Intramuscular delivery of antiangiogenic genes suppresses ... - Nature

Cancer Gene Therapy (2005) 12, 35–45 All rights reserved 0929-1903/05 $30.00

r 2005 Nature Publishing Group

www.nature.com/cgt

Intramuscular delivery of antiangiogenic genes suppresses secondary metastases after removal of primary tumors Xueying Sun,1,2 Haiquan Qiao,3 Hongchi Jiang,3 Xuting Zhi,1 Fengjun Liu,1 Jianli Wang,1 Meng Liu,1 Dianning Dong,1 Jagat R Kanwar,2 Ruian Xu,4 and Geoffrey W Krissansen2 1

Department of Surgery, Qilu Hospital, Shandong University, Jinan 250012, China; 2Department of Molecular Medicine & Pathology, School of Medicine, University of Auckland, Auckland, New Zealand; 3Department of Surgery, The first Clinical College of Harbin Medical University, Harbin 150001, China; and 4Gene therapy Laboratory, Institute of Molecular Biology, University of Hong Kong, Hong Kong.

The success of surgery to remove primary tumors can be compromised by the subsequent outgrowth of metastases. It is recognized that primary tumors secrete antiangiogenic factors that suppress the outgrowth of their daughter metastases. In accord we show here that surgical removal of primary EL-4 lymphomas led to a marked decrease in the levels of circulating angiostatin and endostatin, and promoted the growth of distant nodular tumors. Expression vectors encoding angiostatin and endostatin, formulated with polyN-vinyl pyrrolidone (PVP), were injected into the tibialis and gastrocnemia muscles, leading to expression of angiostatin and endostatin in muscle fibers. High levels of biologically active exogenous proteins were secreted into the circulation. Intramuscular gene therapy with angiostatin and endostatin plasmids significantly inhibited tumor vascularity and induced tumor cell apoptosis, and thereby suppressed the growth of secondary subcutaneous and disseminated metastatic tumors in the lung and liver. Simultaneous intramuscular delivery of both angiostatin and endostatin plasmids significantly prolonged the survival of mice after removal of primary tumors. These results suggest that intramuscular gene transfer of angiostatin and endostatin might serve as a prophylactic cancer-prevention strategy to combat the recurrence of cancer after surgical resection of primary tumors. Cancer Gene Therapy (2005) 12, 35–45. doi:10.1038/sj.cgt.7700766 Published online 15 October 2004 Keywords: angiostatin; endostatin; metastasis; antiangiogenic therapy

umor metastases do not always outgrow at distant sites, but can remain dormant until they are T awakened by a suitable stimulus. They pose a constant risk for tumor recurrence, and are responsible for most cancer patient morbidity and mortality.1,2 One of the recognized stimulators is surgical removal of primary tumors. Primary tumors secrete endogenous antiangiogenic factors such as angiostatin3 and endostatin,4 which hold the growth of metastases in check. Angiostatin and endostatin are two of the most potent angiogenesis inhibitors isolated from the sera and urine of tumorbearing mice. They both inhibit endothelial cell migration and proliferation, and induce the regression of a wide variety of established tumors of mouse, rat and human origin, including metastases.4–9 The tumor-suppressor activity of angiostatin may arise from its ability to inhibit the proliferation of endothelial cells by binding to the a/bsubunits of ATP synthase,10 inducing apoptotic cell death,5 by subverting adhesion plaque formation and thereby inhibiting the migration and tube formation of Received March 27, 2004.

Address correspondence and reprint requests to: Dr Xueying Sun, Department of Surgery, Qilu Hospital of Shandong University, 107 Wenhuaxi Road, Jinan 250012, Shandong Province, China. E-mail: [email protected]

endothelial cells, and/or by downregulating vascular endothelial growth factor (VEGF) expression.11,12 The latter feature of angiostatin is not retained in all tumor models, as it had no effect on VEGF expression in two reports,13,14 and even slightly upregulated VEGF expression in EL-4 tumors.15 On the other hand, endostatin’s antitumor activity depends on its binding to the catalytic domain of matrix metalloproteinase-2,16 blocking VEGFmediated signaling via direct interaction with KDR/Flk1,17 inhibiting Wnt signaling18 and adhesion of endothelial cells to collagen I via a2b1 integrin,19 and attenuating endothelial cell migration.20 These observations indicate that angiostatin and endostatin have potential synergistic antiangiogenic activity, a notion that has already been confirmed.21,22 The clinical application of antiangiogenic therapies has been hindered by high-dose requirements, manufacturing constraints, and the relative instability of the corresponding recombinant proteins,23,24 despite their broad-spectrum action, and low toxicity.25 Gene transfer represents an alternative method to deliver these agents.26,27 Compared to intratumoral or intravenous gene injection, intramuscular gene delivery is more practical and safer for clinical application. Ever since the first report using skeletal muscle as a direct target for expressing foreign transgenes,28 an avalanche of papers has identified a

Antiangiogenic therapy suppresses metastases X Sun et al

36

variety of proteins that can be synthesized and correctly processed by skeletal muscle. The most efficient way of introducing antiangiogenic transgenes into muscle fibers may be via recombinant viral vectors. In a recent report, adeno-associated virus (AAV) mediated intramuscular gene transfer of angiostatin and endostatin resulted in suppression of primary and metastatic tumours.29 However, there are potential benefits in the use of nonviral vectors.30 Viral vectors may cause inflammation and immunological response on repeated injection,31–33 and their safety issues and toxicity may preclude their use in human beings in the near future. Plasmid DNA does not pose the health risks entailed by viral infection, it is easy to propagate on a large scale at high quality, and plasmids are able to carry relatively large DNA sequences. Naked plasmid DNA itself has low immunogenicity and can be administered repeatedly.34 The major drawback of plasmid DNA as a vector is its low efficiency for gene delivery. Nevertheless, cationic substances that are widely used as vehicle diluents, are easily manipulated, safe, and improve gene delivery. Every muscle fiber is surrounded by a layer of mechanically strong and negatively charged basement membrane rich in glycosaminoglycans. In contrast, the commonly used cationic lipid formulations, such as DOTAP and cationic polymers are positively charged. These directly injected lipoplexes can bind to negatively charged extracellular matrix (ECM) components,35 which may be the main reason counting for low transfection efficiency. However, incorporating nonionic polymers such as poly-N-vinyl pyrrolidone (PVP) into plasmid formulations circumvents this problem, and can result in up to a 10-fold enhancement of transgene expression in muscle fibers over that obtained using naked plasmid alone.36,37 Intramuscular administration of endostatin assisted by PVP was found to be effective in inhibiting the growth of metastatic brain tumors.38 Here, we simulate the process of tumor metastasis using EL-4 tumor cells that are relatively antigenic, but become potently immunosuppressive as they grow to form tumors. We evaluate the effect of intramuscular gene delivery of angiostatin and endostatin in suppressing the growth of secondary tumors after removal of primary tumors.

