Characterization of a Human Ovarian Carcinoma Cell Line ...cancerres.aacrjournals.org/content/canres/43/11/5379.full.pdf · Characterization of a Human Ovarian ... hormonal influence

[CANCER RESEARCH 43, 5379-5389, November 1983]

Characterization of a Human Ovarian Carcinoma Cell Line (NIH:OVCAR-3)1

with Androgen and Estrogen Receptors

Thomas C. Hamilton, Robert C. Young, Wilma M. McKoy, Karen R. Grotzinger, John A. Green, Elizabeth W. Chu,Jacqueline Whang-Peng, Alfred M. Rogan, William R. Green, and Robert F. Ozols2

Medicine Branch, Division of Cancer Treatment [T. C. H., R. C. Y., W. M. M., K. R. G., J. A. G., J. W.-P., A. M. R., R. F. 0.], and Laboratory ot Pathology, Division of

Cancer Biology and Diagnosis [E. W. C., W. R. G.], National Cancer Institute, Bethesda, Maryland 20205

ABSTRACT

A cell line, NIH:OVCAR-3, has been established from the

malignant ascites of a patient with progressive adenocarcinomaof the ovary after combination chemotherapy with cyclophospha-mide, Adriamycin, and cisplatin. OVCAR-3 grows as a cobblestone-like monolayer with foci of multilayering, is tumorigenic in

athymic mice, clones in agarose, and has an abnormal karyotypewhich includes a homogeneous staining region and a doubleminute chromosome. The cultured cells and xenografts containcytoplasmic androgen- and estrogen-binding macromolecules

with the specificity of the respective steroid hormone receptors.These components have sedimentation coefficients of 7 to 9S inlow-salt sucrose-density gradients, have dissociation constants

of 250 and 9.6 pM, and are present at concentrations of 30 and28 fmol/mg cytosol protein characteristic of androgen and estrogen receptors, respectively. OVCAR-3 is resistant in vitro toclinically relevant concentrations of Adriamycin (5 x 10~8 M),melphalan (5 x 10~6 M), and cisplatin (5 x 10~7 M) with survival

compared to untreated controls of 43,45, and 77%, respectively.Furthermore, there are multiple histological similarities betweenthe patient's original tumor, the cell line, and the transplantable

tumor.These data indicate that OVCAR-3 may be of use for investi

gations as to the significance of androgens and estrogens andthe mechanisms of cytotoxic drug resistance in ovarian cancer.

INTRODUCTION

The search for new therapeutic modalities in ovarian cancer,including ways to increase the effectiveness of chemotherapy,and investigations into the possible role of hormones in thebiology and treatment of ovarian cancer have been limited, inpart, by the lack of suitable experimental models of the humandisease. Relevant experimental systems are needed in which themajor clinical problem of rapid emergence of primary drug resistance and associated cross-resistance (58) can be directly

investigated. In addition, although there has been an increasedinterest in the potential relevance of hormones in the disease(16) due to the frequent identification of steroid hormone receptors in human ovarian cancer specimens (2, 11, 12, 17, 22-24,

28, 40, 45) and to continued reports of variable successes ofhormonal therapy (2, 13, 32, 44), experimental confirmation ofhormonal influence on the growth of ovarian cancer cells likewiseawaits the development of appropriate model systems.

We have studied previously a murine model of ovarian cancer(34) as well as the direct cloning of human ovarian cancer cells

1NIH:OVCAR-3 refers to NIH ovarian carcinoma cell line 3.2To whom requests for reprints should be addressed.

Received April 28,1983; accepted July 27,1983.

in a double-layer agar system (35, 36). Pharmacological studies

in these systems have provided some support for new therapeutic approaches in ovarian cancer, such as i.p. chemotherapy (37).However, neither of these systems was satisfactory for biochemical investigation of drug resistance and hormonal effects onovarian cancer cells. The murine model was derived from aspontaneous teratoma, and, although the pattern of mÃ©tastasesis similar to that of human ovarian cancer, biological differencesbetween murine teratoma cells and human epithelial cancer cellslimit the clinical relevance of therapeutic studies in this experimental system. Although the direct cloning of human ovariancancer cells has provided clinically relevant information about thepatterns of drug resistance, the small size of tumor colonies(usually 30 to 50 cells) does not allow for direct biochemicalinvestigations on the mechanisms of drug resistance. Accordingly, we have turned to human ovarian cancer cell lines asmodels with which to study endocrine relationships in ovariancancer and the pharmacology of antineoplastic drugs.

Ovarian cancer cell lines can potentially overcome the experimental limitations inherent in both the murine mode! of ovariancancer and the direct agar cloning of human ovarian cancerspecimens. Cell lines may allow for the direct measurement ofhormone effects on malignant cell proliferation and provide invitro correlation between the presence of steroid hormone receptors and the specific hormone response. In addition, cell linesshould be useful for the investigation of clinically relevant biological modifiers, such as MÃ¼llerianinhibition substance (7) andmonoclonal antibodies (1). The availability of adequate quantitiesof proliferating malignant cells would also facilitate the identification of potentially clinically useful dose-response relationships

with antineoplastic drugs as well as a model system in which tocompare the cytotoxicity of drug analogues.

One of the more clinically relevant studies with human celllines relates to the mechanisms of drug resistance and cross-

resistance. Ovarian cancer is clinically noted for the rapid development of primary drug resistance and a broad cross-resistance

(58). It would be of interest to determine if the mechanisms ofcross-resistance in human ovarian cancer are similar to those

reported in animal tumor cell lines, particularly with regard to theinduction of certain membrane glycoproteins (41). Finally, resistant cell lines may be useful to determine the mechanism ofresistance to alkylating agents, anthracyclines, and cisplatin andto investigate ways in which this primary resistance can bemodulated.

Although there have been several human ovarian cancer celllines described in the literature (6, 10, 27, 57), many of thesehave not been fully characterized. In particular, there have beenno previous reports on the presence of steroid hormone receptors in any ovarian cancer cell line, nor have there been any dataon the in vitro sensitivity of these cell lines to the standard

NOVEMBER 1983 5379

on June 23, 2018. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

7. C. Hamilton et al.

chemotherapeutic agents used in the treatment of ovarian cancer. We have established a new human ovarian cancer cell linewhich contains cytoplasmic androgen and estrogen receptor-likemacromolecules. This line was derived from the malignant as-cites of a patient with progressive papillary adenocarcinoma ofthe ovary after combination chemotherapy. In addition to thesteroid receptor status, this line has been characterized withregard to morphology, ultrastructure, karyotype, growth characteristics in vitro (including colony formation in soft agarose),tumorigenicity in athymic nude mice, and in vitro sensitivity tothe chemotherapeutic drugs to which the patient was clinicallyresistant. Furthermore, histological and ultrastructural comparisons are made between the cell line, the nude mouse tumor,and the original specimen.

