 3-wk periodof observation. The change was accompanied by a con-comitant increase in the total number of lavage mac-rophages. Macrophages that were obtained from in-dividual animals 2 wk after bleomycin instillation werecultured, and their effects on fibroblasts were studiedin more detail. These macrophages suppressed bothfibroblast proliferation and collagen production to a

TABLE IIEffects of CMon Cultured Hamster Lung Fibroblasts°

CMconcentration [rH~dThd incorporation irH]hydroxyproline production

% cpm X 10-3/culture cpm X 10-'/yg celllayer protein

0 (Basal medium) 3.65±0.20 2.53±0.5110 2.58±0.181 2.12±0.3225 2.12±0.151 1.81±0.23t

° Fibroblasts were cultured as described in Fig. 1. Values aremean±SEM, n = 4.t Significantly different from basal medium, P < 0.01.

TABLE IIIEffect of Culture Conditions on Detection of

Macrophage-derived Suppressive Activity*

Fibroblast cultureMacrophage culture conditions conditions [H)dThd incorporation

cpm X 10-'/culture

No BAM/10% serum 10% serum 5.79±0.38Adherant BAM/10% serum 10% serum 4.99±0.50t

No BAM/0.4% serum 0.4% serum 0.49±0.06Adherant BAM/0.4% serum 0.4% serum 0.79±0.061

No BAM/no serum no serum 0.09±0.01Adherant BAM/no serum no serum 0.16±0.011

No BAM/10% serum 10% serum 5.67±0.33Suspension BAM/10% serum 10% serum 4.68±0.251§

° Macrophages were cultured in medium containing fetal calf serumconcentrations as indicated. An equal number of BAMwere culturedin polypropylene tubes ("suspension BAM"). Fibroblasts were culturedin basal medium. 2 d before addition of BAM-CM, the fibroblastmedium was changed to contain the serum concentration indicated.After incubation with macrophage CM, fibroblast [3H]dThd incor-poration was measured. Values are mean±SEM, n = 4.t Significantly different from control culture (No BAM), P < 0.05.§ No significant difference from adherant BAMin 10% serum.


TABLE IVAppearance of Macrophage-derived Fibroblast Suppressive

Activity after Intratracheal Bleomycin'

Lavage [HkdThdSource of macrophages macrophages incorporation

%/number cpm X 10-i/X 106 culture

per lung

Basal medium (no macrophages) - 5.70±0.16Control hamster (no bleomycin) 98/1.77 3.72±0.07tHamster, 1 wk after bleomycin 65/1.70 2.48±0.10§Hamster, 2 wk after bleomycin 80/2.92 2.66±0.22§Hamster, 3 wk after bleomycin 80/3.43 1.46±0.10§11e Macrophages were obtained from control hamsters and hamsterspreviously given intratracheal bleomycin. Lavage cells from eachgroup of four to six animals were pooled. The cells were counted,and differential counts obtained from hematoxylin-eosin-stained cy-tocentrifuge preparations. Macrophage cultures contained equalnumbers of macrophages. Fibroblasts were exposed to CM(10% con-centration), and fibroblast [H]dThd incorporation was measured. Val-ues are mean±SEM(n = 4 cultures).I Significantly different from basal medium, P < 0.01.§ Significantly different from control hamster, P < 0.01.11 Significantly different from 1 or 2 wk after bleomycin, P < 0.01.

greater degree than normal hamster macrophages. Fig.3 indicates that CMfrom normal hamster macrophagessuppressed [3H]dThd incorporation by 18%, whilebleomycin-treated hamster macrophage CM resulted

in 57% suppression. Similarly, [3H]hydroxyproline pro-duction was suppressed 44 and 60% by normal andbleomycin-treated hamster macrophage CM, respec-tively.

