The p53-Independent Tumoricidal Activity of an Adenoviral Vector Encoding a p27–p16 Fusion Tumor Suppressor Gene

doi:10.1006/mthe.2000.0106, available online at http://www.idealibrary.com on IDEAL

ARTICLE

The p53-Independent Tumoricidal Activity of an Adenoviral Vector Encoding a p27–p16 Fusion Tumor Suppressor Gene Salil D. Patel,* Annie C. Tran,* Ying Ge,* Marina Moskalenko,* Lisa Tsui,* Gautam Banik,* Warren Tom,* Michael Scott,* Lili Chen,* Melinda Van Roey,* Marianne Rivkin,* Michael Mendez,* Jeno Gyuris,† and James G. McArthur*,1 *Cell Genesys Incorporated, 324 Lakeside Drive, Foster City, California 94404 †GPC Biotech Incorporated, One Kendall Square, Cambridge, Massachusetts 02139 Received for publication May 17, 2000, and accepted in revised form June 27, 2000

We describe here that DE1-adenovirus vectors (AV) expressing a p27–p16 fusion molecule, termed W9, induce tumor cell apoptosis when overexpressed in a wide range of tumor cell types. However, in primary human cells derived from a variety of normal tissues, AV–W9 induced minimal apoptosis. In tumor cells AV–W9 demonstrated 5- to 50-fold greater tumoricidal activity than either of the parental molecules p16 and p27. In these studies, AV–W9 elicited apoptosis independent of the p53 and Rb status of the tumor cells. In several murine tumor models AV–W9 demonstrated p53-independent antitumor activity. It completely prevented tumor formation in two ex vivo models, whereas the parental molecules resulted in partial protection. Furthermore, AV–W9 induced tumor regression or suppressed tumor growth when introduced intratumorally into preestablished tumors in mice. This effect may be mediated through tumor cell apoptosis or antiangiogenic activity of AV–W9. Thus, this novel chimeric molecule is more potent and capable of killing a broader spectrum of tumors than the parental p16 and p27 molecules independent of the tumor cell p53 and phenotype and represents a powerful new therapeutic agent for cancer gene therapy. Key Words: p16; p27; adenovirus cancer gene therapy; antiangiogenic.

INTRODUCTION Progression through the cell cycle is regulated by the family of cyclin-dependent kinases (CDKs). Hypophosphorylated pRb binds and represses the activity of the E2F family of transcription factors. CDK2 and CDK4 are responsible for the phosphorylation of Rb that results in the release of E2F proteins and allows the activation of genes required for entry into the S phase of the cell cycle. The CDKs are holoenzymes that are activated by their association with regulatory cyclin subunits. The activity of the CDKs is, in turn, downregulated by the cyclin-dependent kinase inhibitors (CDKi’s). It is the balance of these activities that determines if the cell proliferates or if cell cycle arrest occurs [for a review, see (1–4)]. The CDKi’s consist of proteins from two structurally distinct families, which have two distinct mechanisms of

1To whom correspondence [email protected].

should

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy 1525-0016/00 $35.00

be

addressed.

E-mail:

inhibiting the activity of the CDK complexes to regulate cell cycle progression. The INK4 family consists of p15, p16, p18, and p19. The binding of p16 to CDK4 appears to prevent the formation of CDK4/cyclinD complexes, whereas its binding to the assembled CDK4/cyclinD complex produces a catalytically inactive ternary complex. The second family of CDKi’s is the CIP/KIP family comprising p21, p27, and p57 (4). p27 binds to the active CDK2/cyclinA and CDK2/cyclinE complexes, which leads to the formation of a catalytically inactive ternary complex (5, 6). Mutations in the p16 gene and downregulation of p27 are strongly associated with tumorigenesis, pointing to the importance of these proteins in regulating the cell cycle (7). In fact, mice that are nullizygous for p16 develop spontaneous tumors by 6 months of age (8), and p27 knockout mice exhibit frank organomegaly with all tissues containing increased numbers of smaller cells and increased incidence of spontaneous pituitary tumors (9–11). Both p16 and p27 have been separately shown to have antiproliferative or apoptotic activity when overex-