MATERIALS AND METHODS

Chemicals and Reagents

[2,4,6,7-3H]Estradiol (90 to 115 Ci/mmol), 17/3-hydroxy-1 7a-methyles-

tra-4,9,11-trien-3-one(methyltrienolone, R 1881), and [17a-mefhy/-3H]R

1881 (70 to 87 Ci/mmol) were purchased from New England Nuclear,Boston, Mass. Unlabeled 5a-dihydrotestosterone, diethylstilbestrol, progesterone, and 9a-fluoro-11fi,21-dihydroxy-16tt,17a-isopropylidene-dioxy-1,4-pregnadiene-3,20-dione (triamcinolone acetonide) were ac

quired from Sigma Chemical Co., St. Louis, Mo. 1-(4-f)-Dimethylamino-ethoxyphenyl)-1,2-diphenylbut-1-ene (tamoxifen) was given by Stuart

Pharmaceuticals (Division of ICI Americas, Inc., Wilmington, Del.). Mel-phalan, Adriamycin, and cisplatin were obtained from the PharmaceuticalResources Branch, Division of Cancer Treatment, National Cancer Institute, Bethesda, Md.

RPM1 1640,3 fetal bovine serum, penicillin.-streptomycin stock solution

(10,000 units/ml; 10,000 Mg/ml), and PBS were from Grand IslandBiological Co., Grand Island, N. Y. EGF and insulin were from Collaborative Research, Inc., Waltham, Mass. Agarose (type VII) was fromSigma, and lymphocyte separation medium was from Bionetics Laboratory Products, Kensington, Md. All other chemicals and reagents wereof the highest purity commercially available.

Tissue Culture Medium

Medium was prepared volumetrlcally from RPMI 1640 powder usingdouble-glass-distilled water. Unless otherwise stated, the formulation

used for growth and passage of cells contained RPMI 1640 at unitconcentration, heat-inactivated fetal bovine serum (2%, v/v), sodium

bicarbonate (24 nriM), insulin (10 Mg/ml), penicillin (100 units/ml), andstreptomycin (100

Initiation and Propagation of the Cell Line

Malignant ascites from a patient with poorly differentiated papillaryadenocarcinoma of the ovary (Fig. 1) collected in heparin (1.0,000 units/liter ascites) was the source of cells used to initiate the OVCAR-3 cellline. The patient developed ascites 8 months after diagnosis and treatment with chemotherapy which included cyclophosphamide, Adriamycin,and cisplatin. The cytological characteristics of the ascites were thoseof malignant cells and consistent with the primary tumor.

Cells were obtained from the ascites by centrifugation at 300 x g for10 min at 4Â°.The cell pellet was then washed once in growth medium,

and the suspension was placed in tissue culture T-flasks. Incubation wasat 37Â° in a humidified atmosphere of 5% (v/v) CO2 in air, and medium

3The abbreviations used are: RPMI 1640, Roswell Park Memorial Institute

Tissue Culture Medium 1640; PBS, phosphate-buffered saline without calcium andmagnesium, but containing KCI (0.2 g/liter), KH2PO4 (0.2 g/liter), Na2HP04 â€¢7H2O(2.2 g/liter), and NaCI (8 g/liter); EGF, epidermal growth factor; OVCAR-3,NIH:OVCAR 3; DM, double minute chromosome; HSR, homogeneously stainingregion.

was changed at intervals of 3 to 4 days. In primary culture, 2 morphologically distinct cell types were observed by phase-contrast microscopy.The minor component consisted of elongated fibroblastoid cells whichwere readily removed by selective treatment with a PBS solution whichcontained trypsin (0.05%. w/v) and EDTA (0.02%, w/v) (33). Subcultureof the remaining epithelioid cells, at split ratios of 1:2 or 1:3, wasaccomplished with trypsin:EDTA (as above); the suspension of detachedcells was diluted with an equal volume of growth medium and centrifugedat 300 x g for 10 min at 4Â°.After centrifugation, the cell pellet was

suspended in growth medium and added to tissue culture flasks. Subculture was at 10- to 14-day intervals.

Tests for Mycoplasma

OVCAR-3 was tested for Mycoplasma by culture on Mycoplasma

agar, Hoechst stain, and fluorescent antibody stain for Mycoplasmahyorhinis. Results were negative in all tests.4

Morphology

Cells growing in tissue culture T-flasks were examined by phase-contrast microscopy. Alternatively, cells were grown on coverslips, fixed,stained in situ, and evaluated by standard light microscopy.

Ultrastructural analysis was accomplished by transmission electronmicroscopy on cells fixed [2.5% (v/v) glutaraldehyde in 0.1 M phosphatebuffer, pH 7.4], postfixed (osmium tetroxide), and embedded whileattached to the growth surface of the tissue culture flask. Tumor tissue,3-mm cubes, obtained from nude mice with the heterotransplanted cells

(see below), was fixed and embedded in the above combination ofmaterials prior to examination by electron microscopy.

Cytogenetic Analysis

After subculture, cells were grown for 48 to 72 hr, at which timemetaphase arrest was initiated by the addition of colchicine (0.2 ^g/ml),and the incubation continued for 1 hr. Chromosome preparations weremade according to the air drying technique of Tjio and Whang (52). Bothconventional Giemsa stain and trypsin:Giemsa banding (46) were performed. Findings are described with nomenclature established at theParis Conference, 1971 (38).

Effect of Antineoplastic Drugs

Cells were subcultured into 25-sq cm tissue culture flasks and, 1 daypostplating, were changed to media which contained carrier, 0.9% NaCIsolution (control), Adriamycin (5 x 10~" M), melphalan (5 x 10~6 M), orcisplatin (5 x 10~7M). After 4 days of drug treatment, cells were harvested

with trypsin:EDTA and counted in a hemocytometer.

% of survival = Cells remaining after drug exposure

Cells in control flaskx 100

Tumorigenicity and Passage in Vivo

OVCAR-3 cells to be tested were harvested with trypsin:EDTA, and

the pellet of cells obtained by centrifugation was suspended in growthmedium (1.3 x 107 cells/ml). Amounts (0.1 ml) were injected s.c. into the

bilateral subscapular areas of nude athymic BALB/c mice (42). Tumorswhich formed were heterotransplanted to subsequent hosts by s.c.surgical implantation of small 3-mm cubes of solid tumor.

