Thorax 1996;51:781-786

Original articles

Effect of exposure to silica on human alveolarmacrophages in supporting growth activity in typeII epithelial cells

B Melloni, 0 Lesur, T Bouhadiba, A Cantin, M Martel, R Begin

Unite de RecherchePuhnonaire,Universite deSherbrooke,Sherbrooke, QuebecCanada J1H 5N4B Melloni0 LesurT BouhadibaA CantinM MartelR Begin

Correspondence to:Dr R Begin.

Received 4 July 1995Returned to authors23 October 1995Revised version received13 December 1995Accepted for publication5 February 1996

AbstractBackground - The proliferative responseof type II cells is an important eventfollowing silica-induced lung injury. Al-veolar macrophages, when activated byfibrogenic agents, secrete various biologi-cal mediators which affect cell growth.Methods - Human alveolar macrophagesfrom normal volunteers were incubated inserum-free medium or in the presence ofincreasing concentrations of silica. Alveo-lar macrophage conditioned media werediluted and added to type II cell culturesfor proliferation studies. Purified type IIpneumocytes were isolated from fetal ratlungs for bioassays. Growth factor activi-ties were partially characterised by sizeexclusion chromatography. Each fraction-ated mitogenic peak was preincubatedwith monoclonal antibody against plateletderived growth factor (PDGF) or antiseraagainst insulin-like growth factor 1(IGF-1) or fibroblast derived growth fac-tor (FGF) in order to study the nature ofeach activity.Results - Conditioned media from alveo-lar macrophages exposed to silica inducedan increase in type II cell DNA synthesisand cell number over that seen when typeII cells were incubated with unstimulatedalveolar macrophage supernatants. Sizeexclusion of alveolar macrophage super-natants exposed to silica showed fourpeaks of type II cell stimulating activitywith apparent molecular weights of 38, 22,16, and 8 kDa. Anti-PDGF antibodysignificantly reduced the activity of thefirst and second peaks, antiserum againstIGF-1 partially reduced the activity of thefirst and fourth peaks, and antiserumagainst FGF reduced only the third peakof activity.Conclusions - Human alveolar macro-phages exposed to silica in vitro releasemitogens for type II pneumocytes includ-ing PDGF-like, IGF-1-like, and FGF-likemolecules. These agents are likely to beinvolved in the epithelial repair and type IIcell hyperplasia observed in silicosis.(Thorax 1996;51:781-786)

Keywords: alveolar macrophages, silicosis, type II cells.

Proliferation of type II pneumocytes is an earlyevent following silica-induced lung injury inanimal models.' 2 Type II cell hyperplasia hasbeen reported after exposure to high levels ofsilica in humans and type II cells have beenfound in the bronchoalveolar lavage (BAL)fluid of acutely exposed workers.3 It is estab-lished that alveolar macrophages are importantcells in the pathogenesis of silicosis by releasingmediators and messengers for other cell typesand for the extracellular matrix.6 Alveolarmacrophages are the main cell found in BALfluid from patients with subclinical or simplesilicosis.i 7

We have studied the direct effect of lowdoses of silica on fetal rat type II cells in vitro,8and have identified growth factor activity inBAL fluid from normal human lung thatstimulates fetal rat type II cell growth in vitro.9This mitogenic activity was clearly enhanced inBAL fluid from patients with early stages ofsilicosis. In addition, similar results were docu-mented with BAL fluid from our experimentalmodel of sheep silicosis. We have also foundthat conditioned media from alveolar macro-phages of control sheep induced type II cellgrowth."0 Supernatants from in vitro or in vivosheep alveolar macrophages exposed to silicasignificantly increased type II pneumocytegrowth compared with those from non-exposed sheep alveolar macrophages. Partialcharacterisation of these mitogenic activitieshave recently been performed."

It was hypothesised that human macro-phages exposed in vitro to low doses of silicacan be stimulated to release an increasedmitogenic activity for type II cells. Factorsinvolved in this macrophage-derived mitogenicactivity are characterised in this study.