161

ARTICLE pressed in malignant cells following adenoviral vectormediated gene transfer (12–15). To create more potent cytostatic and cytocidal agents, the activities of both families, p27KIP1 and p16INK4b, were combined in a series of novel chimeric CDKi’s. Among these chimeric CDKi’s were a series of p27–p16 fusions where the p27 moiety bore both 5′ and 3′ deletions designed to increase the protein’s half-life, eliminate potential phosphorylation sites involved in the negative regulation of CDKi activity, and increase the activity of these molecules (16). These chimeric p27–p16 molecules should interact with the CDK4/cyclinD, CDK2/cyclinA, CDK2/cyclinE, and CDK2/cyclinB complexes and inhibit cell cycle progression at multiple points. One particular molecule, termed W9, was observed to be the most potent inhibitor of primary smooth muscle and endothelial cell proliferation when compared to the parental p16 and p27 molecules or several other alternative p27–p16 fusion proteins (16). Based on these studies, we chose to test the activity of p16, p27, and W9 as antitumor agents for gene therapy using adenoviral vectors. We now demonstrate that adenovirus-mediated delivery and overexpression of the W9 protein induce cell cycle arrest and apoptosis in tumor cells, but do not extensively induce apoptosis in nontumor cells. The apoptotic activity was independent of the Rb and p53 status of the target cells, and this tumoricidal activity was recapitulated in a xenograft murine model where AV–W9 was more effective in inducing tumor regression than AV–p16. Thus, W9 is more potent and capable of killing a broader spectrum of tumors than the parental p16 and p27 molecules and provides a powerful new therapeutic agent for cancer gene therapy.

MATERIALS

AND

METHODS

Construction of AV–CDKi. cDNAs encoding the CDKi’s were amplified by PCR with primers designed to add a Kozak sequence at the initiator ATG along with a unique AvrII restriction site at the 5′ end and a 3′ primer that introduced a unique ApaI restriction site. The amplicons were cloned into the AvrII/SalI sites of pLOX, a vector that has a CMV promoter, AvrII/SalI CMV splice donor intron, paired with the alpha-globin splice acceptor and SV40 poly(A) site configuration within the E1 region of Ad5 (0 to 15.8 mu). This construct was used to generate CMV–CDKi–Ad5 in a yeastbased homologous recombination system. This CMV–CDKi–Ad5 construct was grown in bacteria and cut with I-I-SCE1 (rare cutter) to release the viral genome. The linear genome fragment was used to transfect 293 cells to generate recombinant virus that was purified by standard methods (17). Tumor cell lines. Tumor cell lines were purchased from the American Type Culture Collection (Rockville, MD). We tested prostate- (PC-3, LNCaP, and DU145), breast- (MDA-MB-231 and MDA-MB-468), pancreatic- (Capan-1 and Capan-2), lung- (A549), liver- (Hep-G2), colon- (SW480 and LS174T), ovarian- (SK-OV3), retinoblastoma- (Y79), cervical- (Hela), and kidney-derived tumor cell lines (293). Cell culture conditions were as per the instructions from ATCC except for MDA-MB-231, which was grown in DMEM/High with 10% FCS in a 5% CO2 incubator. Cell lines derived from normal tissue were purchased from Clonetics (Biowhittaker, MD). We tested coronary artery smooth muscle cells (CASMC), coronary artery endothelium cells (CAEC), small airway epithelium cells (SAEC), prostate epithelium cells (PrEC), and prostate stromal cells (PrSC). Cells were grown under the conditions provided by Clonetics.

