Comparison of Production of Transforming Growth Factor ... · [CANCER RESEARCH 47, 6180-6184, December 1, 1987] Comparison of Production of Transforming Growth Factor-/^ and Platelet-derived

[CANCER RESEARCH 47, 6180-6184, December 1, 1987]

Comparison of Production of Transforming Growth Factor-/^ and Platelet-derived

Growth Factor by Normal Human Mesothelial Cells and MesotheliomaCell LinesBrenda I. Gerwin,1 John F. Lechner, Roger R. Reddel, Anita B. Roberts, Keith C. Robbins, Edward W. Gabrielson,2

and Curtis C. HarrisLaboratory of Human Carcinogenesis ¡B.1. G., J. F. L., R. R. R., E. W. G., C. C. H.J, Laboratory of Chemoprevention [A. B. R.], and Laboratory of Cellular andMolecular Biology [K. C. R.J, Division of Cancer Etiology, National Cancer Institute, Bethesda, Maryland 20892

ABSTRACT

Steady state mRNA levels for transforming growth factor fi, platelet-derived growth factor (PDGF) A-chain, and PDGF B-chain were measured in normal human mesothelial cells, SV40 large T-antigen expressing

human mesothelial cells, and human mesothelioma cell lines. The mRNAexpression level for transforming growth factor ti was similar in all threetypes of culture while normal human mesothelial cells secrete moretransforming growth factor ßthan do mesothelioma cell lines or I -antigen expressing mesothelial cells. In contrast, both PDGF A- and H-

chain mRNAs are expressed at higher levels in mesothelioma cell linesthan in normal human mesothelial cells. PDGF-like mitogenic activity

was readily detectable in medium conditioned by a mesothelioma cell lineand was undetectable in conditioned medium from normal cells. Theseresults suggest the hypothesis that PDGF may be an autocrine growthfactor in mesothelioma.

INTRODUCTION

Epidemiológica! investigations indicate that mesotheliomasarise from mesothelial cells as a consequence of exposure toasbestos or other asbestiform fibers (1, 2). In our previous invitro investigations of NHM3 cells, we showed that these cells

are more sensitive than human bronchial fibroblasts or epithelial cells to the cytotoxic (3, 4) and clastogenic (5) effects ofasbestos. In several /// vitro carcinogenesis experiments whereNHM cells were exposed to amosite asbestos and then subcul-tured, nontumorigenic cells with karyotypic abnormalities rapidly became the predominant cell types. In addition, the cellsin the exposed cultures continued to replicate after the controlcultures had ceased to divide. However, it was not possible fromthe data of five experimental trials to recognize a chromosomalabnormality common to or prevalent in the asbestos-alteredcells. Similar observations have been made in Syrian hamster(6, 7) and rat mesothelial cells (8, 9). These observations suggested that mesotheliomas might arise after asbestos exposuredue to the development of a population in which chromosomalalteration conferred a growth advantage. We have, therefore,sought to define the growth requirements of NHM cells as aguide to analyzing the expression of genes which are of significance for cell growth. In the course of these studies, we havedetermined that TGF-/3 and PDGF induce quiescent NHMcells to undergo one round of DNA synthesis and, further,stimulate continuous replication of NHM cells supplied with

Received 4/28/87; revised 8/12/87; accepted 9/8/87.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1To whom requests for reprints should be addressed, at Building 37, Room

2C09, NIH, Bethesda, MD 20892.'Current address: Baltimore VA Medical Center, 3900 Loch Raven Blvd.,

Baltimore, MD 21218.3The abbreviations used are: NHM, normal human mesothelial; I (>!.(.

transforming growth factor-^; PDGF, platelet-derived growth factor; EGF, epidermal growth factor, BSA, bovine serum albumin; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; SDS, sodium lauryl sulfate; SSC, standard salinecitrate (lx SSC = 0.15 M NaCl, 0.015 M sodium citrate); kb, kilobase.

other lipoprotein requirements4 (10). It was, therefore, of interest to compare the expression of TGF-0 and PDGF by NHMcells and by human mesothelioma cell lines.

MATERIALS AND METHODS

Cells and Culture Conditions. Primary cultures of normal humanmesothelial cells were initiated from pleural fluids obtained from donorswith no known malignancies. The cell pellets were obtained by centrif-ugation at 125 x g for 5 min, rinsed in growth medium, and inoculatedat a ratio of one 100-mm dish/50-100 ml of pleural fluid. Semicon-fluent cultures were dissociated using trypsin and expanded or utilizedfor experimental protocols. Normal cells and mesothelioma cell lineswere cultured in identical growth medium consisting of LHC basalmedium (11) with hydrocortisone (4 x 10~7M), zinc-free insulin (8.7 xIO"7 M), EGF (3.3 x 10"' M), transferrin (1.25 x IO"7 M), gentamicin