CM-mediated stimulation of fibroblast endogenousPGE2production and intracellular cAMP. CMfrommacrophages resulted in striking increases in fibroblastPGE2production (Fig. 4). Medium from fibroblasts in-cubated under basal conditions for 24 h contained only12 pg PGE2/culture. After incubation with CM, fibro-blast PGE2 increased several orders of magnitude in adose-related manner. Intracellular fibroblast cAMPwasalso increased after 24 h exposure to CMat 20% con-centration. It should be noted, however, that the valuesof cAMPpresented in Fig. 4 do not represent the max-imum cAMPlevels during this period since peak valuesoccur after 8 h of incubation with CM(Amill-Acosta,S. A., and Clark, J. G., unpublished observations).

Ablation of CM-mediated effects in indomethacinpretreated fibroblast cultures. To establish that CM-mediated stimulation of PG results in increased cAMPand decreased proliferation and collagen production,indomethacin (1 gg/ml), a cyclooxygenase inhibitor,was added to fibroblast cultures 2 h before addition ofCM. The results are shown in Fig. 5. Indomethacincompletely blocked CM-mediated increases in PGE2and cAMP, and suppression of fibroblast proliferationand collagen production was not observed. The CM-mediated increase in [3H]dThd incorporation observed

BASAL NORMAL BLEOMYCINMEDIUM BAlM-CM BSAM-CM

FIGURE 3 Suppression of fibroblast proliferation and collagenproduction by CMfrom normal and bleomycin-treated ham-ster BAM. Fibroblasts were exposed for 24 h to CM(10%concentration in proliferation assay, and 20% in collagenproduction assay) from normal hamster macrophages or mac-rophages of bleomycin-exposed hamsters. The macrophagescultures from normal hamsters and from bleomycin-exposedhamsters contained equal numbers of cells. Fibroblast pro-liferation (left) and collagen production (right) were mea-sured as described in Fig. 1. Values are mean±SD (n = 6animals). BAM-CMfrom bleomycin-exposed hamsters sup-pressed both proliferation and collagen synthesis significantlymore (P < 0.01) than BAM-CMfrom normal hamsters.

0.52

E

a-t '

I

0 10 20 0CM CONCENTRATION(%)

FIGURE 4 CM-mediated stimulation of fibroblast PGE2pro-duction and intracellular cAMP. Fibroblasts were culturedin the presence of BAM-CMin varying concentration for 24h. PGE2 was measured in culture medium (left) and cAMPwas measured in cell layers (right) by radioimmunoassay.Values are mean±SEM(n = 4). CM-exposed fibroblast cul-tures are significantly different from cultures in basal medium(P < 0.05) except for cAMP in 10% CM.

Collagen Production in Pulmonary Fibrosis

.2zO

9?

BASAL NORMAL BLEOblYCINMEDIUM BAM-CM SAM-CM

2087

PGE2 cAMP PROLIFERATION _ COLLAGEN2000 8 --8 2 4

15006E

E E-~1000 .a4 4-C

expse a.baa eim()o A-M()(0 onetaini rlfrto sa npaallCD obatclue 2 0beoeepsr obslmdu rBMC +NO.I h

d. a.

0 0

z500 2 2

0

I

-INDO *INDO 0-iNDO +INDO 0-INDO +INDO -INDO +INDO

FIGURE 5 Ablation of CM-mediated effects on fibroblasts by indomethacin. Fibroblasts wereexposed to basal medium (0) or BAM-CM(U) (10% concentration in proliferation assay and20%'in other assays) as described in Fig. 1 (-INDO). Indomethacin (1.0 jsg/ml) was added toparallel fibroblast cultures 2 h before exposure to basal medium or BAM-CM(+INDO). In theabsence of indomethacin, BAM-CMexposure resulted in stimulation of fibroblast endogenousPGE2and intracellular cAMP, and suppression of fibroblast proliferation and collagen production(P < 0.01). These GM-mediated effects were absent when fibroblasts were cultured with in-domethacin. Values are mean±SD (n = 4).

after CMexposure in indomethacin pretreated culturesseen in Fig. 5 was not a consistently reproducible ob-servation.

