Expression of OVCA1, a Candidate Tumor Suppressor, Is Reduced in Tumors and Inhibits Growth of Ovarian Cancer Cells

Reagents and Cell Lines.

Cell culture reagents were from Life Technologies, Inc. (Gaithersburg, MD); unless otherwise indicated, most other reagents were obtained from Sigma Chemical Co. (St. Louis, MO). Cell lines were obtained from American Type Culture Collection (Manassas, VA) or were derived in our laboratory (HOSE, human ovarian surface epithelial cell lines grown in primary culture; and HIO cell lines, SV40-immortalized human ovarian epithelial cells). A2780 cells were maintained in DMEM supplemented with 10% FCS and 0.2 IU/ml porcine insulin. COS-1, MCF-7, and MDA-MB8 cells were maintained in DMEM supplemented with 10% FCS. T47D cells were maintained in RPMI 1640 supplemented with 10% FCS and 0.2 IU/ml porcine insulin. SKBR3 cells were maintained in McCoy’s 5a medium supplemented with 10% FCS. HOSE cells and HIO cell lines were maintained in a 1:1 mixture of medium 199 and MCDB-105 medium, supplemented with 5% FCS and 0.2 IU/ml porcine insulin. Unless otherwise stated, cells were transfected with Superfect (Qiagen, Chatsworth, CA), as described by the manufacturer. The A2780 clones that stably express OVCA1 were obtained after selection in G418 by standard methods and maintained in DMEM supplemented with 10% FCS and 0.5 mg/ml G418.

SSCP Analysis of OVCA1.

PCR was carried out in a reaction volume of 10 μl containing 100 ng of genomic DNA template, 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 1.5 mm MgCl₂, 0.001% gelatin, 1 μm forward and reverse primers, 60 μm dNTPs, 0.1 μCi of [³²P]dATP (DuPont-NEN, Boston, MA), 5% DMSO, and 0.5 unit of AmpliTaq DNA polymerase (Perkin Elmer Corp., Foster City, CA). Following an initial denaturation step at 94°C for 4 min, DNA was amplified through 20 cycles consisting of 5 s of denaturing at 94°C, 1 min of annealing at 65°C − 0.5°C/cycle, and 1 min of extension at 72°C. The samples were then subjected to an additional 25 cycles, consisting of 1 min of denaturation at 94°C, 1 min of annealing at 55°C, and 1 min of extension at 72°C and a final extension at 72°C for 5 min. PCR products were diluted 1:10 in 95% formamide, 10 mm NaOH, 0.25% bromphenol blue, and 0.25% xylene cyanol. Diluted products were denatured for 5 min at 95°C and flash-cooled on ice. Four μl were loaded onto a 0.5× MDE gel (AT Biochem, Malvern, PA), prepared according to manufacturer’s specifications, and electrophoresed at 6 W for 12–16 h at room temperature in 0.6× TBE [1× TBF, 0.09 m Tris, 0.09 m boric acid, and 0.002 m EDTA (pH 8.0)]. Following electrophoresis, the gel was dried and exposed to autoradiography film at −80°C for 4–24 h. Variant and normal SSCP bands were cut out from the gels after alignment with the autoradiograph, and the DNA was eluted in 100 μl of distilled deoionized H₂O at 37°C for 3 h. Two μl of the eluted DNA were used as template for secondary PCRs, as described above, except that radiolabeled dATP was omitted. Following amplification, the DNA was collected on Wizard resin (Promega, Madison, WI) and eluted in 50 μl of distilled deoionized H₂O, and 50–100 fmol of purified PCR product were subjected to direct sequencing.

Plasmids.

The eukaryotic expression vectors pcDNA3 and pcDNA3-LacZ were obtained from Invitrogen. The HA antibody tag (YPYDVPDYA) was added to the COOH or NH₂ terminus of the OVCA1 cDNA by standard PCR technology, and the resulting tagged cDNAs were subcloned into pcDNA3 and are referred to as pcDNA3-HAOVCA1 or pcDNA3-OVCA1HA, depending on the location of the HA tag. The plasmid pGFP-C1, which expresses green fluorescent protein, was obtained from Clontech. The cDNA of OVCA1 was fused to the COOH terminus of the green fluorescent protein at the BglII site to generate the plasmid pGFP-OVCA1. To prepare a GST fusion of OVCA1 in bacteria, we subcloned the OVCA1 cDNA, containing an NH₂-terminal HA tag, into pGEX-2T (Pharmacia).

Production of Anti-OVCA1 Antibodies.

