Therapy (2003) 10, 1961–1969 & 2003 Publishing Group All rights reserved 0969-7128/03 $25.00 www.nature.com/gt RESEARCH ARTICLE Adenovirus-mediated transfer of the PTEN gene inhibits human colorectal growth in vitro and in vivo

Y Saito1, X Swanson2, AM Mhashilkar2, Y Oida1, R Schrock2, CD Branch1, S Chada2, L Zumstein2 and R Ramesh1 1Department of Thoracic and Cardiovascular Surgery, The University of Texas MD Anderson Cancer Center, Houston, TX, USA; and 2Introgen Therapeutics, Houston, TX, USA

The tumor-suppressor gene PTEN encodes a multifunctional wtPTEN than in normal colon fibroblast cells (CCD-18Co) that is mutated in a variety of human . carrying wtPTEN. This suppression was induced through PTEN inhibits the phosphatidylinositol 3- pathway and downregulation of the Akt/PKB pathway, dephosphorylation downstream functions, including activation of Akt/protein of focal adhesion kinase (FAK) and mitogen-activated protein kinase B (PKB), cell survival, and cell proliferation in tumor kinase (MAPK) and cell-cycle arrest at the G2/M phase, but cells carrying mutant- or deletion-type PTEN. In such tumor not the . Furthermore, treatment of human color- cells, enforced expression of PTEN decreases cell prolifera- ectal tumor xenografts (HT-29, and SW480) with Ad-PTEN tion through cell-cycle arrest at G1 phase accompanied, in resulted in significant (P¼0.01) suppression of tumor growth. some cases, by induction of . More recently, the These results indicate that Ad-PTEN exerts its tumor- tumor-suppressive effect of PTEN has been reported in suppressive effect on colorectal cancer cells through ovarian and thyroid tumors that are wild type for PTEN. In the inhibition of cell-cycle progression and induction of cell present study, we examined the tumor-suppressive effect of death. Thus Ad-PTEN may be a potential therapeutic for PTEN in human colorectal cancer cells that are wild type for treatment of colorectal cancers. PTEN. Adenoviral-mediated transfer of PTEN (Ad-PTEN) Gene Therapy (2003) 10, 1961–1969. doi:10.1038/ suppressed and induced apoptosis significantly in sj.gt.3302100 colorectal cancer cells (DLD-1, HT29, and SW480) carrying

Keywords: cancer; PTEN; apoptosis; ; colorectal

Introduction PI3K. This mechanism inhibits downstream functions mediated by the PI3K pathway, such as activation of PTEN is a tumor-suppressor gene located on human Akt/protein kinase B (PKB), cell survival, and cell 10q23.3.1,2 Deletion and somatic mutation proliferation.15 Thus, overexpression of PTEN in cancer of PTEN occur frequently in glioblastoma, endometrial cells carrying mutant- or deletion-type PTEN can inhibit cancer, prostate cancer, and small-cell lung cancer.3–7 cell proliferation and tumorigenicity via induction of Furthermore, germline mutations of PTEN are the cause cell-cycle arrest at G1 phase and apoptosis.16–19 Although of Cowden disease and Bannayan–Zonana syndrome, studies demonstrating the inhibitory function of PTEN which are autosomal dominant cancer predisposition in tumor cells mutant for PTEN exist, very little disorders.8–10 PTEN functions as both a protein and lipid information is available on the effect of PTEN in tumor phosphatase in cells that are genotypically wild type for cells that are wild-type for PTEN (wtPTEN). More PTEN.11,12 The former activity consists of dephospho- recently, studies using colorectal, ovarian, and thyroid rylation of focal adhesion kinase (FAK) and Shc, which cancer cells that are wtPTEN, demonstrated ectopic are upstream of the p44/42 mitogen-activated protein expression of PTEN resulting in growth inhibition and kinase (MAPK) pathway associated with the inhibition of cell death.20–22 Based on these reports, we investigated cell spreading and migration,13,14 while the latter activity the tumor-suppressive effect of PTEN in colorectal cancer consists of dephosphorylation of phosphatidylinositol- cells that are wtPTEN. triphosphate (PtdIns (3,4,5)P3) and phosphatidylinositol- In the present study, we demonstrate that adenovirus- biphosphate (PtdIns (3,4)P2).11 PTEN also modulates mediated overexpression of PTEN (Ad-PTEN) in color- the phosphatidylinositol 3-kinase (PI3K) pathway by ectal cancer cells and in normal colon fibroblast cells that catalyzing degradation of PtdIns (3,4,5)P3 generated by are wtPTEN results in selective induction of G2 cell-cycle arrest and apoptosis in tumor cells, but not in normal cells. Furthermore, treatment of colorectal tumor xeno- Correspondence: Dr R Ramesh, Department of Thoracic and Cardiovas- cular Surgery, Box 445, The University of Texas MD Anderson Cancer grafts with Ad-PTEN resulted in tumor inhibition. These Center, 1515 Holcombe Blvd, Houston, TX 77030, USA results indicate that Ad-PTEN can be used as a gene Received 15 August 2002; accepted 24 April 2003 therapeutic for the treatment of colorectal cancers. Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1962 Results Inhibition of cell proliferation in colorectal cancer due to overexpression of PTEN Tumor cells (DLD-1, SW480, HT29, and Colo 320 HSR) and normal cells (CCD-18Co) were treated with PBS, Ad- Luc, or Ad-PTEN and examined for exogenous PTEN expression and cell proliferation at various time points. Treatment with Ad-PTEN resulted in exogenous PTEN expression in all the cell lines tested (Figure 1a). However, endogenous PTEN expression was also observed in all the cell lines, indicating that they are wild type for PTEN, except Colo 320 HSR that is mutant for PTEN. Although Colo 320 HSR cell line is mutant for PTEN, detection of PTEN protein is not surprising since it has been previously reported to produce a truncated PTEN protein that lacks the tumor-suppressive activity.23 Furthermore, analysis for cell viability daily on days 1–5 demonstrated significant inhibition of cell proliferation (Po0.01) in all the cancer cell lines treated with Ad-PTEN, when compared with that in control cells treated with Ad-Luc or treated with PBS (Figure 1b). In normal cells (CCD-18Co), Ad-PTEN treatment resulted in moderate inhibition of cell growth compared to cells treated with Ad-Luc and PBS (Figure 1b). However, Ad- PTEN-mediated inhibitory effect on tumor cell prolifera- tion was greater and more pronounced than that observed in normal cells. Minimal to no inhibitory effect on cell proliferation was also observed in normal human colon epithelial cell line (FHC) when treated with Ad- PTEN (data not shown). These results suggest that Ad- PTEN may selectively and more effectively inhibit tumor cells than normal cells.