(CA), Affinity BioReagents, Inc. (Golden, CO), and Pharmingen (CA), respectively.

Angiostatin and endostatin expression vectors A 1.4-kb cDNA encoding full-length mouse angiostatin consisting of the signal peptide and first four kringles of mouse plasminogen, was subcloned into the pcDNA3 expression plasmid (Invitrogen).15 The cDNA encoding a secretion signal from the mouse immunoglobulin k chain was fused to the N-terminus of mouse endostatin by two rounds of PCR. Mouse endostatin cDNA encoding 184 amino acids of collagen XVIII was firstly amplified with the two primers 50 -GGG TAC TGC TGC TCT GGG TTC CAG GTT CCA CTG GTG ACC ATA CTC ATC AGG ACT TTC AG-30 and 50 -GGG GGA TCC CTA TTT GGA GAA AGA GGT CAT G-30 . The resulting PCR product was used as a template for further amplication with the two primers 50 -GAA GCT TAT GGA GAC AGA CAC ACT CCT GCT ATG GGT ACT GCT GCT CTG GG-30 and 50 -GGG GGA TCC CTA TTT GGA GAA AGA GGT CATG-30 . The final PCR product was subcloned into the pcDNA3 expression vector (Invitrogen). This expression vector was driven by a cytomegalovirus (CMV) immediate-early promoter, the most widely used promoter for transgene expression in skeletal muscle.28 DNA sequence analysis confirmed that the cDNA sequences were inserted in the proper reading frame, and no mutations had been incorporated.

Plasmid/PVP formulation PVP (plasdone C-30, Mr 50,000) was kindly supplied by Alchemy Chemicals Ltd. (Auckland, New Zealand). Purified plasmids were formulated at a concentration of 3 mg DNA/ml using a 5% PVP delivery system, as described previously.8,39 Briefly, formulations were made by adding stock solutions in the following order: water, plasmid, polymer, and 5 M NaCl. The plasmid and PVP are allowed to incubate at room temperature for 15 minutes prior to adding salt or lactose for tonicity adjustment. Likewise, sodium citrate buffers in 0.9% NaCl are added after incubating the plasmid and polymers for 15 minutes at room temperature.

Animal experimental protocols Materials and methods

Mice, cell lines and antibodies Male C57BL/6 mice (H-2b), 6–8 weeks old, were obtained from the Animal Research Center, Shandong University, China. The syngeneic (H-2b) EL-4 thymic lymphoma cell line was purchased from the American Type Culture Collection (Rockville, MD). It was cultured at 371C in DMEM medium (Gibco BRL, Grand Island, NY), supplemented with 10% fetal calf serum, 50 U/ml penicillin/streptomycin, 2 mM L-glutamine, and 1 mM pyruvate. The anti-plasminogen, anti-endostatin (PA1601), and anti-CD31 (MEC13.3) antibodies were purchased from Calbiochem-Novabiochem Corporation

Cancer Gene Therapy

All surgical procedures and care administered to the animals were in accordance with institutional guidelines. The primary tumors were established by s.c. injection of 1 106 EL-4 tumor cells in the right flank of C57BL mice. Tumor volumes were estimated by the formula V ¼ p/ 6 a2 b, where a was the short axis and b the long axis. When the primary tumors reached a size 4500 mm3 (B21 days after implantation), they were surgically excised, and the skin was sutured. Then, animals were randomly assigned to a treatment group, where each group contained 10 mice. Secondary local tumors were established by s.c. injection of 1 106 cells into the left flank of mice, whereas disseminated metastatic tumors were induced by


intravenous injection of 1 106 tumor cells into the tail vein. Mice were then intramuscularly injected with DNA/ PVP complexes containing either empty vector, angiostatin, endostatin, or angiostatin plus endostatin plasmids, once a week. A 80 ml solution of DNA/PVP containing 240 mg of plasmid DNA was injected into the tibialis (30 ml) and gastrocnemia (50 ml) muscles. In the combination therapy, angiostatin and endostatin plasmids were injected separately into different sites in the muscles to avoid the potential for one mediator to impact on either the overall level or expression kinetics of the other. All the mice are killed 3 weeks later. The tumor size and numbers of liver and lung surface metastases were counted.

Studies of tumor metastasis and mouse survival Primary s.c. EL-4 tumors do not spontaneously metastase to the lung and liver, hence in order to mimic the metastatic process 1 107 EL-4 cells were intravenously injected via the tail vein following removal of the primary tumor. The latter approach is advantageous as it allowed us to more accurately control the degree of tumor dissemination. While we could have used a spontaneously metastatic tumor type, most such tumors metastasize to the lung but not to the liver. Gene constructs were then injected intramuscularly, as above. The animals were weighed thrice weekly and assessed. Moribund mice were euthanized according to pre-established criteria, namely the presence of two or more of the following premorbid conditions: gross ascites, palpable tumor burden greater than 2 cm in diameter, dehydration, lethargy, emaciation, or weight loss greater than 20% of initial body weight.

Immunohistologic analysis Cryosections (10 mm thickness) prepared from the tibialis muscles following intramuscular gene delivery were incubated overnight with specific antibodies (Abs). They were subsequently incubated for 30 minutes with appropriate secondary antibodies (VECTASTAIN Universal Quick kit, Vector Laboratories, Burlingame, CA), and developed with Sigma FAST DAB (3,30 -diaminobenzidine tetrahydrochloride) and CoCl2 enhancer tablets (Sigma). Sections were counterstained with Mayer’s hematoxylin.

