© 2001 by Oxford University Press
Journal of the National Cancer Institute, Vol. 93, No. 13, 1014-1020,
July 4, 2001
© 2001 Oxford University Press
REPORT |
Effect of Retroviral Endostatin Gene Transfer on Subcutaneous and Intraperitoneal Growth of Murine Tumors
Affiliations of authors: A. L. Feldman, H. R. Alexander, D. Lorang, C. E. Thiruvathukal, E. M. Turner, S. K. Libutti (Surgical Metabolism Section, Surgery Branch, Center for Cancer Research), S. M. Hewitt (Laboratory of Pathology, Center for Cancer Research), National Cancer Institute, Bethesda, MD.
Correspondence to: Steven K. Libutti, M.D., National Institutes of Health, Bldg. 10, Rm. 3C428, Bethesda, MD 20892 (e-mail: Steven_Libutti{at}nih. gov).
 |
ABSTRACT |
|---|
 Top Abstract IntroductionNotesMaterials and MethodsResultsDiscussionReferences |
|---|
Background: Inhibiting tumor angiogenesis is a promising new strategy for treating cancer. Difficulties with the stability, manufacture, and long-term administration of recombinant antiangiogenic proteins have prompted investigators to use gene therapy to generate these proteins in vivo. We investigated whether transfer of the gene encoding the angiogenesis inhibitor endostatin into the murine liver cell line NMuLi could inhibit tumor growth in vivo. Methods: NMuLi cells were transduced with retroviral vectors containing the murine endostatin gene. The presence and function of endostatin in transduced cell supernatants were confirmed by competitive enzyme immunoassay and endothelial cell proliferation assays. Nude mice were given a subcutaneous or intraperitoneal injection with NMuLi cells, control transduced cells (NEF-null), or endostatin-transduced clones (NEF-Endo1 to 4) and were monitored for tumor growth. All statistical tests were two-sided. Results: Supernatants from the clone secreting the lowest amount of endostatin (NEF-Endo4, 28 ng/mL) inhibited endothelial cell proliferation by 6% (95% confidence interval [CI] = 0% to 12%), and those from the clone secreting the highest amount (NEF-Endo1, 223 ng/mL) inhibited endothelial cell proliferation by 20% (95% CI = 13% to 27%). Increased levels of endostatin were detected in tumor lysates, but not serum, of mice given a subcutaneous injection of NEF-Endo1 cells. After 63 days, mice given a subcutaneous injection of parental NMuLi or NEF-null cells had tumor volumes of 2400 mm3 (95% CI = 1478 mm3 to 3300 mm3) and 2700 mm3 (95% CI = 2241 mm3 to 3144 mm3), respectively, compared with mean tumor volumes of less than 30 mm3 in mice given an injection of NEF-Endo clones, a statistically significant difference (P<.001). After 123 days, all 16 mice given an intraperitoneal injection of parental NMuLi or NEF-null cells had died, compared with only three (9%) of 32 mice given an injection of NEF-Endo clones. Conclusions: Retroviral endostatin gene transfer leads to secretion of functional endostatin that is sufficiently active to inhibit tumor growth. Further studies of retroviral endostatin gene transfer for the treatment of cancer are warranted.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionNotesMaterials and MethodsResultsDiscussionReferences |
|---|
Because tumors require angiogenesis for sustained growth (1), inhibiting tumor angiogenesis is a promising new strategy for treating cancer patients. However, difficulties in the stability, manufacture, and long-term administration of recombinant forms of endogenous antiangiogenic proteins have led investigators to develop gene therapy approaches to the antiangiogenic treatment of cancer (2). Endostatin, a 20-kd C-terminal fragment of collagen XVIII (3), is a potent antiangiogenic agent currently being evaluated in clinical trials (4).1 We tested whether retroviral transfer of the endostatin gene could generate sufficient functional endostatin in vivo to inhibit tumor growth.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionNotesMaterials and MethodsResultsDiscussionReferences |
|---|
Generation of Pseudotyped Retroviral Particles
The retroviral vector pCLNCX (5) and the vector pMD.G containing the G protein gene from vesicular stomatitis virus (VSV), as well as the cell line 293GP stably transfected with the retroviral gag and pol elements, were obtained from P. Robbins, National Cancer Institute (NCI), Bethesda, MD. The human elongation factor (EF) 1
promoter was isolated from the plasmid pAd.EF1 (Z. Guo, NCI) by BamHI/HindIII digestion (New England Biolabs Inc., Beverly, MA) and ligated in place of the downstream cytomegalovirus (CMV) promoter in BamHI/HindIII-digested pCLNCX to generate the vector pEF-null (Fig. 1
, a). Generation of the construct ss-mEndo, consisting of the 18-amino acid E3/19K signal sequence followed by the murine endostatin gene cloned from murine liver, has been described previously (6). The ss-mEndo construct was cloned into HindIII/ClaI-digested pEF-null to generate the vector pEF-Endo or into HindIII/ClaI-digested pCLNCX to generate the vector pCMV-Endo.