Induction of apoptosis in colorectal cancer cells due to overexpression of PTEN After treatment with Ad-PTEN, tumor cells (DLD-1, SW480, HT29, and Colo 320 HSR) and normal cells (CCD-18Co) were analyzed for apoptotic changes using fluorescence-activated cell sorter (FACS) and/or Hoechst 33258 staining. At 72 h after Ad-PTEN treatment, an increase in the number of cells in the sub-G0/G1 phase, an indicator of apoptotic changes, was observed in the four cancer cell lines by FACS analysis (Figure 2a). The percent increase in apoptotic cells was between 5 and 35%, depending on the tumor cell type. However, no changes were observed in tumor cells treated with Ad- Luc or PBS. In contrast, no significant induction of apoptosis was observed in normal cells treated with Ad- PTEN compared to cells treated with PBS and Ad-Luc Figure 1 Inhibition of cell proliferation in colorectal cancer cells due to (Figure 2a). To confirm these results further, Hoechst overexpression of PTEN. Tumor cells (DLD-1, HT29, SW480, and Colo 33258 staining was performed 72 h after treatment. 320 HSR) and normal colon fibroblast cells (CCD-18Co) were treated with Tumor cells underwent apoptosis following Ad-PTEN PBS, Ad-Luc, and Ad-PTEN and examined for PTEN protein expression treatment, but normal cells did not. No changes were and cell proliferation. (a) At 72 h after treatment, cells were analyzed for PTEN protein expression by Western blot analysis. Exogenous PTEN observed in any of the cells treated with Ad-Luc (Figure expression as indicated by the intensity was observed in all the cells treated 2b). with Ad-PTEN. However, endogenous PTEN expression was also observed in all the cell types except Colo 320 HSR that were treated with PBS and Ad-Luc indicating that the cells were wild type for PTEN. The Induction of G2/M cell-cycle arrest due to corresponding b-actin levels are shown as a loading control. (b) overexpression of PTEN Measurement of cell proliferation as determined by viability assay on To determine whether Ad-PTEN is capable of inducing days 1, 2, 3, 4, and 5 indicated that tumor cells were significantly m G1 cell-cycle arrest as previously reported,21,22,24,25 (P¼0.01) suppressed by Ad-PTEN ( ) when compared with control cells treated with PBS (E), and Ad-Luc (T). However, CCD-18Co cells were cell-cycle phases were analyzed using a fluorescence- not growth inhibited by both Ad-Luc and Ad-PTEN when compared with activated cell sorter (FACS). Cell-cycle analysis demon- PBS-treated control. Data represented are the average of triplicates. Error strated an increase in the percentage (20–42%) of tumor bars denote standard error (SE).