162

Infection of cells with recombinant adenovirus. Cells were seeded for 4 h at 1  105 cells/well in 12-well plates before being infected for 2 h with virus in an overlay volume of 0.5 ml. The cells were washed once with medium and cultured with 2 ml of fresh medium. The cells were collected 2 to 4 days later for analyses. Apoptosis assay. Apoptosis was assessed by Annexin-V and propidium iodide (PI) labeling (18) or by TUNEL staining (19). TUNEL assay. Cells were resuspended and fixed in 1% formaldehyde in PBS for 1 h. Cells were washed and stored in 70% ethanol. Before starting the TdT reaction, cells were washed twice with PBS. Fifty microliters of TdT reaction mixture (1 TdT buffer, CoCl, biotinylated dUTP, terminal transferase enzyme; Boehringer Mannheim, Indianapolis, IN) was added. The mixture was incubated at 37C for 1 h. Samples were then washed once with PBS and incubated with 2.5 µg/ml fluorosceinated streptavidin in 0.1% Triton X-100, 5% nonfat dry milk at room temperature for 30 min. Samples were washed twice with 3 ml of PBS and analyzed by FACS. Samples that were treated similarly without the addition of terminal transferase enzyme served as negative controls. Annexin-V. Floating and adherent infected cells were harvested and washed with ice-cold DMEM (without serum). Cells (105–106) were resuspended in 495 µl of binding buffer (Coulter) and kept on ice. Five microliters of Annexin-V–FITC was added and cells were incubated in the dark on ice for 10 min. Cells were washed with 4 ml of ice-cold DMEM (without serum) and the pellet was resuspended in 495 µl of binding buffer. Five microliters of propidium iodide (250 µg/ml) was added and the mixture stored on ice until FACS analyses. Cells were collected without compensation. WinList software was used for compensation of the parameters after acquisition. CellQuest software (Becton–Dickinson, CA) was utilized to analyze the results. Intracellular CDKi expression. Harvested cells were washed with 5 ml of PBS containing 1% BSA (PBS/BSA) and resuspended in 0.5 ml of icecold PBS. The cell suspension was added into 5 ml of 1% ultrapure formaldehyde in PBS and placed on ice for 15 min. Cells were washed with PBS/BSA and resuspended in 1 ml of ice-cold PBS. Four milliliters of ice-cold ethanol was carefully layered onto the PBS-containing cell suspension, before vortexing for 1 s and leaving at −20C for at least 1 h. Cells were washed, resuspended in 0.1 ml of PBS/BSA containing 20 µl of anti-p16 or anti-p27 mAb (Pharmingen, CA), and incubated for 45 min at room temperature in the dark. Cells were washed and analyzed by FACS. Tumor prevention models. One million tumor cells were infected at a multiplicity of infection units (m.o.i.) of 50 with different AV–CDKi vectors and then harvested and washed 16 h later. Cell viability was checked by trypan blue staining. The infected cells were introduced subcutaneously into Balb-nude mice, and tumor formation was tracked for 2–3 months. Intratumoral treatment of primary tumors. Balb/c-nude mice were injected subcutaneously with 3  106 DU-145 cells or 1  106 PC-3 cells. When tumors reached a volume of 20–40 mm3, mice were treated by intratumoral injection with a dose of 1.25  1011 viral particles that was delivered using a programmable pump microsyringe (the KDS 100 infusion pump, KD Scientific, Boston, MA) formatted to inject 30 µl per minute. All subsequent injections were performed under the same conditions. Ten mice were utilized per treatment group. Tumor size was measured every week for each mouse and volume assessed. Aortic ring sprouting assay. Aortas were isolated from 4- to 6-week-old rats and placed in Hank’s solution (Gibco BRL). Outer fat was removed with tweezers and the aortas were rinsed several times to remove blood and debris. The aortas were then sliced into very thin rings using a scalpel on a dry dish. Aortic rings were infected with AV–W9 or control virus by immersion in 50 µl of PBS containing the virus, for 60 min at 37°C. The infected ring was placed on gelled Matrigel (Becton–Dickinson) in 24well plates and 200 µl of Matrigel was used to seal it in place with the plate cooling on ice. Plates were incubated at 37°C for 30 min and 1 ml of serum-free endothelial cell growth medium (Clonetics) was added to each well. The plates were incubated at 37°C for 7 days, fixed with 1%

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

ARTICLE glutaraldehyde in PBS at room temperature for 20 min, and stained with Diff Quick Staining Solution II for 10 min at room temperature. Ring sprouting was recorded with a digital camera.

RESULTS AV–CDKi Expression The CDKi genes expressed from the CMV promoter/enhancer were introduced into the E1 region of the E1 and E3 region-deleted adenoviral vector (∆E1/∆E3–AV) and recombinant virus was generated. Each virus was tested for CDKi expression by an intracellular FACS staining protocol of transduced cells using an antibody directed against p16 to detect p16 and W9 and an anti-p27 antibody to detect p27. CDKi protein expres-

sion was readily detected by FACS analysis in all cells. Staining observed with AV–CDKi-transduced Capan-1 and A549 cells is shown (Figs. 1A and 1B). To determine the kinetics of transgene expression, a time-course analysis of CDKi expression was performed over a period of 4 days. Expression was detected beginning at day 1 postinfection and reached peak levels at day 2. The percentage of transgene-positive cells was similar in A549 cells transduced by AV–W9 and AV–p16. In A549 tumor cells the total amount of CDKi protein present, as determined by mean fluorescence, was two- to threefold higher in the AV–W9-transduced cells than the AV–p16-transduced cells. This was not, however, observed to be the case in all cell types tested such as Capan-1 (Fig. 1A) and primary human smooth muscle cells (16).

FIG. 1. AV–CDKi expression. Expression of AV–W9, p16, and p27. Capan-1 (A) and A549 (B) cells were transduced with AV–CDKi at a m.o.i. of 100 and expression was assessed by intracellular FACS analyses after 3 days. Solid lines indicate CDKi expression, and dashed lines represent staining in nontransduced cells. Time course of expression of AV delivered W9 and p16. A549 cells were infected with AV–W9 or AV–p16 (50 m.o.i.) and assessed for CDKi expression over a period of 4 days (C). The percentage of transduced cells is represented by the line graph and the fluorescent intensity is represented by the bar graph. W9, black squares and bars; p16, white triangles and bars.