|
293GP cells were maintained in complete medium consisting of Dulbecco's modified Eagle medium (DMEM) with 10% fetal calf serum (FCS), 100 U/mL of penicillin, 100 µg/mL of streptomycin, 50 µg/mL of gentamicin, 0.5 µg/mL of Fungizone, and 4 mM glutamine (Biofluids, Rockville, MD). Cells were plated in 10-cm tissue culture dishes at a density of 3 x 106 cells/dish and allowed to adhere overnight in a 5% CO2 incubator at 37 °C. Cells then were rinsed with phosphate-buffered saline (PBS), cotransfected with 6 µg of retroviral plasmid and 6 µg of pMD.G (for VSV envelope protein expression) by use of 60 µL of Lipofectamine (Life Technologies, Inc. [GIBCO BRL], Rockville, MD) in 6 mL of serum-free DMEM, and incubated at 37 °C for 3 hours. After the addition of 12 mL of complete medium, cells were further incubated for 24 hours before the medium was replaced with 10 mL of fresh complete medium. After 24 hours, the supernatant containing the retroviral particles was collected, passed through a 0.45-µm (pore size) filter, and stored at -70 °C until ready for use.
Retroviral Transduction of NMuLi Cells
The cell line NMuLi, an epithelioid nonparenchymal cell line derived from the NAMRU mouse liver (7), was obtained from the American Type Culture Collection (Manassas, VA) and passaged in complete medium. We confirmed its ability to form malignant tumors in nude mice (8). For the retroviral transduction, cells were plated in six-well tissue culture plates at a density of 100 000 cells/well and allowed to adhere overnight. The medium then was replaced with 1 mL of fresh complete medium, 1 mL of retroviral supernatant, and 8 µg/mL of hexadimethrine bromide (Sigma Chemical Co., St. Louis, MO) to assist the uptake of viral particles. After incubation at 37 °C for 4 hours, the medium was replaced with 2 mL of fresh complete medium. The transduction was repeated the following day. Because the retroviral vector contains a selectable marker, the neomycin resistance gene, G418 (400 µg/mL; Life Technologies, Inc.), was added to the medium 24 hours after the second transduction. The surviving G418-resistant cells were amplified. To obtain individual clones, the endostatin-transduced cells (NEF-Endo) were plated in limiting dilution in complete medium containing G418 (400 µg/mL). The resultant clones were isolated by use of 8-mm diameter cloning cylinders (Specialty Media, Phillipsburg, NJ) and amplified.
To confirm endostatin expression, parental NMuLi cells, pEF-null-transduced cells (NEF-null), or NEF-Endo clones were plated in six-well plates at a density of 106 cells/well and allowed to adhere overnight. The medium was replaced with 1 mL/well of modified complete medium containing 5% FCS and incubated at 37 °C for 24 hours. Cell supernatants then were collected and passed through a 0.45-µm (pore size) filter. Endostatin concentrations in the supernatants were determined by competitive enzyme immunoassay (EIA) (Cytimmune Sciences, College Park, MD) according to the instructions of the manufacturer. Four endostatin-transduced clones with varying supernatant endostatin concentrations were selected for further study and designated NEF-Endo1 to 4, in decreasing order of endostatin production. The molecular weight of endostatin was determined from the culture supernatants by western blotting (NuPAGE; Novex, San Diego, CA) by use of 570 ng/mL of rabbit antimurine endostatin polyclonal immunogloblin G antibody (from Cytimmune Sciences). The EIA murine endostatin standard was used as a positive control.
Functional Assay of Retrovirally Derived Endostatin
To assess the functional activity of the retrovirally derived endostatin, we tested the culture supernatants in endothelial cell proliferation assays as described previously (6), with slight modifications. Briefly, bovine adrenal capillary endothelial cells (EJG; American Type Culture Collection) were plated in 200 µL of complete medium in collagen I-coated 96-well plates (Biocoat; Becton Dickinson Labware, Bedford, MA) at a density of 1000 cells/well and were incubated overnight at 37 °C. The medium then was aspirated and replaced with 100 µL of cell supernatant or modified complete medium containing 5% FCS. Basic fibroblast growth factor (R&D Systems, Inc., Minneapolis, MN) at a final concentration of 1 ng/mL was added to all wells as a stimulus of proliferation. After incubation at 37 °C for 72 hours, proliferation was analyzed by the WST-1 assay (Roche, Indianapolis, IN) according to the manufacturer's instructions. The inhibition of proliferation by each sample was calculated according to the formula: inhibition (%) = (mean 
ODmedia – ODsample/mean ODmedia) x 100, where OD represents the change in optical density at 450 nm (Multiskan MCC/340 plate reader; Titertek, Huntsville, AL) from before the addition of the WST-1 reagent to the completion of the 4 hours' incubation at 37 °C. Each sample was measured in six individual wells. The experiment was repeated to confirm results.
In Vitro Growth of Parental and Transduced NMuLi Cells
To place in perspective any differences noted in the in vivo growth of parental and transduced cell lines, we compared in vitro growth rates. Five identical 96-well tissue culture plates were prepared. On each plate, NMuLi, NEF-null, and NEF-Endo1 to 4 cells were each plated in eight identical wells at a density of 1000 cells/well in 100 µL of complete medium. WST-1 proliferation assays were performed daily on days 0–4, as described above. Cells were allowed to adhere for 3 hours before the day-0 assay was performed. Proliferation was expressed as the OD as defined above.