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1963 Ad-PTEN did not induce G1 arrest in cancer cells carrying wtPTEN. To confirm that the Ad-PTEN-induced G2 arrest observed in wtPTEN cells was true and not due to experimental conditions used in the present study, we examined the effects of Ad-PTEN in cancer cells (Colo

Figure 2 Induction of apoptosis due to overexpression of PTEN. (a) Tumor cells (DLD-1, HT29, SW480, and Colo 320 HSR) and normal cells (CCD- 18Co) were treated with PBS, Ad-Luc, or Ad-PTEN and harvested at 72 h after treatment. The number of cells at sub-G0/G1 (apoptotic cells) were analyzed by flow cytometry. Cells treated with Ad-PTEN showed an increase in the number of cells in the subG0/G1 phase compared to control cells. Data represent the mean of duplicate experiments. Error bars denote standard error (SE). (b) Analysis of tumor cells (DLD-1, HT29 SW480, and Colo 320 HSR) and normal cells (CCD-18Co) by Hoechst 33258 staining 72 h after treatment with Ad-Luc and Ad-PTEN. Ad-PTEN induced apoptosis in tumor cells but not in normal cells. Arrows indicate apoptotic cells. Magnification was  40 for all cell lines. Figure 3 Induction of G2/M cell-cycle arrest due to overexpression of PTEN. Tumor cells that are wtPTEN (DLD-1, HT29, and SW480) or cells (DLD-1, HT29, and SW480) in the G2/M population mtPTEN (Colo 320 HSR) were treated with PBS, Ad-Luc, Ad-PTEN, or 72 h after treatment with Ad-PTEN (Figure 3) when exposed to 20-mM LY294002 reconstituted in DMSO or 0.2% DMSO compared to control cells treated with Ad-Luc, treated alone. Cells were harvested 72 h after treatment and cell-cycle analysis was performed using flow cytometry. In all, 20 000 events were captured for with PBS, or exposed to 20- mM LY294002 or 0.2% each treatment and the data are shown as histograms. Cells that are dimethyl sulfoxide (DMSO). However, tumor cells wtPTEN underwent G2/M arrest, while cells that are mtPTEN underwent treated with LY294002 were arrested at the G1 phase G1 arrest. The cell-cycle phase is represented on the X-axis. Data were since LY294002 is a PI3K inhibitor. Although one of the generated in duplicate; the average values are shown. Bars denote standard PTEN functions is PI3K inhibition, like LY294002,17,22,26 error (SE).

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1964 320 HSR) that are mutant for PTEN (mtPTEN). Over- sphorylated significantly in DLD-1 cells by Ad-PTEN expression of PTEN demonstrated an increase in the as compared with LY294002 and other controls (Figure percentage (14%) of Colo 320 HSR cancer cells in the G1 5a). In contrast, no significant change in pAkt, pFKHR, population, suggesting that Ad-PTEN selectively induces pGSK3a, and pGSK-3b was observed in Ad-PTEN- G2/M arrest in tumor cells that are wtPTEN (Figure 3). treated CCD-18Co cells compared to control cells (Figure Induction of G2 arrest by Ad-PTEN has also been 5b). These results demonstrate that PTEN downregulates observed in prostate cancer cells (DU145) that are the Akt-kinase pathway and suppressed cell growth and wtPTEN and G1 arrest in prostate cancer cells that are proliferation in cancer cells, but not in normal cells mutant (LNCaP) or null (PC-3) for PTEN (Saito et al, carrying wtPTEN unpublished data). We next examined the phosphorylation status of FAK (pFAK) and p44/42 MAPK associated with cell migra- PTEN induced activation of the caspase cascade in tion and/or focal adhesions in DLD-1 and CCD-18Co 13,14,30,31 colorectal cancer cells cells 72 h after infection. In DLD-1 cells, pFAK To determine the underlying mechanism by which Ad- and p44/42 MAPK were dephosphorylated by Ad- PTEN induced apoptosis in tumor cells, the activation PTEN, but not by Ad-Luc or PBS (Figure 5a). U0126, of caspases was examined in DLD-1 cells by Western which is an MEK 1/2 inhibitor, also inhibited p44/42 blot analysis. This analysis demonstrated activation of MAPK and served as a positive control. No changes in the caspase cascade, including cleavage of caspase-9, pFAK and p44/42 MAPK expression were observed caspase-3, and poly (ADP-ribose) polymerase (PARP) in Ad-PTEN-treated CC18-Co cells compared to cells 72 h after treatment with Ad-PTEN (Figure 4). These treated with Ad-Luc and PBS (Figure 5b). These results results show that PTEN induced apoptosis via a caspase demonstrate that PTEN can selectively inhibit cell cascade in cancer cells with wtPTEN. migration and adhesion through downregulation of the FAK and MAPK pathways in cancer cells carrying wtPTEN, but not in normal cells. Signaling pathways modulated by PTEN overexpression in colorectal cancer cells To determine the signaling pathways that are modulated Association of G2 arrest in cancer cells with by PTEN, we examined the expression of Akt, forkhead downregulation of Cdc25C homolog of rhabdomyosarcoma (FKHR), and glycogen To investigate the mechanisms through which Ad-PTEN synthase kinase-3a/b (GSK-3a/b), which are down- induces G2/M arrest in cells carrying wtPTEN,we stream of the Akt kinase pathway and transcriptional evaluated proteins related to the G2/M cell-cycle regulators leading to apoptotic cell death,27–29 in tumor checkpoint using Western blot analysis. In this analysis, cells (DLD-1) and normal cells (CCD-18Co) by Western DLD-1 and CCD-18C0 cells were exposed to LY294002 or blot analysis (Figure 5). In DLD-1 cells, phospho-Akt treated with Ad-PTEN or Ad-Luc. Total cell lysates were was strongly dephosphorylated by Ad-PTEN and prepared 72 h after exposure or treatment and analyzed. LY294002 (Figure 5a). Surprisingly, increased phosphor- DLD-1 and CCD-18Co cells treated with Ad-PTEN ylation of Akt was observed in cells treated with Ad-Luc. demonstrated reduced expression of both phosphory- Also, phospho-FKHR (pFKHR) and phospho-GSK3a lated and nonphosphorylated Cdc25C and increased (pGSK3a), but not GSK-3b (pGSK-3b), were depho- expression of B1 compared with that in cells that were treated with PBS or treated with Ad-Luc (Figure 6a, b). However, the inhibition of phospho-Cdc25C and total Cdc25C was more significant in DLD-1 cells than in CCD-18Co cells. These results indicate that Cdc25C may have been strongly phosphorylated by PTEN in tumor cells and then excluded from the nucleus through binding with 14-3-3 protein,32,33 resulting in a reduction of the amount of Cdc25C. Treatment with LY294002 showed no changes in the level of Cdc25C and expression. These results are consistent with those of FACS analysis, which showed a significant induction of G2/M arrest caused by PTEN in tumor cells than in normal cells, while the PI3K inhibitor LY294002 induced G1 arrest.