(50 ng/m\), and fetal bovine serum (3%). Lux culture dishes were coatedwith a mixture of human fibronectin (10 ug/ml), Vitrogen collagen (30/jg/nil). and crystallized BSA (10 ug/m\) dissolved in LHC basal medium. A volume of this mixture sufficient to cover the surface of thedishes was added, allowed to stand at 37°Cfor at least 2 h, and removed

by aspiration. Seven established human mesothelioma cell lines wereutilized for this study. These cell lines were DND-10B from Dr. TakaoOhnuma of the Mt. Sinai School of Medicine, N.Y., NY (12); VAMT-1 from Dr. P. Calabresi of Roger Williams Hospital, Providence, RI(13); JMN from Dr. A. Behbehani of the University of Kansas Schoolof Medicine, Kansas City, KA (14); HUT 28 and HUT 290 (15) fromDr. A. Gazdar of Uniformed Services University of the Health Sciences,Bethesda, MD; and MT-1 (16) and MT-3 from Drs. T. Anderson andR. Saxton of the UCLA School of Medicine, Los Angeles, CA. Thesecell lines were cultured in the same way as normal cells after passage.Normal mesothelial cells transformed by transfection with a SV40 largeT-antigen construct (MeT 4A, MeT 5A, and MeT 8A),5 were alsocultured identically. These cell strains exhibited life spans of 60-80population doublings before crisis as compared to the normal cellswhich senesced after 15 population doublings. All T-antigen expressingcells were nontumorigenic after 1 year in athymic nude mice.5

RNA Isolation, Electrophoresis, and Hybridization. RNA was isolatedaccording to the procedure of Chirgwin et al. (17) except that RNAsolutions after CsCl centrifugation were extracted with phenol andchloroform before precipitation with LiCl. Concentrations were determined spectrophotometrically and RNA was denatured by boiling in2.2 M formaldehyde, 50% formamide, and then electrophoresedthrough denaturing 2.2 M formaldehyde, 1.2% agarose gels, and transferred to Gene Screen Plus (Dupont/NEN) nylon membranes by capillary blotting in lOx SSC (18). RNA was immobilized by UV cross-linking for 3 min at 254 nm. Filters were prehybridized for 6 h andthen hybridized with nick-translated complimentary DNA inserts at42°Covernight in 1 M NaCl, 10% dextran sulfate, 1% SDS, 100 ug/m\

denatured salmon sperm DNA, and 50% formamide. Filters werewashed twice at room temperature in 2x SSC, twice at 65"C in 2x

SSC, 1% SDS, and finally, twice at room temperature in 0.1 x SSC.Membranes were then exposed to Kodak Xoiu.it XAR5 Him. Whenreprobing, filters were first boiled for 20 min in 0.1 x SSC-1% SDS toremove specifically hybridized fragments. Probes utilized for this study

4J. F. Lechner, unpublished observations.5 R. Reddel, manuscript in preparation.

6180

on April 5, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TGF-0, PDGF IN HUMAN MESOTHELIAL AND MESOTHELIOMA CELLS

were all fragments of human clones: for TGF-/3, the X/3C, insert described by Derynck et al. (19) and supplied by Dr. Rik Derynck,Genentech, Inc., San Francisco, CA; for c-sis, a human 2.0-kb BamHlfragment, containing all the B chain coding sequences and for PDGFA chain a 1.3-kb EcoRl fragment from the human clone Dl (20). Thehuman PDGF A chain and B chain (c-sis) clones were kindly suppliedby Dr. Christer Betsholtz, University of Uppsala, Uppsala, Sweden.The GAPDH clone used to quantitate RNA expression levels was a ratclone of a constitutive gene (21) kindly provided by Dr. Marc Piechac-zyk, Laboratoire de Biologie Moléculaire,Montpellier, France. In orderto determine expression ratios, representative autoradiograms for thespecified probes were scanned on a Shimadzu CS930 densitometer.Areas for each band were calculated and compared to areas for theGAPDH band in the same lane on the same filter. These ratios areexpressed as arbitrary numbers and have not been normalized to anabsolute concentration standard.

Determination of TGF-0 in Conditioned Medium. Medium containingneither EGF nor fetal calf serum was conditioned for 24 h by cells,collected, and immediately frozen until processed further. BSA wasadded to the thawed medium to a final concentration of 100 ng/ml. Itwas then dialyzed against l M acetic acid using ,\f, 3500 cut-off tubing(Spectrapor; Spectrum Medical Industries, Los Angeles, CA), lyophi-lized to dryness and redissolved at 20-50x in 4 HIMHC1 containing 1mg/ml BSA; this procedure activates all latent TGF-/3 in the medium(22). To determine the proportion of the TGF-0 that was in a latentform, dialysis was carried out against water and the sample was redissolved in neutral Dulbecco's modified Eagle's medium containing 4

mg/ml BSA; aliquots were then assayed as such or first acidified withHC1 (to a final concentration of ISO mMHCl) and reneutralized beforeassay.

T( ¡1-'-,(in the concentrated medium was determined both by a competitive radioreceptor binding assay using [I2sl]-labeled human plateletTGF-0 and AS49 human lung carcinoma cells (23) and by induction ofcolony formation of normal rat kidney cells in soft agar in the presenceof EGF (24). In both assays, TGF-0 was quantitated by comparison of4-6 point dilution curves of the samples with dilution curves of humanplatelet TGF-/3 standard assayed at the same time.