Initiation of the OVCAR-3"" Subline

Solid tumors initiated from injection of OVCAR-3 cells into nude micewere heterotransplanted once, and then the subsequent tumor wasexcised when it measured approximately 1 cm in its maximum plane.

4K. M. Miller (Flow Laboratories), personal communication.

5380 CANCER RESEARCH VOL. 43



Ovarian Carcinoma Cell Line

The tumor was finely minced in a minimum volume of growth mediumand added to a culture flask which contained 30 ml of growth mediumand collagenase (250 units/ml of medium). After 24-hr incubation under

culture conditions, the digest was repeatedly transferred by pipet to aidin further disaggregation of the tissue. A cell pellet obtained from thedigest by centrifugation was resuspended in PBS. The tumor cells in thissuspension were separated from blood cells and host stroma by astepwise fractionation procedure which utilized lymphocyte separationmedium (54). The tumor cell fraction was cultured and passaged asdescribed for the parental OVCAR-3 cell line.

Anchorage-independent Growth

The capacity for anchorage-independent growth of OVCAR-3 andOVCAR-3"" was assessed in medium modified to contain agarose (0.3%,

w/v) and EGF (10 ng/ml) (15, 39). A single-cell suspension of tryp-sin:EDTA-harvested cells was suspended in the above medium (100,000cells/ml), and 1-ml amounts were plated over a layer (1 ml) of 0.6% (w/v) agarose in growth medium in 35-mm tissue culture dishes. Aftersolidification of the cell-containing upper layer, plates were first examined

to ascertain that only single cells were present and then incubated at37Â°in a humidified atmosphere of 5% CO2 in air. At 5- to 7-day intervals,

0.8 ml of growth medium with EGF (10 ng/ml) was layered onto theagarose surface. After 21 days of incubation, cultures were scored forcolonies, which were defined as aggregates containing more than 30cells.

Assessment of Steroid Hormone Receptor Status

Buffer. The buffer used in the following procedures consisted of Tris-

HCI (10 rriM), pH 7.4, containing EDTA (1 mM), dithiothreitol (1 mw),phenylmethylsulfonyl fluoride (1 mw), and sodium molybdate (10 mM).

Protein Estimation. Cytosol protein was measured by a modification(21) of the method of Lowry ef al. (31) with bovine serum albumin asstandard.

Preparation of Cytosol. OVCAR-3 cells grown to confluence were

switched to and maintained for 6 to 8 days in a medium prepared withcharcoal-treated fetal bovine serum (49), after which time, the cells were

harvested with trypsin:EDTA, and the pellet obtained by centrifugationwas washed once with the above medium. Cytosol was then preparedfrom the cell pellet as described previously (18) in detail. In brief, the cellpellet was suspended in buffer and homogenized with a motor-drivenglass-Teflon homogenizer. The homogenate was then centrifuged at100,000 x g for 1 hr at 3Â°.The supernatant was the tumor cell cytosol.

Sedimentation Analysis. Cytosol was labeled with [3H]estradiol or[3H]R 1881 in the absence or presence of various potential competitorsfor 3 to 5 hr at 4Â°and then treated with charcoal (18) prior to analysis

of 200-iil amounts on 4-ml linear (5 to 20%, w/v) gradients of sucrose

prepared in a uniform concentration of buffer. Centrifugation was at150,000 x g for 16 hr at 3Â°.After centrifugation, gradients were frac

tionated (bottom to top, 150-jil fractions) into scintillation vials.In the case of tumors which formed in mice after injection of OVCAR-

3, cytosol was prepared with a homogenate generated in a hand-operated glass-glass homogenizer from tumors removed 3 days post-

ovariectomy. Cytosol was analyzed as described above.Saturation Analysis. Androgen binding was assessed in cytosol from

OVCAR-3-induced mouse tumors with [3H]R 1881 as labeled ligand and

5Â«-dihydrotestosterone as competitor at a 100-fold molar excess. Estrogen binding was studied with [3H]estradiol and diethylstilbestrol. The

assay protocols were as was described previously for assessment ofestrogen binding using charcoal to separate free from bound steroid (17).The data for specific binding (difference between binding in the presenceand absence of unlabeled competitor) were analyzed according to themethod of Scatchard (43).

Measurement of Radioactivity. Samples containing tritium weremixed with 10 ml of scintillation fluid (Aquassure; New England Nuclear)and counted at an efficiency of approximately 40%.

RESULTS

Morphology and Growth Characteristics. OVCAR-3 culturesconsisted of cells with well-defined, closely approximating margins which grew as a monolayer giving the typical cobblestone-like appearance ascribed to epithelium in vitro (Fig. 2). OVCAR-3 has been in continuous culture for >10 months and has beenpassaged 20 times. The cells have a doubling time of 48 hr.Cells have been cryopreserved using a programmed freezer, andcultures were subsequently established from the frozen cells.Additionally, it is noted that OVCAR-3 cultures left at confluencefor extended periods develop foci of "piled-up" cells. These foci

seem morphologically more complex than simple multilayering;observation of carefully handled cultures revealed fragile 3-di-mensional papillary-like structures. Closer examination of this

feature was possible in culture material prepared for electronmicroscopy. One-^m plastic-embedded vertical sections of the

culture examined on the light microscope showed cells in singlelayers continuous with the grape-like clusters of cells (Fig. 4).High-magnification light microscopy revealed the rounded free

surfaces of the cells to have microvilli. The cell cytoplasm wasoften vacuolated, and round intercellular spaces were seen.

Stained coverslip cultures permitted more detailed examination of OVCAR-3. Papanicolaou-stained cells (Fig. 3) appeared

variable in both size and shape and contained large strikinglypleomorphic nuclei with prominent multiple nucleoli. These nuclear abnormalities contributed to an evident alteration in thenucleancytoplasmic ratio for the cells. The cells were also studiedwith regard to their staining characteristics with periodic acid-Schiff pre- and post-diastase treatment and were mucin positive.This was confirmed in mucicarmine-stained coverslip cultures.The OVCAR-3nu subline was morphologically indistinguishable

from the parental OVCAR-3 line.

Electron microscopic examination of the culture material confirmed and expanded on the light microscopic observations (Figs.5 and 6). Ultrastructural examination confirmed the presence ofmicrovilli which were sometimes branched. Cell-cell attachments

were abundant, but true individual desmosomes were uncommon. Junctional complexes demarcated intercellular spaces intowhich long microvilli often protruded, and rare intracellular spaceswere seen. Distinct from intracellular spaces, there were intracellular vacuoles which were membrane bound and containedno discernible contents. The cytoplasm also contained irregularelectron-dense bodies consistent with lysosomes. Small elongated mitochondria with shelf-like cristae were evident in the

cells as were abundant polyribosomes. The Golgi complex wasnot prominent, nor were rough or smooth endoplasmic retÃculaabundant. The nuclei were seen to contain only limited hetero-

chromatin but did contain abundant euchromatin and one to 2nucleoli.