Inhibition of colorectal tumor xenografts by PTEN The ability of Ad-PTEN to inhibit subcutaneous tumor xenografts was evaluated next. Subcutaneous colorectal tumor xenografts established in nude mice were divided into groups (n¼9/group) and treated as follows: Group 1 Figure 4 PTEN overexpression induces activation of the caspase pathway. received PBS, Group 2 received Ad-Luc, and Group 3 DLD-1 tumor cells were treated with PBS, Ad-Luc, or Ad-PTEN. At 72 h received Ad-PTEN. Intratumoral injections of Ad-PTEN after treatment, cells were harvested and examined by Western blot analysis for activation of the caspase cascade. Activation of caspase-9, on HT29 and SW480 tumor xenografts resulted in a caspase-3, and cleavage of PARP, a caspase substrate, was observed in Ad- significant (P¼0.01) growth inhibition compared to PTEN-treated DLD-1 cells, but not in cells treated with PBS and Ad-Luc. control tumors that were treated with PBS or treated The corresponding b-actin levels are shown as a loading control. with Ad-Luc (Figure 7a, b). However, no complete tumor

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1965

Figure 5 Signaling pathways regulated by PTEN overexpression. DLD-1 cells (a) and CCD-18Co (b) cells were treated with PBS, Ad-Luc, Ad-PTEN, 20- mM LY294002, or 10-mM U0126. At 72 h after treatment, cells were harvested and examined by Western blot analysis for signaling pathways regulated by PTEN. (A) Downregulation of phospho-Akt, phospho-FKHR, phospho-GSK3a, phospho-FAK, and inhibition of phospho-p44/42MAPK were observed in DLD-1 cells after Ad-PTEN treatment compared to Ad-Luc- and PBS-treated cells. (b) No change in the expression levels of phospho-Akt, phospho-FKHR, phospho-GSK3a, phospho-FAK, and phospho-p44/42MAPK was observed in CCD-18Co cells after Ad-PTEN treatment compared to Ad-Luc- and PBS- treated cells. The corresponding b-actin levels are shown as a loading control. regression was observed during the course of the difference observed in our results from others is that experiment. No treatment-related toxicity was observed we observed that colorectal tumor cells that are wtPTEN during the course of the experiment. were arrested in G2/M phase after Ad-PTEN treatment. In contrast, previous studies have reported that ectopic expression of PTEN in cancer cells resulted in G1 arrest Discussion that was associated with an increase in p27 expres- sion.17,18,21,35–37 Analysis for p27 expression in colorectal In the present study, we investigated the effects of cancer cells after Ad-PTEN treatment demonstrated no overexpressing PTEN protein in colorectal cancer cells change in expression levels (data not shown). In support that are wtPTEN. Prior to the start of the study, we of our findings is the recent report by Stewart et al,38 who examined the mutational status for PTEN gene in the demonstrated enforced expression of PTEN-inhibited cancer cell lines used in the study. Genetic alterations growth of melanoma cells that are wtPTEN. The were determined by sequencing the coding region using discrepancy in the results is not clear. However, one exon-specific primers. All of the cancer cell lines, except possibility is that the induction of cell-cycle arrest by Colo 320 HSR and CCD18-Co, were sequenced and PTEN may be cell-type-dependent. Alternatively, differ- found to contain no mutations in this gene (data not ences in the mutational changes in different cancers may shown). The Colo 320 HSR cell line has recently been contribute in determining the cell-cycle phases. In fact, as shown to have a 1- bp insertion in exon 7 of the PTEN demonstrated in the present study, overexpression of gene resulting in a nonsense mutation.23 As a conse- PTEN in colorectal cancers that are wtPTEN underwent quence of the nonsense mutation, a truncated PTEN G2 arrest while tumor cells that are mutant for PTEN protein is produced that lacks the COOH-terminal region underwent G1 arrest. Induction of G2 arrest by PTEN has critical for the tumor-suppressive activity. The impor- also been observed in a broad spectrum of tumor cells of tance of the COOH-terminal region for the tumor- the prostate and melanoma that are wtPTEN, but not in suppressive activity of the PTEN protein is well tumor cells that are mutant or null for PTEN (Saito et al, established.34 Correlating with these results was the unpublished data). These results indicate that the detection of exogenous PTEN protein expression in all endogenous status of PTEN may play a role in the cell- the tumor and normal cells used in the present study. cycle arrest and warrants further investigation. Whatever However, as described above, the PTEN protein detected the underlying mechanism, it is evident from the present in Colo 320 HSR cells lacks the tumor-suppressive study that the ability of PTEN to inhibit tumor growth is activity. Furthermore, we showed that overexpression independent of the mutational status of PTEN in cancer of PTEN suppresses the growth and induces apoptosis in cells and is in agreement with the findings previously colorectal cancer cell lines significantly compared to reported by Li et al.20 normal cells. These results are similar to the findings To further understand the mechanism by which PTEN previously reported in colorectal, ovarian, and thyroid induced G2/M arrest, G2/M-associated proteins were cancer that are wtPTEN.20,21,22 However, one major examined. In the G2/M phase, the active Cdc25C