MethodsALVEOLAR MACROPHAGE PREPARATIONSAlveolar macrophages were obtained by bron-choalveolar lavage from 10 normal volunteersas described previously.7 The mean (SD) age ofthe subjects was 28 (3) years and all were non-

781

on 24 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.8.781 on 1 A

ugust 1996. Dow

nloaded from

Melloni, Lesur, Bouhadiba, Cantin, Martel, Begin

smokers and had not been exposed to mineraldusts. The lavage fluid was filtered throughsurgical gauze and centrifuged at 400g for 10minutes. The cell pellet was suspended inEagle's minimal essential medium (MEM;Gibco, Grand Island, New York, USA) supple-mented with 10% fetal bovine serum (FBS)and antibiotics at a concentration of 2 x 106cells. After one hour of adherence the cellswere washed to remove non-adherent cellssuch as lymphocytes. Adherent cells werecultured in 24-well tissue culture plates (Fal-con Inc, Lincoln Park, New Jersey, USA) inserum-free medium supplemented with 2 mMglutamine, with or without mineral dusts (seebelow) and incubated for 24 hours at 37°C in5% C02/air. At the end of the culture periodsupernatants were removed, centrifuged (500gfor 10 minutes), filtered through a 0.22 jim fil-ter (Millex-GS; Millipore Corp, Bedford, Mas-sachusetts, USA) and stored at -70°C untilfurther use. The nature of the adherent cellswas evaluated on a cytocentrifuge preparationstained with Wright-Giemsa. In all cases theresulting adherent cells were ¢ 92% alveolarmacrophages with cell viability of > 95% asassessed by exclusion of trypan blue dye.

EXPOSURE OF ALVEOLAR MACROPHAGES TOMINERAL DUSTSDifferent concentrations of mineral dust parti-cles were suspended in serum-free mediumand added to macrophage cultures for 24hours as reported previously.'0 Silica(Minusil-5; Pennsylvania Glass Co, Pittsburgh,Pennsylvania, USA), titanium dioxide (KronosInc, Paris, France), or aluminum-treatedquartz'2 were heated for two hours at 200°C forsterilisation prior to use. All particles had adiameter of < 5,m. Alveolar macrophageconditioned media were routinely checked forendotoxin contamination by the limulus amoe-bocyte assay (Sigma, St Louis, Missouri,USA). The alveolar macrophage supernatantswere filtered through a 0.45 gm filter andstored at -200C until required for cell prolif-eration assays.

EVALUATION OF SILICA DUST PARTICLES INALVEOLAR MACROPHAGESThe number of particles in alveolar macro-phages was evaluated by polarised light micro-scopy. Quantification of the particles in theformer evaluation was carried out by counting500 cells for each sample with three independ-ent determinations.

DUST CYTOTOXICITY ON ALVEOLARMACROPHAGESLactate dehydrogenase (LDH) release in cul-ture medium was quantified to determine dustcytotoxicity on macrophages after incubationfor 24 hours.'0

ASSAY FOR TYPE II CELL GROWTHPneumocytes were isolated from fetal rat lungsas reported.8 These cells exhibited an immuno-cytochemical staining for cytokeratin 19, anintermediate filament specific to epithelialcells. Type II cells were used at passage 2 and

were rendered quiescent by a 48 hour incuba-tion in "breaking medium" (MEM) containing2% serum (heat inactivated FBS). Condi-tioned media from alveolar macrophages ex-posed or not exposed to mineral dusts wereadded to the pneumocyte cultures.'0 Given thelog phase between DNA synthesis and celldivision, cell growth was evaluated by theincorporation of tritiated thymidine at 24hours and by counting the monolayers at 48hours.89

PARTIAL CHARACTERISATION OF THEMACROPHAGE-DERIVED MITOGENIC ACTIVITIESTo produce sufficient amounts of growthpromoting activity for partial purification,adherent cells from six normal individuals wereincubated in serum-free medium supple-mented with 50 jg/ml silica particles. This lat-ter concentration induced the higher mitogenicactivity on type II cells.

GelfiltrationPooled alveolar macrophage supernatants fromsix healthy volunteers were lyophilised, con-centrated 50-fold, and reconstituted in 1 mlsterile phosphate buffered saline (PBS) as pre-viously reported." Gel filtration was performedon a G-75 SF column (Pharmacia Inc, Baied'urfe, Quebec, Canada) at a flow rate of 0.75ml/min and 2.5 ml fractions were collected.The eluate was monitored by absorbance read-ing at 280 nm. Fifty four fractions were diluted1:2 in MEM supplemented by 2% serum andtested in type II cell DNA synthesis assays.

