c-myc amplification and expression in newly established ...derlying the formation of this neoplasm...

[CANCER RESEARCH 47, 3808-3814, July 15, 1987)

c-myc Amplification and Expression in Newly Established Human OsteosarcomaCell Lines1

l-.mil Bogenmann,2 Homa Moghadam, Yves A. DeClerck, and Allan Mock

Division of Hematology-Oncology, Children! Hospital of Los Angeles, Los Angeles, California 90027

ABSTRACT

Three cell lines were isolated from a patient with osteosarcoma of thefemur. These lines were obtained from the primary neoplasm before(HTLA145) and after (HTLA161) chemotherapy and from a metastasisof the lung (HTLA19S) in the same patient. The three cell lines exhibiteda similar morphology in culture and formed tumors in nude mice whichdemonstrated a histopathology similar to that which had been observedin the patient. High expression of the genes coding for the a-1 and a-2chain of collagen Type I was found in vitro and in s.c. tumors growing innude mice. The c-myc protooncogene was amplified in all three cell linesand extensive expression of c-myc was found in vitro and in vivo. Noheterogeneity in regard to c-myc expression in vivo was detected by insitu localization in tumors growing in nude mice.

INTRODUCTION

Osteogenic sarcoma is a primary malignant tumor of thebone. It often has a "spindle cell" morphology and produces

malignant osteoids (1). The tumor originates mostly in areas ofmaximal bone growth such as the distal femur, the proximaltibia, and the proximal humÃ©rusand its potential to metastasizeearly is responsible for its poor prognosis. Many pathologicalvariants of osteosarcoma with various metastatic potentials andprognosis have been described (2-4).

Studies of human osteosarcoma have been hampered by thedifficulty to isolate and propagate malignant cells in culture.Therefore only few cell lines have been established (5, 6). Thesecell lines have been mainly used to study sensitivity to chemotherapy drugs and their metabolism (7-9).

Restriction fragment length polymorphism analysis of humanosteosarcoma demonstrated a similar genetic mechanism underlying the formation of this neoplasm as observed in retino-blastoma (10). Analysis of oncogene expression in osteosarcoma demonstrated the presence of a message for the c-sisoncogene as well as the expression of a message for a plateletderived growth factor related protein (11, 12). Recently, a newoncogene called met was isolated from a chemically transformedhuman osteosarcoma line (13, 14).

The c-myc protooncogene (15,16) codes for a nuclear proteinand has been shown to be expressed in a variety of different celltypes. It has been implicated in various biological functionssuch as growth control (17-19), cellular transformation (20),and more recently in differentiation of leukemic cells (21-23).

Recently we developed a novel culture system which allowsfor the growth of most solid pediatrie tumors and in whichoncogene amplification and expression can be detected (24,25).Here we report the unique isolation of 3 pediatrie osteosarcomacell lines from the same patient using this in vitro system. Thecell lines were isolated from the primary tumor before and aftersystemic chemotherapy as well as from a lung metastasis. Thecells maintained their histopathological characteristics, and

Received 11/14/86; revised 3/26/87; accepted 4/2/87.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.

1This work was supported by USPHS Grant EY04950 from the National EyeInstitute and from Childrens Hospital of Los Angeles funds.

2To whom requests for reprints should be addressed.

high expression of collagen type I genes was found in vitro andin vivo. DNA amplification of the c-myc protooncogene andoverexpression of the oncogene could be demonstrated in the 3cell lines in vitro and in vivo.

MATERIALS AND METHODS

Isolation of Tumor Cells in Vitro. Tumor specimens were obtainedafter surgery and processed under sterile conditions. The tissue wasmechanically treated with a razor blade and the resulting cell suspensions were washed with Dulbecco's modified Eagle's medium (Gibco)

supplemented with 10% human serum prepared as previously described(25) centrifuged and seeded in Dulbecco's modified Eagle's medium

10% human serum, penicillin (100 units/ml; Gibco), streptomycin (100Mg/ml; Gibco), insulin (10 //n/nil: Sigma), transferrin (SO Â¿ig/ml;Sigma), and 4 mM glutamine (Gibco). Tumor cells were seeded ontoextracellular matrices produced by rat smooth muscle cells in plasticPetri dishes (Falcon Plastics, Oxnard, CA) as previously described (25,26). The medium was changed biweekly. Tumor cell colonies wereisolated after 1-3 weeks and propagated in separate dishes covered withextracellular matrices. The resulting tumor cell populations were subsequently split and parts were viably frozen. Isolated cell lines weregrown after several cell passages in heat-inactivated fetal bovine serum(Irvine Scientific, Irvine, CA) and on regular tissue culture plastic(Falcon Plastics).

RNA and DNA Isolation. A procedure which allows for a combinedisolation of RNA and DNA from the same cell population was used(27). In brief, tumor cells were grown in 100-mm Petri dishes untilsubconfluency. Cells were washed twice with PBS3 and mechanically

removed from the substrate in PBS supplemented with 10 mM EDTA.Cells were subsequently washed twice with cold PBS, resuspended inlysis buffer (0.14 M NaCl-0.01 M Tris-HCl, pH 7.4-1.5 mM MgCl2-0.5% NP40), and nuclei were spun down at 1,500 x g. The supernatant was then extracted three times with an equal mixture ofphenohchloroform/isoamylalcohol (24:1) followed by an extractiontwice with chloroform:isoamylalcohol (24:1), and the RNA was thenprecipitated overnight with ethanol, pelleted at 10,000 x g and resuspended in TE.