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1966

Figure 6 Regulation of -associated proteins by PTEN. Tumor cells (DLD-1) and normal cells (CCD-18Co) were treated with PBS, Ad- Luc, or Ad-PTEN or exposed to 20-mM LY294002. Cells were harvested 72 h after treatment and analyzed for G2 phase-associated proteins by Figure 7 Therapeutic effect of Ad-PTEN on subcutaneous human Western blot analysis. Downregulation of phospho-Cdc25C and upregula- colorectal tumor xenografts. Subcutaneous HT29 (a) and SW480 (b) tion of cyclin B1 expression were observed in Ad-PTEN-treated tumor cells tumor-bearing mice were divided into three groups (n¼9/group) and (a) and normal cells (b), compared to control cells treated with PBS, Ad- treated on alternate days for a total of six doses (3 Â 1010 vp/dose) as Luc, and LY294002. However, the level of phospho-CdC25C down- follows: no treatment (m), Ad-Luc (E), and Ad-PTEN (T). Tumors were regulated by PTEN was more significant in tumor cells than in normal measured using calipers and the statistical significance of tumor volume cells. No significant change in Cdc2 expression was observed among the changes was calculated using Student’s t-test. Tumors treated with Ad- treatment groups in both tumor and normal cells. The corresponding b- PTEN showed significant tumor inhibition (P¼0.01) compared to tumors actin levels are shown as a loading control. that were not treated or treated with Ad-Luc. Each time point represents the mean tumor volume for each group. Bars represent standard error (SE).

phosphatase dephosphorylates Cdc2 on both Thr 14 The mechanism by which PTEN suppressed cell and Tyr 15, and then activates Cdc2/cyclin B1 com- growth and induced apoptosis appears to be dependent plexes, leading to the progression of the cells to upon its phosphatase activity. Requirement of the PTEN the M phase. However, in response to DNA damage phosphatase activity to downregulate the Akt, FAK, and or inhibition of DNA replication, the inhibitory phos- p44/42MAPK pathways and activate the caspase-9 phorylation of cdc2 still remains and the cells are cascade has been previously demonstrated in tumor arrested at the G2 phase.31,39–41 Our data demonstrated cells that are mutant or null for PTEN.13,14,27–31 In that Cdc25C was significantly downregulated, while agreement with these reports are the results from the cyclin B1 expression was upregulated following present study. Treatment of colorectal cancer cells with Ad-PTEN treatment in tumor cells compared to Ad-PTEN resulted in the inhibition of Akt, FAK, and normal cells. Induction of G2/M arrest through inhibi- p44/42 MAPK pathways and activation of caspase-9 and tion of Cdc25C correlated with our cell-cycle analysis, -3. However, Ad-PTEN treatment in normal cells did not where the number of tumor cells that underwent result in the regulation of these pathways, and may G2/M arrest was significantly higher than that in normal explain their inability to undergo apoptotic cell death, cells. Induction of G2/M arrest after Ad-PTEN treatment thereby achieving tumor-selective killing. The ability of through inhibition of Cdc25C has also been observed in tumor cells, but not normal cells that are wtPTEN,to other tumor cell lines that are wtPTEN (unpublished undergo apoptotic cell death is in agreement with the data). findings of Li et al,20 who demonstrated that colorectal