Tumor Formation by Retrovirally Transduced Cells
Animal experiments were conducted according to protocols approved by the National Institutes of Health Animal Care and Use Committee (Bethesda, MD). Eight-week-old female nude mice (Charles River Laboratories, Wilmington, DE) were given an injection of 5 x 105 parental NMuLi cells, NEF-null cells, or NEF-Endo1 to 4 cells. Subcutaneous injections were administered in the right flank in 100 µL of PBS. Tumors were measured in two dimensions by use of calipers at regular intervals, and tumor volumes were calculated according to the following formula: volume = width2 x length x 0.52, where 0.52 is a constant to calculate the volume of an ellipsoid. Intraperitoneal injections were administered in 2 mL of PBS. All animals were followed for survival. Each group consisted of eight mice.
Measurement of Endostatin Levels In Vivo
Mice were given an inoculation of NEF-null or NEF-Endo1 cells subcutaneously or intraperitoneally as described above. Serum samples obtained 45 days (subcutaneous model) or 21 days (intraperitoneal model) after tumor cell inoculation were analyzed for endostatin concentration by EIA. Serum samples also were obtained from non-tumor-bearing animals. Each group consisted of six mice.
To assess in vivo local endostatin expression, we harvested NEF-null or NEF-Endo1 tumors immediately after the mice were killed 38 days after subcutaneous injection as described above. Tumors were snap-frozen in liquid nitrogen and stored at -80 °C until ready for analysis. Tumors then were homogenized at room temperature in a protease inhibitor cocktail (Complete Mini; Roche) dissolved in 10 mM HEPES and 1 mM EDTA by use of a bead homogenizer (FastPrep; Savant Instruments, Holbrook, NY). Homogenates were centrifuged at 7500g for 5 minutes at 4 °C, and supernatants were analyzed for endostatin by EIA. Endostatin concentrations were normalized to total protein concentrations determined by use of a BCA protein assay kit (Pierce Chemical Co., Rockford, IL) and a linear bovine serum albumin standard curve, according to the manufacturer's instructions.
Immunohistochemistry and Histopathologic Evaluation
NEF-null or NEF-Endo1 tumors were harvested immediately after the mice were killed 38 days after subcutaneous injection, as described above. Tumors were snap-frozen in liquid nitrogen and stored at -80 °C. Frozen sections (7 µm) were cut on a cryostat and stained with hematoxylin–eosin (H&E) or antibodies specific for CD31 (PharMingen, San Diego, CA), proliferating cell nuclear antigen (PCNA) (Zymed Laboratories, San Francisco, CA), or caspase-3 (PharMingen). For immunostaining, sections were fixed in acetone for 10 minutes (CD31) or in 4% formalin for 1 hour (PCNA and caspase-3). After endogenous peroxidase activity was blocked by use of 3% hydrogen peroxide in methanol for 10 minutes, sections were incubated for 1 hour in a blocking solution containing 10% normal goat serum. Sections were incubated with primary antibody at 4 °C overnight (CD31 at 1 : 50 dilution or caspase-3 at a concentration of 1 µg/mL) or at room temperature for 1 hour (PCNA at 1 : 50 dilution). Slides were then washed three times in PBS, incubated in biotinylated species-specific appropriate secondary antibody for 1 hour, and exposed to avidin–biotin–peroxidase complex (Vector Laboratories, Inc., Burlingame, CA). Sections were reacted with 0.06% 3,3'-diaminobenzidine (Sigma Chemical Co.) and counterstained with hematoxylin.
H&E and immunostained sections were analyzed by a pathologist (S. M. Hewitt), who was blinded to the identity of the groups. Only good-quality sections with uniform, well-demarcated staining and low background were analyzed. Microvascular density of each tumor was assessed by use of a scoring system (Table 1) based on the mean number of CD31-positive cells per high-power field (hpf, x600). Proliferative and apoptotic cells were analyzed by use of a similar scoring system (Table 1) based on the number of cells per hpf staining positively for PCNA and caspase-3, respectively.
|
Statistical Analysis
Data are presented as means with 95% confidence intervals (CIs). Comparisons between groups were made by use of the Mann–Whitney U test or the Kruskal–Wallis test, where appropriate. The endostatin dose dependence of endothelial cell inhibition was analyzed by use of the Jonckheere–Terpstra trend test. Two-tailed P values <.05 were considered to be statistically significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionNotesMaterials and MethodsResultsDiscussionReferences |
|---|
In Vitro Production of Endostatin
Initial transduction of NMuLi cells with pCMV-Endo did not yield supernatant endostatin concentrations above baseline. We, therefore, replaced the downstream CMV promoter in pCLNCX with the EF1 promoter (Fig. 1
, a) based on favorable results reported with the use of the EF1 promoter in other viral gene delivery systems (9,10). Endostatin levels in supernatants from NEF-Endo1, 2, 3, and 4, NEF-null, and NMuLi cells were 223, 163, 42, 28, 20, and 4 ng/mL per 106 cells, respectively. The NEF-Endo clone supernatants produced specific 20-kd bands on western blot proportional in intensity to the endostatin concentration as measured by EIA (Fig. 1, b).