Determination of PDGF-like Activity in Conditioned Medium. NHM(86-70) or HUT 28 cells were grown as described, washed in phosphate-buffered saline, and incubated for 24 h in medium containing no EGFor fetal calf serum. Conditioned medium was harvested from 4.2 x 10'NHM cells and 9.3 x IO6HUT 28 cells and concentrated by Centricon

filtration so that 1 ml of concentrate represented medium conditionedby 9.3 x IO7 cells. Conditioned medium was tested for its ability to

induce DNA synthesis in quiescent NIH/3T3 fibroblasts as described(25).

RESULTS

Expression of TGF-/8. For the experiments described here,total RNA was extracted from three populations of cells: (a)primary cultures of NHM cells; (¿>)human mesothelioma celllines; and (c) NHM cell strains expressing the SV40 large T-antigen gene.5 The last class of cells was utilized because they

are nontumorigenic but exhibit a greatly extended in vitro lifespan as compared to NHM cells, thus providing a model for anintermediate stage in tumor development. Northern blots containing RNA from these cell cultures were hybridized sequentially to probes for TGF-ßand the constitutively expressedprotein GAPDH (21). Fig. 1 shows autoradiograms representative of these studies. In all cells, when probing with TGF-0complimentary DNA, a single mRNA species was observed atthe expected size of 2.5 kb (19). In addition, a smaller specieswhich has not been characterized, was sometimes observed.Thus, all three categories of mesothelial cell cultures clearlyexpress TGF-/9 mRNA. An expression ratio of TGF-/3 mRNAto GAPDH mRNA was calculated for each cell as described in"Materials and Methods" and arbitrarily normalized to the

T-Normal antigen Mesothelioma

'l 2 3 4M5 6 ?' 8 91011121314

TGF-/3

GAPDH

Fig. 1. Expression of TGF-0 mRNA in nonna! or T-antigen expressing humanmesothelial cells and in mesothelioma cell lines. Total cellular RNA (20 fig/lane)was denatured and analyzed by Northern blotting as described in "Materials andMethods." The filter was hybridized first for TGF-0, stripped, and rehybridized

for GAPDH. Expression ratios are shown in Table 1. Lanes correspond to thefollowing cultures: 7, 86-156; 2. 86-160; 3, P251; 4, 86-127; 5, MeT-4A; 6, MeT-5A; 7, MeT-8A; 8, DND; 9, JMN; 10, MT-3; 11, MT-1; 12, VAMT-1; 13,HUT290; 14, HUT28.

Table 1 Expression of PDGF and TGF-ßmRNAs relative to a constitutive probe

Autoradiograms were analyzed by densitometry and ratios for expression ofthe major band of the indicated mRNA relative to that for GAPDH mRNA werecalculated from different exposures of the filters shown in Figs. 1 and 2.

CelltypeNHM"P25186-12786-15686-16886-5386-6586-75T-antigen

NHMMeT^»AMeT-5AMeT-8AMesotheliomaMT-31*DNDJMNMT-1VAMT-1HUT290HUT28TGF-/321516238102———151163219100272372149794464>834PDGF-A174e—187—3081010017158314914161PDGF-BND*ND——ND—ND113191301003785730>891ND>159

" Since these represented primary cultures, sufficient material was not available

to make all tests on the same cells. Several primary cultures were tested for eachprobe as a partial measure of interindividual variation.

* Not detectable: in every case, no H chain mRNA was detected (Fig. 211).' Not tested.d Ratios were normalized to the MT-3 cell line as 100% since values obtained

for this cell line were always within the linear densitometric range. Actual valuesfor 100% were TGF-/3 = 0.47; PDGF-A = 5.91; PDGF-B = 0.23.

ratio obtained for the mesothelioma cell line MT-3 which wasaccurately quantitated in all measurements (Table 1). Theseexpression ratios for individual cell strains or cell lines havebeen consistent in replicate experiments, although there is greatvariation from cell to cell, particularly for tumor cell lines. Themean values for expression of TGF-0 relative to GAPDH donot differ greatly among these human mesothelial cells andmesotheliomas (Fig. 1; Table 1). The mean values obtained forboth NHM cells [129 ±66 (SD)] and T-antigen transformedNHM cells (178 ±30) are not significantly different. While themean value for the mesotheliomas was much higher (430 ±285), the variation was much wider than in the other two classesof cells.

Measurement of TGF-/S in Conditioned Medium. Medium

6181

on April 5, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TGF-0, PDGF IN HUMAN MESOTHELIAL AND MESOTHELIOMA CELLS

conditioned by the three classes of cells was prepared as described in "Materials and Methods" and assayed for TGF-ß

activity by quantitation of its ability to compete for receptorbinding with a standard of highly purified TGF-0. This assaywas performed after acidification of the medium and is thereforea measure of all inactive and active species secreted by thesecells (22). Variations in cell numbers arose from characteristically different growth rates. The results obtained for the TGF-ßassays are shown in Table 2. As expected from the analysisof TGF-/3 message, all of the cells tested secrete TGF-/3 intotheir medium. However, these data, when compared on thebasis of cell number suggest that NHM cells produce significantly more TGF-/S than either human mesothelioma cell linesor T-antigen transformed NHM cells. In order to test whetherNHM differ from mesothelioma cultures with respect to secretion of active TGF-0, media conditioned by one NHM primaryculture (P222) and one mesothelioma cell line (VAMT-1) were

compared by assaying both with and without acidification asdescribed in "Materials and Methods." The values obtainedindicated that both types of cultures secreted some active TGF-ß(10% or 3.2 pM for NHM; 8% or 9.0 pM for mesothelioma)but that there was no significant difference between the normaland the tumor cells in this respect. Thus, for TGF-0, the levelsof steady-state mRNA production are similar for NHM, T-antigen transformed, and mesothelioma cultures while secretion of TGF-0 protein per cell was greatest in primary NHMcultures.