Characterization of BAM-derived suppressive ac-tivity. Macrophage-derived fibroblast suppressivefactor has an apparent molecular weight of 20,000-30,000 by molecular sieve chromatography. Heatingat 560C for 1 h did not reduce the activity of CM, butheating at 80'C for 1 h resulted in a 50% reduction ofsuppressive activity.

BAMlysates contained suppressive activity that in-dicated that the factor is preformed in the cells (TableV). Cycloheximide reduced macrophage 'H-proteinsynthesis to <5% of control values, but suppressive ac-tivity in CMwas not reduced. In fact, suppression of[3H]dThd incorporation was greater in fibroblast cul-tures exposed to CMfrom cycloheximide-treated mac-rophages. Macrophage release of suppressive activitywas stimulated by Sepharose 4B to levels greater thanpresent in lysates; this suggests that enhanced synthesisand release of suppressive activity is induced by Se-pharose 4B.

DISCUSSION

Lung collagen content is determined at least in partby fibroblast population size and collagen productionrates, parameters which are potentially regulated bystimulatory and inhibitory influences within the al-veolus. This study demonstrates that alveolar macro-phages are capable of releasing a product that sup-

presses fibroblast proliferation and collagen production.Several lines of evidence indicate that this suppressionis mediated by endogenous fibroblast PG productionand cAMPformation. First, fibroblast PGE2productionand cAMP levels are increased in proportion to theconcentration of BAM-CMand the degree of CM-me-diated suppression. Second, when PGE2 production isblocked with indomethacin, a cyclooxygenase inhibitor,CM-mediated increases in cAMP and suppression offibroblast proliferation and collagen production areeliminated. Third, exogenous PGE2, in amounts similarto quantities produced by fibroblasts after exposure toCM, results in increased intracellular cAMP and de-creased fibroblast proliferation and collagen production(7, 17). It has also previously been shown that PG canstimulate adenylate cyclase with resultant increases inintracellular cAMP(18), and that increases in fibroblastcAMP induced by adrenergic agonists result in a se-lective decrease in collagen production similar to thatdemonstrated in this study (19, 20). In addition, di-butyryl cAMP, an analog of cAMP, enters the cell andsuppresses fibroblast proliferation and collagen pro-duction (7, 19). We, therefore, conclude that a mac-rophage-derived factor increases endogenous fibroblastPG production with subsequent cAMP formation thatresults in decreased fibroblast proliferation and collagenproduction.

Fibroblast collagen production after exposure tomacrophage CMwas further examined in experimentsin which we measured proline-specific radioactivity inthe cultured fibroblasts. Weobserved, as has previously


TABLE VMacrophage Production of Fibroblast Suppressive Factor'

PercentCulture conditions I'HldThd incorporation suppression

cpm X 1OV4/culture

Basal medium 4.69±0.31CM(no additions) 3.07±0.161 34.7CM(cycloheximide added prior to dialysis) 3.38±0.304 27.9CM(cycloheximide) 2.62±0.121§ 44.2CM(Sepharose) 2.10±0.13t1 55.3CM(Sepharose + cycloheximide) 1.86±0.114" 60.5Macrophage lysate 2.77±0.111"1 41.0

° Macrophages were cultured in the presence of cycloheximide (100 pg/ml) and/or Sepharose4B (10% wt/vol). Fibroblasts were then exposed to dialyzed basal medium or macrophageCM(10% concentration), and fibroblast [3H]dThd incorporation was measured. Values aremean±SD (n = 3).I Significantly different from basal medium, P < 0.01.§ Significantly different from CM(no additions), P < 0.01."Significantly different from CM (Sepharose), P < 0.05.