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1967 cancer cells that are wtPTEN were sensitive to PTEN, but determined via apoptotic body and/or chromosome not normal cells. It is also possible that tumor cells have condensation. other signaling pathways that are defective, resulting in triggering of apoptotic cell death upon PTEN expression. Cell-cycle analysis An alternate explanation can be that the endogenous Cells were seeded in 10-cm culture dishes (5–10 Â 105 PTEN expressed in the tumor cell lines, but not in the cells/dish) and treated with Ad-PTEN or Ad-Luc, normal cells, is not functional. To gain an insight into this or treated with PBS, 20-mM LY294002 (Cell Signaling possibility further investigation is warranted. Technology, Beverly, MA, USA) or 0.2% DMSO alone; Finally, the ability of Ad-PTEN to suppress the growth LY294002 was reconstituted in DMSO. At specific times of xenograft tumors (HT-29 and SW480) was investi- after treatment, cells were harvested via trypsinization, gated. Ad-PTEN treatment inhibited tumor growth washed once with ice-cold PBS, fixed with 70% ethanol, significantly and supports our in vitro findings presented and stored at À201C. Cells were then washed twice with here. The ability of Ad-PTEN to inhibit human colorectal ice-cold PBS and treated with RNase (30 min at 371C, tumor growth that is wtPTEN has not been previously 500 U/ml; Sigma Chemicals, St. Louis, MO, USA), and documented. In conclusion, the present study demon- DNA was stained with propidium iodide (50 mg/ml; strates the ability of Ad-PTEN to inhibit colorectal cancer Boehringer Mannheim, Indianapolis, IN, USA). The cell- growth, both in vitro and in vivo. Thus, adenovirus- cycle phase and apoptotic rate (cells at sub-G0/G1 mediated transfer of the PTEN gene may be a promising phase) were analyzed using a FACScan (EPICS therapeutic strategy for the treatment of patients with XL-MCL; Beckman Coulter Inc; Fullerton, CA, USA). colorectal cancer. Immunoblot analysis Cells were treated with Ad-PTEN or Ad-Luc, or treated Materials and methods with PBS, 20-mM LY294002, 10-mM U0126 (Cell Signaling Technology Inc., Beverly, MA, USA), or 0.2% DMSO Cell lines and cultures alone. Cells were incubated for indicated number of The colorectal cancer cell lines DLD-1, HT29, SW480, and hours at 371C and then collected to prepare whole-cell Colo 320 HSR and the normal colon fibroblast cell line lysates as previously described.38,42 The following anti- CCD-18Co were obtained from the American Type bodies were used as primary antibodies: PTEN, Cdc25C, Culture Collection (ATCC; Rockville, MD, USA). The Cdc2, and Cyclin B1 (Santa Cruz Biotechnology, Santa cancer cell lines and normal cell line were maintained as Cruz, CA, USA); caspase-9, caspase-3, PARP, and recommended by ATCC. Cells were regularly passaged phospho-FAK (Pharmingen, San Diego, CA, USA), and tested for mycoplasma prior to use in the experi- b-actin (Sigma Chemicals, St Louis, MO, USA); p27Kip1, ments. Akt, Phospho-Akt, phospho-FKHR, phospho-GSK3a/b, p44/42MAPK, and phospho-p44/42MAPK (Cell Signal- Viral production and transduction efficiency ing Technology Inc; Beverly, MA, USA). The proteins Construction and production of adenovirus carrying the were visualized on an enhanced chemiluminescence PTEN (Ad-PTEN) and luciferase (Ad-Luc) gene have (ECL) film (Hyperfilm, Amersham Biosciences, Inc., recently been described.38 Prior to start of the experi- Piscataway, NJ, USA) using ECL Western blot detection ment, transduction efficiency for the cell lines to be used reagent (Amersham Biosciences, Inc., Piscataway, NJ, was determined using an adenoviral vector carrying the USA). green fluorescent protein (Ad-GFP). Treatment of cells with Ad-GFP different viral particle number per cell In vivo tumor xenograft studies (vp/cell) demonstrated that more than 80% of the cells Subcutaneous tumor xenografts (HT29 and SW480) were were transduced at 5000 vp/cell (data not shown). established in nude mice by injecting tumor cells (5 Â 105 Therefore, in all subsequent experiments cells were cells/) in the right dorsal flank. When the tumor treated with 5000 vp/cell. had reached a size of 50–100 mm3, were randomized into three groups (n¼9 animals/group) Cell proliferation assay and treatment was initiated as follows: group 1 received no treatment; group 2 received Ad-Luc (3 Â 1010 vp/ All the cell lines were plated in six-well tissue culture 10 plates at a density of 1 Â 105 cells/well. Tumor cells were dose); and group 3 received Ad-PTEN (3 Â 10 vp/ then treated with Ad-PTEN or Ad-Luc, or treated with dose); all treatments were given on alternate days for a PBS as a mock control. Cells in each treatment group total of six doses. Intratumoral injections were performed were plated in triplicate and cultured for 5 days. Then, at under anesthesia. Tumor measurements were recorded without knowledge of the treatment groups as pre- designated time points, cells were harvested by trypsi- 43 nization and stained with 0.4% trypan blue (GIBCO BRL, viously described. Experiments were terminated when Grand Island, NY, USA) to reveal dead cells. Viable cells tumors showed signs of necrosis. The therapeutic effect were then counted using a hemocytometer. was determined by statistical analysis using Student’s t-test. Apoptotic staining Cells were seeded in six-well tissue culture dishes at a Acknowledgements density of 1 Â 105 cells/well and treated with Ad-PTEN or Ad-Luc. At 72 h after treatment, cells were analyzed We thank Dr Jack Roth for his critical comments, Don for apoptosis using Hoechst 33258 staining (Sigma Norwood for editorial assistance, and Peggy James for Chemicals, St Louis, MO, USA). Apoptotic cells were preparation of this manuscript. This work was supported

