Annals of Surgical Oncology, 9(3):310-316 Published by Lippincott Williams & Wilkins �9 2002 The Society of Surgical Oncology, Inc.

Phosphatase and Tensin Analog Gene Overexpression Engenders Cellular Death in Human Malignant Mesothelioma Cells via

Inhibition of AKT Phosphorylation

Imran Mohiuddin, MD, Xiaobo Cao, MD, Mustafa K. Ozvaran, MD, Lou Zumstein, PhD, Sunil Chada, PhD, and W. Roy Smythe, MD

Background: Abnormal phosphatase and tensin analog (PTEN) gene expression has been noted in neoplasms. The PTEN protein cleaves phosphate groups from cellular growth kinases, inhibiting tumor propagation. A downstream target of PTEN is AKT, a serine-threonine kinase that when activated inhibits apoptosis. We sought to determine whether PTEN overexpression in mesotheli- oma cells would engender hypophosphorylation of AKT and apoptosis.

Methods: Human malignant mesothelioma cell lines REN and 1-45 were transfected with adenoviral vectors AdPTEN and AdBgal (marker gene) at various multiplicities of infection (MOI). Cell viability was measured using a colofimetric assay, and apoptosis was assessed by morphology and subG1 fluorescence-activated cell sorter (FACS) analysis. PTEN protein and AKT phophory- lation were evaluated by Western blot, and AKT kinase activity was evaluated by functional assay.

Results: Increased cellular killing was noted with AdPTEN gene transfer. The ratio of cell killing with AdPTEN to AdBgal widened with increasing MOI and was statistically significant at all MOI. Cells demonstrated apoptosis by morphologic and subG1 FACS analyses. Ceils overexpressing PTEN demonstrated decreased phosphorylated (active) AKT and AKT kinase activity compared with controls.

Conclusions: Overexpression of PTEN engenders apoptosis in mesothelioma by AKT hypophos- phorylation. The forced overexpression of PTEN may prove useful clinically in this treatment- resistant neoplasm.

Key Words: Mesothelioma--PTEN--AKT--Apoptosis.

Malignant mesothel ioma (MPM) is a relatively un- common neoplasm that arises from a variety of mesothe- lial lined and covered tissues, with the pleura most often affected. 1 A definitive epidemiological link to asbestos has been established, and recent studies have implicated exposure to the SV40 tumor virus as well. 1 Although uncommon, this tumor continues to be a frustrating clin- ical problem. Significant controversy currently exists regarding the conventional treatment of MPM, but little

Received March 16, 2001; accepted September 20, 2001. From the Section of Thoracic Molecular Oncology, Department of

Thoracic and Cardiovascular Surgery, The University of Texas M. D. Anderson Cancer Center, Houston, Texas.

Address correspondence and reprint requests to: W. Roy Smythe, MD, Department of Thoracic and Cardiovascular Surgery, The Uni- versity of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Box 109, Houston, TX 77030; Fax: 713-794-4669; E-maih rsmythe @ mdanderson.org.

controversy exists regarding outcome. Although highly selected patients may derive benefit from aggressive therapy, the response to conventional treatment of all comers with MPM has been quite poor5 Response rates to chemotherapy are generally 3 <20%. This unrespon- siveness to conventional therapy has led to an interest in more novel paradigms, such as photodynamic therapy, immunotherapy, and gene therapy approaches. 4-7 The need to find effective treatment is given added urgency by the recent evidence that shows an increasing inci- dence of MPM in many Western countries. In Europe, the number of mesothel ioma deaths in men is expected to double over the next 20 years, with a peak around the

year 2020. 8 Gene therapy for malignant pleural mesothel ioma re-

mains in its infancy. The main approach to date has been a prodrug approach, or gene therapeutics, whereby a

310

PHOSPHATASE AND TENSIN ANALOG GENE AND MESOTHELIOMA 311

protein, usually an enzyme, is delivered into the cell via a viral vector. This enzyme is then able to act on a subsequently delivered relatively inactive prodrug and metabolizes it to an active agent. Smythe et al. 4,9 have described one such treatment with adenoviral transfer of the herpes simplex virus/thymidine kinase gene com- bined with exposure to ganciclovir.