Western blotting The method for detecting the expression of proteins has been previously described.15,40 Briefly, blood samples were collected and the sera were resolved by SDS-PAGE, and electrophoretically transferred to nitrocellulose Hybond C extra membranes. The membranes were incubated with primary antibodies, and subsequently with horseradish peroxidase-conjugated secondary antibodies. They were developed by enhanced chemiluminescence (Amersham International, Buckingham, England), and exposure to X-ray film. Band density was quantified using the Scion Image software (Scion Corporation, Frederick, MD).

37

Assessment of vascularity The methodology to determine tumor vascularity has been described previously.15,40 Briefly, 10 mm frozen tumor sections prepared from subcutaneous nodular tumors 3 weeks after treatment were immunostained with the anti-CD31 antibody, as described above. Stained blood vessels were counted in five blindly chosen random fields (0.155 mm2) at 40 magnification, and the mean of the highest three counts was calculated. The concentric circles method was also used to assess vascularity.

In situ detection of apoptotic cells The method has been described previously.15,41 Briefly, serial sections of 6 mm thickness were prepared from tumors 3 weeks following treatment. TUNEL staining of sections was performed using an in situ apoptosis detection kit from Boehringer Mannheim, Germany, according to the manufacturer’s instruction, and examined by fluorescence microscopy. Adjacent sections were counterstained with hematoxylin and eosin. The total number of apoptotic cells in 10 randomly selected fields was counted. The apoptotic index was calculated as the percentage of positive staining cells, namely AI ¼ number of apoptotic cells 100/total number of nucleated cells.

Statistical analysis Kruskal–Wallis tests were performed to test the significance of the treatment effect in relation to tumor volume and numbers of metastases, whereas log rank tests were performed for survival data. For other data, results were expressed as mean values7standard deviation (SD), and a Student’s t-test was used for evaluating statistical significance. Po.05 was considered to be statistically significant.

Results

Removal of primary tumors promotes the outgrowth of distant nodules and metastases Tumors of different sizes were established in the right flank of mice, and either left to grow or surgically excised. Data are shown only for primary tumors left to grow to B500 mm3. Both groups of tumor-bearing and surgically treated mice were challenged by s.c. injection of 1 106 EL-4 tumor cells in the left flank. The presence of large tumors (4500 mm3 in volume) in the tumor-bearing mice significantly suppressed the growth of subcutaneous tumors in the left flank, compared to the growth of corresponding tumors in surgically treated mice (Fig 1a). Tumors were palpable 3–4 days earlier in mice whose primary tumor had been surgically removed than in mice bearing primary tumors, and on day 21 the average tumor size was 40% larger in surgically treated mice than in those bearing primary tumors (Fig 1a). The secondary tumors of mice bearing large tumors (4500 mm3 in volume) were less vascular than those bearing smaller (o500 mm3) primary tumors, indicating that the size of

Cancer Gene Therapy


38

a

800

Tumor size (mm3)

With primary tumors Primary tumors excised

600

400

200

0

0

5

10

15

20

Days after challenge

Numbers of lung metastases

b 150

100

50

0 With primary tumors

Primary tumors excised

Numbers of liver metastases

Intramuscular gene injection leads to the expression of exogenous angiostatin and endostatin at the site of injection, and in the circulation A DNA/PVP solution containing 240 mg of either angiostatin or endostatin plasmid was intramuscularly injected into C57BL mice. Other dosages of plasmid were tested, but 240 mg of plasmid proved to generate the highest levels of exogenous antiangiogenic factors in circulation, in accord with two previous studies.8,38 After 4 days, the mice were killed and the muscles at the injection sites were surgically excised and sectioned. Immunohistochemical analysis of muscle sections with

c 200 150 100 50 0 With primary tumors

Primary tumors excised

d 0

12

24

36

48

Angiostatin Endostatin

160 Band density

the primary tumor dictated the degree of tumor angiogenesis in the secondary tumors (data now shown). These results are in accord with similar observations made by Sckell et al42 and Guba et al.43 Once again, tumors of different sizes were established in the right flank of mice, and either left to grow or surgically excised. This time both groups of tumorbearing and surgically treated mice were challenged by i.v. injection of 1 106 tumor cells to investigate whether primary EL-4 tumors influenced the growth of tumors that disseminate to the lung and liver. The numbers of tumors on the surface of both the lung and liver were significantly increased 2.5-fold in the mice whose primary tumors had been surgically removed, compared with mice bearing primary tumors (Fig 1b, c). The differences in the growth of the tumor metastases in the two groups of mice correlated with changes in the levels of angiogensis inhibitors in sera detected by Western blotting. Blood samples were collected from mice 0, 12, 24, 36, 48 hours after removal of the primary tumors. The levels of angiostatin and endostatin gradually decreased after the primary tumors were excised, and finally disappeared altogether (Fig 1d).


120 80 40 0

0

Cancer Gene Therapy

12

24 Hours

36

48

Figure 1 Removal of primary tumors deprives the host of endogenous angiogenesis inhibitors, and promotes the growth of secondary tumors. Primary tumors were established by s.c. injection of 1 106 EL-4 tumor cells into the right flank of C57BL/6 mice. Tumors were either left alone, or surgically excised 3 weeks later, when the tumors reached over 500 mm3 in volume. (a) Both groups of mice (i.e. those with primary tumors and those whose tumors had been surgically removed) were rechallenged by s.c. injection of 1 106 tumor cells into the left flank of mice. The mice were monitored and secondary tumors were measured every third day for 3 weeks. (b, c) Both groups of mice (i.e. those with primary tumors and those whose tumors had been surgically removed) were rechallenged by intravenous injection of 1 106 tumor cells. After 3 weeks, the mice were killed, and the number of tumors on the surface of lungs and livers were microscopically counted. (d) Blood samples were collected at 0, 12, 24, 36, and 48 hours, as indicated, after removal of primary tumors from mice that were surgically treated. The level of angiostatin and endostatin in circulation was analyzed by Western blot analysis (upper figure), and the band density was quantified using the Scion Image software, and graphed (lower figure).