The 13-amino acid peptide, RDGPGRGRAPRGC, corresponding to amino acids 20–31 of OVCA1 (where the terminal cysteine was added for conjugation purposes) was synthesized (Research Genetics, Huntsville, AL). Purity of the peptide was confirmed by high-performance liquid chromatography. The peptide was conjugated to malemide activated keyhole limpet hemocyanin (Pierce, Rockford, IL) and used to immunize a New Zealand White rabbit (Cocalico, Reamstown, PA). Two mg of antigenic peptide were covalently linked to Aminolink agarose (Pierce) and used to purify anti-OVCA1 antibody from crude serum by affinity chromatography. The final antibody is referred to as TJ132. The antibody FC21 was produced by immunizing a New Zealand White rabbit (Cocalico) with a bacterially expressed carboxyl terminal portion of OVCA1 (amino acids 330–443). The resulting antiserum was immunoaffinity purified on Aminolink agarose covalently linked to bacterially expressed GST-OVCA1.

Purification of Bacterially Expressed OVCA1.

BL21 bacteria were transformed with pGEX2T-OVCA1. Expression of the fusion protein was induced with 1 mm isopropyl-β-thio-galactopyranoside (Stratagene, La Jolla, CA). The bacteria were lysed by sonication, and GST-OVCA1 was purified from the soluble fraction by binding to glutathione-Sepharose 4B (Pharmacia). Pure OVCA1 was released by digesting with thrombin (Pharmacia), or the GST-OVCA1 fusion was eluted with excess glutathione. PET-OVCA1 (nucleotides 1011–1350) was expressed in BL21 bacteria and purified as an insoluble inclusion body by repeated washing of the insoluble fraction with 1% Triton X-100. The insoluble pellet was solubilized in 8 m urea-2% SDS. The protein (OVCA1 amino acids 330–443) was further purified by SDS-PAGE. The gel slice containing the protein was homogenized and used to immunize rabbits.

Preparation of Protein Extracts from Human Tumor Specimens.

Normal human tissues were obtained from Clontech. Tumors were snap-frozen after surgical removal and stored in liquid nitrogen until use. One g of tumor tissue was rinsed twice with cold PBS and minced finely into small pieces. Tissue pieces were suspended in 1 ml of PBSTDS [0.137 m NaCl, 2.68 mm KCl, 10.6 mm Na₂HPO₄, 1.47 mm K₂H₂PO₄, 1% (v/v) Triton X-100, 0.5% (w/v) deoxycholate, 0.1% (v/v) SDS, 0.004% (w/v) NaF, 100 mg/ml phenylmethylsulfonyl fluoride, 1 mg/ml aprotinin, 1 mg/ml leupeptin, and 2 mm sodium orthovandate, (pH 7.4)] and ground with a Polytron tissue grinder at 300–400 rpm for two 30-s intervals at 4°C. Tissue homogenates were clarified by centrifugation at 100,000 × g for 1 hour at 4°C. Lipid layers were removed, and cytosolic extracts were aliquoted, snap-frozen in liquid nitrogen, and stored at −80°C. Quantitation of protein was achieved using a bicinchoninic acid/copper (II) sulfate assay (Sigma).

SDS-PAGE and Western Blot Analysis.

Fifty μg of total protein extract from tissues or 20 μg of total protein from cell extracts, unless otherwise stated, were separated by standard SDS-PAGE and transferred to Immobilon-P (polyvinylidene difluoride; Millipore, Bedford, MA). The membranes were blocked with 3% BSA and probed with the anti-OVCA1 antibody TJ132, or blocked with 3% dried milk and probed with the indicated antibody. The signal was visualized using anti-rabbit antibodies coupled to HRP (Amersham) and developed using ECL reagents, as recommended by the manufacturer (Amersham).

Subcellular Fractionation.

HOSE cells were homogenized in ice-cold hypotonic homogenization buffer [40 mm Tris (pH 7.4), 1 mm EDTA, 1 mm EGTA, 1 mm DTT, and 10% glycerol]. The nuclei were pelleted by centrifugation at 2500 rpm for 10 min. The supernatant was collected, and insoluble debris was pelleted at 180,000 × g for 30 min to give the cytosol fraction. The nuclear pellet was washed twice with homogenization buffer containing 0.1 m KCl. The nuclear pellet was then extracted with homogenization buffer plus 0.45 m KCl for 1 h on ice, with frequent vortexing. Insoluble debris was pelleted at 180,000 × g for 30 min to obtain the nuclear fraction.

Immunofluorescent Staining and Imaging.