Nuclei were digested with 100 fig/ml proteinase K (BRL, BethesdaResearch Lab, Gaithersburg, MD)-0.1% sodium dodecyl sulfate in TE,overnight at 37Â°Cwith continuous shaking. The DNA was then ex

tracted as mentioned above, ethanol precipitated, spooled, and air dried.DNA was then resuspended in TE at 1 mg/ml concentration.

Restriction Enzyme Digestion. Genomic DNA was digested with theindicated restriction endonucleases (3 units/^g DNA) overnight asrecommended by the supplier (New England Biolabs, Beverly, MA).The following day, additional 3 units,>K DNA were added and incubated for another 4-6 h at 37Â°C.

Gel Electrophoresis and Southern Transfer. Restricted DNAs (10 Mg/lane) were fractionated by electrophoresis through 1% agarose (FMCCorporation, Rockland, ME) gel in Tris-acetate buffer, pH 8.0, overnight at 20 V. Gels were stained with ethidium bromide, photographed,and DNA was transferred to Biodyne filters (ICN Radiochemicals,Irvine, CA) using standard techniques (28). The filters were then backedfor 2 h at 80Â°Cand stored at 4Â°Cuntil use.

Northern Transfer. Total cellular RNA (30 Â¿ig/lane)was electropho-resed on a 1% agarose-formaldehyde gel according to standard techniques (29). RNA was transferred to Biodyne overnight, filters were

3The abbreviations are: PBS, phosphate buffered saline; TE, 10 mM Tris-HCl-1 mM EDTA, pH 7.4; cDNA, complementary DNA; SSC, standard saline citrate(0.1 S M sodium chloride-0.015 M sodium citrate, pH 7.4).

3808

on April 6, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

c-myc IN HUMAN OSTEOSARCOMA CELL LINES

then backed at 80Â°Cfor 2 h and stored at 4Â°Cuntil use. Control stripes

of filters were soaked in 5% acetic acid for 15 min and stained with0.04% mÃ©thylÃ¨neblue in 0.5 M sodium acetate for 5-10 min (30) in

order to determine transfer efficiency.Radiolabeling of DNA Probes. Collagen complementary DNA probes

were kindly provided by Dr. F. Ramirez (Rutgers Medical School) andused as previously described (31). The following clones were used: Hf1131, a 2.2-kilobase cDNA fragment of the pro a-2 chain of humantype I collagen (32); Hf 677, a 1.8-kilobase cDNA fragment coding forthe pro a-1 chain of the human type I collagen (33). The protooncogenec-myc (15) was obtained from American Type Culture Collection,Rockville, MD (American Type Culture Collection No. 41010).

The cDNA probes were radiolabeled with [a-32P]dCTP (3000 Ci/

mmol; Amersham, Arlington Heights, IL). Briefly, 20 Mgof plasmidwere digested with the appropriate restriction endonucleases, fractionated on a 1% low gelling agarose gel, and inserts were then labeled tospecific activities of 10* cpm/ng DNA, according to Feinberg and

Vogelstein (34).Hybridization Conditions. Filters were prehybridized overnight with

the following solution: 50% formamide; 5x SSC; Ix Denhardts (0.02%each of Ficoll, polyvinyl pyrrolidone, and bovine serum albumin); 20mM sodium phosphate, pH 6.5; 0.1% sodium dodecyl sulfate; anddenatured salmon sperm DNA (250 fig/ml) at 42Â°C.The prehybridi-

zation solution was discarded, and the filters were then hybridized for3 days with 32P-labeled DNA probes using the same hybridization

conditions, including 10% dextran sulfate (Sigma) in the hybridizationsolution. Filters were then washed under stringent conditions (0.3xSSC; 65Â°C)and exposed to Kodak XAR-5 films at -80"C.

In Situ Hybridization. In situ hybridization of DNA probes to tissueculture cells and histolÃ³gica! sections was done according to Hafen etal. (35) and Breborowicz and Tomaoki (36). Briefly cells were grownon 13-mm glass coverslips for 4-6 days, washed with PBS, and fixedwith 4% paraformaldehyde in PBS at room temperature for 15 min.Cells were then treated with 1-5 fig/ml of proteinase K in TE for 15min at 37Â°C,washed with PBS, and dehydrated in a series of ethanol.

Control cells were treated with RNase A (100 ng/ml; BoehringerMannheim) in H2O for 30 min at 37'C. Specimens were then prehybridized for l h at 42Â°Cin the following solution: 50% formamide; Ix

Denhardts (see above); 5x SSC; 0.02 M sodium phosphate, pH 6.5; 25HIMdithiothreitol (Bio-Rad, Richmond, CA); and 250 /jg/ntl denaturedsalmon sperm DNA. Specimens were then hybridized with cDNAprobes for 2 days using the same hybridization conditions. Probes werelabeled according to Feinberg and Vogelstein (34) using [a-35S]dATP(1000 Ci/mmol; Amersham) and in general, IO7cpm were hybridizedin 10-15 /il/coverslip. After hybridization, specimens were washed in5x SSC-50% formamide at room temperature, 2x SSC-50% formamide, and finally 2x SSC at room temperature. Specimens were thendehydrated in a series of ethanol, 0.3 M ammonium acetate, and coatedwith an autoradiographic emulsion (NTB-2; Eastman-Kodak, Rochester, NY). Slides were exposed at -80Â°C for 10-14 days, developed,

stained with hematoxylin-eosin, and viewed in a Zeiss standard microscope.