Expression of PDGF A and B Chain mRNA. Total cellularRNA from the same three cell types was further probed forexpression of PDGF A- or B chain in comparison to GAPDH(Fig. 2, A and B). In this case, a marked difference is observedbetween expression by NHM cells and mesotheliomas. ForPDGF A-chain, a low level of mRNA is detected in the NHMcells and in two of three T-antigen transformed NHM isolates.In contrast, the mesothelioma cell lines which express A-chainmessage do so at a much higher level than the normal cells. Inagreement with visual inspection, calculated ratios listed inTable 1 show that the steady state expression of PDGF A-chainmRNA relative to GAPDH is at one level for both NHM cellsand their T-antigen expressing derivatives but that, in general(five of seven), the mesotheliomas express significantly higherlevels of the PDGF A-chain message. The mean value obtainedfor mesotheliomas (90 ±59) is significantly higher than thatfound in NHM cells (11 ±6) or T-antigen modified cells (16 ±10).

It was also of interest to compare the expression of PDGF

Table 2 TGF-ßin media conditioned by mesothelial cells and mesotheliomas

Medium conditioned for 24 h by cells was collected and processed as describedto activate any latent TGF-/3. Amounts of TGF-/3 were determined by a receptorcompetition assay as described in "Materials and Methods."

A T-Normal antigen Mssothelioma

123456789 1011121314

B T"Normal Iantigen I Mesolhelioma

1 2 3 4l5 6 ili 91011121314

CelltypeNHM86-7586-8186-83T-antigen

NHMMeT-»AMeT-5AMeT-8AMesolheliomaDNDHUT

28HUT290JMNVAMT-1Cells

(10«)3.53.55.214.58.522.46.09.037.732.826.5PMTGF-011012015085741804853310420260TGF-/3(10-*),no. ofcells3134296988681310

"'I

GAPDH •.

Fig. 2. Expression of PDGF A- or B-chain mRNA in normal or T-antigenexpressing human mesothelial cells and in mesothelioma cell lines. Analysis wasperformed as described for Fig. 1. The initial probe was for the PDGF A-chainin 2A or for PDGF B-chain (c-sis) in IB. Lanes correspond to the followingcultures: A: I, 86-127; 2, 86-53; 3, P251; 4, 86-65; 5, MeT-4A; 6, MeT-5A; 7,MeT-8A; 8, DND; 9, JMN; JO, MT-3; //, MT-1; 12, VAMT-1; lì,HUT290;14, HUT28. Lanes for B correspond to those for A except for lanes 4, 86-75,10,MT-1, and //, MT-3.

B-chain mRNA (Fig. 2B) to that of PDGF A-chain (Fig. 2A).The PDGF B-chain message was not detectable in the fourNHM cultures shown in Fig. 2B; we have been unable to detectany B-chain mRNA in a total of 10 primary cultures derivedfrom 10 individual donors. In contrast, three T-antigen expressing cell lines derived from NHM cells express detectableamounts of PDGF B-chain mRNA as do five of the sevenhuman mesothelioma cell lines. The expression values of thethree classes of cells appear to be different although the expression levels shown by the mesothelioma cell lines vary widely(Table 1). For example, one cell line, HUT 290, makes nodetectable B-chain message while another, VAMT-1, expressesat a level at least nine times higher than the MT-3 line. It isclear, however, from inspection of Fig. 2 and Table 1 that allhuman mesothelioma cell lines express either PDGF A- or B-chain mRNA and that six of seven produce both.

Presence of PDGF-like Mitogenic Activity in Medium Conditioned by Mesothelioma Cells. To determine whether cells whichexpressed PDGF mRNAs produced PDGF-like activity, medium conditioned for 24 h by the mesothelioma cell line HUT28 or by NHM cells (86-70) was tested for its ability to stimulatepHjthymidine incorporation by quiescent NIH 3T3 cells. Asshown in Fig. 3, medium conditioned by NHM cells does notcontain detectable PDGF activity. The curve described by theNHM data points (•)corresponds to 40 x IO3 cpm, the

background level for this assay. In contrast, the medium conditioned by the mesothelioma cells possessed readily detectablemitogenic activity approximately equivalent to 150 pg PDGF/ml. Recently an IgG preparation capable of potently inhibitingPDGF but not EGF or fibroblast growth factor mitogenicactivity has been described (26). This antibody inhibited themitogenic activity present in HUT 28 conditioned medium asindicated by an arrow to the open circles (O) on Fig. 3. Thisinhibition identifies the mitogen as PDGF-like. This findingstrongly suggests that one or both PDGF mRNAs expressedby HUT 28 cells are translated.