COS-1 cells were transfected with the indicated plasmids using Lipofectamine (Life Technologies, Inc.) or Superfect (Qiagen), as directed by the manufacturer. Forty-eight h after transfection, the cells were visualized. For green fluorescent proteins, the living cells were observed with a Nikon TE 800 upright microscope. For visualizing HA-tagged OVCA1 proteins, the cells were fixed for 20 min in 3.7% formaldehyde-PBS and then for 30 s in methanol. The cells were then stained with an anti-HA rabbit polyclonal antibody Y-11 (Santa Cruz Biotechnology, Santa Cruz, CA) in TBS [50 mm Tris (pH 8.0) and 140 mm NaCl] plus 0.03% Triton X-100 and 10% FCS. The staining was visualized with an antirabbit antibody coupled to Texas Red (Jackson Immunoresearch Laboratory, Inc.). The stained cells were observed on a Biorad MRC 600 laser scanning confocal microscope, using COMOS Version 7.0.1 software. The images were rendered and pseudocolored with Voxel View 2.5.1 (Vital Images) software. Final prints were made using a codonics dye sublimation printer.

Stable Colony Formation Assay.

A2780 cells (2.5 × 10⁵ per 60-cm plate) were cotransfected with 5 μg of pcDNA3-LacZ and 2.5 pmol of pcDNA3 control vector, pcDNA3-HAOVCA1, or p53 expression plasmid or the cyclin D1 expression plasmid. At 24 h posttransfection, G418 (Life Technologies, Inc.) was added to a final concentration of 0.5 mg/ml, or cells were stained for transient β-galactosidase activity. Antibiotic selection was continued for 10–14 days. Colonies were fixed with 0.2% formaldehyde and stained with 0.2% (w/v) crystal violet, and colonies containing >50 cells were scored.

Growth Curve Analysis.

Cells were removed from the flask by trypsinization. The trypsin was inactivated by addition of complete medium to a final volume of 10 ml. One hundred μl of cell suspension were diluted in 20 ml of Isoton solution (Coulter, Miami, FL), and the number of cells quantified on a Z1 Coulter counter (Coulter). Cells (200,000) were plated in triplicate in 35-mm tissue culture dishes and incubated at 37°C and 5.0% CO₂. Cells were counted in 24-h intervals by trypsinization and resuspension of cells in 10 ml of Isoton (Coulter) and counted on the Z1 Coulter counter (Coulter).

Pulse-Chase Labeling.

Cells were seeded into 60-mm dishes and grown until they were 60% confluent. They were starved in minus-methionine medium (ICN) for 30 min, and then Trans35-Label (ICN) was added to 500 μCi/ml and the cells were labeled for 30 min. The radioactive medium was removed, and the cells were washed with large volumes of complete medium and then incubated in complete medium for the indicated times. The cells were then lysed in 100 μl of PBSTDS. Insoluble debris was pelleted, and the lysates were diluted 10-fold into RIPA buffer [10 mm Tris (pH 8.0), 150 mm NaCl, 1% NP40, 0.1% SDS, and 0.5% deoxycholate]. Anti-cyclin D1 antibody (Santa Cruz Biotechnology) was added and the immunoprecipitates were collected on Protein A beads (Life Technologies, Inc.) and washed well with RIPA buffer. The immunoprecipitates were released by boiling in SDS-PAGE loading buffer and were separated by 12% SDS-PAGE. The amount of label incorporated into cyclin D1 was quantitated by Phosphoimager (Fuji).

FACS Analysis of Stable Transfectants.

Cells (500,000) were seeded in 10 ml of complete medium supplemented with 0.5 mg/ml G418. Seventy-two h postseeding, cells were harvested and 1 million cells were prepared for FACS analysis by resuspending cell pellets in 1 ml of staining buffer [3.4 mm sodium citrate, 10 mm NaCl, 0.1% (v/v) NP40, and 75 mm ethidium bromide] and stored at 4°C for no more than 3 days. The cells were filtered through a 37-μm nylon mesh and then analyzed by a flow cytometer (Becton Dickinson, San Jose, CA). Data for 20,000 events were gathered by CellQuest (Becton Dickinson, San Jose, CA) and analyzed by MacCycle (Phoenix Flow Systems, San Diego, CA).

TUNEL Staining.

Cells were plated on coverslips and stained for TUNEL using an in situ cell death detection kit, as recommended by the manufacturer (Boehringer Mannheim).

Mutational Analysis of OVCA1 by SSCP.

