Paradoxical increase in retinoblastoma protein in colorectal carcinomas may protect cells from apoptosis

Paradoxical Increase in Retinoblastoma Protein in Colorectal Carcinomas May Protect Cells from Apoptosis Hirofumi Yamamoto, Jae-Won Soh, Takushi Monden, et al. Clin Cancer Res 1999;5:1805-1815.

Updated Version

Cited Articles Citing Articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/5/7/1805

This article cites 67 articles, 28 of which you can access for free at: http://clincancerres.aacrjournals.org/content/5/7/1805.full.html#ref-list-1 This article has been cited by 16 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/5/7/1805.full.html#related-urls

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

Vol. 5, 1805–1815, July 1999

Clinical Cancer Research 1805

Paradoxical Increase in Retinoblastoma Protein in Colorectal Carcinomas May Protect Cells from Apoptosis1 Hirofumi Yamamoto, Jae-Won Soh, Takushi Monden, Micheal G. Klein, Li Ming Zhang, Haim Shirin, Nadir Arber, Naohiro Tomita, Ira Schieren, C. A. Stein, and I. Bernard Weinstein2 Herbert Irving Comprehensive Cancer Center [H. Y., J-W. S., M. G. K., H. S., C. A. S., I. B. W.], Center for Neurobiology and Behavior [I. S.], Department of Medicine [L. M. Z., C. A. S., I. B. W.], and Department of Pharmacology [C. A. S.], Columbia University, College of Physicians and Surgeons, New York, New York 10032; Department of Surgery II, Osaka University Medical School, Osaka, 565-0871, Japan [H. Y., T. M., N. T.]; and Department of Gastroenterology, Tel Aviv Sourasky Medical Center, Tel Aviv, 64239, Israel [N. A.]

ABSTRACT The retinoblastoma (Rb) gene is inactivated in a variety of human cancers, but in colorectal carcinomas there is frequently increased expression of this gene. This is paradoxical in view of the known role of Rb as a tumor suppressor gene. In the present study, we compared the levels of expression of the Rb protein (pRb) in normal human colorectal mucosa, adenomatous polyps, and carcinomas by immunohistochemistry. In vitro studies were also done to examine the phenotypic effects of an antisense oligodeoxynucleotide (AS-Rb) targeted to Rb mRNA in the HCT116 colon carcinoma cell line that expresses a relatively high level of pRb. The incidence of pRb-positive cells was increased during multistage colorectal carcinogenesis. In vitro treatment of HCT116 cells with AS-Rb decreased the level of pRb by about 70% and also decreased the levels of the cyclin D1 protein and cyclin D1-associated kinase activity. AS-Rb inhibited growth of HCT116 cells and induced apoptosis. Reporter assays indicated about a 17-fold increase in E2F activity. These findings suggest that the increased expression of pRb in colorectal carcinoma cells may provide a homeostatic mechanism that protects them from

Received 12/23/98; revised 3/31/99; accepted 3/31/99. 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 in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by an award for cancer research from the FUSO Pharmaceutical Company (Osaka, Japan; to H. Y.), a Deutshes Krebsforschungszentrum (DKFZ) award (to N. A.), and by National Cancer Institute Grant CA 63467 and an award from the National Foundation for Cancer Research (to I. B. W.). 2 To whom requests for reprints should be addressed, at Herbert Irving Comprehensive Cancer Center, Columbia University, College of Physicians and Surgeons, 701 West 168th Street, New York, NY 10032. Phone (212) 305-6921; Fax: (212) 305-6889; E-mail: weinstein@ cuccfa.ccc.columbia.edu.

growth inhibition and apoptosis, perhaps by counterbalancing potentially toxic effects of excessive E2F activity.

INTRODUCTION Studies done about 10 years ago on cases of hereditary Rbs3 led to the discovery of the Rb gene, the first tumor suppressor gene to be identified (1–3). Subsequent studies indicated that the function of this gene is also lost, at the somatic level, in a variety of nonhereditary human cancers, including cancers of the breast, lung, soft tissue, bladder, and prostate. On the other hand, deletions and loss of expression of the Rb gene are rare in human colorectal carcinomas (for review see Ref. 4). Indeed, in these cancers, and in cell lines derived from these cancers, there is often increased expression of the Rb gene (5– 8), and in some tumors, there are additional copies of the Rb gene due to amplification of the corresponding DNA region on chromosome 13 or nonrandom increases in chromosome 13 (4). This increase in the expression of a tumor suppressor gene in colorectal cancers seems to be paradoxical, and its biological significance is not known. Studies done within the past few years indicate that the protein encoded by the Rb gene, pRb, normally plays a key role as a negative regulator of the G1 to S transition in the cell cycle by binding the transcription factor E2F and preventing it from activating the transcription of genes required for the S phase. Phosphorylation of pRb by cyclin D1/CDK4 or cyclin D1/ CDK6 complexes, and possibly by the cyclin E/CDK2 complex, during the latter part of G1 relieves this inhibitory effect, thus activating E2F (for review see Ref. 9). Human colorectal carcinomas frequently display abnormalities in the expression of several genes that play either positive or negative roles in this pathway. Thus, there is frequently increased expression of the cyclin D1, cyclin E, CDK1, and CDK2 proteins (7, 10, 11), and in a subset of tumors and cell lines, there is amplification of the cyclin E and CDK2 genes (12, 13). Alterations in cyclin dependent kinase inhibitors are also seen in colorectal carcinomas. The p16Ink4 gene is inactivated in about 40% of the cases due to de novo methylation of a 59 CpG island (14), and expression of the p21Waf1 gene is frequently decreased (15, 16). Decreased levels of p27Kip1 due to proteolytic degradation have been seen in a subset of cases, and this is associated with a poor prognosis or high-grade tumors (17, 18). To further explore the role of pRb in colon cancer, in the present study, we have examined the expression of pRb by immunohistochemistry, and we found that there was an increase in the expression of pRb during the multistage process of

3

The abbreviations used are: Rb, retinoblastoma; pRb, Rb protein; oligo, oligodeoxynucleotide; TUNEL, terminal deoxynucleotide transferase; DAPI, 49,6-diamidino-2-phenylindole; AS-Rb, antisense Rb oligo; S-Rb, sense Rb oligo.

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

1806 Antisense to Rb in Colon Cancer Cells

colorectal carcinogenesis. To examine the phenotypic and biochemical effects of decreasing the level of expression of pRb in colon cancer cells, we carried out in vitro studies using the HCT116 human colon carcinoma cell line because it expresses a relatively high level of pRb. Our strategy was to transfect into these cells an antisense phosphorothioate oligonucleotide targeted to the 59 region of Rb mRNA. Our results provide evidence that the increased expression of pRb has a positive selection value in HCT116 colon carcinoma cells because it protects them from growth inhibition and apoptosis. Furthermore, our studies provide evidence that cellular levels of pRb can markedly influence the expression of other proteins involved in cell cycle control.