HistolÃ³gica! sections of nude mouse tumors were treated using thesame protocol with the exception that only 0.1-0.5 Mg/ml of proteinaseK was utilized.

Light and Electronmicroscopy. Tumor tissue was fixed with 2.5%glutaraldehyde in PBS for 2 h at 4Â°C.The specimens were then

postfixed in 1% osmium tetroxide-0. l M cacodylate, pH 7.4 for 2 h at4"( "and dehydrated in a series of ethanol. The tissue was then embeddedin epon/araldyte and polymerized at 60Â°Cfor 3 days. Thick (1 /Â¿m)and

thin (800 A) sections were made with a LKB UMIV microtome.Tumorigenicity in Nude Mice. Tumor cells from the 3 different cell

lines were injected (1 x IO6cells/mouse) into the flank of 4-week old

nude mice (Swiss BALB/c). Tumor growth was monitored biweeklyand the animals were sacrificed 2-3 months after inoculation.

RESULTS

Surgical specimens obtained from the primary tumor (femur)prior to and after chemotherapy and from a lung metastasiswere explanted onto extracellular matrices (R22C1F), andgrowing tumor cell colonies were isolated after 1-3 weeks. Thecell lines were named HTLA145, for cells from the primarytumor before chemotherapy, HTLA161, for cells from theprimary tumor after chemotherapy, and HTLA195 from a lungmetastasis. Explanted tumors initially grew in small colonieswhich easily could be distinguished from normal cells. Thetumor cells were generally large cells with extensive cytoplasmicgranulations surrounding the nucleus (Fig. 1). The nucleus ofthe tumor cells contained various numbers of nucleoli (2-6nucleoli) and often giant multinucleated cells were observed.The colonies also contained numerous round cells which werecontinuously present, suggesting active cell growth. The cellsfrom the lung metastasis (HTLA195) were loosely attached tothe substratum and various sizes of cell aggregates were routinely present.

Growth Analysis in Vitro. The growth of the 3 cell lines isshown in Fig. 2. Cells from each cell line were seeded onto 60-mm dishes at 0.2 x IO6 cells/dish, and duplicate cultures were

counted at indicated time intervals. Plating efficiency determined 24 h after seeding of the cells was best for HTLA161.The 3 cell lines initially grew at approximately the same growthrate. The final cell density per plate was similar for HTLA145and HTLA161 while the cells from the metastasis reached amuch higher cell density due to the presence of large cellaggregates seen by phase contrast microscopy.

Histopathology of Nude Mouse Xenografts. The 3 cell lineswere tested for their tumorigenicity in nude mice (Swiss BALB/c), and the histopathology of the resulting tumors was analyzed.Tumor cells (106/mouse) were injected into the flank of 4 mice/

cell line. All cell lines developed tumors with a latency periodof approximately 6 weeks. The tumors were firm and eventuallykilled the animals within 3 months. Histopathological analysisdemonstrated tightly packed tumor cells organized into islands(Fig. 3A) interspersed with bundles of connective tissue andblood vessels. The nuclei showed numerous nucleoli similar tothat which has been seen in culture. Bone formation was observed in certain areas of the tumors demonstrating the origin

Isolation and Characterization of the Cell Lines. Three tumorcell lines were isolated from a patient with osteosarcoma.

3809

Fig. 1. Phase contrast micrograph of a colony of HTLA145 cells grown onextracellular matrix which can be seen in the background. Note the large size ofthe tumor cells which contained extensive granulation in the cytoplasm andnumerous nucleoli. The cells were tightly packed and many round cells can beseen, bar, 1 mm; x 150.




KHIo.

I 3 5 7 9 II 13

DAYS IN CULTURE

Fig. 2. Growth analysis of the 3 cell lines when grown on plastic. Cells (2 x10s cells/60-mm dish) were seeded in duplicate cultures, trypsinized at indicated

time intervals, and counted in a Coulter Counter. Results are the mean of 2independent experiments. Less than 10% variation was obtained.

of the tumor. No morphological differences were observedbetween the xenografts of the various cell lines, and theirhistopathology was similar to that which was seen in the tumorstaken from the patient (Fig. 3Ã„).

Electron microscopy analysis of the nude mice tumors furtherconfirmed the nature of the neoplasm. The cells showed anextensive rough endoplasmic reticulum suggesting that thesecells were biosynthetically active (Fig. 3C). Large deposits ofglycogen were also present, and extensive extracellular matrix,composed of cross-striated bundles of collagen fibrils surrounding tumor cell islands could be seen.

The nude mice tumors were subsequently explanted in vitroand cell lines (HTLA 145SQ, HTLA 161SQ) were developed.