In Vitro Function of Endostatin
To determine if the endostatin produced by the transduced clones was biologically active, we tested the supernatants in an endothelial cell proliferation assay. Compared with unconditioned medium, supernatants from the transduced clones NEF-Endo 1, 2, 3, and 4, and NEF-null inhibited endothelial cell proliferation by 20% (95% CI = 13% to 27%), 11% (95% CI = 3% to 19%), 9% (95% CI = –1% to 19%), 6% (95% CI = 0% to 12%), and 4% (95% CI = –8% to 17%), respectively (Fig. 1, c). The difference in inhibition between supernatants from NEF-Endo1 and NEF-null was statistically significant (P = .037). The dose-dependent inhibition of endothelial cell proliferation by endostatin also was statistically significant (P = .01).
In Vitro Growth of Parental and Transduced NMuLi Cells
We next compared the in vitro growth of the transduced cells with that of the parental NMuLi cells. All four NEF-Endo clones demonstrated similar patterns of in vitro growth to each other and to NEF-null cells (Fig. 1, d). Parental NMuLi cells grew more slowly in vitro than their transduced, G418-selected counterparts.
Subcutaneous Growth of Parental and Transduced NMuLi Cells
Mice given a subcutaneous injection of NMuLi or NEF-null cells developed rapidly growing tumors and were killed 63 days after injection when the tumor volumes were 2400 mm3 (95% CI = 1478 mm3 to 3300 mm3) and 2700 mm3 (95% CI = 2241 mm3 to 3144 mm3), respectively (Fig. 2, a). The mean tumor volumes were less than 30 mm3 in all four groups given an injection of NEF-Endo clones (P.<001 versus null-transduced tumors). Although the largest endostatin-transduced tumors were observed in the mice that were given an injection of the NEF-Endo4 clone, which produced the lowest amount of endostatin in vitro (NEF-Endo4; Fig. 2, a, inset), the difference between the size of the endostatin-transduced tumors was not statistically significant (P = .52 at day 63). After 122 days of follow-up, mice given an injection of NEF-Endo clones 1, 3, and 4 had tumor volumes of 70 mm3 (95% CI = 0 mm3 to 202 mm3), 300 mm3 (95% CI = 123 mm3 to 419 mm3), and 90 mm3 (95% CI = 0 mm3 to 279 mm3), respectively. Mice given an injection of the NEF-Endo clone 2 had nonpalpable or barely palpable lesions.
|
), NEF-null cells (
), or NEF-Endo clones 1 (•), 2 (
), 3 (
), and 4 (
). Each point represents the mean volume of eight mice. Inset shows expanded y-axis to demonstrate tumor volumes of NEF-Endo clones. Control (NMuLi and NEF-null) mice were killed on day 63, at which time the mean tumor volumes among NEF-Endo groups were less than 30 mm3. Panel b: mouse survival after intraperitoneal inoculation of parental and transduced NMuLi cells. Each group contained eight mice. Median survival times were 58 days for mice receiving NMuLi cells and 56 days for mice receiving NEF-null cells. All control animals died by day 123, at which time only three (9%) of the 32 animals receiving NEF-Endo clones had died. The numbers of mice at risk in each group after 60 days were as follows: NMuLi = 2 (95% confidence interval [CI] = 0.2 to 3.8); NEF-null = 3 (95% CI = 0.6 to 4.7); NEF-Endo1 = 8 (95% CI = 5.0 to 8.0); NEF-Endo2 = 8 (95% CI = 5.0 to 8.0); NEF-Endo3 = 8 (95% CI = 5.0 to 8.0); and NEF-Endo4 = 8 (95% CI = 5.0 to 8.0). The numbers of mice at risk in each group after 120 days were as follows: NMuLi = 0; NEF-null = 1 (95% CI = 0.0 to 4.2); NEF-Endo1 = 7 (95% CI = 4.2 to 7.8); NEF-Endo2 = 6 (95% CI = 3.3 to 7.4); NEF-Endo3 = 8 (95% CI = 5.0 to 8.0); and NEF-Endo4 = 8 (95% CI = 5.0 to 8.0).Intraperitoneal Growth of Parental and Transduced NMuLi Cells
Mice given an intraperitoneal injection of parental NMuLi or NEF-null cells had median survival times of 58 days (95% CI = 49 days to 61 days) and 56 days (95% CI = 46 days to 92 days), respectively, and all mice were dead by day 123 (Fig. 2, b). At this time, only three (9%) of the 32 mice receiving intraperitoneal NEF-Endo clones had died. Autopsies of mice that died revealed massive peritoneal tumor deposits and ascites. Surviving mice were killed and had an autopsy, revealing only occasional, small peritoneal deposits.
In Vivo Endostatin Levels
Serum endostatin levels were similar in mice bearing subcutaneous NEF-Endo1 tumors (mean, 33.8 ng/mL; 95% CI = 18.5 ng/mL to 49.1 ng/mL) and NEF-null tumors (mean, 32.6 ng/mL; 95% CI = 15.8 ng/ml to 49.4 ng/mL), respectively (P = .87). Serum endostatin levels in mice bearing intraperitoneal NEF-Endo1 tumors (mean, 40.0 ng/mL; 95% CI = 7.8 ng/mL to 72.2 ng/mL) were slightly higher than those in mice bearing intraperitoneal NEF-null tumors (mean, 22.2 ng/mL; 95% CI = 18.3 ng/mL to 26.1 ng/mL), but this difference was not statistically significant (P = .42). Serum endostatin levels in non-tumor-bearing mice were 27.1 ng/mL (95% CI = 15.3 ng/mL to 38.9 ng/mL), similar to values we reported previously (11) and not statistically significantly different from any of the other groups tested.