MATERIALS AND METHODS Cell Lines and Tissues. The HCT116 human colon carcinoma and the WI38 human lung fibroblast cell lines were obtained from the American Type Culture Collection. The HCT116 cell line was grown in DMEM, and the fibroblasts were grown in RPMI 1640 plus 10% fetal bovine serum, 100 units/ml penicillin, and 100 mg/ml streptomycin in 5% CO2 at 37°C. Fifty colorectal carcinomas and paired adjacent normal mucosa samples, 50 samples of adenomatous colorectal polyps (37 tubular and 13 tubulo-villous adenomas), and 15 polyps containing focal carcinomas were collected during surgery or during endoscopic polypectomy in the Department of Surgery II in Osaka University Medical School from 1994 –1996. The samples of normal mucosa were cut in the longitudinal direction to examine the top, middle, and bottom parts of the glands, and the polyps and carcinomas were cut across the maximum diameter. These samples were fixed in buffered formalin at 4°C overnight and embedded in paraffin. For Western blot analyses, a piece of tissue was immediately frozen in liquid nitrogen and stored at 280°C. Immunostaining. Immunostaining was performed by an avidin-biotin complex method, as described previously (7). An anti-pRb antibody (G3-245; Pharmingen) was applied to the sections at a dilution of 1:50. A negative control section, to which normal mouse serum had been applied rather than the specific antibody, was included in each staining procedure. The entire series of samples was stained at least twice, with separately prepared sections, and similar results were obtained. With each section, 10 fields were randomly picked under high power magnification of the microscope, more than 700 cells were scored for positive or negative nuclear staining, and the percentage of pRb-positive cells was determined. If a nucleus displayed brown staining, it was considered positive. Construction of Antisense and Sense Oligos to Rb mRNA. An antisense phosphorothioate oligo to Rb mRNA (AS-Rb) was designed to target the translation start site. It was synthesized by standard methods using a 380B DNA synthesizer (Applied Biosystems Inc., Foster City, CA; Ref. 19). As a control, the corresponding S-Rb was synthesized. The sequences of these oligos and the corresponding region of the human Rb mRNA are shown below. Antisense oligo: 39-CAGTACGGCGGGTTTTGG-59; sense oligo: 59-GTCATGCCGCCCAAAACC-39;

Rb mRNA: 59-GUCAUGCCGCCCAAAACC (AUG is the translation start site). These 18 mer oligos are identical or similar to those used in previous studies (20 –22). Transfection of Oligos. Cells were plated at the appropriate numbers, according to the plate size, and grown for 48 h in the standard medium, at which time they were in the exponential growth phase and the monolayer was 40 –50% confluent. They were then rinsed once with serum-free medium and transfected with a 1 mM solution of the indicated oligo together with the lipofectin reagent (Life Technologies, Inc., Gaithersburg, MD), which were premixed for 15 min after dilution, as recommended by the manufacturer, in the medium (DMEM for HCT116 cells and RPMI 1640 for WI38 cells). The amount of lipofectin (19 mg/ml) and volume of medium (3– 4 ml for a 6-cm dish) used were proportional to the plate area. Four hours later, fetal bovine serum was added at a final concentration of 10%. The toxicity of the lipofectin was monitored by cell viability using trypan blue dye exclusion and found to be minimal because the viability for HCT116 and WI38 cells at 48 h after transfection was over 96%. Western Blot Analysis. Western blot analysis was performed, as described previously (23). Equal loading of the protein samples was confirmed either by Coomassie blue staining or by immunoreactivity with an antiactin antibody (Sigma). The intensities of the bands were quantitated with an image scanner (Molecular Dynamics). Antibodies. The following antibodies were used for detection of various human proteins, using the concentrations recommended by the manufacturer: (a) mouse monoclonal antibodies: Rb (G3-245) and p16Ink4 (Pharmingen), and p21Waf1 (Ab-1; Oncogene Science); (b) rabbit polyclonal antibodies: cyclin A, cyclin D1, CDK2, CDK4, and cyclin E (UBI); actin (Sigma); p27Kip1 (C-19) and CDK6 (Santa Cruz). Flow Cytometric Analysis. Flow cytometric analysis was performed, as described previously (23). In Vitro Assay for Cyclin D1, Cyclin E, and CDK2associated Kinase Activities. In vitro assays for cyclin D1-, cyclin E-, and CDK2-associated kinase activities were performed as described previously (23). For the cyclin D1-associated kinase assay, 200 mg of the cell extracts were immunoprecipitated with 2 mg of the cyclin D1 antibody. The GST-Rb fusion protein (1 mg) was incubated with the immunoprecipitate plus 10 mCi [g-32P]ATP for 30 min at 30°C. For the cyclin Eor CDK2-associated kinase assay, 100 mg or 50 mg of total cell extracts, respectively, were immunoprecipitated with 2 mg of the cyclin E antibody or 1 mg of the CDK2 antibody. These immunoprecipitates were incubated with 5 mg of Histone H1 (Sigma) as the substrate and 5 mCi of [g-32P]ATP for 30 min at 30°C. The reaction mixtures were then subjected to SDS-PAGE, and the intensities of phosphorylation of the substrates were determined by autoradiography. In Situ Apoptosis Detection Using the TUNEL Assay and DAPI Staining. For the detection of apoptotic cells, an in situ apoptosis detection kit ApopTag (Oncor, Gaithersburg, MD) was used, as recommended by the manufacturer (20). For the DAPI staining, the cells were incubated with DAPI solution (Boehringer Mannheim, Indianapolis, IN) at a concentration of 1.5 mg/ml PBS for 30 min and then washed in PBS for 2 h. For

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

Clinical Cancer Research 1807

Fig. 1 Immunohistochemical staining for pRb expression with an antihuman pRb antibody in normal colonic mucosa (A), an adenomatous colonic polyp (B), a focal cancer in an adenomatous polyp (C), and a colon carcinoma (D). In the normal mucosa, the positive staining for nuclear pRb was confined to the lower portion of the crypts and the germinal center of lymphoid follicles (A). The upper regions of the normal crypt (not shown here) did not show pRb staining. In the polyps and carcinomas, pRbpositive cells were scattered throughout the lesions. Magnification: A, 366; B, 366; C, 350; D, 350.

quantitation of apoptotic cells, 10 microscopic fields were randomly chosen at 325 magnification, and a series of pictures were taken under light (for whole cells) or fluorescence (for TUNEL-positive cells or for DAPI-positive cells) microscopy. The total number of the cells and the number of apoptotic cells (those with a fluorescent signal by TUNEL staining and those with a nuclear fragmentation or a condensation by DAPI staining) were counted in each field. Over 700 total cells were counted in each sample, and the percentage of the apoptotic cells was calculated. Luciferase Reporter Assay for E2F Activity. E2F activity was determined by a luciferase reporter assay. The luciferase reporter plasmid E2F-Luc contains four repeats of the wild-type E2F site (TTTCGCGC) upstream of a minimal TATA box, followed by the luciferase coding sequence (24). To normalize the transfection efficiency, the E2F-Luc plasmid was cotransfected with a CMV-b gal (galactosidase) plasmid at a ratio of 2:1. At the indicated time, the cell extracts were prepared and assayed for luciferase and b-galactosidase, as described previously (25). The luciferase level in each sample was normalized by the corresponding value for b-gal activity. Colony Efficiency Assay. Cells were seeded at a density of 1 3 105/well in 6-cm dishes. Twenty four hours later, the vectors encoding wild-type E2F (CMV-Ap12-Stu) or mutant type E2F (CMV-Ap12DDS) under CMV promoter and the corresponding control vector (26) were cotransfected with the pBabe plasmid encoding puromycin resistance at a ratio of 5:1, using the lipofectin reagent. Forty-eight hours after transfection, the cells were subplated into 15-cm dishes and then selected in the presence of 0.5 mg/ml puromycin in complete medium for 2 weeks. Individual drug-resistant clones were stained with Giemsa solution (Sigma).

Statistical Analyses. Statistical analyses were performed using In Stat version 2.01. To determine correlations between the expression of pRb and histological types, Fisher’s exact test was used. The differences obtained between the S-Rb and AS-Rb treatment of the HCT116 cells was determined by the unpaired t test.