Expression of Collagen Genes. A major characteristic of os-teogenic sarcoma is its ability to synthesize collagen, mainlytype I which in vivo can be deposited as calcified material(bone). The expression of collagen genes (a-1 and a-2 Type I)was subsequently analyzed using human cDNA probes codingfor the pro a-1 and a-2 chains, respectively.

DNA and RNA from the 3 cell lines including the onesisolated from the nude mice tumors was isolated, analyzed bySouthern and Northern blots, and results are shown in Fig. 4.Genomic DNA, digested with EcoRl enzyme when hybridizedwith Hf 1131 (pro a-1 chain) cDNA showed a different restriction fragment pattern than when hybridized with the Hf 32 (proa-2 chain) cDNA (Fig. 4A). The pattern for the a-2 chain geneof the 3 cell lines was similar, as was the pattern of the a-1

Fig. 3. HistolÃ³gica!and ullrastructural analysis performed on tumors obtainedafter injection of HTLA14S cells into nude mice. 1, islands of tightly packedtumor cells so that their cell boundaries cannot be distinguished. These islandsare separated by bundles of extracellular matrix material and few blood vessels.Note the large nuclei containing various numbers of nucleoli, bar, 200 tun: x 500.B, histolÃ³gica! section of the primary osteosarcoma tumor fixed at time ofdiagnosis. Note the characteristic pattern of the tumor which contained areas ofosteoid formation, bar, I mm; x 100. C, electron micrograph of an osteogenicsarcoma cell. The cell contains an extensive rough endoplasmic reticulum, largeummir.ts of glycogen particles, and few mitochondria. Note that the cell mem- ibranes of 2 adjacent cells are intermingled, bar, l um: x 10,000.

3810




<

iI Â«! I

m i- in<o o>

_i -j _iS E fe

23-1

9.4-6.6-

4.4-

28S-

18S-

Â«Kl)ex 1(1)

6.6-4.4-

2.2-

28 S

18S-^^^^^ ^^^^^f ^^^^^ ^BW i**** _ ^

^ * ^^00 Â» â€”~. te_^

2(02(1)

B

(Fig. 5.-Ã•).These cells demonstrated a uniform labeling of thecollagen gene transcripts in the cytoplasm, and all analyzedcells could be scored as positive. The level of labeling in individual cells was comparable and no significant heterogeneitywas found. Similar results were obtained with histolÃ³gica! section of s.c. nude mice tumors. Tumor cells were significantlylabeled when hybridized with the a-2 chain cDNA probe (Fig.5B). The label was, however, only found in viable tumor cells.No grains were detected in stromal cells or blood vessels demonstrating that the observed extracellular matrix was activelysynthesized by the tumor cells. These results supported theprevious observation that extensive collagen type I synthesisoccurred in these tumors.

Fig. 4. Southern and Northern blot analysis of a-1 and a-2 chains genes ofcollagen T>pe I. A, DNA (10 ^g/lane) from indicated cell lines was restricted withEcoRl enzyme and hybridized with the probe for the a-1 and a-2 chains probes,respectively. The restriction fragment pattern of the a-1 chain gene is differentfrom the a-2 chain gene. There is a restriction fragment length polymorphismpresent in the a-1 chain gene in the cell line isolated from a s.c. tumor of thenude mouse (HTLA161SQ). B, total cellular RNA (30 Â»ig/lane)from the indicatedcell lines was hybridized with the a-1 and a-2 chains genes cDNA probes, andexpression of the 2 genes was found in all cell lines tested.

chain gene for the parental lines (HTLA 145, HTLA161,HTLA 195). However, the cell line isolated from the mouse s.c.tumor (HTLA161SQ) demonstrated a polymorphism for thea-1 chain gene not seen with the a-2 chain gene. Severalrestriction fragments of the a-1 chain gene present in the DNAof the parental cell line did not hybridize in the digested DNAof the HTLA161SQ cells. This result was consistently seen inindependent experiments. The same digested DNAs of thevarious different cell lines were hybridized with the a-2 chaingene probe and no restriction fragment length polymorphismwas found (Fig. 4/4). We therefore concluded that the absenceof labeled restriction fragments in the HTLA161SQ DNA ofa-1 chain gene represents a true restriction fragment lengthpolymorphism rather than a technical artifact.

Expression of the a-1 and a-2 chains genes was analyzed intotal RNA isolated from subconfluent cells (Fig. 4B). All celllines showed expression of the a-1 and a-2 chains genes whichconfirmed previous results obtained by analysis of collagenproteins secreted by the HT 145 cell line (31). The rearrangement of the a-1 chain gene seen in Southern blots (Fig. 4.1)didnot affect the expression of the a-1 chain gene in these cells(Fig. 4B); however, there was more hybridization of the messagefor the a-2 than the a-1 gene in total RNA from theHTLA161SQ cells.

In situ hybridization of cultured HTLA161 cells and s.c.tumors (HTLA161SQ) was also performed to study collagengene expression in single tumor cells in vitro and in vivo.