Biophysical properties of macrophage-derivedmitogenic peaksTo study the properties of the alveolarmacrophage derived growth activity, adjacentfractions with maximum activity for type IIcells were pooled, lyophilised, and resuspendedin medium as described above.Acid treatment: Aliquots of mitogenic peak

were dialysed against 1 M acetic acid (pH 2.5,three changes, 24 hours), followed by dialysisagainst MEM.Heat treatment: Aliquots of mitogenic peak

were incubated at 100°C or 560C for 30minutes and then cooled at room temperature.

Role of reducing agent: Aliquots of mito-genic peak were incubated with 2 mMdithiothreitol for 30 minutes at 37°C anddialysed against 1000 volumes ofMEM (threechanges, 24 hours).

Sensitivity to protease reduction: Aliquots ofmitogenic peak were incubated with 1700 BAEunits of trypsin (type IX, 30 minutes, 37°C)followed by incubation with twice the concen-tration of soybean trypsin inhibitor.

All treated aliquots were tested on type IIcell [3H]-thymidine incorporation and com-pared with an untreated aliquot prepared bythe same process. All treated aliquots and con-trols were filtered prior to assay on type II cellcultures.

Blocking antibodiesPreincubation of 50 ng/ml human recom-binant platelet-derived growth factor antibody(PDGF) with 50 gg/ml anti-PDGF IgG (Anti-

782

on 24 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.8.781 on 1 A

ugust 1996. Dow

nloaded from

Effect of exposure to silica on human alveolar macrophages

20 r

15

10

5

] Negative controlW SilicaC Aluminum-treated silicaE7 Titanium dioxideI Positive control (Triton 1 %)

* *

*

0 10 25 50 100 500 1000 50 100 50 100 Triton1%

Figure 1 Release of lactate dehydrogenase by human alveolar macrophages exposed invitro for 24 hours to silica, aluminum-treated silica, and titanium dioxide. *p < 0. 05compared with control value.

PDGF, Collaborative Research Inc, Bedford,Massachusetts, USA) completely blocked theability of this cytokine to stimulate [3H]-thymidine incorporation by type II cells."G-75 mitogenic peaks were preincubated withmonoclonal anti-PDGF IgG prior to the deter-mination of mitogenic activity. Rabbit polyclo-nal antisera to human recombinant insulin-likegrowth factor 1 and to fibroblast growth factorfrom bovine pituitary extracts (IGF- 1 andFGF, Collaborative Research Inc) were pro-

duced by immunisation of New Zealand whiterabbits with IGF- 1 emulsified in Freund'sadjuvant. Rabbit antiserum to IGF-1 com-

pletely blocked the activity of 1 ig/ml IGF- 1 ontype II cell DNA synthesis at a dilution of1:500, and rabbit antiserum to FGF at a dilu-tion of 1:2000 completely blocked the effect of500 ng/ml FGF on type II cell DNA synthesis.We have previously shown" that each blocking

*

0

oo-,4-0

-

0

,

C;

L-

500 r

400 *

300 e

200

100 e-

o1 1:2 1:5 1:10 1:20 Control

Conditioned media dilutions

Figure 2 Comparison of the effect of supernatants fromunstimulated macrophages (0) and macrophages exposedto silica in vitro (@) on incorporation of f3H]-thymidineinto type II cells. *p < 0. 05 versus unstimulatedmacrophage supernatants. Correlations between thymidineincorporation and cellular proliferation assessed by cellcounting were significant at silica concentrations from 10 to50 ,ugl ml (1 ugl ml: r' = 0.14, 25 ugl ml: r2 = 0. 54, 50jugl ml: r' = 0. 28,p < 0. 05).

antibody or antiserum is specific for the corre-sponding cytokine.

STATISTICAL ANALYSISData are expressed as mean (SE). Statisticalsignificance (p < 0.05) was determined byanalysis of variance (ANOVA) and a Student'st test for comparison with control data.