DISCUSSION

The results presented here clearly demonstrate that normalhuman mesothelial cells synthesize and secrete both active andinactive species of TGF-/3 (Fig. 1; Tables 1 and 2). Severalconclusions can be derived from these analyses. Comparison ofsteady-state message production and total TGF-/3 protein secreted by NHM cells, T-antigen-transformed NHM cells, andmesothelioma cell lines makes it clear that in this cell system,

6182

on April 5, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TGF-iä,PDGF IN HUMAN MESOTHELIAL AND MESOTHELIOMA CELLS

250

200

oi 15°ocu 100

50

ANTI-PDGF ANTISERUMNO ADDITIONS

. l

10 15 20

¡niCONDITIONED MEDIUMFig. 3. Production of PDGF-like activity by a human mesothelioma cell line.

Media conditioned by equal numbers of HUT28 (O) or NHM (•)cells weretested for the presence of PDGF-like activity. NIH/3T3 cells (5 x HP/well) wereplated in 96-well culture dishes and grown to confluence without medium change.Test samples were added alone or with antisera where specified and 16 h latercultures were supplemented with [3H]thymidine (20 Ci/mmol/well; New EnglandNuclear). After S h additional incubation, material insoluble in trichloroaceticacid was measured by scintillation counting. Wells were supplemented with 5 nganti-PDGF IgG (26) together with the addition of test samples. Data representthe average of three separate determinations. Background values are 40,000 cpmand have not been subtracted.

as has also been observed for MCF-7 cells (27), there is noconcordance between steady-state mRNA and protein levels forTGF-/3. In the case of estrogen-inhibitable MCF-7 cells, changeswere observed in levels of active TGF-/3 secreted by the cellswith no corresponding changes in steady-state levels of themRNA transcript (27). Such data, as well as evidence for thepresence of TGF-/3 and its receptors in most cell types make itclear that posttranslational control mechanisms are of criticalimportance for the role of TGF-/3 as a regulator of cell growth.

In this regard, the observation that NHM cells have receptorsfor TGF-/3 (28) and that the active molecule is a mitogen forthese cells (10) suggested that production of TGF-/S by mesothelioma cell lines might provide an autocrine function for thismolecule in tumor growth. However, when conditioned mediafrom NHM and mesothelioma cultures were compared, it wasclear that the tumor cell lines actually produce a lower level ofTGF-/Õper cell than do NHM cultures. Additionally, a normaland mesothelioma culture produced equivalent percentages ofactive TGF-/3. It seems unlikely, therefore, that growth stimulation by TGF-/3 alone could be a decisive factor in the generation of mesothelioma.

Our laboratory has recently found that PDGF from humanplatelets is mitogenic for normal human mesothelial cultures(10) and that these cells have receptors for this growth hormone.4 PDGF from human platelets is composed of two relatedpolypeptide chains, PDGF-A and PDGF-B which are the products of two distinct genes (20, 29-31). PDGF-like activity hasbeen associated with B chain (32-37) and A-chain (38, 39)homodimers as well as heterodimers (40). The data presentedin this report (Fig. 2; Table 1) indicate that the steady-stateexpression level of both the A- and B chain gene is elevated inmesothelioma cell lines as compared to NHM cultures. Thedata of Fig. 3 suggest that this increase is reflected in thesecretion of PDGF activity into the culture medium. Theseobservations do not define the expressed protein as A- or B-chain since the antiserum utilized in Fig. 3 is specific for PDGFas opposed to fibroblast growth factor or EGF and would notdistinguish between A- and B-chain molecules (26). However,

in the human osteosarcoma cell line U-2 OS, which, like mostmesothelioma cell lines expresses both A- and B-chain mRNAtranscripts (20, 41, 42), PDGF-like activity in culture mediumhas been shown to result from A-chain homodimers (38). Incontrast, v-sis encoded PDGF B-chain homodimer as well asthe normal human B-chain product has been shown to be cellassociated (43). Thus, in analogy with these systems, it seemsmost likely that the PDGF-like activity in mesothelioma conditioned medium represents A-chain homodimers. Since noPDGF-like activity was detected in medium conditioned bynormal cells, this secretion of mitogenic activity may be ofimportance in the pathogenesis of mesothelioma.

The present studies show that NHM cells do not expressdetectable levels of PDGF-B mRNA, whereas the majority ofmesothelioma cell lines and NHM cells expressing SV40 largeT antigen contain readily measurable steady-state levels of thistranscript. Previous studies have indicated that expression ofthe normal human PDGF-B gene induces transformation ofmurine NIH/3T3 fibroblasts (44) and that antibodies to PDGFinhibit both proliferation and simian sarcoma virus-inducedmorphological changes (45). Thus, it is clear that the productof the B-chain gene when expressed in a cell for which PDGFis stimulatory can contribute to the oncogenic transformationof that cell. In this respect, the finding of B-chain expressionin mesotheliomas combined with the observations that NHMcells are responsive to PDGF is consistent with involvement ofthe PDGF B-chain independently of or in addition to the A-chain in the genesis of mesothelioma.