The phosphatase and tensin analog (PTEN)/MMAC1/ TEP1 tumor suppressor gene has recently been described in the literature to encode a protein that has a key role in both cell growth and integrin function. This gene has been localized to chromosome 10@3. l~ PTEN dysfunc- tion was initially linked to a group of inherited germline disorders such as Cowden's syndrome. 11 Since then, deleted or mutated PTEN genes have been described in various tumors, including gliomas; endometfial cancers; breast, thyroid, bladder, ovary, and small-cell lung can- cer; and hematological malignancies. 12 The N-terminal domain of PTEN shows extensive homology to tensin and auxillin, key cytoskeletal proteins. In another exon of the gene resides the signature motif of the dual spec- ificity protein phosphatase, which dephosphorylates serine, threonine, and tyrosine residues. Mutations in PTEN are found to be almost exclusively in this phos- phatase domain. 13 One key protein that is dephosphory- lated by wild-type PTEN is the serine-threonine kinase AKT. Phosphorylated AKT, the active form of this ki- nase, promotes cell survival through multiple pathways, many of which are associated with cell growth and survival. Levels of phosphorylated AKT have been re- ported to be increased in multiple tumors. 14

This article describes work evaluating the effects of forced overexpression of the PTEN gene product in MPM cells. We hypothesized that this would engender apoptosis in these treatment-resistant cells via changes in the phosphorylation status of AKT. The results presented here indicate that such forced overexpression does in- deed engender apoptotic cellular death, inhibition of AKT kinase, and subsequent inhibition of AKT phos- phorylation. Although we noted these findings in both cell lines evaluated, they were attenuated somewhat in the p53 wild-type MPM cell line, 1-45.

MATERIALS AND METHODS

Cell Lines and Adenoviral Vectors To study the effect of PTEN transfection in mesothe-

lioma, two well-characterized cell lines were used: 1-45 and REN. 1-45 is a human sarcomatous-type mesotheli- oma, p53 wild type (donated generously by Dr. Joseph Testa of the Fox Chase Institute). REN is a human epithelial-type mesothelioma and p53 mutant (devel-

oped by author W.R.S.). Cell lines were maintained in RPMI media containing 10% (v/v) fetal bovine serum, 1% nonessential amino acids, 100 U/ml penicillin, and 100 /xg/ml of streptomycin. Cells were housed in a humidified incubator with 5% CO 2 at 37~ AdPTEN and Ad-/3-galactosidase (Bgal) are first-generation ade- novirus type 5 vector constructs with E1 and E3 dele- tions. These vectors were created by use of homologous recombination in U937 cells. AdPTEN was provided by Introgen Therapeutics, Inc. (Houston, TX). AdBgal was synthesized in our laboratory. A standard cytomegalovi- ms promoter was used in both constructs.

Adenoviral Transfection Cells were exposed to adenoviral vectors in 96-well

and 6-cm plates. Cells were initially collected, counted manually, and then incubated overnight. Cells were then washed with phosphate-buffered saline (PBS) and trans- fected with varying multiplicities of infection (MOIs) of adenovirus diluted in serum-free RPMI. After 3 hours of exposure to vector, RPMI media supplemented with 20% fetal bovine serum was added in equal volumes to bring the final concentration to 10% serum.

Cell Viability Cell viability studies were performed with the XTT

assay kit (Roche Diagnostics Corporation, Indianapolis, IN). Cells were plated on 96-well plates at a concentra- tion of 3000 cells per well. After overnight incubation, cells were transfected with either AdPTEN or AdBgal marker virus for 3 hours at a wide range of MOIs from 0 to 120. At each time point, 50/zl of XTT reagent was added to all wells and allowed to incubate with the cells for 4 hours. After this incubation period, all plates were read with a colorimetric plate reader system (Dynatech, Inc., Langley, VA). Analysis of data was performed with Microsoft Excel (Microsoft Corp., Redmond, WA). These experiments were performed in triplicate.