RESULTS pRb Expression during Multistage Colorectal Carcinogenesis. The expression of pRb was examined by immunohistochemistry in the following set of colorectal samples: normal mucosa (n 5 50), adenomatous polyps (n 5 50), focal carcinomas within adenomatous polyps (n 5 15), and various stages of carcinomas (n 5 50; Fig. 1 and Table 1). The extent of expression of pRb was divided into three categories, based on the percentage of cells that were positive for pRb, as follows: ,10%, 10 –50%, and .50%. In the normal mucosa samples, pRb-positive nuclei were seen only in the transitional zone of the crypt in the lower portion of the glands and not the upper portions of the crypts, and also in the germinal centers of lymphoid follicles (Fig. 1A). The percentage of pRb-positive epithelial cells in all of the normal mucosa samples was ,10% (Table 1). When compared with the normal mucosa samples, 14 of 50 (28%) adenomatous polyps, 13 of 15 (87%) focal carcinomas in polyps, and 39 of 50 (78%) carcinomas displayed increased expression of pRb. When a cut off of .50% was used to define very high expressors of pRb, none of the normal mucosa samples, 8% of the adenomatous polyps, 47% of the focal carcinomas, and 34% of the carcinomas fit this criterion (Table 1). The differences between the normal mucosa and tumor samples and between the adenomatous polyps and focal

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

1808 Antisense to Rb in Colon Cancer Cells

Table 1 pRb expression during multistage colorectal carcinogenesis In the polyps group of samples, the percentage of pRb-positive cells showed no significant correlation with the following clinicopathologic parameters: location in the colorectum (right side, transverse, left side, or rectum), size (,10 mm, 10 –20 mm, or .20 mm), shape (pedunculate, pseudosessile, or sessile), histology (tubular or tubulovillous). In the carcinomas group of samples, the percentage of pRbpositive cells showed no significant correlation with tumor size (,3.9 cm or .4.0 cm), depth of invasion, Dukes’ stage, site, or metastasis to lymph nodes or the liver. pRb expression (% of cells positive)

Normal mucosa Adenomatous polyps Focal cancers in adenomatous polyps Carcinomas

,10%

10–50%

.50%

Total

50 (100%) 36 (72%) 2 (13%)

0 10 (20%) 6 (40%)

0 4 (8%) 7 (47%)

50 50 15

11 (22%)

22 (44%)

17 (34%)

50

cancers were significant (P 5 ,0.0001 and 0.0018, respectively), but the differences between the focal carcinomas and carcinomas was not significant (P 5 0.381). These findings indicate that during the transition from normal mucosa to adenomatous polyps to carcinomas there is a progressive increase in pRb expression. No significant correlations were found between the extent of pRb expression and other clinicopathologic parameters (Table 1), but we should emphasize that only a relatively small number of cases were available in each data set. Phosphorylation status of pRb was also examined by Western blot analyses with the same antibody that was used for immunostaining. Ten normal mucosa expressed solely underphosphorylated pRb, whereas 9 of 10 cancer tissues exhibited both hyper- and underphosphorylated pRb. In the adenomatous polyps, seven of eight samples displayed mainly the underphosphorylated pRb (data not shown). Effects of AS-Rb on pRb Expression and Cell Growth. An antisense phosphorothioate oligo to Rb mRNA (AS-Rb) was designed to target the translation start site, and the corresponding S-Rb served as a control, as described previously (20 –22). The lipofectin reagent (Life Technologies, Inc.) was used as the vehicle. To assess the self-association properties of these oligos, 0.5 OD of the AS-Rb or S-Rb was subjected to PAGE in a nondenaturing gel (12%), and the gels were stained with ethidium bromide solution. A single band was seen, which migrated at the appropriate rate for a monomer. No higher-order structures were observed (data not shown). To determine the effects of reducing the level of pRb expression in the HCT116 colon cancer cell line, the cells were transfected with 1 mM AS-Rb, 1 mM S-Rb, or only lipofectin, as described in “Materials and Methods.” Cell extracts were harvested 46 h later when the cells were confluent or still growing and examined by Western blot analysis with a pRb antibody. With extracts of the confluent cultures only single Rb bands were detected that were about 110 kDa (underphosphorylated forms of pRb; Fig. 2A, top). The treatment with AS-Rb led to a marked reduction in the level of pRb. Densitometric analysis indicated that the residual level was only about 30% that of the control culture, which was treated only with lipofectin. In contrast, treatment of the cells with S-Rb, a sense Rb oligonucleotide, led to only a slight

Fig. 2 Effects of AS-Rb and S-Rb on the expression of pRb (A) and on growth (B) in HCT116 cells. A, cells were plated at 1–2 3 105 in a 6-cm diameter dish or at 3–5 3 105 in a 10-cm dish (for the confluent cultures 5 3 105 cells/6-cm diameter dish were seeded) and then transfected with 1 mM AS-Rb, S-Rb, or lipofectin alone. At 46 h after transfection, extracts were prepared and examined by Western blot analysis for the level of expression of pRb. B, for the growth studies, 2 3 104 cells were plated in 24-well plates (1.5-cm diameter) in triplicate for each treatment 2 days before transfection, and the cell number was about 8 3 104 at the time of transfection. At 24, 36, and 48 h after transfection, the cells were trypsinized and counted in a Coulter counter to provide growth curves. The data indicate the mean values and SDs for triplicate assays.

(,10%) reduction in the level of pRb. In the growing condition, the control culture treated only with lipofectin displayed a prominent doublet, consisting of both the hyperphosphorylated (higher molecular weight) and underphosphorylated (lower molecular weight) forms of pRb (Fig. 2A, bottom). Similar results were obtained with untreated HCT116 cells (data not shown). The treatment with AS-Rb reduced the total level of pRb by about 70%, and the residual pRb protein was mainly the hyperphosphorylated form. Equal loading of these lanes was confirmed either by Coomassie blue staining or by parallel Western blotting with an antiactin antibody (data not shown).

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

Clinical Cancer Research 1809

Fig. 3 A, morphologic changes in HCT116 cells. a, at 48 h after transfection with AS-Rb, the HCT116 cells displayed cytoplasmic enlargement and irregular shaped cell morphologies. b, comparable cultures transfected with S-Rb exhibited no morphological change. Magnification, 3100. B, in situ TUNEL assays for apoptosis. For these assays, 1–2 3 104 cells were plated in duplicate in 24-well plates (1.5-cm diameter) and were treated with the oligos or only with lipofectin. Forty-eight hours later, floating cells and adherent cells were collected, pooled, and centrifuged. TUNEL assays were performed using an in situ apoptosis detection kit, ApopTag (Oncor, Gaithersburg, MD), as recommended by the manufacturer. A representative fluorescence microscopy photograph of the cells at 48 h after treatment with AS-Rb displays chromatin condensation and nuclear fragmentation characteristic of apoptosis. Magnification, 380. C, cell kinetics in HCT116 cultures treated with the AS-Rb. Exponential cultures of HCT116 cells were transfected with either AS-Rb or S-Rb and, 48 h later, analyzed by DNA flow cytometry, as described in “Materials and Methods.” The figure also indicates the mean percentage of the cell population in each phase of the cell cycle from three independent experiments. Control cultures gave results similar to those seen with the S-Rb-treated culture (data not shown). The AS-Rb-treated cells displayed a subdiploid DNA fraction characteristic of apoptosis.

Growth curves indicated that the culture treated with AS-Rb displayed marked growth inhibition when compared with the lipofectin-treated control culture; but, the growth of the culture treated with S-Rb was similar to that of the control culture (Fig. 2B). These effects of AS-Rb, with respect to causing a decrease in the level of pRb expression and markedly inhibiting the growth of HCT116 cells, were reproduced in three additional experiments and with three independent preparations of AS-Rb and S-Rb. In contrast to the effects obtained with the HCT116 cells, when cultures of WI38 human lung fibroblasts were treated under identical conditions with AS-Rb, a reproducible 1.5–2.0 fold increase in growth was seen when compared with the lipofectin-treated control culture at 48 h after transfection, but the S-Rb had no significant effect on growth (data not shown). The latter result is consistent with previous evidence that AS-Rb stimulates the growth of HEL human lung fibroblasts and keratinocytes (21, 22).