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

163

ARTICLE AV–CDKi Elicits Apoptosis Preferentially in Tumor Cells We tested the effect of adenovirus expressing the different CDKi’s on tumor cell lines using FACS-based TUNEL (19) and Annexin-V labeling (18) assays to detect apoptotic cell death. In all cases, we also tracked transgene expression by intracellular FACS labeling and counted viable cells to assess the effect of AV–CDKi on cell growth. Analysis was performed 2–4 days after virus infection. Levels of apoptosis obtained using prostate-derived PC-3 cells and colon carcinoma-derived SW480 cells infected with AV–CDKi or an AV-null (transgene negative) control virus at a dose range of 0–100 m.o.i. are shown in Figs. 2A and 2B. In both cases shown here and with A549 cells (data not shown), the dose of AV–W9 required to elicit equivalent levels of apoptosis was at least 5-fold lower than AV–p16 and up to 50-fold less than AV–p27. At very high doses, AV-null elicited low levels of death in SW480 cells.

In the Annexin-V labeling assay, it is possible to differentiate early and late apoptotic phases by dual labeling the cells with propidium iodide; cells undergoing early apoptosis are labeled by Annexin-V only, whereas “late apoptotic” cells and necrotic cells are labeled by both reagents. A representative result with AV–W9 and an AV-null control is shown in Fig. 2C. In this example, 72% of the cells are undergoing apoptosis in the AV–W9transduced cells compared to 14.3% of the AV-null-transduced control cells. These cells can be further differentiated into cells undergoing late apoptosis–necrosis (27% for AV–W9 cells vs 8.8% for AV-null cells) and cells just beginning to apoptose (42% for AV–W9 cells vs 4% for AV-null cells). Clearly, cell apoptosis is a continuum and therefore the cumulative effect of the CDKi cannot be easily assessed with these in vitro assays. To determine the breadth of this tumoricidal effect, a panel of 14 tumor cell lines were transduced with AV–W9 or AV–GFP at a m.o.i. of 50 to 100. The relative

FIG. 2. AV–CDKi induces apoptosis in tumor cells. (A) W9 is more effective than AV–p16 or AV–p27 in inducing apoptosis in PC-3 prostate tumor cells. PC-3 cells were infected with Ad-CDKi at m.o.i.’s of 1, 10, 50, and 100 and apoptosis was assessed after 3 days. W9 is more effective than AV–p1 or AV–p27 in inducing apoptosis in SW480 colon carcinoma cells. SW480 cells were infected with AV–CDKi at m.o.i.’s of 1, 10, 50, and 100 and apoptosis was assessed after 3 days. Detection of apoptosis and necrosis by Annexin-V–propidium iodide costaining. A549 cells infected with AV–W9 or AV-null for 3 days were harvested and colabeled with Annexin-V and propidium iodide.

164

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

ARTICLE ability of AV to transduce each line was assessed by determining the percentage of GFP-positive tumor cells by FACS analyses (data not shown). The ability of AV–W9 to induce apoptosis in these tumor lines was assessed using FACS-based TUNEL and Annexin-V labeling assays. As shown in Table 1, AV–W9 induced apoptosis in 12 of the 14 tumor cell lines tested regardless of tumor origin. More importantly, apoptosis appeared to be independent of either Rb or p53 gene status. Apoptosis was detected in cells that were nullizygous for Rb and p53 (DU145), for p53 only (PC-3), and for Rb only (Y79) or normal for both (A549). The genetic typing is based on assays carried out by the NCI Developmental Therapeutics Program [http://epnws1.ncifcrf.gov:2345/dis3d/dtpdata.html] and previously published reports (20–22). W9 protein was expressed in the two tumor cell lines that did not apoptose following transduction with AV–W9. One of these cell lines underwent nonapoptotic cell death (Capan-1) and the second was growth arrested (HepG2). Primary cell cultures derived from normal tissue obtained from human cadavers were also tested for AV–W9-mediated apoptosis. Using the same doses of AV–W9 that were used for the tumor cell study, we observed little or no apoptosis in a series of primary human cells, although these cells were growth arrested. In a previous study examining the impact of W9 overexTABLE 1 Impact of AV-W9 on Normal and Tumor Cells Cell type

Apoptosis

Cell origin

DU145

Yes

PC3

Yes

LnCAP

Yes

Rb status

p53 status

Prostate

Mutant

Mutant

Prostate

Normal

Mutant

Prostate

?

Normal

A549

Yes

Lung

Normal

Normal

SW480

Yes

Colon

?

?

LS174

Yes

Colon

?

?

Y79

Yes

Retinoblastoma

Mutant

Normal

Capan-1

Cell death

Pancreatic

?

?

Capan-2

Yes

Pancreatic

?

?

Hep-G2

Growth arrest

Liver

Normal

Normal

293

Yes

Kidney

Mutant

Mutant

SKOV-3

Yes

Ovarian

?