Tumorigenicity of OVCAR-3. The cultured cells injected s.c.

into immunodeficient heterologous hosts formed expansilemasses in the dermis and subcutis which grew progressively to>1 cm within 10 to 12 weeks in all animals. At this stage, thetumor was excised, prepared for examination by light or electronmicroscopy, or implanted into subsequent hosts. Light microscopy showed that the malignant cells formed glands and papillarystructures (Fig. 7) but did not invade the skeletal muscle beneaththe subcutis. Mitotic figures were common in the tumor. Numerous vacuoles were seen within tumor cells, and these containedmucicarmine-positive material. Grimelius stain for argyrophil (en-

NOVEMBER 1983 5381



T. C. Hamilton et al.

terochromaffin) granules was negative. This pattern of stainingwas the same as in the original solid tumor of the patient. Therewas only limited stromal involvement with no malignant-appear

ing connective tissue cells.Electron microscopy of the tumor growing in nude mice

showed that carcinoma cells had an extensive array of branchingmicrovilli, numerous junctional complexes, and relatively straightborders with peg-like interdigitations (Fig. 8). Although OVCAR-3 cells as a tumor in vivo contained a full complement of organ-

elles consistent with their cultured progenitors, differences indegree of organelle development suggestive of more active cellswere apparent. In contrast to the cultured cells, the Golgi complex was prominent, and the rough endoplasmic reticulum wasmore hiahlv develooed. Polvribosomes were abundant also.Patches of glycogen were seen in some of the tumor cells, andareas of the tumor had cells with large vacuoles, which in somesections were seen to be membrane bound, often empty, butrarely contained reticular, partially eluted material consistent withmucin. Scattered among the predominant tumor cells were darkcells and rare multinucleated cells. At the interface of the tumorand mouse tissue, tumor cells formed a well-developed external

lamina, and the cells resting on this external lamina often hadsmall, pleomorphic dense granules consistent with lysosomes(Fig. 9).

Anchorage-independent Growth. The cloning efficiency ofOVCAR-3 cells was 0.01 % when suspended in semisolid agarose. OVCAR-3nu also demonstrated the capacity for substrate-

independent growth but with a 10-fold-greater cloning efficiency

(0.1%), compared to the parent cell line. The colonies whichformed contained numerous balloon-like cells similar to the ap

pearance of ovarian carcinoma cells after direct cloning of malignant ascites cells in the double-layer agar system (35).

Cytogenetic Analysis. The conventional Giemsa-stained me-

taphases examined revealed 20% of the cells to have chromosome numbers near the 16-ploid region with extensive fragmentation present in 10% of the metaphases. Fifty banded meta-

phases were analyzed. In this group, the number of chromosomes in the majority of cells clustered around 60 to 70 chromosomes, but a range of 54 to 160 chromosomes per cell wasfound. The representative karyotype consistent with that of ahuman female is shown in Fig. 10. Nearly every pair of chromosomes had some abnormalities, either numerical or structural,with origins of some of them not identifiable. All contained oneor 2 small acrocentric chromosomes, and 5% of preparationsalso had an additional small acentric fragment which resembleda single DM. The common chromosome markers are 1p+,1p+q31-, 7p+, 11p+, 12q+, 19q+, 20q+, 20p12-q12- (small

acrocentric), and one small acrocentric with an unknown origin.The additional material on 19q is consistent with a HSR.

Response to Antineoplastic Drugs. Table 1 summarizes theeffects of Adriamycin, melphalan, and cisplatin upon the survivaland proliferation of OVCAR-3 cells. The cells were most resistant

to melphalan, with a 45% survival following continuous exposureto the drug at a concentration 3 times greater than the peakachievable plasma level. At concentrations of cisplatin and Adriamycin approximately one-tenth the peak-achievable plasma

levels, i.e., chosen to reflect clinically relevant drug doses (36),treated cultures were 77 and 43% of untreated controls, respectively. These results of in vitro resistance are in marked contrastto the in vitro sensitivities for these same drugs in ovarian cancercell lines 1847 and 2780, which were established from patients

Table 1Effect of antineoplastic agents on OVCAR-3 survival

Drug M dose % of control

AdriamycinMelphalanCisplatin

5x 10-Â«(0.03)a5x 10-" (1.5)5x10-7(0.15)

434577

a Numbers in parentheses, ng/ml.

previously untreated (8). At the drug concentrations shown inTable 1, there was complete inhibition of colony formation in cellline 2780. Likewise, for cell line 1847, these concentrations ofAdriamycin, melphalan, and cisplatin resulted in 15, <2, and 60%survival, respectively.5

Steroid Hormone Receptor Status. Cytosol prepared fromOVCAR-3 cells when examined by sedimentation analysis was

found to contain a component which bound synthetic (R 1881 )and natural androgens (5Â«-dihydrotestosterone) and with a sedi

mentation coefficient of 7 to 9S (Chart 1>4).A separate anddistinct moiety with similar sedimentation characteristics but withspecificity for estrogenic compounds was present, also, in thecytosol of OVCAR-3 (Chart IB). Under these conditions, apparent androgen receptor-like binding was greater than estrogen

binding site levels based on displaceable tritiated steroid bindingin the 8S region of gradients. Cytosol prepared from OVCAR-3nu

had the same pattern of binding (Chart 2, A and B) as did theparental line, OVCAR-3.

The ability to grow OVCAR-3 in nude mice simplified accu

mulation of sufficient cell mass for more extensive characterization of these steroid hormone receptor-like macromolecules and

had the advantage of a more physiological growth environmentfor the cells than that obtainable in the tissue culture flask. Thespecificities of the androgen and estrogen binding entities observed by sucrose density gradient analysis were more completely investigated in tumor cytosol. Chart 3^ shows that the 7to 9S [3H]R 1881 binding species only bound androgenic com

pounds with the exception of partial competition by progesteroneat a 100-fold molar excess. The component which bound [3H]

estradici was seen to bind nonsteroidal estrogenic and antiestro-

genic molecules but showed no ability to bind the syntheticandrogen receptor ligand, R 1881, or triamcinolone acetonide(Chart 3B). When the specific binding data generated by saturation analysis were analyzed by the method of Scatchard (43),apparent and distinct single classes of binding sites for androgenand estrogen were present with dissociation constants of 250and 9.6 PM,respectively. Saturation analysis, also, demonstratedthat the levels of androgen and estrogen receptor-like binding in

tumor cytosol (30 and 28 fmol/mg cytosol protein, respectively)were consistent with our previous results (17) on fresh humanovarian tumors [range (fmol/mg cytosol protein): androgen, 4 to88; estrogen, 15 to 136].