Effects of AS-Rb on Cell Morphology, Apoptosis, and Cell Cycle Kinetics in HCT116 Cells. Treatment of monolayer cultures of HCT116 cells with AS-Rb under the conditions described above not only inhibited growth but also increased the number of nonadherent cells. At 48 h, the number of floating cells, expressed as percentage of the total cell culture, for the cultures treated with lipofectin alone, S-Rb or AS-Rb, were: 4.8 6 0.01%, 7.3 6 0.36%, and 13.3 6 0.92%, respectively. The difference between AS-Rb- and S-Rb-treated cultures was significant (P 5 0.014). In addition, by 48 h some adherent cells treated with AS-Rb displayed triangular or irregular shaped cell morphologies, and other adherent cells had an enlarged cytoplasm (Fig. 3Aa). The S-Rb-treated adherent cells were round and had a thin cytoplasm (Fig. 3Ab); their morphology was similar to that of cultures treated only with lipofectin or the untreated HCT116 cultures (data not shown). To investigate whether the above-described effects of

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

1810 Antisense to Rb in Colon Cancer Cells

AS-Rb on growth inhibition and morphological changes in HCT116 cells were associated with apoptosis, we carried out in situ TUNEL assays to assess the extent of DNA cleavage. The culture treated with AS-Rb for 48 h displayed numerous positive cells. A representative photograph of the AS-Rb-treated cells is shown in Fig. 3B. The changes of the fluorescein-stained cells were characteristic of apoptosis (i.e., nuclear condensation and fragmentation). The cultures treated with lipofectin or S-Rb displayed only an occasional positive staining cell, and control samples in which PBS was added instead of the TdT enzyme gave negative results (data not shown). The percentage of TUNEL-positive cells was calculated (see “Materials and Methods”), and the values for the cultures treated with lipofectin alone, S-Rb or AS-Rb, were: 1.64 6 0.81%, 2.87 6 0.64%, and 17.39 6 0.47%, respectively. Thus, when compared with the lipofectin-treated culture, S-Rb caused less than a 2-fold increase, but AS-Rb caused about a 10-fold increase in apoptosis. The difference between AS-Rb- and S-Rb-treated cultures was significant (P 5 0.0015). These TUNEL assays were repeated three times, and similar results were obtained. DAPI staining also confirmed this difference in apoptosis (P 5 0.003), with the following values of percentage of apoptotic cells: lipofectin, 2.24 6 0.25%; S-Rb, 4.12 6 0.17%; and AS-Rb, 12.94 6 0.61%. To further assess the mechanism of growth inhibition of HCT116 cells by AS-Rb, we analyzed cell cycle kinetics by DNA flow cytometry of cultures treated for 48 h with either S-Rb or AS-Rb. The results obtained with the S-Rb were similar to those obtained with cultures treated only with lipofectin (data not shown). The mean values and SD for each cell cycle phase were determined from three independent experiments and are shown in Fig. 3C. There was no significant difference in the fraction of cells in each phase of the cell cycle between AS-Rband S-Rb-treated cells. However, the AS-Rb-treated cells displayed an increase in the early sub-G1 DNA fraction, providing further evidence for an increase in apoptotic cells (Fig. 3C). Effects of AS-Rb on the Expression of Other Cell CycleRelated Genes. It was of interest to examine the effects of reducing the level of pRb in HCT116 cells on the expression of various positive and negative regulators of the cell cycle. HCT116 cells were transfected with either S-Rb or AS-Rb, and 24, 32, or 46 h later, extracts were prepared and examined by Western blot analyses for levels of expression of several proteins (Fig. 4A). The intensities of the various bands were quantitated by densitometry. Treatment of the cells with AS-Rb led to about a 50% reduction in the levels of cyclin D1 and p21Waf1 and about a 30% reduction in the levels of p27Kip1, cyclin E, and CDK4. Slight or no major changes were seen in the levels of cyclin A, CDK6, or CDK2. We did not detect the p16Ink4 protein in any of the HCT116 cultures, presumably reflecting inactivation of this gene, although when used as a positive control, WI38 cells gave a strong p16Ink4 band on a Western blot. In vitro kinase assays were also performed with the 46-h protein extracts (Fig. 4B). Assays for cyclin D1-associated kinase activity using a GST-Rb fusion protein as substrate indicated that there was an appreciable reduction of this activity in the AS-Rb-treated cells. This activity was only 48 6 5% that obtained with the extract from the S-Rb-treated cells. On the

Fig. 4 A, effects of AS-Rb on the expression of cyclins, CDKs, and CKIs. HCT116 cells were transfected with S-Rb or AS-Rb, and cell extracts were prepared 24, 32, and 46 h later. Protein (50 mg) was applied to each lane and subjected to PAGE and Western blot analysis with the appropriate antibodies. The relative intensities of the protein bands were determined by densitometry. For additional details see “Materials and Methods.” B, in vitro kinase assays. Cells (3–5 3 105) were plated in a 10-cm dish and then treated with oligos. Forty-six h later, the cells were harvested and lysed with kinase lysis buffer [50 mM HEPES, 150 mM NaCl, 2.5 mM EGTA, 1.0 mM EDTA, 1.0 mM DTT, 0.1% Tween 20, 10% Glycerol, 10 mM glycerophosphate, 1.0 mM NaF, 0.1 mM Na3VO4, 10 mg/ml leupeptin, 10 mg/ml aprotinin, and 1.0 mM PMSF (pH 7.5)]. After sonification, the extracts were clarified at 15,000 3 g for 10 min at 4°C, and the supernatant fraction was collected. In vitro assays for cyclin D1-, cyclin E-, and CDK2-associated kinase activities were performed as described previously (23).

other hand, there was no major change in cyclin-E-associated or CDK2-associated histone H1 kinase activity. The studies shown in Fig. 4 were repeated three times and gave similar results. E2F Reporter Assays. Because it seems that the major function of pRb in regulating the G1 to S transition of the cell cycle is to inhibit activity of the transcription factor E2F, it was of interest to use an E2F-luciferase reporter to examine E2F activity in HCT116 cells cotransfected with S-Rb or AS-Rb. The luciferase activity was normalized by performing simultaneous cotransfection assays for b-galactosidase (see “Materials and Methods”). Treatment with S-Rb caused a slight (,3.5-fold) increase in E2F activity, but treatment with AS-Rb caused about a 17-fold increase in E2F activity (P 5 0.003) (Fig. 5A). Similar results were obtained in two additional studies (data not shown). The finding of a marked increase in E2F activity in the AS-Rb-treated cells suggested that this might be the reason for the induction of apoptosis because enhancement of E2F activity can induce apoptosis in other types of carcinoma (27). Therefore, we examined the direct effect of E2F overexpression on the growth of HCT116 cells, using the following plasmids; CMV wild-type E2F, CMV mutant E2F, and a CMV vector control plasmid. To determine the level of E2F activity that could be induced with the wild-type E2F plasmid, we first carried out

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

Clinical Cancer Research 1811

Fig. 5 A, luciferase reporter assays for E2F activity. The HCT116 cells were plated at 0.5 3 105/well (3.5-cm diameter) in triplicate 48 h before transfection. The E2F-Luc (2 mg) and CMV-b gal (1 mg) plasmids were cotransfected together with either AS-Rb or S-Rb oligos or with only lipofectin. Twenty-four hours after transfection, the cells were washed twice with cold PBS and lysed with 300 ml of reporter lysis buffer (Promega) for 15 min at room temperature. After centrifugation, 30 ml of supernatant fraction was used for the luciferase assays, as described previously (25). b-gal activity was determined using the b-Galactosidase Enzyme Assay System (Promega), as recommended by the manufacturer. The luciferase level in each sample was normalized by the corresponding value for b-gal activity. The increase in E2F activity in the AS-Rb-treated cells when compared with the S-Rb-treated cells was highly significant (P 5 0.0031). B, induction of E2F activity by forced expression of wild-type E2F. One mg of wild type E2F, mutant E2F, or control vector was cotransfected with 1 mg of E2F reporter plasmid and 0.5 mg of the CMV-b gal plasmid. Forty-eight h later, cells were harvested and then examined for the luciferase activity. Luciferase assays indicated that the E2F wild type construct yielded increased E2F activity by about 20-fold.

reporter assays. We found that cotransfection of the wild-type E2F plasmid and the E2F reporter plasmid into HCT116 cells yielded about a 25-fold increase in E2F activity when compared with the vector control plasmid, but this marked induction was not seen by cotransfection with the mutant E2F plasmid (Fig. 5B). In an additional assay, the level of E2F activity using the wild-type E2F plasmid was again about 17-fold higher than that with vector control plasmid. This high level was similar to that induced with AS-Rb on growth (Fig. 5A). We then performed colony efficiency assays to examine the effects of this marked increase in E2F activity. The cells were transfected with each of the above plasmids, transfectants were selected with puromycin for 2 weeks, and the puromycin-resistant colonies were stained. A clear inverse correlation was seen between the ability to form colonies and the levels of E2F activity (Fig. 6). Furthermore, during selection, morphological changes similar to those seen in the AS-Rb-treated cells, including cytoplasmic enlargement and

Fig. 6 Colony efficiency assay with overexpression of wild type, or mutant type E2F. Cells were seeded at a density of 1 3 105/well in 6-cm dishes. Twenty-four h later, the vectors encoding wild-type E2F, or mutant type E2F under CMV promoter, and the corresponding control vector were cotransfected with the pBabe plasmid encoding puromycin resistance, at a ratio of 5:1 (2.5 mg of each plasmid and 0.5 mg of pBabe plasmid), using the lipofectin reagent. Forty-eight hours after transfection, the cells were subplated into 15-cm dishes and then selected in the presence of 0.5 mg/ml of puromycin in complete medium, for 2 weeks. Individual drug-resistant clones were stained with Giemsa solution (Sigma).

irregular-shaped cell morphologies, were frequently seen in the cultures transfected with the wild-type E2F plasmid, but not in the cultures transfected with the mutant E2F or vector control plasmids (data not shown). These assays on colony formation were repeated three times, and similar results were obtained.