Mutant

MDA-MB231

Yes

Breast

Normal

Mutant

MB468

Yes

Breast

Mutant

Mutant

Smooth muscle cells

No

Coronary artery

Normal

Normal

Endothelial

Minimal

Coronary artery

Normal

Normal

Epithelial

Minimal

Lung

Normal

Normal

Stromal

Minimal

Prostate

Normal

Normal

Epithelial

Yes

Prostate

Normal

Normal

Note. AV-W9 elicits apoptosis in tumor cells derived from prostate, lung, colon, breast, and retina regardless of the p53, Rb status of the cells. Some apoptosis is elicited in normal cells, although this is generally lower than in tumor cells.

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

pression in human vascular smooth muscle cells, AV–W9 initially arrested the cells in G2/M phase with a progressive block in G1 phase of the cell cycle (16). Similar results were observed with primary vascular endothelial, lung epithelial, and prostate stromal cells(Table 1, bottom panel). FACS analyses for expression confirmed that >99% of the cells were transduced and expressed W9. In prostate epithelial cells, some apoptosis was detected (data not shown); however, it is possible that the phenotype of these cells was altered by their explantation and culture in vitro.

AV–W9 Prevents Tumor Growth in ex Vivo Models To directly assess the cumulative effect of CDKi on cell survival, AV–CDKi-transduced tumor cells were transplanted into mice and the development of tumors was tracked over time. In these tumor prevention models, 106 PC-3 cells (Fig. 3A) or A549 cells (Fig. 3B) were infected with AV–CDKi at a m.o.i. of 50 and these infected cells were subsequently implanted subcutaneously into Balb/nude mice. As shown in Fig. 3, only the AV–W9 pretreatment completely prevented tumor formation in both models. AV–p16 prevented tumor formation in 80% (8 of 10) of the mice. A separate dose titration of the tumor cells showed that 50,000 cells were sufficient for tumor growth suggesting that the cumulative effect of AV–W9 resulted in at least 95% apoptotic cell death. AV–p27 pretreatment had no impact on the growth of PC-3 tumors and only a modest effect on the growth of A549 tumors in spite of the near 100% transduction efficiency of the input cells by the virus. It is interesting that unlike AV–W9 and AV–p16 pretreatments, the antitumor effect of AV–p27 varied for the different cell lines.

AV–W9 Leads to Tumor Regression in an Intratumoral Injection Treatment Model The ability of AV–W9 to eradicate established tumors in immunodeficient mice was assessed in two tumor treatment models (Fig. 4). In the first model, 3 million DU145 cells were implanted subcutaneously into Balb/nude mice. Approximately 2–3 weeks following the injection of the tumor cells, when the tumors were 20–40 mm3 in size, 1.25  1011 viral particles were administered intratumorally. This treatment was repeated every second day for a total of three injections. Tumor size was assessed weekly and its volume calculated according to the formula previously described (23). The average tumor sizes over an observation period of 10 weeks are shown in Fig. 4A. The growth of AV–W9-treated tumors was significantly delayed compared to AVnull-treated tumors. Even compared to the AV–p16-treated tumors there was an appreciable delay in tumor progression in the AV–W9-treated animals. In a second model, established PC-3 xenograft tumors were treated weekly on consecutive days with two doses of 1.25  1011 viral particles of recombinant virus or PBS.

165

ARTICLE

FIG. 3. AV–W9 prevents tumor formation in two xenograft models. (A) One million A549 cells were infected at a m.o.i. of 50 with different AV–CDKi vectors and then harvested and washed 24 h later. These infected cells were introduced subcutaneously into Balb-nude mice (n = 10), and tumor formation was tracked for 12 weeks. (B) One million PC-3 prostate tumor cells were infected at a m.o.i. of 50 with different AV–CDKi vectors and then harvested and washed 24 h later. These infected cells were introduced subcutaneously into Balb-nude mice (n = 10), and tumor formation was tracked for 12 weeks.

By measuring tumor growth (Fig. 4B), we observed potent antitumor activity in the AV–W9-treated group compared to the untreated and null adenovirus-treated animals. By week 13, all of the mice in the two control arms had to be sacrificed, while 50% (5 of 10) of the AV–W9-treated mice remained alive (Fig. 4C) (P =

0.0017). AV–W9 was also superior to AV–p16 in arresting tumor progression as only 20% of the AV–p16-treated animals remained alive following treatment (P = 0.118). In two of the AV–W9-treated mice, tumors were completely eradicated leaving behind scab-like lesions that did not develop further even following treatment termi-

FIG. 4. AV–W9 slows tumor growth in a prostate xenograft tumor model. (A) Subcutaneous DU-145 xenografts (induced using 3 million cells) were treated with AV–W9, AV–p16, or AV-null by intratumoral injections at days 1, 3, and 5 using 1.25 ↔ 1011 virus particles per injection. The starting volume of the tumors was 20–40 mm3 (n = 10). Tumor size was followed for 10 weeks. (B) Subcutaneous PC-3 xenografts (induced using one million cells) were treated with AV–W9, AV–p16, AV-null, or PBS by intratumoral injections at days indicated using 1.25 ↔ 1011 virus particles per injection. The average starting volume of the tumors was 30 mm3 (n = 10). Tumor progression was tracked for 12 weeks. (C) Kaplan–Meier analysis showing percentage of surviving animals from the experiment described in B.