Subconfluent cultured tumor cells when hybridized with thea-2 chain probe showed an extensive labeling of individual cells

381

Fig. 5. In situ hybridization of the a-2 gene transcripts in cultured HTLA161cells and nude mouse tumor. Cells and tissue sections were prepared for in situlocalization as described in "Materials and Methods." A, extensive labeling of

individual cultured cells in which the autoradiographic grains are almost exclusively present in the cytoplasm of the cells. Note the homogeneous distributionof the label within the cells. All cells show active transcription of the a-2 gene.bar, 200 uni; x 450. B, in situ localization of a-2 gene transcripts in a nude mousetumor. Tumor cells demonstrated significant labeling indicating active transcription of the a-2 chain gene in neoplastic cells while the surrounding stromal andblood vessel cells as well as necrotic areas did not. bar, 200 fmi; x 500.

1




c-myc Amplification and Overexpression in Vitro and in Vivo,c-myc oncogene has been implicated in the growth control ofnormal and neoplastic cells and was found to be amplified in amouse osteosarcoma line infected with polyoma virus (37). Wehave therefore analyzed the protooncogene c-myc in the newlyisolated cell lines. DNA and RNA isolated from the variousparental and nude mouse tumors derived cell lines was probedwith a 0.9-kilobase fragment of the first exon of the c-myc gene,and results are shown in Fig. 6. Restricted DNA from thevarious osteosarcoma lines clearly demonstrated a significant(10-20 times) stronger signal for the c-myc EcoRl restrictionfragment than that which was observed with either normalhuman kidney DNA or retinoblastoma DNA isolated fromcultured cells (Fig. 6/4). Amplification of the c-myc gene wasobserved in all cell lines, including the ones obtained from thenude mice tumors. No amplification was seen in osteosarcomacell DNA when hybridized with a probe for the N-myc gene

w ,- m $< < < <

OE-BQfi

<

ce

12 kb

B

iu> Ss IO

I si I

2.3 kb

c-MYC ^Fig. 6. Southern and Northern blot analysis of the c-myc oncogene in the

various cell lines. A, DNA ( 10 Mg/lane) from the indicated osteosarcoma cell linesas well as from normal human kidney and cultured retinoblastoma cells (RBLA13,RBLA18) was digested with EcoRl and subsequently hybridized with the c-mycprotooncogene. An approximate 12-kilobase (kb) restriction fragment showed a10- to 20-fold stronger labeling in the various osteosarcoma cell lines whencompared to normal DNA or retinoblastoma DNA, suggesting c-myc amplification. B, Southern blot of A was subsequently hybridized with the N-myc probe todemonstrate that approximately equal amounts of digested DNA was presentedin the various lanes. C, total cellular RNA (30 pg/lane) from the various osteosarcoma lines including a human neuroblastoma line (IMR32) was hybridized. Astrong expression of the c-myc protooncogene with an approximate 2.3-kilobasemessage was found in the osteosarcoma cells from the parental lines as well asfrom the nude mouse tumor cell lines. No such expression was seen in theneuroblastoma cell line known to express the N-myc gene.

(Fig. 6Ã„). Similarly, total cellular RNA showed significantexpression of a 2.3-kilobase message for the c-myc gene in theosteosarcoma cell lines but not in a human neuroblastoma line(IMR32) known to overexpress the N-myc gene (Fig. 6C). The2.3-kilobase c-myc message corresponded to that which haspreviously been described (15).

In situ hybridization of c-myc transcripts in cultured cells(HTLA145) and in tumors from the nude mice was also performed (Fig. 7). Almost every cell analyzed in culture waslabeled, demonstrating a uniform expression of the c-myc genein the tumor cell population (Fig. 1A). Similarly, nude micetumors when analyzed in histological section clearly demonstrated tumor cells actively transcribing the c-myc gene. Noexpression of the protooncogene was found in stromal cells orblood vessel cells (Fig. IB). Analysis performed in differentareas of the tumors did not reveal any heterogeneity in regardto levels of c-myc expression again suggesting a uniform expression of the protooncogene c-myc in nude mice tumors derivedfrom the isolated cell lines.

DISCUSSION

The present study extends our previous work with primarypediatrie tumors explanted in vitro. In this paper we describethe isolation of 3 osteosarcoma cell lines from the same patientusing the previously described in vitro system (24, 25). Thesystem allows not only for maintenance of explanted tumorsbut also enables the isolation and development of cell linesfrom rare childhood neoplasms. The successful isolation of 3cell lines from a patient with osteosarcoma opens new opportunities to study this neoplasm with regard to various biologicalfunctions including molecular events underlying its development.

Cellular heterogeneity within a tumor is a known biologicalphenomenon which may have important clinical relevance fortreatment and prognosis. The 3 cell lines were developed fromsmall specimens from the patients malignancy and thereforemay have represented highly selected cell populations. The celllines however demonstrated similar cell morphology in vitroand in vivo (nude mice) comparable to that which was observedin the primary tumor indicating that representative specimenswere obtained from the original neoplasm. The tumorigenicpotential of the isolated cell lines demonstrates the neoplasticorigin of the cells and contrasts the nontumorigenic behaviorof the TE85 line previously developed (5).