Evaluation of Apoptosis Apoptotic cell death was examined by cell morphol-

ogy and cell cycle (sub-G 1) flow cytometry. Fluores- cence-activated cell sorter analysis was performed as follows: cells were plated in 6-cm plates at a concentra- tion of 1 • 106 cells per plate, incubated for 24 hours, and transfected with AdPTEN at an MOI of 60 for 3 hours. Seventy-two hours after transfection, cells were trypsinized, collected by centrifugation, resuspended in PBS, and fixed in 70% ethanol at -20~ overnight. After fixation, the cells were washed in PBS, followed by the addition of propridium iodide and RNase (Boehr- inger Mannheim Co., Indianapolis, IN). Controls with

Ann Surg Oncol, Vol. 9, No. 3, 2002

312 I. M O H I U D D I N E T AL.

AdBgal were prepared in a similar fashion. Cells were then heated in a 37~ water bath for 10 minutes and then kept in the dark. The specimens were then analyzed with the use of an EPICS Profile II flow cytometer (Beckman Coulter, Inc., Miami, FL). An analysis region was set on the basis of the negative controls, and the percentage of labeled cells was calculated from this region. Attention was concentrated on the sub-G 1 population of cells. The area under the curve at this point was used to determine the amount of apoptosis.

PTEN Expression Western blot analysis was performed to confirm over-

expression of PTEN after adenoviral transfer. Cells were plated in 6-cm plates at a concentration of 1 x 106 cells per plate, incubated overnight, and then transfected with AdPTEN at an MOI of 60. Forty-eight hours after trans- fection, total cell lysates were prepared by lysing plated cell monolayers with a sodium dodecyl sulfate-polyac- rylamide gel electrophoresis sample buffer. The protein content of the lysates was then determined by bicincho- ninic acid (BCA) protein assay (Pierce, Rockford, IL). Then, each lane on an sodium dodecyl sulfate-polyac- rylamide gradient (8%-12%) gel was loaded with 5 /xg of cell lysate and electrophoresed to separate proteins by using a monoclonal antibody directed against PTEN (Chemicon, Temecula, CA). After electrophoresis at 100 V for 2 hours, the proteins were transferred to high-bond electrogenerated chemiluminescence membranes. The membranes were then incubated with the primary and secondary antibodies and developed according to the Amersham (Uppsala, Sweden) electrogenerated chemi- luminescence protocol. These experiments were repeated in triplicate.

AKT Phosphorylation Assay To confirm expression of AKT and discern the amount

of phosphorylated AKT, an AKT antibody assay was performed with an assay kit (New England Biolabs, Beverly, MA). Briefly, cells were plated at concentration of 1 X 106 cells per plate in a 6-cm plate, incubated for 24 hours, transfected with either AdPTEN or AdBgal at an MOI of 60, and collected at 48 hours; then, by use of antibodies directed against AKT and phosphorylated AKT provided in the kit, they were electrophoresed as previously described. A commercially available inhibitor to AKT kinase from Cell Signaling Technology (Bev- erly, MA) was used as a positive control at a dose of 50 p.M. By using Optimas software (Media Cybernetics, Inc., Silver Spring, MD), the intensity of bands from each cell line was compared. These experiments were repeated in triplicate.

AKT Kinase Activity AKT kinase activity was detected by using an enzyme

assay kit (New England Biolabs). Briefly, cells were plated in concentrations of 1 X 106 cells per plate in 6-cm plates, incubated for 24 hours, and transfected with either AdFFEN or AdBgal at an MOI of 60. Forty-eight hours after transfection, cells were collected and lysed. An antibody to AKT was used to selectively immuno- precipitate AKT from cell lysates. The resulting immu- noprecipitate was then incubated with a glycogen syn- thetase kinase (GSK)-3 fusion protein (substrate) in the presence of adenosine triphosphate and kinase buffer; this allows immunoprecipitated AKT to phosphorylate GSK-3. Phosphorylation of GSK-3 was measured by Western blotting by using a phospho-GSK-3a/b (Ser21/9) antibody. The intensity of the band is repre- sentative of the amount of AKT kinase activity. These experiments were repeated in triplicate.