166

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

ARTICLE 5  1010 vp

2.5  1010 vp

5  109 vp

W9

p16

Null

FIG. 5. AV–W9 has antiangiogenic activity. Rat aortic rings were treated with AV–W9, AV–p16, or AV-null and placed on Matrigel for 7 days. Rings were stained and visualized for sprouting using a digital camera. Arrows indicate areas of vessel sprouting.

nation at 3 months. The suppression of tumor growth for the entire treatment period suggests that resistance to AV–W9 treatment did not arise.

Antiangiogenic Activity of AV–W9 We have previously demonstrated that adenovirus expressing W9 can inhibit the proliferation of endothelial cells. Since endothelial cell proliferation and migration are critical steps in angiogenesis (24, 25), we examined the antiangiogenic activity of AV–W9 using an aortic ring sprouting assay (26, 27). Virus infection of the rings was achieved by submerging them in a minimal volume of PBS containing the amount of virus indicated for a period of 1 h. As shown in Fig. 5, AV–W9 clearly inhibited the formation of vessel sprouts in this assay. AV-null was used as a control. At lower doses of virus treatment (5  109 viral particles) some sprouting occurs in AV–p16- and AV–W9-treated rings, but this is significantly lower than the amount of sprouting observed with AV-null-treated rings. At higher doses of over 2.5 

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

1010 viral particles both AV–W9 and AV–p16 completely inhibited sprout formation, although at these doses AVnull also had a partial effect. Figure 5 is a representative result; several rings were used for each experimental point and the experiment was repeated several times.

DISCUSSION Gene transfer of CDKi to tumor cells is a viable antitumor approach that has been tested by several groups (14, 15, 28–34). In a comparison of p16, p18, p19, p21, and p27, overexpression of p16, p27, and p18 was demonstrated to elicit apoptosis in tumor cells (12). In this study, we confirm the observation of apoptosis in tumor cells overexpressing p16 and p27. In addition, we show that W9, a molecule combining p16 and p27, was more effective than either parental molecule. AV–W9 elicited apoptosis in numerous tumor cell lines at doses that were 5- to 50-fold lower than those required for AV–p16 and AV–p27. In previous studies, the activities of the individ-

167

ARTICLE ual p16 and p27 molecules were similar and depended on the tumor cell type and setting in vivo (12). In our analyses, AV–W9 was at least 5-fold more active than p16 and p27 in multiple tumor lines, as well as in arresting growth of SMC and EC (16). In addition, the superiority of W9 was demonstrated in several murine tumor therapy models. In 12 of the 14 human tumor-derived cell lines tested we observed apoptosis by day 2 postinfection which typically peaked by day 3. In some lines, such as Capan1, cell death was observed as late as 4 days postinfection. This delay in the appearance of apoptotic death may occur because the expression of CDKi begins approximately 1 day after infection and peaks 3 days postinfection (see Fig. 1). Expression analyses over a range of viral doses revealed that low levels of CDKi expression were insufficient to elicit apoptosis, suggesting that a threshold level of expression is required to induce apoptosis (data not shown). Improved stability may be one of the reasons for the higher activity of W9 compared to p16 (as shown in Fig. 1). It has been previously reported that AV–p16-expressing tumor cells show enhanced cytotoxicity compared to control viruses within cell lines that express functional pRb (12, 35, 36). We observed that the genetic status, especially the Rb and p53 gene status, did not impact apoptosis with AV–W9. Apoptosis was detected in cells that were null for Rb and p53 (DU145), for p53 only (PC3), and for Rb only (Y79) or normal for both (A549). Combining the activities of p27 and p16 together is not enough to produce this potent p53- and Rb-independent cytotoxic activity. Related p27–p16 fusion molecules that were shown biochemically to possess an active p27 domain (16) were far less efficient in inducing tumor cell apoptosis in vitro and were far more sensitive to the underlying genetic background of the tumor cell (Salil Patel, unpublished results). The general sensitivity of neoplastic cells to apoptosis induction, compared to normal cells, is well described (37). Downregulation of cyclin A/cdk2 activity has been proposed to lead to higher levels of E2F activity which would lead to apoptosis in tumor cells but not in normal cells, since tumor cells already have a high level of E2F activity (38, 39). The superior potency observed for AV–W9 in vitro was recapitulated in vivo. In two treatment models employing two different p53 nullizygous human tumor cell lines, AV–W9 suppressed tumor growth more significantly than AV–p16. Not only did AV–W9 treatment prolong survival in treated mice, but in two of the AV–W9 treatment groups the tumors were eradicated. Expression analysis of cells derived from an injected tumor revealed that 10–20% of the tumor cells express W9 following injection (data not shown). This is consistent with the observation of Schreiber et al. using AV–p16 (12). The modest transduction efficiency of intratumoral administration of AV–W9 suggests that a second mechanism of action might contribute to the antitumor effects of AV–W9. We had previously demonstrated that AV–W9