DISCUSSION

The morphological, ultrastructural, and biochemical featuresof OVCAR-3 and the corresponding human tumor grown in the

nude mouse are consistent with a derivation from a malignantcommon epithelial tumor of the human ovary. The demonstrations that OVCAR-3 cells both in culture and as tumors in nudemice possess steroid hormone-binding macromolecules provide

5T. C. Hamilton and R. F. Ozols, manuscript in preparation.





10 20FRACTION TOP

10 20FRACTION

Chart 1. Cytosol prepared from confluent cultures of OVCAR-3, which had beenmaintained for 6 days in medium prepared with charcoal-treated fetal bovine serum,was analyzed by sedimentation through gradients of sucrose after incubation (3hr, 4Â°)with: in A, 5 nM [3H]R 1881 (O) alone or in the presence of unlabeled 5a-dihydrotestosterone (â€¢)at 500 nM, cytosol protein (4.1 mg/ml); and in B, 5 nw [3H]

estradici (O) alone or in the presence of unlabeled diethylstilbestrol (â€¢)at 500 nu,cytosol protein (6.2 mg/ml). After centrifugation, gradients were fractionated bottomto top (shown left to right). Indicators of comparative sedimentation are bovinecatatase (a, sx.w = 11.35), human immunogtobulin G (b, SM.â€ž= 6.9S), and bovineserum albumin (c, S;,0w= 4.6S).

2SW

1500

1000

500

10 20

FRACTION

10 20

FRACTION

Chart 2. Cytosol (protein concentration, 6.3 mg/ml) prepared from confluentcultures of OVCAR-31*1,which had been maintained for 6 days in medium prepared

with charcoal-treated fetal bovine serum, was analyzed by sedimentation throughgradients of sucrose as described in Chart 1. A. [3H]R 1881 (5 nM) alone (O) or inthe presence of 5a-dihydrotestosterone (â€¢)at 500 MM;B, [3H]estradiol (5 nw) alone

(O) or in the presence of diethylstilbestrol (â€¢)at 500 nM. Fractionation and markersof comparative sedimentation are as in Chart 1.

the basis for unique models for the study of the requisite hormones in human ovarian cancer. In addition, since OVCAR-3

was developed from a patient refractory to cytotoxic chemotherapy, it provides a system for the investigation of drug resistance and cross-resistance in ovarian cancer.

Morphologically, cultured cells were epithelial and malignant innature and continued to show the focal presence of mucin.Ultrastructurally, the presence of microvilli, occasional desmo-some-like junctions, and abundant tight junctions confirmed the

epithelial and supported the ovarian character of the cells (9,27).The intercellular and rare intracellular spaces described wereconsistent with the gland-forming propensity of adenocarcinoma

cells (14), and the striking histological similarities between thetumor induced in athymic mice and the original solid tumor of thepatient were noteworthy. The patient's tumor, although classified

I

8504

350

300

200

100

10FRACTION

20 TOPTOP

Charts. Cytosol prepared from OVCAR-3 tumors grown in nude mice wasanalyzed by sedimentation after incubation (3 hr, 4") with: in A, 4 nM [3H]R 1881(O) alone or in the presence of unlabeled R 1881 (â€¢),5Â«-dihydrotestosterone P).triamcinolone acetonide (A), progesterone (A), or diethylstilbestrol (â€¢),all at 400nM, cytosol protein (9.0 mg/ml); and in B, 4 nM [3H]estradiol (â€¢)alone or in the

presence of unlabeled diethylstilbestrol (O, 400 nM), triamcinolone acetonide (A,400 nM), R 1881 (A, 400 nM), or tamoxifen (â€¢.4 ^M), cytosol protein (14.7 mg/ml).Fractionation and markers of comparative sedimentation are as in Chart 1.

as poorly differentiated and often exhibiting as solid sheets ofmalignant cells, did show the occasional tendency to form glandular and papillary structures (Fig. 1) as did the tumor whengrown in nude mice (Fig. 7). The presence of mucin in the tumor,in culture, and in the nude mouse tumor provides further supportfor the biological similarities between the original tumor and theexperimental models.

Epidemiological evidence has long implicated hormones in theetiology of ovarian cancer (16); an example is the increasedincidence of the disease at or near the climacteric. We (17, 19)and others (2,11,12, 22, 24, 28, 40, 45) have reported recentlythe existence of steroid hormone receptors in fresh epithelialovarian tumors. In addition, our previous studies have shownthat the surface germinal epithelium of the rat ovary has at least2 classes of steroid hormone receptors, estrogen and glucocor-

ticoid (18,19). If this pattern is maintained across species boundsas may be expected, then the surface epithelium of the humanovary, the accepted normal progenitor of common epithelialtumors, would have such receptors. The clinical significance ofthese potential control systems is unknown in both the normaland malignant cells. The presence of steroid receptors in ovariantumors, however, has been used as the rationale for hormonaltherapy in the disease (44). In the present study, cell culture andin vivo growth were used as sources of material to demonstratethat binding site concentrations, sedimentation characteristics,specificity, and dissociation constants of OVCAR-3 cytosol were

characteristic of androgen and estrogen receptors in classicaltarget tissues (5, 25, 47). Unfortunately, the receptor status ofthe patient's primary tumor was not established at the time of

diagnosis. We have shown, also, that the peaks of radioactivityseen on sedimentation analysis of 3H-steroid-labeled cytosol are

shifted from the 8S region of low-salt gradients to the 4 to 5Sregion in high-salt (0.6 M KCI) sucrose density gradients and thatthe material which binds [3H]estradiol interacts with a monoclonal

antibody to the estrogen receptor as manifested by a change insedimentation profile of [3H]estradiol-labeled cytosol in the pres

ence of the antibody (20).

NOVEMBER 1983 5383




The characteristics of OVCAR-3 in culture and as xenografts

make these systems suitable for investigations into the biologyof ovarian carcinoma both in vivo and in vitro. Studies on theeffects of androgens, antiandrogens, estrogens, and antiestro-gens on cell growth of OVCAR-3 in vitro and tumor growth in

vivo are in progress. The results of these studies may providefurther insight into the molecular basis for endocrine therapy inovarian cancer. The effects of steroids on macromolecular synthesis are also under investigation. Of particular interest arestudies on the induction of progesterone receptors with estrogenas occurs in breast cancer (26) and has been speculated tooccur in ovarian cancer (22). These model systems should alsoprove useful in the study of other factors of regulatory significance. Indeed, the shift in pattern of binding site concentrationfor androgen and estrogen receptors between the cultured cellsand the mouse tumor indicates that these systems are subjectto regulation by as yet unidentified factors. The impact of thedynamic milieu in vivo on organelle development in OVCAR-3

provides further evidence for the regulation of ovarian tumor cellfunction by unknown physiological agents.