a

pcDNA3

Angiostatin

b

0

3

5

Endostatin

7

9


Band density

200

Angiostatin

160

Endostatin

120 80 40 0 0

3

5 Days

7

9

Figure 2 Intramuscular antiangiogenic gene delivery leads to expression of angiostatin and endostatin at the site of injection as well as in the circulation. (a) A DNA/PVP solution containing 240 mg of plasmid encoding either angiostatin or endostatin was injected into the tibialis and gastrocnemia muscles of C57BL/6 mice. The muscle was surgically excised 4 days later, sectioned, and subjected to immunohistochemical analysis with antiangiostatin and antiendostatin antibodies, as indicated. (b) Serum samples collected from mice at 0, 3, 5, 7, 9 days, as indicated, after intramuscular injection of angiostatin and endostatin plasmids, were Western blotted with antiangiostatin and antiendostatin antibodies, as indicated (upper figure). Band density was quantified using the Scion Image software (lower figure).

specific Abs against angiostatin and endostatin revealed high levels of expression of the exogenous proteins (Fig 2a). Blood samples were collected via the tail vein from mice 0, 3, 5, 7, and 9 days following intramuscular gene injection. As shown in Fig 2b, exogenous angiostatin and endostatin were detected in the sera 3 days after intramuscular gene injection, and their levels rose to a peak on day 7, and then decreased slightly in accord with one previous report.8 Hence, to counter the pharmacokinetic decline in factor expression, angiostatin and endostatin plasmids were injected intramuscularly once a week for the following experiments to achieve a stable level of these exogenous proteins in the circulation. The peak levels of exogenous angiostatin and endostatin in the circulation 7 days after intramuscular gene delivery, as determined by measuring band density in Western blots, were increased 1.3 and 1.8-fold, respectively, compared with the levels in the circulation of mice bearing primary tumors (data not shown or see Figs 1d and 2b).

Intramuscular gene delivery of angiostatin and endostatin results in suppression of secondary tumor nodules An experiment was undertaken to analyze whether intramuscular injection of antiangiogenic DNA/PVP complexes would have therapeutic potential in inhibiting the growth of secondary nodular tumors. In order to mimic the clinical situation, primary tumors were established, and excised prior to dissemination of EL-4 tumors, and treatment. This step of establishing and subsequently removing primary tumors was necessary as we could not discount the possibility that primary tumors secreted long-lasting factors into the circulation capable of influencing subsequent antiangiogenic therapy. Tumors were established in the right flanks of 40 mice, the tumors excised, and each mouse immediately challenged by s.c. injection of EL-4 tumor cells, followed by intramuscular injection of DNA/PVP solutions containing either empty vector (n ¼ 10), angiostatin (n ¼ 10), endostatin (n ¼ 10), or angiostatin plus endostatin (n ¼ 10). Tumors grew rapidly in the control empty vector-treated group, reaching around 580 mm3 in volume in 3 weeks. In contrast, intramuscular gene therapy with either angiostatin or endostatin significantly suppressed the growth of secondary subcutaneous tumors, such that the mean volume of tumors was 350 and 370 mm3 following treatment with angiostatin and endostatin, respectively (Fig 3a). Furthermore, the tumors of mice treated with angiostatin or endostatin were only palpable 3–4 days after tumors were detected in mice treated with empty vectors. Combinational therapy was even more effective such that tumors of mice treated simultaneously with angiostatin and endostatin plasmids were only detectable 3 days after detection of tumors in angiostatin or endostatin monotherapy treated mice, and 6–7 days later than in mice treated with empty vectors. Tumors of mice treated with a combination of angiostatin and endostatin were on average 60% smaller at day 21 than those of mice treated with empty vector, and 40% smaller than those of mice treated by angiostatin or endostatin monotherapy (Fig 3a).

Intramuscular gene delivery of angiostatin and endostatin prevents the outgrowth of disseminated tumors An experiment was undertaken to determine whether intramuscular injection of antiangiogenic DNA/PVP complexes would have therapeutic potential in preventing the outgrowth of disseminated tumors. Tumors were established in the right flanks of 40 mice, the tumors excised, and each mouse challenged by i.v. injection of EL-4 cells, followed by intramuscular injection of DNA/ PVP solutions containing either empty vector (n ¼ 10), angiostatin (n ¼ 10), endostatin (n ¼ 10), or angiostatin plus endostatin (n ¼ 10). All the mice were killed 3 weeks later and the numbers of tumors on the surface of lungs and livers were counted and compared respectively (Fig 3b and c). The mean numbers of tumors on the lung surface were 108, 75, 69, and 36 after treatment with

Cancer Gene Therapy

39


empty vector, angiostatin, endostatin, or angiostatin plus endostatin, respectively. Angiostatin and endostatin monotherapies led to statistically significant (Po.01) reductions (31 and 36%, respectively) in tumor numbers on the lung surface. Furthermore, combinational therapy with both genes resulted in an even greater reduction (67%) in tumor numbers on the lung surface, compared with that in control groups (Po.001). Similarly,

800

Control

Tumor size (mm3)

Angiostatin

600

Endostatin Angiostatin + Endostatin

400 200 0 0

0

15

20

Days after challenge

b Number of lung metastases

10

150

100

Intramuscular gene delivery of angiostatin and endostatin leads to inhibition of tumor vascularization

0 Angiostatin + Endostatin

c Number of liver metastases

Intramuscular gene therapy improves the survival rate of mice bearing disseminated metastases

50

Control Angiostatin Endostatin

200 150

Secondary tumors that were excised 3 weeks after establishment and gene therapy were sectioned, and stained with an anti-CD31 antibody. Representative sections from mice treated with empty vector (control), angiostatin, endostatin, and angiostatin plus endostatin, are shown in Fig 4a. Angiostatin and endostatin

100 50 0 Control Angiostatin Endostatin

d

angiostatin and endostatin monotherapies and the combination therapy resulted in 33, 24, and 51% reductions, respectively, in the mean tumor numbers (102, 115, and 75, respectively) on the liver surface, compared with the mean tumor number (151) in the control group.