168

was a potent inhibitor of endothelial cell proliferation (16). From this result, it might be predicted that AV–W9 should inhibit neoangiogenesis. Indeed, in a commonly employed measurement of antiangiogenesis activity, the rat aortic ring assay, AV–W9 suppressed the formation of vascular sprouts. This suggests that AV–W9 mediates a dual therapeutic effect of antiangiogenesis and apoptotic cell death. In summary, W9 is more potent and capable of killing a broader spectrum of tumors than the parental p16 and p27 molecules and represents a powerful new therapeutic agent for cancer gene therapy. ACKNOWLEDGMENT We thank Tammy Langer for technical assistance.

REFERENCES 1Grana, X., and Reddy, E. P. (1995). Cell cycle control in mammalian cells: Role of cyclins, cyclin dependent kinases (CDKs), growth suppressor genes and cyclin-dependent kinase inhibitors (CKIs). Oncogene 11: 211–219. 2Sherr, C. J. (1995). Mammalian G1 cyclins and cell cycle progression. Proc. Assoc. Am. Physicians 107: 181–186. 3Sherr, C. J., and Roberts, J. M. (1995). Inhibitors of mammalian G1 cyclin-dependent kinases. Genes Dev. 9: 1149–1163. 4Sherr, C. J., and Roberts, J. M. (1999). CDK inhibitors: Positive and negative regulators of G1-phase progression. Genes Dev. 13: 1501–1512. 5Russo, A. A., Jeffrey, P. D., Patten, A. K., Massague, J., and Pavletich, N. P. (1996). Crystal structure of the p27Kip1 cyclin-dependent-kinase inhibitor bound to the cyclin A–Cdk2 complex [see comments]. Nature 382: 325–331. 6Russo, A. A., Tong, L., Lee, J. O., Jeffrey, P. D., and Pavletich, N. P. (1998). Structural basis for inhibition of the cyclin-dependent kinase Cdk6 by the tumour suppressor p16INK4a. Nature 395: 237–243. 7Sherr, C. J. (1996). Cancer cell cycles. Science 274: 1672–1677. 8Serrano, M., Lee, H., Chin, L., Cordon-Cardo, C., Beach, D., and DePinho, R. A. (1996). Role of the INK4a locus in tumor suppression and cell mortality. Cell 85: 27–37. 9Nakayama, K., et al. (1996). Mice lacking p27(Kip1) display increased body size, multiple organ hyperplasia, retinal dysplasia, and pituitary tumors. Cell 85: 707–720. 10Fero, M. L., et al. (1996). A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)-deficient mice. Cell 85: 733–744. 11Kiyokawa, H., et al. (1996). Enhanced growth of mice lacking the cyclin-dependent kinase inhibitor function of p27(Kip1). Cell 85: 721–732. 12Schreiber, M., Muller, W. J., Singh, G., and Graham, F. L. (1999). Comparison of the effectiveness of adenovirus vectors expressing cyclin kinase inhibitors p16INK4A, p18INK4C, p19INK4D, p21(WAF1/CIP1) and p27KIP1 in inducing cell cycle arrest, apoptosis and inhibition of tumorigenicity. Oncogene 18: 1663–1676. 13Sandig, V., Brand, K., Herwig, S., Lukas, J., Bartek, J., and Strauss, M. (1997). Adenovirally transferred p16INK4/CDKN2 and p53 genes cooperate to induce apoptotic tumor cell death. Nat. Med. 3: 313–319. 14Wang, X., Gorospe, M., Huang, Y., and Holbrook, N. J. (1997). p27Kip1 overexpression causes apoptotic death of mammalian cells. Oncogene 15: 2991–2997. 15Katayose, Y., et al. (1997). Promoting apoptosis: A novel activity associated with the cyclin-dependent kinase inhibitor p27. Cancer Res. 57: 5441–5445. 16Lamphere, L., et al. Novel chimeric p16 and p27 molecules with increased anti-proliferative activity for vascular disease gene therapy. J. Mol. Med., in press. 17Graham, F. L., and Prevec, L. (1995). Methods for construction of adenovirus vectors. Mol. Biotechnol. 3: 207–220. 18Vermes, I., Haanen, C., Steffens-Nakken, H., and Reutelingsperger, C. (1995). A novel assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labeled Annexin V. J. Immunol. Methods 184: 39–51. 19Sgonc, R., Boeck, G., Dietrich, H., Gruber, J., Recheis, H., and Wick, G. (1994). Simultaneous determination of cell surface antigens and apoptosis. Trends Genet. 10: 41–42. 20Bookstein, R., Shew, J. Y., Chen, P. L., Scully, P., and Lee, W. H. (1990). Suppression of tumorigenicity of human prostate carcinoma cells by replacing a mutated RB gene. Science 247: 712–715. 21Lee, E. Y., Bookstein, R., Young, L. J., Lin, C. J., Rosenfeld, M. G., and Lee, W. H. (1988). Molecular mechanism of retinoblastoma gene inactivation in retinoblastoma cell line Y79. Proc. Natl. Acad. Sci. USA 85: 6017–6021. 22Kondo, I., et al. (1985). A case report of a patient with retinoblastoma and chromosome 13q deletion: Assignment of a new gene (gene for LCP1) on human chromosome 13. Hum. Genet. 71: 263–266. 23Spang-Thomsen, M., Nielsen, A., and Visfeldt, J. (1980). Growth curves of three