been reported (14), that the specific radioactivity ofthe intracellular free proline pool differs from that ofthe prolyl-tRNA pool. Also, the specific radioactivityin the intracellular free pool decreased after CMex-posure. The data suggest that the fractional decreasemay not be equal in the intracellular and prolyl-tRNApools; this data emphasizes the importance of using anappropriate precursor pool specific activity to calculateactual collagen production rate. Our results also indicatethat the change in proline-specific radioactivity afterCMexposure is associated with increased intracellularfree proline. A possible mechanism for this increase isan increase in endogenously derived proline. It haspreviously been shown that increased fibroblast cAMPafter exposure to adrenergic agonists results in de-creased procollagen production due at least in part toincreased intracellular degradation of procollagen (20).Increased cAMPand subsequent increased intracellulardegradation in CM-exposed cultures could contributeproline to the intracellular free pool, with the resultantdecrease in specific radioactivity which we observed.Wealso found that the decrease in collagen productionwas greater than that of other soluble proteins producedin culture; this indicates that CM-mediated suppressionis selective for collagen. This study did not determinewhether the selective decrease in collagen productionoccurs as the result of increased intracellular degra-dation of collagen and/or decreased levels of procol-lagen messenger RNAor procollagen translation rate.

The active product released from alveolar macro-phages is nondialyzable, heat stable to 56°C, and hasan apparent molecular weight of 20,000-30,000. It doesnot appear to be species restricted since both hamster

and human fibroblasts are similarly suppressed. Thebiological and biochemical characteristics of this mac-rophage product resemble those of the fibroblast-sup-pressive factor derived from hamster lung explant cul-tures which we previously reported (7, see below). In-terleukin 1 also similarly stimulates PGproduction andsuppresses proliferation of rheumatoid synovial cells,but macrophage-derived fibroblast suppressive factoris of greater apparent molecular weight (21, 22). Furtherbiochemical characterization will be necessary to de-termine the relationship, if any, to interleukin 1 derivedand characterized from murine and human mononu-clear cells (23).

Since lysates of lavage macrophages contain sup-pressive activity and cycloheximide did not reduceproduction by cultured macrophages, adherent alveolarmacrophages may release primarily preformed sup-pressive factor during 24 h in culture. However, in-creased suppressive activity induced by Sepharose 4Bexceeded that in lysates; this indicated that productionof macrophage-derived suppressive factor can be stim-ulated in vitro. Suppressive activity was also greaterin macrophage cultures exposed to cycloheximide com-pared with control macrophage cultures. This may rep-resent superinduction such as is observed after acti-nomycin D (24). An alternative explanation for thisobservation is that synthesis and/or release of a mac-rophage-derived fibroblast stimulatory (growth) factoris reduced by cycloheximide, while release of the sup-pressive factor is independent of new protein synthesis.A macrophage-derived stimulatory factor whose pro-duction is inhibited by cycloheximide has recently beendescribed in human cells (25), and we have also detected


fibroblast stimulatory activity in freshly adherent ham-ster macrophages (7); this suggests that the productionof soluble macrophage-derived products may be subjectto different control mechanisms.

The rate of synthesis and/or release of fibroblast-suppressive activity from macrophages could potentiallymodulate lung collagen accumulation in vivo. Alveolarmacrophages obtained from animals after bleomycintreatment released more fibroblast-suppressive activityinto culture medium than macrophages obtained fromnormal hamsters; this suggests that macrophages mightlimit collagen accumulation after bleomycin-inducedinjury. This role is further facilitated by the appearanceof increased numbers of macrophages at 2 and 3 wk."Negative regulation" of collagen production as a com-ponent of the response to fibrogenic injury in lung hasbeen suggested by previous studies. Phan and Thrall(26) detected a soluble factor in lung tissue which se-lectively suppressed collagen production in lung explantcultures. They found that suppressive activity was di-minished in the first week after bleomycin adminis-tration. Werecently described a factor present in lungexplants that increased endogenous fibroblast PG pro-duction and intracellular cAMP level and suppressedfibroblast proliferation and collagen production (7). Wefound that animals treated with bleomycin initially haddecreased levels of this suppressive activity. Relativelyfew lavage macrophages are present at this time whenpolymorphonuclear cell exudation predominates (8).Subsequently, levels of suppressive activity exceedingnormal values were detected at times when lung col-lagen synthesis rate was declining (i.e., 1-2 wk afterbleomycin administration). Although these observationsdo not preclude direct stimulation of collagen produc-tion during the development of pulmonary fibrosis, theysuggest that suppression of collagen synthesis is im-portant in regulating collagen production and in lim-iting collagen accumulation in response to injury. Fail-ure to appropriately inhibit collagen production afterinjury could contribute to excessive accumulation ofcollagen, which is characteristic of pulmonary fibrosis.