Any suggestion that mesothelioma may be generated by autocrine mechanisms (46) must consider the relationship of suchmechanisms to the role of asbestos, which has been establishedas a complete carcinogen in the generation of this tumor (1,2).It has been shown that asbestos is clastogenic for NHM cells(4). The tumor lines used for the present study have beenkaryoiyped." Although multiple abnormalities are present in

these aneuploid cells, no consistent abnormality associated withmesothelioma can be defined. Thus, observation of human cellsmodified by asbestos in vitro (4) of primary human mesotheliomas (47) and of mesothelioma cell lines shows that the tumorcells have abnormal karyotypes but does not suggest a specificlesion which could be associated with mesothelioma. Neitherchromosome 7 nor 22, the locations of the PDGF A- (20) andB-chain (30) genes shows consistent cytogenetic alteration.However, inability to detect genetic changes by light microscopydoes not eliminate the possibility that alterations exist whichare detectable only at the molecular level. Furthermore, alterations at other genetic loci might affect the transcription orprocessing of the specific messages or the translation andsecretion of the final protein products of the genes. Thus, itseems likely that genetic abnormalities generated by asbestosdamage to these sensitive cells (3, 4) may lead to a populationof cells which gains a growth advantage by autocrine secretionof growth factors and/or causes overgrowth by paracrine secretion to responsive mesothelial cells. The data presented heresuggest that PDGF is one growth factor which may have a rolein tumor development by this model.

ACKNOWLEDGMENTS

The authors thank Drs. T. Ohnuma, P. Calabresi, A. Behbehani, A.Gazdar, T. Anderson, and R. Saxton for supplying the human mesothelioma cell lines. We are also grateful to Drs. R. Derynck, C.Betsholtz, and M. Piechaczyk for giving us constructs with which to

' B. I. Gerwin, R. R. Redell, and J. F. Lechner, unpublished data.

6183

on April 5, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TGF-/3, PDGF IN HUMAN MESOTHELIAL AND MESOTHELIOMA CELLS

probe for specific mRNAs. The assistance of Monica Shimmin in thepreparation of this manuscript is much appreciated.

REFERENCES

1. Craighead, J. E., and Mossman, B. T. The pathogenesis of asbestos-associated diseases. N. Engl. J. Med. 306: 1446-1455, 1982.

2. Kannerstein, M., and Churg, J. Mesothelioma in man and experimentalanimal. Environ. Health Perspect., 34: 31-36, 1980.

3. Lechner, J. F., Tokiwa, T., Yeager, H. Jr., and Harris, C. C. Asbestos-associated chromosomal changes in human mesothelial cells. In: E. G. Beckand J. Bignon (eds.), In Vitro Effects of Mineral Dusts, NATO ASI Series,Vol. G3, pp. 197-202. Berlin: Springer-Verlag, 1985.

4. Lechner, J. F., Tokiwa, T., LaVeck, M., Benedict, W. F., Banks-Schlegel, S.,Yeager, H. Jr., Banerjee, A., and Harris, C. C. Asbestos-associated chromosomal changes in human mesothelial cells. Proc. Nati. Acad. Sci. USA, 82:3884-3888, 1985.

5. Linnainmaa, K., Gerwin, B., Pelin, K., Jantunen, K., LaVeck, M., Lechner,J. F., and Harris, C. C. Asbestos-induced mesothelioma and chromosomalabnormalities in human mesothelial cells in vitro. In: Third Joint US-FinnishScience Symposium. The Changing Nature of Work and Workforce, pp.119-122. Cincinnati, OH: NIOSH, 1986.

6. Hesterberg, T. W., and Barrett, J. C. Induction by asbestos fibers of anaphaseabnormalities: mechanism for aneuploidy induction and possibly carcinogen-esis. Carcinogenesis (Lond.), 6: 473-475, 1985.

7. Hesterberg, T. W., Hutu-rick, C. J., Oshimura, M., Brody, A. R., and Barrett,J. C. Role of phagocytosis in Syrian hamster cell transformation and cyto-genetic effects induced by asbestos and short and long glass fibers. CancerRes., 46: 5795-5802, 1986.

8. Jaurand, M. C., Kheuang, L., Magna, L., and Bignon, J. Chromosomalchanges induced by chrysotile fibres or benzo-3,4-pyrene in rat pleuralmesothelial cells. Mutât.Res., 769: 141-148, 1986.

9. Paterour, M. J., Bignon, J., and Jaurand, M. C. In vitro transformation ofrat pleural mesothelial cells by chrysotile fibres and/or benzo[a)pyrene.Carcinogenesis (Lond.), 6: 523-529, 1985.

10. Gabrielson, E. W., Lechner, J. F., Gerwin, B. I., LaVeck, M. A., and Harris,C. C. Growth factors for mesothelial and mesothelioma cells. Chest, 91:17S,1987.

11. Lechner, J. F., and LaVeck, M. A. A serum-free method for culturing normalhuman bronchial epithelial cells at clonal density. J. Tissue Cult. Methods,9:43-48,1985.