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

ARTICLE human malignant tumors transplanted to nude mice. Exp. Cell Biol. 48: 138–154. 24Folkman, J. (1998). Antiangiogenic gene therapy. Proc. Natl. Acad. Sci. USA 95: 9064–9066. 25Claesson-Welsh, L., et al. (1998). Angiostatin induces endothelial cell apoptosis and activation of focal adhesion kinase independently of the integrin-binding motif RGD. Proc. Natl. Acad. Sci. USA 95: 5579–5583. 26Auerbach, R., Auerbach, W., and Polakowski, I. (1991). Assays for angiogenesis: A review. Pharmacol. Ther. 51: 1–11. 27Auerbach, W., and Auerbach, R. (1994). Angiogenesis inhibition: A review. Pharmacol. Ther. 63: 265–311. 28Clayman, G. L., et al. (1996). Gene therapy for head and neck cancer. Comparing the tumor suppressor gene p53 and a cell cycle regulator WAF1/CIP1 (p21). Arch. Otolaryngol. Head Neck Surg. 122: 489–493. 29Craig, C., et al. (1997). A recombinant adenovirus expressing p27Kip1 induces cell cycle arrest and loss of cyclin–Cdk activity in human breast cancer cells. Oncogene 14: 2283–2289. 30Favrot, M., Coll, J. L., Louis, N., and Negoescu, A. (1998). Cell death and cancer: Replacement of apoptotic genes and inactivation of death suppressor genes in therapy. Gene Ther. 5: 728–739. 31Katayose, D., Wersto, R., Cowan, K. H., and Seth, P. (1995). Effects of a recombinant adenovirus expressing WAF1/Cip1 on cell growth, cell cycle, and apoptosis. Cell Growth Differ. 6: 1207–1212.

MOLECULAR THERAPY Vol. 2, No. 2, August 2000 Copyright  The American Society of Gene Therapy

32Katayose, D., Wersto, R., Cowan, K., and Seth, P. (1995). Consequences of p53 gene expression by adenovirus vector on cell cycle arrest and apoptosis in human aortic vascular smooth muscle cells. Biochem. Biophys. Res. Commun. 215: 446–451. 33Katayose, D., Gudas, J., Nguyen, H., Srivastava, S., Cowan, K. H., and Seth, P. (1995). Cytotoxic effects of adenovirus-mediated wild-type p53 protein expression in normal and tumor mammary epithelial cells. Clin. Cancer Res. 1: 889–897. 34Rocco, J. W., et al. (1998). p16INK4A adenovirus-mediated gene therapy for human head and neck squamous cell cancer. Clin. Cancer Res. 4: 1697–1704. 35Sandig, B. V., Brand, K., Herwig, S., Lukas, J., Bartek, J., and Strauss, M. (1997). p16 and p53 genes transferred with the help of adenovirus to induce apoptotic tumor cell death. Ugeskr. Laeger. 159: 6825–6830. 36Craig, C., et al. (1998). Effects of adenovirus-mediated p16INK4A expression on cell cycle arrest are determined by endogenous p16 and Rb status in human cancer cells. Oncogene 16: 265–272. 37Fisher, D. E. (1994). Apoptosis in cancer therapy: Crossing the threshold. Cell 78: 539–542. 38Chen, Y. N., et al. (1999). Selective killing of transformed cells by cyclin/cyclindependent kinase 2 antagonists [see comments]. Proc. Natl. Acad. Sci. USA 96: 4325–4329. 39Lees, J. A., and Weinberg, R. A. (1999). Tossing monkey wrenches into the clock: New ways of treating cancer [comment]. Proc. Natl. Acad. Sci. USA 96: 4221–4223.

169