Abnormalities of chromosome 1 are the most common cyto-

genetic abnormalities in ovarian cancer; they are present in over80% of reported cases (10, 29, 30, 51, 53, 55, 57) and arecommon (30 of 44 patients) in our own studies performed onfresh tissue (56) as well as in the present study on OVCAR-3.

Unfortunately, such changes are not unique and thus cannot beconsidered markers in the classical sense. It was noted, also,that about 50% of the cells examined were missing chromosome6; in cases where the chromosome was present, however, noabnormalities were evident. Of special interest were the observations of a HSR and one DM in the karyotype of OVCAR-3.

Trent and Salmon (53) noted either HSRs or DMs in 20% of theiragar cultures of ovarian cancer specimens, and Woods ef al.(57) found such changes in their ovarian cancer cell lines. Therelevance of such changes in human cancer has yet to bedelineated, but HSRs and DMs have been postulated to have arole in the mechanism of drug resistance (3, 4). It is possible thatthese karyotypic changes in OVCAR-3 relate to mutations favor

able to cancer cell survival in the face of extensive but unsuccessful chemotherapy in the patient from whom this line originated. Studies in progress should clarify whether the pattern ofresistance on OVCAR-3 is related to the cytotoxic drugs to

which the patient was exposed. If this is the case, the mechanisms of drug resistance in OVCAR-3 are likely to be analogous

with those which develop in vivo in the course of treatment ofhuman ovarian cancer. The preliminary results of in vitro drugsensitivity for melphalan, Adriamycin, and cisplatin (Table 1)suggest that OVCAR-3 is resistant to these drugs at clinically

relevant drug levels. The exact magnitude of the resistanceawaits comparative cytotoxicity studies with human ovarian cancer cell lines established from untreated patients and cell linesmade resistant by exposure to drugs in vitro. It is also possiblethat drug resistance in OVCAR-3 may be transitory for some of

the drugs, since it has been reported previously that both drugresistance and DMs can disappear when cells are maintained inthe absence of drug (4). Since OVCAR-3 does appear to be

resistant to those drugs with which the patient was treated, thiscell line should be a useful model in which to study the role ofmembrane glycoproteins in development of resistance and cross-

resistance (41) and potential modulators of drug resistance, suchas verapamil (48) and glutathione (50). Knowledge gained from

such studies in OVCAR-3 and other appropriate model systems

will hopefully lead to a better understanding of ovarian cancerand ultimately to new treatment approaches for this disease inwhich current therapeutic modalities are not successful for themajority of patients.

ACKNOWLEDGMENTS

The authors wish to thank Turid Knutsen for excellent technical assistance andthe secretaries of the Medicine Branch for careful typing of the manuscript. HermanMichelitch and Ralph Isenberg are acknowledged for their expert assistance withelectron microscopy and photography. We also thank Dr. Stuart A. Aaronson(National Cancer Institute) for human ovarian cancer cell lines 1847 and 2780.

REFERENCES

1. Bast, R. C., Freeney, M., Lazarus, H.. Nadler, L. M., Colvin, R. B., and Knapp,R. C. Reactivity of a monoclonal antibody with human ovarian carcinoma. J.Clin. Invest., 68: 1331-1337, 1981.

2. Bergquist, A., Kuliander, S., and Thorell, J. A study of estrogen and progesterone cytosol receptor concentrations in benign and malignant ovarian tumorsand a review of malignant ovarian tumors treated with medroxyprogesteroneacetate. Acta Obstet. Gynecol. Scand. Suppl.. 101: 75-81,1981.

3. Cowan, K. H., Goldsmith, M. E., Levine, R. M., Aitken, S. C., Douglass, E.,Clendeninn, N., Nienhuis, A. W., and Lippman, M. E. Dihydrofolate reducÃasegene amplification and possible rearrangement in estrogen-responsive meth-otrexate-resistant human breast cancer cells. J. Biol. Chem., 257: 15079-15086,1982.

4. Curt, G. A.. Carney, D. N., Cowan, K. H., Jolivet, J., Bailey, B. D., Drake, J.C., Kao-Shan, C. S., Minna, J. D., and Chabner, B. A. Unstable methotrexateresistance in human small-cell carcinoma associated with double minute chromosomes. N. Engl. J. Med., 308: 199-202, 1983.

5. Davies, P., Thomas, P., and Griffiths, K. Measurement of free and occupiedcytoplasmic and nuclear androgen receptor sites in rat ventral prostate gland.J. Endocrinol., 74: 393-404, 1977

6. DiSaia, P. J., Morrow, M., Kanabus, J., Piechal, W.. and Townsend, D. E. Twonew tissue culture cell lines from ovarian cancer. Gynecol. Oncol., 3: 215-219, 1975.

7. Donahoe, P. K., Swann, D. A., Hayashi, A., and Sullivan, M. D. Mullerian ductregression in the embryo correlated with cytotoxic activity against humanovarian cancer. Science (Wash. D. C.). 205: 913-915, 1979.

8. Eva, A., Robbins, K. C.. Andersen, P. R., Srinivasan, A., Tronick, S. R., Reddy,E. P., Ellmore, N. W., Galen, A. T., Lautenberger. J. A.. Papas, T. S., Westin,E. H., Wong-Staal, F., Gallo, R. C., and Aaronson, S. A. Cellular genesanalogous to retroviral one genes are transcribed in human tumour cells.Nature (Lond.), 295: 116-119, 1982.

9. Fenoglio, C. M. Overview article: ultrastructural features of common epithelialtumors of the ovary. Ultrastruct. Pathol., 7. 419-444. 1980.

10. Freedman, R. S., Phil, E., Kusyk, C., Gallager, H. S., and Rutledge, F.Characterization of an ovarian carcinoma cell line. Cancer (Phila.), 42: 2352-2359, 1978.

11. Friberg, L. G., Kullander, S., Persijn, J. P., and Korsten, C. B. On receptors forestrogens (E2) and androgens (DHT) in human endometrial carcinoma andovarian tumors. Acta Obstet. Gynecol. Scand., 57: 261-264, 1978.