SSCP analysis was conducted on 50 ovarian tumors independent of LOH status for markers on 17p13.3 and on 20 breast tumors demonstrating allelic loss of OVCA1 and retention of TP53. Multiple sequence variants were identified throughout the gene (Fig. 1 ⇓ ; Table 1 ⇓ ). These sequence variants were deemed to be polymorphisms because these same alterations were either found in the corresponding germ line or resulted in either conservative or silent amino acid substitutions. The frequency of these putative polymorphisms was determined by SSCP analysis of 100 chromosomes from control individuals (Table 1) ⇓ . In addition, we identified two nonconservative amino acid substitutions: alanine 34 changed to an aspartic acid residue and serine 389 changed to an arginine residue. Each alteration was detected in the germ line of a woman with early-onset breast cancer who reported a family history of the disease. In both cases, the missense mutation/rare polymorphism was retained in the corresponding breast tumor DNA and showed reduction to homozygosity (data not shown). Evaluation of >100 control chromosomes has failed to detect these sequence variants. The individual carrying the A34D missense variant was diagnosed with breast cancer at age 37 and reported a history of one first-degree relative and two second-degree relatives with the disease (ages of onset unknown). The individual carrying the S389R missense variant was diagnosed with breast cancer at age 49. She reported that her mother was affected with breast cancer at age 55 and that two maternal aunts were diagnosed with the disease at 61 and 65 years of age. The functional significance of these mutations is not yet clear; preliminary experiments exploring their effect on the OVCA1 protein are presented below.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Mutational analysis of OVCA1. Each exon and its surrounding intronic regions was amplified by PCR and analyzed for sequence variants by SSCP, as described in “Materials and Methods.” Examples of the SSCP gel patterns are shown for 10 tumor samples for exons 1, 6, and 9. The exon 1 region has several sequence variants, shown in Lanes 1, 5, 6, 7, and 8 (left to right); exon 6 has one sequence variant, shown in Lanes 7, 8, and 9; and exon 9 has one sequence variant, shown in Lanes 5 and 9. All variants seen by SSCP were verified by directly sequencing a separate PCR.

Southern Blot Analysis of OVCA1.

To assess deletions or rearrangements within the OVCA1 gene, we performed Southern blotting of 60 normal/ovarian tumor DNA pairs using the full-length OVCA1 cDNA as the probe. The vast majority of the tumors had lost one copy of the OVCA1 gene. No rearrangements or large interstitial deletions were detected in the remaining copy. However, a 7-kbp EcoRI fragment was observed to be variable in length due to changes in the VNTR, i.e., YNH37.3, which is intragenic to OVCA1 (data not shown). We did not observe any correlation between the length of the VNTR fragment and an increased risk of developing ovarian cancer.

Western Blot Analysis of OVCA1.