Our results indicate that alveolar macrophages,prominent cells in the subacute phase of bleomycin-induced pulmonary fibrosis as well as in pulmonaryfibrosis in man (1, 8), are a cellular source of suppressiveactivity for fibroblasts in the lung. In other studies,mononuclear phagocytes also emerge as secretory cellsthat may influence the formation of collagen by productsthat modulate fibroblast proliferation and collagen pro-duction (21, 25, 27-31). In addition, release of fibro-nectin (32, 33), proteases such as collagenase (34) andelastase (35), and PG (36, 37) by macrophages mayfurther modify the production and deposition of con-nective tissue. However, in vivo interactions betweenmononuclear phagocyte products and fibroblasts remain

uncertain, in part because regulation of these macro-phage secretory products is not well understood. In thisstudy, we included serum in fibroblast cultures so thatwe could examine dividing fibroblasts, a model systemthat in this respect may be appropriate for studies re-lated to injured lung in which fibroblasts proliferate.Whenserum was omitted from the cultures, fibroblastsbecame quiescent and an increase in fibroblast [3H]dThdincorporation was observed after exposure to macro-phage CM; this suggests that a fibroblast-stimulatingfactor is also present. The precise role of serum in thecultures is not known, but it is evident that mononuclearcells and macrophages can release a number of productsin vitro, of which some have opposing effects on fi-broblasts (38). The result in vivo will depend on controlof their release as well as fibroblast "receptivity" tomononuclear cell products. Further characterization ofmacrophage products, regulation of their release, andputative fibroblast receptors may be useful in eluci-dating the role of macrophages in pulmonary fibrosis.

ACKNOWLEDGMENTSWethank Dr. John A. McDonald and Dr. Robert M. Seniorfor helpful suggestions in the preparation of this manuscript.

This work was supported by U.S. Public Health Servicegrant HL16118 from the National Institutes of Health.

REFERENCES1. Crystal, R. G., J. D. Fulmer, W. D. Roberts, M. D. Moss,

B. R. Line, and H. Y. Reynolds. 1976. Idiopathic pul-monary fibrosis: clinical, histologic, radiographic, phys-iologic, scintigraphic, cytologic, and biochemical aspects.Ann. Intern. Med. 85:769-788.

2. Katzenstein, A.-L. A., and F. G. Askin. 1982. In SurgicalPathology of Non-neoplastic Lung Disease, W. B. Saun-ders Co., Philadelphia. 3:43-72.

3. Spencer, H. 1980. Pathogenesis of interstitial fibrosis ofthe lungs. In Pulmonary Diseases and Disorders. A. P.Fishman, editor. McGraw-Hill, Inc., New York. 56:648-652.

4. Starcher, B. C., C. Kuhn, and J. E. Overton. 1978. In-creased elastin and collagen content in the lungs of ham-sters receiving an intratracheal injection of bleomycin.Am. Rev. Respir. Dis. 117:299-305.

5. Snider, G. L., J. A. Hayes, and A. L. Korthy. 1978. Chronicinterstitial pulmonary fibrosis produced in hamsters byendotracheal bleomycin. Pathology and sterology. Am.Rev. Respir. Dis. 117:1099-1108.

6. Clark, J. G., J. E. Overton, B. A. Marino, J. Uitto, andB. C. Starcher. 1980. Collagen biosynthesis in bleomycin-induced pulmonary fibrosis in hamsters. J. Clin. Med.96:943-953.