We further investigated whether intramuscular gene therapy could result in a survival benefit for mice. A total of 40 mice were rechallenged by i.v. injection of 1 107 EL-4 tumor cells after surgical removal of their primary tumors, followed by intramuscular injection of DNA/PVP solutions containing either empty vector (n ¼ 10), angiostatin (n ¼ 10), endostatin (n ¼ 10), or angiostatin þ endostatin plasmids (n ¼ 10). Intramuscular delivery of either the angiostatin or endostatin gene resulted in significant improvement in the survival of mice (Po.05), with a median survival time of 50 or 57 days, respectively, compared to a 33-day median survival time for the control group (Fig 3d). Combinational therapy with both genes gave a profound increase in the median survival time of 79 days, compared with control groups (Po.01), or angiostatin and endostatin monotherapies (Po.05). Three of 10 mice in this group survived for more than 100 days after tumor cell inoculation, whereas all the mice treated with empty vector, angiostatin or endostatin became moribund and were killed (Fig 3d).

a

Angiostatin + Endostatin

120 100

Percent survival

40

Endostatin

Control

80

Angiostatin + Endostatin

60 Angiostatin

40 20 0 0

20

40 Days

Cancer Gene Therapy

60

80

100

Figure 3 Angiostatin and endostatin plasmids injected intramuscularly synergize in suppressing the growth of secondary tumors after removal of primary tumors, and increase the survival rate of mice. Primary tumors were established by s.c. injection of 1 106 EL-4 tumor cells into the right flank of C57BL/6 mice. The tumors were surgically excised 3 weeks later. The mice were challenged by either s.c. injection of 1 106 tumor cells into the left flank of mice (a), or i.v. injection of 1 106 (b, c), or 1 107 (d) tumor cells. A DNA/PVP solution containing 240 mg of plasmids encoding either angiostatin, endostatin, angiostatin plus endostatin, or empty vector (Control) was injected into the tibialis and gastrocnemia muscles of mice in each group. The mice were monitored and secondary tumors in the left flank of mice were measured every three days (a). Alternatively, mice were killed 3 weeks later and the number of tumors on the surface of lungs and livers counted (b, c), where each point represents a single animal and mean tumor numbers are indicated by a transverse line. (d) Alternatively, mice were observed twice weekly, and killed when they became moribund by pre-established criteria. A plot of survival curves is illustrated.


41 Angiostatin + Endostatin

a Angiostatin

c

d **

40

2

+

gi

os

ta

tro An

on

+

C

in at st do

En

An

ta os gi An

en gio do sta st tin at in

tin

l tro on

+

C

en gio do sta st tin at in

in at do

st

ta En

os gi An

An

tin

l tro on C

at in g en io do sta st tin at in

0

0

l

0

10

*

*

4

st

10

20

**

6

An

**

*

*

30

tin

*

Apoptosis Index

*

20

Median distance to CD31-labelled venules

Vessel counts per surface area

30

do

b

Endostatin

En

Control

Figure 4 Intramuscular antiangiogenic therapy inhibits tumor vascularization and increases tumor apoptosis. Secondary tumors (Fig 3a) that were excised 3 weeks after establishment and gene therapy were sectioned, and stained with an anti-CD31 mAb to identify vascular endothelial cells (upper row), or examined for apoptosis using TUNEL (lower row). Representative photographs were from mice treated with empty vector (Control), angiostatin, endostatin, or angiostatin plus endostatin plasmids, as indicated (a). Blood vessels stained with the anti-CD31 mAb were counted in blindly chosen random fields to record mean vessel density (b), or the median distance to the nearest CD31 mAb-labeled blood vessel from an array point was recorded using the concentric circles methods (c). Adjacent sections were stained with hematoxylin/eosin to calculate the apoptotic index (AI) (d). *Significant difference (Po.01) in results obtained with angiostatin and endostatin plasmid monotherapies versus empty vector. **Significant difference (Po.05) in results obtained with combination therapy versus angiostatin or endostatin monotherapy.

monotherapies resulted in a statistically significant (Po.01) 40% reduction in tumor vessel density compared with empty vector (Fig 4b). The median distance to the nearest vessels labeled with the anti-CD31 monoclonal antibody (mAb) from an array of points within tumors treated with angiostatin or endostatin was significantly (Po.01) longer than for tumors treated with empty vector (Fig 4c). The combination therapy resulted in even greater inhibition of tumor angiogenesis as evidenced by decreased vessel density and increased median distance to the nearest labeled vessels, compared with the control tumors, and angiostatin and endostatin monotherapies (Fig 4b and c).

Intramuscular gene therapy leads to enhancement of tumor apoptosis We next examined whether tumors deprived of an adequate vascular network underwent programmed cell death as measured by in situ labeling of fragmented DNA using the TUNEL method. Small numbers of apoptotic cells were detected in tumors treated with empty vector, whereas tumor cell apoptosis was doubled following angiostatin or endostatin treatment, and even tripled in the combinational therapy (Fig 4a). The Apoptosis Index for tumors from either angiostatin- or endostatin-treated mice was significantly (Po.01) higher than that for

Cancer Gene Therapy


42

tumors from the control groups (Fig 4d). The combination therapy resulted in an even greater increase in tumor cell apoptosis, compared with either angiostatin (Po.05) or endostatin (Po.05) monotherapy.

Discussion

In this study, we simulated the process of tumor metastasis using an EL-4 tumor cell system so that the process could be studied in a controlled and quantitative manner, which is not easily achievable using tumors that spontaneously metastasize to different extents. We demonstrate that surgical removal of primary EL-4 tumors leads to a marked decrease of endogenous angiostatin and endostatin, and promotes the outgrowth of distant nodular tumors, and of tumours that have disseminated to the lungs and livers. The presence of the primary tumor dictated the degree of vascularity of the secondary tumor, suggesting that antiangiogenic factors secreted by the primary tumor decrease angiogenesis in distant tumors, and thereby inhibit their growth, as previously demonstrated by Sckell et al42 and Guba et al.43 An alternative possibility for this phenomenon is ‘‘concomitant immunity’’, whereby the primary tumor stimulates an antitumor immune response that suppresses the growth of the secondary tumor. This possibility is not probable as we have previously demonstrated that EL-4 tumors become highly immunosuppressive as they grow.44 In particular, they secrete TGF-b, which downregulates antitumor immunity by inducing IL-10-mediated development of Th2 responses, and inhibition of Th1 responses.45 Thus, the tumorigenicity of EL-4 lymphomas is suppressed by soluble type II TGF-b therapy.46 In our hands, EL-4 cells do not stimulate antitumour immunity sufficient to generate a memory response. EL-4 cells have to be genetically engineered to express T-cell costimulators in order to be immunostimulatory.47 Even then, tumours become so potently immunosuppressive when they reach 0.3 cm in diameter that they can no longer be contained by the latter approach. We have also demonstrated that challenge of mice with large numbers of EL-4 cells suppresses the generation of anti-tumor CTL activity, and accelerates tumour growth.41 Thus, to the contrary, the primary tumor would be expected to assist the growth of the secondary tumor by inhibiting an antitumor immune response. Intramuscular gene delivery of angiostatin and endostatin resulted in elevated levels of expression of both antiangiogenic factors in the circulation, and suppressed the growth of secondary and disseminated tumors by inhibiting tumor angiogenesis and inducing cell apoptosis. Accordingly, the survival rate of mice was increased. Intramuscular gene delivery of a combination of angiostatin and endostatin had the greatest therapeutic benefit, increasing the median survival time from 33 to 79 days, compared to mice treated with empty vector. Our results agree with those of Ponnazhagan et al29 who showed that intramuscular delivery of a bicistronic AAV vector