RESULTS

AdPTEN Transfection Results in Overexpression of PTEN Protein

To confirm that exposure to our adenoviral construct was effectively leading to PTEN overexpression at the posttranslational level, we evaluated both cell lines for PTEN protein expression after transfection by use of a monoclonal PTEN antibody and Western analysis. A small amount of baseline PTEN expression was noted; however, as seen in Fig. 1, there was a dense PTEN band

PTEN

REN 1-45

ACTIN

FIG. 1. Cells from both cell lines were transfected with either Ad- phosphatase and tensin analog (PTEN) or Ad-/3-galactosidase (first- generation adenovirus type 5 vector constructs with E1 and E3 dele- tions). By using monoclonal antibody directed against PTEN, cell lysates were analyzed by Western blot for PTEN expression at 48 hours after transfection. Overexpression of PTEN was seen with both cell lines after AdPTEN transfection.

Ann Surg Oncol, Vol. 9, No. 3, 2002

PHOSPHATASE AND TENSIN ANALOG GENE AND MESOTHELIOMA 313

in both cell lines that corresponded with PTEN overex- pression when compared with marker virus transfection after exposure to AdPTEN.

Transfection of PTEN Engenders Apoptotic Cellular Death in Mesothelioma Cell Lines

To asses the ability of overexpression of the PTEN gene to engender cell death in our mesothelioma cell lines, we transfected both with variable MOIs of AdPTEN and compared it with marker virus transfection. We observed that as early as 48 hours after transfection, there was a significant amount of cellular death in both cell lines. REN seemed to be much more sensitive to PTEN-induced cell death than 1-45 (P = .024). At an MOI of 20, there was a 32.9% • 5.5% and 16.22% • 3.5% increase in cell killing when compared with marker virus for REN and 1-45, respectively, at 72 hours. Both cell lines showed a statistically significant increase in cell killing with PTEN when compared with AdBgal (P = .011 and P = .019, respectively). After studying a time course of cellular death over 5 days at an MOI of 60, both cell lines seemed to reach peak cellular death at 72 hours (Fig. 2A and 2B).

After confirming that PTEN did induce cellular death, we next investigated the amount of apoptosis. Morpho- logical changes (nuclear condensation and fragmenta- tion) associated with apoptosis were noted in both lines after transfection with AdPTEN, but not AdBgal. We quantified the amount of apoptosis by measuring the area under the curve denoting the sub-G 1 population for pro- pridium iodide fluorescence-activated cell sorter analy- sis. Our results confirmed that indeed there was a sig- nificant amount of PTEN-induced apoptosis when compared with marker virus. A higher degree of apopto- sis was noted in the REN cell line when compared with 1-45 (P = .003). At an MOI of 60, the apoptotic fraction by sub-G a quantification in REN was 25.4% • 1.04%, compared with 14.1% • 1.55% for 1-45. When com- pared with marker virus AdBgal, there was a 4.6-fold and 3.2-fold increase, respectively, for REN and 1-45. Levels of apoptosis in both cell lines were statistically significant with respect to marker virus (P = .002 and P = .038 for REN and 1-45, respectively; Fig. 3).