To confirm endostatin production in the NEF-Endo tumors, we measured local endostatin levels in tumor lysates 38 days after subcutaneous injection. Tumor lysates from endostatin- and null-transduced tumor had 38.7 ng (95% CI = 20.7 ng to 56.6 ng) and 11.2 ng (95% CI = 8.8 ng to 13.7 ng) endostatin/mg total protein, respectively (P = .006).
Immunohistologic Characteristics of Transduced Tumors
Evaluation of H&E-stained tumor sections revealed small tumor nodules (mean diameter, 1.7 mm) derived from NEF-Endo1 and NEF-Endo2 cells (Table 1; Fig. 3, a); tumors derived from NEF-null cells were larger (mean diameter, 4.9 mm) and had zones of central necrosis (Fig. 3, b). Mean microvessel density scores for endostatin- and null-transduced tumors were 3 and 4, respectively (Table 1; Fig. 3, c–e). Endostatin-transduced tumors did not have discernible vessels on H&E stains (Fig. 3, a) but did contain CD31-positive endothelial cells as individual capillaries (Fig. 3, c). By contrast, the null-transduced tumors contained easily identifiable vessels on H&E stains, away from zones of necrosis (Fig. 3, b). The CD31 staining highlighted larger vessels as well as individual endothelial cells (Fig. 3, d).
|
Finally, we determined the apoptotic and proliferative scores for the endostatin- and null-transduced tumors. The null-transduced tumors had higher apoptotic and proliferative scores than the endostatin-transduced tumors, with mean apoptosis scores for endostatin- and null-transduced tumors of 1 and 2.3, respectively, and mean proliferative scores of 3 and 4, respectively (Table 1
; Fig. 3, e). |
DISCUSSION |
|---|
|
TopAbstractIntroductionNotesMaterials and MethodsResultsDiscussionReferences |
|---|
The demonstration that tumors require neoangiogenesis for sustained growth (1) has prompted intense investigation into the inhibition of tumor angiogenesis as a strategy for treating cancer patients. More than 40 endogenous inhibitors of angiogenesis have been described previously (2). Of interest, some of these inhibitors are tumor derived, including endostatin, a 20-kd C-terminal fragment of collagen XVIII originally isolated from a murine hemangioendothelioma cell line (3). There are, however, a number of potential limitations to treating cancer with antiangiogenic agents, such as endostatin, in their recombinant forms. First, some of these recombinant biologics, including endostatin, are relatively unstable in vitro and require high doses for antitumor efficacy (3). Second, many of these agents are not cytotoxic to tumor cells themselves and may need to be administered chronically (12). Finally, systems that allow continuous delivery of antiangiogenic agents appear to be therapeutically preferable to the peak/trough kinetics associated with bolus infusion of recombinant proteins (13,14). These limitations have prompted investigators to study gene therapy approaches to deliver antiangiogenic proteins for the treatment of cancer (2).
Previously, we and others (6,15,16) have shown that adenoviral transfer of the endostatin gene to mice leads to high circulating endostatin levels that inhibit tumor growth. Several groups (17–19) also have demonstrated that nonviral endostatin gene transfer can cause an antitumor effect in mice. Both of these approaches, however, are associated with limitations in their ability to be translated for the treatment of human patients: The use of adenoviral vectors is limited because of immunogenicity and toxicity, and nonviral vectors generally are associated with low levels of transgene expression. Because retroviruses have been used safely in human gene therapy trials (20,21), we investigated whether retroviral transfer of the endostatin gene to tumor cells could result in sufficient functional protein to inhibit tumor growth in vivo.
Because endostatin occurs naturally as a cleavage product of collagen XVIII, the coding sequence for endostatin lacks a secretion signal. We previously generated a construct in which the endostatin coding sequence is preceded by the adenoviral protein E3/19K signal sequence (6,22). We cloned this construct into the multiple cloning site of pCLNCX, a Moloney murine leukemia virus-derived retroviral expression plasmid (5), and generated pseudotyped retroviral particles containing the VSV glycoprotein G (23). NMuLi cells transduced with this construct were resistant to G418, but they did not generate high levels of endostatin in their cell supernatants. We hypothesized that this was a promoter-related phenomenon and replaced the downstream CMV promoter in the pCLNCX plasmid with the constitutively active human EF-1 promoter. Cells transduced with this modified construct secreted high endostatin levels, as measured in a competitive EIA.