The synthesis of collagen, mainly Type I, is a major characteristic of osteosarcoma cells and the presence of calcifiedcollagen (bone) is often used as diagnostic criterion (1). Wepreviously demonstrated that primary pediatrie tumors whenexplanted in vitro synthesize and secrete mature collagen chains(38), and the expression of extracellular matrix proteins hasbeen used to characterize human tumor cell lines in vitro (39,40). Collagen biosynthetic profiles, previously determined incultured HTLA145 cells, showed the presence of substantialamounts of collagen Type I (a-\ and a-2 chains) secreted intothe medium (31). The expression of collagen genes variedbetween the different cell lines in culture. These differences inthe amount of collagen mRNA present could be due to differences in message stability in these individual tumor cell linesor as a result of culture conditions at the time of cell harvest.Extensive expression of the pro a-1 and a-2 chains genes forcollagen type I in the osteosarcoma cell lines substantiated thenature of the tumor cells isolated and confirmed the presenceof osteoids in the s.c. nude mice tumors. Analysis of collagen

3812




Fig. 7. In situ localization of c-myc transcripts in cultured HTLA145 cells anda nude mouse tumor. A, actively transcribing HTLAI4S cells in vitro afterhybridizing with the 35S-labeled c-myc probe, bar, 200 Â»im;x 450. B, histolÃ³gica!section of a s.c. HTLA145 tumor hybridized with the c-myc probe. Tumor cellsdemonstrate specific labeling while stromal and blood vessel cells and necroticareas did not show any incorporated labeled probe. Note the uniform distributionof the label in individual neoplastic cells, bar, 200 (.m; x 500.

gene expression in s.c. nude mice tumors using in situ hybridization strongly suggests that the extensive extracellular matrixwithin the tumor was synthesized and secreted by the tumoritself rather than by stromal cells surrounding the neoplasm.

The presence of a restriction fragment length polymorphismin the pro a-1 chain gene in the cell line isolated from s.c. nudemouse tumor is of importance. Such an event could have takenplace during tumor formation in the nude mouse or as a resultof in vitro selection. Genomic rearrangement within humantumors passaged through the nude mouse has not yet beenreported and needs to be analyzed in more detail. However,polymorphism within other structural genes such as the <>-tubulin gene and the vimentin gene (data not shown) could notbe detected in HTLA161SQ cells, and therefore pro a-1 generearrangement represents a random event.

The observation that the c-myc protooncogene was substan

tially amplified in the 3 cell lines isolated adds an additionalmalignancy to the list of human tumors with c-myc involvement(15, 41-44). c-myc amplification and expression was found ina mouse osteosarcoma cell line after superinfection with poly-oma virus (37); however, no report was made with humanosteosarcoma malignancies. The substantial expression of c-myc in all 3 cell lines might have facilitated the growth of thesecells under in vitro conditions and therefore might be responsible for successful establishment of the cell lines, since theprotooncogene c-myc has been implicated in various biologicalfunctions including growth control (17, 19).

c-myc amplification could also have resulted from in vitro

selection. It seems, however, highly unlikely that all 3 cell lineswould independently develop such gene amplification, andtherefore c-myc amplification must have occurred during tumorprogression. It is, however, important to notice that c-mycamplification was also found in cells from the lung metastasis,indicating that amplification occurred prior to dissemination.Similar results were found with small cell carcinoma of the lung(44).

Amplified genes have been shown to be compartmentedwithin double minute chromosomes or homogeneously stainedregions (45, 46). Chromosomal analysis of the 3 cell lines isnow being done to elucidate whether c-myc amplification iscontained within either double minute chromosomes or homogeneously stained regions.

c-myc transcription as seen by in situ hybridization in culturedcells as well as in tumors from the nude mice demonstrated auniform expression of the protooncogene among single cells.This result contrasts with the finding of human neuroblastomawhere extensive heterogeneity of N-myc expression has beendescribed (47). Since c-myc may be a regulatory gene for growthcontrol, tumor cells with high expression of this oncogene mayhave a growth advantage resulting in a homogeneous cell population with regard to c-myc expression.

The successful isolation of cell lines from a rare childhoodneoplasm demonstrates the validity of the in vitro system previously established (25). The cell lines for the first time demonstrate the involvement of c-myc in human osteosarcoma andtherefore represent a unique opportunity to study the role of c-myc in tumor progression.

ACKNOWLEDGMENTS

We would like to thank Dr. William Benedict for the use of the nudemouse colony and Clarita Stutters for typing the manuscript.

REFERENCES

1. Dahlin, D. C. Osteosarcoma of bone and a consideration of prognosticvariables. Cancer Treat. Rep., 62: 189-192, 1978.

2. Ghadially, F. N., and Mehta, P. N. Ultrastructure of osteogenic sarcoma.Cancer (Phila.), 25: 1457-1467, 1970.

3. (Â¿norm.A. M., Abelson, H., and Freis, E. Osteosarcoma: fifteen years later.N. Engl. J. Med., 313: 1637-1643, 1985.

4. Stark, A., Aparisi, T., and Ericsson, J. L. E. Human osteogenic sarcoma:fine structure of the osteoblastic type. Ultrastruct. Pathol., 4:311-329,1983.

5. McAllister, R. M., Gardner, M. B., Green, A. E., Bradt, C., Nichols, W. W.,and Landing, B. H. Cultivation in vitro of cells derived from a humanosteosarcoma. Cancer (Phila.), 27: 397-402, 1971.

6. Smith, S. H., Owens, R. B., Hiller, A. J., Nelson-Rees, W. A., and Johnson,J. O. The biology of human cells in tissue culture. 1. Characterization of cellderived from osteogenic sarcomas. Int. J. Cancer, 17: 219-234, 1976.