ResultsEFFECTS OF IN VITRO SILICA DUST EXPOSURE ONALVEOLAR MACROPHAGESIncreasing concentrations of silica induced ahigher release ofLDH than under control con-ditions (fig 1); low doses of silica did notprovoke a significant release of LDH. Treat-ment with titanium dioxide or aluminum-treated silica did not increase LDH release.

EVALUATION OF SILICA PARTICLES IN ALVEOLARMACROPHAGESDust particles were rarely found in alveolarmacrophages of control patients (2 (1)% oftotal cells). A linear relationship between dustconcentration and the percentage of adherentmacrophages containing silica particles wasfound when silica was incubated with alveolarmacrophages for 24 hours (8 (2)% at 25 ,ug/ml,14 (4)% at 50 pg/ml, 22 (5) % at 100 gg/ml).

EFFECT OF MACROPHAGE SUPERNATANTS ONTYPE II CELL GROWTHThe effects of macrophage conditioned mediawere tested for their ability to induce thegrowth of fetal rat type II cells in vitro. Condi-

600c0

" 5000-

0L-

4

o aU o 400

C >LE 0 300

s200

-0

100

[7 UnstimulatedW Silica= Aluminum-treated silicaF Titanium dioxideA

0 10 25 50 100 50 50

B200

*

CD

0~~~~~

DE LXl:3 a) 150~C >

-o

00

0 10 25 50 100 50 50Conditioned media by dusts (pg/ml)

Figure 3 Effects of supernatants from dust-exposedmacrophages on (A) DNA synthesis of type II cells and(B) numbers of type II cells. *p < 0.05 versusunstimulated macrophage supernatants.

a)en

a)

I1-

n,II .I . ..., ...

T

783

on 24 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.8.781 on 1 A

ugust 1996. Dow

nloaded from

Melloni, Lesur, Bouhadiba, Cantin, Martel, Bigin

kDa31 201 12 4 6 5 0 66

02 C0

i / o

0k // 0.1 C.%\ ,/ \\ ,'x

Co0

0 .0

Ao 10 20 30 iT 0TFraction number

)0

0

c,)

40 50

Figure 4 Size exclusion chromatography on Sephadex G-75 gelfiltration. One ml ofconditioned media from macrophages exposed to silica (concentrated 50 times) was elutedand each resulting fraction was testedfor type II cell [3H]-thymidine incorporation (0).The absorbance offraction protein content was monitored at 280 nm (----). Molecularweight standards (arrows), starting with the highest, are blue dextran, ovalbumin,deoxyribonuclease-1, soybean trypsin inhibitor, cytochrome C, aprotinin, and oxidisedglutathione. Data from one representative experiment out offour are shown.

tioned media from adherent macrophagesstimulated ['H]-thymidine incorporation intoprimary fetal rat type II cells and this effect wassignificantly increased with conditioned mediafrom macrophages incubated in vitro with 50tg/ml silica (fig 2). The amount of growth fac-tor activity for type II cells released dependedon the concentration of silica dust used, withlow concentrations of silica inducing the great-

Table I Effects of various biochemical treatments on the percentage activity remaining inthe peaks isolatedfrom fractionated macrophage conditioned media compared withuntreated controls assessed by ['H]-thymidine incorporation in type II cells

% activity remaining in peak

Treatment 1 2 3 4

Heat (100°C, 15 min) 84 (6) 88 (6) 61 (4) 64 (9)Heat (56°C, 30 min) 96 (5) 92 (3) 79 (5) 88 (7)Reduction (2 mM DTT, 37°C, 30 min) 25 (8) 72 (6) 63 (4) 75 (6)Acid (pH 2.5, 1 M acetic acid, dialysed 24 hours) 85 (3) 82 (5) 73 (5) 48 (8)Trypsin (1700 BAE units/ml, 30 min, 37°C) 39 (5) 44 (7) 33 (4) 32 (5)

Results represent mean (SE) of four experiments.DTT = dithiothreitol.