To verify that the secreted protein measured in the EIA was, in fact, endostatin, we performed western blotting of aliquots from supernatants of the transduced cells. Supernatants from NEF-Endo clones produced bands of equal mobility to that of recombinant murine endostatin and of intensities proportional to the EIA-measured endostatin concentrations. When these supernatants were placed on bovine adrenal capillary endothelial cells, they inhibited endothelial cell proliferation in a dose-dependent fashion, confirming the activity of the secreted endostatin protein. The secreted endostatin was relatively endothelial cell specific, as has been reported previously (3), because the growth rates of the NEF-null cells and the NEF-Endo clones were similar, including NEF-Endo1 cells that secreted the highest level of endostatin.
By contrast with the in vitro growth, transduction of NMuLi cells with the endostatin gene profoundly inhibited their subcutaneous growth in mice. Consistent with the antiangiogenic function of endostatin, NEF-Endo cells were capable of forming tumors in vivo, although these tumors were very small and displayed fewer microvessels than their null-transduced counterparts. The NEF-Endo4 clone, which produced the least endostatin in vitro, initially produced larger tumors than the other NEF-Endo clones tested. However, the inhibition of in vivo tumor growth of all clones was not directly proportional to in vitro endostatin expression. In fact, the growth curves of the four NEF-Endo clones began to diverge after approximately 100 days, although tumors in all groups remained much smaller than tumors in the NEF-null mice, which exceeded 2 cm3. The eventual divergence of the growth curves among the NEF-Endo clones growth may reflect clonal variations unrelated to endostatin. Alternatively, local endostatin production at late time points in vivo may not correspond with in vitro expression by various clones.
Of interest, endostatin transduction of NMuLi cells also inhibited their ability to form aggressive lesions in the peritoneal cavity. Like tumors in other sites, the pathogenesis of tumors of the peritoneal cavity has been shown to be associated with angiogenesis (1,24,25). In addition, several inhibitors of angiogenesis have been reported to inhibit peritoneal tumor growth in rodent models, including protamine (26), TNP-470 (27–29), and agents that inhibit the action of vascular endothelial growth factor (30–32). Our data suggest that endostatin also can inhibit peritoneal tumor growth. Similar to the results seen with our subcutaneous model, the tumor lesions present in mice receiving intraperitoneal NEF-Endo cells were generally small. In mice with subcutaneous or intraperitoneal tumors, circulating endostatin levels were not statistically significantly increased in mice harboring endostatin-transduced tumors. Because of the minimal amount of tumor tissue in these animals and the baseline mean serum endostatin concentration of 27 ng/mL, small amounts of endostatin released into the circulation may not have been detectable by the assay employed. The assay was, however, sensitive enough to detect increased concentrations of endostatin in lysates of endostatin-transduced tumors, confirming that endostatin was being produced locally in these animals.
Our results indicate that retroviral transfer of the endostatin gene to eukaryotic cells does not inhibit their growth in vitro but generates a secreted, functional endostatin protein that inhibits endothelial cells in vitro and tumor formation at multiple sites in vivo. In addition, these data support previous evidence (26–32) that inhibiting angiogenesis is a promising strategy for treating tumors of the peritoneal cavity. However, one limitation of this study is that ex vivo transduction of tumor cells is not a strategy applicable to clinical therapeutic use. Recently, retroviruses secreting another antiangiogenic molecule, interleukin 12 (IL-12), have been used in preclinical models of disseminated peritoneal cancer. Sanches et al. (33) transduced fibroblasts with the IL-12 gene ex vivo, then introduced them into mice with peritoneal ovarian carcinomatosis. Lechanteur et al. (34) used an intraperitoneal injection of retroviral packaging cells carrying the IL-12 gene as part of a combined approach to treating peritoneal colon carcinomatosis in rats. In each case, at least part of the tumor response was because of immune (i.e., nonantiangiogenic) mechanisms, and the immune effects of IL-12 have been associated with considerable toxicity in clinical and preclinical models. By contrast, endostatin appears to impart its antitumor effect solely by inhibiting angiogenesis, with minimal toxicity(3,35). Therefore, further studies using retroviral endostatin gene transfer for the treatment of cancer, including disseminated tumors of the peritoneal cavity, are warranted.
|
NOTES |
|---|
1 An overview of angiogenesis inhibitors in clinical trials is available from URL: http://cancertrials.nci.nih.gov/news/angio/table.html (accessed April 26, 2001).
We thank S. Steinberg, National Cancer Institute (Bethesda, MD), for his help with the statistical analysis.
|
REFERENCES |
|---|
|
TopAbstractIntroductionNotesMaterials and MethodsResultsDiscussionReferences |
|---|
1
Folkman J. What is the evidence that tumors are angiogenesis dependent? [editorial]. J Natl Cancer Inst 1990;82:4–6.
2 Feldman AL, Libutti SK. Progress in antiangiogenic gene therapy of cancer. Cancer 2000;89:1181–94.[CrossRef][ISI][Medline]cancerlit;20458923
3 O'Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, et al. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell 1997;88:277–85.[CrossRef][ISI][Medline]cancerlit;97160848
4 Herbst RS, Lee AT, Tran HT, Abbruzzese JL. Clinical studies of angiogenesis inhibitors: the University of Texas MD Anderson Center Trial of Human Endostatin. Curr Oncol Rep 2001;3:131–40.[Medline]cancerlit;21114649
5
Naviaux RK, Costanzi E, Haas M, Verma IM. The pCL vector system: rapid production of helper-free, high-titer, recombinant retroviruses. J Virol 1996;70:5701–5.