DISCUSSION In the present study, we performed immunostaining of the pRb protein in normal human colonic mucosa, adenomatous polyps, early carcinoma in polyps, and advanced colon cancers. To our knowledge, this is the first immunohistochemical analysis for pRb expression using boundary lesions such as adenoma and carcinoma in adenoma. In the normal mucosa, pRb was located in the transitional zone of the crypt at the lower part of the glands where there is a homeostatic state of equilibrium between cell proliferation and cell differentiation (28). It was of interest that in the normal crypt, lack of expression of pRb was observed in the upper portion of crypt where growth arrest and apoptosis occurs (29, 30). This is consistent with the findings of our in vitro study (Figs. 2– 4), indicating that decreased expression of pRb, induced by AS-Rb, leads to growth inhibition and apoptosis in HCT116 colon cancer cells. Thus, if loss of Rb expression is mechanistically associated with growth inhibition and apoptosis in the normal crypt, the maintenance of increased Rb in the colonic epithelial cell may prevent this negative growth process from occurring. Previous studies (5– 8) have found that there is often increased expression of pRb in human colorectal carcinomas. The present study confirms this finding

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

1812 Antisense to Rb in Colon Cancer Cells

and, furthermore, demonstrates that this increase is progressive during the multistage process of colorectal carcinogenesis because we found a progressive increase in pRb expression during the transition from normal mucosa to adenomatous polyps to carcinomas (Table 1 and Fig. 1). This is surprising because in previous studies (10, 17) we found that during colorectal carcinogenesis there is an increase in staining for the proliferation marker Ki67, and also an increase in the expression of cyclin D1, which would be expected to enhance cell proliferation. Western blot analyses for pRb with the same antibody used for immunostaining indicated that this antibody reacts with both the hyper- and underphosphorylated forms of pRb, as described previously (7, 31). It is known that the percentage of phosphorylation of pRb is higher in colon carcinoma tissues than in normal mucosa, thus suggesting a mechanism for inactivation of pRb in the colon carcinomas. However, in a subset of colon carcinomas, there is a significant increase in the absolute amount of underphosphorylated pRb (6, 7), which could bind to and inactivate E2F. These findings seem paradoxical for a tumor suppressor gene, which in its active form, inhibits the G1 to S transition in the cell cycle (32). In an attempt to understand this paradox, we carried out mechanistic studies in vitro to examine the effects of decreasing the level of pRb expression in the human colon cancer cell line HCT116, which expresses a high level of this protein. For this purpose, we used an antisense oligonucleotide sequence (ASRb) targeted to Rb mRNA because previous investigators had shown the specificity of this AS-Rb in other cell systems (20 – 22). We found that introduction into the HCT116 cell line of the AS-Rb caused about a 70% inhibition in the level of expression of pRb. This was associated with a marked inhibition of cell growth in vitro, changes in cell morphology, induction of apoptosis, altered levels of expression of several cell cycle related proteins, inhibition of cyclin D1 associated kinase activity, and a marked increase in E2F activity. These pleiotropic effects were highly reproducible and seem to be secondary to the reduced levels of pRb because of their time course and the fact that they occurred with relatively low concentrations of AS-Rb (1 mM). A similar sense Rb oligonucleotide, S-Rb, had a slight effect on some of these parameters, perhaps due to partial inhibition of translation of Rb mRNA, but the effects of the AS-Rb were always much more dramatic (Figs. 2–5). In addition, a preliminary study on tumorigenicity in nude mice showed an apparent growth inhibition of HCT116 cells when treated with AS-Rb (data not shown). As previously emphasized (33), studies using antisense oligos must be interpreted with caution because of the possibility of obtaining nonspecific effects with these materials. In the present study, we followed several guidelines (34, 35) to minimize such effects. These included using a relatively short (18 mer) oligomer that does not have apparent higher order structure, using a low concentration of the oligomer (1 mM), and using the cationic lipid carrier lipofectin (which allows us to use a lower concentration of the oligomer and also reduces the possibility of degradation of the oligomer to toxic derivatives). The use of lipofectin as a carrier (36, 37) also permits egress of charged oligos from the intracellular vesicular compartment. Thus, the extracellular concentration of phosphorothioate oligos can be kept low, and nonsequence specificity can be avoided.

Furthermore, it is clear that our AS-Rb exerted the expected biochemical effects because the treated cells displayed a marked decrease in the level of pRb and a marked increase in E2F activity, and these effects were not seen with the S-Rb construct (Figs. 2 and 5). In addition, it is unlikely that the phenotypic effects of the AS-Rb on HCT116 cells were due to nonspecific toxicity because, as mentioned below, treatment of WI38 cells with the same AS-Rb construct actually stimulated the growth of these cells. Previous studies indicated that introduction of a full-length Rb cDNA or micro injection of pRb into Rb-deficient osteosarcoma cells or melanoma cells inhibited growth (32, 38) and that antisense oligos to Rb stimulated the growth of human fibroblast and keratinocytes (21, 22). We also confirmed that transfection of our AS-Rb into WI38 human lung fibroblasts under the identical conditions used with HCT116 cells stimulated growth. These findings are what would be expected with a tumor suppressor gene. The growth inhibition seen when the level of pRb was reduced in HCT116 cells is, therefore, somewhat surprising. This situation is not, however, unique since a previous study indicated that an antisense Rb oligonucleotide also inhibited the growth of rat hepatocytes (20). In addition, although introduction of a wild-type Wilms’ tumor suppressor gene (WT1) cDNA into NIH3T3 cells inhibited proliferation (39), treatment of chronic leukemia cells with an WT1 antisense oligonucleotide inhibited growth (19, 40). Thus, the phenotype effects of decreasing the expression of certain tumor suppressor genes is highly dependent on cell context. The reduction in the level of pRb in the HCT116 cells was associated with changes in several other cell cycle control genes. The most striking change was a decrease in the level of cyclin D1 (Figure 4A). This, plus a decrease in the level of CDK4, probably explains the marked decrease in cyclin D1-associated kinase activity (Fig. 4B). The latter effect may be responsible, at least in part, for the overall growth inhibition seen in the AS-Rb-treated HCT116 cells. Our finding that decreased expression of pRb was associated with decreased expression of cyclin D1 and p27Kip1 is consistent with previous evidence for the existence of feedback control mechanisms that coordinate the levels of these three proteins. Thus, pRb-deficient tumors often have decreased levels of cyclin D1 (41), and increased expression of cyclin D1 is often associated with increased levels of pRb and p27Kip1 in human esophageal and colon carcinoma cells (42, 43). It is of interest that the AS-Rb-treated cells also displayed some reduction in the levels of p21Waf1 (Fig. 4A). The decreased levels of p21Waf1, and possibly CDK4, in the AS-Rbtreated cells might be secondary to the decrease in cyclin D1 because there is evidence that cyclin D1 can induce p21Waf1 expression (44). Furthermore, in unpublished recent studies, we have found that retrovirus-mediated transduction of a cyclin D1 cDNA sequence into HCT116 cells results in a marked increase in the expression of pRb, p27Kip1, p21Waf1, and CDK4. It is also of interest that increased expression of pRb, cyclin D1, and p27Kip1 is often seen in primary colorectal carcinomas (5– 8, 10, 17–18), and concurrent overexpression of cyclin D1 and CDK4 is seen in the intestinal adenomas of patients with familial adenomatous polyposis coli (45). Thus, the abrupt drop in cellular levels of pRb in the HCT116 cells, which occurred in the present study after transfection of AS-Rb, could disrupt a