7. Clark, J. G., K. M. Kostal, and B. A. Marino. 1982. Mod-ulation of collagen production following bleomycin-in-duced pulmonary fibrosis in hamsters: presence of a factorthat increases fibroblast prostaglandin E2 and cAMPandsuppresses fibroblast proliferation and collagen produc-tion. J. Biol. Chem. 257:8098-8105.

8. Clark, J. G., and C. Kuhn. 1982. Bleomycin-induced pul-monary fibrosis in hamsters: effect of neutrophil depletion


on lung collagen synthesis. Am. Rev. Respir. Dis.126:737-739.

9. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randell. 1951. Protein measurement with the Folin phe-nol reagent. J. Biol. Chem. 193:265-275.

10. Burton, K. 1956. A study of the conditions and mech-anisms of the diphenylamine reaction for the colorimetricestimation of deoxyribonucleic acid. Biochem. J. 62:315-323.

11. Clark, J. G., B. C. Starcher, and J. Uitto. 1980. Bleomycin-induced synthesis of type I procollagen by human lungand skin fibroblasts in culture. Biochim. Biophys. Acta.631:359-370.

12. Peterkofsky, B., and R. Diegelmann. 1971. Use of a mix-ture of proteinase-free collagenases for the specific assayof radioactive collagen in the presence of other proteins.Biochemistry. 10:988-994.

13. Airhart, J., J. Kelley, J. E. Braydon, R. B. Low, andW. S. Stirewalt. 1979. An ultramicro method of aminoacid analysis: application to studies of protein metabolismin cultured cells. Anal. Biochem. 96:45-55.

14. Hildebran, J. N., J. Airhart, W. S. Stirewalt, and R. B.Low. 1981. Prolyl-tRNA-based rates of protein and col-lagen synthesis in human lung fibroblasts. Biochem. J.198:249-258.

15. Troll, W., and J. Lindsley. 1955. A photometric methodfor the determination of proline. J. Biol. Chem. 215:655-660.

16. Breul, S. D., K. H. Bradley, A. J. Hance, M. P. Schaffer,R. A. Berg, and R. G. Crystal. 1980. Control of collagenproduction by human diploid lung fibroblasts. J. Biol.Chem. 255:5250-5260.

17. Saltzman, L. E., J. Moss, R. A. Berg, B. Hom, and R. G.Crystal. 1982. Modulation of collagen production by fi-broblasts. Effects of chronic exposure to agonists thatincrease intracellular cyclic AMP. Biochem. J. 204:24-40.

18. Samuelsson, B., M. Goldyne, S. Granstrdm, M. Hamberg,S. Hammarstrbm, and C. Malmsten. 1978. Prostaglandinsand thromboxanes. Annu. Rev. Biochem. 477:997-1029.

19. Baum, B. J., J. Moss, S. D. Breul, and R. G. Crystal.1978. Association in normal human fibroblasts of elevatedlevels of adenosine 3',5' monophosphate with a selectivedecrease in collagen production. J. Biol. Chem.253:3391-3394.

20. Baum, B. J., J. Moss, S. D. Breul, R. A. Berg, and R. G.Crystal. 1980. Effect of cyclic AMPon intracellular deg-radation of newly synthesized collagen. J. Biol. Chem.255:2843-2847.

21. Dayer, J.-M., S. R. Goldring, D. R. Robinson, and S. M.Krane. 1979. Effects of human mononuclear cell factoron cultured rheumatoid synovial cells. Interactions ofprostaglandin E2 and cyclic adenosine 3',5'-monophos-phate. Biochim. Biophys. Acta. 586:87-105.

22. Mizel, S. B., J.-M. Dayer, S. M. Krane, and S. E. Mer-genhagen. 1981. Stimulation of rheumatoid synovial cellcollagenase and prostaglandin production by partiallypurified lymphocyte-activating factor (interleukin 1).Proc. Natl. Acad. Sci. USA. 78:2474-2477.