The Level of Unphosphorylated AKT Is Increased After AdPTEN Exposure, and This Is Related to a Decrease in AKT Kinase Activity

Previous studies have revealed that one of the down- stream effectors of PTEN is the phosphate kinase AKT. Wild-type PTEN normally dephosphorylates AKT. We decided to investigate whether the level of AKT phos-

45 01 .C 40

35

- - 30

r 25 t- 20

m

~t 15

~ lO

u 5 C

m 0

0 10 20 40 60 80

A . MOI

"~-- 4035 1-45 r

C 2 0 '

~ 10 5 f ' ~ / * p=0.024

J

24 48 72 96 120

B Hours After Transfection FIG. 2. (A) Cells from both cell lines were transfected with either Ad-phosphatase and tensin analog (PTEN) or Ad-/3-galactosidase (Bgal) (first-generation adenovirus type 5 vector constructs with E1 and E3 deletions). Cell viability was measured at 72 hours by XTT assay. Cell killing was calculated from controls and was normalized by subtracting baseline AdBgal cell killing from AdPTEN cell killing for each cell line. The increase in the amount of cell killing with AdPTEN from AdBgal was plotted against multiplicity of infection (MOI) of AdPTEN for each cell line. AdPTEN-induced cell killing was seen in both cell lines. However, cell killing in REN was greater than that seen in 1-45. (B) Cells from both cell lines were transfected with either AdPTEN or AdBgal. Cell viability was measured at 24, 48, 72, 96, and 120 hours by XTT assay. Cell killing was calculated from controls and was normalized by subtracting baseline AdBgal cell killing from AdPTEN cell killing for each cell line. The increase in the amount of cell killing with AdPTEN from AdBgal was plotted against hours after inspection. Cell killing increased until 72 hours and then did not increase afterward. For all time points, there was a greater amount of AdPTEN cell killing in REN than in 1-45 (P = .024).

phorylation was altered in our cell lines after exposure to AdPTEN and overexpression of PTEN protein. After transfection with AdPTEN, both cell lines displayed a decrease in the amount of phosphorylated AKT when compared with marker virus transfection. This decrease was less prominent in the 1-45 cell line (Fig. 4).

Ann Surg Oncol, Vol. 9, No. 3, 2002

314 I. M O H I U D D I N E T AL.

30

i i m

25 o ~,. 20 o

<1~ 15 ,4 , , , I

G) 10 O L .

G) 5

0

I I a d - B g a l

I-1 a d - P T E N

r I

R E N 1-45 FIG. 3. Cells from both cell lines were transfected with either Ad- phosphatase and tensin analog (PTEN) or Ad-/3-galactosidase (Bgal) (first-generation adenovirus type 5 vector constructs with E1 and E3 deletions). Apoptosis was measured at 72 hours by propridium iodide flow cytometry analysis. The percentage of cells in the sub-G 1 popu- lation was equated to apoptosis. Percentage apoptosis was plotted against each cell line for AdBgal and AdPTEN. There was a greater amount of apoptosis in the REN cell line (P = .003).

After demonstrating PTEN overexpression-driven hy- pophosphorylation of AKT, we assessed the amount of activity of the actual AKT phosphorylating enzyme by using a substrate-driven immunoprecipitation assay. As noted in Fig. 5, after AdPTEN transfection, there was a decrease in the amount of AKT kinase activity when compared with marker vector (Fig. 5). However, as in the previous assay, the decrease in activity of the AKT kinase after PTEN transfection was greater in the REN cell line than in the 1-45 line.

1-45 P " '

PHOSPHORYLATED AKT

AKT

RATIO .54 .87 .76

R E N P B ,

PHOSPHORYLATED AKT

AKT

RATIO .24 .96 .16

FIG. 4. Cells from both cell lines were transfected with either Ad- phosphatase and tensin analog (PTEN) or Ad-/3-galactosidase (Bgal) (first-generation adenovirus type 5 vector constructs with E1 and E3 deletions). A separate group of cells was treated with kinase inhibitor and used as a positive control. By using polyclonal antibody directed against phosphorylated and unphosphorylated AKT, Western blots were performed at 48 hours after transfection. The ratio of intensity between phosphorylated and unphosphorylated AKT bands was calcu- lated for each experimental group. There was a significant decrease in the amount of phosphorylated to unphosphorylated AKT with PTEN transfection or inhibitor, but not with Bgal transfection. REN was more sensitive to AdPTEN- or inhibitor-induced dephosphorylation than 1-45 (P, AdPTEN; B, AdBgal; I, kinase inhibitor).