12. Holland, J. F. Establishment and characterization of a human mesothelioma.Proc. Am. Assoc. Cancer Res., 20:62, 1979.

13. Tibbetts, L. M. Loss and restoration of tumorigenicity in a human mesothelioma cell line. Proc. Am. Assoc. Cancer Res., 25:45. 1984.

14. Behbehani, A. M. Studies of a human mesothelioma. Hum. Pathol., /3:862-866, 1982.

15. Carney, D. N., Gazdar, A. F., Bepler, G., Guccion, J. G., Marangos, P. J.,Moody, T. W., Zweig, M. H., and Minna, J. D. Establishment and identification of small cell lung cancer cell lines having classic and variant features.Cancer Res., 45: 2913-2923, 1985.

16. Anderson, T. M., Holmes, E. C., Kosaka, C. J., Cheng, L., and Saxton, R.E. Monoclonal antibodies to human malignant mesothelioma. J. Clin, luimunol., 7: 254-261, 1987.

17. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. G., and Rutter, W. J.Isolation of biologically active ribonucleic acid from sources enriched inribonuclease. Biochemistry, 18: 5294-5299, 1979.

18. Maniatis, T., Fritsch, E. F., and Sambrook, J. Molecular Cloning (A Laboratory Manual), pp. 202-203. Cold Spring Harbor, NY: Cold Spring HarborLaboratory, 1982.

19. Derynck, R., Jarrett, J. A., Chen, F. Y., Eaton, D. H., Bell, J. R., Assoian,R. K., Roberts, A. B., Sporn, M. B., and Goeddel, D. V. Human transforminggrowth factor-beta complementary DNA sequence and expression in normaland transformed ceils. Nature (Lond.), 316:701-705, 1985.

20. Betsholtz, C., Johnsson, A., Heldin, C.-H., Westermark, B., Lind, P., Urdea,M. S., Eddy, R., Shows, T. B., Philpott, K., Mellor, A. L., Knott, T. J., andScott, J. cDNA sequence and chromosomal location of human platelet-derived growth factor A-chain and its expression in tumour cell lines. Nature(Lond.), 320:695-699, 1986.

21. Fort, P., Marty, L., Piechaczyk, M., El Sabrouty, S., Dani, C., Jeanteur, P.,and Blanchard, J. M. Various rat adult tissues express only one major mRNAspecies from the glyceraldehyde-3-phosphate-dehydrogenase multigenic family. Nucleic Acids Res., 13: 1431-1442, 1985.

22. Lawrence, D. A., Pircher, R., Kryceve-Martinerie, C., and Jullien, P. Normalembryo fibroblasts release transforming growth factors in a latent form. J.Cell. Physiol., 121: 184-188, 1984.

23. Frolik, C. A., Wakefield, L. A., Smith, D. M., and Sporn, M. B. Characterization of a membrane receptor for transforming growth factor-beta in normalrat kidney fibroblasts. J. Biol. Chem., 259: 10995-11000, 1984.

24. Roberts, A. B., Anzano, M. A., Lamb, L. C., Smith, J. M., and Sporn, M.B. New class of transforming growth factors potentiated by epidermal growth

factor: isolation from non-neoplastic tissues. Proc. Nati. Acad. Sci. USA, 78:5339-5343, 1981.

25. Leal, F., Williams, L. T., Robbins, K. C., and Aaronson, S. A. Evidence thatthe \ •vi'.vgene product transforms by interaction with the receptor for platelet-

derived growth factor. Science (Wash. DC), 230: 327-330, 1985.26. Igarashi, H., Rao, C. D., Siroff, M., Leal, F., Robbins, K. C., and Aaronson,

S. A. Detection of PDGF-2 homodimers in human tumor cells. Oncogene,/: 79-85, 1987.

27. Knabbe, C., Lippman, M. E., Wakefield, L. M., Flanders, K. C., Kasid, A.,Derynck, R., and Dickson, R. B. Evidence that transforming growth factor-beta is a hormonally regulated negative growth factor in human breast cancercells. Cell, «.-417-428, 1987.

28. Wakefield, L. M. An assay for type beta transforming growth factor receptor.In: D. Barnes, D. A. Sirbasku, S. P. Colowick, and N. O. Kaplan, (eds.),Peptide Growth Factors; Methods in Enzymology, Vol. 146, pp. 167-173.Orlando: Academic Press, Inc., 1987.

29. Doolittle, R. F., Hunkapiller, M. W., Devare, S. G., Robbins, K. C., Aaron-son, S. A., and Antoniades, H. N. Simian sarcoma virus oncogene v-sis isderived from the gene (or genes) encoding a platelet-derived growth factor.Science (Wash. DC), 221: 275-277, 1983.

30. Swan, D. C., McBride, O. W., Robbins, K. C., Keithley, D. A., Reddy, E. P.,and Aaronson, S. A. Chromosomal mapping of the simian sarcoma virusoncogene analogue in human cells. Proc. Nati. Acad. Sci. USA, 79: 4691-4695, 1982.

31. Waterfield, M. D., Scrace, G. T., Whittle, N., Stroobant, P., Johnsson, A.,Wasteson, A., Westermark, B., Heldin, C.-H., Huang, J. S., and Deuel, T.F. Platelet-derived growth factor is structurally related to the putative transforming protein p28 sis of simian sarcoma virus. Nature (Lond.), 304: 35-39, 1983.