Cancer Gene Therapy

encoding angiostatin and endostatin prior to s.c. challenge with xenografts of an angiogenesis-dependent human ovarian cancer cell line, SKOV3.ip1. completely inhibited tumorigenesis in an athymic (nude) mouse model. In contrast, intramuscular injection of AAV encoding angiostatin or endostatin alone was not nearly as effective. The apparent increased efficacy of the AAV vector approach (completely inhibited tumorigenesis), compared to naked plasmids used in our study (inhibited, but did not completely block tumorigenesis), may relate to the timing of administration of the antiangiogenic factors. Thus, Ponnazhagan et al29 delivered their AAV vector three weeks prior to inoculation of tumour cells allowing time for angiostatin and endostatin to be expressed, whereas in our case the angiogenic factors were delivered at the time of tumour cell injection. In hindsight, it would be preferable to deliver the angiogenic factors prior to surgical removal of a tumor in order to ensure that serum levels of angiostatin and endostatin have reached a maximum level. Our approach using naked plasmids could potentially be enhanced by electrotransfer of angiostatin and endostatin plasmids into muscle tissues as exemplified by Cichon et al.9 Antiangiogenic therapy is widely regarded as a promising treatment strategy to combat cancer based on the fact that most tumors are dependent on angiogenesis for survival. However, despite many successful animal experiments employing antiangiogenic therapies, disappointing reports have emerged recently, particular when antiangiogenic therapies have been applied to human beings.48,49 The lesson from one study50 is that low dosage or short-term administration of angiogenesis inhibitors may not be beneficial for the treatment of metastases. Although the majority of preclinical and clinical antiangiogenic therapies to date have been conducted with purified antiangiogenic factors, gene therapy may be a more powerful strategy, as the host tissues are engineered to constitutively ‘‘manufacture’’ the antiangiogenic proteins over the long term, which avoids constant administration of these otherwise very expensive agents. Constitutive delivery of antiangiogenic factors in mice has been shown to be more effective than the intermittent peaks of expression obtained by injection of antiangiogenic proteins.8 Although intratumoral, and intravenous vector delivery have been widely investigated, skeletal muscle has some advantages as a target for gene therapy. Firstly, skeletal muscle constitutes about 30% of the normal adult body mass and is easily accessible for nearly all gene delivery approaches; secondly, skeletal muscle has an abundant blood vascular supply, thus providing an efficient transport system for the carriage of secreted protein into the circulation; thirdly, skeletal muscle fibers are terminally differentiated cells, which would provide a stable environment as a ‘‘factory’’ for continuous production of transgenes. Metastatic micrometastases that are maintained in a dormant state due to the secretion of antiangiogenic factors secreted by primary tumors, pose a constant threat. Under conditions of angiogenic suppression, the proliferation and apoptosis of micrometastases is in


balance resulting in viable, but nongrowing, tumors. The mechanism of induction of tumor cell apoptosis by angiostatin and endostatin is unclear. Some studies have demonstrated that angiostatin-mediated inhibition of angiogenesis results in increased tumor cell apoptosis with no direct effect on the rate of tumor cell proliferation.12 Endostatin may act through beta-catenin, an intracellular protein participating in cell adhesion and transcriptional regulation.51 The inhibition of neovascularization by angiostatin or endostatin may restrict the supply of tumor cell survival factors provided either by endothelial cells or by the circulation. Both angiostatin and endostatin have been shown to induce endothelial cell apoptosis.13,52,53 When endothelial-cell apoptosis occurs in a microvessel, tumor cells supported by that vessel subsequently undergo apoptosis. Several studies indicate that angiogenesis inhibitors can induce tumor-cell apoptosis by decreasing the levels of endothelial-cell-derived paracrine factors that promote cell survival. At least 20 such proteins have been reported to be produced by endothelial cells, including PDGF, IL-6 and heparinbinding epithelial growth factor.54 Production of paracrine factors is decreased, in part, because angiogenesis inhibitors can inhibit endothelial-cell proliferation.55 It is unclear whether angiogenesis inhibitors directly decrease endothelial cell production of paracrine factors. The method presented in this study offers the promise of the translation of antiangiogenic therapy into clinical application. Removal of primary tumors by surgery,56 or irradiation57 often results in the vascularization and rapid growth of disseminated micrometastases. The phenomenon can now be explained by the fact that certain tumors produce enzymes that activate angiogenesis inhibitors such as angiostatin,3 endostatin,4,58,59 and antiangiogenic antithrombin,60,61 which in turn prevent the growth of tumors remote from the primary tumor. There are more than 70 antiangiogenic agents currently in clinical trial that follow from the pioneering work performed with angiostatin and endostatin. Chemotherapy is the most common method of preventing and treating disseminated micrometastases. However, its toxic and immune-suppressing features devalue its clinical utility. Antiangiogenic therapy is generally less toxic and less likely to induce acquired drug resistance. We therefore speculate that intramuscular delivery of antiangiogenic expression plasmids could be used in prophylactic cancer prevention therapy for patients who have a high risk of cancer, for example, HBV-infected patients with liver cirrhosis, or alternatively as a therapy to combat the recurrence of cancer after complete surgical resection of primary tumors. An experimental study of spontaneous carcinogen-induced breast cancer in rats revealed that endostatin prevented the onset of breast cancer and also prolonged survival, compared with untreated controls.62 In order to achieve the preventive results, antiangiogenic reagents have to be delivered for prolonged periods at a stable therapeutic level. The method presented here can meet this requirement and may represent a promising strategy to improve the outcome of therapeutic surgery by

replacing the endogenous angiogenesis inhibitors that are lost after removal of primary tumors.