Conceptual translation of OVCA1 predicts a 443-amino acid protein with M_r ∼50,000. An antibody that recognizes 11 amino acids at the NH₂ terminus of OVCA1 was prepared by immunizing rabbits with a peptide. The antiserum was affinity-purified and was designated TJ132. Another antibody that recognizes the COOH terminus of OVCA1 (amino acids 330–443) was prepared by immunoaffinity purification following immunization of rabbits with a bacterially expressed polypeptide and was designated FC21. Both antibodies were able to recognize bacterially expressed OVCA1 by Western blotting (data not shown). In addition, these antibodies were able to recognize a protein of M_r ∼50,000 in extracts prepared from COS-1 cells that had been transiently transfected with pcDNA3-HAOVCA1 and in whole-cell lysates from the ovarian tumor cell line A2780 (Fig. 2A) ⇓ . Recognition of this M_r 50,000 protein could be competed with a molar excess of the antigenic peptide, indicating that the antibodies recognize the authentic OVCA1 protein (data not shown). In addition to the M_r 50,000 protein, both antibodies detected proteins of M_r ∼85,000, as observed in extracts prepared from a variety of sources, including normal human tissues, primary cultures of HOSE cells and a number of cell lines (Figs. 2 ⇓ and 3 ⇓ ; data not shown). The NH₂-terminal antibody TJ132 also recognized proteins of M_r ∼70,000, but these species were variable in amount and presence and were not recognized by antibodies directed against the COOH terminus of the protein. The secondary antibody alone did not recognize any of the three proteins (M_r 50,000, 70,000, and 85,000; data not shown). The identity of the M_r 70,000 and M_r 85,000 proteins is unknown, as are their relationships with the M_r 50,000 OVCA1 protein; however, the available evidence suggests that the M_r 85,000 form is an alternatively spliced or posttranslationally modified form of OVCA1 and that the p70 form is either unrelated to OVCA1 or is a breakdown product of the p85 form. Alternatively, the p85/p70 forms could be unrelated, cross-reacting proteins. However, this is unlikely because completely different anti-OVCA1 antibodies recognize the p85 protein, and recognition of the M_r 85,000 protein by TJ132 can be competed with a molar excess of the antigenic peptide. Because FC21 and TJ132 gave almost identical patterns by Western blotting (Fig. 2B) ⇓ , most of the data shown used only the antibody TJ132.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Characterization of OVCA1 expression. A, extracts from the indicated tissues and cell lines were separated by 10% SDS-PAGE and transferred to Immobilon-P, as described in “Materials and Methods.” The blot was then probed with the anti-OVCA1 antibody TJ132, as described in “Materials and Methods.” Lane IVT, in vitro translated pcDNA3-HAOVCA1; Lane COS, extract of COS-1 cells that had been transfected with pcDNA3-HAOVCA1; Lane A2780, extract of the ovarian tumor cell line A2780; Lane H. ovary, extract of whole human ovary. Arrowheads, three different polypeptides that TJ132 recognizes (p50, p70, and p85). B, 20 μg of each indicated cell line extract were separated in duplicate by 10% PAGE, transferred to Immobilon-P, and probed with the indicated antibodies, as described in “Materials and Methods.” One of the duplicate blots was probed with the anti-OVCA1 antibody TJ132, and the other was probed in parallel with the anti-OVCA1 antibody FC21. OVCAR-5 is a cell line derived from an ovarian tumor, whereas T47D and SKBR3 are cell lines derived from breast tumors. C, 50 μg of extracts from various human tissues (Clontech) were separated by 12% SDS-PAGE and processed for Western blotting, as described in “Materials and Methods.” The blot was probed with the anti-OVCA1 antibody TJ132. Lane H, heart; Lane B, brain; Lane P, placenta; Lane L, lung; Lane Li, liver; Lane S, skeletal muscle; Lane K, kidney; Second Lane P, pancreas; Second Lane S, spleen; Lane T, thymus; Lane M, mammary gland; Lane Te, testis; Lane O, ovary.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Expression of OVCA1 in tumors. A, 50 μg of each sample were separated by 12% SDS-PAGE and transferred to Immobilon-P, as described in “Materials and Methods.” The blot was probed with the anti-OVCA1 antibody TJ132. UPN, extracts from primary tumors; MCF-7 and MDA-MB8, extracts from cell lines derived from breast tumors; N.M.E*, extract from normal breast epithelial cells grown in short-term primary tissue culture. Equal loading of the blots was confirmed by staining with Coomassie Blue for total protein after probing. B, protein extracts (50 μg) from primary ovarian tumors (UPN), normal human ovarian surface epithelial cell lines (HIO), and a normal ovary (Ovary) were separated by 12% SDS-PAGE and transferred to Immobilon-P, as described in “Materials and Methods.” The blot was probed with the anti-OVCA1 antibody TJ132. Equal loading of the blots was confirmed by staining the blots for total protein with Coomassie Blue after probing.

OVCA1 was found to be expressed in many different tissues (Fig. 2C) ⇓ . In some cases, the M_r 70,000 and M_r 85,000 proteins were very prominent, whereas the M_r 50,000 protein was less so (notably the ovary and placenta) and, in other tissues, the p50 form was predominant (liver and thymus). Note that, although extracts from total breast tissue appeared to express little or no OVCA1 (Fig. 2C) ⇓ , breast epithelial cells did express the p50 OVCA1 protein (Fig. 3A ⇓ , N.M.E.*). We explain this apparent discrepancy as being due to epithelial cells making up only a low percentage of the total breast. Analysis of breast and ovarian tumor extracts demonstrated variable expression levels of p50 and an almost complete absence of the p70/p85 species (Fig. 3) ⇓ . Expression levels of p50 were reduced as compared to normal epithelial cells in 21 of 59 ovarian (37%) and 18 of 46 breast (39%) carcinomas. p85 and p70 were not detected in the majority of tumors analyzed (100% of breast tumors and 85% of ovarian tumors) (Fig. 3) ⇓ . No correlation was evident between reduced expression and clinical prognostic factors.

Subcellular Localization of OVCA1.