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

Clinical Cancer Research 1813

complex network of feedback control mechanisms in colon cancer cells. It is also of interest that treatment of the HCT116 cells with AS-Rb induced apoptosis, as assessed by the TUNEL assay, DAPI staining, and DNA flow cytometry (Fig. 3). These results suggests that the high level of expression of pRb in HCT116 colorectal carcinoma cells might have a selective advantage by protecting them from apoptosis. This finding is consistent with evidence that pRb can protect osteosarcoma cells, bladder carcinoma, and hepatic carcinoma cells from apoptosis induced by ceramide, INF-g, ionizing radiation and transforming growth factor-b (20, 46 – 48). Using an E2F-luciferase reporter assay, we found that treatment of HCT116 cells with AS-Rb led to a 17-fold increase in E2F activity as compared with control cultures treated with lipofectin (Fig. 5A). Although E2F can function as either an oncogene or tumor suppressor gene, depending on the cellular environment (49, 50), our findings of the toxicity and morphological changes with forced expression of wild-typeE2F (Figs. 5B and 6) suggest that dysregulation of E2F activity may be responsible for the induction of apoptosis and also contribute to the growth inhibition seen in the AS-Rb-treated HCT116 cells. Similar effects of E2F on apoptosis were seen in several cell lines, including breast and ovarian cancer cells (27, 50, 51). However, we do not know the mechanism that mediates the apoptosis because the AS-Rb-treated HCT116 cells showed no change in the level of expression of p53, Bax, or Bcl-2, although they did show a decrease in Bcl-XL (data not shown). Nor can we rule out the possibility that other transcription factors that bind to pRb (52, 53) may also be involved in the growth inhibition and apoptosis seen with AS-Rb. Our strategy to transfect the AS-Rb into another colon cancer cell line, SW480, to reduce the level of pRb has thus far not been successful, probably because the transfection efficiency using lipofectin was much higher in the HCT116 cells (data not shown). However, the present findings are not limited to HCT116 cells because CMV plasmids encoding wild-type E2F or Rb cDNA in the antisense orientation significantly inhibited colony formation in the SW480 colon cancer cell line, as well as in HCT116 cells (data not shown). Because the status of the p53 gene is wild type in HCT116 cells and mutant in SW480 cells (54, 55), these findings are consistent with a previous study indicating that overexpression of E2F can induce apoptosis in breast cancer cell lines, irrespective of their p53 status (27). Regardless of the precise mechanisms, it is apparent from the present study that a decrease in the level of pRb in HCT116 cells adversely affects their growth and viability, perhaps because the increased level of pRb maintains a homeostatic balance between positive and negative factors that control the cell cycle and cellular proliferation at least in certain colon carcinoma cell lines. Indeed, it seems that a majority of colon cancer cases display overexpression of the E2F-1 protein (56). Homeostatic mechanisms may also explain why a subset of some human tumors express relatively high levels of two additional negative regulators of the cell cycle, p27Kip1 (17–18, 57–59) and p21Waf1 (60 – 62). Evidence is accumulating that increased expression of pRb occurs in subsets of bladder, leukemia, and breast cancer (31, 63– 66). It is also likely that increased expression of pRb in some tumors may be the results of loss of

upstream pRb-regulatory genes such as p16 (67). Taken together, these findings suggest that the multistage process of carcinogenesis does not simply involve the step-wise activation of growth-promoting proto-oncogenes and inactivation of growth inhibitory tumor suppressor genes. The evolving population of tumor cells may also have to upregulate the expression of certain negative-acting genes to maintain a homeostatic balance that favors optimal growth and viability. This concept would help to explain the complex and heterogeneous phenotypes of cancer cells and might also have implications with respect to novel approaches to cancer chemoprevention and therapy.

ACKNOWLEDGMENTS We thank Dr. S. Chellappan (Columbia University, New York, NY) for providing the plasmid used for the E2F luciferase reporter assay; Dr. W-H. Lee (The University of Texas Health Science Center at San Antonio, San Antonio, TX) for providing the wild type and mutant E2F expression plasmids; Y. J. Zhang, E. Okin, and W-Q. Xing for excellent assistance; and other members of the Weinstein Laboratory for helpful advice and discussions. We are also grateful to B. Castro and P. Jean-Louis for valuable assistance in the preparation of this manuscript.

REFERENCES 1. Friend, S. H., Bernards, R., Rogelj, S., Weinberg, R. A., Rapaport, J. M., Albert, D. M., and Dryja, T. P. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature (Lond.), 323: 643– 646, 1986. 2. Lee, W. H., Bookstein, R., Hong, F., Young, L. J., Shew, J. Y., and Lee, E. Y. Human retinoblastoma susceptibility gene: cloning, identification, and sequence. Science, 235: 1394 –1399, 1987. 3. Fung, Y. K., Murphree, A. L., T’Ang, A., Qian, J., Hinrichs, S. H., and Benedict, W. F. Structural evidence for the authenticity of the human retinoblastoma gene. Science (Washington DC), 236: 1657– 1661, 1987. 4. Wildrick, D. M., and Boman, B. M. Does the human retinoblastoma gene have a role in colon cancer? Mol. Carcinog., 10: 1–7, 1994. 5. Gope, R., Christensen, M. A., Thorson, A., Lynch, H. T., Smyrk, T., Hodgson, C., Wildrick, D. M., Gope, M. L., and Boman, B. M. Increased expression of the retinoblastoma gene in human colorectal carcinomas relative to normal colonic mucosa. J. Natl. Cancer Inst., 82: 310 –314, 1990. 6. Gope, R., and Gope, M. L. Abundance and state of phosphorylation of the retinoblastoma susceptibility gene product in human colon cancer. Mol. Cell. Biochem., 110: 123–133, 1992. 7. Yamamoto, H., Monden, T., Ikeda, K., Izawa, H., Fukuda, K., Fukunaga, M., Tomita, N., Shimano, T., Shiozaki, H., and Monden, M. Coexpression of cdk2/cdc2 and retinoblastoma gene products in colorectal cancer. Br. J. Cancer, 71: 1231–1236, 1995. 8. Poller, D. N., Baxter, K. J., and Shepherd, N. A. p53 and Rb1 protein expression: are they prognostically useful in colorectal cancer? Br. J. Cancer, 75: 87–93, 1997. 9. Hunter, T., and Pines, J. Cyclins and cancer II: cyclin D and CDK inhibitors come of age. Cell, 79: 573–582, 1994. 10. Arber, N., Hibshoosh, H., Moss, S. F., Sutter, T., Zhang, Y., Begg, M., Wang, S., Weinstein, I. B., and Holt, P. R. Increased expression of cyclin D1 is an early event in multistage colorectal carcinogenesis. Gastroenterology, 110: 669 – 674, 1996. 11. Yasui, W., Kuniyasu, H., Yokozaki, H., Semba, S., Shimamoto, F., and Tahara, E. Expression of cyclin E in colorectal adenomas and adenocarcinomas: correlation with expression of Ki-67 antigen and p53 protein. Virchows Arch., 429: 13–19, 1996. 12. Kitahara, K., Yasui, W., Kuniyasu, H., Yokozaki, H., Akama, Y., Yunotani, S., Hisatsugu, T., and Tahara, E. Concurrent amplification of