6
Feldman AL, Restifo NP, Alexander HR, Bartlett DL, Hwu P, Seth P, et al. Antiangiogenic gene therapy of cancer utilizing a recombinant adenovirus to elevate systemic endostatin levels in mice. Cancer Res 2000;60:1503–6.
7 Owens RB, Smith HS, Hackett AJ. Epithelial cell culture from normal glandular tissue of mice. J Natl Cancer Inst 1974;53:261–9.
8 Anderson LW, Smith HS. Premalignancy in vitro: progression of an initially benign epithelial cell line to malignancy. Br J Exp Pathol 1979;60:575–81.[ISI][Medline]
9 Guo ZS, Wang LH, Eisensmith RC, Woo SL. Evaluation of promoter strength for hepatic gene expression in vivo following adenovirus-mediated gene transfer. Gene Ther 1996;3:802–10.[ISI][Medline]cancerlit;97029217
10
Salmon P, Kindler V, Ducrey O, Chapuis B, Zubler RH, Trono D. High-level transgene expression in human hematopoietic progenitors and differentiated blood lineages after transduction with improved lentiviral vectors. Blood 2000;96:3392–8.
11
Feldman AL, Libutti SK. Correspondence re: Q-R Chen et al., Liposomes complexed to plasmids encoding angiostatin and endostatin inhibit breast cancer in nude mice [letter]. Cancer Res 2000;60:1463–4.
12 Folkman J. The influence of angiogenesis research on management of patients with breast cancer. Breast Cancer Res Treat 1995;36:109–18.[CrossRef][ISI][Medline]cancerlit;96049796
13
Drixler TA, Rinkes IH, Ritchie ED, van Vroonhoven TJ, Gebbink MF, Voest EE. Continuous administration of angiostatin inhibits accelerated growth of colorectal liver metastases after partial hepatectomy. Cancer Res 2000;60:1761–5.
14 Joki T, Machluf M, Atala A, Zhu J, Seyfried NT, Dunn IF, et al. Continuous release of endostatin from microencapsulated engineered cells for tumor therapy. Nat Biotechnol 2001;19:35–9.[CrossRef][ISI][Medline]cancerlit;20580683
15
Sauter BV, Martinet O, Zhang WJ, Mandeli J, Woo SL. Adenovirus-mediated gene transfer of endostatin in vivo results in high level of transgene expression and inhibition of tumor growth and metastases. Proc Natl Acad Sci U S A 2000;97:4802–7.
16 Chen CT, Lin J, Li Q, Phipps SS, Jakubczak JL, Stewart DA, et al. Antiangiogenic gene therapy for cancer via systemic administration of adenoviral vectors expressing secretable endostatin. Hum Gene Ther 2000;11:1983–96.[CrossRef][ISI][Medline]cancerlit;20476832
17
Yoon LS, Eto H, Lin CM, Nakamura H, Pawlik TM, Song SU, et al. Mouse endostatin inhibits the formation of lung and liver metastases. Cancer Res 1999;59:6251–6.
18 Blezinger P, Wang J, Gondo M, Quezeda A, Mehrens D, French M, et al. Systemic inhibition of tumor growth and tumor metastases by intramuscular administration of the endostatin gene. Nat Biotechnol 1999;17:343–8.[CrossRef][ISI][Medline]cancerlit;99224288
19
Chen QR, Kumar D, Stass SA, Mixson AJ. Liposomes complexed to plasmids encoding angiostatin and endostatin inhibit breast cancer in nude mice. Cancer Res 1999;59:3308–12.
20 Rosenberg SA, Aebersold P, Cornetta K, Kasid A, Morgan RA, Moen R, et al. Gene transfer into humans—immunotherapy of patients with advanced melanoma, using tumor-infiltrating lymphocytes modified by retroviral gene transduction. N Engl J Med 1990;323:570–8.[Abstract]cancerlit;90340355
21
Cavazzana-Calvo M, Hacein-Bey S, de Saint Basile G, Gross F, Yvon E, Nusbaum P, et al. Gene therapy of human severe combined immunodeficiency (SCID)-X1 disease. Science 2000;288:669–72.
22
Persson H, Jornvall H, Zabielski J. Multiple mRNA species for the precursor to an adenovirus-encoded glycoprotein: identification and structure of the signal sequence. Proc Natl Acad Sci U S A 1980;77:6349–53.
23
Emi N, Friedmann T, Yee JK. Pseudotype formation of murine leukemia virus with the G protein of vesicular stomatitis virus. J Virol 1991;65:1202–7.
24
Nagy JA, Morgan ES, Herzberg, KT, Manseau EJ, Dvorak AM, Dvorak HF. Pathogenesis of ascites tumor growth: angiogenesis, vascular remodeling, and stroma formation in the peritoneal lining. Cancer Res 1995;55:376–85.
25
Yoneda J, Kuniyasu H, Crispens MA, Price JE, Bucana CD, Fidler IJ. Expression of angiogenesis-related genes and progression of human ovarian carcinomas in nude mice. J Natl Cancer Inst 1998;90:447–54.