death has not been carefully studied in mesothelioma, but by clinical behavior, it seems somewhat resistant. One group evaluated a number of conventional inducers of apoptosis in mesothelioma cells, including calcium iono- phore and hydrogen peroxide, but found the cells to be resistant.t6 Several unconventional treatments have been shown to lead to apoptosis in mesothelioma cells, such as talc, asbestos, amphotericin, and the anticholesterol drug

DISCUSSION

Although uncommon, MPM continues to be a frustrat- ing clinical problem with no definitively effective con- ventional therapy. In the past decade, however, advances have been made on several fronts that indicate that novel forms of therapy may eventually prove promising. Both experimental and preliminary clinical trial data suggest that gene therapy approaches for MPM have emerged as a class of potentially effective contemporary treatments, and this paradigm certainly deserves additional investigation.IS

The overexpression of PTEN in the MPM cell lines examined here led to the development of apoptotic cel- lular death. The forced induction of programmed cellular

1-45

Ad-PTEN Ad-Bgal

REN

FIG. 5. AKT kinase activity. Cells from both cell lines were trans- fected with either Ad-phosphatase and tensin analog (PTEN) or Ad- /3-galactosidase (Bgal) (first-generation adenovirus type 5 vector con- structs with E1 and E3 deletions). By using a nonradioactive substrate- driven immunoprecipitation assay, AKT ldnase activity was measured at 48 hours after transfection in both experimental groups. AKT kinase activity was decreased after AdPTEN transfection in both groups. The decrease in REN was much more significant than in 1-45.

Ann Surg Oncol, Vol. 9, No. 3, 2002

PHOSPHATASE AND TENSIN ANALOG GENE AND MESOTHELIOMA 315

lovastatin174~ however, no discrete mechanism has been revealed in any of these studies.

Since its discovery, several lines of evidence have confirmed the PTEN gene's role as a tumor suppressor both in vitro and in vivo in various types of malignancy. The PTEN gene (located at 10q23) was discovered and characterized in 1997 by Li et al. e~ This group had determined that a loss of heterozygosity (LOH) at 10q23 was common and that the wild-type 10q23 locus was capable of suppression of tumorigenicity of glioblastoma cells in mice, and they were therefore looking for a candidate tumor suppressor gene. Germline mutations of this gene have been identified and are related to a number of genetic diseases, such as Cowden's disease and the Bannayan-Zonana syndrome.11, 22 Mutations and LOH for this gene are relatively frequent in glioblastoma and small-cell carcinoma but seem to be relatively uncom- mon in non-small-cell lung carcinoma. 23-25 Despite the large number of data available regarding PTEN expres- sion in various tumors, its activity has not been evaluated in mesothelioma. The mutation status in our mesotheli- oma cell lines is unknown, but an LOH is unlikely, because at least the portion of the PTEN protein that is recognized by our antibody via Western blot detects baseline expression in both of our cell lines.

PTEN has been demonstrated to be involved in the regulation of cell proliferation and signal transduction through the AKT pathway. 26 By PTEN's ability to act as a phosphatase, it is able to downregulate phosphorylated or active AKT, thereby inhibiting the tumor cell's ability to communicate and proliferate. We have investigated the effect of forced overexpression of PTEN in two different human MPM cell lines. We demonstrated sig- nificantly increased induction of apoptotic cellular death after AdPTEN exposure and PTEN protein overexpres- sion in both of these cell lines when compared with exposure to the AdBgal marker gene vector. We noted that a decrease in phosphorylation of AKT occurred in both cell lines and that this seems to be related to an inhibition of AKT kinase. Both of these changes were less marked in the 1-45 (p53 wild-type) cell line, and these findings were also correlated with a less impressive rate of apoptotic cell death. One possible reason for the attenuated effect of PTEN on cell killing and apoptosis seen with the 1-45 cell line may be its wild-type p53 status, but reports in the literature regarding this expla- nation are conflicting. Cheng et al. 27 have demonstrated that inactivation of both alleles of p53 and PTEN tumor were found in >91% of nine human glioma cell lines. In another study, by Zhou et al., 28 it was demonstrated that there was no correlation between PTEN and p53 muta- tions in >88 brain tumor cell lines. Another unusual