Acknowledgments

This work is supported in part by grants from the National Natural Scientific Foundation of China (30100178), the Scientific and Technological Committee of Shandong Province (1997BB1CJB1), a Young and Middle-aged Excellent Scientists Prize Fund from Shandong Province, China (9836), the Wellcome Trust (UK), the Maurice and Phyllis Paykel Trust, and the Health Research Council of New Zealand. X Sun is a recipient of a Wellcome Trust Research Leave Fellowship.

References 1. Weinstat-Saslow D, Steeg PS. Angiogenesis and colonization in the tumor metastatic process: basic and applied advances. FASEB J. 1994;8:401–407. 2. Fidler IJ, Ellis LM. The implications of angiogenesis for the biology and therapy of cancer metastasis. Cell. 1994;79: 185–188. 3. O’Reilly MS, Holmgren L, Shing Y, et al. Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell. 1994;79: 315–328. 4. O’Reilly MS, Boehm T, Shing Y, et al. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell. 1997;88:277–285. 5. Holmgren L, O’Reilly MS, Folkman J. Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenesis suppression. Nat Med. 1995;1:149–153. 6. O’Reilly MS, Holmgren L, Chen C, et al. Angiostatin induces and sustains dormancy of human primary tumors in mice. Nat Med. 1996;2:689–692. 7. Dhanabal M, Ramchandran R, Volk R, et al. Endostatin: yeast production, mutants, and antitumor effect in renal cell carcinoma. Cancer Res. 1999;59:189–197. 8. Blezinger P, Wang J, Gondo M, et al. Systemic inhibition of tumor growth and tumor metastases by intramuscular administration of the endostatin gene. Nat Biotechnol. 1999;17:343–348. 9. Cichon T, Jamrozy L, Glogowska J, et al. Electrotransfer of gene encoding endostatin into normal and neoplastic mouse tissues: Inhibition of primary tumor growth and metastatic spread. Cancer Gene Ther. 2002;9:771–777. 10. Moser TL, Stack MS, Asplin I, et al. Angiostatin binds ATP synthase on the surface of human endothelial cells. Proc Natl Acad Sci USA. 1999;96:2811–2816. 11. Kirsch M, Strasser J, Allender R, et al. Angiostatin suppresses malignant glioma growth in vivo. Cancer Res. 1998;58:4654–4659. 12. Joe YA, Hong YK, Chung DS, et al. Inhibition of human malignant glioma growth in vivo by human recombinant plasminogen kringles 1-3. Int J Cancer. 1999;82:694–699. 13. Clasesson-Welsh L, Welsh M, Ito N, et al. Angiostatin induces endothelial cell apoptosis and activation of focal adhesion kinase independently of the integrin-binding motif RGD. Proc Natl Acad Sci USA. 1998;95:5579–5583.

Cancer Gene Therapy

43


44

14. Redlitz A, Daum G, Sage EH. Angiostatin diminishes activation of the mitogen-activated protein kinase ERK-1 and ERK-2 in human dermal microvascular endothelial cells. J Vasc Res. 1999;36:28–34. 15. Sun X, Kanwar JR, Leung E, et al. Angiostatin enhances B7.1-mediated cancer immunotherapy independently of effects on vascular endothelial growth factor expression. Cancer Gene Ther. 2001;8:719–727. 16. Lee SJ, Jang JW, Kim YM, et al. Endostatin binds to the catalytic domain of matrix metalloproteinase-2. FEBS Lett. 2002;519:147–152. 17. Kim YM, Hwang S, Kim YM, et al. Endostatin blocks vascular endothelial growth factor-mediated signaling via direct interaction with KDR/Flk-1. J Biol Chem. 2002;277:27872–27879. 18. Hanai J, Gloy J, Karumanchi SA, et al. Endostatin is a potential inhibitor of Wnt signaling. J Cell Biol. 2002;158:529–539. 19. Furumatsu T, Yamaguchi N, Nishida K, et al. Endostatin inhibits adhesion of endothelial cells to collagen I via alpha(2)beta(1) integrin, a possible cause of prevention of chondrosarcoma growth. J Biochem. 2002;131:619–626. 20. Dixelius J, Cross M, Matsumoto T, et al. Endostatin regulates endothelial cell adhesion and cytoskeletal organization. Cancer Res. 2002;62:1944–1947. 21. Scappaticci FA, Smith R, Pathak A, et al. Combination angiostatin and endostatin gene transfer induces synergistic antiangiogenic activity in vitro and antitumor efficacy in leukemia and solid tumors in mice. Mol Ther. 2001;3: 186–196. 22. Scappaticci FA, Contreras A, Smith R, et al. Statin-AE: a novel angiostatin-endostatin fusion protein with enhanced antiangiogenic and antitumor activity. Angiogenesis. 2001;4:263–268. 23. Putney SD, Burke PA. Improving protein therapeutics with sustained-release formulations. Nat Biotechnol. 1998;16: 153–157. 24. Crystal RG. The body as a manufacturer of endostatin. Nat Biotechnol. 1999;17:336–337. 25. Boehm T, Folkman J, Browder T. Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance. Nature. 1997;390:404–407. 26. Folkman J. Antiangiogenic gene therapy. Proc Natl Acad Sci USA. 1998;95:9064–9066. 27. Kong HL, Crystal RG. Gene therapy strategies for tumor antiangiogenesis. J Natl Cancer Inst. 1998;90:273–286. 28. Wolff JA, Malone RW, Williams P, et al. Direct gene transfer into mouse skeletal muscle in vivo. Science. 1990;247:1465–1468. 29. Ponnazhagan S, Mahendra G, Kumar S, et al. Adenoassociated virus 2-mediated antiangiogenic cancer gene therapy: long-term efficacy of a vector encoding angiostatin and endostatin over vectors encoding a single factor. Cancer Res. 2004;64:1781–1787. 30. Lu QL, Bou-Gharios G, Partridge TA. Non-viral gene delivery in skeletal muscle: a protein factory. Gene Therapy. 2003;10:131–142. 31. Zabner J, Ramsey BW, Meeker DP. Repeat administration of an adenovirus vector encoding cystic fibrosis transmembrane conductance regulator to the nasal epithelium of patients with cystic fibrosis. J Clin Invest. 1996;97: 1504–1511. 32. Fisher KJ, Choi H, Burda J. Recombinant adenovirus deleted of all viral genes for gene therapy of cystic fibrosis. Virology. 1996;217:11–22.