7. Abelson, H. T., and Gorka, C. Absence of hypoxanthine: guanine phospho-ribosyltransferase activity in murine dÃ¼nnosteosarcoma. Cancer Res., 4.1:4098-4101, 1983.

8. Mordoh, J., Chacon, R. D., and Filmus, J. Isolation of tumor cells frompatients with osteosarcoma and analysis of their sensitivity to methotrexate.Cancer Res., 41: 3621-3628, 1981.

3813



C-myc IN HUMAN OSTEOSARCOMA CELL LINES

9. I ln-iii, R., and Lotan, R. Sensitivity of cultured human osteosarcoma andchondrosarcoma cells to retinoic acid. Cancer Res., 42:4771-4775, 1982.

10. Dryja, T. P., Rapaport, J. M., Epstein, J., Goorin, A. M., Weichselbaum,R., Koufos, A., and Cavenee, W. K. Chromosome 13 homozygosity inosteosarcoma without retinoblastoma. Am. J. Hum. Genet., 38:59-66,1986.

11. Graves, D. T., Owen, A. J., Barth, R. K., Tempst, P., Winoto, A., Fors, L.,Hood, L. E., and Antoniades, L. E. Detection of c-si'Ãtranscripts and synthesis

of PDGF-like proteins by human osteosarcoma cells. Science (Wash. DC),226:972-974, 1984.

12. 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.), 319:511-514,1986.

13. Cooper, C. S., Park, M., Blair, D. G., Tainsky, M. A., Huebner, K., Croce,C. M., and Vande Woude, G. G. Molecular cloning of a new transforminggene from a chemically-transformed human cell line. Nature (Lond.), Ml:29-33, 1984.

14. Park, M., Dean, M., Cooper, C. S., Schmidt, M., O'Brien, S. J., Blair, D.

G., and Vonde Woude, G. F. Mechanism of met oncogene activation. Cell,45: 895-904, 1986.

15. Alitalo, K., Schwab, M., Lin, C. C., Varmus, H. E., and Bishop, J. M.Homogeneously staining chromosomal regions contain amplified copies ofan abundantly expressed cellular oncogene (c mir) in malignant neuroendo-crine cells from a human colon carcinoma. Proc. Nati. Acad. Sci. USA, XII:1707-1711, 1983.

16. Hann, S., Abrams, H., Rohrschneider, C., and Eisenmann, R. N. Proteinsencoded by \-myc and c-myc oncogenes: identification and localisation inacute leukemia virus transformation and bursal lymphoma cell lines. Cell,34:789-798, 1983.

17. Campisi, J., Gray, H. E., Pardee, A. B., Dean, M., and Sonnenschein, G. E.Cell-cycle control of c-myc but not c-ros expression is lost following chemicaltransformation. Cell, 36: 341-347, 1984.

18. Hayward, W., Neel, B., and Astrin, S. Activation of a cellular one gene bypromoter insertion of ALV-induced lymphoid leukosis. Nature (Lond.), 290:475-480, 1981.

19. Kelly, K., Cochran, B., Sliles, C., and Leeds, P. Cell-specific regulation ofthe c-myc gene by lymphocyte mitogens and platelet-derived growth factor.Cell, 55:603-610, 1983.

20. Klein, G., and Klein, E. Conditioned tumorigenicity of activated oncogenes.Cancer Res., 46: 3211-3224, 1986.

21. Bentley, D. C., and Groudine, M. A block to elongation is largely responsiblefor decreased transcription of c-myc in differentiated HL 60 cells. Nature(Lond.), 321: 702-706, 1986.

22. Lachman, H. M., and Skoultchi, A. I. Expression of c-myc changes duringdifferentiation of mouse erythroleukemia cells. Nature (Lond.), 310: 592-594, 1984.

23. Westin, E., Wong-Stall, F., Gelmann, E. P., Della Pavera, R., Papas, T. S.,Cantenberger, J. A., Eva, A., Reddy, E. P., Tronick, S. R., Aaaronson, S. A.,and Gallo, R. C. Expression of cellular homologues of retroviral one genesin human hematopoietic cells. Proc. Nati. Acad. Sci. USA, 79: 2490-2494,1982.

24. Bogenmann, E., and Mark, C. Routine growth and differentiation of primaryretinoblastoma cells in culture. J. Nati. Cancer Inst., 70:95-104, 1983.

25. Bogenmann, E., Mark, C., Isaacs, H., Neustein, H. B., DeClerck, Y. A.,Laug, W. E., and Jones, P. A. Invasive properties of primary pediatrieneoplasms in vitro. Cancer Res., 43: 1176-1186, 1983.

26. Jones, P. A., Scott-Burden, T., and Gevers, W. Glycoprotein, elastin andcollagen secretion by rat smooth muscle cells. Proc. Nati. Acad. Sci. USA,76:353-357, 1979.

27. Lee, A. S., Delegeane, A. M., Baker, V., and Chow, P. C. Transcriptional

regulation of two genes specifically induced by glucose starvation in a hamstermutant fibroblast cell line. J. Biol. Chem., 25Â«:597-603, 1983.