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1968 in part by the Texas Higher Education Coordinating 18 Furnari FB, Huang HJ, Cavenee WK. The phosphoinositol

Board ATP/ARP grant (003657-0078-2001; RR), by a phosphatase activity of PTEN mediates a serum-sensitive G1 Career Development award from the University of Texas growth arrest in glioma cells. Cancer Res 1998; 58: 5002–5008. SPORE in Lung Cancer (P50 CA70907; RR), by the 19 Davies MA et al. Adenoviral transgene expression of MMAC/ University of Texas MD Anderson Cancer Center PTEN in human glioma cells inhibits Akt activation and induces Institutional Grant (RR), by the KECK Foundation Fund anoikis. Cancer Res 1998; 58: 5285–5290. for Human Cancer Gene Prevention and Therapy (RR), 20 Minaguchi T et al. Growth suppression of human ovarian cancer by BESCT Lung Cancer Program Grant (DAMD17-01-1- cells by adenovirus-mediated transfer of the PTEN gene. Cancer 0689), by TARGET Lung Cancer grant (DAMD17-01- Res 1999; 59: 6063–6067. 0706), by the University of Texas MD Anderson Cancer 21 Weng LP et al. Transient ectopic expression of PTEN in thyroid Center Support Grant (CA 16672), and by a sponsored cancer cell lines induces cell cycle arrest and cell type-dependent cell death. Hum Mol Genet 2001; 10: 251–258. research agreement with Introgen Therapeutics, Inc. 22 Li J et al. The PTEN/MMAC1 tumor suppressor induces cell death that is rescued by the AKT/protein kinase B . Cancer Res 1998; 58: 5667–5672. References 23 Dicuonzo G et al. Colorectal carcinomas and PTEN/MMAC1 gene mutations. Clin Cancer Res 2001; 7: 4049–4053. 1 Steck PA et al. Identification of a candidate tumour suppressor 24 Gottschalk AR et al. p27Kip1 is required for PTEN-induced G1 gene, MMAC1, at chromosome 10q23.3 that is mutated in growth arrest. Cancer Res 2001; 61: 2105–2111. multiple advanced cancers. Nat Genet 1997; 15: 356–362. 25 Weng LP, Brown JL, Eng C. PTEN coordinates G(1) arrest by 2LiJet al. PTEN, a putative protein tyrosine phosphatase gene down-regulating via its protein phosphatase activity mutated in human brain, breast, and prostate cancer. Science and up-regulating p27 via its lipid phosphatase activity in a 1997; 275: 1943–1947. breast cancer model. Hum Mol Genet 2001; 10: 599–604. 0 3 Rasheed BK et al. PTEN gene mutations are seen in high- 26 Taylor V et al. 5 phospholipid phosphatase SHIP-2 causes grade but not in low-grade gliomas. Cancer Res 1997; 57: protein kinase B inactivation and cell cycle arrest in glioblastoma 4187–4190. cells. Mol Cell Biol 2000; 20: 6860–6871. 4 Wang SI et al. Somatic mutations of PTEN in glioblastoma 27 Nakamura N et al. Forkhead transcription factors are critical multiforme. Cancer Res 1997; 57: 4183–4186. effectors of cell death and cell cycle arrest downstream of PTEN. 5 Tashiro H et al. Mutations in PTEN are frequent in endometrial Mol Cell Biol 2000; 20: 8969–8982. carcinoma but rare in other common gynecological 28 Dijkers PF et al. Forkhead transcription factor FKHR-L1 malignancies. Cancer Res 1977; 57: 3935–3940. modulates cytokine-dependent transcriptional regulation of 6 Cairns P et al. Frequent inactivation of PTEN/MMAC1 in p27(KIP1). Mol Cell Biol 2000; 20: 9138–9148. primary prostate cancer. Cancer Res 1997; 57: 4997–5000. 29 Lu Y et al. The PTEN/MMAC1/TEP tumor suppressor gene 7 Yokomizo A et al. PTEN/MMAC1 mutations identified in small decreases cell growth and induces apoptosis and anoikis in cell, but not in non-small cell lung cancers. Oncogene 1998; 17: breast cancer cells. Oncogene 1999; 18: 7034–7045. 475–479. 