Cancer Gene Therapy

33. Yei S, Mittereder N, Tang K. Adenovirus-mediated gene transfer for cystic fibrosis: quantitative evaluation of repeated in vivo vector administration to the lung. Gene Therapy. 1994;1:192–200. 34. McMahon JM, Wells KE, Bamfo JE, et al. Inflammatory responses following direct injection of plasmid DNA into skeletal muscle. Gene Therapy. 1998;5:1283–1290. 35. Caron NJ, Torrente Y, Camirand G, et al. Intracellular delivery of a Tat-eGFP fusion protein into muscle cells. Mol Ther. 2001;3:310–318. 36. Anwer K, Earle KA, Shi M, et al. Synergistic effect of formulated plasmid and needle-free injection for genetic vaccines. Pharm Res. 1999;16:889–895. 37. Fewell JG, MacLaughlin F, Mehta V, et al. Gene therapy for the treatment of hemophilia B using PINC-formulated plasmid delivered to muscle with electroporation. Mol Ther. 2001;3:574–583. 38. Oga M, Takenaga K, Sato Y, et al. Inhibition of metastatic brain tumor growth by intramuscular administration of the endostatin gene. Int J Oncol. 2003;23:73–79. 39. Mumper RJ, Wang J, Klakamp SL, et al. Protective interactive noncondensing (PINC) polymers for enhanced plasmid distribution and expression in rat skeletal muscle. J Control Release. 1998;52:191–203. 40. Sun X, Kanwar JR, Leung E, et al. Gene transfer of antisense hypoxia inducible factor-1a enhances the therapeutic efficacy of cancer immunotherapy. Gene Therapy. 2001;8:638–645. 41. Kanwar JR, Berg RW, Yang Y, et al. Requirements for ICAM-1 immunogene therapy of lymphoma. Cancer Gene Ther. 2003;10:468–476. 42. Sckell A, Safabakhsh N, Dellian M, et al. Primary tumor size-dependent inhibition of angiogenesis at a secondary site: an intravital microscopic study in mice. Cancer Res. 1998;58:5866–5869. 43. Guba M, Cernaianu G, Koehl G, et al. A primary tumor promotes dormancy of solitary tumor cells before inhibiting angiogenesis. Cancer Res. 2001;61:5575–5579. 44. Kanwar JR, Shen WP, Berg R, et al. Effect of survivin antagonists on the growth of established tumors and B7.1 immunogene therapy. J Natl Cancer Inst. 2001;93: 1541–1552. 45. Maeda H, Shiraishi A. TGF-beta contributes to the shift toward Th2-type responses through direct and IL-10mediated pathways in tumor-bearing mice. J Immunol. 1996;156:73–78. 46. Won J, Kim H, Park EJ, et al. Tumorigenicity of mouse thymoma is suppressed by soluble type II transforming growth factor b receptor therapy. Cancer Res. 1999;59: 1273–1277. 47. Kanwar JR, Berg R, Lehnert K, et al. Taking lessons from dendritic cells: multiple xenogeneic ligands for leukocyte integrins have the potential to stimulate anti-tumour immunity. Gene Therapy. 1999;6:1835–1844. 48. Pawliuk R, Bachelot T, Zurkiya O, et al. Continuous intravascular secretion of endostatin in mice from transduced hematopoietic stem cells. Mol Ther. 2002;5: 345–351. 49. Eisterer W, Jiang X, Bachelot T, et al. Unfulfilled promise of endostatin in a gene therapy–xenotransplant model of human acute lymphocytic leukemia. Mol Ther. 2002;5: 352–359. 50. Hori K, Li HC, Saito S, et al. Increased growth and incidence of lymph node metastases due to the angiogenesis inhibitor AGM-1470. Br J Cancer. 1997;75:1730–1734.


51. Dixelius J, Cross MJ, Matsumoto T. Endostatin action and intracellular signaling: beta-catenin as a potential target? Cancer Lett. 2003;196:1–12. 52. Lucas R, Holmgren L, Garcia I, et al. Multiple forms of angiostatin induce apoptosis in endothelial cells. Blood. 1998;92:4730–4741. 53. Wickstrom SA, Veikkola T, Rehn M. Endostatin-induced modulation of plasminogen activation with concomitant loss of focal adhesions and actin stress fibers in cultured human endothelial cells. Cancer Res. 2002;61:6511–6516. 54. Rak JW, St Croix BD, Kerbel RS. Consequences of angiogenesis for tumor progression, metastasis and cancer therapy. Anti-Cancer Drugs. 1995;6:3–18. 55. Dixelius J, Larsson H, Sasaki T, et al. Endostatin-induced tyrosine kinase signaling through the Shb adaptor protein regulates endothelial cell apoptosis. Blood. 2000;95: 3403–3411. 56. Oehler MK, Bicknell R. The promise of anti-angiogenic cancer therapy. Br J Cancer. 2000;82:749–752.

57. Camphausen K, Moses MA, Beecken WD, et al. Radiation therapy to a primary tumor accelerates metastatic growth in mice. Cancer Res. 2001;61:2207–2211. 58. Wen W, Moses MA, Wiederschain D, et al. The generation of endostatin is mediated by elastase. Cancer Res. 1999;59:6052–6056. 59. Felbor U, Dreier L, Bryant RA, et al. Secreted cathepsin L generates endostatin from collagen XVIII. EMBO J. 2000;19:1187–1194. 60. O’Reilly MS, Pirie-Shepherd S, Lane WS, Folkman J. Antiangiogenic activity of the cleaved conformation of the serpin antithrombin. Science. 1999;285:1926–1928. 61. Kisker O, Onizuka S, Banyard J, et al. Generation of multiple angiogenesis inhibitors by human pancreatic cancer. Cancer Res. 2001;61:7298–7304. 62. Feldman AL, Libutti SK. Liposomes complexed to plasmids encoding angiostatin and endostatin inhibit breast cancer in nude mice. Cancer Res. 1999;59: 3308–3312.

Cancer Gene Therapy

45