26 Heuser LS, Taylor SH, Folkman J. Prevention of carcinomatosis and bloody malignant ascites in the rat by an inhibitor of angiogenesis. J Surg Res 1984;36:244–50.[CrossRef][ISI][Medline]cancerlit;84140239
27 Tsujimoto H, Hagiwara A, Osaki K, Ohyama T, Sakakibara T, Sakuyama A, et al. Therapeutic effects of the angiogenesis inhibitor TNP-470 against carcinomatous peritonitis in mice. Anticancer Drugs 1995;6:438–42.[Medline]cancerlit;95399818
28 Kanai T, Konno H, Tanaka T, Matsumoto K, Baba M, Nakamura S, et al. Effect of angiogenesis inhibitor TNP-470 on the progression of human gastric cancer xenotransplanted into nude mice. Int J Cancer 1997;71:838–41.[CrossRef][ISI][Medline]cancerlit;97323958
29 Yoshikawa T, Yanoma S, Ysuburaya A, Kobayashi O, Sairenji M, Motohashi H, et al. Angiogenesis inhibitor, TNP-470, suppresses growth of peritoneal disseminating foci. Hepatogastroenterology 2000;47:298–302.[Medline]cancerlit;20155214
30 Xu L, Yoneda J, Herrera C, Wood J, Killion JJ, Fidler IJ. Inhibition of malignant ascites and growth of human ovarian carcinoma by oral administration of a potent inhibitor of the vascular endothelial growth factor receptor tyrosine kinases. Int J Oncol 2000;16:445–54.[ISI][Medline]cancerlit;20141533
31 Mori A, Arii S, Furutani M, Mizumoto M, Uchida S, Furuyama H, et al. Soluble Flt-1 gene therapy for peritoneal metastases using HVJ-cationic liposomes. Gene Ther 2000;7:1027–33.[CrossRef][ISI][Medline]cancerlit;20332428
32 Shaheen RM, Tseng WW, Vellagas R, Liu W, Ahmad SA, Jung YD, et al. Effects of an antibody to vascular endothelial growth factor receptor-2 on survival, tumor vascularity, and apoptosis in a murine model of colon carcinomatosis. Int J Oncol 2001;18:221–6.[ISI][Medline]cancerlit;21110930
33 Sanches R, Kuiper M, Penault-Llorca F, Aunoble B, D'Incan C, Bignon YJ. Antitumoral effect of interleukin-12-secreting fibroblasts in a mouse model of ovarian cancer: implications for the use of ovarian cancer biopsy-derived fibroblasts as a vehicle for regional gene therapy. Cancer Gene Ther 2000;7:707–20.[CrossRef][ISI][Medline]cancerlit;20289530
34
Lechanteur C, Delvenne P, Princen F, Lopez M, Fillet G, Gielen J, et al. Combined suicide and cytokine gene therapy for peritoneal carcinomatosis. Gut 2000;47:343–8.
35 Berger AC, Feldman AL, Gnant MF, Kruger EA, Sim BK, Hewitt S, et al. The angiogenesis inhibitor, endostatin, does not affect murine cutaneous wound healing. J Surg Res 2000;91:26–31.[CrossRef][ISI][Medline]cancerlit;20278204
Manuscript received December 15, 2000;
revised May 1, 2001; revised May 9, 2001;
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. L. Coutinho, L. N. de Sousa Andrade, R. Chammas, L. Morganti, N. Schor, and M. H. Bellini Anti-tumor effect of endostatin mediated by retroviral gene transfer in mice bearing renal cell carcinoma FASEB J, October 1, 2007; 21(12): 3153 - 3161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. L. Feldman, W. G. Stetler-Stevenson, N. G. Costouros, V. Knezevic, G. Baibakov, H. R. Alexander Jr., D. Lorang, S. M. Hewitt, D.-W. Seo, M. S. Miller, et al. Modulation of Tumor-host Interactions, Angiogenesis, and Tumor Growth by Tissue Inhibitor of Metalloproteinase 2 via a Novel Mechanism Cancer Res., July 1, 2004; 64(13): 4481 - 4486. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Wickstrom, K. Alitalo, and J. Keski-Oja Endostatin Associates with Lipid Rafts and Induces Reorganization of the Actin Cytoskeleton via Down-regulation of RhoA Activity J. Biol. Chem., September 26, 2003; 278(39): 37895 - 37901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. A. Scappaticci Mechanisms and Future Directions for Angiogenesis-Based Cancer Therapies J. Clin. Oncol., September 15, 2002; 20(18): 3906 - 3927. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Ye, C. Feng, S. Wang, X. Liu, Y. Lin, and M. Li Antiangiogenic and Antitumor Effects of Endostatin on Follicular Thyroid Carcinoma Endocrinology, September 1, 2002; 143(9): 3522 - 3528. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-i. Hanai, J. Gloy, S. A. Karumanchi, S. Kale, J. Tang, G. Hu, B. Chan, R. Ramchandran, V. Jha, V. P. Sukhatme, et al. Endostatin is a potential inhibitor of Wnt signaling J. Cell Biol., August 5, 2002; 158(3): 529 - 539. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. K. Kleinman and G. Liau Gene Therapy for Antiangiogenesis J Natl Cancer Inst, July 4, 2001; 93(13): 965 - 967. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||