To aid in understanding the function of OVCA1, we determined its subcellular localization. COS-1 cells were transfected with either an empty vector pcDNA3 or with pcDNA3-OVCA1HA, which expresses OVCA1 fused to a COOH-terminal HA tag. Immunostaining of transfected cells with an anti-HA antibody (Y-11; Santa Cruz Biotechnology) indicates that OVCA1 is located throughout the cell. A widespread diffuse staining was seen, in addition to strongly staining punctate bodies (Fig. 4 A and B) ⇓ . These bodies were scattered throughout the cell and were heavily clustered around the nucleus. A similar pattern was obtained in immortalized HOSE cells transfected with pcDNA3-OVCA1HA and when the cells were immunostained with the specific anti-OVCA1 antibody, TJ132 (data not shown). To further confirm the localization, we fused OVCA1 to the COOH terminus of the green fluorescent protein. COS-1 cells expressing the green fluorescent protein-OVCA1 fusion again demonstrated a punctate, primarily perinuclear localization of the protein set against a weaker, diffuse staining throughout the cell (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Subcellular localization of OVCA1. COS-1 cells were transiently transfected with pcDNA3 or with pcDNA3-OVCA1HA. Forty-eight h after transfection, the cells were fixed and stained with an anti-HA tag antibody (Y-11; Santa Cruz Biotechnology), as described in “Materials and Methods.” A, cells transfected with pcDNA3; B, cells transfected with pcDNA3-OVCA1HA. At the level of sensitivity needed to obtain high resolution of the OVCA1HA staining, the mock-transfected cells are not visible. C, subcellular fractionation of SV40-immortalized human ovarian surface epithelial cell lines (HIO). Cell lines were fractionated as described in “Materials and Methods,” and 50 μg of the corresponding extracts were separated by 12% SDS-PAGE and transferred to Immobilon-P. The Western blot shown was probed with the anti-OVCA1 antibody TJ132. Lane WCL, whole-cell extract; Lane C, cytoplasmic fraction; Lane N, nuclear fraction.

Fractionation studies confirmed that the M_r 50,000 OVCA1 protein is located throughout the cell (Fig. 4C) ⇓ . However, the M_r 70,000 and M_r 85,000 species appeared to be exclusively located within the nucleus (Fig. 4C) ⇓ .

Suppression of Clonal Outgrowth.

Attempts to generate cell lines that stably expressed OVCA1 were unsuccessful. Very few clones were found to express OVCA1, and those that did expressed only low levels of the protein. This phenomenon was consistently observed in a number of different cell types (RAT-1, U2OS, MCF-7, HIO118, and T47D cells; data not shown). To quantitate this effect, we transfected equimolar amounts of a mammalian expression vector containing the NH₂-terminal HA-tagged OVCA1 open reading frame (pcDNA3-HAOVCA1) and an empty expression vector (pcDNA3) into the ovarian cancer cell line A2780. A2780 cells were chosen for further analysis because they are well-characterized ovarian tumor cells that normally express fairly low levels of OVCA1 p50 and almost no p85/p70 OVCA1 (Fig. 2) ⇓ . As a positive control for growth suppression, an expression vector that expresses wild-type p53 protein was included in some of the colony number experiments. Evaluation of colony formation in the presence of geneticin (G418) consistently resulted in a 50–60% reduction in colony number in cells transfected with the OVCA1 expression construct compared with controls (Fig. 5 A) ⇓ . This effect was reproducibly observed in more than four independent transfection experiments. Suppression of clonal outgrowth was independent of plasmid DNA purity (we tested three different preparations of plasmid DNA) whether equivalent molar amounts or microgram amounts of plasmid were transfected. Furthermore, experiments in which an expression vector containing the gene encoding for the β-galactosidase protein were cotransfected with OVCA1 indicate that the reduction in clonal outgrowth is not an artifact due to differences in transfection efficiency (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Suppression of clonal outgrowth by OVCA1 in A2780 ovarian cancer cells. A2780 cells were transfected with the indicated plasmids and selected for resistance to G418, as described in “Materials and Methods.” Ten to 14 days posttransfection, the colonies were stained and counted. A, the percentage of G418 (Neo^R) colonies that formed relative to the number formed after transfection with pcDNA3 (defined as 100% in each experiment) along the ordinate. Abscissa, data from eight independent transfection experiments. A wild-type TP53 expression vector was included in some experiments as a positive control for colony suppression. B, the total number of colonies obtained after transfection with the indicated plasmids and selection with G418. pcDNA3 is the parent vector; OVCA1HA expresses the wild-type OVCA1 plus a COOH-terminal HA tag; and A34D and S389R refer to the point mutations, discussed in the text, that were introduced into the wild-type OVCA1HA construct using standard PCR technology. Columns, means of three independently repeated experiments; bars, SD.