Table 2 Effects ofpreincubation with blocking antibodies on percentage of mitogenicactivity remaining compared with untreated controls

Inhibition of type II cell DNA synthesis (G/o)

Anti-PDGF Anti-IGF-I Anti-FGF Anti-PDGF +Peaks (50,uglml) (1:500) (1:2000) anti-IGF-I

1 40 (5) 37 (4) 19 (3) 34 (10)2 27 (3) 26 (9) 14 (5) 46 (5)3 6 (2) 15 (3) 50 (6) 3 (2)4 16 (4) 35 (4) 12 (3) 30 (7)

Results represent mean (SE) of four experiments.

est mitogenic effect (fig 3). In comparison,macrophages incubated with inert dust/titanium dioxide or modified dust/aluminium-treated silica did not release a similar mitogenicactivity for type II cells. To determine whetherthe increase in thymidine incorporation wasfollowed by an increase in cell number, type IIcells were counted after 48 hours in cultureand the cell counts were correlated with thymi-dine incorporation (fig 2).

MACROPHAGE CONDITIONED MEDIA SIZEEXCLUSIONMacrophages were incubated for 24 hours inmedium containing 50 gg/ml of silica, the con-centration which induced the greatest releaseof mitogenic activity. Alveolar macrophageconditioned media were separated according totheir molecular weights with a G-75 column.Type II cell growth promoting activity (forexample, [3H]-thymidine incorporation) wasdetected in four peaks (fig 4). The highestmitogenic peaks eluted with apparent molecu-lar weights of 38 kDa (first peak) and 16 kDa(third peak). In addition, two smaller peakswere detected with apparent molecular weightsof 22 kDa (second peak) and 8 kDa (fourthpeak).

PHYSICAL PROPERTIES OF MITOGENIC ACTIVITIESEach peak was subjected to various treatmentsto determine the biochemical properties ofmacrophage growth promoting activities fortype II cells (table 1). Heating the first and sec-ond peaks to 100°C for 10 minutes or to 56°Cfor 30 minutes resulted in only a slightreduction in mitogenic activity compared withthe third and fourth peaks. On the other hand,the first and second peaks were more resistantto acid treatment than the third and fourthpeaks of activity. The reducing agent dithio-threitol significantly altered the activity of thefirst peak. Treatment of each peak with trypsinresulted in a considerable loss of growthpromoting activity.

NEUTRALISING ANTIBODIES IN MITOGENIC PEAKSAntibody against PDGF and antisera againstIGF- 1 and FGF were used to block themitogenic activity of each peak fractionatedfrom silica-exposed macrophage supernatants(table 2). A significant part of the mitogenicactivity of the first and second peaks was inhib-ited by anti-PDGF IgG (40% for the first peakand 27% for the second peak). Anti-IGF-1antiserum substantially reduced the activity ofthe first (37%), second (26%), and fourth(35%) peaks of activity; anti-FGF antiserumaffected the mitogenic activity of all peaks, butonly that of the third peak (50%) to asignificant extent. Several combinations ofanti-PDGF IgG and antisera against IGF-1 orFGF were tested and a synergistic neutralisingeffect on the second peak (46%) was shown bythe combination of anti-PDGF with anti-IGF-1. On the other hand, the addition ofeither anti-PDGF antibody or antisera againstIGF-1 or FGF alone to unfractionated condi-tioned media had no effect on type II cell DNAsynthesis (data not shown).

200 45

0

.0 300Co-

.° 250Co0C.)

a)0)

C

._

E 1250

10

<I 1200

50

784

on 24 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.8.781 on 1 A

ugust 1996. Dow

nloaded from

Effect of exposure to silica on human alveolar macrophages

DiscussionMany recent studies have focused on the abilityof inorganic particles to induce secretion offibroblast mitogenic activities by macrophagesin vitro.'3 14 Macrophages isolated from patientsexposed to silica dust released in vitro analveolar macrophage-derived progressiongrowth factor'5 which is involved in fibroblastregulation and has been shown to be an IGF-1molecule with a molecular weight of 18-25kDa.'6 Type II cell hyperplasia in animal mod-els appears to be an early event following type Iepithelial damage2 and this also appears to be afeature of human silicosis.3 In this study thenature of the mitogenic activities for type IIcells released by human alveolar macrophagesincubated in vitro with low doses of silica wasinvestigated.We have previously explained the choice of