30 Tamura M, Gu J, Takino T, Yamada KM. Tumor suppressor 8 Nelen MR et al. Germline mutations in the PTEN/MMAC1 PTEN inhibition of cell invasion, migration, and growth: gene in patients with Cowden disease. Hum Mol Genet 1997; 6: differential involvement of focal adhesion kinase and p130Cas. 1383–1387. Cancer Res 1999; 59: 442–449. 9 Liaw D et al. Germline mutations of the PTEN gene in Cowden 31 Gu J et al. Shc and FAK differentially regulate cell motility disease, an inherited breast and thyroid cancer syndrome. Nat and directionality modulated by PTEN. J Cell Biol 1999; 146: Genet 1997; 16: 64–67. 389–403. 10 Marsh DJ et al. Germline mutations in PTEN are present in 32 Peng CY et al. Mitotic and G2 checkpoint control: regulation of Bannayan–Zonana syndrome. Nat Genet 1997; 16: 333–334. 14-3-3 protein binding by phosphorylation of Cdc25C on serine- 11 Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1, 216. Science 1997; 277: 1501–1505. dephosphorylates the lipid second messenger, phosphati- 33 Dalal SN, Schweitzer CM, Gan J, DeCaprio JA. Cytoplasmic dylinositol 3,4,5-triphosphate. J Biol Chem 1998; 273: localization of human cdc25C during requires an 13375–13378. intact 14-3-3 binding site. Mol Cell Biol 1999; 19: 4465–4479. 12 Myers MP et al. P-TEN, the tumor suppressor from human 34 Georgescu MM, Kirsch KH, Akagi T, Shishido T, Hanafusa H. chromosome 10q23, is a dual-specificity phosphatase. Proc Natl The tumor-suppressive activity of PTEN is regulated by its Acad Sci USA 1997; 94: 9052–9057. carboxyl-terminal region. Proc Natl Acad Sci USA 1999; 96: 13 Tamura M et al. Inhibition of cell migration, spreading, and 10182–10187. focal adhesions by tumor suppressor PTEN. Science 1998; 280: 35 Furnari FB, Lin H, Huang HS, Cavenee WK. Growth 1614–1617. suppression of glioma cells by PTEN requires a functional 14 Gu J, Tamura M, Yamada KM. Tumor suppressor PTEN inhibits phosphatase catalytic domain. Proc Natl Acad Sci USA 1997; 94: integrin- and growth factor-mediated mitogen-activated protein 12479–12484. (MAP) kinase signaling pathways. J Cell Biol 1998; 143: 36 Robertson GP et al. In vitro loss of heterozygosity targets the 1375–1383. PTEN/MMAC1 gene in melanoma. Proc Natl Acad Sci USA 1998; 15 Stambolic V et al. Negative regulation of PKB/Akt-dependent 95: 9418–9423. cell survival by the tumor suppressor PTEN. Cell 1998; 95: 37 Cheney IW et al. Adenovirus-mediated gene transfer of 29–39. MMAC1/PTEN to glioblastoma cells inhibits entry by 16 Cheney IW et al. Suppression of tumorigenicity of glioblastoma the recruitment of p27Kip1 into /CDK2 complexes. cells by adenovirus-mediated MMAC1/PTEN gene transfer. Cancer Res 1999; 59: 2318–2323. Cancer Res 1998; 58: 2331–2334. 38 Stewart A et al. PI3 kinase blockade by Ad-PTEN inhibits 17 Li D M, Sun H. PTEN/MMAC1/TEP1 suppresses the invasion and induces apoptosis in RGP and metastatic 8 tumorigenicity and induces G1 cell cycle arrest in human melanoma cells. Mol Med 2002; : 451–461. glioblastoma cells. Proc Natl Acad Sci USA 1998; 95: 39 Dunphy WG, Kumagai A. The protein contains an 15406–15411. intrinsic phosphatase activity. Cell 1991; 67: 189–196.

Gene Therapy Inhibition of human colorectal cancer growth by PTEN Y Saito et al 1969 40 Kastan MB et al. Participation of protein in the cellular 42 Saeki T et al. Tumor suppressive effects by adenovirus-mediated response to DNA damage. Cancer Res 1991; 51: 6304–6311. mda-7 gene transfer in non-small cell lung cancer cells in vitro. 41 Sebastian B, Kakizuka A, Hunter T. Cdc25M2 activation Gene Therapy 2000; 35: 2015–2057. of cyclin-dependent by dephosphorylation of 43 Saeki T et al. Inhibition of human lung cancer growth following threonine-14 and tyrosine-15. Proc Natl Acad Sci USA 1993; 90: adenovirus mediated mda-7 in vivo. Oncogene 3521–3524. 2002; 21: 4558–4566.

Gene Therapy