23. Oppenheim, J. J., B. M. Stadler, R. P. Siraganian, M.

Mage, and B. Mathieson. 1982. Properties of interleukin1. Fed. Proc. 41:257-262.

24. Palmiter, R. D., and R. T. Schimke. 1973. Regulation ofprotein synthesis in chick oviduct. III. Mechanism ofovalbumin "superinduction" by actinomycin D. J. Biol.Chem. 248:1502-1512.

25. Bitterman, P. B., S. I. Rennard, G. W. Hunninghake,and R. G. Crystal. 1982. Human alveolar macrophagegrowth factor for fibroblasts. Regulation and partialcharacterization. J. Clin. Invest. 70:806-822.

26. Phan, S. H., and R. S. Thrall. 1982. The role of solublefactors in bleomycin induced pulmonary fibrosis. Am.J. Pathol. 106:156-164.

27. Korn, J. H., P. V. Halushka, and E. C. LeRoy. 1980.Mononuclear cell modulation of connective tissue func-tion. Suppression of fibroblast growth by stimulation ofendogenous prostaglandin production. J. Clin. Invest.65:543-554.

28. Jimenez, S. A., W. McArthur, and J. Rosenbloom. 1979.Inhibition of collagen synthesis by mononuclear cell su-pernates. J. Exp. Med. 150:1421-1431.

29. Leibovich, S. J., and R. Ross. 1976. A macrophage-de-pendent factor that stimulates the proliferation of fibro-blasts in vitro. Am. J. Path. 84:501-513.

30. Goldstein, R. H., K. Miller, J. Glassroth, R. Linscott,G. L. Snider, C. Franzblau, and P. Polgar. 1982. Influenceof asbestos fibers on collagen and prostaglandin produc-tion in fibroblast and macrophage co-cultures. J. Lab.Clin. Med. 100:778-785.

31. Hibbs, M. S., A. E. Postlethwaite, C. L. Mainardi, J. M.Seyer, and A. H. Kang. 1983. Alterations in collagenproduction in mixed mononuclear leukocyte-fibroblastcultures. J. Exp. Med. 157:47-59.

32. Villager, B., T. J. Broekelmann, D. G. Kelley, G. J. Hey-mach, and J. A. McDonald. 1981. Bronchoalveolar fi-bronectin in smokers and non-smokers. Am. Rev. Respir.Dis. 124:652-654.

33. Rennard, S., and R. G. Crystal. 1982. Fibronectin inhuman bronchopulmonary lavage fluid. Elevation in pa-tients with interstitial lung disease. J. Clin. Invest.69:113-122.

34. Werb, Z., and S. Gorden. 1975. Secretion of a specificcollagenase by stimulated macrophages. J. Exp. Med.142:346-360.

35. Rodriguez, R. J., R. R. White, R. M. Senior, and E. A.Levine. 1977. Elastase release from human alveolar mac-rophages: comparison between smokers and non-smokers.Science (Wash. DC). 198:313-314.

36. Demenkoff, J. H., M. D. Ansfield, H. B. Kaltreider, andE. Adam. 1980. Alveolar macrophage suppression of ca-nine bronchoalveolar lymphocytes: the role of prosta-glandin E2 in the inhibition of mitogen-responses. J. Im-munol. 124:1243-1365.

37. Rouzer, C. A., W. A. Scott, A. L. Hamill, and Z. A. Cohn.1982. The synthesis of leukotriene C and other arachi-donic acid metabolites by pulmonary macrophages. J.Exp. Med. 155:720-733.

38. Jalkenen, M., J. Peltonen, and E. Kulonen. 1979. Iso-electric focusing of macrophage culture medium and theeffects on the synthesis of DNAand collagen by fibro-blasts. Acta. Pathol. Microbiol. Scand. Sect. C. Immunol.87:347-352.


bleomycin-induced pulmonary fibrosis hamsters€¦ · in hamsters is associated with collagen...

Documents