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

1814 Antisense to Rb in Colon Cancer Cells

cyclin E and CDK2 genes in colorectal carcinomas. Int. J. Cancer, 62: 25–28, 1995. 13. Leach, F. S., Elledge, S. J., Sherr, C. J., Willson, J. K., Markowitz, S., Kinzler, K. W., and Vogelstein, B. Amplification of cyclin genes in colorectal carcinomas. Cancer Res., 53: 1986 –1989, 1993. 14. Herman, J. G., Merlo, A., Mao, L., Lapidus, R. G., Issa, J. P., Davidson, N. E., Sidransky, D., and Baylin, S. B. Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Res., 55: 4523– 4530, 1995. 15. Matsushita, K., Kobayashi, S., Kato, M., Itoh, Y., Okuyama, K., Sakiyama, S., and Isono, K. Reduced messenger RNA expression level of p21 CIP1 in human colorectal carcinoma tissues and its association with p53 gene mutation. Int. J. Cancer, 69: 259 –264, 1996. 16. Polyak, K., Hamilton, S. R., Vogelstein, B., and Kintzler, K. W. Early alterations of cell-cycle-regulated gene expression in colorectal neoplasia. Am. J. Pathol., 149: 381–387, 1996. 17. Ciaparrone, M., Yamamoto, H., Yao, Y., Sgambato, A., Cattoretti, G., Tomita, N., Monden, T., Rotterdam, H., and Weinstein, I. B. Localization and expression of p27Kip1 in multistage colorectal carcinogenesis. Cancer Res., 58: 114 –122, 1998. 18. Loda, M., Cukor, B., Tam, S. W., Lavin, P., Fiorentino, M., Draetta, G. F., Jessup, J. M., and Pagano, M. Increased proteasome-dependent degradation of the cyclin-dependent kinase inhibitor p27 in aggressive colorectal carcinomas. Nat. Med., 3: 231–234, 1997. 19. Algar, E. M., Khromykh, T., Smith, S. I., Blackburn, D. M., Bryson, G. J., and Smith, P. J. A WT1 antisense oligonucleotide inhibits proliferation and induces apoptosis in myeloid leukaemia cell lines. Oncogene, 12: 1005–1014, 1996. 20. Fan, G., Ma, X., Kren, B. T., and Steer, C. J. The retinoblastoma gene product inhibits TGF-b1 induced apoptosis in primary rat hepatocytes and human HuH-7 hepatoma cells. Oncogene, 12: 1909 –1919, 1996. 21. Strauss, M., Hering, S., Lieber, A., Herrmann, G., Griffin, B. E., and Arnold, W. Stimulation of cell division and fibroblast focus formation by antisense repression of retinoblastoma protein synthesis. Oncogene, 7: 769 –773, 1992. 22. Koike, M., Ishino, K., Ikuta, T., Huh, N., and Kuroki, T. Growth enhancement of normal human keratinocytes by the antisense oligonucleotide of retinoblastoma susceptibility gene. Oncogene, 10: 117–122, 1995. 23. Yamamoto, H., Soh, J. W., Shirin, H., Xing, W. Q., Lim, J. T., Yao, Y., Slosberg, E., Tomita, N., Schieren, I., and Weinstein, I. B. Comparative effects of overexpression of p27Kip1 and p21Cip1/Waf1 on growth and differentiation in human colon carcinoma cells. Oncogene, 18: 103–115, 1999. 24. Wong, K. K., Zou, X., Merrell, K. T., Patel, A. J., Marcu, K. B., Chellappan, S., and Calame, K. v-Abl activates c-myc transcription through the E2F site. Mol. Cell. Biol., 15: 6535– 6544, 1995. 25. Riggs, K. J., Saleque, S., Wong, K. K., Merrell, K. T., Lee, J. S., Shi, Y., and Calame, K. Yin-yang 1 activates the c-myc promoter. Mol. Cell. Biol., 13: 7487–7495, 1993. 26. Shan, B., Zhu, X., Chen, P. L., Durfee, T., Yang, Y., Sharp, D., and Lee, W. H. Molecular cloning of cellular genes encoding retinoblastoma-associated proteins: identification of a gene with properties of the transcription factor E2F. Mol. Cell. Biol., 12: 5620 –5631, 1992. 27. Hunt, K. K., Deng, J., Liu, T. J., Wilson-Heiner, M., Swisher, S. G., Clayman, G., and Hung, M. C. Adenovirus-mediated overexpression of the transcription factor E2F-1 induces apoptosis in human breast and ovarian carcinoma cell lines and does not require p53. Cancer Res., 57: 4722– 4726, 1997. 28. Ali, A. A., Marcus, J. N., Harvey, J. P., Roll, R., Hodgson, C. P., Wildrick, D. M., Chakraborty, A., and Bomam, B. M. RB1 protein in normal and malignant human colorectal tissue and colon cancer cell lines. FASEB J., 7: 931–937, 1993. 29. Moss, S. F., Scholes, J. V., and Holt, P. R. Abnormalities of epithelial apoptosis in multistep colorectal neoplasia demonstrated by ter-

minal deoxyuridine nick end labeling. Dig. Dis. Sci., 41: 2238 –2247, 1996. 30. Doglioni, C., Pelosio, P., Laurino, L., Marci, E., Meggiolaro, E., Favretti, F., and Barbareschi, M. p21/WAF1/CIP1 expression in normal mucosa and in adenomatous and adenocarcinomas of the colon: its relationship with differentiation. J. Pathol., 179: 248 –253, 1996. 31. Wakasugi, E., Kobayashi, T., Tamaki, Y., Nakano, Y., Ito, Y., Miyashiro, I., Komoike, Y., Miyazaki, M., Takeda, T., Monden, T., and Monden, M. Analysis of phosphorylation of pRB and its regulatory proteins in breast cancer. J. Clin. Pathol., 50: 407– 412, 1997. 32. Goodrich, D. W., Wang, N. P., Qian, Y. W., Lee, E. Y., and Lee, W. H. The retinoblastoma gene product regulates progression through the G1 phase of the cell cycle. Cell, 67: 293–302, 1991. 33. Stein, C. A. Does antisense exist? Nat. Med., 1: 1119 –1121, 1995. 34. Stein, C. A., and Krieg, A. Problems in interpretation of data derived from in vitro and in vivo use of antisense oligodeoxynucleotides. Antisense Res. Dev., 4: 67– 69, 1994. 35. Stein, C. A. How to design an antisense oligodeoxynucleotide experiment: a consensus approach. Antisense Nucleic Acid Drug Dev., 8: 129 –132, 1998. 36. Dean, N. M., and McKay, R. Inhibition of protein kinase C-a expression in mice after systemic administration of phosphorothioate antisense oligodeoxynucleotides. Proc. Natl. Acad. Sci. USA, 91: 11762–11766, 1994. 37. Monia, B. P., Johnston, J. F., Geiger, T., Muller, M., and Fabbro, D. Antitumor activity of a phosphorothioate antisense oligodeoxynucleotide targeted against C-raf kinase. Nat. Med., 2: 668 – 675, 1996. 38. Valente, P., Melchiori, A., Paggi, M. G., Masiello, L., Ribatti, D., Santi, L., Takahashi, R., Albini, A., and Noonan, D. M. RB1 oncosuppressor gene overexpression inhibits tumor progression and induces melanogenesis in metastatic melanoma cells. Oncogene, 13: 1169 – 1178, 1996. 39. Kudoh, T., Ishidate, T., Moriyama, M., Toyoshima, K., and Akiyama, T. G1 phase arrest induced by Wilms tumor protein WT1 is abrogated by cyclin/CDK complexes. Proc. Natl. Acad. Sci., USA, 92: 4517– 4521, 1995. 40. Yamagami, T., Sugiyama, H., Inoue, K., Ogawa, H., Tatekawa, T., Hirata, M., Kudou, T., Akiyama, T., Murakami, A., and Maekawa, T. Growth inhibition of human leukemic cells by WT1 antisense oligodeoxynucleotides: implications for the involvement of WT1 in leukemogenesis. Blood, 87: 2878 –2884, 1996. 41. Muller, H., Lukas, J., Schneider, A., Warthoe, P., Bartek, J., Eilers, M., and Strauss, M. Cyclin D1 expression is regulated by the retinoblastoma protein. Proc. Natl. Acad. Sci. USA, 91: 2945–2949, 1994. 42. Doki, Y., Imoto, M., Han, E. K., Sgambato, A., and Weinstein, I. B. Increased expression of the p27KIP1 protein in human esophageal cancer cell lines that overexpress cyclin D1. Carcinogenesis (Lond.), 18: 1139 –1148, 1997. 43. Arber, N., Doki, Y., Han, E. K., Sgambato, A., Zhou, P., Kim, N. H., Delohery, T., Klein, M. G., Holt, P. R., and Weinstein, I. B. Antisense to cyclin D1 inhibits the growth and tumorigenicity of human colon cancer cells. Cancer Res., 57: 1569 –1574, 1997. 44. Hiyama, H., Iavarone, A., LaBaer, J., and Reeves, S. A. Regulated ectopic expression of cyclin D1 induces transcriptional activation of the cdk inhibitor p21 gene without altering cell cycle progression. Oncogene, 14: 2533–2542, 1997. 45. Zhang, T., Nanney, L. B., Luongo, C., Lamps, L., Heppner, K. J., DuBois, R. N., and Beauchamp, R. D. Concurrent overexpression of cyclin D1 and cyclin-dependant kinase 4 (Cdk4) in intestinal adenomas from multiple intestinal neoplasia mice and human familial adenomatous polyposis patients. Cancer Res., 57: 169 –175, 1997. 46. McConkey, D. J., Goodrich, D., Bucana, C., and Klostergaard, J. The human retinoblastoma gene product suppresses ceramide-induced apoptosis in human bladder tumor cells. Oncogene, 13: 1693–1700, 1996.