finding in this study was the inability of the kinase inhibitor to downregulate phosphorylated AKT in the 1-45 line (Fig. 4). This finding raises the possibility of a mutated form of AKT or AKT kinase in this cell line that could also account for the relative resistance to the ef- fects of AdPTEN.

Although this experiment does not unequivocally prove the relationship between overexpression of PTEN, hypophosphorylation of AKT, and apoptosis, what is known about the PTEN/AKT pathway is supportive. The use of an adenoviral vector-mediated dominant negative form of AKT has been compared with PTEN overex- pression in pancreatic carcinoma cell lines, with identical results in regard to both downregulation of phosphory- lated AKT and the induction of apoptosis . 29 The PTEN regulatory pathway in mesothelioma deserves further study, because the forced overexpression of this protein, or other means of AKT kinase inhibition, may prove useful clinically in this apoptosis- and conventional ther- apy-resistant neoplasm.

Acknowledgments: Supported by an unrestricted Physi- cian-Scientist award by the University of Texas M. D. Ander- son Cancer Center (W.R.S.) and by the Win. Keck Center for Gene Therapy.

REFERENCES

1. Carbone M, Rizzo P, Pass H. Simian virus 40: the link with human malignant mesothelioma is well established. Anticancer Res 2000; 20:875-8.

2. Sugarbaker DJ, Flores RM, Jaklitsch MT, et al. Resection margins, extrapleural nodal status, and cell type determine postoperative long-term survival in trimodality therapy of malignant pleural mesothelioma: results in 183 patients. J Thorac Cardiovasc Surg 1999; 117:54-65.

3. Ong T, Vogelzang NJ. Chemotherapy in malignant mesothelioma: a review. J Clin Oncol 1996;14:1007-17.

4. Smythe WR, Hwang HC, Elshami AA, et al. Treatment of exper- imental human mesothelioma using adenovirus transfer of the herpes simplex thymidine kinase gene. Ann Surg 1995;222:78-86.

5. Moskal TL, Doughtery TJ, Urschel JD, et al. Operation and pho- todynamic therapy for pleural mesothelioma: 6-year followup. Ann Thorac Surg 1998;66:11128-33.

6. Christmas TI, Manning LS, Garlepp MJ, Musk AW, Robinson BW. Effect of interferon-alpha 2a on malignant mesothelioma. J Interferon Res 1993;13:9-12.

7. Eggermont AMM, Goey SH, Slingerland R, et al. Clinical and immunological evaluation of intrapleural interleukin-2 in malig- nant pleural mesothelioma: a phase II study (abstract). Proc Am Assoc Cancer Res 1991;32:206-9.

8. Peto J, Decarli A, La Vecchia C, et al. The European mesothelioma epidemic. Br J Cancer 1999;79:666-72.

9. Smythe WR, Hwang HC, Amin KM, et al. Use of recombinant adenovirus to transfer the herpes simplex virus thymidine kinase (HSVtk) gene to thoracic neoplasms: an effective in vitro drug sensitization system. Cancer Res 1994;54:2055-9.

10. Dahia PLM. PTEN, a unique tumor suppressor gene. Endocr Relat Cancer 2000;7:115-29.

11. Liaw D, Marsh DJ, Li J, et al. Germline mutations of the PTEN

Ann Surg Oncol, Vol. 9, No. 3, 2002

316 L MOHIUDDIN ET AL.

gene in Cowden disease, an inherited breast and thyroid cancer syndrome. Nat Genet 1997;16:64-7.