12. Galli, M. C., DeGiovanni, C., Nicoletti, G., Grilli, S., Nanni, P., Prodi, G., Gola,G., Rocchetta, R., and Oriandi, C. The occurrence of multiple steroid hormonereceptors in disease-free and neoplastic human ovary. Cancer (Phila.), 47:1297-1302,1981.

13. Geisler, H. E. Megestrol acetate for the palliation of advanced ovarian carcinoma. Obstet. Gynecol., 61: 95-98, 1983.

14. Ghadially, F. N. Diagnostic Electron Microscopy of Tumours. Boston: Butter-worth, 1980.

15. Hamburger, A. W., White, C. P., and Brown, R. W. Effect of epidermal growthfactor on proliferation of human tumor cells in soft agar. J. Nati. Cancer Inst.,67:825-830,1981.

16. Hamilton, T. C., and Davies, P. Hormonal relationships in ovarian cancer. Rev.Endocr.-Relat. Cancer, 14: 19-22. 1983.

17. Hamilton, T. C., Davies, P., and Griffiths, K. Androgen and oestrogen bindingin cytosols of human ovarian tumours. J. Endocrinol., 90: 421-431, 1981.

18. Hamilton, T. C., Davies, P., and Griffiths, K. Oestrogen receptor-like binding inthe surface germinal epithelium of the rat ovary. J. Endocrinol., 95: 377-385,1982.

19. Hamilton, T. C., Davies, P., and Griffiths, K. Steroid-hormone receptor statusof the normal and neoplastic ovarian surface germinal epithelium. In: G. S.Greenwald and P. F. Terranova (eds.). Factors Regulating Ovarian Function,pp. 81-85. New York: Raven Press, 1983.

20. Hamilton, T. C., McKoy, W., Grotzinger, K. R., Green, J. A., Rogan, A. M.,Kasid, A., and Ozols, R. F. A model system for investigation of androgen andestrogen action in human ovarian carcinoma (abstract). Endocrinology. 108(Suppl.V 184, 1983.

21. Hartree, E. F. Determination of protein: a modification of the Lowry method





that gives a linear photometric response. Anal. Biochem., 48: 422-427,1972.22. Holt, J. A., Caputo, T. A., Kelly, K. M., Greenwald, P., and Chorost, S. Estrogen

and progestin binding in cytosols of ovarian adenocarcinomas. Obstet. Gyne-col., 53:50-58, 1979.

23. Holt, J. A., Lorincz, M. A., and Hospelhom, V. D. Sulfhydryl sensitivity and[125l]-16Â«-iodo-17)3-estradiol binding of estrogen receptor in ovarian epithelialcarcinomas. J. Steroid Biochem., 18: 41-50, 1983.

24. Holt, J. A., Lyttle, C. R., Lorincz, M. A., Stern, S. D., Press, M. F., and Herbst,A. L. Estrogen receptor and peroxidase activity in epithelial ovarian carcinomas.J. Nati. Cancer Inst., 67: 307-317, 1981.

25. Horwitz, K. B., Costlow, M. E., and McGuire, W. L. MCF-7; a human breastcancer cell line with estrogen, androgen, progesterone, and glucocorticoidreceptors. Steroids, 26: 785-795, 1975.

26. Horwitz, K. B., and McGuire, W. L. Estrogen control of progesterone receptorin human breast cancer. J. Biol. Chem., 253: 2223-2228, 1978.

27. loachim, H. L., Dorsett, B. H., Sabbath, M., and Barber, H. K. Electronmicroscopy, tissue cultures, and immunology of ovarian carcinoma. Nati.Cancer Inst. Monogr., 42: 45-62, 1975.

28. Janne, O., Kauppila, A., Syrjala, P., and Vihko, R. Comparison of cytosolestrogen and progestin receptor status in malignant and benign tumors andtumor-like lesions of the human ovary. Int. J. Cancer, 25: 175-179,1980.

29. Kakati, S., Hayata, I., Oshimura, M., and Sandberg, A. A. Chromosomes andcausation of human cancer and leukemia. X. Banding patterns in cancerouseffusions. Cancer (Phila.), 36: 1729-1738,1975.

30. Kusyk, C. J., Seski, J. C., Medlin, W. V., and Edwards, C. L. Progressivechromosome changes associated with different sites of one ovarian carcinoma.J. Nati. Cancer Inst., 66: 1021-1025, 1981.

31. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Proteinmeasurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275,1951.

32. Myers, A. M., Moore, G. E., and Major, F. J. Advanced ovarian cancer:response to antiestrogen therapy. Cancer (Phila.), 48: 2368-2370, 1981.

33. Owens, R. B. Selective cultivation of mammalian epithelial cells. Methods CellBiol., 74:341-355,1976.

34. Ozols, R. F., Grotzinger, K. R., Fisher, R. I., Myers, C. E., and Young, R. C.Kinetic characterization and response to chemotherapy in a transplantablemurine ovarian cancer. Cancer Res., 39: 3202-3208,1979.

35. Ozols, R. F., Willson, J. K. V., Grotzinger, K. R., and Young, R. C. Cloning ofhuman ovarian cancer cells in soft agar from malignant effusions and peritonealwashings. Cancer Res., 40: 2743-2747,1980.

36. Ozols, R. F., Willson, J. K. V., Weitz, M. D., Grotzinger, K. R., Myers, C. E.,and Young, R. C. Inhibition of human ovarian cancer colony formation byAdriamycin and its major metabolites. Cancer Res., 40: 4109-4112, 1980.

37. Ozols, R. F., Young, R. C., Speyer, J. L., Sugarbaker, P. H., Greene, R.,Jenkins, J., and Myers, C. E. Phase I and pharmacological studies of Adriamycin administered intraperitoneally to patients with ovarian cancer. CancerRes., 42: 4265-4269, 1982.

38. Paris Conference (1971). Standardization in human cytogenetics. In: BirthDefects Original Article Series, Vol. 8, No. 7. White Plains, N. Y.: NationalFoundation-March of Dimes, 1972.

39. Pathak, M. A., Matrisian, L. M., Magun, B. E., and Salmon, S. E. Effect ofepidermal growth factor on clonogenic growth of primary human tumor cells.

Int. J. Cancer, 30: 745-750, 1982.

40. Quinn, M. A., Pearce, P., Rome, R., Funder, J. W., Fortune, D., and Pepperell,R. J. Cytoplasmic steroid receptors in ovarian tumors. Br. J. Obstet. Gynaecol.,89: 754-759, 1982.

41. Riordan, J. R., and Ling, V. Purification of P-glycoprotein from plasma membrane vesicles of Chinese hamster ovary cell mutants with reduced colchicinepermeability. J. Biol. Chem., 254: 12701-12705, 1979.