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research

Clinical Cancer Research 1815

47. Berry, D. E., Lu, Y., Schmidt, B., Fallon, P. G., O’Connell, C., Hu, S. X., Xu, H. J., and Blanck, G. Retinoblastoma protein inhibits IFN-g induced apoptosis. Oncogene, 12: 1809 –1819, 1996. 48. Haas-Kogan, D. A., Kogan, S. C., Levi, D., Dazin, P., T’Ang, A., Fung, Y. K., and Israel, M. A. Inhibition of apoptosis by the retinoblastoma gene product. EMBO J., 14: 461– 472, 1995. 49. Zwicker, J., Liu, N., England, K., Lucibello, F. C., and Muller, R. Cell cycle regulation of E2F site occupation in vivo. Science (Washington DC), 271: 1595–1597, 1996. 50. Field, S. J., Tsai, F. Y., Kuo, F., Zubiaga, A. M., Kaelin, W. G., Jr., Livingston, D. M., Orkin, S. H., and Greenberg, M. E. E2F-1 functions in mice to promote apoptosis and suppress proliferation. Cell, 85: 549 –561, 1996. 51. Shan, B. and Lee, W. H. Deregulated expression of E2F-1 induces S-phase entry and leads to apoptosis. Mol. Cell. Biol., 14: 8166 – 8173, 1994. 52. Cuevo, R. S., Garrett, S., and Horowitz, J. M. Detection and functional characterization of p180, a novel cell cycle regulated yeast transcription factor that binds retinoblastoma control elements. J. Biol. Chem., 272: 3813–3822, 1997. 53. Chen, P. L., Riley, D. J., Chen-Kiang, S., and Lee, W. H. Retinoblastoma protein directly interacts with and activates the transcription factor NF-IL6. Proc. Natl. Acad. Sci. USA, 93: 465– 469, 1996. 54. Take, Y., Kumano, M., Teraoka, H., Nishimura, S., and Okuyama, A. DNA-dependent protein kinase inhibitor (OK-1035) suppresses p21 expression in HCT116 cells containing wild-type p53 induced by adriamycin. Biochem. Biophys. Res. Commun., 221: 207–212, 1996. 55. Rodrigues, N. R., Rowan, A., Smith, M. E., Kerr, I. B., Bodmer, W. F., Gannon, J. V., and Lane, D. P. p53 mutations in colorectal cancer. Proc. Natl. Acad. Sci. USA, 87: 7555–7559, 1990. 56. Lemass, H., Ryan, E., Mathuna, P. M., Crowe, J., and O’Keane, J. C. Elevated pRB is associated with elevated E2F-1 but not cyclin D1 or cyclin-dependent kinase 4 in the colorectal carcinoma cell cycle. Gastroenterology, 114: G2612, 1998. 57. Esposito, V., Baldi, A., De Luca, A., Groger, A. M., Loda, M., Giordano, G. G., Caputi, M., Baldi, F., Pagano, M., and Giordano, A. Prognostic role of the cyclin-dependent kinase inhibitor p27 in nonsmall cell lung cancer. Cancer Res., 57: 3381–3385, 1997. 58. Sgambato, A., Zhang, Y. J., Arber, N., Hibshoosh, H., Doki, Y., Ciaparrone, M., Santella, R. M., Weinstein, I. B. Deregulated expression of p27Kip1 in human breast cancers. Clin. Cancer Res., 3: 1879 –1887, 1997.

59. Catzavelos, C., Bhattacharya, N., Ung, Y. C., Wilson, J. A., Roncari, L., Sandhu, C., Shaw, P., Yeger, H., Morava-Protzner, I., Kapusta, L., Franssen, E., Pritchard, K. I., and Slingerland, J. M. Decreased level of the cell-cycle inhibitor p27Kip1 protein: prognostic implications in primary breast cancer. Nat. Med., 3: 227–230, 1997. 60. Barbareschi, M., Caffo, O., Doglioni, C., Fina, P., Marchetti, A., Buttitta, F., Leek, R., Morelli, L., Leonardi, E., Bevilacqua, G., Dalla Palma, P., and Harris, A. L. p21WAF1 immunohistochemical expression in breast carcinoma: correlations with clinicopathological data, oestrogen receptor status, MIB1 expression, p53 gene and protein alterations and relapse-free survival. Br. J. Cancer, 74: 208 –215, 1996. 61. Jung, J. M., Bruner, J. M., Ruan, S., Langford, L. A., Kyritsis, A. P., Kobayashi, T., Levin, V. A., and Zhang, W. Increased levels of p21WAF1/Cip1 in human brain tumors. Oncogene, 11: 2021–2028, 1995. 62. Marchetti, A., Doglioni, C., Barbareschi, M., Buttitta, F., Pellegrini, S., Bertacca, G., Chella, A., Merlo, G., Angeletti, C. A., Dalla Palma, P., and Bevilacqua, G. p21 RNA and protein expression in non-small cell lung carcinomas: evidence of p53-independent expression and association with tumoral differentiation. Oncogene, 12: 1319 –1324, 1996. 63. Kornblau, S. M., Xu, H. J., Zhang, W., Hu, S. X., Beran, M., Smith, T. L., Hester, J., Estey, E., Benedict, W. F., and Deisseroth, A. B. Levels of retinoblastoma protein expression in newly diagnosed acute myelogenous leukemia. Blood, 84: 256 –261, 1994. 64. Cote, R. J., Dunn, M. D., Chatterjee, S. J., Stein, J. P., Shi, S. R., Tran, Q. C., Hu, S. X., Xu, H. J., Groshen, S., Taylor, C. R., Skinner, D. G., and Benedict, W. F. Elevated and absent pRb expression is associated with bladder cancer progression and has cooperative effects with p53. Cancer Res., 58: 1090 –1094, 1998. 65. Trudel, M., Mulligan, L., Cavenee, W., Margolese, R., Cote, J., and Gariepy, G. Retinoblastoma and p53 gene product expression in breast carcinoma: immunohistochemical analysis and clinicopathologic correlation. Hum. Pathol., 23: 1388 –1394, 1992. 66. Grossman, H. B., Liebert, M., Antelo, M., Dinney, C. P., Hu, S. X., Palmer, J. L., Benedict, W. F. p53 and RB expression predict progression in T1 bladder cancer. Clin. Cancer Res., 4: 829 – 834, 1998. 67. Otterson, G. A., Kratzke, R. A., Coxon, A., Kim, Y. W., and Kaye, F. J. Absence of p16INK4 protein is restricted to the subset of lung cancer lines that retains wildtype RB. Oncogene, 9: 3375–3378, 1994.

Downloaded from clincancerres.aacrjournals.org on December 20, 2012 Copyright © 1999 American Association for Cancer Research