32. Bowen-Pope, D. F., Vogel, A., and Ross, R. Production of platelet-derivedgrowth factor-like molecules and reduced expression of platelet-derivedgrowth factor receptors accompany transformation by a wide spectrum ofagents. Proc. Nati. Acad. Sci. USA, 81: 2396-2400. 1984.

33. Deuel, T. F., Huang, J. S., Huang, S. S., Stroobant, P., and Waterfield, M.D. Expression of a platelet-derived growth factor-like protein in simiansarcoma virus transformed cells. Science (Wash. DC), 221:1348-1350,1983.

34. Garrett, J. S., Coughlin, S. R., Niman, H. L., Tremble, P. M., Giels, G. M.,and Williams, L. T. Blockade of autocrine stimulation in simian sarcomavirus-transformed cells reverses down-regulation of platelet-derived growthfactor receptors. Proc. Nati. Acad. Sci. USA, «7:7466-7470, 1984.

35. Johnsson, A., Betsholtz, C., von der Helm, K., Heldin, C.-H., and Westermark, B. Platelet-derived growth factor agonist activity of a secreted form ofthe v-sis oncogene product. Proc. Nati. Acad. Sci. USA, 82: 1721-1725,1985.

36. Robbins, K. C., Antoniades, H. N., Devare, S. G., Hunkapiller, M. W., andAaronson, S. A. Structural and immunological similarities between simiansarcoma virus gene product(s) and human platelet-derived growth factor.Nature (Lond.), 305:605-608, 1983.

37. Owen, A. J., Pantazis, P., and Antoniades, H. N. Simian sarcoma virus-transformed cells secrete a mitogen identical to platelet-derived growth factor.Science (Wash. DC), 225:54-56, 1984.

38. Heldin, C.-H., Johnsson, A., Wennergren, S., Wernstedt, C., Betsholtz, C.,and Westermark, B. A human osteosarcoma cell line secretes a growth factorstructurally related to a homodimer of PDGF A-chains. Nature (Lond.), 3/9:511-514, 1986.

39. Sejersen, T., Betsholtz, C., Sjolund, M., Heldin, C.-H., Westermark, B., andThyberg, J. Rat skeletal myoblasts and arterial smooth muscle cells expressthe gene for the A chain but not the gene for the H-chain (c-sis) of platelet-derived growth factor (PDGF) and produce a PDGF-like protein. Proc. Nati.Acad. Sci. USA, «3:6844-6848, 1986.

40. Johnsson, A., Heldin, C.-H., Wasteson, A., Westermark, B., Deuel, T. F.,Huang, J. S., Seeburg, P. H., Gray, A., Ullrich, A., Scrace, G., Stroobant,P., and Waterfield, M. D. The c-sis gene encodes a precursor of the B-chainof platelet-derived growth factor. EMBO J., 3: 921-928, 1984.

41. Betsholtz, C., Westermark, B., Ek, B., and Heldin, C.-H. Coexpression of aPDGF-like growth factor and PDGF receptors in a human osteosarcoma cellline: implications for autocrine receptor activation. Cell, 39:447-457, 1984.

42. Graves, D. T., Owen, A. J., Barth, R. K., Tempst, P., Winoto, A., Fors, L.,Hood, L. E., and Antoniades, H. N. Detection of c-sis transcripts andsynthesis of PDGF-like proteins by human osteosarcoma cells. Science(Wash. DC), 226:972-974, 1984.

43. Robbins, K. C., Leal, F., Pierce, J. H., and Aaronson, S. A. The v-sis/PDGF-2 transforming gene product localizes to cell membranes but is not a secretoryprotein. EMBO J., 4: 1783-1792, 1985.

44. Gazit, A., Igarash, I. C., Srinivasan, A., Yaniv, A., Tronick, S. R., Robbins,K. C., and Aaronson, S. A. Expression of the normal human sis/PDGF-2coding sequence induces cellular transformation. Cell, 39: 89-97, 1984.

45. Johnsson, A., Betsholtz, C., Heldin, C.-H., and Westermark, B. Antibodiesagainst platelet-derived growth factor inhibit acute transformation by simiansarcoma virus. Nature (Lond.), 3/7: 438-440, 1985.

46. Sporn, M. B., and Todaro, G. J. Autocrine secretion and malignant transformation of cells. N. Engl. J. Med., 303: 878-880, 1980.

47. Gibas, Z., Li, F. P., Animan, K., Bernal, S., Stahel, R., and Sandberg, A. A.Chromosome changes in malignant mesothelioma. Cancer Genet. Cytoge-net., 20: 191-201, 1986.

6184

on April 5, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1987;47:6180-6184. Cancer Res   Brenda I. Gerwin, John F. Lechner, Roger R. Reddel, et al.   Cells and Mesothelioma Cell LinesPlatelet-derived Growth Factor by Normal Human Mesothelial

andβComparison of Production of Transforming Growth Factor-

  Updated version

  http://cancerres.aacrjournals.org/content/47/23/6180

Access the most recent version of this article at:

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  [email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/47/23/6180To request permission to re-use all or part of this article, use this link

on April 5, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from