fetal rat type II cells for bioassays.8 Earlierstudies have shown that fetal rat type II cellswere differentiated at 19 days and expresseddifferent markers of alveolar epithelial cells. 7Most importantly, fetal type II cells are able todivide in cultures and the cell numbers arestrictly correlated with DNA synthesis in ourmodel. Of the well known mitogenic factors,PDGF, FGF, and IGF-1 stimulate fetal type IIcell growth in vitro,9 " and these are releasedspontaneously by activated alveolar macro-phages. "Our results indicate that supernatants from

human alveolar macrophages exposed to silicain vitro induced a significantly greater increasein fetal type II cell growth than that seen withsupernatants from unstimulated adherentmacrophages. This effect is specific for silicaparticles and was not observed with inert dust/titanium dioxide or modified dust/aluminum-treated silica. Importantly, low doses of silicainduced the highest growth factor activity. Thisobservation confirms the results previouslyreported with sheep alveolar macrophagesexposed in vitro or in vivo to silica." Activationof macrophages by silica could stimulate theirproduction of macrophage derived growth fac-tor for type II cells. The percentage ofmacrophages in the BAL fluid that containedparticles of silica dust was 52% in the subjectsexposed to silica and only 6% in controlsubjects.'9 A non-cytotoxic concentration of 50,ug/ml of silica was used for partial characterisa-tion in our experiments. At this concentration,only 14% of the cells contained cytoplasmicmineral particles.Pneumoconioses are characterised by an

alveolitis rich in alveolar macrophages.3 6 7 Therelease of soluble mediators from alveolar mac-rophages in patients with silicosis has beeninvolved in the pathogenesis of inflammationand fibrosis.4 It is well established that theinteraction between macrophages and silica isan early event in the inflammatory response tomineral dust. In addition, alveolar type II cellproliferation is an important process whichoccurs after most types oflung injury includingexposure to silica dust. The mechanisms thatregulate alveolar type II cell proliferation inlung injury are poorly understood. We havedocumented a proliferative activity for fetal

type II cells in BAL fluid from healthyvolunteers in our previous studies9 and haveshown that this activity was significantlyenhanced in BAL fluid from patients exposedto silica which suggests that the mitogenicactivity of epithelial cells in the BAL fluidcould at least partly be released by macro-phages. These findings are consistent with pre-vious reports which have shown that condi-tioned media from rat or human alveolarmacrophages stimulated adult rat type II cellDNA synthesis.20 The BAL fluid from normalrats induces DNA synthesis in adult rat type IIcells.2' A recent study has characterised thenature of the alveolitis in the lower respiratorytract of patients with pneumoconiosis. Alveolarmacrophages from subjects exposed to asbes-tos or silica spontaneously release mitogenicactivity for lung fibroblasts and fibronectin, animportant component of the extracellularmatrix. '" In a recent study of patients with coalworkers'pneumoconiosis the production ofPDGF, IGF-1, and transforming growth factorP(TGF-0) was demonstrated in alveolar mac-rophage supernatants and the role of PDGFand IGF-1 was associated with fibroblastgrowth.22

Partial characterisation of these mitogenicactivities for type II cells was performed andsize exclusion chromatography showed fourdistinct peaks of growth promoting activity inalveolar macrophage supernatants exposed tosilica. Most of the four distinct peaks of activitywere lost with trypsin treatment, suggesting aproteinic nature. The first and second peaks ofapparent molecular weights of 38 kDa and 22kDa, respectively, were resistant to acid andheat and were also partially inhibited by theanti-PDGF IgG. These results suggest thatthese factors could be PDGF-like molecules orthat they are closely related to IGF-1 mol-ecules. IGF-1 is an 18 kDa mediator impli-cated in mineral dust diseases and shares simi-larities with PDGF or IGF-I molecules.6 1518Antiserum against IGF-1 partially inhibitedthe activity of the first and second peaks, andthe combination of anti-PDGF with anti-IGF-1 was synergistic in blocking the activityof the second peak. The third peak of type IIcell mitogenic activity has an apparent molecu-lar weight of 16 kDa and was sensitive to acidand heat and was considerably reduced by pre-incubation with antiserum against FGF. Thisactivity could be an FGF-like molecule or acleavage product of another peptide. Thefourth activity peak has a small molecularweight of 8 kDa which was also acid-sensitiveand partially reduced by preincubation withantiserum against IGF-1. In this area the 7kDa bloodstream form of IGF-1 could beimplicated,'8-23 or other peptides which areknown to induce fetal type II cell growth couldbe suspected such as EGF and TGF-x.9 24The nature of the mitogenic activities for