The A34D and S389R alterations described above, detected in the germ line of women with breast cancer and with a strong family history of the disease, were rebuilt into the pcDNA3-OVCA1HA expression plasmid using standard PCR technology. Both altered proteins were expressed well in transient transfection assays (data not shown). However, both alterations were found to suppress colony formation 50–60%, as compared with controls, similar to wild-type OVCA1 (Fig. 5B) ⇓ .

Growth Kinetics of Stable Transfectants.

To verify that the suppression effect was due to exogenous OVCA1 expression, individual colonies were clonally expanded after G418 selection. A total of seven colonies from pcDNA3 vector control transfected cells and 15 colonies from pcDNA3-HAOVCA1 transfected cells were amplified following selection in G418 for 10 days. All colonies selected from pcDNA3 vector control plates expanded and formed stable cultures. In contrast, only 9 of 15 colonies selected from pcDNA3-HAOVCA1 transfected cells expanded to form a stable culture. Because an in-frame HA epitope was fused to the open reading frame of OVCA1 at the NH₂ terminus, the level of exogenous protein produced in these clones could be monitored. Western blot analysis revealed that there was approximately equimolar expression of exogenous and endogenous OVCA1 in only four of nine stable pcDNA3-HAOVCA1 clones (OV-5, OV-6, OV-9, and OV-13; Fig. 6 ⇓ ).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6.

Maintenance of exogenous OVCA1 expression in stable transfectants of A2780 ovarian cancer cells. A2780 cells were transfected with an HAOVCA1 expression vector and then selected for resistance to G418, as described in “Materials and Methods.” Clones were chosen at random and amplified. After amplification, extracts were prepared from the cells, as described in “Materials and Methods.” Ten μg of each extract were separated by duplicate 10% SDS-PAGE. The gels were transferred to Immobilon-P and probed with the indicated antibodies, as described in “Materials and Methods.” Top, total OVCA1 antigen was detected with the anti-OVCA1 antibody TJ132. Bottom, exogenous OVCA1 antigen was detected with the anti-HA mAB 16B12 (BabCo, Richmond, CA). Lane COS, protein extract prepared from COS cells transiently transfected with an HAOVCA1 expression vector; Lanes A2780, protein extract prepared from the parental cell line; Lanes CMV, cell lines derived from pcDNA3 transfected cells; Lanes OV, cell lines derived from HAOVCA1-transfected cells.

Of the HAOVCA1 transfectants with exogenous expression, no major differences in morphological features were observed when compared to parental A2780 cells (data not shown). Cells were plated at limited dilutions and monitored for growth kinetics. Two independent clones, OV-5 and OV-13, displayed ∼8- and 10-fold reductions in total cell number compared with expression vector controls and parental A2780 cells, respectively. A third clonal line, OV-9, demonstrated a 4-fold reduction in total cell number over the same time interval compared with controls (Fig. 7A) ⇓ . On the basis of these growth curves, the cell doubling times between parental A2780 and OVCA1-expressing stable clones were found to be considerably different. A2780 cells doubled 2–2.5 times during a 24-h period, whereas OV-5, OV-9, and OV-13 doubled ∼1–1.5 times during the same time interval. Consistent with the reduced growth rate, the clones stably expressing OVCA1 had a dramatic reduction in cyclin D1 levels (Fig. 7B) ⇓ . The reduction of steady-state cyclin D1 levels appeared to be primarily due to an increased rate of degradation of cyclin D1 in cells expressing HAOVCA1 compared with the parental cell line (Fig. 7C) ⇓ .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7.

Suppression of growth rate by OVCA1. A, the proliferation of A2780 clonal lines was monitored over time, as described in “Materials and Methods.” Vector (□; CMV-5) is A2780 cells that have stably integrated the plasmid pcDNA3; HAOVCA1-5 (•; OV-5) and HAOVCA1-9 (▴; OV-9) are two lines that stably express OVCA1. The parental cell line, A2780, gave results that were virtually identical to those of the CMV-5 cell line (data not shown), and the stably expressing OVCA1 clone OV-13 gave results similar to that of OV-5 (data not shown). The graph represents the growth rates for the cell lines over the indicated time period and is representative of a number of independent experiments. Ordinate, number of cells (10⁴); abscissa, days in 24-h time points. B, OVCA1 expression (Western blot probed with the anti-HA tag antibody Y-11; Santa Cruz Biotechnology) and cyclin D1 expression (the Western blot showing OVCA1 expression was reprobed with an anti-cyclin D1 antibody; Santa Cruz Biotechnology) in the indicated cell lines: Lane 1, CMV-5; Lane 2, OV-5; Lane 3, OV-9. The parental cell line, A2780, gave results that were virtually identical to those of the CMV-5 cell line (data not shown), and the stably expressing OVCA1 clone OV-13 gave results that were similar to those of OV-5 (data not shown). C, the stability of cyclin D1 was monitored by pulse-chase [³⁵S]methionine labeling of the indicated cells followed by immunoprecipitation of the labeled cyclin D1. The immunoprecipitates were separated on SDS-PAGE and quantitated by Phosphoimager (Fuji). Abscissa, minutes after initiation of the chase; ordinate, relative units of labeling incorporated into cyclin D1. The graph is representative of a number of independent experiments.