42. Rygaard, J. Immunobiology of the mouse mutant "nude." Acta Pathol. Micro-

biol. Scand., 77: 761-762, 1969.43. Scatchard, G. The attractions of proteins for small molecules and ions. Ann.

N. Y. Acad. Sci., 5Ã•:660-672, 1949.44. Schwartz, P. E., Keating, G., MacLusky, N., Naftolin, F., and Eisenfield, A.

Tamoxifen therapy for advanced ovarian cancer. Obstet. Gynecol., 59: 583-588, 1982.

45. Schwartz, P. E., LiVoIsi, V. A., Hildreth, N., MacLusky, N. J., Naftolin, F. N.,and Eisenfeld, A. J. Estrogen receptors in ovarian epithelial carcinoma. Obstet.Gynecol., 59: 229-238,1982.

46. Seabright, M. A. A rapid banding technique for human chromosomes. Lancet,2:971-972, 1971.

47. Seibert, K., and Lippman, M. E. Hormone receptors in breast cancer. Clin.Oncol., 1: 735-794, 1982.

48. Slater, L. M., Murray, S. L., and Wetzel, M. W. Verapamil restoration ofdaunorubicin responsiveness in daunorubicin-resistant Ehrlich asertes carcinomas. J. Clin. Invest., 70: 1131-1134,1982.

49. Stanley, E. R., Palmer, R. E., and Sohn, H. Development of methods for thequantitative in vitro analysis of androgen-dependent and autonomous Shionogicarcinoma 115 cells. Cell, 70: 35-44, 1977.

50. Suzakaka, K., Petra, B. J., and Vistica, D. T. Reduction in glutathione contentof L-PAM resistant L1210 cells confers drug sensitivity. Biochem. Pharmacol.,37: 121-124,1982.

51. Tiepolo, L., and Zuffardi, 0. Identification of normal and abnormal chromosomes in tumor cells. Cytogenet. Cell Genet., 72: 8-16,1973.

52. Tjio, J. H., and Whang, J. Chromosome preparations of bone marrow cellswithout prior in vitro culture or in vivo colchicine administration. Stain Technol.,37: 17-20, 1962.

53. Trent, J. M., and Salmon, S. E. Karyotypic analysis of human ovarian carcinomacells cloned in agar (abstract). Am. J. Hum. Genet., 37: 379, 1979.

54. Vose, B. M. Quantitation of proliferative and cytotoxic precursor cells directedagainst human tumours: limiting dilution analysis in peripheral blood and at thetumour site. Int. J. Cancer, 30: 135-142, 1982.

55. Wake, N., Hreshchyshyn, M. M., Piver, S. M., Matsui, S., and Sandberg, A. A.Specific cytogenetic changes in ovarian cancer involving chromosomes 6 and14. Cancer Res., 40: 4512-4518, 1980.

56. Whang-Peng, J., Knutsen, T., Douglass, E. C., Chu, E., Ozols, R. F., Hogan,W. M., and Young, R. C. Cytogenetic studies in ovarian cancer. Cancer Genet.Cytogenet., in press, 1983.

57. Woods, L. K., Morgan, R. T., Quinn, L. A., Moore, G. E., Semple, T. U, andStedman, K. E. Comparison of four new cell lines from patients with adeno-carcinoma of the ovary. Cancer Res., 39: 4449-4459,1979.

58. Young, R. C., Knapp, R. C., and Perez, C. A. Cancer of the ovary. In: V. T.DeVita, S. Hellman, and S. A. Rosenberg (eds.), Cancer Principles and Practiceof Oncology, pp. 884-913. Philadelphia: J. B. Lippincott Co., 1982.

NOVEMBER 1983 5385




'

.â€¢

Fig. 1. Surgical specimen of ovarian carcinoma. Malignant cells invade the ovarian stroma and form glands. H & E, x 160.Fig. 2. OVCAR-3 cells as a confluent monolayer. Phase contrast, x 118.

Fig. 3. Coverslip culture of OVCAR-3 prepared for light microscopy. Papanicolaou, x 400.





l

.'â€¢â€¢'.â€¢''â€¢'

.. >

A â€¢â€¢â€¢Â«--v"-

Fig. 4. One-^m plastic section shows that malignant cells in vitro form clusters of varying sizes. The lower edge of the section corresponds to the bottom of theculture flask. MÃ©thylÃ¨neblue, Azur II, x 1,080.

Fig. 5. Low-magnification electron micrograph shows microvilli at the free cell surface, x 4,400.

Fig. 6. Cell cluster in vitro. A mitotic figure is present at the upper center. Intercellular spaces are wide and are bounded by junctkxial complexes (arrows), x 9,800.

NOVEMBER 1983 5387



7. C. Hamilton et al.

-. c â€¢ ' "â€¢- * *â€¢- - '

; â€¢â€¢"i-

> "jÂ«r^?%:*'â€¢Â¿Â«';M' ' ' * 'M

â€¢:>>';:^-/. ' '^â€¢:>

Fig. 7. Malignant cells after injection into a nude, athymic mouse form a tumor which invades the subcutis and encroaches on underlying skeletal muscle. H & E, x160. Inset, papillary clusters of malignant cells. H & E, x 320.

Fig. 8. Tumor cells in the nude mouse show well-developed Golgi complex (G) seen in one cell above and below the nucleus and also in an adjacent cell. A functionalcomplex (c) is present, x 10,500.




*,:*â€¢. /

lÃSÃ•CÃIH7 8

Ã¼ li ??W W

H Kill4 o

MMt!-Â«Â«10 11 12

13

*** * AP16

19

Â«Ã¼

21

??

22

17

t

18

XX

Fig. 9. Cells at the invading edge of the tumor in the nude mouse form a well-developed external lamina (arrow). Lysosome-like granules are present in these cells,9,240. Two membrane-bound coalescing vacuoles (v) are shown. A solitary vacuole with flocculent material (m) is also present in this vacuolated cell, x 8,400.

Fig. 10. Representative karyotype of OVCAR-3. Arrows indicate sites of additions or deletions on involved individual chromosomes.

NOVEMBER 1983 5389



1983;43:5379-5389. Cancer Res Thomas C. Hamilton, Robert C. Young, Wilma M. McKoy, et al. (NIH:OVCAR-3) with Androgen and Estrogen ReceptorsCharacterization of a Human Ovarian Carcinoma Cell Line

Updated version

http://cancerres.aacrjournals.org/content/43/11/5379

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/43/11/5379To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



Documents

Characterization of a Human Ovarian Carcinoma Cell Line ...cancerres.aacrjournals.org/content/canres/43/11/5379.full.pdf · Characterization of a Human Ovarian ... hormonal influence