epithelial cells is not completely characterisedby these conventional methods because of dif-ficulties with the induction of secretion ofgrowth factors by alveolar macrophages in theartificial environment of cell culture and thedifferences between species. Furthermore, the

785

on 24 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.8.781 on 1 A

ugust 1996. Dow

nloaded from

Melloni, Lesur, Bouhadiba, Cantin, Martel, Begin

activities of combinations of cytokines are

complex and cannot be predicted from theactivity of each individual cytokine.'"PDGF has been identified as a cationic pro-

tein with an apparent molecular weight of28-35 kDa. It is known to behave as twodistinct peptide chains (chains A and B) thatare synthesised in high molecular weightprecursors and are processed before secretion.On the other hand, PDGF-like activities are

detected in vitro in conditioned media fromvarious cells such as macrophages. There are

many differences in the molecular weights ofPDGF isoforms. Monomeric forms outline theexistence of a growing family of relatedcytokines, the so called PDGFs. The partialpurification of PDGF-like activity can beattributed to the presence of different PDGFisoforms. The characterisation may also beconfused by the presence of circulating bindingproteins which can modify the apparentmolecular weight of these factors.23 25 Hence,purification of IGFs reveals a larger molecularweight in relation to the combination of IGFsto their binding proteins. IGF-1 is a peptide of7.6 kDa in serum and 18 kDa in macrophagesupernatants, but the highest molecular weightof 38 kDa corresponds with the IGF bindingprotein.25 The existence of IGF bindingproteins in conditioned media could explainour results. Secretion of growth factors in vivomay be driven by mediators secreted by othercells such as lymphocytes which participate inthe inflammatory response to silica,26 but thesemediators may not be present in in vitroexperiments with macrophages in BAL fluid.Nevertheless, the results of our study are inagreement with previous data that have shownthat alveolar macrophages exposed to mineraldust release cytokine-like activity.'3-15 22

In conclusion, this study has confirmed thatmitogenic factors for type II cell growth in vitroare produced by macrophages exposed tosilica. Macrophages activated by silica phago-cytosis could play a key regulatory role inalveolar epithelial repair. Of the known growthfactors, PDGF-like, IGF-1-like, and FGF-likemolecules could be implicated in the regula-tion of type II cell growth after lung injury.Further investigations are required to analysethis non-specific response to silica at a molecu-lar level.

This study was supported by MRC Canada.

1 Rennard SE, Bitterman PB, Crystal RG. Response of thelower respiratory tract to injury. Mechanisms of repair ofthe parenchymal cells of the alveolar wall. Chest 1984;6:735-9.

2 Miller BE, Dethloff LA, Hook GER. Progression of type IIcell hypertrophy and hyperplasia during silica-inducedpulmonary inflammation. Lab Invest 1987;57:546-54.

3 Schuyler MR, Gaumer HR, Stankus RP, Kaimal J,Hoffmann E, Salvaggio JE. Bronchoalveolar lavage in sili-cosis. Evidence of type II cell hyperplasia. Lung 1980;157:95-102.

4 Davis GS. Pathogenesis of silicosis: current concepts andhypotheses. Lung 1986;164:139-54.

5 Scheule RK, Holian A. Immunologic aspects of pneumo-coniosis. Exp Lung Res 1991;17:661-85.

6 Rom WN. Relationship of inflammatory cell cytokines todisease severity in individuals with occupational inorganicdust exposure. Am lInd Med 1991;19:15-27.

7 Begin R, Cantin A, Boileau R, Bisson G. Spectrum of alveo-litis in quartz-exposed human subjects. Chest 1987;92:1061-7.