FACS Analysis of Stable Transfectants.

Two main mechanisms, apoptosis and cell cycle arrest, may account for the growth suppression observed in stable clones expressing exogenous OVCA1. To investigate the mechanism of growth suppression, we seeded parental A2780 cells and each of the stable transfectants at an equal number of cells. Seventy-two h postseeding, the cells were harvested, nuclei were stained with ethidium bromide, and cell cycle distribution was measured by FACS analysis. As illustrated in Fig. 8 ⇓ , a 10–20% increase in the number of cells in the G₁ fraction was observed in clones OV-5, OV-9, and OV-13 compared with parental A2780 cells and stable vector control cells, CMV-5. No subdiploid cell peaks suggestive of apoptosis were observed. To further investigate the possibility of apoptosis playing a role in reduced cell number, we subjected clones OV-5 and OV-9 to TUNEL staining and compared them with the vector control cells. There were no TUNEL-positive cells on the vector control cell slides. A total of 1.2% of the OV-9 cells were TUNEL-positive, and 4% of the OV-5 cells were TUNEL-positive, suggesting that, although rates of apoptosis are slightly elevated in A2780 cells stably expressing OVCA1, apoptosis does not fully account for the drastic reduction in growth rates (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 8.

FACS analysis of stable transfectants. Cell cycle distributions were determined 72 h postseeding by flow cytometry, as described in “Materials and Methods.” DNA profiles represent cell number (counts) along the ordinate and DNA content (PI/FL2) along the abscissa. Percentage of cells in each phase of the cell cycle is listed on the right. G₁, both the G₀ and G₁ populations of cells; S, the population of cells in DNA synthesis; G₂, both the G₂ and M populations of cells.

Cyclin D1 Overexpression Can Partially Overcome OVCA1’s Suppression of Clonal Outgrowth.

Reduction of cyclin D1 levels by OVCA1 may be the primary cause of OVCA1’s growth-suppressive effect. To test this theory, we cotransfected A2780 cells with pcDNA3-OVCA1HA and CMV-cyclin D1. The cells were transfected with pcDNA3 alone, pcDNA3-OVCA1HA alone, CMV-cyclin D1 alone, or both pcDNA3-OVCA1HA and CMV-cyclin D1. pcDNA3 and pskII were added as necessary to equalize the amount of selectable marker and plasmid DNA in each transfection. After selection for 14 days in G418, the colonies were fixed, stained, and counted. As noted previously, cells transfected with the OVCA1 expression construct formed ∼50% fewer colonies than did cells transfected with pcDNA3 (Fig. 9) ⇓ . Cells transfected with the cyclin D1 expression vector formed almost as many colonies as did cells transfected with pcDNA3. Cells cotransfected with pcDNA3-OVCA1HA and CMV-cyclin D1 formed ∼75% fewer colonies than did cells transfected with pcDNA3 and almost the same number of colonies formed by cells transfected with CMV-cyclin D1 alone, suggesting that overexpression of cyclin D1 can compensate for overexpression of OVCA1.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 9.

Overexpression of cyclin D1 can override OVCA1’s suppression of clonal outgrowth. A2780 cells were transfected with the indicated plasmids, as described in “Materials and Methods.” Resistant colonies were selected in G418 for 14 days and then they were fixed and stained. Colonies with >50 cells were counted. The experiment was repeated three times. pcDNA-3, transfected with 1 μg of pcDNA3 and 1 μg of pskII; CMV-OVCA1, transfected with 1 μg of pcDNA3-OVCA1HA and 1 μg of pskII; CMV-cyclin D1, transfected with 1 μg of CMV-cyclin D1 (carries no selectable marker) and 1 μg of pcDNA3; CMV-OVCA1+-cyclin D1 was transfected with 1 μg of pcDNA3-OVCA1HA and 1 μg of CMV-cyclin D1.