28. Southern, E. M. Selection of specific sequences among DNA fragmentsseparated by gel electrophoresis. J. Mol. Biol., 98: 503-517, 1975.

29. Lehrach, H., Diamond, D., Wozney, J. M., and Boechker, H. RNA molecularweight determinations by gel electrophoresis under denaturing conditions, acritical re-examination. Biochemistry, 16:4743, 1977.

30. Maniatis, T., Fritsch, E. F., and Sambrook, J. Molecular Cloning. ColdSpring Harbor, NY: Cold Spring Harbor Laboratory, 1982.

31. Chandler, L. A., DeClerck, Y. A., Bogenmann, E., and Jones, P. A. Patternsof DNA methylation and gene expression in human tumor cell lines. CancerRes., 46: 2944-2948, 1986.

32. Myers, J. C., Chu, M. L., Faro, S. H., Clark, W. J., Prockop, D. J., andRamirez, F. Cloning a cDNA for the pro-2 chain of human type I collagen.Proc. Nati. Acad. Sci. USA, 78: 3516-3520, 1981.

33. Chu, M. L., Myers, J. C., Bernard, M. P., Ding, J. F., and Ramirez, F.Cloning and characterization of five overlapping cDNAs specific for thehuman pro- 1(1) collagen chain. Nucleic Acids Res., 10: 5925-5934, 1982.

34. Feinberg, A. P., and Vogelstein, B. A. A technique for radiolabeling DNArestriction endonuclease fragments to high specific activity. Anal. Biochem.,137: 266-267, 1984.

35. Hafen, E., Levine, M., Garber, R. L., and Gehring, W. J. An improved insitu hybridization method for the detection of cellular RNAs in Drosophilatissue sections and its application for localizing transcripts of the homeoticantennapedia gene complex. EMBO J., 2:617-623, 1983.

36. Breborowicz, J., and Tamaoki, T. Detection of messenger RNAs of a-fetoprotein and albumin in a human hepatoma cell line by in situ hybridization. Cancer Res., 45: 1730-1736, 1985.

37. Schwab, M., Ramsay, G., Alitalo, K., Varmus, H. E., Michael Bishop, J.,Martinsson, T., Levan, G., and Levan, A. Amplification and enhancedexpression of the c-myc oncogene in mouse SEW A tumour cells. Nature(Lond.), 315: 345-347, 1985.

38. DeClerck, Y. A., Bogenmann, E., and Jones, P. A. Collagen synthesis byshort term expiants of pediatrie tumors. Cancer Res., 45: 1229-1238, 1985.

39. Alitalo, K., Keski-Oja, J., and Vaheri, A. Extracellular matrix proteinscharacterize human tumor cell lines. Int. J. Cancer, 27: 755-761, 1981.

40. Alitalo, K., Kurkinen, M., Vaheri, A., Virtanen, I., Rhode, H., and Timpl,R. Basal lamina glycoproteins are produced by neuroblastoma cells. Nature(Lond.), 287:465-466, 1980.

41. Rothberg, P. G., Erisman. M. D.. Diehl, R. E.. Rovigatti, U. G., and Astrin,S. M. Structure and expression of the oncogene c-myc in fresh tumor materialfrom patients with hematopoietic malignancies. Mol. Cell. Biol., 4: 1096-1103, 1984.

42. Saksela, K., Bergh, J., Lehto, V-P., Nilsson, K., and Alitalo, K. Amplificationof the c-myc oncogene in a subpopulation of human small cell lung cancer.Cancer Res., 45: 1823-1827, 1985.

43. Trent, J., Meltzer, P., Rosenblum, M., Harsh, G., Kinzler, K., Mashal, R.,Feinberg, A., and Vogelstein, B. Evidence for rearrangement, amplificationand expression of c-myc in human glioblastoma. Proc. Nati. Acad. Sci. USA,Â«3:470-473, 1986.

44. Wong, A. J., Ruppert, J. M., Eggleston, J., Hamilton, S. R., Baylin, S. B.,and Vogelstein, B. Gene amplification of c-myc and N-myc in small cellcarcinoma of the lung. Science (Wash. DC), 233: 461-464, 1986.

45. Kinzler, K. W., Zehnbauer, B. A., Brodeur, G. M., Seeger, R. C., Trent, J.M., Meltzer, P. S., and Vogelstein, B. Amplification units containing humanN-myc and c-myc genes. Proc. Nati. Acad. Sci. USA, Â«3:1031-1035, 1986.

46. Schimke, R. T. Gene amplification in cultured animal cells. Cell, 37: 705-713, 1984.

47. Schwab, M., Ellison. J., Rusch, M., Rosenau, W., Varmus, H. E., and Bishop,J. M. Enhanced expression of the human gene N-myc consequent to amplification of DNA may contribute to malignant progression of neuroblastoma.Proc. Nati. Acad. Sci. USA, 81: 4940-4944, 1984.

3814



1987;47:3808-3814. Cancer Res Emil Bogenmann, Homa Moghadam, Yves A. DeClerck, et al. Human Osteosarcoma Cell Lines

Amplification and Expression in Newly Establishedmycc-

Updated version

http://cancerres.aacrjournals.org/content/47/14/3808

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/47/14/3808To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



c-myc amplification and expression in newly established ...derlying the formation of this neoplasm...

Documents