8 Lesur 0, Cantin A, Tanswell AK, Melloni B, Beaulieu JF,Begin R. Silica exposure induces cytotoxicity and prolif-erative activity of type II pneumocytes. Exp Lung Res1992;18:173-90.

9 Lesur 0, Melloni B, Cantin A, Begin R. Silica-exposed lungfluids have a proliferative activity for type II epithelial cells.A study on human and sheep alveolar fluids. Exp Lung Res1992;18:633-54.

10 Melloni B, Lesur 0, Cantin A, Begin R. Silica-exposedmacrophages release a growth-promoting activity for typeII pneumocytes. 7 Leukoc Biol 1993;53:327-35.

11 Melloni B, Lesur 0, Bouhadiba T, Cantin A, R Begin. Par-tial characterization of the proliferative activity for fetallung epithelial cells produced by silica-exposed alveoalarmacrophages. .7 Leukoc Biol 1993;55:574-80.

12 Begin R, Masse S, Rola-Plesczynski M, Martel M, Desma-rais Y, Geoffroy M, et al. Aluminum lactate treatmentalters the lung's biological activity of quartz. Exp Lung Res1986;10:385-99.

13 Bauman MD, Jetten AM, Bonner JC, Kumar RK, BennetRA, Brody AR. Secretion of a platelet-derived growth fac-tor by rat alveolar macrophages exposed to particulates invitro. EurJt Cell Biol 1990;51:327-34.

14 Kumar RK, 0 Grady R, Li W, Velan GM. Mitogenic activ-ity for fibroblasts induced by silica and titanium dioxideparticles in vitro and in vivo. Int7Exp Pathol 1992;73:573-83.

15 Rom WN, Bitterman PB, Rennard SI, Cantin A, CrystalRG. Characterization of the lower respiratory tract inflam-mation of nonsmoking individuals with interstitial lungdisease associated with chronic inhalation of inorganicdusts. Am Rev Respir Dis 1987;136:1429-34.

16 Rom WN, Basset P, Fells GA, Nukiwa T, Trapnell BC,Crystal RG. Alveolar macrophages release an insulin-likegrowth factor 1-type molecule. _7 Clin Invest 1988;82:1685-93.

17 Post M, Smith BT. Histochemical and immunocytochemi-cal identification of alveolar type II epithelial cells isolatedfrom fetal rat lung. Am Rev Respir Dis 1988;137:525-30.

18 Kelley J. Cytokines of the lung: state of the art. Am RevRespirDis 1990;141:765-88.

19 Takemura T, Rom WN, Ferrans VJ, Crystal RG. Morpho-logic characterization of alveolar macrophages fromsubjects with occupational exposure to inorganic particles.Am Rev Respir Dis 1989;140:1674-85.

20 Leslie CC, McCormick-Shannon K, Robinson PC, MasonRJ. Stimulation of DNA synthesis in rat alveolar type IIcells. Am Rev Respir Dis 1985;132:1246-52.

21 Leslie CC, McCormick-Shannon K, Mason RJ. Bronchoal-veolar lavage fluid from normal rats stimulates DNA syn-thesis in rat alveolar type II cells. Am Rev Respir Dis 1989;139:360-6.

22 Vanhee D, Gosset P, Wallaert B, Voisin C, Tonnel AB.Mechanisms of fibrosis in coal workers' pneumoconiosis.Am 7 Respir Crit Care Med 1994;150:1049-55.

23 King RJ, Jones MB, Minoo P. Regulation of lung cell prolif-eration by polypeptide growth factors. Am _7 Physiol 1989;257:L23-38.

24 Stiles AD, Smith BT, Post M. Reciprocal autocrine andparacrine regulation of growth of mesenchymal and alveo-lar epithelial cells from fetal lung. Exp Lung Res1986;11:165-77.

25 Sara VR, Hall K. Insulin-like growth factors and their bind-ing proteins. Physiol Rev 1992;70:591-614.

26 Li W, Kumar RK, O'Grady R, Velan GM. Role oflymphocytes in silicosis: regulation of secretion ofmacrophage-derived mitogenic activity for fibroblasts. IntJ Exp Pathol 1992;73:793-800.

786

on 24 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.8.781 on 1 A

ugust 1996. Dow

nloaded from