12. Besson A, Robbins S, Yong VW. PTEN/MMAC1/TEP1 in signal transduction and tumorigenesis. Eur J Biochem 1999;263:605-11.

13. Bonneau D, Longy M. Mutations of the human PTEN gene. Hum Mutat 2000; 16:109-11.

14. Tamura M, Gu J, Tran H, Yamada K. PTEN gene and integrin signaling in cancer. J Natl Cancer Inst 1999;91:1820-8.

15. Sterman DH, Treat J, Litzky LA, et al. Adenovirus-mediated herpes simplex virus thymidine kinase/ganciclovir gene therapy in patients with low malignancy: results of a phase I clinical trial in malignant mesothelioma. Hum Gene Ther 1998;9:1083-92.

16. Narasimhan SR, Yang L, Gerwin BI, Broaddus VC. Resistance of pleura1 mesothelioma ceil lines to apoptosis: relation to expression of bcl-2 and bax. Am J Physiol 1998;275:L165-71.

17. Broaddus VC, Yang L, Scavo LM, Ernst JD, Boylan AM. Asbestos induces apoptosis of human and rabbit pleural mesothelial cells via reactive oxygen species. J Clin Invest 1996;98:2050-9.

18. Marklund L, Henriksson R, Grankvist K. Amphotericin B-induced apoptosis and cytotoxicity is prevented by the Na § K § 2C1( )- cotransport blocker bumetanide. Life Sci 2000;66:PL319-24.

19. Nasreen N, Mohammed KA, Dowling PA, Ward MJ, Galffy G, Antony VB. Talc induces apoptosis in human malignant mesothe- lioma ceils in vitro. Am J Respir Crit Care Med 21300;161:595- 600.

20. Rubins JB, Greatens T, Kratzke RA, Tan AT, Polunovsky VA, Bitterman P. Lovastatin induces apoptosis in malignant mesothe- lioma cells. Am J Respir Crit Care Med 1998;175:1616-22.

21. Li J, Yen C, Liaw D, et al. PTEN, a putative protein tyrosine kinase gene mutated in brain, breast and prostate cancer. Science 1997; 275:1943-7.

22. Marsh DJ, Coulon V, Lunetta KL, et al. Mutation spectrum and genotype-phenotype analysis in Cowden disease and Bannayan- Zonana syndrome, two hamartoma syndromes with germline PTEN mutation. Hum Mol Genet 1998;7:507-15.

23. Forgacs E, Biesterveld EJ, Sekido u et al. Mutation analysis of the PTEN/MMAC gene in lung cancer. Oncogene 1998;17:1557-65.

24. Yokomizo A, Tindall D J, Drabkin H, et al. PTEN/MMAC muta- tions identified in small cell but not in non-small cell lung cancers. Oncogene 1998;17:475-9.

25. Teng DH, Hu R, Lin H, et al. MMAC/PTEN mutations in primary tumor specimens and tumor cell lines. Cancer Res 1997;57:5221-5.

26. Li D, Sun H. PTEN/MMAC/TEP1 suppresses the tumorigenicity and induces G1 arrest in human glioblastoma cells. Proc NatlAcad Sci U S A 1998;95:15406-11.

27. Cheng Y, Ng HK, Zhang SF, et al. Genetic alterations in pediatric high-grade astrocytomas. Hum Pathol 1999; 11:1284-90.

28. Zhou XP, Li YJ, Hoang-Xuan K. Mutational analysis of the PTEN gene in gliomas: molecular and pathological correlations. Int J Cancer 1999;84:150-4.

29. Calleja V, Stoll V, Vassaux G, Downward J, Lemoine N. Gene therapy for cancer using dominant negative mutants of the RAS/ PI-3?AKT signalling molecules (abstract No. 603). Presented at the Fourth Annual Meeting of The American Society For Gene Therapy, Seattle, WA, May 30-June 3, 2001.

Ann Surg Oncol, VoL 9, No. 3, 2002